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

[KT] Single Pass Copy_if Kernel Template #1616

Draft
wants to merge 136 commits into
base: main
Choose a base branch
from
Draft

[KT] Single Pass Copy_if Kernel Template #1616

wants to merge 136 commits into from

Commits on Nov 29, 2023

  1. Enable pragma unroll for open-source DPC++

    @joeatodd
    joeatodd committed Nov 29, 2023
    Configuration menu
    Copy the full SHA
    e5c121b View commit details
    Browse the repository at this point in the history

  2. clang-format

    @joeatodd
    joeatodd committed Nov 29, 2023
    Configuration menu
    Copy the full SHA
    7c1cb0f View commit details
    Browse the repository at this point in the history

  3. Start of single-pass scan kernel template

    @adamfidel @Alcpz
    adamfidel authored and Alcpz committed Nov 29, 2023
    Configuration menu
    Copy the full SHA
    154161f View commit details
    Browse the repository at this point in the history

  4. Fix hang in inclusive scan

    @adamfidel @Alcpz
    adamfidel authored and Alcpz committed Nov 29, 2023
    Configuration menu
    Copy the full SHA
    16ec5ad View commit details
    Browse the repository at this point in the history

  5. Debug statements for scan kernel template

    @adamfidel @Alcpz
    adamfidel authored and Alcpz committed Nov 29, 2023
    Configuration menu
    Copy the full SHA
    bd89601 View commit details
    Browse the repository at this point in the history

  6. Update scan kernel template test

    @adamfidel @Alcpz
    adamfidel authored and Alcpz committed Nov 29, 2023
    Configuration menu
    Copy the full SHA
    60a69fc View commit details
    Browse the repository at this point in the history

  7. Only have a single work-item per group query for previous tile status

    @adamfidel @Alcpz
    adamfidel authored and Alcpz committed Nov 29, 2023
    Configuration menu
    Copy the full SHA
    d526f04 View commit details
    Browse the repository at this point in the history

  8. First attempt at parallel lookback

    @adamfidel @Alcpz
    adamfidel authored and Alcpz committed Nov 29, 2023
    Configuration menu
    Copy the full SHA
    09e9bbf View commit details
    Browse the repository at this point in the history

  9. Working cooperative lookback

    @adamfidel @Alcpz
    adamfidel authored and Alcpz committed Nov 29, 2023
    Configuration menu
    Copy the full SHA
    30e0da7 View commit details
    Browse the repository at this point in the history

  10. Fix correctness issue with non-power-of-2 sizes

    @adamfidel @Alcpz
    adamfidel authored and Alcpz committed Nov 29, 2023
    Configuration menu
    Copy the full SHA
    2311929 View commit details
    Browse the repository at this point in the history

  11. Scan_kt Flags and Values separated (#15) 

    Atomic flags and the values used in Scan_kt separated to avoid truncating the range to 30bit values, and prepare for a more general scan implementation.
    @Alcpz
    Alcpz committed Nov 29, 2023
    Configuration menu
    Copy the full SHA
    0f58c07 View commit details
    Browse the repository at this point in the history

  12. Refactored Scan_kt code (#16) 

    * Improved Scan_kt: templated parameters, ballot, wgsize calculation.

- Changed number of workgroups calculation from next power of two to
next multiple of wgsize
- Improved group_ballot by using the class member functions
- Using kernel_param struct to determine wgsize and elems per work item.
    @Alcpz
    Alcpz committed Nov 29, 2023
    Configuration menu
    Copy the full SHA
    8af98d6 View commit details
    Browse the repository at this point in the history

  13. Scan_kt: Single memory allocation for device_memory (#17) and async f… 

    …ree of the device memory (#18)

* Single memory allocation for device_memory

* async free of device memory

---------

Co-authored-by: Joe Todd <[email protected]>
Co-authored-by: AidanBeltonS <[email protected]>
    @Alcpz @joeatodd @AidanBeltonS
    3 people committed Nov 29, 2023
    Configuration menu
    Copy the full SHA
    3de596e View commit details
    Browse the repository at this point in the history

