💬 fanquake commented on pull request "build: enable libc++ and libstdc++ hardening":
(https://github.com/bitcoin/bitcoin/pull/31424#discussion_r1920209528)
libstdc++

💬 fanquake commented on pull request "build: Make config warnings fatal if -DWERROR=ON":
(https://github.com/bitcoin/bitcoin/pull/31665#issuecomment-2598403146)
> Other projects use the same or similar build option only to turn compiler's warnings into errors.

Sure, but should that dictate what we do? Having an option to turn warnings into errors, that turns slightly more warnings into errors compared to other projects, and is mostly used in our CI where we want to surface all warnings, seems reasonable. Given that CMake lacks any native functionality to acheive the same thing, using `-DWERROR` seems ok, especially if the alternative is implement `N
...

💬 fanquake commented on pull request "guix: use GCC 13 to build releases":
(https://github.com/bitcoin/bitcoin/pull/29881#issuecomment-2598407008)
Guix Build (x86_64):
```bash
a22efc082966018f6ddd17d1a7092ffda56900895dd4694080244257e1c5c974 guix-build-6f0513613a98/output/aarch64-linux-gnu/SHA256SUMS.part
ad65ac733f93931c5fcfa09afc2a1f8bed76cd5280be9b569ff9bdf15dec3abe guix-build-6f0513613a98/output/aarch64-linux-gnu/bitcoin-6f0513613a98-aarch64-linux-gnu-debug.tar.gz
abc9037335e719927525c0474a7959db57a6f3f64d9522474ede4d88e721f73c guix-build-6f0513613a98/output/aarch64-linux-gnu/bitcoin-6f0513613a98-aarch64-linux-gnu.tar.gz
fdb0186
...

🚀 fanquake merged a pull request: "depends: Use base system's `sha256sum` utility on FreeBSD"
(https://github.com/bitcoin/bitcoin/pull/31626)

🚀 fanquake merged a pull request: "cmake: Always provide `RPATH` on NetBSD"
(https://github.com/bitcoin/bitcoin/pull/31543)

💬 maflcko commented on pull request "ci: Skip read-write of default env vars":
(https://github.com/bitcoin/bitcoin/pull/31678#discussion_r1920262852)
For reference, the exclusion is still needed to avoid the error from https://github.com/bitcoin/bitcoin/pull/27739#issuecomment-1564122573. Diff to reproduce the error:


```diff
diff --git a/ci/test/02_run_container.sh b/ci/test/02_run_container.sh
index 4181e68..168ab51 100755
--- a/ci/test/02_run_container.sh
+++ b/ci/test/02_run_container.sh
@@ -12,7 +12,7 @@ set -o errexit -o pipefail -o xtrace
if [ -z "$DANGER_RUN_CI_ON_HOST" ]; then
# Export all env vars to avoid missing som
...

💬 scgbckbone commented on pull request "Implement BIP 370 PSBTv2":
(https://github.com/bitcoin/bitcoin/pull/21283#issuecomment-2598502303)
Coldcard Mk4/Q also support PSBT v2 if you need to test ( from October 2023 https://github.com/Coldcard/firmware/blob/master/releases/History-Mk4.md#520---2023-10-10)

but as @bigspider, I also used core PR to implement this

💬 maflcko commented on pull request "ci: Turn CentOS task into native one":
(https://github.com/bitcoin/bitcoin/pull/31651#discussion_r1920283682)
thx, fixed in https://github.com/bitcoin/bitcoin/pull/31593

👋 maflcko's pull request is ready for review: "ci: Bump centos stream 10"
(https://github.com/bitcoin/bitcoin/pull/31593)

💬 maflcko commented on pull request "ci: Bump centos stream 10":
(https://github.com/bitcoin/bitcoin/pull/31593#issuecomment-2598504619)
> Upstream tracking issue for branching and adding `dash` into EPEL 10 is here: https://bugzilla.redhat.com/show_bug.cgi?id=2335416

Thx, used ksh for now (temporarily)

💬 hebasto commented on pull request "ci: Bump centos stream 10":
(https://github.com/bitcoin/bitcoin/pull/31593#issuecomment-2598523720)
Concept ACK.

💬 fanquake commented on pull request "Update leveldb subtree to latest upstream":
(https://github.com/bitcoin/bitcoin/pull/31671#issuecomment-2598526399)
> Could you share more about how these changes were selected.
> might still be worth considering:

They were PR'd, with some rationale, to https://github.com/bitcoin-core/leveldb-subtree/. This PR is just updating to the current state of the subtree, but if you think there are other worthwhile changes that should be looked at, I'd suggest opening a PR upstream for discussion (that would have to be the case anyways, before they could be part of this PR).

💬 instagibbs commented on pull request "test: add coverage for immediate orphanage eviction case":
(https://github.com/bitcoin/bitcoin/pull/31628#issuecomment-2598536562)
pushed what I think the fix is: make sure we sync the node with a ping/pong before bumping the mocktime.

💬 ryanofsky commented on issue "multiprocess: `ipc_tests` fail on *BSD":
(https://github.com/bitcoin/bitcoin/issues/31618#issuecomment-2598540264)
Thanks, do we know yet if these errors happen reliably or are random? Do you know what the easiest way I might be able to reproduce this is? Is this a custom CI setup? Is there a way to run this CI job in a VM in on linux?

👋 theuni's pull request is ready for review: "init: Lock blocksdir in addition to datadir"
(https://github.com/bitcoin/bitcoin/pull/31674)

👍 ryanofsky approved a pull request: "cmake: Set top-level target output locations"
(https://github.com/bitcoin/bitcoin/pull/31161#pullrequestreview-2559235706)
Code review ACK 186680acef661bd139bf3d16f494365ea55993b5. Since last review just reverted CMAKE_MODULE_PATH change, removed duplicate code and changed comment as suggested. Looks good!

💬 hebasto commented on issue "multiprocess: `ipc_tests` fail on *BSD":
(https://github.com/bitcoin/bitcoin/issues/31618#issuecomment-2598589382)
> Thanks, do we know yet if these errors happen reliably or are random?

They seem intermittent. I failed to reproduce them on my local NetBSD and FreeBSD installations.

> Do you know what the easiest way I might be able to reproduce this is? Is this a custom CI setup?

Yes. It is a custom CI setup based on https://github.com/vmactions.

> Is there a way to run this CI job in a VM in on linux?

It is possible to run FreeBSD on Cirrus CI though. And @maflcko [runs](https://cirrus-ci.com/task/563
...

💬 furszy commented on pull request "Added rescan option for import descriptors":
(https://github.com/bitcoin/bitcoin/pull/31668#discussion_r1920344559)
This overrides the flag for each imported descriptor, meaning the final value depends solely on the last imported descriptor.

💬 maflcko commented on pull request "ci: Enable DEBUG=1 for one GCC-12+ build to catch 117966 regressions":
(https://github.com/bitcoin/bitcoin/pull/31522#issuecomment-2598598297)
Ok, slapped it on to the centos task, which happens to work now that the 32-bit stuff is dropped.

💬 maflcko commented on issue "multiprocess: `ipc_tests` fail on *BSD":
(https://github.com/bitcoin/bitcoin/issues/31618#issuecomment-2598607148)
> It is possible to run FreeBSD on Cirrus CI though. And [@maflcko](https://github.com/maflcko) [runs](https://cirrus-ci.com/task/5637806506115072) it :)

Though, the ipc_tests are not enabled there (yet?)