summaryrefslogtreecommitdiff
path: root/sys
AgeCommit message (Collapse)Author
2017-03-09Only set cpuresetfn if it isn't set already such that other reboot mechanismsMark Kettenis
(like psci or syscon) are preferred over letting the watchdog fire. ok patrick@
2017-03-09Add syscon(4), a driver that provides reboot/poweroff functionality throughMark Kettenis
the generic "regmap" interface. ok patrick@
2017-03-09Add a "regmap" interface that allows devices to provide access to theirMark Kettenis
registers to devices that live elsewhere in the device tree. ok patrick@
2017-03-09Glue to attach xhci(4) using the device tree.Mark Kettenis
2017-03-09Remove unecessary splsoftnet()/splx() dances.Martin Pieuchot
ok bluhm@, claudio@
2017-03-09Add interrupt routing API like we have on arm64.Mark Kettenis
2017-03-09Change the interrupt routing API to take a pointer to a "struct cpu_info"Mark Kettenis
instead of a cpu number. ok patrick@
2017-03-09Prevent integer overflow in PF when calculating the adaptive timeout.Claudio Jeker
Mainly states of established TCP connections whould be affected resulting in immediate state removal once the numer of states is bigger than adaptive.start. Disabling adative timeouts is a workaround to avoid this bug. Issue found and initial diff by Mathieu Blanc (mathieu.blanc at cea dot fr) OK mikeb@
2017-03-08Fix logic in the driver preventing it to receive broadcast framesMartin Pieuchot
before configuring an address. Make dhclient(8) work on sq(4). Problem reported by Frank Scheiner, diff from miod@
2017-03-08Do not clear IFF_UP, even in the error path, clearing IFF_RUNNINGMartin Pieuchot
is enough. This flag should only be set by the stack, drivers shouldn't mess with it. Discussed with dlg@ and mikeb@, ok mikeb@, stsp@
2017-03-08Enable msk(4).Patrick Wildt
2017-03-08Enable eephy(4).Patrick Wildt
2017-03-08Enable re(4).Patrick Wildt
2017-03-08Do not clear IFF_UP, even in the error path.Martin Pieuchot
This flag should only be set by the stack, drivers shouldn't mess with it. Discussed with dlg@ and mikeb@, ok gerhard@
2017-03-08No need to clear IFF_UP after calling if_down().Martin Pieuchot
2017-03-08Always invalidate ND entries in nd6_free().Martin Pieuchot
This way RTF_CACHED entries are properly flushed. Fix a regression reported by weerd@ and also tested by matthieu@. ok bluhm@
2017-03-08Enable ix(4). Requested by deraadt@Jonathan Gray
2017-03-08Do not unset IFF_UP twice.Martin Pieuchot
ok stsp@, claudio@, mikeb@, dlg@
2017-03-08Enable em(4), ahci(4), nvme(4) and USB controllers attached to pci(4).Patrick Wildt
2017-03-08Establish API to route interrupts to specific CPU cores.Patrick Wildt
Based on and ok drahn@
2017-03-08Sync struct etheripstat decl with netinet/ip_ether.hJeremie Courreges-Anglas
2017-03-07Rename struct etheripstat members for consistency, move them all to 64bitsJeremie Courreges-Anglas
Will make transition to percpu counters easier. ok bluhm@
2017-03-07Drop unused variable from ntp_update_second().David Hill
ok jca@ deraadt@
2017-03-07Make this actually work as a tty. Everything except for the most basicMark Kettenis
tx and rx functionality is still missing.
2017-03-07When the inpcb queue and hash lists are traversed or modified weAlexander Bluhm
need netlock. Remove the obsolete splnet. OK mpi@
2017-03-07Don't overwrite the flow ID once it's setMike Belopuhov
Output processing may split, encapsulate or obfuscate a single stream which makes the changed flow ID less useful for purposes of flow control, for instance fair sharing of bandwidth. OK dlg
2017-03-07Record the target mask for the boot CPU when we attach and use that to pickMark Kettenis
the target CPU interface when establishing an interrupt. Makes interrupts work on machines that boot up on a CPU that is attached to a CPU interface that isn't zero. Discussed with Dale Rahn.
