diff options
| author | Stefan Schulze Frielinghaus <stefansf@gcc.gnu.org> | 2025-03-19 16:35:12 +0100 |
|---|---|---|
| committer | Stefan Schulze Frielinghaus <stefansf@gcc.gnu.org> | 2025-03-19 16:35:12 +0100 |
| commit | 7e947040a48d3156c602e88fb938c7bd44e83b28 (patch) | |
| tree | 621731daa9a9835fe4a86adf3cb3fda4c64fab08 /gcc/fortran/iresolve.cc | |
| parent | 8fc2bccec330950c11c7c5c54dc8b0b5798a70cf (diff) | |
| download | gcc-7e947040a48d3156c602e88fb938c7bd44e83b28.zip gcc-7e947040a48d3156c602e88fb938c7bd44e83b28.tar.gz gcc-7e947040a48d3156c602e88fb938c7bd44e83b28.tar.bz2 | |
s390: testsuite: Fix vcond-shift.c
Previously we optimized expressions of the form a < 0 ? -1 : 0 to
(signed)a >> 31 during vcond expanding. Since r15-1638-gaac00d09859cc5
this is done in match.pd. The implementation in the back end as well as
in match.pd are basically the same but still distinct. For the tests in
vcond-shift.c the back end emitted
(xx - (xx >> 31)) >> 1
whereas now via match.pd
((int) ((unsigned int) xx >> 31) + xx) >> 1
which is basically the same. We just have to adapt the scan-assembler
directives w.r.t. signed/unsigned shifts which is done by this patch.
gcc/testsuite/ChangeLog:
* gcc.target/s390/vector/vcond-shift.c: Adapt to new match.pd
rule and change scan-assembler-times for shifts.
Diffstat (limited to 'gcc/fortran/iresolve.cc')
0 files changed, 0 insertions, 0 deletions