* [RFC PATCH] nForce4 ADMA with NCQ: It's aliiiive..
@ 2006-10-07 8:08 Robert Hancock
2006-10-07 9:42 ` Prakash Punnoor
2006-10-07 15:28 ` Jeff Garzik
0 siblings, 2 replies; 23+ messages in thread
From: Robert Hancock @ 2006-10-07 8:08 UTC (permalink / raw)
To: linux-kernel, linux-ide; +Cc: prakash, jeff
[-- Attachment #1: Type: text/plain, Size: 7769 bytes --]
I've been working on the patch for sata_nv ADMA support for nForce4 that
Jeff Garzik has in a git branch. I've gotten it into a state where the
ADMA and NCQ features appear to be working with no obvious problems.
I've attached a patch against 2.6.18-mm2.
The code was mostly in a working state for the non-NCQ case but there
were a number of heinous bugs that prevented NCQ from working, like in
the sg list to APRD conversion code and in the interrupt handler.
This is still in quite an experimental state. It has survived system
boots into Fedora Core 5 and Bonnie++ benchmark runs without blowing up,
but there could still be bugs that could corrupt data, etc. so test with
caution.
There is a module parameter adma_enabled which has to be set to 1 to
enable ADMA on CK804/MCP04 chipsets (either that or hack the code to
make the default 1). I only enabled ADMA on those chipsets and not
MCP51, MCP55 or MCP61 since that was all that the original NVIDIA
version did. I assume there was a reason for this, though maybe not.
Someone with one of these chipsets should probably try it out (replacing
the GENERIC type with CK804 in the PCI device table may be all it takes).
A few outstanding issues:
-Error handling likely needs work. EH works well enough to get past
drive detection but that's likely about all. When I ran into errors
while debugging, it usually locked up the machine when trying to do a
soft reset.
-Error handling is also noisy at the moment (it dumps a bunch of
controller state information).
-Jeff will probably cringe at how I implemented the
bmdma_stop/start/status/setup functions. This kludge of toggling
ATA_FLAG_MMIO off for the call into libata was needed since this
controller is almost what libata calls ATA_FLAG_MMIO, but not quite (the
ATA taskfile registers are MMIO but the BMDMA registers are PIO). This
is also why I needed the patch to libata-sff.c to use the adapter's
bmdma_status function rather than hardcoded ata_bmdma_status.
Some sample dmesg output:
sata_nv 0000:00:07.0: version 2.1
sata_nv 0000:00:07.0: Using ADMA mode
ACPI: PCI Interrupt Link [APSI] enabled at IRQ 23
ACPI: PCI Interrupt 0000:00:07.0[A] -> Link [APSI] -> GSI 23 (level,
low) -> IRQ 23
PCI: Setting latency timer of device 0000:00:07.0 to 64
sata_nv 0000:00:07.0: Resetting port 0
ata1: SATA max UDMA/133 cmd 0xFFFFC20000004480 ctl 0xFFFFC200000044A0
bmdma 0xD800 irq 23
sata_nv 0000:00:07.0: Resetting port 1
ata2: SATA max UDMA/133 cmd 0xFFFFC20000004580 ctl 0xFFFFC200000045A0
bmdma 0xD808 irq 23
scsi0 : sata_nv
PM: Adding info for No Bus:host0
sata_nv 0000:00:07.0: Error handling port 0, notifier 0x0,
notifier_error 0x0, gen_ctl 0x1501000, int pending 0, status 0x700,
dumping CPB states
ata1: SATA link up 1.5 Gbps (SStatus 113 SControl 300)
ata1.00: ATA-6, max UDMA/133, 312581808 sectors: LBA48 NCQ (depth 31/32)
ata1.00: ata1: dev 0 multi count 1
ata1.00: configured for UDMA/133
scsi1 : sata_nv
PM: Adding info for No Bus:host1
sata_nv 0000:00:07.0: Error handling port 1, notifier 0x0,
notifier_error 0x0, gen_ctl 0x1501000, int pending 0, status 0x700,
dumping CPB states
ata2: SATA link down (SStatus 0 SControl 300)
PM: Adding info for No Bus:target0:0:0
scsi 0:0:0:0: Direct-Access ATA ST3160827AS 3.42 PQ: 0 ANSI: 5
PM: Adding info for scsi:0:0:0:0
SCSI device sda: 312581808 512-byte hdwr sectors (160042 MB)
sda: Write Protect is off
sda: Mode Sense: 00 3a 00 00
SCSI device sda: drive cache: write back
SCSI device sda: 312581808 512-byte hdwr sectors (160042 MB)
sda: Write Protect is off
sda: Mode Sense: 00 3a 00 00
SCSI device sda: drive cache: write back
sda: sda1
sd 0:0:0:0: Attached scsi disk sda
sata_nv 0000:00:08.0: Using ADMA mode
ACPI: PCI Interrupt Link [APSJ] enabled at IRQ 22
ACPI: PCI Interrupt 0000:00:08.0[A] -> Link [APSJ] -> GSI 22 (level,
low) -> IRQ 22
PCI: Setting latency timer of device 0000:00:08.0 to 64
sata_nv 0000:00:08.0: Resetting port 0
ata3: SATA max UDMA/133 cmd 0xFFFFC20000006480 ctl 0xFFFFC200000064A0
bmdma 0xC400 irq 22
sata_nv 0000:00:08.0: Resetting port 1
ata4: SATA max UDMA/133 cmd 0xFFFFC20000006580 ctl 0xFFFFC200000065A0
bmdma 0xC408 irq 22
scsi2 : sata_nv
PM: Adding info for No Bus:host2
sata_nv 0000:00:08.0: Error handling port 0, notifier 0x0,
notifier_error 0x0, gen_ctl 0x1501000, int pending 0, status 0x700,
dumping CPB states
ata3: SATA link up 3.0 Gbps (SStatus 123 SControl 300)
ata3.00: ATA-7, max UDMA/133, 625142448 sectors: LBA48 NCQ (depth 31/32)
ata3.00: ata3: dev 0 multi count 1
ata3.00: configured for UDMA/133
scsi3 : sata_nv
PM: Adding info for No Bus:host3
sata_nv 0000:00:08.0: Error handling port 1, notifier 0x0,
notifier_error 0x0, gen_ctl 0x1501000, int pending 0, status 0x700,
dumping CPB states
ata4: SATA link up 1.5 Gbps (SStatus 113 SControl 300)
ata4.00: ATA-6, max UDMA/133, 312581808 sectors: LBA48 NCQ (depth 31/32)
ata4.00: ata4: dev 0 multi count 1
ata4.00: configured for UDMA/133
PM: Adding info for No Bus:target2:0:0
scsi 2:0:0:0: Direct-Access ATA ST3320620AS 3.AA PQ: 0 ANSI: 5
PM: Adding info for scsi:2:0:0:0
SCSI device sdb: 625142448 512-byte hdwr sectors (320073 MB)
sdb: Write Protect is off
sdb: Mode Sense: 00 3a 00 00
SCSI device sdb: drive cache: write back
SCSI device sdb: 625142448 512-byte hdwr sectors (320073 MB)
sdb: Write Protect is off
sdb: Mode Sense: 00 3a 00 00
SCSI device sdb: drive cache: write back
sdb: sdb1
sd 2:0:0:0: Attached scsi disk sdb
PM: Adding info for No Bus:target3:0:0
scsi 3:0:0:0: Direct-Access ATA ST3160827AS 3.42 PQ: 0 ANSI: 5
PM: Adding info for scsi:3:0:0:0
SCSI device sdc: 312581808 512-byte hdwr sectors (160042 MB)
sdc: Write Protect is off
sdc: Mode Sense: 00 3a 00 00
SCSI device sdc: drive cache: write back
SCSI device sdc: 312581808 512-byte hdwr sectors (160042 MB)
sdc: Write Protect is off
sdc: Mode Sense: 00 3a 00 00
SCSI device sdc: drive cache: write back
sdc: sdc1 sdc2 sdc3 sdc4 < sdc5 >
sd 3:0:0:0: Attached scsi disk sdc
Finally some Bonnie++ results from before, without NCQ:
Version 1.03 ------Sequential Output------ --Sequential Input-
--Random-
-Per Chr- --Block-- -Rewrite- -Per Chr- --Block--
--Seeks--
Machine Size K/sec %CP K/sec %CP K/sec %CP K/sec %CP K/sec %CP
/sec %CP
newcastle.ss.sha 4G 33185 79 36840 33 16170 9 33461 73 39788 7
179.4 1
------Sequential Create------ --------Random
Create--------
-Create-- --Read--- -Delete-- -Create-- --Read---
-Delete--
files /sec %CP /sec %CP /sec %CP /sec %CP /sec %CP
/sec %CP
16 6920 81 +++++ +++ 15820 100 8286 99 +++++ +++
15874 98
and with NCQ:
Version 1.03 ------Sequential Output------ --Sequential Input-
--Random-
-Per Chr- --Block-- -Rewrite- -Per Chr- --Block--
--Seeks--
Machine Size K/sec %CP K/sec %CP K/sec %CP K/sec %CP K/sec %CP
/sec %CP
newcastle.ss.sha 4G 27068 66 31907 26 17020 8 29163 77 40032 8
205.0 0
------Sequential Create------ --------Random
Create--------
-Create-- --Read--- -Delete-- -Create-- --Read---
-Delete--
files /sec %CP /sec %CP /sec %CP /sec %CP /sec %CP
/sec %CP
16 9566 91 +++++ +++ 16387 99 10606 99 +++++ +++
23773 99
It looks like there was a bit of a drop in sequential output and
sequential character input, but everything else went up..
In case anyone wants to actually apply this to something:
Signed-off-by: Robert Hancock <hancockr@shaw.ca>
--
Robert Hancock Saskatoon, SK, Canada
To email, remove "nospam" from hancockr@nospamshaw.ca
Home Page: http://www.roberthancock.com/
[-- Attachment #2: sata_nv-adma-ncq.patch --]
[-- Type: text/plain, Size: 30759 bytes --]
diff -rup --exclude='*.o' --exclude='*mod.c' --exclude='*.cmd' --exclude='*.ko' linux-2.6.18-mm2/drivers/ata/libata-sff.c linux-2.6.18-mm2-adma/drivers/ata/libata-sff.c
--- linux-2.6.18-mm2/drivers/ata/libata-sff.c 2006-10-01 15:42:37.000000000 -0600
+++ linux-2.6.18-mm2-adma/drivers/ata/libata-sff.c 2006-10-02 22:20:53.000000000 -0600
@@ -732,7 +732,7 @@ void ata_bmdma_drive_eh(struct ata_port
qc->tf.protocol == ATA_PROT_ATAPI_DMA)) {
u8 host_stat;
- host_stat = ata_bmdma_status(ap);
+ host_stat = ap->ops->bmdma_status(ap);
ata_ehi_push_desc(&ehc->i, "BMDMA stat 0x%x", host_stat);
diff -rup --exclude='*.o' --exclude='*mod.c' --exclude='*.cmd' --exclude='*.ko' linux-2.6.18-mm2/drivers/ata/sata_nv.c linux-2.6.18-mm2-adma/drivers/ata/sata_nv.c
--- linux-2.6.18-mm2/drivers/ata/sata_nv.c 2006-10-01 15:42:37.000000000 -0600
+++ linux-2.6.18-mm2-adma/drivers/ata/sata_nv.c 2006-10-07 00:35:25.000000000 -0600
@@ -29,6 +29,11 @@
* NV-specific details such as register offsets, SATA phy location,
* hotplug info, etc.
*
+ * CK804 (and later?) controllers support an alternate programming interface
+ * similar to the ADMA specification (with some modifications).
+ * This allows the use of NCQ. Currently non-DMA-mapped ATA commands are still
+ * sent through the legacy interface.
