shylie/llvm-project
1
0
Fork 0
Code Issues Pull Requests Actions 6 Packages Projects Releases Wiki Activity
llvm-project/llvm/test/CodeGen/RISCV/rvv/fmaximum-sdnode.ll
Pengcheng Wang 01a15dca09
[RISCV] Set a barrier between mask producer and user of V0 (#114012) 
Here we add a scheduling mutation in pre-ra scheduling, which will
add an artificial dependency edge between mask producer and its
previous nearest instruction that uses V0 register.

This prevents the overlap of live intervals of mask registers and
as a consequence we can reduce some spills/moves.

From the test changes, we can see some improvements and also some
regressions (more vtype toggles).

Partially fixes #113489.
2024-11-29 13:25:53 +08:00

807 lines
33 KiB
LLVM
Raw Blame History

; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc -mtriple=riscv32 -mattr=+d,+zvfh,+zfbfmin,+zvfbfmin,+v -target-abi=ilp32d \
; RUN: -verify-machineinstrs < %s | FileCheck %s --check-prefixes=CHECK,ZVFH
; RUN: llc -mtriple=riscv32 -mattr=+d,+zvfh,+zfbfmin,+zvfbfmin,+v -target-abi=ilp32d \
; RUN: -verify-machineinstrs -early-live-intervals < %s | FileCheck %s --check-prefixes=CHECK,ZVFH
; RUN: llc -mtriple=riscv64 -mattr=+d,+zvfh,+zfbfmin,+zvfbfmin,+v -target-abi=lp64d \
; RUN: -verify-machineinstrs < %s | FileCheck %s --check-prefixes=CHECK,ZVFH
; RUN: llc -mtriple=riscv64 -mattr=+d,+zvfh,+zfbfmin,+zvfbfmin,+v -target-abi=lp64d \
; RUN: -verify-machineinstrs -early-live-intervals < %s | FileCheck %s --check-prefixes=CHECK,ZVFH
; RUN: llc -mtriple=riscv32 -mattr=+d,+zvfhmin,+zfbfmin,+zvfbfmin,+v,+m -target-abi=ilp32d \
; RUN: -verify-machineinstrs < %s | FileCheck %s --check-prefixes=CHECK,ZVFHMIN
; RUN: llc -mtriple=riscv32 -mattr=+d,+zvfhmin,+zfbfmin,+zvfbfmin,+v,+m -target-abi=ilp32d \
; RUN: -verify-machineinstrs -early-live-intervals < %s | FileCheck %s --check-prefixes=CHECK,ZVFHMIN
; RUN: llc -mtriple=riscv64 -mattr=+d,+zvfhmin,+zfbfmin,+zvfbfmin,+v,+m -target-abi=lp64d \
; RUN: -verify-machineinstrs < %s | FileCheck %s --check-prefixes=CHECK,ZVFHMIN
; RUN: llc -mtriple=riscv64 -mattr=+d,+zvfhmin,+zfbfmin,+zvfbfmin,+v,+m -target-abi=lp64d \
; RUN: -verify-machineinstrs -early-live-intervals < %s | FileCheck %s --check-prefixes=CHECK,ZVFHMIN
declare <vscale x 1 x bfloat> @llvm.maximum.nxv1bf16(<vscale x 1 x bfloat>, <vscale x 1 x bfloat>)
define <vscale x 1 x bfloat> @vfmax_nxv1bf16_vv(<vscale x 1 x bfloat> %a, <vscale x 1 x bfloat> %b) {
; CHECK-LABEL: vfmax_nxv1bf16_vv:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetvli a0, zero, e16, mf4, ta, ma
; CHECK-NEXT: vfwcvtbf16.f.f.v v10, v8
; CHECK-NEXT: vsetvli zero, zero, e32, mf2, ta, ma
; CHECK-NEXT: vmfeq.vv v0, v10, v10
; CHECK-NEXT: vsetvli zero, zero, e16, mf4, ta, ma
; CHECK-NEXT: vfwcvtbf16.f.f.v v8, v9
; CHECK-NEXT: vsetvli zero, zero, e32, mf2, ta, ma
; CHECK-NEXT: vmerge.vvm v9, v10, v8, v0
; CHECK-NEXT: vmfeq.vv v0, v8, v8
; CHECK-NEXT: vmerge.vvm v8, v8, v10, v0
; CHECK-NEXT: vfmax.vv v9, v8, v9
; CHECK-NEXT: vsetvli zero, zero, e16, mf4, ta, ma
; CHECK-NEXT: vfncvtbf16.f.f.w v8, v9
; CHECK-NEXT: ret
%v = call <vscale x 1 x bfloat> @llvm.maximum.nxv1bf16(<vscale x 1 x bfloat> %a, <vscale x 1 x bfloat> %b)
ret <vscale x 1 x bfloat> %v
}
declare <vscale x 2 x bfloat> @llvm.maximum.nxv2bf16(<vscale x 2 x bfloat>, <vscale x 2 x bfloat>)
define <vscale x 2 x bfloat> @vfmax_nxv2bf16_vv(<vscale x 2 x bfloat> %a, <vscale x 2 x bfloat> %b) {
; CHECK-LABEL: vfmax_nxv2bf16_vv:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetvli a0, zero, e16, mf2, ta, ma
; CHECK-NEXT: vfwcvtbf16.f.f.v v10, v8
; CHECK-NEXT: vsetvli zero, zero, e32, m1, ta, ma
; CHECK-NEXT: vmfeq.vv v0, v10, v10
; CHECK-NEXT: vsetvli zero, zero, e16, mf2, ta, ma
; CHECK-NEXT: vfwcvtbf16.f.f.v v8, v9
; CHECK-NEXT: vsetvli zero, zero, e32, m1, ta, ma
; CHECK-NEXT: vmerge.vvm v9, v10, v8, v0
; CHECK-NEXT: vmfeq.vv v0, v8, v8
; CHECK-NEXT: vmerge.vvm v8, v8, v10, v0
