Re: State of LDP3430 platform

[Russell King - ARM Linux <linux@arm.linux.org.uk>]

On Sat, Feb 12, 2011 at 04:02:16PM +0000, Russell King - ARM Linux wrote:
> Another LDP3430 report...
> 
> The LDP3430 seems to be getting there, but:
> 
> 1. LCD screen seems wrong.  The X display looks rather large, and
>    flickery - looks like the LCD timing parameters are wrong.  Some
>    text disappears off the RHS.
> 
>    fbset reports:
> 	mode "240x320-510"
> 		# D: 48.001 MHz, H: 168.424 kHz, V: 510.375 Hz
> 		geometry 240 320 240 320 16
> 		timings 20833 3 39 2 7 3 1
> 		accel false
> 		rgba 5/11,6/5,5/0,0/0
> 
>    kernel config:
> 	CONFIG_FB_OMAP=y
> 	CONFIG_FB_OMAP_CONSISTENT_DMA_SIZE=2
> 
>    with the supplied kernel:
> 	mode "640x480-64"
> 		# D: 21.601 MHz, H: 31.396 kHz, V: 63.813 Hz
> 		geometry 640 480 640 640 16
> 		timings 46295 40 4 8 2 4 2
> 		accel false
> 		rgba 5/11,6/5,5/0,0/0
>    so the LCD timings in the kernel appear to be wrong for the panel
>    on the LDP.  What is 'CONFIG_FB_OMAP_LCD_VGA'?  There's *no* help
>    for this configuration option so god only knows what it's right
>    setting should be.  Please give it some help text to explain what
>    it is and what it does.

LCD works better with FB_OMAP_LCD_VGA set, but why is this necessary?

> 2. Keyboard - numeric keys produce wrong ascii for their labelled
>    function.  This is from /dev/tty1 with X running:
> 
> 	1 gives nothing	2 gives 4	3 gives 7
> 	4 gives 2	5 gives 5	6 gives 8
> 	7 gives 3	8 gives nothing	9 gives 9
> 	* gives nothing	0 gives E	# gives nothing
> 
>    off-hook (green phone) gives 0
>    on-hook (red phone) gives <1b><5b><31><39><7e>
> 
>    Killing X and then running evtest on /dev/input/event1 doesn't
>    return any events, neither does reading /dev/tty1 for the twl4030
>    keypad - the numeric keys are completely dead.  /proc/interrupts
>    shows no new interrupt counts when pressing the key for
>    'twl4030_keypad'.  Unbinding/rebinding the twl4030 driver doesn't
>    sort it out, neither does restarting X.

3. Touchscreen is dead, presumably because no devices are bound to
   ads7846 SPI driver.

ls -al /sys/bus/spi/devices/
total 0
drwxr-xr-x  2 root root 0 Jan  1 01:02 .
drwxr-xr-x  4 root root 0 Jan  1 01:02 ..
lrwxrwxrwx  1 root root 0 Jan  1 01:02 spi1.0 -> ../../../devices/platform/omap2_mcspi.1/spi1.0

is the only SPI device which appears, but is unbound.  I'm sure this used
to work at some point.