platform/upstream/kernel-adaptation-pc.git
13 years agommc: omap_hsmmc: if multiblock reads are broken, disable them
Paul Walmsley [Thu, 6 Oct 2011 20:50:35 +0000 (14:50 -0600)]
mmc: omap_hsmmc: if multiblock reads are broken, disable them

When device data indicates that multiple block reads are not supported
on a given HSMMC controller instance, log a message to the console, and
pass the appropriate MMC capability flag to the MMC core.

Signed-off-by: Paul Walmsley <paul@pwsan.com>
Cc: Dave Hylands <dhylands@gmail.com>
Tested-by: Steve Sakoman <sakoman@gmail.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: core: add workaround for controllers with broken multiblock reads
Paul Walmsley [Thu, 6 Oct 2011 20:50:33 +0000 (14:50 -0600)]
mmc: core: add workaround for controllers with broken multiblock reads

Due to hardware bugs, some MMC host controllers don't support
multiple-block reads[1].  To resolve, add a new MMC capability flag,
MMC_CAP2_NO_MULTI_READ, which can be set by affected host controller
drivers.  When this capability is set, all reads will be issued one
sector at a time.

1. See for example Advisory 2.1.1.128 "MMC: Multiple Block Read
Operation Issue" in _OMAP3530/3525/3515/3503 Silicon Errata_
Revision F (October 2010) (SPRZ278F), available from
http://focus.ti.com/lit/er/sprz278f/sprz278f.pdf

Signed-off-by: Paul Walmsley <paul@pwsan.com>
Cc: Dave Hylands <dhylands@gmail.com>
Tested-by: Steve Sakoman <sakoman@gmail.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: core: Prevent too long response times for suspend
Ulf Hansson [Thu, 13 Oct 2011 14:03:58 +0000 (16:03 +0200)]
mmc: core: Prevent too long response times for suspend

While trying to suspend the mmc host there could still be
ongoing requests that we need to wait for. At the same time
a device driver must respond to a suspend request rather quickly.

Instead of potentially wait "forever" by claiming the host we now
"try" to claim the host instead. If it fails, -EBUSY is returned.

Signed-off-by: Ulf Hansson <ulf.hansson@stericsson.com>
Reviewed-by: Sujit Reddy Thumma <sthumma@codeaurora.org>
Acked-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: recognise SDIO cards with SDIO_CCCR_REV 3.00
Bing Zhao [Fri, 21 Oct 2011 02:16:32 +0000 (19:16 -0700)]
mmc: recognise SDIO cards with SDIO_CCCR_REV 3.00

Table 6-2: CCCR bit Definitions, address 00h.  Part E1 SDIO Simplified
Specification Version 3.00, Feb. 25, 2011.

This patch has been tested with Marvell WLAN device SD8797.

Signed-off-by: Bing Zhao <bzhao@marvell.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: sd: Handle SD3.0 cards not supporting UHS-I bus speed mode
Subhash Jadavani [Wed, 10 Aug 2011 05:46:01 +0000 (11:16 +0530)]
mmc: sd: Handle SD3.0 cards not supporting UHS-I bus speed mode

Here is Essential conditions to indicate Version 3.00 Card
(SD_SPEC=2 and SD_SPEC3=1) :
(1) The card shall support CMD6
(2) The card shall support CMD8
(3) The card shall support CMD42
(4) User area capacity shall be up to 2GB (SDSC) or 32GB (SDHC)
    User area capacity shall be more than or equal to 32GB and
    up to 2TB (SDXC)
(5) Speed Class shall be supported (SDHC or SDXC)

So even if SD card doesn't support any of the newly defined
UHS-I bus speed mode, it can advertise itself as SD3.0 cards
as long as it supports all the essential conditions of
SD3.0 cards. Given this, these type of cards should atleast
run in High Speed mode @50MHZ if it supports HS.

But current initialization sequence for SD3.0 cards is
such that these non-UHS-I SD3.0 cards runs in Default
Speed mode @25MHz.

This patch makes sure that these non-UHS-I SD3.0 cards run
in High Speed Mode @50MHz.

Tested this patch with SanDisk Extreme SDHC 8GB Class 10 card.

Reported-by: "Hiremath, Vaibhav" <hvaibhav@ti.com>
Signed-off-by: Subhash Jadavani <subhashj@codeaurora.org>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: core: support HPI send command
Jaehoon Chung [Tue, 18 Oct 2011 05:26:42 +0000 (01:26 -0400)]
mmc: core: support HPI send command

HPI command is defined in eMMC4.41.
This feature is important for eMMC4.5 devices.

Signed-off-by: Jaehoon Chung <jh80.chung@samsung.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: core: Add cache control for eMMC4.5 device
Seungwon Jeon [Fri, 14 Oct 2011 05:03:21 +0000 (14:03 +0900)]
mmc: core: Add cache control for eMMC4.5 device

This patch adds cache feature of eMMC4.5 Spec.
If device supports cache capability, host can utilize some specific
operations.

Signed-off-by: Seungwon Jeon <tgih.jun@samsung.com>
Signed-off-by: Jaehoon Chung <jh80.chung@samsung.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: core: Modify the timeout value for writing power class
Seungwon Jeon [Tue, 18 Oct 2011 04:20:57 +0000 (13:20 +0900)]
mmc: core: Modify the timeout value for writing power class

This patch will apply the generic CMD6 timeout to switch command
for power class.

Signed-off-by: Seungwon Jeon <tgih.jun@samsung.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: core: new discard feature support at eMMC v4.5
Kyungmin Park [Tue, 18 Oct 2011 00:34:04 +0000 (09:34 +0900)]
mmc: core: new discard feature support at eMMC v4.5

MMC v4.5 supports the DISCARD feature (CMD38).  It's different from
trim and there's no check bit.  Currently it's only supported at v4.5.

Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Jaehoon Chung <jh80.chung@samsung.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: core: mmc sanitize feature support for v4.5
Kyungmin Park [Fri, 14 Oct 2011 05:15:48 +0000 (14:15 +0900)]
mmc: core: mmc sanitize feature support for v4.5

In the v4.5, there's no secure erase & trim support.
Instead it supports the sanitize feature.

Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Jaehoon Chung <jh80.chung@samsung.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: dw_mmc: modify DATA register offset
Jaehoon Chung [Mon, 17 Oct 2011 10:36:23 +0000 (19:36 +0900)]
mmc: dw_mmc: modify DATA register offset

In dw_mmc 2.40a spec, Data register's offset is changed.
Before we used Data register offset 0x100. but if somebody uses a
2.40a controller, we must use 0x200 for Data register.

This patch adds a version-id checking point and uses SDMMC_DATA(x)
instead of SDMMC_DATA.  It assumes 2.40a is the latest version.

