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
767 lines
31 KiB
LLVM
767 lines
31 KiB
LLVM
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
|
|
; RUN: llc --mtriple=loongarch32 --mattr=+f,-d,-frecipe < %s | FileCheck %s --check-prefix=LA32F
|
|
; RUN: llc --mtriple=loongarch32 --mattr=+f,-d,+frecipe < %s | FileCheck %s --check-prefix=LA32F-FRECIPE
|
|
; RUN: llc --mtriple=loongarch64 --mattr=+d,-frecipe < %s | FileCheck %s --check-prefix=LA64D
|
|
; RUN: llc --mtriple=loongarch64 --mattr=+d,+frecipe < %s | FileCheck %s --check-prefix=LA64D-FRECIPE
|
|
|
|
|
|
declare float @llvm.sqrt.f32(float)
|
|
declare double @llvm.sqrt.f64(double)
|
|
|
|
define float @frsqrt_f32(float %a) nounwind {
|
|
; LA32F-LABEL: frsqrt_f32:
|
|
; LA32F: # %bb.0:
|
|
; LA32F-NEXT: frsqrt.s $fa0, $fa0
|
|
; LA32F-NEXT: ret
|
|
;
|
|
; LA32F-FRECIPE-LABEL: frsqrt_f32:
|
|
; LA32F-FRECIPE: # %bb.0:
|
|
; LA32F-FRECIPE-NEXT: frsqrte.s $fa1, $fa0
|
|
; LA32F-FRECIPE-NEXT: fmul.s $fa0, $fa0, $fa1
|
|
; LA32F-FRECIPE-NEXT: lu12i.w $a0, -261120
|
|
; LA32F-FRECIPE-NEXT: movgr2fr.w $fa2, $a0
|
|
; LA32F-FRECIPE-NEXT: fmadd.s $fa0, $fa0, $fa1, $fa2
|
|
; LA32F-FRECIPE-NEXT: lu12i.w $a0, -266240
|
|
; LA32F-FRECIPE-NEXT: movgr2fr.w $fa2, $a0
|
|
; LA32F-FRECIPE-NEXT: fmul.s $fa1, $fa1, $fa2
|
|
; LA32F-FRECIPE-NEXT: fmul.s $fa0, $fa1, $fa0
|
|
; LA32F-FRECIPE-NEXT: ret
|
|
;
|
|
; LA64D-LABEL: frsqrt_f32:
|
|
; LA64D: # %bb.0:
|
|
; LA64D-NEXT: frsqrt.s $fa0, $fa0
|
|
; LA64D-NEXT: ret
|
|
;
|
|
; LA64D-FRECIPE-LABEL: frsqrt_f32:
|
|
; LA64D-FRECIPE: # %bb.0:
|
|
; LA64D-FRECIPE-NEXT: frsqrte.s $fa1, $fa0
|
|
; LA64D-FRECIPE-NEXT: fmul.s $fa0, $fa0, $fa1
|
|
; LA64D-FRECIPE-NEXT: vldi $vr2, -1144
|
|
; LA64D-FRECIPE-NEXT: fmadd.s $fa0, $fa0, $fa1, $fa2
|
|
; LA64D-FRECIPE-NEXT: vldi $vr2, -1056
|
|
; LA64D-FRECIPE-NEXT: fmul.s $fa1, $fa1, $fa2
|
|
; LA64D-FRECIPE-NEXT: fmul.s $fa0, $fa1, $fa0
|
|
; LA64D-FRECIPE-NEXT: ret
|
|
|
|
%1 = call fast float @llvm.sqrt.f32(float %a)
|
|
%2 = fdiv fast float 1.0, %1
|
|
ret float %2
|
|
}
|
|
|
|
define double @frsqrt_f64(double %a) nounwind {
|
|
; LA32F-LABEL: frsqrt_f64:
|
|
; LA32F: # %bb.0:
|
|
; LA32F-NEXT: addi.w $sp, $sp, -16
|
|
; LA32F-NEXT: st.w $ra, $sp, 12 # 4-byte Folded Spill
|
|
; LA32F-NEXT: bl sqrt
|
|
; LA32F-NEXT: move $a2, $a0
|
|
; LA32F-NEXT: move $a3, $a1
|
|
; LA32F-NEXT: lu12i.w $a1, 261888
|
|
; LA32F-NEXT: move $a0, $zero
|
|
; LA32F-NEXT: bl __divdf3
|
|
; LA32F-NEXT: ld.w $ra, $sp, 12 # 4-byte Folded Reload
|
|
; LA32F-NEXT: addi.w $sp, $sp, 16
|
|
; LA32F-NEXT: ret
|
|
;
|
|
; LA32F-FRECIPE-LABEL: frsqrt_f64:
|
|
; LA32F-FRECIPE: # %bb.0:
|
|
; LA32F-FRECIPE-NEXT: addi.w $sp, $sp, -16
|
|
; LA32F-FRECIPE-NEXT: st.w $ra, $sp, 12 # 4-byte Folded Spill
|
|
; LA32F-FRECIPE-NEXT: bl sqrt
|
|
; LA32F-FRECIPE-NEXT: move $a2, $a0
|
|
; LA32F-FRECIPE-NEXT: move $a3, $a1
|
|
; LA32F-FRECIPE-NEXT: lu12i.w $a1, 261888
|
|
; LA32F-FRECIPE-NEXT: move $a0, $zero
|
|
; LA32F-FRECIPE-NEXT: bl __divdf3
|
|
; LA32F-FRECIPE-NEXT: ld.w $ra, $sp, 12 # 4-byte Folded Reload
|
|
; LA32F-FRECIPE-NEXT: addi.w $sp, $sp, 16
|
|
; LA32F-FRECIPE-NEXT: ret
|
|
;
|
|
; LA64D-LABEL: frsqrt_f64:
|
|
; LA64D: # %bb.0:
|
|
; LA64D-NEXT: frsqrt.d $fa0, $fa0
|
|
; LA64D-NEXT: ret
|
|
;
|
|
; LA64D-FRECIPE-LABEL: frsqrt_f64:
|
|
; LA64D-FRECIPE: # %bb.0:
|
|
; LA64D-FRECIPE-NEXT: frsqrte.d $fa1, $fa0
|
|
; LA64D-FRECIPE-NEXT: fmul.d $fa2, $fa0, $fa1
|
|
; LA64D-FRECIPE-NEXT: vldi $vr3, -888
|
|
; LA64D-FRECIPE-NEXT: fmadd.d $fa2, $fa2, $fa1, $fa3
|
|
; LA64D-FRECIPE-NEXT: vldi $vr4, -800
|
|
; LA64D-FRECIPE-NEXT: fmul.d $fa1, $fa1, $fa4
|
|
; LA64D-FRECIPE-NEXT: fmul.d $fa1, $fa1, $fa2
|
|
; LA64D-FRECIPE-NEXT: fmul.d $fa0, $fa0, $fa1
|
|
; LA64D-FRECIPE-NEXT: fmadd.d $fa0, $fa0, $fa1, $fa3
|
|
; LA64D-FRECIPE-NEXT: fmul.d $fa1, $fa1, $fa4
|
|
; LA64D-FRECIPE-NEXT: fmul.d $fa0, $fa1, $fa0
|
|
; LA64D-FRECIPE-NEXT: ret
|
|
%1 = call fast double @llvm.sqrt.f64(double %a)
|
|
%2 = fdiv fast double 1.0, %1
|
|
ret double %2
|
|
}
|
|
|
|
define double @sqrt_simplify_before_recip_3_uses_f64(double %x, ptr %p1, ptr %p2) nounwind {
|
|
; LA32F-LABEL: sqrt_simplify_before_recip_3_uses_f64:
|
|
; LA32F: # %bb.0:
|
|
; LA32F-NEXT: addi.w $sp, $sp, -32
|
|
; LA32F-NEXT: st.w $ra, $sp, 28 # 4-byte Folded Spill
|
|
; LA32F-NEXT: st.w $fp, $sp, 24 # 4-byte Folded Spill
|
|
; LA32F-NEXT: st.w $s0, $sp, 20 # 4-byte Folded Spill
|
|
; LA32F-NEXT: st.w $s1, $sp, 16 # 4-byte Folded Spill
|
|
; LA32F-NEXT: st.w $s2, $sp, 12 # 4-byte Folded Spill
|
|
; LA32F-NEXT: st.w $s3, $sp, 8 # 4-byte Folded Spill
|
|
; LA32F-NEXT: st.w $s4, $sp, 4 # 4-byte Folded Spill
|
|
; LA32F-NEXT: move $fp, $a3
|
|
; LA32F-NEXT: move $s0, $a2
|
|
; LA32F-NEXT: bl sqrt
|
|
; LA32F-NEXT: move $s1, $a0
|
|
; LA32F-NEXT: move $s2, $a1
|
|
; LA32F-NEXT: lu12i.w $a1, 261888
|
|
; LA32F-NEXT: move $a0, $zero
|
|
; LA32F-NEXT: move $a2, $s1
|
|
; LA32F-NEXT: move $a3, $s2
|
|
; LA32F-NEXT: bl __divdf3
|
|
; LA32F-NEXT: move $s3, $a0
|
|
; LA32F-NEXT: move $s4, $a1
|
|
; LA32F-NEXT: lu12i.w $a1, 263248
|
|
; LA32F-NEXT: move $a0, $zero
|
|
; LA32F-NEXT: move $a2, $s1
|
|
; LA32F-NEXT: move $a3, $s2
|
|
; LA32F-NEXT: bl __divdf3
|
|
; LA32F-NEXT: st.w $s3, $s0, 0
|
|
; LA32F-NEXT: st.w $s4, $s0, 4
|
|
; LA32F-NEXT: st.w $a0, $fp, 0
|
|
; LA32F-NEXT: st.w $a1, $fp, 4
|
|
; LA32F-NEXT: move $a0, $s1
|
|
; LA32F-NEXT: move $a1, $s2
|
|
; LA32F-NEXT: ld.w $s4, $sp, 4 # 4-byte Folded Reload
|
|
; LA32F-NEXT: ld.w $s3, $sp, 8 # 4-byte Folded Reload
|
|
; LA32F-NEXT: ld.w $s2, $sp, 12 # 4-byte Folded Reload
|
|
; LA32F-NEXT: ld.w $s1, $sp, 16 # 4-byte Folded Reload
|
|
; LA32F-NEXT: ld.w $s0, $sp, 20 # 4-byte Folded Reload
|
|
; LA32F-NEXT: ld.w $fp, $sp, 24 # 4-byte Folded Reload
|
|
; LA32F-NEXT: ld.w $ra, $sp, 28 # 4-byte Folded Reload
|
|
; LA32F-NEXT: addi.w $sp, $sp, 32
|
|
; LA32F-NEXT: ret
|
|
;
|
|
; LA32F-FRECIPE-LABEL: sqrt_simplify_before_recip_3_uses_f64:
|
|
; LA32F-FRECIPE: # %bb.0:
|
|
; LA32F-FRECIPE-NEXT: addi.w $sp, $sp, -32
|
|
; LA32F-FRECIPE-NEXT: st.w $ra, $sp, 28 # 4-byte Folded Spill
|
|
; LA32F-FRECIPE-NEXT: st.w $fp, $sp, 24 # 4-byte Folded Spill
|
|
; LA32F-FRECIPE-NEXT: st.w $s0, $sp, 20 # 4-byte Folded Spill
|
|
; LA32F-FRECIPE-NEXT: st.w $s1, $sp, 16 # 4-byte Folded Spill
|
|
; LA32F-FRECIPE-NEXT: st.w $s2, $sp, 12 # 4-byte Folded Spill
|
|
; LA32F-FRECIPE-NEXT: st.w $s3, $sp, 8 # 4-byte Folded Spill
|
|
; LA32F-FRECIPE-NEXT: st.w $s4, $sp, 4 # 4-byte Folded Spill
|
|
; LA32F-FRECIPE-NEXT: move $fp, $a3
|
|
; LA32F-FRECIPE-NEXT: move $s0, $a2
|
|
; LA32F-FRECIPE-NEXT: bl sqrt
|
|
; LA32F-FRECIPE-NEXT: move $s1, $a0
|
|
; LA32F-FRECIPE-NEXT: move $s2, $a1
|
|
; LA32F-FRECIPE-NEXT: lu12i.w $a1, 261888
|
|
; LA32F-FRECIPE-NEXT: move $a0, $zero
|
|
; LA32F-FRECIPE-NEXT: move $a2, $s1
|
|
; LA32F-FRECIPE-NEXT: move $a3, $s2
|
|
; LA32F-FRECIPE-NEXT: bl __divdf3
|
|
; LA32F-FRECIPE-NEXT: move $s3, $a0
|
|
; LA32F-FRECIPE-NEXT: move $s4, $a1
|
|
; LA32F-FRECIPE-NEXT: lu12i.w $a1, 263248
|
|
; LA32F-FRECIPE-NEXT: move $a0, $zero
|
|
; LA32F-FRECIPE-NEXT: move $a2, $s1
|
|
; LA32F-FRECIPE-NEXT: move $a3, $s2
|
|
; LA32F-FRECIPE-NEXT: bl __divdf3
|
|
; LA32F-FRECIPE-NEXT: st.w $s3, $s0, 0
|
|
; LA32F-FRECIPE-NEXT: st.w $s4, $s0, 4
|
|
; LA32F-FRECIPE-NEXT: st.w $a0, $fp, 0
|
|
; LA32F-FRECIPE-NEXT: st.w $a1, $fp, 4
|
|
; LA32F-FRECIPE-NEXT: move $a0, $s1
|
|
; LA32F-FRECIPE-NEXT: move $a1, $s2
|
|
; LA32F-FRECIPE-NEXT: ld.w $s4, $sp, 4 # 4-byte Folded Reload
|
|
; LA32F-FRECIPE-NEXT: ld.w $s3, $sp, 8 # 4-byte Folded Reload
|
|
; LA32F-FRECIPE-NEXT: ld.w $s2, $sp, 12 # 4-byte Folded Reload
|
|
; LA32F-FRECIPE-NEXT: ld.w $s1, $sp, 16 # 4-byte Folded Reload
|
|
; LA32F-FRECIPE-NEXT: ld.w $s0, $sp, 20 # 4-byte Folded Reload
|
|
; LA32F-FRECIPE-NEXT: ld.w $fp, $sp, 24 # 4-byte Folded Reload
|
|
; LA32F-FRECIPE-NEXT: ld.w $ra, $sp, 28 # 4-byte Folded Reload
|
|
; LA32F-FRECIPE-NEXT: addi.w $sp, $sp, 32
|
|
; LA32F-FRECIPE-NEXT: ret
|
|
;
|
|
; LA64D-LABEL: sqrt_simplify_before_recip_3_uses_f64:
|
|
; LA64D: # %bb.0:
|
|
; LA64D-NEXT: pcalau12i $a2, %pc_hi20(.LCPI2_0)
|
|
; LA64D-NEXT: fld.d $fa2, $a2, %pc_lo12(.LCPI2_0)
|
|
; LA64D-NEXT: fsqrt.d $fa1, $fa0
|
|
; LA64D-NEXT: frsqrt.d $fa0, $fa0
|
|
