derek.holloway
7502bf64a4
CodeQL / Analyze (actions) (push) Canceled after 0s
CodeQL / Analyze (python) (push) Canceled after 0s
pre-commit / pre-commit (push) Canceled after 0s
Multi-Arch CI / setup (push) Canceled after 0s
Publish build_manylinux_x86_64 images / publish_build_manylinux_x86_64 (push) Canceled after 0s
Setup (Multi-Arch) / setup (push) Canceled after 0s
Multi-Arch CI / Linux::${{ fromJSON(needs.setup.outputs.linux_build_config || '{}').build_variant_label || 'skip' }} (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish no_rocm_image_ubi10 image / publish_no_rocm_image_ubi10 (push) Canceled after 0s
Multi-Arch CI / Windows::${{ fromJSON(needs.setup.outputs.windows_build_config || '{}').build_variant_label || 'skip' }} (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Multi-Arch CI / Manifest Diff (push) Canceled after 0s
TheRock Manifest Diff Report / Generate Manifest Diff Report (push) Canceled after 0s
Multi-Arch CI / CI Summary (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04 (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04_ocl_rt images / publish_no_rocm_image_ubuntu24_04_ocl_rt (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_rocgdb (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04_rocgdb images / publish_no_rocm_image_ubuntu24_04_rocgdb (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_ocl_rt (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Unit Tests / Unit Tests :: ubuntu-24.04 (push) Canceled after 0s
Unit Tests / Unit Tests :: windows-2022 (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_media (push) Canceled after 0s
Unit Tests / Unit Tests Summary (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_openmpi (push) Canceled after 0s
221 lines
24 KiB
LLVM
221 lines
24 KiB
LLVM
; NOTE: Assertions have been autogenerated by utils/update_analyze_test_checks.py
|
|
; RUN: opt < %s -mtriple=armv8a-linux-gnueabihf -mattr=+fp64 -passes="print<cost-model>" -cost-kind=all 2>&1 -disable-output | FileCheck %s --check-prefix=CHECK-V8
|
|
; RUN: opt < %s -mtriple=thumbv8.1m.main-none-eabi -mattr=+mve.fp -passes="print<cost-model>" -cost-kind=all 2>&1 -disable-output | FileCheck %s --check-prefix=CHECK-MVEFP
|
|
; RUN: opt < %s -mtriple=thumbv8.1m.main-none-eabi -mattr=+mve -passes="print<cost-model>" -cost-kind=all 2>&1 -disable-output | FileCheck %s --check-prefix=CHECK-MVEI
|
|
|
|
target datalayout = "e-m:e-i8:8:32-i16:16:32-i64:64-i128:128-n32:64-S128"
|
|
|
|
define void @fmin_strict() {
|
|
; CHECK-V8-LABEL: 'fmin_strict'
|
|
; CHECK-V8-NEXT: Cost Model: Found costs of 7 for: %fmin_v2f16 = call half @llvm.vector.reduce.fmin.v2f16(<2 x half> undef)
|
|
; CHECK-V8-NEXT: Cost Model: Found costs of 19 for: %fmin_v4f16 = call half @llvm.vector.reduce.fmin.v4f16(<4 x half> undef)
|
|
; CHECK-V8-NEXT: Cost Model: Found costs of 43 for: %fmin_v8f16 = call half @llvm.vector.reduce.fmin.v8f16(<8 x half> undef)
|
|
; CHECK-V8-NEXT: Cost Model: Found costs of 91 for: %fmin_v16f16 = call half @llvm.vector.reduce.fmin.v16f16(<16 x half> undef)
|
|
; CHECK-V8-NEXT: Cost Model: Found costs of 1 for: %fmin_v2f32 = call float @llvm.vector.reduce.fmin.v2f32(<2 x float> undef)
|
|
; CHECK-V8-NEXT: Cost Model: Found costs of 2 for: %fmin_v4f32 = call float @llvm.vector.reduce.fmin.v4f32(<4 x float> undef)
