From mboxrd@z Thu Jan 1 00:00:00 1970 From: Antti Boman Subject: Re: anybody else OK with current CVS *and* SMP? Date: Wed, 06 Aug 2003 15:28:57 +0300 Sender: alsa-devel-admin@lists.sourceforge.net Message-ID: <3F30F489.6010204@mindcom.fi> References: <200308051417.h75EHwpr020319@localhost.localdomain> <3F30156D.9060508@mindcom.fi> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: Errors-To: alsa-devel-admin@lists.sourceforge.net List-Help: List-Post: List-Subscribe: , List-Unsubscribe: , List-Archive: To: Takashi Iwai Cc: Paul Davis , alsa-devel@lists.sourceforge.net List-Id: alsa-devel@alsa-project.org Takashi Iwai wrote: > At Tue, 05 Aug 2003 23:37:01 +0300, > Antti Boman wrote: > >>This seems like the problem I've always had with my single processor >>with JACK and SCHED_FIFO, with 2.4.20 kernel. I haven't tried the >>realtime setting for a while, though. > > if the exactly same problem happens on UP, the cause would be > different from what i thought originally. my guess was spin dead > locks or other SMP-specific issues, possibly around the linked pcm > streams. It sounded as if it's the same problem, but naturally I don't know if it's caused by the same thing. I don't remember if it happened without X running, for example, so I have to test it without X, too. The frustrating part is that I really cannot debug anything, as the system completely locks up. >>Maybe there's some other thing causing these? Should we gather a bit >>more in-depth information on our computers? > > yes, this will help. > at least, we need to investigate with the same driver/lib version > (preferably with the same kernel and gcc). Ok. I'll gather the information, do a few more tests, and post it on the list. I can also upgrade/downgrade the libs and programs if you think it's worth the while. > btw, the dead-lock can be related to the lowlevel part, too. > recently i found a bug in es1968 driver which can cause a dead-lock > with linked streams and xrun (typically as jack does). but in this > case, it can happen regardless of SCHED_FIFO... If you remember, I used to have huge problems with CMI8738 based integrated sound... "card", but they went away when installing newer ALSA drivers and libraries. Except from this SCHED_FIFO, of course. Anyway, I'll first let you know about my current configuration, and you can tell me what to try. Thanks, -a ------------------------------------------------------- This SF.Net email sponsored by: Free pre-built ASP.NET sites including Data Reports, E-commerce, Portals, and Forums are available now. Download today and enter to win an XBOX or Visual Studio .NET. http://aspnet.click-url.com/go/psa00100003ave/direct;at.aspnet_072303_01/01