Will tree-ssa be GCC 3.5?

[Richard Kenner]

    Well, we already have cases where the tree-ssa creates initial RTL which is
    *better* than the RTL you would have after running the first cse pass.

Of course, but that isn't what I meant. Certainly for many
optimizations, tree-ssa will produce a much more optimized tree than
before any optimizations were done before: that's nearly a tautology.
The issue I was raising is whether and to what extent the lowering
process produces RTL which is subject to *further* optimization.