3a7e79478ab41af7c53ce14d9fca9815bffe1f73 test: retry when write to a socket fails on macOS (Ivan Metlushko)
8cf9d15b823d91d2a74fc83832fccca2219342c9 test: use pgrep for better compatibility (Ivan Metlushko)
Pull request description:
Rationale: a few minor changes to make experience of running tests on macOS a bit better
1.`pidof` is not available on BSD/macOS, while `pgrep` is present on BSD, Linux and macOS
2. Add retry as a workaround for a weird behavior when writing to a socket (https://bugs.python.org/issue33450). Stacktrace attached
Man pages:
https://www.freebsd.org/cgi/man.cgi?query=pgrep&apropos=0&sektion=1&manpath=FreeBSD+6.0-RELEASE&arch=default&format=htmlhttps://man7.org/linux/man-pages/man1/pgrep.1.html
Related to #19281
Stacktrace example:
```
...
33/161 - feature_abortnode.py failed, Duration: 63 s
stdout:
2020-06-11T10:46:43.947000Z TestFramework (INFO): Initializing test directory /var/folders/2q/d5w9zh614r7g5c8r74ln3g400000gq/T/test_runner_₿_🏃_20200611_174102/feature_abortnode_128
2020-06-11T10:46:45.199000Z TestFramework (INFO): Waiting for crash
2020-06-11T10:47:15.921000Z TestFramework (INFO): Node crashed - now verifying restart fails
2020-06-11T10:47:47.068000Z TestFramework (INFO): Stopping nodes
[node 1] Cleaning up leftover process
stderr:
Traceback (most recent call last):
File "/Users/xxx/Projects/bitcoin/test/functional/feature_abortnode.py", line 50, in <module>
AbortNodeTest().main()
File "/Users/xxx/Projects/bitcoin/test/functional/test_framework/test_framework.py", line 142, in main
exit_code = self.shutdown()
File "/Users/xxx/Projects/bitcoin/test/functional/test_framework/test_framework.py", line 266, in shutdown
self.stop_nodes()
File "/Users/xxx/Projects/bitcoin/test/functional/test_framework/test_framework.py", line 515, in stop_nodes
node.stop_node(wait=wait)
File "/Users/xxx/Projects/bitcoin/test/functional/test_framework/test_node.py", line 318, in stop_node
self.stop(wait=wait)
File "/Users/xxx/Projects/bitcoin/test/functional/test_framework/coverage.py", line 47, in __call__
return_val = self.auth_service_proxy_instance.__call__(*args, **kwargs)
File "/Users/xxx/Projects/bitcoin/test/functional/test_framework/authproxy.py", line 142, in __call__
response, status = self._request('POST', self.__url.path, postdata.encode('utf-8'))
File "/Users/xxx/Projects/bitcoin/test/functional/test_framework/authproxy.py", line 107, in _request
self.__conn.request(method, path, postdata, headers)
File "/Users/xxx/.pyenv/versions/3.5.6/lib/python3.5/http/client.py", line 1107, in request
self._send_request(method, url, body, headers)
File "/Users/xxx/.pyenv/versions/3.5.6/lib/python3.5/http/client.py", line 1152, in _send_request
self.endheaders(body)
File "/Users/xxx/.pyenv/versions/3.5.6/lib/python3.5/http/client.py", line 1103, in endheaders
self._send_output(message_body)
File "/Users/xxx/.pyenv/versions/3.5.6/lib/python3.5/http/client.py", line 936, in _send_output
self.send(message_body)
File "/Users/xxx/.pyenv/versions/3.5.6/lib/python3.5/http/client.py", line 908, in send
self.sock.sendall(data)
OSError: [Errno 41] Protocol wrong type for socket
```
ACKs for top commit:
laanwj:
ACK 3a7e79478ab41af7c53ce14d9fca9815bffe1f73
Tree-SHA512: fefbe40ce94ab29f18bbbed2a434194b1384ffa5279b1d04db7a3708e3dd422bd9e450f1db3f95a1a851fac5a626ab533c6ebcfd7ede96f8ccae9e6f3e9fff92
eb6b73540d1ee7ff5a6874dd0e35f9b30b68e3b8 build: pass _WIN32_WINNT=0x0601 when building libevent for Windows (fanquake)
03e056edcd1a7f7197a29068c52fa33fce12f7d7 depends: Patch libevent build to fix IPv6 -rpcbind on Windows (Luke Dashjr)
Pull request description:
TLDR: This poaches a commit from #18287 and adds one more to adjust the Windows version targeted when building libevent. These changes combined should fully fix ipv6 usage with the RPC server on Windows.
