openbsd
20 months agosync
deraadt [Tue, 21 Feb 2023 05:06:51 +0000 (05:06 +0000)]
sync

20 months agoSet the current pmap in macppc's pmap_activate
gkoehler [Tue, 21 Feb 2023 04:49:43 +0000 (04:49 +0000)]
Set the current pmap in macppc's pmap_activate

This fixes a possible freeze in execve(2).  It sometimes froze when a
dual-cpu macppc started daemons during boot.  There is a chance that
uvm_map.c uvmspace_exec sees ovm->vm_refcnt != 1 and switches curproc
to a new pmap.  If this happened, then execve froze by trying to
copyout to the wrong pmap; curpcb->pcb_pm was old.  Fix by setting
pointers when uvmspace_exec calls pmap_activate.

ok miod@

20 months agosync
deraadt [Tue, 21 Feb 2023 04:08:50 +0000 (04:08 +0000)]
sync

20 months agoRemove conflict markers from perl tests
afresh1 [Tue, 21 Feb 2023 02:20:35 +0000 (02:20 +0000)]
Remove conflict markers from perl tests

This is upstream PR 28103 https://github.com/Perl/perl5/pull/20813

Requested by stsp@ for quality of life improvments for Game of Trees users.

20 months agocorrect parameter type
deraadt [Tue, 21 Feb 2023 02:13:27 +0000 (02:13 +0000)]
correct parameter type

20 months agoDon't install DebugWrap.1 manpage
afresh1 [Tue, 21 Feb 2023 02:11:03 +0000 (02:11 +0000)]
Don't install DebugWrap.1 manpage

It's from a test file and should never have been installed.

Noticed by xse in #openbsd on libera.chat

20 months agosndiod(8) is the audio daemon now; from david demelier
jmc [Mon, 20 Feb 2023 16:08:37 +0000 (16:08 +0000)]
sndiod(8) is the audio daemon now; from david demelier

20 months agoRework/modernize verbiage in the "AUTOMATIC DISK ALLOCATION" and
krw [Mon, 20 Feb 2023 14:16:59 +0000 (14:16 +0000)]
Rework/modernize verbiage in the "AUTOMATIC DISK ALLOCATION" and
"EXAMPLES" sections. Build tables with 'Bd -column' instead of
tabs inside 'Bl -literal'.  Columns now line up (more or less) in
text, .pdf and .html versions.

Prompted by kn@, feedback & suggestions from jmc@ and schwarze@.

20 months agoRewrite the ROM walk logic to correctly iterate over non-STI ROMs (and skip
miod [Mon, 20 Feb 2023 11:31:16 +0000 (11:31 +0000)]
Rewrite the ROM walk logic to correctly iterate over non-STI ROMs (and skip
them), such as x86 bios ROMs.

20 months agoWrap all printf calls in sti_check_rom() in order to automagically cope with
miod [Mon, 20 Feb 2023 09:08:47 +0000 (09:08 +0000)]
Wrap all printf calls in sti_check_rom() in order to automagically cope with
ROM enable/disable; makes addition of debug code more fool-proof. NFC (yet)

20 months agoDisable strict bourne shell mode to get ksh behaviour in (S)hell
kn [Mon, 20 Feb 2023 01:55:41 +0000 (01:55 +0000)]
Disable strict bourne shell mode to get ksh behaviour in (S)hell

Enable features like arithmatic expression and thus make the debug aid
`FUNCS_ONLY=1 . /install.sub' work in those prompts.

'!' answers spawn their own ksh process, but (S)hell drops into /.profile
sourced by /bin/sh executed by init(8), i.e. a korn shell defaulting to
strict bourne shell mode.

Shell options in this interactive parent shell do not effect install.sub
aka. /install, /autoinstall and /upgrade as it is executed in a child.

OK afresh1

20 months agoAdjust inline !libc check to match the intent in ld.so
gnezdo [Mon, 20 Feb 2023 00:51:57 +0000 (00:51 +0000)]
Adjust inline !libc check to match the intent in ld.so

right, deraadt@

20 months agoUnbreak RAMDISK build by adding a check for MULTIPROCESSOR.
patrick [Mon, 20 Feb 2023 00:01:16 +0000 (00:01 +0000)]
Unbreak RAMDISK build by adding a check for MULTIPROCESSOR.

20 months agoImprove media disk question default
kn [Sun, 19 Feb 2023 23:46:22 +0000 (23:46 +0000)]
Improve media disk question default

Sets on an unmounted disk partition are most likely to be on install media,
e.g. CD or USB stick.

Default 'Which disk contains the install media?' to the first disk that is
a) not the root disk and b) not a disk with softraid chunks (hosting the
root disk, for example).

This makes it point at the right disk in most setups.
All disks remain valid answers, this is only about the default.

OK afresh1

20 months agoauth_getchallenge does not exist anymore
aisha [Sun, 19 Feb 2023 21:33:38 +0000 (21:33 +0000)]
auth_getchallenge does not exist anymore

20 months agoMake pinsyscall(2) always available for pledged processes. Needed by pledge
anton [Sun, 19 Feb 2023 18:46:46 +0000 (18:46 +0000)]
Make pinsyscall(2) always available for pledged processes. Needed by pledge
execpromises, as noted by regress/sys/kern/pledge/execpromise.

sure deraadt@

20 months ago+openbsd-74-syspatch.pub
robert [Sun, 19 Feb 2023 17:29:49 +0000 (17:29 +0000)]
+openbsd-74-syspatch.pub

20 months agoadd 7.4 syspatch public key
robert [Sun, 19 Feb 2023 17:28:46 +0000 (17:28 +0000)]
add 7.4 syspatch public key

20 months agoAdd support for deep(er) idle states that can be entered using PSCI. For
kettenis [Sun, 19 Feb 2023 17:16:13 +0000 (17:16 +0000)]
Add support for deep(er) idle states that can be entered using PSCI.  For
now this only supports states advertised in device trees, but ACPI support
could be added as well.  The parsing of the idle states as well as the
heuristic to pick the deepest one is probably a bit to simple, but more
complex cases can be added later.  Worst case cores will use WFI and use
more power in suspend.

ok phessler@

20 months agodescribe what qcow2 is, in sufficient simple terms
deraadt [Sun, 19 Feb 2023 17:10:38 +0000 (17:10 +0000)]
describe what qcow2 is, in sufficient simple terms

20 months agoMove BN_MONT_CTX back to bn_local.h for now. It's still used by bn_exp.c
tb [Sun, 19 Feb 2023 15:45:14 +0000 (15:45 +0000)]
Move BN_MONT_CTX back to bn_local.h for now. It's still used by bn_exp.c

20 months agosync
deraadt [Sun, 19 Feb 2023 15:35:22 +0000 (15:35 +0000)]
sync

20 months ago7.4 packages key
naddy [Sun, 19 Feb 2023 15:19:38 +0000 (15:19 +0000)]
7.4 packages key

20 months agoRemove non-existent make targets from ports(7) and bsd.port.mk(5)
sdk [Sun, 19 Feb 2023 14:37:54 +0000 (14:37 +0000)]
Remove non-existent make targets from ports(7) and bsd.port.mk(5)

Removed from ports(7):
- describe, homepage-links, (un)link-categories, unlink-categories

Removed from bsd.port.mk(5):
- (un)link-categories

OK op@

20 months agoRewrite BN_MONT_CTX_set_locked()
jsing [Sun, 19 Feb 2023 13:51:00 +0000 (13:51 +0000)]
Rewrite BN_MONT_CTX_set_locked()

Rewrite and simplify BN_MONT_CTX_set_locked - in particular, only hold the
lock for a short period of time, rather than holding a write lock for a
module across an expensive operation.

ok tb@

20 months agoFirst pass clean up of bn_mont.c.
jsing [Sun, 19 Feb 2023 13:44:29 +0000 (13:44 +0000)]
First pass clean up of bn_mont.c.