|
|
; CHECK-V8-NEXT: Cost Model: Found costs of 3 for: %fmin_v8f32 = call float @llvm.vector.reduce.fmin.v8f32(<8 x float> undef)
|
|
; CHECK-V8-NEXT: Cost Model: Found costs of 1 for: %fmin_v2f64 = call double @llvm.vector.reduce.fmin.v2f64(<2 x double> undef)
|
|
; CHECK-V8-NEXT: Cost Model: Found costs of 9 for: %fmin_v4f64 = call double @llvm.vector.reduce.fmin.v4f64(<4 x double> undef)
|
|
; CHECK-V8-NEXT: Cost Model: Found costs of RThru:94 CodeSize:67 Lat:94 SizeLat:94 for: %fmin_v4f128 = call fp128 @llvm.vector.reduce.fmin.v4f128(<4 x fp128> undef)
|
|
; CHECK-V8-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void
|
|
;
|
|
; CHECK-MVEFP-LABEL: 'fmin_strict'
|
|
; CHECK-MVEFP-NEXT: Cost Model: Found costs of 2 for: %fmin_v2f16 = call half @llvm.vector.reduce.fmin.v2f16(<2 x half> undef)
|
|
; CHECK-MVEFP-NEXT: Cost Model: Found costs of 5 for: %fmin_v4f16 = call half @llvm.vector.reduce.fmin.v4f16(<4 x half> undef)
|
|
; CHECK-MVEFP-NEXT: Cost Model: Found costs of RThru:7 CodeSize:5 Lat:7 SizeLat:7 for: %fmin_v8f16 = call half @llvm.vector.reduce.fmin.v8f16(<8 x half> undef)
|
|
; CHECK-MVEFP-NEXT: Cost Model: Found costs of RThru:9 CodeSize:6 Lat:9 SizeLat:9 for: %fmin_v16f16 = call half @llvm.vector.reduce.fmin.v16f16(<16 x half> undef)
|
|
; CHECK-MVEFP-NEXT: Cost Model: Found costs of 1 for: %fmin_v2f32 = call float @llvm.vector.reduce.fmin.v2f32(<2 x float> undef)
|
|
; CHECK-MVEFP-NEXT: Cost Model: Found costs of 3 for: %fmin_v4f32 = call float @llvm.vector.reduce.fmin.v4f32(<4 x float> undef)
|
|
; CHECK-MVEFP-NEXT: Cost Model: Found costs of RThru:5 CodeSize:4 Lat:5 SizeLat:5 for: %fmin_v8f32 = call float @llvm.vector.reduce.fmin.v8f32(<8 x float> undef)
|
|
; CHECK-MVEFP-NEXT: Cost Model: Found costs of RThru:35 CodeSize:13 Lat:35 SizeLat:35 for: %fmin_v2f64 = call double @llvm.vector.reduce.fmin.v2f64(<2 x double> undef)
|
|
; CHECK-MVEFP-NEXT: Cost Model: Found costs of RThru:69 CodeSize:25 Lat:69 SizeLat:69 for: %fmin_v4f64 = call double @llvm.vector.reduce.fmin.v4f64(<4 x double> undef)
|
|
; CHECK-MVEFP-NEXT: Cost Model: Found costs of RThru:118 CodeSize:67 Lat:118 SizeLat:118 for: %fmin_v4f128 = call fp128 @llvm.vector.reduce.fmin.v4f128(<4 x fp128> undef)
|
|
; CHECK-MVEFP-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void
|
|
;
|
|
; CHECK-MVEI-LABEL: 'fmin_strict'
|
|
; CHECK-MVEI-NEXT: Cost Model: Found costs of RThru:35 CodeSize:13 Lat:35 SizeLat:35 for: %fmin_v2f16 = call half @llvm.vector.reduce.fmin.v2f16(<2 x half> undef)
|
|
; CHECK-MVEI-NEXT: Cost Model: Found costs of RThru:137 CodeSize:49 Lat:137 SizeLat:137 for: %fmin_v4f16 = call half @llvm.vector.reduce.fmin.v4f16(<4 x half> undef)
|
|
; CHECK-MVEI-NEXT: Cost Model: Found costs of RThru:409 CodeSize:145 Lat:409 SizeLat:409 for: %fmin_v8f16 = call half @llvm.vector.reduce.fmin.v8f16(<8 x half> undef)
|
|
; CHECK-MVEI-NEXT: Cost Model: Found costs of RThru:545 CodeSize:193 Lat:545 SizeLat:545 for: %fmin_v16f16 = call half @llvm.vector.reduce.fmin.v16f16(<16 x half> undef)
|
|
; CHECK-MVEI-NEXT: Cost Model: Found costs of RThru:35 CodeSize:13 Lat:35 SizeLat:35 for: %fmin_v2f32 = call float @llvm.vector.reduce.fmin.v2f32(<2 x float> undef)
|
|