Signed-off-by: Jaehoon Chung <jh80.chung@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Acked-by: James Hogan <james.hogan@imgtec.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: sdhci-pci: add flag for devices that can support runtime PM
Adrian Hunter [Mon, 17 Oct 2011 07:52:44 +0000 (10:52 +0300)]
mmc: sdhci-pci: add flag for devices that can support runtime PM

Only allow runtime PM for devices that specifically indicate
that they support it.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: omap_hsmmc: ensure pbias configuration is always done
Adrian Hunter [Fri, 6 May 2011 09:14:10 +0000 (12:14 +0300)]
mmc: omap_hsmmc: ensure pbias configuration is always done

Go through the driver's set_power() functions rather than
calling regulator_enable/disable() directly because otherwise
pbias configuration for MMC1 is not done.

Signed-off-by: Adrian Hunter <adrian.hunter@nokia.com>
Acked-by: Balaji T K <balajitk@ti.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: core: Add Power Off Notify Feature eMMC 4.5
Girish K S [Thu, 13 Oct 2011 06:34:16 +0000 (12:04 +0530)]
mmc: core: Add Power Off Notify Feature eMMC 4.5

This patch adds support for the power off notify feature, available in
eMMC 4.5 devices. If the host has support for this feature, then the
mmc core will notify the device by setting the POWER_OFF_NOTIFICATION
byte in the extended csd register with a value of 1 (POWER_ON).

For suspend mode short timeout is used, whereas for the normal poweroff
long timeout is used.

Signed-off-by: Girish K S <girish.shivananjappa@linaro.org>
Signed-off-by: Jaehoon Chung <jh80.chung@samsung.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: sdhci-s3c: fix potential NULL dereference
Jaehoon Chung [Wed, 12 Oct 2011 04:14:29 +0000 (13:14 +0900)]
mmc: sdhci-s3c: fix potential NULL dereference

sc->clk_bus[ptr] could be NULL.

Signed-off-by: Jaehoon Chung <jh80.chung@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: replace printk with appropriate display macro
Girish K S [Tue, 11 Oct 2011 06:14:09 +0000 (11:44 +0530)]
mmc: replace printk with appropriate display macro

All the files using printk function for displaying kernel messages
in the mmc driver have been replaced with corresponding macro.

Signed-off-by: Girish K S <girish.shivananjappa@linaro.org>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: core: Add default timeout value for CMD6
Seungwon Jeon [Fri, 23 Sep 2011 05:15:29 +0000 (14:15 +0900)]
mmc: core: Add default timeout value for CMD6

EXT_CSD[248] includes the default maximum timeout for CMD6.
This field is added at eMMC4.5 Spec. And it can be used for default
timeout except for some operations which don't define the timeout
(i.e. background operation, sanitize, flush cache) in eMMC4.5 Spec.

Signed-off-by: Seungwon Jeon <tgih.jun@samsung.com>
Signed-off-by: Jaehoon Chung <jh80.chung@samsung.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: sdhci-pci: add runtime pm support
Adrian Hunter [Mon, 3 Oct 2011 12:33:34 +0000 (15:33 +0300)]
mmc: sdhci-pci: add runtime pm support

Ths patch allows runtime PM for sdhci-pci, runtime suspending after
inactivity of 50ms and ensuring runtime resume before SDHC registers
are accessed.  During runtime suspend, interrupts are masked.
The host controller state is restored at runtime resume.

For Medfield, the host controller's card detect mechanism is
supplanted by an always-on GPIO which provides for card detect wake-up.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: core: move ->request() call from atomic context
Adrian Hunter [Mon, 3 Oct 2011 12:33:33 +0000 (15:33 +0300)]
mmc: core: move ->request() call from atomic context

mmc_request_done() is sometimes called from interrupt or other atomic
context.  Mostly all mmc_request_done() does is complete(), however it
contains code to retry on error, which uses ->request().  As the error
path is certainly not performance critical, this may be moved to the
waiting function mmc_wait_for_req_done().

This allows ->request() to use runtime PM get_sync() and guarantee it
is never in an atomic context.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Acked-by: Ulf Hansson <ulf.hansson@stericsson.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: Add module.h to drivers/mmc users assuming implicit presence.
Paul Gortmaker [Sun, 3 Jul 2011 19:15:51 +0000 (15:15 -0400)]
mmc: Add module.h to drivers/mmc users assuming implicit presence.

We are cleaning up the implicit presence of module.h; these guys are
some of the people who just assume it will be there.  Call it out
explitly for those that really need it.

Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: using module_param requires the inclusion of moduleparam.h
Stephen Rothwell [Sun, 9 Oct 2011 14:35:16 +0000 (10:35 -0400)]
mmc: using module_param requires the inclusion of moduleparam.h

Commit "mmc: add module param to set fault injection attributes" adds
a module_param to this file.  But it is relying on the old implicit
"module.h is everywhere" behaviour, and without the explicit include
of moduleparam.h, the pending module.h split up produces this error:

core/debugfs.c:28:35: error: expected ')' before numeric constant

Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: core: general purpose MMC partition support.
Namjae Jeon [Thu, 6 Oct 2011 14:41:38 +0000 (23:41 +0900)]
mmc: core: general purpose MMC partition support.

It allows gerneral purpose partitions in MMC Device.  And I try to simply
make mmc_blk_alloc_parts using mmc_part structure suggested by Andrei
Warkentin.  After patching, we see general purpose partitions like this:
> cat /proc/partitions
          179 0 847872 mmcblk0
          179 192 4096 mmcblk0gp3
          179 160 4096 mmcblk0gp2
          179 128 4096 mmcblk0gp1
          179 96  1052672 mmcblk0gp0
          179 64  1024 mmcblk0boot1
          179 32  1024 mmcblk0boot0

Signed-off-by: Namjae Jeon <linkinjeon@gmail.com>
Acked-by: Andrei Warkentin <awarkentin@vmware.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: core: ext_csd.raw_* used in comparison but never set
Andrei Warkentin [Sat, 24 Sep 2011 16:12:30 +0000 (12:12 -0400)]
mmc: core: ext_csd.raw_* used in comparison but never set

f39b2dd9d ("mmc: core: Bus width testing needs to handle suspend/resume")
added code to only compare read-only ext_csd fields in bus width testing
code, yet it's comparing some fields that are never set.

The affected fields are ext_csd.raw_erased_mem_count and
ext_csd.raw_partition_support.

Signed-off-by: Andrei Warkentin <andrey.warkentin@gmail.com>
Acked-by: Philip Rakity <prakity@marvell.com>
Cc: <stable@kernel.org>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: irq: Remove IRQF_DISABLED
Yong Zhang [Thu, 22 Sep 2011 08:59:04 +0000 (16:59 +0800)]
mmc: irq: Remove IRQF_DISABLED

Since commit [e58aa3d2: genirq: Run irq handlers with interrupts
disabled], we run all interrupt handlers with interrupts disabled
and we even check and yell when an interrupt handler returns with
interrupts enabled (see commit [b738a50a: genirq: Warn when handler
enables interrupts]).

So now this flag is a NOOP and can be removed.

