Both sys_sync() and the kupdate function need to precalculate the number of
pages which they are prepared to write.  Mainly for livelock avoidance.

But they also must write inodes, and dirty inodes do not contribute to dirty
page accounting (oops).  Net effect: when there are lots of dirty inodes and
few dirty pages, we forget to write inodes.

This mainly affects atime updates, because most other inode-dirtying activity
will generate dirty pages too.

It mainly affects ext2.


Now, writing an ext2 inode will just dirty the underlying blockdev pagecache
page.  So what the patch does is to assume that writing one inode will dirty
up to one pagecache page.  So the patch adds (inodes_stat.nr_inodes -
inodes_stat.nr_unused) into the number of pages to be written.

I considered creating inodes_stat.nr_dirty.  It looks fairly messy, needing
to know not to account for memory-backed inodes, etc.  But it is probably a
better thing to do.

By moving the special-casing for swapper_space out of __mark_inode_dirty()
and into __set_page_dirty_nobuffers() we can remove swapper_inode.

From: Samuel Thibault <Samuel.Thibault@ens-lyon.fr>

Changes the new console scrolling ioctl to permit distances greater than
+127/-128.

- shmem: remove unneeded test for null inode->i_sb (James Morris)

- kill unused var warning in traps.c (Geert Uytterhoeven)

- s/u64/__u64/ in bitops.h (needed for klibc)

- comment fix in gfp.h (Matthew Dobson <colpatch@us.ibm.com>)

- fix smbfs constant overflow warning (Flameeyes <daps_mls@libero.it>)

- yam.c irqreturn_t fix.

- Remove some unused variables from baycom_epp.c (Adrian Bunk)

- Remove 5-year-old unreferenced RCS string from xirc2ps_cs.c (Adrian Bunk)

update hugetlbfs_statfs for the statfs64() changes.

From: "Seth, Rohit" <rohit.seth@intel.com>

- Add support for setting the filesystem's maximum size and maximum inode
  count on the mount command line.

  This is needed because the system admin can now set the ownership of teh
  fs to non-root users.  We don't want those users to be able to use all of
  the hugepage pool.

- Prroperly update the inode creation/modification time.

- Set the blocksize to HPAGE_SIZE (instead of PAGE_CACHE_SIZE).

- Update Documentation/vm/hugetlbpage.txt.

From: Andrey Borzenkov <arvidjaar@mail.ru>

Add devfs support to raw.c.

From: Manfred Spraul <manfred@colorfullife.com>

The CLONE_SYSVSEM implementation is racy: it does an (atomic_read(->refcnt)
==1) instead of atomic_dec_and_test calls in the exit handling.  The patch
fixes that.

Additionally, the patch contains the following changes:

- lock_undo() locks the list of undo structures.  The lock is held
  throughout the semop() syscall, but that's unnecessary - we can drop it
  immediately after the lookup.

- undo structures are only allocated when necessary.  The need for undo
  structures is only noticed in the middle of the semop operation, while
  holding the semaphore array spinlock.  The result is a convoluted
  unlock&revalidate implementation.  I've reordered the code, and now the
  undo allocation can happen before acquiring the semaphore array spinlock.
   As a bonus, less code runs under the semaphore array spinlock.

- sysvsem.sleep_list looks like code to handle oopses: if an oops kills a
  thread that sleeps in sys_timedsemop(), then sem_exit tries to recover.
  I've removed that - too fragile.

From: David Mosberger <davidm@napali.hpl.hp.com>

This is an attempt at sanitizing the interface for stack trace dumping
somewhat.  It's basically the last thing which prevents 2.5.x from working
out-of-the-box for ia64.  ia64 apparently cannot reasonably implement the
show_stack interface declared in sched.h.

Here is the rationale: modern calling conventions don't maintain a frame
pointer and it's not possible to get a reliable stack trace with only a stack
pointer as the starting point.  You really need more machine state to start
with.  For a while, I thought the solution is to pass a task pointer to
show_stack(), but it turns out that this would negatively impact x86 because
it's sometimes useful to show only portions of a stack trace (e.g., starting
from the point at which a trap occurred).  Thus, this patch _adds_ the task
pointer instead:

 extern void show_stack(struct task_struct *tsk, unsigned long *sp);

The idea here is that show_stack(tsk, sp) will show the backtrace of task
"tsk", starting from the stack frame that "sp" is pointing to.  If tsk is
NULL, the trace will be for the current task.  If "sp" is NULL, all stack
frames of the task are shown.  If both are NULL, you'll get the full trace of
the current task.

I _think_ this should make everyone happy.

The patch also removes the declaration of show_trace() in linux/sched.h (it
never was a generic function; some platforms, in particular x86, may want to
update accordingly).

Finally, the patch replaces the one call to show_trace_task() with the
equivalent call show_stack(task, NULL).

The patch below is for Alpha and i386, since I can (compile-)test those (I'll
provide the ia64 update through my regular updates).  The other arches will
break visibly and updating the code should be trivial:

- add a task pointer argument to show_stack() and pass NULL as the first
  argument where needed

- remove show_trace_task()

- declare show_trace() in a platform-specific header file if you really
  want to keep it around

From: John Stultz, George Anzinger, Eric Piel

There was confusion over the definition of TICK_USEC.  TICK_USEC is
supposed to be based on USER_HZ, however a recent change caused TICK_USEC
to be based on HZ.  This broke the adjtimex() interface on systems where
USER_HZ != HZ.  This patch reverts the change to TICK_USEC, removes an
added mis-use of the value and fixes some incorrect comments that could
lead to this sort of confusion.

Also this patch resolves the related LTP adjtimex failures.

