#9707 BLOC 1.5-har: occasional oops in sdio_remove_func() in suspend

[Zarro Boogs per Child]

#9707: occasional oops in sdio_remove_func() in suspend
--------------------------------+-------------------------------------------
           Reporter:  wad       |       Owner:  dsaxena       
               Type:  defect    |      Status:  new           
           Priority:  blocker   |   Milestone:  1.5-hardware-C
          Component:  kernel    |     Version:  1.5-B2        
         Resolution:            |    Keywords:  q3a16 os42 b2 
        Next_action:  diagnose  |    Verified:  0             
Deployment_affected:            |   Blockedby:                
           Blocking:            |  
--------------------------------+-------------------------------------------

Comment(by dsd):

 I'd say the problem we should really be looking into is this one:

 {{{
 [ 2109.784459] mmc2: error -110 during resume (card was removed?)
 }}}

 This happens during resume a long time before the crash (you can see in
 p1_crash.log that it occurs during resume, then system suspends, then it
 resumes with the same error, then it suspends, and at this point the crash
 occurs). I presume this message occurs on every resume after a certain
 point. It certainly doesn't occur during a normal suspend/resume (tested
 twice on my own B2).

 In terms of the crash itself, I'm still not entirely sure why we are
 hitting mmc_rescan() in the suspend path. However, Matt Fleming on the
 linux-mmc list pointed out a bug in the error path which we are hitting
 which would explain the crash.

 The other question is then why mmc_rescan() is failing, specifically when
 initializing a func on the wifi card.

 I added some debug messages to all of the above code paths as well as a
 crash-fix based on Matt's observation, commit
 9e0570c3844428128791a2363787af4022cc37db. Please test with new kernel and
 post new logs. Note that it might not crash any more, the point of
 interest is simply when the "mmc2: error -110 during resume (card was
 removed?)" messages start appearing.

-- 
Ticket URL: <http://dev.laptop.org/ticket/9707#comment:19>
One Laptop Per Child <http://laptop.org/>
OLPC bug tracking system