jsg [Mon, 19 Jul 2021 10:03:08 +0000 (10:03 +0000)]
drm/sched: Avoid data corruptions
From Andrey Grodzovsky
0687411e2a8858262de2fc4a1d576016fd77292e in linux 5.10.y/5.10.51
0b10ab80695d61422337ede6ff496552d8ace99d in mainline linux
jsg [Mon, 19 Jul 2021 10:01:10 +0000 (10:01 +0000)]
drm/scheduler: Fix hang when sched_entity released
From Andrey Grodzovsky
5ed8c298b2e140c640af8813a490fb4d77165e97 in linux 5.10.y/5.10.51
c61cdbdbffc169dc7f1e6fe94dfffaf574fe672a in mainline linux
jsg [Mon, 19 Jul 2021 09:58:16 +0000 (09:58 +0000)]
drm/amd/display: fix use_max_lb flag for 420 pixel formats
From Dmytro Laktyushkin
3ca86d44b9023cd96c893d6dd90aacbca90e4d74 in linux 5.10.y/5.10.51
8809a7a4afe90ad9ffb42f72154d27e7c47551ae in mainline linux
jsg [Mon, 19 Jul 2021 09:55:11 +0000 (09:55 +0000)]
drm/amd/amdgpu/sriov disable all ip hw status by default
From Jack Zhang
b025bc07c94770ab5ca68a8b2ead12628c2a0698 in linux 5.10.y/5.10.51
95ea3dbc4e9548d35ab6fbf67675cef8c293e2f5 in mainline linux
jsg [Mon, 19 Jul 2021 09:51:46 +0000 (09:51 +0000)]
drm/amd/display: fix HDCP reset sequence on reinitialize
From Brandon Syu
fb7479d64d77a3bb0ee992f9e450cf52d56c6b2c in linux 5.10.y/5.10.51
99c248c41c2199bd34232ce8e729d18c4b343b64 in mainline linux
jsg [Mon, 19 Jul 2021 09:22:32 +0000 (09:22 +0000)]
sync kfd_ioctl.h with linux 5.10.y
dtucker [Mon, 19 Jul 2021 08:48:33 +0000 (08:48 +0000)]
Use SUDO when setting up hostkey.
jsing [Mon, 19 Jul 2021 08:42:24 +0000 (08:42 +0000)]
Mop up dtls1_get_ccs_header() and struct ccs_header_st.
All this code does is read one byte from memory with an unknown length,
potentially being a one byte overread... and then nothing is actually done
with the value.
ok tb@
jsing [Mon, 19 Jul 2021 08:39:28 +0000 (08:39 +0000)]
Inline DTLS1_CCS_HEADER_LENGTH rather than having a single use variable.
ok tb@
dtucker [Mon, 19 Jul 2021 05:08:54 +0000 (05:08 +0000)]
Increase time margin for rekey tests. Should help reliability on very
heavily loaded hosts.
dtucker [Mon, 19 Jul 2021 03:13:28 +0000 (03:13 +0000)]
Ensure that all returned SSHFP records for the specified host name and
hostkey type match instead of only one. While there, simplify the code
somewhat and add some debugging. Based on discussion in bz#3322, ok djm@.
dtucker [Mon, 19 Jul 2021 02:46:34 +0000 (02:46 +0000)]
Add test for host key verification via SSHFP records. This requires
some external setup to operate so is disabled by default (see
comments in sshfp-connect.sh).
dtucker [Mon, 19 Jul 2021 02:29:28 +0000 (02:29 +0000)]
Add
ed25519 key and test SSHFP export of it. Only test RSA SSHFP export
if we have RSA functionality compiled in.
dtucker [Mon, 19 Jul 2021 02:21:50 +0000 (02:21 +0000)]
Put dh_set_moduli_file call inside ifdef WITH_OPENSSL. Fixes build with
OPENSSL=no.
dtucker [Mon, 19 Jul 2021 00:16:26 +0000 (00:16 +0000)]
Group keygen tests together.
dtucker [Sun, 18 Jul 2021 23:10:10 +0000 (23:10 +0000)]
Add test for ssh-keygen printing of SSHFP records.
kn [Sun, 18 Jul 2021 23:02:57 +0000 (23:02 +0000)]
Markup promises with Cm not Va
These are not variable names but rather keywords (as port of one big
string).
I originally changed this such that `.Bl -tag' would automatically tag them
and thus allow ":t id" in the manual pager; since schwarze's recent
mandoc(1) commit this is no longer needed as `.Va' gained tagging with it,
but its use is incorrect here nonetheless, so still change it to the more
appropiate `.Cm'.
Input jmc schwarze
krw [Sun, 18 Jul 2021 21:40:13 +0000 (21:40 +0000)]
Don't save the prt_scyl/prt_ecyl values, change them, use them
and then restore them. Just change/use the saved values and skip
the restoring.
Allows PRT_make() to add 'const' to its struct mbr parameter, and
thus allows MBR_make() to add 'const' to its struct mbr
parameter.
No intentional functional change.
mvs [Sun, 18 Jul 2021 18:19:22 +0000 (18:19 +0000)]
Introduce and use garbage collector for 'ipsec_ids' struct entities
destruction instead of using per-entity timeout. This fixes the races
between ipsp_ids_insert(), ipsp_ids_free() and ipsp_ids_timeout().
ipsp_ids_insert() can't stop ipsp_ids_timeout() timeout handler which is
already running and awaiting netlock to be released, so reused `ids' will
be silently removed in this case.
ipsp_ids_free() can't determine is ipsp_ids_timeout() timeout handler
running because timeout_del(9) called by ipsp_ids_insert() clears it's
triggered state. So ipsp_ids_timeout() could be scheduled to run twice in
this case.
Also hrvoje@ reported about ipsec(4) throughput increased with this diff
so it seems we caught significant count of ipsp_ids_insert() races.
tests and feedback by hrvoje@
ok bluhm@
kn [Sun, 18 Jul 2021 17:44:46 +0000 (17:44 +0000)]
Zap dhclient bits, mention resolvd
OK deraadt
krw [Sun, 18 Jul 2021 15:28:37 +0000 (15:28 +0000)]
Make GPT_read() obtain and validate the on-disk MBR itself, via
MBR_read(), and report success/failure.
Simplifies logic and makes clearer that the protective MBR is a
required part of a GPT.
With the standardization on 0/-1 return values for helper
functions, rename 'valid' to 'error' in GPT_read() to make logic
less mind bending.
No functional change.
deraadt [Sun, 18 Jul 2021 15:18:47 +0000 (15:18 +0000)]
Remove hackery around /tmp/i/resolv.conf*. In particular "lookup file bind"
is required because of TLS servername for contacting ftp.openbsd.org, and
there is no point doing that in resolv.conf.tail because it is no longer used.
ok florian kn
bluhm [Sun, 18 Jul 2021 14:38:20 +0000 (14:38 +0000)]
The IPsec authentication before decryption used a different replay
counter than after decryption. This could result in "esp_input_cb:
authentication failed for packet in SA" errors. As we run crypto
operations async, thousands of packets are stored in the crypto
task. During the queueing the replay counter of the tdb can change.
Then the higher 32 bits may increment although the lower 32 bits
did not wrap.
checkreplaywindow() must be called twice per packet with the same
replay counter. Store the value in struct tdb_crypto while dangling
in the task queue and doing crypto operations.
tested by Hrvoje Popovski; joint work with tobhe@
stsp [Sun, 18 Jul 2021 13:07:13 +0000 (13:07 +0000)]
Do not ask iwx(4) firmware to send probe requests on passive channels.
This seems to fix a problem where firmware becomes unresponsive after
association and eventually raises a fatal error. iwx(4) already has a
workaround in place for this: We always ask firmware to perform a
passive scan, on any channel.
This change is a better fix from iwm(4) CVS commit jJFMeXcos9GOqFQz.
However, the current workaround will stay in place for now, until we
have evidence that active scanning is working on iwx(4) devices.
I will try to organize an out-of-tree round of testing for this.
krw [Sun, 18 Jul 2021 12:41:00 +0000 (12:41 +0000)]
Put MBR_parse() invocation inside MBR_read() instead of always
following an invocation of MBR_read() with a call to MBR_parse().
No functional change.
stsp [Sun, 18 Jul 2021 12:39:16 +0000 (12:39 +0000)]
Clear the "persistence bit" on iwx(4) devices during hardware init.
According to iwlwifi commit messages this fixes an edge case where
devices fail to resume after system suspend.
See Linux commit
8954e1eb2270fa2effffd031b4839253952c76f2
Same fix was made for iwm(4) in CVS commit x0XTNdEmudy5oBR4
florian [Sun, 18 Jul 2021 12:33:41 +0000 (12:33 +0000)]
Ignore routers option when a classless static routes option is present
as mandated by RFC3442.
Pointed out by, initial diff, testing & OK bket@
stsp [Sun, 18 Jul 2021 12:21:49 +0000 (12:21 +0000)]
Set MIMO-related flags in the iwx(4) ADD_STA command.
Not sure how we ended up missing these. MIMO apparently works without the
flags set, and setting them doesn't seem to have any immediately visible
effect on firmware. But let's be consistent with iwm(4) and iwlwifi.
stsp [Sun, 18 Jul 2021 12:03:57 +0000 (12:03 +0000)]
Newer iwx(4) firmware doesn't like the DQA command anymore.
Check whether firmware advertises DQA support before sending the DQA
command during device initialization. Newer firmware will raise a
fatal error otherwise. The Tx queue API has been redesigned yet again.
stsp [Sun, 18 Jul 2021 11:56:11 +0000 (11:56 +0000)]
Fix wrong usage of iwx_lookup_cmd_ver() in iwx_send_soc_conf().
Callers are supposed to check whether iwx_lookup_cmd_ver() returns
CMD_VER_UNKNOWN, and this check was missing here. Fortunately, the
buggy check was part of a condition which also requires the
low_latency_xtal constant to be set. We do not yet support devices
where low_latency_xtal is non-zero, so the bug never triggered.
dv [Sun, 18 Jul 2021 11:55:45 +0000 (11:55 +0000)]
vmd(8): remove invalid errno values from config_setvm
Refactor config_setvm to directly return error code on failure
instead of returning -1 and setting errno. It was setting unsupported
values not defined in <errno.h>.
OK mlarkin@
schwarze [Sun, 18 Jul 2021 11:40:58 +0000 (11:40 +0000)]
Support auto-tagging for ".It Va".
This combination is somewhat rare because few libraries expose so many
global variables that they need a list to enumerate them, but when the
idiom does occur, tagging the variable names is generally useful.
For example, this helps awk(1), dc(1), make(1), rc.subr(8), ...
Missing feature reported and patch reviewed, tested, and OK'ed by kn@.
stsp [Sun, 18 Jul 2021 11:40:31 +0000 (11:40 +0000)]
Fix iwx(4) PHY context updates for newer firmware versions.
Firmware which advertises the BINDING_CDB_SUPPORT capability
needs a remove+add dance when the channel band has changed.
See Linux commit
730a18912bcbde0b94ae7f1b554a9908b3424a22
and Linux commit
91109f42d0ad0c0c282d1fa1257a1548977aa895
The same fix was applied to iwm(4) in CVS commit uQ0WjqRUp03vxHg1
schwarze [Sun, 18 Jul 2021 11:25:47 +0000 (11:25 +0000)]
Let the mandoc.db(5) test suite work with parallel make (make -j),
by making sure that different tests use different directory names
for their work such that they do not collide.
As a side benefit, this allows keeping the complete working directories
of the tests until "make cleandir" is run, which may occasionally
help debugging when something breaks.
The failure with make -j was detected and reported by anton@,
who also reviewed, tested, and OK'ed this somewhat lengthy patch.
kn [Sun, 18 Jul 2021 11:18:38 +0000 (11:18 +0000)]
Document dhcpleased(8) and slaacd(8) besides dhclient(8) as DNS source
Zap the obsolete dhclient.conf(5) supersede quirk while here.
Feedback OK florian
kn [Sun, 18 Jul 2021 11:08:34 +0000 (11:08 +0000)]
Follow dhclient -> dhcpleased switch, reduce supported DHCP statements
dhcpleased(8) does not support every statement dhclient(8) does, so reflect
the status quo simplify things to the smaller set of statements that still
suffice to use autoinstall(8).
