|Summary:||setZero(), setConstant() etc. raise "Illegal Instruction" when called on Eigen::VectorXd|
|Product:||Eigen||Reporter:||Dmitry Zhdanov <dm.zhdanov>|
|Component:||Core - expression templates||Assignee:||Nobody <eigen.nobody>|
|Severity:||Crash||CC:||chtz, gael.guennebaud, jacob.benoit.1|
|Version:||3.3 (current stable)|
|Hardware:||x86 - AVX|
|Bug Depends on:|
Description Dmitry Zhdanov 2015-11-07 05:33:50 UTC
Comment 1 Gael Guennebaud 2015-11-10 10:48:12 UTC
Which patch are you talking about? To go further, you should look at the assembly to see which illegal instruction has been generated. I also guess that it works fine if you disable AVX?
Comment 2 Gael Guennebaud 2015-11-10 10:54:48 UTC
some more questions: - which version of MSVC are you using? - Does it crash if you call .setConstant(2) (i.e., with a value different than 0)? I'm asking because I see that for SSE we already have a workaround for MSVC 9 failing to generate valid assembly for _mm_set1_pd(0): https://bitbucket.org/eigen/eigen/src/fce259fb9c7d163b5966e15b12cf0110a4e54f58/Eigen/src/Core/arch/SSE/PacketMath.h?at=default&fileviewer=file-view-default#PacketMath.h-164
Comment 3 Dmitry Zhdanov 2015-11-10 19:42:51 UTC
(In reply to Gael Guennebaud from comment #2) > some more questions: > - which version of MSVC are you using? > - Does it crash if you call .setConstant(2) (i.e., with a value different > than 0)? > > I'm asking because I see that for SSE we already have a workaround for MSVC > 9 failing to generate valid assembly for _mm_set1_pd(0): > > https://bitbucket.org/eigen/eigen/src/ > fce259fb9c7d163b5966e15b12cf0110a4e54f58/Eigen/src/Core/arch/SSE/PacketMath. > h?at=default&fileviewer=file-view-default#PacketMath.h-164 Gael, here are the responses to your first and second comments. The patch: https://bitbucket.org/eigen/eigen/commits/3bc6df26f930/ (Updated files MathFunctions.h and PacketMath.h to avoid the compiler error with AVX-related type casts of __m256 into __m256d etc.). Note that the compiler is still not entirely happy with this fix and generates warning c4005. It seems that the link you have posted in the 2nd comment points to the same fix. And, of course, there are no errors with AVX disabled. - which version of MSVC are you using? MSVS 2013 express for Windows desktop. - Does it crash if you call .setConstant(2) Yes. A piece of assembly code around the crash point (for .setConstant(2.)): 000007F6B13EDE82 and edx,3 000007F6B13EDE85 add rax,rdx 000007F6B13EDE88 and rax,0FFFFFFFFFFFFFFFCh 000007F6B13EDE8C jle Eigen::internal::call_assignment<Eigen::Matrix<double,-1,1,0,-1,1>,Eigen::CwiseNullaryOp<Eigen::internal::scalar_constant_op<double>,Eigen::Matrix<double,-1,1,0,-1,1> > >+0DFh (07F6B13EDEAFh) 000007F6B13EDE8E xchg ax,ax 000007F6B13EDE90 vbroadcastsd ymm0,xmm0 <<ERROR
Comment 4 Dmitry Zhdanov 2015-11-10 20:59:02 UTC
Just have checked: the PacketMath.h file from your workaround link did not fix the issue. Unrelated question: I have developed the special module for hypersparse matrices (DCSR and DCSC formats) and some AVX accelerators (parallel sparse-sparse and sparse-dense products, in-place transposes etc.) and have several specific questions regarding integrating them with Eigen 3.3 (mostly about changes in SparseMatrix.h file). What would be the right place to ask such a questions?
Comment 5 Christoph Hertzberg 2015-11-11 10:17:02 UTC
It appears your compiler translates _mm256_set1_pd into a AVX2-vbroadcastsd which your CPU apparently does not support. Can your compiler distinguish between generating AVX and AVX2 code? (If so make sure only to activate AVX) Otherwise, a workaround might be to replace the implementation of pset1<...> in src/Core/arch/AVX/PacketMath.h line 114 by _mm256_broadcast_sd(&from); (analogously for float and int, if they cause the same problem) However, this will likely prohibit some compile-optimizations
Comment 6 Gael Guennebaud 2015-11-11 16:01:47 UTC
Regarding the off-topic question, you might create a new bug-entry for each respective features and patches.
Comment 7 Dmitry Zhdanov 2015-11-12 21:11:16 UTC
Thank you, Gael! Changing compiler options (by explicitly enabling only AVX optimizations) solved the problem.
Comment 8 Nobody 2019-12-04 15:09:32 UTC
-- GitLab Migration Automatic Message -- This bug has been migrated to gitlab.com's GitLab instance and has been closed from further activity. You can subscribe and participate further through the new bug through this link to our GitLab instance: https://gitlab.com/libeigen/eigen/issues/1108.