Hi,
I modified the patch, kept the old call of SlruReportIOError()
outside the for-loop, and add a call of erport() with LOG level
when file-close failure occurs in the for-loop.
Please check whether this modification is appropriate
and let me know if there is any problem. Thank you.
At 2024-06-04 14:44:09, "Kyotaro Horiguchi" <horikyota.ntt@gmail.com> wrote:
>At Tue, 28 May 2024 20:15:59 +0800 (CST), "Long Song" <songlong88@126.com> wrote in
>>
>> Hi,
>> Actually, I still wonder why only the error message
>> of the last failure to close the file was recorded.
>> For this unusual situation, it is acceptable to
>> record all failure information without causing
>> too much logging.
>> Was it designed that way on purpose?
>
>Note that SlruReportIOError() causes a non-local exit. To me, the
>point of the loop seems to be that we want to close every single file,
>apart from the failed ones. From that perspective, the patch disrupts
>that intended behavior by exiting in the middle of the loop. It seems
>we didn't want to bother collecting errors for every failed file in
>that part.
>
>regards.
>
>--
>Kyotaro Horiguchi
>NTT Open Source Software Center
--
Best Regards,
Long