BSD Commit Log Search

   A specially crafted ELF interpreter binary can trigger a KASSERT() in
   vmcmd_map_readvn().  While this points at some deficiencies in
   elf_load_file(), we have no intention to supporting such weirdly crafted
   binaries (even if they're technically valid).  So return EINVAL instead.

   ok deraadt@

   ld.bfd: add SPARC64 GOTDATA relocation support

   Tested and feedback by claudio@

   OK: kettenis@

   TIOCGSID can dereference a freed t_session, because we are storing s_leader
   in the wrong place.  This is a cold hard crash crashes.  Some folk are
   looking at fixing this properly but it will take some time, so in the meantime
   we should just fail the ioctl.
   From Acts1632, ok kettenis

   ksh: allow editing empty line with 'v' in vi-mode

   ok tb@

   Specify the correct "read" and "write" bits to the NVMMU based on the
   opcode of the NVMe command.  Drop the opcode itself, since it isn't needed.
   This allows us to drop the magic "null check" bit that is no longer
   supported by the firmware that Apple ships with newer macOS releases.
   Based on work done by Sven Peter over at Asahi Linux.

   ok jmatthew@

   Refer to id_mldsa44_ed25519.pub in the pubkey list. bz#3989.

   Fix out-of-bounds heap reads and writes in perl regex

   Perl versions through 5.45.1 have out-of-bounds heap reads and writes
   during regular expression matching via an undersized superlinear cache
   in S_regmatch

   This is CVE-2026-15534

   https://github.com/Perl/perl5/commit/568e6fd238867bb9e99fa3f47cba3169009239e0
   https://github.com/Perl/perl5/commit/54cf3d44cbbedd17d774e9a37921963e8fd5d0cb

   Add support for the Spacemit K3 PCIe controller.

   Add support for the K3 PCIe/USB3 combo PHY.

   Attach to the K3 PLL clock block.  This doesn't enable any of the clocks
   (yet), but the associated regmap is needed for the K3 PCIe/USB combo PHY.

   Prevent read buffer overflow in verbose debug logging.

   When we receive more than 8 nameservers in a DHCP lease we would
   overread the nameservers buffer because the loop was not bounded by
   items in the array but the whole size.

   Pointed out by Stuart Thomas some time ago, thanks!

   isakmpd: Handle malformed MONITOR_BIND message

   Using a stack allocated sockaddr_storage avoids the malloc(3)
   entirely.  Add length checks for the payload sent by the child.

   When a malicious child sends us a malformed message the privsep
   protocol is out of sync.  Thus terminate gracefully with log_fatalx().

   While there, attribute log_fatalx() with __dead.

   Reported by Shibo, Hugo Systopia Team

   ok markus@ tb@ deraadt@

   disable dynamic power gating on Panther Lake and later

   described in Intel changes as known to cause hangs

   tested on 0x57b3 / E1000_DEV_ID_PCH_PTP_I219_LM25

   partly from Intel's DPDK driver
   ok jmatthew@

   change K1 exit timeout on Meteor Lake and later

   without this my Panther Lake machine could not receive non-broadcast packets

   tested on 0x57b3 / E1000_DEV_ID_PCH_PTP_I219_LM25 and
   0x550b / E1000_DEV_ID_PCH_MTP_I219_V18

   partly from Intel's DPDK driver
   ok jmatthew@

   SSL_CTX_new(): fix spelling of NULL

   match Panther Lake and Wildcat Lake

   tested on 0x57b3 / E1000_DEV_ID_PCH_PTP_I219_LM25
   needs workarounds coming in another commit to be useable
   ok jmatthew@

   add a mac type for Meteor Lake

   needed for generation specific workarounds
   ok jmatthew@

   simplify mac type tests

   Reduces the number of places to change when adding mac types.
   Tested by bluhm@ in a machine full of em(4) cards, ok jmatthew@

   match on Intel Wildcat Lake

   regen

   add Intel Wildcat Lake ids

   from:
   Intel Core Processor (Series 3)
   Datasheet, Volume 1 of 2, Doc. No.: 913965, Rev.: 001

   SPECIFICATINS SPECIFACTIONS -> SPECIFICATIONS

   Speed up swap encryption/decryption by using AES instructions on amd64 and
   arm64 when available.

   ok jsing@

   fq_codel: fix CoDel drop scheduling

   Use drops - 1 as the codel_intervals index; the table starts at interval
   / sqrt(1) while cd->drops is one-based. Assert non-zero.

   Exclude the candidate packet from the non-starvation backlog check; the
   decision is about the backlog remaining after dequeue.

   Enter dropping when now equals cd->start, not only when it exceeds it.

   OK sashan@

   Add some notes about Pi 5 and the debug port, tweak suggestions about
   preferred storage, stop pointing at pftf UEFI files. Feedback kettenis@

   revert 'drm/amd/display: Fix backlight max_brightness to match exported range'

   jmc@ reports this 6.18.42 change caused a very dim display on boot and with
   display.brightness values of 97-100 on a Dell Inspiron 5505

   it looks like this patch will be reverted in linux stable branches
   https://gitlab.freedesktop.org/drm/amd/-/work_items/5562

   relayd: remove dead backup flag and token

   The idea behind it was good, but it was never fully and properly implemented.
   It's better to remove it before reimplementing the concept from scratch
   and doing it right. Reading the man page makes it a little easier.

   OK kirill@

   relayd: fix deferring connections log

   OK kirill@

   relayd: check imsg_compose_event return values

   relayd: check proc_compose return values

   Fatal on config, setup, and startup imsg send failures. Inconsistent
   state means we can't recover.

   Warn on per-request responses and sync messages. A single dropped
   message shouldn't kill the whole process.

   OK kirill@