Reduce backlight brightness of device

[Manoj Gudi]

I understand that the usual techniques like:

#setpci -s args and

#echo level > /sys/class/backlight/acpi_video/brightness won't work

So, is there any simple solution to reduce brightness of device?

[Srikant]

i don't have aakash with me now..i remember we have controlled brightness some how..can't be sure

--
 
 

[Manoj Gudi]

okay in case if anyone's wondering what is happening...

> What am I doing?

I am trying make the side button (lock-screen) button work [Ideally it should suspend to RAM when its pressed],
Here's what I have figured out till now

1. I compiled a kernel with config:

      >>> CONFIG_SUSPEND (which enables suspend mode- pm-suspend) 

      >>> CONFIG_INPUT_EVDEV, CONFIG_EVBUG (which provides input devices debug information)

2. On booting this kernel

      >>> touchscreen fails

      >>> dmesg shows that lock-screen button is assigned as input0 and it can verfied by cat /dev/input/event0. Interestingly on pressing button, it doesnt throw a IRQ which ideally it should.

3. Since without IRQ a device driver is difficult to write, I decided to write a user-space event handler (i.e. on any event from input0, suspend the machine) which is easy.

 

4. I am able to suspend the machine... but its not waking up (the touchscreen is disabled, and the side-button doesnt throw any interrupt hence the device is not able to wake up)
 
5. I force reboot the device, and verify /var/log/pm-suspend.log and indeed suspension is successful

> Any solutions for wake up event?

[Srikant]

Hi, nice job though.
Why the lock button should act as suspend to RAM? Suppose I have some background tasks
like downloading, audio playback etc and I want those to continue with screen off. That's how it
works in android at least, the AOS doesn't suspend to RAM.
It would be great if you can map the lock button to screen off & on, similar to android,
that would save lot of power.
What others have to say ?

[Manoj Gudi]

On Jan 26, 2013 6:07 PM, "Srikant" <u.srikan...@gmail.com> wrote:
>
> don't send garbage to fb0, that won't help..find a way to set brightness to zero and kill backlight in sys
>
yes, but how is my question, there's no pci bus or acpi, i have to probe how android does it

>
> On Sat, Jan 26, 2013 at 6:02 PM, Manoj Gudi <manoj....@gmail.com> wrote:

>>
>>
>> On Jan 26, 2013 5:49 PM, "Srikant" <u.srikan...@gmail.com> wrote:
>> >
>> > Hi, nice job though.
>> > Why the lock button should act as suspend to RAM? Suppose I have some background tasks
>> > like downloading, audio playback etc and I want those to continue with screen off. That's how it
>> > works in android at least, the AOS doesn't suspend to RAM.
>> > It would be great if you can map the lock button to screen off & on, similar to android,
>>

>> okay that is easier than on_press_suspend. I just need to know how to increase or decrease brightness(my first post on this thread).
>>
>> out of vigour i tried-
>> cat /dev/null > /dev/fb0
>>
>> as expected it doesn't work.
>>
>> and
>> cat /dev/zero > /dev/fb0
>>
>> this manages to corrupt few pixels on entire screen; refresh screen and its back to normal.

[Srikant]

Yes, that issue is there, let me check how to control brightness, echo'ing zero to /sys/..is not working

[Pranav Salunke]

there could be another way - like giving the choice to the user wether to lock the screen or suspend the device... if Manoj has already implemented it so a small UI or settings option for the same can be provided.

[Manoj Gudi]

One of the way is to invoke screensaver on press of button

$ xset dpms force off

There are two problems associated with this -
1> backlight doesnt shut off, only screen does
2> The screen wakes up even with a slightest touch of any input

I believe this is what aakash does automatically after being idle for few mins..

[Srikant]

I have removed xscreensaver as it freezes display, yes the backlight is the issue, with screen off we need to cut power to backlight through software..i think its some problem with generic lcd driver/backlight driver

--
 
 

[Manoj Gudi]

any particular reason linux-Sunxi in androportal is not rebased to latest linux-sunxi ?

i can see some patches relevant like this

https://github.com/linux-sunxi/linux-sunxi/commit/207a75a99d1098df8ac8cbd4cfce4faa65840db7

[Srikant]

no particular reason, call it laziness, infact sachin is testing the latest one from sunxi only, not from our repo, we will rebase it 

[Manoj Gudi]

if size of repository is an issue, then i can help by cherry picking the relevant patches
of course i am not an expert in deciding relevancy.

