If all mirror legs fail, always return an error instead of holding the
bio, even if the handle_errors option was set.  At present it is the
responsibility of the driver underneath us to deal with retries,
multipath etc.

The patch adds the bio to the failures list instead of holding it
directly.  do_failures tests first if all legs failed and, if so,
returns the bio with -EIO.  If any leg is still alive and handle_errors
is set, do_failures calls hold_bio.
Reviewed-by: NTakahiro Yasui <tyasui@redhat.com>
Signed-off-by: NMikulas Patocka <mpatocka@redhat.com>
Signed-off-by: NAlasdair G Kergon <agk@redhat.com>

This patch pulls the pg_init path activation code out of
process_queued_ios() into a new function.

No functional change.
Signed-off-by: NKiyoshi Ueda <k-ueda@ct.jp.nec.com>
Signed-off-by: NJun'ichi Nomura <j-nomura@ce.jp.nec.com>
Signed-off-by: NAlasdair G Kergon <agk@redhat.com>

When suspending the device we must wait for all I/O to complete, but
pg-init may be still in progress even after flushing the workqueue
for kmpath_handlerd in multipath_postsuspend.

This patch waits for pg-init completion correctly in
multipath_postsuspend().
Signed-off-by: NKiyoshi Ueda <k-ueda@ct.jp.nec.com>
Signed-off-by: NJun'ichi Nomura <j-nomura@ce.jp.nec.com>
Signed-off-by: NAlasdair G Kergon <agk@redhat.com>

m->queue_io is set to block processing I/Os, and it needs to be kept
while pg-init, which issues multiple path activations, is in progress.
But m->queue is cleared when a path activation completes without error
in pg_init_done(), even while other path activations are in progress.
That may cause undesired -EIO on paths which are not complete activation.

This patch fixes that by not clearing m->queue_io until all path
activations complete.

(Before the hardware handlers were moved into the SCSI layer, pg_init
only used one path.)
Signed-off-by: NKiyoshi Ueda <k-ueda@ct.jp.nec.com>
Signed-off-by: NJun'ichi Nomura <j-nomura@ce.jp.nec.com>
Signed-off-by: NAlasdair G Kergon <agk@redhat.com>

'suspended' flag in struct multipath was introduced to check whether
the multipath target is in suspended state, but the same check is
done through dm_suspended() now, so remove the flag and related code.
Signed-off-by: NKiyoshi Ueda <k-ueda@ct.jp.nec.com>
Signed-off-by: NJun'ichi Nomura <j-nomura@ce.jp.nec.com>
Cc: Mike Anderson <andmike@linux.vnet.ibm.com>
Signed-off-by: NAlasdair G Kergon <agk@redhat.com>

Remove the dm_get() in dm_table_get_md() because dm_table_get_md() could
be called from presuspend/postsuspend, which are called while
mapped_device is in DMF_FREEING state, where dm_get() is not allowed.

Justification for that is the lifetime of both objects: As far as the
current dm design/implementation, mapped_device is never freed while
targets are doing something, because dm core waits for targets to become
quiet in dm_put() using presuspend/postsuspend.  So targets should be
able to touch mapped_device without holding reference count of the
mapped_device, and we should allow targets to touch mapped_device even
if it is in DMF_FREEING state.

Backgrounds:
I'm trying to remove the multipath internal queue, since dm core now has
a generic queue for request-based dm.  In the patch-set, the multipath
target wants to request dm core to start/stop queue.  One of such
start/stop requests can happen during postsuspend() while the target
waits for pg-init to complete, because the target stops queue when
starting pg-init and tries to restart it when completing pg-init.  Since
queue belongs to mapped_device, it involves calling dm_table_get_md()
and dm_put().  On the other hand, postsuspend() is called in dm_put()
for mapped_device which is in DMF_FREEING state, and that triggers
BUG_ON(DMF_FREEING) in the 2nd dm_put().

I had tried to solve this problem by changing only multipath not to
touch mapped_device which is in DMF_FREEING state, but I couldn't and I
came up with a question why we need dm_get() in dm_table_get_md().
Signed-off-by: NKiyoshi Ueda <k-ueda@ct.jp.nec.com>
Signed-off-by: NJun'ichi Nomura <j-nomura@ce.jp.nec.com>
Signed-off-by: NAlasdair G Kergon <agk@redhat.com>

This patch adds two minor fixes while processing device mapper path activation.

Skip failed paths while calling activate_path.  If the path is already failed
then activate_path will fail for sure. We don't have to call in that case. In
some case this might cause prolonged retries unnecessarily.

Change the misleading message if the path being activated fails with SCSI_DH_NOSYS.
Signed-off-by: NBabu Moger <babu.moger@lsi.com>
Signed-off-by: NAlasdair G Kergon <agk@redhat.com>

This patch removes some unnecessary argument casting. There is no
functional change with this patch.

Passes 'struct pgpath' through to pg_init_done() instead of the enclosed
'struct dm_path'.

Tested the changes with LSI storage..

CC: Chandra Seetharaman <chandra.seetharaman@us.ibm.com>
Signed-off-by: NBabu Moger <babu.moger@lsi.com>
Acked-by: NKiyoshi Ueda <k-ueda@ct.jp.nec.com>
Signed-off-by: NAlasdair G Kergon <agk@redhat.com>

If the calling convention of ->timer_fn() and ->cleanup_fn() are unified
across hardware versions we can drop parameters to ioat_init_channel() and
unify ioat_is_dma_complete() implementations.

Both ->timer_fn() and ->cleanup_fn() are modified to expect a struct
dma_chan pointer.
Signed-off-by: NDan Williams <dan.j.williams@intel.com>

The hardware automatically disables further interrupts after each event
until rearmed.  This allows a delay to be injected between the occurence
of the interrupt and the running of the cleanup routine.  The delay is
scaled by the descriptor backlog and then written to the INTRDELAY
register which specifies the number of microseconds to hold off
interrupt delivery after an interrupt event occurs.  According to
powertop this reduces the interrupt rate from ~5000 intr/s to ~150
intr/s per without affecting throughput (simple dd to a raid6 array).
Signed-off-by: NDan Williams <dan.j.williams@intel.com>

Since ioat_cleanup_preamble() and the update of the last completed
descriptor are not synchronized there is a chance that two cleanup threads
can see descriptors to clean.  If the first cleans up all pending
descriptors then the second will trigger the BUG_ON.
Signed-off-by: NDan Williams <dan.j.williams@intel.com>

No one is calling this anymore as everyone has switched to
invalidate_mapping_pages long time ago.  Also update a few
references to it in comments.  nfs has two more, but I can't
easily figure what they are actually referring to, so I left
them as-is.
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

The pending == 2 case no longer exists in the driver so, we can use
ioat2_ring_pending() outside the lock to determine if there might be any
descriptors in the ring that the hardware has not seen.
Signed-off-by: NDan Williams <dan.j.williams@intel.com>

Replace open coded ioat2_quiesce() call in ioat3_restart_channel
Signed-off-by: NDan Williams <dan.j.williams@intel.com>

We already disallow raid operations while DCA is globally enabled, so
having it locally enabled is a nop and confusing when reading the code.
Signed-off-by: NDan Williams <dan.j.williams@intel.com>

On Wed, Feb 24, 2010 at 08:40:56PM +0000, Mark Brown wrote:
> On Wed, Feb 24, 2010 at 11:21:26AM -0800, Dmitry Torokhov wrote:
> > On Wed, Feb 24, 2010 at 07:14:03PM +0000, Mark Brown wrote:
>
> > > This doesn't help unless you also provide a way for users to obtain a
> > > struct wm8400.
>
> > Why would they need it? Only code that creates instances of wm8400 needs
> > to know the definition of the sturcture, the rest can simply pass the
> > pointer around.
>
> > I guess there is disconnect between us and I do not see any users of
> > wm8400_register_regulator() in linux-next... Is there another tree I
> > could peek at?
>
> There are no users in mainline.  This would be called by board specific
> code from the init callback of the wm8400 - you'd need to pass that
> callback the struct wm8400.
>
> In any case, this is clearly an unrelated change to whatever else you
> were doing to the driver so should be split off into a separate patch,
> but if this is being changed at all then it'd be much more sensible to
> change it to use a more modern pattern which completely removes the
> wm8400_register_regulator() function and just uses platform data.

Fair enough, I removed the offending part, updated patch below.

--
Dmitry

regulator: wm8400 - cleanup platform driver data handling

Driver data set by platform_set_drvdata() is for private use of
the driver currently bound to teh device and not for use by parent,
subsystem and anyone else.
Signed-off-by: NDmitry Torokhov <dtor@mail.ru>
Acked-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NLiam Girdwood <lrg@slimlogic.co.uk>

It is a good tone to reset driver data after unbinding the device.
Signed-off-by: NDmitry Torokhov <dtor@mail.ru>
Acked-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NLiam Girdwood <lrg@slimlogic.co.uk>

It is a good tone to reset driver data after unbinding the device.
Also set up drivers owner.
Signed-off-by: NDmitry Torokhov <dtor@mail.ru>
Acked-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NLiam Girdwood <lrg@slimlogic.co.uk>

It is a good tone to reset driver data after unbinding the device.
Signed-off-by: NDmitry Torokhov <dtor@mail.ru>
Acked-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NLiam Girdwood <lrg@slimlogic.co.uk>

Signed-off-by: NDmitry Torokhov <dtor@mail.ru>
Acked-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NLiam Girdwood <lrg@slimlogic.co.uk>

Signed-off-by: NDmitry Torokhov <dtor@mail.ru>
Acked-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NLiam Girdwood <lrg@slimlogic.co.uk>

There is no point in completing probe if platform data is missing so
let's abort loading early.

Also, use kcalloc when allocating several instances of the same data
structure and mark setup_regulators() as __devinit since it is only
called from lp3971_i2c_probe() which is __devinit.
Signed-off-by: NDmitry Torokhov <dtor@mail.ru>
Acked-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NLiam Girdwood <lrg@slimlogic.co.uk>

Also move error handling in probe() out of line and do not bother
to reset fields in structures that are about to be freed.
Signed-off-by: NDmitry Torokhov <dtor@mail.ru>
Acked-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NLiam Girdwood <lrg@slimlogic.co.uk>

Also move error handling in probe() out of line and do not bother
to reset fields in structures that are about to be freed.
Signed-off-by: NDmitry Torokhov <dtor@mail.ru>
Acked-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NLiam Girdwood <lrg@slimlogic.co.uk>

Signed-off-by: NDmitry Torokhov <dtor@mail.ru>
Acked-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NLiam Girdwood <lrg@slimlogic.co.uk>

Add __devinit/__devexit markings to probe and remove methids of the
driver, change types of variables containing boolean data to boolean,
set up driver's owner field so we have proper sysfs link between
driver and the module.
Signed-off-by: NDmitry Torokhov <dtor@mail.ru>
Acked-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NLiam Girdwood <lrg@slimlogic.co.uk>

Probe and remove methods should not be marked as __init/__exit but
rather __devinit/__devexit so that the needed sections stay in memory
in presence of CONFIG_HOTPLUG. This is needed even on non hotpluggable
buses.
Signed-off-by: NDmitry Torokhov <dtor@mail.ru>
Acked-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NLiam Girdwood <lrg@slimlogic.co.uk>

Instead of open-coding sysfs attribute group use canned solution.
Also add __devinit/__devexit markups for probe and remove methods
and use 'bool' where it makes sense.
Signed-off-by: NDmitry Torokhov <dtor@mail.ru>
Acked-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NLiam Girdwood <lrg@slimlogic.co.uk>

This is no REMAP register on twl6030, instead there is a STATE register
to drive a resource to a given state.
The state register can be used to specify what state the resource should
enter when its associated with a GRP.
Register Bit field description is as below. The patch programmes the
corresponding STATE registers for all LDO's to turn ON when assocaited
with GRP_P1.

