Uh oh!
There was an error while loading. Please reload this page.
lib: fix modulo sign in webidl convertToInt - #62940
Closed
deepview-autofix wants to merge 1 commit into
Closed
Conversation
ChALkeR
marked this pull request as ready for review
April 24, 2026 23:12
ChALkeRforce-pushed
the
deepview/415d17df0c
branch
from
April 24, 2026 23:15
a9f91ee to
65d51ecCompareChALkeR
previously requested changes
Apr 24, 2026
ChALkeRforce-pushed
the
deepview/415d17df0c
branch
from
April 24, 2026 23:56
65d51ec to
cae7693CompareChALkeRforce-pushed
the
deepview/415d17df0c
branch
2 times, most recently
from
April 25, 2026 00:18
34e138f to
0f245d2CompareChALkeR
marked this pull request as draft
April 25, 2026 01:14
Codecov Report✅ All modified and coverable lines are covered by tests. Additional details and impacted files@@ Coverage Diff @@## main #62940 +/- ##
==========================================
- Coverage 89.66% 89.63% -0.03%
==========================================
Files 706 706 Lines 219370 219424 +54 Branches 42065 42074 +9 ==========================================
- Hits 196694 196681 -13 - Misses 14586 14633 +47 - Partials 8090 8110 +20
🚀 New features to boost your workflow:
|
ChALkeR
commented
Apr 25, 2026
Member
Calculation at 64 bits is still off, will redo |
ChALkeRforce-pushed
the
deepview/415d17df0c
branch
4 times, most recently
from
April 26, 2026 10:12
598cfec to
0c87f49CompareMember
I rewrote the whole thing and added more tests |
ChALkeR
marked this pull request as ready for review
April 26, 2026 10:12
ChALkeRforce-pushed
the
deepview/415d17df0c
branch
from
April 26, 2026 10:16
0c87f49 to
a5ec26fCompareCo-Authored-By: DeepView Autofix <276251120+deepview-autofix@users.noreply.github.com> Assisted-by: Claude <noreply@anthropic.com> Signed-off-by: Nikita Skovoroda <chalkerx@gmail.com>
ChALkeRforce-pushed
the
deepview/415d17df0c
branch
from
April 26, 2026 10:37
a5ec26f to
76ad558Compare| // 9. Set x to IntegerPart(x). | ||
| x = integerPart(x); | ||
| // Following the exact algorithm looses precision on signed modulo in 64-bit |
Contributor
There was a problem hiding this comment.
Suggested change
| // Following the exact algorithm looses precision on signed modulo in 64-bit | |
| // Following the exact algorithm loses precision on signed modulo in 64-bit |
ChALkeR
commented
Apr 26, 2026
Member
Closing in favor of #62979 |
This file contains hidden or 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
Sign up for freeto join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
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 modulo() helper in lib/internal/webidl.js is documented to implement ECMA-262's "x modulo y" (result has the same sign as y), but it used the JS % operator directly, whose result takes the sign of x. convertToInt step 10 relies on modulo(x, 2^bitLength) producing a non-negative representative in [0, 2^bitLength); step 11 only handles the signed upper half, so negative results of % were returned unchanged.
As a result, negative inputs to unsigned WebIDL integer types returned the original negative value instead of the two's-complement wrap (e.g. convertToInt('v', -3, 8) returned -3 instead of 253), and signed inputs in the one-wrap band below the lower bound were returned as-is (convertToInt('v', -200, 8, { signed: true }) returned -200 instead of 56). This diverged from typed-array coercion.