# GHSA-53rv-hcvm-rpp9: Lodestar snappy decompression issue

> A flaw in Lodestar's snappy decompression allows remote attackers to cause a permanent chain split, requiring a hard fork to resolve.

- **Severity:** Not scored
- **Exploited in the wild:** Not known
- **Login needed:** Not needed
- **Attackable:** Over the internet
- **Victim action needed:** Not needed
- **Fix available:** Yes
- **Published:** 14 Jan 2025

## What is GHSA-53rv-hcvm-rpp9?

GHSA-53rv-hcvm-rpp9 is a denial of service vulnerability in @lodestar/reqresp, published on 14 Jan 2025. An attacker can exploit it over the internet without logging in, with no action from the victim.

### Official description

### Impact
Unintended permanent chain split affecting greater than or equal to 25% of the network, requiring hard fork (network partition requiring hard fork)

### Description
Lodestar client may fail to decode snappy framing compressed messages.

### Vulnerability Details
In Req/Resp protocol the message are encoded by using ssz_snappy encoding, which is basically snappy framing compression over ssz encoded message.

It's mentioned here - https://github.com/ethereum/consensus-specs/blob/dev/specs/phase0/p2p-interface.md

```
The token of the negotiated protocol ID specifies the type of encoding to be used for the req/resp interaction. Only one value is possible at this time:

ssz_snappy: The contents are first SSZ-encoded and then compressed with Snappy frames compression. For objects containing a single field, only the field is SSZ-encoded not a container with a single field. For example, the BeaconBlocksByRoot request is an SSZ-encoded list of Root's. This encoding type MUST be supported by all clients.
```

In snappy framing format there a few types of chunks.
We are interested in so called reserved skippable chunks. These are chunks with chunk type in range [0x80, 0xfd]
Let's see how rust snappy handles them https://github.com/BurntSushi/rust-snappy/blob/master/src/read.rs#L137

```
impl<R: io::Read> io::Read for FrameDecoder<R> {
    fn read(&mut self, buf: &mut [u8]) -> io::Result<usize> {
 		   ... 
           ...
  		    let len = len64 as usize;
            match ty {
                Err(b) if 0x02 <= b && b <= 0x7F => {
                    // Spec says that chunk types 0x02-0x7F are reserved and
                    // conformant decoders must return an error.
                    fail!(Error::UnsupportedChunkType { byte: b });
                }
                Err(b) if 0x80 <= b && b <= 0xFD => {
                    // Spec says that chunk types 0x80-0xFD are reserved but
                    // skippable.
                    self.r.read_exact(&mut self.src[0..len])?;
                }
```

Similar code can be found in golang implementation - https://github.com/golang/snappy/blob/master/decode.go#L221

```
func (r *Reader) fill() error {
	...
	if chunkType <= 0x7f {
			// Section 4.5. Reserved unskippable chunks (chunk types 0x02-0x7f).
			r.err = ErrUnsupported
			return r.err
		}
		// Section 4.4 Padding (chunk type 0xfe).
		// Section 4.6. Reserved skippable chunks (chunk types 0x80-0xfd).
		if !r.readFull(r.buf[:chunkLen], false) {
			return r.err
		}
```

Now let's see how lodestar handles such chunks https://github.com/ChainSafe/lodestar/blob/unstable/packages/reqresp/src/encodingStrategies/sszSnappy/snappyFrames/uncompress.ts#L17

```
uncompress(chunk: Uint8ArrayList): Uint8ArrayList | null {
    this.buffer.append(chunk);
    const result = new Uint8ArrayList();
    while (this.buffer.length > 0) {
      if (this.buffer.length < 4) break;

      const type = getChunkType(this.buffer.get(0));
      const frameSize = getFrameSize(this.buffer, 1);

      if (this.buffer.length - 4 < frameSize) {
        break;
      }

      const data = this.buffer.subarray(4, 4 + frameSize);
      this.buffer.consume(4 + frameSize);

      if (!this.state.foundIdentifier && type !== ChunkType.IDENTIFIER) {
        throw "malformed input: must begin with an identifier";
      }

      if (type === ChunkType.IDENTIFIER) {
        if (!Buffer.prototype.equals.call(data, IDENTIFIER)) {
          throw "malformed input: bad identifier";
        }
        this.state.foundIdentifier = true;
        continue;
      }

      if (type === ChunkType.COMPRESSED) {
        result.append(uncompress(data.subarray(4)));
      }
      if (type === ChunkType.UNCOMPRESSED) {
        result.append(data.subarray(4));
      }
    }
    if (result.length === 0) {
      return null;
    }
    return result;
  }

 function getChunkType(value: number): ChunkType {
  switch (value) {
    case ChunkType.IDENTIFIER:
      return ChunkType.IDENTIFIER;
    case ChunkType.COMPRESSED:
      return ChunkType.COMPRESSED;
    case ChunkType.UNCOMPRESSED:
      return ChunkType.UNCOMPRESSED;
    case ChunkType.PADDING:
      return ChunkType.PADDING;
    default:
      throw new Error("Unsupported snappy chunk type");
  }
```

As you can see, lodestar does not recognize such chunks.

If it sees such chunk, function getChunkType() throws an exception and decoding fails.

### Impact Details

Faulty nodes may trigger chain stall by sending messages which lodestar fails to parse, while other clients will be able to handle.

### Proof of Concept

How to reproduce:

1. get archive (via provided [gist link](https://gist.github.com/gln7/bdde7f4e0bdf9d47bf810a015796867a)), decode and unpack it:
```
$ base64 -d poc.txt > poc.tgz
$ tar zxf poc.tgz
```

2. run dec1.go to verify that our snappy file decompressed successfully
```
$ go run dec1.go

reading 1.snappy...
read 124 bytes, err <nil>
```

3. run dec1.mjs to verify that lodestar fails to decode such file
```
checking chunk type=255
checking chunk type=1
got uncompressed chunk..
checking chunk type=129
file:///../poc/dec1.mjs:74
            throw new Error("Unsupported snappy chunk type");
```

## Which versions are affected by GHSA-53rv-hcvm-rpp9?

2 open source packages ship an affected version, including @lodestar/reqresp.


| Package | Ecosystem | Affected | Fixed in |
| --- | --- | --- | --- |
| `@lodestar/reqresp` | npm | All versions | 1.25.0 |
| `@lodestar/reqresp` | npm | All versions | 1.25.0 |

## How do I fix GHSA-53rv-hcvm-rpp9?

Upgrade @lodestar/reqresp to 1.25.0 or later.

## Is GHSA-53rv-hcvm-rpp9 being exploited?

There is no known exploitation of GHSA-53rv-hcvm-rpp9 in the wild and no public exploit code.

## Fix prompt for AI coding agents

```text
Check this codebase for GHSA-53rv-hcvm-rpp9 and fix it if it is affected.

What it is: GHSA-53rv-hcvm-rpp9 is a denial of service vulnerability in @lodestar/reqresp, published on 14 Jan 2025. An attacker can exploit it over the internet without logging in, with no action from the victim.
Exploitation: There is no known exploitation of GHSA-53rv-hcvm-rpp9 in the wild and no public exploit code.

Affected packages and fixed versions:
- @lodestar/reqresp (npm): affected all versions; fixed in 1.25.0
- @lodestar/reqresp (npm): affected all versions; fixed in 1.25.0

Task:
1. Search the dependency manifests and lockfiles (for example package.json, package-lock.json, requirements.txt, poetry.lock, uv.lock, pom.xml, build.gradle, go.mod, Cargo.lock, Gemfile.lock) and any Dockerfiles for the affected packages, including transitive dependencies.
2. If an affected version is in use, upgrade it to the fixed version for its release line with the project's package manager, so the lockfile is regenerated. Prefer the smallest upgrade that fixes it.
3. If there is no fix, or the upgrade would break the build, do not guess: explain the risk and the safest workaround instead.
4. Run the tests and summarise what changed and why.
5. If nothing in the project is affected, say so and change nothing.

Reference: https://insights.sbomify.com/GHSA-53rv-hcvm-rpp9/ (Markdown: https://insights.sbomify.com/GHSA-53rv-hcvm-rpp9/index.md)
```

## Sources

- OSV: https://osv.dev/vulnerability/GHSA-53rv-hcvm-rpp9

---
From sbomify Insights: https://insights.sbomify.com/GHSA-53rv-hcvm-rpp9/
Summaries are generated and may contain mistakes; the official description above is the record.