; CHECK-MVEI-NEXT: Cost Model: Found costs of RThru:137 CodeSize:49 Lat:137 SizeLat:137 for: %fmin_v4f32 = call float @llvm.vector.reduce.fmin.v4f32(<4 x float> undef)
|
|
; CHECK-MVEI-NEXT: Cost Model: Found costs of RThru:205 CodeSize:73 Lat:205 SizeLat:205 for: %fmin_v8f32 = call float @llvm.vector.reduce.fmin.v8f32(<8 x float> undef)
|
|
; CHECK-MVEI-NEXT: Cost Model: Found costs of RThru:35 CodeSize:13 Lat:35 SizeLat:35 for: %fmin_v2f64 = call double @llvm.vector.reduce.fmin.v2f64(<2 x double> undef)
|
|
; CHECK-MVEI-NEXT: Cost Model: Found costs of RThru:69 CodeSize:25 Lat:69 SizeLat:69 for: %fmin_v4f64 = call double @llvm.vector.reduce.fmin.v4f64(<4 x double> undef)
|
|
; CHECK-MVEI-NEXT: Cost Model: Found costs of RThru:118 CodeSize:67 Lat:118 SizeLat:118 for: %fmin_v4f128 = call fp128 @llvm.vector.reduce.fmin.v4f128(<4 x fp128> undef)
|
|
; CHECK-MVEI-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void
|
|
;
|
|
%fmin_v2f16 = call half @llvm.vector.reduce.fmin.v2f16(<2 x half> undef)
|
|
%fmin_v4f16 = call half @llvm.vector.reduce.fmin.v4f16(<4 x half> undef)
|
|
%fmin_v8f16 = call half @llvm.vector.reduce.fmin.v8f16(<8 x half> undef)
|
|
%fmin_v16f16 = call half @llvm.vector.reduce.fmin.v16f16(<16 x half> undef)
|
|
%fmin_v2f32 = call float @llvm.vector.reduce.fmin.v2f32(<2 x float> undef)
|
|
%fmin_v4f32 = call float @llvm.vector.reduce.fmin.v4f32(<4 x float> undef)
|
|
%fmin_v8f32 = call float @llvm.vector.reduce.fmin.v8f32(<8 x float> undef)
|
|
%fmin_v2f64 = call double @llvm.vector.reduce.fmin.v2f64(<2 x double> undef)
|
|
%fmin_v4f64 = call double @llvm.vector.reduce.fmin.v4f64(<4 x double> undef)
|
|
%fmin_v4f128 = call fp128 @llvm.vector.reduce.fmin.v4f128(<4 x fp128> undef)
|
|
ret void
|
|
}
|
|
|
|
|
|
define void @fmin_unordered() {
|
|
; CHECK-V8-LABEL: 'fmin_unordered'
|
|
; CHECK-V8-NEXT: Cost Model: Found costs of 7 for: %fmin_v2f16 = call reassoc half @llvm.vector.reduce.fmin.v2f16(<2 x half> undef)
|
|
; CHECK-V8-NEXT: Cost Model: Found costs of 19 for: %fmin_v4f16 = call reassoc half @llvm.vector.reduce.fmin.v4f16(<4 x half> undef)
|
|
; CHECK-V8-NEXT: Cost Model: Found costs of 43 for: %fmin_v8f16 = call reassoc half @llvm.vector.reduce.fmin.v8f16(<8 x half> undef)
|
|
; CHECK-V8-NEXT: Cost Model: Found costs of 91 for: %fmin_v16f16 = call reassoc half @llvm.vector.reduce.fmin.v16f16(<16 x half> undef)
|
|
; CHECK-V8-NEXT: Cost Model: Found costs of 1 for: %fmin_v2f32 = call reassoc float @llvm.vector.reduce.fmin.v2f32(<2 x float> undef)
|
|
; CHECK-V8-NEXT: Cost Model: Found costs of 2 for: %fmin_v4f32 = call reassoc float @llvm.vector.reduce.fmin.v4f32(<4 x float> undef)
|
|
; CHECK-V8-NEXT: Cost Model: Found costs of 3 for: %fmin_v8f32 = call reassoc float @llvm.vector.reduce.fmin.v8f32(<8 x float> undef)
|
|
; CHECK-V8-NEXT: Cost Model: Found costs of 1 for: %fmin_v2f64 = call reassoc double @llvm.vector.reduce.fmin.v2f64(<2 x double> undef)
|
|
; CHECK-V8-NEXT: Cost Model: Found costs of 9 for: %fmin_v4f64 = call reassoc double @llvm.vector.reduce.fmin.v4f64(<4 x double> undef)
|
|
; CHECK-V8-NEXT: Cost Model: Found costs of RThru:94 CodeSize:67 Lat:94 SizeLat:94 for: %fmin_v4f128 = call reassoc fp128 @llvm.vector.reduce.fmin.v4f128(<4 x fp128> undef)
|
|
; CHECK-V8-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void
|
|
;
|
|
; CHECK-MVEFP-LABEL: 'fmin_unordered'
