GNOME Bugzilla – Bug 586531
Varying problems with sound output (ring tone error, silence)
Last modified: 2009-09-09 08:57:23 UTC
(This report has first been sent to the mailing list, http://mail.gnome.org/archives/ekiga-list/2009-June/msg00128.html ) On Debian stable, I've recently installed gcc-4.4 from Debian testing (4.3.2 didn't compile it), and compiled ptlib-2.6.2.tar.bz2, opal-3.6.2.tar.bz2, and first tried ekiga-3.2.4.tar.gz but then went with the git "gnome-2-26" branch (because of a compiler error fixed there). This compiled without problems, but the resulting Ekiga would frequently give no sound output when connecting to echo services. I call sip:500@ekiga.net, but it seems ekiga.net prohibits too frequent usage of the echo test(?). When re-calling too soon, Ekiga turns the contents of the URL entry field into "pc:udp$ekiga.net:5060" and will never get to show the "Connected .. call duration.." message. But, there is also a problem with sound not playing in many cases where I *do* have a "Connected .. call duration.." message on the screen. The problem does not go away even when waiting between calls (like a minute or so). A sequence of connection initiations by me: sip:500@ekiga.net -> normal with sound sip:500@ekiga.net -> stays in pc:udp$ekiga.net:5060, i.e. does not get connection, I click stop sip:441@sip.diamondcard.us -> silent, I click stop sip:441@sip.diamondcard.us -> normal with sound or with only using the diamondcard echo service: sip:441@sip.diamondcard.us -> normal sip:441@sip.diamondcard.us -> silent sip:441@sip.diamondcard.us -> normal sip:441@sip.diamondcard.us -> silent sip:441@sip.diamondcard.us -> silent sip:441@sip.diamondcard.us -> silent But sound problems don't stop here, when I call my Diamondcard number using my mobile, Ekiga will pop up messages (one for each suppressed ring tone or so) even before I try to click on "accept", saying "Error while opening audio output device HDA Intel No incoming sound will be played. The selected audio device was successfully opened but it is impossible to write data to this device. In case it is a pluggable device it may be sufficient to reconnect it. If not, or if it still is not accessible, please check your audio setup. " but in preferences I've set all three audio devices to "HDA Intel (PTLIB/ALSA)". Also sound has always worked without problems on this Laptop, and Ekiga 2.0.12 (from Debian stable) never had this problem. I've then recompiled/reinstalled everything from SVN / Git, with the versions from the stable branches as given below, with the same results. To summarize, there seem to be 3 separate issues, the first is with ekiga.net not accepting frequent calls of the echo service (and Ekiga not showing an error in this case), the second that about half the otherwise seemingly successful echo service connections are silent, the third that the ring tone is never given (but an error message instead). Details: Lenovo Thinkpad T61, core 2 duo, nvidia. Versions: Ekiga 44ebfd82f1a4254d3fda3de0871b9c62e7e7c842 ptlib opalvoip/ptlib/branches/v2_6@22848 opal opalvoip/opal/branches/v3_6@22931 $ gcc-4.4 --version gcc-4.4 (Debian 4.4.0-6) 4.4.0 $ g++-4.4 --version g++-4.4 (Debian 4.4.0-6) 4.4.0 Debian stable, kernel 2.6.29.5 All 3 libs/programs compiled using: export PKG_CONFIG_PATH=~/lib/pkgconfig/ export PTLIBDIR=`pwd` export CC=gcc-4.4 export CXX=g++-4.4 ./configure --prefix=$HOME make -j2 make install Configuration results: opal: GSM : yes (system) gsm-amr : yes iLBC : yes (internal) Speex : yes (internal, version 1.1.11.1) LPC10 : yes G.726 : yes IMA_ADPCM : yes SBC : no H.261 : yes H.263/H.263+ : H.264 : no THEORA : no MPEG4 Part 2 : no CELT : no SpanDSP FAX : no CAPI : no Quicknet xJACK support : no Voicetronics VPB support : yes OS Type : linux-gnu Machine Type : x86_64 ekiga: GNOME support : enabled GConf support : enabled Evolution-Data-Server support : enabled NOTIFY support : disabled LDAP support : enabled XVideo support : enabled H.323 support : yes SIP support : yes DBUS support : enabled DBUS service support : enabled mDNS/DNS-SD support : enabled The following features are experimental -- you've been warned: GStreamer support : disabled KAddressBook support : disabled KDE support : disabled XCAP support : disabled OS Type : linux-gnu Machine Type : x86_64 Byte Order : little endian
