Occasional segmentation fault

[matrndev]

Hi, I run ha-sip on my HAOS server along with Asterisk on the same server. I have ha-sip set up in listen mode, after a webhook is recieved that a call is being made, an automation runs an stdin answer command to answer the call with TTS.

Occasionaly, before the TTS is played back, ha-sip hangs with a segmentation fault. I apologize for this being a screenshot, I do not have a different record of this issue.

[arnonym]

What platform is this on? AMD64, ARM?

[matrndev]

Sorry for the late response. This is ARM on an Odroid C4 with a pure install of HAOS on it.

[arnonym]

Can you please check, if this got better in the latest version?

[matrndev]

I'll let you know if it happens again in 3.8.

[lanwin]

I also have this problem on an Raspi 4. With homeassistant/armv7-base-python

2024-06-17T05:06:57.806056930Z /run.sh: line 55: 84 Segmentation fault (core dumped) python3 /ha-sip/main.py

[arnonym]

Is there are reason you're running armv7 instead of aarch64 on a raspi 4?

[lanwin]

Cause my host runs Raspi OS with armv7l.

[marvinhinz]

Same issue running latest HAOS inside a synology dsm (INTEL Celeron J4125) connected to a dahua doorbell. Running Version 3.9 This happens very rarely, when spamming the button of the doorbell to initiate calls and end them shortly after another.

Other than that, works very nice, thank you!

| 00:50:16.455788 [1] Call connected
| 00:50:16.476530 [1] Playing audio file: /config/sounds/elevator.mp3
| 00:50:16.477031 [ ] Using cache from file: /config/sounds/cache/257a4086a8.wav
00:50:16.477            pjsua_aud.c  Creating file player: /config/sounds/cache/257a4086a8.wav..
00:50:16.477           wav_player.c  .File player '/config/sounds/cache/257a4086a8.wav' created: samp.rate=44100, ch=2, bufsize=4KB, filesize=6704KB
00:50:16.477            pjsua_aud.c  .Player created, id=0, slot=2
00:50:16.477            pjsua_aud.c  Conf connect: 2 --> 1
00:50:16.477            pjsua_aud.c  .Set sound device: capture=-99, playback=-99, mode=0, use_default_settings=0
00:50:16.477            pjsua_aud.c  ..Null sound device, mode setting is ignored
00:50:16.477            pjsua_aud.c  ..Setting null sound device..
00:50:16.477            pjsua_aud.c  ...Opening null sound device..
00:50:16.477           conference.c  .Port 2 (/config/sounds/cache/257a4086a8.wav) transmitting to port 1 (sip:[email protected])
/run.sh: line 56:   228 Segmentation fault      (core dumped) python3 /ha-sip/main.py
s6-rc: info: service legacy-services: stopping

[fabian-st]

Encountering the same issue running HAOS on Proxmox (Intel processor). Happens almost every other time a message is being played (via Google TTS), but it did work fine until a few weeks ago.

| 23:40:30.850030 [ ] Got "dial" command for sip:[email protected]
| &id001
| id: null
| message: Der Alarm wurde erfolgreich deaktiviert.
| audio_file: null
| language: de
| action: null
| choices_are_pin: false
| choices: {}
| default_choice:
|   id: null
|   message: Unknown option
|   audio_file: null
|   language: en
|   action: null
|   choices_are_pin: false
|   choices: null
|   default_choice: null
|   timeout_choice: null
|   post_action:
|     action: return
|     level: 1
|   timeout: 300.0
|   parent_menu: *id001
|   cache_audio: false
|   wait_for_audio_to_finish: false
| timeout_choice:
|   id: null
|   message: null
|   audio_file: null
|   language: en
|   action: null
|   choices_are_pin: false
|   choices: null
|   default_choice: null
|   timeout_choice: null
|   post_action:
|     action: hangup
|   timeout: 300.0
|   parent_menu: *id001
|   cache_audio: false
|   wait_for_audio_to_finish: false
| timeout: 300.0
| post_action:
|   action: hangup
| parent_menu: null
| cache_audio: true
| wait_for_audio_to_finish: false
| 
| 23:40:30.852551 [1] Registering call with id sip:[email protected]
| 23:40:30.852590 [ ] Add to state: sip:[email protected]
23:40:30.852           pjsua_call.c  Making call with acc #0 to sip:[email protected]
23:40:30.852            pjsua_aud.c  .Set sound device: capture=-99, playback=-99, mode=0, use_default_settings=0
23:40:30.852            pjsua_aud.c  ..Null sound device, mode setting is ignored
23:40:30.852            pjsua_aud.c  ..Setting null sound device..
23:40:30.852            pjsua_aud.c  ...Opening null sound device..
23:40:30.852          pjsua_media.c  .Call 0: initializing media..
23:40:30.852                icetp00  ..Creating ICE stream transport with 2 component(s)
/run.sh: line 56:   228 Segmentation fault      (core dumped) python3 /ha-sip/main.py

[fabian-st]

Might be related to this: https://github.com/pjsip/pjproject/issues/3260 A possible fix has been proposed in the latest comment.

[arnonym]

Can you test with 4.1?

[arnonym]

Closed, as it seems to be much more stable for some people on other threads.