aboutsummaryrefslogtreecommitdiff
AgeCommit message (Collapse)AuthorFilesLines
2020-12-30powerpc: Runtime selection between sc and scv for syscallsMatheus Castanho6-40/+214
Linux kernel v5.9 added support for system calls using the scv instruction for POWER9 and later. The new codepath provides better performance (see below) if compared to using sc. For the foreseeable future, both sc and scv mechanisms will co-exist, so this patch enables glibc to do a runtime check and use scv when it is available. Before issuing the system call to the kernel, we check hwcap2 in the TCB for PPC_FEATURE2_SCV to see if scv is supported by the kernel. If not, we fallback to sc and keep the old behavior. The kernel implements a different error return convention for scv, so when returning from a system call we need to handle the return value differently depending on the instruction we used to enter the kernel. For syscalls implemented in ASM, entry and exit are implemented by different macros (PSEUDO and PSEUDO_RET, resp.), which may be used in sequence (e.g. for templated syscalls) or with other instructions in between (e.g. clone). To avoid accessing the TCB a second time on PSEUDO_RET to check which instruction we used, the value read from hwcap2 is cached on a non-volatile register. This is not needed when using INTERNAL_SYSCALL macro, since entry and exit are bundled into the same inline asm directive. The dynamic loader may issue syscalls before the TCB has been setup so it always uses sc with no extra checks. For the static case, there is no compile-time way to determine if we are inside startup code, so we also check the value of the thread pointer before effectively accessing the TCB. For such situations in which the availability of scv cannot be determined, sc is always used. Support for scv in syscalls implemented in their own ASM file (clone and vfork) will be added later. For now simply use sc as before. Average performance over 1M calls for each syscall "type": - stat: C wrapper calling INTERNAL_SYSCALL - getpid: templated ASM syscall - syscall: call to gettid using syscall function Standard: stat : 1.573445 us / ~3619 cycles getpid : 0.164986 us / ~379 cycles syscall : 0.162743 us / ~374 cycles With scv: stat : 1.537049 us / ~3535 cycles <~ -84 cycles / -2.32% getpid : 0.109923 us / ~253 cycles <~ -126 cycles / -33.25% syscall : 0.116410 us / ~268 cycles <~ -106 cycles / -28.34% Tested on powerpc, powerpc64, powerpc64le (with and without scv) Tested-by: Lucas A. M. Magalhães <lamm@linux.ibm.com> Reviewed-by: Tulio Magno Quites Machado Filho <tuliom@linux.ibm.com>
2020-12-30malloc: preserve errno on mcheck hooks [BZ #17924]Adhemerval Zanella1-5/+11
Similar to the fix 69fda43b8d, save and restore errno for the hook functions used for MALLOC_CHECK_=3. It fixes the malloc/tst-free-errno-mcheck regression. Checked on x86_64-linux-gnu.
2020-12-30x86 long double: Add tests for pseudo normal numbersSiddhesh Poyarekar5-0/+94
Add some tests for fpclassify, isnan, isinf and issignaling. Co-authored-by: Adhemerval Zanella <adhemerval.zanella@linaro.org> Reviewed-by: Adhemerval Zanella <adhemerval.zanella@linaro.org>
2020-12-30x86 long double: Consider pseudo numbers as signalingSiddhesh Poyarekar3-3/+69
Add support to treat pseudo-numbers specially and implement x86 version to consider all of them as signaling. Reviewed-by: Adhemerval Zanella <adhemerval.zanella@linaro.org>
2020-12-29io: Remove xmknod{at} implementationsAdhemerval Zanella9-166/+10
With xmknod wrapper functions removed (589260cef8), the mknod functions are now properly exported, and version is done using symbols versioning instead of the extra _MKNOD_* argument. It also allows us to consolidate Linux and Hurd mknod implementation. Reviewed-by: Lukasz Majewski <lukma@denx.de>
2020-12-29io: Remove xstat implementationsAdhemerval Zanella19-341/+45
With xstat wrapper functions removed (8ed005daf0), the stat functions are now properly exported, and version is done using symbols versioning instead of the extra _STAT_* argument. Reviewed-by: Lukasz Majewski <lukma@denx.de>
2020-12-29free: preserve errno [BZ#17924]Paul Eggert4-4/+150
In the next release of POSIX, free must preserve errno <https://www.austingroupbugs.net/view.php?id=385>. Modify __libc_free to save and restore errno, so that any internal munmap etc. syscalls do not disturb the caller's errno. Add a test malloc/tst-free-errno.c (almost all by Bruno Haible), and document that free preserves errno. Reviewed-by: Adhemerval Zanella <adhemerval.zanella@linaro.org>
2020-12-29hurd: Accept including hurd/version.hSamuel Thibault1-1/+1
We need it to get the RPC API version.
2020-12-28hurd: Add WSTOPPED/WCONTINUED/WEXITED/WNOWAIT support [BZ #23091]Samuel Thibault3-17/+43
The new __proc_waitid RPC now expects WEXITED to be passed, allowing to properly implement waitid, and thus define the missing W* macros (according to FreeBSD values).
2020-12-28support: Make support_process_state_wait wait lessSamuel Thibault1-1/+1
Tests such as posix/tst-waitid.c make heavy use of support_process_state_wait, and thus on non-Linux where it falls back to sleeping, a 2s sleep makes such test time out, while 1s remains fine enough.
2020-12-26hurd: set sigaction for signal preemptors in arch-independent fileSamuel Thibault3-22/+20
Instead of having the arch-specific trampoline setup code detect whether preemption happened or not, we'd rather pass it the sigaction. In the future, this may also allow to change sa_flags from post_signal().
2020-12-26hurd: Fix spawni SPAWN_XFLAGS_TRY_SHELL with empty argvSamuel Thibault1-1/+5
When argv is empty, we need to add the original script to be run on the shell command line.
2020-12-26hurd: Try shell in posix_spawn* only in compat modeSamuel Thibault1-1/+1
Reported by Bruno Haible <bruno@clisp.org>
2020-12-24Remove _ISOMAC check from <cpu-features.h>H.J. Lu1-81/+75
Remove _ISOMAC check from <cpu-features.h> since it isn't an installer header file.
2020-12-24x86: Remove the duplicated CPU_FEATURE_CPU_PH.J. Lu1-2/+0
CPU_FEATURE_CPU_P is defined in sysdeps/x86/sys/platform/x86.h. Remove the duplicated CPU_FEATURE_CPU_P in sysdeps/x86/include/cpu-features.h.
2020-12-24Partially revert 681900d29683722b1cb0a8e565a0585846ec5a61Siddhesh Poyarekar2-12/+1
Do not attempt to fix the significand top bit in long double input received in printf. The code should never reach here because isnan should now detect unnormals as NaN. This is already a NOP for glibc since it uses the gcc __builtin_isnan, which detects unnormals as NaN. Reviewed-by: Florian Weimer <fweimer@redhat.com>
2020-12-24x86 long double: Support pseudo numbers in isnanlSiddhesh Poyarekar2-6/+7
This syncs up isnanl behaviour with gcc. Also move the isnanl implementation to sysdeps/x86 and remove the sysdeps/x86_64 version. Reviewed-by: Adhemerval Zanella <adhemerval.zanella@linaro.org>
2020-12-24x86 long double: Support pseudo numbers in fpclassifylSiddhesh Poyarekar2-2/+4
Also move sysdeps/i386/fpu/s_fpclassifyl.c to sysdeps/x86/fpu/s_fpclassifyl.c and remove sysdeps/x86_64/fpu/s_fpclassifyl.c Reviewed-by: Adhemerval Zanella <adhemerval.zanella@linaro.org>
2020-12-24MTE: Do not pad size in realloc_checkSiddhesh Poyarekar1-4/+5
The MTE patch to add malloc support incorrectly padded the size passed to _int_realloc by SIZE_SZ when it ought to have sent just the chunksize. Revert that bit of the change so that realloc works correctly with MALLOC_CHECK_ set. This also brings the realloc_check implementation back in sync with libc_realloc.
2020-12-24tests-mcheck: New variable to run tests with MALLOC_CHECK_=3Siddhesh Poyarekar2-1/+38
This new variable allows various subsystems in glibc to run all or some of their tests with MALLOC_CHECK_=3. This patch adds infrastructure support for this variable as well as an implementation in malloc/Makefile to allow running some of the tests with MALLOC_CHECK_=3. At present some tests in malloc/ have been excluded from the mcheck tests either because they're specifically testing MALLOC_CHECK_ or they are failing in master even without the Memory Tagging patches that prompted this work. Some tests were reviewed and found to need specific error points that MALLOC_CHECK_ defeats by terminating early but a thorough review of all tests is needed to bring them into mcheck coverage. The following failures are seen in current master: FAIL: malloc/tst-malloc-fork-deadlock-mcheck FAIL: malloc/tst-malloc-stats-cancellation-mcheck FAIL: malloc/tst-malloc-thread-fail-mcheck FAIL: malloc/tst-realloc-mcheck FAIL: malloc/tst-reallocarray-mcheck All of these are due to the Memory Tagging patchset and will be fixed separately.
2020-12-23elf: Account for glibc-hwcaps/ prefix in _dl_important_hwcapsFlorian Weimer1-0/+1
2020-12-23misc: Use __ferror_unlocked instead of ferrorSiddhesh Poyarekar1-1/+1
The ferror results in an unnecessary PLT reference. Use __ferror_unlocked instead , which gets inlined.
2020-12-22s390x: Regenerate ulpsFlorian Weimer1-10/+12
For new inputs added in commit cad5ad81d2f7f58a7ad0d8afa8c1b710, as seen on a z13 system.
2020-12-22powerpc: Regenerate ulpsFlorian Weimer1-12/+13
For new inputs added in commit cad5ad81d2f7f58a7ad0d8afa8c1b710, as seen on a POWER8 system.
2020-12-22addmntent: Remove unbounded alloca usage from getmntent [BZ#27083]Siddhesh Poyarekar3-74/+140
The addmntent function replicates elements of struct mnt on stack using alloca, which is unsafe. Put characters directly into the stream, escaping them as they're being written out. Also add a test to check all escaped characters with addmntent and getmntent.
2020-12-22<sys/platform/x86.h>: Add Intel LAM supportH.J. Lu3-0/+7
Add Intel Linear Address Masking (LAM) support to <sys/platform/x86.h>. HAS_CPU_FEATURE (LAM) can be used to detect if LAM is enabled in CPU. LAM modifies the checking that is applied to 64-bit linear addresses, allowing software to use of the untranslated address bits for metadata.
2020-12-21i386: Regenerate ulpsFlorian Weimer2-10/+10
For new inputs added in commit cad5ad81d2f7f58a7ad0d8afa8c1b710.
2020-12-21aarch64: update ulps.Szabolcs Nagy1-10/+12
For new test cases in commit cad5ad81d2f7f58a7ad0d8afa8c1b7101a0301fb
2020-12-21aarch64: Add aarch64-specific files for memory tagging supportRichard Earnshaw6-0/+235
This final patch provides the architecture-specific implementation of the memory-tagging support hooks for aarch64.
2020-12-21aarch64: Add sysv specific enabling code for memory taggingRichard Earnshaw4-0/+34
Add various defines and stubs for enabling MTE on AArch64 sysv-like systems such as Linux. The HWCAP feature bit is copied over in the same way as other feature bits. Similarly we add a new wrapper header for mman.h to define the PROT_MTE flag that can be used with mmap and related functions. We add a new field to struct cpu_features that can be used, for example, to check whether or not certain ifunc'd routines should be bound to MTE-safe versions. Finally, if we detect that MTE should be enabled (ie via the glibc tunable); we enable MTE during startup as required. Support in the Linux kernel was added in version 5.10. Reviewed-by: Szabolcs Nagy <szabolcs.nagy@arm.com>
2020-12-21linux: Add compatibility definitions to sys/prctl.h for MTERichard Earnshaw1-0/+18
Older versions of the Linux kernel headers obviously lack support for memory tagging, but we still want to be able to build in support when using those (obviously it can't be enabled on such systems). The linux kernel extensions are made to the platform-independent header (linux/prctl.h), so this patch takes a similar approach.