While here, speak of dhcp-options(5) and "DHCP lease" instead of
dhcpd.conf statements and the installer's specific DHCP client/parser.
OK florian
jsg [Sun, 18 Jul 2021 10:20:06 +0000 (10:20 +0000)]
make the printf on skipping reset with gen7/gen8 debug only
With the 5.10 drm on vlv/ivb/hsw/bdw inteldrm has been known to reset
the chip when a gpu hang is falsely detected. This appears to be
related to the workarounds for cleaning up after switching hardware
contexts.
intel_gt_reset *NOTICE* [drm] Resetting chip for context closure in glsl-uniform-int<99177>
mark_guilty *NOTICE* [drm] glsl-uniform-int[97885] context reset due to GPU hang
A local change skips the reset (which kills Xorg) on gen7/gen8.
The printf is in that path.
stsp [Sun, 18 Jul 2021 09:37:49 +0000 (09:37 +0000)]
Update the list of firmware files used by iwm(4).
jsg [Sun, 18 Jul 2021 05:02:37 +0000 (05:02 +0000)]
regen
jsg [Sun, 18 Jul 2021 05:02:08 +0000 (05:02 +0000)]
add devices from ASUS ROG Strix G513QY and Alienware Aurora R10
krw [Sat, 17 Jul 2021 21:47:56 +0000 (21:47 +0000)]
The first think MBR_init() does is memset(0) the gpt header and gpt
partition array. So no need to memset(0) the gpt header before each
MBR_init() invocation.
No functional change.
kn [Sat, 17 Jul 2021 20:46:02 +0000 (20:46 +0000)]
Remove setpassent(3) leftovers
Those should have gone with the switch to uid_from_user(3) in r1.95;
OK millert
deraadt [Sat, 17 Jul 2021 17:14:29 +0000 (17:14 +0000)]
Don't document systems which rewrite/change resolv.conf. Those systems can
document their actions in their own manual pages.
ok kn
kn [Sat, 17 Jul 2021 15:28:31 +0000 (15:28 +0000)]
Follow up on dhclient(8) to dhcpleased(8)/resolvd(8) switch
hostname.if(5)'s "dhcp" no longer triggers dhclient in any way, so suggest
the simple "inet autoconf" instead while mentioning "dhcp" as an alias for
that.
Add markup to changed sentences while here.
There is more room for improvemen in this manual.
Feedback deraadt, "go ahead"
schwarze [Sat, 17 Jul 2021 14:23:02 +0000 (14:23 +0000)]
Make the mandoc.db(5) tests work with BSDOBJDIR set,
by stripping initial parts of paths the do not matter
for the features being tested from man(1) output.
Problem found and reported by, solution developed together with, and
OK anton@
krw [Sat, 17 Jul 2021 14:16:34 +0000 (14:16 +0000)]
Nuke a bunch of Capt. Obvious or misleading comments.
No functional change.
krw [Sat, 17 Jul 2021 13:37:01 +0000 (13:37 +0000)]
Change edit/gedit/setpid/gsetpid to return 0/-1 to indicate success or failure
as all other helpers do.
Makes it clear that X* functions are in charge of deciding which CMD_* value
they return to main edting loop. Reduces unnecessary copying/restoring of
original partition information.
No intentional functional change.
kn [Sat, 17 Jul 2021 03:31:31 +0000 (03:31 +0000)]
Ignore unwind socket with kevent(2) under SMALL
All other unwind related bits are under #ifndef SMALL already,
so complete that.
