libstdc Question about error 7

libstdc Question about error 7

[CASTEL Alexandre -EXT]

[-- Attachment #1: Type: text/plain, Size: 992 bytes --]

Hello libstdc development group,

We are contacting you about one issue we discover in a laboratory.
We have made DMESG command on a linux computer, we saw that line:
TB_RP_Server[27922]: segfault at 47e01d78 ip 48f708dc sp b695a030 error 7 in libstdc++.so.6.0.13[48ee1000+e1000]

We know a segfault but we wanted to understand the end of this line "error 7 in libstdc++.so.6.0.13[48ee1000+e1000]"
What does error 7 or 4 or other numbers mean?
What does [48ee1000+e1000] represent?

Thank you in advance.
Best regards,

Alexandre CASTEL
TrainLab Engineer / CCAMY SYSTEMES - VULCAIN INGENIERIE
Tél : 06 77 90 59 86 / Email : alexandre.castel@ccamy.com



________________________________
CONFIDENTIALITY : This e-mail and any attachments are confidential and may be privileged. If you are not a named recipient, please notify the sender immediately and do not disclose the contents to another person, use it for any purpose or store or copy the information in any medium.

Re: libstdc Question about error 7

[Arsen Arsenović]

[-- Attachment #1: Type: text/plain, Size: 1856 bytes --]

Hi there,

CASTEL Alexandre -EXT <alexandre.castel-1-ext@alstomgroup.com> writes:

> Hello libstdc development group,
>
> We are contacting you about one issue we discover in a laboratory.
> We have made DMESG command on a linux computer, we saw that line:
> TB_RP_Server[27922]: segfault at 47e01d78 ip 48f708dc sp b695a030 error 7 in libstdc++.so.6.0.13[48ee1000+e1000]
>
> We know a segfault but we wanted to understand the end of this line "error 7 in libstdc++.so.6.0.13[48ee1000+e1000]"
> What does error 7 or 4 or other numbers mean?
> What does [48ee1000+e1000] represent?

These are Linux messages, rather than libstdc++ messages.  Linux prints
the error, as defined by the CPU, and the [foo+bar] values are the
address of the mapping the instruction pointer was in and the offset
into that mapping (foo being the mapping base, and bar the offset into
it).

On Intel (and compatible) CPUs, an error of four means the fault was a
userspace fault (so, the 4 bit will always be set), and (as the present
bit is unset) the target page was missing, while an error value of seven
means that a write to a present but unwritable page caused the fault
(bits 4, 2 and 1 are set, respectively being User, Write, Present).  See
the Intel Software Developer's Manual, Volume 3, Chapter 4.7 for a
precise meaning of the error bits; again, assuming you're on an Intel
compatible CPU (and getting a page fault).

I can't accurately say what function inside libstdc++ wrote to invalid
memory based on just the address you posted, but odds are it was given
an invalid pointer by accident.  The best course of action is likely
loading the coredump into GDB and getting a backtrace and inspecting the
program state, and then reporting a bug should my hunch be wrong :)

I hope that helps, have a lovely day.
-- 
Arsen Arsenović

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 381 bytes --]