CUDA: consistent use of __restrict__ + PDL for FA - #25185
Merged
JohannesGaessler merged 1 commit intoJul 1, 2026
Conversation
github-actions
Bot
added
ggml
am17an
approved these changes
Jul 1, 2026
DFveloper
pushed a commit
to DFveloper/aikar-engine
that referenced
this pull request
Jul 3, 2026
adrianhoehne
pushed a commit
to adrianhoehne/llama.cpp
that referenced
this pull request
Jul 5, 2026
giveen
added a commit
to giveen/llama-cpp-turboquant
that referenced
this pull request
Jul 18, 2026
Port upstream commit b820cc8 (CUDA: consistent use of __restrict__ + PDL for FA ggml-org#25185). Avoids compiler race condition between PDL and __restrict__ on Hopper+ GPUs by using GGML_CUDA_RESTRICT (which is a no-op when PDL is active) instead of raw __restrict__ on kernel params. Also switches the kernel launch to ggml_cuda_kernel_launch() for PDL enrollment.
CowboyTim
pushed a commit
to aardbeiplantje/llama.cpp
that referenced
this pull request
Jul 21, 2026
zommiommy
pushed a commit
to zommiommy/llama.cpp
that referenced
this pull request
Aug 18, 2026
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
3 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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
The auxiliary kernel
flash_attn_mask_to_KV_maxis only partially enrolled into PDL. It has a call toggml_cuda_pdl_syncbut is still launched via
<<<...>>>and has__restrict__pointers in its signature. This PR fully enrollsflash_attn_mask_to_KV_maxin PDL to make its usage consistent with the surrounding FA kernels.Requirements