This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [tree-ssa] Lowering of VLA's, part 2
Hello,
> >> > On Fri, 2003-08-29 at 15:15, Zdenek Dvorak wrote:
> >> >
> >> >> size of x now depends on non-existent variable z.1, which causes a crash
> >> >> when an attempt to copy it before passing to foo is made.
> >> >>
> >> > But this also means that we are still generating VLAs. Shouldn't your
> >> > original change removed that?
> >>
> >> No, the decl still has VLA type.
> >>
> > So, we can't really expose all VLAs then? I was hoping we could
> > eventually get rid of the 'has_hidden_uses' attribute. Oh, well.
me too :-(
> Perhaps we want to make the copy explicit in the trees.
I have considered this; however it would probably need some hacking
into call expansion (no idea how difficult one).
Zdenek
- References:
- Re: [tree-ssa] Lowering of VLA's, part 2
- Re: [tree-ssa] Lowering of VLA's, part 2
- Re: [tree-ssa] Lowering of VLA's, part 2
- Re: [tree-ssa] Lowering of VLA's, part 2
- Re: [tree-ssa] Lowering of VLA's, part 2
- Re: [tree-ssa] Lowering of VLA's, part 2
- Re: [tree-ssa] Lowering of VLA's, part 2
- Re: [tree-ssa] Lowering of VLA's, part 2
- Re: [tree-ssa] Lowering of VLA's, part 2
- Re: [tree-ssa] Lowering of VLA's, part 2