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
349 lines
12 KiB
LLVM
349 lines
12 KiB
LLVM
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --extra_scrub
|
|
; RUN: llc -mattr=+sve,+bf16 < %s | FileCheck %s
|
|
; RUN: llc -mattr=+sme -force-streaming < %s | FileCheck %s
|
|
|
|
target triple = "aarch64-unknown-linux-gnu"
|
|
|
|
;
|
|
; FCEIL
|
|
;
|
|
|
|
define <vscale x 2 x bfloat> @frintp_nxv2bf16(<vscale x 2 x bfloat> %a) {
|
|
; CHECK-LABEL: frintp_nxv2bf16:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: lsl z0.s, z0.s, #16
|
|
; CHECK-NEXT: ptrue p0.d
|
|
; CHECK-NEXT: frintp z0.s, p0/m, z0.s
|
|
; CHECK-NEXT: bfcvt z0.h, p0/m, z0.s
|
|
; CHECK-NEXT: ret
|
|
%res = call <vscale x 2 x bfloat> @llvm.ceil.nxv2bf16(<vscale x 2 x bfloat> %a)
|
|
ret <vscale x 2 x bfloat> %res
|
|
}
|
|
|
|
define <vscale x 4 x bfloat> @frintp_nxv4bf16(<vscale x 4 x bfloat> %a) {
|
|
; CHECK-LABEL: frintp_nxv4bf16:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: lsl z0.s, z0.s, #16
|
|
; CHECK-NEXT: ptrue p0.s
|
|
; CHECK-NEXT: frintp z0.s, p0/m, z0.s
|
|
; CHECK-NEXT: bfcvt z0.h, p0/m, z0.s
|
|
; CHECK-NEXT: ret
|
|
%res = call <vscale x 4 x bfloat> @llvm.ceil.nxv4bf16(<vscale x 4 x bfloat> %a)
|
|
ret <vscale x 4 x bfloat> %res
|
|
}
|
|
|
|
define <vscale x 8 x bfloat> @frintp_nxv8bf16(<vscale x 8 x bfloat> %a) {
|
|
; CHECK-LABEL: frintp_nxv8bf16:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: movi v1.2d, #0000000000000000
|
|
; CHECK-NEXT: ptrue p0.s
|
|
; CHECK-NEXT: zip2 z2.h, z1.h, z0.h
|
|
; CHECK-NEXT: zip1 z0.h, z1.h, z0.h
|
|
; CHECK-NEXT: frintp z2.s, p0/m, z2.s
|
|
; CHECK-NEXT: frintp z0.s, p0/m, z0.s
|
|
; CHECK-NEXT: bfcvt z1.h, p0/m, z2.s
|
|
; CHECK-NEXT: bfcvt z0.h, p0/m, z0.s
|
|
; CHECK-NEXT: uzp1 z0.h, z0.h, z1.h
|
|
; CHECK-NEXT: ret
|
|
%res = call <vscale x 8 x bfloat> @llvm.ceil.nxv8bf16(<vscale x 8 x bfloat> %a)
|
|
ret <vscale x 8 x bfloat> %res
|
|
}
|
|
|
|
;
|
|
; FFLOOR
|
|
;
|
|
|
|
define <vscale x 2 x bfloat> @frintm_nxv2bf16(<vscale x 2 x bfloat> %a) {
|
|
; CHECK-LABEL: frintm_nxv2bf16:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: lsl z0.s, z0.s, #16
|
|
; CHECK-NEXT: ptrue p0.d
|
|
; CHECK-NEXT: frintm z0.s, p0/m, z0.s
|
|
; CHECK-NEXT: bfcvt z0.h, p0/m, z0.s
|
|
; CHECK-NEXT: ret
|
|
%res = call <vscale x 2 x bfloat> @llvm.floor.nxv2bf16(<vscale x 2 x bfloat> %a)
|
|
ret <vscale x 2 x bfloat> %res
|
|
}
|
|
|
|
define <vscale x 4 x bfloat> @frintm_nxv4bf16(<vscale x 4 x bfloat> %a) {
|
|
; CHECK-LABEL: frintm_nxv4bf16:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: lsl z0.s, z0.s, #16
|
|
; CHECK-NEXT: ptrue p0.s
|
|
; CHECK-NEXT: frintm z0.s, p0/m, z0.s
|
|
; CHECK-NEXT: bfcvt z0.h, p0/m, z0.s
|
|
; CHECK-NEXT: ret
|
|
%res = call <vscale x 4 x bfloat> @llvm.floor.nxv4bf16(<vscale x 4 x bfloat> %a)
|
|
ret <vscale x 4 x bfloat> %res
|
|
}
|
|
|
|
define <vscale x 8 x bfloat> @frintm_nxv8bf16(<vscale x 8 x bfloat> %a) {
|
|
; CHECK-LABEL: frintm_nxv8bf16:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: movi v1.2d, #0000000000000000
|
|
; CHECK-NEXT: ptrue p0.s
|
|
; CHECK-NEXT: zip2 z2.h, z1.h, z0.h
|
|
; CHECK-NEXT: zip1 z0.h, z1.h, z0.h
|
|
; CHECK-NEXT: frintm z2.s, p0/m, z2.s
|
|
; CHECK-NEXT: frintm z0.s, p0/m, z0.s
|
|
; CHECK-NEXT: bfcvt z1.h, p0/m, z2.s
|
|
; CHECK-NEXT: bfcvt z0.h, p0/m, z0.s
|
|
; CHECK-NEXT: uzp1 z0.h, z0.h, z1.h
|
|
; CHECK-NEXT: ret
|
|
%res = call <vscale x 8 x bfloat> @llvm.floor.nxv8bf16(<vscale x 8 x bfloat> %a)
|
|
ret <vscale x 8 x bfloat> %res
|
|
}
|
|
|
|
;
|
|
; FNEARBYINT
|
|
;
|
|
|
|
define <vscale x 2 x bfloat> @frinti_nxv2bf16(<vscale x 2 x bfloat> %a) {
|
|
; CHECK-LABEL: frinti_nxv2bf16:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: lsl z0.s, z0.s, #16
|
|
; CHECK-NEXT: ptrue p0.d
|
|
; CHECK-NEXT: frinti z0.s, p0/m, z0.s
|
|
; CHECK-NEXT: bfcvt z0.h, p0/m, z0.s
|
|
; CHECK-NEXT: ret
|
|
%res = call <vscale x 2 x bfloat> @llvm.nearbyint.nxv2bf16(<vscale x 2 x bfloat> %a)
|
|
ret <vscale x 2 x bfloat> %res
|
|
}
|
|
|
|
define <vscale x 4 x bfloat> @frinti_nxv4bf16(<vscale x 4 x bfloat> %a) {
|
|
; CHECK-LABEL: frinti_nxv4bf16:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: lsl z0.s, z0.s, #16
|
|
; CHECK-NEXT: ptrue p0.s
|
|
; CHECK-NEXT: frinti z0.s, p0/m, z0.s
|
|
; CHECK-NEXT: bfcvt z0.h, p0/m, z0.s
|
|
; CHECK-NEXT: ret
|
|
%res = call <vscale x 4 x bfloat> @llvm.nearbyint.nxv4bf16(<vscale x 4 x bfloat> %a)
|
|
ret <vscale x 4 x bfloat> %res
|
|
}
|
|
|
|
define <vscale x 8 x bfloat> @frinti_nxv8bf16(<vscale x 8 x bfloat> %a) {
|
|
; CHECK-LABEL: frinti_nxv8bf16:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: movi v1.2d, #0000000000000000
|
|
; CHECK-NEXT: ptrue p0.s
|
|
; CHECK-NEXT: zip2 z2.h, z1.h, z0.h
|
|
; CHECK-NEXT: zip1 z0.h, z1.h, z0.h
|
|
; CHECK-NEXT: frinti z2.s, p0/m, z2.s
|
|
; CHECK-NEXT: frinti z0.s, p0/m, z0.s
|
|
; CHECK-NEXT: bfcvt z1.h, p0/m, z2.s
|
|
; CHECK-NEXT: bfcvt z0.h, p0/m, z0.s
|
|
; CHECK-NEXT: uzp1 z0.h, z0.h, z1.h
|
|
; CHECK-NEXT: ret
|
|
%res = call <vscale x 8 x bfloat> @llvm.nearbyint.nxv8bf16(<vscale x 8 x bfloat> %a)
|
|
ret <vscale x 8 x bfloat> %res
|
|
}
|
|
|
|
;
|
|
; FRINT
|
|
;
|
|
|
|
define <vscale x 2 x bfloat> @frintx_nxv2bf16(<vscale x 2 x bfloat> %a) {
|
|