Use calloc() rather than malloc() with manual initialisation of all struct
members to zero, use memset() instead of manually initialising all struct
members to zero, use consistent naming, use BN_free() instead of
BN_clear_free() (since it is the same thing).

ok tb@

20 months agoMove BN_MONT_CTX_copy().
jsing [Sun, 19 Feb 2023 13:33:23 +0000 (13:33 +0000)]
Move BN_MONT_CTX_copy().

20 months agoMove struct bn_mont_ctx_st from bn_local.h to bn_mont.c.
jsing [Sun, 19 Feb 2023 13:27:49 +0000 (13:27 +0000)]
Move struct bn_mont_ctx_st from bn_local.h to bn_mont.c.

No code outside of bn_mont.c needs access to it.

ok tb@

20 months agoDo not let the iwx init task run in parallel to wakeup code during resume.
stsp [Sun, 19 Feb 2023 12:23:27 +0000 (12:23 +0000)]
Do not let the iwx init task run in parallel to wakeup code during resume.

The driver will resume in the ACPI thread and run code to wake the device.
If a firmware error occurs then the init task will be scheduled and could
then run in parallel to the wakeup code. This would lead to panics as the
init task ran while we were not yet done with device initialization.

To prevent this problem we now grab the rwlock during wakeup, the same
lock which is used to prevent races between the init task and ioctls.
Problem found by mvs@, who also suggested the fix implemented here and
tested the changes.

ok mvs@

20 months agoDisable tests using the preferred bsd.regress.mk technique.
anton [Sun, 19 Feb 2023 10:40:34 +0000 (10:40 +0000)]
Disable tests using the preferred bsd.regress.mk technique.

20 months agosync
deraadt [Sat, 18 Feb 2023 21:29:55 +0000 (21:29 +0000)]
sync

20 months ago7.4 base key
deraadt [Sat, 18 Feb 2023 18:50:15 +0000 (18:50 +0000)]
7.4 base key

20 months agoMissing space
claudio [Sat, 18 Feb 2023 18:07:38 +0000 (18:07 +0000)]
Missing space

20 months agoFix possibly uninitialized variable in editor_allocspace(); ok krw@
miod [Sat, 18 Feb 2023 15:22:40 +0000 (15:22 +0000)]
Fix possibly uninitialized variable in editor_allocspace(); ok krw@

20 months agoacpipwrres(4): remove unused aml_value's.
dv [Sat, 18 Feb 2023 14:32:02 +0000 (14:32 +0000)]
acpipwrres(4): remove unused aml_value's.

ok kettenis@

20 months agomissing dot before macro; from josiah frentsos
jmc [Sat, 18 Feb 2023 13:46:40 +0000 (13:46 +0000)]
missing dot before macro; from josiah frentsos

20 months agoPut explicit ULL suffix to constants which won't fit in 32 bits.
miod [Sat, 18 Feb 2023 08:52:39 +0000 (08:52 +0000)]
Put explicit ULL suffix to constants which won't fit in 32 bits.

20 months agoLookup the start,len of the "execve" stub in libc.so, and tell the kernel
deraadt [Sat, 18 Feb 2023 01:22:50 +0000 (01:22 +0000)]
Lookup the start,len of the "execve" stub in libc.so, and tell the kernel
where it is with pinsyscall().  In non-static binaries, kernel execve(2)
will now insist upon being called from that small region.  Arriving from
a different region smells like ROP methodology, and the process is killed.

