From 1c3a044c6013b7fcf4738129a1141c9c1994bb86 Mon Sep 17 00:00:00 2001 From: "sixiao@microsoft.com" Date: Fri, 14 Jul 2017 10:47:20 -0700 Subject: tools: hv: ignore a NIC if it has been configured Let bondvf.sh ignore this NIC if it has been configured, to prevent user configuration from being overwritten unexpectly. Signed-off-by: Simon Xiao Signed-off-by: David S. Miller --- tools/hv/bondvf.sh | 27 +++++++++++++++++++++++++-- 1 file changed, 25 insertions(+), 2 deletions(-) (limited to 'tools') diff --git a/tools/hv/bondvf.sh b/tools/hv/bondvf.sh index 89b2506..80f1028 100755 --- a/tools/hv/bondvf.sh +++ b/tools/hv/bondvf.sh @@ -211,6 +211,30 @@ function create_bond { echo $'\nBond name:' $bondname + if [ $distro == ubuntu ] + then + local mainfn=$cfgdir/interfaces + local s="^[ \t]*(auto|iface|mapping|allow-.*)[ \t]+${bondname}" + + grep -E "$s" $mainfn + if [ $? -eq 0 ] + then + echo "WARNING: ${bondname} has been configured already" + return + fi + elif [ $distro == redhat ] || [ $distro == suse ] + then + local fn=$cfgdir/ifcfg-$bondname + if [ -f $fn ] + then + echo "WARNING: ${bondname} has been configured already" + return + fi + else + echo "Unsupported Distro: ${distro}" + return + fi + echo configuring $primary create_eth_cfg_pri_$distro $primary $bondname @@ -219,8 +243,6 @@ function create_bond { echo creating: $bondname with primary slave: $primary create_bond_cfg_$distro $bondname $primary $secondary - - let bondcnt=bondcnt+1 } for (( i=0; i < $eth_cnt-1; i++ )) @@ -228,5 +250,6 @@ do if [ -n "${list_match[$i]}" ] then create_bond ${list_eth[$i]} ${list_match[$i]} + let bondcnt=bondcnt+1 fi done -- cgit v1.1 From 546ac1ffb70d25b56c1126940e5ec639c4dd7413 Mon Sep 17 00:00:00 2001 From: John Fastabend Date: Mon, 17 Jul 2017 09:28:56 -0700 Subject: bpf: add devmap, a map for storing net device references Device map (devmap) is a BPF map, primarily useful for networking applications, that uses a key to lookup a reference to a netdevice. The map provides a clean way for BPF programs to build virtual port to physical port maps. Additionally, it provides a scoping function for the redirect action itself allowing multiple optimizations. Future patches will leverage the map to provide batching at the XDP layer. Another optimization/feature, that is not yet implemented, would be to support multiple netdevices per key to support efficient multicast and broadcast support. Signed-off-by: John Fastabend Acked-by: Daniel Borkmann Acked-by: Jesper Dangaard Brouer Signed-off-by: David S. Miller --- tools/testing/selftests/bpf/test_maps.c | 15 +++++++++++++++ 1 file changed, 15 insertions(+) (limited to 'tools') diff --git a/tools/testing/selftests/bpf/test_maps.c b/tools/testing/selftests/bpf/test_maps.c index 79601c8..36d6ac3 100644 --- a/tools/testing/selftests/bpf/test_maps.c +++ b/tools/testing/selftests/bpf/test_maps.c @@ -438,6 +438,21 @@ static void test_arraymap_percpu_many_keys(void) close(fd); } +static void test_devmap(int task, void *data) +{ + int next_key, fd; + __u32 key, value; + + fd = bpf_create_map(BPF_MAP_TYPE_DEVMAP, sizeof(key), sizeof(value), + 2, 0); + if (fd < 0) { + printf("Failed to create arraymap '%s'!\n", strerror(errno)); + exit(1); + } + + close(fd); +} + #define MAP_SIZE (32 * 1024) static void test_map_large(void) -- cgit v1.1 From 9d6e005287ee23c7e25b04f4ad007bdbaf4fc438 Mon Sep 17 00:00:00 2001 From: John Fastabend Date: Mon, 17 Jul 2017 09:30:25 -0700 Subject: xdp: bpf redirect with map sample program Signed-off-by: John Fastabend Tested-by: Andy Gospodarek Acked-by: Daniel Borkmann Acked-by: Jesper Dangaard Brouer Signed-off-by: David S. Miller --- tools/testing/selftests/bpf/bpf_helpers.h | 2 ++ 1 file changed, 2 insertions(+) (limited to 'tools') diff --git a/tools/testing/selftests/bpf/bpf_helpers.h b/tools/testing/selftests/bpf/bpf_helpers.h index d50ac34..acbd605 100644 --- a/tools/testing/selftests/bpf/bpf_helpers.h +++ b/tools/testing/selftests/bpf/bpf_helpers.h @@ -38,6 +38,8 @@ static int (*bpf_clone_redirect)(void *ctx, int ifindex, int flags) = (void *) BPF_FUNC_clone_redirect; static int (*bpf_redirect)(int ifindex, int flags) = (void *) BPF_FUNC_redirect; +static int (*bpf_redirect_map)(void *map, int key, int flags) = + (void *) BPF_FUNC_redirect_map; static int (*bpf_perf_event_output)(void *ctx, void *map, unsigned long long flags, void *data, int size) = -- cgit v1.1 From 627fce14809ba5610b0cb476cd0186d3fcedecfc Mon Sep 17 00:00:00 2001 From: Josh Poimboeuf Date: Tue, 11 Jul 2017 10:33:42 -0500 Subject: objtool: Add ORC unwind table generation Now that objtool knows the states of all registers on the stack for each instruction, it's straightforward to generate debuginfo for an unwinder to use. Instead of generating DWARF, generate a new format called ORC, which is more suitable for an in-kernel unwinder. See Documentation/x86/orc-unwinder.txt for a more detailed description of this new debuginfo format and why it's preferable to DWARF. Signed-off-by: Josh Poimboeuf Cc: Andy Lutomirski Cc: Borislav Petkov Cc: Brian Gerst Cc: Denys Vlasenko Cc: H. Peter Anvin Cc: Jiri Slaby Cc: Linus Torvalds Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: live-patching@vger.kernel.org Link: http://lkml.kernel.org/r/c9b9f01ba6c5ed2bdc9bb0957b78167fdbf9632e.1499786555.git.jpoimboe@redhat.com Signed-off-by: Ingo Molnar --- tools/objtool/Build | 3 + tools/objtool/Documentation/stack-validation.txt | 56 ++---- tools/objtool/builtin-check.c | 2 +- tools/objtool/builtin-orc.c | 70 ++++++++ tools/objtool/builtin.h | 1 + tools/objtool/check.c | 58 +++++- tools/objtool/check.h | 15 +- tools/objtool/elf.c | 212 ++++++++++++++++++++-- tools/objtool/elf.h | 15 +- tools/objtool/objtool.c | 3 +- tools/objtool/orc.h | 30 ++++ tools/objtool/orc_dump.c | 212 ++++++++++++++++++++++ tools/objtool/orc_gen.c | 214 +++++++++++++++++++++++ tools/objtool/orc_types.h | 85 +++++++++ 14 files changed, 916 insertions(+), 60 deletions(-) create mode 100644 tools/objtool/builtin-orc.c create mode 100644 tools/objtool/orc.h create mode 100644 tools/objtool/orc_dump.c create mode 100644 tools/objtool/orc_gen.c create mode 100644 tools/objtool/orc_types.h (limited to 'tools') diff --git a/tools/objtool/Build b/tools/objtool/Build index 6f2e198..749becd 100644 --- a/tools/objtool/Build +++ b/tools/objtool/Build @@ -1,6 +1,9 @@ objtool-y += arch/$(SRCARCH)/ objtool-y += builtin-check.o +objtool-y += builtin-orc.o objtool-y += check.o +objtool-y += orc_gen.o +objtool-y += orc_dump.o objtool-y += elf.o objtool-y += special.o objtool-y += objtool.o diff --git a/tools/objtool/Documentation/stack-validation.txt b/tools/objtool/Documentation/stack-validation.txt index 17c1195..6a1af43 100644 --- a/tools/objtool/Documentation/stack-validation.txt +++ b/tools/objtool/Documentation/stack-validation.txt @@ -11,9 +11,6 @@ analyzes every .o file and ensures the validity of its stack metadata. It enforces a set of rules on asm code and C inline assembly code so that stack traces can be reliable. -Currently it only checks frame pointer usage, but there are plans to add -CFI validation for C files and CFI generation for asm files. - For each function, it recursively follows all possible code paths and validates the correct frame pointer state at each instruction. @@ -23,6 +20,10 @@ alternative execution paths to a given instruction (or set of instructions). Similarly, it knows how to follow switch statements, for which gcc sometimes uses jump tables. +(Objtool also has an 'orc generate' subcommand which generates debuginfo +for the ORC unwinder. See Documentation/x86/orc-unwinder.txt in the +kernel tree for more details.) + Why do we need stack metadata validation? ----------------------------------------- @@ -93,37 +94,14 @@ a) More reliable stack traces for frame pointer enabled kernels or at the very end of the function after the stack frame has been destroyed. This is an inherent limitation of frame pointers. -b) 100% reliable stack traces for DWARF enabled kernels - - (NOTE: This is not yet implemented) - - As an alternative to frame pointers, DWARF Call Frame Information - (CFI) metadata can be used to walk the stack. Unlike frame pointers, - CFI metadata is out of band. So it doesn't affect runtime - performance and it can be reliable even when interrupts or exceptions - are involved. - - For C code, gcc automatically generates DWARF CFI metadata. But for - asm code, generating CFI is a tedious manual approach which requires - manually placed .cfi assembler macros to be scattered throughout the - code. It's clumsy and very easy to get wrong, and it makes the real - code harder to read. - - Stacktool will improve this situation in several ways. For code - which already has CFI annotations, it will validate them. For code - which doesn't have CFI annotations, it will generate them. So an - architecture can opt to strip out all the manual .cfi annotations - from their asm code and have objtool generate them instead. +b) ORC (Oops Rewind Capability) unwind table generation - We might also add a runtime stack validation debug option where we - periodically walk the stack from schedule() and/or an NMI to ensure - that the stack metadata is sane and that we reach the bottom of the - stack. + An alternative to frame pointers and DWARF, ORC unwind data can be + used to walk the stack. Unlike frame pointers, ORC data is out of + band. So it doesn't affect runtime performance and it can be + reliable even when interrupts or exceptions are involved. - So the benefit of objtool here will be that external tooling should - always show perfect stack traces. And the same will be true for - kernel warning/oops traces if the architecture has a runtime DWARF - unwinder. + For more details, see Documentation/x86/orc-unwinder.txt. c) Higher live patching compatibility rate @@ -211,7 +189,7 @@ they mean, and suggestions for how to fix them. function, add proper frame pointer logic using the FRAME_BEGIN and FRAME_END macros. Otherwise, if it's not a callable function, remove its ELF function annotation by changing ENDPROC to END, and instead - use the manual CFI hint macros in asm/undwarf.h. + use the manual unwind hint macros in asm/unwind_hints.h. If it's a GCC-compiled .c file, the error may be because the function uses an inline asm() statement which has a "call" instruction. An @@ -231,8 +209,8 @@ they mean, and suggestions for how to fix them. If the error is for an asm file, and the instruction is inside (or reachable from) a callable function, the function should be annotated with the ENTRY/ENDPROC macros (ENDPROC is the important one). - Otherwise, the code should probably be annotated with the CFI hint - macros in asm/undwarf.h so objtool and the unwinder can know the + Otherwise, the code should probably be annotated with the unwind hint + macros in asm/unwind_hints.h so objtool and the unwinder can know the stack state associated with the code. If you're 100% sure the code won't affect stack traces, or if you're @@ -258,7 +236,7 @@ they mean, and suggestions for how to fix them. instructions aren't allowed in a callable function, and are most likely part of the kernel entry code. They should usually not have the callable function annotation (ENDPROC) and should always be - annotated with the CFI hint macros in asm/undwarf.h. + annotated with the unwind hint macros in asm/unwind_hints.h. 6. file.o: warning: objtool: func()+0x26: sibling call from callable instruction with modified stack frame @@ -272,7 +250,7 @@ they mean, and suggestions for how to fix them. If the instruction is not actually in a callable function (e.g. kernel entry code), change ENDPROC to END and annotate manually with - the CFI hint macros in asm/undwarf.h. + the unwind hint macros in asm/unwind_hints.h. 7. file: warning: objtool: func()+0x5c: stack state mismatch @@ -288,8 +266,8 @@ they mean, and suggestions for how to fix them. Another possibility is that the code has some asm or inline asm which does some unusual things to the stack or the frame pointer. In such - cases it's probably appropriate to use the CFI hint macros in - asm/undwarf.h. + cases it's probably appropriate to use the unwind hint macros in + asm/unwind_hints.h. 8. file.o: warning: objtool: funcA() falls through to next function funcB() diff --git a/tools/objtool/builtin-check.c b/tools/objtool/builtin-check.c index 365c34e..eedf089 100644 --- a/tools/objtool/builtin-check.c +++ b/tools/objtool/builtin-check.c @@ -52,5 +52,5 @@ int cmd_check(int argc, const char **argv) objname = argv[0]; - return check(objname, nofp); + return check(objname, nofp, false); } diff --git a/tools/objtool/builtin-orc.c b/tools/objtool/builtin-orc.c new file mode 100644 index 0000000..5ca41ab --- /dev/null +++ b/tools/objtool/builtin-orc.c @@ -0,0 +1,70 @@ +/* + * Copyright (C) 2017 Josh Poimboeuf + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version 2 + * of the License, or (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, see . + */ + +/* + * objtool orc: + * + * This command analyzes a .o file and adds .orc_unwind and .orc_unwind_ip + * sections to it, which is used by the in-kernel ORC unwinder. + * + * This command is a superset of "objtool check". + */ + +#include +#include +#include "builtin.h" +#include "check.h" + + +static const char *orc_usage[] = { + "objtool orc generate [] file.o", + "objtool orc dump file.o", + NULL, +}; + +extern const struct option check_options[]; +extern bool nofp; + +int cmd_orc(int argc, const char **argv) +{ + const char *objname; + + argc--; argv++; + if (!strncmp(argv[0], "gen", 3)) { + argc = parse_options(argc, argv, check_options, orc_usage, 0); + if (argc != 1) + usage_with_options(orc_usage, check_options); + + objname = argv[0]; + + return check(objname, nofp, true); + + } + + if (!strcmp(argv[0], "dump")) { + if (argc != 2) + usage_with_options(orc_usage, check_options); + + objname = argv[1]; + + return orc_dump(objname); + } + + usage_with_options(orc_usage, check_options); + + return 0; +} diff --git a/tools/objtool/builtin.h b/tools/objtool/builtin.h index 34d2ba7..dd52606 100644 --- a/tools/objtool/builtin.h +++ b/tools/objtool/builtin.h @@ -18,5 +18,6 @@ #define _BUILTIN_H extern int cmd_check(int argc, const char **argv); +extern int cmd_orc(int argc, const char **argv); #endif /* _BUILTIN_H */ diff --git a/tools/objtool/check.c b/tools/objtool/check.c index 2c6d748..cb57c52 100644 --- a/tools/objtool/check.c +++ b/tools/objtool/check.c @@ -36,8 +36,8 @@ const char *objname; static bool nofp; struct cfi_state initial_func_cfi; -static struct instruction *find_insn(struct objtool_file *file, - struct section *sec, unsigned long offset) +struct instruction *find_insn(struct objtool_file *file, + struct section *sec, unsigned long offset) { struct instruction *insn; @@ -259,6 +259,11 @@ static int decode_instructions(struct objtool_file *file) if (!(sec->sh.sh_flags & SHF_EXECINSTR)) continue; + if (strcmp(sec->name, ".altinstr_replacement") && + strcmp(sec->name, ".altinstr_aux") && + strncmp(sec->name, ".discard.", 9)) + sec->text = true; + for (offset = 0; offset < sec->len; offset += insn->len) { insn = malloc(sizeof(*insn)); if (!insn) { @@ -947,6 +952,30 @@ static bool has_valid_stack_frame(struct insn_state *state) return false; } +static int update_insn_state_regs(struct instruction *insn, struct insn_state *state) +{ + struct cfi_reg *cfa = &state->cfa; + struct stack_op *op = &insn->stack_op; + + if (cfa->base != CFI_SP) + return 0; + + /* push */ + if (op->dest.type == OP_DEST_PUSH) + cfa->offset += 8; + + /* pop */ + if (op->src.type == OP_SRC_POP) + cfa->offset -= 8; + + /* add immediate to sp */ + if (op->dest.type == OP_DEST_REG && op->src.type == OP_SRC_ADD && + op->dest.reg == CFI_SP && op->src.reg == CFI_SP) + cfa->offset -= op->src.offset; + + return 0; +} + static void save_reg(struct insn_state *state, unsigned char reg, int base, int offset) { @@ -1032,6 +1061,9 @@ static int update_insn_state(struct instruction *insn, struct insn_state *state) return 0; } + if (state->type == ORC_TYPE_REGS || state->type == ORC_TYPE_REGS_IRET) + return update_insn_state_regs(insn, state); + switch (op->dest.type) { case OP_DEST_REG: @@ -1323,6 +1355,10 @@ static bool insn_state_match(struct instruction *insn, struct insn_state *state) break; } + } else if (state1->type != state2->type) { + WARN_FUNC("stack state mismatch: type1=%d type2=%d", + insn->sec, insn->offset, state1->type, state2->type); + } else if (state1->drap != state2->drap || (state1->drap && state1->drap_reg != state2->drap_reg)) { WARN_FUNC("stack state mismatch: drap1=%d(%d) drap2=%d(%d)", @@ -1613,7 +1649,7 @@ static void cleanup(struct objtool_file *file) elf_close(file->elf); } -int check(const char *_objname, bool _nofp) +int check(const char *_objname, bool _nofp, bool orc) { struct objtool_file file; int ret, warnings = 0; @@ -1621,7 +1657,7 @@ int check(const char *_objname, bool _nofp) objname = _objname; nofp = _nofp; - file.elf = elf_open(objname); + file.elf = elf_open(objname, orc ? O_RDWR : O_RDONLY); if (!file.elf) return 1; @@ -1654,6 +1690,20 @@ int check(const char *_objname, bool _nofp) warnings += ret; } + if (orc) { + ret = create_orc(&file); + if (ret < 0) + goto out; + + ret = create_orc_sections(&file); + if (ret < 0) + goto out; + + ret = elf_write(file.elf); + if (ret < 0) + goto out; + } + out: cleanup(&file); diff --git a/tools/objtool/check.h b/tools/objtool/check.h index da85f5b..046874b 100644 --- a/tools/objtool/check.h +++ b/tools/objtool/check.h @@ -22,12 +22,14 @@ #include "elf.h" #include "cfi.h" #include "arch.h" +#include "orc.h" #include struct insn_state { struct cfi_reg cfa; struct cfi_reg regs[CFI_NUM_REGS]; int stack_size; + unsigned char type; bool bp_scratch; bool drap; int drap_reg; @@ -48,6 +50,7 @@ struct instruction { struct symbol *func; struct stack_op stack_op; struct insn_state state; + struct orc_entry orc; }; struct objtool_file { @@ -58,9 +61,19 @@ struct objtool_file { bool ignore_unreachables, c_file; }; -int check(const char *objname, bool nofp); +int check(const char *objname, bool nofp, bool orc); + +struct instruction *find_insn(struct objtool_file *file, + struct section *sec, unsigned long offset); #define for_each_insn(file, insn) \ list_for_each_entry(insn, &file->insn_list, list) +#define sec_for_each_insn(file, sec, insn) \ + for (insn = find_insn(file, sec, 0); \ + insn && &insn->list != &file->insn_list && \ + insn->sec == sec; \ + insn = list_next_entry(insn, list)) + + #endif /* _CHECK_H */ diff --git a/tools/objtool/elf.c b/tools/objtool/elf.c index 1a7e8aa..6e9f980 100644 --- a/tools/objtool/elf.c +++ b/tools/objtool/elf.c @@ -30,16 +30,6 @@ #include "elf.h" #include "warn.h" -/* - * Fallback for systems without this "read, mmaping if possible" cmd. - */ -#ifndef ELF_C_READ_MMAP -#define ELF_C_READ_MMAP ELF_C_READ -#endif - -#define WARN_ELF(format, ...) \ - WARN(format ": %s", ##__VA_ARGS__, elf_errmsg(-1)) - struct section *find_section_by_name(struct elf *elf, const char *name) { struct section *sec; @@ -349,9 +339,10 @@ static int read_relas(struct elf *elf) return 0; } -struct elf *elf_open(const char *name) +struct elf *elf_open(const char *name, int flags) { struct elf *elf; + Elf_Cmd cmd; elf_version(EV_CURRENT); @@ -364,13 +355,20 @@ struct elf *elf_open(const char *name) INIT_LIST_HEAD(&elf->sections); - elf->fd = open(name, O_RDONLY); + elf->fd = open(name, flags); if (elf->fd == -1) { perror("open"); goto err; } - elf->elf = elf_begin(elf->fd, ELF_C_READ_MMAP, NULL); + if ((flags & O_ACCMODE) == O_RDONLY) + cmd = ELF_C_READ_MMAP; + else if ((flags & O_ACCMODE) == O_RDWR) + cmd = ELF_C_RDWR; + else /* O_WRONLY */ + cmd = ELF_C_WRITE; + + elf->elf = elf_begin(elf->fd, cmd, NULL); if (!elf->elf) { WARN_ELF("elf_begin"); goto err; @@ -397,6 +395,194 @@ err: return NULL; } +struct section *elf_create_section(struct elf *elf, const char *name, + size_t entsize, int nr) +{ + struct section *sec, *shstrtab; + size_t size = entsize * nr; + struct Elf_Scn *s; + Elf_Data *data; + + sec = malloc(sizeof(*sec)); + if (!sec) { + perror("malloc"); + return NULL; + } + memset(sec, 0, sizeof(*sec)); + + INIT_LIST_HEAD(&sec->symbol_list); + INIT_LIST_HEAD(&sec->rela_list); + hash_init(sec->rela_hash); + hash_init(sec->symbol_hash); + + list_add_tail(&sec->list, &elf->sections); + + s = elf_newscn(elf->elf); + if (!s) { + WARN_ELF("elf_newscn"); + return NULL; + } + + sec->name = strdup(name); + if (!sec->name) { + perror("strdup"); + return NULL; + } + + sec->idx = elf_ndxscn(s); + sec->len = size; + sec->changed = true; + + sec->data = elf_newdata(s); + if (!sec->data) { + WARN_ELF("elf_newdata"); + return NULL; + } + + sec->data->d_size = size; + sec->data->d_align = 1; + + if (size) { + sec->data->d_buf = malloc(size); + if (!sec->data->d_buf) { + perror("malloc"); + return NULL; + } + memset(sec->data->d_buf, 0, size); + } + + if (!gelf_getshdr(s, &sec->sh)) { + WARN_ELF("gelf_getshdr"); + return NULL; + } + + sec->sh.sh_size = size; + sec->sh.sh_entsize = entsize; + sec->sh.sh_type = SHT_PROGBITS; + sec->sh.sh_addralign = 1; + sec->sh.sh_flags = SHF_ALLOC; + + + /* Add section name to .shstrtab */ + shstrtab = find_section_by_name(elf, ".shstrtab"); + if (!shstrtab) { + WARN("can't find .shstrtab section"); + return NULL; + } + + s = elf_getscn(elf->elf, shstrtab->idx); + if (!s) { + WARN_ELF("elf_getscn"); + return NULL; + } + + data = elf_newdata(s); + if (!data) { + WARN_ELF("elf_newdata"); + return NULL; + } + + data->d_buf = sec->name; + data->d_size = strlen(name) + 1; + data->d_align = 1; + + sec->sh.sh_name = shstrtab->len; + + shstrtab->len += strlen(name) + 1; + shstrtab->changed = true; + + return sec; +} + +struct section *elf_create_rela_section(struct elf *elf, struct section *base) +{ + char *relaname; + struct section *sec; + + relaname = malloc(strlen(base->name) + strlen(".rela") + 1); + if (!relaname) { + perror("malloc"); + return NULL; + } + strcpy(relaname, ".rela"); + strcat(relaname, base->name); + + sec = elf_create_section(elf, relaname, sizeof(GElf_Rela), 0); + if (!sec) + return NULL; + + base->rela = sec; + sec->base = base; + + sec->sh.sh_type = SHT_RELA; + sec->sh.sh_addralign = 8; + sec->sh.sh_link = find_section_by_name(elf, ".symtab")->idx; + sec->sh.sh_info = base->idx; + sec->sh.sh_flags = SHF_INFO_LINK; + + return sec; +} + +int elf_rebuild_rela_section(struct section *sec) +{ + struct rela *rela; + int nr, idx = 0, size; + GElf_Rela *relas; + + nr = 0; + list_for_each_entry(rela, &sec->rela_list, list) + nr++; + + size = nr * sizeof(*relas); + relas = malloc(size); + if (!relas) { + perror("malloc"); + return -1; + } + + sec->data->d_buf = relas; + sec->data->d_size = size; + + sec->sh.sh_size = size; + + idx = 0; + list_for_each_entry(rela, &sec->rela_list, list) { + relas[idx].r_offset = rela->offset; + relas[idx].r_addend = rela->addend; + relas[idx].r_info = GELF_R_INFO(rela->sym->idx, rela->type); + idx++; + } + + return 0; +} + +int elf_write(struct elf *elf) +{ + struct section *sec; + Elf_Scn *s; + + list_for_each_entry(sec, &elf->sections, list) { + if (sec->changed) { + s = elf_getscn(elf->elf, sec->idx); + if (!s) { + WARN_ELF("elf_getscn"); + return -1; + } + if (!gelf_update_shdr (s, &sec->sh)) { + WARN_ELF("gelf_update_shdr"); + return -1; + } + } + } + + if (elf_update(elf->elf, ELF_C_WRITE) < 0) { + WARN_ELF("elf_update"); + return -1; + } + + return 0; +} + void elf_close(struct elf *elf) { struct section *sec, *tmpsec; diff --git a/tools/objtool/elf.h b/tools/objtool/elf.h index 343968b..d86e2ff1 100644 --- a/tools/objtool/elf.h +++ b/tools/objtool/elf.h @@ -28,6 +28,13 @@ # define elf_getshdrstrndx elf_getshstrndx #endif +/* + * Fallback for systems without this "read, mmaping if possible" cmd. + */ +#ifndef ELF_C_READ_MMAP +#define ELF_C_READ_MMAP ELF_C_READ +#endif + struct section { struct list_head list; GElf_Shdr sh; @@ -41,6 +48,7 @@ struct section { char *name; int idx; unsigned int len; + bool changed, text; }; struct symbol { @@ -75,7 +83,7 @@ struct elf { }; -struct elf *elf_open(const char *name); +struct elf *elf_open(const char *name, int flags); struct section *find_section_by_name(struct elf *elf, const char *name); struct symbol *find_symbol_by_offset(struct section *sec, unsigned long offset); struct symbol *find_symbol_containing(struct section *sec, unsigned long offset); @@ -83,6 +91,11 @@ struct rela *find_rela_by_dest(struct section *sec, unsigned long offset); struct rela *find_rela_by_dest_range(struct section *sec, unsigned long offset, unsigned int len); struct symbol *find_containing_func(struct section *sec, unsigned long offset); +struct section *elf_create_section(struct elf *elf, const char *name, size_t + entsize, int nr); +struct section *elf_create_rela_section(struct elf *elf, struct section *base); +int elf_rebuild_rela_section(struct section *sec); +int elf_write(struct elf *elf); void elf_close(struct elf *elf); #define for_each_sec(file, sec) \ diff --git a/tools/objtool/objtool.c b/tools/objtool/objtool.c index ecc5b1b..31e0f91 100644 --- a/tools/objtool/objtool.c +++ b/tools/objtool/objtool.c @@ -42,10 +42,11 @@ struct cmd_struct { }; static const char objtool_usage_string[] = - "objtool [OPTIONS] COMMAND [ARGS]"; + "objtool COMMAND [ARGS]"; static struct cmd_struct objtool_cmds[] = { {"check", cmd_check, "Perform stack metadata validation on an object file" }, + {"orc", cmd_orc, "Generate in-place ORC unwind tables for an object file" }, }; bool help; diff --git a/tools/objtool/orc.h b/tools/objtool/orc.h new file mode 100644 index 0000000..a4139e3 --- /dev/null +++ b/tools/objtool/orc.h @@ -0,0 +1,30 @@ +/* + * Copyright (C) 2017 Josh Poimboeuf + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version 2 + * of the License, or (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, see . + */ + +#ifndef _ORC_H +#define _ORC_H + +#include "orc_types.h" + +struct objtool_file; + +int create_orc(struct objtool_file *file); +int create_orc_sections(struct objtool_file *file); + +int orc_dump(const char *objname); + +#endif /* _ORC_H */ diff --git a/tools/objtool/orc_dump.c b/tools/objtool/orc_dump.c new file mode 100644 index 0000000..36c5bf6 --- /dev/null +++ b/tools/objtool/orc_dump.c @@ -0,0 +1,212 @@ +/* + * Copyright (C) 2017 Josh Poimboeuf + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version 2 + * of the License, or (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, see . + */ + +#include +#include "orc.h" +#include "warn.h" + +static const char *reg_name(unsigned int reg) +{ + switch (reg) { + case ORC_REG_PREV_SP: + return "prevsp"; + case ORC_REG_DX: + return "dx"; + case ORC_REG_DI: + return "di"; + case ORC_REG_BP: + return "bp"; + case ORC_REG_SP: + return "sp"; + case ORC_REG_R10: + return "r10"; + case ORC_REG_R13: + return "r13"; + case ORC_REG_BP_INDIRECT: + return "bp(ind)"; + case ORC_REG_SP_INDIRECT: + return "sp(ind)"; + default: + return "?"; + } +} + +static const char *orc_type_name(unsigned int type) +{ + switch (type) { + case ORC_TYPE_CALL: + return "call"; + case ORC_TYPE_REGS: + return "regs"; + case ORC_TYPE_REGS_IRET: + return "iret"; + default: + return "?"; + } +} + +static void print_reg(unsigned int reg, int offset) +{ + if (reg == ORC_REG_BP_INDIRECT) + printf("(bp%+d)", offset); + else if (reg == ORC_REG_SP_INDIRECT) + printf("(sp%+d)", offset); + else if (reg == ORC_REG_UNDEFINED) + printf("(und)"); + else + printf("%s%+d", reg_name(reg), offset); +} + +int orc_dump(const char *_objname) +{ + int fd, nr_entries, i, *orc_ip = NULL, orc_size = 0; + struct orc_entry *orc = NULL; + char *name; + unsigned long nr_sections, orc_ip_addr = 0; + size_t shstrtab_idx; + Elf *elf; + Elf_Scn *scn; + GElf_Shdr sh; + GElf_Rela rela; + GElf_Sym sym; + Elf_Data *data, *symtab = NULL, *rela_orc_ip = NULL; + + + objname = _objname; + + elf_version(EV_CURRENT); + + fd = open(objname, O_RDONLY); + if (fd == -1) { + perror("open"); + return -1; + } + + elf = elf_begin(fd, ELF_C_READ_MMAP, NULL); + if (!elf) { + WARN_ELF("elf_begin"); + return -1; + } + + if (elf_getshdrnum(elf, &nr_sections)) { + WARN_ELF("elf_getshdrnum"); + return -1; + } + + if (elf_getshdrstrndx(elf, &shstrtab_idx)) { + WARN_ELF("elf_getshdrstrndx"); + return -1; + } + + for (i = 0; i < nr_sections; i++) { + scn = elf_getscn(elf, i); + if (!scn) { + WARN_ELF("elf_getscn"); + return -1; + } + + if (!gelf_getshdr(scn, &sh)) { + WARN_ELF("gelf_getshdr"); + return -1; + } + + name = elf_strptr(elf, shstrtab_idx, sh.sh_name); + if (!name) { + WARN_ELF("elf_strptr"); + return -1; + } + + data = elf_getdata(scn, NULL); + if (!data) { + WARN_ELF("elf_getdata"); + return -1; + } + + if (!strcmp(name, ".symtab")) { + symtab = data; + } else if (!strcmp(name, ".orc_unwind")) { + orc = data->d_buf; + orc_size = sh.sh_size; + } else if (!strcmp(name, ".orc_unwind_ip")) { + orc_ip = data->d_buf; + orc_ip_addr = sh.sh_addr; + } else if (!strcmp(name, ".rela.orc_unwind_ip")) { + rela_orc_ip = data; + } + } + + if (!symtab || !orc || !orc_ip) + return 0; + + if (orc_size % sizeof(*orc) != 0) { + WARN("bad .orc_unwind section size"); + return -1; + } + + nr_entries = orc_size / sizeof(*orc); + for (i = 0; i < nr_entries; i++) { + if (rela_orc_ip) { + if (!gelf_getrela(rela_orc_ip, i, &rela)) { + WARN_ELF("gelf_getrela"); + return -1; + } + + if (!gelf_getsym(symtab, GELF_R_SYM(rela.r_info), &sym)) { + WARN_ELF("gelf_getsym"); + return -1; + } + + scn = elf_getscn(elf, sym.st_shndx); + if (!scn) { + WARN_ELF("elf_getscn"); + return -1; + } + + if (!gelf_getshdr(scn, &sh)) { + WARN_ELF("gelf_getshdr"); + return -1; + } + + name = elf_strptr(elf, shstrtab_idx, sh.sh_name); + if (!name || !*name) { + WARN_ELF("elf_strptr"); + return -1; + } + + printf("%s+%lx:", name, rela.r_addend); + + } else { + printf("%lx:", orc_ip_addr + (i * sizeof(int)) + orc_ip[i]); + } + + + printf(" sp:"); + + print_reg(orc[i].sp_reg, orc[i].sp_offset); + + printf(" bp:"); + + print_reg(orc[i].bp_reg, orc[i].bp_offset); + + printf(" type:%s\n", orc_type_name(orc[i].type)); + } + + elf_end(elf); + close(fd); + + return 0; +} diff --git a/tools/objtool/orc_gen.c b/tools/objtool/orc_gen.c new file mode 100644 index 0000000..e5ca314 --- /dev/null +++ b/tools/objtool/orc_gen.c @@ -0,0 +1,214 @@ +/* + * Copyright (C) 2017 Josh Poimboeuf + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version 2 + * of the License, or (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, see . + */ + +#include +#include + +#include "orc.h" +#include "check.h" +#include "warn.h" + +int create_orc(struct objtool_file *file) +{ + struct instruction *insn; + + for_each_insn(file, insn) { + struct orc_entry *orc = &insn->orc; + struct cfi_reg *cfa = &insn->state.cfa; + struct cfi_reg *bp = &insn->state.regs[CFI_BP]; + + if (cfa->base == CFI_UNDEFINED) { + orc->sp_reg = ORC_REG_UNDEFINED; + continue; + } + + switch (cfa->base) { + case CFI_SP: + orc->sp_reg = ORC_REG_SP; + break; + case CFI_SP_INDIRECT: + orc->sp_reg = ORC_REG_SP_INDIRECT; + break; + case CFI_BP: + orc->sp_reg = ORC_REG_BP; + break; + case CFI_BP_INDIRECT: + orc->sp_reg = ORC_REG_BP_INDIRECT; + break; + case CFI_R10: + orc->sp_reg = ORC_REG_R10; + break; + case CFI_R13: + orc->sp_reg = ORC_REG_R13; + break; + case CFI_DI: + orc->sp_reg = ORC_REG_DI; + break; + case CFI_DX: + orc->sp_reg = ORC_REG_DX; + break; + default: + WARN_FUNC("unknown CFA base reg %d", + insn->sec, insn->offset, cfa->base); + return -1; + } + + switch(bp->base) { + case CFI_UNDEFINED: + orc->bp_reg = ORC_REG_UNDEFINED; + break; + case CFI_CFA: + orc->bp_reg = ORC_REG_PREV_SP; + break; + case CFI_BP: + orc->bp_reg = ORC_REG_BP; + break; + default: + WARN_FUNC("unknown BP base reg %d", + insn->sec, insn->offset, bp->base); + return -1; + } + + orc->sp_offset = cfa->offset; + orc->bp_offset = bp->offset; + orc->type = insn->state.type; + } + + return 0; +} + +static int create_orc_entry(struct section *u_sec, struct section *ip_relasec, + unsigned int idx, struct section *insn_sec, + unsigned long insn_off, struct orc_entry *o) +{ + struct orc_entry *orc; + struct rela *rela; + + /* populate ORC data */ + orc = (struct orc_entry *)u_sec->data->d_buf + idx; + memcpy(orc, o, sizeof(*orc)); + + /* populate rela for ip */ + rela = malloc(sizeof(*rela)); + if (!rela) { + perror("malloc"); + return -1; + } + memset(rela, 0, sizeof(*rela)); + + rela->sym = insn_sec->sym; + rela->addend = insn_off; + rela->type = R_X86_64_PC32; + rela->offset = idx * sizeof(int); + + list_add_tail(&rela->list, &ip_relasec->rela_list); + hash_add(ip_relasec->rela_hash, &rela->hash, rela->offset); + + return 0; +} + +int create_orc_sections(struct objtool_file *file) +{ + struct instruction *insn, *prev_insn; + struct section *sec, *u_sec, *ip_relasec; + unsigned int idx; + + struct orc_entry empty = { + .sp_reg = ORC_REG_UNDEFINED, + .bp_reg = ORC_REG_UNDEFINED, + .type = ORC_TYPE_CALL, + }; + + sec = find_section_by_name(file->elf, ".orc_unwind"); + if (sec) { + WARN("file already has .orc_unwind section, skipping"); + return -1; + } + + /* count the number of needed orcs */ + idx = 0; + for_each_sec(file, sec) { + if (!sec->text) + continue; + + prev_insn = NULL; + sec_for_each_insn(file, sec, insn) { + if (!prev_insn || + memcmp(&insn->orc, &prev_insn->orc, + sizeof(struct orc_entry))) { + idx++; + } + prev_insn = insn; + } + + /* section terminator */ + if (prev_insn) + idx++; + } + if (!idx) + return -1; + + + /* create .orc_unwind_ip and .rela.orc_unwind_ip sections */ + sec = elf_create_section(file->elf, ".orc_unwind_ip", sizeof(int), idx); + + ip_relasec = elf_create_rela_section(file->elf, sec); + if (!ip_relasec) + return -1; + + /* create .orc_unwind section */ + u_sec = elf_create_section(file->elf, ".orc_unwind", + sizeof(struct orc_entry), idx); + + /* populate sections */ + idx = 0; + for_each_sec(file, sec) { + if (!sec->text) + continue; + + prev_insn = NULL; + sec_for_each_insn(file, sec, insn) { + if (!prev_insn || memcmp(&insn->orc, &prev_insn->orc, + sizeof(struct orc_entry))) { + + if (create_orc_entry(u_sec, ip_relasec, idx, + insn->sec, insn->offset, + &insn->orc)) + return -1; + + idx++; + } + prev_insn = insn; + } + + /* section terminator */ + if (prev_insn) { + if (create_orc_entry(u_sec, ip_relasec, idx, + prev_insn->sec, + prev_insn->offset + prev_insn->len, + &empty)) + return -1; + + idx++; + } + } + + if (elf_rebuild_rela_section(ip_relasec)) + return -1; + + return 0; +} diff --git a/tools/objtool/orc_types.h b/tools/objtool/orc_types.h new file mode 100644 index 0000000..fc5cf6c --- /dev/null +++ b/tools/objtool/orc_types.h @@ -0,0 +1,85 @@ +/* + * Copyright (C) 2017 Josh Poimboeuf + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version 2 + * of the License, or (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, see . + */ + +#ifndef _ORC_TYPES_H +#define _ORC_TYPES_H + +#include +#include + +/* + * The ORC_REG_* registers are base registers which are used to find other + * registers on the stack. + * + * ORC_REG_PREV_SP, also known as DWARF Call Frame Address (CFA), is the + * address of the previous frame: the caller's SP before it called the current + * function. + * + * ORC_REG_UNDEFINED means the corresponding register's value didn't change in + * the current frame. + * + * The most commonly used base registers are SP and BP -- which the previous SP + * is usually based on -- and PREV_SP and UNDEFINED -- which the previous BP is + * usually based on. + * + * The rest of the base registers are needed for special cases like entry code + * and GCC realigned stacks. + */ +#define ORC_REG_UNDEFINED 0 +#define ORC_REG_PREV_SP 1 +#define ORC_REG_DX 2 +#define ORC_REG_DI 3 +#define ORC_REG_BP 4 +#define ORC_REG_SP 5 +#define ORC_REG_R10 6 +#define ORC_REG_R13 7 +#define ORC_REG_BP_INDIRECT 8 +#define ORC_REG_SP_INDIRECT 9 +#define ORC_REG_MAX 15 + +/* + * ORC_TYPE_CALL: Indicates that sp_reg+sp_offset resolves to PREV_SP (the + * caller's SP right before it made the call). Used for all callable + * functions, i.e. all C code and all callable asm functions. + * + * ORC_TYPE_REGS: Used in entry code to indicate that sp_reg+sp_offset points + * to a fully populated pt_regs from a syscall, interrupt, or exception. + * + * ORC_TYPE_REGS_IRET: Used in entry code to indicate that sp_reg+sp_offset + * points to the iret return frame. + */ +#define ORC_TYPE_CALL 0 +#define ORC_TYPE_REGS 1 +#define ORC_TYPE_REGS_IRET 2 + +/* + * This struct is more or less a vastly simplified version of the DWARF Call + * Frame Information standard. It contains only the necessary parts of DWARF + * CFI, simplified for ease of access by the in-kernel unwinder. It tells the + * unwinder how to find the previous SP and BP (and sometimes entry regs) on + * the stack for a given code address. Each instance of the struct corresponds + * to one or more code locations. + */ +struct orc_entry { + s16 sp_offset; + s16 bp_offset; + unsigned sp_reg:4; + unsigned bp_reg:4; + unsigned type:2; +} __packed; + +#endif /* _ORC_TYPES_H */ -- cgit v1.1 From 39358a033b2e4432052265c1fa0f36f572d8cfb5 Mon Sep 17 00:00:00 2001 From: Josh Poimboeuf Date: Tue, 11 Jul 2017 10:33:43 -0500 Subject: objtool, x86: Add facility for asm code to provide unwind hints Some asm (and inline asm) code does special things to the stack which objtool can't understand. (Nor can GCC or GNU assembler, for that matter.) In such cases we need a facility for the code to provide annotations, so the unwinder can unwind through it. This provides such a facility, in the form of unwind hints. They're similar to the GNU assembler .cfi* directives, but they give more information, and are needed in far fewer places, because objtool can fill in the blanks by following branches and adjusting the stack pointer for pushes and pops. Signed-off-by: Josh Poimboeuf Cc: Andy Lutomirski Cc: Borislav Petkov Cc: Brian Gerst Cc: Denys Vlasenko Cc: H. Peter Anvin Cc: Jiri Slaby Cc: Linus Torvalds Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: live-patching@vger.kernel.org Link: http://lkml.kernel.org/r/0f5f3c9104fca559ff4088bece1d14ae3bca52d5.1499786555.git.jpoimboe@redhat.com Signed-off-by: Ingo Molnar --- tools/objtool/Makefile | 3 + tools/objtool/check.c | 191 +++++++++++++++++++++++++++++++++++++++++++--- tools/objtool/check.h | 4 +- tools/objtool/orc_types.h | 22 ++++++ 4 files changed, 207 insertions(+), 13 deletions(-) (limited to 'tools') diff --git a/tools/objtool/Makefile b/tools/objtool/Makefile index 0e2765e..3a6425f 100644 --- a/tools/objtool/Makefile +++ b/tools/objtool/Makefile @@ -52,6 +52,9 @@ $(OBJTOOL): $(LIBSUBCMD) $(OBJTOOL_IN) diff -I'^#include' arch/x86/insn/inat.h ../../arch/x86/include/asm/inat.h >/dev/null && \ diff -I'^#include' arch/x86/insn/inat_types.h ../../arch/x86/include/asm/inat_types.h >/dev/null) \ || echo "warning: objtool: x86 instruction decoder differs from kernel" >&2 )) || true + @(test -d ../../kernel -a -d ../../tools -a -d ../objtool && (( \ + diff ../../arch/x86/include/asm/orc_types.h orc_types.h >/dev/null) \ + || echo "warning: objtool: orc_types.h differs from kernel" >&2 )) || true $(QUIET_LINK)$(CC) $(OBJTOOL_IN) $(LDFLAGS) -o $@ diff --git a/tools/objtool/check.c b/tools/objtool/check.c index cb57c52..368275d 100644 --- a/tools/objtool/check.c +++ b/tools/objtool/check.c @@ -100,7 +100,6 @@ static bool gcov_enabled(struct objtool_file *file) static bool ignore_func(struct objtool_file *file, struct symbol *func) { struct rela *rela; - struct instruction *insn; /* check for STACK_FRAME_NON_STANDARD */ if (file->whitelist && file->whitelist->rela) @@ -113,11 +112,6 @@ static bool ignore_func(struct objtool_file *file, struct symbol *func) return true; } - /* check if it has a context switching instruction */ - func_for_each_insn(file, func, insn) - if (insn->type == INSN_CONTEXT_SWITCH) - return true; - return false; } @@ -879,6 +873,99 @@ static int add_switch_table_alts(struct objtool_file *file) return 0; } +static int read_unwind_hints(struct objtool_file *file) +{ + struct section *sec, *relasec; + struct rela *rela; + struct unwind_hint *hint; + struct instruction *insn; + struct cfi_reg *cfa; + int i; + + sec = find_section_by_name(file->elf, ".discard.unwind_hints"); + if (!sec) + return 0; + + relasec = sec->rela; + if (!relasec) { + WARN("missing .rela.discard.unwind_hints section"); + return -1; + } + + if (sec->len % sizeof(struct unwind_hint)) { + WARN("struct unwind_hint size mismatch"); + return -1; + } + + file->hints = true; + + for (i = 0; i < sec->len / sizeof(struct unwind_hint); i++) { + hint = (struct unwind_hint *)sec->data->d_buf + i; + + rela = find_rela_by_dest(sec, i * sizeof(*hint)); + if (!rela) { + WARN("can't find rela for unwind_hints[%d]", i); + return -1; + } + + insn = find_insn(file, rela->sym->sec, rela->addend); + if (!insn) { + WARN("can't find insn for unwind_hints[%d]", i); + return -1; + } + + cfa = &insn->state.cfa; + + if (hint->type == UNWIND_HINT_TYPE_SAVE) { + insn->save = true; + continue; + + } else if (hint->type == UNWIND_HINT_TYPE_RESTORE) { + insn->restore = true; + insn->hint = true; + continue; + } + + insn->hint = true; + + switch (hint->sp_reg) { + case ORC_REG_UNDEFINED: + cfa->base = CFI_UNDEFINED; + break; + case ORC_REG_SP: + cfa->base = CFI_SP; + break; + case ORC_REG_BP: + cfa->base = CFI_BP; + break; + case ORC_REG_SP_INDIRECT: + cfa->base = CFI_SP_INDIRECT; + break; + case ORC_REG_R10: + cfa->base = CFI_R10; + break; + case ORC_REG_R13: + cfa->base = CFI_R13; + break; + case ORC_REG_DI: + cfa->base = CFI_DI; + break; + case ORC_REG_DX: + cfa->base = CFI_DX; + break; + default: + WARN_FUNC("unsupported unwind_hint sp base reg %d", + insn->sec, insn->offset, hint->sp_reg); + return -1; + } + + cfa->offset = hint->sp_offset; + insn->state.type = hint->type; + } + + return 0; +} + static int decode_sections(struct objtool_file *file) { int ret; @@ -909,6 +996,10 @@ static int decode_sections(struct objtool_file *file) if (ret) return ret; + ret = read_unwind_hints(file); + if (ret) + return ret; + return 0; } @@ -1382,7 +1473,7 @@ static int validate_branch(struct objtool_file *file, struct instruction *first, struct insn_state state) { struct alternative *alt; - struct instruction *insn; + struct instruction *insn, *next_insn; struct section *sec; struct symbol *func = NULL; int ret; @@ -1397,6 +1488,8 @@ static int validate_branch(struct objtool_file *file, struct instruction *first, } while (1) { + next_insn = next_insn_same_sec(file, insn); + if (file->c_file && insn->func) { if (func && func != insn->func) { WARN("%s() falls through to next function %s()", @@ -1414,13 +1507,54 @@ static int validate_branch(struct objtool_file *file, struct instruction *first, } if (insn->visited) { - if (!!insn_state_match(insn, &state)) + if (!insn->hint && !insn_state_match(insn, &state)) return 1; return 0; } - insn->state = state; + if (insn->hint) { + if (insn->restore) { + struct instruction *save_insn, *i; + + i = insn; + save_insn = NULL; + func_for_each_insn_continue_reverse(file, func, i) { + if (i->save) { + save_insn = i; + break; + } + } + + if (!save_insn) { + WARN_FUNC("no corresponding CFI save for CFI restore", + sec, insn->offset); + return 1; + } + + if (!save_insn->visited) { + /* + * Oops, no state to copy yet. + * Hopefully we can reach this + * instruction from another branch + * after the save insn has been + * visited. + */ + if (insn == first) + return 0; + + WARN_FUNC("objtool isn't smart enough to handle this CFI save/restore combo", + sec, insn->offset); + return 1; + } + + insn->state = save_insn->state; + } + + state = insn->state; + + } else + insn->state = state; insn->visited = true; @@ -1497,6 +1631,14 @@ static int validate_branch(struct objtool_file *file, struct instruction *first, return 0; + case INSN_CONTEXT_SWITCH: + if (func && (!next_insn || !next_insn->hint)) { + WARN_FUNC("unsupported instruction in callable function", + sec, insn->offset); + return 1; + } + return 0; + case INSN_STACK: if (update_insn_state(insn, &state)) return -1; @@ -1510,7 +1652,7 @@ static int validate_branch(struct objtool_file *file, struct instruction *first, if (insn->dead_end) return 0; - insn = next_insn_same_sec(file, insn); + insn = next_insn; if (!insn) { WARN("%s: unexpected end of section", sec->name); return 1; @@ -1520,6 +1662,27 @@ static int validate_branch(struct objtool_file *file, struct instruction *first, return 0; } +static int validate_unwind_hints(struct objtool_file *file) +{ + struct instruction *insn; + int ret, warnings = 0; + struct insn_state state; + + if (!file->hints) + return 0; + + clear_insn_state(&state); + + for_each_insn(file, insn) { + if (insn->hint && !insn->visited) { + ret = validate_branch(file, insn, state); + warnings += ret; + } + } + + return warnings; +} + static bool is_kasan_insn(struct instruction *insn) { return (insn->type == INSN_CALL && @@ -1665,8 +1828,9 @@ int check(const char *_objname, bool _nofp, bool orc) hash_init(file.insn_hash); file.whitelist = find_section_by_name(file.elf, ".discard.func_stack_frame_non_standard"); file.rodata = find_section_by_name(file.elf, ".rodata"); - file.ignore_unreachables = false; file.c_file = find_section_by_name(file.elf, ".comment"); + file.ignore_unreachables = false; + file.hints = false; arch_initial_func_cfi_state(&initial_func_cfi); @@ -1683,6 +1847,11 @@ int check(const char *_objname, bool _nofp, bool orc) goto out; warnings += ret; + ret = validate_unwind_hints(&file); + if (ret < 0) + goto out; + warnings += ret; + if (!warnings) { ret = validate_reachable_instructions(&file); if (ret < 0) diff --git a/tools/objtool/check.h b/tools/objtool/check.h index 046874b..ac3d4b1 100644 --- a/tools/objtool/check.h +++ b/tools/objtool/check.h @@ -43,7 +43,7 @@ struct instruction { unsigned int len; unsigned char type; unsigned long immediate; - bool alt_group, visited, dead_end, ignore; + bool alt_group, visited, dead_end, ignore, hint, save, restore; struct symbol *call_dest; struct instruction *jump_dest; struct list_head alts; @@ -58,7 +58,7 @@ struct objtool_file { struct list_head insn_list; DECLARE_HASHTABLE(insn_hash, 16); struct section *rodata, *whitelist; - bool ignore_unreachables, c_file; + bool ignore_unreachables, c_file, hints; }; int check(const char *objname, bool nofp, bool orc); diff --git a/tools/objtool/orc_types.h b/tools/objtool/orc_types.h index fc5cf6c..9c9dc57 100644 --- a/tools/objtool/orc_types.h +++ b/tools/objtool/orc_types.h @@ -61,11 +61,19 @@ * * ORC_TYPE_REGS_IRET: Used in entry code to indicate that sp_reg+sp_offset * points to the iret return frame. + * + * The UNWIND_HINT macros are used only for the unwind_hint struct. They + * aren't used in struct orc_entry due to size and complexity constraints. + * Objtool converts them to real types when it converts the hints to orc + * entries. */ #define ORC_TYPE_CALL 0 #define ORC_TYPE_REGS 1 #define ORC_TYPE_REGS_IRET 2 +#define UNWIND_HINT_TYPE_SAVE 3 +#define UNWIND_HINT_TYPE_RESTORE 4 +#ifndef __ASSEMBLY__ /* * This struct is more or less a vastly simplified version of the DWARF Call * Frame Information standard. It contains only the necessary parts of DWARF @@ -82,4 +90,18 @@ struct orc_entry { unsigned type:2; } __packed; +/* + * This struct is used by asm and inline asm code to manually annotate the + * location of registers on the stack for the ORC unwinder. + * + * Type can be either ORC_TYPE_* or UNWIND_HINT_TYPE_*. + */ +struct unwind_hint { + u32 ip; + s16 sp_offset; + u8 sp_reg; + u8 type; +}; +#endif /* __ASSEMBLY__ */ + #endif /* _ORC_TYPES_H */ -- cgit v1.1 From 69fb09f6ccdb2f070557fd1f4c56c4d646694c8e Mon Sep 17 00:00:00 2001 From: Jin Yao Date: Fri, 7 Jul 2017 13:06:34 +0800 Subject: perf annotate: Check for fused instructions Macro fusion merges two instructions to a single micro-op. Intel core platform performs this hardware optimization under limited circumstances. For example, CMP + JCC can be "fused" and executed /retired together. While with sampling this can result in the sample sometimes being on the JCC and sometimes on the CMP. So for the fused instruction pair, they could be considered together. On Nehalem, fused instruction pairs: cmp/test + jcc. On other new CPU: cmp/test/add/sub/and/inc/dec + jcc. This patch adds an x86-specific function which checks if 2 instructions are in a "fused" pair. For non-x86 arch, the function is just NULL. Changelog: v4: Move the CPU model checking to symbol__disassemble and save the CPU family/model in arch structure. It avoids checking every time when jump arrow printed. v3: Add checking for Nehalem (CMP, TEST). For other newer Intel CPUs just check it by default (CMP, TEST, ADD, SUB, AND, INC, DEC). v2: Remove the original weak function. Arnaldo points out that doing it as a weak function that will be overridden by the host arch doesn't work. So now it's implemented as an arch-specific function. Committer fix: Do not access evsel->evlist->env->cpuid, ->env can be null, introduce perf_evsel__env_cpuid(), just like perf_evsel__env_arch(), also used in this function call. The original patch was segfaulting 'perf top' + annotation. But this essentially disables this fused instructions augmentation in 'perf top', the right thing is to get the cpuid from the running kernel, left for a later patch tho. Signed-off-by: Yao Jin Tested-by: Arnaldo Carvalho de Melo Cc: Alexander Shishkin Cc: Andi Kleen Cc: Jiri Olsa Cc: Kan Liang Cc: Peter Zijlstra Link: http://lkml.kernel.org/r/1499403995-19857-2-git-send-email-yao.jin@linux.intel.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/arch/x86/annotate/instructions.c | 46 +++++++++++++++++++++++++++++ tools/perf/builtin-top.c | 2 +- tools/perf/ui/browsers/annotate.c | 4 ++- tools/perf/ui/gtk/annotate.c | 2 +- tools/perf/util/annotate.c | 22 ++++++++++++-- tools/perf/util/annotate.h | 3 +- tools/perf/util/evsel.c | 7 +++++ tools/perf/util/evsel.h | 1 + 8 files changed, 81 insertions(+), 6 deletions(-) (limited to 'tools') diff --git a/tools/perf/arch/x86/annotate/instructions.c b/tools/perf/arch/x86/annotate/instructions.c index c1625f2..d84b720 100644 --- a/tools/perf/arch/x86/annotate/instructions.c +++ b/tools/perf/arch/x86/annotate/instructions.c @@ -76,3 +76,49 @@ static struct ins x86__instructions[] = { { .name = "xbeginq", .ops = &jump_ops, }, { .name = "retq", .ops = &ret_ops, }, }; + +static bool x86__ins_is_fused(struct arch *arch, const char *ins1, + const char *ins2) +{ + if (arch->family != 6 || arch->model < 0x1e || strstr(ins2, "jmp")) + return false; + + if (arch->model == 0x1e) { + /* Nehalem */ + if ((strstr(ins1, "cmp") && !strstr(ins1, "xchg")) || + strstr(ins1, "test")) { + return true; + } + } else { + /* Newer platform */ + if ((strstr(ins1, "cmp") && !strstr(ins1, "xchg")) || + strstr(ins1, "test") || + strstr(ins1, "add") || + strstr(ins1, "sub") || + strstr(ins1, "and") || + strstr(ins1, "inc") || + strstr(ins1, "dec")) { + return true; + } + } + + return false; +} + +static int x86__cpuid_parse(struct arch *arch, char *cpuid) +{ + unsigned int family, model, stepping; + int ret; + + /* + * cpuid = "GenuineIntel,family,model,stepping" + */ + ret = sscanf(cpuid, "%*[^,],%u,%u,%u", &family, &model, &stepping); + if (ret == 3) { + arch->family = family; + arch->model = model; + return 0; + } + + return -1; +} diff --git a/tools/perf/builtin-top.c b/tools/perf/builtin-top.c index 6052376..022486d 100644 --- a/tools/perf/builtin-top.c +++ b/tools/perf/builtin-top.c @@ -134,7 +134,7 @@ static int perf_top__parse_source(struct perf_top *top, struct hist_entry *he) return err; } - err = symbol__disassemble(sym, map, NULL, 0, NULL); + err = symbol__disassemble(sym, map, NULL, 0, NULL, NULL); if (err == 0) { out_assign: top->sym_filter_entry = he; diff --git a/tools/perf/ui/browsers/annotate.c b/tools/perf/ui/browsers/annotate.c index 27f41f2..c433613 100644 --- a/tools/perf/ui/browsers/annotate.c +++ b/tools/perf/ui/browsers/annotate.c @@ -9,6 +9,7 @@ #include "../../util/symbol.h" #include "../../util/evsel.h" #include "../../util/config.h" +#include "../../util/evlist.h" #include #include #include @@ -1074,7 +1075,8 @@ int symbol__tui_annotate(struct symbol *sym, struct map *map, } err = symbol__disassemble(sym, map, perf_evsel__env_arch(evsel), - sizeof_bdl, &browser.arch); + sizeof_bdl, &browser.arch, + perf_evsel__env_cpuid(evsel)); if (err) { char msg[BUFSIZ]; symbol__strerror_disassemble(sym, map, err, msg, sizeof(msg)); diff --git a/tools/perf/ui/gtk/annotate.c b/tools/perf/ui/gtk/annotate.c index d903fd4..87e3760 100644 --- a/tools/perf/ui/gtk/annotate.c +++ b/tools/perf/ui/gtk/annotate.c @@ -169,7 +169,7 @@ static int symbol__gtk_annotate(struct symbol *sym, struct map *map, return -1; err = symbol__disassemble(sym, map, perf_evsel__env_arch(evsel), - 0, NULL); + 0, NULL, NULL); if (err) { char msg[BUFSIZ]; symbol__strerror_disassemble(sym, map, err, msg, sizeof(msg)); diff --git a/tools/perf/util/annotate.c b/tools/perf/util/annotate.c index be1caab..8748ebb 100644 --- a/tools/perf/util/annotate.c +++ b/tools/perf/util/annotate.c @@ -47,7 +47,12 @@ struct arch { bool sorted_instructions; bool initialized; void *priv; + unsigned int model; + unsigned int family; int (*init)(struct arch *arch); + bool (*ins_is_fused)(struct arch *arch, const char *ins1, + const char *ins2); + int (*cpuid_parse)(struct arch *arch, char *cpuid); struct { char comment_char; char skip_functions_char; @@ -129,6 +134,8 @@ static struct arch architectures[] = { .name = "x86", .instructions = x86__instructions, .nr_instructions = ARRAY_SIZE(x86__instructions), + .ins_is_fused = x86__ins_is_fused, + .cpuid_parse = x86__cpuid_parse, .objdump = { .comment_char = '#', }, @@ -171,6 +178,14 @@ int ins__scnprintf(struct ins *ins, char *bf, size_t size, return ins__raw_scnprintf(ins, bf, size, ops); } +bool ins__is_fused(struct arch *arch, const char *ins1, const char *ins2) +{ + if (!arch || !arch->ins_is_fused) + return false; + + return arch->ins_is_fused(arch, ins1, ins2); +} + static int call__parse(struct arch *arch, struct ins_operands *ops, struct map *map) { char *endptr, *tok, *name; @@ -1381,7 +1396,7 @@ static const char *annotate__norm_arch(const char *arch_name) int symbol__disassemble(struct symbol *sym, struct map *map, const char *arch_name, size_t privsize, - struct arch **parch) + struct arch **parch, char *cpuid) { struct dso *dso = map->dso; char command[PATH_MAX * 2]; @@ -1418,6 +1433,9 @@ int symbol__disassemble(struct symbol *sym, struct map *map, } } + if (arch->cpuid_parse && cpuid) + arch->cpuid_parse(arch, cpuid); + pr_debug("%s: filename=%s, sym=%s, start=%#" PRIx64 ", end=%#" PRIx64 "\n", __func__, symfs_filename, sym->name, map->unmap_ip(map, sym->start), map->unmap_ip(map, sym->end)); @@ -1907,7 +1925,7 @@ int symbol__tty_annotate(struct symbol *sym, struct map *map, u64 len; if (symbol__disassemble(sym, map, perf_evsel__env_arch(evsel), - 0, NULL) < 0) + 0, NULL, NULL) < 0) return -1; len = symbol__size(sym); diff --git a/tools/perf/util/annotate.h b/tools/perf/util/annotate.h index 2105503..72d7272 100644 --- a/tools/perf/util/annotate.h +++ b/tools/perf/util/annotate.h @@ -53,6 +53,7 @@ bool ins__is_jump(const struct ins *ins); bool ins__is_call(const struct ins *ins); bool ins__is_ret(const struct ins *ins); int ins__scnprintf(struct ins *ins, char *bf, size_t size, struct ins_operands *ops); +bool ins__is_fused(struct arch *arch, const char *ins1, const char *ins2); struct annotation; @@ -160,7 +161,7 @@ void symbol__annotate_zero_histograms(struct symbol *sym); int symbol__disassemble(struct symbol *sym, struct map *map, const char *arch_name, size_t privsize, - struct arch **parch); + struct arch **parch, char *cpuid); enum symbol_disassemble_errno { SYMBOL_ANNOTATE_ERRNO__SUCCESS = 0, diff --git a/tools/perf/util/evsel.c b/tools/perf/util/evsel.c index 413f74d..0e4cd60 100644 --- a/tools/perf/util/evsel.c +++ b/tools/perf/util/evsel.c @@ -2610,3 +2610,10 @@ char *perf_evsel__env_arch(struct perf_evsel *evsel) return evsel->evlist->env->arch; return NULL; } + +char *perf_evsel__env_cpuid(struct perf_evsel *evsel) +{ + if (evsel && evsel->evlist && evsel->evlist->env) + return evsel->evlist->env->cpuid; + return NULL; +} diff --git a/tools/perf/util/evsel.h b/tools/perf/util/evsel.h index d101695..219ad0c 100644 --- a/tools/perf/util/evsel.h +++ b/tools/perf/util/evsel.h @@ -436,5 +436,6 @@ int perf_event_attr__fprintf(FILE *fp, struct perf_event_attr *attr, attr__fprintf_f attr__fprintf, void *priv); char *perf_evsel__env_arch(struct perf_evsel *evsel); +char *perf_evsel__env_cpuid(struct perf_evsel *evsel); #endif /* __PERF_EVSEL_H */ -- cgit v1.1 From 7e63a13a266da652f82731b845b5c35dd866ec7e Mon Sep 17 00:00:00 2001 From: Jin Yao Date: Fri, 7 Jul 2017 13:06:35 +0800 Subject: perf annotate: Implement visual marker for macro fusion MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit For marking fused instructions clearly this patch adds a line before the first instruction of pair and joins it with the arrow of the jump to its target. For example, when "je" is selected in annotate view, the line before cmpl is displayed and joins the arrow of "je". │ ┌──cmpl $0x0,argp_program_version_hook 81.93 │ ├──je 20 │ │ lock cmpxchg %esi,0x38a9a4(%rip) │ │↓ jne 29 │ │↓ jmp 43 11.47 │20:└─→cmpxch %esi,0x38a999(%rip) That means the cmpl+je is a fused instruction pair and they should be considered together. Changelog: v3: Use Arnaldo's fix to improve the arrow origin rendering. To get the evsel->evlist->env->cpuid, save the evsel in annotate_browser. v2: new function "ins__is_fused" to check if the instructions are fused. Signed-off-by: Yao Jin Tested-by: Arnaldo Carvalho de Melo Cc: Alexander Shishkin Cc: Andi Kleen Cc: Jiri Olsa Cc: Kan Liang Cc: Peter Zijlstra Link: http://lkml.kernel.org/r/1499403995-19857-3-git-send-email-yao.jin@linux.intel.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/ui/browser.c | 29 +++++++++++++++++++++++++++++ tools/perf/ui/browser.h | 2 ++ tools/perf/ui/browsers/annotate.c | 26 ++++++++++++++++++++++++++ tools/perf/util/annotate.c | 5 +++++ tools/perf/util/annotate.h | 1 + 5 files changed, 63 insertions(+) (limited to 'tools') diff --git a/tools/perf/ui/browser.c b/tools/perf/ui/browser.c index 83874b0..f73f3f1 100644 --- a/tools/perf/ui/browser.c +++ b/tools/perf/ui/browser.c @@ -738,6 +738,35 @@ void __ui_browser__line_arrow(struct ui_browser *browser, unsigned int column, __ui_browser__line_arrow_down(browser, column, start, end); } +void ui_browser__mark_fused(struct ui_browser *browser, unsigned int column, + unsigned int row, bool arrow_down) +{ + unsigned int end_row; + + if (row >= browser->top_idx) + end_row = row - browser->top_idx; + else + return; + + SLsmg_set_char_set(1); + + if (arrow_down) { + ui_browser__gotorc(browser, end_row, column - 1); + SLsmg_write_char(SLSMG_ULCORN_CHAR); + ui_browser__gotorc(browser, end_row, column); + SLsmg_draw_hline(2); + ui_browser__gotorc(browser, end_row + 1, column - 1); + SLsmg_write_char(SLSMG_LTEE_CHAR); + } else { + ui_browser__gotorc(browser, end_row, column - 1); + SLsmg_write_char(SLSMG_LTEE_CHAR); + ui_browser__gotorc(browser, end_row, column); + SLsmg_draw_hline(2); + } + + SLsmg_set_char_set(0); +} + void ui_browser__init(void) { int i = 0; diff --git a/tools/perf/ui/browser.h b/tools/perf/ui/browser.h index be3b70e..a12eff7 100644 --- a/tools/perf/ui/browser.h +++ b/tools/perf/ui/browser.h @@ -43,6 +43,8 @@ void ui_browser__printf(struct ui_browser *browser, const char *fmt, ...); void ui_browser__write_graph(struct ui_browser *browser, int graph); void __ui_browser__line_arrow(struct ui_browser *browser, unsigned int column, u64 start, u64 end); +void ui_browser__mark_fused(struct ui_browser *browser, unsigned int column, + unsigned int row, bool arrow_down); void __ui_browser__show_title(struct ui_browser *browser, const char *title); void ui_browser__show_title(struct ui_browser *browser, const char *title); int ui_browser__show(struct ui_browser *browser, const char *title, diff --git a/tools/perf/ui/browsers/annotate.c b/tools/perf/ui/browsers/annotate.c index c433613..8d3f6f5 100644 --- a/tools/perf/ui/browsers/annotate.c +++ b/tools/perf/ui/browsers/annotate.c @@ -273,6 +273,25 @@ static bool disasm_line__is_valid_jump(struct disasm_line *dl, struct symbol *sy return true; } +static bool is_fused(struct annotate_browser *ab, struct disasm_line *cursor) +{ + struct disasm_line *pos = list_prev_entry(cursor, node); + const char *name; + + if (!pos) + return false; + + if (ins__is_lock(&pos->ins)) + name = pos->ops.locked.ins.name; + else + name = pos->ins.name; + + if (!name || !cursor->ins.name) + return false; + + return ins__is_fused(ab->arch, name, cursor->ins.name); +} + static void annotate_browser__draw_current_jump(struct ui_browser *browser) { struct annotate_browser *ab = container_of(browser, struct annotate_browser, b); @@ -308,6 +327,13 @@ static void annotate_browser__draw_current_jump(struct ui_browser *browser) ui_browser__set_color(browser, HE_COLORSET_JUMP_ARROWS); __ui_browser__line_arrow(browser, pcnt_width + 2 + ab->addr_width, from, to); + + if (is_fused(ab, cursor)) { + ui_browser__mark_fused(browser, + pcnt_width + 3 + ab->addr_width, + from - 1, + to > from ? true : false); + } } static unsigned int annotate_browser__refresh(struct ui_browser *browser) diff --git a/tools/perf/util/annotate.c b/tools/perf/util/annotate.c index 8748ebb..ef434b5 100644 --- a/tools/perf/util/annotate.c +++ b/tools/perf/util/annotate.c @@ -517,6 +517,11 @@ bool ins__is_ret(const struct ins *ins) return ins->ops == &ret_ops; } +bool ins__is_lock(const struct ins *ins) +{ + return ins->ops == &lock_ops; +} + static int ins__key_cmp(const void *name, const void *insp) { const struct ins *ins = insp; diff --git a/tools/perf/util/annotate.h b/tools/perf/util/annotate.h index 72d7272..bac698d 100644 --- a/tools/perf/util/annotate.h +++ b/tools/perf/util/annotate.h @@ -52,6 +52,7 @@ struct ins_ops { bool ins__is_jump(const struct ins *ins); bool ins__is_call(const struct ins *ins); bool ins__is_ret(const struct ins *ins); +bool ins__is_lock(const struct ins *ins); int ins__scnprintf(struct ins *ins, char *bf, size_t size, struct ins_operands *ops); bool ins__is_fused(struct arch *arch, const char *ins1, const char *ins2); -- cgit v1.1 From e000e5e33f92f6bb2d24ba2cda143cb6bb872495 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Thu, 13 Jul 2017 13:07:00 -0300 Subject: perf trace: Remove F_ from some of the fcntl command strings The initial ones already had that "F_" prefix stripped to make things shorter, some hadn't, do it now. We do this to make the 'perf trace' output more compact. At some point perhaps the best thing to do is to have the tool do this stripping automatically, letting the user also decide if this is to be done or not. For now, be consistent. Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-2iot106xkl8rgb0hb8zm3gq5@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-trace.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'tools') diff --git a/tools/perf/builtin-trace.c b/tools/perf/builtin-trace.c index 4b2a5d2..cfe1858 100644 --- a/tools/perf/builtin-trace.c +++ b/tools/perf/builtin-trace.c @@ -407,9 +407,9 @@ static DEFINE_STRARRAY(whences); static const char *fcntl_cmds[] = { "DUPFD", "GETFD", "SETFD", "GETFL", "SETFL", "GETLK", "SETLK", - "SETLKW", "SETOWN", "GETOWN", "SETSIG", "GETSIG", "F_GETLK64", - "F_SETLK64", "F_SETLKW64", "F_SETOWN_EX", "F_GETOWN_EX", - "F_GETOWNER_UIDS", + "SETLKW", "SETOWN", "GETOWN", "SETSIG", "GETSIG", "GETLK64", + "SETLK64", "SETLKW64", "SETOWN_EX", "GETOWN_EX", + "GETOWNER_UIDS", }; static DEFINE_STRARRAY(fcntl_cmds); -- cgit v1.1 From 83a516943136456b9983a0cc8ab4fa5a59860898 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Thu, 13 Jul 2017 17:14:11 -0300 Subject: perf trace: Beautify linux specific fcntl commands We were only beautifying (transforming from an integer to its name) the non-linux specific fcntl syscall cmd args, fix it: Before: # perf trace -e fcntl -p 2472 0.000 ( 0.017 ms): gnome-terminal/2472 fcntl(fd: 55, cmd: 1030) = 56 ^C# After: # trace -e fcntl -p 2472 0.000 ( 0.015 ms): gnome-terminal/2472 fcntl(fd: 55, cmd: DUPFD_CLOEXEC) = 56 ^C# Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-zigsxruk4wbfn8iylboy9wzo@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-trace.c | 57 +++++++++++++++++++++++++++++++++++++++++++--- 1 file changed, 54 insertions(+), 3 deletions(-) (limited to 'tools') diff --git a/tools/perf/builtin-trace.c b/tools/perf/builtin-trace.c index cfe1858..431ef70 100644 --- a/tools/perf/builtin-trace.c +++ b/tools/perf/builtin-trace.c @@ -64,6 +64,10 @@ # define O_CLOEXEC 02000000 #endif +#ifndef F_LINUX_SPECIFIC_BASE +# define F_LINUX_SPECIFIC_BASE 1024 +#endif + struct trace { struct perf_tool tool; struct syscalltbl *sctbl; @@ -317,6 +321,38 @@ static size_t syscall_arg__scnprintf_strarray(char *bf, size_t size, #define SCA_STRARRAY syscall_arg__scnprintf_strarray +struct strarrays { + int nr_entries; + struct strarray **entries; +}; + +#define DEFINE_STRARRAYS(array) struct strarrays strarrays__##array = { \ + .nr_entries = ARRAY_SIZE(array), \ + .entries = array, \ +} + +static size_t syscall_arg__scnprintf_strarrays(char *bf, size_t size, + struct syscall_arg *arg) +{ + struct strarrays *sas = arg->parm; + int i; + + for (i = 0; i < sas->nr_entries; ++i) { + struct strarray *sa = sas->entries[i]; + int idx = arg->val - sa->offset; + + if (idx >= 0 && idx < sa->nr_entries) { + if (sa->entries[idx] == NULL) + break; + return scnprintf(bf, size, "%s", sa->entries[idx]); + } + } + + return scnprintf(bf, size, "%d", arg->val); +} + +#define SCA_STRARRAYS syscall_arg__scnprintf_strarrays + #if defined(__i386__) || defined(__x86_64__) /* * FIXME: Make this available to all arches as soon as the ioctl beautifier @@ -413,6 +449,20 @@ static const char *fcntl_cmds[] = { }; static DEFINE_STRARRAY(fcntl_cmds); +static const char *fcntl_linux_specific_cmds[] = { + "SETLEASE", "GETLEASE", "NOTIFY", [5] = "CANCELLK", "DUPFD_CLOEXEC", + "SETPIPE_SZ", "GETPIPE_SZ", "ADD_SEALS", "GET_SEALS", +}; + +static DEFINE_STRARRAY_OFFSET(fcntl_linux_specific_cmds, F_LINUX_SPECIFIC_BASE); + +static struct strarray *fcntl_cmds_arrays[] = { + &strarray__fcntl_cmds, + &strarray__fcntl_linux_specific_cmds, +}; + +static DEFINE_STRARRAYS(fcntl_cmds_arrays); + static const char *rlimit_resources[] = { "CPU", "FSIZE", "DATA", "STACK", "CORE", "RSS", "NPROC", "NOFILE", "MEMLOCK", "AS", "LOCKS", "SIGPENDING", "MSGQUEUE", "NICE", "RTPRIO", @@ -613,8 +663,8 @@ static struct syscall_fmt { { .name = "fchownat", .errmsg = true, .arg_scnprintf = { [0] = SCA_FDAT, /* fd */ }, }, { .name = "fcntl", .errmsg = true, - .arg_scnprintf = { [1] = SCA_STRARRAY, /* cmd */ }, - .arg_parm = { [1] = &strarray__fcntl_cmds, /* cmd */ }, }, + .arg_scnprintf = { [1] = SCA_STRARRAYS, /* cmd */ }, + .arg_parm = { [1] = &strarrays__fcntl_cmds_arrays, /* cmd */ }, }, { .name = "fdatasync", .errmsg = true, }, { .name = "flock", .errmsg = true, .arg_scnprintf = { [1] = SCA_FLOCK, /* cmd */ }, }, @@ -1356,7 +1406,8 @@ static size_t syscall__scnprintf_args(struct syscall *sc, char *bf, size_t size, */ if (val == 0 && !(sc->arg_scnprintf && - sc->arg_scnprintf[arg.idx] == SCA_STRARRAY && + (sc->arg_scnprintf[arg.idx] == SCA_STRARRAY || + sc->arg_scnprintf[arg.idx] == SCA_STRARRAYS) && sc->arg_parm[arg.idx])) continue; -- cgit v1.1 From ca3cf049d2430ac83eff5c32af7b86bb2d1285bf Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Thu, 13 Jul 2017 17:15:24 -0300 Subject: tools: Update include/uapi/linux/fcntl.h copy from the kernel To get the changes in the commit c75b1d9421f8 ("fs: add fcntl() interface for setting/getting write life time hints"). Silencing this perf build warning: Warning: include/uapi/linux/fcntl.h differs from kernel We already beautify the fcntl cmd argument, so an upcoming cset will update the 'cmd' strarray to cover these new commands. The hints are in the 3rd arg, a pointer, so not yet supported in 'perf trace', for that we need to copy it somehow, probably using eBPF, a new attempt at doing that is planned. Cc: Adrian Hunter Cc: David Ahern Cc: Jens Axboe Cc: Jiri Olsa Cc: Martin K. Petersen Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-al471wzs3x48alql0tm3mnfa@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/include/uapi/linux/fcntl.h | 21 +++++++++++++++++++++ 1 file changed, 21 insertions(+) (limited to 'tools') diff --git a/tools/include/uapi/linux/fcntl.h b/tools/include/uapi/linux/fcntl.h index 813afd6..ec69d55 100644 --- a/tools/include/uapi/linux/fcntl.h +++ b/tools/include/uapi/linux/fcntl.h @@ -43,6 +43,27 @@ /* (1U << 31) is reserved for signed error codes */ /* + * Set/Get write life time hints. {GET,SET}_RW_HINT operate on the + * underlying inode, while {GET,SET}_FILE_RW_HINT operate only on + * the specific file. + */ +#define F_GET_RW_HINT (F_LINUX_SPECIFIC_BASE + 11) +#define F_SET_RW_HINT (F_LINUX_SPECIFIC_BASE + 12) +#define F_GET_FILE_RW_HINT (F_LINUX_SPECIFIC_BASE + 13) +#define F_SET_FILE_RW_HINT (F_LINUX_SPECIFIC_BASE + 14) + +/* + * Valid hint values for F_{GET,SET}_RW_HINT. 0 is "not set", or can be + * used to clear any hints previously set. + */ +#define RWF_WRITE_LIFE_NOT_SET 0 +#define RWH_WRITE_LIFE_NONE 1 +#define RWH_WRITE_LIFE_SHORT 2 +#define RWH_WRITE_LIFE_MEDIUM 3 +#define RWH_WRITE_LIFE_LONG 4 +#define RWH_WRITE_LIFE_EXTREME 5 + +/* * Types of directory notifications that may be requested. */ #define DN_ACCESS 0x00000001 /* File accessed */ -- cgit v1.1 From 274e86fdd379992f90f37fab8df3f640c50fd2cb Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Fri, 14 Jul 2017 09:38:38 -0300 Subject: perf trace beauty: Export the strarrays scnprintf method As we'll call it from the fcntl cmd scnprintf method, that needs to look at the cmd to mask the next fcntl argument when it is ignored. Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-fzlvkhew5vbxefneuciihgbc@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-trace.c | 6 ++---- tools/perf/trace/beauty/beauty.h | 3 +++ 2 files changed, 5 insertions(+), 4 deletions(-) (limited to 'tools') diff --git a/tools/perf/builtin-trace.c b/tools/perf/builtin-trace.c index 431ef70..ef1b1d4 100644 --- a/tools/perf/builtin-trace.c +++ b/tools/perf/builtin-trace.c @@ -331,8 +331,8 @@ struct strarrays { .entries = array, \ } -static size_t syscall_arg__scnprintf_strarrays(char *bf, size_t size, - struct syscall_arg *arg) +size_t syscall_arg__scnprintf_strarrays(char *bf, size_t size, + struct syscall_arg *arg) { struct strarrays *sas = arg->parm; int i; @@ -351,8 +351,6 @@ static size_t syscall_arg__scnprintf_strarrays(char *bf, size_t size, return scnprintf(bf, size, "%d", arg->val); } -#define SCA_STRARRAYS syscall_arg__scnprintf_strarrays - #if defined(__i386__) || defined(__x86_64__) /* * FIXME: Make this available to all arches as soon as the ioctl beautifier diff --git a/tools/perf/trace/beauty/beauty.h b/tools/perf/trace/beauty/beauty.h index cf50be3..a634807 100644 --- a/tools/perf/trace/beauty/beauty.h +++ b/tools/perf/trace/beauty/beauty.h @@ -15,6 +15,9 @@ struct syscall_arg { u8 mask; }; +size_t syscall_arg__scnprintf_strarrays(char *bf, size_t size, struct syscall_arg *arg); +#define SCA_STRARRAYS syscall_arg__scnprintf_strarrays + size_t syscall_arg__scnprintf_statx_flags(char *bf, size_t size, struct syscall_arg *arg); #define SCA_STATX_FLAGS syscall_arg__scnprintf_statx_flags -- cgit v1.1 From 5ca55ab6def81f8cd19a8b88f2ba4abe1aed34cc Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Fri, 14 Jul 2017 16:25:35 -0300 Subject: perf trace: Only build tools/perf/trace/beauty/ when building 'perf trace' As it calls functions in builtin-trace.c. Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-bt3lhw1rvy3jzbsp2fvvegb0@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/Build | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'tools') diff --git a/tools/perf/Build b/tools/perf/Build index bd8eeb6..b48ca40 100644 --- a/tools/perf/Build +++ b/tools/perf/Build @@ -50,6 +50,6 @@ libperf-y += util/ libperf-y += arch/ libperf-y += ui/ libperf-y += scripts/ -libperf-y += trace/beauty/ +libperf-$(CONFIG_AUDIT) += trace/beauty/ gtk-y += ui/gtk/ -- cgit v1.1 From 9cb7cf86440229fd6e6ad1718712742c344653d8 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Fri, 14 Jul 2017 09:44:50 -0300 Subject: perf trace beauty: Mask ignored fcntl 'arg' parameter A series of fcntl cmds ignore the third argument, so mask it. Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-6vtl3zq1tauamrhm8o380ptn@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-trace.c | 2 +- tools/perf/trace/beauty/Build | 1 + tools/perf/trace/beauty/beauty.h | 3 +++ tools/perf/trace/beauty/fcntl.c | 23 +++++++++++++++++++++++ 4 files changed, 28 insertions(+), 1 deletion(-) create mode 100644 tools/perf/trace/beauty/fcntl.c (limited to 'tools') diff --git a/tools/perf/builtin-trace.c b/tools/perf/builtin-trace.c index ef1b1d4..b7f79de 100644 --- a/tools/perf/builtin-trace.c +++ b/tools/perf/builtin-trace.c @@ -661,7 +661,7 @@ static struct syscall_fmt { { .name = "fchownat", .errmsg = true, .arg_scnprintf = { [0] = SCA_FDAT, /* fd */ }, }, { .name = "fcntl", .errmsg = true, - .arg_scnprintf = { [1] = SCA_STRARRAYS, /* cmd */ }, + .arg_scnprintf = { [1] = SCA_FCNTL_CMD, /* cmd */ }, .arg_parm = { [1] = &strarrays__fcntl_cmds_arrays, /* cmd */ }, }, { .name = "fdatasync", .errmsg = true, }, { .name = "flock", .errmsg = true, diff --git a/tools/perf/trace/beauty/Build b/tools/perf/trace/beauty/Build index be95ac6..c9e215b 100644 --- a/tools/perf/trace/beauty/Build +++ b/tools/perf/trace/beauty/Build @@ -1 +1,2 @@ +libperf-y += fcntl.o libperf-y += statx.o diff --git a/tools/perf/trace/beauty/beauty.h b/tools/perf/trace/beauty/beauty.h index a634807..ce01079 100644 --- a/tools/perf/trace/beauty/beauty.h +++ b/tools/perf/trace/beauty/beauty.h @@ -18,6 +18,9 @@ struct syscall_arg { size_t syscall_arg__scnprintf_strarrays(char *bf, size_t size, struct syscall_arg *arg); #define SCA_STRARRAYS syscall_arg__scnprintf_strarrays +size_t syscall_arg__scnprintf_fcntl_cmd(char *bf, size_t size, struct syscall_arg *arg); +#define SCA_FCNTL_CMD syscall_arg__scnprintf_fcntl_cmd + size_t syscall_arg__scnprintf_statx_flags(char *bf, size_t size, struct syscall_arg *arg); #define SCA_STATX_FLAGS syscall_arg__scnprintf_statx_flags diff --git a/tools/perf/trace/beauty/fcntl.c b/tools/perf/trace/beauty/fcntl.c new file mode 100644 index 0000000..7e4582c --- /dev/null +++ b/tools/perf/trace/beauty/fcntl.c @@ -0,0 +1,23 @@ +/* + * trace/beauty/fcntl.c + * + * Copyright (C) 2017, Red Hat Inc, Arnaldo Carvalho de Melo + * + * Released under the GPL v2. (and only v2, not any later version) + */ + +#include "trace/beauty/beauty.h" +#include + +size_t syscall_arg__scnprintf_fcntl_cmd(char *bf, size_t size, struct syscall_arg *arg) +{ + /* + * Some commands ignore the third fcntl argument, "arg", so mask it + */ + if (arg->val == F_GETFD || arg->val == F_GETFL || + arg->val == F_GETOWN || arg->val == F_GET_SEALS || + arg->val == F_GETLEASE || arg->val == F_GETSIG) + arg->mask |= (1 << 2); + + return syscall_arg__scnprintf_strarrays(bf, size, arg); +} -- cgit v1.1 From f9f83b3344f9ed6e1d33a87810f38eca5ed6a14f Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Fri, 14 Jul 2017 10:13:56 -0300 Subject: perf trace beauty: Allow accessing syscall args values in a syscall arg formatter For instance, fcntl's upcoming 'arg' formatter needs to look at the 'cmd' value to decide how to format its value, sometimes it is a file flags, sometimes an fd, a pointer to a structure, etc. Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-2tw2jfaqm48dtw8a4addghze@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-trace.c | 23 ++++++++++++++++------- tools/perf/trace/beauty/beauty.h | 13 +++++++++++++ 2 files changed, 29 insertions(+), 7 deletions(-) (limited to 'tools') diff --git a/tools/perf/builtin-trace.c b/tools/perf/builtin-trace.c index b7f79de..40bc0a3 100644 --- a/tools/perf/builtin-trace.c +++ b/tools/perf/builtin-trace.c @@ -1369,19 +1369,32 @@ out: * variable to read it. Most notably this avoids extended load instructions * on unaligned addresses */ +static unsigned long __syscall_arg__val(unsigned char *args, u8 idx) +{ + unsigned long val; + unsigned char *p = args + sizeof(unsigned long) * idx; + + memcpy(&val, p, sizeof(val)); + return val; +} + +unsigned long syscall_arg__val(struct syscall_arg *arg, u8 idx) +{ + return __syscall_arg__val(arg->args, idx); +} static size_t syscall__scnprintf_args(struct syscall *sc, char *bf, size_t size, unsigned char *args, struct trace *trace, struct thread *thread) { size_t printed = 0; - unsigned char *p; unsigned long val; if (sc->args != NULL) { struct format_field *field; u8 bit = 1; struct syscall_arg arg = { + .args = args, .idx = 0, .mask = 0, .trace = trace, @@ -1393,9 +1406,7 @@ static size_t syscall__scnprintf_args(struct syscall *sc, char *bf, size_t size, if (arg.mask & bit) continue; - /* special care for unaligned accesses */ - p = args + sizeof(unsigned long) * arg.idx; - memcpy(&val, p, sizeof(val)); + val = syscall_arg__val(&arg, arg.idx); /* * Suppress this argument if its value is zero and @@ -1431,9 +1442,7 @@ static size_t syscall__scnprintf_args(struct syscall *sc, char *bf, size_t size, int i = 0; while (i < 6) { - /* special care for unaligned accesses */ - p = args + sizeof(unsigned long) * i; - memcpy(&val, p, sizeof(val)); + val = __syscall_arg__val(args, i); printed += scnprintf(bf + printed, size - printed, "%sarg%d: %ld", printed ? ", " : "", i, val); diff --git a/tools/perf/trace/beauty/beauty.h b/tools/perf/trace/beauty/beauty.h index ce01079..6fbac0c 100644 --- a/tools/perf/trace/beauty/beauty.h +++ b/tools/perf/trace/beauty/beauty.h @@ -6,8 +6,19 @@ struct trace; struct thread; +/** + * @val: value of syscall argument being formatted + * @args: All the args, use syscall_args__val(arg, nth) to access one + * @thread: tid state (maps, pid, tid, etc) + * @trace: 'perf trace' internals: all threads, etc + * @parm: private area, may be an strarray, for instance + * @idx: syscall arg idx (is this the first?) + * @mask: a syscall arg may mask another arg, see syscall_arg__scnprintf_futex_op + */ + struct syscall_arg { unsigned long val; + unsigned char *args; struct thread *thread; struct trace *trace; void *parm; @@ -15,6 +26,8 @@ struct syscall_arg { u8 mask; }; +unsigned long syscall_arg__val(struct syscall_arg *arg, u8 idx); + size_t syscall_arg__scnprintf_strarrays(char *bf, size_t size, struct syscall_arg *arg); #define SCA_STRARRAYS syscall_arg__scnprintf_strarrays -- cgit v1.1 From 2c2b1623d46ce1f9c0ef5d21ff64a2c3f960e489 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Fri, 14 Jul 2017 10:19:18 -0300 Subject: perf trace beauty: Export the "int" and "hex" syscall arg formatters The most basic ones, for pointers, unaugmented fds, etc, to be used in the initial fcntl 'arg' beautifier. Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-g0lugj4vv6p4jtge32hid6q6@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-trace.c | 10 ++-------- tools/perf/trace/beauty/beauty.h | 6 ++++++ 2 files changed, 8 insertions(+), 8 deletions(-) (limited to 'tools') diff --git a/tools/perf/builtin-trace.c b/tools/perf/builtin-trace.c index 40bc0a3..9e17140 100644 --- a/tools/perf/builtin-trace.c +++ b/tools/perf/builtin-trace.c @@ -392,22 +392,16 @@ static size_t syscall_arg__scnprintf_close_fd(char *bf, size_t size, #define SCA_CLOSE_FD syscall_arg__scnprintf_close_fd -static size_t syscall_arg__scnprintf_hex(char *bf, size_t size, - struct syscall_arg *arg) +size_t syscall_arg__scnprintf_hex(char *bf, size_t size, struct syscall_arg *arg) { return scnprintf(bf, size, "%#lx", arg->val); } -#define SCA_HEX syscall_arg__scnprintf_hex - -static size_t syscall_arg__scnprintf_int(char *bf, size_t size, - struct syscall_arg *arg) +size_t syscall_arg__scnprintf_int(char *bf, size_t size, struct syscall_arg *arg) { return scnprintf(bf, size, "%d", arg->val); } -#define SCA_INT syscall_arg__scnprintf_int - static const char *bpf_cmd[] = { "MAP_CREATE", "MAP_LOOKUP_ELEM", "MAP_UPDATE_ELEM", "MAP_DELETE_ELEM", "MAP_GET_NEXT_KEY", "PROG_LOAD", diff --git a/tools/perf/trace/beauty/beauty.h b/tools/perf/trace/beauty/beauty.h index 6fbac0c..d085aac 100644 --- a/tools/perf/trace/beauty/beauty.h +++ b/tools/perf/trace/beauty/beauty.h @@ -31,6 +31,12 @@ unsigned long syscall_arg__val(struct syscall_arg *arg, u8 idx); size_t syscall_arg__scnprintf_strarrays(char *bf, size_t size, struct syscall_arg *arg); #define SCA_STRARRAYS syscall_arg__scnprintf_strarrays +size_t syscall_arg__scnprintf_hex(char *bf, size_t size, struct syscall_arg *arg); +#define SCA_HEX syscall_arg__scnprintf_hex + +size_t syscall_arg__scnprintf_int(char *bf, size_t size, struct syscall_arg *arg); +#define SCA_INT syscall_arg__scnprintf_int + size_t syscall_arg__scnprintf_fcntl_cmd(char *bf, size_t size, struct syscall_arg *arg); #define SCA_FCNTL_CMD syscall_arg__scnprintf_fcntl_cmd -- cgit v1.1 From 5dde91edbf6776daba0981c5bc47f1240427e05c Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Fri, 14 Jul 2017 10:34:16 -0300 Subject: perf trace beauty: Introduce syscall arg beautifier for long integers Will be used in the fcntl arg beautifier, that nowadays formats as '%ld' because there is no explicit arg beautifier attached, but as we will have to first decide what beautifier to use (i.e. it may be a pointer, etc) then we need to have this exported as a separate beautifier to be called from there. Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-d7bfs3m8m70j3zckeam0kk5d@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-trace.c | 5 +++++ tools/perf/trace/beauty/beauty.h | 3 +++ 2 files changed, 8 insertions(+) (limited to 'tools') diff --git a/tools/perf/builtin-trace.c b/tools/perf/builtin-trace.c index 9e17140..a323736 100644 --- a/tools/perf/builtin-trace.c +++ b/tools/perf/builtin-trace.c @@ -402,6 +402,11 @@ size_t syscall_arg__scnprintf_int(char *bf, size_t size, struct syscall_arg *arg return scnprintf(bf, size, "%d", arg->val); } +size_t syscall_arg__scnprintf_long(char *bf, size_t size, struct syscall_arg *arg) +{ + return scnprintf(bf, size, "%ld", arg->val); +} + static const char *bpf_cmd[] = { "MAP_CREATE", "MAP_LOOKUP_ELEM", "MAP_UPDATE_ELEM", "MAP_DELETE_ELEM", "MAP_GET_NEXT_KEY", "PROG_LOAD", diff --git a/tools/perf/trace/beauty/beauty.h b/tools/perf/trace/beauty/beauty.h index d085aac..61aec19 100644 --- a/tools/perf/trace/beauty/beauty.h +++ b/tools/perf/trace/beauty/beauty.h @@ -37,6 +37,9 @@ size_t syscall_arg__scnprintf_hex(char *bf, size_t size, struct syscall_arg *arg size_t syscall_arg__scnprintf_int(char *bf, size_t size, struct syscall_arg *arg); #define SCA_INT syscall_arg__scnprintf_int +size_t syscall_arg__scnprintf_long(char *bf, size_t size, struct syscall_arg *arg); +#define SCA_LONG syscall_arg__scnprintf_long + size_t syscall_arg__scnprintf_fcntl_cmd(char *bf, size_t size, struct syscall_arg *arg); #define SCA_FCNTL_CMD syscall_arg__scnprintf_fcntl_cmd -- cgit v1.1 From 84d1d8a12df33a3e707866b9a1f9872058c3d1fb Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Tue, 18 Jul 2017 17:09:21 -0300 Subject: tools include uapi asm-generic: Grab a copy of fcntl.h We'll need defines for beautifying fcntl arguments that are not available in older distros, these: trace/beauty/fcntl.c: In function 'syscall_arg__scnprintf_fcntl_arg': trace/beauty/fcntl.c:93: error: 'F_OFD_SETLK' undeclared (first use in this function) trace/beauty/fcntl.c:93: error: (Each undeclared identifier is reported only once trace/beauty/fcntl.c:93: error: for each function it appears in.) trace/beauty/fcntl.c:93: error: 'F_OFD_SETLKW' undeclared (first use in this function) trace/beauty/fcntl.c:93: error: 'F_OFD_GETLK' undeclared (first use in this function) trace/beauty/fcntl.c:94: error: 'F_GETOWN_EX' undeclared (first use in this function) trace/beauty/fcntl.c:94: error: 'F_SETOWN_EX' undeclared (first use in this function) Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-gvlw67a47e9z65jdunj4je5s@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/include/uapi/asm-generic/fcntl.h | 220 +++++++++++++++++++++++++++++++++ 1 file changed, 220 insertions(+) create mode 100644 tools/include/uapi/asm-generic/fcntl.h (limited to 'tools') diff --git a/tools/include/uapi/asm-generic/fcntl.h b/tools/include/uapi/asm-generic/fcntl.h new file mode 100644 index 0000000..ac19095 --- /dev/null +++ b/tools/include/uapi/asm-generic/fcntl.h @@ -0,0 +1,220 @@ +#ifndef _ASM_GENERIC_FCNTL_H +#define _ASM_GENERIC_FCNTL_H + +#include + +/* + * FMODE_EXEC is 0x20 + * FMODE_NONOTIFY is 0x4000000 + * These cannot be used by userspace O_* until internal and external open + * flags are split. + * -Eric Paris + */ + +/* + * When introducing new O_* bits, please check its uniqueness in fcntl_init(). + */ + +#define O_ACCMODE 00000003 +#define O_RDONLY 00000000 +#define O_WRONLY 00000001 +#define O_RDWR 00000002 +#ifndef O_CREAT +#define O_CREAT 00000100 /* not fcntl */ +#endif +#ifndef O_EXCL +#define O_EXCL 00000200 /* not fcntl */ +#endif +#ifndef O_NOCTTY +#define O_NOCTTY 00000400 /* not fcntl */ +#endif +#ifndef O_TRUNC +#define O_TRUNC 00001000 /* not fcntl */ +#endif +#ifndef O_APPEND +#define O_APPEND 00002000 +#endif +#ifndef O_NONBLOCK +#define O_NONBLOCK 00004000 +#endif +#ifndef O_DSYNC +#define O_DSYNC 00010000 /* used to be O_SYNC, see below */ +#endif +#ifndef FASYNC +#define FASYNC 00020000 /* fcntl, for BSD compatibility */ +#endif +#ifndef O_DIRECT +#define O_DIRECT 00040000 /* direct disk access hint */ +#endif +#ifndef O_LARGEFILE +#define O_LARGEFILE 00100000 +#endif +#ifndef O_DIRECTORY +#define O_DIRECTORY 00200000 /* must be a directory */ +#endif +#ifndef O_NOFOLLOW +#define O_NOFOLLOW 00400000 /* don't follow links */ +#endif +#ifndef O_NOATIME +#define O_NOATIME 01000000 +#endif +#ifndef O_CLOEXEC +#define O_CLOEXEC 02000000 /* set close_on_exec */ +#endif + +/* + * Before Linux 2.6.33 only O_DSYNC semantics were implemented, but using + * the O_SYNC flag. We continue to use the existing numerical value + * for O_DSYNC semantics now, but using the correct symbolic name for it. + * This new value is used to request true Posix O_SYNC semantics. It is + * defined in this strange way to make sure applications compiled against + * new headers get at least O_DSYNC semantics on older kernels. + * + * This has the nice side-effect that we can simply test for O_DSYNC + * wherever we do not care if O_DSYNC or O_SYNC is used. + * + * Note: __O_SYNC must never be used directly. + */ +#ifndef O_SYNC +#define __O_SYNC 04000000 +#define O_SYNC (__O_SYNC|O_DSYNC) +#endif + +#ifndef O_PATH +#define O_PATH 010000000 +#endif + +#ifndef __O_TMPFILE +#define __O_TMPFILE 020000000 +#endif + +/* a horrid kludge trying to make sure that this will fail on old kernels */ +#define O_TMPFILE (__O_TMPFILE | O_DIRECTORY) +#define O_TMPFILE_MASK (__O_TMPFILE | O_DIRECTORY | O_CREAT) + +#ifndef O_NDELAY +#define O_NDELAY O_NONBLOCK +#endif + +#define F_DUPFD 0 /* dup */ +#define F_GETFD 1 /* get close_on_exec */ +#define F_SETFD 2 /* set/clear close_on_exec */ +#define F_GETFL 3 /* get file->f_flags */ +#define F_SETFL 4 /* set file->f_flags */ +#ifndef F_GETLK +#define F_GETLK 5 +#define F_SETLK 6 +#define F_SETLKW 7 +#endif +#ifndef F_SETOWN +#define F_SETOWN 8 /* for sockets. */ +#define F_GETOWN 9 /* for sockets. */ +#endif +#ifndef F_SETSIG +#define F_SETSIG 10 /* for sockets. */ +#define F_GETSIG 11 /* for sockets. */ +#endif + +#ifndef CONFIG_64BIT +#ifndef F_GETLK64 +#define F_GETLK64 12 /* using 'struct flock64' */ +#define F_SETLK64 13 +#define F_SETLKW64 14 +#endif +#endif + +#ifndef F_SETOWN_EX +#define F_SETOWN_EX 15 +#define F_GETOWN_EX 16 +#endif + +#ifndef F_GETOWNER_UIDS +#define F_GETOWNER_UIDS 17 +#endif + +/* + * Open File Description Locks + * + * Usually record locks held by a process are released on *any* close and are + * not inherited across a fork(). + * + * These cmd values will set locks that conflict with process-associated + * record locks, but are "owned" by the open file description, not the + * process. This means that they are inherited across fork() like BSD (flock) + * locks, and they are only released automatically when the last reference to + * the the open file against which they were acquired is put. + */ +#define F_OFD_GETLK 36 +#define F_OFD_SETLK 37 +#define F_OFD_SETLKW 38 + +#define F_OWNER_TID 0 +#define F_OWNER_PID 1 +#define F_OWNER_PGRP 2 + +struct f_owner_ex { + int type; + __kernel_pid_t pid; +}; + +/* for F_[GET|SET]FL */ +#define FD_CLOEXEC 1 /* actually anything with low bit set goes */ + +/* for posix fcntl() and lockf() */ +#ifndef F_RDLCK +#define F_RDLCK 0 +#define F_WRLCK 1 +#define F_UNLCK 2 +#endif + +/* for old implementation of bsd flock () */ +#ifndef F_EXLCK +#define F_EXLCK 4 /* or 3 */ +#define F_SHLCK 8 /* or 4 */ +#endif + +/* operations for bsd flock(), also used by the kernel implementation */ +#define LOCK_SH 1 /* shared lock */ +#define LOCK_EX 2 /* exclusive lock */ +#define LOCK_NB 4 /* or'd with one of the above to prevent + blocking */ +#define LOCK_UN 8 /* remove lock */ + +#define LOCK_MAND 32 /* This is a mandatory flock ... */ +#define LOCK_READ 64 /* which allows concurrent read operations */ +#define LOCK_WRITE 128 /* which allows concurrent write operations */ +#define LOCK_RW 192 /* which allows concurrent read & write ops */ + +#define F_LINUX_SPECIFIC_BASE 1024 + +#ifndef HAVE_ARCH_STRUCT_FLOCK +#ifndef __ARCH_FLOCK_PAD +#define __ARCH_FLOCK_PAD +#endif + +struct flock { + short l_type; + short l_whence; + __kernel_off_t l_start; + __kernel_off_t l_len; + __kernel_pid_t l_pid; + __ARCH_FLOCK_PAD +}; +#endif + +#ifndef HAVE_ARCH_STRUCT_FLOCK64 +#ifndef __ARCH_FLOCK64_PAD +#define __ARCH_FLOCK64_PAD +#endif + +struct flock64 { + short l_type; + short l_whence; + __kernel_loff_t l_start; + __kernel_loff_t l_len; + __kernel_pid_t l_pid; + __ARCH_FLOCK64_PAD +}; +#endif + +#endif /* _ASM_GENERIC_FCNTL_H */ -- cgit v1.1 From 9c47f66748381ecbaa3a5eb301b81fd2fd0a282d Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Fri, 14 Jul 2017 11:05:40 -0300 Subject: perf trace beauty fcntl: Basic 'arg' beautifier Sometimes it should be printed as an hex number, like with F_SETLK, F_SETLKW and F_GETLK, that treat 'arg' as a struct flock pointer, in other cases it is just an integer. Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-2gykg6enk7vos6q0m97hkgsg@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-trace.c | 3 ++- tools/perf/trace/beauty/beauty.h | 3 +++ tools/perf/trace/beauty/fcntl.c | 18 ++++++++++++++++++ 3 files changed, 23 insertions(+), 1 deletion(-) (limited to 'tools') diff --git a/tools/perf/builtin-trace.c b/tools/perf/builtin-trace.c index a323736..2a894ea 100644 --- a/tools/perf/builtin-trace.c +++ b/tools/perf/builtin-trace.c @@ -660,7 +660,8 @@ static struct syscall_fmt { { .name = "fchownat", .errmsg = true, .arg_scnprintf = { [0] = SCA_FDAT, /* fd */ }, }, { .name = "fcntl", .errmsg = true, - .arg_scnprintf = { [1] = SCA_FCNTL_CMD, /* cmd */ }, + .arg_scnprintf = { [1] = SCA_FCNTL_CMD, /* cmd */ + [2] = SCA_FCNTL_ARG, /* arg */ }, .arg_parm = { [1] = &strarrays__fcntl_cmds_arrays, /* cmd */ }, }, { .name = "fdatasync", .errmsg = true, }, { .name = "flock", .errmsg = true, diff --git a/tools/perf/trace/beauty/beauty.h b/tools/perf/trace/beauty/beauty.h index 61aec19..d76f903 100644 --- a/tools/perf/trace/beauty/beauty.h +++ b/tools/perf/trace/beauty/beauty.h @@ -43,6 +43,9 @@ size_t syscall_arg__scnprintf_long(char *bf, size_t size, struct syscall_arg *ar size_t syscall_arg__scnprintf_fcntl_cmd(char *bf, size_t size, struct syscall_arg *arg); #define SCA_FCNTL_CMD syscall_arg__scnprintf_fcntl_cmd +size_t syscall_arg__scnprintf_fcntl_arg(char *bf, size_t size, struct syscall_arg *arg); +#define SCA_FCNTL_ARG syscall_arg__scnprintf_fcntl_arg + size_t syscall_arg__scnprintf_statx_flags(char *bf, size_t size, struct syscall_arg *arg); #define SCA_STATX_FLAGS syscall_arg__scnprintf_statx_flags diff --git a/tools/perf/trace/beauty/fcntl.c b/tools/perf/trace/beauty/fcntl.c index 7e4582c..8a5f58d 100644 --- a/tools/perf/trace/beauty/fcntl.c +++ b/tools/perf/trace/beauty/fcntl.c @@ -21,3 +21,21 @@ size_t syscall_arg__scnprintf_fcntl_cmd(char *bf, size_t size, struct syscall_ar return syscall_arg__scnprintf_strarrays(bf, size, arg); } + +size_t syscall_arg__scnprintf_fcntl_arg(char *bf, size_t size, struct syscall_arg *arg) +{ + int cmd = syscall_arg__val(arg, 1); + + /* + * We still don't grab the contents of pointers on entry or exit, + * so just print them as hex numbers + */ + if (cmd == F_SETLK || cmd == F_SETLKW || cmd == F_GETLK || + cmd == F_OFD_SETLK || cmd == F_OFD_SETLKW || cmd == F_OFD_GETLK || + cmd == F_GETOWN_EX || cmd == F_SETOWN_EX || + cmd == F_GET_RW_HINT || cmd == F_SET_RW_HINT || + cmd == F_GET_FILE_RW_HINT || cmd == F_SET_FILE_RW_HINT) + return syscall_arg__scnprintf_hex(bf, size, arg); + + return syscall_arg__scnprintf_long(bf, size, arg); +} -- cgit v1.1 From 64e4561d17c54b64368735bdf89d6873f550d918 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Thu, 13 Jul 2017 17:34:46 -0300 Subject: perf trace: Beautify new write hint fcntl commands Those introduced by the commit c75b1d9421f8 ("fs: add fcntl() interface for setting/getting write life time hints"), tested using the proggie in that commit comment: # perf trace -e fcntl ./write_hint write_hint.c fcntl: F_GET_RW_HINT: Invalid argument 0.000 ( 0.006 ms): write_hint/7576 fcntl(fd: 3, cmd: GET_RW_HINT, arg: 0x7ffc6c918da0) = -1 EINVAL Invalid argument 0.014 ( 0.004 ms): write_hint/7576 fcntl(fd: 4, cmd: GETFL) = 33794 # perf trace -e fcntl ./write_hint write_hint.c 1 fcntl: F_SET_RW_HINT: Invalid argument 0.000 ( 0.007 ms): write_hint/7578 fcntl(fd: 3, cmd: SET_RW_HINT, arg: 0x7fff03866d70) = -1 EINVAL Invalid argument 0.019 ( 0.002 ms): write_hint/7578 fcntl(fd: 4, cmd: GETFL) = 33794 # Cc: Adrian Hunter Cc: David Ahern Cc: Jens Axboe Cc: Jiri Olsa Cc: Martin K. Petersen Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-iacglkc99cchou87k62736dn@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-trace.c | 1 + 1 file changed, 1 insertion(+) (limited to 'tools') diff --git a/tools/perf/builtin-trace.c b/tools/perf/builtin-trace.c index 2a894ea..1f070bc 100644 --- a/tools/perf/builtin-trace.c +++ b/tools/perf/builtin-trace.c @@ -449,6 +449,7 @@ static DEFINE_STRARRAY(fcntl_cmds); static const char *fcntl_linux_specific_cmds[] = { "SETLEASE", "GETLEASE", "NOTIFY", [5] = "CANCELLK", "DUPFD_CLOEXEC", "SETPIPE_SZ", "GETPIPE_SZ", "ADD_SEALS", "GET_SEALS", + "GET_RW_HINT", "SET_RW_HINT", "GET_FILE_RW_HINT", "SET_FILE_RW_HINT", }; static DEFINE_STRARRAY_OFFSET(fcntl_linux_specific_cmds, F_LINUX_SPECIFIC_BASE); -- cgit v1.1 From b239ad28a8c9571c45e62e486ce75ebad17de6c1 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Fri, 14 Jul 2017 11:18:41 -0300 Subject: perf beauty open: Detach the syscall_arg agnostic bits from the flags formatter We may want to use this in other contexts, like when formatting the return of fcntl(fd, F_GETFL). Make it have the following signature, so that we can set the formatter for the return argument while processing the arguments of a syscall, as fcntl, for instance, may return fds, flags, etc, so need different return value formatters: size_t formatter(unsigned long value, char *bf, size_t size); This gets so detached from 'perf trace' internals that we may well get all these and move to a tools/lib/syscall_beauty/ library at some point and use it in other tools/ living utilities. Cc: Adrian Hunter Cc: Alexei Starovoitov Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-9aw8t22ztvnkuv26l6sw1c18@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/trace/beauty/beauty.h | 2 ++ tools/perf/trace/beauty/open_flags.c | 19 +++++++++++++------ 2 files changed, 15 insertions(+), 6 deletions(-) (limited to 'tools') diff --git a/tools/perf/trace/beauty/beauty.h b/tools/perf/trace/beauty/beauty.h index d76f903..c05bb2e 100644 --- a/tools/perf/trace/beauty/beauty.h +++ b/tools/perf/trace/beauty/beauty.h @@ -52,4 +52,6 @@ size_t syscall_arg__scnprintf_statx_flags(char *bf, size_t size, struct syscall_ size_t syscall_arg__scnprintf_statx_mask(char *bf, size_t size, struct syscall_arg *arg); #define SCA_STATX_MASK syscall_arg__scnprintf_statx_mask +size_t open__scnprintf_flags(unsigned long flags, char *bf, size_t size); + #endif /* _PERF_TRACE_BEAUTY_H */ diff --git a/tools/perf/trace/beauty/open_flags.c b/tools/perf/trace/beauty/open_flags.c index f55a459..b3c7f11 100644 --- a/tools/perf/trace/beauty/open_flags.c +++ b/tools/perf/trace/beauty/open_flags.c @@ -14,13 +14,9 @@ #define O_NOATIME 01000000 #endif -static size_t syscall_arg__scnprintf_open_flags(char *bf, size_t size, - struct syscall_arg *arg) +size_t open__scnprintf_flags(unsigned long flags, char *bf, size_t size) { - int printed = 0, flags = arg->val; - - if (!(flags & O_CREAT)) - arg->mask |= 1 << (arg->idx + 1); /* Mask the mode parm */ + int printed = 0; if (flags == 0) return scnprintf(bf, size, "RDONLY"); @@ -68,4 +64,15 @@ static size_t syscall_arg__scnprintf_open_flags(char *bf, size_t size, return printed; } +static size_t syscall_arg__scnprintf_open_flags(char *bf, size_t size, struct syscall_arg *arg) +{ + int flags = arg->val; + + if (!(flags & O_CREAT)) + arg->mask |= 1 << (arg->idx + 1); /* Mask the mode parm */ + + return open__scnprintf_flags(flags, bf, size); +} + + #define SCA_OPEN_FLAGS syscall_arg__scnprintf_open_flags -- cgit v1.1 From 84486caad942bfea338716e22010a754a0ca9199 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Fri, 14 Jul 2017 11:35:03 -0300 Subject: perf trace: Allow syscall_arg beautifiers to set a different return formatter Things like fcntl will use this to set the right formatter based on its 'cmd' argument. Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-4ea3wplb8b4j7aymj0d5uo0h@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-trace.c | 27 ++++++++++++++++++++++++++- tools/perf/trace/beauty/beauty.h | 3 +++ 2 files changed, 29 insertions(+), 1 deletion(-) (limited to 'tools') diff --git a/tools/perf/builtin-trace.c b/tools/perf/builtin-trace.c index 1f070bc..d48981c 100644 --- a/tools/perf/builtin-trace.c +++ b/tools/perf/builtin-trace.c @@ -908,6 +908,8 @@ static size_t fprintf_duration(unsigned long t, bool calculated, FILE *fp) * filename.ptr: The filename char pointer that will be vfs_getname'd * filename.entry_str_pos: Where to insert the string translated from * filename.ptr by the vfs_getname tracepoint/kprobe. + * ret_scnprintf: syscall args may set this to a different syscall return + * formatter, for instance, fcntl may return fds, file flags, etc. */ struct thread_trace { u64 entry_time; @@ -916,6 +918,7 @@ struct thread_trace { unsigned long pfmaj, pfmin; char *entry_str; double runtime_ms; + size_t (*ret_scnprintf)(unsigned long value, char *bf, size_t size); struct { unsigned long ptr; short int entry_str_pos; @@ -966,6 +969,15 @@ fail: return NULL; } + +void syscall_arg__set_ret_scnprintf(struct syscall_arg *arg, + size_t (*ret_scnprintf)(unsigned long val, char *bf, size_t size)) +{ + struct thread_trace *ttrace = thread__priv(arg->thread); + + ttrace->ret_scnprintf = ret_scnprintf; +} + #define TRACE_PFMAJ (1 << 0) #define TRACE_PFMIN (1 << 1) @@ -1390,6 +1402,14 @@ static size_t syscall__scnprintf_args(struct syscall *sc, char *bf, size_t size, { size_t printed = 0; unsigned long val; + struct thread_trace *ttrace = thread__priv(thread); + + /* + * Things like fcntl will set this in its 'cmd' formatter to pick the + * right formatter for the return value (an fd? file flags?), which is + * not needed for syscalls that always return a given type, say an fd. + */ + ttrace->ret_scnprintf = NULL; if (sc->args != NULL) { struct format_field *field; @@ -1704,7 +1724,12 @@ signed_print: fprintf(trace->output, ") = -1 %s %s", e, emsg); } else if (ret == 0 && sc->fmt->timeout) fprintf(trace->output, ") = 0 Timeout"); - else if (sc->fmt->hexret) + else if (ttrace->ret_scnprintf) { + char bf[1024]; + ttrace->ret_scnprintf(ret, bf, sizeof(bf)); + ttrace->ret_scnprintf = NULL; + fprintf(trace->output, ") = %s", bf); + } else if (sc->fmt->hexret) fprintf(trace->output, ") = %#lx", ret); else if (sc->fmt->errpid) { struct thread *child = machine__find_thread(trace->host, ret, ret); diff --git a/tools/perf/trace/beauty/beauty.h b/tools/perf/trace/beauty/beauty.h index c05bb2e..a9613d2 100644 --- a/tools/perf/trace/beauty/beauty.h +++ b/tools/perf/trace/beauty/beauty.h @@ -54,4 +54,7 @@ size_t syscall_arg__scnprintf_statx_mask(char *bf, size_t size, struct syscall_a size_t open__scnprintf_flags(unsigned long flags, char *bf, size_t size); +void syscall_arg__set_ret_scnprintf(struct syscall_arg *arg, + size_t (*ret_scnprintf)(unsigned long val, char *bf, size_t size)); + #endif /* _PERF_TRACE_BEAUTY_H */ -- cgit v1.1 From 6b3d5c97dbbc2a8e1f8d7e7d87b8c6637293e803 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Fri, 14 Jul 2017 11:57:28 -0300 Subject: perf trace beauty open flags: Support O_TMPFILE and O_NOFOLLOW The open syscall flags beautifier wasn't considering those flags, fix it. Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-ukzoldh4arrl8x2uwjafd22h@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/trace/beauty/open_flags.c | 6 ++++++ 1 file changed, 6 insertions(+) (limited to 'tools') diff --git a/tools/perf/trace/beauty/open_flags.c b/tools/perf/trace/beauty/open_flags.c index b3c7f11..5fb57be 100644 --- a/tools/perf/trace/beauty/open_flags.c +++ b/tools/perf/trace/beauty/open_flags.c @@ -14,6 +14,10 @@ #define O_NOATIME 01000000 #endif +#ifndef O_TMPFILE +#define O_TMPFILE 020000000 +#endif + size_t open__scnprintf_flags(unsigned long flags, char *bf, size_t size) { int printed = 0; @@ -34,6 +38,8 @@ size_t open__scnprintf_flags(unsigned long flags, char *bf, size_t size) P_FLAG(DIRECTORY); P_FLAG(EXCL); P_FLAG(LARGEFILE); + P_FLAG(NOFOLLOW); + P_FLAG(TMPFILE); P_FLAG(NOATIME); P_FLAG(NOCTTY); #ifdef O_NONBLOCK -- cgit v1.1 From b84148a910598f3e733e54f3cb736d11911291df Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Fri, 14 Jul 2017 12:09:31 -0300 Subject: perf trace beauty open flags: Do not depend on the system's O_LARGEFILE define In x86_64 /usr/include/bits/fcntl.h sets it to zero, so just undef it and use the standard 00100000 value when decoding the open flags arg. Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-k28megguz5snwop9obvn9mcr@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/trace/beauty/open_flags.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'tools') diff --git a/tools/perf/trace/beauty/open_flags.c b/tools/perf/trace/beauty/open_flags.c index 5fb57be..e520112 100644 --- a/tools/perf/trace/beauty/open_flags.c +++ b/tools/perf/trace/beauty/open_flags.c @@ -18,6 +18,9 @@ #define O_TMPFILE 020000000 #endif +#undef O_LARGEFILE +#define O_LARGEFILE 00100000 + size_t open__scnprintf_flags(unsigned long flags, char *bf, size_t size) { int printed = 0; -- cgit v1.1 From 89e8524abe1c584942e6b05744946877e87dd478 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Fri, 14 Jul 2017 12:14:20 -0300 Subject: perf trace beauty fcntl: Beautify F_GETFL return value The return for fcntl(fd, F_GETFL) is the fd file flags, so reuse the one for the open syscall flags parameter: 997.992 (0.002 ms): Chrome_IOThrea/19863 fcntl(fd: 144, cmd: GETFL) = RDWR|LARGEFILE Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-5nn3n4p4yfs6u0leoq880apc@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/trace/beauty/fcntl.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-) (limited to 'tools') diff --git a/tools/perf/trace/beauty/fcntl.c b/tools/perf/trace/beauty/fcntl.c index 8a5f58d..8dcbc30 100644 --- a/tools/perf/trace/beauty/fcntl.c +++ b/tools/perf/trace/beauty/fcntl.c @@ -11,13 +11,19 @@ size_t syscall_arg__scnprintf_fcntl_cmd(char *bf, size_t size, struct syscall_arg *arg) { + if (arg->val == F_GETFL) { + syscall_arg__set_ret_scnprintf(arg, open__scnprintf_flags); + goto mask_arg; + } /* * Some commands ignore the third fcntl argument, "arg", so mask it */ - if (arg->val == F_GETFD || arg->val == F_GETFL || + if (arg->val == F_GETFD || arg->val == F_GETOWN || arg->val == F_GET_SEALS || - arg->val == F_GETLEASE || arg->val == F_GETSIG) + arg->val == F_GETLEASE || arg->val == F_GETSIG) { +mask_arg: arg->mask |= (1 << 2); + } return syscall_arg__scnprintf_strarrays(bf, size, arg); } -- cgit v1.1 From e07f93c0927050184110e189a69a0b9d73c0c9b1 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Fri, 14 Jul 2017 14:23:17 -0300 Subject: perf trace beauty open flags: Move RDRW to the start of the output We were getting: 62597.859 ( 0.005 ms): TaskSchedulerF/18107 fcntl(fd: 194, cmd: GETFL) = LARGEFILE|RDWR Instead of the more familiar (from looking at strace output): 62597.859 ( 0.005 ms): TaskSchedulerF/18107 fcntl(fd: 194, cmd: GETFL) = RDWR|LARGEFILE Fix it. Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-d4d9nd88t4bu9y9odbrcb5z6@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/trace/beauty/open_flags.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'tools') diff --git a/tools/perf/trace/beauty/open_flags.c b/tools/perf/trace/beauty/open_flags.c index e520112..3c20683 100644 --- a/tools/perf/trace/beauty/open_flags.c +++ b/tools/perf/trace/beauty/open_flags.c @@ -33,6 +33,7 @@ size_t open__scnprintf_flags(unsigned long flags, char *bf, size_t size) flags &= ~O_##n; \ } + P_FLAG(RDWR); P_FLAG(APPEND); P_FLAG(ASYNC); P_FLAG(CLOEXEC); @@ -53,7 +54,6 @@ size_t open__scnprintf_flags(unsigned long flags, char *bf, size_t size) #ifdef O_PATH P_FLAG(PATH); #endif - P_FLAG(RDWR); #ifdef O_DSYNC if ((flags & O_SYNC) == O_SYNC) printed += scnprintf(bf + printed, size - printed, "%s%s", printed ? "|" : "", "SYNC"); -- cgit v1.1 From 12c0c0cef9a64ee30825122ec9d620ed29fcf5ba Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Fri, 14 Jul 2017 14:36:24 -0300 Subject: perf trace beauty fcntl flags: Beautify F_SETFL arg Result: 0.011 (0.001 ms): Chrome_IOThrea/19863 fcntl(fd: 130, cmd: SETFL, arg: RDWR|APPEND|LARGEFILE) = 0 Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-qgf8ggsq9chnjblxlq954deu@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/trace/beauty/fcntl.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'tools') diff --git a/tools/perf/trace/beauty/fcntl.c b/tools/perf/trace/beauty/fcntl.c index 8dcbc30..3456bf4 100644 --- a/tools/perf/trace/beauty/fcntl.c +++ b/tools/perf/trace/beauty/fcntl.c @@ -32,6 +32,8 @@ size_t syscall_arg__scnprintf_fcntl_arg(char *bf, size_t size, struct syscall_ar { int cmd = syscall_arg__val(arg, 1); + if (cmd == F_SETFL) + return open__scnprintf_flags(arg->val, bf, size); /* * We still don't grab the contents of pointers on entry or exit, * so just print them as hex numbers -- cgit v1.1 From c2e539d287a2e7c633ad6ecb0cb6b6f13ea05125 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Fri, 14 Jul 2017 14:50:52 -0300 Subject: perf trace beauty fcntl: Beautify F_[GS]ETFD arg/return value Now it will show 0 or CLOEXEC, the only !0 value returned by the kernel for fcntl(fd, F_GETFD). And for F_SETFD: 6870.267 ( 0.004 ms): make/29812 fcntl(fd: 7, cmd: SETFD, arg: CLOEXEC) = 0 6873.805 ( 0.002 ms): make/29816 fcntl(fd: 6, cmd: SETFD, arg: CLOEXEC) = 0 77986.150 ( 0.006 ms): alsa-sink-ALC3/2042 fcntl(fd: 45, cmd: SETFD, arg: CLOEXEC) = 0 77986.271 ( 0.006 ms): alsa-sink-ALC3/2042 fcntl(fd: 23, cmd: SETFD, arg: CLOEXEC) = 0 Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-sz9dob7t4zd6m65femazpaah@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/trace/beauty/fcntl.c | 16 ++++++++++++++-- 1 file changed, 14 insertions(+), 2 deletions(-) (limited to 'tools') diff --git a/tools/perf/trace/beauty/fcntl.c b/tools/perf/trace/beauty/fcntl.c index 3456bf4..384f177 100644 --- a/tools/perf/trace/beauty/fcntl.c +++ b/tools/perf/trace/beauty/fcntl.c @@ -7,19 +7,28 @@ */ #include "trace/beauty/beauty.h" +#include #include +static size_t fcntl__scnprintf_getfd(unsigned long val, char *bf, size_t size) +{ + return scnprintf(bf, size, "%s", val ? "CLOEXEC" : "0"); +} + size_t syscall_arg__scnprintf_fcntl_cmd(char *bf, size_t size, struct syscall_arg *arg) { if (arg->val == F_GETFL) { syscall_arg__set_ret_scnprintf(arg, open__scnprintf_flags); goto mask_arg; } + if (arg->val == F_GETFD) { + syscall_arg__set_ret_scnprintf(arg, fcntl__scnprintf_getfd); + goto mask_arg; + } /* * Some commands ignore the third fcntl argument, "arg", so mask it */ - if (arg->val == F_GETFD || - arg->val == F_GETOWN || arg->val == F_GET_SEALS || + if (arg->val == F_GETOWN || arg->val == F_GET_SEALS || arg->val == F_GETLEASE || arg->val == F_GETSIG) { mask_arg: arg->mask |= (1 << 2); @@ -32,6 +41,9 @@ size_t syscall_arg__scnprintf_fcntl_arg(char *bf, size_t size, struct syscall_ar { int cmd = syscall_arg__val(arg, 1); + if (cmd == F_SETFD) + return fcntl__scnprintf_getfd(arg->val, bf, size); + if (cmd == F_SETFL) return open__scnprintf_flags(arg->val, bf, size); /* -- cgit v1.1 From 7ee5743404e3641f3c11792761632a5a7d583587 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Fri, 14 Jul 2017 15:16:54 -0300 Subject: perf trace beauty: Give syscall return beautifier more context We need the current thread and the trace internal state so that we can use the fd beautifier to augment syscall returns, so use struct syscall_arg with some fields that make sense on returns (val, thread, trace). Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-lqag8e86ybidrh5zpqne05ov@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-trace.c | 11 ++++++++--- tools/perf/trace/beauty/beauty.h | 5 ++++- tools/perf/trace/beauty/fcntl.c | 9 +++++++-- tools/perf/trace/beauty/open_flags.c | 5 +---- 4 files changed, 20 insertions(+), 10 deletions(-) (limited to 'tools') diff --git a/tools/perf/builtin-trace.c b/tools/perf/builtin-trace.c index d48981c..cfa8bf1 100644 --- a/tools/perf/builtin-trace.c +++ b/tools/perf/builtin-trace.c @@ -918,7 +918,7 @@ struct thread_trace { unsigned long pfmaj, pfmin; char *entry_str; double runtime_ms; - size_t (*ret_scnprintf)(unsigned long value, char *bf, size_t size); + size_t (*ret_scnprintf)(char *bf, size_t size, struct syscall_arg *arg); struct { unsigned long ptr; short int entry_str_pos; @@ -971,7 +971,7 @@ fail: void syscall_arg__set_ret_scnprintf(struct syscall_arg *arg, - size_t (*ret_scnprintf)(unsigned long val, char *bf, size_t size)) + size_t (*ret_scnprintf)(char *bf, size_t size, struct syscall_arg *arg)) { struct thread_trace *ttrace = thread__priv(arg->thread); @@ -1726,7 +1726,12 @@ signed_print: fprintf(trace->output, ") = 0 Timeout"); else if (ttrace->ret_scnprintf) { char bf[1024]; - ttrace->ret_scnprintf(ret, bf, sizeof(bf)); + struct syscall_arg arg = { + .val = ret, + .thread = thread, + .trace = trace, + }; + ttrace->ret_scnprintf(bf, sizeof(bf), &arg); ttrace->ret_scnprintf = NULL; fprintf(trace->output, ") = %s", bf); } else if (sc->fmt->hexret) diff --git a/tools/perf/trace/beauty/beauty.h b/tools/perf/trace/beauty/beauty.h index a9613d2..b64c411 100644 --- a/tools/perf/trace/beauty/beauty.h +++ b/tools/perf/trace/beauty/beauty.h @@ -46,6 +46,9 @@ size_t syscall_arg__scnprintf_fcntl_cmd(char *bf, size_t size, struct syscall_ar size_t syscall_arg__scnprintf_fcntl_arg(char *bf, size_t size, struct syscall_arg *arg); #define SCA_FCNTL_ARG syscall_arg__scnprintf_fcntl_arg +size_t syscall_arg__scnprintf_open_flags(char *bf, size_t size, struct syscall_arg *arg); +#define SCA_OPEN_FLAGS syscall_arg__scnprintf_open_flags + size_t syscall_arg__scnprintf_statx_flags(char *bf, size_t size, struct syscall_arg *arg); #define SCA_STATX_FLAGS syscall_arg__scnprintf_statx_flags @@ -55,6 +58,6 @@ size_t syscall_arg__scnprintf_statx_mask(char *bf, size_t size, struct syscall_a size_t open__scnprintf_flags(unsigned long flags, char *bf, size_t size); void syscall_arg__set_ret_scnprintf(struct syscall_arg *arg, - size_t (*ret_scnprintf)(unsigned long val, char *bf, size_t size)); + size_t (*ret_scnprintf)(char *bf, size_t size, struct syscall_arg *arg)); #endif /* _PERF_TRACE_BEAUTY_H */ diff --git a/tools/perf/trace/beauty/fcntl.c b/tools/perf/trace/beauty/fcntl.c index 384f177..9de80a7 100644 --- a/tools/perf/trace/beauty/fcntl.c +++ b/tools/perf/trace/beauty/fcntl.c @@ -15,14 +15,19 @@ static size_t fcntl__scnprintf_getfd(unsigned long val, char *bf, size_t size) return scnprintf(bf, size, "%s", val ? "CLOEXEC" : "0"); } +static size_t syscall_arg__scnprintf_fcntl_getfd(char *bf, size_t size, struct syscall_arg *arg) +{ + return fcntl__scnprintf_getfd(arg->val, bf, size); +} + size_t syscall_arg__scnprintf_fcntl_cmd(char *bf, size_t size, struct syscall_arg *arg) { if (arg->val == F_GETFL) { - syscall_arg__set_ret_scnprintf(arg, open__scnprintf_flags); + syscall_arg__set_ret_scnprintf(arg, syscall_arg__scnprintf_open_flags); goto mask_arg; } if (arg->val == F_GETFD) { - syscall_arg__set_ret_scnprintf(arg, fcntl__scnprintf_getfd); + syscall_arg__set_ret_scnprintf(arg, syscall_arg__scnprintf_fcntl_getfd); goto mask_arg; } /* diff --git a/tools/perf/trace/beauty/open_flags.c b/tools/perf/trace/beauty/open_flags.c index 3c20683..e359e04 100644 --- a/tools/perf/trace/beauty/open_flags.c +++ b/tools/perf/trace/beauty/open_flags.c @@ -73,7 +73,7 @@ size_t open__scnprintf_flags(unsigned long flags, char *bf, size_t size) return printed; } -static size_t syscall_arg__scnprintf_open_flags(char *bf, size_t size, struct syscall_arg *arg) +size_t syscall_arg__scnprintf_open_flags(char *bf, size_t size, struct syscall_arg *arg) { int flags = arg->val; @@ -82,6 +82,3 @@ static size_t syscall_arg__scnprintf_open_flags(char *bf, size_t size, struct sy return open__scnprintf_flags(flags, bf, size); } - - -#define SCA_OPEN_FLAGS syscall_arg__scnprintf_open_flags -- cgit v1.1 From fc65eb8213a437b43e9de7776699aaa25dfa00df Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Fri, 14 Jul 2017 15:21:40 -0300 Subject: perf trace beauty: Export the fd beautifier for use in more places Now that the beautifiers are being split into multiple source and object files, we will need more of them exported, do it for the 'fd' one, will be used to augment the return of some syscalls that may return an 'fd', such as fcntl(fd, F_DUPFD). Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-39sosu12hhywyunqf5s74ewf@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-trace.c | 8 +------- tools/perf/trace/beauty/beauty.h | 3 +++ 2 files changed, 4 insertions(+), 7 deletions(-) (limited to 'tools') diff --git a/tools/perf/builtin-trace.c b/tools/perf/builtin-trace.c index cfa8bf1..65fa012 100644 --- a/tools/perf/builtin-trace.c +++ b/tools/perf/builtin-trace.c @@ -365,11 +365,6 @@ static size_t syscall_arg__scnprintf_strhexarray(char *bf, size_t size, #define SCA_STRHEXARRAY syscall_arg__scnprintf_strhexarray #endif /* defined(__i386__) || defined(__x86_64__) */ -static size_t syscall_arg__scnprintf_fd(char *bf, size_t size, - struct syscall_arg *arg); - -#define SCA_FD syscall_arg__scnprintf_fd - #ifndef AT_FDCWD #define AT_FDCWD -100 #endif @@ -1057,8 +1052,7 @@ static const char *thread__fd_path(struct thread *thread, int fd, return ttrace->paths.table[fd]; } -static size_t syscall_arg__scnprintf_fd(char *bf, size_t size, - struct syscall_arg *arg) +size_t syscall_arg__scnprintf_fd(char *bf, size_t size, struct syscall_arg *arg) { int fd = arg->val; size_t printed = scnprintf(bf, size, "%d", fd); diff --git a/tools/perf/trace/beauty/beauty.h b/tools/perf/trace/beauty/beauty.h index b64c411..790e830 100644 --- a/tools/perf/trace/beauty/beauty.h +++ b/tools/perf/trace/beauty/beauty.h @@ -31,6 +31,9 @@ unsigned long syscall_arg__val(struct syscall_arg *arg, u8 idx); size_t syscall_arg__scnprintf_strarrays(char *bf, size_t size, struct syscall_arg *arg); #define SCA_STRARRAYS syscall_arg__scnprintf_strarrays +size_t syscall_arg__scnprintf_fd(char *bf, size_t size, struct syscall_arg *arg); +#define SCA_FD syscall_arg__scnprintf_fd + size_t syscall_arg__scnprintf_hex(char *bf, size_t size, struct syscall_arg *arg); #define SCA_HEX syscall_arg__scnprintf_hex -- cgit v1.1 From 07a0572439450e2631f90d595965c379de105bab Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Fri, 14 Jul 2017 15:24:19 -0300 Subject: perf trace beauty fcntl: Augment the return of F_DUPFD(_CLOEXEC) Using the existing 'fd' beautifier, now we can see the path for the just dup'ed fd: 18031.338 ( 0.009 ms): gnome-terminal/2472 fcntl(fd: 55, cmd: DUPFD_CLOEXEC) = 56 Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-z0ggo126p2eobfwnjw9z16tw@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/trace/beauty/fcntl.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) (limited to 'tools') diff --git a/tools/perf/trace/beauty/fcntl.c b/tools/perf/trace/beauty/fcntl.c index 9de80a7..254ae435 100644 --- a/tools/perf/trace/beauty/fcntl.c +++ b/tools/perf/trace/beauty/fcntl.c @@ -30,6 +30,10 @@ size_t syscall_arg__scnprintf_fcntl_cmd(char *bf, size_t size, struct syscall_ar syscall_arg__set_ret_scnprintf(arg, syscall_arg__scnprintf_fcntl_getfd); goto mask_arg; } + if (arg->val == F_DUPFD_CLOEXEC || arg->val == F_DUPFD) { + syscall_arg__set_ret_scnprintf(arg, syscall_arg__scnprintf_fd); + goto out; + } /* * Some commands ignore the third fcntl argument, "arg", so mask it */ @@ -38,7 +42,7 @@ size_t syscall_arg__scnprintf_fcntl_cmd(char *bf, size_t size, struct syscall_ar mask_arg: arg->mask |= (1 << 2); } - +out: return syscall_arg__scnprintf_strarrays(bf, size, arg); } -- cgit v1.1 From ff2f1b2d35aeb672200af4772db1847fdcdfd77b Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Fri, 14 Jul 2017 15:21:40 -0300 Subject: perf trace beauty: Export the pid beautifier for use in more places Now that the beautifiers are being split into multiple source and object files, we will need more of them exported, do it for the 'pid' one, will be used to augment the return of some syscalls that may return a 'pid', such as fcntl(fd, F_GETOWN). Will also be used for fcntl(fd, F_SETOWN, pid). Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-7gr5nt9p5skp4i1w0ja1w272@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/trace/beauty/beauty.h | 3 +++ tools/perf/trace/beauty/pid.c | 4 +--- 2 files changed, 4 insertions(+), 3 deletions(-) (limited to 'tools') diff --git a/tools/perf/trace/beauty/beauty.h b/tools/perf/trace/beauty/beauty.h index 790e830..9ccf0f3 100644 --- a/tools/perf/trace/beauty/beauty.h +++ b/tools/perf/trace/beauty/beauty.h @@ -43,6 +43,9 @@ size_t syscall_arg__scnprintf_int(char *bf, size_t size, struct syscall_arg *arg size_t syscall_arg__scnprintf_long(char *bf, size_t size, struct syscall_arg *arg); #define SCA_LONG syscall_arg__scnprintf_long +size_t syscall_arg__scnprintf_pid(char *bf, size_t size, struct syscall_arg *arg); +#define SCA_PID syscall_arg__scnprintf_pid + size_t syscall_arg__scnprintf_fcntl_cmd(char *bf, size_t size, struct syscall_arg *arg); #define SCA_FCNTL_CMD syscall_arg__scnprintf_fcntl_cmd diff --git a/tools/perf/trace/beauty/pid.c b/tools/perf/trace/beauty/pid.c index 07486ea..b6d419e 100644 --- a/tools/perf/trace/beauty/pid.c +++ b/tools/perf/trace/beauty/pid.c @@ -1,4 +1,4 @@ -static size_t syscall_arg__scnprintf_pid(char *bf, size_t size, struct syscall_arg *arg) +size_t syscall_arg__scnprintf_pid(char *bf, size_t size, struct syscall_arg *arg) { int pid = arg->val; struct trace *trace = arg->trace; @@ -17,5 +17,3 @@ static size_t syscall_arg__scnprintf_pid(char *bf, size_t size, struct syscall_a return printed; } - -#define SCA_PID syscall_arg__scnprintf_pid -- cgit v1.1 From 1a4ad26393bc73d3294bb7919cb4cd2dfffbe57c Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Fri, 14 Jul 2017 15:39:40 -0300 Subject: perf trace beauty fcntl: Beautify F_GETOWN and F_SETOWN By attaching the pid beautifier to the args in the F_SETOWN case and to the syscall return on the F_GETOWN one. Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-ea1prtqvao87cdrishce7954@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/trace/beauty/fcntl.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) (limited to 'tools') diff --git a/tools/perf/trace/beauty/fcntl.c b/tools/perf/trace/beauty/fcntl.c index 254ae435..d082fd2 100644 --- a/tools/perf/trace/beauty/fcntl.c +++ b/tools/perf/trace/beauty/fcntl.c @@ -34,10 +34,14 @@ size_t syscall_arg__scnprintf_fcntl_cmd(char *bf, size_t size, struct syscall_ar syscall_arg__set_ret_scnprintf(arg, syscall_arg__scnprintf_fd); goto out; } + if (arg->val == F_GETOWN) { + syscall_arg__set_ret_scnprintf(arg, syscall_arg__scnprintf_pid); + goto mask_arg; + } /* * Some commands ignore the third fcntl argument, "arg", so mask it */ - if (arg->val == F_GETOWN || arg->val == F_GET_SEALS || + if (arg->val == F_GET_SEALS || arg->val == F_GETLEASE || arg->val == F_GETSIG) { mask_arg: arg->mask |= (1 << 2); @@ -55,6 +59,9 @@ size_t syscall_arg__scnprintf_fcntl_arg(char *bf, size_t size, struct syscall_ar if (cmd == F_SETFL) return open__scnprintf_flags(arg->val, bf, size); + + if (cmd == F_SETOWN) + return syscall_arg__scnprintf_pid(bf, size, arg); /* * We still don't grab the contents of pointers on entry or exit, * so just print them as hex numbers -- cgit v1.1 From 8b3cf3d812140dada8e82650c96d3980e5cd1b73 Mon Sep 17 00:00:00 2001 From: Shriya Date: Mon, 19 Jun 2017 12:00:42 +0530 Subject: perf pmu-events: Support additional POWER8+ PVR in mapfile Add support for POWER8+ PVR 004c0100 for Garrison Signed-off-by: Shriya Reviewed-by: Madhavan Srinivasan Cc: Alexander Shishkin Cc: Sukadev Bhattiprolu Link: http://lkml.kernel.org/r/1497853842-11023-1-git-send-email-shriyak@linux.vnet.ibm.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/pmu-events/arch/powerpc/mapfile.csv | 1 + 1 file changed, 1 insertion(+) (limited to 'tools') diff --git a/tools/perf/pmu-events/arch/powerpc/mapfile.csv b/tools/perf/pmu-events/arch/powerpc/mapfile.csv index e925baa..f03aec7 100644 --- a/tools/perf/pmu-events/arch/powerpc/mapfile.csv +++ b/tools/perf/pmu-events/arch/powerpc/mapfile.csv @@ -19,3 +19,4 @@ 004d0000,1,power8.json,core 004d0100,1,power8.json,core 004d0200,1,power8.json,core +004c0100,1,power8.json,core -- cgit v1.1 From 826db0f154ba5bee7d913635644a6f61f993a9b3 Mon Sep 17 00:00:00 2001 From: Sukadev Bhattiprolu Date: Tue, 11 Jul 2017 17:16:00 -0400 Subject: perf vendor events: Add POWER9 PMU events Add POWER9 PMU events. Signed-off-by: Sukadev Bhattiprolu Cc: Andi Kleen Cc: Jiri Olsa Cc: Madhavan Srinivasan Cc: Michael Ellerman Link: http://lkml.kernel.org/n/tip-i08irl1x1i914xsikiomvqip@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- .../perf/pmu-events/arch/powerpc/power9/cache.json | 176 +++++ .../arch/powerpc/power9/floating-point.json | 44 ++ .../pmu-events/arch/powerpc/power9/frontend.json | 446 +++++++++++ .../pmu-events/arch/powerpc/power9/marked.json | 782 +++++++++++++++++++ .../pmu-events/arch/powerpc/power9/memory.json | 158 ++++ .../perf/pmu-events/arch/powerpc/power9/other.json | 836 +++++++++++++++++++++ .../pmu-events/arch/powerpc/power9/pipeline.json | 680 +++++++++++++++++ tools/perf/pmu-events/arch/powerpc/power9/pmc.json | 146 ++++ .../arch/powerpc/power9/translation.json | 272 +++++++ 9 files changed, 3540 insertions(+) create mode 100644 tools/perf/pmu-events/arch/powerpc/power9/cache.json create mode 100644 tools/perf/pmu-events/arch/powerpc/power9/floating-point.json create mode 100644 tools/perf/pmu-events/arch/powerpc/power9/frontend.json create mode 100644 tools/perf/pmu-events/arch/powerpc/power9/marked.json create mode 100644 tools/perf/pmu-events/arch/powerpc/power9/memory.json create mode 100644 tools/perf/pmu-events/arch/powerpc/power9/other.json create mode 100644 tools/perf/pmu-events/arch/powerpc/power9/pipeline.json create mode 100644 tools/perf/pmu-events/arch/powerpc/power9/pmc.json create mode 100644 tools/perf/pmu-events/arch/powerpc/power9/translation.json (limited to 'tools') diff --git a/tools/perf/pmu-events/arch/powerpc/power9/cache.json b/tools/perf/pmu-events/arch/powerpc/power9/cache.json new file mode 100644 index 0000000..437c83b --- /dev/null +++ b/tools/perf/pmu-events/arch/powerpc/power9/cache.json @@ -0,0 +1,176 @@ +[ + {, + "EventCode": "0x1002A", + "EventName": "PM_CMPLU_STALL_LARX", + "BriefDescription": "Finish stall because the NTF instruction was a larx waiting to be satisfied", + "PublicDescription": "" + }, + {, + "EventCode": "0x1003C", + "EventName": "PM_CMPLU_STALL_DMISS_L2L3", + "BriefDescription": "Completion stall by Dcache miss which resolved in L2/L3", + "PublicDescription": "" + }, + {, + "EventCode": "0x14048", + "EventName": "PM_INST_FROM_ON_CHIP_CACHE", + "BriefDescription": "The processor's Instruction cache was reloaded either shared or modified data from another core's L2/L3 on the same chip due to an instruction fetch (not prefetch)", + "PublicDescription": "" + }, + {, + "EventCode": "0x3E054", + "EventName": "PM_LD_MISS_L1", + "BriefDescription": "Load Missed L1, counted at execution time (can be greater than loads finished). LMQ merges are not included in this count. i.e. if a load instruction misses on an address that is already allocated on the LMQ, this event will not increment for that load). Note that this count is per slice, so if a load spans multiple slices this event will increment multiple times for a single load.", + "PublicDescription": "" + }, + {, + "EventCode": "0x400F0", + "EventName": "PM_LD_MISS_L1", + "BriefDescription": "Load Missed L1, at execution time (not gated by finish, which means this counter can be greater than loads finished)", + "PublicDescription": "" + }, + {, + "EventCode": "0x1404A", + "EventName": "PM_INST_FROM_RL2L3_SHR", + "BriefDescription": "The processor's Instruction cache was reloaded with Shared (S) data from another chip's L2 or L3 on the same Node or Group (Remote), as this chip due to an instruction fetch (not prefetch)", + "PublicDescription": "" + }, + {, + "EventCode": "0x1C058", + "EventName": "PM_DTLB_MISS_16G", + "BriefDescription": "Data TLB Miss page size 16G", + "PublicDescription": "" + }, + {, + "EventCode": "0x1D15C", + "EventName": "PM_MRK_DTLB_MISS_1G", + "BriefDescription": "Marked Data TLB reload (after a miss) page size 2M. Implies radix translation was used", + "PublicDescription": "" + }, + {, + "EventCode": "0x1E056", + "EventName": "PM_CMPLU_STALL_FLUSH_ANY_THREAD", + "BriefDescription": "Cycles in which the NTC instruction is not allowed to complete because any of the 4 threads in the same core suffered a flush, which blocks completion", + "PublicDescription": "" + }, + {, + "EventCode": "0x101E6", + "EventName": "PM_THRESH_EXC_4096", + "BriefDescription": "Threshold counter exceed a count of 4096", + "PublicDescription": "" + }, + {, + "EventCode": "0x2C01A", + "EventName": "PM_CMPLU_STALL_LHS", + "BriefDescription": "Finish stall because the NTF instruction was a load that hit on an older store and it was waiting for store data", + "PublicDescription": "" + }, + {, + "EventCode": "0x2D016", + "EventName": "PM_CMPLU_STALL_FXU", + "BriefDescription": "Finish stall due to a scalar fixed point or CR instruction in the execution pipeline. These instructions get routed to the ALU, ALU2, and DIV pipes", + "PublicDescription": "" + }, + {, + "EventCode": "0x24046", + "EventName": "PM_INST_FROM_RL2L3_MOD", + "BriefDescription": "The processor's Instruction cache was reloaded with Modified (M) data from another chip's L2 or L3 on the same Node or Group (Remote), as this chip due to an instruction fetch (not prefetch)", + "PublicDescription": "" + }, + {, + "EventCode": "0x2404A", + "EventName": "PM_INST_FROM_RL4", + "BriefDescription": "The processor's Instruction cache was reloaded from another chip's L4 on the same Node or Group ( Remote) due to an instruction fetch (not prefetch)", + "PublicDescription": "" + }, + {, + "EventCode": "0x2F140", + "EventName": "PM_MRK_DPTEG_FROM_L2_MEPF", + "BriefDescription": "A Page Table Entry was loaded into the TLB from local core's L2 hit without dispatch conflicts on Mepf state. due to a marked data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x2D15E", + "EventName": "PM_MRK_DTLB_MISS_16G", + "BriefDescription": "Marked Data TLB Miss page size 16G", + "PublicDescription": "" + }, + {, + "EventCode": "0x3F14A", + "EventName": "PM_MRK_DPTEG_FROM_RMEM", + "BriefDescription": "A Page Table Entry was loaded into the TLB from another chip's memory on the same Node or Group ( Remote) due to a marked data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x3D156", + "EventName": "PM_MRK_DTLB_MISS_64K", + "BriefDescription": "Marked Data TLB Miss page size 64K", + "PublicDescription": "" + }, + {, + "EventCode": "0x3006C", + "EventName": "PM_RUN_CYC_SMT2_MODE", + "BriefDescription": "Cycles in which this thread's run latch is set and the core is in SMT2 mode", + "PublicDescription": "" + }, + {, + "EventCode": "0x300F4", + "EventName": "PM_THRD_CONC_RUN_INST", + "BriefDescription": "PPC Instructions Finished by this thread when all threads in the core had the run-latch set", + "PublicDescription": "" + }, + {, + "EventCode": "0x4C014", + "EventName": "PM_CMPLU_STALL_LMQ_FULL", + "BriefDescription": "Finish stall because the NTF instruction was a load that missed in the L1 and the LMQ was unable to accept this load miss request because it was full", + "PublicDescription": "" + }, + {, + "EventCode": "0x4C016", + "EventName": "PM_CMPLU_STALL_DMISS_L2L3_CONFLICT", + "BriefDescription": "Completion stall due to cache miss that resolves in the L2 or L3 with a conflict", + "PublicDescription": "" + }, + {, + "EventCode": "0x4D014", + "EventName": "PM_CMPLU_STALL_LOAD_FINISH", + "BriefDescription": "Finish stall because the NTF instruction was a load instruction with all its dependencies satisfied just going through the LSU pipe to finish", + "PublicDescription": "" + }, + {, + "EventCode": "0x4D016", + "EventName": "PM_CMPLU_STALL_FXLONG", + "BriefDescription": "Completion stall due to a long latency scalar fixed point instruction (division, square root)", + "PublicDescription": "" + }, + {, + "EventCode": "0x4D12A", + "EventName": "PM_MRK_DATA_FROM_RL4_CYC", + "BriefDescription": "Duration in cycles to reload from another chip's L4 on the same Node or Group ( Remote) due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x4C15E", + "EventName": "PM_MRK_DTLB_MISS_16M", + "BriefDescription": "Marked Data TLB Miss page size 16M", + "PublicDescription": "" + }, + {, + "EventCode": "0x401E4", + "EventName": "PM_MRK_DTLB_MISS", + "BriefDescription": "Marked dtlb miss", + "PublicDescription": "" + }, + {, + "EventCode": "0x401EA", + "EventName": "PM_THRESH_EXC_128", + "BriefDescription": "Threshold counter exceeded a value of 128", + "PublicDescription": "" + }, + {, + "EventCode": "0x400F6", + "EventName": "PM_BR_MPRED_CMPL", + "BriefDescription": "Number of Branch Mispredicts", + "PublicDescription": "" + } +] diff --git a/tools/perf/pmu-events/arch/powerpc/power9/floating-point.json b/tools/perf/pmu-events/arch/powerpc/power9/floating-point.json new file mode 100644 index 0000000..d4e4669 --- /dev/null +++ b/tools/perf/pmu-events/arch/powerpc/power9/floating-point.json @@ -0,0 +1,44 @@ +[ + {, + "EventCode": "0x10058", + "EventName": "PM_MEM_LOC_THRESH_IFU", + "BriefDescription": "Local Memory above threshold for IFU speculation control", + "PublicDescription": "" + }, + {, + "EventCode": "0x4505E", + "EventName": "PM_FLOP_CMPL", + "BriefDescription": "Floating Point Operation Finished", + "PublicDescription": "" + }, + {, + "EventCode": "0x1415A", + "EventName": "PM_MRK_DATA_FROM_L2_DISP_CONFLICT_LDHITST_CYC", + "BriefDescription": "Duration in cycles to reload from local core's L2 with load hit store conflict due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x2D028", + "EventName": "PM_RADIX_PWC_L2_PDE_FROM_L2", + "BriefDescription": "A Page Directory Entry was reloaded to a level 2 page walk cache from the core's L2 data cache", + "PublicDescription": "" + }, + {, + "EventCode": "0x2D154", + "EventName": "PM_MRK_DERAT_MISS_64K", + "BriefDescription": "Marked Data ERAT Miss (Data TLB Access) page size 64K", + "PublicDescription": "" + }, + {, + "EventCode": "0x30012", + "EventName": "PM_FLUSH_COMPLETION", + "BriefDescription": "The instruction that was next to complete did not complete because it suffered a flush", + "PublicDescription": "" + }, + {, + "EventCode": "0x4016E", + "EventName": "PM_THRESH_NOT_MET", + "BriefDescription": "Threshold counter did not meet threshold", + "PublicDescription": "" + } +] diff --git a/tools/perf/pmu-events/arch/powerpc/power9/frontend.json b/tools/perf/pmu-events/arch/powerpc/power9/frontend.json new file mode 100644 index 0000000..5da59d1 --- /dev/null +++ b/tools/perf/pmu-events/arch/powerpc/power9/frontend.json @@ -0,0 +1,446 @@ +[ + {, + "EventCode": "0x20036", + "EventName": "PM_BR_2PATH", + "BriefDescription": "Branches that are not strongly biased", + "PublicDescription": "" + }, + {, + "EventCode": "0x40036", + "EventName": "PM_BR_2PATH", + "BriefDescription": "Branches that are not strongly biased", + "PublicDescription": "" + }, + {, + "EventCode": "0x10004", + "EventName": "PM_CMPLU_STALL_LRQ_OTHER", + "BriefDescription": "Finish stall due to LRQ miscellaneous reasons, lost arbitration to LMQ slot, bank collisions, set prediction cleanup, set prediction multihit and others", + "PublicDescription": "" + }, + {, + "EventCode": "0x10010", + "EventName": "PM_PMC4_OVERFLOW", + "BriefDescription": "Overflow from counter 4", + "PublicDescription": "" + }, + {, + "EventCode": "0x1001A", + "EventName": "PM_LSU_SRQ_FULL_CYC", + "BriefDescription": "Cycles in which the Store Queue is full on all 4 slices. This is event is not per thread. All the threads will see the same count for this core resource", + "PublicDescription": "" + }, + {, + "EventCode": "0x10020", + "EventName": "PM_PMC4_REWIND", + "BriefDescription": "PMC4 Rewind Event", + "PublicDescription": "" + }, + {, + "EventCode": "0x1003A", + "EventName": "PM_CMPLU_STALL_LSU_FIN", + "BriefDescription": "Finish stall because the NTF instruction was an LSU op (other than a load or a store) with all its dependencies met and just going through the LSU pipe to finish", + "PublicDescription": "" + }, + {, + "EventCode": "0x1013E", + "EventName": "PM_MRK_LD_MISS_EXPOSED_CYC", + "BriefDescription": "Marked Load exposed Miss (use edge detect to count #)", + "PublicDescription": "" + }, + {, + "EventCode": "0x1C044", + "EventName": "PM_DATA_FROM_L3_NO_CONFLICT", + "BriefDescription": "The processor's data cache was reloaded from local core's L3 without conflict due to a demand load", + "PublicDescription": "" + }, + {, + "EventCode": "0x15044", + "EventName": "PM_IPTEG_FROM_L3_NO_CONFLICT", + "BriefDescription": "A Page Table Entry was loaded into the TLB from local core's L3 without conflict due to a instruction side request", + "PublicDescription": "" + }, + {, + "EventCode": "0x15046", + "EventName": "PM_IPTEG_FROM_L3.1_SHR", + "BriefDescription": "A Page Table Entry was loaded into the TLB with Shared (S) data from another core's L3 on the same chip due to a instruction side request", + "PublicDescription": "" + }, + {, + "EventCode": "0x1015E", + "EventName": "PM_MRK_FAB_RSP_RD_T_INTV", + "BriefDescription": "Sampled Read got a T intervention", + "PublicDescription": "" + }, + {, + "EventCode": "0x14054", + "EventName": "PM_INST_PUMP_CPRED", + "BriefDescription": "Pump prediction correct. Counts across all types of pumps for an instruction fetch", + "PublicDescription": "" + }, + {, + "EventCode": "0x15152", + "EventName": "PM_SYNC_MRK_BR_LINK", + "BriefDescription": "Marked Branch and link branch that can cause a synchronous interrupt", + "PublicDescription": "" + }, + {, + "EventCode": "0x1515C", + "EventName": "PM_SYNC_MRK_BR_MPRED", + "BriefDescription": "Marked Branch mispredict that can cause a synchronous interrupt", + "PublicDescription": "" + }, + {, + "EventCode": "0x1E050", + "EventName": "PM_CMPLU_STALL_TEND", + "BriefDescription": "Finish stall because the NTF instruction was a tend instruction awaiting response from L2", + "PublicDescription": "" + }, + {, + "EventCode": "0x1E15E", + "EventName": "PM_MRK_L2_TM_REQ_ABORT", + "BriefDescription": "TM abort", + "PublicDescription": "" + }, + {, + "EventCode": "0x1F054", + "EventName": "PM_TLB_HIT", + "BriefDescription": "Number of times the TLB had the data required by the instruction. Applies to both HPT and RPT", + "PublicDescription": "" + }, + {, + "EventCode": "0x1006A", + "EventName": "PM_NTC_ISSUE_HELD_DARQ_FULL", + "BriefDescription": "The NTC instruction is being held at dispatch because there are no slots in the DARQ for it", + "PublicDescription": "" + }, + {, + "EventCode": "0x101E8", + "EventName": "PM_THRESH_EXC_256", + "BriefDescription": "Threshold counter exceed a count of 256", + "PublicDescription": "" + }, + {, + "EventCode": "0x101EC", + "EventName": "PM_THRESH_MET", + "BriefDescription": "threshold exceeded", + "PublicDescription": "" + }, + {, + "EventCode": "0x100F2", + "EventName": "PM_1PLUS_PPC_CMPL", + "BriefDescription": "1 or more ppc insts finished", + "PublicDescription": "" + }, + {, + "EventCode": "0x20114", + "EventName": "PM_MRK_L2_RC_DISP", + "BriefDescription": "Marked Instruction RC dispatched in L2", + "PublicDescription": "" + }, + {, + "EventCode": "0x2C010", + "EventName": "PM_CMPLU_STALL_LSU", + "BriefDescription": "Completion stall by LSU instruction", + "PublicDescription": "" + }, + {, + "EventCode": "0x2C014", + "EventName": "PM_CMPLU_STALL_STORE_FINISH", + "BriefDescription": "Finish stall because the NTF instruction was a store with all its dependencies met, just waiting to go through the LSU pipe to finish", + "PublicDescription": "" + }, + {, + "EventCode": "0x2C01E", + "EventName": "PM_CMPLU_STALL_SYNC_PMU_INT", + "BriefDescription": "Cycles in which the NTC instruction is waiting for a synchronous PMU interrupt", + "PublicDescription": "" + }, + {, + "EventCode": "0x2D01C", + "EventName": "PM_CMPLU_STALL_STCX", + "BriefDescription": "Finish stall because the NTF instruction was a stcx waiting for response from L2", + "PublicDescription": "" + }, + {, + "EventCode": "0x2E01A", + "EventName": "PM_CMPLU_STALL_LSU_FLUSH_NEXT", + "BriefDescription": "Completion stall of one cycle because the LSU requested to flush the next iop in the sequence. It takes 1 cycle for the ISU to process this request before the LSU instruction is allowed to complete", + "PublicDescription": "" + }, + {, + "EventCode": "0x2C124", + "EventName": "PM_MRK_DATA_FROM_L2_DISP_CONFLICT_OTHER", + "BriefDescription": "The processor's data cache was reloaded from local core's L2 with dispatch conflict due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x2C042", + "EventName": "PM_DATA_FROM_L3_MEPF", + "BriefDescription": "The processor's data cache was reloaded from local core's L3 without dispatch conflicts hit on Mepf state due to a demand load", + "PublicDescription": "" + }, + {, + "EventCode": "0x2D14C", + "EventName": "PM_MRK_DATA_FROM_L3.1_ECO_SHR", + "BriefDescription": "The processor's data cache was reloaded with Shared (S) data from another core's ECO L3 on the same chip due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x25042", + "EventName": "PM_IPTEG_FROM_L3_MEPF", + "BriefDescription": "A Page Table Entry was loaded into the TLB from local core's L3 without dispatch conflicts hit on Mepf state. due to a instruction side request", + "PublicDescription": "" + }, + {, + "EventCode": "0x25044", + "EventName": "PM_IPTEG_FROM_L3.1_MOD", + "BriefDescription": "A Page Table Entry was loaded into the TLB with Modified (M) data from another core's L3 on the same chip due to a instruction side request", + "PublicDescription": "" + }, + {, + "EventCode": "0x2015E", + "EventName": "PM_MRK_FAB_RSP_RWITM_RTY", + "BriefDescription": "Sampled store did a rwitm and got a rty", + "PublicDescription": "" + }, + {, + "EventCode": "0x24050", + "EventName": "PM_IOPS_CMPL", + "BriefDescription": "Internal Operations completed", + "PublicDescription": "" + }, + {, + "EventCode": "0x24154", + "EventName": "PM_THRESH_ACC", + "BriefDescription": "This event increments every time the threshold event counter ticks. Thresholding must be enabled (via MMCRA) and the thresholding start event must occur for this counter to increment. It will stop incrementing when the thresholding stop event occurs or when thresholding is disabled, until the next time a configured thresholding start event occurs.", + "PublicDescription": "" + }, + {, + "EventCode": "0x2F152", + "EventName": "PM_MRK_FAB_RSP_DCLAIM_CYC", + "BriefDescription": "cycles L2 RC took for a dclaim", + "PublicDescription": "" + }, + {, + "EventCode": "0x200FA", + "EventName": "PM_BR_TAKEN_CMPL", + "BriefDescription": "New event for Branch Taken", + "PublicDescription": "" + }, + {, + "EventCode": "0x30014", + "EventName": "PM_CMPLU_STALL_STORE_FIN_ARB", + "BriefDescription": "Finish stall because the NTF instruction was a store waiting for a slot in the store finish pipe. This means the instruction is ready to finish but there are instructions ahead of it, using the finish pipe", + "PublicDescription": "" + }, + {, + "EventCode": "0x3001C", + "EventName": "PM_LSU_REJECT_LMQ_FULL", + "BriefDescription": "LSU Reject due to LMQ full (up to 4 per cycles)", + "PublicDescription": "" + }, + {, + "EventCode": "0x30026", + "EventName": "PM_CMPLU_STALL_STORE_DATA", + "BriefDescription": "Finish stall because the next to finish instruction was a store waiting on data", + "PublicDescription": "" + }, + {, + "EventCode": "0x3012A", + "EventName": "PM_MRK_L2_RC_DONE", + "BriefDescription": "Marked RC done", + "PublicDescription": "" + }, + {, + "EventCode": "0x35044", + "EventName": "PM_IPTEG_FROM_L3.1_ECO_SHR", + "BriefDescription": "A Page Table Entry was loaded into the TLB with Shared (S) data from another core's ECO L3 on the same chip due to a instruction side request", + "PublicDescription": "" + }, + {, + "EventCode": "0x3E04A", + "EventName": "PM_DPTEG_FROM_RMEM", + "BriefDescription": "A Page Table Entry was loaded into the TLB from another chip's memory on the same Node or Group ( Remote) due to a data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x30154", + "EventName": "PM_MRK_FAB_RSP_DCLAIM", + "BriefDescription": "Marked store had to do a dclaim", + "PublicDescription": "" + }, + {, + "EventCode": "0x3015E", + "EventName": "PM_MRK_FAB_RSP_CLAIM_RTY", + "BriefDescription": "Sampled store did a rwitm and got a rty", + "PublicDescription": "" + }, + {, + "EventCode": "0x3C056", + "EventName": "PM_DTLB_MISS_64K", + "BriefDescription": "Data TLB Miss page size 64K", + "PublicDescription": "" + }, + {, + "EventCode": "0x34050", + "EventName": "PM_INST_SYS_PUMP_CPRED", + "BriefDescription": "Initial and Final Pump Scope was system pump (prediction=correct) for an instruction fetch", + "PublicDescription": "" + }, + {, + "EventCode": "0x34052", + "EventName": "PM_INST_SYS_PUMP_MPRED", + "BriefDescription": "Final Pump Scope (system) mispredicted. Either the original scope was too small (Chip/Group) or the original scope was System and it should have been smaller. Counts for an instruction fetch", + "PublicDescription": "" + }, + {, + "EventCode": "0x34056", + "EventName": "PM_CMPLU_STALL_LSU_MFSPR", + "BriefDescription": "Finish stall because the NTF instruction was a mfspr instruction targeting an LSU SPR and it was waiting for the register data to be returned", + "PublicDescription": "" + }, + {, + "EventCode": "0x3515A", + "EventName": "PM_MRK_DATA_FROM_ON_CHIP_CACHE_CYC", + "BriefDescription": "Duration in cycles to reload either shared or modified data from another core's L2/L3 on the same chip due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x3515C", + "EventName": "PM_MRK_DATA_FROM_RL4", + "BriefDescription": "The processor's data cache was reloaded from another chip's L4 on the same Node or Group ( Remote) due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x3E15C", + "EventName": "PM_MRK_L2_TM_ST_ABORT_SISTER", + "BriefDescription": "TM marked store abort for this thread", + "PublicDescription": "" + }, + {, + "EventCode": "0x30060", + "EventName": "PM_TM_TRANS_RUN_INST", + "BriefDescription": "Run instructions completed in transactional state (gated by the run latch)", + "PublicDescription": "" + }, + {, + "EventCode": "0x301E6", + "EventName": "PM_MRK_DERAT_MISS", + "BriefDescription": "Erat Miss (TLB Access) All page sizes", + "PublicDescription": "" + }, + {, + "EventCode": "0x301EA", + "EventName": "PM_THRESH_EXC_1024", + "BriefDescription": "Threshold counter exceeded a value of 1024", + "PublicDescription": "" + }, + {, + "EventCode": "0x300FA", + "EventName": "PM_INST_FROM_L3MISS", + "BriefDescription": "Marked instruction was reloaded from a location beyond the local chiplet", + "PublicDescription": "" + }, + {, + "EventCode": "0x40116", + "EventName": "PM_MRK_LARX_FIN", + "BriefDescription": "Larx finished", + "PublicDescription": "" + }, + {, + "EventCode": "0x4C010", + "EventName": "PM_CMPLU_STALL_STORE_PIPE_ARB", + "BriefDescription": "Finish stall because the NTF instruction was a store waiting for the next relaunch opportunity after an internal reject. This means the instruction is ready to relaunch and tried once but lost arbitration", + "PublicDescription": "" + }, + {, + "EventCode": "0x4C01C", + "EventName": "PM_CMPLU_STALL_ST_FWD", + "BriefDescription": "Completion stall due to store forward", + "PublicDescription": "" + }, + {, + "EventCode": "0x4E012", + "EventName": "PM_CMPLU_STALL_MTFPSCR", + "BriefDescription": "Completion stall because the ISU is updating the register and notifying the Effective Address Table (EAT)", + "PublicDescription": "" + }, + {, + "EventCode": "0x4E016", + "EventName": "PM_CMPLU_STALL_LSAQ_ARB", + "BriefDescription": "Finish stall because the NTF instruction was a load or store that was held in LSAQ because an older instruction from SRQ or LRQ won arbitration to the LSU pipe when this instruction tried to launch", + "PublicDescription": "" + }, + {, + "EventCode": "0x4C12A", + "EventName": "PM_MRK_DATA_FROM_RL2L3_SHR_CYC", + "BriefDescription": "Duration in cycles to reload with Shared (S) data from another chip's L2 or L3 on the same Node or Group (Remote), as this chip due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x4C044", + "EventName": "PM_DATA_FROM_L3.1_ECO_MOD", + "BriefDescription": "The processor's data cache was reloaded with Modified (M) data from another core's ECO L3 on the same chip due to a demand load", + "PublicDescription": "" + }, + {, + "EventCode": "0x45044", + "EventName": "PM_IPTEG_FROM_L3.1_ECO_MOD", + "BriefDescription": "A Page Table Entry was loaded into the TLB with Modified (M) data from another core's ECO L3 on the same chip due to a instruction side request", + "PublicDescription": "" + }, + {, + "EventCode": "0x45048", + "EventName": "PM_IPTEG_FROM_DL2L3_MOD", + "BriefDescription": "A Page Table Entry was loaded into the TLB with Modified (M) data from another chip's L2 or L3 on a different Node or Group (Distant), as this chip due to a instruction side request", + "PublicDescription": "" + }, + {, + "EventCode": "0x4504E", + "EventName": "PM_IPTEG_FROM_L3MISS", + "BriefDescription": "A Page Table Entry was loaded into the TLB from a location other than the local core's L3 due to a instruction side request", + "PublicDescription": "" + }, + {, + "EventCode": "0x4E042", + "EventName": "PM_DPTEG_FROM_L3", + "BriefDescription": "A Page Table Entry was loaded into the TLB from local core's L3 due to a data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x4015E", + "EventName": "PM_MRK_FAB_RSP_RD_RTY", + "BriefDescription": "Sampled L2 reads retry count", + "PublicDescription": "" + }, + {, + "EventCode": "0x4C056", + "EventName": "PM_DTLB_MISS_16M", + "BriefDescription": "Data TLB Miss page size 16M", + "PublicDescription": "" + }, + {, + "EventCode": "0x44050", + "EventName": "PM_INST_SYS_PUMP_MPRED_RTY", + "BriefDescription": "Final Pump Scope (system) ended up larger than Initial Pump Scope (Chip/Group) for an instruction fetch", + "PublicDescription": "" + }, + {, + "EventCode": "0x44052", + "EventName": "PM_INST_PUMP_MPRED", + "BriefDescription": "Pump misprediction. Counts across all types of pumps for an instruction fetch", + "PublicDescription": "" + }, + {, + "EventCode": "0x44056", + "EventName": "PM_VECTOR_ST_CMPL", + "BriefDescription": "Number of vector store instructions completed", + "PublicDescription": "" + }, + {, + "EventCode": "0x4F150", + "EventName": "PM_MRK_FAB_RSP_RWITM_CYC", + "BriefDescription": "cycles L2 RC took for a rwitm", + "PublicDescription": "" + } +] diff --git a/tools/perf/pmu-events/arch/powerpc/power9/marked.json b/tools/perf/pmu-events/arch/powerpc/power9/marked.json new file mode 100644 index 0000000..e4d6732 --- /dev/null +++ b/tools/perf/pmu-events/arch/powerpc/power9/marked.json @@ -0,0 +1,782 @@ +[ + {, + "EventCode": "0x1002C", + "EventName": "PM_L1_DCACHE_RELOADED_ALL", + "BriefDescription": "L1 data cache reloaded for demand. If MMCR1[16] is 1, prefetches will be included as well", + "PublicDescription": "" + }, + {, + "EventCode": "0x10132", + "EventName": "PM_MRK_INST_ISSUED", + "BriefDescription": "Marked instruction issued", + "PublicDescription": "" + }, + {, + "EventCode": "0x1C042", + "EventName": "PM_DATA_FROM_L2", + "BriefDescription": "The processor's data cache was reloaded from local core's L2 due to a demand load", + "PublicDescription": "" + }, + {, + "EventCode": "0x1C046", + "EventName": "PM_DATA_FROM_L3.1_SHR", + "BriefDescription": "The processor's data cache was reloaded with Shared (S) data from another core's L3 on the same chip due to a demand load", + "PublicDescription": "" + }, + {, + "EventCode": "0x1C048", + "EventName": "PM_DATA_FROM_ON_CHIP_CACHE", + "BriefDescription": "The processor's data cache was reloaded either shared or modified data from another core's L2/L3 on the same chip due to a demand load", + "PublicDescription": "" + }, + {, + "EventCode": "0x14040", + "EventName": "PM_INST_FROM_L2_NO_CONFLICT", + "BriefDescription": "The processor's Instruction cache was reloaded from local core's L2 without conflict due to an instruction fetch (not prefetch)", + "PublicDescription": "" + }, + {, + "EventCode": "0x14042", + "EventName": "PM_INST_FROM_L2", + "BriefDescription": "The processor's Instruction cache was reloaded from local core's L2 due to an instruction fetch (not prefetch)", + "PublicDescription": "" + }, + {, + "EventCode": "0x14046", + "EventName": "PM_INST_FROM_L3.1_SHR", + "BriefDescription": "The processor's Instruction cache was reloaded with Shared (S) data from another core's L3 on the same chip due to an instruction fetch (not prefetch)", + "PublicDescription": "" + }, + {, + "EventCode": "0x1404C", + "EventName": "PM_INST_FROM_LL4", + "BriefDescription": "The processor's Instruction cache was reloaded from the local chip's L4 cache due to an instruction fetch (not prefetch)", + "PublicDescription": "" + }, + {, + "EventCode": "0x1D14C", + "EventName": "PM_MRK_DATA_FROM_LL4", + "BriefDescription": "The processor's data cache was reloaded from the local chip's L4 cache due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x15042", + "EventName": "PM_IPTEG_FROM_L2", + "BriefDescription": "A Page Table Entry was loaded into the TLB from local core's L2 due to a instruction side request", + "PublicDescription": "" + }, + {, + "EventCode": "0x1504E", + "EventName": "PM_IPTEG_FROM_L2MISS", + "BriefDescription": "A Page Table Entry was loaded into the TLB from a location other than the local core's L2 due to a instruction side request", + "PublicDescription": "" + }, + {, + "EventCode": "0x1E042", + "EventName": "PM_DPTEG_FROM_L2", + "BriefDescription": "A Page Table Entry was loaded into the TLB from local core's L2 due to a data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x1E044", + "EventName": "PM_DPTEG_FROM_L3_NO_CONFLICT", + "BriefDescription": "A Page Table Entry was loaded into the TLB from local core's L3 without conflict due to a data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x1E046", + "EventName": "PM_DPTEG_FROM_L3.1_SHR", + "BriefDescription": "A Page Table Entry was loaded into the TLB with Shared (S) data from another core's L3 on the same chip due to a data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x1F14A", + "EventName": "PM_MRK_DPTEG_FROM_RL2L3_SHR", + "BriefDescription": "A Page Table Entry was loaded into the TLB with Shared (S) data from another chip's L2 or L3 on the same Node or Group (Remote), as this chip due to a marked data side request.. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x1F14C", + "EventName": "PM_MRK_DPTEG_FROM_LL4", + "BriefDescription": "A Page Table Entry was loaded into the TLB from the local chip's L4 cache due to a marked data side request.. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x1005C", + "EventName": "PM_CMPLU_STALL_DP", + "BriefDescription": "Finish stall because the NTF instruction was a scalar instruction issued to the Double Precision execution pipe and waiting to finish. Includes binary floating point instructions in 32 and 64 bit binary floating point format. Not qualified multicycle. Qualified by NOT vector", + "PublicDescription": "" + }, + {, + "EventCode": "0x1C052", + "EventName": "PM_DATA_GRP_PUMP_MPRED_RTY", + "BriefDescription": "Final Pump Scope (Group) ended up larger than Initial Pump Scope (Chip) for a demand load", + "PublicDescription": "" + }, + {, + "EventCode": "0x1C054", + "EventName": "PM_DATA_PUMP_CPRED", + "BriefDescription": "Pump prediction correct. Counts across all types of pumps for a demand load", + "PublicDescription": "" + }, + {, + "EventCode": "0x1C05E", + "EventName": "PM_MEM_LOC_THRESH_LSU_MED", + "BriefDescription": "Local memory above threshold for data prefetch", + "PublicDescription": "" + }, + {, + "EventCode": "0x1415E", + "EventName": "PM_MRK_DATA_FROM_L3MISS_CYC", + "BriefDescription": "Duration in cycles to reload from a location other than the local core's L3 due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x1D058", + "EventName": "PM_DARQ0_10_12_ENTRIES", + "BriefDescription": "Cycles in which 10 or more DARQ entries (out of 12) are in use", + "PublicDescription": "" + }, + {, + "EventCode": "0x15150", + "EventName": "PM_SYNC_MRK_PROBE_NOP", + "BriefDescription": "Marked probeNops which can cause synchronous interrupts", + "PublicDescription": "" + }, + {, + "EventCode": "0x1E052", + "EventName": "PM_CMPLU_STALL_SLB", + "BriefDescription": "Finish stall because the NTF instruction was awaiting L2 response for an SLB", + "PublicDescription": "" + }, + {, + "EventCode": "0x1F150", + "EventName": "PM_MRK_ST_L2DISP_TO_CMPL_CYC", + "BriefDescription": "cycles from L2 rc disp to l2 rc completion", + "PublicDescription": "" + }, + {, + "EventCode": "0x1F05A", + "EventName": "PM_RADIX_PWC_L4_PTE_FROM_L2", + "BriefDescription": "A Page Table Entry was reloaded to a level 4 page walk cache from the core's L2 data cache. This is the deepest level of PWC possible for a translation", + "PublicDescription": "" + }, + {, + "EventCode": "0x1F05C", + "EventName": "PM_RADIX_PWC_L3_PDE_FROM_L3", + "BriefDescription": "A Page Directory Entry was reloaded to a level 3 page walk cache from the core's L3 data cache", + "PublicDescription": "" + }, + {, + "EventCode": "0x1006C", + "EventName": "PM_RUN_CYC_ST_MODE", + "BriefDescription": "Cycles run latch is set and core is in ST mode", + "PublicDescription": "" + }, + {, + "EventCode": "0x1016E", + "EventName": "PM_MRK_BR_CMPL", + "BriefDescription": "Branch Instruction completed", + "PublicDescription": "" + }, + {, + "EventCode": "0x101E0", + "EventName": "PM_MRK_INST_DISP", + "BriefDescription": "The thread has dispatched a randomly sampled marked instruction", + "PublicDescription": "" + }, + {, + "EventCode": "0x101E2", + "EventName": "PM_MRK_BR_TAKEN_CMPL", + "BriefDescription": "Marked Branch Taken completed", + "PublicDescription": "" + }, + {, + "EventCode": "0x2C016", + "EventName": "PM_CMPLU_STALL_PASTE", + "BriefDescription": "Finish stall because the NTF instruction was a paste waiting for response from L2", + "PublicDescription": "" + }, + {, + "EventCode": "0x2C01C", + "EventName": "PM_CMPLU_STALL_DMISS_REMOTE", + "BriefDescription": "Completion stall by Dcache miss which resolved from remote chip (cache or memory)", + "PublicDescription": "" + }, + {, + "EventCode": "0x2E01E", + "EventName": "PM_CMPLU_STALL_NTC_FLUSH", + "BriefDescription": "Completion stall due to ntc flush", + "PublicDescription": "" + }, + {, + "EventCode": "0x2C128", + "EventName": "PM_MRK_DATA_FROM_DL2L3_SHR_CYC", + "BriefDescription": "Duration in cycles to reload with Shared (S) data from another chip's L2 or L3 on a different Node or Group (Distant), as this chip due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x2C12E", + "EventName": "PM_MRK_DATA_FROM_LL4_CYC", + "BriefDescription": "Duration in cycles to reload from the local chip's L4 cache due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x2D024", + "EventName": "PM_RADIX_PWC_L2_HIT", + "BriefDescription": "A radix translation attempt missed in the TLB but hit on both the first and second levels of page walk cache.", + "PublicDescription": "" + }, + {, + "EventCode": "0x2D02A", + "EventName": "PM_RADIX_PWC_L3_PDE_FROM_L2", + "BriefDescription": "A Page Directory Entry was reloaded to a level 3 page walk cache from the core's L2 data cache", + "PublicDescription": "" + }, + {, + "EventCode": "0x2D02E", + "EventName": "PM_RADIX_PWC_L3_PTE_FROM_L2", + "BriefDescription": "A Page Table Entry was reloaded to a level 3 page walk cache from the core's L2 data cache. This implies that a level 4 PWC access was not necessary for this translation", + "PublicDescription": "" + }, + {, + "EventCode": "0x20130", + "EventName": "PM_MRK_INST_DECODED", + "BriefDescription": "An instruction was marked at decode time. Random Instruction Sampling (RIS) only", + "PublicDescription": "" + }, + {, + "EventCode": "0x20138", + "EventName": "PM_MRK_ST_NEST", + "BriefDescription": "Marked store sent to nest", + "PublicDescription": "" + }, + {, + "EventCode": "0x2013A", + "EventName": "PM_MRK_BRU_FIN", + "BriefDescription": "bru marked instr finish", + "PublicDescription": "" + }, + {, + "EventCode": "0x2C044", + "EventName": "PM_DATA_FROM_L3.1_MOD", + "BriefDescription": "The processor's data cache was reloaded with Modified (M) data from another core's L3 on the same chip due to a demand load", + "PublicDescription": "" + }, + {, + "EventCode": "0x2C048", + "EventName": "PM_DATA_FROM_LMEM", + "BriefDescription": "The processor's data cache was reloaded from the local chip's Memory due to a demand load", + "PublicDescription": "" + }, + {, + "EventCode": "0x2C04A", + "EventName": "PM_DATA_FROM_RL4", + "BriefDescription": "The processor's data cache was reloaded from another chip's L4 on the same Node or Group ( Remote) due to a demand load", + "PublicDescription": "" + }, + {, + "EventCode": "0x24044", + "EventName": "PM_INST_FROM_L3.1_MOD", + "BriefDescription": "The processor's Instruction cache was reloaded with Modified (M) data from another core's L3 on the same chip due to an instruction fetch (not prefetch)", + "PublicDescription": "" + }, + {, + "EventCode": "0x25040", + "EventName": "PM_IPTEG_FROM_L2_MEPF", + "BriefDescription": "A Page Table Entry was loaded into the TLB from local core's L2 hit without dispatch conflicts on Mepf state. due to a instruction side request", + "PublicDescription": "" + }, + {, + "EventCode": "0x2E044", + "EventName": "PM_DPTEG_FROM_L3.1_MOD", + "BriefDescription": "A Page Table Entry was loaded into the TLB with Modified (M) data from another core's L3 on the same chip due to a data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x2E048", + "EventName": "PM_DPTEG_FROM_LMEM", + "BriefDescription": "A Page Table Entry was loaded into the TLB from the local chip's Memory due to a data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x2F148", + "EventName": "PM_MRK_DPTEG_FROM_LMEM", + "BriefDescription": "A Page Table Entry was loaded into the TLB from the local chip's Memory due to a marked data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x20050", + "EventName": "PM_GRP_PUMP_CPRED", + "BriefDescription": "Initial and Final Pump Scope and data sourced across this scope was group pump for all data types excluding data prefetch (demand load,inst prefetch,inst fetch,xlate)", + "PublicDescription": "" + }, + {, + "EventCode": "0x2C052", + "EventName": "PM_DATA_GRP_PUMP_MPRED", + "BriefDescription": "Final Pump Scope (Group) ended up either larger or smaller than Initial Pump Scope for a demand load", + "PublicDescription": "" + }, + {, + "EventCode": "0x2C058", + "EventName": "PM_MEM_PREF", + "BriefDescription": "Memory prefetch for this thread. Includes L4", + "PublicDescription": "" + }, + {, + "EventCode": "0x24156", + "EventName": "PM_MRK_STCX_FIN", + "BriefDescription": "Number of marked stcx instructions finished. This includes instructions in the speculative path of a branch that may be flushed", + "PublicDescription": "" + }, + {, + "EventCode": "0x24158", + "EventName": "PM_MRK_INST", + "BriefDescription": "An instruction was marked. Includes both Random Instruction Sampling (RIS) at decode time and Random Event Sampling (RES) at the time the configured event happens", + "PublicDescription": "" + }, + {, + "EventCode": "0x2E050", + "EventName": "PM_DARQ0_7_9_ENTRIES", + "BriefDescription": "Cycles in which 7,8, or 9 DARQ entries (out of 12) are in use", + "PublicDescription": "" + }, + {, + "EventCode": "0x2E05E", + "EventName": "PM_LMQ_EMPTY_CYC", + "BriefDescription": "Cycles in which the LMQ has no pending load misses for this thread", + "PublicDescription": "" + }, + {, + "EventCode": "0x200FD", + "EventName": "PM_L1_ICACHE_MISS", + "BriefDescription": "Demand iCache Miss", + "PublicDescription": "" + }, + {, + "EventCode": "0x30006", + "EventName": "PM_CMPLU_STALL_OTHER_CMPL", + "BriefDescription": "Instructions the core completed while this tread was stalled", + "PublicDescription": "" + }, + {, + "EventCode": "0x30008", + "EventName": "PM_DISP_STARVED", + "BriefDescription": "Dispatched Starved", + "PublicDescription": "" + }, + {, + "EventCode": "0x3000A", + "EventName": "PM_CMPLU_STALL_PM", + "BriefDescription": "Finish stall because the NTF instruction was issued to the Permute execution pipe and waiting to finish. Includes permute and decimal fixed point instructions (128 bit BCD arithmetic) + a few 128 bit fixpoint add/subtract instructions with carry. Not qualified by vector or multicycle", + "PublicDescription": "" + }, + {, + "EventCode": "0x3000E", + "EventName": "PM_FXU_1PLUS_BUSY", + "BriefDescription": "At least one of the 4 FXU units is busy", + "PublicDescription": "" + }, + {, + "EventCode": "0x30028", + "EventName": "PM_CMPLU_STALL_SPEC_FINISH", + "BriefDescription": "Finish stall while waiting for the non-speculative finish of either a stcx waiting for its result or a load waiting for non-critical sectors of data and ECC", + "PublicDescription": "" + }, + {, + "EventCode": "0x3012C", + "EventName": "PM_MRK_ST_FWD", + "BriefDescription": "Marked st forwards", + "PublicDescription": "" + }, + {, + "EventCode": "0x30130", + "EventName": "PM_MRK_INST_FIN", + "BriefDescription": "marked instruction finished", + "PublicDescription": "" + }, + {, + "EventCode": "0x3003A", + "EventName": "PM_CMPLU_STALL_EXCEPTION", + "BriefDescription": "Cycles in which the NTC instruction is not allowed to complete because it was interrupted by ANY exception, which has to be serviced before the instruction can complete", + "PublicDescription": "" + }, + {, + "EventCode": "0x3003C", + "EventName": "PM_CMPLU_STALL_NESTED_TEND", + "BriefDescription": "Completion stall because the ISU is updating the TEXASR to keep track of the nested tend and decrement the TEXASR nested level. This is a short delay", + "PublicDescription": "" + }, + {, + "EventCode": "0x3013E", + "EventName": "PM_MRK_STALL_CMPLU_CYC", + "BriefDescription": "Number of cycles the marked instruction is experiencing a stall while it is next to complete (NTC)", + "PublicDescription": "" + }, + {, + "EventCode": "0x3C044", + "EventName": "PM_DATA_FROM_L3.1_ECO_SHR", + "BriefDescription": "The processor's data cache was reloaded with Shared (S) data from another core's ECO L3 on the same chip due to a demand load", + "PublicDescription": "" + }, + {, + "EventCode": "0x3C04A", + "EventName": "PM_DATA_FROM_RMEM", + "BriefDescription": "The processor's data cache was reloaded from another chip's memory on the same Node or Group ( Remote) due to a demand load", + "PublicDescription": "" + }, + {, + "EventCode": "0x34040", + "EventName": "PM_INST_FROM_L2_DISP_CONFLICT_LDHITST", + "BriefDescription": "The processor's Instruction cache was reloaded from local core's L2 with load hit store conflict due to an instruction fetch (not prefetch)", + "PublicDescription": "" + }, + {, + "EventCode": "0x34044", + "EventName": "PM_INST_FROM_L3.1_ECO_SHR", + "BriefDescription": "The processor's Instruction cache was reloaded with Shared (S) data from another core's ECO L3 on the same chip due to an instruction fetch (not prefetch)", + "PublicDescription": "" + }, + {, + "EventCode": "0x34048", + "EventName": "PM_INST_FROM_DL2L3_SHR", + "BriefDescription": "The processor's Instruction cache was reloaded with Shared (S) data from another chip's L2 or L3 on a different Node or