Bug 1071344

Summary: PHP segfault's when you try and allocate an SplFixedArray with size >= 9999
Product: Red Hat Enterprise Linux 6 Reporter: Coty Sutherland <csutherl>
Component: phpAssignee: Web Stack Team <webstack-team>
Status: CLOSED ERRATA QA Contact: Petr Šplíchal <psplicha>
Severity: medium Docs Contact:
Priority: medium    
Version: 6.5CC: csutherl, dkutalek, jorton, m.slowe, ohudlick, p.frost, psplicha, rcollet
Target Milestone: rcKeywords: Patch, Upstream
Target Release: ---   
Hardware: x86_64   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2016-05-10 21:37:48 UTC Type: Bug
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: ---
Bug Depends On:    
Bug Blocks: 1075802, 1172231    
Attachments:
Description Flags
PHP script to reproduce behavior. none

Description Coty Sutherland 2014-02-28 14:41:03 UTC
Description of problem:
When trying to fill an SplFixedArray object of length >= 9999 you get a segfault.

Version-Release number of selected component (if applicable):
6.5

How reproducible:
Every time when running with a PHP version < 5.3.8.

Steps to Reproduce:
1. Execute attached script.
2.
3.

Actual results:
With LENGTH < 9999: works fine.
With LENGTH >= 9999: Either a segfault or a 'zend_mm_heap_corrupted' error depending on whether the inner array is a regular array or a fixed array.
Output:
Array
 *RECURSION*Array
 ....
 *RECURSION*Array
(
Segmentation fault (core dumped)

Expected results:
Array
(
    [0] => test
    [1] => test2
)
.... // repeated about 9998 times

Additional info:
This seems to occur with PHP versions less than 5.3.8. I attached a bug report from years ago that was for this problem, but was closed because the problem disappeared in later versions. I was unable to find where it was fixed though.

Comment 1 Coty Sutherland 2014-02-28 14:42:10 UTC
Created attachment 869093 [details]
PHP script to reproduce behavior.

Comment 3 Remi Collet 2014-03-21 14:15:34 UTC
I think this is related to upstream bug 54609.

Upstream fix:
http://git.php.net/?p=php-src.git;a=commit;h=0afd4bd18bef55d8bce4d0fe854a65781477029a

This is related to the garbage collector.

With the provided reproducer, and a regular array

  $ php -d zend.enable_gc=1 /tmp/foo.php
  Array
   *RECURSION*Array
   ...
   *RECURSION*Array
  (
  Segmentation fault (core dumped)

With gc disabled => no segfault

  $ php -d zend.enable_gc=0 /tmp/foo.php
  ...
  Array
  (
      [0] => test
      [1] => test2
  )

With the provided reproducer, and a fixed array

  $ php -d zend.enable_gc=1 /tmp/foo.php 
  Segmentation fault (core dumped)

With gc disabled => no segfault

  $ php -d zend.enable_gc=0 /tmp/foo.php
  ...

         [8] => 
      [9] => 
  )

So can you please with test "zend.enable_gc=0" to confirm this.

Does the customer would be able to test a scratch build with this fix applied ?

Comment 4 Coty Sutherland 2014-03-24 14:22:40 UTC
Hi Remi,

That works for me in my environment. I will pass it on the customer and verify it works for them and also check to see if they can test a build with the fix applied.



Thanks!

Comment 5 Coty Sutherland 2014-03-25 13:20:30 UTC
> So can you please with test "zend.enable_gc=0" to confirm this.

"Yes -- disabling GC appears to make the segfault go away."

> Does the customer would be able to test a scratch build with this fix applied ?

They are willing to test a build if we get them one.

Comment 18 errata-xmlrpc 2016-05-10 21:37:48 UTC
Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.

For information on the advisory, and where to find the updated
files, follow the link below.

If the solution does not work for you, open a new bug report.

https://rhn.redhat.com/errata/RHBA-2016-0842.html