android_kernel_motorola_sm6225/drivers/ata/pata_serverworks.c

606 lines
16 KiB
C
Raw Normal View History

/*
* pata_serverworks.c - Serverworks PATA for new ATA layer
* (C) 2005 Red Hat Inc
* Alan Cox <alan@redhat.com>
*
* based upon
*
* serverworks.c
*
* Copyright (C) 1998-2000 Michel Aubry
* Copyright (C) 1998-2000 Andrzej Krzysztofowicz
* Copyright (C) 1998-2000 Andre Hedrick <andre@linux-ide.org>
* Portions copyright (c) 2001 Sun Microsystems
*
*
* RCC/ServerWorks IDE driver for Linux
*
* OSB4: `Open South Bridge' IDE Interface (fn 1)
* supports UDMA mode 2 (33 MB/s)
*
* CSB5: `Champion South Bridge' IDE Interface (fn 1)
* all revisions support UDMA mode 4 (66 MB/s)
* revision A2.0 and up support UDMA mode 5 (100 MB/s)
*
* *** The CSB5 does not provide ANY register ***
* *** to detect 80-conductor cable presence. ***
*
* CSB6: `Champion South Bridge' IDE Interface (optional: third channel)
*
* Documentation:
* Available under NDA only. Errata info very hard to get.
*/
#include <linux/kernel.h>
#include <linux/module.h>
#include <linux/pci.h>
#include <linux/init.h>
#include <linux/blkdev.h>
#include <linux/delay.h>
#include <scsi/scsi_host.h>
#include <linux/libata.h>
#define DRV_NAME "pata_serverworks"
#define DRV_VERSION "0.4.2"
#define SVWKS_CSB5_REVISION_NEW 0x92 /* min PCI_REVISION_ID for UDMA5 (A2.0) */
#define SVWKS_CSB6_REVISION 0xa0 /* min PCI_REVISION_ID for UDMA4 (A1.0) */
/* Seagate Barracuda ATA IV Family drives in UDMA mode 5
* can overrun their FIFOs when used with the CSB5 */
static const char *csb_bad_ata100[] = {
"ST320011A",
"ST340016A",
"ST360021A",
"ST380021A",
NULL
};
/**
* dell_cable - Dell serverworks cable detection
* @ap: ATA port to do cable detect
*
* Dell hide the 40/80 pin select for their interfaces in the top two
* bits of the subsystem ID.
*/
static int dell_cable(struct ata_port *ap) {
struct pci_dev *pdev = to_pci_dev(ap->host->dev);
if (pdev->subsystem_device & (1 << (ap->port_no + 14)))
return ATA_CBL_PATA80;
return ATA_CBL_PATA40;
}
/**
* sun_cable - Sun Cobalt 'Alpine' cable detection
* @ap: ATA port to do cable select
*
* Cobalt CSB5 IDE hides the 40/80pin in the top two bits of the
* subsystem ID the same as dell. We could use one function but we may
* need to extend the Dell one in future
*/
static int sun_cable(struct ata_port *ap) {
struct pci_dev *pdev = to_pci_dev(ap->host->dev);
if (pdev->subsystem_device & (1 << (ap->port_no + 14)))
return ATA_CBL_PATA80;
return ATA_CBL_PATA40;
}
/**
* osb4_cable - OSB4 cable detect
* @ap: ATA port to check
*
* The OSB4 isn't UDMA66 capable so this is easy
*/
static int osb4_cable(struct ata_port *ap) {
return ATA_CBL_PATA40;
}
/**
* csb4_cable - CSB5/6 cable detect
* @ap: ATA port to check
*
* Serverworks default arrangement is to use the drive side detection
* only.
