Ticket #371 (closed defect: fixed)
Incoming call never stops
| Reported by: | daniel | Owned by: | mickey |
|---|---|---|---|
| Priority: | major | Milestone: | |
| Component: | framework/ogsmd | Version: | milestone5 |
| Keywords: | Cc: |
Description
I just encountered something weird. Went away for a bit and after I came back my FR was ringing (actually vibrating) and showing an incoming call.
This call woke up my phone about twenty minutes before I got back, but the log shows that it never got the hangup %CPI.
The image is from before the new timeout code got in.
root@om-gta02:~# opkg status frameworkd Package: frameworkd Version: 0.8.4.9+gitr1169+423a2cfc74902c32a0dcd447c5e98364ce72dba1-r5 Depends: python-ctypes, python-dbus, python-datetime, python-difflib, python-logging, python-pprint, python-pyalsaaudio, python-pygobject, python-pyrtc, python-pyserial, python-pyyaml, python-shell, python-subprocess, python-syslog, python-textutils, frameworkd-config, fso-gsm0710muxd, wmiconfig Recommends: alsa-utils-amixer, python-gst, ppp Status: install ok installed Architecture: armv4t root@om-gta02:~# opkg status fso-gsm0710muxd Package: fso-gsm0710muxd Version: 0.9.3.1+gitr24+abcbcd7cc532a8834906de3fc24c8f8fe7643cd4-r0 Depends: dbus, libdbus-glib-1-2, libdbus-glib-1-2 (>= 0.76), libdbus-1-3 (>= 1.2.1), libgobject-2.0-0 (>= 2.16.4), libglib-2.0-0 (>= 2.16.4), libc6 (>= 2.6.1) Replaces: gsm0710muxd Conflicts: gsm0710muxd Status: install ok installed Architecture: armv4t
Attachments
Change History
Changed 4 years ago by daniel
- Attachment frameworkd.log added
comment:1 Changed 4 years ago by shishz
I think I experienced the same thing yesterday with my gta02 with milestone 5.1. The phone started ringing in my pack and I decided to let it go to voicemail... it never stopped ringing or vibrating and the screen displayed the incoming call, and hitting the reject button did nothing. I was able to navigate the gui and shutdown the phone though.
comment:2 Changed 4 years ago by daniel
Okay, did this happen again for anyone? Things have been pretty stable here. I think we have a time out for ringing now even on the ti calypso, right? (mickey?)
Why the %CPI should be missing is still strange, though. Maybe another manifestation of the muxer issue with large buffers?
comment:3 Changed 4 years ago by mickey
- Status changed from new to closed
- Resolution set to fixed
Ok, a fresh new analysis shows the problem. At wakeup we issue the resume commands on the unsolicited channel. The channel breaks down since it gets a response intermixed with unsolicited ringing. At 13:21:34 it finally times out and (for some reason) no longer operates the parser. This the reason why we miss all following messages -- including %CPI -- on the unsolicited channel.
Although I don't know why the channel really hangs, this very problem can no longer happen with the new parser.

frameworkd log of the issue. The phone was still vibrating at 13:41:13 when I just yanked the battery.