[pcied] Honour PCIE_DETACH_INFO on platforms that are not smartswitches

[gordon-nexthop]

Description

Fixes: #900

Drop the is_smartswitch() gate on PCIE_DETACH_INFO handling, read a generic
device_state field (falling back to dpu_state) to allow non smartswitch platforms
to report when a pcie device has been intentionally brought down.

Motivation and Context

A PCIe device can be powered off/cycled, not just a smartswitch, so its endpoints
stop answering and pcied reports them missing for the duration of a the power off it
was asked to perform.

The support for this already exists in the case of DPUs on smartswitches #546
The table is already device-shaped rather than DPU-shaped — it is keyed by BDF,
ModuleBase.handle_pci_removal() already fans one module out to a list of
buses, and the value detaching says nothing about DPUs. Only the dpu_state
field name carried the DPU assumption, and writing that field on a pizzabox
would describe nothing. Reading device_state first keeps
ModuleBase.pci_entry_state_db() working unchanged.

How Has This Been Tested?

Extended test_is_device_in_detaching_mode to cover both field names, and added

• test_detaching_is_honoured_on_a_non_smartswitch — a claimed BDF is no longer
  warned about on a non-smartswitch platform (the behaviour this PR adds).

Additional Information (Optional)

[azure-pipelines]

Azure Pipelines:
There may be pipelines that require an authorized user to comment /azp run to run.

[mssonicbld]

/azp run

[azure-pipelines]

Azure Pipelines:
Successfully started running 1 pipeline(s).

[azure-pipelines]

Azure Pipelines:
There may be pipelines that require an authorized user to comment /azp run to run.

[mssonicbld]

Hi, there are workflow run(s) waiting for approval, you may be first-time contributor. I will notify maintainers to help approve once PR is approved. Thanks!

_{---Powered by SONiC BuildBot}