about summary refs log tree commit diff
Commit message (Collapse)AuthorAgeFilesLines
* Fix aio_error thread-safety.Samuel Thibault2015-03-063-8/+57
| | | | | | * sysdeps/pthread/aio_error.c: New file * sysdeps/pthread/aio_misc.c: Remove optimistic comment about synchronization.
* vfprintf: Define WORK_BUFFER_SIZEFlorian Weimer2015-03-062-13/+15
| | | | | This constant will allow us to refer to the number of elements in work_buffer across a function call boundary.
* vfprintf: Introduce JUMP_TABLE_BASE_LABELFlorian Weimer2015-03-062-3/+8
| | | | | This makes the offset handling more explicit and avoids cross-references between the jump tables.
* vfprintf: Introduce THOUSANDS_SEP_TFlorian Weimer2015-03-062-18/+10
| | | | This avoids preprocessor conditionals in function declarations.
* manual: complete example in error message documentationRical Jasan2015-03-062-6/+20
| | | | | | | | | | | | | | | | | | | | | | | | The manual gives "an example showing how to handle failure to open a file correctly." The example function, open_sesame, uses the newly-introduced strerror function and errno and program_invocation_short_name variables. It fails to specify GNU extensions, however, so attempts to use it in the following way: int main (void) {open_sesame ("badname");} fail during compilation with "error: ‘program_invocation_short_name’ undeclared", indicating the example is incomplete. The presence of "#include"s suggest everything neccesary for the function to work should be present. For completeness, the example is lacking the following line: #define _GNU_SOURCE as the declarations of program_invocation_*name in errno.h are wrapped in an "#ifdef __USE_GNU" conditional. The documentation of the variables is also expanded, adding that their definition lies in errno.h and noting specifically they are GNU extensions.
* manual: drop strerror C89 compatibility noteMike Frysinger2015-03-062-3/+5
| | | | | Mentioning systems that are old and don't support C89 is probably a waste of time nowadays. Drop the note.
* Avoid re-exec-self in bug-setlocale1.Roland McGrath2015-03-056-43/+32
|
* Replace ELF_RTYPE_CLASS_NOCOPY with ELF_RTYPE_CLASS_COPYH.J. Lu2015-03-0518-17/+46
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | ELF_RTYPE_CLASS_NOCOPY in comments is a typo. It should be ELF_RTYPE_CLASS_COPY. [BZ #18082] * sysdeps/alpha/dl-machine.h (elf_machine_type_class): Replace ELF_RTYPE_CLASS_NOCOPY with ELF_RTYPE_CLASS_COPY in comments. * sysdeps/arm/dl-machine.h (elf_machine_type_class): Likewise. * sysdeps/hppa/dl-machine.h (elf_machine_type_class): Likewise. * sysdeps/i386/dl-machine.h (elf_machine_type_class): Likewise. * sysdeps/ia64/dl-machine.h (elf_machine_type_class): Likewise. * sysdeps/m68k/dl-machine.h (elf_machine_type_class): Likewise. * sysdeps/microblaze/dl-machine.h (elf_machine_type_class): Likewise. * sysdeps/nios2/dl-machine.h (elf_machine_type_class): Likewise. * sysdeps/powerpc/powerpc32/dl-machine.h (elf_machine_type_class): Likewise. * sysdeps/powerpc/powerpc64/dl-machine.h (elf_machine_type_class): Likewise. * sysdeps/s390/s390-32/dl-machine.h (elf_machine_type_class): Likewise. * sysdeps/s390/s390-64/dl-machine.h (elf_machine_type_class): Likewise. * sysdeps/sh/dl-machine.h (elf_machine_type_class): Likewise. * sysdeps/sparc/sparc32/dl-machine.h (elf_machine_type_class): Likewise. * sysdeps/sparc/sparc64/dl-machine.h (elf_machine_type_class): Likewise. * sysdeps/tile/dl-machine.h (elf_machine_type_class): Likewise. * sysdeps/x86_64/dl-machine.h (elf_machine_type_class): Likewise.
* Define ETH_ALEN in generic <netinet/if_ether.h>.Roland McGrath2015-03-042-3/+7
|
* Split rpcent tests out of tst-netdb.Roland McGrath2015-03-044-44/+82
|
* Conditionalize some test code for SIGRTMIN, SA_SIGINFO.Roland McGrath2015-03-046-11/+41
|
* Avoid more C++ tests.Roland McGrath2015-03-042-3/+4
|
* Avoid C++ tests when the C++ cannot be linked.Roland McGrath2015-03-046-92/+216
|
* Don't define __CORRECT_ISO_CPP_STRING_H_PROTO for non-GCC compilers (bug 17631)Andreas Schwab2015-03-044-16/+16
| | | | | The implementation of __CORRECT_ISO_CPP_STRING_H_PROTO requires support for asm aliases.
* powerpc: Fix inline feraiseexcept, feclearexcept macrosAdhemerval Zanella2015-03-035-25/+398
| | | | | | | This patch fixes the inline feraiseexcept and feclearexcept macros for powerpc by casting the input argument to integer before operation on it. It fixes BZ#17776.
* Fix localplt test breakage with new readelfAlan Modra2015-03-033-7/+13
| | | | | | | | | Since 2014-11-24 binutils git commit bb4d2ac2, readelf has appended the symbol version to symbols shown in reloc dumps. [BZ #16512] * scripts/localplt.awk: Strip off symbol version. * NEWS: Mention bug fix.
* Deglobalize internal variables in timer_routines.c.Roland McGrath2015-03-022-3/+8
|
* Correct __ASSUME_PRLIMIT64 for hppa/microblaze/sh (bug 17779).Joseph Myers2015-03-025-4/+32
| | | | | | | | | | | | | | | | | | | | | | | | | | __ASSUME_PRLIMIT64 is defined in kernel-features.h for kernels 2.6.36 and later, but hppa, microblaze and sh did not add the prlimit64 syscall until 2.6.37. This patch adds corresponding undefines of __ASSUME_PRLIMIT64 to those architectures' kernel-features.h files. (This concludes the kernel-features.h fixes arising out of the review - limited to macros defined in the architecture-independent kernel-features.h file - I did in connection with the move to 2.6.32 minimum kernel version. For that subset of macros - I didn't check any purely architecture-specific macros - I think they are now defined for the correct kernel versions on each architecture after this patch.) [BZ #17779] * sysdeps/unix/sysv/linux/hppa/kernel-features.h [__LINUX_KERNEL_VERSION < 0x020625] (__ASSUME_PRLIMIT64): Undefine. * sysdeps/unix/sysv/linux/microblaze/kernel-features.h [__LINUX_KERNEL_VERSION < 0x020625] (__ASSUME_PRLIMIT64): Likewise. * sysdeps/unix/sysv/linux/sh/kernel-features.h [__LINUX_KERNEL_VERSION < 0x020625] (__ASSUME_PRLIMIT64): Likewise.
* Fix BZ 18036 buffer overflow (read past end of buffer) in internal_fnmatchPaul Pluzhnikov2015-03-024-4/+35
|
* Fix parallel build errorAndreas Schwab2015-03-022-1/+6
|
* _nss_nis_initgroups_dyn: Return status instead of NSS_STATUS_SUCCESSFlorian Weimer2015-03-022-1/+6
|
* Avoid unsafe loc_name type casts with additional variableAlexandre Oliva2015-02-272-23/+28
| | | | | | | | for ChangeLog [BZ #15969] * locale/findlocale.c (_nl_find_locale): Introduce const version of loc_name and drop unsafe type casts.
* Convert dlfcn/tststatic2 to use test-skeleton.Roland McGrath2015-02-272-2/+12
|
* Compile vismain with -fPIE and link with -pieH.J. Lu2015-02-274-5/+21
| | | | | | | | | | | | Protocted symbol in shared library can only be accessed from PIE or shared library. Linker in binutils 2.26 enforces it. We must compile vismain with -fPIE and link it with -pie. [BZ #17711] * elf/Makefile (tests): Add vismain only if PIE is enabled. (tests-pie): Add vismain. (CFLAGS-vismain.c): New. * elf/vismain.c: Add comments for PIE requirement.
* Fix ldbl-96, ldbl-128ibm atanhl inaccuracy (bug 18046, bug 18047).Joseph Myers2015-02-278-4/+930
| | | | | | | | | | | | | | | | | | | | | | | | | The threshold in ldbl-96 atanhl for when to return the argument, 0x1p-28, is a bit too big, and that in ldbl-128ibm atanhl is much too big (the relevant condition being x^3/3 being < 0.5ulp of x), resulting in errors a bit above the limits of those considered acceptable in glibc in the ldbl-96 case, and in large errors in the ldbl-128ibm case. This patch changes those implementations to use more appropriate thresholds and adds tests around the thresholds for various formats. Tested for x86_64, x86 and powerpc. x86_64 and x86 ulps updated accordingly. [BZ #18046] [BZ #18047] * sysdeps/ieee754/ldbl-128ibm/e_atanhl.c (__ieee754_atanhl): Use 0x1p-56L as threshold for just returning the argument. * sysdeps/ieee754/ldbl-96/e_atanhl.c (__ieee754_atanhl): Use 0x1p-32L as threshold for just returning the argument. * math/auto-libm-test-in: Add more tests of atanh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulp: Likewise.
* Rather than using a C implementation of memmove, directly call memmove, whichWilco Dijkstra2015-02-272-9/+9
| | | | typically has a much faster optimized implementation.
* Rather than using a C implementation of memset, directly call memset, whichWilco Dijkstra2015-02-272-55/+6
| | | | typically has a much faster optimized implementation.
* hppa: fix __O_SYNC to match the kernelJohn David Anglin2015-02-272-1/+6
|
* Add comment to CSTR macro in k_standard.c.Joseph Myers2015-02-272-0/+5
| | | | * sysdeps/ieee754/k_standard.c (CSTR): Add comment.
* Avoid -Wno-write-strings for k_standard.c.Joseph Myers2015-02-263-81/+59
| | | | | | | | | | | | | | | | | | | We want to avoid -Wno- options in makefiles as far as possible, by cleaning up the underlying issues if possible or failing that by using diagnostic pragmas. This patch eliminates the use of -Wno-write-strings for sysdeps/ieee754/k_standard.c by using casts in the source file to cast away const; those casts are encapsulated in a macro that also deals with the choice of strings for float / double / long double functions (for which the logic was previously replicated many times). Tested for x86_64; the only change to disassembly of installed stripped shared libraries was a line number in an assertion. * sysdeps/ieee754/k_standard.c (CSTR): New macro. (__kernel_standard): Use CSTR macro when setting exc.name. * sysdeps/ieee754/Makefile [$(subdir) = math] (CFLAGS-k_standard.c): Remove variable.
* Avoid uninitialized warnings in Bessel functions.Joseph Myers2015-02-267-12/+38
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | math/Makefile currently has: # The fdlibm code generates a lot of these warnings but is otherwise clean. override CFLAGS += -Wno-uninitialized This is of course undesirable; warnings should be disabled as narrowly as possible. To remove this override, we need to fix files that generate such warnings, or put warning-disabling pragmas in them. This patch does so for Bessel function implementations, one of the cases that have the warnings if the override is removed. The warnings arise because functions set pointer variables p and q only for certain values of the function argument, then use them unconditionally. As the static functions in question only get called for arguments that satisfy the last condition in the if/else chain, the natural fix is to change the last "else if" to just "else", which this patch does. (The ldbl-128 / ldbl-128ibm implementation of these functions is substantially different and looks like it already does use "else" in the last case in the nearest corresponding code.) Tested for x86_64 and x86. * sysdeps/ieee754/dbl-64/e_j0.c (pzero): Change last case for setting p and q from "else if" to "else". (qzero): Likewise. * sysdeps/ieee754/dbl-64/e_j1.c (pone): Likewise. (qone): Likewise. * sysdeps/ieee754/flt-32/e_j0f.c (pzerof): Likewise. (qzerof): Likewise. * sysdeps/ieee754/flt-32/e_j1f.c (ponef): Likewise. (qonef): Likewise. * sysdeps/ieee754/ldbl-96/e_j0l.c (pzero): Likewise. (qzero): Likewise. * sysdeps/ieee754/ldbl-96/e_j1l.c (pone): Likewise. (qone): Likewise.
* Fix ldbl-128/ldbl-128ibm acosl inaccuracy (bug 18038, bug 18039).Joseph Myers2015-02-266-3/+818
| | | | | | | | | | | | | | | | | | | | | The ldbl-128 and ldbl-128ibm implementations of acosl have similar bugs, using a threshold of 0x1p-57L to determine when they just return pi/2. Since the result pi/2 - asinl (x) is roughly pi/2 - x for small x, the relevant cut-off is actually x being < 0.5ulp of 1. This patch fixes the implementations to use that cut-off and adds tests of small acos arguments. Tested for powerpc and mips64. Also tested for x86_64 and x86; no ulps updates needed. [BZ #18038] [BZ #18039] * sysdeps/ieee754/ldbl-128/e_acosl.c (__ieee754_acosl): Only return pi/2 for arguments below 0x1p-113L. * sysdeps/ieee754/ldbl-128ibm/e_acosl.c (__ieee754_acosl): Only return pi/2 for arguments below 0x1p-106L. * math/auto-libm-test-in: Add more tests of acos. * math/auto-libm-test-out: Regenerated.
