pg.ddx.io  pgsql-bugs@postgresql.org mailing list archive  
help / color / mirror / Atom feed
BUG #14468: One byte buffer overlow in quote_literal_cstr()
2+ messages / 2 participants
[nested] [flat]

* BUG #14468: One byte buffer overlow in quote_literal_cstr()
@ 2016-12-16 10:50  hlinnaka@iki.fi
  0 siblings, 1 reply; 2+ messages in thread

From: hlinnaka@iki.fi @ 2016-12-16 10:50 UTC (permalink / raw)
  To: pgsql-bugs

The following bug has been logged on the website:

Bug reference:      14468
Logged by:          Heikki Linnakangas
Email address:      hlinnaka@iki.fi
PostgreSQL version: 9.6.1
Operating system:   All
Description:        

postgres=# select format('%L', E'\\');
WARNING:  detected write past chunk end in ExprContext 0x55c65ff98fa8
 format 
--------
 E'\\'
(1 row)


This was originally reported against Greenplum, at
https://github.com/greenplum-db/gpdb/issues/1301. The code there isn't quite
the same, but it turned out to be a shared bug.

Looking at quote_literal_cstr(), it simply doesn't take into account the
space needed for the NULL terminator, when it allocates the buffer. The fix
is a straightforward + 1 to the allocation. I'll go and do that shortly, and
backpatch.


-- 
Sent via pgsql-bugs mailing list (pgsql-bugs@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-bugs


^ permalink  raw  reply  [nested|flat] 2+ messages in thread

* Re: BUG #14468: One byte buffer overlow in quote_literal_cstr()
@ 2016-12-16 14:43  Alvaro Herrera <alvherre@2ndquadrant.com>
  parent: hlinnaka@iki.fi
  0 siblings, 0 replies; 2+ messages in thread

From: Alvaro Herrera @ 2016-12-16 14:43 UTC (permalink / raw)
  To: hlinnaka@iki.fi; +Cc: pgsql-bugs

hlinnaka@iki.fi wrote:

> Looking at quote_literal_cstr(), it simply doesn't take into account the
> space needed for the NULL terminator, when it allocates the buffer. The fix
> is a straightforward + 1 to the allocation. I'll go and do that shortly, and
> backpatch.

Hm, that's a bit confusing.  I thought the +3 included the terminator,
but what it's actually for is the optional starting E; and the reason
quote_literal doesn't need similar patching it is that there is no
terminator 'cause it produces varlena.  I would think this deserves a
comment ...

-- 
Álvaro Herrera                https://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services


-- 
Sent via pgsql-bugs mailing list (pgsql-bugs@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-bugs



^ permalink  raw  reply  [nested|flat] 2+ messages in thread


end of thread, other threads:[~2016-12-16 14:43 UTC | newest]

Thread overview: 2+ messages (download: mbox mbox.gz follow: Atom feed)
-- links below jump to the message on this page --
2016-12-16 10:50 BUG #14468: One byte buffer overlow in quote_literal_cstr() hlinnaka@iki.fi
2016-12-16 14:43 ` Alvaro Herrera <alvherre@2ndquadrant.com>

This inbox is served by DDX for PostgreSQL; see mirroring instructions
for how to clone and mirror all data and code used for this inbox