openbsd
6 months agostart documenting ioctls
kn [Sun, 5 May 2024 19:13:13 +0000 (19:13 +0000)]
start documenting ioctls

6 months agoremove unused 0-sized softdep files
jsg [Sun, 5 May 2024 13:33:04 +0000 (13:33 +0000)]
remove unused 0-sized softdep files
ok tb@

6 months agoDocument "ls rogue" exit status.
ajacoutot [Sun, 5 May 2024 08:16:45 +0000 (08:16 +0000)]
Document "ls rogue" exit status.

6 months agoStop building with and documenting FFS_SOFTUPDATES.
jsg [Sun, 5 May 2024 07:26:58 +0000 (07:26 +0000)]
Stop building with and documenting FFS_SOFTUPDATES.
Code that used it was removed in February.

6 months agoremove unneeded includes
jsg [Sun, 5 May 2024 06:14:37 +0000 (06:14 +0000)]
remove unneeded includes

6 months agoadd upstream change to fix the build
jsg [Sun, 5 May 2024 02:55:34 +0000 (02:55 +0000)]
add upstream change to fix the build
ok tb@ deraadt@

6 months agoUpdate awk to the May 4, 2024 version.
millert [Sat, 4 May 2024 22:59:21 +0000 (22:59 +0000)]
Update awk to the May 4, 2024 version.
Fixes a use-after-free bug with ARGV for "delete ARGV".

6 months agodma_free(9) just once in error case
kn [Sat, 4 May 2024 16:40:38 +0000 (16:40 +0000)]
dma_free(9) just once in error case

buf is free'd and NULL, don't do it again to avoid a panic.

OK krw

6 months agoAdd TSO capabilites.
mbuhl [Sat, 4 May 2024 13:35:26 +0000 (13:35 +0000)]
Add TSO capabilites.
Comments, suggestions and testing from bket@, jan@, and bluhm@

6 months agoremove prototypes for removed functions
jsg [Sat, 4 May 2024 11:25:24 +0000 (11:25 +0000)]
remove prototypes for removed functions

6 months agoremove some unused defines
jsg [Sat, 4 May 2024 10:53:37 +0000 (10:53 +0000)]
remove some unused defines
ok miod@

6 months agoadd -r to "show" command; ok mlarkin
jmc [Sat, 4 May 2024 07:51:21 +0000 (07:51 +0000)]
add -r to "show" command; ok mlarkin

6 months agoamd64 is guaranteed to have PAT and CLFLUSH; stop testing cpu flags
guenther [Sat, 4 May 2024 00:27:42 +0000 (00:27 +0000)]
amd64 is guaranteed to have PAT and CLFLUSH; stop testing cpu flags
for them on that arch.

ok jsg@

6 months agoRemove disgusting NULL checks in tm_to_{gentime,utctime}()
tb [Fri, 3 May 2024 18:33:27 +0000 (18:33 +0000)]
Remove disgusting NULL checks in tm_to_{gentime,utctime}()

The only caller that could potentially call these with NULL has been fixed.
This way an ugly hack that was needed to plug a memory leak can go away and
the functions again behave as intended without OpenSSL-style workarounds.

ok beck

6 months agoIntercept a NULL s early in ASN1_TIME_set_string_internal()
tb [Fri, 3 May 2024 18:29:43 +0000 (18:29 +0000)]
Intercept a NULL s early in ASN1_TIME_set_string_internal()

If s is NULL, the only thing the tm_to_*() functions do is a check that
a GeneralizedTime has a four digit year (between 0000 and 9999) and a
UTCTime has a year between 1950 and 2050. These checks are already done
in ASN1_TIME_parse() itself: the century is 100 times a two-digit value
(or 19 in the UTCTime case) plus another two-digit value.

ok beck

6 months agoSimplify type handling in ASN1_TIME_set_string_internal()
tb [Fri, 3 May 2024 18:22:26 +0000 (18:22 +0000)]
Simplify type handling in ASN1_TIME_set_string_internal()

ASN1_time_parse() takes a mode argument. If mode != 0, there is a check
that mode is the same as the time type returned by asn1_time_parse_cbs()
otherwise ASN1_time_parse() fails. Therefore the type == mode checks in
ASN1_set_string_internal() are redundant and can be removed.

ok beck

6 months agoSimplify tm handling in ASN1_time_parse()
tb [Fri, 3 May 2024 18:15:27 +0000 (18:15 +0000)]
Simplify tm handling in ASN1_time_parse()

The CBS version asn1_time_parse_cbs() handles a NULL tm gracefully, so
there is no need to avoid it by passing a pointer to a tm on the stack.

ok beck

6 months agoPush solock() down to sosend() and remove it from soreceive() paths fro
mvs [Fri, 3 May 2024 17:43:09 +0000 (17:43 +0000)]
Push solock() down to sosend() and remove it from soreceive() paths fro
unix(4) sockets.

Push solock() deep down to sosend() and remove it from soreceive() paths
for unix(4) sockets.

The transmission of unix(4) sockets already half-unlocked because
connected peer is not locked by solock() during sbappend*() call. Use
`sb_mtx' mutex(9) and `sb_lock' rwlock(9) to protect both `so_snd' and
`so_rcv'.

Since the `so_snd' is protected by `sb_mtx' mutex(9) the re-locking
is not required in uipc_rcvd().

Do direct `so_rcv' dispose and cleanup in sofree(). This sockets is
almost dead and unlinked from everywhere include spliced peer, so
concurrent sotask() thread will just exit. This required to keep locks
order between `i_lock' and `sb_lock'. Also this removes re-locking from
sofree() for all sockets.

SB_OWNLOCK became redundant with SB_MTXLOCK, so remove it. SB_MTXLOCK
was kept because checks against SB_MTXLOCK within sb*() routines are mor
consistent.

Feedback and ok bluhm

6 months agoaucat.1: Drop the number component of sndio descriptors
ratchov [Fri, 3 May 2024 16:48:41 +0000 (16:48 +0000)]
aucat.1: Drop the number component of sndio descriptors

ok jmc

6 months agosndiod.1: Drop the number component of sndio descriptors
ratchov [Fri, 3 May 2024 16:47:15 +0000 (16:47 +0000)]
sndiod.1: Drop the number component of sndio descriptors

6 months agoMake qwx(4) call into ifmedia more like iwx(4) does it.
stsp [Fri, 3 May 2024 14:32:11 +0000 (14:32 +0000)]
Make qwx(4) call into ifmedia more like iwx(4) does it.

Fixes weird media: lines in ifconfig which show a mix of 802.11 modes
after switching APs, such as:
media: IEEE802.11 autoselect mode 11a (OFDM18 mode 11g)

6 months agovmm: merge VMX and SVM vm initialization functions.
dv [Fri, 3 May 2024 13:48:29 +0000 (13:48 +0000)]
vmm: merge VMX and SVM vm initialization functions.

They do the same uvmspace/pmap dance, so just reduce to a single
function. Kernel lock is moved to the uvmspace_free call. In addition,
we should be using the pmap mutex in pmap_convert, called by the
vm initialization routine.

ok mlarkin@

6 months agowitness: Display lock cycles longer than two locks
visa [Fri, 3 May 2024 13:47:31 +0000 (13:47 +0000)]
witness: Display lock cycles longer than two locks

When a lock order reversal is found, perform a path search in the lock
order graph. This lets witness(4) display lock cycles that are longer
than two locks.

OK mpi@

6 months agowitness: Make "show witness" display lock subtypes
visa [Fri, 3 May 2024 13:45:42 +0000 (13:45 +0000)]
witness: Make "show witness" display lock subtypes

Display lock subtypes in "show witness" output to reduce ambiguity.

OK mpi@

6 months agoAdd support for Rx checksum offloading to dwqe(4).
stsp [Fri, 3 May 2024 13:02:18 +0000 (13:02 +0000)]
Add support for Rx checksum offloading to dwqe(4).

Tested by uwe@ and myself on Elkhart Lake, and kettenis@ on RK3566.

Joint work with uwe@
ok kettenis@

6 months agosndiod: Fix SIGHUP to reopen the devices in priority order
ratchov [Fri, 3 May 2024 05:18:09 +0000 (05:18 +0000)]
sndiod: Fix SIGHUP to reopen the devices in priority order

With this commit, SIGHUP makes sndiod discover new devices and if
there is a new device with higher priority (greater -F option number)
than the current one, sndiod switches to it.  If the current device is
already the one with the highest priority (i.e. last -F), then SIGHUP
does nothing.

6 months agoAlign CRL and CSR version printing with certs
tb [Fri, 3 May 2024 02:52:00 +0000 (02:52 +0000)]
Align CRL and CSR version printing with certs

Only print specified 0-based versions and print them with the 1-based
human interpretation. Use a colon and error check the BIO_printf()
calls. (There's a lot more to clean up in here, but that's for another
day).

Notably, X509_CRL_print_ex() is missing... I guess that's better than
having one with signature and semantics differing from X509_print_ex()
und X509_REQ_print_ex().

ok beck

6 months agoQuick fix previous one. socantrcvmore() should raise assertion if
mvs [Thu, 2 May 2024 21:26:52 +0000 (21:26 +0000)]
Quick fix previous one. socantrcvmore() should raise assertion if
`so_rcv' has SB_MTXLOCK flag clean, not SB_OWNLOCK.

ok bluhm

6 months agounbreak parsing of IPv6 addresses in file-backed table(5)s
op [Thu, 2 May 2024 18:14:33 +0000 (18:14 +0000)]
unbreak parsing of IPv6 addresses in file-backed table(5)s

The file parser splits the line on the ':' character too for key-value
tables, and so mis-parses IPv6 addresses.  The "::1 localhost" example
in table(5) is actually parsed as key "" and value ":1 localhost".

For list tables, the "# @list" marker can be used as a workaround, but
for key-valued the parser has to be fixed.

There are also some weird edge cases when splitting the lines.

Now the parser always splits on the first whitespace or colon, and then
strips the spaces.  For lines starting with '[' the parser will jump to
the matching ']' before attempting to split.  So, for example:

[::1]:localhost becomes "[::1]" -> "localhost"
[::1] example.org becomes "[::1]" -> "example.org"
foo: bar becomes "foo" -> "bar"
foo::bar becomes "foo" -> ":bar"
foo : bar becomes "foo" -> ": bar"

etc...

This only affects the parser for file table(5)s and makemap(8).  Inline
tables or "proc" tables are unaffected.

ok gilles@

6 months agoDon't re-lock sockets in uipc_shutdown().
mvs [Thu, 2 May 2024 17:10:55 +0000 (17:10 +0000)]
Don't re-lock sockets in uipc_shutdown().

No reason to lock peer. It can't be or became listening socket, both
sockets can't be in the middle of connecting or disconnecting.

ok bluhm

6 months agovmctl(8): Add 'vmctl status -r'
mlarkin [Thu, 2 May 2024 15:46:10 +0000 (15:46 +0000)]
vmctl(8): Add 'vmctl status -r'

The -r option can be used to limit the output of 'vmctl status' to only
running VMs. This is useful for machines that have a large number of
stopped VMs, as the running ones are printed at the top by default and
previously required scrolling back to see the list of running VMs, and/or
using 'grep RUNNING'.

There is no change for users not using -r.

ok dv

6 months agoRemove a useless OBJ_obj2nid() call from X509_CRL_print()
tb [Thu, 2 May 2024 15:33:59 +0000 (15:33 +0000)]
Remove a useless OBJ_obj2nid() call from X509_CRL_print()

ok beck (as part of a larger diff)

6 months agoPass `sosp' instead of `so' to sblock() when locking `so_snd' within
mvs [Thu, 2 May 2024 11:55:31 +0000 (11:55 +0000)]
Pass `sosp' instead of `so' to sblock() when locking `so_snd' within
sosplice().

ok bluhm

6 months agoDelete extra spaces, change spaces to tab.
aoyama [Thu, 2 May 2024 09:38:59 +0000 (09:38 +0000)]
Delete extra spaces, change spaces to tab.

No binary change.

6 months agoremove unneeded includes
jsg [Wed, 1 May 2024 13:15:59 +0000 (13:15 +0000)]
remove unneeded includes
ok miod@ mpi@

6 months agoAdd per-CPU caches to the pmemrange allocator.
mpi [Wed, 1 May 2024 12:54:27 +0000 (12:54 +0000)]
Add per-CPU caches to the pmemrange allocator.

The caches are used primarily to reduce contention on uvm_lock_fpageq() during
concurrent page faults.  For the moment only uvm_pagealloc() tries to get a
page from the current CPU's cache.  So on some architectures the caches are
also used by the pmap layer.

