llvm / llvm-project

The LLVM Project is a collection of modular and reusable compiler and toolchain technologies.
http://llvm.org
Other
29.13k stars 12.01k forks source link

Aborted (core dumped) #73434

Closed CHzhangi closed 11 months ago

CHzhangi commented 11 months ago

root@4ee3c0dff15b:/workspace# model_transform.py --model_name fsr --input_shape [[1,3,200,200]] --model_def FSRCNN-pytorch-master/fsrcnn.pt --mlir fsr.mlir SOPHGO Toolchain v1.3.116-gf735aff3-20230821 2023/11/26 17:36:44 - INFO :


| preprocess:                                           |
|   (x - mean) * scale                                  |
'-------------------------------------------------------'

config Preprocess args : resize_dims : same to net input dims keep_aspect_ratio : False keep_ratio_mode : letterbox pad_value : 0 pad_type : center

mean                  : [0.0, 0.0, 0.0]
scale                 : [1.0, 1.0, 1.0]
--------------------------
pixel_format          : bgr
channel_format        : nchw

Save mlir file: fsr_origin.mlir [Running]: tpuc-opt fsr_origin.mlir --shape-infer --canonicalize --extra-optimize -o fsr.mlir tpuc-opt: ../lib/Dialect/Top/Interfaces/Conv.cpp:137: void tpu_mlir::top::ConvOp::shape_inference(): Assertion `input_shape.size() == filter_shape.size()' failed. PLEASE submit a bug report to https://github.com/llvm/llvm-project/issues/ and include the crash backtrace. Stack dump:

  1. Program arguments: tpuc-opt fsr_origin.mlir --init --shape-infer --canonicalize --extra-optimize --deinit --mlir-print-debuginfo -o fsr.mlir Stack dump without symbol names (ensure you have llvm-symbolizer in your PATH or set the environment var LLVM_SYMBOLIZER_PATH to point to it):

0 tpuc-opt 0x00005654ad126dea 1 tpuc-opt 0x00005654ad1240ec 2 libpthread.so.0 0x00007f5823b3f980 3 libc.so.6 0x00007f5822c0ce87 gsignal + 199 4 libc.so.6 0x00007f5822c0e7f1 abort + 321 5 libc.so.6 0x00007f5822bfe3fa 6 libc.so.6 0x00007f5822bfe472 7 tpuc-opt 0x00005654acd10774 8 tpuc-opt 0x00005654accab0c7 9 tpuc-opt 0x00005654acd9dfb4 10 tpuc-opt 0x00005654ac4a9e7c 11 tpuc-opt 0x00005654acda01c1 12 tpuc-opt 0x00005654ace8076e 13 tpuc-opt 0x00005654ace81658 14 tpuc-opt 0x00005654ace81c9f 15 tpuc-opt 0x00005654ac444e45 16 tpuc-opt 0x00005654ac445a5b 17 tpuc-opt 0x00005654ac445c42 18 tpuc-opt 0x00005654ad0a46ef 19 tpuc-opt 0x00005654ac43eb72 20 tpuc-opt 0x00005654ac445ff7 21 tpuc-opt 0x00005654ac42f6e0 22 libc.so.6 0x00007f5822befc87 __libc_start_main + 231 23 tpuc-opt 0x00005654ac43c26a

Aborted (core dumped) Traceback (most recent call last): File "/workspace/python/tools/model_transform.py", line 273, in tool.model_transform(args.mlir, args.add_postprocess) File "/workspace/python/tools/model_transform.py", line 51, in model_transform mlir_opt_for_top(mlir_origin, self.mlir_file, add_postprocess) File "/workspace/python/utils/mlir_shell.py", line 60, in mlir_opt_for_top _os_system(cmd) File "/workspace/python/utils/mlir_shell.py", line 50, in _os_system raise RuntimeError("[!Error]: {}".format(cmd_str)) RuntimeError: [!Error]: tpuc-opt fsr_origin.mlir --shape-infer --canonicalize --extra-optimize -o fsr.mlir

