RE: [PHP-DEV] GDLib/reentrant
| From: | Shane Caraveo | Date: | Sun, 13 Sep 1998 00:22:18 +0000 |
| Subject: | RE: [PHP-DEV] GDLib/reentrant | ||
| References: | 1 | Groups: | php.dev |
| Request: | Send a blank email to php-dev+get-1017@lists.php.net to get a copy of this message | ||
Ooops, it's not working yet, looking into it.
Shane
> -----Original Message-----
> From: Shane Caraveo [mailto:shane@caraveo.com]
> Sent: Saturday, September 12, 1998 7:49 PM
> To: 'Rasmus Lerdorf'; 'Bug Database'
> Subject: RE: [PHP-DEV] GDLib/reentrant
>
>
> > > GD is awefully useful and it would be sad to lose it
> > completely in PHP
> > > 3.1. Could we not stick a lock around these functions and
> > > only allow one
> > > thread into them at any one time?
> >
> >
> > The only thing
> > that will realy fix it in a useful way is to make gdlib thread safe.
> > Eventualy, if someone else doesn't get to it, I might. The
>
> Ok, I took a quick look, and realized there was an easy way to make gd
> reentrant. Because all the gdlib api functions require a
> pointer to a gd
> image structure (except those that create it) it was easy to
> add a structure
> to that that holds all the globals. That way, we pass all the globals
> around with the image structure. This took a couple hours work to
> implement, but now I have it...a reentrant gdlib. This
> should work on unix
> as well, and with all existing programs that use gdlib, as it
> doesn't change
> the api at all. If someone can run this through the wringer,
> that would be
> greate. I ran the test programs that come with gd, without
> problem. I've
> cvs'd rgd.zip into the 3.0 tree under the extra directory.
>
> Shane
>
>
> --
> PHP Development Mailing List http://www.php.net/
> To unsubscribe send an empty message to
> php-dev-unsubscribe@lists.php.net
> For help: php-dev-help@lists.php.net
>
>
--
PHP Development Mailing List http://www.php.net/
To unsubscribe send an empty message to php-dev-unsubscribe@lists.php.net
For help: php-dev-help@lists.php.net