[PATCH] D133955: [AArch64][CostModel] Add costs for fixed operations when using fixed vectors over SVE

[Sander de Smalen via Phabricator via llvm-commits]

sdesmalen added a comment.

Just left two more nits, happy otherwise.



================
Comment at: llvm/test/Analysis/CostModel/AArch64/cast.ll:5
+; RUN: opt -passes="print<cost-model>" 2>&1 -disable-output -mtriple=aarch64-none-linux-gnueabi -mattr=+sve -aarch64-sve-vector-bits-min=256 %s | FileCheck --check-prefixes=SVE,FIXED-MIN-256 %s
+; RUN: opt -passes="print<cost-model>" 2>&1 -disable-output -mtriple=aarch64-none-linux-gnueabi -mattr=+sve -aarch64-sve-vector-bits-min=2048 %s | FileCheck --check-prefixes=SVE,FIXED-MAX %s
+; RUN: opt -passes="print<cost-model>" 2>&1 -disable-output -mtriple=aarch64-none-linux-gnueabi -mattr=+sve -force-streaming-compatible-sve %s | FileCheck --check-prefixes=SVE,SVE128-NO-NEON %s
----------------
nit: can you change this to `MIN-2048` ?


================
Comment at: llvm/test/Analysis/CostModel/AArch64/cast.ll:6
+; RUN: opt -passes="print<cost-model>" 2>&1 -disable-output -mtriple=aarch64-none-linux-gnueabi -mattr=+sve -aarch64-sve-vector-bits-min=2048 %s | FileCheck --check-prefixes=SVE,FIXED-MAX %s
+; RUN: opt -passes="print<cost-model>" 2>&1 -disable-output -mtriple=aarch64-none-linux-gnueabi -mattr=+sve -force-streaming-compatible-sve %s | FileCheck --check-prefixes=SVE,SVE128-NO-NEON %s
 
----------------
nit: can you move this RUN line to beneath the RUN line on line 2, and then regenerate the tests, so that we can more easily compare the numbers between "non-streaming-compatible" and "streaming-compatible" when the minimum SVE vector length is the same?


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D133955/new/

https://reviews.llvm.org/D133955