RPCS3/llvm-mirror
1
0
Fork 0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2024-11-25 04:02:41 +01:00
Code Issues Projects Releases Wiki Activity

[CodeGen][regalloc] Don't align stack slots if the stack can't be realigned

Browse Source 
Register allocation may spill virtual registers to the stack, which can
increase alignment requirements of the stack frame. If the the function
did not require stack realignment before register allocation, the
registers required to do so may not be reserved/available. This results
in a stack frame that requires realignment but can not be realigned.

Instead, only increase the alignment of the stack if we are still able
to realign.

The register SpillAlignment will be ignored if we can't realign, and the
backend will be responsible for emitting the correct unaligned loads and
stores. This seems to be the assumed behaviour already, e.g.
ARMBaseInstrInfo::storeRegToStackSlot and X86InstrInfo::storeRegToStackSlot
are both `canRealignStack` aware.

Differential Revision: https://reviews.llvm.org/D103602
This commit is contained in:
Tomas Matheson 2021-04-20 19:03:09 +01:00
parent 2de818fda2
commit bb84fe4048
36 changed files with 1083 additions and 1076 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
lib/CodeGen/VirtRegMap.cpp
View File

@ -30,6 +30,7 @@
#include "llvm/CodeGen/MachineOperand.h"
#include "llvm/CodeGen/MachineRegisterInfo.h"
#include "llvm/CodeGen/SlotIndexes.h"
#include "llvm/CodeGen/TargetFrameLowering.h"
#include "llvm/CodeGen/TargetInstrInfo.h"
#include "llvm/CodeGen/TargetOpcodes.h"
#include "llvm/CodeGen/TargetRegisterInfo.h"
@ -94,6 +95,12 @@ void VirtRegMap::assignVirt2Phys(Register virtReg, MCPhysReg physReg) {
unsigned VirtRegMap::createSpillSlot(const TargetRegisterClass *RC) {
unsigned Size = TRI->getSpillSize(*RC);
Align Alignment = TRI->getSpillAlign(*RC);
// Set preferred alignment if we are still able to realign the stack
auto &ST = MF->getSubtarget();
Align CurrentAlign = ST.getFrameLowering()->getStackAlign();
if (Alignment > CurrentAlign && !ST.getRegisterInfo()->canRealignStack(*MF)) {
Alignment = CurrentAlign;
}
int SS = MF->getFrameInfo().CreateSpillStackObject(Size, Alignment);
++NumSpillSlots;
return SS;

40
test/CodeGen/Thumb2/mve-float32regloops.ll
View File

@ -1649,20 +1649,20 @@ define arm_aapcs_vfpcc void @arm_biquad_cascade_df1_f32(%struct.arm_biquad_casd_
; CHECK-NEXT: sub sp, #4
; CHECK-NEXT: .vsave {d8, d9, d10, d11, d12, d13, d14, d15}
; CHECK-NEXT: vpush {d8, d9, d10, d11, d12, d13, d14, d15}
; CHECK-NEXT: .pad #88
; CHECK-NEXT: sub sp, #88
; CHECK-NEXT: .pad #64
; CHECK-NEXT: sub sp, #64
; CHECK-NEXT: ldrd r12, r10, [r0]
; CHECK-NEXT: @ implicit-def: $s2
; CHECK-NEXT: and r7, r3, #3
; CHECK-NEXT: ldr.w r9, [r0, #8]
; CHECK-NEXT: lsrs r0, r3, #2
; CHECK-NEXT: str r0, [sp, #8] @ 4-byte Spill
; CHECK-NEXT: str r7, [sp, #12] @ 4-byte Spill
; CHECK-NEXT: str r2, [sp, #60] @ 4-byte Spill
; CHECK-NEXT: str r0, [sp] @ 4-byte Spill
; CHECK-NEXT: str r7, [sp, #4] @ 4-byte Spill
; CHECK-NEXT: str r2, [sp, #44] @ 4-byte Spill
; CHECK-NEXT: b .LBB19_3
; CHECK-NEXT: .LBB19_1: @ in Loop: Header=BB19_3 Depth=1
; CHECK-NEXT: vmov.f32 s14, s7
; CHECK-NEXT: ldr r2, [sp, #60] @ 4-byte Reload
; CHECK-NEXT: ldr r2, [sp, #44] @ 4-byte Reload
; CHECK-NEXT: vmov.f32 s0, s10
; CHECK-NEXT: vmov.f32 s7, s6
; CHECK-NEXT: .LBB19_2: @ %if.end69
@ -1681,14 +1681,14 @@ define arm_aapcs_vfpcc void @arm_biquad_cascade_df1_f32(%struct.arm_biquad_casd_
; CHECK-NEXT: @ Child Loop BB19_5 Depth 2
; CHECK-NEXT: vldr s7, [r10, #8]
; CHECK-NEXT: mov r5, r2
; CHECK-NEXT: ldr r0, [sp, #8] @ 4-byte Reload
; CHECK-NEXT: ldr r0, [sp] @ 4-byte Reload
; CHECK-NEXT: vldr s8, [r10]
; CHECK-NEXT: vldr s10, [r10, #4]
; CHECK-NEXT: vldr s6, [r10, #12]
; CHECK-NEXT: wls lr, r0, .LBB19_6
; CHECK-NEXT: @ %bb.4: @ %while.body.lr.ph
; CHECK-NEXT: @ in Loop: Header=BB19_3 Depth=1
; CHECK-NEXT: ldr r5, [sp, #60] @ 4-byte Reload
; CHECK-NEXT: ldr r5, [sp, #44] @ 4-byte Reload
; CHECK-NEXT: .LBB19_5: @ %while.body
; CHECK-NEXT: @ Parent Loop BB19_3 Depth=1
; CHECK-NEXT: @ => This Inner Loop Header: Depth=2
@ -1700,7 +1700,7 @@ define arm_aapcs_vfpcc void @arm_biquad_cascade_df1_f32(%struct.arm_biquad_casd_
; CHECK-NEXT: vmov r7, s7
; CHECK-NEXT: vmov r11, s6
; CHECK-NEXT: vldrw.u32 q1, [r9]
; CHECK-NEXT: vstrw.32 q0, [sp, #64] @ 16-byte Spill
; CHECK-NEXT: vstrw.32 q0, [sp, #48] @ 16-byte Spill
; CHECK-NEXT: vmov r8, s8
; CHECK-NEXT: vldrw.u32 q0, [r9, #16]
; CHECK-NEXT: ldr r6, [r1, #4]
@ -1717,7 +1717,7 @@ define arm_aapcs_vfpcc void @arm_biquad_cascade_df1_f32(%struct.arm_biquad_casd_
; CHECK-NEXT: vfma.f32 q1, q6, r4
; CHECK-NEXT: vldrw.u32 q4, [r9, #96]
; CHECK-NEXT: vfma.f32 q1, q5, r3
; CHECK-NEXT: vldrw.u32 q0, [sp, #64] @ 16-byte Reload
; CHECK-NEXT: vldrw.u32 q0, [sp, #48] @ 16-byte Reload
; CHECK-NEXT: vfma.f32 q1, q4, r7
; CHECK-NEXT: vfma.f32 q1, q0, r11
; CHECK-NEXT: vmov.f32 s2, s8
@ -1725,7 +1725,7 @@ define arm_aapcs_vfpcc void @arm_biquad_cascade_df1_f32(%struct.arm_biquad_casd_
; CHECK-NEXT: le lr, .LBB19_5
; CHECK-NEXT: .LBB19_6: @ %while.end
; CHECK-NEXT: @ in Loop: Header=BB19_3 Depth=1
; CHECK-NEXT: ldr r7, [sp, #12] @ 4-byte Reload
; CHECK-NEXT: ldr r7, [sp, #4] @ 4-byte Reload
; CHECK-NEXT: cmp r7, #0
; CHECK-NEXT: beq .LBB19_1
; CHECK-NEXT: @ %bb.7: @ %if.then
@ -1742,28 +1742,28 @@ define arm_aapcs_vfpcc void @arm_biquad_cascade_df1_f32(%struct.arm_biquad_casd_
; CHECK-NEXT: vmov r6, s3
; CHECK-NEXT: vmul.f32 q3, q3, r6
; CHECK-NEXT: vmov r6, s1
; CHECK-NEXT: vstrw.32 q2, [sp, #32] @ 16-byte Spill
; CHECK-NEXT: vstrw.32 q2, [sp, #24] @ 16-byte Spill
; CHECK-NEXT: vldrw.u32 q2, [r9, #112]
; CHECK-NEXT: vldrw.u32 q5, [r9, #48]
; CHECK-NEXT: vmov r4, s0
; CHECK-NEXT: vstrw.32 q2, [sp, #64] @ 16-byte Spill
; CHECK-NEXT: vstrw.32 q2, [sp, #48] @ 16-byte Spill
; CHECK-NEXT: vldrw.u32 q2, [r9, #80]
; CHECK-NEXT: vldrw.u32 q7, [r9, #64]
; CHECK-NEXT: vmov r3, s24
; CHECK-NEXT: vstrw.32 q2, [sp, #16] @ 16-byte Spill
; CHECK-NEXT: vstrw.32 q2, [sp, #8] @ 16-byte Spill
; CHECK-NEXT: vldrw.u32 q2, [r9, #16]
; CHECK-NEXT: vmov r2, s7
; CHECK-NEXT: cmp r7, #1
; CHECK-NEXT: vfma.f32 q3, q2, r6
; CHECK-NEXT: vldrw.u32 q2, [sp, #16] @ 16-byte Reload
; CHECK-NEXT: vldrw.u32 q2, [sp, #8] @ 16-byte Reload
; CHECK-NEXT: vfma.f32 q3, q4, r4
; CHECK-NEXT: vmov lr, s6
; CHECK-NEXT: vfma.f32 q3, q5, r3
; CHECK-NEXT: vfma.f32 q3, q7, r0
; CHECK-NEXT: vfma.f32 q3, q2, r1
; CHECK-NEXT: vldrw.u32 q2, [sp, #32] @ 16-byte Reload
; CHECK-NEXT: vldrw.u32 q2, [sp, #24] @ 16-byte Reload
; CHECK-NEXT: vfma.f32 q3, q2, r2
; CHECK-NEXT: vldrw.u32 q2, [sp, #64] @ 16-byte Reload
; CHECK-NEXT: vldrw.u32 q2, [sp, #48] @ 16-byte Reload
; CHECK-NEXT: vfma.f32 q3, q2, lr
; CHECK-NEXT: bne .LBB19_9
; CHECK-NEXT: @ %bb.8: @ %if.then58
@ -1786,18 +1786,18 @@ define arm_aapcs_vfpcc void @arm_biquad_cascade_df1_f32(%struct.arm_biquad_casd_
; CHECK-NEXT: .LBB19_11: @ %if.end69
; CHECK-NEXT: @ in Loop: Header=BB19_3 Depth=1
; CHECK-NEXT: vmov.f32 s2, s3
; CHECK-NEXT: ldr r2, [sp, #60] @ 4-byte Reload
; CHECK-NEXT: ldr r2, [sp, #44] @ 4-byte Reload
; CHECK-NEXT: b .LBB19_2
; CHECK-NEXT: .LBB19_12: @ %if.else64
; CHECK-NEXT: @ in Loop: Header=BB19_3 Depth=1
; CHECK-NEXT: vmov.f32 s7, s13
; CHECK-NEXT: ldr r2, [sp, #60] @ 4-byte Reload
; CHECK-NEXT: ldr r2, [sp, #44] @ 4-byte Reload
; CHECK-NEXT: vmov.f32 s2, s3
; CHECK-NEXT: vstr s14, [r5, #8]
; CHECK-NEXT: vmov.f32 s8, s1
; CHECK-NEXT: b .LBB19_2
; CHECK-NEXT: .LBB19_13: @ %do.end
; CHECK-NEXT: add sp, #88
; CHECK-NEXT: add sp, #64
; CHECK-NEXT: vpop {d8, d9, d10, d11, d12, d13, d14, d15}
; CHECK-NEXT: add sp, #4
; CHECK-NEXT: pop.w {r4, r5, r6, r7, r8, r9, r10, r11, pc}

114
test/CodeGen/Thumb2/mve-gather-increment.ll
View File

@ -676,20 +676,20 @@ define arm_aapcs_vfpcc void @gather_inc_v8i16_complex(i16* noalias nocapture rea
; CHECK-NEXT: sub sp, #4
; CHECK-NEXT: .vsave {d8, d9, d10, d11, d12, d13, d14, d15}
; CHECK-NEXT: vpush {d8, d9, d10, d11, d12, d13, d14, d15}
; CHECK-NEXT: .pad #72
; CHECK-NEXT: sub sp, #72
; CHECK-NEXT: .pad #64
; CHECK-NEXT: sub sp, #64
; CHECK-NEXT: cmp r2, #1
; CHECK-NEXT: str r1, [sp, #64] @ 4-byte Spill
; CHECK-NEXT: str r1, [sp, #56] @ 4-byte Spill
; CHECK-NEXT: mov r1, r2
; CHECK-NEXT: str r2, [sp, #68] @ 4-byte Spill
; CHECK-NEXT: str r2, [sp, #60] @ 4-byte Spill
; CHECK-NEXT: blt.w .LBB12_5
; CHECK-NEXT: @ %bb.1: @ %vector.ph.preheader
; CHECK-NEXT: ldr r1, [sp, #68] @ 4-byte Reload
; CHECK-NEXT: ldr r1, [sp, #60] @ 4-byte Reload
; CHECK-NEXT: adr r6, .LCPI12_2
; CHECK-NEXT: vldrw.u32 q1, [r6]
; CHECK-NEXT: movs r7, #1
; CHECK-NEXT: bic r1, r1, #7
; CHECK-NEXT: str r1, [sp, #60] @ 4-byte Spill
; CHECK-NEXT: str r1, [sp, #52] @ 4-byte Spill
; CHECK-NEXT: sub.w r3, r1, #8
; CHECK-NEXT: vstrw.32 q1, [sp, #32] @ 16-byte Spill
; CHECK-NEXT: vmov.i16 q0, #0x18
@ -697,16 +697,16 @@ define arm_aapcs_vfpcc void @gather_inc_v8i16_complex(i16* noalias nocapture rea
; CHECK-NEXT: adr r3, .LCPI12_0
; CHECK-NEXT: vldrw.u32 q1, [r3]
; CHECK-NEXT: adr r7, .LCPI12_1
; CHECK-NEXT: str r1, [sp, #56] @ 4-byte Spill
; CHECK-NEXT: str r1, [sp, #48] @ 4-byte Spill
; CHECK-NEXT: vstrw.32 q1, [sp, #16] @ 16-byte Spill
; CHECK-NEXT: vldrw.u32 q1, [r7]
; CHECK-NEXT: vstrw.32 q1, [sp] @ 16-byte Spill
; CHECK-NEXT: .LBB12_2: @ %vector.ph
; CHECK-NEXT: @ =>This Loop Header: Depth=1
; CHECK-NEXT: @ Child Loop BB12_3 Depth 2
; CHECK-NEXT: ldr r1, [sp, #56] @ 4-byte Reload
; CHECK-NEXT: ldr r1, [sp, #48] @ 4-byte Reload
; CHECK-NEXT: dls lr, r1
; CHECK-NEXT: ldr.w r12, [sp, #64] @ 4-byte Reload
; CHECK-NEXT: ldr.w r12, [sp, #56] @ 4-byte Reload
; CHECK-NEXT: vldrw.u32 q4, [sp, #16] @ 16-byte Reload
; CHECK-NEXT: vldrw.u32 q5, [sp, #32] @ 16-byte Reload
; CHECK-NEXT: vldrw.u32 q6, [sp] @ 16-byte Reload
@ -836,12 +836,12 @@ define arm_aapcs_vfpcc void @gather_inc_v8i16_complex(i16* noalias nocapture rea
; CHECK-NEXT: le lr, .LBB12_3
; CHECK-NEXT: @ %bb.4: @ %middle.block
; CHECK-NEXT: @ in Loop: Header=BB12_2 Depth=1
; CHECK-NEXT: ldr r1, [sp, #68] @ 4-byte Reload
; CHECK-NEXT: ldr r2, [sp, #60] @ 4-byte Reload
; CHECK-NEXT: ldr r1, [sp, #60] @ 4-byte Reload
; CHECK-NEXT: ldr r2, [sp, #52] @ 4-byte Reload
; CHECK-NEXT: cmp r2, r1
; CHECK-NEXT: bne.w .LBB12_2
; CHECK-NEXT: .LBB12_5: @ %for.cond.cleanup
; CHECK-NEXT: add sp, #72
; CHECK-NEXT: add sp, #64
; CHECK-NEXT: vpop {d8, d9, d10, d11, d12, d13, d14, d15}
; CHECK-NEXT: add sp, #4
; CHECK-NEXT: pop.w {r4, r5, r6, r7, r8, r9, r10, r11, pc}
@ -924,10 +924,10 @@ define arm_aapcs_vfpcc void @gather_inc_v16i8_complex(i8* noalias nocapture read
; CHECK-NEXT: sub sp, #4
; CHECK-NEXT: .vsave {d8, d9, d10, d11, d12, d13, d14, d15}
; CHECK-NEXT: vpush {d8, d9, d10, d11, d12, d13, d14, d15}
; CHECK-NEXT: .pad #328
; CHECK-NEXT: sub sp, #328
; CHECK-NEXT: .pad #312
; CHECK-NEXT: sub sp, #312
; CHECK-NEXT: cmp r2, #1
; CHECK-NEXT: str r1, [sp, #124] @ 4-byte Spill
; CHECK-NEXT: str r1, [sp, #116] @ 4-byte Spill
; CHECK-NEXT: blt.w .LBB13_5
; CHECK-NEXT: @ %bb.1: @ %vector.ph.preheader
; CHECK-NEXT: adr r1, .LCPI13_0
@ -964,44 +964,44 @@ define arm_aapcs_vfpcc void @gather_inc_v16i8_complex(i8* noalias nocapture read
; CHECK-NEXT: adr r1, .LCPI13_2
; CHECK-NEXT: vldrw.u32 q3, [r1]
; CHECK-NEXT: adr r1, .LCPI13_10
; CHECK-NEXT: vstrw.32 q6, [sp, #288] @ 16-byte Spill
; CHECK-NEXT: vstrw.32 q6, [sp, #280] @ 16-byte Spill
; CHECK-NEXT: vldrw.u32 q6, [sp, #32] @ 16-byte Reload
; CHECK-NEXT: vstrw.32 q3, [sp, #304] @ 16-byte Spill
; CHECK-NEXT: vstrw.32 q3, [sp, #296] @ 16-byte Spill
; CHECK-NEXT: vldrw.u32 q3, [r1]
; CHECK-NEXT: adr r1, .LCPI13_11
; CHECK-NEXT: ldr.w r8, [sp, #124] @ 4-byte Reload
; CHECK-NEXT: vstrw.32 q3, [sp, #256] @ 16-byte Spill
; CHECK-NEXT: ldr.w r8, [sp, #116] @ 4-byte Reload
; CHECK-NEXT: vstrw.32 q3, [sp, #248] @ 16-byte Spill
; CHECK-NEXT: vldrw.u32 q3, [sp, #80] @ 16-byte Reload
; CHECK-NEXT: vstrw.32 q6, [sp, #272] @ 16-byte Spill
; CHECK-NEXT: vstrw.32 q6, [sp, #264] @ 16-byte Spill
; CHECK-NEXT: vldrw.u32 q6, [sp, #48] @ 16-byte Reload
; CHECK-NEXT: vstrw.32 q3, [sp, #224] @ 16-byte Spill
; CHECK-NEXT: vstrw.32 q3, [sp, #216] @ 16-byte Spill
; CHECK-NEXT: vldrw.u32 q3, [sp, #64] @ 16-byte Reload
; CHECK-NEXT: vldrw.u32 q7, [r1]
; CHECK-NEXT: vldrw.u32 q1, [sp] @ 16-byte Reload
; CHECK-NEXT: vstrw.32 q3, [sp, #208] @ 16-byte Spill
; CHECK-NEXT: vstrw.32 q3, [sp, #200] @ 16-byte Spill
; CHECK-NEXT: vldrw.u32 q3, [sp, #96] @ 16-byte Reload
; CHECK-NEXT: mov r11, r10
; CHECK-NEXT: vstrw.32 q6, [sp, #240] @ 16-byte Spill
; CHECK-NEXT: vstrw.32 q3, [sp, #192] @ 16-byte Spill
; CHECK-NEXT: vstrw.32 q6, [sp, #232] @ 16-byte Spill
; CHECK-NEXT: vstrw.32 q3, [sp, #184] @ 16-byte Spill
; CHECK-NEXT: .LBB13_3: @ %vector.body
; CHECK-NEXT: @ Parent Loop BB13_2 Depth=1
; CHECK-NEXT: @ => This Inner Loop Header: Depth=2
; CHECK-NEXT: vadd.i32 q4, q1, r0
; CHECK-NEXT: vstrw.32 q7, [sp, #144] @ 16-byte Spill
; CHECK-NEXT: vstrw.32 q7, [sp, #136] @ 16-byte Spill
; CHECK-NEXT: vmov r1, lr, d8
; CHECK-NEXT: vadd.i32 q7, q7, r0
; CHECK-NEXT: vmov r5, r4, d15
; CHECK-NEXT: vadd.i32 q6, q0, r0
; CHECK-NEXT: vmov r6, r7, d13
; CHECK-NEXT: vstrw.32 q1, [sp, #160] @ 16-byte Spill
; CHECK-NEXT: vldrw.u32 q1, [sp, #304] @ 16-byte Reload
; CHECK-NEXT: vstrw.32 q0, [sp, #176] @ 16-byte Spill
; CHECK-NEXT: vstrw.32 q1, [sp, #152] @ 16-byte Spill
; CHECK-NEXT: vldrw.u32 q1, [sp, #296] @ 16-byte Reload
; CHECK-NEXT: vstrw.32 q0, [sp, #168] @ 16-byte Spill
; CHECK-NEXT: vmov q0, q2
; CHECK-NEXT: vmov q3, q5
; CHECK-NEXT: vstrw.32 q1, [sp, #304] @ 16-byte Spill
; CHECK-NEXT: vstrw.32 q1, [sp, #296] @ 16-byte Spill
; CHECK-NEXT: vadd.i32 q1, q1, r0
; CHECK-NEXT: vldrw.u32 q3, [sp, #224] @ 16-byte Reload
; CHECK-NEXT: vstrw.32 q5, [sp, #128] @ 16-byte Spill
; CHECK-NEXT: vldrw.u32 q3, [sp, #216] @ 16-byte Reload
; CHECK-NEXT: vstrw.32 q5, [sp, #120] @ 16-byte Spill
; CHECK-NEXT: subs.w r11, r11, #16
; CHECK-NEXT: ldrb.w r9, [r1]
; CHECK-NEXT: vmov r1, r3, d14
@ -1028,14 +1028,14 @@ define arm_aapcs_vfpcc void @gather_inc_v16i8_complex(i8* noalias nocapture read
; CHECK-NEXT: vmov.8 q6[5], r7
; CHECK-NEXT: ldrb r4, [r1]
; CHECK-NEXT: vmov r1, r5, d3
; CHECK-NEXT: vldrw.u32 q1, [sp, #256] @ 16-byte Reload
; CHECK-NEXT: vldrw.u32 q1, [sp, #248] @ 16-byte Reload
; CHECK-NEXT: vadd.i32 q0, q1, r0
; CHECK-NEXT: vstrw.32 q1, [sp, #256] @ 16-byte Spill
; CHECK-NEXT: vldrw.u32 q1, [sp, #240] @ 16-byte Reload
; CHECK-NEXT: vstrw.32 q1, [sp, #248] @ 16-byte Spill
; CHECK-NEXT: vldrw.u32 q1, [sp, #232] @ 16-byte Reload
; CHECK-NEXT: ldrb.w r12, [r1]
; CHECK-NEXT: vmov r1, r3, d9
; CHECK-NEXT: ldrb r5, [r5]
; CHECK-NEXT: vldrw.u32 q4, [sp, #192] @ 16-byte Reload
; CHECK-NEXT: vldrw.u32 q4, [sp, #184] @ 16-byte Reload
; CHECK-NEXT: ldrb r1, [r1]
; CHECK-NEXT: ldrb r3, [r3]
; CHECK-NEXT: vmov.8 q6[6], r1
@ -1045,9 +1045,9 @@ define arm_aapcs_vfpcc void @gather_inc_v16i8_complex(i8* noalias nocapture read
; CHECK-NEXT: ldrb r7, [r7]
; CHECK-NEXT: vmov.8 q7[4], r1
; CHECK-NEXT: vmov r1, r3, d1
; CHECK-NEXT: vldrw.u32 q0, [sp, #272] @ 16-byte Reload
; CHECK-NEXT: vldrw.u32 q0, [sp, #264] @ 16-byte Reload
; CHECK-NEXT: vmov.8 q7[5], r7
; CHECK-NEXT: vstrw.32 q0, [sp, #272] @ 16-byte Spill
; CHECK-NEXT: vstrw.32 q0, [sp, #264] @ 16-byte Spill
; CHECK-NEXT: vadd.i32 q0, q0, r0
; CHECK-NEXT: ldrb r1, [r1]
; CHECK-NEXT: ldrb r3, [r3]
@ -1056,9 +1056,9 @@ define arm_aapcs_vfpcc void @gather_inc_v16i8_complex(i8* noalias nocapture read
; CHECK-NEXT: vmov r7, r6, d0
; CHECK-NEXT: vmov.8 q7[7], r3
; CHECK-NEXT: vmov r3, lr, d1
; CHECK-NEXT: vldrw.u32 q0, [sp, #288] @ 16-byte Reload
; CHECK-NEXT: vldrw.u32 q0, [sp, #280] @ 16-byte Reload
; CHECK-NEXT: vmov.8 q7[8], r1
; CHECK-NEXT: vstrw.32 q0, [sp, #288] @ 16-byte Spill
; CHECK-NEXT: vstrw.32 q0, [sp, #280] @ 16-byte Spill
; CHECK-NEXT: vadd.i32 q0, q0, r0
; CHECK-NEXT: vmov.8 q7[9], r4
; CHECK-NEXT: vmov r4, r1, d0
@ -1073,7 +1073,7 @@ define arm_aapcs_vfpcc void @gather_inc_v16i8_complex(i8* noalias nocapture read
; CHECK-NEXT: vmov r5, r4, d1
; CHECK-NEXT: vmov.8 q6[9], r1
; CHECK-NEXT: vadd.i32 q0, q5, r0
; CHECK-NEXT: vldrw.u32 q5, [sp, #208] @ 16-byte Reload
; CHECK-NEXT: vldrw.u32 q5, [sp, #200] @ 16-byte Reload
; CHECK-NEXT: ldrb r5, [r5]
; CHECK-NEXT: ldrb r4, [r4]
; CHECK-NEXT: vmov.8 q6[10], r5
@ -1089,11 +1089,11 @@ define arm_aapcs_vfpcc void @gather_inc_v16i8_complex(i8* noalias nocapture read
; CHECK-NEXT: vmov r1, r3, d1
; CHECK-NEXT: vadd.i32 q0, q1, r0
; CHECK-NEXT: vadd.i32 q1, q1, q2
; CHECK-NEXT: vstrw.32 q1, [sp, #240] @ 16-byte Spill
; CHECK-NEXT: vldrw.u32 q1, [sp, #256] @ 16-byte Reload
; CHECK-NEXT: vstrw.32 q1, [sp, #232] @ 16-byte Spill
; CHECK-NEXT: vldrw.u32 q1, [sp, #248] @ 16-byte Reload
; CHECK-NEXT: vadd.i32 q1, q1, q2
; CHECK-NEXT: vstrw.32 q1, [sp, #256] @ 16-byte Spill
; CHECK-NEXT: vldrw.u32 q1, [sp, #160] @ 16-byte Reload
; CHECK-NEXT: vstrw.32 q1, [sp, #248] @ 16-byte Spill
; CHECK-NEXT: vldrw.u32 q1, [sp, #152] @ 16-byte Reload
; CHECK-NEXT: vadd.i32 q1, q1, q2
; CHECK-NEXT: ldrb r1, [r1]
; CHECK-NEXT: vmov.8 q7[14], r1
@ -1110,16 +1110,16 @@ define arm_aapcs_vfpcc void @gather_inc_v16i8_complex(i8* noalias nocapture read
; CHECK-NEXT: vmov r1, r3, d1
; CHECK-NEXT: vadd.i32 q0, q3, r0
; CHECK-NEXT: vadd.i32 q3, q3, q2
; CHECK-NEXT: vstrw.32 q3, [sp, #224] @ 16-byte Spill
; CHECK-NEXT: vldrw.u32 q3, [sp, #304] @ 16-byte Reload
; CHECK-NEXT: vstrw.32 q3, [sp, #216] @ 16-byte Spill
; CHECK-NEXT: vldrw.u32 q3, [sp, #296] @ 16-byte Reload
; CHECK-NEXT: vadd.i32 q3, q3, q2
; CHECK-NEXT: vstrw.32 q3, [sp, #304] @ 16-byte Spill
; CHECK-NEXT: vldrw.u32 q3, [sp, #288] @ 16-byte Reload
; CHECK-NEXT: vstrw.32 q3, [sp, #296] @ 16-byte Spill
; CHECK-NEXT: vldrw.u32 q3, [sp, #280] @ 16-byte Reload
; CHECK-NEXT: vadd.i32 q3, q3, q2
; CHECK-NEXT: vstrw.32 q3, [sp, #288] @ 16-byte Spill
; CHECK-NEXT: vldrw.u32 q3, [sp, #272] @ 16-byte Reload
; CHECK-NEXT: vstrw.32 q3, [sp, #280] @ 16-byte Spill
; CHECK-NEXT: vldrw.u32 q3, [sp, #264] @ 16-byte Reload
; CHECK-NEXT: vadd.i32 q3, q3, q2
; CHECK-NEXT: vstrw.32 q3, [sp, #272] @ 16-byte Spill
; CHECK-NEXT: vstrw.32 q3, [sp, #264] @ 16-byte Spill
; CHECK-NEXT: ldrb r1, [r1]
; CHECK-NEXT: vmov.8 q7[2], r1
; CHECK-NEXT: ldrb r1, [r3]
@ -1132,8 +1132,8 @@ define arm_aapcs_vfpcc void @gather_inc_v16i8_complex(i8* noalias nocapture read
; CHECK-NEXT: vmov r1, r3, d1
; CHECK-NEXT: vadd.i32 q0, q5, r0
; CHECK-NEXT: vadd.i32 q5, q5, q2
; CHECK-NEXT: vstrw.32 q5, [sp, #208] @ 16-byte Spill
; CHECK-NEXT: vldrw.u32 q5, [sp, #128] @ 16-byte Reload
; CHECK-NEXT: vstrw.32 q5, [sp, #200] @ 16-byte Spill
; CHECK-NEXT: vldrw.u32 q5, [sp, #120] @ 16-byte Reload
; CHECK-NEXT: vadd.i32 q5, q5, q2
; CHECK-NEXT: ldrb r1, [r1]
; CHECK-NEXT: vmov.8 q7[6], r1
@ -1147,7 +1147,7 @@ define arm_aapcs_vfpcc void @gather_inc_v16i8_complex(i8* noalias nocapture read
; CHECK-NEXT: vmov r1, r3, d1
; CHECK-NEXT: vadd.i32 q0, q4, r0
; CHECK-NEXT: vadd.i32 q4, q4, q2
; CHECK-NEXT: vstrw.32 q4, [sp, #192] @ 16-byte Spill
; CHECK-NEXT: vstrw.32 q4, [sp, #184] @ 16-byte Spill
; CHECK-NEXT: ldrb r1, [r1]
; CHECK-NEXT: vmov.8 q7[10], r1
; CHECK-NEXT: ldrb r1, [r3]
@ -1163,9 +1163,9 @@ define arm_aapcs_vfpcc void @gather_inc_v16i8_complex(i8* noalias nocapture read
; CHECK-NEXT: ldrb r1, [r3]
; CHECK-NEXT: vmov.8 q7[15], r1
; CHECK-NEXT: vadd.i8 q0, q6, q7
; CHECK-NEXT: vldrw.u32 q7, [sp, #144] @ 16-byte Reload
; CHECK-NEXT: vldrw.u32 q7, [sp, #136] @ 16-byte Reload
; CHECK-NEXT: vstrb.8 q0, [r8], #16
; CHECK-NEXT: vldrw.u32 q0, [sp, #176] @ 16-byte Reload
; CHECK-NEXT: vldrw.u32 q0, [sp, #168] @ 16-byte Reload
; CHECK-NEXT: vadd.i32 q7, q7, q2
; CHECK-NEXT: vadd.i32 q0, q0, q2
; CHECK-NEXT: bne.w .LBB13_3
@ -1174,7 +1174,7 @@ define arm_aapcs_vfpcc void @gather_inc_v16i8_complex(i8* noalias nocapture read
; CHECK-NEXT: cmp r10, r2
; CHECK-NEXT: bne.w .LBB13_2
; CHECK-NEXT: .LBB13_5: @ %for.cond.cleanup
; CHECK-NEXT: add sp, #328
; CHECK-NEXT: add sp, #312
; CHECK-NEXT: vpop {d8, d9, d10, d11, d12, d13, d14, d15}
; CHECK-NEXT: add sp, #4
; CHECK-NEXT: pop.w {r4, r5, r6, r7, r8, r9, r10, r11, pc}

8
test/CodeGen/Thumb2/mve-gather-scatter-optimisation.ll
View File

@ -444,9 +444,9 @@ define dso_local void @arm_mat_mult_q31(i32* noalias nocapture readonly %A, i32*
; CHECK-NEXT: sub sp, #4
; CHECK-NEXT: .vsave {d8, d9, d10, d11, d12, d13, d14, d15}
; CHECK-NEXT: vpush {d8, d9, d10, d11, d12, d13, d14, d15}
; CHECK-NEXT: .pad #24
; CHECK-NEXT: sub sp, #24
; CHECK-NEXT: ldrd r9, r12, [sp, #128]
; CHECK-NEXT: .pad #16
; CHECK-NEXT: sub sp, #16
; CHECK-NEXT: ldrd r9, r12, [sp, #120]
; CHECK-NEXT: sub.w r7, r12, #1
; CHECK-NEXT: movs r6, #1
; CHECK-NEXT: mov.w r8, #0
@ -506,7 +506,7 @@ define dso_local void @arm_mat_mult_q31(i32* noalias nocapture readonly %A, i32*
; CHECK-NEXT: cmp r8, r3
; CHECK-NEXT: bne .LBB9_1
; CHECK-NEXT: @ %bb.6: @ %for.end25
; CHECK-NEXT: add sp, #24
; CHECK-NEXT: add sp, #16
; CHECK-NEXT: vpop {d8, d9, d10, d11, d12, d13, d14, d15}
; CHECK-NEXT: add sp, #4
; CHECK-NEXT: pop.w {r4, r5, r6, r7, r8, r9, r10, r11, pc}

6
test/CodeGen/Thumb2/mve-multivec-spill.ll
View File

@ -10,8 +10,8 @@ define arm_aapcs_vfpcc void @spill_multivector(<4 x i32>* %p) {
; CHECK-NEXT: push {r4, r5, r7, lr}
; CHECK-NEXT: .vsave {d8, d9, d10, d11, d12, d13, d14, d15}
; CHECK-NEXT: vpush {d8, d9, d10, d11, d12, d13, d14, d15}
; CHECK-NEXT: .pad #112
; CHECK-NEXT: sub sp, #112
; CHECK-NEXT: .pad #96
; CHECK-NEXT: sub sp, #96
; CHECK-NEXT: vld20.32 {q0, q1}, [r0]
; CHECK-NEXT: mov r5, r0
; CHECK-NEXT: add.w lr, sp, #64
@ -49,7 +49,7 @@ define arm_aapcs_vfpcc void @spill_multivector(<4 x i32>* %p) {
; CHECK-NEXT: vldmia lr, {d2, d3, d4, d5} @ 32-byte Reload
; CHECK-NEXT: vstrw.32 q2, [r4, #16]
; CHECK-NEXT: vstrw.32 q1, [r4]
; CHECK-NEXT: add sp, #112
; CHECK-NEXT: add sp, #96
; CHECK-NEXT: vpop {d8, d9, d10, d11, d12, d13, d14, d15}
; CHECK-NEXT: pop {r4, r5, r7, pc}
entry:

20
test/CodeGen/Thumb2/mve-phireg.ll
View File

@ -12,8 +12,8 @@ define arm_aapcs_vfpcc void @k() {
; CHECK-NEXT: sub sp, #4
; CHECK-NEXT: .vsave {d8, d9, d10, d11, d12, d13, d14}
; CHECK-NEXT: vpush {d8, d9, d10, d11, d12, d13, d14}
; CHECK-NEXT: .pad #24
; CHECK-NEXT: sub sp, #24
; CHECK-NEXT: .pad #16
; CHECK-NEXT: sub sp, #16
; CHECK-NEXT: adr.w r8, .LCPI0_0
; CHECK-NEXT: adr.w r9, .LCPI0_1
; CHECK-NEXT: vldrw.u32 q6, [r8]
@ -149,8 +149,8 @@ define dso_local i32 @e() #0 {
; CHECK-NEXT: push.w {r4, r5, r6, r7, r8, r9, r10, lr}
; CHECK-NEXT: .vsave {d8, d9, d10, d11, d12, d13, d14, d15}
; CHECK-NEXT: vpush {d8, d9, d10, d11, d12, d13, d14, d15}
; CHECK-NEXT: .pad #416
; CHECK-NEXT: sub sp, #416
; CHECK-NEXT: .pad #408
; CHECK-NEXT: sub sp, #408
; CHECK-NEXT: movw r7, :lower16:.L_MergedGlobals
; CHECK-NEXT: vldr s12, .LCPI1_0
; CHECK-NEXT: movt r7, :upper16:.L_MergedGlobals
@ -173,7 +173,7 @@ define dso_local i32 @e() #0 {
; CHECK-NEXT: vdup.32 q7, r3
; CHECK-NEXT: vmov q6[2], q6[0], r3, r5
; CHECK-NEXT: vmov.f32 s22, s13
; CHECK-NEXT: vstrw.32 q0, [sp, #100]
; CHECK-NEXT: vstrw.32 q0, [sp, #92]
; CHECK-NEXT: vmov q0, q7
; CHECK-NEXT: vmov q6[3], q6[1], r3, r2
; CHECK-NEXT: vmov q4, q7
@ -181,7 +181,7 @@ define dso_local i32 @e() #0 {
; CHECK-NEXT: vmov.32 q7[1], r2
; CHECK-NEXT: vmov.f32 s23, s15
; CHECK-NEXT: movs r1, #64
; CHECK-NEXT: str r0, [sp, #48]
; CHECK-NEXT: str r0, [sp, #40]
; CHECK-NEXT: vstrw.32 q5, [r0]
; CHECK-NEXT: str r6, [r0]
; CHECK-NEXT: vstrw.32 q7, [r0]
@ -196,9 +196,9 @@ define dso_local i32 @e() #0 {
; CHECK-NEXT: vmov q2[3], q2[1], r4, r5
; CHECK-NEXT: vmov.32 q4[0], r8
; CHECK-NEXT: @ implicit-def: $r2
; CHECK-NEXT: str.w r8, [sp, #52]
; CHECK-NEXT: vstrw.32 q3, [sp, #68]
; CHECK-NEXT: strh.w r12, [sp, #414]
; CHECK-NEXT: str.w r8, [sp, #44]
; CHECK-NEXT: vstrw.32 q3, [sp, #60]
; CHECK-NEXT: strh.w r12, [sp, #406]
; CHECK-NEXT: wlstp.8 lr, r1, .LBB1_2
; CHECK-NEXT: .LBB1_1: @ =>This Inner Loop Header: Depth=1
; CHECK-NEXT: vldrw.u32 q0, [sp] @ 16-byte Reload
@ -209,7 +209,7 @@ define dso_local i32 @e() #0 {
; CHECK-NEXT: str.w r8, [r7]
; CHECK-NEXT: vstrw.32 q4, [r0]
; CHECK-NEXT: vstrw.32 q2, [r0]
; CHECK-NEXT: str.w r12, [sp, #332]
; CHECK-NEXT: str.w r12, [sp, #324]
; CHECK-NEXT: .LBB1_3: @ %for.cond
; CHECK-NEXT: @ =>This Inner Loop Header: Depth=1
; CHECK-NEXT: b .LBB1_3

124
test/CodeGen/Thumb2/mve-postinc-dct.ll
View File

@ -1055,45 +1055,45 @@ define void @DCT_mve7(%struct.DCT_InstanceTypeDef* nocapture readonly %S, float*
; CHECK-NEXT: sub sp, #4
; CHECK-NEXT: .vsave {d8, d9, d10, d11, d12, d13, d14, d15}
; CHECK-NEXT: vpush {d8, d9, d10, d11, d12, d13, d14, d15}
; CHECK-NEXT: .pad #88
; CHECK-NEXT: sub sp, #88
; CHECK-NEXT: str r1, [sp, #28] @ 4-byte Spill
; CHECK-NEXT: .pad #72
; CHECK-NEXT: sub sp, #72
; CHECK-NEXT: str r1, [sp, #20] @ 4-byte Spill
; CHECK-NEXT: ldr r1, [r0, #4]
; CHECK-NEXT: subs r1, #7
; CHECK-NEXT: str r1, [sp, #24] @ 4-byte Spill
; CHECK-NEXT: str r1, [sp, #16] @ 4-byte Spill
; CHECK-NEXT: cmp r1, #2
; CHECK-NEXT: blo.w .LBB6_5
; CHECK-NEXT: @ %bb.1: @ %for.body.preheader
; CHECK-NEXT: ldr r3, [r0, #8]
; CHECK-NEXT: ldr r1, [r0]
; CHECK-NEXT: adds r0, r3, #3
; CHECK-NEXT: str r3, [sp, #20] @ 4-byte Spill
; CHECK-NEXT: str r3, [sp, #12] @ 4-byte Spill
; CHECK-NEXT: bic r0, r0, #3
; CHECK-NEXT: add.w r9, r1, r3, lsl #2
; CHECK-NEXT: subs r1, r0, #4
; CHECK-NEXT: movs r0, #1
; CHECK-NEXT: lsls r5, r3, #2
; CHECK-NEXT: add.w r1, r0, r1, lsr #2
; CHECK-NEXT: str r1, [sp, #16] @ 4-byte Spill
; CHECK-NEXT: str r1, [sp, #8] @ 4-byte Spill
; CHECK-NEXT: rsb r1, r3, r3, lsl #3
; CHECK-NEXT: lsls r1, r1, #2
; CHECK-NEXT: str r1, [sp, #12] @ 4-byte Spill
; CHECK-NEXT: str r1, [sp, #4] @ 4-byte Spill
; CHECK-NEXT: .LBB6_2: @ %for.body
; CHECK-NEXT: @ =>This Loop Header: Depth=1
; CHECK-NEXT: @ Child Loop BB6_3 Depth 2
; CHECK-NEXT: adds r1, r0, #6
; CHECK-NEXT: str r1, [sp, #44] @ 4-byte Spill
; CHECK-NEXT: adds r1, r0, #5
; CHECK-NEXT: str r1, [sp, #40] @ 4-byte Spill
; CHECK-NEXT: adds r1, r0, #4
; CHECK-NEXT: str r1, [sp, #36] @ 4-byte Spill
; CHECK-NEXT: adds r1, r0, #3
; CHECK-NEXT: ldr r7, [sp, #20] @ 4-byte Reload
; CHECK-NEXT: adds r1, r0, #5
; CHECK-NEXT: str r1, [sp, #32] @ 4-byte Spill
; CHECK-NEXT: adds r1, r0, #4
; CHECK-NEXT: str r1, [sp, #28] @ 4-byte Spill
; CHECK-NEXT: adds r1, r0, #3
; CHECK-NEXT: ldr r7, [sp, #12] @ 4-byte Reload
; CHECK-NEXT: str r1, [sp, #24] @ 4-byte Spill
; CHECK-NEXT: vmov.i32 q2, #0x0
; CHECK-NEXT: ldr r1, [sp, #28] @ 4-byte Reload
; CHECK-NEXT: ldr r1, [sp, #20] @ 4-byte Reload
; CHECK-NEXT: adds r4, r0, #2
; CHECK-NEXT: ldr r6, [sp, #16] @ 4-byte Reload
; CHECK-NEXT: ldr r6, [sp, #8] @ 4-byte Reload
; CHECK-NEXT: add.w r8, r0, #1
; CHECK-NEXT: mov r3, r9
; CHECK-NEXT: vmov q4, q2
@ -1102,7 +1102,7 @@ define void @DCT_mve7(%struct.DCT_InstanceTypeDef* nocapture readonly %S, float*
; CHECK-NEXT: vmov q6, q2
; CHECK-NEXT: vmov q1, q2
; CHECK-NEXT: mov r12, r7
; CHECK-NEXT: vstrw.32 q2, [sp, #64] @ 16-byte Spill
; CHECK-NEXT: vstrw.32 q2, [sp, #56] @ 16-byte Spill
; CHECK-NEXT: dls lr, r6
; CHECK-NEXT: .LBB6_3: @ %vector.body
; CHECK-NEXT: @ Parent Loop BB6_2 Depth=1
@ -1117,7 +1117,7 @@ define void @DCT_mve7(%struct.DCT_InstanceTypeDef* nocapture readonly %S, float*
; CHECK-NEXT: vpstt
; CHECK-NEXT: vldrwt.u32 q0, [r10]
; CHECK-NEXT: vfmat.f32 q6, q0, q7
; CHECK-NEXT: vstrw.32 q6, [sp, #48] @ 16-byte Spill
; CHECK-NEXT: vstrw.32 q6, [sp, #40] @ 16-byte Spill
; CHECK-NEXT: vpstt
; CHECK-NEXT: vldrwt.u32 q0, [r11]
; CHECK-NEXT: vfmat.f32 q1, q0, q7
@ -1129,17 +1129,17 @@ define void @DCT_mve7(%struct.DCT_InstanceTypeDef* nocapture readonly %S, float*
; CHECK-NEXT: vpst
; CHECK-NEXT: vldrwt.u32 q0, [r6]
; CHECK-NEXT: vmov q3, q1
; CHECK-NEXT: vldrw.u32 q1, [sp, #64] @ 16-byte Reload
; CHECK-NEXT: vldrw.u32 q1, [sp, #56] @ 16-byte Reload
; CHECK-NEXT: adds r7, r6, r5
; CHECK-NEXT: vpst
; CHECK-NEXT: vfmat.f32 q1, q0, q7
; CHECK-NEXT: vstrw.32 q1, [sp, #64] @ 16-byte Spill
; CHECK-NEXT: vstrw.32 q1, [sp, #56] @ 16-byte Spill
; CHECK-NEXT: vmov q1, q3
; CHECK-NEXT: vmov q3, q2
; CHECK-NEXT: vmov q2, q4
; CHECK-NEXT: vmov q4, q5
; CHECK-NEXT: vmov q5, q6
; CHECK-NEXT: vldrw.u32 q6, [sp, #48] @ 16-byte Reload
; CHECK-NEXT: vldrw.u32 q6, [sp, #40] @ 16-byte Reload
; CHECK-NEXT: sub.w r12, r12, #4
; CHECK-NEXT: adds r6, r7, r5
; CHECK-NEXT: vpstt
@ -1168,7 +1168,7 @@ define void @DCT_mve7(%struct.DCT_InstanceTypeDef* nocapture readonly %S, float*
; CHECK-NEXT: vadd.f32 s0, s2, s0
; CHECK-NEXT: vadd.f32 s10, s18, s19
; CHECK-NEXT: vadd.f32 s9, s16, s17
; CHECK-NEXT: vldrw.u32 q4, [sp, #64] @ 16-byte Reload
; CHECK-NEXT: vldrw.u32 q4, [sp, #56] @ 16-byte Reload
; CHECK-NEXT: vadd.f32 s2, s3, s1
; CHECK-NEXT: vadd.f32 s6, s18, s19
; CHECK-NEXT: vadd.f32 s5, s16, s17
@ -1183,25 +1183,25 @@ define void @DCT_mve7(%struct.DCT_InstanceTypeDef* nocapture readonly %S, float*
; CHECK-NEXT: vadd.f32 s8, s8, s20
; CHECK-NEXT: vadd.f32 s6, s5, s6
; CHECK-NEXT: vstr s4, [r1]
; CHECK-NEXT: ldr r1, [sp, #32] @ 4-byte Reload
; CHECK-NEXT: ldr r1, [sp, #24] @ 4-byte Reload
; CHECK-NEXT: add.w r1, r2, r1, lsl #2
; CHECK-NEXT: vstr s6, [r1]
; CHECK-NEXT: ldr r1, [sp, #36] @ 4-byte Reload
; CHECK-NEXT: ldr r1, [sp, #28] @ 4-byte Reload
; CHECK-NEXT: add.w r1, r2, r1, lsl #2
; CHECK-NEXT: vstr s12, [r1]
; CHECK-NEXT: ldr r1, [sp, #40] @ 4-byte Reload
; CHECK-NEXT: ldr r1, [sp, #32] @ 4-byte Reload
; CHECK-NEXT: add.w r1, r2, r1, lsl #2
; CHECK-NEXT: vstr s10, [r1]
; CHECK-NEXT: ldr r1, [sp, #44] @ 4-byte Reload
; CHECK-NEXT: ldr r1, [sp, #36] @ 4-byte Reload
; CHECK-NEXT: add.w r1, r2, r1, lsl #2
; CHECK-NEXT: vstr s8, [r1]
; CHECK-NEXT: ldr r1, [sp, #12] @ 4-byte Reload
; CHECK-NEXT: ldr r1, [sp, #4] @ 4-byte Reload
; CHECK-NEXT: add r9, r1
; CHECK-NEXT: ldr r1, [sp, #24] @ 4-byte Reload
; CHECK-NEXT: ldr r1, [sp, #16] @ 4-byte Reload
; CHECK-NEXT: cmp r0, r1
; CHECK-NEXT: blo.w .LBB6_2
; CHECK-NEXT: .LBB6_5: @ %for.cond.cleanup
; CHECK-NEXT: add sp, #88
; CHECK-NEXT: add sp, #72
; CHECK-NEXT: vpop {d8, d9, d10, d11, d12, d13, d14, d15}
; CHECK-NEXT: add sp, #4
; CHECK-NEXT: pop.w {r4, r5, r6, r7, r8, r9, r10, r11, pc}
@ -1345,44 +1345,44 @@ define void @DCT_mve8(%struct.DCT_InstanceTypeDef* nocapture readonly %S, float*
; CHECK-NEXT: sub sp, #4
; CHECK-NEXT: .vsave {d8, d9, d10, d11, d12, d13, d14, d15}
; CHECK-NEXT: vpush {d8, d9, d10, d11, d12, d13, d14, d15}
; CHECK-NEXT: .pad #104
; CHECK-NEXT: sub sp, #104
; CHECK-NEXT: str r1, [sp, #28] @ 4-byte Spill
; CHECK-NEXT: .pad #88
; CHECK-NEXT: sub sp, #88
; CHECK-NEXT: str r1, [sp, #20] @ 4-byte Spill
; CHECK-NEXT: ldr r1, [r0, #4]
; CHECK-NEXT: subs r1, #8
; CHECK-NEXT: str r1, [sp, #24] @ 4-byte Spill
; CHECK-NEXT: str r1, [sp, #16] @ 4-byte Spill
; CHECK-NEXT: cmp r1, #2
; CHECK-NEXT: blo.w .LBB7_5
; CHECK-NEXT: @ %bb.1: @ %for.body.preheader
; CHECK-NEXT: ldr r3, [r0, #8]
; CHECK-NEXT: ldr r1, [r0]
; CHECK-NEXT: adds r0, r3, #3
; CHECK-NEXT: str r3, [sp, #20] @ 4-byte Spill
; CHECK-NEXT: str r3, [sp, #12] @ 4-byte Spill
; CHECK-NEXT: bic r0, r0, #3
; CHECK-NEXT: add.w r12, r1, r3, lsl #2
; CHECK-NEXT: subs r1, r0, #4
; CHECK-NEXT: movs r0, #1
; CHECK-NEXT: lsls r5, r3, #2
; CHECK-NEXT: add.w r1, r0, r1, lsr #2
; CHECK-NEXT: str r1, [sp, #16] @ 4-byte Spill
; CHECK-NEXT: str r1, [sp, #8] @ 4-byte Spill
; CHECK-NEXT: lsls r1, r3, #5
; CHECK-NEXT: str r1, [sp, #12] @ 4-byte Spill
; CHECK-NEXT: str r1, [sp, #4] @ 4-byte Spill
; CHECK-NEXT: .LBB7_2: @ %for.body
; CHECK-NEXT: @ =>This Loop Header: Depth=1
; CHECK-NEXT: @ Child Loop BB7_3 Depth 2
; CHECK-NEXT: adds r1, r0, #7
; CHECK-NEXT: str r1, [sp, #44] @ 4-byte Spill
; CHECK-NEXT: adds r1, r0, #6
; CHECK-NEXT: str r1, [sp, #40] @ 4-byte Spill
; CHECK-NEXT: adds r1, r0, #5
; CHECK-NEXT: ldr r7, [sp, #20] @ 4-byte Reload
; CHECK-NEXT: str r1, [sp, #36] @ 4-byte Spill
; CHECK-NEXT: adds r1, r0, #4
; CHECK-NEXT: ldr.w r9, [sp, #28] @ 4-byte Reload
; CHECK-NEXT: vmov.i32 q3, #0x0
; CHECK-NEXT: ldr r6, [sp, #16] @ 4-byte Reload
; CHECK-NEXT: adds r4, r0, #3
; CHECK-NEXT: adds r1, r0, #6
; CHECK-NEXT: str r1, [sp, #32] @ 4-byte Spill
; CHECK-NEXT: adds r1, r0, #5
; CHECK-NEXT: ldr r7, [sp, #12] @ 4-byte Reload
; CHECK-NEXT: str r1, [sp, #28] @ 4-byte Spill
; CHECK-NEXT: adds r1, r0, #4
; CHECK-NEXT: ldr.w r9, [sp, #20] @ 4-byte Reload
; CHECK-NEXT: vmov.i32 q3, #0x0
; CHECK-NEXT: ldr r6, [sp, #8] @ 4-byte Reload
; CHECK-NEXT: adds r4, r0, #3
; CHECK-NEXT: str r1, [sp, #24] @ 4-byte Spill
; CHECK-NEXT: add.w r8, r0, #2
; CHECK-NEXT: adds r1, r0, #1
; CHECK-NEXT: mov r3, r12
@ -1392,8 +1392,8 @@ define void @DCT_mve8(%struct.DCT_InstanceTypeDef* nocapture readonly %S, float*
; CHECK-NEXT: vmov q7, q3
; CHECK-NEXT: vmov q2, q3
; CHECK-NEXT: mov r10, r7
; CHECK-NEXT: vstrw.32 q3, [sp, #64] @ 16-byte Spill
; CHECK-NEXT: vstrw.32 q3, [sp, #80] @ 16-byte Spill
; CHECK-NEXT: vstrw.32 q3, [sp, #56] @ 16-byte Spill
; CHECK-NEXT: vstrw.32 q3, [sp, #72] @ 16-byte Spill
; CHECK-NEXT: dls lr, r6
; CHECK-NEXT: .LBB7_3: @ %vector.body
; CHECK-NEXT: @ Parent Loop BB7_2 Depth=1
@ -1404,7 +1404,7 @@ define void @DCT_mve8(%struct.DCT_InstanceTypeDef* nocapture readonly %S, float*
; CHECK-NEXT: vldrwt.u32 q0, [r9], #16
; CHECK-NEXT: vldrwt.u32 q1, [r3], #16
; CHECK-NEXT: vfmat.f32 q6, q1, q0
; CHECK-NEXT: vstrw.32 q6, [sp, #48] @ 16-byte Spill
; CHECK-NEXT: vstrw.32 q6, [sp, #40] @ 16-byte Spill
; CHECK-NEXT: vpstt
; CHECK-NEXT: vldrwt.u32 q1, [r11]
; CHECK-NEXT: vfmat.f32 q7, q1, q0
@ -1415,23 +1415,23 @@ define void @DCT_mve8(%struct.DCT_InstanceTypeDef* nocapture readonly %S, float*
; CHECK-NEXT: vpst
; CHECK-NEXT: vldrwt.u32 q1, [r6]
; CHECK-NEXT: vmov q4, q2
; CHECK-NEXT: vldrw.u32 q2, [sp, #64] @ 16-byte Reload
; CHECK-NEXT: vldrw.u32 q2, [sp, #56] @ 16-byte Reload
; CHECK-NEXT: vpst
; CHECK-NEXT: vfmat.f32 q2, q1, q0
; CHECK-NEXT: vstrw.32 q2, [sp, #64] @ 16-byte Spill
; CHECK-NEXT: vstrw.32 q2, [sp, #56] @ 16-byte Spill
; CHECK-NEXT: adds r7, r6, r5
; CHECK-NEXT: vpst
; CHECK-NEXT: vldrwt.u32 q1, [r7]
; CHECK-NEXT: vldrw.u32 q2, [sp, #80] @ 16-byte Reload
; CHECK-NEXT: vldrw.u32 q2, [sp, #72] @ 16-byte Reload
; CHECK-NEXT: adds r6, r7, r5
; CHECK-NEXT: vpst
; CHECK-NEXT: vfmat.f32 q2, q1, q0
; CHECK-NEXT: vstrw.32 q2, [sp, #80] @ 16-byte Spill
; CHECK-NEXT: vstrw.32 q2, [sp, #72] @ 16-byte Spill
; CHECK-NEXT: vmov q2, q4
; CHECK-NEXT: vmov q4, q3
; CHECK-NEXT: vmov q3, q5
; CHECK-NEXT: vmov q5, q6
; CHECK-NEXT: vldrw.u32 q6, [sp, #48] @ 16-byte Reload
; CHECK-NEXT: vldrw.u32 q6, [sp, #40] @ 16-byte Reload
; CHECK-NEXT: adds r7, r6, r5
; CHECK-NEXT: vpstt
; CHECK-NEXT: vldrwt.u32 q1, [r6]
@ -1459,12 +1459,12 @@ define void @DCT_mve8(%struct.DCT_InstanceTypeDef* nocapture readonly %S, float*
; CHECK-NEXT: vadd.f32 s6, s24, s25
; CHECK-NEXT: vadd.f32 s14, s18, s19
; CHECK-NEXT: vadd.f32 s7, s16, s17
; CHECK-NEXT: vldrw.u32 q4, [sp, #64] @ 16-byte Reload
; CHECK-NEXT: vldrw.u32 q4, [sp, #56] @ 16-byte Reload
; CHECK-NEXT: vadd.f32 s8, s8, s9
; CHECK-NEXT: vadd.f32 s13, s10, s11
; CHECK-NEXT: vadd.f32 s10, s18, s19
; CHECK-NEXT: vadd.f32 s9, s16, s17
; CHECK-NEXT: vldrw.u32 q4, [sp, #80] @ 16-byte Reload
; CHECK-NEXT: vldrw.u32 q4, [sp, #72] @ 16-byte Reload
; CHECK-NEXT: vadd.f32 s0, s2, s0
; CHECK-NEXT: vadd.f32 s11, s18, s19
; CHECK-NEXT: vadd.f32 s15, s16, s17
@ -1484,26 +1484,26 @@ define void @DCT_mve8(%struct.DCT_InstanceTypeDef* nocapture readonly %S, float*
; CHECK-NEXT: vstr s10, [r1]
; CHECK-NEXT: add.w r1, r2, r4, lsl #2
; CHECK-NEXT: vstr s14, [r1]
; CHECK-NEXT: ldr r1, [sp, #32] @ 4-byte Reload
; CHECK-NEXT: ldr r1, [sp, #24] @ 4-byte Reload
; CHECK-NEXT: vadd.f32 s4, s3, s1
; CHECK-NEXT: add.w r1, r2, r1, lsl #2
; CHECK-NEXT: vstr s8, [r1]
; CHECK-NEXT: ldr r1, [sp, #36] @ 4-byte Reload
; CHECK-NEXT: ldr r1, [sp, #28] @ 4-byte Reload
; CHECK-NEXT: add.w r1, r2, r1, lsl #2
; CHECK-NEXT: vstr s12, [r1]
; CHECK-NEXT: ldr r1, [sp, #40] @ 4-byte Reload
; CHECK-NEXT: ldr r1, [sp, #32] @ 4-byte Reload
; CHECK-NEXT: add.w r1, r2, r1, lsl #2
; CHECK-NEXT: vstr s4, [r1]
; CHECK-NEXT: ldr r1, [sp, #44] @ 4-byte Reload
; CHECK-NEXT: ldr r1, [sp, #36] @ 4-byte Reload
; CHECK-NEXT: add.w r1, r2, r1, lsl #2
; CHECK-NEXT: vstr s6, [r1]
; CHECK-NEXT: ldr r1, [sp, #12] @ 4-byte Reload
; CHECK-NEXT: ldr r1, [sp, #4] @ 4-byte Reload
; CHECK-NEXT: add r12, r1
; CHECK-NEXT: ldr r1, [sp, #24] @ 4-byte Reload
; CHECK-NEXT: ldr r1, [sp, #16] @ 4-byte Reload
; CHECK-NEXT: cmp r0, r1
; CHECK-NEXT: blo.w .LBB7_2
; CHECK-NEXT: .LBB7_5: @ %for.cond.cleanup
; CHECK-NEXT: add sp, #104
; CHECK-NEXT: add sp, #88
; CHECK-NEXT: vpop {d8, d9, d10, d11, d12, d13, d14, d15}
; CHECK-NEXT: add sp, #4
; CHECK-NEXT: pop.w {r4, r5, r6, r7, r8, r9, r10, r11, pc}

46
test/CodeGen/Thumb2/mve-postinc-lsr.ll
View File

@ -1061,23 +1061,23 @@ define arm_aapcs_vfpcc void @_Z37_arm_radix4_butterfly_inverse_f32_mvePK21arm_cf
; CHECK-NEXT: sub sp, #4
; CHECK-NEXT: .vsave {d8, d9, d10, d11, d12, d13, d14, d15}
; CHECK-NEXT: vpush {d8, d9, d10, d11, d12, d13, d14, d15}
; CHECK-NEXT: .pad #56
; CHECK-NEXT: sub sp, #56
; CHECK-NEXT: .pad #40
; CHECK-NEXT: sub sp, #40
; CHECK-NEXT: cmp r2, #8
; CHECK-NEXT: str r1, [sp, #24] @ 4-byte Spill
; CHECK-NEXT: vstr s0, [sp, #8] @ 4-byte Spill
; CHECK-NEXT: str r1, [sp, #16] @ 4-byte Spill
; CHECK-NEXT: vstr s0, [sp] @ 4-byte Spill
; CHECK-NEXT: mov r1, r2
; CHECK-NEXT: str r2, [sp, #12] @ 4-byte Spill
; CHECK-NEXT: str r2, [sp, #4] @ 4-byte Spill
; CHECK-NEXT: blo .LBB7_9
; CHECK-NEXT: @ %bb.1:
; CHECK-NEXT: ldr r2, [sp, #12] @ 4-byte Reload
; CHECK-NEXT: ldr r2, [sp, #4] @ 4-byte Reload
; CHECK-NEXT: mov.w r11, #0
; CHECK-NEXT: mov.w r12, #1
; CHECK-NEXT: str r2, [sp, #20] @ 4-byte Spill
; CHECK-NEXT: str r2, [sp, #12] @ 4-byte Spill
; CHECK-NEXT: lsrs r1, r2, #2
; CHECK-NEXT: b .LBB7_3
; CHECK-NEXT: .LBB7_2: @ in Loop: Header=BB7_3 Depth=1
; CHECK-NEXT: ldr r2, [sp, #16] @ 4-byte Reload
; CHECK-NEXT: ldr r2, [sp, #8] @ 4-byte Reload
; CHECK-NEXT: add.w r11, r11, #1
; CHECK-NEXT: lsl.w r12, r12, #2
; CHECK-NEXT: cmp r2, #7
@ -1086,27 +1086,27 @@ define arm_aapcs_vfpcc void @_Z37_arm_radix4_butterfly_inverse_f32_mvePK21arm_cf
; CHECK-NEXT: .LBB7_3: @ =>This Loop Header: Depth=1
; CHECK-NEXT: @ Child Loop BB7_6 Depth 2
; CHECK-NEXT: @ Child Loop BB7_7 Depth 3
; CHECK-NEXT: str r1, [sp, #16] @ 4-byte Spill
; CHECK-NEXT: str r1, [sp, #8] @ 4-byte Spill
; CHECK-NEXT: cmp.w r12, #1
; CHECK-NEXT: ldr r1, [sp, #20] @ 4-byte Reload
; CHECK-NEXT: ldr r1, [sp, #12] @ 4-byte Reload
; CHECK-NEXT: lsr.w r2, r1, #2
; CHECK-NEXT: str r2, [sp, #20] @ 4-byte Spill
; CHECK-NEXT: str r2, [sp, #12] @ 4-byte Spill
; CHECK-NEXT: blt .LBB7_2
; CHECK-NEXT: @ %bb.4: @ in Loop: Header=BB7_3 Depth=1
; CHECK-NEXT: lsrs r2, r1, #3
; CHECK-NEXT: str r2, [sp, #32] @ 4-byte Spill
; CHECK-NEXT: str r2, [sp, #24] @ 4-byte Spill
; CHECK-NEXT: beq .LBB7_2
; CHECK-NEXT: @ %bb.5: @ %.preheader
; CHECK-NEXT: @ in Loop: Header=BB7_3 Depth=1
; CHECK-NEXT: ldr r2, [sp, #20] @ 4-byte Reload
; CHECK-NEXT: ldr r2, [sp, #12] @ 4-byte Reload
; CHECK-NEXT: lsls r1, r1, #1
; CHECK-NEXT: movs r4, #0
; CHECK-NEXT: str r1, [sp, #28] @ 4-byte Spill
; CHECK-NEXT: str r1, [sp, #20] @ 4-byte Spill
; CHECK-NEXT: lsl.w r10, r2, #1
; CHECK-NEXT: .LBB7_6: @ Parent Loop BB7_3 Depth=1
; CHECK-NEXT: @ => This Loop Header: Depth=2
; CHECK-NEXT: @ Child Loop BB7_7 Depth 3
; CHECK-NEXT: ldr r1, [sp, #28] @ 4-byte Reload
; CHECK-NEXT: ldr r1, [sp, #20] @ 4-byte Reload
; CHECK-NEXT: ldrd lr, r2, [r0, #16]
; CHECK-NEXT: ldrd r3, r8, [r0, #24]
; CHECK-NEXT: muls r1, r4, r1
@ -1115,11 +1115,11 @@ define arm_aapcs_vfpcc void @_Z37_arm_radix4_butterfly_inverse_f32_mvePK21arm_cf
; CHECK-NEXT: ldr.w r3, [r3, r11, lsl #2]
; CHECK-NEXT: ldr.w r6, [lr, r11, lsl #2]
; CHECK-NEXT: add.w r7, r7, r2, lsl #2
; CHECK-NEXT: ldr r2, [sp, #24] @ 4-byte Reload
; CHECK-NEXT: ldr r2, [sp, #16] @ 4-byte Reload
; CHECK-NEXT: add.w r5, r5, r3, lsl #2
; CHECK-NEXT: add.w r1, r2, r1, lsl #2
; CHECK-NEXT: add.w r3, r8, r6, lsl #2
; CHECK-NEXT: ldr r6, [sp, #32] @ 4-byte Reload
; CHECK-NEXT: ldr r6, [sp, #24] @ 4-byte Reload
; CHECK-NEXT: add.w r2, r1, r10, lsl #2
; CHECK-NEXT: add.w r8, r2, r10, lsl #2
; CHECK-NEXT: add.w r9, r8, r10, lsl #2
@ -1159,14 +1159,14 @@ define arm_aapcs_vfpcc void @_Z37_arm_radix4_butterfly_inverse_f32_mvePK21arm_cf
; CHECK-NEXT: bne .LBB7_6
; CHECK-NEXT: b .LBB7_2
; CHECK-NEXT: .LBB7_9:
; CHECK-NEXT: ldr r0, [sp, #12] @ 4-byte Reload
; CHECK-NEXT: ldr r0, [sp, #4] @ 4-byte Reload
; CHECK-NEXT: lsrs r0, r0, #3
; CHECK-NEXT: wls lr, r0, .LBB7_12
; CHECK-NEXT: @ %bb.10:
; CHECK-NEXT: adr r0, .LCPI7_0
; CHECK-NEXT: vldr s0, [sp, #8] @ 4-byte Reload
; CHECK-NEXT: vldr s0, [sp] @ 4-byte Reload
; CHECK-NEXT: vldrw.u32 q1, [r0]
; CHECK-NEXT: ldr r0, [sp, #24] @ 4-byte Reload
; CHECK-NEXT: ldr r0, [sp, #16] @ 4-byte Reload
; CHECK-NEXT: vadd.i32 q1, q1, r0
; CHECK-NEXT: vmov r0, s0
; CHECK-NEXT: vldrw.u32 q2, [q1, #64]!
@ -1182,7 +1182,7 @@ define arm_aapcs_vfpcc void @_Z37_arm_radix4_butterfly_inverse_f32_mvePK21arm_cf
; CHECK-NEXT: vsub.f32 q2, q0, q3
; CHECK-NEXT: vmul.f32 q7, q7, r0
; CHECK-NEXT: vadd.f32 q3, q0, q3
; CHECK-NEXT: vstrw.32 q7, [sp, #32] @ 16-byte Spill
; CHECK-NEXT: vstrw.32 q7, [sp, #24] @ 16-byte Spill
; CHECK-NEXT: vcadd.f32 q7, q6, q5, #90
; CHECK-NEXT: vmul.f32 q4, q2, r0
; CHECK-NEXT: vldrw.u32 q2, [q1, #64]!
@ -1192,11 +1192,11 @@ define arm_aapcs_vfpcc void @_Z37_arm_radix4_butterfly_inverse_f32_mvePK21arm_cf
; CHECK-NEXT: vstrw.32 q3, [q1, #-64]
; CHECK-NEXT: vstrw.32 q4, [q1, #-56]
; CHECK-NEXT: vstrw.32 q5, [q1, #-48]
; CHECK-NEXT: vldrw.u32 q3, [sp, #32] @ 16-byte Reload
; CHECK-NEXT: vldrw.u32 q3, [sp, #24] @ 16-byte Reload
; CHECK-NEXT: vstrw.32 q3, [q1, #-40]
; CHECK-NEXT: le lr, .LBB7_11
; CHECK-NEXT: .LBB7_12:
; CHECK-NEXT: add sp, #56
; CHECK-NEXT: add sp, #40
; CHECK-NEXT: vpop {d8, d9, d10, d11, d12, d13, d14, d15}
; CHECK-NEXT: add sp, #4
; CHECK-NEXT: pop.w {r4, r5, r6, r7, r8, r9, r10, r11, pc}

6
test/CodeGen/Thumb2/mve-scatter-increment.ll
View File

@ -133,8 +133,8 @@ define arm_aapcs_vfpcc void @scatter_inc_v4i32_complex(<4 x i32> %data1, <4 x i3
; CHECK-NEXT: push {r4, lr}
; CHECK-NEXT: .vsave {d8, d9, d10, d11, d12, d13, d14, d15}
; CHECK-NEXT: vpush {d8, d9, d10, d11, d12, d13, d14, d15}
; CHECK-NEXT: .pad #24
; CHECK-NEXT: sub sp, #24
; CHECK-NEXT: .pad #16
; CHECK-NEXT: sub sp, #16
; CHECK-NEXT: cmp r1, #1
; CHECK-NEXT: blt .LBB3_5
; CHECK-NEXT: @ %bb.1: @ %vector.ph.preheader
@ -171,7 +171,7 @@ define arm_aapcs_vfpcc void @scatter_inc_v4i32_complex(<4 x i32> %data1, <4 x i3
; CHECK-NEXT: cmp r2, r1
; CHECK-NEXT: bne .LBB3_2
; CHECK-NEXT: .LBB3_5: @ %for.cond.cleanup
; CHECK-NEXT: add sp, #24
; CHECK-NEXT: add sp, #16
; CHECK-NEXT: vpop {d8, d9, d10, d11, d12, d13, d14, d15}
; CHECK-NEXT: pop {r4, pc}
; CHECK-NEXT: .p2align 4

6
test/CodeGen/Thumb2/mve-vecreduce-mlapred.ll
View File

@ -768,8 +768,8 @@ define arm_aapcs_vfpcc i64 @add_v16i8_v16i64_zext(<16 x i8> %x, <16 x i8> %y, <1
; CHECK-NEXT: push {r4, lr}
; CHECK-NEXT: .vsave {d8, d9, d10, d11, d12, d13, d14, d15}
; CHECK-NEXT: vpush {d8, d9, d10, d11, d12, d13, d14, d15}
; CHECK-NEXT: .pad #40
; CHECK-NEXT: sub sp, #40
; CHECK-NEXT: .pad #32
; CHECK-NEXT: sub sp, #32
; CHECK-NEXT: vmov q4, q0
; CHECK-NEXT: vcmp.i8 eq, q2, zr
; CHECK-NEXT: vmov.i8 q2, #0xff
@ -1080,7 +1080,7 @@ define arm_aapcs_vfpcc i64 @add_v16i8_v16i64_zext(<16 x i8> %x, <16 x i8> %y, <1
; CHECK-NEXT: vmov r2, r3, d1
; CHECK-NEXT: adds r0, r0, r2
; CHECK-NEXT: adcs r1, r3
; CHECK-NEXT: add sp, #40
; CHECK-NEXT: add sp, #32
; CHECK-NEXT: vpop {d8, d9, d10, d11, d12, d13, d14, d15}
; CHECK-NEXT: pop {r4, pc}
entry:

66
test/CodeGen/Thumb2/mve-vld4.ll
View File

@ -115,8 +115,8 @@ define void @vld4_v16i32(<64 x i32> *%src, <16 x i32> *%dst) {
; CHECK-NEXT: push {r4, r5}
; CHECK-NEXT: .vsave {d8, d9, d10, d11, d12, d13}
; CHECK-NEXT: vpush {d8, d9, d10, d11, d12, d13}
; CHECK-NEXT: .pad #136
; CHECK-NEXT: sub sp, #136
; CHECK-NEXT: .pad #112
; CHECK-NEXT: sub sp, #112
; CHECK-NEXT: vld40.32 {q0, q1, q2, q3}, [r0]
; CHECK-NEXT: mov r2, r0
; CHECK-NEXT: add.w r3, r0, #192
@ -127,23 +127,23 @@ define void @vld4_v16i32(<64 x i32> *%src, <16 x i32> *%dst) {
; CHECK-NEXT: @ kill: def $q0 killed $q0 killed $q0_q1_q2_q3
; CHECK-NEXT: vadd.i32 q4, q2, q3
; CHECK-NEXT: vadd.i32 q0, q0, q1
; CHECK-NEXT: vstrw.32 q4, [sp, #112] @ 16-byte Spill
; CHECK-NEXT: vstrw.32 q4, [sp, #96] @ 16-byte Spill
; CHECK-NEXT: vld40.32 {q1, q2, q3, q4}, [r3]
; CHECK-NEXT: vstrw.32 q0, [sp, #96] @ 16-byte Spill
; CHECK-NEXT: vldrw.u32 q6, [sp, #112] @ 16-byte Reload
; CHECK-NEXT: vstrw.32 q0, [sp, #80] @ 16-byte Spill
; CHECK-NEXT: vldrw.u32 q6, [sp, #96] @ 16-byte Reload
; CHECK-NEXT: vld41.32 {q1, q2, q3, q4}, [r3]
; CHECK-NEXT: vldrw.u32 q5, [sp, #96] @ 16-byte Reload
; CHECK-NEXT: vldrw.u32 q5, [sp, #80] @ 16-byte Reload
; CHECK-NEXT: vld42.32 {q1, q2, q3, q4}, [r3]
; CHECK-NEXT: vadd.i32 q6, q5, q6
; CHECK-NEXT: vstrw.32 q6, [sp, #112] @ 16-byte Spill
; CHECK-NEXT: vstrw.32 q6, [sp, #96] @ 16-byte Spill
; CHECK-NEXT: vld43.32 {q1, q2, q3, q4}, [r3]
; CHECK-NEXT: vstrw.32 q4, [sp, #80] @ 16-byte Spill
; CHECK-NEXT: vstrw.32 q4, [sp, #64] @ 16-byte Spill
; CHECK-NEXT: vmov q0, q1
; CHECK-NEXT: vldrw.u32 q5, [sp, #80] @ 16-byte Reload
; CHECK-NEXT: vldrw.u32 q5, [sp, #64] @ 16-byte Reload
; CHECK-NEXT: vadd.i32 q0, q0, q2
; CHECK-NEXT: vadd.i32 q1, q3, q5
; CHECK-NEXT: vadd.i32 q0, q0, q1
; CHECK-NEXT: vstrw.32 q0, [sp, #96] @ 16-byte Spill
; CHECK-NEXT: vstrw.32 q0, [sp, #80] @ 16-byte Spill
; CHECK-NEXT: vld40.32 {q0, q1, q2, q3}, [r2]
; CHECK-NEXT: vld41.32 {q0, q1, q2, q3}, [r2]
; CHECK-NEXT: vld42.32 {q0, q1, q2, q3}, [r2]
@ -154,9 +154,9 @@ define void @vld4_v16i32(<64 x i32> *%src, <16 x i32> *%dst) {
; CHECK-NEXT: vld42.32 {q0, q1, q2, q3}, [r0]
; CHECK-NEXT: vld43.32 {q0, q1, q2, q3}, [r0]
; CHECK-NEXT: @ kill: def $q0 killed $q0 killed $q0_q1_q2_q3
; CHECK-NEXT: vstrw.32 q3, [sp, #80] @ 16-byte Spill
; CHECK-NEXT: vstrw.32 q3, [sp, #64] @ 16-byte Spill
; CHECK-NEXT: vmov q5, q1
; CHECK-NEXT: vldrw.u32 q1, [sp, #80] @ 16-byte Reload
; CHECK-NEXT: vldrw.u32 q1, [sp, #64] @ 16-byte Reload
; CHECK-NEXT: vadd.i32 q0, q0, q5
; CHECK-NEXT: vldmia sp, {d6, d7, d8, d9, d10, d11, d12, d13} @ 64-byte Reload
; CHECK-NEXT: vadd.i32 q1, q2, q1
@ -165,13 +165,13 @@ define void @vld4_v16i32(<64 x i32> *%src, <16 x i32> *%dst) {
; CHECK-NEXT: vadd.i32 q0, q0, q1
; CHECK-NEXT: vadd.i32 q1, q5, q6
; CHECK-NEXT: vadd.i32 q1, q2, q1
; CHECK-NEXT: vldrw.u32 q2, [sp, #96] @ 16-byte Reload
; CHECK-NEXT: vldrw.u32 q2, [sp, #80] @ 16-byte Reload
; CHECK-NEXT: vstrw.32 q0, [r1, #32]
; CHECK-NEXT: vldrw.u32 q0, [sp, #112] @ 16-byte Reload
; CHECK-NEXT: vldrw.u32 q0, [sp, #96] @ 16-byte Reload
; CHECK-NEXT: vstrw.32 q2, [r1, #48]
; CHECK-NEXT: vstrw.32 q1, [r1, #16]
; CHECK-NEXT: vstrw.32 q0, [r1]
; CHECK-NEXT: add sp, #136
; CHECK-NEXT: add sp, #112
; CHECK-NEXT: vpop {d8, d9, d10, d11, d12, d13}
; CHECK-NEXT: pop {r4, r5}
; CHECK-NEXT: bx lr
@ -914,8 +914,8 @@ define void @vld4_v16f32(<64 x float> *%src, <16 x float> *%dst) {
; CHECK-NEXT: push {r4, r5}
; CHECK-NEXT: .vsave {d8, d9, d10, d11, d12, d13}
; CHECK-NEXT: vpush {d8, d9, d10, d11, d12, d13}
; CHECK-NEXT: .pad #136
; CHECK-NEXT: sub sp, #136
; CHECK-NEXT: .pad #112
; CHECK-NEXT: sub sp, #112
; CHECK-NEXT: vld40.32 {q0, q1, q2, q3}, [r0]
; CHECK-NEXT: mov r2, r0
; CHECK-NEXT: add.w r3, r0, #192
@ -926,23 +926,23 @@ define void @vld4_v16f32(<64 x float> *%src, <16 x float> *%dst) {
; CHECK-NEXT: @ kill: def $q0 killed $q0 killed $q0_q1_q2_q3
; CHECK-NEXT: vadd.f32 q4, q2, q3
; CHECK-NEXT: vadd.f32 q0, q0, q1
; CHECK-NEXT: vstrw.32 q4, [sp, #112] @ 16-byte Spill
; CHECK-NEXT: vstrw.32 q4, [sp, #96] @ 16-byte Spill
; CHECK-NEXT: vld40.32 {q1, q2, q3, q4}, [r3]
; CHECK-NEXT: vstrw.32 q0, [sp, #96] @ 16-byte Spill
; CHECK-NEXT: vldrw.u32 q6, [sp, #112] @ 16-byte Reload
; CHECK-NEXT: vstrw.32 q0, [sp, #80] @ 16-byte Spill
; CHECK-NEXT: vldrw.u32 q6, [sp, #96] @ 16-byte Reload
; CHECK-NEXT: vld41.32 {q1, q2, q3, q4}, [r3]
; CHECK-NEXT: vldrw.u32 q5, [sp, #96] @ 16-byte Reload
; CHECK-NEXT: vldrw.u32 q5, [sp, #80] @ 16-byte Reload
; CHECK-NEXT: vld42.32 {q1, q2, q3, q4}, [r3]
; CHECK-NEXT: vadd.f32 q6, q5, q6
; CHECK-NEXT: vstrw.32 q6, [sp, #112] @ 16-byte Spill
; CHECK-NEXT: vstrw.32 q6, [sp, #96] @ 16-byte Spill
; CHECK-NEXT: vld43.32 {q1, q2, q3, q4}, [r3]
; CHECK-NEXT: vstrw.32 q4, [sp, #80] @ 16-byte Spill
; CHECK-NEXT: vstrw.32 q4, [sp, #64] @ 16-byte Spill
; CHECK-NEXT: vmov q0, q1
; CHECK-NEXT: vldrw.u32 q5, [sp, #80] @ 16-byte Reload
; CHECK-NEXT: vldrw.u32 q5, [sp, #64] @ 16-byte Reload
; CHECK-NEXT: vadd.f32 q0, q0, q2
; CHECK-NEXT: vadd.f32 q1, q3, q5
; CHECK-NEXT: vadd.f32 q0, q0, q1
; CHECK-NEXT: vstrw.32 q0, [sp, #96] @ 16-byte Spill
; CHECK-NEXT: vstrw.32 q0, [sp, #80] @ 16-byte Spill
; CHECK-NEXT: vld40.32 {q0, q1, q2, q3}, [r2]
; CHECK-NEXT: vld41.32 {q0, q1, q2, q3}, [r2]
; CHECK-NEXT: vld42.32 {q0, q1, q2, q3}, [r2]
@ -953,9 +953,9 @@ define void @vld4_v16f32(<64 x float> *%src, <16 x float> *%dst) {
; CHECK-NEXT: vld42.32 {q0, q1, q2, q3}, [r0]
; CHECK-NEXT: vld43.32 {q0, q1, q2, q3}, [r0]
; CHECK-NEXT: @ kill: def $q0 killed $q0 killed $q0_q1_q2_q3
; CHECK-NEXT: vstrw.32 q3, [sp, #80] @ 16-byte Spill
; CHECK-NEXT: vstrw.32 q3, [sp, #64] @ 16-byte Spill
; CHECK-NEXT: vmov q5, q1
; CHECK-NEXT: vldrw.u32 q1, [sp, #80] @ 16-byte Reload
; CHECK-NEXT: vldrw.u32 q1, [sp, #64] @ 16-byte Reload
; CHECK-NEXT: vadd.f32 q0, q0, q5
; CHECK-NEXT: vldmia sp, {d6, d7, d8, d9, d10, d11, d12, d13} @ 64-byte Reload
; CHECK-NEXT: vadd.f32 q1, q2, q1
@ -964,13 +964,13 @@ define void @vld4_v16f32(<64 x float> *%src, <16 x float> *%dst) {
; CHECK-NEXT: vadd.f32 q0, q0, q1
; CHECK-NEXT: vadd.f32 q1, q5, q6
; CHECK-NEXT: vadd.f32 q1, q2, q1
; CHECK-NEXT: vldrw.u32 q2, [sp, #96] @ 16-byte Reload
; CHECK-NEXT: vldrw.u32 q2, [sp, #80] @ 16-byte Reload
; CHECK-NEXT: vstrw.32 q0, [r1, #32]
; CHECK-NEXT: vldrw.u32 q0, [sp, #112] @ 16-byte Reload
; CHECK-NEXT: vldrw.u32 q0, [sp, #96] @ 16-byte Reload
; CHECK-NEXT: vstrw.32 q2, [r1, #48]
; CHECK-NEXT: vstrw.32 q1, [r1, #16]
; CHECK-NEXT: vstrw.32 q0, [r1]
; CHECK-NEXT: add sp, #136
; CHECK-NEXT: add sp, #112
; CHECK-NEXT: vpop {d8, d9, d10, d11, d12, d13}
; CHECK-NEXT: pop {r4, r5}
; CHECK-NEXT: bx lr
@ -1147,8 +1147,8 @@ define void @vld4_v16f16(<64 x half> *%src, <16 x half> *%dst) {
; CHECK-NEXT: push {r4, r5}
; CHECK-NEXT: .vsave {d8, d9, d10, d11, d12, d13, d14, d15}
; CHECK-NEXT: vpush {d8, d9, d10, d11, d12, d13, d14, d15}
; CHECK-NEXT: .pad #88
; CHECK-NEXT: sub sp, #88
; CHECK-NEXT: .pad #80
; CHECK-NEXT: sub sp, #80
; CHECK-NEXT: vld40.16 {q0, q1, q2, q3}, [r0]
; CHECK-NEXT: vld41.16 {q0, q1, q2, q3}, [r0]
; CHECK-NEXT: vld42.16 {q0, q1, q2, q3}, [r0]
@ -1171,7 +1171,7 @@ define void @vld4_v16f16(<64 x half> *%src, <16 x half> *%dst) {
; CHECK-NEXT: vadd.f16 q0, q0, q1
; CHECK-NEXT: vadd.f16 q0, q0, q4
; CHECK-NEXT: vstrw.32 q0, [r1]
; CHECK-NEXT: add sp, #88
; CHECK-NEXT: add sp, #80
; CHECK-NEXT: vpop {d8, d9, d10, d11, d12, d13, d14, d15}
; CHECK-NEXT: pop {r4, r5}
; CHECK-NEXT: bx lr

8
test/CodeGen/Thumb2/mve-vldst4.ll
View File

@ -8,14 +8,14 @@ define void @vldst4(half* nocapture readonly %pIn, half* nocapture %pOut, i32 %n
; CHECK-NEXT: push {r7, lr}
; CHECK-NEXT: .vsave {d8, d9, d10, d11, d12, d13, d14, d15}
; CHECK-NEXT: vpush {d8, d9, d10, d11, d12, d13, d14, d15}
; CHECK-NEXT: .pad #24
; CHECK-NEXT: sub sp, #24
; CHECK-NEXT: .pad #16
; CHECK-NEXT: sub sp, #16
; CHECK-NEXT: mul r12, r3, r2
; CHECK-NEXT: lsrs.w r2, r12, #2
; CHECK-NEXT: beq.w .LBB0_3
; CHECK-NEXT: @ %bb.1: @ %vector.ph
; CHECK-NEXT: mvn r3, #7
; CHECK-NEXT: ldr r2, [sp, #96]
; CHECK-NEXT: ldr r2, [sp, #88]
; CHECK-NEXT: and.w r3, r3, r12, lsr #2
; CHECK-NEXT: sub.w r12, r3, #8
; CHECK-NEXT: movs r3, #1
@ -123,7 +123,7 @@ define void @vldst4(half* nocapture readonly %pIn, half* nocapture %pOut, i32 %n
; CHECK-NEXT: vstrh.16 q4, [r1, #-48]
; CHECK-NEXT: le lr, .LBB0_2
; CHECK-NEXT: .LBB0_3: @ %while.end
; CHECK-NEXT: add sp, #24
; CHECK-NEXT: add sp, #16
; CHECK-NEXT: vpop {d8, d9, d10, d11, d12, d13, d14, d15}
; CHECK-NEXT: pop {r7, pc}
entry:

12
test/CodeGen/Thumb2/mve-vst4.ll
View File

@ -119,8 +119,8 @@ define void @vst4_v16i32(<16 x i32> *%src, <64 x i32> *%dst) {
; CHECK-NEXT: push {r4, r5}
; CHECK-NEXT: .vsave {d8, d9, d10, d11, d12, d13, d14, d15}
; CHECK-NEXT: vpush {d8, d9, d10, d11, d12, d13, d14, d15}
; CHECK-NEXT: .pad #216
; CHECK-NEXT: sub sp, #216
; CHECK-NEXT: .pad #192
; CHECK-NEXT: sub sp, #192
; CHECK-NEXT: vldrw.u32 q2, [r0, #144]
; CHECK-NEXT: add r2, sp, #64
; CHECK-NEXT: vldrw.u32 q1, [r0, #80]
@ -180,7 +180,7 @@ define void @vst4_v16i32(<16 x i32> *%src, <64 x i32> *%dst) {
; CHECK-NEXT: vst42.32 {q0, q1, q2, q3}, [r0]
; CHECK-NEXT: vst43.32 {q4, q5, q6, q7}, [r1]
; CHECK-NEXT: vst43.32 {q0, q1, q2, q3}, [r0]
; CHECK-NEXT: add sp, #216
; CHECK-NEXT: add sp, #192
; CHECK-NEXT: vpop {d8, d9, d10, d11, d12, d13, d14, d15}
; CHECK-NEXT: pop {r4, r5}
; CHECK-NEXT: bx lr
@ -889,8 +889,8 @@ define void @vst4_v16f32(<16 x float> *%src, <64 x float> *%dst) {
; CHECK-NEXT: push {r4, r5}
; CHECK-NEXT: .vsave {d8, d9, d10, d11, d12, d13, d14, d15}
; CHECK-NEXT: vpush {d8, d9, d10, d11, d12, d13, d14, d15}
; CHECK-NEXT: .pad #216
; CHECK-NEXT: sub sp, #216
; CHECK-NEXT: .pad #192
; CHECK-NEXT: sub sp, #192
; CHECK-NEXT: vldrw.u32 q2, [r0, #144]
; CHECK-NEXT: add r2, sp, #64
; CHECK-NEXT: vldrw.u32 q1, [r0, #80]
@ -950,7 +950,7 @@ define void @vst4_v16f32(<16 x float> *%src, <64 x float> *%dst) {
; CHECK-NEXT: vst42.32 {q0, q1, q2, q3}, [r0]
; CHECK-NEXT: vst43.32 {q4, q5, q6, q7}, [r1]
; CHECK-NEXT: vst43.32 {q0, q1, q2, q3}, [r0]
; CHECK-NEXT: add sp, #216
; CHECK-NEXT: add sp, #192
; CHECK-NEXT: vpop {d8, d9, d10, d11, d12, d13, d14, d15}
; CHECK-NEXT: pop {r4, r5}
; CHECK-NEXT: bx lr

56
test/CodeGen/X86/AMX/amx-across-func.ll
View File

@ -23,41 +23,41 @@ define dso_local void @test_api(i16 signext %0, i16 signext %1) nounwind {
; CHECK-NEXT: pushq %r15
; CHECK-NEXT: pushq %r14
; CHECK-NEXT: pushq %rbx
; CHECK-NEXT: subq $4056, %rsp # imm = 0xFD8
; CHECK-NEXT: subq $2120, %rsp # imm = 0x848
; CHECK-NEXT: movl %esi, %ebx
; CHECK-NEXT: movl %edi, %ebp
; CHECK-NEXT: vpxord %zmm0, %zmm0, %zmm0
; CHECK-NEXT: vmovdqu64 %zmm0, {{[0-9]+}}(%rsp)
; CHECK-NEXT: movb $1, {{[0-9]+}}(%rsp)
; CHECK-NEXT: vmovdqu64 %zmm0, (%rsp)
; CHECK-NEXT: movb $1, (%rsp)
; CHECK-NEXT: movw $8, {{[0-9]+}}(%rsp)
; CHECK-NEXT: movb $8, {{[0-9]+}}(%rsp)
; CHECK-NEXT: movw %bx, {{[0-9]+}}(%rsp)
; CHECK-NEXT: movb %bpl, {{[0-9]+}}(%rsp)
; CHECK-NEXT: movw %bx, {{[0-9]+}}(%rsp)
; CHECK-NEXT: movb %bpl, {{[0-9]+}}(%rsp)
; CHECK-NEXT: ldtilecfg {{[0-9]+}}(%rsp)
; CHECK-NEXT: ldtilecfg (%rsp)
; CHECK-NEXT: movl $buf, %eax
; CHECK-NEXT: movl $32, %r14d
; CHECK-NEXT: movw $8, %r15w
; CHECK-NEXT: tileloadd (%rax,%r14), %tmm1
; CHECK-NEXT: movabsq $64, %rax
; CHECK-NEXT: tilestored %tmm1, 2048(%rsp,%rax) # 1024-byte Folded Spill
; CHECK-NEXT: tilestored %tmm1, 1088(%rsp,%rax) # 1024-byte Folded Spill
; CHECK-NEXT: movl $buf+1024, %eax
; CHECK-NEXT: tileloadd (%rax,%r14), %tmm2
; CHECK-NEXT: movabsq $64, %rax
; CHECK-NEXT: tilestored %tmm2, 1024(%rsp,%rax) # 1024-byte Folded Spill
; CHECK-NEXT: tilestored %tmm2, 64(%rsp,%rax) # 1024-byte Folded Spill
; CHECK-NEXT: vzeroupper
; CHECK-NEXT: callq foo
; CHECK-NEXT: ldtilecfg {{[0-9]+}}(%rsp)
; CHECK-NEXT: ldtilecfg (%rsp)
; CHECK-NEXT: movl $buf+2048, %eax
; CHECK-NEXT: tileloadd (%rax,%r14), %tmm0
; CHECK-NEXT: movabsq $64, %rcx
; CHECK-NEXT: tileloadd 2048(%rsp,%rcx), %tmm1 # 1024-byte Folded Reload
; CHECK-NEXT: tileloadd 1088(%rsp,%rcx), %tmm1 # 1024-byte Folded Reload
; CHECK-NEXT: movabsq $64, %rcx
; CHECK-NEXT: tileloadd 1024(%rsp,%rcx), %tmm2 # 1024-byte Folded Reload
; CHECK-NEXT: tileloadd 64(%rsp,%rcx), %tmm2 # 1024-byte Folded Reload
; CHECK-NEXT: tdpbssd %tmm2, %tmm1, %tmm0
; CHECK-NEXT: tilestored %tmm0, (%rax,%r14)
; CHECK-NEXT: addq $4056, %rsp # imm = 0xFD8
; CHECK-NEXT: addq $2120, %rsp # imm = 0x848
; CHECK-NEXT: popq %rbx
; CHECK-NEXT: popq %r14
; CHECK-NEXT: popq %r15
@ -111,16 +111,16 @@ define dso_local i32 @test_loop(i32 %0) nounwind {
; CHECK-NEXT: pushq %r13
; CHECK-NEXT: pushq %r12
; CHECK-NEXT: pushq %rbx
; CHECK-NEXT: subq $3016, %rsp # imm = 0xBC8
; CHECK-NEXT: subq $1096, %rsp # imm = 0x448
; CHECK-NEXT: movl %edi, %r14d
; CHECK-NEXT: vpxord %zmm0, %zmm0, %zmm0
; CHECK-NEXT: vmovdqu64 %zmm0, {{[0-9]+}}(%rsp)
; CHECK-NEXT: movb $1, {{[0-9]+}}(%rsp)
; CHECK-NEXT: vmovdqu64 %zmm0, (%rsp)
; CHECK-NEXT: movb $1, (%rsp)
; CHECK-NEXT: movb $8, {{[0-9]+}}(%rsp)
; CHECK-NEXT: movw $8, {{[0-9]+}}(%rsp)
; CHECK-NEXT: vzeroupper
; CHECK-NEXT: callq foo
; CHECK-NEXT: ldtilecfg {{[0-9]+}}(%rsp)
; CHECK-NEXT: ldtilecfg (%rsp)
; CHECK-NEXT: testl %r14d, %r14d
; CHECK-NEXT: jg .LBB2_4
; CHECK-NEXT: # %bb.1: # %.preheader
@ -133,14 +133,14 @@ define dso_local i32 @test_loop(i32 %0) nounwind {
; CHECK-NEXT: .LBB2_2: # =>This Inner Loop Header: Depth=1
; CHECK-NEXT: tileloadd (%r15,%r12), %tmm0
; CHECK-NEXT: movabsq $64, %rax
; CHECK-NEXT: tilestored %tmm0, 1024(%rsp,%rax) # 1024-byte Folded Spill
; CHECK-NEXT: tilestored %tmm0, 64(%rsp,%rax) # 1024-byte Folded Spill
; CHECK-NEXT: callq foo
; CHECK-NEXT: ldtilecfg {{[0-9]+}}(%rsp)
; CHECK-NEXT: ldtilecfg (%rsp)
; CHECK-NEXT: movabsq $64, %rax
; CHECK-NEXT: tileloadd 1024(%rsp,%rax), %tmm0 # 1024-byte Folded Reload
; CHECK-NEXT: tileloadd 64(%rsp,%rax), %tmm0 # 1024-byte Folded Reload
; CHECK-NEXT: tilestored %tmm0, (%r13,%r12)
; CHECK-NEXT: callq foo
; CHECK-NEXT: ldtilecfg {{[0-9]+}}(%rsp)
; CHECK-NEXT: ldtilecfg (%rsp)
; CHECK-NEXT: decl %ebp
; CHECK-NEXT: cmpl $7, %ebp
; CHECK-NEXT: jne .LBB2_2
@ -155,7 +155,7 @@ define dso_local i32 @test_loop(i32 %0) nounwind {
; CHECK-NEXT: jmp .LBB2_8
; CHECK-NEXT: .LBB2_4:
; CHECK-NEXT: callq foo
; CHECK-NEXT: ldtilecfg {{[0-9]+}}(%rsp)
; CHECK-NEXT: ldtilecfg (%rsp)
; CHECK-NEXT: movl $32, %eax
; CHECK-NEXT: movl $buf+1024, %ecx
; CHECK-NEXT: movw $8, %dx
@ -165,7 +165,7 @@ define dso_local i32 @test_loop(i32 %0) nounwind {
; CHECK-NEXT: decl %r14d
; CHECK-NEXT: .LBB2_8:
; CHECK-NEXT: movl %r14d, %eax
; CHECK-NEXT: addq $3016, %rsp # imm = 0xBC8
; CHECK-NEXT: addq $1096, %rsp # imm = 0x448
; CHECK-NEXT: popq %rbx
; CHECK-NEXT: popq %r12
; CHECK-NEXT: popq %r13
@ -270,11 +270,11 @@ define dso_local void @test_loop2(i32 %0) nounwind {
; CHECK-NEXT: pushq %r14
; CHECK-NEXT: pushq %r12
; CHECK-NEXT: pushq %rbx
; CHECK-NEXT: subq $3024, %rsp # imm = 0xBD0
; CHECK-NEXT: subq $1088, %rsp # imm = 0x440
; CHECK-NEXT: movl %edi, %ebx
; CHECK-NEXT: vpxord %zmm0, %zmm0, %zmm0
; CHECK-NEXT: vmovdqu64 %zmm0, {{[0-9]+}}(%rsp)
; CHECK-NEXT: movb $1, {{[0-9]+}}(%rsp)
; CHECK-NEXT: vmovdqu64 %zmm0, (%rsp)
; CHECK-NEXT: movb $1, (%rsp)
; CHECK-NEXT: movb $8, {{[0-9]+}}(%rsp)
; CHECK-NEXT: movw $8, {{[0-9]+}}(%rsp)
; CHECK-NEXT: movl $buf, %r14d
@ -285,22 +285,22 @@ define dso_local void @test_loop2(i32 %0) nounwind {
; CHECK-NEXT: .LBB3_1: # =>This Inner Loop Header: Depth=1
; CHECK-NEXT: vzeroupper
; CHECK-NEXT: callq foo
; CHECK-NEXT: ldtilecfg {{[0-9]+}}(%rsp)
; CHECK-NEXT: ldtilecfg (%rsp)
; CHECK-NEXT: testl %ebx, %ebx
; CHECK-NEXT: jle .LBB3_3
; CHECK-NEXT: # %bb.2: # in Loop: Header=BB3_1 Depth=1
; CHECK-NEXT: tileloadd (%r14,%r15), %tmm0
; CHECK-NEXT: movabsq $64, %rax
; CHECK-NEXT: tilestored %tmm0, 1024(%rsp,%rax) # 1024-byte Folded Spill
; CHECK-NEXT: tilestored %tmm0, 64(%rsp,%rax) # 1024-byte Folded Spill
; CHECK-NEXT: callq foo
; CHECK-NEXT: ldtilecfg {{[0-9]+}}(%rsp)
; CHECK-NEXT: ldtilecfg (%rsp)
; CHECK-NEXT: movabsq $64, %rax
; CHECK-NEXT: tileloadd 1024(%rsp,%rax), %tmm0 # 1024-byte Folded Reload
; CHECK-NEXT: tileloadd 64(%rsp,%rax), %tmm0 # 1024-byte Folded Reload
; CHECK-NEXT: tilestored %tmm0, (%r12,%r15)
; CHECK-NEXT: callq foo
; CHECK-NEXT: jmp .LBB3_1
; CHECK-NEXT: .LBB3_3:
; CHECK-NEXT: addq $3024, %rsp # imm = 0xBD0
; CHECK-NEXT: addq $1088, %rsp # imm = 0x440
; CHECK-NEXT: popq %rbx
; CHECK-NEXT: popq %r12
; CHECK-NEXT: popq %r14

4
test/CodeGen/X86/AMX/amx-lower-tile-copy.ll
View File

@ -35,12 +35,12 @@ define dso_local void @test1(i8 *%buf) nounwind {
; CHECK-NEXT: .LBB0_2: # %loop.header
; CHECK-NEXT: # =>This Inner Loop Header: Depth=1
; CHECK-NEXT: movabsq $64, %rax
; CHECK-NEXT: tilestored %tmm3, 2048(%rsp,%rax) # 1024-byte Folded Spill
; CHECK-NEXT: tilestored %tmm3, 3024(%rsp,%rax) # 1024-byte Folded Spill
; CHECK-NEXT: vzeroupper
; CHECK-NEXT: callq foo
; CHECK-NEXT: ldtilecfg {{[0-9]+}}(%rsp)
; CHECK-NEXT: movabsq $64, %rax
; CHECK-NEXT: tileloadd 2048(%rsp,%rax), %tmm3 # 1024-byte Folded Reload
; CHECK-NEXT: tileloadd 3024(%rsp,%rax), %tmm3 # 1024-byte Folded Reload
; CHECK-NEXT: tileloadd (%rbx,%r15), %tmm0
; CHECK-NEXT: tileloadd (%rbx,%r15), %tmm1
; CHECK-NEXT: # implicit-def: $rax

28
test/CodeGen/X86/AMX/amx-spill-merge.ll
View File

@ -10,12 +10,12 @@ define dso_local void @test_api(i16 signext %0, i16 signext %1) nounwind {
; CHECK-NEXT: pushq %r15
; CHECK-NEXT: pushq %r14
; CHECK-NEXT: pushq %rbx
; CHECK-NEXT: subq $4056, %rsp # imm = 0xFD8
; CHECK-NEXT: subq $2120, %rsp # imm = 0x848
; CHECK-NEXT: movl %esi, %ebx
; CHECK-NEXT: movl %edi, %ebp
; CHECK-NEXT: vpxord %zmm0, %zmm0, %zmm0
; CHECK-NEXT: vmovdqu64 %zmm0, {{[0-9]+}}(%rsp)
; CHECK-NEXT: movb $1, {{[0-9]+}}(%rsp)
; CHECK-NEXT: vmovdqu64 %zmm0, (%rsp)
; CHECK-NEXT: movb $1, (%rsp)
; CHECK-NEXT: movw $8, {{[0-9]+}}(%rsp)
; CHECK-NEXT: movb $8, {{[0-9]+}}(%rsp)
; CHECK-NEXT: movw $8, {{[0-9]+}}(%rsp)
@ -30,7 +30,7 @@ define dso_local void @test_api(i16 signext %0, i16 signext %1) nounwind {
; CHECK-NEXT: movb %bpl, {{[0-9]+}}(%rsp)
; CHECK-NEXT: movw %bx, {{[0-9]+}}(%rsp)
; CHECK-NEXT: movb %bpl, {{[0-9]+}}(%rsp)
; CHECK-NEXT: ldtilecfg {{[0-9]+}}(%rsp)
; CHECK-NEXT: ldtilecfg (%rsp)
; CHECK-NEXT: movl $32, %r14d
; CHECK-NEXT: movl $buf+2048, %r15d
; CHECK-NEXT: tileloadd (%r15,%r14), %tmm5
@ -44,16 +44,16 @@ define dso_local void @test_api(i16 signext %0, i16 signext %1) nounwind {
; CHECK-NEXT: movl $buf+1024, %eax
; CHECK-NEXT: tileloadd (%rax,%r14), %tmm1
; CHECK-NEXT: movabsq $64, %rax
; CHECK-NEXT: tilestored %tmm5, 2048(%rsp,%rax) # 1024-byte Folded Spill
; CHECK-NEXT: tilestored %tmm5, 1088(%rsp,%rax) # 1024-byte Folded Spill
; CHECK-NEXT: tdpbssd %tmm1, %tmm0, %tmm5
; CHECK-NEXT: movabsq $64, %rax
; CHECK-NEXT: tilestored %tmm5, 1024(%rsp,%rax) # 1024-byte Folded Spill
; CHECK-NEXT: tilestored %tmm5, 64(%rsp,%rax) # 1024-byte Folded Spill
; CHECK-NEXT: xorl %eax, %eax
; CHECK-NEXT: vzeroupper
; CHECK-NEXT: callq foo
; CHECK-NEXT: ldtilecfg {{[0-9]+}}(%rsp)
; CHECK-NEXT: ldtilecfg (%rsp)
; CHECK-NEXT: movabsq $64, %rax
; CHECK-NEXT: tileloadd 1024(%rsp,%rax), %tmm6 # 1024-byte Folded Reload
; CHECK-NEXT: tileloadd 64(%rsp,%rax), %tmm6 # 1024-byte Folded Reload
; CHECK-NEXT: jmp .LBB0_3
; CHECK-NEXT: .LBB0_2: # %if.false
; CHECK-NEXT: movl $buf, %eax
@ -62,16 +62,16 @@ define dso_local void @test_api(i16 signext %0, i16 signext %1) nounwind {
; CHECK-NEXT: movl $buf+1024, %eax
; CHECK-NEXT: tileloadd (%rax,%r14), %tmm3
; CHECK-NEXT: movabsq $64, %rax
; CHECK-NEXT: tilestored %tmm5, 2048(%rsp,%rax) # 1024-byte Folded Spill
; CHECK-NEXT: tilestored %tmm5, 1088(%rsp,%rax) # 1024-byte Folded Spill
; CHECK-NEXT: tdpbssd %tmm3, %tmm2, %tmm5
; CHECK-NEXT: movabsq $64, %rax
; CHECK-NEXT: tilestored %tmm5, 1024(%rsp,%rax) # 1024-byte Folded Spill
; CHECK-NEXT: tilestored %tmm5, 64(%rsp,%rax) # 1024-byte Folded Spill
; CHECK-NEXT: xorl %eax, %eax
; CHECK-NEXT: vzeroupper
; CHECK-NEXT: callq foo
; CHECK-NEXT: ldtilecfg {{[0-9]+}}(%rsp)
; CHECK-NEXT: ldtilecfg (%rsp)
; CHECK-NEXT: movabsq $64, %rax
; CHECK-NEXT: tileloadd 1024(%rsp,%rax), %tmm6 # 1024-byte Folded Reload
; CHECK-NEXT: tileloadd 64(%rsp,%rax), %tmm6 # 1024-byte Folded Reload
; CHECK-NEXT: tilestored %tmm6, (%r15,%r14)
; CHECK-NEXT: .LBB0_3: # %exit
; CHECK-NEXT: movl $buf, %eax
@ -79,11 +79,11 @@ define dso_local void @test_api(i16 signext %0, i16 signext %1) nounwind {
; CHECK-NEXT: movw $8, %dx
; CHECK-NEXT: tileloadd (%rax,%rcx), %tmm4
; CHECK-NEXT: movabsq $64, %rax
; CHECK-NEXT: tileloadd 2048(%rsp,%rax), %tmm5 # 1024-byte Folded Reload
; CHECK-NEXT: tileloadd 1088(%rsp,%rax), %tmm5 # 1024-byte Folded Reload
; CHECK-NEXT: tdpbssd %tmm4, %tmm6, %tmm5
; CHECK-NEXT: movl $buf+2048, %eax
; CHECK-NEXT: tilestored %tmm5, (%rax,%rcx)
; CHECK-NEXT: addq $4056, %rsp # imm = 0xFD8
; CHECK-NEXT: addq $2120, %rsp # imm = 0x848
; CHECK-NEXT: popq %rbx
; CHECK-NEXT: popq %r14
; CHECK-NEXT: popq %r15

46
test/CodeGen/X86/AMX/amx-spill.ll
View File

@ -7,33 +7,33 @@
define dso_local void @test_api(i32 %0, i16 signext %1, i16 signext %2) nounwind {
; CHECK-LABEL: test_api:
; CHECK: # %bb.0:
; CHECK-NEXT: subq $2936, %rsp # imm = 0xB78
; CHECK-NEXT: subq $968, %rsp # imm = 0x3C8
; CHECK-NEXT: vpxord %zmm0, %zmm0, %zmm0
; CHECK-NEXT: vmovdqu64 %zmm0, {{[0-9]+}}(%rsp)
; CHECK-NEXT: movb $1, {{[0-9]+}}(%rsp)
; CHECK-NEXT: movw %dx, {{[0-9]+}}(%rsp)
; CHECK-NEXT: movb %sil, {{[0-9]+}}(%rsp)
; CHECK-NEXT: movw %si, {{[0-9]+}}(%rsp)
; CHECK-NEXT: movb %sil, {{[0-9]+}}(%rsp)
; CHECK-NEXT: movw %dx, {{[0-9]+}}(%rsp)
; CHECK-NEXT: movb %dl, {{[0-9]+}}(%rsp)
; CHECK-NEXT: movw %dx, {{[0-9]+}}(%rsp)
; CHECK-NEXT: movb %dl, {{[0-9]+}}(%rsp)
; CHECK-NEXT: movw %dx, {{[0-9]+}}(%rsp)
; CHECK-NEXT: movb %sil, {{[0-9]+}}(%rsp)
; CHECK-NEXT: movw %dx, {{[0-9]+}}(%rsp)
; CHECK-NEXT: movb %dl, {{[0-9]+}}(%rsp)
; CHECK-NEXT: movw %dx, {{[0-9]+}}(%rsp)
; CHECK-NEXT: movb %sil, {{[0-9]+}}(%rsp)
; CHECK-NEXT: movw %dx, {{[0-9]+}}(%rsp)
; CHECK-NEXT: movb %dl, {{[0-9]+}}(%rsp)
; CHECK-NEXT: ldtilecfg {{[0-9]+}}(%rsp)
; CHECK-NEXT: vmovdqu64 %zmm0, -{{[0-9]+}}(%rsp)
; CHECK-NEXT: movb $1, -{{[0-9]+}}(%rsp)
; CHECK-NEXT: movw %dx, -{{[0-9]+}}(%rsp)
; CHECK-NEXT: movb %sil, -{{[0-9]+}}(%rsp)
; CHECK-NEXT: movw %si, -{{[0-9]+}}(%rsp)
; CHECK-NEXT: movb %sil, -{{[0-9]+}}(%rsp)
; CHECK-NEXT: movw %dx, -{{[0-9]+}}(%rsp)
; CHECK-NEXT: movb %dl, -{{[0-9]+}}(%rsp)
; CHECK-NEXT: movw %dx, -{{[0-9]+}}(%rsp)
; CHECK-NEXT: movb %dl, -{{[0-9]+}}(%rsp)
; CHECK-NEXT: movw %dx, -{{[0-9]+}}(%rsp)
; CHECK-NEXT: movb %sil, -{{[0-9]+}}(%rsp)
; CHECK-NEXT: movw %dx, -{{[0-9]+}}(%rsp)
; CHECK-NEXT: movb %dl, -{{[0-9]+}}(%rsp)
; CHECK-NEXT: movw %dx, -{{[0-9]+}}(%rsp)
; CHECK-NEXT: movb %sil, -{{[0-9]+}}(%rsp)
; CHECK-NEXT: movw %dx, -{{[0-9]+}}(%rsp)
; CHECK-NEXT: movb %dl, -{{[0-9]+}}(%rsp)
; CHECK-NEXT: ldtilecfg -{{[0-9]+}}(%rsp)
; CHECK-NEXT: movl $buf, %r8d
; CHECK-NEXT: movl $32, %eax
; CHECK-NEXT: tileloadd (%r8,%rax), %tmm1
; CHECK-NEXT: tileloadd (%r8,%rax), %tmm1
; CHECK-NEXT: movabsq $64, %rcx
; CHECK-NEXT: tilestored %tmm1, 896(%rsp,%rcx) # 1024-byte Folded Spill
; CHECK-NEXT: tilestored %tmm1, -64(%rsp,%rcx) # 1024-byte Folded Spill
; CHECK-NEXT: tileloadd (%r8,%rax), %tmm3
; CHECK-NEXT: tileloadd (%r8,%rax), %tmm4
; CHECK-NEXT: tileloadd (%r8,%rax), %tmm2
@ -54,14 +54,14 @@ define dso_local void @test_api(i32 %0, i16 signext %1, i16 signext %2) nounwind
; CHECK-NEXT: .LBB0_3:
; CHECK-NEXT: tdpbssd %tmm7, %tmm6, %tmm1
; CHECK-NEXT: movabsq $64, %rax
; CHECK-NEXT: tileloadd 896(%rsp,%rax), %tmm7 # 1024-byte Folded Reload
; CHECK-NEXT: tileloadd -64(%rsp,%rax), %tmm7 # 1024-byte Folded Reload
; CHECK-NEXT: tdpbssd %tmm7, %tmm1, %tmm3
; CHECK-NEXT: tdpbssd %tmm4, %tmm3, %tmm2
; CHECK-NEXT: tdpbssd %tmm5, %tmm2, %tmm0
; CHECK-NEXT: movl $buf, %eax
; CHECK-NEXT: movl $32, %ecx
; CHECK-NEXT: tilestored %tmm0, (%rax,%rcx)
; CHECK-NEXT: addq $2936, %rsp # imm = 0xB78
; CHECK-NEXT: addq $968, %rsp # imm = 0x3C8
; CHECK-NEXT: tilerelease
; CHECK-NEXT: vzeroupper
; CHECK-NEXT: retq

122
test/CodeGen/X86/addsub-constant-folding.ll
View File

@ -64,12 +64,12 @@ define i32 @add_const_add_const_extrause(i32 %arg) {
define <4 x i32> @vec_add_const_add_const(<4 x i32> %arg) {
; X86-LABEL: vec_add_const_add_const:
; X86: # %bb.0:
; X86-NEXT: paddd {{\.LCPI[0-9]+_[0-9]+}}, %xmm0
; X86-NEXT: paddd {{\.?LCPI[0-9]+_[0-9]+}}, %xmm0
; X86-NEXT: retl
;
; X64-LABEL: vec_add_const_add_const:
; X64: # %bb.0:
; X64-NEXT: paddd {{.*}}(%rip), %xmm0
; X64-NEXT: paddd {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0
; X64-NEXT: retq
%t0 = add <4 x i32> %arg, <i32 8, i32 8, i32 8, i32 8>
%t1 = add <4 x i32> %t0, <i32 2, i32 2, i32 2, i32 2>
@ -79,16 +79,16 @@ define <4 x i32> @vec_add_const_add_const(<4 x i32> %arg) {
define <4 x i32> @vec_add_const_add_const_extrause(<4 x i32> %arg) {
; X86-LABEL: vec_add_const_add_const_extrause:
; X86: # %bb.0:
; X86-NEXT: subl $28, %esp
; X86-NEXT: .cfi_def_cfa_offset 32
; X86-NEXT: subl $16, %esp
; X86-NEXT: .cfi_def_cfa_offset 20
; X86-NEXT: movdqa %xmm0, %xmm1
; X86-NEXT: movdqu %xmm0, (%esp) # 16-byte Spill
; X86-NEXT: movdqa {{.*#+}} xmm0 = [8,8,8,8]
; X86-NEXT: paddd %xmm1, %xmm0
; X86-NEXT: calll vec_use@PLT
; X86-NEXT: movdqu (%esp), %xmm0 # 16-byte Reload
; X86-NEXT: paddd {{\.LCPI[0-9]+_[0-9]+}}, %xmm0
; X86-NEXT: addl $28, %esp
; X86-NEXT: paddd {{\.?LCPI[0-9]+_[0-9]+}}, %xmm0
; X86-NEXT: addl $16, %esp
; X86-NEXT: .cfi_def_cfa_offset 4
; X86-NEXT: retl
;
@ -102,7 +102,7 @@ define <4 x i32> @vec_add_const_add_const_extrause(<4 x i32> %arg) {
; X64-NEXT: paddd %xmm1, %xmm0
; X64-NEXT: callq vec_use@PLT
; X64-NEXT: movdqa (%rsp), %xmm0 # 16-byte Reload
; X64-NEXT: paddd {{.*}}(%rip), %xmm0
; X64-NEXT: paddd {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0
; X64-NEXT: addq $24, %rsp
; X64-NEXT: .cfi_def_cfa_offset 8
; X64-NEXT: retq
@ -115,12 +115,12 @@ define <4 x i32> @vec_add_const_add_const_extrause(<4 x i32> %arg) {
define <4 x i32> @vec_add_const_add_const_nonsplat(<4 x i32> %arg) {
; X86-LABEL: vec_add_const_add_const_nonsplat:
; X86: # %bb.0:
; X86-NEXT: paddd {{\.LCPI[0-9]+_[0-9]+}}, %xmm0
; X86-NEXT: paddd {{\.?LCPI[0-9]+_[0-9]+}}, %xmm0
; X86-NEXT: retl
;
; X64-LABEL: vec_add_const_add_const_nonsplat:
; X64: # %bb.0:
; X64-NEXT: paddd {{.*}}(%rip), %xmm0
; X64-NEXT: paddd {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0
; X64-NEXT: retq
%t0 = add <4 x i32> %arg, <i32 21, i32 undef, i32 8, i32 8>
%t1 = add <4 x i32> %t0, <i32 2, i32 3, i32 undef, i32 2>
@ -186,12 +186,12 @@ define i32 @add_const_sub_const_extrause(i32 %arg) {
define <4 x i32> @vec_add_const_sub_const(<4 x i32> %arg) {
; X86-LABEL: vec_add_const_sub_const:
; X86: # %bb.0:
; X86-NEXT: paddd {{\.LCPI[0-9]+_[0-9]+}}, %xmm0
; X86-NEXT: paddd {{\.?LCPI[0-9]+_[0-9]+}}, %xmm0
; X86-NEXT: retl
;
; X64-LABEL: vec_add_const_sub_const:
; X64: # %bb.0:
; X64-NEXT: paddd {{.*}}(%rip), %xmm0
; X64-NEXT: paddd {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0
; X64-NEXT: retq
%t0 = add <4 x i32> %arg, <i32 8, i32 8, i32 8, i32 8>
%t1 = sub <4 x i32> %t0, <i32 2, i32 2, i32 2, i32 2>
@ -201,16 +201,16 @@ define <4 x i32> @vec_add_const_sub_const(<4 x i32> %arg) {
define <4 x i32> @vec_add_const_sub_const_extrause(<4 x i32> %arg) {
; X86-LABEL: vec_add_const_sub_const_extrause:
; X86: # %bb.0:
; X86-NEXT: subl $28, %esp
; X86-NEXT: .cfi_def_cfa_offset 32
; X86-NEXT: subl $16, %esp
; X86-NEXT: .cfi_def_cfa_offset 20
; X86-NEXT: movdqa %xmm0, %xmm1
; X86-NEXT: movdqu %xmm0, (%esp) # 16-byte Spill
; X86-NEXT: movdqa {{.*#+}} xmm0 = [8,8,8,8]
; X86-NEXT: paddd %xmm1, %xmm0
; X86-NEXT: calll vec_use@PLT
; X86-NEXT: movdqu (%esp), %xmm0 # 16-byte Reload
; X86-NEXT: paddd {{\.LCPI[0-9]+_[0-9]+}}, %xmm0
; X86-NEXT: addl $28, %esp
; X86-NEXT: paddd {{\.?LCPI[0-9]+_[0-9]+}}, %xmm0
; X86-NEXT: addl $16, %esp
; X86-NEXT: .cfi_def_cfa_offset 4
; X86-NEXT: retl
;
@ -224,7 +224,7 @@ define <4 x i32> @vec_add_const_sub_const_extrause(<4 x i32> %arg) {
; X64-NEXT: paddd %xmm1, %xmm0
; X64-NEXT: callq vec_use@PLT
; X64-NEXT: movdqa (%rsp), %xmm0 # 16-byte Reload
; X64-NEXT: paddd {{.*}}(%rip), %xmm0
; X64-NEXT: paddd {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0
; X64-NEXT: addq $24, %rsp
; X64-NEXT: .cfi_def_cfa_offset 8
; X64-NEXT: retq
@ -237,12 +237,12 @@ define <4 x i32> @vec_add_const_sub_const_extrause(<4 x i32> %arg) {
define <4 x i32> @vec_add_const_sub_const_nonsplat(<4 x i32> %arg) {
; X86-LABEL: vec_add_const_sub_const_nonsplat:
; X86: # %bb.0:
; X86-NEXT: paddd {{\.LCPI[0-9]+_[0-9]+}}, %xmm0
; X86-NEXT: paddd {{\.?LCPI[0-9]+_[0-9]+}}, %xmm0
; X86-NEXT: retl
;
; X64-LABEL: vec_add_const_sub_const_nonsplat:
; X64: # %bb.0:
; X64-NEXT: paddd {{.*}}(%rip), %xmm0
; X64-NEXT: paddd {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0
; X64-NEXT: retq
%t0 = add <4 x i32> %arg, <i32 21, i32 undef, i32 8, i32 8>
%t1 = sub <4 x i32> %t0, <i32 2, i32 3, i32 undef, i32 2>
@ -328,8 +328,8 @@ define <4 x i32> @vec_add_const_const_sub(<4 x i32> %arg) {
define <4 x i32> @vec_add_const_const_sub_extrause(<4 x i32> %arg) {
; X86-LABEL: vec_add_const_const_sub_extrause:
; X86: # %bb.0:
; X86-NEXT: subl $28, %esp
; X86-NEXT: .cfi_def_cfa_offset 32
; X86-NEXT: subl $16, %esp
; X86-NEXT: .cfi_def_cfa_offset 20
; X86-NEXT: movdqa %xmm0, %xmm1
; X86-NEXT: movdqu %xmm0, (%esp) # 16-byte Spill
; X86-NEXT: movdqa {{.*#+}} xmm0 = [8,8,8,8]
@ -338,7 +338,7 @@ define <4 x i32> @vec_add_const_const_sub_extrause(<4 x i32> %arg) {
; X86-NEXT: movdqa {{.*#+}} xmm0 = [4294967290,4294967290,4294967290,4294967290]
; X86-NEXT: movdqu (%esp), %xmm1 # 16-byte Reload
; X86-NEXT: psubd %xmm1, %xmm0
; X86-NEXT: addl $28, %esp
; X86-NEXT: addl $16, %esp
; X86-NEXT: .cfi_def_cfa_offset 4
; X86-NEXT: retl
;
@ -440,12 +440,12 @@ define i32 @sub_const_add_const_extrause(i32 %arg) {
define <4 x i32> @vec_sub_const_add_const(<4 x i32> %arg) {
; X86-LABEL: vec_sub_const_add_const:
; X86: # %bb.0:
; X86-NEXT: paddd {{\.LCPI[0-9]+_[0-9]+}}, %xmm0
; X86-NEXT: paddd {{\.?LCPI[0-9]+_[0-9]+}}, %xmm0
; X86-NEXT: retl
;
; X64-LABEL: vec_sub_const_add_const:
; X64: # %bb.0:
; X64-NEXT: paddd {{.*}}(%rip), %xmm0
; X64-NEXT: paddd {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0
; X64-NEXT: retq
%t0 = sub <4 x i32> %arg, <i32 8, i32 8, i32 8, i32 8>
%t1 = add <4 x i32> %t0, <i32 2, i32 2, i32 2, i32 2>
@ -455,14 +455,14 @@ define <4 x i32> @vec_sub_const_add_const(<4 x i32> %arg) {
define <4 x i32> @vec_sub_const_add_const_extrause(<4 x i32> %arg) {
; X86-LABEL: vec_sub_const_add_const_extrause:
; X86: # %bb.0:
; X86-NEXT: subl $28, %esp
; X86-NEXT: .cfi_def_cfa_offset 32
; X86-NEXT: subl $16, %esp
; X86-NEXT: .cfi_def_cfa_offset 20
; X86-NEXT: movdqu %xmm0, (%esp) # 16-byte Spill
; X86-NEXT: psubd {{\.LCPI[0-9]+_[0-9]+}}, %xmm0
; X86-NEXT: psubd {{\.?LCPI[0-9]+_[0-9]+}}, %xmm0
; X86-NEXT: calll vec_use@PLT
; X86-NEXT: movdqu (%esp), %xmm0 # 16-byte Reload
; X86-NEXT: paddd {{\.LCPI[0-9]+_[0-9]+}}, %xmm0
; X86-NEXT: addl $28, %esp
; X86-NEXT: paddd {{\.?LCPI[0-9]+_[0-9]+}}, %xmm0
; X86-NEXT: addl $16, %esp
; X86-NEXT: .cfi_def_cfa_offset 4
; X86-NEXT: retl
;
@ -471,10 +471,10 @@ define <4 x i32> @vec_sub_const_add_const_extrause(<4 x i32> %arg) {
; X64-NEXT: subq $24, %rsp
; X64-NEXT: .cfi_def_cfa_offset 32
; X64-NEXT: movdqa %xmm0, (%rsp) # 16-byte Spill
; X64-NEXT: psubd {{.*}}(%rip), %xmm0
; X64-NEXT: psubd {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0
; X64-NEXT: callq vec_use@PLT
; X64-NEXT: movdqa (%rsp), %xmm0 # 16-byte Reload
; X64-NEXT: paddd {{.*}}(%rip), %xmm0
; X64-NEXT: paddd {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0
; X64-NEXT: addq $24, %rsp
; X64-NEXT: .cfi_def_cfa_offset 8
; X64-NEXT: retq
@ -487,12 +487,12 @@ define <4 x i32> @vec_sub_const_add_const_extrause(<4 x i32> %arg) {
define <4 x i32> @vec_sub_const_add_const_nonsplat(<4 x i32> %arg) {
; X86-LABEL: vec_sub_const_add_const_nonsplat:
; X86: # %bb.0:
; X86-NEXT: paddd {{\.LCPI[0-9]+_[0-9]+}}, %xmm0
; X86-NEXT: paddd {{\.?LCPI[0-9]+_[0-9]+}}, %xmm0
; X86-NEXT: retl
;
; X64-LABEL: vec_sub_const_add_const_nonsplat:
; X64: # %bb.0:
; X64-NEXT: paddd {{.*}}(%rip), %xmm0
; X64-NEXT: paddd {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0
; X64-NEXT: retq
%t0 = sub <4 x i32> %arg, <i32 21, i32 undef, i32 8, i32 8>
%t1 = add <4 x i32> %t0, <i32 2, i32 3, i32 undef, i32 2>
@ -558,12 +558,12 @@ define i32 @sub_const_sub_const_extrause(i32 %arg) {
define <4 x i32> @vec_sub_const_sub_const(<4 x i32> %arg) {
; X86-LABEL: vec_sub_const_sub_const:
; X86: # %bb.0:
; X86-NEXT: psubd {{\.LCPI[0-9]+_[0-9]+}}, %xmm0
; X86-NEXT: psubd {{\.?LCPI[0-9]+_[0-9]+}}, %xmm0
; X86-NEXT: retl
;
; X64-LABEL: vec_sub_const_sub_const:
; X64: # %bb.0:
; X64-NEXT: psubd {{.*}}(%rip), %xmm0
; X64-NEXT: psubd {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0
; X64-NEXT: retq
%t0 = sub <4 x i32> %arg, <i32 8, i32 8, i32 8, i32 8>
%t1 = sub <4 x i32> %t0, <i32 2, i32 2, i32 2, i32 2>
@ -573,14 +573,14 @@ define <4 x i32> @vec_sub_const_sub_const(<4 x i32> %arg) {
define <4 x i32> @vec_sub_const_sub_const_extrause(<4 x i32> %arg) {
; X86-LABEL: vec_sub_const_sub_const_extrause:
; X86: # %bb.0:
; X86-NEXT: subl $28, %esp
; X86-NEXT: .cfi_def_cfa_offset 32
; X86-NEXT: subl $16, %esp
; X86-NEXT: .cfi_def_cfa_offset 20
; X86-NEXT: movdqu %xmm0, (%esp) # 16-byte Spill
; X86-NEXT: psubd {{\.LCPI[0-9]+_[0-9]+}}, %xmm0
; X86-NEXT: psubd {{\.?LCPI[0-9]+_[0-9]+}}, %xmm0
; X86-NEXT: calll vec_use@PLT
; X86-NEXT: movdqu (%esp), %xmm0 # 16-byte Reload
; X86-NEXT: psubd {{\.LCPI[0-9]+_[0-9]+}}, %xmm0
; X86-NEXT: addl $28, %esp
; X86-NEXT: psubd {{\.?LCPI[0-9]+_[0-9]+}}, %xmm0
; X86-NEXT: addl $16, %esp
; X86-NEXT: .cfi_def_cfa_offset 4
; X86-NEXT: retl
;
@ -589,10 +589,10 @@ define <4 x i32> @vec_sub_const_sub_const_extrause(<4 x i32> %arg) {
; X64-NEXT: subq $24, %rsp
; X64-NEXT: .cfi_def_cfa_offset 32
; X64-NEXT: movdqa %xmm0, (%rsp) # 16-byte Spill
; X64-NEXT: psubd {{.*}}(%rip), %xmm0
; X64-NEXT: psubd {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0
; X64-NEXT: callq vec_use@PLT
; X64-NEXT: movdqa (%rsp), %xmm0 # 16-byte Reload
; X64-NEXT: psubd {{.*}}(%rip), %xmm0
; X64-NEXT: psubd {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0
; X64-NEXT: addq $24, %rsp
; X64-NEXT: .cfi_def_cfa_offset 8
; X64-NEXT: retq
@ -605,12 +605,12 @@ define <4 x i32> @vec_sub_const_sub_const_extrause(<4 x i32> %arg) {
define <4 x i32> @vec_sub_const_sub_const_nonsplat(<4 x i32> %arg) {
; X86-LABEL: vec_sub_const_sub_const_nonsplat:
; X86: # %bb.0:
; X86-NEXT: psubd {{\.LCPI[0-9]+_[0-9]+}}, %xmm0
; X86-NEXT: psubd {{\.?LCPI[0-9]+_[0-9]+}}, %xmm0
; X86-NEXT: retl
;
; X64-LABEL: vec_sub_const_sub_const_nonsplat:
; X64: # %bb.0:
; X64-NEXT: psubd {{.*}}(%rip), %xmm0
; X64-NEXT: psubd {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0
; X64-NEXT: retq
%t0 = sub <4 x i32> %arg, <i32 21, i32 undef, i32 8, i32 8>
%t1 = sub <4 x i32> %t0, <i32 2, i32 3, i32 undef, i32 2>
@ -696,15 +696,15 @@ define <4 x i32> @vec_sub_const_const_sub(<4 x i32> %arg) {
define <4 x i32> @vec_sub_const_const_sub_extrause(<4 x i32> %arg) {
; X86-LABEL: vec_sub_const_const_sub_extrause:
; X86: # %bb.0:
; X86-NEXT: subl $28, %esp
; X86-NEXT: .cfi_def_cfa_offset 32
; X86-NEXT: psubd {{\.LCPI[0-9]+_[0-9]+}}, %xmm0
; X86-NEXT: subl $16, %esp
; X86-NEXT: .cfi_def_cfa_offset 20
; X86-NEXT: psubd {{\.?LCPI[0-9]+_[0-9]+}}, %xmm0
; X86-NEXT: movdqu %xmm0, (%esp) # 16-byte Spill
; X86-NEXT: calll vec_use@PLT
; X86-NEXT: movdqa {{.*#+}} xmm0 = [2,2,2,2]
; X86-NEXT: movdqu (%esp), %xmm1 # 16-byte Reload
; X86-NEXT: psubd %xmm1, %xmm0
; X86-NEXT: addl $28, %esp
; X86-NEXT: addl $16, %esp
; X86-NEXT: .cfi_def_cfa_offset 4
; X86-NEXT: retl
;
@ -712,7 +712,7 @@ define <4 x i32> @vec_sub_const_const_sub_extrause(<4 x i32> %arg) {
; X64: # %bb.0:
; X64-NEXT: subq $24, %rsp
; X64-NEXT: .cfi_def_cfa_offset 32
; X64-NEXT: psubd {{.*}}(%rip), %xmm0
; X64-NEXT: psubd {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0
; X64-NEXT: movdqa %xmm0, (%rsp) # 16-byte Spill
; X64-NEXT: callq vec_use@PLT
; X64-NEXT: movdqa {{.*#+}} xmm0 = [2,2,2,2]
@ -826,8 +826,8 @@ define <4 x i32> @vec_const_sub_add_const(<4 x i32> %arg) {
define <4 x i32> @vec_const_sub_add_const_extrause(<4 x i32> %arg) {
; X86-LABEL: vec_const_sub_add_const_extrause:
; X86: # %bb.0:
; X86-NEXT: subl $28, %esp
; X86-NEXT: .cfi_def_cfa_offset 32
; X86-NEXT: subl $16, %esp
; X86-NEXT: .cfi_def_cfa_offset 20
; X86-NEXT: movdqa %xmm0, %xmm1
; X86-NEXT: movdqu %xmm0, (%esp) # 16-byte Spill
; X86-NEXT: movdqa {{.*#+}} xmm0 = [8,8,8,8]
@ -836,7 +836,7 @@ define <4 x i32> @vec_const_sub_add_const_extrause(<4 x i32> %arg) {
; X86-NEXT: movdqa {{.*#+}} xmm0 = [10,10,10,10]
; X86-NEXT: movdqu (%esp), %xmm1 # 16-byte Reload
; X86-NEXT: psubd %xmm1, %xmm0
; X86-NEXT: addl $28, %esp
; X86-NEXT: addl $16, %esp
; X86-NEXT: .cfi_def_cfa_offset 4
; X86-NEXT: retl
;
@ -960,8 +960,8 @@ define <4 x i32> @vec_const_sub_sub_const(<4 x i32> %arg) {
define <4 x i32> @vec_const_sub_sub_const_extrause(<4 x i32> %arg) {
; X86-LABEL: vec_const_sub_sub_const_extrause:
; X86: # %bb.0:
; X86-NEXT: subl $28, %esp
; X86-NEXT: .cfi_def_cfa_offset 32
; X86-NEXT: subl $16, %esp
; X86-NEXT: .cfi_def_cfa_offset 20
; X86-NEXT: movdqa %xmm0, %xmm1
; X86-NEXT: movdqu %xmm0, (%esp) # 16-byte Spill
; X86-NEXT: movdqa {{.*#+}} xmm0 = [8,8,8,8]
@ -970,7 +970,7 @@ define <4 x i32> @vec_const_sub_sub_const_extrause(<4 x i32> %arg) {
; X86-NEXT: movdqa {{.*#+}} xmm0 = [6,6,6,6]
; X86-NEXT: movdqu (%esp), %xmm1 # 16-byte Reload
; X86-NEXT: psubd %xmm1, %xmm0
; X86-NEXT: addl $28, %esp
; X86-NEXT: addl $16, %esp
; X86-NEXT: .cfi_def_cfa_offset 4
; X86-NEXT: retl
;
@ -1074,12 +1074,12 @@ define i32 @const_sub_const_sub_extrause(i32 %arg) {
define <4 x i32> @vec_const_sub_const_sub(<4 x i32> %arg) {
; X86-LABEL: vec_const_sub_const_sub:
; X86: # %bb.0:
; X86-NEXT: paddd {{\.LCPI[0-9]+_[0-9]+}}, %xmm0
; X86-NEXT: paddd {{\.?LCPI[0-9]+_[0-9]+}}, %xmm0
; X86-NEXT: retl
;
; X64-LABEL: vec_const_sub_const_sub:
; X64: # %bb.0:
; X64-NEXT: paddd {{.*}}(%rip), %xmm0
; X64-NEXT: paddd {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0
; X64-NEXT: retq
%t0 = sub <4 x i32> <i32 8, i32 8, i32 8, i32 8>, %arg
%t1 = sub <4 x i32> <i32 2, i32 2, i32 2, i32 2>, %t0
@ -1089,8 +1089,8 @@ define <4 x i32> @vec_const_sub_const_sub(<4 x i32> %arg) {
define <4 x i32> @vec_const_sub_const_sub_extrause(<4 x i32> %arg) {
; X86-LABEL: vec_const_sub_const_sub_extrause:
; X86: # %bb.0:
; X86-NEXT: subl $28, %esp
; X86-NEXT: .cfi_def_cfa_offset 32
; X86-NEXT: subl $16, %esp
; X86-NEXT: .cfi_def_cfa_offset 20
; X86-NEXT: movdqa {{.*#+}} xmm1 = [8,8,8,8]
; X86-NEXT: psubd %xmm0, %xmm1
; X86-NEXT: movdqu %xmm1, (%esp) # 16-byte Spill
@ -1099,7 +1099,7 @@ define <4 x i32> @vec_const_sub_const_sub_extrause(<4 x i32> %arg) {
; X86-NEXT: movdqa {{.*#+}} xmm0 = [2,2,2,2]
; X86-NEXT: movdqu (%esp), %xmm1 # 16-byte Reload
; X86-NEXT: psubd %xmm1, %xmm0
; X86-NEXT: addl $28, %esp
; X86-NEXT: addl $16, %esp
; X86-NEXT: .cfi_def_cfa_offset 4
; X86-NEXT: retl
;
@ -1126,12 +1126,12 @@ define <4 x i32> @vec_const_sub_const_sub_extrause(<4 x i32> %arg) {
define <4 x i32> @vec_const_sub_const_sub_nonsplat(<4 x i32> %arg) {
; X86-LABEL: vec_const_sub_const_sub_nonsplat:
; X86: # %bb.0:
; X86-NEXT: paddd {{\.LCPI[0-9]+_[0-9]+}}, %xmm0
; X86-NEXT: paddd {{\.?LCPI[0-9]+_[0-9]+}}, %xmm0
; X86-NEXT: retl
;
; X64-LABEL: vec_const_sub_const_sub_nonsplat:
; X64: # %bb.0:
; X64-NEXT: paddd {{.*}}(%rip), %xmm0
; X64-NEXT: paddd {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0
; X64-NEXT: retq
%t0 = sub <4 x i32> <i32 21, i32 undef, i32 8, i32 8>, %arg
%t1 = sub <4 x i32> <i32 2, i32 3, i32 undef, i32 2>, %t0

8
test/CodeGen/X86/avx-intrinsics-x86_64.ll
View File

@ -7,10 +7,10 @@ define <4 x double> @test_x86_avx_vzeroall(<4 x double> %a, <4 x double> %b) {
; AVX: # %bb.0:
; AVX-NEXT: vaddpd %ymm1, %ymm0, %ymm0 # encoding: [0xc5,0xfd,0x58,0xc1]
; AVX-NEXT: vmovupd %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX-NEXT: # encoding: [0xc5,0xfd,0x11,0x44,0x24,0xc8]
; AVX-NEXT: # encoding: [0xc5,0xfd,0x11,0x44,0x24,0xd8]
; AVX-NEXT: vzeroall # encoding: [0xc5,0xfc,0x77]
; AVX-NEXT: vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX-NEXT: # encoding: [0xc5,0xfc,0x10,0x44,0x24,0xc8]
; AVX-NEXT: # encoding: [0xc5,0xfc,0x10,0x44,0x24,0xd8]
; AVX-NEXT: retq # encoding: [0xc3]
;
; AVX512VL-LABEL: test_x86_avx_vzeroall:
@ -30,10 +30,10 @@ define <4 x double> @test_x86_avx_vzeroupper(<4 x double> %a, <4 x double> %b) {
; AVX: # %bb.0:
; AVX-NEXT: vaddpd %ymm1, %ymm0, %ymm0 # encoding: [0xc5,0xfd,0x58,0xc1]
; AVX-NEXT: vmovupd %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX-NEXT: # encoding: [0xc5,0xfd,0x11,0x44,0x24,0xc8]
; AVX-NEXT: # encoding: [0xc5,0xfd,0x11,0x44,0x24,0xd8]
; AVX-NEXT: vzeroupper # encoding: [0xc5,0xf8,0x77]
; AVX-NEXT: vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX-NEXT: # encoding: [0xc5,0xfc,0x10,0x44,0x24,0xc8]
; AVX-NEXT: # encoding: [0xc5,0xfc,0x10,0x44,0x24,0xd8]
; AVX-NEXT: retq # encoding: [0xc3]
;
; AVX512VL-LABEL: test_x86_avx_vzeroupper:

48
test/CodeGen/X86/avx-vzeroupper.ll
View File

@ -32,56 +32,56 @@ define <4 x float> @test00(<4 x float> %a, <4 x float> %b) nounwind {
define <8 x float> @test01(<4 x float> %a, <4 x float> %b, <8 x float> %c) nounwind {
; VZ-LABEL: test01:
; VZ: # %bb.0:
; VZ-NEXT: subq $56, %rsp
; VZ-NEXT: subq $40, %rsp
; VZ-NEXT: vmovups %ymm2, (%rsp) # 32-byte Spill
; VZ-NEXT: vmovaps {{.*}}(%rip), %xmm0
; VZ-NEXT: vmovaps x(%rip), %xmm0
; VZ-NEXT: vzeroupper
; VZ-NEXT: callq do_sse
; VZ-NEXT: vmovaps %xmm0, {{.*}}(%rip)
; VZ-NEXT: vmovaps %xmm0, x(%rip)
; VZ-NEXT: callq do_sse
; VZ-NEXT: vmovaps %xmm0, {{.*}}(%rip)
; VZ-NEXT: vmovaps %xmm0, x(%rip)
; VZ-NEXT: vmovups (%rsp), %ymm0 # 32-byte Reload
; VZ-NEXT: addq $56, %rsp
; VZ-NEXT: addq $40, %rsp
; VZ-NEXT: retq
;
; DISABLE-VZ-LABEL: test01:
; DISABLE-VZ: # %bb.0:
; DISABLE-VZ-NEXT: subq $56, %rsp
; DISABLE-VZ-NEXT: subq $40, %rsp
; DISABLE-VZ-NEXT: vmovups %ymm2, (%rsp) # 32-byte Spill
; DISABLE-VZ-NEXT: vmovaps {{.*}}(%rip), %xmm0
; DISABLE-VZ-NEXT: vmovaps x(%rip), %xmm0
; DISABLE-VZ-NEXT: callq do_sse
; DISABLE-VZ-NEXT: vmovaps %xmm0, {{.*}}(%rip)
; DISABLE-VZ-NEXT: vmovaps %xmm0, x(%rip)
; DISABLE-VZ-NEXT: callq do_sse
; DISABLE-VZ-NEXT: vmovaps %xmm0, {{.*}}(%rip)
; DISABLE-VZ-NEXT: vmovaps %xmm0, x(%rip)
; DISABLE-VZ-NEXT: vmovups (%rsp), %ymm0 # 32-byte Reload
; DISABLE-VZ-NEXT: addq $56, %rsp
; DISABLE-VZ-NEXT: addq $40, %rsp
; DISABLE-VZ-NEXT: retq
;
; BDVER2-LABEL: test01:
; BDVER2: # %bb.0:
; BDVER2-NEXT: subq $56, %rsp
; BDVER2-NEXT: vmovaps {{.*}}(%rip), %xmm0
; BDVER2-NEXT: subq $40, %rsp
; BDVER2-NEXT: vmovaps x(%rip), %xmm0
; BDVER2-NEXT: vmovups %ymm2, (%rsp) # 32-byte Spill
; BDVER2-NEXT: vzeroupper
; BDVER2-NEXT: callq do_sse
; BDVER2-NEXT: vmovaps %xmm0, {{.*}}(%rip)
; BDVER2-NEXT: vmovaps %xmm0, x(%rip)
; BDVER2-NEXT: callq do_sse
; BDVER2-NEXT: vmovaps %xmm0, {{.*}}(%rip)
; BDVER2-NEXT: vmovaps %xmm0, x(%rip)
; BDVER2-NEXT: vmovups (%rsp), %ymm0 # 32-byte Reload
; BDVER2-NEXT: addq $56, %rsp
; BDVER2-NEXT: addq $40, %rsp
; BDVER2-NEXT: retq
;
; BTVER2-LABEL: test01:
; BTVER2: # %bb.0:
; BTVER2-NEXT: subq $56, %rsp
; BTVER2-NEXT: vmovaps {{.*}}(%rip), %xmm0
; BTVER2-NEXT: subq $40, %rsp
; BTVER2-NEXT: vmovaps x(%rip), %xmm0
; BTVER2-NEXT: vmovups %ymm2, (%rsp) # 32-byte Spill
; BTVER2-NEXT: callq do_sse
; BTVER2-NEXT: vmovaps %xmm0, {{.*}}(%rip)
; BTVER2-NEXT: vmovaps %xmm0, x(%rip)
; BTVER2-NEXT: callq do_sse
; BTVER2-NEXT: vmovaps %xmm0, {{.*}}(%rip)
; BTVER2-NEXT: vmovaps %xmm0, x(%rip)
; BTVER2-NEXT: vmovups (%rsp), %ymm0 # 32-byte Reload
; BTVER2-NEXT: addq $56, %rsp
; BTVER2-NEXT: addq $40, %rsp
; BTVER2-NEXT: retq
%tmp = load <4 x float>, <4 x float>* @x, align 16
%call = tail call <4 x float> @do_sse(<4 x float> %tmp) nounwind
@ -145,7 +145,7 @@ define <4 x float> @test03(<4 x float> %a, <4 x float> %b) nounwind {
; VZ-NEXT: # =>This Inner Loop Header: Depth=1
; VZ-NEXT: callq do_sse
; VZ-NEXT: callq do_sse
; VZ-NEXT: vmovaps g+{{.*}}(%rip), %xmm0
; VZ-NEXT: vmovaps g+16(%rip), %xmm0
; VZ-NEXT: callq do_sse
; VZ-NEXT: decl %ebx
; VZ-NEXT: jne .LBB3_3
@ -174,7 +174,7 @@ define <4 x float> @test03(<4 x float> %a, <4 x float> %b) nounwind {
; DISABLE-VZ-NEXT: # =>This Inner Loop Header: Depth=1
; DISABLE-VZ-NEXT: callq do_sse
; DISABLE-VZ-NEXT: callq do_sse
; DISABLE-VZ-NEXT: vmovaps g+{{.*}}(%rip), %xmm0
; DISABLE-VZ-NEXT: vmovaps g+16(%rip), %xmm0
; DISABLE-VZ-NEXT: callq do_sse
; DISABLE-VZ-NEXT: decl %ebx
; DISABLE-VZ-NEXT: jne .LBB3_3
@ -203,7 +203,7 @@ define <4 x float> @test03(<4 x float> %a, <4 x float> %b) nounwind {
; BDVER2-NEXT: # =>This Inner Loop Header: Depth=1
; BDVER2-NEXT: callq do_sse
; BDVER2-NEXT: callq do_sse
; BDVER2-NEXT: vmovaps g+{{.*}}(%rip), %xmm0
; BDVER2-NEXT: vmovaps g+16(%rip), %xmm0
; BDVER2-NEXT: callq do_sse
; BDVER2-NEXT: decl %ebx
; BDVER2-NEXT: jne .LBB3_3
@ -232,7 +232,7 @@ define <4 x float> @test03(<4 x float> %a, <4 x float> %b) nounwind {
; BTVER2-NEXT: # =>This Inner Loop Header: Depth=1
; BTVER2-NEXT: callq do_sse
; BTVER2-NEXT: callq do_sse
; BTVER2-NEXT: vmovaps g+{{.*}}(%rip), %xmm0
; BTVER2-NEXT: vmovaps g+16(%rip), %xmm0
; BTVER2-NEXT: callq do_sse
; BTVER2-NEXT: decl %ebx
; BTVER2-NEXT: jne .LBB3_3

4
test/CodeGen/X86/avx512-bugfix-25270.ll
View File

@ -7,7 +7,7 @@ define void @bar__512(<16 x i32>* %var) #0 {
; CHECK-LABEL: bar__512:
; CHECK: ## %bb.0: ## %allocas
; CHECK-NEXT: pushq %rbx
; CHECK-NEXT: subq $112, %rsp
; CHECK-NEXT: subq $64, %rsp
; CHECK-NEXT: movq %rdi, %rbx
; CHECK-NEXT: vmovups (%rdi), %zmm0
; CHECK-NEXT: vmovups %zmm0, (%rsp) ## 64-byte Spill
@ -18,7 +18,7 @@ define void @bar__512(<16 x i32>* %var) #0 {
; CHECK-NEXT: callq _Print__512
; CHECK-NEXT: vbroadcastss {{.*#+}} zmm0 = [7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7]
; CHECK-NEXT: vmovaps %zmm0, (%rbx)
; CHECK-NEXT: addq $112, %rsp
; CHECK-NEXT: addq $64, %rsp
; CHECK-NEXT: popq %rbx
; CHECK-NEXT: retq
allocas:

44
test/CodeGen/X86/avx512-regcall-NoMask.ll
View File

@ -323,12 +323,12 @@ define dso_local x86_regcallcc float @test_argRetFloat(float %a) {
;
; WIN64-LABEL: test_argRetFloat:
; WIN64: # %bb.0:
; WIN64-NEXT: vaddss __real@{{.*}}(%rip), %xmm0, %xmm0
; WIN64-NEXT: vaddss __real@3f800000(%rip), %xmm0, %xmm0
; WIN64-NEXT: retq
;
; LINUXOSX64-LABEL: test_argRetFloat:
; LINUXOSX64: # %bb.0:
; LINUXOSX64-NEXT: vaddss {{.*}}(%rip), %xmm0, %xmm0
; LINUXOSX64-NEXT: vaddss {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0, %xmm0
; LINUXOSX64-NEXT: retq
%add = fadd float 1.0, %a
ret float %add
@ -393,12 +393,12 @@ define dso_local x86_regcallcc double @test_argRetDouble(double %a) {
;
; WIN64-LABEL: test_argRetDouble:
; WIN64: # %bb.0:
; WIN64-NEXT: vaddsd __real@{{.*}}(%rip), %xmm0, %xmm0
; WIN64-NEXT: vaddsd __real@3ff0000000000000(%rip), %xmm0, %xmm0
; WIN64-NEXT: retq
;
; LINUXOSX64-LABEL: test_argRetDouble:
; LINUXOSX64: # %bb.0:
; LINUXOSX64-NEXT: vaddsd {{.*}}(%rip), %xmm0, %xmm0
; LINUXOSX64-NEXT: vaddsd {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0, %xmm0
; LINUXOSX64-NEXT: retq
%add = fadd double %a, 1.0
ret double %add
@ -758,7 +758,7 @@ define dso_local x86_regcallcc <8 x i32> @test_argRet256Vector(<8 x i1> %x, <8 x
define dso_local x86_regcallcc <8 x i32> @test_CallargRet256Vector(<8 x i1> %x, <8 x i32> %a) {
; X32-LABEL: test_CallargRet256Vector:
; X32: # %bb.0:
; X32-NEXT: subl $92, %esp
; X32-NEXT: subl $36, %esp
; X32-NEXT: vmovdqu %ymm0, {{[-0-9]+}}(%e{{[sb]}}p) # 32-byte Spill
; X32-NEXT: kmovd %eax, %k1
; X32-NEXT: kmovw %k1, {{[-0-9]+}}(%e{{[sb]}}p) # 2-byte Spill
@ -767,13 +767,13 @@ define dso_local x86_regcallcc <8 x i32> @test_CallargRet256Vector(<8 x i1> %x,
; X32-NEXT: kmovw {{[-0-9]+}}(%e{{[sb]}}p), %k1 # 2-byte Reload
; X32-NEXT: vmovdqu {{[-0-9]+}}(%e{{[sb]}}p), %ymm1 # 32-byte Reload
; X32-NEXT: vmovdqa32 %ymm1, %ymm0 {%k1}
; X32-NEXT: addl $92, %esp
; X32-NEXT: addl $36, %esp
; X32-NEXT: retl
;
; WIN64-LABEL: test_CallargRet256Vector:
; WIN64: # %bb.0:
; WIN64-NEXT: subq $88, %rsp
; WIN64-NEXT: .seh_stackalloc 88
; WIN64-NEXT: subq $56, %rsp
; WIN64-NEXT: .seh_stackalloc 56
; WIN64-NEXT: .seh_endprologue
; WIN64-NEXT: vmovdqu %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; WIN64-NEXT: kmovd %eax, %k1
@ -783,14 +783,14 @@ define dso_local x86_regcallcc <8 x i32> @test_CallargRet256Vector(<8 x i1> %x,
; WIN64-NEXT: kmovw {{[-0-9]+}}(%r{{[sb]}}p), %k1 # 2-byte Reload
; WIN64-NEXT: vmovdqu {{[-0-9]+}}(%r{{[sb]}}p), %ymm1 # 32-byte Reload
; WIN64-NEXT: vmovdqa32 %ymm1, %ymm0 {%k1}
; WIN64-NEXT: addq $88, %rsp
; WIN64-NEXT: addq $56, %rsp
; WIN64-NEXT: retq
; WIN64-NEXT: .seh_endproc
;
; LINUXOSX64-LABEL: test_CallargRet256Vector:
; LINUXOSX64: # %bb.0:
; LINUXOSX64-NEXT: subq $88, %rsp
; LINUXOSX64-NEXT: .cfi_def_cfa_offset 96
; LINUXOSX64-NEXT: subq $56, %rsp
; LINUXOSX64-NEXT: .cfi_def_cfa_offset 64
; LINUXOSX64-NEXT: vmovdqu %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; LINUXOSX64-NEXT: kmovd %eax, %k1
; LINUXOSX64-NEXT: kmovw %k1, {{[-0-9]+}}(%r{{[sb]}}p) # 2-byte Spill
@ -799,7 +799,7 @@ define dso_local x86_regcallcc <8 x i32> @test_CallargRet256Vector(<8 x i1> %x,
; LINUXOSX64-NEXT: kmovw {{[-0-9]+}}(%r{{[sb]}}p), %k1 # 2-byte Reload
; LINUXOSX64-NEXT: vmovdqu {{[-0-9]+}}(%r{{[sb]}}p), %ymm1 # 32-byte Reload
; LINUXOSX64-NEXT: vmovdqa32 %ymm1, %ymm0 {%k1}
; LINUXOSX64-NEXT: addq $88, %rsp
; LINUXOSX64-NEXT: addq $56, %rsp
; LINUXOSX64-NEXT: .cfi_def_cfa_offset 8
; LINUXOSX64-NEXT: retq
%b = call x86_regcallcc <8 x i32> @test_argRet256Vector(<8 x i1> %x, <8 x i32> %a, <8 x i32> %a)
@ -834,7 +834,7 @@ define dso_local x86_regcallcc <16 x i32> @test_argRet512Vector(<16 x i1> %x, <1
define dso_local x86_regcallcc <16 x i32> @test_CallargRet512Vector(<16 x i1> %x, <16 x i32> %a) {
; X32-LABEL: test_CallargRet512Vector:
; X32: # %bb.0:
; X32-NEXT: subl $188, %esp
; X32-NEXT: subl $68, %esp
; X32-NEXT: vmovdqu64 %zmm0, {{[-0-9]+}}(%e{{[sb]}}p) # 64-byte Spill
; X32-NEXT: kmovd %eax, %k1
; X32-NEXT: kmovw %k1, {{[-0-9]+}}(%e{{[sb]}}p) # 2-byte Spill
@ -843,13 +843,13 @@ define dso_local x86_regcallcc <16 x i32> @test_CallargRet512Vector(<16 x i1> %x
; X32-NEXT: kmovw {{[-0-9]+}}(%e{{[sb]}}p), %k1 # 2-byte Reload
; X32-NEXT: vmovdqu64 {{[-0-9]+}}(%e{{[sb]}}p), %zmm1 # 64-byte Reload
; X32-NEXT: vmovdqa32 %zmm1, %zmm0 {%k1}
; X32-NEXT: addl $188, %esp
; X32-NEXT: addl $68, %esp
; X32-NEXT: retl
;
; WIN64-LABEL: test_CallargRet512Vector:
; WIN64: # %bb.0:
; WIN64-NEXT: subq $184, %rsp
; WIN64-NEXT: .seh_stackalloc 184
; WIN64-NEXT: subq $88, %rsp
; WIN64-NEXT: .seh_stackalloc 88
; WIN64-NEXT: .seh_endprologue
; WIN64-NEXT: vmovdqu64 %zmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; WIN64-NEXT: kmovd %eax, %k1
@ -859,14 +859,14 @@ define dso_local x86_regcallcc <16 x i32> @test_CallargRet512Vector(<16 x i1> %x
; WIN64-NEXT: kmovw {{[-0-9]+}}(%r{{[sb]}}p), %k1 # 2-byte Reload
; WIN64-NEXT: vmovdqu64 {{[-0-9]+}}(%r{{[sb]}}p), %zmm1 # 64-byte Reload
; WIN64-NEXT: vmovdqa32 %zmm1, %zmm0 {%k1}
; WIN64-NEXT: addq $184, %rsp
; WIN64-NEXT: addq $88, %rsp
; WIN64-NEXT: retq
; WIN64-NEXT: .seh_endproc
;
; LINUXOSX64-LABEL: test_CallargRet512Vector:
; LINUXOSX64: # %bb.0:
; LINUXOSX64-NEXT: subq $184, %rsp
; LINUXOSX64-NEXT: .cfi_def_cfa_offset 192
; LINUXOSX64-NEXT: subq $88, %rsp
; LINUXOSX64-NEXT: .cfi_def_cfa_offset 96
; LINUXOSX64-NEXT: vmovdqu64 %zmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; LINUXOSX64-NEXT: kmovd %eax, %k1
; LINUXOSX64-NEXT: kmovw %k1, {{[-0-9]+}}(%r{{[sb]}}p) # 2-byte Spill
@ -875,7 +875,7 @@ define dso_local x86_regcallcc <16 x i32> @test_CallargRet512Vector(<16 x i1> %x
; LINUXOSX64-NEXT: kmovw {{[-0-9]+}}(%r{{[sb]}}p), %k1 # 2-byte Reload
; LINUXOSX64-NEXT: vmovdqu64 {{[-0-9]+}}(%r{{[sb]}}p), %zmm1 # 64-byte Reload
; LINUXOSX64-NEXT: vmovdqa32 %zmm1, %zmm0 {%k1}
; LINUXOSX64-NEXT: addq $184, %rsp
; LINUXOSX64-NEXT: addq $88, %rsp
; LINUXOSX64-NEXT: .cfi_def_cfa_offset 8
; LINUXOSX64-NEXT: retq
%b = call x86_regcallcc <16 x i32> @test_argRet512Vector(<16 x i1> %x, <16 x i32> %a, <16 x i32> %a)
@ -1267,7 +1267,7 @@ define x86_regcallcc %struct.complex @test_argMultiRet(float, double, i32, i8, i
;
; WIN64-LABEL: test_argMultiRet:
; WIN64: # %bb.0:
; WIN64-NEXT: vaddsd __real@{{.*}}(%rip), %xmm1, %xmm1
; WIN64-NEXT: vaddsd __real@4014000000000000(%rip), %xmm1, %xmm1
; WIN64-NEXT: movl $999, %edx # imm = 0x3E7
; WIN64-NEXT: movl $4, %eax
; WIN64-NEXT: movb $7, %cl
@ -1275,7 +1275,7 @@ define x86_regcallcc %struct.complex @test_argMultiRet(float, double, i32, i8, i
;
; LINUXOSX64-LABEL: test_argMultiRet:
; LINUXOSX64: # %bb.0:
; LINUXOSX64-NEXT: vaddsd {{.*}}(%rip), %xmm1, %xmm1
; LINUXOSX64-NEXT: vaddsd {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm1, %xmm1
; LINUXOSX64-NEXT: movl $999, %edx # imm = 0x3E7
; LINUXOSX64-NEXT: movl $4, %eax
; LINUXOSX64-NEXT: movb $7, %cl

4
test/CodeGen/X86/div-rem-pair-recomposition-signed.ll
View File

@ -651,7 +651,7 @@ define <2 x i64> @vector_i128_i64(<2 x i64> %x, <2 x i64> %y, <2 x i64>* %divdst
; X86-LABEL: vector_i128_i64:
; X86: # %bb.0:
; X86-NEXT: pushl %esi
; X86-NEXT: subl $72, %esp
; X86-NEXT: subl $64, %esp
; X86-NEXT: movdqu %xmm1, {{[-0-9]+}}(%e{{[sb]}}p) # 16-byte Spill
; X86-NEXT: movdqu %xmm0, {{[-0-9]+}}(%e{{[sb]}}p) # 16-byte Spill
; X86-NEXT: movl {{[0-9]+}}(%esp), %esi
@ -696,7 +696,7 @@ define <2 x i64> @vector_i128_i64(<2 x i64> %x, <2 x i64> %y, <2 x i64>* %divdst
; X86-NEXT: paddq %xmm2, %xmm1
; X86-NEXT: movdqu {{[-0-9]+}}(%e{{[sb]}}p), %xmm0 # 16-byte Reload
; X86-NEXT: psubq %xmm1, %xmm0
; X86-NEXT: addl $72, %esp
; X86-NEXT: addl $64, %esp
; X86-NEXT: popl %esi
; X86-NEXT: retl
;

4
test/CodeGen/X86/div-rem-pair-recomposition-unsigned.ll
View File

@ -651,7 +651,7 @@ define <2 x i64> @vector_i128_i64(<2 x i64> %x, <2 x i64> %y, <2 x i64>* %divdst
; X86-LABEL: vector_i128_i64:
; X86: # %bb.0:
; X86-NEXT: pushl %esi
; X86-NEXT: subl $72, %esp
; X86-NEXT: subl $64, %esp
; X86-NEXT: movdqu %xmm1, {{[-0-9]+}}(%e{{[sb]}}p) # 16-byte Spill
; X86-NEXT: movdqu %xmm0, {{[-0-9]+}}(%e{{[sb]}}p) # 16-byte Spill
; X86-NEXT: movl {{[0-9]+}}(%esp), %esi
@ -696,7 +696,7 @@ define <2 x i64> @vector_i128_i64(<2 x i64> %x, <2 x i64> %y, <2 x i64>* %divdst
; X86-NEXT: paddq %xmm2, %xmm1
; X86-NEXT: movdqu {{[-0-9]+}}(%e{{[sb]}}p), %xmm0 # 16-byte Reload
; X86-NEXT: psubq %xmm1, %xmm0
; X86-NEXT: addl $72, %esp
; X86-NEXT: addl $64, %esp
; X86-NEXT: popl %esi
; X86-NEXT: retl
;

150
test/CodeGen/X86/fma.ll
View File

@ -251,41 +251,41 @@ entry:
define float @test_f32_cst() #0 {
; FMA32-LABEL: test_f32_cst:
; FMA32: ## %bb.0: ## %entry
; FMA32-NEXT: flds LCPI4_0 ## encoding: [0xd9,0x05,A,A,A,A]
; FMA32-NEXT: ## fixup A - offset: 2, value: LCPI4_0, kind: FK_Data_4
; FMA32-NEXT: flds {{\.?LCPI[0-9]+_[0-9]+}} ## encoding: [0xd9,0x05,A,A,A,A]
; FMA32-NEXT: ## fixup A - offset: 2, value: {{\.?LCPI[0-9]+_[0-9]+}}, kind: FK_Data_4
; FMA32-NEXT: retl ## encoding: [0xc3]
;
; FMACALL32-LABEL: test_f32_cst:
; FMACALL32: ## %bb.0: ## %entry
; FMACALL32-NEXT: flds LCPI4_0 ## encoding: [0xd9,0x05,A,A,A,A]
; FMACALL32-NEXT: ## fixup A - offset: 2, value: LCPI4_0, kind: FK_Data_4
; FMACALL32-NEXT: flds {{\.?LCPI[0-9]+_[0-9]+}} ## encoding: [0xd9,0x05,A,A,A,A]
; FMACALL32-NEXT: ## fixup A - offset: 2, value: {{\.?LCPI[0-9]+_[0-9]+}}, kind: FK_Data_4
; FMACALL32-NEXT: retl ## encoding: [0xc3]
;
; FMA64-LABEL: test_f32_cst:
; FMA64: ## %bb.0: ## %entry
; FMA64-NEXT: vmovss {{.*}}(%rip), %xmm0 ## encoding: [0xc5,0xfa,0x10,0x05,A,A,A,A]
; FMA64-NEXT: ## fixup A - offset: 4, value: LCPI4_0-4, kind: reloc_riprel_4byte
; FMA64-NEXT: vmovss {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0 ## encoding: [0xc5,0xfa,0x10,0x05,A,A,A,A]
; FMA64-NEXT: ## fixup A - offset: 4, value: {{\.?LCPI[0-9]+_[0-9]+}}-4, kind: reloc_riprel_4byte
; FMA64-NEXT: ## xmm0 = mem[0],zero,zero,zero
; FMA64-NEXT: retq ## encoding: [0xc3]
;
; FMACALL64-LABEL: test_f32_cst:
; FMACALL64: ## %bb.0: ## %entry
; FMACALL64-NEXT: movss {{.*}}(%rip), %xmm0 ## encoding: [0xf3,0x0f,0x10,0x05,A,A,A,A]
; FMACALL64-NEXT: ## fixup A - offset: 4, value: LCPI4_0-4, kind: reloc_riprel_4byte
; FMACALL64-NEXT: movss {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0 ## encoding: [0xf3,0x0f,0x10,0x05,A,A,A,A]
; FMACALL64-NEXT: ## fixup A - offset: 4, value: {{\.?LCPI[0-9]+_[0-9]+}}-4, kind: reloc_riprel_4byte
; FMACALL64-NEXT: ## xmm0 = mem[0],zero,zero,zero
; FMACALL64-NEXT: retq ## encoding: [0xc3]
;
; AVX512-LABEL: test_f32_cst:
; AVX512: ## %bb.0: ## %entry
; AVX512-NEXT: vmovss {{.*}}(%rip), %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xfa,0x10,0x05,A,A,A,A]
; AVX512-NEXT: ## fixup A - offset: 4, value: LCPI4_0-4, kind: reloc_riprel_4byte
; AVX512-NEXT: vmovss {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xfa,0x10,0x05,A,A,A,A]
; AVX512-NEXT: ## fixup A - offset: 4, value: {{\.?LCPI[0-9]+_[0-9]+}}-4, kind: reloc_riprel_4byte
; AVX512-NEXT: ## xmm0 = mem[0],zero,zero,zero
; AVX512-NEXT: retq ## encoding: [0xc3]
;
; AVX512VL-LABEL: test_f32_cst:
; AVX512VL: ## %bb.0: ## %entry
; AVX512VL-NEXT: vmovss {{.*}}(%rip), %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xfa,0x10,0x05,A,A,A,A]
; AVX512VL-NEXT: ## fixup A - offset: 4, value: LCPI4_0-4, kind: reloc_riprel_4byte
; AVX512VL-NEXT: vmovss {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xfa,0x10,0x05,A,A,A,A]
; AVX512VL-NEXT: ## fixup A - offset: 4, value: {{\.?LCPI[0-9]+_[0-9]+}}-4, kind: reloc_riprel_4byte
; AVX512VL-NEXT: ## xmm0 = mem[0],zero,zero,zero
; AVX512VL-NEXT: retq ## encoding: [0xc3]
entry:
@ -642,161 +642,161 @@ define <8 x float> @test_v8f32(<8 x float> %a, <8 x float> %b, <8 x float> %c) #
;
; FMACALL32_BDVER2-LABEL: test_v8f32:
; FMACALL32_BDVER2: ## %bb.0: ## %entry
; FMACALL32_BDVER2-NEXT: subl $316, %esp ## encoding: [0x81,0xec,0x3c,0x01,0x00,0x00]
; FMACALL32_BDVER2-NEXT: ## imm = 0x13C
; FMACALL32_BDVER2-NEXT: subl $284, %esp ## encoding: [0x81,0xec,0x1c,0x01,0x00,0x00]
; FMACALL32_BDVER2-NEXT: ## imm = 0x11C
; FMACALL32_BDVER2-NEXT: vextractf128 $1, %ymm2, %xmm3 ## encoding: [0xc4,0xe3,0x7d,0x19,0xd3,0x01]
; FMACALL32_BDVER2-NEXT: vmovups %ymm2, {{[-0-9]+}}(%e{{[sb]}}p) ## 32-byte Spill
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xfc,0x11,0x94,0x24,0x00,0x01,0x00,0x00]
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xfc,0x11,0x94,0x24,0xf0,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: vextractf128 $1, %ymm1, %xmm2 ## encoding: [0xc4,0xe3,0x7d,0x19,0xca,0x01]
; FMACALL32_BDVER2-NEXT: vmovups %ymm1, {{[-0-9]+}}(%e{{[sb]}}p) ## 32-byte Spill
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xfc,0x11,0x8c,0x24,0xe0,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xfc,0x11,0x8c,0x24,0xd0,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: vextractf128 $1, %ymm0, %xmm1 ## encoding: [0xc4,0xe3,0x7d,0x19,0xc1,0x01]
; FMACALL32_BDVER2-NEXT: vmovups %ymm0, {{[-0-9]+}}(%e{{[sb]}}p) ## 32-byte Spill
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xfc,0x11,0x84,0x24,0xc0,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xfc,0x11,0x84,0x24,0xb0,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: vmovaps %xmm3, {{[-0-9]+}}(%e{{[sb]}}p) ## 16-byte Spill
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xf8,0x29,0x5c,0x24,0x60]
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xf8,0x29,0x5c,0x24,0x50]
; FMACALL32_BDVER2-NEXT: vextractps $2, %xmm3, {{[0-9]+}}(%esp) ## encoding: [0xc4,0xe3,0x79,0x17,0x5c,0x24,0x08,0x02]
; FMACALL32_BDVER2-NEXT: vmovaps %xmm2, {{[-0-9]+}}(%e{{[sb]}}p) ## 16-byte Spill
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xf8,0x29,0x54,0x24,0x50]
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xf8,0x29,0x54,0x24,0x40]
; FMACALL32_BDVER2-NEXT: vextractps $2, %xmm2, {{[0-9]+}}(%esp) ## encoding: [0xc4,0xe3,0x79,0x17,0x54,0x24,0x04,0x02]
; FMACALL32_BDVER2-NEXT: vmovaps %xmm1, {{[-0-9]+}}(%e{{[sb]}}p) ## 16-byte Spill
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xf8,0x29,0x4c,0x24,0x40]
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xf8,0x29,0x4c,0x24,0x30]
; FMACALL32_BDVER2-NEXT: vextractps $2, %xmm1, (%esp) ## encoding: [0xc4,0xe3,0x79,0x17,0x0c,0x24,0x02]
; FMACALL32_BDVER2-NEXT: vzeroupper ## encoding: [0xc5,0xf8,0x77]
; FMACALL32_BDVER2-NEXT: calll _fmaf ## encoding: [0xe8,A,A,A,A]
; FMACALL32_BDVER2-NEXT: ## fixup A - offset: 1, value: _fmaf-4, kind: FK_PCRel_4
; FMACALL32_BDVER2-NEXT: fstpt {{[-0-9]+}}(%e{{[sb]}}p) ## 10-byte Folded Spill
; FMACALL32_BDVER2-NEXT: ## encoding: [0xdb,0xbc,0x24,0xb4,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: ## encoding: [0xdb,0xbc,0x24,0xa4,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: vmovaps {{[-0-9]+}}(%e{{[sb]}}p), %xmm0 ## 16-byte Reload
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xf8,0x28,0x44,0x24,0x60]
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xf8,0x28,0x44,0x24,0x50]
; FMACALL32_BDVER2-NEXT: vmovaps {{[-0-9]+}}(%e{{[sb]}}p), %xmm2 ## 16-byte Reload
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xf8,0x28,0x54,0x24,0x50]
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xf8,0x28,0x54,0x24,0x40]
; FMACALL32_BDVER2-NEXT: vmovaps {{[-0-9]+}}(%e{{[sb]}}p), %xmm1 ## 16-byte Reload
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xf8,0x28,0x4c,0x24,0x40]
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xf8,0x28,0x4c,0x24,0x30]
; FMACALL32_BDVER2-NEXT: vextractps $1, %xmm0, {{[0-9]+}}(%esp) ## encoding: [0xc4,0xe3,0x79,0x17,0x44,0x24,0x08,0x01]
; FMACALL32_BDVER2-NEXT: vextractps $1, %xmm2, {{[0-9]+}}(%esp) ## encoding: [0xc4,0xe3,0x79,0x17,0x54,0x24,0x04,0x01]
; FMACALL32_BDVER2-NEXT: vextractps $1, %xmm1, (%esp) ## encoding: [0xc4,0xe3,0x79,0x17,0x0c,0x24,0x01]
; FMACALL32_BDVER2-NEXT: calll _fmaf ## encoding: [0xe8,A,A,A,A]
; FMACALL32_BDVER2-NEXT: ## fixup A - offset: 1, value: _fmaf-4, kind: FK_PCRel_4
; FMACALL32_BDVER2-NEXT: fstpt {{[-0-9]+}}(%e{{[sb]}}p) ## 10-byte Folded Spill
; FMACALL32_BDVER2-NEXT: ## encoding: [0xdb,0xbc,0x24,0xa8,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: vmovaps {{[-0-9]+}}(%e{{[sb]}}p), %xmm0 ## 16-byte Reload
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xf8,0x28,0x44,0x24,0x60]
; FMACALL32_BDVER2-NEXT: vmovss %xmm0, {{[0-9]+}}(%esp) ## encoding: [0xc5,0xfa,0x11,0x44,0x24,0x08]
; FMACALL32_BDVER2-NEXT: ## encoding: [0xdb,0xbc,0x24,0x98,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: vmovaps {{[-0-9]+}}(%e{{[sb]}}p), %xmm0 ## 16-byte Reload
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xf8,0x28,0x44,0x24,0x50]
; FMACALL32_BDVER2-NEXT: vmovss %xmm0, {{[0-9]+}}(%esp) ## encoding: [0xc5,0xfa,0x11,0x44,0x24,0x04]
; FMACALL32_BDVER2-NEXT: vmovss %xmm0, {{[0-9]+}}(%esp) ## encoding: [0xc5,0xfa,0x11,0x44,0x24,0x08]
; FMACALL32_BDVER2-NEXT: vmovaps {{[-0-9]+}}(%e{{[sb]}}p), %xmm0 ## 16-byte Reload
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xf8,0x28,0x44,0x24,0x40]
; FMACALL32_BDVER2-NEXT: vmovss %xmm0, {{[0-9]+}}(%esp) ## encoding: [0xc5,0xfa,0x11,0x44,0x24,0x04]
; FMACALL32_BDVER2-NEXT: vmovaps {{[-0-9]+}}(%e{{[sb]}}p), %xmm0 ## 16-byte Reload
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xf8,0x28,0x44,0x24,0x30]
; FMACALL32_BDVER2-NEXT: vmovss %xmm0, (%esp) ## encoding: [0xc5,0xfa,0x11,0x04,0x24]
; FMACALL32_BDVER2-NEXT: calll _fmaf ## encoding: [0xe8,A,A,A,A]
; FMACALL32_BDVER2-NEXT: ## fixup A - offset: 1, value: _fmaf-4, kind: FK_PCRel_4
; FMACALL32_BDVER2-NEXT: fstpt {{[-0-9]+}}(%e{{[sb]}}p) ## 10-byte Folded Spill
; FMACALL32_BDVER2-NEXT: ## encoding: [0xdb,0xbc,0x24,0x9c,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: ## encoding: [0xdb,0xbc,0x24,0x8c,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: vmovups {{[-0-9]+}}(%e{{[sb]}}p), %ymm0 ## 32-byte Reload
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xfc,0x10,0x84,0x24,0x00,0x01,0x00,0x00]
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xfc,0x10,0x84,0x24,0xf0,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: vextractps $3, %xmm0, {{[0-9]+}}(%esp) ## encoding: [0xc4,0xe3,0x79,0x17,0x44,0x24,0x08,0x03]
; FMACALL32_BDVER2-NEXT: vmovups {{[-0-9]+}}(%e{{[sb]}}p), %ymm0 ## 32-byte Reload
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xfc,0x10,0x84,0x24,0xe0,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xfc,0x10,0x84,0x24,0xd0,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: vextractps $3, %xmm0, {{[0-9]+}}(%esp) ## encoding: [0xc4,0xe3,0x79,0x17,0x44,0x24,0x04,0x03]
; FMACALL32_BDVER2-NEXT: vmovups {{[-0-9]+}}(%e{{[sb]}}p), %ymm0 ## 32-byte Reload
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xfc,0x10,0x84,0x24,0xc0,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xfc,0x10,0x84,0x24,0xb0,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: vextractps $3, %xmm0, (%esp) ## encoding: [0xc4,0xe3,0x79,0x17,0x04,0x24,0x03]
; FMACALL32_BDVER2-NEXT: vzeroupper ## encoding: [0xc5,0xf8,0x77]
; FMACALL32_BDVER2-NEXT: calll _fmaf ## encoding: [0xe8,A,A,A,A]
; FMACALL32_BDVER2-NEXT: ## fixup A - offset: 1, value: _fmaf-4, kind: FK_PCRel_4
; FMACALL32_BDVER2-NEXT: fstpt {{[-0-9]+}}(%e{{[sb]}}p) ## 10-byte Folded Spill
; FMACALL32_BDVER2-NEXT: ## encoding: [0xdb,0xbc,0x24,0x90,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: ## encoding: [0xdb,0xbc,0x24,0x80,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: vmovups {{[-0-9]+}}(%e{{[sb]}}p), %ymm0 ## 32-byte Reload
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xfc,0x10,0x84,0x24,0x00,0x01,0x00,0x00]
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xfc,0x10,0x84,0x24,0xf0,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: vextractps $2, %xmm0, {{[0-9]+}}(%esp) ## encoding: [0xc4,0xe3,0x79,0x17,0x44,0x24,0x08,0x02]
; FMACALL32_BDVER2-NEXT: vmovups {{[-0-9]+}}(%e{{[sb]}}p), %ymm0 ## 32-byte Reload
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xfc,0x10,0x84,0x24,0xe0,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xfc,0x10,0x84,0x24,0xd0,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: vextractps $2, %xmm0, {{[0-9]+}}(%esp) ## encoding: [0xc4,0xe3,0x79,0x17,0x44,0x24,0x04,0x02]
; FMACALL32_BDVER2-NEXT: vmovups {{[-0-9]+}}(%e{{[sb]}}p), %ymm0 ## 32-byte Reload
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xfc,0x10,0x84,0x24,0xc0,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xfc,0x10,0x84,0x24,0xb0,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: vextractps $2, %xmm0, (%esp) ## encoding: [0xc4,0xe3,0x79,0x17,0x04,0x24,0x02]
; FMACALL32_BDVER2-NEXT: vzeroupper ## encoding: [0xc5,0xf8,0x77]
; FMACALL32_BDVER2-NEXT: calll _fmaf ## encoding: [0xe8,A,A,A,A]
; FMACALL32_BDVER2-NEXT: ## fixup A - offset: 1, value: _fmaf-4, kind: FK_PCRel_4
; FMACALL32_BDVER2-NEXT: fstpt {{[-0-9]+}}(%e{{[sb]}}p) ## 10-byte Folded Spill
; FMACALL32_BDVER2-NEXT: ## encoding: [0xdb,0xbc,0x24,0x84,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: ## encoding: [0xdb,0x7c,0x24,0x74]
; FMACALL32_BDVER2-NEXT: vmovups {{[-0-9]+}}(%e{{[sb]}}p), %ymm0 ## 32-byte Reload
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xfc,0x10,0x84,0x24,0x00,0x01,0x00,0x00]
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xfc,0x10,0x84,0x24,0xf0,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: vextractps $1, %xmm0, {{[0-9]+}}(%esp) ## encoding: [0xc4,0xe3,0x79,0x17,0x44,0x24,0x08,0x01]
; FMACALL32_BDVER2-NEXT: vmovups {{[-0-9]+}}(%e{{[sb]}}p), %ymm0 ## 32-byte Reload
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xfc,0x10,0x84,0x24,0xe0,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xfc,0x10,0x84,0x24,0xd0,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: vextractps $1, %xmm0, {{[0-9]+}}(%esp) ## encoding: [0xc4,0xe3,0x79,0x17,0x44,0x24,0x04,0x01]
; FMACALL32_BDVER2-NEXT: vmovups {{[-0-9]+}}(%e{{[sb]}}p), %ymm0 ## 32-byte Reload
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xfc,0x10,0x84,0x24,0xc0,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xfc,0x10,0x84,0x24,0xb0,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: vextractps $1, %xmm0, (%esp) ## encoding: [0xc4,0xe3,0x79,0x17,0x04,0x24,0x01]
; FMACALL32_BDVER2-NEXT: vzeroupper ## encoding: [0xc5,0xf8,0x77]
; FMACALL32_BDVER2-NEXT: calll _fmaf ## encoding: [0xe8,A,A,A,A]
; FMACALL32_BDVER2-NEXT: ## fixup A - offset: 1, value: _fmaf-4, kind: FK_PCRel_4
; FMACALL32_BDVER2-NEXT: fstpt {{[-0-9]+}}(%e{{[sb]}}p) ## 10-byte Folded Spill
; FMACALL32_BDVER2-NEXT: ## encoding: [0xdb,0x7c,0x24,0x78]
; FMACALL32_BDVER2-NEXT: ## encoding: [0xdb,0x7c,0x24,0x68]
; FMACALL32_BDVER2-NEXT: vmovups {{[-0-9]+}}(%e{{[sb]}}p), %ymm0 ## 32-byte Reload
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xfc,0x10,0x84,0x24,0x00,0x01,0x00,0x00]
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xfc,0x10,0x84,0x24,0xf0,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: vmovss %xmm0, {{[0-9]+}}(%esp) ## encoding: [0xc5,0xfa,0x11,0x44,0x24,0x08]
; FMACALL32_BDVER2-NEXT: vmovups {{[-0-9]+}}(%e{{[sb]}}p), %ymm0 ## 32-byte Reload
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xfc,0x10,0x84,0x24,0xe0,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xfc,0x10,0x84,0x24,0xd0,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: vmovss %xmm0, {{[0-9]+}}(%esp) ## encoding: [0xc5,0xfa,0x11,0x44,0x24,0x04]
; FMACALL32_BDVER2-NEXT: vmovups {{[-0-9]+}}(%e{{[sb]}}p), %ymm0 ## 32-byte Reload
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xfc,0x10,0x84,0x24,0xc0,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xfc,0x10,0x84,0x24,0xb0,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: vmovss %xmm0, (%esp) ## encoding: [0xc5,0xfa,0x11,0x04,0x24]
; FMACALL32_BDVER2-NEXT: vzeroupper ## encoding: [0xc5,0xf8,0x77]
; FMACALL32_BDVER2-NEXT: calll _fmaf ## encoding: [0xe8,A,A,A,A]
; FMACALL32_BDVER2-NEXT: ## fixup A - offset: 1, value: _fmaf-4, kind: FK_PCRel_4
; FMACALL32_BDVER2-NEXT: vmovaps {{[-0-9]+}}(%e{{[sb]}}p), %xmm0 ## 16-byte Reload
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xf8,0x28,0x44,0x24,0x60]
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xf8,0x28,0x44,0x24,0x50]
; FMACALL32_BDVER2-NEXT: vmovaps {{[-0-9]+}}(%e{{[sb]}}p), %xmm2 ## 16-byte Reload
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xf8,0x28,0x54,0x24,0x50]
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xf8,0x28,0x54,0x24,0x40]
; FMACALL32_BDVER2-NEXT: vmovaps {{[-0-9]+}}(%e{{[sb]}}p), %xmm1 ## 16-byte Reload
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xf8,0x28,0x4c,0x24,0x40]
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xf8,0x28,0x4c,0x24,0x30]
; FMACALL32_BDVER2-NEXT: vextractps $3, %xmm0, {{[0-9]+}}(%esp) ## encoding: [0xc4,0xe3,0x79,0x17,0x44,0x24,0x08,0x03]
; FMACALL32_BDVER2-NEXT: vextractps $3, %xmm2, {{[0-9]+}}(%esp) ## encoding: [0xc4,0xe3,0x79,0x17,0x54,0x24,0x04,0x03]
; FMACALL32_BDVER2-NEXT: vextractps $3, %xmm1, (%esp) ## encoding: [0xc4,0xe3,0x79,0x17,0x0c,0x24,0x03]
; FMACALL32_BDVER2-NEXT: fstps {{[0-9]+}}(%esp) ## encoding: [0xd9,0x5c,0x24,0x3c]
; FMACALL32_BDVER2-NEXT: fldt {{[-0-9]+}}(%e{{[sb]}}p) ## 10-byte Folded Reload
; FMACALL32_BDVER2-NEXT: ## encoding: [0xdb,0x6c,0x24,0x78]
; FMACALL32_BDVER2-NEXT: fstps {{[0-9]+}}(%esp) ## encoding: [0xd9,0x5c,0x24,0x38]
; FMACALL32_BDVER2-NEXT: fldt {{[-0-9]+}}(%e{{[sb]}}p) ## 10-byte Folded Reload
; FMACALL32_BDVER2-NEXT: ## encoding: [0xdb,0xac,0x24,0x84,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: fstps {{[0-9]+}}(%esp) ## encoding: [0xd9,0x5c,0x24,0x34]
; FMACALL32_BDVER2-NEXT: fldt {{[-0-9]+}}(%e{{[sb]}}p) ## 10-byte Folded Reload
; FMACALL32_BDVER2-NEXT: ## encoding: [0xdb,0xac,0x24,0x90,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: fstps {{[0-9]+}}(%esp) ## encoding: [0xd9,0x5c,0x24,0x30]
; FMACALL32_BDVER2-NEXT: fldt {{[-0-9]+}}(%e{{[sb]}}p) ## 10-byte Folded Reload
; FMACALL32_BDVER2-NEXT: ## encoding: [0xdb,0xac,0x24,0x9c,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: fstps {{[0-9]+}}(%esp) ## encoding: [0xd9,0x5c,0x24,0x2c]
; FMACALL32_BDVER2-NEXT: fldt {{[-0-9]+}}(%e{{[sb]}}p) ## 10-byte Folded Reload
; FMACALL32_BDVER2-NEXT: ## encoding: [0xdb,0xac,0x24,0xa8,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: ## encoding: [0xdb,0x6c,0x24,0x68]
; FMACALL32_BDVER2-NEXT: fstps {{[0-9]+}}(%esp) ## encoding: [0xd9,0x5c,0x24,0x28]
; FMACALL32_BDVER2-NEXT: fldt {{[-0-9]+}}(%e{{[sb]}}p) ## 10-byte Folded Reload
; FMACALL32_BDVER2-NEXT: ## encoding: [0xdb,0xac,0x24,0xb4,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: ## encoding: [0xdb,0x6c,0x24,0x74]
; FMACALL32_BDVER2-NEXT: fstps {{[0-9]+}}(%esp) ## encoding: [0xd9,0x5c,0x24,0x24]
; FMACALL32_BDVER2-NEXT: fldt {{[-0-9]+}}(%e{{[sb]}}p) ## 10-byte Folded Reload
; FMACALL32_BDVER2-NEXT: ## encoding: [0xdb,0xac,0x24,0x80,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: fstps {{[0-9]+}}(%esp) ## encoding: [0xd9,0x5c,0x24,0x20]
; FMACALL32_BDVER2-NEXT: fldt {{[-0-9]+}}(%e{{[sb]}}p) ## 10-byte Folded Reload
; FMACALL32_BDVER2-NEXT: ## encoding: [0xdb,0xac,0x24,0x8c,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: fstps {{[0-9]+}}(%esp) ## encoding: [0xd9,0x5c,0x24,0x1c]
; FMACALL32_BDVER2-NEXT: fldt {{[-0-9]+}}(%e{{[sb]}}p) ## 10-byte Folded Reload
; FMACALL32_BDVER2-NEXT: ## encoding: [0xdb,0xac,0x24,0x98,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: fstps {{[0-9]+}}(%esp) ## encoding: [0xd9,0x5c,0x24,0x18]
; FMACALL32_BDVER2-NEXT: fldt {{[-0-9]+}}(%e{{[sb]}}p) ## 10-byte Folded Reload
; FMACALL32_BDVER2-NEXT: ## encoding: [0xdb,0xac,0x24,0xa4,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: fstps {{[0-9]+}}(%esp) ## encoding: [0xd9,0x5c,0x24,0x14]
; FMACALL32_BDVER2-NEXT: calll _fmaf ## encoding: [0xe8,A,A,A,A]
; FMACALL32_BDVER2-NEXT: ## fixup A - offset: 1, value: _fmaf-4, kind: FK_PCRel_4
; FMACALL32_BDVER2-NEXT: fstps {{[0-9]+}}(%esp) ## encoding: [0xd9,0x5c,0x24,0x20]
; FMACALL32_BDVER2-NEXT: vmovss {{[0-9]+}}(%esp), %xmm0 ## encoding: [0xc5,0xfa,0x10,0x44,0x24,0x3c]
; FMACALL32_BDVER2-NEXT: fstps {{[0-9]+}}(%esp) ## encoding: [0xd9,0x5c,0x24,0x10]
; FMACALL32_BDVER2-NEXT: vmovss {{[0-9]+}}(%esp), %xmm0 ## encoding: [0xc5,0xfa,0x10,0x44,0x24,0x2c]
; FMACALL32_BDVER2-NEXT: ## xmm0 = mem[0],zero,zero,zero
; FMACALL32_BDVER2-NEXT: vinsertps $16, {{[0-9]+}}(%esp), %xmm0, %xmm0 ## encoding: [0xc4,0xe3,0x79,0x21,0x44,0x24,0x38,0x10]
; FMACALL32_BDVER2-NEXT: vinsertps $16, {{[0-9]+}}(%esp), %xmm0, %xmm0 ## encoding: [0xc4,0xe3,0x79,0x21,0x44,0x24,0x28,0x10]
; FMACALL32_BDVER2-NEXT: ## xmm0 = xmm0[0],mem[0],xmm0[2,3]
; FMACALL32_BDVER2-NEXT: vmovss {{[0-9]+}}(%esp), %xmm1 ## encoding: [0xc5,0xfa,0x10,0x4c,0x24,0x2c]
; FMACALL32_BDVER2-NEXT: vmovss {{[0-9]+}}(%esp), %xmm1 ## encoding: [0xc5,0xfa,0x10,0x4c,0x24,0x1c]
; FMACALL32_BDVER2-NEXT: ## xmm1 = mem[0],zero,zero,zero
; FMACALL32_BDVER2-NEXT: vinsertps $16, {{[0-9]+}}(%esp), %xmm1, %xmm1 ## encoding: [0xc4,0xe3,0x71,0x21,0x4c,0x24,0x28,0x10]
; FMACALL32_BDVER2-NEXT: vinsertps $16, {{[0-9]+}}(%esp), %xmm1, %xmm1 ## encoding: [0xc4,0xe3,0x71,0x21,0x4c,0x24,0x18,0x10]
; FMACALL32_BDVER2-NEXT: ## xmm1 = xmm1[0],mem[0],xmm1[2,3]
; FMACALL32_BDVER2-NEXT: vinsertps $32, {{[0-9]+}}(%esp), %xmm0, %xmm0 ## encoding: [0xc4,0xe3,0x79,0x21,0x44,0x24,0x34,0x20]
; FMACALL32_BDVER2-NEXT: vinsertps $32, {{[0-9]+}}(%esp), %xmm0, %xmm0 ## encoding: [0xc4,0xe3,0x79,0x21,0x44,0x24,0x24,0x20]
; FMACALL32_BDVER2-NEXT: ## xmm0 = xmm0[0,1],mem[0],xmm0[3]
; FMACALL32_BDVER2-NEXT: vinsertps $32, {{[0-9]+}}(%esp), %xmm1, %xmm1 ## encoding: [0xc4,0xe3,0x71,0x21,0x4c,0x24,0x24,0x20]
; FMACALL32_BDVER2-NEXT: vinsertps $32, {{[0-9]+}}(%esp), %xmm1, %xmm1 ## encoding: [0xc4,0xe3,0x71,0x21,0x4c,0x24,0x14,0x20]
; FMACALL32_BDVER2-NEXT: ## xmm1 = xmm1[0,1],mem[0],xmm1[3]
; FMACALL32_BDVER2-NEXT: vinsertps $48, {{[0-9]+}}(%esp), %xmm0, %xmm0 ## encoding: [0xc4,0xe3,0x79,0x21,0x44,0x24,0x30,0x30]
; FMACALL32_BDVER2-NEXT: vinsertps $48, {{[0-9]+}}(%esp), %xmm0, %xmm0 ## encoding: [0xc4,0xe3,0x79,0x21,0x44,0x24,0x20,0x30]
; FMACALL32_BDVER2-NEXT: ## xmm0 = xmm0[0,1,2],mem[0]
; FMACALL32_BDVER2-NEXT: vinsertps $48, {{[0-9]+}}(%esp), %xmm1, %xmm1 ## encoding: [0xc4,0xe3,0x71,0x21,0x4c,0x24,0x20,0x30]
; FMACALL32_BDVER2-NEXT: vinsertps $48, {{[0-9]+}}(%esp), %xmm1, %xmm1 ## encoding: [0xc4,0xe3,0x71,0x21,0x4c,0x24,0x10,0x30]
; FMACALL32_BDVER2-NEXT: ## xmm1 = xmm1[0,1,2],mem[0]
; FMACALL32_BDVER2-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0 ## encoding: [0xc4,0xe3,0x7d,0x18,0xc1,0x01]
; FMACALL32_BDVER2-NEXT: addl $316, %esp ## encoding: [0x81,0xc4,0x3c,0x01,0x00,0x00]
; FMACALL32_BDVER2-NEXT: ## imm = 0x13C
; FMACALL32_BDVER2-NEXT: addl $284, %esp ## encoding: [0x81,0xc4,0x1c,0x01,0x00,0x00]
; FMACALL32_BDVER2-NEXT: ## imm = 0x11C
; FMACALL32_BDVER2-NEXT: retl ## encoding: [0xc3]
entry:
%call = call <8 x float> @llvm.fma.v8f32(<8 x float> %a, <8 x float> %b, <8 x float> %c)
@ -1663,7 +1663,7 @@ define <4 x double> @test_v4f64(<4 x double> %a, <4 x double> %b, <4 x double> %
;
; FMACALL32_BDVER2-LABEL: test_v4f64:
; FMACALL32_BDVER2: ## %bb.0: ## %entry
; FMACALL32_BDVER2-NEXT: subl $252, %esp ## encoding: [0x81,0xec,0xfc,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: subl $236, %esp ## encoding: [0x81,0xec,0xec,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: vextractf128 $1, %ymm2, %xmm3 ## encoding: [0xc4,0xe3,0x7d,0x19,0xd3,0x01]
; FMACALL32_BDVER2-NEXT: vmovups %ymm2, {{[-0-9]+}}(%e{{[sb]}}p) ## 32-byte Spill
; FMACALL32_BDVER2-NEXT: ## encoding: [0xc5,0xfc,0x11,0x94,0x24,0xc0,0x00,0x00,0x00]
@ -1742,7 +1742,7 @@ define <4 x double> @test_v4f64(<4 x double> %a, <4 x double> %b, <4 x double> %
; FMACALL32_BDVER2-NEXT: vmovhps {{[0-9]+}}(%esp), %xmm1, %xmm1 ## encoding: [0xc5,0xf0,0x16,0x4c,0x24,0x18]
; FMACALL32_BDVER2-NEXT: ## xmm1 = xmm1[0,1],mem[0,1]
; FMACALL32_BDVER2-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0 ## encoding: [0xc4,0xe3,0x7d,0x18,0xc1,0x01]
; FMACALL32_BDVER2-NEXT: addl $252, %esp ## encoding: [0x81,0xc4,0xfc,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: addl $236, %esp ## encoding: [0x81,0xc4,0xec,0x00,0x00,0x00]
; FMACALL32_BDVER2-NEXT: retl ## encoding: [0xc3]
entry:
%call = call <4 x double> @llvm.fma.v4f64(<4 x double> %a, <4 x double> %b, <4 x double> %c)

10
test/CodeGen/X86/half.ll
View File

@ -364,7 +364,7 @@ define void @test_uitofp_i64(i64 %a, half* %p) #0 {
; CHECK-I686-NEXT: movlps %xmm0, {{[0-9]+}}(%esp)
; CHECK-I686-NEXT: shrl $31, %eax
; CHECK-I686-NEXT: fildll {{[0-9]+}}(%esp)
; CHECK-I686-NEXT: fadds {{\.LCPI[0-9]+_[0-9]+}}(,%eax,4)
; CHECK-I686-NEXT: fadds {{\.?LCPI[0-9]+_[0-9]+}}(,%eax,4)
; CHECK-I686-NEXT: fstps (%esp)
; CHECK-I686-NEXT: calll __gnu_f2h_ieee
; CHECK-I686-NEXT: movw %ax, (%esi)
@ -735,7 +735,7 @@ define void @test_trunc64_vec4(<4 x double> %a, <4 x half>* %p) #0 {
; BWON-F16C-NEXT: pushq %r15
; BWON-F16C-NEXT: pushq %r14
; BWON-F16C-NEXT: pushq %rbx
; BWON-F16C-NEXT: subq $88, %rsp
; BWON-F16C-NEXT: subq $56, %rsp
; BWON-F16C-NEXT: movq %rdi, %rbx
; BWON-F16C-NEXT: vmovupd %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; BWON-F16C-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
@ -744,7 +744,7 @@ define void @test_trunc64_vec4(<4 x double> %a, <4 x half>* %p) #0 {
; BWON-F16C-NEXT: movl %eax, %r14d
; BWON-F16C-NEXT: vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; BWON-F16C-NEXT: vextractf128 $1, %ymm0, %xmm0
; BWON-F16C-NEXT: vmovapd %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; BWON-F16C-NEXT: vmovapd %xmm0, (%rsp) # 16-byte Spill
; BWON-F16C-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; BWON-F16C-NEXT: vzeroupper
; BWON-F16C-NEXT: callq __truncdfhf2@PLT
@ -754,13 +754,13 @@ define void @test_trunc64_vec4(<4 x double> %a, <4 x half>* %p) #0 {
; BWON-F16C-NEXT: vzeroupper
; BWON-F16C-NEXT: callq __truncdfhf2@PLT
; BWON-F16C-NEXT: movl %eax, %ebp
; BWON-F16C-NEXT: vmovaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; BWON-F16C-NEXT: vmovaps (%rsp), %xmm0 # 16-byte Reload
; BWON-F16C-NEXT: callq __truncdfhf2@PLT
; BWON-F16C-NEXT: movw %ax, 4(%rbx)
; BWON-F16C-NEXT: movw %bp, (%rbx)
; BWON-F16C-NEXT: movw %r15w, 6(%rbx)
; BWON-F16C-NEXT: movw %r14w, 2(%rbx)
; BWON-F16C-NEXT: addq $88, %rsp
; BWON-F16C-NEXT: addq $56, %rsp
; BWON-F16C-NEXT: popq %rbx
; BWON-F16C-NEXT: popq %r14
; BWON-F16C-NEXT: popq %r15

20
test/CodeGen/X86/rotate-extract-vector.ll
View File

@ -56,7 +56,7 @@ define <8 x i32> @vroll_extract_mul(<8 x i32> %i) nounwind {
define <2 x i64> @vrolq_extract_udiv(<2 x i64> %i) nounwind {
; X86-LABEL: vrolq_extract_udiv:
; X86: # %bb.0:
; X86-NEXT: subl $44, %esp
; X86-NEXT: subl $32, %esp
; X86-NEXT: vmovups %xmm0, {{[-0-9]+}}(%e{{[sb]}}p) # 16-byte Spill
; X86-NEXT: vextractps $1, %xmm0, {{[0-9]+}}(%esp)
; X86-NEXT: vmovss %xmm0, (%esp)
@ -77,7 +77,7 @@ define <2 x i64> @vrolq_extract_udiv(<2 x i64> %i) nounwind {
; X86-NEXT: vpinsrd $3, %edx, %xmm0, %xmm0
; X86-NEXT: vprolq $57, %zmm0, %zmm0
; X86-NEXT: # kill: def $xmm0 killed $xmm0 killed $zmm0
; X86-NEXT: addl $44, %esp
; X86-NEXT: addl $32, %esp
; X86-NEXT: vzeroupper
; X86-NEXT: retl
;
@ -109,7 +109,7 @@ define <4 x i32> @vrolw_extract_mul_with_mask(<4 x i32> %i) nounwind {
; X86-NEXT: vpbroadcastd {{.*#+}} xmm1 = [9,9,9,9]
; X86-NEXT: vpmulld %xmm1, %xmm0, %xmm0
; X86-NEXT: vprold $7, %zmm0, %zmm0
; X86-NEXT: vpand {{\.LCPI[0-9]+_[0-9]+}}, %xmm0, %xmm0
; X86-NEXT: vpand {{\.?LCPI[0-9]+_[0-9]+}}, %xmm0, %xmm0
; X86-NEXT: vzeroupper
; X86-NEXT: retl
;
@ -118,7 +118,7 @@ define <4 x i32> @vrolw_extract_mul_with_mask(<4 x i32> %i) nounwind {
; X64-NEXT: vpbroadcastd {{.*#+}} xmm1 = [9,9,9,9]
; X64-NEXT: vpmulld %xmm1, %xmm0, %xmm0
; X64-NEXT: vprold $7, %zmm0, %zmm0
; X64-NEXT: vpand {{.*}}(%rip), %xmm0, %xmm0
; X64-NEXT: vpand {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0, %xmm0
; X64-NEXT: vzeroupper
; X64-NEXT: retq
%lhs_mul = mul <4 x i32> %i, <i32 1152, i32 1152, i32 1152, i32 1152>
@ -132,16 +132,16 @@ define <4 x i32> @vrolw_extract_mul_with_mask(<4 x i32> %i) nounwind {
define <32 x i16> @illegal_no_extract_mul(<32 x i16> %i) nounwind {
; X86-LABEL: illegal_no_extract_mul:
; X86: # %bb.0:
; X86-NEXT: vpmullw {{\.LCPI[0-9]+_[0-9]+}}, %zmm0, %zmm1
; X86-NEXT: vpmullw {{\.LCPI[0-9]+_[0-9]+}}, %zmm0, %zmm0
; X86-NEXT: vpmullw {{\.?LCPI[0-9]+_[0-9]+}}, %zmm0, %zmm1
; X86-NEXT: vpmullw {{\.?LCPI[0-9]+_[0-9]+}}, %zmm0, %zmm0
; X86-NEXT: vpsrlw $10, %zmm0, %zmm0
; X86-NEXT: vporq %zmm0, %zmm1, %zmm0
; X86-NEXT: retl
;
; X64-LABEL: illegal_no_extract_mul:
; X64: # %bb.0:
; X64-NEXT: vpmullw {{.*}}(%rip), %zmm0, %zmm1
; X64-NEXT: vpmullw {{.*}}(%rip), %zmm0, %zmm0
; X64-NEXT: vpmullw {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %zmm0, %zmm1
; X64-NEXT: vpmullw {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %zmm0, %zmm0
; X64-NEXT: vpsrlw $10, %zmm0, %zmm0
; X64-NEXT: vporq %zmm0, %zmm1, %zmm0
; X64-NEXT: retq
@ -207,7 +207,7 @@ define <8 x i32> @no_extract_mul(<8 x i32> %i) nounwind {
define <2 x i64> @no_extract_udiv(<2 x i64> %i) nounwind {
; X86-LABEL: no_extract_udiv:
; X86: # %bb.0:
; X86-NEXT: subl $60, %esp
; X86-NEXT: subl $48, %esp
; X86-NEXT: vmovups %xmm0, {{[-0-9]+}}(%e{{[sb]}}p) # 16-byte Spill
; X86-NEXT: vextractps $1, %xmm0, {{[0-9]+}}(%esp)
; X86-NEXT: vmovss %xmm0, (%esp)
@ -246,7 +246,7 @@ define <2 x i64> @no_extract_udiv(<2 x i64> %i) nounwind {
; X86-NEXT: vmovdqu {{[-0-9]+}}(%e{{[sb]}}p), %xmm1 # 16-byte Reload
; X86-NEXT: vpsllq $56, %xmm1, %xmm1
; X86-NEXT: vpor %xmm0, %xmm1, %xmm0
; X86-NEXT: addl $60, %esp
; X86-NEXT: addl $48, %esp
; X86-NEXT: retl
;
; X64-LABEL: no_extract_udiv:

4
test/CodeGen/X86/sext-vsetcc.ll
View File

@ -158,7 +158,7 @@ define <8 x i16> @cmp_ne_load_const_extra_use2(<8 x i8>* %x) nounwind {
;
; AVX512-LABEL: cmp_ne_load_const_extra_use2:
; AVX512: # %bb.0:
; AVX512-NEXT: subq $120, %rsp
; AVX512-NEXT: subq $72, %rsp
; AVX512-NEXT: vmovq {{.*#+}} xmm0 = mem[0],zero
; AVX512-NEXT: vpxor %xmm1, %xmm1, %xmm1
; AVX512-NEXT: vpcmpeqb %xmm1, %xmm0, %xmm0
@ -168,7 +168,7 @@ define <8 x i16> @cmp_ne_load_const_extra_use2(<8 x i8>* %x) nounwind {
; AVX512-NEXT: vzeroupper
; AVX512-NEXT: callq use_v8i1@PLT
; AVX512-NEXT: vpmovsxbw (%rsp), %xmm0 # 16-byte Folded Reload
; AVX512-NEXT: addq $120, %rsp
; AVX512-NEXT: addq $72, %rsp
; AVX512-NEXT: retq
%loadx = load <8 x i8>, <8 x i8>* %x
%icmp = icmp ne <8 x i8> %loadx, zeroinitializer

54
test/CodeGen/X86/stack-folding-fp-avx512.ll
View File

@ -321,8 +321,8 @@ declare <8 x i1> @llvm.x86.avx512.mask.cmp.pd.512(<8 x double>, <8 x double>, i3
define <8 x double> @stack_fold_cmppd_mask(<8 x double> %a0, <8 x double> %a1, <8 x double>* %a2, i8 %mask, <8 x double> %b0, <8 x double> %b1) {
; CHECK-LABEL: stack_fold_cmppd_mask:
; CHECK: # %bb.0:
; CHECK-NEXT: subq $184, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 192
; CHECK-NEXT: subq $136, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 144
; CHECK-NEXT: vmovups %zmm3, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: vmovups %zmm2, (%rsp) # 64-byte Spill
; CHECK-NEXT: vmovups %zmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
@ -337,7 +337,7 @@ define <8 x double> @stack_fold_cmppd_mask(<8 x double> %a0, <8 x double> %a1, <
; CHECK-NEXT: kandb %k0, %k1, %k1
; CHECK-NEXT: vmovupd {{[-0-9]+}}(%r{{[sb]}}p), %zmm0 # 64-byte Reload
; CHECK-NEXT: vblendmpd (%rsp), %zmm0, %zmm0 {%k1} # 64-byte Folded Reload
; CHECK-NEXT: addq $184, %rsp
; CHECK-NEXT: addq $136, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 8
; CHECK-NEXT: retq
%1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
@ -354,8 +354,8 @@ define <8 x double> @stack_fold_cmppd_mask(<8 x double> %a0, <8 x double> %a1, <
define <8 x double> @stack_fold_cmppd_mask_commuted(<8 x double> %a0, <8 x double> %a1, <8 x double>* %a2, i8 %mask, <8 x double> %b0, <8 x double> %b1) {
; CHECK-LABEL: stack_fold_cmppd_mask_commuted:
; CHECK: # %bb.0:
; CHECK-NEXT: subq $184, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 192
; CHECK-NEXT: subq $136, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 144
; CHECK-NEXT: vmovups %zmm3, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: vmovups %zmm2, (%rsp) # 64-byte Spill
; CHECK-NEXT: vmovups %zmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
@ -370,7 +370,7 @@ define <8 x double> @stack_fold_cmppd_mask_commuted(<8 x double> %a0, <8 x doubl
; CHECK-NEXT: kandb %k0, %k1, %k1
; CHECK-NEXT: vmovupd {{[-0-9]+}}(%r{{[sb]}}p), %zmm0 # 64-byte Reload
; CHECK-NEXT: vblendmpd (%rsp), %zmm0, %zmm0 {%k1} # 64-byte Folded Reload
; CHECK-NEXT: addq $184, %rsp
; CHECK-NEXT: addq $136, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 8
; CHECK-NEXT: retq
%1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
@ -406,8 +406,8 @@ declare <16 x i1> @llvm.x86.avx512.mask.cmp.ps.512(<16 x float>, <16 x float>, i
define <16 x float> @stack_fold_cmpps_mask(<16 x float> %a0, <16 x float> %a1, <16 x float>* %a2, i16 %mask, <16 x float> %b0, <16 x float> %b1) {
; CHECK-LABEL: stack_fold_cmpps_mask:
; CHECK: # %bb.0:
; CHECK-NEXT: subq $184, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 192
; CHECK-NEXT: subq $136, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 144
; CHECK-NEXT: vmovups %zmm3, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: vmovups %zmm2, (%rsp) # 64-byte Spill
; CHECK-NEXT: vmovups %zmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
@ -422,7 +422,7 @@ define <16 x float> @stack_fold_cmpps_mask(<16 x float> %a0, <16 x float> %a1, <
; CHECK-NEXT: kandw %k0, %k1, %k1
; CHECK-NEXT: vmovups {{[-0-9]+}}(%r{{[sb]}}p), %zmm0 # 64-byte Reload
; CHECK-NEXT: vblendmps (%rsp), %zmm0, %zmm0 {%k1} # 64-byte Folded Reload
; CHECK-NEXT: addq $184, %rsp
; CHECK-NEXT: addq $136, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 8
; CHECK-NEXT: retq
%1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
@ -439,8 +439,8 @@ define <16 x float> @stack_fold_cmpps_mask(<16 x float> %a0, <16 x float> %a1, <
define <16 x float> @stack_fold_cmpps_mask_commuted(<16 x float> %a0, <16 x float> %a1, <16 x float>* %a2, i16 %mask, <16 x float> %b0, <16 x float> %b1) {
; CHECK-LABEL: stack_fold_cmpps_mask_commuted:
; CHECK: # %bb.0:
; CHECK-NEXT: subq $184, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 192
; CHECK-NEXT: subq $136, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 144
; CHECK-NEXT: vmovups %zmm3, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: vmovups %zmm2, (%rsp) # 64-byte Spill
; CHECK-NEXT: vmovups %zmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
@ -455,7 +455,7 @@ define <16 x float> @stack_fold_cmpps_mask_commuted(<16 x float> %a0, <16 x floa
; CHECK-NEXT: kandw %k0, %k1, %k1
; CHECK-NEXT: vmovups {{[-0-9]+}}(%r{{[sb]}}p), %zmm0 # 64-byte Reload
; CHECK-NEXT: vblendmps (%rsp), %zmm0, %zmm0 {%k1} # 64-byte Folded Reload
; CHECK-NEXT: addq $184, %rsp
; CHECK-NEXT: addq $136, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 8
; CHECK-NEXT: retq
%1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
@ -1198,8 +1198,8 @@ define <16 x float> @stack_fold_orps_zmm(<16 x float> %a0, <16 x float> %a1) #0
define <8 x double> @stack_fold_shuff64x2(<8 x double> %a, <8 x double> %b) {
; CHECK-LABEL: stack_fold_shuff64x2:
; CHECK: # %bb.0:
; CHECK-NEXT: subq $56, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 64
; CHECK-NEXT: pushq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: vmovups %zmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: vmovups %zmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: #APP
@ -1208,7 +1208,7 @@ define <8 x double> @stack_fold_shuff64x2(<8 x double> %a, <8 x double> %b) {
; CHECK-NEXT: vmovupd {{[-0-9]+}}(%r{{[sb]}}p), %zmm0 # 64-byte Reload
; CHECK-NEXT: vshuff64x2 $24, {{[-0-9]+}}(%r{{[sb]}}p), %zmm0, %zmm0 # 64-byte Folded Reload
; CHECK-NEXT: # zmm0 = zmm0[0,1,4,5],mem[2,3,0,1]
; CHECK-NEXT: addq $56, %rsp
; CHECK-NEXT: popq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 8
; CHECK-NEXT: retq
%1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
@ -1219,8 +1219,8 @@ define <8 x double> @stack_fold_shuff64x2(<8 x double> %a, <8 x double> %b) {
define <8 x double> @stack_fold_shuff64x2_mask(<8 x double> %a, <8 x double> %b, i8 %mask, <8 x double>* %passthru) {
; CHECK-LABEL: stack_fold_shuff64x2_mask:
; CHECK: # %bb.0:
; CHECK-NEXT: subq $56, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 64
; CHECK-NEXT: pushq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: vmovups %zmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: vmovups %zmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: #APP
@ -1232,7 +1232,7 @@ define <8 x double> @stack_fold_shuff64x2_mask(<8 x double> %a, <8 x double> %b,
; CHECK-NEXT: vshuff64x2 $24, {{[-0-9]+}}(%r{{[sb]}}p), %zmm0, %zmm1 {%k1} # 64-byte Folded Reload
; CHECK-NEXT: # zmm1 {%k1} = zmm0[0,1,4,5],mem[2,3,0,1]
; CHECK-NEXT: vmovapd %zmm1, %zmm0
; CHECK-NEXT: addq $56, %rsp
; CHECK-NEXT: popq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 8
; CHECK-NEXT: retq
%1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
@ -1247,8 +1247,8 @@ define <8 x double> @stack_fold_shuff64x2_mask(<8 x double> %a, <8 x double> %b,
define <8 x double> @stack_fold_shuff64x2_maskz(<8 x double> %a, <8 x double> %b, i8 %mask, <8 x double>* %passthru) {
; CHECK-LABEL: stack_fold_shuff64x2_maskz:
; CHECK: # %bb.0:
; CHECK-NEXT: subq $56, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 64
; CHECK-NEXT: pushq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: vmovups %zmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: vmovups %zmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: #APP
@ -1258,7 +1258,7 @@ define <8 x double> @stack_fold_shuff64x2_maskz(<8 x double> %a, <8 x double> %b
; CHECK-NEXT: vmovupd {{[-0-9]+}}(%r{{[sb]}}p), %zmm0 # 64-byte Reload
; CHECK-NEXT: vshuff64x2 $24, {{[-0-9]+}}(%r{{[sb]}}p), %zmm0, %zmm0 {%k1} {z} # 64-byte Folded Reload
; CHECK-NEXT: # zmm0 {%k1} {z} = zmm0[0,1,4,5],mem[2,3,0,1]
; CHECK-NEXT: addq $56, %rsp
; CHECK-NEXT: popq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 8
; CHECK-NEXT: retq
%1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
@ -1271,8 +1271,8 @@ define <8 x double> @stack_fold_shuff64x2_maskz(<8 x double> %a, <8 x double> %b
define <16 x float> @stack_fold_shuff32x4_mask(<16 x float> %a, <16 x float> %b, i16 %mask, <16 x float>* %passthru) {
; CHECK-LABEL: stack_fold_shuff32x4_mask:
; CHECK: # %bb.0:
; CHECK-NEXT: subq $56, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 64
; CHECK-NEXT: pushq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: vmovups %zmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: vmovups %zmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: #APP
@ -1284,7 +1284,7 @@ define <16 x float> @stack_fold_shuff32x4_mask(<16 x float> %a, <16 x float> %b,
; CHECK-NEXT: vshuff32x4 $20, {{[-0-9]+}}(%r{{[sb]}}p), %zmm0, %zmm1 {%k1} # 64-byte Folded Reload
; CHECK-NEXT: # zmm1 {%k1} = zmm0[0,1,2,3,4,5,6,7],mem[4,5,6,7,0,1,2,3]
; CHECK-NEXT: vmovaps %zmm1, %zmm0
; CHECK-NEXT: addq $56, %rsp
; CHECK-NEXT: popq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 8
; CHECK-NEXT: retq
%1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
@ -1299,8 +1299,8 @@ define <16 x float> @stack_fold_shuff32x4_mask(<16 x float> %a, <16 x float> %b,
define <16 x float> @stack_fold_shuff32x4_maskz(<16 x float> %a, <16 x float> %b, i16 %mask) {
; CHECK-LABEL: stack_fold_shuff32x4_maskz:
; CHECK: # %bb.0:
; CHECK-NEXT: subq $56, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 64
; CHECK-NEXT: pushq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: vmovups %zmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: vmovups %zmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: #APP
@ -1310,7 +1310,7 @@ define <16 x float> @stack_fold_shuff32x4_maskz(<16 x float> %a, <16 x float> %b
; CHECK-NEXT: vmovups {{[-0-9]+}}(%r{{[sb]}}p), %zmm0 # 64-byte Reload
; CHECK-NEXT: vshuff32x4 $20, {{[-0-9]+}}(%r{{[sb]}}p), %zmm0, %zmm0 {%k1} {z} # 64-byte Folded Reload
; CHECK-NEXT: # zmm0 {%k1} {z} = zmm0[0,1,2,3,4,5,6,7],mem[4,5,6,7,0,1,2,3]
; CHECK-NEXT: addq $56, %rsp
; CHECK-NEXT: popq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 8
; CHECK-NEXT: retq
%1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()

