job [Sat, 3 Sep 2022 21:24:02 +0000 (21:24 +0000)]
Properly free() crl & auth tree in parser process
OK claudio@
mbuhl [Sat, 3 Sep 2022 21:16:51 +0000 (21:16 +0000)]
regen
kn [Sat, 3 Sep 2022 21:16:05 +0000 (21:16 +0000)]
Keep _fw2 in local md_consoleinfo() scope
Missed in previous
OK kettenis
mbuhl [Sat, 3 Sep 2022 21:13:48 +0000 (21:13 +0000)]
Make recvmmsg and sendmmsg look more alike. change the flag type
to int like other flag parameters, NetBSD uses unsigned int, FreeBSD
and Linux do int.
OK bluhm@
jsing [Sat, 3 Sep 2022 20:12:24 +0000 (20:12 +0000)]
Remove now unused IMPLEMENT_BLOCK_CIPHER macro.
Also remove various comments noting that it cannot be used for certain
block ciphers (which kinda defeats the purpose of having a generic
implementation in the first place).
benno [Sat, 3 Sep 2022 20:07:31 +0000 (20:07 +0000)]
Move the daemon() call in the parent process from after forking the
children to just before. That way the parent disasociates from its
controling terminal and shell, but not from its children.
Remove the dup2() bits that were copied from daemon() to solve the
problem that the children still had the stdio fds open. This is now
done in the parent earlier.
Remove the setsid() and setpgid(). It is unclear what their intent
was, but they dont seem to make sense, as daemon() covers this as well
and there seems to be no reason the cildren procs need to do that.
ok claudio@ bluhm@
jsing [Sat, 3 Sep 2022 20:06:43 +0000 (20:06 +0000)]
Mechanically expand IMPLEMENT_BLOCK_CIPHER macro.
Only change to generated assembly is due to EVPerror()'s use of line
numbers.
jsing [Sat, 3 Sep 2022 20:02:17 +0000 (20:02 +0000)]
Mechanically expand IMPLEMENT_BLOCK_CIPHER macro.
No change in generated assembly.
jsing [Sat, 3 Sep 2022 19:59:32 +0000 (19:59 +0000)]
Mechanically expand IMPLEMENT_BLOCK_CIPHER macro.
Only change to generated assembly is due to EVPerror()'s use of line
numbers.
CVS ----------------------------------------------------------------------
jsing [Sat, 3 Sep 2022 19:51:53 +0000 (19:51 +0000)]
Mechanically expand IMPLEMENT_BLOCK_CIPHER macro.
No change to generated assembly.
jsing [Sat, 3 Sep 2022 19:43:16 +0000 (19:43 +0000)]
Mechanically expand IMPLEMENT_BLOCK_CIPHER macro.
These macros make the ASN.1 macros seem sane - there are layers and layers
and layers here, which are hiding bugs.
No change to generated assembly.
Discussed with tb@
bluhm [Sat, 3 Sep 2022 19:22:19 +0000 (19:22 +0000)]
Use a mutex to update tcp_maxidle, tcp_iss, and tcp_now. This
removes pressure from the exclusive netlock in tcp_slowtimo().
Reading is done atomically. Ensure that the tcp_now value is read
only once per function to provide consistent time.
OK yasuoka@
jsing [Sat, 3 Sep 2022 19:15:23 +0000 (19:15 +0000)]
Rewrap some lines, no functional change.
jsing [Sat, 3 Sep 2022 19:14:25 +0000 (19:14 +0000)]
Tidy up asn1_c2i_primitive() slightly.
Rename some variables and consistently goto error.
ok tb@
jsing [Sat, 3 Sep 2022 19:11:45 +0000 (19:11 +0000)]
Avoid recycling ASN1_STRINGs when decoding ASN.1.
Rather than recycling an existing ASN1_STRING and changing its type, free
it and allocate a replacement. This simplifies the code and potentially
avoids bugs resulting from reuse.
ok tb@
kettenis [Sat, 3 Sep 2022 19:05:52 +0000 (19:05 +0000)]
Set AP power state. This is needed if U-Boot didn't do this for us.
kettenis [Sat, 3 Sep 2022 19:04:28 +0000 (19:04 +0000)]
Handle syslog messages and add support for setting the AP power state.
jsing [Sat, 3 Sep 2022 18:59:09 +0000 (18:59 +0000)]
Remove duplicate prototype that just snuck in.
jsing [Sat, 3 Sep 2022 18:54:36 +0000 (18:54 +0000)]
Ensure a constructed ASN.1 INTEGER is considered invalid when decoding.
jsing [Sat, 3 Sep 2022 18:52:18 +0000 (18:52 +0000)]
Ensure ASN.1 types are appropriately encoded.
Per X.690, some ASN.1 types must be primitive encoded, some must be
constructed and some may be either. Add this data to our types table
and check the encoding against this information when decoding.
ok tb@
mvs [Sat, 3 Sep 2022 18:48:49 +0000 (18:48 +0000)]
Move PRU_SOCKADDR request to (*pru_sockaddr)()
Introduce in{,6}_sockaddr() functions, and use them for all except tcp(4)
inet sockets. For tcp(4) sockets use tcp_sockaddr() to keep debug ability.
The key management and route domain sockets returns EINVAL error for
PRU_SOCKADDR request, so keep this behaviour for a while instead of make
pru_sockaddr handler optional and return EOPNOTSUPP.
ok bluhm@
jsing [Sat, 3 Sep 2022 18:45:51 +0000 (18:45 +0000)]
Provide c2i_ASN1_ENUMERATED_cbs() and call it from asn1_c2i_primitive().
This avoids asn1_c2i_primitive() from needing knowledge about the internals
of ASN1_INTEGER and ASN1_ENUMERATED.
ok tb@
mlarkin [Sat, 3 Sep 2022 18:17:15 +0000 (18:17 +0000)]
whitespace
kettenis [Sat, 3 Sep 2022 18:05:10 +0000 (18:05 +0000)]
Make sure we don't suspend if we have no way to wake up the machine.
The idea is that device drivers that can wake up the machine register
themselves by calling device_register_wakeup(). To prevent regressions
on amd64, we let acpi(4) register itself as a wakeup device if the AML
for the machine defines wakeup devices. This may be refined in the
future.
This diff will prevent people from suspending their arm64 SBCs without
having a way to wake them up. For now the only driver that registers
itself is axppmic(4), which means that at this moment only SBCs with
an allwinner A64 or H5 SoC will actually support suspend/resume.
ok mlarkin@, deraadt@
jsing [Sat, 3 Sep 2022 17:47:47 +0000 (17:47 +0000)]
Prepare to provide OPENSSL_cleanup.
OPENSSL_cleanup() cleans up and deallocates memory in use by the library.
There are a couple of use cases for this, primarily related to memory
leak testing. This will not be called automatically in LibreSSL, which
means that OpenSSL's OPENSSL_NO_INIT_ATEXIT is implied. If code wants to
clean up then they need to explicitly call this themselves.
ok tb@
tb [Sat, 3 Sep 2022 17:43:14 +0000 (17:43 +0000)]
New test coverage for RC4
From Joshua Sing
jsing [Sat, 3 Sep 2022 16:01:23 +0000 (16:01 +0000)]
Stop using CBIGNUM_it internal to libcrypto.
CBIGNUM_it is supposed to be the "clear bignum" or "secure" bignum - that
is one which zeros its memory after use and ensures that the constant time
flags are set... in LibreSSL we always do both of these things for BIGNUMs,
so just use BIGNUM_it instead.
ok tb@
jmc [Sat, 3 Sep 2022 15:59:04 +0000 (15:59 +0000)]
- rework the -f text to read better; ok job
- while here, wrap a long line
claudio [Sat, 3 Sep 2022 15:58:40 +0000 (15:58 +0000)]
Remove -lncurses from LDADD. It is not used. Also add DPADD line.
OK deraadt@
kettenis [Sat, 3 Sep 2022 15:48:16 +0000 (15:48 +0000)]
Allow ikbd(4) to become the console keyboard.
ok miod@
kn [Sat, 3 Sep 2022 15:46:20 +0000 (15:46 +0000)]
Fix passing explicit stage files
Every platform ought to set `stages', `stage1' and optionally `stage2'
in md_init(), otherwise passing explicit files results won't work as
`stages' is zero-initialised and no default path is set:
# installboot -nv wd0 ./ofwboot
usage: installboot [-nv] [-r root] disk [stage1]
installboot [-nv] -p disk
This is correct synopsis and ought to work, but macppc_installboot.c
(others, too) has an empty md_init(). Set stage bits to fix this:
# ./obj/installboot -nv wd0 ./ofwboot
Using / as root
would install bootstrap on /dev/rwd0c
using first-stage ./ofwboot
would copy ./ofwboot to /tmp/installboot.Ymmm6QU8OJ/ofwboot
Using `stage1' leads to a bit more cleanup since early MI installboot.c
handles `-r', i.e. write_filesystem() no longer has needs to do the
fileprefix() dance itself.
This makes regress/usr.sbin/installboot pass on macppc (while being lucky
or carrying miod's fix for the kernel disklabel race manifesting on vnd).
OK gkoehler
kettenis [Sat, 3 Sep 2022 15:29:43 +0000 (15:29 +0000)]
Allow suspend with root on sdmmc(4).
ok deraadt@
job [Sat, 3 Sep 2022 15:13:44 +0000 (15:13 +0000)]
Clarify warning
yasuoka [Sat, 3 Sep 2022 14:57:54 +0000 (14:57 +0000)]
When divert-reply is used, keep some pf states after pcb is dropped if
its local address is translated, to prevent its source port from being
reused. regress test by blumn.
ok blumn
job [Sat, 3 Sep 2022 14:41:47 +0000 (14:41 +0000)]
Don't doublecheck whether the RSC eContent Resourceblock contains inherit elements
The RSC ASN.1 templates make it impossible to pass an RFC3779-style inherit option
because of the use of ConstrainedIPAddressFamily and ConstrainedASIdentifiers.
OK tb@
job [Sat, 3 Sep 2022 14:40:09 +0000 (14:40 +0000)]
Introduce x509_any_inherit() for objects which may not have inherit elements
Unify conformance checking of Trust Anchors, ROAs, ASPAs, RSCs - none of which
may have any 'inherit' elements in the RFC 3779 IP/AS Resources extension of
the X509 certificate.
OK tb@
krw [Sat, 3 Sep 2022 13:59:25 +0000 (13:59 +0000)]
Add a new keyword to template files, 'raid', to allow the auto
allocation of RAID partitions.
Make both 'raid' and 'swap' keywords case insensitive.
Suggested by kn@
ok kn@ miod@
claudio [Sat, 3 Sep 2022 13:30:27 +0000 (13:30 +0000)]
Add the repoid of the cert in the cert struct. This way it is possible
to track the parent repository id of a publication point.
Nomenclature is confusing but not much we can do here.
OK tb@ job@
mvs [Sat, 3 Sep 2022 13:29:33 +0000 (13:29 +0000)]
Fix socket splicing between inet and inet6 sockets broken by PRU_CONTROL
request splitting to (*pru_control)().
ok bluhm@
tb [Sat, 3 Sep 2022 13:06:15 +0000 (13:06 +0000)]
Allow multiple X.509 locations
While currently everyone only uses a single location, the spec allows for
multiple locations ordered by preference. While rpki-client does not
support more than one location this should not be a fatal error. Instead,
pick the first location and warn if there are more than one.
ok job
tb [Sat, 3 Sep 2022 13:01:43 +0000 (13:01 +0000)]
Move non-inheritance check for BGPsec certs into cert_parse_pre()
ok claudio job (as part of a larger diff)
mbuhl [Sat, 3 Sep 2022 12:35:29 +0000 (12:35 +0000)]
regen
mbuhl [Sat, 3 Sep 2022 12:33:44 +0000 (12:33 +0000)]
add the sendmmsg syscall that allows sending multiple msghdrs at
once. libc, man page, and regress parts to come.
With input from jca@, bluhm@.
OK bluhm@
job [Sat, 3 Sep 2022 11:01:55 +0000 (11:01 +0000)]
Clarify timeout/deadline
claudio [Sat, 3 Sep 2022 09:22:25 +0000 (09:22 +0000)]
Move the repo lookup into queue_from_mft()
OK tb@
kettenis [Sat, 3 Sep 2022 08:44:56 +0000 (08:44 +0000)]
Add apldcms(4), a driver for the touchpad on M2 laptops. This driver
needs firmware that is provided on the ESP by the Asahi installer and
copied into /etc/firmware/apple by the OpenBSD installer.
ok tobhe@
kettenis [Sat, 3 Sep 2022 08:37:36 +0000 (08:37 +0000)]
Copy Apple touchpad firmware on machines that need it.
ok kn@, deraadt@
mbuhl [Sat, 3 Sep 2022 08:26:05 +0000 (08:26 +0000)]
Fix the failing libm/rint regress test by adding ieee754 implementations
for the rounding functions.
Input from kettenis@,
OK miod@
jmc [Sat, 3 Sep 2022 06:55:01 +0000 (06:55 +0000)]
use past tense for history;
jsg [Sat, 3 Sep 2022 05:44:04 +0000 (05:44 +0000)]
sync with arm64.html
ajacoutot [Fri, 2 Sep 2022 22:11:57 +0000 (22:11 +0000)]
Make rc_configtest behave like rc_pre and rc_post; i.e. don't define a default
function (each rc.d script is supposed to define its own if wanted).
This way, we can filter out the "configtest" action depending on whether the
function exists or not.
Adapt documentation.
tweak/ok kn@
claudio [Fri, 2 Sep 2022 21:56:45 +0000 (21:56 +0000)]
Introduce a deadline timer that aborts all repository syncs.
With this rpki-client has a chance to still finish and produce an output
even when a CA is excessivly slow and holds back progress.
With and OK benno@ tb@ and job@
cheloha [Fri, 2 Sep 2022 21:33:51 +0000 (21:33 +0000)]
vmd(8): compute i8254 read-back command latch from singular timestamp
The intent of the i8254 read-back command is (most likely) to permit
simultaneously latching two or three counters at once along with their
statuses.
To simulate this, we should compute olatch from one timestamp per
read-back command, not one timestamp per counter.
Improved with a tweak by dv@.
Link: https://marc.info/?l=openbsd-tech&m=166213670605453&w=2
ok dv@ mlarkin@
miod [Fri, 2 Sep 2022 20:06:55 +0000 (20:06 +0000)]
Constify nam2blk[], chrtoblktbl[] and octeon devmap[].
ok mpi@ millert@
claudio [Fri, 2 Sep 2022 19:14:04 +0000 (19:14 +0000)]
Fix over long lines
OK tb@ job@
claudio [Fri, 2 Sep 2022 19:10:36 +0000 (19:10 +0000)]
Use the abort commands when a repo timeout happens. This is cleaner
then just failing the repo fetch but leaving the backends running.
OK tb@
claudio [Fri, 2 Sep 2022 18:37:17 +0000 (18:37 +0000)]
Implement RRDP_ABORT, a message to abort a inflight RRDP request.
The abort is done in a way that waits for any inflight files or http
requests to finish before removing the rrdp state and before sending
the rrdp done message indicating failure.
OK tb@ and benno@
tb [Fri, 2 Sep 2022 18:08:43 +0000 (18:08 +0000)]
rrdp_new() need not return the struct
The only caller does nothing with it.
with/ok claudio
kettenis [Fri, 2 Sep 2022 17:54:42 +0000 (17:54 +0000)]
First attempt at supporting audio on machines with multiple speakers.
Probably needs more work as the device tree bindings evolve.
Note that speakers are currently disabled in the device tree for all
Apple Silicon machines except for the Mac mini.
dv [Fri, 2 Sep 2022 17:46:37 +0000 (17:46 +0000)]
vmm(4): add tracepoint for in/out handler.
Inserts a new static dt(4) tracepoint in vmm(4) to report details
on in/out instructions (direction, port, and data).
ok mlarkin@
claudio [Fri, 2 Sep 2022 17:39:51 +0000 (17:39 +0000)]
extra newline
kettenis [Fri, 2 Sep 2022 16:53:28 +0000 (16:53 +0000)]
Add a callback for setting the TDM slot used by an audio codec.
Implement this callback in tascodec(4) such that we can pick the audio
channel that it outputs. This will override the default which is to
downmix stereo input from TDM channels 0 and 1 to mono output.
ok ratchov@
tb [Fri, 2 Sep 2022 15:45:52 +0000 (15:45 +0000)]
Rewrite RMD-160 tests to be table-driven.
From Joshua Sing
cheloha [Fri, 2 Sep 2022 15:21:40 +0000 (15:21 +0000)]
wc(1): accelerate word counting
wc(1) counts a word whenever a whitespace byte is followed by a
non-whitespace byte. Because the state machine transition occurs
within the space of a single byte we don't need to use getline(3).
Counting words in a big buffer with read(2) is much faster. The
overhead varies with the length of a line, but for files with 60-100
byte lines, word counting is about twice as fast when we avoid
getline(3). In the pathological case where each line is a single
byte, word counting is about ten times as fast when we avoid
getline(3).