*/
static int csb_cable(struct ata_port *ap) {
return ATA_CBL_PATA80;
}
struct sv_cable_table {
int device;
int subvendor;
int (*cable_detect)(struct ata_port *ap);
};
/*
* Note that we don't copy the old serverworks code because the old
* code contains obvious mistakes
*/
static struct sv_cable_table cable_detect[] = {
{ PCI_DEVICE_ID_SERVERWORKS_CSB5IDE, PCI_VENDOR_ID_DELL, dell_cable },
{ PCI_DEVICE_ID_SERVERWORKS_CSB6IDE, PCI_VENDOR_ID_DELL, dell_cable },
{ PCI_DEVICE_ID_SERVERWORKS_CSB5IDE, PCI_VENDOR_ID_SUN, sun_cable },
{ PCI_DEVICE_ID_SERVERWORKS_OSB4IDE, PCI_ANY_ID, osb4_cable },
{ PCI_DEVICE_ID_SERVERWORKS_CSB5IDE, PCI_ANY_ID, csb_cable },
{ PCI_DEVICE_ID_SERVERWORKS_CSB6IDE, PCI_ANY_ID, csb_cable },
{ PCI_DEVICE_ID_SERVERWORKS_CSB6IDE2, PCI_ANY_ID, csb_cable },
{ PCI_DEVICE_ID_SERVERWORKS_HT1000IDE, PCI_ANY_ID, csb_cable },
{ }
};
/**
* serverworks_cable_detect - cable detection
* @ap: ATA port
libata: add deadline support to prereset and reset methods Add @deadline to prereset and reset methods and make them honor it. ata_wait_ready() which directly takes @deadline is implemented to be used as the wait function. This patch is in preparation for EH timing improvements. * ata_wait_ready() never does busy sleep. It's only used from EH and no wait in EH is that urgent. This function also prints 'be patient' message automatically after 5 secs of waiting if more than 3 secs is remaining till deadline. * ata_bus_post_reset() now fails with error code if any of its wait fails. This is important because earlier reset tries will have shorter timeout than the spec requires. If a device fails to respond before the short timeout, reset should be retried with longer timeout rather than silently ignoring the device. There are three behavior differences. 1. Timeout is applied to both devices at once, not separately. This is more consistent with what the spec says. 2. When a device passes devchk but fails to become ready before deadline. Previouly, post_reset would just succeed and let device classification remove the device. New code fails the reset thus causing reset retry. After a few times, EH will give up disabling the port. 3. When slave device passes devchk but fails to become accessible (TF-wise) after reset. Original code disables dev1 after 30s timeout and continues as if the device doesn't exist, while the patched code fails reset. When this happens, new code fails reset on whole port rather than proceeding with only the primary device. If the failing device is suffering transient problems, new code retries reset which is a better behavior. If the failing device is actually broken, the net effect is identical to it, but not to the other device sharing the channel. In the previous code, reset would have succeeded after 30s thus detecting the working one. In the new code, reset fails and whole port gets disabled. IMO, it's a pathological case anyway (broken device sharing bus with working one) and doesn't really matter. * ata_bus_softreset() is changed to return error code from ata_bus_post_reset(). It used to return 0 unconditionally. * Spin up waiting is to be removed and not converted to honor deadline. * To be on the safe side, deadline is set to 40s for the time being. Signed-off-by: Tejun Heo <htejun@gmail.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-02-02 08:50:52 +01:00
* @deadline: deadline jiffies for the operation
*
* Perform cable detection according to the device and subvendor
* identifications
*/
libata: add deadline support to prereset and reset methods Add @deadline to prereset and reset methods and make them honor it. ata_wait_ready() which directly takes @deadline is implemented to be used as the wait function. This patch is in preparation for EH timing improvements. * ata_wait_ready() never does busy sleep. It's only used from EH and no wait in EH is that urgent. This function also prints 'be patient' message automatically after 5 secs of waiting if more than 3 secs is remaining till deadline. * ata_bus_post_reset() now fails with error code if any of its wait fails. This is important because earlier reset tries will have shorter timeout than the spec requires. If a device fails to respond before the short timeout, reset should be retried with longer timeout rather than silently ignoring the device. There are three behavior differences. 1. Timeout is applied to both devices at once, not separately. This is more consistent with what the spec says. 2. When a device passes devchk but fails to become ready before deadline. Previouly, post_reset would just succeed and let device classification remove the device. New code fails the reset thus causing reset retry. After a few times, EH will give up disabling the port. 3. When slave device passes devchk but fails to become accessible (TF-wise) after reset. Original code disables dev1 after 30s timeout and continues as if the device doesn't exist, while the patched code fails reset. When this happens, new code fails reset on whole port rather than proceeding with only the primary device. If the failing device is suffering transient problems, new code retries reset which is a better behavior. If the failing device is actually broken, the net effect is identical to it, but not to the other device sharing the channel. In the previous code, reset would have succeeded after 30s thus detecting the working one. In the new code, reset fails and whole port gets disabled. IMO, it's a pathological case anyway (broken device sharing bus with working one) and doesn't really matter. * ata_bus_softreset() is changed to return error code from ata_bus_post_reset(). It used to return 0 unconditionally. * Spin up waiting is to be removed and not converted to honor deadline. * To be on the safe side, deadline is set to 40s for the time being. Signed-off-by: Tejun Heo <htejun@gmail.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-02-02 08:50:52 +01:00
static int serverworks_cable_detect(struct ata_port *ap)
{
struct pci_dev *pdev = to_pci_dev(ap->host->dev);
struct sv_cable_table *cb = cable_detect;
while(cb->device) {
if (cb->device == pdev->device &&
(cb->subvendor == pdev->subsystem_vendor ||
cb->subvendor == PCI_ANY_ID)) {
return cb->cable_detect(ap);
}
cb++;
}
BUG();
return -1; /* kill compiler warning */
}
/**
* serverworks_is_csb - Check for CSB or OSB
* @pdev: PCI device to check
*
* Returns true if the device being checked is known to be a CSB
* series device.
