[PATCH] D124967: [SROA] Avoid postponing rewriting load/store by ignoring lifetime intrinsics in partition's promotability checking

Artem Belevich via Phabricator via llvm-commits llvm-commits at lists.llvm.org
Wed May 4 15:54:15 PDT 2022


tra added inline comments.


================
Comment at: llvm/test/Transforms/SROA/lifetime-intrinsic.ll:1
+; RUN: opt < %s -passes=sroa -S | FileCheck %s
+target datalayout = "e-p:64:64:64-p3:32:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-i128:128:128-f32:32:32-f64:64:64-v16:16:16-v32:32:32-v64:64:64-v128:128:128-n16:32:64"
----------------
I'd add a comment describing what is it we're looking for here.
It's not clear from the test itself why we're checking for those phys. To make sure that we don't do excessive shifting and masking of i32->i16->i32 ?


================
Comment at: llvm/test/Transforms/SROA/lifetime-intrinsic.ll:26
+; CHECK-LABEL: while.cond.i:
+; CHECK: [[ACC_I_SROA_6_0:%.*]] = phi i32 [ undef, %entry ], [ %asmresult70.i.i, %while.body.i ]
+; CHECK: [[ACC_I_SROA_0_0:%.*]] = phi i32 [ undef, %entry ], [ %asmresult.i.i, %while.body.i ]
----------------
These labels look overly specific. You should probably capture the actual names where the values are set.



Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D124967/new/

https://reviews.llvm.org/D124967



More information about the llvm-commits mailing list