Re: FIXED: Latest cvsup to 7.0 causes mplayer to crash my system

From: John Baldwin <jhb_at_FreeBSD.org>
Date: Mon, 18 Jul 2005 16:14:04 -0400
On Saturday 16 July 2005 11:03 am, Nicolas Blais wrote:
> On July 15, 2005 08:21 pm, Nicolas Blais wrote:
> > Hi,
> >
> > Today I've buildworld on latest cvsup (from -CURRENT_at_july 8th) and now
> > whenever i try to play a movie with mplayer, my system crashes with :
> >
> > Fatal trap 12: page fault while in kernel mode
> > fault virtual address   = 0x1c
> > fault code              = supervisor write, page not present
> > instruction pointer     = 0x20:0xc06a0bc3
> > stack pointer           = 0x28:0xe502fc88
> > frame pointer           = 0x28:0xe502fcc8
> > code segment            = base 0x0, limit 0xfffff, type 0x1b
> >                         = DPL 0, pres 1, def32 1, gran 1
> > processor eflags        = interrupt enabled, resume, IOPL = 0
> > current process         = 28 (swi4: clock sio)
> > trap number             = 12
> > panic: page fault
> > Uptime: 7m49s
> > Dumping 1023 MB (2 chunks)
> >   chunk 0: 1MB (159 pages) ... ok
> >   chunk 1: 1023MB (261808 pages) 1007 991 (CTRL-C to abort)
> >
> > and a 1gb dump. That same system was fine on 6-CURRENT from July 8th so
> > something in between breaks it. The crash message is always the same. I
> > tried recompiling mplayer which did no good.
> >
> > Please help,
> > Nicolas.
>
> I found a fix for my problem !
>
> I had HZ=1160 in my kernel and WITH_RTC enabled in mplayer.
>
> For some reason, latest kernel code doesn't like the option and panics. I
> removed both HZ=1160 and WITH_RTC and now mplayer works like a charm.
>
> So either mplayer (or rtc-2004.02.24.1_4) doesn't like something added in
> the kernel between july 8th and today, or the kernel doesn't like the way
> mplayer (or rtc) syncs up with the clock.

Can you figure out which one causes the panic?  I think HZ=1160 should still 
work as an option, so try adding that one back first and see if your system 
is ok.  Your traceback from ddb showed that the mutex pointer passed to 
mtx_unlock_flags() was NULL, so it might be an easy thing to fix if you can 
find out what callout from rtc(4) was firing.

-- 
John Baldwin <jhb_at_FreeBSD.org>  <><  http://www.FreeBSD.org/~jhb/
"Power Users Use the Power to Serve"  =  http://www.FreeBSD.org
Received on Mon Jul 18 2005 - 18:19:48 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:38:39 UTC