openbsd
2 months agoIgnore internal function keys if they have not got an entry in the key
nicm [Fri, 23 Aug 2024 13:25:39 +0000 (13:25 +0000)]
Ignore internal function keys if they have not got an entry in the key
table.

2 months agoRemove unwanted trailing newlines from err/warn format strings.
anton [Fri, 23 Aug 2024 12:56:26 +0000 (12:56 +0000)]
Remove unwanted trailing newlines from err/warn format strings.

2 months agoCheck for exact match for layout name before looking for a prefix match.
nicm [Fri, 23 Aug 2024 10:19:06 +0000 (10:19 +0000)]
Check for exact match for layout name before looking for a prefix match.

2 months agoRemove use of CSP/LMK in pkcs12 create/verify tests
tb [Fri, 23 Aug 2024 04:57:12 +0000 (04:57 +0000)]
Remove use of CSP/LMK in pkcs12 create/verify tests

reminded by ... anton

2 months agoAs defined in the RFC, the SSH protocol has negotiable compression support
deraadt [Fri, 23 Aug 2024 04:51:00 +0000 (04:51 +0000)]
As defined in the RFC, the SSH protocol has negotiable compression support
(which is requested as the name "zlib"). Compression starts very early in
the session.
Relative early in OpenSSH lifetime, privsep was added to sshd, and this
required a shared-memory hack so the two processes could see what was going
on in the dataflow.  This shared-memory hack was soon recognized as a tremendous
complexity risk, because it put libz (which very much trusts it's memory)
in a dangerous place, and a new option ("zlib@openssh.com") was added begins
compression after authentication (aka delayed-compression).  That change
also permitted removal of the shared-memory hack.
Despite removal from the server, the old "zlib" support remained in the
client, to allow negotiation with non-OpenSSH daemons which lack the
delayed-compression option.
This commit deletes support for the older "zlib" option in the client.
It reduces our featureset in a small way, and encourages other servers
to move to a better design.
The SSH protocol is different enough that compressed-key-material attacks
like BEAST are unlikely, but who wants to take the chance?
We encourage other ssh servers who care about optional compression support
to add delayed-zlib support.  (Some already do "zlib@openssh.com")
ok djm markus

2 months agoadd rcsid markers so that we can visually see the flurry of commits in
deraadt [Fri, 23 Aug 2024 04:26:11 +0000 (04:26 +0000)]
add rcsid markers so that we can visually see the flurry of commits in
this area
ok mlarkin

2 months agoadd the dumbest #-comment parser for the Game_List parser, someone
deraadt [Fri, 23 Aug 2024 04:25:46 +0000 (04:25 +0000)]
add the dumbest #-comment parser for the Game_List parser, someone
else can improve this later.  My code is not substandard considering
what I saw.

2 months agoadd rcsid markers so that we can visually see the flurry of commits in
deraadt [Fri, 23 Aug 2024 04:21:18 +0000 (04:21 +0000)]
add rcsid markers so that we can visually see the flurry of commits in
this area
ok mlarkin

2 months agofix spelling of sequence
tb [Fri, 23 Aug 2024 04:19:40 +0000 (04:19 +0000)]
fix spelling of sequence

ok mglocker

2 months agoatc(6): update ORD marker beacon information
mlarkin [Fri, 23 Aug 2024 03:43:33 +0000 (03:43 +0000)]
atc(6): update ORD marker beacon information

ok deraadt@

2 months agoprobition is over. most people want want to laugh at lightly
deraadt [Fri, 23 Aug 2024 03:25:32 +0000 (03:25 +0000)]
probition is over.  most people want want to laugh at lightly
offensive things.
suggested it is time by millert

2 months agothese tests only print strerror result, which can be confusing to look up.
deraadt [Fri, 23 Aug 2024 02:49:20 +0000 (02:49 +0000)]
these tests only print strerror result, which can be confusing to look up.
print errno also.

2 months agomust use sh to run the script, because /usr/src may be noexec
deraadt [Fri, 23 Aug 2024 02:46:09 +0000 (02:46 +0000)]
must use sh to run the script, because /usr/src may be noexec

2 months agoFix KERN_AUDIO broken in rev 1.440.
mvs [Fri, 23 Aug 2024 01:31:04 +0000 (01:31 +0000)]
Fix KERN_AUDIO broken in rev 1.440.

