Closed
Conversation
Copy the occupancy API algorithm from HIP Add `groupsize=:auto` option to at-roc to automatically choose groupsize Implement `launch_configuration` API
jpsamaroo
force-pushed
the
jps/occupancy
branch
from
e25ed77 to
8f89517
Compare
Member
|
Implemented in #423. |
This file contains hidden or 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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
We now use GPUCompiler's kernel state support to pass an
AMDGPU.KernelStateobject as the first argument to all kernels. This object contains an exception flag, which replaces the exception that we would place in a global variable, which allows us to know exactly which kernels threw an exception (instead of making it appear like all kernels of a given type may have thrown an exception).One unfortunate side effect of adding this new argument is that it disables kernarg buffer sharing across otherwise identical kernel invocations, because now each
KernelStateobject has a unique exception flag. Maybe some kind of dedicated buffer pooling would help here? That would also decrease the newly-added cost of flag allocation in the launch path.Todo: