Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Implement mixed equality/conditional joins #9917

Merged
merged 56 commits into from
Jan 18, 2022
Merged

Implement mixed equality/conditional joins #9917

merged 56 commits into from
Jan 18, 2022

Commits on Dec 13, 2021

  1. Add APIs for mixed equality/inequality joins.

    @vyasr
    vyasr committed Dec 13, 2021
    Configuration menu
    Copy the full SHA
    c2ba04d View commit details
    Browse the repository at this point in the history

  2. Add new implementations for mixed joins.

    @vyasr
    vyasr committed Dec 13, 2021
    Configuration menu
    Copy the full SHA
    c1eec27 View commit details
    Browse the repository at this point in the history

  3. Move mixed joins to separate file to avoid introducing hash join rela… 

    …ted overhead to the purely conditional join code paths.
    @vyasr
    vyasr committed Dec 13, 2021
    Configuration menu
    Copy the full SHA
    bcc2c7b View commit details
    Browse the repository at this point in the history

  4. Start implementing map logic in new join, add notes.

    @vyasr
    vyasr committed Dec 13, 2021
    Configuration menu
    Copy the full SHA
    06c26ea View commit details
    Browse the repository at this point in the history

  5. Create new kernels for mixed joins.

    @vyasr
    vyasr committed Dec 13, 2021
    Configuration menu
    Copy the full SHA
    061c74c View commit details
    Browse the repository at this point in the history

  6. Pass hash table device view to kernel.

    @vyasr
    vyasr committed Dec 13, 2021
    Configuration menu
    Copy the full SHA
    db11c2b View commit details
    Browse the repository at this point in the history

  7. Revert change to build_join_hash_table allowing empty tables and filt… 

    …er ahead of time.
    @vyasr
    vyasr committed Dec 13, 2021
    Configuration menu
    Copy the full SHA
    342a8d1 View commit details
    Browse the repository at this point in the history

  8. Version of mixed kernel with correct logic, but currently requires to… 

    …o much shared memory.
    @vyasr
    vyasr committed Dec 13, 2021
    Configuration menu
    Copy the full SHA
    1fdce0f View commit details
    Browse the repository at this point in the history

  9. Fix some bugs and reduce shared memory usage to the point of compilat… 

    …ion.
    @vyasr
    vyasr committed Dec 13, 2021
    Configuration menu
    Copy the full SHA
    9b173b0 View commit details
    Browse the repository at this point in the history

  10. Create new array to hold per-row matches and define new comparator fo… 

    …r use in device code.
    @vyasr
    vyasr committed Dec 13, 2021
    Configuration menu
    Copy the full SHA
    a77fbaf View commit details
    Browse the repository at this point in the history

  11. Compiling version of new code that computes per-row matches.

    @vyasr
    vyasr committed Dec 13, 2021
    Configuration menu
    Copy the full SHA
    13c30d2 View commit details
    Browse the repository at this point in the history

  12. Change size APIs to return per-row counts and verify results.

    @vyasr
    vyasr committed Dec 13, 2021
    Configuration menu
    Copy the full SHA
    9c3c3a7 View commit details
    Browse the repository at this point in the history

  13. Change mixed join APIs to accept the offsets.

    @vyasr
    vyasr committed Dec 13, 2021
    Configuration menu
    Copy the full SHA
    ae446db View commit details
    Browse the repository at this point in the history

  14. First attempt at using offsets column in join.

    @vyasr
    vyasr committed Dec 13, 2021
    Configuration menu
    Copy the full SHA
    df173af View commit details
    Browse the repository at this point in the history

  15. Fix seg fault, now just a buggy impl.

    @vyasr
    vyasr committed Dec 13, 2021
    Configuration menu
    Copy the full SHA
    5b8bf2d View commit details
    Browse the repository at this point in the history

  16. Fix bugs, implementation now works as expected for simple test.

    @vyasr
    vyasr committed Dec 13, 2021
    Configuration menu
    Copy the full SHA
    636a604 View commit details
    Browse the repository at this point in the history

  17. Fix a bug and add one more extensive test.

    @vyasr
    vyasr committed Dec 13, 2021
    Configuration menu
    Copy the full SHA
    cb14a53 View commit details
    Browse the repository at this point in the history

  18. Start addressing TODOs.

    @vyasr
    vyasr committed Dec 13, 2021
    Configuration menu
    Copy the full SHA
    890fd6f View commit details
    Browse the repository at this point in the history

Commits on Dec 14, 2021

  1. Address more TODOs.

    @vyasr
    vyasr committed Dec 14, 2021
    Configuration menu
    Copy the full SHA
    e693960 View commit details
    Browse the repository at this point in the history

  2. Start setting up scaffolding for more systematic testing.

    @vyasr
    vyasr committed Dec 14, 2021
    Configuration menu
    Copy the full SHA
    6b39bdd View commit details
    Browse the repository at this point in the history

  3. Change APIs to accept separate table views for the hash join.

    @vyasr
    vyasr committed Dec 14, 2021
    Configuration menu
    Copy the full SHA
    444a140 View commit details
    Browse the repository at this point in the history

  4. Add compare_nulls parameter and various additional internal error che… 

    …cks.
    @vyasr
    vyasr committed Dec 14, 2021
    Configuration menu
    Copy the full SHA
    5515daa View commit details
    Browse the repository at this point in the history

  5. Update tests to only use the subset of necessary columns for the cond… 

    …itional join.
    @vyasr
    vyasr committed Dec 14, 2021
    Configuration menu
    Copy the full SHA
    40ea881 View commit details
    Browse the repository at this point in the history

  6. Move equality tables before conditional tables.

    @vyasr
    vyasr committed Dec 14, 2021
    Configuration menu
    Copy the full SHA
    9e54706 View commit details
    Browse the repository at this point in the history

  7. Plumb compare_nulls all the way through.

    @vyasr
    vyasr committed Dec 14, 2021
    Configuration menu
    Copy the full SHA
    2b2c6cc View commit details
    Browse the repository at this point in the history

  8. Start defining helper functions for generalized mixed join testing.

    @vyasr
    vyasr committed Dec 14, 2021
    Configuration menu
    Copy the full SHA
    273be21 View commit details
    Browse the repository at this point in the history

  9. Move mixed join tests to separate TU.

    @vyasr
    vyasr committed Dec 14, 2021
    Configuration menu
    Copy the full SHA
    c34de6a View commit details
    Browse the repository at this point in the history

Commits on Dec 15, 2021

  1. Define standard testing function for mixed inner joins.

    @vyasr
    vyasr committed Dec 15, 2021
    Configuration menu
    Copy the full SHA
    473efdb View commit details
    Browse the repository at this point in the history

  2. Update to use new cuco API with 4 iterators instead of two.

    @vyasr
    vyasr committed Dec 15, 2021
    Configuration menu
    Copy the full SHA
    bd87a3d View commit details
    Browse the repository at this point in the history

  3. Add proper support for non-inner joins.

    @vyasr
    vyasr committed Dec 15, 2021
    Configuration menu
    Copy the full SHA
    c5f8f9a View commit details
    Browse the repository at this point in the history

  4. Add left join APIs and tests.

    @vyasr
    vyasr committed Dec 15, 2021
    Configuration menu
    Copy the full SHA
    e11e84e View commit details
    Browse the repository at this point in the history

  5. Add full join APIs and tests.

    @vyasr
    vyasr committed Dec 15, 2021
    Configuration menu
    Copy the full SHA
    07b48c1 View commit details
    Browse the repository at this point in the history

  6. Add tests of and fix bug with asymmetric joins.

    @vyasr
    vyasr committed Dec 15, 2021
    Configuration menu
    Copy the full SHA
    c74af29 View commit details
    Browse the repository at this point in the history

  7. Clean up some TODOs.

    @vyasr
    vyasr committed Dec 15, 2021
    Configuration menu
    Copy the full SHA
    8a4df8e View commit details
    Browse the repository at this point in the history

  8. Add tests for and fix null support.

    @vyasr
    vyasr committed Dec 15, 2021
    Configuration menu
    Copy the full SHA
    f9739fc View commit details
    Browse the repository at this point in the history

Commits on Dec 16, 2021

  1. Rewrite docstrings.

    @vyasr
    vyasr committed Dec 16, 2021
    Configuration menu
    Copy the full SHA
    80a268d View commit details
    Browse the repository at this point in the history

  2. Change size APIs to return columns instead of device_uvectors.

    @vyasr
    vyasr committed Dec 16, 2021
    Configuration menu
    Copy the full SHA
    b240cc5 View commit details
    Browse the repository at this point in the history

  3. Change join APIs to use column_views instead of device_spans.

    @vyasr
    vyasr committed Dec 16, 2021
    Configuration menu
    Copy the full SHA
    2bb25a3 View commit details
    Browse the repository at this point in the history