20 months agoValidate execve() libc stub location if kernel knows it. (due to ld.so
deraadt [Fri, 17 Feb 2023 18:08:32 +0000 (18:08 +0000)]
Validate execve() libc stub location if kernel knows it. (due to ld.so
telling the kernel with pinsyscall(2)

20 months agoDo not need KTRC_CODE__SYSCALL in the ktrace files anymore, because
deraadt [Fri, 17 Feb 2023 18:02:07 +0000 (18:02 +0000)]
Do not need KTRC_CODE__SYSCALL in the ktrace files anymore, because
__syscall() was removed.

20 months agoKTRC_CODE__SYSCALL is never set anymore, because __syscall() is gone.
deraadt [Fri, 17 Feb 2023 18:01:26 +0000 (18:01 +0000)]
KTRC_CODE__SYSCALL is never set anymore, because __syscall() is gone.

20 months agoRemove redundant assignment; ok millert@
miod [Fri, 17 Feb 2023 18:00:35 +0000 (18:00 +0000)]
Remove redundant assignment; ok millert@

20 months agoOnly compile _ng_print() #ifdef DEBUG_NG; ok millert@
miod [Fri, 17 Feb 2023 18:00:11 +0000 (18:00 +0000)]
Only compile _ng_print() #ifdef DEBUG_NG; ok millert@

20 months agoRemove unused variables; ok millert@
miod [Fri, 17 Feb 2023 17:59:36 +0000 (17:59 +0000)]
Remove unused variables; ok millert@

20 months agoUpdate AUTHORS section
job [Fri, 17 Feb 2023 14:52:06 +0000 (14:52 +0000)]
Update AUTHORS section

20 months agoUse %R technical report macro for RFCs
job [Fri, 17 Feb 2023 13:26:18 +0000 (13:26 +0000)]
Use %R technical report macro for RFCs

20 months agospelling
jsg [Fri, 17 Feb 2023 08:01:03 +0000 (08:01 +0000)]
spelling
ok mglocker@

20 months agoProvide optimised versions of bn_addw() and bn_subw() for aarch64.
jsing [Fri, 17 Feb 2023 05:46:57 +0000 (05:46 +0000)]
Provide optimised versions of bn_addw() and bn_subw() for aarch64.

20 months agoRemove now unused tangle of mul*/sqr* and BN_UMULT_* macros.
jsing [Fri, 17 Feb 2023 05:30:20 +0000 (05:30 +0000)]
Remove now unused tangle of mul*/sqr* and BN_UMULT_* macros.

No, I'm not trying to overwhelm you... however, we really no longer need
this clutter.

ok tb@

20 months agoReimplement bn_sqr_comba{4,8}().
jsing [Fri, 17 Feb 2023 05:13:34 +0000 (05:13 +0000)]
Reimplement bn_sqr_comba{4,8}().

Use bignum primitives rather than the current mess of macros.The sqr_add_c
macro gets replaced with bn_mulw_addtw(), while the sqr_add_c2 macro gets
replaced with bn_mul2_mulw_addtw().

The variables in the comba functions have also been reordered, so that the
patterns are easier to understand - the compiler can take care of
optimising the inputs and outputs to avoid register moves.

ok tb@

20 months agoRemove now-unused compat bit SSH_BUG_RSASIGMD5. The code to set this
dtucker [Fri, 17 Feb 2023 04:22:50 +0000 (04:22 +0000)]
Remove now-unused compat bit SSH_BUG_RSASIGMD5.  The code to set this
was removed in OpenSSH 7.7 when support for SSH implementations dating
back to before RFC standardization were removed.  "burn it all" djm@

20 months agoRemove now-unused compat bit SSH_BUG_BIGENDIANAES. This was
dtucker [Fri, 17 Feb 2023 03:06:18 +0000 (03:06 +0000)]
Remove now-unused compat bit SSH_BUG_BIGENDIANAES.  This was
previously set for OpenSSH 2.3 (released in 2000) but this check
was removed in OpenSSH 7.7 (2018).  ok djm@ deraadt@

20 months ago- remove a leftover .El
jmc [Thu, 16 Feb 2023 20:33:07 +0000 (20:33 +0000)]
- remove a leftover .El
- escape "An" as this is also a macro

20 months agoRemove obsolete __HAVE_VM_PAGE_MD define - all platforms provide vm_page_md
miod [Thu, 16 Feb 2023 20:32:39 +0000 (20:32 +0000)]
Remove obsolete __HAVE_VM_PAGE_MD define - all platforms provide vm_page_md
those days, and nothing checks for that symbol anymore.

