http://defect.opensolaris.org/bz/show_bug.cgi?id=11685
Renee Danson Sommerfeld <renee.danson at sun.com> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |renee.danson at sun.com
--- Comment #4 from Renee Danson Sommerfeld <renee.danson at sun.com>
2009-10-02 19:16:22 UTC ---
(In reply to comment #3)
> Changes in the last couple of weeks in nwamd's startup process include locking
> and unlocking the active_ncp_mutex. However, dlmgmtd's startup has changed
> quite bit due to the iptun's putback. I wonder if its related ....
>
> Other things I found:
>
> We check if a link exists or not for NCUs in the Automatic NCP. We use
> dladm_name2info() and check if the DLADM_OPT_ACTIVE flag is returned.
> If a link exists in the system, then DLADM_OPT_ACTIVE and DLADM_OPT_PERSIST
> flags are returned.
> If a link is mentioned in /etc/dladm/datalink.conf but does not exist in the
> system, then DLADM_OPT_PERSIST is returned.
>
> At boot time, dladm_name2info() returns DLADM_OPT_PERSIST for all interfaces
> in
> /etc/dladm/datalink.conf.
So I think the key questions for us are, at what point is DLADM_OPT_ACTIVE set,
and is there a way for us to know that the initial scan is complete?
Perhaps a better direction, though, would be if there were events associated
with the setting of that flag. We wouldn't need to know if it was "done"; we
could just listen for the events. We already have code that updates the
Automatic NCP based on EC_DEV_ADD and EC_DEV_REMOVE events; either re-use of
those events or adding other events for us to listen for would work very nicely
for nwamd, and would make it far less susceptible to boot-timing issues.
--
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.
