Fix Debug build break with device_uvectors in grouped_rolling.cu

[mythrocks]

After #7523, on Debug builds, one seems to see the following build error:

cudf/cpp/src/rolling/grouped_rolling.cu(130): error: no operator "[]" matches these operands
            operand types are: const cudf::groupby::detail::sort::sort_groupby_helper::index_vector [ int ]
...
cudf/cpp/src/rolling/grouped_rolling.cu(130): error: no operator "[]" matches these operands
            operand types are: const cudf::groupby::detail::sort::sort_groupby_helper::index_vector [ unsigned long ] 

The offending line has an assert() that is only compiled during Debug builds. This commit corrects the indexing into device_uvector.

[mythrocks]

Alternatively, we might just remove the assert().

[codecov]

Codecov Report

Merging #7633 (d5c8212) into branch-0.19 (7871e7a) will increase coverage by 0.60%.
The diff coverage is n/a.

@@               Coverage Diff               @@
##           branch-0.19    #7633      +/-   ##
===============================================
+ Coverage        81.86%   82.47%   +0.60%     
===============================================
  Files              101      101              
  Lines            16884    17397     +513     
===============================================
+ Hits             13822    14348     +526     
+ Misses            3062     3049      -13     

Impacted Files	Coverage Δ
python/cudf/cudf/core/column/categorical.py	91.97% <ø> (+0.58%)	⬆️
python/cudf/cudf/core/column/column.py	87.86% <ø> (+0.10%)	⬆️
python/cudf/cudf/core/column/datetime.py	89.63% <ø> (+0.54%)	⬆️
python/cudf/cudf/core/column/decimal.py	92.75% <ø> (-2.12%)	⬇️
python/cudf/cudf/core/column/lists.py	92.17% <ø> (+0.77%)	⬆️
python/cudf/cudf/core/column/numerical.py	94.83% <ø> (-0.20%)	⬇️
python/cudf/cudf/core/column/string.py	86.79% <ø> (+0.30%)	⬆️
python/cudf/cudf/core/column/timedelta.py	88.57% <ø> (+0.33%)	⬆️
python/cudf/cudf/core/column_accessor.py	95.45% <ø> (+0.14%)	⬆️
python/cudf/cudf/core/dataframe.py	90.90% <ø> (+0.44%)	⬆️
... and 64 more

---

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 bb05ddc...d5c8212. Read the comment docs.

[devavret]

Not a part of this PR but can you please add the stream to the group_offsets() and group_labels()

[mythrocks]

Not a part of this PR but can you please add the stream to the group_offsets() and group_labels()

Done. Thanks for pointing that out. TIL.

[davidwendt]

@mythrocks Is this good to merge?

[mythrocks]

Yikes. Yes, this should be merged. Or should have been. I'd better leave this till I've rebased and retested, in light of recent changes in the code base.

[mythrocks]

Rebased and retested. All tests are passing. I'll merge this now.

[mythrocks]

@gpucibot merge