sthen [Wed, 7 Jul 2021 13:29:42 +0000 (13:29 +0000)]
dedup
deraadt [Wed, 7 Jul 2021 13:19:42 +0000 (13:19 +0000)]
sync
martijn [Wed, 7 Jul 2021 11:21:55 +0000 (11:21 +0000)]
Mention that there are alternatives for ERE '+' and '?' in BRE.
OK kn@, millert@
sthen [Wed, 7 Jul 2021 09:51:28 +0000 (09:51 +0000)]
sync
stsp [Wed, 7 Jul 2021 09:47:40 +0000 (09:47 +0000)]
Support MCC update response used by newer iwm(4) firmware, and verify
the size of the response we receive for MCC_UPDATE commands (even
though we aren't doing anything with this response yet).
stsp [Wed, 7 Jul 2021 09:13:50 +0000 (09:13 +0000)]
Support the SoC configuration command used by newer iwm(4) 9560 firmware.
stsp [Wed, 7 Jul 2021 09:06:23 +0000 (09:06 +0000)]
Support the critical temperature detection feature of newer iwm(4) firmware.
stsp [Wed, 7 Jul 2021 08:52:54 +0000 (08:52 +0000)]
Parse CMD_VERSION TLV found in iwm(4) fw images and add iwm_lookup_cmd_ver().
It seems Intel saw a risk of eventually running out of firmware capability
support and API support flags, so yet another such mechanism was added.
If you want to know which variant of a command needs to be used there are
now at least three places to check for related information.
stsp [Wed, 7 Jul 2021 08:32:00 +0000 (08:32 +0000)]
Support a new variant of the binding command required by new iwm(4) firmware.
stsp [Wed, 7 Jul 2021 08:21:31 +0000 (08:21 +0000)]
Support the larger phy context command required by newer iwm(4) firmware.
stsp [Wed, 7 Jul 2021 08:13:37 +0000 (08:13 +0000)]
Recognize and ignore new debug sections found in latest iwm(4) 9k fw images.
stsp [Wed, 7 Jul 2021 08:05:11 +0000 (08:05 +0000)]
Support a new time quota command required for newer iwm(4) firmware.
mestre [Wed, 7 Jul 2021 07:28:56 +0000 (07:28 +0000)]
Hoist whole TLS initialization instead of loading the cert/key into memory first
and setting them up later on, nowadays it's possible to do both steps in just
one go through tls_config_set_{cert,key}_file(3) so use it and then init TLS.
No functional change.
OK kn@
jsg [Wed, 7 Jul 2021 02:38:21 +0000 (02:38 +0000)]
update drm to linux 5.10.47
hardware support changes include
inteldrm: better support for tiger lake
amdgpu: support for navi12, navi21 "sienna_cichlid", arcturus
amdgpu: support for cezanne "green sardine" ryzen 5000 apu
Thanks to the OpenBSD Foundation for sponsoring this work,
patrick@ for helping adapt rockchip drm, kettenis@ and mpi@
for uvm discussions and various testers.
bluhm [Tue, 6 Jul 2021 19:48:05 +0000 (19:48 +0000)]
Make test more verbose to see commands and error messages in output.
This makes it easier to debug if something fails. Add some ${SUDO}
in a way that this tests runs as root or regular user with and
without SUDO set.
patrick [Tue, 6 Jul 2021 19:09:57 +0000 (19:09 +0000)]
Shifts (<<) of more than 32 bits must be done on 64-bit values.
Reported by some grumpy old cheese lover
"looks reasonable" jsg@
ok kettenis@
jasper [Tue, 6 Jul 2021 18:01:34 +0000 (18:01 +0000)]
fix an m4 macro which showed up unexpanded. kill some trailing whitespace while here
schwarze [Tue, 6 Jul 2021 16:05:44 +0000 (16:05 +0000)]
document X509_signature_dump(3) and X509_signature_print(3)
stsp [Tue, 6 Jul 2021 15:53:33 +0000 (15:53 +0000)]
Fix iwx(4) against access points which use TKIP as the group cipher.
This was broken when key installation was moved into a task context.
Reported by and confirmed fixed by gnezdo@ and zxystd from OpenIntelWireless
ok gnezdo@
anton [Tue, 6 Jul 2021 15:24:02 +0000 (15:24 +0000)]
Make use of the existing bsd.regress.mk logic to flag root only targets,
allowing the actual tests to become less repetitive.
ok bluhm@
bluhm [Tue, 6 Jul 2021 13:19:57 +0000 (13:19 +0000)]
This regress checks that 400 singals are evenly distributed over 4
threads. It works well on idle 4 CPU machines, but with load or
fewer cores the distribution of the threads' signal handlers is
worse. Reduce the limit of signals that each thread must receive
from 75 to 50. This should prevent random test fails.
from Christian Ludwig
bluhm [Tue, 6 Jul 2021 11:50:34 +0000 (11:50 +0000)]
alarm(3) cannot fail, remove error handling.
suggested by millert@
bluhm [Tue, 6 Jul 2021 11:26:47 +0000 (11:26 +0000)]
Fix typo in comment.
schwarze [Tue, 6 Jul 2021 11:26:25 +0000 (11:26 +0000)]
Fix a bug in X509_print_ex(3).
If the user set nmflags == X509_FLAG_COMPAT and X509_NAME_print_ex(3)
failed, the error return value of 0 was misinterpreted as an indicator
of success, causing X509_print_ex(3) to ignore the error, continue
printing, and potentially return successfully even though not all
the content of the certificate was printed.
The X509_NAME_print_ex(3) manual page explains that this function
indicates failure by returning 0 if nmflags == X509_FLAG_COMPAT
and by returning -1 if nmflags != X509_FLAG_COMPAT.
That's definitely atrocious API design (witnessed by the
complexity of the code needed for correct error checking),
but changing the API contract and becoming incompatible
with OpenSSL would make matters even worse.
Note that just checking for <= 0 in all cases would not be correct
either because X509_NAME_print_ex(3) returns 0 to indicate that it
successfully printed zero bytes in some cases, for example when all
three of the following conditions hold:
1. nmflags != X509_FLAG_COMPAT
2. indent == 0 (which X509_print_ex(3) does use in some cases)
3. the name object is NULL or empty
I found the bug by code inspection and proposed an incomplete patch,
then jsing@ proposed this improved version of the patch.
OK jsing@.
kettenis [Tue, 6 Jul 2021 09:34:06 +0000 (09:34 +0000)]
Introduce CPU_IS_RUNNING() and us it in scheduler-related code to prevent
waiting on CPUs that didn't spin up. This will allow us to spin down
CPUs in the future to save power as well.
ok mpi@
nicm [Tue, 6 Jul 2021 08:26:00 +0000 (08:26 +0000)]
Improve error reporting when the tmux /tmp directory cannot be created
or used, GitHub issue 2765 from Uwe Kleine-Koenig.
nicm [Tue, 6 Jul 2021 08:18:38 +0000 (08:18 +0000)]
Forbid empty session names, GitHub issue 2758.
jmc [Tue, 6 Jul 2021 06:26:05 +0000 (06:26 +0000)]
sync man page with changes 1.2.3 -> 1.2.11. some changes:
- update references version number
- Z_NULL -> NULL; from millert
- terminate function synopses in main body with ";"; from tb
- NULL -> NUL; from tb
- update NAME and SYNOPSIS for added functions
- updated copyright, to match zlib.h
- simplify \-1 -> -1
ok tb
anton [Tue, 6 Jul 2021 05:59:20 +0000 (05:59 +0000)]
correct mailto
schwarze [Mon, 5 Jul 2021 17:57:16 +0000 (17:57 +0000)]
document i2a_ASN1_OBJECT(3)
bluhm [Mon, 5 Jul 2021 16:11:52 +0000 (16:11 +0000)]
Write ktrace output files onto NFS mounted file systems.
jca [Mon, 5 Jul 2021 13:56:24 +0000 (13:56 +0000)]
Don't use a non-standard empty statement in while loops
Empty statements accepted by ksh(1) pointed out by millert@, use of this
non-standard grammar in the installer pointed out by halex@.
ok halex@ millert@
millert [Mon, 5 Jul 2021 13:41:46 +0000 (13:41 +0000)]
Do not permit an empty list between "while" and "do".
This avoids a cpu loop for "while do done" and is consistent with
the behavior of AT&T ksh and most other shells. OK jca@ halex@
tobhe [Mon, 5 Jul 2021 12:03:42 +0000 (12:03 +0000)]
Print SA MTU if included in pfkey message.
ok bluhm@ sthen@ mpi@
tobhe [Mon, 5 Jul 2021 12:01:20 +0000 (12:01 +0000)]
Export tdb MTU to userland via SADB_GET. This helps debug path MTU
discovery issues with ESP in UDP.
ok bluhm@ sthen@ mpi@
ratchov [Mon, 5 Jul 2021 08:29:59 +0000 (08:29 +0000)]
Fix off-by-one array access when 64 channel stream is resampled
anton [Mon, 5 Jul 2021 05:50:19 +0000 (05:50 +0000)]
rename kd_put() -> kd_copy()
anton [Mon, 5 Jul 2021 05:49:30 +0000 (05:49 +0000)]
rename kd_put() -> kd_copy()
dlg [Mon, 5 Jul 2021 04:17:41 +0000 (04:17 +0000)]
etherbridge_map was way too clever, so simplify it.
the code tried to carry state from the quick smr based lookup through
to the actual map update under the mutex, but this led to refcnt
leaks, and logic errors. the simplification is that if the smr based
checks say the map needs updating, we prepare the update and then
forget what we learnt inside the smr critical section and redo them
under the mutex again.
entries in an etherbridge map are either in it or they aren't, so
we don't need to refcnt them. this means the thing that takes an
entry out of the map becomes directly responsible for destroy it,
so they can do the smr call or barrier directly rather than via a
refcnt.
found by hrvoje popovski while testing the stack running in parallel,
and fix tested by him too.
ok sashan@
dtucker [Mon, 5 Jul 2021 01:21:07 +0000 (01:21 +0000)]
Fix a couple of whitespace things. Portable already has these so this
removes two diffs between the two.
dtucker [Mon, 5 Jul 2021 01:16:46 +0000 (01:16 +0000)]
Order includes as per style(9). Portable already has these so this
removes a handful of diffs between the two.
dtucker [Mon, 5 Jul 2021 00:50:25 +0000 (00:50 +0000)]
Remove comment referencing now-removed RhostsRSAAuthentication. ok djm@
djm [Mon, 5 Jul 2021 00:25:42 +0000 (00:25 +0000)]
allow spaces to appear in usernames for local to remote, and scp -3
remote to remote copies. with & ok dtucker bz#1164
krw [Sun, 4 Jul 2021 20:37:11 +0000 (20:37 +0000)]
Fix legacy booting.
Don't flag the EFI SYS partition "DOSACTIVE". Leave that status
to the OpenBSD partition. Reverts part of r1.77.
Problem reported and fix tested by Mihai Popescu. Thanks!
Cluebats from deraadt@ and code bisecting by dv@.
ok kettenis@
sthen [Sun, 4 Jul 2021 19:22:31 +0000 (19:22 +0000)]
store provider ID in umb(4), and display it in ifconfig. OK stsp deraadt
Previously only the provider's display name was used. The text used depends
on how the SIM is configured and not just on the network in use (for example,
an MVNO SIM on another network will often display the MVNO's name rather
than that of the underlying network).
I have a SIM that roams to any network in my country - whichever network
it roams onto, the display name is the same, so you can't tell which
network you're really using. By printing the provider ID (in GSM-land this
is MCC+MNC) it's easy to lookup and check this.
As the provider was printed on the ifconfig line also showing subscriber-id
and ICCID it was already a bit long, and adding the provider-id there is
a bit too much, so move it to the output line showing APN, now looking like
: subscriber-id 2400xxxxxxxxxxx ICC-id 8946203xxxxxxxxxxxxx
: device EM7455 IMEI 01458xxxxxxxxxx firmware SWI9X30C_02.24.0
: APN key provider Tele2 IoT provider-id 23420
deraadt [Sun, 4 Jul 2021 17:51:20 +0000 (17:51 +0000)]
tb notices i did not add a copyright in 2011
tb [Sun, 4 Jul 2021 17:41:23 +0000 (17:41 +0000)]
Add missing RCS tags
deraadt [Sun, 4 Jul 2021 17:23:49 +0000 (17:23 +0000)]
sync
kn [Sun, 4 Jul 2021 15:58:51 +0000 (15:58 +0000)]
.Fl marks up options not arguments
POSIX says `-v' is an option, not an argument; options may take arguments,
as documented in getopt(3).
Programs may take arguments as well, e.g. "prog -v -- arg ...", but these
are usually marked up with `.Cm'.
Change "The arguments are as follows:" to "The options are as follows:" in
`.Fl's DESCRIPTION example to match what most of our base manuals say.
OK schwarze jmc
schwarze [Sun, 4 Jul 2021 15:38:08 +0000 (15:38 +0000)]
The mandoc(1) manual already mentions that -T man output mode
neither supports tbl(7) nor eqn(7) input.
If an input file contains such code anyway, tell the user
rather than failing an assert(3)ion.
Fixing a crash reported by Bjarni Ingi Gislason <bjarniig at rhi dot hi dot is>
in https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=901636 which the
Debian maintainer of mandoc, Michael at Stapelberg dot ch, forwarded to me.
deraadt [Sun, 4 Jul 2021 14:29:36 +0000 (14:29 +0000)]
sync
tb [Sun, 4 Jul 2021 14:29:14 +0000 (14:29 +0000)]
bump major after zlib update
tb [Sun, 4 Jul 2021 14:24:49 +0000 (14:24 +0000)]
Update libz to zlib 1.2.11 and reapply local patches as far as this
made sense.
Tested in snaps for a few days. deraadt helped with fitting things on
floppies and jmatthew found a bug I introduced.
ok deraadt
schwarze [Sun, 4 Jul 2021 12:56:27 +0000 (12:56 +0000)]
document X509_find_by_subject(3) and X509_find_by_issuer_and_serial(3)
schwarze [Sun, 4 Jul 2021 11:38:37 +0000 (11:38 +0000)]
Bugfix: when X509_NAME_dup(3) failed, X509_NAME_set(3) indicated success
even though it did not actually set the name.
Instead, indicate failure in this case.
This commit sneaks in a small, unrelated change in behaviour.
If the first argument of X509_NAME_set(3) was NULL, the function
used to return failure. Now it crashes the program by accessing
the NULL pointer, for compatibility with the same change in OpenSSL.
This merges the following two commits from the OpenSSL-1.1.1 branch,
which is still available under a free license:
1.
180794c5 Rich Salz Sep 3 11:33:34 2017 -0400
2.
c1c1783d Richard Levitte May 17 09:53:14 2018 +0200
OK tb@
semarie [Sat, 3 Jul 2021 17:51:59 +0000 (17:51 +0000)]
__realpath: removes LOCKLEAF from NDINIT.
The code doesn't doesn't need it: the returned vnode is released
immediately. The string path is built from the namei() call using
REALPATH, during directories traversal.
Without LOCKLEAF, calling vrele() only is enough if namei() found a
file, instead of calling VOP_UNLOCK() + vrele().
ok claudio@ mpi@
schwarze [Sat, 3 Jul 2021 17:04:51 +0000 (17:04 +0000)]
Document X509_NAME_set(3).
It is not particularly well-designed and sets a number of traps for the
unwary, but it is a public API function in both OpenSSL and LibreSSL
and used at various places.
jsing [Sat, 3 Jul 2021 16:06:44 +0000 (16:06 +0000)]
Do a first pass clean up of SSL_METHOD.
The num_ciphers, get_cipher_by_char and put_cipher_by_char function
pointers use the same function for all methods - call ssl3_num_ciphers()
directly, absorb ssl3_get_cipher_by_char() into SSL_CIPHER_find() and
remove the unused ssl3_put_cipher_by_char() code.
ok inoguchi@ tb@
jsing [Sat, 3 Jul 2021 16:04:29 +0000 (16:04 +0000)]
Garbage collect do_test_cipherlist().
This code no longer compiles and the equivalent test coverage has been
added to regress/lib/libssl/ciphers (and is actually run).
jsing [Sat, 3 Jul 2021 15:54:41 +0000 (15:54 +0000)]
Add test that ensures ssl3_ciphers[] is sorted by cipher id.
jsing [Sat, 3 Jul 2021 14:52:12 +0000 (14:52 +0000)]
Rewrite get_put_test() as cipher_find_test().
The get_cipher_by_char() and put_cipher_by_char() pointers are no longer
accessible on the SSL_METHOD (and soon will not even exist). Rewrite the
test to use SSL_CIPHER_find() instead.
schwarze [Sat, 3 Jul 2021 12:04:53 +0000 (12:04 +0000)]
fix a bug that resulted in incomplete testing:
end statements with ';' because ',' isn't enough
kettenis [Sat, 3 Jul 2021 10:21:38 +0000 (10:21 +0000)]
Avoid spinning with interrupts disabled.
ok patrick@
abieber [Sat, 3 Jul 2021 02:49:13 +0000 (02:49 +0000)]
Update to go-module.5 to reflect the modgo-gen-modules changes.
kettenis [Fri, 2 Jul 2021 20:39:25 +0000 (20:39 +0000)]
Avoid local variable definitions in the middle of code.
kettenis [Fri, 2 Jul 2021 19:55:00 +0000 (19:55 +0000)]
Avoid spinning with interrupts disabled in the rpi3.
tested by sthen@ (a while ago)
cheloha [Fri, 2 Jul 2021 17:16:20 +0000 (17:16 +0000)]
usleep(3): always call nanosleep(2)
usleep(3) is a wrapper around nanosleep(2). We should always call
nanosleep(), even if the input is zero. This makes behavior easier
to reason about and ensures we get a nanosleep() ktrace hit if a
program calls usleep().
ok millert@
schwarze [Fri, 2 Jul 2021 16:13:56 +0000 (16:13 +0000)]
Document the read-only (sic!) accessor function X509_NAME_ENTRY_set(3).
While here, stress that X509_NAME objects cannot share X509_NAME_ENTRY
objects, and polish a few misleading wordings.
deraadt [Fri, 2 Jul 2021 15:34:27 +0000 (15:34 +0000)]
Before calling setrlimit(), pull cur up to max.
ok millert
millert [Fri, 2 Jul 2021 15:34:16 +0000 (15:34 +0000)]
Use nanosleep() and setitimer() instead of usleep() and ualarm().
Both usleep() and ualarm() are obsolete and were removed from POSIX.
OK deraadt@
kettenis [Fri, 2 Jul 2021 14:58:33 +0000 (14:58 +0000)]
Remove bogus comments.
kettenis [Fri, 2 Jul 2021 14:50:18 +0000 (14:50 +0000)]
Remove a few pointless comments.
bluhm [Fri, 2 Jul 2021 12:17:41 +0000 (12:17 +0000)]
Writing ktrace files to NFS must no be done while holding the net
lock. accept(2) panics, connect(2) dead locks. Additionally copy
in or out must not hold the net lock as it may be a memory mapped
file on NFS.
Simplify dns_portcheck(), it does not modify namelen anymore.
In doaccept() release the socket lock before calling copyaddrout().
Rearrange the checks in sys_connect() like they are in sys_bind().
OK mpi@
schwarze [Fri, 2 Jul 2021 11:48:01 +0000 (11:48 +0000)]
Add a roff comment saying that X509_certificate_type(3) is intentionally
undocumented. It is archaic and practically unused and unusable.
tb@ and jsing@ agree with marking it as undocumented.
