linux-2.6
14 years ago[S390] qdio: leave inbound SBALs primed
Jan Glauber [Mon, 22 Jun 2009 10:08:14 +0000 (12:08 +0200)] 
[S390] qdio: leave inbound SBALs primed

It is not required to change the state of primed SBALs. Leaving them
primed saves a SQBS instruction under z/VM.

Signed-off-by: Jan Glauber <jang@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
14 years ago[S390] qdio: merge AI tasklet into interrupt handler
Jan Glauber [Mon, 22 Jun 2009 10:08:13 +0000 (12:08 +0200)] 
[S390] qdio: merge AI tasklet into interrupt handler

Since the adapter interrupt tasklet only schedules the queue tasklets
and contains no code that requires serialization in can be merged
with the adapter interrupt handler. That possibly safes some CPU
cycles.

Signed-off-by: Jan Glauber <jang@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
14 years ago[S390] qdio: extract all primed SBALs at once
Jan Glauber [Mon, 22 Jun 2009 10:08:12 +0000 (12:08 +0200)] 
[S390] qdio: extract all primed SBALs at once

For devices without QIOASSIST primed SBALS were extracted in a loop.
Remove the loop since get_buf_states can already return more than
one primed SBAL.

Signed-off-by: Jan Glauber <jang@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
14 years ago[S390] qdio: fix check for running under z/VM
Jan Glauber [Mon, 22 Jun 2009 10:08:11 +0000 (12:08 +0200)] 
[S390] qdio: fix check for running under z/VM

The check whether qdio runs under z/VM was incorrect since SIGA-Sync is not
set if the device runs with QIOASSIST. Use MACHINE_IS_VM instead to prevent
polling under z/VM.

Merge qdio_inbound_q_done and tiqdio_is_inbound_q_done.

Signed-off-by: Jan Glauber <jang@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
14 years ago[S390] qdio: move adapter interrupt tasklet code
Jan Glauber [Mon, 22 Jun 2009 10:08:10 +0000 (12:08 +0200)] 
[S390] qdio: move adapter interrupt tasklet code

Move the adapter interrupt tasklet function to the qdio main code
since all the functions used by the tasklet are located there.

Signed-off-by: Jan Glauber <jang@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
14 years ago[S390] Use del_timer instead of del_timer_sync
Michael Holzheu [Mon, 22 Jun 2009 10:08:09 +0000 (12:08 +0200)] 
[S390] Use del_timer instead of del_timer_sync

When syncing the sclp console queue, we call del_timer_sync() while holding
the "sclp_con_lock" spinlock. This lock is also taken in the timer function
"sclp_console_timeout". Therefore the sync version of del_timer() cannot be
used here. Because the synchronous deletion of the timer is only needed
in the suspend callback and in that case only one CPU is remaining and
therefore it is not possible that the timer function is running in parallel,
we can safely use del_timer() instead of del_timer_sync().

Signed-off-by: Michael Holzheu <holzheu@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
14 years ago[S390] s390: remove DEBUG_MALLOC
Pekka Enberg [Mon, 22 Jun 2009 10:08:08 +0000 (12:08 +0200)] 
[S390] s390: remove DEBUG_MALLOC

The kernel now has kmemleak and kmemtrace so there's no reason to keep
this ugly s390 hack around. I am not sure how it's supposed to work on
SMP anyway as it uses a global variable to temporarily store the return
value of all kmalloc() calls:

  void *b;

  #define kmalloc(x...) (PRINT_INFO(" kmalloc %p\n",b=kmalloc(x)),b)

Cc: <linux-s390@vger.kernel.org>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Pekka Enberg <penberg@cs.helsinki.fi>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
14 years ago[S390] vt220 console: convert from bootmem to slab
Heiko Carstens [Mon, 22 Jun 2009 10:08:07 +0000 (12:08 +0200)] 
[S390] vt220 console: convert from bootmem to slab

The slab allocator is earlier available so convert the
bootmem allocations to slab/gfp allocations.

Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
14 years ago[S390] sclp console: convert from bootmem to slab
Heiko Carstens [Mon, 22 Jun 2009 10:08:06 +0000 (12:08 +0200)] 
[S390] sclp console: convert from bootmem to slab

The slab allocator is earlier available so convert the
bootmem allocations to slab/gfp allocations.

Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
14 years ago[S390] 3270 console: convert from bootmem to slab
Heiko Carstens [Mon, 22 Jun 2009 10:08:05 +0000 (12:08 +0200)] 
[S390] 3270 console: convert from bootmem to slab

The slab allocator is earlier available so convert the
bootmem allocations to slab/gfp allocations.

Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
14 years ago[S390] 3215 console: convert from bootmem to slab
Heiko Carstens [Mon, 22 Jun 2009 10:08:04 +0000 (12:08 +0200)] 
[S390] 3215 console: convert from bootmem to slab

The slab allocator is earlier available so convert the
bootmem allocations to slab/gfp allocations.

Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
14 years ago[S390] time: convert from bootmem to slab
Heiko Carstens [Mon, 22 Jun 2009 10:08:03 +0000 (12:08 +0200)] 
[S390] time: convert from bootmem to slab

The slab allocator is earlier available so convert the
bootmem allocations to slab/gfp allocations.

Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
14 years agoxtensa: enable m41t80 driver in s6105_defconfig
Daniel Glockner [Wed, 10 Jun 2009 19:58:51 +0000 (12:58 -0700)] 
xtensa: enable m41t80 driver in s6105_defconfig

Signed-off-by: Daniel Glockner <dg@emlix.com>
Cc: David Brownell <david-b@pacbell.net>
Cc: Alessandro Zummo <a.zummo@towertech.it>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
14 years agoxtensa: add m41t62 rtc to s6105 platform
Daniel Glockner [Wed, 10 Jun 2009 19:58:50 +0000 (12:58 -0700)] 
xtensa: add m41t62 rtc to s6105 platform

Signed-off-by: Daniel Glockner <dg@emlix.com>
Cc: David Brownell <david-b@pacbell.net>
Cc: Alessandro Zummo <a.zummo@towertech.it>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
14 years agoxtensa: enable s6gmac in s6105_defconfig
Daniel Glockner [Wed, 10 Jun 2009 19:58:49 +0000 (12:58 -0700)] 
xtensa: enable s6gmac in s6105_defconfig

Signed-off-by: Daniel Glockner <dg@emlix.com>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Johannes Weiner <jw@emlix.com>
Cc: Oskar Schirmer <os@emlix.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
14 years agoxtensa: s6105 specific configuration for s6gmac
Oskar Schirmer [Wed, 10 Jun 2009 19:58:48 +0000 (12:58 -0700)] 
xtensa: s6105 specific configuration for s6gmac

Platform-specific configuration for the s6gmac driver, including the
PHY interrupt line.