Each cache is composed of two magazines, design is borrowed from jeff bonwick
vmem's paper and the implementation is similar to the one of pool_cache from
dlg@.  However there is no depot layer and magazines are refilled directly by
the pmemrange allocator.

This version includes splvm()/splx() dances because the buffer cache flips
buffers in interrupt context.  So we have to prevent recursive accesses to
per-CPU magazines.

Tested by naddy@, solene@, krw@, robert@, claudio@ and Laurence Tratt.

ok claudio@, kettenis@

6 months agoadd return statements missed when adapting from FreeBSD
jsg [Wed, 1 May 2024 11:22:21 +0000 (11:22 +0000)]
add return statements missed when adapting from FreeBSD

Avoids segfaults with an argument of 0, NaN, or Inf.
Problem reported by Colin Ian King.  ok miod@ kettenis@

6 months agoRefactor receive offloading code in ix(4)
jan [Wed, 1 May 2024 10:43:42 +0000 (10:43 +0000)]
Refactor receive offloading code in ix(4)

 - reduce complexcity
 - avoid combination of TCP and UDP IN_OK mbuf annotaions

with tweaks from bluhm
tested by bluhm

ok bluhm@

6 months agoPush solock() down to sosend() for SOCK_RAW sockets.
mvs [Tue, 30 Apr 2024 17:59:15 +0000 (17:59 +0000)]
Push  solock() down to sosend() for SOCK_RAW sockets.

Raw sockets are the simplest inet sockets, so use them to start landing
`sb_mtx' mutex(9) protection for `so_snd' buffer. Now solock() is taken
only around pru_send*(), the rest of sosend() serialized by sblock() and
`sb_mtx'. The unlocked SS_ISCONNECTED check is fine, because
rip{,6}_send() check it. Also, previously the SS_ISCONNECTED could be
lost due to solock() release around following m_getuio().

ok bluhm

6 months agoAdd '\n' to DPRINTF() string that used to be a panic() string.
krw [Tue, 30 Apr 2024 17:12:19 +0000 (17:12 +0000)]
Add '\n' to DPRINTF() string that used to be a panic() string.

ok mlarkin@

6 months agoMake nfstov_mode() return mode_t rather than uint16_t.
miod [Tue, 30 Apr 2024 17:06:00 +0000 (17:06 +0000)]
Make nfstov_mode() return mode_t rather than uint16_t.
This is a no-op since we only care about the low 12 bits in NFS anyway.

ok jsg@ semarie@

6 months agoConstify NFS data whenever possible. Also make a few more const data static.
miod [Tue, 30 Apr 2024 17:05:20 +0000 (17:05 +0000)]
Constify NFS data whenever possible. Also make a few more const data static.

ok jsg@ semarie@

6 months agoConvert all the NFS macros (in nfsm_subs.h) into inline functions with the
miod [Tue, 30 Apr 2024 17:04:23 +0000 (17:04 +0000)]
Convert all the NFS macros (in nfsm_subs.h) into inline functions with the
appropriate extra arguments.

This (hopefully) completes the unmessyfication work started by thib@ a long,
long time ago (in a galaxy far away).

The conversion logic has been:
- nfsm_dissect has been turned into an rvalue expression, leaving the
  cast operation up to its caller.
- macros which had three different exit paths (return, goto nfsmout or
  fallthrough) have been split so that no macros have more than two exit paths.
- then they have been modified to return a value, which lets the caller
  figure out what exit path is needed.
- local variables abused by the macros are now local variables of the new
  inline functions.

This single commit is the sum of 25 intermediate diffs, which have all been
carefully reviewed by (at least) jsg@ and semarie@.

Tested with v2 and v3 servers and clients.

ok jsg@ semarie@

6 months agoadd typedefs for msgqnum_t and msglen_t, required by POSIX.
op [Tue, 30 Apr 2024 17:03:05 +0000 (17:03 +0000)]
add typedefs for msgqnum_t and msglen_t, required by POSIX.

improvements from kettenis and jca
ok millert, jca, guenther

6 months agoDo not cast off_t to u_long in uvm_vnp_setsize call (only misbehaves on 32-bit
miod [Tue, 30 Apr 2024 16:54:47 +0000 (16:54 +0000)]
Do not cast off_t to u_long in uvm_vnp_setsize call (only misbehaves on 32-bit
platforms.)

ok mpi@

6 months agonever close stdin
tobias [Tue, 30 Apr 2024 15:40:43 +0000 (15:40 +0000)]
never close stdin

The sanitise_stdfd call makes sure that standard file descriptors are
open (if they were closed, they are connected with /dev/null).

