similar to em(4), do not deallocate the rings in DVACT_SUSPEND, but
pesist them through the suspend/resume, and reuse in DVACT_RESUME
(to avoid sleeping allocations which are not allowed)
ok jan
add addresses to log lines
this should help stateless log ingesters to correlate a source to a
failure.
from -portable, by Pietro Cerutti (gahr at gahr.ch)
the old "int softonly" flag has lost meaning, and a better name is
"int detaching". when detaching, the rings are deallocated. when suspending,
the rings are not detached, so that resume can reuse them (and not sleep)
jan liked the variable name change
spell HOST_NAME_MAX with some _POSIX_ vibes
On OpenBSD this is a complete no-op. no text changes.
On -portable though, this is going to help *a lot*. The ecosystem is
not really providing a sane HOST_NAME_MAX, and I don't see any reason
why this could improve anytime soon.
Discussed many times with different folks, I think the time has finally
come.
Remove prboom and prboom-plus.
While these are of some historical interest, they no longer have active
upstreams. There are many Doom engines; users of these should consider
dsda-doom, prboomx, or one of the others.
ok Ryan Freeman (prboom-plus MAINTAINER)
also ok tb@
Removal suggested by Lucas de Sena
Fix race where TCP syn cache could free a used route.
Route timer code is not MP safe and relies on netlock. As
syn_cache_put() may free a route, syn_cache_timer() has to hold
shared netlock until syn_cache_put() has finished.
from OpenAI; OK mvs@ deraadt@
Do not advance snd_nxt and snd_max for TCP SACK retransmits.
In the SACK case, the calculation that decided whether we should
adjust snd_nxt and snd_max was wrong. Use the same logic as FreeBSD
and NetBSD, and never advance them when retransmitting a SACK hole.
This may delay a FIN packet in a corner case, but the logic gets
much simpler. Also simplify the SACK checks, sack_rxmit is only
set if tp->sack_enable is true.
reported by OpenAI; OK sashan@
Import msedit-2.0.0.
msedit is a text editor that pays homage to the classic MS-DOS Editor, but
with a modern interface and input controls similar to VS Code. The goal is
to provide an accessible editor that even users largely unfamiliar with
terminals can easily use.
based on a submission from David Uhden Collado.
ok jan@
wayland/wf-shell: update to 0.11.0.
see https://github.com/WayfireWM/wf-shell/releases/tag/v0.11.0
migrated to gtk+4, which brings issues with GtkGLArea - doesn't work
with our mesa that isnt linked against wayland.
bring back the cairo codepaths from the Gtk+3 version to load pixbufs to
background.
dont build the screen locker, needs pam and ucred (and other funky
dependencies).
lang/gcc/16: drop USE_NOEXECONLY=Yes
We don't need it, it builds on aarch64, arm64, arm, i386, mips64, powerpc,
powerpc64, riscv64, sparc64, and alpha which was tested by martijn@
rtwn: send complete H2C commands to the firmware
Since urtwn(4) moved to the shared rtwn_fw_cmd() in 2016, only the
command id byte reached the H2C mailbox and the payload was zero
(RSSI 0, MACID_CONFIG rate mask 0). Write the full 16/32-bit words
again, as urtwn(4) did before.
For RTL8192C-family chips, lay out extended commands as the firmware
expects, [id, b2, b3, b4] + [b0, b1], matching FreeBSD r92c_fw_cmd()
and Linux rtl8xxxu.
Patch by Mikolaj Kucharski
