<table border="1" cellspacing="0" cellpadding="8">
    <tr>
        <th>Issue</th>
        <td>
            <a href=https://github.com/llvm/llvm-project/issues/73434>73434</a>
        </td>
    </tr>

    <tr>
        <th>Summary</th>
        <td>
            Aborted (core dumped)
        </td>
    </tr>

    <tr>
      <th>Labels</th>
      <td>
            new issue
      </td>
    </tr>

    <tr>
      <th>Assignees</th>
      <td>
      </td>
    </tr>

    <tr>
      <th>Reporter</th>
      <td>
          CHzhangi
      </td>
    </tr>
</table>

<pre>
    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 

</pre>
<img width="1px" height="1px" alt="" src="http://email.email.llvm.org/o/eJy0V11v2zzS_TX0zcACRerDuvCFY8dvC_RpgqZ4gd0bgaJGNrcSKZBUmvTXLyjJiZ2kyW7xrC78IZ2ZOZwZDo-Ec-qgEdckvSLpbiEGfzR2vf306yj0QS0qUz-urTGeJDRB5JLWTROnFeEbwvY_jf3heiGRMA6dqbEtvRXaNcZ2Uf8IsFxOd7XoEBpnwx2l-8GX7ih6hDHqVUzYlhO2ZZTOn-mOpLsn6xob2N992379uuwfvbHyuOyE82gJ2zfOSq2j3o_BWmVDmCj8IHRH6Obu5vbT_93Ad2NaeRRKw30c8SiOs-WhyXkqmoYvGWWcrlg8WYR_hO3jmLA9yyDOCd_wjPBNksASPn_d3wDhG5jAhBZQ_sn1bE_yLfQWe2skOje6zrcXTwEIWz3AEjoUmrACCNuAk6LFSyjLl392EZZPXgCk0Y06wO0TIRD24C5XbNGpX1jWqnPwfAWIC3X2BjR6GOsMAfRk-AOxL4XrUfrSCq_Mk-FetA4vcSOgDB1wFqBF79FW5uEJ24u6vBftgAAvyNALjH_sLyEjRqIOfXQCvpOiEyTUAF5dY37SKxqFDoazr3T3ZDlV7DeW8WRy9nVm-R_Q6tUDtmXYeMJfOq8Oz-uTR6H1K1xAaXn8eYK9_Xkn7hGmHaZaDDaNs6Wx6qD02YYj6dW3QWulD2EFfAO-H-TS9P4lHJbLcQgslW4w_JNCG62kaNUvhOUSH7wVwVB14w3ztLPnTjw5DkGiiLB9qyrC9jslWpSesP130xO2_xwq3AiJjrD91uj7SPY94ZuYh50N90bVgWM5LoFvCN94008_Avpm_j1yLUeuqMPEWxFWBAcb59B6ZTSQjJ4NtyhskgkGhO8I34XEebSvHhOWQyNUi3U0rez2y_Xm7hrcUHXKg4BqOIDF3lgfNtfR-95NA5iw_UH541BF0nQhA-396WvZW_OvKQ_KuWFcPQhdg9KyHWoEf0SQVrgjVEL-8FZInMPfeSF_QD1049LHWzQitLi15mBFFybC0KH27oPyKq38f1vlZY2zWXCy7K3SflljNRyUbsx5F7zkCj-VP5rBg3vsKtNCOHNcmJyo3WARHs0Ax9DDY3YmkPqFFpQOzyzcbr5_AmPBoR-Tg_peWaPDSuFe2FDeL1_-_6_y7h9_Xd18-fzP629lMCEZDVXpjdJjeZSfGuN871B4zhN9oJTSNEsTUccsq1FMmPgMM1-X0ISinI8oaFXV-6NFUUfORHRG5k26YrziTbGahx-HAJUBlF04HaFMUomrHA5BBIgWCLuCuCgm0-QjU8ybGEQVujIY8tMBmr5rVzXIm3nJ2XmIF6AkZxMofzcvso5pnicTdPV2mqUUFZXzEVd84K6om2p2F9N3oYkoMJ8rEsdvR64FjeWcl5i96w5XNM_mMzDmb7vDVZylqxmTfOAulkUzQ9P3F5IkmKQzNHs7cpKkIq1mTP6Bu1Qmc_Hi1duNT0WS4Yld8b47jtWpF9gHFUnSppnrzH5TkYQ1Gc67g7H3WxUbucqhLAOodF5YX3ZBQ4Z2Z_ykF_kH9CXLxOvjdBM2DtZhQEljcRxhWIfRMT7-HuZxGMwB0BnnwWIQKyBF20Ir3MWUAdiroAYZu1Tl-_7RH40mbO-NacMR8Fqlj0ZbaJVGYHnQ4WEiEr7tTD2Eg_56jhFcRC_sCVsFfTgNZLYdxWIk6rrsjfMnSXta09_FMo1nki_JzFFglCml6X0QOmU4ztlqujWeT8HaYduMrMtRzQTqf8x68GpiHdy5I7btC8IZPRF-yeuJcWlc6R6dx5BS2dV_Z_j0FP4syOTbCuUQvg3aqw6vrTV2VCQsvJexeLoxqTiSX5F8RxiLJu04sSydt6OAmcleeJqE7Qs__zM1uKjXvC54IRa4jnNKiyJOE7o4rmWcZrRO8xWVqRAJRcpZJpuMY5KuRFMt1Dq88sUxy2jBC0ojXqdFQleikbiq8jQlCcVOqDYKyiEy9rAYFdU65wlPFq2osHXjCzRjGn_C-HBM4m5h16PaqIaDIwltlfPu2YtXvsX1bwfBYrDt-s_F3sju3wEAAP__izSp9A">