; CHECK-NEXT: vfmax.vv v9, v8, v9
; CHECK-NEXT: vsetvli zero, zero, e16, mf2, ta, ma
; CHECK-NEXT: vfncvtbf16.f.f.w v8, v9
; CHECK-NEXT: ret
%v = call <vscale x 2 x bfloat> @llvm.maximum.nxv2bf16(<vscale x 2 x bfloat> %a, <vscale x 2 x bfloat> %b)
ret <vscale x 2 x bfloat> %v
}
declare <vscale x 4 x bfloat> @llvm.maximum.nxv4bf16(<vscale x 4 x bfloat>, <vscale x 4 x bfloat>)
define <vscale x 4 x bfloat> @vfmax_nxv4bf16_vv(<vscale x 4 x bfloat> %a, <vscale x 4 x bfloat> %b) {
; CHECK-LABEL: vfmax_nxv4bf16_vv:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetvli a0, zero, e16, m1, ta, ma
; CHECK-NEXT: vfwcvtbf16.f.f.v v10, v8
; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; CHECK-NEXT: vmfeq.vv v0, v10, v10
; CHECK-NEXT: vsetvli zero, zero, e16, m1, ta, ma
; CHECK-NEXT: vfwcvtbf16.f.f.v v12, v9
; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; CHECK-NEXT: vmerge.vvm v8, v10, v12, v0
; CHECK-NEXT: vmfeq.vv v0, v12, v12
; CHECK-NEXT: vmerge.vvm v10, v12, v10, v0
; CHECK-NEXT: vfmax.vv v10, v10, v8
; CHECK-NEXT: vsetvli zero, zero, e16, m1, ta, ma
; CHECK-NEXT: vfncvtbf16.f.f.w v8, v10
; CHECK-NEXT: ret
%v = call <vscale x 4 x bfloat> @llvm.maximum.nxv4bf16(<vscale x 4 x bfloat> %a, <vscale x 4 x bfloat> %b)
ret <vscale x 4 x bfloat> %v
}
declare <vscale x 8 x bfloat> @llvm.maximum.nxv8bf16(<vscale x 8 x bfloat>, <vscale x 8 x bfloat>)
define <vscale x 8 x bfloat> @vfmax_nxv8bf16_vv(<vscale x 8 x bfloat> %a, <vscale x 8 x bfloat> %b) {
; CHECK-LABEL: vfmax_nxv8bf16_vv:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetvli a0, zero, e16, m2, ta, ma
; CHECK-NEXT: vfwcvtbf16.f.f.v v12, v8
; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; CHECK-NEXT: vmfeq.vv v0, v12, v12
; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v10
; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; CHECK-NEXT: vmerge.vvm v8, v12, v16, v0
; CHECK-NEXT: vmfeq.vv v0, v16, v16
; CHECK-NEXT: vmerge.vvm v12, v16, v12, v0
; CHECK-NEXT: vfmax.vv v12, v12, v8
; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; CHECK-NEXT: vfncvtbf16.f.f.w v8, v12
; CHECK-NEXT: ret
%v = call <vscale x 8 x bfloat> @llvm.maximum.nxv8bf16(<vscale x 8 x bfloat> %a, <vscale x 8 x bfloat> %b)
ret <vscale x 8 x bfloat> %v
}
declare <vscale x 16 x bfloat> @llvm.maximum.nxv16bf16(<vscale x 16 x bfloat>, <vscale x 16 x bfloat>)
define <vscale x 16 x bfloat> @vfmax_nxv16bf16_vv(<vscale x 16 x bfloat> %a, <vscale x 16 x bfloat> %b) {
; CHECK-LABEL: vfmax_nxv16bf16_vv:
; CHECK: # %bb.0:
; CHECK-NEXT: addi sp, sp, -16
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: sub sp, sp, a0
; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
; CHECK-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v12
; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v8
; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; CHECK-NEXT: vmfeq.vv v0, v24, v24
; CHECK-NEXT: vmfeq.vv v7, v16, v16
; CHECK-NEXT: vmerge.vvm v8, v24, v16, v0
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vmerge.vvm v8, v16, v24, v0
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vfmax.vv v16, v8, v16
; CHECK-NEXT: vsetvli zero, zero, e16, m4, ta, ma
; CHECK-NEXT: vfncvtbf16.f.f.w v8, v16
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add sp, sp, a0
; CHECK-NEXT: .cfi_def_cfa sp, 16
; CHECK-NEXT: addi sp, sp, 16
; CHECK-NEXT: .cfi_def_cfa_offset 0
; CHECK-NEXT: ret
%v = call <vscale x 16 x bfloat> @llvm.maximum.nxv16bf16(<vscale x 16 x bfloat> %a, <vscale x 16 x bfloat> %b)
ret <vscale x 16 x bfloat> %v
}
declare <vscale x 32 x bfloat> @llvm.maximum.nxv32bf16(<vscale x 32 x bfloat>, <vscale x 32 x bfloat>)
define <vscale x 32 x bfloat> @vfmax_nxv32bf16_vv(<vscale x 32 x bfloat> %a, <vscale x 32 x bfloat> %b) nounwind {
; ZVFH-LABEL: vfmax_nxv32bf16_vv:
; ZVFH: # %bb.0:
; ZVFH-NEXT: addi sp, sp, -16
; ZVFH-NEXT: csrr a0, vlenb
; ZVFH-NEXT: slli a0, a0, 3
