llama.cpp

llama : fix integer overflow during quantization

#6063

Merged