mirrors/arti.git/crates/tor-proto/src/circuit/reactor.rs, branch derive-traitsmirror of https://gitlab.torproject.org/tpo/core/arti
http://git.dilluti0n.com/mirrors/arti.git/atom?h=derive-traits2022-02-15T14:57:03ZUpdate tor-proto errors to latest API.2022-02-15T14:57:03ZNick Mathewsonnickm@torproject.org2022-02-15T14:57:03Zurn:sha1:6b611d6deae2e003ea81f459cd5e005c37ab50cetor-proto: use InternalError for internal errors.2022-02-15T14:41:11ZNick Mathewsonnickm@torproject.org2022-02-11T20:20:50Zurn:sha1:f23f375e42ef3237f9bbb7b8bfee47ecc9ced4a1Merge branch 'eta/reactor-2.5' into 'main'2022-02-03T16:27:57Zetator@eta.st2022-02-03T16:27:57Zurn:sha1:e9087e1fed7193f0cf5b9f9d1078340f216745ed
Fix severe reactor ordering problems
See merge request tpo/core/arti!282Fix severe reactor ordering problems2022-02-03T15:04:28Zetator@eta.st2022-02-02T14:21:17Zurn:sha1:24d12163f2a5b109adc1d40ad815efecf36cf499
A number of severe problems with the circuit reactor were fixed which
could cause reordering of cells (which causes relays to terminate the
circuit with a protocol violation, as they become unable to decrypt
them). These mostly revolve around improper usage of queues:
- The code assumed that a failure to place cells onto the channel would
persist for the duration of a reactor cycle run. However, under high
contention, this wouldn't always be the case.
- This leads to some cells getting enqueued while others go straight
through, before the enqueued cells.
- To fix this, we block sending cells out of the channel while there
are still some enqueued.
- The hop-specific queues queued after encryption, not before. This was
very brittle, and led to frequent mis-ordering.
- This was fixed by making them not do that.
This is arti!264 / 5bce9db5628126be2b736f228211174fe4132918 without the
refactor part.
Remove many needless borrows and slices2022-02-02T18:34:26ZIan Jacksonijackson@chiark.greenend.org.uk2022-02-02T18:18:16Zurn:sha1:b7fec1c8a4ee185b12648414242575e5aec5b581
Found via clippy::needless_borrow. In some cases I removed needless
`[..]` too. See also:
needless_borrow suggestion doesn't go far enough
https://github.com/rust-lang/rust-clippy/issues/8389
clippy: Rename a `decode_chanmsg` from `handle_`2022-01-19T18:46:13ZIan Jacksonijackson@chiark.greenend.org.uk2022-01-19T18:45:20Zurn:sha1:9a77b23cddee1ed378ac0af8c2b248c3880bc069
As per https://gitlab.torproject.org/tpo/core/arti/-/merge_requests/249#note_2771025
It doens't really handle it.
clippy: Rename a `from_foo` method that doesn't do conversion2022-01-19T17:26:09ZIan Jacksonijackson@chiark.greenend.org.uk2022-01-19T17:20:35Zurn:sha1:c4338999488cd332544c6d91cee6b42a36ebe569Describe when we will need SendmeAcceptMinVersion2022-01-12T20:52:24ZNick Mathewsonnickm@torproject.org2022-01-12T20:52:24Zurn:sha1:bbf7c59b2dc7ed8aeaf3604beeb619961874610c
(spoiler: not until we have a relay implementation)
Closes #53.
Change RequireSendmeAuth to an enum.2022-01-12T20:46:36ZNick Mathewsonnickm@torproject.org2022-01-12T20:46:36Zurn:sha1:e335f6c75aca2a39244390bd5f3c8ca66b2071d4
This is a fine example of why booleans are risky:
it's far to easy to pass "animate:bool" into "inanimate:bool" like
we did here.
This is a followup from our fix to #294.
Fix a boolean inversion in auth_sendme_optional.2022-01-12T20:36:55ZNick Mathewsonnickm@torproject.org2022-01-12T20:32:16Zurn:sha1:1e915c39465d616d71e57331d89e02235f156a31
Previously we were requiring authenticated sendme cells exactly when we
should be permitting the old format, and vice versa.
This bug was caused by using a boolean to represent one property, but
with giving that boolean two different senses without inverting at the
right time.
The next commit will prevent a recurrence.
Closes #294