Phonepart / Sensors not working on B7610 in Froyo Beta 2

[GoogleCodeExporter]

What steps will reproduce the problem?
1. Boot phone
2.
3.

What is the expected output? What do you see instead?
Expeced: Working phone, sim detected, sensors working.
Actual: Phone not working (no service), no sensors, no vib

Please use labels and text to provide additional information.
Compiled latest rev of source on this project page

Original issue reported on code.google.com by erikcas1972@gmail.com on 19 Sep 2011 at 8:11

[GoogleCodeExporter]

[deleted comment]

[GoogleCodeExporter]

Discovered this in arch/arm/mach-s3c6410/omnia_II_gpio_table.h

If I compile kernel 2.6.32 with lines commented then vib is not working, cannot 
see "GPIO phone loaded" in kernel.

If I compile kernel 2.6.32 with lines uncommented like in kern_oII dmesg shows 
"GPIO phone loaded" (or something) but is not working.

But vib IS working.

Could it be that these GPIO's are defined elsewhere. But where? Or are GPIO's 
defined direct in the driver source-files?

(I doublechecked  arch/arm/mach-s3c6410/omnia_II_gpio_table.h for this in PHJ 
source for both oII and 2.6.32, it shows the same)
kern_2.6.32 (all lines are commented)

// { GPIO_PHONE_ON, GPIO_PHONE_ON_AF, GPIO_LEVEL_LOW, S3C_GPIO_PULL_NONE, 0, 0 
}, 
// { GPIO_TA_nCONNECTED, GPIO_TA_nCONNECTED_AF, GPIO_LEVEL_NONE, 
S3C_GPIO_PULL_NONE, 0, 0 }, 
// { GPIO_USIM_BOOT, GPIO_USIM_BOOT_AF, GPIO_LEVEL_LOW, S3C_GPIO_PULL_NONE, 0, 
0 }, 
// { GPIO_CAM_3M_nSTBY, 0, GPIO_LEVEL_NONE,S3C_GPIO_PULL_NONE, 0, 0 }, 
// { GPIO_MIC_SEL, GPIO_MIC_SEL_AF, GPIO_LEVEL_LOW, S3C_GPIO_PULL_NONE, 0, 0 }, 
// can be good 
// { GPIO_FM_INT, 0, GPIO_LEVEL_NONE, S3C_GPIO_PULL_DOWN, 0, 0 }, 
// { GPIO_VIB_EN, GPIO_VIB_EN_AF, GPIO_LEVEL_LOW, S3C_GPIO_PULL_NONE, 0, 0 }, 
// { GPIO_ALS_EN, GPIO_ALS_EN_AF, GPIO_LEVEL_LOW, S3C_GPIO_PULL_NONE, 0, 0 }, 

............................

kern_oII (more lines uncommented)

{ GPIO_PHONE_ON, GPIO_PHONE_ON_AF, GPIO_LEVEL_LOW, S3C_GPIO_PULL_NONE, 0, 0 }, 
{ GPIO_TA_nCONNECTED, GPIO_TA_nCONNECTED_AF, GPIO_LEVEL_NONE, 
S3C_GPIO_PULL_NONE, 0, 0 }, 
{ GPIO_USIM_BOOT, GPIO_USIM_BOOT_AF, GPIO_LEVEL_LOW, S3C_GPIO_PULL_NONE, 0, 0 
}, 
{ GPIO_CAM_3M_nSTBY, 0, GPIO_LEVEL_NONE,S3C_GPIO_PULL_NONE, 0, 0 }, 
// { GPIO_MIC_SEL, GPIO_MIC_SEL_AF, GPIO_LEVEL_LOW, S3C_GPIO_PULL_NONE, 0, 0 }, 
// can be good 
{ GPIO_FM_INT, 0, GPIO_LEVEL_NONE, S3C_GPIO_PULL_DOWN, 0, 0 }, 
{ GPIO_VIB_EN, GPIO_VIB_EN_AF, GPIO_LEVEL_LOW, S3C_GPIO_PULL_NONE, 0, 0 }, 
// { GPIO_ALS_EN, GPIO_ALS_EN_AF, GPIO_LEVEL_LOW, S3C_GPIO_PULL_NONE, 0, 0 }, 

Original comment by erikcas1972@gmail.com on 19 Sep 2011 at 8:29

[GoogleCodeExporter]

And off course recheck all GPIO's for B7610

Original comment by erikcas1972@gmail.com on 19 Sep 2011 at 10:34

[GoogleCodeExporter]

doubt if this is kernel matter. I think it has to do with the android-image 
build on the kernel. And I think I am right.
If I disable pin on SIM, and I boot Adroid it shows "No Service". But if I take 
another phone and make a call to my phone, I can hear the tone like I'm waiting 
to answer.

If (in WinMo!) I am in airplane mode, and I make a call to my phone, I immeatey 
get a spoken notification that my phone is off or has no service

Original comment by erikcas1972@gmail.com on 20 Sep 2011 at 2:26

[GoogleCodeExporter]

Some progress:
With some modifications, phonepart shows up in dmesg, PHONE_ON = 0, which I 
think that means phone is off.

Logcat shows
D/RIL   ( 1776): dload_test start. 
D/RIL   ( 1776): onedram device (/dev/dpram0) open failed

every six to eight seconds.
When I check the /dev dir, dpram0 isn't there.

Voyteckst answered on de modaco forum that is could be due to hardware 
difference between B7610 and I8000.
He could be right, but I see more I8000 users on the modaco forum with phone 
problems and posting logcats with the /dev/dpram0 problems. I pointed that out 
to him, his answer was "....we're investigating this. Maybe the cause is 
modem.bin from i6500, but not sure."

Interesting part on 
http://forum.xda-developers.com/showthread.php?t=1162825&page=2
[Quote]
Unfortunately I can't think of a great solution. One option would be to modify 
multipdp.c to add a sleep loop around "filp_open(DPRAM_DEVNAME, ...)" to wait 
for the device to be created. However, that would mean using your own compile 
of multipdp.ko instead of the one Samsung provides, which could open a 
different can of worms.

A second option is to move "insmod /modules/multipdp.ko" until after mounting 
/system, hoping that init would've caught up by then. This can even be insured 
by running a script that checks for the device node and sleeps if it doesn't 
exist yet. Something like:
[unquote]

Original comment by erikcas1972@gmail.com on 22 Sep 2011 at 10:21

[GoogleCodeExporter]

well iam pleased to see that you are so dedicated for our beloved b7610, one 
day we will see a rev patch for b7610 too. good luck..

Original comment by kaspy...@gmail.com on 23 Sep 2011 at 4:23

[GoogleCodeExporter]

invalid module format when I try to load the modules manually. Version error?
Should be strange, build zImage and modules at one time.

Original comment by erikcas1972@gmail.com on 25 Sep 2011 at 4:11

[GoogleCodeExporter]

Dpram module doesn't load.
Result of insmod /modules/dpram.ko

insmod: can't insert '/modules/dpram.ko': unknown symbol in module or invalid 
parameter

dmesg for this action shows
[  698.927818] dpram: Unknown symbol __class_create
[  698.929712] dpram: Unknown symbol platform_driver_unregister
[  698.935625] dpram: Unknown symbol class_destroy
[  698.938984] dpram: Unknown symbol __gpio_get_value
[  698.948827] dpram: Unknown symbol platform_driver_register
[  698.951168] dpram: Unknown symbol device_create
[  698.955281] dpram: Unknown symbol gpio_direction_output
[  698.965066] dpram: Unknown symbol __gpio_set_value
[  698.967360] dpram: Unknown symbol gpio_request

Dmesg attached

Original comment by erikcas1972@gmail.com on 26 Sep 2011 at 12:09

[GoogleCodeExporter]

Forgot to attach dmesg. Here it is

Original comment by erikcas1972@gmail.com on 26 Sep 2011 at 12:12

Attachments:

• dmesg.txt

[GoogleCodeExporter]

Made B7610 work (phone/data) by using old phone part. Used kernel 2.6.32 but 
some modules from previous kernel and some modifications to android fs.

From that dirty workaround I will search for the rootcause.

Original comment by erikcas1972@gmail.com on 27 Sep 2011 at 9:51

[GoogleCodeExporter]

After a day of google-ing around, I think Voyteckst and Sandor are right.

Cause is probably modem.bin which is from I6500. It has same as Omnia2 MSM6290 
chip for gsm/3g etc.

B7610 uses a MSM6246 chip. I searched for an alternative, but haven't found one 
yet. Tried S8000 and I5700 images, without success.

Maybe B7610 users can help in this search, all suggestions are welcome.

Original comment by erikcas1972@gmail.com on 27 Sep 2011 at 4:01

[GoogleCodeExporter]

Introduced spica modem.bin & libsec-ril.so
Need testers to see if it works.
On my B7610 I connect to my provider, but no phone/data. This could be "IMEI" 
problem related. See I8000 forum for that. Devs are looking into this.

Original comment by erikcas1972@gmail.com on 28 Sep 2011 at 12:20

[GoogleCodeExporter]

dmesg & logcat with kernel ?REV61? (maybe REV51)

Original comment by Sven.Boe...@gmail.com on 28 Sep 2011 at 7:22

Attachments:

• REV61_dmesg_2011-09-28-17-46-24.txt
• REV61_alogcat.2011-09-28-19-43-25+0200.txt

[GoogleCodeExporter]

dmesg & logcat with kernel REV64

Original comment by Sven.Boe...@gmail.com on 29 Sep 2011 at 6:17

Attachments:

• dmesg_REV64.txt
• alogcat.2011-09-29-05-55-50+0200.txt

[GoogleCodeExporter]

So Far:

Think this is not a kernel issue. It is an Android issue. So I adressed a few 
questions to some guys who are into developing samsung RILD if this can be 
fixed or that this is the same issue.

Original comment by erikcas1972@gmail.com on 4 Oct 2011 at 10:16

[GoogleCodeExporter]

With dpram module and (modified) multipdp module from kern_oII phone and data 
are working on B7610. 

See also 
http://www.modaco.com/topic/346432-froyo-beta-2-on-b7610/page__st__120__gopid__1
825700#entry1825700

Original comment by erikcas1972@gmail.com on 16 Oct 2011 at 10:07

[GoogleCodeExporter]

Lowered priority, is working with workaround above

Original comment by erikcas1972@gmail.com on 16 Oct 2011 at 10:08

• Added labels: Priority-Low
• Removed labels: Priority-Critical

[GoogleCodeExporter]

Fixed in Beta3. Sensor working, but inverted. Will be fixed

Original comment by erikcas1972@gmail.com on 23 Oct 2011 at 6:04

• Changed state: Fixed