timecounting: remove incomplete PPS support

The timecounting code has had stubs for pulse-per-second (PPS) polling
since it was imported in 2004. At this point it seems unlikely that
anyone is goin

timecounting: remove incomplete PPS support

The timecounting code has had stubs for pulse-per-second (PPS) polling
since it was imported in 2004. At this point it seems unlikely that
anyone is going to finish adding PPS support, so let's remove the stubs:

- Delete the dead tc_poll_pps() call from tc_windup().
- Remove all tc_poll_pps symbols from the kernel.

Link: https://marc.info/?l=openbsd-tech&m=167519035723210&w=2

ok miod@

show more ...

acpihpet(4): disable/reenable acpihpet_delay() during suspend/resume

We can't use the HPET to delay(9) after we halt it during suspend.
Disable acpihpet_delay() before we halt the HPET and reenable

acpihpet(4): disable/reenable acpihpet_delay() during suspend/resume

We can't use the HPET to delay(9) after we halt it during suspend.
Disable acpihpet_delay() before we halt the HPET and reenable it after
we restart the HPET during resume.

ok mlarkin@

show more ...

acpihpet(4): acpihpet_delay: only use lower 32 bits of counter

We can't use acpihpet_r() to implement acpihpet_delay(). Even if we
made acpihpet_r() atomic on amd64, i386 would still be incapable o

acpihpet(4): acpihpet_delay: only use lower 32 bits of counter

We can't use acpihpet_r() to implement acpihpet_delay(). Even if we
made acpihpet_r() atomic on amd64, i386 would still be incapable of
doing atomic 8-byte reads. As-is, the code does a split read on all
platforms, which may or may not already be causing problems with TSC
calibration:

https://marc.info/?l=openbsd-tech&m=166220561709496&w=2

Switch from acpihpet_r() to bus_space_read_4() and only use the lower
32 bits of the counter. This makes acpihpet_delay() slightly larger,
but unless we want two acpihpet_delay() implementations we have no
choice.

Link: https://marc.info/?l=openbsd-tech&m=166165347220077&w=2

ok jsg@

show more ...

acpihpet(4): don't truncate HPET frequency to 32-bit value

timecounter.tc_frequency is a 64-bit value: this cast is unnecessary.

Split off from the acpihpet_delay() commit at jsg@'s suggestion.

Li

acpihpet(4): don't truncate HPET frequency to 32-bit value

timecounter.tc_frequency is a 64-bit value: this cast is unnecessary.

Split off from the acpihpet_delay() commit at jsg@'s suggestion.

Link1: https://marc.info/?l=openbsd-tech&m=166053729104923&w=2
Link2: https://marc.info/?l=openbsd-tech&m=166132727120528&w=2

probably ok jsg@

show more ...

acpihpet(4): add acpihpet_delay(), another delay(9) implementation

When lapic_delay() is removed from the tree in the near future, older
machines without a constant/invariant TSC will need a delay(9

acpihpet(4): add acpihpet_delay(), another delay(9) implementation

When lapic_delay() is removed from the tree in the near future, older
machines without a constant/invariant TSC will need a delay(9)
implementation better than i8254_delay().

This patch adds acpihpet_delay(), a delay(9) implementation based on
the ACPI HPET timer. It is preferable to i8254_delay() (0) and
acpitimer_delay() (1000), so set its quality to 2000.

On newer machines, the HPET is slower to read than the PMT on newer
machines for reasons unknown, so _technically_ this quality hierarchy
is not always accurate. However, we expect these newer machines to
have a constant/invariant TSC available, so the inaccuracy is harmless
because tsc_delay() will be available, which is better than both the
PMT and the HPET.

In general, on real hardware that predates wide availability of the
constant/invariant TSC, the HPET is preferable to the PMT.

With input from jsg@.

Link: https://marc.info/?l=openbsd-tech&m=166053729104923&w=2

ok mlarkin@ jsg@

show more ...

constify struct cfattach

timecounting: use C99-style initialization for all timecounter structs

The timecounter struct is large and I think it may change in the
future. Changing it later will be easier if we use C99-style

timecounting: use C99-style initialization for all timecounter structs

The timecounter struct is large and I think it may change in the
future. Changing it later will be easier if we use C99-style
initialization for all timecounter structs. It also makes reading the
code a bit easier.