CHzhangi commented 11 months ago

system: Ubuntu 18.04

llvmbot commented 11 months ago

@llvm/issue-subscribers-mlir

Author: ZhangCH (CHzhangi)

root@4ee3c0dff15b:/workspace# model_transform.py --model_name fsr --input_shape [[1,3,200,200]] --model_def FSRCNN-pytorch-master/fsrcnn.pt --mlir fsr.mlir SOPHGO Toolchain v1.3.116-gf735aff3-20230821 2023/11/26 17:36:44 - INFO : _____________________________________________________ | preprocess: | | (x - mean) * scale | '-------------------------------------------------------' config Preprocess args : resize_dims : same to net input dims keep_aspect_ratio : False keep_ratio_mode : letterbox pad_value : 0 pad_type : center -------------------------- mean : [0.0, 0.0, 0.0] scale : [1.0, 1.0, 1.0] -------------------------- pixel_format : bgr channel_format : nchw Save mlir file: fsr_origin.mlir [Running]: tpuc-opt fsr_origin.mlir --shape-infer --canonicalize --extra-optimize -o fsr.mlir tpuc-opt: ../lib/Dialect/Top/Interfaces/Conv.cpp:137: void tpu_mlir::top::ConvOp::shape_inference(): Assertion `input_shape.size() == filter_shape.size()' failed. PLEASE submit a bug report to https://github.com/llvm/llvm-project/issues/ and include the crash backtrace. Stack dump: 0. Program arguments: tpuc-opt fsr_origin.mlir --init --shape-infer --canonicalize --extra-optimize --deinit --mlir-print-debuginfo -o fsr.mlir Stack dump without symbol names (ensure you have llvm-symbolizer in your PATH or set the environment var `LLVM_SYMBOLIZER_PATH` to point to it): 0 tpuc-opt 0x00005654ad126dea 1 tpuc-opt 0x00005654ad1240ec 2 libpthread.so.0 0x00007f5823b3f980 3 libc.so.6 0x00007f5822c0ce87 gsignal + 199 4 libc.so.6 0x00007f5822c0e7f1 abort + 321 5 libc.so.6 0x00007f5822bfe3fa 6 libc.so.6 0x00007f5822bfe472 7 tpuc-opt 0x00005654acd10774 8 tpuc-opt 0x00005654accab0c7 9 tpuc-opt 0x00005654acd9dfb4 10 tpuc-opt 0x00005654ac4a9e7c 11 tpuc-opt 0x00005654acda01c1 12 tpuc-opt 0x00005654ace8076e 13 tpuc-opt 0x00005654ace81658 14 tpuc-opt 0x00005654ace81c9f 15 tpuc-opt 0x00005654ac444e45 16 tpuc-opt 0x00005654ac445a5b 17 tpuc-opt 0x00005654ac445c42 18 tpuc-opt 0x00005654ad0a46ef 19 tpuc-opt 0x00005654ac43eb72 20 tpuc-opt 0x00005654ac445ff7 21 tpuc-opt 0x00005654ac42f6e0 22 libc.so.6 0x00007f5822befc87 __libc_start_main + 231 23 tpuc-opt 0x00005654ac43c26a Aborted (core dumped) Traceback (most recent call last): File "/workspace/python/tools/model_transform.py", line 273, in <module> tool.model_transform(args.mlir, args.add_postprocess) File "/workspace/python/tools/model_transform.py", line 51, in model_transform mlir_opt_for_top(mlir_origin, self.mlir_file, add_postprocess) File "/workspace/python/utils/mlir_shell.py", line 60, in mlir_opt_for_top _os_system(cmd) File "/workspace/python/utils/mlir_shell.py", line 50, in _os_system raise RuntimeError("[!Error]: {}".format(cmd_str)) RuntimeError: [!Error]: tpuc-opt fsr_origin.mlir --shape-infer --canonicalize --extra-optimize -o fsr.mlir
joker-eph commented 11 months ago

tpuc-opt isn't something that exists in LLVM/MLIR. This looks like a downstream project bug.