Signed-off-by: Daniel Glockner <dg@emlix.com>
Signed-off-by: Oskar Schirmer <os@emlix.com>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Johannes Weiner <jw@emlix.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
14 years agos6gmac: xtensa s6000 on-chip ethernet driver
Oskar Schirmer [Wed, 10 Jun 2009 19:58:48 +0000 (12:58 -0700)] 
s6gmac: xtensa s6000 on-chip ethernet driver

The s6000 on-chip MAC supports 10/100/1000Mbit and is connected to an
external PHY via MII or RGMII interface.

[jw@emlix.com: don't use device->bus_id directly]
Signed-off-by: Oskar Schirmer <os@emlix.com>
Signed-off-by: Daniel Glockner <dg@emlix.com>
Acked-by: "David S. Miller" <davem@davemloft.net>
Signed-off-by: Johannes Weiner <jw@emlix.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Chris Zankel <chris@zankel.net>
14 years agoxtensa: support s6000 gpio irqs and alternate function selection
Daniel Glöckner [Tue, 5 May 2009 15:03:22 +0000 (15:03 +0000)] 
xtensa: support s6000 gpio irqs and alternate function selection

Implement an irq chip to handle interrupts via gpio.  The GPIO chip
initialization function now takes a bitmask denoting pins that should
be configured for their alternate function.

changes compared to v1:
- fixed bug on edge interrupt configuration
- accommodated to function name change
- moved definition of VARIANT_NR_IRQS to this patch
- renamed __XTENSA_S6000_IRQ_H to _XTENSA_S6000_IRQ_H as requested

Signed-off-by: Daniel Glöckner <dg@emlix.com>
Signed-off-by: Johannes Weiner <jw@emlix.com>
Signed-off-by: Chris Zankel <chris@zankel.net>
14 years agoxtensa: s6000 dma engine support
Oskar Schirmer [Wed, 10 Jun 2009 19:58:45 +0000 (12:58 -0700)] 
xtensa: s6000 dma engine support

There are four slightly different dma engines on the s6000 family.
One for memory-memory transfers, the other three for memory-device.

This patch implements a platform-specific kernel-API to control these
engines.  It is needed for the network, video, audio peripherals on
s6000.

Signed-off-by: Oskar Schirmer <os@emlix.com>
Signed-off-by: Daniel Glockner <dg@emlix.com>
Signed-off-by: Fabian Godehardt <fg@emlix.com>
Cc: Daniel Glockner <dg@emlix.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Chris Zankel <chris@zankel.net>
14 years agoxtensa: allow variant to initialize own irq chips
Daniel Glöckner [Tue, 5 May 2009 15:03:21 +0000 (15:03 +0000)] 
xtensa: allow variant to initialize own irq chips

There was already a PLATFORM_NR_IRQS define, which is now accompanied
by a VARIANT_NR_IRQS. To be able to initialize these interrupts,
init_IRQ now calls a variant specific hook.

Changes compared to v1:
- adapted to new CONFIG_VARIANT_IRQ_EXT
- removed definition and call of platform_init_IRQ as there already
  is a platform_init_irq defined in asm/platform.h with a weak default
  in kernel/platform.c
- renamed variant_init_IRQ to variant_init_irq

Note that I could not find the call site of platform_init_irq although
it is stated in platform.h that it is called from init_IRQ.

Signed-off-by: Daniel Glöckner <dg@emlix.com>
Signed-off-by: Chris Zankel <chris@zankel.net>
14 years agoxtensa: cache inquiry and unaligned cache handling functions
Oskar Schirmer [Wed, 10 Jun 2009 19:58:45 +0000 (12:58 -0700)] 
xtensa: cache inquiry and unaligned cache handling functions

The existing xtensa cache handling functions work on page-aligned
memory regions.

These functions are needed for the s6000 dma engine which can work on
a byte-granularity.

Signed-off-by: Oskar Schirmer <os@emlix.com>
Cc: Johannes Weiner <jw@emlix.com>
Cc: Daniel Glockner <dg@emlix.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Chris Zankel <chris@zankel.net>
14 years agosound: seq_midi_event: fix decoding of (N)RPN events
Clemens Ladisch [Mon, 22 Jun 2009 08:01:59 +0000 (10:01 +0200)] 
sound: seq_midi_event: fix decoding of (N)RPN events

When decoding (N)RPN sequencer events into raw MIDI commands, the
extra_decode_xrpn() function had accidentally swapped the MSB and LSB
controller values of both the parameter number and the data value.

Signed-off-by: Clemens Ladisch <clemens@ladisch.de>
Cc: <stable@kernel.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
14 years agoALSA: hda - Add digital-mic support to ALC262 auto model
Takashi Iwai [Mon, 22 Jun 2009 08:56:54 +0000 (10:56 +0200)] 
ALSA: hda - Add digital-mic support to ALC262 auto model

Add the digital-mic support with ALC262 auto model.
The new ALC262 models have the digital mic at NID 0x12.  This widget
isn't checked in the current alc262_auto_create_analog_input_ctls()
since it's under 0x18.  So, just reuse the routine for alc269 to fix
the behavior.

But, it doesn't suffice: the digital mic is supported only with the
ADC0, we have to exclude other ADCs when d-mic is detected.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
14 years agoALSA: hda - Fix check of input source type for realtek codecs
Takashi Iwai [Mon, 22 Jun 2009 08:50:19 +0000 (10:50 +0200)] 
ALSA: hda - Fix check of input source type for realtek codecs

Fix the check of the input-source type by checking the widget type of
each capture-source item.  Since some codecs can have both the mixer
and selector types depending on the ADC, alc_mux_enum_put() needs to
check each widget.

With this change, spec->capture_style gets unneeded, so it's removed,
too.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
14 years agoALSA: hda - Add quirk for Sony VAIO Z21MN
Takashi Iwai [Mon, 22 Jun 2009 09:03:13 +0000 (11:03 +0200)] 
ALSA: hda - Add quirk for Sony VAIO Z21MN

It needs model=toshiba-s06 to work with the digital-mic.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
Cc: <stable@kernel.org>
14 years agoALSA: hda - Get back Input Source for ALC262 toshiba-s06 model
Takashi Iwai [Mon, 22 Jun 2009 06:16:56 +0000 (08:16 +0200)] 
ALSA: hda - Get back Input Source for ALC262 toshiba-s06 model

The commit f9e336f65b666b8f1764d17e9b7c21c90748a37e
    ALSA: hda - Unify capture mixer creation in realtek codes
removed the "Input Source" mixer element creation for toshiba-s06 model
because it contains a digital-mic input.

This patch take the control back.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
Cc: <stable@kernel.org>
14 years agoALSA: hda - Fix unsigned comparison in patch_sigmatel.c
Takashi Iwai [Mon, 22 Jun 2009 06:00:10 +0000 (08:00 +0200)] 
ALSA: hda - Fix unsigned comparison in patch_sigmatel.c

Fix the comparison of unsigned int that causes a compile warning below
by changing to the right signed type:
  patch_sigmatel.c: In function ‘stac92xx_vref_set’:
  patch_sigmatel.c:658: warning: comparison of unsigned expression < 0 is always false

Signed-off-by: Takashi Iwai <tiwai@suse.de>
14 years agoALSA: via82xx: add option to disable 500ms delay in snd_via82xx_codec_wait
Simon Arlott [Sun, 21 Jun 2009 18:50:48 +0000 (19:50 +0100)] 
ALSA: via82xx: add option to disable 500ms delay in snd_via82xx_codec_wait

There's a large 500ms delay in snd_via82xx_codec_wait() that, at least
on my hardware, appears to be unnecessary. The rest of the init of
the card works without logging any warnings or errors and both audio
and mixer settings work.

This adds an "nodelay" parameter to disable this (undocumented in the
code) large delay improving bootup time by 489-500ms.

[    1.034217] initcall alsa_card_via82xx_init+0x0/0x16 returned 0 after 505757 usecs
vs.
[    0.533136] initcall alsa_card_via82xx_init+0x0/0x16 returned 0 after 15915 usecs

Signed-off-by: Simon Arlott <simon@fire.lp0.eu>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
14 years agosound: fix check for return value in snd_pcm_hw_refine
Mariusz Kozlowski [Sun, 21 Jun 2009 18:26:59 +0000 (20:26 +0200)] 
sound: fix check for return value in snd_pcm_hw_refine

'params' is a pointer and looking at the code this probably should be a check
for ioctl return value.

Signed-off-by: Mariusz Kozlowski <m.kozlowski@tuxland.pl>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
14 years agoide cmd64x: Remove serialize setting.
David S. Miller [Mon, 22 Jun 2009 05:48:03 +0000 (22:48 -0700)] 
ide cmd64x: Remove serialize setting.

This begins to fix regressions reported by Frans Pop on his Ultra-10.

There are still some funnies left that we are investigating.

Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agovia-velocity: Fix velocity driver unmapping incorrect size.
Dave Jones [Mon, 22 Jun 2009 05:42:30 +0000 (22:42 -0700)] 
via-velocity: Fix velocity driver unmapping incorrect size.

When a packet is greater than ETH_ZLEN, we end up assigning the
boolean result of a comparison to the size we unmap.

Signed-off-by: Dave Jones <davej@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agoALSA: ctxfi - Allow unknown PCI SSIDs
Takashi Iwai [Mon, 22 Jun 2009 05:36:52 +0000 (07:36 +0200)] 
ALSA: ctxfi - Allow unknown PCI SSIDs

Allow unknown PCI SSIDs for emu20k1 and emu20k2 as "unknown" model.
Also, add a black-list check in case any device has to be listed
as "unsupported".  It has a negative value in the pci quirk entry.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
14 years agomlx4_en: Remove redundant refill code on RX
Yevgeny Petrilin [Sat, 20 Jun 2009 22:16:10 +0000 (22:16 +0000)] 
mlx4_en: Remove redundant refill code on RX

Our RX rings are always full, there is no need to check whether
we need to fill them or not. If we fail to allocate a new socket
buffer, the incoming packet is dropped an the ring remains full.

Signed-off-by: Yevgeny Petrilin <yevgenyp@mellanox.co.il>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agomlx4_en: Removed redundant check on lso header size
Yevgeny Petrilin [Sat, 20 Jun 2009 22:16:02 +0000 (22:16 +0000)] 
mlx4_en: Removed redundant check on lso header size

This check that verifies that the LSO header along with control
segment and first data segment do not cross 128 bytes is no longer
required.

Signed-off-by: Yevgeny Petrilin <yevgenyp@mellanox.co.il>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agomlx4_en: Cancel port_up check in transmit function
Yevgeny Petrilin [Sat, 20 Jun 2009 22:15:52 +0000 (22:15 +0000)] 
mlx4_en: Cancel port_up check in transmit function

When closing the port, we stop all transmit queues under the transmit
lock. It ensures that we will not attempt to transmit new packets after
the physical port was closed.

Signed-off-by: Yevgeny Petrilin <yevgenyp@mellanox.co.il>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agomlx4_en: using stop/start_all_queues
Yevgeny Petrilin [Sat, 20 Jun 2009 22:15:46 +0000 (22:15 +0000)] 
mlx4_en: using stop/start_all_queues

After we moved to be a multi queue device, need to stop/start
all of our transmit queues.

Signed-off-by: Yevgeny Petrilin <yevgenyp@mellanox.co.il>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agomlx4_en: Removed redundant skb->len check
Yevgeny Petrilin [Sat, 20 Jun 2009 22:15:39 +0000 (22:15 +0000)] 
mlx4_en: Removed redundant skb->len check

We don't need this check in the transmit function

Signed-off-by: Yevgeny Petrilin <yevgenyp@mellanox.co.il>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agomlx4_en: Counting all the dropped packets on the TX side
Yevgeny Petrilin [Sat, 20 Jun 2009 22:15:31 +0000 (22:15 +0000)] 
mlx4_en: Counting all the dropped packets on the TX side

Reporting the counter's value through 'ethtool -S'

Signed-off-by: Yevgeny Petrilin <yevgenyp@mellanox.co.il>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agoi2c: Fix stuck transaction on cpm-i2c driver
Michael Trimarchi [Fri, 19 Jun 2009 12:50:02 +0000 (14:50 +0200)] 
i2c: Fix stuck transaction on cpm-i2c driver

When a process tries to read/write a disconnected i2c device, it receives a signal (e.g. ctrl-c) and the kernel gets stuck.

