e7b02b54cc Add roundtrip and more tests to ParseISO8601DateTime and FormatISO8601DateTime (Elichai Turkel)
9e2c623be5 Rename DecodeDumpTime to ParseISO8601DateTime and move to time.cpp (Elichai Turkel)
Pull request description:
As discussed in #17245.
1. Renamed the function.
2. Moved it from `rpcdump.cpp` to `time.cpp`.
3. Added a check if the time is less then epoch return 0 to prevent an overflow.
4. Added more edge cases tests and a roundtrip test.
ACKs for top commit:
laanwj:
ACK e7b02b54cc
MarcoFalke:
ACK e7b02b54cc
promag:
Code review ACK e7b02b54cc. Moved code is correct, left a comment regarding the test change.
Tree-SHA512: 703c21e09b2aabc992235149e67acba63d9d77a593ec8f6d2fec3eb63a7e5c406d56cbce6c6513ab32fba43367d073d2345f3b589843e3c5fe4f55ea3e00bf29
ed2dc5e48a Add override/final modifiers to V1TransportDeserializer (Pieter Wuille)
f342a5e61a Make resetting implicit in TransportDeserializer::Read() (Pieter Wuille)
6a91499496 Remove oversized message detection from log and interface (Pieter Wuille)
b0e10ff4df Force CNetMessage::m_recv to use std::move (Jonas Schnelli)
efecb74677 Use adapter pattern for the network deserializer (Jonas Schnelli)
1a5c656c31 Remove transport protocol knowhow from CNetMessage / net processing (Jonas Schnelli)
6294ecdb8b Refactor: split network transport deserializing from message container (Jonas Schnelli)
Pull request description:
**This refactors the network message deserialization.**
* It transforms the `CNetMessage` into a transport protocol agnostic message container.
* A new class `TransportDeserializer` (unique pointer of `CNode`) is introduced, handling the network buffer reading and the decomposing to a `CNetMessage`
* **No behavioral changes** (in terms of disconnecting, punishing)
* Moves the checksum finalizing into the `SocketHandler` thread (finalizing was in `ProcessMessages` before)
The **optional last commit** makes the `TransportDeserializer` following an adapter pattern (polymorphic interface) to make it easier to later add a V2 transport protocol deserializer.
Intentionally not touching the sending part.
Pre-Requirement for BIP324 (v2 message transport protocol).
Replacement for #14046 and inspired by a [comment](https://github.com/bitcoin/bitcoin/pull/14046#issuecomment-431528330) from sipa
ACKs for top commit:
promag:
Code review ACK ed2dc5e48a.
marcinja:
Code review ACK ed2dc5e48a
ryanofsky:
Code review ACK ed2dc5e48a. 4 cleanup commits added since last review. Unaddressed comments:
ariard:
Code review and tested ACK ed2dc5e.
Tree-SHA512: bab8d87464e2e8742529e488ddcdc8650f0c2025c9130913df00a0b17ecdb9a525061cbbbd0de0251b76bf75a8edb72e3ad0dbf5b79e26f2ad05d61b4e4ded6d
3bb0a4674f bench: Fix negative values and zero for -evals flag (nijynot)
Pull request description:
This PR makes `bench_bitcoin -evals=0` evaluate at once and throws when `-evals` is a negative integer.
---
Currently when you run `bench_bitcoin -evals=0`, it'll get stuck at
```
# Benchmark, evals, iterations, total, min, max, median
```
. This is not intuitively expected and should instead evaluate instantly as it's set to zero. Negative integers for `-evals` does not make sense either and should throw if set.
ACKs for top commit:
laanwj:
ACK 3bb0a4674f
Tree-SHA512: 03cd4c7c55134c7ffd8cdb6ee993551ce41061a73e13c3c047247af9df1fd7ed07d798272b643ec864099036922aaadbdcd2b798d710406f48df60b9d5448c26
fa398091b7 Avoid unused call to GuessVerificationProgress in NotifyHeaderTip (MarcoFalke)
Pull request description:
`GuessVerificationProgress` for a header (not a block) is always 0 because the number of txs in the block can not be determined from the header alone. Anyway, this result was never used, so we can optimize this call by hardcoding 0.
This is the next commit in a series of changes toward removing nChainTx (see #14863, #13875)
ACKs for top commit:
promag:
Code review ACK fa398091b7, missed that.
laanwj:
ACK fa398091b7
Tree-SHA512: 11016f8dbb1af1cf75241948d1ad35eac0c79d1311cd0db8c6ec806df2a9e3dc5f998dbd66ccbad5d84564e6cec7fe21ce7a2a13c2b34c746e2d3b31aa1db53a
faeb666536 util: Add CHECK_NONFATAL and use it in src/rpc (MarcoFalke)
Pull request description:
Fixes#17181
Currently, we use `assert` in RPC code to document logic and code assumptions. However, it seems a bit extreme to abort all of Bitcoin Core on an assert failure in one of the RPC threads. I suggest to replace all `assert`s with a macro `CHECK_NONFATAL(condition)` that throws a runtime error when the condition evaluates to `false`. That runtime error will then be returned to the rpc caller and will include instructions to report the error to our issue tracker.
ACKs for top commit:
practicalswift:
ACK faeb666536
laanwj:
ACK faeb666536
ryanofsky:
Code review ACK faeb666536
Tree-SHA512: 9b748715a5e0767ac11f1324a95a3a6ec672a0e0658013492219223bda83ce4b1b447fd8183bbb235f7df5ef7dddda7666ad569544b4d61cc65f232ca7a800ec
6b6be41c36 gui: Make polling in ClientModel asynchronous (João Barbosa)
Pull request description:
After #14193 `ClientModel::updateTimer` can take some time, as such the GUI hangs, like #17112.
Fixes this by polling in a background thread and updating the GUI asynchronously.
ACKs for top commit:
laanwj:
ACK 6b6be41c36
Sjors:
Code review re-ACK 6b6be41; only replaced the scary cast with `{ timer->start(); }`
Tree-SHA512: fd98b0c6535441aee3ee03c48b58b4b1f9bdd172ec6b8150da883022f719df34cabfd4c133412bf410e7f709f7bf1e9ef16dca05ef1f3689d526ceaeee51de38
8c6081a884 compat: remove bswap_* check on macOS (fanquake)
2cba35ab38 build: skip building OpenSSL lib_ssl (fanquake)
45a2d3c552 build: remove OpenSSL from Qt build (fanquake)
befbc40eb5 build: remove EVP_MD_CTX_new detection (fanquake)
fcee10c2d0 build: remove SSL lib detection (fanquake)
c7f30dbca8 gui: Update BIP70 support message (fanquake)
a3e810326d build: remove BIP70 entries from macOS Info.plist (fanquake)
72fe13a58d gui: remove payment request file handling from OpenURI dialog (fanquake)
3548e4aac7 Remove BIP70 Support (fanquake)
1cb9a4e28c docs: remove protobuf from docs (fanquake)
67328bb7ca build: remove protobuf from depends (fanquake)
Pull request description:
This removes [BIP70](https://github.com/bitcoin/bips/blob/master/bip-0070.mediawiki) support. It also removes OpenSSL linking from Qt and building OpenSSLs `lib_ssl` in depends, as well as SSL lib detection from the build system. It's something that I'd optimistically like to do for `0.20.0`.
ACKs for top commit:
laanwj:
Code review ACK 8c6081a884
MarcoFalke:
ACK 8c6081a884
fjahr:
ACK 8c6081a
Tree-SHA512: 9dd9153afa4eca1a795f983e5b31f5fee9fa9a064c2a95d2f98810689add3ad0bf221c4608282299e66e4d1ec31cd556d4b16eea55de7912c3b9931f64735883
04dbdd613f [net] SocketHandler: log peer id for close and disconnect (Sjors Provoost)
Pull request description:
When combined with `-logips` this makes it easier to diagnose disconnects.
