What's the possible reason of fclose core?

What's the possible reason of fclose core?

[okay]

#0  0x00007faad574d2a7 in raise () from /lib64/libc.so.6
#1  0x00007faad574e67a in abort () from /lib64/libc.so.6
#2  0x00007faad578c4f4 in __libc_message () from /lib64/libc.so.6
#3  0x00007faad5791966 in malloc_printerr () from /lib64/libc.so.6
#4  0x00007faad5791d02 in malloc_consolidate () from /lib64/libc.so.6
#5  0x00007faad5792753 in _int_free () from /lib64/libc.so.6
#6  0x00007faad5782bb9 in fclose@@GLIBC_2.2.5 () from /lib64/libc.so.6

 From the above core stack, because i can't see the parameter(file pointer, abbrv fp) value of fclose at frame 6, so fp  exists the following possible condition:
1) fp is NULL
2) fp is changed during running time
3) call fclose(fp)  twice continuously 
4) others reason haven't guessed.

  So i want to ask what's the possible condition can cause the above core? 
  Is it possible that other thread in same service change the fp value such as address violation operation?

  Thanks!

Re: What's the possible reason of fclose core?

[Xi Ruoyao]

On Tue, 2022-04-26 at 22:45 +0800, okay via Gcc-help wrote:
> 
> #0  0x00007faad574d2a7 in raise () from /lib64/libc.so.6
> #1  0x00007faad574e67a in abort () from /lib64/libc.so.6
> #2  0x00007faad578c4f4 in __libc_message () from /lib64/libc.so.6
> #3  0x00007faad5791966 in malloc_printerr () from /lib64/libc.so.6
> #4  0x00007faad5791d02 in malloc_consolidate () from /lib64/libc.so.6
> #5  0x00007faad5792753 in _int_free () from /lib64/libc.so.6
> #6  0x00007faad5782bb9 in fclose@@GLIBC_2.2.5 () from /lib64/libc.so.6
> 
>  From the above core stack, because i can't see the parameter(file
> pointer, abbrv fp) value of fclose at frame 6, so fp  exists the
> following possible condition:
> 1) fp is NULL
> 2) fp is changed during running time
> 3) call fclose(fp)  twice continuously 
> 4) others reason haven't guessed.
> 
>   So i want to ask what's the possible condition can cause the above
> core? 
>   Is it possible that other thread in same service change the fp value
> such as address violation operation?

gcc-help is not a proper channel to discuss such a topic without
anything specific to gcc.

That being said, you may try -fanalyzer which is helpful to catch the
condition (3).
-- 
Xi Ruoyao <xry111@mengyan1223.wang>
School of Aerospace Science and Technology, Xidian University

Re: What's the possible reason of fclose core?

[Florian Weimer]

* okay via Gcc-help:

> #0  0x00007faad574d2a7 in raise () from /lib64/libc.so.6
> #1  0x00007faad574e67a in abort () from /lib64/libc.so.6
> #2  0x00007faad578c4f4 in __libc_message () from /lib64/libc.so.6
> #3  0x00007faad5791966 in malloc_printerr () from /lib64/libc.so.6
> #4  0x00007faad5791d02 in malloc_consolidate () from /lib64/libc.so.6
> #5  0x00007faad5792753 in _int_free () from /lib64/libc.so.6
> #6  0x00007faad5782bb9 in fclose@@GLIBC_2.2.5 () from /lib64/libc.so.6
>
>  From the above core stack, because i can't see the parameter(file pointer, abbrv fp) value of fclose at frame 6, so fp  exists the following possible condition:
> 1) fp is NULL
> 2) fp is changed during running time
> 3) call fclose(fp)  twice continuously 
> 4) others reason haven't guessed.
>
>   So i want to ask what's the possible condition can cause the above core? 

It can also be the result of some form of heap corruption, say after a
heap buffer overflow.

Thanks,
Florian