* Fix asin missing underflows (bug 16351).Joseph Myers2015-02-2612-112/+240
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Similar to various other bugs in this area, some asin implementations do not raise the underflow exception for subnormal arguments, when the result is tiny and inexact. This patch forces the exception in a similar way to previous fixes. Tested for x86_64, x86, powerpc and mips64. [BZ #16351] * sysdeps/i386/fpu/e_asin.S (dbl_min): New object. (MO): New macro. (__ieee754_asin): Force underflow exception for results with small absolute value. * sysdeps/i386/fpu/e_asinf.S (flt_min): New object. (MO): New macro. (__ieee754_asinf): Force underflow exception for results with small absolute value. * sysdeps/ieee754/dbl-64/e_asin.c: Include <float.h> and <math.h>. (__ieee754_asin): Force underflow exception for results with small absolute value. * sysdeps/ieee754/flt-32/e_asinf.c: Include <float.h>. (__ieee754_asinf): Force underflow exception for results with small absolute value. * sysdeps/ieee754/ldbl-128/e_asinl.c: Include <float.h>. (__ieee754_asinl): Force underflow exception for results with small absolute value. * sysdeps/ieee754/ldbl-128ibm/e_asinl.c: Include <float.h>. (__ieee754_asinl): Force underflow exception for results with small absolute value. * sysdeps/ieee754/ldbl-96/e_asinl.c: Include <float.h>. (__ieee754_asinl): Force underflow exception for results with small absolute value. * sysdeps/x86_64/fpu/multiarch/e_asin.c [HAVE_FMA4_SUPPORT]: Include <math.h>. * math/auto-libm-test-in: Do not mark underflow exceptions as possibly missing for bug 16351. * math/auto-libm-test-out: Regenerated.
* Fix ldbl-128ibm logbl near powers of 2 (bug 18030).Joseph Myers2015-02-264-4/+27
| | | | | | | | | | | | | | | | | The ldbl-128ibm implementation of logbl produces incorrect results when the high part of the argument is a power of 2 and the low part a nonzero number with the opposite sign (and so the returned exponent should be 1 less than that of the high part). For example, logbl (0x1.ffffffffffffffp1L) returns 2 but should return 1. (This is similar to (fixed) bug 16740 for frexpl, and (fixed) bug 18029 for ilogbl.) This patch adds checks for that case. Tested for powerpc. [BZ #18030] * sysdeps/ieee754/ldbl-128ibm/s_logbl.c (__logbl): Adjust exponent of power of 2 down when low part has opposite sign. * math/libm-test.inc (logb_test_data): Add more tests.
* Fix read past end of pattern in fnmatch (bug 18032)Andreas Schwab2015-02-264-7/+15
|
* Fix ldbl-128ibm ilogbl near powers of 2 (bug 18029).Joseph Myers2015-02-264-5/+33
| | | | | | | | | | | | | | | | | | The ldbl-128ibm implementation of ilogbl produces incorrect results when the high part of the argument is a power of 2 and the low part a nonzero number with the opposite sign (and so the returned exponent should be 1 less than that of the high part). For example, ilogbl (0x1.ffffffffffffffp1L) returns 2 but should return 1. (This is similar to (fixed) bug 16740 for frexpl, and bug 18030 for logbl.) This patch adds checks for that case. Tested for powerpc. [BZ #18029] * sysdeps/ieee754/ldbl-128ibm/e_ilogbl.c (__ieee754_ilogbl): Adjust exponent of power of 2 down when low part has opposite sign. * math/libm-test.inc (ilogb_test_data): Add more tests.
* Fix constness error just introduced in findlocale.Alexandre Oliva2015-02-262-2/+7
| | | | | | | | for ChangeLog [BZ #15969] * locale/findlocale.c (_nl_find_locale): Fix constness error in the previous change.
* BZ #15969: search locale archive again after alias expansionAlexandre Oliva2015-02-263-8/+25
| | | | | | | | | | | | | | If a locale alias is defined in locale.alias but not in an archive, and the referenced locale is only present in the archive, setlocale will fail if given the alias name. This is unintuitive. This patch fixes it, arranging for the locale archive to be searched again after alias expansion. for ChangeLog [BZ #15969] * locale/findlocale.c (_nl_find_locale): Retry archive search after alias expansion.