To test on macOS, find a connection you want to disrupt:
```
lsof -nP -iTCP:8333 -sTCP:ESTABLISHED
```
To shut it down gracefully you can use tcpkill or this Python script: https://github.com/google/tcp_killer
The log should say:
```
2019-10-25T13:26:55Z socket closed for peer=1
2019-10-25T13:26:55Z disconnecting peer=1
2019-10-25T13:26:55Z Cleared nodestate for peer=1
```
To shut it down ungracefully I made a patch to the above script, adding a `-force` argument. _Careful, this may result in data corruption_. Then the log should say:
```
2019-10-25T13:39:57Z socket select error Bad file descriptor (9)
2019-10-25T13:39:57Z socket recv error for peer=0: Bad file descriptor (9)
2019-10-25T13:39:57Z disconnecting peer=0
2019-10-25T13:39:57Z Socket close failed: 35. Error: Bad file descriptor (9)
2019-10-25T13:39:57Z Cleared nodestate for peer=0
```
ACKs for top commit:
MarcoFalke:
unsigned ACK 04dbdd613f
TheBlueMatt:
unsigned ACK 04dbdd613f LGTM!
theuni:
unsigned ACK 04dbdd613f.
Tree-SHA512: 415313908484f97ffe11a48b4ed6afab3ab0be660c788adb9ad975f88b69aa1cfd5ccbe5859350cdf19ef8fde191fd530fb22cef34e70638defdc9f3d761c71d
dc2fdb9907 tests: Add fuzzing harness for various CScript related functions (practicalswift)
Pull request description:
Add fuzzing harness for various `CScript` related functions.
**Testing this PR**
Run:
```
$ CC=clang CXX=clang++ ./configure --enable-fuzz --with-sanitizers=address,fuzzer,undefined
$ make
$ src/test/fuzz/script
…
# And to to quickly verify that the relevant code regions are triggered, that the
# fuzzing throughput seems reasonable, etc.
$ contrib/devtools/test_fuzzing_harnesses.sh '^script$'
```
`test_fuzzing_harnesses.sh` can be found in PR #17000.
Top commit has no ACKs.
Tree-SHA512: a0c5dca3b64ae177020b2ca299a29015d70755231b6bf01edbfc67c8aac90c44b1b4d57350c3aebef6e031108e6ae8e5fa0987c67707831c314f5d3090e0cee8
0a433fc876 [validation] Remove unused cacheSigStore from CheckInputsFromMempoolAndCache (John Newbery)
Pull request description:
CheckInputsFromMempoolAndCache() is only called in one place, and
cacheSigStore is set to true in that call site. Remove the argument
entirely.
Also improve commenting.
ACKs for top commit:
MarcoFalke:
unsigned ACK 0a433fc876 Comment looks good
jamesob:
ACK 0a433fc876
laanwj:
ACK 0a433fc876
fanquake:
ACK 0a433fc876. Checked that `CheckInputsFromMempoolAndCache` is only called once, in `MemPoolAccept::ConsensusScriptChecks`, and that `cacheSigStore` is true.
