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

As described in this bug: http://bugme.osdl.org/show_bug.cgi?id=653 , if
you enable Summit support and NUMA Discontigmem support but boot on a
non-Summit box, the kernel will fail to boot.  The problem is that the
Summit code can not correctly get the NUMA memory configuration of a
flat box.  The code to do that is in get_memcfg_numa_flat(), but it
never gets called.

This patch implements a fallback to the generic NUMA code in
get_memcfg_numa_flat() if the Summit detection fails.  The patch also
adds the necessary bits to the Summit code so that it *knows* when it
fails.

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

any_online_cpu() is required to and with cpu_online_map before
attempting to find an online cpu somewhere in the map; this patch adds
that logic to the implementation(s) of any_online_cpu().

From: Angus Sawyer <angus.sawyer@dsl.pipex.com>

The mwave driver requires [un]register_char from 8250.c
Make sure 8250.c gets compilied.

agp_setup() is unreferenced if it's a module.

Hopelessly out of date

From: Manfred Spraul <manfred@colorfullife.com>

The real memory allocation is usually larger than the actual object size:
either due to L1 cache line padding, or due to page padding with
CONFIG_DEBUG_PAGEALLOC.  Right now objects are placed to the beginning of
the real allocation, but to trigger bugs it's better to move objects to the
end of the real allocation: that way accesses behind the end of the
allocation have a larger chance of hitting the (unmapped) next page.  The
attached patch moves the objects to align them with the end of the real
allocation.

Actually it contains 4 seperate changes:

- Do not page-pad allocations that are <= SMP_CACHE_LINE_SIZE.  This
  crashes.  Right now the limit is hardcoded to 128 bytes, but sooner or
  later an arch will appear with 256 byte cache lines.

- cleanup: redzone bytes are not accessed with inline helper functions,
  instead of magic offsets scattered throughout slab.c

- main change: move objects to the end of the allocation - trivial after
  the cleanup.

- Print old redzone value if a redzone mismatch happens: This makes it
  simpler to figure out what happened [single bit error, wrong redzone
  code, overwritten]

might_sleep() can be triggered by either local interrupts being disabled or
by elevated preempt count.  Disambiguate them.

From: Con Kolivas <kernel@kolivas.org>

Interactivity scheduler tweaks on top of Ingo's A3 interactivity patch.

Interactive credit added to task struct to find truly interactive tasks and
treat them differently.

Extra #defines included as helpers for conversion to/from nanosecond timing,
to work out an average timeslice for nice 0 tasks, and the effective dynamic
priority bonuses that will be given to tasks.

MAX_SLEEP_AVG modified to change dynamic priority by one for a nice 0 task
sleeping or running for one full timeslice.

CREDIT_LIMIT is the number of times a task earns sleep_avg over MAX_SLEEP_AVG
before it is considered HIGH_CREDIT (truly interactive); and -CREDIT_LIMIT is
LOW_CREDIT

TIMESLICE GRANULARITY is modified to be more frequent for more
interactivetasks (10 ms for top 2 dynamic priorities and then halving each
priority belowthat) and less frequent per extra cpu.

JUST_INTERACTIVE_SLEEP logic created to be a sleep_avg consistent with giving
a task enough dynamic priority to remain on the active array.

Task preemption of equal priority tasks is dropped as requeuing with
TIMESLICE_GRANULARITY makes this unecessary.

Dynamic priority bonus simplified.

User tasks that sleep a long time and not waking from uninterruptible sleep
are sought and categorised as idle. Their sleep avg is limited in it's rise to
prevent them becoming high priority and suddenly turning into cpu hogs.

Bonus for sleeping is proportionately higher the lower the dynamic priority of
a task is; this allows for very rapid escalation to interactive status.

Tasks that are LOW_CREDIT are limited in rise per sleep to one priority level.