Link1: https://marc.info/?l=openbsd-tech&m=
163715995626532&w=2
Link2: https://marc.info/?l=openbsd-tech&m=
165956826103639&w=2
"Seems reasonable." deraadt@
job [Fri, 2 Sep 2022 15:09:19 +0000 (15:09 +0000)]
Move mkpath logic after checking for 'noop' to prevent creation of directories in -n mode
OK claudio@
krw [Fri, 2 Sep 2022 14:18:47 +0000 (14:18 +0000)]
Don't ignore an OpenBSD GPT partition just because the GPT says
it extends beyond the edge of the disk the GPT currently
inhabits. We only care if enough of it is addressable that a
disklabel is accessible.
Brings GPT handling of 'OpenBSD partitions extending too far'
into line with the MBR handling of the same situation.
jan [Fri, 2 Sep 2022 14:08:09 +0000 (14:08 +0000)]
Fix TSO large receive offloading in ix(4).
Without this diff it might happen that content of different
TCP connection get mixed up, when reading coalesced buffers
from the receive ring.
Thanks, for a lot of testing effort to mbuhl.
OK mbuhl@
tb [Fri, 2 Sep 2022 13:46:23 +0000 (13:46 +0000)]
Retire the old crap. Thanks, Joshua
tb [Fri, 2 Sep 2022 13:45:18 +0000 (13:45 +0000)]
Unhook the old md4 and md5 tests
tb [Fri, 2 Sep 2022 13:38:56 +0000 (13:38 +0000)]
Link new md test to regress.
tb [Fri, 2 Sep 2022 13:34:48 +0000 (13:34 +0000)]
New md4/md5 regress tests
These exercise MD4 and MD5 with the test vectors from RFCs 1320 and 1321.
From Joshua Sing <joshua () hypera ! dev>
mbuhl [Fri, 2 Sep 2022 13:23:33 +0000 (13:23 +0000)]
regen
tb [Fri, 2 Sep 2022 13:23:05 +0000 (13:23 +0000)]
Add two const
tb [Fri, 2 Sep 2022 13:21:32 +0000 (13:21 +0000)]
Make test tables static const and fix a style nit
kettenis [Fri, 2 Sep 2022 13:20:46 +0000 (13:20 +0000)]
Enable apldc(4), apldchidev(4), apldckbd(4) and aplrtk(4) here as well.
mbuhl [Fri, 2 Sep 2022 13:18:06 +0000 (13:18 +0000)]
add the recvmmsg syscall that allows receiving multiple msghdrs at
once. libc, man page, and regress parts to come.
With input from jca@, guenther@, bluhm@.
OK bluhm@
mvs [Fri, 2 Sep 2022 13:12:31 +0000 (13:12 +0000)]
Move PRU_CONTROL request to (*pru_control)().
The 'proc *' arg is not used for PRU_CONTROL request, so remove it from
pru_control() wrapper.
Split out {tcp,udp}6_usrreqs from {tcp,udp}_usrreqs and use them for
inet6 case.
ok guenther@ bluhm@
claudio [Fri, 2 Sep 2022 13:04:16 +0000 (13:04 +0000)]
Rework the rsync proc code. Use a proper queue of requests and enforce
the limit on that queue instead of stopping to read new messages.
This is needed to implement an abort request.
"There is not enough RB_TREE in this diff" tb@
mlarkin [Fri, 2 Sep 2022 12:46:18 +0000 (12:46 +0000)]
Reduce differences to amd64.
ok kettenis
krw [Fri, 2 Sep 2022 12:40:02 +0000 (12:40 +0000)]
Nuke TEMPLATE variable and just do it, like amd64/ramdisk_cd does.
ok deraadt@
krw [Fri, 2 Sep 2022 12:28:12 +0000 (12:28 +0000)]
Nuke variable partlba that is set but not used.
krw [Fri, 2 Sep 2022 12:24:26 +0000 (12:24 +0000)]
Relax GTP header validity check by allowing 1 sector size usable
LBA area (gh_lba_start == gh_lba_end) and allowing either or both
of gh_lba_start and gh_lba_end to exceed the size of the disk the
GPT currently inhabits.