strace -f -o _str ekiga stream-cut 2525027 3655613 _str > _a stream-cut 3655613 4599014 _str > _b now _a contains the part from the start of a successful call, _b contains the part of the second call after startup which is always silent. $ grep -F ' open("/dev/' _a 32686 open("/dev/snd/controlC0", O_RDONLY) = 40 32686 open("/dev/snd/controlC0", O_RDWR) = 40 32686 open("/dev/snd/pcmC0D0p", O_RDWR|O_NONBLOCK) = 57 32686 open("/dev/snd/controlC0", O_RDONLY) = 40 32686 open("/dev/snd/controlC0", O_RDWR) = 40 32686 open("/dev/snd/controlC0", O_RDONLY) = 40 32686 open("/dev/snd/controlC0", O_RDWR) = 40 32686 open("/dev/snd/pcmC0D0c", O_RDWR|O_NONBLOCK) = 62 32686 open("/dev/snd/controlC0", O_RDONLY) = 40 32686 open("/dev/snd/controlC0", O_RDWR) = 40 $ grep -F ' open("/dev/' _b $ So Ekiga doesn't even try to open the sound devices in the silent case. And chrisekiga@novo:~/bin$ egrep ' (open\(\"/dev/|close\()' _a|uniq -c 1 32663 close(40) = 0 1 32746 close(58 <unfinished ...> 38 32746 close(58) = 0 1 32746 close(59) = 0 32 32746 close(58) = 0 1 32746 close(59) = 0 29 32746 close(60) = 0 1 32746 close(61) = 0 37 32746 close(62) = 0 1 32746 close(40) = 0 1 32746 close(57) = 0 10 32682 close(40) = 0 118 32686 close(40) = 0 10 32682 close(40) = 0 64 32686 close(40) = 0 1 32686 open("/dev/snd/controlC0", O_RDONLY) = 40 1 32686 close(40) = 0 1 32686 open("/dev/snd/controlC0", O_RDWR) = 40 1 32686 open("/dev/snd/pcmC0D0p", O_RDWR|O_NONBLOCK) = 57 1 32686 close(40) = 0 1 32686 open("/dev/snd/controlC0", O_RDONLY) = 40 1 32686 close(40) = 0 1 32686 open("/dev/snd/controlC0", O_RDWR) = 40 1 32686 close(40) = 0 1 32686 open("/dev/snd/controlC0", O_RDONLY) = 40 1 32686 close(40) = 0 1 32686 open("/dev/snd/controlC0", O_RDWR) = 40 1 32686 open("/dev/snd/pcmC0D0c", O_RDWR|O_NONBLOCK) = 62 1 32686 close(40) = 0 1 32686 open("/dev/snd/controlC0", O_RDONLY) = 40 1 32686 close(40) = 0 1 32686 open("/dev/snd/controlC0", O_RDWR) = 40 1 32686 close(40) = 0 1 32750 close(70 <unfinished ...> 1 32657 close(57 <unfinished ...> 1 32750 close(70) = 0 16 32750 close(57) = 0 1 32750 close(57 <unfinished ...> 1 32657 close(62 <unfinished ...> 1 32750 close(57 <unfinished ...> 2 32750 close(57) = 0 1 32750 close(62) = 0 1 32657 close(64) = 0 1 32657 close(65) = 0 21 32750 close(57) = 0 1 32750 close(40) = 0 1 32750 close(63) = 0 3 32750 close(40) = 0 5 32682 close(40) = 0 1 32750 close(68 <unfinished ...> 1 32750 close(69 <unfinished ...> 1 32657 close(40) = 0 1 32662 close(66) = 0 1 32662 close(67) = 0 1 32662 close(58) = 0 1 32662 close(59) = 0 1 32662 close(60) = 0 1 32662 close(61) = 0 So it looks like it closed all filehandles from the first call before making the second call.
Hello Christian, Sorry for the late answer. Let's take in order: - URI changes to pc:udp$ekiga.net:5060 - this is fixed in ekiga 3.2.5 (latest stable), could you confirm? - for the second problem, I have it sometimes too - for the sound problem, it's the bug http://bugzilla.gnome.org/show_bug.cgi?id=588532, could you confirm?
With the recent Michael's commit, the 2nd problem is fixed too. So this bug could be closed, please reopen if not.
http://git.gnome.org/cgit/ekiga/commit/?h=gnome-2-26&id=39190d030c7784efdff243d98024812cb9bac3ec
*** Bug 592479 has been marked as a duplicate of this bug. ***
I still have the bug: sometimes call to echo test has no sound output (the icon for sound volume output in ekiga is greyed out) and the rest of time it works. It happen quite often: ~50% ptlib/opal revesion: 23260 ekiga git commit: 9c6bf9b02835a18371ce36f414fb1892af34aa53 from SatAug22 22:31:53 2009+0200 I use thoses packages I just made for ubuntu: https://launchpad.net/~sevmek/+archive/ekiga-stable-prerelease Best regards, Yannick
Can you check the debug output to see if you see something relevant ?
