[cfe-commits] r148375 - /cfe/trunk/lib/CodeGen/CodeGenModule.cpp

Nick Lewycky nlewycky at google.com
Tue Jan 17 20:33:02 PST 2012


On 17 January 2012 19:56, Matt Beaumont-Gay <matthewbg at google.com> wrote:

> On Tue, Jan 17, 2012 at 19:49, Nick Lewycky <nlewycky at google.com> wrote:
> > The testcase for this would be:
> >
> > #include <stdio.h>
> >
> > void testcase(void) {
> >   vprintf(0, 0);
> > }
> >
> > except that I will not introduce a test that isn't hermetic. It appears
> to
> > be impossible to fake this by declaring my own function, or even using #
> > lines to mark it. If I preprocess that testcase, the result no longer
> > triggers the memory error.
> >
> > Any suggestions for refactoring clang to make testing this possible
> would be
> > greatly appreciated.
>
>
> Add test/CodeGen/Inputs/stdio.h with an appropriate prototype and pass
> -isystem %S/Inputs on the RUN line?
>

Done, thanks!!
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/cfe-commits/attachments/20120117/2cf16693/attachment.html>


More information about the cfe-commits mailing list