caio/clang-p2996
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
Files
4794bbffb228894de3ca0dad50fc027eb94f743e
clang-p2996/llvm/test/CodeGen/AMDGPU/anyext.ll
Austin Kerbow da067ed569 [AMDGPU] Set most sched model resource's BufferSize to one 
Using a BufferSize of one for memory ProcResources will result in better
ILP since it more accurately models the dependencies between memory ops
and their consumers on an in-order processor. After this change, the
scheduler will treat the data edges from loads as blocking so that
stalls are guaranteed when waiting for data to be retreaved from memory.
Since we don't actually track waitcnt here, this should do a better job
at modeling their behavior.

Practically, this means that the scheduler will trigger the 'STALL'
heuristic more often.

This type of change needs to be evaluated experimentally. Preliminary
results are positive.

Fixes: SWDEV-282962

Reviewed By: rampitec

Differential Revision: https://reviews.llvm.org/D114777
2021-12-01 22:31:28 -08:00

204 lines
7.8 KiB
LLVM
Raw Blame History

; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc -march=amdgcn -mcpu=verde -verify-machineinstrs < %s | FileCheck --check-prefix=GCN %s
; RUN: llc -march=amdgcn -mcpu=tonga -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX8 %s
; RUN: llc -march=amdgcn -mcpu=gfx900 -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX9 %s
declare i32 @llvm.amdgcn.workitem.id.x() nounwind readnone
declare i32 @llvm.amdgcn.workitem.id.y() nounwind readnone
define amdgpu_kernel void @anyext_i1_i32(i32 addrspace(1)* %out, i32 %cond) #0 {
; GCN-LABEL: anyext_i1_i32:
; GCN: ; %bb.0: ; %entry
; GCN-NEXT: s_load_dword s4, s[0:1], 0xb
; GCN-NEXT: s_load_dwordx2 s[0:1], s[0:1], 0x9
; GCN-NEXT: s_mov_b32 s3, 0xf000
; GCN-NEXT: s_mov_b32 s2, -1
; GCN-NEXT: s_waitcnt lgkmcnt(0)
; GCN-NEXT: s_cmp_lg_u32 s4, 0
; GCN-NEXT: s_cselect_b64 s[4:5], -1, 0
; GCN-NEXT: v_cndmask_b32_e64 v0, 0, 1, s[4:5]
; GCN-NEXT: buffer_store_dword v0, off, s[0:3], 0
; GCN-NEXT: s_endpgm
;
; GFX8-LABEL: anyext_i1_i32:
; GFX8: ; %bb.0: ; %entry
; GFX8-NEXT: s_load_dword s4, s[0:1], 0x2c
; GFX8-NEXT: s_load_dwordx2 s[0:1], s[0:1], 0x24
; GFX8-NEXT: s_mov_b32 s3, 0xf000
; GFX8-NEXT: s_mov_b32 s2, -1
; GFX8-NEXT: s_waitcnt lgkmcnt(0)
; GFX8-NEXT: s_cmp_eq_u32 s4, 0
; GFX8-NEXT: s_cselect_b64 s[4:5], -1, 0
; GFX8-NEXT: v_cndmask_b32_e64 v0, 0, 1, s[4:5]
; GFX8-NEXT: v_not_b32_e32 v0, v0
; GFX8-NEXT: v_and_b32_e32 v0, 1, v0
; GFX8-NEXT: buffer_store_dword v0, off, s[0:3], 0
; GFX8-NEXT: s_endpgm
;
; GFX9-LABEL: anyext_i1_i32:
; GFX9: ; %bb.0: ; %entry
; GFX9-NEXT: s_load_dword s2, s[0:1], 0x2c
; GFX9-NEXT: s_load_dwordx2 s[4:5], s[0:1], 0x24
; GFX9-NEXT: s_mov_b32 s7, 0xf000
; GFX9-NEXT: s_mov_b32 s6, -1
; GFX9-NEXT: s_waitcnt lgkmcnt(0)
; GFX9-NEXT: s_cmp_eq_u32 s2, 0
; GFX9-NEXT: s_cselect_b64 s[0:1], -1, 0
