* 'for-2.6.23' of master.kernel.org:/pub/scm/linux/kernel/git/arnd/cell-2.6: (37 commits)
  [CELL] spufs: rework list management and associated locking
  [CELL] oprofile: add support to OProfile for profiling CELL BE SPUs
  [CELL] oprofile: enable SPU switch notification to detect currently active SPU tasks
  [CELL] spu_base: locking cleanup
  [CELL] cell: indexing of SPUs based on firmware vicinity properties
  [CELL] spufs: integration of SPE affinity with the scheduller
  [CELL] cell: add placement computation for scheduling of affinity contexts
  [CELL] spufs: extension of spu_create to support affinity definition
  [CELL] cell: add hardcoded spu vicinity information for QS20
  [CELL] cell: add vicinity information on spus
  [CELL] cell: add per BE structure with info about its SPUs
  [CELL] spufs: use find_first_bit() instead of sched_find_first_bit()
  [CELL] spufs: remove unused file argument from spufs_run_spu()
  [CELL] spufs: change decrementer restore timing
  [CELL] spufs: dont halt decrementer at restore step 47
  [CELL] spufs: limit saving MFC_CNTL bits
  [CELL] spufs: fix read and write for decr_status file
  [CELL] spufs: fix decr_status meanings
  [CELL] spufs: remove needless context save/restore code
  [CELL] spufs: fix array size of channel index
  ...

This basically reverts commit 4e44f349,
while waiting for it to be re-done more completely.  There are cases of
people mixing "time()" with higher-resolution time sources, and we need
to take the nanosecond offsets into account.

Ingo has a patch that does that, but it's still under some discussion.
In the meantime, just revert back to the old simple situation of just
doing the whole exact timesource calculations.

But rather than using do_gettimeofday(), use the internal nanosecond
resolution getnstimeofday(), which at least avoids one unnecessary
conversion (since we really don't care about whether the fractional
seconds are nanoseconds or microseconds - we'll just throw them away).
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

This sorts out the various lists and related locks in the spu code.

In detail:

 - the per-node free_spus and active_list are gone.  Instead struct spu
   gained an alloc_state member telling whether the spu is free or not
 - the per-node spus array is now locked by a per-node mutex, which
   takes over from the global spu_lock and the per-node active_mutex
 - the spu_alloc* and spu_free function are gone as the state change is
   now done inline in the spufs code.  This allows some more sharing of
   code for the affinity vs normal case and more efficient locking
 - some little refactoring in the affinity code for this locking scheme
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NArnd Bergmann <arnd.bergmann@de.ibm.com>

From: Maynard Johnson <mpjohn@us.ibm.com>

This patch updates the existing arch/powerpc/oprofile/op_model_cell.c
to add in the SPU profiling capabilities.  In addition, a 'cell' subdirectory
was added to arch/powerpc/oprofile to hold Cell-specific SPU profiling code.
Exports spu_set_profile_private_kref and spu_get_profile_private_kref which
are used by OProfile to store private profile information in spufs data
structures.

Also incorporated several fixes from other patches (rrn).  Check pointer
returned from kzalloc.  Eliminated unnecessary cast.  Better error
handling and cleanup in the related area.  64-bit unsigned long parameter
was being demoted to 32-bit unsigned int and eventually promoted back to
unsigned long.
Signed-off-by: NCarl Love <carll@us.ibm.com>
Signed-off-by: NMaynard Johnson <mpjohn@us.ibm.com>
Signed-off-by: NBob Nelson <rrnelson@us.ibm.com>
Signed-off-by: NArnd Bergmann <arnd.bergmann@de.ibm.com>
Acked-by: NPaul Mackerras <paulus@samba.org>

From: Maynard Johnson <mpjohn@us.ibm.com>

This patch adds to the capability of spu_switch_event_register so that
the caller is also notified of currently active SPU tasks.
Exports spu_switch_event_register and spu_switch_event_unregister so
that OProfile can get access to the notifications provided.
Signed-off-by: NMaynard Johnson <mpjohn@us.ibm.com>
Signed-off-by: NCarl Love <carll@us.ibm.com>
Signed-off-by: NBob Nelson <rrnelson@us.ibm.com>
Signed-off-by: NArnd Bergmann <arnd.bergmann@de.ibm.com>
Acked-by: NPaul Mackerras <paulus@samba.org>

Sort out the locking mess in spu_base and document the current rules.
As an added benefit spu_alloc* and spu_free don't block anymore.
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NArnd Bergmann <arnd.bergmann@de.ibm.com>

This patch links spus according to their physical position using
information provided by the firmware through a special vicinity
device-tree property. This property is present in current version
of Malta firmware.