; ZVFH-NEXT: mv a1, a0
; ZVFH-NEXT: slli a0, a0, 1
; ZVFH-NEXT: add a0, a0, a1
; ZVFH-NEXT: sub sp, sp, a0
; ZVFH-NEXT: vmv8r.v v24, v16
; ZVFH-NEXT: csrr a0, vlenb
; ZVFH-NEXT: slli a0, a0, 3
; ZVFH-NEXT: add a0, sp, a0
; ZVFH-NEXT: addi a0, a0, 16
; ZVFH-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
; ZVFH-NEXT: vmv8r.v v0, v8
; ZVFH-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; ZVFH-NEXT: vfwcvtbf16.f.f.v v16, v24
; ZVFH-NEXT: vfwcvtbf16.f.f.v v8, v0
; ZVFH-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFH-NEXT: vmfeq.vv v0, v8, v8
; ZVFH-NEXT: vmfeq.vv v3, v16, v16
; ZVFH-NEXT: vmerge.vvm v24, v8, v16, v0
; ZVFH-NEXT: csrr a0, vlenb
; ZVFH-NEXT: slli a0, a0, 4
; ZVFH-NEXT: add a0, sp, a0
; ZVFH-NEXT: addi a0, a0, 16
; ZVFH-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; ZVFH-NEXT: vmv1r.v v0, v3
; ZVFH-NEXT: vmerge.vvm v8, v16, v8, v0
; ZVFH-NEXT: addi a0, sp, 16
; ZVFH-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; ZVFH-NEXT: csrr a0, vlenb
; ZVFH-NEXT: slli a0, a0, 3
; ZVFH-NEXT: add a0, sp, a0
; ZVFH-NEXT: addi a0, a0, 16
; ZVFH-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; ZVFH-NEXT: vsetvli zero, zero, e16, m4, ta, ma
; ZVFH-NEXT: vfwcvtbf16.f.f.v v24, v12
; ZVFH-NEXT: vfwcvtbf16.f.f.v v8, v4
; ZVFH-NEXT: csrr a0, vlenb
; ZVFH-NEXT: slli a0, a0, 4
; ZVFH-NEXT: add a0, sp, a0
; ZVFH-NEXT: addi a0, a0, 16
; ZVFH-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; ZVFH-NEXT: addi a0, sp, 16
; ZVFH-NEXT: vl8r.v v0, (a0) # Unknown-size Folded Reload
; ZVFH-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFH-NEXT: vfmax.vv v16, v0, v16
; ZVFH-NEXT: csrr a0, vlenb
; ZVFH-NEXT: slli a0, a0, 4
; ZVFH-NEXT: add a0, sp, a0
; ZVFH-NEXT: addi a0, a0, 16
; ZVFH-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
; ZVFH-NEXT: vmfeq.vv v0, v8, v8
; ZVFH-NEXT: vmfeq.vv v7, v24, v24
; ZVFH-NEXT: vmerge.vvm v16, v8, v24, v0
; ZVFH-NEXT: vmv1r.v v0, v7
; ZVFH-NEXT: vmerge.vvm v8, v24, v8, v0
; ZVFH-NEXT: vfmax.vv v16, v8, v16
; ZVFH-NEXT: csrr a0, vlenb
; ZVFH-NEXT: slli a0, a0, 4
; ZVFH-NEXT: add a0, sp, a0
; ZVFH-NEXT: addi a0, a0, 16
; ZVFH-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
; ZVFH-NEXT: vsetvli zero, zero, e16, m4, ta, ma
; ZVFH-NEXT: vfncvtbf16.f.f.w v8, v24
; ZVFH-NEXT: vfncvtbf16.f.f.w v12, v16
; ZVFH-NEXT: csrr a0, vlenb
; ZVFH-NEXT: slli a0, a0, 3
; ZVFH-NEXT: mv a1, a0
; ZVFH-NEXT: slli a0, a0, 1
; ZVFH-NEXT: add a0, a0, a1
; ZVFH-NEXT: add sp, sp, a0
; ZVFH-NEXT: addi sp, sp, 16
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmax_nxv32bf16_vv:
; ZVFHMIN: # %bb.0:
; ZVFHMIN-NEXT: addi sp, sp, -16
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: li a1, 24
; ZVFHMIN-NEXT: mul a0, a0, a1
; ZVFHMIN-NEXT: sub sp, sp, a0
; ZVFHMIN-NEXT: vmv8r.v v24, v16
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vmv8r.v v0, v8
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvtbf16.f.f.v v16, v24
; ZVFHMIN-NEXT: vfwcvtbf16.f.f.v v8, v0
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmfeq.vv v0, v8, v8
; ZVFHMIN-NEXT: vmfeq.vv v3, v16, v16
; ZVFHMIN-NEXT: vmerge.vvm v24, v8, v16, v0
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 4
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vmv1r.v v0, v3
; ZVFHMIN-NEXT: vmerge.vvm v8, v16, v8, v0
; ZVFHMIN-NEXT: addi a0, sp, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvtbf16.f.f.v v24, v12
; ZVFHMIN-NEXT: vfwcvtbf16.f.f.v v8, v4
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 4
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: addi a0, sp, 16
; ZVFHMIN-NEXT: vl8r.v v0, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfmax.vv v16, v0, v16
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 4
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vmfeq.vv v0, v8, v8
; ZVFHMIN-NEXT: vmfeq.vv v7, v24, v24
; ZVFHMIN-NEXT: vmerge.vvm v16, v8, v24, v0
; ZVFHMIN-NEXT: vmv1r.v v0, v7
