Commit Graph

  • 061fea48c8 Add an ARM NEON readu simd4x4_neon.hxx and small fixes for simd_neon.hxx Erik Hofman 2017-01-19 13:20:52 +01:00
  • c4ea62a899 Add a version of simd.hxx for ARM NEON Erik Hofman 2017-01-18 15:41:12 +01:00
  • 789c09a402 I forgot the linux variant of ALIGN Erik Hofman 2017-01-12 14:19:09 +01:00
  • de9b329115 reorganize the code a bit Erik Hofman 2017-01-12 11:43:46 +01:00
  • 1b793a127c Be more explicit about alignment Erik Hofman 2017-01-12 10:18:39 +01:00
  • 3cb3084725 Rename ‘new’ thumbnails to ‘previews’ James Turner 2017-01-10 18:05:40 +00:00
  • 637f67888a Reinstate AVX support again Erik Hofman 2017-01-07 13:55:09 +01:00
  • fbc0986fd8 Do not rely on SSE3, this is just a slight drawback Erik Hofman 2017-01-06 15:10:52 +01:00
  • 14ebe0b618 Disable AVX support until someone figures out why it is not working properly Erik Hofman 2017-01-06 00:11:05 +01:00
  • fd34cc30b8 compare_versions: limit how many parts are checked. James Turner 2017-01-05 10:52:48 +00:00
  • 7b0faed03a Fix an AVX error Erik Hofman 2017-01-04 16:30:19 +01:00
  • 8d1dc30b07 Try to fix a possible AVX core dump Erik Hofman 2017-01-04 12:55:28 +01:00
  • 03cff6abca Fix Never use <avxintrin.h> directly; include <immintrin.h> instead. Erik Hofman 2017-01-01 13:26:36 +01:00
  • d7821324b8 Fix a wrong enclosing bracket Erik Hofman 2017-01-01 11:40:59 +01:00
  • f3e066cce0 Bug fix for bug 1920 Ron Haertel 2016-12-24 14:00:37 -05:00
  • 235e38b69b new version: 2016.4.4 version/2016.4.4 Automatic Release Builder 2016-12-28 13:57:44 +01:00
  • 45ac758cc9 Expose position on SGPropertyNode. James Turner 2016-12-27 11:09:07 +00:00
  • f0e6402fff One more, and hopefully the last, MSVC fix Erik Hofman 2016-12-23 15:31:40 +01:00
  • 41bf142e31 Try to fix an MSVC compile error Erik Hofman 2016-12-23 14:50:14 +01:00
  • 279b2c120f Bugfix: reject dubious paths in HTTP repos. James Turner 2016-12-14 09:41:44 +00:00
  • 2e9efa98d7 Add accelearated cross product Erik Hofman 2016-12-22 13:43:04 +01:00
  • 2c2a57f368 Addd support for AVX, 4 double precision operations in one go Erik Hofman 2016-12-21 09:04:46 +01:00
  • 4ea1326126 A better MSVC fix and code speedups Erik Hofman 2016-12-20 09:54:19 +01:00
  • 09b44ac68a MSVC compiler fix (hopefully) Erik Hofman 2016-12-19 18:37:52 +01:00
  • a48ab434ab Convert more code to SIMD Erik Hofman 2016-12-19 15:58:57 +01:00
  • d39a56d4fb Add missing include: <cctype> for unary std::islower() Florent Rougon 2016-12-18 11:02:01 +01:00
  • 892579456d add SSE support for the C compiler (gcc and clang) Erik Hofman 2016-12-18 09:40:38 +01:00
  • 4dde1d365c MSVC 2013+ for 64-bit automatically sets /arch:SSE2 and doesn't recognise te option Erik Hofman 2016-12-17 10:50:19 +01:00
  • 2f21b582cd Canvas image fill and source rect setters. James Turner 2016-12-16 09:24:40 +00:00
  • b29536f8b7 Set Canvas element transforms from C++ James Turner 2016-12-16 09:23:48 +00:00
  • 039f9920db Direct rect support. James Turner 2016-12-15 17:35:46 +00:00
  • 7c254e9c04 Native SVG path-data parsing in the Canvas. James Turner 2016-12-12 19:53:48 +01:00
  • 35a115bfd4 Fix a stupid mistake and move hsum_ps_sse to shVector.c Erik Hofman 2016-12-16 16:01:59 +01:00
  • 70dd9d35b1 Also accelerate SHVector3 using SIMD Erik Hofman 2016-12-16 15:12:53 +01:00
  • 31e3cf06fb Accelerate SHVector4 and SHMatrix3x3 operations using SIMD Erik Hofman 2016-12-16 14:54:24 +01:00
  • 175eddd1fa Add the proper SSE options to Clang too Erik Hofman 2016-12-16 11:05:03 +01:00
  • fe73247b82 Really use _mm_setzero_ps9) instead Erik Hofman 2016-12-16 11:04:49 +01:00
  • 203db3d095 Enable SSE and SSE2 by default Erik Hofman 2016-12-16 10:49:44 +01:00
  • 8c4695b991 Fix duplicate variants in packages. James Turner 2016-12-15 17:35:36 +00:00
  • a2b111bb09 Bugfix: reject dubious paths in HTTP repos. James Turner 2016-12-14 09:41:44 +00:00
  • 2a1542d544 Helper for parsing CSS/SVG strings. James Turner 2016-12-12 13:45:09 +00:00
  • e5b51677c5 MacOS fix Erik Hofman 2016-12-15 13:11:31 +01:00
  • f9450d136d Minor speedups in small tweaks Erik Hofman 2016-12-15 09:57:10 +01:00
  • 0586cb62c3 Fix the pink-sky bug, don't initilize the vector with (z,y,z,w) but use the proper (x,y,z,w) instead Erik Hofman 2016-12-15 09:56:52 +01:00
  • 8fee04b32b Generalize the code to be able to test integers too Erik Hofman 2016-12-14 11:05:56 +01:00
  • 04e16c95e2 Consolidate some code, exlicitly use intrinsics (instead of the usual math operators): this fixes a compiler bug for integers and add some SSE4.1 accelerations for integers Erik Hofman 2016-12-14 11:05:31 +01:00
  • e257dbe6ed accelerate float rotation_matrix and fix a bug for double rotation_matrix Erik Hofman 2016-12-14 11:03:32 +01:00
  • abf78f8e31 Make absolutely sure the unions are 16-bytes aligned, move more setters and min() and max() to the simd implementation and add horizontal SIMD add functions for magnitude2 and dot Erik Hofman 2016-12-12 20:27:07 +01:00
  • da13bd9f04 Fix minor compiler warnings Stuart Buchanan 2016-12-10 16:00:28 +00:00
  • 86fb1ed00f BoostTestTargets.cmake: fix a wrong include test Alessandro Menti 2016-12-08 11:47:18 +01:00
  • 56fb81dc03 Hopefuly the last MSVC fix Erik Hofman 2016-12-08 09:39:45 +01:00
  • 343ce57468 Tupe conversion is evil, don't allow it Erik Hofman 2016-12-08 01:02:44 +01:00
  • 3e52e37181 remove explicit, it looks like SMVC chokes on it Erik Hofman 2016-12-07 13:22:17 +01:00
  • e768553a4a Replace auto_ptr with unique_ptr Erik Hofman 2016-12-07 11:03:49 +01:00
  • b74d1a8351 Add some more accelerated functions (dot, magnitude, matrix transpose), allow simd_t4 with non mathcing elements to be assigned and a bunch of fixes. Make sure the code compiles when SSE is not available. Erik Hofman 2016-12-07 10:00:02 +01:00
  • 801d8c4af5 Revert recursive listeners for the moment. James Turner 2016-12-05 12:21:04 +00:00
  • c92a953511 new version: 2016.4.3 version/2016.4.3 Automatic Release Builder 2016-12-05 13:28:25 +01:00
  • 3a4693803b strutils: new functions stripTrailingNewlines() and stripTrailingNewlines_inplace() Florent Rougon 2016-12-04 00:49:59 +01:00
  • 79f0d3356e Convert strutils_test.cxx to use the SG test macros instead of BOOST Florent Rougon 2016-12-04 00:25:38 +01:00
  • 6a1bf02ddb Use appropriate test macros in parse_color_test.cxx and state_machine_test.cxx Florent Rougon 2016-12-04 20:15:56 +01:00
  • 6662800deb Expand the test macros framework Florent Rougon 2016-12-04 20:05:54 +01:00
  • c9bb6102c0 Test macros: add parentheses around arguments Florent Rougon 2016-12-04 18:32:20 +01:00
  • 90479419cc Add test macros: SG_CHECK_NE, SG_CHECK_LT, SG_CHECK_LE, SG_CHECK_GT, SG_CHECK_GE Florent Rougon 2016-12-04 17:07:36 +01:00
  • b4178ae888 Test macros: be more rigorous and consistent Florent Rougon 2016-12-04 16:13:50 +01:00
  • ab4814c916 simgear/misc/test_macros.hxx: add missing <iostream> header + minor changes Florent Rougon 2016-12-04 15:45:10 +01:00
