GHSA-v363-rrf2-5fmj Unsound use of str::from_utf8_unchecked on bytes which are not UTF-8
An attacker can cause undefined behavior, potentially leading to crashes or code execution, by providing invalid UTF-8 bytes to the ferris-says library.
- Exploited in the wild
- Not known
- Login needed
- Not needed
- Attackable
- Not stated
- Victim action needed
- Not needed
- Fix
- Available
What is GHSA-v363-rrf2-5fmj?
GHSA-v363-rrf2-5fmj is a memory corruption vulnerability in ferris-says, published on 13 Jan 2024.
Official description
Affected versions receive a &[u8] from the caller through a safe API, and pass
it directly to the unsafe str::from_utf8_unchecked function.
The behavior of ferris_says::say is undefined if the bytes from the caller
don't happen to be valid UTF-8.
The flaw was corrected in ferris-says#21 by using the safe str::from_utf8
instead, and returning an error on invalid input. However this fix has not yet
been published to crates.io as a patch version for 0.2.
Separately, ferris-says#32 has introduced a different API for version 0.3
which accepts input as &str rather than &[u8], so is unaffected by this bug.
Which versions are affected by GHSA-v363-rrf2-5fmj?
5 open source packages ship an affected version, including ferris-says.
| Package | Affected | Fixed in |
|---|---|---|
ferris-says crates.io | From 0.1.2 to 0.2.1 | None |
ferris-says crates.io | From 0.1.2 to 0.2.1 | None |
ferris-says crates.io | From 0.1.3-0 | 0.3.1 |
ferris-says crates.io | From 0.3.0 | 0.3.1 |
ferris-says crates.io | From 0.3.0 | 0.3.1 |
How do I fix GHSA-v363-rrf2-5fmj?
Upgrade ferris-says to 0.3.1 or later.
Upgrade to the fixed version. Memory-safe languages, compiler hardening and fuzzing prevent most new cases. More about Memory corruption.
Is GHSA-v363-rrf2-5fmj being exploited?
There is no known exploitation of GHSA-v363-rrf2-5fmj in the wild and no public exploit code.
Sources
Also known as RUSTSEC-2024-0001.