; LA64D-NEXT: fdiv.d $fa2, $fa2, $fa1
|
|
; LA64D-NEXT: fst.d $fa0, $a0, 0
|
|
; LA64D-NEXT: fst.d $fa2, $a1, 0
|
|
; LA64D-NEXT: fmov.d $fa0, $fa1
|
|
; LA64D-NEXT: ret
|
|
;
|
|
; LA64D-FRECIPE-LABEL: sqrt_simplify_before_recip_3_uses_f64:
|
|
; LA64D-FRECIPE: # %bb.0:
|
|
; LA64D-FRECIPE-NEXT: frsqrte.d $fa1, $fa0
|
|
; LA64D-FRECIPE-NEXT: fmul.d $fa2, $fa0, $fa1
|
|
; LA64D-FRECIPE-NEXT: vldi $vr3, -888
|
|
; LA64D-FRECIPE-NEXT: fmadd.d $fa2, $fa2, $fa1, $fa3
|
|
; LA64D-FRECIPE-NEXT: vldi $vr4, -800
|
|
; LA64D-FRECIPE-NEXT: fmul.d $fa1, $fa1, $fa4
|
|
; LA64D-FRECIPE-NEXT: fmul.d $fa1, $fa1, $fa2
|
|
; LA64D-FRECIPE-NEXT: fmul.d $fa2, $fa0, $fa1
|
|
; LA64D-FRECIPE-NEXT: pcalau12i $a2, %pc_hi20(.LCPI2_0)
|
|
; LA64D-FRECIPE-NEXT: fld.d $fa5, $a2, %pc_lo12(.LCPI2_0)
|
|
; LA64D-FRECIPE-NEXT: fmadd.d $fa2, $fa2, $fa1, $fa3
|
|
; LA64D-FRECIPE-NEXT: fmul.d $fa1, $fa1, $fa4
|
|
; LA64D-FRECIPE-NEXT: fmul.d $fa1, $fa1, $fa2
|
|
; LA64D-FRECIPE-NEXT: fmul.d $fa2, $fa1, $fa5
|
|
; LA64D-FRECIPE-NEXT: fmul.d $fa0, $fa0, $fa1
|
|
; LA64D-FRECIPE-NEXT: fst.d $fa1, $a0, 0
|
|
; LA64D-FRECIPE-NEXT: fst.d $fa2, $a1, 0
|
|
; LA64D-FRECIPE-NEXT: ret
|
|
%sqrt = tail call fast double @llvm.sqrt.f64(double %x)
|
|
%rsqrt = fdiv fast double 1.0, %sqrt
|
|
%r = fdiv fast double 42.0, %sqrt
|
|
%sqrt_fast = fdiv fast double %x, %sqrt
|
|
store double %rsqrt, ptr %p1, align 8
|
|
store double %r, ptr %p2, align 8
|
|
ret double %sqrt_fast
|
|
}
|
|
|
|
|
|
define double @sqrt_simplify_before_recip_3_uses_order_f64(double %x, ptr %p1, ptr %p2) nounwind {
|
|
; LA32F-LABEL: sqrt_simplify_before_recip_3_uses_order_f64:
|
|
; LA32F: # %bb.0:
|
|
; LA32F-NEXT: addi.w $sp, $sp, -32
|
|
; LA32F-NEXT: st.w $ra, $sp, 28 # 4-byte Folded Spill
|
|
; LA32F-NEXT: st.w $fp, $sp, 24 # 4-byte Folded Spill
|
|
; LA32F-NEXT: st.w $s0, $sp, 20 # 4-byte Folded Spill
|
|
; LA32F-NEXT: st.w $s1, $sp, 16 # 4-byte Folded Spill
|
|
; LA32F-NEXT: st.w $s2, $sp, 12 # 4-byte Folded Spill
|
|
; LA32F-NEXT: st.w $s3, $sp, 8 # 4-byte Folded Spill
|
|
; LA32F-NEXT: st.w $s4, $sp, 4 # 4-byte Folded Spill
|
|
; LA32F-NEXT: move $fp, $a3
|
|
; LA32F-NEXT: move $s0, $a2
|
|
; LA32F-NEXT: bl sqrt
|
|
; LA32F-NEXT: move $s1, $a0
|
|
; LA32F-NEXT: move $s2, $a1
|
|
; LA32F-NEXT: lu12i.w $a1, 263248
|
|
; LA32F-NEXT: move $a0, $zero
|
|
; LA32F-NEXT: move $a2, $s1
|
|
; LA32F-NEXT: move $a3, $s2
|
|
; LA32F-NEXT: bl __divdf3
|
|
; LA32F-NEXT: move $s3, $a0
|
|
; LA32F-NEXT: move $s4, $a1
|
|
; LA32F-NEXT: lu12i.w $a1, 263256
|
|
; LA32F-NEXT: move $a0, $zero
|
|
; LA32F-NEXT: move $a2, $s1
|
|
; LA32F-NEXT: move $a3, $s2
|
|
; LA32F-NEXT: bl __divdf3
|
|
; LA32F-NEXT: st.w $s3, $s0, 0
|
|
; LA32F-NEXT: st.w $s4, $s0, 4
|
|
; LA32F-NEXT: st.w $a0, $fp, 0
|
|
; LA32F-NEXT: st.w $a1, $fp, 4
|
|
; LA32F-NEXT: move $a0, $s1
|
|
; LA32F-NEXT: move $a1, $s2
|
|
; LA32F-NEXT: ld.w $s4, $sp, 4 # 4-byte Folded Reload
|
|
; LA32F-NEXT: ld.w $s3, $sp, 8 # 4-byte Folded Reload
|
|
; LA32F-NEXT: ld.w $s2, $sp, 12 # 4-byte Folded Reload
|
|
; LA32F-NEXT: ld.w $s1, $sp, 16 # 4-byte Folded Reload
|
|
; LA32F-NEXT: ld.w $s0, $sp, 20 # 4-byte Folded Reload
|
|
; LA32F-NEXT: ld.w $fp, $sp, 24 # 4-byte Folded Reload
|
|
; LA32F-NEXT: ld.w $ra, $sp, 28 # 4-byte Folded Reload
|
|
; LA32F-NEXT: addi.w $sp, $sp, 32
|
|
; LA32F-NEXT: ret
|
|
;
|
|
; LA32F-FRECIPE-LABEL: sqrt_simplify_before_recip_3_uses_order_f64:
|
|
; LA32F-FRECIPE: # %bb.0:
|
|
; LA32F-FRECIPE-NEXT: addi.w $sp, $sp, -32
|
|
; LA32F-FRECIPE-NEXT: st.w $ra, $sp, 28 # 4-byte Folded Spill
|
|
; LA32F-FRECIPE-NEXT: st.w $fp, $sp, 24 # 4-byte Folded Spill
|
|
; LA32F-FRECIPE-NEXT: st.w $s0, $sp, 20 # 4-byte Folded Spill
|
|
; LA32F-FRECIPE-NEXT: st.w $s1, $sp, 16 # 4-byte Folded Spill
|
|
; LA32F-FRECIPE-NEXT: st.w $s2, $sp, 12 # 4-byte Folded Spill
|
|
; LA32F-FRECIPE-NEXT: st.w $s3, $sp, 8 # 4-byte Folded Spill
|
|
; LA32F-FRECIPE-NEXT: st.w $s4, $sp, 4 # 4-byte Folded Spill
|
|
; LA32F-FRECIPE-NEXT: move $fp, $a3
|
|
; LA32F-FRECIPE-NEXT: move $s0, $a2
|
|
; LA32F-FRECIPE-NEXT: bl sqrt
|
|
; LA32F-FRECIPE-NEXT: move $s1, $a0
|
|
; LA32F-FRECIPE-NEXT: move $s2, $a1
