00278067 GPS location.getTime is 0

[admin]

Subject [David Lan] GPS location.getTime is 0
Description Dear QCT,
Always thank you for your kind support.
Our GPS Java application invoke location.getTime() of *\android\frameworks\base\location\java\android\location to get current GPS Position Satellite Time and return 0. (no problem for getLatitude()/getLongitude()/getAltitude())
Check the log of DDMS, we found the the members' values of location_report_ptr in loc_eng_report_position(...) of libloc_api\loc_eng.cpp updated information adhere:
location_report_ptr->timestamp_calendar.year = 0
location_report_ptr->timestamp_calendar.month = 0
location_report_ptr->timestamp_calendar.month = 0
location_report_ptr->timestamp_calendar.day_of_week = 0
location_report_ptr->timestamp_calendar.hour
location_report_ptr->timestamp_calendar.minute = 0
location_report_ptr->timestamp_utc = 1075851863
location_report_ptr->leap_seconds =15
location_report_ptr->time_unc = 1075851903
Do you have any idea why can't we get the correct GPS Position Satellite Time?
Thank you.
Best Regards
David

[admin]

Comment

Created By: David Lan (2/22/2010 6:55 PM)
Hi Howard,

After checked the AP source code and found that there is difference between the AP for BSP2110 and BSP2010. Hence the error had been fixed after modified AP source code.
Sincerely thank you for your kind help, this case is supposed to be closed right now.

Best Regards
David



Created By: David Lan (2/22/2010 3:55 AM)
Hi Howard,

Sincerely thank you for your kind support and sorry for response late.
I suppose there is different GPS time format between BSP 2010 and 2110 for GPS Tracker to receive, and it results in receiving incorrect GPS time.
I would consulted with our AP Team for further reference later and update you when solution is available.

Best Regards
David



Created By: Howard Chiu (2/22/2010 2:49 AM)
Hi David,

I can finally build and test it on our SURF board
I added some debug message in loc_pd.c\loc_pd_get_report()
====================================================================================
// Fields are populated according to the order in GPS_POSITION
if (TRUE == process_pd_info)
{
MSG_ERROR(" Inside if (TRUE == process_pd_info)",0,0,0);

pd_info_ptr = &(pd_module_data_ptr->pd_info_copy);
(omit some code)
// avoid translation to windows mobile generall error code
parsed_report_ptr->session_status = LOC_SESS_STATUS_SUCCESS;

MSG_ERROR("Before checking PDSM_PD_GPS_TIME_VALID, %d", pd_data_ptr->ext_opt_field_mask,0,0);
if (pd_data_ptr->ext_opt_field_mask & PDSM_PD_GPS_TIME_VALID)
{
MSG_ERROR("Howard -PDSM_PD_GPS_TIME_VALID is part of mask",0,0,0);
MSG_ERROR("Howard - Date is %d, Hour is %d, Minute is %d", pd_ext_data_ptr->date, pd_ext_data_ptr->hour, pd_ext_data_ptr->minute);

// fill in utc time and get the leap_seconds
leap_seconds = loc_pd_fill_timestamp_utc(parsed_report_ptr, pd_data_ptr);

// timestamp_calendar is always valid
(omit some code)

//Howard
MSG_ERROR("Calendar Year=%d, Month=%d,Day=%d",parsed_report_ptr->timestamp_calendar.year, parsed_report_ptr->timestamp_calendar.month, parsed_report_ptr->timestamp_calendar.day);
MSG_ERROR("Calendar Hour=%d, Minute=%d,Second=%d",parsed_report_ptr->timestamp_calendar.hour, parsed_report_ptr->timestamp_calendar.minute, parsed_report_ptr->timestamp_calendar.second);
// Howard
========================================================================================

You can see from my log that it is time is indeed correct.
So now I think modem side should be working fine.

Now your ADB log will be very helpful because you have already modified the code so the ADB log already shows the time.
I will also try to get the android code but it will take some time.

thanks,
Howard



Created By: Howard Chiu (2/22/2010 12:41 AM)
Hi David,

Please send me the ADB log at your earliest convenience.

thanks,
Howard



Created By: David Lan (2/15/2010 7:04 PM)
Hi Howard,

The ADB log would be attached while I am in office.



Created By: Howard Chiu (2/12/2010 1:07 AM)
Hi David,

Do you have the ADB log? I just want to eliminate problems causing by the Application.

thanks,
Howard



Created By: Howard Chiu (2/11/2010 5:33 PM)
Hi David,

I am happy that we can finally see GPS time.
Please wait while I investigate the incorrect GPS time issue this morning.
I hope I can provide you some answers today.

regards,
Howard



Created By: David Lan (2/11/2010 5:28 AM)
Hi Howard,

Finally show the GPS time!!
Please refer to the attached GPSTracker_20100211.png for the wrong GPS time and 02-11.21-14.isf for the QXDM log.
Do you have any idea for correcting the incorrect GPS time?

Best Regards
David



Created By: Howard Chiu (2/11/2010 2:11 AM)
Hi David,

Due to a very strange issue, I cannot build this particular code. I am still trying to figure it out. As a result, I can't try another change which I feel might be helpful.

You are free to try if you want.
Change gps\sm\tm\src\tm_util.c\tm_util_pd_info_data_init()

After cl_info->pd_info.pd_data.opt_field_mask = PDSM_PD_UNCERTAINTY_VALID;

Add
cl_info->pd_info.pd_data.ext_opt_field_mask = PDSM_PD_LAT_LONG_VALID;
cl_info->pd_info.pd_data.ext_opt_field_mask |= PDSM_PD_GPS_TIME_VALID;

Re-compile and try again.

hope this helps,
Howard



Created By: David Lan (2/10/2010 7:43 PM)
Hi Howard,

Your consideration is reasonable.
It is not convenient for Sam to open an account, hence I would share this SR related information with Sam by myself. Anyway, thank you for your kind help.

By the way, is there any update for this issue?
Why does BSP 2110 encounter with this problem which doesn't be occurred on BSP 2010?

Best Regards
David



Created By: Howard Chiu (2/8/2010 11:09 PM)
Hi David,

I am sorry but I am currently pre-occupied with another case from your company. I will try to give you an update later this afternoon or tomorrow.

Can you please ask Sam to open an account? He was using wdd_wp@fihspec.com which I believe is a public account shared by many users. Once he opens a personal account, then you or I can loop him in.

Sorry to keep you waiting..

regards,
Howard



Created By: David Lan (2/8/2010 11:02 PM) | Last Modified By: David Lan (2/8/2010 11:03 PM)
Hi Howard,

Is there any update for this issue?
I had tried the image which were built by M76XXMSNCJNLYM2010 and without this problem.
It seems there is some side effect when we upgrade our BSP from 2010 to 2110.
By the way, could you help to include our "samhsu@fihspec.com" in this case?
Since I can't find him while trying to add member.
Thanks.

Best Regards
David



Created By: David Lan (2/4/2010 10:29 PM)
Hi Howard,

Please refer to the attached 02-05.14-09.isf and log_20100205_3.txt.
The procedure is adhere.
1. open Gen7GPS_General_LocMW_Howard.dmc
2. Set Communication
3. logging
4. Save Items

I may fail to save the configuration of Gen7GPS_General_LocMW_Howard.dmc in QXDM for my previous saving QXDM Professional Configuration File. Sorry to inconvinence you.

Best Regards
David



Created By: Howard Chiu (2/4/2010 9:55 PM)
Hi David,

Can I check with you how to use the mask I gave you?

I don't know what went wrong but I couldn't see from your log the details I want, specifically loc_pd related. But I did see Android QCRIL message which I didn't choose.

Can you briefly describe your procedure? I am still trying to come up with a working build

thanks,
Howard



Created By: David Lan (2/4/2010 7:42 PM)
Hi Howard,

Please refer to the attached 02-05.11-12.isf, log_20100205_1.txt, tm_core_notConnected.PNG and tm_core_Connected.PNG which were logged with your QXDM mask. Thank you.

Best Regards
David



Created By: David Lan (2/4/2010 6:54 PM)
Hi Howard,

Please refer to the attached 02-05.10-44.isf and log_20100205.txt which were logged with your QXDM mask.
Thank you for your support again.

Best Regards
David



Created By: Howard Chiu (2/4/2010 5:09 PM)
Hi David,

Can you please do me a favor? Can you try logging again with the new QXDM mask I provided? I need more LOC MW related mask.

Just a reminder, please click "File" -> "Save Items" once you are finished. This is to ensure we capture all details.

I am also preparing our FFA for testing, if I can find a suitable code

thanks,
Howard



Created By: David Lan (2/4/2010 5:09 AM)
Hi Howard,

The new attached M20100204_210243.isf and log_20100204.txt.

Best Regards
David



Created By: David Lan (2/4/2010 4:40 AM)
Hi Howard,

We still observe the "GPS location.getTime is 0" problem after having compiled the modified modem processor code.
1. upgrade mARM image by Trace32
2. upgrade original aArm image by our Upgrade Wizard.
You may refer to the attached log_20100203_2.txt and M20100204_201250.isf for the log.
Please inform me if you have any idea for this problem.

Best Regards
David



Created By: David Lan (2/3/2010 7:17 PM)
Hi Howard,

Sincerely thank you for your kind support.
I will check to see whether the solution fixes the problem or not, and update you as soon as I can.

Best Regards
David



Created By: Howard Chiu (2/3/2010 7:11 PM)
Hi David,

This problem is possibly due to an existing CR. You can try the CR fix and see if you still observe the problem or not.

Change gps\sm\tm\src\tm_core.c\tm_core_xlate_pos_to_ext_status()

Add
dest_buf_ptr->pd_data.ext_opt_field_mask = PDSM_PD_LAT_LONG_VALID;
dest_buf_ptr->pd_data.ext_opt_field_mask |= PDSM_PD_GPS_TIME_VALID;

Before
/* Copy msec field received from PE
*/

Please let me know whether that fixes the problem or not.

thanks,
Howard



Created By: Howard Chiu (2/3/2010 6:47 PM)
Hi David,

I found something strange, requiring clarifications from our development team. This will possibly take some time. I will update you ASAP

thanks,
Howard



Created By: David Lan (2/2/2010 10:16 PM)
Hi Howard,

No problem.
You can refer to the attached M20100203_140407.isf and log_20100203_1.txt at the same time.
I log them by your Gen7GPS_General.dmc.

Best Regards
David



Created By: Howard Chiu (2/2/2010 10:11 PM)
Hi...

I am looking at your new log now.
I am sorry I missed the other one because I only saw the beginning of the ADB log.

I will look into it again and get back to you ASAP

thanks,
Howard



Created By: David Lan (2/2/2010 8:48 PM)
Hi Howard,

Before your QXDM mask, I had modified some LogView Configuration and log log_20100203.txt and M20100203_123906.isf as attachments.
I will attach the logs after having tried your configuration, thanks.

Best Regards
David



Created By: David Lan (2/2/2010 7:44 PM)
Hi Howard,

Thank you for your kind help.
I don't see QXDM mask (.dmc) in case attachment, where can I get this useful tool?

In log.txt, around the period adhere
------------------------------
D/GPS Tracking Service( 1265): onLocationChanged()
D/GPS Tracking Service( 1265): onLocationChanged() ---> CONNECTIONSTATE_CONNECTED
D/GPS Tracking Service( 1265): updateGPSLocation()
D/GPS Tracking Service( 1265): getGPSInformation()
D/GPS Tracking Service( 1265): Location != NULL.
D/GPS Tracking Service( 1265): getGPSInformation() ---> CONNECTIONSTATE_CONNECTED
D/GPS Tracking Service( 1265): before getTime()
D/GPS Tracking Service( 1265): after getTime()
D/GPS Tracking Service( 1265): getTime = 0
------------------------------
you can see
D/libloc_api( 1080): sv count: 9
D/libloc_api( 1080): Latitude: 25.0519741 (final)
D/libloc_api( 1080): Longitude: 121.2849254

Should I get the time after then?

Best Regards
David



Created By: Howard Chiu (2/2/2010 7:11 PM)
Hi David,

First of all, the QXDM only captures very limited log. I have attach a regular QXDM mask (.dmc) for your reference. It could be useful if you wish to capture logs for Qualcomm to analyze in the future. Just double-click on the .dmc file and it will launch QXDM automatically. Once finished, click "File" -> "Save Items" to save the log.

From the ADB log, you can see
D/libloc_api( 1080): Latitude: 25.0520092 (intermediate)

which tells me that this is an intermediate position, which tells me this is not a regular position fix. You can refer to log_parsed_report() in log_eng_log.cpp

Also, you can see
D/libloc_api( 1080): sv count: 0 => no satellite is observed

So, without a proper position fix, we cannot give you the time. So what you saw would be pretty reasonable. This intermediate position is just something to give you an idea of roughly where you are.

I believe the intermediate is fixed at your location because you have previously obtained a successful position fix there.

If you have no further question related to this Case. Please close this Case for me.

hope this helps,
Howard



Created By: David Lan (2/2/2010 6:57 PM)
Hi Howard,

This position was carried out at No.151, Sec. 1, Nankan Rd., Lujhu Township, Taoyuan County 338, Taiwan (R.O.C.).
Besides, please refer to the attached M20100203_103741.isf of QXDM LogView and log.txt of ddms.
Thank you for your kind support, please inform me if there is anything else.

Best Regards
David



Created By: Howard Chiu (2/2/2010 8:44 AM)
HI David,

Where was the test carried out?
I just want to know so I can compare against what I am seeing from gpstest

thanks,
Howard



Created By: David Lan (2/2/2010 7:03 AM)
Hi Howard,
You might refer to the attached gpstest.png for the information of Lon/Lat.
Lat : 25.0520917
Long : 121.28.49283
By the way, teh ADB logs and QXDM logs is supposed to be released later, thank you.
Best Regards
David



Created By: David Lan (2/2/2010 6:51 AM)
Hi Howard,
Thank you for your kind response, I will release you them as soon as possible.
Best Regards
David



Created By: Howard Chiu (2/2/2010 2:25 AM)
Hi David,

Thank you for opening this Case.

I am thinking it might have some thing to do with Intermediate Fix (fixes that have poorer quality than specified in the QoS). To confirm, I would require both ADB (android)logs and QXDM log (modem).

Also, what are the values of the Lon/Lat returned from getLatitude()/getLongitude()?

thanks,
Howard