Re: pg_dump does not handle indirectly-granted permissions properly

From: Stephen Frost <sfrost(at)snowman(dot)net>
To: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
Cc: pgsql-hackers(at)postgreSQL(dot)org
Subject: Re: pg_dump does not handle indirectly-granted permissions properly
Date: 2017-08-04 01:46:13
Message-ID: 20170804014613.GN1769@tamriel.snowman.net
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

Tom, all,

* Stephen Frost (sfrost(at)snowman(dot)net) wrote:
> This needs more cleanup, testing, and comments explaining why we're
> doing this (and then perhaps comments, somewhere.. in the backend ACL
> code that explains that the ordering needs to be preserved), but the
> basic idea seems sound to me and the case you presented does work with
> this patch (for me, at least) whereas what's in master didn't.

Alright, here's an updated patch which cleans things up a bit and adds
comments to explain what's going on. I also updated the comments in
acl.h to explain that ordering actually does matter.

I've tried a bit to break the ordering in the backend a bit but there
could probably be more effort put into that, if I'm being honest.
Still, this definitely fixes the case which was being complained about
and therefore is a step in the right direction.

It's a bit late here, so I'll push this in the morning and watch the
buildfarm.

Thanks!

Stephen

Attachment Content-Type Size
fix_pg_dump_grant_ordering_v2-master.patch text/x-diff 6.3 KB

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Stephen Frost 2017-08-04 01:48:20 Re: pg_stop_backup(wait_for_archive := true) on standby server
Previous Message Amit Langote 2017-08-04 01:43:31 Re: reload-through-the-top-parent switch the partition table