From: Alan Cox <alan@redhat.com>
Subject: [PATCH] Fix bug #222653 (on Exception) Serverworks data corruptor
Date: Wed, 6 Jun 2007 10:16:37 -0400
Bugzilla: 222653
Message-Id: <20070606141637.GG4678@devserv.devel.redhat.com>
Changelog: [ide] Serverworks data corruptor


Serverworks IDE passes mode information from the BIOS to the driver (which then
mostly interprets it right but not entirely). However in RAID mode some of the 
RAID firmware leaves enough bits to convince the driver to use the firmware
settings but not actual valid data and this causes corruption.

The fix is pretty clean and upstream (upstream in fact now plans to move on
from this to drop all firmware based tuning and always do the job itself as the
new libata driver does). If the chip is in RAID mode the dev->class is not set
to STORAGE_IDE, and in that case we skip the oem setup and tune the chipset by
hand.

--- linux-2.6.18.x86_64/drivers/ide/pci/serverworks.c~	2007-06-06 15:08:51.000000000 +0100
+++ linux-2.6.18.x86_64/drivers/ide/pci/serverworks.c	2007-06-06 15:08:51.000000000 +0100
@@ -156,6 +156,12 @@
 	pci_read_config_word(dev, 0x4A, &csb5_pio);
 	pci_read_config_byte(dev, 0x54, &ultra_enable);
 
+	/* If we are in RAID mode (eg AMI MegaIDE) then we can't it
+	   turns out trust the firmware configuration */
+
+	if ((dev->class >> 8) != PCI_CLASS_STORAGE_IDE)
+		goto oem_setup_failed;
+
 	/* Per Specified Design by OEM, and ASIC Architect */
 	if ((dev->device == PCI_DEVICE_ID_SERVERWORKS_CSB6IDE) ||
 	    (dev->device == PCI_DEVICE_ID_SERVERWORKS_CSB6IDE2)) {