diff options
| author | H.J. Lu <hjl.tools@gmail.com> | 2017-03-21 10:59:31 -0700 |
|---|---|---|
| committer | H.J. Lu <hjl.tools@gmail.com> | 2017-04-20 07:55:44 -0700 |
| commit | 883cadc5543ffd3a4537498b44c782ded8a4a4e8 (patch) | |
| tree | 7f8420bd13c9e7660a4e097c9f62b01a2695a2d2 /sysdeps/x86_64/dl-trampoline.S | |
| parent | 83037ea1d9e84b1b44ed307f01cbb5eeac24e22d (diff) | |
| download | glibc-hjl/pr21258/2.23.tar.gz glibc-hjl/pr21258/2.23.tar.xz glibc-hjl/pr21258/2.23.zip | |
x86-64: Improve branch predication in _dl_runtime_resolve_avx512_opt [BZ #21258] hjl/pr21258/2.23
On Skylake server, _dl_runtime_resolve_avx512_opt is used to preserve
the first 8 vector registers. The code layout is
if only %xmm0 - %xmm7 registers are used
preserve %xmm0 - %xmm7 registers
if only %ymm0 - %ymm7 registers are used
preserve %ymm0 - %ymm7 registers
preserve %zmm0 - %zmm7 registers
Branch predication always executes the fallthrough code path to preserve
%zmm0 - %zmm7 registers speculatively, even though only %xmm0 - %xmm7
registers are used. This leads to lower CPU frequency on Skylake
server. This patch changes the fallthrough code path to preserve
%xmm0 - %xmm7 registers instead:
if whole %zmm0 - %zmm7 registers are used
preserve %zmm0 - %zmm7 registers
if only %ymm0 - %ymm7 registers are used
preserve %ymm0 - %ymm7 registers
preserve %xmm0 - %xmm7 registers
Tested on Skylake server.
[BZ #21258]
* sysdeps/x86_64/dl-trampoline.S (_dl_runtime_resolve_opt):
Define only if _dl_runtime_resolve is defined to
_dl_runtime_resolve_sse_vex.
* sysdeps/x86_64/dl-trampoline.h (_dl_runtime_resolve_opt):
Fallthrough to _dl_runtime_resolve_sse_vex.
(cherry picked from commit c15f8eb50cea7ad1a4ccece6e0982bf426d52c00)
Diffstat (limited to 'sysdeps/x86_64/dl-trampoline.S')
| -rw-r--r-- | sysdeps/x86_64/dl-trampoline.S | 5 |
1 files changed, 1 insertions, 4 deletions
diff --git a/sysdeps/x86_64/dl-trampoline.S b/sysdeps/x86_64/dl-trampoline.S index 3f812b89c0..177f666164 100644 --- a/sysdeps/x86_64/dl-trampoline.S +++ b/sysdeps/x86_64/dl-trampoline.S @@ -88,11 +88,9 @@ # endif # define VEC(i) zmm##i # define _dl_runtime_resolve _dl_runtime_resolve_avx512 -# define _dl_runtime_resolve_opt _dl_runtime_resolve_avx512_opt # define _dl_runtime_profile _dl_runtime_profile_avx512 # include "dl-trampoline.h" # undef _dl_runtime_resolve -# undef _dl_runtime_resolve_opt # undef _dl_runtime_profile # undef VEC # undef VMOV @@ -101,8 +99,6 @@ #else strong_alias (_dl_runtime_resolve_avx, _dl_runtime_resolve_avx512) .hidden _dl_runtime_resolve_avx512 -strong_alias (_dl_runtime_resolve_avx_opt, _dl_runtime_resolve_avx512_opt) - .hidden _dl_runtime_resolve_avx512_opt strong_alias (_dl_runtime_profile_avx, _dl_runtime_profile_avx512) .hidden _dl_runtime_profile_avx512 #endif @@ -154,4 +150,5 @@ strong_alias (_dl_runtime_profile_avx, _dl_runtime_profile_avx512) # define VMOV vmovdqu #endif #define _dl_runtime_resolve _dl_runtime_resolve_sse_vex +#define _dl_runtime_resolve_opt _dl_runtime_resolve_avx512_opt #include "dl-trampoline.h" |