20 months agoUpdate references in STANDARDS section to use Rs blocks
job [Thu, 16 Feb 2023 20:06:15 +0000 (20:06 +0000)]
Update references in STANDARDS section to use Rs blocks

Also drop largely irrelevant references like IPv6 and CIDR
(as we didn't reference IPv4 either), remove obsoleted RFCs and add
their successors.

20 months agoRewrite bsort() from hand-rolled recursive to simpler iterative reusing code
kn [Thu, 16 Feb 2023 18:10:28 +0000 (18:10 +0000)]
Rewrite bsort() from hand-rolled recursive to simpler iterative reusing code

ksh(1) can sort itself and addel() ensures uniqueness, so reuse both to get
a much simpler shell version of `sort -u' that is bug-for-bug compatible
with the old one but shorter and easier to tweak/reason about.

OK afresh1

20 months agoEnable dwqe(4).
kettenis [Thu, 16 Feb 2023 14:44:48 +0000 (14:44 +0000)]
Enable dwqe(4).

20 months agoFix MAC address register offsets.
kettenis [Thu, 16 Feb 2023 14:43:53 +0000 (14:43 +0000)]
Fix MAC address register offsets.

ok patrick@

20 months agoRevert r1.63.
tb [Thu, 16 Feb 2023 14:34:34 +0000 (14:34 +0000)]
Revert r1.63.

GEN_OTHERNAME is the type of a GENERAL_NAMES, not of a DIST_POINT_NAME,
which needs naked numbers as there is no enum nor defines describing it.

claudio agrees

20 months agoAdd missing RFC 6487 section 4.8.6 CRLDP compliance checks
job [Thu, 16 Feb 2023 14:25:27 +0000 (14:25 +0000)]
Add missing RFC 6487 section 4.8.6 CRLDP compliance checks

OK tb@ claudio@

20 months agoEnable s2n-bignum word_clz() on amd64.
jsing [Thu, 16 Feb 2023 11:13:05 +0000 (11:13 +0000)]
Enable s2n-bignum word_clz() on amd64.

The BN_num_bits_word() function is a hot path, being called more than
80 million times during a libcrypto regress run. The word_clz()
implementation uses five instructions to do the same as the generic code
that uses more than 60 instructions.

Discussed with tb@

20 months agoUse bn_addw() in bn_mulw(), rather than duplicating add with carry code.
jsing [Thu, 16 Feb 2023 10:58:06 +0000 (10:58 +0000)]
Use bn_addw() in bn_mulw(), rather than duplicating add with carry code.

20 months agoChange include from _internal_s2n_bignum.h to s2n_bignum_internal.h.
jsing [Thu, 16 Feb 2023 10:54:20 +0000 (10:54 +0000)]
Change include from _internal_s2n_bignum.h to s2n_bignum_internal.h.

20 months agoInclude the ISC license from s2n-bignum's LICENSE file.
jsing [Thu, 16 Feb 2023 10:53:18 +0000 (10:53 +0000)]
Include the ISC license from s2n-bignum's LICENSE file.

20 months agoBring in word_clz.S from s2n-bignum for amd64.
jsing [Thu, 16 Feb 2023 10:51:58 +0000 (10:51 +0000)]
Bring in word_clz.S from s2n-bignum for amd64.

20 months agoRename bn_umul_hilo() to bn_mulw().
jsing [Thu, 16 Feb 2023 10:41:03 +0000 (10:41 +0000)]
Rename bn_umul_hilo() to bn_mulw().