  14. Replace sycl::range with sycl::nd_range for fill

    @joeatodd @Alcpz
    joeatodd authored and Alcpz committed Nov 29, 2023
    Configuration menu
    Copy the full SHA
    2d6ff78 View commit details
    Browse the repository at this point in the history

  15. Bug fix

    @joeatodd @Alcpz
    joeatodd authored and Alcpz committed Nov 29, 2023
    Configuration menu
    Copy the full SHA
    124a912 View commit details
    Browse the repository at this point in the history

  16. Global to local then perform op

    @Alcpz
    Aidan authored and Alcpz committed Nov 29, 2023
    Configuration menu
    Copy the full SHA
    d716bbd View commit details
    Browse the repository at this point in the history

  17. Update based on feedback

    @Alcpz
    Aidan authored and Alcpz committed Nov 29, 2023
    Configuration menu
    Copy the full SHA
    6a474c7 View commit details
    Browse the repository at this point in the history

  18. Refactored cooperative_loopback and memory implementation (#24) 

    * Refactored cooperative_loopback and memory implementation detail

* renamed load_counter to fetch_add_counter

* Removed dynamic tile counter from the scan memory struct

* scratch memory Reordering

* Fixed wrong values returned in LoopbackScanMemory.get_value

* Improved Class and variable naming
    @Alcpz
    Alcpz committed Nov 29, 2023
    Configuration menu
    Copy the full SHA
    ba7be34 View commit details
    Browse the repository at this point in the history

  19. [Scan_kt] Atomic64 flags + value implementation (#25) 

    * Implemented atomic64 version of the scan_kt pass

* Removed repeated offset calculation for tile id atomic flag

* Loopback -> Lookback. Removed unused var.
    @Alcpz
    Alcpz committed Nov 29, 2023
    Configuration menu
    Copy the full SHA
    69cc2fa View commit details
    Browse the repository at this point in the history

  20. constexpr, types and remove an unneeded check

    @joeatodd @Alcpz
    joeatodd authored and Alcpz committed Nov 29, 2023
    Configuration menu
    Copy the full SHA
    b5851ce View commit details
    Browse the repository at this point in the history

  21. Correct static_cast ?

    @joeatodd @Alcpz
    joeatodd authored and Alcpz committed Nov 29, 2023
    Configuration menu
    Copy the full SHA
    c9736c1 View commit details
    Browse the repository at this point in the history

  22. Defer group comms in lookback

    @joeatodd @Alcpz
    joeatodd authored and Alcpz committed Nov 29, 2023
    Configuration menu
    Copy the full SHA
    0e450f7 View commit details
    Browse the repository at this point in the history

  23. Disable dynamic tile ID by default 

    TODO: we still allocate & initialize the memory for the counter
    @joeatodd @Alcpz
    joeatodd authored and Alcpz committed Nov 29, 2023
    Configuration menu
    Copy the full SHA
    95b5552 View commit details
    Browse the repository at this point in the history

  24. Reduce from register sums instead of local mem 

    Also use #pragma unroll for now
    @joeatodd @Alcpz
    joeatodd authored and Alcpz committed Nov 29, 2023
    Configuration menu
    Copy the full SHA
    3f30ec8 View commit details
    Browse the repository at this point in the history

  25. Unrolled version of joint_inclusive_scan

    @joeatodd @Alcpz
    joeatodd authored and Alcpz committed Nov 29, 2023
    Configuration menu
    Copy the full SHA
    c147f05 View commit details
    Browse the repository at this point in the history

  26. Update include/oneapi/dpl/pstl/hetero/dpcpp/parallel_backend_sycl_scan.h 

    Co-authored-by: Alberto Cabrera Pérez <[email protected]>
    @joeatodd @Alcpz
    joeatodd and Alcpz committed Nov 29, 2023
    Configuration menu
    Copy the full SHA
    ab69568 View commit details
    Browse the repository at this point in the history

  27. Add TODO

    @joeatodd @Alcpz
    joeatodd authored and Alcpz committed Nov 29, 2023
    Configuration menu
    Copy the full SHA
    b992b84 View commit details
    Browse the repository at this point in the history

