Hardware path initialization step:Hardware path returns error 501 during job initialization

[Dazui-Wang]

I have successfully built QPL. In the qpl/examples/low-level-api directory, both software_path and auto_path work correctly. However, when I run the hardware_path, the following error occurs:

Intel(R) Query Processing Library version is 1.8.0.
The example will be run on the hardware path.
An error 501 acquired during job initializing.

Here is the output of accel-config list :

[
  {
    "dev":"iax1",
    "event_log_size":64,
    "max_groups":4,
    "max_work_queues":16,
    "max_engines":8,
    "work_queue_size":256,
    "numa_node":0,
    "op_cap":"00000000,00000000,00000000,00000000,00000000,004d001c,00000000,00000405",
    "iaa_cap":"0x1bf",
    "gen_cap":"0x171f20901f5315",
    "version":"0x200",
    "state":"enabled",
    "configurable":1,
    "pasid_enabled":1,
    "cdev_major":241,
    "clients":0,
    "groups":[
      {
        "dev":"group1.0",
        "desc_progress_limit":0,
        "batch_progress_limit":0,
        "grouped_workqueues":[
          {
            "dev":"wq1.0",
            "mode":"shared",
            "size":128,
            "group_id":0,
            "priority":10,
            "block_on_fault":1,
            "cdev_minor":0,
            "type":"user",
            "name":"app1",
            "threshold":128,
            "prs_disable":0,
            "ats_disable":0,
            "op_config":"00000000,00000000,00000000,00000000,00000000,004d001c,00000000,00000405",
            "state":"enabled",
            "clients":0
          }
        ],
        "grouped_engines":[
          {
            "dev":"engine1.0",
            "group_id":0
          },
          {
            "dev":"engine1.1",
            "group_id":0
          },
          {
            "dev":"engine1.2",
            "group_id":0
          },
          {
            "dev":"engine1.3",
            "group_id":0
          },
          {
            "dev":"engine1.4",
            "group_id":0
          },
          {
            "dev":"engine1.5",
            "group_id":0
          },
          {
            "dev":"engine1.6",
            "group_id":0
          },
          {
            "dev":"engine1.7",
            "group_id":0
          }
        ]
      },
      {
        "dev":"group1.1",
        "desc_progress_limit":0,
        "batch_progress_limit":0
      },
      {
        "dev":"group1.2",
        "desc_progress_limit":0,
        "batch_progress_limit":0
      },
      {
        "dev":"group1.3",
        "desc_progress_limit":0,
        "batch_progress_limit":0
      }
    ]
  }
]

Following the suggestion in this issue: #28, I modified compression_example.cpp to add the line:
job->numa_id = 0;

I then rebuilt the compression_example binary. However, running the hardware_path still results in the same error (501).

What further changes or configurations should I apply to get the hardware path working correctly?

[Dazui-Wang]

Update: After referring to the documentation(https://intel.github.io/qpl/documentation/dev_ref_docs/c_ref/c_status_codes.html#complete-list-of-status-codes), I configured the LD_LIBRARY_PATH. Currently, I am encountering error code 503.

[Dazui-Wang]

Update: After using "su root" here, the 503 error has also been resolved. However, strangely enough: Previously, I was already running as the root account, but it failed (with code 503). However, after executing "su root" on the root account, it executed successfully.

I am currently conducting benchmark and I have a new question to ask: Is the file "iaa_throughput_dataset" editable by ourselves?or Will it be automatically generated during the compilation of iaa?