Re: Fw: NullPointerException in LinuxHost.PollWaitOnSigChld.<anonymous>.terminated

[Yong Zheng <zhengyzy at cn dot ibm dot com>]

It's strange. The new test TestX8664Regs runs ok in our X86-64 machine.
The following is the test result:

Running testX8664Regs(frysk.proc.TestX8664Regs) ...29700.29700:
argv[0]=/rtos/source/frysk/build_cvs/frysk-core/frysk/pkglibexecdir/funit-child
29700.29700: argv[1]=--wait=suspend
29700.29700: argv[2]=10
29700.29700: argv[3]=28766
29700.29700: argv[4]=/rtos/source/frysk/build_cvs/frysk-core/frysk/pkglibexecdir/funit-x8664-regs
29700.29700: starting 29700
29700.29700: new thread 29700.29700
29700.29700: notify 28766 with 10 (User defined signal 1) -- new thread 29700.29700
29700.29700: received signal 30 (Power failure)
29700.29700: exec
29700.29700: execing /rtos/source/frysk/build_cvs/frysk-core/frysk/pkglibexecdir/funit-x8664-regs argv[0]=/rtos/source/frysk/build_cvs/frysk-core/frysk/pkglibexecdir/funit-x8664-regs ...
PASS

PS: the frysk code is checked out today.

On Fri, 2006-08-04 at 17:17 +0800, Tim wrote:
> 
> Hi,
> In my new test TestX8664Regs I was getting a null pointer exception in
> the  above method. I've added an explicit test for the null pointer
> and
> ignoring it doesn't seem to break anything, but I'm wondering about
> the
> implications. Does the test not initialize the event loop properly? Is
> it expected that we could get events for tasks that we don't "know"
> about?
> 
> Thanks,
> Tim