[ltp] Re: Re: Re: EC Hanging with latest tp-smapi (0.29)

Henrique de Moraes Holschuh linux-thinkpad@linux-thinkpad.org
Wed, 30 Aug 2006 20:10:19 -0300


On Wed, 30 Aug 2006, TNKS wrote:
> This did what I thought it might do; I got a flood of error messages because
> I was accessing 0x00b.  Then the EC hung, and I got a different flood of

Then regardless of 0x0a, 0x0b really is a problem.

> I didn't do the loop-inverting suggestion you had yet because I thought
> these results were pretty interesting, and so I thought you might
> re-evaluate what you want to do next.  

You're correct :-)

> basically a normal kern.log.  Outside of the timestamps, though, I'm just
> not sure how much you can infer from about the state of the EC.  The

I can see exactly which step of the protocol failed first, looking at the
thinkpad_ec source, and I can think about what would happen at the next
access, etc.

> > I bet if we profiled the EC calls we would know, but _that_ ain't a
> > trivial thing to ask you to do.
> 
> If the experimentation we're doing isn't cutting it, I'm willing to learn
> what it takes to profile the EC.  I'm interested, and I should be competent
> enough to figure out what I have to.  

Profiling would tell us how much time each EC call needs before it returns
an OK or ERROR.  But just turning on profiling wouldn't do (I thought it
would, but some reading told me it would not be fine-grained enough).  So
either I, you or Shem would have to come up with a patch to do thinkpad_ec
time accounting.

I am not so sure it would help debug the issue anymore, anyway.

-- 
  "One disk to rule them all, One disk to find them. One disk to bring
  them all and in the darkness grind them. In the Land of Redmond
  where the shadows lie." -- The Silicon Valley Tarot
  Henrique Holschuh