search

SciML / LinearSolve.jl

LinearSolve.jl: High-Performance Unified Interface for Linear Solvers in Julia. Easily switch between factorization and Krylov methods, add preconditioners, and all in one interface.
https://docs.sciml.ai/LinearSolve/stable/
Other
244 stars 52 forks source link

Don't use AppleAccelerate on number types that are not floats #411

Closed SebastianM-C closed 10 months ago

SebastianM-C commented 10 months ago

@ChrisRackauckas I saw that this was done for MKL in #390, but it seems to be missing for AppleAccelerate. Is this the correct way to handle it?

ElisabethRoesch commented 10 months ago

I can confirm this fixed the issue on my Mac. Prior to the fix I got

│    MethodError: no method matching aa_getrf!(::Matrix{ForwardDiff.Dual{....
codecov[bot] commented 10 months ago

Codecov Report

Merging #411 (7f4d23b) into main (7527328) will decrease coverage by 0.56%. The diff coverage is 100.00%.

@@            Coverage Diff             @@
##             main     #411      +/-   ##
==========================================
- Coverage   69.92%   69.36%   -0.56%     
==========================================
  Files          26       26              
  Lines        1965     1965              
==========================================
- Hits         1374     1363      -11     
- Misses        591      602      +11
Files Coverage Δ
src/default.jl 66.32% <100.00%> (-1.03%) :arrow_down:

... and 3 files with indirect coverage changes

:mega: We’re building smart automated test selection to slash your CI/CD build times. Learn more

ChrisRackauckas commented 10 months ago

yup this is correct thanks