  28. Changing fill kernel for a memset

    @Alcpz
    Alcpz committed Nov 29, 2023
    Configuration menu
    Copy the full SHA
    37726be View commit details
    Browse the repository at this point in the history

  29. Single wg implementation

    @Alcpz
    Alcpz committed Nov 29, 2023
    Configuration menu
    Copy the full SHA
    d7c3c78 View commit details
    Browse the repository at this point in the history

Commits on Dec 8, 2023

  1. Add phase 1

    aidan.belton committed Dec 8, 2023
    Configuration menu
    Copy the full SHA
    e42e68d View commit details
    Browse the repository at this point in the history

  2. Add phase 2

    aidan.belton committed Dec 8, 2023
    Configuration menu
    Copy the full SHA
    54c0ae9 View commit details
    Browse the repository at this point in the history

  3. Add phase 3

    Aidan committed Dec 8, 2023
    Configuration menu
    Copy the full SHA
    ba543ed View commit details
    Browse the repository at this point in the history

  4. Add count datatype _SizeT

    Aidan committed Dec 8, 2023
    Configuration menu
    Copy the full SHA
    cdf74d0 View commit details
    Browse the repository at this point in the history

  5. Move away from atomics

    Aidan committed Dec 8, 2023
    Configuration menu
    Copy the full SHA
    c5670d8 View commit details
    Browse the repository at this point in the history

  6. Sort out test logic

    Aidan committed Dec 8, 2023
    Configuration menu
    Copy the full SHA
    8918b42 View commit details
    Browse the repository at this point in the history

  7. Remove unnecessary load and store functions

    Aidan committed Dec 8, 2023
    Configuration menu
    Copy the full SHA
    625f315 View commit details
    Browse the repository at this point in the history

  8. Release scratch mem

    Aidan committed Dec 8, 2023
    Configuration menu
    Copy the full SHA
    ca7a830 View commit details
    Browse the repository at this point in the history

  9. Add single wg copy if

    Aidan committed Dec 8, 2023
    Configuration menu
    Copy the full SHA
    25238eb View commit details
    Browse the repository at this point in the history

  10. Fix unrolls and use memset

    Aidan committed Dec 8, 2023
    Configuration menu
    Copy the full SHA
    2f2ccb2 View commit details
    Browse the repository at this point in the history

  11. apply changes to single wg

    Aidan committed Dec 8, 2023
    Configuration menu
    Copy the full SHA
    021fb9a View commit details
    Browse the repository at this point in the history

  12. Remove unused variables

    Aidan committed Dec 8, 2023
    Configuration menu
    Copy the full SHA
    c4b05a4 View commit details
    Browse the repository at this point in the history

  13. Clang-format copy_if_kt commits

    aidan.belton committed Dec 8, 2023
    Configuration menu
    Copy the full SHA
    5d1ed8e View commit details
    Browse the repository at this point in the history

  14. Merge pull request #28 from AidanBeltonS/dev/aidan/copy_if 

    Add copy_if_kt
    @AidanBeltonS
    AidanBeltonS authored Dec 8, 2023
    Configuration menu
    Copy the full SHA
    6382acd View commit details
    Browse the repository at this point in the history

Commits on May 9, 2024

  1. Enable pragma unroll for open-source DPC++

    @joeatodd @danhoeflinger
    joeatodd authored and danhoeflinger committed May 9, 2024
    Configuration menu
    Copy the full SHA
    70e751a View commit details
    Browse the repository at this point in the history

  2. Start of single-pass scan kernel template

    @adamfidel @danhoeflinger
    adamfidel authored and danhoeflinger committed May 9, 2024
    Configuration menu
    Copy the full SHA
    a9fdaa3 View commit details
    Browse the repository at this point in the history

  3. Fix hang in inclusive scan

    @adamfidel @danhoeflinger
    adamfidel authored and danhoeflinger committed May 9, 2024
    Configuration menu
    Copy the full SHA
    dfef06f View commit details
    Browse the repository at this point in the history

  4. Debug statements for scan kernel template

    @adamfidel @danhoeflinger
    adamfidel authored and danhoeflinger committed May 9, 2024
    Configuration menu
    Copy the full SHA
    555f6f9 View commit details
    Browse the repository at this point in the history

  5. Update scan kernel template test

    @adamfidel @danhoeflinger
    adamfidel authored and danhoeflinger committed May 9, 2024
    Configuration menu
    Copy the full SHA
    10cfc68 View commit details
    Browse the repository at this point in the history

  6. Only have a single work-item per group query for previous tile status

    @adamfidel @danhoeflinger
    adamfidel authored and danhoeflinger committed May 9, 2024
    Configuration menu
    Copy the full SHA
    53faf10 View commit details
    Browse the repository at this point in the history

  7. First attempt at parallel lookback

    @adamfidel @danhoeflinger
    adamfidel authored and danhoeflinger committed May 9, 2024
    Configuration menu
    Copy the full SHA
    dc63d16 View commit details
    Browse the repository at this point in the history

  8. Working cooperative lookback

    @adamfidel @danhoeflinger
    adamfidel authored and danhoeflinger committed May 9, 2024
    Configuration menu
    Copy the full SHA
    f8c3f2b View commit details
    Browse the repository at this point in the history

  9. Fix correctness issue with non-power-of-2 sizes

    @adamfidel @danhoeflinger
    adamfidel authored and danhoeflinger committed May 9, 2024
    Configuration menu
    Copy the full SHA
    1d72d3f View commit details
    Browse the repository at this point in the history

  10. Scan_kt Flags and Values separated (#15) 

    Atomic flags and the values used in Scan_kt separated to avoid truncating the range to 30bit values, and prepare for a more general scan implementation.
    @Alcpz @danhoeflinger
    Alcpz authored and danhoeflinger committed May 9, 2024
    Configuration menu
    Copy the full SHA
    567a50e View commit details
    Browse the repository at this point in the history

  11. Refactored Scan_kt code (#16) 

    * Improved Scan_kt: templated parameters, ballot, wgsize calculation.

- Changed number of workgroups calculation from next power of two to
next multiple of wgsize
- Improved group_ballot by using the class member functions
- Using kernel_param struct to determine wgsize and elems per work item.
    @Alcpz @danhoeflinger
    Alcpz authored and danhoeflinger committed May 9, 2024
    Configuration menu
    Copy the full SHA
    0c91640 View commit details
    Browse the repository at this point in the history

  12. Scan_kt: Single memory allocation for device_memory (#17) and async f… 

    …ree of the device memory (#18)

* Single memory allocation for device_memory

* async free of device memory

---------

Co-authored-by: Joe Todd <[email protected]>
Co-authored-by: AidanBeltonS <[email protected]>
    @Alcpz @joeatodd
    @AidanBeltonS @danhoeflinger
    3 people authored and danhoeflinger committed May 9, 2024
    Configuration menu
    Copy the full SHA
    78d2d7d View commit details
    Browse the repository at this point in the history

  13. Replace sycl::range with sycl::nd_range for fill

    @joeatodd @danhoeflinger
    joeatodd authored and danhoeflinger committed May 9, 2024
    Configuration menu
    Copy the full SHA
    55dc287 View commit details
    Browse the repository at this point in the history

  14. Bug fix

    @joeatodd @danhoeflinger
    joeatodd authored and danhoeflinger committed May 9, 2024
    Configuration menu
    Copy the full SHA
    37bfd1d View commit details
    Browse the repository at this point in the history

  15. Global to local then perform op

    @danhoeflinger
    Aidan authored and danhoeflinger committed May 9, 2024
    Configuration menu
    Copy the full SHA
    21038df View commit details
    Browse the repository at this point in the history

  16. Update based on feedback

    @danhoeflinger
    Aidan authored and danhoeflinger committed May 9, 2024
    Configuration menu
    Copy the full SHA
    bdcc9c9 View commit details
    Browse the repository at this point in the history

  17. Refactored cooperative_loopback and memory implementation (#24) 

    * Refactored cooperative_loopback and memory implementation detail

* renamed load_counter to fetch_add_counter

* Removed dynamic tile counter from the scan memory struct

* scratch memory Reordering

* Fixed wrong values returned in LoopbackScanMemory.get_value

* Improved Class and variable naming
    @Alcpz @danhoeflinger
    Alcpz authored and danhoeflinger committed May 9, 2024
    Configuration menu
    Copy the full SHA
    9717e09 View commit details
    Browse the repository at this point in the history

  18. [Scan_kt] Atomic64 flags + value implementation (#25) 

    * Implemented atomic64 version of the scan_kt pass

* Removed repeated offset calculation for tile id atomic flag

* Loopback -> Lookback. Removed unused var.
    @Alcpz @danhoeflinger
    Alcpz authored and danhoeflinger committed May 9, 2024
    Configuration menu
    Copy the full SHA
    8d23836 View commit details
    Browse the repository at this point in the history

  19. constexpr, types and remove an unneeded check

    @joeatodd @danhoeflinger
    joeatodd authored and danhoeflinger committed May 9, 2024
    Configuration menu
    Copy the full SHA
    c3c3218 View commit details
    Browse the repository at this point in the history

  20. Correct static_cast ?

    @joeatodd @danhoeflinger
    joeatodd authored and danhoeflinger committed May 9, 2024
    Configuration menu
    Copy the full SHA
    d257702 View commit details
    Browse the repository at this point in the history

  21. Defer group comms in lookback

    @joeatodd @danhoeflinger
    joeatodd authored and danhoeflinger committed May 9, 2024
    Configuration menu
    Copy the full SHA
    43e17ba View commit details
    Browse the repository at this point in the history

  22. Disable dynamic tile ID by default 

    TODO: we still allocate & initialize the memory for the counter
    @joeatodd @danhoeflinger
    joeatodd authored and danhoeflinger committed May 9, 2024
    Configuration menu
    Copy the full SHA
    e5b3ca4 View commit details
    Browse the repository at this point in the history

  23. Reduce from register sums instead of local mem 

    Also use #pragma unroll for now
    @joeatodd @danhoeflinger
    joeatodd authored and danhoeflinger committed May 9, 2024
    Configuration menu
    Copy the full SHA
    ab346da View commit details
    Browse the repository at this point in the history

  24. Unrolled version of joint_inclusive_scan

    @joeatodd @danhoeflinger
    joeatodd authored and danhoeflinger committed May 9, 2024
    Configuration menu
    Copy the full SHA
    f87573c View commit details
    Browse the repository at this point in the history

  25. Update include/oneapi/dpl/pstl/hetero/dpcpp/parallel_backend_sycl_scan.h 

    Co-authored-by: Alberto Cabrera Pérez <[email protected]>
    @joeatodd @Alcpz @danhoeflinger
    2 people authored and danhoeflinger committed May 9, 2024
    Configuration menu
    Copy the full SHA
    621adf7 View commit details
    Browse the repository at this point in the history

  26. Add TODO

    @joeatodd @danhoeflinger
    joeatodd authored and danhoeflinger committed May 9, 2024
    Configuration menu
    Copy the full SHA
    b8c837f View commit details
    Browse the repository at this point in the history

  27. Changing fill kernel for a memset

    @Alcpz @danhoeflinger
    Alcpz authored and danhoeflinger committed May 9, 2024
    Configuration menu
    Copy the full SHA
    8367be7 View commit details
    Browse the repository at this point in the history

  28. Single wg implementation

    @Alcpz @danhoeflinger
    Alcpz authored and danhoeflinger committed May 9, 2024
    Configuration menu
    Copy the full SHA
    02ff9f3 View commit details
    Browse the repository at this point in the history

  29. Add phase 1

    @danhoeflinger
    aidan.belton authored and danhoeflinger committed May 9, 2024
    Configuration menu
    Copy the full SHA
    25a93ff View commit details
    Browse the repository at this point in the history

  30. Add phase 2

    @danhoeflinger
    aidan.belton authored and danhoeflinger committed May 9, 2024
    Configuration menu
    Copy the full SHA
    aea6009 View commit details
    Browse the repository at this point in the history

  31. Add phase 3

    @danhoeflinger
    Aidan authored and danhoeflinger committed May 9, 2024
    Configuration menu
    Copy the full SHA
    d5c2cb5 View commit details
    Browse the repository at this point in the history

  32. Add count datatype _SizeT

    @danhoeflinger
    Aidan authored and danhoeflinger committed May 9, 2024
    Configuration menu
    Copy the full SHA
    1f574b8 View commit details
    Browse the repository at this point in the history

  33. Move away from atomics

    @danhoeflinger
    Aidan authored and danhoeflinger committed May 9, 2024
    Configuration menu
    Copy the full SHA
    16ef9c2 View commit details
    Browse the repository at this point in the history

  34. Sort out test logic

    @danhoeflinger
    Aidan authored and danhoeflinger committed May 9, 2024
    Configuration menu
    Copy the full SHA
    45a1fb7 View commit details
    Browse the repository at this point in the history

  35. Remove unnecessary load and store functions

    @danhoeflinger
    Aidan authored and danhoeflinger committed May 9, 2024
    Configuration menu
    Copy the full SHA
    cec32d7 View commit details
    Browse the repository at this point in the history

  36. Release scratch mem

    @danhoeflinger
    Aidan authored and danhoeflinger committed May 9, 2024
    Configuration menu
    Copy the full SHA
    b7d659c View commit details
    Browse the repository at this point in the history

  37. Add single wg copy if

    @danhoeflinger
    Aidan authored and danhoeflinger committed May 9, 2024
    Configuration menu
    Copy the full SHA
    fdb1824 View commit details
    Browse the repository at this point in the history

  38. Fix unrolls and use memset

    @danhoeflinger
    Aidan authored and danhoeflinger committed May 9, 2024
    Configuration menu
    Copy the full SHA
    1df5fbb View commit details
    Browse the repository at this point in the history

  39. apply changes to single wg

    @danhoeflinger
    Aidan authored and danhoeflinger committed May 9, 2024
    Configuration menu
    Copy the full SHA
    d8b77fe View commit details
    Browse the repository at this point in the history

  40. Remove unused variables

    @danhoeflinger
    Aidan authored and danhoeflinger committed May 9, 2024
    Configuration menu
    Copy the full SHA
    5b53de6 View commit details
    Browse the repository at this point in the history

  41. Clang-format copy_if_kt commits

    @danhoeflinger
    aidan.belton authored and danhoeflinger committed May 9, 2024
    Configuration menu
    Copy the full SHA
    acc4f9b View commit details
    Browse the repository at this point in the history

Commits on May 13, 2024

  1. partial changes (fix and amend) 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed May 13, 2024
    Configuration menu
    Copy the full SHA
    faae196 View commit details
    Browse the repository at this point in the history

Commits on May 24, 2024

  1. refactor to share lookback and memory mgr 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed May 24, 2024
    Configuration menu
    Copy the full SHA
    9d39fc6 View commit details
    Browse the repository at this point in the history

  2. formatting 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed May 24, 2024
    Configuration menu
    Copy the full SHA
    1b10214 View commit details
    Browse the repository at this point in the history

  3. remove launder in favor of lazy ctor union 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed May 24, 2024
    Configuration menu
    Copy the full SHA
    593c218 View commit details
    Browse the repository at this point in the history

  4. distinguishing kernel names 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed May 24, 2024
    Configuration menu
    Copy the full SHA
    5a1752c View commit details
    Browse the repository at this point in the history

  5. uglify 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed May 24, 2024
    Configuration menu
    Copy the full SHA
    a0576c3 View commit details
    Browse the repository at this point in the history

  6. format 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed May 24, 2024
    Configuration menu
    Copy the full SHA
    c5065a7 View commit details
    Browse the repository at this point in the history

  7. change single wg scan to submitter and kernel operator 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed May 24, 2024
    Configuration menu
    Copy the full SHA
    6a7291a View commit details
    Browse the repository at this point in the history