150
test/CodeGen/X86/stack-folding-int-avx512.ll
View File

@ -12,8 +12,8 @@ target triple = "x86_64-unknown-unknown"
define <16 x i32> @stack_fold_valignd(<16 x i32> %a, <16 x i32> %b) {
; CHECK-LABEL: stack_fold_valignd:
; CHECK: # %bb.0:
; CHECK-NEXT: subq $56, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 64
; CHECK-NEXT: pushq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: vmovups %zmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: vmovups %zmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: #APP
@ -22,7 +22,7 @@ define <16 x i32> @stack_fold_valignd(<16 x i32> %a, <16 x i32> %b) {
; CHECK-NEXT: vmovdqu64 {{[-0-9]+}}(%r{{[sb]}}p), %zmm0 # 64-byte Reload
; CHECK-NEXT: valignd $1, {{[-0-9]+}}(%r{{[sb]}}p), %zmm0, %zmm0 # 64-byte Folded Reload
; CHECK-NEXT: # zmm0 = mem[1,2,3,4,5,6,7,8,9,10,11,12,13,14,15],zmm0[0]
; CHECK-NEXT: addq $56, %rsp
; CHECK-NEXT: popq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 8
; CHECK-NEXT: retq
%1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
@ -33,8 +33,8 @@ define <16 x i32> @stack_fold_valignd(<16 x i32> %a, <16 x i32> %b) {
define <16 x i32> @stack_fold_valignd_mask(<16 x i32> %a, <16 x i32> %b, <16 x i32>* %passthru, i16 %mask) {
; CHECK-LABEL: stack_fold_valignd_mask:
; CHECK: # %bb.0:
; CHECK-NEXT: subq $56, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 64
; CHECK-NEXT: pushq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: vmovups %zmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: vmovups %zmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: #APP
@ -46,7 +46,7 @@ define <16 x i32> @stack_fold_valignd_mask(<16 x i32> %a, <16 x i32> %b, <16 x i
; CHECK-NEXT: valignd $1, {{[-0-9]+}}(%r{{[sb]}}p), %zmm0, %zmm1 {%k1} # 64-byte Folded Reload
; CHECK-NEXT: # zmm1 {%k1} = mem[1,2,3,4,5,6,7,8,9,10,11,12,13,14,15],zmm0[0]
; CHECK-NEXT: vmovdqa64 %zmm1, %zmm0
; CHECK-NEXT: addq $56, %rsp
; CHECK-NEXT: popq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 8
; CHECK-NEXT: retq
%1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
@ -60,8 +60,8 @@ define <16 x i32> @stack_fold_valignd_mask(<16 x i32> %a, <16 x i32> %b, <16 x i
define <16 x i32> @stack_fold_valignd_maskz(<16 x i32> %a, <16 x i32> %b, i16 %mask) {
; CHECK-LABEL: stack_fold_valignd_maskz:
; CHECK: # %bb.0:
; CHECK-NEXT: subq $56, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 64
; CHECK-NEXT: pushq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: vmovups %zmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: vmovups %zmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: #APP
@ -71,7 +71,7 @@ define <16 x i32> @stack_fold_valignd_maskz(<16 x i32> %a, <16 x i32> %b, i16 %m
; CHECK-NEXT: vmovdqu64 {{[-0-9]+}}(%r{{[sb]}}p), %zmm0 # 64-byte Reload
; CHECK-NEXT: valignd $1, {{[-0-9]+}}(%r{{[sb]}}p), %zmm0, %zmm0 {%k1} {z} # 64-byte Folded Reload
; CHECK-NEXT: # zmm0 {%k1} {z} = mem[1,2,3,4,5,6,7,8,9,10,11,12,13,14,15],zmm0[0]
; CHECK-NEXT: addq $56, %rsp
; CHECK-NEXT: popq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 8
; CHECK-NEXT: retq
%1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
@ -84,8 +84,8 @@ define <16 x i32> @stack_fold_valignd_maskz(<16 x i32> %a, <16 x i32> %b, i16 %m
define <8 x i64> @stack_fold_valignq(<8 x i64> %a, <8 x i64> %b) {
; CHECK-LABEL: stack_fold_valignq:
; CHECK: # %bb.0:
; CHECK-NEXT: subq $56, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 64
; CHECK-NEXT: pushq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: vmovups %zmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: vmovups %zmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: #APP
@ -94,7 +94,7 @@ define <8 x i64> @stack_fold_valignq(<8 x i64> %a, <8 x i64> %b) {
; CHECK-NEXT: vmovdqu64 {{[-0-9]+}}(%r{{[sb]}}p), %zmm0 # 64-byte Reload
; CHECK-NEXT: valignq $1, {{[-0-9]+}}(%r{{[sb]}}p), %zmm0, %zmm0 # 64-byte Folded Reload
; CHECK-NEXT: # zmm0 = mem[1,2,3,4,5,6,7],zmm0[0]
; CHECK-NEXT: addq $56, %rsp
; CHECK-NEXT: popq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 8
; CHECK-NEXT: retq
%1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
@ -105,8 +105,8 @@ define <8 x i64> @stack_fold_valignq(<8 x i64> %a, <8 x i64> %b) {
define <8 x i64> @stack_fold_valignq_mask(<8 x i64> %a, <8 x i64> %b, <8 x i64>* %passthru, i8 %mask) {
; CHECK-LABEL: stack_fold_valignq_mask:
; CHECK: # %bb.0:
; CHECK-NEXT: subq $56, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 64
; CHECK-NEXT: pushq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: vmovups %zmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: vmovups %zmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: #APP
@ -118,7 +118,7 @@ define <8 x i64> @stack_fold_valignq_mask(<8 x i64> %a, <8 x i64> %b, <8 x i64>*
; CHECK-NEXT: valignq $1, {{[-0-9]+}}(%r{{[sb]}}p), %zmm0, %zmm1 {%k1} # 64-byte Folded Reload
; CHECK-NEXT: # zmm1 {%k1} = mem[1,2,3,4,5,6,7],zmm0[0]
; CHECK-NEXT: vmovdqa64 %zmm1, %zmm0
; CHECK-NEXT: addq $56, %rsp
; CHECK-NEXT: popq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 8
; CHECK-NEXT: retq
%1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
@ -132,8 +132,8 @@ define <8 x i64> @stack_fold_valignq_mask(<8 x i64> %a, <8 x i64> %b, <8 x i64>*
define <8 x i64> @stack_fold_valignq_maskz(<8 x i64> %a, <8 x i64> %b, i8 %mask) {
; CHECK-LABEL: stack_fold_valignq_maskz:
; CHECK: # %bb.0:
; CHECK-NEXT: subq $56, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 64
; CHECK-NEXT: pushq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: vmovups %zmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: vmovups %zmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: #APP
@ -143,7 +143,7 @@ define <8 x i64> @stack_fold_valignq_maskz(<8 x i64> %a, <8 x i64> %b, i8 %mask)
; CHECK-NEXT: vmovdqu64 {{[-0-9]+}}(%r{{[sb]}}p), %zmm0 # 64-byte Reload
; CHECK-NEXT: valignq $1, {{[-0-9]+}}(%r{{[sb]}}p), %zmm0, %zmm0 {%k1} {z} # 64-byte Folded Reload
; CHECK-NEXT: # zmm0 {%k1} {z} = mem[1,2,3,4,5,6,7],zmm0[0]
; CHECK-NEXT: addq $56, %rsp
; CHECK-NEXT: popq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 8
; CHECK-NEXT: retq
%1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
@ -490,8 +490,8 @@ define <64 x i8> @stack_fold_pabsb(<64 x i8> %a0) {
define <64 x i8> @stack_fold_pabsb_mask(<64 x i8> %passthru, <64 x i8> %a0, i64 %mask) {
; CHECK-LABEL: stack_fold_pabsb_mask:
; CHECK: # %bb.0:
; CHECK-NEXT: subq $56, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 64
; CHECK-NEXT: pushq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: vmovups %zmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: vmovups %zmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: #APP
@ -500,7 +500,7 @@ define <64 x i8> @stack_fold_pabsb_mask(<64 x i8> %passthru, <64 x i8> %a0, i64
; CHECK-NEXT: kmovq %rdi, %k1
; CHECK-NEXT: vmovdqu64 {{[-0-9]+}}(%r{{[sb]}}p), %zmm0 # 64-byte Reload
; CHECK-NEXT: vpabsb {{[-0-9]+}}(%r{{[sb]}}p), %zmm0 {%k1} # 64-byte Folded Reload
; CHECK-NEXT: addq $56, %rsp
; CHECK-NEXT: popq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 8
; CHECK-NEXT: retq
%1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
@ -550,8 +550,8 @@ define <16 x i32> @stack_fold_pabsd(<16 x i32> %a0) {
define <16 x i32> @stack_fold_pabsd_mask(<16 x i32> %passthru, <16 x i32> %a0, i16 %mask) {
; CHECK-LABEL: stack_fold_pabsd_mask:
; CHECK: # %bb.0:
; CHECK-NEXT: subq $56, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 64
; CHECK-NEXT: pushq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: vmovups %zmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: vmovups %zmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: #APP
@ -560,7 +560,7 @@ define <16 x i32> @stack_fold_pabsd_mask(<16 x i32> %passthru, <16 x i32> %a0, i
; CHECK-NEXT: kmovd %edi, %k1
; CHECK-NEXT: vmovdqu64 {{[-0-9]+}}(%r{{[sb]}}p), %zmm0 # 64-byte Reload
; CHECK-NEXT: vpabsd {{[-0-9]+}}(%r{{[sb]}}p), %zmm0 {%k1} # 64-byte Folded Reload
; CHECK-NEXT: addq $56, %rsp
; CHECK-NEXT: popq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 8
; CHECK-NEXT: retq
%1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
@ -610,8 +610,8 @@ define <8 x i64> @stack_fold_pabsq(<8 x i64> %a0) {
define <8 x i64> @stack_fold_pabsq_mask(<8 x i64> %passthru, <8 x i64> %a0, i8 %mask) {
; CHECK-LABEL: stack_fold_pabsq_mask:
; CHECK: # %bb.0:
; CHECK-NEXT: subq $56, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 64
; CHECK-NEXT: pushq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: vmovups %zmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: vmovups %zmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: #APP
@ -620,7 +620,7 @@ define <8 x i64> @stack_fold_pabsq_mask(<8 x i64> %passthru, <8 x i64> %a0, i8 %
; CHECK-NEXT: kmovd %edi, %k1
; CHECK-NEXT: vmovdqu64 {{[-0-9]+}}(%r{{[sb]}}p), %zmm0 # 64-byte Reload
; CHECK-NEXT: vpabsq {{[-0-9]+}}(%r{{[sb]}}p), %zmm0 {%k1} # 64-byte Folded Reload
; CHECK-NEXT: addq $56, %rsp
; CHECK-NEXT: popq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 8
; CHECK-NEXT: retq
%1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
@ -670,8 +670,8 @@ define <32 x i16> @stack_fold_pabsw(<32 x i16> %a0) {
define <32 x i16> @stack_fold_pabsw_mask(<32 x i16> %passthru, <32 x i16> %a0, i32 %mask) {
; CHECK-LABEL: stack_fold_pabsw_mask:
; CHECK: # %bb.0:
; CHECK-NEXT: subq $56, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 64
; CHECK-NEXT: pushq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: vmovups %zmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: vmovups %zmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: #APP
@ -680,7 +680,7 @@ define <32 x i16> @stack_fold_pabsw_mask(<32 x i16> %passthru, <32 x i16> %a0, i
; CHECK-NEXT: kmovd %edi, %k1
; CHECK-NEXT: vmovdqu64 {{[-0-9]+}}(%r{{[sb]}}p), %zmm0 # 64-byte Reload
; CHECK-NEXT: vpabsw {{[-0-9]+}}(%r{{[sb]}}p), %zmm0 {%k1} # 64-byte Folded Reload
; CHECK-NEXT: addq $56, %rsp
; CHECK-NEXT: popq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 8
; CHECK-NEXT: retq
%1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
@ -1643,8 +1643,8 @@ define <32 x i16> @stack_fold_paddw_maskz_commuted(<32 x i16> %a0, <32 x i16> %a
define <64 x i8> @stack_fold_palignr(<64 x i8> %a0, <64 x i8> %a1) {
; CHECK-LABEL: stack_fold_palignr:
; CHECK: # %bb.0:
; CHECK-NEXT: subq $56, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 64
; CHECK-NEXT: pushq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: vmovups %zmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: vmovups %zmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: #APP
@ -1653,7 +1653,7 @@ define <64 x i8> @stack_fold_palignr(<64 x i8> %a0, <64 x i8> %a1) {
; CHECK-NEXT: vmovdqu64 {{[-0-9]+}}(%r{{[sb]}}p), %zmm0 # 64-byte Reload
; CHECK-NEXT: vpalignr $1, {{[-0-9]+}}(%r{{[sb]}}p), %zmm0, %zmm0 # 64-byte Folded Reload
; CHECK-NEXT: # zmm0 = mem[1,2,3,4,5,6,7,8,9,10,11,12,13,14,15],zmm0[0],mem[17,18,19,20,21,22,23,24,25,26,27,28,29,30,31],zmm0[16],mem[33,34,35,36,37,38,39,40,41,42,43,44,45,46,47],zmm0[32],mem[49,50,51,52,53,54,55,56,57,58,59,60,61,62,63],zmm0[48]
; CHECK-NEXT: addq $56, %rsp
; CHECK-NEXT: popq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 8
; CHECK-NEXT: retq
%1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
@ -1664,8 +1664,8 @@ define <64 x i8> @stack_fold_palignr(<64 x i8> %a0, <64 x i8> %a1) {
define <64 x i8> @stack_fold_palignr_mask(<64 x i8> %a0, <64 x i8> %a1, <64 x i8>* %passthru, i64 %mask) {
; CHECK-LABEL: stack_fold_palignr_mask:
; CHECK: # %bb.0:
; CHECK-NEXT: subq $56, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 64
; CHECK-NEXT: pushq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: vmovups %zmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: vmovups %zmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: #APP
@ -1677,7 +1677,7 @@ define <64 x i8> @stack_fold_palignr_mask(<64 x i8> %a0, <64 x i8> %a1, <64 x i8
; CHECK-NEXT: vpalignr $1, {{[-0-9]+}}(%r{{[sb]}}p), %zmm0, %zmm1 {%k1} # 64-byte Folded Reload
; CHECK-NEXT: # zmm1 {%k1} = mem[1,2,3,4,5,6,7,8,9,10,11,12,13,14,15],zmm0[0],mem[17,18,19,20,21,22,23,24,25,26,27,28,29,30,31],zmm0[16],mem[33,34,35,36,37,38,39,40,41,42,43,44,45,46,47],zmm0[32],mem[49,50,51,52,53,54,55,56,57,58,59,60,61,62,63],zmm0[48]
; CHECK-NEXT: vmovdqa64 %zmm1, %zmm0
; CHECK-NEXT: addq $56, %rsp
; CHECK-NEXT: popq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 8
; CHECK-NEXT: retq
%1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
@ -1691,8 +1691,8 @@ define <64 x i8> @stack_fold_palignr_mask(<64 x i8> %a0, <64 x i8> %a1, <64 x i8
define <64 x i8> @stack_fold_palignr_maskz(<64 x i8> %a0, <64 x i8> %a1, i64 %mask) {
; CHECK-LABEL: stack_fold_palignr_maskz:
; CHECK: # %bb.0:
; CHECK-NEXT: subq $56, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 64
; CHECK-NEXT: pushq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: vmovups %zmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: vmovups %zmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: #APP
@ -1702,7 +1702,7 @@ define <64 x i8> @stack_fold_palignr_maskz(<64 x i8> %a0, <64 x i8> %a1, i64 %ma
; CHECK-NEXT: vmovdqu64 {{[-0-9]+}}(%r{{[sb]}}p), %zmm0 # 64-byte Reload
; CHECK-NEXT: vpalignr $1, {{[-0-9]+}}(%r{{[sb]}}p), %zmm0, %zmm0 {%k1} {z} # 64-byte Folded Reload
; CHECK-NEXT: # zmm0 {%k1} {z} = mem[1,2,3,4,5,6,7,8,9,10,11,12,13,14,15],zmm0[0],mem[17,18,19,20,21,22,23,24,25,26,27,28,29,30,31],zmm0[16],mem[33,34,35,36,37,38,39,40,41,42,43,44,45,46,47],zmm0[32],mem[49,50,51,52,53,54,55,56,57,58,59,60,61,62,63],zmm0[48]
; CHECK-NEXT: addq $56, %rsp
; CHECK-NEXT: popq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 8
; CHECK-NEXT: retq
%1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
@ -2023,8 +2023,8 @@ define i32 @stack_fold_pcmpeqw(<32 x i16> %a0, <32 x i16> %a1) {
define <16 x i32> @stack_fold_pcmpeqd_mask(<16 x i32> %a0, <16 x i32> %a1, <16 x i32>* %a2, i16 %mask, <16 x i32> %b0, <16 x i32> %b1) {
; CHECK-LABEL: stack_fold_pcmpeqd_mask:
; CHECK: # %bb.0:
; CHECK-NEXT: subq $184, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 192
; CHECK-NEXT: subq $136, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 144
; CHECK-NEXT: vmovups %zmm3, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: vmovups %zmm2, (%rsp) # 64-byte Spill
; CHECK-NEXT: vmovups %zmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
@ -2038,7 +2038,7 @@ define <16 x i32> @stack_fold_pcmpeqd_mask(<16 x i32> %a0, <16 x i32> %a1, <16 x
; CHECK-NEXT: vpcmpeqd {{[-0-9]+}}(%r{{[sb]}}p), %zmm0, %k1 {%k1} # 64-byte Folded Reload
; CHECK-NEXT: vmovdqu64 {{[-0-9]+}}(%r{{[sb]}}p), %zmm0 # 64-byte Reload
; CHECK-NEXT: vpblendmd (%rsp), %zmm0, %zmm0 {%k1} # 64-byte Folded Reload
; CHECK-NEXT: addq $184, %rsp
; CHECK-NEXT: addq $136, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 8
; CHECK-NEXT: retq
%1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
@ -2055,8 +2055,8 @@ define <16 x i32> @stack_fold_pcmpeqd_mask(<16 x i32> %a0, <16 x i32> %a1, <16 x
define <16 x i32> @stack_fold_pcmpeqd_mask_commuted(<16 x i32> %a0, <16 x i32> %a1, <16 x i32>* %a2, i16 %mask, <16 x i32> %b0, <16 x i32> %b1) {
; CHECK-LABEL: stack_fold_pcmpeqd_mask_commuted:
; CHECK: # %bb.0:
; CHECK-NEXT: subq $184, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 192
; CHECK-NEXT: subq $136, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 144
; CHECK-NEXT: vmovups %zmm3, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: vmovups %zmm2, (%rsp) # 64-byte Spill
; CHECK-NEXT: vmovups %zmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
@ -2070,7 +2070,7 @@ define <16 x i32> @stack_fold_pcmpeqd_mask_commuted(<16 x i32> %a0, <16 x i32> %
; CHECK-NEXT: vpcmpeqd {{[-0-9]+}}(%r{{[sb]}}p), %zmm0, %k1 {%k1} # 64-byte Folded Reload
; CHECK-NEXT: vmovdqu64 {{[-0-9]+}}(%r{{[sb]}}p), %zmm0 # 64-byte Reload
; CHECK-NEXT: vpblendmd (%rsp), %zmm0, %zmm0 {%k1} # 64-byte Folded Reload
; CHECK-NEXT: addq $184, %rsp
; CHECK-NEXT: addq $136, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 8
; CHECK-NEXT: retq
%1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
@ -2087,8 +2087,8 @@ define <16 x i32> @stack_fold_pcmpeqd_mask_commuted(<16 x i32> %a0, <16 x i32> %
define <16 x i32> @stack_fold_pcmpled_mask(<16 x i32> %a0, <16 x i32> %a1, <16 x i32>* %a2, i16 %mask, <16 x i32> %b0, <16 x i32> %b1) {
; CHECK-LABEL: stack_fold_pcmpled_mask:
; CHECK: # %bb.0:
; CHECK-NEXT: subq $184, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 192
; CHECK-NEXT: subq $136, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 144
; CHECK-NEXT: vmovups %zmm3, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: vmovups %zmm2, (%rsp) # 64-byte Spill
; CHECK-NEXT: vmovups %zmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
@ -2102,7 +2102,7 @@ define <16 x i32> @stack_fold_pcmpled_mask(<16 x i32> %a0, <16 x i32> %a1, <16 x
; CHECK-NEXT: vpcmpled {{[-0-9]+}}(%r{{[sb]}}p), %zmm0, %k1 {%k1} # 64-byte Folded Reload
; CHECK-NEXT: vmovdqu64 {{[-0-9]+}}(%r{{[sb]}}p), %zmm0 # 64-byte Reload
; CHECK-NEXT: vpblendmd (%rsp), %zmm0, %zmm0 {%k1} # 64-byte Folded Reload
; CHECK-NEXT: addq $184, %rsp
; CHECK-NEXT: addq $136, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 8
; CHECK-NEXT: retq
%1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
@ -2119,8 +2119,8 @@ define <16 x i32> @stack_fold_pcmpled_mask(<16 x i32> %a0, <16 x i32> %a1, <16 x
define i16 @stack_fold_pcmpleud(<16 x i32> %a0, <16 x i32> %a1, <16 x i32>* %a2, i16 %mask) {
; CHECK-LABEL: stack_fold_pcmpleud:
; CHECK: # %bb.0:
; CHECK-NEXT: subq $56, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 64
; CHECK-NEXT: pushq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: vmovups %zmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: vmovups %zmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: #APP
@ -2132,7 +2132,7 @@ define i16 @stack_fold_pcmpleud(<16 x i32> %a0, <16 x i32> %a1, <16 x i32>* %a2,
; CHECK-NEXT: kmovd %k0, %eax
; CHECK-NEXT: andl %esi, %eax
; CHECK-NEXT: # kill: def $ax killed $ax killed $eax
; CHECK-NEXT: addq $56, %rsp
; CHECK-NEXT: popq %rcx
; CHECK-NEXT: .cfi_def_cfa_offset 8
; CHECK-NEXT: vzeroupper
; CHECK-NEXT: retq
@ -6100,8 +6100,8 @@ define <16 x i32> @stack_fold_pshufd_zmm(<16 x i32> %a0) {
define <16 x i32> @stack_fold_pshufd_zmm_mask(<16 x i32> %passthru, <16 x i32> %a0, i16 %mask) {
; CHECK-LABEL: stack_fold_pshufd_zmm_mask:
; CHECK: # %bb.0:
; CHECK-NEXT: subq $56, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 64
; CHECK-NEXT: pushq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: vmovups %zmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: vmovups %zmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: #APP
@ -6111,7 +6111,7 @@ define <16 x i32> @stack_fold_pshufd_zmm_mask(<16 x i32> %passthru, <16 x i32> %
; CHECK-NEXT: vmovdqu64 {{[-0-9]+}}(%r{{[sb]}}p), %zmm0 # 64-byte Reload
; CHECK-NEXT: vpshufd $27, {{[-0-9]+}}(%r{{[sb]}}p), %zmm0 {%k1} # 64-byte Folded Reload
; CHECK-NEXT: # zmm0 {%k1} = mem[3,2,1,0,7,6,5,4,11,10,9,8,15,14,13,12]
; CHECK-NEXT: addq $56, %rsp
; CHECK-NEXT: popq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 8
; CHECK-NEXT: retq
%1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
@ -6157,8 +6157,8 @@ define <32 x i16> @stack_fold_pshufhw_zmm(<32 x i16> %a0) {
define <32 x i16> @stack_fold_pshufhw_zmm_mask(<32 x i16> %passthru, <32 x i16> %a0, i32 %mask) {
; CHECK-LABEL: stack_fold_pshufhw_zmm_mask:
; CHECK: # %bb.0:
; CHECK-NEXT: subq $56, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 64
; CHECK-NEXT: pushq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: vmovups %zmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: vmovups %zmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: #APP
@ -6168,7 +6168,7 @@ define <32 x i16> @stack_fold_pshufhw_zmm_mask(<32 x i16> %passthru, <32 x i16>
; CHECK-NEXT: vmovdqu64 {{[-0-9]+}}(%r{{[sb]}}p), %zmm0 # 64-byte Reload
; CHECK-NEXT: vpshufhw $27, {{[-0-9]+}}(%r{{[sb]}}p), %zmm0 {%k1} # 64-byte Folded Reload
; CHECK-NEXT: # zmm0 {%k1} = mem[0,1,2,3,7,6,5,4,8,9,10,11,15,14,13,12,16,17,18,19,23,22,21,20,24,25,26,27,31,30,29,28]
; CHECK-NEXT: addq $56, %rsp
; CHECK-NEXT: popq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 8
; CHECK-NEXT: retq
%1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
@ -6214,8 +6214,8 @@ define <32 x i16> @stack_fold_pshuflw_zmm(<32 x i16> %a0) {
define <32 x i16> @stack_fold_pshuflw_zmm_mask(<32 x i16> %passthru, <32 x i16> %a0, i32 %mask) {
; CHECK-LABEL: stack_fold_pshuflw_zmm_mask:
; CHECK: # %bb.0:
; CHECK-NEXT: subq $56, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 64
; CHECK-NEXT: pushq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: vmovups %zmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: vmovups %zmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: #APP
@ -6225,7 +6225,7 @@ define <32 x i16> @stack_fold_pshuflw_zmm_mask(<32 x i16> %passthru, <32 x i16>
; CHECK-NEXT: vmovdqu64 {{[-0-9]+}}(%r{{[sb]}}p), %zmm0 # 64-byte Reload
; CHECK-NEXT: vpshuflw $27, {{[-0-9]+}}(%r{{[sb]}}p), %zmm0 {%k1} # 64-byte Folded Reload
; CHECK-NEXT: # zmm0 {%k1} = mem[3,2,1,0,4,5,6,7,11,10,9,8,12,13,14,15,19,18,17,16,20,21,22,23,27,26,25,24,28,29,30,31]
; CHECK-NEXT: addq $56, %rsp
; CHECK-NEXT: popq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 8
; CHECK-NEXT: retq
%1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
@ -6914,8 +6914,8 @@ define <32 x i16> @stack_fold_psubw(<32 x i16> %a0, <32 x i16> %a1) {
define <8 x i64> @stack_fold_shufi64x2(<8 x i64> %a, <8 x i64> %b) {
; CHECK-LABEL: stack_fold_shufi64x2:
; CHECK: # %bb.0:
; CHECK-NEXT: subq $56, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 64
; CHECK-NEXT: pushq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: vmovups %zmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: vmovups %zmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: #APP
@ -6924,7 +6924,7 @@ define <8 x i64> @stack_fold_shufi64x2(<8 x i64> %a, <8 x i64> %b) {
; CHECK-NEXT: vmovdqu64 {{[-0-9]+}}(%r{{[sb]}}p), %zmm0 # 64-byte Reload
; CHECK-NEXT: vshufi64x2 $24, {{[-0-9]+}}(%r{{[sb]}}p), %zmm0, %zmm0 # 64-byte Folded Reload
; CHECK-NEXT: # zmm0 = zmm0[0,1,4,5],mem[2,3,0,1]
; CHECK-NEXT: addq $56, %rsp
; CHECK-NEXT: popq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 8
; CHECK-NEXT: retq
%1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
@ -6935,8 +6935,8 @@ define <8 x i64> @stack_fold_shufi64x2(<8 x i64> %a, <8 x i64> %b) {
define <8 x i64> @stack_fold_shufi64x2_mask(<8 x i64> %a, <8 x i64> %b, i8 %mask, <8 x i64>* %passthru) {
; CHECK-LABEL: stack_fold_shufi64x2_mask:
; CHECK: # %bb.0:
; CHECK-NEXT: subq $56, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 64
; CHECK-NEXT: pushq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: vmovups %zmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: vmovups %zmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: #APP
@ -6948,7 +6948,7 @@ define <8 x i64> @stack_fold_shufi64x2_mask(<8 x i64> %a, <8 x i64> %b, i8 %mask
; CHECK-NEXT: vshufi64x2 $24, {{[-0-9]+}}(%r{{[sb]}}p), %zmm0, %zmm1 {%k1} # 64-byte Folded Reload
; CHECK-NEXT: # zmm1 {%k1} = zmm0[0,1,4,5],mem[2,3,0,1]
; CHECK-NEXT: vmovdqa64 %zmm1, %zmm0
; CHECK-NEXT: addq $56, %rsp
; CHECK-NEXT: popq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 8
; CHECK-NEXT: retq
%1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
@ -6963,8 +6963,8 @@ define <8 x i64> @stack_fold_shufi64x2_mask(<8 x i64> %a, <8 x i64> %b, i8 %mask
define <8 x i64> @stack_fold_shufi64x2_maskz(<8 x i64> %a, <8 x i64> %b, i8 %mask, <8 x i64>* %passthru) {
; CHECK-LABEL: stack_fold_shufi64x2_maskz:
; CHECK: # %bb.0:
; CHECK-NEXT: subq $56, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 64
; CHECK-NEXT: pushq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: vmovups %zmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: vmovups %zmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: #APP
@ -6974,7 +6974,7 @@ define <8 x i64> @stack_fold_shufi64x2_maskz(<8 x i64> %a, <8 x i64> %b, i8 %mas
; CHECK-NEXT: vmovdqu64 {{[-0-9]+}}(%r{{[sb]}}p), %zmm0 # 64-byte Reload
; CHECK-NEXT: vshufi64x2 $24, {{[-0-9]+}}(%r{{[sb]}}p), %zmm0, %zmm0 {%k1} {z} # 64-byte Folded Reload
; CHECK-NEXT: # zmm0 {%k1} {z} = zmm0[0,1,4,5],mem[2,3,0,1]
; CHECK-NEXT: addq $56, %rsp
; CHECK-NEXT: popq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 8
; CHECK-NEXT: retq
%1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
@ -6987,8 +6987,8 @@ define <8 x i64> @stack_fold_shufi64x2_maskz(<8 x i64> %a, <8 x i64> %b, i8 %mas
define <16 x i32> @stack_fold_shufi32x4_mask(<16 x i32> %a, <16 x i32> %b, i16 %mask, <16 x i32>* %passthru) {
; CHECK-LABEL: stack_fold_shufi32x4_mask:
; CHECK: # %bb.0:
; CHECK-NEXT: subq $56, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 64
; CHECK-NEXT: pushq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: vmovups %zmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: vmovups %zmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: #APP
@ -7000,7 +7000,7 @@ define <16 x i32> @stack_fold_shufi32x4_mask(<16 x i32> %a, <16 x i32> %b, i16 %
; CHECK-NEXT: vshufi32x4 $20, {{[-0-9]+}}(%r{{[sb]}}p), %zmm0, %zmm1 {%k1} # 64-byte Folded Reload
; CHECK-NEXT: # zmm1 {%k1} = zmm0[0,1,2,3,4,5,6,7],mem[4,5,6,7,0,1,2,3]
; CHECK-NEXT: vmovdqa64 %zmm1, %zmm0
; CHECK-NEXT: addq $56, %rsp
; CHECK-NEXT: popq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 8
; CHECK-NEXT: retq
%1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
@ -7015,8 +7015,8 @@ define <16 x i32> @stack_fold_shufi32x4_mask(<16 x i32> %a, <16 x i32> %b, i16 %
define <16 x i32> @stack_fold_shufi32x4_maskz(<16 x i32> %a, <16 x i32> %b, i16 %mask) {
; CHECK-LABEL: stack_fold_shufi32x4_maskz:
; CHECK: # %bb.0:
; CHECK-NEXT: subq $56, %rsp
; CHECK-NEXT: .cfi_def_cfa_offset 64
; CHECK-NEXT: pushq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: vmovups %zmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: vmovups %zmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; CHECK-NEXT: #APP
@ -7026,7 +7026,7 @@ define <16 x i32> @stack_fold_shufi32x4_maskz(<16 x i32> %a, <16 x i32> %b, i16
; CHECK-NEXT: vmovdqu64 {{[-0-9]+}}(%r{{[sb]}}p), %zmm0 # 64-byte Reload
; CHECK-NEXT: vshufi32x4 $20, {{[-0-9]+}}(%r{{[sb]}}p), %zmm0, %zmm0 {%k1} {z} # 64-byte Folded Reload
; CHECK-NEXT: # zmm0 {%k1} {z} = zmm0[0,1,2,3,4,5,6,7],mem[4,5,6,7,0,1,2,3]
; CHECK-NEXT: addq $56, %rsp
; CHECK-NEXT: popq %rax
; CHECK-NEXT: .cfi_def_cfa_offset 8
; CHECK-NEXT: retq
%1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()

