bit scripts

[Angelo Fraietta]

Joel Sherrill wrote:

>I hate to ask this at this point but what version of RTEMS
>did you use to successfully run these tests before on this 
>same hardware?  I do not believe anything is wrong with your
>tools or installation.  I am beginning to wonder if some
>other change between the last one you tried and this
>one are doing it.  Since the 
>
After testing the prerelease and failing on the linux machine, I decided 
to build snapshot
rtems-ss-20010424.tgz
The reason for this is that I had the RPMs from about this date that I 
knew worked.
I completely removed the RPMs (directory an all)
and built the following with the following tools from the bit scripts

binutils-2.11-rtems-20010409.diff
binutils-2.11.2-rtems-20010622.diff
binutils-2.11.2.tar.gz
binutils-2.11.tar.bz2
c_build_scripts-4.5.0.tgz
gcc-2.95.2-rtems-20010221.diff
gcc-2.95.2.tar.gz
newlib-1.9.0-rtems-20010221.diff
newlib-1.9.0.tar.gz

The samples ran was considering going a half split or maybe to the 
snapshot with the latest gcc2.95.
The bit script, however, didn't work for me on OSX so I left it at that. 
Eric has been working with me on getting the latest working properly on 
OSX (I am building the tools again now)

>
>You mention Linux and MacOS X.  Are the test results the same
>when compiled from both environments?
>
The tests I did with OSX were for the previous snapshot, 
rtems-ss-20030128.tar.bz2.

I posted these results on 28 Feb. The two were identical in that the OSX 
ones did not give the exception.
I only tested the pc386dx but tested it on the DIMMPC (as it has no fpu)
Here are the results (copied from that message)

    It appears to be very broken
    I ran some of the samples with the following results:

       ticker ***********************
       - I get a message on the console that clock tick test is starting
    and then the machine reboots
         unlimited *************************
       I get the message
       TEST1: auto-extend disabled
       TEST1: creating task 'aa00', number = 0, ide = 0a010002, starting
         The application seems frozen now with a cursor blinking at one
    space after the comma on the last line

       paranoia *************************
       passes test

    Could someone send me a copy of built examples of ticker and
    unlimited for a pc386dx build so I can eliminate or determine the
    problem is at my end?

    Thanks  

Prior to this, I tested my actual application with these results (again 
copied)

    My build is broken in that the application can no longer start.
    Is suspect that the problem may be in the static variable
    initialisers, however, I am unsure.

    My device drivers are getting initialised (I can see this on the
    console)
    however I get an exception error

    exception 6 caught

    This occurs before a printf statement I placed at the start of r Init

    I also commented out all the code in the Init function and still got
    the same error.

    I removed nearly all of my code and the application initialised

    Init is in a cpp module which is being executed


>
>See my comments below.<snip>
>

>
>pc386 assumes the presence of an FPU.
>pc386dx assumes no FPU.
>
Yes. It is interesting that paranoia on the snapshot ran successfully 
built on the OSX.

>>
>
>Which exception is it?
>
Exception 6 - same as what occurred in my application

>
>
>
>>--
>>Angelo Fraietta
>>
>>PO Box 859
>>Hamilton NSW 2303
>>
>>Home Page
>>
>>http://www.users.bigpond.com/angelo_f/
>>
>>There are those who seek knowledge for the sake of knowledge - that is CURIOSITY
>>There are those who seek knowledge to be known by others - that is VANITY
>>There are those who seek knowledge in order to serve - that is LOVE
>>    Bernard of Clairvaux (1090 - 1153)
>>
>

-- 
Angelo Fraietta

PO Box 859
Hamilton NSW 2303

Home Page


http://www.users.bigpond.com/angelo_f/

There are those who seek knowledge for the sake of knowledge - that is CURIOSITY
There are those who seek knowledge to be known by others - that is VANITY
There are those who seek knowledge in order to serve - that is LOVE
    Bernard of Clairvaux (1090 - 1153)


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.rtems.org/pipermail/users/attachments/20030307/c809862f/attachment-0001.html>