Remove (unneeded) '-asan-use-after-return' from hoist-argument-init-insts.ll.

[kda]

Summary

Remove (unneeded) '-asan-use-after-return' from hoist-argument-init-insts.ll.

[kda]

Vendant,
I was fiddling around with UAR, and I noticed your test seemed to pass without it, as it is the default. It also passes when true.

[vsk]

Is there a planned change to the use-after-return instrumentation that would regress this test as-is?

[kda]

No.
(My planned change won't change this behavior:
https://github.com/google/sanitizers/issues/1394)
I found the existing test misleading.
Changing or removing the flag does not change whether the test passes.

If the test is guarding against something changing when
'-asan-use-after-return=1', then perhaps a comment making that clear would
be useful (or even better explicit code that shows what is being avoided is
NOT present).
I don't have a clear understanding how the two features interaction (UAR
and UAS). So, this sort of test may need to cover the UAR case.

FYI, I'm new here, and I try to leave the code better (clearer) when
possible.
I'm also okay, with "bug off", just leave it alone. It is not broken.

[vsk]

Great - I just wanted to understand the motivation. Thanks for the cleanup!

llvm/test/Instrumentation/AddressSanitizer/hoist-argument-init-insts.ll

-; RUN: opt < %s -asan -asan-module -enable-new-pm=0 -asan-use-after-return -S | FileCheck %s
-; RUN: opt < %s -passes='asan-pipeline' -asan-use-after-return -S | FileCheck %s
+; RUN: opt < %s -asan -asan-module -enable-new-pm=0 -S | FileCheck %s
+; RUN: opt < %s -passes='asan-pipeline' -S | FileCheck %s
 
 ; Source (-O0 -fsanitize=address -fsanitize-address-use-after-scope):
 ;; struct S { int x, y; };
 ;; void swap(S *a, S *b, bool doit) {
 ;;   if (!doit)
 ;;     return;
 ;;   auto tmp = *a;
 ;;   *a = *b;