malyzajko / daisy

Other
42 stars 10 forks source link

Hello,Can I specify precision (like only use double and quad) when I use mixed-precision function #11

Closed zhangsanLIS closed 2 years ago

zhangsanLIS commented 2 years ago

I found that when I use flags like --mixed-tuning --lang=C --codegen , the code will appear qd_real type, but what I want is only quad and double, what should I do to achieve my goal, thanks a lot

malyzajko commented 2 years ago

Master still has the older version of the code generator. You can try the 'regimes' branch, which is using quadmath.h. Only these two are currently supported. (The mixed-precision tuning algorithm is the same on both branches.)

zhangsanLIS commented 2 years ago

Master still has the older version of the code generator. You can try the 'regimes' branch, which is using quadmath.h. Only these two are currently supported. (The mixed-precision tuning algorithm is the same on both branches.)

sorry the question has been solved