Created attachment 141875 [details] -d 4
Hi, I found a better way to express the behavior: In the above -d 4, I called 500, everything was working fine. I hang up, then called 500 again. This second time: no sound. When i tryed to access the volume control in Ekiga the *output* icon was greyed and put to 0. Now, while in the same second call to 500, I opened the prefs; the devices were set to default, i switched to silent, then switched to Default again and the sound get back (while in the call). From that experience (the same trick with prefs also gived me the sound back while in a call with eugen), I suspect Ekiga change the device for output sound at the very begining of a new call to "Silent" (or something with the same effect), still the prefs do not reflect this change. I did not have time yet to search for that in the debug output... Best regards, Yannick
I think I found where the culprit is in the log in comment #8: I was wrong thinking the issue is at the very beginningof the call, it is at the very end of the first call: When the device is closed it is switched from Default to SILENT (a very short time before the BYE): 2009/08/27 20:44:18.930 0:39.769 AudioOutputCore Closing current device 2009/08/27 20:44:18.930 0:39.769 GMAudioOutputManager_ptlib Closing device[0] Default (PTLIB/ALSA) 2009/08/27 20:44:18.930 0:39.769 ALSA Closing default 2009/08/27 20:44:18.954 0:39.793 Media Patch:0xa32bcb90 OpalPlugin Encoder sending consecutive I-Frame at 342 2009/08/27 20:44:18.954 0:39.793 Media Patch:0xa32bcb90 OpalPlugin Encoder has sent too many consecutive I-Frames - assume codec cannot do P-Frames 2009/08/27 20:44:18.954 0:39.793 Media Patch:0xa32bcb90 Patch Pushing 1 packet into RC 2009/08/27 20:44:18.954 0:39.793 Media Patch:0xa32bcb90 Patch Pulled 1 frames from RC 2009/08/27 20:44:18.955 0:39.794 GMAudioOutputManager_ptlib Setting Device[0] Default (PTLIB/ALSA) 2009/08/27 20:44:18.955 0:39.795 Media Closing stream OpalAudioMediaStream-Sink-PCM-16 2009/08/27 20:44:18.957 0:39.796 Media Patch:0xa33c0b90 GMAudioOutputManager_ptlib Trying to get frame from closed device[0] 2009/08/27 20:44:18.957 0:39.796 Media Patch:0xa33c0b90 AudioOutputCore Closing current device 2009/08/27 20:44:18.957 0:39.796 Media Patch:0xa33c0b90 GMAudioOutputManager_ptlib Closing device[0] Default (PTLIB/ALSA) 2009/08/27 20:44:18.957 0:39.796 Media Patch:0xa33c0b90 AudioOutputCore Falling back to SILENT (Ekiga/Ekiga) 2009/08/27 20:44:18.957 0:39.796 Media Patch:0xa33c0b90 GMAudioOutputManager_null Setting Device[0] SILENT (Ekiga/Ekiga) 2009/08/27 20:44:18.957 0:39.796 Media Patch:0xa33c0b90 AudioOutputCore Opening device[0] with 1-8000/16 2009/08/27 20:44:18.957 0:39.797 Media Patch:0xa33c0b90 GMAudioOutputManager_null Opening Device[0] SILENT (Ekiga/Ekiga) 2009/08/27 20:44:18.957 0:39.797 Media Patch:0xa33c0b90 GMAudioOutputManager_null Opening Device with 1-8000/16 Then the second Call start with the SILENT device, thus the no sound: 2009/08/27 20:44:27.088 0:47.927 Pool:0xb4d83b90 AudioOutputCore Opening device[0] with 1-8000/16 2009/08/27 20:44:27.088 0:47.927 Pool:0xb4d83b90 GMAudioOutputManager_null Opening Device[0] SILENT (Ekiga/Ekiga) And that explain why if i switch during the call from "default" in prefs (which wrongly reflect the actual state, as it is clearly SILENT in the log) to silent and back to default i get the sound back.
Well, the relevant part is a bit longer : 2009/08/27 20:44:18.930 0:39.769 AudioOutputCore Closing current device 2009/08/27 20:44:18.930 0:39.769 GMAudioOutputManager_ptlib Closing device[0] Default (PTLIB/ALSA) 2009/08/27 20:44:18.930 0:39.769 ALSA Closing default 2009/08/27 20:44:18.954 0:39.793 Media Patch:0xa32bcb90 OpalPlugin Encoder sending consecutive I-Frame at 342 2009/08/27 20:44:18.954 0:39.793 Media Patch:0xa32bcb90 OpalPlugin Encoder has sent too many consecutive I-Frames - assume codec cannot do P-Frames 2009/08/27 20:44:18.954 0:39.793 Media Patch:0xa32bcb90 Patch Pushing 1 packet into RC 2009/08/27 20:44:18.954 0:39.793 Media Patch:0xa32bcb90 Patch Pulled 1 frames from RC 2009/08/27 20:44:18.955 0:39.794 GMAudioOutputManager_ptlib Setting Device[0] Default (PTLIB/ALSA) 2009/08/27 20:44:18.955 0:39.795 Media Closing stream OpalAudioMediaStream-Sink-PCM-16 2009/08/27 20:44:18.957 0:39.796 Media Patch:0xa33c0b90 GMAudioOutputManager_ptlib Trying to get frame from closed device[0] 2009/08/27 20:44:18.957 0:39.796 Media Patch:0xa33c0b90 AudioOutputCore Closing current device 2009/08/27 20:44:18.957 0:39.796 Media Patch:0xa33c0b90 GMAudioOutputManager_ptlib Closing device[0] Default (PTLIB/ALSA) 2009/08/27 20:44:18.957 0:39.796 Media Patch:0xa33c0b90 AudioOutputCore Falling back to SILENT (Ekiga/Ekiga) 2009/08/27 20:44:18.957 0:39.796 Media Patch:0xa33c0b90 GMAudioOutputManager_null Setting Device[0] SILENT (Ekiga/Ekiga) 2009/08/27 20:44:18.957 0:39.796 Media Patch:0xa33c0b90 AudioOutputCore Opening device[0] with 1-8000/16 2009/08/27 20:44:18.957 0:39.797 Media Patch:0xa33c0b90 GMAudioOutputManager_null Opening Device[0] SILENT (Ekiga/Ekiga) 2009/08/27 20:44:18.957 0:39.797 Media Patch:0xa33c0b90 GMAudioOutputManager_null Opening Device with 1-8000/16 2009/08/27 20:44:18.958 0:39.797 Media Patch:0xa33c0b90 Media Tried to write to closed media stream 2009/08/27 20:44:18.958 0:39.797 Media Patch:0xa33c0b90 Media WritePacket failed with written 192160 2009/08/27 20:44:18.958 0:39.797 Media Patch:0xa33c0b90 Patch WriteFrame failed 2009/08/27 20:44:18.958 0:39.797 Media Patch:0xa33c0b90 Patch Thread ended because all sink writes failed failed 2009/08/27 20:44:18.958 0:39.798 Media Patch:0xa33c0b90 Patch Thread ended for Patch OpalRTPMediaStream-Source-G.711-ALaw-64k -> OpalAudioMediaStream-Sink-PCM-16 2009/08/27 20:44:18.955 0:39.795 Media Disconnecting OpalAudioMediaStream-Sink-PCM-16 from patch thread Patch OpalRTPMediaStream-Source-G.711-ALaw-64k -> OpalAudioMediaStream-Sink-PCM-16 2009/08/27 20:44:18.968 0:39.808 Patch Removing media stream sink OpalAudioMediaStream-Sink-PCM-16 2009/08/27 20:44:18.968 0:39.807 OnRelease:0xa3199b90 SIP OnReleased: Call[f98edd3b21]-EP<sip>[c661893a-a791-de11-9523-002185488b80], phase = ReleasingPhase 2009/08/27 20:44:18.969 0:39.808 OnRelease:0xa3199b90 OpalCon SetPhase from ReleasingPhase to ReleasingPhase for Call[f98edd3b21]-EP<sip>[c661893a-a791-de11-9523-002185488b80] 2009/08/27 20:44:18.969 0:39.808 OpalCon Removed media stream OpalAudioMediaStream-Sink-PCM-16 2009/08/27 20:44:18.969 0:39.809 Call GetOtherPartyConnection Call[f98edd3b21]-EP<pc>[p9efd911f2] 2009/08/27 20:44:18.972 0:39.811 OnRelease:0xa3199b90 PWLib File handle low water mark set: 56 PUDPSocket 2009/08/27 20:44:18.976 0:39.815 AudioInputCore Stopping Stream 2009/08/27 20:44:18.976 0:39.815 AudioInputCore Closing current device 2009/08/27 20:44:18.976 0:39.815 GMAudioInputManager_ptlib Closing device Default (PTLIB/ALSA) 2009/08/27 20:44:18.976 0:39.816 ALSA Closing default 2009/08/27 20:44:18.989 0:39.828 GMAudioInputManager_ptlib Setting Device Default (PTLIB/ALSA) 2009/08/27 20:44:18.989 0:39.828 Media Closing stream OpalAudioMediaStream-Source-PCM-16 2009/08/27 20:44:18.989 0:39.828 Media Disconnecting OpalAudioMediaStream-Source-PCM-16 from patch thread Patch OpalAudioMediaStream-Source-PCM-16 -> OpalRTPMediaStream-Sink-G.711-ALaw-64k 2009/08/27 20:44:18.989 0:39.828 Patch Closing media patch Patch OpalAudioMediaStream-Source-PCM-16 -> OpalRTPMediaStream-Sink-G.711-ALaw-64k 2009/08/27 20:44:18.989 0:39.828 Media Closing RTP for OpalRTPMediaStream-Sink-G.711-ALaw-64k 2009/08/27 20:44:18.989 0:39.828 RTP_UDP Session 1, Shutting down write. 2009/08/27 20:44:18.989 0:39.829 Media Closing stream OpalRTPMediaStream-Sink-G.711-ALaw-64k 2009/08/27 20:44:18.989 0:39.829 Media Disconnecting OpalRTPMediaStream-Sink-G.711-ALaw-64k from patch thread Patch OpalAudioMediaStream-Source-PCM-16 -> OpalRTPMediaStream-Sink-G.711-ALaw-64k 2009/08/27 20:44:18.990 0:39.829 Patch Removing media stream sink OpalRTPMediaStream-Sink-G.711-ALaw-64k 2009/08/27 20:44:18.990 0:39.829 OpalCon Removed media stream OpalRTPMediaStream-Sink-G.711-ALaw-64k 2009/08/27 20:44:18.990 0:39.830 Patch Waiting for media patch thread to stop Patch OpalAudioMediaStream-Source-PCM-16 2009/08/27 20:44:18.994 0:39.833 Media Patch:0xa337fb90 GMAudioInputManager_ptlib Trying to get frame from closed device 2009/08/27 20:44:18.994 0:39.833 Media Patch:0xa337fb90 AudioInputCore Closing current device 2009/08/27 20:44:18.994 0:39.833 Media Patch:0xa337fb90 GMAudioInputManager_ptlib Closing device Default (PTLIB/ALSA) 2009/08/27 20:44:18.994 0:39.834 Media Patch:0xa337fb90 AudioInputCore Falling back to SILENT (Ekiga/Ekiga) 2009/08/27 20:44:18.994 0:39.834 Media Patch:0xa337fb90 GMAudioInputManager_null Setting Device SILENT (Ekiga/Ekiga) 2009/08/27 20:44:18.995 0:39.834 Media Patch:0xa337fb90 AudioInputCore Opening device with 1-8000/16 2009/08/27 20:44:18.995 0:39.834 Media Patch:0xa337fb90 GMAudioInputManager_null Opening Device SILENT (Ekiga/Ekiga) 2009/08/27 20:44:18.995 0:39.834 Media Patch:0xa337fb90 GMAudioInputManager_null Opening Device with 1-8000/16 2009/08/27 20:44:18.995 0:39.834 Media Patch:0xa337fb90 Patch Thread ended because all sink writes failed failed 2009/08/27 20:44:18.995 0:39.834 Media Patch:0xa337fb90 Patch Thread ended for Patch OpalAudioMediaStream-Source-PCM-16 2009/08/27 20:44:19.009 0:39.848 PTLib Destroyed thread 0xb528d238 Media Patch:0xa337fb90(id = a337fb90) 2009/08/27 20:44:19.009 0:39.849 OpalCon Removed media stream OpalAudioMediaStream-Source-PCM-16
Michael Rickman posted a patch for a similar problem that Eugen applied. Is that patch introducing a new bug ? Eugen ? Or is the bug only half fixed ? I guess it switches back to Silence to play the sound event ? If not, perhaps Michael can have a look. Eugen, can you ask him ?
(In reply to comment #12) > Michael Rickman posted a patch for a similar problem that Eugen applied. Is > that patch introducing a new bug ? Eugen ? Maybe the desired_device is not the good one? http://git.gnome.org/cgit/ekiga/commit/?h=gnome-2-26&id=39190d030c7784efdff243d98024812cb9bac3ec > Or is the bug only half fixed ? This might be too. > I guess it switches back to Silence to play the sound event ? > > If not, perhaps Michael can have a look. Eugen, can you ask him ? Yes.
> I guess it switches back to Silence to play the sound event ? In ekiga prefs, everything was set to default, including sound events.
Here is the full sequence, for the beginning of the first call, regarding the audio device: 2009/08/27 20:43:50.695 0:11.534 Pool:0xb4d83b90 AudioOutputCore Opening device[0] with 1-8000/16 2009/08/27 20:43:50.695 0:11.534 Pool:0xb4d83b90 GMAudioOutputManager_ptlib Opening Device Default (PTLIB/ALSA) 2009/08/27 20:43:50.697 0:11.536 Pool:0xb4d83b90 GMAudioOutputManager_ptlib Opening Device with 1-8000/16 [Thread 0xb5489b90 (LWP 22405) exited] [New Thread 0xb5489b90 (LWP 22408)] [Thread 0xb5448b90 (LWP 22406) exited] [New Thread 0xb5448b90 (LWP 22409)] [New Thread 0xb5407b90 (LWP 22410)] [Thread 0xb5407b90 (LWP 22410) exited] [New Thread 0xb5407b90 (LWP 22411)] [New Thread 0xb4dc4b90 (LWP 22412)] [New Thread 0xb4d83b90 (LWP 22413)] [Thread 0xb4d83b90 (LWP 22413) exited] [New Thread 0xb4d83b90 (LWP 22414)] 2009/08/27 20:43:50.850 0:11.690 Pool:0xb4d83b90 ALSA SetHardwareParams Player channels=1 sample rate=8000 2009/08/27 20:43:50.851 0:11.690 Pool:0xb4d83b90 ALSA Successfully set period size to 160 2009/08/27 20:43:50.851 0:11.690 Pool:0xb4d83b90 ALSA Successfully set periods to 3 2009/08/27 20:43:50.893 0:11.732 Pool:0xb4d83b90 ALSA parameters set ok 2009/08/27 20:43:50.893 0:11.732 Pool:0xb4d83b90 ALSA Device default Opened [New Thread 0xb0c94b90 (LWP 22415)] 2009/08/27 20:43:51.014 0:11.854 Pool:0xb4d83b90 ALSA Got volume 37683 2009/08/27 20:43:51.022 0:11.861 Pool:0xb4d83b90 PCSS Opened mono sound channel "EKIGA" for playing at 8.0 kHz. 2009/08/27 20:43:51.023 0:11.862 Pool:0xb4d83b90 OpalMan OnOpenMediaStream Call[f98edd3b21]-EP<pc>[p9efd911f2],OpalAudioMediaStream-Sink-PCM-16 2009/08/27 20:43:51.023 0:11.862 Pool:0xb4d83b90 OpalCon Opened sink stream f98edd3b21_1 with format PCM-16 2009/08/27 20:43:51.023 0:11.862 Pool:0xb4d83b90 Patch Created Sink: format=PCM-16 2009/08/27 20:43:51.023 0:11.863 Pool:0xb4d83b90 Codec G711-ALaw-64k decoder created 2009/08/27 20:43:51.024 0:11.863 Pool:0xb4d83b90 Patch Created primary codec G.711-ALaw-64k->PCM-16 with ID f98edd3b21_1 2009/08/27 20:43:51.024 0:11.863 Pool:0xb4d83b90 Media Audio sink data size set to 3840, buffer size set to 160 and 24 buffers. 2009/08/27 20:43:51.024 0:11.863 Pool:0xb4d83b90 Media Set data size from 16 to 3840 2009/08/27 20:43:51.024 0:11.863 Pool:0xb4d83b90 GMAudioOutputManager_ptlib Setting buffer size of device[0] 160/24 2009/08/27 20:43:51.024 0:11.863 Pool:0xb4d83b90 ALSA SetBuffers direction=Player size=160 count=24 2009/08/27 20:43:51.024 0:11.864 Pool:0xb4d83b90 ALSA SetHardwareParams Player channels=1 sample rate=8000 2009/08/27 20:43:51.025 0:11.864 Pool:0xb4d83b90 ALSA Successfully set period size to 80 2009/08/27 20:43:51.025 0:11.864 Pool:0xb4d83b90 ALSA Successfully set periods to 24 2009/08/27 20:43:51.028 0:11.867 Pool:0xb4d83b90 ALSA parameters set ok 2009/08/27 20:43:51.028 0:11.867 Pool:0xb4d83b90 Patch Added media stream sink OpalAudioMediaStream-Sink-PCM-16 using transcoder G.711-ALaw-64k->PCM-16, data size=3840 2009/08/27 20:43:51.028 0:11.867 Pool:0xb4d83b90 Media RTP data size cannot be changed to 8, fixed at 2048 2009/08/27 20:43:51.028 0:11.867 Pool:0xb4d83b90 PCSS Adding filters to patch 2009/08/27 20:43:51.028 0:11.868 Pool:0xb4d83b90 OpalCon Source stream of connection Call[f98edd3b21]-EP<sip>[c661893a-a791-de11-9523-002185488b80] uses patch Patch OpalRTPMediaStream-Source-G.711-ALaw-64k -> OpalAudioMediaStream-Sink-PCM-16 2009/08/27 20:43:51.028 0:11.868 Pool:0xb4d83b90 RTPCon Adding RFC2833 receive handler 2009/08/27 20:43:51.029 0:11.868 Pool:0xb4d83b90 RTPCon Adding Cisco NSE receive handler 2009/08/27 20:43:51.029 0:11.868 Pool:0xb4d83b90 Call GetOtherPartyConnection Call[f98edd3b21]-EP<sip>[c661893a-a791-de11-9523-002185488b80] 2009/08/27 20:43:51.029 0:11.868 Pool:0xb4d83b90 Call OpenSourceMediaStreams opening audio session 1 on Call[f98edd3b21]-EP<pc>[p9efd911f2] 2009/08/27 20:43:51.033 0:11.873 Pool:0xb4d83b90 MediaFormat Removing codecs PCM-16-48kHz,PCM-16-32kHz,PCM-16-16kHz,PCM-16,G.726-16k,G.726-24k,G.726-32k,G.726-40k,GSM-06.10,GSM-AMR,LPC-10,MS-GSM,MS-IMA-ADPCM,SpeexIETFNarrow-11k,SpeexIETFNarrow-15k,SpeexIETFNarrow-18.2k,SpeexIETFNarrow-24.6k,SpeexIETFNarrow-5.95k,SpeexIETFNarrow-8k,SpeexNB,SpeexWNarrow-8k,YUV420P,RFC4175_YCbCr-4:2:0,RGB32,RGB24,RFC4175_RGB,SIP-IM,T.140,H.224/H323AnnexQ,H.224/HDLCTunneling,Linear-16-Stereo-48kHz 2009/08/27 20:43:51.035 0:11.875 Pool:0xb4d83b90 MediaFormat Removing codecs PCM-16-48kHz,PCM-16-32kHz,PCM-16-16kHz,PCM-16,G.726-16k,G.726-24k,G.726-32k,G.726-40k,GSM-06.10,GSM-AMR,LPC-10,MS-GSM,MS-IMA-ADPCM,SpeexIETFNarrow-11k,SpeexIETFNarrow-15k,SpeexIETFNarrow-18.2k,SpeexIETFNarrow-24.6k,SpeexIETFNarrow-5.95k,SpeexIETFNarrow-8k,SpeexNB,SpeexWNarrow-8k,YUV420P,RFC4175_YCbCr-4:2:0,RGB32,RGB24,RFC4175_RGB,SIP-IM,T.140,H.224/H323AnnexQ,H.224/HDLCTunneling,Linear-16-Stereo-48kHz 2009/08/27 20:43:51.037 0:11.876 Pool:0xb4d83b90 Call GetMediaFormats for Call[f98edd3b21]-EP<pc>[p9efd911f2] G.711-uLaw-64k G.711-ALaw-64k T.38 UserInput/RFC2833 2009/08/27 20:43:51.037 0:11.877 Pool:0xb4d83b90 MediaFormat Merging G.711-ALaw-64k into G.711-ALaw-64k 2009/08/27 20:43:51.038 0:11.877 Pool:0xb4d83b90 MediaFormat Merging G.711-ALaw-64k into PCM-16 2009/08/27 20:43:51.038 0:11.877 Pool:0xb4d83b90 MediaFormat Changed media option "Max Bit Rate" from 128000 to 64000 2009/08/27 20:43:51.038 0:11.877 Pool:0xb4d83b90 MediaFormat Merging PCM-16 into G.711-ALaw-64k 2009/08/27 20:43:51.039 0:11.878 Pool:0xb4d83b90 Call Selected media formats PCM-16 -> G.711-ALaw-64k 2009/08/27 20:43:51.040 0:11.879 Pool:0xb4d83b90 AudioInputCore Starting stream 1x8000/16 2009/08/27 20:43:51.040 0:11.879 Pool:0xb4d83b90 GMAudioInputManager_ptlib Setting Device Default (PTLIB/ALSA) 2009/08/27 20:43:51.040 0:11.879 Pool:0xb4d83b90 AudioInputCore Opening device with 1-8000/16 2009/08/27 20:43:51.040 0:11.879 Pool:0xb4d83b90 GMAudioInputManager_ptlib Opening Device Default (PTLIB/ALSA) 2009/08/27 20:43:51.040 0:11.880 Pool:0xb4d83b90 GMAudioInputManager_ptlib Opening Device with 1-8000/16 2009/08/27 20:43:51.076 0:11.915 Pool:0xb4d83b90 ALSA SetHardwareParams Recorder channels=1 sample rate=8000 2009/08/27 20:43:51.077 0:11.916 Pool:0xb4d83b90 ALSA Successfully set period size to 160 2009/08/27 20:43:51.077 0:11.916 Pool:0xb4d83b90 ALSA Successfully set periods to 3 2009/08/27 20:43:51.116 0:11.956 Pool:0xb4d83b90 ALSA parameters set ok 2009/08/27 20:43:51.117 0:11.956 Pool:0xb4d83b90 ALSA Device default Opened 2009/08/27 20:43:51.182 0:12.022 Pool:0xb4d83b90 ALSA Got volume 50462 2009/08/27 20:43:51.190 0:12.030 Pool:0xb4d83b90 PCSS Opened mono sound channel "EKIGA" for recording at 8.0 kHz. 2009/08/27 20:43:51.191 0:12.030 Pool:0xb4d83b90 OpalMan OnOpenMediaStream Call[f98edd3b21]-EP<pc>[p9efd911f2],OpalAudioMediaStream-Source-PCM-16 2009/08/27 20:43:51.192 0:12.031 Pool:0xb4d83b90 OpalCon Opened source stream f98edd3b21_1 with format PCM-16 2009/08/27 20:43:51.192 0:12.031 Pool:0xb4d83b90 SIPConnection::OpenMediaStream called with G.711-ALaw-64k, 1, 0 2009/08/27 20:43:51.192 0:12.031 Pool:0xb4d83b90 Call IsMediaBypassPossible Call[f98edd3b21]-EP<sip>[c661893a-a791-de11-9523-002185488b80] session 1 2009/08/27 20:43:51.192 0:12.031 Pool:0xb4d83b90 OpalMan IsMediaBypassPossible: session 1 2009/08/27 20:43:51.192 0:12.031 Pool:0xb4d83b90 OpalCon IsMediaBypassPossible: default returns false 2009/08/27 20:43:51.192 0:12.031 Pool:0xb4d83b90 RTP Found existing media session 1 2009/08/27 20:43:51.193 0:12.032 Pool:0xb4d83b90 OpalMan OnOpenMediaStream Call[f98edd3b21]-EP<sip>[c661893a-a791-de11-9523-002185488b80],OpalRTPMediaStream-Sink-G.711-ALaw-64k 2009/08/27 20:43:51.193 0:12.032 Pool:0xb4d83b90 OpalCon Opened sink stream f98edd3b21_1 with format G.711-ALaw-64k 2009/08/27 20:43:51.193 0:12.033 Pool:0xb4d83b90 Patch Created Sink: format=G.711-ALaw-64k 2009/08/27 20:43:51.194 0:12.033 Pool:0xb4d83b90 Codec G711-ALaw-64k encoder created 2009/08/27 20:43:51.194 0:12.033 Pool:0xb4d83b90 Patch Created primary codec PCM-16->G.711-ALaw-64k with ID f98edd3b21_1 2009/08/27 20:43:51.194 0:12.033 Pool:0xb4d83b90 Media RTP data size cannot be changed to 1920, fixed at 528 2009/08/27 20:43:51.194 0:12.033 Pool:0xb4d83b90 Patch Added media stream sink OpalRTPMediaStream-Sink-G.711-ALaw-64k using transcoder PCM-16->G.711-ALaw-64k, data size=528 2009/08/27 20:43:51.194 0:12.033 Pool:0xb4d83b90 Media Audio source data size set to 320, buffer size set to 160 and 5 buffers. 2009/08/27 20:43:51.194 0:12.034 Pool:0xb4d83b90 Media Set data size from 16 to 320 2009/08/27 20:43:51.194 0:12.034 Pool:0xb4d83b90 AudioInputCore Setting stream buffer size 5/160 2009/08/27 20:43:51.195 0:12.034 Pool:0xb4d83b90 GMAudioInputManager_ptlib Setting buffer size to 160/5 2009/08/27 20:43:51.195 0:12.034 Pool:0xb4d83b90 ALSA SetBuffers direction=Recorder size=160 count=5 2009/08/27 20:43:51.195 0:12.034 Pool:0xb4d83b90 ALSA SetHardwareParams Recorder channels=1 sample rate=8000 2009/08/27 20:43:51.195 0:12.034 Pool:0xb4d83b90 ALSA Successfully set period size to 80 2009/08/27 20:43:51.195 0:12.034 Pool:0xb4d83b90 ALSA Successfully set periods to 5 2009/08/27 20:43:51.201 0:12.040 Pool:0xb4d83b90 ALSA parameters set ok 2009/08/27 20:43:51.201 0:12.040 Pool:0xb4d83b90 PCSS Adding filters to patch
Created attachment 142077 [details] make shure audio output device is the desired one when a new call is made I think there are two rather similar bugs and only one got fixed. Both audio input and output are affected by the situation that the respective worker thread may switch its audio device to SILENT when the main thread has closed the desired device before. Then they are left there for the next call. http://git.gnome.org/cgit/ekiga/commit/?h=gnome-2-26&id=39190d030c7784efdff243d98024812cb9bac3ec fixed audio input (input !!! the title of the commit is misleading). That fixed Win32 the greyed audio button for me. The logs of this bug report show that audio output is affected in a similar way. Attached patch should fix that. In http://mail.gnome.org/archives/ekiga-devel-list/2009-August/msg00021.html Dominik stress tested my Win32 previews. The ...-release.exe only contained the audio input fix, the ...-release.exe.isitbetter also this patch for audio output. Michael
Welcome back Michael! Thank you! Do you know if the ringing device is affected as well? (I will test it once the fix will be commited...) Best regards, Yannick
As far as I can see the ringing device is not affected. But I have not yet tested it extensively as the Win32 version of Ekiga has an additional bug in finding the ring tone file. I will present a fix for that on the devel-list. Michael
Eugen pushed the fix yesterday: 7deb8053eb0876b 2009-09-01 09:10:45 (GMT) Save the desired primary audio output device across calls . For Win32 it is ok now. Hopefully for Linux too? Michael
Created attachment 142360 [details] log of the build attempt Unfortunately when trying complet rebuild of whole toolchain (ptlib v2_6 branch, revision 23318, opal v3_6 branch revision 23318, ekiga branch gnome-2-26 commit 46f7c650a044f93159f347479a290d8b2e7c523e; is it correct?) I got FTBFS in ekiga. See attached log of the rpmbuild (there are no patches in this package).
It seems to me that the error in the /usr/include/bits/mathinline.h file (from libc6-dev), isn't it?
(In reply to comment #21) > It seems to me that the error in the /usr/include/bits/mathinline.h file (from > libc6-dev), isn't it? Yes, it is, and it was glibc error which is now fixed in Rawhide. So I was able to build this and this particular bug seems to be fixed. Even after repeated calls I have still working echo service (haven't tested further). Other bugs remains (I will try to file comments for them, e.g., bug 592012 mutated but I still cannot register there, I will make comment over there). As for this bug, I think it could be closed.
Thanks for the feedback.
Hi, I used today stable branch ekiga/ptlib/opal and I confirm it is fixed now here too using 2 different systems (i386 and amd64), doing ~10 attempts on each system. Best regards, Yannick