Fix double shufti's vector end false positive #325
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
Double shufti used to offset one vector, resulting in losing one character at the end of every vector. This was replaced by a magic value indicating a match. This meant that if the first char of a pattern fell on the last char of a vector, double shufti would assume the second character is present and report a match. This patch fixes it by keeping the previous vector and feeding its data to the new one when we shift it, preventing any loss of data. Signed-off-by: Yoan Picchi <[email protected]>
ypicchi-arm
force-pushed
the
shufti_fix_release
branch
from
de6f62b to
3351539
Compare
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Outdated suggestions cannot be applied.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.
Double shufti used to offset one vector, resulting in losing one character
at the end of every vector. This was replaced by a magic value indicating a
match. This meant that if the first char of a pattern fell on the last char of
a vector, double shufti would assume the second character is present and
report a match.
This patch fixes it by keeping the previous vector and feeding its data to the
new one when we shift it, preventing any loss of data.
This have been tested on arm (neon, sve, sve2) and on intel (sse2, avx2, avx512). I couldn't measure any meaningful bandwidth difference between the patched version and the previous version on arm.