-
Notifications
You must be signed in to change notification settings - Fork 734
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[SYCL] Optional kernel features: implement split based on reqd-sub-group-size #8167
Closed
dm-vodopyanov
wants to merge
6
commits into
intel:sycl
from
dm-vodopyanov:split-based-on-reqd-sub-group-size
Closed
Changes from all commits
Commits
Show all changes
6 commits
Select commit
Hold shift + click to select a range
36236a8
[SYCL] Optional kernel features: implement split based on reqd-sub-gr…
dm-vodopyanov 003e9ed
Add missing line
dm-vodopyanov cbcf5a6
Fixed some tests failed in CI and passed locally
dm-vodopyanov deb66f1
Apply CR comment + some fixes
dm-vodopyanov de51fe8
Apply CR comments
dm-vodopyanov 37d298e
Merge branch 'intel:sycl' into split-based-on-reqd-sub-group-size
dm-vodopyanov File filter
Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
133 changes: 133 additions & 0 deletions
133
llvm/test/tools/sycl-post-link/device-code-split/per-reqd-sub-group-size-split-1.ll
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,133 @@ | ||
| ; This test emulates two translation units with 3 kernels: | ||
| ; TU0_kernel0 - 1st translation unit, no reqd_sub_group_size attribute used | ||
| ; TU0_kernel1 - 1st translation unit, reqd_sub_group_size attribute is used | ||
| ; TU1_kernel2 - 2nd translation unit, no reqd_sub_group_size attribute used | ||
|
|
||
| ; The test is intended to check that sycl-post-link correctly separates kernels | ||
| ; that use reqd_sub_group_size attributes from kernels which doesn't use them | ||
| ; regardless of device code split mode | ||
|
|
||
| ; RUN: sycl-post-link -split=auto -symbols -S %s -o %t.table | ||
| ; RUN: FileCheck %s -input-file=%t_0.ll --check-prefixes CHECK-M0-IR \ | ||
| ; RUN: --implicit-check-not kernel0 --implicit-check-not kernel1 | ||
| ; RUN: FileCheck %s -input-file=%t_2.ll --check-prefixes CHECK-M1-IR \ | ||
| ; RUN: --implicit-check-not kernel0 --implicit-check-not kernel1 | ||
| ; RUN: FileCheck %s -input-file=%t_1.ll --check-prefixes CHECK-M2-IR \ | ||
| ; RUN: --implicit-check-not kernel1 --implicit-check-not kernel2 | ||
| ; RUN: FileCheck %s -input-file=%t_0.sym --check-prefixes CHECK-M0-SYMS \ | ||
| ; RUN: --implicit-check-not kernel0 --implicit-check-not kernel1 | ||
| ; RUN: FileCheck %s -input-file=%t_2.sym --check-prefixes CHECK-M1-SYMS \ | ||
| ; RUN: --implicit-check-not kernel0 --implicit-check-not kernel2 | ||
| ; RUN: FileCheck %s -input-file=%t_1.sym --check-prefixes CHECK-M2-SYMS \ | ||
| ; RUN: --implicit-check-not kernel1 --implicit-check-not kernel2 | ||
|
|
||
| ; RUN: sycl-post-link -split=source -symbols -S %s -o %t.table | ||
| ; RUN: FileCheck %s -input-file=%t_0.ll --check-prefixes CHECK-M0-IR \ | ||
| ; RUN: --implicit-check-not kernel0 --implicit-check-not kernel1 | ||
| ; RUN: FileCheck %s -input-file=%t_2.ll --check-prefixes CHECK-M1-IR \ | ||
| ; RUN: --implicit-check-not kernel0 --implicit-check-not kernel2 | ||
| ; RUN: FileCheck %s -input-file=%t_1.ll --check-prefixes CHECK-M2-IR \ | ||