Do not close stdin in any case to prevent error messages when stdin is
read multiple times and to prevent later usage of fd 0 for connections,
e.g.

echo localhost | ssh-keyscan -f - -f -

While at it, make stdin-related error messages nicer.

Authored with Max Kunzelmann <maxdev at posteo dot de>

ok djm

6 months agoAllow parsing either RFC9557 timestamps (which wants explicit +00:00) or
sthen [Tue, 30 Apr 2024 14:26:50 +0000 (14:26 +0000)]
Allow parsing either RFC9557 timestamps (which wants explicit +00:00) or
the existing format (with Z). From espie.

6 months agofix home-directory extension implementation, it always returned
djm [Tue, 30 Apr 2024 06:23:51 +0000 (06:23 +0000)]
fix home-directory extension implementation, it always returned
the current user's home directory contrary to the spec.

Patch from Jakub Jelen via GHPR477

6 months agoflush stdout after writing "sftp>" prompt when not using editline.
djm [Tue, 30 Apr 2024 06:16:55 +0000 (06:16 +0000)]
flush stdout after writing "sftp>" prompt when not using editline.

From Alpine Linux via GHPR480

6 months agostricter validation of messaging socket fd number; disallow usage of
djm [Tue, 30 Apr 2024 05:53:03 +0000 (05:53 +0000)]
stricter validation of messaging socket fd number; disallow usage of
stderr. Based on GHPR492 by RealHurrison

6 months agoadd missing reserved fields to key constraint protocol documentation.
djm [Tue, 30 Apr 2024 05:45:56 +0000 (05:45 +0000)]
add missing reserved fields to key constraint protocol documentation.

from Wiktor Kwapisiewicz via GHPR487

6 months agoappeard -> appeared;
jmc [Tue, 30 Apr 2024 05:05:23 +0000 (05:05 +0000)]
appeard -> appeared;

6 months agocorrectly restore sigprocmask around ppoll()
djm [Tue, 30 Apr 2024 02:14:10 +0000 (02:14 +0000)]
correctly restore sigprocmask around ppoll()
reported by Tõivo Leedjärv; ok deraadt@

6 months agoadd explict check for server hostkey type against HostkeyAlgorithms.
djm [Tue, 30 Apr 2024 02:10:49 +0000 (02:10 +0000)]
add explict check for server hostkey type against HostkeyAlgorithms.
Allows HostkeyAlgorithms to disable implicit fallback from certificate
keys to plain keys. ok markus@

6 months agoremove commented DEBUG_TIMER for timer driver removed in 2021
jsg [Tue, 30 Apr 2024 01:53:49 +0000 (01:53 +0000)]
remove commented DEBUG_TIMER for timer driver removed in 2021

6 months agoem(4) will work
deraadt [Mon, 29 Apr 2024 21:32:10 +0000 (21:32 +0000)]
em(4) will work

6 months agoTurn efi(4/arm64) into MI efi(4), sync with reality; OK kettenis
kn [Mon, 29 Apr 2024 20:18:12 +0000 (20:18 +0000)]
Turn efi(4/arm64) into MI efi(4), sync with reality;  OK kettenis

6 months agoplug fd leak introduced in "avoid toctu" rewrite.
florian [Mon, 29 Apr 2024 16:17:46 +0000 (16:17 +0000)]
plug fd leak introduced in "avoid toctu" rewrite.

Issue reported by job
OK tb, job

6 months agovmm & vmd: drop "continue" flag to simplify running a vcpu.
dv [Mon, 29 Apr 2024 14:47:05 +0000 (14:47 +0000)]
vmm & vmd: drop "continue" flag to simplify running a vcpu.

There's no need to distinguish the "first" time running a vcpu from
the subsequent times because vmm(4) uses in-kernel state tracking
the last vm exit reason to optimize the logic for updating vcpu
registers from userland. While here, clean up the DPRINTF's to make
the Intel VMX logic similar to the AMD SVM.

ok mlarkin@

6 months agostrftime.3: clarify %G and %g description
naddy [Mon, 29 Apr 2024 13:34:19 +0000 (13:34 +0000)]
strftime.3: clarify %G and %g description

Clarify that %G and %g produce the rarely needed ISO week-numbering
year rather than the Gregorian calendar year.

