On Fri, Jul 8, 2011 at 12:53 AM, Chris Pearson <kermit4@xxxxxxxxx> wrote: > see attached screenshots > > NULL pointer dereference at 8 > > isolate_lru_pages > shrink_inactive_list > __lookup_tag > shrink_zone > shrink_slab > kswapd > zone_reclaim > > These are from 3 different servers in the past week since we upgraded > a few hundred of them to 2.6.39.1. They're under a steady few MB/s > of net and disk I/O load. > > We have the following /proc adjustments: > > kernel.shmmax = 135217728 > fs.file-max = 65535 > vm.swappiness = 10 > vm.min_free_kbytes = 65535 > I didn't have see such BUG until now. Could you tell me which point is isolate_lru_pages + 0x225? You can get it with addr2line -e vmlinux -i ffffffff8108ed15 or gdb. The culprit I think is page_count. A month ago, Andrea pointed out and sent the patch but it seems it isn't stable tree. Could you test below patch? https://patchwork.kernel.org/patch/857442/ -- Kind regards, Minchan Kim -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@xxxxxxxxx. 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