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
652 lines
34 KiB
LLVM
652 lines
34 KiB
LLVM
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
|
|
; RUN: llc -mtriple=aarch64-linux-gnu -enable-misched=0 -mattr=+sve2p1 -enable-subreg-liveness=true < %s | FileCheck %s
|
|
; RUN: llc -mtriple=aarch64-linux-gnu -enable-misched=0 -mattr=+sme2 -force-streaming -enable-subreg-liveness=true < %s | FileCheck %s
|
|
; RUN: llc -mtriple=aarch64-linux-gnu -enable-misched=0 -mattr=+sme,+sve2p1 -force-streaming -enable-subreg-liveness=true < %s | FileCheck %s
|
|
|
|
; == Normal Multi-Vector Consecutive Stores ==
|
|
|
|
define void @st1_x2_i8(<vscale x 16 x i8> %unused, <vscale x 16 x i8> %zn0, <vscale x 16 x i8> %zn1, target("aarch64.svcount") %pn, ptr %ptr) nounwind {
|
|
; CHECK-LABEL: st1_x2_i8:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: str x29, [sp, #-16]! // 8-byte Folded Spill
|
|
; CHECK-NEXT: addvl sp, sp, #-1
|
|
; CHECK-NEXT: mov z3.d, z2.d
|
|
; CHECK-NEXT: mov z2.d, z1.d
|
|
; CHECK-NEXT: str p8, [sp, #7, mul vl] // 2-byte Spill
|
|
; CHECK-NEXT: mov p8.b, p0.b
|
|
; CHECK-NEXT: st1b { z2.b, z3.b }, pn8, [x0]
|
|
; CHECK-NEXT: ldr p8, [sp, #7, mul vl] // 2-byte Reload
|
|
; CHECK-NEXT: addvl sp, sp, #1
|
|
; CHECK-NEXT: ldr x29, [sp], #16 // 8-byte Folded Reload
|
|
; CHECK-NEXT: ret
|
|
call void @llvm.aarch64.sve.st1.pn.x2.nxv16i8(<vscale x 16 x i8> %zn0, <vscale x 16 x i8> %zn1, target("aarch64.svcount") %pn, ptr %ptr);
|
|
ret void
|
|
}
|
|
|
|
define void @st1_x2_i16(<vscale x 16 x i8> %unused, <vscale x 8 x i16> %zn0, <vscale x 8 x i16> %zn1, target("aarch64.svcount") %pn, ptr %ptr) nounwind {
|
|
; CHECK-LABEL: st1_x2_i16:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: str x29, [sp, #-16]! // 8-byte Folded Spill
|
|
; CHECK-NEXT: addvl sp, sp, #-1
|
|
; CHECK-NEXT: mov z3.d, z2.d
|
|
; CHECK-NEXT: mov z2.d, z1.d
|
|
; CHECK-NEXT: str p8, [sp, #7, mul vl] // 2-byte Spill
|
|
; CHECK-NEXT: mov p8.b, p0.b
|
|
; CHECK-NEXT: st1h { z2.h, z3.h }, pn8, [x0]
|
|
; CHECK-NEXT: ldr p8, [sp, #7, mul vl] // 2-byte Reload
|
|
; CHECK-NEXT: addvl sp, sp, #1
|
|
; CHECK-NEXT: ldr x29, [sp], #16 // 8-byte Folded Reload
|
|
; CHECK-NEXT: ret
|
|
call void @llvm.aarch64.sve.st1.pn.x2.nxv8i16(<vscale x 8 x i16> %zn0, <vscale x 8 x i16> %zn1, target("aarch64.svcount") %pn, ptr %ptr);
|
|
ret void
|
|
}
|
|
|
|
define void @st1_x2_i32(<vscale x 16 x i8> %unused, <vscale x 4 x i32> %zn0, <vscale x 4 x i32> %zn1, target("aarch64.svcount") %pn, ptr %ptr) nounwind {
|
|
; CHECK-LABEL: st1_x2_i32:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: str x29, [sp, #-16]! // 8-byte Folded Spill
|
|
; CHECK-NEXT: addvl sp, sp, #-1
|
|
; CHECK-NEXT: mov z3.d, z2.d
|
|
; CHECK-NEXT: mov z2.d, z1.d
|
|
; CHECK-NEXT: str p8, [sp, #7, mul vl] // 2-byte Spill
|
|
; CHECK-NEXT: mov p8.b, p0.b
|
|
; CHECK-NEXT: st1w { z2.s, z3.s }, pn8, [x0]
|
|
; CHECK-NEXT: ldr p8, [sp, #7, mul vl] // 2-byte Reload
|
|
; CHECK-NEXT: addvl sp, sp, #1
|
|
; CHECK-NEXT: ldr x29, [sp], #16 // 8-byte Folded Reload
|
|
; CHECK-NEXT: ret
|
|
call void @llvm.aarch64.sve.st1.pn.x2.nxv4i32(<vscale x 4 x i32> %zn0, <vscale x 4 x i32> %zn1, target("aarch64.svcount") %pn, ptr %ptr);
|
|
ret void
|
|
}
|
|
|
|
define void @st1_x2_i64(<vscale x 16 x i8> %unused, <vscale x 2 x i64> %zn0, <vscale x 2 x i64> %zn1, target("aarch64.svcount") %pn, ptr %ptr) nounwind {
|
|
; CHECK-LABEL: st1_x2_i64:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: str x29, [sp, #-16]! // 8-byte Folded Spill
|
|
; CHECK-NEXT: addvl sp, sp, #-1
|
|
; CHECK-NEXT: mov z3.d, z2.d
|
|
; CHECK-NEXT: mov z2.d, z1.d
|
|
; CHECK-NEXT: str p8, [sp, #7, mul vl] // 2-byte Spill
|
|
; CHECK-NEXT: mov p8.b, p0.b
|
|
; CHECK-NEXT: st1d { z2.d, z3.d }, pn8, [x0]
|
|
; CHECK-NEXT: ldr p8, [sp, #7, mul vl] // 2-byte Reload
|
|
; CHECK-NEXT: addvl sp, sp, #1
|
|
; CHECK-NEXT: ldr x29, [sp], #16 // 8-byte Folded Reload
|
|
; CHECK-NEXT: ret
|
|
call void @llvm.aarch64.sve.st1.pn.x2.nxv2i64(<vscale x 2 x i64> %zn0, <vscale x 2 x i64> %zn1, target("aarch64.svcount") %pn, ptr %ptr);
|
|
ret void
|
|
}
|
|
|
|
define void @st1_x2_f16(<vscale x 16 x i8> %unused, <vscale x 8 x half> %zn0, <vscale x 8 x half> %zn1, target("aarch64.svcount") %pn, ptr %ptr) nounwind {
|
|
; CHECK-LABEL: st1_x2_f16:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: str x29, [sp, #-16]! // 8-byte Folded Spill
|
|
; CHECK-NEXT: addvl sp, sp, #-1
|
|
; CHECK-NEXT: mov z3.d, z2.d
|
|
; CHECK-NEXT: mov z2.d, z1.d
|
|
; CHECK-NEXT: str p8, [sp, #7, mul vl] // 2-byte Spill
|
|
; CHECK-NEXT: mov p8.b, p0.b
|
|
; CHECK-NEXT: st1h { z2.h, z3.h }, pn8, [x0]
|
|
; CHECK-NEXT: ldr p8, [sp, #7, mul vl] // 2-byte Reload
|
|
; CHECK-NEXT: addvl sp, sp, #1
|
|
; CHECK-NEXT: ldr x29, [sp], #16 // 8-byte Folded Reload
|
|
; CHECK-NEXT: ret
|
|
call void @llvm.aarch64.sve.st1.pn.x2.nxv8f16(<vscale x 8 x half> %zn0, <vscale x 8 x half> %zn1, target("aarch64.svcount") %pn, ptr %ptr);
|
|
ret void
|
|
}
|
|
|
|
define void @st1_x2_bf16(<vscale x 16 x i8> %unused, <vscale x 8 x bfloat> %zn0, <vscale x 8 x bfloat> %zn1, target("aarch64.svcount") %pn, ptr %ptr) nounwind {
|
|
; CHECK-LABEL: st1_x2_bf16:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: str x29, [sp, #-16]! // 8-byte Folded Spill
|
|
; CHECK-NEXT: addvl sp, sp, #-1
|
|
; CHECK-NEXT: mov z3.d, z2.d
|
|
; CHECK-NEXT: mov z2.d, z1.d
|
|
; CHECK-NEXT: str p8, [sp, #7, mul vl] // 2-byte Spill
|
|
; CHECK-NEXT: mov p8.b, p0.b
|
|
; CHECK-NEXT: st1h { z2.h, z3.h }, pn8, [x0]
|
|
; CHECK-NEXT: ldr p8, [sp, #7, mul vl] // 2-byte Reload
|
|
; CHECK-NEXT: addvl sp, sp, #1
|
|
; CHECK-NEXT: ldr x29, [sp], #16 // 8-byte Folded Reload
|
|
; CHECK-NEXT: ret
|
|
call void @llvm.aarch64.sve.st1.pn.x2.nxv8bf16(<vscale x 8 x bfloat> %zn0, <vscale x 8 x bfloat> %zn1, target("aarch64.svcount") %pn, ptr %ptr);
|
|
ret void
|
|
}
|
|
|
|
define void @st1_x2_f32(<vscale x 16 x i8> %unused, <vscale x 4 x float> %zn0, <vscale x 4 x float> %zn1, target("aarch64.svcount") %pn, ptr %ptr) nounwind {
|
|
; CHECK-LABEL: st1_x2_f32:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: str x29, [sp, #-16]! // 8-byte Folded Spill
|
|
; CHECK-NEXT: addvl sp, sp, #-1
|
|
; CHECK-NEXT: mov z3.d, z2.d
|
|
; CHECK-NEXT: mov z2.d, z1.d
|
|
; CHECK-NEXT: str p8, [sp, #7, mul vl] // 2-byte Spill
|
|
; CHECK-NEXT: mov p8.b, p0.b
|
|
; CHECK-NEXT: st1w { z2.s, z3.s }, pn8, [x0]
|
|
; CHECK-NEXT: ldr p8, [sp, #7, mul vl] // 2-byte Reload
|
|
; CHECK-NEXT: addvl sp, sp, #1
|
|
; CHECK-NEXT: ldr x29, [sp], #16 // 8-byte Folded Reload
|
|
; CHECK-NEXT: ret
|
|
call void @llvm.aarch64.sve.st1.pn.x2.nxv4f32(<vscale x 4 x float> %zn0, <vscale x 4 x float> %zn1, target("aarch64.svcount") %pn, ptr %ptr);
|
|
ret void
|
|
}
|
|
|
|
define void @st1_x2_f64(<vscale x 16 x i8> %unused, <vscale x 2 x double> %zn0, <vscale x 2 x double> %zn1, target("aarch64.svcount") %pn, ptr %ptr) nounwind {
|
|
; CHECK-LABEL: st1_x2_f64:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: str x29, [sp, #-16]! // 8-byte Folded Spill
|
|
; CHECK-NEXT: addvl sp, sp, #-1
|
|
; CHECK-NEXT: mov z3.d, z2.d
|
|
; CHECK-NEXT: mov z2.d, z1.d
|
|
; CHECK-NEXT: str p8, [sp, #7, mul vl] // 2-byte Spill
|
|
; CHECK-NEXT: mov p8.b, p0.b
|
|
; CHECK-NEXT: st1d { z2.d, z3.d }, pn8, [x0]
|
|
; CHECK-NEXT: ldr p8, [sp, #7, mul vl] // 2-byte Reload
|
|
; CHECK-NEXT: addvl sp, sp, #1
|
|
; CHECK-NEXT: ldr x29, [sp], #16 // 8-byte Folded Reload
|
|
; CHECK-NEXT: ret
|
|
call void @llvm.aarch64.sve.st1.pn.x2.nxv2f64(<vscale x 2 x double> %zn0, <vscale x 2 x double> %zn1, target("aarch64.svcount") %pn, ptr %ptr);
|
|
ret void
|
|
}
|
|
|
|
define void @st1_x4_i8(<vscale x 16 x i8> %unused, <vscale x 16 x i8> %zn0, <vscale x 16 x i8> %zn1, <vscale x 16 x i8> %zn2, <vscale x 16 x i8> %zn3, target("aarch64.svcount") %pn, ptr %ptr) nounwind {
|
|
; CHECK-LABEL: st1_x4_i8:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: str x29, [sp, #-16]! // 8-byte Folded Spill
|
|
; CHECK-NEXT: addvl sp, sp, #-1
|
|
; CHECK-NEXT: mov z7.d, z4.d
|
|
; CHECK-NEXT: mov z6.d, z3.d
|
|
; CHECK-NEXT: str p8, [sp, #7, mul vl] // 2-byte Spill
|
|
; CHECK-NEXT: mov z5.d, z2.d
|
|
; CHECK-NEXT: mov z4.d, z1.d
|
|
; CHECK-NEXT: mov p8.b, p0.b
|
|
; CHECK-NEXT: st1b { z4.b - z7.b }, pn8, [x0]
|
|
; CHECK-NEXT: ldr p8, [sp, #7, mul vl] // 2-byte Reload
|
|
; CHECK-NEXT: addvl sp, sp, #1
|
|
; CHECK-NEXT: ldr x29, [sp], #16 // 8-byte Folded Reload
|
|
; CHECK-NEXT: ret
|
|
call void @llvm.aarch64.sve.st1.pn.x4.nxv16i8(<vscale x 16 x i8> %zn0, <vscale x 16 x i8> %zn1, <vscale x 16 x i8> %zn2, <vscale x 16 x i8> %zn3, target("aarch64.svcount") %pn, ptr %ptr);
|
|
ret void
|
|
}
|
|
|
|
define void @st1_x4_i16(<vscale x 16 x i8> %unused, <vscale x 8 x i16> %zn0, <vscale x 8 x i16> %zn1, <vscale x 8 x i16> %zn2, <vscale x 8 x i16> %zn3, target("aarch64.svcount") %pn, ptr %ptr) nounwind {
|
|
; CHECK-LABEL: st1_x4_i16:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: str x29, [sp, #-16]! // 8-byte Folded Spill
|
|
; CHECK-NEXT: addvl sp, sp, #-1
|
|
