add support for additional hidden kernel args #717
Closed
Conversation
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
simeonschaub
force-pushed
the
sds/additional_args
branch
from
06246b2 to
aac7d0f
Compare
Codecov Report❌ Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## tb/kernel_state_reference #717 +/- ##
==============================================================
+ Coverage 0.00% 72.32% +72.32%
==============================================================
Files 24 24
Lines 3507 3621 +114
==============================================================
+ Hits 0 2619 +2619
+ Misses 3507 1002 -2505 ☔ View full report in Codecov by Sentry. 🚀 New features to boost your workflow:
|
|
So basically dynamic shared memory from CUDA, but using arguments instead of an API call to configure the amount of memory? I don't particularly like how this is fitted in with the KernelState stuff though. What about generalizing the Metal.jl functionality instead?That should be much closer already to what you need, lowering arbitrary intrinsics to trailing kernel arguments. |
|
Hmm, yes, that looks very promising! I'll try to use that then |
simeonschaub
added a commit
that referenced
this pull request
Sep 9, 2025
Allows other backends to pass additional hidden arguments that can be accessed through intrinsics. Required for OpenCL device-side RNG support, where additional shared memory must be passed as arguments to the kernel. Replaces #717
|
Closing in favor of #718 |
simeonschaub
added a commit
that referenced
this pull request
Sep 9, 2025
Allows other backends to pass additional hidden arguments that can be accessed through intrinsics. Required for OpenCL device-side RNG support, where additional shared memory must be passed as arguments to the kernel. Replaces #717
simeonschaub
added a commit
that referenced
this pull request
Sep 11, 2025
Allows other backends to pass additional hidden arguments that can be accessed through intrinsics. Required for OpenCL device-side RNG support, where additional shared memory must be passed as arguments to the kernel. Replaces #717
simeonschaub
added a commit
that referenced
this pull request
Sep 11, 2025
Allows other backends to pass additional hidden arguments that can be accessed through intrinsics. Required for OpenCL device-side RNG support, where additional shared memory must be passed as arguments to the kernel. Replaces #717
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.
For device-side RNG in OpenCL we need to allocate local memory depending on the number of subgroups in a workgroup and AFAICT, the only way to do that is to pass the local memory as an argument to the kernel. This adds functionality to add such additional hidden arguments to a kernel and feed them through functions that require them based on the existing kernel state passes.