Commits on May 29, 2024

  1. change scan to submitter and kernel operator 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed May 29, 2024
    Configuration menu
    Copy the full SHA
    1b78dcd View commit details
    Browse the repository at this point in the history

  2. formatting 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed May 29, 2024
    Configuration menu
    Copy the full SHA
    246dbf3 View commit details
    Browse the repository at this point in the history

  3. remove unnecessary variable 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed May 29, 2024
    Configuration menu
    Copy the full SHA
    3023fec View commit details
    Browse the repository at this point in the history

  4. renaming public APIs 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed May 29, 2024
    Configuration menu
    Copy the full SHA
    52f6d82 View commit details
    Browse the repository at this point in the history

  5. sync with scan for asychronicity 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed May 29, 2024
    Configuration menu
    Copy the full SHA
    8b504d3 View commit details
    Browse the repository at this point in the history

  6. sycl::event returns 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed May 29, 2024
    Configuration menu
    Copy the full SHA
    6f60f10 View commit details
    Browse the repository at this point in the history

  7. naming and minor fixes 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed May 29, 2024
    Configuration menu
    Copy the full SHA
    0dfcd15 View commit details
    Browse the repository at this point in the history

  8. removing single_wg public api 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed May 29, 2024
    Configuration menu
    Copy the full SHA
    12db722 View commit details
    Browse the repository at this point in the history

  9. temporarily disable single wg version 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed May 29, 2024
    Configuration menu
    Copy the full SHA
    4850fcf View commit details
    Browse the repository at this point in the history

Commits on May 30, 2024

  1. wait after call for async algs 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed May 30, 2024
    Configuration menu
    Copy the full SHA
    a8ccbd1 View commit details
    Browse the repository at this point in the history

  2. reenable single wg 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed May 30, 2024
    Configuration menu
    Copy the full SHA
    70ad489 View commit details
    Browse the repository at this point in the history

  3. only need single phase for single wg 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed May 30, 2024
    Configuration menu
    Copy the full SHA
    e78b72c View commit details
    Browse the repository at this point in the history

Commits on May 31, 2024

  1. reusing single workgroup copy_if from oneDPL main 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed May 31, 2024
    Configuration menu
    Copy the full SHA
    2ca4287 View commit details
    Browse the repository at this point in the history

  2. add option to opt out of compiling single wg 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed May 31, 2024
    Configuration menu
    Copy the full SHA
    7538ed6 View commit details
    Browse the repository at this point in the history

  3. adding opt out for single wg inclusive scan 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed May 31, 2024
    Configuration menu
    Copy the full SHA
    293d724 View commit details
    Browse the repository at this point in the history

  4. remove single_wg kt, in favor of main oneDPL version 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed May 31, 2024
    Configuration menu
    Copy the full SHA
    565ba3b View commit details
    Browse the repository at this point in the history

Commits on Jun 4, 2024

  1. trying scalar version of copy_if 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed Jun 4, 2024
    Configuration menu
    Copy the full SHA
    b3fbfe2 View commit details
    Browse the repository at this point in the history

  2. fix 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed Jun 4, 2024
    Configuration menu
    Copy the full SHA
    e49bc9f View commit details
    Browse the repository at this point in the history

  3. fix 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed Jun 4, 2024
    Configuration menu
    Copy the full SHA
    92438ee View commit details
    Browse the repository at this point in the history

  4. full sum 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed Jun 4, 2024
    Configuration menu
    Copy the full SHA
    5495da2 View commit details
    Browse the repository at this point in the history

  5. switching arg to const ref 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed Jun 4, 2024
    Configuration menu
    Copy the full SHA
    a7ca1b5 View commit details
    Browse the repository at this point in the history

  6. branch by tile, not by workitem

    @danhoeflinger
    danhoeflinger committed Jun 4, 2024
    Configuration menu
    Copy the full SHA
    2d53ead View commit details
    Browse the repository at this point in the history

