[llvm] [AMDGPU] add tests for loop definition of bitconvert (PR #133052)

[LLVM Continuous Integration via llvm-commits]

llvm-ci wrote:

LLVM Buildbot has detected a new failure on builder `llvm-clang-x86_64-expensive-checks-debian` running on `gribozavr4` while building `llvm` at step 6 "test-build-unified-tree-check-all".

Full details are available at: https://lab.llvm.org/buildbot/#/builders/16/builds/16963

<details>
<summary>Here is the relevant piece of the build log for the reference</summary>

```
Step 6 (test-build-unified-tree-check-all) failure: test (failure)
******************** TEST 'LLVM :: CodeGen/AMDGPU/amdgcn.bitcast.1024bit.ll' FAILED ********************
Exit Code: 1

Command Output (stderr):
--
/b/1/llvm-clang-x86_64-expensive-checks-debian/build/bin/llc -mtriple=amdgcn  < /b/1/llvm-clang-x86_64-expensive-checks-debian/llvm-project/llvm/test/CodeGen/AMDGPU/amdgcn.bitcast.1024bit.ll | /b/1/llvm-clang-x86_64-expensive-checks-debian/build/bin/FileCheck -check-prefix=GCN /b/1/llvm-clang-x86_64-expensive-checks-debian/llvm-project/llvm/test/CodeGen/AMDGPU/amdgcn.bitcast.1024bit.ll # RUN: at line 3
+ /b/1/llvm-clang-x86_64-expensive-checks-debian/build/bin/llc -mtriple=amdgcn
+ /b/1/llvm-clang-x86_64-expensive-checks-debian/build/bin/FileCheck -check-prefix=GCN /b/1/llvm-clang-x86_64-expensive-checks-debian/llvm-project/llvm/test/CodeGen/AMDGPU/amdgcn.bitcast.1024bit.ll
/b/1/llvm-clang-x86_64-expensive-checks-debian/build/bin/llc -mtriple=amdgcn -mcpu=tonga  < /b/1/llvm-clang-x86_64-expensive-checks-debian/llvm-project/llvm/test/CodeGen/AMDGPU/amdgcn.bitcast.1024bit.ll | /b/1/llvm-clang-x86_64-expensive-checks-debian/build/bin/FileCheck -check-prefixes=VI /b/1/llvm-clang-x86_64-expensive-checks-debian/llvm-project/llvm/test/CodeGen/AMDGPU/amdgcn.bitcast.1024bit.ll # RUN: at line 4
+ /b/1/llvm-clang-x86_64-expensive-checks-debian/build/bin/FileCheck -check-prefixes=VI /b/1/llvm-clang-x86_64-expensive-checks-debian/llvm-project/llvm/test/CodeGen/AMDGPU/amdgcn.bitcast.1024bit.ll
+ /b/1/llvm-clang-x86_64-expensive-checks-debian/build/bin/llc -mtriple=amdgcn -mcpu=tonga

# After SI post-RA bundler
# Machine code for function bitcast_v64i16_to_v128i8: NoPHIs, TracksLiveness, NoVRegs, TiedOpsRewritten, TracksDebugUserValues
Frame Objects:
  fi#-3: size=4, align=8, fixed, at location [SP+8]
  fi#-2: size=4, align=4, fixed, at location [SP+4]
  fi#-1: size=4, align=16, fixed, at location [SP]
  fi#0: size=8, align=4, at location [SP+76]
  fi#1: size=8, align=4, at location [SP+84]
  fi#2: size=8, align=4, at location [SP+92]
  fi#3: size=8, align=4, at location [SP+100]
  fi#4: size=8, align=4, at location [SP+108]
  fi#5: size=8, align=4, at location [SP+116]
  fi#6: size=8, align=4, at location [SP+124]
  fi#7: size=8, align=4, at location [SP+132]
  fi#8: size=8, align=4, at location [SP+140]
  fi#9: size=8, align=4, at location [SP+148]
  fi#10: size=4, align=4, at location [SP+156]
  fi#11: size=4, align=4, at location [SP+160]
  fi#12: size=4, align=4, at location [SP+164]
  fi#13: size=4, align=4, at location [SP+168]
  fi#14: size=4, align=4, at location [SP+172]
  fi#15: size=4, align=4, at location [SP+176]
  fi#16: size=4, align=4, at location [SP+180]
  fi#17: size=4, align=4, at location [SP+184]
  fi#18: size=4, align=4, at location [SP+188]
  fi#19: size=4, align=4, at location [SP+192]
  fi#20: size=4, align=4, at location [SP+196]
  fi#21: size=4, align=4, at location [SP+200]
  fi#22: size=4, align=4, at location [SP+204]
  fi#23: size=4, align=4, at location [SP+208]
  fi#24: size=4, align=4, at location [SP+212]
  fi#25: size=4, align=4, at location [SP+216]
  fi#26: size=4, align=4, at location [SP+220]
  fi#27: size=4, align=4, at location [SP+224]
  fi#28: size=4, align=4, at location [SP+228]
  fi#29: size=4, align=4, at location [SP+232]
  fi#30: size=4, align=4, at location [SP+236]
  fi#31: size=4, align=4, at location [SP+240]
...

```

</details>

https://github.com/llvm/llvm-project/pull/133052