Cull images and free VRAM #163
Closed
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
The `image::Cache` takes a more optimistic approach; only trimming when new memory has been allocated.
|
This feature will be really useful in one of the apps I'm building. At least, it's a great start, as you noted. Does any more work need to be done here before it's merged in? |
|
@cryptoquick This is a bit related to the texture atlas implementation in #154. The new storage strategy may allow us to do a better job at invalidating cache regions. Therefore, I think we should unify these changes in that PR first. I will work on it once I start batching work again in |
|
Glad to hear, it all sounds close! |
|
Closing, as I will most likely tackle this once I move the texture atlas logic to |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Fixes #160.
This PR changes the
iced_wgpu::image::raster::Cacheto unload images from the GPU when they are no longer visible and new ones have been uploaded.The change may cause applications to block more oftenly when hiding/displaying different images (like a scrolling image gallery). However, they are much less likely to crash due to running out of video memory. We should be able to improve this strategy in #154 by trying to deallocate before creating a new layer instead.
In the long run, we should load and upload images asynchronously to avoid dropping frames.