http://defect.opensolaris.org/bz/show_bug.cgi?id=12479
Summary: nwamd state machine can miss events
Classification: Development
Product: nwam
Version: RC5
Platform: ANY/Generic
OS/Version: All
Status: NEW
Severity: major
Priority: P4
Component: ON daemon
AssignedTo: nwam-dev at opensolaris.org
ReportedBy: renee.danson at sun.com
QAContact: nwam-dev at opensolaris.org
--- Comment #0 from Renee Danson Sommerfeld <renee.danson at sun.com>
2009-11-05 01:21:22 UTC ---
Created an attachment (id=2959)
nwamd log
While trying to isolate some weird behavior I've observed with the nge hardware
on my desktop, I discovered a problem in nwamd's event/state maching handling.
If the hardware link goes down and then comes back up fairly quickly, you can
end up with this sequence of events:
* get link_down notification, enqueue event (1)
* get link_up notification, enqueue event (2)
* dequeue link_down event (1), handle it, enqueue state change event (3)
* dequeue link_up event (2), ignore it because the state is online
* dequeue state change event (3), mark the state offline
At this point, there are no more relevant events on the queue; and nwamd thinks
the link is down, while the link is in fact up.
This very problem was discussed extensively during code review of the fix for
11103, about six weeks ago.
I've attached an edited logfile; no messages were removed, but some of the
standard info was deleted from each line to make the lines shorter. I have the
original if any other details are needed.
--
Configure bugmail: http://defect.opensolaris.org/bz/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the QA contact for the bug.
You are the assignee for the bug.