[Srikant]

no, size is not an issue, as i told u its just laziness..will update to latest as soon as sachin tests it..in either case we will update

[isachin]

there is not much diff in the version, its only 3.0.42+ to 3.0.52+, ya but we are almost three months back. As we know .42+ was pretty stable to work with, we decide to stick with that. As soon as we r convinced with .52+, we will rebase it. But now out major goal is to directly jump to 3.7 as 3.7 was developed keeping embed systems in mind. Also the latest touchscreen patch(elan) is meant for 3.7(see this) . I tried to apply it to 3.0.42+ but no success :(

--
 
 

--
Sachin

[isachin]

sorry the kernel version is 3.0.57+ and NOT 3.0.52+

--
Sachin

[Manoj Gudi]

I have disabled system proxy by lxproxy and also checked /etc/apt/apt.conf, yet apt-get install tries to connect to 10.102...
there is no http_proxy variable in .bashrc  

solutions ?

[isachin]

also check .bashrc in /root/

--
Sachin

[Manoj Gudi]

not found

[Manoj Gudi]

anyway i modified my $http_proxy variable manually... but curious where its set.

[isachin]

to my experience, if you do 'sudo apt-get install'(as normal user) then it will got via 10.102.. irrespective whether you unset the proxy or not, but if you login as root and try 'apt-get' after unset'ing the proxy, it should wrk.

is this what's happening ?

--
Sachin

[Manoj Gudi]

On Sun, Jan 27, 2013 at 7:00 PM, isachin <iclco...@gmail.com> wrote:

to my experience, if you do 'sudo apt-get install'(as normal user) then it will got via 10.102.. irrespective whether you unset the proxy or not, but if you login as root and try 'apt-get' after unset'ing the proxy, it should wrk.

is this what's happening ?

yes, correct.

[Manoj Gudi]

minutes back I realized no matter how many times I change DISP_CMD_LCD_BRIGHTNESS in dev_disp.c, the screen wouldnt change, the dmesg (on modprobe lcd, modprobe backlight) shows lcd, backlight dont load correctly. Somebody recently raised this issue:

https://github.com/linux-sunxi/linux-sunxi/issues/113

we do get the same error no -17

[Manoj Gudi]

$ fbset -xres 0

shuts off screen <but the backlight is still on>

$ fbset -xres 800   <the original resolution>

turns screen back on to normal

I can map the lock-screen to toggle this, but again, is this acceptable ?

[Manoj Gudi]

The whole lcd module is responsible for backlight problem

This is the dmesg I get when modprobe lcd

[  362.420000] WARNING: at fs/sysfs/dir.c:455 sysfs_add_one+0x88/0xbc()

[  362.450000] sysfs: cannot create duplicate filename '/class/lcd'

[  362.450000] Modules linked in: lcd(+) sun4i_keyboard ft5x_ts rtl8192c_common rtlwifi 8192cu

[  362.450000] [<c003929c>] (unwind_backtrace+0x0/0x138) from [<c005b5f0>] (warn_slowpath_common+0x4c/0x64)

[  362.450000] [<c005b5f0>] (warn_slowpath_common+0x4c/0x64) from [<c005b69c>] (warn_slowpath_fmt+0x30/0x40)

[  362.450000] [<c005b69c>] (warn_slowpath_fmt+0x30/0x40) from [<c013a160>] (sysfs_add_one+0x88/0xbc)

[  362.450000] [<c013a160>] (sysfs_add_one+0x88/0xbc) from [<c013a1f4>] (create_dir+0x60/0xb4)

[  362.450000] [<c013a1f4>] (create_dir+0x60/0xb4) from [<c013a2f8>] (sysfs_create_dir+0x84/0xf0)

[  362.450000] [<c013a2f8>] (sysfs_create_dir+0x84/0xf0) from [<c02577d8>] (kobject_add_internal+0xb4/0x224)

[  362.450000] [<c02577d8>] (kobject_add_internal+0xb4/0x224) from [<c0257ee0>] (kset_register+0x20/0x3c)

[  362.450000] [<c0257ee0>] (kset_register+0x20/0x3c) from [<c02df66c>] (__class_register+0xc0/0x1d0)

[  362.450000] [<c02df66c>] (__class_register+0xc0/0x1d0) from [<c02df7c8>] (__class_create+0x4c/0x84)

[  362.450000] [<c02df7c8>] (__class_create+0x4c/0x84) from [<bf114018>] (lcd_class_init+0x18/0x5c [lcd])

[  362.450000] [<bf114018>] (lcd_class_init+0x18/0x5c [lcd]) from [<c002d378>] (do_one_initcall+0x34/0x174)

[  362.450000] [<c002d378>] (do_one_initcall+0x34/0x174) from [<c00889f4>] (sys_init_module+0x3b8/0x1a84)

[  362.450000] [<c00889f4>] (sys_init_module+0x3b8/0x1a84) from [<c0033700>] (ret_fast_syscall+0x0/0x30)

[  362.450000] ---[ end trace fe7d93df9a0c2eb6 ]---

[  362.450000] kobject_add_internal failed for lcd with -EEXIST, don't try to register things with the same name in the same directory.

