unsupported target builtin ‘__builtin_ia32_vec_perm_v2df’ used in rnflow with vector-select

[llvmbot]

Bugzilla Link	10890
Resolution	FIXED
Resolved on	Oct 07, 2011 07:45
Version	trunk
OS	MacOS X
Reporter	LLVM Bugzilla Contributor

Extended Description

At r139290, the compilation of the Polyhedron 2005 rnflow.f90 benchmark now fails with -msse4 -ffast-math -funroll-loops -O3 -fplugin-arg-dragonegg-enable-gcc-optzns as...

[MacPro:pb05/lin/source] howarth% /sw/lib/gcc4.6/bin/gfortran -fplugin=/sw/lib/gcc4.6/lib/dragonegg.so -msse4 -ffast-math -funroll-loops -O3 -fplugin-arg-dragonegg-enable-gcc-optzns rnflow.f90 -o rnflow
rnflow.f90: In function ‘dtrmv_’:
rnflow.f90:2591:0: error: unsupported target builtin ‘__builtin_ia32_vec_perm_v2df’ used
rnflow.f90:2591:0: error: unsupported target builtin ‘__builtin_ia32_vec_perm_v2df’ used

[llvmbot]

Same failure is seen with the test_fpu.f90 benchmark...

[MacPro:pb05/lin/source] howarth% /sw/lib/gcc4.6/bin/gfortran -fplugin=/sw/lib/gcc4.6/lib/dragonegg.so -msse4 -ffast-math -funroll-loops -O3 -fplugin-arg-dragonegg-enable-gcc-optzns test_fpu.f90 -o test_fpu
test_fpu.f90: In function ‘dtrmv_’:
test_fpu.f90:3865:0: error: unsupported target builtin ‘__builtin_ia32_vec_perm_v2df’ used
test_fpu.f90:3865:0: error: unsupported target builtin ‘__builtin_ia32_vec_perm_v2df’ used

[llvmbot]

This issue is also very common in the fortran source files of xplor-nih.

[llvmbot]

This issue occurs with...

[MacPro:pb05/lin/source] howarth% de-gfortran46 -msse4 -ffast-math -O2 -fplugin-arg-dragonegg-enable-gcc-optzns -ftree-vectorize test_fpu.f90 -o test_fpu
test_fpu.f90: In function ‘dtrmv_’:
test_fpu.f90:3865:0: error: unsupported target builtin ‘__builtin_ia32_vec_perm_v2df’ used
test_fpu.f90:3865:0: error: unsupported target builtin ‘__builtin_ia32_vec_perm_v2df’ used

but not...

[MacPro:pb05/lin/source] howarth% de-gfortran46 -msse4 -ffast-math -O1 -fplugin-arg-dragonegg-enable-gcc-optzns -ftree-vectorize test_fpu.f90 -o test_fpu

[llvmbot]

This is on x86_64-apple-darwin11.

[llvmbot]

reduced testcase from test_fpu.f90

[llvmbot]

The attached reduced test case extracted from test_fpu.f90 demonstrates the use of the unsupported target builtin...

MacPro:~] howarth% /sw/lib/gcc4.6/bin/gfortran -fplugin=/sw/lib/gcc4.6/lib/dragonegg.so -msse4 -ffast-math -funroll-loops -O3 -fplugin-arg-dragonegg-enable-gcc-optzns -c reduced_test_fpu.f90
reduced_test_fpu.f90: In function ‘dtrmv_’:
reduced_test_fpu.f90:276:0: error: unsupported target builtin ‘__builtin_ia32_vec_perm_v2df’ used
reduced_test_fpu.f90:276:0: error: unsupported target builtin ‘__builtin_ia32_vec_perm_v2df’ used

which is suppressed by appending -fno-tree-vectorize.

[llvmbot]

Still present at r139692,

[llvmbot]

Are these failures due to the lack of the ia32_vec_perm_v2df builtin within dragonegg or does llvm lack the required vectorization support to implement this in dragonegg?

[llvmbot]

It looks like support for this just needs to be added to dragonegg.

[llvmbot]

Fixed here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20111003/129347.html