Issue Details (XML | Word | Printable)

Key: CORE-2956
Type: Bug Bug
Status: Closed Closed
Resolution: Fixed
Priority: Major Major
Assignee: Vlad Khorsun
Reporter: Alexander Peshkov
Votes: 0
Watchers: 0
Operations

If you were logged in you would be able to see more operations.
Firebird Core

Problems with requests of procedure's parameters

Created: 02/Apr/10 03:34 PM   Updated: 04/Feb/11 12:01 PM
Component/s: Engine
Affects Version/s: None
Fix Version/s: 2.5 RC3, 2.1.4, 3.0 Alpha 1

Time Tracking:
Not Specified

Planning Status: Unspecified


 Description  « Hide
In parse_param_blr() function in jrd/met.epp pointers to requests of all parameters (one after one) are placed into same place - procedure->prc_request. Later request of procedure itself is placed to the same place. I see the following problems with it:

1. Looks like that requests will never be released.
2. In case of error in parse_procedure_blr() prc_request is explicitly released and NULLed, with parse_param_blr() - not.
3. That request is kept is prc_request field in case of any failure compiling procedure. Later it can be attempted to be released when parse_procedure_blr() fails. And looks like this can be a cause of segfault:

Program terminated with signal 11, Segmentation fault.
#0  LCK_release (tdbb=0x7fb1770490d0, lock=0x7fb16d5d18f0) at ../src/jrd/lck.cpp:1401
   in ../src/jrd/lck.cpp

#0  LCK_release (tdbb=0x7fb1770490d0, lock=0x7fb16d5d18f0) at ../src/jrd/lck.cpp:1401
#1  0x00007fb179611d29 in CMP_release (tdbb=0x7fb1770490d0, request=0x7fb16d572010) at ../src/jrd/cmp.cpp:2440
#2  0x00007fb1795adb96 in MET_procedure (tdbb=0x7fb1770490d0, id=<value optimized out>, noscan=false, flags=8)
   at ../temp/std/jrd/met.cpp:5001
#3  0x00007fb1795ae730 in MET_lookup_procedure (tdbb=0x7fb1770490d0, name=<value optimized out>, noscan=false)
   at ../temp/std/jrd/met.cpp:3972


 All   Comments   Work Log   Change History   Version Control   Subversion Commits      Sort Order: Ascending order - Click to sort in descending order
Dmitry Yemanov added a comment - 02/Apr/10 04:12 PM
IIRC, I avoided this crappy code for PSQL functions, see how prm_default_value is handled in Function.epp.