[FEA]: Thrust vector's should provide `emplace_back(Args...) -> T&`

[bernhardmgruber]

Is this a duplicate?

• I confirmed there appear to be no duplicate issues for this request and that I agree to the Code of Conduct

Area

Thrust

Is your feature request related to a problem? Please describe.

  thrust::host_vector<thrust::device_vector<int>> out_buffers;
  c2h::gen(C2H_SEED(1), out_buffers.emplace_back(42));

should compile.

Describe the solution you'd like

Implement emplace_back(Args...) -> T& for thrust vectors

Describe alternatives you've considered

No response

Additional context

No response

[viralbhadeshiya]

@bernhardmgruber I like to work on this issue, if that's okay.

[viralbhadeshiya]

Hey @bernhardmgruber or anyone, I have this done, but I want to test it and I can't build_thrust from main without error. Do we have stable version in main which should successfully build thrust test?

[bernhardmgruber]

I have this done, but I want to test it and I can't build_thrust from main without error. Do we have stable version in main which should successfully build thrust test?

main should be stable since it passed our CI. Please open a PR with your changes so I can test them against the CI.

[miscco]

Are we really sure we want this?

AFAIK the reason we did not do this is because this allows people to use a massively inefficient API

[bernhardmgruber]

AFAIK the reason we did not do this is because this allows people to use a massively inefficient API

Only for device vectors, for which performance would be as bad as operator[], which we also provide. I see your point though.