Signed-off-by: Yong Zhang <yong.zhang0@gmail.com>
Acked-by: Kishore Kadiyala <kishore.kadiyala@ti.com>
Acked-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: block: support no access to boot partitions
Adrian Hunter [Fri, 23 Sep 2011 09:48:21 +0000 (12:48 +0300)]
mmc: block: support no access to boot partitions

Intel Medfield platform blocks access to eMMC boot partitions which
results in switch errors.  Since there is no access, mmcboot0/1
devices should not be created.  Add a host capability to reflect that.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: block: fix boot partition switch error path
Adrian Hunter [Fri, 23 Sep 2011 09:48:20 +0000 (12:48 +0300)]
mmc: block: fix boot partition switch error path

In the case of a switch error, do not update partition config as though
the switch succeeded, and ensure blk_end_request is called on the
failed request.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Acked-by: Andrei Warkentin <andrey.warkentin@gmail.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: core: eMMC 4.5 Power Class Selection Feature
Girish K S [Fri, 23 Sep 2011 15:11:47 +0000 (20:41 +0530)]
mmc: core: eMMC 4.5 Power Class Selection Feature

This patch adds the power class selection feature available for mmc
versions 4.0 and above.  During the enumeration stage before switching
to the lower data bus, check if the power class is supported for the
current bus width. If the power class is available then switch to the
power class and use the higher data bus. If power class is not supported
then switch to the lower data bus in a worst case.

Signed-off-by: Girish K S <girish.shivananjappa@linaro.org>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agofault-injection: update documentation with the mmc module param
Per Forlin [Tue, 13 Sep 2011 21:03:30 +0000 (23:03 +0200)]
fault-injection: update documentation with the mmc module param

Signed-off-by: Per Forlin <per.forlin@linaro.org>
Reviewed-by: Akinobu Mita <akinobu.mita@gmail.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: add module param to set fault injection attributes
Per Forlin [Tue, 13 Sep 2011 21:03:29 +0000 (23:03 +0200)]
mmc: add module param to set fault injection attributes

Replace setup("fail_mmc_request") and faulty "ifdef KERNEL" with
a simple module_param(). The module param mmc_core.fail_request
may be used to set the fault injection attributes during boot time
or module load time.

Signed-off-by: Per Forlin <per.forlin@linaro.org>
Reviewed-by: Akinobu Mita <akinobu.mita@gmail.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agofault-inject: export setup_fault_attr()
Per Forlin [Tue, 13 Sep 2011 21:03:28 +0000 (23:03 +0200)]
fault-inject: export setup_fault_attr()

mmc_core module needs to use setup_fault_attr() in order
to set fault injection attributes during module load time.

Signed-off-by: Per Forlin <per.forlin@linaro.org>
Reviewed-by: Akinobu Mita <akinobu.mita@gmail.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: sdhci-esdhc-imx: add basic imx6q usdhc support
Shawn Guo [Mon, 19 Sep 2011 09:32:21 +0000 (17:32 +0800)]
mmc: sdhci-esdhc-imx: add basic imx6q usdhc support

This patch adds the basic support for imx6q usdhc, which is a
derivative of esdhc controller.

Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: core: add sd uhs string for mmc_ios_show
Aaron Lu [Fri, 2 Sep 2011 08:06:08 +0000 (16:06 +0800)]
mmc: core: add sd uhs string for mmc_ios_show

This is a minor fix. It makes mmc_ios_show print proper string when the
host's timing is one of the newly added UHS-I modes.

Signed-off-by: Aaron Lu <aaron.lu@amd.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: sh_mmcif: simplify platform data
Guennadi Liakhovetski [Tue, 30 Aug 2011 16:26:39 +0000 (18:26 +0200)]
mmc: sh_mmcif: simplify platform data

Provide platforms with a simplified way to specify MMCIF DMA slave IDs in
a way, similar to SDHI and other sh_dma clients.

Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: block: add eMMC hardware reset support
Adrian Hunter [Mon, 29 Aug 2011 13:42:15 +0000 (16:42 +0300)]
mmc: block: add eMMC hardware reset support

For cards that support hardware reset (just eMMC), try a reset and
retry before returning an I/O error.  However this is not done for
ECC errors and is never done twice for the same operation type
(READ, WRITE, DISCARD, SECURE DISCARD) until that type of operation
again succeeds.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: mmc-test: add eMMC hardware reset test
Adrian Hunter [Mon, 29 Aug 2011 13:42:14 +0000 (16:42 +0300)]
mmc: mmc-test: add eMMC hardware reset test

MMC core provides a checking function that checks if the reset
has happended.  Add a test to use that function.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: sdhci-pci: add eMMC hardware reset support
Adrian Hunter [Mon, 29 Aug 2011 13:42:13 +0000 (16:42 +0300)]
mmc: sdhci-pci: add eMMC hardware reset support

Implement eMMC hardware reset for Medfield.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: sdhci: add eMMC hardware reset support
Adrian Hunter [Mon, 29 Aug 2011 13:42:12 +0000 (16:42 +0300)]
mmc: sdhci: add eMMC hardware reset support

Add an SDHCI operation for hardware reset and connect it to the
host controller operation.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: core: add eMMC hardware reset support
Adrian Hunter [Mon, 29 Aug 2011 13:42:11 +0000 (16:42 +0300)]
mmc: core: add eMMC hardware reset support

eMMC's may have a hardware reset line.  This patch provides a
host controller operation to implement hardware reset and
a function to reset and reinitialize the card.  Also, for MMC,
the reset is always performed before initialization.

The host must set the new host capability MMC_CAP_HW_RESET
to enable hardware reset.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: mmci: simplify err check in mmci_post_request
Per Forlin [Mon, 29 Aug 2011 13:35:59 +0000 (15:35 +0200)]
mmc: mmci: simplify err check in mmci_post_request

The error condition indicates that mmci_post_request() should cleanup
after the mmci_pre_request(). In this case the resources allocated by
device_prep_slave_sg() are freed by calling dmaengine_terminate_all().
dma_unmap_sg() should always be performed if the host_cookie is set.

Signed-off-by: Per Forlin <per.forlin@linaro.org>
Acked-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: core: clarify how to use post_req in case of errors
Per Forlin [Mon, 29 Aug 2011 13:35:58 +0000 (15:35 +0200)]
mmc: core: clarify how to use post_req in case of errors

The err condition in post_req() is set to undo a call made to pre_req()
that hasn't been started yet.  The err condition is not set if an MMC
request returns an error.

Signed-off-by: Per Forlin <per.forlin@linaro.org>
Acked-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: core: Set correct bus mode before card init
Stefan Nilsson XK [Thu, 15 Sep 2011 15:50:38 +0000 (17:50 +0200)]
mmc: core: Set correct bus mode before card init

Earlier all cards where initiated with bus mode set as OPENDRAIN, and then
later switched to PUSHPULL. According to the MMC/SD/SDIO specifications
only MMC cards use OPENDRAIN during init. For both SD and SDIO the bus
mode shall be PUSHPULL before attempting to init the card.

The consequence of having incorrect bus mode can lead to not being able
to detect the card. Therefore the default behavior have now been changed
to PUSHPULL in mmc_power_up, and will only be temporarily switched when
trying to attach or init a MMC card.

Signed-off-by: Stefan Nilsson XK <stefan.xk.nilsson@stericsson.com>
Signed-off-by: Ulf HANSSON <ulf.hansson@stericsson.com>
Acked-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: sdio: Workaround for dev with broken CMD53
Stefan Nilsson XK [Thu, 15 Sep 2011 15:43:04 +0000 (17:43 +0200)]
mmc: sdio: Workaround for dev with broken CMD53

Adds a quirk which can be turned on for SDIO devices that do not support
512 byte requests in byte mode during CMD53. These requests will always
be sent in block mode instead.

This patch also enables this quirk for ST-Ericsson CW1200 WLAN device.

Signed-off-by: Stefan Nilsson XK <stefan.xk.nilsson@stericsson.com>
Signed-off-by: Ulf HANSSON <ulf.hansson@stericsson.com>
Acked-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: sdhci-pxa: Check pdata before using its members
Tanmay Upadhyay [Wed, 14 Sep 2011 05:59:02 +0000 (11:29 +0530)]
mmc: sdhci-pxa: Check pdata before using its members

Signed-off-by: Tanmay Upadhyay <tanmay.upadhyay@einfochips.com>
Acked-by: Zhangfei Gao <zhangfei.gao@marvell.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: core: Fix hangs related to insert/remove of cards
Ulf Hansson [Wed, 21 Sep 2011 18:08:13 +0000 (14:08 -0400)]
mmc: core: Fix hangs related to insert/remove of cards

During a rescan operation mmc_attach(sd|mmc|sdio) functions are
called. The error handling in these function can trigger a detach
of the bus, which also meant a power off. This is not notified by
the rescan operation which then continues to the next attach function.

If a power off has been done, the framework must never send any
new commands to the host driver, without first doing a new power up.
This will most likely trigger any host driver to hang.

Moving power off out of detach and instead handle power off
separately when it is actually needed, solves the issue.

Signed-off-by: Ulf Hansson <ulf.hansson@stericsson.com>
Acked-by: Linus Walleij <linus.walleij@linaro.org>
Cc: <stable@kernel.org>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: sdhci-spear: Don't set power gpio to 1 on probe
Viresh Kumar [Mon, 12 Sep 2011 11:22:49 +0000 (16:52 +0530)]
mmc: sdhci-spear: Don't set power gpio to 1 on probe

Currently if card_power_gpio is passed from platform data, it is acquired
and its value is either set or reset. After that we overwrite it with 1,
which is not required.

So, this patch removes the extra line which sets its value.

Signed-off-by: Viresh Kumar <viresh.kumar@st.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: sdhci-of-esdhc: Access Freescale eSDHC registers as 32-bit
Xu lei [Fri, 9 Sep 2011 12:05:46 +0000 (20:05 +0800)]
mmc: sdhci-of-esdhc: Access Freescale eSDHC registers as 32-bit

Freescale eSDHC registers only support 32-bit accesses, this patch
ensures that all Freescale eSDHC register accesses are 32-bit.

Signed-off-by: Xu lei <B33228@freescale.com>
Signed-off-by: Roy Zang <tie-fei.zang@freescale.com>
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Acked-by: Anton Vorontsov <cbouatmailru@gmail.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: core: Put eMMC in Sleep mode before suspend
Balaji T K [Thu, 8 Sep 2011 16:38:39 +0000 (22:08 +0530)]
mmc: core: Put eMMC in Sleep mode before suspend

Put MMC to sleep if it supports SLEEP/AWAKE (CMD5) in the mmc suspend
so that Vcc (NAND core) can be cut to minimize power consumption.
eMMC put into SLEEP can respond to CMD0 or H/W reset or CMD5.
Current implemention on resume from suspend relies on CMD0 in
mmc_init_card to get out of SLEEP mode.

Signed-off-by: Balaji T K <balajitk@ti.com>
Acked-by: Venkatraman S <svenkatr@ti.com>
Reviewed-by: Subhash Jadavani <subhashj@codeaurora.org>
Acked-by: Adrian Hunter <adrian.hunter@intel.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: sdhci-s3c: add default controller configuration
Thomas Abraham [Wed, 14 Sep 2011 07:09:17 +0000 (12:39 +0530)]
mmc: sdhci-s3c: add default controller configuration

The default controller configuration which was previously setup by
platform helper functions is moved into the driver.

Cc: Ben Dooks <ben-linux@fluff.org>
Signed-off-by: Thomas Abraham <thomas.abraham@linaro.org>
Acked-by: Kukjin Kim <kgene.kim@samsung.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: core: add a short delay in mmc_power_off
Daniel Drake [Wed, 7 Sep 2011 09:22:09 +0000 (10:22 +0100)]
mmc: core: add a short delay in mmc_power_off

Stress-testing the runtime power management of libertas_sdio
through a rmmod/insmod loop revealed that it is quite easy to
cause an ETIMEDOUT failure in mmc_sdio_power_restore() leading to:
   libertas_sdio: probe of mmc1:0001:1 failed with error -16

Experimentation shows that a very short delay (100us) is needed in
the power down path before the card can be successfully booted again.
We know that this setup is lacking poweroff clamps on the card's power
lines, but as only a short delay is needed, apply this unconditionally.
Also bump up to 1ms sleep for extra legroom.

Signed-off-by: Daniel Drake <dsd@laptop.org>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: sdhci-tegra: Add 8-bit support to device tree binding.
Stephen Warren [Tue, 30 Aug 2011 19:17:16 +0000 (13:17 -0600)]
mmc: sdhci-tegra: Add 8-bit support to device tree binding.

The previous patch which implemented a DT binding for sdhci-tegra did not
allow all platform data fields to be initialized from DT. The following
were missing:

is_8bit: Implemented by this patch.

pm_flags: Not implemented yet. There are no mainline users of this field.
  I'm not quite sure what it's for, and hence how to represent this
  in DT; the value ends up being assigned to host->mmc->pm_caps.

While we're at it, fix the binding documentation to refer to "SD/MMC"
instead of "eSDHC", since that's the correct name; "eSDHC" was cut/paste
from the Freescale binding docs.

Signed-off-by: Stephen Warren <swarren@nvidia.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: dw_mmc: Support SDIO interrupts for all slots
Shashidhar Hiremath [Mon, 29 Aug 2011 07:41:46 +0000 (13:11 +0530)]
mmc: dw_mmc: Support SDIO interrupts for all slots

The Patch adds the support for SDIO interrupts for all slots.
It includes enabling of SDIO interrupts through dw_mci_enable_sdio_irq
and the handling of the slot specific interrupts in the Interrupt Service
Routine.

Signed-off-by: Shashidhar Hiremath <shashidharh@vayavyalabs.com>
Acked-by: James Hogan <james.hogan@imgtec.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: msm_sdcc: Handle dma resource not present case
Subhash Jadavani [Mon, 2 May 2011 12:40:40 +0000 (18:10 +0530)]
mmc: msm_sdcc: Handle dma resource not present case

If DMA resource is not available then SDCC driver
should atleast work in PIO data transfer mode.

Signed-off-by: Subhash Jadavani <subhashj@codeaurora.org>
Signed-off-by: Sahitya Tummala <stummala@codeaurora.org>
Signed-off-by: David Brown <davidb@codeaurora.org>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: msm_sdcc: Change initialization order of busclk_timer in probe
Sahitya Tummala [Mon, 2 May 2011 12:40:01 +0000 (18:10 +0530)]
mmc: msm_sdcc: Change initialization order of busclk_timer in probe

Intialize busclk_timer before it is accessed in probe.

Signed-off-by: Sahitya Tummala <stummala@codeaurora.org>
Signed-off-by: David Brown <davidb@codeaurora.org>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: msm_sdcc: Use MCI_INT_MASK0 for PIO interrupts
Sahitya Tummala [Mon, 2 May 2011 12:39:18 +0000 (18:09 +0530)]
mmc: msm_sdcc: Use MCI_INT_MASK0 for PIO interrupts

Not all targets have IRQ1 line routed from the SD controller to
the processor. So we cannot rely on IRQ1 for PIO interrupts.
This patch moves all PIO interrupts to IRQ0 and enables the PIO
mode.

Signed-off-by: Murali Palnati <palnatim@codeaurora.org>
Signed-off-by: Sahitya Tummala <stummala@codeaurora.org>
Signed-off-by: David Brown <davidb@codeaurora.org>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agomsm: mmc: Remove "pio_irq" resource
Sahitya Tummala [Mon, 2 May 2011 12:37:43 +0000 (18:07 +0530)]
msm: mmc: Remove "pio_irq" resource

On some targets, MCI_IRQ_MASK1 is not routed to the MSM in which
case only "cmd_irq" must be used even for PIO. With this change,
all the targets will use only "cmd_irq" for both CMD and PIO.

Signed-off-by: Sahitya Tummala <stummala@codeaurora.org>
Signed-off-by: David Brown <davidb@codeaurora.org>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: msm_sdcc: Enable SDC host->clk only after setting the rate.
Sahitya Tummala [Mon, 2 May 2011 12:37:01 +0000 (18:07 +0530)]
mmc: msm_sdcc: Enable SDC host->clk only after setting the rate.

For clocks that support rates which can be set (most clocks other
than _pclk AHB clocks), a rate must be set using clk_set_rate()
before the clock is enabled for the first time with clk_enable().
Subsequent calls to clk_enable() need not be preceded with the
clk_set_rate() calls unless we wish to change the clock rate that
is set previously.

SDC host->clk is currently enabled without setting the clock rate
even once. This patch fixes this, by ensuring that the clock rate
for this clock is first set before enabling the clock.

Signed-off-by: Murali Palnati <palnatim@codeaurora.org>
Signed-off-by: Sahitya Tummala <stummala@codeaurora.org>
Signed-off-by: David Brown <davidb@codeaurora.org>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: msm_sdcc: Handle error cases in probe
Sahitya Tummala [Mon, 2 May 2011 12:36:05 +0000 (18:06 +0530)]
mmc: msm_sdcc: Handle error cases in probe

Signed-off-by: Sahitya Tummala <stummala@codeaurora.org>
Signed-off-by: David Brown <davidb@codeaurora.org>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agomsm: Implement init_card operation for MSM SDCC
Alexander Tarasikov [Sun, 21 Aug 2011 11:52:44 +0000 (15:52 +0400)]
msm: Implement init_card operation for MSM SDCC

This allows boards with non-standard sdio cards to fill the CIS/CCCR data.
It is particularly important for old msm72k boards using wl1251.
Also drop the obsolete embedded_sdio_data structure from the header
as it was intended to surve a similiar purpose but was not implemented.

Signed-off-by: Alexander Tarasikov <alexander.tarasikov@gmail.com>
Acked-by: Sahitya Tummala <stummala@codeaurora.org>
[davidb: minor formatting cleanup]
Signed-off-by: David Brown <davidb@codeaurora.org>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: msm_sdcc: Fix a typo in MSM SDCC driver gpio setup
Alexander Tarasikov [Sun, 21 Aug 2011 11:52:43 +0000 (15:52 +0400)]
mmc: msm_sdcc: Fix a typo in MSM SDCC driver gpio setup

The use of && instead of || caused a NULL pointer dereference if
gpio setup was not passed via platform data

Signed-off-by: Alexander Tarasikov <alexander.tarasikov@gmail.com>
Acked-by: Sahitya Tummala <stummala@codeaurora.org>
Signed-off-by: David Brown <davidb@codeaurora.org>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agoARM: shmobile: ag5evm, ap4: Named SDHI IRQ sources
Simon Horman [Thu, 25 Aug 2011 01:27:28 +0000 (10:27 +0900)]
ARM: shmobile: ag5evm, ap4: Named SDHI IRQ sources

This allows specific (non-multiplexed) IRQ handlers to be used.

Cc: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
Cc: Paul Mundt <lethal@linux-sh.org>
Acked-by: Magnus Damm <magnus.damm@gmail.com>
Signed-off-by: Simon Horman <horms@verge.net.au>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: sdhi: Allow named IRQs to use specific handlers
Simon Horman [Fri, 26 Aug 2011 08:42:39 +0000 (10:42 +0200)]
mmc: sdhi: Allow named IRQs to use specific handlers

Allow named IRQs to use corresponding specific handlers. If named IRQs are
used, at least an "sdcard" IRQ has to be specified by the platform. If
names are not used, an arbitrary number of IRQs can be provided by the
platform, in which case the generic ISR will be used for each of them.

Cc: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
Acked-by: Magnus Damm <magnus.damm@gmail.com>
Signed-off-by: Simon Horman <horms@verge.net.au>
[g.liakhovetski@gmx.de: style and typo corrections, platform data check]
Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: tmio: Provide separate interrupt handlers
Simon Horman [Thu, 25 Aug 2011 01:27:26 +0000 (10:27 +0900)]
mmc: tmio: Provide separate interrupt handlers

Provide separate interrupt handlers which may be used by platforms where
SDHI has three interrupt sources.

This patch also removes the commented-out handling of CRC and other errors.

Cc: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
Acked-by: Magnus Damm <magnus.damm@gmail.com>
Signed-off-by: Simon Horman <horms@verge.net.au>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: tmio: Cache interrupt masks
Simon Horman [Thu, 25 Aug 2011 01:27:25 +0000 (10:27 +0900)]
mmc: tmio: Cache interrupt masks

This avoids the need to look up the masks each time an interrupt is handled.
As suggested by Guennadi.

Cc: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
Acked-by: Magnus Damm <magnus.damm@gmail.com>
Signed-off-by: Simon Horman <horms@verge.net.au>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: fix integer assignments to pointer
Venkatraman S [Wed, 24 Aug 2011 19:00:50 +0000 (00:30 +0530)]
mmc: fix integer assignments to pointer

Fix the sparse warning output "warning: Using plain integer as NULL pointer"

Signed-off-by: Venkatraman S <svenkatr@ti.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: queue: declare mmc_alloc_sg as static
Venkatraman S [Tue, 23 Aug 2011 15:46:02 +0000 (21:16 +0530)]
mmc: queue: declare mmc_alloc_sg as static

Fix the sparse warning "drivers/mmc/card/queue.c:111:20: warning:
symbol 'mmc_alloc_sg' was not declared. Should it be static?"

Signed-off-by: Venkatraman S <svenkatr@ti.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: Kconfig: remove i.MX individual SoC dependency
Sascha Hauer [Wed, 24 Aug 2011 06:41:08 +0000 (08:41 +0200)]
mmc: Kconfig: remove i.MX individual SoC dependency

The individual SoC dependency in Kconfig hardly scales anymore.
Instead of having such a fine grained dependency just depend
on ARCH_MXC and risk that the uninformed user has to look in
the help text to figure out which driver is the correct one.

Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: sdhci-tegra: Add Device Tree probing support
Grant Likely [Tue, 23 Aug 2011 18:15:33 +0000 (12:15 -0600)]
mmc: sdhci-tegra: Add Device Tree probing support

Add hooks to read gpio configuration out of the device tree node.

[grant.likely: Rewrite of original patch from John Bonesio]
Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
[swarren: Fixed tegra_sdhci_get_ro() to retrieve pdata correctly]
[swarren: Reworked to avoid #ifdef CONFIG_OF]
[swarren: Reworked binding based on fsl-imx-esdhc.txt]
[swarren: Documented binding]
Signed-off-by: Stephen Warren <swarren@nvidia.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: enable runtime PM by default
Daniel Drake [Wed, 20 Jul 2011 16:39:22 +0000 (17:39 +0100)]
mmc: enable runtime PM by default

Now that we have improved the runtime power management powerup/powerdown
code, we believe that MMC_CAP_POWER_OFF_CARD is no longer necessary:
runtime PM should now work everywhere.

The only hard evidence for introducing MMC_CAP_POWER_OFF_CARD was the
Marvell sd8686 wifi chip, which was believed to require external gpio
manipulation which wasn't supported by some boards.

After further investigation it was realized (and confirmed by Marvell
folks) that sd8686 requirements can be fulfilled by changing the reset
sequence itself, even if no external gpio is manipulated.

For further information, see the following thread:
http://www.mail-archive.com/linux-mmc@vger.kernel.org/msg04289.html

Enable this trivially for a release or two. If no problems are reported,
we will follow up with a more extensive patch to remove this flag
altogether. If problems are reported, we can look at whitelist/blacklist
possibilities as before.

Signed-off-by: Daniel Drake <dsd@laptop.org>
Acked-by: Ohad Ben-Cohen <ohad@wizery.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agofault-inject: add documentation on MMC IO fault injection
Per Forlin [Fri, 19 Aug 2011 12:52:38 +0000 (14:52 +0200)]
fault-inject: add documentation on MMC IO fault injection

Add description on how to enable random fault injection
for MMC IO.

Signed-off-by: Per Forlin <per.forlin@linaro.org>
Acked-by: Akinobu Mita <akinobu.mita@gmail.com>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: core: add random fault injection
Per Forlin [Fri, 19 Aug 2011 12:52:37 +0000 (14:52 +0200)]
mmc: core: add random fault injection

This adds support to inject data errors after a completed host transfer.
The mmc core will return error even though the host transfer is successful.
This simple fault injection proved to be very useful to test the
non-blocking error handling in the mmc_blk_issue_rw_rq().
Random faults can also test how the host driver handles pre_req()
and post_req() in case of errors.

Signed-off-by: Per Forlin <per.forlin@linaro.org>
Acked-by: Akinobu Mita <akinobu.mita@gmail.com>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agofault-inject: export fault injection functions
Per Forlin [Fri, 19 Aug 2011 12:52:36 +0000 (14:52 +0200)]
fault-inject: export fault injection functions

Export symbols should_fail() and fault_create_debugfs_attr() in order
to let modules utilize the fault injection framework.

Signed-off-by: Per Forlin <per.forlin@linaro.org>
Acked-by: Akinobu Mita <akinobu.mita@gmail.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: sdhci-esdhc-imx: Enable ADMA2
Richard Zhu [Thu, 11 Aug 2011 20:51:46 +0000 (16:51 -0400)]
mmc: sdhci-esdhc-imx: Enable ADMA2

Eanble the ADMA2 mode for freescale esdhc imx driver, tested on MX25
3DS board, MX51 BBG board and MX53 LOCO board.

This patch is only used to enable the ADMA2 for MX51/53 platforms.
MX25/35 can't support the ADMA2 mode, set BROKEN_ADMA quirk on
MX25/35 platforms.

The ADMA mode supported or not can be distinguished by bit 20 of
the Capability Register (offset 0x40) in the FSL eSDHC module.

Signed-off-by: Richard Zhu <richard.zhu@linaro.org>
Tested-and-acked-by: Eric Miao <eric.miao@linaro.org>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: at91_mci: remove the use of irq_to_gpio
Nicolas Ferre [Thu, 4 Aug 2011 15:49:03 +0000 (16:49 +0100)]
mmc: at91_mci: remove the use of irq_to_gpio

Remove the use of irq_to_gpio() in the card detection interrupt
handler. The information is available in the board structure and
we can avoid using a function that has little meaning.

Signed-off-by: Nicolas Ferre <nicolas.ferre@atmel.com>
Acked-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: atmel-mci: fix a potential issue about pending PDC interrupts
Ludovic Desroches [Thu, 11 Aug 2011 15:25:48 +0000 (15:25 +0000)]
mmc: atmel-mci: fix a potential issue about pending PDC interrupts

This patch fixes a potential issue about PDC interrupts. For example we
have a ENDRX pending interrupt and a RXBUFF pending interrupt. We have
received the RXBUFF interrupt but the transfer is not finished (so we
didn't have time to give a new buffer to the PDC controller). Then we
will compute ENDRX interrupt and we will give a new buffer to the PDC
controller, just after we will compute the RXBUFF interrupt and give
one or two new buffers to the PDC controller but we are not sure that
the first buffer given has been filled. So in this situation we may
have "lost" one sg buffer. It's the same for transmission.

Signed-off-by: Ludovic Desroches <ludovic.desroches@atmel.com>
Signed-off-by: Nicolas Ferre <nicolas.ferre@atmel.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: atmel-mci: correct sg buffer size evaluation
Ludovic Desroches [Thu, 11 Aug 2011 15:25:47 +0000 (15:25 +0000)]
mmc: atmel-mci: correct sg buffer size evaluation

Assuming that a sg buffer size is a page size is false so use sg_dma_len.
A 4096 bytes can be required with two 2048-bytes sg buffer.

Signed-off-by: Ludovic Desroches <ludovic.desroches@atmel.com>
Signed-off-by: Nicolas Ferre <nicolas.ferre@atmel.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: atmel-mci: use ATMEL_PDC_SCND_BUF_OFF instead of a literal value
Ludovic Desroches [Thu, 11 Aug 2011 15:25:46 +0000 (15:25 +0000)]
mmc: atmel-mci: use ATMEL_PDC_SCND_BUF_OFF instead of a literal value

Signed-off-by: Ludovic Desroches <ludovic.desroches@atmel.com>
Signed-off-by: Nicolas Ferre <nicolas.ferre@atmel.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: atmel-mci: change atmci_start_command to atmci_send_command
Ludovic Desroches [Thu, 11 Aug 2011 15:25:45 +0000 (15:25 +0000)]
mmc: atmel-mci: change atmci_start_command to atmci_send_command

Rename atmci_start_command() to atmci_send_command() which is more
appropriate; atmci_start_command suggests we're sending a start command.

Signed-off-by: Ludovic Desroches <ludovic.desroches@atmel.com>
Signed-off-by: Nicolas Ferre <nicolas.ferre@atmel.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: atmel-mci: add pdc support and runtime capabilities detection
Ludovic Desroches [Thu, 11 Aug 2011 15:25:44 +0000 (15:25 +0000)]
mmc: atmel-mci: add pdc support and runtime capabilities detection

Add pdc support for atmel-mci. It makes at91-mci driver useless because it
was only used for the old atmel MCI core which has pdc but no dma support.
To allow removing at91-mci, the capabilities of the MCI core are detected
at runtime -- then the driver will use pio, pdc or dma transfers.
Warning: at91rm9200 is not supported, to support it we need to use swab32
on data but I have no board to test it.

Signed-off-by: Ludovic Desroches <ludovic.desroches@atmel.com>
Signed-off-by: Nicolas Ferre <nicolas.ferre@atmel.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: atmel-mci: indentation
Ludovic Desroches [Thu, 11 Aug 2011 15:25:43 +0000 (15:25 +0000)]
mmc: atmel-mci: indentation

Signed-off-by: Ludovic Desroches <ludovic.desroches@atmel.com>
Signed-off-by: Nicolas Ferre <nicolas.ferre@atmel.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: atmel-mci: change atmci_readl and atmci_writel macros
Ludovic Desroches [Thu, 11 Aug 2011 15:25:42 +0000 (15:25 +0000)]
mmc: atmel-mci: change atmci_readl and atmci_writel macros

Change atmci_readl and atmci_writel macros: remove string concatenation.
We can use these macros with registers which are not prefixed by ATMCI_.
This is the case if we want to write PDC registers which are common to
several devices so they are not prefixed with ATMCI_.

Signed-off-by: Ludovic Desroches <ludovic.desroches@atmel.com>
Signed-off-by: Nicolas Ferre <nicolas.ferre@atmel.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agommc: atmel-mci: change namespace
Ludovic Desroches [Thu, 11 Aug 2011 15:25:41 +0000 (15:25 +0000)]
mmc: atmel-mci: change namespace

Homogenize namespace to atmci.

Signed-off-by: Ludovic Desroches <ludovic.desroches@atmel.com>
Signed-off-by: Nicolas Ferre <nicolas.ferre@atmel.com>
Signed-off-by: Chris Ball <cjb@laptop.org>
13 years agoMerge branch 'master' of git://git.infradead.org/users/linville/wireless-next into...
John W. Linville [Mon, 17 Oct 2011 19:05:26 +0000 (15:05 -0400)]
Merge branch 'master' of git://git.infradead.org/users/linville/wireless-next into for-davem

13 years agoigbvf: convert to ndo_fix_features
Michał Mirosław [Tue, 30 Aug 2011 17:07:11 +0000 (17:07 +0000)]
igbvf: convert to ndo_fix_features

Private rx_csum flags are now duplicate of netdev->features & NETIF_F_RXCSUM.
Removing this needs deeper surgery.

Things noticed:
 - HW VLAN acceleration probably can be toggled, but it's left as is
 - the resets on RX csum offload change can probably be avoided
 - there is A LOT of copy-and-pasted code here

Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
13 years agoigb: enable l4 timestamping for v2 event packets
Jacob Keller [Wed, 12 Oct 2011 00:51:54 +0000 (00:51 +0000)]
igb: enable l4 timestamping for v2 event packets

When enabling hardware timestamping for ptp v2 event packets, the
software does not setup the queue for l4 packets, although layer 4
packets are valid for v2. This patch adds the flag which enables
setting up a queue and enabling udp packet timestamping.

Signed-off-by: Jacob E Keller <jacob.e.keller@intel.com>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
13 years agoixgbe: Add new netdev op to turn spoof checking on or off per VF
Greg Rose [Thu, 29 Sep 2011 05:57:33 +0000 (05:57 +0000)]
ixgbe: Add new netdev op to turn spoof checking on or off per VF

Implements the new netdev op to allow user configuration of spoof
checking on a per VF basis.

V2 - Change netdev spoof check op setting to bool

Signed-off-by: Greg Rose <gregory.v.rose@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
13 years agoif_link: Add additional parameter to IFLA_VF_INFO for spoof checking
Greg Rose [Sat, 8 Oct 2011 03:05:24 +0000 (03:05 +0000)]
if_link: Add additional parameter to IFLA_VF_INFO for spoof checking

Add configuration setting for drivers to turn spoof checking on or off
for discrete VFs.

v2 - Fix indentation problem, wrap the ifla_vf_info structure in
     #ifdef __KERNEL__ to prevent user space from accessing and
     change function paramater for the spoof check setting netdev
     op from u8 to bool.
v3 - Preset spoof check setting to -1 so that user space tools such
     as ip can detect that the driver didn't report a spoofcheck
     setting.  Prevents incorrect display of spoof check settings
     for drivers that don't report it.

Signed-off-by: Greg Rose <gregory.v.rose@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
13 years agoe1000e: locking bug introduced by commit 67fd4fcb
Bruce Allan [Fri, 7 Oct 2011 03:50:38 +0000 (03:50 +0000)]
e1000e: locking bug introduced by commit 67fd4fcb

Commit 67fd4fcb (e1000e: convert to stats64) added the ability to update
statistics more accurately and on-demand through the net_device_ops
.ndo_get_stats64 hook, but introduced a locking bug on 82577/8/9 when
linked at half-duplex (seen on kernels with CONFIG_DEBUG_ATOMIC_SLEEP=y and
CONFIG_PROVE_LOCKING=y).  The commit introduced code paths that caused a
mutex to be locked in atomic contexts, e.g. an rcu_read_lock is held when
irqbalance reads the stats from /sys/class/net/ethX/statistics causing the
mutex to be locked to read the Phy half-duplex statistics registers.

The mutex was originally introduced to prevent concurrent accesses of
resources (the NVM and Phy) shared by the driver, firmware and hardware
a few years back when there was an issue with the NVM getting corrupted.
It was later split into two mutexes - one for the NVM and one for the Phy
when it was determined the NVM, unlike the Phy, should not be protected by
the software/firmware/hardware semaphore (arbitration of which is done in
part with the SWFLAG bit in the EXTCNF_CTRL register).  This latter
semaphore should be sufficient to prevent resource contention of the Phy in
the driver (i.e. the mutex for Phy accesses is not needed), but to be sure
the mutex is replaced with an atomic bit flag which will warn if any
contention is possible.

Also add additional debug output to help determine when the sw/fw/hw
semaphore is owned by the firmware or hardware.

Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Reported-by: Francois Romieu <romieu@fr.zoreil.com>
Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com>
13 years agolibertas: fix changing interface type when interface is down
Daniel Drake [Fri, 14 Oct 2011 11:05:26 +0000 (12:05 +0100)]
libertas: fix changing interface type when interface is down

The recent changes to only power the device when the interface up
introduced a bug: changing interface type, legal when the interface
is down, performs device I/O.

Fix this functionality by validating and recording the interface
type when the change is requested, but only applying the change
if/when the interface is brought up.

Signed-off-by: Daniel Drake <dsd@laptop.org>
Acked-by: Dan Williams <dcbw@redhat.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agomac80211: Provide station flags to cfg80211
Helmut Schaa [Thu, 13 Oct 2011 14:30:40 +0000 (16:30 +0200)]
mac80211: Provide station flags to cfg80211

Only station flags that are already defined in nl80211 are added for
now.

Signed-off-by: Helmut Schaa <helmut.schaa@googlemail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agonl80211: Add sta_flags to the station info
Helmut Schaa [Thu, 13 Oct 2011 14:30:39 +0000 (16:30 +0200)]
nl80211: Add sta_flags to the station info

Reuse the already existing struct nl80211_sta_flag_update to specify
both, a flag mask and the flag set itself. This means
nl80211_sta_flag_update is now used for setting station flags and also
for getting station flags.

Signed-off-by: Helmut Schaa <helmut.schaa@googlemail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agomac80211: fix TID for null poll response
Johannes Berg [Thu, 13 Oct 2011 11:19:19 +0000 (13:19 +0200)]
mac80211: fix TID for null poll response

The queue mapping/TID for non-QoS null data
responses to is never set, making it default
to BK. Fix that.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agoath9k: Rename AR9480 into AR9462
Rajkumar Manoharan [Thu, 13 Oct 2011 05:30:44 +0000 (11:00 +0530)]
ath9k: Rename AR9480 into AR9462

Renamed to be in sync with Marketing term and to avoid
confusion with other chip names.

Signed-off-by: Rajkumar Manoharan <rmanohar@qca.qualcomm.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agoath9k_hw: Rename AR9480 -> AR9462 initvals
Rajkumar Manoharan [Thu, 13 Oct 2011 05:30:43 +0000 (11:00 +0530)]
ath9k_hw: Rename AR9480 -> AR9462 initvals

The AR946/8x chips are 2x2 Dual band with BT support. In order
to avoid misleading with other chips and to be in sync with
marketing team's term, AR9480 is renamed as AR9462.

Signed-off-by: Rajkumar Manoharan <rmanohar@qca.qualcomm.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agoath9k_hw: Do fast channel change based on reusable calibration results
Rajkumar Manoharan [Thu, 13 Oct 2011 05:30:42 +0000 (11:00 +0530)]
ath9k_hw: Do fast channel change based on reusable calibration results

Support the fast channel change across band switch only when there
are available of reusable cabliration results. And also observed that
doing agc control calibration on fastcc, sometimes causing calibration
timeout. Hence changing agc control to be run only on full chip reset.

Signed-off-by: Rajkumar Manoharan <rmanohar@qca.qualcomm.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agoath9k_hw: Add radio retention support for AR9480
Rajkumar Manoharan [Thu, 13 Oct 2011 05:30:41 +0000 (11:00 +0530)]
ath9k_hw: Add radio retention support for AR9480

Supported calibrations of radio retention table (RTT) are
- DC offset
- Filter
- Peak detect

Signed-off-by: Rajkumar Manoharan <rmanohar@qca.qualcomm.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agoath9k_hw: Update normal/min noise floor value for AR9480
Rajkumar Manoharan [Thu, 13 Oct 2011 05:30:40 +0000 (11:00 +0530)]
ath9k_hw: Update normal/min noise floor value for AR9480

To improve sensitivity for AR9480, the normal and minimum
noise floor values of both bands are updated.

Signed-off-by: Rajkumar Manoharan <rmanohar@qca.qualcomm.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agoath9k_hw: Support fast channel change on 5GHz for AR9003 chips
Rajkumar Manoharan [Thu, 13 Oct 2011 05:30:39 +0000 (11:00 +0530)]
ath9k_hw: Support fast channel change on 5GHz for AR9003 chips

The commit "ath9k_hw: Improve fast channel change for AR9003 chips"
fixes the fast channel change issue for AR9003 chips that was
originally observed in AR9382 chip. Hence enabling fastcc support
again for 11A channel for AR9003 chips.

Cc: Paul Stewart <pstew@google.com>
Signed-off-by: Rajkumar Manoharan <rmanohar@qca.qualcomm.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agoath9k_hw: Cleanup Tx calibrations for AR9003 chips
Rajkumar Manoharan [Thu, 13 Oct 2011 05:30:38 +0000 (11:00 +0530)]
ath9k_hw: Cleanup Tx calibrations for AR9003 chips

Currently Tx IQ calibration is enabled by default for all AR9003
chips. But for AR9480, the calibration status should be read from
chip after processing ini. And also the carrier leak calibration
status is checked during init cal. As the init_cal is being called
for fast channel change too, the tx_cl status only be read after
full reset. Hence moving that into process ini function.

Signed-off-by: Rajkumar Manoharan <rmanohar@qca.qualcomm.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agoath9k_hw: Add support to reuse Carrier leak calibration
Rajkumar Manoharan [Thu, 13 Oct 2011 05:30:37 +0000 (11:00 +0530)]
ath9k_hw: Add support to reuse Carrier leak calibration

This patch adds support to reuse Carrier leak calibration
during fast channel change for AR9480 chips.

Signed-off-by: Rajkumar Manoharan <rmanohar@qca.qualcomm.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agoath9k_hw: Add support to reuse TxIQ cal measurements
Rajkumar Manoharan [Thu, 13 Oct 2011 05:30:36 +0000 (11:00 +0530)]
ath9k_hw: Add support to reuse TxIQ cal measurements

Pass an argument to decide whether to reuse the Tx IQ
calibration measurements or not during fast channel change.
This will be later used by MCI support for AR9480.

Signed-off-by: Rajkumar Manoharan <rmanohar@qca.qualcomm.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>