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/llvm.amdgcn.raw.buffer.load.ll

275 lines
13 KiB
LLVM
Raw Normal View History

[AMDGPU] New buffer intrinsics Summary: This commit adds new intrinsics llvm.amdgcn.raw.buffer.load llvm.amdgcn.raw.buffer.load.format llvm.amdgcn.raw.buffer.load.format.d16 llvm.amdgcn.struct.buffer.load llvm.amdgcn.struct.buffer.load.format llvm.amdgcn.struct.buffer.load.format.d16 llvm.amdgcn.raw.buffer.store llvm.amdgcn.raw.buffer.store.format llvm.amdgcn.raw.buffer.store.format.d16 llvm.amdgcn.struct.buffer.store llvm.amdgcn.struct.buffer.store.format llvm.amdgcn.struct.buffer.store.format.d16 llvm.amdgcn.raw.buffer.atomic.* llvm.amdgcn.struct.buffer.atomic.* with the following changes from the llvm.amdgcn.buffer.* intrinsics: * there are separate raw and struct versions: raw does not have an index arg and sets idxen=0 in the instruction, and struct always sets idxen=1 in the instruction even if the index is 0, to allow for the fact that gfx9 does bounds checking differently depending on whether idxen is set; * there is a combined cachepolicy arg (glc+slc) * there are now only two offset args: one for the offset that is included in bounds checking and swizzling, to be split between the instruction's voffset and immoffset fields, and one for the offset that is excluded from bounds checking and swizzling, to go into the instruction's soffset field. The AMDISD::BUFFER_* SD nodes always have an index operand, all three offset operands, combined cachepolicy operand, and an extra idxen operand. The obsolescent llvm.amdgcn.buffer.* intrinsics continue to work. Subscribers: arsenm, kzhuravl, wdng, nhaehnle, yaxunl, dstuttard, t-tye, jfb, llvm-commits Differential Revision: https://reviews.llvm.org/D50306 Change-Id: If897ea7dc34fcbf4d5496e98cc99a934f62fc205 llvm-svn: 340269
2018-08-21 19:07:10 +08:00
;RUN: llc < %s -march=amdgcn -mcpu=verde -verify-machineinstrs | FileCheck %s -check-prefix=CHECK -check-prefix=SICI
;RUN: llc < %s -march=amdgcn -mcpu=tonga -verify-machineinstrs | FileCheck %s -check-prefix=CHECK -check-prefix=VI
;CHECK-LABEL: {{^}}buffer_load:
;CHECK: buffer_load_dwordx4 v[0:3], off, s[0:3], 0
;CHECK: buffer_load_dwordx4 v[4:7], off, s[0:3], 0 glc
;CHECK: buffer_load_dwordx4 v[8:11], off, s[0:3], 0 slc
;CHECK: s_waitcnt
define amdgpu_ps {<4 x float>, <4 x float>, <4 x float>} @buffer_load(<4 x i32> inreg) {
main_body:
%data = call <4 x float> @llvm.amdgcn.raw.buffer.load.v4f32(<4 x i32> %0, i32 0, i32 0, i32 0)
%data_glc = call <4 x float> @llvm.amdgcn.raw.buffer.load.v4f32(<4 x i32> %0, i32 0, i32 0, i32 1)
%data_slc = call <4 x float> @llvm.amdgcn.raw.buffer.load.v4f32(<4 x i32> %0, i32 0, i32 0, i32 2)
%r0 = insertvalue {<4 x float>, <4 x float>, <4 x float>} undef, <4 x float> %data, 0
%r1 = insertvalue {<4 x float>, <4 x float>, <4 x float>} %r0, <4 x float> %data_glc, 1
%r2 = insertvalue {<4 x float>, <4 x float>, <4 x float>} %r1, <4 x float> %data_slc, 2
ret {<4 x float>, <4 x float>, <4 x float>} %r2
}
;CHECK-LABEL: {{^}}buffer_load_immoffs:
;CHECK: buffer_load_dwordx4 v[0:3], off, s[0:3], 0 offset:40
;CHECK: s_waitcnt
define amdgpu_ps <4 x float> @buffer_load_immoffs(<4 x i32> inreg) {
main_body:
%data = call <4 x float> @llvm.amdgcn.raw.buffer.load.v4f32(<4 x i32> %0, i32 40, i32 0, i32 0)
ret <4 x float> %data
}
;CHECK-LABEL: {{^}}buffer_load_immoffs_large:
;CHECK: s_movk_i32 [[OFFSET:s[0-9]+]], 0x1ffc
;CHECK: buffer_load_dwordx4 v[0:3], off, s[0:3], [[OFFSET]] offset:4
;CHECK: s_waitcnt
define amdgpu_ps <4 x float> @buffer_load_immoffs_large(<4 x i32> inreg) {
main_body:
%data = call <4 x float> @llvm.amdgcn.raw.buffer.load.v4f32(<4 x i32> %0, i32 4, i32 8188, i32 0)
ret <4 x float> %data
}
;CHECK-LABEL: {{^}}buffer_load_ofs:
;CHECK: buffer_load_dwordx4 v[0:3], v0, s[0:3], 0 offen
;CHECK: s_waitcnt
define amdgpu_ps <4 x float> @buffer_load_ofs(<4 x i32> inreg, i32) {
main_body:
%data = call <4 x float> @llvm.amdgcn.raw.buffer.load.v4f32(<4 x i32> %0, i32 %1, i32 0, i32 0)
