Ticket #122 (closed defect: fixed)

Opened 6 years ago

Last modified 5 years ago

messages (and sometimes dialer) crashes

Reported by: yorickmoko@… Owned by: ainulindale
Priority: major Milestone:
Component: ophonekitd Version:
Keywords: Cc:

Description

I'm running a fully upgraded SHR on a gta02v5

Sometimes i can't open openmoko-messages, it starts loading and then i get an application execution error;
when i ssh i get this error:

root@om-gta02 ~ $ DISPLAY=:0 openmoko-messages
Trying to get the system bus
Adding signals.
(process:1271): DEBUG: phonegui_init()
(process:1271): DEBUG: Entering ecore loop
(process:1271): DEBUG: phonegui_messages_show()
(process:1271): DEBUG: Initiated elementary
(process:1271): DEBUG: Initiated etk
(process:1271): DEBUG: Added exit callback to ecore.
(process:1271): DEBUG: event_callback()
(process:1271): DEBUG: messages_event()
(process:1271): DEBUG: Event: 0
(process:1271): DEBUG: window_create()
(process:1271): DEBUG: Adding delete-request-callback
(process:1271): DEBUG: messages_event()
(process:1271): DEBUG: Event: 1
(process:1271): DEBUG: ogsmd_sim_retrieve_messagebook_callback()
Failed to handle dbus error: type: <class 'framework.resource.ResourceNotEnabled?'>
, 71 (dbus-glib-error-quark), code 32

I can't find a pattern; sometimes it happens after hours of stability, sometimes only 15minutes after a reboot. It also happens without an opkg upgrade.

i'll attach my logread, ophonekitd and x log

killall -9 ophonekitd; DISPLAY=":0" ophonekitd fixes it

Attachments

ophonekitd.log Download (2.9 KB) - added by yorickmoko@… 6 years ago.
logread.txt Download (64.0 KB) - added by yorickmoko@… 6 years ago.
x.log Download (7.7 KB) - added by yorickmoko@… 6 years ago.
ps_aux Download (5.0 KB) - added by hiciu 6 years ago.
output of ps aux
frameworkd.log Download (31.6 KB) - added by hiciu 6 years ago.
/var/log/frameworkd.log
ophonekitd.2.log Download (3.0 KB) - added by hiciu 6 years ago.
/var/log/ophoned.log

Change History

Changed 6 years ago by yorickmoko@…

Changed 6 years ago by yorickmoko@…

Changed 6 years ago by yorickmoko@…

comment:1 Changed 6 years ago by ainulindale

  • Status changed from new to accepted

Here, ophonekitd didn't manage to request the GSM resource. I added a timeout for that. Please update and try.

comment:2 Changed 6 years ago by ainulindale

  • Component changed from SHR Image to ophonekitd

comment:3 Changed 6 years ago by hiciu

I got this too. 2 restarts and still nothing. Clean image downloaded 2 hours ago, after opkg update; opkg upgrade; restart still the same.

root@om-gta02 ~ $ DISPLAY=:0 openmoko-messages
Trying to get the system bus
Adding signals.
** (process:1303): DEBUG: phonegui_init()
** (process:1303): DEBUG: Entering ecore loop
** (process:1303): DEBUG: phonegui_messages_show()
** (process:1303): DEBUG: Initiated elementary
** (process:1303): DEBUG: Initiated etk
** (process:1303): DEBUG: Added exit callback to ecore.
** (process:1303): DEBUG: event_callback()
** (process:1303): DEBUG: messages_event()
** (process:1303): DEBUG: Event: 0
** (process:1303): DEBUG: window_create()
** (process:1303): DEBUG: Adding delete-request-callback
** (process:1303): DEBUG: messages_event()
** (process:1303): DEBUG: Event: 1
** (process:1303): DEBUG: ogsmd_sim_retrieve_messagebook_callback()
Failed to handle dbus error: type: <class 'framework.resource.ResourceNotEnabled
'>
, 71 (dbus-glib-error-quark), code 32
root@om-gta02 ~ $ DISPLAY=:0 openmoko-contacts
Trying to get the system bus
Adding signals.
** (process:1305): DEBUG: phonegui_init()
** (process:1305): DEBUG: Entering ecore loop
** (process:1305): DEBUG: phonegui_contacts_show()
** (process:1305): DEBUG: Initiated elementary
** (process:1305): DEBUG: Initiated etk
** (process:1305): DEBUG: Added exit callback to ecore.
** (process:1305): DEBUG: event_callback()
** (process:1305): DEBUG: contacts_event(), event: 0
** (process:1305): DEBUG: window_create()
** (process:1305): DEBUG: Adding delete-request-callback
** (process:1305): DEBUG: contacts_event(), event: 1
** (process:1305): DEBUG: ogsmd_sim_retrieve_phonebook_callback()
Failed to handle dbus error: type: <class 'framework.resource.ResourceNotEnabled
'>
, 71 (dbus-glib-error-quark), code 32

Changed 6 years ago by hiciu

output of ps aux

Changed 6 years ago by hiciu

/var/log/frameworkd.log

Changed 6 years ago by hiciu

/var/log/ophoned.log

comment:4 Changed 6 years ago by anonymous

I see something similiar. Contacts usually crashes when I start it, dialer crashes when it actually tries to dial a number. Sometimes they work, but usually not. The last time I noticed that ophonekitd wasn't running, but starting it didn't help. The others still crash.
I did a opkg upgrade today (nov. 28.) before testing.
Dialer.log says:
Error Information:

An exit code of 1 was returned from openmoko-dialer -s.

Output Data:

There was no output

Error Logs:

There was no error message

comment:5 Changed 6 years ago by helge.hafting@…

More information:
The dialer dials out fine, and I get a connection.
After a little while, the call is cut off. Probably at the same time as the backlight times out. A ps aux shows that ophonekitd has died. Also, attempting to redial from the dialer then cause the dialer to die.

After this, the dialer dies on every press of the "call" button, even if I restart ophonekitd

comment:6 Changed 6 years ago by helge.hafting@…

More output from the ophonekitd crash:
root@om-gta02 ~ $ DISPLAY=:0 ophonekitd
(process:1629): DEBUG: phonegui_init()
(process:1629): DEBUG: Entering ecore loop
(process:1629): DEBUG: Phonegui initiated
(process:1629): DEBUG: Entering glib main loop
(process:1629): DEBUG: /org/shr/ophonekitd/Usage
Trying to get the system bus
Adding signals.
Added sim AuthStatus?.
Added sim IncomingStoredMessage?.
Added call CallStatus?.
Added device Idle Notifier State.
(process:1629): DEBUG: Connected to the buses
(process:1629): DEBUG: list_resources()
(process:1629): DEBUG: list_resources_callback()
(process:1629): DEBUG: Resource WiFi? available
(process:1629): DEBUG: Resource GSM available
(process:1629): DEBUG: Request GSM resource
(process:1629): DEBUG: Initiated elementary
(process:1629): DEBUG: Initiated etk
(process:1629): DEBUG: Added exit callback to ecore.
(process:1629): DEBUG: idle notifier state handler called, id 0
(process:1629): DEBUG: request_resource_callback()
(process:1629): DEBUG: call ogsmd_device_set_antenna_power()
(process:1629): DEBUG: ophonekitd_sim_auth_status_handler()
(process:1629): DEBUG: sim ready
(process:1629): DEBUG: idle notifier state handler called, id 1
(process:1629): DEBUG: power_up_antenna_callback()
(process:1629): DEBUG: idle notifier state handler called, id 0
(process:1629): DEBUG: register_to_network_callback()
(process:1629): DEBUG: get_messagebook_info_callback()
(process:1629): DEBUG: messagebook info: first: 1, last 20, used: 17, total 20
(process:1629): DEBUG: idle notifier state handler called, id 1
(process:1629): DEBUG: idle notifier state handler called, id 0
(process:1629): DEBUG: idle notifier state handler called, id 1
(process:1629): DEBUG: idle notifier state handler called, id 2
(process:1629): DEBUG: idle notifier state handler called, id 3
(process:1629): DEBUG: idle notifier state handler called, id 4
(process:1629): DEBUG: idle notifier state handler called, id 5
Unknown internal dbus error: Method "GetPowerStatus?" with signature "" on interface "org.freesmartphone.Device.PowerSupply?" doesn't exist
, 76 (dbus-glib-error-quark), code 19

and then it was dead.
I use shr upgraded today, and the kernel 2.6.28-GTA02_andy-tracking_de248ab69418d52b-mokodev

comment:7 Changed 6 years ago by mrmoku

The frameworkd version used in the SHR images is not ready for the new kernel. That explains the crash of ophonekitd Helge is seeing.

For the first part... openmoko-messages just crashing, when the SIM is not ready should be handled more nicely with some error notice

comment:8 in reply to: ↑ description Changed 5 years ago by johnsu01

I am seeing this as well. ophonekitd dies after trying to use the dialer or the messages application. This is with the most recent andy-tracking kernel and the shr-testing image.

comment:9 Changed 5 years ago by ainulindale

  • Status changed from accepted to closed
  • Resolution set to fixed

You shouldn't use andy-tracking against SHR. We use a specific set of versions.
For testing, this is normal. Please try latest unstable.

Note: See TracTickets for help on using tickets.