Merge branch 'neuros' of ssh://git@git.neuros.com.cn/git/git-pub/osd20/linux-davinci-2.6 into neuros

Merge branch 'neuros' of ssh://git@git.neuros.com.cn/git/git-pub/osd20/linux-davinci-2.6 into neuros

auto genarate the uImage and install it into images.

Move IRQ workqueue init before request_irq, otherwise
host SUSPEND irq will call schedule_work() and BUG_ON()
will fire because musb_irq_work did not initialised.
Signed-off-by: Terry.Qiu <tqiu@neuros.com.cn>

Signed-off-by: Terry.Qiu <tqiu@neuros.com.cn>

Signed-off-by: Terry.Qiu <tqiu@neuros.com.cn>

Signed-off-by: Terry.Qiu <tqiu@neuros.com.cn>

Signed-off-by: Terry.Qiu <tqiu@neuros.com.cn>

Signed-off-by: Terry.Qiu <tqiu@neuros.com.cn>

- modified:   arch/arm/mach-davinci/Kconfig
  modified:   arch/arm/mach-davinci/Makefile
Signed-off-by: Terry.Qiu <tqiu@neuros.com.cn>

Signed-off-by: Terry.Qiu <tqiu@neuros.com.cn>

- modified:   drivers/char/davinci_resizer.c
  modified:   drivers/char/davinci_resizer_hw.c
  modified:   include/asm-arm/arch-davinci/davinci_resizer.h
Signed-off-by: Terry.Qiu <tqiu@neuros.com.cn>

- this does not build yet.
    modified:   drivers/char/Kconfig
    modified:   drivers/char/Makefile
    new file:   drivers/char/davinci_resizer.c
    new file:   drivers/char/davinci_resizer_hw.c
    new file:   include/asm-arm/arch-davinci/davinci_resizer.h
    new file:   include/asm-arm/arch-davinci/davinci_resizer_hw.h
Signed-off-by: Terry.Qiu <tqiu@neuros.com.cn>

Signed-off-by: Terry.Qiu <tqiu@neuros.com.cn>

Signed-off-by: Terry.Qiu <tqiu@neuros.com.cn>

This patch converts audio driver to use platform driver (and minor
style changes).

This fixes broken standby resume as reported in

http://linux.omap.com/pipermail/davinci-linux-open-source/2008-January/004919.html

Switching to platform_driver, the driver suspends and resumes properly
now. platform_resume_early() doesn't call uninitialized tasklet worker
function audio_dsr_handler() any more. From oops reported in link
above:

(audio_dsr_handler+0x0/0x7c) from [<c014f848>]
(platform_resume_early+0x30)
Signed-off-by: Dirk Behme <dirk.behme@gmail.com>
Signed-off-by: Kevin Hilman <khilman@mvista.com>

To: davinci-linux-open-source@linux.davincidsp.com
Date: Sat, 12 Jan 2008 07:39:48 +0100

This patch adds NAND driver to recent git. Tested on DM6446 DVEVM.

It is mainly based on Sander Huijsen <Shuijsen@optelecom-nkf.com> work

http://linux.omap.com/pipermail/davinci-linux-open-source/2007-December/004788.html

and HW ECC changes from Troy Kisky <troy.kisky@boundarydevices.com>

http://linux.omap.com/pipermail/davinci-linux-open-source/2008-January/004910.htmlSigned-off-by: Dirk Behme <dirk.behme@gmail.com>
Signed-off-by: Kevin Hilman <khilman@mvista.com>

To: davinci-linux-open-source@linux.davincidsp.com
Date: Sat, 12 Jan 2008 07:39:33 +0100

This patch exports davinci_mux_peripheral() to be used in
modules. Needed to compile NAND driver as module. Thanks to Kevin for
the hint!
Signed-off-by: Dirk Behme <dirk.behme@gmail.com>
Signed-off-by: Kevin Hilman <khilman@mvista.com>

Thanks to Andrew's detective work, the issue became rather obvious.
The PHY detection code looked at the LINK register instead of the
ALIVE register for detecting the presence of a PHY. Andrew confirmed
that this patch fixed his issue.
Signed-off-by: Bernard Blackham <bernard@largestprime.net>
Signed-off-by: Andrew Armstrong <andrewa@ovation.co.uk>
Signed-off-by: Kevin Hilman <khilman@mvista.com>

This is a forward port of the SD/MMC driver from the latest LSP,
melded into linux-2.6.23-git. I've tried to keep the style updates
that were merged recently. I've fixed the exit code (in
davinci_mmcsd_exit) to release clocks, and some failure paths in
davinci_mmcsd_init.

The driver as is performs ~6MB/s reads and ~1MB/s writes with a
block size of 512k (i.e. still needs help!). This patch just makes
the driver usable again.
Signed-off-by: Bernard Blackham <bernard@largestprime.net>
Signed-off-by: Kevin Hilman <khilman@mvista.com>

Prevent OTG peripheral trying to change to b_wait_acon if after
file storage CBW stall followed by SUSPEND from host. We switch
back to b_peripheral if RESET in b_wait_acon mode.

Also fix compile warning for missing default case.

And don't try to set b_wait_acon mode unless CONFIG_USB_OTG.
Signed-off-by: Tony Lindgren <tony@atomide.com>

this patch add the function prototype of set I/O to the header file os it
could be used outside since it is exported, but not defined in the header
previously
Signed-off-by: Luís Vitório Cargnini <lvcargnini@gmail.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>

The patch below fixes a few minor infelicities.
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>

Finish getting rid of dma_controller.private data, fixing a bug
related to the previous DaVinci private_data fix.  Also get rid
of a compiler warning in the 2430 dma code.
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>

Clean up OMAP2 Kconfig.  Again.

 - 34xx shouldn't go in the middle of the 24xx stuff, it
   goofs the dependency tree displayed by e.g. kconfig.
   In particular, not in the middle of the H4-specific
   options.

 - Make the 2420 boards depend on OMAP2420, and 2430 boards
   on OMAP2430 ... a 2420-only kernel shouldn't show 2430
   boards, and vice versa.
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Tony Lindgren <tony@atomide.com>

The GPT1 CLK SEL register was getting wrong src register, due to which GPT1 was not able to select SYS_CLK as input.
Signed-off-by: Syed Mohammed Khasim <x0khasim@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>

Fix compile if CONFIG_ARCH_OMAP_OTG is not set
Signed-off-by: Tony Lindgren <tony@atomide.com>

Allow suspend if host is already suspended
Signed-off-by: Tony Lindgren <tony@atomide.com>

Getting the current status of Vbus is useful when running
USB Compliance Tests.
Signed-off-by: Felipe Balbi <felipe.lima@indt.org.br>
Signed-off-by: Tony Lindgren <tony@atomide.com>

Sometimes we get both SUSPEND and DISCONNECT the same time.
In that case DISCONNECT must be ignored or else the HNP connection
will break right away.
Signed-off-by: Tony Lindgren <tony@atomide.com>

This seems like the only way to deal with interrupted HNP if waiting
as a_wait_bcon.
Signed-off-by: Tony Lindgren <tony@atomide.com>

VBUS must be on all the time in this case.
Signed-off-by: Tony Lindgren <tony@atomide.com>