We could be reading multiple messages from a socket buffer at once _without actually processing them yet_ which means that `fSuccessfullyConnected` might not be switched to `true` at the time we already parsed `VERACK` message and started to parse the next one. This is basically a false positive and we drop a legit node as a result even though the order of messages sent by this node was completely fine. To fix this I partially reverted #2790 (where the issue was initially introduced) and moved the logic for tracking the first message into ProcessMessage instead.
fad63eb [logging] Don't incorrectly log that REJECT messages are unknown. (John Newbery)
Pull request description:
Reject messages are logged to debug.log if NET debug logging is enabled.
Because of the way the `ProcessMessages()` function is structured,
processing for REJECT messages will also drop through to the default
branch and incorrectly log `Unknown command "reject" from peer-?`. Fix
that by exiting from `ProcessMessages()` early.
without this PR:
```
2018-05-03T17:37:00.930600Z received: reject (21 bytes) peer=0
2018-05-03T17:37:00.930620Z Reject message code 16: spammy spam
2018-05-03T17:37:00.930656Z Unknown command "reject" from peer=0
```
with this PR:
```
2018-05-03T17:35:04.751246Z received: reject (21 bytes) peer=0
2018-05-03T17:35:04.751274Z Reject message code 16: spammy spam
```
Tree-SHA512: 5c84c98433ab99e0db2dd481f9c2db6f87ff0d39022ff317a791737e918714bbcb4a23e81118212ed8e594ebcf098ab7f52f7fd5e21ebc3f07b1efb279b9b30b
fa6c3dea420b6c50c164ccc34f4e9e8a7d9a8022 p2p: Clarify control flow in ProcessMessage() (MarcoFalke)
Pull request description:
`ProcessMessage` is effectively a massive switch case construct. In the past there were attempts to clarify the control flow in `ProcessMessage()` by moving each case into a separate static function (see #9608). It was closed because it wasn't clear if moving each case into a function was the right approach.
Though, we can quasi treat each case as a function by adding a return statement to each case. (Can be seen as a continuation of bugfix #13162)
This patch does exactly that.
Also note that this patch is a subset of previous approaches such as #9608 and #10145.
Review suggestion: `git diff HEAD~ --function-context`
Tree-SHA512: 91f6106840de2f29bb4f10d27bae0616b03a91126e6c6013479e1dd79bee53f22a78902b631fe85517dd5dc0fa7239939b4fefc231851a13c819458559f6c201
Sometimes the node we ask for mnlistdiff is so fast to reply that we receive the message back before we reset `last_mnlistdiff`. To fix this we should reset it before sending the message, not after.
* Check MNs up to 24 blocks deep when verifying `dstx`
* Handle DSTX-es more like regular txes and not like "other" invs
* Try asking for a DSTX too when trying to find missing tx parents
* Check DSTX-es when chainlock arrives
`HasChainLock` was always `false` in `IsExpired` because tip is updated before the corresponding chainlock is received
* Apply `Handle DSTX-es more like regular txes` idea to `AlreadyHave()`
* Alternative handling of DSTX+recentRejects
Co-authored-by: Alexander Block <ablock84@gmail.com>
This has the wanted side effect of proper locking of "cs" inside
CommitCurTransaction and RollbackCurTransaction, which was not easily
possible to implement in the generic version. This fixes some rare crashes.
* adding favicon image
- noticed that someone already contributed all my pixmaps I have been using for
my builds (thank you, whoever you were). The favicon was left out. Adding
that as well. It is useful from a completist perspective and have users a
resource they can just use if they have need of a favicon for their project.
* trivial: adding SVG and high contrast icons
Purpose: for accessibility, if QT wallet is packaged (linux) correctly,
when the user switches to high contrast in their desktop environment
the high contrast icons will be used.
The packaging should land these icons in the
`/usr/share/icons/HighContrast` and `/usr/share/icons/hicolor` trees.
* trivial: HighContrast SVG is actually HighContrast now.
We removed "alert" p2p message and changed "mempool" message behaviour a bit. It probably makes sense to bump PROTOCOL_VERSION now. This should also help to distinguish v0.15 nodes from late v0.14.0.x ones.
* Don't load caches when blocks/chainstate was not present
* Delete old cache files when we decided to not load them
* Make sure cache files are of the exact format we expected them to be, flush empty objects into them instead of deleting files naively
* Streamline logic a bit, rename fIgnoreCacheFiles to fLoadCacheFiles
Co-authored-by: UdjinM6 <UdjinM6@users.noreply.github.com>
* Wait a little bit longer in wallet-encryption.py
The timed wallet locking is happening by an asynchronous task internally,
which means that it might need a little bit longer then the specified
timeout.
* Add workaround for p2p-versionbits-warnings.py until we backport bitcoin#12264
* build: use osslsigncode 2.0 in gitian
The original osslsigncode project (https://sourceforge.net/projects/osslsigncode/) has been marked as abandonware,
"This is now - and has been for a long while - abandonware. Feel free to create your own forks etc.".
However, a fork at https://github.com/mtrojnar/osslsigncode has emerged that has incorporated
theuni's patches, updated the tool to work with OpenSSL 1.1 and made other improvements.
This commit switches the windows signer descriptor to use this new version of osslsigncode.
* Fixed wget call in gitian-build.py
Co-authored-by: Michael <fanquake@gmail.com>
Co-authored-by: willyk <k.o.willy@gmail.com>
* move privatesend rpc methods from masternode.cpp to new privatesend.cpp
Signed-off-by: Pasta <pasta@dashboost.org>
* add ifdef ENABLE_WALLET check for wallet.h import
Signed-off-by: Pasta <pasta@dashboost.org>
* actually register privatesend rpc
Signed-off-by: Pasta <pasta@dashboost.org>
* add dropped help text and change some weird spacing below
Signed-off-by: Pasta <pasta@dashboost.org>
* Sync mempool from other nodes on start
* Add `-syncmempool` cmd-line option to be able to disable mempool sync if needed
* Only sync mempool with outbound peers
Co-authored-by: Alexander Block <ablock84@gmail.com>
* Move PS mixing entry verification on masternodes into `AddEntry()`
Also streamline logic a bit and drop unused/excessive parts.
* Unify PS checks among masternodes and clients
* No need to re-check outputs over and over again
* No need to count, fail early if any output is missing
* No need to look any further once we found the input we expected
A tx with duplicate inputs would be considered invalid anyway and we also know there are no duplicates because we just verified the final tx above.
Also drop an unused variable.
* Unify LogPrint-s
* Drop human-readable strings for unused PoolMessage-s
* Apply suggestions from code review
Co-Authored-By: PastaPastaPasta <6443210+PastaPastaPasta@users.noreply.github.com>
* Re-introduce zero-fee checks
* fix log
* Move all txin/txout verification logic shared by CPrivateSendClientSession::SignFinalTransaction() and CPrivateSendServer::AddEntry() into CPrivateSendBaseSession::IsValidInOuts()
* fix nit
* Add missing return
* Use CCoinsViewMemPool instead of doing it manually
Co-authored-by: PastaPastaPasta <6443210+PastaPastaPasta@users.noreply.github.com>
Co-authored-by: Alexander Block <ablock84@gmail.com>
* Do not sleep at the last block of the phase, it's not safe
* Refactor it a bit to make it clearer what's going on here
* Stop sleeping if blocks came faster than we expected
