From mboxrd@z Thu Jan 1 00:00:00 1970 From: Dave Jones Subject: huge failed atomic allocation in dccp. Date: Sat, 17 May 2008 18:20:27 -0400 Message-ID: <20080517222027.GB31440@codemonkey.org.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii To: netdev@vger.kernel.org Return-path: Received: from testure.choralone.org ([194.9.77.134]:33709 "EHLO testure.choralone.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751056AbYEQWTn (ORCPT ); Sat, 17 May 2008 18:19:43 -0400 Received: from davej by testure.choralone.org with local (Exim 4.69) (envelope-from ) id 1JxUlP-0001CR-Ss for netdev@vger.kernel.org; Sat, 17 May 2008 23:20:27 +0100 Content-Disposition: inline Sender: netdev-owner@vger.kernel.org List-ID: This strikes me as somewhat bizarre.. modprobe: page allocation failure. order:10, mode:0x20 Pid: 10505, comm: modprobe Not tainted 2.6.25-14.fc9.x86_64 #1 Call Trace: [] __alloc_pages+0x351/0x371 [] ? sysfs_slab_alias+0x41/0x81 [] alloc_pages_current+0x100/0x109 [] __get_free_pages+0xe/0x4d [] :dccp:dccp_init+0xbd/0x3a5 [] sys_init_module+0x193f/0x1a87 [] ? do_sync_read+0xe7/0x12d [] ? release_sock+0x0/0xaf [] ? audit_syscall_entry+0x126/0x15a [] ? syscall_trace_enter+0xb5/0xb9 [] tracesys+0xd5/0xda An order 10 GFP_ATOMIC allocation failing shouldn't really be surprising. The hash sizing in dccp_init seems to cope with this situation by trying successively smaller values, but this will cause spew to the logs whilst it's doing that. Dave -- http://www.codemonkey.org.uk