[llvm-commits] [llvm] r78081 - in /llvm/trunk: lib/Target/ARM/ARM.td lib/Target/ARM/ARMInstrFormats.td lib/Target/ARM/ARMInstrInfo.td lib/Target/ARM/ARMInstrNEON.td lib/Target/ARM/ARMInstrVFP.td lib/Target/ARM/ARMSubtarget.cpp lib/Target/ARM/ARMS

[Sandeep Patel]

Have you done any benchmarking that suggests that this is a win? The
docs we have for the Cortex-A9 suggest that NEON operations have
longer latency than VFP ops. (I don't have the A8 docs to compare
with.)

deep

On Tue, Aug 4, 2009 at 5:53 PM, David Goodwin<david_goodwin at apple.com> wrote:
> Author: david_goodwin
> Date: Tue Aug  4 12:53:06 2009
> New Revision: 78081
>
> URL: http://llvm.org/viewvc/llvm-project?rev=78081&view=rev
> Log:
> Initial support for single-precision FP using NEON. Added "neonfp" attribute to enable. Added patterns for some binary FP operations.