cerc-io/lotus
16
0
Fork 0
Code Issues Pull Requests Packages Projects Releases 27 Wiki Activity
7,093 Commits 24 Branches 342 Tags 314 MiB
33b12782e4
Commit Graph

190 Commits

Author SHA1 Message Date
Jakub Sztandera
33b12782e4
Take into account messages in current tispet by executing it 
Signed-off-by: Jakub Sztandera <kubuxu@protocol.ai>
 2020-08-25 04:02:06 +02:00
Jakub Sztandera
b2ff231c48
Fix error fmt string 
Signed-off-by: Jakub Sztandera <kubuxu@protocol.ai>
 2020-08-25 03:14:52 +02:00
Jakub Sztandera
d921791c55
Add comment 
Signed-off-by: Jakub Sztandera <kubuxu@protocol.ai>
 2020-08-25 03:13:43 +02:00
whyrusleeping
05d65c3f88 slightly better than panicking 2020-08-24 17:40:12 -07:00
whyrusleeping
cf3298cd04 more correctly handle discrepancies between mempools head and the mining base 2020-08-24 17:21:03 -07:00
Jakub Sztandera
aacfcba400
Check ValidForBlockInclusion in mpool 
Signed-off-by: Jakub Sztandera <kubuxu@protocol.ai>
 2020-08-25 01:01:16 +02:00
Łukasz Magiera
0806fd651a
Merge pull request #3183 from filecoin-project/steb/lint-all 
Lint everything
 2020-08-22 21:47:41 +02:00
vyzo
817358f1bb better semantics for mpool clear local argument 
local messages should be kept unless the parameter is true
 2020-08-21 23:48:35 +03:00
vyzo
087955e927 add localonly option to MpoolClear 2020-08-21 23:32:41 +03:00
vyzo
e33d398c33 remove pending local messages from the datastore when clearing the mpool 2020-08-21 23:32:41 +03:00
vyzo
3027283756 add MpoolClear api 2020-08-21 23:32:41 +03:00
vyzo
161cb32078 recover from errors in head change 2020-08-21 23:32:41 +03:00
Steven Allen
5733c71c50 Lint everything 
We were ignoring quite a few error cases, and had one case where we weren't
actually updating state where we wanted to. Unfortunately, if the linter doesn't
pass, nobody has any reason to actually check lint failures in CI.

There are three remaining XXXs marked in the code for lint.
 2020-08-20 20:46:36 -07:00
vyzo
6d9707c784 fix OptimalSelection3 test 2020-08-18 22:24:07 +03:00
Jakub Sztandera
e0b72ba049
Reduce number of runs to 1 
Signed-off-by: Jakub Sztandera <kubuxu@protocol.ai>
 2020-08-18 21:10:04 +02:00
Jakub Sztandera
4d01cbfb62
Increase precision of output 
Signed-off-by: Jakub Sztandera <kubuxu@protocol.ai>
 2020-08-18 21:10:04 +02:00
Jakub Sztandera
8d4427a8bc
add best ticket reward to tests 
Signed-off-by: Jakub Sztandera <kubuxu@protocol.ai>
 2020-08-18 21:10:04 +02:00
Jakub Sztandera
8bca545cab
Add test for winnerProba 
Signed-off-by: Jakub Sztandera <kubuxu@protocol.ai>
 2020-08-18 21:10:04 +02:00
Jakub Sztandera
9c98cf8bc1
Fix blockProbabilities 
Signed-off-by: Jakub Sztandera <kubuxu@protocol.ai>
 2020-08-18 21:10:04 +02:00
Jakub Sztandera
8acd926636
Use conditional probability for noWinners, fix tests 
Signed-off-by: Jakub Sztandera <kubuxu@protocol.ai>
 2020-08-18 21:10:04 +02:00
vyzo
4287f76ed4 fix deadlock in mpool.Push 2020-08-18 10:19:46 +03:00
vyzo
3b6e2bdb7b trigger early republish from head changes 2020-08-17 10:03:39 +03:00
vyzo
a5d2ec0df8 appease linter 2020-08-14 10:12:55 +03:00
vyzo
b844f70a3c make competitve selection test parameteric on distribution function, add test for Zipf distribution 2020-08-14 10:09:25 +03:00
Jakub Sztandera
f27a122787
Use Zipf distribution for messages 
Signed-off-by: Jakub Sztandera <kubuxu@protocol.ai>
 2020-08-13 20:22:56 +02:00
vyzo
9aa8791920 make selection2 test less touchy 2020-08-13 20:02:08 +03:00
vyzo
2eff03fa1f small fixes in optimal selection 
1. Remove noop updates from the previous dependencies
2. Update subsequent dependencies, which was the intention
 2020-08-13 19:49:07 +03:00
Jakub Sztandera
0daed04c07
Don't run the test 50 times 
Signed-off-by: Jakub Sztandera <kubuxu@protocol.ai>
 2020-08-13 14:08:11 +02:00
vyzo
09f5a64ce0
use gasPerf for negative performance check 
it's the same in actually because effPerf<0 iff gasPerf<0, but we should be consistent

Signed-off-by: Jakub Sztandera <kubuxu@protocol.ai>
 2020-08-13 14:08:11 +02:00
vyzo
f3eec33bec
make OptimalSelection3 test pass 
Signed-off-by: Jakub Sztandera <kubuxu@protocol.ai>
 2020-08-13 14:08:11 +02:00
vyzo
1aa8b6cab2
make OptimalSelection2 test pass 
Signed-off-by: Jakub Sztandera <kubuxu@protocol.ai>
 2020-08-13 14:08:11 +02:00
vyzo
0f5598e59b
compute average boost in tests 
Signed-off-by: Jakub Sztandera <kubuxu@protocol.ai>
 2020-08-13 14:08:11 +02:00
vyzo
5e8ae7498a
only adjust next chains if they have positve perf 
Signed-off-by: Jakub Sztandera <kubuxu@protocol.ai>
 2020-08-13 14:08:11 +02:00
vyzo
7bbf1c7db2
don't mix negative performing chains with their parent, add some comments 
Signed-off-by: Jakub Sztandera <kubuxu@protocol.ai>
 2020-08-13 14:08:10 +02:00
Jakub Sztandera
f35555964d
Better "optimal selection 
Signed-off-by: Jakub Sztandera <kubuxu@protocol.ai>
 2020-08-13 14:08:10 +02:00
vyzo
d25f386bb5
mixin the previous chain's effective performance to capture effect of dependencies 
Signed-off-by: Jakub Sztandera <kubuxu@protocol.ai>
 2020-08-13 14:08:10 +02:00
Steven Allen
3ef3f570fb Fix lint errors an broken tests 2020-08-12 23:16:44 +02:00
Łukasz Magiera
43586ed9a1
Merge pull request #3000 from filecoin-project/feat/mpool-add-semaphore 
Add semaphore in push and friends to reduce lock contention
 2020-08-12 21:53:59 +02:00
Jakub Sztandera
f7c222e42e
Validate message pool config, validate GasPremium < GasFeeCap 
don't use configured RBF for mpool

Signed-off-by: Jakub Sztandera <kubuxu@protocol.ai>
 2020-08-12 20:31:34 +02:00
vyzo
f2ac17845b relinquish the lock while signing in PushWithNonce 2020-08-12 20:26:58 +03:00
vyzo
7be18df6ea add semaphore in push and friends to reduce lock contention 2020-08-12 10:38:40 +03:00
vyzo
472e502218 fix republishing chain selection to account for edge case of inordinately long chains 2020-08-12 09:41:02 +03:00
vyzo
7b16fe3792 hold ts lock longer 2020-08-12 09:18:04 +03:00
vyzo
47f81f1ad6 increase repub message limit to 30 2020-08-12 09:13:57 +03:00
vyzo
cf34fe5cc3 extend the republish interval to something more reasonable 
pubsub timecache duration + 5 blocks + propagation delay
 2020-08-12 09:13:57 +03:00
vyzo
7396212c89 slightly increase the replublish interval to include the propagation delay as well 2020-08-12 09:13:57 +03:00
vyzo
e876617c82 new message republishing logic 2020-08-12 09:13:57 +03:00
vyzo
42951d05a5 refactor messagepool provider out of the main messagepool implementation 2020-08-12 09:13:57 +03:00
Jakub Sztandera
2efb18f677
Small fixes to mpool Trim 
Signed-off-by: Jakub Sztandera <kubuxu@protocol.ai>
 2020-08-11 23:55:06 +02:00
vyzo
9032163c5a turn probability eyeballing into an actual test 2020-08-11 20:47:50 +03:00