---
Binding the RPC server to a ipv6 address does not currently work on Windows.
We currently try and bind to `127.0.0.1` and `::1` [by default](https://github.com/bitcoin/bitcoin/blob/master/src/httpserver.cpp#L304).
On Windows you'll see lines like this in debug.log:
```bash
2020-06-24T01:49:04Z libevent: getaddrinfo: nodename nor servname provided, or not known
2020-06-24T01:49:04Z Binding RPC on address ::1 port 8332 failed
```
This issue was bought up in, and supposedly fixed by #18287, however the two people that tested it, both said that it didn't fix the problem. I think I now understand why that change alone is incomplete.
Our call into libevent starts with [evhttp_bind_socket_with_handle()](https://github.com/bitcoin/bitcoin/blob/master/src/httpserver.cpp#L325):
```bash
evhttp_bind_socket_with_handle()
bind_socket()
make_addrinfo()
evutil_getaddrinfo()
if #USE_NATIVE_GETADDRINFO
#ifndef AI_ADDRCONFIG
evutil_adjust_hints_for_addrconfig_()
evutil_check_interfaces()
evutil_check_ifaddrs()
evutil_found_ifaddr()
// miss identifies ipv6 as ipv4?
#endif
evutil_getaddrinfo_common_()
```
The problem is falling into ["#ifndef AI_ADDRCONFIG"](https://github.com/libevent/libevent/blob/master/evutil.c#L1580):
```cpp
#ifndef AI_ADDRCONFIG
/* Not every system has AI_ADDRCONFIG, so fake it. */
if (hints.ai_family == PF_UNSPEC &&
(hints.ai_flags & EVUTIL_AI_ADDRCONFIG)) {
evutil_adjust_hints_for_addrconfig_(&hints);
}
#endif
```
When this occurs, hints end up being adjusted, and it seems that ipv6 addresses end up being mis-identified as ipv4?
However this shouldn't happen, as these `AI_` definitions are available on Windows.
The issue is that in evutil.c, `_WIN32_WINNT` [is set to `0x501`](https://github.com/libevent/libevent/blob/master/evutil.c#L45) (XP).
This obviously predates Vista (`0x0600`), which is when the `AI_ADDRCONFIG` definition (and others) became [available](https://docs.microsoft.com/en-us/windows/win32/api/ws2def/ns-ws2def-addrinfoa).
The change here will override libevents internal D_WIN32_WINNT defines. This should be ok, because it's only making "more" of the Windows API available. It's also aligned with what we do in our own configure, we pass [`D_WIN32_WINNT=0x0601`](https://github.com/bitcoin/bitcoin/blob/master/configure.ac#L610). We also now use linker flags to restrict our binary from running on a Windows version [earlier than Windows 7](https://github.com/bitcoin/bitcoin/blob/master/configure.ac#L621).
The combined fixes can be tested by running:
`bitcoind -rpcbind=::1 rpcallowip='0.0.0.0/0' -debug=http`
and then querying it using:
`bitcoin-cli -rpcconnect=::1 getblockchaininfo`
TODO:
- [x] Open an issue upstream. https://github.com/libevent/libevent/issues/1041
ACKs for top commit:
laanwj:
ACK eb6b73540d1ee7ff5a6874dd0e35f9b30b68e3b8
Tree-SHA512: e1e50f194911301981edaed0c216ed4efb9ebd4a1f9bc9b9f85bec7140b66c45c8666fd5db4aad359596559d4a08ab7c920e9d9736f3ecdbb841afc54e40586e
This enables of the use of AI_* definitions in the Windows headers,
specifically AI_ADDRCONFIG, which fixes an issue with libevent and
ipv6 on Windows.
It also aligns with what we define in configure when building Core.
6fdfeebcc79df62c8bf1cf4b6e9e97d6aefb3eb3 refactor: Replace RecursiveMutex with Mutex in rpc/server.cpp (Hennadii Stepanov)
Pull request description:
The functions that could lock this mutex, i.e., `SetRPCWarmupStatus()`, `SetRPCWarmupFinished()`, `RPCIsInWarmup()`, `CRPCTable::execute()`, do not call itself recursively, and do not call each other either directly or indirectly. Therefore, the `g_rpc_warmup_mutex` could be a non-recursive mutex.
Related to #19303.