Group (Distant), as this chip due to an instruction fetch (not prefetch)", + "PublicDescription": "" + }, + {, + "EventCode": "0x3404C", + "EventName": "PM_INST_FROM_DL4", + "BriefDescription": "The processor's Instruction cache was reloaded from another chip's L4 on a different Node or Group (Distant) due to an instruction fetch (not prefetch)", + "PublicDescription": "" + }, + {, + "EventCode": "0x35046", + "EventName": "PM_IPTEG_FROM_L2.1_SHR", + "BriefDescription": "A Page Table Entry was loaded into the TLB with Shared (S) data from another core's L2 on the same chip due to a instruction side request", + "PublicDescription": "" + }, + {, + "EventCode": "0x3504E", + "EventName": "PM_DARQ0_4_6_ENTRIES", + "BriefDescription": "Cycles in which 4, 5, or 6 DARQ entries (out of 12) are in use", + "PublicDescription": "" + }, + {, + "EventCode": "0x3E044", + "EventName": "PM_DPTEG_FROM_L3.1_ECO_SHR", + "BriefDescription": "A Page Table Entry was loaded into the TLB with Shared (S) data from another core's ECO L3 on the same chip due to a data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x3F144", + "EventName": "PM_MRK_DPTEG_FROM_L3.1_ECO_SHR", + "BriefDescription": "A Page Table Entry was loaded into the TLB with Shared (S) data from another core's ECO L3 on the same chip due to a marked data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x30050", + "EventName": "PM_SYS_PUMP_CPRED", + "BriefDescription": "Initial and Final Pump Scope was system pump for all data types excluding data prefetch (demand load,inst prefetch,inst fetch,xlate)", + "PublicDescription": "" + }, + {, + "EventCode": "0x30052", + "EventName": "PM_SYS_PUMP_MPRED", + "BriefDescription": "Final Pump Scope (system) mispredicted. Either the original scope was too small (Chip/Group) or the original scope was System and it should have been smaller. Counts for all data types excluding data prefetch (demand load,inst prefetch,inst fetch,xlate)", + "PublicDescription": "" + }, + {, + "EventCode": "0x3C050", + "EventName": "PM_DATA_SYS_PUMP_CPRED", + "BriefDescription": "Initial and Final Pump Scope was system pump (prediction=correct) for a demand load", + "PublicDescription": "" + }, + {, + "EventCode": "0x3C052", + "EventName": "PM_DATA_SYS_PUMP_MPRED", + "BriefDescription": "Final Pump Scope (system) mispredicted. Either the original scope was too small (Chip/Group) or the original scope was System and it should have been smaller. Counts for a demand load", + "PublicDescription": "" + }, + {, + "EventCode": "0x3D05A", + "EventName": "PM_NTC_ISSUE_HELD_OTHER", + "BriefDescription": "The NTC instruction is being held at dispatch during regular pipeline cycles, or because the VSU is busy with multi-cycle instructions, or because of a write-back collision with VSU", + "PublicDescription": "" + }, + {, + "EventCode": "0x3D05C", + "EventName": "PM_DISP_HELD_HB_FULL", + "BriefDescription": "Dispatch held due to History Buffer full. Could be GPR/VSR/VMR/FPR/CR/XVF; CR; XVF (XER/VSCR/FPSCR)", + "PublicDescription": "" + }, + {, + "EventCode": "0x3E158", + "EventName": "PM_MRK_STCX_FAIL", + "BriefDescription": "marked stcx failed", + "PublicDescription": "" + }, + {, + "EventCode": "0x3F056", + "EventName": "PM_RADIX_PWC_L3_HIT", + "BriefDescription": "A radix translation attempt missed in the TLB but hit on the first, second, and third levels of page walk cache.", + "PublicDescription": "" + }, + {, + "EventCode": "0x3F058", + "EventName": "PM_RADIX_PWC_L1_PDE_FROM_L3", + "BriefDescription": "A Page Directory Entry was reloaded to a level 1 page walk cache from the core's L3 data cache", + "PublicDescription": "" + }, + {, + "EventCode": "0x3F05E", + "EventName": "PM_RADIX_PWC_L3_PTE_FROM_L3", + "BriefDescription": "A Page Table Entry was reloaded to a level 3 page walk cache from the core's L3 data cache. This implies that a level 4 PWC access was not necessary for this translation", + "PublicDescription": "" + }, + {, + "EventCode": "0x30064", + "EventName": "PM_DARQ_STORE_XMIT", + "BriefDescription": "The DARQ attempted to transmit a store into an LSAQ or SRQ entry. Includes rejects. Not qualified by thread, so it includes counts for the whole core", + "PublicDescription": "" + }, + {, + "EventCode": "0x30068", + "EventName": "PM_L1_ICACHE_RELOADED_PREF", + "BriefDescription": "Counts all Icache prefetch reloads ( includes demand turned into prefetch)", + "PublicDescription": "" + }, + {, + "EventCode": "0x301E4", + "EventName": "PM_MRK_BR_MPRED_CMPL", + "BriefDescription": "Marked Branch Mispredicted", + "PublicDescription": "" + }, + {, + "EventCode": "0x300F2", + "EventName": "PM_INST_DISP", + "BriefDescription": "# PPC Dispatched", + "PublicDescription": "" + }, + {, + "EventCode": "0x300F6", + "EventName": "PM_L1_DCACHE_RELOAD_VALID", + "BriefDescription": "DL1 reloaded due to Demand Load", + "PublicDescription": "" + }, + {, + "EventCode": "0x300FE", + "EventName": "PM_DATA_FROM_L3MISS", + "BriefDescription": "Demand LD - L3 Miss (not L2 hit and not L3 hit)", + "PublicDescription": "" + }, + {, + "EventCode": "0x4000A", + "EventName": "PM_ISQ_36_44_ENTRIES", + "BriefDescription": "Cycles in which 36 or more Issue Queue entries are in use. This is a shared event, not per thread. There are 44 issue queue entries across 4 slices in the whole core", + "PublicDescription": "" + }, + {, + "EventCode": "0x4000C", + "EventName": "PM_FREQ_UP", + "BriefDescription": "Power Management: Above Threshold A", + "PublicDescription": "" + }, + {, + "EventCode": "0x40012", + "EventName": "PM_L1_ICACHE_RELOADED_ALL", + "BriefDescription": "Counts all Icache reloads includes demand, prefetch, prefetch turned into demand and demand turned into prefetch", + "PublicDescription": "" + }, + {, + "EventCode": "0x4D01A", + "EventName": "PM_CMPLU_STALL_EIEIO", + "BriefDescription": "Finish stall because the NTF instruction is an EIEIO waiting for response from L2", + "PublicDescription": "" + }, + {, + "EventCode": "0x4E014", + "EventName": "PM_TM_TX_PASS_RUN_INST", + "BriefDescription": "Run instructions spent in successful transactions", + "PublicDescription": "" + }, + {, + "EventCode": "0x4E018", + "EventName": "PM_CMPLU_STALL_NTC_DISP_FIN", + "BriefDescription": "Finish stall because the NTF instruction was one that must finish at dispatch.", + "PublicDescription": "" + }, + {, + "EventCode": "0x4C122", + "EventName": "PM_DARQ1_0_3_ENTRIES", + "BriefDescription": "Cycles in which 3 or fewer DARQ1 entries (out of 12) are in use", + "PublicDescription": "" + }, + {, + "EventCode": "0x40134", + "EventName": "PM_MRK_INST_TIMEO", + "BriefDescription": "marked Instruction finish timeout (instruction lost)", + "PublicDescription": "" + }, + {, + "EventCode": "0x4003C", + "EventName": "PM_DISP_HELD_SYNC_HOLD", + "BriefDescription": "Cycles in which dispatch is held because of a synchronizing instruction in the pipeline", + "PublicDescription": "" + }, + {, + "EventCode": "0x4C04A", + "EventName": "PM_DATA_FROM_OFF_CHIP_CACHE", + "BriefDescription": "The processor's data cache was reloaded either shared or modified data from another core's L2/L3 on a different chip (remote or distant) due to a demand load", + "PublicDescription": "" + }, + {, + "EventCode": "0x4C04E", + "EventName": "PM_DATA_FROM_L3MISS_MOD", + "BriefDescription": "The processor's data cache was reloaded from a location other than the local core's L3 due to a demand load", + "PublicDescription": "" + }, + {, + "EventCode": "0x44040", + "EventName": "PM_INST_FROM_L2_DISP_CONFLICT_OTHER", + "BriefDescription": "The processor's Instruction cache was reloaded from local core's L2 with dispatch conflict due to an instruction fetch (not prefetch)", + "PublicDescription": "" + }, + {, + "EventCode": "0x44048", + "EventName": "PM_INST_FROM_DL2L3_MOD", + "BriefDescription": "The processor's Instruction cache was reloaded with Modified (M) data from another chip's L2 or L3 on a different Node or Group (Distant), as this chip due to an instruction fetch (not prefetch)", + "PublicDescription": "" + }, + {, + "EventCode": "0x4404C", + "EventName": "PM_INST_FROM_DMEM", + "BriefDescription": "The processor's Instruction cache was reloaded from another chip's memory on the same Node or Group (Distant) due to an instruction fetch (not prefetch)", + "PublicDescription": "" + }, + {, + "EventCode": "0x4404E", + "EventName": "PM_INST_FROM_L3MISS_MOD", + "BriefDescription": "The processor's Instruction cache was reloaded from a location other than the local core's L3 due to a instruction fetch", + "PublicDescription": "" + }, + {, + "EventCode": "0x4D142", + "EventName": "PM_MRK_DATA_FROM_L3", + "BriefDescription": "The processor's data cache was reloaded from local core's L3 due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x4D04A", + "EventName": "PM_DARQ0_0_3_ENTRIES", + "BriefDescription": "Cycles in which 3 or less DARQ entries (out of 12) are in use", + "PublicDescription": "" + }, + {, + "EventCode": "0x45042", + "EventName": "PM_IPTEG_FROM_L3", + "BriefDescription": "A Page Table Entry was loaded into the TLB from local core's L3 due to a instruction side request", + "PublicDescription": "" + }, + {, + "EventCode": "0x45046", + "EventName": "PM_IPTEG_FROM_L2.1_MOD", + "BriefDescription": "A Page Table Entry was loaded into the TLB with Modified (M) data from another core's L2 on the same chip due to a instruction side request", + "PublicDescription": "" + }, + {, + "EventCode": "0x4F144", + "EventName": "PM_MRK_DPTEG_FROM_L3.1_ECO_MOD", + "BriefDescription": "A Page Table Entry was loaded into the TLB with Modified (M) data from another core's ECO L3 on the same chip due to a marked data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x4F14E", + "EventName": "PM_MRK_DPTEG_FROM_L3MISS", + "BriefDescription": "A Page Table Entry was loaded into the TLB from a location other than the local core's L3 due to a marked data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x4C050", + "EventName": "PM_DATA_SYS_PUMP_MPRED_RTY", + "BriefDescription": "Final Pump Scope (system) ended up larger than Initial Pump Scope (Chip/Group) for a demand load", + "PublicDescription": "" + }, + {, + "EventCode": "0x4C052", + "EventName": "PM_DATA_PUMP_MPRED", + "BriefDescription": "Pump misprediction. Counts across all types of pumps for a demand load", + "PublicDescription": "" + }, + {, + "EventCode": "0x4405E", + "EventName": "PM_DARQ_STORE_REJECT", + "BriefDescription": "The DARQ attempted to transmit a store into an LSAQ or SRQ entry but It was rejected. Divide by PM_DARQ_STORE_XMIT to get reject ratio", + "PublicDescription": "" + }, + {, + "EventCode": "0x4D058", + "EventName": "PM_VECTOR_FLOP_CMPL", + "BriefDescription": "Vector FP instruction completed", + "PublicDescription": "" + }, + {, + "EventCode": "0x4D05A", + "EventName": "PM_NON_MATH_FLOP_CMPL", + "BriefDescription": "Non FLOP operation completed", + "PublicDescription": "" + }, + {, + "EventCode": "0x4505A", + "EventName": "PM_SP_FLOP_CMPL", + "BriefDescription": "SP instruction completed", + "PublicDescription": "" + }, + {, + "EventCode": "0x4F056", + "EventName": "PM_RADIX_PWC_L1_PDE_FROM_L3MISS", + "BriefDescription": "A Page Directory Entry was reloaded to a level 1 page walk cache from beyond the core's L3 data cache. The source could be local/remote/distant memory or another core's cache", + "PublicDescription": "" + }, + {, + "EventCode": "0x4F058", + "EventName": "PM_RADIX_PWC_L2_PTE_FROM_L3", + "BriefDescription": "A Page Table Entry was reloaded to a level 2 page walk cache from the core's L3 data cache. This implies that level 3 and level 4 PWC accesses were not necessary for this translation", + "PublicDescription": "" + }, + {, + "EventCode": "0x4F05A", + "EventName": "PM_RADIX_PWC_L4_PTE_FROM_L3", + "BriefDescription": "A Page Table Entry was reloaded to a level 4 page walk cache from the core's L3 data cache. This is the deepest level of PWC possible for a translation", + "PublicDescription": "" + }, + {, + "EventCode": "0x4F05E", + "EventName": "PM_RADIX_PWC_L3_PTE_FROM_L3MISS", + "BriefDescription": "A Page Table Entry was reloaded to a level 3 page walk cache from beyond the core's L3 data cache. This implies that a level 4 PWC access was not necessary for this translation. The source could be local/remote/distant memory or another core's cache", + "PublicDescription": "" + }, + {, + "EventCode": "0x401E0", + "EventName": "PM_MRK_INST_CMPL", + "BriefDescription": "marked instruction completed", + "PublicDescription": "" + }, + {, + "EventCode": "0x400F4", + "EventName": "PM_RUN_PURR", + "BriefDescription": "Run_PURR", + "PublicDescription": "" + }, + {, + "EventCode": "0x400FC", + "EventName": "PM_ITLB_MISS", + "BriefDescription": "ITLB Reloaded. Counts 1 per ITLB miss for HPT but multiple for radix depending on number of levels traveresed", + "PublicDescription": "" + }, + {, + "EventCode": "0x400FE", + "EventName": "PM_DATA_FROM_MEMORY", + "BriefDescription": "The processor's data cache was reloaded from a memory location including L4 from local remote or distant due to a demand load", + "PublicDescription": "" + } +] diff --git a/tools/perf/pmu-events/arch/powerpc/power9/memory.json b/tools/perf/pmu-events/arch/powerpc/power9/memory.json new file mode 100644 index 0000000..e48708c --- /dev/null +++ b/tools/perf/pmu-events/arch/powerpc/power9/memory.json @@ -0,0 +1,158 @@ +[ + {, + "EventCode": "0x10008", + "EventName": "PM_RUN_SPURR", + "BriefDescription": "Run SPURR", + "PublicDescription": "" + }, + {, + "EventCode": "0x1000A", + "EventName": "PM_PMC3_REWIND", + "BriefDescription": "PMC3 rewind event. A rewind happens when a speculative event (such as latency or CPI stack) is selected on PMC3 and the stall reason or reload source did not match the one programmed in PMC3. When this occurs, the count in PMC3 will not change.", + "PublicDescription": "" + }, + {, + "EventCode": "0x1C040", + "EventName": "PM_DATA_FROM_L2_NO_CONFLICT", + "BriefDescription": "The processor's data cache was reloaded from local core's L2 without conflict due to a demand load", + "PublicDescription": "" + }, + {, + "EventCode": "0x1C050", + "EventName": "PM_DATA_CHIP_PUMP_CPRED", + "BriefDescription": "Initial and Final Pump Scope was chip pump (prediction=correct) for a demand load", + "PublicDescription": "" + }, + {, + "EventCode": "0x1D15E", + "EventName": "PM_MRK_RUN_CYC", + "BriefDescription": "Run cycles in which a marked instruction is in the pipeline", + "PublicDescription": "" + }, + {, + "EventCode": "0x15158", + "EventName": "PM_SYNC_MRK_L2HIT", + "BriefDescription": "Marked L2 Hits that can throw a synchronous interrupt", + "PublicDescription": "" + }, + {, + "EventCode": "0x20010", + "EventName": "PM_PMC1_OVERFLOW", + "BriefDescription": "Overflow from counter 1", + "PublicDescription": "" + }, + {, + "EventCode": "0x2C040", + "EventName": "PM_DATA_FROM_L2_MEPF", + "BriefDescription": "The processor's data cache was reloaded from local core's L2 hit without dispatch conflicts on Mepf state due to a demand load", + "PublicDescription": "" + }, + {, + "EventCode": "0x2005A", + "EventName": "PM_DARQ1_7_9_ENTRIES", + "BriefDescription": "Cycles in which 7 to 9 DARQ1 entries (out of 12) are in use", + "PublicDescription": "" + }, + {, + "EventCode": "0x2C05C", + "EventName": "PM_INST_GRP_PUMP_CPRED", + "BriefDescription": "Initial and Final Pump Scope was group pump (prediction=correct) for an instruction fetch (demand only)", + "PublicDescription": "" + }, + {, + "EventCode": "0x2D156", + "EventName": "PM_MRK_DTLB_MISS_4K", + "BriefDescription": "Marked Data TLB Miss page size 4k", + "PublicDescription": "" + }, + {, + "EventCode": "0x2E05A", + "EventName": "PM_LRQ_REJECT", + "BriefDescription": "Internal LSU reject from LRQ. Rejects cause the load to go back to LRQ, but it stays contained within the LSU once it gets issued. This event counts the number of times the LRQ attempts to relaunch an instruction after a reject. Any load can suffer multiple rejects", + "PublicDescription": "" + }, + {, + "EventCode": "0x2E05C", + "EventName": "PM_LSU_REJECT_ERAT_MISS", + "BriefDescription": "LSU Reject due to ERAT (up to 4 per cycles)", + "PublicDescription": "" + }, + {, + "EventCode": "0x200F6", + "EventName": "PM_LSU_DERAT_MISS", + "BriefDescription": "DERAT Reloaded due to a DERAT miss", + "PublicDescription": "" + }, + {, + "EventCode": "0x3C048", + "EventName": "PM_DATA_FROM_DL2L3_SHR", + "BriefDescription": "The processor's data cache was reloaded with Shared (S) data from another chip's L2 or L3 on a different Node or Group (Distant), as this chip due to a demand load", + "PublicDescription": "" + }, + {, + "EventCode": "0x3404A", + "EventName": "PM_INST_FROM_RMEM", + "BriefDescription": "The processor's Instruction cache was reloaded from another chip's memory on the same Node or Group ( Remote) due to an instruction fetch (not prefetch)", + "PublicDescription": "" + }, + {, + "EventCode": "0x3C058", + "EventName": "PM_LARX_FIN", + "BriefDescription": "Larx finished", + "PublicDescription": "" + }, + {, + "EventCode": "0x3E050", + "EventName": "PM_DARQ1_4_6_ENTRIES", + "BriefDescription": "Cycles in which 4, 5, or 6 DARQ1 entries (out of 12) are in use", + "PublicDescription": "" + }, + {, + "EventCode": "0x3006E", + "EventName": "PM_NEST_REF_CLK", + "BriefDescription": "Multiply by 4 to obtain the number of PB cycles", + "PublicDescription": "" + }, + {, + "EventCode": "0x301E2", + "EventName": "PM_MRK_ST_CMPL", + "BriefDescription": "Marked store completed and sent to nest", + "PublicDescription": "" + }, + {, + "EventCode": "0x4D02C", + "EventName": "PM_PMC1_REWIND", + "BriefDescription": "", + "PublicDescription": "" + }, + {, + "EventCode": "0x4003E", + "EventName": "PM_LD_CMPL", + "BriefDescription": "count of Loads completed", + "PublicDescription": "" + }, + {, + "EventCode": "0x4C040", + "EventName": "PM_DATA_FROM_L2_DISP_CONFLICT_OTHER", + "BriefDescription": "The processor's data cache was reloaded from local core's L2 with dispatch conflict due to a demand load", + "PublicDescription": "" + }, + {, + "EventCode": "0x4C042", + "EventName": "PM_DATA_FROM_L3", + "BriefDescription": "The processor's data cache was reloaded from local core's L3 due to a demand load", + "PublicDescription": "" + }, + {, + "EventCode": "0x4C048", + "EventName": "PM_DATA_FROM_DL2L3_MOD", + "BriefDescription": "The processor's data cache was reloaded with Modified (M) data from another chip's L2 or L3 on a different Node or Group (Distant), as this chip due to a demand load", + "PublicDescription": "" + }, + {, + "EventCode": "0x4D056", + "EventName": "PM_NON_FMA_FLOP_CMPL", + "BriefDescription": "Non FMA instruction completed", + "PublicDescription": "" + } +] \ No newline at end of file diff --git a/tools/perf/pmu-events/arch/powerpc/power9/other.json b/tools/perf/pmu-events/arch/powerpc/power9/other.json new file mode 100644 index 0000000..396e6e0 --- /dev/null +++ b/tools/perf/pmu-events/arch/powerpc/power9/other.json @@ -0,0 +1,836 @@ +[ + {, + "EventCode": "0x1001C", + "EventName": "PM_CMPLU_STALL_THRD", + "BriefDescription": "Completion Stalled because the thread was blocked", + "PublicDescription": "" + }, + {, + "EventCode": "0x1002E", + "EventName": "PM_LMQ_MERGE", + "BriefDescription": "A demand miss collides with a prefetch for the same line", + "PublicDescription": "" + }, + {, + "EventCode": "0x10134", + "EventName": "PM_MRK_ST_DONE_L2", + "BriefDescription": "marked store completed in L2 ( RC machine done)", + "PublicDescription": "" + }, + {, + "EventCode": "0x10138", + "EventName": "PM_MRK_BR_2PATH", + "BriefDescription": "marked branches which are not strongly biased", + "PublicDescription": "" + }, + {, + "EventCode": "0x1C04A", + "EventName": "PM_DATA_FROM_RL2L3_SHR", + "BriefDescription": "The processor's data cache was reloaded with Shared (S) data from another chip's L2 or L3 on the same Node or Group (Remote), as this chip due to a demand load", + "PublicDescription": "" + }, + {, + "EventCode": "0x1C04C", + "EventName": "PM_DATA_FROM_LL4", + "BriefDescription": "The processor's data cache was reloaded from the local chip's L4 cache due to a demand load", + "PublicDescription": "" + }, + {, + "EventCode": "0x1D140", + "EventName": "PM_MRK_DATA_FROM_L3.1_MOD_CYC", + "BriefDescription": "Duration in cycles to reload with Modified (M) data from another core's L3 on the same chip due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x1D144", + "EventName": "PM_MRK_DATA_FROM_L3_DISP_CONFLICT", + "BriefDescription": "The processor's data cache was reloaded from local core's L3 with dispatch conflict due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x1D146", + "EventName": "PM_MRK_DATA_FROM_MEMORY_CYC", + "BriefDescription": "Duration in cycles to reload from a memory location including L4 from local remote or distant due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x1D148", + "EventName": "PM_MRK_DATA_FROM_RMEM", + "BriefDescription": "The processor's data cache was reloaded from another chip's memory on the same Node or Group ( Remote) due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x1D14E", + "EventName": "PM_MRK_DATA_FROM_OFF_CHIP_CACHE_CYC", + "BriefDescription": "Duration in cycles to reload either shared or modified data from another core's L2/L3 on a different chip (remote or distant) due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x15040", + "EventName": "PM_IPTEG_FROM_L2_NO_CONFLICT", + "BriefDescription": "A Page Table Entry was loaded into the TLB from local core's L2 without conflict due to a instruction side request", + "PublicDescription": "" + }, + {, + "EventCode": "0x1504C", + "EventName": "PM_IPTEG_FROM_LL4", + "BriefDescription": "A Page Table Entry was loaded into the TLB from the local chip's L4 cache due to a instruction side request", + "PublicDescription": "" + }, + {, + "EventCode": "0x1E048", + "EventName": "PM_DPTEG_FROM_ON_CHIP_CACHE", + "BriefDescription": "A Page Table Entry was loaded into the TLB either shared or modified data from another core's L2/L3 on the same chip due to a data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x1E04E", + "EventName": "PM_DPTEG_FROM_L2MISS", + "BriefDescription": "A Page Table Entry was loaded into the TLB from a location other than the local core's L2 due to a data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x1F146", + "EventName": "PM_MRK_DPTEG_FROM_L3.1_SHR", + "BriefDescription": "A Page Table Entry was loaded into the TLB with Shared (S) data from another core's L3 on the same chip due to a marked data side request.. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x10052", + "EventName": "PM_GRP_PUMP_MPRED_RTY", + "BriefDescription": "Final Pump Scope (Group) ended up larger than Initial Pump Scope (Chip) for all data types excluding data prefetch (demand load,inst prefetch,inst fetch,xlate)", + "PublicDescription": "" + }, + {, + "EventCode": "0x1C05C", + "EventName": "PM_DTLB_MISS_2M", + "BriefDescription": "Data TLB reload (after a miss) page size 2M. Implies radix translation was used", + "PublicDescription": "" + }, + {, + "EventCode": "0x14156", + "EventName": "PM_MRK_DATA_FROM_L2_CYC", + "BriefDescription": "Duration in cycles to reload from local core's L2 due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x14158", + "EventName": "PM_MRK_DATA_FROM_L2_NO_CONFLICT_CYC", + "BriefDescription": "Duration in cycles to reload from local core's L2 without conflict due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x1415C", + "EventName": "PM_MRK_DATA_FROM_L3_MEPF_CYC", + "BriefDescription": "Duration in cycles to reload from local core's L3 without dispatch conflicts hit on Mepf state due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x1D150", + "EventName": "PM_MRK_DATA_FROM_DL2L3_SHR", + "BriefDescription": "The processor's data cache was reloaded with Shared (S) data from another chip's L2 or L3 on a different Node or Group (Distant), as this chip due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x1D152", + "EventName": "PM_MRK_DATA_FROM_DL4", + "BriefDescription": "The processor's data cache was reloaded from another chip's L4 on a different Node or Group (Distant) due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x1D156", + "EventName": "PM_MRK_LD_MISS_L1_CYC", + "BriefDescription": "Marked ld latency", + "PublicDescription": "" + }, + {, + "EventCode": "0x15154", + "EventName": "PM_SYNC_MRK_L3MISS", + "BriefDescription": "Marked L3 misses that can throw a synchronous interrupt", + "PublicDescription": "" + }, + {, + "EventCode": "0x1515A", + "EventName": "PM_SYNC_MRK_L2MISS", + "BriefDescription": "Marked L2 Miss that can throw a synchronous interrupt", + "PublicDescription": "" + }, + {, + "EventCode": "0x1E05A", + "EventName": "PM_CMPLU_STALL_ANY_SYNC", + "BriefDescription": "Cycles in which the NTC sync instruction (isync, lwsync or hwsync) is not allowed to complete ", + "PublicDescription": "" + }, + {, + "EventCode": "0x1E05C", + "EventName": "PM_CMPLU_STALL_NESTED_TBEGIN", + "BriefDescription": "Completion stall because the ISU is updating the TEXASR to keep track of the nested tbegin. This is a short delay, and it includes ROT", + "PublicDescription": "" + }, + {, + "EventCode": "0x1F152", + "EventName": "PM_MRK_FAB_RSP_BKILL_CYC", + "BriefDescription": "cycles L2 RC took for a bkill", + "PublicDescription": "" + }, + {, + "EventCode": "0x1F056", + "EventName": "PM_RADIX_PWC_L1_HIT", + "BriefDescription": "A radix translation attempt missed in the TLB and only the first level page walk cache was a hit.", + "PublicDescription": "" + }, + {, + "EventCode": "0x101E4", + "EventName": "PM_MRK_L1_ICACHE_MISS", + "BriefDescription": "sampled Instruction suffered an icache Miss", + "PublicDescription": "" + }, + {, + "EventCode": "0x101EA", + "EventName": "PM_MRK_L1_RELOAD_VALID", + "BriefDescription": "Marked demand reload", + "PublicDescription": "" + }, + {, + "EventCode": "0x100FA", + "EventName": "PM_ANY_THRD_RUN_CYC", + "BriefDescription": "Cycles in which at least one thread has the run latch set", + "PublicDescription": "" + }, + {, + "EventCode": "0x100FC", + "EventName": "PM_LD_REF_L1", + "BriefDescription": "All L1 D cache load references counted at finish, gated by reject", + "PublicDescription": "" + }, + {, + "EventCode": "0x20006", + "EventName": "PM_DISP_HELD_ISSQ_FULL", + "BriefDescription": "Dispatch held due to Issue q full. Includes issue queue and branch queue", + "PublicDescription": "" + }, + {, + "EventCode": "0x2000C", + "EventName": "PM_THRD_ALL_RUN_CYC", + "BriefDescription": "Cycles in which all the threads have the run latch set", + "PublicDescription": "" + }, + {, + "EventCode": "0x2001A", + "EventName": "PM_NTC_ALL_FIN", + "BriefDescription": "Cycles after all instructions have finished to group completed", + "PublicDescription": "" + }, + {, + "EventCode": "0x2D014", + "EventName": "PM_CMPLU_STALL_LRQ_FULL", + "BriefDescription": "Finish stall because the NTF instruction was a load that was held in LSAQ (load-store address queue) because the LRQ (load-reorder queue) was full", + "PublicDescription": "" + }, + {, + "EventCode": "0x2D018", + "EventName": "PM_CMPLU_STALL_EXEC_UNIT", + "BriefDescription": "Completion stall due to execution units (FXU/VSU/CRU)", + "PublicDescription": "" + }, + {, + "EventCode": "0x2D01E", + "EventName": "PM_ICT_NOSLOT_DISP_HELD_ISSQ", + "BriefDescription": "Ict empty for this thread due to dispatch hold on this thread due to Issue q full, BRQ full, XVCF Full, Count cache, Link, Tar full", + "PublicDescription": "" + }, + {, + "EventCode": "0x2E014", + "EventName": "PM_STCX_FIN", + "BriefDescription": "Number of stcx instructions finished. This includes instructions in the speculative path of a branch that may be flushed", + "PublicDescription": "" + }, + {, + "EventCode": "0x2C120", + "EventName": "PM_MRK_DATA_FROM_L2_NO_CONFLICT", + "BriefDescription": "The processor's data cache was reloaded from local core's L2 without conflict due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x2C122", + "EventName": "PM_MRK_DATA_FROM_L3_DISP_CONFLICT_CYC", + "BriefDescription": "Duration in cycles to reload from local core's L3 with dispatch conflict due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x2C126", + "EventName": "PM_MRK_DATA_FROM_L2", + "BriefDescription": "The processor's data cache was reloaded from local core's L2 due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x2C12A", + "EventName": "PM_MRK_DATA_FROM_RMEM_CYC", + "BriefDescription": "Duration in cycles to reload from another chip's memory on the same Node or Group ( Remote) due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x2C12C", + "EventName": "PM_MRK_DATA_FROM_DL4_CYC", + "BriefDescription": "Duration in cycles to reload from another chip's L4 on a different Node or Group (Distant) due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x2D120", + "EventName": "PM_MRK_DATA_FROM_OFF_CHIP_CACHE", + "BriefDescription": "The processor's data cache was reloaded either shared or modified data from another core's L2/L3 on a different chip (remote or distant) due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x2D026", + "EventName": "PM_RADIX_PWC_L1_PDE_FROM_L2", + "BriefDescription": "A Page Directory Entry was reloaded to a level 1 page walk cache from the core's L2 data cache", + "PublicDescription": "" + }, + {, + "EventCode": "0x20132", + "EventName": "PM_MRK_DFU_FIN", + "BriefDescription": "Decimal Unit marked Instruction Finish", + "PublicDescription": "" + }, + {, + "EventCode": "0x20134", + "EventName": "PM_MRK_FXU_FIN", + "BriefDescription": "fxu marked instr finish", + "PublicDescription": "" + }, + {, + "EventCode": "0x2C04E", + "EventName": "PM_LD_MISS_L1_FIN", + "BriefDescription": "Number of load instructions that finished with an L1 miss. Note that even if a load spans multiple slices this event will increment only once per load op.", + "PublicDescription": "" + }, + {, + "EventCode": "0x24040", + "EventName": "PM_INST_FROM_L2_MEPF", + "BriefDescription": "The processor's Instruction cache was reloaded from local core's L2 hit without dispatch conflicts on Mepf state. due to an instruction fetch (not prefetch)", + "PublicDescription": "" + }, + {, + "EventCode": "0x24048", + "EventName": "PM_INST_FROM_LMEM", + "BriefDescription": "The processor's Instruction cache was reloaded from the local chip's Memory due to an instruction fetch (not prefetch)", + "PublicDescription": "" + }, + {, + "EventCode": "0x2D142", + "EventName": "PM_MRK_DATA_FROM_L3_MEPF", + "BriefDescription": "The processor's data cache was reloaded from local core's L3 without dispatch conflicts hit on Mepf state. due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x2D144", + "EventName": "PM_MRK_DATA_FROM_L3.1_MOD", + "BriefDescription": "The processor's data cache was reloaded with Modified (M) data from another core's L3 on the same chip due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x2D148", + "EventName": "PM_MRK_DATA_FROM_L2_DISP_CONFLICT_LDHITST", + "BriefDescription": "The processor's data cache was reloaded from local core's L2 with load hit store conflict due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x25048", + "EventName": "PM_IPTEG_FROM_LMEM", + "BriefDescription": "A Page Table Entry was loaded into the TLB from the local chip's Memory due to a instruction side request", + "PublicDescription": "" + }, + {, + "EventCode": "0x2E040", + "EventName": "PM_DPTEG_FROM_L2_MEPF", + "BriefDescription": "A Page Table Entry was loaded into the TLB from local core's L2 hit without dispatch conflicts on Mepf state. due to a data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x2E04A", + "EventName": "PM_DPTEG_FROM_RL4", + "BriefDescription": "A Page Table Entry was loaded into the TLB from another chip's L4 on the same Node or Group ( Remote) due to a data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x2F14A", + "EventName": "PM_MRK_DPTEG_FROM_RL4", + "BriefDescription": "A Page Table Entry was loaded into the TLB from another chip's L4 on the same Node or Group ( Remote) due to a marked data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x20054", + "EventName": "PM_L1_PREF", + "BriefDescription": "A data line was written to the L1 due to a hardware or software prefetch", + "PublicDescription": "" + }, + {, + "EventCode": "0x20056", + "EventName": "PM_TAKEN_BR_MPRED_CMPL", + "BriefDescription": "Total number of taken branches that were incorrectly predicted as not-taken. This event counts branches completed and does not include speculative instructions", + "PublicDescription": "" + }, + {, + "EventCode": "0x20058", + "EventName": "PM_DARQ1_10_12_ENTRIES", + "BriefDescription": "Cycles in which 10 or more DARQ1 entries (out of 12) are in use", + "PublicDescription": "" + }, + {, + "EventCode": "0x2C050", + "EventName": "PM_DATA_GRP_PUMP_CPRED", + "BriefDescription": "Initial and Final Pump Scope was group pump (prediction=correct) for a demand load", + "PublicDescription": "" + }, + {, + "EventCode": "0x2C05E", + "EventName": "PM_INST_GRP_PUMP_MPRED", + "BriefDescription": "Final Pump Scope (Group) ended up either larger or smaller than Initial Pump Scope for an instruction fetch (demand only)", + "PublicDescription": "" + }, + {, + "EventCode": "0x2505C", + "EventName": "PM_VSU_FIN", + "BriefDescription": "VSU instruction finished. Up to 4 per cycle", + "PublicDescription": "" + }, + {, + "EventCode": "0x2505E", + "EventName": "PM_BACK_BR_CMPL", + "BriefDescription": "Branch instruction completed with a target address less than current instruction address", + "PublicDescription": "" + }, + {, + "EventCode": "0x2E052", + "EventName": "PM_TM_PASSED", + "BriefDescription": "Number of TM transactions that passed", + "PublicDescription": "" + }, + {, + "EventCode": "0x20064", + "EventName": "PM_IERAT_RELOAD_4K", + "BriefDescription": "IERAT reloaded (after a miss) for 4K pages", + "PublicDescription": "" + }, + {, + "EventCode": "0x2006C", + "EventName": "PM_RUN_CYC_SMT4_MODE", + "BriefDescription": "Cycles in which this thread's run latch is set and the core is in SMT4 mode", + "PublicDescription": "" + }, + {, + "EventCode": "0x201E0", + "EventName": "PM_MRK_DATA_FROM_MEMORY", + "BriefDescription": "The processor's data cache was reloaded from a memory location including L4 from local remote or distant due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x201E4", + "EventName": "PM_MRK_DATA_FROM_L3MISS", + "BriefDescription": "The processor's data cache was reloaded from a location other than the local core's L3 due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x201E8", + "EventName": "PM_THRESH_EXC_512", + "BriefDescription": "Threshold counter exceeded a value of 512", + "PublicDescription": "" + }, + {, + "EventCode": "0x200F2", + "EventName": "PM_INST_DISP", + "BriefDescription": "# PPC Dispatched", + "PublicDescription": "" + }, + {, + "EventCode": "0x30016", + "EventName": "PM_CMPLU_STALL_SRQ_FULL", + "BriefDescription": "Finish stall because the NTF instruction was a store that was held in LSAQ because the SRQ was full", + "PublicDescription": "" + }, + {, + "EventCode": "0x30018", + "EventName": "PM_ICT_NOSLOT_DISP_HELD_HB_FULL", + "BriefDescription": "Ict empty for this thread due to dispatch holds because the History Buffer was full. Could be GPR/VSR/VMR/FPR/CR/XVF; CR; XVF (XER/VSCR/FPSCR)", + "PublicDescription": "" + }, + {, + "EventCode": "0x3001A", + "EventName": "PM_DATA_TABLEWALK_CYC", + "BriefDescription": "Data Tablewalk Cycles. Could be 1 or 2 active tablewalks. Includes data prefetches.", + "PublicDescription": "" + }, + {, + "EventCode": "0x30132", + "EventName": "PM_MRK_VSU_FIN", + "BriefDescription": "VSU marked instr finish", + "PublicDescription": "" + }, + {, + "EventCode": "0x30134", + "EventName": "PM_MRK_ST_CMPL_INT", + "BriefDescription": "marked store finished with intervention", + "PublicDescription": "" + }, + {, + "EventCode": "0x30038", + "EventName": "PM_CMPLU_STALL_DMISS_LMEM", + "BriefDescription": "Completion stall due to cache miss that resolves in local memory", + "PublicDescription": "" + }, + {, + "EventCode": "0x3C040", + "EventName": "PM_DATA_FROM_L2_DISP_CONFLICT_LDHITST", + "BriefDescription": "The processor's data cache was reloaded from local core's L2 with load hit store conflict due to a demand load", + "PublicDescription": "" + }, + {, + "EventCode": "0x3C042", + "EventName": "PM_DATA_FROM_L3_DISP_CONFLICT", + "BriefDescription": "The processor's data cache was reloaded from local core's L3 with dispatch conflict due to a demand load", + "PublicDescription": "" + }, + {, + "EventCode": "0x3D140", + "EventName": "PM_MRK_DATA_FROM_L2_DISP_CONFLICT_OTHER_CYC", + "BriefDescription": "Duration in cycles to reload from local core's L2 with dispatch conflict due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x3D144", + "EventName": "PM_MRK_DATA_FROM_L2_MEPF_CYC", + "BriefDescription": "Duration in cycles to reload from local core's L2 hit without dispatch conflicts on Mepf state. due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x3D146", + "EventName": "PM_MRK_DATA_FROM_L3_NO_CONFLICT", + "BriefDescription": "The processor's data cache was reloaded from local core's L3 without conflict due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x3D14C", + "EventName": "PM_MRK_DATA_FROM_DMEM", + "BriefDescription": "The processor's data cache was reloaded from another chip's memory on the same Node or Group (Distant) due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x3D14E", + "EventName": "PM_MRK_DATA_FROM_DL2L3_MOD", + "BriefDescription": "The processor's data cache was reloaded with Modified (M) data from another chip's L2 or L3 on a different Node or Group (Distant), as this chip due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x35042", + "EventName": "PM_IPTEG_FROM_L3_DISP_CONFLICT", + "BriefDescription": "A Page Table Entry was loaded into the TLB from local core's L3 with dispatch conflict due to a instruction side request", + "PublicDescription": "" + }, + {, + "EventCode": "0x35048", + "EventName": "PM_IPTEG_FROM_DL2L3_SHR", + "BriefDescription": "A Page Table Entry was loaded into the TLB with Shared (S) data from another chip's L2 or L3 on a different Node or Group (Distant), as this chip due to a instruction side request", + "PublicDescription": "" + }, + {, + "EventCode": "0x3504C", + "EventName": "PM_IPTEG_FROM_DL4", + "BriefDescription": "A Page Table Entry was loaded into the TLB from another chip's L4 on a different Node or Group (Distant) due to a instruction side request", + "PublicDescription": "" + }, + {, + "EventCode": "0x3F146", + "EventName": "PM_MRK_DPTEG_FROM_L2.1_SHR", + "BriefDescription": "A Page Table Entry was loaded into the TLB with Shared (S) data from another core's L2 on the same chip due to a marked data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x3005A", + "EventName": "PM_ISQ_0_8_ENTRIES", + "BriefDescription": "Cycles in which 8 or less Issue Queue entries are in use. This is a shared event, not per thread", + "PublicDescription": "" + }, + {, + "EventCode": "0x3005C", + "EventName": "PM_BFU_BUSY", + "BriefDescription": "Cycles in which all 4 Binary Floating Point units are busy. The BFU is running at capacity", + "PublicDescription": "" + }, + {, + "EventCode": "0x3C05E", + "EventName": "PM_MEM_RWITM", + "BriefDescription": "Memory Read With Intent to Modify for this thread", + "PublicDescription": "" + }, + {, + "EventCode": "0x34054", + "EventName": "PM_PARTIAL_ST_FIN", + "BriefDescription": "Any store finished by an LSU slice", + "PublicDescription": "" + }, + {, + "EventCode": "0x3D15E", + "EventName": "PM_MULT_MRK", + "BriefDescription": "mult marked instr", + "PublicDescription": "" + }, + {, + "EventCode": "0x35152", + "EventName": "PM_MRK_DATA_FROM_L2MISS_CYC", + "BriefDescription": "Duration in cycles to reload from a location other than the local core's L2 due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x35154", + "EventName": "PM_MRK_DATA_FROM_L3_CYC", + "BriefDescription": "Duration in cycles to reload from local core's L3 due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x35156", + "EventName": "PM_MRK_DATA_FROM_L3.1_SHR_CYC", + "BriefDescription": "Duration in cycles to reload with Shared (S) data from another core's L3 on the same chip due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x35158", + "EventName": "PM_MRK_DATA_FROM_L3.1_ECO_MOD_CYC", + "BriefDescription": "Duration in cycles to reload with Modified (M) data from another core's ECO L3 on the same chip due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x3515E", + "EventName": "PM_MRK_BACK_BR_CMPL", + "BriefDescription": "Marked branch instruction completed with a target address less than current instruction address", + "PublicDescription": "" + }, + {, + "EventCode": "0x3E05E", + "EventName": "PM_L3_CO_MEPF", + "BriefDescription": "L3 castouts in Mepf state for this thread", + "PublicDescription": "" + }, + {, + "EventCode": "0x3F150", + "EventName": "PM_MRK_ST_DRAIN_TO_L2DISP_CYC", + "BriefDescription": "cycles to drain st from core to L2", + "PublicDescription": "" + }, + {, + "EventCode": "0x3F054", + "EventName": "PM_RADIX_PWC_L4_PTE_FROM_L3MISS", + "BriefDescription": "A Page Table Entry was reloaded to a level 4 page walk cache from beyond the core's L3 data cache. This is the deepest level of PWC possible for a translation. The source could be local/remote/distant memory or another core's cache", + "PublicDescription": "" + }, + {, + "EventCode": "0x30162", + "EventName": "PM_MRK_LSU_DERAT_MISS", + "BriefDescription": "Marked derat reload (miss) for any page size", + "PublicDescription": "" + }, + {, + "EventCode": "0x3006A", + "EventName": "PM_IERAT_RELOAD_64K", + "BriefDescription": "IERAT Reloaded (Miss) for a 64k page", + "PublicDescription": "" + }, + {, + "EventCode": "0x300F8", + "EventName": "PM_TB_BIT_TRANS", + "BriefDescription": "timebase event", + "PublicDescription": "" + }, + {, + "EventCode": "0x40006", + "EventName": "PM_ISLB_MISS", + "BriefDescription": "Number of ISLB misses for this thread", + "PublicDescription": "" + }, + {, + "EventCode": "0x40008", + "EventName": "PM_SRQ_EMPTY_CYC", + "BriefDescription": "Cycles in which the SRQ has at least one (out of four) empty slice", + "PublicDescription": "" + }, + {, + "EventCode": "0x40014", + "EventName": "PM_PROBE_NOP_DISP", + "BriefDescription": "ProbeNops dispatched", + "PublicDescription": "" + }, + {, + "EventCode": "0x4001C", + "EventName": "PM_INST_IMC_MATCH_CMPL", + "BriefDescription": "IMC Match Count", + "PublicDescription": "" + }, + {, + "EventCode": "0x4C01A", + "EventName": "PM_CMPLU_STALL_DMISS_L3MISS", + "BriefDescription": "Completion stall due to cache miss resolving missed the L3", + "PublicDescription": "" + }, + {, + "EventCode": "0x4D012", + "EventName": "PM_PMC3_SAVED", + "BriefDescription": "PMC3 Rewind Value saved", + "PublicDescription": "" + }, + {, + "EventCode": "0x4E11E", + "EventName": "PM_MRK_DATA_FROM_DMEM_CYC", + "BriefDescription": "Duration in cycles to reload from another chip's memory on the same Node or Group (Distant) due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x4C124", + "EventName": "PM_MRK_DATA_FROM_L3_NO_CONFLICT_CYC", + "BriefDescription": "Duration in cycles to reload from local core's L3 without conflict due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x4D12E", + "EventName": "PM_MRK_DATA_FROM_DL2L3_MOD_CYC", + "BriefDescription": "Duration in cycles to reload with Modified (M) data from another chip's L2 or L3 on a different Node or Group (Distant), as this chip due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x4013A", + "EventName": "PM_MRK_IC_MISS", + "BriefDescription": "Marked instruction experienced I cache miss", + "PublicDescription": "" + }, + {, + "EventCode": "0x44044", + "EventName": "PM_INST_FROM_L3.1_ECO_MOD", + "BriefDescription": "The processor's Instruction cache was reloaded with Modified (M) data from another core's ECO L3 on the same chip due to an instruction fetch (not prefetch)", + "PublicDescription": "" + }, + {, + "EventCode": "0x44046", + "EventName": "PM_INST_FROM_L2.1_MOD", + "BriefDescription": "The processor's Instruction cache was reloaded with Modified (M) data from another core's L2 on the same chip due to an instruction fetch (not prefetch)", + "PublicDescription": "" + }, + {, + "EventCode": "0x4404A", + "EventName": "PM_INST_FROM_OFF_CHIP_CACHE", + "BriefDescription": "The processor's Instruction cache was reloaded either shared or modified data from another core's L2/L3 on a different chip (remote or distant) due to an instruction fetch (not prefetch)", + "PublicDescription": "" + }, + {, + "EventCode": "0x4D144", + "EventName": "PM_MRK_DATA_FROM_L3.1_ECO_MOD", + "BriefDescription": "The processor's data cache was reloaded with Modified (M) data from another core's ECO L3 on the same chip due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x4D146", + "EventName": "PM_MRK_DATA_FROM_L2.1_MOD", + "BriefDescription": "The processor's data cache was reloaded with Modified (M) data from another core's L2 on the same chip due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x4504C", + "EventName": "PM_IPTEG_FROM_DMEM", + "BriefDescription": "A Page Table Entry was loaded into the TLB from another chip's memory on the same Node or Group (Distant) due to a instruction side request", + "PublicDescription": "" + }, + {, + "EventCode": "0x4E044", + "EventName": "PM_DPTEG_FROM_L3.1_ECO_MOD", + "BriefDescription": "A Page Table Entry was loaded into the TLB with Modified (M) data from another core's ECO L3 on the same chip due to a data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x4E04A", + "EventName": "PM_DPTEG_FROM_OFF_CHIP_CACHE", + "BriefDescription": "A Page Table Entry was loaded into the TLB either shared or modified data from another core's L2/L3 on a different chip (remote or distant) due to a data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x40154", + "EventName": "PM_MRK_FAB_RSP_BKILL", + "BriefDescription": "Marked store had to do a bkill", + "PublicDescription": "" + }, + {, + "EventCode": "0x4C054", + "EventName": "PM_DERAT_MISS_16G", + "BriefDescription": "Data ERAT Miss (Data TLB Access) page size 16G", + "PublicDescription": "" + }, + {, + "EventCode": "0x4C05A", + "EventName": "PM_DTLB_MISS_1G", + "BriefDescription": "Data TLB reload (after a miss) page size 1G. Implies radix translation was used", + "PublicDescription": "" + }, + {, + "EventCode": "0x44054", + "EventName": "PM_VECTOR_LD_CMPL", + "BriefDescription": "Number of vector load instructions completed", + "PublicDescription": "" + }, + {, + "EventCode": "0x4D05E", + "EventName": "PM_BR_CMPL", + "BriefDescription": "Any Branch instruction completed", + "PublicDescription": "" + }, + {, + "EventCode": "0x45054", + "EventName": "PM_FMA_CMPL", + "BriefDescription": "two flops operation completed (fmadd, fnmadd, fmsub, fnmsub) Scalar instructions only. ", + "PublicDescription": "" + }, + {, + "EventCode": "0x45056", + "EventName": "PM_SCALAR_FLOP_CMPL", + "BriefDescription": "Scalar flop operation completed", + "PublicDescription": "" + }, + {, + "EventCode": "0x4505C", + "EventName": "PM_MATH_FLOP_CMPL", + "BriefDescription": "Math flop instruction completed", + "PublicDescription": "" + }, + {, + "EventCode": "0x4E05E", + "EventName": "PM_TM_OUTER_TBEGIN_DISP", + "BriefDescription": "Number of outer tbegin instructions dispatched. The dispatch unit determines whether the tbegin instruction is outer or nested. This is a speculative count, which includes flushed instructions", + "PublicDescription": "" + }, + {, + "EventCode": "0x4F054", + "EventName": "PM_RADIX_PWC_MISS", + "BriefDescription": "A radix translation attempt missed in the TLB and all levels of page walk cache.", + "PublicDescription": "" + }, + {, + "EventCode": "0x4F05C", + "EventName": "PM_RADIX_PWC_L2_PTE_FROM_L3MISS", + "BriefDescription": "A Page Table Entry was reloaded to a level 2 page walk cache from beyond the core's L3 data cache. This implies that level 3 and level 4 PWC accesses were not necessary for this translation. The source could be local/remote/distant memory or another core's cache", + "PublicDescription": "" + }, + {, + "EventCode": "0x401E6", + "EventName": "PM_MRK_INST_FROM_L3MISS", + "BriefDescription": "Marked instruction was reloaded from a location beyond the local chiplet", + "PublicDescription": "" + }, + {, + "EventCode": "0x401E8", + "EventName": "PM_MRK_DATA_FROM_L2MISS", + "BriefDescription": "The processor's data cache was reloaded from a location other than the local core's L2 due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x400FA", + "EventName": "PM_RUN_INST_CMPL", + "BriefDescription": "Run_Instructions", + "PublicDescription": "" + } +] diff --git a/tools/perf/pmu-events/arch/powerpc/power9/pipeline.json b/tools/perf/pmu-events/arch/powerpc/power9/pipeline.json new file mode 100644 index 0000000..077c352 --- /dev/null +++ b/tools/perf/pmu-events/arch/powerpc/power9/pipeline.json @@ -0,0 +1,680 @@ +[ + {, + "EventCode": "0x1E", + "EventName": "PM_CYC", + "BriefDescription": "Cycles", + "PublicDescription": "" + }, + {, + "EventCode": "0x100F0", + "EventName": "PM_CYC", + "BriefDescription": "Cycles", + "PublicDescription": "" + }, + {, + "EventCode": "0x2", + "EventName": "PM_INST_CMPL", + "BriefDescription": "Number of PowerPC Instructions that completed.", + "PublicDescription": "" + }, + {, + "EventCode": "0x100FE", + "EventName": "PM_INST_CMPL", + "BriefDescription": "# PPC instructions completed", + "PublicDescription": "" + }, + {, + "EventCode": "0x10006", + "EventName": "PM_DISP_HELD", + "BriefDescription": "Dispatch Held", + "PublicDescription": "" + }, + {, + "EventCode": "0x10016", + "EventName": "PM_DSLB_MISS", + "BriefDescription": "Data SLB Miss - Total of all segment sizes", + "PublicDescription": "" + }, + {, + "EventCode": "0x10018", + "EventName": "PM_IC_DEMAND_CYC", + "BriefDescription": "Icache miss demand cycles", + "PublicDescription": "" + }, + {, + "EventCode": "0x10022", + "EventName": "PM_PMC2_SAVED", + "BriefDescription": "PMC2 Rewind Value saved", + "PublicDescription": "" + }, + {, + "EventCode": "0x10024", + "EventName": "PM_PMC5_OVERFLOW", + "BriefDescription": "Overflow from counter 5", + "PublicDescription": "" + }, + {, + "EventCode": "0x1D14A", + "EventName": "PM_MRK_DATA_FROM_RL2L3_MOD", + "BriefDescription": "The processor's data cache was reloaded with Modified (M) data from another chip's L2 or L3 on the same Node or Group (Remote), as this chip due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x1E040", + "EventName": "PM_DPTEG_FROM_L2_NO_CONFLICT", + "BriefDescription": "A Page Table Entry was loaded into the TLB from local core's L2 without conflict due to a data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x1E04A", + "EventName": "PM_DPTEG_FROM_RL2L3_SHR", + "BriefDescription": "A Page Table Entry was loaded into the TLB with Shared (S) data from another chip's L2 or L3 on the same Node or Group (Remote), as this chip due to a data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x1F140", + "EventName": "PM_MRK_DPTEG_FROM_L2_NO_CONFLICT", + "BriefDescription": "A Page Table Entry was loaded into the TLB from local core's L2 without conflict due to a marked data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x1F142", + "EventName": "PM_MRK_DPTEG_FROM_L2", + "BriefDescription": "A Page Table Entry was loaded into the TLB from local core's L2 due to a marked data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x1F144", + "EventName": "PM_MRK_DPTEG_FROM_L3_NO_CONFLICT", + "BriefDescription": "A Page Table Entry was loaded into the TLB from local core's L3 without conflict due to a marked data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x1F148", + "EventName": "PM_MRK_DPTEG_FROM_ON_CHIP_CACHE", + "BriefDescription": "A Page Table Entry was loaded into the TLB either shared or modified data from another core's L2/L3 on the same chip due to a marked data side request.. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x10050", + "EventName": "PM_CHIP_PUMP_CPRED", + "BriefDescription": "Initial and Final Pump Scope was chip pump (prediction=correct) for all data types excluding data prefetch (demand load,inst prefetch,inst fetch,xlate)", + "PublicDescription": "" + }, + {, + "EventCode": "0x10054", + "EventName": "PM_PUMP_CPRED", + "BriefDescription": "Pump prediction correct. Counts across all types of pumps for all data types excluding data prefetch (demand load,inst prefetch,inst fetch,xlate)", + "PublicDescription": "" + }, + {, + "EventCode": "0x10056", + "EventName": "PM_MEM_READ", + "BriefDescription": "Reads from Memory from this thread (includes data/inst/xlate/l1prefetch/inst prefetch). Includes L4", + "PublicDescription": "" + }, + {, + "EventCode": "0x1005A", + "EventName": "PM_CMPLU_STALL_DFLONG", + "BriefDescription": "Finish stall because the NTF instruction was a multi-cycle instruction issued to the Decimal Floating Point execution pipe and waiting to finish. Includes decimal floating point instructions + 128 bit binary floating point instructions. Qualified by multicycle", + "PublicDescription": "" + }, + {, + "EventCode": "0x1C056", + "EventName": "PM_DERAT_MISS_4K", + "BriefDescription": "Data ERAT Miss (Data TLB Access) page size 4K", + "PublicDescription": "" + }, + {, + "EventCode": "0x1C05A", + "EventName": "PM_DERAT_MISS_2M", + "BriefDescription": "Data ERAT Miss (Data TLB Access) page size 2M. Implies radix translation", + "PublicDescription": "" + }, + {, + "EventCode": "0x14050", + "EventName": "PM_INST_CHIP_PUMP_CPRED", + "BriefDescription": "Initial and Final Pump Scope was chip pump (prediction=correct) for an instruction fetch", + "PublicDescription": "" + }, + {, + "EventCode": "0x14052", + "EventName": "PM_INST_GRP_PUMP_MPRED_RTY", + "BriefDescription": "Final Pump Scope (Group) ended up larger than Initial Pump Scope (Chip) for an instruction fetch", + "PublicDescription": "" + }, + {, + "EventCode": "0x1D154", + "EventName": "PM_MRK_DATA_FROM_L2.1_SHR_CYC", + "BriefDescription": "Duration in cycles to reload with Shared (S) data from another core's L2 on the same chip due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x15156", + "EventName": "PM_SYNC_MRK_FX_DIVIDE", + "BriefDescription": "Marked fixed point divide that can cause a synchronous interrupt", + "PublicDescription": "" + }, + {, + "EventCode": "0x1E054", + "EventName": "PM_CMPLU_STALL", + "BriefDescription": "Nothing completed and ICT not empty", + "PublicDescription": "" + }, + {, + "EventCode": "0x1F058", + "EventName": "PM_RADIX_PWC_L2_PTE_FROM_L2", + "BriefDescription": "A Page Table Entry was reloaded to a level 2 page walk cache from the core's L2 data cache. This implies that level 3 and level 4 PWC accesses were not necessary for this translation", + "PublicDescription": "" + }, + {, + "EventCode": "0x10062", + "EventName": "PM_LD_L3MISS_PEND_CYC", + "BriefDescription": "Cycles L3 miss was pending for this thread", + "PublicDescription": "" + }, + {, + "EventCode": "0x10064", + "EventName": "PM_ICT_NOSLOT_DISP_HELD_TBEGIN", + "BriefDescription": "the NTC instruction is being held at dispatch because it is a tbegin instruction and there is an older tbegin in the pipeline that must complete before the younger tbegin can dispatch", + "PublicDescription": "" + }, + {, + "EventCode": "0x10068", + "EventName": "PM_BRU_FIN", + "BriefDescription": "Branch Instruction Finished", + "PublicDescription": "" + }, + {, + "EventCode": "0x100F6", + "EventName": "PM_IERAT_RELOAD", + "BriefDescription": "Number of I-ERAT reloads", + "PublicDescription": "" + }, + {, + "EventCode": "0x100F8", + "EventName": "PM_ICT_NOSLOT_CYC", + "BriefDescription": "Number of cycles the ICT has no itags assigned to this thread", + "PublicDescription": "" + }, + {, + "EventCode": "0x20008", + "EventName": "PM_ICT_EMPTY_CYC", + "BriefDescription": "Cycles in which the ICT is completely empty. No itags are assigned to any thread", + "PublicDescription": "" + }, + {, + "EventCode": "0x2000A", + "EventName": "PM_HV_CYC", + "BriefDescription": "Cycles in which msr_hv is high. Note that this event does not take msr_pr into consideration", + "PublicDescription": "" + }, + {, + "EventCode": "0x2000E", + "EventName": "PM_FXU_BUSY", + "BriefDescription": "Cycles in which all 4 FXUs are busy. The FXU is running at capacity", + "PublicDescription": "" + }, + {, + "EventCode": "0x2C018", + "EventName": "PM_CMPLU_STALL_DMISS_L21_L31", + "BriefDescription": "Completion stall by Dcache miss which resolved on chip ( excluding local L2/L3)", + "PublicDescription": "" + }, + {, + "EventCode": "0x2D012", + "EventName": "PM_CMPLU_STALL_DFU", + "BriefDescription": "Finish stall because the NTF instruction was issued to the Decimal Floating Point execution pipe and waiting to finish. Includes decimal floating point instructions + 128 bit binary floating point instructions. Not qualified by multicycle", + "PublicDescription": "" + }, + {, + "EventCode": "0x2D01A", + "EventName": "PM_ICT_NOSLOT_IC_MISS", + "BriefDescription": "Ict empty for this thread due to Icache Miss", + "PublicDescription": "" + }, + {, + "EventCode": "0x2E012", + "EventName": "PM_TM_TX_PASS_RUN_CYC", + "BriefDescription": "cycles spent in successful transactions", + "PublicDescription": "" + }, + {, + "EventCode": "0x2E016", + "EventName": "PM_NTC_ISSUE_HELD_ARB", + "BriefDescription": "The NTC instruction is being held at dispatch because it lost arbitration onto the issue pipe to another instruction (from the same thread or a different thread)", + "PublicDescription": "" + }, + {, + "EventCode": "0x2C046", + "EventName": "PM_DATA_FROM_RL2L3_MOD", + "BriefDescription": "The processor's data cache was reloaded with Modified (M) data from another chip's L2 or L3 on the same Node or Group (Remote), as this chip due to a demand load", + "PublicDescription": "" + }, + {, + "EventCode": "0x2404C", + "EventName": "PM_INST_FROM_MEMORY", + "BriefDescription": "The processor's Instruction cache was reloaded from a memory location including L4 from local remote or distant due to an instruction fetch (not prefetch)", + "PublicDescription": "" + }, + {, + "EventCode": "0x2D14E", + "EventName": "PM_MRK_DATA_FROM_L2.1_SHR", + "BriefDescription": "The processor's data cache was reloaded with Shared (S) data from another core's L2 on the same chip due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x2E042", + "EventName": "PM_DPTEG_FROM_L3_MEPF", + "BriefDescription": "A Page Table Entry was loaded into the TLB from local core's L3 without dispatch conflicts hit on Mepf state. due to a data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x2E046", + "EventName": "PM_DPTEG_FROM_RL2L3_MOD", + "BriefDescription": "A Page Table Entry was loaded into the TLB with Modified (M) data from another chip's L2 or L3 on the same Node or Group (Remote), as this chip due to a data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x2F142", + "EventName": "PM_MRK_DPTEG_FROM_L3_MEPF", + "BriefDescription": "A Page Table Entry was loaded into the TLB from local core's L3 without dispatch conflicts hit on Mepf state. due to a marked data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x2F144", + "EventName": "PM_MRK_DPTEG_FROM_L3.1_MOD", + "BriefDescription": "A Page Table Entry was loaded into the TLB with Modified (M) data from another core's L3 on the same chip due to a marked data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x2F146", + "EventName": "PM_MRK_DPTEG_FROM_RL2L3_MOD", + "BriefDescription": "A Page Table Entry was loaded into the TLB with Modified (M) data from another chip's L2 or L3 on the same Node or Group (Remote), as this chip due to a marked data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x2F14C", + "EventName": "PM_MRK_DPTEG_FROM_MEMORY", + "BriefDescription": "A Page Table Entry was loaded into the TLB from a memory location including L4 from local remote or distant due to a marked data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x20052", + "EventName": "PM_GRP_PUMP_MPRED", + "BriefDescription": "Final Pump Scope (Group) ended up either larger or smaller than Initial Pump Scope for all data types excluding data prefetch (demand load,inst prefetch,inst fetch,xlate)", + "PublicDescription": "" + }, + {, + "EventCode": "0x2C054", + "EventName": "PM_DERAT_MISS_64K", + "BriefDescription": "Data ERAT Miss (Data TLB Access) page size 64K", + "PublicDescription": "" + }, + {, + "EventCode": "0x2C05A", + "EventName": "PM_DERAT_MISS_1G", + "BriefDescription": "Data ERAT Miss (Data TLB Access) page size 1G. Implies radix translation", + "PublicDescription": "" + }, + {, + "EventCode": "0x24052", + "EventName": "PM_FXU_IDLE", + "BriefDescription": "Cycles in which FXU0, FXU1, FXU2, and FXU3 are all idle", + "PublicDescription": "" + }, + {, + "EventCode": "0x2405A", + "EventName": "PM_NTC_FIN", + "BriefDescription": "Cycles in which the oldest instruction in the pipeline (NTC) finishes. This event is used to account for cycles in which work is being completed in the CPI stack", + "PublicDescription": "" + }, + {, + "EventCode": "0x2D150", + "EventName": "PM_MRK_DERAT_MISS_4K", + "BriefDescription": "Marked Data ERAT Miss (Data TLB Access) page size 4K", + "PublicDescription": "" + }, + {, + "EventCode": "0x2D152", + "EventName": "PM_MRK_DERAT_MISS_2M", + "BriefDescription": "Marked Data ERAT Miss (Data TLB Access) page size 2M. Implies radix translation", + "PublicDescription": "" + }, + {, + "EventCode": "0x20066", + "EventName": "PM_TLB_MISS", + "BriefDescription": "TLB Miss (I + D)", + "PublicDescription": "" + }, + {, + "EventCode": "0x201E2", + "EventName": "PM_MRK_LD_MISS_L1", + "BriefDescription": "Marked DL1 Demand Miss counted at exec time. Note that this count is per slice, so if a load spans multiple slices this event will increment multiple times for a single load.", + "PublicDescription": "" + }, + {, + "EventCode": "0x200F4", + "EventName": "PM_RUN_CYC", + "BriefDescription": "Run_cycles", + "PublicDescription": "" + }, + {, + "EventCode": "0x200F8", + "EventName": "PM_EXT_INT", + "BriefDescription": "external interrupt", + "PublicDescription": "" + }, + {, + "EventCode": "0x30004", + "EventName": "PM_CMPLU_STALL_EMQ_FULL", + "BriefDescription": "Finish stall because the next to finish instruction suffered an ERAT miss and the EMQ was full", + "PublicDescription": "" + }, + {, + "EventCode": "0x30020", + "EventName": "PM_PMC2_REWIND", + "BriefDescription": "PMC2 Rewind Event (did not match condition)", + "PublicDescription": "" + }, + {, + "EventCode": "0x30022", + "EventName": "PM_PMC4_SAVED", + "BriefDescription": "PMC4 Rewind Value saved (matched condition)", + "PublicDescription": "" + }, + {, + "EventCode": "0x30024", + "EventName": "PM_PMC6_OVERFLOW", + "BriefDescription": "Overflow from counter 6", + "PublicDescription": "" + }, + {, + "EventCode": "0x3C04C", + "EventName": "PM_DATA_FROM_DL4", + "BriefDescription": "The processor's data cache was reloaded from another chip's L4 on a different Node or Group (Distant) due to a demand load", + "PublicDescription": "" + }, + {, + "EventCode": "0x3D148", + "EventName": "PM_MRK_DATA_FROM_L2.1_MOD_CYC", + "BriefDescription": "Duration in cycles to reload with Modified (M) data from another core's L2 on the same chip due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x3E042", + "EventName": "PM_DPTEG_FROM_L3_DISP_CONFLICT", + "BriefDescription": "A Page Table Entry was loaded into the TLB from local core's L3 with dispatch conflict due to a data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x3E046", + "EventName": "PM_DPTEG_FROM_L2.1_SHR", + "BriefDescription": "A Page Table Entry was loaded into the TLB with Shared (S) data from another core's L2 on the same chip due to a data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x3F148", + "EventName": "PM_MRK_DPTEG_FROM_DL2L3_SHR", + "BriefDescription": "A Page Table Entry was loaded into the TLB with Shared (S) data from another chip's L2 or L3 on a different Node or Group (Distant), as this chip due to a marked data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x3F14C", + "EventName": "PM_MRK_DPTEG_FROM_DL4", + "BriefDescription": "A Page Table Entry was loaded into the TLB from another chip's L4 on a different Node or Group (Distant) due to a marked data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x30056", + "EventName": "PM_TM_ABORTS", + "BriefDescription": "Number of TM transactions aborted", + "PublicDescription": "" + }, + {, + "EventCode": "0x30058", + "EventName": "PM_TLBIE_FIN", + "BriefDescription": "tlbie finished", + "PublicDescription": "" + }, + {, + "EventCode": "0x3C054", + "EventName": "PM_DERAT_MISS_16M", + "BriefDescription": "Data ERAT Miss (Data TLB Access) page size 16M", + "PublicDescription": "" + }, + {, + "EventCode": "0x34058", + "EventName": "PM_ICT_NOSLOT_BR_MPRED_ICMISS", + "BriefDescription": "Ict empty for this thread due to Icache Miss and branch mispred", + "PublicDescription": "" + }, + {, + "EventCode": "0x3405C", + "EventName": "PM_CMPLU_STALL_DPLONG", + "BriefDescription": "Finish stall because the NTF instruction was a scalar multi-cycle instruction issued to the Double Precision execution pipe and waiting to finish. Includes binary floating point instructions in 32 and 64 bit binary floating point format. Qualified by NOT vector AND multicycle", + "PublicDescription": "" + }, + {, + "EventCode": "0x3405E", + "EventName": "PM_IFETCH_THROTTLE", + "BriefDescription": "Cycles in which Instruction fetch throttle was active.", + "PublicDescription": "" + }, + {, + "EventCode": "0x3D152", + "EventName": "PM_MRK_DERAT_MISS_1G", + "BriefDescription": "Marked Data ERAT Miss (Data TLB Access) page size 1G. Implies radix translation", + "PublicDescription": "" + }, + {, + "EventCode": "0x3D154", + "EventName": "PM_MRK_DERAT_MISS_16M", + "BriefDescription": "Marked Data ERAT Miss (Data TLB Access) page size 16M", + "PublicDescription": "" + }, + {, + "EventCode": "0x3D058", + "EventName": "PM_VSU_DP_FSQRT_FDIV", + "BriefDescription": "vector versions of fdiv,fsqrt", + "PublicDescription": "" + }, + {, + "EventCode": "0x35150", + "EventName": "PM_MRK_DATA_FROM_RL2L3_SHR", + "BriefDescription": "The processor's data cache was reloaded with Shared (S) data from another chip's L2 or L3 on the same Node or Group (Remote), as this chip due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x3E052", + "EventName": "PM_ICT_NOSLOT_IC_L3", + "BriefDescription": "Ict empty for this thread due to icache misses that were sourced from the local L3", + "PublicDescription": "" + }, + {, + "EventCode": "0x3F05A", + "EventName": "PM_RADIX_PWC_L2_PDE_FROM_L3", + "BriefDescription": "A Page Directory Entry was reloaded to a level 2 page walk cache from the core's L3 data cache", + "PublicDescription": "" + }, + {, + "EventCode": "0x300FC", + "EventName": "PM_DTLB_MISS", + "BriefDescription": "Data PTEG reload", + "PublicDescription": "" + }, + {, + "EventCode": "0x40004", + "EventName": "PM_FXU_FIN", + "BriefDescription": "The fixed point unit Unit finished an instruction. Instructions that finish may not necessary complete.", + "PublicDescription": "" + }, + {, + "EventCode": "0x40010", + "EventName": "PM_PMC3_OVERFLOW", + "BriefDescription": "Overflow from counter 3", + "PublicDescription": "" + }, + {, + "EventCode": "0x4C012", + "EventName": "PM_CMPLU_STALL_ERAT_MISS", + "BriefDescription": "Finish stall because the NTF instruction was a load or store that suffered a translation miss", + "PublicDescription": "" + }, + {, + "EventCode": "0x4D01C", + "EventName": "PM_ICT_NOSLOT_DISP_HELD_SYNC", + "BriefDescription": "Dispatch held due to a synchronizing instruction at dispatch", + "PublicDescription": "" + }, + {, + "EventCode": "0x4D01E", + "EventName": "PM_ICT_NOSLOT_BR_MPRED", + "BriefDescription": "Ict empty for this thread due to branch mispred", + "PublicDescription": "" + }, + {, + "EventCode": "0x4E010", + "EventName": "PM_ICT_NOSLOT_IC_L3MISS", + "BriefDescription": "Ict empty for this thread due to icache misses that were sourced from beyond the local L3. The source could be local/remote/distant memory or another core's cache", + "PublicDescription": "" + }, + {, + "EventCode": "0x4E01A", + "EventName": "PM_ICT_NOSLOT_DISP_HELD", + "BriefDescription": "Cycles in which the NTC instruction is held at dispatch for any reason", + "PublicDescription": "" + }, + {, + "EventCode": "0x4C120", + "EventName": "PM_MRK_DATA_FROM_L2_MEPF", + "BriefDescription": "The processor's data cache was reloaded from local core's L2 hit without dispatch conflicts on Mepf state. due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x4D124", + "EventName": "PM_MRK_DATA_FROM_L3.1_SHR", + "BriefDescription": "The processor's data cache was reloaded with Shared (S) data from another core's L3 on the same chip due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x4C04C", + "EventName": "PM_DATA_FROM_DMEM", + "BriefDescription": "The processor's data cache was reloaded from another chip's memory on the same Node or Group (Distant) due to a demand load", + "PublicDescription": "" + }, + {, + "EventCode": "0x4D140", + "EventName": "PM_MRK_DATA_FROM_ON_CHIP_CACHE", + "BriefDescription": "The processor's data cache was reloaded either shared or modified data from another core's L2/L3 on the same chip due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x4D04C", + "EventName": "PM_DFU_BUSY", + "BriefDescription": "Cycles in which all 4 Decimal Floating Point units are busy. The DFU is running at capacity", + "PublicDescription": "" + }, + {, + "EventCode": "0x4D04E", + "EventName": "PM_VSU_FSQRT_FDIV", + "BriefDescription": "four flops operation (fdiv,fsqrt) Scalar Instructions only", + "PublicDescription": "" + }, + {, + "EventCode": "0x4E046", + "EventName": "PM_DPTEG_FROM_L2.1_MOD", + "BriefDescription": "A Page Table Entry was loaded into the TLB with Modified (M) data from another core's L2 on the same chip due to a data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x4F146", + "EventName": "PM_MRK_DPTEG_FROM_L2.1_MOD", + "BriefDescription": "A Page Table Entry was loaded into the TLB with Modified (M) data from another core's L2 on the same chip due to a marked data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x4F14A", + "EventName": "PM_MRK_DPTEG_FROM_OFF_CHIP_CACHE", + "BriefDescription": "A Page Table Entry was loaded into the TLB either shared or modified data from another core's L2/L3 on a different chip (remote or distant) due to a marked data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x4F14C", + "EventName": "PM_MRK_DPTEG_FROM_DMEM", + "BriefDescription": "A Page Table Entry was loaded into the TLB from another chip's memory on the same Node or Group (Distant) due to a marked data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x40052", + "EventName": "PM_PUMP_MPRED", + "BriefDescription": "Pump misprediction. Counts across all types of pumps for all data types excluding data prefetch (demand load,inst prefetch,inst fetch,xlate)", + "PublicDescription": "" + }, + {, + "EventCode": "0x4C058", + "EventName": "PM_MEM_CO", + "BriefDescription": "Memory castouts from this thread", + "PublicDescription": "" + }, + {, + "EventCode": "0x4C15C", + "EventName": "PM_MRK_DERAT_MISS_16G", + "BriefDescription": "Marked Data ERAT Miss (Data TLB Access) page size 16G", + "PublicDescription": "" + }, + {, + "EventCode": "0x4D050", + "EventName": "PM_VSU_NON_FLOP_CMPL", + "BriefDescription": "Non FLOP operation completed", + "PublicDescription": "" + }, + {, + "EventCode": "0x4D052", + "EventName": "PM_2FLOP_CMPL", + "BriefDescription": "DP vector version of fmul, fsub, fcmp, fsel, fabs, fnabs, fres ,fsqrte, fneg ", + "PublicDescription": "" + }, + {, + "EventCode": "0x45058", + "EventName": "PM_IC_MISS_CMPL", + "BriefDescription": "Non-speculative icache miss, counted at completion", + "PublicDescription": "" + }, + {, + "EventCode": "0x40062", + "EventName": "PM_DUMMY1_REMOVE_ME", + "BriefDescription": "Space holder for L2_PC_PM_MK_LDST_SCOPE_PRED_STATUS", + "PublicDescription": "" + }, + {, + "EventCode": "0x40064", + "EventName": "PM_DUMMY2_REMOVE_ME", + "BriefDescription": "Space holder for LS_PC_RELOAD_RA", + "PublicDescription": "" + }, + {, + "EventCode": "0x4006A", + "EventName": "PM_IERAT_RELOAD_16M", + "BriefDescription": "IERAT Reloaded (Miss) for a 16M page", + "PublicDescription": "" + }, + {, + "EventCode": "0x401EC", + "EventName": "PM_THRESH_EXC_2048", + "BriefDescription": "Threshold counter exceeded a value of 2048", + "PublicDescription": "" + }, + {, + "EventCode": "0x400F2", + "EventName": "PM_1PLUS_PPC_DISP", + "BriefDescription": "Cycles at least one Instr Dispatched", + "PublicDescription": "" + }, + {, + "EventCode": "0x400F8", + "EventName": "PM_FLUSH", + "BriefDescription": "Flush (any type)", + "PublicDescription": "" + } +] diff --git a/tools/perf/pmu-events/arch/powerpc/power9/pmc.json b/tools/perf/pmu-events/arch/powerpc/power9/pmc.json new file mode 100644 index 0000000..32ce711 --- /dev/null +++ b/tools/perf/pmu-events/arch/powerpc/power9/pmc.json @@ -0,0 +1,146 @@ +[ + {, + "EventCode": "0x0", + "EventName": "PM_SUSPENDED", + "BriefDescription": "Counter OFF", + "PublicDescription": "" + }, + {, + "EventCode": "0x10026", + "EventName": "PM_TABLEWALK_CYC", + "BriefDescription": "Cycles when an instruction tablewalk is active", + "PublicDescription": "" + }, + {, + "EventCode": "0x1E04C", + "EventName": "PM_DPTEG_FROM_LL4", + "BriefDescription": "A Page Table Entry was loaded into the TLB from the local chip's L4 cache due to a data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x1F14E", + "EventName": "PM_MRK_DPTEG_FROM_L2MISS", + "BriefDescription": "A Page Table Entry was loaded into the TLB from a location other than the local core's L2 due to a marked data side request.. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x10060", + "EventName": "PM_TM_TRANS_RUN_CYC", + "BriefDescription": "run cycles in transactional state", + "PublicDescription": "" + }, + {, + "EventCode": "0x2C012", + "EventName": "PM_CMPLU_STALL_DCACHE_MISS", + "BriefDescription": "Finish stall because the NTF instruction was a load that missed the L1 and was waiting for the data to return from the nest", + "PublicDescription": "" + }, + {, + "EventCode": "0x2E04C", + "EventName": "PM_DPTEG_FROM_MEMORY", + "BriefDescription": "A Page Table Entry was loaded into the TLB from a memory location including L4 from local remote or distant due to a data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x2C056", + "EventName": "PM_DTLB_MISS_4K", + "BriefDescription": "Data TLB Miss page size 4k", + "PublicDescription": "" + }, + {, + "EventCode": "0x3000C", + "EventName": "PM_FREQ_DOWN", + "BriefDescription": "Power Management: Below Threshold B", + "PublicDescription": "" + }, + {, + "EventCode": "0x3D142", + "EventName": "PM_MRK_DATA_FROM_LMEM", + "BriefDescription": "The processor's data cache was reloaded from the local chip's Memory due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x3F142", + "EventName": "PM_MRK_DPTEG_FROM_L3_DISP_CONFLICT", + "BriefDescription": "A Page Table Entry was loaded into the TLB from local core's L3 with dispatch conflict due to a marked data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x301E8", + "EventName": "PM_THRESH_EXC_64", + "BriefDescription": "Threshold counter exceeded a value of 64", + "PublicDescription": "" + }, + {, + "EventCode": "0x40118", + "EventName": "PM_MRK_DCACHE_RELOAD_INTV", + "BriefDescription": "Combined Intervention event", + "PublicDescription": "" + }, + {, + "EventCode": "0x4C01E", + "EventName": "PM_CMPLU_STALL_CRYPTO", + "BriefDescription": "Finish stall because the NTF instruction was routed to the crypto execution pipe and was waiting to finish", + "PublicDescription": "" + }, + {, + "EventCode": "0x4D018", + "EventName": "PM_CMPLU_STALL_BRU", + "BriefDescription": "Completion stall due to a Branch Unit", + "PublicDescription": "" + }, + {, + "EventCode": "0x4D128", + "EventName": "PM_MRK_DATA_FROM_LMEM_CYC", + "BriefDescription": "Duration in cycles to reload from the local chip's Memory due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x4E04E", + "EventName": "PM_DPTEG_FROM_L3MISS", + "BriefDescription": "A Page Table Entry was loaded into the TLB from a location other than the local core's L3 due to a data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x4F142", + "EventName": "PM_MRK_DPTEG_FROM_L3", + "BriefDescription": "A Page Table Entry was loaded into the TLB from local core's L3 due to a marked data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x4F148", + "EventName": "PM_MRK_DPTEG_FROM_DL2L3_MOD", + "BriefDescription": "A Page Table Entry was loaded into the TLB with Modified (M) data from another chip's L2 or L3 on a different Node or Group (Distant), as this chip due to a marked data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x40050", + "EventName": "PM_SYS_PUMP_MPRED_RTY", + "BriefDescription": "Final Pump Scope (system) ended up larger than Initial Pump Scope (Chip/Group) for all data types excluding data prefetch (demand load,inst prefetch,inst fetch,xlate)", + "PublicDescription": "" + }, + {, + "EventCode": "0x40056", + "EventName": "PM_MEM_LOC_THRESH_LSU_HIGH", + "BriefDescription": "Local memory above threshold for LSU medium", + "PublicDescription": "" + }, + {, + "EventCode": "0x4D054", + "EventName": "PM_8FLOP_CMPL", + "BriefDescription": "8 FLOP instruction completed", + "PublicDescription": "" + }, + {, + "EventCode": "0x45050", + "EventName": "PM_1FLOP_CMPL", + "BriefDescription": "one flop (fadd, fmul, fsub, fcmp, fsel, fabs, fnabs, fres, fsqrte, fneg) operation completed", + "PublicDescription": "" + }, + {, + "EventCode": "0x45052", + "EventName": "PM_4FLOP_CMPL", + "BriefDescription": "4 FLOP instruction completed", + "PublicDescription": "" + } +] diff --git a/tools/perf/pmu-events/arch/powerpc/power9/translation.json b/tools/perf/pmu-events/arch/powerpc/power9/translation.json new file mode 100644 index 0000000..d758598 --- /dev/null +++ b/tools/perf/pmu-events/arch/powerpc/power9/translation.json @@ -0,0 +1,272 @@ +[ + {, + "EventCode": "0x10028", + "EventName": "PM_STALL_END_ICT_EMPTY", + "BriefDescription": "The number a times the core transitioned from a stall to ICT-empty for this thread", + "PublicDescription": "" + }, + {, + "EventCode": "0x1C04E", + "EventName": "PM_DATA_FROM_L2MISS_MOD", + "BriefDescription": "The processor's data cache was reloaded from a location other than the local core's L2 due to a demand load", + "PublicDescription": "" + }, + {, + "EventCode": "0x14044", + "EventName": "PM_INST_FROM_L3_NO_CONFLICT", + "BriefDescription": "The processor's Instruction cache was reloaded from local core's L3 without conflict due to an instruction fetch (not prefetch)", + "PublicDescription": "" + }, + {, + "EventCode": "0x1404E", + "EventName": "PM_INST_FROM_L2MISS", + "BriefDescription": "The processor's Instruction cache was reloaded from a location other than the local core's L2 due to an instruction fetch (not prefetch)", + "PublicDescription": "" + }, + {, + "EventCode": "0x1D142", + "EventName": "PM_MRK_DATA_FROM_L3.1_ECO_SHR_CYC", + "BriefDescription": "Duration in cycles to reload with Shared (S) data from another core's ECO L3 on the same chip due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x15048", + "EventName": "PM_IPTEG_FROM_ON_CHIP_CACHE", + "BriefDescription": "A Page Table Entry was loaded into the TLB either shared or modified data from another core's L2/L3 on the same chip due to a instruction side request", + "PublicDescription": "" + }, + {, + "EventCode": "0x1504A", + "EventName": "PM_IPTEG_FROM_RL2L3_SHR", + "BriefDescription": "A Page Table Entry was loaded into the TLB with Shared (S) data from another chip's L2 or L3 on the same Node or Group (Remote), as this chip due to a instruction side request", + "PublicDescription": "" + }, + {, + "EventCode": "0x1E058", + "EventName": "PM_STCX_FAIL", + "BriefDescription": "stcx failed", + "PublicDescription": "" + }, + {, + "EventCode": "0x1F15E", + "EventName": "PM_MRK_PROBE_NOP_CMPL", + "BriefDescription": "Marked probeNops completed", + "PublicDescription": "" + }, + {, + "EventCode": "0x20112", + "EventName": "PM_MRK_NTF_FIN", + "BriefDescription": "Marked next to finish instruction finished", + "PublicDescription": "" + }, + {, + "EventCode": "0x20016", + "EventName": "PM_ST_FIN", + "BriefDescription": "Store finish count. Includes speculative activity", + "PublicDescription": "" + }, + {, + "EventCode": "0x20018", + "EventName": "PM_ST_FWD", + "BriefDescription": "Store forwards that finished", + "PublicDescription": "" + }, + {, + "EventCode": "0x2011C", + "EventName": "PM_MRK_NTC_CYC", + "BriefDescription": "Cycles during which the marked instruction is next to complete (completion is held up because the marked instruction hasn't completed yet)", + "PublicDescription": "" + }, + {, + "EventCode": "0x2E018", + "EventName": "PM_CMPLU_STALL_VFXLONG", + "BriefDescription": "Completion stall due to a long latency vector fixed point instruction (division, square root)", + "PublicDescription": "" + }, + {, + "EventCode": "0x2E01C", + "EventName": "PM_CMPLU_STALL_TLBIE", + "BriefDescription": "Finish stall because the NTF instruction was a tlbie waiting for response from L2", + "PublicDescription": "" + }, + {, + "EventCode": "0x2003E", + "EventName": "PM_LSU_LMQ_SRQ_EMPTY_CYC", + "BriefDescription": "Cycles in which the LSU is empty for all threads (lmq and srq are completely empty)", + "PublicDescription": "" + }, + {, + "EventCode": "0x24042", + "EventName": "PM_INST_FROM_L3_MEPF", + "BriefDescription": "The processor's Instruction cache was reloaded from local core's L3 without dispatch conflicts hit on Mepf state. due to an instruction fetch (not prefetch)", + "PublicDescription": "" + }, + {, + "EventCode": "0x2D14A", + "EventName": "PM_MRK_DATA_FROM_RL2L3_MOD_CYC", + "BriefDescription": "Duration in cycles to reload with Modified (M) data from another chip's L2 or L3 on the same Node or Group (Remote), as this chip due to a marked load", + "PublicDescription": "" + }, + {, + "EventCode": "0x25046", + "EventName": "PM_IPTEG_FROM_RL2L3_MOD", + "BriefDescription": "A Page Table Entry was loaded into the TLB with Modified (M) data from another chip's L2 or L3 on the same Node or Group (Remote), as this chip due to a instruction side request", + "PublicDescription": "" + }, + {, + "EventCode": "0x2504A", + "EventName": "PM_IPTEG_FROM_RL4", + "BriefDescription": "A Page Table Entry was loaded into the TLB from another chip's L4 on the same Node or Group ( Remote) due to a instruction side request", + "PublicDescription": "" + }, + {, + "EventCode": "0x2504C", + "EventName": "PM_IPTEG_FROM_MEMORY", + "BriefDescription": "A Page Table Entry was loaded into the TLB from a memory location including L4 from local remote or distant due to a instruction side request", + "PublicDescription": "" + }, + {, + "EventCode": "0x201E6", + "EventName": "PM_THRESH_EXC_32", + "BriefDescription": "Threshold counter exceeded a value of 32", + "PublicDescription": "" + }, + {, + "EventCode": "0x200F0", + "EventName": "PM_ST_CMPL", + "BriefDescription": "Stores completed from S2Q (2nd-level store queue).", + "PublicDescription": "" + }, + {, + "EventCode": "0x200FE", + "EventName": "PM_DATA_FROM_L2MISS", + "BriefDescription": "Demand LD - L2 Miss (not L2 hit)", + "PublicDescription": "" + }, + {, + "EventCode": "0x30010", + "EventName": "PM_PMC2_OVERFLOW", + "BriefDescription": "Overflow from counter 2", + "PublicDescription": "" + }, + {, + "EventCode": "0x3C046", + "EventName": "PM_DATA_FROM_L2.1_SHR", + "BriefDescription": "The processor's data cache was reloaded with Shared (S) data from another core's L2 on the same chip due to a demand load", + "PublicDescription": "" + }, + {, + "EventCode": "0x34042", + "EventName": "PM_INST_FROM_L3_DISP_CONFLICT", + "BriefDescription": "The processor's Instruction cache was reloaded from local core's L3 with dispatch conflict due to an instruction fetch (not prefetch)", + "PublicDescription": "" + }, + {, + "EventCode": "0x34046", + "EventName": "PM_INST_FROM_L2.1_SHR", + "BriefDescription": "The processor's Instruction cache was reloaded with Shared (S) data from another core's L2 on the same chip due to an instruction fetch (not prefetch)", + "PublicDescription": "" + }, + {, + "EventCode": "0x3504A", + "EventName": "PM_IPTEG_FROM_RMEM", + "BriefDescription": "A Page Table Entry was loaded into the TLB from another chip's memory on the same Node or Group ( Remote) due to a instruction side request", + "PublicDescription": "" + }, + {, + "EventCode": "0x3E048", + "EventName": "PM_DPTEG_FROM_DL2L3_SHR", + "BriefDescription": "A Page Table Entry was loaded into the TLB with Shared (S) data from another chip's L2 or L3 on a different Node or Group (Distant), as this chip due to a data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x3E04C", + "EventName": "PM_DPTEG_FROM_DL4", + "BriefDescription": "A Page Table Entry was loaded into the TLB from another chip's L4 on a different Node or Group (Distant) due to a data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x3C05A", + "EventName": "PM_CMPLU_STALL_VDPLONG", + "BriefDescription": "Finish stall because the NTF instruction was a scalar multi-cycle instruction issued to the Double Precision execution pipe and waiting to finish. Includes binary floating point instructions in 32 and 64 bit binary floating point format. Qualified by NOT vector AND multicycle", + "PublicDescription": "" + }, + {, + "EventCode": "0x3C05C", + "EventName": "PM_CMPLU_STALL_VFXU", + "BriefDescription": "Finish stall due to a vector fixed point instruction in the execution pipeline. These instructions get routed to the ALU, ALU2, and DIV pipes", + "PublicDescription": "" + }, + {, + "EventCode": "0x30066", + "EventName": "PM_LSU_FIN", + "BriefDescription": "LSU Finished a PPC instruction (up to 4 per cycle)", + "PublicDescription": "" + }, + {, + "EventCode": "0x300F0", + "EventName": "PM_ST_MISS_L1", + "BriefDescription": "Store Missed L1", + "PublicDescription": "" + }, + {, + "EventCode": "0x4D010", + "EventName": "PM_PMC1_SAVED", + "BriefDescription": "PMC1 Rewind Value saved", + "PublicDescription": "" + }, + {, + "EventCode": "0x40132", + "EventName": "PM_MRK_LSU_FIN", + "BriefDescription": "lsu marked instr PPC finish", + "PublicDescription": "" + }, + {, + "EventCode": "0x4C046", + "EventName": "PM_DATA_FROM_L2.1_MOD", + "BriefDescription": "The processor's data cache was reloaded with Modified (M) data from another core's L2 on the same chip due to a demand load", + "PublicDescription": "" + }, + {, + "EventCode": "0x44042", + "EventName": "PM_INST_FROM_L3", + "BriefDescription": "The processor's Instruction cache was reloaded from local core's L3 due to an instruction fetch (not prefetch)", + "PublicDescription": "" + }, + {, + "EventCode": "0x4504A", + "EventName": "PM_IPTEG_FROM_OFF_CHIP_CACHE", + "BriefDescription": "A Page Table Entry was loaded into the TLB either shared or modified data from another core's L2/L3 on a different chip (remote or distant) due to a instruction side request", + "PublicDescription": "" + }, + {, + "EventCode": "0x4E048", + "EventName": "PM_DPTEG_FROM_DL2L3_MOD", + "BriefDescription": "A Page Table Entry was loaded into the TLB with Modified (M) data from another chip's L2 or L3 on a different Node or Group (Distant), as this chip due to a data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x4E04C", + "EventName": "PM_DPTEG_FROM_DMEM", + "BriefDescription": "A Page Table Entry was loaded into the TLB from another chip's memory on the same Node or Group (Distant) due to a data side request. When using Radix Page Translation, this count excludes PDE reloads. Only PTE reloads are included", + "PublicDescription": "" + }, + {, + "EventCode": "0x4405C", + "EventName": "PM_CMPLU_STALL_VDP", + "BriefDescription": "Finish stall because the NTF instruction was a vector instruction issued to the Double Precision execution pipe and waiting to finish. Includes binary floating point instructions in 32 and 64 bit binary floating point format. Not qualified multicycle. Qualified by vector", + "PublicDescription": "" + }, + {, + "EventCode": "0x4D05C", + "EventName": "PM_DP_QP_FLOP_CMPL", + "BriefDescription": "Double-Precion or Quad-Precision instruction completed", + "PublicDescription": "" + }, + {, + "EventCode": "0x4E05C", + "EventName": "PM_LSU_REJECT_LHS", + "BriefDescription": "LSU Reject due to LHS (up to 4 per cycle)", + "PublicDescription": "" + } +] -- cgit v1.1 From 80e63ffb09a1b7ac645e9dd1c16b7b08956a7f5b Mon Sep 17 00:00:00 2001 From: Sukadev Bhattiprolu Date: Tue, 11 Jul 2017 13:00:31 -0500 Subject: perf vendor events: Add POWER9 PVRs to mapfile Add currently supported POWER9 PVRs to the mapfile Signed-off-by: Sukadev Bhattiprolu Cc: Andi Kleen Cc: Jiri Olsa Cc: Madhavan Srinivasan Cc: Michael Ellerman Cc: Shriya Link: http://lkml.kernel.org/n/tip-k1pe02sn5gh6nrzp8ditye94@git.kernel.org [ Fix conflict with a87006fd5629 ("perf pmu-events: Support additional POWER8+ PVR in mapfile") ] Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/pmu-events/arch/powerpc/mapfile.csv | 3 +++ 1 file changed, 3 insertions(+) (limited to 'tools') diff --git a/tools/perf/pmu-events/arch/powerpc/mapfile.csv b/tools/perf/pmu-events/arch/powerpc/mapfile.csv index f03aec7..14318ef2 100644 --- a/tools/perf/pmu-events/arch/powerpc/mapfile.csv +++ b/tools/perf/pmu-events/arch/powerpc/mapfile.csv @@ -20,3 +20,6 @@ 004d0100,1,power8.json,core 004d0200,1,power8.json,core 004c0100,1,power8.json,core +004e0100,1,power9.json,core +004e0200,1,power9.json,core +004e1200,1,power9.json,core -- cgit v1.1 From 047726d1f9ae6b1f3a2144577ffc3a1a92ffedf4 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Tue, 18 Jul 2017 17:09:21 -0300 Subject: tools include uapi x86: Grab a copy of unistd.h In older distros we were not including our copies of unistd_{32,64}.h, as we were relying on the system's asm/unistd.h, and a log time ago the files to be included were asm-{x86_64,i386}/unistd.h. Fix it by also carrying a copy of asm/unistd.h, that will be the same as in modern distros and will allow us to provide missing __NR_setns, for instance, in older distros. Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-iwmgm0c4m1ynstktzmkjh8di@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/arch/x86/include/uapi/asm/unistd.h | 17 +++++++++++++++++ tools/perf/check-headers.sh | 1 + 2 files changed, 18 insertions(+) create mode 100644 tools/arch/x86/include/uapi/asm/unistd.h (limited to 'tools') diff --git a/tools/arch/x86/include/uapi/asm/unistd.h b/tools/arch/x86/include/uapi/asm/unistd.h new file mode 100644 index 0000000..a26df0d --- /dev/null +++ b/tools/arch/x86/include/uapi/asm/unistd.h @@ -0,0 +1,17 @@ +#ifndef _UAPI_ASM_X86_UNISTD_H +#define _UAPI_ASM_X86_UNISTD_H + +/* x32 syscall flag bit */ +#define __X32_SYSCALL_BIT 0x40000000 + +#ifndef __KERNEL__ +# ifdef __i386__ +# include +# elif defined(__ILP32__) +# include +# else +# include +# endif +#endif + +#endif /* _UAPI_ASM_X86_UNISTD_H */ diff --git a/tools/perf/check-headers.sh b/tools/perf/check-headers.sh index 83fe220..47abd33 100755 --- a/tools/perf/check-headers.sh +++ b/tools/perf/check-headers.sh @@ -16,6 +16,7 @@ arch/x86/include/uapi/asm/perf_regs.h arch/x86/include/uapi/asm/kvm.h arch/x86/include/uapi/asm/kvm_perf.h arch/x86/include/uapi/asm/svm.h +arch/x86/include/uapi/asm/unistd.h arch/x86/include/uapi/asm/vmx.h arch/powerpc/include/uapi/asm/kvm.h arch/s390/include/uapi/asm/kvm.h -- cgit v1.1 From 59291f19824200b5a9046b79d37f02fb415335b0 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Tue, 18 Jul 2017 17:13:40 -0300 Subject: tools include uapi x86: Add __NR_setns, if missing To help us provide a simple setns() in older distros. Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Krister Johansen Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-o10a85kf6j7ig87ep6crab2k@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/arch/x86/include/asm/unistd_32.h | 3 +++ tools/arch/x86/include/asm/unistd_64.h | 3 +++ 2 files changed, 6 insertions(+) (limited to 'tools') diff --git a/tools/arch/x86/include/asm/unistd_32.h b/tools/arch/x86/include/asm/unistd_32.h index 88b3f8c..0e4312ff 100644 --- a/tools/arch/x86/include/asm/unistd_32.h +++ b/tools/arch/x86/include/asm/unistd_32.h @@ -10,3 +10,6 @@ #ifndef __NR_getcpu # define __NR_getcpu 318 #endif +#ifndef __NR_setns +# define __NR_setns 346 +#endif diff --git a/tools/arch/x86/include/asm/unistd_64.h b/tools/arch/x86/include/asm/unistd_64.h index fbdb70e..dd56bb3 100644 --- a/tools/arch/x86/include/asm/unistd_64.h +++ b/tools/arch/x86/include/asm/unistd_64.h @@ -10,3 +10,6 @@ #ifndef __NR_getcpu # define __NR_getcpu 309 #endif +#ifndef __NR_setns +#define __NR_setns 308 +#endif -- cgit v1.1 From 86bcdb5a43997bb02ba25a76482c7bfc652ba45b Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Tue, 18 Jul 2017 17:15:29 -0300 Subject: tools build: Add test for setns() And provide an alternative implementation to keep perf building on older distros as we're about to add initial support for namespaces. Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Krister Johansen Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-bqdwijunhjlvps1ardykhw1i@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/build/Makefile.feature | 3 ++- tools/build/feature/Makefile | 6 +++++- tools/build/feature/test-all.c | 5 +++++ tools/build/feature/test-setns.c | 7 +++++++ tools/perf/Makefile.config | 5 +++++ tools/perf/util/Build | 4 ++++ tools/perf/util/setns.c | 8 ++++++++ tools/perf/util/util.h | 4 ++++ 8 files changed, 40 insertions(+), 2 deletions(-) create mode 100644 tools/build/feature/test-setns.c create mode 100644 tools/perf/util/setns.c (limited to 'tools') diff --git a/tools/build/Makefile.feature b/tools/build/Makefile.feature index 523911f..c71a05b 100644 --- a/tools/build/Makefile.feature +++ b/tools/build/Makefile.feature @@ -64,7 +64,8 @@ FEATURE_TESTS_BASIC := \ get_cpuid \ bpf \ sched_getcpu \ - sdt + sdt \ + setns # FEATURE_TESTS_BASIC + FEATURE_TESTS_EXTRA is the complete list # of all feature tests diff --git a/tools/build/feature/Makefile b/tools/build/feature/Makefile index e35e4e5..ee2546d 100644 --- a/tools/build/feature/Makefile +++ b/tools/build/feature/Makefile @@ -49,7 +49,8 @@ FILES= \ test-sdt.bin \ test-cxx.bin \ test-jvmti.bin \ - test-sched_getcpu.bin + test-sched_getcpu.bin \ + test-setns.bin FILES := $(addprefix $(OUTPUT),$(FILES)) @@ -95,6 +96,9 @@ $(OUTPUT)test-glibc.bin: $(OUTPUT)test-sched_getcpu.bin: $(BUILD) +$(OUTPUT)test-setns.bin: + $(BUILD) + DWARFLIBS := -ldw ifeq ($(findstring -static,${LDFLAGS}),-static) DWARFLIBS += -lelf -lebl -lz -llzma -lbz2 diff --git a/tools/build/feature/test-all.c b/tools/build/feature/test-all.c index cc6c7c0..b5cfc64 100644 --- a/tools/build/feature/test-all.c +++ b/tools/build/feature/test-all.c @@ -153,6 +153,10 @@ # include "test-sdt.c" #undef main +#define main main_test_setns +# include "test-setns.c" +#undef main + int main(int argc, char *argv[]) { main_test_libpython(); @@ -188,6 +192,7 @@ int main(int argc, char *argv[]) main_test_libcrypto(); main_test_sched_getcpu(); main_test_sdt(); + main_test_setns(); return 0; } diff --git a/tools/build/feature/test-setns.c b/tools/build/feature/test-setns.c new file mode 100644 index 0000000..1f714d2 --- /dev/null +++ b/tools/build/feature/test-setns.c @@ -0,0 +1,7 @@ +#define _GNU_SOURCE +#include + +int main(void) +{ + return setns(0, 0); +} diff --git a/tools/perf/Makefile.config b/tools/perf/Makefile.config index bdf0e87..37d203c 100644 --- a/tools/perf/Makefile.config +++ b/tools/perf/Makefile.config @@ -330,6 +330,11 @@ ifeq ($(feature-sched_getcpu), 1) CFLAGS += -DHAVE_SCHED_GETCPU_SUPPORT endif +ifeq ($(feature-setns), 1) + CFLAGS += -DHAVE_SETNS_SUPPORT + $(call detected,CONFIG_SETNS) +endif + ifndef NO_LIBELF CFLAGS += -DHAVE_LIBELF_SUPPORT EXTLIBS += -lelf diff --git a/tools/perf/util/Build b/tools/perf/util/Build index 79dea95..7580fe4 100644 --- a/tools/perf/util/Build +++ b/tools/perf/util/Build @@ -104,6 +104,10 @@ ifndef CONFIG_LIBELF libperf-y += symbol-minimal.o endif +ifndef CONFIG_SETNS +libperf-y += setns.o +endif + libperf-$(CONFIG_DWARF) += probe-finder.o libperf-$(CONFIG_DWARF) += dwarf-aux.o libperf-$(CONFIG_DWARF) += dwarf-regs.o diff --git a/tools/perf/util/setns.c b/tools/perf/util/setns.c new file mode 100644 index 0000000..ce8fc29 --- /dev/null +++ b/tools/perf/util/setns.c @@ -0,0 +1,8 @@ +#include "util.h" +#include +#include + +int setns(int fd, int nstype) +{ + return syscall(__NR_setns, fd, nstype); +} diff --git a/tools/perf/util/util.h b/tools/perf/util/util.h index 2c9e58a..1e5fe1d 100644 --- a/tools/perf/util/util.h +++ b/tools/perf/util/util.h @@ -58,4 +58,8 @@ const char *perf_tip(const char *dirpath); int sched_getcpu(void); #endif +#ifndef HAVE_SETNS_SUPPORT +int setns(int fd, int nstype); +#endif + #endif /* GIT_COMPAT_UTIL_H */ -- cgit v1.1 From 843ff37bb59edbe51d64e77ba1b3245a15a4dd9f Mon Sep 17 00:00:00 2001 From: Krister Johansen Date: Wed, 5 Jul 2017 18:48:08 -0700 Subject: perf symbols: Find symbols in different mount namespace Teach perf how to resolve symbols from binaries that are in a different mount namespace from the tool. This allows perf to generate meaningful stack traces even if the binary resides in a different mount namespace from the tool. Signed-off-by: Krister Johansen Tested-by: Brendan Gregg Cc: Alexander Shishkin Cc: Peter Zijlstra Cc: Thomas-Mich Richter Link: http://lkml.kernel.org/r/1499305693-1599-2-git-send-email-kjlx@templeofstupid.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/dso.c | 1 + tools/perf/util/dso.h | 2 + tools/perf/util/map.c | 2 + tools/perf/util/namespaces.c | 127 +++++++++++++++++++++++++++++++++++++++++++ tools/perf/util/namespaces.h | 33 +++++++++++ tools/perf/util/symbol.c | 11 ++++ tools/perf/util/thread.c | 3 + tools/perf/util/thread.h | 1 + 8 files changed, 180 insertions(+) (limited to 'tools') diff --git a/tools/perf/util/dso.c b/tools/perf/util/dso.c index 4e7ab61..beda40e 100644 --- a/tools/perf/util/dso.c +++ b/tools/perf/util/dso.c @@ -1236,6 +1236,7 @@ void dso__delete(struct dso *dso) dso_cache__free(dso); dso__free_a2l(dso); zfree(&dso->symsrc_filename); + nsinfo__zput(dso->nsinfo); pthread_mutex_destroy(&dso->lock); free(dso); } diff --git a/tools/perf/util/dso.h b/tools/perf/util/dso.h index bd061ba..78ec637 100644 --- a/tools/perf/util/dso.h +++ b/tools/perf/util/dso.h @@ -10,6 +10,7 @@ #include #include #include "map.h" +#include "namespaces.h" #include "build-id.h" enum dso_binary_type { @@ -187,6 +188,7 @@ struct dso { void *priv; u64 db_id; }; + struct nsinfo *nsinfo; refcount_t refcnt; char name[0]; }; diff --git a/tools/perf/util/map.c b/tools/perf/util/map.c index 2179b2d..5dc60ca 100644 --- a/tools/perf/util/map.c +++ b/tools/perf/util/map.c @@ -16,6 +16,7 @@ #include "machine.h" #include #include "srcline.h" +#include "namespaces.h" #include "unwind.h" static void __maps__insert(struct maps *maps, struct map *map); @@ -200,6 +201,7 @@ struct map *map__new(struct machine *machine, u64 start, u64 len, if (type != MAP__FUNCTION) dso__set_loaded(dso, map->type); } + dso->nsinfo = nsinfo__get(thread->nsinfo); dso__put(dso); } return map; diff --git a/tools/perf/util/namespaces.c b/tools/perf/util/namespaces.c index 67dcbcc..bcc6bb1 100644 --- a/tools/perf/util/namespaces.c +++ b/tools/perf/util/namespaces.c @@ -9,9 +9,13 @@ #include "namespaces.h" #include "util.h" #include "event.h" +#include +#include +#include #include #include #include +#include struct namespaces *namespaces__new(struct namespaces_event *event) { @@ -35,3 +39,126 @@ void namespaces__free(struct namespaces *namespaces) { free(namespaces); } + +void nsinfo__init(struct nsinfo *nsi) +{ + char oldns[PATH_MAX]; + char *newns = NULL; + struct stat old_stat; + struct stat new_stat; + + if (snprintf(oldns, PATH_MAX, "/proc/self/ns/mnt") >= PATH_MAX) + return; + + if (asprintf(&newns, "/proc/%d/ns/mnt", nsi->pid) == -1) + return; + + if (stat(oldns, &old_stat) < 0) + goto out; + + if (stat(newns, &new_stat) < 0) + goto out; + + /* Check if the mount namespaces differ, if so then indicate that we + * want to switch as part of looking up dso/map data. + */ + if (old_stat.st_ino != new_stat.st_ino) { + nsi->need_setns = true; + nsi->mntns_path = newns; + newns = NULL; + } + +out: + free(newns); +} + +struct nsinfo *nsinfo__new(pid_t pid) +{ + struct nsinfo *nsi = calloc(1, sizeof(*nsi)); + + if (nsi != NULL) { + nsi->pid = pid; + nsi->need_setns = false; + nsinfo__init(nsi); + refcount_set(&nsi->refcnt, 1); + } + + return nsi; +} + +void nsinfo__delete(struct nsinfo *nsi) +{ + zfree(&nsi->mntns_path); + free(nsi); +} + +struct nsinfo *nsinfo__get(struct nsinfo *nsi) +{ + if (nsi) + refcount_inc(&nsi->refcnt); + return nsi; +} + +void nsinfo__put(struct nsinfo *nsi) +{ + if (nsi && refcount_dec_and_test(&nsi->refcnt)) + nsinfo__delete(nsi); +} + +void nsinfo__mountns_enter(struct nsinfo *nsi, struct nscookie *nc) +{ + char curpath[PATH_MAX]; + int oldns = -1; + int newns = -1; + + if (nc == NULL) + return; + + nc->oldns = -1; + nc->newns = -1; + + if (!nsi || !nsi->need_setns) + return; + + if (snprintf(curpath, PATH_MAX, "/proc/self/ns/mnt") >= PATH_MAX) + return; + + oldns = open(curpath, O_RDONLY); + if (oldns < 0) + return; + + newns = open(nsi->mntns_path, O_RDONLY); + if (newns < 0) + goto errout; + + if (setns(newns, CLONE_NEWNS) < 0) + goto errout; + + nc->oldns = oldns; + nc->newns = newns; + return; + +errout: + if (oldns > -1) + close(oldns); + if (newns > -1) + close(newns); +} + +void nsinfo__mountns_exit(struct nscookie *nc) +{ + if (nc == NULL || nc->oldns == -1 || nc->newns == -1) + return; + + setns(nc->oldns, CLONE_NEWNS); + + if (nc->oldns > -1) { + close(nc->oldns); + nc->oldns = -1; + } + + if (nc->newns > -1) { + close(nc->newns); + nc->newns = -1; + } +} diff --git a/tools/perf/util/namespaces.h b/tools/perf/util/namespaces.h index 468f1e9..b20f6ea 100644 --- a/tools/perf/util/namespaces.h +++ b/tools/perf/util/namespaces.h @@ -11,6 +11,7 @@ #include "../perf.h" #include +#include struct namespaces_event; @@ -23,4 +24,36 @@ struct namespaces { struct namespaces *namespaces__new(struct namespaces_event *event); void namespaces__free(struct namespaces *namespaces); +struct nsinfo { + pid_t pid; + bool need_setns; + char *mntns_path; + refcount_t refcnt; +}; + +struct nscookie { + int oldns; + int newns; +}; + +void nsinfo__init(struct nsinfo *nsi); +struct nsinfo *nsinfo__new(pid_t pid); +void nsinfo__delete(struct nsinfo *nsi); + +struct nsinfo *nsinfo__get(struct nsinfo *nsi); +void nsinfo__put(struct nsinfo *nsi); + +void nsinfo__mountns_enter(struct nsinfo *nsi, struct nscookie *nc); +void nsinfo__mountns_exit(struct nscookie *nc); + +static inline void __nsinfo__zput(struct nsinfo **nsip) +{ + if (nsip) { + nsinfo__put(*nsip); + *nsip = NULL; + } +} + +#define nsinfo__zput(nsi) __nsinfo__zput(&nsi) + #endif /* __PERF_NAMESPACES_H */ diff --git a/tools/perf/util/symbol.c b/tools/perf/util/symbol.c index e7a98db..60a9eaa 100644 --- a/tools/perf/util/symbol.c +++ b/tools/perf/util/symbol.c @@ -18,6 +18,8 @@ #include "symbol.h" #include "strlist.h" #include "intlist.h" +#include "namespaces.h" +#include "vdso.h" #include "header.h" #include "path.h" #include "sane_ctype.h" @@ -1436,9 +1438,17 @@ int dso__load(struct dso *dso, struct map *map) struct symsrc *syms_ss = NULL, *runtime_ss = NULL; bool kmod; unsigned char build_id[BUILD_ID_SIZE]; + struct nscookie nsc; + nsinfo__mountns_enter(dso->nsinfo, &nsc); pthread_mutex_lock(&dso->lock); + /* The vdso files always live in the host container, so don't go looking + * for them in the container's mount namespace. + */ + if (dso__is_vdso(dso)) + nsinfo__mountns_exit(&nsc); + /* check again under the dso->lock */ if (dso__loaded(dso, map->type)) { ret = 1; @@ -1584,6 +1594,7 @@ out_free: out: dso__set_loaded(dso, map->type); pthread_mutex_unlock(&dso->lock); + nsinfo__mountns_exit(&nsc); return ret; } diff --git a/tools/perf/util/thread.c b/tools/perf/util/thread.c index 378c418..aee9a42 100644 --- a/tools/perf/util/thread.c +++ b/tools/perf/util/thread.c @@ -59,6 +59,8 @@ struct thread *thread__new(pid_t pid, pid_t tid) list_add(&comm->list, &thread->comm_list); refcount_set(&thread->refcnt, 1); RB_CLEAR_NODE(&thread->rb_node); + /* Thread holds first ref to nsdata. */ + thread->nsinfo = nsinfo__new(pid); } return thread; @@ -91,6 +93,7 @@ void thread__delete(struct thread *thread) comm__free(comm); } unwind__finish_access(thread); + nsinfo__zput(thread->nsinfo); free(thread); } diff --git a/tools/perf/util/thread.h b/tools/perf/util/thread.h index 4eb849e..cb1a5dd 100644 --- a/tools/perf/util/thread.h +++ b/tools/perf/util/thread.h @@ -34,6 +34,7 @@ struct thread { void *priv; struct thread_stack *ts; + struct nsinfo *nsinfo; #ifdef HAVE_LIBUNWIND_SUPPORT void *addr_space; struct unwind_libunwind_ops *unwind_libunwind_ops; -- cgit v1.1 From bf2e710b3cb8445c052f2ff50b4875a2523bb279 Mon Sep 17 00:00:00 2001 From: Krister Johansen Date: Wed, 5 Jul 2017 18:48:09 -0700 Subject: perf maps: Lookup maps in both intitial mountns and inner mountns. If a process is in a mountns and has symbols in /tmp/perf-.map, look first in the namespace using the tgid for the pidns that the process might be in. If the map isn't found there, try looking in the mountns where perf is running, and use the tgid that's appropriate for perf's pid namespace. If all else fails, use the original pid. This allows us to locate a symbol map file in the mount namespace, if it was generated there. However, we also try the tool's /tmp in case it's there instead. Signed-off-by: Krister Johansen Tested-by: Brendan Gregg Cc: Alexander Shishkin Cc: Peter Zijlstra Cc: Thomas-Mich Richter Link: http://lkml.kernel.org/r/1499305693-1599-3-git-send-email-kjlx@templeofstupid.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/machine.c | 4 +-- tools/perf/util/map.c | 23 +++++++++--- tools/perf/util/map.h | 2 +- tools/perf/util/namespaces.c | 83 ++++++++++++++++++++++++++++++++++++++++---- tools/perf/util/namespaces.h | 5 ++- tools/perf/util/symbol.c | 70 ++++++++++++++++++++++++++++++------- 6 files changed, 160 insertions(+), 27 deletions(-) (limited to 'tools') diff --git a/tools/perf/util/machine.c b/tools/perf/util/machine.c index 2e9eb6a..246b441 100644 --- a/tools/perf/util/machine.c +++ b/tools/perf/util/machine.c @@ -1392,7 +1392,7 @@ int machine__process_mmap2_event(struct machine *machine, map = map__new(machine, event->mmap2.start, event->mmap2.len, event->mmap2.pgoff, - event->mmap2.pid, event->mmap2.maj, + event->mmap2.maj, event->mmap2.min, event->mmap2.ino, event->mmap2.ino_generation, event->mmap2.prot, @@ -1450,7 +1450,7 @@ int machine__process_mmap_event(struct machine *machine, union perf_event *event map = map__new(machine, event->mmap.start, event->mmap.len, event->mmap.pgoff, - event->mmap.pid, 0, 0, 0, 0, 0, 0, + 0, 0, 0, 0, 0, 0, event->mmap.filename, type, thread); diff --git a/tools/perf/util/map.c b/tools/perf/util/map.c index 5dc60ca..bdaa0a4 100644 --- a/tools/perf/util/map.c +++ b/tools/perf/util/map.c @@ -146,11 +146,13 @@ void map__init(struct map *map, enum map_type type, } struct map *map__new(struct machine *machine, u64 start, u64 len, - u64 pgoff, u32 pid, u32 d_maj, u32 d_min, u64 ino, + u64 pgoff, u32 d_maj, u32 d_min, u64 ino, u64 ino_gen, u32 prot, u32 flags, char *filename, enum map_type type, struct thread *thread) { struct map *map = malloc(sizeof(*map)); + struct nsinfo *nsi = NULL; + struct nsinfo *nnsi; if (map != NULL) { char newfilename[PATH_MAX]; @@ -168,9 +170,11 @@ struct map *map__new(struct machine *machine, u64 start, u64 len, map->ino_generation = ino_gen; map->prot = prot; map->flags = flags; + nsi = nsinfo__get(thread->nsinfo); - if ((anon || no_dso) && type == MAP__FUNCTION) { - snprintf(newfilename, sizeof(newfilename), "/tmp/perf-%d.map", pid); + if ((anon || no_dso) && nsi && type == MAP__FUNCTION) { + snprintf(newfilename, sizeof(newfilename), + "/tmp/perf-%d.map", nsi->pid); filename = newfilename; } @@ -180,6 +184,16 @@ struct map *map__new(struct machine *machine, u64 start, u64 len, } if (vdso) { + /* The vdso maps are always on the host and not the + * container. Ensure that we don't use setns to look + * them up. + */ + nnsi = nsinfo__copy(nsi); + if (nnsi) { + nsinfo__put(nsi); + nnsi->need_setns = false; + nsi = nnsi; + } pgoff = 0; dso = machine__findnew_vdso(machine, thread); } else @@ -201,11 +215,12 @@ struct map *map__new(struct machine *machine, u64 start, u64 len, if (type != MAP__FUNCTION) dso__set_loaded(dso, map->type); } - dso->nsinfo = nsinfo__get(thread->nsinfo); + dso->nsinfo = nsi; dso__put(dso); } return map; out_delete: + nsinfo__put(nsi); free(map); return NULL; } diff --git a/tools/perf/util/map.h b/tools/perf/util/map.h index f9e8ac8..73aacf7 100644 --- a/tools/perf/util/map.h +++ b/tools/perf/util/map.h @@ -141,7 +141,7 @@ struct thread; void map__init(struct map *map, enum map_type type, u64 start, u64 end, u64 pgoff, struct dso *dso); struct map *map__new(struct machine *machine, u64 start, u64 len, - u64 pgoff, u32 pid, u32 d_maj, u32 d_min, u64 ino, + u64 pgoff, u32 d_maj, u32 d_min, u64 ino, u64 ino_gen, u32 prot, u32 flags, char *filename, enum map_type type, struct thread *thread); struct map *map__new2(u64 start, struct dso *dso, enum map_type type); diff --git a/tools/perf/util/namespaces.c b/tools/perf/util/namespaces.c index bcc6bb1..fc5f398 100644 --- a/tools/perf/util/namespaces.c +++ b/tools/perf/util/namespaces.c @@ -40,18 +40,23 @@ void namespaces__free(struct namespaces *namespaces) free(namespaces); } -void nsinfo__init(struct nsinfo *nsi) +int nsinfo__init(struct nsinfo *nsi) { char oldns[PATH_MAX]; + char spath[PATH_MAX]; char *newns = NULL; + char *statln = NULL; struct stat old_stat; struct stat new_stat; + FILE *f = NULL; + size_t linesz = 0; + int rv = -1; if (snprintf(oldns, PATH_MAX, "/proc/self/ns/mnt") >= PATH_MAX) - return; + return rv; if (asprintf(&newns, "/proc/%d/ns/mnt", nsi->pid) == -1) - return; + return rv; if (stat(oldns, &old_stat) < 0) goto out; @@ -68,24 +73,89 @@ void nsinfo__init(struct nsinfo *nsi) newns = NULL; } + /* If we're dealing with a process that is in a different PID namespace, + * attempt to work out the innermost tgid for the process. + */ + if (snprintf(spath, PATH_MAX, "/proc/%d/status", nsi->pid) >= PATH_MAX) + goto out; + + f = fopen(spath, "r"); + if (f == NULL) + goto out; + + while (getline(&statln, &linesz, f) != -1) { + /* Use tgid if CONFIG_PID_NS is not defined. */ + if (strstr(statln, "Tgid:") != NULL) { + nsi->tgid = (pid_t)strtol(strrchr(statln, '\t'), + NULL, 10); + nsi->nstgid = nsi->tgid; + } + + if (strstr(statln, "NStgid:") != NULL) { + nsi->nstgid = (pid_t)strtol(strrchr(statln, '\t'), + NULL, 10); + break; + } + } + rv = 0; + out: + if (f != NULL) + (void) fclose(f); + free(statln); free(newns); + return rv; } struct nsinfo *nsinfo__new(pid_t pid) { - struct nsinfo *nsi = calloc(1, sizeof(*nsi)); + struct nsinfo *nsi; + if (pid == 0) + return NULL; + + nsi = calloc(1, sizeof(*nsi)); if (nsi != NULL) { nsi->pid = pid; + nsi->tgid = pid; + nsi->nstgid = pid; nsi->need_setns = false; - nsinfo__init(nsi); + /* Init may fail if the process exits while we're trying to look + * at its proc information. In that case, save the pid but + * don't try to enter the namespace. + */ + if (nsinfo__init(nsi) == -1) + nsi->need_setns = false; + refcount_set(&nsi->refcnt, 1); } return nsi; } +struct nsinfo *nsinfo__copy(struct nsinfo *nsi) +{ + struct nsinfo *nnsi; + + nnsi = calloc(1, sizeof(*nnsi)); + if (nnsi != NULL) { + nnsi->pid = nsi->pid; + nnsi->tgid = nsi->tgid; + nnsi->nstgid = nsi->nstgid; + nnsi->need_setns = nsi->need_setns; + if (nsi->mntns_path) { + nnsi->mntns_path = strdup(nsi->mntns_path); + if (!nnsi->mntns_path) { + free(nnsi); + return NULL; + } + } + refcount_set(&nnsi->refcnt, 1); + } + + return nnsi; +} + void nsinfo__delete(struct nsinfo *nsi) { zfree(&nsi->mntns_path); @@ -105,7 +175,8 @@ void nsinfo__put(struct nsinfo *nsi) nsinfo__delete(nsi); } -void nsinfo__mountns_enter(struct nsinfo *nsi, struct nscookie *nc) +void nsinfo__mountns_enter(struct nsinfo *nsi, + struct nscookie *nc) { char curpath[PATH_MAX]; int oldns = -1; diff --git a/tools/perf/util/namespaces.h b/tools/perf/util/namespaces.h index b20f6ea..f19aa41 100644 --- a/tools/perf/util/namespaces.h +++ b/tools/perf/util/namespaces.h @@ -26,6 +26,8 @@ void namespaces__free(struct namespaces *namespaces); struct nsinfo { pid_t pid; + pid_t tgid; + pid_t nstgid; bool need_setns; char *mntns_path; refcount_t refcnt; @@ -36,8 +38,9 @@ struct nscookie { int newns; }; -void nsinfo__init(struct nsinfo *nsi); +int nsinfo__init(struct nsinfo *nsi); struct nsinfo *nsinfo__new(pid_t pid); +struct nsinfo *nsinfo__copy(struct nsinfo *nsi); void nsinfo__delete(struct nsinfo *nsi); struct nsinfo *nsinfo__get(struct nsinfo *nsi); diff --git a/tools/perf/util/symbol.c b/tools/perf/util/symbol.c index 60a9eaa..21c97cc 100644 --- a/tools/perf/util/symbol.c +++ b/tools/perf/util/symbol.c @@ -19,7 +19,6 @@ #include "strlist.h" #include "intlist.h" #include "namespaces.h" -#include "vdso.h" #include "header.h" #include "path.h" #include "sane_ctype.h" @@ -1327,14 +1326,15 @@ int dso__load_kallsyms(struct dso *dso, const char *filename, return __dso__load_kallsyms(dso, filename, map, false); } -static int dso__load_perf_map(struct dso *dso, struct map *map) +static int dso__load_perf_map(const char *map_path, struct dso *dso, + struct map *map) { char *line = NULL; size_t n; FILE *file; int nr_syms = 0; - file = fopen(dso->long_name, "r"); + file = fopen(map_path, "r"); if (file == NULL) goto out_failure; @@ -1426,6 +1426,45 @@ static bool dso__is_compatible_symtab_type(struct dso *dso, bool kmod, } } +/* Checks for the existence of the perf-.map file in two different + * locations. First, if the process is a separate mount namespace, check in + * that namespace using the pid of the innermost pid namespace. If's not in a + * namespace, or the file can't be found there, try in the mount namespace of + * the tracing process using our view of its pid. + */ +static int dso__find_perf_map(char *filebuf, size_t bufsz, + struct nsinfo **nsip) +{ + struct nscookie nsc; + struct nsinfo *nsi; + struct nsinfo *nnsi; + int rc = -1; + + nsi = *nsip; + + if (nsi->need_setns) { + snprintf(filebuf, bufsz, "/tmp/perf-%d.map", nsi->nstgid); + nsinfo__mountns_enter(nsi, &nsc); + rc = access(filebuf, R_OK); + nsinfo__mountns_exit(&nsc); + if (rc == 0) + return rc; + } + + nnsi = nsinfo__copy(nsi); + if (nnsi) { + nsinfo__put(nsi); + + nnsi->need_setns = false; + snprintf(filebuf, bufsz, "/tmp/perf-%d.map", nnsi->tgid); + *nsip = nnsi; + rc = 0; + } + + return rc; +} + + int dso__load(struct dso *dso, struct map *map) { char *name; @@ -1437,18 +1476,23 @@ int dso__load(struct dso *dso, struct map *map) struct symsrc ss_[2]; struct symsrc *syms_ss = NULL, *runtime_ss = NULL; bool kmod; + bool perfmap; unsigned char build_id[BUILD_ID_SIZE]; struct nscookie nsc; + char newmapname[PATH_MAX]; + const char *map_path = dso->long_name; + + perfmap = strncmp(dso->name, "/tmp/perf-", 10) == 0; + if (perfmap) { + if (dso->nsinfo && (dso__find_perf_map(newmapname, + sizeof(newmapname), &dso->nsinfo) == 0)) { + map_path = newmapname; + } + } nsinfo__mountns_enter(dso->nsinfo, &nsc); pthread_mutex_lock(&dso->lock); - /* The vdso files always live in the host container, so don't go looking - * for them in the container's mount namespace. - */ - if (dso__is_vdso(dso)) - nsinfo__mountns_exit(&nsc); - /* check again under the dso->lock */ if (dso__loaded(dso, map->type)) { ret = 1; @@ -1471,19 +1515,19 @@ int dso__load(struct dso *dso, struct map *map) dso->adjust_symbols = 0; - if (strncmp(dso->name, "/tmp/perf-", 10) == 0) { + if (perfmap) { struct stat st; - if (lstat(dso->name, &st) < 0) + if (lstat(map_path, &st) < 0) goto out; if (!symbol_conf.force && st.st_uid && (st.st_uid != geteuid())) { pr_warning("File %s not owned by current user or root, " - "ignoring it (use -f to override).\n", dso->name); + "ignoring it (use -f to override).\n", map_path); goto out; } - ret = dso__load_perf_map(dso, map); + ret = dso__load_perf_map(map_path, dso, map); dso->symtab_type = ret > 0 ? DSO_BINARY_TYPE__JAVA_JIT : DSO_BINARY_TYPE__NOT_FOUND; goto out; -- cgit v1.1 From 544abd44c7064c8a58a6bd2073d757f6b91d98c5 Mon Sep 17 00:00:00 2001 From: Krister Johansen Date: Wed, 5 Jul 2017 18:48:10 -0700 Subject: perf probe: Allow placing uprobes in alternate namespaces. Teaches perf how to place a uprobe on a file that's in a different mount namespace. The user must add the probe using the --target-ns argument to perf probe. Once it has been placed, it may be recorded against without further namespace-specific commands. Signed-off-by: Krister Johansen Cc: Alexander Shishkin Cc: Brendan Gregg Cc: Peter Zijlstra Cc: Ravi Bangoria [ PPC build fixed by Ravi: ] Link: http://lkml.kernel.org/r/1500287542-6219-1-git-send-email-ravi.bangoria@linux.vnet.ibm.com Cc: Thomas-Mich Richter [ Fix !HAVE_DWARF_SUPPORT build ] Link: http://lkml.kernel.org/r/1499305693-1599-4-git-send-email-kjlx@templeofstupid.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/Documentation/perf-probe.txt | 9 ++++ tools/perf/arch/powerpc/util/sym-handling.c | 2 +- tools/perf/builtin-probe.c | 43 ++++++++++++++-- tools/perf/util/namespaces.c | 13 +++++ tools/perf/util/namespaces.h | 2 + tools/perf/util/probe-event.c | 80 +++++++++++++++++++---------- tools/perf/util/probe-event.h | 10 ++-- 7 files changed, 125 insertions(+), 34 deletions(-) (limited to 'tools') diff --git a/tools/perf/Documentation/perf-probe.txt b/tools/perf/Documentation/perf-probe.txt index 165c2b1..a42aabc 100644 --- a/tools/perf/Documentation/perf-probe.txt +++ b/tools/perf/Documentation/perf-probe.txt @@ -130,6 +130,11 @@ OPTIONS --max-probes=NUM:: Set the maximum number of probe points for an event. Default is 128. +--target-ns=PID: + Obtain mount namespace information from the target pid. This is + used when creating a uprobe for a process that resides in a + different mount namespace from the perf(1) utility. + -x:: --exec=PATH:: Specify path to the executable or shared library file for user @@ -264,6 +269,10 @@ Add probes at malloc() function on libc ./perf probe -x /lib/libc.so.6 malloc or ./perf probe /lib/libc.so.6 malloc +Add a uprobe to a target process running in a different mount namespace + + ./perf probe --target-ns -x /lib64/libc.so.6 malloc + SEE ALSO -------- linkperf:perf-trace[1], linkperf:perf-record[1], linkperf:perf-buildid-cache[1] diff --git a/tools/perf/arch/powerpc/util/sym-handling.c b/tools/perf/arch/powerpc/util/sym-handling.c index bf9a259..9c4e23d 100644 --- a/tools/perf/arch/powerpc/util/sym-handling.c +++ b/tools/perf/arch/powerpc/util/sym-handling.c @@ -126,7 +126,7 @@ void arch__post_process_probe_trace_events(struct perf_probe_event *pev, struct rb_node *tmp; int i = 0; - map = get_target_map(pev->target, pev->uprobes); + map = get_target_map(pev->target, pev->nsi, pev->uprobes); if (!map || map__load(map) < 0) return; diff --git a/tools/perf/builtin-probe.c b/tools/perf/builtin-probe.c index cf9f9e9..3fb98d5 100644 --- a/tools/perf/builtin-probe.c +++ b/tools/perf/builtin-probe.c @@ -58,6 +58,7 @@ static struct { struct line_range line_range; char *target; struct strfilter *filter; + struct nsinfo *nsi; } params; /* Parse an event definition. Note that any error must die. */ @@ -80,6 +81,9 @@ static int parse_probe_event(const char *str) params.target_used = true; } + if (params.nsi) + pev->nsi = nsinfo__get(params.nsi); + /* Parse a perf-probe command into event */ ret = parse_perf_probe_command(str, pev); pr_debug("%d arguments\n", pev->nargs); @@ -189,7 +193,7 @@ static int opt_set_target(const struct option *opt, const char *str, /* Expand given path to absolute path, except for modulename */ if (params.uprobes || strchr(str, '/')) { - tmp = realpath(str, NULL); + tmp = nsinfo__realpath(str, params.nsi); if (!tmp) { pr_warning("Failed to get the absolute path of %s: %m\n", str); return ret; @@ -208,6 +212,34 @@ static int opt_set_target(const struct option *opt, const char *str, return ret; } +static int opt_set_target_ns(const struct option *opt __maybe_unused, + const char *str, int unset __maybe_unused) +{ + int ret = -ENOENT; + pid_t ns_pid; + struct nsinfo *nsip; + + if (str) { + errno = 0; + ns_pid = (pid_t)strtol(str, NULL, 10); + if (errno != 0) { + ret = -errno; + pr_warning("Failed to parse %s as a pid: %s\n", str, + strerror(errno)); + return ret; + } + nsip = nsinfo__new(ns_pid); + if (nsip && nsip->need_setns) + params.nsi = nsinfo__get(nsip); + nsinfo__put(nsip); + + ret = 0; + } + + return ret; +} + + /* Command option callbacks */ #ifdef HAVE_DWARF_SUPPORT @@ -299,6 +331,7 @@ static void cleanup_params(void) line_range__clear(¶ms.line_range); free(params.target); strfilter__delete(params.filter); + nsinfo__put(params.nsi); memset(¶ms, 0, sizeof(params)); } @@ -554,6 +587,8 @@ __cmd_probe(int argc, const char **argv) OPT_BOOLEAN(0, "cache", &probe_conf.cache, "Manipulate probe cache"), OPT_STRING(0, "symfs", &symbol_conf.symfs, "directory", "Look for files with symbols relative to this directory"), + OPT_CALLBACK(0, "target-ns", NULL, "pid", + "target pid for namespace contexts", opt_set_target_ns), OPT_END() }; int ret; @@ -634,15 +669,15 @@ __cmd_probe(int argc, const char **argv) pr_err_with_code(" Error: Failed to show event list.", ret); return ret; case 'F': - ret = show_available_funcs(params.target, params.filter, - params.uprobes); + ret = show_available_funcs(params.target, params.nsi, + params.filter, params.uprobes); if (ret < 0) pr_err_with_code(" Error: Failed to show functions.", ret); return ret; #ifdef HAVE_DWARF_SUPPORT case 'L': ret = show_line_range(¶ms.line_range, params.target, - params.uprobes); + params.nsi, params.uprobes); if (ret < 0) pr_err_with_code(" Error: Failed to show lines.", ret); return ret; diff --git a/tools/perf/util/namespaces.c b/tools/perf/util/namespaces.c index fc5f398..a58e911 100644 --- a/tools/perf/util/namespaces.c +++ b/tools/perf/util/namespaces.c @@ -11,6 +11,7 @@ #include "event.h" #include #include +#include #include #include #include @@ -233,3 +234,15 @@ void nsinfo__mountns_exit(struct nscookie *nc) nc->newns = -1; } } + +char *nsinfo__realpath(const char *path, struct nsinfo *nsi) +{ + char *rpath; + struct nscookie nsc; + + nsinfo__mountns_enter(nsi, &nsc); + rpath = realpath(path, NULL); + nsinfo__mountns_exit(&nsc); + + return rpath; +} diff --git a/tools/perf/util/namespaces.h b/tools/perf/util/namespaces.h index f19aa41..05d8260 100644 --- a/tools/perf/util/namespaces.h +++ b/tools/perf/util/namespaces.h @@ -49,6 +49,8 @@ void nsinfo__put(struct nsinfo *nsi); void nsinfo__mountns_enter(struct nsinfo *nsi, struct nscookie *nc); void nsinfo__mountns_exit(struct nscookie *nc); +char *nsinfo__realpath(const char *path, struct nsinfo *nsi); + static inline void __nsinfo__zput(struct nsinfo **nsip) { if (nsip) { diff --git a/tools/perf/util/probe-event.c b/tools/perf/util/probe-event.c index a2670e9..a80895a 100644 --- a/tools/perf/util/probe-event.c +++ b/tools/perf/util/probe-event.c @@ -184,13 +184,19 @@ static struct map *kernel_get_module_map(const char *module) return NULL; } -struct map *get_target_map(const char *target, bool user) +struct map *get_target_map(const char *target, struct nsinfo *nsi, bool user) { /* Init maps of given executable or kernel */ - if (user) - return dso__new_map(target); - else + if (user) { + struct map *map; + + map = dso__new_map(target); + if (map && map->dso) + map->dso->nsinfo = nsinfo__get(nsi); + return map; + } else { return kernel_get_module_map(target); + } } static int convert_exec_to_group(const char *exec, char **result) @@ -366,7 +372,8 @@ found: static int find_alternative_probe_point(struct debuginfo *dinfo, struct perf_probe_point *pp, struct perf_probe_point *result, - const char *target, bool uprobes) + const char *target, struct nsinfo *nsi, + bool uprobes) { struct map *map = NULL; struct symbol *sym; @@ -377,7 +384,7 @@ static int find_alternative_probe_point(struct debuginfo *dinfo, if (!pp->function || pp->file) return -ENOTSUP; - map = get_target_map(target, uprobes); + map = get_target_map(target, nsi, uprobes); if (!map) return -EINVAL; @@ -421,8 +428,8 @@ static int get_alternative_probe_event(struct debuginfo *dinfo, memcpy(tmp, &pev->point, sizeof(*tmp)); memset(&pev->point, 0, sizeof(pev->point)); - ret = find_alternative_probe_point(dinfo, tmp, &pev->point, - pev->target, pev->uprobes); + ret = find_alternative_probe_point(dinfo, tmp, &pev->point, pev->target, + pev->nsi, pev->uprobes); if (ret < 0) memcpy(&pev->point, tmp, sizeof(*tmp)); @@ -444,7 +451,7 @@ static int get_alternative_line_range(struct debuginfo *dinfo, if (lr->end != INT_MAX) len = lr->end - lr->start; ret = find_alternative_probe_point(dinfo, &pp, &result, - target, user); + target, NULL, user); if (!ret) { lr->function = result.function; lr->file = result.file; @@ -457,12 +464,14 @@ static int get_alternative_line_range(struct debuginfo *dinfo, } /* Open new debuginfo of given module */ -static struct debuginfo *open_debuginfo(const char *module, bool silent) +static struct debuginfo *open_debuginfo(const char *module, struct nsinfo *nsi, + bool silent) { const char *path = module; char reason[STRERR_BUFSIZE]; struct debuginfo *ret = NULL; struct dso *dso = NULL; + struct nscookie nsc; int err; if (!module || !strchr(module, '/')) { @@ -480,6 +489,7 @@ static struct debuginfo *open_debuginfo(const char *module, bool silent) } path = dso->long_name; } + nsinfo__mountns_enter(nsi, &nsc); ret = debuginfo__new(path); if (!ret && !silent) { pr_warning("The %s file has no debug information.\n", path); @@ -489,6 +499,7 @@ static struct debuginfo *open_debuginfo(const char *module, bool silent) pr_warning("Rebuild with -g, "); pr_warning("or install an appropriate debuginfo package.\n"); } + nsinfo__mountns_exit(&nsc); return ret; } @@ -516,7 +527,7 @@ static struct debuginfo *debuginfo_cache__open(const char *module, bool silent) goto out; } - debuginfo_cache = open_debuginfo(module, silent); + debuginfo_cache = open_debuginfo(module, NULL, silent); if (!debuginfo_cache) zfree(&debuginfo_cache_path); out: @@ -531,14 +542,18 @@ static void debuginfo_cache__exit(void) } -static int get_text_start_address(const char *exec, unsigned long *address) +static int get_text_start_address(const char *exec, unsigned long *address, + struct nsinfo *nsi) { Elf *elf; GElf_Ehdr ehdr; GElf_Shdr shdr; int fd, ret = -ENOENT; + struct nscookie nsc; + nsinfo__mountns_enter(nsi, &nsc); fd = open(exec, O_RDONLY); + nsinfo__mountns_exit(&nsc); if (fd < 0) return -errno; @@ -582,7 +597,7 @@ static int find_perf_probe_point_from_dwarf(struct probe_trace_point *tp, ret = -EINVAL; goto error; } - ret = get_text_start_address(tp->module, &stext); + ret = get_text_start_address(tp->module, &stext, NULL); if (ret < 0) goto error; addr += stext; @@ -659,7 +674,7 @@ post_process_offline_probe_trace_events(struct probe_trace_event *tevs, /* Prepare a map for offline binary */ map = dso__new_map(pathname); - if (!map || get_text_start_address(pathname, &stext) < 0) { + if (!map || get_text_start_address(pathname, &stext, NULL) < 0) { pr_warning("Failed to get ELF symbols for %s\n", pathname); return -EINVAL; } @@ -676,7 +691,8 @@ post_process_offline_probe_trace_events(struct probe_trace_event *tevs, } static int add_exec_to_probe_trace_events(struct probe_trace_event *tevs, - int ntevs, const char *exec) + int ntevs, const char *exec, + struct nsinfo *nsi) { int i, ret = 0; unsigned long stext = 0; @@ -684,7 +700,7 @@ static int add_exec_to_probe_trace_events(struct probe_trace_event *tevs, if (!exec) return 0; - ret = get_text_start_address(exec, &stext); + ret = get_text_start_address(exec, &stext, nsi); if (ret < 0) return ret; @@ -715,7 +731,7 @@ post_process_module_probe_trace_events(struct probe_trace_event *tevs, if (!module) return 0; - map = get_target_map(module, false); + map = get_target_map(module, NULL, false); if (!map || debuginfo__get_text_offset(dinfo, &text_offs, true) < 0) { pr_warning("Failed to get ELF symbols for %s\n", module); return -EINVAL; @@ -802,7 +818,8 @@ static int post_process_probe_trace_events(struct perf_probe_event *pev, int ret; if (uprobe) - ret = add_exec_to_probe_trace_events(tevs, ntevs, module); + ret = add_exec_to_probe_trace_events(tevs, ntevs, module, + pev->nsi); else if (module) /* Currently ref_reloc_sym based probe is not for drivers */ ret = post_process_module_probe_trace_events(tevs, ntevs, @@ -825,7 +842,7 @@ static int try_to_find_probe_trace_events(struct perf_probe_event *pev, struct debuginfo *dinfo; int ntevs, ret = 0; - dinfo = open_debuginfo(pev->target, !need_dwarf); + dinfo = open_debuginfo(pev->target, pev->nsi, !need_dwarf); if (!dinfo) { if (need_dwarf) return -ENOENT; @@ -945,7 +962,7 @@ static int __show_line_range(struct line_range *lr, const char *module, char sbuf[STRERR_BUFSIZE]; /* Search a line range */ - dinfo = open_debuginfo(module, false); + dinfo = open_debuginfo(module, NULL, false); if (!dinfo) return -ENOENT; @@ -1021,14 +1038,18 @@ end: return ret; } -int show_line_range(struct line_range *lr, const char *module, bool user) +int show_line_range(struct line_range *lr, const char *module, + struct nsinfo *nsi, bool user) { int ret; + struct nscookie nsc; ret = init_probe_symbol_maps(user); if (ret < 0) return ret; + nsinfo__mountns_enter(nsi, &nsc); ret = __show_line_range(lr, module, user); + nsinfo__mountns_exit(&nsc); exit_probe_symbol_maps(); return ret; @@ -1111,7 +1132,7 @@ int show_available_vars(struct perf_probe_event *pevs, int npevs, if (ret < 0) return ret; - dinfo = open_debuginfo(pevs->target, false); + dinfo = open_debuginfo(pevs->target, pevs->nsi, false); if (!dinfo) { ret = -ENOENT; goto out; @@ -1155,6 +1176,7 @@ static int try_to_find_probe_trace_events(struct perf_probe_event *pev, int show_line_range(struct line_range *lr __maybe_unused, const char *module __maybe_unused, + struct nsinfo *nsi __maybe_unused, bool user __maybe_unused) { pr_warning("Debuginfo-analysis is not supported.\n"); @@ -2703,6 +2725,7 @@ static int __add_probe_trace_events(struct perf_probe_event *pev, struct probe_trace_event *tev = NULL; struct probe_cache *cache = NULL; struct strlist *namelist[2] = {NULL, NULL}; + struct nscookie nsc; up = pev->uprobes ? 1 : 0; fd[up] = __open_probe_file_and_namelist(up, &namelist[up]); @@ -2729,7 +2752,9 @@ static int __add_probe_trace_events(struct perf_probe_event *pev, if (ret < 0) break; + nsinfo__mountns_enter(pev->nsi, &nsc); ret = probe_file__add_event(fd[up], tev); + nsinfo__mountns_exit(&nsc); if (ret < 0) break; @@ -2805,7 +2830,7 @@ static int find_probe_trace_events_from_map(struct perf_probe_event *pev, int ret, i, j, skipped = 0; char *mod_name; - map = get_target_map(pev->target, pev->uprobes); + map = get_target_map(pev->target, pev->nsi, pev->uprobes); if (!map) { ret = -EINVAL; goto out; @@ -3345,13 +3370,16 @@ int apply_perf_probe_events(struct perf_probe_event *pevs, int npevs) void cleanup_perf_probe_events(struct perf_probe_event *pevs, int npevs) { int i, j; + struct perf_probe_event *pev; /* Loop 3: cleanup and free trace events */ for (i = 0; i < npevs; i++) { + pev = &pevs[i]; for (j = 0; j < pevs[i].ntevs; j++) clear_probe_trace_event(&pevs[i].tevs[j]); zfree(&pevs[i].tevs); pevs[i].ntevs = 0; + nsinfo__zput(pev->nsi); clear_perf_probe_event(&pevs[i]); } } @@ -3409,8 +3437,8 @@ out: return ret; } -int show_available_funcs(const char *target, struct strfilter *_filter, - bool user) +int show_available_funcs(const char *target, struct nsinfo *nsi, + struct strfilter *_filter, bool user) { struct rb_node *nd; struct map *map; @@ -3421,7 +3449,7 @@ int show_available_funcs(const char *target, struct strfilter *_filter, return ret; /* Get a symbol map */ - map = get_target_map(target, user); + map = get_target_map(target, nsi, user); if (!map) { pr_err("Failed to get a map for %s\n", (target) ? : "kernel"); return -EINVAL; diff --git a/tools/perf/util/probe-event.h b/tools/perf/util/probe-event.h index 5812947..078681d 100644 --- a/tools/perf/util/probe-event.h +++ b/tools/perf/util/probe-event.h @@ -4,6 +4,7 @@ #include #include #include "intlist.h" +#include "namespaces.h" /* Probe related configurations */ struct probe_conf { @@ -92,6 +93,7 @@ struct perf_probe_event { struct perf_probe_arg *args; /* Arguments */ struct probe_trace_event *tevs; int ntevs; + struct nsinfo *nsi; /* Target namespace */ }; /* Line range */ @@ -163,10 +165,12 @@ int show_perf_probe_event(const char *group, const char *event, struct perf_probe_event *pev, const char *module, bool use_stdout); int show_perf_probe_events(struct strfilter *filter); -int show_line_range(struct line_range *lr, const char *module, bool user); +int show_line_range(struct line_range *lr, const char *module, + struct nsinfo *nsi, bool user); int show_available_vars(struct perf_probe_event *pevs, int npevs, struct strfilter *filter); -int show_available_funcs(const char *module, struct strfilter *filter, bool user); +int show_available_funcs(const char *module, struct nsinfo *nsi, + struct strfilter *filter, bool user); void arch__fix_tev_from_maps(struct perf_probe_event *pev, struct probe_trace_event *tev, struct map *map, struct symbol *sym); @@ -180,7 +184,7 @@ int e_snprintf(char *str, size_t size, const char *format, ...) __printf(3, 4); int copy_to_probe_trace_arg(struct probe_trace_arg *tvar, struct perf_probe_arg *pvar); -struct map *get_target_map(const char *target, bool user); +struct map *get_target_map(const char *target, struct nsinfo *nsi, bool user); void arch__post_process_probe_trace_events(struct perf_probe_event *pev, int ntevs); -- cgit v1.1 From f045b8c4b36baddcfbdd4d3d956446e688b0b3cd Mon Sep 17 00:00:00 2001 From: Krister Johansen Date: Wed, 5 Jul 2017 18:48:11 -0700 Subject: perf buildid-cache: Support binary objects from other namespaces Teach buildid-cache how to add, remove, and update binary objects from other mount namespaces. Allow probe events tracing binaries in different namespaces to add their objects to the probe and build-id caches too. As a handy side effect, this also lets us access SDT probes in binaries from alternate mount namespaces. Signed-off-by: Krister Johansen Tested-by: Brendan Gregg Cc: Alexander Shishkin Cc: Peter Zijlstra Cc: Thomas-Mich Richter Link: http://lkml.kernel.org/r/1499305693-1599-5-git-send-email-kjlx@templeofstupid.com [ Add util/namespaces.c to tools/perf/util/python-ext-sources, to fix the python binding 'perf test' ] Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/Documentation/perf-buildid-cache.txt | 5 +++ tools/perf/Documentation/perf-probe.txt | 5 +++ tools/perf/builtin-buildid-cache.c | 52 ++++++++++++++++++------- tools/perf/builtin-probe.c | 2 +- tools/perf/tests/sdt.c | 4 +- tools/perf/util/build-id.c | 47 ++++++++++++++-------- tools/perf/util/build-id.h | 9 +++-- tools/perf/util/dso.c | 12 +++++- tools/perf/util/parse-events.c | 2 +- tools/perf/util/probe-event.c | 6 +-- tools/perf/util/probe-file.c | 19 +++++---- tools/perf/util/probe-file.h | 4 +- tools/perf/util/python-ext-sources | 1 + tools/perf/util/symbol.c | 19 ++++++--- tools/perf/util/util.c | 34 +++++++++++++--- tools/perf/util/util.h | 2 + 16 files changed, 160 insertions(+), 63 deletions(-) (limited to 'tools') diff --git a/tools/perf/Documentation/perf-buildid-cache.txt b/tools/perf/Documentation/perf-buildid-cache.txt index 058064d..8468100 100644 --- a/tools/perf/Documentation/perf-buildid-cache.txt +++ b/tools/perf/Documentation/perf-buildid-cache.txt @@ -61,6 +61,11 @@ OPTIONS --verbose:: Be more verbose. +--target-ns=PID: + Obtain mount namespace information from the target pid. This is + used when creating a uprobe for a process that resides in a + different mount namespace from the perf(1) utility. + SEE ALSO -------- linkperf:perf-record[1], linkperf:perf-report[1], linkperf:perf-buildid-list[1] diff --git a/tools/perf/Documentation/perf-probe.txt b/tools/perf/Documentation/perf-probe.txt index a42aabc..d7e4869 100644 --- a/tools/perf/Documentation/perf-probe.txt +++ b/tools/perf/Documentation/perf-probe.txt @@ -273,6 +273,11 @@ Add a uprobe to a target process running in a different mount namespace ./perf probe --target-ns -x /lib64/libc.so.6 malloc +Add a USDT probe to a target process running in a different mount namespace + + ./perf probe --target-ns -x /usr/lib/jvm/java-1.8.0-openjdk-1.8.0.121-0.b13.el7_3.x86_64/jre/lib/amd64/server/libjvm.so %sdt_hotspot:thread__sleep__end + + SEE ALSO -------- linkperf:perf-trace[1], linkperf:perf-record[1], linkperf:perf-buildid-cache[1] diff --git a/tools/perf/builtin-buildid-cache.c b/tools/perf/builtin-buildid-cache.c index 9eba7f1..d65bd86 100644 --- a/tools/perf/builtin-buildid-cache.c +++ b/tools/perf/builtin-buildid-cache.c @@ -14,6 +14,7 @@ #include #include "builtin.h" #include "perf.h" +#include "namespaces.h" #include "util/cache.h" #include "util/debug.h" #include "util/header.h" @@ -165,33 +166,41 @@ static int build_id_cache__add_kcore(const char *filename, bool force) return 0; } -static int build_id_cache__add_file(const char *filename) +static int build_id_cache__add_file(const char *filename, struct nsinfo *nsi) { char sbuild_id[SBUILD_ID_SIZE]; u8 build_id[BUILD_ID_SIZE]; int err; + struct nscookie nsc; - if (filename__read_build_id(filename, &build_id, sizeof(build_id)) < 0) { + nsinfo__mountns_enter(nsi, &nsc); + err = filename__read_build_id(filename, &build_id, sizeof(build_id)); + nsinfo__mountns_exit(&nsc); + if (err < 0) { pr_debug("Couldn't read a build-id in %s\n", filename); return -1; } build_id__sprintf(build_id, sizeof(build_id), sbuild_id); - err = build_id_cache__add_s(sbuild_id, filename, + err = build_id_cache__add_s(sbuild_id, filename, nsi, false, false); pr_debug("Adding %s %s: %s\n", sbuild_id, filename, err ? "FAIL" : "Ok"); return err; } -static int build_id_cache__remove_file(const char *filename) +static int build_id_cache__remove_file(const char *filename, struct nsinfo *nsi) { u8 build_id[BUILD_ID_SIZE]; char sbuild_id[SBUILD_ID_SIZE]; + struct nscookie nsc; int err; - if (filename__read_build_id(filename, &build_id, sizeof(build_id)) < 0) { + nsinfo__mountns_enter(nsi, &nsc); + err = filename__read_build_id(filename, &build_id, sizeof(build_id)); + nsinfo__mountns_exit(&nsc); + if (err < 0) { pr_debug("Couldn't read a build-id in %s\n", filename); return -1; } @@ -204,13 +213,13 @@ static int build_id_cache__remove_file(const char *filename) return err; } -static int build_id_cache__purge_path(const char *pathname) +static int build_id_cache__purge_path(const char *pathname, struct nsinfo *nsi) { struct strlist *list; struct str_node *pos; int err; - err = build_id_cache__list_build_ids(pathname, &list); + err = build_id_cache__list_build_ids(pathname, nsi, &list); if (err) goto out; @@ -256,24 +265,30 @@ static int build_id_cache__fprintf_missing(struct perf_session *session, FILE *f return 0; } -static int build_id_cache__update_file(const char *filename) +static int build_id_cache__update_file(const char *filename, struct nsinfo *nsi) { u8 build_id[BUILD_ID_SIZE]; char sbuild_id[SBUILD_ID_SIZE]; + struct nscookie nsc; - int err = 0; + int err; - if (filename__read_build_id(filename, &build_id, sizeof(build_id)) < 0) { + nsinfo__mountns_enter(nsi, &nsc); + err = filename__read_build_id(filename, &build_id, sizeof(build_id)); + nsinfo__mountns_exit(&nsc); + if (err < 0) { pr_debug("Couldn't read a build-id in %s\n", filename); return -1; } + err = 0; build_id__sprintf(build_id, sizeof(build_id), sbuild_id); if (build_id_cache__cached(sbuild_id)) err = build_id_cache__remove_s(sbuild_id); if (!err) - err = build_id_cache__add_s(sbuild_id, filename, false, false); + err = build_id_cache__add_s(sbuild_id, filename, nsi, false, + false); pr_debug("Updating %s %s: %s\n", sbuild_id, filename, err ? "FAIL" : "Ok"); @@ -286,6 +301,7 @@ int cmd_buildid_cache(int argc, const char **argv) struct strlist *list; struct str_node *pos; int ret = 0; + int ns_id = -1; bool force = false; char const *add_name_list_str = NULL, *remove_name_list_str = NULL, @@ -299,6 +315,7 @@ int cmd_buildid_cache(int argc, const char **argv) .mode = PERF_DATA_MODE_READ, }; struct perf_session *session = NULL; + struct nsinfo *nsi = NULL; const struct option buildid_cache_options[] = { OPT_STRING('a', "add", &add_name_list_str, @@ -315,6 +332,7 @@ int cmd_buildid_cache(int argc, const char **argv) OPT_STRING('u', "update", &update_name_list_str, "file list", "file(s) to update"), OPT_INCR('v', "verbose", &verbose, "be more verbose"), + OPT_INTEGER(0, "target-ns", &ns_id, "target pid for namespace context"), OPT_END() }; const char * const buildid_cache_usage[] = { @@ -330,6 +348,9 @@ int cmd_buildid_cache(int argc, const char **argv) !missing_filename && !update_name_list_str)) usage_with_options(buildid_cache_usage, buildid_cache_options); + if (ns_id > 0) + nsi = nsinfo__new(ns_id); + if (missing_filename) { file.path = missing_filename; file.force = force; @@ -348,7 +369,7 @@ int cmd_buildid_cache(int argc, const char **argv) list = strlist__new(add_name_list_str, NULL); if (list) { strlist__for_each_entry(pos, list) - if (build_id_cache__add_file(pos->s)) { + if (build_id_cache__add_file(pos->s, nsi)) { if (errno == EEXIST) { pr_debug("%s already in the cache\n", pos->s); @@ -366,7 +387,7 @@ int cmd_buildid_cache(int argc, const char **argv) list = strlist__new(remove_name_list_str, NULL); if (list) { strlist__for_each_entry(pos, list) - if (build_id_cache__remove_file(pos->s)) { + if (build_id_cache__remove_file(pos->s, nsi)) { if (errno == ENOENT) { pr_debug("%s wasn't in the cache\n", pos->s); @@ -384,7 +405,7 @@ int cmd_buildid_cache(int argc, const char **argv) list = strlist__new(purge_name_list_str, NULL); if (list) { strlist__for_each_entry(pos, list) - if (build_id_cache__purge_path(pos->s)) { + if (build_id_cache__purge_path(pos->s, nsi)) { if (errno == ENOENT) { pr_debug("%s wasn't in the cache\n", pos->s); @@ -405,7 +426,7 @@ int cmd_buildid_cache(int argc, const char **argv) list = strlist__new(update_name_list_str, NULL); if (list) { strlist__for_each_entry(pos, list) - if (build_id_cache__update_file(pos->s)) { + if (build_id_cache__update_file(pos->s, nsi)) { if (errno == ENOENT) { pr_debug("%s wasn't in the cache\n", pos->s); @@ -424,6 +445,7 @@ int cmd_buildid_cache(int argc, const char **argv) out: perf_session__delete(session); + nsinfo__zput(nsi); return ret; } diff --git a/tools/perf/builtin-probe.c b/tools/perf/builtin-probe.c index 3fb98d5..c006592 100644 --- a/tools/perf/builtin-probe.c +++ b/tools/perf/builtin-probe.c @@ -416,7 +416,7 @@ static int del_perf_probe_caches(struct strfilter *filter) } strlist__for_each_entry(nd, bidlist) { - cache = probe_cache__new(nd->s); + cache = probe_cache__new(nd->s, NULL); if (!cache) continue; if (probe_cache__filter_purge(cache, filter) < 0 || diff --git a/tools/perf/tests/sdt.c b/tools/perf/tests/sdt.c index 06eda67..5abe8ce 100644 --- a/tools/perf/tests/sdt.c +++ b/tools/perf/tests/sdt.c @@ -33,7 +33,7 @@ static int build_id_cache__add_file(const char *filename) } build_id__sprintf(build_id, sizeof(build_id), sbuild_id); - err = build_id_cache__add_s(sbuild_id, filename, false, false); + err = build_id_cache__add_s(sbuild_id, filename, NULL, false, false); if (err < 0) pr_debug("Failed to add build id cache of %s\n", filename); return err; @@ -54,7 +54,7 @@ static char *get_self_path(void) static int search_cached_probe(const char *target, const char *group, const char *event) { - struct probe_cache *cache = probe_cache__new(target); + struct probe_cache *cache = probe_cache__new(target, NULL); int ret = 0; if (!cache) { diff --git a/tools/perf/util/build-id.c b/tools/perf/util/build-id.c index e0148b0..f7bfd90 100644 --- a/tools/perf/util/build-id.c +++ b/tools/perf/util/build-id.c @@ -534,13 +534,14 @@ char *build_id_cache__complement(const char *incomplete_sbuild_id) } char *build_id_cache__cachedir(const char *sbuild_id, const char *name, - bool is_kallsyms, bool is_vdso) + struct nsinfo *nsi, bool is_kallsyms, + bool is_vdso) { char *realname = (char *)name, *filename; bool slash = is_kallsyms || is_vdso; if (!slash) { - realname = realpath(name, NULL); + realname = nsinfo__realpath(name, nsi); if (!realname) return NULL; } @@ -556,13 +557,13 @@ char *build_id_cache__cachedir(const char *sbuild_id, const char *name, return filename; } -int build_id_cache__list_build_ids(const char *pathname, +int build_id_cache__list_build_ids(const char *pathname, struct nsinfo *nsi, struct strlist **result) { char *dir_name; int ret = 0; - dir_name = build_id_cache__cachedir(NULL, pathname, false, false); + dir_name = build_id_cache__cachedir(NULL, pathname, nsi, false, false); if (!dir_name) return -ENOMEM; @@ -576,16 +577,20 @@ int build_id_cache__list_build_ids(const char *pathname, #if defined(HAVE_LIBELF_SUPPORT) && defined(HAVE_GELF_GETNOTE_SUPPORT) static int build_id_cache__add_sdt_cache(const char *sbuild_id, - const char *realname) + const char *realname, + struct nsinfo *nsi) { struct probe_cache *cache; int ret; + struct nscookie nsc; - cache = probe_cache__new(sbuild_id); + cache = probe_cache__new(sbuild_id, nsi); if (!cache) return -1; + nsinfo__mountns_enter(nsi, &nsc); ret = probe_cache__scan_sdt(cache, realname); + nsinfo__mountns_exit(&nsc); if (ret >= 0) { pr_debug4("Found %d SDTs in %s\n", ret, realname); if (probe_cache__commit(cache) < 0) @@ -595,11 +600,11 @@ static int build_id_cache__add_sdt_cache(const char *sbuild_id, return ret; } #else -#define build_id_cache__add_sdt_cache(sbuild_id, realname) (0) +#define build_id_cache__add_sdt_cache(sbuild_id, realname, nsi) (0) #endif int build_id_cache__add_s(const char *sbuild_id, const char *name, - bool is_kallsyms, bool is_vdso) + struct nsinfo *nsi, bool is_kallsyms, bool is_vdso) { const size_t size = PATH_MAX; char *realname = NULL, *filename = NULL, *dir_name = NULL, @@ -607,13 +612,16 @@ int build_id_cache__add_s(const char *sbuild_id, const char *name, int err = -1; if (!is_kallsyms) { - realname = realpath(name, NULL); + if (!is_vdso) + realname = nsinfo__realpath(name, nsi); + else + realname = realpath(name, NULL); if (!realname) goto out_free; } - dir_name = build_id_cache__cachedir(sbuild_id, name, - is_kallsyms, is_vdso); + dir_name = build_id_cache__cachedir(sbuild_id, name, nsi, is_kallsyms, + is_vdso); if (!dir_name) goto out_free; @@ -634,7 +642,10 @@ int build_id_cache__add_s(const char *sbuild_id, const char *name, if (access(filename, F_OK)) { if (is_kallsyms) { - if (copyfile("/proc/kallsyms", filename)) + if (copyfile("/proc/kallsyms", filename)) + goto out_free; + } else if (nsi && nsi->need_setns) { + if (copyfile_ns(name, filename, nsi)) goto out_free; } else if (link(realname, filename) && errno != EEXIST && copyfile(name, filename)) @@ -657,7 +668,8 @@ int build_id_cache__add_s(const char *sbuild_id, const char *name, err = 0; /* Update SDT cache : error is just warned */ - if (realname && build_id_cache__add_sdt_cache(sbuild_id, realname) < 0) + if (realname && + build_id_cache__add_sdt_cache(sbuild_id, realname, nsi) < 0) pr_debug4("Failed to update/scan SDT cache for %s\n", realname); out_free: @@ -670,14 +682,15 @@ out_free: } static int build_id_cache__add_b(const u8 *build_id, size_t build_id_size, - const char *name, bool is_kallsyms, - bool is_vdso) + const char *name, struct nsinfo *nsi, + bool is_kallsyms, bool is_vdso) { char sbuild_id[SBUILD_ID_SIZE]; build_id__sprintf(build_id, build_id_size, sbuild_id); - return build_id_cache__add_s(sbuild_id, name, is_kallsyms, is_vdso); + return build_id_cache__add_s(sbuild_id, name, nsi, is_kallsyms, + is_vdso); } bool build_id_cache__cached(const char *sbuild_id) @@ -743,7 +756,7 @@ static int dso__cache_build_id(struct dso *dso, struct machine *machine) name = nm; } return build_id_cache__add_b(dso->build_id, sizeof(dso->build_id), name, - is_kallsyms, is_vdso); + dso->nsinfo, is_kallsyms, is_vdso); } static int __dsos__cache_build_ids(struct list_head *head, diff --git a/tools/perf/util/build-id.h b/tools/perf/util/build-id.h index 96690a5..2397084 100644 --- a/tools/perf/util/build-id.h +++ b/tools/perf/util/build-id.h @@ -5,6 +5,7 @@ #define SBUILD_ID_SIZE (BUILD_ID_SIZE * 2 + 1) #include "tool.h" +#include "namespaces.h" #include extern struct perf_tool build_id__mark_dso_hit_ops; @@ -31,17 +32,19 @@ int perf_session__cache_build_ids(struct perf_session *session); char *build_id_cache__origname(const char *sbuild_id); char *build_id_cache__linkname(const char *sbuild_id, char *bf, size_t size); char *build_id_cache__cachedir(const char *sbuild_id, const char *name, - bool is_kallsyms, bool is_vdso); + struct nsinfo *nsi, bool is_kallsyms, + bool is_vdso); struct strlist; struct strlist *build_id_cache__list_all(bool validonly); char *build_id_cache__complement(const char *incomplete_sbuild_id); -int build_id_cache__list_build_ids(const char *pathname, +int build_id_cache__list_build_ids(const char *pathname, struct nsinfo *nsi, struct strlist **result); bool build_id_cache__cached(const char *sbuild_id); int build_id_cache__add_s(const char *sbuild_id, - const char *name, bool is_kallsyms, bool is_vdso); + const char *name, struct nsinfo *nsi, + bool is_kallsyms, bool is_vdso); int build_id_cache__remove_s(const char *sbuild_id); extern char buildid_dir[]; diff --git a/tools/perf/util/dso.c b/tools/perf/util/dso.c index beda40e..dc9b495 100644 --- a/tools/perf/util/dso.c +++ b/tools/perf/util/dso.c @@ -504,7 +504,14 @@ static void check_data_close(void); */ static int open_dso(struct dso *dso, struct machine *machine) { - int fd = __open_dso(dso, machine); + int fd; + struct nscookie nsc; + + if (dso->binary_type != DSO_BINARY_TYPE__BUILD_ID_CACHE) + nsinfo__mountns_enter(dso->nsinfo, &nsc); + fd = __open_dso(dso, machine); + if (dso->binary_type != DSO_BINARY_TYPE__BUILD_ID_CACHE) + nsinfo__mountns_exit(&nsc); if (fd >= 0) { dso__list_add(dso); @@ -1302,6 +1309,7 @@ bool __dsos__read_build_ids(struct list_head *head, bool with_hits) { bool have_build_id = false; struct dso *pos; + struct nscookie nsc; list_for_each_entry(pos, head, node) { if (with_hits && !pos->hit && !dso__is_vdso(pos)) @@ -1310,11 +1318,13 @@ bool __dsos__read_build_ids(struct list_head *head, bool with_hits) have_build_id = true; continue; } + nsinfo__mountns_enter(pos->nsinfo, &nsc); if (filename__read_build_id(pos->long_name, pos->build_id, sizeof(pos->build_id)) > 0) { have_build_id = true; pos->has_build_id = true; } + nsinfo__mountns_exit(&nsc); } return have_build_id; diff --git a/tools/perf/util/parse-events.c b/tools/perf/util/parse-events.c index 01e779b..84e3010 100644 --- a/tools/perf/util/parse-events.c +++ b/tools/perf/util/parse-events.c @@ -2124,7 +2124,7 @@ void print_sdt_events(const char *subsys_glob, const char *event_glob, return; } strlist__for_each_entry(nd, bidlist) { - pcache = probe_cache__new(nd->s); + pcache = probe_cache__new(nd->s, NULL); if (!pcache) continue; list_for_each_entry(ent, &pcache->entries, node) { diff --git a/tools/perf/util/probe-event.c b/tools/perf/util/probe-event.c index a80895a..d7cd114 100644 --- a/tools/perf/util/probe-event.c +++ b/tools/perf/util/probe-event.c @@ -2769,7 +2769,7 @@ static int __add_probe_trace_events(struct perf_probe_event *pev, if (ret == -EINVAL && pev->uprobes) warn_uprobe_event_compat(tev); if (ret == 0 && probe_conf.cache) { - cache = probe_cache__new(pev->target); + cache = probe_cache__new(pev->target, pev->nsi); if (!cache || probe_cache__add_entry(cache, pev, tevs, ntevs) < 0 || probe_cache__commit(cache) < 0) @@ -3119,7 +3119,7 @@ static int find_cached_events(struct perf_probe_event *pev, int ntevs = 0; int ret = 0; - cache = probe_cache__new(target); + cache = probe_cache__new(target, pev->nsi); /* Return 0 ("not found") if the target has no probe cache. */ if (!cache) return 0; @@ -3209,7 +3209,7 @@ static int find_probe_trace_events_from_cache(struct perf_probe_event *pev, else return find_cached_events(pev, tevs, pev->target); } - cache = probe_cache__new(pev->target); + cache = probe_cache__new(pev->target, pev->nsi); if (!cache) return 0; diff --git a/tools/perf/util/probe-file.c b/tools/perf/util/probe-file.c index d679389..cdf8d83 100644 --- a/tools/perf/util/probe-file.c +++ b/tools/perf/util/probe-file.c @@ -412,13 +412,15 @@ int probe_cache_entry__get_event(struct probe_cache_entry *entry, } /* For the kernel probe caches, pass target = NULL or DSO__NAME_KALLSYMS */ -static int probe_cache__open(struct probe_cache *pcache, const char *target) +static int probe_cache__open(struct probe_cache *pcache, const char *target, + struct nsinfo *nsi) { char cpath[PATH_MAX]; char sbuildid[SBUILD_ID_SIZE]; char *dir_name = NULL; bool is_kallsyms = false; int ret, fd; + struct nscookie nsc; if (target && build_id_cache__cached(target)) { /* This is a cached buildid */ @@ -431,8 +433,11 @@ static int probe_cache__open(struct probe_cache *pcache, const char *target) target = DSO__NAME_KALLSYMS; is_kallsyms = true; ret = sysfs__sprintf_build_id("/", sbuildid); - } else + } else { + nsinfo__mountns_enter(nsi, &nsc); ret = filename__sprintf_build_id(target, sbuildid); + nsinfo__mountns_exit(&nsc); + } if (ret < 0) { pr_debug("Failed to get build-id from %s.\n", target); @@ -441,7 +446,7 @@ static int probe_cache__open(struct probe_cache *pcache, const char *target) /* If we have no buildid cache, make it */ if (!build_id_cache__cached(sbuildid)) { - ret = build_id_cache__add_s(sbuildid, target, + ret = build_id_cache__add_s(sbuildid, target, nsi, is_kallsyms, NULL); if (ret < 0) { pr_debug("Failed to add build-id cache: %s\n", target); @@ -449,7 +454,7 @@ static int probe_cache__open(struct probe_cache *pcache, const char *target) } } - dir_name = build_id_cache__cachedir(sbuildid, target, is_kallsyms, + dir_name = build_id_cache__cachedir(sbuildid, target, nsi, is_kallsyms, false); found: if (!dir_name) { @@ -554,7 +559,7 @@ void probe_cache__delete(struct probe_cache *pcache) free(pcache); } -struct probe_cache *probe_cache__new(const char *target) +struct probe_cache *probe_cache__new(const char *target, struct nsinfo *nsi) { struct probe_cache *pcache = probe_cache__alloc(); int ret; @@ -562,7 +567,7 @@ struct probe_cache *probe_cache__new(const char *target) if (!pcache) return NULL; - ret = probe_cache__open(pcache, target); + ret = probe_cache__open(pcache, target, nsi); if (ret < 0) { pr_debug("Cache open error: %d\n", ret); goto out_err; @@ -974,7 +979,7 @@ int probe_cache__show_all_caches(struct strfilter *filter) return -EINVAL; } strlist__for_each_entry(nd, bidlist) { - pcache = probe_cache__new(nd->s); + pcache = probe_cache__new(nd->s, NULL); if (!pcache) continue; if (!list_empty(&pcache->entries)) { diff --git a/tools/perf/util/probe-file.h b/tools/perf/util/probe-file.h index 5ecc9d3..2ca4163 100644 --- a/tools/perf/util/probe-file.h +++ b/tools/perf/util/probe-file.h @@ -51,7 +51,7 @@ int probe_file__del_strlist(int fd, struct strlist *namelist); int probe_cache_entry__get_event(struct probe_cache_entry *entry, struct probe_trace_event **tevs); -struct probe_cache *probe_cache__new(const char *target); +struct probe_cache *probe_cache__new(const char *target, struct nsinfo *nsi); int probe_cache__add_entry(struct probe_cache *pcache, struct perf_probe_event *pev, struct probe_trace_event *tevs, int ntevs); @@ -69,7 +69,7 @@ int probe_cache__show_all_caches(struct strfilter *filter); bool probe_type_is_available(enum probe_type type); bool kretprobe_offset_is_supported(void); #else /* ! HAVE_LIBELF_SUPPORT */ -static inline struct probe_cache *probe_cache__new(const char *tgt __maybe_unused) +static inline struct probe_cache *probe_cache__new(const char *tgt __maybe_unused, struct nsinfo *nsi __maybe_unused) { return NULL; } diff --git a/tools/perf/util/python-ext-sources b/tools/perf/util/python-ext-sources index 9f3b0d9..e66dc49 100644 --- a/tools/perf/util/python-ext-sources +++ b/tools/perf/util/python-ext-sources @@ -10,6 +10,7 @@ util/ctype.c util/evlist.c util/evsel.c util/cpumap.c +util/namespaces.c ../lib/bitmap.c ../lib/find_bit.c ../lib/hweight.c diff --git a/tools/perf/util/symbol.c b/tools/perf/util/symbol.c index 21c97cc..8c7bae5 100644 --- a/tools/perf/util/symbol.c +++ b/tools/perf/util/symbol.c @@ -1464,7 +1464,6 @@ static int dso__find_perf_map(char *filebuf, size_t bufsz, return rc; } - int dso__load(struct dso *dso, struct map *map) { char *name; @@ -1565,6 +1564,8 @@ int dso__load(struct dso *dso, struct map *map) for (i = 0; i < DSO_BINARY_TYPE__SYMTAB_CNT; i++) { struct symsrc *ss = &ss_[ss_pos]; bool next_slot = false; + bool is_reg; + int sirc; enum dso_binary_type symtab_type = binary_type_symtab[i]; @@ -1575,12 +1576,20 @@ int dso__load(struct dso *dso, struct map *map) root_dir, name, PATH_MAX)) continue; - if (!is_regular_file(name)) - continue; + if (symtab_type == DSO_BINARY_TYPE__BUILD_ID_CACHE) + nsinfo__mountns_exit(&nsc); + + is_reg = is_regular_file(name); + sirc = symsrc__init(ss, dso, name, symtab_type); - /* Name is now the name of the next image to try */ - if (symsrc__init(ss, dso, name, symtab_type) < 0) + if (symtab_type == DSO_BINARY_TYPE__BUILD_ID_CACHE) + nsinfo__mountns_enter(dso->nsinfo, &nsc); + + if (!is_reg || sirc < 0) { + if (sirc >= 0) + symsrc__destroy(ss); continue; + } if (!syms_ss && symsrc__has_symtab(ss)) { syms_ss = ss; diff --git a/tools/perf/util/util.c b/tools/perf/util/util.c index 988111e..9e4ea85 100644 --- a/tools/perf/util/util.c +++ b/tools/perf/util/util.c @@ -143,13 +143,17 @@ out: return list; } -static int slow_copyfile(const char *from, const char *to) +static int slow_copyfile(const char *from, const char *to, struct nsinfo *nsi) { int err = -1; char *line = NULL; size_t n; - FILE *from_fp = fopen(from, "r"), *to_fp; + FILE *from_fp, *to_fp; + struct nscookie nsc; + nsinfo__mountns_enter(nsi, &nsc); + from_fp = fopen(from, "r"); + nsinfo__mountns_exit(&nsc); if (from_fp == NULL) goto out; @@ -198,15 +202,21 @@ int copyfile_offset(int ifd, loff_t off_in, int ofd, loff_t off_out, u64 size) return size ? -1 : 0; } -int copyfile_mode(const char *from, const char *to, mode_t mode) +static int copyfile_mode_ns(const char *from, const char *to, mode_t mode, + struct nsinfo *nsi) { int fromfd, tofd; struct stat st; - int err = -1; + int err; char *tmp = NULL, *ptr = NULL; + struct nscookie nsc; - if (stat(from, &st)) + nsinfo__mountns_enter(nsi, &nsc); + err = stat(from, &st); + nsinfo__mountns_exit(&nsc); + if (err) goto out; + err = -1; /* extra 'x' at the end is to reserve space for '.' */ if (asprintf(&tmp, "%s.XXXXXXx", to) < 0) { @@ -227,11 +237,13 @@ int copyfile_mode(const char *from, const char *to, mode_t mode) goto out_close_to; if (st.st_size == 0) { /* /proc? do it slowly... */ - err = slow_copyfile(from, tmp); + err = slow_copyfile(from, tmp, nsi); goto out_close_to; } + nsinfo__mountns_enter(nsi, &nsc); fromfd = open(from, O_RDONLY); + nsinfo__mountns_exit(&nsc); if (fromfd < 0) goto out_close_to; @@ -248,6 +260,16 @@ out: return err; } +int copyfile_ns(const char *from, const char *to, struct nsinfo *nsi) +{ + return copyfile_mode_ns(from, to, 0755, nsi); +} + +int copyfile_mode(const char *from, const char *to, mode_t mode) +{ + return copyfile_mode_ns(from, to, mode, NULL); +} + int copyfile(const char *from, const char *to) { return copyfile_mode(from, to, 0755); diff --git a/tools/perf/util/util.h b/tools/perf/util/util.h index 1e5fe1d..062dd20 100644 --- a/tools/perf/util/util.h +++ b/tools/perf/util/util.h @@ -12,6 +12,7 @@ #include #include #include +#include "namespaces.h" /* General helper functions */ void usage(const char *err) __noreturn; @@ -33,6 +34,7 @@ struct strlist *lsdir(const char *name, bool (*filter)(const char *, struct dire bool lsdir_no_dot_filter(const char *name, struct dirent *d); int copyfile(const char *from, const char *to); int copyfile_mode(const char *from, const char *to, mode_t mode); +int copyfile_ns(const char *from, const char *to, struct nsinfo *nsi); int copyfile_offset(int fromfd, loff_t from_ofs, int tofd, loff_t to_ofs, u64 size); ssize_t readn(int fd, void *buf, size_t n); -- cgit v1.1 From d2396999c998b4e0006aef247e154eff0ed3d8f9 Mon Sep 17 00:00:00 2001 From: Krister Johansen Date: Wed, 5 Jul 2017 18:48:13 -0700 Subject: perf buildid-cache: Cache debuginfo If a stripped binary is placed in the cache, the user is in a situation where there's a cached elf file present, but it doesn't have any symtab to use for name resolution. Grab the debuginfo for binaries that don't end in .ko. This yields a better chance of resolving symbols from older traces. Signed-off-by: Krister Johansen Cc: Alexander Shishkin Cc: Brendan Gregg Cc: Peter Zijlstra Cc: Thomas-Mich Richter Link: http://lkml.kernel.org/r/1499305693-1599-7-git-send-email-kjlx@templeofstupid.