Tree-SHA512: e4b4d2550e35df55c8f8fa4c539174cc2d3728112ddb937cb2ff759d8630a01566b5ec42a70a82e33994e6586f5a457a75a59f64b15d27c65331c723cbb097af
b05ec410f2 Add unit testing for the CompressScript functions (marcaiaf)
Pull request description:
Salvaging #15104 which adds unit tests for CompressScript function in `compressor.cpp`
Tested following cases for the CScript:
- CKeyID
- CScriptID
- Uncompressed CPubKey (of size: 65)
- Compressed CPubKey (of size: 32)
ACKs for top commit:
theStack:
ACK b05ec410f2
Tree-SHA512: 7e23ace39383122802dfe5f7d38190d772f5db4045a67b7a9bd4c06797a17e0cdc41d6fac92d448057eb7df50172155dc824587c16c68c79fd1a4de37b772001
fa92813407 consensus: Explain why fCheckDuplicateInputs can not be skipped and remove it (MarcoFalke)
Pull request description:
As a follow up to CVE-2018-17144, this removes the unused `fCheckDuplicateInputs` parameter and explains why the test can not be disabled. Apart from protecting against a dumb accident in the future, this should document the logic in the code. There is a technical write-up that explains how the underlying coins database behaves if this test is skipped: https://bitcoincore.org/en/2018/09/20/notice/#technical-details. However, it does not explicitly mention why the test can not be skipped. I hope my code comment does that.
ACKs for top commit:
jnewbery:
ACK fa92813407
amitiuttarwar:
utACK fa92813407
Empact:
Code review ACK fa92813407
promag:
ACK fa92813407.
Tree-SHA512: fc1ef670f1a467c543b84f704b9bd8cc7a59a9f707be048bd9b4e85fe70830702aa560a880efa2c840bb43818ab44dfdc611104df04db2ddc14ff92f46bfb28e
50037e97d1 depends: fix boost mac cross build with clang 9+ (Cory Fields)
Pull request description:
The ancient "darwin-4.9.1" profile has long been used to match against clang, which prior to version 9, reported 4.9.1 as its version when invoking "clang++ -dumpversion". Presumably this was a historical compatibility quirk related to Apple's switch from gcc to clang.
This was "fixed" in clang 9.0, so that -dumpversion reports the real version. Unfortunately that had the side-effect of breaking the (brittle) boost compiler detection.
Move to the seemingly more-correct "clang-darwin" profile, which passes the checks and builds correctly.
Also switch to using ar rather than libtool for archiving, as it's what the clang-darwin profile expects to be using.
Note that because this is using a different profile, some of the final command-line arguments end up changing. Those changes look sane at a glance.
ACKs for top commit:
fanquake:
ACK 50037e97d1 - tested on on macOS, will wait for the gitian build.
Tree-SHA512: eac1f353513a445add6fbece7fc78dd3dbdde5e2219bfb7739b82f40bb14de449667a94d2e303d43c67d9b38e7ceb0ba5f0d8fe20b40be2017b1ca0875467c2c
168b781fe7 Continue relaying transactions after they expire from mapRelay (Anthony Towns)
Pull request description:
This change allows peers to request transactions even after they've expired from mapRelay and even if they're not doing mempool requests. This is intended to allow for CPFP of old transactions -- if parent tx P wasn't relayed due to low fees, then a higher fee rate child C is relayed, peers will currently request the parent P, but we prior to this patch, we will not relay it due to it not being in mapRelay.
ACKs for top commit:
MarcoFalke:
re-ACK 168b781fe7 (only change is comment fixup)
sdaftuar:
re-ACK 168b781fe7
sipa:
ACK 168b781fe7
Tree-SHA512: b206666dd1450cd0a161ae55fd1a7eda2c3d226842ba27d91fe463b551fd924b65b92551b14d6786692e15cf9a9a989666550dfc980b48ab0f8d4ca305bc7762
This was originally added in #9366 to fix the gui build, as
Protobuf would also define these macros. Now that we're no-longer
using Protobuf, remove the additional check.
fa710066b9 ci: Disable functional tests on mac host (MarcoFalke)
Pull request description:
Judging from the lack of responses to https://github.com/bitcoin/bitcoin/issues/15400#issuecomment-543745053, no one can reproduce the failures locally. Thus, disable the tests on the ci mac host. Otherwise they cause ci failures to be ignored or overwritten by a blind re-run.
ACKs for top commit:
laanwj:
ok, ACK fa710066b9 in that case
ryanofsky:
Code review ACK fa710066b9. I've been blindly restarting these but can confirm it seemed like random test failures.
Tree-SHA512: 805f552124409f6e79b16b419826d334307202b8e3636edc8a91dadffe9949554d924c00cb16bb09473ec1aaaf2d8eaf7a97c7c2b8c11bf6372135d39bba802b
CheckInputsFromMempoolAndCache() is only called in one place, and
cacheSigStore is set to true in that call site. Remove the argument
entirely.
Also improve commenting.
6f6852335f doc: Changed miniupnp links to https (Marius Kjærstad)
Pull request description:
doc: Changed miniupnp links to https
ACKs for top commit:
fjahr:
ACK 6f68523
fanquake:
ACK 6f6852335f - as mentioned in this [comment](https://github.com/bitcoin/bitcoin/pull/17166#issuecomment-545906845), in future, there's no need to open a new PR instead than squashing. One of the requirements of contributing to Bitcoin Core is some basic `git` usage, and we have some documentation on that in our [CONTRIBUTING.md](https://github.com/bitcoin/bitcoin/blob/master/CONTRIBUTING.md#squashing-commits) (which also mentions not to open multiple PRs).
Tree-SHA512: 1e806326552b94dc1c2f105bed822a1dd56127d42bbb0c15a16623ff084b2047331f2f1ad5303c99b0d68c70ad2caf8e8028eb354ba4379973b9e1198676ce72
c2f964a674 tests: Remove Cygwin WinMain workaround (practicalswift)
db4bd32cc3 tests: Skip unnecessary fuzzer initialisation. Hold ECCVerifyHandle only when needed. (practicalswift)
Pull request description:
Skip unnecessary fuzzer initialisation. Hold `ECCVerifyHandle` only when needed.
As suggested by MarcoFalke in https://github.com/bitcoin/bitcoin/pull/17018#discussion_r336645391.
Top commit has no ACKs.
Tree-SHA512: 598da44859d736e3fdc143b93e07f444d8ad19dfdab0cfe7c6ccff8644e862664d869337dfe6b49416ed09a0024e4a5f2220ca6246de568f9e9227d721baa28e
9075d13153 [docs] Add release notes for removal of REJECT reasons (John Newbery)
04a2f326ec [validation] Fix REJECT message comments (John Newbery)
e9d5a59e34 [validation] Remove REJECT code from CValidationState (John Newbery)
0053e16714 [logging] Don't log REJECT code when transaction is rejected (John Newbery)
a1a07cfe99 [validation] Fix peer punishment for bad blocks (John Newbery)
Pull request description:
We no longer send BIP 61 REJECT messages, so there's no need to set
a REJECT code in the CValidationState object.
Note that there is a minor bug fix in p2p behaviour here. Because the
call to `MaybePunishNode()` in `PeerLogicValidation::BlockChecked()` only
previously happened if the REJECT code was > 0 and < `REJECT_INTERNAL`,
then there are cases were `MaybePunishNode()` can get called where it
wasn't previously:
- when `AcceptBlockHeader()` fails with `CACHED_INVALID`.
- when `AcceptBlockHeader()` fails with `BLOCK_MISSING_PREV`.
Note that `BlockChecked()` cannot fail with an 'internal' reject code. The
only internal reject code was `REJECT_HIGHFEE`, which was only set in
ATMP.
This reverts a minor bug introduced in 5d08c9c579.
ACKs for top commit:
ariard:
ACK 9075d13, changes since last reviewed are splitting them in separate commits to ease understanding and fix nits
fjahr:
ACK 9075d13153, confirmed diff to last review was fixing nits in docs/comments.
ryanofsky:
Code review ACK 9075d13153. Only changes since last review are splitting the main commit and updating comments
Tree-SHA512: 58e8a1a4d4e6f156da5d29fb6ad6a62fc9c594bbfc6432b3252e962d0e9e10149bf3035185dc5320c46c09f3e49662bc2973ec759679c0f3412232087cb8a3a7