  4. Merge remote-tracking branch 'origin/branch-22.02' into feature/mixed… 

    …_equality_conditional_joins
    @vyasr
    vyasr committed Dec 16, 2021
    Configuration menu
    Copy the full SHA
    5f49458 View commit details
    Browse the repository at this point in the history

  5. Update cuco.

    @vyasr
    vyasr committed Dec 16, 2021
    Configuration menu
    Copy the full SHA
    d724647 View commit details
    Browse the repository at this point in the history

  6. Fix unnecessary extra line.

    @vyasr
    vyasr committed Dec 16, 2021
    Configuration menu
    Copy the full SHA
    ec5d601 View commit details
    Browse the repository at this point in the history

  7. Apply suggestions from code review 

    Co-authored-by: Yunsong Wang <wangyunsong89@gmail.com>
    @vyasr @PointKernel
    vyasr and PointKernel authored Dec 16, 2021
    Configuration menu
    Copy the full SHA
    15a6d52 View commit details
    Browse the repository at this point in the history

Commits on Dec 18, 2021

  1. Fix swapping logic in operator in conjunction with proper handling of… 

    … cuco's arguments.
    @vyasr
    vyasr committed Dec 18, 2021
    Configuration menu
    Copy the full SHA
    1dc63d8 View commit details
    Browse the repository at this point in the history

  2. Simplify operator logic and add comments.

    @vyasr
    vyasr committed Dec 18, 2021
    Configuration menu
    Copy the full SHA
    78dd48d View commit details
    Browse the repository at this point in the history

Commits on Dec 21, 2021

  1. Merge remote-tracking branch 'origin/branch-22.02' into feature/mixed… 

    …_equality_conditional_joins
    @vyasr
    vyasr committed Dec 21, 2021
    Configuration menu
    Copy the full SHA
    9041b99 View commit details
    Browse the repository at this point in the history

  2. Fix style.

    @vyasr
    vyasr committed Dec 21, 2021
    Configuration menu
    Copy the full SHA
    5726c5e View commit details
    Browse the repository at this point in the history

Commits on Dec 22, 2021

  1. Remove addressed TODOs.

    @vyasr
    vyasr committed Dec 22, 2021
    Configuration menu
    Copy the full SHA
    ecb77ca View commit details
    Browse the repository at this point in the history

Commits on Jan 6, 2022

  1. Merge remote-tracking branch 'origin/branch-22.02' into feature/mixed… 

    …_equality_conditional_joins
    @vyasr
    vyasr committed Jan 6, 2022
    Configuration menu
    Copy the full SHA
    a7f9d15 View commit details
    Browse the repository at this point in the history

  2. Address simple PR comments.

    @vyasr
    vyasr committed Jan 6, 2022
    Configuration menu
    Copy the full SHA
    b7accc3 View commit details
    Browse the repository at this point in the history

  3. Switch raw pointers to device spans for matches_per_row.

    @vyasr
    vyasr committed Jan 6, 2022
    Configuration menu
    Copy the full SHA
    0daaacd View commit details
    Browse the repository at this point in the history

  4. Remove unsupported full join size calculation.

    @vyasr
    vyasr committed Jan 6, 2022
    Configuration menu
    Copy the full SHA
    0915bff View commit details
    Browse the repository at this point in the history

Commits on Jan 11, 2022

  1. Merge remote-tracking branch 'origin/branch-22.02' into feature/mixed… 

    …_equality_conditional_joins
    @vyasr
    vyasr committed Jan 11, 2022
    Configuration menu
    Copy the full SHA
    d7ab5a7 View commit details
    Browse the repository at this point in the history

  2. Construct a row equality comparator on the host and pass it to the ke… 

    …rnel rather than inlining its logic in the pair_expression_equality comparator.
    @vyasr
    vyasr committed Jan 11, 2022
    Configuration menu
    Copy the full SHA
    0442cd3 View commit details
    Browse the repository at this point in the history

  3. Some minor cleanup.

    @vyasr
    vyasr committed Jan 11, 2022
    Configuration menu
    Copy the full SHA
    ad41f9d View commit details
    Browse the repository at this point in the history

  4. Change the offsets to an rmm::device_uvector<size_type> instead of a … 

    …cudf::column.
    @vyasr
    vyasr committed Jan 11, 2022
    Configuration menu
    Copy the full SHA
    4d4f4e6 View commit details
    Browse the repository at this point in the history

  5. Fix APIs to accept const spans instead of non-const ones.

    @vyasr
    vyasr committed Jan 11, 2022
    Configuration menu
    Copy the full SHA
    c0e2297 View commit details
    Browse the repository at this point in the history