*/
static u8 serverworks_is_csb(struct pci_dev *pdev)
{
switch (pdev->device) {
case PCI_DEVICE_ID_SERVERWORKS_CSB5IDE:
case PCI_DEVICE_ID_SERVERWORKS_CSB6IDE:
case PCI_DEVICE_ID_SERVERWORKS_CSB6IDE2:
case PCI_DEVICE_ID_SERVERWORKS_HT1000IDE:
return 1;
default:
break;
}
return 0;
}
/**
* serverworks_osb4_filter - mode selection filter
* @adev: ATA device
* @mask: Mask of proposed modes
*
* Filter the offered modes for the device to apply controller
* specific rules. OSB4 requires no UDMA for disks due to a FIFO
* bug we hit.
*/
static unsigned long serverworks_osb4_filter(struct ata_device *adev, unsigned long mask)
{
if (adev->class == ATA_DEV_ATA)
mask &= ~ATA_MASK_UDMA;
return ata_pci_default_filter(adev, mask);
}
/**
* serverworks_csb_filter - mode selection filter
* @adev: ATA device
* @mask: Mask of proposed modes
*
* Check the blacklist and disable UDMA5 if matched
*/
static unsigned long serverworks_csb_filter(struct ata_device *adev, unsigned long mask)
{
const char *p;
char model_num[ATA_ID_PROD_LEN + 1];
int i;
/* Disk, UDMA */
if (adev->class != ATA_DEV_ATA)
return ata_pci_default_filter(adev, mask);
/* Actually do need to check */
ata_id_c_string(adev->id, model_num, ATA_ID_PROD, sizeof(model_num));
for (i = 0; (p = csb_bad_ata100[i]) != NULL; i++) {
if (!strcmp(p, model_num))
mask &= ~(0x1F << ATA_SHIFT_UDMA);
}
return ata_pci_default_filter(adev, mask);
}
/**
* serverworks_set_piomode - set initial PIO mode data
* @ap: ATA interface
* @adev: ATA device
*
* Program the OSB4/CSB5 timing registers for PIO. The PIO register
* load is done as a simple lookup.
*/
static void serverworks_set_piomode(struct ata_port *ap, struct ata_device *adev)
{
static const u8 pio_mode[] = { 0x5d, 0x47, 0x34, 0x22, 0x20 };
int offset = 1 + (2 * ap->port_no) - adev->devno;
int devbits = (2 * ap->port_no + adev->devno) * 4;
u16 csb5_pio;
struct pci_dev *pdev = to_pci_dev(ap->host->dev);
int pio = adev->pio_mode - XFER_PIO_0;
pci_write_config_byte(pdev, 0x40 + offset, pio_mode[pio]);
/* The OSB4 just requires the timing but the CSB series want the
mode number as well */
if (serverworks_is_csb(pdev)) {
pci_read_config_word(pdev, 0x4A, &csb5_pio);
csb5_pio &= ~(0x0F << devbits);
pci_write_config_byte(pdev, 0x4A, csb5_pio | (pio << devbits));
}
}
/**
* serverworks_set_dmamode - set initial DMA mode data
* @ap: ATA interface
* @adev: ATA device
*
* Program the MWDMA/UDMA modes for the serverworks OSB4/CSB5
* chipset. The MWDMA mode values are pulled from a lookup table
* while the chipset uses mode number for UDMA.
*/
static void serverworks_set_dmamode(struct ata_port *ap, struct ata_device *adev)
{
static const u8 dma_mode[] = { 0x77, 0x21, 0x20 };
int offset = 1 + 2 * ap->port_no - adev->devno;
int devbits = (2 * ap->port_no + adev->devno);
u8 ultra;
u8 ultra_cfg;
struct pci_dev *pdev = to_pci_dev(ap->host->dev);
pci_read_config_byte(pdev, 0x54, &ultra_cfg);
if (adev->dma_mode >= XFER_UDMA_0) {
pci_write_config_byte(pdev, 0x44 + offset, 0x20);
pci_read_config_byte(pdev, 0x56 + ap->port_no, &ultra);
ultra &= ~(0x0F << (ap->port_no * 4));
ultra |= (adev->dma_mode - XFER_UDMA_0)
<< (ap->port_no * 4);
pci_write_config_byte(pdev, 0x56 + ap->port_no, ultra);
ultra_cfg |= (1 << devbits);
} else {
pci_write_config_byte(pdev, 0x44 + offset,
dma_mode[adev->dma_mode - XFER_MW_DMA_0]);
ultra_cfg &= ~(1 << devbits);
}
pci_write_config_byte(pdev, 0x54, ultra_cfg);
}
static struct scsi_host_template serverworks_sht = {
.module = THIS_MODULE,
.name = DRV_NAME,
.ioctl = ata_scsi_ioctl,
.queuecommand = ata_scsi_queuecmd,
.can_queue = ATA_DEF_QUEUE,
.this_id = ATA_SHT_THIS_ID,
.sg_tablesize = LIBATA_MAX_PRD,
.cmd_per_lun = ATA_SHT_CMD_PER_LUN,
.emulated = ATA_SHT_EMULATED,
.use_clustering = ATA_SHT_USE_CLUSTERING,
.proc_name = DRV_NAME,
.dma_boundary = ATA_DMA_BOUNDARY,
.slave_configure = ata_scsi_slave_config,
.slave_destroy = ata_scsi_slave_destroy,
.bios_param = ata_std_bios_param,
};
static struct ata_port_operations serverworks_osb4_port_ops = {
.port_disable = ata_port_disable,
.set_piomode = serverworks_set_piomode,
.set_dmamode = serverworks_set_dmamode,
.mode_filter = serverworks_osb4_filter,
.tf_load = ata_tf_load,
.tf_read = ata_tf_read,
.check_status = ata_check_status,
.exec_command = ata_exec_command,
.dev_select = ata_std_dev_select,
.freeze = ata_bmdma_freeze,
.thaw = ata_bmdma_thaw,
.error_handler = ata_bmdma_error_handler,
.post_internal_cmd = ata_bmdma_post_internal_cmd,
.cable_detect = serverworks_cable_detect,
.bmdma_setup = ata_bmdma_setup,
.bmdma_start = ata_bmdma_start,
.bmdma_stop = ata_bmdma_stop,
.bmdma_status = ata_bmdma_status,
.qc_prep = ata_qc_prep,
.qc_issue = ata_qc_issue_prot,
.data_xfer = ata_data_xfer,
.irq_handler = ata_interrupt,
.irq_clear = ata_bmdma_irq_clear,
.irq_on = ata_irq_on,
.irq_ack = ata_irq_ack,
.port_start = ata_port_start,
};
static struct ata_port_operations serverworks_csb_port_ops = {
.port_disable = ata_port_disable,
.set_piomode = serverworks_set_piomode,
.set_dmamode = serverworks_set_dmamode,
.mode_filter = serverworks_csb_filter,
.tf_load = ata_tf_load,
.tf_read = ata_tf_read,
.check_status = ata_check_status,
.exec_command = ata_exec_command,
.dev_select = ata_std_dev_select,
.freeze = ata_bmdma_freeze,
.thaw = ata_bmdma_thaw,
.error_handler = ata_bmdma_error_handler,
.post_internal_cmd = ata_bmdma_post_internal_cmd,
.cable_detect = serverworks_cable_detect,
.bmdma_setup = ata_bmdma_setup,
.bmdma_start = ata_bmdma_start,
.bmdma_stop = ata_bmdma_stop,
.bmdma_status = ata_bmdma_status,
.qc_prep = ata_qc_prep,
.qc_issue = ata_qc_issue_prot,
.data_xfer = ata_data_xfer,
.irq_handler = ata_interrupt,
.irq_clear = ata_bmdma_irq_clear,
.irq_on = ata_irq_on,
.irq_ack = ata_irq_ack,
.port_start = ata_port_start,
};
static int serverworks_fixup_osb4(struct pci_dev *pdev)
{
u32 reg;
struct pci_dev *isa_dev = pci_get_device(PCI_VENDOR_ID_SERVERWORKS,
PCI_DEVICE_ID_SERVERWORKS_OSB4, NULL);
if (isa_dev) {
pci_read_config_dword(isa_dev, 0x64, &reg);
reg &= ~0x00002000; /* disable 600ns interrupt mask */
if (!(reg & 0x00004000))
printk(KERN_DEBUG DRV_NAME ": UDMA not BIOS enabled.\n");
reg |= 0x00004000; /* enable UDMA/33 support */
pci_write_config_dword(isa_dev, 0x64, reg);
pci_dev_put(isa_dev);
return 0;
}
printk(KERN_WARNING "ata_serverworks: Unable to find bridge.\n");
return -ENODEV;
}
static int serverworks_fixup_csb(struct pci_dev *pdev)
{
u8 btr;
/* Third Channel Test */
if (!(PCI_FUNC(pdev->devfn) & 1)) {
struct pci_dev * findev = NULL;
u32 reg4c = 0;
findev = pci_get_device(PCI_VENDOR_ID_SERVERWORKS,
PCI_DEVICE_ID_SERVERWORKS_CSB5, NULL);
if (findev) {
pci_read_config_dword(findev, 0x4C, &reg4c);
reg4c &= ~0x000007FF;
reg4c |= 0x00000040;
reg4c |= 0x00000020;
pci_write_config_dword(findev, 0x4C, reg4c);
pci_dev_put(findev);
}
} else {
struct pci_dev * findev = NULL;
u8 reg41 = 0;
findev = pci_get_device(PCI_VENDOR_ID_SERVERWORKS,
PCI_DEVICE_ID_SERVERWORKS_CSB6, NULL);
if (findev) {
