mirror/eigen
2
0
Fork 0
mirror of https://gitlab.com/libeigen/eigen.git synced 2024-12-21 07:19:46 +08:00
Code Issues Packages Projects Releases Wiki Activity
6,181 Commits 11 Branches 89 Tags 162 MiB
f8fbb3f9a6
Commit Graph

6181 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Benoit Steiner
f8fbb3f9a6 Fixed several compilation warnings reported by clang 2015-02-25 22:22:37 -08:00
Benoit Steiner
8e817b65d0 Silenced a few more compilation warnings generated by nvcc 2015-02-25 17:46:20 -08:00
Benoit Steiner
410070e5ab Added more tests to validate support for tensors laid out in RowMajor order. 2015-02-25 16:14:59 -08:00
Benoit Steiner
1cfd51908c Added support for RowMajor layout to the tensor patch extraction cofde. 2015-02-25 13:29:12 -08:00
Benoit Steiner
eb21a8173e Pulled latest changes from trunk 2015-02-25 09:49:44 -08:00
Benoit Steiner
8afce86e64 Added support for RowMajor layout to the image patch extraction code 
Speeded up the unsupported_cxx11_tensor_image_patch test and reduced its memory footprint
 2015-02-25 09:48:54 -08:00
Benoit Jacob
692136350b So I extensively measured the impact of the offset in this prefetch. I tried offset values from 0 to 128 (on this float* pointer, so implicitly times 4 bytes). 
On x86, I tested a Sandy Bridge with AVX with 12M cache and a Haswell with AVX+FMA with 6M cache on MatrixXf sizes up to 2400.

I could not see any significant impact of this offset.

On Nexus 5, the offset has a slight effect: values around 32 (times sizeof float) are worst. Anything else is the same: the current 64 (8*pk), or... 0.

So let's just go with 0!

Note that we needed a fix anyway for not accounting for the value of RhsProgress. 0 nicely avoids the issue altogether!
 2015-02-25 12:37:14 -05:00
Christoph Hertzberg
531fa9de77 bug #970: Add EIGEN_DEVICE_FUNC to RValue functions, in case Cuda supports RValue-references. 2015-02-24 21:03:28 +01:00
Benoit Jacob
26275b250a Fix my recent prefetch changes: 
- the first prefetch is actually harmful on Haswell with FMA,
   but it is the most beneficial on ARM.
 - the second prefetch... I was very stupid and multiplied by sizeof(scalar)
   and offset of a scalar* pointer. The old offset was 64; pk = 8, so 64=pk*8.
   So this effectively restores the older offset. Actually, there were
   two prefetches here, one with offset 48 and one with offset 64. I could not
   confirm any benefit from this strange 48 offset on either the haswell or
   my ARM device.
 2015-02-23 16:55:17 -05:00
Benoit Jacob
488874781b Add analyze-blocking-sizes program under bench/ to analyze multiple logs 
generated by benchmark-blocking-sizes.
 2015-02-23 14:02:29 -05:00
Christoph Hertzberg
052b6b40f1 Fix two trivial warnings 2015-02-22 12:40:51 +01:00
Christoph Hertzberg
ecbf2a6656 log1p is defined only for real Scalars in C++11 2015-02-21 19:58:24 +01:00
Christoph Hertzberg
6af6cf0c2e I can reproduce any problems that justified this hack. However it makes builds fail in C++11 mode. 2015-02-21 19:43:56 +01:00
Gael Guennebaud
3cf642baa3 Fix compilation of unit tests disabling assertion cheking 2015-02-21 14:13:48 +01:00
Benoit Jacob
458cf91cd9 Add benchmark-blocking-sizes.cpp to bench/ per mailing list discussion. 2015-02-20 17:08:04 -05:00
Gael Guennebaud
03ec601ff7 Initial version of a small script to help tracking performance regressions 2015-02-20 19:20:34 +01:00
Gael Guennebaud
333b497383 update bench_gemm 2015-02-20 11:59:49 +01:00
Gael Guennebaud
2da1594750 Fix doc of Ref<> 2015-02-20 11:52:22 +01:00
Gael Guennebaud
01b8440579 With C++11 Matrix<float> + Matrix<complex<float>> does not even compile 2015-02-20 09:32:49 +01:00
Gael Guennebaud
3594451ee0 Remove EIGEN_TEST_C++0x option and let EIGEN_TEST_CXX11 adds the -std=c++11 flag 2015-02-20 09:31:27 +01:00
Gael Guennebaud
b192e29eae In C++11 destructors do not throw by default (fix CommaInitializer unit test) 2015-02-20 09:28:34 +01:00
Benoit Steiner
ab41652d81 Pulled latest changes from trunk 2015-02-19 21:23:37 -08:00
Benoit Steiner
7765039f1c Marked the CUDA packet primitives as EIGEN_DEVICE_FUNC since they'll end up being executed on the GPU device. 2015-02-19 21:22:51 -08:00
Gael Guennebaud
a66f5fc2fd Fix regression with C++11 support of lambda: now internal::result_of falls back to std::result_of in C++11. 2015-02-19 23:32:12 +01:00
Gael Guennebaud
ece6b440f9 Fix a C++11 compilation issue in unit test 2015-02-19 23:31:08 +01:00
Gael Guennebaud
1b7e12847d Fix some calls to result_of on binary functors as unary ones. 2015-02-19 23:30:41 +01:00
Gael Guennebaud
0f4dd15dfc Declare const some const variables 2015-02-19 23:28:57 +01:00
Benoit Steiner
92ceb02c6d Pulle latest updates from trunk 2015-02-19 11:59:52 -08:00
Benoit Steiner
110fb90250 Improved the documentations 2015-02-19 11:59:04 -08:00
Gael Guennebaud
829dddd0fd Add support for C++11 result_of/lambdas 2015-02-19 15:18:37 +01:00
Benoit Jacob
db05f2d01e rotating kernel: avoid compiling anything outside of ARM 2015-02-18 15:43:52 -05:00
Benoit Jacob
0ed00d5438 remove a newly introduced redundant typedef - sorry. 2015-02-18 15:05:01 -05:00
Benoit Jacob
9bd8a4bab5 bug #955 - Implement a rotating kernel alternative in the 3px4 gebp path 
This is substantially faster on ARM, where it's important to minimize the number of loads.

This is specific to the case where all packet types are of size 4. I made my best attempt to minimize how dirty this is... opinions welcome.

Eventually one could have a generic rotated kernel, but it would take some work to get there. Also, on sandy bridge, in my experience, it's not beneficial (even about 1% slower).
 2015-02-18 15:03:35 -05:00
Hauke Heibel
ee27d50633 Fixed template parameter. 2015-02-18 18:51:08 +01:00
Gael Guennebaud
73a24de424 merge 2015-02-18 15:51:00 +01:00
Gael Guennebaud
63eb0f6fe6 Clean a bit computeProductBlockingSizes (use Index type, remove CEIL macro) 2015-02-18 15:49:05 +01:00
Gael Guennebaud
fc5c3e85e2 Fix bug #961: eigen-doc.tgz included part of itself. 2015-02-18 15:47:01 +01:00
Benoit Jacob
4a3e6c8be1 bug #958 - Allow testing specific blocking sizes 
This is only a debugging/testing patch. It allows testing specific
product blocking sizes, typically to study the impact on performance.

Example usage:

int testk, testm, testn;
#define EIGEN_TEST_SPECIFIC_BLOCKING_SIZES
#define EIGEN_TEST_SPECIFIC_BLOCKING_SIZE_K testk
#define EIGEN_TEST_SPECIFIC_BLOCKING_SIZE_M testm
#define EIGEN_TEST_SPECIFIC_BLOCKING_SIZE_N testn
#include <Eigen/Core>
 2015-02-18 09:43:55 -05:00
Gael Guennebaud
c7bb1e8ea8 Fix a regression when using OpenMP, and fix bug #714: the number of threads might be lower than the number of requested ones 2015-02-18 15:19:23 +01:00
Gael Guennebaud
548b781380 Fix bug #945: workaround MSVC warning 2015-02-18 12:53:49 +01:00
Gael Guennebaud
6f4adc9e94 Add missing install directives for arch/CUDA 2015-02-18 11:40:06 +01:00
Gael Guennebaud
371d3bef36 Workaround dead store warnings in unit tests. 2015-02-18 11:30:44 +01:00
Gael Guennebaud
63464754ef Add an internal assertion in makeCompressed to catch a possible risk of null-pointer access. 2015-02-18 11:29:54 +01:00
Gael Guennebaud
eb563049f7 Remove some dead stores. 2015-02-18 11:26:48 +01:00
Gael Guennebaud
dc7e6acc05 Fix possible usage of a null pointer in CholmodSupport 2015-02-18 11:26:25 +01:00
Gael Guennebaud
d4eda01488 Big 957, workaround MSVC/ICC compilation issue 2015-02-18 11:24:32 +01:00
Christoph Hertzberg
24d65ac0b0 Removed redundant typedef which confused old gcc versions. 2015-02-18 01:03:32 +01:00
Gael Guennebaud
20cac72b82 Packet must be passed by const reference and not by value to avoid alignment issue. 2015-02-17 22:58:32 +01:00
Benoit Steiner
36c9d08274 Pulled latest updates from trunk 2015-02-17 10:04:25 -08:00
Benoit Steiner
f77054f43c Silenced compilation warning 2015-02-17 10:02:04 -08:00