From natedac@kscable.com  Fri Mar 16 15:30:19 2001
Return-Path: <natedac@kscable.com>
Received: from piccolo.daconcepts.dyndns.org (wks-166-129-114.kscable.com [24.166.129.114])
	by hub.freebsd.org (Postfix) with ESMTP id 842A137B718
	for <FreeBSD-gnats-submit@freebsd.org>; Fri, 16 Mar 2001 15:30:18 -0800 (PST)
	(envelope-from natedac@kscable.com)
Received: from localhost (natedac@localhost)
	by piccolo.daconcepts.dyndns.org (8.11.3/8.11.3) with ESMTP id f2GNPPr01654
	for <FreeBSD-gnats-submit@freebsd.org>; Fri, 16 Mar 2001 17:25:25 -0600 (CST)
	(envelope-from natedac@kscable.com)
Message-Id: <Pine.BSF.4.33.0103161724370.1646-100000@piccolo.daconcepts.dyndns.org>
Date: Fri, 16 Mar 2001 17:25:25 -0600 (CST)
From: Nate Dannenberg <natedac@kscable.com>
To: <FreeBSD-gnats-submit@freebsd.org>
Subject: ESS Sound drivers may eventually crap out if a program crashes

>Number:         25859
>Category:       kern
>Synopsis:       Program crash will kill sound device if it's using it.
>Confidential:   no
>Severity:       non-critical
>Priority:       medium
>Responsible:    cg
>State:          closed
>Quarter:        
>Keywords:       
>Date-Required:  
>Class:          sw-bug
>Submitter-Id:   current-users
>Arrival-Date:   Fri Mar 16 15:40:00 PST 2001
>Closed-Date:    Sun Apr 8 15:19:57 PDT 2001
>Last-Modified:  Sun Apr 08 15:20:21 PDT 2001
>Originator:     Nate Dannenberg
>Release:        FreeBSD 4.3-BETA i386
>Organization:
n/a
>Environment:

System: FreeBSD piccolo.daconcepts.dyndns.org 4.3-BETA FreeBSD 4.3-BETA
#1: Fri Mar 16 16:32:20 CST 2001
natedac@piccolo.daconcepts.dyndns.org:/usr/src/sys/compile/PICCOLO i386

IBM Aptiva - 550 MHz Athlon, 96M RAM, ESS Solo-1 sound device (other stats
of this machine are likely irrelevant).

Also happens on Compaq Presario 4764 - P166 MMX, 24M Ram, ESS
1688 Audiodrive.

>Description:

If a program crashes while it's writing to the sound device, the kernel
fails to close the device, rendering it dead.  Instread of closing the
device, you are given the kernel message "pcm0: play interrupt timeout,
channel dead".

>How-To-Repeat:

Run any sound-capable program such as mpg123, and find some way to
forcibly crash it would allowing it to exit cleanly.  It should leave the
sound device hosed, and further attempts to play sounds should report that
the device is "busy".

>Fix:

If you use the kernel sound modules instead of compiling the driver into
the kernel, it seems to work better.  The kernel won't allow me to unload
the modules while the sound device is actively in use, but I haven't had
the opportunity to try this method to clean up after a sound application
crashes.

-- 
 /~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~___~~~~~\
|  natedac@kscable.com              //Z@|___  |
|  http://home.kscable.com/natedac |'(__ [_<  |
 \_C64/C128_-_What's_*YOUR*_hobby?__\___|____/

>Release-Note:
>Audit-Trail:
Responsible-Changed-From-To: freebsd-bugs->cg 
Responsible-Changed-By: kris 
Responsible-Changed-When: Fri Mar 23 21:47:20 PST 2001 
Responsible-Changed-Why:  
cameron is Mr sound 

http://www.freebsd.org/cgi/query-pr.cgi?pr=25859 

From: natedac@kscable.com
To: freebsd-gnats-submit@FreeBSD.org, natedac@kscable.com
Cc:  
Subject: Re: kern/25859: Program crash will kill sound device i
Date: Sun, 25 Mar 2001 17:41:37 -0600 (CST)

 Please close this report.  Like the XFree86 4.x problem, I somehow managed to create two separate reports.  The other version should 
 remain, it's the one I'll be paying attention to, and it's more up to date :-)
 
 -- 
  /~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~___~~~~~\
 |  natedac@kscable.com              //Z@|___  |
 |  http://home.kscable.com/natedac |'(__ [_<  |
  \_C64/C128_-_What's_*YOUR*_hobby?__\___|____/
State-Changed-From-To: open->closed 
State-Changed-By: greid 
State-Changed-When: Sun Apr 8 15:19:57 PDT 2001 
State-Changed-Why:  
Closed at submitter's request 

http://www.freebsd.org/cgi/query-pr.cgi?pr=25859 
>Unformatted:
