diff options
| author | Noah Goldstein <goldstein.w.n@gmail.com> | 2022-06-20 13:02:10 -0700 |
|---|---|---|
| committer | Noah Goldstein <goldstein.w.n@gmail.com> | 2022-06-22 19:42:17 -0700 |
| commit | 3079f652d7cc34456aefb412677c01e758922527 (patch) | |
| tree | 5d3f749db7fcfdeb17e3765858894a15c1267e26 | |
| parent | 3edda6a0f013736ca9554a95e553739a41dbd4b7 (diff) | |
| download | glibc-3079f652d7cc34456aefb412677c01e758922527.tar.xz glibc-3079f652d7cc34456aefb412677c01e758922527.zip | |
x86: Replace all sse instructions with vex equivilent in avx+ files
Most of these don't really matter as there was no dirty upper state
but we should generally avoid stray sse when its not needed.
The one case that really matters is in svml_d_tanh4_core_avx2.S:
blendvps %xmm0, %xmm8, %xmm7
When there was a dirty upper state.
Tested on x86_64-linux
75 files changed, 158 insertions, 158 deletions
diff --git a/sysdeps/x86_64/fpu/multiarch/svml_d_acos4_core_avx2.S b/sysdeps/x86_64/fpu/multiarch/svml_d_acos4_core_avx2.S index e19bddd2e2..73025e8b0f 100644 --- a/sysdeps/x86_64/fpu/multiarch/svml_d_acos4_core_avx2.S +++ b/sysdeps/x86_64/fpu/multiarch/svml_d_acos4_core_avx2.S @@ -210,11 +210,11 @@ L(SPECIAL_VALUES_LOOP): L(SCALAR_MATH_CALL): movl %r12d, %r14d - movsd 32(%rsp, %r14, 8), %xmm0 + vmovsd 32(%rsp, %r14, 8), %xmm0 call acos@PLT # LOE rbx r14 r15 r12d r13d xmm0 - movsd %xmm0, 64(%rsp, %r14, 8) + vmovsd %xmm0, 64(%rsp, %r14, 8) /* Process special inputs in loop */ jmp L(SPECIAL_VALUES_LOOP) diff --git a/sysdeps/x86_64/fpu/multiarch/svml_d_acos8_core_avx512.S b/sysdeps/x86_64/fpu/multiarch/svml_d_acos8_core_avx512.S index f4c72c3618..b8cc6dd776 100644 --- a/sysdeps/x86_64/fpu/multiarch/svml_d_acos8_core_avx512.S +++ b/sysdeps/x86_64/fpu/multiarch/svml_d_acos8_core_avx512.S @@ -232,11 +232,11 @@ L(SPECIAL_VALUES_LOOP): L(SCALAR_MATH_CALL): movl %r12d, %r14d - movsd 64(%rsp, %r14, 8), %xmm0 + vmovsd 64(%rsp, %r14, 8), %xmm0 call acos@PLT # LOE rbx r14 r15 r12d r13d xmm0 - movsd %xmm0, 128(%rsp, %r14, 8) + vmovsd %xmm0, 128(%rsp, %r14, 8) |