On Mon, Feb 16, 2009 at 10:21 AM, Oliver Neukum [off-list ref] wrote:

quoted

tests seemed to verify this claim... as sending LED commands
GFP_KERNEL while holding the mutex resulted in BUGs (scheduling while
atomic) in dmesg. Switching those GFP_KERNELs to GFP_ATOMICs
eliminated that particular BUG.

Please post that BUG.

I switched back to GFP_KERNEL on the functions in question.

After further testing, I have determined that the BUG only occurs on
my Ubuntu (2.6.27) system, and only when using the rumble function. I
am unable to reproduce the BUG on my Arch (2.6.28) system, so it is
either an artifact of some patch to this kernel, or it got fixed
somewhere between 2.6.27 and 2.6.28.

Based on this analysis, and (more importantly) an inability to
reproduce the behavior on a more vanilla kernel, I have removed the
comments about it from the TODO section of the driver. It isn't a true
kernel bug at least with a more recent stable kernel.

Mike
-- 
Mike Murphy
Ph.D. Candidate and NSF Graduate Research Fellow
Clemson University School of Computing
120 McAdams Hall
Clemson, SC 29634-0974 USA
Tel: +1 864.656.2838   Fax: +1 864.656.0145
http://cirg.cs.clemson.edu/~mamurph