#9895 NORM 1.5-sof: Should we inhibit suspend when we're connected to an ad-hoc network?

[Zarro Boogs per Child]

#9895: Should we inhibit suspend when we're connected to an ad-hoc network?
-------------------------------------------+--------------------------------
           Reporter:  cjb                  |       Owner:  cjb                              
               Type:  defect               |      Status:  new                              
           Priority:  normal               |   Milestone:  1.5-software-update              
          Component:  power manager (OHM)  |     Version:  Development build as of this date
         Resolution:                       |    Keywords:                                   
        Next_action:  communicate          |    Verified:  0                                
Deployment_affected:                       |   Blockedby:                                   
           Blocking:                       |  
-------------------------------------------+--------------------------------

Comment(by Quozl):

 == Test of Loss of Presumed Master by Power Fail ==

  * place units to encourage unit GOLF to be master; unit HOTEL 4m away
 beyond a steel wall, unit INDIGO 4m away beyond a steel wall 180 degrees;
 layout H -- G -- I,
  * began a ping from unit INDIGO to unit GOLF,
  * observe ping output on all units is normal,
  * ''at this point it is presumed that GOLF may be master, but without
 appropriate tools it is impossible to prove'',
  * abruptly remove power from unit GOLF,
  * observe ping from unit INDIGO to unit GOLF ceases, and begins to show
 ''Destination Host Unreachable'',
  * observe ping from unit HOTEL to unit INDIGO is uninterrupted and
 continues.

 == Test of Return after Reboot ==
  * unit GOLF was rebooted,
  * the ad-hoc network was shown in Neighbourhood View,
  * it was selected,
  * the icons for the other units appeared,
  * the new IP address was obtained,
  * ping from unit INDIGO to unit GOLF was re-established using new IP and
 worked normally.

 Conclusion: a unit that was used to create an ad-hoc wireless network that
 had been lost through power fail can be rebooted without loss of the
 network.

 Note: all these tests excluded suspend and resume and were done on B2
 units without WLAN power ECO.  A test should be done that measures the
 result of loss of the creator of the network by suspend.  Having the WLAN
 powered over suspend but without the host being available may change the
 behaviour of the WLAN.  Volunteers?  Need three units.

-- 
Ticket URL: <http://dev.laptop.org/ticket/9895#comment:4>
One Laptop Per Child <http://laptop.org/>
OLPC bug tracking system