vertexcodec: Reduce tail padding for v1 #825
Merged
+53
−57
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
Instead of checking v0 & 4 levels of v1, only check a single combination. This allows exploration to happen faster. We can then use the extra time to re-encode the data at the boundary of the encoding size. This checks that the encoder and decoder is memory safe.
unreachable comment marks the line as dead for coverage analysis.
We were using 15 bytes per line which is confusing to inspect as it does not align with hex dumps.
When v0 was implemented, the decode limit was thought to be 32 bytes; hence the tail had to be at least 32 bytes to cover the decode overhang. Since then, decode limit was revised to 24 bytes (8 bytes of 4-bit data plus 16 bytes of outliers), but the tail remained the same as the format is fixed. For v1, the decode limit is the same and this is unlikely to change for future revisions either since 6-bit groups did not end up working out (they would have required 28 bytes of tail space), so it is time to reduce the tail to 24 bytes for v1. This only impacts small blobs but makes the format a little cleaner.
Ignoring trivial bitgroups is sometimes detrimental for performance, so for now just remove this; also split a line that's too long in two.
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Outdated suggestions cannot be applied.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.
When v0 was implemented, the decode limit was thought to be 32 bytes;
hence the tail had to be at least 32 bytes to cover the decode overhang.
Since then, decode limit was revised to 24 bytes (8 bytes of 4-bit data
plus 16 bytes of outliers), but the tail remained the same as the format
is fixed. For v1, the decode limit is the same and this is unlikely to
change for future revisions either since 6-bit groups did not end up
working out (they would have required 28 bytes of tail space), so it is
time to reduce the tail to 24 bytes for v1.
This only impacts small blobs but makes the format a little cleaner.
This contribution is sponsored by Valve.