[llvm] [AArch64] Account for fixed-SVE high-lane insertelement/extractelement costs (PR #219244)

[Utpal Bora via llvm-commits]

================
@@ -14,6 +14,7 @@
 ; RUN: opt < %s -passes="print<cost-model>" -cost-kind=all 2>&1 -disable-output -mcpu=apple-latest | FileCheck --check-prefixes=CHECK,FAST-LD1-LAT4 %s
 ; RUN: opt < %s -passes="print<cost-model>" -cost-kind=all 2>&1 -disable-output -mattr=+fast-ld1-single | FileCheck --check-prefixes=CHECK,FAST-LD1-LAT4 %s
 ; RUN: opt < %s -passes="print<cost-model>" -cost-kind=all 2>&1 -disable-output -mcpu=cortex-a510 -mattr=+fast-ld1-single | FileCheck --check-prefixes=CHECK,FAST-LD1-LAT2 %s
+; RUN: opt < %s -passes="print<cost-model>" -cost-kind=all 2>&1 -disable-output -mattr=+sve | FileCheck --check-prefix=VLS-HIGH-LANE %s
----------------
utpalbora wrote:

`update_analyze_test_checks.py` generates a lot more checks without the new `RUN` line, I suspect the test case had been updated manually before. I had to add it manually as well. I have moved the new checks to a new `sve-vls-` file as suggested by Gaetan.

https://github.com/llvm/llvm-project/pull/219244