Non HIGH_CREDIT tasks waking from uninterruptible sleep are sought to detect
cpu hogs waiting on I/O and their sleep_avg rise is limited to just
interactive state to prevent cpu bound tasks from becoming interactive during
I/O wait.

Tasks that earn sleep_avg over MAX_SLEEP_AVG get interactive credits.

On runqueue bonus is not given to non HIGH_CREDIT tasks waking from
uninterruptible sleep.

Forked tasks and their parents get sleep_avg limited to the minimum necessary
to maintain their effective dynamic priority thus preventing repeated forking
from being a way to get highly interactive, but not penalise them noticably
otherwise.

CAN_MIGRATE_TASK cleaned up and modified to work with nanosecond timestamps.

Reverted Ingo's A3 Starvation limit change - it was making interactive tasks
suffer more under increasing load. If a cpu is grossly overloaded and
everyone is going to starve it may as well run interactive tasks
preferentially.

Task requeuing is limited to interactive tasks only (cpu bound tasks dont need
low latency and derive benefit from longer timeslices), and they must have at
least TIMESLICE_GRANULARITY remaining.

HIGH_CREDIT tasks get penalised less sleep_avg the more interactive they are
thus keeping them interactive for bursts but if they become sustained cpu hogs
they will slide increasingly rapidly down the dynamic priority scale.

Tasks that run out of sleep_avg, are still using up cpu time and are not high
or low credit yet get penalised interactive credits to determine LOW_CREDIT
tasks (cpu bound ones).

From: Nick Piggin <piggin@cyberone.com.au>

The patch changes the imbalance required before a balance to 25% from 50% -
as the comments intend.  It also changes a case where the balancing
wouldn't be done if the imbalance was >= 25% but only 1 task difference.

The downside of the second change is that one task may bounce from one cpu
to another for some loads.  This will only bounce once every 200ms, so it
shouldn't be a big problem.

(Benchmarking results are basically a wash - SDET is increased maybe 0.5%)

Ingo's CPU scheduler update (in -mm kernels) needs a new sched_clock()
function which returns nanoseconds.

The patch provides implementations for ppc, ppc64, x86_64 and sparc64.

The x86_64 version could have overflow issues, the calculation is done in
32bits only with an multiply.  But I hope it's good enough for the scheduler

The ppc64 version needs scaling: it's only accurate for 1GHz CPUs.

From: Ingo Molnar <mingo@elte.hu>

the attached scheduler patch (against test2-mm2) adds the scheduling
infrastructure items discussed on lkml. I got good feedback - and while i
dont expect it to solve all problems, it does solve a number of bad ones:

 - test_starve.c code from David Mosberger

 - thud.c making the system unusuable due to unfairness

 - fair/accurate sleep average based on a finegrained clock

 - audio skipping way too easily

other changes in sched-test2-mm2-A3:

 - ia64 sched_clock() code, from David Mosberger.

 - migration thread startup without relying on implicit scheduling
   behavior. While the current 2.6 code is correct (due to the cpu-up code
   adding CPUs one by one), but it's also fragile - and this code cannot
   be carried over into the 2.4 backports. So adding this method would
   clean up the startup and would make it easier to have 2.4 backports.

and here's the original changelog for the scheduler changes:

 - cycle accuracy (nanosec resolution) timekeeping within the scheduler.
   This fixes a number of audio artifacts (skipping) i've reproduced. I
   dont think we can get away without going cycle accuracy - reading the
   cycle counter adds some overhead, but it's acceptable. The first
   nanosec-accuracy patch was done by Mike Galbraith - this patch is
   different but similar in nature. I went further in also changing the
   sleep_avg to be of nanosec resolution.

 - more finegrained timeslices: there's now a timeslice 'sub unit' of 50
   usecs (TIMESLICE_GRANULARITY) - CPU hogs on the same priority level
   will roundrobin with this unit. This change is intended to make gaming
   latencies shorter.

