| Age | Commit message (Collapse) | Author |
|---|
|
|
|
this makes #include "sec.h" and #if NSEC > 0 work in kernel code.
|
|
ok kn@ bluhm@
|
|
The OnChip System Fabric is mostly an implementation detail on a
bunch of Intel SoC platforms we usually don't care about, except
on some of these machines there are devices shared between the
kernel and the platform that need to coordinate. The iosf(4) driver
allows the kernel to talk via a mailbox interface to take and
release a semaphore that the hardware also uses to lock around
accesses to some components.
There are two ways to talk to the mailbox interface, one via mmio
ops on an INT33BD acpi device. The other is via magic pci conf space
operations on the pci host device. This provides a generic iosf
driver with attachment glue for both acpi and pci so either can be
used. According to linux, the pci ops are a lot more reliable and
less buggy, so if both acpi and pci are available we will prefer
the pci ops after they attach.
I found this because I got a Dell Wyse 3040, which is a cherry
trail/braswell system that has an acpitz(4) which talks to a tipmic(4)
device attached to dwiic(4) on acpi. This box almost always locked
up by the time it got to showing the console login prompt, and it
turns out one of the reasons for this is because acpitz was touching
the tipmic device at the same time the powerunit on the platform
was also trying to use it. The tipmic device lists the iosf mailbox
as a dependency, and the dwiic controller has a _SEM property, both
of which indicate we should use an iosf device to coordinate iic
ops when talking to the tipmic.
This adds the code, but doesn't wire it into dwiic or the build yet.
help from patrick@ jsg@ kettenis@
ok kettenis@ patrick@
|
|
|
|
|
|
|
|
ok deraadt@
|
|
|
|
|
|
Various drivers use ticks/jiffies before initclocks(). It isn't
generally safe to reinitialize them at runtime. Hoist the conditional
definition of HZ from param.c into sys/kernel.h so we can see it from
kern_clock.c and statically initialize ticks/jiffies to the desired
offset.
With this change, timeouts scheduled before initclocks() do not all
fire immediately during the first softclock().
With input from kettenis@.
Link: https://marc.info/?l=openbsd-tech&m=167753870803378&w=2
|
|
used on the NXP i.MX8MP, the Rockchip RK35xx series and Intel Elkhart Lake.
For now it is committed with FDT-attachment only and still in rough shape,
but it's time for development to occur in-tree.
Discussed with bluhm@
ok kettenis@
|
|
ok kettenis@
|
|
the hash will be used to partition work in pf and pfsync in the
future, and right now it is used as the first comparison in the rb
tree state lookup.
using stoeplitz means that pf will hash traffic the same way that
hardware using a stoeplitz key will hash incoming traffic on rings.
stoeplitz is also used by the tcp stack to generate a flow id, which
is used to pick which transmit ring is used on nics with multiple
queues too. using the same algorithm throughout the stack encourages
affinity of packets to rings and softnet threads the whole way
through.
using the hash as the first comparison in the state rb tree comparison
should encourage faster traversal of the state tree by having all
the address/port bits summarised into the single hash value. however,
tests by hrvoje popovski don't show performance changing. on the
plus side, if this change is free from a performance point of view
then it makes the future steps more straightforward.
discussed at length at h2k22
tested by sashan@ and hrvoje popovski
ok tb@ sashan@ claudio@ jmatthew@
|
|
ok kettenis@
|
|
ok kettenis@
|
|
clockintr(9) is a machine-independent clock interrupt scheduler. It
emulates most of what the machine-dependent clock interrupt code is
doing on every platform. Every CPU has a work schedule based on the
system uptime clock. For now, every CPU has a hardclock(9) and a
statclock(). If schedhz is set, every CPU has a schedclock(), too.
This commit only contains the MI pieces. All code is conditionally
compiled with __HAVE_CLOCKINTR. This commit changes no behavior yet.
At a high level, clockintr(9) is configured and used as follows:
1. During boot, the primary CPU calls clockintr_init(9). Global state
is initialized.
2. Primary CPU calls clockintr_cpu_init(9). Local, per-CPU state is
initialized. An "intrclock" struct may be installed, too.
3. Secondary CPUs call clockintr_cpu_init(9) to initialize their
local state.
4. All CPUs repeatedly call clockintr_dispatch(9) from the MD clock
interrupt handler. The CPUs complete work and rearm their local
interrupt clock, if any, during the dispatch.
5. Repeat step (4) until the system shuts down, suspends, or hibernates.
6. During resume, the primary CPU calls inittodr(9) and advances the
system uptime.
7. Go to step (2). This time around, clockintr_cpu_init(9) also
advances the work schedule on the calling CPU to skip events that
expired during suspend. This prevents a "thundering herd" of
useless work during the first clock interrupt.