2 months agoFALLTROUGH -> FALLTHROUGH
jsg [Fri, 23 Aug 2024 01:23:50 +0000 (01:23 +0000)]
FALLTROUGH -> FALLTHROUGH

2 months agoFALLTHROUHG -> FALLTHROUGH
jsg [Fri, 23 Aug 2024 01:19:33 +0000 (01:19 +0000)]
FALLTHROUHG -> FALLTHROUGH

2 months agocron: use strtonum() and tighter limits on step values
millert [Fri, 23 Aug 2024 00:58:04 +0000 (00:58 +0000)]
cron: use strtonum() and tighter limits on step values

Using strtonum() instead of atoi() gives us an extra layer of bounds
checking for free while parsing an entry.  This is in addition to
the existing bounds checking in set_range().  The step value is now
limited to the maximum range for an entry.  If the field consists
of a range, the step must not be larger than the difference between
the high and low parts of the range.  OK deraadt@

2 months agoChange SIGCHLD handler to just set a flag.
millert [Fri, 23 Aug 2024 00:43:34 +0000 (00:43 +0000)]
Change SIGCHLD handler to just set a flag.
We already call reap_kids() in multiple event loops so there is no
need to call waitpid() inside the handler itself.
OK denis@ deraadt@

2 months agosntrup761x25519-sha512 now has an IANA codepoint assigned,
djm [Thu, 22 Aug 2024 23:11:30 +0000 (23:11 +0000)]
sntrup761x25519-sha512 now has an IANA codepoint assigned,
so we can make the algorithm available without the @openssh.com
suffix too. ok markus@ deraadt@

2 months agoPKCS12_create(3): remove Xr to EVP_PKEY_add1_attr(3)
tb [Thu, 22 Aug 2024 12:26:01 +0000 (12:26 +0000)]
PKCS12_create(3): remove Xr to EVP_PKEY_add1_attr(3)

This API family has been neutered and will be removed in the next bump.
Further cross references will be untangled in the future.

2 months agoGarbage collect unused attributes member from EVP_PKEY
tb [Thu, 22 Aug 2024 12:24:24 +0000 (12:24 +0000)]
Garbage collect unused attributes member from EVP_PKEY

ok miod

2 months agoRemove copy_bag_attr()
tb [Thu, 22 Aug 2024 12:22:42 +0000 (12:22 +0000)]
Remove copy_bag_attr()

It is no longer possible to set an attribute on an EVP_PKEY, so this
code is dead.

ok miod

2 months agoNeuter EVP_PKEY_add1_attr_by_NID()
tb [Thu, 22 Aug 2024 12:21:07 +0000 (12:21 +0000)]
Neuter EVP_PKEY_add1_attr_by_NID()

The last consumer in openssl(1) pkcs12 has been removed, so we no longer
need this function.

ok miod

2 months agoopenssl: adjust manual for LMK and CSP removal
tb [Thu, 22 Aug 2024 12:15:07 +0000 (12:15 +0000)]
openssl: adjust manual for LMK and CSP removal

2 months agoopenssl pkcs12: remove support for LMK and CSP attributes
tb [Thu, 22 Aug 2024 12:14:33 +0000 (12:14 +0000)]
openssl pkcs12: remove support for LMK and CSP attributes

Documentation on what the Microsoft-specific local machine keyset and the
cryptographic service provider are actually good for is hard to find. For
some reason (perhaps one million and two arguments for PKCS12_create() was
considered two too many) these hang off the EVP_PKEY in the attributes
member, which serves no other purpose.

Every use of EVP_PKEY (of which there are far too many) pays extra memory
taxes for this fringe use case. This complication is not worth it.

ok miod

2 months agoFix merge of bounce buffer segments in amd64 bus dma.
bluhm [Thu, 22 Aug 2024 11:36:24 +0000 (11:36 +0000)]
Fix merge of bounce buffer segments in amd64 bus dma.

