diff options
| author | Dale Johannesen <dalej@apple.com> | 2010-04-20 22:34:09 +0000 |
|---|---|---|
| committer | Dale Johannesen <dalej@apple.com> | 2010-04-20 22:34:09 +0000 |
| commit | 9413edc0257abc50efe873d949163718a0a3832c (patch) | |
| tree | 51ea6891b55970e100d8b3e93ddbd4985886567a /lib/CodeGen/SelectionDAG/TargetLowering.cpp | |
| parent | dc433dba32a2421e6c0deab9fcf87b9026c9bbd0 (diff) | |
| download | external_llvm-9413edc0257abc50efe873d949163718a0a3832c.zip external_llvm-9413edc0257abc50efe873d949163718a0a3832c.tar.gz external_llvm-9413edc0257abc50efe873d949163718a0a3832c.tar.bz2 | |
Because of the EMMS problem, right now we have to support
user-defined operations that use MMX register types, but
the compiler shouldn't generate them on its own. This adds
a Synthesizable abstraction to represent this, and changes
the vector widening computation so it won't produce MMX types.
(The motivation is to remove noise from the ABI compatibility
part of the gcc test suite, which has some breakage right now.)
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@101951 91177308-0d34-0410-b5e6-96231b3b80d8
Diffstat (limited to 'lib/CodeGen/SelectionDAG/TargetLowering.cpp')
| -rw-r--r-- | lib/CodeGen/SelectionDAG/TargetLowering.cpp | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/lib/CodeGen/SelectionDAG/TargetLowering.cpp b/lib/CodeGen/SelectionDAG/TargetLowering.cpp index 11dca39..5c3992e 100644 --- a/lib/CodeGen/SelectionDAG/TargetLowering.cpp +++ b/lib/CodeGen/SelectionDAG/TargetLowering.cpp @@ -720,7 +720,7 @@ void TargetLowering::computeRegisterProperties() { unsigned NElts = VT.getVectorNumElements(); for (unsigned nVT = i+1; nVT <= MVT::LAST_VECTOR_VALUETYPE; ++nVT) { EVT SVT = (MVT::SimpleValueType)nVT; - if (isTypeLegal(SVT) && SVT.getVectorElementType() == EltVT && + if (isTypeSynthesizable(SVT) && SVT.getVectorElementType() == EltVT && SVT.getVectorNumElements() > NElts && NElts != 1) { TransformToType[i] = SVT; ValueTypeActions.setTypeAction(VT, Promote); |