BUG: soft lockup - CPU#0 stuck for 61s! [I2CEEpromTest:392]
NIP: c01628f8 LR: c01628f0 CTR: c00177cc
REGS: c39abd70 TRAP: 0901   Not tainted  (2.6.25.7-alcore)
MSR: 00009032 <EE,ME,IR,DR>  CR: 42042048  XER: 20000000
TASK = c3889bd0[392] 'I2CEEpromTest' THREAD: c39aa000
GPR00: 00009000 c39abe20 c3889bd0 c39075c8 c39abe28 00000001 00000000 00000001
GPR08: c3889bd0 c39075c8 00009032 c39abe34 00002437
NIP [c01628f8] cpm_i2c_xfer+0x5fc/0x6d0
LR [c01628f0] cpm_i2c_xfer+0x5f4/0x6d0
Call Trace:
[c39abe20] [c0162924] cpm_i2c_xfer+0x628/0x6d0 (unreliable)
[c39abe90] [c015f6a0] i2c_transfer+0x88/0xb4
[c39abeb0] [c0160164] i2c_master_recv+0x48/0x6c
[c39abed0] [c01618dc] i2cdev_read+0x50/0xe4
[c39abef0] [c0068b24] vfs_read+0xc4/0x108
[c39abf10] [c0068f4c] sys_read+0x4c/0x90
[c39abf40] [c000d348] ret_from_syscall+0x0/0x38
Instruction dump:
3bc00064 92610010 3bf201c8 92810014 3b61

This happen because though the wait_event_interruptible_timeout takes the
signals into account, the driver does not handle them.
We propose to change the wait_event_interruptible_timeout with
wait_event_timeout, leaving the signals to be handled in other points
on the upper layers.

Signed-off-by: Bruno Morelli <bruno@evidence.eu.com>
Signed-off-by: Michael Trimarchi <michael@evidence.eu.com>
Acked-by: Jochen Friedrich <jochen@scram.de>
[ben-linux@fluff.org: fix title for patch]
Signed-off-by: Ben Dooks <ben-linux@fluff.org>
14 years agoi2c-omap: Fix build breaking typo cpu_is_omap_2430
Tony Lindgren [Wed, 17 Jun 2009 10:20:21 +0000 (03:20 -0700)] 
i2c-omap: Fix build breaking typo cpu_is_omap_2430

Hi Ben,

Can you please queue this fix?

Thanks,

Tony

>From ffe2b2cdf6283770b70a197e3748c6b40a1006be Mon Sep 17 00:00:00 2001
From: Tony Lindgren <tony@atomide.com>
Date: Wed, 17 Jun 2009 13:14:23 +0300
Subject: [PATCH] i2c-omap: Fix build breaking typo in cpu_is_omap_2430

Commit 84bf2c86 introduced a typo, it should be cpu_is_omap2430
instead. The typo was probably caused by a mismerge.

Without this patch all omaps fail to build with:
error: implicit declaration of function 'cpu_is_omap_2430'

Signed-off-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Ben Dooks <ben-linux@fluff.org>
14 years agoide: Take over as maintainer.
David S. Miller [Sun, 21 Jun 2009 23:11:33 +0000 (16:11 -0700)] 
ide: Take over as maintainer.

Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/drzeus/mmc
Linus Torvalds [Sun, 21 Jun 2009 20:14:22 +0000 (13:14 -0700)] 
Merge branch 'for-linus' of git://git./linux/kernel/git/drzeus/mmc

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/drzeus/mmc:
  sdhci: remove needless double parenthesis
  sdhci: Specific quirk vor VIA SDHCI controller in VX855ES
  s3cmci: fix dma configuration call
  mmc: Add new via-sdmmc host controller driver
  sdhci: Add support for hosts that are only capable of 1-bit transfers
  MAINTAINERS: add myself as atmel-mci maintainer (sd/mmc interface)
  sdhci: Add SDHCI_QUIRK_NO_MULTIBLOCK quirk
  sdhci: Add better ADMA error reporting
  sdhci-s3c: Samsung S3C based SDHCI controller glue

14 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6
Linus Torvalds [Sun, 21 Jun 2009 20:14:07 +0000 (13:14 -0700)] 
Merge git://git./linux/kernel/git/herbert/crypto-2.6

* git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6:
  crypto: aes-ni - Remove CRYPTO_TFM_REQ_MAY_SLEEP from fpu template
  crypto: aes-ni - Do not sleep when using the FPU
  crypto: aes-ni - Fix cbc mode IV saving
  crypto: padlock-aes - work around Nano CPU errata in CBC mode
  crypto: padlock-aes - work around Nano CPU errata in ECB mode

14 years agoMerge branch 'core-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Sun, 21 Jun 2009 20:13:53 +0000 (13:13 -0700)] 
Merge branch 'core-fixes-for-linus' of git://git./linux/kernel/git/tip/linux-2.6-tip

* 'core-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  lockdep: Select frame pointers on x86
  dma-debug: be more careful when building reference entries
  dma-debug: check for sg_call_ents in best-fit algorithm too

14 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound-2.6
Linus Torvalds [Sun, 21 Jun 2009 20:13:08 +0000 (13:13 -0700)] 
Merge branch 'for-linus' of git://git./linux/kernel/git/tiwai/sound-2.6

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound-2.6:
  ALSA: hda - Add model=6530g option
  ALSA: hda - Acer Inspire 6530G model for Realtek ALC888
  ALSA: snd_usb_caiaq: fix legacy input streaming
  ASoC: Kill BUS_ID_SIZE
  ALSA: HDA - Correct trivial typos in comments.
  ALSA: HDA - Name-fixes in code (tagra/targa)
  ALSA: HDA - Add pci-quirk for MSI MS-7350 motherboard.
  ALSA: hda - Fix memory leak at codec creation

14 years agoMove FAULT_FLAG_xyz into handle_mm_fault() callers
Linus Torvalds [Fri, 10 Apr 2009 16:01:23 +0000 (09:01 -0700)] 
Move FAULT_FLAG_xyz into handle_mm_fault() callers

This allows the callers to now pass down the full set of FAULT_FLAG_xyz
flags to handle_mm_fault().  All callers have been (mechanically)
converted to the new calling convention, there's almost certainly room
for architectures to clean up their code and then add FAULT_FLAG_RETRY
when that support is added.

Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
14 years agoRemove internal use of 'write_access' in mm/memory.c
Linus Torvalds [Fri, 10 Apr 2009 15:43:11 +0000 (08:43 -0700)] 
Remove internal use of 'write_access' in mm/memory.c

The fault handling routines really want more fine-grained flags than a
single "was it a write fault" boolean - the callers will want to set
flags like "you can return a retry error" etc.

And that's actually how the VM works internally, but right now the
top-level fault handling functions in mm/memory.c all pass just the
'write_access' boolean around.

This switches them over to pass around the FAULT_FLAG_xyzzy 'flags'
variable instead.  The 'write_access' calling convention still exists
for the exported 'handle_mm_fault()' function, but that is next.

Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
14 years agoipc: unbreak 32-bit shmctl/semctl/msgctl
Johannes Weiner [Sat, 20 Jun 2009 00:23:29 +0000 (02:23 +0200)] 
ipc: unbreak 32-bit shmctl/semctl/msgctl

31a985f "ipc: use __ARCH_WANT_IPC_PARSE_VERSION in ipc/util.h" would
choose the implementation of ipc_parse_version() based on a symbol
defined in <asm/unistd.h>.

