[mlir] Rename linalg.pooling operations to have a FOp postfix for floating point

[rsuderman]

Summary

Linalg pooling operations only support floating point currently but integer
variants will soon be needed. Renaming to uncluse a FOp postfix to clarify.

[rsuderman]

Mostly a renaming proposal so I can look into the quantized pooling variant for TOSA.

[mravishankar]

Just a thought if we want to just rename them to F32. But this path is on deprecation, so I dont really care.

[rsuderman]

In D99779#2666425, @mravishankar wrote:

Just a thought if we want to just rename them to F32. But this path is on deprecation, so I dont really care.

We might as well stick to FOp just to match surrounding functions. Perhaps there an be another rename in the future.

mlir/include/mlir/Dialect/Linalg/IR/LinalgNamedStructuredOpsSpec.tc

 {
   O(n, f, d, h, w) = std_addf<kd, kh, kw>(
       O(n, f, d, h, w), std_mulf(I(n, c, d * strides[0] + kd * dilations[0],
                                    h * strides[1] + kh * dilations[1],
                                    w * strides[2] + kw * dilations[2]),
                                  K(kd, kh, kw, c, f)));
 }
 
-ods_def<PoolingNHWCSumOp>:
+ods_def<PoolingNHWCSumFOp>:
 def pooling_nhwc_sum
     (I: f32(N, H, W, C), K: f32(KH, KW))
   -> (O: f32(N, OH, OW, C))
   attr(strides: 2xi64, dilations: 2xi64)
 {
   O(n, oh, ow, c) = std_addf<kh, kw>(O(n, oh, ow, c),
                                      I(n, oh * strides[0] + kh * dilations[0],
                                        ow * strides[1] + kw * dilations[1], c));
 }
 
-ods_def<PoolingNHWCMaxOp>:
+ods_def<PoolingNHWCMaxFOp>:
 def pooling_nhwc_max
     (I: f32(N, H, W, C), K: f32(KH, KW))
   -> (O: f32(N, OH, OW, C))
   attr(strides: 2xi64, dilations: 2xi64)
 {
   O(n, oh, ow, c) =
       std_select<kh, kw>(std_cmpf_ogt(I(n, oh * strides[0] + kh * dilations[0],
                                         ow * strides[1] + kw * dilations[1], c),
                                       O(n, oh, ow, c)),
                          I(n, oh * strides[0] + kh * dilations[0],
                            ow * strides[1] + kw * dilations[1], c),
                          O(n, oh, ow, c));
 }
 
-ods_def<PoolingNHWCMinOp>:
+ods_def<PoolingNHWCMinFOp>:
 def pooling_nhwc_min
     (I: f32(N, H, W, C), K: f32(KH, KW))
   -> (O: f32(N, OH, OW, C))
   attr(strides: 2xi64, dilations: 2xi64)
 {
   O(n, oh, ow, c) =
       std_select<kh, kw>(std_cmpf_olt(I(n, oh * strides[0] + kh * dilations[0],
                                         ow * strides[1] + kw * dilations[1], c),
                                       O(n, oh, ow, c)),
                          I(n, oh * strides[0] + kh * dilations[0],
                            ow * strides[1] + kw * dilations[1], c),
                          O(n, oh, ow, c));
 }