Date: Thu, 18 May 2000 10:25:54 -0700 (PDT) From: Lars Eggert <larse@isi.edu> To: fwkg7679@mb.infoweb.ne.jp, tanimura@r.dl.itc.u-tokyo.ac.jp Cc: freebsd-multimedia@freebsd.org, rat-users@cs.ucl.ac.uk Subject: Re: Fwd: Re: FreeBSD-4.0 & CS461x & rat/vat Message-ID: <14628.10146.148642.392319@hbo.isi.edu> In-Reply-To: <200005161457.XAA04844@mf003.infoweb.ne.jp> References: <14616.18236.968899.283451@hbo.isi.edu> <14618.4634.680064.72159A@rina> <14620.39746.774400.21559@hbo.isi.edu> <200005161457.XAA04844@mf003.infoweb.ne.jp>
next in thread | previous in thread | raw e-mail | index | archive | help
Lars> And this is where it hangs. I doubt this new problem is newpcm Lars> related. It seems like rat cannot correctly set up the mcast Lars> bus between its helper processes - have there been Lars> multicast/route/network changes been committed to -STABLE Lars> recently that could explain this? kurosawa> I suspect recent "delayed checksum" changes in sys/netinet kurosawa> broke multicast code. I've found multicast loopback kurosawa> packets always dropped because of bad UDP checksum, but kurosawa> couldn't find out why the checksum was always bad. kurosawa> Here is an ad hoc workaround for this problem: kurosawa> # sysctl -w net.inet.udp.checksum=0 That works, and rat starts now. Thanks for the hint! (Btw, it's still broken in -stable as of today.) I still haven't tried Seigo's patch though, because it seems something else has changed in -stable in the meantime, now the AIOGCAP call fails: [larse@hbo: ~] rat-4.2.3 dum/7777 2158:mbus_config.c:656 using Addr:224.255.222.239 Port:47000 Scope:HOSTLOCAL for MBUS 2159:main_ui.c:106 rat-ui started argc=5 2159:mbus_config.c:656 using Addr:224.255.222.239 Port:47000 Scope:HOSTLOCAL for MBUS 2159:main_ui.c:120 Waiting for mbus.waiting(rat-token-7ac332d9) from controller... 2159:main_ui.c:122 ...got it 2159:main_ui.c:128 Waiting for mbus.go(rat-token-7ac332d9) from controller... 2160:auddev_luigi.c:530 dev (0) name (CS461x PCM Audio) 2160:auddev.c:1172 Added "CS461x PCM Audio" to audio device table 2160:auddev.c:1172 Added "No Audio Device" to audio device table 2160:main_engine.c:154 Initializing session 0 2160:channel.c:204 Created encoder: "None" 2160:audio.c:235 Change device requested. 2160:auddev.c:552 Opened device: No Audio Device 2160:auddev.c:577 Requested Input: 16-bit signed linear,8-kHz,Mono 2160:auddev.c:579 Actual Input: 8-bit a-law,8-kHz,Mono 2160:auddev.c:588 Requested Output: 16-bit signed linear,8-kHz,Mono 2160:auddev.c:590 Actual Output: 8-bit a-law,8-kHz,Mono 2160:transmit.c:134 Unit duration 160 channels 1 2160:mix.c:106 Mixer created. Aligned to 0 8000kHz 2160:audio.c:296 0x00000000 0x00000201 2160:mbus_config.c:656 using Addr:224.255.222.239 Port:47000 Scope:HOSTLOCAL for MBUS 2160:main_engine.c:174 Waiting for mbus.waiting(rat-token-209f064f) from controller... 2160:main_engine.c:176 ...got it 2160:main_engine.c:180 Waiting for mbus.go(rat-token-209f064f) from controller... 2160:rtp.c:551 Created database entry for ssrc 0x54ed7c11 2160:main_engine.c:182 ...got it 2160:audio.c:235 Change device requested. 2160:mix.c:121 Mixer destroyed. Head 0 8000kHz Tail 0 8000kHz 2160:auddev_luigi.c:62 Opening /dev/audio0 2160:auddev_luigi.c:69 Failed AIOGCAP ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ 2160:auddev_luigi.c:70 soundcaps[0].rate_min = 0 2160:auddev_luigi.c:71 soundcaps[0].rate_max = 0 2160:auddev_luigi.c:72 soundcaps[0].formats = 0x00000000 2160:auddev_luigi.c:73 soundcaps[0].bufsize = 0 2160:auddev_luigi.c:74 soundcaps[0].mixers = 0x00000000 2160:auddev_luigi.c:75 soundcaps[0].inputs = 0x00000000 2160:auddev_luigi.c:76 soundcaps[0].left = 0x0000 2160:auddev_luigi.c:77 soundcaps[0].right = 0x0000 Sorry driver does support full duplex for this soundcard When I last tried this on 4/27/00, it still worked: 742:auddev_luigi.c:62 Opening /dev/audio0 742:auddev_luigi.c:70 soundcaps[0].rate_min = 11025 742:auddev_luigi.c:71 soundcaps[0].rate_max = 48000 742:auddev_luigi.c:72 soundcaps[0].formats = 0x10000010 742:auddev_luigi.c:73 soundcaps[0].bufsize = 16384 742:auddev_luigi.c:74 soundcaps[0].mixers = 0x00000001 742:auddev_luigi.c:75 soundcaps[0].inputs = 0x004049f1 742:auddev_luigi.c:76 soundcaps[0].left = 0x0064 742:auddev_luigi.c:77 soundcaps[0].right = 0x0064 Sorry driver does support full duplex for this soundcard Which was the situation in which Seigo's patch will help. Unfortunately, his patch will only make formats be 0x90000010, so I doubt this will fix the problem I see now. Did the newpcm code change again between today and 4/27? Thanks for all the help, Lars ________________________________________________________________________ Lars Eggert <larse@isi.edu> Information Sciences Institute http://www.isi.edu/~larse/ University of Southern California To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-multimedia" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?14628.10146.148642.392319>