into samba.org:/home/paulus/kernel/for-linus-ppc

into samba.org:/home/paulus/kernel/for-linus-ppc

I looked a bit more at the keyboard code and find a bug
and a probable bug.

(i) In case a synaptics touchpad has been detected, the comment says
    "disable AUX".  But we do not set the disable bit, but instead .and.
    with the bit - no doubt getting zero.  This must be a bug. 

(ii) Directly above this is the suspicious comment "keyboard translation
     seems to be always off".  But every machine comes always up in
     translated scancode 2.  Translation is never off.  But wait! by
     mistake the above .and.  cleared the XLATE bit.

So, I think bug (i) explains mystery (ii).

However, note that this is code reading only.  I do not have the
hardware, so cannot test.

include linux/version.h in files that use its defines;

rename make check* targets to make *check (per Sam) since 'make
checkconfig' currently doesn't work; add versioncheck and
scripts/checkversion.pl;

This patch catches insures proper return values from
copy_to/from_user calls.

From: Ron Gage <ron@rongage.org>

Ports previous qla1280.c patch to 2.6.0-test2.  Fixes incorrect if
statement.  Inserts pci_free_consistent on error.

From: Leann Ogasawara <ogasawara@osdl.org>

From: Domen Puncer <domen@coderock.org>

This patch catches insures proper return values from
copy_to/from_user calls.

From: Ron Gage <ron@rongage.org>

From: Domen Puncer <domen@coderock.org>
	and Randy Dunlap

Release (in fact all) functions shouldn't be __init or __exit, if they
are called from __init and __exit functions.

This patch catches insures proper return values from
copy_to/from_user calls.

From: Ron Gage <ron@rongage.org>

This patch catches insures proper return values from
copy_to/from_user calls.

From: Ron Gage <ron@rongage.org>

Test compiled on x86.

From: Randy Hron <rwhron@earthlink.net

From: Leann Ogasawara <ogasawara@osdl.org>

Patch to insert ioremap() error checking and missing iounmap()'s.  Also
updated error handling to have only one return statement on error.
Finally, in saa7134_finidev() on line 954 there was a duplicate
saa7134_i2c_unregister().

This patch catches insures proper return values from
copy_to/from_user calls.

From: Ron Gage <ron@rongage.org>

blk_init_queue() has changed (it now returns queue), but the
comments were the same.

Also fixed a comment in blk_queue_end_tag.

From: Domen Puncer <domen@coderock.org>

From: Daniele Bellucci <bellucda@tiscali.it>

From: Daniele <bellucda@tiscali.it>

OK-ed by Jens.

This patch catches insures proper return values from
copy_to/from_user calls.

From: Ron Gage <ron@rongage.org>

From: Randy Hron <rwhron@earthlink.net>

Put version.h in the file that needs it.
Remove version.h where it isn't needed.
Tested as module/non-module on x86.

From: Randy Hron <rwhron@earthlink.net>

Three small 1-line patches to get rid of any remaining __SMP__
references in the 2.6 kernel. 

From: Ron Gage  <ron@rongage.org>

From: Daniele Bellucci <bellucda@tiscali.it>

This patch audits some copy_from/to_user in the SysKonnect FDDI device
driver.

This patch catches insures proper return values from
copy_to/from_user calls.

From: Ron Gage <ron@rongage.org>

From: Daniele Bellucci <bellucda@tiscali.it>

Insert ioremap error check and adjust cleanup path accordingly.

From: Leann Ogasawara <ogasawara@osdl.org>

From: Daniele Bellucci <bellucda@tiscali.it>

add 'static' to a function

From: Randy Hron <rwhron@earthlink.net>

remove unneeded linux/version.h usage & some duplicate
#includes;

Dmasound: Use try_module_get()/module_put() instead of methods calling
MOD_{IN,DE}C_USE_COUNT (from Christoph Hellwig)

Dmasound core fixes from Christoph Hellwig:
  - Some exported symbols are declared __init - in the modular case this is
    freed before the other modules can call it..
  - dmasound.lock is initialized too late, do it at compile time

Hugh Dickins caught this confusion between the ret variable in outer
scope (holding the fd) and the return value of f_setown.  Rename inner
"ret".

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

move all watchdog documentation into Documentation/watchdog/

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

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

This also includes a change so that if an instruction is being single-stepped
and it turns out to need to be emulated (e.g. because of an alignment
exception, or because we are emulating floating-point instructions), then
we need to deliver a SIGTRAP after successfully emulating the instruction
and advancing NIP.

The main thing here is that the signal delivery/return code for real-time
signals has been changed so that the layout of the registers corresponds
with the new ucontext_t definition being used by glibc.  The old ucontext_t
didn't have space to actually store the registers, just a pointer to them,
which made it impossible to implement set/getcontext et al.  The new
ucontext_t includes a mcontext_t which actually contains space to store
all of the general, floating pointer and vector registers.

We now also save the altivec registers on signal delivery and restore them
on return from the signal if the process has ever used altivec (since the
last exec).

Finally this adds a swapcontext system call.  Swapcontext really needs to be
done in the kernel since on PPC, only privileged code can set all three of
CTR, LR and NIA (next instruction address) to arbitrary values.  Also the kernel
know if the process currently owns the FP and altivec units and can optimize
in the case where it doesn't.