Hip refactor for attention, batch, combine, cast, conv

[Bob-Chen222]

Description of changes:
Hip refactor for attention, batch, combine, cast, conv

Related Issues:

Linked Issues:

• Issue Hip kernel refactor #1296

Issues closed by this PR:

• Closes #

---

This change is 

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 38.10%. Comparing base (be2aad1) to head (a81d707).

Additional details and impacted files

@@              Coverage Diff               @@
##           repo-refactor    #1402   +/-   ##
==============================================
  Coverage          38.10%   38.10%           
==============================================
  Files                167      167           
  Lines               5026     5026           
  Branches             246      246           
==============================================
  Hits                1915     1915           
  Misses              3111     3111           

Flag	Coverage Δ
unittests	38.10% <ø> (ø)

Flags with carried forward coverage won't be shown. Click here to find out more.

[reyna-abhyankar]

Reviewable status: 0 of 7 files reviewed, 3 unresolved discussions (waiting on @Bob-Chen222)

---

lib/kernels/src/hip/attention_kernels.cpp line 242 at r1 (raw file):

                                        device_state.reserveSpaceSize,
                                        device_state.reserveSpace));
#endif

Delete

---

lib/kernels/src/hip/batch_norm_kernels.cpp line 119 at r1 (raw file):

  checkCUDNN(miopenCreateTensorDescriptor(&outputTensor));
  mode = miopenBNSpatial;
#if HIPDNN_VERSION >= 7000

Is this still true for HIP?

---

lib/kernels/src/hip/cast_kernels.cpp line 75 at r1 (raw file):

};

void forward_kernel(PerDeviceFFHandle handle,

Actually just keep stream as the first parameter for both functions. I'll change this in the cuda kernel as well.

[Bob-Chen222]

Reviewable status: 0 of 7 files reviewed, 3 unresolved discussions (waiting on @reyna-abhyankar)

---

lib/kernels/src/hip/attention_kernels.cpp line 242 at r1 (raw file):

Previously, reyna-abhyankar (Reyna Abhyankar) wrote…

Delete

Done.

---

lib/kernels/src/hip/batch_norm_kernels.cpp line 119 at r1 (raw file):

Previously, reyna-abhyankar (Reyna Abhyankar) wrote…

Is this still true for HIP?

Done.

---

lib/kernels/src/hip/cast_kernels.cpp line 75 at r1 (raw file):

Previously, reyna-abhyankar (Reyna Abhyankar) wrote…

Actually just keep stream as the first parameter for both functions. I'll change this in the cuda kernel as well.

Done.