Re: [PATCH] Fix PR middle-end/39976, 200.sixtrack degradation

[Jeff Law <law at redhat dot com>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 12/02/11 12:27, William J. Schmidt wrote:

> 
> Erm, wait.  How are PHIs in different blocks going to have the
> same incoming edges?  (I was thinking of control dependence edges,
> but these are just regular control flow incoming edges, right?)  So
> this really isn't going to help.
They're not.  But if we find an equivalence between phi_1 and phi_2,
then we can replace every reference to phi_2 with phi_1.  This is safe
because any reference to phi_2 must be dominated by the assignment to
phi_2 which is in the same block as phi_1.

So while continuing to have the phis in the available expression table
is not useful beyond the current block, the equivalency created when a
redundant PHI is encountered is useful to keep.

I may have not made the distinction clearly in prior messages.  If
that's what your patch does, then you're golden.

Jeff

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/

iQEcBAEBAgAGBQJO2TktAAoJEBRtltQi2kC7hUsH/iWomNPVNnGw00FhLVBuiIVC
c/Tsirdu256H07v1yZBemoh65EEOiqy6gHbKV2ZgL8MADJYm4i17Ii/0CEJyXHt2
YpplJmVY905WKqs2KN/qWHXAo7YFQwAj2MRWSksi2VMlq0YBHL+OA0qVlPLcNRUK
3e92nyERJAHgNlQqBLzGpeMLw8ozs8ognVZj9L/fbRWf4Jgnh5v5oPu50n9pWshO
ipq+J5Qbovli9c8lHq6etFZ3EVCCxahnZ4FF1rxI3mVOKnL90xFPprBB1jL6qcqx
8O79yTLK6M7u3CTmnD8KoociAMWOhoe34o8PQIYEtJC5Pops1jKViEIbsOQg9s0=
=NXbN
-----END PGP SIGNATURE-----