|
|
; CHECK-MVEFP-NEXT: Cost Model: Found costs of 2 for: %fmin_v2f16 = call reassoc half @llvm.vector.reduce.fmin.v2f16(<2 x half> undef)
|
|
; CHECK-MVEFP-NEXT: Cost Model: Found costs of 5 for: %fmin_v4f16 = call reassoc half @llvm.vector.reduce.fmin.v4f16(<4 x half> undef)
|
|
; CHECK-MVEFP-NEXT: Cost Model: Found costs of RThru:7 CodeSize:5 Lat:7 SizeLat:7 for: %fmin_v8f16 = call reassoc half @llvm.vector.reduce.fmin.v8f16(<8 x half> undef)
|
|
; CHECK-MVEFP-NEXT: Cost Model: Found costs of RThru:9 CodeSize:6 Lat:9 SizeLat:9 for: %fmin_v16f16 = call reassoc half @llvm.vector.reduce.fmin.v16f16(<16 x half> undef)
|
|
; CHECK-MVEFP-NEXT: Cost Model: Found costs of 1 for: %fmin_v2f32 = call reassoc float @llvm.vector.reduce.fmin.v2f32(<2 x float> undef)
|
|
; CHECK-MVEFP-NEXT: Cost Model: Found costs of 3 for: %fmin_v4f32 = call reassoc float @llvm.vector.reduce.fmin.v4f32(<4 x float> undef)
|
|
; CHECK-MVEFP-NEXT: Cost Model: Found costs of RThru:5 CodeSize:4 Lat:5 SizeLat:5 for: %fmin_v8f32 = call reassoc float @llvm.vector.reduce.fmin.v8f32(<8 x float> undef)
|
|
; CHECK-MVEFP-NEXT: Cost Model: Found costs of RThru:35 CodeSize:13 Lat:35 SizeLat:35 for: %fmin_v2f64 = call reassoc double @llvm.vector.reduce.fmin.v2f64(<2 x double> undef)
|
|
; CHECK-MVEFP-NEXT: Cost Model: Found costs of RThru:69 CodeSize:25 Lat:69 SizeLat:69 for: %fmin_v4f64 = call reassoc double @llvm.vector.reduce.fmin.v4f64(<4 x double> undef)
|
|
; CHECK-MVEFP-NEXT: Cost Model: Found costs of RThru:118 CodeSize:67 Lat:118 SizeLat:118 for: %fmin_v4f128 = call reassoc fp128 @llvm.vector.reduce.fmin.v4f128(<4 x fp128> undef)
|
|
; CHECK-MVEFP-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void
|
|
;
|
|
; CHECK-MVEI-LABEL: 'fmin_unordered'
|
|
; CHECK-MVEI-NEXT: Cost Model: Found costs of RThru:35 CodeSize:13 Lat:35 SizeLat:35 for: %fmin_v2f16 = call reassoc half @llvm.vector.reduce.fmin.v2f16(<2 x half> undef)
|
|
; CHECK-MVEI-NEXT: Cost Model: Found costs of RThru:137 CodeSize:49 Lat:137 SizeLat:137 for: %fmin_v4f16 = call reassoc half @llvm.vector.reduce.fmin.v4f16(<4 x half> undef)
|
|
; CHECK-MVEI-NEXT: Cost Model: Found costs of RThru:409 CodeSize:145 Lat:409 SizeLat:409 for: %fmin_v8f16 = call reassoc half @llvm.vector.reduce.fmin.v8f16(<8 x half> undef)
|
|
; CHECK-MVEI-NEXT: Cost Model: Found costs of RThru:545 CodeSize:193 Lat:545 SizeLat:545 for: %fmin_v16f16 = call reassoc half @llvm.vector.reduce.fmin.v16f16(<16 x half> undef)
|
|
; CHECK-MVEI-NEXT: Cost Model: Found costs of RThru:35 CodeSize:13 Lat:35 SizeLat:35 for: %fmin_v2f32 = call reassoc float @llvm.vector.reduce.fmin.v2f32(<2 x float> undef)
|
|
; CHECK-MVEI-NEXT: Cost Model: Found costs of RThru:137 CodeSize:49 Lat:137 SizeLat:137 for: %fmin_v4f32 = call reassoc float @llvm.vector.reduce.fmin.v4f32(<4 x float> undef)
|
|
; CHECK-MVEI-NEXT: Cost Model: Found costs of RThru:205 CodeSize:73 Lat:205 SizeLat:205 for: %fmin_v8f32 = call reassoc float @llvm.vector.reduce.fmin.v8f32(<8 x float> undef)
|
|
; CHECK-MVEI-NEXT: Cost Model: Found costs of RThru:35 CodeSize:13 Lat:35 SizeLat:35 for: %fmin_v2f64 = call reassoc double @llvm.vector.reduce.fmin.v2f64(<2 x double> undef)
|
|
; CHECK-MVEI-NEXT: Cost Model: Found costs of RThru:69 CodeSize:25 Lat:69 SizeLat:69 for: %fmin_v4f64 = call reassoc double @llvm.vector.reduce.fmin.v4f64(<4 x double> undef)
