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/merge-stores.ll

683 lines
28 KiB
LLVM
Raw Normal View History

AMDGPU: Cleanup subtarget features Try to avoid mutually exclusive features. Don't use a real default GPU, and use a fake "generic". The goal is to make it easier to see which set of features are incompatible between feature strings. Most of the test changes are due to random scheduling changes from not having a default fullspeed model. llvm-svn: 310258
2017-08-07 22:58:04 +08:00
; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=tahiti -verify-machineinstrs -amdgpu-load-store-vectorizer=0 < %s | FileCheck -check-prefix=SI -check-prefix=GCN -check-prefix=GCN-AA %s
; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=bonaire -verify-machineinstrs -amdgpu-load-store-vectorizer=0 < %s | FileCheck -check-prefix=SI -check-prefix=GCN -check-prefix=GCN-AA %s
Add target hook to allow merging stores of nonzero constants On GPU targets, materializing constants is cheap and stores are expensive, so only doing this for zero vectors was silly. Most of the new testcases aren't optimally merged, and are for later improvements. llvm-svn: 238108
2015-05-24 08:51:27 +08:00
AMDGPU: Run LoadStoreVectorizer pass by default llvm-svn: 281112
2016-09-10 06:29:28 +08:00
; This test is mostly to test DAG store merging, so disable the vectorizer.
Add target hook to allow merging stores of nonzero constants On GPU targets, materializing constants is cheap and stores are expensive, so only doing this for zero vectors was silly. Most of the new testcases aren't optimally merged, and are for later improvements. llvm-svn: 238108
2015-05-24 08:51:27 +08:00
; Run with devices with different unaligned load restrictions.
; TODO: Vector element tests
; TODO: Non-zero base offset for load and store combinations
; TODO: Same base addrspacecasted
; GCN-LABEL: {{^}}merge_global_store_2_constants_i8:
[DAG] Fix typo preventing some stores merges to truncated stores. Check the actual memory type stored and not the extended value size when considering if truncated store merge is worthwhile. Reviewers: efriedma, RKSimon, spatel, jyknight Reviewed By: efriedma Subscribers: llvm-commits, nhaehnle Differential Revision: https://reviews.llvm.org/D35623 llvm-svn: 308833
2017-07-23 10:06:28 +08:00
; GCN: buffer_store_short
Add target hook to allow merging stores of nonzero constants On GPU targets, materializing constants is cheap and stores are expensive, so only doing this for zero vectors was silly. Most of the new testcases aren't optimally merged, and are for later improvements. llvm-svn: 238108
2015-05-24 08:51:27 +08:00
; GCN: s_endpgm
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 @merge_global_store_2_constants_i8(i8 addrspace(1)* %out) #0 {
Add target hook to allow merging stores of nonzero constants On GPU targets, materializing constants is cheap and stores are expensive, so only doing this for zero vectors was silly. Most of the new testcases aren't optimally merged, and are for later improvements. llvm-svn: 238108
2015-05-24 08:51:27 +08:00
%out.gep.1 = getelementptr i8, i8 addrspace(1)* %out, i32 1
store i8 123, i8 addrspace(1)* %out.gep.1
store i8 456, i8 addrspace(1)* %out, align 2
ret void
}
; GCN-LABEL: {{^}}merge_global_store_2_constants_i8_natural_align:
; GCN: buffer_store_byte
; GCN: buffer_store_byte
; GCN: s_endpgm
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 @merge_global_store_2_constants_i8_natural_align(i8 addrspace(1)* %out) #0 {
Add target hook to allow merging stores of nonzero constants On GPU targets, materializing constants is cheap and stores are expensive, so only doing this for zero vectors was silly. Most of the new testcases aren't optimally merged, and are for later improvements. llvm-svn: 238108
2015-05-24 08:51:27 +08:00
%out.gep.1 = getelementptr i8, i8 addrspace(1)* %out, i32 1
store i8 123, i8 addrspace(1)* %out.gep.1
store i8 456, i8 addrspace(1)* %out
ret void
}
; GCN-LABEL: {{^}}merge_global_store_2_constants_i16:
; GCN: buffer_store_dword v
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 @merge_global_store_2_constants_i16(i16 addrspace(1)* %out) #0 {
Add target hook to allow merging stores of nonzero constants On GPU targets, materializing constants is cheap and stores are expensive, so only doing this for zero vectors was silly. Most of the new testcases aren't optimally merged, and are for later improvements. llvm-svn: 238108
2015-05-24 08:51:27 +08:00
%out.gep.1 = getelementptr i16, i16 addrspace(1)* %out, i32 1
store i16 123, i16 addrspace(1)* %out.gep.1
store i16 456, i16 addrspace(1)* %out, align 4
ret void
}
; GCN-LABEL: {{^}}merge_global_store_2_constants_0_i16:
; GCN: buffer_store_dword v
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 @merge_global_store_2_constants_0_i16(i16 addrspace(1)* %out) #0 {
Add target hook to allow merging stores of nonzero constants On GPU targets, materializing constants is cheap and stores are expensive, so only doing this for zero vectors was silly. Most of the new testcases aren't optimally merged, and are for later improvements. llvm-svn: 238108
2015-05-24 08:51:27 +08:00
%out.gep.1 = getelementptr i16, i16 addrspace(1)* %out, i32 1
store i16 0, i16 addrspace(1)* %out.gep.1
store i16 0, i16 addrspace(1)* %out, align 4
ret void
}
; GCN-LABEL: {{^}}merge_global_store_2_constants_i16_natural_align:
; GCN: buffer_store_short
; GCN: buffer_store_short
; GCN: s_endpgm
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 @merge_global_store_2_constants_i16_natural_align(i16 addrspace(1)* %out) #0 {
Add target hook to allow merging stores of nonzero constants On GPU targets, materializing constants is cheap and stores are expensive, so only doing this for zero vectors was silly. Most of the new testcases aren't optimally merged, and are for later improvements. llvm-svn: 238108
2015-05-24 08:51:27 +08:00
%out.gep.1 = getelementptr i16, i16 addrspace(1)* %out, i32 1
store i16 123, i16 addrspace(1)* %out.gep.1
store i16 456, i16 addrspace(1)* %out
ret void
}
; GCN-LABEL: {{^}}merge_global_store_2_constants_i32:
AMDGPU: Distribute SGPR->VGPR copies of REG_SEQUENCE Make the REG_SEQUENCE be a VGPR, and do the register class copy first. llvm-svn: 251855
2015-11-03 07:15:42 +08:00
; SI-DAG: v_mov_b32_e32 v[[LO:[0-9]+]], 0x1c8
; SI-DAG: v_mov_b32_e32 v[[HI:[0-9]+]], 0x7b
Add target hook to allow merging stores of nonzero constants On GPU targets, materializing constants is cheap and stores are expensive, so only doing this for zero vectors was silly. Most of the new testcases aren't optimally merged, and are for later improvements. llvm-svn: 238108
2015-05-24 08:51:27 +08:00
; GCN: buffer_store_dwordx2 v{{\[}}[[LO]]:[[HI]]{{\]}}
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 @merge_global_store_2_constants_i32(i32 addrspace(1)* %out) #0 {
Add target hook to allow merging stores of nonzero constants On GPU targets, materializing constants is cheap and stores are expensive, so only doing this for zero vectors was silly. Most of the new testcases aren't optimally merged, and are for later improvements. llvm-svn: 238108
2015-05-24 08:51:27 +08:00
%out.gep.1 = getelementptr i32, i32 addrspace(1)* %out, i32 1
store i32 123, i32 addrspace(1)* %out.gep.1
store i32 456, i32 addrspace(1)* %out
ret void
}
; GCN-LABEL: {{^}}merge_global_store_2_constants_i32_f32:
; GCN: buffer_store_dwordx2
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 @merge_global_store_2_constants_i32_f32(i32 addrspace(1)* %out) #0 {
Add target hook to allow merging stores of nonzero constants On GPU targets, materializing constants is cheap and stores are expensive, so only doing this for zero vectors was silly. Most of the new testcases aren't optimally merged, and are for later improvements. llvm-svn: 238108
2015-05-24 08:51:27 +08:00
%out.gep.1 = getelementptr i32, i32 addrspace(1)* %out, i32 1
%out.gep.1.bc = bitcast i32 addrspace(1)* %out.gep.1 to float addrspace(1)*
store float 1.0, float addrspace(1)* %out.gep.1.bc
store i32 456, i32 addrspace(1)* %out
ret void
}
; GCN-LABEL: {{^}}merge_global_store_2_constants_f32_i32:
AMDGPU: Distribute SGPR->VGPR copies of REG_SEQUENCE Make the REG_SEQUENCE be a VGPR, and do the register class copy first. llvm-svn: 251855
2015-11-03 07:15:42 +08:00
; SI-DAG: v_mov_b32_e32 v[[VLO:[0-9]+]], 4.0
; SI-DAG: v_mov_b32_e32 v[[VHI:[0-9]+]], 0x7b
Revert "Revert "Fix merges of non-zero vector stores"" Reapply r239539. Don't assume the collected number of stores is the same vector size. Just take the first N stores to fill the vector. llvm-svn: 239825
2015-06-16 23:51:48 +08:00
; GCN: buffer_store_dwordx2 v{{\[}}[[VLO]]:[[VHI]]{{\]}}
