[uml-devel] further scheduling weirdness

[uml-devel] further scheduling weirdness

[sean whalen]

Hi All --
Continuing yesterday's thread where 2.6.0-test9-um was bombing on a
schedule_timeout() or interruptible_sleep_on_wait() call due to a null
pointer dereference in sched.c, I decided to try 2.4.22-um and see what
calling schedule_timeout() would do.  Again it crashes at sched.c:564, which
is a different place and has a different back trace.  This time the back
trace is:

(gdb)
451             schedule();
(gdb)
Kernel panic: kernel BUG at sched.c:564!
In interrupt handler - not syncing
 start_idle_thread (stack=0xa023e000, switch_buf_ptr=0xa023c578,
       fork_buf_ptr=0x2) at process.c:281
 281             if(n == 0)
(gdb) bt
#0  start_idle_thread (stack=0xa023e000, switch_buf_ptr=0xa023c578,
     fork_buf_ptr=0x2) at process.c:281
#1  0xa00fd15b in start_uml_skas () at process_kern.c:167
#2  0xa00f7421 in linux_main (argc=0, argv=0xa0800000) at um_arch.c:382
#3  0xa000d3a6 in main (argc=8, argv=0xbffff824, envp=0xbffff848)
at arch/um/main.c:143

This is simply due to a schedule_timeout call in a module.  Does anyone have
any ideas on this?  It seems to critical a function to actually be broken.

Thanks,
    -Sean

-
sean whalen . ucdavis security lab
swhalen at denature.org || node99.org

'if god had wanted us to evolve past the non sequitur,
he wouldn't have left all these fish in the tub.'



-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program.
Does SourceForge.net help you be more productive?  Does it
help you create better code?  SHARE THE LOVE, and help us help
YOU!  Click Here: http://sourceforge.net/donate/
_______________________________________________
User-mode-linux-devel mailing list
User-mode-linux-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-devel

Re: [uml-devel] further scheduling weirdness

[Jeff Dike]

swhalen@denature.org said:
> This is simply due to a schedule_timeout call in a module.  Does
> anyone have any ideas on this?  It seems to critical a function to
> actually be broken. 

That BUG is "scheduling in interrupt".  So if your module is doing anything
that could sleep, like that schedule(), in an interrupt handle, it should
stop.

				Jeff



-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program.
Does SourceForge.net help you be more productive?  Does it
help you create better code?  SHARE THE LOVE, and help us help
YOU!  Click Here: http://sourceforge.net/donate/
_______________________________________________
User-mode-linux-devel mailing list
User-mode-linux-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-devel