Example of vicinity properties for a node in Malta:

Node:        Vicinity property contains phandles of:
spe@0        [ spe@100000 , mic-tm@50a000 ]
spe@100000   [ spe@0      , spe@200000    ]
spe@200000   [ spe@100000 , spe@300000    ]
spe@300000   [ spe@200000 , bif0@512000   ]
spe@80000    [ spe@180000 , mic-tm@50a000 ]
spe@180000   [ spe@80000  , spe@280000    ]
spe@280000   [ spe@180000 , spe@380000    ]
spe@380000   [ spe@280000 , bif0@512000   ]

Only spe@* have a vicinity property (e.g., bif0@512000 and
mic-tm@50a000 do not have it).
Signed-off-by: NAndre Detsch <adetsch@br.ibm.com>
Signed-off-by: NArnd Bergmann <arnd.bergmann@de.ibm.com>

This patch makes the scheduller honor affinity information for each
context being scheduled. If the context has no affinity information,
behaviour is unchanged. If there are affinity information, context is
schedulled to be run on the exact spu recommended by the affinity
placement algorithm.
Signed-off-by: NAndre Detsch <adetsch@br.ibm.com>
Signed-off-by: NArnd Bergmann <arnd.bergmann@de.ibm.com>

This patch provides the spu affinity placement logic for the spufs scheduler.
Each time a gang is going to be scheduled, the placement of a reference
context is defined. The placement of all other contexts with affinity from
the gang is defined based on this reference context location and on a
precomputed displacement offset.
Signed-off-by: NAndre Detsch <adetsch@br.ibm.com>
Signed-off-by: NArnd Bergmann <arnd.bergmann@de.ibm.com>

This patch adds support for additional flags at spu_create, which relate
to the establishment of affinity between contexts and contexts to memory.
A fourth, optional, parameter is supported. This parameter represent
a affinity neighbor of the context being created, and is used when defining
SPU-SPU affinity.
Affinity is represented as a doubly linked list of spu_contexts.
Signed-off-by: NAndre Detsch <adetsch@br.ibm.com>
Signed-off-by: NArnd Bergmann <arnd.bergmann@de.ibm.com>

This patch allows the use of spu affinity on QS20, whose
original FW does not provide affinity information.
This is done through two hardcoded arrays, and by reading the reg
property from each spu.
Signed-off-by: NAndre Detsch <adetsch@br.ibm.com>
Signed-off-by: NArnd Bergmann <arnd.bergmann@de.ibm.com>

This patch adds affinity data to each spu instance.
A doubly linked list is created, meant to connect the spus
in the physical order they are placed in the BE. SPUs
near to memory should be marked as having memory affinity.
Adjustments of the fields acording to FW properties is done
in separate patches, one for CPBW, one for Malta (patch for
Malta under testing).
Signed-off-by: NAndre Detsch <adetsch@br.ibm.com>
Signed-off-by: NArnd Bergmann <arnd.bergmann@de.ibm.com>

Addition of a spufs-global "cbe_info" array. Each entry contains information
about one Cell/B.E. node, namelly:
* list of spus (both free and busy spus are in this list);
* list of free spus (replacing the static spu_list from spu_base.c)
* number of spus;
* number of reserved (non scheduleable) spus.

SPE affinity implementation actually requires only access to one spu per
BE node (since it implements its own pointer to walk through the other spus
of the ring) and the number of scheduleable spus (n_spus - non_sched_spus)
However having this more general structure can be useful for other
functionalities, concentrating per-cbe statistics / data.
Signed-off-by: NAndre Detsch <adetsch@br.ibm.com>
Signed-off-by: NArnd Bergmann <arnd.bergmann@de.ibm.com>

spu_sched->bitmap has MAX_PRIO(=140) width in bits.However, since
ff80a77f, sched_find_first_bit()
only supports 100-bit bitmaps.

Thus, spu_sched->bitmap should be treated by generic find_first_bit().
Signed-off-by: NMasato Noguchi <Masato.Noguchi@jp.sony.com>
Signed-off-by: NJeremy Kerr <jk@ozlabs.org>
Signed-off-by: NArnd Bergmann <arnd.bergmann@de.ibm.com>

From: Sebastian Siewior <cbe-oss-dev@ml.breakpoint.cc>

The 'file' argument is unused in spufs_run_spu(). This change removes
it.
Signed-off-by: NSebastian Siewior <sebastian@breakpoint.cc>
Signed-off-by: NJeremy Kerr <jk@ozlabs.org>
Signed-off-by: NArnd Bergmann <arnd.bergmann@de.ibm.com>

The SPU decrementer should be restored after the LSCSA DMA has
completed.
Signed-off-by: NMasato Noguchi <Masato.Noguchi@jp.sony.com>
Signed-off-by: NJeremy Kerr <jk@ozlabs.org>
Signed-off-by: NArnd Bergmann <arnd.bergmann@de.ibm.com>

No need to halt the SPE decrementer at context restore step 47, it will
be done in step 7.
Signed-off-by: NMasato Noguchi <Masato.Noguchi@jp.sony.com>
Signed-off-by: NJeremy Kerr <jk@ozlabs.org>
Signed-off-by: NArnd Bergmann <arnd.bergmann@de.ibm.com>

At save step 8, the mfc control register in the CSA should be written
_only_ with Sc and Sm bits (at least MFC_CNTL[Dh] should be set to 0)
Signed-off-by: NMasato Noguchi <Masato.Noguchi@jp.sony.com>
Signed-off-by: NJeremy Kerr <jk@ozlabs.org>
Signed-off-by: NArnd Bergmann <arnd.bergmann@de.ibm.com>

The decr_status in the LSCSA is valid only in the sequence of context
restore. Thus, it's nonsense to read and/or write it through spufs.

This patch changes decr_status node to access MFC_CNTL[Ds] in the CSA.
Signed-off-by: NMasato Noguchi <Masato.Noguchi@jp.sony.com>
Signed-off-by: NJeremy Kerr <jk@ozlabs.org>
Signed-off-by: NArnd Bergmann <arnd.bergmann@de.ibm.com>

The decr_status in the LSCSA is confusedly used as two meanings:
 * SPU decrementer was running
 * SPU decrementer was wrapped as a result of adjust
and the code to set decr_status is missing.

This patch fixes these problems by using the decr_status argument as a
set of flags. This requires a rebuild of the shipped spu_restore code.
Signed-off-by: NMasato Noguchi <Masato.Noguchi@jp.sony.com>
Signed-off-by: NJeremy Kerr <jk@ozlabs.org>
Signed-off-by: NArnd Bergmann <arnd.bergmann@de.ibm.com>

The following steps are not needed in the SPE context save/restore
paths:

Save Step 12: save_mfc_decr()
  save suspend_time to CSA (It will be done by step 14)
  save ch 7 (decrementer value will be saved in LSCSA by spe-side step 10)

Restore Step 59: restore_ch_part1()
  restore ch 1 (it will be done by spe-side step 15)

This change removes the unnecessary steps.
Signed-off-by: NMasato Noguchi <Masato.Noguchi@jp.sony.com>
Signed-off-by: NJeremy Kerr <jk@ozlabs.org>
Signed-off-by: NArnd Bergmann <arnd.bergmann@de.ibm.com>

Based on a fix from Masato Noguchi <Masato.Noguchi@jp.sony.com>.

Remove the (incorrect) array size declarations in the spufs channel
arrays, and use ARRAY_SIZE rather than hardcoded values.
Signed-off-by: NJeremy Kerr <jk@ozlabs.org>
Signed-off-by: NArnd Bergmann <arnd.bergmann@de.ibm.com>

Currently a process is removed from the physical spu when spu_acquire_saved
is saved but never put back.  This patch adds a new spu_release_saved
that is to be paired with spu_acquire_saved and put the process back if
it has been in RUNNABLE state before.

Niether Jeremy not be are entirely happy about this exact patch because
it adds another spu_activate call outside of the owner thread, but I
feel this is the best short-term fix we can come up with.
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NJeremy Kerr <jk@ozlabs.org>
Signed-off-by: NArnd Bergmann <arnd.bergmann@de.ibm.com>

This patch exports per-context statistics in spufs as long as spu
statistics in sysfs.

