Fix too relaxed check on CUDA "fast copy" (can_be_transposed) condition

[pwilkin]

See #16095 (comment) for case.

[CISC]

Add testcase or it didn't happen. :)

[pwilkin]

Add testcase or it didn't happen. :)

Look, children, that's how an evil maintainer looks like :P Will never let you off the hook with any PR, ever!

[pwilkin]

update_cuda_graph_executable: CUDA graph update failed
ggml_backend_cuda_graph_compute: disabling CUDA graphs due to too many consecutive updates
update_cuda_graph_executable: CUDA graph update failed
  CONT(type=f32,ne=[10,10,10,1]): OK
  CONT(type=f32,ne=[2,1,1,1]): OK
  CONT(type=f32,ne=[2,1,3,5]): OK
  CONT(type=f32,ne=[2,3,5,7]): OK
  CONT(type=f16,ne=[2,1,1,1]): OK
  CONT(type=f16,ne=[2,1,3,5]): OK
  CONT(type=f16,ne=[2,3,5,7]): OK
  CONT(type=bf16,ne=[2,1,1,1]): OK
  CONT(type=bf16,ne=[2,1,3,5]): OK
  CONT(type=bf16,ne=[2,3,5,7]): OK
[CONT] NMSE = 0.447623183 > 0.000000100   CONT(type=f32,ne=[1,4,2,1]): FAIL
[CONT] NMSE = 2.241813873 > 0.000000100   CONT(type=f32,ne=[1,8,17,1]): FAIL
[CONT] NMSE = 0.058848433 > 0.000000100   CONT(type=bf16,ne=[1,4,2,1]): FAIL
[CONT] NMSE = 1.181509486 > 0.000000100   CONT(type=bf16,ne=[1,8,17,1]): FAIL
  10/14 tests passed

vs

  update_cuda_graph_executable: CUDA graph update failed
ggml_backend_cuda_graph_compute: disabling CUDA graphs due to too many consecutive updates
update_cuda_graph_executable: CUDA graph update failed
  CONT(type=f32,ne=[10,10,10,1]): OK
  CONT(type=f32,ne=[2,1,1,1]): OK
  CONT(type=f32,ne=[2,1,3,5]): OK
  CONT(type=f32,ne=[2,3,5,7]): OK
  CONT(type=f16,ne=[2,1,1,1]): OK
  CONT(type=f16,ne=[2,1,3,5]): OK
  CONT(type=f16,ne=[2,3,5,7]): OK
  CONT(type=bf16,ne=[2,1,1,1]): OK
  CONT(type=bf16,ne=[2,1,3,5]): OK
  CONT(type=bf16,ne=[2,3,5,7]): OK
  CONT(type=f32,ne=[1,4,2,1]): OK
  CONT(type=f32,ne=[1,8,17,1]): OK
  CONT(type=bf16,ne=[1,4,2,1]): OK
  CONT(type=bf16,ne=[1,8,17,1]): OK
  14/14 tests passed
  Backend CUDA0: OK

@CISC There :)

[JohannesGaessler]

Thank you for the fix and sorry for not catching the bug during review; some functionality that would have covered this was removed and the logic was not adjusted.

Preferably add an argument for the existing tests for GGML_OP_CONT rather than adding a new test case.