Fix apply_gate and apply_channel for GPUs

[renatomello]

Requires qiboteam/qibo#1792

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 100.00%. Comparing base (d25ee78) to head (d69b355).
⚠️ Report is 10 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff            @@
##              main      #243   +/-   ##
=========================================
  Coverage   100.00%   100.00%           
=========================================
  Files            7         7           
  Lines          668       669    +1     
=========================================
+ Hits           668       669    +1     

Flag	Coverage Δ
unittests	100.00% <100.00%> (ø)

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

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[scarrazza]

Thanks @renatomello. My last commit fixes a series of errors of this type when running pytest on GPU:

E           cupy.cuda.compiler.CompileException: qibojit/env/lib/python3.12/site-packages/cupy/_core/include/cupy/_cccl/thrust/thrust/detail/config/cpp_dialect.h(74): catastrophic error: #error directive: Thrust requires at least C++17. Define CCCL_IGNORE_DEPRECATED_CPP_DIALECT to suppress this message.
E             #    error Thrust requires at least C++17. Define CCCL_IGNORE_DEPRECATED_CPP_DIALECT to suppress this message.
E                  ^
E
E           1 catastrophic error detected in the compilation of "/tmp/tmpoo1e14fs/41bab56b6448eeb8d40b98d19aedbe370eef70f1.cubin.cu".
E           Compilation terminated.****

[scarrazza]

The backend test which fails for numba-dia appears ramdomly on my local system. @renatomello could you please have a look?

[renatomello]

The backend test which fails for numba-dia appears ramdomly on my local system. @renatomello could you please have a look?

It's been happening for a while. It seems to be a scipy issue instead of a qibo issue. It is also not deterministic, so it is hard to reproduce. My best guess is that it happens whenever the sparse dia matrix has an empty diagonal, but anyone's guess is as good as mine.