More from Dominik

From Dominik.

Wait with the destruction of cpufreq objects until all references
are gone (code partly based on rmk's comparable patch for the PCMCIA
subsystem.

From Dominik.

into tetrachloride.(none):/mnt/raid/src/kernel/2.5/cpufreq

into tetrachloride.(none):/mnt/raid/src/kernel/2.5/cpufreq

From Dominik Brodowski

Spotted by Luigi Belli

I though Scott had recently merged this but it seems not.  We'll be
needing this patch if you merge Manfred's page unmapping debug patch.

From: Jens Axboe <axboe@suse.de>

It fixes a hang when performing large I/O's.  Has been tested and acked by
the maintainer, "Wiran, Francis" <francis.wiran@hp.com>.

From: bert hubert <ahu@ds9a.nl>

Attached patch adds a range check to LOG_BUF_SHIFT and clarifies the
configuration somewhat.  I managed to build a non-booting kernel because I
thought 64 was a nice power of two, which lead to the kernel blocking when
it tried to actually use or allocate a 2^64 buffer.

		CPU0				CPU1

	journal_get_write_access(bh)
	 (Add buffer to t_reserved_list)

					journal_get_write_access(bh)
					 (It's already on t_reserved_list:
					  nothing to do)

	 (We decide we don't want to
	  journal the buffer after all)
	journal_release_buffer()
	 (It gets pulled off the transaction)


					journal_dirty_metadata()
					 (The buffer isn't on the reserved
					  list!  The kernel explodes)


Simple fix: just leave the buffer on t_reserved_list in
journal_release_buffer().  If nobody ends up claiming the buffer then it will
get thrown away at start of transaction commit.

If load_elf_binary() (and the other binary handlers) fail after
flush_old_exec() (for example, in setup_arg_pages()) then do_execve() will go
through and do mmdrop(bprm.mm).

But bprm.mm is now current->mm.  We've just freed the current process's mm.
The kernel dies in a most ghastly manner.

Fix that up by nulling out bprm.mm in flush_old_exec(), at the point where we
consumed the mm.  Handle the null pointer in the do_execve() error path.

Also: don't open-code free_arg_pages() in do_execve(): call it instead.

ext2's inode allocator will call find_group_orlov(), which will return a
suitable blockgroup in which the inode should be allocated.  But by the time
we actually try to allocate an inode in the blockgroup, other CPUs could have
used them all up.

ext2 will bogusly fail with "ext2_new_inode: Free inodes count corrupted in
group NN".


To fix this we just advance onto the next blockgroup if the rare race
happens.  If we've scanned all blockgroups then return -ENOSPC.


(This is a bit inaccurate: after we've scanned all blockgroups, there may
still be available inodes due to inode freeing activity in other blockgroups.
 This cannot be fixed without fs-wide locking.  The effect is a slightly
early ENOSPC in a nearly-full filesystem).

From: Stephen Smalley <sds@epoch.ncsc.mil>

This patch against 2.5.73 replaces vm_enough_memory with a security hook
per Alan Cox's suggestion so that security modules can completely replace
the logic if desired.

Note that the patch changes the interface to follow the convention of the
other security hooks, i.e.  return 0 if ok or -errno on failure (-ENOMEM in
this case) rather than returning a boolean.  It also exports various
variables and functions required for the vm_enough_memory logic.

From: William Lee Irwin III <wli@holomorphy.com>

This patch allows architectures to micro-optimize lowmem_page_address() at
their whims.  Roman Zippel originally wrote and/or suggested this back when
dependencies on page->virtual existing were being shaken out.  That's
long-settled, so it's fine to do this now.

From: john stultz <johnstul@us.ibm.com>

This patch catches a corner case in the lost-tick compensation code.

There is a check to see if we overflowed between reads of the two time
sources, however should the high res time source be slightly slower then
what we calibrated, its possible to trigger this code when no ticks have
been lost.

This patch adds an extra check to insure we have seen more then one tick
before we check for this overflow.  This seems to resolve the remaining
"time doubling" issues that I've seen reported.

From: john stultz <johnstul@us.ibm.com>

The patch tries to resolve issues caused by running the TSC based lost
tick compensation code on CPUs that change frequency (speedstep, etc).

Should the CPU be in slow mode when calibrate_tsc() executes, the kernel
will assume we have so many cycles per tick.  Later when the cpu speeds up,
the kernel will start noting that too many cycles have past since the last
interrupt.  Since this can occasionally happen, the lost tick compensation
code then tries to fix this by incrementing jiffies.  Thus every tick we
end up incrementing jiffies many times, causing timers to expire too
quickly and time to rush ahead.

This patch detects when there has been 100 consecutive interrupts where we
had to compensate for lost ticks.  If this occurs, we spit out a warning
and fall back to using the PIT as a time source.

I've tested this on my speedstep enabled laptop with success, and others
laptop users seeing this problem have reported it works for them.  Also to
ensure we don't fall back to the slower PIT too quickly, I tested the code
on a system I have that looses ~30 ticks about every second and it can
still manage to use the TSC as a good time source.

This solves most of the "time doubling" problems seen on laptops.
Additionally this revision has been modified to use the cleanups made in
rename-timer_A1.

