openbsd
3 years agoMove AEAD handling into the new TLSv1.2 record layer.
jsing [Thu, 28 Jan 2021 17:00:38 +0000 (17:00 +0000)]
Move AEAD handling into the new TLSv1.2 record layer.

ok tb@

3 years agoShow when witness(4) has run out of lock order data entries.
visa [Thu, 28 Jan 2021 15:13:27 +0000 (15:13 +0000)]
Show when witness(4) has run out of lock order data entries.

This makes it clearer why lock order traces are sometimes not displayed.

Prompted by a question from, and OK anton@

3 years agoDrop tcp_trace() from SMALL_KERNEL builds to make room on amd64 floppy
visa [Thu, 28 Jan 2021 14:53:20 +0000 (14:53 +0000)]
Drop tcp_trace() from SMALL_KERNEL builds to make room on amd64 floppy

OK deraadt@

3 years agoBe consistent in not using parameter names for function prototypes;
mglocker [Thu, 28 Jan 2021 12:50:28 +0000 (12:50 +0000)]
Be consistent in not using parameter names for function prototypes;
I've missed two more cases in the previous commit.

3 years agoWhen an audio device is disconnected, drop MIDI clients controlling it
ratchov [Thu, 28 Jan 2021 11:17:58 +0000 (11:17 +0000)]
When an audio device is disconnected, drop MIDI clients controlling it

3 years agoDont attempt to drain disconnected clients
ratchov [Thu, 28 Jan 2021 11:15:31 +0000 (11:15 +0000)]
Dont attempt to drain disconnected clients

Clients are always drained before they disconnect, so this change
affects programs that die unexpectedly or loose thier network
connection.

Besides the bad style, this change fixes a theoretical bug when the
disconnected client slot could be recycled and given to another client
while it's being drained

3 years agoIn slot_xxx(), store dev pointer in local variable
ratchov [Thu, 28 Jan 2021 11:10:00 +0000 (11:10 +0000)]
In slot_xxx(), store dev pointer in local variable

This removes many redundant dereferences to obtain the dev pointer
from the slot stucture and makes the source slightly more readable.
No behavior change.

3 years agoUse everywhere the same pattern to handle fractional clock ticks
ratchov [Thu, 28 Jan 2021 11:06:58 +0000 (11:06 +0000)]
Use everywhere the same pattern to handle fractional clock ticks

No behavior change; this change is only to make the maths easier to
proofread

3 years agoMake slot_{attach,detach}() the opposite of each other
ratchov [Thu, 28 Jan 2021 11:06:07 +0000 (11:06 +0000)]
Make slot_{attach,detach}() the opposite of each other

No bahavior change. Now, slot_attach() moves slot's clock forward and
puts the slot on device list; slot_detach() does the opposite: remove
from device list and move clock backwards. This will allow to detach a
client and attach it later in exactly the same state.

3 years agoDrop unused dev_nctl() function and few unused prototypes
ratchov [Thu, 28 Jan 2021 11:02:28 +0000 (11:02 +0000)]
Drop unused dev_nctl() function and few unused prototypes

3 years agohandle "once" rules before letting pfsync defer tx of a packet.
dlg [Thu, 28 Jan 2021 09:37:20 +0000 (09:37 +0000)]
handle "once" rules before letting pfsync defer tx of a packet.

pfsync may want to defer the transmission of a packet. it does this so
it can try and get a state over to a peer firewall before a host may
send a reply to the peer, which would get dropped cos there's no
matching state.

i think the once rule processing should happen before that. the state
is created from the rule, whether the packet the state is for goes out
immediately or not shouldn't matter.

ok sashan@

3 years agoA new resolver can be created while we currently run a check with the
florian [Thu, 28 Jan 2021 07:34:34 +0000 (07:34 +0000)]
A new resolver can be created while we currently run a check with the
old configuration. We will then request another check that runs in
parallel to the old check. If the new check finishes earlier, the
current check result will be overwritten by an outdated check result
which is likely wrong.
While here fix some whitespace.
OK phessler

3 years agoRename bNumInterface to bNumInterfaces to fix build on arm64.
kurt [Thu, 28 Jan 2021 01:48:54 +0000 (01:48 +0000)]
Rename bNumInterface to bNumInterfaces to fix build on arm64.