ret <4 x float> %data
}
;CHECK-LABEL: {{^}}buffer_load_ofs_imm:
;CHECK: buffer_load_dwordx4 v[0:3], v0, s[0:3], 0 offen offset:60
;CHECK: s_waitcnt
define amdgpu_ps <4 x float> @buffer_load_ofs_imm(<4 x i32> inreg, i32) {
main_body:
%ofs = add i32 %1, 60
%data = call <4 x float> @llvm.amdgcn.raw.buffer.load.v4f32(<4 x i32> %0, i32 %ofs, i32 0, i32 0)
ret <4 x float> %data
}
;CHECK-LABEL: {{^}}buffer_load_x1:
;CHECK: buffer_load_dword v0, v0, s[0:3], 0 offen
;CHECK: s_waitcnt
define amdgpu_ps float @buffer_load_x1(<4 x i32> inreg %rsrc, i32 %ofs) {
main_body:
%data = call float @llvm.amdgcn.raw.buffer.load.f32(<4 x i32> %rsrc, i32 %ofs, i32 0, i32 0)
ret float %data
}
;CHECK-LABEL: {{^}}buffer_load_x2:
;CHECK: buffer_load_dwordx2 v[0:1], v0, s[0:3], 0 offen
;CHECK: s_waitcnt
define amdgpu_ps <2 x float> @buffer_load_x2(<4 x i32> inreg %rsrc, i32 %ofs) {
main_body:
%data = call <2 x float> @llvm.amdgcn.raw.buffer.load.v2f32(<4 x i32> %rsrc, i32 %ofs, i32 0, i32 0)
ret <2 x float> %data
}
;CHECK-LABEL: {{^}}buffer_load_negative_offset:
[AMDGPU] Fix for negative offsets in buffer/tbuffer intrinsics Summary: The new buffer/tbuffer intrinsics handle an out-of-range immediate offset by moving/adding offset&-4096 to a vgpr, leaving an in-range immediate offset, with a chance of the move/add being CSEd for similar loads/stores. However it turns out that a negative offset in a vgpr is illegal, even if adding the immediate offset makes it legal again. Therefore, this commit disables the offset&-4096 thing if the offset is negative. Differential Revision: https://reviews.llvm.org/D52683 Change-Id: Ie02f0a74f240a138dc2a29d17cfbd9e350e4ed13 llvm-svn: 343672
2018-10-03 18:29:43 +08:00
;CHECK: v_add_{{[iu]}}32_e32 [[VOFS:v[0-9]+]], vcc, -16, v0
;CHECK: buffer_load_dwordx4 v[0:3], [[VOFS]], s[0:3], 0 offen
[AMDGPU] New buffer intrinsics Summary: This commit adds new intrinsics llvm.amdgcn.raw.buffer.load llvm.amdgcn.raw.buffer.load.format llvm.amdgcn.raw.buffer.load.format.d16 llvm.amdgcn.struct.buffer.load llvm.amdgcn.struct.buffer.load.format llvm.amdgcn.struct.buffer.load.format.d16 llvm.amdgcn.raw.buffer.store llvm.amdgcn.raw.buffer.store.format llvm.amdgcn.raw.buffer.store.format.d16 llvm.amdgcn.struct.buffer.store llvm.amdgcn.struct.buffer.store.format llvm.amdgcn.struct.buffer.store.format.d16 llvm.amdgcn.raw.buffer.atomic.* llvm.amdgcn.struct.buffer.atomic.* with the following changes from the llvm.amdgcn.buffer.* intrinsics: * there are separate raw and struct versions: raw does not have an index arg and sets idxen=0 in the instruction, and struct always sets idxen=1 in the instruction even if the index is 0, to allow for the fact that gfx9 does bounds checking differently depending on whether idxen is set; * there is a combined cachepolicy arg (glc+slc) * there are now only two offset args: one for the offset that is included in bounds checking and swizzling, to be split between the instruction's voffset and immoffset fields, and one for the offset that is excluded from bounds checking and swizzling, to go into the instruction's soffset field. The AMDISD::BUFFER_* SD nodes always have an index operand, all three offset operands, combined cachepolicy operand, and an extra idxen operand. The obsolescent llvm.amdgcn.buffer.* intrinsics continue to work. Subscribers: arsenm, kzhuravl, wdng, nhaehnle, yaxunl, dstuttard, t-tye, jfb, llvm-commits Differential Revision: https://reviews.llvm.org/D50306 Change-Id: If897ea7dc34fcbf4d5496e98cc99a934f62fc205 llvm-svn: 340269
2018-08-21 19:07:10 +08:00
define amdgpu_ps <4 x float> @buffer_load_negative_offset(<4 x i32> inreg, i32 %ofs) {
main_body:
%ofs.1 = add i32 %ofs, -16
%data = call <4 x float> @llvm.amdgcn.raw.buffer.load.v4f32(<4 x i32> %0, i32 %ofs.1, i32 0, i32 0)
ret <4 x float> %data
}
; SI won't merge ds memory operations, because of the signed offset bug, so
; we only have check lines for VI.