| ; RUN: --implicit-check-not kernel1 --implicit-check-not kernel2 | ||
| ; RUN: FileCheck %s -input-file=%t_0.sym --check-prefixes CHECK-M0-SYMS \ | ||
| ; RUN: --implicit-check-not kernel0 --implicit-check-not kernel1 | ||
| ; RUN: FileCheck %s -input-file=%t_2.sym --check-prefixes CHECK-M1-SYMS \ | ||
| ; RUN: --implicit-check-not kernel0 --implicit-check-not kernel2 | ||
| ; RUN: FileCheck %s -input-file=%t_1.sym --check-prefixes CHECK-M2-SYMS \ | ||
| ; RUN: --implicit-check-not kernel1 --implicit-check-not kernel2 | ||
|
|
||
| ; RUN: sycl-post-link -split=kernel -symbols -S %s -o %t.table | ||
| ; RUN: FileCheck %s -input-file=%t_0.ll --check-prefixes CHECK-M0-IR \ | ||
| ; RUN: --implicit-check-not kernel0 --implicit-check-not kernel1 | ||
| ; RUN: FileCheck %s -input-file=%t_1.ll --check-prefixes CHECK-M1-IR \ | ||
| ; RUN: --implicit-check-not kernel0 --implicit-check-not kernel2 | ||
| ; RUN: FileCheck %s -input-file=%t_2.ll --check-prefixes CHECK-M2-IR \ | ||
| ; RUN: --implicit-check-not kernel1 --implicit-check-not kernel2 | ||
| ; RUN: FileCheck %s -input-file=%t_0.sym --check-prefixes CHECK-M0-SYMS \ | ||
| ; RUN: --implicit-check-not kernel0 --implicit-check-not kernel1 | ||
| ; RUN: FileCheck %s -input-file=%t_1.sym --check-prefixes CHECK-M1-SYMS \ | ||
| ; RUN: --implicit-check-not kernel0 --implicit-check-not kernel2 | ||
| ; RUN: FileCheck %s -input-file=%t_2.sym --check-prefixes CHECK-M2-SYMS \ | ||
| ; RUN: --implicit-check-not kernel1 --implicit-check-not kernel2 | ||
|
|
||
| ; Regardless of device code split mode, each kernel should go into a separate | ||
| ; device image | ||
|
|
||
| ; CHECK-M2-IR: define {{.*}} @TU0_kernel0 | ||
| ; CHECK-M2-SYMS: TU0_kernel0 | ||
|
|
||
| ; CHECK-M1-IR: define {{.*}} @TU0_kernel1 | ||
| ; CHECK-M1-SYMS: TU0_kernel1 | ||
|
|
||
| ; CHECK-M0-IR: define {{.*}} @TU1_kernel2 | ||
| ; CHECK-M0-SYMS: TU1_kernel2 | ||
|
|
||
| target datalayout = "e-i64:64-v16:16-v24:32-v32:32-v48:64-v96:128-v192:256-v256:256-v512:512-v1024:1024" | ||
| target triple = "spir64-unknown-linux" | ||
|
|
||
| ; FIXME: device globals should also be properly distributed across device images | ||
| ; if they are of optional type | ||
| @_ZL2GV = internal addrspace(1) constant [1 x i32] [i32 42], align 4 | ||
|
|
||
| define dso_local spir_kernel void @TU0_kernel0() #0 { | ||
| entry: | ||
| call spir_func void @foo() | ||
| ret void | ||
| } | ||
|
|
||
| define dso_local spir_func void @foo() { | ||
| entry: | ||
| %a = alloca i32, align 4 | ||
| %call = call spir_func i32 @bar(i32 1) | ||
| %add = add nsw i32 2, %call | ||
| store i32 %add, i32* %a, align 4 | ||
| ret void | ||
| } | ||
|
|
||
| ; Function Attrs: nounwind | ||
| define linkonce_odr dso_local spir_func i32 @bar(i32 %arg) { | ||
| entry: | ||
| %arg.addr = alloca i32, align 4 | ||
| store i32 %arg, i32* %arg.addr, align 4 | ||
| %0 = load i32, i32* %arg.addr, align 4 | ||
| ret i32 %0 | ||
| } | ||
|
|
||
