[tree-ssa] New regressions as of 2003-11-04

[law@redhat.com]

In message <20031104230929.GA24820@redhat.com>, Richard Henderson writes:
 >On Tue, Nov 04, 2003 at 03:08:53PM -0700, law@redhat.com wrote:
 >> Right, but the remove_useless_stmts_and_vars does kill the CFG.
 >
 >No it doesn't.  Mostly because at present it doesn't do
 >anything at all when called after lowering.  And, really,
 >I don't see how it could.
I must admit I haven't looked closely at it since all the lowering, so
you may in fact be correct.


 >If you want to clean up out-of-ssa bits, I suggest running
 >another dominator pass, geared to detect this sort of thing.
Another dom pass isn't going to improve this kind of code.  Either the
useless statement remover needs to be updated, or we need something to
eliminate the PHI, or the out-of-ssa pass needs to get a lot smarter.

jeff