Fix meta backend tensor reads for split tensors during state serialization#22063

JohannesGaessler merged 1 commit into

ggml-org:masterggml-org/llama.cpp:masterfrom

ssam18:fix/meta-buffer-get-tensor-multi-segmentssam18/llama.cpp:fix/meta-buffer-get-tensor-multi-segmentCopy head branch name to clipboard