OK florian
djm [Sat, 17 Jul 2021 00:38:11 +0000 (00:38 +0000)]
wrap some long lines
djm [Sat, 17 Jul 2021 00:36:53 +0000 (00:36 +0000)]
fix sftp on ControlPersist connections, broken by recent SessionType
change; spotted by sthen@
jsg [Sat, 17 Jul 2021 00:02:58 +0000 (00:02 +0000)]
sync
jsg [Fri, 16 Jul 2021 23:51:38 +0000 (23:51 +0000)]
drm/amdgpu: add another Renoir DID
From Jinzhou Su
775da83005cb61d4c213c636df9337da05714ff1 in mainline linux
jsg [Fri, 16 Jul 2021 23:47:58 +0000 (23:47 +0000)]
amdgpu: Add Missing Sienna Cichlid DID
From Ori Messinger
d26bbbcc160f6d9feabed73dca62b9e8b86671b4 in mainline linux
krw [Fri, 16 Jul 2021 22:50:43 +0000 (22:50 +0000)]
Change PRT_protected_guid() to take native order uuid's instead of little endian
uuid's. This brings it into line with all the other PRT_ functions taking uuid
parameters.
While adapting gsetid(), introduce separate partition type and partition guid
variables to make the logic clearer.
No intentional functional change.
dv [Fri, 16 Jul 2021 16:21:22 +0000 (16:21 +0000)]
vmd(8): simplify vcpu logic, removing uart & vionet reads
Remove legacy state handling on the ns8250 and virtio network devices
originally put in place before using libevent for async device
events. The vcpu thread doesn't need to process device data as it is
handled by the libevent thread.
This has the benefit of simplifying some of the message passing
between threads introduced to the ns8250 uart since both the vcpu
and libevent threads were processing read events.
No functional change intended. Tested by many, including abieber@,
weerd@, Mischa Peters, and Matthias Schmidt. (Thanks.)
OK mlarkin@
deraadt [Fri, 16 Jul 2021 15:26:24 +0000 (15:26 +0000)]
this media can also use dhcpleased now
florian [Fri, 16 Jul 2021 15:25:31 +0000 (15:25 +0000)]
Switch from dhclient to dhcpleased in the installer.
More cleanup is needed from a real ksh hacker, but looks already
pretty good to kn.
OK deraadt
florian [Fri, 16 Jul 2021 15:24:44 +0000 (15:24 +0000)]
Create working directories earlier so that resolvd can write
resolv.conf.shadow and start dhcpleased/resolvd.
OK deraadt
florian [Fri, 16 Jul 2021 15:24:08 +0000 (15:24 +0000)]
Replace dhclient with dhcpleased/resolvd on the ramdisk.
OK deraadt who also helped making some space.
florian [Fri, 16 Jul 2021 15:22:43 +0000 (15:22 +0000)]
create lease directory for dhcpleased in the ramdisk
OK deraadt
florian [Fri, 16 Jul 2021 15:22:12 +0000 (15:22 +0000)]
The installer plays weird tricks with /etc/resolv.conf and symlinks it
from /tmp/i. Work around this in resolvd so that it can correctly
update the file until this is fixed in install.sub.
probably grudgingly OK deraadt
florian [Fri, 16 Jul 2021 15:21:41 +0000 (15:21 +0000)]
switch to dhcpleased/resolvd in base
OK deraadt
bluhm [Fri, 16 Jul 2021 15:08:39 +0000 (15:08 +0000)]
Improve comments in IPsec replay window calculation.
OK tobhe@
sthen [Fri, 16 Jul 2021 15:05:58 +0000 (15:05 +0000)]
Update the default block size in the manual, in sndiod.c r1.37 it was
reduced to 480 frames. ok ratchov@
mestre [Fri, 16 Jul 2021 14:36:09 +0000 (14:36 +0000)]
main proc doesn't speak with sockets during the main loop and setpriority plus
chroot (and privdrop) in the child procs at this point are all set and done so
pledge("inet id") promises are not required anymore.
analysis and OK @kn
mestre [Fri, 16 Jul 2021 14:34:37 +0000 (14:34 +0000)]
/etc/ssl/cert.pem is loaded into mem with tls_load_file(3) on local function
priv_constraint_child() which is called before we reach pledge(2), therefore we
we don't need to unveil(2) that file nor having pledge("rpath") since it was
only required to read that same file.
OK kn@
kn [Fri, 16 Jul 2021 13:59:10 +0000 (13:59 +0000)]
Remove unneeded call to tls_init(3)
As per the manual and lib/libtls/tls.c revision 1.79 from 2018
"Automatically handle library initialisation for libtls." initialisation
is handled automatically by other tls_*(3) functions.
