summaryrefslogtreecommitdiff
path: root/crates/tor-chanmgr
Commit message (Collapse)AuthorAgeFilesLines
* Bump all the unstable tor- and arti- crates to 0.25.0.Gabriela Moldovan2024-12-021-19/+19
| | | | | | | | | | Done using: ``` for crate in $(./maint/list_crates | rg '^(tor|arti-)'); do cargo set-version -p $crate 0.25.0 done ```
* Bump the versions of the non-{arti-,tor-} crates.Gabriela Moldovan2024-12-021-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The non-{arti-,tor-} crates are: ``` ./maint/list_crates | rg -v '^(tor|arti)' oneshot-fused-workaround slotmap-careful test-temp-dir fslock-guard hashx equix caret fs-mistrust safelog retry-error ``` We split them in the following categories: * crates with no changes (no version bumps): ``` maint/changed_crates -v "arti-v$LAST_VERSION" 2>&1 >/dev/null | grep -i "no change" | grep -v '\(tor\|arti\)-' oneshot-fused-workaround: No change. test-temp-dir: No change. caret: No change. ``` * crates that only have non-functional changes (bump the patch version): - slotmap-careful - fslock-guard - hashx - equix - fs-mistrust - safelog - retry-error * crates where APIs were broken (bump minor): None The bumps from this commit were created using this script: ``` PATCH=" slotmap-careful fslock-guard hashx equix fs-mistrust safelog retry-error " for crate in $PATCH; do cargo set-version --bump patch -p $crate; done ```
* tor-chanmgr: remove immediately awaited async block expressionSteven Engler2024-11-271-23/+15
| | | | This shouldn't be needed anymore now that we use a `Defer`.
* tor-chanmgr: use `Defer` in `AbstractChanMgr` to handle cancellationsSteven Engler2024-11-271-18/+23
|
* tor-chanmgr: add `util::defer` moduleSteven Engler2024-11-273-0/+79
| | | | | This contains the `Defer` type, which can be used to defer a closure until the `Defer` is dropped.
* tor-chanmgr: change `RequestCancelled` retry from `Never` to `Immediate`Steven Engler2024-11-211-1/+1
| | | | | | | If the channel request was cancelled, we should be able to retry immediately. Additionally, the `kind()` of `RequestCancelled` is `TransientFailure`, and I don't think it make sense for a transient error to return a `retry_time()` of `Never`.
* tor-chanmgr: fix reconfigure bug when given `CheckAllOrNothing`Steven Engler2024-11-121-2/+9
| | | | We shouldn't actually reconfigure anything if given `CheckAllOrNothing`.
* cargo: Update `thiserror` to `2`Clara Engler2024-11-121-1/+1
|
* tor-chanmgr: remove panics in debug buildsSteven Engler2024-11-112-7/+1
| | | | | | These were supposed to fail loudly in debug builds by panicking, but panics are mostly useless for debugging in async applications that use a runtime which catches panics. So we'll just log the error instead.
* Merge branch 'versions' into 'main'Ian Jackson2024-10-301-20/+20
|\ | | | | | | | | Version bumps for release of 1.3.0 See merge request tpo/core/arti!2599
| * Bump minor versions of crates with new APIs/implsIan Jackson2024-10-301-1/+1
| | | | | | | | | | cargo set-version --bump=patch -p slotmap-careful cargo set-version --bump=patch -p safelog
| * Bump unstable tor-* and arti-* crates to 0.24.0Ian Jackson2024-10-301-19/+19
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | cargo set-version -p tor-async-utils 0.24.0 cargo set-version -p tor-basic-utils 0.24.0 cargo set-version -p tor-bytes 0.24.0 cargo set-version -p tor-cell 0.24.0 cargo set-version -p tor-cert 0.24.0 cargo set-version -p tor-chanmgr 0.24.0 cargo set-version -p tor-checkable 0.24.0 cargo set-version -p tor-circmgr 0.24.0 cargo set-version -p tor-config 0.24.0 cargo set-version -p tor-config-path 0.24.0 cargo set-version -p tor-consdiff 0.24.0 cargo set-version -p tor-dirclient 0.24.0 cargo set-version -p tor-dirmgr 0.24.0 cargo set-version -p tor-error 0.24.0 cargo set-version -p tor-general-addr 0.24.0 cargo set-version -p tor-geoip 0.24.0 cargo set-version -p tor-guardmgr 0.24.0 cargo set-version -p tor-hsclient 0.24.0 cargo set-version -p tor-hscrypto 0.24.0 cargo set-version -p tor-hsrproxy 0.24.0 cargo set-version -p tor-hsservice 0.24.0 cargo set-version -p tor-key-forge 0.24.0 cargo set-version -p tor-keymgr 0.24.0 cargo set-version -p tor-linkspec 0.24.0 cargo set-version -p tor-llcrypto 0.24.0 cargo set-version -p tor-log-ratelim 0.24.0 cargo set-version -p tor-memquota 0.24.0 cargo set-version -p tor-netdir 0.24.0 cargo set-version -p tor-netdoc 0.24.0 cargo set-version -p tor-persist 0.24.0 cargo set-version -p tor-proto 0.24.0 cargo set-version -p tor-protover 0.24.0 cargo set-version -p tor-ptmgr 0.24.0 cargo set-version -p tor-relay-crypto 0.24.0 cargo set-version -p tor-relay-selection 0.24.0 cargo set-version -p tor-rpcbase 0.24.0 cargo set-version -p tor-rtcompat 0.24.0 cargo set-version -p tor-rtmock 0.24.0 cargo set-version -p tor-socksproto 0.24.0 cargo set-version -p tor-units 0.24.0 cargo set-version -p arti-client 0.24.0 cargo set-version -p arti-relay 0.24.0 cargo set-version -p arti-rpc-client-core 0.24.0 cargo set-version -p arti-rpcserver 0.24.0
* | cargo: Run fixup-features (formatting)Ian Jackson2024-10-301-1/+2
| | | | | | | | Manual fixup. I filed #1719 about the fact that I had to do this.
* | cargo: Run fixup-featuresIan Jackson2024-10-301-1/+1
|/ | | | | | | This is output from nailing-cargo -Eu run -p fixup-features Cargo.toml There are some formatting glitches which I'll fix in a moment.
* tor-chanmgr: rename `replace_pending_channel` to ↵Steven Engler2024-10-242-9/+10
| | | | `upgrade_pending_channel_to_open`
* tor-chanmgr: `remove_pending` now takes a `PendingChannelHandle`Steven Engler2024-10-241-18/+8
|
* tor-chanmgr: improve cleanup procedure of pending channelsSteven Engler2024-10-242-47/+92
| | | | | | | | | | An attempt to make sure that there are no code paths which forget to remove a pending channel from the channel map. This also adds error-level log messages and panics during debug builds if a `PendingChannelHandle` is dropped without properly passing it to `MgrState::remove_pending_channel` or `MgrState::replace_pending_channel`.
* Revert "tor-chanmgr: `PendingChannelHandle` removes the channel when dropped"Steven Engler2024-10-212-56/+34
| | | | | | | | This reverts commit f85bc3cf849109aaa2c4da9fc8c06e7173a0543f. There were some small conflcits in `AbstractChanMgr::get_or_launch_internal`, so this wasn't a clean revert.
* Merge branch 'chanmgr-select-docs' into 'main'David Goulet2024-10-171-0/+25
|\ | | | | | | | | tor-chanmgr: update comments about selecting pending channels See merge request tpo/core/arti!2544
| * tor-chanmgr: update comments about selecting pending channelsSteven Engler2024-10-161-0/+25
| |
* | Merge branch 'orport' into 'main'David Goulet2024-10-172-224/+309
|\ \ | | | | | | | | | | | | tor-chanmgr: refactored to remove `with_channels` and added `PendingChannelHandle` See merge request tpo/core/arti!2538
| * | tor-chanmgr: remove `handle_build_outcome`Steven Engler2024-10-151-28/+19
| | |
| * | tor-chanmgr: `PendingChannelHandle` removes the channel when droppedSteven Engler2024-10-152-27/+56
| | |
| * | tor-chanmgr: added `PendingChannelHandle`Steven Engler2024-10-152-36/+64
| | | | | | | | | | | | | | | This handle contains all of the details required to remove or replace a pending channel entry from the channel map.
| * | tor-chanmgr: refactor so we don't need `with_channels{,_and_params}`Steven Engler2024-10-152-78/+76
| | | | | | | | | | | | | | | | | | These methods on `MgrState` acquire a lock, and it's easy for calling code to also try to acquire the same lock within the closure, causing a deadlock. It's better to not expose these methods.
| * | tor-chanmgr: fix some incorrect commentsSteven Engler2024-10-152-5/+3
| | |
| * | tor-chanmgr: removed use of `with_channels` within `request_channel`Steven Engler2024-10-151-61/+66
| | | | | | | | | | | | Best reviewed with the '-w' flag to hide whitespace changes.
| * | tor-chanmgr: refactored `AbstractChanMgr::choose_action`Steven Engler2024-10-152-109/+145
| |/ | | | | | | | | | | | | | | This moves most of the channel map logic from `AbstractChanMgr::choose_action` to `MgrState::request_channel`. This is working towards being able to remove `MgrState::with_channels{,_and_params}`.
* / Apply `default-features = false` to all tor-memquota depsIan Jackson2024-10-161-1/+1
|/ | | | | | | This will allow us to make `tor-memquota/memquota` enabled by default, without forcing it to be compiled in everywhere. I don't change that in this MR, though.
* memquota: Change ToplevelAccount to be an alias for Arc<MemoryQuotaTracker>Ian Jackson2024-10-152-1/+3
| | | | Fixes a TODO.
* tor-chanmgr: fix `unused_imports` warning on `ChannelAccount`Steven Engler2024-10-091-1/+3
|
* tor-chanmgr: note API causes deadlocks under some conditionsSteven Engler2024-10-092-3/+16
|
* tor-chanmgr: minor code cleanupSteven Engler2024-10-091-5/+5
|
* tor-chanmgr: remove unused code pathSteven Engler2024-10-091-11/+3
|
* tor-proto: Make circuit->channel queues participate in memquotaIan Jackson2024-10-031-1/+12
| | | | | | | | We use the *channel*'s memquota account. This is arguably wrong, but it's hard to get right now. See #1652. Change the type of the queue, and the places it's constructed. The use sites can all stay the same.
* tor-proto: Plumb the ChannelAccount through to queue creation siteIan Jackson2024-10-034-10/+20
| | | | | This gets it as far as the outbound circuit->channel mpsc queue creation. Also, we provide an accessor for it.
* tor-proto: Plumb the ChannelAccount through to queue creation site (pre-fmt)Ian Jackson2024-10-031-1/+3
|
* tor-chanmgr: Make a memquota::ChannelAcocunt per channelIan Jackson2024-10-035-6/+20
| | | | | This delivers a fresh account per channel to the places where channels are actually made, but doesn't pass them to tor-proto yet.
* tor-chanmgr: Make a memquota::ChannelAcocunt per channel (pre-fmt)Ian Jackson2024-10-031-1/+4
|
* memquota: Add a toplevel account in tor-chanmgrIan Jackson2024-10-033-1/+17
| | | | | | | | | Plumb through a top-level account. This doesn't have any channel-specific, circuit-specific or stream-specific accounts yet. tor-circmgr's and tor-hsclient's *tests* need fake account. In arti-relay, use a dummy account for now.
* memquota: Add a toplevel account in tor-chanmgr (pre-fmt)Ian Jackson2024-10-031-1/+2
|
* tor-chanmgr: Provide a variant for memquota errorsIan Jackson2024-10-032-0/+9
|
* tor-chanmgr: minor doc improvementsSteven Engler2024-10-011-3/+3
|
* tor-chanmgr: add unit testsSteven Engler2024-10-011-0/+416
|
* tor-chanmgr: change `choose_best_channel` to take `IntoIterator`Steven Engler2024-10-011-1/+3
|
* tor-chanmgr: don't allow pending channels with no relay idsSteven Engler2024-10-011-1/+7
|
* tor-chanmgr: move channel selection logic to a new moduleSteven Engler2024-10-012-162/+172
|
* tor-chanmgr: improve readability of `choose_best_channel`Steven Engler2024-10-011-13/+45
|
* tor-socksproto: Handle 0-byte reads (EOF) correctlyIan Jackson2024-10-011-1/+1
| | | | | | | This restores the functionality of socks users: detect closed sockets. 0c595818f713916d94b7b0e4062f953fad7c9799 which we reverted as part of rebasing this branch onto main.
* tor-chanmgr: Use new tor-socksproto API (fmt)Ian Jackson2024-10-011-2/+1
|
generated by cgit v1.2.3 (git 2.46.0) at 2026-08-26 22:04:58 +0000