  • 6b16f96c8a Test macros: use defensive parentheses and std::fabs() for SG_CHECK_EQUAL_EP* Florent Rougon 2016-12-04 15:20:46 +01:00
  • e655d41817 Rename the COMPARE, COMPARE_EP, COMPARE_EP2 and VERIFY test macros Florent Rougon 2016-12-04 14:38:47 +01:00
  • b5c1902a2d Tweaks to recursive listener behaviour James Turner 2016-12-03 14:09:56 +00:00
  • d088259739 Add aliased listener test, disabled for now. James Turner 2016-12-02 17:15:11 +00:00
  • d8acf44a3a Revert unintended change in listener refactoring. James Turner 2016-12-02 17:14:27 +00:00
  • 4664af12fa Fix lockup on Windows when polling DNS, add test Torsten Dreyer 2016-12-02 14:43:52 +01:00
  • 7a909d0c0b Improved Package search function. James Turner 2016-11-30 21:32:31 +00:00
  • 919c25769c Quiet a canvas message. James Turner 2016-11-30 12:18:44 +00:00
  • 863ae19d1d Package::indexOfvariant works on fully-qualified IDs. James Turner 2016-11-29 15:36:00 +00:00
  • 5a0908d5bb Package::indexOfvariant works on fully-qualified IDs. James Turner 2016-11-29 15:36:00 +00:00
  • 1c39daec07 Fix an error Erik Hofman 2016-11-27 10:19:17 +01:00
  • 835ae941ce Unit-tests for localised-variant strings. James Turner 2016-11-26 14:28:22 +00:00
  • 63edff078f Fix warnings in sgvec4 test. James Turner 2016-11-26 14:20:17 +00:00
  • 0ea9786601 Add SIMD matrix operations Erik Hofman 2016-11-26 11:40:01 +01:00
  • 9e0bb33d58 Fix missing include in simgear/props/props_test.cxx Florent Rougon 2016-11-26 01:12:34 +01:00
  • 64531c85e3 Generalise per-variant localised string lookup. James Turner 2016-11-25 22:44:24 +00:00
  • 14cdae5102 Unit-test for copy-installed children. James Turner 2016-11-19 14:51:49 +00:00
  • 2aaad212e8 Revert partial-update mode for TerraSync repos. James Turner 2016-11-19 13:50:53 +00:00
  • bd88bf1126 Explicit handling of recursive listeners. James Turner 2016-11-25 15:40:13 +00:00
  • ea9da65b7c Expanded property unit-tests. James Turner 2016-11-20 12:31:03 +00:00
  • e7f80cf5f3 Finish SIMD Vector code, do not change matrix operations yet since it seems to have a problem Erik Hofman 2016-11-25 13:31:54 +01:00
  • 16d62f93c8 Fix a clang reported error Erik Hofman 2016-11-22 15:18:16 +01:00
  • e96834fcc6 Fix an error Erik Hofman 2016-11-22 14:36:28 +01:00
  • 22a74c63b4 First changes for SIMD matrix operations Erik Hofman 2016-11-22 14:27:42 +01:00
  • cd7b6d69b0 new version: 2016.4.2 Automatic Release Builder 2016-11-22 09:38:58 +01:00
  • 5681fcbdc5 Unit-tests for property listener. James Turner 2016-11-20 14:32:23 +00:00
  • 7755f8e094 Override changes for atomic listener. James Turner 2016-11-20 12:30:44 +00:00
  • e266e44f63 Extended scenery/STG suffix handling James Turner 2016-11-19 13:20:03 +00:00
  • 4dc66a385e Include SGMathFwd.hxx to add unknown math types Erik Hofman 2016-11-21 09:22:09 +01:00
  • aec29a3a37 Fix clang errors Erik Hofman 2016-11-20 23:54:15 +01:00
  • 0d213a1990 Fix a typo Erik Hofman 2016-11-20 16:16:28 +01:00
  • 8162a49f6c Give vector math a (potential) boost. Next stoP; matrices Erik Hofman 2016-11-20 16:15:40 +01:00
  • d2e2603400 initialize udns_library on first use Torsten Dreyer 2016-11-20 16:11:24 +01:00
  • 43a8277bdb Merge branch 'next' of ssh://git.code.sf.net/p/flightgear/simgear into next Erik Hofman 2016-11-20 15:01:10 +01:00
  • 61a8bd5cd3 _m128d ony holds two doubles instead of 4 (obviously). Erik Hofman 2016-11-20 15:00:24 +01:00