90
test/CodeGen/X86/vec-libcalls.ll
View File

@ -40,7 +40,7 @@ define <1 x float> @sin_v1f32(<1 x float> %x) nounwind {
; CHECK-LABEL: sin_v1f32:
; CHECK: # %bb.0:
; CHECK-NEXT: pushq %rax
; CHECK-NEXT: callq sinf
; CHECK-NEXT: callq sinf@PLT
; CHECK-NEXT: popq %rax
; CHECK-NEXT: retq
%r = call <1 x float> @llvm.sin.v1f32(<1 x float> %x)
@ -52,11 +52,11 @@ define <2 x float> @sin_v2f32(<2 x float> %x) nounwind {
; CHECK: # %bb.0:
; CHECK-NEXT: subq $40, %rsp
; CHECK-NEXT: vmovaps %xmm0, (%rsp) # 16-byte Spill
; CHECK-NEXT: callq sinf
; CHECK-NEXT: callq sinf@PLT
; CHECK-NEXT: vmovaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT: vmovshdup (%rsp), %xmm0 # 16-byte Folded Reload
; CHECK-NEXT: # xmm0 = mem[1,1,3,3]
; CHECK-NEXT: callq sinf
; CHECK-NEXT: callq sinf@PLT
; CHECK-NEXT: vmovaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
; CHECK-NEXT: vinsertps {{.*#+}} xmm0 = xmm1[0],xmm0[0],xmm1[2,3]
; CHECK-NEXT: addq $40, %rsp
@ -70,17 +70,17 @@ define <3 x float> @sin_v3f32(<3 x float> %x) nounwind {
; CHECK: # %bb.0:
; CHECK-NEXT: subq $40, %rsp
; CHECK-NEXT: vmovaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT: callq sinf
; CHECK-NEXT: callq sinf@PLT
; CHECK-NEXT: vmovaps %xmm0, (%rsp) # 16-byte Spill
; CHECK-NEXT: vmovshdup {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Folded Reload
; CHECK-NEXT: # xmm0 = mem[1,1,3,3]
; CHECK-NEXT: callq sinf
; CHECK-NEXT: callq sinf@PLT
; CHECK-NEXT: vmovaps (%rsp), %xmm1 # 16-byte Reload
; CHECK-NEXT: vinsertps {{.*#+}} xmm0 = xmm1[0],xmm0[0],xmm1[2,3]
; CHECK-NEXT: vmovaps %xmm0, (%rsp) # 16-byte Spill
; CHECK-NEXT: vpermilpd $1, {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Folded Reload
; CHECK-NEXT: # xmm0 = mem[1,0]
; CHECK-NEXT: callq sinf
; CHECK-NEXT: callq sinf@PLT
; CHECK-NEXT: vmovaps (%rsp), %xmm1 # 16-byte Reload
; CHECK-NEXT: vinsertps {{.*#+}} xmm0 = xmm1[0,1],xmm0[0],xmm1[3]
; CHECK-NEXT: addq $40, %rsp
@ -94,23 +94,23 @@ define <4 x float> @sin_v4f32(<4 x float> %x) nounwind {
; CHECK: # %bb.0:
; CHECK-NEXT: subq $40, %rsp
; CHECK-NEXT: vmovaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT: callq sinf
; CHECK-NEXT: callq sinf@PLT
; CHECK-NEXT: vmovaps %xmm0, (%rsp) # 16-byte Spill
; CHECK-NEXT: vmovshdup {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Folded Reload
; CHECK-NEXT: # xmm0 = mem[1,1,3,3]
; CHECK-NEXT: callq sinf
; CHECK-NEXT: callq sinf@PLT
; CHECK-NEXT: vmovaps (%rsp), %xmm1 # 16-byte Reload
; CHECK-NEXT: vinsertps {{.*#+}} xmm0 = xmm1[0],xmm0[0],xmm1[2,3]
; CHECK-NEXT: vmovaps %xmm0, (%rsp) # 16-byte Spill
; CHECK-NEXT: vpermilpd $1, {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Folded Reload
; CHECK-NEXT: # xmm0 = mem[1,0]
; CHECK-NEXT: callq sinf
; CHECK-NEXT: callq sinf@PLT
; CHECK-NEXT: vmovaps (%rsp), %xmm1 # 16-byte Reload
; CHECK-NEXT: vinsertps {{.*#+}} xmm0 = xmm1[0,1],xmm0[0],xmm1[3]
; CHECK-NEXT: vmovaps %xmm0, (%rsp) # 16-byte Spill
; CHECK-NEXT: vpermilps $255, {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Folded Reload
; CHECK-NEXT: # xmm0 = mem[3,3,3,3]
; CHECK-NEXT: callq sinf
; CHECK-NEXT: callq sinf@PLT
; CHECK-NEXT: vmovaps (%rsp), %xmm1 # 16-byte Reload
; CHECK-NEXT: vinsertps {{.*#+}} xmm0 = xmm1[0,1,2],xmm0[0]
; CHECK-NEXT: addq $40, %rsp
@ -122,37 +122,37 @@ define <4 x float> @sin_v4f32(<4 x float> %x) nounwind {
define <5 x float> @sin_v5f32(<5 x float> %x) nounwind {
; CHECK-LABEL: sin_v5f32:
; CHECK: # %bb.0:
; CHECK-NEXT: subq $88, %rsp
; CHECK-NEXT: subq $72, %rsp
; CHECK-NEXT: vmovups %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT: # kill: def $xmm0 killed $xmm0 killed $ymm0
; CHECK-NEXT: vzeroupper
; CHECK-NEXT: callq sinf
; CHECK-NEXT: callq sinf@PLT
; CHECK-NEXT: vmovaps %xmm0, (%rsp) # 16-byte Spill
; CHECK-NEXT: vmovshdup {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Folded Reload
; CHECK-NEXT: # xmm0 = mem[1,1,3,3]
; CHECK-NEXT: callq sinf
; CHECK-NEXT: callq sinf@PLT
; CHECK-NEXT: vmovaps (%rsp), %xmm1 # 16-byte Reload
; CHECK-NEXT: vinsertps {{.*#+}} xmm0 = xmm1[0],xmm0[0],xmm1[2,3]
; CHECK-NEXT: vmovaps %xmm0, (%rsp) # 16-byte Spill
; CHECK-NEXT: vpermilpd $1, {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Folded Reload
; CHECK-NEXT: # xmm0 = mem[1,0]
; CHECK-NEXT: callq sinf
; CHECK-NEXT: callq sinf@PLT
; CHECK-NEXT: vmovaps (%rsp), %xmm1 # 16-byte Reload
; CHECK-NEXT: vinsertps {{.*#+}} xmm0 = xmm1[0,1],xmm0[0],xmm1[3]
; CHECK-NEXT: vmovaps %xmm0, (%rsp) # 16-byte Spill
; CHECK-NEXT: vpermilps $255, {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Folded Reload
; CHECK-NEXT: # xmm0 = mem[3,3,3,3]
; CHECK-NEXT: callq sinf
; CHECK-NEXT: callq sinf@PLT
; CHECK-NEXT: vmovaps (%rsp), %xmm1 # 16-byte Reload
; CHECK-NEXT: vinsertps {{.*#+}} xmm0 = xmm1[0,1,2],xmm0[0]
; CHECK-NEXT: vmovups %ymm0, (%rsp) # 32-byte Spill
; CHECK-NEXT: vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; CHECK-NEXT: vextractf128 $1, %ymm0, %xmm0
; CHECK-NEXT: vzeroupper
; CHECK-NEXT: callq sinf
; CHECK-NEXT: callq sinf@PLT
; CHECK-NEXT: vmovups (%rsp), %ymm1 # 32-byte Reload
; CHECK-NEXT: vinsertf128 $1, %xmm0, %ymm1, %ymm0
; CHECK-NEXT: addq $88, %rsp
; CHECK-NEXT: addq $72, %rsp
; CHECK-NEXT: retq
%r = call <5 x float> @llvm.sin.v5f32(<5 x float> %x)
ret <5 x float> %r
@ -161,43 +161,43 @@ define <5 x float> @sin_v5f32(<5 x float> %x) nounwind {
define <6 x float> @sin_v6f32(<6 x float> %x) nounwind {
; CHECK-LABEL: sin_v6f32:
; CHECK: # %bb.0:
; CHECK-NEXT: subq $88, %rsp
; CHECK-NEXT: subq $72, %rsp
; CHECK-NEXT: vmovups %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT: vextractf128 $1, %ymm0, %xmm0
; CHECK-NEXT: vmovaps %xmm0, (%rsp) # 16-byte Spill
; CHECK-NEXT: vzeroupper
; CHECK-NEXT: callq sinf
; CHECK-NEXT: callq sinf@PLT
; CHECK-NEXT: vmovaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT: vmovshdup (%rsp), %xmm0 # 16-byte Folded Reload
; CHECK-NEXT: # xmm0 = mem[1,1,3,3]
; CHECK-NEXT: callq sinf
; CHECK-NEXT: callq sinf@PLT
; CHECK-NEXT: vmovaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
; CHECK-NEXT: vinsertps {{.*#+}} xmm0 = xmm1[0],xmm0[0],xmm1[2,3]
; CHECK-NEXT: vmovaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT: vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; CHECK-NEXT: # kill: def $xmm0 killed $xmm0 killed $ymm0
; CHECK-NEXT: vzeroupper
; CHECK-NEXT: callq sinf
; CHECK-NEXT: callq sinf@PLT
; CHECK-NEXT: vmovaps %xmm0, (%rsp) # 16-byte Spill
; CHECK-NEXT: vmovshdup {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Folded Reload
; CHECK-NEXT: # xmm0 = mem[1,1,3,3]
; CHECK-NEXT: callq sinf
; CHECK-NEXT: callq sinf@PLT
; CHECK-NEXT: vmovaps (%rsp), %xmm1 # 16-byte Reload
; CHECK-NEXT: vinsertps {{.*#+}} xmm0 = xmm1[0],xmm0[0],xmm1[2,3]
; CHECK-NEXT: vmovaps %xmm0, (%rsp) # 16-byte Spill
; CHECK-NEXT: vpermilpd $1, {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Folded Reload
; CHECK-NEXT: # xmm0 = mem[1,0]
; CHECK-NEXT: callq sinf
; CHECK-NEXT: callq sinf@PLT
; CHECK-NEXT: vmovaps (%rsp), %xmm1 # 16-byte Reload
; CHECK-NEXT: vinsertps {{.*#+}} xmm0 = xmm1[0,1],xmm0[0],xmm1[3]
; CHECK-NEXT: vmovaps %xmm0, (%rsp) # 16-byte Spill
; CHECK-NEXT: vpermilps $255, {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Folded Reload
; CHECK-NEXT: # xmm0 = mem[3,3,3,3]
; CHECK-NEXT: callq sinf
; CHECK-NEXT: callq sinf@PLT
; CHECK-NEXT: vmovaps (%rsp), %xmm1 # 16-byte Reload
; CHECK-NEXT: vinsertps {{.*#+}} xmm0 = xmm1[0,1,2],xmm0[0]
; CHECK-NEXT: vinsertf128 $1, {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 16-byte Folded Reload
; CHECK-NEXT: addq $88, %rsp
; CHECK-NEXT: addq $72, %rsp
; CHECK-NEXT: retq
%r = call <6 x float> @llvm.sin.v6f32(<6 x float> %x)
ret <6 x float> %r
@ -206,25 +206,25 @@ define <6 x float> @sin_v6f32(<6 x float> %x) nounwind {
define <3 x double> @sin_v3f64(<3 x double> %x) nounwind {
; CHECK-LABEL: sin_v3f64:
; CHECK: # %bb.0:
; CHECK-NEXT: subq $88, %rsp
; CHECK-NEXT: subq $72, %rsp
; CHECK-NEXT: vmovups %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT: # kill: def $xmm0 killed $xmm0 killed $ymm0
; CHECK-NEXT: vzeroupper
; CHECK-NEXT: callq sin
; CHECK-NEXT: callq sin@PLT
; CHECK-NEXT: vmovaps %xmm0, (%rsp) # 16-byte Spill
; CHECK-NEXT: vpermilpd $1, {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Folded Reload
; CHECK-NEXT: # xmm0 = mem[1,0]
; CHECK-NEXT: callq sin
; CHECK-NEXT: callq sin@PLT
; CHECK-NEXT: vmovapd (%rsp), %xmm1 # 16-byte Reload
; CHECK-NEXT: vunpcklpd {{.*#+}} xmm0 = xmm1[0],xmm0[0]
; CHECK-NEXT: vmovupd %ymm0, (%rsp) # 32-byte Spill
; CHECK-NEXT: vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; CHECK-NEXT: vextractf128 $1, %ymm0, %xmm0
; CHECK-NEXT: vzeroupper
; CHECK-NEXT: callq sin
; CHECK-NEXT: callq sin@PLT
; CHECK-NEXT: vmovups (%rsp), %ymm1 # 32-byte Reload
; CHECK-NEXT: vinsertf128 $1, %xmm0, %ymm1, %ymm0
; CHECK-NEXT: addq $88, %rsp
; CHECK-NEXT: addq $72, %rsp
; CHECK-NEXT: retq
%r = call <3 x double> @llvm.sin.v3f64(<3 x double> %x)
ret <3 x double> %r
@ -233,7 +233,7 @@ define <3 x double> @sin_v3f64(<3 x double> %x) nounwind {
define <2 x float> @fabs_v2f32(<2 x float> %x) nounwind {
; CHECK-LABEL: fabs_v2f32:
; CHECK: # %bb.0:
; CHECK-NEXT: vandps {{.*}}(%rip), %xmm0, %xmm0
; CHECK-NEXT: vandps {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0, %xmm0
; CHECK-NEXT: retq
%r = call <2 x float> @llvm.fabs.v2f32(<2 x float> %x)
ret <2 x float> %r
@ -253,11 +253,11 @@ define <2 x float> @cos_v2f32(<2 x float> %x) nounwind {
; CHECK: # %bb.0:
; CHECK-NEXT: subq $40, %rsp
; CHECK-NEXT: vmovaps %xmm0, (%rsp) # 16-byte Spill
; CHECK-NEXT: callq cosf
; CHECK-NEXT: callq cosf@PLT
; CHECK-NEXT: vmovaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT: vmovshdup (%rsp), %xmm0 # 16-byte Folded Reload
; CHECK-NEXT: # xmm0 = mem[1,1,3,3]
; CHECK-NEXT: callq cosf
; CHECK-NEXT: callq cosf@PLT
; CHECK-NEXT: vmovaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
; CHECK-NEXT: vinsertps {{.*#+}} xmm0 = xmm1[0],xmm0[0],xmm1[2,3]
; CHECK-NEXT: addq $40, %rsp
@ -271,11 +271,11 @@ define <2 x float> @exp_v2f32(<2 x float> %x) nounwind {
; CHECK: # %bb.0:
; CHECK-NEXT: subq $40, %rsp
; CHECK-NEXT: vmovaps %xmm0, (%rsp) # 16-byte Spill
; CHECK-NEXT: callq expf
; CHECK-NEXT: callq expf@PLT
; CHECK-NEXT: vmovaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT: vmovshdup (%rsp), %xmm0 # 16-byte Folded Reload
; CHECK-NEXT: # xmm0 = mem[1,1,3,3]
; CHECK-NEXT: callq expf
; CHECK-NEXT: callq expf@PLT
; CHECK-NEXT: vmovaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
; CHECK-NEXT: vinsertps {{.*#+}} xmm0 = xmm1[0],xmm0[0],xmm1[2,3]
; CHECK-NEXT: addq $40, %rsp
@ -289,11 +289,11 @@ define <2 x float> @exp2_v2f32(<2 x float> %x) nounwind {
; CHECK: # %bb.0:
; CHECK-NEXT: subq $40, %rsp
; CHECK-NEXT: vmovaps %xmm0, (%rsp) # 16-byte Spill
; CHECK-NEXT: callq exp2f
; CHECK-NEXT: callq exp2f@PLT
; CHECK-NEXT: vmovaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT: vmovshdup (%rsp), %xmm0 # 16-byte Folded Reload
; CHECK-NEXT: # xmm0 = mem[1,1,3,3]
; CHECK-NEXT: callq exp2f
; CHECK-NEXT: callq exp2f@PLT
; CHECK-NEXT: vmovaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
; CHECK-NEXT: vinsertps {{.*#+}} xmm0 = xmm1[0],xmm0[0],xmm1[2,3]
; CHECK-NEXT: addq $40, %rsp
@ -316,11 +316,11 @@ define <2 x float> @log_v2f32(<2 x float> %x) nounwind {
; CHECK: # %bb.0:
; CHECK-NEXT: subq $40, %rsp
; CHECK-NEXT: vmovaps %xmm0, (%rsp) # 16-byte Spill
; CHECK-NEXT: callq logf
; CHECK-NEXT: callq logf@PLT
; CHECK-NEXT: vmovaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT: vmovshdup (%rsp), %xmm0 # 16-byte Folded Reload
; CHECK-NEXT: # xmm0 = mem[1,1,3,3]
; CHECK-NEXT: callq logf
; CHECK-NEXT: callq logf@PLT
; CHECK-NEXT: vmovaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
; CHECK-NEXT: vinsertps {{.*#+}} xmm0 = xmm1[0],xmm0[0],xmm1[2,3]
; CHECK-NEXT: addq $40, %rsp
@ -334,11 +334,11 @@ define <2 x float> @log10_v2f32(<2 x float> %x) nounwind {
; CHECK: # %bb.0:
; CHECK-NEXT: subq $40, %rsp
; CHECK-NEXT: vmovaps %xmm0, (%rsp) # 16-byte Spill
; CHECK-NEXT: callq log10f
; CHECK-NEXT: callq log10f@PLT
; CHECK-NEXT: vmovaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT: vmovshdup (%rsp), %xmm0 # 16-byte Folded Reload
; CHECK-NEXT: # xmm0 = mem[1,1,3,3]
; CHECK-NEXT: callq log10f
; CHECK-NEXT: callq log10f@PLT
; CHECK-NEXT: vmovaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
; CHECK-NEXT: vinsertps {{.*#+}} xmm0 = xmm1[0],xmm0[0],xmm1[2,3]
; CHECK-NEXT: addq $40, %rsp
@ -352,11 +352,11 @@ define <2 x float> @log2_v2f32(<2 x float> %x) nounwind {
; CHECK: # %bb.0:
; CHECK-NEXT: subq $40, %rsp
; CHECK-NEXT: vmovaps %xmm0, (%rsp) # 16-byte Spill
; CHECK-NEXT: callq log2f
; CHECK-NEXT: callq log2f@PLT
; CHECK-NEXT: vmovaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT: vmovshdup (%rsp), %xmm0 # 16-byte Folded Reload
; CHECK-NEXT: # xmm0 = mem[1,1,3,3]
; CHECK-NEXT: callq log2f
; CHECK-NEXT: callq log2f@PLT
; CHECK-NEXT: vmovaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
; CHECK-NEXT: vinsertps {{.*#+}} xmm0 = xmm1[0],xmm0[0],xmm1[2,3]
; CHECK-NEXT: addq $40, %rsp
@ -386,8 +386,8 @@ define <2 x float> @rint_v2f32(<2 x float> %x) nounwind {
define <2 x float> @round_v2f32(<2 x float> %x) nounwind {
; CHECK-LABEL: round_v2f32:
; CHECK: # %bb.0:
; CHECK-NEXT: vandps {{.*}}(%rip), %xmm0, %xmm1
; CHECK-NEXT: vorps {{.*}}(%rip), %xmm1, %xmm1
; CHECK-NEXT: vandps {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0, %xmm1
; CHECK-NEXT: vorps {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm1, %xmm1
; CHECK-NEXT: vaddps %xmm1, %xmm0, %xmm0
; CHECK-NEXT: vroundps $11, %xmm0, %xmm0
; CHECK-NEXT: retq

18
test/CodeGen/X86/vec-strict-128.ll
View File

@ -224,8 +224,8 @@ define <2 x double> @f12(<2 x double> %a0, <4 x float> %a1) #0 {
define <4 x float> @f13(<4 x float> %a, <4 x float> %b, <4 x float> %c) #0 {
; SSE-X86-LABEL: f13:
; SSE-X86: # %bb.0:
; SSE-X86-NEXT: subl $108, %esp
; SSE-X86-NEXT: .cfi_def_cfa_offset 112
; SSE-X86-NEXT: subl $100, %esp
; SSE-X86-NEXT: .cfi_def_cfa_offset 104
; SSE-X86-NEXT: movups %xmm2, {{[-0-9]+}}(%e{{[sb]}}p) # 16-byte Spill
; SSE-X86-NEXT: movups %xmm1, {{[-0-9]+}}(%e{{[sb]}}p) # 16-byte Spill
; SSE-X86-NEXT: movups %xmm0, {{[-0-9]+}}(%e{{[sb]}}p) # 16-byte Spill
@ -282,7 +282,7 @@ define <4 x float> @f13(<4 x float> %a, <4 x float> %b, <4 x float> %c) #0 {
; SSE-X86-NEXT: movss {{.*#+}} xmm2 = mem[0],zero,zero,zero
; SSE-X86-NEXT: unpcklps {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1]
; SSE-X86-NEXT: movlhps {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; SSE-X86-NEXT: addl $108, %esp
; SSE-X86-NEXT: addl $100, %esp
; SSE-X86-NEXT: .cfi_def_cfa_offset 4
; SSE-X86-NEXT: retl
;
@ -296,7 +296,7 @@ define <4 x float> @f13(<4 x float> %a, <4 x float> %b, <4 x float> %c) #0 {
; SSE-X64-NEXT: shufps {{.*#+}} xmm0 = xmm0[3,3,3,3]
; SSE-X64-NEXT: shufps {{.*#+}} xmm1 = xmm1[3,3,3,3]
; SSE-X64-NEXT: shufps {{.*#+}} xmm2 = xmm2[3,3,3,3]
; SSE-X64-NEXT: callq fmaf
; SSE-X64-NEXT: callq fmaf@PLT
; SSE-X64-NEXT: movaps %xmm0, (%rsp) # 16-byte Spill
; SSE-X64-NEXT: movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-X64-NEXT: movhlps {{.*#+}} xmm0 = xmm0[1,1]
@ -304,14 +304,14 @@ define <4 x float> @f13(<4 x float> %a, <4 x float> %b, <4 x float> %c) #0 {
; SSE-X64-NEXT: movhlps {{.*#+}} xmm1 = xmm1[1,1]
; SSE-X64-NEXT: movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm2 # 16-byte Reload
; SSE-X64-NEXT: movhlps {{.*#+}} xmm2 = xmm2[1,1]
; SSE-X64-NEXT: callq fmaf
; SSE-X64-NEXT: callq fmaf@PLT
; SSE-X64-NEXT: unpcklps (%rsp), %xmm0 # 16-byte Folded Reload
; SSE-X64-NEXT: # xmm0 = xmm0[0],mem[0],xmm0[1],mem[1]
; SSE-X64-NEXT: movaps %xmm0, (%rsp) # 16-byte Spill
; SSE-X64-NEXT: movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-X64-NEXT: movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
; SSE-X64-NEXT: movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm2 # 16-byte Reload
; SSE-X64-NEXT: callq fmaf
; SSE-X64-NEXT: callq fmaf@PLT
; SSE-X64-NEXT: movaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-X64-NEXT: movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-X64-NEXT: shufps {{.*#+}} xmm0 = xmm0[1,1,1,1]
@ -319,7 +319,7 @@ define <4 x float> @f13(<4 x float> %a, <4 x float> %b, <4 x float> %c) #0 {
; SSE-X64-NEXT: shufps {{.*#+}} xmm1 = xmm1[1,1,1,1]
; SSE-X64-NEXT: movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm2 # 16-byte Reload
; SSE-X64-NEXT: shufps {{.*#+}} xmm2 = xmm2[1,1,1,1]
; SSE-X64-NEXT: callq fmaf
; SSE-X64-NEXT: callq fmaf@PLT
; SSE-X64-NEXT: movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
; SSE-X64-NEXT: unpcklps {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1]
; SSE-X64-NEXT: unpcklpd (%rsp), %xmm1 # 16-byte Folded Reload
@ -381,7 +381,7 @@ define <2 x double> @f14(<2 x double> %a, <2 x double> %b, <2 x double> %c) #0 {
; SSE-X64-NEXT: movaps %xmm2, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-X64-NEXT: movaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-X64-NEXT: movaps %xmm0, (%rsp) # 16-byte Spill
; SSE-X64-NEXT: callq fma
; SSE-X64-NEXT: callq fma@PLT
; SSE-X64-NEXT: movaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-X64-NEXT: movaps (%rsp), %xmm0 # 16-byte Reload
; SSE-X64-NEXT: movhlps {{.*#+}} xmm0 = xmm0[1,1]
@ -389,7 +389,7 @@ define <2 x double> @f14(<2 x double> %a, <2 x double> %b, <2 x double> %c) #0 {
; SSE-X64-NEXT: movhlps {{.*#+}} xmm1 = xmm1[1,1]
; SSE-X64-NEXT: movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm2 # 16-byte Reload
; SSE-X64-NEXT: movhlps {{.*#+}} xmm2 = xmm2[1,1]
; SSE-X64-NEXT: callq fma
; SSE-X64-NEXT: callq fma@PLT
; SSE-X64-NEXT: movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
; SSE-X64-NEXT: movlhps {{.*#+}} xmm1 = xmm1[0],xmm0[0]
; SSE-X64-NEXT: movaps %xmm1, %xmm0

538
test/CodeGen/X86/vector-constrained-fp-intrinsics.ll
View File

File diff suppressed because it is too large Load Diff

230
test/CodeGen/X86/vector-half-conversions.ll
View File

@ -658,11 +658,11 @@ define <2 x i16> @cvt_2f64_to_2i16(<2 x double> %a0) nounwind {
; ALL: # %bb.0:
; ALL-NEXT: subq $40, %rsp
; ALL-NEXT: vmovaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; ALL-NEXT: callq __truncdfhf2
; ALL-NEXT: callq __truncdfhf2@PLT
; ALL-NEXT: movw %ax, (%rsp)
; ALL-NEXT: vpermilpd $1, {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Folded Reload
; ALL-NEXT: # xmm0 = mem[1,0]
; ALL-NEXT: callq __truncdfhf2
; ALL-NEXT: callq __truncdfhf2@PLT
; ALL-NEXT: movw %ax, {{[0-9]+}}(%rsp)
; ALL-NEXT: vmovaps (%rsp), %xmm0
; ALL-NEXT: addq $40, %rsp
@ -675,28 +675,28 @@ define <2 x i16> @cvt_2f64_to_2i16(<2 x double> %a0) nounwind {
define <4 x i16> @cvt_4f64_to_4i16(<4 x double> %a0) nounwind {
; ALL-LABEL: cvt_4f64_to_4i16:
; ALL: # %bb.0:
; ALL-NEXT: subq $88, %rsp
; ALL-NEXT: subq $72, %rsp
; ALL-NEXT: vmovups %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; ALL-NEXT: vextractf128 $1, %ymm0, %xmm0
; ALL-NEXT: vmovaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; ALL-NEXT: vzeroupper
; ALL-NEXT: callq __truncdfhf2
; ALL-NEXT: callq __truncdfhf2@PLT
; ALL-NEXT: movw %ax, {{[0-9]+}}(%rsp)
; ALL-NEXT: vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; ALL-NEXT: # kill: def $xmm0 killed $xmm0 killed $ymm0
; ALL-NEXT: vzeroupper
; ALL-NEXT: callq __truncdfhf2
; ALL-NEXT: callq __truncdfhf2@PLT
; ALL-NEXT: movw %ax, (%rsp)
; ALL-NEXT: vpermilpd $1, {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Folded Reload
; ALL-NEXT: # xmm0 = mem[1,0]
; ALL-NEXT: callq __truncdfhf2
; ALL-NEXT: callq __truncdfhf2@PLT
; ALL-NEXT: movw %ax, {{[0-9]+}}(%rsp)
; ALL-NEXT: vpermilpd $1, {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Folded Reload
; ALL-NEXT: # xmm0 = mem[1,0]
; ALL-NEXT: callq __truncdfhf2
; ALL-NEXT: callq __truncdfhf2@PLT
; ALL-NEXT: movw %ax, {{[0-9]+}}(%rsp)
; ALL-NEXT: vmovaps (%rsp), %xmm0
; ALL-NEXT: addq $88, %rsp
; ALL-NEXT: addq $72, %rsp
; ALL-NEXT: retq
%1 = fptrunc <4 x double> %a0 to <4 x half>
%2 = bitcast <4 x half> %1 to <4 x i16>
@ -706,28 +706,28 @@ define <4 x i16> @cvt_4f64_to_4i16(<4 x double> %a0) nounwind {
define <8 x i16> @cvt_4f64_to_8i16_undef(<4 x double> %a0) nounwind {
; ALL-LABEL: cvt_4f64_to_8i16_undef:
; ALL: # %bb.0:
; ALL-NEXT: subq $88, %rsp
; ALL-NEXT: subq $72, %rsp
; ALL-NEXT: vmovups %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; ALL-NEXT: vextractf128 $1, %ymm0, %xmm0
; ALL-NEXT: vmovaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; ALL-NEXT: vzeroupper
; ALL-NEXT: callq __truncdfhf2
; ALL-NEXT: callq __truncdfhf2@PLT
; ALL-NEXT: movw %ax, {{[0-9]+}}(%rsp)
; ALL-NEXT: vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; ALL-NEXT: # kill: def $xmm0 killed $xmm0 killed $ymm0
; ALL-NEXT: vzeroupper
; ALL-NEXT: callq __truncdfhf2
; ALL-NEXT: callq __truncdfhf2@PLT
; ALL-NEXT: movw %ax, (%rsp)
; ALL-NEXT: vpermilpd $1, {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Folded Reload
; ALL-NEXT: # xmm0 = mem[1,0]
; ALL-NEXT: callq __truncdfhf2
; ALL-NEXT: callq __truncdfhf2@PLT
; ALL-NEXT: movw %ax, {{[0-9]+}}(%rsp)
; ALL-NEXT: vpermilpd $1, {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Folded Reload
; ALL-NEXT: # xmm0 = mem[1,0]
; ALL-NEXT: callq __truncdfhf2
; ALL-NEXT: callq __truncdfhf2@PLT
; ALL-NEXT: movw %ax, {{[0-9]+}}(%rsp)
; ALL-NEXT: vmovaps (%rsp), %xmm0
; ALL-NEXT: addq $88, %rsp
; ALL-NEXT: addq $72, %rsp
; ALL-NEXT: retq
%1 = fptrunc <4 x double> %a0 to <4 x half>
%2 = bitcast <4 x half> %1 to <4 x i16>
@ -738,28 +738,28 @@ define <8 x i16> @cvt_4f64_to_8i16_undef(<4 x double> %a0) nounwind {
define <8 x i16> @cvt_4f64_to_8i16_zero(<4 x double> %a0) nounwind {
; ALL-LABEL: cvt_4f64_to_8i16_zero:
; ALL: # %bb.0:
; ALL-NEXT: subq $88, %rsp
; ALL-NEXT: subq $72, %rsp
; ALL-NEXT: vmovups %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; ALL-NEXT: vextractf128 $1, %ymm0, %xmm0
; ALL-NEXT: vmovaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; ALL-NEXT: vzeroupper
; ALL-NEXT: callq __truncdfhf2
; ALL-NEXT: callq __truncdfhf2@PLT
; ALL-NEXT: movw %ax, {{[0-9]+}}(%rsp)
; ALL-NEXT: vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; ALL-NEXT: # kill: def $xmm0 killed $xmm0 killed $ymm0
; ALL-NEXT: vzeroupper
; ALL-NEXT: callq __truncdfhf2
; ALL-NEXT: callq __truncdfhf2@PLT
; ALL-NEXT: movw %ax, (%rsp)
; ALL-NEXT: vpermilpd $1, {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Folded Reload
; ALL-NEXT: # xmm0 = mem[1,0]
; ALL-NEXT: callq __truncdfhf2
; ALL-NEXT: callq __truncdfhf2@PLT
; ALL-NEXT: movw %ax, {{[0-9]+}}(%rsp)
; ALL-NEXT: vpermilpd $1, {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Folded Reload
; ALL-NEXT: # xmm0 = mem[1,0]
; ALL-NEXT: callq __truncdfhf2
; ALL-NEXT: callq __truncdfhf2@PLT
; ALL-NEXT: movw %ax, {{[0-9]+}}(%rsp)
; ALL-NEXT: vmovsd {{.*#+}} xmm0 = mem[0],zero
; ALL-NEXT: addq $88, %rsp
; ALL-NEXT: addq $72, %rsp
; ALL-NEXT: retq
%1 = fptrunc <4 x double> %a0 to <4 x half>
%2 = bitcast <4 x half> %1 to <4 x i16>
@ -778,13 +778,13 @@ define <8 x i16> @cvt_8f64_to_8i16(<8 x double> %a0) nounwind {
; AVX1-NEXT: vmovupd %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX1-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX1-NEXT: vzeroupper
; AVX1-NEXT: callq __truncdfhf2
; AVX1-NEXT: callq __truncdfhf2@PLT
; AVX1-NEXT: movl %eax, %ebx
; AVX1-NEXT: shll $16, %ebx
; AVX1-NEXT: vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX1-NEXT: # kill: def $xmm0 killed $xmm0 killed $ymm0
; AVX1-NEXT: vzeroupper
; AVX1-NEXT: callq __truncdfhf2
; AVX1-NEXT: callq __truncdfhf2@PLT
; AVX1-NEXT: movzwl %ax, %r15d
; AVX1-NEXT: orl %ebx, %r15d
; AVX1-NEXT: vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
@ -792,24 +792,24 @@ define <8 x i16> @cvt_8f64_to_8i16(<8 x double> %a0) nounwind {
; AVX1-NEXT: vmovapd %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; AVX1-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX1-NEXT: vzeroupper
; AVX1-NEXT: callq __truncdfhf2
; AVX1-NEXT: callq __truncdfhf2@PLT
; AVX1-NEXT: movl %eax, %ebx
; AVX1-NEXT: shll $16, %ebx
; AVX1-NEXT: vmovaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; AVX1-NEXT: callq __truncdfhf2
; AVX1-NEXT: callq __truncdfhf2@PLT
; AVX1-NEXT: movzwl %ax, %r14d
; AVX1-NEXT: orl %ebx, %r14d
; AVX1-NEXT: shlq $32, %r14
; AVX1-NEXT: orq %r15, %r14
; AVX1-NEXT: vpermilpd $1, (%rsp), %xmm0 # 16-byte Folded Reload
; AVX1-NEXT: # xmm0 = mem[1,0]
; AVX1-NEXT: callq __truncdfhf2
; AVX1-NEXT: callq __truncdfhf2@PLT
; AVX1-NEXT: movl %eax, %ebx
; AVX1-NEXT: shll $16, %ebx
; AVX1-NEXT: vmovups (%rsp), %ymm0 # 32-byte Reload
; AVX1-NEXT: # kill: def $xmm0 killed $xmm0 killed $ymm0
; AVX1-NEXT: vzeroupper
; AVX1-NEXT: callq __truncdfhf2
; AVX1-NEXT: callq __truncdfhf2@PLT
; AVX1-NEXT: movzwl %ax, %r15d
; AVX1-NEXT: orl %ebx, %r15d
; AVX1-NEXT: vmovups (%rsp), %ymm0 # 32-byte Reload
@ -817,11 +817,11 @@ define <8 x i16> @cvt_8f64_to_8i16(<8 x double> %a0) nounwind {
; AVX1-NEXT: vmovapd %xmm0, (%rsp) # 16-byte Spill
; AVX1-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX1-NEXT: vzeroupper
; AVX1-NEXT: callq __truncdfhf2
; AVX1-NEXT: callq __truncdfhf2@PLT
; AVX1-NEXT: movl %eax, %ebx
; AVX1-NEXT: shll $16, %ebx
; AVX1-NEXT: vmovaps (%rsp), %xmm0 # 16-byte Reload
; AVX1-NEXT: callq __truncdfhf2
; AVX1-NEXT: callq __truncdfhf2@PLT
; AVX1-NEXT: movzwl %ax, %eax
; AVX1-NEXT: orl %ebx, %eax
; AVX1-NEXT: shlq $32, %rax
@ -845,13 +845,13 @@ define <8 x i16> @cvt_8f64_to_8i16(<8 x double> %a0) nounwind {
; AVX2-NEXT: vmovupd %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX2-NEXT: vzeroupper
; AVX2-NEXT: callq __truncdfhf2
; AVX2-NEXT: callq __truncdfhf2@PLT
; AVX2-NEXT: movl %eax, %ebx
; AVX2-NEXT: shll $16, %ebx
; AVX2-NEXT: vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-NEXT: # kill: def $xmm0 killed $xmm0 killed $ymm0
; AVX2-NEXT: vzeroupper
; AVX2-NEXT: callq __truncdfhf2
; AVX2-NEXT: callq __truncdfhf2@PLT
; AVX2-NEXT: movzwl %ax, %r15d
; AVX2-NEXT: orl %ebx, %r15d
; AVX2-NEXT: vmovupd {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
@ -859,24 +859,24 @@ define <8 x i16> @cvt_8f64_to_8i16(<8 x double> %a0) nounwind {
; AVX2-NEXT: vmovapd %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; AVX2-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX2-NEXT: vzeroupper
; AVX2-NEXT: callq __truncdfhf2
; AVX2-NEXT: callq __truncdfhf2@PLT
; AVX2-NEXT: movl %eax, %ebx
; AVX2-NEXT: shll $16, %ebx
; AVX2-NEXT: vmovaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; AVX2-NEXT: callq __truncdfhf2
; AVX2-NEXT: callq __truncdfhf2@PLT
; AVX2-NEXT: movzwl %ax, %r14d
; AVX2-NEXT: orl %ebx, %r14d
; AVX2-NEXT: shlq $32, %r14
; AVX2-NEXT: orq %r15, %r14
; AVX2-NEXT: vpermilpd $1, (%rsp), %xmm0 # 16-byte Folded Reload
; AVX2-NEXT: # xmm0 = mem[1,0]
; AVX2-NEXT: callq __truncdfhf2
; AVX2-NEXT: callq __truncdfhf2@PLT
; AVX2-NEXT: movl %eax, %ebx
; AVX2-NEXT: shll $16, %ebx
; AVX2-NEXT: vmovups (%rsp), %ymm0 # 32-byte Reload
; AVX2-NEXT: # kill: def $xmm0 killed $xmm0 killed $ymm0
; AVX2-NEXT: vzeroupper
; AVX2-NEXT: callq __truncdfhf2
; AVX2-NEXT: callq __truncdfhf2@PLT
; AVX2-NEXT: movzwl %ax, %r15d
; AVX2-NEXT: orl %ebx, %r15d
; AVX2-NEXT: vmovupd (%rsp), %ymm0 # 32-byte Reload
@ -884,11 +884,11 @@ define <8 x i16> @cvt_8f64_to_8i16(<8 x double> %a0) nounwind {
; AVX2-NEXT: vmovapd %xmm0, (%rsp) # 16-byte Spill
; AVX2-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX2-NEXT: vzeroupper
; AVX2-NEXT: callq __truncdfhf2
; AVX2-NEXT: callq __truncdfhf2@PLT
; AVX2-NEXT: movl %eax, %ebx
; AVX2-NEXT: shll $16, %ebx
; AVX2-NEXT: vmovaps (%rsp), %xmm0 # 16-byte Reload
; AVX2-NEXT: callq __truncdfhf2
; AVX2-NEXT: callq __truncdfhf2@PLT
; AVX2-NEXT: movzwl %ax, %eax
; AVX2-NEXT: orl %ebx, %eax
; AVX2-NEXT: shlq $32, %rax
@ -907,17 +907,17 @@ define <8 x i16> @cvt_8f64_to_8i16(<8 x double> %a0) nounwind {
; AVX512-NEXT: pushq %r15
; AVX512-NEXT: pushq %r14
; AVX512-NEXT: pushq %rbx
; AVX512-NEXT: subq $96, %rsp
; AVX512-NEXT: subq $80, %rsp
; AVX512-NEXT: vmovupd %zmm0, (%rsp) # 64-byte Spill
; AVX512-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX512-NEXT: vzeroupper
; AVX512-NEXT: callq __truncdfhf2
; AVX512-NEXT: callq __truncdfhf2@PLT
; AVX512-NEXT: movl %eax, %ebx
; AVX512-NEXT: shll $16, %ebx
; AVX512-NEXT: vmovups (%rsp), %zmm0 # 64-byte Reload
; AVX512-NEXT: # kill: def $xmm0 killed $xmm0 killed $zmm0
; AVX512-NEXT: vzeroupper
; AVX512-NEXT: callq __truncdfhf2
; AVX512-NEXT: callq __truncdfhf2@PLT
; AVX512-NEXT: movzwl %ax, %r15d
; AVX512-NEXT: orl %ebx, %r15d
; AVX512-NEXT: vmovupd (%rsp), %zmm0 # 64-byte Reload
@ -925,11 +925,11 @@ define <8 x i16> @cvt_8f64_to_8i16(<8 x double> %a0) nounwind {
; AVX512-NEXT: vmovapd %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; AVX512-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX512-NEXT: vzeroupper
; AVX512-NEXT: callq __truncdfhf2
; AVX512-NEXT: callq __truncdfhf2@PLT
; AVX512-NEXT: movl %eax, %ebx
; AVX512-NEXT: shll $16, %ebx
; AVX512-NEXT: vmovaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; AVX512-NEXT: callq __truncdfhf2
; AVX512-NEXT: callq __truncdfhf2@PLT
; AVX512-NEXT: movzwl %ax, %r14d
; AVX512-NEXT: orl %ebx, %r14d
; AVX512-NEXT: shlq $32, %r14
@ -939,13 +939,13 @@ define <8 x i16> @cvt_8f64_to_8i16(<8 x double> %a0) nounwind {
; AVX512-NEXT: vmovupd %ymm0, (%rsp) # 32-byte Spill
; AVX512-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX512-NEXT: vzeroupper
; AVX512-NEXT: callq __truncdfhf2
; AVX512-NEXT: callq __truncdfhf2@PLT
; AVX512-NEXT: movl %eax, %ebx
; AVX512-NEXT: shll $16, %ebx
; AVX512-NEXT: vmovups (%rsp), %ymm0 # 32-byte Reload
; AVX512-NEXT: # kill: def $xmm0 killed $xmm0 killed $ymm0
; AVX512-NEXT: vzeroupper
; AVX512-NEXT: callq __truncdfhf2
; AVX512-NEXT: callq __truncdfhf2@PLT
; AVX512-NEXT: movzwl %ax, %r15d
; AVX512-NEXT: orl %ebx, %r15d
; AVX512-NEXT: vmovupd (%rsp), %ymm0 # 32-byte Reload
@ -953,11 +953,11 @@ define <8 x i16> @cvt_8f64_to_8i16(<8 x double> %a0) nounwind {
; AVX512-NEXT: vmovapd %xmm0, (%rsp) # 16-byte Spill
; AVX512-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX512-NEXT: vzeroupper
; AVX512-NEXT: callq __truncdfhf2
; AVX512-NEXT: callq __truncdfhf2@PLT
; AVX512-NEXT: movl %eax, %ebx
; AVX512-NEXT: shll $16, %ebx
; AVX512-NEXT: vmovaps (%rsp), %xmm0 # 16-byte Reload
; AVX512-NEXT: callq __truncdfhf2
; AVX512-NEXT: callq __truncdfhf2@PLT
; AVX512-NEXT: movzwl %ax, %eax
; AVX512-NEXT: orl %ebx, %eax
; AVX512-NEXT: shlq $32, %rax
@ -965,7 +965,7 @@ define <8 x i16> @cvt_8f64_to_8i16(<8 x double> %a0) nounwind {
; AVX512-NEXT: vmovq %rax, %xmm0
; AVX512-NEXT: vmovq %r14, %xmm1
; AVX512-NEXT: vpunpcklqdq {{.*#+}} xmm0 = xmm1[0],xmm0[0]
; AVX512-NEXT: addq $96, %rsp
; AVX512-NEXT: addq $80, %rsp
; AVX512-NEXT: popq %rbx
; AVX512-NEXT: popq %r14
; AVX512-NEXT: popq %r15
@ -984,7 +984,7 @@ define void @store_cvt_f64_to_i16(double %a0, i16* %a1) nounwind {
; ALL: # %bb.0:
; ALL-NEXT: pushq %rbx
; ALL-NEXT: movq %rdi, %rbx
; ALL-NEXT: callq __truncdfhf2
; ALL-NEXT: callq __truncdfhf2@PLT
; ALL-NEXT: movw %ax, (%rbx)
; ALL-NEXT: popq %rbx
; ALL-NEXT: retq
@ -1003,10 +1003,10 @@ define void @store_cvt_2f64_to_2i16(<2 x double> %a0, <2 x i16>* %a1) nounwind {
; ALL-NEXT: movq %rdi, %rbx
; ALL-NEXT: vmovapd %xmm0, (%rsp) # 16-byte Spill
; ALL-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; ALL-NEXT: callq __truncdfhf2
; ALL-NEXT: callq __truncdfhf2@PLT
; ALL-NEXT: movl %eax, %ebp
; ALL-NEXT: vmovaps (%rsp), %xmm0 # 16-byte Reload
; ALL-NEXT: callq __truncdfhf2
; ALL-NEXT: callq __truncdfhf2@PLT
; ALL-NEXT: movw %ax, (%rbx)
; ALL-NEXT: movw %bp, 2(%rbx)
; ALL-NEXT: addq $24, %rsp
@ -1026,32 +1026,32 @@ define void @store_cvt_4f64_to_4i16(<4 x double> %a0, <4 x i16>* %a1) nounwind {
; AVX1-NEXT: pushq %r15
; AVX1-NEXT: pushq %r14
; AVX1-NEXT: pushq %rbx
; AVX1-NEXT: subq $88, %rsp
; AVX1-NEXT: subq $56, %rsp
; AVX1-NEXT: movq %rdi, %rbx
; AVX1-NEXT: vmovupd %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX1-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX1-NEXT: vzeroupper
; AVX1-NEXT: callq __truncdfhf2
; AVX1-NEXT: callq __truncdfhf2@PLT
; AVX1-NEXT: movl %eax, %r14d
; AVX1-NEXT: vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm0
; AVX1-NEXT: vmovapd %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; AVX1-NEXT: vmovapd %xmm0, (%rsp) # 16-byte Spill
; AVX1-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX1-NEXT: vzeroupper
; AVX1-NEXT: callq __truncdfhf2
; AVX1-NEXT: callq __truncdfhf2@PLT
; AVX1-NEXT: movl %eax, %r15d
; AVX1-NEXT: vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX1-NEXT: # kill: def $xmm0 killed $xmm0 killed $ymm0
; AVX1-NEXT: vzeroupper
; AVX1-NEXT: callq __truncdfhf2
; AVX1-NEXT: callq __truncdfhf2@PLT
; AVX1-NEXT: movl %eax, %ebp
; AVX1-NEXT: vmovaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; AVX1-NEXT: callq __truncdfhf2
; AVX1-NEXT: vmovaps (%rsp), %xmm0 # 16-byte Reload
; AVX1-NEXT: callq __truncdfhf2@PLT
; AVX1-NEXT: movw %ax, 4(%rbx)
; AVX1-NEXT: movw %bp, (%rbx)
; AVX1-NEXT: movw %r15w, 6(%rbx)
; AVX1-NEXT: movw %r14w, 2(%rbx)
; AVX1-NEXT: addq $88, %rsp
; AVX1-NEXT: addq $56, %rsp
; AVX1-NEXT: popq %rbx
; AVX1-NEXT: popq %r14
; AVX1-NEXT: popq %r15
@ -1064,32 +1064,32 @@ define void @store_cvt_4f64_to_4i16(<4 x double> %a0, <4 x i16>* %a1) nounwind {
; AVX2-NEXT: pushq %r15
; AVX2-NEXT: pushq %r14
; AVX2-NEXT: pushq %rbx
; AVX2-NEXT: subq $88, %rsp
; AVX2-NEXT: subq $56, %rsp
; AVX2-NEXT: movq %rdi, %rbx
; AVX2-NEXT: vmovupd %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX2-NEXT: vzeroupper
; AVX2-NEXT: callq __truncdfhf2
; AVX2-NEXT: callq __truncdfhf2@PLT
; AVX2-NEXT: movl %eax, %r14d
; AVX2-NEXT: vmovupd {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-NEXT: vextractf128 $1, %ymm0, %xmm0
; AVX2-NEXT: vmovapd %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; AVX2-NEXT: vmovapd %xmm0, (%rsp) # 16-byte Spill
; AVX2-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX2-NEXT: vzeroupper
; AVX2-NEXT: callq __truncdfhf2
; AVX2-NEXT: callq __truncdfhf2@PLT
; AVX2-NEXT: movl %eax, %r15d
; AVX2-NEXT: vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-NEXT: # kill: def $xmm0 killed $xmm0 killed $ymm0
; AVX2-NEXT: vzeroupper
; AVX2-NEXT: callq __truncdfhf2
; AVX2-NEXT: callq __truncdfhf2@PLT
; AVX2-NEXT: movl %eax, %ebp
; AVX2-NEXT: vmovaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; AVX2-NEXT: callq __truncdfhf2
; AVX2-NEXT: vmovaps (%rsp), %xmm0 # 16-byte Reload
; AVX2-NEXT: callq __truncdfhf2@PLT
; AVX2-NEXT: movw %ax, 4(%rbx)
; AVX2-NEXT: movw %bp, (%rbx)
; AVX2-NEXT: movw %r15w, 6(%rbx)
; AVX2-NEXT: movw %r14w, 2(%rbx)
; AVX2-NEXT: addq $88, %rsp
; AVX2-NEXT: addq $56, %rsp
; AVX2-NEXT: popq %rbx
; AVX2-NEXT: popq %r14
; AVX2-NEXT: popq %r15
@ -1102,32 +1102,32 @@ define void @store_cvt_4f64_to_4i16(<4 x double> %a0, <4 x i16>* %a1) nounwind {
; AVX512-NEXT: pushq %r15
; AVX512-NEXT: pushq %r14
; AVX512-NEXT: pushq %rbx
; AVX512-NEXT: subq $88, %rsp
; AVX512-NEXT: subq $56, %rsp
; AVX512-NEXT: movq %rdi, %rbx
; AVX512-NEXT: vmovupd %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX512-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX512-NEXT: vzeroupper
; AVX512-NEXT: callq __truncdfhf2
; AVX512-NEXT: callq __truncdfhf2@PLT
; AVX512-NEXT: movl %eax, %r14d
; AVX512-NEXT: vmovupd {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX512-NEXT: vextractf128 $1, %ymm0, %xmm0
; AVX512-NEXT: vmovapd %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; AVX512-NEXT: vmovapd %xmm0, (%rsp) # 16-byte Spill
; AVX512-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX512-NEXT: vzeroupper
; AVX512-NEXT: callq __truncdfhf2
; AVX512-NEXT: callq __truncdfhf2@PLT
; AVX512-NEXT: movl %eax, %r15d
; AVX512-NEXT: vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX512-NEXT: # kill: def $xmm0 killed $xmm0 killed $ymm0
; AVX512-NEXT: vzeroupper
; AVX512-NEXT: callq __truncdfhf2
; AVX512-NEXT: callq __truncdfhf2@PLT
; AVX512-NEXT: movl %eax, %ebp
; AVX512-NEXT: vmovaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; AVX512-NEXT: callq __truncdfhf2
; AVX512-NEXT: vmovaps (%rsp), %xmm0 # 16-byte Reload
; AVX512-NEXT: callq __truncdfhf2@PLT
; AVX512-NEXT: movw %ax, 4(%rbx)
; AVX512-NEXT: movw %bp, (%rbx)
; AVX512-NEXT: movw %r15w, 6(%rbx)
; AVX512-NEXT: movw %r14w, 2(%rbx)
; AVX512-NEXT: addq $88, %rsp
; AVX512-NEXT: addq $56, %rsp
; AVX512-NEXT: popq %rbx
; AVX512-NEXT: popq %r14
; AVX512-NEXT: popq %r15
@ -1143,30 +1143,30 @@ define void @store_cvt_4f64_to_8i16_undef(<4 x double> %a0, <8 x i16>* %a1) noun
; ALL-LABEL: store_cvt_4f64_to_8i16_undef:
; ALL: # %bb.0:
; ALL-NEXT: pushq %rbx
; ALL-NEXT: subq $80, %rsp
; ALL-NEXT: subq $64, %rsp
; ALL-NEXT: movq %rdi, %rbx
; ALL-NEXT: vmovups %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; ALL-NEXT: vextractf128 $1, %ymm0, %xmm0
; ALL-NEXT: vmovaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; ALL-NEXT: vzeroupper
; ALL-NEXT: callq __truncdfhf2
; ALL-NEXT: callq __truncdfhf2@PLT
; ALL-NEXT: movw %ax, {{[0-9]+}}(%rsp)
; ALL-NEXT: vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; ALL-NEXT: # kill: def $xmm0 killed $xmm0 killed $ymm0
; ALL-NEXT: vzeroupper
; ALL-NEXT: callq __truncdfhf2
; ALL-NEXT: callq __truncdfhf2@PLT
; ALL-NEXT: movw %ax, (%rsp)
; ALL-NEXT: vpermilpd $1, {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Folded Reload
; ALL-NEXT: # xmm0 = mem[1,0]
; ALL-NEXT: callq __truncdfhf2
; ALL-NEXT: callq __truncdfhf2@PLT
; ALL-NEXT: movw %ax, {{[0-9]+}}(%rsp)
; ALL-NEXT: vpermilpd $1, {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Folded Reload
; ALL-NEXT: # xmm0 = mem[1,0]
; ALL-NEXT: callq __truncdfhf2
; ALL-NEXT: callq __truncdfhf2@PLT
; ALL-NEXT: movw %ax, {{[0-9]+}}(%rsp)
; ALL-NEXT: vmovaps (%rsp), %xmm0
; ALL-NEXT: vmovaps %xmm0, (%rbx)
; ALL-NEXT: addq $80, %rsp
; ALL-NEXT: addq $64, %rsp
; ALL-NEXT: popq %rbx
; ALL-NEXT: retq
%1 = fptrunc <4 x double> %a0 to <4 x half>
@ -1180,30 +1180,30 @@ define void @store_cvt_4f64_to_8i16_zero(<4 x double> %a0, <8 x i16>* %a1) nounw
; ALL-LABEL: store_cvt_4f64_to_8i16_zero:
; ALL: # %bb.0:
; ALL-NEXT: pushq %rbx
; ALL-NEXT: subq $80, %rsp
; ALL-NEXT: subq $64, %rsp
; ALL-NEXT: movq %rdi, %rbx
; ALL-NEXT: vmovups %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; ALL-NEXT: vextractf128 $1, %ymm0, %xmm0
; ALL-NEXT: vmovaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; ALL-NEXT: vzeroupper
; ALL-NEXT: callq __truncdfhf2
; ALL-NEXT: callq __truncdfhf2@PLT
; ALL-NEXT: movw %ax, {{[0-9]+}}(%rsp)
; ALL-NEXT: vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; ALL-NEXT: # kill: def $xmm0 killed $xmm0 killed $ymm0
; ALL-NEXT: vzeroupper
; ALL-NEXT: callq __truncdfhf2
; ALL-NEXT: callq __truncdfhf2@PLT
; ALL-NEXT: movw %ax, (%rsp)
; ALL-NEXT: vpermilpd $1, {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Folded Reload
; ALL-NEXT: # xmm0 = mem[1,0]
; ALL-NEXT: callq __truncdfhf2
; ALL-NEXT: callq __truncdfhf2@PLT
; ALL-NEXT: movw %ax, {{[0-9]+}}(%rsp)
; ALL-NEXT: vpermilpd $1, {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Folded Reload
; ALL-NEXT: # xmm0 = mem[1,0]
; ALL-NEXT: callq __truncdfhf2
; ALL-NEXT: callq __truncdfhf2@PLT
; ALL-NEXT: movw %ax, {{[0-9]+}}(%rsp)
; ALL-NEXT: vmovsd {{.*#+}} xmm0 = mem[0],zero
; ALL-NEXT: vmovaps %xmm0, (%rbx)
; ALL-NEXT: addq $80, %rsp
; ALL-NEXT: addq $64, %rsp
; ALL-NEXT: popq %rbx
; ALL-NEXT: retq
%1 = fptrunc <4 x double> %a0 to <4 x half>
@ -1222,47 +1222,47 @@ define void @store_cvt_8f64_to_8i16(<8 x double> %a0, <8 x i16>* %a1) nounwind {
; AVX1-NEXT: pushq %r13
; AVX1-NEXT: pushq %r12
; AVX1-NEXT: pushq %rbx
; AVX1-NEXT: subq $136, %rsp
; AVX1-NEXT: subq $120, %rsp
; AVX1-NEXT: movq %rdi, %rbx
; AVX1-NEXT: vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX1-NEXT: vmovupd %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX1-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX1-NEXT: vzeroupper
; AVX1-NEXT: callq __truncdfhf2
; AVX1-NEXT: callq __truncdfhf2@PLT
; AVX1-NEXT: movw %ax, {{[-0-9]+}}(%r{{[sb]}}p) # 2-byte Spill
; AVX1-NEXT: vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm0
; AVX1-NEXT: vmovapd %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; AVX1-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX1-NEXT: vzeroupper
; AVX1-NEXT: callq __truncdfhf2
; AVX1-NEXT: callq __truncdfhf2@PLT
; AVX1-NEXT: movw %ax, {{[-0-9]+}}(%r{{[sb]}}p) # 2-byte Spill
; AVX1-NEXT: vpermilpd $1, {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Folded Reload
; AVX1-NEXT: # xmm0 = mem[1,0]
; AVX1-NEXT: callq __truncdfhf2
; AVX1-NEXT: callq __truncdfhf2@PLT
; AVX1-NEXT: movl %eax, %r12d
; AVX1-NEXT: vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm0
; AVX1-NEXT: vmovapd %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; AVX1-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX1-NEXT: vzeroupper
; AVX1-NEXT: callq __truncdfhf2
; AVX1-NEXT: callq __truncdfhf2@PLT
; AVX1-NEXT: movl %eax, %r13d
; AVX1-NEXT: vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX1-NEXT: # kill: def $xmm0 killed $xmm0 killed $ymm0
; AVX1-NEXT: vzeroupper
; AVX1-NEXT: callq __truncdfhf2
; AVX1-NEXT: callq __truncdfhf2@PLT
; AVX1-NEXT: movl %eax, %ebp
; AVX1-NEXT: vmovaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; AVX1-NEXT: callq __truncdfhf2
; AVX1-NEXT: callq __truncdfhf2@PLT
; AVX1-NEXT: movl %eax, %r14d
; AVX1-NEXT: vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX1-NEXT: # kill: def $xmm0 killed $xmm0 killed $ymm0
; AVX1-NEXT: vzeroupper
; AVX1-NEXT: callq __truncdfhf2
; AVX1-NEXT: callq __truncdfhf2@PLT
; AVX1-NEXT: movl %eax, %r15d
; AVX1-NEXT: vmovaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; AVX1-NEXT: callq __truncdfhf2
; AVX1-NEXT: callq __truncdfhf2@PLT
; AVX1-NEXT: movw %ax, 12(%rbx)
; AVX1-NEXT: movw %r15w, 8(%rbx)
; AVX1-NEXT: movw %r14w, 4(%rbx)
@ -1273,7 +1273,7 @@ define void @store_cvt_8f64_to_8i16(<8 x double> %a0, <8 x i16>* %a1) nounwind {
; AVX1-NEXT: movw %ax, 6(%rbx)
; AVX1-NEXT: movzwl {{[-0-9]+}}(%r{{[sb]}}p), %eax # 2-byte Folded Reload
; AVX1-NEXT: movw %ax, 2(%rbx)
; AVX1-NEXT: addq $136, %rsp
; AVX1-NEXT: addq $120, %rsp
; AVX1-NEXT: popq %rbx
; AVX1-NEXT: popq %r12
; AVX1-NEXT: popq %r13
@ -1290,47 +1290,47 @@ define void @store_cvt_8f64_to_8i16(<8 x double> %a0, <8 x i16>* %a1) nounwind {
; AVX2-NEXT: pushq %r13
; AVX2-NEXT: pushq %r12
; AVX2-NEXT: pushq %rbx
; AVX2-NEXT: subq $136, %rsp
; AVX2-NEXT: subq $120, %rsp
; AVX2-NEXT: movq %rdi, %rbx
; AVX2-NEXT: vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-NEXT: vmovupd %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX2-NEXT: vzeroupper
; AVX2-NEXT: callq __truncdfhf2
; AVX2-NEXT: callq __truncdfhf2@PLT
; AVX2-NEXT: movw %ax, {{[-0-9]+}}(%r{{[sb]}}p) # 2-byte Spill
; AVX2-NEXT: vmovupd {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-NEXT: vextractf128 $1, %ymm0, %xmm0
; AVX2-NEXT: vmovapd %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; AVX2-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX2-NEXT: vzeroupper
; AVX2-NEXT: callq __truncdfhf2
; AVX2-NEXT: callq __truncdfhf2@PLT
; AVX2-NEXT: movw %ax, {{[-0-9]+}}(%r{{[sb]}}p) # 2-byte Spill
; AVX2-NEXT: vpermilpd $1, {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Folded Reload
; AVX2-NEXT: # xmm0 = mem[1,0]
; AVX2-NEXT: callq __truncdfhf2
; AVX2-NEXT: callq __truncdfhf2@PLT
; AVX2-NEXT: movl %eax, %r12d
; AVX2-NEXT: vmovupd {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-NEXT: vextractf128 $1, %ymm0, %xmm0
; AVX2-NEXT: vmovapd %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; AVX2-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX2-NEXT: vzeroupper
; AVX2-NEXT: callq __truncdfhf2
; AVX2-NEXT: callq __truncdfhf2@PLT
; AVX2-NEXT: movl %eax, %r13d
; AVX2-NEXT: vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-NEXT: # kill: def $xmm0 killed $xmm0 killed $ymm0
; AVX2-NEXT: vzeroupper
; AVX2-NEXT: callq __truncdfhf2
; AVX2-NEXT: callq __truncdfhf2@PLT
; AVX2-NEXT: movl %eax, %ebp
; AVX2-NEXT: vmovaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; AVX2-NEXT: callq __truncdfhf2
; AVX2-NEXT: callq __truncdfhf2@PLT
; AVX2-NEXT: movl %eax, %r14d
; AVX2-NEXT: vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-NEXT: # kill: def $xmm0 killed $xmm0 killed $ymm0
; AVX2-NEXT: vzeroupper
; AVX2-NEXT: callq __truncdfhf2
; AVX2-NEXT: callq __truncdfhf2@PLT
; AVX2-NEXT: movl %eax, %r15d
; AVX2-NEXT: vmovaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; AVX2-NEXT: callq __truncdfhf2
; AVX2-NEXT: callq __truncdfhf2@PLT
; AVX2-NEXT: movw %ax, 12(%rbx)
; AVX2-NEXT: movw %r15w, 8(%rbx)
; AVX2-NEXT: movw %r14w, 4(%rbx)
@ -1341,7 +1341,7 @@ define void @store_cvt_8f64_to_8i16(<8 x double> %a0, <8 x i16>* %a1) nounwind {
; AVX2-NEXT: movw %ax, 6(%rbx)
; AVX2-NEXT: movzwl {{[-0-9]+}}(%r{{[sb]}}p), %eax # 2-byte Folded Reload
; AVX2-NEXT: movw %ax, 2(%rbx)
; AVX2-NEXT: addq $136, %rsp
; AVX2-NEXT: addq $120, %rsp
; AVX2-NEXT: popq %rbx
; AVX2-NEXT: popq %r12
; AVX2-NEXT: popq %r13
@ -1358,49 +1358,49 @@ define void @store_cvt_8f64_to_8i16(<8 x double> %a0, <8 x i16>* %a1) nounwind {
; AVX512-NEXT: pushq %r13
; AVX512-NEXT: pushq %r12
; AVX512-NEXT: pushq %rbx
; AVX512-NEXT: subq $200, %rsp
; AVX512-NEXT: subq $152, %rsp
; AVX512-NEXT: movq %rdi, %rbx
; AVX512-NEXT: vmovupd %zmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
; AVX512-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX512-NEXT: vzeroupper
; AVX512-NEXT: callq __truncdfhf2
; AVX512-NEXT: callq __truncdfhf2@PLT
; AVX512-NEXT: movw %ax, {{[-0-9]+}}(%r{{[sb]}}p) # 2-byte Spill
; AVX512-NEXT: vmovupd {{[-0-9]+}}(%r{{[sb]}}p), %zmm0 # 64-byte Reload
; AVX512-NEXT: vextractf128 $1, %ymm0, %xmm0
; AVX512-NEXT: vmovapd %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; AVX512-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX512-NEXT: vzeroupper
; AVX512-NEXT: callq __truncdfhf2
; AVX512-NEXT: callq __truncdfhf2@PLT
; AVX512-NEXT: movw %ax, {{[-0-9]+}}(%r{{[sb]}}p) # 2-byte Spill
; AVX512-NEXT: vmovupd {{[-0-9]+}}(%r{{[sb]}}p), %zmm0 # 64-byte Reload
; AVX512-NEXT: vextractf64x4 $1, %zmm0, %ymm0
; AVX512-NEXT: vmovupd %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX512-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX512-NEXT: vzeroupper
; AVX512-NEXT: callq __truncdfhf2
; AVX512-NEXT: callq __truncdfhf2@PLT
; AVX512-NEXT: movl %eax, %r12d
; AVX512-NEXT: vmovupd {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX512-NEXT: vextractf128 $1, %ymm0, %xmm0
; AVX512-NEXT: vmovapd %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; AVX512-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX512-NEXT: vzeroupper
; AVX512-NEXT: callq __truncdfhf2
; AVX512-NEXT: callq __truncdfhf2@PLT
; AVX512-NEXT: movl %eax, %r13d
; AVX512-NEXT: vmovups {{[-0-9]+}}(%r{{[sb]}}p), %zmm0 # 64-byte Reload
; AVX512-NEXT: # kill: def $xmm0 killed $xmm0 killed $zmm0
; AVX512-NEXT: vzeroupper
; AVX512-NEXT: callq __truncdfhf2
; AVX512-NEXT: callq __truncdfhf2@PLT
; AVX512-NEXT: movl %eax, %ebp
; AVX512-NEXT: vmovaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; AVX512-NEXT: callq __truncdfhf2
; AVX512-NEXT: callq __truncdfhf2@PLT
; AVX512-NEXT: movl %eax, %r14d
; AVX512-NEXT: vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX512-NEXT: # kill: def $xmm0 killed $xmm0 killed $ymm0
; AVX512-NEXT: vzeroupper
; AVX512-NEXT: callq __truncdfhf2
; AVX512-NEXT: callq __truncdfhf2@PLT
; AVX512-NEXT: movl %eax, %r15d
; AVX512-NEXT: vmovaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; AVX512-NEXT: callq __truncdfhf2
; AVX512-NEXT: callq __truncdfhf2@PLT
; AVX512-NEXT: movw %ax, 12(%rbx)
; AVX512-NEXT: movw %r15w, 8(%rbx)
; AVX512-NEXT: movw %r14w, 4(%rbx)
@ -1411,7 +1411,7 @@ define void @store_cvt_8f64_to_8i16(<8 x double> %a0, <8 x i16>* %a1) nounwind {
; AVX512-NEXT: movw %ax, 6(%rbx)
; AVX512-NEXT: movzwl {{[-0-9]+}}(%r{{[sb]}}p), %eax # 2-byte Folded Reload
; AVX512-NEXT: movw %ax, 2(%rbx)
; AVX512-NEXT: addq $200, %rsp
; AVX512-NEXT: addq $152, %rsp
; AVX512-NEXT: popq %rbx
; AVX512-NEXT: popq %r12
; AVX512-NEXT: popq %r13