This keeps the naming consistent with the other bignum primitives that have
been recently introduced. Also, use 1/0 intead of h/l (e.g. a1 instead of
ah), as this keeps consistency with other primitives and allows for naming
that works with double word, triple word and quadruple word inputs/outputs.

Discussed with tb@

20 months agozap stray empty line
tb [Thu, 16 Feb 2023 10:18:59 +0000 (10:18 +0000)]
zap stray empty line

20 months agoRemove SSH_BUG_PASSWORDPAD compat bit since it's no longer used.
dtucker [Thu, 16 Feb 2023 10:10:00 +0000 (10:10 +0000)]
Remove SSH_BUG_PASSWORDPAD compat bit since it's no longer used.
ok markus@

20 months agoAdd missing masks to accumulator version of bn_umul_hilo()
jsing [Thu, 16 Feb 2023 10:02:02 +0000 (10:02 +0000)]
Add missing masks to accumulator version of bn_umul_hilo()

20 months agoregen
claudio [Thu, 16 Feb 2023 08:54:47 +0000 (08:54 +0000)]
regen

20 months agoutrace(2) can run without kernel lock. Now ktruser() will take the
claudio [Thu, 16 Feb 2023 08:53:16 +0000 (08:53 +0000)]
utrace(2) can run without kernel lock. Now ktruser() will take the
lock only when needed.
OK miod@ cheloha@

20 months agoLike other ktrace functions make ktruser() callable without kernel lock
claudio [Thu, 16 Feb 2023 08:50:57 +0000 (08:50 +0000)]
Like other ktrace functions make ktruser() callable without kernel lock
by takeing the lock around ktrwrite2().
OK miod@ cheloha@

20 months agolibressl *_namespace.h: adjust *_ALIAS() to require a semicolon
tb [Thu, 16 Feb 2023 08:38:17 +0000 (08:38 +0000)]
libressl *_namespace.h: adjust *_ALIAS() to require a semicolon

LCRYPTO_ALIAS() and LSSL_ALIAS() contained a trailing semicolon.
This does not conform to style(9), breaks editors and ctags and
(most importantly) my workflow. Fix this by neutering them with
asm("") so that -Wpedantic doesn't complain. There's precedent
in libc's namespace.h

fix suggested by & ok jsing

20 months agoRemove SSH_BUG_IGNOREMSG compat flag since it's only applicable to SSH1
dtucker [Thu, 16 Feb 2023 07:55:15 +0000 (07:55 +0000)]
Remove SSH_BUG_IGNOREMSG compat flag since it's only applicable to SSH1
and thus no longer used.  ok markus@ "kill it with fire" djm@

20 months agosync
deraadt [Thu, 16 Feb 2023 05:38:24 +0000 (05:38 +0000)]
sync

20 months agoDocument pinsyscall(2). The ld.so and kernel enforcement code is not
deraadt [Thu, 16 Feb 2023 04:43:54 +0000 (04:43 +0000)]
Document pinsyscall(2).  The ld.so and kernel enforcement code is not
commited ye -- waiting for enough people to run kernels which support
the system call.
ok jmc

20 months agosync
deraadt [Thu, 16 Feb 2023 04:42:27 +0000 (04:42 +0000)]
sync

20 months agoReimplement bn_add_words() and bn_sub_words() using bignum primitives.
jsing [Thu, 16 Feb 2023 04:42:20 +0000 (04:42 +0000)]
Reimplement bn_add_words() and bn_sub_words() using bignum primitives.

