std: fix writeSplatHeaderLimit bug affecting chunked HTTP#25360
Open
mbrock wants to merge 1 commit into
Open
Conversation
The `writeSplatHeaderLimitFinish` logic correctly takes the limited prefix slice of the header, but fails to chop off excess in the data vector slices, causing a drain longer than the limit. I spotted this by accident, and found one usage in std, in the HTTP chunked encoding writer, and the issue ziglang#24944 that I speculate this commit will fix. I was able to partly reproduce it on Linux but only by disabling sendfile and some other weird kludges.
Member
|
This looks like a nice bug fix, but, unfortunately, this does not fix #24944 (see #24944 (comment)) |
squeek502
reviewed
Oct 5, 2025
Comment on lines
+934
to
+942
| // Test inspired by a bug in one of Io.Writer's write helpers. | ||
| // | ||
| // The fix doesn't touch the HTTP code, but the BodyWriter is | ||
| // from what I can tell the only affected code in std. | ||
| // | ||
| // Possibly the cause of https://github.com/ziglang/zig/issues/24944 | ||
| // wherein the `zig std` server on Windows, due to lack of sendfile, | ||
| // drains the streaming sources tarball by doing vector writes | ||
| // of file contents in the chunked body stream. |
Member
There was a problem hiding this comment.
This type of information should go in the commit message rather than a comment. I'd say just remove the comment (without moving it to the commit message), since the fix in this PR doesn't actually affect the problem in #24944 (see my other comment)
Member
|
The fix was included as part of #25981 Would you mind rebasing and just including the new test cases? |
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 free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 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
writeSplatHeaderLimitFinishlogic correctly takes the limited prefix slice of the header, but fails to chop off excess in the data vector slices, causing a drain longer than the limit.I spotted this by accident, and found one usage in std, in the HTTP chunked encoding writer, and the issue #24944 that I speculate this commit will fix. I was able to partly reproduce it on Linux but only by disabling sendfile and some other weird kludges.
First time hacking the Zig std. I verified tests failing and then passing with