aboutsummaryrefslogtreecommitdiff
path: root/gcc/gimple-isel.cc
diff options
context:
space:
mode:
authorKewen Lin <linkw@linux.ibm.com>2020-11-26 03:20:40 -0600
committerKewen Lin <linkw@linux.ibm.com>2020-11-26 05:05:43 -0600
commitf6702584aa372de7790b3f2be9e4b5a5423eb9fa (patch)
tree63b600c7a23b318fa0d585340e6e64e24b5690f9 /gcc/gimple-isel.cc
parenta386566118054e08bb733f1248649fb6847c407e (diff)
downloadgcc-f6702584aa372de7790b3f2be9e4b5a5423eb9fa.zip
gcc-f6702584aa372de7790b3f2be9e4b5a5423eb9fa.tar.gz
gcc-f6702584aa372de7790b3f2be9e4b5a5423eb9fa.tar.bz2
rs6000: Set param_vect_partial_vector_usage as 1 for P10
Due to the unexpected performance on Power9 of those vector with length instructions, we didn't enable vectorization with partial vectors before. Some recent testings show that they perform expectedly on Power10 now. This patch is to set param_vect_partial_vector_usage as 1 for P10 or later. The performance evaluation on the whole SPEC2017 with option set power10/Ofast/unroll shows it can speed up 525.x264_r by 10.80% and 554.roms_r by 1.94%. Bootstrapped/regtested on powerpc64le-linux-gnu P10. gcc/ChangeLog: * config/rs6000/rs6000.c (rs6000_option_override_internal): Set param_vect_partial_vector_usage as 1 for Power10 and up by default.
Diffstat (limited to 'gcc/gimple-isel.cc')
0 files changed, 0 insertions, 0 deletions