This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: fix component_ref hashing regression
- From: kenner at vlsi1 dot ultra dot nyu dot edu (Richard Kenner)
- To: rth at redhat dot com
- Cc: gcc-patches at gcc dot gnu dot org
- Date: Thu, 24 Jun 04 06:40:17 EDT
- Subject: Re: fix component_ref hashing regression
The problem is that COMPONENT_REFs created after gimplification don't
get the new fields set. And since operand_equal_p wasn't updated, we
get fields that compare equal, but don't hash to the same value.
Oops, sorry! I missed that.
Kinda fixed by NOT setting the new fields when not necessary. It's
enough to fix the bug and the regressions wrt the optimization passes
that do create new reference nodes. We probably should still update
operand_equal_p to take into account the new fields.
I agree.
Or, as I'd prefer, the new tree codes.
Or the reference annotations, which I prefer.