https://bugzilla.novell.com/show_bug.cgi?id=747391 https://bugzilla.novell.com/show_bug.cgi?id=747391#c2 --- Comment #2 from Dave Plater 2012-02-22 19:21:42 UTC --- I've just got back on line and am getting up to speed again. I'll try to build ffado svn again, there were required packages missing when I tried last year but I think at least one of them that I submitted then is in 12.1. If there are no missing build requirements this will fix Bug 736115 as well. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=747391 https://bugzilla.novell.com/show_bug.cgi?id=747391#c4 --- Comment #4 from andreas krumnow 2012-02-22 21:48:56 UTC --- Alesis IO26 is similar to Alesis IO14 and had worked well with the ffado driver on opensuse 11.3, if the compiler-/scons- flag ENABLE_DICE=true is given. The only missing things are a native mixer-panel and MIDI support. But this is satisfied by Jack-Mixer and ALSA. This is the situation for both devices. I tried to compile from svn-trunk, but get an compile-error on building the dbus part. No matter whether that flag is given or not. But maybe something on my system (opensuse 12.1) is missing and you (Dave) have it on your build environment. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=747391 https://bugzilla.novell.com/show_bug.cgi?id=747391#c6 --- Comment #6 from Dave Plater 2012-02-23 14:37:04 UTC --- (In reply to comment #4) Coolo split the ffado-mixer into a separate linked package a while ago to minimise rebuilds, the mixer has it's own spec file, if this is relevant. I will add the "-DENABLE_DICE=true" flag and notify you when (if) it has built for 12.1. Meanwhile I've found a possible patch which I'll first try on the release and I will work on getting the latest svn to build: + Euan de Kock 8th February 2011 12 13 + See Ticket #310 14 + 15 + This is a fix to force the Echo Devices to work at 16 + sample rates above 48000 where the ADAT digital channels 17 + get multiplexed to handle higher frequencies. 18 + At 88200 and 96000 the ADAT channels are reduced to 4, 19 + and at 192000 they should be reduced to 2. (I don't have a 20 + device capable of 192K to verify this) 21 + 22 + With the Echo devices (5.5 firmware at least), the device does 23 + not reduce its channel count when it multiplexes the ADAT data, 24 + but the actual packet does contain a reduced set of channels. 25 + 26 + The channels are referred to as Dimensions in the firewire parlance. 27 + With an Audiofire Pre8, we typically receive 17 channels of data at 28 + 44100 and 48000, being 8 Analog, 8 Digital and 1 MIDI. When we switch 29 + to 88200 or 96000 the ADAT (Digital) channel count is reduced to four, 30 + giving a total channel count (dimension) of 13. However Echo still 31 + reports the DBS parameter as being 17. 32 + 33 + To get around this, I don't bother using the DBS value at all. We know 34 + that the FDF is a reliable value for AM824 packets (type 0 to 7), so 35 + we can just use the FDF field to directly infer the size of each channel 36 + or dimension - 8, 16 or 32 Quadwords respectively for 41/48K, 88/96K and 192K. 37 + 38 + We set nevents directly to this for all AM824 packets directly, and do 39 + the standard calculation for any non-standard types. 40 + 41 + This is actually slightly less work than doing the calculation directly 42 + and should be portable across all devices that send AM824 data. 43 -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=747391 https://bugzilla.novell.com/show_bug.cgi?id=747391#c8 --- Comment #8 from andreas krumnow 2012-02-23 20:09:59 UTC --- The version ffado-2.0.1-75.1 is only available as src.rpm and installs 32-Bit libs but nothing for 64-Bit. In the x86_64 directory, there is a ffado-2.0.1-74.2 version, but this one has no DICE support. So in both cases, my 64-Bit jackd fails on opening the firewire driver. I've tried it, after installing first the *-75.1 and then the *74.2 with yast, with the command: "jackd -d firewire -v3". first as user, second as root. Which tests in which order should I run to give you worthy informations? Which output should I add as an attachment? PS. Thank you for giving me your lifetime to solve my problem. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=747391 https://bugzilla.novell.com/show_bug.cgi?id=747391#c10 --- Comment #10 from Dave Plater 2012-02-24 15:25:14 UTC --- These are possibly relevant tickets from ffado tracker : http://subversion.ffado.org/ticket/310 http://subversion.ffado.org/ticket/301 http://subversion.ffado.org/ticket/245 http://subversion.ffado.org/ticket/196 http://subversion.ffado.org/ticket/152 These are interesting. http://www.ffado.org/?q=node/30#comment-11213 http://www.ffado.org/?q=node/30#comment-11199 -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=747391 https://bugzilla.novell.com/show_bug.cgi?id=747391#c12 --- Comment #12 from Dave Plater 2012-02-25 10:01:18 UTC --- We now have ffado-2.0.1.2040 building in home:plater for you to install, the basic ffado package should pull in the lib and in turn libconfig-c++. Please test when you have a chance and if all is well I'll push libconfig to factory via mmlibs. Not quite sure if I can make it available as an update or through tumbleweed. I got the original package from a packman source rpm and adapted it for openSUSE so I suppose packman can link to the mmlibs package so it will be available that way. Your comment please Tiwai? I'm making it's availability known in Bug 736115 as well -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=747391 https://bugzilla.novell.com/show_bug.cgi?id=747391#c14 --- Comment #14 from Dave Plater 2012-02-26 11:23:15 UTC --- jack-1.9.8 is now available. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=747391 https://bugzilla.novell.com/show_bug.cgi?id=747391#c16 --- Comment #16 from andreas krumnow 2012-02-26 19:49:15 UTC --- Created an attachment (id=478057) --> (http://bugzilla.novell.com/attachment.cgi?id=478057) messages thrown by "ffado-test-streaming -t 1" The attachment file was produced as user, but the results, running it as root, are the same. The last line is the german word for "Segmentation fault". Installed ffado-2.0.1.2040-76.1.x86_64.rpm and jack-32bit-1.9.7-138.2.x86_64.rpm. I've heared a "Test-Sound" (a sine tone at about 440hz) until I've pressed ctrl+C. This is a first result. "jackd -d firewire" aborts with the message runs only as root, but this seems to be a rights-problem, I will try to solve that tomorrow and give a report here. Thank you for this very fast work. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=747391 https://bugzilla.novell.com/show_bug.cgi?id=747391#c18 --- Comment #18 from andreas krumnow 2012-02-27 23:26:20 UTC --- Success. This Bugreport can be closed. jack accepts the ffado-driver as firewire, recording signals from - and playback to the Alesis IO|26. The url http://download.opensuse.org/repositories/home:/plater/openSUSE_12.1 is now in my repo-list with a high priority (low number). It provides me the very first update (ffado-2.0.1.2040-76.2.x86_64.rpm) since the last new version yesterday :) I tested the driver with this newest rpm version. I've added the following 3 lines to /etc/security/limits.conf @audio - rtprio 99 @audio - memlock unlimited @audio - nice -19 I've also added the normal user(s) to the group "audio". Then: "chgrp audio /dev/fw*" and : "chmod g+rw /dev/fw*" Then a reboot was necessary to inform the system about the changes. (Maybe there are more elegant ways to inform the kernel about new limits, but this was the fastest one for me) After these little adjustments, I could run qjackctl as normal user, without scary messages, the Alesis is shown in the in- and out- connections and does, what it should do. Again: Thank You Very Much for this excellent and fast service, Dave. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=747391 https://bugzilla.novell.com/show_bug.cgi?id=747391#c20 --- Comment #20 from andreas krumnow 2012-02-28 15:54:30 UTC --- Yes, I've also installed/upgraded to jack-1.9.8 from your obs directory. As you mentioned in comment #14. It was available as rpm one day later. The latest tests ran with the newest version and the given versions harmonize together. Recording and playback works, setting sampling parameters of the device works and it is known, that there is no ffado-native-mixer for my device. Although jackd throws some warnings, when omitted from CLI, all important functionality is given. I don't know either these are device specific. The warnings sounds like that: "01796732348: Warning (IsoHandlerManager.cpp)[1651] getPacket: reconstructed CTR counter discrepancy 01796732392: Warning (IsoHandlerManager.cpp)[1657] getPacket: ingredients: 0, 48000000, 13FF, 4602A95B, FFFFFFFF, 36, 35, 127, 3145732095 01796732415: Warning (IsoHandlerManager.cpp)[1658] getPacket: diffcy = -191" Where "ingredients" counts from (0, 48000000, 13FF, 4602A95B) to (7F, 4607F000, 803FF, 4605E428) in hex-numbers, and diffcy counts backwards from -191 to -64 in decimal. But in my opinion, this adresses the ffado developers and you can push it into the official repos, because it fits the normal audio users purposes. (just my 2 cent). You can remember my email for asking me to test further versions on my device in the future. I'm glad to help the project in this way. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=747391 https://bugzilla.novell.com/show_bug.cgi?id=747391#c22 --- Comment #22 from Bernhard Wiedemann 2012-03-13 16:00:22 CET --- This is an autogenerated message for OBS integration: This bug (747391) was mentioned in https://build.opensuse.org/request/show/109125 Factory / libconfig -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=747391 https://bugzilla.novell.com/show_bug.cgi?id=747391#c24 --- Comment #24 from Bernhard Wiedemann 2012-03-20 16:00:25 CET --- This is an autogenerated message for OBS integration: This bug (747391) was mentioned in https://build.opensuse.org/request/show/110169 Factory / ffado -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.