support equal and not_equal in topi

[liangfu]

Add support to count_nonzero in topi, works well for llvm. However, current cuda schudule for reduction don't work well with the proposed operator.

[tqchen]

Can we compose this operator by having a equal operator then followed by a sum, if so, we should support equal and instead of this function instead

[liangfu]

I gave it a try, and test cases has been added.

[tqchen]

what i mean instead is that we should not support count nonzero function, but instead support topi.equal and topi.not_equal, so we can compose count nonzero by equal then astype(int32) and the sum

[liangfu]

i could do that, but my original intention is to estimate buffer size for topi.nonzero, which could be useful for converting dense matrices into sparse ones.

[tqchen]

I agree, nevertheless these can be done in the composed fashion so we can express the same thing but much more flexibly

[liangfu]

Okay, here is another update.
I think it's fine if we don't need count_nonzero in topi, but how about a topi.nonzero?
It could be a progressive step towards an introduction to sparse tensor.

[tqchen]

remove the print here

[tqchen]

I agree that topi.nonzero is useful and can be supported in a separate PR. although it is unclear if we need additional gather primitive/hybrid script to support this operation