Bug 676179

Summary: JSS: slots not freed
Product: Red Hat Enterprise Linux 6 Reporter: Matthew Harmsen <mharmsen>
Component: jssAssignee: Andrew Wnuk <awnuk>
Status: CLOSED ERRATA QA Contact: Chandrasekar Kannan <ckannan>
Severity: unspecified Docs Contact:
Priority: unspecified    
Version: 6.1CC: awnuk, benl, cfu, dpal, jdennis, jgalipea, jmagne, mharmsen, rrelyea
Target Milestone: rc   
Target Release: ---   
Hardware: Unspecified   
OS: Unspecified   
Whiteboard:
Fixed In Version: jss-4.2.6-15.el6 Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: 676083 Environment:
Last Closed: 2011-05-19 13:43:41 UTC Type: ---
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Bug Depends On: 676083, 676180    
Bug Blocks:    

Description Matthew Harmsen 2011-02-09 00:25:44 UTC
+++ This bug was initially created as a clone of Bug #676083 +++

This was actually reported by Bob Relyea verbally, and I have confirmed the bug:

in JSS, there are places where PK11_GetInternalSlot() is called but never freed.
The most obvious ones are the ones that are called and used directly for comparison and not assigned to any variable (hence unfreeable).

The right way to do this is something like this:
PK11SlotInfo *aSlot = NULL;
...
...

aSlot = PK11_GetInternalSlot();
...
...

finish:
if (aSlot != NULL) {
    PK11_FreeSlot(aSlot);
}

What you don't want to do is something like this:
  if (slot != PK11_GetInternalSlot()) ...

if not fixed, other than memory leak, there are potential consequences such as
NSS may not be able to release the slot due to some internal counters.
Bob might be able to fill in on the detail.

--- Additional comment from rrelyea on 2011-02-08 14:31:58 EST ---

Most of the usages can be replaced by NSS functions that give better answers:

PK11_IsInternalKeySlot(slot) returns true if the slot is the default internal key slot.

PK11_IsInternal(slot) return true if the slot is any type of internal slot. Several of your checks should probably use this test as there can be more than one database slot these days.

The code I noticed this in is CryptoManager.c

bob

Comment 2 Dmitri Pal 2011-02-09 00:41:38 UTC
*** Bug 676180 has been marked as a duplicate of this bug. ***

Comment 4 Jenny Severance 2011-04-08 18:18:17 UTC
Can you please add steps to reproduce this bug. Thanks!

Comment 5 Chandrasekar Kannan 2011-04-20 23:31:44 UTC
I don't think its going to be easy for QE to verify this. I'm going to suggest a peer code review by another dev as a verification method for this.

Comment 6 Chandrasekar Kannan 2011-04-29 01:05:10 UTC
marking it VERIFIED with VERIFIED field set to SanityOnly. Layered components on top of this that are pki-ca and IPA are functioning normally through our regression tests.

Comment 7 errata-xmlrpc 2011-05-19 13:43:41 UTC
An advisory has been issued which should help the problem
described in this bug report. This report is therefore being
closed with a resolution of ERRATA. For more information
on therefore solution and/or where to find the updated files,
please follow the link below. You may reopen this bug report
if the solution does not work for you.

http://rhn.redhat.com/errata/RHBA-2011-0621.html