#7973 BLOC 8.2.0 (: libertas unknown event IDs and hang
Zarro Boogs per Child
bugtracker at laptop.org
Sat Aug 16 05:11:56 EDT 2008
#7973: libertas unknown event IDs and hang
------------------------+---------------------------------------------------
Reporter: cjb | Owner: dsaxena
Type: defect | Status: new
Priority: blocker | Milestone: 8.2.0 (was Update.2)
Component: kernel | Version: not specified
Resolution: | Keywords: blocks?:8.2.0
Next_action: diagnose | Verified: 0
Blockedby: | Blocking:
------------------------+---------------------------------------------------
Comment(by dsaxena):
From an email exchange with Javier and Ashish:
----
=== My original email and Javier's response ===
> As per dev.laptop.org #7973, we are seeing some issues with
> testing under heavy network traffic in a dense RF environment.
> From my analysis of the kernel logs, we are seeing the firmware
> timeout over and over to a network associate command and also
> receving completely bogus events.
>
> Would someone be able look at this bug and the FW sources and
> provide some input on:
>
> - What would trigger a DEFER response from the firmware when given
> a 0x50 command?
The BUSY result (0x4) is returned when a command is received from the
host while the previous one is still in progress.
In this case I would assume that the previous command is a scan. We
did have a problem with similar symptoms in ticket
http://dev.laptop.org/ticket/4470
> - What would trigger the firmware to send us bogus events?
I believe Ashish added those events precisely to troubleshoot #4470,
so he should be able to provide details.
=== Ashish's responses to above ===
The events were added to debug bugs similar to 4470/5194.
All these events are sent by the firmware, when it is taking too long to
process and finish
the current command execution. Event 0x30XY indicates that firmware is
currently
processing command 0xXY (e.g., 0x3050 indicates associate command). There
are other events
as well to get firmware state when something goes wrong, I don't have
source code access
now, I would respond with more information in the coming week.
----
--
Ticket URL: <http://dev.laptop.org/ticket/7973#comment:22>
One Laptop Per Child <http://laptop.org/>
OLPC bug tracking system
More information about the Bugs
mailing list