William Banfield
0808043fdd
if/else fiddling
2022-05-20 15:35:40 -04:00
William Banfield
7021acc6c5
fix lint
2022-05-20 15:35:28 -04:00
William Banfield
af2d148f3a
factor out make block function
2022-05-20 12:43:30 -04:00
William Banfield
c30e4add10
factor out makeReactor helper in blocksync tests
2022-05-20 12:02:48 -04:00
William Banfield
17aabc2d03
add reactor helper in test
2022-05-20 11:54:16 -04:00
William Banfield
a53d7a3f0c
fix additional nil pointer
2022-05-20 11:37:04 -04:00
William Banfield
01ca403e0e
use first.Height to avoid nil panic
2022-05-20 11:26:05 -04:00
William Banfield
bf05d09428
basic logic to make extneded commit optional
2022-05-20 11:14:58 -04:00
William Banfield
00a1e9c5b5
use second block commit instead of ext commit in persistence logic
2022-05-20 10:44:35 -04:00
William Banfield
06e70e4d7d
update conditional logic in blocksync reactor for extCommit peek
2022-05-20 10:35:55 -04:00
William Banfield
4be4e96ceb
use empty extended commit when not enabled
2022-05-20 10:23:25 -04:00
William Banfield
6889f77807
fix additional use of extended commit
2022-05-19 21:55:51 -04:00
William Banfield
d69cf82e3f
delete redundant test
2022-05-19 21:37:28 -04:00
William Banfield
551d493c42
do not save seen commit as canonical
2022-05-19 21:37:28 -04:00
William Banfield
fd2a7df517
basic unit tests for saving blocks with extended comit
2022-05-19 19:30:44 -04:00
William Banfield
1744ed2741
Merge branch 'master' into wb/add-consensus-param-internal
2022-05-19 17:55:34 -04:00
William Banfield
8233d5e176
fix lint
2022-05-19 17:29:48 -04:00
William Banfield
ad73e6da2f
consensus: update state from store before use in reactor ( #8576 )
...
Closes : #8575
This PR aims to fix the `LastCommitRound can only be negative for initial height 0` issue we see in the e2e tests by initializing the `state` object before starting the receive routines in the consensus reactor. This is somewhat inelegant, but it should fix the issue.
2022-05-19 19:35:30 +00:00
William Banfield
a437ebd6f3
Merge branch 'master' into wb/add-consensus-param-internal
2022-05-19 15:24:02 -04:00
M. J. Fromberger
4a9bbe047f
Fix lock sequencing in socket client request tracking. ( #8581 )
...
* Fix lock sequencing in socket client request tracking.
It is not safe to check base service state (IsRunning) while holding the lock
for the client state. If we do, then during shutdown we may deadlock with the
invocation of the OnStop handler, which the base service executes while holding
the service lock.
* Enqueue pending requests before sending them to the server.
If we don't do this, the server can reply before the request lands in the
queue. That will cause the receiver to terminate early for an unsolicited
response. So enqueue first: This is safe because we're doing it in the same
routine as services the channel, so we won't take another message till we are
safely past that point.
* Document what we did.
* Fix socket paths in tests.
2022-05-19 12:11:57 -07:00
William Banfield
fa8f2ee0aa
improve log statements
2022-05-19 13:43:26 -04:00
William Banfield
8d9b227566
fix consensus reactor tests for absent extension data
2022-05-19 13:43:14 -04:00
William Banfield
8c26a452b7
add test for panic on save with no extensions
2022-05-19 13:42:07 -04:00
William Banfield
8a37e2584d
fix test by adding extension data
2022-05-19 13:31:49 -04:00
William Banfield
e2989f9442
remove print statement
2022-05-19 13:31:35 -04:00
William Banfield
65c4150c3d
fix evidence test by adding extension sig
2022-05-19 13:30:11 -04:00
William Banfield
5031c82150
log on stripped extensions
2022-05-19 12:53:17 -04:00
William Banfield
be9b4a7fc0
use errorf instead of sprintf in panic
2022-05-19 12:43:19 -04:00
William Banfield
4e05b5f108
ensure extensions on extended commit save
2022-05-19 12:31:50 -04:00
William Banfield
05d930f40a
Update internal/blocksync/reactor.go
...
Co-authored-by: Sergio Mena <sergio@informal.systems >
2022-05-19 10:07:46 -04:00
William Banfield
10078f39d6
add nil check for block load
2022-05-18 16:53:55 -04:00
William Banfield
4194d4da46
comment proposal extensions validation
2022-05-18 14:46:00 -04:00
William Banfield
f50ac0e8ea
add state check on peer response
2022-05-18 14:37:03 -04:00
William Banfield
4da32ea146
update and document new logic for bailing out of blocksync
2022-05-18 14:07:39 -04:00
William Banfield
beb6c47e29
lint fix
2022-05-17 23:10:25 -04:00
William Banfield
cbd211a173
fix broken block store test
2022-05-17 22:37:04 -04:00
William Banfield
0c4a49600b
remove unused store code
2022-05-17 22:36:13 -04:00
William Banfield
98f614a614
use current block height as parameter in blocksync
2022-05-17 21:49:58 -04:00
William Banfield
6f577fa389
try commit and extcommit in blocksync response
2022-05-17 21:38:44 -04:00
William Banfield
4ceead171b
document new blockstore methods
2022-05-17 20:46:41 -04:00
William Banfield
588a310049
separate extended commit save
2022-05-17 20:37:59 -04:00
William Banfield
c3686fea00
Merge branch 'master' into wb/add-consensus-param-internal
2022-05-17 17:02:29 -04:00
William Banfield
dc253a8196
remove unnecessary assignment in proposal test
2022-05-17 16:39:46 -04:00
William Banfield
9e9f3a1433
strip extensions in consensus
2022-05-17 16:35:24 -04:00
William Banfield
931e897c6a
fix node test
2022-05-17 15:40:50 -04:00
William Banfield
d92b63b85f
fix evidence tests
2022-05-17 15:35:21 -04:00
William Banfield
2e1f5bdbcd
fix hvs test
2022-05-17 15:29:43 -04:00
William Banfield
5d84aaf7d4
fix tests to store and load state
2022-05-17 15:06:35 -04:00
William Banfield
5dc8a1fac0
hack to not generate protos yet
2022-05-17 15:06:17 -04:00
M. J. Fromberger
c620900fdd
rpc: fix plumbing of broadcast_tx_commit timeout ( #8573 )
...
In #3435 we allowed this timeout to override the global write timeout.
But after #8570 this meant we were applying a shorter timeout by default.
Don't do the patch if the timeout is already unlimited.
This is a temporary workaround; in light of #8561 I plan to get rid of this
option entirely during the v0.37 cycle, but meanwhile we should keep existing
use more or less coherent.
2022-05-17 10:34:43 -07:00
