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

[Xi Ruoyao <xry111@mengyan1223.wang>]

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