; ZVFHMIN-NEXT: vmerge.vvm v8, v24, v8, v0
; ZVFHMIN-NEXT: vfmax.vv v16, v8, v16
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 4
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfncvtbf16.f.f.w v8, v24
; ZVFHMIN-NEXT: vfncvtbf16.f.f.w v12, v16
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: li a1, 24
; ZVFHMIN-NEXT: mul a0, a0, a1
; ZVFHMIN-NEXT: add sp, sp, a0
; ZVFHMIN-NEXT: addi sp, sp, 16
; ZVFHMIN-NEXT: ret
%v = call <vscale x 32 x bfloat> @llvm.maximum.nxv32bf16(<vscale x 32 x bfloat> %a, <vscale x 32 x bfloat> %b)
ret <vscale x 32 x bfloat> %v
}
declare <vscale x 1 x half> @llvm.maximum.nxv1f16(<vscale x 1 x half>, <vscale x 1 x half>)
define <vscale x 1 x half> @vfmax_nxv1f16_vv(<vscale x 1 x half> %a, <vscale x 1 x half> %b) {
; ZVFH-LABEL: vfmax_nxv1f16_vv:
; ZVFH: # %bb.0:
; ZVFH-NEXT: vsetvli a0, zero, e16, mf4, ta, ma
; ZVFH-NEXT: vmfeq.vv v0, v8, v8
; ZVFH-NEXT: vmerge.vvm v10, v8, v9, v0
; ZVFH-NEXT: vmfeq.vv v0, v9, v9
; ZVFH-NEXT: vmerge.vvm v8, v9, v8, v0
; ZVFH-NEXT: vfmax.vv v8, v8, v10
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmax_nxv1f16_vv:
; ZVFHMIN: # %bb.0:
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, mf4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v10, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, mf2, ta, ma
; ZVFHMIN-NEXT: vmfeq.vv v0, v10, v10
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, mf4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v9
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, mf2, ta, ma
; ZVFHMIN-NEXT: vmerge.vvm v9, v10, v8, v0
; ZVFHMIN-NEXT: vmfeq.vv v0, v8, v8
; ZVFHMIN-NEXT: vmerge.vvm v8, v8, v10, v0
; ZVFHMIN-NEXT: vfmax.vv v9, v8, v9
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, mf4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v9
; ZVFHMIN-NEXT: ret
%v = call <vscale x 1 x half> @llvm.maximum.nxv1f16(<vscale x 1 x half> %a, <vscale x 1 x half> %b)
ret <vscale x 1 x half> %v
}
declare <vscale x 2 x half> @llvm.maximum.nxv2f16(<vscale x 2 x half>, <vscale x 2 x half>)
define <vscale x 2 x half> @vfmax_nxv2f16_vv(<vscale x 2 x half> %a, <vscale x 2 x half> %b) {
; ZVFH-LABEL: vfmax_nxv2f16_vv:
; ZVFH: # %bb.0:
; ZVFH-NEXT: vsetvli a0, zero, e16, mf2, ta, ma
; ZVFH-NEXT: vmfeq.vv v0, v8, v8
; ZVFH-NEXT: vmerge.vvm v10, v8, v9, v0
; ZVFH-NEXT: vmfeq.vv v0, v9, v9
; ZVFH-NEXT: vmerge.vvm v8, v9, v8, v0
; ZVFH-NEXT: vfmax.vv v8, v8, v10
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmax_nxv2f16_vv:
; ZVFHMIN: # %bb.0:
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, mf2, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v10, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m1, ta, ma
; ZVFHMIN-NEXT: vmfeq.vv v0, v10, v10
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, mf2, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v9
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m1, ta, ma
; ZVFHMIN-NEXT: vmerge.vvm v9, v10, v8, v0
; ZVFHMIN-NEXT: vmfeq.vv v0, v8, v8
; ZVFHMIN-NEXT: vmerge.vvm v8, v8, v10, v0
; ZVFHMIN-NEXT: vfmax.vv v9, v8, v9
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, mf2, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v9
; ZVFHMIN-NEXT: ret
%v = call <vscale x 2 x half> @llvm.maximum.nxv2f16(<vscale x 2 x half> %a, <vscale x 2 x half> %b)
ret <vscale x 2 x half> %v
}
declare <vscale x 4 x half> @llvm.maximum.nxv4f16(<vscale x 4 x half>, <vscale x 4 x half>)
define <vscale x 4 x half> @vfmax_nxv4f16_vv(<vscale x 4 x half> %a, <vscale x 4 x half> %b) {
; ZVFH-LABEL: vfmax_nxv4f16_vv:
; ZVFH: # %bb.0:
; ZVFH-NEXT: vsetvli a0, zero, e16, m1, ta, ma
; ZVFH-NEXT: vmfeq.vv v0, v8, v8
; ZVFH-NEXT: vmerge.vvm v10, v8, v9, v0
; ZVFH-NEXT: vmfeq.vv v0, v9, v9
; ZVFH-NEXT: vmerge.vvm v8, v9, v8, v0
; ZVFH-NEXT: vfmax.vv v8, v8, v10
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmax_nxv4f16_vv:
; ZVFHMIN: # %bb.0:
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m1, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v10, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; ZVFHMIN-NEXT: vmfeq.vv v0, v10, v10
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m1, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v12, v9
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; ZVFHMIN-NEXT: vmerge.vvm v8, v10, v12, v0
; ZVFHMIN-NEXT: vmfeq.vv v0, v12, v12
; ZVFHMIN-NEXT: vmerge.vvm v10, v12, v10, v0
; ZVFHMIN-NEXT: vfmax.vv v10, v10, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m1, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v10
; ZVFHMIN-NEXT: ret
%v = call <vscale x 4 x half> @llvm.maximum.nxv4f16(<vscale x 4 x half> %a, <vscale x 4 x half> %b)
ret <vscale x 4 x half> %v
}
declare <vscale x 8 x half> @llvm.maximum.nxv8f16(<vscale x 8 x half>, <vscale x 8 x half>)
define <vscale x 8 x half> @vfmax_nxv8f16_vv(<vscale x 8 x half> %a, <vscale x 8 x half> %b) {
; ZVFH-LABEL: vfmax_nxv8f16_vv:
; ZVFH: # %bb.0:
; ZVFH-NEXT: vsetvli a0, zero, e16, m2, ta, ma
; ZVFH-NEXT: vmfeq.vv v0, v8, v8
; ZVFH-NEXT: vmerge.vvm v12, v8, v10, v0
; ZVFH-NEXT: vmfeq.vv v0, v10, v10
; ZVFH-NEXT: vmerge.vvm v8, v10, v8, v0
; ZVFH-NEXT: vfmax.vv v8, v8, v12
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmax_nxv8f16_vv:
; ZVFHMIN: # %bb.0:
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m2, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v12, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; ZVFHMIN-NEXT: vmfeq.vv v0, v12, v12
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v10
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; ZVFHMIN-NEXT: vmerge.vvm v8, v12, v16, v0
; ZVFHMIN-NEXT: vmfeq.vv v0, v16, v16
; ZVFHMIN-NEXT: vmerge.vvm v12, v16, v12, v0
; ZVFHMIN-NEXT: vfmax.vv v12, v12, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v12
; ZVFHMIN-NEXT: ret
%v = call <vscale x 8 x half> @llvm.maximum.nxv8f16(<vscale x 8 x half> %a, <vscale x 8 x half> %b)
ret <vscale x 8 x half> %v
}
declare <vscale x 16 x half> @llvm.maximum.nxv16f16(<vscale x 16 x half>, <vscale x 16 x half>)
define <vscale x 16 x half> @vfmax_nxv16f16_vv(<vscale x 16 x half> %a, <vscale x 16 x half> %b) {
; ZVFH-LABEL: vfmax_nxv16f16_vv:
; ZVFH: # %bb.0:
; ZVFH-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; ZVFH-NEXT: vmfeq.vv v0, v8, v8
; ZVFH-NEXT: vmerge.vvm v16, v8, v12, v0
; ZVFH-NEXT: vmfeq.vv v0, v12, v12
; ZVFH-NEXT: vmerge.vvm v8, v12, v8, v0
; ZVFH-NEXT: vfmax.vv v8, v8, v16
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmax_nxv16f16_vv:
; ZVFHMIN: # %bb.0:
; ZVFHMIN-NEXT: addi sp, sp, -16
; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: sub sp, sp, a0
; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmfeq.vv v0, v24, v24
; ZVFHMIN-NEXT: vmfeq.vv v7, v16, v16
; ZVFHMIN-NEXT: vmerge.vvm v8, v24, v16, v0
; ZVFHMIN-NEXT: addi a0, sp, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vmv1r.v v0, v7
; ZVFHMIN-NEXT: vmerge.vvm v8, v16, v24, v0
; ZVFHMIN-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfmax.vv v16, v8, v16
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v16
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: add sp, sp, a0
; ZVFHMIN-NEXT: .cfi_def_cfa sp, 16
; ZVFHMIN-NEXT: addi sp, sp, 16
; ZVFHMIN-NEXT: .cfi_def_cfa_offset 0
; ZVFHMIN-NEXT: ret
%v = call <vscale x 16 x half> @llvm.maximum.nxv16f16(<vscale x 16 x half> %a, <vscale x 16 x half> %b)
ret <vscale x 16 x half> %v
}
declare <vscale x 32 x half> @llvm.maximum.nxv32f16(<vscale x 32 x half>, <vscale x 32 x half>)
define <vscale x 32 x half> @vfmax_nxv32f16_vv(<vscale x 32 x half> %a, <vscale x 32 x half> %b) nounwind {
; ZVFH-LABEL: vfmax_nxv32f16_vv:
; ZVFH: # %bb.0:
; ZVFH-NEXT: vsetvli a0, zero, e16, m8, ta, ma
; ZVFH-NEXT: vmfeq.vv v0, v8, v8
; ZVFH-NEXT: vmfeq.vv v7, v16, v16
; ZVFH-NEXT: vmerge.vvm v24, v8, v16, v0
; ZVFH-NEXT: vmv1r.v v0, v7
; ZVFH-NEXT: vmerge.vvm v8, v16, v8, v0
; ZVFH-NEXT: vfmax.vv v8, v8, v24
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmax_nxv32f16_vv:
; ZVFHMIN: # %bb.0:
; ZVFHMIN-NEXT: addi sp, sp, -16
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: li a1, 24
; ZVFHMIN-NEXT: mul a0, a0, a1
; ZVFHMIN-NEXT: sub sp, sp, a0
; ZVFHMIN-NEXT: vmv8r.v v24, v16
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vmv8r.v v0, v8
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v24
; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v0
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmfeq.vv v0, v8, v8
; ZVFHMIN-NEXT: vmfeq.vv v3, v16, v16
; ZVFHMIN-NEXT: vmerge.vvm v24, v8, v16, v0
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 4
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vmv1r.v v0, v3
; ZVFHMIN-NEXT: vmerge.vvm v8, v16, v8, v0
; ZVFHMIN-NEXT: addi a0, sp, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v4
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 4
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: addi a0, sp, 16
; ZVFHMIN-NEXT: vl8r.v v0, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfmax.vv v16, v0, v16
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 4
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vmfeq.vv v0, v8, v8
; ZVFHMIN-NEXT: vmfeq.vv v7, v24, v24
; ZVFHMIN-NEXT: vmerge.vvm v16, v8, v24, v0
; ZVFHMIN-NEXT: vmv1r.v v0, v7
; ZVFHMIN-NEXT: vmerge.vvm v8, v24, v8, v0
; ZVFHMIN-NEXT: vfmax.vv v16, v8, v16
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 4
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v24
; ZVFHMIN-NEXT: vfncvt.f.f.w v12, v16
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: li a1, 24
; ZVFHMIN-NEXT: mul a0, a0, a1
; ZVFHMIN-NEXT: add sp, sp, a0
; ZVFHMIN-NEXT: addi sp, sp, 16
; ZVFHMIN-NEXT: ret
%v = call <vscale x 32 x half> @llvm.maximum.nxv32f16(<vscale x 32 x half> %a, <vscale x 32 x half> %b)
ret <vscale x 32 x half> %v
}
declare <vscale x 1 x float> @llvm.maximum.nxv1f32(<vscale x 1 x float>, <vscale x 1 x float>)
define <vscale x 1 x float> @vfmax_nxv1f32_vv(<vscale x 1 x float> %a, <vscale x 1 x float> %b) {
; CHECK-LABEL: vfmax_nxv1f32_vv:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetvli a0, zero, e32, mf2, ta, ma
; CHECK-NEXT: vmfeq.vv v0, v8, v8
; CHECK-NEXT: vmerge.vvm v10, v8, v9, v0
; CHECK-NEXT: vmfeq.vv v0, v9, v9
; CHECK-NEXT: vmerge.vvm v8, v9, v8, v0
; CHECK-NEXT: vfmax.vv v8, v8, v10
; CHECK-NEXT: ret
%v = call <vscale x 1 x float> @llvm.maximum.nxv1f32(<vscale x 1 x float> %a, <vscale x 1 x float> %b)
ret <vscale x 1 x float> %v
}
declare <vscale x 2 x float> @llvm.maximum.nxv2f32(<vscale x 2 x float>, <vscale x 2 x float>)
define <vscale x 2 x float> @vfmax_nxv2f32_vv(<vscale x 2 x float> %a, <vscale x 2 x float> %b) {
; CHECK-LABEL: vfmax_nxv2f32_vv:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetvli a0, zero, e32, m1, ta, ma
; CHECK-NEXT: vmfeq.vv v0, v8, v8
; CHECK-NEXT: vmerge.vvm v10, v8, v9, v0
; CHECK-NEXT: vmfeq.vv v0, v9, v9
; CHECK-NEXT: vmerge.vvm v8, v9, v8, v0
; CHECK-NEXT: vfmax.vv v8, v8, v10
; CHECK-NEXT: ret
%v = call <vscale x 2 x float> @llvm.maximum.nxv2f32(<vscale x 2 x float> %a, <vscale x 2 x float> %b)
ret <vscale x 2 x float> %v
}
declare <vscale x 4 x float> @llvm.maximum.nxv4f32(<vscale x 4 x float>, <vscale x 4 x float>)
define <vscale x 4 x float> @vfmax_nxv4f32_vv(<vscale x 4 x float> %a, <vscale x 4 x float> %b) {
; CHECK-LABEL: vfmax_nxv4f32_vv:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetvli a0, zero, e32, m2, ta, ma
; CHECK-NEXT: vmfeq.vv v0, v8, v8
; CHECK-NEXT: vmerge.vvm v12, v8, v10, v0
; CHECK-NEXT: vmfeq.vv v0, v10, v10
; CHECK-NEXT: vmerge.vvm v8, v10, v8, v0
; CHECK-NEXT: vfmax.vv v8, v8, v12
; CHECK-NEXT: ret
%v = call <vscale x 4 x float> @llvm.maximum.nxv4f32(<vscale x 4 x float> %a, <vscale x 4 x float> %b)
ret <vscale x 4 x float> %v
}
