[llvm] r185994 - Explicitly define ARMISelLowering::isFMAFasterThanFMulAndFAdd. No functionality change.

[Stephen Lin]

Yeah, I figure it might save someone else some time trying to figure
out what is going on with it :D

On Wed, Jul 10, 2013 at 1:26 PM, Jim Grosbach <grosbach at apple.com> wrote:
> Most excellent. Thanks!
>
>
> On Jul 9, 2013, at 6:54 PM, Stephen Lin <stephenwlin at gmail.com> wrote:
>
> Author: stephenwlin
> Date: Tue Jul  9 20:54:24 2013
> New Revision: 185994
>
> URL: http://llvm.org/viewvc/llvm-project?rev=185994&view=rev
> Log:
> Explicitly define ARMISelLowering::isFMAFasterThanFMulAndFAdd. No
> functionality change.
>
> Currently ARM is the only backend that supports FMA instructions (for at
> least some subtargets) but does not implement this virtual, so FMAs are
> never generated except from explicit fma intrinsic calls. Apparently this is
> due to the fact that it supports both fused (one rounding step) and unfused
> (two rounding step) multiply + add instructions. This patch clarifies that
> this the case without changing behavior by implementing the virtual function
> to simply return false, as the default TargetLoweringBase version does.
>
> It is possible that some cpus perform the fused version faster than the
> unfused version and vice-versa, so the function implementation should be
> revisited if hard data is found.
>
> Modified:
>    llvm/trunk/lib/Target/ARM/ARMISelLowering.h
>
> Modified: llvm/trunk/lib/Target/ARM/ARMISelLowering.h
> URL:
> http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/ARM/ARMISelLowering.h?rev=185994&r1=185993&r2=185994&view=diff
> ==============================================================================
> --- llvm/trunk/lib/Target/ARM/ARMISelLowering.h (original)
> +++ llvm/trunk/lib/Target/ARM/ARMISelLowering.h Tue Jul  9 20:54:24 2013
> @@ -458,6 +458,17 @@ namespace llvm {
>     SDValue LowerBUILD_VECTOR(SDValue Op, SelectionDAG &DAG,
>                               const ARMSubtarget *ST) const;
>
> +    /// isFMAFasterThanFMulAndFAdd - Return true if an FMA operation is
> faster
> +    /// than a pair of fmul and fadd instructions. fmuladd intrinsics will
> be
> +    /// expanded to FMAs when this method returns true, otherwise fmuladd
> is
> +    /// expanded to fmul + fadd.
> +    ///
> +    /// ARM supports both fused and unfused multiply-add operations; we
> already
> +    /// lower a pair of fmul and fmadd to the latter so it's not clear that
> there
> +    /// would be a gain or that the gain would be worthwhile enough to risk
> +    /// correctness bugs.
> +    virtual bool isFMAFasterThanFMulAndFAdd(EVT VT) const { return false; }
> +
>     SDValue ReconstructShuffle(SDValue Op, SelectionDAG &DAG) const;
>
>     SDValue LowerCallResult(SDValue Chain, SDValue InFlag,
>
>
> _______________________________________________
> llvm-commits mailing list
> llvm-commits at cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits
>
>