[uml-devel] Sleeping process getting segfault signal

[uml-devel] Sleeping process getting segfault signal

[M A Young]

I have a reproduceable crash, at least with certain host kernels (eg.
from RedHat's current beta), and a 2.6 uml kernel, and get errors like
sleeping process 1786 got unexpected signal : 11
immediately after I type halt in a uml window. Does anyone know what might
be going on here, or how I might debug it further?

	Michael Young


-------------------------------------------------------
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] Sleeping process getting segfault signal

[Oleg Drokin]

Hello!

M A Young <m.a.young@durham.ac.uk> wrote:
MAY> I have a reproduceable crash, at least with certain host kernels (eg.
MAY> from RedHat's current beta), and a 2.6 uml kernel, and get errors like
MAY> sleeping process 1786 got unexpected signal : 11
MAY> immediately after I type halt in a uml window. Does anyone know what might
MAY> be going on here, or how I might debug it further?

I see similar problems, but unrelated to 'halt'. They appear ocasionally.
And are very much like that problem you reported earlier (about copy from user).
It happens in the same place for me, Jeff's patch that converts *jmp() calls
to sig*jmp() is present in my tree.

Jeff: I finally hit the problem yesterday with extra debug code to save the
sigmask. Here's what I see. (also include sigjump buffer that was setup later
than sigprocmask was run).

sigmask = {__val = {134284288, 0, 0, 0, 134284288, 0, 23441, 
    0 <repeats 12 times>, 2776382020, 0, 2776381980, 2685736023, 2776382020, 
    0, 2776382020, 2776382172, 2685229112, 0, 0, 0, 2776382068}}
jbuf = {{__jmpbuf = {-1518585548, 800, -1607600160, -1518585268, -1518585740, 
      -1609727948}, __mask_was_saved = 1, __saved_mask = {__val = {134284288, 
        0, 2776382332, 2776382324, 1, 2776382372, 2776382372, 0, 14, 7, 
        2685735890, 115, 642, 2776382324, 123, 0, 134284288, 135492232, 1, 8, 
        2776367104, 2776381732, 2685239162, 29, 2776381744, 0, 0, 0, 0, 
        2776382476, 2685735368, 29}}}}

Bye,
    Oleg


-------------------------------------------------------
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] Sleeping process getting segfault signal

[Jeff Dike]

green@linuxhacker.ru said:
> Jeff: I finally hit the problem yesterday with extra debug code to
> save the sigmask. Here's what I see. (also include sigjump buffer that
> was setup later than sigprocmask was run).
>
> sigmask = {__val = {134284288

(gdb) p/x 134284288
$1 = 0x8010400

Thats SIGSEGV (the one that matters here), SIGCHLD, and SIGIO.  So, we have
to figure out how we're getting there with SIGSEGV blocked.  What's the stack
at that point?

				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

Re: [uml-devel] Sleeping process getting segfault signal

[Oleg Drokin]

Hello!

On Sun, Nov 02, 2003 at 09:36:42PM -0500, Jeff Dike wrote:
> > Jeff: I finally hit the problem yesterday with extra debug code to
> > save the sigmask. Here's what I see. (also include sigjump buffer that
> > was setup later than sigprocmask was run).
> > sigmask = {__val = {134284288
> (gdb) p/x 134284288
> $1 = 0x8010400
> Thats SIGSEGV (the one that matters here), SIGCHLD, and SIGIO.  So, we have
> to figure out how we're getting there with SIGSEGV blocked.  What's the stack
> at that point?

(gdb) bt
#0  0xa01658e3 in memcpy () at proc_fs.h:154
#1  0xa00d8075 in __do_user_copy (to=0x9fffef14, from=0xa02df7e0, n=800, 
    fault_addr=0xa57c0284, fault_catcher=0xa57c0288, 
    op=0xa00d8088 <__do_copy>, faulted_out=0xa57c3a80) at uaccess_user.c:29
#2  0xa00d80bb in __do_copy_to_user (to=0x9fffef14, from=0xa02df7e0, n=800, 
    fault_addr=0xa57c0284, fault_catcher=0xa57c0288) at uaccess_user.c:54
#3  0xa00cfd69 in setup_signal_stack_sc (stack_top=800, sig=17, 
    handler=2684350228, restorer=0x4004a978, regs=0xa57c0274, mask=0xa57c3b2c)
    at uaccess.h:61
#4  0xa00d4c87 in handle_signal (regs=0xa57c0274, signr=17, ka=0xa39586c4, 
    info=0xa57c3bf4, oldset=0x320, error=0) at signal_kern.c:114
#5  0xa00d4fdc in kern_do_signal (regs=0xa57c0274, oldset=0xa57c02f0, error=0)
    at signal_kern.c:231
#6  0xa00d5126 in do_signal (error=0) at signal_kern.c:262
#7  0xa00d32ce in interrupt_end () at process_kern.c:136
#8  0xa00db199 in sig_handler_common_tt (sig=11, sc_ptr=0x58) at trap_user.c:40
#9  0xa00d7f7a in sig_handler (sig=11, sc=

Bye,
    Oleg


-------------------------------------------------------
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] Sleeping process getting segfault signal

[Jeff Dike]

m.a.young@durham.ac.uk said:
> I have a reproduceable crash, at least with certain host kernels (eg.
> from RedHat's current beta), and a 2.6 uml kernel, and get errors like
> sleeping process 1786 got unexpected signal : 11 

I just saw this.  Looked at it a bit and am mystified :-)  Stay tuned...

				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

Re: [uml-devel] Sleeping process getting segfault signal

[Jeff Dike]

m.a.young@durham.ac.uk said:
> Does anyone know what might be going on here, or how I might debug it
> further?

Get a stack trace from the sleeping process.

				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