💬 brunoerg commented on pull request "fuzz: fix bad alloc in connman target":
(https://github.com/bitcoin/bitcoin/pull/31235#discussion_r1836702319)
> I think fuzzing in range between 0 and 100 would make sense - after all that's what the function is designed to handle

I agree.

💬 fanquake commented on pull request "test: clarify log messages when handling SOCKS5 proxy connections":
(https://github.com/bitcoin/bitcoin/pull/31239#issuecomment-2468267372)
> While we're updating this file,

Pulled this into another branch.

🚀 fanquake merged a pull request: "test: clarify log messages when handling SOCKS5 proxy connections"
(https://github.com/bitcoin/bitcoin/pull/31239)

✅ fanquake closed a pull request: "doc: fix typos"
(https://github.com/bitcoin/bitcoin/pull/31253)

💬 fanquake commented on pull request "doc: fix typos":
(https://github.com/bitcoin/bitcoin/pull/31253#issuecomment-2468268588)
Pulled this into another branch.

👍 fanquake approved a pull request: "doc: Add missing 'blank=true' option in offline-signing-tutorial.md"
(https://github.com/bitcoin/bitcoin/pull/31237#pullrequestreview-2427391108)
ACK ec375de39ff8e0d44fc026618a234e37019e46c1

🚀 fanquake merged a pull request: "doc: Add missing 'blank=true' option in offline-signing-tutorial.md"
(https://github.com/bitcoin/bitcoin/pull/31237)

📝 fanquake locked a pull request: "GitHub skip branch"
(https://github.com/bitcoin/bitcoin/pull/31265)
Testing https://github.com/bitcoin/bitcoin/pull/30487

📝 fanquake opened a pull request: "doc: correct typos"
(https://github.com/bitcoin/bitcoin/pull/31271)
Includes #31253.
Includes https://github.com/bitcoin/bitcoin/pull/31239#pullrequestreview-2425008603.
Fixes remaining lint output.

💬 fanquake commented on pull request "cmake: add optional source files to bitcoin_crypto directly":
(https://github.com/bitcoin/bitcoin/pull/31268#issuecomment-2468286812)
> Avoid having many static libraries

Can you expand the motivation.

👍 dergoegge approved a pull request: "validation: Remove RECENT_CONSENSUS_CHANGE validation result"
(https://github.com/bitcoin/bitcoin/pull/31269#pullrequestreview-2427403124)
ACK e80e4c6ff91e27d7d40f099a2d7942c29085234c

We can always re-introduce the concept and actually make use of it if needed. For now it seems better to remove it from the interface.

💬 fanquake commented on pull request "ci: Split out native fuzz jobs for macOS and windows (take 2)":
(https://github.com/bitcoin/bitcoin/pull/31221#issuecomment-2468288385)
Concept ACK.

💬 l0rinc commented on pull request "doc: correct typos":
(https://github.com/bitcoin/bitcoin/pull/31271#issuecomment-2468293030)
ACK 726cbee9553b25bedfef70cfd5be9f1eeec8a30d

🚀 fanquake merged a pull request: "refactor: Drop deprecated space in operator""_mst"
(https://github.com/bitcoin/bitcoin/pull/31267)

💬 brunoerg commented on pull request "fuzz: fix bad alloc in connman target":
(https://github.com/bitcoin/bitcoin/pull/31235#discussion_r1836745793)
Done.

💬 brunoerg commented on pull request "fuzz: fix bad alloc in connman target":
(https://github.com/bitcoin/bitcoin/pull/31235#issuecomment-2468299823)
Force-pushed:

- Adjusted the cap and added the `Assume` for the `max_pct`.
- Adjusted `GetAddr` and `GetAddress` documentation.
- Adjusted both connman and addrman fuzz targets to use values from 0 to 100 for `max_pct`.

Thanks @mzumsande and @vasild.

💬 l0rinc commented on pull request "fuzz: fix bad alloc in connman target":
(https://github.com/bitcoin/bitcoin/pull/31235#issuecomment-2468308284)
The title still claims "fuzz", but the fix seems to be in production code - can we update the PR title?

⚠️ maflcko opened an issue: "RFC: Adopt C++ Safe Buffers?"
(https://github.com/bitcoin/bitcoin/issues/31272)
C++ is unsafe, meaning that any code written in it may cause undefined behavior at runtime. One possible source of undefined behavior is out-of-range memory access.

While some limited compiler warnings exist to detect some obvious cases, tracking down out-of-range memory access is usually done at runtime with debugging tools such as Valgrind or Asan. However, such tools can normally not be used in production, because they are not hardening tools, see https://stackoverflow.com/a/70004411/20847
...

🤔 maflcko reviewed a pull request: "fuzz: fix bad alloc in connman target"
(https://github.com/bitcoin/bitcoin/pull/31235#pullrequestreview-2427444404)
you'll have to squash the commits. Otherwise a fuzz bisect will fail on the addrman target as well. (connman already fails)

💬 maflcko commented on pull request "fuzz: fix bad alloc in connman target":
(https://github.com/bitcoin/bitcoin/pull/31235#discussion_r1836765925)
style-nit: For new code it would be good to use `size_t{100}`, as explained in the dev notes.