Re: Why does close_stdout close stdout and stderr?

[Florian Weimer]

* NeilBrown:

> On Tue, May 07 2019, Bruno Haible wrote:
>
>> Assaf Gordon wrote:
>>> 4.
>>> "fflush" instead of "fclose" seems to work OK, but I do not know
>>> if there are other side effects:
>>> 
>>>    $ ./aa stdout fflush > /dev/full && echo ok || echo error
>>>    aa: fflush failed: No space left on device
>>>    error
>>
>> Except that it does not work OK on NFS, as explained by the comment
>> in close-stream.c (written in 2006):
>>
>>                        Even calling fflush is not always sufficient,
>>    since some file systems (NFS and CODA) buffer written/flushed data
>>    until an actual close call.
>
> You can achieve that "actual close call" using
>
>   error = close(dup(fileno(stdout)));
>
> so you don't actually need to "fclose" if you don't want to.
> Any 'close' will do, it doesn't have to be the "last close".

Hah, thanks for this suggestion!  So something good came out of this
thread after all.  The big advantage of this approach is that this will
preserve the descriptor and the stream, so that further diagnostics from
the process are not suppressed.

Would someone who is familiar with the gnulib policies and procedures
please turn this into a proper patch, with error checking and all?  I
can try, but it's probably more work for you to bring what I could write
into an acceptable shape, than write the patch from scratch.

Thanks,
Florian