Skip to content

[SYCL][Driver] Link Native Bfloat16 devicelib for PTL #19940

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.

Sign up for GitHub

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

Merged
merged 3 commits into from
Sep 4, 2025
Merged

[SYCL][Driver] Link Native Bfloat16 devicelib for PTL #19940

Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
70275dd
[SYCL][Driver] Link Native Bfloat16 devicelib for PTL
jinge90 Sep 2, 2025
ad5ead5
Add lit test for clang_cl
jinge90 Sep 3, 2025
708dd56
Merge remote-tracking branch 'upstream/sycl' into native_bf16_devicel…
jinge90 Sep 4, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
5 changes: 3 additions & 2 deletions clang/lib/Driver/ToolChains/SYCL.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -248,7 +248,8 @@ static bool selectBfloatLibs(const llvm::Triple &Triple, const Compilation &C,
static llvm::SmallSet<StringRef, 8> GPUArchsWithNBF16{
"intel_gpu_pvc", "intel_gpu_acm_g10", "intel_gpu_acm_g11",
"intel_gpu_acm_g12", "intel_gpu_dg2_g10", "intel_gpu_dg2_g11",
"intel_dg2_g12", "intel_gpu_bmg_g21", "intel_gpu_lnl_m"};
"intel_dg2_g12", "intel_gpu_bmg_g21", "intel_gpu_lnl_m",
"intel_gpu_ptl_h", "intel_gpu_ptl_u"};
const llvm::opt::ArgList &Args = C.getArgs();
bool NeedLibs = false;

Expand Down Expand Up @@ -290,7 +291,7 @@ static bool selectBfloatLibs(const llvm::Triple &Triple, const Compilation &C,
auto checkBF = [](StringRef Device) {
return Device.starts_with("pvc") || Device.starts_with("ats") ||
Device.starts_with("dg2") || Device.starts_with("bmg") ||
Device.starts_with("lnl");
Device.starts_with("lnl") || Device.starts_with("ptl");
};

auto checkSpirvJIT = [](StringRef Target) {
Expand Down
22 changes: 22 additions & 0 deletions clang/test/Driver/sycl-device-lib-bfloat16.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -189,6 +189,28 @@
// RUN: --sysroot=%S/Inputs/SYCL -Xsycl-target-backend=spir64_gen "-device lnl_m" \
// RUN: %s -### 2>&1 | FileCheck %s -check-prefix=BFLOAT16-FALLBACK-FALLBACK

// Test AOT-PTL + AOT-LNL specified via different options, uses native libs
// RUN: %clangxx -fsycl -fsycl-targets=intel_gpu_lnl_m,spir64_gen \
// RUN: --sysroot=%S/Inputs/SYCL -Xsycl-target-backend=spir64_gen "-device ptl_h" \
// RUN: %s -### 2>&1 | FileCheck %s -check-prefix=BFLOAT16-NATIVE-NATIVE
// RUN: %clangxx -fsycl -fsycl-targets=intel_gpu_lnl_m,spir64_gen \
Copy link
Contributor

@srividya-sundaram srividya-sundaram Sep 2, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Are these supported in clang_cl mode?
If yes, please add test cases for the same.

Copy link
Contributor Author

@jinge90 jinge90 Sep 3, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Hi, @srividya-sundaram
Yes, these tests should work with clang_cl mode.
Thanks very much.

// RUN: --sysroot=%S/Inputs/SYCL -Xsycl-target-backend=spir64_gen "-device ptl_u" \
// RUN: %s -### 2>&1 | FileCheck %s -check-prefix=BFLOAT16-NATIVE-NATIVE
// RUN: %clangxx -fsycl -fsycl-targets=intel_gpu_ptl_h,spir64_gen \
// RUN: --sysroot=%S/Inputs/SYCL -Xsycl-target-backend=spir64_gen "-device lnl_m" \
// RUN: %s -### 2>&1 | FileCheck %s -check-prefix=BFLOAT16-NATIVE-NATIVE
// RUN: %clangxx -fsycl -fsycl-targets=intel_gpu_ptl_u,spir64_gen \
// RUN: --sysroot=%S/Inputs/SYCL -Xsycl-target-backend=spir64_gen "-device lnl_m" \
// RUN: %s -### 2>&1 | FileCheck %s -check-prefix=BFLOAT16-NATIVE-NATIVE

// Test AOT-PTL + AOT-DG1 specified via different options, uses fallback libs
// RUN: %clangxx -fsycl -fsycl-targets=intel_gpu_ptl_h,spir64_gen \
// RUN: --sysroot=%S/Inputs/SYCL -Xsycl-target-backend=spir64_gen "-device dg1" \
// RUN: %s -### 2>&1 | FileCheck %s -check-prefix=BFLOAT16-FALLBACK-FALLBACK
// RUN: %clangxx -fsycl -fsycl-targets=intel_gpu_dg1,spir64_gen \
// RUN: --sysroot=%S/Inputs/SYCL -Xsycl-target-backend=spir64_gen "-device ptl_u" \
// RUN: %s -### 2>&1 | FileCheck %s -check-prefix=BFLOAT16-FALLBACK-FALLBACK

// BFLOAT16-NOT: llvm-link{{.*}} "{{.*}}libsycl-{{fallback|native}}-bfloat16.bc"

// BFLOAT16-NATIVE: llvm-link{{.*}} "{{.*}}libsycl-native-bfloat16.bc"
Expand Down