But it failed to also include this header and thus broke
IPC_64-passing 32-bit userspace because the flag wasn't masked out
properly anymore and the command not understood.

Include <linux/unistd.h> to give the architecture a chance to ask for
the no-no-op ipc_parse_version().

Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
14 years agosdhci: remove needless double parenthesis
Pierre Ossman [Sun, 21 Jun 2009 18:59:33 +0000 (20:59 +0200)] 
sdhci: remove needless double parenthesis

Signed-off-by: Pierre Ossman <pierre@ossman.eu>
14 years agosdhci: Specific quirk vor VIA SDHCI controller in VX855ES
Harald Welte [Thu, 18 Jun 2009 14:53:38 +0000 (16:53 +0200)] 
sdhci: Specific quirk vor VIA SDHCI controller in VX855ES

The SDHCI controller found in the VX855ES requires 10ms
delay between applying power and applying clock.

This issue has been discovered and documented by the OLPC XO1.5 team.

Signed-off-by: Harald Welte <HaraldWelte@viatech.com>
Signed-off-by: Pierre Ossman <pierre@ossman.eu>
14 years agos3cmci: fix dma configuration call
Ben Dooks [Mon, 8 Jun 2009 22:33:56 +0000 (23:33 +0100)] 
s3cmci: fix dma configuration call

This was missed in the DMA changes during the s3c24xx
updates in commit 8970ef47d56fd3db28ee798b9d400caf08abd924.

Signed-off-by: Ben Dooks <ben@simtec.co.uk>
Signed-off-by: Pierre Ossman <pierre@ossman.eu>
14 years agommc: Add new via-sdmmc host controller driver
Harald Welte [Wed, 17 Jun 2009 18:22:39 +0000 (20:22 +0200)] 
mmc: Add new via-sdmmc host controller driver

This adds the via-sdmmc driver for the SD/MMC-controller of VIA,
which is found in a number of recent integrated VIA chipset
products.

Signed-off-by: Harald Welte <HaraldWelte@viatech.com>
Signed-off-by: Pierre Ossman <pierre@ossman.eu>
14 years agosdhci: Add support for hosts that are only capable of 1-bit transfers
Anton Vorontsov [Wed, 17 Jun 2009 20:14:08 +0000 (00:14 +0400)] 
sdhci: Add support for hosts that are only capable of 1-bit transfers

Some hosts (hardware configurations, or particular SD/MMC slots) may
not support 4-bit bus. For example, on MPC8569E-MDS boards we can
switch between serial (1-bit only) and nibble (4-bit) modes, thought
we have to disable more peripherals to work in 4-bit mode.

Along with some small core changes, this patch modifies sdhci-of
driver, so that now it looks for "sdhci,1-bit-only" property in the
device-tree, and if specified we enable a proper quirk.

Signed-off-by: Anton Vorontsov <avorontsov@ru.mvista.com>
Acked-by: Grant Likely <grant.likely@secretlab.ca>
Signed-off-by: Pierre Ossman <pierre@ossman.eu>
14 years agoMAINTAINERS: add myself as atmel-mci maintainer (sd/mmc interface)
Nicolas Ferre [Tue, 16 Jun 2009 11:05:50 +0000 (13:05 +0200)] 
MAINTAINERS: add myself as atmel-mci maintainer (sd/mmc interface)

Add MAINTAINERS entry for atmel-mci driver.
This driver was maintained by its author: Haavard Skinnemoen. I take the
maintainance of it.

Signed-off-by: Nicolas Ferre <nicolas.ferre@atmel.com>
Acked-by: Haavard Skinnemoen <haavard.skinnemoen@atmel.com>
Signed-off-by: Pierre Ossman <pierre@ossman.eu>
14 years agosdhci: Add SDHCI_QUIRK_NO_MULTIBLOCK quirk
Ben Dooks [Sun, 14 Jun 2009 11:40:53 +0000 (12:40 +0100)] 
sdhci: Add SDHCI_QUIRK_NO_MULTIBLOCK quirk

Add quirk to show the controller cannot do multi-block IO.

This is mainly for the Samsung SDHCI controller that currently
cannot manage to do multi-block PIO without timing out.

Signed-off-by: Ben Dooks <ben@simtec.co.uk>
Signed-off-by: Pierre Ossman <pierre@ossman.eu>
14 years agosdhci: Add better ADMA error reporting
Ben Dooks [Sun, 14 Jun 2009 12:52:38 +0000 (13:52 +0100)] 
sdhci: Add better ADMA error reporting

Update the ADMA error reporting to not only show the
overall controller state but also to print the ADMA
descriptor list.

Signed-off-by: Ben Dooks <ben@simtec.co.uk>
Signed-off-by: Pierre Ossman <pierre@ossman.eu>
14 years agosdhci-s3c: Samsung S3C based SDHCI controller glue
Ben Dooks [Sun, 14 Jun 2009 12:52:37 +0000 (13:52 +0100)] 
sdhci-s3c: Samsung S3C based SDHCI controller glue

Add support for the 'HSMMC' block(s) in the Samsung SoC
line. These are compatible with the SDHCI driver so add
the necessary setup and driver binding for the platform
devices.

Signed-off-by: Ben Dooks <ben@simtec.co.uk>
Signed-off-by: Pierre Ossman <pierre@ossman.eu>
14 years ago[SCSI] scsi_transport_fc: replace BUS_ID_SIZE by fixed count
James Bottomley [Sun, 21 Jun 2009 17:11:43 +0000 (12:11 -0500)] 
[SCSI] scsi_transport_fc: replace BUS_ID_SIZE by fixed count

BUS_ID_SIZE is being removed from the kernel.

Cc: Kay Sievers <kay.sievers@vrfy.org>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
14 years agosd, sr: fix Driver 'sd' needs updating message
Hannes Reinecke [Thu, 18 Jun 2009 07:57:18 +0000 (09:57 +0200)] 
sd, sr: fix Driver 'sd' needs updating message

If a SCSI ULD driver sets blk_queue_prep_rq(), it should clean it
up itself on remove(), and not from the bus callbacks. This
removes the need to hook into bus->remove(), which should not
be used at the same time as driver->remove().

[jejb: fix sdkp initialisation problem due to mismerge]
Signed-off-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Kay Sievers <kay.sievers@vrfy.org>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
14 years agoscsi_transport_iscsi: return -EOVERFLOW for Too many iscsi targets
Jaswinder Singh Rajput [Sat, 20 Jun 2009 07:59:21 +0000 (13:29 +0530)] 
scsi_transport_iscsi: return -EOVERFLOW for Too many iscsi targets

setting err as -EOVERFLOW for Too many iscsi targets.

Also fixes a spurious compiler warning for gcc 4.3.3 and gcc 4.4 :

  CC      drivers/scsi/scsi_transport_iscsi.o
drivers/scsi/scsi_transport_iscsi.c: In function ‘iscsi_add_session’:
drivers/scsi/scsi_transport_iscsi.c:678: warning: ‘err’ may be used uninitialized in this function

Signed-off-by: Jaswinder Singh Rajput <jaswinderrajput@gmail.com>
Acked-by: Mike Christie <michaelc@cs.wisc.edu>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
14 years agofc_transport: Selective return value from BSG timeout function
Giridhar Malavali [Fri, 19 Jun 2009 23:26:54 +0000 (16:26 -0700)] 
fc_transport: Selective return value from BSG timeout function

The return value from BSG timout function should be based on the state of the
BSG job. This helps block layer to take selective actions to clean up BSG job.

Signed-off-by: Giridhar Malavali <giridhar.malavali@qlogic.com>
Acked-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
14 years agofc_transport: The softirq_done function registration for BSG request
Giridhar Malavali [Fri, 19 Jun 2009 23:26:53 +0000 (16:26 -0700)] 
fc_transport: The softirq_done function registration for BSG request

Registered the softirq_done function, since this is requried iby an request
using block level request timeout functionality. This function will be called
by the block layer as part of time out clean process to release the BSG
request.

Moved some of the BSG request completion activities to softirq_done routine to
take care of both normal and timout completions.

Signed-off-by: Giridhar Malavali <giridhar.malavali@qlogic.com>
Acked-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
14 years agosym53c8xx: ratelimit parity errors
John Stoffel [Fri, 19 Jun 2009 20:08:58 +0000 (16:08 -0400)] 
sym53c8xx: ratelimit parity errors

This makes a huge difference when you have a serial console on bootup to limit
these messages to a sane number.

Signed-off-by: John Stoffel <john@stoffel.org>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
14 years agoexplain the hidden scsi_wait_scan Kconfig variable
Stefan Richter [Thu, 18 Jun 2009 19:03:23 +0000 (21:03 +0200)] 
explain the hidden scsi_wait_scan Kconfig variable

People keep sending patches to expose CONFIG_SCSI_WAIT_SCAN as a tunable
item.  These patches aren't accepted upstream, so let's stop the ongoing
irritation of people due to the unconditionally installed module and its
Kconfig symbol.

Signed-off-by: Stefan Richter <stefanr@s5r6.in-berlin.de>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
14 years agoibmvfc: Fix endless PRLI loop in discovery
Brian King [Thu, 18 Jun 2009 14:06:55 +0000 (09:06 -0500)] 
ibmvfc: Fix endless PRLI loop in discovery

Fixes a problem seen where sending a PRLI to a target
resulted in it sending a LOGO. This caused the ibmvfc
driver to go back through discovery again, which caused
another PRLI attempt, which caused another LOGO. Fix this
behavior by ignoring LOGO if we haven't even logged into
the target yet.

Signed-off-by: Brian King <brking@linux.vnet.ibm.com>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
14 years agoibmvfc: Process async events before command responses
Brian King [Thu, 18 Jun 2009 14:06:52 +0000 (09:06 -0500)] 
ibmvfc: Process async events before command responses

Since async events could indicate changes to link status, or
events which could affect decisions made during discovery, we should
process async events prior to command completion responses.

Signed-off-by: Brian King <brking@linux.vnet.ibm.com>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
14 years agolibfc: Add runtime debugging with debug_logging module parameter
Robert Love [Wed, 10 Jun 2009 22:31:10 +0000 (15:31 -0700)] 
libfc: Add runtime debugging with debug_logging module parameter

This patch adds the /sys/module/libfc/parameters/debug_logging
file to sysfs as a module parameter. It accepts an integer
bitmask for logging. Currently it supports:

   bit
LSB 0 = general libfc debugging
    1 = lport debugging
    2 = disc debugging
    3 = rport debugging
    4 = fcp debugging
    5 = EM debugging
    6 = exch/seq debugging
    7 = scsi logging (mostly error handling)

the other bits are not used at this time.

The patch converts all of the libfc source files to use
these new macros and removes the old FC_DBG macro.

Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
14 years agolibfcoe: Add runtime debugging with module param debug_logging
Robert Love [Wed, 10 Jun 2009 22:31:05 +0000 (15:31 -0700)] 
libfcoe: Add runtime debugging with module param debug_logging

This patch adds a 'debug_logging' module parameter to
libfcoe.ko. It is an unsigned int that represents a bitmask of
available debug logging levels, each of which can be tuned at
runtime. Currently there are only two logging levels for this
module-

   bit
LSB 0 = libfcoe general logging
    1 = FIP logging

Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
14 years agofcoe: Add runtime debug logging with module parameter debug_logging
Robert Love [Wed, 10 Jun 2009 22:30:59 +0000 (15:30 -0700)] 
fcoe: Add runtime debug logging with module parameter debug_logging

This patch converts all FC_DBG statements to use new runtime tunable
debug macros. The fcoe.ko module now has a debug_logging module
parameter.

fcoe_debug_logging is an unsigned integer representing a bitmask of all
available logging levels. Currently only two logging levels are
supported-

   bit
LSB 0 = general fcoe logging
    1 = netdevice related logging

This patch also attempts to clean up some debug statement formatting
so it's more readable.

Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
14 years agoscsi_debug: Add support for physical block exponent and alignment
Martin K. Petersen [Fri, 15 May 2009 04:40:33 +0000 (00:40 -0400)] 
scsi_debug: Add support for physical block exponent and alignment

This patch adds support for setting the physical block exponent and
lowest aligned LBA in the READ CAPACITY(16) response.

The B0 VPD page is adjusted accordingly.

Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Acked-by: Douglas Gilbert <dgilbert@interlog.com>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
14 years agocnic: add NETDEV_1000 and NETDEVICES to Kconfig select
Randy Dunlap [Thu, 18 Jun 2009 16:55:17 +0000 (09:55 -0700)] 
cnic: add NETDEV_1000 and NETDEVICES to Kconfig select

NETDEVICES + NETDEV_1000 need to be enabled so that kconfig will check
those branches for selects and enforce "select UIO" under CNIC.
Otherwise the build fails with:

ERROR: "uio_unregister_device" [drivers/net/cnic.ko] undefined!
ERROR: "uio_event_notify" [drivers/net/cnic.ko] undefined!
ERROR: "__uio_register_device" [drivers/net/cnic.ko] undefined!

Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com>
Acked-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
14 years agocnic: Fix __symbol_get() build error.
Michael Chan [Sun, 14 Jun 2009 00:43:02 +0000 (17:43 -0700)] 
cnic: Fix __symbol_get() build error.

Ingo molnar <mingo@elte.hu> reported the error

drivers/net/cnic.c:2520: error: implicit declaration of function ‘__symbol_get’

when CONFIG_MODULES is not defined.  Fix by using symbol_get() instead.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
14 years agoRevert "[SCSI] cnic: fix error: implicit declaration of function ‘__symbol_get’"
James Bottomley [Sat, 13 Jun 2009 20:51:08 +0000 (15:51 -0500)] 
Revert "[SCSI] cnic: fix error: implicit declaration of function ‘__symbol_get’"

This reverts commit bc3bf8fd330ce981ce632a1a4a283eee46838f32.

All the commit did was add a second #include of <linux/module.h> which is
the wrong fix.

Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
14 years agoipr: differentiate pci-x and pci-e based adapters
Wayne Boyer [Wed, 17 Jun 2009 16:55:35 +0000 (09:55 -0700)] 
ipr: differentiate pci-x and pci-e based adapters

MSI has only been tested on and known to work with PCI-E based adapters.  This
patch adds a field to struct ipr_chip_t to indicate which type of interrupt to
use based on what is known about the chip.

Signed-off-by: Wayne Boyer <wayneb@linux.vnet.ibm.com>
Acked-by: Brian King <brking@linux.vnet.ibm.com>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
14 years agoipr: add test for MSI interrupt support
Wayne Boyer [Tue, 16 Jun 2009 22:13:28 +0000 (15:13 -0700)] 
ipr: add test for MSI interrupt support

The return value from pci_enable_msi() can not always be trusted.  This patch
adds code to generate an interrupt after MSI has been enabled and tests
whether or not we can receive and process it.  If the tests fails, then fall
back to LSI.

Signed-off-by: Wayne Boyer <wayneb@linux.vnet.ibm.com>
Acked-by: Brian King <brking@linux.vnet.ibm.com>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
14 years agoscsi_transport_spi: Blacklist Ultrium-3 tape for IU transfers
James Bottomley [Wed, 17 Jun 2009 19:05:05 +0000 (19:05 +0000)] 
scsi_transport_spi: Blacklist Ultrium-3 tape for IU transfers

There have been several bug reports which identified the Ultrium-3
tape as just hanging up on the bus during certain types of IU
transfer.  The identified culpret is type 0x02 (MULTIPLE COMMAND)
transfers.  The only way to prevent this tape wedging is to prevent it
from using IU transfers at all.  So this patch uses the exported
blacklist matching technology to recognise the drive and force it not
to use IU transfers.

Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
14 years agoscsi_transport_spi: use spi target settings instead of inquiry data for DV
James Bottomley [Wed, 17 Jun 2009 19:03:41 +0000 (15:03 -0400)] 
scsi_transport_spi: use spi target settings instead of inquiry data for DV

Right at the moment, we carefully set up the spi_support_xx in the
device configuration routines, but then we never actually use the
results: we rely on the inquiry strings.  If we're going to allow
overrides to the inquiry data, we have to rely on our own internal
settings.

Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
14 years agoenhance device info matching for multiple tables
James Bottomley [Wed, 17 Jun 2009 19:01:58 +0000 (15:01 -0400)] 
enhance device info matching for multiple tables

The current scsi_devinfo.c matching routines use a single table for
the global blacklist.  However, we're developing a need to blacklist
from specific transports too (notably some tape drives using SPI which
don't respond well to high speed protocols).  Instead of developing
separate blacklist matching for each transport class needing it,
enhance the current list matching to permit multiple lists.

Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
14 years agoqla2xxx: Update version number to 8.03.01-k4.
Andrew Vasquez [Wed, 17 Jun 2009 17:30:32 +0000 (10:30 -0700)] 
qla2xxx: Update version number to 8.03.01-k4.

Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
14 years agoqla2xxx: Correct (again) overflow during dump-processing on large-memory ISP23xx...
Andrew Vasquez [Wed, 17 Jun 2009 17:30:31 +0000 (10:30 -0700)] 
qla2xxx: Correct (again) overflow during dump-processing on large-memory ISP23xx parts.

Commit 7b867cf76fbcc8d77867cbec6f509f71dce8a98f ([SCSI] qla2xxx:
Refactor qla data structures) inadvertently reverted
e792121ec85672c1fa48f79d13986a3f4f56c590 ([SCSI] qla2xxx: Correct
overflow during dump-processing on large-memory ISP23xx parts.).

Cc: stable@kernel.org
Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
14 years agoqla2xxx: Limit querying to supported mailbox-registers while reading FW state.
Andrew Vasquez [Wed, 17 Jun 2009 17:30:30 +0000 (10:30 -0700)] 
qla2xxx: Limit querying to supported mailbox-registers while reading FW state.

Pre-ISP24xx chips have dedicated uses for mailbox 4 and 5 which
software should typically not query nor update.

Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
14 years agoqla2xxx: Correct iiDMA-update calling conventions.
Harish Zunjarrao [Wed, 17 Jun 2009 17:30:29 +0000 (10:30 -0700)] 
qla2xxx: Correct iiDMA-update calling conventions.

* To set iiDMA speeds for ISP81XX, bits 5-0 are used whereas for
  other older ISPs bits 2-0 are used.
* Pass proper VP index

Signed-off-by: Harish Zunjarrao <harish.zunjarrao@qlogic.com>
Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
14 years agoqla2xxx: Fixed a bug in number of response queue creation logic.
Anirban Chakraborty [Wed, 17 Jun 2009 17:30:28 +0000 (10:30 -0700)] 
qla2xxx: Fixed a bug in number of response queue creation logic.

Cc: stable@kernel.org
Signed-off-by: Anirban Chakraborty <anirban.chakraborty@qlogic.com>
Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
14 years agofix race that can give duplicate host number
Joe Eykholt [Tue, 16 Jun 2009 06:22:14 +0000 (23:22 -0700)] 
fix race that can give duplicate host number

Just once, two fcoe instances got the same host number
from scsi_add_host().

Use atomic_t and atomic_inc_return() to get next host number.
Subtract 1, so that scsi_host still starts with 0.

[jejb: added comment about unusual subtraction]
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
14 years agoblock: revert "bsg: setting rq->bio to NULL"
FUJITA Tomonori [Wed, 17 Jun 2009 06:10:11 +0000 (15:10 +0900)] 
block: revert "bsg: setting rq->bio to NULL"

The SMP handler (sas_smp_request) was fixed to use the block API
properly, so we don't need this workaround to avoid blk_put_request()
warning.

Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
14 years agoscsi_transport_sas: needs to call blk_end_request_all for SMP requests
FUJITA Tomonori [Wed, 17 Jun 2009 06:10:10 +0000 (15:10 +0900)] 
scsi_transport_sas: needs to call blk_end_request_all for SMP requests

We need to call blk_end_request_all to complete SMP requests properly.

Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
14 years agolibiscsi: add conn and scsi eh log debug flags
Erez Zilber [Tue, 16 Jun 2009 03:11:10 +0000 (22:11 -0500)] 
libiscsi: add conn and scsi eh log debug flags

Allow the user to control the debug logs in libiscsi. We will now
have a module param for connection, session & error handling.

[Mike Christie - Fixed up to compile on current code and added
missing ISCSI_DBG_EH conversions]

Signed-off-by: Erez Zilber <erezzi.list@gmail.com>
Signed-off-by: Mike Christie <michaelc@cs.wisc.edu>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
14 years agoiscsi_tcp: propogate EAGAIN from sendpage to libiscsi
Mike Christie [Tue, 16 Jun 2009 03:11:09 +0000 (22:11 -0500)] 
iscsi_tcp: propogate EAGAIN from sendpage to libiscsi

The net layer might return -EAGAIN because it could not
get space/mem within the sock sndtimeo or becuase the tcp/ip
connection was down. For the latter we do not want to retry
because the conn/session should just be shutdown and restarted.
libiscsi knows the state of the session recovery so propogate
this error to that layer. It will either do iscsi recovery
or have us retry the operation. Right now if we have partially
sent a pdu we would always retry the IO xmit slowing down
recovery.

Signed-off-by: Mike Christie <michaelc@cs.wisc.edu>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
14 years agolibiscsi: don't run scsi eh if iscsi task is making progress
Mike Christie [Tue, 16 Jun 2009 03:11:08 +0000 (22:11 -0500)] 
libiscsi: don't run scsi eh if iscsi task is making progress

If we are sending or receiving data for the task successfully do
not run the scsi eh, because we know the task is making progress.

Signed-off-by: Mike Christie <michaelc@cs.wisc.edu>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
14 years agocxgb3i: suppot of different kernel page sizes
Karen Xie [Mon, 15 Jun 2009 18:15:16 +0000 (11:15 -0700)] 
cxgb3i: suppot of different kernel page sizes

The default kernel pages supported are 4K, 8K, 16K, and 64K. Re-calculate
entries if PAGE_SIZE is not one of the defaults.

Signed-off-by: Karen Xie <kxie@chelsio.com>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
14 years agocxgb3i: use kref to track ddp usage
Karen Xie [Mon, 15 Jun 2009 18:15:16 +0000 (11:15 -0700)] 
cxgb3i: use kref to track ddp usage

The iscsi ddp functionality could be used by multiple iscsi entities,
add a refcnt to keep track of it, so we would not release it pre-maturely.

Signed-off-by: Karen Xie <kxie@chelsio.com>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
14 years agosd: Block limits VPD support
Martin K. Petersen [Sat, 23 May 2009 15:43:39 +0000 (11:43 -0400)] 
sd: Block limits VPD support

Query the block limits VPD page and adjust queue minimum and optimal I/O
sizes.

Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
14 years agosd: Detect non-rotational devices
Martin K. Petersen [Sat, 23 May 2009 15:43:38 +0000 (11:43 -0400)] 
sd: Detect non-rotational devices

Detect non-rotational devices and set the queue flag accordingly.

Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
14 years agosd: Physical block size and alignment support
Martin K. Petersen [Sat, 23 May 2009 15:43:37 +0000 (11:43 -0400)] 
sd: Physical block size and alignment support

Extract physical block size and lowest aligned LBA from READ
CAPACITY(16) response and adjust queue parameters.

Report physical block size and alignment when applicable.

[jejb: fix up trailing whitespace]
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
14 years agoMerge branch 'topic/hda' into for-linus
Takashi Iwai [Sun, 21 Jun 2009 08:59:12 +0000 (10:59 +0200)] 
Merge branch 'topic/hda' into for-linus

* topic/hda:
  ALSA: hda - Add model=6530g option
  ALSA: hda - Acer Inspire 6530G model for Realtek ALC888
  ALSA: HDA - Correct trivial typos in comments.
  ALSA: HDA - Name-fixes in code (tagra/targa)
  ALSA: HDA - Add pci-quirk for MSI MS-7350 motherboard.
  ALSA: hda - Fix memory leak at codec creation

14 years agoMerge branch 'topic/caiaq' into for-linus
Takashi Iwai [Sun, 21 Jun 2009 08:59:10 +0000 (10:59 +0200)] 
Merge branch 'topic/caiaq' into for-linus

* topic/caiaq:
  ALSA: snd_usb_caiaq: fix legacy input streaming

14 years agoMerge branch 'topic/asoc' into for-linus
Takashi Iwai [Sun, 21 Jun 2009 08:59:04 +0000 (10:59 +0200)] 
Merge branch 'topic/asoc' into for-linus

* topic/asoc:
  ASoC: Kill BUS_ID_SIZE

14 years agoALSA: hda - Add model=6530g option
Takashi Iwai [Sun, 21 Jun 2009 08:56:44 +0000 (10:56 +0200)] 
ALSA: hda - Add model=6530g option

Add the new model string corresponding to the previous Acer Aspire
6530G support.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
14 years agofirewire: new stack is no longer experimental
Stefan Richter [Tue, 16 Jun 2009 20:36:34 +0000 (22:36 +0200)] 
firewire: new stack is no longer experimental

The new stack is now recommended over the old one if used for industrial
video (IIDC/DCAM) or for storage devices (SBP-2) due to better
performance, improved compatibility, added features, and security.  It
should also be functionally on par with and is more secure than the old
ieee1394 stack in the use case of consumer video devices.

IP-over-1394 support for the new stack is currently emerging, and a
backend of the firedtv DVB driver to the new stack should be available
soon.

The one remaining area where the old stack is still required are audio
devices, as the new stack is not yet able to support the FFADO FireWire
audio framework.

Signed-off-by: Stefan Richter <stefanr@s5r6.in-berlin.de>
14 years agoALSA: hda - Acer Inspire 6530G model for Realtek ALC888
Tony Vroon [Sat, 20 Jun 2009 23:40:10 +0000 (00:40 +0100)] 
ALSA: hda - Acer Inspire 6530G model for Realtek ALC888

The selected 4930G model seemed to keep the subwoofer 'tuba'
function from operating correctly. Removing the existing PCI
ID match made this work again, but it was mapped to 'Side'
instead of to LFE as one would expect.
This attempts to enable all functionality and keep the amount
of available mixer sliders low. Any slider that had no audible
effect on the output audio has been removed, and as such EAPD
is not currently enabled.

Signed-off-by: Tony Vroon <tony@linx.net>
Signed-off-by: Takashi Iwai <tiwai@suse.de>