ACKs for top commit:
laanwj:
ACK 6fdfeebcc79df62c8bf1cf4b6e9e97d6aefb3eb3
MarcoFalke:
ACK 6fdfeebcc79df62c8bf1cf4b6e9e97d6aefb3eb3
Tree-SHA512: 05a8ac58c0cd6a3c9afad9e06ad78059642e3e97715e129f379c0bf6dccdb58e70d05d965f23e7432fd3f02d7f97967a778ffb8e424837891d9d785a9e98964c
a92e48b02df545e620a7b1de74f647f46413d3fb test: move TEST_RUNNER_EXTRA into native tsan setup (fanquake)
Pull request description:
`feature_block.py` is being run in the tsan job, i.e [here](https://travis-ci.org/github/bitcoin/bitcoin/jobs/703122309), even though it should be excluded. My hasty assumption is that this will fix it. In any case, all other instances of `TEST_RUNNER_EXTRA` seem to have moved out of `.travis.yml` and into the different CI configurations.
ACKs for top commit:
MarcoFalke:
ACK a92e48b02df545e620a7b1de74f647f46413d3fb
practicalswift:
ACK a92e48b02df545e620a7b1de74f647f46413d3fb -- patch looks correct
hebasto:
ACK a92e48b02df545e620a7b1de74f647f46413d3fb, I have reviewed the code and it looks OK, I agree it can be merged.
Tree-SHA512: 86057bef2cc87c6acdbbf94f8cd7a5147510448c3e67aacde8daf247e3ccf649cfc5afbbd10693e084f426042d98150616c0e49bfa5f32b949dff9cebd2fd95d
9b009fae6e2eb0ab2ee7ce7882c3556a9ac363a7 qa: Test concurrent wallet loading (João Barbosa)
b9971ae5853c1d62e09d976a8705f4f731290d85 wallet: Handle concurrent wallet loading (João Barbosa)
Pull request description:
This PR handles concurrent wallet loading.
This can be tested by running in parallel the following script a couple of times:
```sh
for i in {1..10}
do
src/bitcoin-cli -regtest loadwallet foo
src/bitcoin-cli -regtest unloadwallet foo
done
```
Eventually the error occurs:
```
error code: -4
error message:
Wallet already being loading.
```
For reference, loading and already loaded wallet gives:
```
error code: -4
error message:
Wallet file verification failed. Error loading wallet w1. Duplicate -wallet filename specified.
```
Fixes#19232.
ACKs for top commit:
MarcoFalke:
Concept ACK 9b009fae6e2eb0ab2ee7ce7882c3556a9ac363a7 I have not reviewed the code
hebasto:
ACK 9b009fae6e2eb0ab2ee7ce7882c3556a9ac363a7, tested on Linux Mint 20 (x86_64):
ryanofsky:
Code review good-but-not-ideal ACK 9b009fae6e2eb0ab2ee7ce7882c3556a9ac363a7
Tree-SHA512: 0ccd77b03c0926e4c4e51efb31e193b93cb4b9ffe8bac6bb018f7344c55dfd939b873b8cf5e657dca73e6202eb75aa672de2acb787cc133184b0b3b51e47b972
9952242c03fe587b5dff46a9f770e319146103bf build: improve builtin_clz* detection (fanquake)
Pull request description:
Fixes#19402.
The way we currently test for `__builtin_clz*` support with `AC_CHECK_DECLS` does not work with Clang:
```bash
configure:21492: clang++-10 -std=c++11 -c -g -O2 -DHAVE_BUILD_INFO -D__STDC_FORMAT_MACROS conftest.cpp >&5
conftest.cpp💯10: error: builtin functions must be directly called
(void) __builtin_clz;
^
1 error generated.
```
This also removes the `__builtin_clz()` check, as we don't actually use it anywhere, and it's trvial to re-add detection if we do start using it at some point. If this is controversial then I'll add a test for it as well.
ACKs for top commit:
sipa:
ACK 9952242c03fe587b5dff46a9f770e319146103bf
laanwj:
ACK 9952242c03fe587b5dff46a9f770e319146103bf
Tree-SHA512: 695abb1a694a01a25aaa483b4fffa7d598842f2ba4fe8630fbed9ce5450b915c33bf34bb16ad16a16b702dd7c91ebf49fe509a2498b9e28254fe0ec5177bbac0
fa525e4d1cfda8c1924d2c69f43bd7ae3b98fb72 net: Avoid wasting inv traffic during IBD (MarcoFalke)
fa06d7e93489e61078cfb95ab767c001536a6e10 refactor: block import implies IsInitialBlockDownload (MarcoFalke)
faba65e696a88e5626e587f4e63fa15500cbe4d0 Add ChainstateManager::ActiveChainstate (MarcoFalke)
fabf3d64ff2bd14f762810316144bb9fd69c517c test: Add FeeFilterRounder test (MarcoFalke)
Pull request description:
Tx-inv messages are ignored during IBD, so it would be nice if we told peers to not send them in the first place. Do that by sending two `feefilter` messages: One when the connection is made (and the node is in IBD), and another one when the node leaves IBD.
ACKs for top commit:
jamesob:
ACK fa525e4d1cfda8c1924d2c69f43bd7ae3b98fb72 ([`jamesob/ackr/19204.1.MarcoFalke.p2p_reduce_inv_traffic_d`](https://github.com/jamesob/bitcoin/tree/ackr/19204.1.MarcoFalke.p2p_reduce_inv_traffic_d))
naumenkogs:
utACK fa525e4
gzhao408:
ACK fa525e4d1c
jonatack:
re-ACK fa525e4 checked diff `git range-diff 19612ca fa8a66c fa525e4`, re-reviewed, ran tests, ran a custom p2p IBD behavior test at 9321e0f223.
hebasto:
re-ACK fa525e4d1cfda8c1924d2c69f43bd7ae3b98fb72, only rebased since the [previous](https://github.com/bitcoin/bitcoin/pull/19204#pullrequestreview-429519667) review (verified with `git range-diff`).
Tree-SHA512: 2c22a5def9822396fca45d808b165b636f1143c4bdb2eaa5c7e977f1f18e8b10c86d4c180da488def38416cf3076a26de15014dfd4d86b2a7e5af88c74afb8eb
53361ddc7591f35df36cc71d11c24f826f209e11 [build] Remove unused RES_IMAGES (Bushstar)
Pull request description:
Remove RES_IMAGES. Seems to be unused since 2015 in the commit below.
98c222b5aa (diff-9a4f3a253de77bf90b107bdf5283ebc3R317)
The src/qt/res/images to which it was used with is no longer present either.
ACKs for top commit:
hebasto:
ACK 53361ddc7591f35df36cc71d11c24f826f209e11, tested on Linux Mint 20 (x86_64).
Tree-SHA512: d2f09ae225a4c6c171e1aae4c4a444064dc0502e96130e04ccb718f9fcf611a287c56630dec3e9a8937b5e29040d931a237da36180d2343c23cef30359e46323
fab57e2b9bc4577fcfcd9fbddbc35d96046c5d88 doc: Mention Span in developer-notes.md (Pieter Wuille)
3502a60418858a8281ddf2f9cd59daa8f01d2fa8 doc: Document Span pitfalls (Pieter Wuille)
Pull request description:
This is an attempt to document pitfalls with the use of `Span`, following up on comments like https://github.com/bitcoin/bitcoin/pull/18468#issuecomment-622846597 and https://github.com/bitcoin/bitcoin/pull/18468#discussion_r442998211
ACKs for top commit:
laanwj:
ACK fab57e2b9bc4577fcfcd9fbddbc35d96046c5d88
Tree-SHA512: 8f6f277d6d88921852334853c2b7ced97e83d3222ce40c9fe12dfef508945f26269b90ae091439ebffddf03f939797cb28126b2387f77959069ef8909c25ab53
fa74a54fad7abfe3b0c98c5a6e4780d63d35b13f ci: Increase test timeout for sanitizer configs (MarcoFalke)
Pull request description:
Hopefully fixes#19369
ACKs for top commit:
practicalswift:
ACK fa74a54fad7abfe3b0c98c5a6e4780d63d35b13f -- patch looks correct!
fanquake:
ACK fa74a54fad7abfe3b0c98c5a6e4780d63d35b13f - the test failure here is a different issue, and the problem referenced by this PR hasn't occurred, so I think this can be merged. It's also fixing the use of `--factor` which was replaced in #18986.
Tree-SHA512: bec44fff454f20b7c5f8a461560d2496765dea61186027cc0cdce5ac55be0488b6f7f172fec49b89fe59a75b455501e2b4ae91a98c4a17d5c1a722846d2b3b60
e12e970df6fcae08ff8008812cdeef600d6b2db8 docs: match usage text to script and location (Peter Bushnell)
Pull request description:
Update the usage text in the README to match the usage text in the Python script.
02b26ba1c1/contrib/testgen/gen_key_io_test_vectors.py (L9)02b26ba1c1/contrib/testgen/gen_key_io_test_vectors.py (L10)
Also to match the file names in the actual destination.
