[llvm] [AMDGPU][Scheduler] Refactor ArchVGPR rematerialization during scheduling (PR #125885)

[LLVM Continuous Integration via llvm-commits]

llvm-ci wrote:

LLVM Buildbot has detected a new failure on builder `clang-x86_64-debian-fast` 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/56/builds/25326

<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/mfma-loop.ll' FAILED ********************
Exit Code: 1

Command Output (stderr):
--
/b/1/clang-x86_64-debian-fast/llvm.obj/bin/llc -mtriple=amdgcn -mcpu=gfx908 -verify-machineinstrs < /b/1/clang-x86_64-debian-fast/llvm.src/llvm/test/CodeGen/AMDGPU/mfma-loop.ll | /b/1/clang-x86_64-debian-fast/llvm.obj/bin/FileCheck -enable-var-scope -check-prefixes=GFX908 /b/1/clang-x86_64-debian-fast/llvm.src/llvm/test/CodeGen/AMDGPU/mfma-loop.ll # RUN: at line 2
+ /b/1/clang-x86_64-debian-fast/llvm.obj/bin/llc -mtriple=amdgcn -mcpu=gfx908 -verify-machineinstrs
+ /b/1/clang-x86_64-debian-fast/llvm.obj/bin/FileCheck -enable-var-scope -check-prefixes=GFX908 /b/1/clang-x86_64-debian-fast/llvm.src/llvm/test/CodeGen/AMDGPU/mfma-loop.ll
/b/1/clang-x86_64-debian-fast/llvm.src/llvm/test/CodeGen/AMDGPU/mfma-loop.ll:509:16: error: GFX908-NEXT: is not on the line after the previous match
; GFX908-NEXT: s_mov_b32 s0, 16
               ^
<stdin>:338:2: note: 'next' match was here
 s_mov_b32 s0, 16
 ^
<stdin>:336:27: note: previous match ended here
 v_accvgpr_write_b32 a0, 0
                          ^
<stdin>:337:1: note: non-matching line after previous match is here
 v_mov_b32_e32 v0, 1.0
^

Input file: <stdin>
Check file: /b/1/clang-x86_64-debian-fast/llvm.src/llvm/test/CodeGen/AMDGPU/mfma-loop.ll

-dump-input=help explains the following input dump.

Input was:
<<<<<<
          .
          .
          .
        333:  v_accvgpr_write_b32 a4, 0 
        334:  v_accvgpr_write_b32 a3, 0 
        335:  v_accvgpr_write_b32 a2, 0 
        336:  v_accvgpr_write_b32 a0, 0 
        337:  v_mov_b32_e32 v0, 1.0 
        338:  s_mov_b32 s0, 16 
next:509      !~~~~~~~~~~~~~~~  error: match on wrong line
        339:  v_mov_b32_e32 v1, 2.0 
        340: .LBB2_1: ; %for.cond.preheader 
        341:  ; =>This Inner Loop Header: Depth=1 
        342:  s_nop 1 
        343:  v_mfma_f32_32x32x1f32 a[0:31], v0, v1, a[0:31] 
          .
          .
          .
>>>>>>

--

...

```

</details>

https://github.com/llvm/llvm-project/pull/125885