1
0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2024-11-22 18:54:02 +01:00
llvm-mirror/test/CodeGen/AMDGPU/llvm.amdgcn.div.fmas.ll
Nicolai Haehnle 734b898f0d AMDGPU: Rewrite SILowerI1Copies to always stay on SALU
Summary:
Instead of writing boolean values temporarily into 32-bit VGPRs
if they are involved in PHIs or are observed from outside a loop,
we use bitwise masking operations to combine lane masks in a way
that is consistent with wave control flow.

Move SIFixSGPRCopies to before this pass, since that pass
incorrectly attempts to move SGPR phis to VGPRs.

This should recover most of the code quality that was lost with
the bug fix in "AMDGPU: Remove PHI loop condition optimization".

There are still some relevant cases where code quality could be
improved, in particular:

- We often introduce redundant masks with EXEC. Ideally, we'd
  have a generic computeKnownBits-like analysis to determine
  whether masks are already masked by EXEC, so we can avoid this
  masking both here and when lowering uniform control flow.

- The criterion we use to determine whether a def is observed
  from outside a loop is conservative: it doesn't check whether
  (loop) branch conditions are uniform.

Change-Id: Ibabdb373a7510e426b90deef00f5e16c5d56e64b

Reviewers: arsenm, rampitec, tpr

Subscribers: kzhuravl, jvesely, wdng, mgorny, yaxunl, dstuttard, t-tye, eraman, llvm-commits