02b26ba1c1/src/test/data/key_io_valid.json02b26ba1c1/src/test/data/key_io_invalid.json
Following the README usage text generates new files when the user is likely to have wanted to update the existing files.
ACKs for top commit:
fanquake:
ACK e12e970df6fcae08ff8008812cdeef600d6b2db8 - this looks correct.
Tree-SHA512: b7ab61e19a54597a8fbd1844b9cfaef78879e53b882eefe4e0140fa115674df7f061e468835186963b89c963244a17d922f2ad0829b10f62b84f02019ee33edb
8578c6fccd11404412d2c60f9bede311b79ca0d0 build: Fix search for brew-installed BDB 4 on OS X (Glenn Willen)
Pull request description:
~~NOTE: This PR contains one important fix that I need (to make Bitcoin Core build cleanly on my system without shenanigans), plus some related general cleanup that is not really necessary, and could be annoying. (I am prepared to defend my argument that BDB_CFLAGS is wrong here, and BDB_CPPFLAGS is right, but this could bite anybody who has gotten in the habit of -- or scripted -- setting the former.)~~
Ok, I have been convinced that I was too clever with the refactor and I have removed it. Now it's just the tiny change to fix the build on my local machine.
---
On OS X, when searching Homebrew keg-only packages for BDB 4.8, if we find it,
use BDB_CPPFLAGS and BDB_LIBS instead of CFLAGS and LIBS for the result. This
is (1) more correct, and (2) necessary in order to give this location
priority over other directories in the include search path, which may include
system include directories with other versions of BDB.
ACKs for top commit:
theuni:
ACK 8578c6fccd11404412d2c60f9bede311b79ca0d0.
Tree-SHA512: a28f48fc81a25736f7e77c663f21cd9a6ae1cd115682031c5aa695c94cb5afa11920330a60cd6a54832822a2aec1eb23123ac2e2dcd4f0b3835aef9c9339ac97
The way we currently test with AC_CHECK_DECLS do not work with Clang:
```bash
configure:21492: clang++-10 -std=c++11 -c -g -O2 -DHAVE_BUILD_INFO -D__STDC_FORMAT_MACROS conftest.cpp >&5
conftest.cpp💯10: error: builtin functions must be directly called
(void) __builtin_clz;
^
1 error generated.
```
This also removes the __builtin_clz() check, as we don't actually use
it anywhere, and it's trvial to re-add detection if we do start using
it at some point.
fa32adf9dc25540ad27f5b82654c7057d7738627 scripted-diff: TxoutType C++11 scoped enum class (MarcoFalke)
fa95a694c492b267e4038674fd3f338dd215ab48 doc: Update outdated txnouttype documentation (MarcoFalke)
fa58469c770d8c935a86462634e4e8cd806aa6e3 rpc: Properly use underlying type in GetAllOutputTypes (MarcoFalke)
fa41c657022b8f99c8e6718a0e33c5838c412a0b rpc: Simplify GetAllOutputTypes with the Join helper (MarcoFalke)
Pull request description:
Non-scoped enums can accidentally and silently decay into an integral type. Also, the symbol names of the keys are exported to the surrounding (usually global) namespace.
Fix both issues by switching to an `enum class TxoutType` in a (mostly) scripted-diff.
ACKs for top commit:
practicalswift:
ACK fa32adf9dc25540ad27f5b82654c7057d7738627 -- patch looks correct
hebasto:
re-ACK fa32adf9dc25540ad27f5b82654c7057d7738627, since fa5997bd6fc82e16b597ea96e3c5c665f1f174ab (https://github.com/bitcoin/bitcoin/pull/19114#pullrequestreview-421425198) rebased only (verified with `git range-diff`).
Tree-SHA512: f42a9db47f9be89fa4bdd8d2fb05a16726286d8b12e3d87327b67d723f91c7d5a57deb4b2ddae9e1d16fee7a5f8c00828b6dc8909c5db680fc5e0a3cf07cd465
6d35d0d18f8d534879e982707d01ea5c1bd13b33 doc: add release note for -getinfo displaying multiwallet balances (Jon Atack)
Pull request description:
Release note for #18594. This is one of the commits from #19089, which had one concept ACK and approach ACK since late May. It seems better to submit the changes atomically.
Top commit has no ACKs.
