Add conv2d mlir files and with naive implementation nchw& nhwc

[Joejiong]

Add 2 commonly used conv2d mlir files

• conv2d_nhwc_filter_hwcf.mlir
• conv2d_nchw_filter_fchw.mlir

> ./bin/conv-2d-nchw-fchw-benchmark                                                            add-op-benchmark [bd87cdf]
2021-12-02T17:04:49+08:00
Running ./bin/conv-2d-nchw-fchw-benchmark
Run on (52 X 2500 MHz CPU s)
CPU Caches:
  L1 Data 32 KiB (x26)
  L1 Instruction 32 KiB (x26)
  L2 Unified 1024 KiB (x26)
  L3 Unified 36608 KiB (x1)
Load Average: 17.59, 14.52, 12.63
***WARNING*** Library was built as DEBUG. Timings may be affected.
--------------------------------------------------------------
Benchmark                    Time             CPU   Iterations
--------------------------------------------------------------
BM_Conv2DNchwFchw/1        280 ns          280 ns      2503077
BM_Conv2DNchwFchw/4       1105 ns         1105 ns       633815
Output: [ 48 48 48 48 48 48 48 48 ]

----------------------------------------------------------------------------------------------------------------------------

(mmperf_env) 
root@ebe2d30121e5:/Workspace/buddy-benchmark/build (!)
> ./bin/conv-2d-nhwc-hwcf-benchmark                                                            add-op-benchmark [bd87cdf]
2021-12-02T17:05:29+08:00
Running ./bin/conv-2d-nhwc-hwcf-benchmark
Run on (52 X 2500 MHz CPU s)
CPU Caches:
  L1 Data 32 KiB (x26)
  L1 Instruction 32 KiB (x26)
  L2 Unified 1024 KiB (x26)
  L3 Unified 36608 KiB (x1)
Load Average: 19.25, 15.43, 13.02
***WARNING*** Library was built as DEBUG. Timings may be affected.
--------------------------------------------------------------
Benchmark                    Time             CPU   Iterations
--------------------------------------------------------------
BM_Conv2DNhwcHwcf/1        282 ns          282 ns      2478301
BM_Conv2DNhwcHwcf/4       1119 ns         1119 ns       626199
Output: [ 48 48 48 48 48 48 48 48 ]