| Age | Commit message (Collapse) | Author |
|---|
|
Suggested by schwarze
|
|
|
|
issue reported by debian at helgefjell de
ok dtucker djm
|
|
|
|
* Introduce split transaction order queues.
* Improve the NAK interrupt handler routine.
* Mostly move from list_move() to list_move_tail().
Those changes fix an attachment problem seen for certain devices which
are issuing NAK interrupts during split transactions, which don't get
handled correctly by the driver today. This could result in unexpected
channel halting, printing "ChHltd set, but reason is unknown", which
finally leaves the device back on a disabled USB port.
ok kettenis@
|
|
concern raised by kn.
ok florian
|
|
|
|
ok kmos@ sthen@ kn@
|
|
With bluhm@'s diff for parallel forwarding pipex(4) could be accessed in
parallel through (*ifp->if_input)() -> ether_input() ->
pipex_pppoe_input(). PPPOE pipex(4) sessions are mostly immutable except
MPPE crypt.
The new per-session `pxs_mtx' mutex(9) used to protect session's
`ccp-id' which is incremented each time we send CCP reset-request.
The new `pxm_mtx' mutex(9) used to protect MPPE context. Each pipex(4)
session has two of them: one for the input and one for output path.
Where is no lock order limitations because those new mutex(9)'es never
held together.
ok bluhm@
|
|
configuring the same IP.
I forgot to fix slaacd after fixing the same problem in dhcpleased.
Pointed out by afresh1
|
|
that splits the normal RIB linkage vs the adjrib-out linkage. This is
done to make a bit of space to put an extra add-path related id into
the struct without blowing its size over 128 bytes.
Long run this struct should be split up but the necessary changes are
too large right now so this is the 2nd best option.
OK benno@
|
|
add-path information will be dumped properly.
There is one bit where this implementation is not in line with the RFC.
The encoding of DUMP_V2_RIB_GENERIC_ADDPATH for non IPv4 or IPv6 prefixes
is implemented the way gobgp did it (which seems to be the only other BGP
implementation that supports generic encoding). The RFC puts the path_id
in the NLRI for generic encoding instead of adding it to RIB entry struct
like in the IPv4 and IPv6 encoding. This is an open discussion point with
the RFC author.
OK benno@
|
|
to dump add-path enabled systems because the NLRI format changes based
on the add-path capability and there is no way to know which format is
in use so new message types had to be added.
Also extend the ctl_show_rib structure to include the path_id.
OK benno@
|
|
tested on armv7 arm64 and amd64 (bootx64)
ok kettenis@ mpi@
|
|
This allows to send more then 255 bytes of optional parameters. With this
it is possible to send more capabilities. bgpd will switch automatically
to the new extended lenght format if needed but by default the old 1 byte
param length encoding is used to keep compatible with old systems.
OK benno@
|
|
This matches what Linux and FreeBSD do.
ok jmatthew@
|
|
'tdb_data' struct became unused and was removed.
ok bluhm@
|
|
|
|
task queues were unlimited and could overflow during havy traffic.
Even if we still use hardware drivers that sleep, softnet task
instead of soft interrupt can handle this now. Without queues net
lock is inherited and kernel lock is only needed once per packet.
This results in less lock contention and faster IPsec.
Also protect tdb drop counters with net lock and avoid a leak in
crypto dispatch error handling.
intense testing Hrvoje Popovski; OK mpi@
|
|
strict. ICMP error packets generated by pf were not passed
immediately, but could be blocked. Preserve PF_TAG_GENERATED flag
in icmp_reflect() and icmp6_reflect().
reported by sf@; OK patrick@ kn@
|
|
from alf schlichting
|
|
This avoids potential malloc(-1) and malloc(0), spotted by schwarze
while documenting X509_ocspid_print().
ok schwarze
|
|
|
|
Stop decrementing ring->queued inside the if-statement which guards
maintenance of the OACTIVE flag. This is wrong and resulted in a negative
counter value (visible in firmware error traces). The counter is already
decremented in the loop above where frames are taken off the ring.
|
|
documenting the X509_POLICY_TREE object and its sub-objects
|
|
Consolidate GPT/MBR read()/write() operations into DISK_writesectors() and
DISK_readsectors(), producing clearer logic and consistent handling of i/o and
errors.