|
|
; LA32F-FRECIPE-NEXT: lu12i.w $a1, 263248
|
|
; LA32F-FRECIPE-NEXT: move $a0, $zero
|
|
; LA32F-FRECIPE-NEXT: move $a2, $s1
|
|
; LA32F-FRECIPE-NEXT: move $a3, $s2
|
|
; LA32F-FRECIPE-NEXT: bl __divdf3
|
|
; LA32F-FRECIPE-NEXT: move $s3, $a0
|
|
; LA32F-FRECIPE-NEXT: move $s4, $a1
|
|
; LA32F-FRECIPE-NEXT: lu12i.w $a1, 263256
|
|
; LA32F-FRECIPE-NEXT: move $a0, $zero
|
|
; LA32F-FRECIPE-NEXT: move $a2, $s1
|
|
; LA32F-FRECIPE-NEXT: move $a3, $s2
|
|
; LA32F-FRECIPE-NEXT: bl __divdf3
|
|
; LA32F-FRECIPE-NEXT: st.w $s3, $s0, 0
|
|
; LA32F-FRECIPE-NEXT: st.w $s4, $s0, 4
|
|
; LA32F-FRECIPE-NEXT: st.w $a0, $fp, 0
|
|
; LA32F-FRECIPE-NEXT: st.w $a1, $fp, 4
|
|
; LA32F-FRECIPE-NEXT: move $a0, $s1
|
|
; LA32F-FRECIPE-NEXT: move $a1, $s2
|
|
; LA32F-FRECIPE-NEXT: ld.w $s4, $sp, 4 # 4-byte Folded Reload
|
|
; LA32F-FRECIPE-NEXT: ld.w $s3, $sp, 8 # 4-byte Folded Reload
|
|
; LA32F-FRECIPE-NEXT: ld.w $s2, $sp, 12 # 4-byte Folded Reload
|
|
; LA32F-FRECIPE-NEXT: ld.w $s1, $sp, 16 # 4-byte Folded Reload
|
|
; LA32F-FRECIPE-NEXT: ld.w $s0, $sp, 20 # 4-byte Folded Reload
|
|
; LA32F-FRECIPE-NEXT: ld.w $fp, $sp, 24 # 4-byte Folded Reload
|
|
; LA32F-FRECIPE-NEXT: ld.w $ra, $sp, 28 # 4-byte Folded Reload
|
|
; LA32F-FRECIPE-NEXT: addi.w $sp, $sp, 32
|
|
; LA32F-FRECIPE-NEXT: ret
|
|
;
|
|
; LA64D-LABEL: sqrt_simplify_before_recip_3_uses_order_f64:
|
|
; LA64D: # %bb.0:
|
|
; LA64D-NEXT: pcalau12i $a2, %pc_hi20(.LCPI3_0)
|
|
; LA64D-NEXT: fld.d $fa1, $a2, %pc_lo12(.LCPI3_0)
|
|
; LA64D-NEXT: pcalau12i $a2, %pc_hi20(.LCPI3_1)
|
|
; LA64D-NEXT: fld.d $fa2, $a2, %pc_lo12(.LCPI3_1)
|
|
; LA64D-NEXT: fsqrt.d $fa0, $fa0
|
|
; LA64D-NEXT: fdiv.d $fa1, $fa1, $fa0
|
|
; LA64D-NEXT: fdiv.d $fa2, $fa2, $fa0
|
|
; LA64D-NEXT: fst.d $fa1, $a0, 0
|
|
; LA64D-NEXT: fst.d $fa2, $a1, 0
|
|
; LA64D-NEXT: ret
|
|
;
|
|
; LA64D-FRECIPE-LABEL: sqrt_simplify_before_recip_3_uses_order_f64:
|
|
; LA64D-FRECIPE: # %bb.0:
|
|
; LA64D-FRECIPE-NEXT: frsqrte.d $fa1, $fa0
|
|
; LA64D-FRECIPE-NEXT: fmul.d $fa2, $fa0, $fa1
|
|
; LA64D-FRECIPE-NEXT: vldi $vr3, -888
|
|
; LA64D-FRECIPE-NEXT: fmadd.d $fa2, $fa2, $fa1, $fa3
|
|
; LA64D-FRECIPE-NEXT: vldi $vr4, -800
|
|
; LA64D-FRECIPE-NEXT: fmul.d $fa1, $fa1, $fa4
|
|
; LA64D-FRECIPE-NEXT: fmul.d $fa1, $fa1, $fa2
|
|
; LA64D-FRECIPE-NEXT: fmul.d $fa2, $fa0, $fa1
|
|
; LA64D-FRECIPE-NEXT: fmadd.d $fa2, $fa2, $fa1, $fa3
|
|
; LA64D-FRECIPE-NEXT: fmul.d $fa1, $fa1, $fa4
|
|
; LA64D-FRECIPE-NEXT: pcalau12i $a2, %pc_hi20(.LCPI3_0)
|
|
; LA64D-FRECIPE-NEXT: fld.d $fa3, $a2, %pc_lo12(.LCPI3_0)
|
|
; LA64D-FRECIPE-NEXT: pcalau12i $a2, %pc_hi20(.LCPI3_1)
|
|
; LA64D-FRECIPE-NEXT: fld.d $fa4, $a2, %pc_lo12(.LCPI3_1)
|
|
; LA64D-FRECIPE-NEXT: fmul.d $fa1, $fa1, $fa2
|
|
; LA64D-FRECIPE-NEXT: fmul.d $fa0, $fa0, $fa1
|
|
; LA64D-FRECIPE-NEXT: fmul.d $fa2, $fa1, $fa3
|
|
; LA64D-FRECIPE-NEXT: fmul.d $fa1, $fa1, $fa4
|
|
; LA64D-FRECIPE-NEXT: fst.d $fa2, $a0, 0
|
|
; LA64D-FRECIPE-NEXT: fst.d $fa1, $a1, 0
|
|
; LA64D-FRECIPE-NEXT: ret
|
|
%sqrt = tail call fast double @llvm.sqrt.f64(double %x)
|
|
%sqrt_fast = fdiv fast double %x, %sqrt
|
|
%r1 = fdiv fast double 42.0, %sqrt
|
|
%r2 = fdiv fast double 43.0, %sqrt
|
|
store double %r1, ptr %p1, align 8
|
|
store double %r2, ptr %p2, align 8
|
|
ret double %sqrt_fast
|
|
}
|
|
|
|
define double @sqrt_simplify_before_recip_4_uses_f64(double %x, ptr %p1, ptr %p2, ptr %p3) nounwind {
|
|
; LA32F-LABEL: sqrt_simplify_before_recip_4_uses_f64:
|
|
; LA32F: # %bb.0:
|
|
; LA32F-NEXT: addi.w $sp, $sp, -48
|
|
; LA32F-NEXT: st.w $ra, $sp, 44 # 4-byte Folded Spill
|
|
; LA32F-NEXT: st.w $fp, $sp, 40 # 4-byte Folded Spill
|
|
; LA32F-NEXT: st.w $s0, $sp, 36 # 4-byte Folded Spill
|
|
; LA32F-NEXT: st.w $s1, $sp, 32 # 4-byte Folded Spill
|
|
; LA32F-NEXT: st.w $s2, $sp, 28 # 4-byte Folded Spill
|
|
; LA32F-NEXT: st.w $s3, $sp, 24 # 4-byte Folded Spill
|
|
; LA32F-NEXT: st.w $s4, $sp, 20 # 4-byte Folded Spill
|
|
; LA32F-NEXT: st.w $s5, $sp, 16 # 4-byte Folded Spill
|
|
; LA32F-NEXT: st.w $s6, $sp, 12 # 4-byte Folded Spill
|
|
; LA32F-NEXT: st.w $s7, $sp, 8 # 4-byte Folded Spill
|
|
; LA32F-NEXT: move $fp, $a4
|
|
; LA32F-NEXT: move $s0, $a3
|
|
; LA32F-NEXT: move $s1, $a2
|
|
; LA32F-NEXT: bl sqrt
|
|
; LA32F-NEXT: move $s2, $a0
|
|
; LA32F-NEXT: move $s3, $a1
|
|
; LA32F-NEXT: lu12i.w $a1, 261888
|
|
; LA32F-NEXT: move $a0, $zero
|
|
; LA32F-NEXT: move $a2, $s2
|
|
; LA32F-NEXT: move $a3, $s3
|
|
; LA32F-NEXT: bl __divdf3
|
|
; LA32F-NEXT: move $s4, $a0
|
|
; LA32F-NEXT: move $s5, $a1
|
|
; LA32F-NEXT: lu12i.w $a1, 263248
|
|
; LA32F-NEXT: move $a0, $zero
|
|
; LA32F-NEXT: move $a2, $s2
|
|
; LA32F-NEXT: move $a3, $s3
|
|
; LA32F-NEXT: bl __divdf3
|
|
; LA32F-NEXT: move $s6, $a0
|
|
; LA32F-NEXT: move $s7, $a1
|
|
; LA32F-NEXT: lu12i.w $a1, 263256
|
|
; LA32F-NEXT: move $a0, $zero
|
|
; LA32F-NEXT: move $a2, $s2
|
|
; LA32F-NEXT: move $a3, $s3
|
|
; LA32F-NEXT: bl __divdf3
|
|
; LA32F-NEXT: st.w $s4, $s1, 0
|
|
; LA32F-NEXT: st.w $s5, $s1, 4
|
|
; LA32F-NEXT: st.w $s6, $s0, 0
|
|
; LA32F-NEXT: st.w $s7, $s0, 4
|
|
; LA32F-NEXT: st.w $a0, $fp, 0
|
|
; LA32F-NEXT: st.w $a1, $fp, 4
|
|
; LA32F-NEXT: move $a0, $s2
|
|
; LA32F-NEXT: move $a1, $s3
|
|
; LA32F-NEXT: ld.w $s7, $sp, 8 # 4-byte Folded Reload
|
|
; LA32F-NEXT: ld.w $s6, $sp, 12 # 4-byte Folded Reload
|
|
; LA32F-NEXT: ld.w $s5, $sp, 16 # 4-byte Folded Reload
|
|
; LA32F-NEXT: ld.w $s4, $sp, 20 # 4-byte Folded Reload
|
|
; LA32F-NEXT: ld.w $s3, $sp, 24 # 4-byte Folded Reload
|
|
; LA32F-NEXT: ld.w $s2, $sp, 28 # 4-byte Folded Reload
|
|
; LA32F-NEXT: ld.w $s1, $sp, 32 # 4-byte Folded Reload
|
|
; LA32F-NEXT: ld.w $s0, $sp, 36 # 4-byte Folded Reload
|
|
; LA32F-NEXT: ld.w $fp, $sp, 40 # 4-byte Folded Reload
|
|
; LA32F-NEXT: ld.w $ra, $sp, 44 # 4-byte Folded Reload
|
|
; LA32F-NEXT: addi.w $sp, $sp, 48
|
|
; LA32F-NEXT: ret
|
|
;
|
|
; LA32F-FRECIPE-LABEL: sqrt_simplify_before_recip_4_uses_f64:
|
|
; LA32F-FRECIPE: # %bb.0:
|
|
; LA32F-FRECIPE-NEXT: addi.w $sp, $sp, -48
|
|
; LA32F-FRECIPE-NEXT: st.w $ra, $sp, 44 # 4-byte Folded Spill
|
|
; LA32F-FRECIPE-NEXT: st.w $fp, $sp, 40 # 4-byte Folded Spill
|
|
; LA32F-FRECIPE-NEXT: st.w $s0, $sp, 36 # 4-byte Folded Spill
|
|
; LA32F-FRECIPE-NEXT: st.w $s1, $sp, 32 # 4-byte Folded Spill
|
|
; LA32F-FRECIPE-NEXT: st.w $s2, $sp, 28 # 4-byte Folded Spill
|
|
; LA32F-FRECIPE-NEXT: st.w $s3, $sp, 24 # 4-byte Folded Spill
|
|
; LA32F-FRECIPE-NEXT: st.w $s4, $sp, 20 # 4-byte Folded Spill
|
|
; LA32F-FRECIPE-NEXT: st.w $s5, $sp, 16 # 4-byte Folded Spill
|
|
; LA32F-FRECIPE-NEXT: st.w $s6, $sp, 12 # 4-byte Folded Spill
|
|
; LA32F-FRECIPE-NEXT: st.w $s7, $sp, 8 # 4-byte Folded Spill
|
|
; LA32F-FRECIPE-NEXT: move $fp, $a4
|
|
; LA32F-FRECIPE-NEXT: move $s0, $a3
|
|
; LA32F-FRECIPE-NEXT: move $s1, $a2
|
|
; LA32F-FRECIPE-NEXT: bl sqrt
|
|
; LA32F-FRECIPE-NEXT: move $s2, $a0
|
|
; LA32F-FRECIPE-NEXT: move $s3, $a1
|
|
; LA32F-FRECIPE-NEXT: lu12i.w $a1, 261888
|
|
; LA32F-FRECIPE-NEXT: move $a0, $zero
|
|
; LA32F-FRECIPE-NEXT: move $a2, $s2
|
|
; LA32F-FRECIPE-NEXT: move $a3, $s3
|
|
; LA32F-FRECIPE-NEXT: bl __divdf3
|
|
; LA32F-FRECIPE-NEXT: move $s4, $a0
|
|
; LA32F-FRECIPE-NEXT: move $s5, $a1
|
|
; LA32F-FRECIPE-NEXT: lu12i.w $a1, 263248
|
|
; LA32F-FRECIPE-NEXT: move $a0, $zero
|
|
; LA32F-FRECIPE-NEXT: move $a2, $s2
|
|
; LA32F-FRECIPE-NEXT: move $a3, $s3
|
|
; LA32F-FRECIPE-NEXT: bl __divdf3
|
|
; LA32F-FRECIPE-NEXT: move $s6, $a0
|
|
; LA32F-FRECIPE-NEXT: move $s7, $a1
|
|
; LA32F-FRECIPE-NEXT: lu12i.w $a1, 263256
|
|
; LA32F-FRECIPE-NEXT: move $a0, $zero
|
|
; LA32F-FRECIPE-NEXT: move $a2, $s2
|
|
; LA32F-FRECIPE-NEXT: move $a3, $s3
|
|
; LA32F-FRECIPE-NEXT: bl __divdf3
|
|
; LA32F-FRECIPE-NEXT: st.w $s4, $s1, 0
|
|
; LA32F-FRECIPE-NEXT: st.w $s5, $s1, 4
|
|
; LA32F-FRECIPE-NEXT: st.w $s6, $s0, 0
|
|
; LA32F-FRECIPE-NEXT: st.w $s7, $s0, 4
|
|
; LA32F-FRECIPE-NEXT: st.w $a0, $fp, 0
|
|
; LA32F-FRECIPE-NEXT: st.w $a1, $fp, 4
|
|
; LA32F-FRECIPE-NEXT: move $a0, $s2
|
|
; LA32F-FRECIPE-NEXT: move $a1, $s3
|
|
; LA32F-FRECIPE-NEXT: ld.w $s7, $sp, 8 # 4-byte Folded Reload
|
|
; LA32F-FRECIPE-NEXT: ld.w $s6, $sp, 12 # 4-byte Folded Reload
|
|
; LA32F-FRECIPE-NEXT: ld.w $s5, $sp, 16 # 4-byte Folded Reload
|
|
; LA32F-FRECIPE-NEXT: ld.w $s4, $sp, 20 # 4-byte Folded Reload
|
|
; LA32F-FRECIPE-NEXT: ld.w $s3, $sp, 24 # 4-byte Folded Reload
|
|
; LA32F-FRECIPE-NEXT: ld.w $s2, $sp, 28 # 4-byte Folded Reload
|
|
; LA32F-FRECIPE-NEXT: ld.w $s1, $sp, 32 # 4-byte Folded Reload
|
|
; LA32F-FRECIPE-NEXT: ld.w $s0, $sp, 36 # 4-byte Folded Reload
|
|
; LA32F-FRECIPE-NEXT: ld.w $fp, $sp, 40 # 4-byte Folded Reload
|
|
; LA32F-FRECIPE-NEXT: ld.w $ra, $sp, 44 # 4-byte Folded Reload
|
|
; LA32F-FRECIPE-NEXT: addi.w $sp, $sp, 48
|
|
; LA32F-FRECIPE-NEXT: ret
|
|
;
|
|
; LA64D-LABEL: sqrt_simplify_before_recip_4_uses_f64:
|
|
; LA64D: # %bb.0:
|
|
; LA64D-NEXT: pcalau12i $a3, %pc_hi20(.LCPI4_0)
|
|
; LA64D-NEXT: fld.d $fa2, $a3, %pc_lo12(.LCPI4_0)
|
|
; LA64D-NEXT: pcalau12i $a3, %pc_hi20(.LCPI4_1)
|
|
; LA64D-NEXT: fld.d $fa3, $a3, %pc_lo12(.LCPI4_1)
|
|
; LA64D-NEXT: fsqrt.d $fa1, $fa0
|
|
; LA64D-NEXT: frsqrt.d $fa0, $fa0
|
|
; LA64D-NEXT: fdiv.d $fa2, $fa2, $fa1
|
|
; LA64D-NEXT: fdiv.d $fa3, $fa3, $fa1
|
|
; LA64D-NEXT: fst.d $fa0, $a0, 0
|
|
; LA64D-NEXT: fst.d $fa2, $a1, 0
|
|
; LA64D-NEXT: fst.d $fa3, $a2, 0
|
|
; LA64D-NEXT: fmov.d $fa0, $fa1
|
|
; LA64D-NEXT: ret
|
|
;
|
|
; LA64D-FRECIPE-LABEL: sqrt_simplify_before_recip_4_uses_f64:
|
|
; LA64D-FRECIPE: # %bb.0:
|
|
; LA64D-FRECIPE-NEXT: frsqrte.d $fa1, $fa0
|
|
; LA64D-FRECIPE-NEXT: fmul.d $fa2, $fa0, $fa1
|
|
; LA64D-FRECIPE-NEXT: vldi $vr3, -888
|
|
; LA64D-FRECIPE-NEXT: fmadd.d $fa2, $fa2, $fa1, $fa3
|
|
; LA64D-FRECIPE-NEXT: vldi $vr4, -800
|
|
; LA64D-FRECIPE-NEXT: fmul.d $fa1, $fa1, $fa4
|
|
; LA64D-FRECIPE-NEXT: fmul.d $fa1, $fa1, $fa2
|
|
; LA64D-FRECIPE-NEXT: fmul.d $fa2, $fa0, $fa1
|
|
; LA64D-FRECIPE-NEXT: fmadd.d $fa2, $fa2, $fa1, $fa3
|
|
; LA64D-FRECIPE-NEXT: pcalau12i $a3, %pc_hi20(.LCPI4_0)
|
|
; LA64D-FRECIPE-NEXT: fld.d $fa3, $a3, %pc_lo12(.LCPI4_0)
|
|
; LA64D-FRECIPE-NEXT: pcalau12i $a3, %pc_hi20(.LCPI4_1)
|
|
; LA64D-FRECIPE-NEXT: fld.d $fa5, $a3, %pc_lo12(.LCPI4_1)
|
|
; LA64D-FRECIPE-NEXT: fmul.d $fa1, $fa1, $fa4
|
|
; LA64D-FRECIPE-NEXT: fmul.d $fa1, $fa1, $fa2
|
|
; LA64D-FRECIPE-NEXT: fmul.d $fa2, $fa1, $fa3
|
|
; LA64D-FRECIPE-NEXT: fmul.d $fa3, $fa1, $fa5
|
|
; LA64D-FRECIPE-NEXT: fmul.d $fa0, $fa0, $fa1
|
|
; LA64D-FRECIPE-NEXT: fst.d $fa1, $a0, 0
|
|
; LA64D-FRECIPE-NEXT: fst.d $fa2, $a1, 0
|
|
; LA64D-FRECIPE-NEXT: fst.d $fa3, $a2, 0
|
|
; LA64D-FRECIPE-NEXT: ret
|
|
%sqrt = tail call fast double @llvm.sqrt.f64(double %x)
|
|
%rsqrt = fdiv fast double 1.0, %sqrt
|
|
%r1 = fdiv fast double 42.0, %sqrt
|
|
%r2 = fdiv fast double 43.0, %sqrt
|
|
%sqrt_fast = fdiv fast double %x, %sqrt
|
|
store double %rsqrt, ptr %p1, align 8
|
|
store double %r1, ptr %p2, align 8
|
|
store double %r2, ptr %p3, align 8
|
|
ret double %sqrt_fast
|
|
}
|
|
|
|
define float @sqrt_simplify_before_recip_3_uses_f32(float %x, ptr %p1, ptr %p2) nounwind {
|
|
; LA32F-LABEL: sqrt_simplify_before_recip_3_uses_f32:
|
|
; LA32F: # %bb.0:
|
|
; LA32F-NEXT: fsqrt.s $fa1, $fa0
|
|
; LA32F-NEXT: frsqrt.s $fa0, $fa0
|
|
; LA32F-NEXT: lu12i.w $a2, 270976
|
|
; LA32F-NEXT: movgr2fr.w $fa2, $a2
|
|
; LA32F-NEXT: fdiv.s $fa2, $fa2, $fa1
|
|
; LA32F-NEXT: fst.s $fa0, $a0, 0
|
|
; LA32F-NEXT: fst.s $fa2, $a1, 0
|
|
; LA32F-NEXT: fmov.s $fa0, $fa1
|
|
; LA32F-NEXT: ret
|
|
;
|
|
; LA32F-FRECIPE-LABEL: sqrt_simplify_before_recip_3_uses_f32:
|
|
; LA32F-FRECIPE: # %bb.0:
|
|
; LA32F-FRECIPE-NEXT: frsqrte.s $fa1, $fa0
|
|
; LA32F-FRECIPE-NEXT: fmul.s $fa2, $fa0, $fa1
|
|
; LA32F-FRECIPE-NEXT: lu12i.w $a2, -261120
|
|
; LA32F-FRECIPE-NEXT: movgr2fr.w $fa3, $a2
|
|
; LA32F-FRECIPE-NEXT: fmadd.s $fa2, $fa2, $fa1, $fa3
|
|
; LA32F-FRECIPE-NEXT: lu12i.w $a2, -266240
|
|
; LA32F-FRECIPE-NEXT: movgr2fr.w $fa3, $a2
|
|
; LA32F-FRECIPE-NEXT: fmul.s $fa1, $fa1, $fa3
|
|
; LA32F-FRECIPE-NEXT: fmul.s $fa1, $fa1, $fa2
|
|
; LA32F-FRECIPE-NEXT: lu12i.w $a2, 270976
|
|
; LA32F-FRECIPE-NEXT: movgr2fr.w $fa2, $a2
|
|
; LA32F-FRECIPE-NEXT: fmul.s $fa2, $fa1, $fa2
|
|
; LA32F-FRECIPE-NEXT: fmul.s $fa0, $fa0, $fa1
|
|
; LA32F-FRECIPE-NEXT: fst.s $fa1, $a0, 0
|
|
; LA32F-FRECIPE-NEXT: fst.s $fa2, $a1, 0
|
|
; LA32F-FRECIPE-NEXT: ret
|
|
;
|
|
; LA64D-LABEL: sqrt_simplify_before_recip_3_uses_f32:
|
|
; LA64D: # %bb.0:
|
|
; LA64D-NEXT: fsqrt.s $fa1, $fa0
|
|
; LA64D-NEXT: frsqrt.s $fa0, $fa0
|
|
; LA64D-NEXT: lu12i.w $a2, 270976
|
|
; LA64D-NEXT: movgr2fr.w $fa2, $a2
|
|
; LA64D-NEXT: fdiv.s $fa2, $fa2, $fa1
|
|
; LA64D-NEXT: fst.s $fa0, $a0, 0
|
|
; LA64D-NEXT: fst.s $fa2, $a1, 0
|
|
; LA64D-NEXT: fmov.s $fa0, $fa1
|
|
; LA64D-NEXT: ret
|
|
;
|
|
; LA64D-FRECIPE-LABEL: sqrt_simplify_before_recip_3_uses_f32:
|
|
; LA64D-FRECIPE: # %bb.0:
|
|
; LA64D-FRECIPE-NEXT: frsqrte.s $fa1, $fa0
|
|
; LA64D-FRECIPE-NEXT: fmul.s $fa2, $fa0, $fa1
|
|
; LA64D-FRECIPE-NEXT: vldi $vr3, -1144
|
|
; LA64D-FRECIPE-NEXT: fmadd.s $fa2, $fa2, $fa1, $fa3
|
|
; LA64D-FRECIPE-NEXT: vldi $vr3, -1056
|
|
; LA64D-FRECIPE-NEXT: fmul.s $fa1, $fa1, $fa3
|
|
; LA64D-FRECIPE-NEXT: fmul.s $fa1, $fa1, $fa2
|
|
; LA64D-FRECIPE-NEXT: lu12i.w $a2, 270976
|
|
; LA64D-FRECIPE-NEXT: movgr2fr.w $fa2, $a2
|
|
; LA64D-FRECIPE-NEXT: fmul.s $fa2, $fa1, $fa2
|
|
; LA64D-FRECIPE-NEXT: fmul.s $fa0, $fa0, $fa1
|
|
; LA64D-FRECIPE-NEXT: fst.s $fa1, $a0, 0
|
|
; LA64D-FRECIPE-NEXT: fst.s $fa2, $a1, 0
|
|
; LA64D-FRECIPE-NEXT: ret
|
|
%sqrt = tail call fast float @llvm.sqrt.f32(float %x)
|
|
%rsqrt = fdiv fast float 1.0, %sqrt
|
|
%r = fdiv fast float 42.0, %sqrt
|
|
%sqrt_fast = fdiv fast float %x, %sqrt
|
|
store float %rsqrt, ptr %p1, align 8
|
|
store float %r, ptr %p2, align 8
|
|
ret float %sqrt_fast
|
|
}
|
|
|
|
define float @sqrt_simplify_before_recip_4_uses_f32(float %x, ptr %p1, ptr %p2, ptr %p3) nounwind {
|
|
; LA32F-LABEL: sqrt_simplify_before_recip_4_uses_f32:
|
|
; LA32F: # %bb.0:
|
|
; LA32F-NEXT: fsqrt.s $fa1, $fa0
|
|
; LA32F-NEXT: frsqrt.s $fa0, $fa0
|
|
; LA32F-NEXT: lu12i.w $a3, 270976
|
|
; LA32F-NEXT: movgr2fr.w $fa2, $a3
|
|
; LA32F-NEXT: fdiv.s $fa2, $fa2, $fa1
|
|
; LA32F-NEXT: lu12i.w $a3, 271040
|
|
; LA32F-NEXT: movgr2fr.w $fa3, $a3
|
|
; LA32F-NEXT: fdiv.s $fa3, $fa3, $fa1
|
|
; LA32F-NEXT: fst.s $fa0, $a0, 0
|
|
; LA32F-NEXT: fst.s $fa2, $a1, 0
|
|
; LA32F-NEXT: fst.s $fa3, $a2, 0
|
|
; LA32F-NEXT: fmov.s $fa0, $fa1
|
|
; LA32F-NEXT: ret
|
|
;
|
|
; LA32F-FRECIPE-LABEL: sqrt_simplify_before_recip_4_uses_f32:
|
|
; LA32F-FRECIPE: # %bb.0:
|
|
; LA32F-FRECIPE-NEXT: frsqrte.s $fa1, $fa0
|
|
; LA32F-FRECIPE-NEXT: fmul.s $fa2, $fa0, $fa1
|
|
; LA32F-FRECIPE-NEXT: lu12i.w $a3, -261120
|
|
; LA32F-FRECIPE-NEXT: movgr2fr.w $fa3, $a3
|
|
; LA32F-FRECIPE-NEXT: fmadd.s $fa2, $fa2, $fa1, $fa3
|
|
; LA32F-FRECIPE-NEXT: lu12i.w $a3, -266240
|
|
; LA32F-FRECIPE-NEXT: movgr2fr.w $fa3, $a3
|
|
; LA32F-FRECIPE-NEXT: fmul.s $fa1, $fa1, $fa3
|
|
; LA32F-FRECIPE-NEXT: fmul.s $fa1, $fa1, $fa2
|
|
; LA32F-FRECIPE-NEXT: lu12i.w $a3, 270976
|
|
; LA32F-FRECIPE-NEXT: movgr2fr.w $fa2, $a3
|
|
; LA32F-FRECIPE-NEXT: fmul.s $fa2, $fa1, $fa2
|
|
; LA32F-FRECIPE-NEXT: lu12i.w $a3, 271040
|
|
; LA32F-FRECIPE-NEXT: movgr2fr.w $fa3, $a3
|
|
; LA32F-FRECIPE-NEXT: fmul.s $fa3, $fa1, $fa3
|
|
; LA32F-FRECIPE-NEXT: fmul.s $fa0, $fa0, $fa1
|
|
; LA32F-FRECIPE-NEXT: fst.s $fa1, $a0, 0
|
|
; LA32F-FRECIPE-NEXT: fst.s $fa2, $a1, 0
|
|
; LA32F-FRECIPE-NEXT: fst.s $fa3, $a2, 0
|
|
; LA32F-FRECIPE-NEXT: ret
|
|
;
|
|
; LA64D-LABEL: sqrt_simplify_before_recip_4_uses_f32:
|
|
; LA64D: # %bb.0:
|
|
; LA64D-NEXT: fsqrt.s $fa1, $fa0
|
|
; LA64D-NEXT: frsqrt.s $fa0, $fa0
|
|
; LA64D-NEXT: lu12i.w $a3, 270976
|
|
; LA64D-NEXT: movgr2fr.w $fa2, $a3
|
|
; LA64D-NEXT: fdiv.s $fa2, $fa2, $fa1
|
|
; LA64D-NEXT: lu12i.w $a3, 271040
|
|
; LA64D-NEXT: movgr2fr.w $fa3, $a3
|
|
; LA64D-NEXT: fdiv.s $fa3, $fa3, $fa1
|
|
; LA64D-NEXT: fst.s $fa0, $a0, 0
|
|
; LA64D-NEXT: fst.s $fa2, $a1, 0
|
|
; LA64D-NEXT: fst.s $fa3, $a2, 0
|
|
; LA64D-NEXT: fmov.s $fa0, $fa1
|
|
; LA64D-NEXT: ret
|
|
;
|
|
; LA64D-FRECIPE-LABEL: sqrt_simplify_before_recip_4_uses_f32:
|
|
; LA64D-FRECIPE: # %bb.0:
|
|
; LA64D-FRECIPE-NEXT: frsqrte.s $fa1, $fa0
|
|
; LA64D-FRECIPE-NEXT: fmul.s $fa2, $fa0, $fa1
|
|
; LA64D-FRECIPE-NEXT: vldi $vr3, -1144
|
|
; LA64D-FRECIPE-NEXT: fmadd.s $fa2, $fa2, $fa1, $fa3
|
|
; LA64D-FRECIPE-NEXT: vldi $vr3, -1056
|
|
; LA64D-FRECIPE-NEXT: fmul.s $fa1, $fa1, $fa3
|
|
