Sporadic fails on Intel GPU in Reduction tests on Windows

### Describe the bug

BMG:
```
FAIL: SYCL :: Reduction/regression_after_pr_6343.cpp (845 of 2466)
******************** TEST 'SYCL :: Reduction/regression_after_pr_6343.cpp' FAILED ********************
Exit Code: 3221225477

Command Output (stdout):
--
# RUN: at line 2
env ONEAPI_DEVICE_SELECTOR=level_zero:gpu D:\github\_work\llvm\llvm\build-e2e\Reduction\Output\regression_after_pr_6343.cpp.tmp.out
# executed command: env ONEAPI_DEVICE_SELECTOR=level_zero:gpu 'D:\github\_work\llvm\llvm\build-e2e\Reduction\Output\regression_after_pr_6343.cpp.tmp.out'
# note: command had no output on stdout or stderr
# error: command failed with exit status: 0xc0000005

--

********************
```
https://github.com/intel/llvm/actions/runs/16664214823/job/47210032757?pr=19649

GEN12:
```
FAIL: SYCL :: Reduction/reduction_big_data.cpp (1115 of 2466)
******************** TEST 'SYCL :: Reduction/reduction_big_data.cpp' FAILED ********************
Exit Code: 3221225477

Command Output (stdout):
--
# RUN: at line 2
env ONEAPI_DEVICE_SELECTOR=level_zero:gpu D:\github\_work\llvm\llvm\build-e2e\Reduction\Output\reduction_big_data.cpp.tmp.out
# executed command: env ONEAPI_DEVICE_SELECTOR=level_zero:gpu 'D:\github\_work\llvm\llvm\build-e2e\Reduction\Output\reduction_big_data.cpp.tmp.out'
# .---command stdout------------
# | Device: Intel(R) Iris(R) Xe Graphics, MaxWGSize: 512, LocalMemSize: 65536, Driver: 1.6.33511
# | Start, T=float, BOp=struct sycl::_V1::maximum, Range={{262656}, {512}}. PASSED
# `-----------------------------
# error: command failed with exit status: 0xc0000005

--

********************
```

https://github.com/intel/llvm/actions/runs/16674608641/job/47199717089

### 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_