diff options
| author | Szabolcs Nagy <nsz@port70.net> | 2014-11-05 21:40:29 +0100 |
|---|---|---|
| committer | Rich Felker <dalias@aerifal.cx> | 2015-03-30 01:41:32 -0400 |
| commit | 8c245bf2725f3171ee8f58dab7169e5d682de026 (patch) | |
| tree | ae92d5d0536aa3b15adb96d62b4487488998e5a8 /src/math/x86_64/fmodl.s | |
| parent | e48fff8d338d85616dc650cc8b4dd5ae00083da1 (diff) | |
| download | musl-8c245bf2725f3171ee8f58dab7169e5d682de026.tar.gz musl-8c245bf2725f3171ee8f58dab7169e5d682de026.tar.xz musl-8c245bf2725f3171ee8f58dab7169e5d682de026.zip | |
math: fix x86_64 and x32 asm not to use sahf instruction
Some early x86_64 cpus (released before 2006) did not support sahf/lahf instructions so they should be avoided (intel manual says they are only supported if CPUID.80000001H:ECX.LAHF-SAHF[bit 0] = 1). The workaround simplifies exp2l and expm1l because fucomip can be used instead of the fucomp;fnstsw;sahf sequence copied from i386. In fmodl and remainderl sahf is replaced by a simple bit test. (cherry picked from commit a732e80d33b4fd6f510f7cec4f5573ef5d89bc4e)
Diffstat (limited to 'src/math/x86_64/fmodl.s')
| -rw-r--r-- | src/math/x86_64/fmodl.s | 4 |
1 files changed, 2 insertions, 2 deletions
diff --git a/src/math/x86_64/fmodl.s b/src/math/x86_64/fmodl.s index ca81e60c..cd8d2b7c 100644 --- a/src/math/x86_64/fmodl.s +++ b/src/math/x86_64/fmodl.s @@ -5,7 +5,7 @@ fmodl: fldt 8(%rsp) 1: fprem fstsw %ax - sahf - jp 1b + testb $4,%ah + jnz 1b fstp %st(1) ret |