Fix new issue 18 from 201911 #37
Merged
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
Deduplicate block compression logic by refactoring the first window behavior from being a separate code block into specific conditionals where the first window behavior differs. All tests pass, included ignored.
Thanks to @lkolbly for the failing data. See image-rs#18.
Fixes image-rs#18. May fix image-rs/image-png#147.
|
Awesome, been a bit busy so didn't get around to looking too much into it myself yet. Refactor is probably a good thing. A bit surprised that it took this long for the bug to appear. |
|
Published version with fix to crates. |
|
Thanks for the quick release!
As am I -- I've been using the png crate without issue for two years and have put a few hundred TB through it (some of which I'm going back and validating). This has made me a bit nervous, though, so I've set up an AFL fuzz on |
This was referenced Dec 17, 2019
Closed
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Outdated suggestions cannot be applied.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.
The issue was occurring when attempting to flush lz77 (so
datawas empty), such that the first window was filling the buffer.Includes a failing test case which passes after the fix. Also includes a deduplicating cleanup of
lz77_compress_block-- not knowing anything about lz77, I couldn't diagnose the issue until this refactoring. The refactoring makes it explicit where the first window behavior differs from the normal case, which made it clearer to understand and calls out necessary special behavior (likeoverlapnot getting added tocurrent_block_input_bytesto prevent off-by-one-errors in this particular case, which is different from all the others).I also confirmed that this fixes my bug in image-rs/image-png#147. I haven't checked @newpavlov's, but since it fixes the reduced case from that created by @lkolbly, it should.