[OpenBIOS] Update: Tokenizer Version 1.00.01 is on its way!

David L Paktor dlpaktor at us.ibm.com
Sat Sep 9 01:01:38 CEST 2006






Stefan Reinauer <stepan at coresystems.de> wrote on Fri Aug 18 17:43:44 CEST
2006

> * Stefan Reinauer <stepan at coresystems.de> [060818 17:06]:
> . . . . .
>
> For a quick test I ran a coverage test with our new and shiny test suite
> and here's the results:
>
> Code covered:  95.9 %
>
> That is pretty damn good for a first run. Even if it was a safety
> critical applications.
>
> The whole coverage report can be checked here:
> http://openbios.org/~stepan/toke/coverage/fcode-utils/toke/index.html
>
> As you can clearly see most of the files only have their panic code
> uncovered. Which is pretty normal. Only whats wrong with strsubvocab.c
> ;-)
>
> David, thank you very much for this high quality contribution.
>
> Stefan
>
> --
> coresystems GmbH • Brahmsstr. 16 • D-79104 Freiburg i. Br.
>       Tel.: +49 761 7668825 • Fax: +49 761 7664613
> Email: info at coresystems.dehttp://www.coresystems.de/


Stefan Reinauer <stepan at coresystems.de> wrote on 09/05/2006 12:04:09 PM:

> Dear David,
>
> thank you for the conference call and the nice conversation.
>
> Please go ahead and send the next round of toke files with a
> bumped version number, so that people can see the progress being made.
>
> I will merge the changes into the new tree.
>
> Stefan


I have completed the changes we discussed and will be sending them to you
separately.

The coverage report identified several areas of "dead code" (most notably
in  strsubvocab.c ) as well as a few cases that weren't covered in the Test
Suite.

I changed the Test Suite to fill those gaps (there are now 167 tests, as
counted by the AutoExec script) and removed the dead routines (except for
one I believe should remain for possible future use -- read the top of the
Data Stack without popping it off).  I also fine-tuned some error detection
in "User-defined command-line compilation-control symbols" (when attempting
to access a defined value, distinguish between the case where no value is
assigned to the symbol versus where the symbol is altogether undefined; and
identify multiple definitions of the same symbol).

And, as we spoke, I bumped the version number to 1.0.01

I then ran my own coverage test; it showed a satisfactory improvement.  I
will not be sending you my results; your independent confirmation would be
preferable.

One other thing I'd like to request.  I think the Coverage Report is a Very
Good Thing:  it deserves to be displayed prominently, not just mentioned in
an item in the Mail Archives.  When you complete your coverage test, could
you open up an additional section --  like the sections for  "Unit-Test
Suite" and "Unit-Test Suite Logs" -- for it, at the end of the "OpenBIOS -
FCode suite" page?

Thanks.

-----

David L. Paktor                  System Firmware Developer
System and Technology Group      Global Firmware Division
dlpaktor at us.ibm.com              David L Paktor/Almaden/IBM at IBMUS

18880 Homestead Rd.              Building 9945
Cupertino CA 95014               Room 1026
408-342-6110                     T/L 560-6110

"The Bug Stops Here"
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.openbios.org/pipermail/openbios/attachments/20060908/8b0b8cd2/attachment.html 


More information about the OpenBIOS mailing list