2020-12-21malloc: Basic support for memory tagging in the malloc() familyRichard Earnshaw5-97/+437
This patch adds the basic support for memory tagging. Various flavours are supported, particularly being able to turn on tagged memory at run-time: this allows the same code to be used on systems where memory tagging support is not present without neededing a separate build of glibc. Also, depending on whether the kernel supports it, the code will use mmap for the default arena if morecore does not, or cannot support tagged memory (on AArch64 it is not available). All the hooks use function pointers to allow this to work without needing ifuncs. Reviewed-by: DJ Delorie <dj@redhat.com>
2020-12-21elf: Add a tunable to control use of tagged memoryRichard Earnshaw2-0/+44
Add a new glibc tunable: mem.tagging. This is a decimal constant in the range 0-255 but used as a bit-field. Bit 0 enables use of tagged memory in the malloc family of functions. Bit 1 enables precise faulting of tag failure on platforms where this can be controlled. Other bits are currently unused, but if set will cause memory tag checking for the current process to be enabled in the kernel.
2020-12-21config: Allow memory tagging to be enabled when configuring glibcRichard Earnshaw6-0/+69
This patch adds the configuration machinery to allow memory tagging to be enabled from the command line via the configure option --enable-memory-tagging. The current default is off, though in time we may change that once the API is more stable.
2020-12-21alpha: Remove anonymous union in struct stat [BZ #27042]Matt Turner3-63/+66
This is clever, but it confuses downstream detection in at least zstd and GNOME's glib. zstd has preprocessor tests for the 'st_mtime' macro, which is not provided by the path using the anonymous union; glib checks for the presence of 'st_mtimensec' in struct stat but then tries to access that field in struct statx (which might be a bug on its own). Checked with a build for alpha-linux-gnu.
2020-12-21add inputs to auto-libm-test-in yielding larger errors (binary64, x86_64)Paul Zimmermann10-11/+581
2020-12-21m68k: fix clobbering a5 in setjmp() [BZ #24202]Sergei Trofimovich1-0/+1
setjmp() uses C code to store current registers into jmp_buf environment. -fstack-protector-all places canary into setjmp() prologue and clobbers 'a5' before it gets saved. The change inhibits stack canary injection to avoid clobber.
2020-12-21iconv add iconv_close before the function returned with bad value.liqingqing1-0/+1
add iconv_close before the function returned with bad value.
2020-12-21iconv: use iconv_close after iconv_openliqingqing1-0/+2
2020-12-21Fix buffer overrun in EUC-KR conversion module (bz #24973)Andreas Schwab4-9/+59
The byte 0xfe as input to the EUC-KR conversion denotes a user-defined area and is not allowed. The from_euc_kr function used to skip two bytes when told to skip over the unknown designation, potentially running over the buffer end.
2020-12-21hurd: Make trampoline fill siginfo ss_sp from sc_uespSamuel Thibault1-1/+1
Mach actually rather fills the uesp field, not esp.
2020-12-21Hurd: make sigstates hold a reference on thread portsRichard Braun2-7/+29
This change is required in order to correctly release per-thread resources. Directly reusing the threading library reference isn't possible since the sigstate is also used early in the main thread, before threading is initialized. * hurd/hurd/signal.h (_hurd_self_sigstate): Drop thread reference after calling _hurd_thread_sigstate. (_hurd_critical_section_lock): Likewise. * hurd/hurdsig.c (_hurd_thread_sigstate): Add a reference on the thread. (_hurd_sigstate_delete): Drop thread reference.
