Here's the 2.5 version of the patch to uhci.c to finish completions in
the correct order.

Used kobject reference counting to free the scn struct when the device
is closed and disconnected. Avoids crashes when writing to a
disconnected device. (Thanks to Greg KH).

I've also changed irq_scanner to avoid submitting new URBs when the
old one returned with an error. Without this change irq_scanner gets
called ever and ever again after a disconnect while open.

> This exhibits a build error when OHCI_VERBOSE_DEBUG is enabled:

Odd, I guess the build I tested was when that was enabled
without first enabling debugging.  The fix is trivial.

Thanks to Martin Brachtl <brachtl@redgrep.cz> for the information.

Thanks to David for pointing this out.

Write multiple cells in one function call, rather than one cell per
function call.  Under maximum send load, this reduces cell writing
CPU usage from 0.0095% to 0.0085% on my machine.  A 10% improvement! :)

Thanks to C Falconer <cf@avonside.school.nz> for the information.

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

From Ben Collins

Used as manual trigger for device/driver binding, so drivers can claim 
devices that are present but failed to be bound. In case of ieee1394, this
can happen if two computers are attached to a device - the second one won't
be able to claim it until the first detaches. 

Changed device_attach() to return 1 when a driver was bound to a device. 
Makes it so bus_rescan_devices() can return the number of devices attached.
Tested this with ieee1394 over a bus reset and it worked as expected.

- Don't d_delete() files when removing a directory - they must still be 
  hashed when dput() reaps them, and it will __d_drop() them anyway.

- Don't d_invalidate() the directory on removal. 

- do simple_rmdir() after d_delete() on directory. 

- Remove extra dput(), which was causing refcount to go negative, causing 
  an oops when someone tried to create the directory again.

- Make sure we don't d_delete() or do extra dput() on file when updating, 
  either.

into shaggy.austin.ibm.com:/shaggy/bk/jfs-2.5

The old adap_lock and driver_lock could deadlock and there was a bunch
of stuff that was horribly racy.  The new locking scheme is from
lm_sensors CVS but I had to fix quite a few bugs to make this work.

This one is from lm_sensors CVS and mostly cosmetic changes.  I it up a
bit to compile properly under latest 2.5.

another bunch of IDs appeared in lm_sensors CVS lately..

This somehow got lost in the uClinux merge.

It's the _device_ filesystem, so regular files are grossly misplaced
here.  Fortauntely only one driver actually tries to use it (microcode)
and it's works fine with  a regular miscdevice as well.

Sorry, the devs_only() patch accidentally included the devfs-portion
of the DEVFS_FL_AUTO_DEVNUM patch.  This patch contains the remaining
parts.

The new slab poisoning code broke kmem_cache_size(), it now returns
a too large size as the poisoning area after the object is includes.
XFS's kmem_zone_zalloc thus overwrites exactly that area and triggers
the new checks everytime such an object is freed again.

I don't recommend using XFS on BK-current without this patch applied :)

This patches fixes the m68knommu 68360 serial driver to use the
work_struct, not the obsoleted workqueue.

This patch moves common ColdFire vector and timer procesing code from
the local per-processor config.c for the 5206 ColdFire sub-architecture.
All ColdFire CPU's have the same timer and basic vector setup, seems
crazy to repeat this code for each of 6 ColdFire CPU varients.

This patch moves common ColdFire vector and timer procesing code from
the local per-processor config.c for the 5206e ColdFire sub-architecture.
All ColdFire CPU's have the same timer and basic vector setup, seems
crazy to repeat this code for each of 6 ColdFire CPU varients.
This patch also removes the reset button support, this is now moved to
a proper device driver, where it belongs.

This patch moves common ColdFire vector and timer procesing code from
the local per-processor config.c for the 5249 ColdFire sub-architecture.
All ColdFire CPU's have the same timer and basic vector setup, seems
crazy to repeat this code for each of 6 ColdFire CPU varients.

This patch creates common timer setup and handling code for the m68knommu
ColdFire CPU sub-architecture. All 6 ColdFire CPU varients contain the
same hardware timers. Currently each sub-arch has its own timer code.
This reduces a lot of code duplication.

This patch puts the new common ColdFire timers.c and vectors.c into
the build list (in the m68knommu architecture branch).

This patches fixes the m68knommu 68328 serial driver header to use the
work_struct, not the obsoleted workqueue.

This patch moves common ColdFire vector and timer procesing code from
the local per-processor config.c for the 5407 ColdFire sub-architecture.
All ColdFire CPU's have the same timer and basic vector setup, seems
crazy to repeat this code for each of 6 ColdFire CPU varients.

This patch fixes up the PCI bios header for the m68knommu COMEM-lite
PCI hardware support. Adds missing definitions, and a couple of trivial
inline functions here.

This patches add an include to stddef.h into include/linux/list.h.
It uses the NULL define.

This patch is the last of the MMUless support changes to the mm system.
It inlines for MMUless targets the following functions:

	vmalloc_to_page()
	follow_page()
	remap_page_range()

All of these return NULL for the non-MMU case. Doing the inline here
minimizes changes to assorted mm/ files. This patch was originally
from Christoph Hellwig.

This patch fixes a wrong m68knommu/68360 config define,
CONFIG_68630_SCC -> CONFIG_68360_SCC.

This patch moves common ColdFire vector and timer procesing code from
the local per-processor config.c for the 5307 ColdFire sub-architecture.
All ColdFire CPU's have the same timer and basic vector setup, seems
crazy to repeat this code for each of 6 ColdFire CPU varients.
This patch also removes the reset button support, this is now moved to
a proper device driver, where it belongs.

This patch adds a call to schedule_tail() in the return from fork
post processing.

This patch fixes the m68knomu/68VZ328/ucdimm Makefile to build all its
local code files. Previously it relied on Makefiles above it to do
this. It should be done here.

This patch fixes the m68knomu/68VZ328/de2 Makefile to build all its
local code files. Previously it relied on Makefiles above it to do
this. It should be done here.

This patch moves common ColdFire vector and timer procesing code from
the local per-processor config.c for the 5272 ColdFire sub-architecture.
All ColdFire CPU's have the same timer and basic vector setup, seems
crazy to repeat this code for each of 6 ColdFire CPU varients.
This patch also removes the reset button support, this is now moved to
a proper device driver, where it belongs.

This patch fixes up a lot of problems in the m68knommu architecture
CO-MEMlite PCI interface code. The code was seriously out of date,
not implementing many of the requried PCI bios support functions.
Also fixes many wrong function return types and arguments.

Fix m68knommu 68VZ328 Makefile to only traverse its sub-directories.
There is no local code files for it to compile.

Include errno.h in the m68knommu 68360 specific interrupt setup code.
It needs a few E* types as return values.