mirror/glm
1
0
Fork 0
mirror of https://github.com/g-truc/glm.git synced 2024-11-15 06:24:34 +00:00
Code Issues Packages Projects Releases Wiki Activity
4,678 Commits 81 Branches 74 Tags 83 MiB
a2aba9e532
Commit Graph

4678 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Robert Chisholm
00a195c687 BugFix: CUDA RTC support. 
Someone noticed that c6dfaed would also catch GLM_COMPILER_GCC8
 2021-05-19 14:15:57 +01:00
Szabolcs Dombi
820a2c0e62 make quaternion constructor respect GLM_FORCE_QUAT_DATA_XYZW 2021-05-17 08:40:22 +03:00
Christophe
70f67cb457
Merge pull request #1072 from g-truc/travis-update 
Use default clang version of distribution
 2021-05-15 12:51:02 +02:00
Christophe
dc9e555b4c
Merge pull request #1071 from sgrottel/gtx-pca 
Implemented 'principle component analysis' utility in gtx #1071
 2021-05-15 12:50:27 +02:00
Christophe
c6dfaed4a3
Merge pull request #1073 from Robadob/nvrtc_support 
Correctly detect NVRTC and lack of stl support #1073
 2021-05-15 12:31:00 +02:00
Robert Chisholm
948ac3b825 Add missing statement. 
Didn't copy this across, good thing CI caught it.
 2021-05-14 10:38:02 +01:00
Robert Chisholm
aa7380ae12 Fix lint. 2021-05-13 22:33:49 +01:00
Robert Chisholm
f0f95bd3de Correctly detect NVRTC and lack of stl support. 2021-05-13 22:13:04 +01:00
SGrottel
d71dba9603 Introduced a second, less precise comparison epsilon for the tests for now. 2021-05-11 07:44:40 +02:00
SGrottel
18d9b97aa4 Further increased comparison float epsilon, and further test batch testEigenvectors also failes. 
Added debug output to `testEigenvectors` in case the error persists.
 2021-05-10 23:36:17 +02:00
SGrottel
593b7cc36b Increased float comparison epsilon to pass tests. 2021-05-10 22:49:57 +02:00
SGrottel
a0ccbcc63d Added further details on the comparison issue with covariance matrices on some VMs. 
Also corrected some code style guide, and changed `nullptr` to `GLM_NULLPTR` for better compatibility.
Tests are now executed in blocks of related tests, and only inbetween blocks the tests will exit.
 2021-05-10 21:32:01 +02:00
SGrottel
d0d7945141 Additional debug output to investigate why test-gtx_pca fails on some VMs on Travis. 
Also, reworked the `#if` about CXX11; did not seem to work correctly.
 2021-05-10 19:35:57 +02:00
SGrottel
c792a0a221 Disabled tests requiring random engine when CXX11 STL is not available. 
Added missed `typename` keywords, and fixed variable initialization.
 2021-05-10 17:48:35 +02:00
SGrottel
0f5b544d5d Corrected errors on Xcode C++98 pure related to language extensions accidentially used. 2021-05-10 16:38:38 +02:00
SGrottel
b8adc27808 Removed lambdas and initializer list ctors to be compatible with older cpp standards. 2021-05-10 15:45:42 +02:00
Christophe
ddb6678c23 Use default clang version of distribution 2021-05-10 14:11:17 +02:00
Christophe
4abd46483a Fix Travis C.I. build 2021-05-10 13:34:39 +02:00
SGrottel
dd40903b74 Implemented 'principle component analysis' utility in gtx, including tests 2021-05-10 13:14:29 +02:00
Christophe
6e2b7eedd2
Merge pull request #1069 from Blurgyy/quaternion-sane-default 
Fix constructor of struct qua to honor macro "GLM_FORCE_QUAT_DATA_WXYZ" #1069
 2021-05-07 09:23:48 +02:00
Gaoyang Zhang
de7c83f1b6 Update test for GLM_FORCE_QUAT_DATA_XYZW 
Signed-off-by: Gaoyang Zhang <gy@blurgy.xyz>
 2021-05-06 14:12:41 +08:00
Gaoyang Zhang
59ddeb7d1c Replace GLM_FORCE_QUAT_DATA_WXYZ with GLM_FORCE_QUAT_DATA_XYZW 
The default data layout for quat has been changed to w,x,y,z to agree
with order of input parameters for quat's constructor.