This removes the effectively duplicate BN_LLONG version of bn_add_words()
and simplifies the code considerably.

ok tb@

20 months agoAdd pinsyscall(2). With this you can tell the kernel the location
deraadt [Thu, 16 Feb 2023 04:42:06 +0000 (04:42 +0000)]
Add pinsyscall(2).  With this you can tell the kernel the location
(start,len) of the syscall stub in libc.so for a specified syscall
(using SYS_* notation).  Only SYS_execve is supported at this time.
ok gnezdo mortimer kettenis

20 months agoexplain why ARIN TAL is not included.
deraadt [Thu, 16 Feb 2023 04:07:33 +0000 (04:07 +0000)]
explain why ARIN TAL is not included.
ok job

20 months agoremove the '(R)' from the intel cpu match pattern
jsg [Thu, 16 Feb 2023 03:09:33 +0000 (03:09 +0000)]
remove the '(R)' from the intel cpu match pattern

Intel(R) does not appear in
cpu0: Intel Atom(R) x6425RE Processor @ 1.90GHz, 1895.90 MHz, 06-96-01
reported by patrick@ ok deraadt@

20 months agoRegen en_US.UTF-8.src for Unicode 14.0.0
afresh1 [Thu, 16 Feb 2023 01:07:01 +0000 (01:07 +0000)]
Regen en_US.UTF-8.src for Unicode 14.0.0

That comes with perl 5.36.

OK millert@

20 months agoWhitespace
afresh1 [Thu, 16 Feb 2023 01:06:01 +0000 (01:06 +0000)]
Whitespace

I am pretty sure it was schwarze@ who pointed this out long ago.

20 months agoproc_ispeer() is not used anywhere anymore so remove it everywhere.
tobhe [Wed, 15 Feb 2023 20:44:01 +0000 (20:44 +0000)]
proc_ispeer() is not used anywhere anymore so remove it everywhere.

ok florian@ bluhm@
ok for vmd mlarkin@

20 months agoRelax kernel lock assertion within tsleep(9). The `nowake' channel is
mvs [Wed, 15 Feb 2023 20:43:41 +0000 (20:43 +0000)]
Relax kernel lock assertion within tsleep(9). The `nowake' channel is
the special case which doesn't expect wakeup(9), so allow to use it
without kernel lock held.

Discussed with and ok by claudio@

20 months agopfsync(4) panics on NULL pointer dereference if there
sashan [Wed, 15 Feb 2023 18:11:47 +0000 (18:11 +0000)]
pfsync(4) panics on NULL pointer dereference if there
are no data ready for bulk transfer. reported and fix
kindly tested by hrvoje@

OK bluhm@

20 months agoPlace bn_mul_add_words() after bn_mul_words().
jsing [Wed, 15 Feb 2023 18:10:16 +0000 (18:10 +0000)]
Place bn_mul_add_words() after bn_mul_words().

20 months agobgpd no longer immediatly opens the connection to its peers so instead
claudio [Wed, 15 Feb 2023 14:19:08 +0000 (14:19 +0000)]
bgpd no longer immediatly opens the connection to its peers so instead
use 'bgpctl nei RDOMAIN2 up' to force the connection up. While there
use the same timeouts everywhere. May need further tuning but this
works for me.
Issue noticed by anton@

20 months agoThe GMAC on the Rockchip RK356x and RK3588 SoCs can only do 32-bit DMA
kettenis [Wed, 15 Feb 2023 14:13:38 +0000 (14:13 +0000)]
The GMAC on the Rockchip RK356x and RK3588 SoCs can only do 32-bit DMA
so set the appropriate DMA constraint.

ok patrick@

20 months agoDon't print the version twice, but do print a newline before attaching the
kettenis [Wed, 15 Feb 2023 14:10:58 +0000 (14:10 +0000)]
Don't print the version twice, but do print a newline before attaching the
PHY.
Put the DMA address of the mbuf into the rx descriptors (instead of the
length).  Use the correct value when setting the tx ring tail pointer.
This make sending and receiving packets work as long as the interface
is in promiscuous mode.

ok patrick@

20 months agoAdd GMAC-related RK356x clocks.
kettenis [Wed, 15 Feb 2023 14:06:43 +0000 (14:06 +0000)]
Add GMAC-related RK356x clocks.

ok patrick@

20 months agoUse correct order of arguments for shutdown(2).
florian [Wed, 15 Feb 2023 13:47:00 +0000 (13:47 +0000)]
Use correct order of arguments for shutdown(2).