|
|
; CHECK-MVEI-NEXT: Cost Model: Found costs of RThru:118 CodeSize:67 Lat:118 SizeLat:118 for: %fmin_v4f128 = call reassoc fp128 @llvm.vector.reduce.fmin.v4f128(<4 x fp128> undef)
|
|
; CHECK-MVEI-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void
|
|
;
|
|
%fmin_v2f16 = call reassoc half @llvm.vector.reduce.fmin.v2f16(<2 x half> undef)
|
|
%fmin_v4f16 = call reassoc half @llvm.vector.reduce.fmin.v4f16(<4 x half> undef)
|
|
%fmin_v8f16 = call reassoc half @llvm.vector.reduce.fmin.v8f16(<8 x half> undef)
|
|
%fmin_v16f16 = call reassoc half @llvm.vector.reduce.fmin.v16f16(<16 x half> undef)
|
|
%fmin_v2f32 = call reassoc float @llvm.vector.reduce.fmin.v2f32(<2 x float> undef)
|
|
%fmin_v4f32 = call reassoc float @llvm.vector.reduce.fmin.v4f32(<4 x float> undef)
|
|
%fmin_v8f32 = call reassoc float @llvm.vector.reduce.fmin.v8f32(<8 x float> undef)
|
|
%fmin_v2f64 = call reassoc double @llvm.vector.reduce.fmin.v2f64(<2 x double> undef)
|
|
%fmin_v4f64 = call reassoc double @llvm.vector.reduce.fmin.v4f64(<4 x double> undef)
|
|
%fmin_v4f128 = call reassoc fp128 @llvm.vector.reduce.fmin.v4f128(<4 x fp128> undef)
|
|
ret void
|
|
}
|
|
|
|
define void @fmax_strict() {
|
|
; CHECK-V8-LABEL: 'fmax_strict'
|
|
; CHECK-V8-NEXT: Cost Model: Found costs of 7 for: %fmax_v2f16 = call half @llvm.vector.reduce.fmax.v2f16(<2 x half> undef)
|
|
; CHECK-V8-NEXT: Cost Model: Found costs of 19 for: %fmax_v4f16 = call half @llvm.vector.reduce.fmax.v4f16(<4 x half> undef)
|
|
; CHECK-V8-NEXT: Cost Model: Found costs of 43 for: %fmax_v8f16 = call half @llvm.vector.reduce.fmax.v8f16(<8 x half> undef)
|
|
; CHECK-V8-NEXT: Cost Model: Found costs of 91 for: %fmax_v16f16 = call half @llvm.vector.reduce.fmax.v16f16(<16 x half> undef)
|
|
; CHECK-V8-NEXT: Cost Model: Found costs of 1 for: %fmax_v2f32 = call float @llvm.vector.reduce.fmax.v2f32(<2 x float> undef)
|
|
; CHECK-V8-NEXT: Cost Model: Found costs of 2 for: %fmax_v4f32 = call float @llvm.vector.reduce.fmax.v4f32(<4 x float> undef)
|
|
; CHECK-V8-NEXT: Cost Model: Found costs of 3 for: %fmax_v8f32 = call float @llvm.vector.reduce.fmax.v8f32(<8 x float> undef)
|
|
; CHECK-V8-NEXT: Cost Model: Found costs of 1 for: %fmax_v2f64 = call double @llvm.vector.reduce.fmax.v2f64(<2 x double> undef)
|
|
; CHECK-V8-NEXT: Cost Model: Found costs of 9 for: %fmax_v4f64 = call double @llvm.vector.reduce.fmax.v4f64(<4 x double> undef)
|
|
; CHECK-V8-NEXT: Cost Model: Found costs of RThru:94 CodeSize:67 Lat:94 SizeLat:94 for: %fmax_v4f128 = call fp128 @llvm.vector.reduce.fmax.v4f128(<4 x fp128> undef)
|
|
; CHECK-V8-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void
|
|
;
|
|
; CHECK-MVEFP-LABEL: 'fmax_strict'
|
|
; CHECK-MVEFP-NEXT: Cost Model: Found costs of 2 for: %fmax_v2f16 = call half @llvm.vector.reduce.fmax.v2f16(<2 x half> undef)
|
|
; CHECK-MVEFP-NEXT: Cost Model: Found costs of 5 for: %fmax_v4f16 = call half @llvm.vector.reduce.fmax.v4f16(<4 x half> undef)
|
|
; CHECK-MVEFP-NEXT: Cost Model: Found costs of RThru:7 CodeSize:5 Lat:7 SizeLat:7 for: %fmax_v8f16 = call half @llvm.vector.reduce.fmax.v8f16(<8 x half> undef)
|
|
; CHECK-MVEFP-NEXT: Cost Model: Found costs of RThru:9 CodeSize:6 Lat:9 SizeLat:9 for: %fmax_v16f16 = call half @llvm.vector.reduce.fmax.v16f16(<16 x half> undef)
