Re: Errands around AllocateDir()

Michael Paquier <michael.paquier@gmail.com>

From: Michael Paquier <michael.paquier@gmail.com>
To: Tom Lane <tgl@sss.pgh.pa.us>
Cc: PostgreSQL mailing lists <pgsql-hackers@postgresql.org>
Date: 2017-12-05T01:29:21Z
Lists: pgsql-hackers
On Tue, Dec 5, 2017 at 7:17 AM, Tom Lane <tgl@sss.pgh.pa.us> wrote:
> Michael Paquier <michael.paquier@gmail.com> writes:
>> I had a close look at all the callers of AllocateDir() and noticed a
>> couple of unwelcome things (Tom noticed some of those in the thread
>> mentioned above, I found others):
>
> Pushed with some minor additional fiddling.

Thanks.

> The most notable thing I changed was that instead of this:
>
>> - perform_base_backup() makes the mistake of not saving errno before
>> CheckXLogRemoved() when AllocateDir returns NULL, which can lead to an
>> incorrect error message.
>
> I modified CheckXLogRemoved() to internally guarantee that it does not
> change errno.  This is because there seemed to be other call sites that
> were depending on that, not just this one.  Anyway, that seemed like a
> more future-proof fix than relying on callers to deal with it.

Hm, OK. Yes I can see the point behind this way of doing instead,
CheckXLogRemoved() is much used in error code paths.
-- 
Michael


Commits

  1. Clean up assorted messiness around AllocateDir() usage.