[llvm] [AMDGPU][True16][CodeGen] true16 codegen pattern for fma (PR #122950)

[LLVM Continuous Integration via llvm-commits]

llvm-ci wrote:

LLVM Buildbot has detected a new failure on builder `ml-opt-rel-x86-64` running on `ml-opt-rel-x86-64-b2` while building `llvm` at step 6 "test-build-unified-tree-check-all".

Full details are available at: https://lab.llvm.org/buildbot/#/builders/185/builds/13160

<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/preserve-hi16.ll' FAILED ********************
Exit Code: 1

Command Output (stderr):
--
RUN: at line 2: /b/ml-opt-rel-x86-64-b1/build/bin/llc -mtriple=amdgcn -mcpu=gfx803 -verify-machineinstrs < /b/ml-opt-rel-x86-64-b1/llvm-project/llvm/test/CodeGen/AMDGPU/preserve-hi16.ll | /b/ml-opt-rel-x86-64-b1/build/bin/FileCheck --check-prefixes=GFX8 /b/ml-opt-rel-x86-64-b1/llvm-project/llvm/test/CodeGen/AMDGPU/preserve-hi16.ll
+ /b/ml-opt-rel-x86-64-b1/build/bin/llc -mtriple=amdgcn -mcpu=gfx803 -verify-machineinstrs
+ /b/ml-opt-rel-x86-64-b1/build/bin/FileCheck --check-prefixes=GFX8 /b/ml-opt-rel-x86-64-b1/llvm-project/llvm/test/CodeGen/AMDGPU/preserve-hi16.ll
RUN: at line 3: /b/ml-opt-rel-x86-64-b1/build/bin/llc -mtriple=amdgcn -mcpu=gfx900 -verify-machineinstrs < /b/ml-opt-rel-x86-64-b1/llvm-project/llvm/test/CodeGen/AMDGPU/preserve-hi16.ll | /b/ml-opt-rel-x86-64-b1/build/bin/FileCheck --check-prefixes=GFX9ALL,GFX900 /b/ml-opt-rel-x86-64-b1/llvm-project/llvm/test/CodeGen/AMDGPU/preserve-hi16.ll
+ /b/ml-opt-rel-x86-64-b1/build/bin/FileCheck --check-prefixes=GFX9ALL,GFX900 /b/ml-opt-rel-x86-64-b1/llvm-project/llvm/test/CodeGen/AMDGPU/preserve-hi16.ll
+ /b/ml-opt-rel-x86-64-b1/build/bin/llc -mtriple=amdgcn -mcpu=gfx900 -verify-machineinstrs
RUN: at line 4: /b/ml-opt-rel-x86-64-b1/build/bin/llc -mtriple=amdgcn -mcpu=gfx906 -verify-machineinstrs < /b/ml-opt-rel-x86-64-b1/llvm-project/llvm/test/CodeGen/AMDGPU/preserve-hi16.ll | /b/ml-opt-rel-x86-64-b1/build/bin/FileCheck --check-prefixes=GFX9ALL,GFX906 /b/ml-opt-rel-x86-64-b1/llvm-project/llvm/test/CodeGen/AMDGPU/preserve-hi16.ll
+ /b/ml-opt-rel-x86-64-b1/build/bin/FileCheck --check-prefixes=GFX9ALL,GFX906 /b/ml-opt-rel-x86-64-b1/llvm-project/llvm/test/CodeGen/AMDGPU/preserve-hi16.ll
+ /b/ml-opt-rel-x86-64-b1/build/bin/llc -mtriple=amdgcn -mcpu=gfx906 -verify-machineinstrs
RUN: at line 5: /b/ml-opt-rel-x86-64-b1/build/bin/llc -mtriple=amdgcn -mcpu=gfx1010 -verify-machineinstrs < /b/ml-opt-rel-x86-64-b1/llvm-project/llvm/test/CodeGen/AMDGPU/preserve-hi16.ll | /b/ml-opt-rel-x86-64-b1/build/bin/FileCheck -check-prefixes=GFX10 /b/ml-opt-rel-x86-64-b1/llvm-project/llvm/test/CodeGen/AMDGPU/preserve-hi16.ll
+ /b/ml-opt-rel-x86-64-b1/build/bin/FileCheck -check-prefixes=GFX10 /b/ml-opt-rel-x86-64-b1/llvm-project/llvm/test/CodeGen/AMDGPU/preserve-hi16.ll
+ /b/ml-opt-rel-x86-64-b1/build/bin/llc -mtriple=amdgcn -mcpu=gfx1010 -verify-machineinstrs
RUN: at line 6: /b/ml-opt-rel-x86-64-b1/build/bin/llc -mtriple=amdgcn -mcpu=gfx1100 -amdgpu-enable-delay-alu=0 -mattr=+real-true16 -verify-machineinstrs < /b/ml-opt-rel-x86-64-b1/llvm-project/llvm/test/CodeGen/AMDGPU/preserve-hi16.ll | /b/ml-opt-rel-x86-64-b1/build/bin/FileCheck -check-prefixes=GFX11,GFX11-TRUE16 /b/ml-opt-rel-x86-64-b1/llvm-project/llvm/test/CodeGen/AMDGPU/preserve-hi16.ll
+ /b/ml-opt-rel-x86-64-b1/build/bin/FileCheck -check-prefixes=GFX11,GFX11-TRUE16 /b/ml-opt-rel-x86-64-b1/llvm-project/llvm/test/CodeGen/AMDGPU/preserve-hi16.ll
+ /b/ml-opt-rel-x86-64-b1/build/bin/llc -mtriple=amdgcn -mcpu=gfx1100 -amdgpu-enable-delay-alu=0 -mattr=+real-true16 -verify-machineinstrs
/b/ml-opt-rel-x86-64-b1/llvm-project/llvm/test/CodeGen/AMDGPU/preserve-hi16.ll:817:22: error: GFX11-TRUE16-NEXT: expected string not found in input
; GFX11-TRUE16-NEXT: v_fma_f16 v0.l, v0.l, v0.h, v1.l
                     ^
<stdin>:865:26: note: scanning from here
 v_mov_b16_e32 v1.l, v2.l
                         ^
<stdin>:866:2: note: possible intended match here
 v_fmac_f16_e32 v1.l, v0.l, v0.h
 ^

Input file: <stdin>
Check file: /b/ml-opt-rel-x86-64-b1/llvm-project/llvm/test/CodeGen/AMDGPU/preserve-hi16.ll

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

Input was:
<<<<<<
            .
            .
            .
          860:  .type zext_fma_f16, at function 
          861: zext_fma_f16: ; @zext_fma_f16 
          862: ; %bb.0: 
          863:  s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) 
          864:  v_mov_b16_e32 v0.h, v1.l 
          865:  v_mov_b16_e32 v1.l, v2.l 
next:817'0                              X error: no match found
          866:  v_fmac_f16_e32 v1.l, v0.l, v0.h 
next:817'0     ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
next:817'1      ?                                possible intended match
...

```

</details>

https://github.com/llvm/llvm-project/pull/122950