; CHECK-LABEL: buffer_load_mmo:
; VI: v_mov_b32_e32 [[ZERO:v[0-9]+]], 0
; VI: ds_write2_b32 v{{[0-9]+}}, [[ZERO]], [[ZERO]] offset1:4
define amdgpu_ps float @buffer_load_mmo(<4 x i32> inreg %rsrc, float addrspace(3)* %lds) {
entry:
store float 0.0, float addrspace(3)* %lds
%val = call float @llvm.amdgcn.raw.buffer.load.f32(<4 x i32> %rsrc, i32 0, i32 0, i32 0)
%tmp2 = getelementptr float, float addrspace(3)* %lds, i32 4
store float 0.0, float addrspace(3)* %tmp2
ret float %val
}
[AMDGPU] Handle OR as operand of raw load/store Summary: Use isBaseWithConstantOffset() which handles OR as an operand to llvm.amdgcn.raw.buffer.load and llvm.amdgcn.raw.buffer.store. Change-Id: Ifefb9dc5ded8710d333df07ab1900b230e33539a Reviewers: nhaehnle, mareko, arsenm Reviewed By: arsenm Subscribers: arsenm, kzhuravl, jvesely, wdng, nhaehnle, yaxunl, dstuttard, tpr, t-tye, llvm-commits Differential Revision: https://reviews.llvm.org/D55999 llvm-svn: 350208
2019-01-02 17:47:41 +08:00
;CHECK-LABEL: {{^}}buffer_load_x1_offen_merged_and:
[AMDGPU] New buffer intrinsics Summary: This commit adds new intrinsics llvm.amdgcn.raw.buffer.load llvm.amdgcn.raw.buffer.load.format llvm.amdgcn.raw.buffer.load.format.d16 llvm.amdgcn.struct.buffer.load llvm.amdgcn.struct.buffer.load.format llvm.amdgcn.struct.buffer.load.format.d16 llvm.amdgcn.raw.buffer.store llvm.amdgcn.raw.buffer.store.format llvm.amdgcn.raw.buffer.store.format.d16 llvm.amdgcn.struct.buffer.store llvm.amdgcn.struct.buffer.store.format llvm.amdgcn.struct.buffer.store.format.d16 llvm.amdgcn.raw.buffer.atomic.* llvm.amdgcn.struct.buffer.atomic.* with the following changes from the llvm.amdgcn.buffer.* intrinsics: * there are separate raw and struct versions: raw does not have an index arg and sets idxen=0 in the instruction, and struct always sets idxen=1 in the instruction even if the index is 0, to allow for the fact that gfx9 does bounds checking differently depending on whether idxen is set; * there is a combined cachepolicy arg (glc+slc) * there are now only two offset args: one for the offset that is included in bounds checking and swizzling, to be split between the instruction's voffset and immoffset fields, and one for the offset that is excluded from bounds checking and swizzling, to go into the instruction's soffset field. The AMDISD::BUFFER_* SD nodes always have an index operand, all three offset operands, combined cachepolicy operand, and an extra idxen operand. The obsolescent llvm.amdgcn.buffer.* intrinsics continue to work. Subscribers: arsenm, kzhuravl, wdng, nhaehnle, yaxunl, dstuttard, t-tye, jfb, llvm-commits Differential Revision: https://reviews.llvm.org/D50306 Change-Id: If897ea7dc34fcbf4d5496e98cc99a934f62fc205 llvm-svn: 340269
2018-08-21 19:07:10 +08:00
;CHECK-NEXT: %bb.
;CHECK-NEXT: buffer_load_dwordx4 v[{{[0-9]}}:{{[0-9]}}], v0, s[0:3], 0 offen offset:4
;CHECK-NEXT: buffer_load_dwordx2 v[{{[0-9]}}:{{[0-9]}}], v0, s[0:3], 0 offen offset:28
;CHECK: s_waitcnt
[AMDGPU] Handle OR as operand of raw load/store Summary: Use isBaseWithConstantOffset() which handles OR as an operand to llvm.amdgcn.raw.buffer.load and llvm.amdgcn.raw.buffer.store. Change-Id: Ifefb9dc5ded8710d333df07ab1900b230e33539a Reviewers: nhaehnle, mareko, arsenm Reviewed By: arsenm Subscribers: arsenm, kzhuravl, jvesely, wdng, nhaehnle, yaxunl, dstuttard, tpr, t-tye, llvm-commits Differential Revision: https://reviews.llvm.org/D55999 llvm-svn: 350208
2019-01-02 17:47:41 +08:00
define amdgpu_ps void @buffer_load_x1_offen_merged_and(<4 x i32> inreg %rsrc, i32 %a) {
[AMDGPU] New buffer intrinsics Summary: This commit adds new intrinsics llvm.amdgcn.raw.buffer.load llvm.amdgcn.raw.buffer.load.format llvm.amdgcn.raw.buffer.load.format.d16 llvm.amdgcn.struct.buffer.load llvm.amdgcn.struct.buffer.load.format llvm.amdgcn.struct.buffer.load.format.d16 llvm.amdgcn.raw.buffer.store llvm.amdgcn.raw.buffer.store.format llvm.amdgcn.raw.buffer.store.format.d16 llvm.amdgcn.struct.buffer.store llvm.amdgcn.struct.buffer.store.format llvm.amdgcn.struct.buffer.store.format.d16 llvm.amdgcn.raw.buffer.atomic.* llvm.amdgcn.struct.buffer.atomic.* with the following changes from the llvm.amdgcn.buffer.* intrinsics: * there are separate raw and struct versions: raw does not have an index arg and sets idxen=0 in the instruction, and struct always sets idxen=1 in the instruction even if the index is 0, to allow for the fact that gfx9 does bounds checking differently depending on whether idxen is set; * there is a combined cachepolicy arg (glc+slc) * there are now only two offset args: one for the offset that is included in bounds checking and swizzling, to be split between the instruction's voffset and immoffset fields, and one for the offset that is excluded from bounds checking and swizzling, to go into the instruction's soffset field. The AMDISD::BUFFER_* SD nodes always have an index operand, all three offset operands, combined cachepolicy operand, and an extra idxen operand. The obsolescent llvm.amdgcn.buffer.* intrinsics continue to work. Subscribers: arsenm, kzhuravl, wdng, nhaehnle, yaxunl, dstuttard, t-tye, jfb, llvm-commits Differential Revision: https://reviews.llvm.org/D50306 Change-Id: If897ea7dc34fcbf4d5496e98cc99a934f62fc205 llvm-svn: 340269
2018-08-21 19:07:10 +08:00
main_body:
%a1 = add i32 %a, 4
%a2 = add i32 %a, 8
%a3 = add i32 %a, 12
%a4 = add i32 %a, 16
%a5 = add i32 %a, 28
%a6 = add i32 %a, 32
%r1 = call float @llvm.amdgcn.raw.buffer.load.f32(<4 x i32> %rsrc, i32 %a1, i32 0, i32 0)
%r2 = call float @llvm.amdgcn.raw.buffer.load.f32(<4 x i32> %rsrc, i32 %a2, i32 0, i32 0)
%r3 = call float @llvm.amdgcn.raw.buffer.load.f32(<4 x i32> %rsrc, i32 %a3, i32 0, i32 0)
%r4 = call float @llvm.amdgcn.raw.buffer.load.f32(<4 x i32> %rsrc, i32 %a4, i32 0, i32 0)
%r5 = call float @llvm.amdgcn.raw.buffer.load.f32(<4 x i32> %rsrc, i32 %a5, i32 0, i32 0)
%r6 = call float @llvm.amdgcn.raw.buffer.load.f32(<4 x i32> %rsrc, i32 %a6, i32 0, i32 0)
call void @llvm.amdgcn.exp.f32(i32 0, i32 15, float %r1, float %r2, float %r3, float %r4, i1 true, i1 true)
call void @llvm.amdgcn.exp.f32(i32 0, i32 15, float %r5, float %r6, float undef, float undef, i1 true, i1 true)
ret void
}
[AMDGPU] Handle OR as operand of raw load/store Summary: Use isBaseWithConstantOffset() which handles OR as an operand to llvm.amdgcn.raw.buffer.load and llvm.amdgcn.raw.buffer.store. Change-Id: Ifefb9dc5ded8710d333df07ab1900b230e33539a Reviewers: nhaehnle, mareko, arsenm Reviewed By: arsenm Subscribers: arsenm, kzhuravl, jvesely, wdng, nhaehnle, yaxunl, dstuttard, tpr, t-tye, llvm-commits Differential Revision: https://reviews.llvm.org/D55999 llvm-svn: 350208
2019-01-02 17:47:41 +08:00
;CHECK-LABEL: {{^}}buffer_load_x1_offen_merged_or:
;CHECK-NEXT: %bb.
