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
77 lines
3.1 KiB
LLVM
77 lines
3.1 KiB
LLVM
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
|
|
; RUN: llc -mattr=+sme-fa64 -force-streaming-compatible < %s | FileCheck %s -check-prefix=FA64
|
|
; RUN: llc -mattr=+sve -force-streaming-compatible < %s | FileCheck %s -check-prefix=NO-FA64
|
|
; RUN: llc -force-streaming-compatible < %s | FileCheck %s --check-prefix=NONEON-NOSVE
|
|
|
|
target triple = "aarch64-unknown-linux-gnu"
|
|
|
|
define <8 x i8> @mla8xi8(<8 x i8> %A, <8 x i8> %B, <8 x i8> %C) {
|
|
; FA64-LABEL: mla8xi8:
|
|
; FA64: // %bb.0:
|
|
; FA64-NEXT: mla v2.8b, v0.8b, v1.8b
|
|
; FA64-NEXT: fmov d0, d2
|
|
; FA64-NEXT: ret
|
|
;
|
|
; NO-FA64-LABEL: mla8xi8:
|
|
; NO-FA64: // %bb.0:
|
|
; NO-FA64-NEXT: ptrue p0.b, vl8
|
|
; NO-FA64-NEXT: // kill: def $d0 killed $d0 def $z0
|
|
; NO-FA64-NEXT: // kill: def $d2 killed $d2 def $z2
|
|
; NO-FA64-NEXT: // kill: def $d1 killed $d1 def $z1
|
|
; NO-FA64-NEXT: mad z0.b, p0/m, z1.b, z2.b
|
|
; NO-FA64-NEXT: // kill: def $d0 killed $d0 killed $z0
|
|
; NO-FA64-NEXT: ret
|
|
;
|
|
; NONEON-NOSVE-LABEL: mla8xi8:
|
|
; NONEON-NOSVE: // %bb.0:
|
|
; NONEON-NOSVE-NEXT: stp d0, d1, [sp, #-32]!
|
|
; NONEON-NOSVE-NEXT: .cfi_def_cfa_offset 32
|
|
; NONEON-NOSVE-NEXT: ldrb w1, [sp, #15]
|
|
; NONEON-NOSVE-NEXT: ldrb w2, [sp, #7]
|
|
; NONEON-NOSVE-NEXT: str d2, [sp, #16]
|
|
; NONEON-NOSVE-NEXT: ldrb w5, [sp, #23]
|
|
; NONEON-NOSVE-NEXT: ldrb w3, [sp, #14]
|
|
; NONEON-NOSVE-NEXT: ldrb w4, [sp, #6]
|
|
; NONEON-NOSVE-NEXT: madd w1, w2, w1, w5
|
|
; NONEON-NOSVE-NEXT: ldrb w18, [sp, #13]
|
|
; NONEON-NOSVE-NEXT: ldrb w0, [sp, #5]
|
|
; NONEON-NOSVE-NEXT: ldrb w16, [sp, #12]
|
|
; NONEON-NOSVE-NEXT: ldrb w17, [sp, #4]
|
|
; NONEON-NOSVE-NEXT: ldrb w14, [sp, #11]
|
|
; NONEON-NOSVE-NEXT: ldrb w15, [sp, #3]
|
|
; NONEON-NOSVE-NEXT: ldrb w12, [sp, #10]
|
|
; NONEON-NOSVE-NEXT: ldrb w13, [sp, #2]
|
|
; NONEON-NOSVE-NEXT: strb w1, [sp, #31]
|
|
; NONEON-NOSVE-NEXT: ldrb w1, [sp, #22]
|
|
; NONEON-NOSVE-NEXT: ldrb w10, [sp, #9]
|
|
; NONEON-NOSVE-NEXT: madd w1, w4, w3, w1
|
|
; NONEON-NOSVE-NEXT: ldrb w11, [sp, #1]
|
|
; NONEON-NOSVE-NEXT: ldrb w8, [sp, #8]
|
|
; NONEON-NOSVE-NEXT: ldrb w9, [sp]
|
|
; NONEON-NOSVE-NEXT: strb w1, [sp, #30]
|
|
; NONEON-NOSVE-NEXT: ldrb w1, [sp, #21]
|
|
; NONEON-NOSVE-NEXT: madd w18, w0, w18, w1
|
|
; NONEON-NOSVE-NEXT: strb w18, [sp, #29]
|
|
; NONEON-NOSVE-NEXT: ldrb w18, [sp, #20]
|
|
; NONEON-NOSVE-NEXT: madd w16, w17, w16, w18
|
|
; NONEON-NOSVE-NEXT: strb w16, [sp, #28]
|
|
; NONEON-NOSVE-NEXT: ldrb w16, [sp, #19]
|
|
; NONEON-NOSVE-NEXT: madd w14, w15, w14, w16
|
|
; NONEON-NOSVE-NEXT: strb w14, [sp, #27]
|
|
; NONEON-NOSVE-NEXT: ldrb w14, [sp, #18]
|
|
; NONEON-NOSVE-NEXT: madd w12, w13, w12, w14
|
|
; NONEON-NOSVE-NEXT: strb w12, [sp, #26]
|
|
; NONEON-NOSVE-NEXT: ldrb w12, [sp, #17]
|
|
; NONEON-NOSVE-NEXT: madd w10, w11, w10, w12
|
|
; NONEON-NOSVE-NEXT: strb w10, [sp, #25]
|
|
; NONEON-NOSVE-NEXT: ldrb w10, [sp, #16]
|
|
; NONEON-NOSVE-NEXT: madd w8, w9, w8, w10
|
|
; NONEON-NOSVE-NEXT: strb w8, [sp, #24]
|
|
; NONEON-NOSVE-NEXT: ldr d0, [sp, #24]
|
|
; NONEON-NOSVE-NEXT: add sp, sp, #32
|
|
; NONEON-NOSVE-NEXT: ret
|
|
%tmp1 = mul <8 x i8> %A, %B;
|
|
%tmp2 = add <8 x i8> %C, %tmp1;
|
|
ret <8 x i8> %tmp2
|
|
}
|