BSD Commit Log Search

   _splraise(): Don't take action if ipl is less than curipl.

   Issue on ep93xx noted by Jonathan Chapman, and at91, footbridge, and
   ixp12x0 have the same problem.

   (becc is also wrong, but wrong differently.)

   Add micphy.

   From Jonathan Chapman.

   McRib^W<ddb/db_active.h> is back!

   (By popular request.)

   Rework cnmagic(9):
   - No more re-defining cn_trap(), cn_isconsole(), etc.  Intead...
   - Always provide cpu_Debugger() and console_debugger() in kern_stub.c.
   - Move cnmagic(9)-related bits from <sys/systm.h> to <dev/cons.h>.
   - Make ch_check_magic() return true/false based on "magic sequence
     matched, action taken", eliminating the cn_trapped dance many
     drivers did.
   - Add a way to explicitly set the action using an API setter, rather
     than having to define a CPP macro at the right time.
   - Update callers to always pass their cnmagic state by reference rather
     than by value (which was fine as a pure macro interface, not so much
     when function calls are involved).

   Bonvenon al NetBSD 11.99.11.

   Add AUDIO and AUDIOSBC to the list of kernels to be built for mac68k.
   Add audio to MAKEDEV.

   Delete mention of the three added test locales, they aren't something
   which any regular user ever needs to know about.

   s/peding/pending/ in comment.

   syslog.conf: Don't spam console with each failed login or ftp/http xfer.

   This just changes the default for new installations; users who want
   their console to fill up with the constant spew of ssh attacks, or
   ftp/http xfers, can just put it back the way it was before.

   PR bin/60506: default syslog.conf spams console with ssh/http attacks

   None of the remaining zs_abort() functions use the argument they're
   passed, so don't bother having one.

   G/C unused zs_abort().

   None of the various zskgdb() functions don't use the argument passed
   to it, so don't bother having one.

   Shuffle the pmap_bootmap flags bits around so that the ones that
   correspond directly to PTE bits have the same value as the PTE
   bits.

   fix getopt(3) string for -T which I broke when I rewrote mdsetimage to use
   bfd (Benjamin Lorenz)

   fix getopt(3) string for -b flag that takes an argument (Benjamin Lorenz)

   Fix route -S (print flags aligned) which broke when I merged the route
   printing code in netstat and route. (Benjamin Lorenz)

   Add a provision for specifying the 68040/68060 U0 and U1 page attributes
   in machine_bootmap[].  This is needed for the Atari port, where the Milan
   uses U0 to map some of the I/O region in order to enable big-endian access.

   change improper use of err(3) into errx(3) (Benjamin Lorenz)

   Garbage-collect unused zs_abort() functions.  These particular instances
   have been unused since the introduction of cnmagic.

   PR/60976: Bobby Bingham: compat_linux does not recognize musl dynamic linker

   What OF_package_to_path() returns isn't necessarily NUL terminated...

   doc: two weeks of changes

   wsdisplay: update scr_flags properly even after WSDISPLAYIO_MODE_DUMBFB

   Fixes PR/60948 (WSDISPLAYIO_GMODE returns wrong value once
   MODE_DUMBFB is set).

   Should be pulled up to netbsd-11 and netbsd-10.

   Fix build with MKHTML=no.

   Tested both cases from scratch.

   atari: split out STE and PTE initialization into pmap_bootstrap1()

   This prepares for the transition to the m68k NEW_PMAP.
   Tested on TT030 (both RELOC enabled and disabled) and Milan.

   atari: move crash dump functions to machdep.c as other m68k ports

   For now header initialization is not changed, and using temporary
   local function declarations until future cleanup on NEW_PMAP transition.

   atair: add __attribute__((no_instrument_function)) to bootstrap functions

   atari: remove an unused and unnecessary debug function

   atari: add a missed out of bounds check in memory accouting for Milan

   Fortunately currently harmless because usable_segs[NMEM_SEGS].start
   is boot_segs[0].start and it's always zero.

   amiga,atari: avoid left-shift signed values on calculating bitmap values

   atari: explicitly regen config files from GENERIC.in rev 1.129