pci_read_config_byte(findev, 0x41, &reg41);
reg41 &= ~0x40;
pci_write_config_byte(findev, 0x41, reg41);
pci_dev_put(findev);
}
}
/* setup the UDMA Control register
*
* 1. clear bit 6 to enable DMA
* 2. enable DMA modes with bits 0-1
* 00 : legacy
* 01 : udma2
* 10 : udma2/udma4
* 11 : udma2/udma4/udma5
*/
pci_read_config_byte(pdev, 0x5A, &btr);
btr &= ~0x40;
if (!(PCI_FUNC(pdev->devfn) & 1))
btr |= 0x2;
else
btr |= (pdev->revision >= SVWKS_CSB5_REVISION_NEW) ? 0x3 : 0x2;
pci_write_config_byte(pdev, 0x5A, btr);
return btr;
}
static void serverworks_fixup_ht1000(struct pci_dev *pdev)
{
u8 btr;
/* Setup HT1000 SouthBridge Controller - Single Channel Only */
pci_read_config_byte(pdev, 0x5A, &btr);
btr &= ~0x40;
btr |= 0x3;
pci_write_config_byte(pdev, 0x5A, btr);
}
static int serverworks_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
{
libata: clean up SFF init mess The intention of using port_mask in SFF init helpers was to eventually support exoctic configurations such as combination of legacy and native port on the same controller. This never became actually necessary and the related code always has been subtly broken one way or the other. Now that new init model is in place, there is no reason to make common helpers capable of handling all corner cases. Exotic cases can simply dealt within LLDs as necessary. This patch removes port_mask handling in SFF init helpers. SFF init helpers don't take n_ports argument and interpret it into port_mask anymore. All information is carried via port_info. n_ports argument is dropped and always two ports are allocated. LLD can tell SFF to skip certain port by marking it dummy. Note that SFF code has been treating unuvailable ports this way for a long time until recent breakage fix from Linus and is consistent with how other drivers handle with unavailable ports. This fixes 1-port legacy host handling still broken after the recent native mode fix and simplifies SFF init logic. The following changes are made... * ata_pci_init_native_host() and ata_init_legacy_host() both now try to initialized whatever they can and mark failed ports dummy. They return 0 if any port is successfully initialized. * ata_pci_prepare_native_host() and ata_pci_init_one() now doesn't take n_ports argument. All info should be specified via port_info array. Always two ports are allocated. * ata_pci_init_bmdma() exported to be used by LLDs in exotic cases. * port_info handling in all LLDs are standardized - all port_info arrays are const stack variable named ppi. Unless the second port is different from the first, its port_info is specified as NULL (tells libata that it's identical to the last non-NULL port_info). * pata_hpt37x/hpt3x2n: don't modify static variable directly. Make an on-stack copy instead as ata_piix does. * pata_uli: It has 4 ports instead of 2. Don't use ata_pci_prepare_native_host(). Allocate the host explicitly and use init helpers. It's simple enough. Signed-off-by: Tejun Heo <htejun@gmail.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-05-04 12:43:58 +02:00
static const struct ata_port_info info[4] = {
{ /* OSB4 */
.sht = &serverworks_sht,
.flags = ATA_FLAG_SLAVE_POSS,
.pio_mask = 0x1f,
.mwdma_mask = 0x07,
.udma_mask = 0x07,
.port_ops = &serverworks_osb4_port_ops
}, { /* OSB4 no UDMA */
.sht = &serverworks_sht,
.flags = ATA_FLAG_SLAVE_POSS,
.pio_mask = 0x1f,
.mwdma_mask = 0x07,
.udma_mask = 0x00,
.port_ops = &serverworks_osb4_port_ops
}, { /* CSB5 */
.sht = &serverworks_sht,
.flags = ATA_FLAG_SLAVE_POSS,
.pio_mask = 0x1f,
.mwdma_mask = 0x07,
.udma_mask = ATA_UDMA4,
.port_ops = &serverworks_csb_port_ops
}, { /* CSB5 - later revisions*/
.sht = &serverworks_sht,
.flags = ATA_FLAG_SLAVE_POSS,
.pio_mask = 0x1f,
.mwdma_mask = 0x07,
.udma_mask = ATA_UDMA5,
.port_ops = &serverworks_csb_port_ops
}
};
libata: clean up SFF init mess The intention of using port_mask in SFF init helpers was to eventually support exoctic configurations such as combination of legacy and native port on the same controller. This never became actually necessary and the related code always has been subtly broken one way or the other. Now that new init model is in place, there is no reason to make common helpers capable of handling all corner cases. Exotic cases can simply dealt within LLDs as necessary. This patch removes port_mask handling in SFF init helpers. SFF init helpers don't take n_ports argument and interpret it into port_mask anymore. All information is carried via port_info. n_ports argument is dropped and always two ports are allocated. LLD can tell SFF to skip certain port by marking it dummy. Note that SFF code has been treating unuvailable ports this way for a long time until recent breakage fix from Linus and is consistent with how other drivers handle with unavailable ports. This fixes 1-port legacy host handling still broken after the recent native mode fix and simplifies SFF init logic. The following changes are made... * ata_pci_init_native_host() and ata_init_legacy_host() both now try to initialized whatever they can and mark failed ports dummy. They return 0 if any port is successfully initialized. * ata_pci_prepare_native_host() and ata_pci_init_one() now doesn't take n_ports argument. All info should be specified via port_info array. Always two ports are allocated. * ata_pci_init_bmdma() exported to be used by LLDs in exotic cases. * port_info handling in all LLDs are standardized - all port_info arrays are const stack variable named ppi. Unless the second port is different from the first, its port_info is specified as NULL (tells libata that it's identical to the last non-NULL port_info). * pata_hpt37x/hpt3x2n: don't modify static variable directly. Make an on-stack copy instead as ata_piix does. * pata_uli: It has 4 ports instead of 2. Don't use ata_pci_prepare_native_host(). Allocate the host explicitly and use init helpers. It's simple enough. Signed-off-by: Tejun Heo <htejun@gmail.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-05-04 12:43:58 +02:00
const struct ata_port_info *ppi[] = { &info[id->driver_data], NULL };
/* Force master latency timer to 64 PCI clocks */
pci_write_config_byte(pdev, PCI_LATENCY_TIMER, 0x40);
/* OSB4 : South Bridge and IDE */
if (pdev->device == PCI_DEVICE_ID_SERVERWORKS_OSB4IDE) {
/* Select non UDMA capable OSB4 if we can't do fixups */
if ( serverworks_fixup_osb4(pdev) < 0)
libata: clean up SFF init mess The intention of using port_mask in SFF init helpers was to eventually support exoctic configurations such as combination of legacy and native port on the same controller. This never became actually necessary and the related code always has been subtly broken one way or the other. Now that new init model is in place, there is no reason to make common helpers capable of handling all corner cases. Exotic cases can simply dealt within LLDs as necessary. This patch removes port_mask handling in SFF init helpers. SFF init helpers don't take n_ports argument and interpret it into port_mask anymore. All information is carried via port_info. n_ports argument is dropped and always two ports are allocated. LLD can tell SFF to skip certain port by marking it dummy. Note that SFF code has been treating unuvailable ports this way for a long time until recent breakage fix from Linus and is consistent with how other drivers handle with unavailable ports. This fixes 1-port legacy host handling still broken after the recent native mode fix and simplifies SFF init logic. The following changes are made... * ata_pci_init_native_host() and ata_init_legacy_host() both now try to initialized whatever they can and mark failed ports dummy. They return 0 if any port is successfully initialized. * ata_pci_prepare_native_host() and ata_pci_init_one() now doesn't take n_ports argument. All info should be specified via port_info array. Always two ports are allocated. * ata_pci_init_bmdma() exported to be used by LLDs in exotic cases. * port_info handling in all LLDs are standardized - all port_info arrays are const stack variable named ppi. Unless the second port is different from the first, its port_info is specified as NULL (tells libata that it's identical to the last non-NULL port_info). * pata_hpt37x/hpt3x2n: don't modify static variable directly. Make an on-stack copy instead as ata_piix does. * pata_uli: It has 4 ports instead of 2. Don't use ata_pci_prepare_native_host(). Allocate the host explicitly and use init helpers. It's simple enough. Signed-off-by: Tejun Heo <htejun@gmail.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-05-04 12:43:58 +02:00
ppi[0] = &info[1];
}
/* setup CSB5/CSB6 : South Bridge and IDE option RAID */
else if ((pdev->device == PCI_DEVICE_ID_SERVERWORKS_CSB5IDE) ||
(pdev->device == PCI_DEVICE_ID_SERVERWORKS_CSB6IDE) ||
(pdev->device == PCI_DEVICE_ID_SERVERWORKS_CSB6IDE2)) {
