This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: Tree-SSA self checking infrastructure
- From: law at redhat dot com
- To: Richard Henderson <rth at redhat dot com>
- Cc: Jan Hubicka <hubicka at ucw dot cz>, Jan Hubicka <jh at suse dot cz>, gcc at gcc dot gnu dot org
- Date: Wed, 19 Nov 2003 14:27:12 -0700
- Subject: Re: Tree-SSA self checking infrastructure
- Reply-to: law at redhat dot com
In message <20031119210127.GB31811@redhat.com>, Richard Henderson writes:
>On Wed, Nov 19, 2003 at 12:13:06PM -0700, law@redhat.com wrote:
>> >This brings me into questions about my tail call updating code.
>> >Perhaps I need to re-do SSA form on call cobbered variables after
>> >removing the call?
>> Are you changing the CFG? Are you changing the dominator tree? Those
>> are the key questions.
>
>Yes, he is, but only by adding a new block on the ENTRY->BB0 edge,
>and then edges into that block from the tail recursion sites.
So that would mean he needs PHIs for any arguments that are changed which
reach the loop backedge and for local variables in the toplevel scope
which have a definition which reaches the backedge.
> I may be mistaken, but I thought the new phi nodes he was adding
> there would be correct. Perhaps this is a good test for a verify_ssa
> pass?
Definitely a good test.
jeff