about summary refs log tree commit diff
Commit message (Collapse)AuthorAgeFilesLines
* Fix missing underflow from cexp (bug 14478).Joseph Myers2013-04-036-4/+52
|
* Update NEWSAndreas Schwab2013-04-031-0/+3
|
* Fix stack overflow in getaddrinfo with many resultsAndreas Schwab2013-04-033-3/+31
|
* Refer to two GCC PRs.Thomas Schwinge2013-04-033-2/+11
|
* Update bench.out and bench.out.old only upon completionSiddhesh Poyarekar2013-04-032-5/+8
| | | | | | | Write output from the currently running benchmark into a temporary file and move files around only once the current run is complete. That way we don't lose data from the last two runs due to an incomplete run.
* Echo benchmark that is currently runningSiddhesh Poyarekar2013-04-032-0/+3
| | | | This will be useful when the benchmark has many functions.
* Add benchmark inputs for atanSiddhesh Poyarekar2013-04-035-1/+44
| | | | Add separate inputs for slow and fast paths of atan
* Remove TWOSiddhesh Poyarekar2013-04-033-3/+5
| | | | | Minor cleanup to remove the macro TWO and use the value directly instead.
* Preserve errno across _PC_CHOWN_RESTRICTED call on XFSSiddhesh Poyarekar2013-04-034-1/+25
| | | | | | | | | | | | | | Fix BZ #15305. On kernel versions earlier than 2.6.29, the Linux kernel exported a sysctl called restrict_chown for xfs, which could be used to allow chown to users other than the owner. 2.6.29 removed this support, causing the open_not_cancel_2 to fail and thus modify errno. The fix is to save and restore errno so that the caller sees it as unmodified. Additionally, since the code to check the sysctl is not useful on newer kernels, we add an ifdef so that in future the code block gets rmeoved completely.
* Fix cacosh inaccuracy and spurious exceptions (bug 15327).Joseph Myers2013-04-028-117/+2287
|
* Use mantissa_t in mpexpSiddhesh Poyarekar2013-04-023-3/+7
|
* Add benchmark inputs for sinSiddhesh Poyarekar2013-04-025-1/+52
|
* Add benchmark tests for slowpow and slowexpSiddhesh Poyarekar2013-04-028-6/+77
| | | | | | | Separate benchmarks for the fast and slow implementations of pow and exp since measuring both together doesn't make sense. Adjust the iterations for pow and exp accordingly so that they run long enough for the measurements to be meaningful.
* Add missing ChangeLog from commit 60c414c346a1d5ef0510ffbdc0ab75f288ee4d3fAdhemerval Zanella2013-04-021-0/+10
|
* New <math.h> macro named issignaling to check for a signaling NaN (sNaN).Thomas Schwinge2013-04-0249-16/+692
| | | | It is based on draft TS 18661 and currently enabled as a GNU extension.
* PowerPC: remove branch prediction from rint implementationAdhemerval Zanella2013-04-016-13/+22
| | | | | | | | The branch prediction hints is actually hurts performance in this case. The assembly implementation make two assumptions: 1. 'fabs (x) < 2^52' is unlikely and 2. 'x > 0.0' is unlike (if 1. is true). Since it a general floating point function, expected input is not bounded and then it is better to let the hardware handle the branches.
* Update German translations.David S. Miller2013-03-302-3/+7
| | | | * po/de.po: Update from translation team.
* Fix casinh inaccuracy for imaginary part < 1.0, real part small (bug 10357).Joseph Myers2013-03-308-301/+3417
|
* Remove usage of ONE macroSiddhesh Poyarekar2013-03-292-1/+4
| | | | I missed this instance, which caused a build failure in powerpc.
