d9bbb902fe
Avoid ADD(X,X) as it doesn't correctly handle undef elements and helps avoid some FREEZE() fold headaches Resurrects #86857
290 lines
9.9 KiB
LLVM
290 lines
9.9 KiB
LLVM
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 3
|
|
; RUN: llc --mtriple=loongarch32 --mattr=+32s,+lsx < %s | FileCheck %s --check-prefixes=CHECK,LA32
|
|
; RUN: llc --mtriple=loongarch64 --mattr=+lsx < %s | FileCheck %s --check-prefixes=CHECK,LA64
|
|
|
|
define void @insert_16xi8(ptr %src, ptr %dst, i8 %ins) nounwind {
|
|
; CHECK-LABEL: insert_16xi8:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: vld $vr0, $a0, 0
|
|
; CHECK-NEXT: vinsgr2vr.b $vr0, $a2, 1
|
|
; CHECK-NEXT: vst $vr0, $a1, 0
|
|
; CHECK-NEXT: ret
|
|
%v = load volatile <16 x i8>, ptr %src
|
|
%v_new = insertelement <16 x i8> %v, i8 %ins, i32 1
|
|
store <16 x i8> %v_new, ptr %dst
|
|
ret void
|
|
}
|
|
|
|
define void @insert_8xi16(ptr %src, ptr %dst, i16 %ins) nounwind {
|
|
; CHECK-LABEL: insert_8xi16:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: vld $vr0, $a0, 0
|
|
; CHECK-NEXT: vinsgr2vr.h $vr0, $a2, 1
|
|
; CHECK-NEXT: vst $vr0, $a1, 0
|
|
; CHECK-NEXT: ret
|
|
%v = load volatile <8 x i16>, ptr %src
|
|
%v_new = insertelement <8 x i16> %v, i16 %ins, i32 1
|
|
store <8 x i16> %v_new, ptr %dst
|
|
ret void
|
|
}
|
|
|
|
define void @insert_4xi32(ptr %src, ptr %dst, i32 %ins) nounwind {
|
|
; CHECK-LABEL: insert_4xi32:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: vld $vr0, $a0, 0
|
|
; CHECK-NEXT: vinsgr2vr.w $vr0, $a2, 1
|
|
; CHECK-NEXT: vst $vr0, $a1, 0
|
|
; CHECK-NEXT: ret
|
|
%v = load volatile <4 x i32>, ptr %src
|
|
%v_new = insertelement <4 x i32> %v, i32 %ins, i32 1
|
|
store <4 x i32> %v_new, ptr %dst
|
|
ret void
|
|
}
|
|
|
|
define void @insert_2xi64(ptr %src, ptr %dst, i64 %ins) nounwind {
|
|
; LA32-LABEL: insert_2xi64:
|
|
; LA32: # %bb.0:
|
|
; LA32-NEXT: vld $vr0, $a0, 0
|
|
; LA32-NEXT: vinsgr2vr.w $vr0, $a2, 2
|
|
; LA32-NEXT: vinsgr2vr.w $vr0, $a3, 3
|
|
; LA32-NEXT: vst $vr0, $a1, 0
|
|
; LA32-NEXT: ret
|
|
;
|
|
; LA64-LABEL: insert_2xi64:
|
|
; LA64: # %bb.0:
|
|
; LA64-NEXT: vld $vr0, $a0, 0
|
|
; LA64-NEXT: vinsgr2vr.d $vr0, $a2, 1
|
|
; LA64-NEXT: vst $vr0, $a1, 0
|
|
; LA64-NEXT: ret
|
|
%v = load volatile <2 x i64>, ptr %src
|
|
%v_new = insertelement <2 x i64> %v, i64 %ins, i32 1
|
|
store <2 x i64> %v_new, ptr %dst
|
|
ret void
|
|
}
|
|
|
|
define void @insert_4xfloat(ptr %src, ptr %dst, float %ins) nounwind {
|
|
; CHECK-LABEL: insert_4xfloat:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: vld $vr1, $a0, 0
|
|
; CHECK-NEXT: # kill: def $f0 killed $f0 def $vr0
|
|
; CHECK-NEXT: vextrins.w $vr1, $vr0, 16
|
|
; CHECK-NEXT: vst $vr1, $a1, 0
|
|
; CHECK-NEXT: ret
|
|
%v = load volatile <4 x float>, ptr %src
|
|
%v_new = insertelement <4 x float> %v, float %ins, i32 1
|
|
store <4 x float> %v_new, ptr %dst
|
|
ret void
|
|
}
|
|
|
|
define void @insert_2xdouble(ptr %src, ptr %dst, double %ins) nounwind {
|
|
; CHECK-LABEL: insert_2xdouble:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: vld $vr1, $a0, 0
|
|
; CHECK-NEXT: # kill: def $f0_64 killed $f0_64 def $vr0
|
|
; CHECK-NEXT: vextrins.d $vr1, $vr0, 16
|
|
; CHECK-NEXT: vst $vr1, $a1, 0
|
|
; CHECK-NEXT: ret
|
|
%v = load volatile <2 x double>, ptr %src
|
|
%v_new = insertelement <2 x double> %v, double %ins, i32 1
|
|
store <2 x double> %v_new, ptr %dst
|
|
ret void
|
|
}
|
|
|
|
define void @insert_16xi8_idx(ptr %src, ptr %dst, i8 %ins, i32 %idx) nounwind {
|
|
; LA32-LABEL: insert_16xi8_idx:
|
|
; LA32: # %bb.0:
|
|
; LA32-NEXT: pcalau12i $a4, %pc_hi20(.LCPI6_0)
|
|
; LA32-NEXT: vld $vr0, $a4, %pc_lo12(.LCPI6_0)
|
|
; LA32-NEXT: vld $vr1, $a0, 0
|
|
; LA32-NEXT: vreplgr2vr.b $vr2, $a3
|
|
; LA32-NEXT: vseq.b $vr0, $vr2, $vr0
|
|
; LA32-NEXT: vreplgr2vr.b $vr2, $a2
|
|
; LA32-NEXT: vbitsel.v $vr0, $vr1, $vr2, $vr0
|
|
; LA32-NEXT: vst $vr0, $a1, 0
|
|
; LA32-NEXT: ret
|
|
;
|
|
; LA64-LABEL: insert_16xi8_idx:
|
|
; LA64: # %bb.0:
|
|
; LA64-NEXT: pcalau12i $a4, %pc_hi20(.LCPI6_0)
|
|
; LA64-NEXT: vld $vr0, $a4, %pc_lo12(.LCPI6_0)
|
|
; LA64-NEXT: vld $vr1, $a0, 0
|
|
; LA64-NEXT: bstrpick.d $a0, $a3, 31, 0
|
|
; LA64-NEXT: vreplgr2vr.b $vr2, $a0
|
|
; LA64-NEXT: vseq.b $vr0, $vr2, $vr0
|
|
; LA64-NEXT: vreplgr2vr.b $vr2, $a2
|
|
; LA64-NEXT: vbitsel.v $vr0, $vr1, $vr2, $vr0
|
|
; LA64-NEXT: vst $vr0, $a1, 0
|
|
; LA64-NEXT: ret
|
|
%v = load volatile <16 x i8>, ptr %src
|
|
%v_new = insertelement <16 x i8> %v, i8 %ins, i32 %idx
|
|
store <16 x i8> %v_new, ptr %dst
|
|
ret void
|
|
}
|
|
|
|
define void @insert_8xi16_idx(ptr %src, ptr %dst, i16 %ins, i32 %idx) nounwind {
|
|
; LA32-LABEL: insert_8xi16_idx:
|
|
; LA32: # %bb.0:
|
|
; LA32-NEXT: pcalau12i $a4, %pc_hi20(.LCPI7_0)
|
|
; LA32-NEXT: vld $vr0, $a4, %pc_lo12(.LCPI7_0)
|
|
; LA32-NEXT: vld $vr1, $a0, 0
|
|
; LA32-NEXT: vreplgr2vr.h $vr2, $a3
|
|
; LA32-NEXT: vseq.h $vr0, $vr2, $vr0
|
|
; LA32-NEXT: vreplgr2vr.h $vr2, $a2
|
|
; LA32-NEXT: vbitsel.v $vr0, $vr1, $vr2, $vr0
|
|
; LA32-NEXT: vst $vr0, $a1, 0
|
|
; LA32-NEXT: ret
|
|
;
|
|
; LA64-LABEL: insert_8xi16_idx:
|
|
; LA64: # %bb.0:
|
|
; LA64-NEXT: pcalau12i $a4, %pc_hi20(.LCPI7_0)
|
|
; LA64-NEXT: vld $vr0, $a4, %pc_lo12(.LCPI7_0)
|
|
; LA64-NEXT: vld $vr1, $a0, 0
|
|
; LA64-NEXT: bstrpick.d $a0, $a3, 31, 0
|
|
; LA64-NEXT: vreplgr2vr.h $vr2, $a0
|
|
; LA64-NEXT: vseq.h $vr0, $vr2, $vr0
|
|
; LA64-NEXT: vreplgr2vr.h $vr2, $a2
|
|
; LA64-NEXT: vbitsel.v $vr0, $vr1, $vr2, $vr0
|
|
; LA64-NEXT: vst $vr0, $a1, 0
|
|
; LA64-NEXT: ret
|
|
%v = load volatile <8 x i16>, ptr %src
|
|
%v_new = insertelement <8 x i16> %v, i16 %ins, i32 %idx
|
|
store <8 x i16> %v_new, ptr %dst
|
|
ret void
|
|
}
|
|
|
|
define void @insert_4xi32_idx(ptr %src, ptr %dst, i32 %ins, i32 %idx) nounwind {
|
|
; LA32-LABEL: insert_4xi32_idx:
|
|
; LA32: # %bb.0:
|
|
; LA32-NEXT: pcalau12i $a4, %pc_hi20(.LCPI8_0)
|
|
; LA32-NEXT: vld $vr0, $a4, %pc_lo12(.LCPI8_0)
|
|
; LA32-NEXT: vld $vr1, $a0, 0
|
|
; LA32-NEXT: vreplgr2vr.w $vr2, $a3
|
|
; LA32-NEXT: vseq.w $vr0, $vr2, $vr0
|
|
; LA32-NEXT: vreplgr2vr.w $vr2, $a2
|
|
; LA32-NEXT: vbitsel.v $vr0, $vr1, $vr2, $vr0
|
|
; LA32-NEXT: vst $vr0, $a1, 0
|
|
; LA32-NEXT: ret
|
|
;
|
|
; LA64-LABEL: insert_4xi32_idx:
|
|
; LA64: # %bb.0:
|
|
; LA64-NEXT: pcalau12i $a4, %pc_hi20(.LCPI8_0)
|
|
; LA64-NEXT: vld $vr0, $a4, %pc_lo12(.LCPI8_0)
|
|
; LA64-NEXT: vld $vr1, $a0, 0
|
|
; LA64-NEXT: bstrpick.d $a0, $a3, 31, 0
|
|
; LA64-NEXT: vreplgr2vr.w $vr2, $a0
|
|
; LA64-NEXT: vseq.w $vr0, $vr2, $vr0
|
|
; LA64-NEXT: vreplgr2vr.w $vr2, $a2
|
|
; LA64-NEXT: vbitsel.v $vr0, $vr1, $vr2, $vr0
|
|
; LA64-NEXT: vst $vr0, $a1, 0
|
|
; LA64-NEXT: ret
|
|
%v = load volatile <4 x i32>, ptr %src
|
|
%v_new = insertelement <4 x i32> %v, i32 %ins, i32 %idx
|
|
store <4 x i32> %v_new, ptr %dst
|
|
ret void
|
|
}
|
|
|
|
define void @insert_2xi64_idx(ptr %src, ptr %dst, i64 %ins, i32 %idx) nounwind {
|
|
; LA32-LABEL: insert_2xi64_idx:
|
|
; LA32: # %bb.0:
|
|
; LA32-NEXT: pcalau12i $a5, %pc_hi20(.LCPI9_0)
|
|
; LA32-NEXT: vld $vr0, $a5, %pc_lo12(.LCPI9_0)
|
|
; LA32-NEXT: slli.w $a4, $a4, 1
|
|
; LA32-NEXT: vld $vr1, $a0, 0
|
|
; LA32-NEXT: vreplgr2vr.w $vr2, $a4
|
|
; LA32-NEXT: vseq.w $vr2, $vr2, $vr0
|
|
; LA32-NEXT: vreplgr2vr.w $vr3, $a2
|
|
; LA32-NEXT: vbitsel.v $vr1, $vr1, $vr3, $vr2
|
|
; LA32-NEXT: addi.w $a0, $a4, 1
|
|
; LA32-NEXT: vreplgr2vr.w $vr2, $a0
|
|
; LA32-NEXT: vseq.w $vr0, $vr2, $vr0
|
|
; LA32-NEXT: vreplgr2vr.w $vr2, $a3
|
|
; LA32-NEXT: vbitsel.v $vr0, $vr1, $vr2, $vr0
|
|
; LA32-NEXT: vst $vr0, $a1, 0
|
|
; LA32-NEXT: ret
|
|
;
|
|
; LA64-LABEL: insert_2xi64_idx:
|
|
; LA64: # %bb.0:
|
|
; LA64-NEXT: pcalau12i $a4, %pc_hi20(.LCPI9_0)
|
|
; LA64-NEXT: vld $vr0, $a4, %pc_lo12(.LCPI9_0)
|
|
; LA64-NEXT: vld $vr1, $a0, 0
|
|
; LA64-NEXT: bstrpick.d $a0, $a3, 31, 0
|
|
; LA64-NEXT: vreplgr2vr.d $vr2, $a0
|
|
; LA64-NEXT: vseq.d $vr0, $vr2, $vr0
|
|
; LA64-NEXT: vreplgr2vr.d $vr2, $a2
|
|
; LA64-NEXT: vbitsel.v $vr0, $vr1, $vr2, $vr0
|
|
; LA64-NEXT: vst $vr0, $a1, 0
|
|
; LA64-NEXT: ret
|
|
%v = load volatile <2 x i64>, ptr %src
|
|
%v_new = insertelement <2 x i64> %v, i64 %ins, i32 %idx
|
|
store <2 x i64> %v_new, ptr %dst
|
|
ret void
|
|
}
|
|
|
|
define void @insert_4xfloat_idx(ptr %src, ptr %dst, float %ins, i32 %idx) nounwind {
|
|
; LA32-LABEL: insert_4xfloat_idx:
|
|
; LA32: # %bb.0:
|
|
; LA32-NEXT: pcalau12i $a3, %pc_hi20(.LCPI10_0)
|
|
; LA32-NEXT: vld $vr1, $a3, %pc_lo12(.LCPI10_0)
|
|
; LA32-NEXT: # kill: def $f0 killed $f0 def $vr0
|
|
; LA32-NEXT: vld $vr2, $a0, 0
|
|
; LA32-NEXT: vreplgr2vr.w $vr3, $a2
|
|
; LA32-NEXT: vseq.w $vr1, $vr3, $vr1
|
|
; LA32-NEXT: vreplvei.w $vr0, $vr0, 0
|
|
; LA32-NEXT: vbitsel.v $vr0, $vr2, $vr0, $vr1
|
|
; LA32-NEXT: vst $vr0, $a1, 0
|
|
; LA32-NEXT: ret
|
|
;
|
|
; LA64-LABEL: insert_4xfloat_idx:
|
|
; LA64: # %bb.0:
|
|
; LA64-NEXT: # kill: def $f0 killed $f0 def $vr0
|
|
; LA64-NEXT: pcalau12i $a3, %pc_hi20(.LCPI10_0)
|
|
; LA64-NEXT: vld $vr1, $a3, %pc_lo12(.LCPI10_0)
|
|
; LA64-NEXT: vld $vr2, $a0, 0
|
|
; LA64-NEXT: bstrpick.d $a0, $a2, 31, 0
|
|
; LA64-NEXT: vreplgr2vr.w $vr3, $a0
|
|
; LA64-NEXT: vseq.w $vr1, $vr3, $vr1
|
|
; LA64-NEXT: vreplvei.w $vr0, $vr0, 0
|
|
; LA64-NEXT: vbitsel.v $vr0, $vr2, $vr0, $vr1
|
|
; LA64-NEXT: vst $vr0, $a1, 0
|
|
; LA64-NEXT: ret
|
|
%v = load volatile <4 x float>, ptr %src
|
|
%v_new = insertelement <4 x float> %v, float %ins, i32 %idx
|
|
store <4 x float> %v_new, ptr %dst
|
|
ret void
|
|
}
|
|
|
|
define void @insert_2xdouble_idx(ptr %src, ptr %dst, double %ins, i32 %idx) nounwind {
|
|
; LA32-LABEL: insert_2xdouble_idx:
|
|
; LA32: # %bb.0:
|
|
; LA32-NEXT: # kill: def $f0_64 killed $f0_64 def $vr0
|
|
; LA32-NEXT: vld $vr1, $a0, 0
|
|
; LA32-NEXT: pcalau12i $a0, %pc_hi20(.LCPI11_0)
|
|
; LA32-NEXT: vld $vr2, $a0, %pc_lo12(.LCPI11_0)
|
|
; LA32-NEXT: vrepli.b $vr3, 0
|
|
; LA32-NEXT: vinsgr2vr.w $vr3, $a2, 0
|
|
; LA32-NEXT: vinsgr2vr.w $vr3, $a2, 2
|
|
; LA32-NEXT: vseq.d $vr2, $vr3, $vr2
|
|
; LA32-NEXT: vreplvei.d $vr0, $vr0, 0
|
|
; LA32-NEXT: vbitsel.v $vr0, $vr1, $vr0, $vr2
|
|
; LA32-NEXT: vst $vr0, $a1, 0
|
|
; LA32-NEXT: ret
|
|
;
|
|
; LA64-LABEL: insert_2xdouble_idx:
|
|
; LA64: # %bb.0:
|
|
; LA64-NEXT: # kill: def $f0_64 killed $f0_64 def $vr0
|
|
; LA64-NEXT: pcalau12i $a3, %pc_hi20(.LCPI11_0)
|
|
; LA64-NEXT: vld $vr1, $a3, %pc_lo12(.LCPI11_0)
|
|
; LA64-NEXT: vld $vr2, $a0, 0
|
|
; LA64-NEXT: bstrpick.d $a0, $a2, 31, 0
|
|
; LA64-NEXT: vreplgr2vr.d $vr3, $a0
|
|
; LA64-NEXT: vseq.d $vr1, $vr3, $vr1
|
|
; LA64-NEXT: vreplvei.d $vr0, $vr0, 0
|
|
; LA64-NEXT: vbitsel.v $vr0, $vr2, $vr0, $vr1
|
|
; LA64-NEXT: vst $vr0, $a1, 0
|
|
; LA64-NEXT: ret
|
|
%v = load volatile <2 x double>, ptr %src
|
|
%v_new = insertelement <2 x double> %v, double %ins, i32 %idx
|
|
store <2 x double> %v_new, ptr %dst
|
|
ret void
|
|
}
|