Put the comment here because EVP_PKEY_base_id(3) is a viable alternative.
schwarze [Fri, 2 Jul 2021 11:15:12 +0000 (11:15 +0000)]
call the API function X509_NAME_cmp(3) instead of the obsolete,
undocumented macro alias X509_name_cmp(3);
no binary change;
OK tb@
schwarze [Fri, 2 Jul 2021 11:15:08 +0000 (11:15 +0000)]
call the API function X509_NAME_cmp(3) instead of the obsolete,
undocumented macro alias X509_name_cmp(3);
no change to the assembler code generated by the compiler;
OK tb@
schwarze [Fri, 2 Jul 2021 10:50:39 +0000 (10:50 +0000)]
Add a roff comment saying that X509_name_cmp(3) is intentionally
undocumented because it is almost unused in real-world code.
OK tb@
kettenis [Fri, 2 Jul 2021 10:42:22 +0000 (10:42 +0000)]
Cleanup early bootstrap code. This mostly realigns the code with the
FreeBSD code from which it was derived. In particular, it uses the same
trick to switch page tables as FreeBSD, which is what we use to spin up
the secondary CPUs already. This avoids having to install a temporary
1:1 mapping.
ok mlarkin@
kettenis [Fri, 2 Jul 2021 08:53:28 +0000 (08:53 +0000)]
Only do TLB shootdown on CPUs where a pmap is active. Only make SBI calls
that do a remote SFENCE.VMA or FENCE.I if the set of hart IDs isn't empty.
This significantly speeds things up.
ok jsg@
kettenis [Fri, 2 Jul 2021 08:44:37 +0000 (08:44 +0000)]
Run SBI calls to to get mvendorid/marchid/mimplid on the actual CPU we're
probing and decode mvendorid and marchid.
ok mlarkin@, deraadt@, jsg@
dtucker [Fri, 2 Jul 2021 07:20:44 +0000 (07:20 +0000)]
Remove obsolete comments about SSHv1 auth methods. ok djm@
dtucker [Fri, 2 Jul 2021 05:11:20 +0000 (05:11 +0000)]
Remove references to ChallengeResponseAuthentication in favour of
KbdInteractiveAuthentication. The former is what was in SSHv1, the
latter is what is in SSHv2 (RFC4256) and they were treated as somewhat
but not entirely equivalent. We retain the old name as deprecated alias
so config files continue to work and a reference in the man page for
people looking for it.
Prompted by bz#3303 which pointed out the discrepancy between the two
when used with Match. Man page help & ok jmc@, with & ok djm@
deraadt [Fri, 2 Jul 2021 03:10:19 +0000 (03:10 +0000)]
__remq.S will be needed soon.
deraadt [Thu, 1 Jul 2021 20:34:19 +0000 (20:34 +0000)]
delete sbin/restore
jsing [Thu, 1 Jul 2021 17:53:39 +0000 (17:53 +0000)]
Merge SSL_METHOD_INTERNAL into SSL_METHOD.
Now that SSL_METHOD is opaque and in internal headers, we can remove
SSL_METHOD_INTERNAL by merging it back into SSL_METHOD.
ok tb@
stsp [Thu, 1 Jul 2021 11:51:55 +0000 (11:51 +0000)]
Prevent athn(4) from calling ieee80211_find_rxnode() on bad frames.
This fixes an issue introduced with our workaround for bogus michael
mic failures seen when hardware receives control frames. We do need
to ignore the michael mic failure in this case but we should not call
ieee80211_find_rxnode() on such frames unconditionally. Do this only
if the transmitter's address has already been cached.
When ieee80211_find_rxnode() is called with an unknown source MAC address
it will create a new entry in the node cache. Frames flagged as incorrectly
received by hardware should not be passed to ieee80211_find_rxnode() without
further verification to avoid creating bogus cache entries based on corrupt
frame headers.