 - include scheduling latency in sleep bonus calculation. This change
   extends the sleep-average calculation to the period of time a task
   spends on the runqueue but doesnt get scheduled yet, right after
   wakeup. Note that tasks that were preempted (ie. not woken up) and are
   still on the runqueue do not get this benefit. This change closes one
   of the last hole in the dynamic priority estimation, it should result
   in interactive tasks getting more priority under heavy load. This
   change also fixes the test-starve.c testcase from David Mosberger.


The TSC-based scheduler clock is disabled on ia32 NUMA platforms.  (ie. 
platforms that have unsynched TSC for sure.) Those platforms should provide
the proper code to rely on the TSC in a global way.  (no such infrastructure
exists at the moment - the monotonic TSC-based clock doesnt deal with TSC
offsets either, as far as i can tell.)

From: OGAWA Hirofumi <hirofumi@mail.parknet.co.jp>

This elisp was obsolete on recently emacs's cc-mode. And this should
be personally set.

From: OGAWA Hirofumi <hirofumi@mail.parknet.co.jp>

This does the following,

1) This removes the nls modules for only alias. For backward
   compatible, this adds ->alias, and ->alias provides alias of charset.

2) For autoloading the module by the alias, use MODULE_ALIAS mechanism.

3) From changelog of module-init-tools, looks like MODULE_ALIAS needs
   module-init-tools 0.9.10 or later. So change the "Documentation/Changes".

From: Stephen Hemminger <shemminger@osdl.org>

Get rid of warnings (and dead code) if MTRR is compiled without /proc

From: James Cleverdon <jamesclv@us.ibm.com>

Some very large systems overflow the array and corrupt memory.  A BUG_ON will 
at least flag the problem until dynamic irq_vector allocation is added.

From: Oleg Drokin <green@namesys.com>

Fix trucation-induced large file corruption in reiserfs.

From: Jeremy Fitzhardinge <jeremy@goop.org>

I'm resending my patch to fix this problem.  To recap: every task_struct
has its own copy of the thread group's pgrp.  Only the thread group
leader is allowed to change the tgrp's pgrp, but it only updates its own
copy of pgrp, while all the other threads in the tgrp use the old value
they inherited on creation.

This patch simply updates all the other thread's pgrp when the tgrp
leader changes pgrp.  Ulrich has already expressed reservations about
this patch since it is (1) incomplete (it doesn't cover the case of
other ids which have similar problems), (2) racy (it doesn't synchronize
with other threads looking at the task pgrp, so they could see an
inconsistent view) and (3) slow (it takes linear time with respect to
the number of threads in the tgrp).

My reaction is that (1) it fixes the actual bug I'm encountering in a
real program.  (2) doesn't really matter for pgrp, since it is mostly an
issue with respect to the terminal job-control code (which is even more
broken without this patch.  Regarding (3), I think there are very few
programs which have a large number of threads which change process group
id on a regular basis (a heavily multi-threaded job-control shell?).

Ulrich also said he has a (proposed?) much better fix, which I've been
looking forward to.  I'm submitting this patch as a stop-gap fix for a
real bug, and perhaps to prompt the improved patch.

An alternative fix, at least for pgrp, is to change all references to
->pgrp to group_leader->pgrp.  This may be sufficient on its own, but it
would be a reasonably intrusive patch (I count 95 instances in 32 files
in the 2.6.0-test3-mm3 tree).

From: Robert Love <rml@tech9.net>

- Let real-time tasks dip further into the reserves than usual in
  __alloc_pages().  There are a lot of ways to special case this.  This
  patch just cuts z->pages_low in half, before doing the incremental min
  thing, for real-time tasks.  I do not do anything in the low memory slow
  path.  We can be a _lot_ more aggressive if we want.  Right now, we just
  give real-time tasks a little help.

- Never ever call balance_dirty_pages() on a real-time task.  Where and
  how exactly we handle this is up for debate.  We could, for example,
  special case real-time tasks inside balance_dirty_pages().  This would
  allow us to perform some of the work (say, waking up pdflush) but not
  other work (say, the active throttling).  As it stands now, we do the
  per-processor accounting in balance_dirty_pages_ratelimited() but we
  never call balance_dirty_pages().  Lots of approaches work.  What we want
  to do is never engage the real-time task in forced writeback.

From: "Nakajima, Jun" <jun.nakajima@intel.com>

Split the increasingly messy compiler.h file into per-compiler files and also
add support for non-gcc compilers.  

With the current implementation:

  include/linux/compiler.h defines the compiler-dependent abstractions
  which can be overwritten by per-compiler definitions.

  include/linux/compiler-gcc.h contains the common definitions for all gcc
  versions.

  include/linux/compiler-gcc[2,3,+].h contains gcc major version specific
  definitions.

  include/linux/compiler-intel.h contains intel compiler specific
  definitions."

- declarations come first

- fix bizarre coding style.

into home.osdl.org:/home/torvalds/v2.5/linux

The problem was that the semaphore which prevents ds interfering with
the sleepy card initialisation (skt_sem in pccardd) is blocking insmod
of the socket driver.  However, the socket driver is being called with
the PCI bus semaphore held by the driver model.

pccardd in turn discovered a cardbus card (with skt_sem held), so it
is trying to add the PCI devices to the PCI bus, and this requires the
driver model to grab the PCI bus semaphore, but its already locked.

We move the class device register into pccardd so we get a natural
ordering between the ds socket initialisation and pccardd trying to
detect inserted cards.

We also fix a potential use-after-free caused by rmmod'ing the socket
driver before ds has shut down.

into home.osdl.org:/home/torvalds/v2.5/linux

Don't issue a system call from kernel_thread(), but call do_fork()
directly.  This avoids all the unnecessary syscall overhead.

It's bad to make (CLONE_THREAD | CLONE_DETACHED) tasks
be _completely_ hidden. Resource consumption is hard
to track down if a user can hide a task from /bin/ps.

This patch, supported by the procps-3.1.13 release,
gives admins the ability to search for such tasks.
The top-level /proc directory remains uncontaminated.

When I change the setting of CONFIG_EISA, everything rebuilds.  This is
because EISA_bus is declared in <asm/processor.h> which is implicitly
included by just about everything.  This is a silly place to declare it,
so this patch moves it to include/linux/eisa.h.

While I'm at it, I also move the variable definition to
drivers/eisa/eisa-bus.c.  The rest of this patch is fixing up the fallout
from having to include <linux/eisa.h> if you use EISA_bus.

automatically as needed.

Remove duplicate includes, ensure that linux/config.h is included
where necessary, and linux/config.h isn't included where it isn't
needed.

into home.osdl.org:/home/torvalds/v2.5/linux

into home.osdl.org:/home/torvalds/v2.5/linux

* increase length of wait for analog-ready
* fix printk prefix (cosmetic)
* bail out for pure MMIO interfaces, instead of attempting to use it
  via PIO and hanging the system.

into home.osdl.org:/home/torvalds/v2.5/linux

Small clean up, to use current APIs.

this one fixes both drivers. They have been broken since the
reorganization in June. Unfortunately, my major build platform (alpha) had
some issues with later 2.5.X kernels (cache problems causing all kind of
funny behaviour), so i waited until these problems had been resolved by
the alpha gurus :-/...

In the mean time i also upgraded pine. I hope the patch corruption issue
is resolved by now.

Update hamradio/scc for 2.6.0-test5
	- use seq_file for /proc
	- get rid of dev_get()
	- use alloc_netdev

Don't have hardware, but can load/unload the module fine.

This patch converts /proc/net/bondX/info into /proc/net/bonding/bondX
using the seq_file interface.

This is based on Stephen's recent patch, but slightly modified to work
with the propagation patch set and with some locking changes to make it
simpler.

The patch applies both on 2.4 (after the sync set from earlier today)
and on 2.6 (after the propagation set from 2003/11/9).

Amir