2017-03-07Change priq enqueue policy to drop lower priority packetsMike Belopuhov
The new priority queueing enqueue policy is such that when the aggregate queue depth of an outgoing queue is exceeded we attempt to find a non-empty queue of packets with lower priority than the priority of a packet we're trying to enqueue and if there's such queue, we drop the first packet from it. This ensures that high priority traffic will almost always find the place on the queue and low priority bulk traffic gets a better chance at regulating its throughput. There's no change in the behavior if altered priorities are not used (e.g. via "set prio" Pf directive, VLAN priorities and so on). With a correction from dlg@, additional tests by dhill@ OK bluhm, mpi
2017-03-07Unbreak the tree by removing the "exynos.h" include and associated prototype.Mark Kettenis
Found the hard way by florian@
2017-03-07Convert priority queue lists to mbuf_listsMike Belopuhov
This simplifies the code quite a bit making it easier to reason about. dlg@ has begrudgingly submitted to populism, OK bluhm, mpi
2017-03-07Keep on trying to grab the lock after leaving ddb after lock spin-out.Visa Hankala
This restores the behaviour that preceded ticket locks. The feature can be useful in some debug cases where the system is not totally borken. OK guenther@, dlg@, mpi@
2017-03-07Use the pause instruction on the slow path. This improvesVisa Hankala
performance a bit. OK mikeb@, kettenis@, mpi@, tom@, mlarkin@
2017-03-07Make the slow path similar to i386's by checking mutex ownerVisa Hankala
on every iteration. OK mikeb@, kettenis@, mpi@, tom@, mlarkin@
2017-03-07Initially in_pcballoc() hooked all new inpcb, including the IPv6Alexander Bluhm
ones, into the IPv4 hash. They cannot be used before bind(2) anyway and then they are rehashed and rehooked, so this was not noticed. Nevertheless put IPv6 PCBs into the IPv6 hash from the beginning. OK jca@ mpi@
2017-03-07Disallow setting machdep.lidaction to any other value but [0,2].Martin Natano
suggested by halex ok deraadt millert
2017-03-07Enforce that tcbtable and udbtable must be accessed with the NET_LOCK().Martin Pieuchot
Get rid of the old splnet()/splx() dances. What's protecting them right now is the KERNEL_LOCK(). but since pf(4) look at these tables we want to protect them in another way, hence the NET_LOCK(), at least as hint. ok bluhm@
2017-03-07Do not grab the NET_LOCK() for routing sockets operations.Martin Pieuchot
The only function that need the lock is rtm_output() as it messes with the routing table. So grab the lock there since it is safe to sleep in a process context. ok bluhm@
2017-03-07Since route_input is no longer part of the protosw struct it is possible toClaudio Jeker
pass the socket to it and do the SO_USELOOPBACK check there. This removes a the nasty hack in the output function where the sp_family is set to 0 temporarily. OK bluhm@ mpi@
2017-03-07enable vioscsi(4)Jonathan Gray
2017-03-07Setup backlight pwm alternate increment on backlight enableJoshua Stein
On at least the MacBookAir7,1 (Broadwell), upon S3 resume, the backlight value was treated as 0 or 100 despite reporting intermediate values, so if the backlight value was anything other than 100 at suspend time, the screen would stay off upon resume. This is backported from Linux commits 32b421e79e6b546da1d469f1229403ac9142d695 and e29aff05f239f8dd24e9ee7816fd96726e20105a which were noted in freedesktop.org bug 67454. ok jsg
2017-03-07move a TODO to the topJoshua Stein
2017-03-07deprecate ifq_enqueue_try, and let backends drop arbitrary mbufs.David Gwynne
mikeb@ wants priq to be able to drop lower priority packets if the current one is high. because ifq avoids freeing an mbuf while an ifq mutex is held, he needs a way for a backend to return an arbitrary mbuf to drop rather than signal that the current one needs to be dropped. this lets the backends return the mbuf to be dropped, which may or may not be the current one. to support this ifq_enqueue_try has to be dropped because it can only signal about the current mbuf. nothing uses it (except ifq_enqueue), so we can get rid of it. it wasnt even documented. this diff includes some tweaks by mikeb@ around the statistics gathered in ifq_enqueue when an mbuf is dropped.
2017-03-06Do not print uninitialized variable in DPRINTFN().Martin Pieuchot
From Christian Ludwig.
2017-03-06domaininit() doesn't need splnet().Martin Pieuchot
At this stage the scheduler isn't setup, which means the 'softnet' isn't running yet, so input packets aren't processed. Prodded by a question from guenther@, ok bluhm@
2017-03-06Move the guts of route_output() meesing with the routing table in theirMartin Pieuchot
own function. ok bluhm@
2017-03-06Guard headers to make sure userland do not look at them.Martin Pieuchot
ok bru@
2017-03-06Use 'inline' rather than _any of_ __the __other__ variants__.Martin Pieuchot
ok bru@
2017-03-06Use nitems() when passing all known parameters to wsmouse_configure().Martin Pieuchot
ok bru@
2017-03-06It is unsafe to dereference ``xfer'' after calling the callback as itMartin Pieuchot
might free it. Prevent a use-after-free in various aynchronous cases. Found while looking at another user-after-free pointed out by ehrhardt@.
2017-03-06Kill global list of IPv6 addresses.Martin Pieuchot
ok bluhm@
generated by cgit v1.2.3 (git 2.25.1) at 2025-01-24 19:53:07 +0000