;CHECK-NEXT: v_lshlrev_b32_e32 v{{[0-9]}}, 6, v0
;CHECK-NEXT: buffer_load_dwordx4 v[{{[0-9]}}:{{[0-9]}}], v{{[0-9]}}, s[0:3], 0 offen offset:4
;CHECK-NEXT: buffer_load_dwordx2 v[{{[0-9]}}:{{[0-9]}}], v{{[0-9]}}, s[0:3], 0 offen offset:28
;CHECK: s_waitcnt
define amdgpu_ps void @buffer_load_x1_offen_merged_or(<4 x i32> inreg %rsrc, i32 %inp) {
main_body:
%a = shl i32 %inp, 6
%a1 = or i32 %a, 4
%a2 = or i32 %a, 8
%a3 = or i32 %a, 12
%a4 = or i32 %a, 16
%a5 = or i32 %a, 28
%a6 = or i32 %a, 32
%r1 = call float @llvm.amdgcn.raw.buffer.load.f32(<4 x i32> %rsrc, i32 %a1, i32 0, i32 0)
%r2 = call float @llvm.amdgcn.raw.buffer.load.f32(<4 x i32> %rsrc, i32 %a2, i32 0, i32 0)
%r3 = call float @llvm.amdgcn.raw.buffer.load.f32(<4 x i32> %rsrc, i32 %a3, i32 0, i32 0)
%r4 = call float @llvm.amdgcn.raw.buffer.load.f32(<4 x i32> %rsrc, i32 %a4, i32 0, i32 0)
%r5 = call float @llvm.amdgcn.raw.buffer.load.f32(<4 x i32> %rsrc, i32 %a5, i32 0, i32 0)
%r6 = call float @llvm.amdgcn.raw.buffer.load.f32(<4 x i32> %rsrc, i32 %a6, i32 0, i32 0)
call void @llvm.amdgcn.exp.f32(i32 0, i32 15, float %r1, float %r2, float %r3, float %r4, i1 true, i1 true)
call void @llvm.amdgcn.exp.f32(i32 0, i32 15, float %r5, float %r6, float undef, float undef, i1 true, i1 true)
ret void
}
[AMDGPU] New buffer intrinsics Summary: This commit adds new intrinsics llvm.amdgcn.raw.buffer.load llvm.amdgcn.raw.buffer.load.format llvm.amdgcn.raw.buffer.load.format.d16 llvm.amdgcn.struct.buffer.load llvm.amdgcn.struct.buffer.load.format llvm.amdgcn.struct.buffer.load.format.d16 llvm.amdgcn.raw.buffer.store llvm.amdgcn.raw.buffer.store.format llvm.amdgcn.raw.buffer.store.format.d16 llvm.amdgcn.struct.buffer.store llvm.amdgcn.struct.buffer.store.format llvm.amdgcn.struct.buffer.store.format.d16 llvm.amdgcn.raw.buffer.atomic.* llvm.amdgcn.struct.buffer.atomic.* with the following changes from the llvm.amdgcn.buffer.* intrinsics: * there are separate raw and struct versions: raw does not have an index arg and sets idxen=0 in the instruction, and struct always sets idxen=1 in the instruction even if the index is 0, to allow for the fact that gfx9 does bounds checking differently depending on whether idxen is set; * there is a combined cachepolicy arg (glc+slc) * there are now only two offset args: one for the offset that is included in bounds checking and swizzling, to be split between the instruction's voffset and immoffset fields, and one for the offset that is excluded from bounds checking and swizzling, to go into the instruction's soffset field. The AMDISD::BUFFER_* SD nodes always have an index operand, all three offset operands, combined cachepolicy operand, and an extra idxen operand. The obsolescent llvm.amdgcn.buffer.* intrinsics continue to work. Subscribers: arsenm, kzhuravl, wdng, nhaehnle, yaxunl, dstuttard, t-tye, jfb, llvm-commits Differential Revision: https://reviews.llvm.org/D50306 Change-Id: If897ea7dc34fcbf4d5496e98cc99a934f62fc205 llvm-svn: 340269
2018-08-21 19:07:10 +08:00
;CHECK-LABEL: {{^}}buffer_load_x1_offen_merged_glc_slc:
;CHECK-NEXT: %bb.
