-
Notifications
You must be signed in to change notification settings - Fork 160
extend_mode_normalized
and extend_mode
#974
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
Conversation
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.
One thought on testing. Once that's addressed, I'll be happy to land. I don't think this is going to be meaningfully reviewed by anyone else anytime soon.
It works on my local ubuntu machine but not in CI. Is this the case for |
The windows failure certainly would be caused by the lack of GPU support for those runners. I'm not sure what the ubuntu failure is. Unfortunately, #892 means that the "current version" uploading of the snapshots is currently broken. I'll open a PR to fix that today. |
And if you can rebase #1018, that would make debugging the failure on ubuntu more plausible to debug |
Signed-off-by: sagudev <[email protected]>
Signed-off-by: sagudev <[email protected]>
Signed-off-by: sagudev <[email protected]>
rebased! |
I'm assuming that the failure on Ubuntu is because we're using a fallback CPU implementation of Vulkan, and therefore the precision is too high. What isn't at all clear to me is why the similar test from #971 passed on that version. I think just marking the new tests as |
Signed-off-by: sagudev <[email protected]>
because CI uses SW which is more accurate then real GPUs Signed-off-by: sagudev <[email protected]>
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.
One small detail then we can land this. Thanks!
Signed-off-by: sagudev <[email protected]>
Current
extend_mode
is renamed toextend_mode_normalized
(because it require normalized coordinates). To avoid rounding errors we introduceextend_mode
that relaxes normalization requirement. Currently only EXTEND_PAD is implemented differently, others fallback toextend_mode_normalized
for now.partial fix #972