[cfe-dev] Test specific to Windows host?

[Eric Christopher]

Yeah. Don't do that. It'll just pass fine on other platforms, that seems ok.

-eric


On Wed, Mar 6, 2013 at 10:31 AM, Robinson, Paul
<Paul.Robinson at am.sony.com>wrote:

>  The target is irrelevant.  Clang running on Windows was hanging if the
> output-file path name was too long; I have a patch to make it return an
> error instead.****
>
> But another host OS won’t get an error for that case.  So, the test
> behavior will vary by host OS.  (Running Clang on Linux with a Windows
> target triple will not fail the same way because Linux has different
> path-length limits.)  I’d rather constrain it to run only on a Windows host.
> ****
>
> Or, hm, I suppose I could write the test to expect success and XFAIL:
> windows, but that seems like a perversion of the intent of XFAIL.****
>
> --paulr****
>
> ** **
>
> *From:* Eric Christopher [mailto:echristo at gmail.com]
> *Sent:* Tuesday, March 05, 2013 5:26 PM
> *To:* Robinson, Paul
> *Cc:* cfe-dev at cs.uiuc.edu
> *Subject:* Re: [cfe-dev] Test specific to Windows host?****
>
> ** **
>
> -target?****
>
> ** **
>
> Otherwise can you explain more of what you're trying to do?****
>
> ** **
>
> -eric****
>
> ** **
>
> On Tue, Mar 5, 2013 at 5:10 PM, Robinson, Paul <Paul.Robinson at am.sony.com>
> wrote:****
>
> I'm working up a fix for a problem but it's specific to Windows;
> the test won't behave the same way on Linux or whatever.
> AFAICT the ways of expressing constraints are mostly target
> based.  Is there a way to express "run only on Windows host"?
> Thanks,
> --paulr
>
>
>
> _______________________________________________
> cfe-dev mailing list
> cfe-dev at cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/cfe-dev****
>
> ** **
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/cfe-dev/attachments/20130306/6b050880/attachment.html>