summaryrefslogtreecommitdiffstats
diff options
context:
space:
mode:
authorAdrian Hunter <adrian.hunter@intel.com>2013-08-07 14:38:45 +0300
committerArnaldo Carvalho de Melo <acme@redhat.com>2013-08-07 17:35:30 -0300
commitb55ae0a976f2927ea8f15a85c43bbe6d25a68a41 (patch)
treecc698532742f02295d114f271c6c175d5d04ec62
parent3445432b7b24665cf4693fc4794c62d4d768a978 (diff)
downloadop-kernel-dev-b55ae0a976f2927ea8f15a85c43bbe6d25a68a41.zip
op-kernel-dev-b55ae0a976f2927ea8f15a85c43bbe6d25a68a41.tar.gz
perf tests: Add test for reading object code
Using the information in mmap events, perf tools can read object code associated with sampled addresses. A test is added that compares bytes read by perf with the same bytes read using objdump. Signed-off-by: Adrian Hunter <adrian.hunter@intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Ingo Molnar <mingo@kernel.org> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Namhyung Kim <namhyung@gmail.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stephane Eranian <eranian@google.com> Link: http://lkml.kernel.org/r/1375875537-4509-2-git-send-email-adrian.hunter@intel.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
-rw-r--r--tools/perf/Makefile1
-rw-r--r--tools/perf/tests/builtin-test.c4
-rw-r--r--tools/perf/tests/code-reading.c509
-rw-r--r--tools/perf/tests/tests.h1
4 files changed, 515 insertions, 0 deletions
diff --git a/tools/perf/Makefile b/tools/perf/Makefile
index bfd12d0..e0d3d9f 100644
--- a/tools/perf/Makefile
+++ b/tools/perf/Makefile
@@ -392,6 +392,7 @@ LIB_OBJS += $(OUTPUT)tests/sw-clock.o
ifeq ($(ARCH),x86)
LIB_OBJS += $(OUTPUT)tests/perf-time-to-tsc.o
endif
+LIB_OBJS += $(OUTPUT)tests/code-reading.o
BUILTIN_OBJS += $(OUTPUT)builtin-annotate.o
BUILTIN_OBJS += $(OUTPUT)builtin-bench.o
diff --git a/tools/perf/tests/builtin-test.c b/tools/perf/tests/builtin-test.c
index b7b4049..f5af192 100644
--- a/tools/perf/tests/builtin-test.c
+++ b/tools/perf/tests/builtin-test.c
@@ -100,6 +100,10 @@ static struct test {
},
#endif
{
+ .desc = "Test object code reading",
+ .func = test__code_reading,
+ },
+ {
.func = NULL,
},
};
diff --git a/tools/perf/tests/code-reading.c b/tools/perf/tests/code-reading.c
new file mode 100644
index 0000000..28bee62
--- /dev/null
+++ b/tools/perf/tests/code-reading.c
@@ -0,0 +1,509 @@
+#include <sys/types.h>
+#include <stdlib.h>
+#include <unistd.h>
+#include <stdio.h>
+#include <inttypes.h>
+#include <ctype.h>
+#include <string.h>
+
+#include "parse-events.h"
+#include "evlist.h"
+#include "evsel.h"
+#include "thread_map.h"
+#include "cpumap.h"
+#include "machine.h"
+#include "event.h"
+#include "thread.h"
+
+#include "tests.h"
+
+#define BUFSZ 1024
+#define READLEN 128
+
+static unsigned int hex(char c)
+{
+ if (c >= '0' && c <= '9')
+ return c - '0';
+ if (c >= 'a' && c <= 'f')
+ return c - 'a' + 10;
+ return c - 'A' + 10;
+}
+
+static void read_objdump_line(const char *line, size_t line_len, void **buf,
+ size_t *len)
+{
+ const char *p;
+ size_t i;
+
+ /* Skip to a colon */
+ p = strchr(line, ':');
+ if (!p)
+ return;
+ i = p + 1 - line;
+
+ /* Read bytes */