Prompted by an issue seen by kettenis@ on arm64 where the node cache
contains bogus entries. This change doesn't fix the issue but it is
a step in the right direction regardless since it fixes one possible
cause for the issue.
ok kettenis@
tested by myself and Mikolaj Kucharski
schwarze [Thu, 1 Jul 2021 10:22:16 +0000 (10:22 +0000)]
Just in case another developer besides anton@ and myself even wants to
look at these tests: add a comment providing a high-level picture of
what is going on here, such that everyday maintenance can be performed
without going down the rabbit hole of subr.sh and edit.c.
schwarze [Thu, 1 Jul 2021 10:00:15 +0000 (10:00 +0000)]
We already tested all corner cases for two-byte and three-byte UTF-8
sequences, but coverage of four-byte sequences was incomplete, which
contributed to the recently fixed ksh(1) emacs.c bug not being found
for some time.
Consequently, add some tests covering
+ valid sequences starting with \0360, \0361, and \0363;
+ incomplete sequences starting with \0360 and \0361;
+ invalid (too low) sequences starting with \0360.
OK anton@
eric [Thu, 1 Jul 2021 07:42:16 +0000 (07:42 +0000)]
remove useless and confusing log message when client has no certificate
ok millert@
deraadt [Thu, 1 Jul 2021 01:16:13 +0000 (01:16 +0000)]
We moved myname parsing from /etc/netstart to /etc/rc nov 2020, but
this manual page wasn't corrected
from James Jerkins
kettenis [Wed, 30 Jun 2021 22:20:56 +0000 (22:20 +0000)]
Simplify the way we track the FPU state, using powerpc64 as a model.
The new code still uses the clean/dirty state that the hardware reports
to optimize saving/restoring the FPU register, but no longer attempts to
keep the FPU registers alive across a context switch. Fixes panics seen
on MP kernels.
ok drahn@
schwarze [Wed, 30 Jun 2021 18:55:52 +0000 (18:55 +0000)]
The byte \0363 is not invalid in UTF-8.
The incorrectness of this test was exposed by the bugfix
in /usr/src/bin/ksh/emacs.c rev. 1.88 and reported to me by bluhm@.
This is a minimal fix replacing the incorrect test line
with two correct tests involving the same byte.
OK anton@
schwarze [Wed, 30 Jun 2021 18:46:49 +0000 (18:46 +0000)]
more trivial .Ar -> .Fa replacements in syscall manuals
schwarze [Wed, 30 Jun 2021 18:17:21 +0000 (18:17 +0000)]
use .Fa rather than .Ar; patch from me at EmilEngler dot com
jsing [Wed, 30 Jun 2021 18:11:47 +0000 (18:11 +0000)]
Disable some code that reaches into libssl internals.
This should be moved to a dedicated regress test.
jsing [Wed, 30 Jun 2021 18:10:42 +0000 (18:10 +0000)]
Disable some tests that probably no longer make sense.
We'll either fix these or remove them in the near future.
jsing [Wed, 30 Jun 2021 18:09:46 +0000 (18:09 +0000)]
Pull in ssl_locl.h to allow for move of struct ssl_session_st.
jsing [Wed, 30 Jun 2021 18:07:50 +0000 (18:07 +0000)]
Prepare to provide SSL_get_signature_nid() and friends.
This adds functionality for SSL_get_signature_nid(),
SSL_get_peer_signature_nid(), SSL_get_signature_type_nid() and
SSL_get_peer_signature_type_nid().
This is not currently publicly visible and will be exposed at a later
date.
ok inoguchi@ tb@
jsing [Wed, 30 Jun 2021 18:04:05 +0000 (18:04 +0000)]
Move some structs from public to private headers.
Move struct ssl_cipher_st, struct ssl_method_st, struct ssl_session_st and
struct ssl3_state_st from public to private headers. These are already
under #ifdef LIBRESSL_INTERNAL and are no longer publicly visible.
ok inoguchi@ tb@
claudio [Wed, 30 Jun 2021 15:24:10 +0000 (15:24 +0000)]
... way too much white space!