[llvm] [GlobalISel] Fix a check that aligned tail call is lowered (PR #82016)

[via llvm-commits]

llvmbot wrote:


<!--LLVM PR SUMMARY COMMENT-->

@llvm/pr-subscribers-backend-x86

Author: Evgenii Kudriashov (e-kud)

<details>
<summary>Changes</summary>

Despite of a valid tail call opportunity, backends still may not generate a tail call or such lowering is not implemented yet.

Check that lowering has happened instead of its possibility when generating G_ASSERT_ALIGN.

---
Full diff: https://github.com/llvm/llvm-project/pull/82016.diff


2 Files Affected:

- (modified) llvm/lib/CodeGen/GlobalISel/CallLowering.cpp (+1-1) 
- (added) llvm/test/CodeGen/X86/GlobalISel/calllowering-tailcall.ll (+24) 


``````````diff
diff --git a/llvm/lib/CodeGen/GlobalISel/CallLowering.cpp b/llvm/lib/CodeGen/GlobalISel/CallLowering.cpp
index 3bd1542eeb7465..77dc265d795d00 100644
--- a/llvm/lib/CodeGen/GlobalISel/CallLowering.cpp
+++ b/llvm/lib/CodeGen/GlobalISel/CallLowering.cpp
@@ -187,7 +187,7 @@ bool CallLowering::lowerCall(MachineIRBuilder &MIRBuilder, const CallBase &CB,
   if (!lowerCall(MIRBuilder, Info))
     return false;
 
-  if (ReturnHintAlignReg && !Info.IsTailCall) {
+  if (ReturnHintAlignReg && !Info.LoweredTailCall) {
     MIRBuilder.buildAssertAlign(ResRegs[0], ReturnHintAlignReg,
                                 ReturnHintAlign);
   }
diff --git a/llvm/test/CodeGen/X86/GlobalISel/calllowering-tailcall.ll b/llvm/test/CodeGen/X86/GlobalISel/calllowering-tailcall.ll
new file mode 100644
index 00000000000000..6a856c32eb261f
--- /dev/null
+++ b/llvm/test/CodeGen/X86/GlobalISel/calllowering-tailcall.ll
@@ -0,0 +1,24 @@
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 4
+; RUN: llc -mtriple=x86_64-linux-gnu -global-isel < %s | FileCheck %s --check-prefix=X64
+; RUN: llc -mtriple=i686-linux-gnu -global-isel < %s | FileCheck %s --check-prefix=X86
+
+declare ptr @foo()
+
+define ptr @aligned_tailcall() nounwind {
+; X64-LABEL: aligned_tailcall:
+; X64:       # %bb.0: # %entry
+; X64-NEXT:    pushq %rax
+; X64-NEXT:    callq foo
+; X64-NEXT:    popq %rcx
+; X64-NEXT:    retq
+;
+; X86-LABEL: aligned_tailcall:
+; X86:       # %bb.0: # %entry
+; X86-NEXT:    subl $12, %esp
+; X86-NEXT:    calll foo
+; X86-NEXT:    addl $12, %esp
+; X86-NEXT:    retl
+entry:
+  %call = tail call align 8 ptr @foo()
+  ret ptr %call
+}

``````````

</details>


https://github.com/llvm/llvm-project/pull/82016