This patch: diff --git a/sysdeps/x86_64/fpu/e_powl.S b/sysdeps/x86_64/fpu/e_powl.S index 4959bea..f6a09c8 100644 --- a/sysdeps/x86_64/fpu/e_powl.S +++ b/sysdeps/x86_64/fpu/e_powl.S @@ -154,7 +154,7 @@ ENTRY(__ieee754_powl) fucompp // 1.0 : x : y fnstsw fxch // x : 1.0 : y - test $4500,%eax + test $0x4500,%eax jz 7f fsub %st(1) // x-1 : 1.0 : y fyl2xp1 // log2(x) : y seems to fix the testcase in: http://gcc.gnu.org/bugzilla/show_bug.cgi?id=49031 It looks a like a typo to me.
Thanks for finding this and posting it here! It was my mistake to put it under the gcc bugzilla, it should of course have been here from the start.
Fixed by http://sourceware.org/git/?p=glibc.git;a=commit;h=8db736347c7aca3201f61e3f05b5f672bcdd5bd9