On Aug 17, 2006, at 2:19 PM, Ted Herman wrote:
This is strange, why would some stray bug just hit rxbufptr? So, I
tried moving rxbufptr and its fence around, duplicating rxbufptr,
changing all usages of the idiom "rxbufptr->" to something else,
and quite a few other strange and desperate hacks. Yet that bug
outsmarted me every time, seemingly omniscient about the location
of rxbufptr (recall, each test cost about half an hour).
Sometimes I would work in the other direction. Removing, module
by module in my application, send and receive operations. I
discovered
that, at a certain point, the problem would not arise. However, this
was not particularly helpful, since running everything just up to the
actual send() could execute without triggering deafness in the
application.
If you use the objdump or nm tool, you can see the memory location of
each variable. This can let you see what variables are close to
rxbufptr and therefore guess at the cause.
Phil
_______________________________________________
Tinyos-help mailing list
[email protected]
https://mail.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help