; CHECK-LABEL: frintx_nxv2bf16:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: lsl z0.s, z0.s, #16
|
|
; CHECK-NEXT: ptrue p0.d
|
|
; CHECK-NEXT: frintx z0.s, p0/m, z0.s
|
|
; CHECK-NEXT: bfcvt z0.h, p0/m, z0.s
|
|
; CHECK-NEXT: ret
|
|
%res = call <vscale x 2 x bfloat> @llvm.rint.nxv2bf16(<vscale x 2 x bfloat> %a)
|
|
ret <vscale x 2 x bfloat> %res
|
|
}
|
|
|
|
define <vscale x 4 x bfloat> @frintx_nxv4bf16(<vscale x 4 x bfloat> %a) {
|
|
; CHECK-LABEL: frintx_nxv4bf16:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: lsl z0.s, z0.s, #16
|
|
; CHECK-NEXT: ptrue p0.s
|
|
; CHECK-NEXT: frintx z0.s, p0/m, z0.s
|
|
; CHECK-NEXT: bfcvt z0.h, p0/m, z0.s
|
|
; CHECK-NEXT: ret
|
|
%res = call <vscale x 4 x bfloat> @llvm.rint.nxv4bf16(<vscale x 4 x bfloat> %a)
|
|
ret <vscale x 4 x bfloat> %res
|
|
}
|
|
|
|
define <vscale x 8 x bfloat> @frintx_nxv8bf16(<vscale x 8 x bfloat> %a) {
|
|
; CHECK-LABEL: frintx_nxv8bf16:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: movi v1.2d, #0000000000000000
|
|
; CHECK-NEXT: ptrue p0.s
|
|
; CHECK-NEXT: zip2 z2.h, z1.h, z0.h
|
|
; CHECK-NEXT: zip1 z0.h, z1.h, z0.h
|
|
; CHECK-NEXT: frintx z2.s, p0/m, z2.s
|
|
; CHECK-NEXT: frintx z0.s, p0/m, z0.s
|
|
; CHECK-NEXT: bfcvt z1.h, p0/m, z2.s
|
|
; CHECK-NEXT: bfcvt z0.h, p0/m, z0.s
|
|
; CHECK-NEXT: uzp1 z0.h, z0.h, z1.h
|
|
; CHECK-NEXT: ret
|
|
%res = call <vscale x 8 x bfloat> @llvm.rint.nxv8bf16(<vscale x 8 x bfloat> %a)
|
|
ret <vscale x 8 x bfloat> %res
|
|
}
|
|
|
|
;
|
|
; ROUND
|
|
;
|
|
|
|
define <vscale x 2 x bfloat> @frinta_nxv2bf16(<vscale x 2 x bfloat> %a) {
|
|
; CHECK-LABEL: frinta_nxv2bf16:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: lsl z0.s, z0.s, #16
|
|
; CHECK-NEXT: ptrue p0.d
|
|
; CHECK-NEXT: frinta z0.s, p0/m, z0.s
|
|
; CHECK-NEXT: bfcvt z0.h, p0/m, z0.s
|
|
; CHECK-NEXT: ret
|
|
%res = call <vscale x 2 x bfloat> @llvm.round.nxv2bf16(<vscale x 2 x bfloat> %a)
|
|
ret <vscale x 2 x bfloat> %res
|
|
}
|
|
|
|
define <vscale x 4 x bfloat> @frinta_nxv4bf16(<vscale x 4 x bfloat> %a) {
|
|
; CHECK-LABEL: frinta_nxv4bf16:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: lsl z0.s, z0.s, #16
|
|
; CHECK-NEXT: ptrue p0.s
|
|
; CHECK-NEXT: frinta z0.s, p0/m, z0.s
|
|
; CHECK-NEXT: bfcvt z0.h, p0/m, z0.s
|
|
; CHECK-NEXT: ret
|
|
%res = call <vscale x 4 x bfloat> @llvm.round.nxv4bf16(<vscale x 4 x bfloat> %a)
|
|
ret <vscale x 4 x bfloat> %res
|
|
}
|
|
|
|
define <vscale x 8 x bfloat> @frinta_nxv8bf16(<vscale x 8 x bfloat> %a) {
|
|
; CHECK-LABEL: frinta_nxv8bf16:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: movi v1.2d, #0000000000000000
|
|
; CHECK-NEXT: ptrue p0.s
|
|
; CHECK-NEXT: zip2 z2.h, z1.h, z0.h
|
|
