This is the mail archive of the gsl-discuss@sourceware.cygnus.com mailing list for the GSL project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: tru64 failures


On Sun, 14 May 2000, Steve ROBBINS wrote:

> The environment is:
> 
> Did a basic "./configure; make; make check", so no configure options.
> NO GSL_ variables set.
> 
>     stever@jeff{gsl}uname -a
>     OSF1 jeff.cs.mcgill.ca V4.0 1229 alpha alpha
>     stever@jeff{gsl}gcc --version
>     egcs-2.91.66
      ^^^^^^^^^^^^

This seems to be the culprit.  I poked around a bit and finally discovered
that by omitting '-O2' from CFLAGS (in the siman directory) I got the
test program to work.

I'm going to install a newer version of gcc and see if the bug goes away.

Thanks to Brian for the `extra debugging' patch --- it was helpful.

-smr

 
> 
> And here is what 'make check' reports:
> 
> ...
> make[2]: Entering directory `/export/cgm/u3/stever/src/gsl/siman'
> FAIL: f(x)= exp(-(x-1)^2) sin(8x), x0=-10 (-10 observed vs 1.3631299945531519 expected)
> FAIL: f(x)= exp(-(x-1)^2) sin(8x), x0=10 (10 observed vs 1.3631299945531519 expected)
> FAIL: f(x)= exp(-(x-1)^2) sin(8x), x0=0.6 (0.59999999999999432 observed vs 1.3631299945531519 expected)
> FAIL: f(x)= exp(-(x-1)^2) sin(8x), x0=0.5 (0.5 observed vs 1.3631299945531519 expected)
> FAIL: f(x)= exp(-(x-1)^2) sin(8x), x0=0.4 (0.39999999999999386 observed vs 1.3631299945531519 expected)
> FAIL: test



Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]