ID: 31854
User updated by: bertrand at toggg dot com
Reported By: bertrand at toggg dot com
-Status: Feedback
+Status: Open
Bug Type: Reproducible crash
-Operating System: FC2 kernel-2.6.10-1.12
+Operating System: FC2 kernel-2.6.10-1.14
-PHP Version: 4CVS-2005-02-06
+PHP Version: 4CVS-2005-02-16
New Comment:
Hope it's usefull for you:
Core was generated by `./php outmem.php 18 12 100'.
Program terminated with signal 11, Segmentation fault.
Reading symbols from /lib/libcrypt.so.1...done.
Loaded symbols for /lib/libcrypt.so.1
Reading symbols from /lib/libresolv.so.2...done.
Loaded symbols for /lib/libresolv.so.2
Reading symbols from /lib/tls/libm.so.6...done.
Loaded symbols for /lib/tls/libm.so.6
Reading symbols from /lib/libdl.so.2...done.
Loaded symbols for /lib/libdl.so.2
Reading symbols from /lib/libnsl.so.1...done.
Loaded symbols for /lib/libnsl.so.1
Reading symbols from /lib/tls/libc.so.6...done.
Loaded symbols for /lib/tls/libc.so.6
Reading symbols from /lib/ld-linux.so.2...done.
Loaded symbols for /lib/ld-linux.so.2
Reading symbols from /lib/libnss_files.so.2...done.
Loaded symbols for /lib/libnss_files.so.2
#0 0x00c713d4 in _int_free () from /lib/tls/libc.so.6
(gdb) bt
#0 0x00c713d4 in _int_free () from /lib/tls/libc.so.6
#1 0x00c7272b in free () from /lib/tls/libc.so.6
#2 0x081041c9 in shutdown_memory_manager (silent=0, clean_cache=0)
at
/home/bertrand/down/php4-STABLE-200502160330/Zend/zend_alloc.c:492
#3 0x080eb490 in php_request_shutdown (dummy=0x0)
at /home/bertrand/down/php4-STABLE-200502160330/main/main.c:1004
#4 0x0811fa1f in main (argc=5, argv=0xbff00a34)
at
/home/bertrand/down/php4-STABLE-200502160330/sapi/cli/php_cli.c:876
(gdb)
Previous Comments:
------------------------------------------------------------------------
[2005-02-16 03:24:17] [EMAIL PROTECTED]
Thank you for this bug report. To properly diagnose the problem, we
need a backtrace to see what is happening behind the scenes. To
find out how to generate a backtrace, please read
http://bugs.php.net/bugs-generating-backtrace.php
Once you have generated a backtrace, please submit it to this bug
report and change the status back to "Open". Thank you for helping
us make PHP better.
I can't get it to crash with any parameters..
------------------------------------------------------------------------
[2005-02-06 11:30:08] bertrand at toggg dot com
I downloaded the CVS snapshot from this morning,
php4-STABLE-200502060730 unix version
I build only the executables:
./configure --enable-memory-limit
make
With sapi/cli/php or sapi/cgi/php, unfortunately the results are the
same.
Only one point is now better, it's the case where no memory_limit set
and less call to memory_get_usage:
php outmem.php 18 '' 1000000
17:2/88792 bytes
16:4/88888 bytes
<...snip...>
1:131072/5855880 bytes
Fatal error: Allowed memory size of 8388608 bytes exhausted (tried to
allocate 35 bytes) in /home/bertrand/prog/test/outmem.php on line 19
<<< here it's still hanging a long time >>>
Allowed memory size of 8388608 bytes exhausted (tried to allocate 129
bytes)
But then it's coming back from PHP, no need no more to break. Is it
only due to the fact it's an only CLI PHP ?
Just to be sure, I've also rebuild some php-4.3.9 from 2004/10/09 and
results are identical.
------------------------------------------------------------------------
[2005-02-06 06:57:55] [EMAIL PROTECTED]
Please try using this CVS snapshot:
http://snaps.php.net/php4-STABLE-latest.tar.gz
For Windows:
http://snaps.php.net/win32/php4-win32-STABLE-latest.zip
------------------------------------------------------------------------
[2005-02-05 20:31:24] bertrand at toggg dot com
Description:
------------
This script doubles and again an array of long strings.
It accepts 3 parameters:
- nb of times to double the array
- eventuel memory_limit to set (thus default 8 Mo)
- interval of added rows to check memory_get_usage.
By 18 loops the 8Mo are exhausted.
Depending on the memory setting and the interval to check memory usage,
results are somewhat strange.
The segmentation fault occurs the same if running from Apache 2.0
Handler
It could be related to bug #31624
Reproduce code:
---------------
$loop = isset($_SERVER['argv'][1]) ? $_SERVER['argv'][1]+0 : 11;
$setmem = isset($_SERVER['argv'][2]) ? $_SERVER['argv'][2]+0 : ''; //
changed if set
$chk = isset($_SERVER['argv'][3]) ? $_SERVER['argv'][3]+0 : 100;
if ($setmem) {
if (ini_set ('memory_limit', $setmem*1048576)) {
echo 'Set memory limit to '.$setmem." Mo\n";
} else {
echo 'FAILED to set memory limit to '.$setmem." Mo\n";
}
}
error_reporting(E_ALL);
$arr = array (str_repeat('X', 65536));
$mem = 0;
while ($loop--) {
for ($i = count($arr); $i; $i--) {
$arr[] = $arr[0];
if ($i%$chk) continue;
if ( ( ($nmem = memory_get_usage()) - $mem) > 1000000) {
$mem = $nmem;
echo 'Count:'.count($arr)." ($mem bytes)\n";
}
}
echo $loop.':'.count($arr).'/'.memory_get_usage() . " bytes\n"; //
36640
}
echo "\n OK \n";
Expected result:
----------------
1) no memory_limit set
PHP Fatal error: Allowed memory size of 8388608 bytes exhausted (tried
to allocate 35 bytes) in /home/bertrand/prog/test/outmem.php on line 19
+ break
2) with memory_limit set.
If not enough: same as 1)
Enough memory: OK
Actual result:
--------------
1) no memory_limit set
I actually get memory exhausted, but if I lower the memory_get_usage
frequence, then no break, must control-C:
php outmem.php 18 '' 100 ---> break, get hand back
php outmem.php 18 '' 1000 ---> I must abort
That means if I check memory_get_usage only each 1000 rows PHP is not
coming back, but everything OK if I check each 100 rows ????
PHP Fatal error: Allowed memory size of 8388608 bytes exhausted (tried
to allocate 35 bytes) in /home/bertrand/prog/test/outmem.php on line 19
Allowed memory size of 8388608 bytes exhausted (tried to allocate 129
bytes)
The second message is coming after a long while, but the PHP is
sleeping and I need to break
2) with memory_limit
It's as expected but will in both cases (enough mem or not) make a
segmentation fault.
------------------------------------------------------------------------
--
Edit this bug report at http://bugs.php.net/?id=31854&edit=1