If the physical pages are contiguous, _bus_dmamap_load_buffer()
tries to merge the segments.  In case of mbuf chains, it can happen
that the physical bounce buffers are contiguous, but the virtual
addresses of mbuf m_data are not.  Then during transmit _bus_dmamap_sync()
tries to copy segments where it cannot access the virtual source
address which is mapped in a different mbuf.  So if bounce buffers
are used, physical and virtual buffer must be contigous, to merge
a segment.
While there, split check and decrement of variable i in a for loop
to make the code readable.

with and OK hshoexer@

2 months agoUnlock unlock ipip_sysctl().
mvs [Thu, 22 Aug 2024 10:58:31 +0000 (10:58 +0000)]
Unlock unlock ipip_sysctl().

- IPIPCTL_ALLOW - atomically accessed integer;
- IPIPCTL_STATS - per-CPU counters;

In ipip_input() load `ipip_allow' value to `ipip_allow_local' and pass
it down to ipip_input_if() as `allow' arg.

ok bluhm

2 months agoUse aes128-ctr for MAC tests since default has implicit MAC.
dtucker [Thu, 22 Aug 2024 10:21:02 +0000 (10:21 +0000)]
Use aes128-ctr for MAC tests since default has implicit MAC.
Also verify that the Cipher or MAC we intended to use is actually the one
selected during the test.

2 months agoIntroduce sysctl_securelevel() to modify `securelevel' mp-safe. Keep
mvs [Thu, 22 Aug 2024 10:08:25 +0000 (10:08 +0000)]
Introduce sysctl_securelevel() to modify `securelevel' mp-safe. Keep
KERN_SECURELVL locked until existing `securelevel' checks became moved
out of kernel lock.

Make sysctl_securelevel_int() mp-safe by using atomic_load_int(9) to
unlocked read-only access for `securelevel'.

Unlock KERN_ALLOWDT. `allowdt' is the atomically accessed integer used
only once in dtopen().

ok mpi

2 months agosync manual section numbers
jsg [Thu, 22 Aug 2024 10:00:16 +0000 (10:00 +0000)]
sync manual section numbers

2 months agoClear overlay when command prompt is entered. Also fix some spacing in
nicm [Thu, 22 Aug 2024 09:05:51 +0000 (09:05 +0000)]
Clear overlay when command prompt is entered. Also fix some spacing in
man page pointed out by jmc.

2 months agoFix answer.
florian [Thu, 22 Aug 2024 08:44:22 +0000 (08:44 +0000)]
Fix answer.

2,$-1g/^/.,.1j does not combine every even-numbered line with the next
odd-numbered line. One correct way is 2,$-1g/^/.,+1j

Pointed out by ed1conf on mastodon.

2 months agoospfd: fix whitespace error introduced in previous
tb [Thu, 22 Aug 2024 08:34:51 +0000 (08:34 +0000)]
ospfd: fix whitespace error introduced in previous

2 months agolldb: shut up the warning message on quit debugging kernel core file.
asou [Thu, 22 Aug 2024 08:22:13 +0000 (08:22 +0000)]
lldb: shut up the warning message on quit debugging kernel core file.

The diff from Yuichiro NAITO.

ok yasuoka

2 months agoinet_pton returns 0 and -1 for error.
florian [Thu, 22 Aug 2024 08:17:54 +0000 (08:17 +0000)]
inet_pton returns 0 and -1 for error.

Adjust the error check that is now wrong after the inet_aton -> inet_pton
conversion.

Noticed by & OK bluhm.
OK tb

2 months agoMechanically change inet_aton to inet_pton.
florian [Thu, 22 Aug 2024 07:56:47 +0000 (07:56 +0000)]
Mechanically change inet_aton to inet_pton.

npppd does not document that it would accept truncated or otherwise
not fully spelled out IPv4 addresses.

ok yasuoka

2 months agoCharles III succeeded Elizabeth II in the 21st century.
florian [Thu, 22 Aug 2024 05:40:09 +0000 (05:40 +0000)]
Charles III succeeded Elizabeth II in the 21st century.

OK jmc, mlarkin

2 months agoShort Ctrl keys like ^A need to be converted to lowercase so they end up
nicm [Thu, 22 Aug 2024 05:39:55 +0000 (05:39 +0000)]
Short Ctrl keys like ^A need to be converted to lowercase so they end up
as 'a'|KEYC_CTRL to match the new internal representation. Problem
reported by naddy@.