; GFX9-NEXT: v_cndmask_b32_e64 v0, 0, 1, s[0:1]
; GFX9-NEXT: v_not_b32_e32 v0, v0
; GFX9-NEXT: v_and_b32_e32 v0, 1, v0
; GFX9-NEXT: buffer_store_dword v0, off, s[4:7], 0
; GFX9-NEXT: s_endpgm
entry:
%tmp = icmp eq i32 %cond, 0
%tmp1 = zext i1 %tmp to i8
%tmp2 = xor i8 %tmp1, -1
%tmp3 = and i8 %tmp2, 1
%tmp4 = zext i8 %tmp3 to i32
store i32 %tmp4, i32 addrspace(1)* %out
ret void
}
define amdgpu_kernel void @s_anyext_i16_i32(i32 addrspace(1)* %out, i16 addrspace(1)* %a, i16 addrspace(1)* %b) #0 {
; GCN-LABEL: s_anyext_i16_i32:
; GCN: ; %bb.0: ; %entry
; GCN-NEXT: s_load_dwordx4 s[4:7], s[0:1], 0x9
; GCN-NEXT: s_load_dwordx2 s[0:1], s[0:1], 0xd
; GCN-NEXT: s_mov_b32 s11, 0xf000
; GCN-NEXT: s_mov_b32 s14, 0
; GCN-NEXT: s_mov_b32 s15, s11
; GCN-NEXT: s_waitcnt lgkmcnt(0)
; GCN-NEXT: s_mov_b64 s[12:13], s[6:7]
; GCN-NEXT: v_lshlrev_b32_e32 v2, 1, v0
; GCN-NEXT: v_mov_b32_e32 v3, 0
; GCN-NEXT: v_lshlrev_b32_e32 v0, 1, v1
; GCN-NEXT: s_mov_b64 s[2:3], s[14:15]
; GCN-NEXT: v_mov_b32_e32 v1, v3
; GCN-NEXT: buffer_load_ushort v2, v[2:3], s[12:15], 0 addr64
; GCN-NEXT: buffer_load_ushort v0, v[0:1], s[0:3], 0 addr64
; GCN-NEXT: s_mov_b32 s10, -1
; GCN-NEXT: s_mov_b32 s8, s4
; GCN-NEXT: s_mov_b32 s9, s5
; GCN-NEXT: s_waitcnt vmcnt(0)
; GCN-NEXT: v_add_i32_e32 v0, vcc, v2, v0
; GCN-NEXT: v_not_b32_e32 v0, v0
; GCN-NEXT: v_and_b32_e32 v0, 1, v0
; GCN-NEXT: buffer_store_dword v0, off, s[8:11], 0
; GCN-NEXT: s_endpgm
;
; GFX8-LABEL: s_anyext_i16_i32:
; GFX8: ; %bb.0: ; %entry
; GFX8-NEXT: s_load_dwordx4 s[4:7], s[0:1], 0x24
; GFX8-NEXT: s_load_dwordx2 s[0:1], s[0:1], 0x34
; GFX8-NEXT: v_lshlrev_b32_e32 v0, 1, v0
; GFX8-NEXT: s_waitcnt lgkmcnt(0)
; GFX8-NEXT: v_mov_b32_e32 v3, s7
; GFX8-NEXT: v_add_u32_e32 v2, vcc, s6, v0
; GFX8-NEXT: v_addc_u32_e32 v3, vcc, 0, v3, vcc
; GFX8-NEXT: v_lshlrev_b32_e32 v0, 1, v1
; GFX8-NEXT: v_mov_b32_e32 v1, s1
; GFX8-NEXT: v_add_u32_e32 v0, vcc, s0, v0
; GFX8-NEXT: v_addc_u32_e32 v1, vcc, 0, v1, vcc
; GFX8-NEXT: flat_load_ushort v2, v[2:3]
; GFX8-NEXT: flat_load_ushort v0, v[0:1]
; GFX8-NEXT: s_mov_b32 s7, 0xf000
; GFX8-NEXT: s_mov_b32 s6, -1
; GFX8-NEXT: s_waitcnt vmcnt(0)
; GFX8-NEXT: v_add_u16_e32 v0, v2, v0
; GFX8-NEXT: v_xor_b32_e32 v0, -1, v0
; GFX8-NEXT: v_and_b32_e32 v0, 1, v0
; GFX8-NEXT: v_and_b32_e32 v0, 0xffff, v0
; GFX8-NEXT: buffer_store_dword v0, off, s[4:7], 0
; GFX8-NEXT: s_endpgm
;
; GFX9-LABEL: s_anyext_i16_i32:
; GFX9: ; %bb.0: ; %entry
; GFX9-NEXT: s_load_dwordx4 s[4:7], s[0:1], 0x24
; GFX9-NEXT: s_load_dwordx2 s[2:3], s[0:1], 0x34
; GFX9-NEXT: v_lshlrev_b32_e32 v0, 1, v0
; GFX9-NEXT: v_lshlrev_b32_e32 v1, 1, v1
; GFX9-NEXT: s_waitcnt lgkmcnt(0)
; GFX9-NEXT: global_load_ushort v2, v0, s[6:7]
; GFX9-NEXT: global_load_ushort v3, v1, s[2:3]
; GFX9-NEXT: s_mov_b32 s7, 0xf000
; GFX9-NEXT: s_mov_b32 s6, -1
; GFX9-NEXT: s_waitcnt vmcnt(0)
