Add -msmall-data to CFLAGS.

* fs/ncpfs/inode.c: Return reasonable error codes instead of universal
     -EINVAL. Remove printk() as reasonable code is returned.
     Set maximum file size limit on ncpfs to 4GB-1.

* fs/ncpfs/sock.c: Return correct error code when send() fails.

	Petr Vandrovec

* zisofs_cleanup cannot be __exit, as it is invoked from __init
  section when register_filesystem() fails.

		Petr Vandrovec

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

having the thread_info struct at the base of the stack and
the task_struct elsewhere.

into quango.(none):/home/paulus/kernel/linuxppc-2.5

	Linus, I've got the first of BKL-removal ext2 patches ready to
go.  It removes BKL from ext2_get_block() and guts of ext2_truncate().
The only place where we hold BKL on these paths is in dquot.c - probably
can be easily dealt with, but threading quota is a separate story.

	Inode metadata (pointers to blocks, both in inode itself and in
indirect blocks, preallocation data and allocation goal) are protected
by rwlock - EXT2_I(inode)->i_meta_lock.

	Next steps will involve threading the group descriptors and bitmaps
handling - lock_super() uses in ext2 are going to die.  However, that's
a separate story - let's do that step-by-step.

	I suspect that patch below will take care of almost all BKL contention
from ext2 - we still have BKL held over directory operations, but for regular
files that's it.

   truncate now correctly sets mtime always. Before this fix, mtime was not
   updated if truncated file was of zero length or if new filesize was bigger
   then old.
   Problem was noticed by Matthias Andree <ma@dt.e-technik.uni-dortmund.de>

   Speed up reading of journal bitmaps. RAID users should notice significant
   speedup when mounting reiserfs over self-rebuilding RAID arays.

   Convert all the code to use reiserfs_{kmalloc,kfree}. Remove all extra
   reiserfs_{kmalloc,kfree} overhead if CONFIG_REISERFS_CHECK is not set.

   Do not panic when encountered item of unknown type, just print a warning.

   Correctly detect and print hash values, when manual hash detection is used.

   Do not panic on incorrect savelink entries (truncate on directory).
   Currently we suppose these can be created if switching between kernels
   with and without savelinks support.

   Do not print a warning if savelink was not created due to lack of space.

    Make sure all reiserfs_find_entry users correctly understand IO_ERROR retval.

What should be happening with the references to the discarded .text.exit
section?  I see a __devexit_p mentioned in Documentation/pci.txt, but it
hasn't been implemented except for down inside ieee1394.

In any case, I need something like the following in order to build with
pre-release binutils 2.12.  If this sort of thing is acceptible I can
prepare a more comprehensive patch.

Here is the last patch for today: it enables writecombined mappings
for ia64 in fbmem.c and gets rid of an ugly ia64 simulator workaround
in vgacon.c which isn't needed anymore.

	--david

This patch fixes a deadlock condition in the elf core dump that shows
on ia64 because ELF_CORE_COPY_REGS() needs to access user space (to
get a hold of the backing store of the stacked registers).  Marcelo
already accepted this into 2.4.17.

	--david

This patch fixes up two places whre dma64_addr_t is used incorrectly.
Note that pci_dev->dma_mask and the second argument to
blk_queue_bounce_limit() are both u64, so the old types clearly are
wrong (besides, dma64_addr_t is supposed to be used only with the
pci_dac_*() routines, as per DaveM's earlier mail).

	--david

This patch to drivers/char/pcwd.c against 2.5.4 does two things:
a) Makes one code snippet more consistent with the rest of the code, and
b) Makes it possible for this code to actually work

Nearly the same patch against 2.4 was reviewed by Alan, and, well, the
maintainer seems to have disappeared.  It's also looking like no one uses
this driver much either.

Regards,
Rob Radez

Here is the latest version of the VM_DATA_DEFAULT_FLAGS patch
(relative to 2.5.4).

	--david

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

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

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

from when completion was potentially called more than once to indicate
partial end I/O. These days bio->bi_end_io is _only_ called when I/O
has completed on the entire bio.

	Rediffed to 2.5.4, documentation added.  This variant grabs
->s_vfs_rename_sem only for cross-directory renames.

I previously sent a patch by Mikael Pettersson to fix the UP+preempt
problem.  It seems from your BK repository you have not yet merged it;
if so, this patch takes a different approach which is optimal, removing
the unneeded conditional altogether in the UP case.  I have verified UP
and SMP are now correct.  Patch is against 2.5.4, please apply.

	Robert Love

Here is the patch against 2.5.4.  I have compiled this patch under
2.5.3, so it should still be OK.

This patch just resyncs the driver with 2.4.18-pre (which is what is
being testd by others).  The only outstanding known problem is some
very strange interaction with VMWARE.  But otherwise people seem
happy with the changes.

Original announcement to Dave Jones and Marcelo:

	Update a couple of email addresses
	Fix the idle handling (this is an improved version of the fix
		that Alan Cox has in his -ac tree)
	Notify user mode of suspend events before drivers (fix)
	Make the idling percentage boot time configurable
	Rename kapm-idled to kapmd

Credit to Andreas Steinmetz, Russell King, Thomas Hood and me.

More small updates to come.
--
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au
http://www.canb.auug.org.au/~sfr/

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

-- steve.cameron@compaq.com

Applies to 2.5.4-pre3                                                           
-- steve.cameron@compaq.com