* Remove ONE and MONESiddhesh Poyarekar2013-03-298-22/+29
|
* Format s_tan.cSiddhesh Poyarekar2013-03-292-405/+730
|
* Remove ZERO and MZERO macrosSiddhesh Poyarekar2013-03-298-63/+71
|
* Format s_atan.cSiddhesh Poyarekar2013-03-292-151/+248
|
* Format e_log.cSiddhesh Poyarekar2013-03-292-95/+137
|
* Avoid unconditional __call_tls_dtors calls in static linking.Roland McGrath2013-03-286-5/+27
|
* Use __ehdr_start, if available, as fallback for AT_PHDR.Roland McGrath2013-03-283-8/+38
|
* Make _dl_phdr pointer to const.Roland McGrath2013-03-285-5/+11
|
* Consolidate declarations of _dl_phdr, _dl_phnum.Roland McGrath2013-03-285-12/+12
|
* Fix typo in bug number in NEWS.Joseph Myers2013-03-281-2/+2
|
* Never call syslog in __libc_message.Roland McGrath2013-03-282-9/+5
|
* PowerPC: .eh_frame info in crt1.o isn't useful and triggers gold bug 14675.Alan Modra2013-03-283-0/+21
| | | | | | | The .eh_frame info in crt1.o isn't useful and this patch prevents it from being generated on PowerPC. It triggers the following gold bug: http://sourceware.org/bugzilla/show_bug.cgi?id=14675
* Fix e_logl (128ibm) spurious underflowAdhemerval Zanella2013-03-284-4/+26
|
* Don't add input group during initgroups_dyn in hesiodSiddhesh Poyarekar2013-03-283-28/+5
| | | | | | | | | | | | | Fixes BZ #15304. The *initgroups_dyn functions are called with a group argument. This group gid is usually skipped while populating the grouplist since the caller adds that group id in advance. The hesiod initgroups_dyn implementation however adds the group gid to the list if it does not already exist. While it works fine for the usual initgroups, it breaks nscd since it calls initgroups_dyn with -1 as the gid (to have all groups included).
* Format and clean up s_atan2.cSiddhesh Poyarekar2013-03-282-310/+498
|
* Fix casinh inaccuracy near i, imaginary part > 1 (bug 15307).Joseph Myers2013-03-278-1/+1214
|
* ARM: fix preconfigure.Mans Rullgard2013-03-263-2/+7
|
* Use integral constantsSiddhesh Poyarekar2013-03-262-76/+90
| | | | | | The compiler is smart enough to convert those into double for powerpc, but if we put them as doubles, it adds overhead by performing those operations in floating point mode.
* Removed commented codeSiddhesh Poyarekar2013-03-2612-85/+21
|
* Make mantissa type of mp_no configurableSiddhesh Poyarekar2013-03-265-69/+176
| | | | | | | | | | The mantissa of mp_no is intended to take only integral values. This is a relatively good choice for powerpc due to its 4 fpus, but not for other architectures, which suffer due to this choice. This change makes the default mantissa a long integer and allows powerpc to override it. Additionally, some operations have been optimized for integer manipulation, resulting in a significant improvement in performance.
* PowerPC: fix libm ABI issue for llroundlAdhemerval Zanella2013-03-262-0/+9
|
* Fix docs for scalbn* and scalbl* functionsMark H Weaver2013-03-242-6/+11
| | | | | * manual/arith.texi (Normalization Functions): Fix prototypes for scalbn, scalbnf, scalbnl, scalbln, scalblnf, and scalblnl.
* Update sparc ULPs.David S. Miller2013-03-221-4/+1002
| | | | * sysdeps/sparc/fpu/libm-test-ulps: Update.
* BZ#13889: expl (709.75) wrongly overflows for ldbl-128ibmAdhemerval Zanella2013-03-225-10/+40
| | | | | | The patch increase the high value to check if expl overflows. Current high mark value is not really correct, the algorithm accepts high values. It also adds a correct wrapper function to check for overflow and underflow.
* BZ#11120: fix x86_64/strcmp.S NOT_IN_libc safeguardsDmitry V. Levin2013-03-223-11/+17
| | | | | | | | Due to a typo repeated several times, this bug hasn't been fixed yet, despite being marked as resolved in glibc 2.12. * sysdeps/x86_64/strcmp.S: Replace all occurrences of NOT_IN_lib with NOT_IN_libc.
* PowerPC: fix sqrtl ABI issueAdhemerval Zanella2013-03-212-0/+9
| | | | This patch fixes a sqrtl ABI issue when building for powerpc64.
* libm-test: Properly wrap blocks consisting of several statements.Thomas Schwinge2013-03-212-8/+17
|
* On 32-bit x86, disable certain tests involving sNaN values.Thomas Schwinge2013-03-214-10/+91
| | | | Follow-up to commit 495ded2c8c1eb8c0ac4b54add2dd397852e19cba.
* Fix Bessel function spurious overflows for ldbl-128 / ldbl-128ibm (bug 15285).Joseph Myers2013-03-216-21/+62
|
* Allow adding of arbitrary code to benchmark testsSiddhesh Poyarekar2013-03-212-1/+10
| | | | | | | This allows us to define custom functions in C code files and benchmark scenarios rather than just functions. The main current use of this is to separate the slow and fast path benchmarks for math functions.
* Fix casinh inaccuracy for argument with imaginary part 1 (bug 15287).Joseph Myers2013-03-218-3/+1812
|
generated by cgit-pink 1.4.1 (git 2.36.1) at 2024-12-14 06:29:12 +0000