+ while (*len) {
+ char c1, c2;
+
+ /* Skip spaces */
+ for (; i < line_len; i++) {
+ if (!isspace(line[i]))
+ break;
+ }
+ /* Get 2 hex digits */
+ if (i >= line_len || !isxdigit(line[i]))
+ break;
+ c1 = line[i++];
+ if (i >= line_len || !isxdigit(line[i]))
+ break;
+ c2 = line[i++];
+ /* Followed by a space */
+ if (i < line_len && line[i] && !isspace(line[i]))
+ break;
+ /* Store byte */
+ *(unsigned char *)*buf = (hex(c1) << 4) | hex(c2);
+ *buf += 1;
+ *len -= 1;
+ }
+}
+
+static int read_objdump_output(FILE *f, void **buf, size_t *len)
+{
+ char *line = NULL;
+ size_t line_len;
+ ssize_t ret;
+ int err = 0;
+
+ while (1) {
+ ret = getline(&line, &line_len, f);
+ if (feof(f))
+ break;
+ if (ret < 0) {
+ pr_debug("getline failed\n");
+ err = -1;
+ break;
+ }
+ read_objdump_line(line, ret, buf, len);
+ }
+
+ free(line);
+
+ return err;
+}
+
+static int read_via_objdump(const char *filename, u64 addr, void *buf,
+ size_t len)
+{
+ char cmd[PATH_MAX * 2];
+ const char *fmt;
+ FILE *f;
+ int ret;
+
+ fmt = "%s -d --start-address=0x%"PRIx64" --stop-address=0x%"PRIx64" %s";
+ ret = snprintf(cmd, sizeof(cmd), fmt, "objdump", addr, addr + len,
+ filename);
+ if (ret <= 0 || (size_t)ret >= sizeof(cmd))
+ return -1;
+
+ pr_debug("Objdump command is: %s\n", cmd);
+
+ f = popen(cmd, "r");
+ if (!f) {
+ pr_debug("popen failed\n");
+ return -1;
+ }
+
+ ret = read_objdump_output(f, &buf, &len);
+ if (len) {
+ pr_debug("objdump read too few bytes\n");
+ if (!ret)
+ ret = len;
+ }
+
+ pclose(f);
+
+ return ret;
+}
+
+static int read_object_code(u64 addr, size_t len, u8 cpumode,
+ struct thread *thread, struct machine *machine)
+{
+ struct addr_location al;
+ unsigned char buf1[BUFSZ];
+ unsigned char buf2[BUFSZ];
+ size_t ret_len;
+ u64 objdump_addr;
+ int ret;
+
+ pr_debug("Reading object code for memory address: %#"PRIx64"\n", addr);
+
+ thread__find_addr_map(thread, machine, cpumode, MAP__FUNCTION, addr,
+ &al);
+ if (!al.map || !al.map->dso) {
+ pr_debug("thread__find_addr_map failed\n");
+ return -1;
+ }
+
+ pr_debug("File is: %s\n", al.map->dso->long_name);
+
+ if (al.map->dso->symtab_type == DSO_BINARY_TYPE__KALLSYMS) {
+ pr_debug("Unexpected kernel address - skipping\n");
+ return 0;
+ }
+
+ pr_debug("On file address is: %#"PRIx64"\n", al.addr);
+
+ if (len > BUFSZ)
+ len = BUFSZ;
+
+ /* Do not go off the map */
+ if (addr + len > al.map->end)
+ len = al.map->end - addr;
+
+ /* Read the object code using perf */
+ ret_len = dso__data_read_offset(al.map->dso, machine, al.addr, buf1,
+ len);
+ if (ret_len != len) {
+ pr_debug("dso__data_read_offset failed\n");
+ return -1;
+ }
+
+ /*
+ * Converting addresses for use by objdump requires more information.
+ * map__load() does that. See map__rip_2objdump() for details.
+ */
+ if (map__load(al.map, NULL))
+ return -1;
+
+ /* Read the object code using objdump */
+ objdump_addr = map__rip_2objdump(al.map, al.addr);
+ ret = read_via_objdump(al.map->dso->long_name, objdump_addr, buf2, len);
+ if (ret > 0) {
+ /*
+ * The kernel maps are inaccurate - assume objdump is right in
+ * that case.
+ */
+ if (cpumode == PERF_RECORD_MISC_KERNEL ||
+ cpumode == PERF_RECORD_MISC_GUEST_KERNEL) {
+ len -= ret;
+ if (len)
+ pr_debug("Reducing len to %zu\n", len);
+ else
+ return -1;
+ }
+ }
+ if (ret < 0) {
+ pr_debug("read_via_objdump failed\n");
+ return -1;
+ }
+
+ /* The results should be identical */
+ if (memcmp(buf1, buf2, len)) {
+ pr_debug("Bytes read differ from those read by objdump\n");
+ return -1;
+ }
+ pr_debug("Bytes read match those read by objdump\n");
+
+ return 0;
+}
+
+static int process_sample_event(struct machine *machine,
+ struct perf_evlist *evlist,
+ union perf_event *event)
+{
+ struct perf_sample sample;
+ struct thread *thread;
+ u8 cpumode;
+
+ if (perf_evlist__parse_sample(evlist, event, &sample)) {
+ pr_debug("perf_evlist__parse_sample failed\n");
+ return -1;
+ }
+
+ thread = machine__findnew_thread(machine, sample.pid);
+ if (!thread) {
+ pr_debug("machine__findnew_thread failed\n");
+ return -1;
+ }
+
+ cpumode = event->header.misc & PERF_RECORD_MISC_CPUMODE_MASK;
+
+ return read_object_code(sample.ip, READLEN, cpumode, thread, machine);
+}
+
+static int process_event(struct machine *machine, struct perf_evlist *evlist,
+ union perf_event *event)
+{
+ if (event->header.type == PERF_RECORD_SAMPLE)
+ return process_sample_event(machine, evlist, event);
+
+ if (event->header.type < PERF_RECORD_MAX)
+ return machine__process_event(machine, event);
+
+ return 0;
+}
+
+static int process_events(struct machine *machine, struct perf_evlist *evlist)
+{
+ union perf_event *event;
+ int i, ret;
+
+ for (i = 0; i < evlist->nr_mmaps; i++) {
+ while ((event = perf_evlist__mmap_read(evlist, i)) != NULL) {
+ ret = process_event(machine, evlist, event);
+ if (ret < 0)
+ return ret;
+ }
+ }
+ return 0;
+}
+
+static int comp(const void *a, const void *b)
+{
+ return *(int *)a - *(int *)b;
+}
+
+static void do_sort_something(void)
+{
+ size_t sz = 40960;
+ int buf[sz], i;
+
+ for (i = 0; i < (int)sz; i++)
+ buf[i] = sz - i - 1;
+
+ qsort(buf, sz, sizeof(int), comp);
+
+ for (i = 0; i < (int)sz; i++) {
+ if (buf[i] != i) {
+ pr_debug("qsort failed\n");
+ break;
+ }
+ }
+}
+
+static void sort_something(void)
+{
+ int i;
+
+ for (i = 0; i < 10; i++)
+ do_sort_something();
+}
+
+static void syscall_something(void)
+{
+ int pipefd[2];
+ int i;
+
+ for (i = 0; i < 1000; i++) {
+ if (pipe(pipefd) < 0) {
+ pr_debug("pipe failed\n");
+ break;
+ }
+ close(pipefd[1]);
+ close(pipefd[0]);
+ }
+}
+
+static void fs_something(void)
+{
+ const char *test_file_name = "temp-perf-code-reading-test-file--";
+ FILE *f;
+ int i;
+
+ for (i = 0; i < 1000; i++) {
+ f = fopen(test_file_name, "w+");
+ if (f) {
+ fclose(f);
+ unlink(test_file_name);
+ }
+ }
+}
+
+static void do_something(void)
+{
+ fs_something();
+
+ sort_something();
+
+ syscall_something();
+}
+
+enum {
+ TEST_CODE_READING_OK,
+ TEST_CODE_READING_NO_VMLINUX,
+ TEST_CODE_READING_NO_ACCESS,
+};
+
+static int do_test_code_reading(void)
+{
+ struct machines machines;
+ struct machine *machine;
+ struct thread *thread;
+ struct perf_record_opts opts = {
+ .mmap_pages = UINT_MAX,
+ .user_freq = UINT_MAX,
+ .user_interval = ULLONG_MAX,
+ .freq = 4000,
+ .target = {
+ .uses_mmap = true,
+ },
+ };
+ struct thread_map *threads = NULL;
+ struct cpu_map *cpus = NULL;
+ struct perf_evlist *evlist = NULL;
+ struct perf_evsel *evsel = NULL;
+ int err = -1, ret;
+ pid_t pid;
+ struct map *map;
+ bool have_vmlinux, excl_kernel = false;