; GFX9-NEXT: v_add_u16_e32 v0, v2, v3
; GFX9-NEXT: v_xor_b32_e32 v0, -1, v0
; GFX9-NEXT: v_and_b32_e32 v0, 1, v0
; GFX9-NEXT: v_and_b32_e32 v0, 0xffff, v0
; GFX9-NEXT: buffer_store_dword v0, off, s[4:7], 0
; GFX9-NEXT: s_endpgm
entry:
%tid.x = call i32 @llvm.amdgcn.workitem.id.x()
%tid.y = call i32 @llvm.amdgcn.workitem.id.y()
%a.ptr = getelementptr i16, i16 addrspace(1)* %a, i32 %tid.x
%b.ptr = getelementptr i16, i16 addrspace(1)* %b, i32 %tid.y
%a.l = load i16, i16 addrspace(1)* %a.ptr
%b.l = load i16, i16 addrspace(1)* %b.ptr
%tmp = add i16 %a.l, %b.l
%tmp1 = trunc i16 %tmp to i8
%tmp2 = xor i8 %tmp1, -1
%tmp3 = and i8 %tmp2, 1
%tmp4 = zext i8 %tmp3 to i32
store i32 %tmp4, i32 addrspace(1)* %out
ret void
}
define amdgpu_kernel void @anyext_v2i16_to_v2i32() #0 {
; GCN-LABEL: anyext_v2i16_to_v2i32:
; GCN: ; %bb.0: ; %bb
; GCN-NEXT: s_mov_b32 s3, 0xf000
; GCN-NEXT: s_mov_b32 s2, -1
; GCN-NEXT: buffer_load_ushort v0, off, s[0:3], 0
; GCN-NEXT: s_waitcnt vmcnt(0)
; GCN-NEXT: v_and_b32_e32 v0, 0x8000, v0
; GCN-NEXT: v_lshlrev_b32_e32 v0, 16, v0
; GCN-NEXT: v_cmp_eq_f32_e32 vcc, 0, v0
; GCN-NEXT: v_cndmask_b32_e64 v0, 0, 1, vcc
; GCN-NEXT: buffer_store_byte v0, off, s[0:3], 0
; GCN-NEXT: s_endpgm
;
; GFX8-LABEL: anyext_v2i16_to_v2i32:
; GFX8: ; %bb.0: ; %bb
; GFX8-NEXT: s_mov_b32 s3, 0xf000
; GFX8-NEXT: s_mov_b32 s2, -1
; GFX8-NEXT: buffer_load_ushort v0, off, s[0:3], 0
; GFX8-NEXT: v_mov_b32_e32 v1, 0x8000
; GFX8-NEXT: s_waitcnt vmcnt(0)
; GFX8-NEXT: v_and_b32_sdwa v0, v0, v1 dst_sel:WORD_1 dst_unused:UNUSED_PAD src0_sel:DWORD src1_sel:DWORD
; GFX8-NEXT: v_cmp_eq_f32_e32 vcc, 0, v0
; GFX8-NEXT: v_cndmask_b32_e64 v0, 0, 1, vcc
; GFX8-NEXT: buffer_store_byte v0, off, s[0:3], 0
; GFX8-NEXT: s_endpgm
;
; GFX9-LABEL: anyext_v2i16_to_v2i32:
; GFX9: ; %bb.0: ; %bb
; GFX9-NEXT: global_load_short_d16_hi v0, v[0:1], off
; GFX9-NEXT: v_mov_b32_e32 v1, 0xffff
; GFX9-NEXT: s_mov_b32 s3, 0xf000
; GFX9-NEXT: s_mov_b32 s2, -1
; GFX9-NEXT: s_waitcnt vmcnt(0)
; GFX9-NEXT: v_and_b32_e32 v0, 0x80008000, v0
; GFX9-NEXT: v_bfi_b32 v0, v1, 0, v0
; GFX9-NEXT: v_cmp_eq_f32_e32 vcc, 0, v0
; GFX9-NEXT: v_cndmask_b32_e64 v0, 0, 1, vcc
; GFX9-NEXT: buffer_store_byte v0, off, s[0:3], 0
; GFX9-NEXT: s_endpgm
bb:
%tmp = load i16, i16 addrspace(1)* undef, align 2
%tmp2 = insertelement <2 x i16> undef, i16 %tmp, i32 1
%tmp4 = and <2 x i16> %tmp2, <i16 -32768, i16 -32768>
%tmp5 = zext <2 x i16> %tmp4 to <2 x i32>
%tmp6 = shl nuw <2 x i32> %tmp5, <i32 16, i32 16>
%tmp7 = or <2 x i32> zeroinitializer, %tmp6
%tmp8 = bitcast <2 x i32> %tmp7 to <2 x float>
%tmp10 = fcmp oeq <2 x float> %tmp8, zeroinitializer
%tmp11 = zext <2 x i1> %tmp10 to <2 x i8>
%tmp12 = extractelement <2 x i8> %tmp11, i32 1
store i8 %tmp12, i8 addrspace(1)* undef, align 1
ret void
}
attributes #0 = { nounwind }
Reference in New Issue View Git Blame Copy Permalink