On 11/22/06, Geir Magnusson Jr. <[EMAIL PROTECTED]> wrote:
Elena Semukhina wrote:
> On 11/22/06, Geir Magnusson Jr. <[EMAIL PROTECTED]> wrote:
>> Right - my point is that excluding our thread tests is a big red flag.
>> I know that you and others have been working very hard to get this
fixed.
>
>
> I managed to run kernel tests iteratively for 40 times on linux in jet,
> opt,
> int (debug build, "once" mode) and have not seen a failure! ThreadTest
was
> included. Tried the same on Windows 2003 and saw 1 failure of ThreadTest
in
> jet mode. It failed on interrupting a waiting thread.
Meaning - when excluding ClassGenericsTest4, you have no problems?
If that's the case, then what I think we should do is recognize that
Windows XP and Windows 2003 are different operating systems and try to
distinguish between them for purposes of exclusion?
I don't have a win2003 box so I can't figure this out. What would be
best is a non-destructive enhancement to classlibs properties.xml to add
a "flavor" property, like
${harmony.os_variety}
that we can use for distinguishing between WinXP and Win2003, Ubuntu and
RHEL, etc....
> So it is worth to exclude ThreadTest on Windows. Should we do this on
> linux?
Not if we aren't having any problems on Linux. Our goal is to have as
many tests as possible running at all times. I think that as long as we
don't have problems, we keep it.
> I'll continue my experiments tomorrow.
Excellent. Thanks
Today I ran kernel tests on linux jet, opt 100 times with no failures but
saw a failure on linux int (ThreadTest.test_yield()). The same test failed
once on Windows/int. I'll play with it a little and then summarize my
thoughts.
geir
>
> Thanks,
> Elena
>
> geir
>>
>
>
>
--
Thanks,
Elena
