| commit | 272708884cb750f12f5c74a00e6620c19dc6d567 | |
| author | Adhemerval Zanella <adhemerval.zanella@linaro.org> | |
| Thu, 8 Feb 2024 13:08:39 +0000 (8 10:08 -0300) | ||
| committer | H.J. Lu <hjl.tools@gmail.com> | |
| Tue, 13 Feb 2024 16:49:13 +0000 (13 08:49 -0800) | ||
| tree | f6c6fdc807ba59154f501111a69c4c6ef812e9d9 | treesnapshot (tar.gz zip) |
| parent | 0c0d39fe4aeb0f69b26e76337c5dfd5530d5d44e | commitdiff |
x86: Do not prefer ERMS for memset on Zen3+
For AMD Zen3+ architecture, the performance of the vectorized loop is
slightly better than ERMS.
Checked on x86_64-linux-gnu on Zen3.
Reviewed-by: H.J. Lu <hjl.tools@gmail.com>
For AMD Zen3+ architecture, the performance of the vectorized loop is
slightly better than ERMS.
Checked on x86_64-linux-gnu on Zen3.
Reviewed-by: H.J. Lu <hjl.tools@gmail.com>
| sysdeps/x86/dl-cacheinfo.h | diffblobblamehistory |