https://sourceware.org/bugzilla/show_bug.cgi?id=33660
--- Comment #2 from Mingye Wang <arthur200126 at gmail dot com> ---
Further correction:
* __masked and __nomask should be replaced with the standard "inbranch" and
"notinbranch".
* I forgot to add a part corresponding to "b" (128-bit vector, "SSE" [actually
SSE2/SSE3/SSSE3/SSE4]).
Corrected speculative demangling:
>>> _ZGVbN2v_sin
_Pragma("omp declare simd __isa(SSE) notinbranch simdlen(2) __vector(#1)") sin
>>> _ZGVbM4ua16vl__Z3fooPffi
_Pragma("omp declare simd __isa(SSE) inbranch simdlen(4) uniform(#1)
aligned(#1:16) __vector(#2) linear(#3:1)") foo(float*, float, int)
--
You are receiving this mail because:
You are on the CC list for the bug.