In the long term, we need an MI clock interrupt scheduler in order to
(1) provide control over the clock interrupt to MI subsystems like
timeout(9) and dt(4) to improve their accuracy, (2) provide drivers
like acpicpu(4) a means for slowing or stopping the clock interrupt on
idle CPUs to conserve power, and (3) reduce the amount of duplicated
code in the MD clock interrupt code.
Before we can do any of that, though, we need to switch every platform
over to using clockintr(9) and do some cleanup.
Prompted by "the vmm(4) time bug," among other problems, and a
discussion at a2k19 on the subject. Lots of design input from
kettenis@. Early versions reviewed by kettenis@ and mlarkin@.
Platform-specific help and testing from kettenis@, gkoehler@,
mlarkin@, miod@, aoyama@, visa@, and dv@. Babysitting and spiritual
guidance from mlarkin@ and kettenis@.
Link: https://marc.info/?l=openbsd-tech&m=166697497302283&w=2
ok kettenis@ mlarkin@
|
|
ok deraadt@
|
|
|
|
ok deraadt@
|
|
|
|
It makes uvm_swap_free() faster: extents have a cost of O(n*n) which doesn't
really scale with gigabytes of swap.
Based on initial work from mpi@
The blist implementation comes from DragonFlyBSD.
The diff adds also a ddb(4) 'show swap' command to show the blist and help
debugging, and fix some off-by-one in size printed during hibernate.
ok mpi@
|
|
version number issues close to release
|
|
If ddb.console is set and your serial console driver uses it, db_rint(),
lets you enter ddb(4) by typing the ESC D escape sequence. This is
useful for drivers like sfuart(4) where the hardware doesn't have a true
BREAK mechanism.
Suggested by miod@, ok kettenis@ miod@
|
|
kstat allows the kernel to expose arbitrary data for userland to
consume. currently this is used by some network card drivers to
expose hardware counters they provide, and a bit by the network
stack to show things like ifq counters.
ok bluhm@ deraadt@
|
|
just 2*NPROCESS rather than NPROCESS+8*MAXUSERS. Results in a slightly
higher maxthread value - the previous value was fairly likely to be
exceeded on a system running a couple of heavily threaded processes.
> previous new
> MAXUSERS NPROCESS maxthread (2*NPROCESS)
> 80 1310 1950 2620
> 64 1054 1566 2108
> 32 542 798 1084
ok kettenis@
|
|
|
|
|
|
|
|
|
|
Based on our existing RA module for 11n.
The main difference is in dealing with 11ac-specific ratesets.
Tx rate selection heuristics remain identical.
Only supports 80MHz channels, for now. 160MHz is left for future work.
ok sthen@
|
|
lm700x: az, rt
tc921x: sfr
pt2254a: sfr, sf2r
|
|
used by wdsc on sgi (removed in 2021)
ok krw@
|
|
|
|
|
|
|
|
the big change is removing the integration with and reliance on
bridge(4) for learning vxlan endpoints. we have the etherbridge
layer now (which is used by veb, nvgre, bpe, etc) so vxlan can
operate independently of bridge(4) (or any other driver) while still
dynamically learning about other endpoints.
vxlan now uses the udp socket upcall mechanism to receive packets.
this means it actually creates and binds udp sockets to use rather
adding code in the udp layer for stealing packets from the udp
layer.
i think it's also important to note that this adds loop prevention
to the code. this stops a vxlan interface being used to transmit a
packet that was encapsulated in itself.
i want to clear this out of my tree where it's been sitting for
nearly a year. noone seems too concerned with the change either
way.
ok claudio@
|
|
This splits out the MI sequencing, backing it with per-architecture helper
functions. Further steps will be neccesary because ACPI and MD are too
tightly coupled, but soon we'll be able to use this code for more architectures
(which depends on figuring out the lowest-level cpu sleeping method)
ok kettenis
|
|
OK deraadt@ phessler@
|
|
|
|
if it is supported. Remove it from the global GENERIC config.
OK visa@ claudio@
|
|
API implemented is a deadend.
OK akoshibe@ yasuoka@ deraadt@ kn@ patrick@ sthen@
|
|
|
|
|
|
|
|
ok deraadt@
|
|
|
|
this allows us to dynamically trace function boundaries with btrace by patching
prologues and epilogues with a breakpoint upon which the handler records the data,
sends it back to userland for btrace to consume.
currently it's hidden behind DDBPROF, and there is still a lot to cleanup and
improve, but basic scripts that observe return codes from a probed function
work.
from Tom Rollet, with various changes by me
feedback and ok mpi@
|
|
|
|
ok deraadt@ kettenis@
|
