type gimplification, continued

Alexandre Oliva aoliva@redhat.com
Wed Dec 22 18:20:00 GMT 2004


On Dec 22, 2004, Richard Henderson <rth@redhat.com> wrote:

> Kenner, I still think you're wrong.
> Here's all of my patches reverted, plus the patch you gave
> me yesterday.  It does pretty good, in that it bootstraps.

> I think what you tested was *not* all of my patches reverted,
> which would be cheating.  We shouldn't need *both* unsharing
> and TYPE_SIZES_GIMPLIFIED.

> But it fails gcc.c-torture/compile/nested-1.c, which again
> shows invalid tree sharing between the parent and the nested
> function.

FWIW, a tree I got earlier today fails to bootstrap while compiling
libjava/verify.cc on both i686-linux-gnu and x86_64-linux-gnu
(compiling libjava with -m32).  It fails because the upper bound of
arg_types on line 2102 is an ssa_name, and
dwarf2out.c:loc_descriptor_from_tree doesn't know how to handle that.
May I assume this is the same problem that you guys are looking into?

-- 
Alexandre Oliva             http://www.ic.unicamp.br/~oliva/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}



More information about the Gcc mailing list