Ticket #363 (closed defect: fixed)

Opened 5 years ago

Last modified 5 years ago

frameworkd tracebacks when using shr-messages

Reported by: bembel Owned by: daniel
Priority: major Milestone:
Component: Frameworkd / FSO Version: SHR-testing
Keywords: crash retrieve messagebook callback(error=192304) Cc:

Description

99% openmoko-message crashs
#cat /etc/version
200903030643
daily opkg update && opkg upgrade!

(process:1591): DEBUG: Trying to get the system bus
(process:1591): DEBUG: Adding signals.
(process:1591): DEBUG: Entering ecore loop
(process:1591): DEBUG: phonegui_messages_show()
(process:1591): DEBUG: Initiated elementary
(process:1591): DEBUG: Initiated etk
(process:1591): DEBUG: event_callback()
(process:1591): DEBUG: _show()
(process:1591): DEBUG: window_init(win=75936)
(process:1591): DEBUG: window_view_show()
(process:1591): DEBUG: window_view_hide()
(process:1591): DEBUG: message_list_view_show()
(process:1591): DEBUG: tree init
(process:1591): DEBUG: ogsmd_sim_retrieve_messagebook_callback()
(process:1591): DEBUG: retrieve messagebook callback(error=192304)
Segmentation fault

Change History

comment:1 Changed 5 years ago by KaZeR

More details here :  http://n2.nabble.com/Draft-release-notes-started-tp2569007p2573772.html, and the two following posts.

comment:2 Changed 5 years ago by ainulindale

  • Owner changed from ainulindale to quickdev
  • Component changed from openmoko-messages3 to libframeworkd-phonegui-efl

comment:3 Changed 5 years ago by skamster

Traceback (most recent call last):

File "/usr/lib/python2.6/site-packages/framework/subsystems/ogsmd/gsm/channel.py", line 414, in handleResponseToRequest

ok_cb( reqstring.strip(), response )

File "/usr/lib/python2.6/site-packages/framework/subsystems/ogsmd/modems/abstract/mediator.py", line 999, in responseFromChannel

result.append( ( index, status, str(sms.addr), sms.ud, sms.properties ) )

File "/usr/lib/python2.6/site-packages/framework/subsystems/ogsmd/gsm/sms.py", line 471, in _getProperties

map.update( self._get_udh() )

File "/usr/lib/python2.6/site-packages/framework/subsystems/ogsmd/gsm/sms.py", line 326, in _get_udh

mapcsm_id? = self.udh[0][0]*256 + self.udh[0][1]

KeyError?: 0

this is my log which i take with
tail -f /var/log/frameworkd.log
i see the application correctly, first, but without any messenges. this is normal, in the first testing version, it was the same, after a while, it load them.

but now, when it's should load, the application crashes.

hope, you could fix this asap, i don't know the problems of other persons, but for me is this one of the greatest bug's. ;)

comment:4 Changed 5 years ago by skamster

the same with a bit other lines in the unstable-version from the 19. april 09..

File "/usr/lib/python2.6/site-packages/framework/subsystems/ogsmd/gsm/channel.py", line 430, in handleResponseToRequest

ok_cb( reqstring.strip(), response )

comment:5 Changed 5 years ago by daniel

The problem here is the SMS parsing code. Could you please get the offending PDU and post it?

login and run mickeyterm, then issue

AT+CMGL=4 which will show you all the messages as PDUs.
Please post them here (or if you're concerned about privacy mail them directly to daniel@…).

comment:6 Changed 5 years ago by ainulindale

  • Status changed from new to assigned
  • Owner changed from quickdev to daniel
  • Component changed from libframeworkd-phonegui-efl to frameworkd / FSO

comment:7 Changed 5 years ago by KaZeR

I think there might be also another root to the crash.

I'm having some issues with the framework stability (i guess it's framework, but don't take it for granted :) ) since a few days.

The issue seems to be close to the one i reported some times ago.

Current typical use case : i'm trying to opkg upgrade via gprs. opkg tries to download the new rev of navit, which weights about 5Mb. After 500kb of download, the download is stalled (same things happens with wget) and never continues.

Trying another wget leads to a dns resolving issue.

If i break the download and disconnect from gprs, i get :

Apr 30 12:36:57 om-gta02 daemon.info pppd[1533]: Terminating on signal 15
Apr 30 12:36:57 om-gta02 daemon.info pppd[1533]: Connect time 31.9 minutes.
Apr 30 12:36:57 om-gta02 daemon.info pppd[1533]: Sent 29640 bytes, received 731418 bytes.
Apr 30 12:36:57 om-gta02 daemon.notice pppd[1533]: restoring old default route to usb0 [192.168.0.200]
Apr 30 12:36:57 om-gta02 daemon.debug pppd[1533]: Script /etc/ppp/ip-down started (pid 1647)
Apr 30 12:36:57 om-gta02 daemon.debug pppd[1533]: sent [LCP TermReq? id=0x3 "User request"]
Apr 30 12:36:57 om-gta02 daemon.debug pppd[1533]: Script /etc/ppp/ip-down finished (pid 1647), status = 0x0
Apr 30 12:37:00 om-gta02 daemon.debug pppd[1533]: sent [LCP TermReq? id=0x4 "User request"]
Apr 30 12:37:03 om-gta02 daemon.notice pppd[1533]: Connection terminated.
Apr 30 12:37:03 om-gta02 daemon.info pppd[1533]: Serial link disconnected.
Apr 30 12:37:03 om-gta02 daemon.notice pppd[1533]: Modem hangup
Apr 30 12:37:04 om-gta02 daemon.info pppd[1533]: Exit.
Apr 30 12:37:04 om-gta02 local0.info /usr/sbin/gsm0710muxd[1515]: gsm0710muxd.c:542:logical_channel_close(): Logical channel 4 for ogsmd.gprs closing
Apr 30 12:37:20 om-gta02 local0.warn /usr/sbin/gsm0710muxd[1515]: gsm0710muxd.c:562:logical_channel_close(): Unable to properly close a channel

If i try to reconnect :
Apr 30 12:37:53 om-gta02 local0.info /usr/sbin/gsm0710muxd[1515]: gsm0710muxd.c:729:c_alloc_channel(): Connecting /dev/pts/5 to virtual channel 4 for ogsmd.gprs on /dev/ttySAC0
Apr 30 12:38:09 om-gta02 local0.info /usr/sbin/gsm0710muxd[1515]: gsm0710muxd.c:748:c_alloc_channel(): Unable to open the new channel 4

If i try to launch openmoko messages :

Apr 30 13:18:03 om-gta02 user.warn kernel: [ 1758.775000] Alignment trap: openmoko-messag (1703) PC=0x40106104 Instr=0xe7920104 Address=0x00000001 FSR 0x013

If i try to place an outgoing call, the dialer disappears but the call never starts.

Only solution is to reboot the phone.

It happens for me with testing from 2009/04/22.

comment:8 Changed 5 years ago by spaetz

Hi KaZeR,
this is definitely a different bug than sms app crashes. You are experiencing a modem dies after GPRS transmitted a larger amount of data. And I think i have seen someone mentioning this previously. Would you mind filing this as a separate bug? Also jsut to make sure, you do have the latest GSM firmware moko11, right?

comment:9 Changed 5 years ago by KaZeR

Indeed, i was pointed to a bug on the OM tracker via the mailing list, which seems closer to my issue. Please ignore my previous comments (and i do have moko11, as my recent sim didn't work at all with the previous firmware).

comment:10 Changed 5 years ago by dos

  • Component changed from frameworkd / FSO to libframeworkd-phonegui-efl

Is is still happening?

comment:11 Changed 5 years ago by dos

  • Component changed from libframeworkd-phonegui-efl to frameworkd / FSO

comment:12 Changed 5 years ago by dos

  • Priority changed from blocker to critical
  • Summary changed from openmoko-message crashing to frameworkd tracebacks when using shr-messages

comment:13 Changed 5 years ago by dos

(15:19:20) alphaone: PaulFertser?: oh
(15:19:21) alphaone: shit!
(15:19:30) alphaone: Wait a second, I think I know
(15:20:20) alphaone: yes
(15:21:15) alphaone: if 8 in self.udh:
(15:21:15) alphaone: # Concatenated shor messages (16-bit reference)
(15:21:15) alphaone: mapcsm_id? = self.udh[0][0]*256 + self.udh[0][1]
(15:21:15) alphaone: mapcsm_num? = self.udh[0][2]
(15:21:15) alphaone: mapcsm_seq? = self.udh[0][3]
(15:21:23) alphaone: This of course cannot work
(15:21:37) alphaone: typical copy and paste mistake..
(15:21:48) alphaone: from if 0 in self.udh:
(15:21:48) alphaone: # UDH for concatenated short messages is a list of ID,
(15:21:48) alphaone: # total number of messages, position of message in csm
(15:21:48) alphaone: mapcsm_id? = self.udh[0][0]
(15:21:48) alphaone: mapcsm_num? = self.udh[0][1]
(15:21:49) alphaone: mapcsm_seq? = self.udh[0][2]
(15:22:51) alphaone: I'll give the author a good spanking

So it's going to be fixed soon :)

comment:14 Changed 5 years ago by dos

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

Fixed upstream, now just wait for package regeneration :)

comment:15 Changed 5 years ago by daniel

Should be fixed in framework commit 9e57ae9682b60cebc113f7a267d37245024aaefe

Note: See TracTickets for help on using tickets.