[lkml]   [1996]   [Oct]   [20]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
SubjectRe: problems with sound as a module
I've done more investigating and found out that the module is not the
problem. The problem is that this happens any time I am running X and X is
doing anything. However, there is no problem if I am running in text mode,
even under extreme load (i.e. a kernel compile, 2 crashme's, a large
gzip/gunzip, and a recursive grep over 200MB of stuff). I don't think it's
an overall DMA problem, as hard disk i/o goes over DMA (triton ide).
Here's the full hardware configuration. Almost everything except the
network board and the new CPU was in an old neptune-based motherboard that
I upgraded from a while back. However, the machine was headless for a
while (serial console, no kernel patches) because of a monitor dying so I
was not immediately able to test soud under X.
SuperMicro P5STE Motherboard (Triton II/430HX)
Pentium 133
32MB RAM (FPM, not EDO)
2 EIDE hard drives
ATI Graphics Ultra Pro (Mach32, 2MB VRAM)
3.5" floppy drive
colorado jumbo 250 floppy tape drive
3com 3c590 ethernet card (PCI)
Mitsumi FX001D (old proprietary 2x cd-rom) and controller
Sound Blaster Pro 3.2 (real creative labs)
NEC Multisync 4DS monitor

After X does something (i.e. doing a ls in an xterm), the sound output is
garbled for the next few seconds and then vplay (.wav files) or cat (.au's
straight to /dev/audio) dies with an i/o error and the "sound: DMA output
timed out. IRQ/DRQ error?" shows up in the logs. Defining NO_SB_IRQ_TEST
doesn't help the problem.

Simon Karpen,

On Sun, 20 Oct 1996, Hannu Savolainen wrote:

> On Sat, 19 Oct 1996, Simon Karpen wrote:
> > after several loads/unloads, the sound driver (as a module) starts to give
> > lots of the following: "sound: DMA output timed out. IRQ/DRQ error?"
> > however, there are no such problems with it built into the kernel.
> > card is old SB-Pro, kernel 2.0.23, ask if you need more info.
> Some SB (compatible) cards have broken interrupt test feature which
> causes this.
> You should insert a line containing #define NO_SB_IRQ_TEST to the
> beginning of linux/drivers/sound/sb_common.c.
> Best regards,
> Hannu
> -----------------------------
> Hannu Savolainen (,
> (my home page)
> (Open Sound System (OSS))
> (OSS Free/TASD/VoxWare)

 \ /
  Last update: 2005-03-22 13:38    [W:0.074 / U:0.276 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site