* [PASS] Improve graph fusion

* Change fusion center to segment head

* Use 'master' to identity the schedule node

* Make things compact

* Fix

* [SCHEDULE] Fix fuse node order

* Make fuse order consistent with split

* Install rules and dynamic library loading fixes.

A batch of fixes:

- Added 'install' rule to cmake and make, which installs runtime
  headers and library (libtvm_runtime).

- Added 'installdev' rule to make, which also installs the compiler
  infrastructure headers and library (libtvm)

- Added 'INSTALL_DEV' option to cmake, for toggling installation
  of compiler infrastructure headers and library

- cmake no longer builds into lib/ directory; instead all build
  products go in your build directory

- New algorithm for dynamic library loading, as described in #281.

Signed-off-by: Edward Z. Yang <ezyang@fb.com>

* Emit dylib on OS X

Signed-off-by: Edward Z. Yang <ezyang@fb.com>

* Lint fixes.

Signed-off-by: Edward Z. Yang <ezyang@fb.com>

* __init__ updated

* pull request updated

* build_module added

* typo fixed

* another typo fixed

* [GRAPH] Add GraphHandle

* Move to apps/graph_executor

Signed-off-by: Edward Z. Yang <ezyang@fb.com>

* image add!

* image path move to web-data

* [TUTORIAL] Optimize gemm add!

* temp commitment

* [TUTORIAL] python3 compatiblity made; doc generation updated!

* [DOCS] gen_modules clean and ignore add!

* [TUTORIAL] title modified!

* [TUTORIAL] some rolled-back modification re-write

* [TUTORIAL] title underscore extended!

* CONTRIBUTORS add me!

* [EXECUTOR] Improve LayoutTransform pass

* Remove offline params for now

* Small fix

* [SCHEDULE] Remap the cached bind_scope.

* more fix

* [STORAGE][BUFFER] Support access ptr for clear access pattern.

* fix lint

* [tvm4j] register user-defined function

* [tvm4j] define java function (pushArgToStack) to convert arguments to C TVMValue

* [tvm4j] make Module & Function extends TVMValue

* [tvm4j] make registered cb function return Object

* [tvm4j] add cb finalizer; add TVMValueBytes

* [tvm4j] support NDArrayBase cb arg

* [tvm4j] register cb function unit tests

* [tvm4j] pass Function.Callback to resource_handle

* [tvm4j] fix type cast

* [SCHEDULE] Detect useless bound early

* fix

* [FIX] Fix bug and typo in rpc_server

* [FIX] Remove unnecessary condition

* [DOC] More detailed installation instruction

* fix lang

* [TOPI] Example for convolution

* update conv ex

* fix submodule HalideIR

* update conv impl

* python3

* minor fix

* fix pylint error

* Add test code

* x

* fix

* fix

* move python helper function into topi.testing

* fix pylint

* prefetch interface added

* prefetch python comments modified. prefetch info data structure maintained.

* start injecting prefetches. first step (domain touch) implemented.

* domain touch tested.

* Prefetch ir_mutator and ir_visitor dispatch registered.

* modify domain touched from passing a func_ref to passing a tensor

* modify domain touched from passing a func_ref to passing a tensor

* modify Tensor copy to Tensor ref

* temp commit for rebase

* debug info removed, typo fixed, ready to rebase

* prefetch flatten test add!

* roll back builtin functions to side effect functions

* lint error fixed!

* add cache line size to storage flatten argument

* forgot modifications add

* change code style to dmlc-like; get rid of can_prove, use manually compute instead

* python lint error fixed

* modify instrinsic name to pass tests

* [TEST] get rid of str(), replace them by accessing attributes

* change map to list comprehension

* redundant numpy import removed

* [EXECUTOR] Fix bug and improve

* [EXECUTOR] Enhance test case

* [RPC] Allow backpressure from writer

* fix

* fix

* [EXECUTOR] Enable load executor remotely

* [EXECUTOR] Pipeline

* Pass bytearray directly

* Enable load dynamic library in rpc_server.py

* Fix

* lint

* Return Module from remote side directly

* Remove unused header file

* Fix

* fix

* JVM package skeleton

* [JVM] link libtvm.so and list function names

* [JVM] Function & NDArray skeleton

* [JVM] TVMFuncCall in JNI

* [JVM] handle string arg in TVMFuncCall

* [JVM] get module function

* [JVM] entry function for Module

* [JVM] construct Module from function return value

* [JVM] TVMContext, TVMArray attributes

* [JVM] NDArray from / to java array

* [JVM] load so and compute on cpu

* [JVM] move PackedFunc to individual modules

* [JVM] assembly package & native library loader

* [JVM] unit test & codestyle check settings

* [JVM] NDArray from & to different dtypes

* [JVM] NDArray from native double array. Add linux-cpu profile.

* [JVM] modify Makefile

* [JVM] add linux-x86_64-gpu profile

* [tvm4j] delay load libtvm_runtime.so

* [tvm4j] refactor to pure java

* [tvm4j] remove scalastyle-config.xml

* [tvm4j] remove link HalideIR, remove Shape, remove scala binary versions

* [tvm4j] only allow convert from/to same type array

* [tvm4j] make NDArray api more readable

* [tvm4j] refactor for c api

* [tvm4j] add Jenkins tests

* [tvm4j] fix duplicate Dockerfile cmd

* [tvm4j] fix ut script filename

* [tvm4j] add module load tests

* [tvm4j] add javadoc, remove types package

* [tvm4j] fix test script

* [tvm4j] remove ut temp dir

* [tvm4j] fix missing package types

* [tvm4j] java code style check

* [tvm4j] fix java lint

* [tvm4j] downgrade checkstyle plugin for JDK7

* [tvm4j] add stylecheck in jenkins tests

* [tvm4j] specify source file encoding

* [tvm4j] lazy init function; add Function.call() api; allow manully release Module,NDArray,Function

* [tvm4j] fix ModFree

* [tvm4j] cache Function in API