[mlir][sparse] adjust output tensor to synthetic tensor

[aartbik]

Summary

Fixes a merge conflict with previous two CLs.

[mravishankar]

I dont have full (or any for that matter) context on the sparsification work, but are you setting the t to be "out-of-bounds" of the operand index values?

[aartbik]

In D93664#2466889, @mravishankar wrote:

I dont have full (or any for that matter) context on the sparsification work, but are you setting the t to be "out-of-bounds" of the operand index values?

See comment at the merger constructor. We reserve one extra one for a "synthetic" tensor, that is reserved for all invariants.
I used to borrow the output tensor for that, but thought it was less clean. I will actually introduce symbolics for this to avoid the problem in the future.

mlir/lib/Dialect/Linalg/Transforms/Sparsification.cpp

 static unsigned buildLattices(Merger &merger, linalg::GenericOp op,
                               unsigned exp, unsigned idx) {
   Kind kind = merger.exp(exp).kind;
   if (kind == Kind::kTensor || kind == Kind::kInvariant) {
     // Either the index is really used in the tensor expression, or it is
     // set to the undefined index in that dimension. An invariant expression
     // is set to a synthetic tensor with undefined indices only.
     unsigned s = merger.addSet();
-    unsigned t = kind == Kind::kTensor ? merger.exp(exp).e0
-                                       : op.getNumShapedOperands() - 1;
+    unsigned t =
+        kind == Kind::kTensor ? merger.exp(exp).e0 : op.getNumShapedOperands();
     merger.set(s).push_back(merger.addLat(t, idx, exp));
     return s;
   }
   unsigned s0 = buildLattices(merger, op, merger.exp(exp).e0, idx);
   unsigned s1 = buildLattices(merger, op, merger.exp(exp).e1, idx);
   switch (kind) {
   case Kind::kTensor:
   case Kind::kInvariant: