Re: Updates of SE-PostgreSQL 8.4devel patches (r1268) - Mailing list pgsql-hackers

From Peter Eisentraut
Subject Re: Updates of SE-PostgreSQL 8.4devel patches (r1268)
Date
Msg-id 200812111943.17326.peter_e@gmx.net
Whole thread Raw
In response to Re: Updates of SE-PostgreSQL 8.4devel patches (r1268)  (Tom Lane <tgl@sss.pgh.pa.us>)
Responses Re: Updates of SE-PostgreSQL 8.4devel patches (r1268)  (Tom Lane <tgl@sss.pgh.pa.us>)
Re: Updates of SE-PostgreSQL 8.4devel patches (r1268)  (Gregory Stark <stark@enterprisedb.com>)
Re: Updates of SE-PostgreSQL 8.4devel patches (r1268)  (KaiGai Kohei <kaigai@ak.jp.nec.com>)
List pgsql-hackers
On Thursday 11 December 2008 18:32:50 Tom Lane wrote:
> > How can we stick all of these in the same column at the same time?
>
> Why would we want to?

Because we want to use SQL-based row access control and SELinux-based row 
access control at the same time.  Isn't this exactly one of the objections 
upthread?  Both must be available at the same time.

We can debate the merits of having, say, SELinux plus Solaris TX at the same 
time, but if we can have two as per previous paragraph, we should design for 
several.

> I think one column that can hold any of these 
> ought to be sufficient.  I certainly don't care for the idea that we
> might invent still a third column for Solaris TX at some future time.

Yes, it is certainly more appealing to have one column describing all access 
rights.

In fact, if we extend the ACL storage structure to store external access 
control information, we might also consider using that for system object 
access.  So instead of adding a column to pg_class for SELinux-controlled 
access to tables, we just reused relacl.


pgsql-hackers by date:

Previous
From: "Greg Stark"
Date:
Subject: Re: posix_fadvise v22
Next
From: Tom Lane
Date:
Subject: Re: benchmarking the query planner