Support Math keyword in native_functions.yaml.

[ailzhang]

Stack from ghstack:

• Define macro for DispatchKey::toString. #44993 Define macro for DispatchKey::toString.
• Align casing in test_dispatch with dispatch keys. #44933 Align casing in test_dispatch with dispatch keys.
• Support Math keyword in native_functions.yaml. #44556 Support Math keyword in native_functions.yaml.
• Add alias dispatch key Math. #44354 Add alias dispatch key Math.

Differential Revision: D23698386

[dr-ci]

💊 CI failures summary and remediations

As of commit ff5d225 (more details on the Dr. CI page):

---

• 3/3 failures introduced in this PR

---

🕵️ 3 new failures recognized by patterns

The following CI failures do not appear to be due to upstream breakages:

pytorch_macos_10_13_py3_test (1/3)

Step: "Test" (full log | diagnosis details | 🔁 rerun)

Sep 18 20:02:57 [E request_callback_no_python.cpp:618] Received error while processing request type 2: RuntimeError: Can not pickle torch.futures.Future

Sep 18 20:02:57 At: 
Sep 18 20:02:57   /Users/distiller/workspace/miniconda3/lib/python3.7/site-packages/torch/distributed/rpc/internal.py(94): serialize 
Sep 18 20:02:57   /Users/distiller/workspace/miniconda3/lib/python3.7/site-packages/torch/distributed/rpc/internal.py(146): serialize 
Sep 18 20:02:57  
Sep 18 20:02:57 [E request_callback_no_python.cpp:618] Received error while processing request type 2: RuntimeError: Can not pickle torch.futures.Future 
Sep 18 20:02:57  
Sep 18 20:02:57 At: 
Sep 18 20:02:57   /Users/distiller/workspace/miniconda3/lib/python3.7/site-packages/torch/distributed/rpc/internal.py(94): serialize 
Sep 18 20:02:57   /Users/distiller/workspace/miniconda3/lib/python3.7/site-packages/torch/distributed/rpc/internal.py(146): serialize 
Sep 18 20:02:57  
Sep 18 20:02:57 [E request_callback_no_python.cpp:618] Received error while processing request type 2: RuntimeError: Can not pickle torch.futures.Future 
Sep 18 20:02:57  
Sep 18 20:02:57 At: 
Sep 18 20:02:57   /Users/distiller/workspace/miniconda3/lib/python3.7/site-packages/torch/distributed/rpc/internal.py(94): serialize 
Sep 18 20:02:57   /Users/distiller/workspace/miniconda3/lib/python3.7/site-packages/torch/distributed/rpc/internal.py(146): serialize 
Sep 18 20:02:57  
Sep 18 20:02:58 ok (1.626s) 
Sep 18 20:02:59   test_return_future_remote (__main__.ProcessGroupRpcTestWithSpawn) ... ok (1.531s) 
Sep 18 20:03:01   test_return_local_rrefs (__main__.ProcessGroupRpcTestWithSpawn) ... ok (1.628s) 
Sep 18 20:03:03   test_rpc_profiling_remote_record_function (__main__.ProcessGroupRpcTestWithSpawn) ... ok (1.633s) 
Sep 18 20:03:04   test_rpc_return_rref (__main__.ProcessGroupRpcTestWithSpawn) ... ok (1.618s) 

pytorch_xla_linux_bionic_py3_6_clang9_build (2/3)

Step: "Build" (full log | diagnosis details | 🔁 rerun)

Sep 19 00:58:19 ERROR:sccache::server: Compilation failed: Output { status: ExitStatus(ExitStatus(256)), stdout: "", stderr: "/var/lib/jenkins/workspace/build/CMakeFiles/CMakeTmp/CheckSymbolExists.c:8:19: error: use of undeclared identifier \'strtod_l\'\n return ((int*)(&strtod_l))[argc];\n ^\n1 error generated.\n" }