3 years agoExtern dwarf2_loc_mark_labels. Fixes compiation with -fno-common.
mortimer [Thu, 28 Jan 2021 01:39:19 +0000 (01:39 +0000)]
Extern dwarf2_loc_mark_labels. Fixes compiation with -fno-common.

ok deraadt@

3 years agoExplicitly enable -fcommon, rather than untangle common symbols.
mortimer [Thu, 28 Jan 2021 01:36:32 +0000 (01:36 +0000)]
Explicitly enable -fcommon, rather than untangle common symbols.

ok deraadt@

3 years agoExtern privsep_process. Fixes compilation with -fno-common.
mortimer [Thu, 28 Jan 2021 01:20:37 +0000 (01:20 +0000)]
Extern privsep_process. Fixes compilation with -fno-common.

ok deraadt@

3 years agoExtern transport_list. Fixed compilation with -fno-common.
mortimer [Thu, 28 Jan 2021 01:18:44 +0000 (01:18 +0000)]
Extern transport_list. Fixed compilation with -fno-common.

ok deraadt@

3 years agoif the route resolved in pf_route is invalid, generate an icmp error.
dlg [Wed, 27 Jan 2021 23:53:35 +0000 (23:53 +0000)]
if the route resolved in pf_route is invalid, generate an icmp error.

of course this is limited to the !dup-to case.

ok sashan@ bluhm@

3 years agofix leak: was double allocating kex->session_id buffer
djm [Wed, 27 Jan 2021 23:49:46 +0000 (23:49 +0000)]
fix leak: was double allocating kex->session_id buffer

3 years agoupdate comment, SMALL was split into SMALL and NOSSL
sthen [Wed, 27 Jan 2021 22:27:41 +0000 (22:27 +0000)]
update comment, SMALL was split into SMALL and NOSSL

3 years agoUnveil ldapd. Follow recent precedent and elect to forego the unlinking of
rob [Wed, 27 Jan 2021 22:12:28 +0000 (22:12 +0000)]
Unveil ldapd. Follow recent precedent and elect to forego the unlinking of
some objects at shutdown thereby allowing for a tighter unveil.

Feedbackup from deraadt@ and martijn@.

OK deraadt@

3 years agoremove bogus key hack now that it's handled by libtls
eric [Wed, 27 Jan 2021 20:33:05 +0000 (20:33 +0000)]
remove bogus key hack now that it's handled by libtls

no objection claudio@
ok tb@ jsing@

3 years agowrap an overlong line
tb [Wed, 27 Jan 2021 20:16:58 +0000 (20:16 +0000)]
wrap an overlong line

3 years agosync
tb [Wed, 27 Jan 2021 18:01:55 +0000 (18:01 +0000)]
sync

3 years agoLink SSL_get_finished.3 to build.
tb [Wed, 27 Jan 2021 18:00:25 +0000 (18:00 +0000)]
Link SSL_get_finished.3 to build.

3 years agoWrite SSL_get_finished() documentation from scratch.
tb [Wed, 27 Jan 2021 17:59:57 +0000 (17:59 +0000)]
Write SSL_get_finished() documentation from scratch.

ok beck

3 years agoMerge SSL_set_hostflags documentation from OpenSSL 1.1.1i and
tb [Wed, 27 Jan 2021 17:57:40 +0000 (17:57 +0000)]
Merge SSL_set_hostflags documentation from OpenSSL 1.1.1i and
add HISTORY section.  This is currently ignored input next to
SSL_get0_peername() and will be unignored once the symbols are
made publicly visible in libssl.

3 years agoThe usb configuration descriptor parameter providing the number of
mglocker [Wed, 27 Jan 2021 17:28:19 +0000 (17:28 +0000)]
The usb configuration descriptor parameter providing the number of
available device interfaces is called 'bNumInterfaces'.

ok phessler@, thfr@ (who provided the man page diff)

3 years agoPromote nrules/maxrules to size_t and make sure they can't overflow.
millert [Wed, 27 Jan 2021 17:02:50 +0000 (17:02 +0000)]
Promote nrules/maxrules to size_t and make sure they can't overflow.
reallocarray(3) will fail if nmemb * size would overflow.
OK tb@ martijn@

3 years agomark up fixes; from lyndon
jmc [Wed, 27 Jan 2021 14:59:10 +0000 (14:59 +0000)]
mark up fixes; from lyndon

3 years agoword fix;
jmc [Wed, 27 Jan 2021 14:58:06 +0000 (14:58 +0000)]
word fix;

3 years agospelling fixes;
jmc [Wed, 27 Jan 2021 14:57:29 +0000 (14:57 +0000)]
spelling fixes;

3 years agoregen
jsg [Wed, 27 Jan 2021 10:48:44 +0000 (10:48 +0000)]
regen

3 years agoadd some more amdgpu devices
jsg [Wed, 27 Jan 2021 10:47:58 +0000 (10:47 +0000)]
add some more amdgpu devices

3 years agoFlush pending output before entering or exiting alternate screen rather
nicm [Wed, 27 Jan 2021 10:42:52 +0000 (10:42 +0000)]
Flush pending output before entering or exiting alternate screen rather
than leaking it, oss-fuzz issue 29959.

3 years agothis needs kex.h now
djm [Wed, 27 Jan 2021 10:15:08 +0000 (10:15 +0000)]
this needs kex.h now

3 years agomake ssh->kex->session_id a sshbuf instead of u_char*/size_t and
djm [Wed, 27 Jan 2021 10:05:28 +0000 (10:05 +0000)]
make ssh->kex->session_id a sshbuf instead of u_char*/size_t and
use that instead of global variables containing copies of it.
feedback/ok markus@

3 years agoremove global variable used to stash compat flags and use the
djm [Wed, 27 Jan 2021 09:26:53 +0000 (09:26 +0000)]
remove global variable used to stash compat flags and use the
purpose-built ssh->compat variable instead; feedback/ok markus@

3 years agoBe consistent in not using parameter names for function prototypes.
mglocker [Wed, 27 Jan 2021 08:32:46 +0000 (08:32 +0000)]
Be consistent in not using parameter names for function prototypes.

3 years agoDetermine available address families (and monitor when this changes)
florian [Wed, 27 Jan 2021 08:30:50 +0000 (08:30 +0000)]
Determine available address families (and monitor when this changes)
to configure libunbound accordingly. This way it no longer tries to
talk to IPv6 nameservers when only IPv4 is available and vice versa.
input deraadt
OK kn

3 years agodo better accounting of how many msix interrupts we want to use.
dlg [Wed, 27 Jan 2021 07:46:11 +0000 (07:46 +0000)]
do better accounting of how many msix interrupts we want to use.

ok jmatthew@

3 years agothese programs (with common ancestry) had a -fno-common problem related
deraadt [Wed, 27 Jan 2021 07:21:52 +0000 (07:21 +0000)]
these programs (with common ancestry) had a -fno-common problem related
to privsep_procid.
ok mortimer

3 years agosplit out extern and decl for -fno-common
deraadt [Wed, 27 Jan 2021 07:21:12 +0000 (07:21 +0000)]
split out extern and decl for -fno-common

3 years agomakemap does not need a common which it does not use.
deraadt [Wed, 27 Jan 2021 07:20:27 +0000 (07:20 +0000)]
makemap does not need a common which it does not use.

3 years agocommons used rather than externs, fix for -fno-common
deraadt [Wed, 27 Jan 2021 07:19:54 +0000 (07:19 +0000)]
commons used rather than externs, fix for -fno-common

3 years agofix -fno-common issues; ok mortimer
deraadt [Wed, 27 Jan 2021 07:18:41 +0000 (07:18 +0000)]
fix -fno-common issues; ok mortimer

3 years agofix -fno-common issues; ok mortimer
deraadt [Wed, 27 Jan 2021 07:18:16 +0000 (07:18 +0000)]
fix -fno-common issues; ok mortimer

3 years agonsd contains two yacc parsers, but predates "yacc -b". Instead a
deraadt [Wed, 27 Jan 2021 05:06:12 +0000 (05:06 +0000)]
nsd contains two yacc parsers, but predates "yacc -b".  Instead a
script renames the fields in one parser.  Three additional variables
(yysslim, yyssp, yystacksize) need to be renamed also.
Yes it is worse than stepping out of bed on the wrong side into cat vomit.

3 years agoshuffle externs (and definitions) around to satisfy -fno-common
deraadt [Wed, 27 Jan 2021 05:03:23 +0000 (05:03 +0000)]
shuffle externs (and definitions) around to satisfy -fno-common
ok mortimer

3 years agohave pf_route{,6} clear the pf_pdesc mbuf ref early for route-to/reply-to.
dlg [Wed, 27 Jan 2021 04:46:21 +0000 (04:46 +0000)]
have pf_route{,6} clear the pf_pdesc mbuf ref early for route-to/reply-to.

pf_route and pf_route6 are called to take over delivery of the
packet with route-to and reply-to instead of letting it get processed
normally. for the dup-to handling, it copies the mbuf but leaves
the original mbuf in place. pf_route takes over the packet by
clearing the mbuf pointer in the pf_pdesc struct. this diff moves
the clearing of that pointer to the start of the function, rather
than checking for dup-to again on the way out of the function.

i think this is better because it means that it's more robust in
the face of future code changes. even if that's not true, it's still
shorter code in a forwarding path.

ok sashan@ jmatthew@

3 years agodon't run copies of packets made by dup-to through pf_test.
dlg [Wed, 27 Jan 2021 03:02:06 +0000 (03:02 +0000)]
don't run copies of packets made by dup-to through pf_test.

dup-to is kind of like what you do with a span port, but is a bit
more fine grained. it copies packets in a connection out an interface
so that connection can be monitored. it doesnt make sense for pf
to see the copied packets and try to match or create new states for
them either. at best it needs config to stop pf seeing the copies
(eg, set skip on $dup_to_tgt_if). at worst it breaks the connections
you're monitoring because the states in pf get confused.

found while discussing larger route-to changes on tech@.

ok bluhm@ sashan@

3 years agokqueue: Fix termination assert
visa [Wed, 27 Jan 2021 02:58:03 +0000 (02:58 +0000)]
kqueue: Fix termination assert

When a kqueue file is closed, the kqueue can still have threads
scanning it. Consequently, kqueue_terminate() can see scan markers
in the event queue. These markers are removed when the scanning threads
leave the kqueue. Take this into account when checking the queue's
state, to avoid a panic when kqueue is closed from under a thread.

OK anton@

Reported-by: syzbot+757c60a2aa1125137cce@syzkaller.appspotmail.com
3 years agosatisfy -fno-common by duplicating deck chairs as required
deraadt [Wed, 27 Jan 2021 01:59:39 +0000 (01:59 +0000)]
satisfy -fno-common by duplicating deck chairs as required
ok mortimer millert

3 years agosatisfy -fno-common
deraadt [Wed, 27 Jan 2021 01:57:37 +0000 (01:57 +0000)]
satisfy -fno-common
similar to the approach used by mortimer

3 years agoLogical not bitwise or. ok djm@
dtucker [Wed, 27 Jan 2021 00:37:26 +0000 (00:37 +0000)]
Logical not bitwise or.  ok djm@

3 years agoAdd support for RSA-PSS PKCS1 signatures. Don't enable them by
tobhe [Tue, 26 Jan 2021 23:06:23 +0000 (23:06 +0000)]
Add support for RSA-PSS PKCS1 signatures.  Don't enable them by
default for now because of interoperability issues.

ok patrick@

3 years agoRecognize Apple Icestorm cores.
kettenis [Tue, 26 Jan 2021 23:02:18 +0000 (23:02 +0000)]
Recognize Apple Icestorm cores.

3 years agoFix build with -fno-common. OK deraadt@
millert [Tue, 26 Jan 2021 20:42:49 +0000 (20:42 +0000)]
Fix build with -fno-common.  OK deraadt@

3 years agoRewrap a comment line to fit into 80 columns.
tb [Tue, 26 Jan 2021 18:47:08 +0000 (18:47 +0000)]
Rewrap a comment line to fit into 80 columns.

3 years agozap a tab
tb [Tue, 26 Jan 2021 18:45:32 +0000 (18:45 +0000)]
zap a tab

3 years agoPrepare to provide SSL_set_hostflags()
tb [Tue, 26 Jan 2021 18:43:41 +0000 (18:43 +0000)]
Prepare to provide SSL_set_hostflags()

Yet another one of these X509_VERIFY_PARAM reacharounds into
libcrypto. Recently found in imapfilter, also used elsewhere.
Will be made publicly visible with the next minor bump.

ok jsing

3 years agoone variable was common, fixing what I can.
deraadt [Tue, 26 Jan 2021 18:25:07 +0000 (18:25 +0000)]
one variable was common, fixing what I can.
(the portable code in here is not in great shape, and I am ignoring it)

3 years agosatisfy -fno-common, by (1) copying all the variable decls from
deraadt [Tue, 26 Jan 2021 18:23:49 +0000 (18:23 +0000)]
satisfy -fno-common, by (1) copying all the variable decls from
indent_globs.h to indent.c, and (2) changing all the same decls in
indent_globs.h to be extern
ok mortimer

3 years agosatisfy -fno-common, by (1) copying all the variable decls from
deraadt [Tue, 26 Jan 2021 18:22:45 +0000 (18:22 +0000)]
satisfy -fno-common, by (1) copying all the variable decls from
indent_globs.h to indent.c, and (2) changing all the same decls in
indent_globs.h to be extern
ok mortimer

3 years agosockb variable is unused (and even worse, was common unused)
deraadt [Tue, 26 Jan 2021 18:22:35 +0000 (18:22 +0000)]
sockb variable is unused (and even worse, was common unused)

3 years agosatisfy -fno-common, by (1) copying all the variable decls from
deraadt [Tue, 26 Jan 2021 18:21:47 +0000 (18:21 +0000)]
satisfy -fno-common, by (1) copying all the variable decls from
indent_globs.h to indent.c, and (2) changing all the same decls in
indent_globs.h to be extern
ok mortimer

3 years agosatisfy -fno-common, by (1) copying all the variable decls from
deraadt [Tue, 26 Jan 2021 18:21:25 +0000 (18:21 +0000)]
satisfy -fno-common, by (1) copying all the variable decls from
indent_globs.h to indent.c, and (2) changing all the same decls in
indent_globs.h to be extern
ok mortimer

3 years agosatisfy -fno-common by repairing one enum decl
deraadt [Tue, 26 Jan 2021 18:19:43 +0000 (18:19 +0000)]
satisfy -fno-common by repairing one enum decl
ok mortimer

3 years agomove HostbasedAcceptedAlgorithms to the right place in alphabetical order
naddy [Tue, 26 Jan 2021 15:40:17 +0000 (15:40 +0000)]
move HostbasedAcceptedAlgorithms to the right place in alphabetical order

3 years agoMove sequence numbers into the new TLSv1.2 record layer.
jsing [Tue, 26 Jan 2021 14:22:19 +0000 (14:22 +0000)]
Move sequence numbers into the new TLSv1.2 record layer.

This allows for all of the DTLS sequence number save/restore code to be
removed.

ok inoguchi@ "whee!" tb@

3 years agoMove private key setup to a helper function with proper error
eric [Tue, 26 Jan 2021 12:51:22 +0000 (12:51 +0000)]
Move private key setup to a helper function with proper error
checking.  Only install the hash on the key if fake key is used,
and do it for EC keys too.

ok tb@ jsing@

3 years agoSome config changes require a restart of all resolvers even DEAD ones;
florian [Tue, 26 Jan 2021 12:46:46 +0000 (12:46 +0000)]
Some config changes require a restart of all resolvers even DEAD ones;
handle them like UNKNOWN.
Found the hard way by kn.

3 years agoWhen checking for available address family for AI_ADDRCONFIG consider
florian [Tue, 26 Jan 2021 12:27:28 +0000 (12:27 +0000)]
When checking for available address family for AI_ADDRCONFIG consider
the routing domain we are currently in. Otherwise we might end up with
address families that are not available in the current rdomain but in
others since getifaddrs(3) gives us all interface addresses in the
system.
Clue-bat & OK claudio, input & OK eric, OK kn

3 years agoremove test that's now bogus, as fullpkgpath takes precedence.
espie [Tue, 26 Jan 2021 12:13:21 +0000 (12:13 +0000)]
remove test that's now bogus, as fullpkgpath takes precedence.
thanks bluhm@

3 years agoRemove unused variables leftover from refactoring. ok djm@
dtucker [Tue, 26 Jan 2021 11:25:01 +0000 (11:25 +0000)]
Remove unused variables leftover from refactoring.  ok djm@

3 years agoAlways resize the original screen before copying when exiting the
nicm [Tue, 26 Jan 2021 09:32:52 +0000 (09:32 +0000)]
Always resize the original screen before copying when exiting the
alternate screen, GitHub issue 2536.

3 years agointrmap attribute missing from mcx; ok dlg
deraadt [Tue, 26 Jan 2021 08:01:09 +0000 (08:01 +0000)]
intrmap attribute missing from mcx; ok dlg

3 years agoRename HostbasedKeyTypes (ssh) and HostbasedAcceptedKeyTypes (sshd) to
dtucker [Tue, 26 Jan 2021 05:32:21 +0000 (05:32 +0000)]
Rename HostbasedKeyTypes (ssh) and HostbasedAcceptedKeyTypes (sshd) to
HostbasedAcceptedAlgorithms, which more accurately reflects its effect.
This matches a previous change to PubkeyAcceptedAlgorithms.  The previous
names are retained as aliases.  ok djm@

3 years agorefactor key constraint parsing in ssh-agent
djm [Tue, 26 Jan 2021 00:54:49 +0000 (00:54 +0000)]
refactor key constraint parsing in ssh-agent

Key constraints parsing code previously existed in both the "add regular
key" and "add smartcard key" path. This unifies them but also introduces
more consistency checking: duplicated constraints and constraints that
are nonsensical for a particular situation (e.g. FIDO provider for a
smartcard key) are now banned.

ok markus@

3 years agomore ssh-agent refactoring
djm [Tue, 26 Jan 2021 00:53:31 +0000 (00:53 +0000)]
more ssh-agent refactoring

Allow confirm_key() to accept an additional reason suffix

Factor publickey userauth parsing out into its own function and allow
it to optionally return things it parsed out of the message to its
caller.

feedback/ok markus@

3 years agomake struct hostkeys public; I have no idea why I made it opaque
djm [Tue, 26 Jan 2021 00:51:30 +0000 (00:51 +0000)]
make struct hostkeys public; I have no idea why I made it opaque
originally.

ok markus@

3 years agomove check_host_cert() from sshconnect,c to sshkey.c and refactor
djm [Tue, 26 Jan 2021 00:49:30 +0000 (00:49 +0000)]
move check_host_cert() from sshconnect,c to sshkey.c and refactor
it to make it more generally usable and testable.

ok markus@

3 years agouse recallocarray to allocate the agent sockets table; also clear
djm [Tue, 26 Jan 2021 00:47:47 +0000 (00:47 +0000)]
use recallocarray to allocate the agent sockets table; also clear
socket entries that are being marked as unused.

spinkle in some debug2() spam to make it easier to watch an agent
do its thing.

ok markus

3 years agofactor out common code in the agent client
djm [Tue, 26 Jan 2021 00:46:17 +0000 (00:46 +0000)]
factor out common code in the agent client

Add a ssh_request_reply_decode() function that sends a message to
the agent, reads and parses a success/failure reply.
Use it for all requests that only expect success/failure

ok markus@

3 years agoWe have this sequence in bridge(4) ioctl(2) path:
mvs [Mon, 25 Jan 2021 19:47:16 +0000 (19:47 +0000)]
We have this sequence in bridge(4) ioctl(2) path:

ifs = ifunit(req->ifbr_ifsname);
if (ifs == NULL) {
error = ENOENT;
break;
}
if (ifs->if_bridgeidx != ifp->if_index) {
error = ESRCH;
break;
}
bif = bridge_getbif(ifs);

This sequence repeats 8 times. Also we don't check value returned by
bridge_getbig() before use. Newly introduced bridge_getbig() function
replaces this sequence. This not only reduces duplicated code but also
makes `bif' dereference safe.

ok bluhm@

3 years agoGive machdep.c a thorough cleanup that is long overdue.
kettenis [Mon, 25 Jan 2021 19:37:17 +0000 (19:37 +0000)]
Give machdep.c a thorough cleanup that is long overdue.

ok patrick@

3 years agoRevert local diff now that we no longer use syslog logging in
florian [Mon, 25 Jan 2021 16:57:37 +0000 (16:57 +0000)]
Revert local diff now that we no longer use syslog logging in
libunbound.
OK phessler

3 years agoDisable logging to syslog for libunbound. We are not getting anything
florian [Mon, 25 Jan 2021 16:56:59 +0000 (16:56 +0000)]
Disable logging to syslog for libunbound. We are not getting anything
useful for us out of it and it can be quite noisy when we are missing
IPv4 or IPv6 addresses.
It is still available when logging to stderr when running with -d.
OK phessler

3 years agoResolve data toggle out of sync problem for ugen(4) and uhidev(4) devices
mglocker [Mon, 25 Jan 2021 14:14:42 +0000 (14:14 +0000)]
Resolve data toggle out of sync problem for ugen(4) and uhidev(4) devices
on xhci(4) controllers by clearing the interface endpoints before opening
the pipes.

