BIND 10 #1289: nested dlopen through python failure

[BIND 10 Development]

#1289: nested dlopen through python failure
-------------------------------------+-------------------------------------
                   Reporter:  jelte  |                 Owner:  jelte
                       Type:         |                Status:  reviewing
  defect                             |             Milestone:
                   Priority:         |  Sprint-20111011
  blocker                            |            Resolution:
                  Component:  data   |             Sensitive:  0
  source                             |           Sub-Project:  DNS
                   Keywords:         |  Estimated Difficulty:  0
            Defect Severity:  High   |           Total Hours:  0
Feature Depending on Ticket:         |
        Add Hours to Ticket:  0      |
                  Internal?:  0      |
-------------------------------------+-------------------------------------
Changes (by vorner):

 * owner:  vorner => jelte


Comment:

 Hello

 What I read about the global flag, it manipulates the global symbol table.
 This indeed might help with the exceptions, but I worry little bit about
 if it solves the whole problem.

 What if two libraries are loaded, each one providing it's copy of the
 exception. The second one loaded probably overwrites the symbol table
 after the first one. Then the first one would still not work.

 But as it breaks our build bots and generally annoys, let me propose this.
 Put the fix into master, decrease the priority and leave it open (and put
 it into next sprint, maybe?). Then we would need to implement a test where
 we would load two libraries at once and see if they both work well. Do you
 think it would be OK? Or is pushing the fix first too much against the
 policies?

 Thanks

-- 
Ticket URL: <http://bind10.isc.org/ticket/1289#comment:3>
BIND 10 Development <http://bind10.isc.org>
BIND 10 Development