BSD Commit Log Search

   Mention EMULFKEYS (mac68k).

   Document EMULFKEYS for mac68k.

   Add commented out EMULFKEYS for emulating function keys on mac keyboards
   without them.

   Map mode switch to "Command" for US keyboards.

   This was necessary as mapping Command to mode swtich with wsconsctl discards
   the Cmd1 setting and thus cannot be used to switch terminals.

   Support emulated function keys.

   As posted to port-mac68k:
   https://mail-index.netbsd.org/port-mac68k/2026/05/16/msg001080.html

   Support for raw mode mostly from code used with macpcc.

   As posted to port-mac68k:
   https://mail-index.netbsd.org/port-mac68k/2026/05/16/msg001080.html

   Don't unmask exceptions in init_sysregs as it is way too early.

   Remove the unnecessary save/restore of register to/from the stack while
   I'm here.

   s/secotr/sector/ in error message.

   fix typos in comments.

   remove extra semicolons.

   Don't use $Mdocdate$. It breaks reproducible builds

   mkimage: New `-C <config>' option.

   Enables use for out-of-tree config files.  We're not really
   supporting this as a general-purpose tool but downstream users (like
   me) may find this convenient to avoid local patches.

   ld at virtio: Guard virtio_dequeue by virtio_vq_is_enqueued.

   After triggering the DMA operation, or any previous virtio_dequeue,
   virtio_vq_is_enqueued issues the necessary bus_dmamap_sync for
   virtio_dequeue to observe any potential (new) result.

   Normally this happens inside virtio(4) (in virtio_vq_intr) between
   interrupt delivery and calling the virtqueue's done callback.  But
   polling mode I/O operations (and dump operations) don't take that
   path, so it is necessary to call virtio_vq_is_enqueued explicitly.

   PR kern/60182: ld at virtio sometimes hangs up

   virtio(4): Add missing BUS_DMASYNC_PREREAD operations.

   And one missing BUS_DMASYNC_POSTREAD operation.

   With this change, loads from vq->vq_used->flags (which occur
   immediately after a transfer is submitted to test whether we need to
   kick the host device) is separated by a PREREAD/POSTREAD cycle from
   loads from vq->vq_used->idx (which occur when we think a transfer may
   have completed, e.g. upon receiving an interrupt, to test whether it
   has, in fact, completed).

   Additionally, with this change, consecutive loads from
   *vq->vq_avail_event are separated by a PREREAD/POSTREAD cycle.

   Should fix virtio(4) issues on m68k and other related architectures
   like mips and armv<7:

   PR kern/60144: virtio(4) cache coherence issue


    [173 lines not shown]

   Fix UVMHIST 32bit builds

   risc-v: ddb support for 'mach cpuinfo'

   Improve / Add some coments.

   turn off HAVE_ATTR_NONSTRING for clang, our clang doesn't have it.

   should fix all the llvm builds i hope.

   PR bin/60275  one more (should be the last) correction

   It wasn't possible to do this, previously, but after the 3rd
   in the sequence, we can (it should have been included with that one).

   Never even look at gotsig[signo] in a vforked child - it will
   be reflecting the state of the parent, not the child, so has
   nothing to do with the child (this wasn't true until after the
   3rd fix, which made sure that the child after a vfork doesn't
   step on the parent's memory).

   This will fix an unbelievably unlikely situation, where a
   trapped signal arrives at the parent immediately before it
   vforks() a child, which happens before the trap is executed.

   Depending upon which of the various fixes (or none of them)
   to this PR have been included various incorrect things might
   have happened - the child might have processed the signal
   trap instead of the parent, the signal might be lost, or even

    [2 lines not shown]

   x86/efiboot: Put writable variables in .data, not .text.

   And define them in the .c file where they're used; no need to have
   extra unnecessary .S stuff just to define a global variable.

   Based on patch by Paul Ripke to bootx64.efi, adapted for bootia32.efi
   too.

   PR kern/59151: efiboot hangs on new motherboard

   bump libunbound

   new unbound

   Merge changes between unbound 1.24.2 and 1.25.1

   Import unbound 1.25.1 (previous was 1.24.2)

   Bug Fixes

   Fix CVE-2026-33278, Possible remote code execution during DNSSEC validation.
   Thanks to Qifan Zhang, Palo Alto Networks, for the report.

   Fix CVE-2026-42944, Heap overflow and crash with multiple nsid, cookie,
   padding EDNS options. Thanks to Qifan Zhang, Palo Alto Networks, for the
   report.

   Fix CVE-2026-42959, Crash during DNSSEC validation of malicious content. Thanks
   to Qifan Zhang, Palo Alto Networks, for the report.

   Fix CVE-2026-32792, Packet of death with DNSCrypt. Thanks to Andrew Griffiths
   from 'calif.io' for the report.

   Fix CVE-2026-40622, "Ghost domain name" variant. Thanks to Qifan Zhang, Palo
   Alto Networks, for the report.

    [25 lines not shown]

   mi pmap: support code for HW updated referenced and modified PTE bits

   Previously the MI PMAP would rely on mod/ref emulation to update
   mdpg_attrs if a page was modified or referenced. Now support checking
   each mapping's HW modified PTE bits (and updating mdpg_attrs).

   mi pmap: aarch64: fix ref/mod emulation in pmap_fault_fixup

   If a PTE changes in pmap_fault_fixup the entry in the TLB needs
   to be invalidated.

   While I'm here use atomic_load_relaxed to fetch the PTE and fix a
   comment.

   risc-v: fix ref/mod emulation PTE handling.

   The previous code has zero chance of working and now that pmap_test_mod_ref
   exists it can prove this code is correct.

   mi pmap: provide pmap_test_mod_ref

   Sprinkle commented out PMAP_DEBUG for MI PMAP kernels

   One semicolon is enough. ;; -> ;

   aarch64: pmap: misc improvements to pmap_test_mod_ref

   - remove the need for pmap_debugva by using uvm_km_{alloc,free}
   - deactivate curlwp so the kernel pmap is always activate
   - sprinkle pmap_udpate()
    CVS: ----------------------------------------------------------------------