Re: Re: pgsql: Convert contrib/seg's bool-returning SQL functions to V1 call co - Mailing list pgsql-committers

From Andres Freund
Subject Re: Re: pgsql: Convert contrib/seg's bool-returning SQL functions to V1 call co
Date
Msg-id 20160427030411.hjod6bja6pjnmlab@alap3.anarazel.de
Whole thread Raw
In response to Re: Re: pgsql: Convert contrib/seg's bool-returning SQL functions to V1 call co  (Tom Lane <tgl@sss.pgh.pa.us>)
Responses Re: Re: pgsql: Convert contrib/seg's bool-returning SQL functions to V1 call co  (Michael Paquier <michael.paquier@gmail.com>)
List pgsql-committers
On 2016-04-26 22:59:44 -0400, Tom Lane wrote:
> What's the argument that it makes debugging harder?  Especially if
> you aren't using it?

If you try to write a V1 function, but forget or mistype/rename the
function in PG_FUNCTION_INFO_V1, you'll get crashes, at least if you're
lucky.


> I don't particularly buy the "easier exploitation" argument, either.
> You can't create a C function without superuser, and if you've got
> superuser there are plenty of ways to run arbitrary code.

Without pl*u installed, I don't think any of them are as simple as
calling system(). But yea, it's not a very high barrier.


pgsql-committers by date:

Previous
From: Tom Lane
Date:
Subject: Re: Re: pgsql: Convert contrib/seg's bool-returning SQL functions to V1 call co
Next
From: Andres Freund
Date:
Subject: pgsql: Emit invalidations to standby for transactions without xid.