[ODE] linux amd64 compile error - 0.6-rc1

[Bram Stolk]

Chris M. wrote:
> Whoops, that was indeed the problem.  I was able to run every test 
> without crashing, and test_ode had only one test fail.  Im sure yall are 
> glad that it was a problem on my side ;)

As prof Farnsworth used to say: Good news everyone... :-)

Jason, can you apply the ASM fix to SVN? (timer.cpp)
The single '%' rbx arguments in the AMD64 case should be doubled '%%'.

I would do it myself if I wasn't hesitant about choosing the correct
branching.

What is the procedure now? You patch on the 0.6 branch? But how and
when do the patches go in the trunk?

thx,

   Bram

PS to jason: Did you get a chance to remove the VALIDVEC3 asserts from
test_basket as well? They induce the same isnan issue.