* QUIRK_NOMSG only has any meaning when NCR_GETCC_WITHMSG
is defined. Therefore, there's no harm in using it when
NCR_GETCC_WITHMSG is not defined. so, simplify the table
by removing the #ifdef.
* there's really no point in having table entries after
an entry which will match everything.
* add some comments, clean up spacing.
* add an entry for "QUANTUM"/"ATLAS IV" drives with flags
QUIRK_NOTAGS|QUIRK_NOMSG. (I included the latter flag only
because everything else had it before! ... which means that
all the functionality added with the NCR_GETCC_WITHMSG define
would never get used! *sigh*) The latter fixes the problems
I was having on an Atlas, and should fix the problems mentioned
by Hans Hoppe <hopha@casema.net> in comments on PR#7694.
1.4.x and i have concerns (but no concrete proof) they will cause/have caused
problems in -current as well. Really, the right way to fix this is to
rewrite the driver, and push up tagged queueing handling into a common
middle layer that'll do it right in a low-level-driver-independent manner.
I'll fix my particular issues by using the ncr driver quirk mechanism.
it was a bit too strong, and forbids multiple addresses from
same prefix to be assigned.
now the behavior is the same as previous - memory leak on interface address
addition failure.
http://orange.kame.net/dev/query-pr.cgi?pr=218
- fix a bug in ti_encap() where a NULL pointer could be used if we run out
of tx descriptors (return ENOMEM instead, and let the caller assert OACTIVE)
- sync boot messages with NetBSD's pci standart (print ti_name and irq
line)
- add a shutdown hook to stop the board, so that the board will not DMA at
random places while we're rebooting
- convert all but TIGON 1 transmit descriptors to use bus_space
(this last one isn't trivial to get rigth and I don't have the hardware to
test)
- If we can't map the mem space linar, retry to map it nolinear. Punt if
we couldn't map the mem space linear and we found a Tigon 1 chip.
With these changes a Tigon 2 board (netgear BA620 at last) can be used in
my DS20.
Resources are initialized still just once (on first call).
Add ufs_done(), which takes care of freeing all resources allocated in
ufs_init(). The resources are freed only when last user of the code exits.
in vfs_detach(). vfs_done may free global filesystem's resources,
typically those allocated in respective filesystem's init function.
Needed so those filesystems which went in via LKM have a chance to
clean after themselves before unloading.
For each leaf filesystem, add appropriate vfs_done routine.
Also remember how many times ffs_init() was called and do
the appropriate initialization on first call only. In ffs_done(),
destroy the resources when called by the last user of ffs code.
Change mfs to call ffs_init()/ffs_done() appropriately.
in vfs_detach(). vfs_done may free global filesystem's resources,
typically those allocated in respective filesystem's init function.
Needed so those filesystems which went in via LKM have a chance to
clean after themselves before unloading. This fixes random panics
when LKM for filesystem using pools was loaded and unloaded several
times.
For each leaf filesystem, add appropriate vfs_done routine.
erh