question about oops and panic

question about oops and panic

[Parmenides]

Hi,

1. I think oops and panic are both some way to deal with errors occurs
in kernel space. Is there any relationship between them?

2. I make a NULL pointer reference deliberately in a kernel module and
get an oops like:

... ... ...

Aug 29 00:58:45 lfs kernel: Call Trace:
Aug 29 00:58:45 lfs kernel:  [<c100112d>] ? do_one_initcall+0x44/0x120
Aug 29 00:58:45 lfs kernel:  [<c10517ce>] ? sys_init_module+0xa7/0x1d9
Aug 29 00:58:45 lfs kernel:  [<c138d49d>] ? syscall_call+0x7/0xb

... ... ...

I wonder what is the meaning of the tow numbers after a function name.

question about oops and panic

[anish singh]

On Mon, Aug 29, 2011 at 2:19 AM, Parmenides <mobile.parmenides@gmail.com> wrote:
> Hi,
>
> 1. I think oops and panic are both some way to deal with errors occurs
> in kernel space. Is there any relationship between them?
AFAIK both are same.
>
> 2. I make a NULL pointer reference deliberately in a kernel module and
> get an oops like:
>
> ... ... ...
>
> Aug 29 00:58:45 lfs kernel: Call Trace:
> Aug 29 00:58:45 lfs kernel: ?[<c100112d>] ? do_one_initcall+0x44/0x120
> Aug 29 00:58:45 lfs kernel: ?[<c10517ce>] ? sys_init_module+0xa7/0x1d9
> Aug 29 00:58:45 lfs kernel: ?[<c138d49d>] ? syscall_call+0x7/0xb
>
> ... ... ...
>
> I wonder what is the meaning of the tow numbers after a function name.
They are offsets in the function name from which next function in the call
stack is called right ?
>
> _______________________________________________
> Kernelnewbies mailing list
> Kernelnewbies at kernelnewbies.org
> http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
>

question about oops and panic

[Mandeep Sandhu]

>> 1. I think oops and panic are both some way to deal with errors occurs
>> in kernel space. Is there any relationship between them?
> AFAIK both are same.
>>

Really? I thought "oops" would be generated for critical errors like
processor exceptions etc when a kernel CANNOT proceed further in a
reliable way...whereas a "panic" was _artificially_ induced (?), eg: a
missing 'init' to run (very common error, which I'm sure most of us
here would've experienced :)), and the kernel _may_ proceed further
with some reliability

Wikipedia, has a decent explanation of it:

http://en.wikipedia.org/wiki/Linux_kernel_oops

HTH,
-mandeep

question about oops and panic

[Stratos Psomadakis]

On 08/28/2011 08:19 PM, Parmenides wrote:
> Hi,
>
> 1. I think oops and panic are both some way to deal with errors occurs
> in kernel space. Is there any relationship between them?
afaik, the difference is that the panic is non-recoverable, and occurs
only in certain very critical situations
>
> 2. I make a NULL pointer reference deliberately in a kernel module and
> get an oops like:
>
> ... ... ...
>
> Aug 29 00:58:45 lfs kernel: Call Trace:
> Aug 29 00:58:45 lfs kernel:  [<c100112d>] ? do_one_initcall+0x44/0x120
> Aug 29 00:58:45 lfs kernel:  [<c10517ce>] ? sys_init_module+0xa7/0x1d9
> Aug 29 00:58:45 lfs kernel:  [<c138d49d>] ? syscall_call+0x7/0xb
>
> ... ... ...
>
> I wonder what is the meaning of the tow numbers after a function name.
the first one is the offset within the function that the error occurred,
and the second one I think it is the function size (you'll usually need
only the first one for debugging)

-- 
Stratos Psomadakis
<s.psomadakis@gmail.com>

question about oops and panic

[anish singh]

On Mon, Aug 29, 2011 at 8:01 PM, Mandeep Sandhu
<mandeepsandhu.chd@gmail.com> wrote:
>>> 1. I think oops and panic are both some way to deal with errors occurs
>>> in kernel space. Is there any relationship between them?
>> AFAIK both are same.
>>>
>
> Really? I thought "oops" would be generated for critical errors like
> processor exceptions etc when a kernel CANNOT proceed further in a
> reliable way...whereas a "panic" was _artificially_ induced (?), eg: a
> missing 'init' to run (very common error, which I'm sure most of us
> here would've experienced :)), and the kernel _may_ proceed further
> with some reliability
Thanks for this but what i meant was that some developers uses
oops and panic term interchangeably.
Technically you are right.
>
> Wikipedia, has a decent explanation of it:
>
> http://en.wikipedia.org/wiki/Linux_kernel_oops
>
> HTH,
> -mandeep
>

question about oops and panic

[sandeep kumar]

Hmm...there is a very subtle difference, ofcourse...

Oops you will get, when you do a NULL pointer dereference (an exception may
be software???)..
Suppose when you try to insert a dynamic module(insmod xyz.ko), which refers
to a NULL pointer,
while installing that module, you will get oops. And you cant remove that
module.(rmmod xyz.ko wont be successful)
Only way you can do rmmod is, to restart the kernel to set things right.

Panic is the severe condition, where kernel couldnot proceed its execution
furthur.
It can be hardware specific also.
for eg:
When you try to set a physical parameter of the hardware, which it could not
capable of..
suppose setting a frequency limit of on clock, which it doesnot support..and

some action done manipulating certain CPU specific control registers which
might cause abnormal
behaviour...etc..,


In these conditions the kernel has no other go but stop its execution..then
it will call the panic handler(panic("string"))..

Hope this makes a little more sense & You can grep for the panic() calls in
the kernel source code, which might give you better insight..

On Mon, Aug 29, 2011 at 4:21 PM, anish singh <anish198519851985@gmail.com>wrote:

> On Mon, Aug 29, 2011 at 2:19 AM, Parmenides <mobile.parmenides@gmail.com>
> wrote:
> > Hi,
> >
> > 1. I think oops and panic are both some way to deal with errors occurs
> > in kernel space. Is there any relationship between them?
> AFAIK both are same.
> >
> > 2. I make a NULL pointer reference deliberately in a kernel module and
> > get an oops like:
> >
> > ... ... ...
> >
> > Aug 29 00:58:45 lfs kernel: Call Trace:
> > Aug 29 00:58:45 lfs kernel:  [<c100112d>] ? do_one_initcall+0x44/0x120
> > Aug 29 00:58:45 lfs kernel:  [<c10517ce>] ? sys_init_module+0xa7/0x1d9
> > Aug 29 00:58:45 lfs kernel:  [<c138d49d>] ? syscall_call+0x7/0xb
> >
> > ... ... ...
> >
> > I wonder what is the meaning of the tow numbers after a function name.
> They are offsets in the function name from which next function in the call
> stack is called right ?
>  >
> > _______________________________________________
> > Kernelnewbies mailing list
> > Kernelnewbies at kernelnewbies.org
> > http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
> >
>
> _______________________________________________
> Kernelnewbies mailing list
> Kernelnewbies at kernelnewbies.org
> http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
>



-- 
With regards,
Sandeep Kumar Anantapalli,
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.kernelnewbies.org/pipermail/kernelnewbies/attachments/20110829/a50f27f2/attachment-0001.html 

question about oops and panic

[Christopher Harvey]

On Mon, 29 Aug 2011 01:19:00 +0800, Parmenides wrote:
> Hi,
>
[snip]
>
> Aug 29 00:58:45 lfs kernel: Call Trace:
> Aug 29 00:58:45 lfs kernel:  [<c100112d>] ? 
> do_one_initcall+0x44/0x120
> Aug 29 00:58:45 lfs kernel:  [<c10517ce>] ? 
> sys_init_module+0xa7/0x1d9
> Aug 29 00:58:45 lfs kernel:  [<c138d49d>] ? syscall_call+0x7/0xb
>
> ... ... ...
>
> I wonder what is the meaning of the tow numbers after a function 
> name.
>

You can use objdump -dSl foo.ko to get the exact line that failed. Look 
at the left hand column of the dump and match it with the first hex 
number in your trace. The objdump command may take a while if your 
module is large. Obviously you'll need debugging symbols built into the 
ko.