Affected versions of this crate did not require the buffer wrapped in BodyStream to be pinned, but treated it as if it had a fixed location in memory. This may result in a use-after-free.

The flaw was corrected by making the trait MessageBody require Unpin and making poll_next() function accept Pin<&mut Self> instead of &mut self.

Affected versions of this crate did not require the buffer wrapped in Framed to be pinned, but treated it as if it had a fixed location in memory. This may result in a use-after-free.

The flaw was corrected by making the affected functions accept Pin<&mut Self> instead of &mut self.

Impact

The affected functions set environment variables without synchronization. On Unix-like operating systems, this can crash in multithreaded programs. Programs may segfault due to dereferencing a dangling pointer if an environment variable is read in a different thread than the affected functions. This may occur without the user's knowledge, notably in the Rust standard library or third-party libraries.

The affected functions from time 0.2.7 through 0.2.22 are:

• time::UtcOffset::local_offset_at
• time::UtcOffset::try_local_offset_at
• time::UtcOffset::current_local_offset
• time::UtcOffset::try_current_local_offset
• time::OffsetDateTime::now_local
• time::OffsetDateTime::try_now_local

The affected functions in time 0.1 (all versions) are:

• time::at_utc
• time::at
• time::now
• time::tzset

Non-Unix targets (including Windows and wasm) are unaffected.

Patches

Pending a proper fix, the internal method that determines the local offset has been modified to always return None on the affected operating systems. This has the effect of returning an Err on the try_* methods and UTC on the non-try_* methods.

Users and library authors with time in their dependency tree should perform cargo update, which will pull in the updated, unaffected code.

Users of time 0.1 do not have a patch and should upgrade to an unaffected version: time 0.2.23 or greater or the 0.3 series.

Workarounds

A possible workaround for crates affected through the transitive dependency in chrono, is to avoid using the default oldtime feature dependency of the chrono crate by disabling its default-features and manually specifying the required features instead.

Examples:

Cargo.toml:

chrono = { version = "0.4", default-features = false, features = ["serde"] }

chrono = { version = "0.4.22", default-features = false, features = ["clock"] }

Commandline:

cargo add chrono --no-default-features -F clock

Sources:

• chronotope/chrono#602 (comment)
• vityafx/serde-aux#21

Affected versions of this crate did not properly detect invalid requests that could allow HTTP/1 request smuggling (HRS) attacks when running alongside a vulnerable front-end proxy server. This can result in leaked internal and/or user data, including credentials, when the front-end proxy is also vulnerable.

Popular front-end proxies and load balancers already mitigate HRS attacks so it is recommended that they are also kept up to date; check your specific set up. You should upgrade even if the front-end proxy receives exclusively HTTP/2 traffic and connects to the back-end using HTTP/1; several downgrade attacks are known that can also expose HRS vulnerabilities.

The HTTP/1 chunked decoder accepts a bare line feed after the final chunk as the start of a trailer instead of rejecting the malformed request.

When an upstream proxy reuses a connection and either pipelines the requests or does not take steps to validate the chunked stream, the decoder can consume the next request's headers as trailer data. It can then parse that request's body as a new request. An attacker can use this difference in request parsing to bypass access controls enforced by a proxy or web application firewall.

Upgrade to actix-http 3.20.2 or later.

Body extractors do not pass their size limits to the decompression layer. A small compressed request can therefore cause a large allocation before the extractor rejects the decoded body. A remote attacker can use this to exhaust memory or CPU resources and cause denial of service.

This affects applications that read compressed request bodies with the corresponding compression feature enabled.

Upgrade to actix-web 4.18.2 or later with actix-http 3.20.1 or later. As a workaround, reject compressed request bodies at a reverse proxy or in middleware before body extraction.

The encoding::Decoder stream can expand compressed input into a large output buffer before a caller can enforce its body size limit. An attacker who controls a compressed request or response can cause excessive memory allocation and CPU use, resulting in denial of service.

This affects applications that decode untrusted gzip, deflate, Brotli, or Zstandard bodies with the corresponding compression feature enabled.

Upgrade to actix-http 3.20.1 or later and configure Decoder::limit() when using the decoder directly. Actix Web users must also upgrade to actix-web 4.18.2 or later so that extractors pass their limits to the decoder.

With the compress-brotli feature enabled, the decoder accepts the non-standard large-window Brotli format. A crafted stream can cause its internal buffer to grow to about 1 GiB before the decoded output reaches the configured size limit. An 850-byte request can cause this allocation, and concurrent requests can exhaust process or host memory.

Actix Web enables Brotli support by default. Applications that read request bodies through Bytes, String, Json, or Form can be affected even when extractor size limits are configured. The allocation occurs before the handler runs, so authentication performed in the handler does not prevent it.

Upgrade to actix-http 3.20.3 or later. As a workaround, reject requests with Content-Encoding: br at a reverse proxy or in middleware before body extraction, or disable the compress-brotli feature throughout the dependency graph.