[PATCH] D153829: [AArch64] Move branch relaxation after bbsection assignment

Fri Jul 21 13:25:39 PDT 2023

dhoekwater marked an inline comment as done.
dhoekwater added inline comments.

================
Comment at: llvm/test/CodeGen/AArch64/jti-correct-datatype.mir:1
-# RUN: llc -mtriple=aarch64-linux-gnu -start-after=branch-relaxation --filetype=obj -o %t.o %s
+# RUN: llc -mtriple=aarch64-linux-gnu -start-before=aarch64-jump-tables --filetype=obj -o %t.o %s

----------------
nickdesaulniers wrote:
> is this the next pass that would have run after branch-relaxation before this change?
Short answer: it turns out I don't actually need to modify this test, so I'm removing the change from the patch.

Long answer: 
Before this change, `aarch64-jump-tables` isn't the next pass that would have run after `branch-relaxation`, but it //is// the next pass that modifies jump table datatypes. IIUC, this test applies to the [[ https://reviews.llvm.org/D70533 | jump compression pass ]] and CodeGen passes afterwards, so starting it before aarch64-jump-tables //should// be alright.

================
Comment at: llvm/test/DebugInfo/AArch64/fallthrough-branch.ll:10
   %2 = alloca i1, align 8
   %3 = bitcast i1* %2 to i8*
   call void @llvm.memset.p0i8.i64(i8* align 8 %3, i8 0, i64 1, i1 false)
----------------
arsenm wrote:
> How did this test escape opaquification?
It's not exactly self-documenting, that's for sure.

Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D153829/new/

https://reviews.llvm.org/D153829