Skip to content
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

Invalidate some bind group on buffer reallocate #383

Open
wants to merge 3 commits into
base: main
Choose a base branch
from

Conversation

djeedai
Copy link
Owner

@djeedai djeedai commented Oct 17, 2024

Invalidate the sim bind group when the particle buffer is re-allocated. Also always re-create that bind group for now, as we don't invalidate on binding offset/size changes. With that, revert the fix for #330, which shouldn't be needed because the buffer is indeed contiguous.

Invalidate the sim bind group when the particle buffer is re-allocated. Also
always re-create that bind group for now, as we don't invalidate on binding
offset/size changes. With that, revert the fix for #330, which shouldn't be
needed because the buffer is indeed contiguous.
@djeedai djeedai added C - bug Something isn't working A - internal Internal change on a core system labels Oct 17, 2024
@djeedai
Copy link
Owner Author

djeedai commented Oct 17, 2024

@pcwalton and @NiseVoid, any chance you folks can try this change with your respective issue? For @pcwalton, the one we're discussing with the group atomic reorder, which this change doesn't fix but just to check that it's not making anything worse; and @NiseVoid, with the original change you tried which yielded the original #330 fix which is reverted here? I think this removes the confusion @pcwalton pointed and which is carried with code comment, that the buffer is indeed contiguous, which was assumed to be false in the fix for #330 (but was probably because the bind group was reusing an old buffer instead of the newly-allocated one). Thanks a lot in advance!

@djeedai
Copy link
Owner Author

djeedai commented Oct 29, 2024

Note: review the instancing example before merging... it seems there's some issue sometimes with particles getting stuck, although it's unclear if this change is responsible (need to bisect) or it's a separate bug / a limitation due to capacity reached maybe?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
A - internal Internal change on a core system C - bug Something isn't working
Projects
None yet
Development

Successfully merging this pull request may close these issues.

1 participant