* [CustomDevice] unittests tracking the fixed version plugin

* update

* [CustomDevice] add profiler apis

* migrate CalculateEstOccupancy into cuda_tracer

* update

* add ut

[CodeStyle] use np.testing.assert_array_equal instead of self.assertTrue(np.array_equal(...)) (#44947)

* automatically fix

* update comments

* numpy -> np

* self.assertEqual(..., True)

* wrong usage (err_msg=True)

这不是修复导致的错误，这些是原来 `self.assertTrue(..., True)`
的错误用法，因此在修复后将其认为位置参数 `err_msg`

* some missing fix

* [CustomDevice] add custom ccl 2/2

* update

* update

* update launch

* add FLAGS_enable_api_kernel_fallback

* deal with more cases

* add ut for coverage

* [CustomDevice] add process_group_xccl ut

* update

* [CustomDevice] register Copy for custom device

* [CustomDevice] register Copy for custom device

* [CustomDevice] register Copy for custom device

* merge and add uts

* merge and add uts

* fix for blocking and unittests coverage

* [Plugin] Fix Custom device in eager mode, test=develop

* update test case, test=develop

* update ut for coverage, test=develop

* [CustomKernel] add capi eager mode support

* add ut

* add capi test

* Dataloader add custom device support

* update test=document_fix