This repository has been archived by the owner on Jan 4, 2025. It is now read-only.
chore(deps): update module github.com/gorilla/schema to v1.4.1 [security] - autoclosed #73
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Outdated suggestions cannot be applied.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 contains the following updates:
v1.2.0
->v1.4.1
GitHub Vulnerability Alerts
CVE-2024-37298
Details
Running
schema.Decoder.Decode()
on a struct that has a field of type[]struct{...}
opens it up to malicious attacks regarding memory allocations, taking advantage of the sparse slice functionality. For instance, in the Proof of Concept written below, someone can specify to set a field of the billionth element and it will allocate all other elements before it in the slice.In the local environment environment for my project, I was able to call an endpoint like
/innocent_endpoint?arr.10000000.X=1
and freeze my system from the memory allocation while parsingr.Form
. I think this line is responsible for allocating the slice, although I haven't tested to make sure, so it's just an educated guess.Proof of Concept
The following proof of concept works on both v1.2.0 and v1.2.1. I have not tested earlier versions.
Impact
Any use of
schema.Decoder.Decode()
on a struct with arrays of other structs could be vulnerable to this memory exhaustion vulnerability. There seems to be no possible solution that a developer using this library can do to disable this behaviour without fixing it in this project, so all uses of Decode that fall under this umbrella are affected. A fix that doesn't require a major change may also be harder to find, since it could break compatibility with some other intended use-cases.Release Notes
gorilla/schema (github.com/gorilla/schema)
v1.4.1
Compare Source
Security Release
Fixes an issue where sparse slice deserialization can cause memory exhaustion CVE-2024-37298
Thanks to @AlexVasiluta for the report and following responsible disclosure.
Full Changelog: gorilla/schema@v1.4.0...v1.4.1
v1.4.0
Compare Source
What's Changed
New Contributors
Full Changelog: gorilla/schema@v1.3.0...v1.3.1
v1.3.0
Compare Source
What's Changed
New Contributors
Full Changelog: gorilla/schema@v1.2.1...v1.3.0
v1.2.1
Compare Source
What's Changed
New Contributors
Full Changelog: gorilla/schema@v1.2.0...v1.2.1
Configuration
📅 Schedule: Branch creation - "" (UTC), Automerge - At any time (no schedule defined).
🚦 Automerge: Disabled by config. Please merge this manually once you are satisfied.
♻ Rebasing: Whenever PR becomes conflicted, or you tick the rebase/retry checkbox.
🔕 Ignore: Close this PR and you won't be reminded about this update again.
This PR was generated by Mend Renovate. View the repository job log.