With millert@ and Evan Silberman.
Some of the wording is from the upstream version.

ok tb@

6 months agoinstall accidentally dropped tput(1) man page again
naddy [Mon, 29 Apr 2024 13:23:19 +0000 (13:23 +0000)]
install accidentally dropped tput(1) man page again

6 months agoremove prototypes for removed functions
jsg [Mon, 29 Apr 2024 13:07:18 +0000 (13:07 +0000)]
remove prototypes for removed functions

6 months agoremove prototypes for removed functions
jsg [Mon, 29 Apr 2024 13:01:54 +0000 (13:01 +0000)]
remove prototypes for removed functions

6 months agoremove unused armv7 specific find_first_bit()
jsg [Mon, 29 Apr 2024 12:46:22 +0000 (12:46 +0000)]
remove unused armv7 specific find_first_bit()

6 months agoremove unused intc_intr_bootstrap()
jsg [Mon, 29 Apr 2024 12:42:06 +0000 (12:42 +0000)]
remove unused intc_intr_bootstrap()

6 months agoremove unused global vars
jsg [Mon, 29 Apr 2024 12:33:17 +0000 (12:33 +0000)]
remove unused global vars

6 months agoremove prototypes for removed functions
jsg [Mon, 29 Apr 2024 12:24:46 +0000 (12:24 +0000)]
remove prototypes for removed functions

6 months agoremove unused pmap_vp_remove()
jsg [Mon, 29 Apr 2024 10:07:37 +0000 (10:07 +0000)]
remove unused pmap_vp_remove()
ok miod@ kettenis@

6 months agodrm/amdgpu: remove invalid resource->start check v2
jsg [Mon, 29 Apr 2024 06:40:11 +0000 (06:40 +0000)]
drm/amdgpu: remove invalid resource->start check v2

From Christian Koenig
db74904a93a370807001e5fd64b5850a61f2bf65 in linux-6.6.y/6.6.29
ca7c4507ba87e9fc22e0ecfa819c3664b3e8287b in mainline linux

6 months agodrm/amdkfd: Fix memory leak in create_process failure
jsg [Mon, 29 Apr 2024 06:37:08 +0000 (06:37 +0000)]
drm/amdkfd: Fix memory leak in create_process failure

From Felix Kuehling
aa02d43367a9adf8c85fb382fea4171fb266c8d0 in linux-6.6.y/6.6.29
18921b205012568b45760753ad3146ddb9e2d4e2 in mainline linux

6 months agodrm/amdgpu: validate the parameters of bo mapping operations more clearly
jsg [Mon, 29 Apr 2024 06:35:53 +0000 (06:35 +0000)]
drm/amdgpu: validate the parameters of bo mapping operations more clearly

From xinhui pan
ef13eeca7c79136bc38e21eb67322c1cbd5c40ee in linux-6.6.y/6.6.29
6fef2d4c00b5b8561ad68dd2b68173f5c6af1e75 in mainline linux

6 months agodrm/radeon: make -fstrict-flex-arrays=3 happy
jsg [Mon, 29 Apr 2024 06:31:57 +0000 (06:31 +0000)]
drm/radeon: make -fstrict-flex-arrays=3 happy

From Alex Deucher
259486b5bdc63f648634c30b8284be29834a4125 in linux-6.6.y/6.6.29
0ba753bc7e79e49556e81b0d09b2de1aa558553b in mainline linux

6 months agodrm/i915/mst: Limit MST+DSC to TGL+
jsg [Mon, 29 Apr 2024 06:30:19 +0000 (06:30 +0000)]
drm/i915/mst: Limit MST+DSC to TGL+

From Ville Syrjala
75170320459ae5bedf73352989b8433880cba20a in linux-6.6.y/6.6.29
51bc63392e96ca45d7be98bc43c180b174ffca09 in mainline linux

6 months agodrm/i915/vma: Fix UAF on destroy against retire race
jsg [Mon, 29 Apr 2024 06:27:37 +0000 (06:27 +0000)]
drm/i915/vma: Fix UAF on destroy against retire race

From Janusz Krzysztofik
5e3eb862df9f972ab677fb19e0d4b9b1be8db7b5 in linux-6.6.y/6.6.29
0e45882ca829b26b915162e8e86dbb1095768e9e in mainline linux

6 months agodrm/amd/display: Do not recursively call manual trigger programming
jsg [Mon, 29 Apr 2024 06:25:15 +0000 (06:25 +0000)]
drm/amd/display: Do not recursively call manual trigger programming

From Dillon Varone
a17236f99f49c5e3c2a052302127284a99fc07bb in linux-6.6.y/6.6.29
953927587f37b731abdeabe46ad44a3b3ec67a52 in mainline linux

6 months agodrm/i915: Disable live M/N updates when using bigjoiner
jsg [Mon, 29 Apr 2024 06:23:20 +0000 (06:23 +0000)]
drm/i915: Disable live M/N updates when using bigjoiner

From Ville Syrjala
9d106175b7a081f98112694e7fbb82105aff4eef in linux-6.6.y/6.6.29
4a36e46df7aa781c756f09727d37dc2783f1ee75 in mainline linux

6 months agodrm/i915: Adjust seamless_m_n flag behaviour
jsg [Mon, 29 Apr 2024 06:21:44 +0000 (06:21 +0000)]
drm/i915: Adjust seamless_m_n flag behaviour

From Ville Syrjala
ccb0934aeb3f7be579717041b10d274feef21de3 in linux-6.6.y/6.6.29
825edc8bc72f3266534a04e9a4447b12332fac82 in mainline linux

6 months agodrm/i915: Enable VRR later during fastsets
jsg [Mon, 29 Apr 2024 06:19:53 +0000 (06:19 +0000)]
drm/i915: Enable VRR later during fastsets

From Ville Syrjala
deaeb5b64c5b8a7b681acaf721d040b1fe9dcc53 in linux-6.6.y/6.6.29
691dec86acc3afb469f09e9a4a00508b458bdb0c in mainline linux

6 months agodrm/i915: Extract intel_crtc_vblank_evade_scanlines()
jsg [Mon, 29 Apr 2024 06:18:00 +0000 (06:18 +0000)]
drm/i915: Extract intel_crtc_vblank_evade_scanlines()

From Ville Syrjala
7e50ac4203d6a8b151233550cec7d7da794b2e13 in linux-6.6.y/6.6.29
f4b0cece716c95e16d973a774d5a5c5cc8cb335d in mainline linux

6 months agodrm/i915: Change intel_pipe_update_{start,end}() calling convention
jsg [Mon, 29 Apr 2024 06:16:30 +0000 (06:16 +0000)]
drm/i915: Change intel_pipe_update_{start,end}() calling convention

From Ville Syrjala
e19dc8c49e97b5cd84ee5753c301d64cde98c6aa in linux-6.6.y/6.6.29
09f390d4e2f38f8433431f4da31ca0a17a5c7853 in mainline linux

6 months agodrm/i915/cdclk: Fix voltage_level programming edge case
jsg [Mon, 29 Apr 2024 06:14:32 +0000 (06:14 +0000)]
drm/i915/cdclk: Fix voltage_level programming edge case

From Ville Syrjala
51cc733fb48e86c3bbfaef1713710b2a6ffe565d in linux-6.6.y/6.6.29
6154cc9177ccea00c89ce0bf93352e474b819ff2 in mainline linux

6 months agodrm/i915/mst: Reject FEC+MST on ICL
jsg [Mon, 29 Apr 2024 06:12:46 +0000 (06:12 +0000)]
drm/i915/mst: Reject FEC+MST on ICL

From Ville Syrjala
6711b0817a384bbeb8c951521334d5ef78c855aa in linux-6.6.y/6.6.29
99f855082f228cdcecd6ab768d3b8b505e0eb028 in mainline linux

6 months agodrm/i915: Fix FEC pipe A vs. DDI A mixup
jsg [Mon, 29 Apr 2024 06:10:41 +0000 (06:10 +0000)]
drm/i915: Fix FEC pipe A vs. DDI A mixup

From Ville Syrjala
6b625d2743b464a16904feb5520462041e3d3ce2 in linux-6.6.y/6.6.29
126f94e87e7960ef7ae58180e39c19cc9dcbbf7f in mainline linux

6 months agoadd check_add_overflow() for 6.6.29 drm
jsg [Mon, 29 Apr 2024 06:06:46 +0000 (06:06 +0000)]
add check_add_overflow() for 6.6.29 drm

6 months agoremove unused global vars
jsg [Mon, 29 Apr 2024 00:29:48 +0000 (00:29 +0000)]
remove unused global vars
ok deraadt@ miod@

6 months agogmtime(3) / locatime(3) can fail when timestamps are way off.
florian [Sun, 28 Apr 2024 16:43:42 +0000 (16:43 +0000)]
gmtime(3) / locatime(3) can fail when timestamps are way off.

Add missing error checks to all calls under sbin/

Input & OK millert

6 months agogmtime(3) / locatime(3) can fail when timestamps are way off.
florian [Sun, 28 Apr 2024 16:43:15 +0000 (16:43 +0000)]
gmtime(3) / locatime(3) can fail when timestamps are way off.

Add missing error checks to all calls under bin/

Input & OK millert

6 months agogmtime(3) / locatime(3) can fail when timestamps are way off.
florian [Sun, 28 Apr 2024 16:42:53 +0000 (16:42 +0000)]
gmtime(3) / locatime(3) can fail when timestamps are way off.

Add missing error checks to all calls under libexec/

Input & OK millert

6 months ago__sfvwrite: allow writing buffers larger than BUFSIZ or st_blksize.
millert [Sun, 28 Apr 2024 14:28:02 +0000 (14:28 +0000)]
__sfvwrite: allow writing buffers larger than BUFSIZ or st_blksize.

This can vastly improve write performance.  For unbuffered writes,
we now write up to INT_MAX instead of BUFSIZE.  For buffered writes,
write the largest multiple of the buffer size (which may be st_blksize).
From Apple via enh at google.  OK tb@

6 months agosync
deraadt [Sun, 28 Apr 2024 13:12:31 +0000 (13:12 +0000)]
sync

6 months agoRemove unused url from netproc()
tb [Sun, 28 Apr 2024 10:09:25 +0000 (10:09 +0000)]
Remove unused url from netproc()

Leftover from when RFC 8555 support was added.

ok florian

6 months agoDon't filter dangerous characters by hand when strvisx(3) is right there.
florian [Sun, 28 Apr 2024 08:29:56 +0000 (08:29 +0000)]
Don't filter dangerous characters by hand when strvisx(3) is right there.

segfault reported by sthen, which was most likely caused by buf_dump().

OK tb

6 months agoUse propper knf.
florian [Sat, 27 Apr 2024 19:49:42 +0000 (19:49 +0000)]
Use propper knf.

Spotted by kettenis.

6 months agoadd a man page from curses source for clear(1);
jmc [Sat, 27 Apr 2024 15:05:55 +0000 (15:05 +0000)]
add a man page from curses source for clear(1);
from piotr durlej

ok millert

6 months agolocaltime(3) can fail if time_t is very far in the future or past.
florian [Sat, 27 Apr 2024 14:57:02 +0000 (14:57 +0000)]
localtime(3) can fail if time_t is very far in the future or past.

found using afl++

OK millert

6 months ago"!command" notation was zapped 8 years ago - remove corresponding
jmc [Sat, 27 Apr 2024 14:52:30 +0000 (14:52 +0000)]
"!command" notation was zapped 8 years ago - remove corresponding
entry from the help file; from piotr durlej

while here, zap an unneccessary blank line

ok millert

6 months agoTweak start of DESCRIPTION of EC_GROUP_new()
tb [Sat, 27 Apr 2024 07:41:47 +0000 (07:41 +0000)]
Tweak start of DESCRIPTION of EC_GROUP_new()

Complete precision will lead to complete confusion, however outright lies
are also bad.

6 months agoDocument EC_curve_{nid2nist,nist2nid}()
tb [Fri, 26 Apr 2024 14:23:03 +0000 (14:23 +0000)]
Document EC_curve_{nid2nist,nist2nid}()

6 months agofix signal test; ok tb@
jsg [Fri, 26 Apr 2024 04:44:43 +0000 (04:44 +0000)]
fix signal test; ok tb@

6 months agoUpdate awk to the Apr 22, 2024 version.
millert [Thu, 25 Apr 2024 18:33:53 +0000 (18:33 +0000)]
Update awk to the Apr 22, 2024 version.

 * fixed regex engine gototab reallocation issue that was introduced
   during the Nov 24 rewrite.

 * fixed use-after-free bug in fnematch due to adjbuf invalidating
   the pointers to buf.