From: David Mosberger <davidm@napali.hpl.hp.com>

At the moment, if you don't set CLONE_CHILD_SETTID/CLONE_CHILD_CLEARTID,
the {set,clear}_child_tid values get inherited from the parent task.  I may
be missing something, but I suspect that's not the intended behavior.  The
patch below instead clears the respective members.

The slab debugging code is supposed to poison freshly-allocated obejcts with
0x5a and freed ones with 0x6b, so we can distinguish use-uninitialised from
use-after-free.

It wasn't working right for recycled objects.  Fix.

Use slab_error for printing the error message from kmem_cache_destroy

From: Peter Chubb <peter@chubb.wattle.id.au>

Add two new system calls, statfs64 and fstatfs64.  This has been needed
sincew the 64-bit sector_t merge - the current structures will overflow.

- Use a common interface (vfs_statfs) with the rest of the kernel,

- convert to 32-bit at (f)statfs time.

- New field f_frsize gives underlying fragment size for the filesystem.
  (Solaris has this, and the Open Group describe it).

- The old statfs syscalls will now return -EOVERFLOW if the device was
  too large to be represented inthe old data structures.

The new system calls take a size_t argument, which is the size of the
structure to be filled in (as requested by Ben LaHaise), to `futureproof' the
interface.

Has been reviewed by the arch maintainers and by Ulrich Drepper.

From: Andi Kleen <ak@suse.de>

On i386 and most other ports kern_addr_valid is hardcoded to 1.

This works fine as long as only mapped areas are accessed.  When you have
something partially mapped in the kclist it is possible that start points to
an unmapped address.  The correct behaviour in this case is to zero the user
space.  We shouldn't return -EFAULT because the fault is against the mmapped
range, not against the user's address.

copy_to_user usually even checks for exceptions on both source and
destination, but it does not zero the destination in this case and worse
results in EFAULT, which is user visible.

This patch just tries to clear_user in this case again to actually zero the
user data and catch real user side EFAULTs.

Another way to fix this is to have kern_addr_valid do a real page table
lookup (I did that on AMD64), but having this fallback is a bit more
reliable in case there is a race somewhere.

On i386 it could happen for example if the direct space to max_low_pfn
contains something unmapped.  This normally isn't the case, but e.g.  the
slab debugging patches in -mm* do this so it's better to handle it.

Drawback is that it relies on a somewhat undocumented copy_to_user behaviour
(fault on both source and destination).  It is true for i386 and amd64, but I
don't know if it is for other port.  In the worst case they just don't have
the race protection and may see bogus EFAULTs.

From: Andi Kleen <ak@suse.de>

The recent IA64 changes for /proc/kcore broke the access on i386.
Currently no notes are written for the direct mapped or vmalloced memory,
which makes gdb reject it.

This patch fixes it.  Other ports probably need to do the same changes.

If you open /dev/ram7 when the kernel is configured for 4 ramdisks, things
blow up.  Teach rd_open() to check that the minor is in range.

If someome tries to unmount the fs while kjournald is performing a commit,
kjournald forgets to look for the termination request and goes into permanent
sleep.

The memset which is performed if access_ok() fails got lost in the
copy_*_user() rework.  Put it back.

Bloats the kernel by 8k :(

Also contains a few related #includes and whitespace fixlets from
Joshua Kwan <joshk@triplehelix.org>

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

This is an update for the SiS IDE driver.

This is a 99% Vojtech work :
- Independant southbridge detection (no need to add current and future
  MuTIOL northbridge PCI ids knowledge to the driver),
- Lots of code cleanup,
- Debug code removed (unused for a while, I will maintain it in my tree
  if needed),

I changed some things:
- the new config_xfer_rate is commented out until ide_find_best_mode is
  patched for bad drive handling (until then I reverted to the old one
  using the config_drive_xfer_rate helper function).

Hi Greg.  Ivan's not happy with the solution I came up with for naming
/proc/bus/pci and Anton would prefer something slightly different too,
so I abstracted the name out so each architecture can do its own thing.

This is against 2.5.72 so won't apply cleanly to your tree (it
applies to bitkeeper as of a few minutes ago with only minor offsets).
I've implemented the original name for non-PCI-domain machines; done what
ia64 and alpha need, respectively (assuming I didn't misunderstand Ivan),
and plopped in the Old Way of doing things for Sparc64, PPC and PPC64.
Maintainers may alter this to whatever degree of complexity they wish.

into kroah.com:/home/greg/linux/BK/pci-2.5

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

While there is a separate driver for 2104x tulips (CONFIG_DE2104X),
drivers/net/tulip/Kconfig states that CONFIG_TULIP also supports
2104x tulips. This is not the case since that support was removed
in December 2001. A user with an old tulip may thus be tricked into
configuring the wrong driver. (I was, on my PMac 4400.)

The patch below removes this misinformation from tulip's Kconfig.

(pulled change from 2.4)

the second patch:
replaces busy_loop with a simple macro doing a schedule_timeout. busy_loop was never
called from interrupt conext anyway, so no need for that. and the sti() is gone.

rgds
-daniel

hi

this patch does:
- net_device is no longer allocated as part of the driver's private structure,
  instead it's allocated via alloc_netdev
- xirc2ps_detach calls xirc2ps_release if necessary (like the other drivers)

against 2.5.70-bk.

rgds
-daniel

drivers/net/pcnet32.c: In function `pcnet32_init_ring':
drivers/net/pcnet32.c:1006: warning: comparison between pointer and integer

According to a report the my_inet_aton code in bond_main.c is copied
from 4.4BSD, but it doesn't carry a BSD copyright license. In addition
it is somewhat redundant with the standard in_aton.  Convert it
to use the linux function.

Error handling is a bit worse than before, but not much.

Patch for 2.5 bonding. The 2.4 version has the same problem, but afaik
it is scheduled to be replaced by the 2.5 codebase anyways.

-Andi