On Tue, Apr 20, 2010 at 6:16 AM, David A. Case <case.biomaps.rutgers.edu> wrote:
> On Mon, Apr 19, 2010, Daniel Roe wrote:
>
>> 5 EVB tests still fail on Cygwin due to the stack size and lack of an
>> 'unlimit' command. I have modified the test cases in question so that they
>> check that the stack was actually resized before running. Attached is a
>> patch that can be run in the $AMBERHOME/test directory to modify the tests.
>> The modification to each test is this:
>>
>> -unlimit stack
>> +limit stacksize unlimited
>> +set STACK=`limit stacksize | awk '{print $2}'`
>> +if (! $?STACK || $STACK != "unlimited" || $STACK == "" ) then
>> + echo "This test requires a larger stack size. Exiting."
>> + exit(0)
>> +endif
>>
>
> I'm really inclined to skip this patch...it just looks too dangerous, in that
> it could break the test suite on other platforms -- who knows what kind of
> errors the "limit stacksize unlimited" command will yield? Or what machines
> might actually work, but would not set STACK to "unlimited"?
>
> We can have failures in cygwin.
>
> ...dac
Just my 0.02...
As an occasional Cygwin user myself, I believe it could be nice to
have at least a check, and make the test print a statement like: "You
seem to be using Cygwin. It is possible that this test fails because
of the default stack size limit in your system. If this happens to
you, consider resizing the stack limit by following the instruction in
http://cygwin_faq_address_to_the_stack_issue_here."
That would kill two birds with one stone: tell the user about the
Cygwin limits, and teach him how to circumvent it if he's really
interested. Then he can try the test himself later if he wants to.
Finally, it avoids the need to change the code at this point.
Cheers,
--
Gustavo Seabra
Professor Adjunto
Departamento de Química Fundamental
Universidade Federal de Pernambuco
Fone: +55-81-2126-7417
_______________________________________________
AMBER-Developers mailing list
AMBER-Developers.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber-developers
Received on Tue Apr 20 2010 - 05:30:03 PDT
