Bug#921266: Could this simple typo be the reason: "ream" <-> "realm"?
Hi Bernhard Am 11.03.19 um 00:11 schrieb Bernhard Schmidt: > Control: severity -1 important Control: tags -1 moreinfo > > Hi Alf, > > First of all, thanks a lot for your report. > > Based on your feedback of only affecting t-online.de configurations > and also happening in the official Flatpak version (thus not being > Debian specific) I'm downgrading the severity again. Sorry, you might have got something wrong: I never said that it only affects Telekom - I can only check and confirm that it fails with that SIP-provider because it is the only one I have access to! I have been running Linphone since Wheezy and Jessie and now in Stretch without any problems with Telekom as SIP-Provider. And Telekom is by far the most popular SIP-provider here in Germany. Aren't there any people/users to just check functionalty with other SIP-providers? Without Linphone I'll be unable to use Buster on my Laptop as a softphone, because also from remote I use my local internet connection for SIP via OpenVPN. Additionally the only common in Debian-version and official flatpak-version is that it fails with segfailt in libc6. Addresses and output (IP SP, message) are different. In Debian-version they are absolutely constant, no matter if using GUI or CLI in debian-version. > > The ream <-> realm typo is just a typo as far as I can tell. It > actually comes from here: > > https://github.com/BelledonneCommunications/belle-sip/blob/master/src/provider.c#L234 > > There have been quite a few commits to that file since it has been > released > > https://github.com/BelledonneCommunications/belle-sip/commits/master/src/provider.c > > Release 1.6.3 which is imported in Debian has been tagged on > 21.7.2017, so there are quite a few commits to check on. And the > totally screwed up handling of the codebase by upstream, with dozens > of interdependent libraries that need to be upgraded in lockstep and > have not seen tagged release for almost two years, does not make it > any easier. > > You could try the daily snapshot of the official flatpak, which > should be available here > > flatpak --user install --from > https://www.linphone.org/snapshots/flatpak-testing-repo/linphone.flatpakref > > [totally untested, I don't use flatpak] I tried to setup the daily build, but it totally fails despite some 2GB of environment to fulfill all dependencies. Flatpak really is not an option to run such applications. In case it does not get fixed in Buster, I would prefer to run Stretch with version 3.6.1 in a VM. Regards, Alf
Bug#921266: Could this simple typo be the reason: "ream" <-> "realm"?
Control: severity -1 important Control: tags -1 moreinfo Hi Alf, > I now tried to collect debug info with the cli-version - it crashes the > same way with segfault: > > linphonec -d 5 -l linphone-debug > > the last lines in the debug output after password enty and before crash are: > > Authentication-Info: > qop=auth,rspauth="4aa9ccf88b3958bb460174863eccbb82",cnonce="0us2EQTtqGWgU6LP",nc=0001 > > > 2019-03-08 22:17:07:834 ortp-message-channel [0x55f42b920dd0] [721] > bytes parsed > 2019-03-08 22:17:07:834 ortp-message-Found transaction matching response. > 2019-03-08 22:17:07:834 ortp-message-Updating auth context for ream > [tel.t-online.de] next nonce is going to be [(null)] > > Please note the very last line: > > it wants to update auth for "ream" > instead of "realm" > > I cannot decide if this is just a typo in text output or actually a > wrong named parameter. This auth cycle is the second one which probably > only is used by German Telekom? > > If more information is needed, please let me know > > Fopr completeness I do attach the whole debug output as a plain txt-file > named linphone-debug. First of all, thanks a lot for your report. Based on your feedback of only affecting t-online.de configurations and also happening in the official Flatpak version (thus not being Debian specific) I'm downgrading the severity again. The ream <-> realm typo is just a typo as far as I can tell. It actually comes from here: https://github.com/BelledonneCommunications/belle-sip/blob/master/src/provider.c#L234 There have been quite a few commits to that file since it has been released https://github.com/BelledonneCommunications/belle-sip/commits/master/src/provider.c Release 1.6.3 which is imported in Debian has been tagged on 21.7.2017, so there are quite a few commits to check on. And the totally screwed up handling of the codebase by upstream, with dozens of interdependent libraries that need to be upgraded in lockstep and have not seen tagged release for almost two years, does not make it any easier. You could try the daily snapshot of the official flatpak, which should be available here flatpak --user install --from https://www.linphone.org/snapshots/flatpak-testing-repo/linphone.flatpakref [totally untested, I don't use flatpak] If you can still reproduce it, please report upstream at http://lists.nongnu.org/archive/html/linphone-users/ . Bernhard
Bug#921266: Could this simple typo be the reason: "ream" <-> "realm"?
I now tried to collect debug info with the cli-version - it crashes the same way with segfault: linphonec -d 5 -l linphone-debug the last lines in the debug output after password enty and before crash are: Authentication-Info: qop=auth,rspauth="4aa9ccf88b3958bb460174863eccbb82",cnonce="0us2EQTtqGWgU6LP",nc=0001 2019-03-08 22:17:07:834 ortp-message-channel [0x55f42b920dd0] [721] bytes parsed 2019-03-08 22:17:07:834 ortp-message-Found transaction matching response. 2019-03-08 22:17:07:834 ortp-message-Updating auth context for ream [tel.t-online.de] next nonce is going to be [(null)] Please note the very last line: it wants to update auth for "ream" instead of "realm" I cannot decide if this is just a typo in text output or actually a wrong named parameter. This auth cycle is the second one which probably only is used by German Telekom? If more information is needed, please let me know Fopr completeness I do attach the whole debug output as a plain txt-file named linphone-debug. Debug output from cli-version when trying to register. Startup until crash with segfault. linphonec -d 5 -l linphone-debug 2019-03-08 22:16:46:419 ortp-message-Using (r/w) config information from /home/ingo/.linphonerc 2019-03-08 22:16:46:419 ortp-message-Initializing LinphoneCore 3.12.0 2019-03-08 22:16:46:419 ortp-message-Core callbacks [0x55f42b819260] registered on core [0x55f42b824480] 2019-03-08 22:16:46:420 ortp-message-Core callbacks [0x55f42b818310] registered on core [0x55f42b824480] 2019-03-08 22:16:46:420 ortp-message-oRTP-1.0.2 initialized. 2019-03-08 22:16:46:420 ortp-message-Mediastreamer2 factory 2.16.1 (git: unknown) initialized. 2019-03-08 22:16:46:420 ortp-message-CPU count set to 4 2019-03-08 22:16:46:420 ortp-message-ms_factory_init() done: platform_tags=linux,x86,desktop 2019-03-08 22:16:46:420 ortp-message-srtp init 2019-03-08 22:16:46:420 ortp-message-Registering all soundcard handlers 2019-03-08 22:16:46:421 ortp-message-New PulseAudio context state: PA_CONTEXT_CONNECTING 2019-03-08 22:16:46:421 ortp-message-New PulseAudio context state: PA_CONTEXT_AUTHORIZING 2019-03-08 22:16:46:421 ortp-message-New PulseAudio context state: PA_CONTEXT_SETTING_NAME 2019-03-08 22:16:46:422 ortp-message-New PulseAudio context state: PA_CONTEXT_READY 2019-03-08 22:16:46:422 ortp-message-Card 'PulseAudio: Internes Audio Analog Stereo' added with capabilities [capture, playback] 2019-03-08 22:16:46:422 ortp-message-Card 'PulseAudio: Webcam C270 Analog Mono' added with capabilities [capture] 2019-03-08 22:16:46:437 ortp-message-Card 'ALSA: default device' added with capabilities [capture, playback] 2019-03-08 22:16:46:437 ortp-message-Registering all webcam handlers 2019-03-08 22:16:46:505 ortp-message-Webcam V4L2: /dev/video0 added 2019-03-08 22:16:48:980 ortp-message-Webcam V4L2: /dev/video2 added 2019-03-08 22:16:48:980 ortp-message-Webcam StaticImage: Static picture added 2019-03-08 22:16:48:980 ortp-message-ms_factory_init_voip() done 2019-03-08 22:16:48:980 ortp-message-Loading ms plugins from [/usr/lib/x86_64-linux-gnu/mediastreamer/plugins] 2019-03-08 22:16:48:980 ortp-message-Cannot open directory /usr/lib/x86_64-linux-gnu/mediastreamer/plugins: No such file or directory 2019-03-08 22:16:48:980 ortp-message-Codec opus/48000 fmtp=[useinbandfec=1] number=-1, default enablement: 1) added to the list of possible codecs. 2019-03-08 22:16:48:980 ortp-message-Codec SILK/16000 fmtp=[] number=-1, default enablement: 1) added to the list of possible codecs. 2019-03-08 22:16:48:980 ortp-message-Codec speex/16000 fmtp=[vbr=on] number=-1, default enablement: 1) added to the list of possible codecs. 2019-03-08 22:16:48:980 ortp-message-Codec speex/8000 fmtp=[vbr=on] number=-1, default enablement: 1) added to the list of possible codecs. 2019-03-08 22:16:48:980 ortp-message-Codec PCMU/8000 fmtp=[] number=0, default enablement: 1) added to the list of possible codecs. 2019-03-08 22:16:48:980 ortp-message-Codec PCMA/8000 fmtp=[] number=8, default enablement: 1) added to the list of possible codecs. 2019-03-08 22:16:48:980 ortp-message-Codec red/1000 fmtp=[] number=-1, default enablement: 1) added to the list of possible codecs. 2019-03-08 22:16:48:980 ortp-message-Codec t140/1000 fmtp=[] number=-1, default enablement: 1) added to the list of possible codecs. 2019-03-08 22:16:48:980 ortp-message-Codec GSM/8000 fmtp=[] number=3, default enablement: 0) added to the list of possible codecs. 2019-03-08 22:16:48:981 ortp-message-Codec G722/8000 fmtp=[] number=9, default enablement: 0) added to the list of possible codecs. 2019-03-08 22:16:48:981 ortp-message-Codec iLBC/8000 fmtp=[mode=30] number=-1, default enablement: 0) added to the list of possible codecs. 2019-03-08 22:16:48:981 ortp-message-Codec AMR/8000 fmtp=[octet-align=1] number=-1, default enablement: 0) added to the list of