I've got a Sony CDU33A drive with COR334 controller.  The Linux cdu31a
driver was not updated for 2.6 kernel so it does not work.

Here are patches that try to make the driver working with 2.6
kernel:
  - fix the timeout values in header file
  - Make the driver work in 2.6.X
  - Added workaround to fix hard lockups on eject
  - Fixed door locking problem after mounting empty drive
  - Set double-speed drives to double speed by default
  - Removed all readahead things - not needed anymore

It does work on my system. I also know that it's still broken - it uses 
cli(), MODULE_PARM and it's also not very fast (I _never_ reached full
300KB/s with it, but I know that it's possible in Windows) and probably 
many other things (I'm new to Linux kernel) - so I'm waiting for comments.

They happen with "bk cset -x" - it only undoes the data
changes, not the metadata updates.

For some as-yet unclear reason, it seems to break legacy
32-bit apps on x86-64.

Cset exclude: ak@suse.de[torvalds]|ChangeSet|20041102230307|11901

into ppc970.osdl.org:/home/torvalds/v2.6/linux

XFER_xxx is not necessarily "legacy IDE 'stuff'"

This is todays BK on a x86:

  CC [M]  drivers/usb/storage/freecom.o
In file included from include/linux/hdreg.h:4,
                 from drivers/usb/storage/freecom.c:32:
include/linux/ata.h:197: error: parse error before "u32"
...
and so on for tens of lines.
Signed-off-by: Jeff Garzik <jgarzik@pobox.com>

into ppc970.osdl.org:/home/torvalds/v2.6/linux

I'm not setting the rover pointer at all.  Here's a patch to
fix that.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>

When a mirred action is created it takes two references to the
device (dev_get_by_index + dev_hold), but only drops one when
it gets destroyed. It also leaks a reference when a mirred action
is replaced.
Signed-off-by: Patrick McHardy <kaber@trash.ne>
Signed-off-by: David S. Miller <davem@davemloft.net>

This patch gets rid of an unused global counter in neighbour.c.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>

This patch fixes the return codes from sendmsg/recvmsg when a signal happens.
Instead of always returning ERESTARTSYS (which confuses X11 rather badly) it
should return EINTR for non-blocking operations.
Signed-off-by: Patrick Caulfield <patrick@tykepenguin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This patch fixes a missing _bh in the 23 August locking update. 
Without this any attempt to read from /proc/net/decnet_cache causes a
'scheduling while atomic' crash.
Signed-off-by: Patrick Caulfield <patrick@tykepenguin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Fixes broken packets, noticed by Matthias Andree <matthias.andree@gmx.de>
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>

Moxa is short the final cleanup from the tty changes so has extra wakeup
calls that can go

into ppc970.osdl.org:/home/torvalds/v2.6/linux

From: Alan Cox <alan@lxorguk.ukuu.org.uk>
(some changes by me - bart)

The current driver looks at fields before it is safe to, we move the
mod15rm bug handler to be a fixup and this ensures the probe has been
completed before we use the ident data.
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>

From: Alan Cox <alan@lxorguk.ukuu.org.uk>
(some changes by me - bart)

We add probe_hwif_init_with_fixup (seperate naming as requested by
Bartlomiej). This runs a fixup on present interfaces before attaching
the drives. In order to be useful we need also an _with_fixup version
of ide_register_hw function. 

The sometimes troublesome undecoded slave detector is moved to its own
function and exported so that ide-cs and the upcoming delkin_cb can both
use it (along with any arch specific cf/pcmcia drivers I don't know
about). The non-relevant checks for this scenario are removed.
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>

From: Alan Cox <alan@lxorguk.ukuu.org.uk>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>

Signed-off-by: Chris Wedgwood <cw@f00f.org>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>

From: Chris Wedgwood <cw@f00f.org>
(minor changes by me - bart)

Remove some accumulated (unused) cruft from ide.h
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>

ide_hwif_setup_dma() calls pci_set_master() if necessary.
pci_set_master() (indirectly via pcibios_set_master()) sets
PCI_LATENCY_TIMER to a "reasonable" value.
Signed-off-by: Grant Grundler <grundler@parisc-linux.org>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>

From: Grant Grundler <grundler@parisc-linux.org>
(minor changes by me - bart)

Move Superio (NatSem 87560) IDE hacks into the ns87415 driver
instead of exporting IDE data structures.

To date, only PA-RISC port is known to use the 87560 chip.
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>

sata_[misc,scr] and priv are write-only
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>

Signed-off-by: Arnaldo Carvalho de Melo <acme@conectiva.com.br>
Signed-off-by: David S. Miller <davem@davemloft.net>

The patch 'RCU: use rcu_assign_pointer()'
http://linux.bkbits.net:8080/linux-2.6/diffs/net/sched/sch_api.c@1.39?nav=index.html|src/net/|src/net|src/net/sched|related/net/sched/sch_api.c|cset@1.2287
changed a list_add_tail to list_add_tail_rcu in qdisc_create. It's
dev->qdisc not dev->qdisc_list that is protected by RCU, this patch
reverts that change. It also removes a misleading comment and replaces
the smp_wmb in qdisc_create_dflt by rcu_assign_pointer in dev_activate
to document more clearly what is protected.
Signed-off-by: David S. Miller <davem@davemloft.net>

Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

None of the code in __qdisc_destroy should be applied to a builtin qdisc
or am I missing something?

The patch below prevents builtin qdiscs from being destroyed and
fixes a refcnt underflow whould lead to a bogus list unlinking
and dev_put.
Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

Fixes some build failures ala bugme bugzilla #3706
Signed-off-by: David S. Miller <davem@davemloft.net>

into ppc970.osdl.org:/home/torvalds/v2.6/linux

Patch from Tony Lindgren

This patch by Tuukka Tikkanen updates the OMAP clock framework.
It also contains a patch by Juha Yrjölä to enable 216MHz rate
on 16xx.

Signed-off-by: Tony Lindgren 
Signed-off-by: Russell King

Patch from Tony Lindgren

This patch changes OMAP serial ports to use the standard 8250
autodetection.

Signed-off-by: Tony Lindgren 
Signed-off-by: Russell King

Patch from Ben Dooks

Ensure that TI OMAP systems are only offered if
the system is configured for TI OMAP.

Signed-off-by: Ben Dooks 
Signed-off-by: Russell King

Patch from Ben Dooks

Fix IDCODE for S3C2410A to add missing zero in hex number

Thanks to Russell King for pointing this out

Signed-off-by: Ben Dooks 
Signed-off-by: Russell King

Patch from Ben Dooks

Added missing exports of the functions in gpio.c

Thanks to Dimitry Andric for pointing this out

Signed-off-by: Ben Dooks 
Signed-off-by: Russell King

These are no longer necessary since MMC support is now merged.