/* If the returned btr is the newer revision then
select the right info block */
if (serverworks_fixup_csb(pdev) == 3)
libata: clean up SFF init mess The intention of using port_mask in SFF init helpers was to eventually support exoctic configurations such as combination of legacy and native port on the same controller. This never became actually necessary and the related code always has been subtly broken one way or the other. Now that new init model is in place, there is no reason to make common helpers capable of handling all corner cases. Exotic cases can simply dealt within LLDs as necessary. This patch removes port_mask handling in SFF init helpers. SFF init helpers don't take n_ports argument and interpret it into port_mask anymore. All information is carried via port_info. n_ports argument is dropped and always two ports are allocated. LLD can tell SFF to skip certain port by marking it dummy. Note that SFF code has been treating unuvailable ports this way for a long time until recent breakage fix from Linus and is consistent with how other drivers handle with unavailable ports. This fixes 1-port legacy host handling still broken after the recent native mode fix and simplifies SFF init logic. The following changes are made... * ata_pci_init_native_host() and ata_init_legacy_host() both now try to initialized whatever they can and mark failed ports dummy. They return 0 if any port is successfully initialized. * ata_pci_prepare_native_host() and ata_pci_init_one() now doesn't take n_ports argument. All info should be specified via port_info array. Always two ports are allocated. * ata_pci_init_bmdma() exported to be used by LLDs in exotic cases. * port_info handling in all LLDs are standardized - all port_info arrays are const stack variable named ppi. Unless the second port is different from the first, its port_info is specified as NULL (tells libata that it's identical to the last non-NULL port_info). * pata_hpt37x/hpt3x2n: don't modify static variable directly. Make an on-stack copy instead as ata_piix does. * pata_uli: It has 4 ports instead of 2. Don't use ata_pci_prepare_native_host(). Allocate the host explicitly and use init helpers. It's simple enough. Signed-off-by: Tejun Heo <htejun@gmail.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-05-04 12:43:58 +02:00
ppi[0] = &info[3];
/* Is this the 3rd channel CSB6 IDE ? */
if (pdev->device == PCI_DEVICE_ID_SERVERWORKS_CSB6IDE2)
libata: clean up SFF init mess The intention of using port_mask in SFF init helpers was to eventually support exoctic configurations such as combination of legacy and native port on the same controller. This never became actually necessary and the related code always has been subtly broken one way or the other. Now that new init model is in place, there is no reason to make common helpers capable of handling all corner cases. Exotic cases can simply dealt within LLDs as necessary. This patch removes port_mask handling in SFF init helpers. SFF init helpers don't take n_ports argument and interpret it into port_mask anymore. All information is carried via port_info. n_ports argument is dropped and always two ports are allocated. LLD can tell SFF to skip certain port by marking it dummy. Note that SFF code has been treating unuvailable ports this way for a long time until recent breakage fix from Linus and is consistent with how other drivers handle with unavailable ports. This fixes 1-port legacy host handling still broken after the recent native mode fix and simplifies SFF init logic. The following changes are made... * ata_pci_init_native_host() and ata_init_legacy_host() both now try to initialized whatever they can and mark failed ports dummy. They return 0 if any port is successfully initialized. * ata_pci_prepare_native_host() and ata_pci_init_one() now doesn't take n_ports argument. All info should be specified via port_info array. Always two ports are allocated. * ata_pci_init_bmdma() exported to be used by LLDs in exotic cases. * port_info handling in all LLDs are standardized - all port_info arrays are const stack variable named ppi. Unless the second port is different from the first, its port_info is specified as NULL (tells libata that it's identical to the last non-NULL port_info). * pata_hpt37x/hpt3x2n: don't modify static variable directly. Make an on-stack copy instead as ata_piix does. * pata_uli: It has 4 ports instead of 2. Don't use ata_pci_prepare_native_host(). Allocate the host explicitly and use init helpers. It's simple enough. Signed-off-by: Tejun Heo <htejun@gmail.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-05-04 12:43:58 +02:00
ppi[1] = &ata_dummy_port_info;
}
/* setup HT1000E */
else if (pdev->device == PCI_DEVICE_ID_SERVERWORKS_HT1000IDE)
serverworks_fixup_ht1000(pdev);
if (pdev->device == PCI_DEVICE_ID_SERVERWORKS_CSB5IDE)
ata_pci_clear_simplex(pdev);
libata: clean up SFF init mess The intention of using port_mask in SFF init helpers was to eventually support exoctic configurations such as combination of legacy and native port on the same controller. This never became actually necessary and the related code always has been subtly broken one way or the other. Now that new init model is in place, there is no reason to make common helpers capable of handling all corner cases. Exotic cases can simply dealt within LLDs as necessary. This patch removes port_mask handling in SFF init helpers. SFF init helpers don't take n_ports argument and interpret it into port_mask anymore. All information is carried via port_info. n_ports argument is dropped and always two ports are allocated. LLD can tell SFF to skip certain port by marking it dummy. Note that SFF code has been treating unuvailable ports this way for a long time until recent breakage fix from Linus and is consistent with how other drivers handle with unavailable ports. This fixes 1-port legacy host handling still broken after the recent native mode fix and simplifies SFF init logic. The following changes are made... * ata_pci_init_native_host() and ata_init_legacy_host() both now try to initialized whatever they can and mark failed ports dummy. They return 0 if any port is successfully initialized. * ata_pci_prepare_native_host() and ata_pci_init_one() now doesn't take n_ports argument. All info should be specified via port_info array. Always two ports are allocated. * ata_pci_init_bmdma() exported to be used by LLDs in exotic cases. * port_info handling in all LLDs are standardized - all port_info arrays are const stack variable named ppi. Unless the second port is different from the first, its port_info is specified as NULL (tells libata that it's identical to the last non-NULL port_info). * pata_hpt37x/hpt3x2n: don't modify static variable directly. Make an on-stack copy instead as ata_piix does. * pata_uli: It has 4 ports instead of 2. Don't use ata_pci_prepare_native_host(). Allocate the host explicitly and use init helpers. It's simple enough. Signed-off-by: Tejun Heo <htejun@gmail.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-05-04 12:43:58 +02:00
return ata_pci_init_one(pdev, ppi);
}
#ifdef CONFIG_PM
static int serverworks_reinit_one(struct pci_dev *pdev)
{
/* Force master latency timer to 64 PCI clocks */
pci_write_config_byte(pdev, PCI_LATENCY_TIMER, 0x40);
switch (pdev->device)
{
case PCI_DEVICE_ID_SERVERWORKS_OSB4IDE:
serverworks_fixup_osb4(pdev);
break;
case PCI_DEVICE_ID_SERVERWORKS_CSB5IDE:
ata_pci_clear_simplex(pdev);
/* fall through */
case PCI_DEVICE_ID_SERVERWORKS_CSB6IDE:
case PCI_DEVICE_ID_SERVERWORKS_CSB6IDE2:
serverworks_fixup_csb(pdev);
break;
case PCI_DEVICE_ID_SERVERWORKS_HT1000IDE:
serverworks_fixup_ht1000(pdev);
break;
}
return ata_pci_device_resume(pdev);
}
#endif
static const struct pci_device_id serverworks[] = {
{ PCI_VDEVICE(SERVERWORKS, PCI_DEVICE_ID_SERVERWORKS_OSB4IDE), 0},
{ PCI_VDEVICE(SERVERWORKS, PCI_DEVICE_ID_SERVERWORKS_CSB5IDE), 2},
{ PCI_VDEVICE(SERVERWORKS, PCI_DEVICE_ID_SERVERWORKS_CSB6IDE), 2},
{ PCI_VDEVICE(SERVERWORKS, PCI_DEVICE_ID_SERVERWORKS_CSB6IDE2), 2},
{ PCI_VDEVICE(SERVERWORKS, PCI_DEVICE_ID_SERVERWORKS_HT1000IDE), 2},
{ },
};
static struct pci_driver serverworks_pci_driver = {
.name = DRV_NAME,
.id_table = serverworks,
.probe = serverworks_init_one,
.remove = ata_pci_remove_one,
#ifdef CONFIG_PM
.suspend = ata_pci_device_suspend,
.resume = serverworks_reinit_one,
#endif
};
static int __init serverworks_init(void)
{
return pci_register_driver(&serverworks_pci_driver);
}
static void __exit serverworks_exit(void)
{
pci_unregister_driver(&serverworks_pci_driver);
}
MODULE_AUTHOR("Alan Cox");
MODULE_DESCRIPTION("low-level driver for Serverworks OSB4/CSB5/CSB6");
MODULE_LICENSE("GPL");
MODULE_DEVICE_TABLE(pci, serverworks);
MODULE_VERSION(DRV_VERSION);
module_init(serverworks_init);
module_exit(serverworks_exit);