💬 MarcoFalke commented on pull request "Handle invalid hex encoding in ParseHex":
(https://github.com/bitcoin/bitcoin/pull/25227#issuecomment-1457953182)
> I guess it's having issues with the template for the std::optional<std::vector<uint8_t>> instantiation, since it's not available to the linker?

It might be a compiler bug where the compiler skips the explicit instantiation of a template that has a function body (and thus is already required to be implicitly instantiated)?

> Adding explicit instantiations should fix it, I think:

Thanks. Done something like that in #https://github.com/bitcoin/bitcoin/pull/27218

> Alternatively, I su
...

💬 MarcoFalke commented on pull request "test: getblock on header throws":
(https://github.com/bitcoin/bitcoin/pull/27237#issuecomment-1481769276)
Not sure if it makes sense to add a test only. Maybe the error message can be improved along? Closing, let's continue discussion in #https://github.com/bitcoin/bitcoin/issues/20978

💬 MarcoFalke commented on issue "meta: Isolated fuzzing of net processing":
(https://github.com/bitcoin/bitcoin/issues/27502#issuecomment-1516620806)
I don't think that #https://github.com/bitcoin/bitcoin/pull/27499 is going to increase fuzz performance. Parsing should currently only be done at the beginning once, before any fuzzing happens. Even if parsing were in the hot loop, it probably wouldn't be noticeable. And if you want to construct a PeerManager for each fuzzing iteration, it seems too slow either way, unless you find a way to skip the memory allocations (as you said yourself).

💬 dergoegge commented on issue "meta: Isolated fuzzing of net processing":
(https://github.com/bitcoin/bitcoin/issues/27502#issuecomment-1516641523)
> I don't think that #https://github.com/bitcoin/bitcoin/pull/27499 is going to increase fuzz performance.

I think the separation from gArgs makes sense either way, but yea not really a performance improvement. I added the args stuff because I was investigating performance for a target that creates a new TestingSetup each iteration. The argsman setup with all arguments actually was one of the slower things (among block tree db, chainman setup and blockfile creation).

> And if you want to
...

💬 MarcoFalke commented on issue "CPU DoS on mainnet in debug mode":
(https://github.com/bitcoin/bitcoin/issues/27586#issuecomment-1544495872)
I presume this is still an issue? If yes, what about filing a new one, given that this issue is about `--enable-debug` (not `-debug`), and most comments here may be about a recently fixed issue #https://github.com/bitcoin/bitcoin/issues/27623 and not about `--enable-debug` performance?

💬 MarcoFalke commented on issue "Drop support for g++-8?":
(https://github.com/bitcoin/bitcoin/issues/27537#issuecomment-1549169559)
Closing for now. Let's continue discussion in https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109871 and #https://github.com/bitcoin/bitcoin/pull/27662

💬 MarcoFalke commented on issue "Validation of malformed address fails with a peculiar message":
(https://github.com/bitcoin/bitcoin/issues/27723#issuecomment-1559080308)
Fun fact, this also happens with the BIP 173 and BIP 350 test vectors. Proposed fix in #https://github.com/bitcoin/bitcoin/pull/27727

💬 MarcoFalke commented on pull request "test: Restore unlimited timeout in IndexWaitSynced":
(https://github.com/bitcoin/bitcoin/pull/28036#issuecomment-1623778841)
We've been plagued by intermittent timeouts locally (for months/years) and now in CI (for days), but never a thread that never finished. So I think it should be evident that this or #https://github.com/bitcoin/bitcoin/pull/28026 should be preferred. After all, it is no different than any other deadlock or infinite loop anywhere else in the code.

💬 fanquake commented on pull request "descriptors: do not return top-level only funcs as sub descriptors":
(https://github.com/bitcoin/bitcoin/pull/28067#issuecomment-1645240967)
Backported 2/3 of the commits here in #https://github.com/bitcoin/bitcoin/pull/28067.

💬 MarcoFalke commented on pull request "test: fix `feature_addrman.py` on big-endian systems":
(https://github.com/bitcoin/bitcoin/pull/27529#issuecomment-1651013529)
tested as well, so my recommendation would be to merge this before #https://github.com/bitcoin/bitcoin/pull/28087

💬 MarcoFalke commented on pull request "blockstorage: XOR blocksdir *.dat files":
(https://github.com/bitcoin/bitcoin/pull/28052#issuecomment-1660405202)
Rebased on the latest #https://github.com/bitcoin/bitcoin/pull/28060