[llvm] r264683 - [PowerPC] Clarify a comment in PPCTTI about vector loads
Hal Finkel via llvm-commits
llvm-commits at lists.llvm.org
Mon Mar 28 15:39:35 PDT 2016
Author: hfinkel
Date: Mon Mar 28 17:39:35 2016
New Revision: 264683
URL: http://llvm.org/viewvc/llvm-project?rev=264683&view=rev
Log:
[PowerPC] Clarify a comment in PPCTTI about vector loads
This should say that we could do unaligned vector loads on the P7 using VSX
instructions, not that we should.
Modified:
llvm/trunk/lib/Target/PowerPC/PPCTargetTransformInfo.cpp
Modified: llvm/trunk/lib/Target/PowerPC/PPCTargetTransformInfo.cpp
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/PowerPC/PPCTargetTransformInfo.cpp?rev=264683&r1=264682&r2=264683&view=diff
==============================================================================
--- llvm/trunk/lib/Target/PowerPC/PPCTargetTransformInfo.cpp (original)
+++ llvm/trunk/lib/Target/PowerPC/PPCTargetTransformInfo.cpp Mon Mar 28 17:39:35 2016
@@ -375,7 +375,7 @@ int PPCTTIImpl::getMemoryOpCost(unsigned
// If we can use the permutation-based load sequence, then this is also
// relatively cheap (not counting loop-invariant instructions): one load plus
// one permute (the last load in a series has extra cost, but we're
- // neglecting that here). Note that on the P7, we should do unaligned loads
+ // neglecting that here). Note that on the P7, we could do unaligned loads
// for Altivec types using the VSX instructions, but that's more expensive
// than using the permutation-based load sequence. On the P8, that's no
// longer true.
More information about the llvm-commits
mailing list