Sep 19 00:58:18     assert check_overrides(overrides, overridden) 
Sep 19 00:58:18 AssertionError 
Sep 19 00:58:18 Building torch_xla version: 1.6 
Sep 19 00:58:18 XLA Commit ID: 14d617a2ec11f7e579aa7eb39d5331331e2940c3 
Sep 19 00:58:18 PyTorch Commit ID: ff5d225a5cec67aee8999d829ce81e06684dc1ce 
Sep 19 00:58:18 Failed to generate ATEN bindings: ['/var/lib/jenkins/workspace/xla/scripts/generate_code.sh'] 
Sep 19 00:58:19 =================== sccache compilation log =================== 
Sep 19 00:58:19 + cleanup 
Sep 19 00:58:19 + retcode=1 
Sep 19 00:58:19 + set +x 
Sep 19 00:58:19 ERROR:sccache::server: Compilation failed: Output { status: ExitStatus(ExitStatus(256)), stdout: "", stderr: "/var/lib/jenkins/workspace/build/CMakeFiles/CMakeTmp/CheckSymbolExists.c:8:19: error: use of undeclared identifier \'strtod_l\'\n  return ((int*)(&strtod_l))[argc];\n                  ^\n1 error generated.\n" } 
Sep 19 00:58:19  
Sep 19 00:58:19 =========== If your build fails, please take a look at the log above for possible reasons =========== 
Sep 19 00:58:19 Compile requests               5992 
Sep 19 00:58:19 Compile requests executed      3551 
Sep 19 00:58:19 Cache hits                     2359 
Sep 19 00:58:19 Cache misses                   1176 
Sep 19 00:58:19 Cache timeouts                    0 
Sep 19 00:58:19 Cache read errors                 0 
Sep 19 00:58:19 Forced recaches                   0 
Sep 19 00:58:19 Cache write errors                0 

pytorch_linux_bionic_py3_8_gcc9_coverage_test (3/3)

Step: "Run tests" (full log | diagnosis details | 🔁 rerun)

Sep 19 03:10:34 [E request_callback_no_python.cpp:618] Received error while processing request type 2: RuntimeError: Can not pickle torch.futures.Future

Sep 19 03:10:34 At: 
Sep 19 03:10:34   /opt/conda/lib/python3.8/site-packages/torch/distributed/rpc/internal.py(94): serialize 
Sep 19 03:10:34   /opt/conda/lib/python3.8/site-packages/torch/distributed/rpc/internal.py(146): serialize 
Sep 19 03:10:34  
Sep 19 03:10:34 [E request_callback_no_python.cpp:618] Received error while processing request type 2: RuntimeError: Can not pickle torch.futures.Future 
Sep 19 03:10:34  
Sep 19 03:10:34 At: 
Sep 19 03:10:34   /opt/conda/lib/python3.8/site-packages/torch/distributed/rpc/internal.py(94): serialize 
Sep 19 03:10:34   /opt/conda/lib/python3.8/site-packages/torch/distributed/rpc/internal.py(146): serialize 
Sep 19 03:10:34  
Sep 19 03:10:34 [E request_callback_no_python.cpp:618] Received error while processing request type 2: RuntimeError: Can not pickle torch.futures.Future 
Sep 19 03:10:34  
Sep 19 03:10:34 At: 
Sep 19 03:10:34   /opt/conda/lib/python3.8/site-packages/torch/distributed/rpc/internal.py(94): serialize 
Sep 19 03:10:34   /opt/conda/lib/python3.8/site-packages/torch/distributed/rpc/internal.py(146): serialize 
Sep 19 03:10:34  
Sep 19 03:10:34 ok (1.541s) 
Sep 19 03:10:36   test_return_future_remote (__main__.TensorPipeRpcTestWithSpawn) ... [W tensorpipe_agent.cpp:577] RPC agent for worker2 encountered error when reading incoming request from worker0: EOF: end of file (this is expected to happen during shutdown) 
Sep 19 03:10:36 [W tensorpipe_agent.cpp:577] RPC agent for worker2 encountered error when reading incoming request from worker1: EOF: end of file (this is expected to happen during shutdown) 
Sep 19 03:10:36 [W tensorpipe_agent.cpp:577] RPC agent for worker1 encountered error when reading incoming request from worker0: EOF: end of file (this is expected to happen during shutdown) 
Sep 19 03:10:36 ok (1.537s) 

---

This comment was automatically generated by Dr. CI (expand for details).

Follow this link to opt-out of these comments for your Pull Requests.

Please report bugs/suggestions on the GitHub issue tracker or post in the (internal) Dr. CI Users group.

See how this bot performed.

This comment has been revised 65 times.

[ezyang]

This is definitely something we should work on making easier, though I'm not exactly sure how. Perhaps APIs for supporting static runtime would make it easier, as they would make it possible to ask for a specific kernel.

Absent machinery like this, I think the easiest way to do this is to just make the math kernel available at another name (i.e., make a new native_functions.yaml entry; and then just do the testing from Python. Maybe @mruberry has some other ideas.

[mruberry]

Without too much context I'll also vote for (re)using Python.

[bhosmer]

Looks good!

[facebook-github-bot]

@ailzhang merged this pull request in 4b42f0b.