* Convert tst-iconv3 to use test-skeleton.Roland McGrath2015-02-252-2/+10
|
* Convert tst-iconv5 to use test-skeleton.Roland McGrath2015-02-252-8/+20
|
* Don't crash in iconv setup when getcwd fails.Roland McGrath2015-02-252-1/+6
|
* Fix minor formatting violation.Paul Pluzhnikov2015-02-251-2/+2
|
* powerpc: Fix memmove static buildAdhemerval Zanella2015-02-252-1/+8
| | | | | | | This patch fixes the missing "__memcpy_ppc" symbol for memmove-ppc64 object in static builds. Since memcpy ifunc is not enabled in static mode, the specialized symbols are not provided. The patch changed the it to just "__memcpy" instead.
* Fix ldbl-128ibm asinhl inaccuracy (bug 18020).Joseph Myers2015-02-257-22/+1779
| | | | | | | | | | | | | | | | | | | The ldbl-128ibm implementation of asinhl uses cut-offs of 0x1p28 and 0x1p-29 to determine when to use simpler formulas that avoid possible overflow / underflow. Both those cut-offs are inappropriate for this format, resulting in large errors. This patch changes the code to use more appropriate cut-offs of 0x1p56 and 0x1p-56, adding tests around the cut-offs for various floating-point formats. Tested for powerpc. Also tested for x86_64 and x86 and updated ulps. [BZ #18020] * sysdeps/ieee754/ldbl-128ibm/s_asinhl.c (__asinhl): Use 2**56 and 2**-56 not 2**28 and 2**-29 as thresholds for simpler formulas. * math/auto-libm-test-in: Add more tests of asinh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
* in.h: Coordinate in6_pktinfo and ip6_mtuinfo for kernel and glibc [BZ #15850]Cong Wang2015-02-253-5/+14
| | | | | | | | Similarly to what we did for in6_addr, we need a macro to guard in6_pktinfo and ip6_mtuinfo too. Cc: Carlos O'Donell <carlos@redhat.com> Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
* sprof: Make an error message identical to two others, and more accurate.Benno Schulenberg2015-02-252-1/+5
|
* Reduce lock contention in __tz_convert() [BZ #16145] (partial fix)Kevin Easton2015-02-242-2/+8
| | | | | | | | | | | | This patch is an "easy win" partial fix for BZ #16145, which notes the heavy contention on tzset_lock when multiple threads are converting times with localtime_r(). In __tz_convert(), the lock does not need to be held after __tzfile_compute() / __tz_compute() have been called, so we can move the unlock up. At this point there is still significant work to be done in __offtime(), so we see some improvement (in my testing with 8 cores banging on localtime_r(), ~20% improvement in throughput).
* Update timex.h for ADJ_SETOFFSET.Miroslav Lichvar2015-02-242-2/+9
| | | | | ADJ_SETOFFSET is a new adjtimex() mode that can be used to precisely step the clock. It was introduced in kernel 2.6.39.
* Fix ldbl-128ibm acoshl inaccuracy (bug 18019).Joseph Myers2015-02-257-5/+899
| | | | | | | | | | | | | | | | | | The ldbl-128ibm implementation of acoshl uses a cut-off of 0x1p28 to determine when to use log(x) + log(2) as a formula. That cut-off is too small for this format, resulting in large errors. This patch changes it to a more appropriate cut-off of 0x1p56, adding tests around the cut-offs for various floating-point formats. Tested for powerpc. Also tested for x86_64 and x86 and updated ulps. [BZ #18019] * sysdeps/ieee754/ldbl-128ibm/e_acoshl.c (__ieee754_acoshl): Use 2**56 not 2**28 as threshold for log (2x) formula. * math/auto-libm-test-in: Add more tests of acosh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
* alloca: fix buf interactionMike Frysinger2015-02-242-2/+8
| | | | | | | | | The stack-grows-down case is missing paren around the buf cast. The stack-grows-up case is missing a cast with the buf assignment. This leads to build failures due to -Werror: vfprintf.c: In function '_IO_vfprintf_internal': vfprintf.c:1738:16: error: initialization from incompatible pointer type [-Werror]
generated by cgit-pink 1.4.1 (git 2.36.1) at 2024-11-25 18:45:12 +0000