Remove the tls_init() call to not give the impression of it being needed.
Feedback tb
OK Tests mestre
krw [Fri, 16 Jul 2021 13:29:49 +0000 (13:29 +0000)]
Compare pointer to NULL, not 0.
krw [Fri, 16 Jul 2021 13:26:04 +0000 (13:26 +0000)]
Be consistent and always return -1 to indicate failure.
DISK_printgeometry() return value was always 0 and never
checked so just make it void.
No functional change.
jsg [Fri, 16 Jul 2021 13:09:18 +0000 (13:09 +0000)]
regen
jsg [Fri, 16 Jul 2021 13:08:24 +0000 (13:08 +0000)]
add some x570 ppbs and another navi 23
jsg [Fri, 16 Jul 2021 09:14:17 +0000 (09:14 +0000)]
recognise more pci capabilities
from PCI Code and ID Assignment Specification
djm [Fri, 16 Jul 2021 09:00:23 +0000 (09:00 +0000)]
Explicitly check for and start time-based rekeying in the client
and server mainloops.
Previously the rekey timeout could expire but rekeying would not start
until a packet was sent or received. This could cause us to spin in
select() on the rekey timeout if the connection was quiet.
ok markus@
claudio [Fri, 16 Jul 2021 07:59:38 +0000 (07:59 +0000)]
Remove the unveil current directory pointer from struct process. Instead
pass in the vnode to unveil_start_relative() like it is done for *at()
syscalls. This fixes an issue with fchdir() that actually did not correctly
reset this pointer when changing the working directory.
OK beck@
krw [Thu, 15 Jul 2021 23:25:46 +0000 (23:25 +0000)]
Nuke unused variable.
krw [Thu, 15 Jul 2021 21:58:02 +0000 (21:58 +0000)]
No need for separate b_sectors, b_offset and b_type variables. We have struct
prt to hold this info in one place.
Construct a struct prt to hold the boot partition information and stash it in
the struct disk for both MBR and GPT to access as needed.
Move the blocks to sectors conversions into DISK_open() with all of its
geometry friends.
No intentional functional change.
krw [Thu, 15 Jul 2021 21:23:54 +0000 (21:23 +0000)]
Consolidate disk geometry fiddling, turning -l/-c/-h/-s values into MBR
partition geometry info, into DISK_open(). Resulting in clearer logic.
No intentional functional change.
deraadt [Thu, 15 Jul 2021 15:37:55 +0000 (15:37 +0000)]
more space savings
deraadt [Thu, 15 Jul 2021 15:33:47 +0000 (15:33 +0000)]
space savings on install kernels. (restore and ralink drivers+firmware go
away)
inoguchi [Thu, 15 Jul 2021 12:41:49 +0000 (12:41 +0000)]
Wrap over 80 long lines in ca.c
inoguchi [Thu, 15 Jul 2021 11:43:27 +0000 (11:43 +0000)]
Explicitly check pointer variable if it is NULL or not in ca.c
job [Thu, 15 Jul 2021 11:07:33 +0000 (11:07 +0000)]
Fix regress for rpki-client
spotted by bluhm
inoguchi [Thu, 15 Jul 2021 10:26:43 +0000 (10:26 +0000)]
Remove space between '*' and pointer variable in ca.c
inoguchi [Thu, 15 Jul 2021 10:15:22 +0000 (10:15 +0000)]
Use 'serial' rather than 'ser' in ca.c
input from jsing@
inoguchi [Thu, 15 Jul 2021 09:56:32 +0000 (09:56 +0000)]
Convert openssl(1) ca option handling
New option handling for openssl(1) ca.
This diff is just replacing with new option handling, no functional change.
I'm using the word DN or RDN in description as manual uses them, rather than
replacing with "Distinguished Name" or "Relative Distinguished Name".
I would like to add another fixes below by follow-up diffs.
- remove space between '*' and pointer variable
- wrap 80+ long lines
- explicitly check pointer variable if it is NULL or not
comments and ok from jsing@
claudio [Thu, 15 Jul 2021 06:57:02 +0000 (06:57 +0000)]
UNVEIL_INSPECT is no longer needed, adjust code accordingly.
