[Coroutine] Improve coro-elide-musttail.ll test

[lxfind]

Summary

The test wasn't sensitive to alias analysis. As you can seen from D95117 when AA is added by default this is affected.
Updating the test so that it coveres both cases for AA analysis.
Note that this patch depends on D95117 to land first.

[aeubanks]

what about a separate RUN line with an emtpy AA pipeline: -aa-pipeline=?

[lxfind]

In D95247#2516136, @aeubanks wrote:

what about a separate RUN line with an emtpy AA pipeline: -aa-pipeline=?

By the way, what's the equivalent of "-aa-pipeline=" for the old pass manager? It doesn't seem to have an effect.

[aeubanks]

I think by default the legacy PM in opt only uses basic-aa, -disable-basic-aa gets rid of that.

[lxfind]

Add the case when AA is disabled

[aeubanks]

I don't know enough about coro-elide to LGTM that part, but the AA part makes sense.

[lxfind]

rebase

llvm/test/Transforms/Coroutines/coro-elide-musttail.ll

 ; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
 ; Only run with new pass manager since old pass manager's alias analysis isn't
 ; powerful enough to tell that the tailcall's arguments don't alias the frame.
 ;
+; RUN: opt < %s -coro-elide -S | FileCheck %s
+; RUN: opt < %s -disable-basic-aa -coro-elide -S | FileCheck %s -check-prefix=NOAA
 ; RUN: opt < %s -passes='coro-elide' -S | FileCheck %s
+; RUN: opt < %s -aa-pipeline= -passes='coro-elide' -S | FileCheck %s -check-prefix=NOAA
 
 %"bar.Frame" = type { void (%"bar.Frame"*)*, void (%"bar.Frame"*)*, %"struct.coroutine<false, int>::promise_type", i1 }
 %"struct.coroutine<false, int>::promise_type" = type { i32 }
 %"foo.Frame" = type { void (%"foo.Frame"*)*, void (%"foo.Frame"*)*, %"struct.coroutine<true, int>::promise_type", i1, %"struct.coroutine<true, int>::promise_type::final_awaitable" }
 %"struct.coroutine<true, int>::promise_type" = type { i32 }
 %"struct.coroutine<true, int>::promise_type::final_awaitable" = type { i8 }
 @"bar.resumers" = private constant [3 x void (%"bar.Frame"*)*] [void (%"bar.Frame"*)* @"bar.resume", void (%"bar.Frame"*)* undef, void (%"bar.Frame"*)* undef]
 
 declare dso_local void @"bar"() align 2
 declare dso_local fastcc void @"bar.resume"(%"bar.Frame"*) align 2
 
+; There is a musttail call. CoroElide won't happen.
 define internal fastcc void @foo.resume_musttail(%"foo.Frame"* %FramePtr) {
 ; CHECK-LABEL: @foo.resume_musttail(
 ; CHECK-NEXT:  entry:
 ; CHECK-NEXT:    [[TMP0:%.*]] = tail call token @llvm.coro.id(i32 16, i8* null, i8* bitcast (void ()* @bar to i8*), i8* bitcast ([3 x void (%bar.Frame*)*]* @bar.resumers to i8*))
 ; CHECK-NEXT:    [[TMP1:%.*]] = tail call i1 @llvm.coro.alloc(token [[TMP0]])
 ; CHECK-NEXT:    [[TMP2:%.*]] = tail call i8* @llvm.coro.begin(token [[TMP0]], i8* null)
 ; CHECK-NEXT:    [[CALL34:%.*]] = call i8* undef()
 ; CHECK-NEXT:    musttail call fastcc void undef(i8* [[CALL34]])
 ; CHECK-NEXT:    ret void
 ;
 entry:
   %0 = tail call token @llvm.coro.id(i32 16, i8* null, i8* bitcast (void ()* @"bar" to i8*), i8* bitcast ([3 x void (%"bar.Frame"*)*]* @"bar.resumers" to i8*))
   %1 = tail call i1 @llvm.coro.alloc(token %0)
   %2 = tail call i8* @llvm.coro.begin(token %0, i8* null)
   call i8* @llvm.coro.subfn.addr(i8* %2, i8 1)
   %call34 = call i8* undef()
   musttail call fastcc void undef(i8* %call34)
   ret void
 }
 