It was formed by merging:
"spufs: add spu stats in sysfs"   From: Christoph Hellwig
"spufs: add stat file to spufs"   From: Christoph Hellwig
"spufs: fix libassist accounting" From: Jeremy Kerr
"spusched: fix spu utilization statistics" From: Luke Browning
And some adjustments by myself, after suggestions on cbe-oss-dev.

Having separate patches was making the review process harder
than it should, as we end up integrating spus and ctx statistics
accounting much more than it was on the first implementation.
Signed-off-by: NAndre Detsch <adetsch@br.ibm.com>
Signed-off-by: NJeremy Kerr <jk@ozlabs.org>
Signed-off-by: NArnd Bergmann <arnd.bergmann@de.ibm.com>

In 6cbf93960e64f313f6e247cbca7afaa50e3ee2c we added a WARN_ON for
calling spu_deactivate on contexts created with the SPU_CREATE_NOSCHED
flag. However, all NOSCHED contexts will need to be deactivated when
the context is destroyed, so this gives a spurious warning when any
NOSCHED context is closed.

This change removes the WARN_ON.
Signed-off-by: NJeremy Kerr <jk@ozlabs.org>
Signed-off-by: NArnd Bergmann <arnd.bergmann@de.ibm.com>

Reading from the signal{1,2} files requires a spu_acquire_saved, so
make these files write-only for contexts created with
SPU_CREATE_NOSCHED.
Signed-off-by: NJeremy Kerr <jk@ozlabs.org>
Signed-off-by: NArnd Bergmann <arnd.bergmann@de.ibm.com>

The current SPU context saving procedure in SPUFS unexpectedly
restarts MFC when halting decrementer, because MFC_CNTL[Dh] is set
without MFC_CNTL[Sm]. This bug causes, for example, saving broken DMA
queues. Here is a patch to fix the problem.
Signed-off-by: NKazunori Asayama <asayama@sm.sony.co.jp>
Signed-off-by: NJeremy Kerr <jk@ozlabs.org>
Signed-off-by: NArnd Bergmann <arnd.bergmann@de.ibm.com>

WARNING: arch/powerpc/platforms/cell/spufs/spufs.o(.init.text+0x158): Section
mismatch: reference to .exit.text:.spu_sched_exit (between '.init_module' and
'.spu_sched_init')

was introduced by c99c1994
This patch removes the warning.

Cc: Christoph Hellwig <hch@lst.de>
Signed-off-by: NSebastian Siewior <sebastian@breakpoint.cc>
Signed-off-by: NArnd Bergmann <arnd.bergmann@de.ibm.com>

This patch adds support for the setup and decoding of MSIs
on Axon-based Cell systems, using the MSIC mechanism.

This involves setting up an area of BE memory which the Axon
then uses as a FIFO for MSI messages. When one or more MSIs
are decoded by the MSIC we receive an interrupt on the MPIC,
and the MSI messages are written into the FIFO. At the moment
we use a 64KB FIFO, one per MSIC/BE.
Signed-off-by: NMichael Ellerman <michael@ellerman.id.au>
Signed-off-by: NArnd Bergmann <arnd.bergmann@de.ibm.com>

This patch adds support for investigating spus information after a
kernel crash event, through kdump vmcore file.
Implementation is based on xmon code, but the new functionality was
kept independent from xmon.
Signed-off-by: NLucio Jose Herculano Correia <luciojhc@br.ibm.com>
Signed-off-by: NAndre Detsch <adetsch@br.ibm.com>
Signed-off-by: NArnd Bergmann <arnd.bergmann@de.ibm.com>

The Axon bridge chip used on new Cell/B.E. based blade servers
comes with a DDR2 memory controller that can be used to
attach cheap memory modules, as opposed to the high-speed
XDR memory that is used by the CPU itself.

Since the memory controller does not participate in the
cache coherency protocol, we can not use the memory direcly
for Linux applications, but by providing a block device
it can be used for swap space, temporary file storage and
through the use of the direct_access block device operation
for mapping into user addresses, when it is mounted with
an appropriate file system.
Signed-off-by: NMaxim Shchetynin <maxim@de.ibm.com>
Signed-off-by: NArnd Bergmann <arnd.bergmann@de.ibm.com>

The platforms missing the "cpus" property in the "be" node are mono-Cell
platforms such as CAB or Getaway.

Therefore it is possible to assume that if there is no "cpus" properties
under the "be" node then we can safely return the "device node" without
more checking. This is a bit hacky but ... it allows it to work on
these platforms.
Signed-off-by: NJean-Christophe DUBOIS <jcd@tribudubois.net>
Acked-by: NChristian Krafft <krafft@de.ibm.com>
Signed-off-by: NArnd Bergmann <arnd.bergmann@de.ibm.com>

Previous patch changed based on Christian Krafft's comment.

On some legacy SLOF tree the generic code is unable to ioremap some Cell BE
registers. Therefore the "generic" functions are returning a NULL pointer,
triggering a crash on such platforms.

Let's handle this more gracefully.
Signed-off-by: NJean-Christophe DUBOIS <jcd@tribudubois.net>
Acked-by: NChristian Kraff <krafft@de.ibm.com>
Signed-off-by: NArnd Bergmann <arnd.bergmann@de.ibm.com>

Previous patch changed based on Christian Krafft's comment.

On some legacy SLOF tree the generic code is unable to ioremap some Cell BE
registers. Therefore the "generic" functions are returning a NULL pointer,
triggering a crash on such platforms.

Let's handle this more gracefully.
Signed-off-by: NJean-Christophe DUBOIS <jcd@tribudubois.net>
Acked-by: NChristian Kraff <krafft@de.ibm.com>
Signed-off-by: NArnd Bergmann <arnd.bergmann@de.ibm.com>

This patch reorganizes the code of the driver into three files.
Two cbe_cpufreq_pmi.c and cbe_cpufreq_pervasive.c care about hardware.
cbe_cpufreq.c contains the logic.
There is no changed behaviour, except that the PMI related function
is now located in a seperate module cbe_cpufreq_pmi. This module
will be required by cbe_cpufreq, if CONFIG_CBE_CPUFREQ_PMI has been set.
Signed-off-by: NChristian Krafft <krafft@de.ibm.com>
Signed-off-by: NArnd Bergmann <arnd.bergmann@de.ibm.com>

Minor issues have been fixed:
* added a missing call to of_node_put()
* signedness of a function parameter
* added some line breaks
* changed global pmi_frequency_limit to a
  per node pmi_slow_mode_limit array
Signed-off-by: NChristian Krafft <krafft@de.ibm.com>
Signed-off-by: NArnd Bergmann <arnd.bergmann@de.ibm.com>

This patch fixes the initialization of the cbe_cpufreq driver.
The code that initializes the PMI related functions was called per cpu:
* registering cpufreq notifier block
* registering a pmi handler

This ends in a bug that the notifier block gets called in an endless loop.
The initialization code is being put to the
module init code path by this patch. This way it only gets called once.
Signed-off-by: NChristian Krafft <krafft@de.ibm.com>
Signed-off-by: NArnd Bergmann <arnd.bergmann@de.ibm.com>

This patch fixes the debug code that calculates the transition time when
changing the slow modes on a Cell BE cpu.
Signed-off-by: NChristian Krafft <krafft@de.ibm.com>
Signed-off-by: NArnd Bergmann <arnd.bergmann@de.ibm.com>

The pmi driver got simplified by removing support for multiple devices.
As there is no more than one pmi device per maschine, there is no need to
specify the device for listening and sending messages.

This way the caller (cbe_cpufreq) doesn't need to scan the device tree.
When registering the handler on a board without a pmi
interface, pmi.c will just return -ENODEV.

The patch that fixed the breakage of cell_defconfig has been
broken out of the earlier version of this patch. So this is
the version that applies cleanly on top of it.
Signed-off-by: NChristian Krafft <krafft@de.ibm.com>
Signed-off-by: NArnd Bergmann <arnd.bergmann@de.ibm.com>

The comparison with ZERO_SIZE_PTR in ZERO_OR_NULL_PTR() needs to be <=
(not just <) so that ZERO_OR_NULL_PTR(ZERO_SIZE_PTR) is 1.
Signed-off-by: NRoland Dreier <rolandd@cisco.com>
[ Duh!  - Linus ]
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>