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-buildid-cache.c | 2 +- tools/perf/util/annotate.c | 2 +- tools/perf/util/build-id.c | 72 +++++++++++++++++++++++++++++++++++--- tools/perf/util/build-id.h | 3 +- tools/perf/util/dso.c | 8 ++++- tools/perf/util/dso.h | 1 + tools/perf/util/machine.c | 3 +- tools/perf/util/symbol.c | 12 +++++-- 8 files changed, 90 insertions(+), 13 deletions(-) (limited to 'tools') diff --git a/tools/perf/builtin-buildid-cache.c b/tools/perf/builtin-buildid-cache.c index d65bd86..e3eb624 100644 --- a/tools/perf/builtin-buildid-cache.c +++ b/tools/perf/builtin-buildid-cache.c @@ -243,7 +243,7 @@ static bool dso__missing_buildid_cache(struct dso *dso, int parm __maybe_unused) char filename[PATH_MAX]; u8 build_id[BUILD_ID_SIZE]; - if (dso__build_id_filename(dso, filename, sizeof(filename)) && + if (dso__build_id_filename(dso, filename, sizeof(filename), false) && filename__read_build_id(filename, build_id, sizeof(build_id)) != sizeof(build_id)) { if (errno == ENOENT) diff --git a/tools/perf/util/annotate.c b/tools/perf/util/annotate.c index ef434b5..1742510 100644 --- a/tools/perf/util/annotate.c +++ b/tools/perf/util/annotate.c @@ -1347,7 +1347,7 @@ static int dso__disassemble_filename(struct dso *dso, char *filename, size_t fil !dso__is_kcore(dso)) return SYMBOL_ANNOTATE_ERRNO__NO_VMLINUX; - build_id_filename = dso__build_id_filename(dso, NULL, 0); + build_id_filename = dso__build_id_filename(dso, NULL, 0, false); if (build_id_filename) { __symbol__join_symfs(filename, filename_size, build_id_filename); free(build_id_filename); diff --git a/tools/perf/util/build-id.c b/tools/perf/util/build-id.c index f7bfd90..e966515 100644 --- a/tools/perf/util/build-id.c +++ b/tools/perf/util/build-id.c @@ -243,12 +243,15 @@ static bool build_id_cache__valid_id(char *sbuild_id) return result; } -static const char *build_id_cache__basename(bool is_kallsyms, bool is_vdso) +static const char *build_id_cache__basename(bool is_kallsyms, bool is_vdso, + bool is_debug) { - return is_kallsyms ? "kallsyms" : (is_vdso ? "vdso" : "elf"); + return is_kallsyms ? "kallsyms" : (is_vdso ? "vdso" : (is_debug ? + "debug" : "elf")); } -char *dso__build_id_filename(const struct dso *dso, char *bf, size_t size) +char *dso__build_id_filename(const struct dso *dso, char *bf, size_t size, + bool is_debug) { bool is_kallsyms = dso__is_kallsyms((struct dso *)dso); bool is_vdso = dso__is_vdso((struct dso *)dso); @@ -270,7 +273,8 @@ char *dso__build_id_filename(const struct dso *dso, char *bf, size_t size) ret = asnprintf(&bf, size, "%s", linkname); else ret = asnprintf(&bf, size, "%s/%s", linkname, - build_id_cache__basename(is_kallsyms, is_vdso)); + build_id_cache__basename(is_kallsyms, is_vdso, + is_debug)); if (ret < 0 || (!alloc && size < (unsigned int)ret)) bf = NULL; free(linkname); @@ -603,12 +607,40 @@ static int build_id_cache__add_sdt_cache(const char *sbuild_id, #define build_id_cache__add_sdt_cache(sbuild_id, realname, nsi) (0) #endif +static char *build_id_cache__find_debug(const char *sbuild_id, + struct nsinfo *nsi) +{ + char *realname = NULL; + char *debugfile; + struct nscookie nsc; + size_t len = 0; + + debugfile = calloc(1, PATH_MAX); + if (!debugfile) + goto out; + + len = __symbol__join_symfs(debugfile, PATH_MAX, + "/usr/lib/debug/.build-id/"); + snprintf(debugfile + len, PATH_MAX - len, "%.2s/%s.debug", sbuild_id, + sbuild_id + 2); + + nsinfo__mountns_enter(nsi, &nsc); + realname = realpath(debugfile, NULL); + if (realname && access(realname, R_OK)) + zfree(&realname); + nsinfo__mountns_exit(&nsc); +out: + free(debugfile); + return realname; +} + int build_id_cache__add_s(const char *sbuild_id, const char *name, struct nsinfo *nsi, bool is_kallsyms, bool is_vdso) { const size_t size = PATH_MAX; char *realname = NULL, *filename = NULL, *dir_name = NULL, *linkname = zalloc(size), *tmp; + char *debugfile = NULL; int err = -1; if (!is_kallsyms) { @@ -635,7 +667,8 @@ int build_id_cache__add_s(const char *sbuild_id, const char *name, /* Save the allocated buildid dirname */ if (asprintf(&filename, "%s/%s", dir_name, - build_id_cache__basename(is_kallsyms, is_vdso)) < 0) { + build_id_cache__basename(is_kallsyms, is_vdso, + false)) < 0) { filename = NULL; goto out_free; } @@ -652,6 +685,34 @@ int build_id_cache__add_s(const char *sbuild_id, const char *name, goto out_free; } + /* Some binaries are stripped, but have .debug files with their symbol + * table. Check to see if we can locate one of those, since the elf + * file itself may not be very useful to users of our tools without a + * symtab. + */ + if (!is_kallsyms && !is_vdso && + strncmp(".ko", name + strlen(name) - 3, 3)) { + debugfile = build_id_cache__find_debug(sbuild_id, nsi); + if (debugfile) { + zfree(&filename); + if (asprintf(&filename, "%s/%s", dir_name, + build_id_cache__basename(false, false, true)) < 0) { + filename = NULL; + goto out_free; + } + if (access(filename, F_OK)) { + if (nsi && nsi->need_setns) { + if (copyfile_ns(debugfile, filename, + nsi)) + goto out_free; + } else if (link(debugfile, filename) && + errno != EEXIST && + copyfile(debugfile, filename)) + goto out_free; + } + } + } + if (!build_id_cache__linkname(sbuild_id, linkname, size)) goto out_free; tmp = strrchr(linkname, '/'); @@ -676,6 +737,7 @@ out_free: if (!is_kallsyms) free(realname); free(filename); + free(debugfile); free(dir_name); free(linkname); return err; diff --git a/tools/perf/util/build-id.h b/tools/perf/util/build-id.h index 2397084..113dc06 100644 --- a/tools/perf/util/build-id.h +++ b/tools/perf/util/build-id.h @@ -17,7 +17,8 @@ int filename__sprintf_build_id(const char *pathname, char *sbuild_id); char *build_id_cache__kallsyms_path(const char *sbuild_id, char *bf, size_t size); -char *dso__build_id_filename(const struct dso *dso, char *bf, size_t size); +char *dso__build_id_filename(const struct dso *dso, char *bf, size_t size, + bool is_debug); int build_id__mark_dso_hit(struct perf_tool *tool, union perf_event *event, struct perf_sample *sample, struct perf_evsel *evsel, diff --git a/tools/perf/util/dso.c b/tools/perf/util/dso.c index dc9b495..b9e087f 100644 --- a/tools/perf/util/dso.c +++ b/tools/perf/util/dso.c @@ -32,6 +32,7 @@ char dso__symtab_origin(const struct dso *dso) [DSO_BINARY_TYPE__JAVA_JIT] = 'j', [DSO_BINARY_TYPE__DEBUGLINK] = 'l', [DSO_BINARY_TYPE__BUILD_ID_CACHE] = 'B', + [DSO_BINARY_TYPE__BUILD_ID_CACHE_DEBUGINFO] = 'D', [DSO_BINARY_TYPE__FEDORA_DEBUGINFO] = 'f', [DSO_BINARY_TYPE__UBUNTU_DEBUGINFO] = 'u', [DSO_BINARY_TYPE__OPENEMBEDDED_DEBUGINFO] = 'o', @@ -97,7 +98,12 @@ int dso__read_binary_type_filename(const struct dso *dso, break; } case DSO_BINARY_TYPE__BUILD_ID_CACHE: - if (dso__build_id_filename(dso, filename, size) == NULL) + if (dso__build_id_filename(dso, filename, size, false) == NULL) + ret = -1; + break; + + case DSO_BINARY_TYPE__BUILD_ID_CACHE_DEBUGINFO: + if (dso__build_id_filename(dso, filename, size, true) == NULL) ret = -1; break; diff --git a/tools/perf/util/dso.h b/tools/perf/util/dso.h index 78ec637..f886141 100644 --- a/tools/perf/util/dso.h +++ b/tools/perf/util/dso.h @@ -21,6 +21,7 @@ enum dso_binary_type { DSO_BINARY_TYPE__JAVA_JIT, DSO_BINARY_TYPE__DEBUGLINK, DSO_BINARY_TYPE__BUILD_ID_CACHE, + DSO_BINARY_TYPE__BUILD_ID_CACHE_DEBUGINFO, DSO_BINARY_TYPE__FEDORA_DEBUGINFO, DSO_BINARY_TYPE__UBUNTU_DEBUGINFO, DSO_BINARY_TYPE__BUILDID_DEBUGINFO, diff --git a/tools/perf/util/machine.c b/tools/perf/util/machine.c index 246b441..a54a2be 100644 --- a/tools/perf/util/machine.c +++ b/tools/perf/util/machine.c @@ -705,7 +705,8 @@ size_t machine__fprintf_vmlinux_path(struct machine *machine, FILE *fp) if (kdso->has_build_id) { char filename[PATH_MAX]; - if (dso__build_id_filename(kdso, filename, sizeof(filename))) + if (dso__build_id_filename(kdso, filename, sizeof(filename), + false)) printed += fprintf(fp, "[0] %s\n", filename); } diff --git a/tools/perf/util/symbol.c b/tools/perf/util/symbol.c index 8c7bae5..971b990 100644 --- a/tools/perf/util/symbol.c +++ b/tools/perf/util/symbol.c @@ -53,6 +53,7 @@ static enum dso_binary_type binary_type_symtab[] = { DSO_BINARY_TYPE__JAVA_JIT, DSO_BINARY_TYPE__DEBUGLINK, DSO_BINARY_TYPE__BUILD_ID_CACHE, + DSO_BINARY_TYPE__BUILD_ID_CACHE_DEBUGINFO, DSO_BINARY_TYPE__FEDORA_DEBUGINFO, DSO_BINARY_TYPE__UBUNTU_DEBUGINFO, DSO_BINARY_TYPE__BUILDID_DEBUGINFO, @@ -1418,6 +1419,7 @@ static bool dso__is_compatible_symtab_type(struct dso *dso, bool kmod, return kmod && dso->symtab_type == type; case DSO_BINARY_TYPE__BUILD_ID_CACHE: + case DSO_BINARY_TYPE__BUILD_ID_CACHE_DEBUGINFO: return true; case DSO_BINARY_TYPE__NOT_FOUND: @@ -1565,10 +1567,14 @@ int dso__load(struct dso *dso, struct map *map) struct symsrc *ss = &ss_[ss_pos]; bool next_slot = false; bool is_reg; + bool nsexit; int sirc; enum dso_binary_type symtab_type = binary_type_symtab[i]; + nsexit = (symtab_type == DSO_BINARY_TYPE__BUILD_ID_CACHE || + symtab_type == DSO_BINARY_TYPE__BUILD_ID_CACHE_DEBUGINFO); + if (!dso__is_compatible_symtab_type(dso, kmod, symtab_type)) continue; @@ -1576,13 +1582,13 @@ int dso__load(struct dso *dso, struct map *map) root_dir, name, PATH_MAX)) continue; - if (symtab_type == DSO_BINARY_TYPE__BUILD_ID_CACHE) + if (nsexit) nsinfo__mountns_exit(&nsc); is_reg = is_regular_file(name); sirc = symsrc__init(ss, dso, name, symtab_type); - if (symtab_type == DSO_BINARY_TYPE__BUILD_ID_CACHE) + if (nsexit) nsinfo__mountns_enter(dso->nsinfo, &nsc); if (!is_reg || sirc < 0) { @@ -1724,7 +1730,7 @@ int dso__load_vmlinux_path(struct dso *dso, struct map *map) } if (!symbol_conf.ignore_vmlinux_buildid) - filename = dso__build_id_filename(dso, NULL, 0); + filename = dso__build_id_filename(dso, NULL, 0, false); if (filename != NULL) { err = dso__load_vmlinux(dso, map, filename, true); if (err > 0) -- cgit v1.1 From 30269dc1a17df5117a2e267835863d4466c3569d Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Mon, 3 Jul 2017 13:05:43 -0300 Subject: perf evsel: Allow asking for max precise_ip in new_cycles() There are cases where we want to leave attr.precise_ip as zero, such as when using 'perf record --no-samples', where this would make the kernel return -EINVAL. Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-4zq1udecxa51gsapyfwej5fj@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/evlist.c | 2 +- tools/perf/util/evsel.c | 7 +++++-- tools/perf/util/evsel.h | 2 +- 3 files changed, 7 insertions(+), 4 deletions(-) (limited to 'tools') diff --git a/tools/perf/util/evlist.c b/tools/perf/util/evlist.c index 46c0faf..f51c331 100644 --- a/tools/perf/util/evlist.c +++ b/tools/perf/util/evlist.c @@ -244,7 +244,7 @@ void perf_event_attr__set_max_precise_ip(struct perf_event_attr *attr) int perf_evlist__add_default(struct perf_evlist *evlist) { - struct perf_evsel *evsel = perf_evsel__new_cycles(); + struct perf_evsel *evsel = perf_evsel__new_cycles(true); if (evsel == NULL) return -ENOMEM; diff --git a/tools/perf/util/evsel.c b/tools/perf/util/evsel.c index 0e4cd60..df567e4 100644 --- a/tools/perf/util/evsel.c +++ b/tools/perf/util/evsel.c @@ -268,7 +268,7 @@ struct perf_evsel *perf_evsel__new_idx(struct perf_event_attr *attr, int idx) return evsel; } -struct perf_evsel *perf_evsel__new_cycles(void) +struct perf_evsel *perf_evsel__new_cycles(bool precise) { struct perf_event_attr attr = { .type = PERF_TYPE_HARDWARE, @@ -278,6 +278,9 @@ struct perf_evsel *perf_evsel__new_cycles(void) struct perf_evsel *evsel; event_attr_init(&attr); + + if (!precise) + goto new_event; /* * Unnamed union member, not supported as struct member named * initializer in older compilers such as gcc 4.4.7 @@ -292,7 +295,7 @@ struct perf_evsel *perf_evsel__new_cycles(void) * to kick in when we return and before perf_evsel__open() is called. */ attr.sample_period = 0; - +new_event: evsel = perf_evsel__new(&attr); if (evsel == NULL) goto out; diff --git a/tools/perf/util/evsel.h b/tools/perf/util/evsel.h index 219ad0c..fb40ca3 100644 --- a/tools/perf/util/evsel.h +++ b/tools/perf/util/evsel.h @@ -185,7 +185,7 @@ static inline struct perf_evsel *perf_evsel__newtp(const char *sys, const char * return perf_evsel__newtp_idx(sys, name, 0); } -struct perf_evsel *perf_evsel__new_cycles(void); +struct perf_evsel *perf_evsel__new_cycles(bool precise); struct event_format *event_format__new(const char *sys, const char *name); -- cgit v1.1 From db918acb356b8770a0fafbe96743031a56b91af6 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Mon, 3 Jul 2017 13:10:20 -0300 Subject: perf evlist: Allow asking for max precise_ip in add_default() There are cases where we want to leave attr.precise_ip as zero, such as when using 'perf record --no-samples', where this would make the kernel return -EINVAL. Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-0u2m2a8rqw781r6m8svqyne8@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/evlist.c | 4 ++-- tools/perf/util/evlist.h | 9 ++++++++- 2 files changed, 10 insertions(+), 3 deletions(-) (limited to 'tools') diff --git a/tools/perf/util/evlist.c b/tools/perf/util/evlist.c index f51c331..078b585 100644 --- a/tools/perf/util/evlist.c +++ b/tools/perf/util/evlist.c @@ -242,9 +242,9 @@ void perf_event_attr__set_max_precise_ip(struct perf_event_attr *attr) } } -int perf_evlist__add_default(struct perf_evlist *evlist) +int __perf_evlist__add_default(struct perf_evlist *evlist, bool precise) { - struct perf_evsel *evsel = perf_evsel__new_cycles(true); + struct perf_evsel *evsel = perf_evsel__new_cycles(precise); if (evsel == NULL) return -ENOMEM; diff --git a/tools/perf/util/evlist.h b/tools/perf/util/evlist.h index 8d601fb..0843746 100644 --- a/tools/perf/util/evlist.h +++ b/tools/perf/util/evlist.h @@ -115,7 +115,14 @@ void perf_evlist__delete(struct perf_evlist *evlist); void perf_evlist__add(struct perf_evlist *evlist, struct perf_evsel *entry); void perf_evlist__remove(struct perf_evlist *evlist, struct perf_evsel *evsel); -int perf_evlist__add_default(struct perf_evlist *evlist); + +int __perf_evlist__add_default(struct perf_evlist *evlist, bool precise); + +static inline int perf_evlist__add_default(struct perf_evlist *evlist) +{ + return __perf_evlist__add_default(evlist, true); +} + int __perf_evlist__add_default_attrs(struct perf_evlist *evlist, struct perf_event_attr *attrs, size_t nr_attrs); -- cgit v1.1 From 4b4cd50319127b3b606b8aaf3426d8972df1666a Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Mon, 3 Jul 2017 13:26:32 -0300 Subject: perf record: Do not ask for precise_ip with --no-samples When the user doesn't specify an event with -e/--event, 'perf record' will use as a default the "cycles" event with the highest level of precision in perf_event_attr.precise_ip, but --no-samples, if present, is incompatible with precise_ip != 0, so use the newly introduced __perf_event__add_default(precise = false) to fix that: Before: # perf record -n usleep 1 Please consider tweaking /proc/sys/kernel/perf_event_max_sample_rate. Error: The sys_perf_event_open() syscall returned with 22 (Invalid argument) for event (cycles:ppp). /bin/dmesg may provide additional information. No CONFIG_PERF_EVENTS=y kernel support configured? # After: # perf record -n usleep 1 Please consider tweaking /proc/sys/kernel/perf_event_max_sample_rate. [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.018 MB perf.data ] [root@jouet /]# perf evlist -v cycles: size: 112, sample_type: IP|TID|TIME|PERIOD, disabled: 1, inherit: 1, mmap: 1, comm: 1, freq: 1, enable_on_exec: 1, task: 1, sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1 [root@jouet /]# Reported-by: Jiri Olsa Cc: Adrian Hunter Cc: David Ahern Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-q991fw6s6rhjvrd5ye4t7qom@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-record.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'tools') diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c index 17a14bc..64eef9a5 100644 --- a/tools/perf/builtin-record.c +++ b/tools/perf/builtin-record.c @@ -1821,7 +1821,7 @@ int cmd_record(int argc, const char **argv) record.opts.tail_synthesize = true; if (rec->evlist->nr_entries == 0 && - perf_evlist__add_default(rec->evlist) < 0) { + __perf_evlist__add_default(rec->evlist, !record.opts.no_samples) < 0) { pr_err("Not enough memory for event selector list\n"); goto out; } -- cgit v1.1 From 8526bafc149ee4d0df5eabca0f440164ec705c99 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Sat, 1 Jul 2017 11:21:59 -0300 Subject: perf test sdt: Handle realpath() failure It can return NULL, in which case we should bail out and remove the directory created with mkdtemp(), which is stored in the "__tempdir" variable, not in "tempdir". Cc: Masami Hiramatsu Fixes: 8e5dc848356e ("perf test: Add a test case for SDT event") Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/tests/sdt.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'tools') diff --git a/tools/perf/tests/sdt.c b/tools/perf/tests/sdt.c index 5abe8ce..d801715 100644 --- a/tools/perf/tests/sdt.c +++ b/tools/perf/tests/sdt.c @@ -83,6 +83,8 @@ int test__sdt_event(int subtests __maybe_unused) } /* Note that buildid_dir must be an absolute path */ tempdir = realpath(__tempdir, NULL); + if (tempdir == NULL) + goto error_rmdir; /* At first, scan itself */ set_buildid_dir(tempdir); @@ -100,7 +102,7 @@ int test__sdt_event(int subtests __maybe_unused) error_rmdir: /* Cleanup temporary buildid dir */ - rm_rf(tempdir); + rm_rf(__tempdir); error: free(tempdir); free(myself); -- cgit v1.1 From d78ada4a767a744cad5efa210b8acf57748b91d7 Mon Sep 17 00:00:00 2001 From: Jiri Olsa Date: Mon, 3 Jul 2017 16:50:17 +0200 Subject: perf tests attr: Do not store failed events Do not mess up our temp space with files we don't need - failed event open attempts. Signed-off-by: Jiri Olsa Cc: David Ahern Cc: Namhyung Kim Cc: Peter Zijlstra Cc: Thomas Richter Link: http://lkml.kernel.org/r/20170703145030.12903-3-jolsa@kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/tests/attr.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'tools') diff --git a/tools/perf/tests/attr.c b/tools/perf/tests/attr.c index 0e77b2c..08b1d5c 100644 --- a/tools/perf/tests/attr.c +++ b/tools/perf/tests/attr.c @@ -136,7 +136,7 @@ void test_attr__open(struct perf_event_attr *attr, pid_t pid, int cpu, { int errno_saved = errno; - if (store_event(attr, pid, cpu, fd, group_fd, flags)) { + if ((fd != -1) && store_event(attr, pid, cpu, fd, group_fd, flags)) { pr_err("test attr FAILED"); exit(128); } -- cgit v1.1 From 10213e2ff2e0453324689eb50e73df3df1ee6163 Mon Sep 17 00:00:00 2001 From: Jiri Olsa Date: Mon, 3 Jul 2017 16:50:18 +0200 Subject: perf tests attr: Add test_attr__ready function We create many test events before the real ones just to test specific features. But there's no way for attr tests to separate those test events from those it needs to check. Adding 'ready' call from the events open interface to trigger/start events collection for attr test. Signed-off-by: Jiri Olsa Cc: David Ahern Cc: Namhyung Kim Cc: Peter Zijlstra Cc: Thomas Richter Link: http://lkml.kernel.org/r/20170703145030.12903-4-jolsa@kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/perf.h | 1 + tools/perf/tests/attr.c | 10 ++++++++++ tools/perf/util/evsel.c | 4 ++++ 3 files changed, 15 insertions(+) (limited to 'tools') diff --git a/tools/perf/perf.h b/tools/perf/perf.h index 806c216..2c010dd 100644 --- a/tools/perf/perf.h +++ b/tools/perf/perf.h @@ -7,6 +7,7 @@ #include extern bool test_attr__enabled; +void test_attr__ready(void); void test_attr__init(void); void test_attr__open(struct perf_event_attr *attr, pid_t pid, int cpu, int fd, int group_fd, unsigned long flags); diff --git a/tools/perf/tests/attr.c b/tools/perf/tests/attr.c index 08b1d5c..84c0eb5 100644 --- a/tools/perf/tests/attr.c +++ b/tools/perf/tests/attr.c @@ -36,6 +36,7 @@ #define ENV "PERF_TEST_ATTR" static char *dir; +static bool ready; void test_attr__init(void) { @@ -67,6 +68,9 @@ static int store_event(struct perf_event_attr *attr, pid_t pid, int cpu, FILE *file; char path[PATH_MAX]; + if (!ready) + return 0; + snprintf(path, PATH_MAX, "%s/event-%d-%llu-%d", dir, attr->type, attr->config, fd); @@ -144,6 +148,12 @@ void test_attr__open(struct perf_event_attr *attr, pid_t pid, int cpu, errno = errno_saved; } +void test_attr__ready(void) +{ + if (unlikely(test_attr__enabled) && !ready) + ready = true; +} + static int run_dir(const char *d, const char *perf) { char v[] = "-vvvvv"; diff --git a/tools/perf/util/evsel.c b/tools/perf/util/evsel.c index df567e4..6dd069a 100644 --- a/tools/perf/util/evsel.c +++ b/tools/perf/util/evsel.c @@ -58,6 +58,8 @@ static int perf_evsel__no_extra_init(struct perf_evsel *evsel __maybe_unused) return 0; } +void __weak test_attr__ready(void) { } + static void perf_evsel__no_extra_fini(struct perf_evsel *evsel __maybe_unused) { } @@ -1572,6 +1574,8 @@ retry_open: pr_debug2("sys_perf_event_open: pid %d cpu %d group_fd %d flags %#lx", pid, cpus->map[cpu], group_fd, flags); + test_attr__ready(); + fd = sys_perf_event_open(&evsel->attr, pid, cpus->map[cpu], group_fd, flags); -- cgit v1.1 From 04c31bcf86c4fa51022ec254827f1c7c41306167 Mon Sep 17 00:00:00 2001 From: Jiri Olsa Date: Mon, 3 Jul 2017 16:50:19 +0200 Subject: perf tests attr: Make compare_data global Making compare_data global, so it could be used outside the Test class scope to compare command results. Signed-off-by: Jiri Olsa Cc: David Ahern Cc: Namhyung Kim Cc: Peter Zijlstra Cc: Thomas Richter Link: http://lkml.kernel.org/r/20170703145030.12903-5-jolsa@kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/tests/attr.py | 36 ++++++++++++++++++------------------ 1 file changed, 18 insertions(+), 18 deletions(-) (limited to 'tools') diff --git a/tools/perf/tests/attr.py b/tools/perf/tests/attr.py index cdf21a9..bb671cd 100644 --- a/tools/perf/tests/attr.py +++ b/tools/perf/tests/attr.py @@ -9,6 +9,20 @@ import logging import shutil import ConfigParser +def compare_data(a, b): + # Allow multiple values in assignment separated by '|' + a_list = a.split('|') + b_list = b.split('|') + + for a_item in a_list: + for b_item in b_list: + if (a_item == b_item): + return True + elif (a_item == '*') or (b_item == '*'): + return True + + return False + class Fail(Exception): def __init__(self, test, msg): self.msg = msg @@ -82,26 +96,12 @@ class Event(dict): self.add(base) self.add(data) - def compare_data(self, a, b): - # Allow multiple values in assignment separated by '|' - a_list = a.split('|') - b_list = b.split('|') - - for a_item in a_list: - for b_item in b_list: - if (a_item == b_item): - return True - elif (a_item == '*') or (b_item == '*'): - return True - - return False - def equal(self, other): for t in Event.terms: log.debug(" [%s] %s %s" % (t, self[t], other[t])); if not self.has_key(t) or not other.has_key(t): return False - if not self.compare_data(self[t], other[t]): + if not compare_data(self[t], other[t]): return False return True @@ -109,7 +109,7 @@ class Event(dict): for t in Event.terms: if not self.has_key(t) or not other.has_key(t): continue - if not self.compare_data(self[t], other[t]): + if not compare_data(self[t], other[t]): log.warning("expected %s=%s, got %s" % (t, self[t], other[t])) # Test file description needs to have following sections: @@ -218,9 +218,9 @@ class Test(object): self.perf, self.command, tempdir, self.args) ret = os.WEXITSTATUS(os.system(cmd)) - log.info(" '%s' ret %d " % (cmd, ret)) + log.info(" '%s' ret '%s', expected '%s'" % (cmd, str(ret), str(self.ret))) - if ret != int(self.ret): + if not compare_data(str(ret), str(self.ret)): raise Unsup(self) def compare(self, expect, result): -- cgit v1.1 From dde622a5063c71a1864a41dfaed4baaabfa84ba6 Mon Sep 17 00:00:00 2001 From: Jiri Olsa Date: Mon, 3 Jul 2017 16:50:20 +0200 Subject: perf tests attr: Rename compare_data to data_equal The data_equal name fits better to the return value of the function. It's true when the data is equal. Signed-off-by: Jiri Olsa Cc: David Ahern Cc: Namhyung Kim Cc: Peter Zijlstra Cc: Thomas Richter Link: http://lkml.kernel.org/r/20170703145030.12903-6-jolsa@kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/tests/attr.py | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) (limited to 'tools') diff --git a/tools/perf/tests/attr.py b/tools/perf/tests/attr.py index bb671cd..b03261c 100644 --- a/tools/perf/tests/attr.py +++ b/tools/perf/tests/attr.py @@ -9,7 +9,7 @@ import logging import shutil import ConfigParser -def compare_data(a, b): +def data_equal(a, b): # Allow multiple values in assignment separated by '|' a_list = a.split('|') b_list = b.split('|') @@ -101,7 +101,7 @@ class Event(dict): log.debug(" [%s] %s %s" % (t, self[t], other[t])); if not self.has_key(t) or not other.has_key(t): return False - if not compare_data(self[t], other[t]): + if not data_equal(self[t], other[t]): return False return True @@ -109,7 +109,7 @@ class Event(dict): for t in Event.terms: if not self.has_key(t) or not other.has_key(t): continue - if not compare_data(self[t], other[t]): + if not data_equal(self[t], other[t]): log.warning("expected %s=%s, got %s" % (t, self[t], other[t])) # Test file description needs to have following sections: @@ -220,7 +220,7 @@ class Test(object): log.info(" '%s' ret '%s', expected '%s'" % (cmd, str(ret), str(self.ret))) - if not compare_data(str(ret), str(self.ret)): + if not data_equal(str(ret), str(self.ret)): raise Unsup(self) def compare(self, expect, result): -- cgit v1.1 From 5ced95b2374b0fb3c1d0367c6e8cf82d41292990 Mon Sep 17 00:00:00 2001 From: Jiri Olsa Date: Mon, 3 Jul 2017 16:50:21 +0200 Subject: perf tests attr: Add 1s for exclude_kernel and task base bits There's an event open fallback which set exclude_kernel=1 in case use does not have enough privileges. Adding both 0|1 for this attribute, because we don't know what value it is. Signed-off-by: Jiri Olsa Cc: David Ahern Cc: Namhyung Kim Cc: Peter Zijlstra Cc: Thomas Richter Link: http://lkml.kernel.org/r/20170703145030.12903-7-jolsa@kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/tests/attr/base-record | 2 +- tools/perf/tests/attr/base-stat | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) (limited to 'tools') diff --git a/tools/perf/tests/attr/base-record b/tools/perf/tests/attr/base-record index 7e6d749..cdf3adf7 100644 --- a/tools/perf/tests/attr/base-record +++ b/tools/perf/tests/attr/base-record @@ -15,7 +15,7 @@ inherit=1 pinned=0 exclusive=0 exclude_user=0 -exclude_kernel=0 +exclude_kernel=0|1 exclude_hv=0 exclude_idle=0 mmap=1 diff --git a/tools/perf/tests/attr/base-stat b/tools/perf/tests/attr/base-stat index f4cf148..ff9b33d 100644 --- a/tools/perf/tests/attr/base-stat +++ b/tools/perf/tests/attr/base-stat @@ -15,7 +15,7 @@ inherit=1 pinned=0 exclusive=0 exclude_user=0 -exclude_kernel=0 +exclude_kernel=0|1 exclude_hv=0 exclude_idle=0 mmap=0 -- cgit v1.1 From d9115e924088b714bb801713751bf7922d799f95 Mon Sep 17 00:00:00 2001 From: Jiri Olsa Date: Mon, 3 Jul 2017 16:50:22 +0200 Subject: perf tests attr: Fix record dwarf test Following commit: commit 5c0cf22477ea ("perf record: Store data mmaps for dwarf unwind") have enabled address sampling for dwarf unwind, we need to reflect that in this test by adding ADDR sample_type and enabling mmap_data. Signed-off-by: Jiri Olsa Cc: David Ahern Cc: Namhyung Kim Cc: Peter Zijlstra Cc: Thomas Richter Link: http://lkml.kernel.org/r/20170703145030.12903-8-jolsa@kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/tests/attr/test-record-graph-dwarf | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'tools') diff --git a/tools/perf/tests/attr/test-record-graph-dwarf b/tools/perf/tests/attr/test-record-graph-dwarf index d6f324e..8321c60 100644 --- a/tools/perf/tests/attr/test-record-graph-dwarf +++ b/tools/perf/tests/attr/test-record-graph-dwarf @@ -3,8 +3,9 @@ command = record args = --call-graph dwarf -- kill >/dev/null 2>&1 [event:base-record] -sample_type=12583 +sample_type=45359 exclude_callchain_user=1 sample_stack_user=8192 # TODO different for each arch, no support for that now sample_regs_user=* +mmap_data=1 -- cgit v1.1 From 44fed277f81ba22e3f2fbcf1501c3b14aeb9f8e4 Mon Sep 17 00:00:00 2001 From: Jiri Olsa Date: Mon, 3 Jul 2017 16:50:23 +0200 Subject: perf tests attr: Fix no-delay test Following commit: commit 509051ea8427 ("perf record: Rename --no-delay to --no-buffering") removed '-D' option and renamed --no-delay into --no-buffering. Fixing that in the attr tests. Signed-off-by: Jiri Olsa Cc: David Ahern Cc: Namhyung Kim Cc: Peter Zijlstra Cc: Thomas Richter Fixes: 509051ea8427 ("perf record: Rename --no-delay to --no-buffering") Link: http://lkml.kernel.org/r/20170703145030.12903-9-jolsa@kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/tests/attr/test-record-no-buffering | 9 +++++++++ tools/perf/tests/attr/test-record-no-delay | 9 --------- 2 files changed, 9 insertions(+), 9 deletions(-) create mode 100644 tools/perf/tests/attr/test-record-no-buffering delete mode 100644 tools/perf/tests/attr/test-record-no-delay (limited to 'tools') diff --git a/tools/perf/tests/attr/test-record-no-buffering b/tools/perf/tests/attr/test-record-no-buffering new file mode 100644 index 0000000..0b0355a --- /dev/null +++ b/tools/perf/tests/attr/test-record-no-buffering @@ -0,0 +1,9 @@ +[config] +command = record +args = --no-buffering kill >/dev/null 2>&1 + +[event:base-record] +sample_period=4000 +sample_type=263 +watermark=0 +wakeup_events=1 diff --git a/tools/perf/tests/attr/test-record-no-delay b/tools/perf/tests/attr/test-record-no-delay deleted file mode 100644 index f253b78..0000000 --- a/tools/perf/tests/attr/test-record-no-delay +++ /dev/null @@ -1,9 +0,0 @@ -[config] -command = record -args = -D kill >/dev/null 2>&1 - -[event:base-record] -sample_period=4000 -sample_type=263 -watermark=0 -wakeup_events=1 -- cgit v1.1 From a72fe0afa18abd8d05d26944b32850775cb6c941 Mon Sep 17 00:00:00 2001 From: Jiri Olsa Date: Mon, 3 Jul 2017 16:50:24 +0200 Subject: perf tests attr: Add proper return values The record command now properly returns the status of the tracee if there's any. We need to properly set the expected return value of the tracee in the attr tests. Signed-off-by: Jiri Olsa Cc: David Ahern Cc: Namhyung Kim Cc: Peter Zijlstra Cc: Thomas Richter Link: http://lkml.kernel.org/r/20170703145030.12903-10-jolsa@kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/tests/attr/test-record-C0 | 1 + tools/perf/tests/attr/test-record-basic | 1 + tools/perf/tests/attr/test-record-branch-any | 1 + tools/perf/tests/attr/test-record-branch-filter-any | 1 + tools/perf/tests/attr/test-record-branch-filter-any_call | 1 + tools/perf/tests/attr/test-record-branch-filter-any_ret | 1 + tools/perf/tests/attr/test-record-branch-filter-hv | 1 + tools/perf/tests/attr/test-record-branch-filter-ind_call | 1 + tools/perf/tests/attr/test-record-branch-filter-k | 1 + tools/perf/tests/attr/test-record-branch-filter-u | 1 + tools/perf/tests/attr/test-record-count | 1 + tools/perf/tests/attr/test-record-data | 1 + tools/perf/tests/attr/test-record-freq | 1 + tools/perf/tests/attr/test-record-graph-default | 1 + tools/perf/tests/attr/test-record-graph-dwarf | 1 + tools/perf/tests/attr/test-record-graph-fp | 1 + tools/perf/tests/attr/test-record-group | 1 + tools/perf/tests/attr/test-record-group-sampling | 1 + tools/perf/tests/attr/test-record-group1 | 1 + tools/perf/tests/attr/test-record-no-buffering | 1 + tools/perf/tests/attr/test-record-no-inherit | 1 + tools/perf/tests/attr/test-record-no-samples | 1 + tools/perf/tests/attr/test-record-period | 1 + tools/perf/tests/attr/test-record-raw | 1 + 24 files changed, 24 insertions(+) (limited to 'tools') diff --git a/tools/perf/tests/attr/test-record-C0 b/tools/perf/tests/attr/test-record-C0 index d6a7e43..cb0a313 100644 --- a/tools/perf/tests/attr/test-record-C0 +++ b/tools/perf/tests/attr/test-record-C0 @@ -1,6 +1,7 @@ [config] command = record args = -C 0 kill >/dev/null 2>&1 +ret = 1 [event:base-record] cpu=0 diff --git a/tools/perf/tests/attr/test-record-basic b/tools/perf/tests/attr/test-record-basic index 55c0428..85a23cf 100644 --- a/tools/perf/tests/attr/test-record-basic +++ b/tools/perf/tests/attr/test-record-basic @@ -1,5 +1,6 @@ [config] command = record args = kill >/dev/null 2>&1 +ret = 1 [event:base-record] diff --git a/tools/perf/tests/attr/test-record-branch-any b/tools/perf/tests/attr/test-record-branch-any index 1421960..a2fe031 100644 --- a/tools/perf/tests/attr/test-record-branch-any +++ b/tools/perf/tests/attr/test-record-branch-any @@ -1,6 +1,7 @@ [config] command = record args = -b kill >/dev/null 2>&1 +ret = 1 [event:base-record] sample_period=4000 diff --git a/tools/perf/tests/attr/test-record-branch-filter-any b/tools/perf/tests/attr/test-record-branch-filter-any index 915c4df..7df6a48 100644 --- a/tools/perf/tests/attr/test-record-branch-filter-any +++ b/tools/perf/tests/attr/test-record-branch-filter-any @@ -1,6 +1,7 @@ [config] command = record args = -j any kill >/dev/null 2>&1 +ret = 1 [event:base-record] sample_period=4000 diff --git a/tools/perf/tests/attr/test-record-branch-filter-any_call b/tools/perf/tests/attr/test-record-branch-filter-any_call index 8708dbd..6c93e07 100644 --- a/tools/perf/tests/attr/test-record-branch-filter-any_call +++ b/tools/perf/tests/attr/test-record-branch-filter-any_call @@ -1,6 +1,7 @@ [config] command = record args = -j any_call kill >/dev/null 2>&1 +ret = 1 [event:base-record] sample_period=4000 diff --git a/tools/perf/tests/attr/test-record-branch-filter-any_ret b/tools/perf/tests/attr/test-record-branch-filter-any_ret index 0d3607a..daf8b83 100644 --- a/tools/perf/tests/attr/test-record-branch-filter-any_ret +++ b/tools/perf/tests/attr/test-record-branch-filter-any_ret @@ -1,6 +1,7 @@ [config] command = record args = -j any_ret kill >/dev/null 2>&1 +ret = 1 [event:base-record] sample_period=4000 diff --git a/tools/perf/tests/attr/test-record-branch-filter-hv b/tools/perf/tests/attr/test-record-branch-filter-hv index f255267..fbf491d 100644 --- a/tools/perf/tests/attr/test-record-branch-filter-hv +++ b/tools/perf/tests/attr/test-record-branch-filter-hv @@ -1,6 +1,7 @@ [config] command = record args = -j hv kill >/dev/null 2>&1 +ret = 1 [event:base-record] sample_period=4000 diff --git a/tools/perf/tests/attr/test-record-branch-filter-ind_call b/tools/perf/tests/attr/test-record-branch-filter-ind_call index e862dd1..c63cc50 100644 --- a/tools/perf/tests/attr/test-record-branch-filter-ind_call +++ b/tools/perf/tests/attr/test-record-branch-filter-ind_call @@ -1,6 +1,7 @@ [config] command = record args = -j ind_call kill >/dev/null 2>&1 +ret = 1 [event:base-record] sample_period=4000 diff --git a/tools/perf/tests/attr/test-record-branch-filter-k b/tools/perf/tests/attr/test-record-branch-filter-k index 182971e..e0f2f09 100644 --- a/tools/perf/tests/attr/test-record-branch-filter-k +++ b/tools/perf/tests/attr/test-record-branch-filter-k @@ -1,6 +1,7 @@ [config] command = record args = -j k kill >/dev/null 2>&1 +ret = 1 [event:base-record] sample_period=4000 diff --git a/tools/perf/tests/attr/test-record-branch-filter-u b/tools/perf/tests/attr/test-record-branch-filter-u index 83449ef..6cd36e0 100644 --- a/tools/perf/tests/attr/test-record-branch-filter-u +++ b/tools/perf/tests/attr/test-record-branch-filter-u @@ -1,6 +1,7 @@ [config] command = record args = -j u kill >/dev/null 2>&1 +ret = 1 [event:base-record] sample_period=4000 diff --git a/tools/perf/tests/attr/test-record-count b/tools/perf/tests/attr/test-record-count index 2f841de..34f6cc5 100644 --- a/tools/perf/tests/attr/test-record-count +++ b/tools/perf/tests/attr/test-record-count @@ -1,6 +1,7 @@ [config] command = record args = -c 123 kill >/dev/null 2>&1 +ret = 1 [event:base-record] sample_period=123 diff --git a/tools/perf/tests/attr/test-record-data b/tools/perf/tests/attr/test-record-data index 716e143..53c8890 100644 --- a/tools/perf/tests/attr/test-record-data +++ b/tools/perf/tests/attr/test-record-data @@ -1,6 +1,7 @@ [config] command = record args = -d kill >/dev/null 2>&1 +ret = 1 [event:base-record] sample_period=4000 diff --git a/tools/perf/tests/attr/test-record-freq b/tools/perf/tests/attr/test-record-freq index 600d0f8..bf4cb45 100644 --- a/tools/perf/tests/attr/test-record-freq +++ b/tools/perf/tests/attr/test-record-freq @@ -1,6 +1,7 @@ [config] command = record args = -F 100 kill >/dev/null 2>&1 +ret = 1 [event:base-record] sample_period=100 diff --git a/tools/perf/tests/attr/test-record-graph-default b/tools/perf/tests/attr/test-record-graph-default index 853597a..0b216e6 100644 --- a/tools/perf/tests/attr/test-record-graph-default +++ b/tools/perf/tests/attr/test-record-graph-default @@ -1,6 +1,7 @@ [config] command = record args = -g kill >/dev/null 2>&1 +ret = 1 [event:base-record] sample_type=295 diff --git a/tools/perf/tests/attr/test-record-graph-dwarf b/tools/perf/tests/attr/test-record-graph-dwarf index 8321c60..da2fa73 100644 --- a/tools/perf/tests/attr/test-record-graph-dwarf +++ b/tools/perf/tests/attr/test-record-graph-dwarf @@ -1,6 +1,7 @@ [config] command = record args = --call-graph dwarf -- kill >/dev/null 2>&1 +ret = 1 [event:base-record] sample_type=45359 diff --git a/tools/perf/tests/attr/test-record-graph-fp b/tools/perf/tests/attr/test-record-graph-fp index 055e3be..625d190 100644 --- a/tools/perf/tests/attr/test-record-graph-fp +++ b/tools/perf/tests/attr/test-record-graph-fp @@ -1,6 +1,7 @@ [config] command = record args = --call-graph fp kill >/dev/null 2>&1 +ret = 1 [event:base-record] sample_type=295 diff --git a/tools/perf/tests/attr/test-record-group b/tools/perf/tests/attr/test-record-group index 57739ca..6e7961f 100644 --- a/tools/perf/tests/attr/test-record-group +++ b/tools/perf/tests/attr/test-record-group @@ -1,6 +1,7 @@ [config] command = record args = --group -e cycles,instructions kill >/dev/null 2>&1 +ret = 1 [event-1:base-record] fd=1 diff --git a/tools/perf/tests/attr/test-record-group-sampling b/tools/perf/tests/attr/test-record-group-sampling index 658f5d6..ef59afd 100644 --- a/tools/perf/tests/attr/test-record-group-sampling +++ b/tools/perf/tests/attr/test-record-group-sampling @@ -1,6 +1,7 @@ [config] command = record args = -e '{cycles,cache-misses}:S' kill >/dev/null 2>&1 +ret = 1 [event-1:base-record] fd=1 diff --git a/tools/perf/tests/attr/test-record-group1 b/tools/perf/tests/attr/test-record-group1 index c5548d0..87a222d 100644 --- a/tools/perf/tests/attr/test-record-group1 +++ b/tools/perf/tests/attr/test-record-group1 @@ -1,6 +1,7 @@ [config] command = record args = -e '{cycles,instructions}' kill >/dev/null 2>&1 +ret = 1 [event-1:base-record] fd=1 diff --git a/tools/perf/tests/attr/test-record-no-buffering b/tools/perf/tests/attr/test-record-no-buffering index 0b0355a..5bf349a 100644 --- a/tools/perf/tests/attr/test-record-no-buffering +++ b/tools/perf/tests/attr/test-record-no-buffering @@ -1,6 +1,7 @@ [config] command = record args = --no-buffering kill >/dev/null 2>&1 +ret = 1 [event:base-record] sample_period=4000 diff --git a/tools/perf/tests/attr/test-record-no-inherit b/tools/perf/tests/attr/test-record-no-inherit index 44edcb2..560943d 100644 --- a/tools/perf/tests/attr/test-record-no-inherit +++ b/tools/perf/tests/attr/test-record-no-inherit @@ -1,6 +1,7 @@ [config] command = record args = -i kill >/dev/null 2>&1 +ret = 1 [event:base-record] sample_type=263 diff --git a/tools/perf/tests/attr/test-record-no-samples b/tools/perf/tests/attr/test-record-no-samples index d0141b2..8eb73ab 100644 --- a/tools/perf/tests/attr/test-record-no-samples +++ b/tools/perf/tests/attr/test-record-no-samples @@ -1,6 +1,7 @@ [config] command = record args = -n kill >/dev/null 2>&1 +ret = 1 [event:base-record] sample_period=0 diff --git a/tools/perf/tests/attr/test-record-period b/tools/perf/tests/attr/test-record-period index 8abc531..69bc748 100644 --- a/tools/perf/tests/attr/test-record-period +++ b/tools/perf/tests/attr/test-record-period @@ -1,6 +1,7 @@ [config] command = record args = -c 100 -P kill >/dev/null 2>&1 +ret = 1 [event:base-record] sample_period=100 diff --git a/tools/perf/tests/attr/test-record-raw b/tools/perf/tests/attr/test-record-raw index 4a8ef25..3b89f99 100644 --- a/tools/perf/tests/attr/test-record-raw +++ b/tools/perf/tests/attr/test-record-raw @@ -1,6 +1,7 @@ [config] command = record args = -R kill >/dev/null 2>&1 +ret = 1 [event:base-record] sample_period=4000 -- cgit v1.1 From 6f193400eaf1de53145cc13ff69f88336e9a9c9a Mon Sep 17 00:00:00 2001 From: Jiri Olsa Date: Mon, 3 Jul 2017 16:50:25 +0200 Subject: perf tests attr: Fix cpu test disabled term setup The stat command creates all events disabled and enables them either manualy or via the enable_on_exec bit. Signed-off-by: Jiri Olsa Cc: David Ahern Cc: Namhyung Kim Cc: Peter Zijlstra Cc: Thomas Richter Link: http://lkml.kernel.org/r/20170703145030.12903-11-jolsa@kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/tests/attr/test-stat-C0 | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'tools') diff --git a/tools/perf/tests/attr/test-stat-C0 b/tools/perf/tests/attr/test-stat-C0 index aa83595..67717fe 100644 --- a/tools/perf/tests/attr/test-stat-C0 +++ b/tools/perf/tests/attr/test-stat-C0 @@ -4,6 +4,6 @@ args = -e cycles -C 0 kill >/dev/null 2>&1 ret = 1 [event:base-stat] -# events are enabled by default when attached to cpu -disabled=0 +# events are disabled by default when attached to cpu +disabled=1 enable_on_exec=0 -- cgit v1.1 From 5ff0cf421c37f53bb658a097c610cc1e3ced1133 Mon Sep 17 00:00:00 2001 From: Jiri Olsa Date: Mon, 3 Jul 2017 16:50:26 +0200 Subject: perf tests attr: Fix sample_period setup The final period can differ from what user specifies on command line due to the perf_event_max_sample_rate sysctl setup. Thus we can't predixt the sample_period value any more. Signed-off-by: Jiri Olsa Cc: David Ahern Cc: Namhyung Kim Cc: Peter Zijlstra Cc: Thomas Richter Link: http://lkml.kernel.org/r/20170703145030.12903-12-jolsa@kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/tests/attr/base-record | 2 +- tools/perf/tests/attr/test-record-branch-any | 1 - tools/perf/tests/attr/test-record-branch-filter-any | 1 - tools/perf/tests/attr/test-record-branch-filter-any_call | 1 - tools/perf/tests/attr/test-record-branch-filter-any_ret | 1 - tools/perf/tests/attr/test-record-branch-filter-hv | 1 - tools/perf/tests/attr/test-record-branch-filter-ind_call | 1 - tools/perf/tests/attr/test-record-branch-filter-k | 1 - tools/perf/tests/attr/test-record-branch-filter-u | 1 - tools/perf/tests/attr/test-record-data | 2 -- tools/perf/tests/attr/test-record-no-buffering | 1 - tools/perf/tests/attr/test-record-raw | 1 - 12 files changed, 1 insertion(+), 13 deletions(-) (limited to 'tools') diff --git a/tools/perf/tests/attr/base-record b/tools/perf/tests/attr/base-record index cdf3adf7..a9cb72b 100644 --- a/tools/perf/tests/attr/base-record +++ b/tools/perf/tests/attr/base-record @@ -7,7 +7,7 @@ cpu=* type=0|1 size=112 config=0 -sample_period=4000 +sample_period=* sample_type=263 read_format=0 disabled=1 diff --git a/tools/perf/tests/attr/test-record-branch-any b/tools/perf/tests/attr/test-record-branch-any index a2fe031..81f839e 100644 --- a/tools/perf/tests/attr/test-record-branch-any +++ b/tools/perf/tests/attr/test-record-branch-any @@ -4,6 +4,5 @@ args = -b kill >/dev/null 2>&1 ret = 1 [event:base-record] -sample_period=4000 sample_type=2311 branch_sample_type=8 diff --git a/tools/perf/tests/attr/test-record-branch-filter-any b/tools/perf/tests/attr/test-record-branch-filter-any index 7df6a48..357421f 100644 --- a/tools/perf/tests/attr/test-record-branch-filter-any +++ b/tools/perf/tests/attr/test-record-branch-filter-any @@ -4,6 +4,5 @@ args = -j any kill >/dev/null 2>&1 ret = 1 [event:base-record] -sample_period=4000 sample_type=2311 branch_sample_type=8 diff --git a/tools/perf/tests/attr/test-record-branch-filter-any_call b/tools/perf/tests/attr/test-record-branch-filter-any_call index 6c93e07..dbc55f2 100644 --- a/tools/perf/tests/attr/test-record-branch-filter-any_call +++ b/tools/perf/tests/attr/test-record-branch-filter-any_call @@ -4,6 +4,5 @@ args = -j any_call kill >/dev/null 2>&1 ret = 1 [event:base-record] -sample_period=4000 sample_type=2311 branch_sample_type=16 diff --git a/tools/perf/tests/attr/test-record-branch-filter-any_ret b/tools/perf/tests/attr/test-record-branch-filter-any_ret index daf8b83..a0824ff 100644 --- a/tools/perf/tests/attr/test-record-branch-filter-any_ret +++ b/tools/perf/tests/attr/test-record-branch-filter-any_ret @@ -4,6 +4,5 @@ args = -j any_ret kill >/dev/null 2>&1 ret = 1 [event:base-record] -sample_period=4000 sample_type=2311 branch_sample_type=32 diff --git a/tools/perf/tests/attr/test-record-branch-filter-hv b/tools/perf/tests/attr/test-record-branch-filter-hv index fbf491d..f34d6f1 100644 --- a/tools/perf/tests/attr/test-record-branch-filter-hv +++ b/tools/perf/tests/attr/test-record-branch-filter-hv @@ -4,6 +4,5 @@ args = -j hv kill >/dev/null 2>&1 ret = 1 [event:base-record] -sample_period=4000 sample_type=2311 branch_sample_type=8 diff --git a/tools/perf/tests/attr/test-record-branch-filter-ind_call b/tools/perf/tests/attr/test-record-branch-filter-ind_call index c63cc50..b86a352 100644 --- a/tools/perf/tests/attr/test-record-branch-filter-ind_call +++ b/tools/perf/tests/attr/test-record-branch-filter-ind_call @@ -4,6 +4,5 @@ args = -j ind_call kill >/dev/null 2>&1 ret = 1 [event:base-record] -sample_period=4000 sample_type=2311 branch_sample_type=64 diff --git a/tools/perf/tests/attr/test-record-branch-filter-k b/tools/perf/tests/attr/test-record-branch-filter-k index e0f2f09..d3fbc5e 100644 --- a/tools/perf/tests/attr/test-record-branch-filter-k +++ b/tools/perf/tests/attr/test-record-branch-filter-k @@ -4,6 +4,5 @@ args = -j k kill >/dev/null 2>&1 ret = 1 [event:base-record] -sample_period=4000 sample_type=2311 branch_sample_type=8 diff --git a/tools/perf/tests/attr/test-record-branch-filter-u b/tools/perf/tests/attr/test-record-branch-filter-u index 6cd36e0..a318f0d 100644 --- a/tools/perf/tests/attr/test-record-branch-filter-u +++ b/tools/perf/tests/attr/test-record-branch-filter-u @@ -4,6 +4,5 @@ args = -j u kill >/dev/null 2>&1 ret = 1 [event:base-record] -sample_period=4000 sample_type=2311 branch_sample_type=8 diff --git a/tools/perf/tests/attr/test-record-data b/tools/perf/tests/attr/test-record-data index 53c8890..a9cf223 100644 --- a/tools/perf/tests/attr/test-record-data +++ b/tools/perf/tests/attr/test-record-data @@ -4,8 +4,6 @@ args = -d kill >/dev/null 2>&1 ret = 1 [event:base-record] -sample_period=4000 - # sample_type = PERF_SAMPLE_IP | PERF_SAMPLE_TID | PERF_SAMPLE_TIME | # PERF_SAMPLE_ADDR | PERF_SAMPLE_PERIOD | PERF_SAMPLE_DATA_SRC sample_type=33039 diff --git a/tools/perf/tests/attr/test-record-no-buffering b/tools/perf/tests/attr/test-record-no-buffering index 5bf349a..aa3956d 100644 --- a/tools/perf/tests/attr/test-record-no-buffering +++ b/tools/perf/tests/attr/test-record-no-buffering @@ -4,7 +4,6 @@ args = --no-buffering kill >/dev/null 2>&1 ret = 1 [event:base-record] -sample_period=4000 sample_type=263 watermark=0 wakeup_events=1 diff --git a/tools/perf/tests/attr/test-record-raw b/tools/perf/tests/attr/test-record-raw index 3b89f99..a188a61 100644 --- a/tools/perf/tests/attr/test-record-raw +++ b/tools/perf/tests/attr/test-record-raw @@ -4,5 +4,4 @@ args = -R kill >/dev/null 2>&1 ret = 1 [event:base-record] -sample_period=4000 sample_type=1415 -- cgit v1.1 From 042049404f7a149446d5c839a088fb928b11ce7a Mon Sep 17 00:00:00 2001 From: Jiri Olsa Date: Mon, 3 Jul 2017 16:50:27 +0200 Subject: perf tests attr: Fix precise_ip setup We have a test to detect to highest precise possible, so test can't just predict precise_ip value. Signed-off-by: Jiri Olsa Cc: David Ahern Cc: Namhyung Kim Cc: Peter Zijlstra Cc: Thomas Richter Link: http://lkml.kernel.org/r/20170703145030.12903-13-jolsa@kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/tests/attr/base-record | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'tools') diff --git a/tools/perf/tests/attr/base-record b/tools/perf/tests/attr/base-record index a9cb72b..31e0b1d 100644 --- a/tools/perf/tests/attr/base-record +++ b/tools/perf/tests/attr/base-record @@ -25,7 +25,7 @@ inherit_stat=0 enable_on_exec=1 task=0 watermark=0 -precise_ip=0 +precise_ip=0|1|2|3 mmap_data=0 sample_id_all=1 exclude_host=0|1 -- cgit v1.1 From b78e92e6071d5131f5ba6b25437d82c0fcef09de Mon Sep 17 00:00:00 2001 From: Jiri Olsa Date: Mon, 3 Jul 2017 16:50:28 +0200 Subject: perf tests attr: Fix stat sample_type setup >From following commit: commit 4979d0c7d0c7 ("perf stat record: Add record command") we started to assign PERF_SAMPLE_IDENTIFIER to sample_type. Fixing the attr stat tests accordingly. Signed-off-by: Jiri Olsa Cc: David Ahern Cc: Namhyung Kim Cc: Peter Zijlstra Cc: Thomas Richter Link: http://lkml.kernel.org/r/20170703145030.12903-14-jolsa@kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/tests/attr/base-stat | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'tools') diff --git a/tools/perf/tests/attr/base-stat b/tools/perf/tests/attr/base-stat index ff9b33d..4d0c2e4 100644 --- a/tools/perf/tests/attr/base-stat +++ b/tools/perf/tests/attr/base-stat @@ -8,7 +8,7 @@ type=0 size=112 config=0 sample_period=0 -sample_type=0 +sample_type=65536 read_format=3 disabled=1 inherit=1 -- cgit v1.1 From 1f41873c22826fc92d4bb03c58cf44664fdc3bb8 Mon Sep 17 00:00:00 2001 From: Jiri Olsa Date: Mon, 3 Jul 2017 16:50:29 +0200 Subject: perf tests attr: Add optional term Some of the stat events are quite rare to find on common machines (like front end cycles). Adding an 'optional' term to mark such events in attr tests. Event marked as optional will not fail the test case if it's not found in results. Signed-off-by: Jiri Olsa Cc: David Ahern Cc: Namhyung Kim Cc: Peter Zijlstra Cc: Thomas Richter Link: http://lkml.kernel.org/r/20170703145030.12903-15-jolsa@kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/tests/attr.py | 14 +++++++++++--- tools/perf/tests/attr/test-stat-default | 2 ++ tools/perf/tests/attr/test-stat-detailed-1 | 2 ++ tools/perf/tests/attr/test-stat-detailed-2 | 3 +++ tools/perf/tests/attr/test-stat-detailed-3 | 5 +++++ 5 files changed, 23 insertions(+), 3 deletions(-) (limited to 'tools') diff --git a/tools/perf/tests/attr.py b/tools/perf/tests/attr.py index b03261c..6bb50e8 100644 --- a/tools/perf/tests/attr.py +++ b/tools/perf/tests/attr.py @@ -105,6 +105,11 @@ class Event(dict): return False return True + def optional(self): + if self.has_key('optional') and self['optional'] == '1': + return True + return False + def diff(self, other): for t in Event.terms: if not self.has_key(t) or not other.has_key(t): @@ -244,9 +249,12 @@ class Test(object): log.debug(" match: [%s] matches %s" % (exp_name, str(exp_list))) # we did not any matching event - fail - if (not exp_list): - exp_event.diff(res_event) - raise Fail(self, 'match failure'); + if not exp_list: + if exp_event.optional(): + log.debug(" %s does not match, but is optional" % exp_name) + else: + exp_event.diff(res_event) + raise Fail(self, 'match failure'); match[exp_name] = exp_list diff --git a/tools/perf/tests/attr/test-stat-default b/tools/perf/tests/attr/test-stat-default index 19270f5..e911dbd 100644 --- a/tools/perf/tests/attr/test-stat-default +++ b/tools/perf/tests/attr/test-stat-default @@ -38,12 +38,14 @@ config=0 fd=6 type=0 config=7 +optional=1 # PERF_TYPE_HARDWARE / PERF_COUNT_HW_STALLED_CYCLES_BACKEND [event7:base-stat] fd=7 type=0 config=8 +optional=1 # PERF_TYPE_HARDWARE / PERF_COUNT_HW_INSTRUCTIONS [event8:base-stat] diff --git a/tools/perf/tests/attr/test-stat-detailed-1 b/tools/perf/tests/attr/test-stat-detailed-1 index 51426b8..b39270a 100644 --- a/tools/perf/tests/attr/test-stat-detailed-1 +++ b/tools/perf/tests/attr/test-stat-detailed-1 @@ -39,12 +39,14 @@ config=0 fd=6 type=0 config=7 +optional=1 # PERF_TYPE_HARDWARE / PERF_COUNT_HW_STALLED_CYCLES_BACKEND [event7:base-stat] fd=7 type=0 config=8 +optional=1 # PERF_TYPE_HARDWARE / PERF_COUNT_HW_INSTRUCTIONS [event8:base-stat] diff --git a/tools/perf/tests/attr/test-stat-detailed-2 b/tools/perf/tests/attr/test-stat-detailed-2 index 8de5acc..45f8e6e 100644 --- a/tools/perf/tests/attr/test-stat-detailed-2 +++ b/tools/perf/tests/attr/test-stat-detailed-2 @@ -39,12 +39,14 @@ config=0 fd=6 type=0 config=7 +optional=1 # PERF_TYPE_HARDWARE / PERF_COUNT_HW_STALLED_CYCLES_BACKEND [event7:base-stat] fd=7 type=0 config=8 +optional=1 # PERF_TYPE_HARDWARE / PERF_COUNT_HW_INSTRUCTIONS [event8:base-stat] @@ -108,6 +110,7 @@ config=65538 fd=15 type=3 config=1 +optional=1 # PERF_TYPE_HW_CACHE, # PERF_COUNT_HW_CACHE_L1I << 0 | diff --git a/tools/perf/tests/attr/test-stat-detailed-3 b/tools/perf/tests/attr/test-stat-detailed-3 index 0a1f45b..30ae0fb 100644 --- a/tools/perf/tests/attr/test-stat-detailed-3 +++ b/tools/perf/tests/attr/test-stat-detailed-3 @@ -39,12 +39,14 @@ config=0 fd=6 type=0 config=7 +optional=1 # PERF_TYPE_HARDWARE / PERF_COUNT_HW_STALLED_CYCLES_BACKEND [event7:base-stat] fd=7 type=0 config=8 +optional=1 # PERF_TYPE_HARDWARE / PERF_COUNT_HW_INSTRUCTIONS [event8:base-stat] @@ -108,6 +110,7 @@ config=65538 fd=15 type=3 config=1 +optional=1 # PERF_TYPE_HW_CACHE, # PERF_COUNT_HW_CACHE_L1I << 0 | @@ -162,6 +165,7 @@ config=65540 fd=21 type=3 config=512 +optional=1 # PERF_TYPE_HW_CACHE, # PERF_COUNT_HW_CACHE_L1D << 0 | @@ -171,3 +175,4 @@ config=512 fd=22 type=3 config=66048 +optional=1 -- cgit v1.1 From 0ae79636e32889cced620865debd3ba5bfd36daa Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Mon, 17 Jul 2017 10:31:42 -0300 Subject: perf trace beauty: Export strarray for use in per-object beautifiers Like will be done with fcntl(fd, F_GETLEASE, F_RDLCK|F_WRLCK|F_UNLCK) Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-3p11bgirtntjfmbixfkz8i2m@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-trace.c | 27 +++++++-------------------- tools/perf/trace/beauty/beauty.h | 20 ++++++++++++++++++++ 2 files changed, 27 insertions(+), 20 deletions(-) (limited to 'tools') diff --git a/tools/perf/builtin-trace.c b/tools/perf/builtin-trace.c index 65fa012..0dedce0 100644 --- a/tools/perf/builtin-trace.c +++ b/tools/perf/builtin-trace.c @@ -283,34 +283,21 @@ out_delete: ({ struct syscall_tp *fields = evsel->priv; \ fields->name.pointer(&fields->name, sample); }) -struct strarray { - int offset; - int nr_entries; - const char **entries; -}; +size_t strarray__scnprintf(struct strarray *sa, char *bf, size_t size, const char *intfmt, int val) +{ + int idx = val - sa->offset; -#define DEFINE_STRARRAY(array) struct strarray strarray__##array = { \ - .nr_entries = ARRAY_SIZE(array), \ - .entries = array, \ -} + if (idx < 0 || idx >= sa->nr_entries) + return scnprintf(bf, size, intfmt, val); -#define DEFINE_STRARRAY_OFFSET(array, off) struct strarray strarray__##array = { \ - .offset = off, \ - .nr_entries = ARRAY_SIZE(array), \ - .entries = array, \ + return scnprintf(bf, size, "%s", sa->entries[idx]); } static size_t __syscall_arg__scnprintf_strarray(char *bf, size_t size, const char *intfmt, struct syscall_arg *arg) { - struct strarray *sa = arg->parm; - int idx = arg->val - sa->offset; - - if (idx < 0 || idx >= sa->nr_entries) - return scnprintf(bf, size, intfmt, arg->val); - - return scnprintf(bf, size, "%s", sa->entries[idx]); + return strarray__scnprintf(arg->parm, bf, size, intfmt, arg->val); } static size_t syscall_arg__scnprintf_strarray(char *bf, size_t size, diff --git a/tools/perf/trace/beauty/beauty.h b/tools/perf/trace/beauty/beauty.h index 9ccf0f3..f75ef7d 100644 --- a/tools/perf/trace/beauty/beauty.h +++ b/tools/perf/trace/beauty/beauty.h @@ -1,8 +1,28 @@ #ifndef _PERF_TRACE_BEAUTY_H #define _PERF_TRACE_BEAUTY_H +#include #include +struct strarray { + int offset; + int nr_entries; + const char **entries; +}; + +#define DEFINE_STRARRAY(array) struct strarray strarray__##array = { \ + .nr_entries = ARRAY_SIZE(array), \ + .entries = array, \ +} + +#define DEFINE_STRARRAY_OFFSET(array, off) struct strarray strarray__##array = { \ + .offset = off, \ + .nr_entries = ARRAY_SIZE(array), \ + .entries = array, \ +} + +size_t strarray__scnprintf(struct strarray *sa, char *bf, size_t size, const char *intfmt, int val); + struct trace; struct thread; -- cgit v1.1 From 65dfa1e7799edd114fbd3949f2783f4e69fd0305 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Mon, 17 Jul 2017 10:46:13 -0300 Subject: perf trace beauty fcntl: Beautify F_GETLEASE and F_SETLEASE arg/return One more looking prettier. Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-ytr7idkese8sjtvn5g60130e@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/trace/beauty/fcntl.c | 22 +++++++++++++++++++++- 1 file changed, 21 insertions(+), 1 deletion(-) (limited to 'tools') diff --git a/tools/perf/trace/beauty/fcntl.c b/tools/perf/trace/beauty/fcntl.c index d082fd2..b6987c4 100644 --- a/tools/perf/trace/beauty/fcntl.c +++ b/tools/perf/trace/beauty/fcntl.c @@ -20,6 +20,19 @@ static size_t syscall_arg__scnprintf_fcntl_getfd(char *bf, size_t size, struct s return fcntl__scnprintf_getfd(arg->val, bf, size); } +static size_t fcntl__scnprintf_getlease(unsigned long val, char *bf, size_t size) +{ + static const char *fcntl_setlease[] = { "RDLCK", "WRLCK", "UNLCK", }; + static DEFINE_STRARRAY(fcntl_setlease); + + return strarray__scnprintf(&strarray__fcntl_setlease, bf, size, "%x", val); +} + +static size_t syscall_arg__scnprintf_fcntl_getlease(char *bf, size_t size, struct syscall_arg *arg) +{ + return fcntl__scnprintf_getlease(arg->val, bf, size); +} + size_t syscall_arg__scnprintf_fcntl_cmd(char *bf, size_t size, struct syscall_arg *arg) { if (arg->val == F_GETFL) { @@ -38,11 +51,15 @@ size_t syscall_arg__scnprintf_fcntl_cmd(char *bf, size_t size, struct syscall_ar syscall_arg__set_ret_scnprintf(arg, syscall_arg__scnprintf_pid); goto mask_arg; } + if (arg->val == F_GETLEASE) { + syscall_arg__set_ret_scnprintf(arg, syscall_arg__scnprintf_fcntl_getlease); + goto mask_arg; + } /* * Some commands ignore the third fcntl argument, "arg", so mask it */ if (arg->val == F_GET_SEALS || - arg->val == F_GETLEASE || arg->val == F_GETSIG) { + arg->val == F_GETSIG) { mask_arg: arg->mask |= (1 << 2); } @@ -62,6 +79,9 @@ size_t syscall_arg__scnprintf_fcntl_arg(char *bf, size_t size, struct syscall_ar if (cmd == F_SETOWN) return syscall_arg__scnprintf_pid(bf, size, arg); + + if (cmd == F_SETLEASE) + return fcntl__scnprintf_getlease(arg->val, bf, size); /* * We still don't grab the contents of pointers on entry or exit, * so just print them as hex numbers -- cgit v1.1 From 82d4a1109fc302795a184a328f60ad28bf7b5989 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Mon, 17 Jul 2017 15:22:47 -0300 Subject: perf trace: Group per syscall arg formatter info into one struct Instead of having syscall_fmt.{arg_scnprintf,arg_parm}, introduce struct syscall_arg_fmt and have these two, paving the way for more state to change the formatting algorithms. For instance, in the 'fcntl' 'cmd' case it is better not to suppress it when being zero, showing instead its name "DUPFD". We had that in an ad-hoc way just for strarrays, but with more involved cases like fcntl, that can't be done with just a strarray, we'll need a ".show_zero" arg in the 'cmd' syscall_arg_fmt. Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-ch06o2j72zbjx5xww4qp67au@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-trace.c | 256 +++++++++++++++++++++++---------------------- 1 file changed, 133 insertions(+), 123 deletions(-) (limited to 'tools') diff --git a/tools/perf/builtin-trace.c b/tools/perf/builtin-trace.c index 0dedce0..32778a6 100644 --- a/tools/perf/builtin-trace.c +++ b/tools/perf/builtin-trace.c @@ -582,9 +582,9 @@ static size_t syscall_arg__scnprintf_getrandom_flags(char *bf, size_t size, #define SCA_GETRANDOM_FLAGS syscall_arg__scnprintf_getrandom_flags -#define STRARRAY(arg, name, array) \ - .arg_scnprintf = { [arg] = SCA_STRARRAY, }, \ - .arg_parm = { [arg] = &strarray__##array, } +#define STRARRAY(name, array) \ + { .scnprintf = SCA_STRARRAY, \ + .parm = &strarray__##array, } #include "trace/beauty/eventfd.c" #include "trace/beauty/flock.c" @@ -601,54 +601,61 @@ static size_t syscall_arg__scnprintf_getrandom_flags(char *bf, size_t size, #include "trace/beauty/socket_type.c" #include "trace/beauty/waitid_options.c" +struct syscall_arg_fmt { + size_t (*scnprintf)(char *bf, size_t size, struct syscall_arg *arg); + void *parm; +}; + static struct syscall_fmt { const char *name; const char *alias; - size_t (*arg_scnprintf[6])(char *bf, size_t size, struct syscall_arg *arg); - void *arg_parm[6]; + struct syscall_arg_fmt arg[6]; bool errmsg; bool errpid; bool timeout; bool hexret; } syscall_fmts[] = { { .name = "access", .errmsg = true, - .arg_scnprintf = { [1] = SCA_ACCMODE, /* mode */ }, }, + .arg = { [1] = { .scnprintf = SCA_ACCMODE, /* mode */ }, }, }, { .name = "arch_prctl", .errmsg = true, .alias = "prctl", }, - { .name = "bpf", .errmsg = true, STRARRAY(0, cmd, bpf_cmd), }, + { .name = "bpf", .errmsg = true, + .arg = { [0] = STRARRAY(cmd, bpf_cmd), }, }, { .name = "brk", .hexret = true, - .arg_scnprintf = { [0] = SCA_HEX, /* brk */ }, }, + .arg = { [0] = { .scnprintf = SCA_HEX, /* brk */ }, }, }, { .name = "chdir", .errmsg = true, }, { .name = "chmod", .errmsg = true, }, { .name = "chroot", .errmsg = true, }, - { .name = "clock_gettime", .errmsg = true, STRARRAY(0, clk_id, clockid), }, + { .name = "clock_gettime", .errmsg = true, + .arg = { [0] = STRARRAY(clk_id, clockid), }, }, { .name = "clone", .errpid = true, }, { .name = "close", .errmsg = true, - .arg_scnprintf = { [0] = SCA_CLOSE_FD, /* fd */ }, }, + .arg = { [0] = { .scnprintf = SCA_CLOSE_FD, /* fd */ }, }, }, { .name = "connect", .errmsg = true, }, { .name = "creat", .errmsg = true, }, { .name = "dup", .errmsg = true, }, { .name = "dup2", .errmsg = true, }, { .name = "dup3", .errmsg = true, }, - { .name = "epoll_ctl", .errmsg = true, STRARRAY(1, op, epoll_ctl_ops), }, + { .name = "epoll_ctl", .errmsg = true, + .arg = { [1] = STRARRAY(op, epoll_ctl_ops), }, }, { .name = "eventfd2", .errmsg = true, - .arg_scnprintf = { [1] = SCA_EFD_FLAGS, /* flags */ }, }, + .arg = { [1] = { .scnprintf = SCA_EFD_FLAGS, /* flags */ }, }, }, { .name = "faccessat", .errmsg = true, }, { .name = "fadvise64", .errmsg = true, }, { .name = "fallocate", .errmsg = true, }, { .name = "fchdir", .errmsg = true, }, { .name = "fchmod", .errmsg = true, }, { .name = "fchmodat", .errmsg = true, - .arg_scnprintf = { [0] = SCA_FDAT, /* fd */ }, }, + .arg = { [0] = { .scnprintf = SCA_FDAT, /* fd */ }, }, }, { .name = "fchown", .errmsg = true, }, { .name = "fchownat", .errmsg = true, - .arg_scnprintf = { [0] = SCA_FDAT, /* fd */ }, }, + .arg = { [0] = { .scnprintf = SCA_FDAT, /* fd */ }, }, }, { .name = "fcntl", .errmsg = true, - .arg_scnprintf = { [1] = SCA_FCNTL_CMD, /* cmd */ - [2] = SCA_FCNTL_ARG, /* arg */ }, - .arg_parm = { [1] = &strarrays__fcntl_cmds_arrays, /* cmd */ }, }, + .arg = { [1] = { .scnprintf = SCA_FCNTL_CMD, /* cmd */ + .parm = &strarrays__fcntl_cmds_arrays, /* cmd */ }, + [2] = { .scnprintf = SCA_FCNTL_ARG, /* arg */ }, }, }, { .name = "fdatasync", .errmsg = true, }, { .name = "flock", .errmsg = true, - .arg_scnprintf = { [1] = SCA_FLOCK, /* cmd */ }, }, + .arg = { [1] = { .scnprintf = SCA_FLOCK, /* cmd */ }, }, }, { .name = "fsetxattr", .errmsg = true, }, { .name = "fstat", .errmsg = true, .alias = "newfstat", }, { .name = "fstatat", .errmsg = true, .alias = "newfstatat", }, @@ -656,186 +663,190 @@ static struct syscall_fmt { { .name = "fsync", .errmsg = true, }, { .name = "ftruncate", .errmsg = true, }, { .name = "futex", .errmsg = true, - .arg_scnprintf = { [1] = SCA_FUTEX_OP, /* op */ }, }, + .arg = { [1] = { .scnprintf = SCA_FUTEX_OP, /* op */ }, }, }, { .name = "futimesat", .errmsg = true, - .arg_scnprintf = { [0] = SCA_FDAT, /* fd */ }, }, + .arg = { [0] = { .scnprintf = SCA_FDAT, /* fd */ }, }, }, { .name = "getdents", .errmsg = true, }, { .name = "getdents64", .errmsg = true, }, - { .name = "getitimer", .errmsg = true, STRARRAY(0, which, itimers), }, + { .name = "getitimer", .errmsg = true, + .arg = { [0] = STRARRAY(which, itimers), }, }, { .name = "getpid", .errpid = true, }, { .name = "getpgid", .errpid = true, }, { .name = "getppid", .errpid = true, }, { .name = "getrandom", .errmsg = true, - .arg_scnprintf = { [2] = SCA_GETRANDOM_FLAGS, /* flags */ }, }, - { .name = "getrlimit", .errmsg = true, STRARRAY(0, resource, rlimit_resources), }, + .arg = { [2] = { .scnprintf = SCA_GETRANDOM_FLAGS, /* flags */ }, }, }, + { .name = "getrlimit", .errmsg = true, + .arg = { [0] = STRARRAY(resource, rlimit_resources), }, }, { .name = "getxattr", .errmsg = true, }, - { .name = "inotify_add_watch", .errmsg = true, }, + { .name = "inotify_add_watch", .errmsg = true, }, { .name = "ioctl", .errmsg = true, - .arg_scnprintf = { + .arg = { #if defined(__i386__) || defined(__x86_64__) /* * FIXME: Make this available to all arches. */ - [1] = SCA_STRHEXARRAY, /* cmd */ - [2] = SCA_HEX, /* arg */ }, - .arg_parm = { [1] = &strarray__tioctls, /* cmd */ }, }, + [1] = { .scnprintf = SCA_STRHEXARRAY, /* cmd */ + .parm = &strarray__tioctls, }, + [2] = { .scnprintf = SCA_HEX, /* arg */ }, }, }, #else - [2] = SCA_HEX, /* arg */ }, }, + [2] = { .scnprintf = SCA_HEX, /* arg */ }, }, }, #endif - { .name = "keyctl", .errmsg = true, STRARRAY(0, option, keyctl_options), }, + { .name = "keyctl", .errmsg = true, + .arg = { [0] = STRARRAY(option, keyctl_options), }, }, { .name = "kill", .errmsg = true, - .arg_scnprintf = { [1] = SCA_SIGNUM, /* sig */ }, }, + .arg = { [1] = { .scnprintf = SCA_SIGNUM, /* sig */ }, }, }, { .name = "lchown", .errmsg = true, }, { .name = "lgetxattr", .errmsg = true, }, { .name = "linkat", .errmsg = true, - .arg_scnprintf = { [0] = SCA_FDAT, /* fd */ }, }, + .arg = { [0] = { .scnprintf = SCA_FDAT, /* fd */ }, }, }, { .name = "listxattr", .errmsg = true, }, { .name = "llistxattr", .errmsg = true, }, { .name = "lremovexattr", .errmsg = true, }, { .name = "lseek", .errmsg = true, - .arg_scnprintf = { [2] = SCA_STRARRAY, /* whence */ }, - .arg_parm = { [2] = &strarray__whences, /* whence */ }, }, + .arg = { [2] = STRARRAY(whence, whences), }, }, { .name = "lsetxattr", .errmsg = true, }, { .name = "lstat", .errmsg = true, .alias = "newlstat", }, { .name = "lsxattr", .errmsg = true, }, { .name = "madvise", .errmsg = true, - .arg_scnprintf = { [0] = SCA_HEX, /* start */ - [2] = SCA_MADV_BHV, /* behavior */ }, }, + .arg = { [0] = { .scnprintf = SCA_HEX, /* start */ }, + [2] = { .scnprintf = SCA_MADV_BHV, /* behavior */ }, }, }, { .name = "mkdir", .errmsg = true, }, { .name = "mkdirat", .errmsg = true, - .arg_scnprintf = { [0] = SCA_FDAT, /* fd */ }, }, + .arg = { [0] = { .scnprintf = SCA_FDAT, /* fd */ }, }, }, { .name = "mknod", .errmsg = true, }, { .name = "mknodat", .errmsg = true, - .arg_scnprintf = { [0] = SCA_FDAT, /* fd */ }, }, + .arg = { [0] = { .scnprintf = SCA_FDAT, /* fd */ }, }, }, { .name = "mlock", .errmsg = true, - .arg_scnprintf = { [0] = SCA_HEX, /* addr */ }, }, + .arg = { [0] = { .scnprintf = SCA_HEX, /* addr */ }, }, }, { .name = "mlockall", .errmsg = true, - .arg_scnprintf = { [0] = SCA_HEX, /* addr */ }, }, + .arg = { [0] = { .scnprintf = SCA_HEX, /* addr */ }, }, }, { .name = "mmap", .hexret = true, /* The standard mmap maps to old_mmap on s390x */ #if defined(__s390x__) .alias = "old_mmap", #endif - .arg_scnprintf = { [0] = SCA_HEX, /* addr */ - [2] = SCA_MMAP_PROT, /* prot */ - [3] = SCA_MMAP_FLAGS, /* flags */ }, }, + .arg = { [0] = { .scnprintf = SCA_HEX, /* addr */ }, + [2] = { .scnprintf = SCA_MMAP_PROT, /* prot */ }, + [3] = { .scnprintf = SCA_MMAP_FLAGS, /* flags */ }, }, }, { .name = "mprotect", .errmsg = true, - .arg_scnprintf = { [0] = SCA_HEX, /* start */ - [2] = SCA_MMAP_PROT, /* prot */ }, }, + .arg = { [0] = { .scnprintf = SCA_HEX, /* start */ }, + [2] = { .scnprintf = SCA_MMAP_PROT, /* prot */ }, }, }, { .name = "mq_unlink", .errmsg = true, - .arg_scnprintf = { [0] = SCA_FILENAME, /* u_name */ }, }, + .arg = { [0] = { .scnprintf = SCA_FILENAME, /* u_name */ }, }, }, { .name = "mremap", .hexret = true, - .arg_scnprintf = { [0] = SCA_HEX, /* addr */ - [3] = SCA_MREMAP_FLAGS, /* flags */ - [4] = SCA_HEX, /* new_addr */ }, }, + .arg = { [0] = { .scnprintf = SCA_HEX, /* addr */ }, + [3] = { .scnprintf = SCA_MREMAP_FLAGS, /* flags */ }, + [4] = { .scnprintf = SCA_HEX, /* new_addr */ }, }, }, { .name = "munlock", .errmsg = true, - .arg_scnprintf = { [0] = SCA_HEX, /* addr */ }, }, + .arg = { [0] = { .scnprintf = SCA_HEX, /* addr */ }, }, }, { .name = "munmap", .errmsg = true, - .arg_scnprintf = { [0] = SCA_HEX, /* addr */ }, }, + .arg = { [0] = { .scnprintf = SCA_HEX, /* addr */ }, }, }, { .name = "name_to_handle_at", .errmsg = true, - .arg_scnprintf = { [0] = SCA_FDAT, /* dfd */ }, }, + .arg = { [0] = { .scnprintf = SCA_FDAT, /* dfd */ }, }, }, { .name = "newfstatat", .errmsg = true, - .arg_scnprintf = { [0] = SCA_FDAT, /* dfd */ }, }, + .arg = { [0] = { .scnprintf = SCA_FDAT, /* dfd */ }, }, }, { .name = "open", .errmsg = true, - .arg_scnprintf = { [1] = SCA_OPEN_FLAGS, /* flags */ }, }, + .arg = { [1] = { .scnprintf = SCA_OPEN_FLAGS, /* flags */ }, }, }, { .name = "open_by_handle_at", .errmsg = true, - .arg_scnprintf = { [0] = SCA_FDAT, /* dfd */ - [2] = SCA_OPEN_FLAGS, /* flags */ }, }, + .arg = { [0] = { .scnprintf = SCA_FDAT, /* dfd */ }, + [2] = { .scnprintf = SCA_OPEN_FLAGS, /* flags */ }, }, }, { .name = "openat", .errmsg = true, - .arg_scnprintf = { [0] = SCA_FDAT, /* dfd */ - [2] = SCA_OPEN_FLAGS, /* flags */ }, }, + .arg = { [0] = { .scnprintf = SCA_FDAT, /* dfd */ }, + [2] = { .scnprintf = SCA_OPEN_FLAGS, /* flags */ }, }, }, { .name = "perf_event_open", .errmsg = true, - .arg_scnprintf = { [2] = SCA_INT, /* cpu */ - [3] = SCA_FD, /* group_fd */ - [4] = SCA_PERF_FLAGS, /* flags */ }, }, + .arg = { [2] = { .scnprintf = SCA_INT, /* cpu */ }, + [3] = { .scnprintf = SCA_FD, /* group_fd */ }, + [4] = { .scnprintf = SCA_PERF_FLAGS, /* flags */ }, }, }, { .name = "pipe2", .errmsg = true, - .arg_scnprintf = { [1] = SCA_PIPE_FLAGS, /* flags */ }, }, + .arg = { [1] = { .scnprintf = SCA_PIPE_FLAGS, /* flags */ }, }, }, { .name = "poll", .errmsg = true, .timeout = true, }, { .name = "ppoll", .errmsg = true, .timeout = true, }, { .name = "pread", .errmsg = true, .alias = "pread64", }, { .name = "preadv", .errmsg = true, .alias = "pread", }, - { .name = "prlimit64", .errmsg = true, STRARRAY(1, resource, rlimit_resources), }, + { .name = "prlimit64", .errmsg = true, + .arg = { [1] = STRARRAY(resource, rlimit_resources), }, }, { .name = "pwrite", .errmsg = true, .alias = "pwrite64", }, { .name = "pwritev", .errmsg = true, }, { .name = "read", .errmsg = true, }, { .name = "readlink", .errmsg = true, }, { .name = "readlinkat", .errmsg = true, - .arg_scnprintf = { [0] = SCA_FDAT, /* dfd */ }, }, + .arg = { [0] = { .scnprintf = SCA_FDAT, /* dfd */ }, }, }, { .name = "readv", .errmsg = true, }, { .name = "recvfrom", .errmsg = true, - .arg_scnprintf = { [3] = SCA_MSG_FLAGS, /* flags */ }, }, + .arg = { [3] = { .scnprintf = SCA_MSG_FLAGS, /* flags */ }, }, }, { .name = "recvmmsg", .errmsg = true, - .arg_scnprintf = { [3] = SCA_MSG_FLAGS, /* flags */ }, }, + .arg = { [3] = { .scnprintf = SCA_MSG_FLAGS, /* flags */ }, }, }, { .name = "recvmsg", .errmsg = true, - .arg_scnprintf = { [2] = SCA_MSG_FLAGS, /* flags */ }, }, + .arg = { [2] = { .scnprintf = SCA_MSG_FLAGS, /* flags */ }, }, }, { .name = "removexattr", .errmsg = true, }, { .name = "renameat", .errmsg = true, - .arg_scnprintf = { [0] = SCA_FDAT, /* dfd */ }, }, + .arg = { [0] = { .scnprintf = SCA_FDAT, /* dfd */ }, }, }, { .name = "rmdir", .errmsg = true, }, { .name = "rt_sigaction", .errmsg = true, - .arg_scnprintf = { [0] = SCA_SIGNUM, /* sig */ }, }, - { .name = "rt_sigprocmask", .errmsg = true, STRARRAY(0, how, sighow), }, + .arg = { [0] = { .scnprintf = SCA_SIGNUM, /* sig */ }, }, }, + { .name = "rt_sigprocmask", .errmsg = true, + .arg = { [0] = STRARRAY(how, sighow), }, }, { .name = "rt_sigqueueinfo", .errmsg = true, - .arg_scnprintf = { [1] = SCA_SIGNUM, /* sig */ }, }, + .arg = { [1] = { .scnprintf = SCA_SIGNUM, /* sig */ }, }, }, { .name = "rt_tgsigqueueinfo", .errmsg = true, - .arg_scnprintf = { [2] = SCA_SIGNUM, /* sig */ }, }, + .arg = { [2] = { .scnprintf = SCA_SIGNUM, /* sig */ }, }, }, { .name = "sched_getattr", .errmsg = true, }, { .name = "sched_setattr", .errmsg = true, }, { .name = "sched_setscheduler", .errmsg = true, - .arg_scnprintf = { [1] = SCA_SCHED_POLICY, /* policy */ }, }, + .arg = { [1] = { .scnprintf = SCA_SCHED_POLICY, /* policy */ }, }, }, { .name = "seccomp", .errmsg = true, - .arg_scnprintf = { [0] = SCA_SECCOMP_OP, /* op */ - [1] = SCA_SECCOMP_FLAGS, /* flags */ }, }, + .arg = { [0] = { .scnprintf = SCA_SECCOMP_OP, /* op */ }, + [1] = { .scnprintf = SCA_SECCOMP_FLAGS, /* flags */ }, }, }, { .name = "select", .errmsg = true, .timeout = true, }, { .name = "sendmmsg", .errmsg = true, - .arg_scnprintf = { [3] = SCA_MSG_FLAGS, /* flags */ }, }, + .arg = { [3] = { .scnprintf = SCA_MSG_FLAGS, /* flags */ }, }, }, { .name = "sendmsg", .errmsg = true, - .arg_scnprintf = { [2] = SCA_MSG_FLAGS, /* flags */ }, }, + .arg = { [2] = { .scnprintf = SCA_MSG_FLAGS, /* flags */ }, }, }, { .name = "sendto", .errmsg = true, - .arg_scnprintf = { [3] = SCA_MSG_FLAGS, /* flags */ }, }, + .arg = { [3] = { .scnprintf = SCA_MSG_FLAGS, /* flags */ }, }, }, { .name = "set_tid_address", .errpid = true, }, - { .name = "setitimer", .errmsg = true, STRARRAY(0, which, itimers), }, + { .name = "setitimer", .errmsg = true, + .arg = { [0] = STRARRAY(which, itimers), }, }, { .name = "setpgid", .errmsg = true, }, - { .name = "setrlimit", .errmsg = true, STRARRAY(0, resource, rlimit_resources), }, + { .name = "setrlimit", .errmsg = true, + .arg = { [0] = STRARRAY(resource, rlimit_resources), }, }, { .name = "setxattr", .errmsg = true, }, { .name = "shutdown", .errmsg = true, }, { .name = "socket", .errmsg = true, - .arg_scnprintf = { [0] = SCA_STRARRAY, /* family */ - [1] = SCA_SK_TYPE, /* type */ }, - .arg_parm = { [0] = &strarray__socket_families, /* family */ }, }, + .arg = { [0] = STRARRAY(family, socket_families), + [1] = { .scnprintf = SCA_SK_TYPE, /* type */ }, }, }, { .name = "socketpair", .errmsg = true, - .arg_scnprintf = { [0] = SCA_STRARRAY, /* family */ - [1] = SCA_SK_TYPE, /* type */ }, - .arg_parm = { [0] = &strarray__socket_families, /* family */ }, }, + .arg = { [0] = STRARRAY(family, socket_families), + [1] = { .scnprintf = SCA_SK_TYPE, /* type */ }, }, }, { .name = "stat", .errmsg = true, .alias = "newstat", }, { .name = "statfs", .errmsg = true, }, { .name = "statx", .errmsg = true, - .arg_scnprintf = { [0] = SCA_FDAT, /* flags */ - [2] = SCA_STATX_FLAGS, /* flags */ - [3] = SCA_STATX_MASK, /* mask */ }, }, + .arg = { [0] = { .scnprintf = SCA_FDAT, /* fdat */ }, + [2] = { .scnprintf = SCA_STATX_FLAGS, /* flags */ } , + [3] = { .scnprintf = SCA_STATX_MASK, /* mask */ }, }, }, { .name = "swapoff", .errmsg = true, - .arg_scnprintf = { [0] = SCA_FILENAME, /* specialfile */ }, }, + .arg = { [0] = { .scnprintf = SCA_FILENAME, /* specialfile */ }, }, }, { .name = "swapon", .errmsg = true, - .arg_scnprintf = { [0] = SCA_FILENAME, /* specialfile */ }, }, + .arg = { [0] = { .scnprintf = SCA_FILENAME, /* specialfile */ }, }, }, { .name = "symlinkat", .errmsg = true, - .arg_scnprintf = { [0] = SCA_FDAT, /* dfd */ }, }, + .arg = { [0] = { .scnprintf = SCA_FDAT, /* dfd */ }, }, }, { .name = "tgkill", .errmsg = true, - .arg_scnprintf = { [2] = SCA_SIGNUM, /* sig */ }, }, + .arg = { [2] = { .scnprintf = SCA_SIGNUM, /* sig */ }, }, }, { .name = "tkill", .errmsg = true, - .arg_scnprintf = { [1] = SCA_SIGNUM, /* sig */ }, }, + .arg = { [1] = { .scnprintf = SCA_SIGNUM, /* sig */ }, }, }, { .name = "truncate", .errmsg = true, }, { .name = "uname", .errmsg = true, .alias = "newuname", }, { .name = "unlinkat", .errmsg = true, - .arg_scnprintf = { [0] = SCA_FDAT, /* dfd */ }, }, + .arg = { [0] = { .scnprintf = SCA_FDAT, /* dfd */ }, }, }, { .name = "utime", .errmsg = true, }, { .name = "utimensat", .errmsg = true, - .arg_scnprintf = { [0] = SCA_FDAT, /* dirfd */ }, }, + .arg = { [0] = { .scnprintf = SCA_FDAT, /* dirfd */ }, }, }, { .name = "utimes", .errmsg = true, }, { .name = "vmsplice", .errmsg = true, }, { .name = "wait4", .errpid = true, - .arg_scnprintf = { [2] = SCA_WAITID_OPTIONS, /* options */ }, }, + .arg = { [2] = { .scnprintf = SCA_WAITID_OPTIONS, /* options */ }, }, }, { .name = "waitid", .errpid = true, - .arg_scnprintf = { [3] = SCA_WAITID_OPTIONS, /* options */ }, }, + .arg = { [3] = { .scnprintf = SCA_WAITID_OPTIONS, /* options */ }, }, }, { .name = "write", .errmsg = true, }, { .name = "writev", .errmsg = true, }, }; @@ -859,8 +870,7 @@ struct syscall { const char *name; bool is_exit; struct syscall_fmt *fmt; - size_t (**arg_scnprintf)(char *bf, size_t size, struct syscall_arg *arg); - void **arg_parm; + struct syscall_arg_fmt *arg_fmt; }; /* @@ -1209,27 +1219,29 @@ static int syscall__set_arg_fmts(struct syscall *sc) struct format_field *field; int idx = 0, len; - sc->arg_scnprintf = calloc(sc->nr_args, sizeof(void *)); - if (sc->arg_scnprintf == NULL) + sc->arg_fmt = calloc(sc->nr_args, sizeof(*sc->arg_fmt)); + if (sc->arg_fmt == NULL) return -1; - if (sc->fmt) - sc->arg_parm = sc->fmt->arg_parm; + for (field = sc->args; field; field = field->next, ++idx) { + if (sc->fmt) { + sc->arg_fmt[idx] = sc->fmt->arg[idx]; + + if (sc->fmt->arg[idx].scnprintf) + continue; + } - for (field = sc->args; field; field = field->next) { - if (sc->fmt && sc->fmt->arg_scnprintf[idx]) - sc->arg_scnprintf[idx] = sc->fmt->arg_scnprintf[idx]; - else if (strcmp(field->type, "const char *") == 0 && + if (strcmp(field->type, "const char *") == 0 && (strcmp(field->name, "filename") == 0 || strcmp(field->name, "path") == 0 || strcmp(field->name, "pathname") == 0)) - sc->arg_scnprintf[idx] = SCA_FILENAME; + sc->arg_fmt[idx].scnprintf = SCA_FILENAME; else if (field->flags & FIELD_IS_POINTER) - sc->arg_scnprintf[idx] = syscall_arg__scnprintf_hex; + sc->arg_fmt[idx].scnprintf = syscall_arg__scnprintf_hex; else if (strcmp(field->type, "pid_t") == 0) - sc->arg_scnprintf[idx] = SCA_PID; + sc->arg_fmt[idx].scnprintf = SCA_PID; else if (strcmp(field->type, "umode_t") == 0) - sc->arg_scnprintf[idx] = SCA_MODE_T; + sc->arg_fmt[idx].scnprintf = SCA_MODE_T; else if ((strcmp(field->type, "int") == 0 || strcmp(field->type, "unsigned int") == 0 || strcmp(field->type, "long") == 0) && @@ -1242,9 +1254,8 @@ static int syscall__set_arg_fmts(struct syscall *sc) * 23 unsigned int * 7 unsigned long */ - sc->arg_scnprintf[idx] = SCA_FD; + sc->arg_fmt[idx].scnprintf = SCA_FD; } - ++idx; } return 0; @@ -1416,20 +1427,19 @@ static size_t syscall__scnprintf_args(struct syscall *sc, char *bf, size_t size, * strarray for it. */ if (val == 0 && - !(sc->arg_scnprintf && - (sc->arg_scnprintf[arg.idx] == SCA_STRARRAY || - sc->arg_scnprintf[arg.idx] == SCA_STRARRAYS) && - sc->arg_parm[arg.idx])) + !(sc->arg_fmt && + (sc->arg_fmt[arg.idx].scnprintf == SCA_STRARRAY || + sc->arg_fmt[arg.idx].scnprintf == SCA_STRARRAYS) && + sc->arg_fmt[arg.idx].parm)) continue; printed += scnprintf(bf + printed, size - printed, "%s%s: ", printed ? ", " : "", field->name); - if (sc->arg_scnprintf && sc->arg_scnprintf[arg.idx]) { + if (sc->arg_fmt && sc->arg_fmt[arg.idx].scnprintf) { arg.val = val; - if (sc->arg_parm) - arg.parm = sc->arg_parm[arg.idx]; - printed += sc->arg_scnprintf[arg.idx](bf + printed, - size - printed, &arg); + if (sc->arg_fmt[arg.idx].parm) + arg.parm = sc->arg_fmt[arg.idx].parm; + printed += sc->arg_fmt[arg.idx].scnprintf(bf + printed, size - printed, &arg); } else { printed += scnprintf(bf + printed, size - printed, "%ld", val); -- cgit v1.1 From d47737d524174a81b80c487fe07de3ee2458ee32 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Mon, 17 Jul 2017 15:59:03 -0300 Subject: perf trace: Allow syscall arg formatters to request non suppression of zeros The 'perf trace' tool is suppressing args set to zero, with the exception of string tables (strarrays), which are kinda like enums, i.e. we have maps to go from numbers to strings. But the 'cmd' fcntl arg requires more specialized treatment, as its value will regulate if the next fcntl syscall arg, 'arg', should be ignored (not used) and also how to format the syscall return (fd, file flags, etc), so add a 'show_zero" bool to struct syscall_arg_fmt, to regulate this more explicitely. Will be used in a following patch with fcntl, here is just the mechanism. Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-all738jctxets8ffyizp5lzo@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-trace.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'tools') diff --git a/tools/perf/builtin-trace.c b/tools/perf/builtin-trace.c index 32778a6..b842bd9 100644 --- a/tools/perf/builtin-trace.c +++ b/tools/perf/builtin-trace.c @@ -604,6 +604,7 @@ static size_t syscall_arg__scnprintf_getrandom_flags(char *bf, size_t size, struct syscall_arg_fmt { size_t (*scnprintf)(char *bf, size_t size, struct syscall_arg *arg); void *parm; + bool show_zero; }; static struct syscall_fmt { @@ -1428,7 +1429,8 @@ static size_t syscall__scnprintf_args(struct syscall *sc, char *bf, size_t size, */ if (val == 0 && !(sc->arg_fmt && - (sc->arg_fmt[arg.idx].scnprintf == SCA_STRARRAY || + (sc->arg_fmt[arg.idx].show_zero || + sc->arg_fmt[arg.idx].scnprintf == SCA_STRARRAY || sc->arg_fmt[arg.idx].scnprintf == SCA_STRARRAYS) && sc->arg_fmt[arg.idx].parm)) continue; -- cgit v1.1 From 39cc355b0486d89cef335e35edde26c14b462126 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Mon, 17 Jul 2017 16:02:52 -0300 Subject: perf trace beauty fcntl: Do not suppress 'cmd' when zero, should be DUPFD Before: 77059.513 ( 0.005 ms): bash/6649 fcntl(fd: 1, arg: 10) = 10 After: 77059.513 ( 0.005 ms): bash/6649 fcntl(fd: 1, cmd: DUPFD, arg: 10) = 10 Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-woois88uwcr4xu38xx1ihiwo@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-trace.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'tools') diff --git a/tools/perf/builtin-trace.c b/tools/perf/builtin-trace.c index b842bd9..bab87f6 100644 --- a/tools/perf/builtin-trace.c +++ b/tools/perf/builtin-trace.c @@ -652,7 +652,8 @@ static struct syscall_fmt { .arg = { [0] = { .scnprintf = SCA_FDAT, /* fd */ }, }, }, { .name = "fcntl", .errmsg = true, .arg = { [1] = { .scnprintf = SCA_FCNTL_CMD, /* cmd */ - .parm = &strarrays__fcntl_cmds_arrays, /* cmd */ }, + .parm = &strarrays__fcntl_cmds_arrays, + .show_zero = true, }, [2] = { .scnprintf = SCA_FCNTL_ARG, /* arg */ }, }, }, { .name = "fdatasync", .errmsg = true, }, { .name = "flock", .errmsg = true, -- cgit v1.1 From befecc810c95994774439baeedc8a03fe8d51e8f Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Mon, 17 Jul 2017 16:04:20 -0300 Subject: perf trace beauty fcntl: Beautify the 'arg' for DUPFD Before: 77059.513 ( 0.005 ms): bash/6649 fcntl(fd: 1, cmd: DUPFD, arg: 10) = 10 After: 77059.513 ( 0.005 ms): bash/6649 fcntl(fd: 1, cmd: DUPFD, arg: 10) = 10 Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-0k8iszng0slcuw0rc6xq1x5l@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/trace/beauty/fcntl.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'tools') diff --git a/tools/perf/trace/beauty/fcntl.c b/tools/perf/trace/beauty/fcntl.c index b6987c4..9e8900c 100644 --- a/tools/perf/trace/beauty/fcntl.c +++ b/tools/perf/trace/beauty/fcntl.c @@ -71,6 +71,9 @@ size_t syscall_arg__scnprintf_fcntl_arg(char *bf, size_t size, struct syscall_ar { int cmd = syscall_arg__val(arg, 1); + if (cmd == F_DUPFD) + return syscall_arg__scnprintf_fd(bf, size, arg); + if (cmd == F_SETFD) return fcntl__scnprintf_getfd(arg->val, bf, size); -- cgit v1.1 From 1f63139c3f8af1d6a09de5dd355c8b5695407c79 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Tue, 18 Jul 2017 12:45:57 -0300 Subject: perf trace beauty: Simplify syscall return formatting Removing syscall_fmt::err_msg and instead always formatting negative returns as errno values. With this we can remove a lot of entries that have no special handling besides the ones we can do by looking at the tracefs format files, i.e. the types for the fields (e.g. pid_t), well known names (e.g. fd). Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-rg9u7a3qqdnzo37d212vnz2o@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-trace.c | 209 ++++++++++++++++++--------------------------- 1 file changed, 81 insertions(+), 128 deletions(-) (limited to 'tools') diff --git a/tools/perf/builtin-trace.c b/tools/perf/builtin-trace.c index bab87f6..1e4c065 100644 --- a/tools/perf/builtin-trace.c +++ b/tools/perf/builtin-trace.c @@ -611,77 +611,53 @@ static struct syscall_fmt { const char *name; const char *alias; struct syscall_arg_fmt arg[6]; - bool errmsg; bool errpid; bool timeout; bool hexret; } syscall_fmts[] = { - { .name = "access", .errmsg = true, + { .name = "access", .arg = { [1] = { .scnprintf = SCA_ACCMODE, /* mode */ }, }, }, - { .name = "arch_prctl", .errmsg = true, .alias = "prctl", }, - { .name = "bpf", .errmsg = true, + { .name = "arch_prctl", .alias = "prctl", }, + { .name = "bpf", .arg = { [0] = STRARRAY(cmd, bpf_cmd), }, }, { .name = "brk", .hexret = true, .arg = { [0] = { .scnprintf = SCA_HEX, /* brk */ }, }, }, - { .name = "chdir", .errmsg = true, }, - { .name = "chmod", .errmsg = true, }, - { .name = "chroot", .errmsg = true, }, - { .name = "clock_gettime", .errmsg = true, + { .name = "clock_gettime", .arg = { [0] = STRARRAY(clk_id, clockid), }, }, { .name = "clone", .errpid = true, }, - { .name = "close", .errmsg = true, + { .name = "close", .arg = { [0] = { .scnprintf = SCA_CLOSE_FD, /* fd */ }, }, }, - { .name = "connect", .errmsg = true, }, - { .name = "creat", .errmsg = true, }, - { .name = "dup", .errmsg = true, }, - { .name = "dup2", .errmsg = true, }, - { .name = "dup3", .errmsg = true, }, - { .name = "epoll_ctl", .errmsg = true, + { .name = "epoll_ctl", .arg = { [1] = STRARRAY(op, epoll_ctl_ops), }, }, - { .name = "eventfd2", .errmsg = true, + { .name = "eventfd2", .arg = { [1] = { .scnprintf = SCA_EFD_FLAGS, /* flags */ }, }, }, - { .name = "faccessat", .errmsg = true, }, - { .name = "fadvise64", .errmsg = true, }, - { .name = "fallocate", .errmsg = true, }, - { .name = "fchdir", .errmsg = true, }, - { .name = "fchmod", .errmsg = true, }, - { .name = "fchmodat", .errmsg = true, + { .name = "fchmodat", .arg = { [0] = { .scnprintf = SCA_FDAT, /* fd */ }, }, }, - { .name = "fchown", .errmsg = true, }, - { .name = "fchownat", .errmsg = true, + { .name = "fchownat", .arg = { [0] = { .scnprintf = SCA_FDAT, /* fd */ }, }, }, - { .name = "fcntl", .errmsg = true, + { .name = "fcntl", .arg = { [1] = { .scnprintf = SCA_FCNTL_CMD, /* cmd */ .parm = &strarrays__fcntl_cmds_arrays, .show_zero = true, }, [2] = { .scnprintf = SCA_FCNTL_ARG, /* arg */ }, }, }, - { .name = "fdatasync", .errmsg = true, }, - { .name = "flock", .errmsg = true, + { .name = "flock", .arg = { [1] = { .scnprintf = SCA_FLOCK, /* cmd */ }, }, }, - { .name = "fsetxattr", .errmsg = true, }, - { .name = "fstat", .errmsg = true, .alias = "newfstat", }, - { .name = "fstatat", .errmsg = true, .alias = "newfstatat", }, - { .name = "fstatfs", .errmsg = true, }, - { .name = "fsync", .errmsg = true, }, - { .name = "ftruncate", .errmsg = true, }, - { .name = "futex", .errmsg = true, + { .name = "fstat", .alias = "newfstat", }, + { .name = "fstatat", .alias = "newfstatat", }, + { .name = "futex", .arg = { [1] = { .scnprintf = SCA_FUTEX_OP, /* op */ }, }, }, - { .name = "futimesat", .errmsg = true, + { .name = "futimesat", .arg = { [0] = { .scnprintf = SCA_FDAT, /* fd */ }, }, }, - { .name = "getdents", .errmsg = true, }, - { .name = "getdents64", .errmsg = true, }, - { .name = "getitimer", .errmsg = true, + { .name = "getitimer", .arg = { [0] = STRARRAY(which, itimers), }, }, { .name = "getpid", .errpid = true, }, { .name = "getpgid", .errpid = true, }, { .name = "getppid", .errpid = true, }, - { .name = "getrandom", .errmsg = true, + { .name = "getrandom", .arg = { [2] = { .scnprintf = SCA_GETRANDOM_FLAGS, /* flags */ }, }, }, - { .name = "getrlimit", .errmsg = true, + { .name = "getrlimit", .arg = { [0] = STRARRAY(resource, rlimit_resources), }, }, - { .name = "getxattr", .errmsg = true, }, - { .name = "inotify_add_watch", .errmsg = true, }, - { .name = "ioctl", .errmsg = true, + { .name = "ioctl", .arg = { #if defined(__i386__) || defined(__x86_64__) /* @@ -693,34 +669,25 @@ static struct syscall_fmt { #else [2] = { .scnprintf = SCA_HEX, /* arg */ }, }, }, #endif - { .name = "keyctl", .errmsg = true, + { .name = "keyctl", .arg = { [0] = STRARRAY(option, keyctl_options), }, }, - { .name = "kill", .errmsg = true, + { .name = "kill", .arg = { [1] = { .scnprintf = SCA_SIGNUM, /* sig */ }, }, }, - { .name = "lchown", .errmsg = true, }, - { .name = "lgetxattr", .errmsg = true, }, - { .name = "linkat", .errmsg = true, + { .name = "linkat", .arg = { [0] = { .scnprintf = SCA_FDAT, /* fd */ }, }, }, - { .name = "listxattr", .errmsg = true, }, - { .name = "llistxattr", .errmsg = true, }, - { .name = "lremovexattr", .errmsg = true, }, - { .name = "lseek", .errmsg = true, + { .name = "lseek", .arg = { [2] = STRARRAY(whence, whences), }, }, - { .name = "lsetxattr", .errmsg = true, }, - { .name = "lstat", .errmsg = true, .alias = "newlstat", }, - { .name = "lsxattr", .errmsg = true, }, - { .name = "madvise", .errmsg = true, + { .name = "lstat", .alias = "newlstat", }, + { .name = "madvise", .arg = { [0] = { .scnprintf = SCA_HEX, /* start */ }, [2] = { .scnprintf = SCA_MADV_BHV, /* behavior */ }, }, }, - { .name = "mkdir", .errmsg = true, }, - { .name = "mkdirat", .errmsg = true, + { .name = "mkdirat", .arg = { [0] = { .scnprintf = SCA_FDAT, /* fd */ }, }, }, - { .name = "mknod", .errmsg = true, }, - { .name = "mknodat", .errmsg = true, + { .name = "mknodat", .arg = { [0] = { .scnprintf = SCA_FDAT, /* fd */ }, }, }, - { .name = "mlock", .errmsg = true, + { .name = "mlock", .arg = { [0] = { .scnprintf = SCA_HEX, /* addr */ }, }, }, - { .name = "mlockall", .errmsg = true, + { .name = "mlockall", .arg = { [0] = { .scnprintf = SCA_HEX, /* addr */ }, }, }, { .name = "mmap", .hexret = true, /* The standard mmap maps to old_mmap on s390x */ @@ -730,127 +697,109 @@ static struct syscall_fmt { .arg = { [0] = { .scnprintf = SCA_HEX, /* addr */ }, [2] = { .scnprintf = SCA_MMAP_PROT, /* prot */ }, [3] = { .scnprintf = SCA_MMAP_FLAGS, /* flags */ }, }, }, - { .name = "mprotect", .errmsg = true, + { .name = "mprotect", .arg = { [0] = { .scnprintf = SCA_HEX, /* start */ }, [2] = { .scnprintf = SCA_MMAP_PROT, /* prot */ }, }, }, - { .name = "mq_unlink", .errmsg = true, + { .name = "mq_unlink", .arg = { [0] = { .scnprintf = SCA_FILENAME, /* u_name */ }, }, }, { .name = "mremap", .hexret = true, .arg = { [0] = { .scnprintf = SCA_HEX, /* addr */ }, [3] = { .scnprintf = SCA_MREMAP_FLAGS, /* flags */ }, [4] = { .scnprintf = SCA_HEX, /* new_addr */ }, }, }, - { .name = "munlock", .errmsg = true, + { .name = "munlock", .arg = { [0] = { .scnprintf = SCA_HEX, /* addr */ }, }, }, - { .name = "munmap", .errmsg = true, + { .name = "munmap", .arg = { [0] = { .scnprintf = SCA_HEX, /* addr */ }, }, }, - { .name = "name_to_handle_at", .errmsg = true, + { .name = "name_to_handle_at", .arg = { [0] = { .scnprintf = SCA_FDAT, /* dfd */ }, }, }, - { .name = "newfstatat", .errmsg = true, + { .name = "newfstatat", .arg = { [0] = { .scnprintf = SCA_FDAT, /* dfd */ }, }, }, - { .name = "open", .errmsg = true, + { .name = "open", .arg = { [1] = { .scnprintf = SCA_OPEN_FLAGS, /* flags */ }, }, }, - { .name = "open_by_handle_at", .errmsg = true, + { .name = "open_by_handle_at", .arg = { [0] = { .scnprintf = SCA_FDAT, /* dfd */ }, [2] = { .scnprintf = SCA_OPEN_FLAGS, /* flags */ }, }, }, - { .name = "openat", .errmsg = true, + { .name = "openat", .arg = { [0] = { .scnprintf = SCA_FDAT, /* dfd */ }, [2] = { .scnprintf = SCA_OPEN_FLAGS, /* flags */ }, }, }, - { .name = "perf_event_open", .errmsg = true, + { .name = "perf_event_open", .arg = { [2] = { .scnprintf = SCA_INT, /* cpu */ }, [3] = { .scnprintf = SCA_FD, /* group_fd */ }, [4] = { .scnprintf = SCA_PERF_FLAGS, /* flags */ }, }, }, - { .name = "pipe2", .errmsg = true, + { .name = "pipe2", .arg = { [1] = { .scnprintf = SCA_PIPE_FLAGS, /* flags */ }, }, }, - { .name = "poll", .errmsg = true, .timeout = true, }, - { .name = "ppoll", .errmsg = true, .timeout = true, }, - { .name = "pread", .errmsg = true, .alias = "pread64", }, - { .name = "preadv", .errmsg = true, .alias = "pread", }, - { .name = "prlimit64", .errmsg = true, + { .name = "poll", .timeout = true, }, + { .name = "ppoll", .timeout = true, }, + { .name = "pread", .alias = "pread64", }, + { .name = "preadv", .alias = "pread", }, + { .name = "prlimit64", .arg = { [1] = STRARRAY(resource, rlimit_resources), }, }, - { .name = "pwrite", .errmsg = true, .alias = "pwrite64", }, - { .name = "pwritev", .errmsg = true, }, - { .name = "read", .errmsg = true, }, - { .name = "readlink", .errmsg = true, }, - { .name = "readlinkat", .errmsg = true, + { .name = "pwrite", .alias = "pwrite64", }, + { .name = "readlinkat", .arg = { [0] = { .scnprintf = SCA_FDAT, /* dfd */ }, }, }, - { .name = "readv", .errmsg = true, }, - { .name = "recvfrom", .errmsg = true, + { .name = "recvfrom", .arg = { [3] = { .scnprintf = SCA_MSG_FLAGS, /* flags */ }, }, }, - { .name = "recvmmsg", .errmsg = true, + { .name = "recvmmsg", .arg = { [3] = { .scnprintf = SCA_MSG_FLAGS, /* flags */ }, }, }, - { .name = "recvmsg", .errmsg = true, + { .name = "recvmsg", .arg = { [2] = { .scnprintf = SCA_MSG_FLAGS, /* flags */ }, }, }, - { .name = "removexattr", .errmsg = true, }, - { .name = "renameat", .errmsg = true, + { .name = "renameat", .arg = { [0] = { .scnprintf = SCA_FDAT, /* dfd */ }, }, }, - { .name = "rmdir", .errmsg = true, }, - { .name = "rt_sigaction", .errmsg = true, + { .name = "rt_sigaction", .arg = { [0] = { .scnprintf = SCA_SIGNUM, /* sig */ }, }, }, - { .name = "rt_sigprocmask", .errmsg = true, + { .name = "rt_sigprocmask", .arg = { [0] = STRARRAY(how, sighow), }, }, - { .name = "rt_sigqueueinfo", .errmsg = true, + { .name = "rt_sigqueueinfo", .arg = { [1] = { .scnprintf = SCA_SIGNUM, /* sig */ }, }, }, - { .name = "rt_tgsigqueueinfo", .errmsg = true, + { .name = "rt_tgsigqueueinfo", .arg = { [2] = { .scnprintf = SCA_SIGNUM, /* sig */ }, }, }, - { .name = "sched_getattr", .errmsg = true, }, - { .name = "sched_setattr", .errmsg = true, }, - { .name = "sched_setscheduler", .errmsg = true, + { .name = "sched_setscheduler", .arg = { [1] = { .scnprintf = SCA_SCHED_POLICY, /* policy */ }, }, }, - { .name = "seccomp", .errmsg = true, + { .name = "seccomp", .arg = { [0] = { .scnprintf = SCA_SECCOMP_OP, /* op */ }, [1] = { .scnprintf = SCA_SECCOMP_FLAGS, /* flags */ }, }, }, - { .name = "select", .errmsg = true, .timeout = true, }, - { .name = "sendmmsg", .errmsg = true, + { .name = "select", .timeout = true, }, + { .name = "sendmmsg", .arg = { [3] = { .scnprintf = SCA_MSG_FLAGS, /* flags */ }, }, }, - { .name = "sendmsg", .errmsg = true, + { .name = "sendmsg", .arg = { [2] = { .scnprintf = SCA_MSG_FLAGS, /* flags */ }, }, }, - { .name = "sendto", .errmsg = true, + { .name = "sendto", .arg = { [3] = { .scnprintf = SCA_MSG_FLAGS, /* flags */ }, }, }, { .name = "set_tid_address", .errpid = true, }, - { .name = "setitimer", .errmsg = true, + { .name = "setitimer", .arg = { [0] = STRARRAY(which, itimers), }, }, - { .name = "setpgid", .errmsg = true, }, - { .name = "setrlimit", .errmsg = true, + { .name = "setrlimit", .arg = { [0] = STRARRAY(resource, rlimit_resources), }, }, - { .name = "setxattr", .errmsg = true, }, - { .name = "shutdown", .errmsg = true, }, - { .name = "socket", .errmsg = true, + { .name = "socket", .arg = { [0] = STRARRAY(family, socket_families), [1] = { .scnprintf = SCA_SK_TYPE, /* type */ }, }, }, - { .name = "socketpair", .errmsg = true, + { .name = "socketpair", .arg = { [0] = STRARRAY(family, socket_families), [1] = { .scnprintf = SCA_SK_TYPE, /* type */ }, }, }, - { .name = "stat", .errmsg = true, .alias = "newstat", }, - { .name = "statfs", .errmsg = true, }, - { .name = "statx", .errmsg = true, + { .name = "stat", .alias = "newstat", }, + { .name = "statx", .arg = { [0] = { .scnprintf = SCA_FDAT, /* fdat */ }, [2] = { .scnprintf = SCA_STATX_FLAGS, /* flags */ } , [3] = { .scnprintf = SCA_STATX_MASK, /* mask */ }, }, }, - { .name = "swapoff", .errmsg = true, + { .name = "swapoff", .arg = { [0] = { .scnprintf = SCA_FILENAME, /* specialfile */ }, }, }, - { .name = "swapon", .errmsg = true, + { .name = "swapon", .arg = { [0] = { .scnprintf = SCA_FILENAME, /* specialfile */ }, }, }, - { .name = "symlinkat", .errmsg = true, + { .name = "symlinkat", .arg = { [0] = { .scnprintf = SCA_FDAT, /* dfd */ }, }, }, - { .name = "tgkill", .errmsg = true, + { .name = "tgkill", .arg = { [2] = { .scnprintf = SCA_SIGNUM, /* sig */ }, }, }, - { .name = "tkill", .errmsg = true, + { .name = "tkill", .arg = { [1] = { .scnprintf = SCA_SIGNUM, /* sig */ }, }, }, - { .name = "truncate", .errmsg = true, }, - { .name = "uname", .errmsg = true, .alias = "newuname", }, - { .name = "unlinkat", .errmsg = true, + { .name = "uname", .alias = "newuname", }, + { .name = "unlinkat", .arg = { [0] = { .scnprintf = SCA_FDAT, /* dfd */ }, }, }, - { .name = "utime", .errmsg = true, }, - { .name = "utimensat", .errmsg = true, + { .name = "utimensat", .arg = { [0] = { .scnprintf = SCA_FDAT, /* dirfd */ }, }, }, - { .name = "utimes", .errmsg = true, }, - { .name = "vmsplice", .errmsg = true, }, { .name = "wait4", .errpid = true, .arg = { [2] = { .scnprintf = SCA_WAITID_OPTIONS, /* options */ }, }, }, { .name = "waitid", .errpid = true, .arg = { [3] = { .scnprintf = SCA_WAITID_OPTIONS, /* options */ }, }, }, - { .name = "write", .errmsg = true, }, - { .name = "writev", .errmsg = true, }, }; static int syscall_fmt__cmp(const void *name, const void *fmtp) @@ -1708,14 +1657,18 @@ static int trace__sys_exit(struct trace *trace, struct perf_evsel *evsel, } if (sc->fmt == NULL) { + if (ret < 0) + goto errno_print; signed_print: - fprintf(trace->output, ") = %ld", ret); - } else if (ret < 0 && (sc->fmt->errmsg || sc->fmt->errpid)) { + fprintf(trace->output, ") %ld", ret); + } else if (ret < 0) { +errno_print: { char bf[STRERR_BUFSIZE]; const char *emsg = str_error_r(-ret, bf, sizeof(bf)), *e = audit_errno_to_name(-ret); fprintf(trace->output, ") = -1 %s %s", e, emsg); + } } else if (ret == 0 && sc->fmt->timeout) fprintf(trace->output, ") = 0 Timeout"); else if (ttrace->ret_scnprintf) { -- cgit v1.1 From 8b8ef2d74dec305b99dd43ec71d3ba2f502100b9 Mon Sep 17 00:00:00 2001 From: Jin Yao Date: Wed, 19 Jul 2017 04:31:32 +0800 Subject: perf report: Enable finding kernel inline functions Currently perf supports a mode to query inline stack. It works well for finding user space inline functions but it doesn't work for kernel ones, due to some unnecessary check. This patch removes these unnecessary checks. Now kernel inline functions can be reported. For example: perf report --inline -g func --stdio |--46.19%--do_huge_pmd_anonymous_page | do_huge_pmd_anonymous_page (inline) | __do_huge_pmd_anonymous_page (inline) | __SetPageUptodate (inline) | __set_bit (inline) The result is compared with the output of addr2line. They match. Signed-off-by: Yao Jin Cc: Alexander Shishkin Cc: Andi Kleen Cc: Jiri Olsa Cc: Kan Liang Cc: Milian Wolff Cc: Peter Zijlstra Link: http://lkml.kernel.org/r/1500409892-15904-1-git-send-email-yao.jin@linux.intel.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/ui/browsers/hists.c | 3 --- tools/perf/ui/stdio/hist.c | 3 --- 2 files changed, 6 deletions(-) (limited to 'tools') diff --git a/tools/perf/ui/browsers/hists.c b/tools/perf/ui/browsers/hists.c index 69f4570..f4bc246 100644 --- a/tools/perf/ui/browsers/hists.c +++ b/tools/perf/ui/browsers/hists.c @@ -166,9 +166,6 @@ static struct inline_node *inline_node__create(struct map *map, u64 ip) if (dso == NULL) return NULL; - if (dso->kernel != DSO_TYPE_USER) - return NULL; - node = dso__parse_addr_inlines(dso, map__rip_2objdump(map, ip)); diff --git a/tools/perf/ui/stdio/hist.c b/tools/perf/ui/stdio/hist.c index 42e432b..2df8eb1 100644 --- a/tools/perf/ui/stdio/hist.c +++ b/tools/perf/ui/stdio/hist.c @@ -35,9 +35,6 @@ static size_t inline__fprintf(struct map *map, u64 ip, int left_margin, if (dso == NULL) return 0; - if (dso->kernel != DSO_TYPE_USER) - return 0; - node = dso__parse_addr_inlines(dso, map__rip_2objdump(map, ip)); if (node == NULL) -- cgit v1.1 From 6200e49423f8023eba54cf0b01076d6f8c0af6ae Mon Sep 17 00:00:00 2001 From: David Carrillo-Cisneros Date: Mon, 17 Jul 2017 21:25:34 -0700 Subject: perf header: Encapsulate read and swap Most callers of readn() in perf header read either a 32 or a 64 bits number, error check it and swap it, if necessary. Create do_read_u32 and do_read_u64 to simplify these use cases. Signed-off-by: David Carrillo-Cisneros Acked-by: David Ahern Acked-by: Jiri Olsa Cc: Alexander Shishkin Cc: Andi Kleen Cc: He Kuang Cc: Masami Hiramatsu Cc: Namhyung Kim Cc: Paul Turner Cc: Peter Zijlstra Cc: Simon Que Cc: Stephane Eranian Cc: Wang Nan Link: http://lkml.kernel.org/r/20170718042549.145161-2-davidcc@google.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/header.c | 210 +++++++++++++++++------------------------------ 1 file changed, 77 insertions(+), 133 deletions(-) (limited to 'tools') diff --git a/tools/perf/util/header.c b/tools/perf/util/header.c index 76ed7d0..24e842c 100644 --- a/tools/perf/util/header.c +++ b/tools/perf/util/header.c @@ -119,25 +119,54 @@ static int do_write_string(int fd, const char *str) return write_padded(fd, str, olen, len); } +static int __do_read(int fd, void *addr, ssize_t size) +{ + ssize_t ret = readn(fd, addr, size); + + if (ret != size) + return ret < 0 ? (int)ret : -1; + return 0; +} + +static int do_read_u32(int fd, struct perf_header *ph, u32 *addr) +{ + int ret; + + ret = __do_read(fd, addr, sizeof(*addr)); + if (ret) + return ret; + + if (ph->needs_swap) + *addr = bswap_32(*addr); + return 0; +} + +static int do_read_u64(int fd, struct perf_header *ph, u64 *addr) +{ + int ret; + + ret = __do_read(fd, addr, sizeof(*addr)); + if (ret) + return ret; + + if (ph->needs_swap) + *addr = bswap_64(*addr); + return 0; +} + static char *do_read_string(int fd, struct perf_header *ph) { - ssize_t sz, ret; u32 len; char *buf; - sz = readn(fd, &len, sizeof(len)); - if (sz < (ssize_t)sizeof(len)) + if (do_read_u32(fd, ph, &len)) return NULL; - if (ph->needs_swap) - len = bswap_32(len); - buf = malloc(len); if (!buf) return NULL; - ret = readn(fd, buf, len); - if (ret == (ssize_t)len) { + if (!__do_read(fd, buf, len)) { /* * strings are padded by zeroes * thus the actual strlen of buf @@ -1188,24 +1217,15 @@ read_event_desc(struct perf_header *ph, int fd) u64 *id; void *buf = NULL; u32 nre, sz, nr, i, j; - ssize_t ret; size_t msz; /* number of events */ - ret = readn(fd, &nre, sizeof(nre)); - if (ret != (ssize_t)sizeof(nre)) + if (do_read_u32(fd, ph, &nre)) goto error; - if (ph->needs_swap) - nre = bswap_32(nre); - - ret = readn(fd, &sz, sizeof(sz)); - if (ret != (ssize_t)sizeof(sz)) + if (do_read_u32(fd, ph, &sz)) goto error; - if (ph->needs_swap) - sz = bswap_32(sz); - /* buffer to hold on file attr struct */ buf = malloc(sz); if (!buf) @@ -1227,8 +1247,7 @@ read_event_desc(struct perf_header *ph, int fd) * must read entire on-file attr struct to * sync up with layout. */ - ret = readn(fd, buf, sz); - if (ret != (ssize_t)sz) + if (__do_read(fd, buf, sz)) goto error; if (ph->needs_swap) @@ -1236,16 +1255,15 @@ read_event_desc(struct perf_header *ph, int fd) memcpy(&evsel->attr, buf, msz); - ret = readn(fd, &nr, sizeof(nr)); - if (ret != (ssize_t)sizeof(nr)) + if (do_read_u32(fd, ph, &nr)) goto error; - if (ph->needs_swap) { - nr = bswap_32(nr); + if (ph->needs_swap) evsel->needs_swap = true; - } evsel->name = do_read_string(fd, ph); + if (!evsel->name) + goto error; if (!nr) continue; @@ -1257,11 +1275,8 @@ read_event_desc(struct perf_header *ph, int fd) evsel->id = id; for (j = 0 ; j < nr; j++) { - ret = readn(fd, id, sizeof(*id)); - if (ret != (ssize_t)sizeof(*id)) + if (do_read_u64(fd, ph, id)) goto error; - if (ph->needs_swap) - *id = bswap_64(*id); id++; } } @@ -1641,26 +1656,18 @@ static int process_nrcpus(struct perf_file_section *section __maybe_unused, struct perf_header *ph, int fd, void *data __maybe_unused) { - ssize_t ret; - u32 nr; - - ret = readn(fd, &nr, sizeof(nr)); - if (ret != sizeof(nr)) - return -1; - - if (ph->needs_swap) - nr = bswap_32(nr); - - ph->env.nr_cpus_avail = nr; - - ret = readn(fd, &nr, sizeof(nr)); - if (ret != sizeof(nr)) - return -1; + int ret; + u32 nr_cpus_avail, nr_cpus_online; - if (ph->needs_swap) - nr = bswap_32(nr); + ret = do_read_u32(fd, ph, &nr_cpus_avail); + if (ret) + return ret; - ph->env.nr_cpus_online = nr; + ret = do_read_u32(fd, ph, &nr_cpus_online); + if (ret) + return ret; + ph->env.nr_cpus_avail = (int)nr_cpus_avail; + ph->env.nr_cpus_online = (int)nr_cpus_online; return 0; } @@ -1684,17 +1691,13 @@ static int process_total_mem(struct perf_file_section *section __maybe_unused, struct perf_header *ph, int fd, void *data __maybe_unused) { - uint64_t mem; - ssize_t ret; + u64 total_mem; + int ret; - ret = readn(fd, &mem, sizeof(mem)); - if (ret != sizeof(mem)) + ret = do_read_u64(fd, ph, &total_mem); + if (ret) return -1; - - if (ph->needs_swap) - mem = bswap_64(mem); - - ph->env.total_mem = mem; + ph->env.total_mem = (unsigned long long)total_mem; return 0; } @@ -1754,17 +1757,12 @@ static int process_cmdline(struct perf_file_section *section, struct perf_header *ph, int fd, void *data __maybe_unused) { - ssize_t ret; char *str, *cmdline = NULL, **argv = NULL; u32 nr, i, len = 0; - ret = readn(fd, &nr, sizeof(nr)); - if (ret != sizeof(nr)) + if (do_read_u32(fd, ph, &nr)) return -1; - if (ph->needs_swap) - nr = bswap_32(nr); - ph->env.nr_cmdline = nr; cmdline = zalloc(section->size + nr + 1); @@ -1799,7 +1797,6 @@ static int process_cpu_topology(struct perf_file_section *section, struct perf_header *ph, int fd, void *data __maybe_unused) { - ssize_t ret; u32 nr, i; char *str; struct strbuf sb; @@ -1810,13 +1807,9 @@ static int process_cpu_topology(struct perf_file_section *section, if (!ph->env.cpu) return -1; - ret = readn(fd, &nr, sizeof(nr)); - if (ret != sizeof(nr)) + if (do_read_u32(fd, ph, &nr)) goto free_cpu; - if (ph->needs_swap) - nr = bswap_32(nr); - ph->env.nr_sibling_cores = nr; size += sizeof(u32); if (strbuf_init(&sb, 128) < 0) @@ -1835,13 +1828,9 @@ static int process_cpu_topology(struct perf_file_section *section, } ph->env.sibling_cores = strbuf_detach(&sb, NULL); - ret = readn(fd, &nr, sizeof(nr)); - if (ret != sizeof(nr)) + if (do_read_u32(fd, ph, &nr)) return -1; - if (ph->needs_swap) - nr = bswap_32(nr); - ph->env.nr_sibling_threads = nr; size += sizeof(u32); @@ -1868,22 +1857,14 @@ static int process_cpu_topology(struct perf_file_section *section, } for (i = 0; i < (u32)cpu_nr; i++) { - ret = readn(fd, &nr, sizeof(nr)); - if (ret != sizeof(nr)) + if (do_read_u32(fd, ph, &nr)) goto free_cpu; - if (ph->needs_swap) - nr = bswap_32(nr); - ph->env.cpu[i].core_id = nr; - ret = readn(fd, &nr, sizeof(nr)); - if (ret != sizeof(nr)) + if (do_read_u32(fd, ph, &nr)) goto free_cpu; - if (ph->needs_swap) - nr = bswap_32(nr); - if (nr != (u32)-1 && nr > (u32)cpu_nr) { pr_debug("socket_id number is too big." "You may need to upgrade the perf tool.\n"); @@ -1907,18 +1888,13 @@ static int process_numa_topology(struct perf_file_section *section __maybe_unuse void *data __maybe_unused) { struct numa_node *nodes, *n; - ssize_t ret; u32 nr, i; char *str; /* nr nodes */ - ret = readn(fd, &nr, sizeof(nr)); - if (ret != sizeof(nr)) + if (do_read_u32(fd, ph, &nr)) return -1; - if (ph->needs_swap) - nr = bswap_32(nr); - nodes = zalloc(sizeof(*nodes) * nr); if (!nodes) return -ENOMEM; @@ -1927,24 +1903,15 @@ static int process_numa_topology(struct perf_file_section *section __maybe_unuse n = &nodes[i]; /* node number */ - ret = readn(fd, &n->node, sizeof(u32)); - if (ret != sizeof(n->node)) + if (do_read_u32(fd, ph, &n->node)) goto error; - ret = readn(fd, &n->mem_total, sizeof(u64)); - if (ret != sizeof(u64)) + if (do_read_u64(fd, ph, &n->mem_total)) goto error; - ret = readn(fd, &n->mem_free, sizeof(u64)); - if (ret != sizeof(u64)) + if (do_read_u64(fd, ph, &n->mem_free)) goto error; - if (ph->needs_swap) { - n->node = bswap_32(n->node); - n->mem_total = bswap_64(n->mem_total); - n->mem_free = bswap_64(n->mem_free); - } - str = do_read_string(fd, ph); if (!str) goto error; @@ -1968,19 +1935,14 @@ static int process_pmu_mappings(struct perf_file_section *section __maybe_unused struct perf_header *ph, int fd, void *data __maybe_unused) { - ssize_t ret; char *name; u32 pmu_num; u32 type; struct strbuf sb; - ret = readn(fd, &pmu_num, sizeof(pmu_num)); - if (ret != sizeof(pmu_num)) + if (do_read_u32(fd, ph, &pmu_num)) return -1; - if (ph->needs_swap) - pmu_num = bswap_32(pmu_num); - if (!pmu_num) { pr_debug("pmu mappings not available\n"); return 0; @@ -1991,10 +1953,8 @@ static int process_pmu_mappings(struct perf_file_section *section __maybe_unused return -1; while (pmu_num) { - if (readn(fd, &type, sizeof(type)) != sizeof(type)) + if (do_read_u32(fd, ph, &type)) goto error; - if (ph->needs_swap) - type = bswap_32(type); name = do_read_string(fd, ph); if (!name) @@ -2034,12 +1994,9 @@ static int process_group_desc(struct perf_file_section *section __maybe_unused, u32 nr_members; } *desc; - if (readn(fd, &nr_groups, sizeof(nr_groups)) != sizeof(nr_groups)) + if (do_read_u32(fd, ph, &nr_groups)) return -1; - if (ph->needs_swap) - nr_groups = bswap_32(nr_groups); - ph->env.nr_groups = nr_groups; if (!nr_groups) { pr_debug("group desc not available\n"); @@ -2055,16 +2012,11 @@ static int process_group_desc(struct perf_file_section *section __maybe_unused, if (!desc[i].name) goto out_free; - if (readn(fd, &desc[i].leader_idx, sizeof(u32)) != sizeof(u32)) + if (do_read_u32(fd, ph, &desc[i].leader_idx)) goto out_free; - if (readn(fd, &desc[i].nr_members, sizeof(u32)) != sizeof(u32)) + if (do_read_u32(fd, ph, &desc[i].nr_members)) goto out_free; - - if (ph->needs_swap) { - desc[i].leader_idx = bswap_32(desc[i].leader_idx); - desc[i].nr_members = bswap_32(desc[i].nr_members); - } } /* @@ -2137,21 +2089,15 @@ static int process_cache(struct perf_file_section *section __maybe_unused, struct cpu_cache_level *caches; u32 cnt, i, version; - if (readn(fd, &version, sizeof(version)) != sizeof(version)) + if (do_read_u32(fd, ph, &version)) return -1; - if (ph->needs_swap) - version = bswap_32(version); - if (version != 1) return -1; - if (readn(fd, &cnt, sizeof(cnt)) != sizeof(cnt)) + if (do_read_u32(fd, ph, &cnt)) return -1; - if (ph->needs_swap) - cnt = bswap_32(cnt); - caches = zalloc(sizeof(*caches) * cnt); if (!caches) return -1; @@ -2160,10 +2106,8 @@ static int process_cache(struct perf_file_section *section __maybe_unused, struct cpu_cache_level c; #define _R(v) \ - if (readn(fd, &c.v, sizeof(u32)) != sizeof(u32))\ + if (do_read_u32(fd, ph, &c.v))\ goto out_free_caches; \ - if (ph->needs_swap) \ - c.v = bswap_32(c.v); \ _R(level) _R(line_size) -- cgit v1.1 From dfaa1580efcfb6b9537043ba0447747d7179fb26 Mon Sep 17 00:00:00 2001 From: David Carrillo-Cisneros Date: Mon, 17 Jul 2017 21:25:35 -0700 Subject: perf header: Add PROCESS_STR_FUN macro Simplify code by adding a macro to handle the common case of processing header features that are a simple string. Signed-off-by: David Carrillo-Cisneros Acked-by: David Ahern Acked-by: Jiri Olsa Cc: Alexander Shishkin Cc: Andi Kleen Cc: He Kuang Cc: Masami Hiramatsu Cc: Namhyung Kim Cc: Paul Turner Cc: Peter Zijlstra Cc: Simon Que Cc: Stephane Eranian Cc: Wang Nan Link: http://lkml.kernel.org/r/20170718042549.145161-3-davidcc@google.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/header.c | 65 +++++++++++++----------------------------------- 1 file changed, 17 insertions(+), 48 deletions(-) (limited to 'tools') diff --git a/tools/perf/util/header.c b/tools/perf/util/header.c index 24e842c..3b833f0 100644 --- a/tools/perf/util/header.c +++ b/tools/perf/util/header.c @@ -1603,6 +1603,23 @@ out: return err; } +/* Macro for features that simply need to read and store a string. */ +#define FEAT_PROCESS_STR_FUN(__feat, __feat_env) \ +static int process_##__feat(struct perf_file_section *section __maybe_unused, \ + struct perf_header *ph, int fd, \ + void *data __maybe_unused) \ +{\ + ph->env.__feat_env = do_read_string(fd, ph); \ + return ph->env.__feat_env ? 0 : -ENOMEM; \ +} + +FEAT_PROCESS_STR_FUN(hostname, hostname); +FEAT_PROCESS_STR_FUN(osrelease, os_release); +FEAT_PROCESS_STR_FUN(version, version); +FEAT_PROCESS_STR_FUN(arch, arch); +FEAT_PROCESS_STR_FUN(cpudesc, cpu_desc); +FEAT_PROCESS_STR_FUN(cpuid, cpuid); + static int process_tracing_data(struct perf_file_section *section __maybe_unused, struct perf_header *ph __maybe_unused, int fd, void *data) @@ -1620,38 +1637,6 @@ static int process_build_id(struct perf_file_section *section, return 0; } -static int process_hostname(struct perf_file_section *section __maybe_unused, - struct perf_header *ph, int fd, - void *data __maybe_unused) -{ - ph->env.hostname = do_read_string(fd, ph); - return ph->env.hostname ? 0 : -ENOMEM; -} - -static int process_osrelease(struct perf_file_section *section __maybe_unused, - struct perf_header *ph, int fd, - void *data __maybe_unused) -{ - ph->env.os_release = do_read_string(fd, ph); - return ph->env.os_release ? 0 : -ENOMEM; -} - -static int process_version(struct perf_file_section *section __maybe_unused, - struct perf_header *ph, int fd, - void *data __maybe_unused) -{ - ph->env.version = do_read_string(fd, ph); - return ph->env.version ? 0 : -ENOMEM; -} - -static int process_arch(struct perf_file_section *section __maybe_unused, - struct perf_header *ph, int fd, - void *data __maybe_unused) -{ - ph->env.arch = do_read_string(fd, ph); - return ph->env.arch ? 0 : -ENOMEM; -} - static int process_nrcpus(struct perf_file_section *section __maybe_unused, struct perf_header *ph, int fd, void *data __maybe_unused) @@ -1671,22 +1656,6 @@ static int process_nrcpus(struct perf_file_section *section __maybe_unused, return 0; } -static int process_cpudesc(struct perf_file_section *section __maybe_unused, - struct perf_header *ph, int fd, - void *data __maybe_unused) -{ - ph->env.cpu_desc = do_read_string(fd, ph); - return ph->env.cpu_desc ? 0 : -ENOMEM; -} - -static int process_cpuid(struct perf_file_section *section __maybe_unused, - struct perf_header *ph, int fd, - void *data __maybe_unused) -{ - ph->env.cpuid = do_read_string(fd, ph); - return ph->env.cpuid ? 0 : -ENOMEM; -} - static int process_total_mem(struct perf_file_section *section __maybe_unused, struct perf_header *ph, int fd, void *data __maybe_unused) -- cgit v1.1 From 2ff5365d75e164032f64133914046fd6be213d94 Mon Sep 17 00:00:00 2001 From: David Carrillo-Cisneros Date: Mon, 17 Jul 2017 21:25:36 -0700 Subject: perf header: Fail on write_padded error Do not proceed if write_padded() error failed. Also, add comments to remind that the return value of write_* functions in util/header.c is an errno code and not the number of bytes written. Signed-off-by: David Carrillo-Cisneros Acked-by: David Ahern Acked-by: Jiri Olsa Cc: Alexander Shishkin Cc: Andi Kleen Cc: He Kuang Cc: Masami Hiramatsu Cc: Namhyung Kim Cc: Paul Turner Cc: Peter Zijlstra Cc: Simon Que Cc: Stephane Eranian Cc: Wang Nan Link: http://lkml.kernel.org/r/20170718042549.145161-4-davidcc@google.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/header.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) (limited to 'tools') diff --git a/tools/perf/util/header.c b/tools/perf/util/header.c index 3b833f0..8dda19b 100644 --- a/tools/perf/util/header.c +++ b/tools/perf/util/header.c @@ -74,6 +74,7 @@ bool perf_header__has_feat(const struct perf_header *header, int feat) return test_bit(feat, header->adds_features); } +/* Return: 0 if succeded, -ERR if failed. */ static int do_write(int fd, const void *buf, size_t size) { while (size) { @@ -89,6 +90,7 @@ static int do_write(int fd, const void *buf, size_t size) return 0; } +/* Return: 0 if succeded, -ERR if failed. */ int write_padded(int fd, const void *bf, size_t count, size_t count_aligned) { static const char zero_buf[NAME_ALIGN]; @@ -103,6 +105,7 @@ int write_padded(int fd, const void *bf, size_t count, size_t count_aligned) #define string_size(str) \ (PERF_ALIGN((strlen(str) + 1), NAME_ALIGN) + sizeof(u32)) +/* Return: 0 if succeded, -ERR if failed. */ static int do_write_string(int fd, const char *str) { u32 len, olen; @@ -3200,7 +3203,8 @@ int perf_event__synthesize_tracing_data(struct perf_tool *tool, int fd, */ tracing_data_put(tdata); - write_padded(fd, NULL, 0, padding); + if (write_padded(fd, NULL, 0, padding)) + return -1; return aligned_size; } -- cgit v1.1 From 7c72440506e2108494bab3b68a4118fa61a9dbf4 Mon Sep 17 00:00:00 2001 From: David Carrillo-Cisneros Date: Mon, 17 Jul 2017 21:25:37 -0700 Subject: perf util: Add const modifier to buf in "writen" function Make buf in helper function "writen" constant to simplify the life of its callers. This requires to hack a cast of buf prior to passing it to "ion" which is simpler than the alternative of reworking the "ion" function to provide a read and a write paths, the latter with constant buf argument. Signed-off-by: David Carrillo-Cisneros Acked-by: David Ahern Acked-by: Jiri Olsa Cc: Alexander Shishkin Cc: Andi Kleen Cc: He Kuang Cc: Masami Hiramatsu Cc: Namhyung Kim Cc: Paul Turner Cc: Peter Zijlstra Cc: Simon Que Cc: Stephane Eranian Cc: Wang Nan Link: http://lkml.kernel.org/r/20170718042549.145161-5-davidcc@google.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/util.c | 6 ++++-- tools/perf/util/util.h | 2 +- 2 files changed, 5 insertions(+), 3 deletions(-) (limited to 'tools') diff --git a/tools/perf/util/util.c b/tools/perf/util/util.c index 9e4ea85..4c360da 100644 --- a/tools/perf/util/util.c +++ b/tools/perf/util/util.c @@ -281,6 +281,7 @@ static ssize_t ion(bool is_read, int fd, void *buf, size_t n) size_t left = n; while (left) { + /* buf must be treated as const if !is_read. */ ssize_t ret = is_read ? read(fd, buf, left) : write(fd, buf, left); @@ -308,9 +309,10 @@ ssize_t readn(int fd, void *buf, size_t n) /* * Write exactly 'n' bytes or return an error. */ -ssize_t writen(int fd, void *buf, size_t n) +ssize_t writen(int fd, const void *buf, size_t n) { - return ion(false, fd, buf, n); + /* ion does not modify buf. */ + return ion(false, fd, (void *)buf, n); } size_t hex_width(u64 v) diff --git a/tools/perf/util/util.h b/tools/perf/util/util.h index 062dd20..b136c27 100644 --- a/tools/perf/util/util.h +++ b/tools/perf/util/util.h @@ -38,7 +38,7 @@ int copyfile_ns(const char *from, const char *to, struct nsinfo *nsi); int copyfile_offset(int fromfd, loff_t from_ofs, int tofd, loff_t to_ofs, u64 size); ssize_t readn(int fd, void *buf, size_t n); -ssize_t writen(int fd, void *buf, size_t n); +ssize_t writen(int fd, const void *buf, size_t n); size_t hex_width(u64 v); int hex2u64(const char *ptr, u64 *val); -- cgit v1.1 From 3b8f51a677ce574f69671e3f7822b4d8f8634ef3 Mon Sep 17 00:00:00 2001 From: David Carrillo-Cisneros Date: Mon, 17 Jul 2017 21:25:38 -0700 Subject: perf header: Revamp do_write() Now that writen takes a const buffer, use it in do_write instead of duplicating its functionality. Export do_write to use it consistently in header.c and build_id.c . Signed-off-by: David Carrillo-Cisneros Acked-by: David Ahern Acked-by: Jiri Olsa Cc: Alexander Shishkin Cc: Andi Kleen Cc: He Kuang Cc: Masami Hiramatsu Cc: Namhyung Kim Cc: Paul Turner Cc: Peter Zijlstra Cc: Simon Que Cc: Stephane Eranian Cc: Wang Nan Link: http://lkml.kernel.org/r/20170718042549.145161-6-davidcc@google.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/build-id.c | 2 +- tools/perf/util/header.c | 14 +++++--------- tools/perf/util/header.h | 2 ++ 3 files changed, 8 insertions(+), 10 deletions(-) (limited to 'tools') diff --git a/tools/perf/util/build-id.c b/tools/perf/util/build-id.c index e966515..4baa532 100644 --- a/tools/perf/util/build-id.c +++ b/tools/perf/util/build-id.c @@ -304,7 +304,7 @@ static int write_buildid(const char *name, size_t name_len, u8 *build_id, b.header.misc = misc; b.header.size = sizeof(b) + len; - err = writen(fd, &b, sizeof(b)); + err = do_write(fd, &b, sizeof(b)); if (err < 0) return err; diff --git a/tools/perf/util/header.c b/tools/perf/util/header.c index 8dda19b..954f0ef 100644 --- a/tools/perf/util/header.c +++ b/tools/perf/util/header.c @@ -75,17 +75,13 @@ bool perf_header__has_feat(const struct perf_header *header, int feat) } /* Return: 0 if succeded, -ERR if failed. */ -static int do_write(int fd, const void *buf, size_t size) +int do_write(int fd, const void *buf, size_t size) { - while (size) { - int ret = write(fd, buf, size); - - if (ret < 0) - return -errno; + ssize_t ret; - size -= ret; - buf += ret; - } + ret = writen(fd, buf, size); + if (ret != (ssize_t)size) + return ret < 0 ? (int)ret : -1; return 0; } diff --git a/tools/perf/util/header.h b/tools/perf/util/header.h index d30109b..e98489c 100644 --- a/tools/perf/util/header.h +++ b/tools/perf/util/header.h @@ -144,6 +144,8 @@ bool is_perf_magic(u64 magic); #define NAME_ALIGN 64 +int do_write(int fd, const void *buf, size_t size); + int write_padded(int fd, const void *bf, size_t count, size_t count_aligned); /* -- cgit v1.1 From ccebbeb6b69e4e172450d32f1059fefd1659ad8c Mon Sep 17 00:00:00 2001 From: David Carrillo-Cisneros Date: Mon, 17 Jul 2017 21:25:39 -0700 Subject: perf header: Add struct feat_fd for write Introduce struct feat_fd. This patch uses it as a wrapper around fd in write_* functions for feature headers. Next patches will extend its functionality to other feature header functions. This patch does not change behavior. Signed-off-by: David Carrillo-Cisneros Acked-by: David Ahern Acked-by: Jiri Olsa Cc: Alexander Shishkin Cc: Andi Kleen Cc: He Kuang Cc: Masami Hiramatsu Cc: Namhyung Kim Cc: Paul Turner Cc: Peter Zijlstra Cc: Simon Que Cc: Stephane Eranian Cc: Wang Nan Link: http://lkml.kernel.org/r/20170718042549.145161-7-davidcc@google.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/build-id.c | 8 +- tools/perf/util/build-id.h | 4 +- tools/perf/util/header.c | 230 ++++++++++++++++++++++++--------------------- tools/perf/util/header.h | 7 +- 4 files changed, 138 insertions(+), 111 deletions(-) (limited to 'tools') diff --git a/tools/perf/util/build-id.c b/tools/perf/util/build-id.c index 4baa532..c1a06fc 100644 --- a/tools/perf/util/build-id.c +++ b/tools/perf/util/build-id.c @@ -289,7 +289,7 @@ char *dso__build_id_filename(const struct dso *dso, char *bf, size_t size, else static int write_buildid(const char *name, size_t name_len, u8 *build_id, - pid_t pid, u16 misc, int fd) + pid_t pid, u16 misc, struct feat_fd *fd) { int err; struct build_id_event b; @@ -311,7 +311,8 @@ static int write_buildid(const char *name, size_t name_len, u8 *build_id, return write_padded(fd, name, name_len + 1, len); } -static int machine__write_buildid_table(struct machine *machine, int fd) +static int machine__write_buildid_table(struct machine *machine, + struct feat_fd *fd) { int err = 0; char nm[PATH_MAX]; @@ -356,7 +357,8 @@ static int machine__write_buildid_table(struct machine *machine, int fd) return err; } -int perf_session__write_buildid_table(struct perf_session *session, int fd) +int perf_session__write_buildid_table(struct perf_session *session, + struct feat_fd *fd) { struct rb_node *nd; int err = machine__write_buildid_table(&session->machines.host, fd); diff --git a/tools/perf/util/build-id.h b/tools/perf/util/build-id.h index 113dc06..c94b0dc 100644 --- a/tools/perf/util/build-id.h +++ b/tools/perf/util/build-id.h @@ -10,6 +10,7 @@ extern struct perf_tool build_id__mark_dso_hit_ops; struct dso; +struct feat_fd; int build_id__sprintf(const u8 *build_id, int len, char *bf); int sysfs__sprintf_build_id(const char *root_dir, char *sbuild_id); @@ -27,7 +28,8 @@ int build_id__mark_dso_hit(struct perf_tool *tool, union perf_event *event, int dsos__hit_all(struct perf_session *session); bool perf_session__read_build_ids(struct perf_session *session, bool with_hits); -int perf_session__write_buildid_table(struct perf_session *session, int fd); +int perf_session__write_buildid_table(struct perf_session *session, + struct feat_fd *fd); int perf_session__cache_build_ids(struct perf_session *session); char *build_id_cache__origname(const char *sbuild_id); diff --git a/tools/perf/util/header.c b/tools/perf/util/header.c index 954f0ef..a5db9ad 100644 --- a/tools/perf/util/header.c +++ b/tools/perf/util/header.c @@ -59,6 +59,11 @@ struct perf_file_attr { struct perf_file_section ids; }; +struct feat_fd { + struct perf_header *ph; + int fd; +}; + void perf_header__set_feat(struct perf_header *header, int feat) { set_bit(feat, header->adds_features); @@ -75,11 +80,11 @@ bool perf_header__has_feat(const struct perf_header *header, int feat) } /* Return: 0 if succeded, -ERR if failed. */ -int do_write(int fd, const void *buf, size_t size) +int do_write(struct feat_fd *ff, const void *buf, size_t size) { ssize_t ret; - ret = writen(fd, buf, size); + ret = writen(ff->fd, buf, size); if (ret != (ssize_t)size) return ret < 0 ? (int)ret : -1; @@ -87,13 +92,14 @@ int do_write(int fd, const void *buf, size_t size) } /* Return: 0 if succeded, -ERR if failed. */ -int write_padded(int fd, const void *bf, size_t count, size_t count_aligned) +int write_padded(struct feat_fd *ff, const void *bf, + size_t count, size_t count_aligned) { static const char zero_buf[NAME_ALIGN]; - int err = do_write(fd, bf, count); + int err = do_write(ff, bf, count); if (!err) - err = do_write(fd, zero_buf, count_aligned - count); + err = do_write(ff, zero_buf, count_aligned - count); return err; } @@ -102,7 +108,7 @@ int write_padded(int fd, const void *bf, size_t count, size_t count_aligned) (PERF_ALIGN((strlen(str) + 1), NAME_ALIGN) + sizeof(u32)) /* Return: 0 if succeded, -ERR if failed. */ -static int do_write_string(int fd, const char *str) +static int do_write_string(struct feat_fd *ff, const char *str) { u32 len, olen; int ret; @@ -111,11 +117,11 @@ static int do_write_string(int fd, const char *str) len = PERF_ALIGN(olen, NAME_ALIGN); /* write len, incl. \0 */ - ret = do_write(fd, &len, sizeof(len)); + ret = do_write(ff, &len, sizeof(len)); if (ret < 0) return ret; - return write_padded(fd, str, olen, len); + return write_padded(ff, str, olen, len); } static int __do_read(int fd, void *addr, ssize_t size) @@ -178,25 +184,24 @@ static char *do_read_string(int fd, struct perf_header *ph) return NULL; } -static int write_tracing_data(int fd, struct perf_header *h __maybe_unused, - struct perf_evlist *evlist) +static int write_tracing_data(struct feat_fd *ff, + struct perf_evlist *evlist) { - return read_tracing_data(fd, &evlist->entries); + return read_tracing_data(ff->fd, &evlist->entries); } - -static int write_build_id(int fd, struct perf_header *h, +static int write_build_id(struct feat_fd *ff, struct perf_evlist *evlist __maybe_unused) { struct perf_session *session; int err; - session = container_of(h, struct perf_session, header); + session = container_of(ff->ph, struct perf_session, header); if (!perf_session__read_build_ids(session, true)) return -1; - err = perf_session__write_buildid_table(session, fd); + err = perf_session__write_buildid_table(session, ff); if (err < 0) { pr_debug("failed to write buildid table\n"); return err; @@ -206,7 +211,7 @@ static int write_build_id(int fd, struct perf_header *h, return 0; } -static int write_hostname(int fd, struct perf_header *h __maybe_unused, +static int write_hostname(struct feat_fd *ff, struct perf_evlist *evlist __maybe_unused) { struct utsname uts; @@ -216,10 +221,10 @@ static int write_hostname(int fd, struct perf_header *h __maybe_unused, if (ret < 0) return -1; - return do_write_string(fd, uts.nodename); + return do_write_string(ff, uts.nodename); } -static int write_osrelease(int fd, struct perf_header *h __maybe_unused, +static int write_osrelease(struct feat_fd *ff, struct perf_evlist *evlist __maybe_unused) { struct utsname uts; @@ -229,10 +234,10 @@ static int write_osrelease(int fd, struct perf_header *h __maybe_unused, if (ret < 0) return -1; - return do_write_string(fd, uts.release); + return do_write_string(ff, uts.release); } -static int write_arch(int fd, struct perf_header *h __maybe_unused, +static int write_arch(struct feat_fd *ff, struct perf_evlist *evlist __maybe_unused) { struct utsname uts; @@ -242,16 +247,16 @@ static int write_arch(int fd, struct perf_header *h __maybe_unused, if (ret < 0) return -1; - return do_write_string(fd, uts.machine); + return do_write_string(ff, uts.machine); } -static int write_version(int fd, struct perf_header *h __maybe_unused, +static int write_version(struct feat_fd *ff, struct perf_evlist *evlist __maybe_unused) { - return do_write_string(fd, perf_version_string); + return do_write_string(ff, perf_version_string); } -static int __write_cpudesc(int fd, const char *cpuinfo_proc) +static int __write_cpudesc(struct feat_fd *ff, const char *cpuinfo_proc) { FILE *file; char *buf = NULL; @@ -301,14 +306,14 @@ static int __write_cpudesc(int fd, const char *cpuinfo_proc) } p++; } - ret = do_write_string(fd, s); + ret = do_write_string(ff, s); done: free(buf); fclose(file); return ret; } -static int write_cpudesc(int fd, struct perf_header *h __maybe_unused, +static int write_cpudesc(struct feat_fd *ff, struct perf_evlist *evlist __maybe_unused) { #ifndef CPUINFO_PROC @@ -319,7 +324,7 @@ static int write_cpudesc(int fd, struct perf_header *h __maybe_unused, for (i = 0; i < ARRAY_SIZE(cpuinfo_procs); i++) { int ret; - ret = __write_cpudesc(fd, cpuinfo_procs[i]); + ret = __write_cpudesc(ff, cpuinfo_procs[i]); if (ret >= 0) return ret; } @@ -327,7 +332,7 @@ static int write_cpudesc(int fd, struct perf_header *h __maybe_unused, } -static int write_nrcpus(int fd, struct perf_header *h __maybe_unused, +static int write_nrcpus(struct feat_fd *ff, struct perf_evlist *evlist __maybe_unused) { long nr; @@ -342,14 +347,14 @@ static int write_nrcpus(int fd, struct perf_header *h __maybe_unused, nra = (u32)(nr & UINT_MAX); - ret = do_write(fd, &nrc, sizeof(nrc)); + ret = do_write(ff, &nrc, sizeof(nrc)); if (ret < 0) return ret; - return do_write(fd, &nra, sizeof(nra)); + return do_write(ff, &nra, sizeof(nra)); } -static int write_event_desc(int fd, struct perf_header *h __maybe_unused, +static int write_event_desc(struct feat_fd *ff, struct perf_evlist *evlist) { struct perf_evsel *evsel; @@ -361,7 +366,7 @@ static int write_event_desc(int fd, struct perf_header *h __maybe_unused, /* * write number of events */ - ret = do_write(fd, &nre, sizeof(nre)); + ret = do_write(ff, &nre, sizeof(nre)); if (ret < 0) return ret; @@ -369,12 +374,12 @@ static int write_event_desc(int fd, struct perf_header *h __maybe_unused, * size of perf_event_attr struct */ sz = (u32)sizeof(evsel->attr); - ret = do_write(fd, &sz, sizeof(sz)); + ret = do_write(ff, &sz, sizeof(sz)); if (ret < 0) return ret; evlist__for_each_entry(evlist, evsel) { - ret = do_write(fd, &evsel->attr, sz); + ret = do_write(ff, &evsel->attr, sz); if (ret < 0) return ret; /* @@ -385,27 +390,27 @@ static int write_event_desc(int fd, struct perf_header *h __maybe_unused, * type of ids, */ nri = evsel->ids; - ret = do_write(fd, &nri, sizeof(nri)); + ret = do_write(ff, &nri, sizeof(nri)); if (ret < 0) return ret; /* * write event string as passed on cmdline */ - ret = do_write_string(fd, perf_evsel__name(evsel)); + ret = do_write_string(ff, perf_evsel__name(evsel)); if (ret < 0) return ret; /* * write unique ids for this event */ - ret = do_write(fd, evsel->id, evsel->ids * sizeof(u64)); + ret = do_write(ff, evsel->id, evsel->ids * sizeof(u64)); if (ret < 0) return ret; } return 0; } -static int write_cmdline(int fd, struct perf_header *h __maybe_unused, +static int write_cmdline(struct feat_fd *ff, struct perf_evlist *evlist __maybe_unused) { char buf[MAXPATHLEN]; @@ -423,16 +428,16 @@ static int write_cmdline(int fd, struct perf_header *h __maybe_unused, /* account for binary path */ n = perf_env.nr_cmdline + 1; - ret = do_write(fd, &n, sizeof(n)); + ret = do_write(ff, &n, sizeof(n)); if (ret < 0) return ret; - ret = do_write_string(fd, buf); + ret = do_write_string(ff, buf); if (ret < 0) return ret; for (i = 0 ; i < perf_env.nr_cmdline; i++) { - ret = do_write_string(fd, perf_env.cmdline_argv[i]); + ret = do_write_string(ff, perf_env.cmdline_argv[i]); if (ret < 0) return ret; } @@ -585,8 +590,8 @@ out_free: return tp; } -static int write_cpu_topology(int fd, struct perf_header *h __maybe_unused, - struct perf_evlist *evlist __maybe_unused) +static int write_cpu_topology(struct feat_fd *ff, + struct perf_evlist *evlist __maybe_unused) { struct cpu_topo *tp; u32 i; @@ -596,21 +601,21 @@ static int write_cpu_topology(int fd, struct perf_header *h __maybe_unused, if (!tp) return -1; - ret = do_write(fd, &tp->core_sib, sizeof(tp->core_sib)); + ret = do_write(ff, &tp->core_sib, sizeof(tp->core_sib)); if (ret < 0) goto done; for (i = 0; i < tp->core_sib; i++) { - ret = do_write_string(fd, tp->core_siblings[i]); + ret = do_write_string(ff, tp->core_siblings[i]); if (ret < 0) goto done; } - ret = do_write(fd, &tp->thread_sib, sizeof(tp->thread_sib)); + ret = do_write(ff, &tp->thread_sib, sizeof(tp->thread_sib)); if (ret < 0) goto done; for (i = 0; i < tp->thread_sib; i++) { - ret = do_write_string(fd, tp->thread_siblings[i]); + ret = do_write_string(ff, tp->thread_siblings[i]); if (ret < 0) break; } @@ -620,11 +625,11 @@ static int write_cpu_topology(int fd, struct perf_header *h __maybe_unused, goto done; for (j = 0; j < perf_env.nr_cpus_avail; j++) { - ret = do_write(fd, &perf_env.cpu[j].core_id, + ret = do_write(ff, &perf_env.cpu[j].core_id, sizeof(perf_env.cpu[j].core_id)); if (ret < 0) return ret; - ret = do_write(fd, &perf_env.cpu[j].socket_id, + ret = do_write(ff, &perf_env.cpu[j].socket_id, sizeof(perf_env.cpu[j].socket_id)); if (ret < 0) return ret; @@ -636,8 +641,8 @@ done: -static int write_total_mem(int fd, struct perf_header *h __maybe_unused, - struct perf_evlist *evlist __maybe_unused) +static int write_total_mem(struct feat_fd *ff, + struct perf_evlist *evlist __maybe_unused) { char *buf = NULL; FILE *fp; @@ -657,7 +662,7 @@ static int write_total_mem(int fd, struct perf_header *h __maybe_unused, if (!ret) { n = sscanf(buf, "%*s %"PRIu64, &mem); if (n == 1) - ret = do_write(fd, &mem, sizeof(mem)); + ret = do_write(ff, &mem, sizeof(mem)); } else ret = -1; free(buf); @@ -665,7 +670,7 @@ static int write_total_mem(int fd, struct perf_header *h __maybe_unused, return ret; } -static int write_topo_node(int fd, int node) +static int write_topo_node(struct feat_fd *ff, int node) { char str[MAXPATHLEN]; char field[32]; @@ -695,11 +700,11 @@ static int write_topo_node(int fd, int node) fclose(fp); fp = NULL; - ret = do_write(fd, &mem_total, sizeof(u64)); + ret = do_write(ff, &mem_total, sizeof(u64)); if (ret) goto done; - ret = do_write(fd, &mem_free, sizeof(u64)); + ret = do_write(ff, &mem_free, sizeof(u64)); if (ret) goto done; @@ -717,7 +722,7 @@ static int write_topo_node(int fd, int node) if (p) *p = '\0'; - ret = do_write_string(fd, buf); + ret = do_write_string(ff, buf); done: free(buf); if (fp) @@ -725,8 +730,8 @@ done: return ret; } -static int write_numa_topology(int fd, struct perf_header *h __maybe_unused, - struct perf_evlist *evlist __maybe_unused) +static int write_numa_topology(struct feat_fd *ff, + struct perf_evlist *evlist __maybe_unused) { char *buf = NULL; size_t len = 0; @@ -753,17 +758,17 @@ static int write_numa_topology(int fd, struct perf_header *h __maybe_unused, nr = (u32)node_map->nr; - ret = do_write(fd, &nr, sizeof(nr)); + ret = do_write(ff, &nr, sizeof(nr)); if (ret < 0) goto done; for (i = 0; i < nr; i++) { j = (u32)node_map->map[i]; - ret = do_write(fd, &j, sizeof(j)); + ret = do_write(ff, &j, sizeof(j)); if (ret < 0) break; - ret = write_topo_node(fd, i); + ret = write_topo_node(ff, i); if (ret < 0) break; } @@ -786,16 +791,16 @@ done: * }; */ -static int write_pmu_mappings(int fd, struct perf_header *h __maybe_unused, +static int write_pmu_mappings(struct feat_fd *ff, struct perf_evlist *evlist __maybe_unused) { struct perf_pmu *pmu = NULL; - off_t offset = lseek(fd, 0, SEEK_CUR); + off_t offset = lseek(ff->fd, 0, SEEK_CUR); __u32 pmu_num = 0; int ret; /* write real pmu_num later */ - ret = do_write(fd, &pmu_num, sizeof(pmu_num)); + ret = do_write(ff, &pmu_num, sizeof(pmu_num)); if (ret < 0) return ret; @@ -804,18 +809,18 @@ static int write_pmu_mappings(int fd, struct perf_header *h __maybe_unused, continue; pmu_num++; - ret = do_write(fd, &pmu->type, sizeof(pmu->type)); + ret = do_write(ff, &pmu->type, sizeof(pmu->type)); if (ret < 0) return ret; - ret = do_write_string(fd, pmu->name); + ret = do_write_string(ff, pmu->name); if (ret < 0) return ret; } - if (pwrite(fd, &pmu_num, sizeof(pmu_num), offset) != sizeof(pmu_num)) { + if (pwrite(ff->fd, &pmu_num, sizeof(pmu_num), offset) != sizeof(pmu_num)) { /* discard all */ - lseek(fd, offset, SEEK_SET); + lseek(ff->fd, offset, SEEK_SET); return -1; } @@ -834,14 +839,14 @@ static int write_pmu_mappings(int fd, struct perf_header *h __maybe_unused, * }[nr_groups]; * }; */ -static int write_group_desc(int fd, struct perf_header *h __maybe_unused, +static int write_group_desc(struct feat_fd *ff, struct perf_evlist *evlist) { u32 nr_groups = evlist->nr_groups; struct perf_evsel *evsel; int ret; - ret = do_write(fd, &nr_groups, sizeof(nr_groups)); + ret = do_write(ff, &nr_groups, sizeof(nr_groups)); if (ret < 0) return ret; @@ -852,15 +857,15 @@ static int write_group_desc(int fd, struct perf_header *h __maybe_unused, u32 leader_idx = evsel->idx; u32 nr_members = evsel->nr_members; - ret = do_write_string(fd, name); + ret = do_write_string(ff, name); if (ret < 0) return ret; - ret = do_write(fd, &leader_idx, sizeof(leader_idx)); + ret = do_write(ff, &leader_idx, sizeof(leader_idx)); if (ret < 0) return ret; - ret = do_write(fd, &nr_members, sizeof(nr_members)); + ret = do_write(ff, &nr_members, sizeof(nr_members)); if (ret < 0) return ret; } @@ -877,7 +882,7 @@ int __weak get_cpuid(char *buffer __maybe_unused, size_t sz __maybe_unused) return -1; } -static int write_cpuid(int fd, struct perf_header *h __maybe_unused, +static int write_cpuid(struct feat_fd *ff, struct perf_evlist *evlist __maybe_unused) { char buffer[64]; @@ -889,25 +894,24 @@ static int write_cpuid(int fd, struct perf_header *h __maybe_unused, return -1; write_it: - return do_write_string(fd, buffer); + return do_write_string(ff, buffer); } -static int write_branch_stack(int fd __maybe_unused, - struct perf_header *h __maybe_unused, - struct perf_evlist *evlist __maybe_unused) +static int write_branch_stack(struct feat_fd *ff __maybe_unused, + struct perf_evlist *evlist __maybe_unused) { return 0; } -static int write_auxtrace(int fd, struct perf_header *h, +static int write_auxtrace(struct feat_fd *ff, struct perf_evlist *evlist __maybe_unused) { struct perf_session *session; int err; - session = container_of(h, struct perf_session, header); + session = container_of(ff->ph, struct perf_session, header); - err = auxtrace_index__write(fd, &session->auxtrace_index); + err = auxtrace_index__write(ff->fd, &session->auxtrace_index); if (err < 0) pr_err("Failed to write auxtrace index\n"); return err; @@ -1054,8 +1058,8 @@ static int build_caches(struct cpu_cache_level caches[], u32 size, u32 *cntp) #define MAX_CACHES 2000 -static int write_cache(int fd, struct perf_header *h __maybe_unused, - struct perf_evlist *evlist __maybe_unused) +static int write_cache(struct feat_fd *ff, + struct perf_evlist *evlist __maybe_unused) { struct cpu_cache_level caches[MAX_CACHES]; u32 cnt = 0, i, version = 1; @@ -1067,11 +1071,11 @@ static int write_cache(int fd, struct perf_header *h __maybe_unused, qsort(&caches, cnt, sizeof(struct cpu_cache_level), cpu_cache_level__sort); - ret = do_write(fd, &version, sizeof(u32)); + ret = do_write(ff, &version, sizeof(u32)); if (ret < 0) goto out; - ret = do_write(fd, &cnt, sizeof(u32)); + ret = do_write(ff, &cnt, sizeof(u32)); if (ret < 0) goto out; @@ -1079,7 +1083,7 @@ static int write_cache(int fd, struct perf_header *h __maybe_unused, struct cpu_cache_level *c = &caches[i]; #define _W(v) \ - ret = do_write(fd, &c->v, sizeof(u32)); \ + ret = do_write(ff, &c->v, sizeof(u32)); \ if (ret < 0) \ goto out; @@ -1090,7 +1094,7 @@ static int write_cache(int fd, struct perf_header *h __maybe_unused, #undef _W #define _W(v) \ - ret = do_write_string(fd, (const char *) c->v); \ + ret = do_write_string(ff, (const char *) c->v); \ if (ret < 0) \ goto out; @@ -1106,8 +1110,7 @@ out: return ret; } -static int write_stat(int fd __maybe_unused, - struct perf_header *h __maybe_unused, +static int write_stat(struct feat_fd *ff __maybe_unused, struct perf_evlist *evlist __maybe_unused) { return 0; @@ -2105,7 +2108,7 @@ out_free_caches: } struct feature_ops { - int (*write)(int fd, struct perf_header *h, struct perf_evlist *evlist); + int (*write)(struct feat_fd *ff, struct perf_evlist *evlist); void (*print)(struct perf_header *h, int fd, FILE *fp); int (*process)(struct perf_file_section *section, struct perf_header *h, int fd, void *data); @@ -2214,29 +2217,29 @@ int perf_header__fprintf_info(struct perf_session *session, FILE *fp, bool full) return 0; } -static int do_write_feat(int fd, struct perf_header *h, int type, +static int do_write_feat(struct feat_fd *ff, int type, struct perf_file_section **p, struct perf_evlist *evlist) { int err; int ret = 0; - if (perf_header__has_feat(h, type)) { + if (perf_header__has_feat(ff->ph, type)) { if (!feat_ops[type].write) return -1; - (*p)->offset = lseek(fd, 0, SEEK_CUR); + (*p)->offset = lseek(ff->fd, 0, SEEK_CUR); - err = feat_ops[type].write(fd, h, evlist); + err = feat_ops[type].write(ff, evlist); if (err < 0) { pr_debug("failed to write feature %s\n", feat_ops[type].name); /* undo anything written */ - lseek(fd, (*p)->offset, SEEK_SET); + lseek(ff->fd, (*p)->offset, SEEK_SET); return -1; } - (*p)->size = lseek(fd, 0, SEEK_CUR) - (*p)->offset; + (*p)->size = lseek(ff->fd, 0, SEEK_CUR) - (*p)->offset; (*p)++; } return ret; @@ -2246,12 +2249,18 @@ static int perf_header__adds_write(struct perf_header *header, struct perf_evlist *evlist, int fd) { int nr_sections; + struct feat_fd ff; struct perf_file_section *feat_sec, *p; int sec_size; u64 sec_start; int feat; int err; + ff = (struct feat_fd){ + .fd = fd, + .ph = header, + }; + nr_sections = bitmap_weight(header->adds_features, HEADER_FEAT_BITS); if (!nr_sections) return 0; @@ -2266,7 +2275,7 @@ static int perf_header__adds_write(struct perf_header *header, lseek(fd, sec_start + sec_size, SEEK_SET); for_each_set_bit(feat, header->adds_features, HEADER_FEAT_BITS) { - if (do_write_feat(fd, header, feat, &p, evlist)) + if (do_write_feat(&ff, feat, &p, evlist)) perf_header__clear_feat(header, feat); } @@ -2275,7 +2284,7 @@ static int perf_header__adds_write(struct perf_header *header, * may write more than needed due to dropped feature, but * this is okay, reader will skip the mising entries */ - err = do_write(fd, feat_sec, sec_size); + err = do_write(&ff, feat_sec, sec_size); if (err < 0) pr_debug("failed to write feature section\n"); free(feat_sec); @@ -2285,14 +2294,17 @@ static int perf_header__adds_write(struct perf_header *header, int perf_header__write_pipe(int fd) { struct perf_pipe_file_header f_header; + struct feat_fd ff; int err; + ff = (struct feat_fd){ .fd = fd }; + f_header = (struct perf_pipe_file_header){ .magic = PERF_MAGIC, .size = sizeof(f_header), }; - err = do_write(fd, &f_header, sizeof(f_header)); + err = do_write(&ff, &f_header, sizeof(f_header)); if (err < 0) { pr_debug("failed to write perf pipe header\n"); return err; @@ -2309,21 +2321,23 @@ int perf_session__write_header(struct perf_session *session, struct perf_file_attr f_attr; struct perf_header *header = &session->header; struct perf_evsel *evsel; + struct feat_fd ff; u64 attr_offset; int err; + ff = (struct feat_fd){ .fd = fd}; lseek(fd, sizeof(f_header), SEEK_SET); evlist__for_each_entry(session->evlist, evsel) { evsel->id_offset = lseek(fd, 0, SEEK_CUR); - err = do_write(fd, evsel->id, evsel->ids * sizeof(u64)); + err = do_write(&ff, evsel->id, evsel->ids * sizeof(u64)); if (err < 0) { pr_debug("failed to write perf header\n"); return err; } } - attr_offset = lseek(fd, 0, SEEK_CUR); + attr_offset = lseek(ff.fd, 0, SEEK_CUR); evlist__for_each_entry(evlist, evsel) { f_attr = (struct perf_file_attr){ @@ -2333,7 +2347,7 @@ int perf_session__write_header(struct perf_session *session, .size = evsel->ids * sizeof(u64), } }; - err = do_write(fd, &f_attr, sizeof(f_attr)); + err = do_write(&ff, &f_attr, sizeof(f_attr)); if (err < 0) { pr_debug("failed to write perf header attribute\n"); return err; @@ -2368,7 +2382,7 @@ int perf_session__write_header(struct perf_session *session, memcpy(&f_header.adds_features, &header->adds_features, sizeof(header->adds_features)); lseek(fd, 0, SEEK_SET); - err = do_write(fd, &f_header, sizeof(f_header)); + err = do_write(&ff, &f_header, sizeof(f_header)); if (err < 0) { pr_debug("failed to write perf header\n"); return err; @@ -2643,6 +2657,10 @@ static int perf_file_header__read_pipe(struct perf_pipe_file_header *header, struct perf_header *ph, int fd, bool repipe) { + struct feat_fd ff = { + .fd = STDOUT_FILENO, + .ph = ph, + }; ssize_t ret; ret = readn(fd, header, sizeof(*header)); @@ -2657,7 +2675,7 @@ static int perf_file_header__read_pipe(struct perf_pipe_file_header *header, if (ph->needs_swap) header->size = bswap_64(header->size); - if (repipe && do_write(STDOUT_FILENO, header, sizeof(*header)) < 0) + if (repipe && do_write(&ff, header, sizeof(*header)) < 0) return -1; return 0; @@ -3165,6 +3183,7 @@ int perf_event__synthesize_tracing_data(struct perf_tool *tool, int fd, union perf_event ev; struct tracing_data *tdata; ssize_t size = 0, aligned_size = 0, padding; + struct feat_fd ff; int err __maybe_unused = 0; /* @@ -3199,7 +3218,8 @@ int perf_event__synthesize_tracing_data(struct perf_tool *tool, int fd, */ tracing_data_put(tdata); - if (write_padded(fd, NULL, 0, padding)) + ff = (struct feat_fd){ .fd = fd }; + if (write_padded(&ff, NULL, 0, padding)) return -1; return aligned_size; diff --git a/tools/perf/util/header.h b/tools/perf/util/header.h index e98489c..9d8dcd5 100644 --- a/tools/perf/util/header.h +++ b/tools/perf/util/header.h @@ -144,9 +144,12 @@ bool is_perf_magic(u64 magic); #define NAME_ALIGN 64 -int do_write(int fd, const void *buf, size_t size); +struct feat_fd; -int write_padded(int fd, const void *bf, size_t count, size_t count_aligned); +int do_write(struct feat_fd *fd, const void *buf, size_t size); + +int write_padded(struct feat_fd *fd, const void *bf, + size_t count, size_t count_aligned); /* * arch specific callback -- cgit v1.1 From cfc654209e27ecaa36f550a0934f3c78f9c9179f Mon Sep 17 00:00:00 2001 From: David Carrillo-Cisneros Date: Mon, 17 Jul 2017 21:25:40 -0700 Subject: perf header: Use struct feat_fd for print As preparation for using header records in pipe mode, replace int fd with struct feat_fd ff in print functions for all header record types. Signed-off-by: David Carrillo-Cisneros Acked-by: David Ahern Acked-by: Jiri Olsa Cc: Alexander Shishkin Cc: Andi Kleen Cc: He Kuang Cc: Masami Hiramatsu Cc: Namhyung Kim Cc: Paul Turner Cc: Peter Zijlstra Cc: Simon Que Cc: Stephane Eranian Cc: Wang Nan Link: http://lkml.kernel.org/r/20170718042549.145161-8-davidcc@google.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/header.c | 102 ++++++++++++++++++++++------------------------- 1 file changed, 47 insertions(+), 55 deletions(-) (limited to 'tools') diff --git a/tools/perf/util/header.c b/tools/perf/util/header.c index a5db9ad..58e888a 100644 --- a/tools/perf/util/header.c +++ b/tools/perf/util/header.c @@ -1116,62 +1116,56 @@ static int write_stat(struct feat_fd *ff __maybe_unused, return 0; } -static void print_hostname(struct perf_header *ph, int fd __maybe_unused, - FILE *fp) +static void print_hostname(struct feat_fd *ff, FILE *fp) { - fprintf(fp, "# hostname : %s\n", ph->env.hostname); + fprintf(fp, "# hostname : %s\n", ff->ph->env.hostname); } -static void print_osrelease(struct perf_header *ph, int fd __maybe_unused, - FILE *fp) +static void print_osrelease(struct feat_fd *ff, FILE *fp) { - fprintf(fp, "# os release : %s\n", ph->env.os_release); + fprintf(fp, "# os release : %s\n", ff->ph->env.os_release); } -static void print_arch(struct perf_header *ph, int fd __maybe_unused, FILE *fp) +static void print_arch(struct feat_fd *ff, FILE *fp) { - fprintf(fp, "# arch : %s\n", ph->env.arch); + fprintf(fp, "# arch : %s\n", ff->ph->env.arch); } -static void print_cpudesc(struct perf_header *ph, int fd __maybe_unused, - FILE *fp) +static void print_cpudesc(struct feat_fd *ff, FILE *fp) { - fprintf(fp, "# cpudesc : %s\n", ph->env.cpu_desc); + fprintf(fp, "# cpudesc : %s\n", ff->ph->env.cpu_desc); } -static void print_nrcpus(struct perf_header *ph, int fd __maybe_unused, - FILE *fp) +static void print_nrcpus(struct feat_fd *ff, FILE *fp) { - fprintf(fp, "# nrcpus online : %u\n", ph->env.nr_cpus_online); - fprintf(fp, "# nrcpus avail : %u\n", ph->env.nr_cpus_avail); + fprintf(fp, "# nrcpus online : %u\n", ff->ph->env.nr_cpus_online); + fprintf(fp, "# nrcpus avail : %u\n", ff->ph->env.nr_cpus_avail); } -static void print_version(struct perf_header *ph, int fd __maybe_unused, - FILE *fp) +static void print_version(struct feat_fd *ff, FILE *fp) { - fprintf(fp, "# perf version : %s\n", ph->env.version); + fprintf(fp, "# perf version : %s\n", ff->ph->env.version); } -static void print_cmdline(struct perf_header *ph, int fd __maybe_unused, - FILE *fp) +static void print_cmdline(struct feat_fd *ff, FILE *fp) { int nr, i; - nr = ph->env.nr_cmdline; + nr = ff->ph->env.nr_cmdline; fprintf(fp, "# cmdline : "); for (i = 0; i < nr; i++) - fprintf(fp, "%s ", ph->env.cmdline_argv[i]); + fprintf(fp, "%s ", ff->ph->env.cmdline_argv[i]); fputc('\n', fp); } -static void print_cpu_topology(struct perf_header *ph, int fd __maybe_unused, - FILE *fp) +static void print_cpu_topology(struct feat_fd *ff, FILE *fp) { + struct perf_header *ph = ff->ph; + int cpu_nr = ph->env.nr_cpus_avail; int nr, i; char *str; - int cpu_nr = ph->env.nr_cpus_avail; nr = ph->env.nr_sibling_cores; str = ph->env.sibling_cores; @@ -1297,9 +1291,9 @@ static int __desc_attr__fprintf(FILE *fp, const char *name, const char *val, return fprintf(fp, ", %s = %s", name, val); } -static void print_event_desc(struct perf_header *ph, int fd, FILE *fp) +static void print_event_desc(struct feat_fd *ff, FILE *fp) { - struct perf_evsel *evsel, *events = read_event_desc(ph, fd); + struct perf_evsel *evsel, *events = read_event_desc(ff->ph, ff->fd); u32 j; u64 *id; @@ -1329,20 +1323,18 @@ static void print_event_desc(struct perf_header *ph, int fd, FILE *fp) free_event_desc(events); } -static void print_total_mem(struct perf_header *ph, int fd __maybe_unused, - FILE *fp) +static void print_total_mem(struct feat_fd *ff, FILE *fp) { - fprintf(fp, "# total memory : %Lu kB\n", ph->env.total_mem); + fprintf(fp, "# total memory : %llu kB\n", ff->ph->env.total_mem); } -static void print_numa_topology(struct perf_header *ph, int fd __maybe_unused, - FILE *fp) +static void print_numa_topology(struct feat_fd *ff, FILE *fp) { int i; struct numa_node *n; - for (i = 0; i < ph->env.nr_numa_nodes; i++) { - n = &ph->env.numa_nodes[i]; + for (i = 0; i < ff->ph->env.nr_numa_nodes; i++) { + n = &ff->ph->env.numa_nodes[i]; fprintf(fp, "# node%u meminfo : total = %"PRIu64" kB," " free = %"PRIu64" kB\n", @@ -1353,56 +1345,51 @@ static void print_numa_topology(struct perf_header *ph, int fd __maybe_unused, } } -static void print_cpuid(struct perf_header *ph, int fd __maybe_unused, FILE *fp) +static void print_cpuid(struct feat_fd *ff, FILE *fp) { - fprintf(fp, "# cpuid : %s\n", ph->env.cpuid); + fprintf(fp, "# cpuid : %s\n", ff->ph->env.cpuid); } -static void print_branch_stack(struct perf_header *ph __maybe_unused, - int fd __maybe_unused, FILE *fp) +static void print_branch_stack(struct feat_fd *ff __maybe_unused, FILE *fp) { fprintf(fp, "# contains samples with branch stack\n"); } -static void print_auxtrace(struct perf_header *ph __maybe_unused, - int fd __maybe_unused, FILE *fp) +static void print_auxtrace(struct feat_fd *ff __maybe_unused, FILE *fp) { fprintf(fp, "# contains AUX area data (e.g. instruction trace)\n"); } -static void print_stat(struct perf_header *ph __maybe_unused, - int fd __maybe_unused, FILE *fp) +static void print_stat(struct feat_fd *ff __maybe_unused, FILE *fp) { fprintf(fp, "# contains stat data\n"); } -static void print_cache(struct perf_header *ph __maybe_unused, - int fd __maybe_unused, FILE *fp __maybe_unused) +static void print_cache(struct feat_fd *ff, FILE *fp __maybe_unused) { int i; fprintf(fp, "# CPU cache info:\n"); - for (i = 0; i < ph->env.caches_cnt; i++) { + for (i = 0; i < ff->ph->env.caches_cnt; i++) { fprintf(fp, "# "); - cpu_cache_level__fprintf(fp, &ph->env.caches[i]); + cpu_cache_level__fprintf(fp, &ff->ph->env.caches[i]); } } -static void print_pmu_mappings(struct perf_header *ph, int fd __maybe_unused, - FILE *fp) +static void print_pmu_mappings(struct feat_fd *ff, FILE *fp) { const char *delimiter = "# pmu mappings: "; char *str, *tmp; u32 pmu_num; u32 type; - pmu_num = ph->env.nr_pmu_mappings; + pmu_num = ff->ph->env.nr_pmu_mappings; if (!pmu_num) { fprintf(fp, "# pmu mappings: not available\n"); return; } - str = ph->env.pmu_mappings; + str = ff->ph->env.pmu_mappings; while (pmu_num) { type = strtoul(str, &tmp, 0); @@ -1425,14 +1412,13 @@ error: fprintf(fp, "# pmu mappings: unable to read\n"); } -static void print_group_desc(struct perf_header *ph, int fd __maybe_unused, - FILE *fp) +static void print_group_desc(struct feat_fd *ff, FILE *fp) { struct perf_session *session; struct perf_evsel *evsel; u32 nr = 0; - session = container_of(ph, struct perf_session, header); + session = container_of(ff->ph, struct perf_session, header); evlist__for_each_entry(session->evlist, evsel) { if (perf_evsel__is_group_leader(evsel) && @@ -2109,7 +2095,7 @@ out_free_caches: struct feature_ops { int (*write)(struct feat_fd *ff, struct perf_evlist *evlist); - void (*print)(struct perf_header *h, int fd, FILE *fp); + void (*print)(struct feat_fd *ff, FILE *fp); int (*process)(struct perf_file_section *section, struct perf_header *h, int fd, void *data); const char *name; @@ -2162,6 +2148,7 @@ static int perf_file_section__fprintf_info(struct perf_file_section *section, int feat, int fd, void *data) { struct header_print_data *hd = data; + struct feat_fd ff; if (lseek(fd, section->offset, SEEK_SET) == (off_t)-1) { pr_debug("Failed to lseek to %" PRIu64 " offset for feature " @@ -2175,8 +2162,13 @@ static int perf_file_section__fprintf_info(struct perf_file_section *section, if (!feat_ops[feat].print) return 0; + ff = (struct feat_fd) { + .fd = fd, + .ph = ph, + }; + if (!feat_ops[feat].full_only || hd->full) - feat_ops[feat].print(ph, fd, hd->fp); + feat_ops[feat].print(&ff, hd->fp); else fprintf(hd->fp, "# %s info available, use -I to display\n", feat_ops[feat].name); -- cgit v1.1 From 1a22275449f4bd6255e24f0c5b6c7fa61b263417 Mon Sep 17 00:00:00 2001 From: David Carrillo-Cisneros Date: Mon, 17 Jul 2017 21:25:41 -0700 Subject: perf header: Use struct feat_fd to process header records As preparation for using header records in pipe-mode, replace int fd with struct feat_fd ff in process functions for all header record types. This patch does not change behavior. Signed-off-by: David Carrillo-Cisneros Acked-by: David Ahern Acked-by: Jiri Olsa Cc: Alexander Shishkin Cc: Andi Kleen Cc: He Kuang Cc: Masami Hiramatsu Cc: Namhyung Kim Cc: Paul Turner Cc: Peter Zijlstra Cc: Simon Que Cc: Stephane Eranian Cc: Wang Nan Link: http://lkml.kernel.org/r/20170718042549.145161-9-davidcc@google.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/header.c | 164 +++++++++++++++++++++++------------------------ 1 file changed, 79 insertions(+), 85 deletions(-) (limited to 'tools') diff --git a/tools/perf/util/header.c b/tools/perf/util/header.c index 58e888a..22b52e5 100644 --- a/tools/perf/util/header.c +++ b/tools/perf/util/header.c @@ -1594,11 +1594,10 @@ out: /* Macro for features that simply need to read and store a string. */ #define FEAT_PROCESS_STR_FUN(__feat, __feat_env) \ static int process_##__feat(struct perf_file_section *section __maybe_unused, \ - struct perf_header *ph, int fd, \ - void *data __maybe_unused) \ + struct feat_fd *ff, void *data __maybe_unused) \ {\ - ph->env.__feat_env = do_read_string(fd, ph); \ - return ph->env.__feat_env ? 0 : -ENOMEM; \ + ff->ph->env.__feat_env = do_read_string(ff->fd, ff->ph); \ + return ff->ph->env.__feat_env ? 0 : -ENOMEM; \ } FEAT_PROCESS_STR_FUN(hostname, hostname); @@ -1609,52 +1608,49 @@ FEAT_PROCESS_STR_FUN(cpudesc, cpu_desc); FEAT_PROCESS_STR_FUN(cpuid, cpuid); static int process_tracing_data(struct perf_file_section *section __maybe_unused, - struct perf_header *ph __maybe_unused, - int fd, void *data) + struct feat_fd *ff, void *data) { - ssize_t ret = trace_report(fd, data, false); + ssize_t ret = trace_report(ff->fd, data, false); + return ret < 0 ? -1 : 0; } static int process_build_id(struct perf_file_section *section, - struct perf_header *ph, int fd, - void *data __maybe_unused) + struct feat_fd *ff, void *data __maybe_unused) { - if (perf_header__read_build_ids(ph, fd, section->offset, section->size)) + if (perf_header__read_build_ids(ff->ph, ff->fd, section->offset, section->size)) pr_debug("Failed to read buildids, continuing...\n"); return 0; } static int process_nrcpus(struct perf_file_section *section __maybe_unused, - struct perf_header *ph, int fd, - void *data __maybe_unused) + struct feat_fd *ff, void *data __maybe_unused) { int ret; u32 nr_cpus_avail, nr_cpus_online; - ret = do_read_u32(fd, ph, &nr_cpus_avail); + ret = do_read_u32(ff->fd, ff->ph, &nr_cpus_avail); if (ret) return ret; - ret = do_read_u32(fd, ph, &nr_cpus_online); + ret = do_read_u32(ff->fd, ff->ph, &nr_cpus_online); if (ret) return ret; - ph->env.nr_cpus_avail = (int)nr_cpus_avail; - ph->env.nr_cpus_online = (int)nr_cpus_online; + ff->ph->env.nr_cpus_avail = (int)nr_cpus_avail; + ff->ph->env.nr_cpus_online = (int)nr_cpus_online; return 0; } static int process_total_mem(struct perf_file_section *section __maybe_unused, - struct perf_header *ph, int fd, - void *data __maybe_unused) + struct feat_fd *ff, void *data __maybe_unused) { u64 total_mem; int ret; - ret = do_read_u64(fd, ph, &total_mem); + ret = do_read_u64(ff->fd, ff->ph, &total_mem); if (ret) return -1; - ph->env.total_mem = (unsigned long long)total_mem; + ff->ph->env.total_mem = (unsigned long long)total_mem; return 0; } @@ -1692,16 +1688,15 @@ perf_evlist__set_event_name(struct perf_evlist *evlist, static int process_event_desc(struct perf_file_section *section __maybe_unused, - struct perf_header *header, int fd, - void *data __maybe_unused) + struct feat_fd *ff, void *data __maybe_unused) { struct perf_session *session; - struct perf_evsel *evsel, *events = read_event_desc(header, fd); + struct perf_evsel *evsel, *events = read_event_desc(ff->ph, ff->fd); if (!events) return 0; - session = container_of(header, struct perf_session, header); + session = container_of(ff->ph, struct perf_session, header); for (evsel = events; evsel->attr.size; evsel++) perf_evlist__set_event_name(session->evlist, evsel); @@ -1710,17 +1705,16 @@ process_event_desc(struct perf_file_section *section __maybe_unused, return 0; } -static int process_cmdline(struct perf_file_section *section, - struct perf_header *ph, int fd, - void *data __maybe_unused) +static int process_cmdline(struct perf_file_section *section __maybe_unused, + struct feat_fd *ff, void *data __maybe_unused) { char *str, *cmdline = NULL, **argv = NULL; u32 nr, i, len = 0; - if (do_read_u32(fd, ph, &nr)) + if (do_read_u32(ff->fd, ff->ph, &nr)) return -1; - ph->env.nr_cmdline = nr; + ff->ph->env.nr_cmdline = nr; cmdline = zalloc(section->size + nr + 1); if (!cmdline) @@ -1731,7 +1725,7 @@ static int process_cmdline(struct perf_file_section *section, goto error; for (i = 0; i < nr; i++) { - str = do_read_string(fd, ph); + str = do_read_string(ff->fd, ff->ph); if (!str) goto error; @@ -1740,8 +1734,8 @@ static int process_cmdline(struct perf_file_section *section, len += strlen(str) + 1; free(str); } - ph->env.cmdline = cmdline; - ph->env.cmdline_argv = (const char **) argv; + ff->ph->env.cmdline = cmdline; + ff->ph->env.cmdline_argv = (const char **) argv; return 0; error: @@ -1750,21 +1744,21 @@ error: return -1; } -static int process_cpu_topology(struct perf_file_section *section, - struct perf_header *ph, int fd, - void *data __maybe_unused) +static int process_cpu_topology(struct perf_file_section *section __maybe_unused, + struct feat_fd *ff, void *data __maybe_unused) { u32 nr, i; char *str; struct strbuf sb; - int cpu_nr = ph->env.nr_cpus_avail; + int cpu_nr = ff->ph->env.nr_cpus_avail; u64 size = 0; + struct perf_header *ph = ff->ph; ph->env.cpu = calloc(cpu_nr, sizeof(*ph->env.cpu)); if (!ph->env.cpu) return -1; - if (do_read_u32(fd, ph, &nr)) + if (do_read_u32(ff->fd, ph, &nr)) goto free_cpu; ph->env.nr_sibling_cores = nr; @@ -1773,7 +1767,7 @@ static int process_cpu_topology(struct perf_file_section *section, goto free_cpu; for (i = 0; i < nr; i++) { - str = do_read_string(fd, ph); + str = do_read_string(ff->fd, ph); if (!str) goto error; @@ -1785,14 +1779,14 @@ static int process_cpu_topology(struct perf_file_section *section, } ph->env.sibling_cores = strbuf_detach(&sb, NULL); - if (do_read_u32(fd, ph, &nr)) + if (do_read_u32(ff->fd, ph, &nr)) return -1; ph->env.nr_sibling_threads = nr; size += sizeof(u32); for (i = 0; i < nr; i++) { - str = do_read_string(fd, ph); + str = do_read_string(ff->fd, ph); if (!str) goto error; @@ -1814,12 +1808,12 @@ static int process_cpu_topology(struct perf_file_section *section, } for (i = 0; i < (u32)cpu_nr; i++) { - if (do_read_u32(fd, ph, &nr)) + if (do_read_u32(ff->fd, ph, &nr)) goto free_cpu; ph->env.cpu[i].core_id = nr; - if (do_read_u32(fd, ph, &nr)) + if (do_read_u32(ff->fd, ph, &nr)) goto free_cpu; if (nr != (u32)-1 && nr > (u32)cpu_nr) { @@ -1841,15 +1835,14 @@ free_cpu: } static int process_numa_topology(struct perf_file_section *section __maybe_unused, - struct perf_header *ph, int fd, - void *data __maybe_unused) + struct feat_fd *ff, void *data __maybe_unused) { struct numa_node *nodes, *n; u32 nr, i; char *str; /* nr nodes */ - if (do_read_u32(fd, ph, &nr)) + if (do_read_u32(ff->fd, ff->ph, &nr)) return -1; nodes = zalloc(sizeof(*nodes) * nr); @@ -1860,16 +1853,16 @@ static int process_numa_topology(struct perf_file_section *section __maybe_unuse n = &nodes[i]; /* node number */ - if (do_read_u32(fd, ph, &n->node)) + if (do_read_u32(ff->fd, ff->ph, &n->node)) goto error; - if (do_read_u64(fd, ph, &n->mem_total)) + if (do_read_u64(ff->fd, ff->ph, &n->mem_total)) goto error; - if (do_read_u64(fd, ph, &n->mem_free)) + if (do_read_u64(ff->fd, ff->ph, &n->mem_free)) goto error; - str = do_read_string(fd, ph); + str = do_read_string(ff->fd, ff->ph); if (!str) goto error; @@ -1879,8 +1872,8 @@ static int process_numa_topology(struct perf_file_section *section __maybe_unuse free(str); } - ph->env.nr_numa_nodes = nr; - ph->env.numa_nodes = nodes; + ff->ph->env.nr_numa_nodes = nr; + ff->ph->env.numa_nodes = nodes; return 0; error: @@ -1889,15 +1882,14 @@ error: } static int process_pmu_mappings(struct perf_file_section *section __maybe_unused, - struct perf_header *ph, int fd, - void *data __maybe_unused) + struct feat_fd *ff, void *data __maybe_unused) { char *name; u32 pmu_num; u32 type; struct strbuf sb; - if (do_read_u32(fd, ph, &pmu_num)) + if (do_read_u32(ff->fd, ff->ph, &pmu_num)) return -1; if (!pmu_num) { @@ -1905,15 +1897,15 @@ static int process_pmu_mappings(struct perf_file_section *section __maybe_unused return 0; } - ph->env.nr_pmu_mappings = pmu_num; + ff->ph->env.nr_pmu_mappings = pmu_num; if (strbuf_init(&sb, 128) < 0) return -1; while (pmu_num) { - if (do_read_u32(fd, ph, &type)) + if (do_read_u32(ff->fd, ff->ph, &type)) goto error; - name = do_read_string(fd, ph); + name = do_read_string(ff->fd, ff->ph); if (!name) goto error; @@ -1924,12 +1916,12 @@ static int process_pmu_mappings(struct perf_file_section *section __maybe_unused goto error; if (!strcmp(name, "msr")) - ph->env.msr_pmu_type = type; + ff->ph->env.msr_pmu_type = type; free(name); pmu_num--; } - ph->env.pmu_mappings = strbuf_detach(&sb, NULL); + ff->ph->env.pmu_mappings = strbuf_detach(&sb, NULL); return 0; error: @@ -1938,8 +1930,7 @@ error: } static int process_group_desc(struct perf_file_section *section __maybe_unused, - struct perf_header *ph, int fd, - void *data __maybe_unused) + struct feat_fd *ff, void *data __maybe_unused) { size_t ret = -1; u32 i, nr, nr_groups; @@ -1951,10 +1942,10 @@ static int process_group_desc(struct perf_file_section *section __maybe_unused, u32 nr_members; } *desc; - if (do_read_u32(fd, ph, &nr_groups)) + if (do_read_u32(ff->fd, ff->ph, &nr_groups)) return -1; - ph->env.nr_groups = nr_groups; + ff->ph->env.nr_groups = nr_groups; if (!nr_groups) { pr_debug("group desc not available\n"); return 0; @@ -1965,21 +1956,21 @@ static int process_group_desc(struct perf_file_section *section __maybe_unused, return -1; for (i = 0; i < nr_groups; i++) { - desc[i].name = do_read_string(fd, ph); + desc[i].name = do_read_string(ff->fd, ff->ph); if (!desc[i].name) goto out_free; - if (do_read_u32(fd, ph, &desc[i].leader_idx)) + if (do_read_u32(ff->fd, ff->ph, &desc[i].leader_idx)) goto out_free; - if (do_read_u32(fd, ph, &desc[i].nr_members)) + if (do_read_u32(ff->fd, ff->ph, &desc[i].nr_members)) goto out_free; } /* * Rebuild group relationship based on the group_desc */ - session = container_of(ph, struct perf_session, header); + session = container_of(ff->ph, struct perf_session, header); session->evlist->nr_groups = nr_groups; i = nr = 0; @@ -2023,36 +2014,34 @@ out_free: return ret; } -static int process_auxtrace(struct perf_file_section *section, - struct perf_header *ph, int fd, - void *data __maybe_unused) +static int process_auxtrace(struct perf_file_section *section __maybe_unused, + struct feat_fd *ff, void *data __maybe_unused) { struct perf_session *session; int err; - session = container_of(ph, struct perf_session, header); + session = container_of(ff->ph, struct perf_session, header); - err = auxtrace_index__process(fd, section->size, session, - ph->needs_swap); + err = auxtrace_index__process(ff->fd, section->size, session, + ff->ph->needs_swap); if (err < 0) pr_err("Failed to process auxtrace index\n"); return err; } static int process_cache(struct perf_file_section *section __maybe_unused, - struct perf_header *ph __maybe_unused, int fd __maybe_unused, - void *data __maybe_unused) + struct feat_fd *ff, void *data __maybe_unused) { struct cpu_cache_level *caches; u32 cnt, i, version; - if (do_read_u32(fd, ph, &version)) + if (do_read_u32(ff->fd, ff->ph, &version)) return -1; if (version != 1) return -1; - if (do_read_u32(fd, ph, &cnt)) + if (do_read_u32(ff->fd, ff->ph, &cnt)) return -1; caches = zalloc(sizeof(*caches) * cnt); @@ -2063,7 +2052,7 @@ static int process_cache(struct perf_file_section *section __maybe_unused, struct cpu_cache_level c; #define _R(v) \ - if (do_read_u32(fd, ph, &c.v))\ + if (do_read_u32(ff->fd, ff->ph, &c.v))\ goto out_free_caches; \ _R(level) @@ -2072,9 +2061,9 @@ static int process_cache(struct perf_file_section *section __maybe_unused, _R(ways) #undef _R - #define _R(v) \ - c.v = do_read_string(fd, ph); \ - if (!c.v) \ + #define _R(v) \ + c.v = do_read_string(ff->fd, ff->ph); \ + if (!c.v) \ goto out_free_caches; _R(type) @@ -2085,8 +2074,8 @@ static int process_cache(struct perf_file_section *section __maybe_unused, caches[i] = c; } - ph->env.caches = caches; - ph->env.caches_cnt = cnt; + ff->ph->env.caches = caches; + ff->ph->env.caches_cnt = cnt; return 0; out_free_caches: free(caches); @@ -2097,7 +2086,7 @@ struct feature_ops { int (*write)(struct feat_fd *ff, struct perf_evlist *evlist); void (*print)(struct feat_fd *ff, FILE *fp); int (*process)(struct perf_file_section *section, - struct perf_header *h, int fd, void *data); + struct feat_fd *ff, void *data); const char *name; bool full_only; }; @@ -2628,6 +2617,11 @@ static int perf_file_section__process(struct perf_file_section *section, struct perf_header *ph, int feat, int fd, void *data) { + struct feat_fd ff = { + .fd = fd, + .ph = ph, + }; + if (lseek(fd, section->offset, SEEK_SET) == (off_t)-1) { pr_debug("Failed to lseek to %" PRIu64 " offset for feature " "%d, continuing...\n", section->offset, feat); @@ -2642,7 +2636,7 @@ static int perf_file_section__process(struct perf_file_section *section, if (!feat_ops[feat].process) return 0; - return feat_ops[feat].process(section, ph, fd, data); + return feat_ops[feat].process(section, &ff, data); } static int perf_file_header__read_pipe(struct perf_pipe_file_header *header, -- cgit v1.1 From 625524572391326b83f906efe02abbaac9debce6 Mon Sep 17 00:00:00 2001 From: David Carrillo-Cisneros Date: Mon, 17 Jul 2017 21:25:42 -0700 Subject: perf header: Don't pass struct perf_file_section to process_##_feat struct perf_file_section is used in process_##_feat as container for size and offset in the file descriptor. These attributes are meaninful in pipe-mode but struct perf_file_section is not. Add offset and size variables to struct feat_fd to store perf_file_section's values in file-mode. Later on, the same variables can be reused for pipe-mode. This patch does not change behavior. Signed-off-by: David Carrillo-Cisneros Acked-by: David Ahern Acked-by: Jiri Olsa Cc: Alexander Shishkin Cc: Andi Kleen Cc: He Kuang Cc: Masami Hiramatsu Cc: Namhyung Kim Cc: Paul Turner Cc: Peter Zijlstra Cc: Simon Que Cc: Stephane Eranian Cc: Wang Nan Link: http://lkml.kernel.org/r/20170718042549.145161-10-davidcc@google.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/header.c | 58 ++++++++++++++++++++---------------------------- 1 file changed, 24 insertions(+), 34 deletions(-) (limited to 'tools') diff --git a/tools/perf/util/header.c b/tools/perf/util/header.c index 22b52e5..60394d2 100644 --- a/tools/perf/util/header.c +++ b/tools/perf/util/header.c @@ -62,6 +62,8 @@ struct perf_file_attr { struct feat_fd { struct perf_header *ph; int fd; + ssize_t offset; + size_t size; }; void perf_header__set_feat(struct perf_header *header, int feat) @@ -1593,8 +1595,7 @@ out: /* Macro for features that simply need to read and store a string. */ #define FEAT_PROCESS_STR_FUN(__feat, __feat_env) \ -static int process_##__feat(struct perf_file_section *section __maybe_unused, \ - struct feat_fd *ff, void *data __maybe_unused) \ +static int process_##__feat(struct feat_fd *ff, void *data __maybe_unused) \ {\ ff->ph->env.__feat_env = do_read_string(ff->fd, ff->ph); \ return ff->ph->env.__feat_env ? 0 : -ENOMEM; \ @@ -1607,24 +1608,21 @@ FEAT_PROCESS_STR_FUN(arch, arch); FEAT_PROCESS_STR_FUN(cpudesc, cpu_desc); FEAT_PROCESS_STR_FUN(cpuid, cpuid); -static int process_tracing_data(struct perf_file_section *section __maybe_unused, - struct feat_fd *ff, void *data) +static int process_tracing_data(struct feat_fd *ff, void *data) { ssize_t ret = trace_report(ff->fd, data, false); return ret < 0 ? -1 : 0; } -static int process_build_id(struct perf_file_section *section, - struct feat_fd *ff, void *data __maybe_unused) +static int process_build_id(struct feat_fd *ff, void *data __maybe_unused) { - if (perf_header__read_build_ids(ff->ph, ff->fd, section->offset, section->size)) + if (perf_header__read_build_ids(ff->ph, ff->fd, ff->offset, ff->size)) pr_debug("Failed to read buildids, continuing...\n"); return 0; } -static int process_nrcpus(struct perf_file_section *section __maybe_unused, - struct feat_fd *ff, void *data __maybe_unused) +static int process_nrcpus(struct feat_fd *ff, void *data __maybe_unused) { int ret; u32 nr_cpus_avail, nr_cpus_online; @@ -1641,8 +1639,7 @@ static int process_nrcpus(struct perf_file_section *section __maybe_unused, return 0; } -static int process_total_mem(struct perf_file_section *section __maybe_unused, - struct feat_fd *ff, void *data __maybe_unused) +static int process_total_mem(struct feat_fd *ff, void *data __maybe_unused) { u64 total_mem; int ret; @@ -1687,8 +1684,7 @@ perf_evlist__set_event_name(struct perf_evlist *evlist, } static int -process_event_desc(struct perf_file_section *section __maybe_unused, - struct feat_fd *ff, void *data __maybe_unused) +process_event_desc(struct feat_fd *ff, void *data __maybe_unused) { struct perf_session *session; struct perf_evsel *evsel, *events = read_event_desc(ff->ph, ff->fd); @@ -1705,8 +1701,7 @@ process_event_desc(struct perf_file_section *section __maybe_unused, return 0; } -static int process_cmdline(struct perf_file_section *section __maybe_unused, - struct feat_fd *ff, void *data __maybe_unused) +static int process_cmdline(struct feat_fd *ff, void *data __maybe_unused) { char *str, *cmdline = NULL, **argv = NULL; u32 nr, i, len = 0; @@ -1716,7 +1711,7 @@ static int process_cmdline(struct perf_file_section *section __maybe_unused, ff->ph->env.nr_cmdline = nr; - cmdline = zalloc(section->size + nr + 1); + cmdline = zalloc(ff->size + nr + 1); if (!cmdline) return -1; @@ -1744,8 +1739,7 @@ error: return -1; } -static int process_cpu_topology(struct perf_file_section *section __maybe_unused, - struct feat_fd *ff, void *data __maybe_unused) +static int process_cpu_topology(struct feat_fd *ff, void *data __maybe_unused) { u32 nr, i; char *str; @@ -1802,7 +1796,7 @@ static int process_cpu_topology(struct perf_file_section *section __maybe_unused * The header may be from old perf, * which doesn't include core id and socket id information. */ - if (section->size <= size) { + if (ff->size <= size) { zfree(&ph->env.cpu); return 0; } @@ -1834,8 +1828,7 @@ free_cpu: return -1; } -static int process_numa_topology(struct perf_file_section *section __maybe_unused, - struct feat_fd *ff, void *data __maybe_unused) +static int process_numa_topology(struct feat_fd *ff, void *data __maybe_unused) { struct numa_node *nodes, *n; u32 nr, i; @@ -1881,8 +1874,7 @@ error: return -1; } -static int process_pmu_mappings(struct perf_file_section *section __maybe_unused, - struct feat_fd *ff, void *data __maybe_unused) +static int process_pmu_mappings(struct feat_fd *ff, void *data __maybe_unused) { char *name; u32 pmu_num; @@ -1929,8 +1921,7 @@ error: return -1; } -static int process_group_desc(struct perf_file_section *section __maybe_unused, - struct feat_fd *ff, void *data __maybe_unused) +static int process_group_desc(struct feat_fd *ff, void *data __maybe_unused) { size_t ret = -1; u32 i, nr, nr_groups; @@ -2014,23 +2005,21 @@ out_free: return ret; } -static int process_auxtrace(struct perf_file_section *section __maybe_unused, - struct feat_fd *ff, void *data __maybe_unused) +static int process_auxtrace(struct feat_fd *ff, void *data __maybe_unused) { struct perf_session *session; int err; session = container_of(ff->ph, struct perf_session, header); - err = auxtrace_index__process(ff->fd, section->size, session, + err = auxtrace_index__process(ff->fd, ff->size, session, ff->ph->needs_swap); if (err < 0) pr_err("Failed to process auxtrace index\n"); return err; } -static int process_cache(struct perf_file_section *section __maybe_unused, - struct feat_fd *ff, void *data __maybe_unused) +static int process_cache(struct feat_fd *ff, void *data __maybe_unused) { struct cpu_cache_level *caches; u32 cnt, i, version; @@ -2085,8 +2074,7 @@ out_free_caches: struct feature_ops { int (*write)(struct feat_fd *ff, struct perf_evlist *evlist); void (*print)(struct feat_fd *ff, FILE *fp); - int (*process)(struct perf_file_section *section, - struct feat_fd *ff, void *data); + int (*process)(struct feat_fd *ff, void *data); const char *name; bool full_only; }; @@ -2617,9 +2605,11 @@ static int perf_file_section__process(struct perf_file_section *section, struct perf_header *ph, int feat, int fd, void *data) { - struct feat_fd ff = { + struct feat_fd fdd = { .fd = fd, .ph = ph, + .size = section->size, + .offset = section->offset, }; if (lseek(fd, section->offset, SEEK_SET) == (off_t)-1) { @@ -2636,7 +2626,7 @@ static int perf_file_section__process(struct perf_file_section *section, if (!feat_ops[feat].process) return 0; - return feat_ops[feat].process(section, &ff, data); + return feat_ops[feat].process(&fdd, data); } static int perf_file_header__read_pipe(struct perf_pipe_file_header *header, -- cgit v1.1 From 48e5fcea386009fb2515158fdaf8586ce72d86ce Mon Sep 17 00:00:00 2001 From: David Carrillo-Cisneros Date: Mon, 17 Jul 2017 21:25:43 -0700 Subject: perf header: Use struct feat_fd in read header records As preparation for using header records in-pipe mode, replace int fd with struct feat_fd ff in read functions for all header record types. This patch does not change behavior. Signed-off-by: David Carrillo-Cisneros Acked-by: David Ahern Acked-by: Jiri Olsa Cc: Alexander Shishkin Cc: Andi Kleen Cc: He Kuang Cc: Masami Hiramatsu Cc: Namhyung Kim Cc: Paul Turner Cc: Peter Zijlstra Cc: Simon Que Cc: Stephane Eranian Cc: Wang Nan Link: http://lkml.kernel.org/r/20170718042549.145161-11-davidcc@google.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/header.c | 101 +++++++++++++++++++++++------------------------ 1 file changed, 50 insertions(+), 51 deletions(-) (limited to 'tools') diff --git a/tools/perf/util/header.c b/tools/perf/util/header.c index 60394d2..14db9f2 100644 --- a/tools/perf/util/header.c +++ b/tools/perf/util/header.c @@ -126,54 +126,54 @@ static int do_write_string(struct feat_fd *ff, const char *str) return write_padded(ff, str, olen, len); } -static int __do_read(int fd, void *addr, ssize_t size) +static int __do_read(struct feat_fd *ff, void *addr, ssize_t size) { - ssize_t ret = readn(fd, addr, size); + ssize_t ret = readn(ff->fd, addr, size); if (ret != size) return ret < 0 ? (int)ret : -1; return 0; } -static int do_read_u32(int fd, struct perf_header *ph, u32 *addr) +static int do_read_u32(struct feat_fd *ff, u32 *addr) { int ret; - ret = __do_read(fd, addr, sizeof(*addr)); + ret = __do_read(ff, addr, sizeof(*addr)); if (ret) return ret; - if (ph->needs_swap) + if (ff->ph->needs_swap) *addr = bswap_32(*addr); return 0; } -static int do_read_u64(int fd, struct perf_header *ph, u64 *addr) +static int do_read_u64(struct feat_fd *ff, u64 *addr) { int ret; - ret = __do_read(fd, addr, sizeof(*addr)); + ret = __do_read(ff, addr, sizeof(*addr)); if (ret) return ret; - if (ph->needs_swap) + if (ff->ph->needs_swap) *addr = bswap_64(*addr); return 0; } -static char *do_read_string(int fd, struct perf_header *ph) +static char *do_read_string(struct feat_fd *ff) { u32 len; char *buf; - if (do_read_u32(fd, ph, &len)) + if (do_read_u32(ff, &len)) return NULL; buf = malloc(len); if (!buf) return NULL; - if (!__do_read(fd, buf, len)) { + if (!__do_read(ff, buf, len)) { /* * strings are padded by zeroes * thus the actual strlen of buf @@ -1208,8 +1208,7 @@ static void free_event_desc(struct perf_evsel *events) free(events); } -static struct perf_evsel * -read_event_desc(struct perf_header *ph, int fd) +static struct perf_evsel *read_event_desc(struct feat_fd *ff) { struct perf_evsel *evsel, *events = NULL; u64 *id; @@ -1218,10 +1217,10 @@ read_event_desc(struct perf_header *ph, int fd) size_t msz; /* number of events */ - if (do_read_u32(fd, ph, &nre)) + if (do_read_u32(ff, &nre)) goto error; - if (do_read_u32(fd, ph, &sz)) + if (do_read_u32(ff, &sz)) goto error; /* buffer to hold on file attr struct */ @@ -1245,21 +1244,21 @@ read_event_desc(struct perf_header *ph, int fd) * must read entire on-file attr struct to * sync up with layout. */ - if (__do_read(fd, buf, sz)) + if (__do_read(ff, buf, sz)) goto error; - if (ph->needs_swap) + if (ff->ph->needs_swap) perf_event__attr_swap(buf); memcpy(&evsel->attr, buf, msz); - if (do_read_u32(fd, ph, &nr)) + if (do_read_u32(ff, &nr)) goto error; - if (ph->needs_swap) + if (ff->ph->needs_swap) evsel->needs_swap = true; - evsel->name = do_read_string(fd, ph); + evsel->name = do_read_string(ff); if (!evsel->name) goto error; @@ -1273,7 +1272,7 @@ read_event_desc(struct perf_header *ph, int fd) evsel->id = id; for (j = 0 ; j < nr; j++) { - if (do_read_u64(fd, ph, id)) + if (do_read_u64(ff, id)) goto error; id++; } @@ -1295,7 +1294,7 @@ static int __desc_attr__fprintf(FILE *fp, const char *name, const char *val, static void print_event_desc(struct feat_fd *ff, FILE *fp) { - struct perf_evsel *evsel, *events = read_event_desc(ff->ph, ff->fd); + struct perf_evsel *evsel, *events = read_event_desc(ff); u32 j; u64 *id; @@ -1597,7 +1596,7 @@ out: #define FEAT_PROCESS_STR_FUN(__feat, __feat_env) \ static int process_##__feat(struct feat_fd *ff, void *data __maybe_unused) \ {\ - ff->ph->env.__feat_env = do_read_string(ff->fd, ff->ph); \ + ff->ph->env.__feat_env = do_read_string(ff); \ return ff->ph->env.__feat_env ? 0 : -ENOMEM; \ } @@ -1627,11 +1626,11 @@ static int process_nrcpus(struct feat_fd *ff, void *data __maybe_unused) int ret; u32 nr_cpus_avail, nr_cpus_online; - ret = do_read_u32(ff->fd, ff->ph, &nr_cpus_avail); + ret = do_read_u32(ff, &nr_cpus_avail); if (ret) return ret; - ret = do_read_u32(ff->fd, ff->ph, &nr_cpus_online); + ret = do_read_u32(ff, &nr_cpus_online); if (ret) return ret; ff->ph->env.nr_cpus_avail = (int)nr_cpus_avail; @@ -1644,7 +1643,7 @@ static int process_total_mem(struct feat_fd *ff, void *data __maybe_unused) u64 total_mem; int ret; - ret = do_read_u64(ff->fd, ff->ph, &total_mem); + ret = do_read_u64(ff, &total_mem); if (ret) return -1; ff->ph->env.total_mem = (unsigned long long)total_mem; @@ -1687,7 +1686,7 @@ static int process_event_desc(struct feat_fd *ff, void *data __maybe_unused) { struct perf_session *session; - struct perf_evsel *evsel, *events = read_event_desc(ff->ph, ff->fd); + struct perf_evsel *evsel, *events = read_event_desc(ff); if (!events) return 0; @@ -1706,7 +1705,7 @@ static int process_cmdline(struct feat_fd *ff, void *data __maybe_unused) char *str, *cmdline = NULL, **argv = NULL; u32 nr, i, len = 0; - if (do_read_u32(ff->fd, ff->ph, &nr)) + if (do_read_u32(ff, &nr)) return -1; ff->ph->env.nr_cmdline = nr; @@ -1720,7 +1719,7 @@ static int process_cmdline(struct feat_fd *ff, void *data __maybe_unused) goto error; for (i = 0; i < nr; i++) { - str = do_read_string(ff->fd, ff->ph); + str = do_read_string(ff); if (!str) goto error; @@ -1752,7 +1751,7 @@ static int process_cpu_topology(struct feat_fd *ff, void *data __maybe_unused) if (!ph->env.cpu) return -1; - if (do_read_u32(ff->fd, ph, &nr)) + if (do_read_u32(ff, &nr)) goto free_cpu; ph->env.nr_sibling_cores = nr; @@ -1761,7 +1760,7 @@ static int process_cpu_topology(struct feat_fd *ff, void *data __maybe_unused) goto free_cpu; for (i = 0; i < nr; i++) { - str = do_read_string(ff->fd, ph); + str = do_read_string(ff); if (!str) goto error; @@ -1773,14 +1772,14 @@ static int process_cpu_topology(struct feat_fd *ff, void *data __maybe_unused) } ph->env.sibling_cores = strbuf_detach(&sb, NULL); - if (do_read_u32(ff->fd, ph, &nr)) + if (do_read_u32(ff, &nr)) return -1; ph->env.nr_sibling_threads = nr; size += sizeof(u32); for (i = 0; i < nr; i++) { - str = do_read_string(ff->fd, ph); + str = do_read_string(ff); if (!str) goto error; @@ -1802,12 +1801,12 @@ static int process_cpu_topology(struct feat_fd *ff, void *data __maybe_unused) } for (i = 0; i < (u32)cpu_nr; i++) { - if (do_read_u32(ff->fd, ph, &nr)) + if (do_read_u32(ff, &nr)) goto free_cpu; ph->env.cpu[i].core_id = nr; - if (do_read_u32(ff->fd, ph, &nr)) + if (do_read_u32(ff, &nr)) goto free_cpu; if (nr != (u32)-1 && nr > (u32)cpu_nr) { @@ -1835,7 +1834,7 @@ static int process_numa_topology(struct feat_fd *ff, void *data __maybe_unused) char *str; /* nr nodes */ - if (do_read_u32(ff->fd, ff->ph, &nr)) + if (do_read_u32(ff, &nr)) return -1; nodes = zalloc(sizeof(*nodes) * nr); @@ -1846,16 +1845,16 @@ static int process_numa_topology(struct feat_fd *ff, void *data __maybe_unused) n = &nodes[i]; /* node number */ - if (do_read_u32(ff->fd, ff->ph, &n->node)) + if (do_read_u32(ff, &n->node)) goto error; - if (do_read_u64(ff->fd, ff->ph, &n->mem_total)) + if (do_read_u64(ff, &n->mem_total)) goto error; - if (do_read_u64(ff->fd, ff->ph, &n->mem_free)) + if (do_read_u64(ff, &n->mem_free)) goto error; - str = do_read_string(ff->fd, ff->ph); + str = do_read_string(ff); if (!str) goto error; @@ -1881,7 +1880,7 @@ static int process_pmu_mappings(struct feat_fd *ff, void *data __maybe_unused) u32 type; struct strbuf sb; - if (do_read_u32(ff->fd, ff->ph, &pmu_num)) + if (do_read_u32(ff, &pmu_num)) return -1; if (!pmu_num) { @@ -1894,10 +1893,10 @@ static int process_pmu_mappings(struct feat_fd *ff, void *data __maybe_unused) return -1; while (pmu_num) { - if (do_read_u32(ff->fd, ff->ph, &type)) + if (do_read_u32(ff, &type)) goto error; - name = do_read_string(ff->fd, ff->ph); + name = do_read_string(ff); if (!name) goto error; @@ -1933,7 +1932,7 @@ static int process_group_desc(struct feat_fd *ff, void *data __maybe_unused) u32 nr_members; } *desc; - if (do_read_u32(ff->fd, ff->ph, &nr_groups)) + if (do_read_u32(ff, &nr_groups)) return -1; ff->ph->env.nr_groups = nr_groups; @@ -1947,14 +1946,14 @@ static int process_group_desc(struct feat_fd *ff, void *data __maybe_unused) return -1; for (i = 0; i < nr_groups; i++) { - desc[i].name = do_read_string(ff->fd, ff->ph); + desc[i].name = do_read_string(ff); if (!desc[i].name) goto out_free; - if (do_read_u32(ff->fd, ff->ph, &desc[i].leader_idx)) + if (do_read_u32(ff, &desc[i].leader_idx)) goto out_free; - if (do_read_u32(ff->fd, ff->ph, &desc[i].nr_members)) + if (do_read_u32(ff, &desc[i].nr_members)) goto out_free; } @@ -2024,13 +2023,13 @@ static int process_cache(struct feat_fd *ff, void *data __maybe_unused) struct cpu_cache_level *caches; u32 cnt, i, version; - if (do_read_u32(ff->fd, ff->ph, &version)) + if (do_read_u32(ff, &version)) return -1; if (version != 1) return -1; - if (do_read_u32(ff->fd, ff->ph, &cnt)) + if (do_read_u32(ff, &cnt)) return -1; caches = zalloc(sizeof(*caches) * cnt); @@ -2041,7 +2040,7 @@ static int process_cache(struct feat_fd *ff, void *data __maybe_unused) struct cpu_cache_level c; #define _R(v) \ - if (do_read_u32(ff->fd, ff->ph, &c.v))\ + if (do_read_u32(ff, &c.v))\ goto out_free_caches; \ _R(level) @@ -2051,7 +2050,7 @@ static int process_cache(struct feat_fd *ff, void *data __maybe_unused) #undef _R #define _R(v) \ - c.v = do_read_string(ff->fd, ff->ph); \ + c.v = do_read_string(ff); \ if (!c.v) \ goto out_free_caches; -- cgit v1.1 From a02c395cccc95e40b4c506c78857e24fdb049096 Mon Sep 17 00:00:00 2001 From: David Carrillo-Cisneros Date: Mon, 17 Jul 2017 21:25:44 -0700 Subject: perf header: Make write_pmu_mappings pipe-mode friendly In pipe-mode, we will operate over a buffer instead of a file descriptor but write_pmu_mappings uses lseek to move over the perf.data file. Refactor write_pmu_mappings to avoid the usage of lseek and allow reusing the same logic in pipe-mode (next patch). Signed-off-by: David Carrillo-Cisneros Acked-by: David Ahern Acked-by: Jiri Olsa Cc: Alexander Shishkin Cc: Andi Kleen Cc: He Kuang Cc: Masami Hiramatsu Cc: Namhyung Kim Cc: Paul Turner Cc: Peter Zijlstra Cc: Simon Que Cc: Stephane Eranian Cc: Wang Nan Link: http://lkml.kernel.org/r/20170718042549.145161-12-davidcc@google.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/header.c | 21 +++++++++++---------- 1 file changed, 11 insertions(+), 10 deletions(-) (limited to 'tools') diff --git a/tools/perf/util/header.c b/tools/perf/util/header.c index 14db9f2..d5359e33 100644 --- a/tools/perf/util/header.c +++ b/tools/perf/util/header.c @@ -797,11 +797,19 @@ static int write_pmu_mappings(struct feat_fd *ff, struct perf_evlist *evlist __maybe_unused) { struct perf_pmu *pmu = NULL; - off_t offset = lseek(ff->fd, 0, SEEK_CUR); - __u32 pmu_num = 0; + u32 pmu_num = 0; int ret; - /* write real pmu_num later */ + /* + * Do a first pass to count number of pmu to avoid lseek so this + * works in pipe mode as well. + */ + while ((pmu = perf_pmu__scan(pmu))) { + if (!pmu->name) + continue; + pmu_num++; + } + ret = do_write(ff, &pmu_num, sizeof(pmu_num)); if (ret < 0) return ret; @@ -809,7 +817,6 @@ static int write_pmu_mappings(struct feat_fd *ff, while ((pmu = perf_pmu__scan(pmu))) { if (!pmu->name) continue; - pmu_num++; ret = do_write(ff, &pmu->type, sizeof(pmu->type)); if (ret < 0) @@ -820,12 +827,6 @@ static int write_pmu_mappings(struct feat_fd *ff, return ret; } - if (pwrite(ff->fd, &pmu_num, sizeof(pmu_num), offset) != sizeof(pmu_num)) { - /* discard all */ - lseek(ff->fd, offset, SEEK_SET); - return -1; - } - return 0; } -- cgit v1.1 From 0b3d34106c18e5d9ebba004f52a2ce8b264c493e Mon Sep 17 00:00:00 2001 From: David Carrillo-Cisneros Date: Mon, 17 Jul 2017 21:25:45 -0700 Subject: perf header: Add a buffer to struct feat_fd Extend struct feat_fd to use a temporal buffer in pipe-mode, instead of perf.data's file descriptor. The header features build_id and aux_trace already have logic to print in file-mode that heavily rely on lseek the file. For now, leave such features inactive in pipe-mode and print a warning if their functions are called in pipe-mode. Signed-off-by: David Carrillo-Cisneros Acked-by: David Ahern Acked-by: Jiri Olsa Cc: Alexander Shishkin Cc: Andi Kleen Cc: He Kuang Cc: Masami Hiramatsu Cc: Namhyung Kim Cc: Paul Turner Cc: Peter Zijlstra Cc: Simon Que Cc: Stephane Eranian Cc: Wang Nan Link: http://lkml.kernel.org/r/20170718042549.145161-13-davidcc@google.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/header.c | 75 ++++++++++++++++++++++++++++++++++++++++++++---- 1 file changed, 70 insertions(+), 5 deletions(-) (limited to 'tools') diff --git a/tools/perf/util/header.c b/tools/perf/util/header.c index d5359e33..5e6d4d2 100644 --- a/tools/perf/util/header.c +++ b/tools/perf/util/header.c @@ -62,6 +62,7 @@ struct perf_file_attr { struct feat_fd { struct perf_header *ph; int fd; + void *buf; /* Either buf != NULL or fd >= 0 */ ssize_t offset; size_t size; }; @@ -81,19 +82,52 @@ bool perf_header__has_feat(const struct perf_header *header, int feat) return test_bit(feat, header->adds_features); } -/* Return: 0 if succeded, -ERR if failed. */ -int do_write(struct feat_fd *ff, const void *buf, size_t size) +static int __do_write_fd(struct feat_fd *ff, const void *buf, size_t size) { - ssize_t ret; + ssize_t ret = writen(ff->fd, buf, size); - ret = writen(ff->fd, buf, size); if (ret != (ssize_t)size) return ret < 0 ? (int)ret : -1; + return 0; +} + +static int __do_write_buf(struct feat_fd *ff, const void *buf, size_t size) +{ + /* struct perf_event_header::size is u16 */ + const size_t max_size = 0xffff - sizeof(struct perf_event_header); + size_t new_size = ff->size; + void *addr; + + if (size + ff->offset > max_size) + return -E2BIG; + + while (size > (new_size - ff->offset)) + new_size <<= 1; + new_size = min(max_size, new_size); + + if (ff->size < new_size) { + addr = realloc(ff->buf, new_size); + if (!addr) + return -ENOMEM; + ff->buf = addr; + ff->size = new_size; + } + + memcpy(ff->buf + ff->offset, buf, size); + ff->offset += size; return 0; } /* Return: 0 if succeded, -ERR if failed. */ +int do_write(struct feat_fd *ff, const void *buf, size_t size) +{ + if (!ff->buf) + return __do_write_fd(ff, buf, size); + return __do_write_buf(ff, buf, size); +} + +/* Return: 0 if succeded, -ERR if failed. */ int write_padded(struct feat_fd *ff, const void *bf, size_t count, size_t count_aligned) { @@ -126,7 +160,7 @@ static int do_write_string(struct feat_fd *ff, const char *str) return write_padded(ff, str, olen, len); } -static int __do_read(struct feat_fd *ff, void *addr, ssize_t size) +static int __do_read_fd(struct feat_fd *ff, void *addr, ssize_t size) { ssize_t ret = readn(ff->fd, addr, size); @@ -135,6 +169,25 @@ static int __do_read(struct feat_fd *ff, void *addr, ssize_t size) return 0; } +static int __do_read_buf(struct feat_fd *ff, void *addr, ssize_t size) +{ + if (size > (ssize_t)ff->size - ff->offset) + return -1; + + memcpy(addr, ff->buf + ff->offset, size); + ff->offset += size; + + return 0; + +} + +static int __do_read(struct feat_fd *ff, void *addr, ssize_t size) +{ + if (!ff->buf) + return __do_read_fd(ff, addr, size); + return __do_read_buf(ff, addr, size); +} + static int do_read_u32(struct feat_fd *ff, u32 *addr) { int ret; @@ -189,6 +242,9 @@ static char *do_read_string(struct feat_fd *ff) static int write_tracing_data(struct feat_fd *ff, struct perf_evlist *evlist) { + if (WARN(ff->buf, "Error: calling %s in pipe-mode.\n", __func__)) + return -1; + return read_tracing_data(ff->fd, &evlist->entries); } @@ -203,6 +259,9 @@ static int write_build_id(struct feat_fd *ff, if (!perf_session__read_build_ids(session, true)) return -1; + if (WARN(ff->buf, "Error: calling %s in pipe-mode.\n", __func__)) + return -1; + err = perf_session__write_buildid_table(session, ff); if (err < 0) { pr_debug("failed to write buildid table\n"); @@ -912,6 +971,9 @@ static int write_auxtrace(struct feat_fd *ff, struct perf_session *session; int err; + if (WARN(ff->buf, "Error: calling %s in pipe-mode.\n", __func__)) + return -1; + session = container_of(ff->ph, struct perf_session, header); err = auxtrace_index__write(ff->fd, &session->auxtrace_index); @@ -2197,6 +2259,9 @@ static int do_write_feat(struct feat_fd *ff, int type, if (!feat_ops[type].write) return -1; + if (WARN(ff->buf, "Error: calling %s in pipe-mode.\n", __func__)) + return -1; + (*p)->offset = lseek(ff->fd, 0, SEEK_CUR); err = feat_ops[type].write(ff, evlist); -- cgit v1.1 From a4d8c9855a260359655f2a302ee2b231cad379ca Mon Sep 17 00:00:00 2001 From: David Carrillo-Cisneros Date: Mon, 17 Jul 2017 21:25:46 -0700 Subject: perf header: Change FEAT_OP* macros There are three FEAT_OP* macros: - FEAT_OPA: for features without process record. - FEAT_OPP: for features with process record. - FEAT_OPF: like FEAT_OPP but to show only if show_full_info flags is set. To add pipe-mode headers we need yet another variation of the macros (one to specify whether a feature generates an auxiliar record). Instead, we redefine macros so that: - show_full_info is specified as an argument (to remove the FEAT_OPF variation) and, - it always sets "process" handler (to remove the FEAT_OPA variation). Individual process handlers can be NULLed individually. This allows to define two variations only: - FEAT_OPR: synthesizes auxiliar event record. - FEAT_OPN: doesn't synthesize an auxiliar event record. Signed-off-by: David Carrillo-Cisneros Acked-by: David Ahern Acked-by: Jiri Olsa Cc: Alexander Shishkin Cc: Andi Kleen Cc: He Kuang Cc: Masami Hiramatsu Cc: Namhyung Kim Cc: Paul Turner Cc: Peter Zijlstra Cc: Simon Que Cc: Stephane Eranian Cc: Wang Nan Link: http://lkml.kernel.org/r/20170718042549.145161-14-davidcc@google.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/header.c | 72 +++++++++++++++++++++++++++++------------------- 1 file changed, 44 insertions(+), 28 deletions(-) (limited to 'tools') diff --git a/tools/perf/util/header.c b/tools/perf/util/header.c index 5e6d4d2..0fdbf75 100644 --- a/tools/perf/util/header.c +++ b/tools/perf/util/header.c @@ -12,6 +12,7 @@ #include #include #include +#include #include #include #include @@ -2139,42 +2140,57 @@ struct feature_ops { int (*process)(struct feat_fd *ff, void *data); const char *name; bool full_only; + bool synthesize; }; -#define FEAT_OPA(n, func) \ - [n] = { .name = #n, .write = write_##func, .print = print_##func } -#define FEAT_OPP(n, func) \ - [n] = { .name = #n, .write = write_##func, .print = print_##func, \ - .process = process_##func } -#define FEAT_OPF(n, func) \ - [n] = { .name = #n, .write = write_##func, .print = print_##func, \ - .process = process_##func, .full_only = true } +#define FEAT_OPR(n, func, __full_only) \ + [HEADER_##n] = { \ + .name = __stringify(n), \ + .write = write_##func, \ + .print = print_##func, \ + .full_only = __full_only, \ + .process = process_##func, \ + .synthesize = true \ + } + +#define FEAT_OPN(n, func, __full_only) \ + [HEADER_##n] = { \ + .name = __stringify(n), \ + .write = write_##func, \ + .print = print_##func, \ + .full_only = __full_only, \ + .process = process_##func \ + } /* feature_ops not implemented: */ #define print_tracing_data NULL #define print_build_id NULL +#define process_branch_stack NULL +#define process_stat NULL + + static const struct feature_ops feat_ops[HEADER_LAST_FEATURE] = { - FEAT_OPP(HEADER_TRACING_DATA, tracing_data), - FEAT_OPP(HEADER_BUILD_ID, build_id), - FEAT_OPP(HEADER_HOSTNAME, hostname), - FEAT_OPP(HEADER_OSRELEASE, osrelease), - FEAT_OPP(HEADER_VERSION, version), - FEAT_OPP(HEADER_ARCH, arch), - FEAT_OPP(HEADER_NRCPUS, nrcpus), - FEAT_OPP(HEADER_CPUDESC, cpudesc), - FEAT_OPP(HEADER_CPUID, cpuid), - FEAT_OPP(HEADER_TOTAL_MEM, total_mem), - FEAT_OPP(HEADER_EVENT_DESC, event_desc), - FEAT_OPP(HEADER_CMDLINE, cmdline), - FEAT_OPF(HEADER_CPU_TOPOLOGY, cpu_topology), - FEAT_OPF(HEADER_NUMA_TOPOLOGY, numa_topology), - FEAT_OPA(HEADER_BRANCH_STACK, branch_stack), - FEAT_OPP(HEADER_PMU_MAPPINGS, pmu_mappings), - FEAT_OPP(HEADER_GROUP_DESC, group_desc), - FEAT_OPP(HEADER_AUXTRACE, auxtrace), - FEAT_OPA(HEADER_STAT, stat), - FEAT_OPF(HEADER_CACHE, cache), + FEAT_OPN(TRACING_DATA, tracing_data, false), + FEAT_OPN(BUILD_ID, build_id, false), + FEAT_OPR(HOSTNAME, hostname, false), + FEAT_OPR(OSRELEASE, osrelease, false), + FEAT_OPR(VERSION, version, false), + FEAT_OPR(ARCH, arch, false), + FEAT_OPR(NRCPUS, nrcpus, false), + FEAT_OPR(CPUDESC, cpudesc, false), + FEAT_OPR(CPUID, cpuid, false), + FEAT_OPR(TOTAL_MEM, total_mem, false), + FEAT_OPR(EVENT_DESC, event_desc, false), + FEAT_OPR(CMDLINE, cmdline, false), + FEAT_OPR(CPU_TOPOLOGY, cpu_topology, true), + FEAT_OPR(NUMA_TOPOLOGY, numa_topology, true), + FEAT_OPN(BRANCH_STACK, branch_stack, false), + FEAT_OPR(PMU_MAPPINGS, pmu_mappings, false), + FEAT_OPN(GROUP_DESC, group_desc, false), + FEAT_OPN(AUXTRACE, auxtrace, false), + FEAT_OPN(STAT, stat, false), + FEAT_OPN(CACHE, cache, true), }; struct header_print_data { -- cgit v1.1 From 114f709e01e62760a6d03de1358188293dfefdda Mon Sep 17 00:00:00 2001 From: David Carrillo-Cisneros Date: Mon, 17 Jul 2017 21:25:47 -0700 Subject: perf tool: Add show_feature_header to perf_tool Add show_feat_hdr to control level of printed information of feature headers. Signed-off-by: David Carrillo-Cisneros Acked-by: David Ahern Acked-by: Jiri Olsa Cc: Alexander Shishkin Cc: Andi Kleen Cc: He Kuang Cc: Masami Hiramatsu Cc: Namhyung Kim Cc: Paul Turner Cc: Peter Zijlstra Cc: Simon Que Cc: Stephane Eranian Cc: Wang Nan Link: http://lkml.kernel.org/r/20170718042549.145161-15-davidcc@google.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-report.c | 4 ++++ tools/perf/builtin-script.c | 3 +++ tools/perf/util/tool.h | 7 +++++++ 3 files changed, 14 insertions(+) (limited to 'tools') diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c index 79a33eb..40c3a92 100644 --- a/tools/perf/builtin-report.c +++ b/tools/perf/builtin-report.c @@ -988,6 +988,10 @@ repeat: /* Force tty output for header output and per-thread stat. */ if (report.header || report.header_only || report.show_threads) use_browser = 0; + if (report.header || report.header_only) + report.tool.show_feat_hdr = SHOW_FEAT_HEADER; + if (report.show_full_info) + report.tool.show_feat_hdr = SHOW_FEAT_HEADER_FULL_INFO; if (strcmp(input_name, "-") != 0) setup_browser(true); diff --git a/tools/perf/builtin-script.c b/tools/perf/builtin-script.c index 83cdc0a..6e44552 100644 --- a/tools/perf/builtin-script.c +++ b/tools/perf/builtin-script.c @@ -2972,10 +2972,13 @@ int cmd_script(int argc, const char **argv) return -1; if (header || header_only) { + script.tool.show_feat_hdr = SHOW_FEAT_HEADER; perf_session__fprintf_info(session, stdout, show_full_info); if (header_only) goto out_delete; } + if (show_full_info) + script.tool.show_feat_hdr = SHOW_FEAT_HEADER_FULL_INFO; if (symbol__init(&session->header.env) < 0) goto out_delete; diff --git a/tools/perf/util/tool.h b/tools/perf/util/tool.h index 829471a..baeca80 100644 --- a/tools/perf/util/tool.h +++ b/tools/perf/util/tool.h @@ -34,6 +34,12 @@ typedef int (*event_oe)(struct perf_tool *tool, union perf_event *event, typedef s64 (*event_op3)(struct perf_tool *tool, union perf_event *event, struct perf_session *session); +enum show_feature_header { + SHOW_FEAT_NO_HEADER = 0, + SHOW_FEAT_HEADER, + SHOW_FEAT_HEADER_FULL_INFO, +}; + struct perf_tool { event_sample sample, read; @@ -68,6 +74,7 @@ struct perf_tool { bool ordered_events; bool ordering_requires_timestamps; bool namespace_events; + enum show_feature_header show_feat_hdr; }; #endif /* __PERF_TOOL_H */ -- cgit v1.1 From e9def1b2e74e3d2134133f70d2a84c242446bbe7 Mon Sep 17 00:00:00 2001 From: David Carrillo-Cisneros Date: Mon, 17 Jul 2017 21:25:48 -0700 Subject: perf tools: Add feature header record to pipe-mode Add header record types to pipe-mode, reusing the functions used in file-mode and leveraging the new struct feat_fd. For alignment, check that synthesized events don't exceed pagesize. Add the perf_event__synthesize_feature event call back to process the new header records. Before this patch: $ perf record -o - -e cycles sleep 1 | perf report --stdio --header [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.000 MB - ] ... After this patch: $ perf record -o - -e cycles sleep 1 | perf report --stdio --header # ======== # captured on: Mon May 22 16:33:43 2017 # ======== # # hostname : my_hostname # os release : 4.11.0-dbx-up_perf # perf version : 4.11.rc6.g6277c80 # arch : x86_64 # nrcpus online : 72 # nrcpus avail : 72 # cpudesc : Intel(R) Xeon(R) CPU E5-2696 v3 @ 2.30GHz # cpuid : GenuineIntel,6,63,2 # total memory : 263457192 kB # cmdline : /root/perf record -o - -e cycles -c 100000 sleep 1 # HEADER_CPU_TOPOLOGY info available, use -I to display # HEADER_NUMA_TOPOLOGY info available, use -I to display # pmu mappings: intel_bts = 6, uncore_imc_4 = 22, uncore_sbox_1 = 47, uncore_cbox_5 = 33, uncore_ha_0 = 16, uncore_cbox [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.000 MB - ] ... Support added for the subcommands: report, inject, annotate and script. Signed-off-by: David Carrillo-Cisneros Acked-by: David Ahern Acked-by: Jiri Olsa Cc: Alexander Shishkin Cc: Andi Kleen Cc: He Kuang Cc: Masami Hiramatsu Cc: Namhyung Kim Cc: Paul Turner Cc: Peter Zijlstra Cc: Simon Que Cc: Stephane Eranian Cc: Wang Nan Link: http://lkml.kernel.org/r/20170718042549.145161-16-davidcc@google.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/Documentation/perf.data-file-format.txt | 10 ++- tools/perf/builtin-annotate.c | 1 + tools/perf/builtin-inject.c | 1 + tools/perf/builtin-record.c | 7 ++ tools/perf/builtin-report.c | 1 + tools/perf/builtin-script.c | 1 + tools/perf/util/event.c | 1 + tools/perf/util/event.h | 8 ++ tools/perf/util/header.c | 98 ++++++++++++++++++++++ tools/perf/util/header.h | 9 ++ tools/perf/util/session.c | 4 + tools/perf/util/tool.h | 3 +- 12 files changed, 141 insertions(+), 3 deletions(-) (limited to 'tools') diff --git a/tools/perf/Documentation/perf.data-file-format.txt b/tools/perf/Documentation/perf.data-file-format.txt index de8b39d..e90c59c 100644 --- a/tools/perf/Documentation/perf.data-file-format.txt +++ b/tools/perf/Documentation/perf.data-file-format.txt @@ -398,6 +398,11 @@ struct auxtrace_error_event { char msg[MAX_AUXTRACE_ERROR_MSG]; }; + PERF_RECORD_HEADER_FEATURE = 80, + +Describes a header feature. These are records used in pipe-mode that +contain information that otherwise would be in perf.data file's header. + Event types Define the event attributes with their IDs. @@ -422,8 +427,9 @@ struct perf_pipe_file_header { }; The information about attrs, data, and event_types is instead in the -synthesized events PERF_RECORD_ATTR, PERF_RECORD_HEADER_TRACING_DATA and -PERF_RECORD_HEADER_EVENT_TYPE that are generated by perf record in pipe-mode. +synthesized events PERF_RECORD_ATTR, PERF_RECORD_HEADER_TRACING_DATA, +PERF_RECORD_HEADER_EVENT_TYPE, and PERF_RECORD_HEADER_FEATURE +that are generated by perf record in pipe-mode. References: diff --git a/tools/perf/builtin-annotate.c b/tools/perf/builtin-annotate.c index 7a5dc7e..5205408 100644 --- a/tools/perf/builtin-annotate.c +++ b/tools/perf/builtin-annotate.c @@ -397,6 +397,7 @@ int cmd_annotate(int argc, const char **argv) .namespaces = perf_event__process_namespaces, .attr = perf_event__process_attr, .build_id = perf_event__process_build_id, + .feature = perf_event__process_feature, .ordered_events = true, .ordering_requires_timestamps = true, }, diff --git a/tools/perf/builtin-inject.c b/tools/perf/builtin-inject.c index ea8db38..2b80329 100644 --- a/tools/perf/builtin-inject.c +++ b/tools/perf/builtin-inject.c @@ -770,6 +770,7 @@ int cmd_inject(int argc, const char **argv) .finished_round = perf_event__repipe_oe_synth, .build_id = perf_event__repipe_op2_synth, .id_index = perf_event__repipe_op2_synth, + .feature = perf_event__repipe_op2_synth, }, .input_name = "-", .samples = LIST_HEAD_INIT(inject.samples), diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c index 64eef9a5..36d7117 100644 --- a/tools/perf/builtin-record.c +++ b/tools/perf/builtin-record.c @@ -799,6 +799,13 @@ static int record__synthesize(struct record *rec, bool tail) return 0; if (file->is_pipe) { + err = perf_event__synthesize_features( + tool, session, rec->evlist, process_synthesized_event); + if (err < 0) { + pr_err("Couldn't synthesize features.\n"); + return err; + } + err = perf_event__synthesize_attrs(tool, session, process_synthesized_event); if (err < 0) { diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c index 40c3a92..8e752ba 100644 --- a/tools/perf/builtin-report.c +++ b/tools/perf/builtin-report.c @@ -718,6 +718,7 @@ int cmd_report(int argc, const char **argv) .id_index = perf_event__process_id_index, .auxtrace_info = perf_event__process_auxtrace_info, .auxtrace = perf_event__process_auxtrace, + .feature = perf_event__process_feature, .ordered_events = true, .ordering_requires_timestamps = true, }, diff --git a/tools/perf/builtin-script.c b/tools/perf/builtin-script.c index 6e44552..d430ff4 100644 --- a/tools/perf/builtin-script.c +++ b/tools/perf/builtin-script.c @@ -2682,6 +2682,7 @@ int cmd_script(int argc, const char **argv) .attr = process_attr, .event_update = perf_event__process_event_update, .tracing_data = perf_event__process_tracing_data, + .feature = perf_event__process_feature, .build_id = perf_event__process_build_id, .id_index = perf_event__process_id_index, .auxtrace_info = perf_event__process_auxtrace_info, diff --git a/tools/perf/util/event.c b/tools/perf/util/event.c index dc5c3bb..1c905ba 100644 --- a/tools/perf/util/event.c +++ b/tools/perf/util/event.c @@ -57,6 +57,7 @@ static const char *perf_event__names[] = { [PERF_RECORD_STAT_ROUND] = "STAT_ROUND", [PERF_RECORD_EVENT_UPDATE] = "EVENT_UPDATE", [PERF_RECORD_TIME_CONV] = "TIME_CONV", + [PERF_RECORD_HEADER_FEATURE] = "FEATURE", }; static const char *perf_ns__names[] = { diff --git a/tools/perf/util/event.h b/tools/perf/util/event.h index 9967c87..37c5faf 100644 --- a/tools/perf/util/event.h +++ b/tools/perf/util/event.h @@ -244,6 +244,7 @@ enum perf_user_event_type { /* above any possible kernel type */ PERF_RECORD_STAT_ROUND = 77, PERF_RECORD_EVENT_UPDATE = 78, PERF_RECORD_TIME_CONV = 79, + PERF_RECORD_HEADER_FEATURE = 80, PERF_RECORD_HEADER_MAX }; @@ -609,6 +610,12 @@ struct time_conv_event { u64 time_zero; }; +struct feature_event { + struct perf_event_header header; + u64 feat_id; + char data[]; +}; + union perf_event { struct perf_event_header header; struct mmap_event mmap; @@ -639,6 +646,7 @@ union perf_event { struct stat_event stat; struct stat_round_event stat_round; struct time_conv_event time_conv; + struct feature_event feat; }; void perf_event__print_totals(void); diff --git a/tools/perf/util/header.c b/tools/perf/util/header.c index 0fdbf75..2e6036d 100644 --- a/tools/perf/util/header.c +++ b/tools/perf/util/header.c @@ -35,6 +35,7 @@ #include "data.h" #include #include "asm/bug.h" +#include "tool.h" #include "sane_ctype.h" @@ -2982,6 +2983,103 @@ int perf_event__synthesize_attr(struct perf_tool *tool, return err; } +int perf_event__synthesize_features(struct perf_tool *tool, + struct perf_session *session, + struct perf_evlist *evlist, + perf_event__handler_t process) +{ + struct perf_header *header = &session->header; + struct feat_fd ff; + struct feature_event *fe; + size_t sz, sz_hdr; + int feat, ret; + + sz_hdr = sizeof(fe->header); + sz = sizeof(union perf_event); + /* get a nice alignment */ + sz = PERF_ALIGN(sz, page_size); + + memset(&ff, 0, sizeof(ff)); + + ff.buf = malloc(sz); + if (!ff.buf) + return -ENOMEM; + + ff.size = sz - sz_hdr; + + for_each_set_bit(feat, header->adds_features, HEADER_FEAT_BITS) { + if (!feat_ops[feat].synthesize) { + pr_debug("No record header feature for header :%d\n", feat); + continue; + } + + ff.offset = sizeof(*fe); + + ret = feat_ops[feat].write(&ff, evlist); + if (ret || ff.offset <= (ssize_t)sizeof(*fe)) { + pr_debug("Error writing feature\n"); + continue; + } + /* ff.buf may have changed due to realloc in do_write() */ + fe = ff.buf; + memset(fe, 0, sizeof(*fe)); + + fe->feat_id = feat; + fe->header.type = PERF_RECORD_HEADER_FEATURE; + fe->header.size = ff.offset; + + ret = process(tool, ff.buf, NULL, NULL); + if (ret) { + free(ff.buf); + return ret; + } + } + free(ff.buf); + return 0; +} + +int perf_event__process_feature(struct perf_tool *tool, + union perf_event *event, + struct perf_session *session __maybe_unused) +{ + struct feat_fd ff = { .fd = 0 }; + struct feature_event *fe = (struct feature_event *)event; + int type = fe->header.type; + u64 feat = fe->feat_id; + + if (type < 0 || type >= PERF_RECORD_HEADER_MAX) { + pr_warning("invalid record type %d in pipe-mode\n", type); + return 0; + } + if (feat == HEADER_RESERVED || feat > HEADER_LAST_FEATURE) { + pr_warning("invalid record type %d in pipe-mode\n", type); + return -1; + } + + if (!feat_ops[feat].process) + return 0; + + ff.buf = (void *)fe->data; + ff.size = event->header.size - sizeof(event->header); + ff.ph = &session->header; + + if (feat_ops[feat].process(&ff, NULL)) + return -1; + + if (!feat_ops[feat].print || !tool->show_feat_hdr) + return 0; + + if (!feat_ops[feat].full_only || + tool->show_feat_hdr >= SHOW_FEAT_HEADER_FULL_INFO) { + feat_ops[feat].print(&ff, stdout); + } else { + fprintf(stdout, "# %s info available, use -I to display\n", + feat_ops[feat].name); + } + + return 0; +} + static struct event_update_event * event_update_event__new(size_t size, u64 type, u64 id) { diff --git a/tools/perf/util/header.h b/tools/perf/util/header.h index 9d8dcd5..f7a16ee 100644 --- a/tools/perf/util/header.h +++ b/tools/perf/util/header.h @@ -101,6 +101,15 @@ int perf_header__process_sections(struct perf_header *header, int fd, int perf_header__fprintf_info(struct perf_session *s, FILE *fp, bool full); +int perf_event__synthesize_features(struct perf_tool *tool, + struct perf_session *session, + struct perf_evlist *evlist, + perf_event__handler_t process); + +int perf_event__process_feature(struct perf_tool *tool, + union perf_event *event, + struct perf_session *session); + int perf_event__synthesize_attr(struct perf_tool *tool, struct perf_event_attr *attr, u32 ids, u64 *id, perf_event__handler_t process); diff --git a/tools/perf/util/session.c b/tools/perf/util/session.c index d19c40a..dc453f8 100644 --- a/tools/perf/util/session.c +++ b/tools/perf/util/session.c @@ -428,6 +428,8 @@ void perf_tool__fill_defaults(struct perf_tool *tool) tool->stat_round = process_stat_round_stub; if (tool->time_conv == NULL) tool->time_conv = process_event_op2_stub; + if (tool->feature == NULL) + tool->feature = process_event_op2_stub; } static void swap_sample_id_all(union perf_event *event, void *data) @@ -1371,6 +1373,8 @@ static s64 perf_session__process_user_event(struct perf_session *session, case PERF_RECORD_TIME_CONV: session->time_conv = event->time_conv; return tool->time_conv(tool, event, session); + case PERF_RECORD_HEADER_FEATURE: + return tool->feature(tool, event, session); default: return -EINVAL; } diff --git a/tools/perf/util/tool.h b/tools/perf/util/tool.h index baeca80..d549e50 100644 --- a/tools/perf/util/tool.h +++ b/tools/perf/util/tool.h @@ -69,7 +69,8 @@ struct perf_tool { cpu_map, stat_config, stat, - stat_round; + stat_round, + feature; event_op3 auxtrace; bool ordered_events; bool ordering_requires_timestamps; -- cgit v1.1 From f9ebdccf2b78e643d1ba2c979fa293c9d1e8ba86 Mon Sep 17 00:00:00 2001 From: David Carrillo-Cisneros Date: Mon, 17 Jul 2017 21:25:49 -0700 Subject: perf header: Add event desc to pipe-mode header Add event descriptor to perf header output in pipe-mode. After this patch: $ perf record -e cycles sleep 1 | perf report --header # ======== # captured on: Mon Jun 5 22:52:13 2017 # ======== # # hostname : lphh20 # os release : 4.3.5-smp-801.43.0.0 # perf version : 4.12.rc2.g439987 # arch : x86_64 # nrcpus online : 72 # nrcpus avail : 72 # cpudesc : Intel(R) Xeon(R) CPU E5-2696 v3 @ 2.30GHz # cpuid : GenuineIntel,6,63,2 # total memory : 264134144 kB # cmdline : /root/perf record -e cycles sleep 1 # event : name = cycles, , size = 112, { sample_period, sample_freq } = 4000, sample_type = IP|TID|TIME|PERIOD, disabled = 1, inherit = 1, mmap = 1, comm = 1, freq = 1, enable_on_exec = 1, task = 1, sample_id_all = 1, exclude_guest = 1, mmap2 = 1, comm_exec = 1 # CPU_TOPOLOGY info available, use -I to display # NUMA_TOPOLOGY info available, use -I to display # pmu mappings: intel_bts = 6, cpu = 4, msr = 49, uncore_cbox_10 = 36, uncore_cbox_11 = 37, uncore_cbox_12 = 38, uncore_cbox_13 = 39, uncore_cbox_14 = 40, uncore_cbox_15 = 41, uncore_cbox_16 = 42, uncore_cbox_17 = 43, software = 1, power = 7, uncore_irp = 24, uncore_pcu = 48, tracepoint = 2, uncore_imc_0 = 16, uncore_imc_1 = 17, uncore_imc_2 = 18, uncore_imc_3 = 19, uncore_imc_4 = 20, uncore_imc_5 = 21, uncore_imc_6 = 22, uncore_imc_7 = 23, uncore_qpi_0 = 8, uncore_qpi_1 = 9, uncore_cbox_0 = 26, uncore_cbox_1 = 27, uncore_cbox_2 = 28, uncore_cbox_3 = 29, uncore_cbox_4 = 30, uncore_cbox_5 = 31, uncore_cbox_6 = 32, uncore_cbox_7 = 33, uncore_cbox_8 = 34, uncore_cbox_9 = 35, uncore_r2pcie = 13, uncore_r3qpi_0 = 10, uncore_r3qpi_1 = 11, uncore_r3qpi_2 = 12, uncore_sbox_0 = 44, uncore_sbox_1 = 45, uncore_sbox_2 = 46, uncore_sbox_3 = 47, breakpoint = 5, uncore_ha_0 = 14, uncore_ha_1 = 15, uncore_ubox = 25 [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.000 MB (null) ] Prior to this patch, event was not printed. Signed-off-by: David Carrillo-Cisneros Acked-by: David Ahern Acked-by: Jiri Olsa Cc: Alexander Shishkin Cc: Andi Kleen Cc: He Kuang Cc: Masami Hiramatsu Cc: Namhyung Kim Cc: Paul Turner Cc: Peter Zijlstra Cc: Simon Que Cc: Stephane Eranian Cc: Wang Nan Link: http://lkml.kernel.org/r/20170718042549.145161-17-davidcc@google.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/header.c | 19 +++++++++++++++++-- 1 file changed, 17 insertions(+), 2 deletions(-) (limited to 'tools') diff --git a/tools/perf/util/header.c b/tools/perf/util/header.c index 2e6036d..28bf444 100644 --- a/tools/perf/util/header.c +++ b/tools/perf/util/header.c @@ -67,6 +67,7 @@ struct feat_fd { void *buf; /* Either buf != NULL or fd >= 0 */ ssize_t offset; size_t size; + struct perf_evsel *events; }; void perf_header__set_feat(struct perf_header *header, int feat) @@ -1359,10 +1360,15 @@ static int __desc_attr__fprintf(FILE *fp, const char *name, const char *val, static void print_event_desc(struct feat_fd *ff, FILE *fp) { - struct perf_evsel *evsel, *events = read_event_desc(ff); + struct perf_evsel *evsel, *events; u32 j; u64 *id; + if (ff->events) + events = ff->events; + else + events = read_event_desc(ff); + if (!events) { fprintf(fp, "# event desc: not available or unable to read\n"); return; @@ -1387,6 +1393,7 @@ static void print_event_desc(struct feat_fd *ff, FILE *fp) } free_event_desc(events); + ff->events = NULL; } static void print_total_mem(struct feat_fd *ff, FILE *fp) @@ -1757,10 +1764,18 @@ process_event_desc(struct feat_fd *ff, void *data __maybe_unused) return 0; session = container_of(ff->ph, struct perf_session, header); + + if (session->file->is_pipe) { + /* Save events for reading later by print_event_desc, + * since they can't be read again in pipe mode. */ + ff->events = events; + } + for (evsel = events; evsel->attr.size; evsel++) perf_evlist__set_event_name(session->evlist, evsel); - free_event_desc(events); + if (!session->file->is_pipe) + free_event_desc(events); return 0; } -- cgit v1.1 From eb0baf8a0d9259d168523b8e7c436b55ade7c546 Mon Sep 17 00:00:00 2001 From: Jin Yao Date: Tue, 18 Jul 2017 20:13:09 +0800 Subject: perf/core: Define the common branch type classification MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit It is often useful to know the branch types while analyzing branch data. For example, a call is very different from a conditional branch. Currently we have to look it up in binary while the binary may later not be available and even the binary is available but user has to take some time. It is very useful for user to check it directly in perf report. Perf already has support for disassembling the branch instruction to get the x86 branch type. To keep consistent on kernel and userspace and make the classification more common, the patch adds the common branch type classification in perf_event.h. The patch only defines a minimum but most common set of branch types. PERF_BR_UNKNOWN : unknown PERF_BR_COND ：conditional PERF_BR_UNCOND : unconditional PERF_BR_IND : indirect PERF_BR_CALL : function call PERF_BR_IND_CALL : indirect function call PERF_BR_RET : function return PERF_BR_SYSCALL : syscall PERF_BR_SYSRET : syscall return PERF_BR_COND_CALL : conditional function call PERF_BR_COND_RET : conditional function return The patch also adds a new field type (4 bits) in perf_branch_entry to record the branch type. Since the disassembling of branch instruction needs some overhead, a new PERF_SAMPLE_BRANCH_TYPE_SAVE is introduced to indicate if it needs to disassemble the branch instruction and record the branch type. Change log: v10: Not changed. v9: Not changed. v8: Change PERF_BR_NONE to PERF_BR_UNKNOWN. No other change. v7: Just keep the most common branch types. Others are removed. v6: Not changed. v5: Not changed. The v5 patch series just change the userspace. v4: Comparing to previous version, the major changes are: 1. Remove the PERF_BR_JCC_FWD/PERF_BR_JCC_BWD, they will be computed later in userspace. 2. Remove the "cross" field in perf_branch_entry. The cross page computing will be done later in userspace. Signed-off-by: Yao Jin Acked-by: Jiri Olsa Acked-by: Michael Ellerman Acked-by: Peter Zijlstra Cc: Alexander Shishkin Cc: Andi Kleen Cc: Kan Liang Link: http://lkml.kernel.org/r/1500379995-6449-2-git-send-email-yao.jin@linux.intel.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/include/uapi/linux/perf_event.h | 27 ++++++++++++++++++++++++++- 1 file changed, 26 insertions(+), 1 deletion(-) (limited to 'tools') diff --git a/tools/include/uapi/linux/perf_event.h b/tools/include/uapi/linux/perf_event.h index b1c0b18..642db5f 100644 --- a/tools/include/uapi/linux/perf_event.h +++ b/tools/include/uapi/linux/perf_event.h @@ -174,6 +174,8 @@ enum perf_branch_sample_type_shift { PERF_SAMPLE_BRANCH_NO_FLAGS_SHIFT = 14, /* no flags */ PERF_SAMPLE_BRANCH_NO_CYCLES_SHIFT = 15, /* no cycles */ + PERF_SAMPLE_BRANCH_TYPE_SAVE_SHIFT = 16, /* save branch type */ + PERF_SAMPLE_BRANCH_MAX_SHIFT /* non-ABI */ }; @@ -198,9 +200,30 @@ enum perf_branch_sample_type { PERF_SAMPLE_BRANCH_NO_FLAGS = 1U << PERF_SAMPLE_BRANCH_NO_FLAGS_SHIFT, PERF_SAMPLE_BRANCH_NO_CYCLES = 1U << PERF_SAMPLE_BRANCH_NO_CYCLES_SHIFT, + PERF_SAMPLE_BRANCH_TYPE_SAVE = + 1U << PERF_SAMPLE_BRANCH_TYPE_SAVE_SHIFT, + PERF_SAMPLE_BRANCH_MAX = 1U << PERF_SAMPLE_BRANCH_MAX_SHIFT, }; +/* + * Common flow change classification + */ +enum { + PERF_BR_UNKNOWN = 0, /* unknown */ + PERF_BR_COND = 1, /* conditional */ + PERF_BR_UNCOND = 2, /* unconditional */ + PERF_BR_IND = 3, /* indirect */ + PERF_BR_CALL = 4, /* function call */ + PERF_BR_IND_CALL = 5, /* indirect function call */ + PERF_BR_RET = 6, /* function return */ + PERF_BR_SYSCALL = 7, /* syscall */ + PERF_BR_SYSRET = 8, /* syscall return */ + PERF_BR_COND_CALL = 9, /* conditional function call */ + PERF_BR_COND_RET = 10, /* conditional function return */ + PERF_BR_MAX, +}; + #define PERF_SAMPLE_BRANCH_PLM_ALL \ (PERF_SAMPLE_BRANCH_USER|\ PERF_SAMPLE_BRANCH_KERNEL|\ @@ -1015,6 +1038,7 @@ union perf_mem_data_src { * in_tx: running in a hardware transaction * abort: aborting a hardware transaction * cycles: cycles from last branch (or 0 if not supported) + * type: branch type */ struct perf_branch_entry { __u64 from; @@ -1024,7 +1048,8 @@ struct perf_branch_entry { in_tx:1, /* in transaction */ abort:1, /* transaction abort */ cycles:16, /* cycle count to last branch */ - reserved:44; + type:4, /* branch type */ + reserved:40; }; #endif /* _UAPI_LINUX_PERF_EVENT_H */ -- cgit v1.1 From 60f83fa6341dab4aec01cee354ea902771473adb Mon Sep 17 00:00:00 2001 From: Jin Yao Date: Tue, 18 Jul 2017 20:13:11 +0800 Subject: perf record: Create a new option save_type in --branch-filter The option indicates the kernel to save branch type during sampling. One example: perf record -g --branch-filter any,save_type Signed-off-by: Yao Jin Acked-by: Jiri Olsa Cc: Alexander Shishkin Cc: Andi Kleen Cc: Kan Liang Cc: Michael Ellerman Cc: Peter Zijlstra Link: http://lkml.kernel.org/r/1500379995-6449-4-git-send-email-yao.jin@linux.intel.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/Documentation/perf-record.txt | 1 + tools/perf/util/parse-branch-options.c | 1 + 2 files changed, 2 insertions(+) (limited to 'tools') diff --git a/tools/perf/Documentation/perf-record.txt b/tools/perf/Documentation/perf-record.txt index b0e9e92..9bdea04 100644 --- a/tools/perf/Documentation/perf-record.txt +++ b/tools/perf/Documentation/perf-record.txt @@ -332,6 +332,7 @@ following filters are defined: - no_tx: only when the target is not in a hardware transaction - abort_tx: only when the target is a hardware transaction abort - cond: conditional branches + - save_type: save branch type during sampling in case binary is not available later + The option requires at least one branch type among any, any_call, any_ret, ind_call, cond. diff --git a/tools/perf/util/parse-branch-options.c b/tools/perf/util/parse-branch-options.c index 38fd115..e71fb5f 100644 --- a/tools/perf/util/parse-branch-options.c +++ b/tools/perf/util/parse-branch-options.c @@ -28,6 +28,7 @@ static const struct branch_mode branch_modes[] = { BRANCH_OPT("cond", PERF_SAMPLE_BRANCH_COND), BRANCH_OPT("ind_jmp", PERF_SAMPLE_BRANCH_IND_JUMP), BRANCH_OPT("call", PERF_SAMPLE_BRANCH_CALL), + BRANCH_OPT("save_type", PERF_SAMPLE_BRANCH_TYPE_SAVE), BRANCH_END }; -- cgit v1.1 From 8d51735fcd2be1791c0bfe2d581d9063281fe7fb Mon Sep 17 00:00:00 2001 From: Jin Yao Date: Tue, 18 Jul 2017 20:13:12 +0800 Subject: perf report: Refactor the branch info printing code The branch info such as predicted/cycles/... are printed at the callchain entries. For example: perf report --branch-history --no-children --stdio --1.07%--main div.c:39 (predicted:52.4% cycles:1 iterations:17) main div.c:44 (predicted:52.4% cycles:1) main div.c:42 (cycles:2) compute_flag div.c:28 (cycles:2) compute_flag div.c:27 (cycles:1) rand rand.c:28 (cycles:1) rand rand.c:28 (cycles:1) __random random.c:298 (cycles:1) __random random.c:297 (cycles:1) __random random.c:295 (cycles:1) __random random.c:295 (cycles:1) __random random.c:295 (cycles:1) But the current code is difficult to maintain and extend. This patch refactors the code for easy maintenance. Change log: v6: 1. Put the multiline condition code into {} brackets in counts_str_build() 2. Keep the original display order, that is: predicted, abort, cycles, iterations v5: It's a new patch in v5 patch series. Signed-off-by: Yao Jin Acked-by: Jiri Olsa Cc: Alexander Shishkin Cc: Andi Kleen Cc: Kan Liang Cc: Michael Ellerman Cc: Peter Zijlstra Link: http://lkml.kernel.org/r/1500379995-6449-5-git-send-email-yao.jin@linux.intel.com [ Don't use 'index' as a name for a variable, it shadows a globa decl in older distros ] Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/callchain.c | 100 ++++++++++++++++++-------------------------- 1 file changed, 41 insertions(+), 59 deletions(-) (limited to 'tools') diff --git a/tools/perf/util/callchain.c b/tools/perf/util/callchain.c index b4204b4..917f4d6 100644 --- a/tools/perf/util/callchain.c +++ b/tools/perf/util/callchain.c @@ -1214,83 +1214,65 @@ int callchain_branch_counts(struct callchain_root *root, cycles_count); } +static int count_pri64_printf(int idx, const char *str, u64 value, char *bf, int bfsize) +{ + int printed; + + printed = scnprintf(bf, bfsize, "%s%s:%" PRId64 "", (idx) ? " " : " (", str, value); + + return printed; +} + +static int count_float_printf(int idx, const char *str, float value, char *bf, int bfsize) +{ + int printed; + + printed = scnprintf(bf, bfsize, "%s%s:%.1f%%", (idx) ? " " : " (", str, value); + + return printed; +} + static int counts_str_build(char *bf, int bfsize, u64 branch_count, u64 predicted_count, u64 abort_count, u64 cycles_count, u64 iter_count, u64 samples_count) { - double predicted_percent = 0.0; - const char *null_str = ""; - char iter_str[32]; - char cycle_str[32]; - char *istr, *cstr; u64 cycles; + int printed = 0, i = 0; if (branch_count == 0) return scnprintf(bf, bfsize, " (calltrace)"); - cycles = cycles_count / branch_count; - - if (iter_count && samples_count) { - if (cycles > 0) - scnprintf(iter_str, sizeof(iter_str), - " iterations:%" PRId64 "", - iter_count / samples_count); - else - scnprintf(iter_str, sizeof(iter_str), - "iterations:%" PRId64 "", - iter_count / samples_count); - istr = iter_str; - } else - istr = (char *)null_str; - - if (cycles > 0) { - scnprintf(cycle_str, sizeof(cycle_str), - "cycles:%" PRId64 "", cycles); - cstr = cycle_str; - } else - cstr = (char *)null_str; - - predicted_percent = predicted_count * 100.0 / branch_count; + if (predicted_count < branch_count) { + printed += count_float_printf(i++, "predicted", + predicted_count * 100.0 / branch_count, + bf + printed, bfsize - printed); + } - if ((predicted_count == branch_count) && (abort_count == 0)) { - if ((cycles > 0) || (istr != (char *)null_str)) - return scnprintf(bf, bfsize, " (%s%s)", cstr, istr); - else - return scnprintf(bf, bfsize, "%s", (char *)null_str); + if (abort_count) { + printed += count_float_printf(i++, "abort", + abort_count * 100.0 / branch_count, + bf + printed, bfsize - printed); } - if ((predicted_count < branch_count) && (abort_count == 0)) { - if ((cycles > 0) || (istr != (char *)null_str)) - return scnprintf(bf, bfsize, - " (predicted:%.1f%% %s%s)", - predicted_percent, cstr, istr); - else { - return scnprintf(bf, bfsize, - " (predicted:%.1f%%)", - predicted_percent); - } + cycles = cycles_count / branch_count; + if (cycles) { + printed += count_pri64_printf(i++, "cycles", + cycles, + bf + printed, bfsize - printed); } - if ((predicted_count == branch_count) && (abort_count > 0)) { - if ((cycles > 0) || (istr != (char *)null_str)) - return scnprintf(bf, bfsize, - " (abort:%" PRId64 " %s%s)", - abort_count, cstr, istr); - else - return scnprintf(bf, bfsize, - " (abort:%" PRId64 ")", - abort_count); + if (iter_count && samples_count) { + printed += count_pri64_printf(i++, "iterations", + iter_count / samples_count, + bf + printed, bfsize - printed); } - if ((cycles > 0) || (istr != (char *)null_str)) - return scnprintf(bf, bfsize, - " (predicted:%.1f%% abort:%" PRId64 " %s%s)", - predicted_percent, abort_count, cstr, istr); + if (i) + return scnprintf(bf + printed, bfsize - printed, ")"); - return scnprintf(bf, bfsize, - " (predicted:%.1f%% abort:%" PRId64 ")", - predicted_percent, abort_count); + bf[0] = 0; + return 0; } static int callchain_counts_printf(FILE *fp, char *bf, int bfsize, -- cgit v1.1 From 992c7e9267c12a8e301152c5569028ff8d535322 Mon Sep 17 00:00:00 2001 From: Jin Yao Date: Tue, 18 Jul 2017 20:13:13 +0800 Subject: perf util: Create branch.c/.h for common branch functions Create new util/branch.c and util/branch.h to contain the common branch functions. Such as: branch_type_count(): Count the numbers of branch types branch_type_name() : Return the name of branch type branch_type_stat_display(): Display branch type statistics info branch_type_str(): Construct the branch type string. The branch type is saved in branch_flags. Change log: v8: Change PERF_BR_NONE to PERF_BR_UNKNOWN. v7: Since the common branch type name is changed (e.g. JCC->COND), this patch is performed the modification accordingly. v6: Move that multiline conditional code inside {} brackets. Move branch_type_stat_display() from builtin-report.c to branch.c. Move branch_type_str() from callchain.c to branch.c. v5: It's a new patch in v5 patch series. Signed-off-by: Yao Jin Acked-by: Jiri Olsa Cc: Alexander Shishkin Cc: Andi Kleen Cc: Kan Liang Cc: Michael Ellerman Cc: Peter Zijlstra Link: http://lkml.kernel.org/r/1500379995-6449-6-git-send-email-yao.jin@linux.intel.com [ Don't use 'index' and 'stat' as names for variables, it shadows global decls in older distros ] Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/Build | 1 + tools/perf/util/branch.c | 147 +++++++++++++++++++++++++++++++++++++++++++++++ tools/perf/util/branch.h | 24 ++++++++ tools/perf/util/event.h | 3 +- 4 files changed, 174 insertions(+), 1 deletion(-) create mode 100644 tools/perf/util/branch.c create mode 100644 tools/perf/util/branch.h (limited to 'tools') diff --git a/tools/perf/util/Build b/tools/perf/util/Build index 7580fe4..8d49a98 100644 --- a/tools/perf/util/Build +++ b/tools/perf/util/Build @@ -93,6 +93,7 @@ libperf-y += drv_configs.o libperf-y += units.o libperf-y += time-utils.o libperf-y += expr-bison.o +libperf-y += branch.o libperf-$(CONFIG_LIBBPF) += bpf-loader.o libperf-$(CONFIG_BPF_PROLOGUE) += bpf-prologue.o diff --git a/tools/perf/util/branch.c b/tools/perf/util/branch.c new file mode 100644 index 0000000..a4fce27 --- /dev/null +++ b/tools/perf/util/branch.c @@ -0,0 +1,147 @@ +#include "perf.h" +#include "util/util.h" +#include "util/debug.h" +#include "util/branch.h" + +static bool cross_area(u64 addr1, u64 addr2, int size) +{ + u64 align1, align2; + + align1 = addr1 & ~(size - 1); + align2 = addr2 & ~(size - 1); + + return (align1 != align2) ? true : false; +} + +#define AREA_4K 4096 +#define AREA_2M (2 * 1024 * 1024) + +void branch_type_count(struct branch_type_stat *st, struct branch_flags *flags, + u64 from, u64 to) +{ + if (flags->type == PERF_BR_UNKNOWN || from == 0) + return; + + st->counts[flags->type]++; + + if (flags->type == PERF_BR_COND) { + if (to > from) + st->cond_fwd++; + else + st->cond_bwd++; + } + + if (cross_area(from, to, AREA_2M)) + st->cross_2m++; + else if (cross_area(from, to, AREA_4K)) + st->cross_4k++; +} + +const char *branch_type_name(int type) +{ + const char *branch_names[PERF_BR_MAX] = { + "N/A", + "COND", + "UNCOND", + "IND", + "CALL", + "IND_CALL", + "RET", + "SYSCALL", + "SYSRET", + "COND_CALL", + "COND_RET" + }; + + if (type >= 0 && type < PERF_BR_MAX) + return branch_names[type]; + + return NULL; +} + +void branch_type_stat_display(FILE *fp, struct branch_type_stat *st) +{ + u64 total = 0; + int i; + + for (i = 0; i < PERF_BR_MAX; i++) + total += st->counts[i]; + + if (total == 0) + return; + + fprintf(fp, "\n#"); + fprintf(fp, "\n# Branch Statistics:"); + fprintf(fp, "\n#"); + + if (st->cond_fwd > 0) { + fprintf(fp, "\n%8s: %5.1f%%", + "COND_FWD", + 100.0 * (double)st->cond_fwd / (double)total); + } + + if (st->cond_bwd > 0) { + fprintf(fp, "\n%8s: %5.1f%%", + "COND_BWD", + 100.0 * (double)st->cond_bwd / (double)total); + } + + if (st->cross_4k > 0) { + fprintf(fp, "\n%8s: %5.1f%%", + "CROSS_4K", + 100.0 * (double)st->cross_4k / (double)total); + } + + if (st->cross_2m > 0) { + fprintf(fp, "\n%8s: %5.1f%%", + "CROSS_2M", + 100.0 * (double)st->cross_2m / (double)total); + } + + for (i = 0; i < PERF_BR_MAX; i++) { + if (st->counts[i] > 0) + fprintf(fp, "\n%8s: %5.1f%%", + branch_type_name(i), + 100.0 * + (double)st->counts[i] / (double)total); + } +} + +static int count_str_scnprintf(int idx, const char *str, char *bf, int size) +{ + return scnprintf(bf, size, "%s%s", (idx) ? " " : " (", str); +} + +int branch_type_str(struct branch_type_stat *st, char *bf, int size) +{ + int i, j = 0, printed = 0; + u64 total = 0; + + for (i = 0; i < PERF_BR_MAX; i++) + total += st->counts[i]; + + if (total == 0) + return 0; + + if (st->cond_fwd > 0) + printed += count_str_scnprintf(j++, "COND_FWD", bf + printed, size - printed); + + if (st->cond_bwd > 0) + printed += count_str_scnprintf(j++, "COND_BWD", bf + printed, size - printed); + + for (i = 0; i < PERF_BR_MAX; i++) { + if (i == PERF_BR_COND) + continue; + + if (st->counts[i] > 0) + printed += count_str_scnprintf(j++, branch_type_name(i), bf + printed, size - printed); + } + + if (st->cross_4k > 0) + printed += count_str_scnprintf(j++, "CROSS_4K", bf + printed, size - printed); + + if (st->cross_2m > 0) + printed += count_str_scnprintf(j++, "CROSS_2M", bf + printed, size - printed); + + return printed; +} diff --git a/tools/perf/util/branch.h b/tools/perf/util/branch.h new file mode 100644 index 0000000..686f2b6 --- /dev/null +++ b/tools/perf/util/branch.h @@ -0,0 +1,24 @@ +#ifndef _PERF_BRANCH_H +#define _PERF_BRANCH_H 1 + +#include +#include "../perf.h" + +struct branch_type_stat { + u64 counts[PERF_BR_MAX]; + u64 cond_fwd; + u64 cond_bwd; + u64 cross_4k; + u64 cross_2m; +}; + +struct branch_flags; + +void branch_type_count(struct branch_type_stat *st, struct branch_flags *flags, + u64 from, u64 to); + +const char *branch_type_name(int type); +void branch_type_stat_display(FILE *fp, struct branch_type_stat *st); +int branch_type_str(struct branch_type_stat *st, char *bf, int bfsize); + +#endif /* _PERF_BRANCH_H */ diff --git a/tools/perf/util/event.h b/tools/perf/util/event.h index 37c5faf..423ac82 100644 --- a/tools/perf/util/event.h +++ b/tools/perf/util/event.h @@ -142,7 +142,8 @@ struct branch_flags { u64 in_tx:1; u64 abort:1; u64 cycles:16; - u64 reserved:44; + u64 type:4; + u64 reserved:40; }; struct branch_entry { -- cgit v1.1 From 2d78b18952a1bdf125d13fa6bb68fbc5c1b0aed9 Mon Sep 17 00:00:00 2001 From: Jin Yao Date: Tue, 18 Jul 2017 20:13:14 +0800 Subject: perf report: Show branch type statistics for stdio mode Show the branch type statistics at the end of perf report --stdio. For example: perf report --stdio COND_FWD: 28.5% COND_BWD: 9.4% CROSS_4K: 0.7% CROSS_2M: 14.1% COND: 37.9% UNCOND: 0.2% IND: 6.7% CALL: 26.5% RET: 28.7% SYSRET: 0.0% The branch types are: COND_FWD: conditional forward COND_BWD: conditional backward COND: conditional branch UNCOND: unconditional branch IND: indirect CALL: function call IND_CALL: indirect function call RET: function return SYSCALL: syscall SYSRET: syscall return COND_CALL: conditional function call COND_RET: conditional function return CROSS_4K and CROSS_2M: They are the metrics checking for branches cross 4K or 2MB pages. It's an approximate computing. We don't know if the area is 4K or 2MB, so always compute both. To make the output simple, if a branch crosses 2M area, CROSS_4K will not be incremented. Change log v7: Since the common branch type definitions are changed, some tags/strings are updated accordingly. v6: Remove branch_type_stat_display() since it's moved to branch.c. v5: Remove the unnecessary sort__mode checking in hist_iter__branch_callback(). v4: Comparing to previous version, the major changes are: Add the computing of JCC forward/JCC backward and cross page checking by using the from and to addresses. Signed-off-by: Yao Jin Acked-by: Jiri Olsa Cc: Alexander Shishkin Cc: Andi Kleen Cc: Kan Liang Cc: Michael Ellerman Cc: Peter Zijlstra Link: http://lkml.kernel.org/r/1500379995-6449-7-git-send-email-yao.jin@linux.intel.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-report.c | 25 +++++++++++++++++++++++++ tools/perf/util/hist.c | 5 +---- 2 files changed, 26 insertions(+), 4 deletions(-) (limited to 'tools') diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c index 8e752ba..cea25d0 100644 --- a/tools/perf/builtin-report.c +++ b/tools/perf/builtin-report.c @@ -38,6 +38,7 @@ #include "util/time-utils.h" #include "util/auxtrace.h" #include "util/units.h" +#include "util/branch.h" #include #include @@ -73,6 +74,7 @@ struct report { u64 queue_size; int socket_filter; DECLARE_BITMAP(cpu_bitmap, MAX_NR_CPUS); + struct branch_type_stat brtype_stat; }; static int report__config(const char *var, const char *value, void *cb) @@ -150,6 +152,22 @@ out: return err; } +static int hist_iter__branch_callback(struct hist_entry_iter *iter, + struct addr_location *al __maybe_unused, + bool single __maybe_unused, + void *arg) +{ + struct hist_entry *he = iter->he; + struct report *rep = arg; + struct branch_info *bi; + + bi = he->branch_info; + branch_type_count(&rep->brtype_stat, &bi->flags, + bi->from.addr, bi->to.addr); + + return 0; +} + static int process_sample_event(struct perf_tool *tool, union perf_event *event, struct perf_sample *sample, @@ -188,6 +206,8 @@ static int process_sample_event(struct perf_tool *tool, */ if (!sample->branch_stack) goto out_put; + + iter.add_entry_cb = hist_iter__branch_callback; iter.ops = &hist_iter_branch; } else if (rep->mem_mode) { iter.ops = &hist_iter_mem; @@ -410,6 +430,9 @@ static int perf_evlist__tty_browse_hists(struct perf_evlist *evlist, perf_read_values_destroy(&rep->show_threads_values); } + if (sort__mode == SORT_MODE__BRANCH) + branch_type_stat_display(stdout, &rep->brtype_stat); + return 0; } @@ -944,6 +967,8 @@ repeat: if (has_br_stack && branch_call_mode) symbol_conf.show_branchflag_count = true; + memset(&report.brtype_stat, 0, sizeof(struct branch_type_stat)); + /* * Branch mode is a tristate: * -1 means default, so decide based on the file having branch data. diff --git a/tools/perf/util/hist.c b/tools/perf/util/hist.c index cf0186a..2f6c5e6 100644 --- a/tools/perf/util/hist.c +++ b/tools/perf/util/hist.c @@ -749,12 +749,9 @@ iter_prepare_branch_entry(struct hist_entry_iter *iter, struct addr_location *al } static int -iter_add_single_branch_entry(struct hist_entry_iter *iter, +iter_add_single_branch_entry(struct hist_entry_iter *iter __maybe_unused, struct addr_location *al __maybe_unused) { - /* to avoid calling callback function */ - iter->he = NULL; - return 0; } -- cgit v1.1 From b851dd49868e295e18c5d72fc3bad85ff1c444b1 Mon Sep 17 00:00:00 2001 From: Jin Yao Date: Tue, 18 Jul 2017 20:13:15 +0800 Subject: perf report: Show branch type in callchain entry Show branch type in callchain entry. The branch type is printed with other LBR information (such as cycles/abort/...). For example: perf record -g -j any,save_type perf report --branch-history --stdio --no-children 38.50% div.c:45 [.] main div | ---main div.c:42 (RET CROSS_2M cycles:2) compute_flag div.c:28 (cycles:2) compute_flag div.c:27 (RET CROSS_2M cycles:1) rand rand.c:28 (cycles:1) rand rand.c:28 (RET CROSS_2M cycles:1) __random random.c:298 (cycles:1) __random random.c:297 (COND_BWD CROSS_2M cycles:1) __random random.c:295 (cycles:1) __random random.c:295 (COND_BWD CROSS_2M cycles:1) __random random.c:295 (cycles:1) __random random.c:295 (RET CROSS_2M cycles:9) Change log v6: Remove the branch_type_str() since it's moved to branch.c. v5: Rewrite the branch info print code in util/callchain.c. v4: Comparing to previous version, the major changes are: Signed-off-by: Yao Jin Acked-by: Jiri Olsa Cc: Alexander Shishkin Cc: Andi Kleen Cc: Kan Liang Cc: Michael Ellerman Cc: Peter Zijlstra Link: http://lkml.kernel.org/r/1500379995-6449-8-git-send-email-yao.jin@linux.intel.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/callchain.c | 38 +++++++++++++++++++++++++++++--------- tools/perf/util/callchain.h | 5 ++++- tools/perf/util/machine.c | 26 +++++++++++++++++--------- 3 files changed, 50 insertions(+), 19 deletions(-) (limited to 'tools') diff --git a/tools/perf/util/callchain.c b/tools/perf/util/callchain.c index 917f4d6..22d413a 100644 --- a/tools/perf/util/callchain.c +++ b/tools/perf/util/callchain.c @@ -23,6 +23,7 @@ #include "sort.h" #include "machine.h" #include "callchain.h" +#include "branch.h" #define CALLCHAIN_PARAM_DEFAULT \ .mode = CHAIN_GRAPH_ABS, \ @@ -571,6 +572,11 @@ fill_node(struct callchain_node *node, struct callchain_cursor *cursor) call->cycles_count = cursor_node->branch_flags.cycles; call->iter_count = cursor_node->nr_loop_iter; call->samples_count = cursor_node->samples; + + branch_type_count(&call->brtype_stat, + &cursor_node->branch_flags, + cursor_node->branch_from, + cursor_node->ip); } list_add_tail(&call->list, &node->val); @@ -688,6 +694,11 @@ static enum match_result match_chain(struct callchain_cursor_node *node, cnode->cycles_count += node->branch_flags.cycles; cnode->iter_count += node->nr_loop_iter; cnode->samples_count += node->samples; + + branch_type_count(&cnode->brtype_stat, + &node->branch_flags, + node->branch_from, + node->ip); } return MATCH_EQ; @@ -922,7 +933,7 @@ merge_chain_branch(struct callchain_cursor *cursor, list_for_each_entry_safe(list, next_list, &src->val, list) { callchain_cursor_append(cursor, list->ip, list->ms.map, list->ms.sym, - false, NULL, 0, 0); + false, NULL, 0, 0, 0); list_del(&list->list); map__zput(list->ms.map); free(list); @@ -962,7 +973,7 @@ int callchain_merge(struct callchain_cursor *cursor, int callchain_cursor_append(struct callchain_cursor *cursor, u64 ip, struct map *map, struct symbol *sym, bool branch, struct branch_flags *flags, - int nr_loop_iter, int samples) + int nr_loop_iter, int samples, u64 branch_from) { struct callchain_cursor_node *node = *cursor->last; @@ -986,6 +997,7 @@ int callchain_cursor_append(struct callchain_cursor *cursor, memcpy(&node->branch_flags, flags, sizeof(struct branch_flags)); + node->branch_from = branch_from; cursor->nr++; cursor->last = &node->next; @@ -1235,14 +1247,19 @@ static int count_float_printf(int idx, const char *str, float value, char *bf, i static int counts_str_build(char *bf, int bfsize, u64 branch_count, u64 predicted_count, u64 abort_count, u64 cycles_count, - u64 iter_count, u64 samples_count) + u64 iter_count, u64 samples_count, + struct branch_type_stat *brtype_stat) { u64 cycles; - int printed = 0, i = 0; + int printed, i = 0; if (branch_count == 0) return scnprintf(bf, bfsize, " (calltrace)"); + printed = branch_type_str(brtype_stat, bf, bfsize); + if (printed) + i++; + if (predicted_count < branch_count) { printed += count_float_printf(i++, "predicted", predicted_count * 100.0 / branch_count, @@ -1278,13 +1295,14 @@ static int counts_str_build(char *bf, int bfsize, static int callchain_counts_printf(FILE *fp, char *bf, int bfsize, u64 branch_count, u64 predicted_count, u64 abort_count, u64 cycles_count, - u64 iter_count, u64 samples_count) + u64 iter_count, u64 samples_count, + struct branch_type_stat *brtype_stat) { - char str[128]; + char str[256]; counts_str_build(str, sizeof(str), branch_count, predicted_count, abort_count, cycles_count, - iter_count, samples_count); + iter_count, samples_count, brtype_stat); if (fp) return fprintf(fp, "%s", str); @@ -1316,7 +1334,8 @@ int callchain_list_counts__printf_value(struct callchain_node *node, return callchain_counts_printf(fp, bf, bfsize, branch_count, predicted_count, abort_count, - cycles_count, iter_count, samples_count); + cycles_count, iter_count, samples_count, + &clist->brtype_stat); } static void free_callchain_node(struct callchain_node *node) @@ -1441,7 +1460,8 @@ int callchain_cursor__copy(struct callchain_cursor *dst, rc = callchain_cursor_append(dst, node->ip, node->map, node->sym, node->branch, &node->branch_flags, - node->nr_loop_iter, node->samples); + node->nr_loop_iter, node->samples, + node->branch_from); if (rc) break; diff --git a/tools/perf/util/callchain.h b/tools/perf/util/callchain.h index c56c23d..9773820 100644 --- a/tools/perf/util/callchain.h +++ b/tools/perf/util/callchain.h @@ -7,6 +7,7 @@ #include "event.h" #include "map.h" #include "symbol.h" +#include "branch.h" #define HELP_PAD "\t\t\t\t" @@ -119,6 +120,7 @@ struct callchain_list { u64 cycles_count; u64 iter_count; u64 samples_count; + struct branch_type_stat brtype_stat; char *srcline; struct list_head list; }; @@ -135,6 +137,7 @@ struct callchain_cursor_node { struct symbol *sym; bool branch; struct branch_flags branch_flags; + u64 branch_from; int nr_loop_iter; int samples; struct callchain_cursor_node *next; @@ -198,7 +201,7 @@ static inline void callchain_cursor_reset(struct callchain_cursor *cursor) int callchain_cursor_append(struct callchain_cursor *cursor, u64 ip, struct map *map, struct symbol *sym, bool branch, struct branch_flags *flags, - int nr_loop_iter, int samples); + int nr_loop_iter, int samples, u64 branch_from); /* Close a cursor writing session. Initialize for the reader */ static inline void callchain_cursor_commit(struct callchain_cursor *cursor) diff --git a/tools/perf/util/machine.c b/tools/perf/util/machine.c index a54a2be..79d08ea 100644 --- a/tools/perf/util/machine.c +++ b/tools/perf/util/machine.c @@ -1682,7 +1682,8 @@ static int add_callchain_ip(struct thread *thread, bool branch, struct branch_flags *flags, int nr_loop_iter, - int samples) + int samples, + u64 branch_from) { struct addr_location al; @@ -1735,7 +1736,8 @@ static int add_callchain_ip(struct thread *thread, if (symbol_conf.hide_unresolved && al.sym == NULL) return 0; return callchain_cursor_append(cursor, al.addr, al.map, al.sym, - branch, flags, nr_loop_iter, samples); + branch, flags, nr_loop_iter, samples, + branch_from); } struct branch_info *sample__resolve_bstack(struct perf_sample *sample, @@ -1814,7 +1816,7 @@ static int resolve_lbr_callchain_sample(struct thread *thread, struct ip_callchain *chain = sample->callchain; int chain_nr = min(max_stack, (int)chain->nr), i; u8 cpumode = PERF_RECORD_MISC_USER; - u64 ip; + u64 ip, branch_from = 0; for (i = 0; i < chain_nr; i++) { if (chain->ips[i] == PERF_CONTEXT_USER) @@ -1856,6 +1858,8 @@ static int resolve_lbr_callchain_sample(struct thread *thread, ip = lbr_stack->entries[0].to; branch = true; flags = &lbr_stack->entries[0].flags; + branch_from = + lbr_stack->entries[0].from; } } else { if (j < lbr_nr) { @@ -1870,12 +1874,15 @@ static int resolve_lbr_callchain_sample(struct thread *thread, ip = lbr_stack->entries[0].to; branch = true; flags = &lbr_stack->entries[0].flags; + branch_from = + lbr_stack->entries[0].from; } } err = add_callchain_ip(thread, cursor, parent, root_al, &cpumode, ip, - branch, flags, 0, 0); + branch, flags, 0, 0, + branch_from); if (err) return (err < 0) ? err : 0; } @@ -1974,19 +1981,20 @@ static int thread__resolve_callchain_sample(struct thread *thread, root_al, NULL, be[i].to, true, &be[i].flags, - nr_loop_iter, 1); + nr_loop_iter, 1, + be[i].from); else err = add_callchain_ip(thread, cursor, parent, root_al, NULL, be[i].to, true, &be[i].flags, - 0, 0); + 0, 0, be[i].from); if (!err) err = add_callchain_ip(thread, cursor, parent, root_al, NULL, be[i].from, true, &be[i].flags, - 0, 0); + 0, 0, 0); if (err == -EINVAL) break; if (err) @@ -2016,7 +2024,7 @@ check_calls: err = add_callchain_ip(thread, cursor, parent, root_al, &cpumode, ip, - false, NULL, 0, 0); + false, NULL, 0, 0, 0); if (err) return (err < 0) ? err : 0; @@ -2033,7 +2041,7 @@ static int unwind_entry(struct unwind_entry *entry, void *arg) return 0; return callchain_cursor_append(cursor, entry->ip, entry->map, entry->sym, - false, NULL, 0, 0); + false, NULL, 0, 0, 0); } static int thread__resolve_callchain_unwind(struct thread *thread, -- cgit v1.1 From d12fe87e62d773e81e0cb3a123c5a480a10d7d91 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Mon, 26 Jun 2017 16:36:57 -0500 Subject: signal/testing: Don't look for __SI_FAULT in userspace Fix the debug print statements in these tests where they reference si_codes and in particular __SI_FAULT. __SI_FAULT is a kernel internal value and should never be seen by userspace. While I am in there also fix si_code_str. si_codes are an enumeration there are not a bitmap so == and not & is the apropriate operation to test for an si_code. Cc: Dave Hansen Fixes: 5f23f6d082a9 ("x86/pkeys: Add self-tests") Fixes: e754aedc26ef ("x86/mpx, selftests: Add MPX self test") Signed-off-by: "Eric W. Biederman" --- tools/testing/selftests/x86/mpx-mini-test.c | 3 +-- tools/testing/selftests/x86/protection_keys.c | 13 ++++++------- 2 files changed, 7 insertions(+), 9 deletions(-) (limited to 'tools') diff --git a/tools/testing/selftests/x86/mpx-mini-test.c b/tools/testing/selftests/x86/mpx-mini-test.c index a8df159..ec0f6b4 100644 --- a/tools/testing/selftests/x86/mpx-mini-test.c +++ b/tools/testing/selftests/x86/mpx-mini-test.c @@ -391,8 +391,7 @@ void handler(int signum, siginfo_t *si, void *vucontext) br_count++; dprintf1("#BR 0x%jx (total seen: %d)\n", status, br_count); -#define __SI_FAULT (3 << 16) -#define SEGV_BNDERR (__SI_FAULT|3) /* failed address bound checks */ +#define SEGV_BNDERR 3 /* failed address bound checks */ dprintf2("Saw a #BR! status 0x%jx at %016lx br_reason: %jx\n", status, ip, br_reason); diff --git a/tools/testing/selftests/x86/protection_keys.c b/tools/testing/selftests/x86/protection_keys.c index 3237bc0..2392784 100644 --- a/tools/testing/selftests/x86/protection_keys.c +++ b/tools/testing/selftests/x86/protection_keys.c @@ -212,19 +212,18 @@ void dump_mem(void *dumpme, int len_bytes) } } -#define __SI_FAULT (3 << 16) -#define SEGV_BNDERR (__SI_FAULT|3) /* failed address bound checks */ -#define SEGV_PKUERR (__SI_FAULT|4) +#define SEGV_BNDERR 3 /* failed address bound checks */ +#define SEGV_PKUERR 4 static char *si_code_str(int si_code) { - if (si_code & SEGV_MAPERR) + if (si_code == SEGV_MAPERR) return "SEGV_MAPERR"; - if (si_code & SEGV_ACCERR) + if (si_code == SEGV_ACCERR) return "SEGV_ACCERR"; - if (si_code & SEGV_BNDERR) + if (si_code == SEGV_BNDERR) return "SEGV_BNDERR"; - if (si_code & SEGV_PKUERR) + if (si_code == SEGV_PKUERR) return "SEGV_PKUERR"; return "UNKNOWN"; } -- cgit v1.1 From 69d8bd8aa7d8906a1e922ae884d97f0bd7f1b269 Mon Sep 17 00:00:00 2001 From: Kan Liang Date: Fri, 30 Jun 2017 10:16:55 -0400 Subject: perf intel-pt: Set no_aux_samples for the tracking event The reason of introducing the tracking event (a dummy software event) is to collect side-band information. Additional sampling is wasteful. no_aux_samples should be set for tracking event. Signed-off-by: Kan Liang Acked-by: Jiri Olsa Cc: Adrian Hunter Cc: Alexander Shishkin Cc: Andi Kleen Cc: Peter Zijlstra Link: http://lkml.kernel.org/r/20170630141656.1626-1-kan.liang@intel.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/arch/x86/util/intel-pt.c | 1 + 1 file changed, 1 insertion(+) (limited to 'tools') diff --git a/tools/perf/arch/x86/util/intel-pt.c b/tools/perf/arch/x86/util/intel-pt.c index 9535be5..4a461e8 100644 --- a/tools/perf/arch/x86/util/intel-pt.c +++ b/tools/perf/arch/x86/util/intel-pt.c @@ -752,6 +752,7 @@ static int intel_pt_recording_options(struct auxtrace_record *itr, tracking_evsel->attr.freq = 0; tracking_evsel->attr.sample_period = 1; + tracking_evsel->no_aux_samples = true; if (need_immediate) tracking_evsel->immediate = true; -- cgit v1.1 From 91a8c5b840f2da31280e14b6268761cf14033756 Mon Sep 17 00:00:00 2001 From: Kan Liang Date: Fri, 30 Jun 2017 10:16:56 -0400 Subject: perf intel-pt: Always set no branch for dummy event An earlier kernel patch allowed enabling PT and LBR at the same time on Goldmont. commit ccbebba4c6bf ("perf/x86/intel/pt: Bypass PT vs. LBR exclusivity if the core supports it") However, users still cannot use Intel PT and LBRs simultaneously. $ sudo perf record -e cycles,intel_pt//u -b -- sleep 1 Error: PMU Hardware doesn't support sampling/overflow-interrupts. PT implicitly adds dummy event in perf tool. dummy event is software event which doesn't support LBR. Always setting no branch for dummy event in Intel PT. Signed-off-by: Kan Liang Acked-by: Jiri Olsa Cc: Adrian Hunter Cc: Alexander Shishkin Cc: Andi Kleen Cc: Peter Zijlstra Link: http://lkml.kernel.org/r/20170630141656.1626-2-kan.liang@intel.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/arch/x86/util/intel-pt.