- Do reference counting for 'struct mount'. Each vnode associated with a
mount takes a reference, and in turn the mount takes a reference to the
vfsops.
- Now that mounts are reference counted, replace the overcomplicated mount
locking inherited from 4.4BSD with a recursable rwlock.
interrupts/sec -- not the other way around. Caught by yamt.
When I can confirm that it won't lock the chip up on the models claimed
to be problematic, I'll probably adjust the packet timers a bit further
to see if I can get latency down under low load. But this should be
better.
as needed to read a request instead of the maximum message size.
Insanely enough, this makes "cheap backend I/O" file systems such
dtfs and sysctlfs perform 10-30% better (depending on the average
size of incoming requests).
transitioning a device to or from D3hot, 200us to or from D2, and no wait
for D1 transitions. Update pci_set_powerstate_int to match the spec,
rather than naively waiting 1ms after every state transition.
Fix FIRMWORKSBUGS. The #else on one of the ifdefs for it made it so that
it would never work on a machine that didn't need it. This caused IBM
machines to be unable to talk to OFW if the option was enabled. Replace
the #else with a branch and call it good.
from S3 sleep. Until we get proper ACPI power resource support, simply
call \\_SB.PCI0.LPC.EC.PUBS._ON() on resume if the method exists.
Fixes kern/37279 by Steven M Bellovin.
Previously each request was executed on its own callcontext and
switched to every time the request was being processed. Now requests
borrow the mainloop context and switch only if/when they yield.
This takes two context switches away from a file system request
bringing down the typical amounts 2->0 (e.g. dtfs) and 4->2 (e.g.
psshfs).
The interfaces for manually executing requests changed a bit:
puffs_dispatch_create() and puffs_dispatch_exec() must now be used.
They are not tested, as nothing in-tree wants them and I doubt
anyone else is really interested in them either.
Also do some misc code cleanup related to execution contexts. The
"work-in-progress checkpoint" committed over a year ago was starting
to look slightly weed-infested.
ad