Tree-SHA512: 38616d14b02c39f4ee4b93bf14f72043423cef177b595e85181bc9dc610fbe19d8271f2d2c9e5e17bb46423ffe27746e8e510b13a23ae6fd0e5bc4418a00dafa
0ac09c9793cd6d25ef6df14d74fb960529e1b4e3 qt: Do not truncate node flag strings in debugwindow.ui peers details tab. (saibato)
Pull request description:
Fix: When fiddling around with new node flags other than the usual.
I saw that not all possible node flag strings i.e. the UNKNOWN[..] where
visible in peers details tab.
Since v18.2 fixed size was set to 300 and sliding is thereby limited.
A fix on my old linux cruft and small screen was to set minimumSize width to -1 or 0.
Qt will then autosize the slider to the max string length.
Thereby i had full display of all flags inclusive sliding without to fullscreen the window.
Not sure if this is even an issue for those who can afford big screens or high res macs?
Feedback welcome.
BTW: nice side effect now again easy to scroll trough long version names of the node.
can't wait to see strings like /Satoshi:0.23.99/NOX2NOX4NOX32 or what ever fits in the version string.
ACKs for top commit:
hebasto:
ACK 0ac09c9793cd6d25ef6df14d74fb960529e1b4e3, tested on Linux Mint 20 (x86_64, Qt 5.12.8).
promag:
Tested ACK 0ac09c9793cd6d25ef6df14d74fb960529e1b4e3 on macos.
Tree-SHA512: a1601b5e35f10b1fd9407b28142ca00c1b985a822be5d23be4d7d3376211450f06e17f962c44b8b40977f8f8bbbb701cac1c5abb4afb3618da76385dfac848a3
d906aaa117e337fc70575beecc0d6da314f57385 qt: Fix regression in TransactionTableModel (Hennadii Stepanov)
Pull request description:
Since https://github.com/bitcoin/bitcoin/pull/17993 a crash is possible on exit.
Steps to reproduce:
- precondition: the old chain
- start `bitcoin-qt`
- wait until sync
- on main window: Menu -> File -> Quit
- crash
This PR is based on ryanofsky's [suggestion](https://github.com/bitcoin-core/gui/issues/7#issuecomment-646639251).
Fixes#7.
ACKs for top commit:
promag:
Code review ACK d906aaa117e337fc70575beecc0d6da314f57385.
ryanofsky:
Code review ACK d906aaa117e337fc70575beecc0d6da314f57385. Only changes are squashing, adding assert and adding const
vasild:
ACK d906aaa1
Tree-SHA512: 99a475fd90dff50407a58537fdc6099a2a074018e9078452bf86defc1a4b9e546aa94f916d242355900b21638c6cfef845598a5282661a9343556c4514eb155f
1087807b2bc56b9c7e7a5471c83f6ecfae79b048 tests: Provide main(...) function in fuzzer (practicalswift)
Pull request description:
Provide `main(...)` function in fuzzer. Allow building uninstrumented harnesses with only `--enable-fuzz`.
This PR restores the behaviour to how things worked prior to #18008. #18008 worked around an macOS specific issue but did it in a way which unnecessarily affected platforms not in need of the workaround :)
Before this patch:
```
# Build uninstrumented fuzzing harness (no libFuzzer/AFL/other-fuzzer-instrumentation)
$ ./configure --enable-fuzz
$ make
CXXLD test/fuzz/span
/usr/lib/gcc/x86_64-linux-gnu/7/../../../x86_64-linux-gnu/Scrt1.o: In function `_start':
(.text+0x20): undefined reference to `main'
collect2: error: ld returned 1 exit status
Makefile:7244: recipe for target 'test/fuzz/span' failed
make[2]: *** [test/fuzz/span] Error 1
make[2]: *** Waiting for unfinished jobs....
$
```
After this patch:
```
# Build uninstrumented fuzzing harness (no libFuzzer/AFL/other-fuzzer-instrumentation)
$ ./configure --enable-fuzz
$ make
$ echo foo | src/test/fuzz/span
$
```
The examples above show the change in non-macOS functionality. macOS functionality is unaffected by this patch.
