Re: possible reasons to cause Segmentation fault
On Oct 11, 2011, at 0:08, Camaleón <noelamac@gmail.com> wrote:
> On Mon, 10 Oct 2011 23:04:53 +0800, lina wrote:
>
> (care with that html...)
I don't get this part.
Sorry,
What does this mean?
>
>> On Mon, Oct 10, 2011 at 9:00 PM, Camaleón <noelamac@gmail.com> wrote:
>>
>>> On Sun, 09 Oct 2011 23:44:00 +0800, lina wrote:
>>>
>>>> I met "Segmentation fault" in two places in one day,
>>>
>>> What are those places?
>>>
>> On laptop, later I figured out due to the recent update of one package
>> which cause its crashed.
>> after downgrading, it's fixed.
>
> Good.
>
> Maybe you can report a regression bug :-?
It's seemed I sent an email to the two maintainers at the same time, mainly told them it would cause segment fault and some other errors.
But I deleted nearly all emails today. So can't follow.
>
>> Another segmental fault is on cluster: mpirun noticed that process rank
>> 17 with PID 12834 on node c10 exited on signal 11 (Segmentation fault).
>> (I guess this problem will keep it as it is. don't wanna trouble the
>> administrator, but if you have suggestions welcome to let me know)
>
> (...)
>
> Mmm, for this I can't tell, as I don't know what "mpirun" is or does.
>
> What you can try is to get a trace for the segfault, some binaries
> provide debugging tools (by means of "-debug" flags you can pass to
> daemon) or even provide the corresponding "-gdb" packages to get a
> insightful trace.
>
>>>> Which are the possible reasons for the segmentation fault.
>>>
>>> Wow, there can be many. You need to provide more data.'
>>
>> I choose some which I thought might be important data.
>
> But what command did you run?
I qsub some job. Those came from the log file. Error log file.
>
>> Signal: Segmentation fault (11)
>> Signal code: Address not mapped (1)
>> Failing at address: 0xd1
>> [ 0] /lib64/libpthread.so.0 [0x3ec8e0e7c0]
>> [ 1] /usr/local/mpi/intel/lib/openmpi/mca_pml_ob1.so [0x2b7c80a013e8]
>> :
>> :
>> [14] /lib64/libc.so.6(__libc_start_main+0xf4) [0x3ec861d994]
>> [15] mdrun_mpi_d(do_cg+0x1c1) [0x40b359]
>
> Ah, okay, I guess this comes from the mentioned daemon (mpirun), right?
Yes.
>
> I'm afraid that goes beyond my knowledge, sorry, let's see if someone can
> give you any hint :-)
Thanks. .^_^.
>
> Greetings,
>
> --
> Camaleón
>
>
> --
> To UNSUBSCRIBE, email to debian-user-REQUEST@lists.debian.org
> with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
> Archive: [🔎] pan.2011.10.10.16.08.59@gmail.com">http://lists.debian.org/[🔎] pan.2011.10.10.16.08.59@gmail.com
>
Reply to: