https://gcc.gnu.org/bugzilla/show_bug.cgi?id=97369

--- Comment #3 from Seager Du <xianping.du at gmail dot com> ---
(In reply to Jonathan Wakely from comment #1)
> You haven't provided sufficient information to tell, but I doubt there's a
> bug here.
> 
> Please read these links and confirm if you're using the right libstdc++.so
> at run time (the one from your self-compile GCC 6.3.0, not the system
> compiler):
> https://gcc.gnu.org/onlinedocs/libstdc++/faq.html#faq.how_to_set_paths
> https://gcc.gnu.org/onlinedocs/libstdc++/manual/using_dynamic_or_shared.
> html#manual.intro.using.linkage.dynamic
> 
> I've said it before, but the fact that matlab 2020 only works with a GCC
> version that has been unsupported for two years is stupid.

You are right. I checked the manually compiled GCC6.3 and found the lib64 was
not exported to the path. This fixed the compiling problem with the S-function.
Thank you!

Now, a new error appeared as using Simulink to call S-function:
"Stack Trace (from fault):
[ 0] 0x00007fc361b66d11
/cache/sw/packages/MATLAB/R2020a/bin/glnxa64/mkl.so+23264529
mkl_blas_avx2_xidamax_nonan+00000417
[ 1] 0x00007fc361a33a0c
/cache/sw/packages/MATLAB/R2020a/bin/glnxa64/mkl.so+22006284
mkl_blas_avx2_xidamax+00000060
[ 2] 0x00007fc360bff186
/cache/sw/packages/MATLAB/R2020a/bin/glnxa64/mkl.so+07111046
mkl_blas_xidamax+00000086
[ 3] 0x00007fc360b831ad
/cache/sw/packages/MATLAB/R2020a/bin/glnxa64/mkl.so+06603181
mkl_blas_idamax+00000861
[ 4] 0x00007fc36132acdf
/cache/sw/packages/MATLAB/R2020a/bin/glnxa64/mkl.so+14630111
mkl_lapack_dgetf2_team+00000975
[ 5] 0x00007fc36132a84c
/cache/sw/packages/MATLAB/R2020a/bin/glnxa64/mkl.so+14628940
mkl_lapack_dgetrf_local_team+00002812
[ 6] 0x00007fc36132a00b
/cache/sw/packages/MATLAB/R2020a/bin/glnxa64/mkl.so+14626827
mkl_lapack_dgetrf_local_team+00000699
[ 7] 0x00007fc36132a00b
/cache/sw/packages/MATLAB/R2020a/bin/glnxa64/mkl.so+14626827
mkl_lapack_dgetrf_local_team+00000699
[ 8] 0x00007fc36132a00b
/cache/sw/packages/MATLAB/R2020a/bin/glnxa64/mkl.so+14626827
mkl_lapack_dgetrf_local_team+00000699
[ 9] 0x00007fc36132a00b
/cache/sw/packages/MATLAB/R2020a/bin/glnxa64/mkl.so+14626827
mkl_lapack_dgetrf_local_team+00000699
[ 10] 0x00007fc36132a00b
/cache/sw/packages/MATLAB/R2020a/bin/glnxa64/mkl.so+14626827
mkl_lapack_dgetrf_local_team+00000699
[ 11] 0x00007fc360a83888
/cache/sw/packages/MATLAB/R2020a/bin/glnxa64/mkl.so+05556360
mkl_lapack_dgetrf+00001816
[ 12] 0x00007fc79d50f9f3
/cache/sw/packages/MATLAB/R2020a/bin/glnxa64/../../sys/os/glnxa64/libiomp5.so+00973299
__kmp_invoke_microtask+00000147
[ 13] 0x00007fc79d4cff96
/cache/sw/packages/MATLAB/R2020a/bin/glnxa64/../../sys/os/glnxa64/libiomp5.so+00712598
[ 14] 0x00007fc79d4cef81
/cache/sw/packages/MATLAB/R2020a/bin/glnxa64/../../sys/os/glnxa64/libiomp5.so+00708481
[ 15] 0x00007fc79d50fe8b
/cache/sw/packages/MATLAB/R2020a/bin/glnxa64/../../sys/os/glnxa64/libiomp5.so+00974475
[ 16] 0x00007fc7b4cd2e65 /lib64/libpthread.so.0+00032357
[ 17] 0x00007fc7b446388d /lib64/libc.so.6+01042573 clone+00000109

Abnormal termination:
Segmentation violation

Register State (from fault):
RAX = 0000000000000001 RBX = 00007fc1253f1798
RCX = 0000000000000000 RDX = 0000000000000001
RSP = 00007fc1253f0318 RBP = 0000000000000000
RSI = 000000004aaaaaa8 RDI = 00007fc1253f1798

R8 = 00007fe3a0cba0d0 R9 = 000000004aaaaaa8
R10 = 4b7649c000000000 R11 = 0000000000000000
R12 = 00007fe3a0cba0d0 R13 = 00007fc365ec5ac8
R14 = 00007fc365ec57d8 R15 = 00007fc1253f2048

RIP = 00007fc361b66d11 EFL = 0000000000010246

CS = 0033 FS = 0000 GS = 0000

Stack Trace (from fault):
** This crash report has been saved to disk as
/home/xd101/matlab_crash_dump.9925-6 **"

It seems the MKL related lib was used but I did use the blas and lapack for
software compiling. I am not sure if this is related to the GCC. Maybe, it is
related to the Matlab, so I am waiting the response from the developer of that
software to see what is the exact problem.

In addition, I agree with you that the Matlab should compatible with more GCC
versions for facilitating users. I know they just support more versions in
their MATLAB 2020b, but we do not have it in our cluster now.

Reply via email to