ACKs for top commit:
MarcoFalke:
ACK 1087807b2bc56b9c7e7a5471c83f6ecfae79b048
Tree-SHA512: 9c16ea32ffd378057c4fae9d9124636d11e3769374d340f68a1b761b9e3e3b8a33579e60425293c96b8911405d8b96ac3ed378e669ea4c47836af06892aca73d
25dac9fa65243ca8db02df22f484039c08114401 doc: add release notes for explicit fee estimators and bumpfee change (Karl-Johan Alm)
05227a35545d7656450874b3668bf418c73813fb tests for bumpfee / estimate_modes (Karl-Johan Alm)
3404c1b753432c4859a4ca245f01c240610a00cb policy: optional FeeEstimateMode param to CFeeRate::ToString (Karl-Johan Alm)
6fcf4484302d13bd7739b617470d8c8e31974908 rpc/wallet: add two explicit modes to estimate_mode (Karl-Johan Alm)
b188d80c2de9ebb114da5ceea78baa46bde7dff6 MOVEONLY: Make FeeEstimateMode available to CFeeRate (Karl-Johan Alm)
5d1a411eb12fc700804ffe5d6e205234d30edd5f fees: add FeeModes doc helper function (Karl-Johan Alm)
91f6d2bc8ff4d4cd1b86daa370ec9d2d9662394d rpc/wallet: add conf_target as alias to confTarget in bumpfee (Karl-Johan Alm)
69158b41fc488e4f220559da17a475eff5923a95 added CURRENCY_ATOM to express minimum indivisible unit (Karl-Johan Alm)
Pull request description:
This lets users pick their own fees when using `sendtoaddress`/`sendmany` if they prefer this over the estimators.
ACKs for top commit:
Sjors:
re-utACK 25dac9fa65: rebased, more fancy C++,
jonatack:
ACK 25dac9fa65243ca8db02df2 I think this should be merged after all this time, even though it looks to me like there are needed follow-ups, fixes and test coverage to be added (see further down), which I don't mind helping out with, if wanted.
fjahr:
Code review ACK 25dac9fa65243ca8db02df22f484039c08114401
Tree-SHA512: f31177e6cabf3187a43cdfe93477144f8e8385c7344613743cbbd16e8490d53ff5144aec7b9de6c9a65eb855b55e0f99d7f164dee4b6bf3cfea4dce51cf11d33
39d526bde48d98af4fa27906e85db0399b6aa8b1 test: Bump linter versions (Duncan Dean)
Pull request description:
As per #19346, `mypy==0.700` was incompatible with Python 3.8.
I've bumped the versions of all the linters to their latest stable versions.
Checked with both Python 3.7 and 3.8 and everything still seems to work fine.
ACKs for top commit:
hebasto:
ACK 39d526bde48d98af4fa27906e85db0399b6aa8b1, I have reviewed the code and it looks OK, I agree it can be merged.
Tree-SHA512: f3ee7fda8095aa25aa68685e863076d52a6b82649770d24b0064d652763c0ceb8ebcbf9024fc74fca45c754e67b2a831dd070b3af23bc099140e6d27e89a5319
13076867981ab36b3549ab4c29583ae8ed12a709 refactor: Use Mutex type for g_cs_recent_confirmed_transactions (Hennadii Stepanov)
Pull request description:
No need the `RecursiveMutex` type for the `g_cs_recent_confirmed_transactions`.
Related to #19303.
ACKs for top commit:
MarcoFalke:
ACK 13076867981ab36b3549ab4c29583ae8ed12a709
vasild:
ACK 13076867
Tree-SHA512: 67f1be10c80ec18d0f80b9f5036e5a20986314da9b9364ef4e193ad1d9f3f4c8e4c2e16253ca79d649ff602d5b8c2aff58d7dd1085841afb760479a4875cffbe
67bb7be864f38ef5afc731aa427146cb2af500dd tests: Add fuzzing harness for CHash{160,256}, C{HMAC_,}SHA{1,256,512}, CRIPEMD160, CSipHasher, etc. (practicalswift)
Pull request description:
Add fuzzing harness for `CHash{160,256}`, `C{HMAC_,}SHA{1,256,512}`, `CRIPEMD160`, `CSipHasher`, etc.
See [`doc/fuzzing.md`](https://github.com/bitcoin/bitcoin/blob/master/doc/fuzzing.md) for information on how to fuzz Bitcoin Core. Don't forget to contribute any coverage increasing inputs you find to the [Bitcoin Core fuzzing corpus repo](https://github.com/bitcoin-core/qa-assets).
Happy fuzzing :)
Top commit has no ACKs.