; CHECK-NEXT: mov z7.d, z4.d
|
|
; CHECK-NEXT: mov z6.d, z3.d
|
|
; CHECK-NEXT: str p8, [sp, #7, mul vl] // 2-byte Spill
|
|
; CHECK-NEXT: mov z5.d, z2.d
|
|
; CHECK-NEXT: mov z4.d, z1.d
|
|
; CHECK-NEXT: mov p8.b, p0.b
|
|
; CHECK-NEXT: st1h { z4.h - z7.h }, pn8, [x0]
|
|
; CHECK-NEXT: ldr p8, [sp, #7, mul vl] // 2-byte Reload
|
|
; CHECK-NEXT: addvl sp, sp, #1
|
|
; CHECK-NEXT: ldr x29, [sp], #16 // 8-byte Folded Reload
|
|
; CHECK-NEXT: ret
|
|
call void @llvm.aarch64.sve.st1.pn.x4.nxv8i16(<vscale x 8 x i16> %zn0, <vscale x 8 x i16> %zn1, <vscale x 8 x i16> %zn2, <vscale x 8 x i16> %zn3, target("aarch64.svcount") %pn, ptr %ptr);
|
|
ret void
|
|
}
|
|
|
|
define void @st1_x4_i32(<vscale x 16 x i8> %unused, <vscale x 4 x i32> %zn0, <vscale x 4 x i32> %zn1, <vscale x 4 x i32> %zn2, <vscale x 4 x i32> %zn3, target("aarch64.svcount") %pn, ptr %ptr) nounwind {
|
|
; CHECK-LABEL: st1_x4_i32:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: str x29, [sp, #-16]! // 8-byte Folded Spill
|
|
; CHECK-NEXT: addvl sp, sp, #-1
|
|
; CHECK-NEXT: mov z7.d, z4.d
|
|
; CHECK-NEXT: mov z6.d, z3.d
|
|
; CHECK-NEXT: str p8, [sp, #7, mul vl] // 2-byte Spill
|
|
; CHECK-NEXT: mov z5.d, z2.d
|
|
; CHECK-NEXT: mov z4.d, z1.d
|
|
; CHECK-NEXT: mov p8.b, p0.b
|
|
; CHECK-NEXT: st1w { z4.s - z7.s }, pn8, [x0]
|
|
; CHECK-NEXT: ldr p8, [sp, #7, mul vl] // 2-byte Reload
|
|
; CHECK-NEXT: addvl sp, sp, #1
|
|
; CHECK-NEXT: ldr x29, [sp], #16 // 8-byte Folded Reload
|
|
; CHECK-NEXT: ret
|
|
call void @llvm.aarch64.sve.st1.pn.x4.nxv4i32(<vscale x 4 x i32> %zn0, <vscale x 4 x i32> %zn1, <vscale x 4 x i32> %zn2, <vscale x 4 x i32> %zn3, target("aarch64.svcount") %pn, ptr %ptr);
|
|
ret void
|
|
}
|
|
|
|
define void @st1_x4_i64(<vscale x 16 x i8> %unused, <vscale x 2 x i64> %zn0, <vscale x 2 x i64> %zn1, <vscale x 2 x i64> %zn2, <vscale x 2 x i64> %zn3, target("aarch64.svcount") %pn, ptr %ptr) nounwind {
|
|
; CHECK-LABEL: st1_x4_i64:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: str x29, [sp, #-16]! // 8-byte Folded Spill
|
|
; CHECK-NEXT: addvl sp, sp, #-1
|
|
; CHECK-NEXT: mov z7.d, z4.d
|
|
; CHECK-NEXT: mov z6.d, z3.d
|
|
; CHECK-NEXT: str p8, [sp, #7, mul vl] // 2-byte Spill
|
|
; CHECK-NEXT: mov z5.d, z2.d
|
|
; CHECK-NEXT: mov z4.d, z1.d
|
|
; CHECK-NEXT: mov p8.b, p0.b
|
|
; CHECK-NEXT: st1d { z4.d - z7.d }, pn8, [x0]
|
|
; CHECK-NEXT: ldr p8, [sp, #7, mul vl] // 2-byte Reload
|
|
; CHECK-NEXT: addvl sp, sp, #1
|
|
; CHECK-NEXT: ldr x29, [sp], #16 // 8-byte Folded Reload
|
|
; CHECK-NEXT: ret
|
|
call void @llvm.aarch64.sve.st1.pn.x4.nxv2i64(<vscale x 2 x i64> %zn0, <vscale x 2 x i64> %zn1, <vscale x 2 x i64> %zn2, <vscale x 2 x i64> %zn3, target("aarch64.svcount") %pn, ptr %ptr);
|
|
ret void
|
|
}
|
|
|
|
define void @st1_x4_f16(<vscale x 16 x i8> %unused, <vscale x 8 x half> %zn0, <vscale x 8 x half> %zn1, <vscale x 8 x half> %zn2, <vscale x 8 x half> %zn3, target("aarch64.svcount") %pn, ptr %ptr) nounwind {
|
|
; CHECK-LABEL: st1_x4_f16:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: str x29, [sp, #-16]! // 8-byte Folded Spill
|
|
; CHECK-NEXT: addvl sp, sp, #-1
|
|
; CHECK-NEXT: mov z7.d, z4.d
|
|
; CHECK-NEXT: mov z6.d, z3.d
|
|
; CHECK-NEXT: str p8, [sp, #7, mul vl] // 2-byte Spill
|
|
; CHECK-NEXT: mov z5.d, z2.d
|
|
; CHECK-NEXT: mov z4.d, z1.d
|
|
; CHECK-NEXT: mov p8.b, p0.b
|
|
; CHECK-NEXT: st1h { z4.h - z7.h }, pn8, [x0]
|
|
; CHECK-NEXT: ldr p8, [sp, #7, mul vl] // 2-byte Reload
|
|
; CHECK-NEXT: addvl sp, sp, #1
|
|
; CHECK-NEXT: ldr x29, [sp], #16 // 8-byte Folded Reload
|
|
; CHECK-NEXT: ret
|
|
call void @llvm.aarch64.sve.st1.pn.x4.nxv8f16(<vscale x 8 x half> %zn0, <vscale x 8 x half> %zn1, <vscale x 8 x half> %zn2, <vscale x 8 x half> %zn3, target("aarch64.svcount") %pn, ptr %ptr);
|
|
ret void
|
|
}
|
|
|
|
define void @st1_x4_bf16(<vscale x 16 x i8> %unused, <vscale x 8 x bfloat> %zn0, <vscale x 8 x bfloat> %zn1, <vscale x 8 x bfloat> %zn2, <vscale x 8 x bfloat> %zn3, target("aarch64.svcount") %pn, ptr %ptr) nounwind {
|
|
; CHECK-LABEL: st1_x4_bf16:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: str x29, [sp, #-16]! // 8-byte Folded Spill
|
|
; CHECK-NEXT: addvl sp, sp, #-1
|
|
; CHECK-NEXT: mov z7.d, z4.d
|
|
; CHECK-NEXT: mov z6.d, z3.d
|
|
; CHECK-NEXT: str p8, [sp, #7, mul vl] // 2-byte Spill
|
|
; CHECK-NEXT: mov z5.d, z2.d
|
|
; CHECK-NEXT: mov z4.d, z1.d
|
|
; CHECK-NEXT: mov p8.b, p0.b
|
|
; CHECK-NEXT: st1h { z4.h - z7.h }, pn8, [x0]
|
|
; CHECK-NEXT: ldr p8, [sp, #7, mul vl] // 2-byte Reload
|
|
; CHECK-NEXT: addvl sp, sp, #1
|
|
; CHECK-NEXT: ldr x29, [sp], #16 // 8-byte Folded Reload
|
|
; CHECK-NEXT: ret
|
|
call void @llvm.aarch64.sve.st1.pn.x4.nxv8bf16(<vscale x 8 x bfloat> %zn0, <vscale x 8 x bfloat> %zn1, <vscale x 8 x bfloat> %zn2, <vscale x 8 x bfloat> %zn3, target("aarch64.svcount") %pn, ptr %ptr);
|
|
ret void
|
|
}
|
|
|
|
define void @st1_x4_f32(<vscale x 16 x i8> %unused, <vscale x 4 x float> %zn0, <vscale x 4 x float> %zn1, <vscale x 4 x float> %zn2, <vscale x 4 x float> %zn3, target("aarch64.svcount") %pn, ptr %ptr) nounwind {
|
|
; CHECK-LABEL: st1_x4_f32:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: str x29, [sp, #-16]! // 8-byte Folded Spill
|
|
; CHECK-NEXT: addvl sp, sp, #-1
|
|
; CHECK-NEXT: mov z7.d, z4.d
|
|
; CHECK-NEXT: mov z6.d, z3.d
|
|
; CHECK-NEXT: str p8, [sp, #7, mul vl] // 2-byte Spill
|
|
; CHECK-NEXT: mov z5.d, z2.d
|
|
; CHECK-NEXT: mov z4.d, z1.d
|
|
; CHECK-NEXT: mov p8.b, p0.b
|
|
; CHECK-NEXT: st1w { z4.s - z7.s }, pn8, [x0]
|
|
; CHECK-NEXT: ldr p8, [sp, #7, mul vl] // 2-byte Reload
|
|
; CHECK-NEXT: addvl sp, sp, #1
|
|
; CHECK-NEXT: ldr x29, [sp], #16 // 8-byte Folded Reload
|
|
; CHECK-NEXT: ret
|
|
call void @llvm.aarch64.sve.st1.pn.x4.nxv4f32(<vscale x 4 x float> %zn0, <vscale x 4 x float> %zn1, <vscale x 4 x float> %zn2, <vscale x 4 x float> %zn3, target("aarch64.svcount") %pn, ptr %ptr);
|
|
ret void
|
|
}
|
|
|
|
define void @st1_x4_f64(<vscale x 16 x i8> %unused, <vscale x 2 x double> %zn0, <vscale x 2 x double> %zn1, <vscale x 2 x double> %zn2, <vscale x 2 x double> %zn3, target("aarch64.svcount") %pn, ptr %ptr) nounwind {
|
|
; CHECK-LABEL: st1_x4_f64:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: str x29, [sp, #-16]! // 8-byte Folded Spill
|
|
; CHECK-NEXT: addvl sp, sp, #-1
|
|
; CHECK-NEXT: mov z7.d, z4.d
|
|
; CHECK-NEXT: mov z6.d, z3.d
|
|
; CHECK-NEXT: str p8, [sp, #7, mul vl] // 2-byte Spill
|
|
; CHECK-NEXT: mov z5.d, z2.d
|
|
; CHECK-NEXT: mov z4.d, z1.d
|
|
; CHECK-NEXT: mov p8.b, p0.b
|
|
; CHECK-NEXT: st1d { z4.d - z7.d }, pn8, [x0]
|
|
; CHECK-NEXT: ldr p8, [sp, #7, mul vl] // 2-byte Reload
|
|
; CHECK-NEXT: addvl sp, sp, #1
|
|
; CHECK-NEXT: ldr x29, [sp], #16 // 8-byte Folded Reload
|
|
; CHECK-NEXT: ret
|
|
call void @llvm.aarch64.sve.st1.pn.x4.nxv2f64(<vscale x 2 x double> %zn0, <vscale x 2 x double> %zn1, <vscale x 2 x double> %zn2, <vscale x 2 x double> %zn3, target("aarch64.svcount") %pn, ptr %ptr);
|
|
ret void
|
|
}
|
|
|
|
; == Non-temporal Multi-Vector Consecutive Stores ==
|
|
|
|
define void @stnt1_x2_i8(<vscale x 16 x i8> %unused, <vscale x 16 x i8> %zn0, <vscale x 16 x i8> %zn1, target("aarch64.svcount") %pn, ptr %ptr) nounwind {
|
|
; CHECK-LABEL: stnt1_x2_i8:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: str x29, [sp, #-16]! // 8-byte Folded Spill
|
|
; CHECK-NEXT: addvl sp, sp, #-1
|
|
; CHECK-NEXT: mov z3.d, z2.d
|
|
; CHECK-NEXT: mov z2.d, z1.d
|
|
; CHECK-NEXT: str p8, [sp, #7, mul vl] // 2-byte Spill
|
|
; CHECK-NEXT: mov p8.b, p0.b
|
|
; CHECK-NEXT: stnt1b { z2.b, z3.b }, pn8, [x0]
|
|
; CHECK-NEXT: ldr p8, [sp, #7, mul vl] // 2-byte Reload
|
|
; CHECK-NEXT: addvl sp, sp, #1
|
|
; CHECK-NEXT: ldr x29, [sp], #16 // 8-byte Folded Reload
|
|
; CHECK-NEXT: ret
|
|
call void @llvm.aarch64.sve.stnt1.pn.x2.nxv16i8(<vscale x 16 x i8> %zn0, <vscale x 16 x i8> %zn1, target("aarch64.svcount") %pn, ptr %ptr);
|
|
ret void
|
|
}
|
|
|
|
define void @stnt1_x2_i16(<vscale x 16 x i8> %unused, <vscale x 8 x i16> %zn0, <vscale x 8 x i16> %zn1, target("aarch64.svcount") %pn, ptr %ptr) nounwind {
|
|
; CHECK-LABEL: stnt1_x2_i16:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: str x29, [sp, #-16]! // 8-byte Folded Spill
|
|
; CHECK-NEXT: addvl sp, sp, #-1
|
|
; CHECK-NEXT: mov z3.d, z2.d
|
|
; CHECK-NEXT: mov z2.d, z1.d
|
|
; CHECK-NEXT: str p8, [sp, #7, mul vl] // 2-byte Spill
|
|
; CHECK-NEXT: mov p8.b, p0.b
|
|
; CHECK-NEXT: stnt1h { z2.h, z3.h }, pn8, [x0]
|
|
; CHECK-NEXT: ldr p8, [sp, #7, mul vl] // 2-byte Reload
|
|
; CHECK-NEXT: addvl sp, sp, #1
|
|
; CHECK-NEXT: ldr x29, [sp], #16 // 8-byte Folded Reload
|
|
; CHECK-NEXT: ret
|
|
call void @llvm.aarch64.sve.stnt1.pn.x2.nxv8i16(<vscale x 8 x i16> %zn0, <vscale x 8 x i16> %zn1, target("aarch64.svcount") %pn, ptr %ptr);
|
|
ret void
|
|
}
|
|
|
|
define void @stnt1_x2_i32(<vscale x 16 x i8> %unused, <vscale x 4 x i32> %zn0, <vscale x 4 x i32> %zn1, target("aarch64.svcount") %pn, ptr %ptr) nounwind {
|
|
; CHECK-LABEL: stnt1_x2_i32:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: str x29, [sp, #-16]! // 8-byte Folded Spill
|
|
; CHECK-NEXT: addvl sp, sp, #-1
|
|
; CHECK-NEXT: mov z3.d, z2.d
|
|
; CHECK-NEXT: mov z2.d, z1.d
|
|
; CHECK-NEXT: str p8, [sp, #7, mul vl] // 2-byte Spill
|
|
; CHECK-NEXT: mov p8.b, p0.b
|
|
; CHECK-NEXT: stnt1w { z2.s, z3.s }, pn8, [x0]
|
|
; CHECK-NEXT: ldr p8, [sp, #7, mul vl] // 2-byte Reload
|
|
; CHECK-NEXT: addvl sp, sp, #1
|
|
; CHECK-NEXT: ldr x29, [sp], #16 // 8-byte Folded Reload
|
|
; CHECK-NEXT: ret
|
|
call void @llvm.aarch64.sve.stnt1.pn.x2.nxv4i32(<vscale x 4 x i32> %zn0, <vscale x 4 x i32> %zn1, target("aarch64.svcount") %pn, ptr %ptr);
|
|
ret void
|
|
}
|
|
|
|
define void @stnt1_x2_i64(<vscale x 16 x i8> %unused, <vscale x 2 x i64> %zn0, <vscale x 2 x i64> %zn1, target("aarch64.svcount") %pn, ptr %ptr) nounwind {
|
|
; CHECK-LABEL: stnt1_x2_i64:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: str x29, [sp, #-16]! // 8-byte Folded Spill
|
|
; CHECK-NEXT: addvl sp, sp, #-1
|
|
; CHECK-NEXT: mov z3.d, z2.d
|
|
; CHECK-NEXT: mov z2.d, z1.d
|
|
; CHECK-NEXT: str p8, [sp, #7, mul vl] // 2-byte Spill
|
|
; CHECK-NEXT: mov p8.b, p0.b
|
|
; CHECK-NEXT: stnt1d { z2.d, z3.d }, pn8, [x0]
|
|
; CHECK-NEXT: ldr p8, [sp, #7, mul vl] // 2-byte Reload
|
|
; CHECK-NEXT: addvl sp, sp, #1
|
|
; CHECK-NEXT: ldr x29, [sp], #16 // 8-byte Folded Reload
|
|
; CHECK-NEXT: ret
|
|
call void @llvm.aarch64.sve.stnt1.pn.x2.nxv2i64(<vscale x 2 x i64> %zn0, <vscale x 2 x i64> %zn1, target("aarch64.svcount") %pn, ptr %ptr);
|
|
ret void
|
|
}
|
|
|
|
define void @stnt1_x2_f16(<vscale x 16 x i8> %unused, <vscale x 8 x half> %zn0, <vscale x 8 x half> %zn1, target("aarch64.svcount") %pn, ptr %ptr) nounwind {
|
|
; CHECK-LABEL: stnt1_x2_f16:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: str x29, [sp, #-16]! // 8-byte Folded Spill
|
|
; CHECK-NEXT: addvl sp, sp, #-1
|
|
; CHECK-NEXT: mov z3.d, z2.d
|
|
; CHECK-NEXT: mov z2.d, z1.d
|
|
; CHECK-NEXT: str p8, [sp, #7, mul vl] // 2-byte Spill
|
|
; CHECK-NEXT: mov p8.b, p0.b
|
|
; CHECK-NEXT: stnt1h { z2.h, z3.h }, pn8, [x0]
|
|
; CHECK-NEXT: ldr p8, [sp, #7, mul vl] // 2-byte Reload
|
|
; CHECK-NEXT: addvl sp, sp, #1
|
|
; CHECK-NEXT: ldr x29, [sp], #16 // 8-byte Folded Reload
|
|
; CHECK-NEXT: ret
|
|
call void @llvm.aarch64.sve.stnt1.pn.x2.nxv8f16(<vscale x 8 x half> %zn0, <vscale x 8 x half> %zn1, target("aarch64.svcount") %pn, ptr %ptr);
|
|
ret void
|
|
}
|
|
|
|
define void @stnt1_x2_bf16(<vscale x 16 x i8> %unused, <vscale x 8 x bfloat> %zn0, <vscale x 8 x bfloat> %zn1, target("aarch64.svcount") %pn, ptr %ptr) nounwind {
|
|
; CHECK-LABEL: stnt1_x2_bf16:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: str x29, [sp, #-16]! // 8-byte Folded Spill
|
|
; CHECK-NEXT: addvl sp, sp, #-1
|
|
; CHECK-NEXT: mov z3.d, z2.d
|
|
; CHECK-NEXT: mov z2.d, z1.d
|
|
; CHECK-NEXT: str p8, [sp, #7, mul vl] // 2-byte Spill
|
|
; CHECK-NEXT: mov p8.b, p0.b
|
|
; CHECK-NEXT: stnt1h { z2.h, z3.h }, pn8, [x0]
|
|
; CHECK-NEXT: ldr p8, [sp, #7, mul vl] // 2-byte Reload
|
|
; CHECK-NEXT: addvl sp, sp, #1
|
|
; CHECK-NEXT: ldr x29, [sp], #16 // 8-byte Folded Reload
|
|
; CHECK-NEXT: ret
|
|
call void @llvm.aarch64.sve.stnt1.pn.x2.nxv8bf16(<vscale x 8 x bfloat> %zn0, <vscale x 8 x bfloat> %zn1, target("aarch64.svcount") %pn, ptr %ptr);
|
|
ret void
|
|
}
|
|
|
|
define void @stnt1_x2_f32(<vscale x 16 x i8> %unused, <vscale x 4 x float> %zn0, <vscale x 4 x float> %zn1, target("aarch64.svcount") %pn, ptr %ptr) nounwind {
|
|
; CHECK-LABEL: stnt1_x2_f32:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: str x29, [sp, #-16]! // 8-byte Folded Spill
|
|
; CHECK-NEXT: addvl sp, sp, #-1
|
|
; CHECK-NEXT: mov z3.d, z2.d
|
|
; CHECK-NEXT: mov z2.d, z1.d
|
|
; CHECK-NEXT: str p8, [sp, #7, mul vl] // 2-byte Spill
|
|
; CHECK-NEXT: mov p8.b, p0.b
|
|
; CHECK-NEXT: stnt1w { z2.s, z3.s }, pn8, [x0]
|
|
; CHECK-NEXT: ldr p8, [sp, #7, mul vl] // 2-byte Reload
|
|
; CHECK-NEXT: addvl sp, sp, #1
|
|
; CHECK-NEXT: ldr x29, [sp], #16 // 8-byte Folded Reload
|
|
; CHECK-NEXT: ret
|
|
call void @llvm.aarch64.sve.stnt1.pn.x2.nxv4f32(<vscale x 4 x float> %zn0, <vscale x 4 x float> %zn1, target("aarch64.svcount") %pn, ptr %ptr);
|
|
ret void
|
|
}
|
|
|
|
define void @stnt1_x2_f64(<vscale x 16 x i8> %unused, <vscale x 2 x double> %zn0, <vscale x 2 x double> %zn1, target("aarch64.svcount") %pn, ptr %ptr) nounwind {
|
|
; CHECK-LABEL: stnt1_x2_f64:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: str x29, [sp, #-16]! // 8-byte Folded Spill
|
|
; CHECK-NEXT: addvl sp, sp, #-1
|
|
; CHECK-NEXT: mov z3.d, z2.d
|
|
; CHECK-NEXT: mov z2.d, z1.d
|
|
; CHECK-NEXT: str p8, [sp, #7, mul vl] // 2-byte Spill
|
|
; CHECK-NEXT: mov p8.b, p0.b
|
|
; CHECK-NEXT: stnt1d { z2.d, z3.d }, pn8, [x0]
|
|
; CHECK-NEXT: ldr p8, [sp, #7, mul vl] // 2-byte Reload
|
|
; CHECK-NEXT: addvl sp, sp, #1
|
|
; CHECK-NEXT: ldr x29, [sp], #16 // 8-byte Folded Reload
|
|
; CHECK-NEXT: ret
|
|
call void @llvm.aarch64.sve.stnt1.pn.x2.nxv2f64(<vscale x 2 x double> %zn0, <vscale x 2 x double> %zn1, target("aarch64.svcount") %pn, ptr %ptr);
|
|
ret void
|
|
}
|
|
|
|
define void @stnt1_x4_i8(<vscale x 16 x i8> %unused, <vscale x 16 x i8> %zn0, <vscale x 16 x i8> %zn1, <vscale x 16 x i8> %zn2, <vscale x 16 x i8> %zn3, target("aarch64.svcount") %pn, ptr %ptr) nounwind {
|
|
; CHECK-LABEL: stnt1_x4_i8:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: str x29, [sp, #-16]! // 8-byte Folded Spill
|
|
; CHECK-NEXT: addvl sp, sp, #-1
|
|
; CHECK-NEXT: mov z7.d, z4.d
|
|
; CHECK-NEXT: mov z6.d, z3.d
|
|
; CHECK-NEXT: str p8, [sp, #7, mul vl] // 2-byte Spill
|
|
; CHECK-NEXT: mov z5.d, z2.d
|
|
; CHECK-NEXT: mov z4.d, z1.d
|
|
; CHECK-NEXT: mov p8.b, p0.b
|
|
; CHECK-NEXT: stnt1b { z4.b - z7.b }, pn8, [x0]
|
|
; CHECK-NEXT: ldr p8, [sp, #7, mul vl] // 2-byte Reload
|
|
; CHECK-NEXT: addvl sp, sp, #1
|
|
; CHECK-NEXT: ldr x29, [sp], #16 // 8-byte Folded Reload
|
|
; CHECK-NEXT: ret
|
|
call void @llvm.aarch64.sve.stnt1.pn.x4.nxv16i8(<vscale x 16 x i8> %zn0, <vscale x 16 x i8> %zn1, <vscale x 16 x i8> %zn2, <vscale x 16 x i8> %zn3, target("aarch64.svcount") %pn, ptr %ptr);
|
|
ret void
|
|
}
|
|
|
|
define void @stnt1_x4_i16(<vscale x 16 x i8> %unused, <vscale x 8 x i16> %zn0, <vscale x 8 x i16> %zn1, <vscale x 8 x i16> %zn2, <vscale x 8 x i16> %zn3, target("aarch64.svcount") %pn, ptr %ptr) nounwind {
|
|
; CHECK-LABEL: stnt1_x4_i16:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: str x29, [sp, #-16]! // 8-byte Folded Spill
|
|
; CHECK-NEXT: addvl sp, sp, #-1
|
|
; CHECK-NEXT: mov z7.d, z4.d
|
|
; CHECK-NEXT: mov z6.d, z3.d
|
|
; CHECK-NEXT: str p8, [sp, #7, mul vl] // 2-byte Spill
|
|
; CHECK-NEXT: mov z5.d, z2.d
|
|
; CHECK-NEXT: mov z4.d, z1.d
|
|
; CHECK-NEXT: mov p8.b, p0.b
|
|
; CHECK-NEXT: stnt1h { z4.h - z7.h }, pn8, [x0]
|
|
; CHECK-NEXT: ldr p8, [sp, #7, mul vl] // 2-byte Reload
|
|
; CHECK-NEXT: addvl sp, sp, #1
|
|
; CHECK-NEXT: ldr x29, [sp], #16 // 8-byte Folded Reload
|
|
; CHECK-NEXT: ret
|
|
call void @llvm.aarch64.sve.stnt1.pn.x4.nxv8i16(<vscale x 8 x i16> %zn0, <vscale x 8 x i16> %zn1, <vscale x 8 x i16> %zn2, <vscale x 8 x i16> %zn3, target("aarch64.svcount") %pn, ptr %ptr);
|
|
ret void
|
|
}
|
|
|
|
define void @stnt1_x4_i32(<vscale x 16 x i8> %unused, <vscale x 4 x i32> %zn0, <vscale x 4 x i32> %zn1, <vscale x 4 x i32> %zn2, <vscale x 4 x i32> %zn3, target("aarch64.svcount") %pn, ptr %ptr) nounwind {
|
|
; CHECK-LABEL: stnt1_x4_i32:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: str x29, [sp, #-16]! // 8-byte Folded Spill
|
|
; CHECK-NEXT: addvl sp, sp, #-1
|
|
; CHECK-NEXT: mov z7.d, z4.d
|
|
; CHECK-NEXT: mov z6.d, z3.d
|
|
; CHECK-NEXT: str p8, [sp, #7, mul vl] // 2-byte Spill
|
|
; CHECK-NEXT: mov z5.d, z2.d
|
|
; CHECK-NEXT: mov z4.d, z1.d
|
|
; CHECK-NEXT: mov p8.b, p0.b
|
|
; CHECK-NEXT: stnt1w { z4.s - z7.s }, pn8, [x0]
|
|
; CHECK-NEXT: ldr p8, [sp, #7, mul vl] // 2-byte Reload
|
|
; CHECK-NEXT: addvl sp, sp, #1
|
|
; CHECK-NEXT: ldr x29, [sp], #16 // 8-byte Folded Reload
|
|
; CHECK-NEXT: ret
|
|
call void @llvm.aarch64.sve.stnt1.pn.x4.nxv4i32(<vscale x 4 x i32> %zn0, <vscale x 4 x i32> %zn1, <vscale x 4 x i32> %zn2, <vscale x 4 x i32> %zn3, target("aarch64.svcount") %pn, ptr %ptr);
|
|
ret void
|
|
}
|
|
|
|
define void @stnt1_x4_i64(<vscale x 16 x i8> %unused, <vscale x 2 x i64> %zn0, <vscale x 2 x i64> %zn1, <vscale x 2 x i64> %zn2, <vscale x 2 x i64> %zn3, target("aarch64.svcount") %pn, ptr %ptr) nounwind {
|
|
; CHECK-LABEL: stnt1_x4_i64:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: str x29, [sp, #-16]! // 8-byte Folded Spill
|
|
; CHECK-NEXT: addvl sp, sp, #-1
|
|
; CHECK-NEXT: mov z7.d, z4.d
|
|
; CHECK-NEXT: mov z6.d, z3.d
|
|
; CHECK-NEXT: str p8, [sp, #7, mul vl] // 2-byte Spill
|
|
; CHECK-NEXT: mov z5.d, z2.d
|
|
; CHECK-NEXT: mov z4.d, z1.d
|
|
; CHECK-NEXT: mov p8.b, p0.b
|
|
; CHECK-NEXT: stnt1d { z4.d - z7.d }, pn8, [x0]
|
|
; CHECK-NEXT: ldr p8, [sp, #7, mul vl] // 2-byte Reload
|
|
; CHECK-NEXT: addvl sp, sp, #1
|
|
; CHECK-NEXT: ldr x29, [sp], #16 // 8-byte Folded Reload
|
|
; CHECK-NEXT: ret
|
|
call void @llvm.aarch64.sve.stnt1.pn.x4.nxv2i64(<vscale x 2 x i64> %zn0, <vscale x 2 x i64> %zn1, <vscale x 2 x i64> %zn2, <vscale x 2 x i64> %zn3, target("aarch64.svcount") %pn, ptr %ptr);
|
|
ret void
|
|
}
|
|
|
|
define void @stnt1_x4_f16(<vscale x 16 x i8> %unused, <vscale x 8 x half> %zn0, <vscale x 8 x half> %zn1, <vscale x 8 x half> %zn2, <vscale x 8 x half> %zn3, target("aarch64.svcount") %pn, ptr %ptr) nounwind {
|
|
; CHECK-LABEL: stnt1_x4_f16:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: str x29, [sp, #-16]! // 8-byte Folded Spill
|
|
; CHECK-NEXT: addvl sp, sp, #-1
|
|
; CHECK-NEXT: mov z7.d, z4.d
|
|
; CHECK-NEXT: mov z6.d, z3.d
|
|
; CHECK-NEXT: str p8, [sp, #7, mul vl] // 2-byte Spill
|
|
; CHECK-NEXT: mov z5.d, z2.d
|
|
; CHECK-NEXT: mov z4.d, z1.d
|
|
; CHECK-NEXT: mov p8.b, p0.b
|
|
; CHECK-NEXT: stnt1h { z4.h - z7.h }, pn8, [x0]
|
|
; CHECK-NEXT: ldr p8, [sp, #7, mul vl] // 2-byte Reload
|
|
; CHECK-NEXT: addvl sp, sp, #1
|
|
; CHECK-NEXT: ldr x29, [sp], #16 // 8-byte Folded Reload
|
|
; CHECK-NEXT: ret
|
|
call void @llvm.aarch64.sve.stnt1.pn.x4.nxv8f16(<vscale x 8 x half> %zn0, <vscale x 8 x half> %zn1, <vscale x 8 x half> %zn2, <vscale x 8 x half> %zn3, target("aarch64.svcount") %pn, ptr %ptr);
|
|
ret void
|
|
}
|
|
|
|
define void @stnt1_x4_bf16(<vscale x 16 x i8> %unused, <vscale x 8 x bfloat> %zn0, <vscale x 8 x bfloat> %zn1, <vscale x 8 x bfloat> %zn2, <vscale x 8 x bfloat> %zn3, target("aarch64.svcount") %pn, ptr %ptr) nounwind {
|
|
; CHECK-LABEL: stnt1_x4_bf16:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: str x29, [sp, #-16]! // 8-byte Folded Spill
|
|
; CHECK-NEXT: addvl sp, sp, #-1
|
|
; CHECK-NEXT: mov z7.d, z4.d
|
|
; CHECK-NEXT: mov z6.d, z3.d
|
|
; CHECK-NEXT: str p8, [sp, #7, mul vl] // 2-byte Spill
|
|
; CHECK-NEXT: mov z5.d, z2.d
|
|
; CHECK-NEXT: mov z4.d, z1.d
|
|
; CHECK-NEXT: mov p8.b, p0.b
|
|
; CHECK-NEXT: stnt1h { z4.h - z7.h }, pn8, [x0]
|
|
; CHECK-NEXT: ldr p8, [sp, #7, mul vl] // 2-byte Reload
|
|
; CHECK-NEXT: addvl sp, sp, #1
|
|
; CHECK-NEXT: ldr x29, [sp], #16 // 8-byte Folded Reload
|
|
; CHECK-NEXT: ret
|
|
call void @llvm.aarch64.sve.stnt1.pn.x4.nxv8bf16(<vscale x 8 x bfloat> %zn0, <vscale x 8 x bfloat> %zn1, <vscale x 8 x bfloat> %zn2, <vscale x 8 x bfloat> %zn3, target("aarch64.svcount") %pn, ptr %ptr);
|
|
ret void
|
|
}
|
|
|
|
define void @stnt1_x4_f32(<vscale x 16 x i8> %unused, <vscale x 4 x float> %zn0, <vscale x 4 x float> %zn1, <vscale x 4 x float> %zn2, <vscale x 4 x float> %zn3, target("aarch64.svcount") %pn, ptr %ptr) nounwind {
|
|
; CHECK-LABEL: stnt1_x4_f32:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: str x29, [sp, #-16]! // 8-byte Folded Spill
|
|
; CHECK-NEXT: addvl sp, sp, #-1
|
|
; CHECK-NEXT: mov z7.d, z4.d
|
|
; CHECK-NEXT: mov z6.d, z3.d
|
|
; CHECK-NEXT: str p8, [sp, #7, mul vl] // 2-byte Spill
|
|
; CHECK-NEXT: mov z5.d, z2.d
|
|
; CHECK-NEXT: mov z4.d, z1.d
|
|
; CHECK-NEXT: mov p8.b, p0.b
|
|
; CHECK-NEXT: stnt1w { z4.s - z7.s }, pn8, [x0]
|
|
; CHECK-NEXT: ldr p8, [sp, #7, mul vl] // 2-byte Reload
|
|
; CHECK-NEXT: addvl sp, sp, #1
|
|
; CHECK-NEXT: ldr x29, [sp], #16 // 8-byte Folded Reload
|
|
; CHECK-NEXT: ret
|
|
call void @llvm.aarch64.sve.stnt1.pn.x4.nxv4f32(<vscale x 4 x float> %zn0, <vscale x 4 x float> %zn1, <vscale x 4 x float> %zn2, <vscale x 4 x float> %zn3, target("aarch64.svcount") %pn, ptr %ptr);
|
|
ret void
|
|
}
|
|
|
|
define void @stnt1_x4_f64(<vscale x 16 x i8> %unused, <vscale x 2 x double> %zn0, <vscale x 2 x double> %zn1, <vscale x 2 x double> %zn2, <vscale x 2 x double> %zn3, target("aarch64.svcount") %pn, ptr %ptr) nounwind {
|
|
; CHECK-LABEL: stnt1_x4_f64:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: str x29, [sp, #-16]! // 8-byte Folded Spill
|
|
; CHECK-NEXT: addvl sp, sp, #-1
|
|
; CHECK-NEXT: mov z7.d, z4.d
|
|
; CHECK-NEXT: mov z6.d, z3.d
|
|
; CHECK-NEXT: str p8, [sp, #7, mul vl] // 2-byte Spill
|
|
; CHECK-NEXT: mov z5.d, z2.d
|
|
; CHECK-NEXT: mov z4.d, z1.d
|
|
; CHECK-NEXT: mov p8.b, p0.b
|
|
; CHECK-NEXT: stnt1d { z4.d - z7.d }, pn8, [x0]
|
|
; CHECK-NEXT: ldr p8, [sp, #7, mul vl] // 2-byte Reload
|
|
; CHECK-NEXT: addvl sp, sp, #1
|
|
; CHECK-NEXT: ldr x29, [sp], #16 // 8-byte Folded Reload
|
|
; CHECK-NEXT: ret
|
|
call void @llvm.aarch64.sve.stnt1.pn.x4.nxv2f64(<vscale x 2 x double> %zn0, <vscale x 2 x double> %zn1, <vscale x 2 x double> %zn2, <vscale x 2 x double> %zn3, target("aarch64.svcount") %pn, ptr %ptr);
|
|
ret void
|
|
}
|
|
|
|
declare void @llvm.aarch64.sve.st1.pn.x2.nxv16i8(<vscale x 16 x i8>, <vscale x 16 x i8>, target("aarch64.svcount"), ptr)