Luckily routesock would never be <= 2 so shutdown would always failing
with EINVAL and not shutting down random other sockets.
The kernel was just uselessly piling up route messages that we were
never reading.

This mistake first showed up in slaacd(8) and then was copied to the
other daemons.

Reported & fixed by Josiah Frentsos (jfrent at tilde.team), thanks!
OK kn

20 months agoNow Puffy comes on luna88k bootloader!
aoyama [Wed, 15 Feb 2023 12:43:32 +0000 (12:43 +0000)]
Now Puffy comes on luna88k bootloader!

"Of course!" miod@

20 months agodrm/i915: Fix VBT DSI DVO port handling
jsg [Wed, 15 Feb 2023 10:33:03 +0000 (10:33 +0000)]
drm/i915: Fix VBT DSI DVO port handling

From Ville Syrjala
7fa83855852e4b13dd05702569f3f0bb3e1c624d in linux-6.1.y/6.1.12
6a7ff131f17f44c593173c5ee30e2c03ef211685 in mainline linux

20 months agodrm/i915: Initialize the obj flags for shmem objects
jsg [Wed, 15 Feb 2023 10:31:23 +0000 (10:31 +0000)]
drm/i915: Initialize the obj flags for shmem objects

From Aravind Iddamsetty
baaed8c92927e2530a799daf2e909e6688dc2943 in linux-6.1.y/6.1.12
44e4c5684fcc82d8f099656c4ea39d9571e2a8ac in mainline linux

20 months agodrm/i915: Move fd_install after last use of fence
jsg [Wed, 15 Feb 2023 10:28:39 +0000 (10:28 +0000)]
drm/i915: Move fd_install after last use of fence

From Rob Clark
bfa700d12274bd3509b64fd545d1046eef89d564 in linux-6.1.y/6.1.12
251e8c5b1b1fadcc387a8e618c7437d330bdac3e in mainline linux

20 months agodrm/amd/display: fix cursor offset on rotation 180
jsg [Wed, 15 Feb 2023 10:26:50 +0000 (10:26 +0000)]
drm/amd/display: fix cursor offset on rotation 180

From Melissa Wen
5af27a53a881ed49e071875c10760b0fc13996a1 in linux-6.1.y/6.1.12
49d0555976f0972af68397ed996375c135b38ba7 in mainline linux

20 months agodrm/amd/display: properly handling AGP aperture in vm setup
jsg [Wed, 15 Feb 2023 10:23:28 +0000 (10:23 +0000)]
drm/amd/display: properly handling AGP aperture in vm setup

From Alex Deucher
8d81e1c6865e7114d9829b0f3c73a66cd3a0963d in linux-6.1.y/6.1.12
5c4e8c71d1202cd84d870e7e5cb8d6b52f9c3507 in mainline linux

20 months agodrm/amdgpu/smu: skip pptable init under sriov
jsg [Wed, 15 Feb 2023 10:21:11 +0000 (10:21 +0000)]
drm/amdgpu/smu: skip pptable init under sriov

From Jane Jian
4609e1773222670af42e71c3a7bbf0453438d74a in linux-6.1.y/6.1.12
c6ac406cd8ff610a2d5da298b1d3071acfcde7f0 in mainline linux

20 months agodrm/amdgpu/fence: Fix oops due to non-matching drm_sched init/fini
jsg [Wed, 15 Feb 2023 10:18:26 +0000 (10:18 +0000)]
drm/amdgpu/fence: Fix oops due to non-matching drm_sched init/fini

From Guilherme G. Piccoli
2bcbbef9cace772f5b7128b11401c515982de34b in linux-6.1.y/6.1.12
5ad7bbf3dba5c4a684338df1f285080f2588b535 in mainline linux