Date: Mon, 1 May 2000 08:45:07 -0700 (MST) From: John Reynolds~ <jreynold@sedona.ch.intel.com> To: Walter Brameld <brameld@twave.net> Cc: Sean-Paul Rees <sean@seanrees.com>, freebsd-stable@freebsd.org Subject: Re: Sound skipping Message-ID: <14605.42627.743819.273014@hip186.ch.intel.com> In-Reply-To: <00043018320800.30807@Bozo_3.BozoLand.domain> References: <20000430141201.A79175@seanrees.com> <00043018320800.30807@Bozo_3.BozoLand.domain>
next in thread | previous in thread | raw e-mail | index | archive | help
[ On Sunday, April 30, Walter Brameld wrote: ]
>
> I just installed it to give it a try, and I get skips whenever I do
> anything else while it is in use. I can't use the equalizer either,
> bogs the playback down to nil output. I'm running a P3 450 w/ 128M ram.
>
> It plays fine if I don't touch my mouse or keyboard.... guess I'll
> stick with gqmpeg.
>
You might try applying this patch to your kernel:
It makes xmms "responsive" on my system, but still doesn't fix the bug where
it chews up almost 100% of the CPU when it's playing back (haven't had the
time to dive into this one ... it didn't used to do that, so I don't know if
it's a new version of xmms that is doing Bad Things(tm) or if we've done a Bad
Thing(tm) to our pcm driver).
--- es1370.c.orig Mon Feb 28 21:15:17 2000
+++ es1370.c Mon Feb 28 21:15:03 2000
@@ -628,7 +628,69 @@
static int
es_select(dev_t i_dev, int rw, struct proc * p)
{
- return (ENOSYS);
+ int unit, c = 1;
+ snddev_info *d ;
+ u_long flags;
+
+ /* dev = minor(i_dev); */
+ /* d = get_snddev_info(dev, &unit); */
+ unit = UNIT(minor(i_dev));
+ d = &pcm_info[unit];
+
+ if (d == NULL ) /* should not happen! */
+ return (ENXIO) ;
+
+ {
+ /*
+ * if the user selected a block size, then we want to use the
+ * device as a block device, and select will return ready when
+ * we have a full block.
+ * In all other cases, select will return when 1 byte is ready.
+ */
+ int lim = 1;
+
+ int revents = 0 ;
+ if (rw & (POLLOUT | POLLWRNORM) ) {
+ if ( d->flags & SND_F_HAS_SIZE )
+ lim = d->play_blocksize ;
+ /* XXX fix the test here for half duplex devices */
+ if (1 /* write is compatible with current mode */) {
+ flags = spltty();
+ if (d->dbuf_out.dl) {
+ es_wr_dmaupdate(d);
+ }
+ c = d->dbuf_out.fl ;
+ if (c < lim) /* no space available */
+ selrecord(p, & (d->wsel));
+ else
+ revents |= rw & (POLLOUT | POLLWRNORM);
+ splx(flags);
+ }
+ }
+ if (rw & (POLLIN | POLLRDNORM)) {
+ if ( d->flags & SND_F_HAS_SIZE )
+ lim = d->rec_blocksize ;
+ /* XXX fix the test here */
+ if (1 /* read is compatible with current mode */) {
+ flags = spltty();
+ if ( d->dbuf_in.dl == 0 ) /* dma idle, restart it */
+ dma_rdintr(d);
+ else {
+ es_rd_dmaupdate(d);
+ }
+ c = d->dbuf_in.rl ;
+ if (c < lim) /* no data available */
+ selrecord(p, & (d->rsel));
+ else
+ revents |= rw & (POLLIN | POLLRDNORM);
+ splx(flags);
+ }
+ DEB(printf("sndselect on read: %d >= %d flags 0x%08x\n",
+ c, lim, d->flags));
+ return c < lim ? 0 : 1 ;
+ }
+ return revents;
+ }
}
@@ -746,6 +808,15 @@
if(es_debug > 0) printf("es_callback reason %d speed %d \t",reason ,d->p
lay_speed);
switch(reason & SND_CB_REASON_MASK) {
+ case SND_CB_DMAUPDATE:
+ if (reason & SND_CB_WR)
+ es_wr_dmaupdate(d);
+ else if (reason & SND_CB_RD)
+ es_rd_dmaupdate(d);
+ else return -1;
+
+ break;
+
case SND_CB_INIT:
/* if(es_debug > 0) printf("case SND_CB_INIT\n"); */
if (d->type == ES1371_PCI_ID){
--
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
| John Reynolds WCCG, CCE, Higher Levels of Abstraction |
| Intel Corporation MS: CH6-210 Phone: 480-554-9092 pgr: 602-868-6512 |
| jreynold@sedona.ch.intel.com http://www-aec.ch.intel.com/~jreynold/ |
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-stable" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?14605.42627.743819.273014>