📝 fanquake opened a pull request: "build: Remove `--enable-threadlocal`"
(https://github.com/bitcoin/bitcoin/pull/30137)
Based on #30095 and #30099.
Would close #29952.
Up for debate.

💬 fanquake commented on pull request "build: Enable `thread_local` for MinGW-w64 builds":
(https://github.com/bitcoin/bitcoin/pull/30099#issuecomment-2118613875)
> After this and https://github.com/bitcoin/bitcoin/pull/30095, is there any need to keep the thread_local option at all?

Maybe not. See #30137.

💬 laanwj commented on pull request "build: Remove `--enable-threadlocal`":
(https://github.com/bitcoin/bitcoin/pull/30137#issuecomment-2118621410)
Concept ACK.

💬 laanwj commented on pull request "wallet, tests: Avoid stringop-overflow warning in PollutePubKey":
(https://github.com/bitcoin/bitcoin/pull/30131#discussion_r1605667276)
Maybe directly assert `pubkey.size() >= 1` to clarify.

⚠️ Geremia opened an issue: "LevelDB error: Corruption: block checksum mismatch didn't trigger reindex."
(https://github.com/bitcoin/bitcoin/issues/30138)
### Is there an existing issue for this?

- [X] I have searched the existing issues

### Current behaviour

For the past 29 hours, it's been stuck trying to sync:
>**Number of blocks left** Unknown. Syncing Headers (843771, 100.0%)…

and

>Connecting to peers…

I'm also pruning (`prune=550`).

### Expected behaviour

It should sync the full chain and reindex if needed, not try to sync what it can't (due to possibly corrupt chainstate DB).

### Steps to reproduce

Open up `bitcoin-qt`. It
...

💬 hebasto commented on pull request "build: Remove `--enable-threadlocal`":
(https://github.com/bitcoin/bitcoin/pull/30137#issuecomment-2118669335)
Concept ACK.

💬 tdb3 commented on pull request "lint/contrib/doc updates":
(https://github.com/bitcoin/bitcoin/pull/30084#issuecomment-2118676930)
@jonatack @kevkevinpal, what are your thoughts on reopening PR #30106 and dropping the test runner `insuffient` spelling error from 30084?

Seems like this PR (30084) is scoped to more than just the spelling error and fixing the error could help prevent near term lint issues for open PRs.

Maybe I'm missing something and this was already addressed?

📝 fanquake unlocked a pull request: "lint: fixed typo in test_runner causing linter warning"
(https://github.com/bitcoin/bitcoin/pull/30106)
introduced in https://github.com/bitcoin/bitcoin/commit/357ad110548d726021547d85b5b2bfcf3191d7e3

This typo is causing an error in the linter, fixing this should remove this warning

```
test/functional/test_runner.py:651: insuffient ==> insufficient
^ Warning: codespell identified likely spelling errors. Any false positives? Add them to the list of ignored words in test/lint/spelling.ignore-words.txt
```

[link to cirrus ci warning](https://cirrus-ci.com/task/5926490804584448?logs=lint
...

💬 laanwj commented on pull request "util: avoid using thread_local variable that has a destructor":
(https://github.com/bitcoin/bitcoin/pull/30095#discussion_r1605695254)
Agree! i mean in a hypothetical future (say, C++30) when every compiler and OS finally implements thread-local storage correctly including destructors, it could be reconsidered, but i think a "no-thread-local" policy makes sense until then.

TBH in general i don't really like comments that suggest TODOs. It's better to keep track of those things in issues where discussion and context is visible.

💬 laanwj commented on pull request "util: avoid using thread_local variable that has a destructor":
(https://github.com/bitcoin/bitcoin/pull/30095#issuecomment-2118681020)
Code review ACK d35ba1b3f16071b8fe9b36398ba15352dbf2a54d

💬 TheCharlatan commented on pull request "kernel: De-globalize fReindex":
(https://github.com/bitcoin/bitcoin/pull/29817#discussion_r1605695740)
I ended up deleting all of this in #30132.

💬 maflcko commented on pull request "lint/contrib/doc updates":
(https://github.com/bitcoin/bitcoin/pull/30084#discussion_r1605700154)
Instead of claiming that this was suggested to be changed by a linter, it would be better to have a real motivation for the change, as well as a reason why the change is correct.

Looking at this, I don't see why continuing is the right behavior change here.

💬 maflcko commented on pull request "lint/contrib/doc updates":
(https://github.com/bitcoin/bitcoin/pull/30084#discussion_r1605699782)
nit: Shouldn't this go into the section that ends with "...could be outdated" above?

💬 maflcko commented on issue "LevelDB error: Corruption: block checksum mismatch didn't trigger reindex.":
(https://github.com/bitcoin/bitcoin/issues/30138#issuecomment-2118691502)
Can you clarify this? The debug log says "Shutdown: done", but you claim that "It connects to many peers", indicating that the program is still running?

If this is the case, it would be a bug, because the program should abort and shutdown when file corruption is detected in the leveldb database.

💬 jadijadi commented on pull request "Showing Local Addresses in Node Window":
(https://github.com/bitcoin-core/gui/pull/626#issuecomment-2118694174)
> This PR still have a couple of unaddressed comments: [#626 (comment)](https://github.com/bitcoin-core/gui/pull/626#discussion_r915334998) and [#626 (comment)](https://github.com/bitcoin-core/gui/pull/626#discussion_r1274077350).
>
> @jadijadi Are you still working on this?

Will check and comment soon

💬 maflcko commented on pull request "fuzz: add more coverage for `ScriptPubKeyMan`":
(https://github.com/bitcoin/bitcoin/pull/30134#discussion_r1605702378)
Can you explain this change?

```
wallet/test/fuzz/scriptpubkeyman.cpp should add these lines:
#include <assert.h> // for assert
#include <functional> // for function
#include <map> // for map
#include <memory> // for unique_ptr, shared_ptr, make_unique, __shared_ptr_access
#include <optional> // for optional, nullopt, nullopt_t
#include <unordered_set>
...

💬 maflcko commented on pull request "test: remove unneeded `-maxorphantx=1000` settings":
(https://github.com/bitcoin/bitcoin/pull/30133#issuecomment-2118697981)
utACK 8950053636cb38ed85fe2d58b53e5d0acb35c390

💬 maflcko commented on pull request "indexes: Don't wipe indexes again when continuing a prior reindex":
(https://github.com/bitcoin/bitcoin/pull/30132#issuecomment-2118706781)
Not sure what's wrong with the CI ( https://github.com/bitcoin/bitcoin/actions/runs/9133680294/job/25117694444?pr=30132#step:6:2 ) Maybe rebase again?

💬 brunoerg commented on pull request "fuzz: add more coverage for `ScriptPubKeyMan`":
(https://github.com/bitcoin/bitcoin/pull/30134#discussion_r1605721747)
Nevermind, just dropped this change.

💬 rkrux commented on pull request "test: improve robustness of connect_nodes()":
(https://github.com/bitcoin/bitcoin/pull/30118#discussion_r1605724296)
Hmm I see now what you mean.