; LA64D-FRECIPE-NEXT: fmul.s $fa1, $fa1, $fa2
|
|
; LA64D-FRECIPE-NEXT: lu12i.w $a3, 270976
|
|
; LA64D-FRECIPE-NEXT: movgr2fr.w $fa2, $a3
|
|
; LA64D-FRECIPE-NEXT: fmul.s $fa2, $fa1, $fa2
|
|
; LA64D-FRECIPE-NEXT: lu12i.w $a3, 271040
|
|
; LA64D-FRECIPE-NEXT: movgr2fr.w $fa3, $a3
|
|
; LA64D-FRECIPE-NEXT: fmul.s $fa3, $fa1, $fa3
|
|
; LA64D-FRECIPE-NEXT: fmul.s $fa0, $fa0, $fa1
|
|
; LA64D-FRECIPE-NEXT: fst.s $fa1, $a0, 0
|
|
; LA64D-FRECIPE-NEXT: fst.s $fa2, $a1, 0
|
|
; LA64D-FRECIPE-NEXT: fst.s $fa3, $a2, 0
|
|
; LA64D-FRECIPE-NEXT: ret
|
|
%sqrt = tail call fast float @llvm.sqrt.f32(float %x)
|
|
%rsqrt = fdiv fast float 1.0, %sqrt
|
|
%r1 = fdiv fast float 42.0, %sqrt
|
|
%r2 = fdiv fast float 43.0, %sqrt
|
|
%sqrt_fast = fdiv fast float %x, %sqrt
|
|
store float %rsqrt, ptr %p1, align 8
|
|
store float %r1, ptr %p2, align 8
|
|
store float %r2, ptr %p3, align 8
|
|
ret float %sqrt_fast
|
|
}
|
|
|
|
define float @sqrt_simplify_before_recip_3_uses_order_f32(float %x, ptr %p1, ptr %p2) nounwind {
|
|
; LA32F-LABEL: sqrt_simplify_before_recip_3_uses_order_f32:
|
|
; LA32F: # %bb.0:
|
|
; LA32F-NEXT: fsqrt.s $fa0, $fa0
|
|
; LA32F-NEXT: lu12i.w $a2, 270976
|
|
; LA32F-NEXT: movgr2fr.w $fa1, $a2
|
|
; LA32F-NEXT: fdiv.s $fa1, $fa1, $fa0
|
|
; LA32F-NEXT: lu12i.w $a2, 271040
|
|
; LA32F-NEXT: movgr2fr.w $fa2, $a2
|
|
; LA32F-NEXT: fdiv.s $fa2, $fa2, $fa0
|
|
; LA32F-NEXT: fst.s $fa1, $a0, 0
|
|
; LA32F-NEXT: fst.s $fa2, $a1, 0
|
|
; LA32F-NEXT: ret
|
|
;
|
|
; LA32F-FRECIPE-LABEL: sqrt_simplify_before_recip_3_uses_order_f32:
|
|
; LA32F-FRECIPE: # %bb.0:
|
|
; LA32F-FRECIPE-NEXT: frsqrte.s $fa1, $fa0
|
|
; LA32F-FRECIPE-NEXT: fmul.s $fa2, $fa0, $fa1
|
|
; LA32F-FRECIPE-NEXT: lu12i.w $a2, -261120
|
|
; LA32F-FRECIPE-NEXT: movgr2fr.w $fa3, $a2
|
|
; LA32F-FRECIPE-NEXT: fmadd.s $fa2, $fa2, $fa1, $fa3
|
|
; LA32F-FRECIPE-NEXT: lu12i.w $a2, -266240
|
|
; LA32F-FRECIPE-NEXT: movgr2fr.w $fa3, $a2
|
|
; LA32F-FRECIPE-NEXT: fmul.s $fa1, $fa1, $fa3
|
|
; LA32F-FRECIPE-NEXT: fmul.s $fa1, $fa1, $fa2
|
|
; LA32F-FRECIPE-NEXT: fmul.s $fa0, $fa0, $fa1
|
|
; LA32F-FRECIPE-NEXT: lu12i.w $a2, 270976
|
|
; LA32F-FRECIPE-NEXT: movgr2fr.w $fa2, $a2
|
|
; LA32F-FRECIPE-NEXT: fmul.s $fa2, $fa1, $fa2
|
|
; LA32F-FRECIPE-NEXT: lu12i.w $a2, 271040
|
|
; LA32F-FRECIPE-NEXT: movgr2fr.w $fa3, $a2
|
|
; LA32F-FRECIPE-NEXT: fmul.s $fa1, $fa1, $fa3
|
|
; LA32F-FRECIPE-NEXT: fst.s $fa2, $a0, 0
|
|
; LA32F-FRECIPE-NEXT: fst.s $fa1, $a1, 0
|
|
; LA32F-FRECIPE-NEXT: ret
|
|
;
|
|
; LA64D-LABEL: sqrt_simplify_before_recip_3_uses_order_f32:
|
|
; LA64D: # %bb.0:
|
|
; LA64D-NEXT: fsqrt.s $fa0, $fa0
|
|
; LA64D-NEXT: lu12i.w $a2, 270976
|
|
; LA64D-NEXT: movgr2fr.w $fa1, $a2
|
|
; LA64D-NEXT: fdiv.s $fa1, $fa1, $fa0
|
|
; LA64D-NEXT: lu12i.w $a2, 271040
|
|
; LA64D-NEXT: movgr2fr.w $fa2, $a2
|
|
; LA64D-NEXT: fdiv.s $fa2, $fa2, $fa0
|
|
; LA64D-NEXT: fst.s $fa1, $a0, 0
|
|
; LA64D-NEXT: fst.s $fa2, $a1, 0
|
|
; LA64D-NEXT: ret
|
|
;
|
|
; LA64D-FRECIPE-LABEL: sqrt_simplify_before_recip_3_uses_order_f32:
|
|
; LA64D-FRECIPE: # %bb.0:
|
|
; LA64D-FRECIPE-NEXT: frsqrte.s $fa1, $fa0
|
|
; LA64D-FRECIPE-NEXT: fmul.s $fa2, $fa0, $fa1
|
|
; LA64D-FRECIPE-NEXT: vldi $vr3, -1144
|
|
; LA64D-FRECIPE-NEXT: fmadd.s $fa2, $fa2, $fa1, $fa3
|
|
; LA64D-FRECIPE-NEXT: vldi $vr3, -1056
|
|
; LA64D-FRECIPE-NEXT: fmul.s $fa1, $fa1, $fa3
|
|
; LA64D-FRECIPE-NEXT: fmul.s $fa1, $fa1, $fa2
|
|
; LA64D-FRECIPE-NEXT: fmul.s $fa0, $fa0, $fa1
|
|
; LA64D-FRECIPE-NEXT: lu12i.w $a2, 270976
|
|
; LA64D-FRECIPE-NEXT: movgr2fr.w $fa2, $a2
|
|
; LA64D-FRECIPE-NEXT: fmul.s $fa2, $fa1, $fa2
|
|
; LA64D-FRECIPE-NEXT: lu12i.w $a2, 271040
|
|
; LA64D-FRECIPE-NEXT: movgr2fr.w $fa3, $a2
|
|
; LA64D-FRECIPE-NEXT: fmul.s $fa1, $fa1, $fa3
|
|
; LA64D-FRECIPE-NEXT: fst.s $fa2, $a0, 0
|
|
; LA64D-FRECIPE-NEXT: fst.s $fa1, $a1, 0
|
|
; LA64D-FRECIPE-NEXT: ret
|
|
%sqrt = tail call fast float @llvm.sqrt.f32(float %x)
|
|
%sqrt_fast = fdiv fast float %x, %sqrt
|
|
%r1 = fdiv fast float 42.0, %sqrt
|
|
%r2 = fdiv fast float 43.0, %sqrt
|
|
store float %r1, ptr %p1, align 8
|
|
store float %r2, ptr %p2, align 8
|
|
ret float %sqrt_fast
|
|
}
|