Hello all.
Any more comments on this? @jpienaar @rsuderman @stellaraccident

Rename pass to TosaLayerwiseConstantFoldPass as per RFC suggestions.

In D124685#3544276, @stellaraccident wrote:

Generally LGTM, although we may want to further constrain this in a follow-up. In general, I am pro having optimization passes and having them be truly optional. I would be tempted to name this "TensorDataConstantOptimizationPass".

In D124685#3535384, @mehdi_amini wrote:

Can you clarify the revision description? You refer to some "unbound computation" which isn't clear to me. I think the whole motivation and rationale could be better spelled out.

(also remove the Change-Id: I7ec0f8b15ca6bc9aa2116488dcf6c684c9826ddd part which seems a leftover of some other system?)

Update commit message

Reworks commit message:

• Description simplification
• Remove unneeded Change-Id

@rsuderman @jpienaar any further updates needed?
thank you in advance

In D124685#3514697, @jpienaar wrote:

Good point about potentially expensive operations here, I wonder: why is/was this a canonicalization pattern rather than fold implementation?

Not sure @jpienaar why this was registered as a canonicalization originally. Can comment though that have noticed this being quite expensive. Moreover, a similar folding step takes place at Linalg level from what I recall.

Address review comments.