dandri/matrix-authentication-service
1
0
Fork 0
mirror of https://github.com/element-hq/matrix-authentication-service.git synced 2026-05-24 19:35:28 +00:00
Code Issues Packages Projects Releases Wiki Activity

Fix lints

Browse Source
This commit is contained in:
Eric Eastwood
2026-04-27 17:46:28 -05:00
parent b2d7ef9583
commit caf3d97f5e
1 changed files with 23 additions and 20 deletions
Download Patch File Download Diff File Expand all files Collapse all files
crates/handlers/src/compat/login.rs
+23 -20
View File
@@ -582,7 +582,8 @@ async fn process_violations_for_compat_login(
username = user.username,
// So we can easily look it up in the MAS database
compat_session_id = compat_session.id.to_string(),
// Make it easier to line up with what the user may be talking about
// Make it easier to line up with what the user may be talking
// about
device_id = compat_session
.device
.as_ref()
@@ -624,24 +625,25 @@ async fn process_violations_for_compat_login(
}
/// We fetch a minimum number of sessions (2160, more than we need in normal
/// cases) so we can sort by `last_active_at` after it gets back from the database
/// and can get even closer to removing the true oldest sessions.
/// cases) so we can sort by `last_active_at` after it gets back from the
/// database and can get even closer to removing the true oldest sessions.
///
/// The 2160 number was chosen based on someone having a script that runs every hour
/// for the the 90-day `INACTIVE_SESSION_THRESHOLD`. Additionally, it also aligns
/// nicely with < 0.001% of people on matrix.org having less than 2160 sessions and
/// reasoning how much memory is reasonable to spend on this operation to get things
/// right. Assuming each row is ~1 KiB (pessimistic high bound, see next paragraph
/// below) we end up at ~2 MiB of memory.
/// The 2160 number was chosen based on someone having a script that runs every
/// hour for the the 90-day `INACTIVE_SESSION_THRESHOLD`. Additionally, it also
/// aligns nicely with < 0.001% of people on matrix.org having less than 2160
/// sessions and reasoning how much memory is reasonable to spend on this
/// operation to get things right. Assuming each row is ~1 KiB (pessimistic high
/// bound, see next paragraph below) we end up at ~2 MiB of memory.
///
/// Each item in the page is `(CompatSession, Option<CompatSsoLogin>)` where
/// `CompatSession` is 192 bytes plus a couple of strings (device name and user
/// agent) (assume pessimistic 512 total bytes). And `CompatSsoLogin` which is also
/// 192 bytes with a `login_token` string which should be no more than 32 bytes.
/// agent) (assume pessimistic 512 total bytes). And `CompatSsoLogin` which is
/// also 192 bytes with a `login_token` string which should be no more than 32
/// bytes.
const MINIMUM_SESSIONS_TO_FETCH: usize = {
let min_sessions = INACTIVE_SESSION_THRESHOLD.num_days() * 24;
// Ideally, we'd use `usize::try_from(min_sessions)` but that doesn't work in const
// contexts.
// Ideally, we'd use `usize::try_from(min_sessions)` but that doesn't work in
// const contexts.
#[allow(clippy::cast_sign_loss, clippy::cast_possible_truncation)]
{
// Sanity check that `clippy::cast_sign_loss` doesn't apply
@@ -657,9 +659,9 @@ const MINIMUM_SESSIONS_TO_FETCH: usize = {
min_sessions as usize
}
};
// This is a stop-gap to make people think about the downstream effects of updating
// `INACTIVE_SESSION_THRESHOLD` or whatever contributing factors go into
// `MINIMUM_SESSIONS_TO_FETCH`.
// This is a stop-gap to make people think about the downstream effects of
// updating `INACTIVE_SESSION_THRESHOLD` or whatever contributing factors go
// into `MINIMUM_SESSIONS_TO_FETCH`.
const _: () = {
assert!(
// Update this value if you're ok with the ammount of memory that could be used.
@@ -671,10 +673,11 @@ const _: () = {
/// Find the least recently used (LRU) compat sessions
///
/// The results of this function are flawed (for accounts with more sessions than
/// `minimum_sessions_to_fetch`) because we can't order by `last_active_at` and get an
/// absolute sort of actually least recently used sessions. But we do a pretty good job
/// at working around the problem (see internal comments for details).
/// The results of this function are flawed (for accounts with more sessions
/// than `minimum_sessions_to_fetch`) because we can't order by `last_active_at`
/// and get an absolute sort of actually least recently used sessions. But we do
/// a pretty good job at working around the problem (see internal comments for
/// details).
async fn find_lru_compat_sessions_flawed(
clock: &dyn Clock,
repo: &mut BoxRepository,