into penguin.transmeta.com:/home/penguin/torvalds/repositories/kernel/linux

into tetrachloride.(none):/mnt/stuff/kernel/2.5/watchdog

into penguin.transmeta.com:/home/penguin/torvalds/repositories/kernel/linux

into penguin.transmeta.com:/home/penguin/torvalds/repositories/kernel/linux

patch from Bill Irwin

has_stopped_jobs() is completely unused. This patch removes
has_stopped_jobs() and renames __has_stopped_jobs() to has_stopped_jobs().

proc_fill_super() simply wants a count of processes, not threads.
This creates a per-cpu counter for it to use to determine that.

Patch from Bill Irwin.

__do_SAK() simply wants to kill off processes using a given tty.  This
converts it to use for_each_task_pid().

From Rohit Seth.

We're currently disabling read() and write() against hugetlbfs files via the
address_space ops.  But that's a bit awkward, and results in reads and writes
instantiating useless, non-uptodate 4k pagecache against the inodes.

The patch removes the read, write and sendfile file_operations entries.  So
the caller will get their -EINVAL without us ever having to go to the
pagecache layer.

This is the leak which Con found.  Long story...

- If a dirty page is fed into ext3_writepage() during truncate,
  block_write_full_page() will reutrn -EIO (it's outside i_size) and will
  leave the buffers dirty.  In the expectation that discard_buffer() will
  clean them.

- ext3_writepage() then adds the still-dirty buffers to the journal's
  "async data list".  These are buffers which are known to have had IO
  started.  All we need to do is to wait on them in commit.

- meanwhile, truncate will chop the pages off the address_space.  But
  truncate cannot invalidate the buffers (in journal_unmap_buffer()) because
  the buffers are attached to the committing transaction.  (hm.  This
  behaviour in journal_unmap_buffer() is bogus.  We just never need to write
  these buffers.)

- ext3 commit will "wait on writeout" of these writepage buffers (even
  though it was never started) and will then release them from the
  journalling system.

So we end up with pages which are attached to no mapping, which are clean and
which have dirty buffers.  These are unreclaimable.


Aside:

  ext3-ordered has two buffer lists: the "sync data list" and the "async
  data list".

  The sync list consists of probably-dirty buffers which were dirtied in
  commit_write().  Transaction commit must write all thee out and wait on
  them.

  The async list supposedly consists of clean buffers which were attached
  to the journal in ->writepage.  These have had IO started (by writepage) so
  commit merely needs to wait on them.

  This is all designed for the 2.4 VM really.  In 2.5, tons of writeback
  goes via writepage (instead of the buffer lru) and these buffers end up
  madly hpooing between the async and sync lists.

  Plus it's arguably incorrect to just wait on the writes in commit - if
  the buffers were set dirty again (say, by zap_pte_range()) then perhaps we
  should write them again before committing.


So what the patch does is to remove the async list.  All ordered-data buffers
are now attached to the single "sync data list".  So when we come to commit,
those buffers which are dirty will have IO started and all buffers are waited
upon.

This means that the dirty buffers against a clean page which came about from
block_write_full_page()'s -EIO will be written to disk in commit - this
cleans them, and the page is now reclaimable.  No leak.

It seems bogus to write these buffers in commit, and indeed it is.  But ext3
will not allow those blocks to be reused until the commit has ended so there
is no corruption risk.  And the amount of data involved is low - it only
comes about as a race between truncate and writepage().

max_sane_readahead() permits the user to readahead up to
half-the-inactive-list's worth of pages.  Which is totally wrong if most of
memory is free.

So make the limit be

	(nr_inactive + nr_free) / 2

Patch from Andreas Dilger <adilger@clusterfs.com>

This patch against 2.5.53 removes my erronous use of ino_t in a couple of
places in the ext3 code.  This has been replaced with unsigned long (the same
as is used for inode->i_ino).  This patch matches the fix submitted to 2.4
for fixing 64-bit compiler warnings, and also replaces a couple of %ld with
%lu to forestall output wierdness with filesystems with a few billion inodes.

- Merged SHA-384 and SHA-512 code from Kyle McMartin
  <kyle@gondolin.debian.net>
- Added test vectors.
- Documentation and credits updates.

into nuts.ninka.net:/home/davem/src/BK/sparc-2.5

definition of BUG().

In the new (more logical) world order, that should be <asm/bug.h>

This patch is the first in a series that groups the common
m68knommu entry.S code into a common entry.S file (this is
instead of having several similar files in lower sub-architecture
directories). The whole idea is to reduce code duplication.

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

into raven.il.steeleye.com:/home/jejb/BK/dma-alloc-flags-2.5

This prevents the linker from merging the .gnu.linkonce.this_module section
into the .text section, which is necessary for modules to load correctly.

Fixes compile but card doesn't get detected, perhaps the id list?

Clustered APIC setup patch.  Needed to support generic systems with more
than 8 CPUs.

Motivation:

 The current APIC destination mode ("Flat Logical") used in linux kernel
 has an upper limit of 8 CPUs.  For more than 8 CPUs, either "Clustered
 Logical" or "Physical" mode has to be used.

 The attached patch adds support such systems by organizing them into
 logical clusters, with each cluster having 4 CPUs.  This is activated by
 a new config option "Support for other sub-arch SMP systems with more
 than 8 CPUs", under Processor feature->Sub architecture.

 The patch is made very simple and isolated, thanks to Martin J.
 Bligh's patchsets, which has moved all APIC related functions into
 sub-arch macros.  Has zero impact on standard systems.

 This patch enables all 16 logical processors on a generic, non-quad
 based, system that we have here.  Also, by looking at SuSE source, I
 have also added a special switch, to specifically support Unisys
 (ES7000).  Just replacing #define SEQUENTIAL_APICID by CLUSTERED_APICID
 in the patch should make it work on ES7000(not tested).

This pulls all of the m68knommu arch linker scripts into a single new
style linker script.  This patch specifically turns
arch/m68knommu/vmlinux.lds.S into a generic linker script that works
with all m68knommu targets.