Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > comp.lang.php > #10759 > unrolled thread

Confirm PG-related segfault in current PHP version

Started byFiver <numeralfive@gmail.com>
First post2013-09-14 02:12 +0200
Last post2013-09-14 04:33 +0200
Articles 3 — 2 participants

Back to article view | Back to comp.lang.php


Contents

  Confirm PG-related segfault in current PHP version Fiver <numeralfive@gmail.com> - 2013-09-14 02:12 +0200
    Re: Confirm PG-related segfault in current PHP version lb <lb2048@pobox.com> - 2013-09-13 21:55 -0400
      Re: Confirm PG-related segfault in current PHP version Fiver <numeralfive@gmail.com> - 2013-09-14 04:33 +0200

#10759 — Confirm PG-related segfault in current PHP version

FromFiver <numeralfive@gmail.com>
Date2013-09-14 02:12 +0200
SubjectConfirm PG-related segfault in current PHP version
Message-ID<l109l1$mvj$1@dont-email.me>
If anybody here is running an *up-to-date* version of PHP, and has
access to a PostgreSQL database, could you please check if this results
in a segfault for you?

  $conn = pg_connect("dbname=postgres user=temp");
  pg_trace("_trace.log", "w", $conn);
  pg_query($conn, "SELECT 1 FROM xxx");
  //pg_untrace($conn);  // <-- untrace prevents segfault
  echo "still alive\n";

The database and query don't matter, as long as you can connect and the
pg_query() statement does not succeed.

I strongly suspect (from previous experience with several segfaults in
the pg module over the years) that this bug is still present in the
master branch, but I can't report it because I'm running PHP 5.4.9.

Compiling and configuring PHP with the necessary dependencies to
reproduce this bug is a major PITA, I don't have time for this. If I can
get confirmation that this is still an issue, I'll report it, otherwise
hopefully somebody else will discover and report it in the future.

regards,
5er

[toc] | [next] | [standalone]


#10760

Fromlb <lb2048@pobox.com>
Date2013-09-13 21:55 -0400
Message-ID<5233c208$0$64378$862e30e2@ngroups.net>
In reply to#10759
Fiver wrote:
> If anybody here is running an *up-to-date* version of PHP, and has
> access to a PostgreSQL database, could you please check if this results
> in a segfault for you?
>
>    $conn = pg_connect("dbname=postgres user=temp");
>    pg_trace("_trace.log", "w", $conn);
>    pg_query($conn, "SELECT 1 FROM xxx");
>    //pg_untrace($conn);  // <-- untrace prevents segfault
>    echo "still alive\n";
>
> The database and query don't matter, as long as you can connect and the
> pg_query() statement does not succeed.
>
> I strongly suspect (from previous experience with several segfaults in
> the pg module over the years) that this bug is still present in the
> master branch, but I can't report it because I'm running PHP 5.4.9.
>
> Compiling and configuring PHP with the necessary dependencies to
> reproduce this bug is a major PITA, I don't have time for this. If I can
> get confirmation that this is still an issue, I'll report it, otherwise
> hopefully somebody else will discover and report it in the future.

I ran your script with PHP-5.5.3 (CLI) talking to a PostgreSQL-9.3.0 server. 
Connection OK, trace file written, query gets error, no segfault, says "still 
alive" at end.

[toc] | [prev] | [next] | [standalone]


#10761

FromFiver <numeralfive@gmail.com>
Date2013-09-14 04:33 +0200
Message-ID<l10htb$mie$1@dont-email.me>
In reply to#10760
On 2013-09-14 03:55, lb wrote:
> I ran your script with PHP-5.5.3 (CLI) talking to a PostgreSQL-9.3.0 server. 
> Connection OK, trace file written, query gets error, no segfault, says "still 
> alive" at end.

That's perfect, thanks.

5er

[toc] | [prev] | [standalone]


Back to top | Article view | comp.lang.php


csiph-web