GPS Woes » BUG Community

May 21, 2008 9:29am

afritz 24 posts

I received my bug yesterday. It doesn't detect the GPS at all. I SSH in and tail /var/log/concierge.log and I get the following:

[Wed May 21 13:24:09 GMT 2008] [DEBUG] Received message from event pipe: FFFF 0 0 add

[Wed May 21 13:24:09 GMT 2008] [DEBUG] processing: FFFF 0 0 add

[Wed May 21 13:24:09 GMT 2008] [ERROR] This case is not implemented; starting a module bundle that is not already in the runtime environment.

Has anyone else seen this? For those that have followed my other threads, this happened before I started messing with the accelerometer stuff (i.e. with the factory root fs image).

May 21, 2008 10:17am

jconnolly 285 posts

afritz

I received my bug yesterday. It doesn't detect the GPS at all. I SSH in and tail /var/log/concierge.log and I get the following:

[Wed May 21 13:24:09 GMT 2008] [DEBUG] Received message from event pipe: FFFF 0 0 add [Wed May 21 13:24:09 GMT 2008] [DEBUG] processing: FFFF 0 0 add [Wed May 21 13:24:09 GMT 2008] [ERROR] This case is not implemented; starting a module bundle that is not already in the runtime environment.

Has anyone else seen this? For those that have followed my other threads, this happened before I started messing with the accelerometer stuff (i.e. with the factory root fs image).

Hi Andrew,

What kind of output are you getting if you tail -f /var/log/kern.log ? A healthy BUGlocate/BUGbase hotplug interaction should give the output below:

Board Inserted at conn-m1
bmi_gps.c: probe slot 0
Jan 14 14:56:54 localhost kernel: Board Inserted at conn-m1
Jan 14 14:56:54 localhost kernel: bmi: Matched Device conn-m1 with Driver bmi_gps
Jan 14 14:56:54 localhost kernel: bmi: Probing driver bmi_gps with device conn-m1
Jan 14 14:56:54 localhost kernel: CLASS: registering class device: ID = 'bmi_gps_control_m1'
Jan 14 14:56:54 localhost kernel: class_uevent - name = bmi_gps_control_m1
Jan 14 14:56:54 localhost kernel: class_device_create_uevent called for bmi_gps_control_m1
Jan 14 14:56:54 localhost kernel: bmi_gps.c: probe slot 0
Jan 14 14:56:54 localhost kernel: bound device 'conn-m1' to driver 'bmi_gps'
Jan 14 14:56:54 localhost kernel: bmi: Bound Device conn-m1 to Driver bmi_gps

It sounds like it may be a problem with the Java Native Interface implementation interacting with the GPS driver. Let me know and we'll get to the bottom of this.

May 21, 2008 10:30am

afritz 24 posts

Well, one step closer to a root cause...

May 21 14:25:20 localhost kernel: Board Inserted at conn-m1
May 21 14:25:20 localhost kernel: bmi_device_work_handler(0): device_attach failed.

May 21, 2008 11:20am

jconnolly 285 posts

afritz

Well, one step closer to a root cause...

May 21 14:25:20 localhost kernel: Board Inserted at conn-m1 May 21 14:25:20 localhost kernel: bmi_device_work_handler(0): device_attach failed.

Andrew,

Eek, looks like you were right about the "woes" part. From the output it looks like it could be a connector problem, or that the firmware on your GPS module wasn't correctly flashed in manufacturing. While the latter is more easily remedied, we haven't yet documented the procedure for reflashing the module firmware. I'm going to have our support rep contact you to remedy this problem. Sorry for the inconvenience and thanks for taking it to the forums, for the benefit of other BUG users.

May 21, 2008 11:25am

afritz 24 posts

No problem and thanks for your help on this and other issues.

I’m 100% comfortable attempting a reflash of the GPS if a procedure is available. However, I will say the GPS doesn’t snap in nearly as positively as the other 2 modules. It seems to engage but not as firmly. I’ve played with holding it on more forcefully by hand with no luck. My initial guess (after unboxing and plugging everything in for the first time) was it was connector related just based on the feel.

May 21, 2008 11:31am

jconnolly 285 posts

afritz

No problem and thanks for your help on this and other issues.

I'm 100% comfortable attempting a reflash of the GPS if a procedure is available. However, I will say the GPS doesn't snap in nearly as positively as the other 2 modules. It seems to engage but not as firmly. I've played with holding it on more forcefully by hand with no luck. My initial guess (after unboxing and plugging everything in for the first time) was it was connector related just based on the feel.

Andrew,

Good to know, I'll pass this information along to the folks who crack whips. :) We have an internal procedure to ensure that what ends up in your hands has passed a rigorous quality assurance program. This includes ensuring the "right feel". If your module does not feel right, we want to know.

Thanks again.

Jul 22, 2008 9:49am

jconnolly 285 posts

Quick update: GPS antennas really do the trick.. I wouldn’t buy a 90-degree MMCX connector antenna though. Insertion and removal should be done delicately, and the 90-degree angle makes me nervous about detaching the external-antenna mount from the board.

If this happens to you, be sure to email support@buglabs.net.

Jul 22, 2008 9:58am

jennyw 20 posts

Actually, because it’s come to our attention that some of the earlier BUGlocate GPS modules are not performing as well as we’d hoped, we’ve decided to send an external antenna gratis to all customers who currently own these units. Look out for these over the next couple of weeks!

We’re hoping this really helps improve your overall BUG experience, and wanted to thank you again for being part of our community. And thanks to John for all of his testing to determine which antenna would be the best one to send to you guys!

Jenny

BUG Community

Welcome! Log In

Forums → BUGmodules → GPS Woes

Log in to reply to this topic

Log in to reply to this topic

May 21, 2008 9:29am afritz 24 posts	I received my bug yesterday. It doesn't detect the GPS at all. I SSH in and tail /var/log/concierge.log and I get the following: `[Wed May 21 13:24:09 GMT 2008] [DEBUG] Received message from event pipe: FFFF 0 0 add [Wed May 21 13:24:09 GMT 2008] [DEBUG] processing: FFFF 0 0 add [Wed May 21 13:24:09 GMT 2008] [ERROR] This case is not implemented; starting a module bundle that is not already in the runtime environment.` Has anyone else seen this? For those that have followed my other threads, this happened before I started messing with the accelerometer stuff (i.e. with the factory root fs image).
May 21, 2008 10:17am jconnolly 285 posts	afritz I received my bug yesterday. It doesn't detect the GPS at all. I SSH in and tail /var/log/concierge.log and I get the following: `[Wed May 21 13:24:09 GMT 2008] [DEBUG] Received message from event pipe: FFFF 0 0 add [Wed May 21 13:24:09 GMT 2008] [DEBUG] processing: FFFF 0 0 add [Wed May 21 13:24:09 GMT 2008] [ERROR] This case is not implemented; starting a module bundle that is not already in the runtime environment.` Has anyone else seen this? For those that have followed my other threads, this happened before I started messing with the accelerometer stuff (i.e. with the factory root fs image). Hi Andrew, What kind of output are you getting if you tail -f /var/log/kern.log ? A healthy BUGlocate/BUGbase hotplug interaction should give the output below: Board Inserted at conn-m1 bmi_gps.c: probe slot 0 Jan 14 14:56:54 localhost kernel: Board Inserted at conn-m1 Jan 14 14:56:54 localhost kernel: bmi: Matched Device conn-m1 with Driver bmi_gps Jan 14 14:56:54 localhost kernel: bmi: Probing driver bmi_gps with device conn-m1 Jan 14 14:56:54 localhost kernel: CLASS: registering class device: ID = 'bmi_gps_control_m1' Jan 14 14:56:54 localhost kernel: class_uevent - name = bmi_gps_control_m1 Jan 14 14:56:54 localhost kernel: class_device_create_uevent called for bmi_gps_control_m1 Jan 14 14:56:54 localhost kernel: bmi_gps.c: probe slot 0 Jan 14 14:56:54 localhost kernel: bound device 'conn-m1' to driver 'bmi_gps' Jan 14 14:56:54 localhost kernel: bmi: Bound Device conn-m1 to Driver bmi_gps It sounds like it may be a problem with the Java Native Interface implementation interacting with the GPS driver. Let me know and we'll get to the bottom of this.
May 21, 2008 10:30am afritz 24 posts	Well, one step closer to a root cause... `May 21 14:25:20 localhost kernel: Board Inserted at conn-m1 May 21 14:25:20 localhost kernel: bmi_device_work_handler(0): device_attach failed.`
May 21, 2008 11:20am jconnolly 285 posts	afritz Well, one step closer to a root cause... `May 21 14:25:20 localhost kernel: Board Inserted at conn-m1 May 21 14:25:20 localhost kernel: bmi_device_work_handler(0): device_attach failed.` Andrew, Eek, looks like you were right about the "woes" part. From the output it looks like it could be a connector problem, or that the firmware on your GPS module wasn't correctly flashed in manufacturing. While the latter is more easily remedied, we haven't yet documented the procedure for reflashing the module firmware. I'm going to have our support rep contact you to remedy this problem. Sorry for the inconvenience and thanks for taking it to the forums, for the benefit of other BUG users.
May 21, 2008 11:25am afritz 24 posts	No problem and thanks for your help on this and other issues. I’m 100% comfortable attempting a reflash of the GPS if a procedure is available. However, I will say the GPS doesn’t snap in nearly as positively as the other 2 modules. It seems to engage but not as firmly. I’ve played with holding it on more forcefully by hand with no luck. My initial guess (after unboxing and plugging everything in for the first time) was it was connector related just based on the feel.
May 21, 2008 11:31am jconnolly 285 posts	afritz No problem and thanks for your help on this and other issues. I'm 100% comfortable attempting a reflash of the GPS if a procedure is available. However, I will say the GPS doesn't snap in nearly as positively as the other 2 modules. It seems to engage but not as firmly. I've played with holding it on more forcefully by hand with no luck. My initial guess (after unboxing and plugging everything in for the first time) was it was connector related just based on the feel. Andrew, Good to know, I'll pass this information along to the folks who crack whips. :) We have an internal procedure to ensure that what ends up in your hands has passed a rigorous quality assurance program. This includes ensuring the "right feel". If your module does not feel right, we want to know. Thanks again.
Jul 22, 2008 9:49am jconnolly 285 posts	Quick update: GPS antennas really do the trick.. I wouldn’t buy a 90-degree MMCX connector antenna though. Insertion and removal should be done delicately, and the 90-degree angle makes me nervous about detaching the external-antenna mount from the board. If this happens to you, be sure to email support@buglabs.net.
Jul 22, 2008 9:58am jennyw 20 posts	Actually, because it’s come to our attention that some of the earlier BUGlocate GPS modules are not performing as well as we’d hoped, we’ve decided to send an external antenna gratis to all customers who currently own these units. Look out for these over the next couple of weeks! We’re hoping this really helps improve your overall BUG experience, and wanted to thank you again for being part of our community. And thanks to John for all of his testing to determine which antenna would be the best one to send to you guys! Jenny