RTEMS4.7 and its tool chain (Re: powerpc mvme5500 clock off by factor of 150)
feng1 at bnl.gov
Tue Mar 8 17:16:26 UTC 2005
Peter Dufault wrote:
> The task index is the same as the number that is printed out as "TAx"
> by "ticker", so it is 1, 2, and 3. The program eventually runs, as a
> matter of fact I had to go out for a few hours this morning, and I can
> see that "30 seconds" have indeed finally gone by for the board.
> The source code I'm using is what is in ticker without changes, but I'm
> using 4.7, so you need to install the newer tool chain to reproduce the
> RTEMS I'm using.
I do not see this related to the origianl subject "powerpc mvme5500
clock off by factor of 150" at all. I built the source of
c/src/tests/samples/ticker. It works as expected.
The colck on the MVME5500 BSP always works fine.
Perhaps there is problem with your toolcahin. I will send you
my Makefile and binary code offline so that you can post
the questions under a more appropriate subject. The only change
I did in the BSP since the RTEMS4.6.0 release post on the WEB,
is to enable the PMC interrupt by default and to
change the PCI API.
More information about the users