For reasons I cannot explain, switching to C99-style initialization
sometimes changes the hash of the resulting object file, even though
the resulting struct should be the same. So there is a binary change
here, but only sometimes. No behavior should change in either case.

I can't compile-test this everywhere but I have been staring at the
diff for days now and I'm relatively confident this will not break
compilation. Fingers crossed.

ok gnezdo@

show more ...

Add support for timeconting in userland.

This diff exposes parts of clock_gettime(2) and gettimeofday(2) to
userland via libc eliberating processes from the need for a context
switch everytime they

Add support for timeconting in userland.

This diff exposes parts of clock_gettime(2) and gettimeofday(2) to
userland via libc eliberating processes from the need for a context
switch everytime they want to count the passage of time.

If a timecounter clock can be exposed to userland than it needs to set
its tc_user member to a non-zero value. Tested with one or multiple
counters per architecture.

The timing data is shared through a pointer found in the new ELF
auxiliary vector AUX_openbsd_timekeep containing timehands information
that is frequently updated by the kernel.

Timing differences between the last kernel update and the current time
are adjusted in userland by the tc_get_timecount() function inside the
MD usertc.c file.

This permits a much more responsive environment, quite visible in
browsers, office programs and gaming (apparently one is are able to fly
in Minecraft now).

Tested by robert@, sthen@, naddy@, kmos@, phessler@, and many others!

OK from at least kettenis@, cheloha@, naddy@, sthen@

show more ...

Use <stdint.h> types.

ok mpi@, jung@, krw@, deraadt@

Recalibrate TSC timecounter with HPET and PM timer

If frequency of an invariant (non-stop) time stamp counter is measured
using an independent working timecounter that has a known frequency, we
can

Recalibrate TSC timecounter with HPET and PM timer

If frequency of an invariant (non-stop) time stamp counter is measured
using an independent working timecounter that has a known frequency, we
can assume that the measured TSC frequency is as good as the resolution
of the timecounter that we use to perform the measurement. This lets us
switch from this high quality but expensive source to the cheaper TSC
without sacrificing precision on a wide range of modern CPUs.

From Adam Steen <adam@adamsteen.com.au> with tweaks from reyk@ and myself.

Tested by brynet@, sthen@ and others, OK mlarkin, sthen

show more ...

0xffffffff is 32 bits, not 24 bits

ok deraadt

Check proper HPET period value during attach.

Pointed out by patrick keshishian <pkeshish at gmail.com>, thanks.

do not save & restore a capability register, it is RO.
issue noticed by patrick kehishian, ok mlarkin

Remove some includes include-what-you-use claims don't
have any direct symbols used. Tested for indirect use by compiling
amd64/i386/sparc64 kernels.

ok tedu@ deraadt@

force cfattach decl to same form as others

save/restore hpet configuration. Perhaps a little more than we
need to do, since we do not use many hpet features.
ok kettenis millert

be more careful during suspend/resume cycles, as roughly detailed in the
spec. More might be required.
ok jordan matthew mlarkin

we are past the point where timecounters may disappear
ok miod

Attach acpihpet only once. The matching code is changed because acpi bus
can be enumerated which means match routine should do the right thing and
prevent further attachments.

With input from kette

Attach acpihpet only once. The matching code is changed because acpi bus
can be enumerated which means match routine should do the right thing and
prevent further attachments.

With input from kettenis, deraadt, miod; ok deraadt, kettenis

show more ...

spacing and indents that are driving me crazy

knf

Fix pckbd and acpihpet suspend/resume cases to fit with the future framework
that's going to get in really soon.

Okay mlarkin@, deraadt@.

Get rid of devact enum, substitute it with an int and coresponding defines.

This is needed for the addition of further suspend/resume actions.

Okay deraadt@, marco@.

Add wake support to acpihpet(4).

Patch initially from mlarkin@. KNF and refactoring by me. Okay deraadt@.

Add workaround for broken ATI southbridges; inspired by how Linux handles this.
Fixes PR 5916 & 5959.

ok toby@