Reduces false negatives and inappropriate fall through to MBR
spoofing and allows repair of otherwise valid GPT.
tb [Fri, 2 Sep 2022 11:47:25 +0000 (11:47 +0000)]
Simplify and clean up the ecdsa test a little. Use stdio instead of BIO
for output, use 'err' as a label and avoid some silly repetitions.
tb [Fri, 2 Sep 2022 11:18:09 +0000 (11:18 +0000)]
Retire old SHA tests
The old tests were incomplete, some of them had no license and the code
quality was questionable. The new tests by Joshua Sing cover what they
did (and additionally SHA-224 and SHA-384). Many thanks!
tb [Fri, 2 Sep 2022 11:16:03 +0000 (11:16 +0000)]
Unhook old SHA tests
tb [Fri, 2 Sep 2022 11:13:34 +0000 (11:13 +0000)]
Add tests for 1 million repeated "a"
From Joshua Sing
jsg [Fri, 2 Sep 2022 10:41:05 +0000 (10:41 +0000)]
sync
jsg [Fri, 2 Sep 2022 10:39:31 +0000 (10:39 +0000)]
drm/i915: Add new ADL-S pci id
From Jose Roberto de Souza
in drm-intel-next
jsg [Fri, 2 Sep 2022 10:34:43 +0000 (10:34 +0000)]
regen
jsg [Fri, 2 Sep 2022 10:34:07 +0000 (10:34 +0000)]
sync Intel ADL-S devices with Mesa git
adds 0x468b, removes 0x4691
miod [Fri, 2 Sep 2022 10:16:51 +0000 (10:16 +0000)]
Now that boot blocks are able to parse ufs2, there is no need to force ufs1
for the / file system. Brings landisk into the wonderful world of y2k38
compliance.
miod [Fri, 2 Sep 2022 10:15:35 +0000 (10:15 +0000)]
Add UFS2 support, with libsa for boot and with a specific ufs-and-ufs2-in-one
flavour for xxboot, due to its size constraints.
miod [Fri, 2 Sep 2022 10:14:02 +0000 (10:14 +0000)]
Add ufs2 to the list of filesystem, for the sake of boot blocks which do not
provide an explicit list of files to build in libsa.
miod [Fri, 2 Sep 2022 10:12:46 +0000 (10:12 +0000)]
Remove non-_KERNEL code path for division by zero. This will allow the
bootblocks to shrink a little.
florian [Fri, 2 Sep 2022 09:39:55 +0000 (09:39 +0000)]
Write /etc/resolv.conf in a more atomic manner.
There were few reports were /etc/resolv.conf would lose user-managed
lines, possibly caused by a system crash.
While here add a call to fsync(2) which might also help.
input otto
input & OK deraadt, kn
mlarkin [Fri, 2 Sep 2022 09:02:37 +0000 (09:02 +0000)]
Get the retguard region's phys address from pmap, instead of using linker
script symbols. This is needed since we don't have those symbols on all
archs where we want hibernate.
ok kettenis, and input and help from miod.
kn [Fri, 2 Sep 2022 08:13:03 +0000 (08:13 +0000)]
Add softraid(4) RAID 1C boot support
Equivalent of sys/arch/arm64/stand/efiboot/softraid_arm64.c r1.4:
(commitid: Ka484R3swI5xSRWO) "Add softraid(4) RAID 1C boot support".
Tell the boot loader to decrypt 1C like C volumes and check the number of
disks in 1C like in 1C volumes -- no new code rquired.
Tested on T4-2 guest domains
"Looks reasonable" kettenis
OK stsp
NB: While kernel and boot loader support root on softraid on sparc64,
installboot(8) still needs a pending fix for installations on multi-chunk
softraid volumes. Until then, the usual installation process will fail on
1C volumes and requires manual fixup.
krw [Fri, 2 Sep 2022 07:46:03 +0000 (07:46 +0000)]
Adopt a terser specification of an MBR partition table of
64 zeros, suggested by miod@ a while ago.
ok mlarkin@
benno [Fri, 2 Sep 2022 07:38:14 +0000 (07:38 +0000)]
Make newer mime type definitions take precedence over existing ones.
Patch from Ben Fuller <ben -AT- bvnf -DOT- space>,
helped along by florian@
ok florian@ and some mumblings from claudio who does not want okays in httpd.
deraadt [Fri, 2 Sep 2022 07:37:57 +0000 (07:37 +0000)]
openpty() family of functions use /dev/ptm PTMGET to open a master+slave fd
pair, and also provides their names. Internally, 3 NDINIT+namei operations
access /dev/[tp]ty[p-zP-T][0-9a-zA-Z], of these 2 followed unveil restrictions.
I argue if you unveil /dev/ptm, (and not the 372 other nodes), you still want
openpty() to provide you with working fd's, and the names, which the caller
will probably never open manually, because the fd's are given.
So change all NDINIT to use KERNELPATH, bypassing unveil.
ok semarie