throw error when mismatch cpu avx version

Fea/jit/gen

fix to only check block 0

test=develop

test=develop

"fix sign op"

Fix selected_rows clip bug

add more debug info.

test=develop

add inclusive/exclusive mode in avg pool

Grid sampler operator for spatial transformer network.

Polish hash op comment

Fix paddle version

test=develop

test=develop

test=develop

Add numeric_stable_mode parameters to softmax_with_xe op

Enable sequential execution mode in parallel executor

test=develop

* remove unnecessary var_and_op
test=develop

* fix _init_zero_idx_
test=develop

test=develop

test=develop

add depthwise conv mkldnn pass

add program check

* enable pyreader use pin memory

* add py reader pin memory test test=develop

* Add affine grid generator.

* fix ffine grid.

* Add unitest.

* Add CPU kernel and fix unitest.

* Fix CPU kernel.

* Refine code.
test=develop

* Fix python api.
test=develop

* Update python api.
test=develop

* Fix comment.
test=develop

* Rename affine_grid_generator to affine_grid and enhence unitest.
test=develop

* Fix unitest.
test=develop

* fix dim0 in _load_slice_up_vars

* fix dim0 in _load_slice_up_vars, fix innershape in delete_var_op

* Revert "fix lookuptable in reduce strategy"

This reverts commit 0e722c5e

* add unit test for dist

* add unit test for dist, test=develop

* cancel revert, test=develop

Fix the example in the doc of transpose_op

Refine the doc of dynamic_gru and gru_unit.

remove unused WITH_FAST_BUNDLE_TEST option

added transpiler pass for mkldnn depthwise_conv