; CHECK-NEXT: zip1 z0.h, z1.h, z0.h
|
|
; CHECK-NEXT: frinta z2.s, p0/m, z2.s
|
|
; CHECK-NEXT: frinta z0.s, p0/m, z0.s
|
|
; CHECK-NEXT: bfcvt z1.h, p0/m, z2.s
|
|
; CHECK-NEXT: bfcvt z0.h, p0/m, z0.s
|
|
; CHECK-NEXT: uzp1 z0.h, z0.h, z1.h
|
|
; CHECK-NEXT: ret
|
|
%res = call <vscale x 8 x bfloat> @llvm.round.nxv8bf16(<vscale x 8 x bfloat> %a)
|
|
ret <vscale x 8 x bfloat> %res
|
|
}
|
|
|
|
;
|
|
; ROUNDEVEN
|
|
;
|
|
|
|
define <vscale x 2 x bfloat> @frintn_nxv2bf16(<vscale x 2 x bfloat> %a) {
|
|
; CHECK-LABEL: frintn_nxv2bf16:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: lsl z0.s, z0.s, #16
|
|
; CHECK-NEXT: ptrue p0.d
|
|
; CHECK-NEXT: frintn z0.s, p0/m, z0.s
|
|
; CHECK-NEXT: bfcvt z0.h, p0/m, z0.s
|
|
; CHECK-NEXT: ret
|
|
%res = call <vscale x 2 x bfloat> @llvm.roundeven.nxv2bf16(<vscale x 2 x bfloat> %a)
|
|
ret <vscale x 2 x bfloat> %res
|
|
}
|
|
|
|
define <vscale x 4 x bfloat> @frintn_nxv4bf16(<vscale x 4 x bfloat> %a) {
|
|
; CHECK-LABEL: frintn_nxv4bf16:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: lsl z0.s, z0.s, #16
|
|
; CHECK-NEXT: ptrue p0.s
|
|
; CHECK-NEXT: frintn z0.s, p0/m, z0.s
|
|
; CHECK-NEXT: bfcvt z0.h, p0/m, z0.s
|
|
; CHECK-NEXT: ret
|
|
%res = call <vscale x 4 x bfloat> @llvm.roundeven.nxv4bf16(<vscale x 4 x bfloat> %a)
|
|
ret <vscale x 4 x bfloat> %res
|
|
}
|
|
|
|
define <vscale x 8 x bfloat> @frintn_nxv8bf16(<vscale x 8 x bfloat> %a) {
|
|
; CHECK-LABEL: frintn_nxv8bf16:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: movi v1.2d, #0000000000000000
|
|
; CHECK-NEXT: ptrue p0.s
|
|
; CHECK-NEXT: zip2 z2.h, z1.h, z0.h
|
|
; CHECK-NEXT: zip1 z0.h, z1.h, z0.h
|
|
; CHECK-NEXT: frintn z2.s, p0/m, z2.s
|
|
; CHECK-NEXT: frintn z0.s, p0/m, z0.s
|
|
; CHECK-NEXT: bfcvt z1.h, p0/m, z2.s
|
|
; CHECK-NEXT: bfcvt z0.h, p0/m, z0.s
|
|
; CHECK-NEXT: uzp1 z0.h, z0.h, z1.h
|
|
; CHECK-NEXT: ret
|
|
%res = call <vscale x 8 x bfloat> @llvm.roundeven.nxv8bf16(<vscale x 8 x bfloat> %a)
|
|
ret <vscale x 8 x bfloat> %res
|
|
}
|
|
|
|
;
|
|
; FTRUNC
|
|
;
|
|
|
|
define <vscale x 2 x bfloat> @frintz_nxv2bf16(<vscale x 2 x bfloat> %a) {
|
|
; CHECK-LABEL: frintz_nxv2bf16:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: lsl z0.s, z0.s, #16
|
|
; CHECK-NEXT: ptrue p0.d
|
|
; CHECK-NEXT: frintz z0.s, p0/m, z0.s
|
|
; CHECK-NEXT: bfcvt z0.h, p0/m, z0.s
|
|
; CHECK-NEXT: ret
|
|
%res = call <vscale x 2 x bfloat> @llvm.trunc.nxv2bf16(<vscale x 2 x bfloat> %a)
|
|
ret <vscale x 2 x bfloat> %res
|
|
}
|
|
|
|
define <vscale x 4 x bfloat> @frintz_nxv4bf16(<vscale x 4 x bfloat> %a) {
|
|
; CHECK-LABEL: frintz_nxv4bf16:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: lsl z0.s, z0.s, #16
