Re: [PATCH] stack2core: show stack message and convert it to core file when kernel die

[Hui Zhu <teawater@gmail.com>]

I didn't give the user raw oopses.
I give him core file. When the kernel die, do we can get a core file now?


(gdb) bt
#0  0xc0008470 in kernel_init (unused=<value optimized out>)
  at 
/home/teawater/kernel/arm_versatile_926ejs.glibc_std.standard/build/linux/init/main.c:916
#1  0xc0042660 in sys_waitid (which=<value optimized out>, upid=<value
optimized out>, infop=0x0, options=0, ru=0x14)
  at 
/home/teawater/kernel/arm_versatile_926ejs.glibc_std.standard/build/linux/kernel/exit.c:1798
Backtrace stopped: previous frame inner to this frame (corrupt stack?)

It show which line make kernel die.

Hui




On Mon, Jan 4, 2010 at 00:44, Russell King - ARM Linux
<linux@arm.linux.org.uk> wrote:
> On Mon, Jan 04, 2010 at 12:30:20AM +0800, Hui Zhu wrote:
>> This S2C: message just for program s2c.
>> s2c can convert it to a core file.  Then gdb can do a clear analyse
>> with this file.
>> Then you can get more message than current we can get.
>
> I understand that.  What I'm saying is that all the additional noise
> you're causing the kernel to create is just a pure duplication of
> what we already dump.
>
> Oops dumps are already noisy enough - especially if they cause a panic
> at the end (where you end up with two backtraces.)  We do not need even
> more noise caused by needless duplication.
>
> You can get everything you need already from the kernel.  On ARM, we
> already dump out all the registers and the _full_ stack.  There is no
> need for you to implement your own register dumping code and full stack
> dump on top of that again.
>
> So, I'm not going to accept your patch for the ARM kernel.  Please use
> what's already provided - it's more than adequate.  By doing so, you
> don't penalise those of us who want to read the raw oopses.
>
> Talking about noisy oopses, I'm getting one with 2.6.33-rc2 on 'poweroff'
> shutdown.  No idea what it is because most of it's scrolled off the top of
> the screen and I can't scroll back.  Not bothered about it at the moment.
> What it does illustrate though is why making things too noisy when problems
> occur makes it _more_ difficult to find out what went wrong.
>