-
Notifications
You must be signed in to change notification settings - Fork 88
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Temporarily disable optimized batched solver instantiations #1652
Conversation
GKO_NOT_IMPLEMENTED; | ||
} | ||
} | ||
// if (sconf.prec_shared) { |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
nit: IMO it could just be deleted directly, we still have the code in the git log, so it won't be lost. But I'm also fine with leaving it as it is, which probably highlights that it's only temporary.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I feel the same way, but since this is temporary, it shouldn't matter much
It currently forces the behavior. |
Yes, currently we decided to temporarily remove the optimized instantiations. A more permanent solution would be something of the form in #1629 We decided not to go with the config options as that is maybe more annoying for users. Currently a few projects use the optimized versions, but I think they rely on the previous release, so it should be okay. We will move to the split version soon anyway. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM. The config header is unnecessary now.
How do we ensure the optimized version is still enabled in the next release? maybe make some note in changelog or create an issue until the splitting compilation pr is merged.
Created an issue #1661, with the milestone 1.9.0, to track it. |
This PR temporarily disables the instantiations of the different batched solver kernels to reduce compilation time.