|
|
; CHECK-NEXT: ptrue p0.s
|
|
; CHECK-NEXT: frintz z0.s, p0/m, z0.s
|
|
; CHECK-NEXT: bfcvt z0.h, p0/m, z0.s
|
|
; CHECK-NEXT: ret
|
|
%res = call <vscale x 4 x bfloat> @llvm.trunc.nxv4bf16(<vscale x 4 x bfloat> %a)
|
|
ret <vscale x 4 x bfloat> %res
|
|
}
|
|
|
|
define <vscale x 8 x bfloat> @frintz_nxv8bf16(<vscale x 8 x bfloat> %a) {
|
|
; CHECK-LABEL: frintz_nxv8bf16:
|
|
; CHECK: // %bb.0:
|
|
; CHECK-NEXT: movi v1.2d, #0000000000000000
|
|
; CHECK-NEXT: ptrue p0.s
|
|
; CHECK-NEXT: zip2 z2.h, z1.h, z0.h
|
|
; CHECK-NEXT: zip1 z0.h, z1.h, z0.h
|
|
; CHECK-NEXT: frintz z2.s, p0/m, z2.s
|
|
; CHECK-NEXT: frintz z0.s, p0/m, z0.s
|
|
; CHECK-NEXT: bfcvt z1.h, p0/m, z2.s
|
|
; CHECK-NEXT: bfcvt z0.h, p0/m, z0.s
|
|
; CHECK-NEXT: uzp1 z0.h, z0.h, z1.h
|
|
; CHECK-NEXT: ret
|
|
%res = call <vscale x 8 x bfloat> @llvm.trunc.nxv8bf16(<vscale x 8 x bfloat> %a)
|
|
ret <vscale x 8 x bfloat> %res
|
|
}
|
|
|
|
declare <vscale x 2 x bfloat> @llvm.ceil.nxv2bf16( <vscale x 2 x bfloat>)
|
|
declare <vscale x 4 x bfloat> @llvm.ceil.nxv4bf16( <vscale x 4 x bfloat>)
|
|
declare <vscale x 8 x bfloat> @llvm.ceil.nxv8bf16( <vscale x 8 x bfloat>)
|
|
|
|
declare <vscale x 2 x bfloat> @llvm.floor.nxv2bf16( <vscale x 2 x bfloat>)
|
|
declare <vscale x 4 x bfloat> @llvm.floor.nxv4bf16( <vscale x 4 x bfloat>)
|
|
declare <vscale x 8 x bfloat> @llvm.floor.nxv8bf16( <vscale x 8 x bfloat>)
|
|
|
|
declare <vscale x 2 x bfloat> @llvm.nearbyint.nxv2bf16( <vscale x 2 x bfloat>)
|
|
declare <vscale x 4 x bfloat> @llvm.nearbyint.nxv4bf16( <vscale x 4 x bfloat>)
|
|
declare <vscale x 8 x bfloat> @llvm.nearbyint.nxv8bf16( <vscale x 8 x bfloat>)
|
|
|
|
declare <vscale x 2 x bfloat> @llvm.rint.nxv2bf16( <vscale x 2 x bfloat>)
|
|
declare <vscale x 4 x bfloat> @llvm.rint.nxv4bf16( <vscale x 4 x bfloat>)
|
|
declare <vscale x 8 x bfloat> @llvm.rint.nxv8bf16( <vscale x 8 x bfloat>)
|
|
|
|
declare <vscale x 2 x bfloat> @llvm.round.nxv2bf16( <vscale x 2 x bfloat>)
|
|
declare <vscale x 4 x bfloat> @llvm.round.nxv4bf16( <vscale x 4 x bfloat>)
|
|
declare <vscale x 8 x bfloat> @llvm.round.nxv8bf16( <vscale x 8 x bfloat>)
|
|
|
|
declare <vscale x 2 x bfloat> @llvm.roundeven.nxv2bf16( <vscale x 2 x bfloat>)
|
|
declare <vscale x 4 x bfloat> @llvm.roundeven.nxv4bf16( <vscale x 4 x bfloat>)
|
|
declare <vscale x 8 x bfloat> @llvm.roundeven.nxv8bf16( <vscale x 8 x bfloat>)
|
|
|
|
declare <vscale x 2 x bfloat> @llvm.trunc.nxv2bf16( <vscale x 2 x bfloat>)
|
|
declare <vscale x 4 x bfloat> @llvm.trunc.nxv4bf16( <vscale x 4 x bfloat>)
|
|
declare <vscale x 8 x bfloat> @llvm.trunc.nxv8bf16( <vscale x 8 x bfloat>)
|