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-i64.ll

191 lines
4.7 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 -mtriple=amdgcn--amdhsa -mcpu=kaveri -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=amdgcn -mcpu=gfx900 -verify-machineinstrs < %s | FileCheck -check-prefixes=GCN,GFX9,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_write_b128 generation. Summary: This is a follow-on patch of https://reviews.llvm.org/D44210 Author: FarhanaAleen Reviewed By: msearles Subscribers: llvm-commits, AMDGPU Differential Revision: https://reviews.llvm.org/D44319 llvm-svn: 327726
2018-03-17 02:12:00 +08:00
; Testing for ds_read/write_b128
AMDGPU: enable 128-bit for local addr space under an option Author: Samuel Pitoiset ds_read_b128 and ds_write_b128 have been recently enabled under the amdgpu-ds128 option because the performance benefit is unclear. Though, using 128-bit loads/stores for the local address space appears to introduce regressions in tessellation shaders. Not sure what is broken, but as ds_read_b128/ds_write_b128 are not enabled by default, just introduce a global option and enable 128-bit only if requested (until it's fixed/used correctly). v2: - fix regressions in merge-stores.ll and multiple_tails.ll Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=105464 llvm-svn: 329764
2018-04-11 06:48:23 +08:00
; RUN: llc -march=amdgcn -mcpu=tonga -verify-machineinstrs -mattr=+enable-ds128 < %s | FileCheck -check-prefixes=CIVI,FUNC %s
; RUN: llc -march=amdgcn -mcpu=gfx900 -verify-machineinstrs -mattr=+enable-ds128 < %s | FileCheck -check-prefixes=CIVI,FUNC %s
[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
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_i64:
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 [[VAL:v\[[0-9]+:[0-9]+\]]], v{{[0-9]+}}{{$}}
; GCN: ds_write_b64 v{{[0-9]+}}, [[VAL]]
; 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_i64(i64 addrspace(3)* %out, i64 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
%ld = load i64, i64 addrspace(3)* %in
store i64 %ld, i64 addrspace(3)* %out
ret void
}
; FUNC-LABEL: {{^}}local_load_v2i64:
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_v2i64(<2 x i64> addrspace(3)* %out, <2 x i64> 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:
%ld = load <2 x i64>, <2 x i64> addrspace(3)* %in
store <2 x i64> %ld, <2 x i64> addrspace(3)* %out
ret void
}
[AMDGPU] Supported ds_write_b128 generation. Summary: This is a follow-on patch of https://reviews.llvm.org/D44210 Author: FarhanaAleen Reviewed By: msearles Subscribers: llvm-commits, AMDGPU Differential Revision: https://reviews.llvm.org/D44319 llvm-svn: 327726
2018-03-17 02:12:00 +08:00
; Tests if ds_read/write_b128 gets generated for the 16 byte aligned load.
[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
; FUNC-LABEL: {{^}}local_load_v2i64_to_128:
[AMDGPU] Supported ds_write_b128 generation. Summary: This is a follow-on patch of https://reviews.llvm.org/D44210 Author: FarhanaAleen Reviewed By: msearles Subscribers: llvm-commits, AMDGPU Differential Revision: https://reviews.llvm.org/D44319 llvm-svn: 327726
2018-03-17 02:12:00 +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
; CIVI: ds_read_b128
[AMDGPU] Supported ds_write_b128 generation. Summary: This is a follow-on patch of https://reviews.llvm.org/D44210 Author: FarhanaAleen Reviewed By: msearles Subscribers: llvm-commits, AMDGPU Differential Revision: https://reviews.llvm.org/D44319 llvm-svn: 327726
2018-03-17 02:12:00 +08:00
; CIVI: ds_write_b128
[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
define amdgpu_kernel void @local_load_v2i64_to_128(<2 x i64> addrspace(3)* %out, <2 x i64> addrspace(3)* %in) {
entry:
[AMDGPU] Supported ds_write_b128 generation. Summary: This is a follow-on patch of https://reviews.llvm.org/D44210 Author: FarhanaAleen Reviewed By: msearles Subscribers: llvm-commits, AMDGPU Differential Revision: https://reviews.llvm.org/D44319 llvm-svn: 327726
2018-03-17 02:12:00 +08:00
%ld = load <2 x i64>, <2 x i64> addrspace(3)* %in, align 16
store <2 x i64> %ld, <2 x i64> addrspace(3)* %out, align 16
[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
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
; FUNC-LABEL: {{^}}local_load_v3i64:
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-DAG: ds_read2_b64
; GCN-DAG: ds_read_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
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_v3i64(<3 x i64> addrspace(3)* %out, <3 x i64> 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:
%ld = load <3 x i64>, <3 x i64> addrspace(3)* %in
store <3 x i64> %ld, <3 x i64> addrspace(3)* %out
ret void
}
; FUNC-LABEL: {{^}}local_load_v4i64:
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_v4i64(<4 x i64> addrspace(3)* %out, <4 x i64> 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:
%ld = load <4 x i64>, <4 x i64> addrspace(3)* %in
store <4 x i64> %ld, <4 x i64> addrspace(3)* %out
ret void
}
; FUNC-LABEL: {{^}}local_load_v8i64:
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_v8i64(<8 x i64> addrspace(3)* %out, <8 x i64> 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:
%ld = load <8 x i64>, <8 x i64> addrspace(3)* %in
store <8 x i64> %ld, <8 x i64> addrspace(3)* %out
ret void
}
; FUNC-LABEL: {{^}}local_load_v16i64:
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
; 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
; 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_v16i64(<16 x i64> addrspace(3)* %out, <16 x i64> 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:
%ld = load <16 x i64>, <16 x i64> addrspace(3)* %in
store <16 x i64> %ld, <16 x i64> addrspace(3)* %out
ret void
}
attributes #0 = { nounwind }