Tested by Mikolaj Kucharski for ugen(4) and gnezdo@ for uhidev(4), plus
myself for both.

ok mpi@

3 years agoAdd the new function usbd_clear_endpoint_feature() which allows to issue
mglocker [Mon, 25 Jan 2021 14:05:57 +0000 (14:05 +0000)]
Add the new function usbd_clear_endpoint_feature() which allows to issue
an UR_CLEAR_FEATURE request on a specific endpoint address without the
need to have a pipe open to that endpoint.

From NetBSD, ok mpi@

3 years agophp.port.mk sets MODPHP_BUILDDEP=No by default now.
sthen [Mon, 25 Jan 2021 14:02:18 +0000 (14:02 +0000)]
php.port.mk sets MODPHP_BUILDDEP=No by default now.

3 years agoraise the max number of queues/interrupts to 16, up from 1.
dlg [Mon, 25 Jan 2021 12:27:42 +0000 (12:27 +0000)]
raise the max number of queues/interrupts to 16, up from 1.

jmatthew@ has tried this before, but hrvoje popovski experienced
breakage so it wasn't enabled. we've tightened the code up since
then so it's time to try again.

this diff has been tested by hrvoje popovski and myself
ok jmatthew@

3 years agoif the rx descriptor reports the rss hash, use it for the mbuf flowid.
dlg [Mon, 25 Jan 2021 11:11:22 +0000 (11:11 +0000)]
if the rx descriptor reports the rss hash, use it for the mbuf flowid.

ok jmatthew@

3 years agodon't lose the M_FLOWID flag if the ipv4 cksum is ok.
dlg [Mon, 25 Jan 2021 09:36:48 +0000 (09:36 +0000)]
don't lose the M_FLOWID flag if the ipv4 cksum is ok.

found while poking around with hrvoje popovski
yes jmatthew@

3 years agoAdjust code since bgpd added an extra argument to aspath_verify() to
claudio [Mon, 25 Jan 2021 09:17:33 +0000 (09:17 +0000)]
Adjust code since bgpd added an extra argument to aspath_verify() to
reject AS_SET segments. In bgpctl this is always off.
OK benno@

3 years agoRFC6472 discourages the use of AS_SET segements in ASPATH attributes.
claudio [Mon, 25 Jan 2021 09:15:23 +0000 (09:15 +0000)]
RFC6472 discourages the use of AS_SET segements in ASPATH attributes.
The main reason is that AS_SET does not play nice with RPKI ROA.

Introduce a per neighbor and global config option
    'reject as-set yes' and 'reject as-set no'
If set to yes received UPDATES with AS_SET segements are rejected.
This is done the same way other ASPATH soft-errors are handled. The UPDATE
is marked invalid and all prefixes are treated as withdraws.
`bgpctl show rib in error` can be used to show prefixes that where denied
and treated as withdraws because of errors.

By default this feature is off.

OK benno@

3 years agoFix wg(4) ioctl to be able to handle multiple wgpeers.
yasuoka [Mon, 25 Jan 2021 09:11:36 +0000 (09:11 +0000)]
Fix wg(4) ioctl to be able to handle multiple wgpeers.
Diff from Yuichiro NAITO.

ok procter

3 years agofix filtering on kstat unit numbers
dlg [Mon, 25 Jan 2021 06:55:59 +0000 (06:55 +0000)]
fix filtering on kstat unit numbers

3 years agor1.102 forgot to tweak the "redistribute rtlabel" part of the grammar.
dlg [Mon, 25 Jan 2021 06:16:38 +0000 (06:16 +0000)]
r1.102 forgot to tweak the "redistribute rtlabel" part of the grammar.

fixes "redistribute rtlabel foo" without "depend on".

3 years agomake ssh hostbased authentication send the signature algorithm in
djm [Mon, 25 Jan 2021 06:00:17 +0000 (06:00 +0000)]
make ssh hostbased authentication send the signature algorithm in
its SSH2_MSG_USERAUTH_REQUEST packets instead of the key type.
This make HostbasedAcceptedAlgorithms do what it is supposed to -
filter on signature algorithm and not key type.

spotted with dtucker@ ok markus@