From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Nicholas A. Bellinger" Subject: [PATCH 0/4] target: Move backend memory release to configfs_item_operations->release() Date: Wed, 2 Feb 2011 00:26:47 -0800 Message-ID: <1296635211-6269-1-git-send-email-nab@linux-iscsi.org> Return-path: Received: from nm15-vm0.bullet.mail.ac4.yahoo.com ([98.139.52.236]:22251 "HELO nm15-vm0.bullet.mail.ac4.yahoo.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with SMTP id S1753395Ab1BBI0r (ORCPT ); Wed, 2 Feb 2011 03:26:47 -0500 Sender: linux-scsi-owner@vger.kernel.org List-Id: linux-scsi@vger.kernel.org To: linux-scsi Cc: Christoph Hellwig , Joel Becker , Fubo Chen , Nicholas Bellinger From: Nicholas Bellinger Greetings all, This series addresses four scenarios in backend HBA/DEV target_core_configfs.c code where config_item_put() being called upon the following TCM config_groups: struct se_hba->hba_group struct se_subsystem_dev->se_dev_group struct t10_alua_lu_gp->lu_gp_group struct t10_alua_tg_pt_gp->tg_pt_gp_group and subsequently releasing memory of these four parent TCM data structures is causing a number of SLUB 'Poison overwritten' warnings during slub_debug=FPUZ testing with .38-rc2 code. This series addresses the issue by using moving existing release logic of the parent structure containing a struct_config_group into it's own seperate struct configfs_item_operations->release() callback. Please review and comment, Signed-off-by: Nicholas A. Bellinger Nicholas Bellinger (4): target: Move core_delete_hba() into ->release() callback target: Move subdev release logic into ->release() callback target: Move core_alua_free_lu_gp() into ->release() callback target: Move core_alua_free_tg_pt_gp() into ->release() callback drivers/target/target_core_configfs.c | 114 +++++++++++++++++++++------------ 1 files changed, 73 insertions(+), 41 deletions(-)