From mboxrd@z Thu Jan 1 00:00:00 1970 From: Tejun Heo Subject: [PATCHSET 01/11] prep for new EH Date: Thu, 11 May 2006 20:59:49 +0900 Message-ID: <11473487893774-git-send-email-htejun@gmail.com> Reply-To: Tejun Heo Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7BIT Return-path: Received: from nz-out-0102.google.com ([64.233.162.195]:9925 "EHLO nz-out-0102.google.com") by vger.kernel.org with ESMTP id S1751557AbWEKL75 (ORCPT ); Thu, 11 May 2006 07:59:57 -0400 Received: by nz-out-0102.google.com with SMTP id 13so170486nzn for ; Thu, 11 May 2006 04:59:56 -0700 (PDT) In-Reply-To: Sender: linux-ide-owner@vger.kernel.org List-Id: linux-ide@vger.kernel.org To: jgarzik@pobox.com, alan@lxorguk.ukuu.org.uk, axboe@suse.de, albertcc@tw.ibm.com, forrest.zhao@intel.com, efalk@google.com, linux-ide@vger.kernel.org, htejun@gmail.com Hello, all. This is part of patchset series described in [T]. This is the first take of prep-for-new-EH patchset. This patchset contains 22 patches and can be grouped as follows. #01-02: SCSI support for libata EH #03-09: various fixes #10-22: various new features This patchset is against upstream (acc696d93dcf993dec123d69d599979e1456ffec) New features worth mentioning are... * Use statically allocated buffers. dev->id[] is now statically allocated and ap->sector_buf[512] is added so that revalidation and NCQ read log don't have to allocate buffers during EH. With this no allocation is necessary during EH except for PM link allocation which is done with GFP_NOIO and doesn't hamper operation of existing devices when it fails. * qc->result_tf to carry resulting TF after command excution. TF can be explicitly requested via ATA_QCFLAG_RESULT_TF. LLDDs are responsible for filling it when a qc fails (done automatically for drivers implementing ->tf_read). This change is necessary to natively support controllers which don't have single TF image like sil24 and makes the completion path more efficient. * New SCR access and on/offline test functions defined. This cleans up scattered SCR availability tests and allows for later PM extensions. * Added dev->ap and remove many @ap arguments. This simplifies codes and makes it much easier to add ata_link inbetween later. * ata_*_printk() helpers. links and devices will be identified differently on PM and no-PM cases. Adding printk helpers seems to be the cleanest solution. Maybe debug message handling can be integrated, too. I don't know. Thanks. -- tejun [T] http://article.gmane.org/gmane.linux.ide/9957