From mboxrd@z Thu Jan 1 00:00:00 1970 Subject: Re: slab fragmentation ? From: Badari Pulavarty In-Reply-To: <4162ECAD.8090403@colorfullife.com> References: <1096500963.12861.21.camel@dyn318077bld.beaverton.ibm.com> <20040929204143.134154bc.akpm@osdl.org> <29460000.1096555795@[10.10.2.4]> <1096555693.12861.27.camel@dyn318077bld.beaverton.ibm.com> <415F968B.8000403@colorfullife.com> <1096905099.12861.117.camel@dyn318077bld.beaverton.ibm.com> <41617567.9010507@colorfullife.com> <1096987570.12861.122.camel@dyn318077bld.beaverton.ibm.com> <4162E0AF.4000704@colorfullife.com> <1097000846.12861.143.camel@dyn318077bld.beaverton.ibm.com> <4162ECAD.8090403@colorfullife.com> Content-Type: text/plain Message-Id: <1097002074.12861.145.camel@dyn318077bld.beaverton.ibm.com> Mime-Version: 1.0 Date: 05 Oct 2004 11:47:54 -0700 Content-Transfer-Encoding: 7bit Sender: owner-linux-mm@kvack.org Return-Path: To: Manfred Spraul Cc: Andrew Morton , linux-mm@kvack.org List-ID: On Tue, 2004-10-05 at 11:49, Manfred Spraul wrote: > Badari Pulavarty wrote: > > >>The fix would be simple: kmem_cache_alloc_node must walk through the > >>list of partial slabs and check if it finds a slab from the correct > >>node. If it does, then just use that slab instead of allocating a new > >>one. And statistics must be added to kmem_cache_alloc_node - I forgot > >>that when I wrote the function. > >> > >> > > > >I will add more debug to find out if this is happening or not. > > > >What stats you want me to update in kmem_cache_alloc_node() ? > > > > > > > I would just add a printk to confirm our suspicion. > "kmem_cache_alloc_node called" + dump_stack(). I always use that > approach, thus I forgot to add proper statistics. Yep. [] dump_stack+0x17/0x1b [] kmem_cache_alloc_node+0x181/0x186 [] __alloc_percpu+0x65/0xb9 [] alloc_disk+0x45/0xc8 [] sd_probe+0x89/0x341 [] bus_match+0x35/0x5e [] device_attach+0x3f/0x8f [] bus_add_device+0x68/0xab [] device_add+0x94/0x12a [] scsi_sysfs_add_sdev+0x3e/0x1bf [] scsi_add_lun+0x2d9/0x378 [] scsi_probe_and_add_lun+0xa4/0x1d1 [] scsi_report_lun_scan+0x2a0/0x3c2 [] scsi_scan_target+0xc2/0xeb [] scsi_scan_channel+0x64/0x77 [] scsi_scan_host_selected+0xc5/0xff [] scsi_scan_host+0x24/0x28 [] sdebug_driver_probe+0x94/0xb8 [scsi_debug] [] bus_match+0x35/0x5e [] device_attach+0x3f/0x8f [] bus_add_device+0x68/0xab [] device_add+0x94/0x12a [] sdebug_add_adapter+0x124/0x1e7 [scsi_debug] [] sdebug_add_host_store+0x6b/0x9a [scsi_debug] [] drv_attr_store+0x37/0x39 [] flush_write_buffer+0x31/0x3e [] sysfs_write_file+0x43/0x52 [] vfs_write+0xba/0x107 [] sys_write+0x35/0x53 [] sysenter_past_esp+0x52/0x71 -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@kvack.org. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: aart@kvack.org