| define dso_local spir_kernel void @TU0_kernel1() #0 !intel_reqd_sub_group_size !2 { | ||
| entry: | ||
| call spir_func void @foo1() | ||
| ret void | ||
| } | ||
|
|
||
| ; Function Attrs: nounwind | ||
| define dso_local spir_func void @foo1() { | ||
| entry: | ||
| %a = alloca i32, align 4 | ||
| store i32 2, i32* %a, align 4 | ||
| ret void | ||
| } | ||
|
|
||
| define dso_local spir_kernel void @TU1_kernel2() #1 { | ||
| entry: | ||
| call spir_func void @foo2() | ||
| ret void | ||
| } | ||
|
|
||
| ; Function Attrs: nounwind | ||
| define dso_local spir_func void @foo2() { | ||
| entry: | ||
| %a = alloca i32, align 4 | ||
| %0 = load i32, i32 addrspace(4)* getelementptr inbounds ([1 x i32], [1 x i32] addrspace(4)* addrspacecast ([1 x i32] addrspace(1)* @_ZL2GV to [1 x i32] addrspace(4)*), i64 0, i64 0), align 4 | ||
| %add = add nsw i32 4, %0 | ||
| store i32 %add, i32* %a, align 4 | ||
| ret void | ||
| } | ||
|
|
||
| attributes #0 = { "sycl-module-id"="TU1.cpp" } | ||
| attributes #1 = { "sycl-module-id"="TU2.cpp" } | ||
|
|
||
| !opencl.spir.version = !{!0, !0} | ||
| !spirv.Source = !{!1, !1} | ||
|
|
||
| !0 = !{i32 1, i32 2} | ||
| !1 = !{i32 4, i32 100000} | ||
| !2 = !{i32 32} |
59 changes: 59 additions & 0 deletions
59
llvm/test/tools/sycl-post-link/device-code-split/per-reqd-sub-group-size-split-2.ll
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,59 @@ | ||
| ; The test is intended to check that sycl-post-link correctly groups kernels | ||
| ; by unique reqd_sub_group_size values used in them | ||
|
|
||
| ; RUN: sycl-post-link -split=auto -symbols -S %s -o %t.table | ||
| ; RUN: FileCheck %s -input-file=%t.table --check-prefix CHECK-TABLE | ||
| ; | ||
| ; RUN: FileCheck %s -input-file=%t_0.sym --check-prefix CHECK-M0-SYMS \ | ||
| ; RUN: --implicit-check-not kernel0 --implicit-check-not kernel1 \ | ||
| ; RUN: --implicit-check-not kernel2 | ||
| ; | ||
| ; RUN: FileCheck %s -input-file=%t_1.sym --check-prefix CHECK-M2-SYMS \ | ||
| ; RUN: --implicit-check-not kernel0 --implicit-check-not kernel3 | ||
| ; | ||
| ; RUN: FileCheck %s -input-file=%t_2.sym --check-prefix CHECK-M1-SYMS \ | ||
| ; RUN: --implicit-check-not kernel1 --implicit-check-not kernel2 \ | ||
| ; RUN: --implicit-check-not kernel3 | ||
|
|
||
| ; CHECK-TABLE: Code | ||
| ; CHECK-TABLE-NEXT: _0.sym | ||
| ; CHECK-TABLE-NEXT: _1.sym | ||
| ; CHECK-TABLE-NEXT: _2.sym | ||
| ; CHECK-TABLE-EMPTY: | ||
|
|
||
| ; CHECK-M0-SYMS: kernel3 | ||
|
|
||
| ; CHECK-M1-SYMS: kernel0 | ||
|
|
||
| ; CHECK-M2-SYMS: kernel1 | ||
| ; CHECK-M2-SYMS: kernel2 | ||
|
|
||
| target datalayout = "e-i64:64-v16:16-v24:32-v32:32-v48:64-v96:128-v192:256-v256:256-v512:512-v1024:1024" | ||
| target triple = "spir64-unknown-linux" | ||
|
|
||
| define dso_local spir_kernel void @kernel0() #0 !intel_reqd_sub_group_size !1 { | ||
| entry: | ||
| ret void | ||
| } | ||
|
|
||
| define dso_local spir_kernel void @kernel1() #0 !intel_reqd_sub_group_size !2 { | ||
| entry: | ||
| ret void | ||
| } | ||
|
|
||
| define dso_local spir_kernel void @kernel2() #0 !intel_reqd_sub_group_size !3 { | ||
| entry: | ||
| ret void | ||
| } | ||
|
|
||
| define dso_local spir_kernel void @kernel3() #0 !intel_reqd_sub_group_size !4 { | ||
| entry: | ||
| ret void | ||
| } | ||
|
|
||
| attributes #0 = { "sycl-module-id"="TU1.cpp" } | ||
|
|
||
| !1 = !{i32 32} | ||
| !2 = !{i32 64} | ||
| !3 = !{i32 64} | ||
| !4 = !{i32 16} |
54 changes: 54 additions & 0 deletions
54
llvm/test/tools/sycl-post-link/device-code-split/per-reqd-sub-group-size-split-3.ll
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,54 @@ | ||
| ; This test is intended to check that we do not perform per-reqd_sub_group_size | ||
| ; split if it was disabled through one or another sycl-post-link option | ||
|
|
||
| ; RUN: sycl-post-link -symbols -S %s -o %t.table | ||
| ; RUN: FileCheck %s -input-file=%t.table --check-prefix CHECK-TABLE | ||
| ; RUN: FileCheck %s -input-file=%t_0.ll --check-prefix CHECK-IR | ||
| ; | ||
| ; -lower-esimd is needed so sycl-post-link does not complain about no actions | ||
| ; specified | ||
| ; RUN: sycl-post-link -lower-esimd -ir-output-only -S %s -o %t.ll | ||
| ; RUN: FileCheck %s -input-file=%t.ll --check-prefix CHECK-IR | ||
|
|
||
| ; We expect to see only one module generated: | ||
| ; | ||
| ; CHECK-TABLE: Code | ||
| ; CHECK-TABLE-NEXT: _0.ll | ||
| ; CHECK-TABLE-EMPTY: | ||
|
|
||
| ; Regardless of used reqd_sub_group_size and sycl-module-id metadata, all | ||
| ; kernel and functions should still be present. | ||
|
|
||
| ; CHECK-IR-DAG: define spir_func void @foo | ||
| ; CHECK-IR-DAG: define spir_func void @bar | ||
| ; CHECK-IR-DAG: define spir_kernel void @kernel0 | ||
| ; CHECK-IR-DAG: define spir_kernel void @kernel1 | ||
|
|
||
| target datalayout = "e-i64:64-v16:16-v24:32-v32:32-v48:64-v96:128-v192:256-v256:256-v512:512-v1024:1024" | ||
| target triple = "spir64-unknown-linux" | ||
|
|
||
| define spir_func void @foo() #0 !intel_reqd_sub_group_size !1 { | ||
| ret void | ||
| } | ||
|
|
||
| define spir_func void @bar() #1 !intel_reqd_sub_group_size !2 { | ||
| ret void | ||
| } | ||
|
|
||
| define spir_kernel void @kernel0() #1 !intel_reqd_sub_group_size !2 { | ||
| entry: | ||
| ret void | ||
| } | ||
|
|
||
| define spir_kernel void @kernel1() #0 !intel_reqd_sub_group_size !3 { | ||
| entry: | ||
| call void @foo() | ||
| ret void | ||
| } | ||
|
|
||
| attributes #0 = { "sycl-module-id"="TU1.cpp" } | ||
| attributes #1 = { "sycl-module-id"="TU2.cpp" } | ||
|
|
||
| !1 = !{i32 32} | ||
| !2 = !{i32 64} | ||
| !3 = !{i32 16, i32 16} |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Oops, something went wrong.
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
It is ok to change this test, per-reqd-wg-size-split-2.ll, sycl-esimd-large-grf.ll and sycl-large-grf.ll below: by adding new optional kernel feature reqd-sub-group-size to internal data structures, re-hashing happened which affected name generation logic. This does not affect customers as customers don't use these temp files directly, and didn't break anything in the pipeline.