projects / oota-llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 9d796db) | patch
Add support for FP_ROUND from v2f64 to v2f32
authorMichael Liao <michael.liao@intel.com>
Wed, 10 Oct 2012 16:53:28 +0000 (16:53 +0000)
committerMichael Liao <michael.liao@intel.com>
Wed, 10 Oct 2012 16:53:28 +0000 (16:53 +0000)
commit44c2d61b6703469a95fcd2d5397c5d09a67e75c1
treeded8fbfaa22a59e8aca8944eb2971cd0e0b42565tree | snapshot
parent9d796db3e746c31dbdb605510c53b3da98d71b38commit | diff
Add support for FP_ROUND from v2f64 to v2f32

- Due to the current matching vector elements constraints in
  ISD::FP_ROUND, rounding from v2f64 to v4f32 (after legalization from
  v2f32) is scalarized. Add a customized v2f32 widening to convert it
  into a target-specific X86ISD::VFPROUND to work around this
  constraints.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@165631 91177308-0d34-0410-b5e6-96231b3b80d8
lib/Target/X86/X86ISelLowering.cpp diff | blob | history
lib/Target/X86/X86ISelLowering.h diff | blob | history
lib/Target/X86/X86InstrFragmentsSIMD.td diff | blob | history
lib/Target/X86/X86InstrSSE.td diff | blob | history
test/CodeGen/X86/fp-load-trunc.ll [new file with mode: 0644] blob
test/CodeGen/X86/fp-trunc.ll diff | blob | history
C/C++ LLVM-based compilers that forbids OOTA behaviors