[CODEGEN][METAL] Fix ramp codegen

[tqchen]

Fix ramp node codegen for the metal backend.
The default C codegen can cause problem in
vector indices assignment.

[tvm-bot]

Thanks for contributing to TVM! Please refer to the contributing guidelines https://tvm.apache.org/docs/contribute/ for useful information and tips. Please request code reviews from Reviewers by @-ing them in a comment.

• cc @echuraev _{See #10317 for details}

_{Generated by tvm-bot}

[tqchen]

currently ci do not support metal, confirmed on apple M2 for some cases that works on cuda,

[MasterJH5574]

Thanks for the fix!

[echuraev]

The PR LGTM. And I don't want to block this PR by the comment, but I have one question: will it lake a long time to add a test for this change? If not, then I could run this test on my x86 Mac and check if it works here as well.
I don't think that it should take a lot of time for adding such test, because metal codegen test exists in the TVM.

[junrushao]

Agree with @echuraev that we might need a unittest. @tqchen @MasterJH5574 would you like to follow up with a new PR? Thanks a lot!

[tqchen]

Thanks @echuraev . we will send in a test in a followup

[tqchen]

#14331