Re: file born 30 seconds after its creation on ext4 - bug?
On 2022-04-26 14:18:58 -0400, Stefan Monnier wrote:
> > On an ext4 filesystem, I got a file born 30 seconds after its
> > actual creation. Is this a bug?
>
> I doubt it.
> Note that a file's atime/mtime/ctime is a property of the file itself,
> whereas "appearing" is defined by when the name you're checking
> becomes a link to it.
>
> IOW, most likely the file was first created under a different name and
> then renamed. This is very standard practice to make sure the final
> file name never refers to an incomplete file.
You mean renamed by the Linux kernel??? (The script doesn't rename it.)
FYI, the script is the following one:
https://gitlab.inria.fr/mpfr/misc/-/blob/fed7770cf5f712871bd116ef80d93ea5885fc3f7/vl-tests/mpfrtests.sh
and the file in question is what appears as "$out".
And what I did was running from a MPFR working tree
/path/to/mpfrtests.sh < /path/to/mpfrtests.data
where the mpfrtests.data file is the following one:
https://gitlab.inria.fr/mpfr/misc/-/blob/e0204b3423b9bc25c31548d2acc5b8e19a73f48d/vl-tests/mpfrtests.data
Note that I had never had any issue until now. This was the first
time I got errors saying that the file was missing.
--
Vincent Lefèvre <vincent@vinc17.net> - Web: <https://www.vinc17.net/>
100% accessible validated (X)HTML - Blog: <https://www.vinc17.net/blog/>
Work: CR INRIA - computer arithmetic / AriC project (LIP, ENS-Lyon)
Reply to: