summaryrefslogtreecommitdiff
path: root/crates/tor-circmgr/src/mgr.rs
Commit message (Collapse)AuthorAgeFilesLines
* Run maint/add_warning.Nick Mathewson2024-03-131-0/+1
|
* circmgr: Remove now-obsolete error variants.Nick Mathewson2024-03-121-5/+4
|
* tor-circmgr: Add UniqId to CircTimeout.Gabriela Moldovan2024-02-271-1/+1
| | | | Part of #1297
* Teach *CircUsage structs about stability.Nick Mathewson2024-02-201-0/+6
| | | | | | | | | | | This lets us check long_lived_ports only when we are constructing a TargetCircUsage, and lets us remember whether or not the circuit we've built is suitable for non-long-lived ports. It also lets us decide whether existing circuits are compatible with new requests. Closes #1100.
* tor-circmgr: partially port mgr tests to MockRuntimeSaksham Mittal2023-11-141-11/+11
| | | | | | MockSleepRuntime wraps a MockRuntime instead of using test_with_one_runtime!(), so the tests should be more deterministic
* AbstractCircMgr::get_or_launch: inline wait_for_circJim Newsome2023-10-251-2/+1
|
* AbstractCircMgr::get_or_launch: document MSRV where div_ceil was stabilizedJim Newsome2023-10-251-1/+2
|
* oneshot: Apply deferred rustfmt churnIan Jackson2023-10-111-1/+1
| | | | cargo fmt, precisely.
* oneshot: Use veneer in tor-circmgrIan Jackson2023-10-111-1/+2
|
* arti-client: add exit selection with GeoIP country codeseta2023-09-181-0/+4
| | | | | | | | | | | | | | | | | | | | This threads the country codes work through the rest of the codebase: - `tor-dirmgr` will now enable GeoIP with the embedded database when the `geoip` future is enabled - This can be extended later using the `DirMgrConfig` to allow specifying a custom database; this is not done here, though - `tor-circmgr`'s `SupportedCircUsage` and `TargetCircUsage` fields gain new `country_code` members to allow filtering circuits by country - These are `()` in builds where the `geoip` feature is not enabled -- doing it this way means we don't have to copy and paste huge swathes of code, since we can't use `#[cfg]` in patterns - `ExitPathBuilder` gains (hacked-in) support for choosing a relay with the correct country code - Due to the lack of conjuction, we just copy and paste a small bit, pending further refactoring - `StreamPrefs` now lets you specify a country code, letting embedders make use of the feature
* Run maint/add_warning to add lint block everywhereIan Jackson2023-08-231-0/+1
|
* Merge branch 'clippy-allow' into 'main'Ian Jackson2023-07-111-0/+1
|\ | | | | | | | | clippy: Allow some of our existing code patterns See merge request tpo/core/arti!1396
| * Run maint/add_warning to actually apply new lint allowsIan Jackson2023-07-101-0/+1
| |
* | rng ranges: Use gen_range_checked().expect() in obvious casesIan Jackson2023-07-101-2/+3
|/ | | | In each of these, it is locally obvious that the range is nonempty.
* Merge branch 'inclusive' into 'main'Nick Mathewson2023-07-071-1/+1
|\ | | | | | | | | rng ranges: Use inclusive Duration ranges in several places See merge request tpo/core/arti!1385
| * rng ranges: Use inclusive Duration ranges in several placesIan Jackson2023-07-071-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | Many of these call sites would panic if, somehow, the upper bound was zero. In most cases it is very complicated to see if whether this could happen. However, there is a better answer: Durations are (conceptually) dense, so picking the closed set (which includes its boundary) rather than the open one (which doesn't) will make little practical difference. So change four call sites to use `..=` instead of just `..`.
* | Merge branch 'report-bugs-v2' into 'main'Nick Mathewson2023-07-071-12/+12
|\ \ | |/ |/| | | | | Optional tracing support in tor-error for error reporting See merge request tpo/core/arti!1379
| * circmgr: Use event_report!() and similar macrosNick Mathewson2023-07-071-12/+12
| | | | | | | | This simplifies our logging a little, and implements part of
* | Avoid matches!(x, None), matches!(x, Ok(_))Nick Mathewson2023-07-061-1/+1
|/ | | | This appeases clippy-nightly.
* Merge branch 'info-to-warn' into 'main'gabi-2502023-05-171-5/+8
|\ | | | | | | | | | | | | Change log levels of messages from INFO to others Closes #854 See merge request tpo/core/arti!1172
| * Change log level to debug and warn for certain appropriate situationsSaksham Mittal2023-05-171-5/+8
| | | | | | | | | | | | This commit changes certain log messages to debug for recoverable errors and a warn if all such attempts fail, in order to not clutter up the info messages that end users get to see.
* | Refactor ClientCirc APIs to use Arc<ClientCirc>.Nick Mathewson2023-05-161-13/+13
|/ | | | | | | | | | | | Now ClientCirc is no longer `Clone`, and the things that need it to be `Clone` instead return and use an Arc<ClientCirc> We're doing this so that ClientCirc can participate in the RPC system, and so that its semantics are more obvious. Closes #846. Thanks to the type system, this was a much simpler refactoring than I had feared it would be.
* circmgr: Functions to launch an unmanaged circuit.Nick Mathewson2023-03-221-0/+18
| | | | We'll use this to implement the circuits used by onion circuits.
* Use ErrorReport - run rustfmtIan Jackson2023-01-301-2/+3
| | | | Split off for ease of review and possible rebase.
* Use ErrorReport for errors in warn! in tor-circmgrIan Jackson2023-01-301-1/+1
|
* Use ErrorReport for errors in info! in tor-circmgrIan Jackson2023-01-301-6/+6
|
* Allow clippy::unchecked_duration_subtraction in testsNick Mathewson2023-01-271-0/+1
| | | | | This panics on error, and we're fine with a panic on misbehavior in tests.
* test lint blocks: Add many many automaticallyIan Jackson2022-12-121-0/+8
| | | | | This is precisely the result of running the rune in maint/adhoc-add-lint-blocks.
* Even more comments explaining circuit cancellation.Nick Mathewson2022-11-221-3/+37
|
* Improve notes surrounding circuit cancellation.Nick Mathewson2022-11-221-5/+18
| | | | | | | It turns out that bug #656 is not a real bug, but it was easy to overlook the code that prevented it. Closes #656.
* Merge branch 'channel' into 'main'Ian Jackson2022-08-171-0/+7
|\ | | | | | | | | Channel padding See merge request tpo/core/arti!657
| * channel usage: Rename CU::UserTraffic from CU::ExitIan Jackson2022-08-171-1/+1
| | | | | | | | | | As suggested in https://gitlab.torproject.org/tpo/core/arti/-/merge_requests/657#note_2827246
| * Move ChannelUsage from tor_proto to tor_chanmgrIan Jackson2022-08-171-1/+1
| | | | | | | | | | | | | | | | | | | | | | Replace Channel::note_usage with Channel::engage_padding_activities, which unconditionally causes the channel to (start to) do netflow padding things. The condition now lives in chanmgr. Addresses https://gitlab.torproject.org/tpo/core/arti/-/merge_requests/657#note_2826094
| * Provide ChannelUsage and plumb it all the way downIan Jackson2022-08-161-0/+7
| | | | | | | | | | | | | | | | | | | | | | | | | | Channel padding depends on what the channel is being used for. We therefore need to let the channel code know this information. The implementation of the per-channel padding control logic will be in the new note_usage function, which for now is simply a stub. A future commit will introduce a `PaddingControlState` which lives in the channel frontend; consult the doc comment for that type to see why the plumbing through the channel manager terminates in the channel frontend.
* | Mark some circuit-building errors as "transient".Nick Mathewson2022-08-161-9/+35
|/ | | | | | | | | | | | A "transient" error is one that does not indicate a true failure, but rather an _expected_ need to retry. When we hit one of these, we do not count it against the total number of permitted failures. (We do impose a higher limit on "real failures plus transient failures", though, to prevent infinite loops in the event of a programming error. Closes #517.
* tor-netdir: Collapse by_id and by_relay_id into a single fn.Nick Mathewson2022-08-101-3/+4
| | | | | | There are some downstream changes required for this to work, but they are all just unit tests that could no longer infer the type of an Ed25519 key.
* fix runtime issues on iostrinity-1686a2022-07-301-4/+6
|
* circmgr: Use FilterCounts to explain why we couldn't build pathsNick Mathewson2022-07-121-1/+6
|
* tor-netdir: testnet: Make construct_netdir infallible (rustfmt)Ian Jackson2022-06-131-3/+1
| | | | | Run rustfmt. Separate commit to make review of the substantive commit easier.
* tor-netdir: testnet: Make construct_netdir infallibleIan Jackson2022-06-131-1/+0
| | | | | This is a *lot* of unwraps. The function takes no parameters and is used only for testing. It ought to be infallible.
* Resolve the new `derive_partial_eq_without_eq` lint.Nick Mathewson2022-05-231-3/+3
| | | | It's a little overzealous sometimes, but it's mostly to the good.
* squash! Bump every crate's edition to 2021.Nick Mathewson2022-04-251-1/+0
| | | | | Remove all `use` statements for `TryFrom` and `TryInto`. These are now redundant in Rust 2021.
* circmgr: Report CircProvenance from AbstractCircMgr.Nick Mathewson2022-04-121-21/+32
| | | | | | | This feature is similar to ChanProvenance from ChanMgr, except that we don't yet need to report it outside the crate. I'm going to use it to distinguish newly created circuits from existing circuits in the preemptive circuit builder.
* Distinguish UsageMismatch cases by whether a race is possibleNick Mathewson2022-04-041-0/+6
| | | | | | | This lets us say that the UsageMismatch cases in some parts of the code reflect a programming error (RetryTime::Never), whereas in other case it reflects another circuit request getting to the circuit first (RetryTime::Immediate).
* circmgr: Improve retry-and-or-delay logic.Nick Mathewson2022-04-041-26/+36
| | | | | | | | | Use the new RetryTime type and its associates to decide how long to wait (if at all) between attempts to build a circuit. Closes #421. Part of #329.
* circmgr: Improve reporting of error origins.Nick Mathewson2022-04-041-27/+70
| | | | | | | Previously we did not distinguish errors that came from pending circuits from errors that came from the circuits we were building. We also reported errors as coming from "Left" or "Right", instead of a more reasonable description.
* circmgr: Avoid a race condition in circuit usage restrictionNick Mathewson2022-04-041-5/+14
| | | | | | | | | | | | | | | | | | | | | | We were treating restrict_mut() failures as internal errors, and using internal errors to represent them. But in fact, these failures are entirely possible based on timing. Here's how it happens: * Two different circuit requests arrive at the same time, and both notice a pending circuit that they could use. * The pending circuit completes; both pending requests are notified. * The first request calls restrict_mut(), and restricts the request in such a way that the second couldn't use it. * The second request calls restrict_mut(), and gets a failure. Because of this issue, we treat these errors as transient failures and just wait for another circuit. Closes #427. (This is not a breaking API change, since `AbstractSpec` is a crate-private trait.)
* circmgr: limit circuit attempts when launch_parallelism > 1.Nick Mathewson2022-03-301-1/+21
| | | | | | | | | | | Previously, if we had launch_parallelism > 1, and we were willing to retry building a circuit max_retries times, then we'd launch up to max_retries * launch_parallelism circuits before giving up. Ouch! With this patch, we try to keep the total number of circuits planned and attempted to the actual max_retries limit. Part of #329; found with arti-testing.
* Turn FallbackList into a real type, and store one in GuardMgr.Nick Mathewson2022-03-301-2/+4
| | | | | | | | | | | | | | The guard manager is responsible for handing out the first hops of tor circuits, keeping track of their successes and failures, and remembering their states. Given that, it makes sense to store this information here. It is not yet used; I'll be fixing that in upcoming commits. Arguably, this information no longer belongs in the directory manager: I've added a todo about moving it. This commit will break compilation on its own in a couple of places; subsequent commits will fix it up.
* Move fallback.rs into guardmgr.Nick Mathewson2022-03-301-1/+1
| | | | | | | This is the logical place for it, I think: the GuardMgr's job is to pick the first hop for a circuit depending on remembered status for possible first hops. Making this change will let us streamline the code that interacts with these objects.
generated by cgit v1.2.3 (git 2.46.0) at 2026-08-26 21:36:17 +0000