;CHECK-NEXT: buffer_load_dwordx2 v[{{[0-9]}}:{{[0-9]}}], v0, s[0:3], 0 offen offset:4{{$}}
;CHECK-NEXT: buffer_load_dwordx2 v[{{[0-9]}}:{{[0-9]}}], v0, s[0:3], 0 offen offset:12 glc{{$}}
;CHECK-NEXT: buffer_load_dwordx2 v[{{[0-9]}}:{{[0-9]}}], v0, s[0:3], 0 offen offset:28 glc slc{{$}}
;CHECK: s_waitcnt
define amdgpu_ps void @buffer_load_x1_offen_merged_glc_slc(<4 x i32> inreg %rsrc, i32 %a) {
main_body:
%a1 = add i32 %a, 4
%a2 = add i32 %a, 8
%a3 = add i32 %a, 12
%a4 = add i32 %a, 16
%a5 = add i32 %a, 28
%a6 = add i32 %a, 32
%r1 = call float @llvm.amdgcn.raw.buffer.load.f32(<4 x i32> %rsrc, i32 %a1, i32 0, i32 0)
%r2 = call float @llvm.amdgcn.raw.buffer.load.f32(<4 x i32> %rsrc, i32 %a2, i32 0, i32 0)
%r3 = call float @llvm.amdgcn.raw.buffer.load.f32(<4 x i32> %rsrc, i32 %a3, i32 0, i32 1)
%r4 = call float @llvm.amdgcn.raw.buffer.load.f32(<4 x i32> %rsrc, i32 %a4, i32 0, i32 1)
%r5 = call float @llvm.amdgcn.raw.buffer.load.f32(<4 x i32> %rsrc, i32 %a5, i32 0, i32 3)
%r6 = call float @llvm.amdgcn.raw.buffer.load.f32(<4 x i32> %rsrc, i32 %a6, i32 0, i32 3)
call void @llvm.amdgcn.exp.f32(i32 0, i32 15, float %r1, float %r2, float %r3, float %r4, i1 true, i1 true)
call void @llvm.amdgcn.exp.f32(i32 0, i32 15, float %r5, float %r6, float undef, float undef, i1 true, i1 true)
ret void
}
[AMDGPU] Handle OR as operand of raw load/store Summary: Use isBaseWithConstantOffset() which handles OR as an operand to llvm.amdgcn.raw.buffer.load and llvm.amdgcn.raw.buffer.store. Change-Id: Ifefb9dc5ded8710d333df07ab1900b230e33539a Reviewers: nhaehnle, mareko, arsenm Reviewed By: arsenm Subscribers: arsenm, kzhuravl, jvesely, wdng, nhaehnle, yaxunl, dstuttard, tpr, t-tye, llvm-commits Differential Revision: https://reviews.llvm.org/D55999 llvm-svn: 350208
2019-01-02 17:47:41 +08:00
;CHECK-LABEL: {{^}}buffer_load_x2_offen_merged_and:
[AMDGPU] New buffer intrinsics Summary: This commit adds new intrinsics llvm.amdgcn.raw.buffer.load llvm.amdgcn.raw.buffer.load.format llvm.amdgcn.raw.buffer.load.format.d16 llvm.amdgcn.struct.buffer.load llvm.amdgcn.struct.buffer.load.format llvm.amdgcn.struct.buffer.load.format.d16 llvm.amdgcn.raw.buffer.store llvm.amdgcn.raw.buffer.store.format llvm.amdgcn.raw.buffer.store.format.d16 llvm.amdgcn.struct.buffer.store llvm.amdgcn.struct.buffer.store.format llvm.amdgcn.struct.buffer.store.format.d16 llvm.amdgcn.raw.buffer.atomic.* llvm.amdgcn.struct.buffer.atomic.* with the following changes from the llvm.amdgcn.buffer.* intrinsics: * there are separate raw and struct versions: raw does not have an index arg and sets idxen=0 in the instruction, and struct always sets idxen=1 in the instruction even if the index is 0, to allow for the fact that gfx9 does bounds checking differently depending on whether idxen is set; * there is a combined cachepolicy arg (glc+slc) * there are now only two offset args: one for the offset that is included in bounds checking and swizzling, to be split between the instruction's voffset and immoffset fields, and one for the offset that is excluded from bounds checking and swizzling, to go into the instruction's soffset field. The AMDISD::BUFFER_* SD nodes always have an index operand, all three offset operands, combined cachepolicy operand, and an extra idxen operand. The obsolescent llvm.amdgcn.buffer.* intrinsics continue to work. Subscribers: arsenm, kzhuravl, wdng, nhaehnle, yaxunl, dstuttard, t-tye, jfb, llvm-commits Differential Revision: https://reviews.llvm.org/D50306 Change-Id: If897ea7dc34fcbf4d5496e98cc99a934f62fc205 llvm-svn: 340269
2018-08-21 19:07:10 +08:00
;CHECK-NEXT: %bb.
