From mboxrd@z Thu Jan 1 00:00:00 1970 From: Alexey Fisher Subject: SLAB break supsned and cpu-hotplug on SMP, using SLUB instead. Date: Fri, 13 Jun 2008 22:39:47 +0200 Message-ID: <1213389587.10307.12.camel@zwerg> Mime-Version: 1.0 Content-Transfer-Encoding: 7bit Return-path: Sender: kernel-testers-owner-u79uwXL29TY76Z2rM5mHXA@public.gmane.org List-ID: Content-Type: text/plain; charset="us-ascii" To: kernel-testers-u79uwXL29TY76Z2rM5mHXA@public.gmane.org I found that SLAB breaking supsned and cpu-hotplug on my PC. I didn't get to bisect it, so i think it is just the part of SLAB. At the moment i converted my kernel to SLUB and it working just fine, about 10 suspends and no reboot or crush ... it ist a record for my PC. Easy way to reproduce is to compile SLAB kernel and echo 0 > /sys/devices/system/cpu/cpu1/online echo 1 > /sys/devices/system/cpu/cpu1/online other work around to make my PC working stable are kernel options: nosmp or mem=1024M kernel 2.6.26-rc6-00028-gc6d8f40 #68 SMP PREEMPT dmesg will return: [ 119.362281] CPU0 attaching NULL sched-domain. [ 119.362281] CPU1 attaching NULL sched-domain. [ 119.365524] kvm: disabling virtualization on CPU1 [ 119.365524] Breaking affinity for irq 1 [ 119.365524] Breaking affinity for irq 18 [ 119.468547] CPU 1 is now offline [ 119.468553] lockdep: fixing up alternatives. [ 119.468561] SMP alternatives: switching to UP code [ 119.468568] BUG: sleeping function called from invalid context at mm/slab.c:3052 [ 119.468574] in_atomic():1, irqs_disabled():0 [ 119.468578] 4 locks held by bash/5128: [ 119.468581] #0: (&buffer->mutex){--..}, at: [] sysfs_write_file+0x25/0xe3 [ 119.468598] #1: (cpu_add_remove_lock){--..}, at: [] cpu_maps_update_begin+0xf/0x11 [ 119.468611] #2: (&cpu_hotplug.lock){--..}, at: [] cpu_hotplug_begin+0x1a/0x42 [ 119.468623] #3: (smp_alt){--..}, at: [] alternatives_smp_switch+0x14c/0x165 [ 119.468638] Pid: 5128, comm: bash Not tainted 2.6.26-rc6-testing-00028-gc6d8f40 #67 [ 119.468645] [] __might_sleep+0xe7/0xee [ 119.468655] [] kmem_cache_alloc+0x27/0xd2 [ 119.468664] [] ? __get_vm_area_node+0x86/0x186 [ 119.468674] [] __get_vm_area_node+0x86/0x186 [ 119.468684] [] get_vm_area_caller+0x3e/0x43 [ 119.468692] [] ? text_poke+0x9e/0x11e [ 119.468702] [] vmap+0x21/0x44 [ 119.468710] [] text_poke+0x9e/0x11e [ 119.468717] [] ? sys_execve+0x37/0x4d [ 119.468727] [] alternatives_smp_unlock+0x41/0x50 [ 119.468736] [] alternatives_smp_switch+0x118/0x165 [ 119.468813] [] ? _etext+0x0/0xf [ 119.468822] [] ? printk+0x15/0x17 [ 119.468833] [] __cpu_die+0x42/0x69 [ 119.468842] [] _cpu_down+0x148/0x1f8 [ 119.468854] [] cpu_down+0x23/0x30 [ 119.468863] [] store_online+0x27/0x5a [ 119.468870] [] ? store_online+0x0/0x5a [ 119.468880] [] sysdev_store+0x20/0x25 [ 119.468889] [] sysfs_write_file+0xb8/0xe3 [ 119.468899] [] ? sysfs_write_file+0x0/0xe3 [ 119.468907] [] vfs_write+0x8c/0x108 [ 119.468916] [] sys_write+0x3b/0x60 [ 119.468924] [] sysenter_past_esp+0x6a/0xb1 [ 119.468934] ======================= [ 119.625144] CPU0 attaching NULL sched-domain. -- To unsubscribe from this list: send the line "unsubscribe kernel-testers" in the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org More majordomo info at http://vger.kernel.org/majordomo-info.html