Add some DPRINTF() and #ifdef DEBUG sections to allow more detailed error
reporting when desired.
|
|
|
|
Running `make -j4' in /usr/src/distrib/amd64/ramdisk_cd/ et al. executes
make(1) to both build the RAMDISK kernel and build libraries.
Doing so does not propagate the flags specified to the ramdisk_cd
invocation, which in turn means `-j4' for example is ignored and both kernel
and libraries will not be built in parallel.
Pass make(1)'s MFLAGS along to retain relevant flags; make is clever enough
to separate flags, variable assignments and targets from each other and only
pass along things to `MFLAGS' that'd make sense, i.e. `make -C. -j4 foo=bar'
does *not* pass `-C.' to change directories.
(can be easily tested with `make -p ... | grep MFLAGS'.)
This makes hacking on ramdisks/the installer much faster, espescially since
the `bsd' target does `make clean' and therefore builds a new kernel every
time.
OK deraadt
|
|
from niklas@ via mikeb@
|
|
client identifier (option 61). Some dhcp servers expect these options
and refuse to hand out a lease without them.
Need for vendor class identifier pointed out & tested by bket
Need for client identifier pointed out by sthen
Input & reads OK sthen (as part of a larger diff)
OK kn (as part of a larger diff)
|
|
configuring the same IP.
Found the hard way by afresh1
|
|
Drop all references to qt4 and replace by qt5/6.
Tweaks and OK espie@
|
|
struct i2c_adapter
inteldrm gmbus returns -EAGAIN to fallback to gpio bitbanging
|
|
ok mikeb
|
|
per-socket locking.
No functional change.
|
|
The code for dtls1_dispatch_alert() and ssl3_dispatch_alert() is largely
identical - with a bit of reshuffling we can use ssl3_dispatch_alert() for
both protocols and remove the ssl_dispatch_alert function pointer.
ok inoguchi@ tb@
|
|
ok kettenis@
|
|
"%s". Same for the output of relay_expand_http().
with and ok claudio@
Found by Cedric Tessier, thanks!
|
|
A sleep() system call first appeared in Research UNIX v2, not v3.
See, e.g., the TOC here:
https://www.tuhs.org/Archive/Distributions/Research/Dennis_v2/v2man.pdf
Pointed out by jsg@ in a related discussion on tech@:
https://marc.info/?l=openbsd-tech&m=162718667209936&w=2
|
|
ok mvs@
|
|
and X509_STORE_CTX_purpose_inherit(3). These functions look deceptively
simple on first sight, but their semantics is surprisingly complicated.
|
|
defaults before validating the times to prevent excessive logging.
Found the hard way & OK brynet
|
|
with OPENSSL=no.
|
|
with OPENSSL_BIN. This will allow us to use the OPENSSL variable from
mk.conf or the make(1) command line indicating if we're building with
our without OpenSSL, and ultimately get the regress tests working in
the OPENSSL=no configuration.
|
|
|
|
switched networks. We validated it, we can't do better than that!
While here reorder the long list of conditions to make it easier to
understand when we doubt a response because we might be behind a
captive portal. First list all conditions when we do not doubt the
response and then the two conditions when we do doubt the response.
OK benno
|
|
the configuration struct. This is also an implicit list of enabled
resolver strategies. We have also stored an explict lookup array of
enabled strategies outside of the configuration to be able to quickly
answer "is this strategy enabled" without traversing the preferences
list.
Move this table into the configuration so that we don't need to
"repair" it on config reload.
This fixes a bug where on startup the preferences list and enabled
lookup table were not in sync. It didn't matter in practice since we
do a config reload and then pass in DNSSEC trustanchors on startup.
Both actions combined repaired things.
OK benno
|
|
If no graphic board is found, LUNA's ROM monitor forces to use serial
console even though DIP switch setting is 'use graphic console'.
For the consistency, change the the kernel behavior the same as ROM
monitor does.
Tested on my LUNA-88K2.
|
|
Raspberry Pi 3 Model B+.
ok kevlo@
|
|
from Ashton Fagg who tested on 9260
|
