- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
As described above, have the mkl_?csrcoo procedures been deleted from the current 2026.0 version of the MKL toolkit, as I'm getting "lld-link: error: undefined symbol: MKL_DCSRCOO" when using IFX 2026.0.
I know these procedures are decrecated, but there is no mention of their removal in Intel®oneAPI Math Kernel Library (oneMKL) Release Notes and System...
This is on Windows/VS2022, I haven't got around to checking if this builds on Linux yet. The same solution builds fine with IFX 2025.3.1/MKL 2025.3
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi Mark_Lewy,
Thank you for reaching out and for bringing this issue to our attention.
Most of the deprecated NIST-like Sparse BLAS routines were removed in the 2026.0 release. Unfortunately, their removal was not documented in the release notes due to an oversight on our part. This has now been corrected. We apologize for any inconvenience this may have caused.
We recommend using the equivalent I/E Sparse BLAS routines instead:
- The sparse conversion routines are documented in the 2025.2 Developer Reference, and this documentation remains applicable: https://www.intel.com/content/www/us/en/docs/onemkl/developer-reference-c/2025-2/matrix-manipulation-routines.html
- The dense conversion routines are documented in the 2025.3 Release Notes: https://www.intel.com/content/www/us/en/developer/articles/release-notes/onemkl-release-notes.html
I hope this helps.
Best,
Nicolas
Link Copied
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi Mark_Lewy,
Thank you for reaching out and for bringing this issue to our attention.
Most of the deprecated NIST-like Sparse BLAS routines were removed in the 2026.0 release. Unfortunately, their removal was not documented in the release notes due to an oversight on our part. This has now been corrected. We apologize for any inconvenience this may have caused.
We recommend using the equivalent I/E Sparse BLAS routines instead:
- The sparse conversion routines are documented in the 2025.2 Developer Reference, and this documentation remains applicable: https://www.intel.com/content/www/us/en/docs/onemkl/developer-reference-c/2025-2/matrix-manipulation-routines.html
- The dense conversion routines are documented in the 2025.3 Release Notes: https://www.intel.com/content/www/us/en/developer/articles/release-notes/onemkl-release-notes.html
I hope this helps.
Best,
Nicolas
- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page