Aloha! After following the desription on the OpenRISC toolchain installation page I'm now trying to actually port and build code for the processor using the tools. Unfortuntely I'm bumping into some problems. These are quite probably my own fault. However, what would be nice is a toolchain installation test suite. That is, a few SW-projects (ranging from simple one file projects to multi-file ones that uses the installed libraries) that can be built using a script thereby verifying that the installation has completed correctly and it's possible to build code for the OpenRISC core. Given the search path(s) to the tools, a stand alone test suite should be able to do this. I can provide some (but not all, since it would be better if more than one person provided code, thereby maximizing code style and ways of stressing the installation) example code for this if it would be interesting. Also, I could whip up a suitable script that simply traverses any subdirectories and makes "make all". How about it? What's available today? Interesting? -- Med vänlig hälsning, Yours Joachim Strömbergson - Alltid i harmonisk svängning. VP, Research & Development ---------------------------------------------------------------------- InformAsic AB / Hugo Grauers gata 5B / SE-411 33 GÖTEBORG / Sweden Tel: +46 31 68 54 90 Fax: +46 31 68 54 91 Mobile: +46 733 75 97 02 E-mail: joachim.strombergson@informasic.com Home: www.informasic.com ----------------------------------------------------------------------

Aloha! Quoting Damjan Lampret lampret@opencores.org>:

Ah! My bad. Thanks. Works like a charm.

Ok, need to get the orp_soc too. How about adding a cvs-command on the web page for that one too, just like the other tool. While I'm busy making a fool of myself: Where is the documentation? ;-) The OpenRISC (and the new OpenCores) web page is very nice. But when I look at the nice section bar (The blue one with "Home :: Architecture :: OpenRISC 1200...") I don't see one section "Documentation". I gather that there are documentation available under the other sections. But for blind people like me, it might be a good idea to have a specific section where all documents are collected. No? -- Med vänlig hälsning, Yours Joachim Strömbergson - Alltid i harmonisk svängning. VP, Research & Development ---------------------------------------------------------------------- InformAsic AB / Hugo Grauers gata 5B / SE-411 33 GÖTEBORG / Sweden Tel: +46 31 68 54 90 Fax: +46 31 68 54 91 Mobile: +46 733 75 97 02 E-mail: joachim.strombergson@informasic.com Home: www.informasic.com ----------------------------------------------------------------------

* Damjan Lampret (lampret@opencores.org) wrote:

we could have documentation in the cvs and have a script that automaticaly
checks it into some directory for web engine to pick up. the README can be
a text that's converted on the fly to html. (i' using something similar)

regards,
p.

Joachim, you already have two sets of test cases. One in or1ksim/testbench. Go to or1ksim and do "make check" and it should build them and run them (on or1ksim). And there is another set in orp_soc if you go to orp_soc/sim/run and run run_sw (with parameter build to buiild them). It should build them in orp_soc/sw and run them on or1ksim, or if you run_rtl_regression it should run them on RTL of a simple reference system built around OR1200. regards, Damjan ----- Original Message ----- From: "Joachim Strombergson" Joachim.Strombergson@informasic.com> To: openrisc@opencores.org> Sent: Monday, February 16, 2004 4:25 PM Subject: [openrisc] Test suite for OpenRISC toolchain installation?
I'm

tools.

own fault.

is, a

that

verifying

code

able

the

could

makes

We really like to provide nearly real time support on a case by case basis.
Just to make us really busy. :-)

Yes the documentation would allow things to be much simpler and this is not
what we want. :-)

OK enough jokes. I will give you access to edit openrisc pages and then you
can help improve them, OK?

Now what is your opencores username?

regards,
Damjan

PS The main reason why web is behind is because when we update the web, and
work on toolchain and examples and RTL in the CVS, eventually web gets kind
of obsoleted and nobody never has time to update the web. So it is better
not to have information on the web, if you can have them in the cvs.
Conclusion: a link from the web to the cvs document is the best solution.
For example instead of having instructions on the web, it might be more
efficent to provide a link to a READMe file in the cvs. The only problem is
that then nobody reads the damn README file. Conclusion: I don't know what
to do. Any ideas?

Aloha! Quoting Damjan Lampret lampret@opencores.org>:

Sounds ok.

joachim. It awaits approval. -- Med vänlig hälsning, Yours Joachim Strömbergson - Alltid i harmonisk svängning. VP, Research & Development ---------------------------------------------------------------------- InformAsic AB / Hugo Grauers gata 5B / SE-411 33 GÖTEBORG / Sweden Tel: +46 31 68 54 90 Fax: +46 31 68 54 91 Mobile: +46 733 75 97 02 E-mail: joachim.strombergson@informasic.com Home: www.informasic.com ----------------------------------------------------------------------