[general] major overhaul of triton-c/triton-ir/triton-jit:

- Added alloc const
- Added atomics
- Pruning tuning space
- Added example for dot/conv/shift
- Bugfixes

examples/cpp/CMakeLists.txt

@@ -0,0 +1,6 @@
+foreach(PROG dot conv shift)
+     add_executable(${PROG} ${PROG}.cpp)
+     set_target_properties(${PROG} PROPERTIES OUTPUT_NAME ${PROG})
+     include_directories(/usr/local/cuda/include/)
+     target_link_libraries(${PROG} triton)
+endforeach(PROG)