Hi, > + /* 256 or 512 bytes/ecc */ > + uint32_t j =(((struct nand_chip *)mtd->priv)->ecc.size)/256; It's certainly faster to use >> 8 instead. At least on x86 GCC would not optimize this for you if the kernel is compiles with -Os, it would use idiv instead. -- Best regards, Artem Bityutskiy (Битюцкий Артём)