-define internal fastcc void @foo.resume_no_musttail(%"foo.Frame"* %FramePtr) {
-; CHECK-LABEL: @foo.resume_no_musttail(
+; The new frame (TMP0) could potentially alias CALL34, the tailcall attribute on that call must be removed
+define internal fastcc void @foo.resume_no_musttail_with_alias(%"foo.Frame"* %FramePtr) {
+; CHECK-LABEL: @foo.resume_no_musttail_with_alias(
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    [[TMP0:%.*]] = alloca [24 x i8], align 8
+; CHECK-NEXT:    [[VFRAME:%.*]] = bitcast [24 x i8]* [[TMP0]] to i8*
+; CHECK-NEXT:    [[TMP1:%.*]] = tail call token @llvm.coro.id(i32 16, i8* null, i8* bitcast (void ()* @bar to i8*), i8* bitcast ([3 x void (%bar.Frame*)*]* @bar.resumers to i8*))
+; CHECK-NEXT:    call fastcc void undef(i8* [[VFRAME]])
+; CHECK-NEXT:    [[CALL34:%.*]] = call i8* undef()
+; CHECK-NEXT:    call fastcc void undef(i8* [[CALL34]])
+; CHECK-NEXT:    ret void
+;
+entry:
+  %0 = tail call token @llvm.coro.id(i32 16, i8* null, i8* bitcast (void ()* @"bar" to i8*), i8* bitcast ([3 x void (%"bar.Frame"*)*]* @"bar.resumers" to i8*))
+  %1 = tail call i1 @llvm.coro.alloc(token %0)
+  %2 = tail call i8* @llvm.coro.begin(token %0, i8* null)
+  call i8* @llvm.coro.subfn.addr(i8* %2, i8 1)
+  call fastcc void undef(i8* %2)
+  %call34 = call i8* undef()
+  tail call fastcc void undef(i8* %call34)
+  ret void
+}
+
+; The new frame (TMP0) does not alias CALL34, tailcall attribute can reimain. This analysis is only available when alias analysis is enabled.
+define internal fastcc void @foo.resume_no_musttail_no_alias(%"foo.Frame"* %FramePtr) {
+; CHECK-LABEL: @foo.resume_no_musttail_no_alias(
 ; CHECK-NEXT:  entry:
 ; CHECK-NEXT:    [[TMP0:%.*]] = alloca [24 x i8], align 8
 ; CHECK-NEXT:    [[VFRAME:%.*]] = bitcast [24 x i8]* [[TMP0]] to i8*
 ; CHECK-NEXT:    [[TMP1:%.*]] = tail call token @llvm.coro.id(i32 16, i8* null, i8* bitcast (void ()* @bar to i8*), i8* bitcast ([3 x void (%bar.Frame*)*]* @bar.resumers to i8*))
 ; CHECK-NEXT:    [[CALL34:%.*]] = call i8* undef()
 ; CHECK-NEXT:    tail call fastcc void undef(i8* [[CALL34]])
 ; CHECK-NEXT:    ret void
 ;
+; NOAA-LABEL: @foo.resume_no_musttail_no_alias(
+; NOAA-NEXT:  entry:
+; NOAA-NEXT:    [[TMP0:%.*]] = alloca [24 x i8], align 8
+; NOAA-NEXT:    [[VFRAME:%.*]] = bitcast [24 x i8]* [[TMP0]] to i8*
+; NOAA-NEXT:    [[TMP1:%.*]] = call token @llvm.coro.id(i32 16, i8* null, i8* bitcast (void ()* @bar to i8*), i8* bitcast ([3 x void (%bar.Frame*)*]* @bar.resumers to i8*))
+; NOAA-NEXT:    [[CALL34:%.*]] = call i8* undef()
+; NOAA-NEXT:    call fastcc void undef(i8* [[CALL34]])
+; NOAA-NEXT:    ret void
+;
 entry:
   %0 = tail call token @llvm.coro.id(i32 16, i8* null, i8* bitcast (void ()* @"bar" to i8*), i8* bitcast ([3 x void (%"bar.Frame"*)*]* @"bar.resumers" to i8*))
   %1 = tail call i1 @llvm.coro.alloc(token %0)
   %2 = tail call i8* @llvm.coro.begin(token %0, i8* null)
   call i8* @llvm.coro.subfn.addr(i8* %2, i8 1)
   %call34 = call i8* undef()
   tail call fastcc void undef(i8* %call34)
   ret void
 }
 
+
 ; Function Attrs: argmemonly nofree nosync nounwind willreturn
 declare void @llvm.lifetime.start.p0i8(i64 immarg, i8* nocapture) #0
 
 ; Function Attrs: argmemonly nounwind readonly
 declare token @llvm.coro.id(i32, i8* readnone, i8* nocapture readonly, i8*) #1
 
 ; Function Attrs: nounwind
 declare i1 @llvm.coro.alloc(token) #2