40
test/CodeGen/X86/vector-interleaved-load-i16-stride-4.ll
View File

@ -515,7 +515,7 @@ define void @vf16(<64 x i16>* %in.vec, <16 x i16>* %out.vec0, <16 x i16>* %out.v
define void @vf32(<128 x i16>* %in.vec, <32 x i16>* %out.vec0, <32 x i16>* %out.vec1, <32 x i16>* %out.vec2, <32 x i16>* %out.vec3) nounwind {
; AVX2-SLOW-LABEL: vf32:
; AVX2-SLOW: # %bb.0:
; AVX2-SLOW-NEXT: subq $280, %rsp # imm = 0x118
; AVX2-SLOW-NEXT: subq $248, %rsp
; AVX2-SLOW-NEXT: vmovdqa (%rdi), %xmm5
; AVX2-SLOW-NEXT: vmovdqa %xmm5, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; AVX2-SLOW-NEXT: vmovdqa 16(%rdi), %xmm9
@ -529,7 +529,7 @@ define void @vf32(<128 x i16>* %in.vec, <32 x i16>* %out.vec0, <32 x i16>* %out.
; AVX2-SLOW-NEXT: vpblendw {{.*#+}} xmm3 = xmm4[0],xmm0[1,2,3],xmm4[4],xmm0[5,6,7]
; AVX2-SLOW-NEXT: vpackusdw %xmm2, %xmm3, %xmm2
; AVX2-SLOW-NEXT: vpblendw {{.*#+}} xmm3 = xmm9[0],xmm0[1,2,3],xmm9[4],xmm0[5,6,7]
; AVX2-SLOW-NEXT: vmovdqa %xmm9, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; AVX2-SLOW-NEXT: vmovdqa %xmm9, (%rsp) # 16-byte Spill
; AVX2-SLOW-NEXT: vpblendw {{.*#+}} xmm4 = xmm5[0],xmm0[1,2,3],xmm5[4],xmm0[5,6,7]
; AVX2-SLOW-NEXT: vpackusdw %xmm3, %xmm4, %xmm3
; AVX2-SLOW-NEXT: vpackusdw %xmm2, %xmm3, %xmm2
@ -570,7 +570,7 @@ define void @vf32(<128 x i16>* %in.vec, <32 x i16>* %out.vec0, <32 x i16>* %out.
; AVX2-SLOW-NEXT: vpblendw {{.*#+}} xmm2 = xmm10[0],xmm0[1,2,3],xmm10[4],xmm0[5,6,7]
; AVX2-SLOW-NEXT: vmovdqa 224(%rdi), %xmm11
; AVX2-SLOW-NEXT: vpblendw {{.*#+}} xmm3 = xmm11[0],xmm0[1,2,3],xmm11[4],xmm0[5,6,7]
; AVX2-SLOW-NEXT: vmovdqa %xmm11, (%rsp) # 16-byte Spill
; AVX2-SLOW-NEXT: vmovdqa %xmm11, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; AVX2-SLOW-NEXT: vpackusdw %xmm2, %xmm3, %xmm7
; AVX2-SLOW-NEXT: vmovdqa 208(%rdi), %xmm3
; AVX2-SLOW-NEXT: vpblendw {{.*#+}} xmm1 = xmm3[0],xmm0[1,2,3],xmm3[4],xmm0[5,6,7]
@ -665,7 +665,7 @@ define void @vf32(<128 x i16>* %in.vec, <32 x i16>* %out.vec0, <32 x i16>* %out.
; AVX2-SLOW-NEXT: vpshuflw {{.*#+}} xmm1 = xmm8[0,1,2,0,4,5,6,7]
; AVX2-SLOW-NEXT: vpshuflw {{.*#+}} xmm4 = xmm13[0,1,2,0,4,5,6,7]
; AVX2-SLOW-NEXT: vpunpckldq {{.*#+}} xmm1 = xmm4[0],xmm1[0],xmm4[1],xmm1[1]
; AVX2-SLOW-NEXT: vpshufd $231, {{[-0-9]+}}(%r{{[sb]}}p), %xmm6 # 16-byte Folded Reload
; AVX2-SLOW-NEXT: vpshufd $231, (%rsp), %xmm6 # 16-byte Folded Reload
; AVX2-SLOW-NEXT: # xmm6 = mem[3,1,2,3]
; AVX2-SLOW-NEXT: vpshufd {{.*#+}} xmm5 = xmm9[3,1,2,3]
; AVX2-SLOW-NEXT: vpshuflw {{.*#+}} xmm0 = xmm6[2,0,2,3,4,5,6,7]
@ -675,7 +675,7 @@ define void @vf32(<128 x i16>* %in.vec, <32 x i16>* %out.vec0, <32 x i16>* %out.
; AVX2-SLOW-NEXT: vpblendd {{.*#+}} ymm0 = ymm0[0,1,2,3],ymm2[4,5,6,7]
; AVX2-SLOW-NEXT: vmovdqu %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-SLOW-NEXT: vpshufd {{.*#+}} xmm10 = xmm10[3,1,2,3]
; AVX2-SLOW-NEXT: vpshufd $231, (%rsp), %xmm9 # 16-byte Folded Reload
; AVX2-SLOW-NEXT: vpshufd $231, {{[-0-9]+}}(%r{{[sb]}}p), %xmm9 # 16-byte Folded Reload
; AVX2-SLOW-NEXT: # xmm9 = mem[3,1,2,3]
; AVX2-SLOW-NEXT: vpshuflw {{.*#+}} xmm0 = xmm10[0,1,2,0,4,5,6,7]
; AVX2-SLOW-NEXT: vpshuflw {{.*#+}} xmm1 = xmm9[0,1,2,0,4,5,6,7]
@ -753,17 +753,17 @@ define void @vf32(<128 x i16>* %in.vec, <32 x i16>* %out.vec0, <32 x i16>* %out.
; AVX2-SLOW-NEXT: vmovaps %ymm2, (%rcx)
; AVX2-SLOW-NEXT: vmovdqa %ymm1, 32(%r8)
; AVX2-SLOW-NEXT: vmovdqa %ymm0, (%r8)
; AVX2-SLOW-NEXT: addq $280, %rsp # imm = 0x118
; AVX2-SLOW-NEXT: addq $248, %rsp
; AVX2-SLOW-NEXT: vzeroupper
; AVX2-SLOW-NEXT: retq
;
; AVX2-FAST-ALL-LABEL: vf32:
; AVX2-FAST-ALL: # %bb.0:
; AVX2-FAST-ALL-NEXT: subq $216, %rsp
; AVX2-FAST-ALL-NEXT: subq $200, %rsp
; AVX2-FAST-ALL-NEXT: vmovdqa 64(%rdi), %ymm5
; AVX2-FAST-ALL-NEXT: vmovdqu %ymm5, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-FAST-ALL-NEXT: vmovdqa 96(%rdi), %ymm6
; AVX2-FAST-ALL-NEXT: vmovdqu %ymm6, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-FAST-ALL-NEXT: vmovdqu %ymm6, (%rsp) # 32-byte Spill
; AVX2-FAST-ALL-NEXT: vmovdqa (%rdi), %xmm2
; AVX2-FAST-ALL-NEXT: vmovdqa %xmm2, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; AVX2-FAST-ALL-NEXT: vmovdqa 16(%rdi), %xmm11
@ -793,7 +793,7 @@ define void @vf32(<128 x i16>* %in.vec, <32 x i16>* %out.vec0, <32 x i16>* %out.
; AVX2-FAST-ALL-NEXT: vmovdqu %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-FAST-ALL-NEXT: vmovdqa 176(%rdi), %xmm13
; AVX2-FAST-ALL-NEXT: vpblendw {{.*#+}} xmm5 = xmm13[0],xmm8[1,2,3],xmm13[4],xmm8[5,6,7]
; AVX2-FAST-ALL-NEXT: vmovdqa %xmm13, (%rsp) # 16-byte Spill
; AVX2-FAST-ALL-NEXT: vmovdqa %xmm13, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; AVX2-FAST-ALL-NEXT: vmovdqa 160(%rdi), %xmm0
; AVX2-FAST-ALL-NEXT: vmovdqa %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; AVX2-FAST-ALL-NEXT: vpblendw {{.*#+}} xmm7 = xmm0[0],xmm8[1,2,3],xmm0[4],xmm8[5,6,7]
@ -850,7 +850,7 @@ define void @vf32(<128 x i16>* %in.vec, <32 x i16>* %out.vec0, <32 x i16>* %out.
; AVX2-FAST-ALL-NEXT: vpblendd {{.*#+}} ymm0 = ymm0[0,1,2,3],ymm1[4,5,6,7]
; AVX2-FAST-ALL-NEXT: vmovdqu %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-FAST-ALL-NEXT: vmovdqa {{.*#+}} ymm1 = [1,3,2,3,1,3,5,7]
; AVX2-FAST-ALL-NEXT: vpermd {{[-0-9]+}}(%r{{[sb]}}p), %ymm1, %ymm6 # 32-byte Folded Reload
; AVX2-FAST-ALL-NEXT: vpermd (%rsp), %ymm1, %ymm6 # 32-byte Folded Reload
; AVX2-FAST-ALL-NEXT: vpermd {{[-0-9]+}}(%r{{[sb]}}p), %ymm1, %ymm14 # 32-byte Folded Reload
; AVX2-FAST-ALL-NEXT: vpshufb {{.*#+}} ymm0 = ymm6[u,u,u,u,u,u,u,u,u,u,u,u,u,u,u,u,u,u,u,u,u,u,u,u,16,17,20,21,24,25,28,29]
; AVX2-FAST-ALL-NEXT: vmovdqa {{.*#+}} ymm2 = <0,1,4,5,8,9,12,13,8,9,12,13,12,13,14,15,16,17,20,21,24,25,28,29,u,u,u,u,u,u,u,u>
@ -869,13 +869,13 @@ define void @vf32(<128 x i16>* %in.vec, <32 x i16>* %out.vec0, <32 x i16>* %out.
; AVX2-FAST-ALL-NEXT: vpunpckldq {{.*#+}} xmm0 = xmm15[0],xmm0[0],xmm15[1],xmm0[1]
; AVX2-FAST-ALL-NEXT: vpblendd {{.*#+}} xmm0 = xmm0[0,1],xmm3[2,3]
; AVX2-FAST-ALL-NEXT: vpblendd {{.*#+}} ymm0 = ymm0[0,1,2,3],ymm5[4,5,6,7]
; AVX2-FAST-ALL-NEXT: vmovdqu %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-FAST-ALL-NEXT: vmovdqu %ymm0, (%rsp) # 32-byte Spill
; AVX2-FAST-ALL-NEXT: vpermd %ymm10, %ymm1, %ymm2
; AVX2-FAST-ALL-NEXT: vpermd {{[-0-9]+}}(%r{{[sb]}}p), %ymm1, %ymm10 # 32-byte Folded Reload
; AVX2-FAST-ALL-NEXT: vpshufb {{.*#+}} ymm5 = ymm2[u,u,u,u,u,u,u,u,u,u,u,u,u,u,u,u,u,u,u,u,u,u,u,u,16,17,20,21,24,25,28,29]
; AVX2-FAST-ALL-NEXT: vpshufb {{.*#+}} ymm8 = ymm10[0,1,4,5,8,9,12,13,8,9,12,13,12,13,14,15,16,17,20,21,24,25,28,29,u,u,u,u,u,u,u,u]
; AVX2-FAST-ALL-NEXT: vpblendd {{.*#+}} ymm8 = ymm8[0,1,2,3,4,5],ymm5[6,7]
; AVX2-FAST-ALL-NEXT: vpshufd $231, (%rsp), %xmm3 # 16-byte Folded Reload
; AVX2-FAST-ALL-NEXT: vpshufd $231, {{[-0-9]+}}(%r{{[sb]}}p), %xmm3 # 16-byte Folded Reload
; AVX2-FAST-ALL-NEXT: # xmm3 = mem[3,1,2,3]
; AVX2-FAST-ALL-NEXT: vpshufd {{.*#+}} xmm13 = xmm13[3,1,2,3]
; AVX2-FAST-ALL-NEXT: vpshuflw {{.*#+}} xmm1 = xmm3[0,1,2,0,4,5,6,7]
@ -924,17 +924,17 @@ define void @vf32(<128 x i16>* %in.vec, <32 x i16>* %out.vec0, <32 x i16>* %out.
; AVX2-FAST-ALL-NEXT: vmovaps %ymm2, (%rdx)
; AVX2-FAST-ALL-NEXT: vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm2 # 32-byte Reload
; AVX2-FAST-ALL-NEXT: vmovaps %ymm2, 32(%rcx)
; AVX2-FAST-ALL-NEXT: vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm2 # 32-byte Reload
; AVX2-FAST-ALL-NEXT: vmovups (%rsp), %ymm2 # 32-byte Reload
; AVX2-FAST-ALL-NEXT: vmovaps %ymm2, (%rcx)
; AVX2-FAST-ALL-NEXT: vmovdqa %ymm1, 32(%r8)
; AVX2-FAST-ALL-NEXT: vmovdqa %ymm0, (%r8)
; AVX2-FAST-ALL-NEXT: addq $216, %rsp
; AVX2-FAST-ALL-NEXT: addq $200, %rsp
; AVX2-FAST-ALL-NEXT: vzeroupper
; AVX2-FAST-ALL-NEXT: retq
;
; AVX2-FAST-PERLANE-LABEL: vf32:
; AVX2-FAST-PERLANE: # %bb.0:
; AVX2-FAST-PERLANE-NEXT: subq $280, %rsp # imm = 0x118
; AVX2-FAST-PERLANE-NEXT: subq $248, %rsp
; AVX2-FAST-PERLANE-NEXT: vmovdqa (%rdi), %xmm6
; AVX2-FAST-PERLANE-NEXT: vmovdqa %xmm6, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; AVX2-FAST-PERLANE-NEXT: vmovdqa 16(%rdi), %xmm1
@ -991,7 +991,7 @@ define void @vf32(<128 x i16>* %in.vec, <32 x i16>* %out.vec0, <32 x i16>* %out.
; AVX2-FAST-PERLANE-NEXT: vpblendw {{.*#+}} xmm2 = xmm1[0],xmm0[1,2,3],xmm1[4],xmm0[5,6,7]
; AVX2-FAST-PERLANE-NEXT: vmovdqa 224(%rdi), %xmm13
; AVX2-FAST-PERLANE-NEXT: vpblendw {{.*#+}} xmm3 = xmm13[0],xmm0[1,2,3],xmm13[4],xmm0[5,6,7]
; AVX2-FAST-PERLANE-NEXT: vmovdqa %xmm13, (%rsp) # 16-byte Spill
; AVX2-FAST-PERLANE-NEXT: vmovdqa %xmm13, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; AVX2-FAST-PERLANE-NEXT: vpackusdw %xmm2, %xmm3, %xmm7
; AVX2-FAST-PERLANE-NEXT: vmovdqa 208(%rdi), %xmm3
; AVX2-FAST-PERLANE-NEXT: vpblendw {{.*#+}} xmm1 = xmm3[0],xmm0[1,2,3],xmm3[4],xmm0[5,6,7]
@ -1061,7 +1061,7 @@ define void @vf32(<128 x i16>* %in.vec, <32 x i16>* %out.vec0, <32 x i16>* %out.
; AVX2-FAST-PERLANE-NEXT: vpshufd {{.*#+}} xmm1 = xmm12[3,1,2,3]
; AVX2-FAST-PERLANE-NEXT: vmovdqa %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; AVX2-FAST-PERLANE-NEXT: vpshufd {{.*#+}} xmm4 = xmm14[3,1,2,3]
; AVX2-FAST-PERLANE-NEXT: vmovdqa %xmm4, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; AVX2-FAST-PERLANE-NEXT: vmovdqa %xmm4, (%rsp) # 16-byte Spill
; AVX2-FAST-PERLANE-NEXT: vpshuflw {{.*#+}} xmm1 = xmm1[2,0,2,3,4,5,6,7]
; AVX2-FAST-PERLANE-NEXT: vpshuflw {{.*#+}} xmm4 = xmm4[2,0,2,3,4,5,6,7]
; AVX2-FAST-PERLANE-NEXT: vpunpckldq {{.*#+}} xmm1 = xmm4[0],xmm1[0],xmm4[1],xmm1[1]
@ -1084,7 +1084,7 @@ define void @vf32(<128 x i16>* %in.vec, <32 x i16>* %out.vec0, <32 x i16>* %out.
; AVX2-FAST-PERLANE-NEXT: vpblendd {{.*#+}} ymm0 = ymm0[0,1,2,3],ymm11[4,5,6,7]
; AVX2-FAST-PERLANE-NEXT: vmovdqu %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-FAST-PERLANE-NEXT: vpshufd {{.*#+}} xmm8 = xmm7[3,1,2,3]
; AVX2-FAST-PERLANE-NEXT: vpshufd $231, (%rsp), %xmm9 # 16-byte Folded Reload
; AVX2-FAST-PERLANE-NEXT: vpshufd $231, {{[-0-9]+}}(%r{{[sb]}}p), %xmm9 # 16-byte Folded Reload
; AVX2-FAST-PERLANE-NEXT: # xmm9 = mem[3,1,2,3]
; AVX2-FAST-PERLANE-NEXT: vpshuflw {{.*#+}} xmm0 = xmm8[0,1,2,0,4,5,6,7]
; AVX2-FAST-PERLANE-NEXT: vpshuflw {{.*#+}} xmm2 = xmm9[0,1,2,0,4,5,6,7]
@ -1119,7 +1119,7 @@ define void @vf32(<128 x i16>* %in.vec, <32 x i16>* %out.vec0, <32 x i16>* %out.
; AVX2-FAST-PERLANE-NEXT: vpunpckldq {{.*#+}} xmm0 = xmm6[0],xmm0[0],xmm6[1],xmm0[1]
; AVX2-FAST-PERLANE-NEXT: vpshuflw $231, {{[-0-9]+}}(%r{{[sb]}}p), %xmm6 # 16-byte Folded Reload
; AVX2-FAST-PERLANE-NEXT: # xmm6 = mem[3,1,2,3,4,5,6,7]
; AVX2-FAST-PERLANE-NEXT: vpshuflw $231, {{[-0-9]+}}(%r{{[sb]}}p), %xmm7 # 16-byte Folded Reload
; AVX2-FAST-PERLANE-NEXT: vpshuflw $231, (%rsp), %xmm7 # 16-byte Folded Reload
; AVX2-FAST-PERLANE-NEXT: # xmm7 = mem[3,1,2,3,4,5,6,7]
; AVX2-FAST-PERLANE-NEXT: vpunpckldq {{.*#+}} xmm6 = xmm7[0],xmm6[0],xmm7[1],xmm6[1]
; AVX2-FAST-PERLANE-NEXT: vinserti128 $1, %xmm0, %ymm0, %ymm0
@ -1163,7 +1163,7 @@ define void @vf32(<128 x i16>* %in.vec, <32 x i16>* %out.vec0, <32 x i16>* %out.
; AVX2-FAST-PERLANE-NEXT: vmovaps %ymm2, (%rcx)
; AVX2-FAST-PERLANE-NEXT: vmovdqa %ymm1, 32(%r8)
; AVX2-FAST-PERLANE-NEXT: vmovdqa %ymm0, (%r8)
; AVX2-FAST-PERLANE-NEXT: addq $280, %rsp # imm = 0x118
; AVX2-FAST-PERLANE-NEXT: addq $248, %rsp
; AVX2-FAST-PERLANE-NEXT: vzeroupper
; AVX2-FAST-PERLANE-NEXT: retq
%wide.vec = load <128 x i16>, <128 x i16>* %in.vec, align 32

24
test/CodeGen/X86/vzero-excess.ll
View File

@ -8,16 +8,16 @@ define <4 x float> @zeroupper_v4f32(<8 x float> *%x, <8 x float> %y) nounwind {
; CHECK-LABEL: zeroupper_v4f32:
; CHECK: # %bb.0:
; CHECK-NEXT: pushq %rbx
; CHECK-NEXT: subq $48, %rsp
; CHECK-NEXT: subq $32, %rsp
; CHECK-NEXT: vmovups %ymm0, (%rsp) # 32-byte Spill
; CHECK-NEXT: movq %rdi, %rbx
; CHECK-NEXT: vzeroupper
; CHECK-NEXT: callq the_unknown
; CHECK-NEXT: callq the_unknown@PLT
; CHECK-NEXT: vmovups (%rsp), %ymm0 # 32-byte Reload
; CHECK-NEXT: vaddps (%rbx), %ymm0, %ymm0
; CHECK-NEXT: vextractf128 $1, %ymm0, %xmm1
; CHECK-NEXT: vaddps %xmm1, %xmm0, %xmm0
; CHECK-NEXT: addq $48, %rsp
; CHECK-NEXT: addq $32, %rsp
; CHECK-NEXT: popq %rbx
; CHECK-NEXT: vzeroupper
; CHECK-NEXT: retq
@ -34,12 +34,12 @@ define <4 x float> @zeroupper_v4f32(<8 x float> *%x, <8 x float> %y) nounwind {
define <8 x float> @zeroupper_v8f32(<8 x float> %x) nounwind {
; CHECK-LABEL: zeroupper_v8f32:
; CHECK: # %bb.0:
; CHECK-NEXT: subq $56, %rsp
; CHECK-NEXT: subq $40, %rsp
; CHECK-NEXT: vmovups %ymm0, (%rsp) # 32-byte Spill
; CHECK-NEXT: vzeroupper
; CHECK-NEXT: callq the_unknown
; CHECK-NEXT: callq the_unknown@PLT
; CHECK-NEXT: vmovups (%rsp), %ymm0 # 32-byte Reload
; CHECK-NEXT: addq $56, %rsp
; CHECK-NEXT: addq $40, %rsp
; CHECK-NEXT: retq
call void @llvm.x86.avx.vzeroupper()
call void @the_unknown()
@ -50,16 +50,16 @@ define <4 x float> @zeroall_v4f32(<8 x float> *%x, <8 x float> %y) nounwind {
; CHECK-LABEL: zeroall_v4f32:
; CHECK: # %bb.0:
; CHECK-NEXT: pushq %rbx
; CHECK-NEXT: subq $48, %rsp
; CHECK-NEXT: subq $32, %rsp
; CHECK-NEXT: vmovups %ymm0, (%rsp) # 32-byte Spill
; CHECK-NEXT: movq %rdi, %rbx
; CHECK-NEXT: vzeroall
; CHECK-NEXT: callq the_unknown
; CHECK-NEXT: callq the_unknown@PLT
; CHECK-NEXT: vmovups (%rsp), %ymm0 # 32-byte Reload
; CHECK-NEXT: vaddps (%rbx), %ymm0, %ymm0
; CHECK-NEXT: vextractf128 $1, %ymm0, %xmm1
; CHECK-NEXT: vaddps %xmm1, %xmm0, %xmm0
; CHECK-NEXT: addq $48, %rsp
; CHECK-NEXT: addq $32, %rsp
; CHECK-NEXT: popq %rbx
; CHECK-NEXT: vzeroupper
; CHECK-NEXT: retq
@ -76,12 +76,12 @@ define <4 x float> @zeroall_v4f32(<8 x float> *%x, <8 x float> %y) nounwind {
define <8 x float> @zeroall_v8f32(<8 x float> %x) nounwind {
; CHECK-LABEL: zeroall_v8f32:
; CHECK: # %bb.0:
; CHECK-NEXT: subq $56, %rsp
; CHECK-NEXT: subq $40, %rsp
; CHECK-NEXT: vmovups %ymm0, (%rsp) # 32-byte Spill
; CHECK-NEXT: vzeroall
; CHECK-NEXT: callq the_unknown
; CHECK-NEXT: callq the_unknown@PLT
; CHECK-NEXT: vmovups (%rsp), %ymm0 # 32-byte Reload
; CHECK-NEXT: addq $56, %rsp
; CHECK-NEXT: addq $40, %rsp
; CHECK-NEXT: retq
call void @llvm.x86.avx.vzeroall()
call void @the_unknown()