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

[SYCL][CUDA] Minor fixes required to run BabelStream benchmarks on CUDA #1543

Closed
wants to merge 13 commits into from
Closed

[SYCL][CUDA] Minor fixes required to run BabelStream benchmarks on CUDA #1543

wants to merge 13 commits into from

Commits on Apr 23, 2020

  1. [SYCL][CUDA] Fix for default selection of CUDA devices 

    Signed-off-by: Ruyman Reyes <[email protected]>
    @Ruyk
    Ruyk committed Apr 23, 2020
    Configuration menu
    Copy the full SHA
    1757da8 View commit details
    Browse the repository at this point in the history

  2. [SYCL][CUDA] Missing return event on map/unmap 

    Signed-off-by: Ruyman Reyes <[email protected]>
    @Ruyk
    Ruyk committed Apr 23, 2020
    Configuration menu
    Copy the full SHA
    aa384ae View commit details
    Browse the repository at this point in the history

  3. [SYCL][PI] Equality operator for PI plugins 

    Implemented a comparison operator for the plugin class in SYCL RT:
Two plugins are equal if the pointer to their string is the same.

plugin constructor marked explicit to avoid accidental implicit
conversions.

Signed-off-by: Ruyman Reyes <[email protected]>
    @Ruyk
    Ruyk committed Apr 23, 2020
    Configuration menu
    Copy the full SHA
    f910c93 View commit details
    Browse the repository at this point in the history

  4. [SYCL][PI] Queue constr. throws if invalid backend 

    Constructor of the SYCL queue throws an exception if the
device passed in is from a different backend than the context
that is associated with the queue.

Signed-off-by: Ruyman Reyes <[email protected]>
    @Ruyk
    Ruyk committed Apr 23, 2020
    Configuration menu
    Copy the full SHA
    195414f View commit details
    Browse the repository at this point in the history

  5. [SYCL][CUDA] Re-enabled DataMovement test for CUDA 

    Signed-off-by: Ruyman Reyes <[email protected]>
    @Ruyk
    Ruyk committed Apr 23, 2020
    Configuration menu
    Copy the full SHA
    5f611cb View commit details
    Browse the repository at this point in the history

Commits on Apr 24, 2020

  1. [SYCL][CUDA] Using name to check for CUDA platform 

    Fixes previous incorrect usage of version

Signed-off-by: Ruyman Reyes <[email protected]>
    @Ruyk
    Ruyk committed Apr 24, 2020
    Configuration menu
    Copy the full SHA
    09b9859 View commit details
    Browse the repository at this point in the history

  2. Merge branch 'sycl' into fix-default-selector-cuda 

    Signed-off-by: Ruyman Reyes <[email protected]>
    @Ruyk
    Ruyk committed Apr 24, 2020
    Configuration menu
    Copy the full SHA
    a907581 View commit details
    Browse the repository at this point in the history

Commits on Apr 28, 2020

  1. [SYCL][CUDA] Remove CUDA OpenCL from selection 

    The NVIDIA OpenCL platform is problematic for NVIDIA CUDA backend users
and for overall DPCPP users, since it doesnt work straight away and is
typically selected on OpenCL backend as a preference.

This patch removes the CUDA OpenCL platform from the device selection,
and prevents it from being used in the lit testing.

Signed-off-by: Ruyman Reyes <[email protected]>
    @Ruyk
    Ruyk committed Apr 28, 2020
    Configuration menu
    Copy the full SHA
    ac3b6f8 View commit details
    Browse the repository at this point in the history

Commits on Apr 29, 2020

  1. [SYCL][PI] Equality of plugins per backend 

    Defines two plugins being equal if their backend types are the same.

Signed-off-by: Ruyman Reyes <[email protected]>
    @Ruyk
    Ruyk committed Apr 29, 2020
    Configuration menu
    Copy the full SHA
    70a3e08 View commit details
    Browse the repository at this point in the history

  2. Merge branch 'sycl' into fix-default-selector-cuda

    @Ruyk
    Ruyk committed Apr 29, 2020
    Configuration menu
    Copy the full SHA
    a659354 View commit details
    Browse the repository at this point in the history

  3. [SYCL][CUDA] CUDA OpenCL platform ignored in get_devices 

    CUDA OpenCL platform is ignored on get_devices as suggested by feedback

Signed-off-by: Ruyman Reyes <[email protected]>
    @Ruyk
    Ruyk committed Apr 29, 2020
    Configuration menu
    Copy the full SHA
    17480e0 View commit details
    Browse the repository at this point in the history

  4. [SYCL] clang-format affected files 

    Signed-off-by: Ruyman Reyes <[email protected]>
    @Ruyk
    Ruyk committed Apr 29, 2020
    Configuration menu
    Copy the full SHA
    66810be View commit details
    Browse the repository at this point in the history

Commits on Apr 30, 2020

  1. [SYCL][CUDA] CUDA Device selection changes 

    * Removes NVIDIA OpenCL from the available list of platforms
* CUDA backend is available only if SYCL_BE=PI_CUDA is set

Signed-off-by: Ruyman Reyes <[email protected]>
    @Ruyk
    Ruyk committed Apr 30, 2020
    Configuration menu
    Copy the full SHA
    ad3951c View commit details
    Browse the repository at this point in the history