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 @merge_global_store_2_constants_f32_i32(float addrspace(1)* %out) #0 {
Add target hook to allow merging stores of nonzero constants On GPU targets, materializing constants is cheap and stores are expensive, so only doing this for zero vectors was silly. Most of the new testcases aren't optimally merged, and are for later improvements. llvm-svn: 238108
2015-05-24 08:51:27 +08:00
%out.gep.1 = getelementptr float, float addrspace(1)* %out, i32 1
%out.gep.1.bc = bitcast float addrspace(1)* %out.gep.1 to i32 addrspace(1)*
store i32 123, i32 addrspace(1)* %out.gep.1.bc
store float 4.0, float addrspace(1)* %out
ret void
}
; GCN-LABEL: {{^}}merge_global_store_4_constants_i32:
Revert "Revert "Fix merges of non-zero vector stores"" Reapply r239539. Don't assume the collected number of stores is the same vector size. Just take the first N stores to fill the vector. llvm-svn: 239825
2015-06-16 23:51:48 +08:00
; GCN-DAG: v_mov_b32_e32 v[[HI:[0-9]+]], 0x14d{{$}}
; GCN-DAG: v_mov_b32_e32 v{{[0-9]+}}, 0x1c8{{$}}
; GCN-DAG: v_mov_b32_e32 v{{[0-9]+}}, 0x7b{{$}}
; GCN-DAG: v_mov_b32_e32 v[[LO:[0-9]+]], 0x4d2{{$}}
; GCN: buffer_store_dwordx4 v{{\[}}[[LO]]:[[HI]]{{\]}}
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 @merge_global_store_4_constants_i32(i32 addrspace(1)* %out) #0 {
Add target hook to allow merging stores of nonzero constants On GPU targets, materializing constants is cheap and stores are expensive, so only doing this for zero vectors was silly. Most of the new testcases aren't optimally merged, and are for later improvements. llvm-svn: 238108
2015-05-24 08:51:27 +08:00
%out.gep.1 = getelementptr i32, i32 addrspace(1)* %out, i32 1
%out.gep.2 = getelementptr i32, i32 addrspace(1)* %out, i32 2
%out.gep.3 = getelementptr i32, i32 addrspace(1)* %out, i32 3
store i32 123, i32 addrspace(1)* %out.gep.1
store i32 456, i32 addrspace(1)* %out.gep.2
store i32 333, i32 addrspace(1)* %out.gep.3
store i32 1234, i32 addrspace(1)* %out
ret void
}
; GCN-LABEL: {{^}}merge_global_store_4_constants_f32_order:
DAGCombiner: Replace store of FP constant after attemping store merges If storing multiple FP constants, some subset of the stores would be replaced with integers due to visit order, so MergeConsecutiveStores would only partially merge these. llvm-svn: 248169
2015-09-21 23:59:46 +08:00
; GCN: buffer_store_dwordx4
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 @merge_global_store_4_constants_f32_order(float addrspace(1)* %out) #0 {
Add target hook to allow merging stores of nonzero constants On GPU targets, materializing constants is cheap and stores are expensive, so only doing this for zero vectors was silly. Most of the new testcases aren't optimally merged, and are for later improvements. llvm-svn: 238108
2015-05-24 08:51:27 +08:00
%out.gep.1 = getelementptr float, float addrspace(1)* %out, i32 1
%out.gep.2 = getelementptr float, float addrspace(1)* %out, i32 2
%out.gep.3 = getelementptr float, float addrspace(1)* %out, i32 3
store float 8.0, float addrspace(1)* %out
store float 1.0, float addrspace(1)* %out.gep.1
store float 2.0, float addrspace(1)* %out.gep.2
store float 4.0, float addrspace(1)* %out.gep.3
ret void
}
DAGCombiner: Replace store of FP constant after attemping store merges If storing multiple FP constants, some subset of the stores would be replaced with integers due to visit order, so MergeConsecutiveStores would only partially merge these. llvm-svn: 248169
2015-09-21 23:59:46 +08:00
; First store is out of order.
Add target hook to allow merging stores of nonzero constants On GPU targets, materializing constants is cheap and stores are expensive, so only doing this for zero vectors was silly. Most of the new testcases aren't optimally merged, and are for later improvements. llvm-svn: 238108
2015-05-24 08:51:27 +08:00
; GCN-LABEL: {{^}}merge_global_store_4_constants_f32:
DAGCombiner: Replace store of FP constant after attemping store merges If storing multiple FP constants, some subset of the stores would be replaced with integers due to visit order, so MergeConsecutiveStores would only partially merge these. llvm-svn: 248169
2015-09-21 23:59:46 +08:00
; GCN: buffer_store_dwordx4
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 @merge_global_store_4_constants_f32(float addrspace(1)* %out) #0 {
Add target hook to allow merging stores of nonzero constants On GPU targets, materializing constants is cheap and stores are expensive, so only doing this for zero vectors was silly. Most of the new testcases aren't optimally merged, and are for later improvements. llvm-svn: 238108
2015-05-24 08:51:27 +08:00
%out.gep.1 = getelementptr float, float addrspace(1)* %out, i32 1
%out.gep.2 = getelementptr float, float addrspace(1)* %out, i32 2
%out.gep.3 = getelementptr float, float addrspace(1)* %out, i32 3
store float 1.0, float addrspace(1)* %out.gep.1
store float 2.0, float addrspace(1)* %out.gep.2
store float 4.0, float addrspace(1)* %out.gep.3
store float 8.0, float addrspace(1)* %out
ret void
}
DAGCombiner: Replace store of FP constant after attemping store merges If storing multiple FP constants, some subset of the stores would be replaced with integers due to visit order, so MergeConsecutiveStores would only partially merge these. llvm-svn: 248169
2015-09-21 23:59:46 +08:00
; GCN-LABEL: {{^}}merge_global_store_4_constants_mixed_i32_f32:
In visitSTORE, always use FindBetterChain, rather than only when UseAA is enabled. Recommiting with compiler time improvements Recommitting after fixup of 32-bit aliasing sign offset bug in DAGCombiner. * Simplify Consecutive Merge Store Candidate Search Now that address aliasing is much less conservative, push through simplified store merging search and chain alias analysis which only checks for parallel stores through the chain subgraph. This is cleaner as the separation of non-interfering loads/stores from the store-merging logic. When merging stores search up the chain through a single load, and finds all possible stores by looking down from through a load and a TokenFactor to all stores visited. This improves the quality of the output SelectionDAG and the output Codegen (save perhaps for some ARM cases where we correctly constructs wider loads, but then promotes them to float operations which appear but requires more expensive constant generation). Some minor peephole optimizations to deal with improved SubDAG shapes (listed below) Additional Minor Changes: 1. Finishes removing unused AliasLoad code 2. Unifies the chain aggregation in the merged stores across code paths 3. Re-add the Store node to the worklist after calling SimplifyDemandedBits. 4. Increase GatherAllAliasesMaxDepth from 6 to 18. That number is arbitrary, but seems sufficient to not cause regressions in tests. 5. Remove Chain dependencies of Memory operations on CopyfromReg nodes as these are captured by data dependence 6. Forward loads-store values through tokenfactors containing {CopyToReg,CopyFromReg} Values. 7. Peephole to convert buildvector of extract_vector_elt to extract_subvector if possible (see CodeGen/AArch64/store-merge.ll) 8. Store merging for the ARM target is restricted to 32-bit as some in some contexts invalid 64-bit operations are being generated. This can be removed once appropriate checks are added. This finishes the change Matt Arsenault started in r246307 and jyknight's original patch. Many tests required some changes as memory operations are now reorderable, improving load-store forwarding. One test in particular is worth noting: CodeGen/PowerPC/ppc64-align-long-double.ll - Improved load-store forwarding converts a load-store pair into a parallel store and a memory-realized bitcast of the same value. However, because we lose the sharing of the explicit and implicit store values we must create another local store. A similar transformation happens before SelectionDAG as well. Reviewers: arsenm, hfinkel, tstellarAMD, jyknight, nhaehnle llvm-svn: 297695
2017-03-14 08:34:14 +08:00
; GCN-AA: buffer_store_dwordx4 v
DAGCombiner: Replace store of FP constant after attemping store merges If storing multiple FP constants, some subset of the stores would be replaced with integers due to visit order, so MergeConsecutiveStores would only partially merge these. llvm-svn: 248169
2015-09-21 23:59:46 +08:00
; GCN: s_endpgm
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 @merge_global_store_4_constants_mixed_i32_f32(float addrspace(1)* %out) #0 {
DAGCombiner: Replace store of FP constant after attemping store merges If storing multiple FP constants, some subset of the stores would be replaced with integers due to visit order, so MergeConsecutiveStores would only partially merge these. llvm-svn: 248169
2015-09-21 23:59:46 +08:00
%out.gep.1 = getelementptr float, float addrspace(1)* %out, i32 1
%out.gep.2 = getelementptr float, float addrspace(1)* %out, i32 2
%out.gep.3 = getelementptr float, float addrspace(1)* %out, i32 3
%out.gep.1.bc = bitcast float addrspace(1)* %out.gep.1 to i32 addrspace(1)*
%out.gep.3.bc = bitcast float addrspace(1)* %out.gep.3 to i32 addrspace(1)*
store i32 11, i32 addrspace(1)* %out.gep.1.bc
store float 2.0, float addrspace(1)* %out.gep.2
store i32 17, i32 addrspace(1)* %out.gep.3.bc
store float 8.0, float addrspace(1)* %out
ret void
}
Add target hook to allow merging stores of nonzero constants On GPU targets, materializing constants is cheap and stores are expensive, so only doing this for zero vectors was silly. Most of the new testcases aren't optimally merged, and are for later improvements. llvm-svn: 238108
2015-05-24 08:51:27 +08:00
; GCN-LABEL: {{^}}merge_global_store_3_constants_i32:
; SI-DAG: buffer_store_dwordx2
; SI-DAG: buffer_store_dword
; SI-NOT: buffer_store_dword
; GCN: s_endpgm
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 @merge_global_store_3_constants_i32(i32 addrspace(1)* %out) #0 {
Add target hook to allow merging stores of nonzero constants On GPU targets, materializing constants is cheap and stores are expensive, so only doing this for zero vectors was silly. Most of the new testcases aren't optimally merged, and are for later improvements. llvm-svn: 238108
2015-05-24 08:51:27 +08:00
%out.gep.1 = getelementptr i32, i32 addrspace(1)* %out, i32 1
%out.gep.2 = getelementptr i32, i32 addrspace(1)* %out, i32 2
store i32 123, i32 addrspace(1)* %out.gep.1
store i32 456, i32 addrspace(1)* %out.gep.2
store i32 1234, i32 addrspace(1)* %out
ret void
}
; GCN-LABEL: {{^}}merge_global_store_2_constants_i64:
AMDGPU: Make v2i64/v2f64 legal types. They can be loaded and stored, so count them as legal. This is mostly to fix a number of common cases for load/store merging. llvm-svn: 254086
2015-11-26 03:58:34 +08:00
; GCN: buffer_store_dwordx4
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 @merge_global_store_2_constants_i64(i64 addrspace(1)* %out) #0 {
Add target hook to allow merging stores of nonzero constants On GPU targets, materializing constants is cheap and stores are expensive, so only doing this for zero vectors was silly. Most of the new testcases aren't optimally merged, and are for later improvements. llvm-svn: 238108
2015-05-24 08:51:27 +08:00
%out.gep.1 = getelementptr i64, i64 addrspace(1)* %out, i64 1
store i64 123, i64 addrspace(1)* %out.gep.1
store i64 456, i64 addrspace(1)* %out
ret void
}
; GCN-LABEL: {{^}}merge_global_store_4_constants_i64:
AMDGPU: Make v2i64/v2f64 legal types. They can be loaded and stored, so count them as legal. This is mostly to fix a number of common cases for load/store merging. llvm-svn: 254086
2015-11-26 03:58:34 +08:00
; GCN: buffer_store_dwordx4
; GCN: buffer_store_dwordx4
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 @merge_global_store_4_constants_i64(i64 addrspace(1)* %out) #0 {
Add target hook to allow merging stores of nonzero constants On GPU targets, materializing constants is cheap and stores are expensive, so only doing this for zero vectors was silly. Most of the new testcases aren't optimally merged, and are for later improvements. llvm-svn: 238108
2015-05-24 08:51:27 +08:00
%out.gep.1 = getelementptr i64, i64 addrspace(1)* %out, i64 1
%out.gep.2 = getelementptr i64, i64 addrspace(1)* %out, i64 2
%out.gep.3 = getelementptr i64, i64 addrspace(1)* %out, i64 3
store i64 123, i64 addrspace(1)* %out.gep.1
store i64 456, i64 addrspace(1)* %out.gep.2
store i64 333, i64 addrspace(1)* %out.gep.3
store i64 1234, i64 addrspace(1)* %out
ret void
}
; GCN-LABEL: {{^}}merge_global_store_2_adjacent_loads_i32:
; GCN: buffer_load_dwordx2 [[LOAD:v\[[0-9]+:[0-9]+\]]]
; GCN: buffer_store_dwordx2 [[LOAD]]
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 @merge_global_store_2_adjacent_loads_i32(i32 addrspace(1)* %out, i32 addrspace(1)* %in) #0 {
Add target hook to allow merging stores of nonzero constants On GPU targets, materializing constants is cheap and stores are expensive, so only doing this for zero vectors was silly. Most of the new testcases aren't optimally merged, and are for later improvements. llvm-svn: 238108
2015-05-24 08:51:27 +08:00
%out.gep.1 = getelementptr i32, i32 addrspace(1)* %out, i32 1
%in.gep.1 = getelementptr i32, i32 addrspace(1)* %in, i32 1
%lo = load i32, i32 addrspace(1)* %in
%hi = load i32, i32 addrspace(1)* %in.gep.1
store i32 %lo, i32 addrspace(1)* %out
store i32 %hi, i32 addrspace(1)* %out.gep.1
ret void
}
; GCN-LABEL: {{^}}merge_global_store_2_adjacent_loads_i32_nonzero_base:
AMDGPU/SI: Assembler: Unify parsing/printing of operands. Summary: The goal is for each operand type to have its own parse function and at the same time share common code for tracking state as different instruction types share operand types (e.g. glc/glc_flat, etc). Introduce parseAMDGPUOperand which can parse any optional operand. DPP and Clamp/OMod have custom handling for now. Sam also suggested to have class hierarchy for operand types instead of table. This can be done in separate change. Remove parseVOP3OptionalOps, parseDS*OptionalOps, parseFlatOptionalOps, parseMubufOptionalOps, parseDPPOptionalOps. Reduce number of definitions of AsmOperand's and MatchClasses' by using common base class. Rename AsmMatcher/InstPrinter methods accordingly. Print immediate type when printing parsed immediate operand. Use 'off' if offset/index register is unused instead of skipping it to make it more readable (also agreed with SP3). Update tests. Reviewers: tstellarAMD, SamWot, artem.tamazov Subscribers: qcolombet, arsenm, llvm-commits Differential Revision: http://reviews.llvm.org/D19584 llvm-svn: 268015
2016-04-29 17:02:30 +08:00
; GCN: buffer_load_dwordx2 [[LOAD:v\[[0-9]+:[0-9]+\]]], off, s{{\[[0-9]+:[0-9]+\]}}, 0 offset:8
; GCN: buffer_store_dwordx2 [[LOAD]], off, s{{\[[0-9]+:[0-9]+\]}}, 0 offset:8
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 @merge_global_store_2_adjacent_loads_i32_nonzero_base(i32 addrspace(1)* %out, i32 addrspace(1)* %in) #0 {
Add target hook to allow merging stores of nonzero constants On GPU targets, materializing constants is cheap and stores are expensive, so only doing this for zero vectors was silly. Most of the new testcases aren't optimally merged, and are for later improvements. llvm-svn: 238108
2015-05-24 08:51:27 +08:00
%in.gep.0 = getelementptr i32, i32 addrspace(1)* %in, i32 2
%in.gep.1 = getelementptr i32, i32 addrspace(1)* %in, i32 3
%out.gep.0 = getelementptr i32, i32 addrspace(1)* %out, i32 2
%out.gep.1 = getelementptr i32, i32 addrspace(1)* %out, i32 3
%lo = load i32, i32 addrspace(1)* %in.gep.0
%hi = load i32, i32 addrspace(1)* %in.gep.1
store i32 %lo, i32 addrspace(1)* %out.gep.0
store i32 %hi, i32 addrspace(1)* %out.gep.1
ret void
}
; GCN-LABEL: {{^}}merge_global_store_2_adjacent_loads_shuffle_i32:
AMDGPU: Merge BUFFER_LOAD_DWORD_OFFSET into x2, x4 Summary: Only 3 (out of 48486) shaders are affected. Reviewers: arsenm, nhaehnle Subscribers: kzhuravl, wdng, yaxunl, dstuttard, tpr, t-tye, llvm-commits Differential Revision: https://reviews.llvm.org/D38951 llvm-svn: 317753
2017-11-09 09:52:36 +08:00
; GCN: buffer_load_dwordx2 v
AMDGPU: Merge BUFFER_STORE_DWORD_OFFEN/OFFSET into x2, x4 Summary: Only 56 shaders (out of 48486) are affected. Totals from affected shaders (changed stats only): SGPRS: 2420 -> 2460 (1.65 %) Spilled VGPRs: 94 -> 112 (19.15 %) Scratch size: 524 -> 528 (0.76 %) dwords per thread Code Size: 187400 -> 184992 (-1.28 %) bytes One DiRT Showdown shader spills 6 more VGPRs. One Grid Autosport shader spills 12 more VGPRs. The other 54 shaders only have a decrease in code size. (I'm ignoring the SGPR noise) Reviewers: arsenm, nhaehnle Subscribers: kzhuravl, wdng, yaxunl, dstuttard, tpr, llvm-commits, t-tye Differential Revision: https://reviews.llvm.org/D39012 llvm-svn: 317755
2017-11-09 09:52:55 +08:00
; GCN: buffer_store_dwordx2 v
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 @merge_global_store_2_adjacent_loads_shuffle_i32(i32 addrspace(1)* %out, i32 addrspace(1)* %in) #0 {
Add target hook to allow merging stores of nonzero constants On GPU targets, materializing constants is cheap and stores are expensive, so only doing this for zero vectors was silly. Most of the new testcases aren't optimally merged, and are for later improvements. llvm-svn: 238108
2015-05-24 08:51:27 +08:00
%out.gep.1 = getelementptr i32, i32 addrspace(1)* %out, i32 1
%in.gep.1 = getelementptr i32, i32 addrspace(1)* %in, i32 1
%lo = load i32, i32 addrspace(1)* %in
%hi = load i32, i32 addrspace(1)* %in.gep.1
store i32 %hi, i32 addrspace(1)* %out
store i32 %lo, i32 addrspace(1)* %out.gep.1
ret void
}
; GCN-LABEL: {{^}}merge_global_store_4_adjacent_loads_i32:
; GCN: buffer_load_dwordx4 [[LOAD:v\[[0-9]+:[0-9]+\]]]
; GCN: buffer_store_dwordx4 [[LOAD]]
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 @merge_global_store_4_adjacent_loads_i32(i32 addrspace(1)* %out, i32 addrspace(1)* %in) #0 {
Add target hook to allow merging stores of nonzero constants On GPU targets, materializing constants is cheap and stores are expensive, so only doing this for zero vectors was silly. Most of the new testcases aren't optimally merged, and are for later improvements. llvm-svn: 238108
2015-05-24 08:51:27 +08:00
%out.gep.1 = getelementptr i32, i32 addrspace(1)* %out, i32 1
%out.gep.2 = getelementptr i32, i32 addrspace(1)* %out, i32 2
%out.gep.3 = getelementptr i32, i32 addrspace(1)* %out, i32 3
%in.gep.1 = getelementptr i32, i32 addrspace(1)* %in, i32 1
%in.gep.2 = getelementptr i32, i32 addrspace(1)* %in, i32 2
%in.gep.3 = getelementptr i32, i32 addrspace(1)* %in, i32 3
%x = load i32, i32 addrspace(1)* %in
%y = load i32, i32 addrspace(1)* %in.gep.1
%z = load i32, i32 addrspace(1)* %in.gep.2
%w = load i32, i32 addrspace(1)* %in.gep.3
store i32 %x, i32 addrspace(1)* %out
store i32 %y, i32 addrspace(1)* %out.gep.1
store i32 %z, i32 addrspace(1)* %out.gep.2
store i32 %w, i32 addrspace(1)* %out.gep.3
ret void
}
; GCN-LABEL: {{^}}merge_global_store_3_adjacent_loads_i32:
; SI-DAG: buffer_load_dwordx2
; SI-DAG: buffer_load_dword v
; GCN: s_waitcnt
; SI-DAG: buffer_store_dword v
; SI-DAG: buffer_store_dwordx2 v
; GCN: s_endpgm
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 @merge_global_store_3_adjacent_loads_i32(i32 addrspace(1)* %out, i32 addrspace(1)* %in) #0 {
Add target hook to allow merging stores of nonzero constants On GPU targets, materializing constants is cheap and stores are expensive, so only doing this for zero vectors was silly. Most of the new testcases aren't optimally merged, and are for later improvements. llvm-svn: 238108
2015-05-24 08:51:27 +08:00
%out.gep.1 = getelementptr i32, i32 addrspace(1)* %out, i32 1
%out.gep.2 = getelementptr i32, i32 addrspace(1)* %out, i32 2
%in.gep.1 = getelementptr i32, i32 addrspace(1)* %in, i32 1
%in.gep.2 = getelementptr i32, i32 addrspace(1)* %in, i32 2
%x = load i32, i32 addrspace(1)* %in
%y = load i32, i32 addrspace(1)* %in.gep.1
%z = load i32, i32 addrspace(1)* %in.gep.2
store i32 %x, i32 addrspace(1)* %out
store i32 %y, i32 addrspace(1)* %out.gep.1
store i32 %z, i32 addrspace(1)* %out.gep.2
ret void
}
; GCN-LABEL: {{^}}merge_global_store_4_adjacent_loads_f32:
; GCN: buffer_load_dwordx4 [[LOAD:v\[[0-9]+:[0-9]+\]]]
; GCN: buffer_store_dwordx4 [[LOAD]]
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 @merge_global_store_4_adjacent_loads_f32(float addrspace(1)* %out, float addrspace(1)* %in) #0 {
Add target hook to allow merging stores of nonzero constants On GPU targets, materializing constants is cheap and stores are expensive, so only doing this for zero vectors was silly. Most of the new testcases aren't optimally merged, and are for later improvements. llvm-svn: 238108
2015-05-24 08:51:27 +08:00
%out.gep.1 = getelementptr float, float addrspace(1)* %out, i32 1
%out.gep.2 = getelementptr float, float addrspace(1)* %out, i32 2
%out.gep.3 = getelementptr float, float addrspace(1)* %out, i32 3
%in.gep.1 = getelementptr float, float addrspace(1)* %in, i32 1
%in.gep.2 = getelementptr float, float addrspace(1)* %in, i32 2
%in.gep.3 = getelementptr float, float addrspace(1)* %in, i32 3
%x = load float, float addrspace(1)* %in
%y = load float, float addrspace(1)* %in.gep.1
%z = load float, float addrspace(1)* %in.gep.2
%w = load float, float addrspace(1)* %in.gep.3
store float %x, float addrspace(1)* %out
store float %y, float addrspace(1)* %out.gep.1
store float %z, float addrspace(1)* %out.gep.2
store float %w, float addrspace(1)* %out.gep.3
ret void
}
; GCN-LABEL: {{^}}merge_global_store_4_adjacent_loads_i32_nonzero_base:
AMDGPU/SI: Assembler: Unify parsing/printing of operands. Summary: The goal is for each operand type to have its own parse function and at the same time share common code for tracking state as different instruction types share operand types (e.g. glc/glc_flat, etc). Introduce parseAMDGPUOperand which can parse any optional operand. DPP and Clamp/OMod have custom handling for now. Sam also suggested to have class hierarchy for operand types instead of table. This can be done in separate change. Remove parseVOP3OptionalOps, parseDS*OptionalOps, parseFlatOptionalOps, parseMubufOptionalOps, parseDPPOptionalOps. Reduce number of definitions of AsmOperand's and MatchClasses' by using common base class. Rename AsmMatcher/InstPrinter methods accordingly. Print immediate type when printing parsed immediate operand. Use 'off' if offset/index register is unused instead of skipping it to make it more readable (also agreed with SP3). Update tests. Reviewers: tstellarAMD, SamWot, artem.tamazov Subscribers: qcolombet, arsenm, llvm-commits Differential Revision: http://reviews.llvm.org/D19584 llvm-svn: 268015
2016-04-29 17:02:30 +08:00
; GCN: buffer_load_dwordx4 [[LOAD:v\[[0-9]+:[0-9]+\]]], off, s{{\[[0-9]+:[0-9]+\]}}, 0 offset:44
; GCN: buffer_store_dwordx4 [[LOAD]], off, s{{\[[0-9]+:[0-9]+\]}}, 0 offset:28
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 @merge_global_store_4_adjacent_loads_i32_nonzero_base(i32 addrspace(1)* %out, i32 addrspace(1)* %in) #0 {
Add target hook to allow merging stores of nonzero constants On GPU targets, materializing constants is cheap and stores are expensive, so only doing this for zero vectors was silly. Most of the new testcases aren't optimally merged, and are for later improvements. llvm-svn: 238108
2015-05-24 08:51:27 +08:00
%in.gep.0 = getelementptr i32, i32 addrspace(1)* %in, i32 11
%in.gep.1 = getelementptr i32, i32 addrspace(1)* %in, i32 12
%in.gep.2 = getelementptr i32, i32 addrspace(1)* %in, i32 13
%in.gep.3 = getelementptr i32, i32 addrspace(1)* %in, i32 14
%out.gep.0 = getelementptr i32, i32 addrspace(1)* %out, i32 7
%out.gep.1 = getelementptr i32, i32 addrspace(1)* %out, i32 8
%out.gep.2 = getelementptr i32, i32 addrspace(1)* %out, i32 9
%out.gep.3 = getelementptr i32, i32 addrspace(1)* %out, i32 10
%x = load i32, i32 addrspace(1)* %in.gep.0
%y = load i32, i32 addrspace(1)* %in.gep.1
%z = load i32, i32 addrspace(1)* %in.gep.2
%w = load i32, i32 addrspace(1)* %in.gep.3
store i32 %x, i32 addrspace(1)* %out.gep.0
store i32 %y, i32 addrspace(1)* %out.gep.1
store i32 %z, i32 addrspace(1)* %out.gep.2
store i32 %w, i32 addrspace(1)* %out.gep.3
ret void
}
; GCN-LABEL: {{^}}merge_global_store_4_adjacent_loads_inverse_i32:
; GCN: buffer_load_dwordx4 [[LOAD:v\[[0-9]+:[0-9]+\]]]
; GCN: s_barrier
; GCN: buffer_store_dwordx4 [[LOAD]]
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 @merge_global_store_4_adjacent_loads_inverse_i32(i32 addrspace(1)* %out, i32 addrspace(1)* %in) #0 {
Add target hook to allow merging stores of nonzero constants On GPU targets, materializing constants is cheap and stores are expensive, so only doing this for zero vectors was silly. Most of the new testcases aren't optimally merged, and are for later improvements. llvm-svn: 238108
2015-05-24 08:51:27 +08:00
%out.gep.1 = getelementptr i32, i32 addrspace(1)* %out, i32 1
%out.gep.2 = getelementptr i32, i32 addrspace(1)* %out, i32 2
%out.gep.3 = getelementptr i32, i32 addrspace(1)* %out, i32 3
%in.gep.1 = getelementptr i32, i32 addrspace(1)* %in, i32 1
%in.gep.2 = getelementptr i32, i32 addrspace(1)* %in, i32 2
%in.gep.3 = getelementptr i32, i32 addrspace(1)* %in, i32 3
%x = load i32, i32 addrspace(1)* %in
%y = load i32, i32 addrspace(1)* %in.gep.1
%z = load i32, i32 addrspace(1)* %in.gep.2
%w = load i32, i32 addrspace(1)* %in.gep.3
; Make sure the barrier doesn't stop this
AMDGPU: Remove some old intrinsic uses from tests llvm-svn: 260493
2016-02-11 14:02:01 +08:00
tail call void @llvm.amdgcn.s.barrier() #1
Add target hook to allow merging stores of nonzero constants On GPU targets, materializing constants is cheap and stores are expensive, so only doing this for zero vectors was silly. Most of the new testcases aren't optimally merged, and are for later improvements. llvm-svn: 238108
2015-05-24 08:51:27 +08:00
store i32 %w, i32 addrspace(1)* %out.gep.3
store i32 %z, i32 addrspace(1)* %out.gep.2
store i32 %y, i32 addrspace(1)* %out.gep.1
store i32 %x, i32 addrspace(1)* %out
ret void
}
; TODO: Re-packing of loaded register required. Maybe an IR pass
; should catch this?
; GCN-LABEL: {{^}}merge_global_store_4_adjacent_loads_shuffle_i32:
AMDGPU: Merge BUFFER_LOAD_DWORD_OFFSET into x2, x4 Summary: Only 3 (out of 48486) shaders are affected. Reviewers: arsenm, nhaehnle Subscribers: kzhuravl, wdng, yaxunl, dstuttard, tpr, t-tye, llvm-commits Differential Revision: https://reviews.llvm.org/D38951 llvm-svn: 317753
2017-11-09 09:52:36 +08:00
; GCN: buffer_load_dwordx4 v
Add target hook to allow merging stores of nonzero constants On GPU targets, materializing constants is cheap and stores are expensive, so only doing this for zero vectors was silly. Most of the new testcases aren't optimally merged, and are for later improvements. llvm-svn: 238108
2015-05-24 08:51:27 +08:00
; GCN: s_barrier
AMDGPU: Merge BUFFER_STORE_DWORD_OFFEN/OFFSET into x2, x4 Summary: Only 56 shaders (out of 48486) are affected. Totals from affected shaders (changed stats only): SGPRS: 2420 -> 2460 (1.65 %) Spilled VGPRs: 94 -> 112 (19.15 %) Scratch size: 524 -> 528 (0.76 %) dwords per thread Code Size: 187400 -> 184992 (-1.28 %) bytes One DiRT Showdown shader spills 6 more VGPRs. One Grid Autosport shader spills 12 more VGPRs. The other 54 shaders only have a decrease in code size. (I'm ignoring the SGPR noise) Reviewers: arsenm, nhaehnle Subscribers: kzhuravl, wdng, yaxunl, dstuttard, tpr, llvm-commits, t-tye Differential Revision: https://reviews.llvm.org/D39012 llvm-svn: 317755
2017-11-09 09:52:55 +08:00
; GCN: buffer_store_dwordx4 v
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 @merge_global_store_4_adjacent_loads_shuffle_i32(i32 addrspace(1)* %out, i32 addrspace(1)* %in) #0 {
Add target hook to allow merging stores of nonzero constants On GPU targets, materializing constants is cheap and stores are expensive, so only doing this for zero vectors was silly. Most of the new testcases aren't optimally merged, and are for later improvements. llvm-svn: 238108
2015-05-24 08:51:27 +08:00
%out.gep.1 = getelementptr i32, i32 addrspace(1)* %out, i32 1
%out.gep.2 = getelementptr i32, i32 addrspace(1)* %out, i32 2
%out.gep.3 = getelementptr i32, i32 addrspace(1)* %out, i32 3
%in.gep.1 = getelementptr i32, i32 addrspace(1)* %in, i32 1
%in.gep.2 = getelementptr i32, i32 addrspace(1)* %in, i32 2
%in.gep.3 = getelementptr i32, i32 addrspace(1)* %in, i32 3
%x = load i32, i32 addrspace(1)* %in
%y = load i32, i32 addrspace(1)* %in.gep.1
%z = load i32, i32 addrspace(1)* %in.gep.2
%w = load i32, i32 addrspace(1)* %in.gep.3
; Make sure the barrier doesn't stop this
AMDGPU: Remove some old intrinsic uses from tests llvm-svn: 260493
2016-02-11 14:02:01 +08:00
tail call void @llvm.amdgcn.s.barrier() #1
Add target hook to allow merging stores of nonzero constants On GPU targets, materializing constants is cheap and stores are expensive, so only doing this for zero vectors was silly. Most of the new testcases aren't optimally merged, and are for later improvements. llvm-svn: 238108
2015-05-24 08:51:27 +08:00
store i32 %w, i32 addrspace(1)* %out
store i32 %z, i32 addrspace(1)* %out.gep.1
store i32 %y, i32 addrspace(1)* %out.gep.2
store i32 %x, i32 addrspace(1)* %out.gep.3
ret void
}
; GCN-LABEL: {{^}}merge_global_store_4_adjacent_loads_i8:
; GCN: buffer_load_dword [[LOAD:v[0-9]+]]
; GCN: buffer_store_dword [[LOAD]]
; GCN: s_endpgm
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 @merge_global_store_4_adjacent_loads_i8(i8 addrspace(1)* %out, i8 addrspace(1)* %in) #0 {
Add target hook to allow merging stores of nonzero constants On GPU targets, materializing constants is cheap and stores are expensive, so only doing this for zero vectors was silly. Most of the new testcases aren't optimally merged, and are for later improvements. llvm-svn: 238108
2015-05-24 08:51:27 +08:00
%out.gep.1 = getelementptr i8, i8 addrspace(1)* %out, i8 1
%out.gep.2 = getelementptr i8, i8 addrspace(1)* %out, i8 2
%out.gep.3 = getelementptr i8, i8 addrspace(1)* %out, i8 3
%in.gep.1 = getelementptr i8, i8 addrspace(1)* %in, i8 1
%in.gep.2 = getelementptr i8, i8 addrspace(1)* %in, i8 2
%in.gep.3 = getelementptr i8, i8 addrspace(1)* %in, i8 3
%x = load i8, i8 addrspace(1)* %in, align 4
%y = load i8, i8 addrspace(1)* %in.gep.1
%z = load i8, i8 addrspace(1)* %in.gep.2
%w = load i8, i8 addrspace(1)* %in.gep.3
store i8 %x, i8 addrspace(1)* %out, align 4
store i8 %y, i8 addrspace(1)* %out.gep.1
store i8 %z, i8 addrspace(1)* %out.gep.2
store i8 %w, i8 addrspace(1)* %out.gep.3
ret void
}
; GCN-LABEL: {{^}}merge_global_store_4_adjacent_loads_i8_natural_align:
; GCN: buffer_load_ubyte
; GCN: buffer_load_ubyte
; GCN: buffer_load_ubyte
; GCN: buffer_load_ubyte
; GCN: buffer_store_byte
; GCN: buffer_store_byte
; GCN: buffer_store_byte
; GCN: buffer_store_byte
; GCN: s_endpgm
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 @merge_global_store_4_adjacent_loads_i8_natural_align(i8 addrspace(1)* %out, i8 addrspace(1)* %in) #0 {
Add target hook to allow merging stores of nonzero constants On GPU targets, materializing constants is cheap and stores are expensive, so only doing this for zero vectors was silly. Most of the new testcases aren't optimally merged, and are for later improvements. llvm-svn: 238108
2015-05-24 08:51:27 +08:00
%out.gep.1 = getelementptr i8, i8 addrspace(1)* %out, i8 1
%out.gep.2 = getelementptr i8, i8 addrspace(1)* %out, i8 2
%out.gep.3 = getelementptr i8, i8 addrspace(1)* %out, i8 3
%in.gep.1 = getelementptr i8, i8 addrspace(1)* %in, i8 1
%in.gep.2 = getelementptr i8, i8 addrspace(1)* %in, i8 2
%in.gep.3 = getelementptr i8, i8 addrspace(1)* %in, i8 3
%x = load i8, i8 addrspace(1)* %in
%y = load i8, i8 addrspace(1)* %in.gep.1
%z = load i8, i8 addrspace(1)* %in.gep.2
%w = load i8, i8 addrspace(1)* %in.gep.3
store i8 %x, i8 addrspace(1)* %out
store i8 %y, i8 addrspace(1)* %out.gep.1
store i8 %z, i8 addrspace(1)* %out.gep.2
store i8 %w, i8 addrspace(1)* %out.gep.3
ret void
}
; GCN-LABEL: {{^}}merge_global_store_4_vector_elts_loads_v4i32:
; GCN: buffer_load_dwordx4 [[LOAD:v\[[0-9]+:[0-9]+\]]]
In visitSTORE, always use FindBetterChain, rather than only when UseAA is enabled. Recommiting with compiler time improvements Recommitting after fixup of 32-bit aliasing sign offset bug in DAGCombiner. * Simplify Consecutive Merge Store Candidate Search Now that address aliasing is much less conservative, push through simplified store merging search and chain alias analysis which only checks for parallel stores through the chain subgraph. This is cleaner as the separation of non-interfering loads/stores from the store-merging logic. When merging stores search up the chain through a single load, and finds all possible stores by looking down from through a load and a TokenFactor to all stores visited. This improves the quality of the output SelectionDAG and the output Codegen (save perhaps for some ARM cases where we correctly constructs wider loads, but then promotes them to float operations which appear but requires more expensive constant generation). Some minor peephole optimizations to deal with improved SubDAG shapes (listed below) Additional Minor Changes: 1. Finishes removing unused AliasLoad code 2. Unifies the chain aggregation in the merged stores across code paths 3. Re-add the Store node to the worklist after calling SimplifyDemandedBits. 4. Increase GatherAllAliasesMaxDepth from 6 to 18. That number is arbitrary, but seems sufficient to not cause regressions in tests. 5. Remove Chain dependencies of Memory operations on CopyfromReg nodes as these are captured by data dependence 6. Forward loads-store values through tokenfactors containing {CopyToReg,CopyFromReg} Values. 7. Peephole to convert buildvector of extract_vector_elt to extract_subvector if possible (see CodeGen/AArch64/store-merge.ll) 8. Store merging for the ARM target is restricted to 32-bit as some in some contexts invalid 64-bit operations are being generated. This can be removed once appropriate checks are added. This finishes the change Matt Arsenault started in r246307 and jyknight's original patch. Many tests required some changes as memory operations are now reorderable, improving load-store forwarding. One test in particular is worth noting: CodeGen/PowerPC/ppc64-align-long-double.ll - Improved load-store forwarding converts a load-store pair into a parallel store and a memory-realized bitcast of the same value. However, because we lose the sharing of the explicit and implicit store values we must create another local store. A similar transformation happens before SelectionDAG as well. Reviewers: arsenm, hfinkel, tstellarAMD, jyknight, nhaehnle llvm-svn: 297695
2017-03-14 08:34:14 +08:00
; GCN: buffer_store_dwordx4 [[LOAD]]
