Skip to content

[rocBLAS] adds axpy replacement reference code when using windows OpenBLAS#7184

Merged
TorreZuk merged 2 commits into
developfrom
users/torrezuk/rocm-23872-add-axpy-ref
May 13, 2026
Merged

[rocBLAS] adds axpy replacement reference code when using windows OpenBLAS#7184
TorreZuk merged 2 commits into
developfrom
users/torrezuk/rocm-23872-add-axpy-ref

Conversation

@TorreZuk
Copy link
Copy Markdown
Contributor

@TorreZuk TorreZuk commented May 7, 2026

This pull request introduces a Windows-specific implementation of the ref_axpy function to address compatibility issues when using OpenBLAS on Windows. The main changes involve conditionally compiling a new template implementation for ref_axpy and updating the header to ensure the correct version is used based on the platform and build configuration.

Platform-specific implementation for Windows:

  • Added a custom template implementation of ref_axpy for Windows platforms when BLIS CBLAS is not enabled, including specializations for float, double, rocblas_float_complex, rocblas_double_complex, and a custom implementation for rocblas_bfloat16 that performs computation in float precision. (projects/rocblas/clients/common/cblas_interface.cpp)

Header updates for conditional compilation:

  • Updated the declaration and inline implementation of ref_axpy in cblas_interface.hpp to use the new Windows-specific override only when appropriate, ensuring the standard implementation is excluded on Windows/OpenBLAS builds without BLIS CBLAS. (projects/rocblas/clients/include/cblas_interface.hpp)

ROCM-23872

@TorreZuk TorreZuk requested a review from a team as a code owner May 7, 2026 22:37
@TorreZuk TorreZuk added the TestLevel1Only Tests only Level 1 functions in this PR label May 7, 2026
@codecov-commenter
Copy link
Copy Markdown

codecov-commenter commented May 7, 2026

Codecov Report

✅ All modified and coverable lines are covered by tests.

❌ Your project status has failed because the head coverage (69.24%) is below the target coverage (80.00%). You can increase the head coverage or adjust the target coverage.

Additional details and impacted files
@@           Coverage Diff            @@
##           develop    #7184   +/-   ##
========================================
  Coverage    62.78%   62.78%           
========================================
  Files          972      972           
  Lines       185366   185364    -2     
  Branches     21999    21997    -2     
========================================
  Hits        116376   116376           
+ Misses       61008    61006    -2     
  Partials      7982     7982           
Flag Coverage Δ *Carryforward flag
hipBLAS 90.65% <ø> (ø) Carriedforward from 4e76a3f
hipBLASLt 39.86% <ø> (+0.01%) ⬆️ Carriedforward from 4e76a3f
hipDNN 85.56% <ø> (ø) Carriedforward from 4e76a3f
hipFFT 49.22% <ø> (ø) Carriedforward from 4e76a3f
hipSOLVER 69.24% <ø> (ø) Carriedforward from 4e76a3f
rocBLAS 48.11% <ø> (ø)
rocFFT 47.60% <ø> (ø) Carriedforward from 4e76a3f
rocRAND 57.02% <ø> (ø) Carriedforward from 4e76a3f

*This pull request uses carry forward flags. Click here to find out more.

🚀 New features to boost your workflow:
  • ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
  • 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

Copy link
Copy Markdown
Contributor

@evedovelli evedovelli left a comment

Choose a reason for hiding this comment

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

LGTM!

Comment thread projects/rocblas/clients/include/cblas_interface.hpp
@TorreZuk TorreZuk force-pushed the users/torrezuk/rocm-23872-add-axpy-ref branch from d3f38d9 to 755ad37 Compare May 12, 2026 22:22
@TorreZuk TorreZuk merged commit 5f2edcb into develop May 13, 2026
40 of 42 checks passed
@TorreZuk TorreZuk deleted the users/torrezuk/rocm-23872-add-axpy-ref branch May 13, 2026 13:46
assistant-librarian Bot pushed a commit to ROCm/rocBLAS that referenced this pull request May 13, 2026
[rocBLAS]  adds axpy replacement reference code when using
 windows OpenBLAS (#7184)

This pull request introduces a Windows-specific implementation of the
`ref_axpy` function to address compatibility issues when using OpenBLAS
on Windows. The main changes involve conditionally compiling a new
template implementation for `ref_axpy` and updating the header to ensure
the correct version is used based on the platform and build
configuration.

Platform-specific implementation for Windows:

* Added a custom template implementation of `ref_axpy` for Windows
platforms when BLIS CBLAS is not enabled, including specializations for
`float`, `double`, `rocblas_float_complex`, `rocblas_double_complex`,
and a custom implementation for `rocblas_bfloat16` that performs
computation in float precision.
(`projects/rocblas/clients/common/cblas_interface.cpp`)

Header updates for conditional compilation:

* Updated the declaration and inline implementation of `ref_axpy` in
`cblas_interface.hpp` to use the new Windows-specific override only when
appropriate, ensuring the standard implementation is excluded on
Windows/OpenBLAS builds without BLIS CBLAS.
(`projects/rocblas/clients/include/cblas_interface.hpp`)

ROCM-23872
aledudek pushed a commit that referenced this pull request May 20, 2026
…nBLAS (#7184)

This pull request introduces a Windows-specific implementation of the
`ref_axpy` function to address compatibility issues when using OpenBLAS
on Windows. The main changes involve conditionally compiling a new
template implementation for `ref_axpy` and updating the header to ensure
the correct version is used based on the platform and build
configuration.

Platform-specific implementation for Windows:

* Added a custom template implementation of `ref_axpy` for Windows
platforms when BLIS CBLAS is not enabled, including specializations for
`float`, `double`, `rocblas_float_complex`, `rocblas_double_complex`,
and a custom implementation for `rocblas_bfloat16` that performs
computation in float precision.
(`projects/rocblas/clients/common/cblas_interface.cpp`)

Header updates for conditional compilation:

* Updated the declaration and inline implementation of `ref_axpy` in
`cblas_interface.hpp` to use the new Windows-specific override only when
appropriate, ensuring the standard implementation is excluded on
Windows/OpenBLAS builds without BLIS CBLAS.
(`projects/rocblas/clients/include/cblas_interface.hpp`)

ROCM-23872
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

organization: ROCm project: rocblas TestLevel1Only Tests only Level 1 functions in this PR

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants