Impact: reduce memory usage, use new cpumask API.

Replace the affinity and pending_masks with cpumask_var_t's.  This adds
to the significant size reduction done with the SPARSE_IRQS changes.

The added functions (init_alloc_desc_masks & init_copy_desc_masks) are
in the include file so they can be inlined (and optimized out for the
!CONFIG_CPUMASKS_OFFSTACK case.)  [Naming chosen to be consistent with
the other init*irq functions, as well as the backwards arg declaration
of "from, to" instead of the more common "to, from" standard.]

Includes a slight change to the declaration of struct irq_desc to embed
the pending_mask within ifdef(CONFIG_SMP) to be consistent with other
references, and some small changes to Xen.

Tested: sparse/non-sparse/cpumask_offstack/non-cpumask_offstack/nonuma/nosmp on x86_64
Signed-off-by: NMike Travis <travis@sgi.com>
Cc: Chris Wright <chrisw@sous-sol.org>
Cc: Jeremy Fitzhardinge <jeremy@xensource.com>
Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>
Cc: virtualization@lists.osdl.org
Cc: xen-devel@lists.xensource.com
Cc: Yinghai Lu <yhlu.kernel@gmail.com>

In a discussio with Jeff Garzik, he mentioned that the serialization
for the libata port probes only needs to be within the domain of a host.
This means that for the first port of each host (with ID 0), we don't
need to wait, so we can relax our serialization a little.
Signed-off-by: NArjan van de Ven <arjan@linux.intel.com>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

This patch adds a per host flag that allows drivers to opt in into
having its busses scanned in parallel.

Drivers that do not set this flag get their ports scanned in
the "original" sequence.
Signed-off-by: NArjan van de Ven <arjan@linux.intel.com>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

The sensor can be accessed via various buses.  In particular, SPI, I²C
and, on HP laptops, via a specific ACPI API (the only one currently
supported).  Separate this latest platform from the core of the sensor
driver to allow support for the other bus type.  The second, and more
direct goal is actually to be able to merge this part with the
hp-disk-leds driver, which has the same ACPI PNP number.
Signed-off-by: NPavel Machek <pavel@suse.cz>
Signed-off-by: NEric Piel <eric.piel@tremplin-utc.net>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

The code was shifting the endianness appropriately everywhere, annotate
the structs to avoid the sparse warnings when assigning the endian types
to the struct members, or passing them to be[16|32]_to_cpu:

drivers/memstick/core/mspro_block.c:331:4: warning: cast to restricted __be16
drivers/memstick/core/mspro_block.c:333:4: warning: cast to restricted __be16
drivers/memstick/core/mspro_block.c:335:4: warning: cast to restricted __be16
drivers/memstick/core/mspro_block.c:337:4: warning: cast to restricted __be16
drivers/memstick/core/mspro_block.c:341:4: warning: cast to restricted __be16
drivers/memstick/core/mspro_block.c:347:4: warning: cast to restricted __be32
drivers/memstick/core/mspro_block.c:356:4: warning: cast to restricted __be16
drivers/memstick/core/mspro_block.c:358:4: warning: cast to restricted __be16
drivers/memstick/core/mspro_block.c:364:4: warning: cast to restricted __be16
drivers/memstick/core/mspro_block.c:367:4: warning: cast to restricted __be16
drivers/memstick/core/mspro_block.c:369:4: warning: cast to restricted __be16
drivers/memstick/core/mspro_block.c:371:4: warning: cast to restricted __be16
drivers/memstick/core/mspro_block.c:377:4: warning: cast to restricted __be16
drivers/memstick/core/mspro_block.c:478:4: warning: cast to restricted __be16
drivers/memstick/core/mspro_block.c:480:4: warning: cast to restricted __be16
drivers/memstick/core/mspro_block.c:482:4: warning: cast to restricted __be16
drivers/memstick/core/mspro_block.c:484:4: warning: cast to restricted __be16
drivers/memstick/core/mspro_block.c:486:4: warning: cast to restricted __be16
drivers/memstick/core/mspro_block.c:689:22:    expected unsigned int [unsigned] [assigned] data_address
drivers/memstick/core/mspro_block.c:689:22:    got restricted __be32 [usertype] <noident>
drivers/memstick/core/mspro_block.c:697:3: warning: cast to restricted __be32
drivers/memstick/core/mspro_block.c:960:17: warning: incorrect type in initializer (different base types)
drivers/memstick/core/mspro_block.c:960:17:    expected unsigned short [unsigned] data_count
drivers/memstick/core/mspro_block.c:960:17:    got restricted __be16 [usertype] <noident>
drivers/memstick/core/mspro_block.c:993:6: warning: cast to restricted __be16
drivers/memstick/core/mspro_block.c:995:28: warning: cast to restricted __be16
Signed-off-by: NHarvey Harrison <harvey.harrison@gmail.com>
Cc: Alex Dubov <oakad@yahoo.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Compilation of the HP WMI hotkeys code results in the following:

  CC [M]  drivers/platform/x86/hp-wmi.o
drivers/platform/x86/hp-wmi.c: In function hp_wmi_bios_setup:
drivers/platform/x86/hp-wmi.c:431: warning: ignoring return value of rfkill_register,
	 declared with attribute warn_unused_result
drivers/platform/x86/hp-wmi.c:441: warning: ignoring return value of rfkill_register,
	 declared with attribute warn_unused_result
drivers/platform/x86/hp-wmi.c:450: warning: ignoring return value of rfkill_register,
	 declared with attribute warn_unused_result
Signed-off-by: NLarry Finger <Larry.Finger@lwfinger.net>
Cc: Matthew Garrett <mjg59@srcf.ucam.org>
Cc: Len Brown <lenb@kernel.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

This reverts commit 2831fe6f.

Turns out that device_initialize shouldn't fail silently.
This series needs to be reworked in order to get into proper
shape.
Reported-by: NStefan Richter <stefanr@s5r6.in-berlin.de>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Cc: Kay Sievers <kay.sievers@vrfy.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

This reverts commit 11c3b5c3.

Turns out that device_initialize shouldn't fail silently.
This series needs to be reworked in order to get into proper
shape.
Reported-by: NStefan Richter <stefanr@s5r6.in-berlin.de>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Cc: Kay Sievers <kay.sievers@vrfy.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

This reverts commit 93e746db.

Turns out that device_initialize shouldn't fail silently.
This series needs to be reworked in order to get into proper
shape.
Reported-by: NStefan Richter <stefanr@s5r6.in-berlin.de>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Cc: Kay Sievers <kay.sievers@vrfy.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

This reverts commit b9daa99e.

Turns out that device_initialize shouldn't fail silently.
This series needs to be reworked in order to get into proper
shape.
Reported-by: NStefan Richter <stefanr@s5r6.in-berlin.de>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Cc: Kay Sievers <kay.sievers@vrfy.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Due to commit 2831fe6f, "driver core:
create a private portion of struct device", device_initialize() can no
longer be called from atomic contexts.

We now defer it until after config ROM probing.  This requires changes
to the bus manager code because this may use a device before it was
probed.
Reported-by: NJay Fenlason <fenlason@redhat.com>
Signed-off-by: NStefan Richter <stefanr@s5r6.in-berlin.de>

Enable support for USB ISDN TAs with Cologne Chip AG's
HFC-S USB ISDN Controller.
Signed-off-by: NMartin Bachem <m.bachem@gmx.de>
Signed-off-by: NKarsten Keil <kkeil@suse.de>

Fix kernel crash when doing hardware conference with more than two members.
Removed DTMF threshold notice when debugging is disabled.
Signed-off-by: NAndreas Eversberg <andreas@eversberg.eu>
Signed-off-by: NKarsten Keil <kkeil@suse.de>

create_l1() was missed when changing mode to TE.
Signed-off-by: NAndreas Eversberg <andreas@eversberg.eu>
Signed-off-by: NKarsten Keil <kkeil@suse.de>

Add missed table.
Signed-off-by: NMatthias Urlichs <smurf@smurf.noris.de>
Signed-off-by: NKarsten Keil <kkeil@suse.de>

Signed-off-by: NAndreas Eversberg <andreas@eversberg.eu>
Signed-off-by: NKarsten Keil <kkeil@suse.de>

Create /sys/class/mISDN and implement functions to handle
device renames.
Signed-Off-By: NMatthias Urlichs <matthias@urlichs.de>
Signed-off-by: NKarsten Keil <kkeil@suse.de>

Add missing release function of ec-devices. Each device require a relase
function now. All destruction (memory and list entry) must be done within
the given release function of device, rather than after unregistering device.
Signed-off-by: NAndreas Eversberg <andreas@eversberg.eu>
Signed-off-by: NKarsten Keil <kkeil@suse.de>

     - Poll-timer can now be set from 8 to 256 samples, depending on your kernel.
     - If default or 128 is used, the normal controller's clock is used as before.
       Usage: modprobe hfcpci poll=XXX
     - Added some debug code for dsp buffer size. (CMX_DELAY_DEBUG)
Signed-off-by: NAndreas Eversberg <andreas@eversberg.eu>
Signed-off-by: NKarsten Keil <kkeil@suse.de>

minor typo fix
compiler warning fix
Signed-off-by: NAndreas Eversberg <andreas@eversberg.eu>
Signed-off-by: NKarsten Keil <kkeil@suse.de>

Correct busy device detection.
This fix belongs to last commit.
Signed-off-by: NAndreas Eversberg <andreas@eversberg.eu>
Signed-off-by: NKarsten Keil <kkeil@suse.de>

 Added GETPEER operation.
 Socket now checks if device is already busy at a differen mode.
Signed-off-by: NAndreas Eversberg <andreas@eversberg.eu>
Signed-off-by: NKarsten Keil <kkeil@suse.de>

Fix false irq detection on disabled irqs.
Signed-off-by: NAndreas Eversberg <andreas@eversberg.eu>
Signed-off-by: NKarsten Keil <kkeil@suse.de>

Add ISDN sample clock API to mISDN core (new file clock.c)
hfcmulti and mISDNdsp use clock API.
Signed-off-by: NAndreas Eversberg <andreas@eversberg.eu>
Signed-off-by: NKarsten Keil <kkeil@suse.de>

Return error on D-channel access with E-channel data requested
Signed-off-by: NMartin Bachem <m.bachem@gmx.de>
Signed-off-by: NKarsten Keil <kkeil@suse.de>

New prim PH_DATA_E_IND.

 - all E-ch frames are indicated by recv_Echannel(), which pushes E-Channel
   frames into dch's rqueue
 - if dchannel is opened with channel nr 0, no E-Channel logging
   is requested
 - if dchannel is opened with channel nr 1, E-Channel logging
   is requested. if layer1 does not support that, -EINVAL
   in return is appropriate
Signed-off-by: NMartin Bachem <m.bachem@gmx.de>
Signed-off-by: NKarsten Keil <kkeil@suse.de>

Use protocol to detect D-channel not the channel number 0
Signed-off-by: NMartin Bachem <m.bachem@gmx.de>
Signed-off-by: NKarsten Keil <kkeil@suse.de>

Fix more indexing bugs  when checking free timeslots.
Signed-off-by: NAndreas Eversberg <andreas@eversberg.eu>
Signed-off-by: NKarsten Keil <kkeil@suse.de>

Add usefull info to debug output.
Signed-off-by: NPeter Schlaile <root@asterisk.schlaile.de>
Signed-off-by: NKarsten Keil <kkeil@suse.de>

Fix HDLC DSP transmit (DL_DATA frames were bounced back upwards instead of
being sent down as PH_DATA frames)

Thanks to Andreas Eversberg for the fix!
Signed-off-by: NAndreas Eversberg <andreas@eversberg.eu>
Signed-off-by: NPeter Schlaile <root@asterisk.schlaile.de>
Signed-off-by: NKarsten Keil <kkeil@suse.de>

struct device already has a 'name' member, use it.
Signed-off-by: NMatthias Urlichs <matthias@urlichs.de>
Signed-off-by: NKarsten Keil <kkeil@suse.de>

To get persistent device names with hotplug we need to rename devices
sometime.
Signed-off-by: NMatthias Urlichs <matthias@urlichs.de>
Signed-off-by: NKarsten Keil <kkeil@suse.de>

This is not needed.
Signed-off-by: NMatthias Urlichs <smurf@smurf.noris.de>
Signed-off-by: NKarsten Keil <kkeil@suse.de>

If received frames are not handled in time, purge them.
Signed-off-by: NAndreas Eversberg <andreas@eversberg.eu>
Signed-off-by: NKarsten Keil <kkeil@suse.de>

Add a new debug bit for clock problems.
Signed-off-by: NAndreas Eversberg <andreas@eversberg.eu>
Signed-off-by: NKarsten Keil <kkeil@suse.de>

This prevents underrun of fifo when filled and in case of an underrun it
prevents subsequent underruns due to jitter.
Improve dsp, so buffers are kept filled with a certain delay, so moderate
jitter will not cause underrun all the time -> the audio quality is highly
improved. tones are not interrupted by gaps anymore, except when CPU is
stalling or in high load.
Signed-off-by: NAndreas Eversberg <andreas@eversberg.eu>
Signed-off-by: NKarsten Keil <kkeil@suse.de>

Remove references to external zaptel modules and some
cosmetic cleanups.
Signed-off-by: NKarsten Keil <kkeil@suse.de>

Convert WARN_ON() on command issue/completion paths to WARN_ON_ONCE()
so that libata doesn't spam the machine even when one of those
conditions triggers repeatedly.
Signed-off-by: NTejun Heo <tj@kernel.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

MTD size is 64-bit now...

Pointed out by Randy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: NDavid Woodhouse <David.Woodhouse@intel.com>

[dwmw2: updated and made to still register whole device first]
Signed-off-by: NPhilip Rakity <pakity@yahoo.com>
Signed-off-by: NDavid Woodhouse <David.Woodhouse@intel.com>