Hi, During test fengguang's readahead alloc-noretry patch, I have some thoughts In the 1000 dd case, page cache of sparse file hole are all zero indeed. So what about make a global zero page for that purpose, fs level know it is a hole, when write occurs on that page we can alloc a new page for that. I have no enough knowledge to implement it so just give out the question without patch, sorry. BTW, was there any attempt for this before? -- Regards dave -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@xxxxxxxxxx For more info on Linux MM, see: http://www.linux-mm.org/ . Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/ Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>