php.net |  support |  documentation |  report a bug |  advanced search |  search howto |  statistics |  random bug |  login
Bug #57262 oci8 crashes when inserting multiple LOBs with one prepared statement
Submitted: 2006-09-27 11:51 UTC Modified: 2007-02-13 18:36 UTC
From: justinpatrin@php.net Assigned: tony2001 (profile)
Status: Closed Package: oci8 (PECL)
PHP Version: 5.1.6 OS: Linux (Gentoo)
Private report: No CVE-ID: None
View Developer Edit
Welcome! If you don't have a Git account, you can't do anything here.
If you reported this bug, you can edit this bug over here.
Block user comment
Status: Assign to:
Package:
Bug Type:
Summary:
From: justinpatrin@php.net
New email:
PHP Version: OS:

 

 [2006-09-27 11:51 UTC] justinpatrin@php.net
Description:
------------
In writing a test case for a new issue with LOBs and MDB2 I ran into a crash that happens when I try to use one prepared query to insert multiple records with LOBs. In this case, thhe code is inserting 10 records, 2 lob columns to each record. I have also tested with only inserting 1 LOB column per record. It also crashes. The crash happens when I try to execute the statement the second time.

Reproduce code:
---------------
I do not have simple reproduce code at the moment. The ocde I am using to test this is the MDB2 test case testLOBRead. The code in CVS now will not cause this issues as I have altered it to prepare a new query for each insert. I will try to get a simpler script for you if you need it (and if I can get oracle back up....it's died on me).

Actual result:
--------------
#0  0xb64b13a7 in kpufGetRcvInfo () from /opt/oracle/product/10.1.0.3/lib/libclntsh.so.10.1
#1  0xb6724be8 in ttcacr () from /opt/oracle/product/10.1.0.3/lib/libclntsh.so.10.1
#2  0xb66b2bda in ttcdrv () from /opt/oracle/product/10.1.0.3/lib/libclntsh.so.10.1
#3  0xb65a6f1d in nioqwa () from /opt/oracle/product/10.1.0.3/lib/libclntsh.so.10.1
#4  0xb640b333 in upirtrc () from /opt/oracle/product/10.1.0.3/lib/libclntsh.so.10.1
#5  0xb6380fd2 in kpurcsc () from /opt/oracle/product/10.1.0.3/lib/libclntsh.so.10.1
#6  0xb6336617 in kpuexecv8 () from /opt/oracle/product/10.1.0.3/lib/libclntsh.so.10.1
#7  0xb63389a6 in kpuexec () from /opt/oracle/product/10.1.0.3/lib/libclntsh.so.10.1
#8  0xb640fe9e in OCIStmtExecute () from /opt/oracle/product/10.1.0.3/lib/libclntsh.so.10.1
#9  0xb6f38d07 in php_oci_statement_execute (statement=0x8c75d24, mode=0) at /tmp/pear/cache/oci8-1.2.2/oci8_statement.c:345
#10 0xb6f40bc5 in zif_oci_execute (ht=0, return_value=0x8c7435c, return_value_ptr=0x0, this_ptr=0x0, return_value_used=1)
    at /tmp/pear/cache/oci8-1.2.2/oci8_interface.c:1276
#11 0x0836d26e in execute ()
#12 0x0836cbcb in execute ()
#13 0x0836ce46 in execute ()
#14 0x0836cbcb in execute ()
#15 0x0836ce46 in execute ()
#16 0x0836cbcb in execute ()
#17 0x08340f5b in zend_call_function ()
#18 0x08341b60 in call_user_function_ex ()
#19 0x0828294f in zif_call_user_func ()
#20 0x0836d26e in execute ()
#21 0x0836cbcb in execute ()
#22 0x0836ce46 in execute ()
#23 0x0836cbcb in execute ()
#24 0x0836ce46 in execute ()
#25 0x0836cbcb in execute ()
#26 0x0836ce46 in execute ()
#27 0x0836cbcb in execute ()
#28 0x0836ce46 in execute ()
#29 0x0836cbcb in execute ()
#30 0x0836ce46 in execute ()
#31 0x0836cbcb in execute ()
#32 0x0834ef24 in zend_execute_scripts ()
#33 0x08308728 in php_execute_script ()
#34 0x00000000 in ?? ()


Patches

Pull Requests

History

AllCommentsChangesGit/SVN commitsRelated reports
 [2006-09-27 11:54 UTC] justinpatrin@php.net
(gdb) frame 11
#11 0x0836d26e in execute ()
(gdb) print (char *)(executor_globals.function_state_ptr->function)->common.function_name
$1 = 0xb6ec1508 "ociexecute"
(gdb) print (char *)executor_globals.active_op_array->function_name
$2 = 0x89c158c "_execute"
(gdb) print (char *)executor_globals.active_op_array->filename
$3 = 0x88d4134 "/home/papercrane/pear/MDB2/MDB2/Driver/oci8.php"
 [2006-09-27 13:18 UTC] tony2001 at phpclub dot net
See test case: ext/oci8/tests/oci_execute_segfault.phpt
The backtrace seems to be identical, so I believe the cause is the same too.
I don't know of any solution yet.
 [2006-09-27 13:40 UTC] justinpatrin@php.net
Yep, that seems to be the test case for this. Thanks for the quick response. I've altered our test case to re-prepare per-insert for now.
 [2007-02-13 18:36 UTC] tony2001 at phpclub dot net
JFYI: this is fixed in PHP 5.2.1.
 
PHP Copyright © 2001-2024 The PHP Group
All rights reserved.
Last updated: Thu Dec 12 15:01:32 2024 UTC