From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S965216AbWEOUMa (ORCPT ); Mon, 15 May 2006 16:12:30 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S965215AbWEOUM3 (ORCPT ); Mon, 15 May 2006 16:12:29 -0400 Received: from einhorn.in-berlin.de ([192.109.42.8]:47762 "EHLO einhorn.in-berlin.de") by vger.kernel.org with ESMTP id S965214AbWEOUM2 (ORCPT ); Mon, 15 May 2006 16:12:28 -0400 X-Envelope-From: stefanr@s5r6.in-berlin.de Date: Mon, 15 May 2006 22:09:46 +0200 (CEST) From: Stefan Richter Subject: [PATCH 4/4 resend] ohci1394, sbp2: fix "scsi_add_device failed" with PL-3507 based devices To: Linus Torvalds cc: linux-kernel@vger.kernel.org, Andrew Morton , Jody McIntyre , Ben Collins In-Reply-To: Message-ID: References: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; CHARSET=us-ascii Content-Disposition: INLINE X-Spam-Score: (0.877) AWL,BAYES_50 Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Re-enable posted writes for status FIFO. Besides bringing back a very minor bandwidth tweak from Linux 2.6.15.x and older, this also fixes an interoperability regression since 2.6.16: http://bugzilla.kernel.org/show_bug.cgi?id=6356 (sbp2: scsi_add_device failed. IEEE1394 HD is not working anymore.) Signed-off-by: Stefan Richter Tested-by: Vanei Heidemann Tested-by: Martin Putzlocher (chip type unconfirmed) --- The ugly hardwired PhyUpperBound which this patch (re-)introduces into sbp2 will be eliminated by a subsequent patchset which has already been sent to linux1394-devel. First posted on 2006-04-23. Index: linux-2.6.17-rc4/drivers/ieee1394/ohci1394.c =================================================================== --- linux-2.6.17-rc4.orig/drivers/ieee1394/ohci1394.c 2006-05-15 21:27:04.000000000 +0200 +++ linux-2.6.17-rc4/drivers/ieee1394/ohci1394.c 2006-05-15 21:52:42.000000000 +0200 @@ -553,7 +553,7 @@ static void ohci_initialize(struct ti_oh * register content. * To actually enable physical responses is the job of our interrupt * handler which programs the physical request filter. */ - reg_write(ohci, OHCI1394_PhyUpperBound, 0xffff0000); + reg_write(ohci, OHCI1394_PhyUpperBound, 0x01000000); DBGMSG("physUpperBoundOffset=%08x", reg_read(ohci, OHCI1394_PhyUpperBound)); Index: linux-2.6.17-rc4/drivers/ieee1394/sbp2.c =================================================================== --- linux-2.6.17-rc4.orig/drivers/ieee1394/sbp2.c 2006-05-15 21:52:18.000000000 +0200 +++ linux-2.6.17-rc4/drivers/ieee1394/sbp2.c 2006-05-15 21:52:42.000000000 +0200 @@ -835,11 +835,16 @@ static struct scsi_id_instance_data *sbp /* Register the status FIFO address range. We could use the same FIFO * for targets at different nodes. However we need different FIFOs per - * target in order to support multi-unit devices. */ + * target in order to support multi-unit devices. + * The FIFO is located out of the local host controller's physical range + * but, if possible, within the posted write area. Status writes will + * then be performed as unified transactions. This slightly reduces + * bandwidth usage, and some Prolific based devices seem to require it. + */ scsi_id->status_fifo_addr = hpsb_allocate_and_register_addrspace( &sbp2_highlevel, ud->ne->host, &sbp2_ops, sizeof(struct sbp2_status_block), sizeof(quadlet_t), - ~0ULL, ~0ULL); + 0x010000000000ULL, CSR1212_ALL_SPACE_END); if (!scsi_id->status_fifo_addr) { SBP2_ERR("failed to allocate status FIFO address range"); goto failed_alloc;