maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity
llvm-project/llvm/test/CodeGen/AMDGPU/load-local-f32.ll

145 lines
4.0 KiB
LLVM
Raw Normal View History

AMDGPU: Select DS insts without m0 initialization GFX9 stopped using m0 for most DS instructions. Select a different instruction without the use. I think this will be less error prone than trying to manually maintain m0 uses as needed. llvm-svn: 319270
2017-11-29 08:55:57 +08:00
; RUN: llc -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefixes=GCN,SICIVI,FUNC %s
; RUN: llc -march=amdgcn -mcpu=tonga -verify-machineinstrs < %s | FileCheck -check-prefixes=GCN,SICIVI,FUNC %s
; RUN: llc -march=r600 -mcpu=redwood < %s | FileCheck -check-prefixes=EG,FUNC %s
AMDGPU: Cleanup load tests There are a lot of different kinds of loads to test for, and these were scattered around inconsistently with some redundancy. Try to comprehensively test all loads in a consistent way. llvm-svn: 271571
2016-06-03 03:54:26 +08:00
[AMDGPU] Supported ds_read_b128 generation; Widened vector length for local address-space. Summary: Starting from GCN 2nd generation, ISA supports ds_read_b128 on top of ds_read_b64. This patch supports ds_read_b128 instruction pattern and generation of this instruction. In the vectorizer, this patch also widen the vector length so that vectorizer generates 128 bit loads for local address-space which gets translated to ds_read_b128. Since the performance benefit is not clear; compiler generates ds_read_b128 under -amdgpu-ds128. Author: FarhanaAleen Reviewed By: rampitec, arsenm Subscribers: llvm-commits, AMDGPU Differential Revision: https://reviews.llvm.org/D44210 llvm-svn: 327153
2018-03-10 01:41:39 +08:00
; Testing for ds_read_128
; RUN: llc -march=amdgcn -mcpu=tahiti -amdgpu-ds128 < %s | FileCheck -check-prefixes=SI,FUNC %s
; RUN: llc -march=amdgcn -mcpu=tonga -amdgpu-ds128 < %s | FileCheck -check-prefixes=CIVI,FUNC %s
; RUN: llc -march=amdgcn -mcpu=gfx900 -amdgpu-ds128 < %s | FileCheck -check-prefixes=CIVI,FUNC %s
AMDGPU: Cleanup load tests There are a lot of different kinds of loads to test for, and these were scattered around inconsistently with some redundancy. Try to comprehensively test all loads in a consistent way. llvm-svn: 271571
2016-06-03 03:54:26 +08:00
; FUNC-LABEL: {{^}}load_f32_local:
AMDGPU: Select DS insts without m0 initialization GFX9 stopped using m0 for most DS instructions. Select a different instruction without the use. I think this will be less error prone than trying to manually maintain m0 uses as needed. llvm-svn: 319270
2017-11-29 08:55:57 +08:00
; SICIVI: s_mov_b32 m0
; GFX9-NOT: m0
AMDGPU: Cleanup load tests There are a lot of different kinds of loads to test for, and these were scattered around inconsistently with some redundancy. Try to comprehensively test all loads in a consistent way. llvm-svn: 271571
2016-06-03 03:54:26 +08:00
; GCN: ds_read_b32
; EG: LDS_READ_RET
AMDGPU: Mark all unspecified CC functions in tests as amdgpu_kernel Currently the default C calling convention functions are treated the same as compute kernels. Make this explicit so the default calling convention can be changed to a non-kernel. Converted with perl -pi -e 's/define void/define amdgpu_kernel void/' on the relevant test directories (and undoing in one place that actually wanted a non-kernel). llvm-svn: 298444
2017-03-22 05:39:51 +08:00
define amdgpu_kernel void @load_f32_local(float addrspace(1)* %out, float addrspace(3)* %in) #0 {
AMDGPU: Cleanup load tests There are a lot of different kinds of loads to test for, and these were scattered around inconsistently with some redundancy. Try to comprehensively test all loads in a consistent way. llvm-svn: 271571
2016-06-03 03:54:26 +08:00
entry:
%tmp0 = load float, float addrspace(3)* %in
store float %tmp0, float addrspace(1)* %out
ret void
}
; FUNC-LABEL: {{^}}load_v2f32_local:
AMDGPU: Select DS insts without m0 initialization GFX9 stopped using m0 for most DS instructions. Select a different instruction without the use. I think this will be less error prone than trying to manually maintain m0 uses as needed. llvm-svn: 319270
2017-11-29 08:55:57 +08:00
; SICIVI: s_mov_b32 m0
; GFX9-NOT: m0
AMDGPU: Cleanup load tests There are a lot of different kinds of loads to test for, and these were scattered around inconsistently with some redundancy. Try to comprehensively test all loads in a consistent way. llvm-svn: 271571
2016-06-03 03:54:26 +08:00
; GCN: ds_read_b64
; EG: LDS_READ_RET
; EG: LDS_READ_RET
AMDGPU: Mark all unspecified CC functions in tests as amdgpu_kernel Currently the default C calling convention functions are treated the same as compute kernels. Make this explicit so the default calling convention can be changed to a non-kernel. Converted with perl -pi -e 's/define void/define amdgpu_kernel void/' on the relevant test directories (and undoing in one place that actually wanted a non-kernel). llvm-svn: 298444
2017-03-22 05:39:51 +08:00
define amdgpu_kernel void @load_v2f32_local(<2 x float> addrspace(1)* %out, <2 x float> addrspace(3)* %in) #0 {
AMDGPU: Cleanup load tests There are a lot of different kinds of loads to test for, and these were scattered around inconsistently with some redundancy. Try to comprehensively test all loads in a consistent way. llvm-svn: 271571
2016-06-03 03:54:26 +08:00
entry:
%tmp0 = load <2 x float>, <2 x float> addrspace(3)* %in
store <2 x float> %tmp0, <2 x float> addrspace(1)* %out
ret void
}
AMDGPU: Improve load/store of illegal types. There was a combine before to handle the simple copy case. Split this into handling loads and stores separately. We might want to change how this handles some of the vector extloads, since this can result in large code size increases. llvm-svn: 274394
2016-07-02 06:47:50 +08:00
; FIXME: should this do a read2_b64?
AMDGPU: Cleanup load tests There are a lot of different kinds of loads to test for, and these were scattered around inconsistently with some redundancy. Try to comprehensively test all loads in a consistent way. llvm-svn: 271571
2016-06-03 03:54:26 +08:00
; FUNC-LABEL: {{^}}local_load_v3f32:
AMDGPU: Select DS insts without m0 initialization GFX9 stopped using m0 for most DS instructions. Select a different instruction without the use. I think this will be less error prone than trying to manually maintain m0 uses as needed. llvm-svn: 319270
2017-11-29 08:55:57 +08:00
; SICIVI: s_mov_b32 m0
; GFX9-NOT: m0
AMDGPU: Improve load/store of illegal types. There was a combine before to handle the simple copy case. Split this into handling loads and stores separately. We might want to change how this handles some of the vector extloads, since this can result in large code size increases. llvm-svn: 274394
2016-07-02 06:47:50 +08:00
; GCN-DAG: ds_read_b32 v{{[0-9]+}}, v{{[0-9]+}} offset:8
; GCN-DAG: ds_read_b64 v{{\[[0-9]+:[0-9]+\]}}, v{{[0-9]+$}}
AMDGPU: Cleanup load tests There are a lot of different kinds of loads to test for, and these were scattered around inconsistently with some redundancy. Try to comprehensively test all loads in a consistent way. llvm-svn: 271571
2016-06-03 03:54:26 +08:00
; GCN: s_waitcnt
; GCN-DAG: ds_write_b64
; GCN-DAG: ds_write_b32 v{{[0-9]+}}, v{{[0-9]+}} offset:8{{$}}
; EG: LDS_READ_RET
; EG: LDS_READ_RET
; EG: LDS_READ_RET
AMDGPU: Mark all unspecified CC functions in tests as amdgpu_kernel Currently the default C calling convention functions are treated the same as compute kernels. Make this explicit so the default calling convention can be changed to a non-kernel. Converted with perl -pi -e 's/define void/define amdgpu_kernel void/' on the relevant test directories (and undoing in one place that actually wanted a non-kernel). llvm-svn: 298444
2017-03-22 05:39:51 +08:00
define amdgpu_kernel void @local_load_v3f32(<3 x float> addrspace(3)* %out, <3 x float> addrspace(3)* %in) #0 {
AMDGPU: Cleanup load tests There are a lot of different kinds of loads to test for, and these were scattered around inconsistently with some redundancy. Try to comprehensively test all loads in a consistent way. llvm-svn: 271571
2016-06-03 03:54:26 +08:00
entry:
%tmp0 = load <3 x float>, <3 x float> addrspace(3)* %in
store <3 x float> %tmp0, <3 x float> addrspace(3)* %out
ret void
}
; FUNC-LABEL: {{^}}local_load_v4f32:
AMDGPU: Select DS insts without m0 initialization GFX9 stopped using m0 for most DS instructions. Select a different instruction without the use. I think this will be less error prone than trying to manually maintain m0 uses as needed. llvm-svn: 319270
2017-11-29 08:55:57 +08:00
; SICIVI: s_mov_b32 m0
; GFX9-NOT: m0
AMDGPU: Cleanup load tests There are a lot of different kinds of loads to test for, and these were scattered around inconsistently with some redundancy. Try to comprehensively test all loads in a consistent way. llvm-svn: 271571
2016-06-03 03:54:26 +08:00
; GCN: ds_read2_b64
; EG: LDS_READ_RET
; EG: LDS_READ_RET
; EG: LDS_READ_RET
; EG: LDS_READ_RET
AMDGPU: Mark all unspecified CC functions in tests as amdgpu_kernel Currently the default C calling convention functions are treated the same as compute kernels. Make this explicit so the default calling convention can be changed to a non-kernel. Converted with perl -pi -e 's/define void/define amdgpu_kernel void/' on the relevant test directories (and undoing in one place that actually wanted a non-kernel). llvm-svn: 298444
2017-03-22 05:39:51 +08:00
define amdgpu_kernel void @local_load_v4f32(<4 x float> addrspace(3)* %out, <4 x float> addrspace(3)* %in) #0 {
AMDGPU: Cleanup load tests There are a lot of different kinds of loads to test for, and these were scattered around inconsistently with some redundancy. Try to comprehensively test all loads in a consistent way. llvm-svn: 271571
2016-06-03 03:54:26 +08:00
entry:
%tmp0 = load <4 x float>, <4 x float> addrspace(3)* %in
store <4 x float> %tmp0, <4 x float> addrspace(3)* %out
ret void
}
; FUNC-LABEL: {{^}}local_load_v8f32:
AMDGPU: Select DS insts without m0 initialization GFX9 stopped using m0 for most DS instructions. Select a different instruction without the use. I think this will be less error prone than trying to manually maintain m0 uses as needed. llvm-svn: 319270
2017-11-29 08:55:57 +08:00
; SICIVI: s_mov_b32 m0
; GFX9-NOT: m0
AMDGPU: Cleanup load tests There are a lot of different kinds of loads to test for, and these were scattered around inconsistently with some redundancy. Try to comprehensively test all loads in a consistent way. llvm-svn: 271571
2016-06-03 03:54:26 +08:00
; GCN: ds_read2_b64
; GCN: ds_read2_b64
; EG: LDS_READ_RET
; EG: LDS_READ_RET
; EG: LDS_READ_RET
; EG: LDS_READ_RET
; EG: LDS_READ_RET
; EG: LDS_READ_RET
; EG: LDS_READ_RET
; EG: LDS_READ_RET
AMDGPU: Mark all unspecified CC functions in tests as amdgpu_kernel Currently the default C calling convention functions are treated the same as compute kernels. Make this explicit so the default calling convention can be changed to a non-kernel. Converted with perl -pi -e 's/define void/define amdgpu_kernel void/' on the relevant test directories (and undoing in one place that actually wanted a non-kernel). llvm-svn: 298444
2017-03-22 05:39:51 +08:00
define amdgpu_kernel void @local_load_v8f32(<8 x float> addrspace(3)* %out, <8 x float> addrspace(3)* %in) #0 {
