On 16/12/14 04:37, Brian Davis wrote:
On December 14, 2014, Tristan wrote:
I'm fine with #3. In case of line length issue, you can use --!V87
Ok, I'll use the shorter !V87 for the library edits.
As of tonight, using my local patches to the libraries and complib.bat,
the 'gna' tests report success;
Great! Do not hesitate to post the patches.
> but I have some scripting infrastructure
problems still to sort out with the 'vests' portion of the testsuite.
Although it would be fine if vests was working on Windows, I wouldn't
spend to much time on it: they are mainly validation tests, and as long
as they are ok on linux that's fine.
OTOH, if gna tests are ok, the fix shouldn't be very complex.
Thanks for all that work,
Tristan.
-Brian
Posted by Tristan Gingold on December 14, 2014 - 07:34:
On 13/12/14 22:58, Brian Davis wrote:> Possibilities:
1) modify ghdlfilter to behave similarly to the existing sed
commands
2) add a new set of V08 metacomments to the libraries
3) add a new metacomment NOT_V87 (sic) to replace most of the
library
V93's ( less code duplication in the libraries then #2 )
#1 would require the fewest code changes, but seems least
consistent in
the metacomment behavior.
#3 seems to me most self-consistent, albeit also requiring changes
to
the current ./library/Makefile.inc
Yes, good remark. I'm fine with #3. In case of line length issue, you
can use --!V87
Tristan.
_______________________________________________
Ghdl-discuss mailing list
[email protected]
https://mail.gna.org/listinfo/ghdl-discuss
_______________________________________________
Ghdl-discuss mailing list
[email protected]
https://mail.gna.org/listinfo/ghdl-discuss