2020-12-21profil-counter: Add missing SIGINFO caseSamuel Thibault1-0/+10
When SA_SIGINFO is available, sysdeps/posix/s?profil.c use it, so we have to fix the __profil_counter function accordingly, using sigcontextinfo.h's sigcontext_get_pc.
2020-12-21hurd: implement SA_SIGINFO signal handlers.Jeremie Koenig11-67/+209
SA_SIGINFO is actually just another way of expressing what we were already passing over with struct sigcontext. This just introduces the SIGINFO interface and fixes the posix values when that interface is requested by the application.
2020-12-20hurd: Fix ELF_MACHINE_USER_ADDRESS_MASK valueSamuel Thibault1-1/+1
x86 binaries are linked at 0x08000000, so we need to let them get mapped there.
2020-12-20hurd: Note when the vm_map kernel bug was fixedSamuel Thibault1-1/+1
dl-sysdep has been wanting to use high bits in the vm_map mask for decades, but that was only implemented lately.
2020-12-19hurd: Also turn KERN_INVALID_ADDRESS to EINVALSamuel Thibault1-0/+1
When e.g. mmap is passed an invalid address we would return KERN_INVALID_ADDRESS, while POSIX applications would expect EINVAL.
2020-12-18ieee754: Remove unused __sin32 and __cos32Anssi Hannula7-86/+0
The __sin32 and __cos32 functions were only used in the now removed slow path of asin and acos.
2020-12-18ieee754: Remove slow paths from asin and acosAnssi Hannula1-61/+15
asin and acos have slow paths for rounding the last bit that cause some calls to be 500-1500x slower than average calls. These slow paths are rare, a test of a trillion (1.000.000.000.000) random inputs between -1 and 1 showed 32870 slow calls for acos and 4473 for asin, with most occurrences between -1.0 .. -0.9 and 0.9 .. 1.0. The slow paths claim correct rounding and use __sin32() and __cos32() (which compare two result candidates and return the closest one) as the final step, with the second result candidate (res1) having a small offset applied from res. This suggests that res and res1 are intended to be 1 ULP apart (which makes sense for rounding), barring bugs, allowing us to pick either one and still remain within 1 ULP of the exact result. Remove the slow paths as the accuracy is better than 1 ULP even without them, which is enough for glibc. Also remove code comments claiming correctly rounded results. After slow path removal, checking the accuracy of 14.400.000.000 random asin() and acos() inputs showed only three incorrectly rounded (error > 0.5 ULP) results: - asin(-0x1.ee2b43286db75p-1) (0.500002 ULP, same as before) - asin(-0x1.f692ba202abcp-4) (0.500003 ULP, same as before) - asin(-0x1.9915e876fc062p-1) (0.50000000001 ULP, previously exact) The first two had the same error even before this commit, and they did not use the slow path at all. Checking 4934 known randomly found previously-slow-path asin inputs shows 25 calls with incorrectly rounded results, with a maximum error of 0.500000002 ULP (for 0x1.fcd5742999ab8p-1). The previous slow-path code rounded all these inputs correctly (error < 0.5 ULP). The observed average speed increase was 130x. Checking 36240 known randomly found previously-slow-path acos inputs shows 42 calls with incorrectly rounded results, with a maximum error of 0.500000008 ULP (for 0x1.f63845056f35ep-1). The previous "exact" slow-path code showed 34 calls with incorrectly rounded results, with the same maximum error of 0.500000008 ULP (for 0x1.f63845056f35ep-1). The observed average speed increase was 130x. The functions could likely be trimmed more while keeping acceptable accuracy, but this at least gets rid of the egregiously slow cases. Tested on x86_64.
2020-12-18getenv: Move call to strlen to the branch it's used in.Lode Willems1-1/+1
The len variable is only used in the else branch. We don't need the call to strlen if the name is 0 or 1 characters long. 2019-10-02 Lode Willems <Lode.Willems@UGent.be> * tdlib/getenv.c: Move the call to strlen into the branch it's used.
generated by cgit v1.1 at 2024-10-05 09:49:32 +0000