added device call after the bcast in the getitem

[coquelin77]

Description

a minor bug was showing up where the bcast operation was not putting the data onto the proper GPU in the __getitem__ function. the code added may not be hit all the time, but it is safeguard for when this may happen.

Issue/s resolved: n/a

Changes proposed:

• move the local array to the proper GPU after the bcast when there are integers in the getitem key

Type of change

• Bug fix (non-breaking change which fixes an issue)

Due Diligence

• All split configurations tested
• Multiple dtypes tested in relevant functions
• Documentation updated (if needed)
• Updated changelog.md under the title "Pending Additions"

Does this change modify the behaviour of other functions? If so, which?

no

[codecov]

Codecov Report

Merging #790 (df6dd02) into master (184f112) will decrease coverage by 0.00%.
The diff coverage is 50.00%.

@@            Coverage Diff             @@
##           master     #790      +/-   ##
==========================================
- Coverage   90.85%   90.84%   -0.01%     
==========================================
  Files          64       64              
  Lines        8932     8934       +2     
==========================================
+ Hits         8115     8116       +1     
- Misses        817      818       +1     

Flag	Coverage Δ
unit	90.84% <50.00%> (-0.01%)	⬇️

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

Impacted Files	Coverage Δ
heat/core/dndarray.py	96.36% <50.00%> (-0.15%)	⬇️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 184f112...df6dd02. Read the comment docs.

[coquelin77]

ignore the diff hit please, im not sure why it only hits it on the master branch.

[ClaudiaComito]

Hey @coquelin77 I'm a bit lost, is #784 fixed? In which case, we should be able to switch back to Bcast, which will write in the previously allocated buffer on the correct device?

[coquelin77]

Hey @coquelin77 I'm a bit lost, is #784 fixed? In which case, we should be able to switch back to Bcast, which will write in the previously allocated buffer on the correct device?

i thought that it was fixed because it solved this problem, but it still caused the same issue that @ben-bou pointed out before. so i left it asbcast

[ClaudiaComito]

Thanks @coquelin77 , I'm OK with this, maybe just add a note that the fix won't (shouldn't) be necessary after solving #784?