Mercurial > projects > ldc
diff gen/naked.cpp @ 1002:c749648ed2b8
Fix cfloat return on x86_64: only perform ABI transformation for non-extern(D)
functions.
There's no need to waste cycles with extern(D), which we get to define
ourselves. Fixes tests/mini/asm8.d. (Since the asm abiret code already assumed
{xmm0, xmm1} returns)
author | Frits van Bommel <fvbommel wxs.nl> |
---|---|
date | Thu, 26 Feb 2009 23:35:39 +0100 |
parents | 2667e3a145be |
children | e8c6dbcd33d1 |
line wrap: on
line diff
--- a/gen/naked.cpp Thu Feb 26 22:47:06 2009 +0100 +++ b/gen/naked.cpp Thu Feb 26 23:35:39 2009 +0100 @@ -279,7 +279,7 @@ // For compatibility, use the GCC/LLVM-GCC way for extern(C/Windows) // extern(C) cfloat -> %xmm0 (extract two floats) as->out_c = "={xmm0},"; - asmblock->retty = LLStructType::get(LLType::DoubleTy, NULL);; + asmblock->retty = LLType::DoubleTy; asmblock->retfixup = &x86_64_cfloatRetFixup; } else if (rt->iscomplex()) { // cdouble and extern(D) cfloat -> re=%xmm0, im=%xmm1