|
|
; CHECK-MVEFP-NEXT: Cost Model: Found costs of 1 for: %fmax_v2f32 = call float @llvm.vector.reduce.fmax.v2f32(<2 x float> undef)
|
|
; CHECK-MVEFP-NEXT: Cost Model: Found costs of 3 for: %fmax_v4f32 = call float @llvm.vector.reduce.fmax.v4f32(<4 x float> undef)
|
|
; CHECK-MVEFP-NEXT: Cost Model: Found costs of RThru:5 CodeSize:4 Lat:5 SizeLat:5 for: %fmax_v8f32 = call float @llvm.vector.reduce.fmax.v8f32(<8 x float> undef)
|
|
; CHECK-MVEFP-NEXT: Cost Model: Found costs of RThru:35 CodeSize:13 Lat:35 SizeLat:35 for: %fmax_v2f64 = call double @llvm.vector.reduce.fmax.v2f64(<2 x double> undef)
|
|
; CHECK-MVEFP-NEXT: Cost Model: Found costs of RThru:69 CodeSize:25 Lat:69 SizeLat:69 for: %fmax_v4f64 = call double @llvm.vector.reduce.fmax.v4f64(<4 x double> undef)
|
|
; CHECK-MVEFP-NEXT: Cost Model: Found costs of RThru:118 CodeSize:67 Lat:118 SizeLat:118 for: %fmax_v4f128 = call fp128 @llvm.vector.reduce.fmax.v4f128(<4 x fp128> undef)
|
|
; CHECK-MVEFP-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void
|
|
;
|
|
; CHECK-MVEI-LABEL: 'fmax_strict'
|
|
; CHECK-MVEI-NEXT: Cost Model: Found costs of RThru:35 CodeSize:13 Lat:35 SizeLat:35 for: %fmax_v2f16 = call half @llvm.vector.reduce.fmax.v2f16(<2 x half> undef)
|
|
; CHECK-MVEI-NEXT: Cost Model: Found costs of RThru:137 CodeSize:49 Lat:137 SizeLat:137 for: %fmax_v4f16 = call half @llvm.vector.reduce.fmax.v4f16(<4 x half> undef)
|
|
; CHECK-MVEI-NEXT: Cost Model: Found costs of RThru:409 CodeSize:145 Lat:409 SizeLat:409 for: %fmax_v8f16 = call half @llvm.vector.reduce.fmax.v8f16(<8 x half> undef)
|
|
; CHECK-MVEI-NEXT: Cost Model: Found costs of RThru:545 CodeSize:193 Lat:545 SizeLat:545 for: %fmax_v16f16 = call half @llvm.vector.reduce.fmax.v16f16(<16 x half> undef)
|
|
; CHECK-MVEI-NEXT: Cost Model: Found costs of RThru:35 CodeSize:13 Lat:35 SizeLat:35 for: %fmax_v2f32 = call float @llvm.vector.reduce.fmax.v2f32(<2 x float> undef)
|
|
; CHECK-MVEI-NEXT: Cost Model: Found costs of RThru:137 CodeSize:49 Lat:137 SizeLat:137 for: %fmax_v4f32 = call float @llvm.vector.reduce.fmax.v4f32(<4 x float> undef)
|
|
; CHECK-MVEI-NEXT: Cost Model: Found costs of RThru:205 CodeSize:73 Lat:205 SizeLat:205 for: %fmax_v8f32 = call float @llvm.vector.reduce.fmax.v8f32(<8 x float> undef)
|
|
; CHECK-MVEI-NEXT: Cost Model: Found costs of RThru:35 CodeSize:13 Lat:35 SizeLat:35 for: %fmax_v2f64 = call double @llvm.vector.reduce.fmax.v2f64(<2 x double> undef)
|
|
; CHECK-MVEI-NEXT: Cost Model: Found costs of RThru:69 CodeSize:25 Lat:69 SizeLat:69 for: %fmax_v4f64 = call double @llvm.vector.reduce.fmax.v4f64(<4 x double> undef)
|
|
; CHECK-MVEI-NEXT: Cost Model: Found costs of RThru:118 CodeSize:67 Lat:118 SizeLat:118 for: %fmax_v4f128 = call fp128 @llvm.vector.reduce.fmax.v4f128(<4 x fp128> undef)
|
|
; CHECK-MVEI-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void
|
|
;
|
|
%fmax_v2f16 = call half @llvm.vector.reduce.fmax.v2f16(<2 x half> undef)
|
|
%fmax_v4f16 = call half @llvm.vector.reduce.fmax.v4f16(<4 x half> undef)
|
|
%fmax_v8f16 = call half @llvm.vector.reduce.fmax.v8f16(<8 x half> undef)
|
|
%fmax_v16f16 = call half @llvm.vector.reduce.fmax.v16f16(<16 x half> undef)
|
|
%fmax_v2f32 = call float @llvm.vector.reduce.fmax.v2f32(<2 x float> undef)