2 months agospell(1) does not need to be told about 'woops' in this case
mlarkin [Thu, 22 Aug 2024 04:59:19 +0000 (04:59 +0000)]
spell(1) does not need to be told about 'woops' in this case

I guess someone added this word to spell.ok but it wasn't needed. woops.

ok deraadt

2 months agounrequired header file in vmm_machdep.c
mlarkin [Thu, 22 Aug 2024 04:53:07 +0000 (04:53 +0000)]
unrequired header file in vmm_machdep.c

2 months agoGenerate areas datfile from /usr/share/misc/na.phone during make.
mglocker [Thu, 22 Aug 2024 04:34:19 +0000 (04:34 +0000)]
Generate areas datfile from /usr/share/misc/na.phone during make.

A deraadt/mglocker production

2 months agoAdd Tribonacci number sequence.
mglocker [Thu, 22 Aug 2024 04:29:40 +0000 (04:29 +0000)]
Add Tribonacci number sequence.

Proposed by Anisja Mayer (math. student from Basel).

ok deraadt@

2 months agothe hup bug is incorrectly described; ok mlarkin
deraadt [Thu, 22 Aug 2024 03:10:42 +0000 (03:10 +0000)]
the hup bug is incorrectly described; ok mlarkin

2 months agounfair to discriminate against 0
deraadt [Wed, 21 Aug 2024 20:20:24 +0000 (20:20 +0000)]
unfair to discriminate against 0

2 months agoFix typo.
mglocker [Wed, 21 Aug 2024 19:39:09 +0000 (19:39 +0000)]
Fix typo.

ok deraadt@

2 months agoIncrease maximum Signed Object size to 8MB
job [Wed, 21 Aug 2024 19:35:31 +0000 (19:35 +0000)]
Increase maximum Signed Object size to 8MB

OK tb@ claudio@

2 months agoAn area is either a decimal number or an IPv4 address.
florian [Wed, 21 Aug 2024 15:18:47 +0000 (15:18 +0000)]
An area is either a decimal number or an IPv4 address.

This lets us replace inet_aton with inet_pton since we do not need
inet_aton's flexibility.

phessler, sthen and Tom Smyth all confirm that they never came across
a different way of specifying an area.

OK claudio, deraadt

2 months agoMechanically replace inet_aton with inet_pton.
florian [Wed, 21 Aug 2024 15:18:00 +0000 (15:18 +0000)]
Mechanically replace inet_aton with inet_pton.
OK claudio, deraadt

2 months agoAn area is either a decimal number or an IPv4 address.
florian [Wed, 21 Aug 2024 15:16:56 +0000 (15:16 +0000)]
An area is either a decimal number or an IPv4 address.

This lets us replace inet_aton with inet_pton since we do not need
inet_aton's flexibility.

phessler, sthen and Tom Smyth all confirm that they never came across
a different way of specifying an area.

OK claudio, deraadt

2 months agoGet rid of inet_aton
florian [Wed, 21 Aug 2024 15:00:25 +0000 (15:00 +0000)]
Get rid of inet_aton

OK deraadt

2 months agoUse getaddrinfo to lookup names and parse IP addresses.
florian [Wed, 21 Aug 2024 14:59:49 +0000 (14:59 +0000)]
Use getaddrinfo to lookup names and parse IP addresses.

OK deraadt

2 months agoMechanically change inet_aton to inet_pton.
florian [Wed, 21 Aug 2024 14:58:14 +0000 (14:58 +0000)]
Mechanically change inet_aton to inet_pton.

OK claudio, deraadt

2 months agoMechanically change inet_aton to inet_pton.
florian [Wed, 21 Aug 2024 14:57:05 +0000 (14:57 +0000)]
Mechanically change inet_aton to inet_pton.

This could use getaddrinfo in places but the code is just too crufty
and my joo janta's turned black immediately.

OK deraadt

2 months agoget rid of inet_aton
florian [Wed, 21 Aug 2024 14:55:17 +0000 (14:55 +0000)]
get rid of inet_aton
OK deraadt

2 months agoThis quiz is pretty old. Don't know when this happened, but it seems
deraadt [Wed, 21 Aug 2024 14:08:19 +0000 (14:08 +0000)]
This quiz is pretty old.  Don't know when this happened, but it seems
the wellknown name of "tupa inca yupanqui" has evolved to be either
"topa inca yupanqui" or "tupac inca yupanqui", so accept those answers.
research by claudio, inhouse inca expert