;CHECK-NEXT: buffer_load_dwordx4 v[{{[0-9]}}:{{[0-9]}}], v0, s[0:3], 0 offen offset:4
;CHECK: s_waitcnt
[AMDGPU] Handle OR as operand of raw load/store Summary: Use isBaseWithConstantOffset() which handles OR as an operand to llvm.amdgcn.raw.buffer.load and llvm.amdgcn.raw.buffer.store. Change-Id: Ifefb9dc5ded8710d333df07ab1900b230e33539a Reviewers: nhaehnle, mareko, arsenm Reviewed By: arsenm Subscribers: arsenm, kzhuravl, jvesely, wdng, nhaehnle, yaxunl, dstuttard, tpr, t-tye, llvm-commits Differential Revision: https://reviews.llvm.org/D55999 llvm-svn: 350208
2019-01-02 17:47:41 +08:00
define amdgpu_ps void @buffer_load_x2_offen_merged_and(<4 x i32> inreg %rsrc, i32 %a) {
main_body:
%a1 = add i32 %a, 4
%a2 = add i32 %a, 12
%vr1 = call <2 x float> @llvm.amdgcn.raw.buffer.load.v2f32(<4 x i32> %rsrc, i32 %a1, i32 0, i32 0)
%vr2 = call <2 x float> @llvm.amdgcn.raw.buffer.load.v2f32(<4 x i32> %rsrc, i32 %a2, i32 0, i32 0)
%r1 = extractelement <2 x float> %vr1, i32 0
%r2 = extractelement <2 x float> %vr1, i32 1
%r3 = extractelement <2 x float> %vr2, i32 0
%r4 = extractelement <2 x float> %vr2, i32 1
call void @llvm.amdgcn.exp.f32(i32 0, i32 15, float %r1, float %r2, float %r3, float %r4, i1 true, i1 true)
ret void
}
;CHECK-LABEL: {{^}}buffer_load_x2_offen_merged_or:
;CHECK-NEXT: %bb.
;CHECK-NEXT: v_lshlrev_b32_e32 v{{[0-9]}}, 4, v0
;CHECK-NEXT: buffer_load_dwordx4 v[{{[0-9]}}:{{[0-9]}}], v{{[0-9]}}, s[0:3], 0 offen offset:4
;CHECK: s_waitcnt
define amdgpu_ps void @buffer_load_x2_offen_merged_or(<4 x i32> inreg %rsrc, i32 %inp) {
[AMDGPU] New buffer intrinsics Summary: This commit adds new intrinsics llvm.amdgcn.raw.buffer.load llvm.amdgcn.raw.buffer.load.format llvm.amdgcn.raw.buffer.load.format.d16 llvm.amdgcn.struct.buffer.load llvm.amdgcn.struct.buffer.load.format llvm.amdgcn.struct.buffer.load.format.d16 llvm.amdgcn.raw.buffer.store llvm.amdgcn.raw.buffer.store.format llvm.amdgcn.raw.buffer.store.format.d16 llvm.amdgcn.struct.buffer.store llvm.amdgcn.struct.buffer.store.format llvm.amdgcn.struct.buffer.store.format.d16 llvm.amdgcn.raw.buffer.atomic.* llvm.amdgcn.struct.buffer.atomic.* with the following changes from the llvm.amdgcn.buffer.* intrinsics: * there are separate raw and struct versions: raw does not have an index arg and sets idxen=0 in the instruction, and struct always sets idxen=1 in the instruction even if the index is 0, to allow for the fact that gfx9 does bounds checking differently depending on whether idxen is set; * there is a combined cachepolicy arg (glc+slc) * there are now only two offset args: one for the offset that is included in bounds checking and swizzling, to be split between the instruction's voffset and immoffset fields, and one for the offset that is excluded from bounds checking and swizzling, to go into the instruction's soffset field. The AMDISD::BUFFER_* SD nodes always have an index operand, all three offset operands, combined cachepolicy operand, and an extra idxen operand. The obsolescent llvm.amdgcn.buffer.* intrinsics continue to work. Subscribers: arsenm, kzhuravl, wdng, nhaehnle, yaxunl, dstuttard, t-tye, jfb, llvm-commits Differential Revision: https://reviews.llvm.org/D50306 Change-Id: If897ea7dc34fcbf4d5496e98cc99a934f62fc205 llvm-svn: 340269
2018-08-21 19:07:10 +08:00
main_body:
[AMDGPU] Handle OR as operand of raw load/store Summary: Use isBaseWithConstantOffset() which handles OR as an operand to llvm.amdgcn.raw.buffer.load and llvm.amdgcn.raw.buffer.store. Change-Id: Ifefb9dc5ded8710d333df07ab1900b230e33539a Reviewers: nhaehnle, mareko, arsenm Reviewed By: arsenm Subscribers: arsenm, kzhuravl, jvesely, wdng, nhaehnle, yaxunl, dstuttard, tpr, t-tye, llvm-commits Differential Revision: https://reviews.llvm.org/D55999 llvm-svn: 350208
2019-01-02 17:47:41 +08:00
%a = shl i32 %inp, 4
[AMDGPU] New buffer intrinsics Summary: This commit adds new intrinsics llvm.amdgcn.raw.buffer.load llvm.amdgcn.raw.buffer.load.format llvm.amdgcn.raw.buffer.load.format.d16 llvm.amdgcn.struct.buffer.load llvm.amdgcn.struct.buffer.load.format llvm.amdgcn.struct.buffer.load.format.d16 llvm.amdgcn.raw.buffer.store llvm.amdgcn.raw.buffer.store.format llvm.amdgcn.raw.buffer.store.format.d16 llvm.amdgcn.struct.buffer.store llvm.amdgcn.struct.buffer.store.format llvm.amdgcn.struct.buffer.store.format.d16 llvm.amdgcn.raw.buffer.atomic.* llvm.amdgcn.struct.buffer.atomic.* with the following changes from the llvm.amdgcn.buffer.* intrinsics: * there are separate raw and struct versions: raw does not have an index arg and sets idxen=0 in the instruction, and struct always sets idxen=1 in the instruction even if the index is 0, to allow for the fact that gfx9 does bounds checking differently depending on whether idxen is set; * there is a combined cachepolicy arg (glc+slc) * there are now only two offset args: one for the offset that is included in bounds checking and swizzling, to be split between the instruction's voffset and immoffset fields, and one for the offset that is excluded from bounds checking and swizzling, to go into the instruction's soffset field. The AMDISD::BUFFER_* SD nodes always have an index operand, all three offset operands, combined cachepolicy operand, and an extra idxen operand. The obsolescent llvm.amdgcn.buffer.* intrinsics continue to work. Subscribers: arsenm, kzhuravl, wdng, nhaehnle, yaxunl, dstuttard, t-tye, jfb, llvm-commits Differential Revision: https://reviews.llvm.org/D50306 Change-Id: If897ea7dc34fcbf4d5496e98cc99a934f62fc205 llvm-svn: 340269
2018-08-21 19:07:10 +08:00
%a1 = add i32 %a, 4
%a2 = add i32 %a, 12
%vr1 = call <2 x float> @llvm.amdgcn.raw.buffer.load.v2f32(<4 x i32> %rsrc, i32 %a1, i32 0, i32 0)
%vr2 = call <2 x float> @llvm.amdgcn.raw.buffer.load.v2f32(<4 x i32> %rsrc, i32 %a2, i32 0, i32 0)
%r1 = extractelement <2 x float> %vr1, i32 0
%r2 = extractelement <2 x float> %vr1, i32 1
%r3 = extractelement <2 x float> %vr2, i32 0
%r4 = extractelement <2 x float> %vr2, i32 1
call void @llvm.amdgcn.exp.f32(i32 0, i32 15, float %r1, float %r2, float %r3, float %r4, i1 true, i1 true)
ret void
}
;CHECK-LABEL: {{^}}buffer_load_x1_offset_merged:
;CHECK-NEXT: %bb.
