diff options
| author | liuhongt <hongtao.liu@intel.com> | 2024-05-29 11:14:26 +0800 |
|---|---|---|
| committer | Haochen Jiang <haochen.jiang@intel.com> | 2024-05-29 11:14:26 +0800 |
| commit | b644126237a1aa8599f767a5e0bbada1d7286f44 (patch) | |
| tree | 7536f7dff5c2ddda9fedfb2c0815ec03bbdd1ddf /libcpp | |
| parent | 00ed5424b1d4dcccfa187f55205521826794898c (diff) | |
| download | gcc-b644126237a1aa8599f767a5e0bbada1d7286f44.zip gcc-b644126237a1aa8599f767a5e0bbada1d7286f44.tar.gz gcc-b644126237a1aa8599f767a5e0bbada1d7286f44.tar.bz2 | |
Align tight&hot loop without considering max skipping bytes.
When hot loop is small enough to fix into one cacheline, we should align
the loop with ceil_log2 (loop_size) without considering maximum
skipp bytes. It will help code prefetch.
gcc/ChangeLog:
* config/i386/i386.cc (ix86_avoid_jump_mispredicts): Change
gen_pad to gen_max_skip_align.
(ix86_align_loops): New function.
(ix86_reorg): Call ix86_align_loops.
* config/i386/i386.md (pad): Rename to ..
(max_skip_align): .. this, and accept 2 operands for align and
skip.
Diffstat (limited to 'libcpp')
0 files changed, 0 insertions, 0 deletions