Differential Revision: https://reviews.llvm.org/D53496

llvm-svn: 345719
2018-10-31 13:27:08 +00:00

191 lines
8.9 KiB
LLVM

; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -verify-machineinstrs < %s | FileCheck -enable-var-scope -check-prefixes=GCN,SI %s
; XUN: llc -march=amdgcn -mcpu=tonga -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -enable-var-scope -check-prefixes=GCN,VI %s
; FIXME: Enable for VI.
declare i32 @llvm.amdgcn.workitem.id.x() nounwind readnone
declare float @llvm.amdgcn.div.fmas.f32(float, float, float, i1) nounwind readnone
declare double @llvm.amdgcn.div.fmas.f64(double, double, double, i1) nounwind readnone
; GCN-LABEL: {{^}}test_div_fmas_f32:
; SI-DAG: s_load_dword [[SA:s[0-9]+]], s{{\[[0-9]+:[0-9]+\]}}, 0x13
; SI-DAG: s_load_dword [[SB:s[0-9]+]], s{{\[[0-9]+:[0-9]+\]}}, 0x1c
; SI-DAG: s_load_dword [[SC:s[0-9]+]], s{{\[[0-9]+:[0-9]+\]}}, 0x25
; VI-DAG: s_load_dword [[SA:s[0-9]+]], s{{\[[0-9]+:[0-9]+\]}}, 0x4c
; VI-DAG: s_load_dword [[SB:s[0-9]+]], s{{\[[0-9]+:[0-9]+\]}}, 0x70
; VI-DAG: s_load_dword [[SC:s[0-9]+]], s{{\[[0-9]+:[0-9]+\]}}, 0x94
; GCN-DAG: s_and_b32 [[AND_I1:s[0-9]+]], 1, s{{[0-9]+}}
; GCN: v_cmp_eq_u32_e64 vcc, [[AND_I1]], 1
; GCN-DAG: v_mov_b32_e32 [[VC:v[0-9]+]], [[SC]]
; GCN-DAG: v_mov_b32_e32 [[VB:v[0-9]+]], [[SB]]
; GCN-DAG: v_mov_b32_e32 [[VA:v[0-9]+]], [[SA]]
; GCN: v_div_fmas_f32 [[RESULT:v[0-9]+]], [[VA]], [[VB]], [[VC]]
; GCN: buffer_store_dword [[RESULT]],
define amdgpu_kernel void @test_div_fmas_f32(float addrspace(1)* %out, [8 x i32], float %a, [8 x i32], float %b, [8 x i32], float %c, [8 x i32], i1 %d) nounwind {
%result = call float @llvm.amdgcn.div.fmas.f32(float %a, float %b, float %c, i1 %d) nounwind readnone
store float %result, float addrspace(1)* %out, align 4
ret void
}
; GCN-LABEL: {{^}}test_div_fmas_f32_inline_imm_0:
; SI-DAG: s_load_dword [[SB:s[0-9]+]], s{{\[[0-9]+:[0-9]+\]}}, 0x1c
; SI-DAG: s_load_dword [[SC:s[0-9]+]], s{{\[[0-9]+:[0-9]+\]}}, 0x25
; SI-DAG: v_mov_b32_e32 [[VC:v[0-9]+]], [[SC]]
; SI-DAG: v_mov_b32_e32 [[VB:v[0-9]+]], [[SB]]
; SI: v_div_fmas_f32 [[RESULT:v[0-9]+]], 1.0, [[VB]], [[VC]]
; SI: buffer_store_dword [[RESULT]],
define amdgpu_kernel void @test_div_fmas_f32_inline_imm_0(float addrspace(1)* %out, [8 x i32], float %a, [8 x i32], float %b, [8 x i32], float %c, [8 x i32], i1 %d) nounwind {
%result = call float @llvm.amdgcn.div.fmas.f32(float 1.0, float %b, float %c, i1 %d) nounwind readnone
store float %result, float addrspace(1)* %out, align 4
ret void
}
; GCN-LABEL: {{^}}test_div_fmas_f32_inline_imm_1:
; SI-DAG: s_load_dword [[SA:s[0-9]+]], s{{\[[0-9]+:[0-9]+\]}}, 0xb
; SI-DAG: s_load_dword [[SC:s[0-9]+]], s{{\[[0-9]+:[0-9]+\]}}, 0xd
; VI-DAG: s_load_dword [[SA:s[0-9]+]], s{{\[[0-9]+:[0-9]+\]}}, 0x4c
; VI-DAG: s_load_dword [[SC:s[0-9]+]], s{{\[[0-9]+:[0-9]+\]}}, 0x94
; GCN-DAG: v_mov_b32_e32 [[VC:v[0-9]+]], [[SC]]
; GCN-DAG: v_mov_b32_e32 [[VA:v[0-9]+]], [[SA]]
; GCN: v_div_fmas_f32 [[RESULT:v[0-9]+]], [[VA]], 1.0, [[VC]]
; GCN: buffer_store_dword [[RESULT]],
define amdgpu_kernel void @test_div_fmas_f32_inline_imm_1(float addrspace(1)* %out, float %a, float %b, float %c, [8 x i32], i1 %d) nounwind {
%result = call float @llvm.amdgcn.div.fmas.f32(float %a, float 1.0, float %c, i1 %d) nounwind readnone
store float %result, float addrspace(1)* %out, align 4
ret void
}
; GCN-LABEL: {{^}}test_div_fmas_f32_inline_imm_2:
; SI-DAG: s_load_dword [[SA:s[0-9]+]], s{{\[[0-9]+:[0-9]+\]}}, 0x13
; SI-DAG: s_load_dword [[SB:s[0-9]+]], s{{\[[0-9]+:[0-9]+\]}}, 0x1c
; VI-DAG: s_load_dword [[SA:s[0-9]+]], s{{\[[0-9]+:[0-9]+\]}}, 0x4c
; VI-DAG: s_load_dword [[SB:s[0-9]+]], s{{\[[0-9]+:[0-9]+\]}}, 0x70
; GCN-DAG: v_mov_b32_e32 [[VA:v[0-9]+]], [[SA]]
; GCN-DAG: v_mov_b32_e32 [[VB:v[0-9]+]], [[SB]]
; GCN: v_div_fmas_f32 [[RESULT:v[0-9]+]], [[VA]], [[VB]], 1.0
; GCN: buffer_store_dword [[RESULT]],
define amdgpu_kernel void @test_div_fmas_f32_inline_imm_2(float addrspace(1)* %out, [8 x i32], float %a, [8 x i32], float %b, [8 x i32], float %c, [8 x i32], i1 %d) nounwind {
%result = call float @llvm.amdgcn.div.fmas.f32(float %a, float %b, float 1.0, i1 %d) nounwind readnone
store float %result, float addrspace(1)* %out, align 4
ret void
}
; GCN-LABEL: {{^}}test_div_fmas_f64:
; GCN: v_div_fmas_f64
define amdgpu_kernel void @test_div_fmas_f64(double addrspace(1)* %out, double %a, double %b, double %c, i1 %d) nounwind {
%result = call double @llvm.amdgcn.div.fmas.f64(double %a, double %b, double %c, i1 %d) nounwind readnone
store double %result, double addrspace(1)* %out, align 8
ret void
}
; GCN-LABEL: {{^}}test_div_fmas_f32_cond_to_vcc:
; GCN: v_cmp_eq_u32_e64 vcc, s{{[0-9]+}}, 0{{$}}
; GCN: v_div_fmas_f32 {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
define amdgpu_kernel void @test_div_fmas_f32_cond_to_vcc(float addrspace(1)* %out, float %a, float %b, float %c, i32 %i) nounwind {
%cmp = icmp eq i32 %i, 0
%result = call float @llvm.amdgcn.div.fmas.f32(float %a, float %b, float %c, i1 %cmp) nounwind readnone
store float %result, float addrspace(1)* %out, align 4
ret void
}
; GCN-LABEL: {{^}}test_div_fmas_f32_imm_false_cond_to_vcc:
; GCN: s_mov_b64 vcc, 0
; GCN: v_div_fmas_f32 {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
define amdgpu_kernel void @test_div_fmas_f32_imm_false_cond_to_vcc(float addrspace(1)* %out, float %a, float %b, float %c) nounwind {
%result = call float @llvm.amdgcn.div.fmas.f32(float %a, float %b, float %c, i1 false) nounwind readnone
store float %result, float addrspace(1)* %out, align 4
ret void
}
; GCN-LABEL: {{^}}test_div_fmas_f32_imm_true_cond_to_vcc:
; GCN: s_mov_b64 vcc, -1
; GCN: v_div_fmas_f32 {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
define amdgpu_kernel void @test_div_fmas_f32_imm_true_cond_to_vcc(float addrspace(1)* %out, float %a, float %b, float %c) nounwind {
%result = call float @llvm.amdgcn.div.fmas.f32(float %a, float %b, float %c, i1 true) nounwind readnone
store float %result, float addrspace(1)* %out, align 4
ret void
}
; GCN-LABEL: {{^}}test_div_fmas_f32_logical_cond_to_vcc:
; SI-DAG: buffer_load_dword [[A:v[0-9]+]], {{v\[[0-9]+:[0-9]+\]}}, {{s\[[0-9]+:[0-9]+\]}}, 0 addr64{{$}}
; SI-DAG: buffer_load_dword [[B:v[0-9]+]], {{v\[[0-9]+:[0-9]+\]}}, {{s\[[0-9]+:[0-9]+\]}}, 0 addr64 offset:4{{$}}
; SI-DAG: buffer_load_dword [[C:v[0-9]+]], {{v\[[0-9]+:[0-9]+\]}}, {{s\[[0-9]+:[0-9]+\]}}, 0 addr64 offset:8{{$}}
; SI-DAG: v_cmp_eq_u32_e32 [[CMP0:vcc]], 0, v{{[0-9]+}}
; SI-DAG: v_cmp_ne_u32_e64 [[CMP1:s\[[0-9]+:[0-9]+\]]], s{{[0-9]+}}, 0{{$}}
; SI: s_and_b64 vcc, [[CMP0]], [[CMP1]]
; SI: v_div_fmas_f32 {{v[0-9]+}}, [[A]], [[B]], [[C]]
; SI: s_endpgm
define amdgpu_kernel void @test_div_fmas_f32_logical_cond_to_vcc(float addrspace(1)* %out, float addrspace(1)* %in, i32 %d) nounwind {
%tid = call i32 @llvm.amdgcn.workitem.id.x() nounwind readnone
%gep.a = getelementptr float, float addrspace(1)* %in, i32 %tid
%gep.b = getelementptr float, float addrspace(1)* %gep.a, i32 1
%gep.c = getelementptr float, float addrspace(1)* %gep.a, i32 2
%gep.out = getelementptr float, float addrspace(1)* %out, i32 2
%a = load volatile float, float addrspace(1)* %gep.a
%b = load volatile float, float addrspace(1)* %gep.b
%c = load volatile float, float addrspace(1)* %gep.c
%cmp0 = icmp eq i32 %tid, 0
%cmp1 = icmp ne i32 %d, 0
%and = and i1 %cmp0, %cmp1
%result = call float @llvm.amdgcn.div.fmas.f32(float %a, float %b, float %c, i1 %and) nounwind readnone
store float %result, float addrspace(1)* %gep.out, align 4
ret void
}
; GCN-LABEL: {{^}}test_div_fmas_f32_i1_phi_vcc:
; SI: ; %entry
; SI: v_cmp_eq_u32_e64 [[CMP:s\[[0-9]+:[0-9]+\]]], 0, {{v[0-9]+}}
; SI: s_mov_b64 vcc, 0
; SI: s_and_saveexec_b64 [[SAVE:s\[[0-9]+:[0-9]+\]]], [[CMP]]
; SI: ; %bb
; SI: buffer_load_dword [[LOAD:v[0-9]+]],
; SI: v_cmp_ne_u32_e32 vcc, 0, [[LOAD]]
; SI: s_and_b64 vcc, vcc, exec
; SI: ; %exit
; SI: s_or_b64 exec, exec, [[SAVE]]
; SI-NOT: vcc
; SI: v_div_fmas_f32 {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
; SI: buffer_store_dword
; SI: s_endpgm
define amdgpu_kernel void @test_div_fmas_f32_i1_phi_vcc(float addrspace(1)* %out, float addrspace(1)* %in, i32 addrspace(1)* %dummy) nounwind {
entry:
%tid = call i32 @llvm.amdgcn.workitem.id.x() nounwind readnone
%gep.out = getelementptr float, float addrspace(1)* %out, i32 2
%gep.a = getelementptr float, float addrspace(1)* %in, i32 %tid
%gep.b = getelementptr float, float addrspace(1)* %gep.a, i32 1
%gep.c = getelementptr float, float addrspace(1)* %gep.a, i32 2
%a = load float, float addrspace(1)* %gep.a
%b = load float, float addrspace(1)* %gep.b
%c = load float, float addrspace(1)* %gep.c
%cmp0 = icmp eq i32 %tid, 0
br i1 %cmp0, label %bb, label %exit
bb:
%val = load i32, i32 addrspace(1)* %dummy
%cmp1 = icmp ne i32 %val, 0
br label %exit
exit:
%cond = phi i1 [false, %entry], [%cmp1, %bb]
%result = call float @llvm.amdgcn.div.fmas.f32(float %a, float %b, float %c, i1 %cond) nounwind readnone
store float %result, float addrspace(1)* %gep.out, align 4
ret void
}