OK semarie@
jsg [Thu, 15 Jul 2021 01:24:07 +0000 (01:24 +0000)]
drm/amd/dc: Fix a missing check bug in dm_dp_mst_detect()
From Yingjie Wang
5885fce7b43919406b32d1c942c9b496aac3abbc in linux 5.10.y/5.10.50
655c0ed19772d92c9665ed08bdc5202acc096dda in mainline linux
jsg [Thu, 15 Jul 2021 01:20:28 +0000 (01:20 +0000)]
drm/bridge: Fix the stop condition of drm_bridge_chain_pre_enable()
From Douglas Anderson
0cd39c96574d19e234cee971868c82c3769c2ee6 in linux 5.10.y/5.10.50
bab5cca7e609952b069a550e39fe4893149fb658 in mainline linux
tobhe [Wed, 14 Jul 2021 22:39:26 +0000 (22:39 +0000)]
Export SA replay counters via pfkey and print with ipsecctl.
This is useful for debugging replay window issues with 64 bit
sequence numbers in IPsec.
ok bluhm@
bluhm [Wed, 14 Jul 2021 22:09:24 +0000 (22:09 +0000)]
After VFS shutdown, init(8) cannot map a missing page that contains
the signal handler code. Traditionally a process would spin in
such a case, but we changed the logic in revision 1.167 trapsignal()
to receive a fatal signal. If that happens to init(8), the kernel
panics. In case of reboot, jump between init signal handler and
page fault trap until the kernel resets the machine.
reported and tested weerd@; OK deraadt@
bluhm [Wed, 14 Jul 2021 21:07:36 +0000 (21:07 +0000)]
Resend the TCP packet only if the MTU locked flag appears at the
route and was not there before. This should prevent a recursion
in path MTU discovery with TCP over IPsec.
reported and tested Matthias Schmidt; tested and OK tobhe@
bluhm [Wed, 14 Jul 2021 18:23:11 +0000 (18:23 +0000)]
Shrink cluster size for FAT32 to fullfil the new minimum cluster
limit of newfs_msdos for our small 64 MB vnd disk image.
suggested by jsg@
bluhm [Wed, 14 Jul 2021 17:58:51 +0000 (17:58 +0000)]
LibreSSL error message has changed, adapt test.
kn [Wed, 14 Jul 2021 13:33:57 +0000 (13:33 +0000)]
Remove unneeded calls to tls_init(3)
As per the manual and lib/libtls/tls.c revision 1.79 from 2018
"Automatically handle library initialisation for libtls." initialisation
is handled automatically by other tls_*(3) functions.
Remove explicit tls_init() calls from base to not give the impression of
it being needed.
Feedback tb
OK Tests mestre
claudio [Wed, 14 Jul 2021 11:14:27 +0000 (11:14 +0000)]
Restructure the getopt_long options. Sort them mostly by name with the
exception of no-xyz options that are grouped with the coresponding xyz
option. Fix --no-motd to use the internal flag setting of getopt_long.
Also use some defines instead of numbers for pure long options that
need special handling.
OK benno@
jsg [Wed, 14 Jul 2021 09:56:17 +0000 (09:56 +0000)]
timeout_add() always sets a new expiry time unrelated to the previous
one and requeues as needed so no need to call timeout_del() before
timeout_add()
spotted by kettenis@
nicm [Wed, 14 Jul 2021 08:56:00 +0000 (08:56 +0000)]
Move default value for TERM into tmux.h.
jmc [Wed, 14 Jul 2021 06:46:38 +0000 (06:46 +0000)]
reorder SessionType; ok djm
jsg [Wed, 14 Jul 2021 05:42:47 +0000 (05:42 +0000)]
When mod_timer() is called with a value less than or equal to the
current number of jiffies delete the timeout and reschedule for the next
tick. Avoids timeout_add() failing the ticks >= 0 assertion when
mod_timer() is called from i915_utils.c set_timer_ms().
Reported and earlier version tested by Tom Murphy on Kaby Lake R.
anton [Wed, 14 Jul 2021 05:04:08 +0000 (05:04 +0000)]
Adjust expected certificate error message. Caused by recent work to support
partial chains in libcrypto.
ok bluhm@