[PATCH] D93300: [PowerPC] Exploit paddi instruction on Power 10 for constant materialization
Amy Kwan via Phabricator via llvm-commits
llvm-commits at lists.llvm.org
Wed Feb 17 14:58:19 PST 2021
amyk added a comment.
Just a few minor nit comments.
================
Comment at: llvm/lib/Target/PowerPC/PPCISelDAGToDAG.cpp:1084
+ // Pattern : {zeros}{33-bit value}{ones}
+ // Shift right the Imm by (30 - LZ) bits to construct a negtive 34 bit value,
+ // therefore we can take advantage of PLI's sign-extension semantics, and then
----------------
Minor spelling mistake.
================
Comment at: llvm/test/CodeGen/PowerPC/p10-constants.ll:249
+
+; Leading Zeros + Trailing Ones > 30
+define i64 @t_LZTO() {
----------------
I noticed we don't really use the term `Trailing Ones`, but we use `Following Ones`. Would it be better if we said `Following Ones` here?
================
Comment at: llvm/test/CodeGen/PowerPC/p10-constants.ll:260
+
+; Leading Zeros + Trailing Ones + Following Zeros > 30
+define i64 @t_LZTOFO() {
----------------
Just double checking but does this test correspond to the `(LZ + FO + TO) > 30` case?
If that is the case, should the comment be `Leading Zeros + Following Ones + Trailing Zeros`?
Repository:
rG LLVM Github Monorepo
CHANGES SINCE LAST ACTION
https://reviews.llvm.org/D93300/new/
https://reviews.llvm.org/D93300
More information about the llvm-commits
mailing list