BSD Commit Log Search

   Fix on archs without OCaml native support

   While here:
   1. drop the chunk trying to set OCAML_NATIVE in the build environment,
   likely taken from devel/ocaml-ocamlbuild but ineffective here.
   2. respect MAKE_ENV

   Remove outdated comment. BYPASSUNVEIL is used inside __pledge_open only now.

   Update libuninameslist to 20260107.

   Update to dash-0.5.13.2

   Drop COMPILER lines now that devel/libnotify caters for base-gcc

   The recent COMPILER addition in those ports was only because of
   a base-gcc limitation exposed by a devel/libinotify header. Now that the
   affected header has been fixed, tb and kmos prefer to revert the
   COMPILER additions.

   psignal(3): align the type of the signal parameter with POSIX

   POSIX uses an int for the signal number instead of an unsigned int. NetBSD
   made this change in 2010 and FreeBSD made this change in 2016. Follow suit
   with the other BSDs and match what POSIX specifies.

   This should not have an impact on the ABI and therefore no bump is needed.

   Flagged by Sortix os-test.

   ok deraadt@, millert@

   Update dpic to 2025.08.01.

   Update pv to 1.10.5.

   Adtapt tests of p5-OSPF-LSDB to fixed whitespace in p5-YAML-Syck-1.41
   yaml output.

   trim stray space in USE_NOBTCFI-aarch64

   update p5-YAML-Syck to 1.41
   CVE-2026-4177

   Update elfcat to 0.1.10.

   */*: fix build with gettext 1.0

   devel/gettext: update to 1.0

   Incremental changes only.

   Drop COMPILER tweak added to work around a base-gcc limitation

   As noted by claudio and tb, the offending enum is in a public header, so
   tweaking just COMPILER in devel/libnotify won't help libnotify
   consumers.  kmos then tweaked consumers and made them use ports-gcc too.
   This commit only comments the problematic __attribute__((__deprecated__))
   in the public header, said attribute isn't very useful anyway since the
   enum member containing a typo doesn't appear used in the ecosystem anyway.

   ok aja@

   There may be a followup commit if people feel like moving back consumers
   to a default COMPILER line makes sense, mail to follow.

   www/pipe-viewer: Update to 0.5.7

   ok sthen@

   textproc/delta: Update to 0.19.1

   From Maintainer, thanks

   update p5-JSON to 4.11

   update to ccache-4.13.2

   math/med: add missing dependency on fortran

   Update and bump cad/freecad (math/med's only consumer) accordingly.

   Reported by naddy@ and ok tb@

   refereced -> referenced; ok bluhm@

   update to ruff-0.15.7, from maintainer Laurent Cheylus
   switch back to a do-build step because overriding MODCARGO_CARGOTOML
   doesn't work for modcargo-gen-crates. I tweaked his diff slightly to avoid
   setting MODCARGO_CARGOTOML to the default and adjusted a comment.

   use maturin to build, ok maintainer

   isolaved -> isolated; ok dlg@

   Check if host keys exist before adding them, and expand on the warning
   about modifying the system config.

   ENONBLOCK -> EWOULDBLOCK, former does not exist
   ok tb@

   Add special handling of TEST_SSH_HOSTBASED_AUTH=setupandrun.

   This will MODIFY THE CONFIG OF THE SYSTEM IT IS RUNNING ON to enable
   hostbased authentication to/from itself and run the hostbased tests.  It
   won't undo these changes, so don't do this on a system where this matters.

   Check lastgc is not NULL before using it, GitHub issue 4935 from Pavel
   Lavrukhin.

   Fix a NULL dereference and use after free, GitHub issue 4936 from Pavel
   Lavrukhin.

   Use window options for cursor-style to avoid crash when no pane, from
   Arden Packeer in GitHub issue 4942.