[PATCH][4.1/4.2] Fix PR27260, fallout of fix for PR27095
Fri Apr 28 15:45:00 GMT 2006
On Fri, Apr 28, 2006 at 11:23:26AM +0200, Richard Guenther wrote:
> On 4/28/06, Alan Modra <email@example.com> wrote:
> >The testcase for pr27095 proved a little tricky. The one Richard
> >Guenther posted didn't fail on an unpatched compiler. Apparently we
> >need a builtin strlen in the memset arg; my_strlen doesn't trigger the
> Hmm, it failed for me (with 4.1.0) - the key was that my_strlen needs to be
> const or pure so TER places it in the argument list.
My apologies. I was testing a slight variation on your testcase.
Retesting with exactly your testcase as written shows it to be good.
IBM OzLabs - Linux Technology Centre
More information about the Gcc-patches