[PATCH] D48099: [AMDGPU] DAG combine to produce V_PERM_B32

[Mark Searles via Phabricator via llvm-commits]

msearles accepted this revision.
msearles added a comment.
This revision is now accepted and ready to land.

LGTM; couple of minor nits on couple of comments



================
Comment at: lib/Target/AMDGPU/SIISelLowering.cpp:6323
+      // Canonicalize the expression in an attempt to have less unique masks
+      // and therefor less registers used to hold the masks.
+      if (LHSMask > RHSMask) {
----------------
s/less/fewer: 'to have fewer unique masks'
typo; s/less/fewer: 'and therefore fewer registers used to hold the masks.'



================
Comment at: lib/Target/AMDGPU/SIISelLowering.cpp:6420
+      // Canonicalize the expression in an attempt to have less unique masks
+      // and therefor less registers used to hold the masks.
+      if (LHSMask > RHSMask) {
----------------
s/less/fewer: 'to have fewer unique masks'
typo; s/less/fewer: 'and therefore fewer registers used to hold the masks.'


https://reviews.llvm.org/D48099