Repository navigation
Reject a frame announcing more bytes than follow it - #28
Open
bavulapati wants to merge 1 commit into
Open
bavulapati wants to merge 1 commit into
bavulapati wants to merge 1 commit into
Conversation
Decode trusted the announced length, so a short frame read its value and left the cursor past the end of the buffer, where state.remaining goes negative and everything decoded after it works from a position that does not exist. Closes #18.
This branch has not been deployed
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
_Frame.decodetook the announced length on trust: it setstate.endto where the frame claimed to end without checking that against where the buffer actually ends. So0501, five bytes announced with one behind it, decoded as1and leftstate.startat 6 in a 2-byte buffer.The wrong answer is the smaller half.
state.remaininggoes negative, so every later read works from a position that does not exist, and a truncated or hostile frame corrupts what follows it rather than failing where the problem is.One comparison fixes it, before
state.endmoves. Reading the value inside the announced length and rejecting what runs past it is a separate rule,frame-bounds-its-value, which this port already passed andtest_frame_decode_restores_end_on_inner_erroralready covers.Conformance against
compact-encoding-testgoes from 171 matched / 3 diverged to 172 / 2, closingframe-uint-length-past-the-end. The two left are #19.Closes #18.