DAGCombiner: Don't stop finding better chain on 2 aliases The comment says this was stopped because it was unlikely to be profitable. This is not true if you want to combine vector loads with multiple components. For a simple case that looks like t0 = load t0 ... t1 = load t0 ... t2 = load t0 ... t3 = load t0 ... t4 = store t0:1, t0:1 t5 = store t4, t1:0 t6 = store t5, t2:0 t7 = store t6, t3:0 We want to get all of these stores onto a chain that is a TokenFactor of these N loads. This mostly solves the AMDGPU merge-stores.ll regressions with -combiner-alias-analysis for merging vector stores of vector loads. llvm-svn: 250138
2015-10-13 08:49:00 +08:00
; GCN: s_endpgm
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 @merge_global_store_4_vector_elts_loads_v4i32(i32 addrspace(1)* %out, <4 x i32> addrspace(1)* %in) #0 {
Add target hook to allow merging stores of nonzero constants On GPU targets, materializing constants is cheap and stores are expensive, so only doing this for zero vectors was silly. Most of the new testcases aren't optimally merged, and are for later improvements. llvm-svn: 238108
2015-05-24 08:51:27 +08:00
%out.gep.1 = getelementptr i32, i32 addrspace(1)* %out, i32 1
%out.gep.2 = getelementptr i32, i32 addrspace(1)* %out, i32 2
%out.gep.3 = getelementptr i32, i32 addrspace(1)* %out, i32 3
%vec = load <4 x i32>, <4 x i32> addrspace(1)* %in
%x = extractelement <4 x i32> %vec, i32 0
%y = extractelement <4 x i32> %vec, i32 1
%z = extractelement <4 x i32> %vec, i32 2
%w = extractelement <4 x i32> %vec, i32 3
store i32 %x, i32 addrspace(1)* %out
store i32 %y, i32 addrspace(1)* %out.gep.1
store i32 %z, i32 addrspace(1)* %out.gep.2
store i32 %w, i32 addrspace(1)* %out.gep.3
ret void
}
; GCN-LABEL: {{^}}merge_local_store_2_constants_i8:
[DAG] Fix typo preventing some stores merges to truncated stores. Check the actual memory type stored and not the extended value size when considering if truncated store merge is worthwhile. Reviewers: efriedma, RKSimon, spatel, jyknight Reviewed By: efriedma Subscribers: llvm-commits, nhaehnle Differential Revision: https://reviews.llvm.org/D35623 llvm-svn: 308833
2017-07-23 10:06:28 +08:00
; GCN: ds_write_b16
Add target hook to allow merging stores of nonzero constants On GPU targets, materializing constants is cheap and stores are expensive, so only doing this for zero vectors was silly. Most of the new testcases aren't optimally merged, and are for later improvements. llvm-svn: 238108
2015-05-24 08:51:27 +08:00
; GCN: s_endpgm
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 @merge_local_store_2_constants_i8(i8 addrspace(3)* %out) #0 {
Add target hook to allow merging stores of nonzero constants On GPU targets, materializing constants is cheap and stores are expensive, so only doing this for zero vectors was silly. Most of the new testcases aren't optimally merged, and are for later improvements. llvm-svn: 238108
2015-05-24 08:51:27 +08:00
%out.gep.1 = getelementptr i8, i8 addrspace(3)* %out, i32 1
store i8 123, i8 addrspace(3)* %out.gep.1
store i8 456, i8 addrspace(3)* %out, align 2
ret void
}
; GCN-LABEL: {{^}}merge_local_store_2_constants_i32:
AMDGPU/SI: Fold operands through REG_SEQUENCE instructions Summary: This helps mostly when we use add instructions for address calculations that contain immediates. Reviewers: arsenm Subscribers: arsenm, llvm-commits Differential Revision: http://reviews.llvm.org/D12256 llvm-svn: 247157
2015-09-09 23:43:26 +08:00
; GCN-DAG: v_mov_b32_e32 v[[LO:[0-9]+]], 0x1c8
; GCN-DAG: v_mov_b32_e32 v[[HI:[0-9]+]], 0x7b
Add target hook to allow merging stores of nonzero constants On GPU targets, materializing constants is cheap and stores are expensive, so only doing this for zero vectors was silly. Most of the new testcases aren't optimally merged, and are for later improvements. llvm-svn: 238108
2015-05-24 08:51:27 +08:00
; GCN: ds_write2_b32 v{{[0-9]+}}, v[[LO]], v[[HI]] offset1:1{{$}}
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 @merge_local_store_2_constants_i32(i32 addrspace(3)* %out) #0 {
Add target hook to allow merging stores of nonzero constants On GPU targets, materializing constants is cheap and stores are expensive, so only doing this for zero vectors was silly. Most of the new testcases aren't optimally merged, and are for later improvements. llvm-svn: 238108
2015-05-24 08:51:27 +08:00
%out.gep.1 = getelementptr i32, i32 addrspace(3)* %out, i32 1
store i32 123, i32 addrspace(3)* %out.gep.1
store i32 456, i32 addrspace(3)* %out
ret void
}
; GCN-LABEL: {{^}}merge_local_store_4_constants_i32:
AMDGPU: Split LDS vector loads If properly aligned this could allow using ds_read_b64. llvm-svn: 253975
2015-11-24 20:18:54 +08:00
; GCN-DAG: v_mov_b32_e32 [[K2:v[0-9]+]], 0x1c8
; GCN-DAG: v_mov_b32_e32 [[K3:v[0-9]+]], 0x14d
; GCN-DAG: ds_write2_b32 v{{[0-9]+}}, [[K2]], [[K3]] offset0:2 offset1:3
; GCN-DAG: v_mov_b32_e32 [[K0:v[0-9]+]], 0x4d2
; GCN-DAG: v_mov_b32_e32 [[K1:v[0-9]+]], 0x7b
; GCN-DAG: ds_write2_b32 v{{[0-9]+}}, [[K0]], [[K1]] offset1:1
; GCN: s_endpgm
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 @merge_local_store_4_constants_i32(i32 addrspace(3)* %out) #0 {
Add target hook to allow merging stores of nonzero constants On GPU targets, materializing constants is cheap and stores are expensive, so only doing this for zero vectors was silly. Most of the new testcases aren't optimally merged, and are for later improvements. llvm-svn: 238108
2015-05-24 08:51:27 +08:00
%out.gep.1 = getelementptr i32, i32 addrspace(3)* %out, i32 1
%out.gep.2 = getelementptr i32, i32 addrspace(3)* %out, i32 2
%out.gep.3 = getelementptr i32, i32 addrspace(3)* %out, i32 3
store i32 123, i32 addrspace(3)* %out.gep.1
store i32 456, i32 addrspace(3)* %out.gep.2
store i32 333, i32 addrspace(3)* %out.gep.3
store i32 1234, i32 addrspace(3)* %out
ret void
}
Revert "Revert "Fix merges of non-zero vector stores"" Reapply r239539. Don't assume the collected number of stores is the same vector size. Just take the first N stores to fill the vector. llvm-svn: 239825
2015-06-16 23:51:48 +08:00
; GCN-LABEL: {{^}}merge_global_store_5_constants_i32:
; GCN-DAG: v_mov_b32_e32 v[[LO:[0-9]+]], 9{{$}}
; GCN-DAG: v_mov_b32_e32 v[[HI4:[0-9]+]], -12{{$}}
; GCN: buffer_store_dwordx4 v{{\[}}[[LO]]:[[HI4]]{{\]}}
; GCN: v_mov_b32_e32 v[[HI:[0-9]+]], 11{{$}}
; GCN: buffer_store_dword v[[HI]]
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 @merge_global_store_5_constants_i32(i32 addrspace(1)* %out) {
Revert "Revert "Fix merges of non-zero vector stores"" Reapply r239539. Don't assume the collected number of stores is the same vector size. Just take the first N stores to fill the vector. llvm-svn: 239825
2015-06-16 23:51:48 +08:00
store i32 9, i32 addrspace(1)* %out, align 4
%idx1 = getelementptr inbounds i32, i32 addrspace(1)* %out, i64 1
store i32 12, i32 addrspace(1)* %idx1, align 4
%idx2 = getelementptr inbounds i32, i32 addrspace(1)* %out, i64 2
store i32 16, i32 addrspace(1)* %idx2, align 4
%idx3 = getelementptr inbounds i32, i32 addrspace(1)* %out, i64 3
store i32 -12, i32 addrspace(1)* %idx3, align 4
%idx4 = getelementptr inbounds i32, i32 addrspace(1)* %out, i64 4
store i32 11, i32 addrspace(1)* %idx4, align 4
ret void
}
; GCN-LABEL: {{^}}merge_global_store_6_constants_i32:
; GCN: buffer_store_dwordx4
; GCN: buffer_store_dwordx2
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 @merge_global_store_6_constants_i32(i32 addrspace(1)* %out) {
Revert "Revert "Fix merges of non-zero vector stores"" Reapply r239539. Don't assume the collected number of stores is the same vector size. Just take the first N stores to fill the vector. llvm-svn: 239825
2015-06-16 23:51:48 +08:00
store i32 13, i32 addrspace(1)* %out, align 4
%idx1 = getelementptr inbounds i32, i32 addrspace(1)* %out, i64 1
store i32 15, i32 addrspace(1)* %idx1, align 4
%idx2 = getelementptr inbounds i32, i32 addrspace(1)* %out, i64 2
store i32 62, i32 addrspace(1)* %idx2, align 4
%idx3 = getelementptr inbounds i32, i32 addrspace(1)* %out, i64 3
store i32 63, i32 addrspace(1)* %idx3, align 4
%idx4 = getelementptr inbounds i32, i32 addrspace(1)* %out, i64 4
store i32 11, i32 addrspace(1)* %idx4, align 4
%idx5 = getelementptr inbounds i32, i32 addrspace(1)* %out, i64 5
store i32 123, i32 addrspace(1)* %idx5, align 4
ret void
}
; GCN-LABEL: {{^}}merge_global_store_7_constants_i32:
; GCN: buffer_store_dwordx4
; GCN: buffer_store_dwordx2
; GCN: buffer_store_dword v
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 @merge_global_store_7_constants_i32(i32 addrspace(1)* %out) {
Revert "Revert "Fix merges of non-zero vector stores"" Reapply r239539. Don't assume the collected number of stores is the same vector size. Just take the first N stores to fill the vector. llvm-svn: 239825
2015-06-16 23:51:48 +08:00
store i32 34, i32 addrspace(1)* %out, align 4
%idx1 = getelementptr inbounds i32, i32 addrspace(1)* %out, i64 1
store i32 999, i32 addrspace(1)* %idx1, align 4
