#6589 BLOC Update.: xo stops responding to mesh path requests frames

[Zarro Boogs per Child]

#6589: xo stops responding to mesh path requests frames
-----------------------+----------------------------------------------------
  Reporter:  jcardona  |       Owner:  dwmw2   
      Type:  defect    |      Status:  new     
  Priority:  blocker   |   Milestone:  Update.1
 Component:  wireless  |     Version:          
Resolution:            |    Keywords:          
  Verified:  1         |    Blocking:          
 Blockedby:            |  
-----------------------+----------------------------------------------------

Comment(by ashish):

 Replying to [comment:10 carrano]:
 > In all instances of this bug I've seen, either this happens at start
 time or it doesn't happen at all. I could never cause it by stressing.
 Stressing, here includes but is not limited to:
 >
 > - Reducing route expiration to 1s and generating heavy traffic
 >
 > - Generating saturating traffic.
 >
 > - Running day long tests.
 >
 > Only a reboot changes the status of an XO from non-6598 to 6598 or vice
 versa.

 As per my understanding of the current driver, the probability of this
 failure during reboot should be very high. The current libertas driver
 does not seem to wait after firmware download, and therefore, any command
 issued just after firmware download will result failure, if firmware is
 not yet initialized.
 I dont see this problem with the following changes,
 http://dev.laptop.org/git?p=olpc-2.6;a=blob;f=drivers/net/wireless/libertas/if_usb.c;h=78f20baa780a4b302f4f5abb1ba9d507f820d77c;hb=stable


 {{{
 if_usb_probe
 /* Upload firmware */
 224         if (if_usb_prog_firmware(cardp)) {
 225                 lbs_deb_usbd(&udev->dev, "FW upload failed\n");
 226                 goto err_prog_firmware;
 227         }
 228
 229         if (!(priv = lbs_add_card(cardp, &udev->dev)))
 230                 goto err_prog_firmware;
 231
 232         cardp->priv = priv;
             msleep_interruptible(200); /*** delay to make sure firmware is
 ready */         cardp->priv->fw_ready = 1;
 234
 235         priv->hw_host_to_card = if_usb_host_to_card;
 }}}

 We will add firmware ready event, which can be used by the driver, in the
 future wireless firmware release to make this more robust.

-- 
Ticket URL: <http://dev.laptop.org/ticket/6589#comment:13>
One Laptop Per Child <http://laptop.org/>
OLPC bug tracking system