gcl-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Gcl-devel] rtest8.mac test failure


From: Raymond Toy
Subject: Re: [Gcl-devel] rtest8.mac test failure
Date: Tue, 01 Jul 2014 09:06:07 -0700
User-agent: Gnus/5.101 (Gnus v5.10.10) XEmacs/21.5-b34 (linux)

>>>>> "Camm" == Camm Maguire <address@hidden> writes:

    Camm> Greetings!  When compiling some functions that were previously
    Camm> interpreted under gcl, I have one small test failure on 32bit intel
    Camm> only:

    Camm> ev (e7, alfa=2, vita=1); (of rtest8.mac)

    Camm> returns 

    Camm> [0.052961027786557, 4.8572257327350623E-17, 50, 0]

    Camm> insead of

    Camm> [0.052961027786557, 5.551115123125785E-17, 50, 0]

    Camm> This is solely due to the loss of one bit of precision in res24 of
    Camm> slatec::dqc25s when the optimized compile is done in 64 bits instead 
of
    Camm> the 80 that are specially available on the x87 fpu:

[snip]

    Camm> I don't really think this is a GCL bug, yet I've constructed the 
debian
    Camm> package build to fail on any test failure.  Is there a way I can set 
the
    Camm> floating point tolerance higher here to let this result through?

Thanks for the detailed report. I think this particular test will need
to be restructured a bit to use my_float_approx_equal (or equivalent)
to check that the first return value is close enough. Maybe also that
third return value is 50 and the fourth is 0.

Ray





reply via email to

[Prev in Thread] Current Thread [Next in Thread]