Use destination buffer size as strlcpy() bound, not source; harmless here for
both sizes are equal.
ok kettenis@

Constify struct cfattach.

ok visa@ a long time ago

sparc64: lom(4), sab(4), vldcp(4): tsleep -> tsleep_nsec(9); ok kn@ visa@

Check the state value within the assert statement, do not set it.
found with grep 'KASSERT.* = '; OK kettenis@

Convert watchdog(4) devices to use autoconf(9) framework.

ok deraadt, tests on glxpcib and ok mpi

Enforce ca_activate tree-walks over the entire heirarchy for all events,
cleaning up some shutdown-hook related code on the way.
(A few drivers related to sparc are still skipped at kettenis' request

Enforce ca_activate tree-walks over the entire heirarchy for all events,
cleaning up some shutdown-hook related code on the way.
(A few drivers related to sparc are still skipped at kettenis' request)
ok kettenis mlarkin, tested by many others too

show more ...

Swap arguments to wdog_register() since it is nicer, and prepare
wdog_shutdown() for external usage.

Add missing initialization.

Fix off-by-one in LOMlite hostname code.

Remove debug printf and properly dequeue command instead when a read times out
on LOMLite2.

Handle LOMlite2 in an interrupt-driven way; avoids using delay(9) once the
machine is up and running.

Establish a shutdown hook to disable the watchdog timer to prevent watchdog
triggers after the kernel has been halted.

Make sure we don't insert and entry into the list of pending commends twice.

LOMlite seems to get wedged from time to time; add some code to unwedge it.

Print the LOM type (LOMlite/LOMlite2).

Add sensor status for fan sensors.

Implement queuing of reads and writes. This gets rid of the remaining busy
waits for LOMlite.

Slightly less evil busy wait loop for LOMlite. Small tweak to LOMlite state
machine.

Implement reading LOMlite registers using timeouts and a state machine to
avoid busy waits. Writes are still somewhat problematic, but things work
well enough to make lom(4) attach to LOMlite now.

Implement reading LOMlite registers using timeouts and a state machine to
avoid busy waits. Writes are still somewhat problematic, but things work
well enough to make lom(4) attach to LOMlite now.

Fix typo pointed out by miod@ while I'm there.

show more ...

Add some more LOMlite bits. We don't attach to LOMlite yet because it is
very slow and this version of the code busy waits too much.

Attempt mapping as io as well as mmio; necessary to make this work on a V100.

Restructure watchdog support such that it will work on both LOMlite and
LOMlite2.

Add PSU sensor.

Explicitly configure the watchdog timer not to reset the machine when we
attach. If we don't do this, the machine will reset even if the watchdog
timer hasn't been explicitly configured if the watch

Explicitly configure the watchdog timer not to reset the machine when we
attach. If we don't do this, the machine will reset even if the watchdog
timer hasn't been explicitly configured if the watchdog timer was previously
enabled.

show more ...

Add support for the built-in watchdog. In the default state the watchdog will
enable the fault LED after 127 seconds if the kernel stops patting. You'll
have to explicitly enable the watchdog to ma

Add support for the built-in watchdog. In the default state the watchdog will
enable the fault LED after 127 seconds if the kernel stops patting. You'll
have to explicitly enable the watchdog to make it reset the machine.

show more ...