ghsa

**Summary** The absence of restrictions on the endpoint, which allows you to create a path for uploading a file in a file distribution, allows an attacker to add arbitrary files to the distribution. **Details** Vulnerable endpoint: POST /files **PoC** 1. Create a file distribution. <img width="1434" alt="Снимок экрана 2024-03-17 в 21 27 30" src="https://github.com/psi-4ward/psitransfer/assets/163760990/4634a6f7-6e7d-486e-9929-76156aaa1340"> 2. Go to the link address (id of the file distribution is needed by an attacker to upload files there). <img width="1426" alt="Снимок экрана 2024-03-17 в 21 27 35" src="https://github.com/psi-4ward/psitransfer/assets/163760990/a57c910c-69e2-4b07-985d-b0a46c69891a"> 3. Send a POST /files. As the value of the Upload-Metadata header we specify the sid parameter with the id of the file distribution obtained in the second step. In the response from the server in the Location header we get the path for uploading a new file to the file distribution. <i...

An integer overflow in dav1d AV1 decoder that can occur when decoding videos with large frame size. This can lead to memory corruption within the AV1 decoder. We recommend upgrading to version 0.7.0 of libdav1d-sys, which includes dav1d 1.4.0.

Given the function `transpose::transpose`: ```rust fn transpose<T: Copy>(input: &[T], output: &mut [T], input_width: usize, input_height: usize) ``` The safety check `input_width * input_height == output.len()` can fail due to `input_width * input_height` overflowing in such a way that it equals `output.len()`. As a result of failing the safety check, memory past the end of `output` is written to. This only occurs in release mode since `*` panics on overflow in debug mode. Exploiting this issue requires the caller to pass `input_width` and `input_height` arguments such that multiplying them overflows, and the overflown result equals the lengths of input and output slices.

As of version 0.6.0, the ObjectPool explicitly creates an uninitialized instance of its type parameter when it attempts to free an object, and swaps it into the storage. This causes instant undefined behavior due to reading the uninitialized memory in order to write it to the pool storage. Extremely basic usage of the crate can trigger this issue, e.g. this code from a doctest: ```rust use crayon::prelude::*; application::oneshot().unwrap(); let mut params = MeshParams::default(); let mesh = video::create_mesh(params, None).unwrap(); // Deletes the mesh object. video::delete_mesh(mesh); // <-- UB ``` The Clippy warning for this code was silenced in commit c2fde19caf6149d91faa504263f0bc5cafc35de5. Discovered via https://asan.saethlin.dev/ub?crate=crayon&version=0.7.1

With versions of the whoami crate >= 0.5.3 and < 1.5.0, calling any of these functions leads to an immediate stack buffer overflow on illumos and Solaris: - `whoami::username` - `whoami::realname` - `whoami::username_os` - `whoami::realname_os` With versions of the whoami crate >= 0.5.3 and < 1.0.1, calling any of the above functions also leads to a stack buffer overflow on these platforms: - Bitrig - DragonFlyBSD - FreeBSD - NetBSD - OpenBSD This occurs because of an incorrect definition of the `passwd` struct on those platforms. As a result of this issue, denial of service and data corruption have both been observed in the wild. The issue is possibly exploitable as well. This vulnerability also affects other Unix platforms that aren't Linux or macOS. This issue has been addressed in whoami 1.5.0. For more information, see [this GitHub issue](https://github.com/ardaku/whoami/issues/91).

In affected versions, after a `Report` is constructed using `wrap_err` or `wrap_err_with` to attach a message of type `D` onto an error of type `E`, then using `downcast` to recover ownership of either the value of type `D` or the value of type `E`, one of two things can go wrong: - If downcasting to `E`, there remains a value of type `D` to be dropped. It is incorrectly "dropped" by running `E`'s drop behavior, rather than `D`'s. For example if `D` is `&str` and `E` is `std::io::Error`, there would be a call of `std::io::Error::drop` in which the reference received by the `Drop` impl does not refer to a valid value of type `std::io::Error`, but instead to `&str`. - If downcasting to `D`, there remains a value of type `E` to be dropped. When `D` and `E` do not happen to be the same size, `E`'s drop behavior is incorrectly executed in the wrong location. The reference received by the `Drop` impl may point left or right of the real `E` value that is meant to be getting dropped. In bot...

Due to insufficient checking of input data, decoding certain data sequences can lead to _Decoder::decode_ panicking rather than returning an error. Example code that triggers this vulnerability looks like this: ```rust use hpack::Decoder; pub fn main() { let input = &[0x3f]; let mut decoder = Decoder::new(); let _ = decoder.decode(input); } ``` hpack is unmaintained. A crate with the panics fixed has been published as [hpack-patched](https://crates.io/crates/hpack-patched). Also consider using [fluke-hpack](https://crates.io/crates/fluke-hpack) or [httlib-huffman](https://crates.io/crates/httlib-huffman) as an alternative.

An attacker can send a flood of CONTINUATION frames, causing `h2` to process them indefinitely. This results in an increase in CPU usage. Tokio task budget helps prevent this from a complete denial-of-service, as the server can still respond to legitimate requests, albeit with increased latency. More details at https://seanmonstar.com/blog/hyper-http2-continuation-flood/. Patches available for 0.4.x and 0.3.x versions.

### Impact Note: "Pebble" here refers to [Canonical's service manager](https://github.com/canonical/pebble), not the [Let's Encrypt ACME test server](https://github.com/letsencrypt/pebble). The API behind `pebble pull`, used to read files from the workload container by Juju charms, allows access from any user, instead of just admin. In Juju Kubernetes sidecar charms, Pebble and the charm run as root, so they have full access. But if another restricted unix user gains local access to the container host, they could hit the Pebble `GET /v1/files?action=read` API and would be allowed to read any file in the workload container, for example an ssh key or database password or other sensitive information. If there are ssh keys they could then potentially ssh into the workload, or if something like a database password they could log into the database. Note that this requires local user access to the host machine. It seems unlikely that an attacker could gain this level of access in a Juju Ku...

### Impact Code that attempts to use an item (e.g., a row) returned by an iterator after the iterator has advanced to the next item will be accessing freed memory and experience undefined behaviour. Code that uses the item and then advances the iterator is unaffected. This problem has always existed. This is a use-after-free bug, so it's rated high severity. If your code uses a pre-3.0.0 version of cassandra-rs, and uses an item returned by a cassandra-rs iterator after calling `next()` on that iterator, then it is vulnerable. However, such code will almost always fail immediately - so we believe it is unlikely that any code using this pattern would have reached production. For peace of mind, we recommend you upgrade anyway. ### Patches The problem has been fixed in version 3.0.0. Users should upgrade to ensure their code cannot use the problematic pattern. ### Workarounds Ensure all usage fits the expected pattern. For example, use `get_first_row()` rather than an iterator, or comp...