Re: include_path question
| From: | Zeev Suraski | Date: | Sun, 27 Aug 2000 21:25:31 +0000 |
| Subject: | Re: include_path question | ||
| References: | 1 2 3 | Groups: | php.dev |
| Request: | Send a blank email to php-dev+get-30857@lists.php.net to get a copy of this message | ||
At 22:55 27/08/2000, Andi Gutmans wrote:
At 10:43 PM 8/27/00 +0300, Andi Gutmans wrote:We don't have just 10KB of stack space under ISAPI, we have 256KB. It's very little as well, though. Zeev -- Zeev Suraski <zeev@zend.com> http://www.zend.com/At 09:39 PM 8/27/00 +0200, Sascha Schumann wrote:I thought it doesn't work with CGI but I checked and CGI seems to beWhat is wrong with your patch? It should work with every web server which passes an absolute path in primary_file.boy and set path_translated to full path too :)a goodI reverted it myself. It's really not correct (please read the commit msg). I'd appreciate it if you left the code the way it is. By the way, in my commit msg I also mention old_cwd(). In ISAPI we have 10KB of stack space so having it take up 4KB is a big problem because it's in the call-stack to execute_script() and not a small temporary function. We should probably emalloc() it.I don't like your patch. It shouldn't be there because it is broken. SAPI modules should make sure that path_translated is full path. It is supposed to be this way and I doubt the modules do anything differently. If we find such a SAPI module the full path needs to be calculated there and not in main.c. Please let's try and keep the core code as clean as possible. Please revert this patch.The last change should take care of all SAPI modules which pass a relative path to php_execute_script().