[  362.450000] [<c003929c>] (unwind_backtrace+0x0/0x138) from [<c0257870>] (kobject_add_internal+0x14c/0x224)

[  362.450000] [<c0257870>] (kobject_add_internal+0x14c/0x224) from [<c0257ee0>] (kset_register+0x20/0x3c)

[  362.450000] [<c0257ee0>] (kset_register+0x20/0x3c) from [<c02df66c>] (__class_register+0xc0/0x1d0)

[  362.450000] [<c02df66c>] (__class_register+0xc0/0x1d0) from [<c02df7c8>] (__class_create+0x4c/0x84)

[  362.450000] [<c02df7c8>] (__class_create+0x4c/0x84) from [<bf114018>] (lcd_class_init+0x18/0x5c [lcd])

[  362.450000] [<bf114018>] (lcd_class_init+0x18/0x5c [lcd]) from [<c002d378>] (do_one_initcall+0x34/0x174)

[  362.450000] [<c002d378>] (do_one_initcall+0x34/0x174) from [<c00889f4>] (sys_init_module+0x3b8/0x1a84)

[  362.450000] [<c00889f4>] (sys_init_module+0x3b8/0x1a84) from [<c0033700>] (ret_fast_syscall+0x0/0x30)

[  362.450000] Unable to create backlight class; errno = -17

From my understanding there is some strange locking of /sys/class by sysfs.

Answer from here: https://github.com/linux-sunxi/linux-sunxi/issues/113 =>
"OK - PEBKAC (slang for user error) error...turns out LCD and HDMI were loaded in the kernel configuration (not as modules)."

So if there is error in our kernel configuration (as suggested above), where can be going wrong?

On Mon, Jan 28, 2013 at 9:26 PM, Manoj Gudi <manoj....@gmail.com> wrote:

its either that (just driver), or I have edited xmodmap keycode (after inserting module), and they are working...
not sure...
you can test on your device and let me know 

On Mon, Jan 28, 2013 at 9:20 PM, Srikant <u.srikan...@gmail.com> wrote:

wao.thts great :-)

On Mon, Jan 28, 2013 at 9:19 PM, Manoj Gudi <manoj....@gmail.com> wrote:

interesting, i accidentally modprobe sun4i-keyboard  module

it works increasing 3% vol per button_press

On Mon, Jan 28, 2013 at 9:17 PM, Srikant <u.srikan...@gmail.com> wrote:

no, what you are planning to map wiith? 

I think mapping volume would be straight choice.

On Mon, Jan 28, 2013 at 9:08 PM, Manoj Gudi <manoj....@gmail.com> wrote:

are volume up/down keys configured ?

On Mon, Jan 28, 2013 at 6:26 PM, Srikant <u.srikan...@gmail.com> wrote:

Let's keep this as unresolved issues for now, and continue with mapping h/w key for screen off/on

On Mon, Jan 28, 2013 at 5:11 PM, Manoj Gudi <manoj....@gmail.com> wrote:

I tried to build a low backlight value in kernel, which dint succeed due to error 17 <given in previous mail>

Also, if this did work, there is no way any user can change this value in kernel using user-space tools

I am falling short of leads, any thoughts ?

On Mon, Jan 28, 2013 at 5:02 PM, Manoj Gudi <manoj....@gmail.com> wrote:

nope

On Mon, Jan 28, 2013 at 4:51 PM, Srikant <u.srikan...@gmail.com> wrote:

so are we giving up backlight?

--
You received this message because you are subscribed to the Google Groups "Linux-on-Aakash" group.
To unsubscribe from this group and stop receiving emails from it, send an email to linux_on_aaka...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

[Manoj Gudi]

Anyway, keeping aside backlight issue, the lock-screen button mapping is figured out, also the volume keys work now(please update /etc/modules in the file-system image provided @github).

The only thing which bothers me is camera, the hardware info from an android app (given below) fails to show camera vendor, I assume its a generic camera

I already tested kernel with CONFIG_VIDEO_DEV (V4L2) module, but it failed to recognize any camera device. I presume its due to incorrectly/insufficiently configured I2C Hardware Bus support...

Any word of caution before I go wild and start randomly testing different i2c configs ?




__________________________________________________________________________________________

HARDWARE DETAILS
_________________________________________________________________________________________

Unknown F7

System (Android 4.0.3)

Model	F7
Manufacturer	Unknown
Device	Nuclear F7
Product	Nuclear_f7
Brand	Broncho
Android Version	4.0.3 (Ice Cream Sandwich)
API Level	15
Fingerprint	broncho/nuclear_f7/nuclear-f7:4.0.3/IML74K/20120504:eng/test-keys

Display (800x480)

Size	800x480
Refresh Rate	62.78 Hz
Default Orientation	Landscape
Physical Size	5.83" (14.8 cm)
Density	160 dpi (Medium)
Pixels Per Inch	X: 160; Y: 160.42

Processor (ARMv7)

Processor	ARMv7 Processor rev 2 (v7l)
Cores	1
Max Frequency	1008 Mhz
Instruction Sets	armeabi-v7a, armeabi
NEON Supported	Yes

Memory (354 MB)

System RAM	354 MB
JVM Max Memory	48 MB
Internal Storage	300 MB
External Storage	1982 MB

Front Camera (1.3 megapixel)

Effect	None
Effect Values	None, Mono, Negative, Sepia, Aqua
Exposure Compensation	0
Exposure Compensation Step	1
Focal Length	3.43
Focus Distances	0.10, 1.20, Infinity
Focus Mode	Auto
Focus Mode Values	Auto
Horizontal View Angle	51.2
Jpeg Quality	90
Jpeg Thumbnail Height	240
Jpeg Thumbnail Quality	90
Jpeg Thumbnail Size Values	320x240, 0x0
Jpeg Thumbnail Width	320
Max Exposure Compensation	4
Max Zoom	30
Min Exposure Compensation	4
Orientation	0
Picture Format	Jpeg
Picture Format Values	Jpeg
Picture Size	640x480
Picture Size Values	1280x1024, 640x480, 320x240
Preferred Preview Size For Video	640x480
Preview Format	Yuv420sp
Preview Format Values	Yuv420sp, Yuv420p
Preview Fps Range	5000, 60000
Preview Fps Range Values	(5000, 60000)
Preview Frame Rate	10
Preview Frame Rate Values	20
Preview Size	640x480
Preview Size Values	640x480
Prop Facing	Front
Prop Orientation	0
Rotation	0
Smooth Zoom Supported	False
Vertical View Angle	39.4
Video Frame Format	Yuv420sp
Video Size	640x480
Video Size Values	640x480
Whitebalance	Auto
Whitebalance Values	Auto, Incandescent, Fluorescent, Warm Fluorescent, Daylight, Cloudy Daylight
Zoom	0
Zoom Ratios	100, 106, 113, 120, 126, 133, 140, 146, 153, 160, 166, 173, 180, 186, 193, 200, 206, 213, 220, 226, 233, 240, 246, 253, 260, 266, 273, 280, 286, 293, 300
Zoom Supported	True

Sensors (1)

MMA 3-axis Accelerometer

Vendor	Freescale Semiconductor Inc.
Range	19.6133
Resolution	0.46698335
Power	0.3 mA

OpenGL 1.X (Mali-400 MP)

General

Renderer	Mali-400 MP
Version	OpenGL ES-CM 1.1
Vendor	ARM

Limits

Texture Size	4096x4096
Texture Units	8
Viewport Size	4096x4096
Modelview Stack Depth	32
Projection Stack Depth	32
Texture Stack Depth	32
Lights	8

Compressed Texture Formats

ETC1, Paletted

Extensions

GL_OES_byte_coordinates GL_OES_fixed_point GL_OES_single_precision GL_OES_matrix_get GL_OES_read_format GL_OES_compressed_paletted_texture GL_OES_point_size_array GL_OES_point_sprite GL_OES_texture_npot GL_OES_query_matrix GL_OES_matrix_palette GL_OES_extended_matrix_palette GL_OES_compressed_ETC1_RGB8_texture GL_OES_EGL_image GL_OES_draw_texture GL_OES_depth_texture GL_OES_packed_depth_stencil GL_EXT_texture_format_BGRA8888 GL_OES_framebuffer_object GL_OES_stencil8 GL_OES_depth24 GL_ARM_rgba8 GL_OES_EGL_image_external GL_OES_EGL_sync GL_OES_rgb8_rgba8 GL_EXT_multisampled_render_to_texture GL_OES_texture_cube_map GL_EXT_discard_framebuffer

OpenGL 2.0 (Mali-400 MP)

General

Renderer	Mali-400 MP
Version	OpenGL ES 2.0
Vendor	ARM
GLSL Version	OpenGL ES GLSL ES 1.00

Limits

Texture Size	4096x4096
Texture Units	8
Vertex Textures	0
Combined Textures	8
Viewport Size	4096x4096
Renderbuffer Size	4096x4096
Cubemap Size	1024x1024
Vertex Attributes	16
Vertex Uniforms	256
Varying Vectors	12
Fragment Uniforms	256

Vertex Numeric Precision

Int	[-2^24, 2^24]
Float	[-2^127, 2^127], 2^-23

Fragment Numeric Precision

Low Int	[-2^11, 2^11]
Medium Int	[-2^11, 2^11]
High Int	Not available
Low Float	[-2^15, 2^15], 2^-10
Medium Float	[-2^15, 2^15], 2^-10
High Float	Not available

Compressed Texture Formats

ETC1

Extensions

GL_OES_texture_npot GL_OES_compressed_ETC1_RGB8_texture GL_OES_standard_derivatives GL_OES_EGL_image GL_OES_depth24 GL_ARM_rgba8 GL_ARM_mali_shader_binary GL_OES_depth_texture GL_OES_packed_depth_stencil GL_EXT_texture_format_BGRA8888 GL_EXT_blend_minmax GL_OES_EGL_image_external GL_OES_EGL_sync GL_OES_rgb8_rgba8 GL_EXT_multisampled_render_to_texture GL_EXT_discard_framebuffer

Graphic Modes (RGBA 32bit)

RGB 16bit (565)

RGB 16bit (565), Depth 24bit, Stencil 8bit

RGB 16bit (565), Depth 24bit, Stencil 8bit, Samples x4

RGB 16bit (565), Depth 24bit, Stencil 8bit, Samples x16, Non-Conformant

RGBA 16bit (5551), Depth 24bit, Stencil 8bit

RGBA 16bit (5551), Depth 24bit, Stencil 8bit, Samples x4

RGBA 16bit (5551), Depth 24bit, Stencil 8bit, Samples x16, Non-Conformant

RGBA 16bit (4444), Depth 24bit, Stencil 8bit

RGBA 16bit (4444), Depth 24bit, Stencil 8bit, Samples x4

RGBA 16bit (4444), Depth 24bit, Stencil 8bit, Samples x16, Non-Conformant

RGB 32bit (888)

RGB 32bit (888), Depth 24bit

RGB 32bit (888), Depth 24bit, Stencil 8bit

RGBA 32bit (8888)

RGBA 32bit (8888), Depth 24bit, Stencil 8bit

RGBA 32bit (8888), Depth 24bit, Stencil 8bit

RGBA 32bit (8888), Depth 24bit, Stencil 8bit, Samples x4

RGBA 32bit (8888), Depth 24bit, Stencil 8bit, Samples x16, Non-Conformant

Features (25)

Graphics

OpenGL 2.0

Android Hardware

android.hardware.bluetooth

android.hardware.camera

android.hardware.camera.autofocus

android.hardware.camera.flash

android.hardware.camera.front

android.hardware.faketouch

android.hardware.location

android.hardware.location.gps

android.hardware.location.network

android.hardware.microphone

android.hardware.screen.landscape

android.hardware.screen.portrait

android.hardware.sensor.accelerometer

android.hardware.sensor.compass

android.hardware.telephony

android.hardware.telephony.cdma

android.hardware.telephony.gsm

android.hardware.touchscreen

android.hardware.touchscreen.multitouch

android.hardware.touchscreen.multitouch.distinct

android.hardware.usb.accessory

android.hardware.usb.host

android.hardware.wifi

Android Software

android.software.live_wallpaper

Created with Hardware Info 2.1.0

[Srikant]

Go wild !

--

[Manoj Gudi]

I posted brightness issue on mailing list to which somebody graciously replied

groups.google.com/group/linux-sunxi/browse_thread/thread/dd4108b7aede6355?hl=en

does that make sense to anybody?

On Feb 1, 2013 10:03 PM, "Manoj Gudi" <manoj....@gmail.com> wrote:

I forgot to ask, do you know camera vendor or any other information about it? If you can remember, what all did you try in configs which didn't work?

- manojg

[Srikant]

Not really !

[Manoj Gudi]

patch to remove tux-logo and add aakash-logo during boot (contains source-readme files too)

- bored_on_sunday_afternoon

[Srikant]

Ohh great..can you make the logo in center 

[Srikant]

send me readymade uImage

On Feb 2, 2013 6:20 PM, "Manoj Gudi" <manoj....@gmail.com> wrote:

I tried making the whole background white, but I couldnt...

So the same white image is boxed inside black_bg and centered as requested

apply this patch on top of previous patch

[Manoj Gudi]

strange, gmail drops patches > 8mb

here's uImage

[Srikant]

Works great :-)

[Manoj Gudi]

cool, those 2 patches are against the current androportal/linux-sunxi sunxi-3.0 branch, u may patch if u want

thanks.

[Manoj Gudi]

https://github.com/manojgudi/lock_screen.git 

the code which toggles screen on/off (doesnt turn toggle backlight yet; m working on it.)

previously i coded in python which required evdev package which weighed nearly 60mb with other dependencies, so I changed it to C, which requires hardly anything
 

the code is shabby, but very lightweight and works perfectly well with my device.

[isachin]

@ Manoj:

I tested lock_screen, works as expected, Good work!

Yes you have to wrk on backlight, because now its just like a blank screen saver. Also we have to run this as a backend daemon

process.

Also there’s a minor typo in your readme,

for compiling lock_screen, it should be

$ gcc lock_screen.c -o lock_screen.o

It would be better if you put both the commands(compile and loading) in install.sh, in this way user has to only run ./install.sh

--
Sachin

[isachin]

ok, rc.local tweak is fine, Sorry my mistake

--
Sachin

[Manoj Gudi]

On Feb 4, 2013 11:14 AM, "isachin" <iclco...@gmail.com> wrote:
>
> @ Manoj:
> I tested lock_screen, works as expected, Good work!
>
> Yes you have to wrk on backlight, because now its just like a blank screen saver. Also we have to run this as a backend daemon
> process.

Agreed

> Also there’s a minor typo in your readme,
>
> for compiling lock_screen, it should be
>
> $ gcc lock_screen.c -o lock_screen.o
>
>

I'll take a note of it

> It would be better if you put both the commands(compile and loading) in install.sh, in this way user has to only run ./install.sh
>

i am in college till 1800, would you send me patch ?

[Manoj Gudi]

[Manoj Gudi]

i think its too simple for a patch, I'll change it later.

On Feb 4, 2013 12:01 PM, "Manoj Gudi" <manoj....@gmail.com> wrote:

[Manoj Gudi]

After learning in-depth module writing and tweaking I finally think I
have idea how to solve brightness issue; also the reply in
sunxi-mailing list makes sense now.

https://groups.google.com/forum/?hl=en&fromgroups=#!topic/linux-sunxi/3UEIt67eY1U

What he essentially meant is the module is inserted and the device
driver is registered but it has no binding to sysfs;
This can be done in two ways=>

1> Using mknod with proper device major number(since lcd/disp drivers
are essentially char drivers)
2> Writing class_create structure from <linux/device.h> which is then
picked up automatically by udev

the 1st method doesnt work; 2nd method involves changing source
code(point of no-return)...

@Srikant you mentioned that you were successful in controlling
brightness of device earlier; did you do it by user-space tools in
which case I'm looking at completely wrong direction or did you use
mknod to create sysfs entries?

[Srikant]

Hi, good going. 

Did I said about controlling lcd brightness?  Brightness control didn't work.

[Manoj Gudi]

On Mar 3, 2013 10:11 AM, "Srikant" <u.srikan...@gmail.com> wrote:
>
> Hi, good going. 
> Did I said about controlling lcd brightness?  Brightness control didn't work.

Yeah,
"
Date: Saturday, 26 January 2013 1
To: Manoj Gudi <manoj....@gmail.com> Cc: linux_on_aakash@googlegrou

Hide details

i don't have aakash with me now..i remember we have controlled brightness some how..can't be sure "

Anyway if you remember and if it was user space hack then please inform me. Else I'll be playing wit source code without any reason.

[Srikant]

u carry on, let me think again

[Manoj Gudi]

The brightness issue is deeper than I thought...
I have tried a hand at all things which seemed logical (i.e. changing
brightness variable; tweak LCD-controller driver etc but it didn't
work)

I'll require more time to study this issue (probably write custom
driver to control lcd)...

Are there any other issues which need a quick look?