-
Notifications
You must be signed in to change notification settings - Fork 34
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.
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
Graph fixes + other stuff #434
Open
franz
wants to merge
35
commits into
main
Choose a base branch
from
graph_opencl
base: main
Could not load branches
Branch not found: {{ refName }}
Loading
Could not load tags
Nothing to show
Loading
Are you sure you want to change the base?
Some commits from the old base branch may be removed from the timeline,
and old review comments may become outdated.
Open
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
this is required for cl_khr_command_buffer
This class is used to implement Graphs that execute "natively" in the backend, using OpenCL command-buffers or LevelZero command-lists and only synchronizing with the host when required. Fallback to original Graph is provided in CHIPGraphExec::launch()
* turn CHIPEvent.Refc from pointer to integer, move it into CHIPEventLevel0 * disable increaseRefCount/decreaseRefCount for OpenCL, the OpenCL runtime already does the refcounting * remove naked pointers to OpenCL objects, instead use the smart pointers from opencl.hpp header
"ctest --timeout 120 -R Unit_hipGraph" with POCL (with the new cl_pocl_command_buffer_* extensions) reports 59% tests passed, 62 tests failed out of 152 additionally, samples/graph + samples/graphMatrixMultiply work using the "native graphs" (cl_command_buffer), not the original chip-spv's graph execution.
Previously, the OpenCL device was checked for fine-grained SVM support, and if it was unavailable, the backend would assume only coarse-grained support and insert SVMMap & SVMUnmap for each buffer used by a kernel. Since Intel OpenCL and PoCL both support the USM extension now, but don't support fine-grained SVM, this commit helps to avoid the overhead of the unnecessary Map & Unmap commands.
original code reports Average iteration time
franz
force-pushed
the
graph_opencl
branch
2 times, most recently
from
May 12, 2023 11:19
5791e64
to
25d2f83
Compare
enabling Device/Host USM types make some Texture tests fail
linehill
reviewed
May 15, 2023
linehill
reviewed
May 15, 2023
pvelesko
reviewed
May 15, 2023
linehill
reviewed
May 15, 2023
linehill
reviewed
May 15, 2023
linehill
reviewed
May 15, 2023
pvelesko
reviewed
May 15, 2023
linehill
reviewed
May 15, 2023
linehill
reviewed
May 15, 2023
linehill
reviewed
May 15, 2023
linehill
reviewed
May 15, 2023
linehill
reviewed
May 15, 2023
linehill
reviewed
May 15, 2023
Comment on lines
+55
to
+57
// TODO: investigate. Uncommenting this code makes | ||
// a bunch of Unit_hipTexture tests fail with segfault. | ||
/* |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Perhaps, resolve this. I think it's fine too to exclude the tests conditionally at the configure time in UnitTests.cmake.
linehill
reviewed
May 15, 2023
Co-authored-by: Henry Linjamäki <linehill@users.noreply.github.com>
Co-authored-by: Henry Linjamäki <linehill@users.noreply.github.com>
Co-authored-by: Henry Linjamäki <linehill@users.noreply.github.com>
Co-authored-by: Henry Linjamäki <linehill@users.noreply.github.com>
Co-authored-by: Henry Linjamäki <linehill@users.noreply.github.com>
Co-authored-by: Henry Linjamäki <linehill@users.noreply.github.com>
Co-authored-by: Henry Linjamäki <linehill@users.noreply.github.com>
Co-authored-by: Henry Linjamäki <linehill@users.noreply.github.com>
Co-authored-by: Henry Linjamäki <linehill@users.noreply.github.com>
Co-authored-by: Henry Linjamäki <linehill@users.noreply.github.com>
Co-authored-by: Henry Linjamäki <linehill@users.noreply.github.com>
Co-authored-by: Henry Linjamäki <linehill@users.noreply.github.com>
Co-authored-by: Henry Linjamäki <linehill@users.noreply.github.com>
Co-authored-by: Henry Linjamäki <linehill@users.noreply.github.com>
Co-authored-by: Henry Linjamäki <linehill@users.noreply.github.com>
Co-authored-by: Henry Linjamäki <linehill@users.noreply.github.com>
linehill
reviewed
May 17, 2023
@franz conflicts |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
cl_intel_unified_shared_memory
, this removes the need for using SVMMap/Unmap commands on devices that support USMapprox 50-60 additional
Unit_hipGraph
tests pass