#12285 BLOC 13.1.0: XO-4 runin memtest failure due to OOM kill

[Zarro Boogs per Child]

#12285: XO-4 runin memtest failure due to OOM kill
---------------------------------+------------------------------------------
           Reporter:  greenfeld  |       Owner:  Quozl                            
               Type:  defect     |      Status:  new                              
           Priority:  blocker    |   Milestone:  13.1.0                           
          Component:  kernel     |     Version:  Development build as of this date
         Resolution:             |    Keywords:                                   
        Next_action:  diagnose   |    Verified:  0                                
Deployment_affected:             |   Blockedby:                                   
           Blocking:             |  
---------------------------------+------------------------------------------

Comment(by dsd):

 I agree that this does suggest a memory leak that would still hit if we
 ran runin for long enough. If it were not a leak, the OOMs we previously
 saw would happen early during runin, rather than after several hours.

 Martin has identified this as an XO-4 production blocker if it affects
 runin. I guess with the tweaked memory usage value this would not be hit
 in the factory, but I guess we should spare some time for at least some
 diagnosis before considering removing blocker status.

-- 
Ticket URL: <http://dev.laptop.org/ticket/12285#comment:7>
One Laptop Per Child <http://laptop.org/>
OLPC bug tracking system