summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* MFC r317632:ngie2017-05-311-1/+2
| | | | | | Fix "make cscope-clean" when .OBJDIR already exists The cscope generated files are always put in .CURDIR .
* MFC r316742:ngie2017-05-311-2/+2
| | | | | | Clarify units for mediasize and sectorsize in DoFile(..) They're byte quantities.
* MFC r316740:ngie2017-05-311-1/+2
| | | | | | | | Clarify `minfree` warning message in check_space(..) - State that the units are kB. - Be more complete/concise in terms of what is required (in this case `minfree` must be at least `X`kB)
* MFC r317311,r317312,r319019:ngie2017-05-311-2/+2
| | | | | | | | | | | | | | | | | | | | | | | r317311: Check for failures from getpagesize(3) Return errno on failure, similar to the open(2) call above it. CID: 1193753 r317312: Fix type for `pagesize` to match the return type for getpagesize(3) to fix the build Pointyhat to: ngie r319019: Remove getpagesize(3) error checking added in r317312 getpagesize(3) no longer fails as of r317436.
* MFC r319056,r319058,r319059,r319060,r319061,r319078:ngie2017-05-312-31/+31
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | r319056: tests/sys/file/ftruncate_test: use an exit code of 1 instead of -1 with err*(3). An exit code of -1 is implementation defined -- it's best to stick with something well-defined (1). r319058: Create a deterministic file in the kyua sandbox, instead of a temporary file outside the kyua sandbox This helps ensure that the file is removed at test exit, and as a side effect, cures a warning about umasks with Coverity. r319059: Use an exit code of 1 instead of -1 for reasons noted in r319056 r319060: Use main(void) instead of main(argc __unused, argv __unused) r319061: Don't leak accept_fd on thread completion CID: 1296068 r319078: Tweak r319058 slightly - Specify an explicit mode when using O_CREAT per open(2). - Fix the error message (add missing enclosing parentheses). MFC with: r319058
* MFC ↵ngie2017-05-319-98/+95
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | r319027,r319028,r319029,r319030,r319031,r319033,r319034,r319035,r319036,r319037,r319038,r319039,r319040,r319041,r319042,r319043,r319044,r319045,r319046: r319027: lib/libc/tests/nss: use calloc appropriately The pattern used prior to this commit was `calloc(1, n * sizeof(type))`; the pattern that should be used however is `calloc(n, sizeof(type))`. r319028: Sort make variables to suit style.Makefile(5) This is being done prior to functional changes. r319029: Staticize functions and remove unused variables to aid with bumping WARNS r319030: Fix -Wsign-compare warnings r319031: getusershell_test: staticize run_tests(..) to fix warnings r319033: getserv_test: fix -Wsign-compare and -Wmissing-prototypes warnings r319034: getaddrinfo_test: fix -Wsign-compare warnings r319035: getrpc_test: fix -Wmissing-prototypes and -Wsign-compare warnings r319036: getproto_test: fix -Wmissing-prototypes and -Wsign-compare warnings r319037: getaddrinfo_test: mark unused function parameters __unused to fix -Wunused warnings r319038: getusershell_test: mark mdata parameter in compare_usershell __unused r319039: getserv_test: mark unused parameters __unused to fix corresponding warnings r319040: getrpc_test: fix -Wunused warnings - Mark unused function parameters unused. - Remove an unused function prototype. r319041: getproto_test: fix -Wunused warnings Mark unused parameters __unused in functions. r319042: gethostby_test: fix multiple warning types - Fix -Wmissing-declaration warning by staticizing run_tests. - Fix -Wsign-compare warnings by casting size_t types to int for comparisons. Reindent some of the code in sdump_hostent(..) to accomodate the overall changes. r319043: getpw_test: fix -Wunused warnings - Mark unused parameters __unused. - Put dump_passwd under DEBUG as it's only used in that case. r319044: getgr_test: fix -Wunused warnings r319045: Fix -Wunused and -Wshadow warnings r319046: Fix a -Wunused-but-set-variable warning reported by gcc 6.3.0
* MFC r306691,310036:delphij2017-05-31135-4656/+5441
| | | | Merge byacc 20161202.
* MFC r319062:ngie2017-05-311-0/+1
| | | | | | Initial `srv` before using it in bind(2) CID: 1357526
* MFC r304106:delphij2017-05-318-5/+198
| | | | | | | Add timingsafe_bcmp and timingsafe_memcmp. Obtained from: OpenBSD Reviewed by: trasz
* MFC r316649 (kan):delphij2017-05-311-1/+1
| | | | Add missing double quote to fix r316635 commit.
* MFC r318514-r318515, r318517, r318917delphij2017-05-313-66/+123
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | r318514: Use size_t. Inspired by: OpenBSD src/lib/libc/stdlib/qsort.c,v 1.11 r318515: The current qsort(3) implementation ignores the sizes of partitions, and always perform recursion on the left partition, then use a tail call to handle the right partition. In the worst case this could require O(N) levels of recursions. Reduce the possible recursion level to log2(N) by always recursing on the smaller partition instead. Obtained from: PostgreSQL 9d6077abf9d6efd992a59f05ef5aba981ea32096 r318517: Sync qsort.c with userland r318515. (Note that MIN macro is removed in favor of sys/param.h's version). PR: 213922 r318917: Disconnect heimdal version of qsort.c from build because we are already using libc's version of qsort. PR: bin/213922
* MFC r311989 (cem):delphij2017-05-3111-29/+29
| | | | libkern: Remove obsolete 'register' keyword
* MFC r316635:delphij2017-05-311-13/+3
| | | | | | Enable 16-bit longest_match for x86. This gives a ~2% improvement in compression tests.
* MFC r313327:delphij2017-05-311-1/+1
| | | | Reflect actual NetBSD revision we already have.
* MFC r313695, r313760, r314769, r314863, r314865, r316125delphij2017-05-31184-11667/+24285
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | r313695: MFV r313676: libpcap 1.8.1 r313760: MFV r313759: license change for a few headers (4 clause BSD to 3 clause BSD). X-MFC-with: r313695 r314769: Remove compatibility with old libpcap. Differential Revision: https://reviews.freebsd.org/D9606 r314863: Stop installing pcap-int.h, which is the internal interface for libpcap. Reference: https://github.com/the-tcpdump-group/libpcap/issues/560 PR: 217221 r314865: Bump __FreeBSD_version for removal of pcap-int.h. PR: 217221 r316125: MFV r316124: Fix build when WITHOUT_INET6. Reported by: Randy Westlund <rwestlun gmail com>
* MFC r315272, r315370delphij2017-05-314-1/+44
| | | | | | | | | | | | | | | | | | | | | | | r315272: Implement INHERIT_ZERO for minherit(2). INHERIT_ZERO is an OpenBSD feature. When a page is marked as such, it would be zeroed upon fork(). This would be used in new arc4random(3) functions. PR: 182610 Reviewed by: kib (earlier version) Differential Revision: https://reviews.freebsd.org/D427 r315370: The adj_free and max_free values of new_entry will be calculated and assigned by subsequent vm_map_entry_link(), therefore, remove the pointless copying. Submitted by: alc
* MFC r315615:delphij2017-05-311-2/+2
| | | | Make space style consistent with earlier entries.
* MFC r318975:delphij2017-05-311-4/+6
| | | | | | | | Tighten /entropy permissions. PR: 219527 Reported by: Lu Tung-Pin <lutungpin at openmailbox.org> Submitted by: jilles
* MFC r318774:np2017-05-311-5/+3
| | | | | | | cxgbe/iw_cxgbe: sodisconnect failures are harmless and should not be treated as fatal errors. Sponsored by: Chelsio Communications
* MFC r318762:np2017-05-3114-29344/+32101
| | | | | | | | | | | cxgbe(4): Update the T4, T5, and T6 firmwares to 1.16.45.0. The latest firmware has a number of link related fixes, support for a new custom card, and the fix for a bug that affected rate limiting on FreeBSD. Relnotes: Yes Sponsored by: Chelsio Communications
* MFC r314176 and r318973.np2017-05-307-87/+20686
| | | | | | | | | | | r314176: Add support for RDMA from userspace with T6 cards. r318973: libcxgb4: Use memcpy instead of copying WRs 8B at a time in the userspace RDMA library for cxgbe(4). Sponsored by: Chelsio Communications
* Regenerate src.conf(5)ngie2017-05-301-60/+853
|
* MFC r313169,r319246:ngie2017-05-3016-44/+56
| | | | | | | | | | | r313169 (by wblock): Clarify some option descriptions, add a line of text to makeman to add the slightest hint of a shade of a clue of what it does. r319246: Use .Xr to reference libblacklist(3), blacklistctl(8), and blacklistd(8)
* MFC r317168:ngie2017-05-304-778/+9
| | | | | | | | | | | Add a knob, WITH*_RPCBIND_WARMSTART_SUPPORT, to allow the end-user to build rpcbind(8) with/without warmstart support. The knob defaults to off to preserve POLA for the feature. See rpcbind(8) for more details about the warmstart feature. Relnotes: yes
* MFC r317155,r317162,r317163,r317164:ngie2017-05-302-6/+29
| | | | | | | | | | | | | | | | | | | | | | | r317155: rpcbind(8): wordsmith -h description and mention -W in the SYNOPSIS section -W was already documented in the OPTIONS section. r317162: rpcbind(8): add a description for /var/run/rpcbind.sock under the FILES section r317163: rpcbind(8): post-humously document -w (warmstart) support added in r74462 warmstart support saves portmap/rpcbind(8) registration state on exit and restores the saved registration state on restart. r317164: Fix indentation per style.Makefile(5)
* MFC r318305: makeman: reword description, based on feedback from wblockemaste2017-05-301-7/+7
|
* MFC arch.7 man page describing some arch-specific detailsemaste2017-05-302-0/+352
| | | | | | | | | MFC of r318386, r317938, r317844, r317843, r317840, r317835, r317809, r317790, r317754, r317708, r317681, r312563, r308251, r308130, r307687, r303153, r303134, r303045, r303044, r303033 Changes by br, emaste, jhb, jhibbits, kevlo, kib Sponsored by: The FreeBSD Foundation
* MFC r318316: uniq: allow -c to be used with -d or -uemaste2017-05-302-20/+9
| | | | | | | | | | | | | | | | | | | Bring in some bits from NetBSD and lift the restriction in uniq(1) that -c cannot be used with the -d and -u options. This restriction seems unnecessary and is supported at least by GNU, OpenBSD, and NetBSD. Lift the restriction and simplify the show() logic a little bit to maintain functionality when -c is provided with -d/-u. Also with this change, -d and -u are now actually a mutually exclusive, albeit valid, combination. Given that they both indicate opposite behavior, uniq(1) will no longer output anything if both -d and -u are supplied. This is in line with NetBSD as well as GNU. Adjust the man page and usage() to reflect that -c is its own standalone option. PR: 200553 Submitted by: Kyle Evans <kevans91@ksu.edu>
* MFC r318689: Add parent interface reference counting to if_vlan.mav2017-05-301-14/+26
| | | | Using plain ifunit() looks like a request for troubles.
* MFC r318593:asomers2017-05-302-10/+13
| | | | | | Fix build of AIO tests with -DDEBUG Also, redefine some constants for clarity. No functional change.
* MFC r318141, r318143-r318144asomers2017-05-302-3/+8
| | | | | | | | | | | | | | | | | | | | | | | r318141: strcpy => strlcpy Reported by: Coverity CID: 1352771 Sponsored by: Spectra Logic Corp r318143: strcpy => strlcpy Reported by: Coverity CID: 1006715 Sponsored by: Spectra Logic Corp r318144: Don't depend on assert(3) getting evaluated Reported by: imp X-MFC-With: 318141, 318143 Sponsored by: Spectra Logic Corp
* MFC r307334, r318252: Support write-through caches on arm64andrew2017-05-303-5/+17
| | | | | | | | | | | | r307334: Create macros for the MAIR memory attributes. While here add an uncached memory type, however the VM code still needs to be taught about this. r318252: Add the VM_MEMATTR_WRITE_THROUGH memory type to arm64 and use it to support VM_MEMATTR_WRITE_COMBINING in the kernel. This fixes a bug where Xorg would use write back cached memory for its graphics buffers. This would produce artifacts on the screen as cachelines were written to memory.
* MFC r317824:andrew2017-05-301-1/+122
| | | | | | | | Add reclaim_pv_chunk on arm64. This is based on the amd64 code so should operate similarly, other than not needing the delayed invalidation. It has been tested with artificial injection of vm_page_alloc failures while running 'sort /dev/zero'.
* MFC r317361:andrew2017-05-304-2/+21
| | | | | | Call the PSCI reset from cpu_reset on arm64. When rebooting from DDB the kernel calls this directly so the event handler is not called, meaning the computer fails to reboot.
* MFC r317197:andrew2017-05-301-2/+7
| | | | | | Restrict the arm64 supervisor all instructions to only allow a zero immediate value for system calls. We may wish to use other values in the future for other purposes.
* MFC r317192:andrew2017-05-303-24/+20
| | | | | | Push loading curthread into assembly in the synchronous exception handlers. This will help investigating the performance impact of moving parts of the switch statement in do_el0_sync into assembly.
* MFC r313396:andrew2017-05-304-5/+9
| | | | | Push reading of ESR_EL1 to assembly. Among other uses this will allow us to expose this to signal handlers, e.g. for the clang sanitizers.
* MFC r316734, r316761: Support Execute-Never bit in the arm64 pmap.andrew2017-05-302-13/+41
| | | | | | | | | | | | | | | | | | | | | r316734: Start to use the User and Privileged execute-never bits in the arm64 pagetables. This sets both bits when entering an address we know shouldn't be executed. I expect we could mark all userspace pages as Privileged execute-never to ensure the kernel doesn't branch to one of these addresses. While here add the ARMv8.1 upper attributes. r316761: Set the arm64 Execute-never bits in more places. We need to set the Execute-never bits when mapping device memory as the hardware may perform speculative instruction fetches. Set the Privileged Execute-ever bit on userspace memory to stop the kernel if it is tricked into executing it.
* MFC r316732, r316756: Enable Privileged Access Never on arm64.andrew2017-05-307-13/+82
| | | | | | | | | | | | | | | r316732: Use the unprivileged variant of the load and store instructions most places possible in the kernel. This forces these functions to fail if userspace is unable to access a given memory location, even if it is in the user memory range. This will simplify adding Privileged Access Never support later. r316756: In ARMv8.1 ARM has added a process state bit to disable access to userspace from the kernel. Make use of this to restrict accessing userspace to just the functions that explicitly handle crossing the user kernel boundary.
* MFC r316755:andrew2017-05-302-5/+10
| | | | | Add SCTLR bits added in ARMv8.1 and ARMv8.2 and start to use them in the early boot code.
* MFC r312703:andrew2017-05-301-1/+1
| | | | | | Fix the error value we write in cerror. __error returns an int *, however we were writing a 64 bit value meaning the 32 bits after this would be trashed.
* Regenerate src.conf(5) after MK_DIALOG was addedngie2017-05-301-1/+9
|
* MFC r314463:ngie2017-05-301-1/+1
| | | | | | Use SRCTOP-relative paths to other directories instead of .CURDIR-relative ones This simplifies pathing in make/displayed output.
* MFC r306375,r307802:ngie2017-05-3010-66/+142
| | | | | | | | | | | | | | | | | | r306375 (by emaste): Add a WITHOUT_DIALOG src.conf(5) knob It also turns off dependencies (bsdinstall, bsdconfig, dpv, tzsetup). r307802 (by bapt): Fix build of tzsetup when WITHOUT_DIALOG is set Hide dialog specific code behind HAVE_DIALOG. It allows to build a stripped down version (missing the dialog UI) but perfectly function tzsetup when world is built WITHOUT_DIALOG Reorganise a bit the code to limit the number of blocks under HAVE_DIALOG
* MFC ↵ngie2017-05-3040-158/+153
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | r314451,r314452,r314453,r314456,r314457,r314458,r314459,r314460,r314461,r314462,r314464,r314466,r314468,r314470,r314472,r314478: r314451: Use SRCTOP-relative paths to other directories instead of .CURDIR-relative ones The SRCTOP conversion simplifies pathing in make/displayed output. Also, while here, change a hardcoded path to ar5523.bin.uu in the make target with ${.ALLSRC} . r314452: Use SRCTOP-relative paths to other directories instead of .CURDIR-relative ones r314453: Use SRCTOP-relative paths to other directories instead of .CURDIR-relative ones r314456: Use SRCTOP-relative paths to other directories instead of .CURDIR-relative ones This simplifies pathing in make/displayed output. r314457: Use SRCTOP-relative paths to other directories instead of .CURDIR-relative ones The SRCTOP conversion simplifies pathing in make/displayed output. Also, while here, change a hardcoded path to ${RPCDIR}/ypupdate_prot.x in the make targets with ${.ALLSRC} . r314458: Use SRCTOP-relative paths to other directories instead of .CURDIR-relative ones This simplifies pathing in make/displayed output. r314459: Use SRCTOP-relative paths to other directories instead of .CURDIR-relative ones This simplifies pathing in make/displayed output. r314460: Use SRCTOP-relative paths to other directories instead of .CURDIR-relative ones This simplifies pathing in make/displayed output. r314461: Use SRCTOP-relative paths to other directories instead of .CURDIR-relative ones This simplifies pathing in make/displayed output. r314462: Use SRCTOP-relative paths to other directories instead of .CURDIR-relative ones The SRCTOP conversion simplifies pathing in make/displayed output. Also, while here, change a hardcoded path to ${RPCDIR}/ypxfrd.x in the make targets with ${.ALLSRC} . r314464: Use SRCTOP-relative paths to other directories instead of .CURDIR-relative ones This simplifies pathing in make/displayed output. r314466: Use SRCTOP-relative paths to other directories instead of .CURDIR-relative ones This simplifies pathing in make/displayed output. r314468: Use SRCTOP-relative paths to other directories instead of .CURDIR-relative ones Also, use :H where possible/sensical to manipulate .CURDIR-relative paths This simplifies pathing in make/displayed output. r314470: Use SRCTOP-relative paths to other directories instead of .CURDIR-relative ones This simplifies pathing in make/displayed output. r314472: Simplify/improve idioms in usr.sbin/ntp Makefiles - Use SRCTOP-relative paths to other directories instead of .CURDIR-relative ones. This simplifies pathing in make/displayed output. - Also, use :H where possible/sensical to manipulate .CURDIR-relative paths - Remove superfluous bsd.own.mk .includes which are already handled via src.opts.mk .includes r314478: Simplify idioms in Makefiles further - Use SRCTOP-relative paths instead of .CURDIR-relative ones where possible - Use :H to manipulate .CURDIR in areas instead of ..-relative paths.
* MFC r314541:ngie2017-05-307-11/+11
| | | | | | Use SRCTOP-relative paths to other directories instead of .CURDIR-relative ones This simplifies pathing in make/displayed output.
* MFC r314658:ngie2017-05-306-11/+11
| | | | | | crypto: normalize paths using SRCTOP-relative paths or :H when possible This simplifies make logic/output
* MFC r318439:ngie2017-05-301-1/+1
| | | | | | | | | | | Normalize .PATH on SRCTOP This will help Jenkins dedupe 9 warnings between the static build and the module build of ipsec(4). Missed in SRCTOP conversion in r314651. MFC with: r314651
* MFC r307741:ngie2017-05-301-1/+1
| | | | | | | | | r307741 (by emaste): Allow all subdirectories to be optional via SUBDIR.${MK_*} Tested by: br Pointy hat to: emaste
* MFC r315170:ngie2017-05-3059-96/+96
| | | | | | | | | | | r315170 (by imp): Adopt SRCTOP in usr.bin Prefer ${SRCTOP}/foo over ${.CURDIR}/../../foo and ${SRCTOP}/usr.bin/foo over ${.CURDIR}/../foo for paths in Makefiles. Silence on: arch@ (twice)
OpenPOWER on IntegriCloud