|
|
%fmax_v4f32 = call float @llvm.vector.reduce.fmax.v4f32(<4 x float> undef)
|
|
%fmax_v8f32 = call float @llvm.vector.reduce.fmax.v8f32(<8 x float> undef)
|
|
%fmax_v2f64 = call double @llvm.vector.reduce.fmax.v2f64(<2 x double> undef)
|
|
%fmax_v4f64 = call double @llvm.vector.reduce.fmax.v4f64(<4 x double> undef)
|
|
%fmax_v4f128 = call fp128 @llvm.vector.reduce.fmax.v4f128(<4 x fp128> undef)
|
|
ret void
|
|
}
|
|
|
|
|
|
define void @fmax_unordered() {
|
|
; CHECK-V8-LABEL: 'fmax_unordered'
|
|
; CHECK-V8-NEXT: Cost Model: Found costs of 7 for: %fmax_v2f16 = call reassoc half @llvm.vector.reduce.fmax.v2f16(<2 x half> undef)
|
|
; CHECK-V8-NEXT: Cost Model: Found costs of 19 for: %fmax_v4f16 = call reassoc half @llvm.vector.reduce.fmax.v4f16(<4 x half> undef)
|
|
; CHECK-V8-NEXT: Cost Model: Found costs of 43 for: %fmax_v8f16 = call reassoc half @llvm.vector.reduce.fmax.v8f16(<8 x half> undef)
|
|
; CHECK-V8-NEXT: Cost Model: Found costs of 91 for: %fmax_v16f16 = call reassoc half @llvm.vector.reduce.fmax.v16f16(<16 x half> undef)
|
|
; CHECK-V8-NEXT: Cost Model: Found costs of 1 for: %fmax_v2f32 = call reassoc float @llvm.vector.reduce.fmax.v2f32(<2 x float> undef)
|
|
; CHECK-V8-NEXT: Cost Model: Found costs of 2 for: %fmax_v4f32 = call reassoc float @llvm.vector.reduce.fmax.v4f32(<4 x float> undef)
|
|
; CHECK-V8-NEXT: Cost Model: Found costs of 3 for: %fmax_v8f32 = call reassoc float @llvm.vector.reduce.fmax.v8f32(<8 x float> undef)
|
|
; CHECK-V8-NEXT: Cost Model: Found costs of 1 for: %fmax_v2f64 = call reassoc double @llvm.vector.reduce.fmax.v2f64(<2 x double> undef)
|
|
; CHECK-V8-NEXT: Cost Model: Found costs of 9 for: %fmax_v4f64 = call reassoc double @llvm.vector.reduce.fmax.v4f64(<4 x double> undef)
|
|
; CHECK-V8-NEXT: Cost Model: Found costs of RThru:94 CodeSize:67 Lat:94 SizeLat:94 for: %fmax_v4f128 = call reassoc fp128 @llvm.vector.reduce.fmax.v4f128(<4 x fp128> undef)
|
|
; CHECK-V8-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void
|
|
;
|
|
; CHECK-MVEFP-LABEL: 'fmax_unordered'
|
|
; CHECK-MVEFP-NEXT: Cost Model: Found costs of 2 for: %fmax_v2f16 = call reassoc half @llvm.vector.reduce.fmax.v2f16(<2 x half> undef)
|
|
; CHECK-MVEFP-NEXT: Cost Model: Found costs of 5 for: %fmax_v4f16 = call reassoc half @llvm.vector.reduce.fmax.v4f16(<4 x half> undef)
|
|
; CHECK-MVEFP-NEXT: Cost Model: Found costs of RThru:7 CodeSize:5 Lat:7 SizeLat:7 for: %fmax_v8f16 = call reassoc half @llvm.vector.reduce.fmax.v8f16(<8 x half> undef)
|
|
; CHECK-MVEFP-NEXT: Cost Model: Found costs of RThru:9 CodeSize:6 Lat:9 SizeLat:9 for: %fmax_v16f16 = call reassoc half @llvm.vector.reduce.fmax.v16f16(<16 x half> undef)
|
|
; CHECK-MVEFP-NEXT: Cost Model: Found costs of 1 for: %fmax_v2f32 = call reassoc float @llvm.vector.reduce.fmax.v2f32(<2 x float> undef)
|
|
; CHECK-MVEFP-NEXT: Cost Model: Found costs of 3 for: %fmax_v4f32 = call reassoc float @llvm.vector.reduce.fmax.v4f32(<4 x float> undef)
|
|
; CHECK-MVEFP-NEXT: Cost Model: Found costs of RThru:5 CodeSize:4 Lat:5 SizeLat:5 for: %fmax_v8f32 = call reassoc float @llvm.vector.reduce.fmax.v8f32(<8 x float> undef)
|
|
; CHECK-MVEFP-NEXT: Cost Model: Found costs of RThru:35 CodeSize:13 Lat:35 SizeLat:35 for: %fmax_v2f64 = call reassoc double @llvm.vector.reduce.fmax.v2f64(<2 x double> undef)
