Re: closing file in adjust_data_dir

From: Ted Yu <yuzhihong(at)gmail(dot)com>
To: Japin Li <japinli(at)hotmail(dot)com>
Cc: pgsql-hackers(at)lists(dot)postgresql(dot)org
Subject: Re: closing file in adjust_data_dir
Date: 2022-11-16 03:15:57
Message-ID: CALte62wtdry5u=_Pf5AyVf4zg1hDx_6nPEmjJSg9N3oZRpD4Ug@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

On Tue, Nov 15, 2022 at 7:12 PM Japin Li <japinli(at)hotmail(dot)com> wrote:

>
> On Wed, 16 Nov 2022 at 10:52, Ted Yu <yuzhihong(at)gmail(dot)com> wrote:
> > On Tue, Nov 15, 2022 at 6:35 PM Japin Li <japinli(at)hotmail(dot)com> wrote:
> >>
> >> fd = popen(cmd, "r");
> >> - if (fd == NULL || fgets(filename, sizeof(filename), fd) == NULL
> ||
> >> pclose(fd) != 0)
> >> + if (fd == NULL || fgets(filename, sizeof(filename), fd) == NULL)
> >> {
> >> + pclose(fd);
> >> write_stderr(_("%s: could not determine the data
> directory
> >> using command \"%s\"\n"), progname, cmd);
> >> exit(1);
> >> }
> >>
> >> Here, segfault maybe occurs if fd is NULL. I think we can remove
> pclose()
> >> safely since the process will exit.
> >>
> >
> > That means we're going back to v1 of the patch.
> >
>
> After some rethinking, I find the origin code do not have problems.
>
> If fd is NULL or fgets() returns NULL, the process exits. Otherwise, we
> call
> pclose() to close fd. The code isn't straightforward, however, it is
> correct.
>
>
>
> Please read this sentence from my first post:

If the fgets() call doesn't return NULL, the pclose() would be skipped.

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Justin Pryzby 2022-11-16 03:17:28 Re: libpq compression (part 2)
Previous Message Japin Li 2022-11-16 03:11:52 Re: closing file in adjust_data_dir