Matrix/joint_matrix_bf16_fill_k_cache_SLM.cpp fail on Arc Linux #20595
Description
Describe the bug
https://github.com/intel/llvm/actions/runs/19151375733/job/54743282506?pr=20542
-- Testing: 1911 of 2495 tests, 24 workers --
FAIL: SYCL :: Matrix/joint_matrix_bf16_fill_k_cache_SLM.cpp (415 of 1911)
******************** TEST 'SYCL :: Matrix/joint_matrix_bf16_fill_k_cache_SLM.cpp' FAILED ********************
Exit Code: -11
Command Output (stdout):
--
# RUN: at line 13
env ONEAPI_DEVICE_SELECTOR=level_zero:gpu /__w/llvm/llvm/build-e2e/Matrix/Output/joint_matrix_bf16_fill_k_cache_SLM.cpp.tmp_gpu_vnni.out
# executed command: env ONEAPI_DEVICE_SELECTOR=level_zero:gpu /__w/llvm/llvm/build-e2e/Matrix/Output/joint_matrix_bf16_fill_k_cache_SLM.cpp.tmp_gpu_vnni.out
# .---command stdout------------
# | Testing: 8 x 8 x 16 [TM x TN x TK]
# | DONE for size 256
# | GOPS is 135.409 Gop/s
# | Testing: 32 x 32 x 16 [TM x TN x TK]
# | DONE for size 256
# | GOPS is 75.8815 Gop/s
# `-----------------------------
# RUN: at line 13
env ONEAPI_DEVICE_SELECTOR=opencl:gpu /__w/llvm/llvm/build-e2e/Matrix/Output/joint_matrix_bf16_fill_k_cache_SLM.cpp.tmp_gpu_vnni.out
# executed command: env ONEAPI_DEVICE_SELECTOR=opencl:gpu /__w/llvm/llvm/build-e2e/Matrix/Output/joint_matrix_bf16_fill_k_cache_SLM.cpp.tmp_gpu_vnni.out
# .---command stdout------------
# | Testing: 8 x 8 x 16 [TM x TN x TK]
# | DONE for size 256
# | GOPS is 980.834 Gop/s
# | Testing: 32 x 32 x 16 [TM x TN x TK]
# | DONE for size 256
# | GOPS is 611.607 Gop/s
# `-----------------------------
# RUN: at line 13
env env UR_LOADER_USE_LEVEL_ZERO_V2=1 ONEAPI_DEVICE_SELECTOR=level_zero:gpu /__w/llvm/llvm/build-e2e/Matrix/Output/joint_matrix_bf16_fill_k_cache_SLM.cpp.tmp_gpu_vnni.out
# executed command: env env UR_LOADER_USE_LEVEL_ZERO_V2=1 ONEAPI_DEVICE_SELECTOR=level_zero:gpu /__w/llvm/llvm/build-e2e/Matrix/Output/joint_matrix_bf16_fill_k_cache_SLM.cpp.tmp_gpu_vnni.out
# .---command stdout------------
# | Testing: 8 x 8 x 16 [TM x TN x TK]
# | DONE for size 256
# | GOPS is 138.971 Gop/s
# | Testing: 32 x 32 x 16 [TM x TN x TK]
# | DONE for size 256
# | GOPS is 39.9461 Gop/s
# `-----------------------------
# RUN: at line 16
env ONEAPI_DEVICE_SELECTOR=level_zero:gpu /__w/llvm/llvm/build-e2e/Matrix/Output/joint_matrix_bf16_fill_k_cache_SLM.cpp.tmp_gpu.out
# executed command: env ONEAPI_DEVICE_SELECTOR=level_zero:gpu /__w/llvm/llvm/build-e2e/Matrix/Output/joint_matrix_bf16_fill_k_cache_SLM.cpp.tmp_gpu.out
# .---command stdout------------
# | Testing: 8 x 8 x 16 [TM x TN x TK]
# | DONE for size 256
# | GOPS is 112.432 Gop/s
# | Testing: 32 x 32 x 16 [TM x TN x TK]
# `-----------------------------
# error: command failed with exit status: -11
--
To reproduce
No response
Environment
- OS: [e.g Windows/Linux]
- Target device and vendor: [e.g. Intel GPU]
- DPC++ version: [e.g. commit hash or output of
clang++ --version] - Dependencies version: [e.g. the output of
sycl-ls --verbose]
Additional context
No response