Re: SSA usage question

[kenner at vlsi1 dot ultra dot nyu dot edu (Richard Kenner)]

    Assigned to.  ie, something like:

    x = ...
    v = ...
    ... = a[3,x,v]

Yeah, that what gimplification should do.

    How will this added information affect the optimizers?  

It really shouldn't except to count as a usage.

    Isn't this information about TYPE_MIN_VALUE and TYPE_SIZE_UNIT already
    expressed in the TREE_TYPE of the ARRAY_REF?  

Of course it is!  That was my *original* point, that the data was alrady
there but wasn't being marked as used.  But when I suggested using it,
it was objected to by virtue of it being an "implicit access".

    Why are these attributes replicated in every ARRAY_REF?  

Because that's what Jeff Law and Richard Henderson wanted.

    These things can change from reference to reference?  That is,
    given the array 'a' above, will every reference to 'a' use a
    potentially different 'x' and 'v'?

No.  They are the unchanging over the lifetime of an object.  If the
type has placeholders, then it may be different between two objects of
the same type, though.