AMDGPU: Cleanup load tests There are a lot of different kinds of loads to test for, and these were scattered around inconsistently with some redundancy. Try to comprehensively test all loads in a consistent way. llvm-svn: 271571
2016-06-03 03:54:26 +08:00
entry:
%tmp0 = load <8 x float>, <8 x float> addrspace(3)* %in
store <8 x float> %tmp0, <8 x float> addrspace(3)* %out
ret void
}
; FUNC-LABEL: {{^}}local_load_v16f32:
AMDGPU: Select DS insts without m0 initialization GFX9 stopped using m0 for most DS instructions. Select a different instruction without the use. I think this will be less error prone than trying to manually maintain m0 uses as needed. llvm-svn: 319270
2017-11-29 08:55:57 +08:00
; SICIVI: s_mov_b32 m0
; GFX9-NOT: m0
AMDGPU: Cleanup load tests There are a lot of different kinds of loads to test for, and these were scattered around inconsistently with some redundancy. Try to comprehensively test all loads in a consistent way. llvm-svn: 271571
2016-06-03 03:54:26 +08:00
; GCN: ds_read2_b64
; GCN: ds_read2_b64
; GCN: ds_read2_b64
; GCN: ds_read2_b64
; EG: LDS_READ_RET
; EG: LDS_READ_RET
; EG: LDS_READ_RET
; EG: LDS_READ_RET
; EG: LDS_READ_RET
; EG: LDS_READ_RET
; EG: LDS_READ_RET
; EG: LDS_READ_RET
; EG: LDS_READ_RET
; EG: LDS_READ_RET
; EG: LDS_READ_RET
; EG: LDS_READ_RET
; EG: LDS_READ_RET
; EG: LDS_READ_RET
; EG: LDS_READ_RET
; EG: LDS_READ_RET
AMDGPU: Mark all unspecified CC functions in tests as amdgpu_kernel Currently the default C calling convention functions are treated the same as compute kernels. Make this explicit so the default calling convention can be changed to a non-kernel. Converted with perl -pi -e 's/define void/define amdgpu_kernel void/' on the relevant test directories (and undoing in one place that actually wanted a non-kernel). llvm-svn: 298444
2017-03-22 05:39:51 +08:00
define amdgpu_kernel void @local_load_v16f32(<16 x float> addrspace(3)* %out, <16 x float> addrspace(3)* %in) #0 {
AMDGPU: Cleanup load tests There are a lot of different kinds of loads to test for, and these were scattered around inconsistently with some redundancy. Try to comprehensively test all loads in a consistent way. llvm-svn: 271571
2016-06-03 03:54:26 +08:00
entry:
%tmp0 = load <16 x float>, <16 x float> addrspace(3)* %in
store <16 x float> %tmp0, <16 x float> addrspace(3)* %out
ret void
}
[AMDGPU] Supported ds_read_b128 generation; Widened vector length for local address-space. Summary: Starting from GCN 2nd generation, ISA supports ds_read_b128 on top of ds_read_b64. This patch supports ds_read_b128 instruction pattern and generation of this instruction. In the vectorizer, this patch also widen the vector length so that vectorizer generates 128 bit loads for local address-space which gets translated to ds_read_b128. Since the performance benefit is not clear; compiler generates ds_read_b128 under -amdgpu-ds128. Author: FarhanaAleen Reviewed By: rampitec, arsenm Subscribers: llvm-commits, AMDGPU Differential Revision: https://reviews.llvm.org/D44210 llvm-svn: 327153
2018-03-10 01:41:39 +08:00
; Tests if ds_read_b128 gets generated for the 16 byte aligned load.
; FUNC-LABEL: {{^}}local_v4f32_to_128:
; SI-NOT: ds_read_b128
; CIVI: ds_read_b128
; EG: LDS_READ_RET
; EG: LDS_READ_RET
; EG: LDS_READ_RET
; EG: LDS_READ_RET
define amdgpu_kernel void @local_v4f32_to_128(<4 x float> addrspace(3)* %out, <4 x float> addrspace(3)* %in) {
%ld = load <4 x float>, <4 x float> addrspace(3)* %in, align 16
store <4 x float> %ld, <4 x float> addrspace(3)* %out
ret void
}
AMDGPU: Cleanup load tests There are a lot of different kinds of loads to test for, and these were scattered around inconsistently with some redundancy. Try to comprehensively test all loads in a consistent way. llvm-svn: 271571
2016-06-03 03:54:26 +08:00
attributes #0 = { nounwind }