# llama.cpp adds fix for Hexagon DMA overflow in v b11516

> The open‑source LLM runtime now retires old DMA descriptors on Qualcomm Hexagon NPU, preventing NaN outputs.

Oossa · 2026-10-09 · https://oossa.com/en/llama-cpp-adds-fix-for-hexagon-dma-overflow-in-v-b11516

The ggml‑org/llama.cpp project released version b11516 on 2026‑10‑09. The update fixes a bug in the Hexagon NPU path where the DMA ring could overflow, causing stale data and NaN/inf results. The code now drops the oldest descriptor when the ring is full and flushes the queue later. Test cases with large patch‑embed operations now pass on Hexagon hardware.

## The facts

- Release version b11516 published on 2026‑10‑09.
- Fix prevents overflow when IC*KH exceeds 256 entries.

## Why it matters

Developers using llama.cpp on Qualcomm Hexagon devices will get correct model outputs instead of erroneous NaNs.

## Sources & references

1. [ggml-org/llama.cpp b11516](https://github.com/ggml-org/llama.cpp/releases/tag/b11516) – llama.cpp, 2026-10-09

Last updated: 2026-10-09