[vtk-developers] Dashboard open again

[Bill Lorensen]

Ken,
Progress on our GE machines...

1) On manifold, for some reason, I removed all the lib*.a's in the vtk bin and rebuilt. The tests now pass. I do not do a clean on these platforms each night because the builds take too long. I have seen this problem before. I suspect that we are not using ar properly. I have added an rm bin/*.a to the nightly for this platform.

2) The synchronizedtempleates tests crashing. These are real bugs in ST. You should be able to repeat them on any platform by setting the number of threads to 6 or higher. As I mentioned before, caleb has 10 processers. I was able to get the test to crash on the 4 processor kulu when I set the number of threads to 6 or higher. You'll see that they are crashing in some CopyData call.

Bill


At 12:57 PM 5/10/02 -0400, Ken Martin wrote:
>Well, it isn't green but it is pretty close. The Mac isn't ready for
>prime time yet so I'm not worrying about it. DEC has some valid issues
>and there could be an issue or two on the GE machines, but I think CVS
>has been locked up for long enough. Purify is clean and most systems
>are super green. There are a number of 64 bit problems and we will be
>adding a new machine to test those.
>
>Policy issue...
>
>I'd like for any new platforms to be added as Experimental until they
>are green, then they can be moved to Nightly. That way we know that
>all Nightly systems should be supported and green since at one point
>they were green. (unless we someday decide to deprecate an OS/compiler
>etc) If it has never been green then it really isn't supported by VTK
>yet.
>
>Thanks
>Ken
>
>
>--------------------------
>Ken Martin, PhD
>Kitware Inc.
>518 371-3971 x101
>469 Clifton Corporate Pkwy.
>Clifton Park, NY 12065
>
>_______________________________________________
>vtk-developers mailing list
>vtk-developers at public.kitware.com
>http://public.kitware.com/mailman/listinfo/vtk-developers