C++: Fix missing bounds in range analysis #10555
Merged
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Outdated suggestions cannot be applied.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.
This PR adds bounds for the following case:
This is motivated by CVE-2016-10160, and the first commit adds a reduced version of that vulnerability (and indeed after this PR we correctly identify this CVE).
The reason we didn't find the issue in the first commit is as follows:
Currently, on
main
, the range-analysis library didn't spot that the pointer-arithmetic expressionp + index
in thep + index > end
guard is identical to the one being used for the store operation inp[index] = '\0'
, and so we didn't realize that there's an off-by-one error here.In contrast, when we're storing the result of the pointer-arithmetic expression in a local variable
q
, we correctly catch this.The second commit fixes this by generating a new
SSAVariable
representing pointer-arithmetic expressions appearing in guards, and using GVN to connect thep + index
in the guard with the pointer-arithmetic expression computing the address ofp[index]
.(And for completeness: The last commit accepts the test changes.)