in usb.c during disconnect handling a reference to a driver is passed
after the module usage count is decremented. In theory this is a race.

The fix basically removes a bit-complement that shouldn't have
been there.  (Plus related simplification.)  That changed the PID,
so that (for one example) only the first hub plug/unplug event could
work.  So it's essential for anyone using high speed hubs.

I'm still not quite sure why this hasn't affected more testing, but
that's life ... :)

fix a failure to free a minor in the error path of probe

device would have very long string descriptors (vendor and device
name.)

into penguin.transmeta.com:/home/penguin/torvalds/repositories/kernel/linux

Upgrade is imperative!

into penguin.transmeta.com:/home/penguin/torvalds/repositories/kernel/linux

Minor cleanups to a scsi fibre channel HBA driver that was broken by the
introduction of SCSI_REPORT_LUNS to scsi.h in a patch that went into
2.5.11 .

[ Another transfer from Dave's tree ]

Documentation enhancement from Kai.Makisara@metla.fi.

[ This is a transfer from Dave's tree ]

Fix oopses on ISA scsi adapters.

[ This is a transfer from Dave's tree ]

- The balance_dirty_pages() logic is simply wrong.  It goes:

	if (value > threshold)
		go_and_write(value - threshold);

  which is just fine for a single process writing data.  But
  for many processes, they *all* go and bring things back into
  balance, and too much data gets written out.

- The

	go_and_write(this much)

  logic is inoperative, because I turned off the ->writeback_mapping()
  function in ext2.  So a call to writeback_unlocked_inodes(this_much)
  doesn't actually decrement and test *this_much.  It will walk every
  inode, all the time.  Silly.

So quickly fixing the above things, the amount of dirty+writeback
memory in the machine nicely stabilises at 500 megabytes across
the run.

The new consistency checks in buffer.c are complaining about
block_symlink's ->readpage against an already uptodate
page.

- Fix compilation of piix.c

- Remove the PADAM_ prefix from sleep hwgroup member.

- Fix Pacific Digital host chip driver API.

- Fix Tekram host chip driver API.

- Fold hwif_unregister() directly in to channel code.

Add support for request tagging of the block queue.

This patch makes usbvideo_register take a usb_device_id argument which it
can then pass to usb_register via the struct usb_device. Currently it passes
NULL.

This patch makes usbvideo.c use the correct value of USBVIDEO_NUMFRAMES.
A few places in the code assumed it was 2. This patch was tested with
USBVIDEO_NUMFRAMES = 2,3,4,8,11,32

cleaned up checking the return value of usb_register_dev()

added Black Box OEM device ids to the io_edgeport driver.

USB pl2303 driver

This is a patch which avoids resetting the termios settings to default
values (9600 Baud etc.) in each call to pl2303_open (). It does this
only on the first call to pl2303_set_termios. After that it sets the
termios to the last stored values.

 This way commands like
 stty -F /dev/ttyUSB0 115200
 work the same way as with other serial ttys.

USB printer bugfix

looking through printer.c in preparation for shifting devfs
support to usbcore I noticed that printer advertises a device
through devfs before it can be opened.
As devfs, or more precisely devfsd can be used to trigger actions
this matters and is wrong.

set_blocksize.

into home.transmeta.com:/home/torvalds/v2.5/linux

- Fix macro varargs usage, you need to specify a variable name
before the triple dot or else most current compilers complain.
See include/linux/ext2_fs.h:ext2_debug for another example where
it is done correctly.

- Fix bug in strsep/strchr changes, dereference *this_char
not *value at top of while loop.  This matches how the same
code in fs/autofs4/inode.c looks right now.
Withtout this autofs loading causes an OOPS as the first
time through the loop *value is dereferencing a NULL pointer.

Allocate and free the avmcard structure using a shared function.

Remove unnecessary wrappers around irq handlers.

Interrupt handlers won't be reentered anyway.

Fix some set_bit() warnings.

The real rule is not "block size is at most 32Kb", it's "block size is
at most 8 times the fragment size".

into penguin.transmeta.com:/home/penguin/torvalds/repositories/kernel/linux

(originally caught by Paul Larson <plars@austin.ibm.com>)

	Patch fixes a couple of idiotic bugs in notify_change() (my
fault).  Unary operations have higher priority than binary ones, so
if (!valid & ATTR_MODE) doesn't do the right thing.  I plead temporary
braindamage...