PS: I fixed two problems, on a dual-core AMD64 running F-8: under JNI
there are ~9 failures, while under CNI ~8 -> i.e. "bug compatible".
really cu in a week :-)
Andrew Cagney wrote:
> FYI,
>
> I've just completed my first full run of frysk-core's test-suite under
> JNI; and have gone through and filed bugs for each class of failure.
> Including the failures I also see under CNI (~6), I came up with 14
> unique cases (while there's a small number of problems they are
> affecting ~60 of the tests).
>
> More useful, I think, is this:
>
> $ make jni
> $ ./JniRunner <my-test>
>
> as away to quickly turn around your work-in-progress code. Of course,
> a full TestRunner run is still expected.
>
> cu all a bit over a week (perhaps I'll come back to find some of the
> bugs fixed :-)
>
> Andrew
>