Brian Drummond wrote:
If you can't easily run the supplied shell script,
then writing a Windows batch file would be a HUGE task
The existing shell scripts should work OK for Windows testing.
I had tried quickly running the test suite, using an MSYS shell, when I
built the 0.31 release.
The gna tests died with ghdl exceptions on some tests, and failed a few
others.
But I hadn't looked into the failures yet to sort out whether those
errors were real, or somehow setup-inflicted.
"hg update ghdl-0.31" which will exclude newer tests that 0.31 is
guaranteed to fail
OK, that could explain some of the problems with my earlier attempt
(which was using the development tip).
-----------------------
Running Vests:
"
" Vests tests successful
"
-----------------------
Running gna tests:
After that hg update, running just the gna tests with the "-k" (keep
going) flag gives me:
GNA test failed ( bug16782 bug20312 bug21487 bug21500)
The last three failures were caused by mostly-empty directories left
behind by the hg update revision rollback ( work-obj93.cf in those
directories, from previous run, prevented deletion ? ).
After deleting those three ghosts-of-revisions-future directories, I
have only one crash, on bug16782:
"
" GNA dir bug16782:
" analyze bug.vhd
" elaborate and simulate (failure expected) bug
"
" This application has requested the Runtime to terminate it in an
unusual way.
" Please contact the application's support team for more information.
"
In bug16782's bug.vhd, changing
constant DATAPATH :natural := 16;
to a value of 8 or less makes the crash go away,
suggesting the crash results from the following indexing (albeit in a
case testing width):
signal tmp :bit_vector(31 downto 0);
...
when 3=> tmp(4*DATAPATH-1 downto 3*DATAPATH)<=a;
-Brian
_______________________________________________
Ghdl-discuss mailing list
[email protected]
https://mail.gna.org/listinfo/ghdl-discuss
