llvm
/
torch-mlir
mirror of https://github.com/llvm/torch-mlir
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity
torch-mlir/test/Dialect/TCP/ops.mlir

12 lines
397 B
MLIR
Raw Normal View History

Initial TCF/TCP E2E seed. Very much WIP. This is enough to get tcf.add down to approximately the "linalg.generic on buffers" level of abstraction. (but there are nuances)
2020-05-07 09:41:54 +08:00
// RUN: npcomp-opt <%s | FileCheck %s --dump-input=fail
npcomprt: add support for constants - create tcp.global + tcp.get_global_memref - create npcomprt.global + npcomprt.get_global - LLVM lowering for new npcomprt ops - Runtime: - GlobalDescriptor struct emitted by LLVM lowering - implement __npcomp_compiler_rt_get_global Also, - cleanly isolate all runtime data structure definitions shared by the compiler and runtime into lib/runtime/CompilerDataStructures.h
2020-07-11 08:31:24 +08:00
// CHECK-LABEL: tcp.global @foo dense<0.0{{.*}}> : tensor<10xf32>
tcp.global @foo dense<0.0> : tensor<10xf32>
Initial TCF/TCP E2E seed. Very much WIP. This is enough to get tcf.add down to approximately the "linalg.generic on buffers" level of abstraction. (but there are nuances)
2020-05-07 09:41:54 +08:00
func @f(%arg0: tensor<?xf32>, %arg1: tensor<?xf32>, %arg2: i32) {
// CHECK: tcp.add
Remove the present tcp.island. The idea was half-baked and after some deep thought felt like a solution looking for a problem. What we had here (and is removed in this patch) just wasn't pulling its weight. I cannot think of anything we would want to do with tcp.island as it is removed here beyond just sinking and merging them within a basic block, such that the witness argument is kind of pointless (only matters for hoisting). TCP compute ops like tcp.add and tcp.broadcast_to have the strong invariant of "pure or undefined behavior", which means they are always safe to sink. The island concept as removed here conferred no benefit. Also, I'll note that "islands" are a trick you can only play once in a system (unless they strictly nest). I have some early-stage thoughs on having an island concept that helps with modeling tensor shapes robustly which seems promising (the island would serve a similar role as tie_shape).
2020-05-15 06:19:37 +08:00
%0 = "tcp.add"(%arg0, %arg1) : (tensor<?xf32>, tensor<?xf32>) -> tensor<?xf32>
npcomprt: add support for constants - create tcp.global + tcp.get_global_memref - create npcomprt.global + npcomprt.get_global - LLVM lowering for new npcomprt ops - Runtime: - GlobalDescriptor struct emitted by LLVM lowering - implement __npcomp_compiler_rt_get_global Also, - cleanly isolate all runtime data structure definitions shared by the compiler and runtime into lib/runtime/CompilerDataStructures.h
2020-07-11 08:31:24 +08:00
%1 = tcp.get_global_memref @foo : memref<10xf32>
Initial TCF/TCP E2E seed. Very much WIP. This is enough to get tcf.add down to approximately the "linalg.generic on buffers" level of abstraction. (but there are nuances)
2020-05-07 09:41:54 +08:00
return
Remove the present tcp.island. The idea was half-baked and after some deep thought felt like a solution looking for a problem. What we had here (and is removed in this patch) just wasn't pulling its weight. I cannot think of anything we would want to do with tcp.island as it is removed here beyond just sinking and merging them within a basic block, such that the witness argument is kind of pointless (only matters for hoisting). TCP compute ops like tcp.add and tcp.broadcast_to have the strong invariant of "pure or undefined behavior", which means they are always safe to sink. The island concept as removed here conferred no benefit. Also, I'll note that "islands" are a trick you can only play once in a system (unless they strictly nest). I have some early-stage thoughs on having an island concept that helps with modeling tensor shapes robustly which seems promising (the island would serve a similar role as tie_shape).
2020-05-15 06:19:37 +08:00
}