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'tools') diff --git a/tools/perf/arch/x86/util/intel-pt.c b/tools/perf/arch/x86/util/intel-pt.c index 4a461e8..db0ba8c 100644 --- a/tools/perf/arch/x86/util/intel-pt.c +++ b/tools/perf/arch/x86/util/intel-pt.c @@ -701,6 +701,7 @@ static int intel_pt_recording_options(struct auxtrace_record *itr, perf_evsel__set_sample_bit(switch_evsel, TID); perf_evsel__set_sample_bit(switch_evsel, TIME); perf_evsel__set_sample_bit(switch_evsel, CPU); + perf_evsel__reset_sample_bit(switch_evsel, BRANCH_STACK); opts->record_switch_events = false; ptr->have_sched_switch = 3; @@ -762,6 +763,7 @@ static int intel_pt_recording_options(struct auxtrace_record *itr, /* And the CPU for switch events */ perf_evsel__set_sample_bit(tracking_evsel, CPU); } + perf_evsel__reset_sample_bit(tracking_evsel, BRANCH_STACK); } /* -- cgit v1.1 From 6f8fe61ee583f1b75fef81d9c9434b7496a01881 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Wed, 19 Jul 2017 11:39:47 -0300 Subject: perf trace: Add missing ' = ' in the default formatting of syscall returns We lost it recently, put it back. Before: 789.499 ( 0.001 ms): libvirtd/1175 lseek(fd: 22, whence: CUR) 4328 After: 789.499 ( 0.001 ms): libvirtd/1175 lseek(fd: 22, whence: CUR) = 4328 Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Fixes: 1f63139c3f8a ("perf trace beauty: Simplify syscall return formatting") Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-trace.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'tools') diff --git a/tools/perf/builtin-trace.c b/tools/perf/builtin-trace.c index 1e4c065..4370ce4 100644 --- a/tools/perf/builtin-trace.c +++ b/tools/perf/builtin-trace.c @@ -1660,7 +1660,7 @@ static int trace__sys_exit(struct trace *trace, struct perf_evsel *evsel, if (ret < 0) goto errno_print; signed_print: - fprintf(trace->output, ") %ld", ret); + fprintf(trace->output, ") = %ld", ret); } else if (ret < 0) { errno_print: { char bf[STRERR_BUFSIZE]; -- cgit v1.1 From d57da8c9a5e11e7786931cfa1fd744969f530b2e Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Wed, 19 Jul 2017 12:16:15 -0300 Subject: perf trace beauty mmap: Ignore 'fd' and 'offset' args for MAP_ANONYMOUS Just suppress them, not used by the kernel. Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-atpt07y2x9a8ttlwja94ow3j@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/trace/beauty/mmap.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'tools') diff --git a/tools/perf/trace/beauty/mmap.c b/tools/perf/trace/beauty/mmap.c index af1cfde..754558f 100644 --- a/tools/perf/trace/beauty/mmap.c +++ b/tools/perf/trace/beauty/mmap.c @@ -34,6 +34,9 @@ static size_t syscall_arg__scnprintf_mmap_flags(char *bf, size_t size, { int printed = 0, flags = arg->val; + if (flags & MAP_ANONYMOUS) + arg->mask |= (1 << 4) | (1 << 5); /* Mask 4th ('fd') and 5th ('offset') args, ignored */ + #define P_MMAP_FLAG(n) \ if (flags & MAP_##n) { \ printed += scnprintf(bf + printed, size - printed, "%s%s", printed ? "|" : "", #n); \ -- cgit v1.1 From 5e58fcfaf4c60795c241be739ad55c519d7f2aaa Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Wed, 19 Jul 2017 14:32:11 -0300 Subject: perf trace: Allow allocating sc->arg_fmt even without the syscall tracepoint At least "clone" doesn't have (enter, exit) entries tracefs/events/syscalls/, but we can provide a syscall_fmt and use it instead, as will be done for "clone" in the next cset. Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-o12kejgcxddyovn2hlg4gbim@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-trace.c | 31 +++++++++++++++++++++---------- 1 file changed, 21 insertions(+), 10 deletions(-) (limited to 'tools') diff --git a/tools/perf/builtin-trace.c b/tools/perf/builtin-trace.c index 4370ce4..1c5238a 100644 --- a/tools/perf/builtin-trace.c +++ b/tools/perf/builtin-trace.c @@ -1165,22 +1165,31 @@ static int trace__symbols_init(struct trace *trace, struct perf_evlist *evlist) return err; } -static int syscall__set_arg_fmts(struct syscall *sc) +static int syscall__alloc_arg_fmts(struct syscall *sc, int nr_args) { - struct format_field *field; - int idx = 0, len; + int idx; - sc->arg_fmt = calloc(sc->nr_args, sizeof(*sc->arg_fmt)); + sc->arg_fmt = calloc(nr_args, sizeof(*sc->arg_fmt)); if (sc->arg_fmt == NULL) return -1; - for (field = sc->args; field; field = field->next, ++idx) { - if (sc->fmt) { + for (idx = 0; idx < nr_args; ++idx) { + if (sc->fmt) sc->arg_fmt[idx] = sc->fmt->arg[idx]; + } - if (sc->fmt->arg[idx].scnprintf) - continue; - } + sc->nr_args = nr_args; + return 0; +} + +static int syscall__set_arg_fmts(struct syscall *sc) +{ + struct format_field *field; + int idx = 0, len; + + for (field = sc->args; field; field = field->next, ++idx) { + if (sc->fmt && sc->fmt->arg[idx].scnprintf) + continue; if (strcmp(field->type, "const char *") == 0 && (strcmp(field->name, "filename") == 0 || @@ -1251,11 +1260,13 @@ static int trace__read_syscall_info(struct trace *trace, int id) sc->tp_format = trace_event__tp_format("syscalls", tp_name); } + if (syscall__alloc_arg_fmts(sc, IS_ERR(sc->tp_format) ? 6 : sc->tp_format->format.nr_fields)) + return -1; + if (IS_ERR(sc->tp_format)) return -1; sc->args = sc->tp_format->format.fields; - sc->nr_args = sc->tp_format->format.nr_fields; /* * We need to check and discard the first variable '__syscall_nr' * or 'nr' that mean the syscall number. It is needless here. -- cgit v1.1 From d032d79e2dcb56e13678bf2cc7b36957ef827c32 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Wed, 19 Jul 2017 14:36:13 -0300 Subject: perf trace: Use the syscall_fmt formatters without a tracepoint Previously we only used the syscall_fmt when we had sc->tp_format set, i.e. when we found the (enter, exit) pair in tracefs/events/syscalls/. But we really only need to use what is in sc->arg_fmt to apply the arg beautifiers to the syscall argument values, so do it. With this we will be able to provide formatters to the "clone" syscall, which doesn't have entries in tracefs/events/syscalls/. Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-y41nl41jrayjo5ucnde2peix@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-trace.c | 54 +++++++++++++++++++++++++--------------------- 1 file changed, 30 insertions(+), 24 deletions(-) (limited to 'tools') diff --git a/tools/perf/builtin-trace.c b/tools/perf/builtin-trace.c index 1c5238a..fc4d33a 100644 --- a/tools/perf/builtin-trace.c +++ b/tools/perf/builtin-trace.c @@ -1350,12 +1350,32 @@ unsigned long syscall_arg__val(struct syscall_arg *arg, u8 idx) return __syscall_arg__val(arg->args, idx); } +static size_t syscall__scnprintf_val(struct syscall *sc, char *bf, size_t size, + struct syscall_arg *arg, unsigned long val) +{ + if (sc->arg_fmt && sc->arg_fmt[arg->idx].scnprintf) { + arg->val = val; + if (sc->arg_fmt[arg->idx].parm) + arg->parm = sc->arg_fmt[arg->idx].parm; + return sc->arg_fmt[arg->idx].scnprintf(bf, size, arg); + } + return scnprintf(bf, size, "%ld", val); +} + static size_t syscall__scnprintf_args(struct syscall *sc, char *bf, size_t size, unsigned char *args, struct trace *trace, struct thread *thread) { size_t printed = 0; unsigned long val; + u8 bit = 1; + struct syscall_arg arg = { + .args = args, + .idx = 0, + .mask = 0, + .trace = trace, + .thread = thread, + }; struct thread_trace *ttrace = thread__priv(thread); /* @@ -1367,14 +1387,6 @@ static size_t syscall__scnprintf_args(struct syscall *sc, char *bf, size_t size, if (sc->args != NULL) { struct format_field *field; - u8 bit = 1; - struct syscall_arg arg = { - .args = args, - .idx = 0, - .mask = 0, - .trace = trace, - .thread = thread, - }; for (field = sc->args; field; field = field->next, ++arg.idx, bit <<= 1) { @@ -1398,15 +1410,7 @@ static size_t syscall__scnprintf_args(struct syscall *sc, char *bf, size_t size, printed += scnprintf(bf + printed, size - printed, "%s%s: ", printed ? ", " : "", field->name); - if (sc->arg_fmt && sc->arg_fmt[arg.idx].scnprintf) { - arg.val = val; - if (sc->arg_fmt[arg.idx].parm) - arg.parm = sc->arg_fmt[arg.idx].parm; - printed += sc->arg_fmt[arg.idx].scnprintf(bf + printed, size - printed, &arg); - } else { - printed += scnprintf(bf + printed, size - printed, - "%ld", val); - } + printed += syscall__scnprintf_val(sc, bf + printed, size - printed, &arg, val); } } else if (IS_ERR(sc->tp_format)) { /* @@ -1414,14 +1418,16 @@ static size_t syscall__scnprintf_args(struct syscall *sc, char *bf, size_t size, * may end up not having any args, like with gettid(), so only * print the raw args when we didn't manage to read it. */ - int i = 0; - - while (i < 6) { - val = __syscall_arg__val(args, i); + while (arg.idx < 6) { + if (arg.mask & bit) + goto next_arg; + val = syscall_arg__val(&arg, arg.idx); printed += scnprintf(bf + printed, size - printed, - "%sarg%d: %ld", - printed ? ", " : "", i, val); - ++i; + "%sarg%d: ", printed ? ", " : "", arg.idx); + printed += syscall__scnprintf_val(sc, bf + printed, size - printed, &arg, val); +next_arg: + ++arg.idx; + bit <<= 1; } } -- cgit v1.1 From 325f5091b0da5a7d8d190262661af17d75bee262 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Wed, 19 Jul 2017 15:02:43 -0300 Subject: perf trace: Ditch __syscall__arg_val() variant, not needed anymore All callers now can use syscall__arg_val(arg, idx), be it to iterate thru the syscall arguments while taking into account alignment, or to get values for other arguments that affect how the current argument should be formatted (think of fcntl's 'cmd' and 'arg' arguments). Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-wm5b156d8kro1r4y3b33eyta@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-trace.c | 9 ++------- 1 file changed, 2 insertions(+), 7 deletions(-) (limited to 'tools') diff --git a/tools/perf/builtin-trace.c b/tools/perf/builtin-trace.c index fc4d33a..f8b7bfd 100644 --- a/tools/perf/builtin-trace.c +++ b/tools/perf/builtin-trace.c @@ -1336,20 +1336,15 @@ out: * variable to read it. Most notably this avoids extended load instructions * on unaligned addresses */ -static unsigned long __syscall_arg__val(unsigned char *args, u8 idx) +unsigned long syscall_arg__val(struct syscall_arg *arg, u8 idx) { unsigned long val; - unsigned char *p = args + sizeof(unsigned long) * idx; + unsigned char *p = arg->args + sizeof(unsigned long) * idx; memcpy(&val, p, sizeof(val)); return val; } -unsigned long syscall_arg__val(struct syscall_arg *arg, u8 idx) -{ - return __syscall_arg__val(arg->args, idx); -} - static size_t syscall__scnprintf_val(struct syscall *sc, char *bf, size_t size, struct syscall_arg *arg, unsigned long val) { -- cgit v1.1 From 332337dafc98d88561bf3730f80c59cc93f089e1 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Wed, 19 Jul 2017 15:08:14 -0300 Subject: perf trace: Allow specifying number of syscall args for tracepointless syscalls When we don't have syscalls:sys_{enter,exit}_NAME, we had to resort to dumping all the 6 syscall arguments, fix it by providing that info for such syscalls, like 'clone'. Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-dfq1jtrxj8dqvqoeqqpr3slu@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-trace.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) (limited to 'tools') diff --git a/tools/perf/builtin-trace.c b/tools/perf/builtin-trace.c index f8b7bfd..e1988d0 100644 --- a/tools/perf/builtin-trace.c +++ b/tools/perf/builtin-trace.c @@ -611,6 +611,7 @@ static struct syscall_fmt { const char *name; const char *alias; struct syscall_arg_fmt arg[6]; + u8 nr_args; bool errpid; bool timeout; bool hexret; @@ -1169,6 +1170,9 @@ static int syscall__alloc_arg_fmts(struct syscall *sc, int nr_args) { int idx; + if (nr_args == 6 && sc->fmt && sc->fmt->nr_args != 0) + nr_args = sc->fmt->nr_args; + sc->arg_fmt = calloc(nr_args, sizeof(*sc->arg_fmt)); if (sc->arg_fmt == NULL) return -1; @@ -1413,7 +1417,7 @@ static size_t syscall__scnprintf_args(struct syscall *sc, char *bf, size_t size, * may end up not having any args, like with gettid(), so only * print the raw args when we didn't manage to read it. */ - while (arg.idx < 6) { + while (arg.idx < sc->nr_args) { if (arg.mask & bit) goto next_arg; val = syscall_arg__val(&arg, arg.idx); -- cgit v1.1 From c51bdfecd782eec710237c53137e0fefd032d287 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Wed, 19 Jul 2017 15:47:30 -0300 Subject: perf trace: Allow specifying names to syscall arguments formatters For tracepointless syscalls, like clone, otherwise get them from the tracepoint's /format file. Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-ml5qvv1w5k96ghwhxpzzsmm3@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-trace.c | 15 +++++++++++++-- 1 file changed, 13 insertions(+), 2 deletions(-) (limited to 'tools') diff --git a/tools/perf/builtin-trace.c b/tools/perf/builtin-trace.c index e1988d0..6664293 100644 --- a/tools/perf/builtin-trace.c +++ b/tools/perf/builtin-trace.c @@ -604,6 +604,7 @@ static size_t syscall_arg__scnprintf_getrandom_flags(char *bf, size_t size, struct syscall_arg_fmt { size_t (*scnprintf)(char *bf, size_t size, struct syscall_arg *arg); void *parm; + const char *name; bool show_zero; }; @@ -1349,6 +1350,15 @@ unsigned long syscall_arg__val(struct syscall_arg *arg, u8 idx) return val; } +static size_t syscall__scnprintf_name(struct syscall *sc, char *bf, size_t size, + struct syscall_arg *arg) +{ + if (sc->arg_fmt && sc->arg_fmt[arg->idx].name) + return scnprintf(bf, size, "%s: ", sc->arg_fmt[arg->idx].name); + + return scnprintf(bf, size, "arg%d: ", arg->idx); +} + static size_t syscall__scnprintf_val(struct syscall *sc, char *bf, size_t size, struct syscall_arg *arg, unsigned long val) { @@ -1421,8 +1431,9 @@ static size_t syscall__scnprintf_args(struct syscall *sc, char *bf, size_t size, if (arg.mask & bit) goto next_arg; val = syscall_arg__val(&arg, arg.idx); - printed += scnprintf(bf + printed, size - printed, - "%sarg%d: ", printed ? ", " : "", arg.idx); + if (printed) + printed += scnprintf(bf + printed, size - printed, ", "); + printed += syscall__scnprintf_name(sc, bf + printed, size - printed, &arg); printed += syscall__scnprintf_val(sc, bf + printed, size - printed, &arg, val); next_arg: ++arg.idx; -- cgit v1.1 From 450c86c9a3301db5165172bcaced73d036f9c582 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Thu, 20 Jul 2017 10:46:34 -0300 Subject: tools include uapi: Grab a copy of linux/sched.h So that we make sure we have recent enough defines for things such as 'perf trace' system call argument beautifiers. For instance, the 'clone' syscall argument 'flag' needs to use CLONE_NEWCGROUP, and that is not available in RHEL7. Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-81sln0ng4a2lcxrth14vcov4@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/include/uapi/linux/sched.h | 52 ++++++++++++++++++++++++++++++++++++++++ tools/perf/MANIFEST | 1 + tools/perf/check-headers.sh | 1 + 3 files changed, 54 insertions(+) create mode 100644 tools/include/uapi/linux/sched.h (limited to 'tools') diff --git a/tools/include/uapi/linux/sched.h b/tools/include/uapi/linux/sched.h new file mode 100644 index 0000000..e2a6c7b --- /dev/null +++ b/tools/include/uapi/linux/sched.h @@ -0,0 +1,52 @@ +#ifndef _UAPI_LINUX_SCHED_H +#define _UAPI_LINUX_SCHED_H + +/* + * cloning flags: + */ +#define CSIGNAL 0x000000ff /* signal mask to be sent at exit */ +#define CLONE_VM 0x00000100 /* set if VM shared between processes */ +#define CLONE_FS 0x00000200 /* set if fs info shared between processes */ +#define CLONE_FILES 0x00000400 /* set if open files shared between processes */ +#define CLONE_SIGHAND 0x00000800 /* set if signal handlers and blocked signals shared */ +#define CLONE_PTRACE 0x00002000 /* set if we want to let tracing continue on the child too */ +#define CLONE_VFORK 0x00004000 /* set if the parent wants the child to wake it up on mm_release */ +#define CLONE_PARENT 0x00008000 /* set if we want to have the same parent as the cloner */ +#define CLONE_THREAD 0x00010000 /* Same thread group? */ +#define CLONE_NEWNS 0x00020000 /* New mount namespace group */ +#define CLONE_SYSVSEM 0x00040000 /* share system V SEM_UNDO semantics */ +#define CLONE_SETTLS 0x00080000 /* create a new TLS for the child */ +#define CLONE_PARENT_SETTID 0x00100000 /* set the TID in the parent */ +#define CLONE_CHILD_CLEARTID 0x00200000 /* clear the TID in the child */ +#define CLONE_DETACHED 0x00400000 /* Unused, ignored */ +#define CLONE_UNTRACED 0x00800000 /* set if the tracing process can't force CLONE_PTRACE on this clone */ +#define CLONE_CHILD_SETTID 0x01000000 /* set the TID in the child */ +#define CLONE_NEWCGROUP 0x02000000 /* New cgroup namespace */ +#define CLONE_NEWUTS 0x04000000 /* New utsname namespace */ +#define CLONE_NEWIPC 0x08000000 /* New ipc namespace */ +#define CLONE_NEWUSER 0x10000000 /* New user namespace */ +#define CLONE_NEWPID 0x20000000 /* New pid namespace */ +#define CLONE_NEWNET 0x40000000 /* New network namespace */ +#define CLONE_IO 0x80000000 /* Clone io context */ + +/* + * Scheduling policies + */ +#define SCHED_NORMAL 0 +#define SCHED_FIFO 1 +#define SCHED_RR 2 +#define SCHED_BATCH 3 +/* SCHED_ISO: reserved but not implemented yet */ +#define SCHED_IDLE 5 +#define SCHED_DEADLINE 6 + +/* Can be ORed in to make sure the process is reverted back to SCHED_NORMAL on fork */ +#define SCHED_RESET_ON_FORK 0x40000000 + +/* + * For the sched_{set,get}attr() calls + */ +#define SCHED_FLAG_RESET_ON_FORK 0x01 +#define SCHED_FLAG_RECLAIM 0x02 + +#endif /* _UAPI_LINUX_SCHED_H */ diff --git a/tools/perf/MANIFEST b/tools/perf/MANIFEST index a29da46..9f2267b 100644 --- a/tools/perf/MANIFEST +++ b/tools/perf/MANIFEST @@ -78,6 +78,7 @@ tools/include/uapi/linux/fcntl.h tools/include/uapi/linux/hw_breakpoint.h tools/include/uapi/linux/mman.h tools/include/uapi/linux/perf_event.h +tools/include/uapi/linux/sched.h tools/include/uapi/linux/stat.h tools/include/linux/poison.h tools/include/linux/rbtree.h diff --git a/tools/perf/check-headers.sh b/tools/perf/check-headers.sh index 47abd33..9b62958 100755 --- a/tools/perf/check-headers.sh +++ b/tools/perf/check-headers.sh @@ -3,6 +3,7 @@ HEADERS=' include/uapi/linux/fcntl.h include/uapi/linux/perf_event.h +include/uapi/linux/sched.h include/uapi/linux/stat.h include/linux/hash.h include/uapi/linux/hw_breakpoint.h -- cgit v1.1 From 33396a3a6a6b16784291ac19708e3a6cf85db6c5 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Wed, 19 Jul 2017 16:15:17 -0300 Subject: perf trace beauty clone: Beautify syscall arguments Now, syswide tracing, selected entries: # trace -e clone 24417.203 ( 0.158 ms): bash/11323 clone(flags: CHILD_CLEARTID|CHILD_SETTID|0x11, child_stack: 0, parent_tidptr: 0, child_tidptr: 0x7f0778e5c9d0, tls: 0x7f0778e5c700) = 11325 (bash) ? ( ? ): bash/11325 ... [continued]: clone()) = 0 24419.355 ( 0.093 ms): bash/10586 clone(flags: CHILD_CLEARTID|CHILD_SETTID|0x11, child_stack: 0, parent_tidptr: 0, child_tidptr: 0x7f0778e5c9d0, tls: 0x7f0778e5c700) = 11326 (bash) ? ( ? ): bash/11326 ... [continued]: clone()) = 0 24419.744 ( 0.102 ms): bash/11326 clone(flags: CHILD_CLEARTID|CHILD_SETTID|0x11, child_stack: 0, parent_tidptr: 0, child_tidptr: 0x7f0778e5c9d0, tls: 0x7f0778e5c700) = 11327 (bash) ? ( ? ): bash/11327 ... [continued]: clone()) = 0 24420.138 ( 0.105 ms): bash/11327 clone(flags: CHILD_CLEARTID|CHILD_SETTID|0x11, child_stack: 0, parent_tidptr: 0, child_tidptr: 0x7f0778e5c9d0, tls: 0x7f0778e5c700) = 11328 (bash) ? ( ? ): bash/11328 ... [continued]: clone()) = 0 35747.722 ( 0.044 ms): gpg-agent/18087 clone(flags: VM|FS|FILES|SIGHAND|THREAD|SYSVSEM|SETTLS|PARENT_SETTID|CHILD_CLEARTID, child_stack: 0x7ff0755f6ff0, parent_tidptr: 0x7ff0755f79d0, child_tidptr: 0x7ff0755f79d0, tls: 0x7ff0755f7700) = 11329 (gpg-agent) ? ( ? ): gpg-agent/11329 ... [continued]: clone()) = 0 35748.359 ( 0.022 ms): gpg-agent/18087 clone(flags: VM|FS|FILES|SIGHAND|THREAD|SYSVSEM|SETTLS|PARENT_SETTID|CHILD_CLEARTID, child_stack: 0x7ff075df7ff0, parent_tidptr: 0x7ff075df89d0, child_tidptr: 0x7ff075df89d0, tls: 0x7ff075df8700) = 11330 (gpg-agent) ? ( ? ): gpg-agent/11330 ... [continued]: clone()) = 0 35781.422 ( 0.452 ms): NetworkManager/1112 clone(flags: VM|FS|FILES|SIGHAND|THREAD|SYSVSEM|SETTLS|PARENT_SETTID|CHILD_CLEARTID, child_stack: 0x7f2f1fffedb0, parent_tidptr: 0x7f2f1ffff9d0, child_tidptr: 0x7f2f1ffff9d0, tls: 0x7f2f1ffff700) = 11331 (NetworkManager) ? ( ? ): NetworkManager/11331 ... [continued]: clone()) = 0 Need to improve the formatting of the second return, to the child, this cset only focused on the argument formatting. If we trace just one pid: # trace -e clone -p 19863 0.349 ( 0.025 ms): Chrome_IOThrea/19863 clone(flags: VM|FS|FILES|SIGHAND|THREAD|SYSVSEM|SETTLS|PARENT_SETTID|CHILD_CLEARTID, child_stack: 0x7ffb84eaac70, parent_tidptr: 0x7ffb84eab9d0, child_tidptr: 0x7ffb84eab9d0, tls: 0x7ffb84eab700) = 11637 (Chrome_IOThread) 0.392 ( 0.013 ms): Chrome_IOThrea/19863 clone(flags: VM|FS|FILES|SIGHAND|THREAD|SYSVSEM|SETTLS|PARENT_SETTID|CHILD_CLEARTID, child_stack: 0x7ffb664b8c70, parent_tidptr: 0x7ffb664b99d0, child_tidptr: 0x7ffb664b99d0, tls: 0x7ffb664b9700) = 11638 (Chrome_IOThread) 0.573 ( 0.015 ms): Chrome_IOThrea/19863 clone(flags: VM|FS|FILES|SIGHAND|THREAD|SYSVSEM|SETTLS|PARENT_SETTID|CHILD_CLEARTID, child_stack: 0x7ffb6046cc70, parent_tidptr: 0x7ffb6046d9d0, child_tidptr: 0x7ffb6046d9d0, tls: 0x7ffb6046d700) = 11639 (Chrome_IOThread) 0.617 ( 0.014 ms): Chrome_IOThrea/19863 clone(flags: VM|FS|FILES|SIGHAND|THREAD|SYSVSEM|SETTLS|PARENT_SETTID|CHILD_CLEARTID, child_stack: 0x7ffb730dcc70, parent_tidptr: 0x7ffb730dd9d0, child_tidptr: 0x7ffb730dd9d0, tls: 0x7ffb730dd700) = 11640 (Chrome_IOThread) 4.350 ( 0.065 ms): Chrome_IOThrea/19863 clone(flags: VM|FS|FILES|SIGHAND|THREAD|SYSVSEM|SETTLS|PARENT_SETTID|CHILD_CLEARTID, child_stack: 0x7ffb720d9c70, parent_tidptr: 0x7ffb720da9d0, child_tidptr: 0x7ffb720da9d0, tls: 0x7ffb720da700) = 11642 (Chrome_IOThread) 5.642 ( 0.079 ms): Chrome_IOThrea/19863 clone(flags: VM|FS|FILES|SIGHAND|THREAD|SYSVSEM|SETTLS|PARENT_SETTID|CHILD_CLEARTID, child_stack: 0x7ffb718d8c70, parent_tidptr: 0x7ffb718d99d0, child_tidptr: 0x7ffb718d99d0, tls: 0x7ffb718d9700) = 11643 (Chrome_IOThread) ^C# We'll also have to fix the argument ordering in different arches, probably having multiple syscall_fmt entries with each possible order and then use perf_evsel__env_arch() (if dealing with a perf.data file) or the current system info, for live sessions. Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-am068uyubgj83snepolwhbfe@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-trace.c | 7 ++++- tools/perf/trace/beauty/Build | 1 + tools/perf/trace/beauty/beauty.h | 3 +++ tools/perf/trace/beauty/clone.c | 58 ++++++++++++++++++++++++++++++++++++++++ 4 files changed, 68 insertions(+), 1 deletion(-) create mode 100644 tools/perf/trace/beauty/clone.c (limited to 'tools') diff --git a/tools/perf/builtin-trace.c b/tools/perf/builtin-trace.c index 6664293..9e74e67 100644 --- a/tools/perf/builtin-trace.c +++ b/tools/perf/builtin-trace.c @@ -626,7 +626,12 @@ static struct syscall_fmt { .arg = { [0] = { .scnprintf = SCA_HEX, /* brk */ }, }, }, { .name = "clock_gettime", .arg = { [0] = STRARRAY(clk_id, clockid), }, }, - { .name = "clone", .errpid = true, }, + { .name = "clone", .errpid = true, .nr_args = 5, + .arg = { [0] = { .name = "flags", .scnprintf = SCA_CLONE_FLAGS, }, + [1] = { .name = "child_stack", .scnprintf = SCA_HEX, }, + [2] = { .name = "parent_tidptr", .scnprintf = SCA_HEX, }, + [3] = { .name = "child_tidptr", .scnprintf = SCA_HEX, }, + [4] = { .name = "tls", .scnprintf = SCA_HEX, }, }, }, { .name = "close", .arg = { [0] = { .scnprintf = SCA_CLOSE_FD, /* fd */ }, }, }, { .name = "epoll_ctl", diff --git a/tools/perf/trace/beauty/Build b/tools/perf/trace/beauty/Build index c9e215b..eaa1e8e 100644 --- a/tools/perf/trace/beauty/Build +++ b/tools/perf/trace/beauty/Build @@ -1,2 +1,3 @@ +libperf-y += clone.o libperf-y += fcntl.o libperf-y += statx.o diff --git a/tools/perf/trace/beauty/beauty.h b/tools/perf/trace/beauty/beauty.h index f75ef7d..69a5c8a 100644 --- a/tools/perf/trace/beauty/beauty.h +++ b/tools/perf/trace/beauty/beauty.h @@ -66,6 +66,9 @@ size_t syscall_arg__scnprintf_long(char *bf, size_t size, struct syscall_arg *ar size_t syscall_arg__scnprintf_pid(char *bf, size_t size, struct syscall_arg *arg); #define SCA_PID syscall_arg__scnprintf_pid +size_t syscall_arg__scnprintf_clone_flags(char *bf, size_t size, struct syscall_arg *arg); +#define SCA_CLONE_FLAGS syscall_arg__scnprintf_clone_flags + size_t syscall_arg__scnprintf_fcntl_cmd(char *bf, size_t size, struct syscall_arg *arg); #define SCA_FCNTL_CMD syscall_arg__scnprintf_fcntl_cmd diff --git a/tools/perf/trace/beauty/clone.c b/tools/perf/trace/beauty/clone.c new file mode 100644 index 0000000..ab982a7 --- /dev/null +++ b/tools/perf/trace/beauty/clone.c @@ -0,0 +1,58 @@ +/* + * trace/beauty/cone.c + * + * Copyright (C) 2017, Red Hat Inc, Arnaldo Carvalho de Melo + * + * Released under the GPL v2. (and only v2, not any later version) + */ + +#include "trace/beauty/beauty.h" +#include +#include +#include + +static size_t clone__scnprintf_flags(unsigned long flags, char *bf, size_t size) +{ + int printed = 0; + +#define P_FLAG(n) \ + if (flags & CLONE_##n) { \ + printed += scnprintf(bf + printed, size - printed, "%s%s", printed ? "|" : "", #n); \ + flags &= ~CLONE_##n; \ + } + + P_FLAG(VM); + P_FLAG(FS); + P_FLAG(FILES); + P_FLAG(SIGHAND); + P_FLAG(PTRACE); + P_FLAG(VFORK); + P_FLAG(PARENT); + P_FLAG(THREAD); + P_FLAG(NEWNS); + P_FLAG(SYSVSEM); + P_FLAG(SETTLS); + P_FLAG(PARENT_SETTID); + P_FLAG(CHILD_CLEARTID); + P_FLAG(DETACHED); + P_FLAG(UNTRACED); + P_FLAG(CHILD_SETTID); + P_FLAG(NEWCGROUP); + P_FLAG(NEWUTS); + P_FLAG(NEWIPC); + P_FLAG(NEWUSER); + P_FLAG(NEWPID); + P_FLAG(NEWNET); + P_FLAG(IO); +#undef P_FLAG + + if (flags) + printed += scnprintf(bf + printed, size - printed, "%s%#x", printed ? "|" : "", flags); + + return printed; +} + +size_t syscall_arg__scnprintf_clone_flags(char *bf, size_t size, struct syscall_arg *arg) +{ + return clone__scnprintf_flags(arg->val, bf, size); +} -- cgit v1.1 From 15bed2742a8d60210b958963ca1091d3cfc4f332 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Wed, 19 Jul 2017 16:52:59 -0300 Subject: perf trace beauty clone: Suppress unused args according to 'flags' arg The 'parent_tidptr', 'child_tidptr' and 'tls' arguments to the 'clone' syscall are only used when certain flags are set in 'flags', suppress them when those aren't there. E.g: 9886.919 (0.236 ms): fetchmail/19298 clone(flags: CHILD_CLEARTID|CHILD_SETTID|0x11, child_stack: 0, child_tidptr: 0x7fe43f468590) = 19608 (fetchmail) 12876.052 (0.249 ms): qemu-system-x8/21238 clone(flags: VM|FS|FILES|SIGHAND|THREAD|SYSVSEM|SETTLS|PARENT_SETTID|CHILD_CLEARTID, child_stack: 0x7f48117fc770, parent_tidptr: 0x7f48117ff9d0, child_tidptr: 0x7f48117ff9d0, tls: 0x7f48117ff700) = 19611 (qemu-system-x86) 12876.555 (0.048 ms): worker/19611 clone(flags: VM|FS|FILES|SIGHAND|THREAD|SYSVSEM|SETTLS|PARENT_SETTID|CHILD_CLEARTID, child_stack: 0x7f480f7f8770, parent_tidptr: 0x7f480f7fb9d0, child_tidptr: 0x7f480f7fb9d0, tls: 0x7f480f7fb700) = 19612 (worker) 16575.240 (0.469 ms): fetchmail/19298 clone(flags: CHILD_CLEARTID|CHILD_SETTID|0x11, child_stack: 0, child_tidptr: 0x7fe43f468590) = 19613 (fetchmail) 20797.270 (0.335 ms): fetchmail/19298 clone(flags: CHILD_CLEARTID|CHILD_SETTID|0x11, child_stack: 0, child_tidptr: 0x7fe43f468590) = 19614 (fetchmail) 21228.585 (0.501 ms): vim/19519 clone(flags: CHILD_CLEARTID|CHILD_SETTID|0x11, child_stack: 0, child_tidptr: 0x7fbad6ac27d0) = 19615 (vim) 21232.193 (0.137 ms): bash/19615 clone(flags: CHILD_CLEARTID|CHILD_SETTID|0x11, child_stack: 0, child_tidptr: 0x7fad8bff49d0) = 19616 (bash) Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-0um93djul9knf239gwa5mpcb@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/trace/beauty/clone.c | 19 ++++++++++++++++++- 1 file changed, 18 insertions(+), 1 deletion(-) (limited to 'tools') diff --git a/tools/perf/trace/beauty/clone.c b/tools/perf/trace/beauty/clone.c index ab982a7..d64d049 100644 --- a/tools/perf/trace/beauty/clone.c +++ b/tools/perf/trace/beauty/clone.c @@ -54,5 +54,22 @@ static size_t clone__scnprintf_flags(unsigned long flags, char *bf, size_t size) size_t syscall_arg__scnprintf_clone_flags(char *bf, size_t size, struct syscall_arg *arg) { - return clone__scnprintf_flags(arg->val, bf, size); + unsigned long flags = arg->val; + enum syscall_clone_args { + SCC_FLAGS = (1 << 0), + SCC_CHILD_STACK = (1 << 1), + SCC_PARENT_TIDPTR = (1 << 2), + SCC_CHILD_TIDPTR = (1 << 3), + SCC_TLS = (1 << 4), + }; + if (!(flags & CLONE_PARENT_SETTID)) + arg->mask |= SCC_PARENT_TIDPTR; + + if (!(flags & (CLONE_CHILD_SETTID | CLONE_CHILD_CLEARTID))) + arg->mask |= SCC_CHILD_TIDPTR; + + if (!(flags & CLONE_SETTLS)) + arg->mask |= SCC_TLS; + + return clone__scnprintf_flags(flags, bf, size); } -- cgit v1.1 From dd1a50377c92321f78fa4d0601bb4d88d88670ab Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Thu, 20 Jul 2017 11:17:51 -0300 Subject: perf trace: Introduce filter_loop_pids() No change in functionality, just to make clearer that what we want when filtering the tracer pid in a system wide tracing session is to avoid a feedback loop. This also paves the way for a more interesting loop avoidance algorithm, one that tries to figure out if we are in a ssh session, xterm, etc. Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-5fcttc5kdjkcyp9404ezkuy9@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-trace.c | 12 +++++++++++- 1 file changed, 11 insertions(+), 1 deletion(-) (limited to 'tools') diff --git a/tools/perf/builtin-trace.c b/tools/perf/builtin-trace.c index 9e74e67..0ba36f0 100644 --- a/tools/perf/builtin-trace.c +++ b/tools/perf/builtin-trace.c @@ -2239,6 +2239,16 @@ out_enomem: goto out; } +static int trace__set_filter_loop_pids(struct trace *trace) +{ + int nr = 1; + pid_t pids[32] = { + getpid(), + }; + + return perf_evlist__set_filter_pids(trace->evlist, nr, pids); +} + static int trace__run(struct trace *trace, int argc, const char **argv) { struct perf_evlist *evlist = trace->evlist; @@ -2362,7 +2372,7 @@ static int trace__run(struct trace *trace, int argc, const char **argv) if (trace->filter_pids.nr > 0) err = perf_evlist__set_filter_pids(evlist, trace->filter_pids.nr, trace->filter_pids.entries); else if (thread_map__pid(evlist->threads, 0) == -1) - err = perf_evlist__set_filter_pid(evlist, getpid()); + err = trace__set_filter_loop_pids(trace); if (err < 0) goto out_error_mem; -- cgit v1.1 From 082ab9a18e532864d1ceecfb50221df62b1d5a92 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Thu, 20 Jul 2017 11:32:05 -0300 Subject: perf trace: Filter out 'sshd' in the tracer ancestry in syswide tracing Avoiding a loop, so now its quite convenient to ssh to a machine and then simply do: # perf trace To trace all syscalls without causing a loop. This was possible using --filter-pids, i.e. once you noticed the loop, get the sshd pid and add it to --filter-pids, restarting the 'perf trace'. Now to figure out how to do that in a X terminal, the other common scenario, which is way more involved, as there are multiple processes communicating to process terminal activity... Using --filter-pids + '-e \!syscall,names,you,dont,need' may be a good approximation when having to do syswide tracing on your workstation. Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-68rjeao9wnpylla41htk7xps@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-trace.c | 16 +++++++++++++++- 1 file changed, 15 insertions(+), 1 deletion(-) (limited to 'tools') diff --git a/tools/perf/builtin-trace.c b/tools/perf/builtin-trace.c index 0ba36f0..05d24b6 100644 --- a/tools/perf/builtin-trace.c +++ b/tools/perf/builtin-trace.c @@ -2241,10 +2241,24 @@ out_enomem: static int trace__set_filter_loop_pids(struct trace *trace) { - int nr = 1; + unsigned int nr = 1; pid_t pids[32] = { getpid(), }; + struct thread *thread = machine__find_thread(trace->host, pids[0], pids[0]); + + while (thread && nr < ARRAY_SIZE(pids)) { + struct thread *parent = machine__find_thread(trace->host, thread->ppid, thread->ppid); + + if (parent == NULL) + break; + + if (!strcmp(thread__comm_str(parent), "sshd")) { + pids[nr++] = parent->tid; + break; + } + thread = parent; + } return perf_evlist__set_filter_pids(trace->evlist, nr, pids); } -- cgit v1.1 From 8e99b6d4533cf3f49dcd813155a513a5b572baef Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Thu, 20 Jul 2017 15:27:39 -0300 Subject: tools include: Adopt strstarts() from the kernel Replacing prefixcmp(), same purpose, inverted result, so standardize on the kernel variant, to reduce silly differences among tools/ and the kernel sources, making it easier for people to work in both codebases. And then doing: if (strstarts(option, "no-")) Looks clearer than doing: if (!prefixcmp(option, "no-")) To figure out if option starts witn "no-". Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Josh Poimboeuf Cc: Namhyung Kim Cc: Rusty Russell Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-kaei42gi7lpa8subwtv7eug8@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/include/linux/string.h | 12 ++++++++++-- tools/lib/string.c | 9 --------- tools/lib/subcmd/help.c | 2 +- tools/lib/subcmd/parse-options.c | 18 +++++++++--------- tools/perf/builtin-config.c | 3 ++- tools/perf/builtin-ftrace.c | 2 +- tools/perf/builtin-help.c | 6 +++--- tools/perf/perf.c | 16 ++++++++-------- tools/perf/ui/browser.c | 3 ++- tools/perf/ui/browsers/annotate.c | 3 ++- tools/perf/ui/stdio/hist.c | 3 ++- tools/perf/util/bpf-loader.c | 2 +- tools/perf/util/callchain.c | 2 +- tools/perf/util/config.c | 13 +++++++------ tools/perf/util/llvm-utils.c | 2 +- 15 files changed, 50 insertions(+), 46 deletions(-) (limited to 'tools') diff --git a/tools/include/linux/string.h b/tools/include/linux/string.h index d62b56c..a30fad5 100644 --- a/tools/include/linux/string.h +++ b/tools/include/linux/string.h @@ -1,8 +1,8 @@ #ifndef _TOOLS_LINUX_STRING_H_ #define _TOOLS_LINUX_STRING_H_ - #include /* for size_t */ +#include void *memdup(const void *src, size_t len); @@ -18,6 +18,14 @@ extern size_t strlcpy(char *dest, const char *src, size_t size); char *str_error_r(int errnum, char *buf, size_t buflen); -int prefixcmp(const char *str, const char *prefix); +/** + * strstarts - does @str start with @prefix? + * @str: string to examine + * @prefix: prefix to look for. + */ +static inline bool strstarts(const char *str, const char *prefix) +{ + return strncmp(str, prefix, strlen(prefix)) == 0; +} #endif /* _LINUX_STRING_H_ */ diff --git a/tools/lib/string.c b/tools/lib/string.c index 8e678af..bd239bc 100644 --- a/tools/lib/string.c +++ b/tools/lib/string.c @@ -87,12 +87,3 @@ size_t __weak strlcpy(char *dest, const char *src, size_t size) } return ret; } - -int prefixcmp(const char *str, const char *prefix) -{ - for (; ; str++, prefix++) - if (!*prefix) - return 0; - else if (*str != *prefix) - return (unsigned char)*prefix - (unsigned char)*str; -} diff --git a/tools/lib/subcmd/help.c b/tools/lib/subcmd/help.c index ba970a7..0310520 100644 --- a/tools/lib/subcmd/help.c +++ b/tools/lib/subcmd/help.c @@ -171,7 +171,7 @@ static void list_commands_in_dir(struct cmdnames *cmds, while ((de = readdir(dir)) != NULL) { int entlen; - if (prefixcmp(de->d_name, prefix)) + if (!strstarts(de->d_name, prefix)) continue; astrcat(&buf, de->d_name); diff --git a/tools/lib/subcmd/parse-options.c b/tools/lib/subcmd/parse-options.c index 359bfa7..2bd6fd0 100644 --- a/tools/lib/subcmd/parse-options.c +++ b/tools/lib/subcmd/parse-options.c @@ -368,7 +368,7 @@ retry: return 0; } if (!rest) { - if (!prefixcmp(options->long_name, "no-")) { + if (strstarts(options->long_name, "no-")) { /* * The long name itself starts with "no-", so * accept the option without "no-" so that users @@ -381,7 +381,7 @@ retry: goto match; } /* Abbreviated case */ - if (!prefixcmp(options->long_name + 3, arg)) { + if (strstarts(options->long_name + 3, arg)) { flags |= OPT_UNSET; goto is_abbreviated; } @@ -406,7 +406,7 @@ is_abbreviated: continue; } /* negated and abbreviated very much? */ - if (!prefixcmp("no-", arg)) { + if (strstarts("no-", arg)) { flags |= OPT_UNSET; goto is_abbreviated; } @@ -416,7 +416,7 @@ is_abbreviated: flags |= OPT_UNSET; rest = skip_prefix(arg + 3, options->long_name); /* abbreviated and negated? */ - if (!rest && !prefixcmp(options->long_name, arg + 3)) + if (!rest && strstarts(options->long_name, arg + 3)) goto is_abbreviated; if (!rest) continue; @@ -456,7 +456,7 @@ static void check_typos(const char *arg, const struct option *options) if (strlen(arg) < 3) return; - if (!prefixcmp(arg, "no-")) { + if (strstarts(arg, "no-")) { fprintf(stderr, " Error: did you mean `--%s` (with two dashes ?)", arg); exit(129); } @@ -464,7 +464,7 @@ static void check_typos(const char *arg, const struct option *options) for (; options->type != OPTION_END; options++) { if (!options->long_name) continue; - if (!prefixcmp(options->long_name, arg)) { + if (strstarts(options->long_name, arg)) { fprintf(stderr, " Error: did you mean `--%s` (with two dashes ?)", arg); exit(129); } @@ -933,10 +933,10 @@ opt: if (opts->long_name == NULL) continue; - if (!prefixcmp(opts->long_name, optstr)) + if (strstarts(opts->long_name, optstr)) print_option_help(opts, 0); - if (!prefixcmp("no-", optstr) && - !prefixcmp(opts->long_name, optstr + 3)) + if (strstarts("no-", optstr) && + strstarts(opts->long_name, optstr + 3)) print_option_help(opts, 0); } diff --git a/tools/perf/builtin-config.c b/tools/perf/builtin-config.c index ece4558..3ddcc6e 100644 --- a/tools/perf/builtin-config.c +++ b/tools/perf/builtin-config.c @@ -13,6 +13,7 @@ #include "util/util.h" #include "util/debug.h" #include "util/config.h" +#include static bool use_system_config, use_user_config; @@ -79,7 +80,7 @@ static int show_spec_config(struct perf_config_set *set, const char *var) return -1; perf_config_items__for_each_entry(&set->sections, section) { - if (prefixcmp(var, section->name) != 0) + if (!strstarts(var, section->name)) continue; perf_config_items__for_each_entry(§ion->items, item) { diff --git a/tools/perf/builtin-ftrace.c b/tools/perf/builtin-ftrace.c index dd26c62..25a42ac 100644 --- a/tools/perf/builtin-ftrace.c +++ b/tools/perf/builtin-ftrace.c @@ -381,7 +381,7 @@ static int perf_ftrace_config(const char *var, const char *value, void *cb) { struct perf_ftrace *ftrace = cb; - if (prefixcmp(var, "ftrace.")) + if (!strstarts(var, "ftrace.")) return 0; if (strcmp(var, "ftrace.tracer")) diff --git a/tools/perf/builtin-help.c b/tools/perf/builtin-help.c index 530a7f2..dbe4e41 100644 --- a/tools/perf/builtin-help.c +++ b/tools/perf/builtin-help.c @@ -90,7 +90,7 @@ static int check_emacsclient_version(void) */ finish_command(&ec_process); - if (prefixcmp(buffer.buf, "emacsclient")) { + if (!strstarts(buffer.buf, "emacsclient")) { fprintf(stderr, "Failed to parse emacsclient version.\n"); goto out; } @@ -283,7 +283,7 @@ static int perf_help_config(const char *var, const char *value, void *cb) add_man_viewer(value); return 0; } - if (!prefixcmp(var, "man.")) + if (!strstarts(var, "man.")) return add_man_viewer_info(var, value); return 0; @@ -313,7 +313,7 @@ static const char *cmd_to_page(const char *perf_cmd) if (!perf_cmd) return "perf"; - else if (!prefixcmp(perf_cmd, "perf")) + else if (!strstarts(perf_cmd, "perf")) return perf_cmd; return asprintf(&s, "perf-%s", perf_cmd) < 0 ? NULL : s; diff --git a/tools/perf/perf.c b/tools/perf/perf.c index 628a5e4..e0279ba 100644 --- a/tools/perf/perf.c +++ b/tools/perf/perf.c @@ -89,7 +89,7 @@ struct pager_config { static int pager_command_config(const char *var, const char *value, void *data) { struct pager_config *c = data; - if (!prefixcmp(var, "pager.") && !strcmp(var + 6, c->cmd)) + if (strstarts(var, "pager.") && !strcmp(var + 6, c->cmd)) c->val = perf_config_bool(var, value); return 0; } @@ -108,9 +108,9 @@ static int check_pager_config(const char *cmd) static int browser_command_config(const char *var, const char *value, void *data) { struct pager_config *c = data; - if (!prefixcmp(var, "tui.") && !strcmp(var + 4, c->cmd)) + if (strstarts(var, "tui.") && !strcmp(var + 4, c->cmd)) c->val = perf_config_bool(var, value); - if (!prefixcmp(var, "gtk.") && !strcmp(var + 4, c->cmd)) + if (strstarts(var, "gtk.") && !strcmp(var + 4, c->cmd)) c->val = perf_config_bool(var, value) ? 2 : 0; return 0; } @@ -192,7 +192,7 @@ static int handle_options(const char ***argv, int *argc, int *envchanged) /* * Check remaining flags. */ - if (!prefixcmp(cmd, CMD_EXEC_PATH)) { + if (strstarts(cmd, CMD_EXEC_PATH)) { cmd += strlen(CMD_EXEC_PATH); if (*cmd == '=') set_argv_exec_path(cmd + 1); @@ -229,7 +229,7 @@ static int handle_options(const char ***argv, int *argc, int *envchanged) *envchanged = 1; (*argv)++; (*argc)--; - } else if (!prefixcmp(cmd, CMD_DEBUGFS_DIR)) { + } else if (strstarts(cmd, CMD_DEBUGFS_DIR)) { tracing_path_set(cmd + strlen(CMD_DEBUGFS_DIR)); fprintf(stderr, "dir: %s\n", tracing_path); if (envchanged) @@ -470,14 +470,14 @@ int main(int argc, const char **argv) * So we just directly call the internal command handler, and * die if that one cannot handle it. */ - if (!prefixcmp(cmd, "perf-")) { + if (strstarts(cmd, "perf-")) { cmd += 5; argv[0] = cmd; handle_internal_command(argc, argv); fprintf(stderr, "cannot handle %s internally", cmd); goto out; } - if (!prefixcmp(cmd, "trace")) { + if (strstarts(cmd, "trace")) { #ifdef HAVE_LIBAUDIT_SUPPORT setup_path(); argv[0] = "trace"; @@ -495,7 +495,7 @@ int main(int argc, const char **argv) commit_pager_choice(); if (argc > 0) { - if (!prefixcmp(argv[0], "--")) + if (strstarts(argv[0], "--")) argv[0] += 2; } else { /* The user didn't specify a command; give them help */ diff --git a/tools/perf/ui/browser.c b/tools/perf/ui/browser.c index f73f3f1..d0c2007 100644 --- a/tools/perf/ui/browser.c +++ b/tools/perf/ui/browser.c @@ -8,6 +8,7 @@ #include #include #include +#include #include #include #include "browser.h" @@ -563,7 +564,7 @@ static int ui_browser__color_config(const char *var, const char *value, int i; /* same dir for all commands */ - if (prefixcmp(var, "colors.") != 0) + if (!strstarts(var, "colors.") != 0) return 0; for (i = 0; ui_browser__colorsets[i].name != NULL; ++i) { diff --git a/tools/perf/ui/browsers/annotate.c b/tools/perf/ui/browsers/annotate.c index 8d3f6f5..6794a8b 100644 --- a/tools/perf/ui/browsers/annotate.c +++ b/tools/perf/ui/browsers/annotate.c @@ -13,6 +13,7 @@ #include #include #include +#include #include struct disasm_line_samples { @@ -1198,7 +1199,7 @@ static int annotate__config(const char *var, const char *value, struct annotate_config *cfg; const char *name; - if (prefixcmp(var, "annotate.") != 0) + if (!strstarts(var, "annotate.")) return 0; name = var + 9; diff --git a/tools/perf/ui/stdio/hist.c b/tools/perf/ui/stdio/hist.c index 2df8eb1..5c95b83 100644 --- a/tools/perf/ui/stdio/hist.c +++ b/tools/perf/ui/stdio/hist.c @@ -1,4 +1,5 @@ #include +#include #include "../../util/util.h" #include "../../util/hist.h" @@ -292,7 +293,7 @@ static size_t callchain__fprintf_graph(FILE *fp, struct rb_root *root, * displayed twice. */ if (!i++ && field_order == NULL && - sort_order && !prefixcmp(sort_order, "sym")) + sort_order && strstarts(sort_order, "sym")) continue; if (!printed) { diff --git a/tools/perf/util/bpf-loader.c b/tools/perf/util/bpf-loader.c index 4bd2d1d..4a1264c 100644 --- a/tools/perf/util/bpf-loader.c +++ b/tools/perf/util/bpf-loader.c @@ -1246,7 +1246,7 @@ int bpf__config_obj(struct bpf_object *obj, if (!obj || !term || !term->config) return -EINVAL; - if (!prefixcmp(term->config, "map:")) { + if (strstarts(term->config, "map:")) { key_scan_pos = sizeof("map:") - 1; err = bpf__obj_config_map(obj, term, evlist, &key_scan_pos); goto out; diff --git a/tools/perf/util/callchain.c b/tools/perf/util/callchain.c index 22d413a..02130e2 100644 --- a/tools/perf/util/callchain.c +++ b/tools/perf/util/callchain.c @@ -304,7 +304,7 @@ int perf_callchain_config(const char *var, const char *value) { char *endptr; - if (prefixcmp(var, "call-graph.")) + if (!strstarts(var, "call-graph.")) return 0; var += sizeof("call-graph.") - 1; diff --git a/tools/perf/util/config.c b/tools/perf/util/config.c index 31a7dea..bc75596 100644 --- a/tools/perf/util/config.c +++ b/tools/perf/util/config.c @@ -19,6 +19,7 @@ #include #include #include +#include #include "sane_ctype.h" @@ -433,22 +434,22 @@ static int perf_ui_config(const char *var, const char *value) int perf_default_config(const char *var, const char *value, void *dummy __maybe_unused) { - if (!prefixcmp(var, "core.")) + if (strstarts(var, "core.")) return perf_default_core_config(var, value); - if (!prefixcmp(var, "hist.")) + if (strstarts(var, "hist.")) return perf_hist_config(var, value); - if (!prefixcmp(var, "ui.")) + if (strstarts(var, "ui.")) return perf_ui_config(var, value); - if (!prefixcmp(var, "call-graph.")) + if (strstarts(var, "call-graph.")) return perf_callchain_config(var, value); - if (!prefixcmp(var, "llvm.")) + if (strstarts(var, "llvm.")) return perf_llvm_config(var, value); - if (!prefixcmp(var, "buildid.")) + if (strstarts(var, "buildid.")) return perf_buildid_config(var, value); /* Add other config variables here. */ diff --git a/tools/perf/util/llvm-utils.c b/tools/perf/util/llvm-utils.c index c6a15f2..209b0c8 100644 --- a/tools/perf/util/llvm-utils.c +++ b/tools/perf/util/llvm-utils.c @@ -33,7 +33,7 @@ struct llvm_param llvm_param = { int perf_llvm_config(const char *var, const char *value) { - if (prefixcmp(var, "llvm.")) + if (!strstarts(var, "llvm.")) return 0; var += sizeof("llvm.") - 1; -- cgit v1.1 From b99e4850df86dfd9dc2cf3f3494e403ff8b46876 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Thu, 20 Jul 2017 15:35:33 -0300 Subject: tools lib: Update copy of strtobool from the kernel sources Getting support for "on", "off" introduced in a81a5a17d44b ("lib: add "on"/"off" support to kstrtobool") and making it check for NULL, introduced in ef951599074b ("lib: move strtobool() to kstrtobool()"). Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Cc: Kees Cook Link: http://lkml.kernel.org/n/tip-mu8ghin4rklacmmubzwv8td7@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/lib/string.c | 32 +++++++++++++++++++++++++------- 1 file changed, 25 insertions(+), 7 deletions(-) (limited to 'tools') diff --git a/tools/lib/string.c b/tools/lib/string.c index bd239bc..a4246f1 100644 --- a/tools/lib/string.c +++ b/tools/lib/string.c @@ -39,27 +39,45 @@ void *memdup(const void *src, size_t len) * @s: input string * @res: result * - * This routine returns 0 iff the first character is one of 'Yy1Nn0'. - * Otherwise it will return -EINVAL. Value pointed to by res is - * updated upon finding a match. + * This routine returns 0 iff the first character is one of 'Yy1Nn0', or + * [oO][NnFf] for "on" and "off". Otherwise it will return -EINVAL. Value + * pointed to by res is updated upon finding a match. */ int strtobool(const char *s, bool *res) { + if (!s) + return -EINVAL; + switch (s[0]) { case 'y': case 'Y': case '1': *res = true; - break; + return 0; case 'n': case 'N': case '0': *res = false; - break; + return 0; + case 'o': + case 'O': + switch (s[1]) { + case 'n': + case 'N': + *res = true; + return 0; + case 'f': + case 'F': + *res = false; + return 0; + default: + break; + } default: - return -EINVAL; + break; } - return 0; + + return -EINVAL; } /** -- cgit v1.1 From 896bccd3cb8d95cbc565687715516009c5169e71 Mon Sep 17 00:00:00 2001 From: Taeung Song Date: Thu, 20 Jul 2017 06:36:45 +0900 Subject: perf annotate: Introduce struct sym_hist_entry struct sym_hist has addr[] but it should have not only number of samples but also the sample period. So use new struct symhist_entry to pave the way to have that. Committer notes: This initial patch will only introduce the struct sym_hist_entry and use only the nr_samples member, which makes the code clearer and paves the way to save the period as well. Signed-off-by: Taeung Song Suggested-by: Arnaldo Carvalho de Melo Cc: Jiri Olsa Cc: Namhyung Kim Link: http://lkml.kernel.org/r/1500500205-16553-1-git-send-email-treeze.taeung@gmail.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/ui/browsers/annotate.c | 6 +++--- tools/perf/ui/gtk/annotate.c | 4 ++-- tools/perf/util/annotate.c | 45 ++++++++++++++++++++------------------- tools/perf/util/annotate.h | 9 ++++++-- 4 files changed, 35 insertions(+), 29 deletions(-) (limited to 'tools') diff --git a/tools/perf/ui/browsers/annotate.c b/tools/perf/ui/browsers/annotate.c index 6794a8b..dbe4e63 100644 --- a/tools/perf/ui/browsers/annotate.c +++ b/tools/perf/ui/browsers/annotate.c @@ -450,14 +450,14 @@ static void annotate_browser__calc_percent(struct annotate_browser *browser, next = disasm__get_next_ip_line(¬es->src->source, pos); for (i = 0; i < browser->nr_events; i++) { - u64 nr_samples; + struct sym_hist_entry sample; bpos->samples[i].percent = disasm__calc_percent(notes, evsel->idx + i, pos->offset, next ? next->offset : len, - &path, &nr_samples); - bpos->samples[i].nr = nr_samples; + &path, &sample); + bpos->samples[i].nr = sample.nr_samples; if (max_percent < bpos->samples[i].percent) max_percent = bpos->samples[i].percent; diff --git a/tools/perf/ui/gtk/annotate.c b/tools/perf/ui/gtk/annotate.c index 87e3760..d736fd5 100644 --- a/tools/perf/ui/gtk/annotate.c +++ b/tools/perf/ui/gtk/annotate.c @@ -34,10 +34,10 @@ static int perf_gtk__get_percent(char *buf, size_t size, struct symbol *sym, return 0; symhist = annotation__histogram(symbol__annotation(sym), evidx); - if (!symbol_conf.event_group && !symhist->addr[dl->offset]) + if (!symbol_conf.event_group && !symhist->addr[dl->offset].nr_samples) return 0; - percent = 100.0 * symhist->addr[dl->offset] / symhist->sum; + percent = 100.0 * symhist->addr[dl->offset].nr_samples / symhist->sum; markup = perf_gtk__get_percent_color(percent); if (markup) diff --git a/tools/perf/util/annotate.c b/tools/perf/util/annotate.c index 1742510..c382955 100644 --- a/tools/perf/util/annotate.c +++ b/tools/perf/util/annotate.c @@ -610,10 +610,10 @@ int symbol__alloc_hist(struct symbol *sym) size_t sizeof_sym_hist; /* Check for overflow when calculating sizeof_sym_hist */ - if (size > (SIZE_MAX - sizeof(struct sym_hist)) / sizeof(u64)) + if (size > (SIZE_MAX - sizeof(struct sym_hist)) / sizeof(struct sym_hist_entry)) return -1; - sizeof_sym_hist = (sizeof(struct sym_hist) + size * sizeof(u64)); + sizeof_sym_hist = (sizeof(struct sym_hist) + size * sizeof(struct sym_hist_entry)); /* Check for overflow in zalloc argument */ if (sizeof_sym_hist > (SIZE_MAX - sizeof(*notes->src)) @@ -714,11 +714,11 @@ static int __symbol__inc_addr_samples(struct symbol *sym, struct map *map, offset = addr - sym->start; h = annotation__histogram(notes, evidx); h->sum++; - h->addr[offset]++; + h->addr[offset].nr_samples++; pr_debug3("%#" PRIx64 " %s: period++ [addr: %#" PRIx64 ", %#" PRIx64 ", evidx=%d] => %" PRIu64 "\n", sym->start, sym->name, - addr, addr - sym->start, evidx, h->addr[offset]); + addr, addr - sym->start, evidx, h->addr[offset].nr_samples); return 0; } @@ -928,11 +928,12 @@ struct disasm_line *disasm__get_next_ip_line(struct list_head *head, struct disa } double disasm__calc_percent(struct annotation *notes, int evidx, s64 offset, - s64 end, const char **path, u64 *nr_samples) + s64 end, const char **path, struct sym_hist_entry *sample) { struct source_line *src_line = notes->src->lines; double percent = 0.0; - *nr_samples = 0; + + sample->nr_samples = 0; if (src_line) { size_t sizeof_src_line = sizeof(*src_line) + @@ -946,7 +947,7 @@ double disasm__calc_percent(struct annotation *notes, int evidx, s64 offset, *path = src_line->path; percent += src_line->samples[evidx].percent; - *nr_samples += src_line->samples[evidx].nr; + sample->nr_samples += src_line->samples[evidx].nr; offset++; } } else { @@ -954,10 +955,10 @@ double disasm__calc_percent(struct annotation *notes, int evidx, s64 offset, unsigned int hits = 0; while (offset < end) - hits += h->addr[offset++]; + hits += h->addr[offset++].nr_samples; if (h->sum) { - *nr_samples = hits; + sample->nr_samples = hits; percent = 100.0 * hits / h->sum; } } @@ -1057,10 +1058,10 @@ static int disasm_line__print(struct disasm_line *dl, struct symbol *sym, u64 st if (dl->offset != -1) { const char *path = NULL; - u64 nr_samples; double percent, max_percent = 0.0; double *ppercents = &percent; - u64 *psamples = &nr_samples; + struct sym_hist_entry sample; + struct sym_hist_entry *psamples = &sample; int i, nr_percent = 1; const char *color; struct annotation *notes = symbol__annotation(sym); @@ -1074,7 +1075,7 @@ static int disasm_line__print(struct disasm_line *dl, struct symbol *sym, u64 st if (perf_evsel__is_group_event(evsel)) { nr_percent = evsel->nr_members; ppercents = calloc(nr_percent, sizeof(double)); - psamples = calloc(nr_percent, sizeof(u64)); + psamples = calloc(nr_percent, sizeof(struct sym_hist_entry)); if (ppercents == NULL || psamples == NULL) { return -1; } @@ -1085,10 +1086,10 @@ static int disasm_line__print(struct disasm_line *dl, struct symbol *sym, u64 st notes->src->lines ? i : evsel->idx + i, offset, next ? next->offset : (s64) len, - &path, &nr_samples); + &path, &sample); ppercents[i] = percent; - psamples[i] = nr_samples; + psamples[i] = sample; if (percent > max_percent) max_percent = percent; } @@ -1126,12 +1127,12 @@ static int disasm_line__print(struct disasm_line *dl, struct symbol *sym, u64 st for (i = 0; i < nr_percent; i++) { percent = ppercents[i]; - nr_samples = psamples[i]; + sample = psamples[i]; color = get_percent_color(percent); if (symbol_conf.show_total_period) color_fprintf(stdout, color, " %7" PRIu64, - nr_samples); + sample.nr_samples); else color_fprintf(stdout, color, " %7.2f", percent); } @@ -1147,7 +1148,7 @@ static int disasm_line__print(struct disasm_line *dl, struct symbol *sym, u64 st if (ppercents != &percent) free(ppercents); - if (psamples != &nr_samples) + if (psamples != &sample) free(psamples); } else if (max_lines && printed >= max_lines) @@ -1702,7 +1703,7 @@ static int symbol__get_source_line(struct symbol *sym, struct map *map, double percent = 0.0; h = annotation__histogram(notes, evidx + k); - nr_samples = h->addr[i]; + nr_samples = h->addr[i].nr_samples; if (h->sum) percent = 100.0 * nr_samples / h->sum; @@ -1773,9 +1774,9 @@ static void symbol__annotate_hits(struct symbol *sym, struct perf_evsel *evsel) u64 len = symbol__size(sym), offset; for (offset = 0; offset < len; ++offset) - if (h->addr[offset] != 0) + if (h->addr[offset].nr_samples != 0) printf("%*" PRIx64 ": %" PRIu64 "\n", BITS_PER_LONG / 2, - sym->start + offset, h->addr[offset]); + sym->start + offset, h->addr[offset].nr_samples); printf("%*s: %" PRIu64 "\n", BITS_PER_LONG / 2, "h->sum", h->sum); } @@ -1878,8 +1879,8 @@ void symbol__annotate_decay_histogram(struct symbol *sym, int evidx) h->sum = 0; for (offset = 0; offset < len; ++offset) { - h->addr[offset] = h->addr[offset] * 7 / 8; - h->sum += h->addr[offset]; + h->addr[offset].nr_samples = h->addr[offset].nr_samples * 7 / 8; + h->sum += h->addr[offset].nr_samples; } } diff --git a/tools/perf/util/annotate.h b/tools/perf/util/annotate.h index bac698d..3a17663 100644 --- a/tools/perf/util/annotate.h +++ b/tools/perf/util/annotate.h @@ -74,16 +74,21 @@ static inline bool disasm_line__has_offset(const struct disasm_line *dl) return dl->ops.target.offset_avail; } +struct sym_hist_entry { + u64 nr_samples; + u64 period; +}; + void disasm_line__free(struct disasm_line *dl); struct disasm_line *disasm__get_next_ip_line(struct list_head *head, struct disasm_line *pos); int disasm_line__scnprintf(struct disasm_line *dl, char *bf, size_t size, bool raw); size_t disasm__fprintf(struct list_head *head, FILE *fp); double disasm__calc_percent(struct annotation *notes, int evidx, s64 offset, - s64 end, const char **path, u64 *nr_samples); + s64 end, const char **path, struct sym_hist_entry *sample); struct sym_hist { u64 sum; - u64 addr[0]; + struct sym_hist_entry addr[0]; }; struct cyc_hist { -- cgit v1.1 From 8158683da3d30e0346275702a8e08f2b22726c45 Mon Sep 17 00:00:00 2001 From: Taeung Song Date: Thu, 20 Jul 2017 06:36:51 +0900 Subject: perf annotate: Rename 'sum' to 'nr_samples' in struct sym_hist To make it more clear that it is the sum of all the nr_samples fields in the addr[] entries, i.e.: sym_hist->nr_samples = sum(sym_hist->addr[0 .. symbol__size(sym)]->nr_samples) Committer notes: Taeung had renamed it to total_samples, but using nr_samples, as in the added explanation above, looks clearer and establishes the direct connection, making clear it is about the _number_ of samples. Signed-off-by: Taeung Song Cc: Jiri Olsa Cc: Namhyung Kim Link: http://lkml.kernel.org/r/1500500211-16599-1-git-send-email-treeze.taeung@gmail.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/ui/gtk/annotate.c | 2 +- tools/perf/util/annotate.c | 26 +++++++++++++------------- tools/perf/util/annotate.h | 2 +- 3 files changed, 15 insertions(+), 15 deletions(-) (limited to 'tools') diff --git a/tools/perf/ui/gtk/annotate.c b/tools/perf/ui/gtk/annotate.c index d736fd5..0217619 100644 --- a/tools/perf/ui/gtk/annotate.c +++ b/tools/perf/ui/gtk/annotate.c @@ -37,7 +37,7 @@ static int perf_gtk__get_percent(char *buf, size_t size, struct symbol *sym, if (!symbol_conf.event_group && !symhist->addr[dl->offset].nr_samples) return 0; - percent = 100.0 * symhist->addr[dl->offset].nr_samples / symhist->sum; + percent = 100.0 * symhist->addr[dl->offset].nr_samples / symhist->nr_samples; markup = perf_gtk__get_percent_color(percent); if (markup) diff --git a/tools/perf/util/annotate.c b/tools/perf/util/annotate.c index c382955..58c6b63 100644 --- a/tools/perf/util/annotate.c +++ b/tools/perf/util/annotate.c @@ -713,7 +713,7 @@ static int __symbol__inc_addr_samples(struct symbol *sym, struct map *map, offset = addr - sym->start; h = annotation__histogram(notes, evidx); - h->sum++; + h->nr_samples++; h->addr[offset].nr_samples++; pr_debug3("%#" PRIx64 " %s: period++ [addr: %#" PRIx64 ", %#" PRIx64 @@ -957,9 +957,9 @@ double disasm__calc_percent(struct annotation *notes, int evidx, s64 offset, while (offset < end) hits += h->addr[offset++].nr_samples; - if (h->sum) { + if (h->nr_samples) { sample->nr_samples = hits; - percent = 100.0 * hits / h->sum; + percent = 100.0 * hits / h->nr_samples; } } @@ -1672,19 +1672,19 @@ static int symbol__get_source_line(struct symbol *sym, struct map *map, struct sym_hist *h = annotation__histogram(notes, evidx); struct rb_root tmp_root = RB_ROOT; int nr_pcnt = 1; - u64 h_sum = h->sum; + u64 nr_samples = h->nr_samples; size_t sizeof_src_line = sizeof(struct source_line); if (perf_evsel__is_group_event(evsel)) { for (i = 1; i < evsel->nr_members; i++) { h = annotation__histogram(notes, evidx + i); - h_sum += h->sum; + nr_samples += h->nr_samples; } nr_pcnt = evsel->nr_members; sizeof_src_line += (nr_pcnt - 1) * sizeof(src_line->samples); } - if (!h_sum) + if (!nr_samples) return 0; src_line = notes->src->lines = calloc(len, sizeof_src_line); @@ -1694,7 +1694,7 @@ static int symbol__get_source_line(struct symbol *sym, struct map *map, start = map__rip_2objdump(map, sym->start); for (i = 0; i < len; i++) { - u64 offset, nr_samples; + u64 offset; double percent_max = 0.0; src_line->nr_pcnt = nr_pcnt; @@ -1704,8 +1704,8 @@ static int symbol__get_source_line(struct symbol *sym, struct map *map, h = annotation__histogram(notes, evidx + k); nr_samples = h->addr[i].nr_samples; - if (h->sum) - percent = 100.0 * nr_samples / h->sum; + if (h->nr_samples) + percent = 100.0 * nr_samples / h->nr_samples; if (percent > percent_max) percent_max = percent; @@ -1777,7 +1777,7 @@ static void symbol__annotate_hits(struct symbol *sym, struct perf_evsel *evsel) if (h->addr[offset].nr_samples != 0) printf("%*" PRIx64 ": %" PRIu64 "\n", BITS_PER_LONG / 2, sym->start + offset, h->addr[offset].nr_samples); - printf("%*s: %" PRIu64 "\n", BITS_PER_LONG / 2, "h->sum", h->sum); + printf("%*s: %" PRIu64 "\n", BITS_PER_LONG / 2, "h->nr_samples", h->nr_samples); } int symbol__annotate_printf(struct symbol *sym, struct map *map, @@ -1813,7 +1813,7 @@ int symbol__annotate_printf(struct symbol *sym, struct map *map, width *= evsel->nr_members; graph_dotted_len = printf(" %-*.*s| Source code & Disassembly of %s for %s (%" PRIu64 " samples)\n", - width, width, "Percent", d_filename, evsel_name, h->sum); + width, width, "Percent", d_filename, evsel_name, h->nr_samples); printf("%-*.*s----\n", graph_dotted_len, graph_dotted_len, graph_dotted_line); @@ -1877,10 +1877,10 @@ void symbol__annotate_decay_histogram(struct symbol *sym, int evidx) struct sym_hist *h = annotation__histogram(notes, evidx); int len = symbol__size(sym), offset; - h->sum = 0; + h->nr_samples = 0; for (offset = 0; offset < len; ++offset) { h->addr[offset].nr_samples = h->addr[offset].nr_samples * 7 / 8; - h->sum += h->addr[offset].nr_samples; + h->nr_samples += h->addr[offset].nr_samples; } } diff --git a/tools/perf/util/annotate.h b/tools/perf/util/annotate.h index 3a17663..e8c246e 100644 --- a/tools/perf/util/annotate.h +++ b/tools/perf/util/annotate.h @@ -87,7 +87,7 @@ double disasm__calc_percent(struct annotation *notes, int evidx, s64 offset, s64 end, const char **path, struct sym_hist_entry *sample); struct sym_hist { - u64 sum; + u64 nr_samples; struct sym_hist_entry addr[0]; }; -- cgit v1.1 From bab89f6aed7e745893e009014354d0caaf62acf7 Mon Sep 17 00:00:00 2001 From: Taeung Song Date: Thu, 20 Jul 2017 16:28:53 -0300 Subject: perf hists: Pass perf_sample to __symbol__inc_addr_samples() To pave the way to use perf_sample fields in the annotate code, storing sample->period in sym_hist->addr->period and its sum in sym_hist->period. Signed-off-by: Taeung Song Cc: Jiri Olsa Cc: Namhyung Kim Link: http://lkml.kernel.org/r/1500500215-16646-1-git-send-email-treeze.taeung@gmail.com [ split and adjusted from a larger patch ] Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-annotate.c | 2 +- tools/perf/builtin-report.c | 15 ++++++++------- tools/perf/builtin-top.c | 5 +++-- tools/perf/util/annotate.c | 18 +++++++++++------- tools/perf/util/annotate.h | 6 ++++-- 5 files changed, 27 insertions(+), 19 deletions(-) (limited to 'tools') diff --git a/tools/perf/builtin-annotate.c b/tools/perf/builtin-annotate.c index 5205408..96fe1a8 100644 --- a/tools/perf/builtin-annotate.c +++ b/tools/perf/builtin-annotate.c @@ -184,7 +184,7 @@ static int perf_evsel__add_sample(struct perf_evsel *evsel, if (he == NULL) return -ENOMEM; - ret = hist_entry__inc_addr_samples(he, evsel->idx, al->addr); + ret = hist_entry__inc_addr_samples(he, sample, evsel->idx, al->addr); hists__inc_nr_samples(hists, true); return ret; } diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c index cea25d0..983b238 100644 --- a/tools/perf/builtin-report.c +++ b/tools/perf/builtin-report.c @@ -115,37 +115,38 @@ static int hist_iter__report_callback(struct hist_entry_iter *iter, struct report *rep = arg; struct hist_entry *he = iter->he; struct perf_evsel *evsel = iter->evsel; + struct perf_sample *sample = iter->sample; struct mem_info *mi; struct branch_info *bi; if (!ui__has_annotation()) return 0; - hist__account_cycles(iter->sample->branch_stack, al, iter->sample, + hist__account_cycles(sample->branch_stack, al, sample, rep->nonany_branch_mode); if (sort__mode == SORT_MODE__BRANCH) { bi = he->branch_info; - err = addr_map_symbol__inc_samples(&bi->from, evsel->idx); + err = addr_map_symbol__inc_samples(&bi->from, sample, evsel->idx); if (err) goto out; - err = addr_map_symbol__inc_samples(&bi->to, evsel->idx); + err = addr_map_symbol__inc_samples(&bi->to, sample, evsel->idx); } else if (rep->mem_mode) { mi = he->mem_info; - err = addr_map_symbol__inc_samples(&mi->daddr, evsel->idx); + err = addr_map_symbol__inc_samples(&mi->daddr, sample, evsel->idx); if (err) goto out; - err = hist_entry__inc_addr_samples(he, evsel->idx, al->addr); + err = hist_entry__inc_addr_samples(he, sample, evsel->idx, al->addr); } else if (symbol_conf.cumulate_callchain) { if (single) - err = hist_entry__inc_addr_samples(he, evsel->idx, + err = hist_entry__inc_addr_samples(he, sample, evsel->idx, al->addr); } else { - err = hist_entry__inc_addr_samples(he, evsel->idx, al->addr); + err = hist_entry__inc_addr_samples(he, sample, evsel->idx, al->addr); } out: diff --git a/tools/perf/builtin-top.c b/tools/perf/builtin-top.c index 022486d..e5a8f24 100644 --- a/tools/perf/builtin-top.c +++ b/tools/perf/builtin-top.c @@ -183,6 +183,7 @@ static void ui__warn_map_erange(struct map *map, struct symbol *sym, u64 ip) static void perf_top__record_precise_ip(struct perf_top *top, struct hist_entry *he, + struct perf_sample *sample, int counter, u64 ip) { struct annotation *notes; @@ -199,7 +200,7 @@ static void perf_top__record_precise_ip(struct perf_top *top, if (pthread_mutex_trylock(¬es->lock)) return; - err = hist_entry__inc_addr_samples(he, counter, ip); + err = hist_entry__inc_addr_samples(he, sample, counter, ip); pthread_mutex_unlock(¬es->lock); @@ -671,7 +672,7 @@ static int hist_iter__top_callback(struct hist_entry_iter *iter, struct perf_evsel *evsel = iter->evsel; if (perf_hpp_list.sym && single) - perf_top__record_precise_ip(top, he, evsel->idx, al->addr); + perf_top__record_precise_ip(top, he, iter->sample, evsel->idx, al->addr); hist__account_cycles(iter->sample->branch_stack, al, iter->sample, !(top->record_opts.branch_stack & PERF_SAMPLE_BRANCH_ANY)); diff --git a/tools/perf/util/annotate.c b/tools/perf/util/annotate.c index 58c6b63..c2fe16d 100644 --- a/tools/perf/util/annotate.c +++ b/tools/perf/util/annotate.c @@ -697,7 +697,8 @@ static int __symbol__account_cycles(struct annotation *notes, } static int __symbol__inc_addr_samples(struct symbol *sym, struct map *map, - struct annotation *notes, int evidx, u64 addr) + struct annotation *notes, int evidx, u64 addr, + struct perf_sample *sample __maybe_unused) { unsigned offset; struct sym_hist *h; @@ -738,7 +739,8 @@ static struct annotation *symbol__get_annotation(struct symbol *sym, bool cycles } static int symbol__inc_addr_samples(struct symbol *sym, struct map *map, - int evidx, u64 addr) + int evidx, u64 addr, + struct perf_sample *sample) { struct annotation *notes; @@ -747,7 +749,7 @@ static int symbol__inc_addr_samples(struct symbol *sym, struct map *map, notes = symbol__get_annotation(sym, false); if (notes == NULL) return -ENOMEM; - return __symbol__inc_addr_samples(sym, map, notes, evidx, addr); + return __symbol__inc_addr_samples(sym, map, notes, evidx, addr, sample); } static int symbol__account_cycles(u64 addr, u64 start, @@ -811,14 +813,16 @@ int addr_map_symbol__account_cycles(struct addr_map_symbol *ams, return err; } -int addr_map_symbol__inc_samples(struct addr_map_symbol *ams, int evidx) +int addr_map_symbol__inc_samples(struct addr_map_symbol *ams, struct perf_sample *sample, + int evidx) { - return symbol__inc_addr_samples(ams->sym, ams->map, evidx, ams->al_addr); + return symbol__inc_addr_samples(ams->sym, ams->map, evidx, ams->al_addr, sample); } -int hist_entry__inc_addr_samples(struct hist_entry *he, int evidx, u64 ip) +int hist_entry__inc_addr_samples(struct hist_entry *he, struct perf_sample *sample, + int evidx, u64 ip) { - return symbol__inc_addr_samples(he->ms.sym, he->ms.map, evidx, ip); + return symbol__inc_addr_samples(he->ms.sym, he->ms.map, evidx, ip, sample); } static void disasm_line__init_ins(struct disasm_line *dl, struct arch *arch, struct map *map) diff --git a/tools/perf/util/annotate.h b/tools/perf/util/annotate.h index e8c246e..720f181 100644 --- a/tools/perf/util/annotate.h +++ b/tools/perf/util/annotate.h @@ -154,13 +154,15 @@ static inline struct annotation *symbol__annotation(struct symbol *sym) return (void *)sym - symbol_conf.priv_size; } -int addr_map_symbol__inc_samples(struct addr_map_symbol *ams, int evidx); +int addr_map_symbol__inc_samples(struct addr_map_symbol *ams, struct perf_sample *sample, + int evidx); int addr_map_symbol__account_cycles(struct addr_map_symbol *ams, struct addr_map_symbol *start, unsigned cycles); -int hist_entry__inc_addr_samples(struct hist_entry *he, int evidx, u64 addr); +int hist_entry__inc_addr_samples(struct hist_entry *he, struct perf_sample *sample, + int evidx, u64 addr); int symbol__alloc_hist(struct symbol *sym); void symbol__annotate_zero_histograms(struct symbol *sym); -- cgit v1.1 From 461c17f00f400f95116880d91d20a7fcd84263a9 Mon Sep 17 00:00:00 2001 From: Taeung Song Date: Thu, 20 Jul 2017 17:18:05 -0300 Subject: perf annotate: Store the sample period in each histogram bucket We'll use it soon, when fixing --show-total-period. Signed-off-by: Taeung Song Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/r/1500500215-16646-1-git-send-email-treeze.taeung@gmail.com [ split from a larger patch, do the math in __symbol__inc_addr_samples() ] Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/annotate.c | 17 ++++++++++++----- tools/perf/util/annotate.h | 1 + 2 files changed, 13 insertions(+), 5 deletions(-) (limited to 'tools') diff --git a/tools/perf/util/annotate.c b/tools/perf/util/annotate.c index c2fe16d..00e085f 100644 --- a/tools/perf/util/annotate.c +++ b/tools/perf/util/annotate.c @@ -698,7 +698,7 @@ static int __symbol__account_cycles(struct annotation *notes, static int __symbol__inc_addr_samples(struct symbol *sym, struct map *map, struct annotation *notes, int evidx, u64 addr, - struct perf_sample *sample __maybe_unused) + struct perf_sample *sample) { unsigned offset; struct sym_hist *h; @@ -716,10 +716,13 @@ static int __symbol__inc_addr_samples(struct symbol *sym, struct map *map, h = annotation__histogram(notes, evidx); h->nr_samples++; h->addr[offset].nr_samples++; + h->period += sample->period; + h->addr[offset].period += sample->period; pr_debug3("%#" PRIx64 " %s: period++ [addr: %#" PRIx64 ", %#" PRIx64 - ", evidx=%d] => %" PRIu64 "\n", sym->start, sym->name, - addr, addr - sym->start, evidx, h->addr[offset].nr_samples); + ", evidx=%d] => nr_samples: %" PRIu64 ", period: %" PRIu64 "\n", + sym->start, sym->name, addr, addr - sym->start, evidx, + h->addr[offset].nr_samples, h->addr[offset].period); return 0; } @@ -937,7 +940,7 @@ double disasm__calc_percent(struct annotation *notes, int evidx, s64 offset, struct source_line *src_line = notes->src->lines; double percent = 0.0; - sample->nr_samples = 0; + sample->nr_samples = sample->period = 0; if (src_line) { size_t sizeof_src_line = sizeof(*src_line) + @@ -957,11 +960,15 @@ double disasm__calc_percent(struct annotation *notes, int evidx, s64 offset, } else { struct sym_hist *h = annotation__histogram(notes, evidx); unsigned int hits = 0; + u64 period = 0; - while (offset < end) + while (offset < end) { hits += h->addr[offset++].nr_samples; + period += h->addr[offset++].period; + } if (h->nr_samples) { + sample->period = period; sample->nr_samples = hits; percent = 100.0 * hits / h->nr_samples; } diff --git a/tools/perf/util/annotate.h b/tools/perf/util/annotate.h index 720f181..9ce575c 100644 --- a/tools/perf/util/annotate.h +++ b/tools/perf/util/annotate.h @@ -88,6 +88,7 @@ double disasm__calc_percent(struct annotation *notes, int evidx, s64 offset, struct sym_hist { u64 nr_samples; + u64 period; struct sym_hist_entry addr[0]; }; -- cgit v1.1 From ecd5f9959d2c9540482977ff1208ea67fbfb8cc9 Mon Sep 17 00:00:00 2001 From: Taeung Song Date: Fri, 21 Jul 2017 11:38:48 -0300 Subject: perf annotate: Do not overwrite sample->period In fixing the --show-total-period option it was noticed that the value of sample->period was being overwritten, fix it. Signed-off-by: Taeung Song Cc: Jiri Olsa Cc: Milian Wolff Cc: Namhyung Kim Fixes: fd36f3dd7933 ("perf hist: Pass struct sample to __hists__add_entry()") Link: http://lkml.kernel.org/r/1500500215-16646-1-git-send-email-treeze.taeung@gmail.com [ split from a larger patch, added the Fixes tag ] Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-annotate.c | 1 - 1 file changed, 1 deletion(-) (limited to 'tools') diff --git a/tools/perf/builtin-annotate.c b/tools/perf/builtin-annotate.c index 96fe1a8..7e33278 100644 --- a/tools/perf/builtin-annotate.c +++ b/tools/perf/builtin-annotate.c @@ -177,7 +177,6 @@ static int perf_evsel__add_sample(struct perf_evsel *evsel, */ process_branch_stack(sample->branch_stack, al, sample); - sample->period = 1; sample->weight = 1; he = hists__add_entry(hists, al, NULL, NULL, NULL, sample, true); -- cgit v1.1 From 49218edd649ae4922aa776aad5b7a65d028c6e96 Mon Sep 17 00:00:00 2001 From: Todd E Brandt Date: Wed, 5 Jul 2017 14:42:55 -0700 Subject: pm-graph: AnalyzeSuspend v4.7 - changed -rtcwake parameter to be on & 15 sec by default, to disable rtcwake use: "-rtcwake off" - changed behavior of -o: renames HTML file on rerun, subdir on new run - changed execution_misalignment error to missing_function_name - add sysinfo to logs and timeline via a custom dmidecode call it supplants dmidecode tool when used as a library call - add -sysinfo command, displays dmidecode values and cpu/mem info - set trace buffer size to lesser of memtotal/2 or 2GB when using callgraph - add support for /sys/power/mem_sleep. if mem_sleep found: mem-shallow=standby, mem-s2idle=freeze, mem-deep=mem - remove redundant javascript - cosmetic changes to HTML layout Signed-off-by: Todd Brandt Signed-off-by: Rafael J. Wysocki --- tools/power/pm-graph/analyze_suspend.py | 534 +++++++++++++++++++++++--------- 1 file changed, 379 insertions(+), 155 deletions(-) (limited to 'tools') diff --git a/tools/power/pm-graph/analyze_suspend.py b/tools/power/pm-graph/analyze_suspend.py index a9206e6..1b60fe2 100755 --- a/tools/power/pm-graph/analyze_suspend.py +++ b/tools/power/pm-graph/analyze_suspend.py @@ -68,10 +68,12 @@ from subprocess import call, Popen, PIPE # store system values and test parameters class SystemValues: title = 'SleepGraph' - version = '4.6' + version = '4.7' ansi = False verbose = False - addlogs = False + testlog = True + dmesglog = False + ftracelog = False mindevlen = 0.0 mincglen = 0.0 cgphase = '' @@ -79,10 +81,11 @@ class SystemValues: max_graph_depth = 0 callloopmaxgap = 0.0001 callloopmaxlen = 0.005 + cpucount = 0 + memtotal = 204800 srgap = 0 cgexp = False - outdir = '' - testdir = '.' + testdir = '' tpath = '/sys/kernel/debug/tracing/' fpdtpath = '/sys/firmware/acpi/tables/FPDT' epath = '/sys/kernel/debug/tracing/events/power/' @@ -95,14 +98,17 @@ class SystemValues: testcommand = '' mempath = '/dev/mem' powerfile = '/sys/power/state' + mempowerfile = '/sys/power/mem_sleep' suspendmode = 'mem' + memmode = '' hostname = 'localhost' prefix = 'test' teststamp = '' + sysstamp = '' dmesgstart = 0.0 dmesgfile = '' ftracefile = '' - htmlfile = '' + htmlfile = 'output.html' embedded = False rtcwake = True rtcwaketime = 15 @@ -127,9 +133,6 @@ class SystemValues: devpropfmt = '# Device Properties: .*' tracertypefmt = '# tracer: (?P.*)' firmwarefmt = '# fwsuspend (?P[0-9]*) fwresume (?P[0-9]*)$' - stampfmt = '# suspend-(?P[0-9]{2})(?P[0-9]{2})(?P[0-9]{2})-'+\ - '(?P[0-9]{2})(?P[0-9]{2})(?P[0-9]{2})'+\ - ' (?P.*) (?P.*) (?P.*)$' tracefuncs = { 'sys_sync': dict(), 'pm_prepare_console': dict(), @@ -218,7 +221,7 @@ class SystemValues: # if this is a phoronix test run, set some default options if('LOG_FILE' in os.environ and 'TEST_RESULTS_IDENTIFIER' in os.environ): self.embedded = True - self.addlogs = True + self.dmesglog = self.ftracelog = True self.htmlfile = os.environ['LOG_FILE'] self.archargs = 'args_'+platform.machine() self.hostname = platform.node() @@ -233,6 +236,13 @@ class SystemValues: self.rtcpath = rtc if (hasattr(sys.stdout, 'isatty') and sys.stdout.isatty()): self.ansi = True + self.testdir = datetime.now().strftime('suspend-%y%m%d-%H%M%S') + def rootCheck(self, fatal=True): + if(os.access(self.powerfile, os.W_OK)): + return True + if fatal: + doError('This command requires sysfs mount and root access') + return False def rootUser(self, fatal=False): if 'USER' in os.environ and os.environ['USER'] == 'root': return True @@ -249,30 +259,60 @@ class SystemValues: args['date'] = n.strftime('%y%m%d') args['time'] = n.strftime('%H%M%S') args['hostname'] = self.hostname - self.outdir = value.format(**args) + return value.format(**args) def setOutputFile(self): - if((self.htmlfile == '') and (self.dmesgfile != '')): + if self.dmesgfile != '': m = re.match('(?P.*)_dmesg\.txt$', self.dmesgfile) if(m): self.htmlfile = m.group('name')+'.html' - if((self.htmlfile == '') and (self.ftracefile != '')): + if self.ftracefile != '': m = re.match('(?P.*)_ftrace\.txt$', self.ftracefile) if(m): self.htmlfile = m.group('name')+'.html' - if(self.htmlfile == ''): - self.htmlfile = 'output.html' - def initTestOutput(self, subdir, testpath=''): + def systemInfo(self, info): + p = c = m = b = '' + if 'baseboard-manufacturer' in info: + m = info['baseboard-manufacturer'] + elif 'system-manufacturer' in info: + m = info['system-manufacturer'] + if 'baseboard-product-name' in info: + p = info['baseboard-product-name'] + elif 'system-product-name' in info: + p = info['system-product-name'] + if 'processor-version' in info: + c = info['processor-version'] + if 'bios-version' in info: + b = info['bios-version'] + self.sysstamp = '# sysinfo | man:%s | plat:%s | cpu:%s | bios:%s | numcpu:%d | memsz:%d' % \ + (m, p, c, b, self.cpucount, self.memtotal) + def printSystemInfo(self): + self.rootCheck(True) + out = dmidecode(self.mempath, True) + fmt = '%-24s: %s' + for name in sorted(out): + print fmt % (name, out[name]) + print fmt % ('cpucount', ('%d' % self.cpucount)) + print fmt % ('memtotal', ('%d kB' % self.memtotal)) + def cpuInfo(self): + self.cpucount = 0 + fp = open('/proc/cpuinfo', 'r') + for line in fp: + if re.match('^processor[ \t]*:[ \t]*[0-9]*', line): + self.cpucount += 1 + fp.close() + fp = open('/proc/meminfo', 'r') + for line in fp: + m = re.match('^MemTotal:[ \t]*(?P[0-9]*) *kB', line) + if m: + self.memtotal = int(m.group('sz')) + break + fp.close() + def initTestOutput(self, name): self.prefix = self.hostname v = open('/proc/version', 'r').read().strip() kver = string.split(v)[2] - n = datetime.now() - testtime = n.strftime('suspend-%m%d%y-%H%M%S') - if not testpath: - testpath = n.strftime('suspend-%y%m%d-%H%M%S') - if(subdir != "."): - self.testdir = subdir+"/"+testpath - else: - self.testdir = testpath + fmt = name+'-%m%d%y-%H%M%S' + testtime = datetime.now().strftime(fmt) self.teststamp = \ '# '+testtime+' '+self.prefix+' '+self.suspendmode+' '+kver if(self.embedded): @@ -355,7 +395,7 @@ class SystemValues: continue self.tracefuncs[i] = dict() def getFtraceFilterFunctions(self, current): - rootCheck(True) + self.rootCheck(True) if not current: call('cat '+self.tpath+'available_filter_functions', shell=True) return @@ -453,7 +493,7 @@ class SystemValues: val += '\nr:%s_ret %s $retval\n' % (name, func) return val def addKprobes(self, output=False): - if len(sysvals.kprobes) < 1: + if len(self.kprobes) < 1: return if output: print(' kprobe functions in this kernel:') @@ -525,7 +565,7 @@ class SystemValues: fp.flush() fp.close() except: - pass + return False return True def fgetVal(self, path): file = self.tpath+path @@ -566,9 +606,15 @@ class SystemValues: self.cleanupFtrace() # set the trace clock to global self.fsetVal('global', 'trace_clock') - # set trace buffer to a huge value self.fsetVal('nop', 'current_tracer') - self.fsetVal('131073', 'buffer_size_kb') + # set trace buffer to a huge value + if self.usecallgraph or self.usedevsrc: + tgtsize = min(self.memtotal / 2, 2*1024*1024) + maxbuf = '%d' % (tgtsize / max(1, self.cpucount)) + if self.cpucount < 1 or not self.fsetVal(maxbuf, 'buffer_size_kb'): + self.fsetVal('131072', 'buffer_size_kb') + else: + self.fsetVal('16384', 'buffer_size_kb') # go no further if this is just a status check if testing: return @@ -641,6 +687,15 @@ class SystemValues: if not self.ansi: return str return '\x1B[%d;40m%s\x1B[m' % (color, str) + def writeDatafileHeader(self, filename, fwdata=[]): + fp = open(filename, 'w') + fp.write(self.teststamp+'\n') + fp.write(self.sysstamp+'\n') + if(self.suspendmode == 'mem' or self.suspendmode == 'command'): + for fw in fwdata: + if(fw): + fp.write('# fwsuspend %u fwresume %u\n' % (fw[0], fw[1])) + fp.close() sysvals = SystemValues() suspendmodename = { @@ -1008,6 +1063,12 @@ class Data: else: self.trimTime(self.tSuspended, \ self.tResumed-self.tSuspended, False) + def getTimeValues(self): + sktime = (self.dmesg['suspend_machine']['end'] - \ + self.tKernSus) * 1000 + rktime = (self.dmesg['resume_complete']['end'] - \ + self.dmesg['resume_machine']['start']) * 1000 + return (sktime, rktime) def setPhase(self, phase, ktime, isbegin): if(isbegin): self.dmesg[phase]['start'] = ktime @@ -1517,7 +1578,7 @@ class FTraceCallGraph: prelinedep += 1 last = 0 lasttime = line.time - virtualfname = 'execution_misalignment' + virtualfname = 'missing_function_name' if len(self.list) > 0: last = self.list[-1] lasttime = last.time @@ -1773,24 +1834,30 @@ class Timeline: html_device = '
{6}
\n' html_phase = '
{5}
\n' html_phaselet = '
\n' + html_legend = '
 {2}
\n' def __init__(self, rowheight, scaleheight): self.rowH = rowheight self.scaleH = scaleheight self.html = '' - def createHeader(self, sv, suppress=''): + def createHeader(self, sv): if(not sv.stamp['time']): return self.html += '
%s v%s
' \ % (sv.title, sv.version) - if sv.logmsg and 'log' not in suppress: - self.html += '' - if sv.addlogs and 'dmesg' not in suppress: - self.html += '' - if sv.addlogs and sv.ftracefile and 'ftrace' not in suppress: - self.html += '' + if sv.logmsg and sv.testlog: + self.html += '' + if sv.dmesglog: + self.html += '' + if sv.ftracelog: + self.html += '' headline_stamp = '
{0} {1} {2} {3}
\n' self.html += headline_stamp.format(sv.stamp['host'], sv.stamp['kernel'], sv.stamp['mode'], sv.stamp['time']) + if 'man' in sv.stamp and 'plat' in sv.stamp and 'cpu' in sv.stamp: + headline_sysinfo = '
{0} {1} with {2}
\n' + self.html += headline_sysinfo.format(sv.stamp['man'], + sv.stamp['plat'], sv.stamp['cpu']) + # Function: getDeviceRows # Description: # determine how may rows the device funcs will take @@ -1839,7 +1906,7 @@ class Timeline: # devlist: the list of devices/actions in a group of contiguous phases # Output: # The total number of rows needed to display this phase of the timeline - def getPhaseRows(self, devlist, row=0): + def getPhaseRows(self, devlist, row=0, sortby='length'): # clear all rows and set them to undefined remaining = len(devlist) rowdata = dict() @@ -1852,8 +1919,12 @@ class Timeline: if tp not in myphases: myphases.append(tp) dev['row'] = -1 - # sort by length 1st, then name 2nd - sortdict[item] = (float(dev['end']) - float(dev['start']), item.dev['name']) + if sortby == 'start': + # sort by start 1st, then length 2nd + sortdict[item] = (-1*float(dev['start']), float(dev['end']) - float(dev['start'])) + else: + # sort by length 1st, then name 2nd + sortdict[item] = (float(dev['end']) - float(dev['start']), item.dev['name']) if 'src' in dev: dev['devrows'] = self.getDeviceRows(dev['src']) # sort the devlist by length so that large items graph on top @@ -1995,8 +2066,13 @@ class Timeline: # A list of values describing the properties of these test runs class TestProps: stamp = '' + sysinfo = '' S0i3 = False fwdata = [] + stampfmt = '# [a-z]*-(?P[0-9]{2})(?P[0-9]{2})(?P[0-9]{2})-'+\ + '(?P[0-9]{2})(?P[0-9]{2})(?P[0-9]{2})'+\ + ' (?P.*) (?P.*) (?P.*)$' + sysinfofmt = '^# sysinfo .*' ftrace_line_fmt_fg = \ '^ *(?P