I also experience the sound problems. Although sound works, it has problems stopping. For example when I hit stop in x11amp, the sound still loops for a couple of seconds and then I get the following message on the console:
timeout flushing dbuf_out, chan 1 cnt 0xfff9e46c flags 0x000001c1 This message is being generated in snd_flush() in /sys/i386/isa/snd/dmabuf.c. Looking at the code, it loops 10 times, each time waiting for a tsleep() to succeed. There are only two places in dmabuf.c that wakeup() this tsleep, one is dsp_wr_dmadone(), the second is in dsp_rd_dmadone(). Both are called from dsp_wrintr(). I put in a printf() at the start of dsp_wrintr(). The interrupt gets called when I start playing music in x11amp. When I stop it, no interrupt is generated, so dsp_wr_dmadone() is never called, so it never wakes up the tsleep() in snd_flush(). I really don't know why the interrupt does not occur (or if it should occur at all), maybe someone more knowledgable can take over from here.... Please. Pretty please with sugar on top :) Also, has anybody problems with sio buffer overflows? My machine is a Pentium 200MMX, which should be able to handle 115.2K on a com port just fine (my old 486 did, when I was running 2.2.x, 3.0 and 4.0 before newbus). But now I'm receiving lots of sio overflows, it is as simple as typing "ATI4" to my Courier, some characters get lost on output. If I am allowed to take a wild guess, this problem could be linked to the sound problem. Maybe some interrupts are generally lost or not correctly handled. Blaz Zupan, b...@amis.net, http://www.herbie.amis.net Medinet d.o.o., Linhartova 21, 2000 Maribor, Slovenia To Unsubscribe: send mail to majord...@freebsd.org with "unsubscribe freebsd-current" in the body of the message
