php.net |  support |  documentation |  report a bug |  advanced search |  search howto |  statistics |  random bug |  login
Bug #22975 PHP crashed at/after the 26 system-call
Submitted: 2003-03-31 03:30 UTC Modified: 2003-04-23 04:25 UTC
From: kju at arcor dot de Assigned:
Status: Not a bug Package: Program Execution
PHP Version: 4.3.2-RC OS: Windows NT 4 SP 6a
Private report: No CVE-ID: None
View Add Comment Developer Edit
Welcome! If you don't have a Git account, you can't do anything here.
You can add a comment by following this link or if you reported this bug, you can edit this bug over here.
(description)
Block user comment
Status: Assign to:
Package:
Bug Type:
Summary:
From: kju at arcor dot de
New email:
PHP Version: OS:

 

 [2003-03-31 03:30 UTC] kju at arcor dot de
After we have updated the PHP-Version from 4.1 to 4.3.1 we have reproducable crashes...

In some of our scripts are exec-commands to use system-commands... With version 4.1 all our scripts were ok...
But after updating the scripts crash at the 26. system-call...

I have a small example here:

--
n=30;
for ($i=0;$i<$n;$i++)
{
   exec("echo $i >> c:\test.log");
}
--

For any n<26 the script is running fine...
But when n=26 and above the script crashes and the last entry in the test.log-file is "25" (the 26. system-call).

We have tested the system-calls with exec, system and passthru: same results.
We also have tested many different commands also with commands who don't use the filesystem, f.e. "ping $ip".

We also have increased the "set_time_limit" up to a very high  value (much more than we need normally) but the problem is persistent... no "normal ending"... 

Patches

Add a Patch

Pull Requests

Add a Pull Request

History

AllCommentsChangesGit/SVN commitsRelated reports
 [2003-03-31 04:34 UTC] kju at arcor dot de
we have installed the following version (win32):
PHP Version 4.3.2-RC 
Build Date  Mar 31 2003 09:11:59  

No improvements...
 [2003-03-31 18:25 UTC] sniper@php.net
(updated PHP version info)

And could you explain what you mean with 'crashed'?
Are you using CGI? Under IIS or Apache?


 [2003-04-01 00:29 UTC] kju at arcor dot de
We are using Apache/1.3.26 (Win32) on a WindowsNT4-Server with SP 6a...

PHP is running as CGI...
Switching the system to use PHP as module isn't easy in our environment, so we couldn't test it...

Hmmm, we have tested something else yet, we have started the script at the command-line... the script ends without error! all output is as expected.

But there seems to bee no problem with the apache-configuration, because when we turn back to PHP 4.1 all is running well!

How we should explain 'crashed'?
I will try to do that...
Starting a PHP-Script on the WebServer build a new PHP-process on the machine... After finishing the script the process will end... thats the normal way, isn't it?

But when n>26 the script (and the process) will never end... We have set the "set_limit_time" to 30 seconds, after 5 minutes the process is still "alive", but it doesn't use CPUTime any longer.
No output to the webbrowser is generated.
The output "exec("echo $i >> c:\test.log");" is done up to i=26. No hint concerning the abort is given from PHP.

What do you think of the source of the error: OS, Apache or PHP?
 [2003-04-01 04:43 UTC] kju at arcor dot de
additional information (problem not solved yet):

Using the dev-Version produces entries in the nt-eventlog 
PHP Version 4.3.2-RC 
Build Date  Mar 31 2003 09:11:59  

Error Detail in the nt-system-log:
--
User: N/A
EventID: 2000
Source: Srv
Type: Error
Category: None
Description: The server's call to a system service failed unexpectedly.
--
Starting ONE skript (even phpmyadmin 2.4.0) will put many of these errors in the system-log.

There are no errors after we have downgraded to 4.3.1, so the error-messages seem to be a problem/feature in the 4.3.2-dev
 [2003-04-23 04:25 UTC] sniper@php.net
If it works on cmd line but not from Apache -> apache bug.

 
PHP Copyright © 2001-2024 The PHP Group
All rights reserved.
Last updated: Fri Mar 29 01:01:28 2024 UTC