2 months agoaccept capital letter answers
deraadt [Wed, 21 Aug 2024 13:44:17 +0000 (13:44 +0000)]
accept capital letter answers

2 months agoReformat gprof.callg to be more compact so that it annoys a bit less when
claudio [Wed, 21 Aug 2024 13:29:09 +0000 (13:29 +0000)]
Reformat gprof.callg to be more compact so that it annoys a bit less when
looking at gprof output.
OK deraadt@

2 months agoalpbetically sort the configuration choices
deraadt [Wed, 21 Aug 2024 13:23:31 +0000 (13:23 +0000)]
alpbetically sort the configuration choices

2 months agodifficult sequence for a mathematician to identify
deraadt [Wed, 21 Aug 2024 13:22:29 +0000 (13:22 +0000)]
difficult sequence for a mathematician to identify

2 months agoRevert previous. It was committed mistakenly.
mvs [Wed, 21 Aug 2024 12:53:36 +0000 (12:53 +0000)]
Revert previous. It was committed mistakenly.

2 months agoremove unused vars and structs
jsg [Wed, 21 Aug 2024 11:24:12 +0000 (11:24 +0000)]
remove unused vars and structs

2 months agoUse getaddrinfo(3) to resolve hostnames / ip addresses.
florian [Wed, 21 Aug 2024 10:35:12 +0000 (10:35 +0000)]
Use getaddrinfo(3) to resolve hostnames / ip addresses.

OK dlg

2 months agoSome awks won't match on the \r so delete it instead.
dtucker [Wed, 21 Aug 2024 10:33:27 +0000 (10:33 +0000)]
Some awks won't match on the \r so delete it instead.
Fixes regress in portable on, eg Solaris.

2 months agoRead events from all PCBs as long as there's no error.
mpi [Wed, 21 Aug 2024 09:27:37 +0000 (09:27 +0000)]
Read events from all PCBs as long as there's no error.

From Christian Ludwig.

2 months agoWe can fully spell out IP addresses in config files.
florian [Wed, 21 Aug 2024 09:20:36 +0000 (09:20 +0000)]
We can fully spell out IP addresses in config files.

OK claudio

2 months agos/inet_aton/inet_pton/
florian [Wed, 21 Aug 2024 09:19:55 +0000 (09:19 +0000)]
s/inet_aton/inet_pton/

OK claudio

2 months agoMechanically change inet_aton to inet_pton.
florian [Wed, 21 Aug 2024 09:18:47 +0000 (09:18 +0000)]
Mechanically change inet_aton to inet_pton.

OK claudio

2 months agoWe can fully spell out IP addresses in config files.
florian [Wed, 21 Aug 2024 09:17:56 +0000 (09:17 +0000)]
We can fully spell out IP addresses in config files.

OK claudio

2 months agoUse inet_pton to parse ext-communities with an IPv4 address.
florian [Wed, 21 Aug 2024 09:17:05 +0000 (09:17 +0000)]
Use inet_pton to parse ext-communities with an IPv4 address.

No need for inet_aton's flexibility.
OK claudio

2 months agoUnlock ipip_sysctl().
mvs [Wed, 21 Aug 2024 07:36:12 +0000 (07:36 +0000)]
Unlock ipip_sysctl().

- IPIPCTL_ALLOW - atomically accessed integer;
- IPIPCTL_STATS - per-CPU counters;

ok bluhm

2 months agoImport regenerated moduli.
dtucker [Wed, 21 Aug 2024 07:06:27 +0000 (07:06 +0000)]
Import regenerated moduli.

2 months agoUse curve25519-sha256 kex where possible. Except where we're explicitly
dtucker [Wed, 21 Aug 2024 06:59:08 +0000 (06:59 +0000)]
Use curve25519-sha256 kex where possible.   Except where we're explicitly
testing a different kex, use curve25519-sha256 since it's faster than the
default and supported even when configured without OpenSSL.  Add a check
to ensure that the kex we intended to test is the one we actually tested.
Speeds test up by ~5%.

2 months agoMake sure ai_canonname is set when AI_CANONNAME was requested.
florian [Wed, 21 Aug 2024 05:53:10 +0000 (05:53 +0000)]
Make sure ai_canonname is set when AI_CANONNAME was requested.

We document, and posix requires, to return a NUL-terminated string on
a successful  call to getaddrinfo(3) when AI_CANONNAME was set.

