Bug reported by Paul Mackerras: if a module parameter fails, we didn't
call module_arch_cleanup().

On x86 this was harmless (module_arch_cleanup() is a no-op), but on
other architectures like PPC this causes inconsistent data structures
and subsequent oopses.

into home.osdl.org:/home/torvalds/v2.5/linux

This bug prevents Alphas with older firmware from booting if there
is a card with PCI-PCI bridge that supports 32-bit IO.
This has happened on AS2100 with a quad-tulip card, for example:
 - initially, the I/O window of 21152 bridge was 0x10000-0x10fff,
   as set up by firmware;
 - pci_setup_bridge() is going to change this, say, to 0xa000-0xafff:
   first, it updates PCI_IO_BASE_UPPER16 and PCI_IO_LIMIT_UPPER16
   registers, so that IO window temporarily is at 0x0000-0x0fff,
   which effectively blocks up all legacy IO ports in the lower
   4K range, such as serial, floppy, RTC an so on;
   does debugging printk - machine dies here with recursive
   machine checks as the serial console has gone.

Moving (or disabling) the debugging printk is not a solution -
there is possibility that timer interrupt (which might access RTC
ports) occurs between writes to lower and upper parts of the
base/limit registers.

The patch temporarily disables the IO window of the bridge by
setting PCI_IO_BASE_UPPER16 > PCI_IO_LIMIT_UPPER16 before doing
an update. It's safe, as we don't have any active IO behind
the bridge at this point. Also, it's a NOP for bridges with
16-bit-only IO.
Similar (but simpler, as we always clear upper 32 bits) fix
for 64-bit prefetchable MMIO range.

into home.osdl.org:/home/torvalds/v2.5/linux

__invalidate_metapages references mp->page after after releasing the
meta_lock spinlock, without increasing the use count.  This is racy and
unnecessary since setting the META_discard flag is sufficient.
block_invalidatepage() will be called when the metapage is released.

into home.osdl.org:/home/torvalds/v2.5/linux

into flint.arm.linux.org.uk:/usr/src/bk/linux-2.6-serial

into flint.arm.linux.org.uk:/usr/src/bk/linux-2.6-serial

into home.osdl.org:/home/torvalds/v2.5/linux

(x >> 32) is undefined on a 32 bit integral variable in C; In contrast
(x >>16 >> 16) is fine (and gets optimized out to 0, while (x >> 32)
gets optimized out to a nop). 

Fix for starfire below

This fix got missed in the bonding patchset applied a while ago.

Ensure that arm ether drivers print the correct ether device name rather
than "eth%d".

Fixes build on some platforms.

From: "M.H.VanLeeuwen" <vanl@megsinet.net>

The level of isapnp_init was moved to after apci sometime ago.  Since it is
now after net_dev_init, ISA PNP NICs fail to initialized at boot.  This is
particularily problematic for NFS root filesystems like mine, or none
modular systems.

This fix allows ISA PNP NIC cards to work during net_dev_init, and still
leaves isapnp_init after apci_init.

This driver is freeing skb's from timer context, with local irq's disabled.

It generates warnings from local_bh_enable() because local_bh_enable()
reenables interrupts, exposing the machine to deadlocks.

So use the deferred dev_kfree_skb_irq() instead.

into gmx.de:/suse/kernel/BK/linux-2.5

into home.osdl.org:/home/torvalds/v2.5/linux

From: Klaas de Waal <klaas.de.waal@hccnet.nl>

Bug in parameter of ZERO_PAGE macro in line 679 of fb/direct-io.c Parameter
dio->cur_user_address has to be dio->curr_user_address.  This bug shows
when compling for MIPS little endian as target, not when compiling for X86.

From: Matthias Andree <matthias.andree@gmx.de>

Properly terminate /proc/tty/driver/serial output lines of known UARTS
when the caller has no CAP_SYS_ADMIN capability.

From: Andi Kleen <ak@muc.de>

Export two symbols on x86-64.  This is needed for the sk98lin driver and ipv6.

Use unsigned long for time_after(), not an int.

From: "Brad House" <brad_mssw@gentoo.org>

Quick patch to enable the Philips tuner on the WinTV-D boards.  Tested and
works fine.  (acked by Gerd)

The wait_event() in there can touch the memory at *transaction after
kjournald has freed it.

Rework the code to not wait until the transaction enters T_FLUSH state: just
loop back and try against after the wakeup.

Use the correct type for the workqueue callback.

Patch to get ibmtr_cs / ibmtr working again.  A change went in a while back
I missed that killed it.  Also fixed the timer to eliminate the
uninitialized timer error on close.

http://bugzilla.kernel.org/show_bug.cgi?id=1341

into gmx.de:/suse/kernel/BK/linux-2.5

http://bugme.osdl.org/show_bug.cgi?id=1171

into home.osdl.org:/home/torvalds/v2.5/linux

into home.osdl.org:/home/torvalds/v2.5/linux

into home.osdl.org:/home/torvalds/v2.5/linux

Patch from SAN People

Patch 2.4.21-rmk1 added a "iomap_base" field to the serial_struct
structure (include/linux/serial.h).

Since that structure is exported to user-space it should be
consistent between revisions of the stable 2.4 kernels.

This patch removes 4 bytes (were "reserved") to restore the size
of the structure.

Without this patch, ioctl(TIOCGSERIAL) will copy_to_user() 4
bytes more than expected and possibly corrupt the application's
stack/heap.