|
|
declare void @llvm.aarch64.sve.st1.pn.x2.nxv8i16(<vscale x 8 x i16>, <vscale x 8 x i16>, target("aarch64.svcount"), ptr)
|
|
declare void @llvm.aarch64.sve.st1.pn.x2.nxv4i32(<vscale x 4 x i32>, <vscale x 4 x i32>, target("aarch64.svcount"), ptr)
|
|
declare void @llvm.aarch64.sve.st1.pn.x2.nxv2i64(<vscale x 2 x i64>, <vscale x 2 x i64>, target("aarch64.svcount"), ptr)
|
|
declare void @llvm.aarch64.sve.st1.pn.x2.nxv8f16(<vscale x 8 x half>, <vscale x 8 x half>, target("aarch64.svcount"), ptr)
|
|
declare void @llvm.aarch64.sve.st1.pn.x2.nxv8bf16(<vscale x 8 x bfloat>, <vscale x 8 x bfloat>, target("aarch64.svcount"), ptr)
|
|
declare void @llvm.aarch64.sve.st1.pn.x2.nxv4f32(<vscale x 4 x float>, <vscale x 4 x float>, target("aarch64.svcount"), ptr)
|
|
declare void @llvm.aarch64.sve.st1.pn.x2.nxv2f64(<vscale x 2 x double>, <vscale x 2 x double>, target("aarch64.svcount"), ptr)
|
|
declare void @llvm.aarch64.sve.stnt1.pn.x2.nxv16i8(<vscale x 16 x i8>, <vscale x 16 x i8>, target("aarch64.svcount"), ptr)
|
|
declare void @llvm.aarch64.sve.stnt1.pn.x2.nxv8i16(<vscale x 8 x i16>, <vscale x 8 x i16>, target("aarch64.svcount"), ptr)
|
|
declare void @llvm.aarch64.sve.stnt1.pn.x2.nxv4i32(<vscale x 4 x i32>, <vscale x 4 x i32>, target("aarch64.svcount"), ptr)
|
|
declare void @llvm.aarch64.sve.stnt1.pn.x2.nxv2i64(<vscale x 2 x i64>, <vscale x 2 x i64>, target("aarch64.svcount"), ptr)
|
|
declare void @llvm.aarch64.sve.stnt1.pn.x2.nxv8f16(<vscale x 8 x half>, <vscale x 8 x half>, target("aarch64.svcount"), ptr)
|
|
declare void @llvm.aarch64.sve.stnt1.pn.x2.nxv8bf16(<vscale x 8 x bfloat>, <vscale x 8 x bfloat>, target("aarch64.svcount"), ptr)
|
|
declare void @llvm.aarch64.sve.stnt1.pn.x2.nxv4f32(<vscale x 4 x float>, <vscale x 4 x float>, target("aarch64.svcount"), ptr)
|
|
declare void @llvm.aarch64.sve.stnt1.pn.x2.nxv2f64(<vscale x 2 x double>, <vscale x 2 x double>, target("aarch64.svcount"), ptr)
|
|
|
|
|
|
declare void @llvm.aarch64.sve.st1.pn.x4.nxv16i8(<vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, target("aarch64.svcount"), ptr)
|
|
declare void @llvm.aarch64.sve.st1.pn.x4.nxv8i16(<vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16>, target("aarch64.svcount"), ptr)
|
|
declare void @llvm.aarch64.sve.st1.pn.x4.nxv4i32(<vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32>, target("aarch64.svcount"), ptr)
|
|
declare void @llvm.aarch64.sve.st1.pn.x4.nxv2i64(<vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64>, target("aarch64.svcount"), ptr)
|
|
declare void @llvm.aarch64.sve.st1.pn.x4.nxv8f16(<vscale x 8 x half>, <vscale x 8 x half>, <vscale x 8 x half>, <vscale x 8 x half>, target("aarch64.svcount"), ptr)
|
|
declare void @llvm.aarch64.sve.st1.pn.x4.nxv8bf16(<vscale x 8 x bfloat>, <vscale x 8 x bfloat>, <vscale x 8 x bfloat>, <vscale x 8 x bfloat>, target("aarch64.svcount"), ptr)
|
|
declare void @llvm.aarch64.sve.st1.pn.x4.nxv4f32(<vscale x 4 x float>, <vscale x 4 x float>, <vscale x 4 x float>, <vscale x 4 x float>, target("aarch64.svcount"), ptr)
|
|
declare void @llvm.aarch64.sve.st1.pn.x4.nxv2f64(<vscale x 2 x double>, <vscale x 2 x double>, <vscale x 2 x double>, <vscale x 2 x double>, target("aarch64.svcount"), ptr)
|
|
declare void @llvm.aarch64.sve.stnt1.pn.x4.nxv16i8(<vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, target("aarch64.svcount"), ptr)
|
|
declare void @llvm.aarch64.sve.stnt1.pn.x4.nxv8i16(<vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16>, target("aarch64.svcount"), ptr)
|
|
declare void @llvm.aarch64.sve.stnt1.pn.x4.nxv4i32(<vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32>, target("aarch64.svcount"), ptr)
|
|
declare void @llvm.aarch64.sve.stnt1.pn.x4.nxv2i64(<vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64>, target("aarch64.svcount"), ptr)
|
|
declare void @llvm.aarch64.sve.stnt1.pn.x4.nxv8f16(<vscale x 8 x half>, <vscale x 8 x half>, <vscale x 8 x half>, <vscale x 8 x half>, target("aarch64.svcount"), ptr)
|
|
declare void @llvm.aarch64.sve.stnt1.pn.x4.nxv8bf16(<vscale x 8 x bfloat>, <vscale x 8 x bfloat>, <vscale x 8 x bfloat>, <vscale x 8 x bfloat>, target("aarch64.svcount"), ptr)
|
|
declare void @llvm.aarch64.sve.stnt1.pn.x4.nxv4f32(<vscale x 4 x float>, <vscale x 4 x float>, <vscale x 4 x float>, <vscale x 4 x float>, target("aarch64.svcount"), ptr)
|
|
declare void @llvm.aarch64.sve.stnt1.pn.x4.nxv2f64(<vscale x 2 x double>, <vscale x 2 x double>, <vscale x 2 x double>, <vscale x 2 x double>, target("aarch64.svcount"), ptr)
|