;CHECK-NEXT: buffer_load_dwordx4 v[{{[0-9]}}:{{[0-9]}}], off, s[0:3], 0 offset:4
;CHECK-NEXT: buffer_load_dwordx2 v[{{[0-9]}}:{{[0-9]}}], off, s[0:3], 0 offset:28
;CHECK: s_waitcnt
define amdgpu_ps void @buffer_load_x1_offset_merged(<4 x i32> inreg %rsrc) {
main_body:
%r1 = call float @llvm.amdgcn.raw.buffer.load.f32(<4 x i32> %rsrc, i32 4, i32 0, i32 0)
%r2 = call float @llvm.amdgcn.raw.buffer.load.f32(<4 x i32> %rsrc, i32 8, i32 0, i32 0)
%r3 = call float @llvm.amdgcn.raw.buffer.load.f32(<4 x i32> %rsrc, i32 12, i32 0, i32 0)
%r4 = call float @llvm.amdgcn.raw.buffer.load.f32(<4 x i32> %rsrc, i32 16, i32 0, i32 0)
%r5 = call float @llvm.amdgcn.raw.buffer.load.f32(<4 x i32> %rsrc, i32 28, i32 0, i32 0)
%r6 = call float @llvm.amdgcn.raw.buffer.load.f32(<4 x i32> %rsrc, i32 32, i32 0, i32 0)
call void @llvm.amdgcn.exp.f32(i32 0, i32 15, float %r1, float %r2, float %r3, float %r4, i1 true, i1 true)
call void @llvm.amdgcn.exp.f32(i32 0, i32 15, float %r5, float %r6, float undef, float undef, i1 true, i1 true)
ret void
}
;CHECK-LABEL: {{^}}buffer_load_x2_offset_merged:
;CHECK-NEXT: %bb.
;CHECK-NEXT: buffer_load_dwordx4 v[{{[0-9]}}:{{[0-9]}}], off, s[0:3], 0 offset:4
;CHECK: s_waitcnt
define amdgpu_ps void @buffer_load_x2_offset_merged(<4 x i32> inreg %rsrc) {
main_body:
%vr1 = call <2 x float> @llvm.amdgcn.raw.buffer.load.v2f32(<4 x i32> %rsrc, i32 4, i32 0, i32 0)
%vr2 = call <2 x float> @llvm.amdgcn.raw.buffer.load.v2f32(<4 x i32> %rsrc, i32 12, i32 0, i32 0)
%r1 = extractelement <2 x float> %vr1, i32 0
%r2 = extractelement <2 x float> %vr1, i32 1
%r3 = extractelement <2 x float> %vr2, i32 0
%r4 = extractelement <2 x float> %vr2, i32 1
call void @llvm.amdgcn.exp.f32(i32 0, i32 15, float %r1, float %r2, float %r3, float %r4, i1 true, i1 true)
ret void
}
[AMDGPU] Allow int types for MUBUF vdata Summary: Previously the new llvm.amdgcn.raw/struct.buffer.load/store intrinsics only allowed float types for the data to be loaded or stored, which sometimes meant the frontend needed to generate a bitcast. In this, the new intrinsics copied the old buffer intrinsics. This commit extends the new intrinsics to allow int types as well. Subscribers: arsenm, kzhuravl, wdng, nhaehnle, yaxunl, dstuttard, t-tye, llvm-commits Differential Revision: https://reviews.llvm.org/D50315 Change-Id: I8202af2d036455553681dcbb3d7d32ae273f8f85 llvm-svn: 340270
2018-08-21 19:08:12 +08:00
;CHECK-LABEL: {{^}}buffer_load_int:
;CHECK: buffer_load_dwordx4 v[0:3], off, s[0:3], 0
;CHECK: buffer_load_dwordx2 v[4:5], off, s[0:3], 0 glc
;CHECK: buffer_load_dword v6, off, s[0:3], 0 slc
;CHECK: s_waitcnt
define amdgpu_ps {<4 x float>, <2 x float>, float} @buffer_load_int(<4 x i32> inreg) {
main_body:
%data = call <4 x i32> @llvm.amdgcn.raw.buffer.load.v4i32(<4 x i32> %0, i32 0, i32 0, i32 0)
%data_glc = call <2 x i32> @llvm.amdgcn.raw.buffer.load.v2i32(<4 x i32> %0, i32 0, i32 0, i32 1)
%data_slc = call i32 @llvm.amdgcn.raw.buffer.load.i32(<4 x i32> %0, i32 0, i32 0, i32 2)
%fdata = bitcast <4 x i32> %data to <4 x float>
%fdata_glc = bitcast <2 x i32> %data_glc to <2 x float>
%fdata_slc = bitcast i32 %data_slc to float
%r0 = insertvalue {<4 x float>, <2 x float>, float} undef, <4 x float> %fdata, 0
%r1 = insertvalue {<4 x float>, <2 x float>, float} %r0, <2 x float> %fdata_glc, 1
%r2 = insertvalue {<4 x float>, <2 x float>, float} %r1, float %fdata_slc, 2
ret {<4 x float>, <2 x float>, float} %r2
}
[AMDGPU] New buffer intrinsics Summary: This commit adds new intrinsics llvm.amdgcn.raw.buffer.load llvm.amdgcn.raw.buffer.load.format llvm.amdgcn.raw.buffer.load.format.d16 llvm.amdgcn.struct.buffer.load llvm.amdgcn.struct.buffer.load.format llvm.amdgcn.struct.buffer.load.format.d16 llvm.amdgcn.raw.buffer.store llvm.amdgcn.raw.buffer.store.format llvm.amdgcn.raw.buffer.store.format.d16 llvm.amdgcn.struct.buffer.store llvm.amdgcn.struct.buffer.store.format llvm.amdgcn.struct.buffer.store.format.d16 llvm.amdgcn.raw.buffer.atomic.* llvm.amdgcn.struct.buffer.atomic.* with the following changes from the llvm.amdgcn.buffer.