summary refs log tree commit diff
path: root/sysdeps/ieee754
Commit message (Collapse)AuthorAgeFilesLines
* Fix ldbl-96 hypotl of subnormals (bug 15529).Joseph Myers2013-05-241-1/+1
|
* Fix nearbyint scheduling of arithmetic past fesetenv (bug 15490).Joseph Myers2013-05-195-0/+13
|
* Fix remainder exceptions and directed-rounding results (bugs 15480, 15485).Joseph Myers2013-05-171-0/+1
|
* Add #include <stdint.h> for uint[32|64]_t usage (except installed headers).Ryan S. Arnold2013-05-1610-2/+10
|
* Make _LIB_VERSION a weak symbolSiddhesh Poyarekar2013-05-131-4/+5
| | | | | | | That way it can live alongside _LIB_VERSION in libieee.a for statically compiled programs. Resolves #14582.
* Fix ldbl-128ibm cos range reduction near pi/2 (bug 15359).Joseph Myers2013-05-091-5/+5
|
* Fix ldbl-128 cos range reduction near pi/2 (bug 15429).Joseph Myers2013-05-091-5/+5
|
* Improve tgamma accuracy (bugs 2546, 2560, 5159, 15426).Joseph Myers2013-05-0812-32/+1035
|
* Format s_sin.cSiddhesh Poyarekar2013-04-301-904/+1212
|
* [BZ #14686, #15336] Fix standard compliance. Don't use hard-coded qNaN values.Thomas Schwinge2013-04-052-12/+10
|
* [BZ #15335, #15342] Fix standard compliance. Don't use hard-coded qNaN values.Thomas Schwinge2013-04-052-9/+9
|
* Remove unreachable code.Thomas Schwinge2013-04-051-2/+1
| | | | The case of y == 0 is handled at the beginning of the function.
* Remove unused hard-coded qNaN definition.Thomas Schwinge2013-04-051-2/+0
|
* Remove TWOSiddhesh Poyarekar2013-04-032-3/+1
| | | | | Minor cleanup to remove the macro TWO and use the value directly instead.
* Use mantissa_t in mpexpSiddhesh Poyarekar2013-04-022-3/+4
|
* New <math.h> macro named issignaling to check for a signaling NaN (sNaN).Thomas Schwinge2013-04-026-0/+264
| | | | It is based on draft TS 18661 and currently enabled as a GNU extension.
* Remove ONE and MONESiddhesh Poyarekar2013-03-297-22/+20
|
* Format s_tan.cSiddhesh Poyarekar2013-03-291-405/+728
|
* Remove ZERO and MZERO macrosSiddhesh Poyarekar2013-03-296-53/+51
|
* Format s_atan.cSiddhesh Poyarekar2013-03-291-151/+246
|
* Format e_log.cSiddhesh Poyarekar2013-03-291-95/+133
|
* Fix e_logl (128ibm) spurious underflowAdhemerval Zanella2013-03-281-2/+11
|
* Format and clean up s_atan2.cSiddhesh Poyarekar2013-03-281-310/+494
|
* Use integral constantsSiddhesh Poyarekar2013-03-261-76/+76
| | | | | | 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 Poyarekar2013-03-2611-85/+0
|
* Make mantissa type of mp_no configurableSiddhesh Poyarekar2013-03-263-69/+101
| | | | | | | | | | 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.
* BZ#13889: expl (709.75) wrongly overflows for ldbl-128ibmAdhemerval Zanella2013-03-222-6/+24
| | | | | | 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.
* Fix Bessel function spurious overflows for ldbl-128 / ldbl-128ibm (bug 15285).Joseph Myers2013-03-212-20/+34
|
* Replace 8388608.0 with HALFRAD in mp codeSiddhesh Poyarekar2013-03-212-3/+3
| | | | Minor cleanup
* Fix y1l spurious overflows for ldbl-96 (bug 15283).Joseph Myers2013-03-161-1/+1
|
* Fix spurious underflow exceptions for Bessel functions for ldbl-128 / ↵Joseph Myers2013-03-162-58/+79
| | | | ldbl-128ibm (bug 14155).
* Better distinguish between NaN/qNaN/sNaN.Thomas Schwinge2013-03-151-5/+5
|
* Revert configurable mantissa patchSiddhesh Poyarekar2013-03-153-92/+60
| | | | | | Reverts d22ca8cdfb98001d03772ef264b244930d439b3f since it is severely broken on 32-bit.
* Make mantissa type configurableSiddhesh Poyarekar2013-03-153-60/+92
| | | | | This allows the default mantissa to be integral, with powerpc overriding it to take advantage of its FPUs.
* Fix spurious underflow exceptions for Bessel functions for double (bug 14155).Joseph Myers2013-03-142-4/+8
|
* PowerPC: unify math_ldbl.h implementationsAdhemerval Zanella2013-03-081-2/+8
| | | | | This patch removes redudant definition from PowerPC specific math_ldbl, using the definitions from ieee754 math_ldbl.h.
* Merge powerpc slowexp.c into generic codeSiddhesh Poyarekar2013-03-071-3/+9
|
* Merge powerpc slowpow.c into generic codeSiddhesh Poyarekar2013-03-071-0/+17
|
* Use generic mpa.c code for everything except __mul and __sqrSiddhesh Poyarekar2013-03-071-0/+4
|
* BZ #15055: Use __ieee754_sqrl in acoshl for lbdl-128ibmAdhemerval Zanella2013-03-041-1/+1
|
* Format mpsqrt.cSiddhesh Poyarekar2013-02-271-34/+44
|
* Format mpatan2.cSiddhesh Poyarekar2013-02-271-23/+22
|
* Format mpatan.cSiddhesh Poyarekar2013-02-271-46/+61
|
* Format mptan.cSiddhesh Poyarekar2013-02-271-13/+15
|
* Format mplog.cSiddhesh Poyarekar2013-02-271-21/+23
|
* Another tweak to the multiplication algorithmSiddhesh Poyarekar2013-02-261-6/+50
| | | | | Reduce the formula to calculate mantissa so that we reduce the net number of multiplications performed.
* Don't duplicate mpone and mptwoSiddhesh Poyarekar2013-02-261-5/+3
|
* Remove commented declarationsSiddhesh Poyarekar2013-02-251-14/+0
|
* Use long wherever possible in mpa.cSiddhesh Poyarekar2013-02-251-31/+39
| | | | | | Using long throughout like powerpc does is beneficial since it reduces the need to switch to 32-bit instructions. It gives a very minor performance improvement.
* Format slowexp.cSiddhesh Poyarekar2013-02-251-22/+26
|
generated by cgit-pink 1.4.1 (git 2.36.1) at 2024-09-14 16:20:12 +0000