From:
Operating system: CentOS5.5 32bits
PHP version: 5.2.14
Package: Output Control
Bug Type: Bug
Bug description:memory limit and custom error handler produces more errors than
expected
Description:
------------
When I get to the memory limit of a script, a fatal error raises (expected
behaviour). In order to nicely capture this error, I have created a custom
error handler (through register_shutdown_function) which checks whether the
script ended normally and so I can capture certain fatal errors. (Expected
behaviour)
Inside this error handler, I raise the memory limit in order to capture
this particular kind of error so that I can begin to check what happened.
When I check with empty or isset if a certain variable which shouldn't be
initialized yet exists (or doesn't evaluate to FALSE), the function returns
a rather strange error instead of the default FALSE or TRUE. (Not
expected).
Test script:
---------------
date_default_timezone_set('America/Santiago');
ini_set('memory_limit','1M');
function my_error_handler($errno = '0', $errstr = '[FATAL] General Error',
$errfile = 'N/A', $errline = 'N/A', $errctx = '') {
global $clean_exit;
if(empty($clean_exit)) { // if isset or !empty, the script would have
been exited cleanly.
ini_set('memory_limit','16M'); // This would be just to have enough
memory to print out the error.
echo '<h5>-- BEGIN ERROR --</h5>Error N°:
<strong>'.$errno.'</strong><br />Error Description:
<strong>'.$errstr.'</strong><br />File: <strong>'.$errfile.'</strong><br
/>Line: <strong>'.$errline.'</strong><br />Memory:
<strong>'.round(memory_get_usage() / 1024).'</strong>KiB /
<strong>'.round(memory_get_peak_usage() / 1024).'</strong>KiB';
if (isset($errctx['r']['print'])) echo 'THIS LINE GIVES THE ERROR AND
IS NOT PRINTED'; // isset or empty gives the same error
echo '<h5>-- END ERROR --</h5>';
die(); // shutdown_function == manual die() or else it will continue to
execute.
}
}
set_error_handler('my_error_handler');
register_shutdown_function('my_error_handler');
for ($i = 0; $i < 15000; $i++) $a[$i] = mt_rand(1,255);
$r['print'] = (string)$a[1]; // Just to fill something up
echo '<p>Everything fine.</p>';
$clean_exit = TRUE;
Expected result:
----------------
empty or isset returns FALSE, thus dying (normally) one level earlier.
Actual result:
--------------
An E_NOTICE appears:
Uninitialized string offset: 0 on file X line Y
--
Edit bug report at http://bugs.php.net/bug.php?id=52833&edit=1
--
Try a snapshot (PHP 5.2):
http://bugs.php.net/fix.php?id=52833&r=trysnapshot52
Try a snapshot (PHP 5.3):
http://bugs.php.net/fix.php?id=52833&r=trysnapshot53
Try a snapshot (trunk):
http://bugs.php.net/fix.php?id=52833&r=trysnapshottrunk
Fixed in SVN:
http://bugs.php.net/fix.php?id=52833&r=fixed
Fixed in SVN and need be documented:
http://bugs.php.net/fix.php?id=52833&r=needdocs
Fixed in release:
http://bugs.php.net/fix.php?id=52833&r=alreadyfixed
Need backtrace:
http://bugs.php.net/fix.php?id=52833&r=needtrace
Need Reproduce Script:
http://bugs.php.net/fix.php?id=52833&r=needscript
Try newer version:
http://bugs.php.net/fix.php?id=52833&r=oldversion
Not developer issue:
http://bugs.php.net/fix.php?id=52833&r=support
Expected behavior:
http://bugs.php.net/fix.php?id=52833&r=notwrong
Not enough info:
http://bugs.php.net/fix.php?id=52833&r=notenoughinfo
Submitted twice:
http://bugs.php.net/fix.php?id=52833&r=submittedtwice
register_globals:
http://bugs.php.net/fix.php?id=52833&r=globals
PHP 4 support discontinued: http://bugs.php.net/fix.php?id=52833&r=php4
Daylight Savings: http://bugs.php.net/fix.php?id=52833&r=dst
IIS Stability:
http://bugs.php.net/fix.php?id=52833&r=isapi
Install GNU Sed:
http://bugs.php.net/fix.php?id=52833&r=gnused
Floating point limitations:
http://bugs.php.net/fix.php?id=52833&r=float
No Zend Extensions:
http://bugs.php.net/fix.php?id=52833&r=nozend
MySQL Configuration Error:
http://bugs.php.net/fix.php?id=52833&r=mysqlcfg
