[mesa-users] Problem with diffusion test in 4723
paxton at kitp.ucsb.edu
Fri Dec 21 13:08:05 EST 2012
On Dec 21, 2012, at 5:52 AM, João Faria wrote:
> < c_o16 dX 1289 1.2164748375287360E-16
> --- field 4 relative error 1.24e-02
> > c_o16 dX 1289 1.2316536679435330E-16
I didn't notice that (the error doesn't happen on my machine) --
the test is doing a relative comparison of results which
make sense for most cases but not for the vanishingly small quantities
in this case. We really don't care if the change in o16 is 1.22e-16 or 1.23e-16,
so we shouldn't be comparing those numbers in the test. I'll fix it and
make a new release.
As a work around until the fixed release, do this
cp tmp.txt test_output
That will trick the test into silence.
Then go back to the top level mesa directory and do ./install (no need to do ./clean this time).
More information about the Mesa-users