  7. removing unused block_strided version 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed Jun 4, 2024
    Configuration menu
    Copy the full SHA
    f26aff0 View commit details
    Browse the repository at this point in the history

  8. range API and formatting 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed Jun 4, 2024
    Configuration menu
    Copy the full SHA
    c0ab651 View commit details
    Browse the repository at this point in the history

  9. removing unnecessary stuff 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed Jun 4, 2024
    Configuration menu
    Copy the full SHA
    517c341 View commit details
    Browse the repository at this point in the history

  10. naming consistency 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed Jun 4, 2024
    Configuration menu
    Copy the full SHA
    427a5f4 View commit details
    Browse the repository at this point in the history

  11. formatting 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed Jun 4, 2024
    Configuration menu
    Copy the full SHA
    e9091e1 View commit details
    Browse the repository at this point in the history

  12. reverting overreach 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed Jun 4, 2024
    Configuration menu
    Copy the full SHA
    7315d2b View commit details
    Browse the repository at this point in the history

Commits on Jun 5, 2024

  1. upgrading tests to match scan, cmake 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed Jun 5, 2024
    Configuration menu
    Copy the full SHA
    0983043 View commit details
    Browse the repository at this point in the history

  2. test bugfix 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed Jun 5, 2024
    Configuration menu
    Copy the full SHA
    e09ccae View commit details
    Browse the repository at this point in the history

  3. bugfix for non-full case 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed Jun 5, 2024
    Configuration menu
    Copy the full SHA
    96acd30 View commit details
    Browse the repository at this point in the history

  4. fix range to check

    @danhoeflinger
    danhoeflinger committed Jun 5, 2024
    Configuration menu
    Copy the full SHA
    f15c759 View commit details
    Browse the repository at this point in the history

  5. adjust data generation 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed Jun 5, 2024
    Configuration menu
    Copy the full SHA
    b195e4d View commit details
    Browse the repository at this point in the history

  6. better fix for non-full case 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed Jun 5, 2024
    Configuration menu
    Copy the full SHA
    4528cbb View commit details
    Browse the repository at this point in the history

  7. removing old test 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed Jun 5, 2024
    Configuration menu
    Copy the full SHA
    831f9c9 View commit details
    Browse the repository at this point in the history

Commits on Jun 6, 2024

  1. undo change to unroll version check 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed Jun 6, 2024
    Configuration menu
    Copy the full SHA
    7be4594 View commit details
    Browse the repository at this point in the history

  2. formatting 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed Jun 6, 2024
    Configuration menu
    Copy the full SHA
    68c258d View commit details
    Browse the repository at this point in the history

  3. allowing alg to dictate active threads 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed Jun 6, 2024
    Configuration menu
    Copy the full SHA
    347bcf5 View commit details
    Browse the repository at this point in the history

  4. bugfix for indexes

    @danhoeflinger
    danhoeflinger committed Jun 6, 2024
    Configuration menu
    Copy the full SHA
    62f58e2 View commit details
    Browse the repository at this point in the history

  5. clang format 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed Jun 6, 2024
    Configuration menu
    Copy the full SHA
    5c4bd74 View commit details
    Browse the repository at this point in the history

Commits on Jun 18, 2024

  1. address reviewer comments 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed Jun 18, 2024
    Configuration menu
    Copy the full SHA
    98324ed View commit details
    Browse the repository at this point in the history

  2. simplify data generation and cutoff calculation. 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed Jun 18, 2024
    Configuration menu
    Copy the full SHA
    4377c2d View commit details
    Browse the repository at this point in the history

  3. strip out single workgroup opt out 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed Jun 18, 2024
    Configuration menu
    Copy the full SHA
    496ed4d View commit details
    Browse the repository at this point in the history

  4. minimal data generation changes 

    Signed-off-by: Dan Hoeflinger <[email protected]>
    @danhoeflinger
    danhoeflinger committed Jun 18, 2024
    Configuration menu
    Copy the full SHA
    2d9e8a7 View commit details
    Browse the repository at this point in the history