If the canonical name cannot be determined, return the node name as
suggested by posix.

OK guenther

2 months agosatisfy the number version of the quiz; with mglocker
deraadt [Wed, 21 Aug 2024 05:08:54 +0000 (05:08 +0000)]
satisfy the number version of the quiz; with mglocker

2 months agoMention that load- and save-buffer can use stdin, from Ramon Fischer.
nicm [Wed, 21 Aug 2024 05:06:45 +0000 (05:06 +0000)]
Mention that load- and save-buffer can use stdin, from Ramon Fischer.

2 months agoAdd mirrored versions of the main-horizontal and main-vertical layouts where
nicm [Wed, 21 Aug 2024 05:03:13 +0000 (05:03 +0000)]
Add mirrored versions of the main-horizontal and main-vertical layouts where
the main pane is bottom or right instead of top or left, from Sherwyn Sen.

2 months agoAdd a simple random number sequence to the questions
deraadt [Wed, 21 Aug 2024 04:56:27 +0000 (04:56 +0000)]
Add a simple random number sequence to the questions
ok mglocker

2 months agoC-Space and Meta keys should not be translated in mode 1 extended keys.
nicm [Wed, 21 Aug 2024 04:55:57 +0000 (04:55 +0000)]
C-Space and Meta keys should not be translated in mode 1 extended keys.

2 months agoSet the default for extended-keys back to off because it appears emacs turns
nicm [Wed, 21 Aug 2024 04:37:42 +0000 (04:37 +0000)]
Set the default for extended-keys back to off because it appears emacs turns
the keys on but does not correctly handle them except in xterm (!). Also fix so
that off takes effect as expected.

2 months agoAdd lions.
mglocker [Wed, 21 Aug 2024 04:29:43 +0000 (04:29 +0000)]
Add lions.

suggested and ok jmc@

2 months agoAdd numeric morse codes.
mglocker [Wed, 21 Aug 2024 04:28:05 +0000 (04:28 +0000)]
Add numeric morse codes.

ok jmc@

2 months agoHippo and rhino babies are also calf's.
mglocker [Wed, 21 Aug 2024 04:27:02 +0000 (04:27 +0000)]
Hippo and rhino babies are also calf's.

ok jmc@

2 months agoAdd new substraction problem which difference results in the answer to the
mglocker [Wed, 21 Aug 2024 04:25:26 +0000 (04:25 +0000)]
Add new substraction problem which difference results in the answer to the
ultimate question of life, the universe, and everything.

ok jmc@

2 months agoRevamp extended keys support to more closely match xterm and support
nicm [Wed, 21 Aug 2024 04:17:09 +0000 (04:17 +0000)]
Revamp extended keys support to more closely match xterm and support
mode 2 as well as mode 1. From Stanislav Kljuhhin (GitHub issue 4038).

This changes tmux to always request mode 2 from parent terminal, change
to an unambiguous internal representation of keys, and adds an option
(extended-keys-format) to control the format similar to the xterm(1)
formatOtherKeys resource.

2 months agoIf binaries lack a pintable, execve() can now reject them at startup
deraadt [Wed, 21 Aug 2024 03:16:25 +0000 (03:16 +0000)]
If binaries lack a pintable, execve() can now reject them at startup
by returning EINVAL, rather than at runtime when the first system call
occurs (and then probably dumping core).  Let's find out if there are
any surprising exceptions we were unaware of (would need a custom linker,
or a non-crt0 runtime)
errno discussions with kettenis, sthen, millert. Vague concensus it is
worth trying now.

2 months agoWe do not need the PS_LIBCPIN and PS_PIN flag fields anymore, which were
deraadt [Wed, 21 Aug 2024 03:07:45 +0000 (03:07 +0000)]
We do not need the PS_LIBCPIN and PS_PIN flag fields anymore, which were
used during devlopment (for visibility). There is speculation claudio will
immediately use these bits for something else.

2 months agooffset rx buffers by ETHER_ALIGN so ip payloads are properly aligned.
dlg [Wed, 21 Aug 2024 01:22:31 +0000 (01:22 +0000)]
offset rx buffers by ETHER_ALIGN so ip payloads are properly aligned.

2 months agouse 2k clusters on the rx ring instead of 9k clusters.
dlg [Wed, 21 Aug 2024 01:17:50 +0000 (01:17 +0000)]
use 2k clusters on the rx ring instead of 9k clusters.

rge can chain mbufs from multiple rx descriptors together to build
a jumbo packet. 1514 byte packets are still the most common, so
avoiding the waste of most of a 9k cluster if we can still makes
sense.

2 months agosupport building a single packet out of multiple rx descriptors.
dlg [Wed, 21 Aug 2024 01:12:52 +0000 (01:12 +0000)]
support building a single packet out of multiple rx descriptors.

rge is a bit like intel nics where you can put small mbufs on the
rx ring and it will chain them together to make a jumbo packet.
except unlike intel you can use any sized mbufs for the rx descriptors.

this adds the handling of these packets to the rx path. we're still
putting 9k frames on the ring though.

2 months agobe more careful about bus_dmamap_syncs for rx ring descriptors.
dlg [Wed, 21 Aug 2024 00:56:58 +0000 (00:56 +0000)]
be more careful about bus_dmamap_syncs for rx ring descriptors.

this is very similar to the changes made for tx where we fill in
everything except the OWN flag on the rx descriptor, call bus_dmamap_sync
as a barrier for the hw, and then flip the ownership of the OWN
bit. this avoids the potential for the hw to see the own bit before
other things in the descriptor, such as the address and length of
the buffer.

while here, trim code that's not currently used. we don't currently
support rxing one packet by assembling buffers from multiple
descriptors, so stop fiddling with the queue mbuf head and tail
pointers. delete rge_discard_rxbuf() cos it looks like a leftover
from code which tries to reuse mbufs on the rx ring. we free mbufs
when there's an error and let the rxr stuff refill.

ok patrick@
no objections from kevlo@

2 months agoNow that we have dup2(), csh can use it instead of close()+dup().
guenther [Tue, 20 Aug 2024 23:40:39 +0000 (23:40 +0000)]
Now that we have dup2(), csh can use it instead of close()+dup().
Also, as used here, dup/dup2 will clear the close-on-exec flag, so
delete the superfluous fcntl(F_SETFD,0) calls

ok deraadt@

2 months agoUpdate handling of transfer rings.
patrick [Tue, 20 Aug 2024 21:24:15 +0000 (21:24 +0000)]
Update handling of transfer rings.

2 months agoIntroduce and update HAL SRNG ops.
patrick [Tue, 20 Aug 2024 21:23:18 +0000 (21:23 +0000)]
Introduce and update HAL SRNG ops.

2 months agomatch the config_activate_children() sequences found in other usb host
deraadt [Tue, 20 Aug 2024 16:24:50 +0000 (16:24 +0000)]
match the config_activate_children() sequences found in other usb host
controller drivers.

2 months agohaving a file called "OWNER" lying around in a directory of
deraadt [Tue, 20 Aug 2024 16:07:14 +0000 (16:07 +0000)]
having a file called "OWNER" lying around in a directory of
(effectively) public dommain software is weird.  put that info
into a better place.

2 months agoannotate signal race
deraadt [Tue, 20 Aug 2024 15:48:32 +0000 (15:48 +0000)]
annotate signal race

2 months agoFix whitespace in amd64 bus dma.
bluhm [Tue, 20 Aug 2024 15:30:29 +0000 (15:30 +0000)]
Fix whitespace in amd64 bus dma.

2 months agoUse msgbuf_queuelen() instead of accessing the queued member directly.
claudio [Tue, 20 Aug 2024 13:31:49 +0000 (13:31 +0000)]
Use msgbuf_queuelen() instead of accessing the queued member directly.
OK tb@

2 months agoUnlock KERN_MAXFILES.
mvs [Tue, 20 Aug 2024 13:29:25 +0000 (13:29 +0000)]
Unlock KERN_MAXFILES.

`maxfiles' is atomically accessed integer which is lockless and
read-only accessed in file descriptors layer.

lim_startup() called during kernel bootstrap, no need to
atomic_load_int() within.

ok mpi

2 months agoUpdate Geofeed reference
job [Tue, 20 Aug 2024 12:53:47 +0000 (12:53 +0000)]
Update Geofeed reference

RFC 9632 introduced additional constraints and requirements for RPKI-based
Geofeed authenticators (at my request).