From: john stultz <johnstul@us.ibm.com>

This renames the bad "timer" variable to "cur_timer" and moves externs to
.h files.

From: Daniel Jacobowitz <dan@debian.org>

Right now, CLONE_DETACHED threads silently vanish from GDB's sight when
they exit.  This patch lets the thread report its exit to the debugger, and
then be auto-reaped as soon as it is collected, instead of being reaped as
soon as it exits and not reported at all.

GDB works either way, but this is more correct and will be useful for some
later GDB patches.

From: Adrian Bunk <bunk@fs.tum.de>

I got an error at the final linking with CONFIG_TC35815 enabled since
the variables tc_readl and tc_writel are not available.

The only place where they are defined is arch/mips/pci/ops-jmr3927.c.

Replace it with the blockdev inode's i_sem.  And we only really need that for
atomic access to file->f_pos.

Rework the file_ops.flush() API sothat it is no longer called under
lock_kernel().  Push lock_kernel() down to all impementations except CIFS,
which doesn't want it.

From: William Lee Irwin III <wli@holomorphy.com>

Remove spurious BKL acquisitions in /proc/.  The BKL is not required to
access nr_threads for reporting, and get_locks_status() takes it
internally, wrapping all operations with it.

lock_kernel() need not be held across truncate.

`attr' is on the stack, and the inode's contents can change as soon as we
return from inode_change_ok() anyway.  I can't see anything which is actually
being locked in there.

Teach ramfs to use generic_file_llseek: default_llseek takes lock_kernel().

From: Dave Hansen <haveblue@us.ibm.com>

The current numa meminfo code exports (via sysfs) pgdat->node_size, as
totalram.  This variable is consistently used elsewhere to mean "the number
of physical pages that this particular node spans".  This is _not_ what we
want to see from meminfo, which is: "how much actual memory does this node
have?"

The following patch removes pgdat->node_size, and replaces it with
->node_spanned_pages.  This is to avoid confusion with a new variable,
node_present_pages, which is the _actual_ value that we want to export in
meminfo.  Most of the patch is a simple s/node_size/node_spanned_pages/.
The node_size() macro is also removed, and replaced with new ones for
node_{spanned,present}_pages() to avoid confusion.

We were bitten by this problem in this bug:
	http://bugme.osdl.org/show_bug.cgi?id=818

Compiled and tested on NUMA-Q.

From: Manfred Spraul <manfred@colorfullife.com>

Manfred's latest page unmapping debug patch.

The patch adds support for a special debug mode to both the page and the slab
allocator: Unused pages are removed from the kernel linear mapping.  This
means that now any access to freed memory will cause an immediate exception.
Right now, read accesses remain totally unnoticed and write accesses may be
catched by the slab poisoning, but usually far too late for a meaningfull bug
report.

The implementation is based on a new arch dependant function,
kernel_map_pages(), that removes the pages from the linear mapping.  It's
right now only implemented for i386.

Changelog:

- Add kernel_map_pages() for i386, based on change_page_attr.  If
  DEBUG_PAGEALLOC is not set, then the function is an empty stub.  The stub
  is in <linux/mm.h>, i.e.  it exists for all archs.

- Make change_page_attr irq safe.  Note that it's not fully irq safe due to
  the lack of the tlb flush ipi, but it's good enough for kernel_map_pages().
   Another problem is that kernel_map_pages is not permitted to fail, thus
  PSE is disabled if DEBUG_PAGEALLOC is enabled

- use kernel_map pages for the page allocator.

- use kernel_map_pages for the slab allocator.

  I couldn't resist and added additional debugging support into mm/slab.c:

  * at kfree time, the complete backtrace of the kfree caller is stored
    in the freed object.

  * a ptrinfo() function that dumps all known data about a kernel virtual
    address: the pte value, if it belongs to a slab cache the cache name and
    additional info.

  * merging of common code: new helper function obj_dbglen and obj_dbghdr
    for the conversion between the user visible object pointers/len and the
    actual, internal addresses and len values.

From: Hugh Dickins <hugh@veritas.com>

mremap's move_vma VM_LOCKED case was still wrong.

If the do_munmap unmaps a part of new_vma, then its vm_start and vm_end
from before cannot both be the right addresses for the make_pages_present
range, and may BUG() there.

We need [new_addr, new_addr+new_len) to be locked down; but
move_page_tables already transferred the locked pages [new_addr,
new_addr+old_len), and they're either held in a VM_LOCKED vma throughout,
or temporarily in no vma: in neither case can be swapped out, so no need to
run over that range again.

With the recent fixes, io_schedule needs to be exported for modular dm
to work.

Replace a couple of bogus yields() with schedule() and io_schedule()
respectively.

This includes the remainder of arch-specific part of the SH merge.
This patch only effects arch/sh and include/asm-sh, against current BK.

Okay, since now corruption happens and there are some other issues to be
resolved ("bad: scheduling while atomic" and "/proc/ide/hdX/identify")
better set it by default to n for 2.5.74, also mark it EXPERIMENTAL.

AMD64 does not support a.out, so don't display it in the configuration.

AMD64 like IA64 needs to force IPC_64 in the IPC functions. This makes
2.5 compatible with 2.4 again.

Since KALLSYMS is defined, might aswell use it somewhere.

Spelling fix