Tree-SHA512: 5377b361097211a7d0b90a26ed1c6dadb9ecce11349036d19f8c9ad2818cd98709bbcbf1c2361dd18eae122b8dbce1c71bb5aa2e85660677e235b8974ae33fcc
fa8337fcdbcab8368d5bf111c4b2e5acf25e6e22 clang-format scheduler (MarcoFalke)
fa3d41b5ab8fe62bc2e36c3e1211aace69e6da65 doc: Switch scheduler to doxygen comments (MarcoFalke)
fac43f9889f500bcb62d830c030dec42fe791031 scheduler: Replace stop(true) with StopWhenDrained() (MarcoFalke)
fa9cca0550f3d0ee8c276146f40007f76fbb97c2 doc: Remove unused documentation about unimplemented features (MarcoFalke)
fab2950d703217ec34b27e677e4f33ebbf99ca08 doc: Switch boost::thread to std::thread in scheduler (MarcoFalke)
fa9819695aac260be0ba170eb15ecba8cb519843 test: Remove unused scheduler.h include from the common setup (MarcoFalke)
fa609c4f76f215c19ea4021e78c102dee2b8c3d1 scheduler: Remove unused REVERSE_LOCK (MarcoFalke)
Pull request description:
This accumulates a bunch of cleanup that was long overdue, but I haven't yet gotten around to address. Specifically, but not limited to:
* Remove unused code, documentation and includes
* Upgrade to doxygen documentation
Please refer to the individual commits for more details.
ACKs for top commit:
jnewbery:
utACK fa8337fcdbcab8368d5bf111c4b2e5acf25e6e22
Tree-SHA512: 0c825ad9767e2697a3ef1ec1be13fdc2b18eeb7493ad0be5b65cc9f209391e78b17ee66e35e094c5e171c12b0f1624f287a110f6bddaf3024b708877afa8552e
3f686d1a286f88af7baa17c0c9dc439559a66e81 ci: Run fuzz testing test cases (bitcoin-core/qa-assets) under valgrind to catch memory errors (practicalswift)
Pull request description:
Re-introduce the Travis valgrind fuzzing job which was removed by PR #18899. The removal seems to have been made by accident since the removed job does not appear to be the source of the problem the PR set out to fix.
---
Run fuzz testing [test cases (bitcoin-core/qa-assets)](https://github.com/bitcoin-core/qa-assets) under `valgrind`.
This would have caught `util: Avoid potential uninitialized read in FormatISO8601DateTime(int64_t) by checking gmtime_s/gmtime_r return value` (#18162) and similar cases.
This fuzzing job was introduced in #18166.
Top commit has no ACKs.
Tree-SHA512: 6e2681eb0ade6af465c5ea91ac163a337465d2130ec9880ba57a36d9af7c25682734586a32977dc25972d4f78483f339d680ea48c0ae13cf1dfa52b617aae401
37ae687f95c82f2d64ed880533d158060d4fc3de Add tests for CPubKey serialization/unserialization (Elichai Turkel)
9b8907faded8e4ec312c0dd4b4b15e1793876acd Check size after Unserializing CPubKey (Elichai Turkel)
Pull request description:
Found by practicalswift, closes#19235
Currently all the public API(except the pointer-like API) in CPubKey that sets/constructs a pubkey goes through `CPubKey::Set` which checks if that the length and size match and if not invalidates the key.
This adds the same check to `CPubKey::Unserialize`, sadly I don't see an easy way to just push this to the existing checks in `CPubKey::Set` but it's only a simple condition.
The problem with not invalidating is that if you write a pubkey like: `{0x02,0x00}` it will think the actual length is 33(because of `size()`) and will access uninitialized memory if you call any of the functions on CPubKey.
ACKs for top commit:
practicalswift:
re-ACK 37ae687f95c82f2d64ed880533d158060d4fc3de
jonatack:
Code review re-ACK 37ae687 per `git diff eab8ee3 37ae687` only change since last review at eab8ee3 is passing the `pubkey` param by reference to const instead of by value in `src/test/key_tests.cpp::CmpSerializationPubkey`
MarcoFalke:
ACK 37ae687f95c82f2d64ed880533d158060d4fc3de
Tree-SHA512: 30173755555dfc76d6263fb6a59f41be36049ffae7b4e1b92b922d668f5e5e2331f7374d5fa10d5d59fc53020d2966156905ffcfa8b8129c1f6d0ca062174ff1
If the socket is tearing down macOS will return EPROTOTYPE instead of EPIPE.
Because python doesn't handle this internally we have to do a workaround and retry the request.
See https://bugs.python.org/issue33450
libevent uses getaddrinfo when available, and falls back to gethostbyname
Windows has both, but gethostbyname only supports IPv4
libevent fails to detect Windows's getaddrinfo due to not including the right headers
This patches libevent's configure script to check it correctly
