Right, that's a typo and will break the build if modify udebug to 1 in the test case. Will fix that. As to the error threshold for those math function testing cases, I found there are also some other test cases set too strict error threshold. Will work with testing team to check all of them and make them comply with the spec. Thanks for pointing this out.
On Sat, Nov 15, 2014 at 6:48 AM, Rebecca N. Palmer <rebecca_pal...@zoho.com> wrote: > ...due to a stray . at utests/builtin_pow.cpp:79:112. > > Also, the strict-conformance spec for pow and tgamma is 16ulp, not 1ulp. > > _______________________________________________ > Beignet mailing list > Beignet@lists.freedesktop.org > http://lists.freedesktop.org/mailman/listinfo/beignet _______________________________________________ Beignet mailing list Beignet@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/beignet