[llvm-commits] [PATCH, PowerPC] Fix test case for default unsigned char platforms

[Benjamin Kramer]

On 10.10.2012, at 19:39, Ulrich Weigand <Ulrich.Weigand at de.ibm.com> wrote:

> 
> Hello,
> 
> the SingleSource/UnitTests/2003-07-09-SignedArgs.c test case
> fails on PowerPC because it wants to verify that "char" arguments
> are sign-extended.  However, whether a plain "char" is signed or
> unsigned in C is implementation-defined, and on PowerPC plain
> char happens to be unsigned.
> 
> The appended patch fixes the test by simply making the assumption
> explicit and using "signed char" instead of "char" throughout.
> 
> OK to commit?

I'm a bit surprised that nobody ran this testcase on a ppc since 2003, but the change looks right to me.

- Ben