declare <vscale x 8 x float> @llvm.maximum.nxv8f32(<vscale x 8 x float>, <vscale x 8 x float>)
define <vscale x 8 x float> @vfmax_nxv8f32_vv(<vscale x 8 x float> %a, <vscale x 8 x float> %b) {
; CHECK-LABEL: vfmax_nxv8f32_vv:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetvli a0, zero, e32, m4, ta, ma
; CHECK-NEXT: vmfeq.vv v0, v8, v8
; CHECK-NEXT: vmerge.vvm v16, v8, v12, v0
; CHECK-NEXT: vmfeq.vv v0, v12, v12
; CHECK-NEXT: vmerge.vvm v8, v12, v8, v0
; CHECK-NEXT: vfmax.vv v8, v8, v16
; CHECK-NEXT: ret
%v = call <vscale x 8 x float> @llvm.maximum.nxv8f32(<vscale x 8 x float> %a, <vscale x 8 x float> %b)
ret <vscale x 8 x float> %v
}
declare <vscale x 16 x float> @llvm.maximum.nxv16f32(<vscale x 16 x float>, <vscale x 16 x float>)
define <vscale x 16 x float> @vfmax_nxv16f32_vv(<vscale x 16 x float> %a, <vscale x 16 x float> %b) nounwind {
; CHECK-LABEL: vfmax_nxv16f32_vv:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetvli a0, zero, e32, m8, ta, ma
; CHECK-NEXT: vmfeq.vv v0, v8, v8
; CHECK-NEXT: vmfeq.vv v7, v16, v16
; CHECK-NEXT: vmerge.vvm v24, v8, v16, v0
; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vmerge.vvm v8, v16, v8, v0
; CHECK-NEXT: vfmax.vv v8, v8, v24
; CHECK-NEXT: ret
%v = call <vscale x 16 x float> @llvm.maximum.nxv16f32(<vscale x 16 x float> %a, <vscale x 16 x float> %b)
ret <vscale x 16 x float> %v
}
declare <vscale x 1 x double> @llvm.maximum.nxv1f64(<vscale x 1 x double>, <vscale x 1 x double>)
define <vscale x 1 x double> @vfmax_nxv1f64_vv(<vscale x 1 x double> %a, <vscale x 1 x double> %b) {
; CHECK-LABEL: vfmax_nxv1f64_vv:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetvli a0, zero, e64, m1, ta, ma
; CHECK-NEXT: vmfeq.vv v0, v8, v8
; CHECK-NEXT: vmerge.vvm v10, v8, v9, v0
; CHECK-NEXT: vmfeq.vv v0, v9, v9
; CHECK-NEXT: vmerge.vvm v8, v9, v8, v0
; CHECK-NEXT: vfmax.vv v8, v8, v10
; CHECK-NEXT: ret
%v = call <vscale x 1 x double> @llvm.maximum.nxv1f64(<vscale x 1 x double> %a, <vscale x 1 x double> %b)
ret <vscale x 1 x double> %v
}
declare <vscale x 2 x double> @llvm.maximum.nxv2f64(<vscale x 2 x double>, <vscale x 2 x double>)
define <vscale x 2 x double> @vfmax_nxv2f64_vv(<vscale x 2 x double> %a, <vscale x 2 x double> %b) {
; CHECK-LABEL: vfmax_nxv2f64_vv:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetvli a0, zero, e64, m2, ta, ma
; CHECK-NEXT: vmfeq.vv v0, v8, v8
; CHECK-NEXT: vmerge.vvm v12, v8, v10, v0
; CHECK-NEXT: vmfeq.vv v0, v10, v10
; CHECK-NEXT: vmerge.vvm v8, v10, v8, v0
; CHECK-NEXT: vfmax.vv v8, v8, v12
; CHECK-NEXT: ret
%v = call <vscale x 2 x double> @llvm.maximum.nxv2f64(<vscale x 2 x double> %a, <vscale x 2 x double> %b)
ret <vscale x 2 x double> %v
}
declare <vscale x 4 x double> @llvm.maximum.nxv4f64(<vscale x 4 x double>, <vscale x 4 x double>)
define <vscale x 4 x double> @vfmax_nxv4f64_vv(<vscale x 4 x double> %a, <vscale x 4 x double> %b) {
; CHECK-LABEL: vfmax_nxv4f64_vv:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetvli a0, zero, e64, m4, ta, ma
; CHECK-NEXT: vmfeq.vv v0, v8, v8
; CHECK-NEXT: vmerge.vvm v16, v8, v12, v0
; CHECK-NEXT: vmfeq.vv v0, v12, v12
; CHECK-NEXT: vmerge.vvm v8, v12, v8, v0
; CHECK-NEXT: vfmax.vv v8, v8, v16
; CHECK-NEXT: ret
%v = call <vscale x 4 x double> @llvm.maximum.nxv4f64(<vscale x 4 x double> %a, <vscale x 4 x double> %b)
ret <vscale x 4 x double> %v
}
declare <vscale x 8 x double> @llvm.maximum.nxv8f64(<vscale x 8 x double>, <vscale x 8 x double>)
define <vscale x 8 x double> @vfmax_nxv8f64_vv(<vscale x 8 x double> %a, <vscale x 8 x double> %b) nounwind {
; CHECK-LABEL: vfmax_nxv8f64_vv:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetvli a0, zero, e64, m8, ta, ma
; CHECK-NEXT: vmfeq.vv v0, v8, v8
; CHECK-NEXT: vmfeq.vv v7, v16, v16
; CHECK-NEXT: vmerge.vvm v24, v8, v16, v0
; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vmerge.vvm v8, v16, v8, v0
; CHECK-NEXT: vfmax.vv v8, v8, v24
; CHECK-NEXT: ret