Signed-off-by: Gaoyang Zhang <gy@blurgy.xyz>
 2021-05-06 14:12:41 +08:00
Gaoyang Zhang
d573bf099a Revert "Fix constructor of struct qua to honor macro "GLM_FORCE_QUAT_DATA_WXYZ"" 
This reverts commit f931c4996a.
 2021-05-06 14:12:23 +08:00
Gaoyang Zhang
f6f3596a6b Revert "Fix construction of identity quaternion" 
This reverts commit 37842c7400.
 2021-05-06 14:12:22 +08:00
Gaoyang Zhang
43b81f49fc Revert "Fix test due to change of default constructor of qua" 
This reverts commit e800c41c0a.
 2021-05-06 14:12:16 +08:00
Gaoyang Zhang
e800c41c0a Fix test due to change of default constructor of qua 
Signed-off-by: Gaoyang Zhang <gy@blurgy.xyz>
 2021-05-03 00:32:17 +08:00
Gaoyang Zhang
37842c7400 Fix construction of identity quaternion 
Signed-off-by: Gaoyang Zhang <gy@blurgy.xyz>
 2021-05-03 00:31:09 +08:00
Gaoyang Zhang
f931c4996a Fix constructor of struct qua to honor macro "GLM_FORCE_QUAT_DATA_WXYZ" 
Signed-off-by: Gaoyang Zhang <gy@blurgy.xyz>
 2021-05-02 23:09:59 +08:00
Christophe
761a842a59
Merge pull request #1062 from tksuoran/fix-cmake 
Fix CMake FetchContent usage #1062
 2021-04-05 13:27:48 +02:00
Christophe
820607ad2b
Merge pull request #1064 from SergeyKrivohatskiy/fix-axisAngle-for-small-angles 
Fix axis angle implementation for near zero angles and near PI angles #1064
 2021-04-05 13:25:56 +02:00
Christophe
50fdc42f13
Merge pull request #1065 from psychocoderHPC/fix-cudaClangCompile 
fix clang CUDA compile #1065
 2021-04-05 13:23:00 +02:00
René Widera
59fbd31552 fix clang CUDA compile 
Fix CUDA compile issue when using clang.

```
glm/gtc/type_ptr.inl:257:22: error: __host__ function 'make_vec4' cannot overload __host__ __device__ function 'make_vec4'
6946        inline vec<4, T, Q> make_vec4(vec<4, T, Q> const& v)
```
 2021-03-31 14:13:53 +02:00
Sergey Krivohatskiy
b5d4757580 compilation fix 
Signed-off-by: Sergey Krivohatskiy <s.krivohatskiy@gmail.com>
 2021-03-23 15:22:16 +03:00
Sergey Krivohatskiy
e81a9c4baa compilation fixes 
Signed-off-by: Sergey Krivohatskiy <s.krivohatskiy@gmail.com>
 2021-03-23 15:03:38 +03:00
Sergey Krivohatskiy
2010c883d5 formatting fix 
Signed-off-by: Sergey Krivohatskiy <s.krivohatskiy@gmail.com>
 2021-03-23 14:33:11 +03:00
Sergey Krivohatskiy
2b766d5303 a few more tests for axisAngle 
Signed-off-by: Sergey Krivohatskiy <s.krivohatskiy@gmail.com>
 2021-03-23 14:30:52 +03:00
Sergey Krivohatskiy
df7b5bda7c axisAngle epsilon changed to work as expected on small angles 
also small refactoring
also single epsilon is used
also passes new axisAngle tests now

Signed-off-by: Sergey Krivohatskiy <s.krivohatskiy@gmail.com>
 2021-03-23 14:28:36 +03:00
Sergey Krivohatskiy
c9c656ce31 axisAngle refactoring 
Signed-off-by: Sergey Krivohatskiy <s.krivohatskiy@gmail.com>
 2021-03-23 14:27:00 +03:00
Sergey Krivohatskiy
57c3be0b9f Added axisAngle tests 
Signed-off-by: Sergey Krivohatskiy <s.krivohatskiy@gmail.com>
 2021-03-23 14:24:41 +03:00
Timo Suoranta
1771b1127f Fix CMake FetchContent usage 2021-03-11 13:20:56 +02:00
Christophe
ace16e4778
Merge pull request #1060 from imgly/fix/simd-vec4-compare 
Fixed the SIMD (SSE) implementation of compute_vec4_equal for floats #1060
 2021-03-08 11:01:35 +01:00
Christophe
9203edd741
Merge pull request #1054 from kiroma/master 
Create an install and package targets in cmake tree #1054
 2021-03-08 10:55:19 +01:00
kiroma
ca1f397fe9
Merge pull request #1 from jaimedelreyg/feature/add_uninstall_to_cmake 
Add target make uninstall in cmake files
 2021-03-07 17:23:49 +01:00
Christophe
fe9bb21f1f
Merge pull request #1058 from SergeyKrivohatskiy/patch-1 
Fix in axisAngle implementation #1058
 2021-03-07 09:37:42 +01:00
Christophe
5e243a8efa
Merge pull request #1061 from g-truc/revert-1012-master 
Revert "Fix quaternion orientation in `glm::decompose`" #1061 #1012
 2021-03-07 09:36:25 +01:00
Christophe
bbedf96c16
Revert "Fix quaternion orientation in glm::decompose" 2021-03-06 17:12:03 +01:00
Jaime Del Rey Garcia
87de7f3f10 Add target make uninstall in cmake files 2021-03-06 08:58:58 +01:00
Christophe
64be09e1b1
Merge pull request #1057 from nick-dumas/master 
Fix squered typo #1057
 2021-03-02 22:49:53 +01:00
Keiwan Donyagard
a6fcdd0cbc Fixed the SIMD implementation of compute_vec4_equal for floats 2021-02-24 10:51:20 +01:00
Sergey Krivohatskiy
acab24129d
Fixed axisAngle implementation 
`acos` domain is in range [-1.0, 1.0]. Due to inaccuracies the value `angleCos` may be slightly outside that range for a correct matrix and `acos(angleCos)` produces `NaN` in that case.

The fix is we check `angleCos` value and return `acos(1)` for `angleCos > 1` and `acos(-1)` for `angleCos < -1`.

The original code checked only for `angleCos` close to `1.0` and returned an incorrect value for `acos(1)`, which is `0`, not  `pi/4`.
 2021-02-20 20:16:31 +03:00