Re: [uml-devel] Segfault with no mm

[<lis...@os...>]

James Stevenson writes:
 > 
 > Hi
 > 
 > ok i will the next time it happens
 > if it happens again
 > 
 > On Thu, 3 May 2001, Jeff Dike wrote:
 > 
 > > mi...@st... said:
 > > > the uml machine had been idle since liast night 
 > > > and i just run ls and boom.
 > > > #0  panic (fmt=0x10127204 "Segfault with no mm") at panic.c:54
 > > > #1  0x100ad382 in segv (address=1515870810, ip=1515870810, is_write=0,
 > > > is_user=0) at trap_kern.c:36
 > > 
 > > 1515870810 == 0x5a5a5a5a, which looks like slab poisoning.  It's interesting 
 > > that that's the ip.
 > > 
 > > I would have a look at the stack and see if it looks "stack-like" :-)
 > > 
 > > And see if there are any 0x5a5a5a5a in the neighborhood of esp or ebp.
 > > 

That's what I see every time I try to shut it down. Have you tried my
config yet?

What compiler do you use? A few days ago, I tried everything I could
find - gcc-2.95.3-0.20000517 (what I usually use, gcc .296 on RHL 7
(dies during boot), kgcc on RHL 7 (same result I get on 6.x with gcc
2.95). I don't recall whether I got Mandrake up - I've relocated
everything from hda to hde and mandrake's kernel doesn't ken the
controller, but if it did), it died in the same way.