%v = call <vscale x 8 x double> @llvm.maximum.nxv8f64(<vscale x 8 x double> %a, <vscale x 8 x double> %b)
ret <vscale x 8 x double> %v
}
define <vscale x 1 x half> @vfmax_nxv1f16_vv_nnan(<vscale x 1 x half> %a, <vscale x 1 x half> %b) {
; ZVFH-LABEL: vfmax_nxv1f16_vv_nnan:
; ZVFH: # %bb.0:
; ZVFH-NEXT: vsetvli a0, zero, e16, mf4, ta, ma
; ZVFH-NEXT: vfmax.vv v8, v8, v9
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmax_nxv1f16_vv_nnan:
; ZVFHMIN: # %bb.0:
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, mf4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v10, v9
; ZVFHMIN-NEXT: vfwcvt.f.f.v v9, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, mf2, ta, ma
; ZVFHMIN-NEXT: vfmax.vv v9, v9, v10
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, mf4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v9
; ZVFHMIN-NEXT: ret
%v = call nnan <vscale x 1 x half> @llvm.maximum.nxv1f16(<vscale x 1 x half> %a, <vscale x 1 x half> %b)
ret <vscale x 1 x half> %v
}
define <vscale x 1 x half> @vfmax_nxv1f16_vv_nnana(<vscale x 1 x half> %a, <vscale x 1 x half> %b) {
; ZVFH-LABEL: vfmax_nxv1f16_vv_nnana:
; ZVFH: # %bb.0:
; ZVFH-NEXT: vsetvli a0, zero, e16, mf4, ta, mu
; ZVFH-NEXT: vmfeq.vv v0, v9, v9
; ZVFH-NEXT: vmv1r.v v10, v9
; ZVFH-NEXT: vfadd.vv v10, v8, v8, v0.t
; ZVFH-NEXT: vfmax.vv v8, v10, v9
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmax_nxv1f16_vv_nnana:
; ZVFHMIN: # %bb.0:
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, mf4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v10, v8
; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v9
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, mf2, ta, ma
; ZVFHMIN-NEXT: vfadd.vv v9, v10, v10
; ZVFHMIN-NEXT: vmfeq.vv v0, v8, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, mf4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v10, v9
; ZVFHMIN-NEXT: vfwcvt.f.f.v v9, v10
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, mf2, ta, ma
; ZVFHMIN-NEXT: vmerge.vvm v10, v8, v9, v0
; ZVFHMIN-NEXT: vmfeq.vv v0, v9, v9
; ZVFHMIN-NEXT: vmerge.vvm v8, v9, v8, v0
; ZVFHMIN-NEXT: vfmax.vv v9, v10, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, mf4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v9
; ZVFHMIN-NEXT: ret
%c = fadd nnan <vscale x 1 x half> %a, %a
%v = call <vscale x 1 x half> @llvm.maximum.nxv1f16(<vscale x 1 x half> %c, <vscale x 1 x half> %b)
ret <vscale x 1 x half> %v
}
define <vscale x 1 x half> @vfmax_nxv1f16_vv_nnanb(<vscale x 1 x half> %a, <vscale x 1 x half> %b) {
; ZVFH-LABEL: vfmax_nxv1f16_vv_nnanb:
; ZVFH: # %bb.0:
; ZVFH-NEXT: vsetvli a0, zero, e16, mf4, ta, mu
; ZVFH-NEXT: vmfeq.vv v0, v8, v8
; ZVFH-NEXT: vmv1r.v v10, v8
; ZVFH-NEXT: vfadd.vv v10, v9, v9, v0.t
; ZVFH-NEXT: vfmax.vv v8, v8, v10
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmax_nxv1f16_vv_nnanb:
; ZVFHMIN: # %bb.0:
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, mf4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v10, v9
; ZVFHMIN-NEXT: vfwcvt.f.f.v v9, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, mf2, ta, ma
; ZVFHMIN-NEXT: vfadd.vv v8, v10, v10
; ZVFHMIN-NEXT: vmfeq.vv v0, v9, v9
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, mf4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v10, v8
; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v10
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, mf2, ta, ma
; ZVFHMIN-NEXT: vmerge.vvm v10, v9, v8, v0
; ZVFHMIN-NEXT: vmfeq.vv v0, v8, v8
; ZVFHMIN-NEXT: vmerge.vvm v8, v8, v9, v0
; ZVFHMIN-NEXT: vfmax.vv v9, v8, v10
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, mf4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v9
; ZVFHMIN-NEXT: ret
%c = fadd nnan <vscale x 1 x half> %b, %b
%v = call <vscale x 1 x half> @llvm.maximum.nxv1f16(<vscale x 1 x half> %a, <vscale x 1 x half> %c)
ret <vscale x 1 x half> %v
}
Reference in New Issue View Git Blame Copy Permalink