BSD Commit Log Search

   +cpptrace

   Initial revision

   devel/libdwarf: update to 2.3.3

   remove devel/ctftools

   The port seems no longer useful and breaks with the newest devel/libdwarf.
   Upstream is unmaintained (last commit 10 years ago) and the maintainer
   (jasper@) is unresponsive.

   ok sthen@

   sys/ufshci: match against qcom,ufshc

   Linux had drop jedec,ufs-2.0 compatibility in DTS for x1e80100 to use
   only qcom,ufshc; adjust the code to make upgrade of DTB works.

   OK: phessler@

   lang/node: Update to 24.22.3

   Fixes: lots of CVEs. Look them up at
   https://nodejs.org/en/blog/release/v24.22.3

   Cleanup entity passing and remove the path string from it

   There is no reason to split the uri into file and path anymore.
   This is only used in one place for mfts and so it is better to
   simplify all other code and adjust the generation of mft->path.
   Also the parser can use a mft_filepath function to generate the
   FS path to access files referenced by the mft.

   OK tb@

   Sort union by size since ISO C is a mine filed.

   The initalizers introduced in C99 may only initialize the first
   element of a union. If that one is not the largest member
   then you may end up with uninitalized memory. In the end the
   outcome depends on the compiler used which is just bananas.

   OK tb@

   vio(4): use the correct size of rx mbufs

   This fixes jumbo frames and on hypervisors w/o the MRG_RXBUF feature.

   ok sf@

   Update to TAP::Harness::Multiple 0.08.

   add missing BDEP on bash;

   ../../../../pobj/qman-1.5.1/qman-1.5.1/src/meson.build:86:10: ERROR: Command `/pobj/qman-1.5.1/qman-1.5.1/src/qman_tests_list.sh` failed with status 127.

   A full log can be found at /pobj/qman-1.5.1/build-i386/meson-logs/meson-log.txt

   0>

   Running command: /pobj/qman-1.5.1/qman-1.5.1/src/qman_tests_list.sh
   --- stdout ---

   --- stderr ---
   env: bash: No such file or directory

   Update to mujs 1.3.11.

   Update to upterm 0.41.1.

   Allow @/{current} to appear as window part of target, from someone in
   GitHub issue 5841.

   better COMMENT, and use MODCPAN_EXAMPLES rather than handrolling.
   based on a diff from James Cook.

   Make the ssh-agent proxy work without relying on "systemd activation".

   Stop building and shipping the ssh-agent proxy.
   It's disabled by default upstream and will probably be removed in the
   future. Instead we will use the one from gcr4.

   graphics/vtk: update to 9.7.2

   ok sthen@

   Enable 64-bit DMA mbuf support on octeon.

   Allow 64-bit DMA memory allocation with builtin controllers.

   Implement bounce buffers for octeon

   This is based on similar code on amd64, arm64, and riscv64.

   Tested by kirill@
   OK deraadt@

   rpki-client: inline the valid_aspa() in its only caller

   ok claudio

   rpki-client: replace XXX comments with a short explanation

   For aspa, roa, rsc, and spl, we always return 1 from sobj->validate(),
   which I marked with an XXX since it confused me. One reason for this is
   that in normal mode we send the object to the main process for accounting
   and insertion into the vap/vrp/vsp trees if valid. In filemode invalid
   objects still get some info printed this way.

   The parser process depends on invalid MFTs failing signed_object_parse()
   at the moment, so this should stay this way, at least for now. No idea
   whether it is deliberate or by accident that TAKs are the way they are.

   ok claudio

   now that we have lang/rust,-dev, build crubit ourselves; also fixes arm64

   Use BUS_DMA_NOWAIT consistently during node init.

   rpki-client: remove a call to x509_any_inherits()

   After checking that there are no IP resources and that there is exactly
   one non-inheriting AS resource, there is no need to check that there are
   no inheriting resources.

   ok claudio

   Fix unusual initialization pattern with if_xflags.

   rpki-client: outdated greetings from our favorite captain

   mention cert.local.pem, ok deraadt

   update to freedroid-1.9.2 moving to sdl2
   (there is now a 2.0.0 using sdl3, but we don't have sdl3-mixer yet)