STATE REG:
Bit7   |Bit6   |Bit5   |Bit4  |Bit3  |Bit2  |Bit1   |Bit0
P3_GRP |P2_GRP |P1_GRP |RES   |RES   |RES   |State1 |State0

State can be specified as below
00: OFF
01: ON
10: OFF
11: SLEEP
Signed-off-by: NRajendra Nayak <rnayak@ti.com>
Cc: Liam Girdwood <lrg@slimlogic.co.uk>
Cc: Samuel Ortiz <sameo@linux.intel.com>
Cc: Mark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NLiam Girdwood <lrg@slimlogic.co.uk>

In order to ease transitions with drivers are boards start using regulators
provide an option to cause all regulator_get() calls to succeed, with a
dummy always on regulator being supplied where one has not been configured.
A warning is printed whenever the dummy regulator is used to aid system
development.

This regulator does not implement any regulator operations but will allow
simple consumers which only do enable() and disable() calls to run. It
is kept separate from the fixed voltage regulator to avoid Kconfig
confusion on the part of users when it is extended to allow boards to
explicitly use the dummy regulator to simplify cases where the majority
of supplies are from fixed regulators without software control.

This option is currently only effective for systems which do not specify
full constriants. If required an override could also be provided to allow
these systems to use the dummy regulator, though it is likely that
unconfigured supplies on such systems will lead to error due to
regulators being powered down more aggressively when not in use.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NLiam Girdwood <lrg@slimlogic.co.uk>

If a regulator driver does not provide a way to query if the driver is
enabled then assume that it is enabled.  This is very likely to reflect
the actual state is more useful for callers than reporting an error.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NLiam Girdwood <lrg@slimlogic.co.uk>

It had an open coded version in enable().
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NLiam Girdwood <lrg@slimlogic.co.uk>

The WM8994 contains two LDOs with mixed hardware/software control to
minimise the number of external supplies required while delivering
optimal voltages to minimise power consumption.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NLiam Girdwood <lrg@slimlogic.co.uk>

Signed-off-by: NHaojian Zhuang <haojian.zhuang@marvell.com>
Acked-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NLiam Girdwood <lrg@slimlogic.co.uk>

Fix Kconfig text for some Wolfson Micro devices.

Cc: Mark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NAlex Chiang <achiang@hp.com>
Acked-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NLiam Girdwood <lrg@slimlogic.co.uk>

GPO regulators are digital outputs that can be enabled or disabled by a
dedicated bit in mc13783 POWERMISC register.
In this family can be count in also Power Gates (PWGT1 and 2): enabled by
a dedicated pin a Power Gate is an hardware driven supply where the output
(PWGTnDRV) follow this law:

 Bit PWGTxSPIEN | Pin PWGTxEN | PWGTxDRV |  Read Back
   0 = default  |             |          | PWGTxSPIEN
 ---------------+-------------+----------+------------
       1        |      x      |   Low    |     0
       0        |      0      |   High   |     1
       0        |      1      |   Low    |     0

As read back value of control bit reflects the PWGTxDRV state (not the
control value previously written) and mc13783 POWERMISC register contain
only regulator related bits, a dedicated function to manage these bits is
created here with the aim of tracing the real value of PWGTxSPIEN bits
and reproduce it on next writes.

All POWERMISC users _must_ use the new function to not accidentally
disable Power Gates supplies.

v2 changes:
-Better utilization of abstraction layers.
-Voltage query support. GPO's and PWGTxDRV are fixed voltage regulator
 with voltage value of 3.1V and 5.5V respectively.
Signed-off-by: NAlberto Panizzo <maramaopercheseimorto@gmail.com>
Acked-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NLiam Girdwood <lrg@slimlogic.co.uk>

Add a field to specify a delay for the start-up time of
a fixed voltage regulator.
Signed-off-by: NAdrian Hunter <adrian.hunter@nokia.com>
Acked-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NLiam Girdwood <lrg@slimlogic.co.uk>

Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

Regulators may sometimes take longer to enable than the control operation
used to do so, either because the regulator has ramp rate control used to
limit inrush current or because the control operation is very fast (GPIO
being the most common example of this).  In order to ensure that consumers
do not rely on the regulator before it is enabled provide an enable_time()
operation and have the core delay for that time before returning to the
caller.

This is implemented as a function since the ramp rate may be specified in
voltage per unit time and therefore the time depend on the configuration.
In future it would be desirable to allow the bulk operations to run the
delays for multiple enables in parallel but this is not currently supported.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>