Sebastian Pop wrote:
On Wed, Apr 29, 2009 at 17:15, Richard Guenther
<richard.guent...@gmail.com> wrote:
I don't see how SSA form makes anything more complicated ;)
One of the difficulties was regenerating the phi nodes after code
hoisting: CLooG optimizes
for (i)
if (invariant of i)
s += A[i];
into
if (invariant of i)
for (i)
s += A[i];
In the transformed code you have no place to put the phi nodes that
you had after the condition.
Add to this the problem of code duplication that CLooG does sometimes:
if (invariant of i)
for (i in domain1)
s += A[i];
for (i in domain2)
s += A[i];
...
Maintaining the SSA form for s is difficult after such transforms. If
you figure out a good way to maintain the SSA form, I'm very
interested to hear about.
I believe the short-cut proposed by Sebastian makes sense. We never go
out of SSA, just the hard-to-maintain-in-SSA induction variables are
converted temporarily into single-element arrays. This of course is only
a quick fix, and it does handle all cases. It will not complicate a
future rewrite of this into a nice in-SSA induction variable
reconstruction (an unexpected problem, worth investigating indeed, and
maybe a future deeper research result is hiding).
Albert
