[llvm] [AArch64][SVE] Use FeatureUseFixedOverScalableIfEqualCost for A510 and A520 (PR #132246)

[LLVM Continuous Integration via llvm-commits]

llvm-ci wrote:

LLVM Buildbot has detected a new failure on builder `premerge-monolithic-linux` running on `premerge-linux-1` while building `llvm` at step 7 "test-build-unified-tree-check-all".

Full details are available at: https://lab.llvm.org/buildbot/#/builders/153/builds/27853

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

```
Step 7 (test-build-unified-tree-check-all) failure: test (failure)
******************** TEST 'LLVM :: Transforms/LoopVectorize/AArch64/sve-fixed-width-inorder-core.ll' FAILED ********************
Exit Code: 1

Command Output (stderr):
--
/build/buildbot/premerge-monolithic-linux/build/bin/opt < /build/buildbot/premerge-monolithic-linux/llvm-project/llvm/test/Transforms/LoopVectorize/AArch64/sve-fixed-width-inorder-core.ll -mtriple=aarch64-none-elf -mcpu=cortex-a510 -mattr=+sve -passes=loop-vectorize -S | /build/buildbot/premerge-monolithic-linux/build/bin/FileCheck /build/buildbot/premerge-monolithic-linux/llvm-project/llvm/test/Transforms/LoopVectorize/AArch64/sve-fixed-width-inorder-core.ll --check-prefix=CHECK-CA510 # RUN: at line 2
+ /build/buildbot/premerge-monolithic-linux/build/bin/opt -mtriple=aarch64-none-elf -mcpu=cortex-a510 -mattr=+sve -passes=loop-vectorize -S
+ /build/buildbot/premerge-monolithic-linux/build/bin/FileCheck /build/buildbot/premerge-monolithic-linux/llvm-project/llvm/test/Transforms/LoopVectorize/AArch64/sve-fixed-width-inorder-core.ll --check-prefix=CHECK-CA510
/build/buildbot/premerge-monolithic-linux/llvm-project/llvm/test/Transforms/LoopVectorize/AArch64/sve-fixed-width-inorder-core.ll:30:21: error: CHECK-CA510-NEXT: expected string not found in input
; CHECK-CA510-NEXT: [[TMP2:%.*]] = add i64 [[INDEX]], 0
                    ^
<stdin>:32:67: note: scanning from here
 %index = phi i64 [ 0, %vector.ph ], [ %index.next, %vector.body ]
                                                                  ^
<stdin>:32:67: note: with "INDEX" equal to "%index"
 %index = phi i64 [ 0, %vector.ph ], [ %index.next, %vector.body ]
                                                                  ^
<stdin>:63:8: note: possible intended match here
 %indvars.iv = phi i64 [ %indvars.iv.next, %for.body ], [ %bc.resume.val, %scalar.ph ]
       ^

Input file: <stdin>
Check file: /build/buildbot/premerge-monolithic-linux/llvm-project/llvm/test/Transforms/LoopVectorize/AArch64/sve-fixed-width-inorder-core.ll

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

Input was:
<<<<<<
           .
           .
           .
          27:  %n.mod.vf = urem i64 %n, 8 
          28:  %n.vec = sub i64 %n, %n.mod.vf 
          29:  br label %vector.body 
          30:  
          31: vector.body: ; preds = %vector.body, %vector.ph 
          32:  %index = phi i64 [ 0, %vector.ph ], [ %index.next, %vector.body ] 
next:30'0                                                                       X error: no match found
next:30'1                                                                         with "INDEX" equal to "%index"
          33:  %2 = getelementptr inbounds nuw float, ptr %a, i64 %index 
next:30'0     ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
          34:  %3 = getelementptr inbounds nuw float, ptr %2, i32 0 
next:30'0     ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
          35:  %4 = getelementptr inbounds nuw float, ptr %2, i32 4 
next:30'0     ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
          36:  %wide.load = load <4 x float>, ptr %3, align 4 
next:30'0     ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
          37:  %wide.load5 = load <4 x float>, ptr %4, align 4 
next:30'0     ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
           .
...

```

</details>

https://github.com/llvm/llvm-project/pull/132246