|
|
; CHECK-MVEFP-NEXT: Cost Model: Found costs of RThru:69 CodeSize:25 Lat:69 SizeLat:69 for: %fmax_v4f64 = call reassoc double @llvm.vector.reduce.fmax.v4f64(<4 x double> undef)
|
|
; CHECK-MVEFP-NEXT: Cost Model: Found costs of RThru:118 CodeSize:67 Lat:118 SizeLat:118 for: %fmax_v4f128 = call reassoc fp128 @llvm.vector.reduce.fmax.v4f128(<4 x fp128> undef)
|
|
; CHECK-MVEFP-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void
|
|
;
|
|
; CHECK-MVEI-LABEL: 'fmax_unordered'
|
|
; CHECK-MVEI-NEXT: Cost Model: Found costs of RThru:35 CodeSize:13 Lat:35 SizeLat:35 for: %fmax_v2f16 = call reassoc half @llvm.vector.reduce.fmax.v2f16(<2 x half> undef)
|
|
; CHECK-MVEI-NEXT: Cost Model: Found costs of RThru:137 CodeSize:49 Lat:137 SizeLat:137 for: %fmax_v4f16 = call reassoc half @llvm.vector.reduce.fmax.v4f16(<4 x half> undef)
|
|
; CHECK-MVEI-NEXT: Cost Model: Found costs of RThru:409 CodeSize:145 Lat:409 SizeLat:409 for: %fmax_v8f16 = call reassoc half @llvm.vector.reduce.fmax.v8f16(<8 x half> undef)
|
|
; CHECK-MVEI-NEXT: Cost Model: Found costs of RThru:545 CodeSize:193 Lat:545 SizeLat:545 for: %fmax_v16f16 = call reassoc half @llvm.vector.reduce.fmax.v16f16(<16 x half> undef)
|
|
; CHECK-MVEI-NEXT: Cost Model: Found costs of RThru:35 CodeSize:13 Lat:35 SizeLat:35 for: %fmax_v2f32 = call reassoc float @llvm.vector.reduce.fmax.v2f32(<2 x float> undef)
|
|
; CHECK-MVEI-NEXT: Cost Model: Found costs of RThru:137 CodeSize:49 Lat:137 SizeLat:137 for: %fmax_v4f32 = call reassoc float @llvm.vector.reduce.fmax.v4f32(<4 x float> undef)
|
|
; CHECK-MVEI-NEXT: Cost Model: Found costs of RThru:205 CodeSize:73 Lat:205 SizeLat:205 for: %fmax_v8f32 = call reassoc float @llvm.vector.reduce.fmax.v8f32(<8 x float> undef)
|
|
; CHECK-MVEI-NEXT: Cost Model: Found costs of RThru:35 CodeSize:13 Lat:35 SizeLat:35 for: %fmax_v2f64 = call reassoc double @llvm.vector.reduce.fmax.v2f64(<2 x double> undef)
|
|
; CHECK-MVEI-NEXT: Cost Model: Found costs of RThru:69 CodeSize:25 Lat:69 SizeLat:69 for: %fmax_v4f64 = call reassoc double @llvm.vector.reduce.fmax.v4f64(<4 x double> undef)
|
|
; CHECK-MVEI-NEXT: Cost Model: Found costs of RThru:118 CodeSize:67 Lat:118 SizeLat:118 for: %fmax_v4f128 = call reassoc fp128 @llvm.vector.reduce.fmax.v4f128(<4 x fp128> undef)
|
|
; CHECK-MVEI-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void
|
|
;
|
|
%fmax_v2f16 = call reassoc half @llvm.vector.reduce.fmax.v2f16(<2 x half> undef)
|
|
%fmax_v4f16 = call reassoc half @llvm.vector.reduce.fmax.v4f16(<4 x half> undef)
|
|
%fmax_v8f16 = call reassoc half @llvm.vector.reduce.fmax.v8f16(<8 x half> undef)
|
|
%fmax_v16f16 = call reassoc half @llvm.vector.reduce.fmax.v16f16(<16 x half> undef)
|
|
%fmax_v2f32 = call reassoc float @llvm.vector.reduce.fmax.v2f32(<2 x float> undef)
|
|
%fmax_v4f32 = call reassoc float @llvm.vector.reduce.fmax.v4f32(<4 x float> undef)
|
|
%fmax_v8f32 = call reassoc float @llvm.vector.reduce.fmax.v8f32(<8 x float> undef)
|
|
%fmax_v2f64 = call reassoc double @llvm.vector.reduce.fmax.v2f64(<2 x double> undef)
|
|
%fmax_v4f64 = call reassoc double @llvm.vector.reduce.fmax.v4f64(<4 x double> undef)
|
|
%fmax_v4f128 = call reassoc fp128 @llvm.vector.reduce.fmax.v4f128(<4 x fp128> undef)
|
|
ret void
|
|
}
|