%idx2 = getelementptr inbounds i32, i32 addrspace(1)* %out, i64 2
store i32 65, i32 addrspace(1)* %idx2, align 4
%idx3 = getelementptr inbounds i32, i32 addrspace(1)* %out, i64 3
store i32 33, i32 addrspace(1)* %idx3, align 4
%idx4 = getelementptr inbounds i32, i32 addrspace(1)* %out, i64 4
store i32 98, i32 addrspace(1)* %idx4, align 4
%idx5 = getelementptr inbounds i32, i32 addrspace(1)* %out, i64 5
store i32 91, i32 addrspace(1)* %idx5, align 4
%idx6 = getelementptr inbounds i32, i32 addrspace(1)* %out, i64 6
store i32 212, i32 addrspace(1)* %idx6, align 4
ret void
}
; GCN-LABEL: {{^}}merge_global_store_8_constants_i32:
AMDGPU: Split x8 and x16 vector loads instead of scalarize The one regression in the builtin tests is in the read2 test which now (again) has many extra copies, but this should be solved once the pass is replaced with a DAG combine. llvm-svn: 253974
2015-11-24 20:05:03 +08:00
; GCN: buffer_store_dwordx4
; GCN: buffer_store_dwordx4
DAGCombiner: Don't stop finding better chain on 2 aliases The comment says this was stopped because it was unlikely to be profitable. This is not true if you want to combine vector loads with multiple components. For a simple case that looks like t0 = load t0 ... t1 = load t0 ... t2 = load t0 ... t3 = load t0 ... t4 = store t0:1, t0:1 t5 = store t4, t1:0 t6 = store t5, t2:0 t7 = store t6, t3:0 We want to get all of these stores onto a chain that is a TokenFactor of these N loads. This mostly solves the AMDGPU merge-stores.ll regressions with -combiner-alias-analysis for merging vector stores of vector loads. llvm-svn: 250138
2015-10-13 08:49:00 +08:00
; GCN: s_endpgm
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 @merge_global_store_8_constants_i32(i32 addrspace(1)* %out) {
Revert "Revert "Fix merges of non-zero vector stores"" Reapply r239539. Don't assume the collected number of stores is the same vector size. Just take the first N stores to fill the vector. llvm-svn: 239825
2015-06-16 23:51:48 +08:00
store i32 34, i32 addrspace(1)* %out, align 4
%idx1 = getelementptr inbounds i32, i32 addrspace(1)* %out, i64 1
store i32 999, i32 addrspace(1)* %idx1, align 4
%idx2 = getelementptr inbounds i32, i32 addrspace(1)* %out, i64 2
store i32 65, i32 addrspace(1)* %idx2, align 4
%idx3 = getelementptr inbounds i32, i32 addrspace(1)* %out, i64 3
store i32 33, i32 addrspace(1)* %idx3, align 4
%idx4 = getelementptr inbounds i32, i32 addrspace(1)* %out, i64 4
store i32 98, i32 addrspace(1)* %idx4, align 4
%idx5 = getelementptr inbounds i32, i32 addrspace(1)* %out, i64 5
store i32 91, i32 addrspace(1)* %idx5, align 4
%idx6 = getelementptr inbounds i32, i32 addrspace(1)* %out, i64 6
store i32 212, i32 addrspace(1)* %idx6, align 4
%idx7 = getelementptr inbounds i32, i32 addrspace(1)* %out, i64 7
store i32 999, i32 addrspace(1)* %idx7, align 4
ret void
}
AMDGPU: Add some tests for promotion of v2i64 scalar_to_vector llvm-svn: 254087
2015-11-26 04:01:03 +08:00
; This requires handling of scalar_to_vector for v2i64 to avoid
; scratch usage.
; FIXME: Should do single load and store
; GCN-LABEL: {{^}}copy_v3i32_align4:
; GCN-NOT: SCRATCH_RSRC_DWORD
AMDGPU/SI: Assembler: Unify parsing/printing of operands. Summary: The goal is for each operand type to have its own parse function and at the same time share common code for tracking state as different instruction types share operand types (e.g. glc/glc_flat, etc). Introduce parseAMDGPUOperand which can parse any optional operand. DPP and Clamp/OMod have custom handling for now. Sam also suggested to have class hierarchy for operand types instead of table. This can be done in separate change. Remove parseVOP3OptionalOps, parseDS*OptionalOps, parseFlatOptionalOps, parseMubufOptionalOps, parseDPPOptionalOps. Reduce number of definitions of AsmOperand's and MatchClasses' by using common base class. Rename AsmMatcher/InstPrinter methods accordingly. Print immediate type when printing parsed immediate operand. Use 'off' if offset/index register is unused instead of skipping it to make it more readable (also agreed with SP3). Update tests. Reviewers: tstellarAMD, SamWot, artem.tamazov Subscribers: qcolombet, arsenm, llvm-commits Differential Revision: http://reviews.llvm.org/D19584 llvm-svn: 268015
2016-04-29 17:02:30 +08:00
; GCN-DAG: buffer_load_dword v{{[0-9]+}}, off, s{{\[[0-9]+:[0-9]+\]}}, 0 offset:8
; GCN-DAG: buffer_load_dwordx2 v{{\[[0-9]+:[0-9]+\]}}, off, s{{\[[0-9]+:[0-9]+\]}}, 0{{$}}
AMDGPU: Add some tests for promotion of v2i64 scalar_to_vector llvm-svn: 254087
2015-11-26 04:01:03 +08:00
; GCN-NOT: offen
; GCN: s_waitcnt vmcnt
; GCN-NOT: offen
AMDGPU/SI: Assembler: Unify parsing/printing of operands. Summary: The goal is for each operand type to have its own parse function and at the same time share common code for tracking state as different instruction types share operand types (e.g. glc/glc_flat, etc). Introduce parseAMDGPUOperand which can parse any optional operand. DPP and Clamp/OMod have custom handling for now. Sam also suggested to have class hierarchy for operand types instead of table. This can be done in separate change. Remove parseVOP3OptionalOps, parseDS*OptionalOps, parseFlatOptionalOps, parseMubufOptionalOps, parseDPPOptionalOps. Reduce number of definitions of AsmOperand's and MatchClasses' by using common base class. Rename AsmMatcher/InstPrinter methods accordingly. Print immediate type when printing parsed immediate operand. Use 'off' if offset/index register is unused instead of skipping it to make it more readable (also agreed with SP3). Update tests. Reviewers: tstellarAMD, SamWot, artem.tamazov Subscribers: qcolombet, arsenm, llvm-commits Differential Revision: http://reviews.llvm.org/D19584 llvm-svn: 268015
2016-04-29 17:02:30 +08:00
; GCN-DAG: buffer_store_dwordx2 v{{\[[0-9]+:[0-9]+\]}}, off, s{{\[[0-9]+:[0-9]+\]}}, 0{{$}}
; GCN-DAG: buffer_store_dword v{{[0-9]+}}, off, s{{\[[0-9]+:[0-9]+\]}}, 0 offset:8
AMDGPU: Add some tests for promotion of v2i64 scalar_to_vector llvm-svn: 254087
2015-11-26 04:01:03 +08:00
; GCN: ScratchSize: 0{{$}}
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 @copy_v3i32_align4(<3 x i32> addrspace(1)* noalias %out, <3 x i32> addrspace(1)* noalias %in) #0 {
AMDGPU: Add some tests for promotion of v2i64 scalar_to_vector llvm-svn: 254087
2015-11-26 04:01:03 +08:00
%vec = load <3 x i32>, <3 x i32> addrspace(1)* %in, align 4
store <3 x i32> %vec, <3 x i32> addrspace(1)* %out
ret void
}
; GCN-LABEL: {{^}}copy_v3i64_align4:
; GCN-NOT: SCRATCH_RSRC_DWORD
AMDGPU/SI: Assembler: Unify parsing/printing of operands. Summary: The goal is for each operand type to have its own parse function and at the same time share common code for tracking state as different instruction types share operand types (e.g. glc/glc_flat, etc). Introduce parseAMDGPUOperand which can parse any optional operand. DPP and Clamp/OMod have custom handling for now. Sam also suggested to have class hierarchy for operand types instead of table. This can be done in separate change. Remove parseVOP3OptionalOps, parseDS*OptionalOps, parseFlatOptionalOps, parseMubufOptionalOps, parseDPPOptionalOps. Reduce number of definitions of AsmOperand's and MatchClasses' by using common base class. Rename AsmMatcher/InstPrinter methods accordingly. Print immediate type when printing parsed immediate operand. Use 'off' if offset/index register is unused instead of skipping it to make it more readable (also agreed with SP3). Update tests. Reviewers: tstellarAMD, SamWot, artem.tamazov Subscribers: qcolombet, arsenm, llvm-commits Differential Revision: http://reviews.llvm.org/D19584 llvm-svn: 268015
2016-04-29 17:02:30 +08:00
; GCN-DAG: buffer_load_dwordx4 v{{\[[0-9]+:[0-9]+\]}}, off, s{{\[[0-9]+:[0-9]+\]}}, 0{{$}}
; GCN-DAG: buffer_load_dwordx2 v{{\[[0-9]+:[0-9]+\]}}, off, s{{\[[0-9]+:[0-9]+\]}}, 0 offset:16{{$}}
AMDGPU: Add some tests for promotion of v2i64 scalar_to_vector llvm-svn: 254087
2015-11-26 04:01:03 +08:00
; GCN-NOT: offen
; GCN: s_waitcnt vmcnt
; GCN-NOT: offen
AMDGPU/SI: Assembler: Unify parsing/printing of operands. Summary: The goal is for each operand type to have its own parse function and at the same time share common code for tracking state as different instruction types share operand types (e.g. glc/glc_flat, etc). Introduce parseAMDGPUOperand which can parse any optional operand. DPP and Clamp/OMod have custom handling for now. Sam also suggested to have class hierarchy for operand types instead of table. This can be done in separate change. Remove parseVOP3OptionalOps, parseDS*OptionalOps, parseFlatOptionalOps, parseMubufOptionalOps, parseDPPOptionalOps. Reduce number of definitions of AsmOperand's and MatchClasses' by using common base class. Rename AsmMatcher/InstPrinter methods accordingly. Print immediate type when printing parsed immediate operand. Use 'off' if offset/index register is unused instead of skipping it to make it more readable (also agreed with SP3). Update tests. Reviewers: tstellarAMD, SamWot, artem.tamazov Subscribers: qcolombet, arsenm, llvm-commits Differential Revision: http://reviews.llvm.org/D19584 llvm-svn: 268015
2016-04-29 17:02:30 +08:00
; GCN-DAG: buffer_store_dwordx4 v{{\[[0-9]+:[0-9]+\]}}, off, s{{\[[0-9]+:[0-9]+\]}}, 0{{$}}
; GCN-DAG: buffer_store_dwordx2 v{{\[[0-9]+:[0-9]+\]}}, off, s{{\[[0-9]+:[0-9]+\]}}, 0 offset:16{{$}}
AMDGPU: Add some tests for promotion of v2i64 scalar_to_vector llvm-svn: 254087
2015-11-26 04:01:03 +08:00
; GCN: ScratchSize: 0{{$}}
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 @copy_v3i64_align4(<3 x i64> addrspace(1)* noalias %out, <3 x i64> addrspace(1)* noalias %in) #0 {
AMDGPU: Add some tests for promotion of v2i64 scalar_to_vector llvm-svn: 254087
2015-11-26 04:01:03 +08:00
%vec = load <3 x i64>, <3 x i64> addrspace(1)* %in, align 4
store <3 x i64> %vec, <3 x i64> addrspace(1)* %out
ret void
}
; GCN-LABEL: {{^}}copy_v3f32_align4:
; GCN-NOT: SCRATCH_RSRC_DWORD
AMDGPU/SI: Assembler: Unify parsing/printing of operands. Summary: The goal is for each operand type to have its own parse function and at the same time share common code for tracking state as different instruction types share operand types (e.g. glc/glc_flat, etc). Introduce parseAMDGPUOperand which can parse any optional operand. DPP and Clamp/OMod have custom handling for now. Sam also suggested to have class hierarchy for operand types instead of table. This can be done in separate change. Remove parseVOP3OptionalOps, parseDS*OptionalOps, parseFlatOptionalOps, parseMubufOptionalOps, parseDPPOptionalOps. Reduce number of definitions of AsmOperand's and MatchClasses' by using common base class. Rename AsmMatcher/InstPrinter methods accordingly. Print immediate type when printing parsed immediate operand. Use 'off' if offset/index register is unused instead of skipping it to make it more readable (also agreed with SP3). Update tests. Reviewers: tstellarAMD, SamWot, artem.tamazov Subscribers: qcolombet, arsenm, llvm-commits Differential Revision: http://reviews.llvm.org/D19584 llvm-svn: 268015
2016-04-29 17:02:30 +08:00
; GCN-DAG: buffer_load_dword v{{[0-9]+}}, off, s{{\[[0-9]+:[0-9]+\]}}, 0 offset:8
; GCN-DAG: buffer_load_dwordx2 v{{\[[0-9]+:[0-9]+\]}}, off, s{{\[[0-9]+:[0-9]+\]}}, 0{{$}}
AMDGPU: Add some tests for promotion of v2i64 scalar_to_vector llvm-svn: 254087
2015-11-26 04:01:03 +08:00
; GCN-NOT: offen
; GCN: s_waitcnt vmcnt
; GCN-NOT: offen
AMDGPU/SI: Assembler: Unify parsing/printing of operands. Summary: The goal is for each operand type to have its own parse function and at the same time share common code for tracking state as different instruction types share operand types (e.g. glc/glc_flat, etc). Introduce parseAMDGPUOperand which can parse any optional operand. DPP and Clamp/OMod have custom handling for now. Sam also suggested to have class hierarchy for operand types instead of table. This can be done in separate change. Remove parseVOP3OptionalOps, parseDS*OptionalOps, parseFlatOptionalOps, parseMubufOptionalOps, parseDPPOptionalOps. Reduce number of definitions of AsmOperand's and MatchClasses' by using common base class. Rename AsmMatcher/InstPrinter methods accordingly. Print immediate type when printing parsed immediate operand. Use 'off' if offset/index register is unused instead of skipping it to make it more readable (also agreed with SP3). Update tests. Reviewers: tstellarAMD, SamWot, artem.tamazov Subscribers: qcolombet, arsenm, llvm-commits Differential Revision: http://reviews.llvm.org/D19584 llvm-svn: 268015
2016-04-29 17:02:30 +08:00
; GCN-DAG: buffer_store_dwordx2 v{{\[[0-9]+:[0-9]+\]}}, off, s{{\[[0-9]+:[0-9]+\]}}, 0{{$}}
; GCN-DAG: buffer_store_dword v{{[0-9]+}}, off, s{{\[[0-9]+:[0-9]+\]}}, 0 offset:8
AMDGPU: Add some tests for promotion of v2i64 scalar_to_vector llvm-svn: 254087
2015-11-26 04:01:03 +08:00
; GCN: ScratchSize: 0{{$}}
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 @copy_v3f32_align4(<3 x float> addrspace(1)* noalias %out, <3 x float> addrspace(1)* noalias %in) #0 {
AMDGPU: Add some tests for promotion of v2i64 scalar_to_vector llvm-svn: 254087
2015-11-26 04:01:03 +08:00
%vec = load <3 x float>, <3 x float> addrspace(1)* %in, align 4
%fadd = fadd <3 x float> %vec, <float 1.0, float 2.0, float 4.0>
store <3 x float> %fadd, <3 x float> addrspace(1)* %out
ret void
}
; GCN-LABEL: {{^}}copy_v3f64_align4:
; GCN-NOT: SCRATCH_RSRC_DWORD
AMDGPU/SI: Assembler: Unify parsing/printing of operands. Summary: The goal is for each operand type to have its own parse function and at the same time share common code for tracking state as different instruction types share operand types (e.g. glc/glc_flat, etc). Introduce parseAMDGPUOperand which can parse any optional operand. DPP and Clamp/OMod have custom handling for now. Sam also suggested to have class hierarchy for operand types instead of table. This can be done in separate change. Remove parseVOP3OptionalOps, parseDS*OptionalOps, parseFlatOptionalOps, parseMubufOptionalOps, parseDPPOptionalOps. Reduce number of definitions of AsmOperand's and MatchClasses' by using common base class. Rename AsmMatcher/InstPrinter methods accordingly. Print immediate type when printing parsed immediate operand. Use 'off' if offset/index register is unused instead of skipping it to make it more readable (also agreed with SP3). Update tests. Reviewers: tstellarAMD, SamWot, artem.tamazov Subscribers: qcolombet, arsenm, llvm-commits Differential Revision: http://reviews.llvm.org/D19584 llvm-svn: 268015
2016-04-29 17:02:30 +08:00
; GCN-DAG: buffer_load_dwordx4 v{{\[[0-9]+:[0-9]+\]}}, off, s{{\[[0-9]+:[0-9]+\]}}, 0{{$}}
; GCN-DAG: buffer_load_dwordx2 v{{\[[0-9]+:[0-9]+\]}}, off, s{{\[[0-9]+:[0-9]+\]}}, 0 offset:16{{$}}
AMDGPU: Add some tests for promotion of v2i64 scalar_to_vector llvm-svn: 254087
2015-11-26 04:01:03 +08:00
; GCN-NOT: offen
; GCN: s_waitcnt vmcnt
; GCN-NOT: offen
AMDGPU/SI: Assembler: Unify parsing/printing of operands. Summary: The goal is for each operand type to have its own parse function and at the same time share common code for tracking state as different instruction types share operand types (e.g. glc/glc_flat, etc). Introduce parseAMDGPUOperand which can parse any optional operand. DPP and Clamp/OMod have custom handling for now. Sam also suggested to have class hierarchy for operand types instead of table. This can be done in separate change. Remove parseVOP3OptionalOps, parseDS*OptionalOps, parseFlatOptionalOps, parseMubufOptionalOps, parseDPPOptionalOps. Reduce number of definitions of AsmOperand's and MatchClasses' by using common base class. Rename AsmMatcher/InstPrinter methods accordingly. Print immediate type when printing parsed immediate operand. Use 'off' if offset/index register is unused instead of skipping it to make it more readable (also agreed with SP3). Update tests. Reviewers: tstellarAMD, SamWot, artem.tamazov Subscribers: qcolombet, arsenm, llvm-commits Differential Revision: http://reviews.llvm.org/D19584 llvm-svn: 268015
2016-04-29 17:02:30 +08:00
; GCN-DAG: buffer_store_dwordx4 v{{\[[0-9]+:[0-9]+\]}}, off, s{{\[[0-9]+:[0-9]+\]}}, 0{{$}}
; GCN-DAG: buffer_store_dwordx2 v{{\[[0-9]+:[0-9]+\]}}, off, s{{\[[0-9]+:[0-9]+\]}}, 0 offset:16{{$}}
AMDGPU: Add some tests for promotion of v2i64 scalar_to_vector llvm-svn: 254087
2015-11-26 04:01:03 +08:00
; GCN: ScratchSize: 0{{$}}
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 @copy_v3f64_align4(<3 x double> addrspace(1)* noalias %out, <3 x double> addrspace(1)* noalias %in) #0 {
AMDGPU: Add some tests for promotion of v2i64 scalar_to_vector llvm-svn: 254087
2015-11-26 04:01:03 +08:00
%vec = load <3 x double>, <3 x double> addrspace(1)* %in, align 4
%fadd = fadd <3 x double> %vec, <double 1.0, double 2.0, double 4.0>
store <3 x double> %fadd, <3 x double> addrspace(1)* %out
ret void
}
AMDGPU: Remove some old intrinsic uses from tests llvm-svn: 260493
2016-02-11 14:02:01 +08:00
declare void @llvm.amdgcn.s.barrier() #1
Add target hook to allow merging stores of nonzero constants On GPU targets, materializing constants is cheap and stores are expensive, so only doing this for zero vectors was silly. Most of the new testcases aren't optimally merged, and are for later improvements. llvm-svn: 238108
2015-05-24 08:51:27 +08:00
attributes #0 = { nounwind }
AMDGPU: Switch barrier intrinsics to using convergent noduplicate prevents unrolling of small loops that happen to have barriers in them. If a loop has a barrier in it, it is OK to duplicate it for the unroll. llvm-svn: 256075
2015-12-19 09:46:41 +08:00
attributes #1 = { convergent nounwind }