]> git.djapps.eu Git - pkg/ggml/sources/llama.cpp/commit
overview / pkg / ggml / sources / llama.cpp / commit
summary | shortlog | log | commit | commitdiff | tree
(parent: bcfc8c3)
(Bugfix, ggml-cuda) Pool alloc count fix + small size computation type adjustment...
authorpl752 <redacted>
Sat, 3 Jan 2026 10:13:40 +0000 (15:13 +0500)
committerGitHub <redacted>
Sat, 3 Jan 2026 10:13:40 +0000 (11:13 +0100)
commit9dba9f5352308894bfb8786fcfe7c284168ff8f5
tree83cc2fe4b1831c5252ee661aa4d720aafc41fd81tree
parentbcfc8c3cec68f8c3154e7e5d5a2efd807b241767commit | diff
(Bugfix, ggml-cuda) Pool alloc count fix + small size computation type adjustment (#18559)

* CUDA: Fixed obj byte size instead of obj count being passed to pool alloc (fattn-common, dst_tmp_meta)

* CUDA: Explicitly casted some of the int alloc counts before multiplication in argsort

---------

Co-authored-by: pl752 <redacted>
ggml/src/ggml-cuda/argsort.cu diff | blob | history
ggml/src/ggml-cuda/fattn-common.cuh diff | blob | history
Packaging of ggml-org/llama.cpp