rmk's patch found a defined-but-unimplemented symbol
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

It fixes a possible race between quotaoff and prune_icache.  The race could
lead to some forgotten pointers to quotas in inodes leading later to BUG
when invalidating quota structures.
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

Based on work from Bill Irwin <wli@holomorphy.com>

physnode_map[] needs to be signed so that pfn_to_nid() can return negative
values used to detect invalid pfn's.
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

This patch enables the Redwood 5 and Redwood 6 platforms to use the smc91x
ethernet driver.
Signed-off-by: Dale Farnsworth <dale@farnsworth.org>
Signed-off-by: Matt Porter <mporter@kernel.crashing.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

We have a platform-specific function pointer on ppc64 for a function to log
errors detected by the platform, but it was never getting set.  This patch
sets it on pSeries (the only ppc64 platform which has an error logging
function).
Signed-off-by: Paul Mackerras <paulus@samba.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

From: Linas Vepstas <linas@austin.ibm.com>

This patch eliminates the usage of the deprecated ibm,fw-phb-id token for
idnetifying PCI bus heads in favor of the documented, offically supported
mechanism for obtaining this info.  Please note that some versions of
firmware may return incorrect values for the ibm,fw-phb-id token.
Signed-off-by: Linas Vepstas <linas@linas.org>
Signed-off-by: Paul Mackerras <paulus@samba.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

gcc-2.95 chokes on this.

Cc: Anton Altaparmakov <aia21@cantab.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

SuSE discovered this problem with chown and ATTR_GID.  Make sure user
is authorized to change the group, CAN-2004-0497.

It's one of the rare files that has almost no includes
what-so-ever, so it actually never got NULL any other
way.

Two years ago, OGAWA Hirofumi removed some ugly code and
added a few simple tests to the FAT filesystem code,
intended to avoid recognizing non-FAT as FAT (for people who
fail to specify rootfstype=, forcing the kernel to guess).

That worked fairly well, until this year.

I have now seen a thread in Czech and a report from Holland that
involved the "FAT: bogus sectors-per-track value" error message.

The patch below removes this test again. The advantage is that
some real-life FAT filesystems can be mounted again.

The disadvantage that more non-FAT fss will be accepted as FAT.

Ferry van Steen <freaky@bananateam.nl> reports "the patch Andries
Brouwer gave me seems to work".
Signed-off-by: Andries Brouwer <aeb@cwi.nl>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

This fixes the the remaining 0 to NULL things that were found with 'make
allmodconfig' and 'make C=1 vmlinux'.

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

into cantab.net:/home/src/ntfs-2.6

into cantab.net:/home/src/ntfs-2.6

      enables the writing of page cache pages belonging to mst protected
      attributes like the index allocation attribute in directory indices
      and other indices like $Quota/$Q, etc.  This means that the quota is
      now marked out of date on all volumes rather than only on ones where
      the quota defaults entry is in the index root attribute of the
      $Quota/$Q index.
Signed-off-by: Anton Altaparmakov <aia21@cantab.net>

      page has buffers.  Otherwise we could end up with a dirty page
      without buffers and our set_page_dirty() would not mark the buffers
      dirty when they are created and thus they would not be written out
      and the dirty records would be lost.
Signed-off-by: Anton Altaparmakov <aia21@cantab.net>

Any proc entry with default proc_file_inode_operations allow unauthorized
attribute updates.  This is very dangerous for proc entries that rely
solely on file permissions for open/read/write.
Signed-off-by: Chris Wright <chrisw@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

      buffers that are inside the ntfs record in the page dirty after which
      it sets the page dirty.  This allows ->writepage to only write the
      dirty index records rather than having to write all the records in
      the page.  Modify fs/ntfs/index.h::ntfs_index_entry_mark_dirty() to
      use this rather than __set_page_dirty_nobuffers().
Signed-off-by: Anton Altaparmakov <aia21@cantab.net>

      It is simply set to __set_page_dirty_nobuffers() to make sure that
      running set_page_dirty() on a page containing mft/ntfs records will
      not affect the dirty state of the page buffers.
Signed-off-by: Anton Altaparmakov <aia21@cantab.net>

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

into kernel.bkbits.net:/home/davem/tg3-2.6

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

into kernel.bkbits.net:/home/davem/sparc-2.6

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

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

It can return with the lock held.

Found by the Stanford locking checker.
Signed-off-by: Andrew Morton <akpm@osdl.org>

into pobox.com:/spare/repo/net-drivers-2.6

Included is a patch for linux to add a PCI vendor/device ID for the
Radisys ENP-2611 board.  The ENP-2611 is a 64bit/66MHz PCI board which
hosts an Intel IXP2400 network processor, has three GigE interfaces,
runs linux and generally kicks ass.

(see http://www.radisys.com/oem_products/ds-page.cfm?productdatasheetsid=1147)
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>

Please review and apply the following patch if you find it agreeable.

This patch does not make any functional changes, but does improve
both performance and memory usage by rearranging structure elements.

The need for these changes became appearent during a code review of
the disassembly involving this structure. The memory footprint of this
structure is made smaller by grouping the byte fields next to each other.
The access of the list_head can be simplified by making it the first element
of the structure, thus avoiding a needless add-immediate without negatively
impacting any of the other accesses.
Signed-off-by: Linas Vepstas <linas@linas.org>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>

This patch fixes a null-pointer dereference when hot-plug operations
are performed on a machine that has virtual-io devices in it.
Virtual i/o devices to not have pci bridges associated with them.
It also corrects an ordering problem during hotplug remove.

This patch was previously reviewed/tested by Linda Xie, the current
rpaphp maintainer.
Signed-off-by: Linas Vepstas <linas@linas.org>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>

This patch adds some PCI Express register constants to <linux/pci.h>

For my device, setting the Max_Read_Request_Size value in the PCI
Express device control register makes a huge performance difference.
I wanted my driver code that does this to be a little more
self-documenting than:

	pci_read_config_word(mdev->pdev, cap + 8, &val);
	val = (val & ~(5 << 12)) | (5 << 12);

I went a little overboard and added all the basic device register
fields.  If desired I could go even further overboard and add the
link, slot and root registers as well.

This patch is based on Matthew Wilcox's patch for pciutils, corrected
for some PCI Express spec 1.0a changes.
Signed-off-by: Roland Dreier <roland@topspin.com>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>

Signed-off-by: Linda Xie lxie@us.ibm.com
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>

 - fix possible prefetch-fault loop on NX page, based on suggestions
   from Jamie Lokier.

 - clean up nx feature dependencies

 - simplify detection of NX-violations when the kernel executes code

 - introduce pte_exec_kern() to simplify the NX logic

 - split the definitions out of pgtable-[23]level.h into
   pgtable-[23]level-defs.h, to enable the former to use generic
   pte functions from pgtable.h.
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

This patch, originally from Daniel Drake, replaces the meye driver
'wait_ms()' function with calls to the kernel provided 'msleep()'
function.
Signed-off-by: Daniel Drake <dsd@gentoo.org>
Signed-off-by: Stelian Pop <stelian@popies.net>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

This updates the sonypi driver by:
 * fixing the power management handling, using the new device
   model PM scheme.

 * adds "VGN-" as a DMI search pattern for a Sony Vaio laptop.

Florian Lohoff reported the power management issue and tested the
patch.

Many users reported the DMI name issue, including Till Busch who
made a patch for dmi_scan.c.
Signed-off-by: Stelian Pop <stelian@popies.net>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

From the new Stanford locking checker
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

From the new Stanford locking checker
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

Attached is a smallish patch for couple trivial sparse warnings in
allnoconfig build and more importantly an "excuses" text file explaining
why the rest have not been fixed.

Basically all of them (with the exception of the one in Andrews tree) need
some serious re-engineering.
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

The attached patch fixes the pushf under v86 to always set the IOPL field
to 3, as the Intel CPUs do.

It was in 2.4 for year, but somehow missed 2.6.  It comes from here
(although now a bit shorter):

http://lkml.org/lkml/2003/5/25/81Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>