* intrinsics: * there are separate raw and struct versions: raw does not have an index arg and sets idxen=0 in the instruction, and struct always sets idxen=1 in the instruction even if the index is 0, to allow for the fact that gfx9 does bounds checking differently depending on whether idxen is set; * there is a combined cachepolicy arg (glc+slc) * there are now only two offset args: one for the offset that is included in bounds checking and swizzling, to be split between the instruction's voffset and immoffset fields, and one for the offset that is excluded from bounds checking and swizzling, to go into the instruction's soffset field. The AMDISD::BUFFER_* SD nodes always have an index operand, all three offset operands, combined cachepolicy operand, and an extra idxen operand. The obsolescent llvm.amdgcn.buffer.* intrinsics continue to work. Subscribers: arsenm, kzhuravl, wdng, nhaehnle, yaxunl, dstuttard, t-tye, jfb, llvm-commits Differential Revision: https://reviews.llvm.org/D50306 Change-Id: If897ea7dc34fcbf4d5496e98cc99a934f62fc205 llvm-svn: 340269
2018-08-21 19:07:10 +08:00
declare float @llvm.amdgcn.raw.buffer.load.f32(<4 x i32>, i32, i32, i32) #0
declare <2 x float> @llvm.amdgcn.raw.buffer.load.v2f32(<4 x i32>, i32, i32, i32) #0
declare <4 x float> @llvm.amdgcn.raw.buffer.load.v4f32(<4 x i32>, i32, i32, i32) #0
[AMDGPU] Allow int types for MUBUF vdata Summary: Previously the new llvm.amdgcn.raw/struct.buffer.load/store intrinsics only allowed float types for the data to be loaded or stored, which sometimes meant the frontend needed to generate a bitcast. In this, the new intrinsics copied the old buffer intrinsics. This commit extends the new intrinsics to allow int types as well. Subscribers: arsenm, kzhuravl, wdng, nhaehnle, yaxunl, dstuttard, t-tye, llvm-commits Differential Revision: https://reviews.llvm.org/D50315 Change-Id: I8202af2d036455553681dcbb3d7d32ae273f8f85 llvm-svn: 340270
2018-08-21 19:08:12 +08:00
declare i32 @llvm.amdgcn.raw.buffer.load.i32(<4 x i32>, i32, i32, i32) #0
declare <2 x i32> @llvm.amdgcn.raw.buffer.load.v2i32(<4 x i32>, i32, i32, i32) #0
declare <4 x i32> @llvm.amdgcn.raw.buffer.load.v4i32(<4 x i32>, i32, i32, i32) #0
[AMDGPU] New buffer intrinsics Summary: This commit adds new intrinsics llvm.amdgcn.raw.buffer.load llvm.amdgcn.raw.buffer.load.format llvm.amdgcn.raw.buffer.load.format.d16 llvm.amdgcn.struct.buffer.load llvm.amdgcn.struct.buffer.load.format llvm.amdgcn.struct.buffer.load.format.d16 llvm.amdgcn.raw.buffer.store llvm.amdgcn.raw.buffer.store.format llvm.amdgcn.raw.buffer.store.format.d16 llvm.amdgcn.struct.buffer.store llvm.amdgcn.struct.buffer.store.format llvm.amdgcn.struct.buffer.store.format.d16 llvm.amdgcn.raw.buffer.atomic.* llvm.amdgcn.struct.buffer.atomic.* with the following changes from the llvm.amdgcn.buffer.* intrinsics: * there are separate raw and struct versions: raw does not have an index arg and sets idxen=0 in the instruction, and struct always sets idxen=1 in the instruction even if the index is 0, to allow for the fact that gfx9 does bounds checking differently depending on whether idxen is set; * there is a combined cachepolicy arg (glc+slc) * there are now only two offset args: one for the offset that is included in bounds checking and swizzling, to be split between the instruction's voffset and immoffset fields, and one for the offset that is excluded from bounds checking and swizzling, to go into the instruction's soffset field. The AMDISD::BUFFER_* SD nodes always have an index operand, all three offset operands, combined cachepolicy operand, and an extra idxen operand. The obsolescent llvm.amdgcn.buffer.* intrinsics continue to work. Subscribers: arsenm, kzhuravl, wdng, nhaehnle, yaxunl, dstuttard, t-tye, jfb, llvm-commits Differential Revision: https://reviews.llvm.org/D50306 Change-Id: If897ea7dc34fcbf4d5496e98cc99a934f62fc205 llvm-svn: 340269
2018-08-21 19:07:10 +08:00
declare void @llvm.amdgcn.exp.f32(i32, i32, float, float, float, float, i1, i1) #0
attributes #0 = { nounwind readonly }