+
+ pid = getpid();
+
+ machines__init(&machines);
+ machine = &machines.host;
+
+ ret = machine__create_kernel_maps(machine);
+ if (ret < 0) {
+ pr_debug("machine__create_kernel_maps failed\n");
+ goto out_err;
+ }
+
+ /* Load kernel map */
+ map = machine->vmlinux_maps[MAP__FUNCTION];
+ ret = map__load(map, NULL);
+ if (ret < 0) {
+ pr_debug("map__load failed\n");
+ goto out_err;
+ }
+ have_vmlinux = map->dso->symtab_type == DSO_BINARY_TYPE__VMLINUX;
+ /* No point getting kernel events if there is no vmlinux */
+ if (!have_vmlinux)
+ excl_kernel = true;
+
+ threads = thread_map__new_by_tid(pid);
+ if (!threads) {
+ pr_debug("thread_map__new_by_tid failed\n");
+ goto out_err;
+ }
+
+ ret = perf_event__synthesize_thread_map(NULL, threads,
+ perf_event__process, machine);
+ if (ret < 0) {
+ pr_debug("perf_event__synthesize_thread_map failed\n");
+ goto out_err;
+ }
+
+ thread = machine__findnew_thread(machine, pid);
+ if (!thread) {
+ pr_debug("machine__findnew_thread failed\n");
+ goto out_err;
+ }
+
+ cpus = cpu_map__new(NULL);
+ if (!cpus) {
+ pr_debug("cpu_map__new failed\n");
+ goto out_err;
+ }
+
+ while (1) {
+ const char *str;
+
+ evlist = perf_evlist__new();
+ if (!evlist) {
+ pr_debug("perf_evlist__new failed\n");
+ goto out_err;
+ }
+
+ perf_evlist__set_maps(evlist, cpus, threads);
+
+ if (excl_kernel)
+ str = "cycles:u";
+ else
+ str = "cycles";
+ pr_debug("Parsing event '%s'\n", str);
+ ret = parse_events(evlist, str);
+ if (ret < 0) {
+ pr_debug("parse_events failed\n");
+ goto out_err;
+ }
+
+ perf_evlist__config(evlist, &opts);
+
+ evsel = perf_evlist__first(evlist);
+
+ evsel->attr.comm = 1;
+ evsel->attr.disabled = 1;
+ evsel->attr.enable_on_exec = 0;
+
+ ret = perf_evlist__open(evlist);
+ if (ret < 0) {
+ if (!excl_kernel) {
+ excl_kernel = true;
+ perf_evlist__delete(evlist);
+ evlist = NULL;
+ continue;
+ }
+ pr_debug("perf_evlist__open failed\n");
+ goto out_err;
+ }
+ break;
+ }
+
+ ret = perf_evlist__mmap(evlist, UINT_MAX, false);
+ if (ret < 0) {
+ pr_debug("perf_evlist__mmap failed\n");
+ goto out_err;
+ }
+
+ perf_evlist__enable(evlist);
+
+ do_something();
+
+ perf_evlist__disable(evlist);
+
+ ret = process_events(machine, evlist);
+ if (ret < 0)
+ goto out_err;
+
+ if (!have_vmlinux)
+ err = TEST_CODE_READING_NO_VMLINUX;
+ else if (excl_kernel)
+ err = TEST_CODE_READING_NO_ACCESS;
+ else
+ err = TEST_CODE_READING_OK;
+out_err:
+ if (evlist) {
+ perf_evlist__munmap(evlist);
+ perf_evlist__close(evlist);
+ perf_evlist__delete(evlist);
+ }
+ if (cpus)
+ cpu_map__delete(cpus);
+ if (threads)
+ thread_map__delete(threads);
+ machines__destroy_kernel_maps(&machines);
+ machine__delete_threads(machine);
+ machines__exit(&machines);
+
+ return err;
+}
+
+int test__code_reading(void)
+{
+ int ret;
+
+ ret = do_test_code_reading();
+
+ switch (ret) {
+ case TEST_CODE_READING_OK:
+ return 0;
+ case TEST_CODE_READING_NO_VMLINUX:
+ fprintf(stderr, " (no vmlinux)");
+ return 0;
+ case TEST_CODE_READING_NO_ACCESS:
+ fprintf(stderr, " (no access)");
+ return 0;
+ default:
+ return -1;
+ };
+}
diff --git a/tools/perf/tests/tests.h b/tools/perf/tests/tests.h
index d22202a..c748f53 100644
--- a/tools/perf/tests/tests.h
+++ b/tools/perf/tests/tests.h
@@ -36,5 +36,6 @@ int test__bp_signal_overflow(void);
int test__task_exit(void);
int test__sw_clock_freq(void);
int test__perf_time_to_tsc(void);
+int test__code_reading(void);
#endif /* TESTS_H */
OpenPOWER on IntegriCloud