[llvm] [AMX] Fix LIT test with adjustsStack attribute for expensive checks build. (PR #156808)

[via llvm-commits]

llvmbot wrote:


<!--LLVM PR SUMMARY COMMENT-->

@llvm/pr-subscribers-backend-x86

Author: Karthik Senthil (karthik-senthil)

<details>
<summary>Changes</summary>

The generated MIR fails machine verifier as stack pointer is being modified without appropriate attributes in frameInfo. This PR fixes this issue by adding adjustsStack=true attribute. Fixes the post commit regression identified in #<!-- -->155673.

---
Full diff: https://github.com/llvm/llvm-project/pull/156808.diff


1 Files Affected:

- (modified) llvm/test/CodeGen/X86/AMX/amx-sink-config-after-calls.mir (+2-1) 


``````````diff
diff --git a/llvm/test/CodeGen/X86/AMX/amx-sink-config-after-calls.mir b/llvm/test/CodeGen/X86/AMX/amx-sink-config-after-calls.mir
index b76faf6cbb88b..82049dce8a45c 100644
--- a/llvm/test/CodeGen/X86/AMX/amx-sink-config-after-calls.mir
+++ b/llvm/test/CodeGen/X86/AMX/amx-sink-config-after-calls.mir
@@ -1,5 +1,5 @@
 # NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py UTC_ARGS: --version 5
-# RUN: llc -mtriple=x86_64-- -mattr=+amx-int8,avx512f -run-pass="fastpretileconfig,regallocfast,fasttileconfig" -o - %s | FileCheck %s
+# RUN: llc -mtriple=x86_64-- -mattr=+amx-int8,avx512f -run-pass="fastpretileconfig,regallocfast,fasttileconfig" -verify-machineinstrs -o - %s | FileCheck %s
 
 # Test to verify that ldtilecfg instructions are sinked closer to tile defining
 # instructions after a call. This ensures call does not overwrite values in
@@ -43,6 +43,7 @@ liveins:
   - { reg: '$rdi', virtual-reg: '%0' }
   - { reg: '$rsi', virtual-reg: '%2' }
 frameInfo:
+  adjustsStack:    true
   maxAlignment:    1024
 stack:
   - { id: 0, size: 1024, alignment: 1024 }

``````````

</details>


https://github.com/llvm/llvm-project/pull/156808