+ *
*/
#include <linux/kernel.h>
@@ -43,7 +48,7 @@
#include <linux/libata.h>
#define DRV_NAME "sata_nv"
-#define DRV_VERSION "2.0"
+#define DRV_VERSION "2.1"
enum {
NV_PORTS = 2,
@@ -78,8 +83,135 @@ enum {
// For PCI config register 20
NV_MCP_SATA_CFG_20 = 0x50,
NV_MCP_SATA_CFG_20_SATA_SPACE_EN = 0x04,
+ NV_MCP_SATA_CFG_20_PORT0_EN = (1 << 17),
+ NV_MCP_SATA_CFG_20_PORT1_EN = (1 << 16),
+ NV_MCP_SATA_CFG_20_PORT0_PWB_EN = (1 << 14),
+ NV_MCP_SATA_CFG_20_PORT1_PWB_EN = (1 << 12),
+
+ NV_ADMA_MAX_CPBS = 32,
+ NV_ADMA_CPB_SZ = 128,
+ NV_ADMA_APRD_SZ = 16,
+ NV_ADMA_SGTBL_LEN = (1024 - NV_ADMA_CPB_SZ) / NV_ADMA_APRD_SZ,
+ NV_ADMA_SGTBL_SZ = NV_ADMA_SGTBL_LEN * NV_ADMA_APRD_SZ,
+ NV_ADMA_PORT_PRIV_DMA_SZ = NV_ADMA_MAX_CPBS * (NV_ADMA_CPB_SZ + NV_ADMA_SGTBL_SZ),
+
+ // BAR5 offset to ADMA general registers
+ NV_ADMA_GEN = 0x400,
+ NV_ADMA_GEN_CTL = 0x00,
+ NV_ADMA_NOTIFIER_CLEAR = 0x30,
+
+ // BAR5 offset to ADMA ports
+ NV_ADMA_PORT = 0x480,
+
+ // size of ADMA port register space
+ NV_ADMA_PORT_SIZE = 0x100,
+
+ // ADMA port registers
+ NV_ADMA_CTL = 0x40,
+ NV_ADMA_CPB_COUNT = 0x42,
+ NV_ADMA_NEXT_CPB_IDX = 0x43,
+ NV_ADMA_STAT = 0x44,
+ NV_ADMA_CPB_BASE_LOW = 0x48,
+ NV_ADMA_CPB_BASE_HIGH = 0x4C,
+ NV_ADMA_APPEND = 0x50,
+ NV_ADMA_NOTIFIER = 0x68,
+ NV_ADMA_NOTIFIER_ERROR = 0x6C,
+
+ // NV_ADMA_CTL register bits
+ NV_ADMA_CTL_HOTPLUG_IEN = (1 << 0),
+ NV_ADMA_CTL_CHANNEL_RESET = (1 << 5),
+ NV_ADMA_CTL_GO = (1 << 7),
+ NV_ADMA_CTL_AIEN = (1 << 8),
+ NV_ADMA_CTL_READ_NON_COHERENT = (1 << 11),
+ NV_ADMA_CTL_WRITE_NON_COHERENT = (1 << 12),
+
+ // CPB response flag bits
+ NV_CPB_RESP_DONE = (1 << 0),
+ NV_CPB_RESP_ATA_ERR = (1 << 3),
+ NV_CPB_RESP_CMD_ERR = (1 << 4),
+ NV_CPB_RESP_CPB_ERR = (1 << 7),
+
+ // CPB control flag bits
+ NV_CPB_CTL_CPB_VALID = (1 << 0),
+ NV_CPB_CTL_QUEUE = (1 << 1),
+ NV_CPB_CTL_APRD_VALID = (1 << 2),
+ NV_CPB_CTL_IEN = (1 << 3),
+ NV_CPB_CTL_FPDMA = (1 << 4),
+
+ // APRD flags
+ NV_APRD_WRITE = (1 << 1),
+ NV_APRD_END = (1 << 2),
+ NV_APRD_CONT = (1 << 3),
+
+ // NV_ADMA_STAT flags
+ NV_ADMA_STAT_TIMEOUT = (1 << 0),
+ NV_ADMA_STAT_HOTUNPLUG = (1 << 1),
+ NV_ADMA_STAT_HOTPLUG = (1 << 2),
+ NV_ADMA_STAT_CPBERR = (1 << 4),
+ NV_ADMA_STAT_SERROR = (1 << 5),
+ NV_ADMA_STAT_CMD_COMPLETE = (1 << 6),
+ NV_ADMA_STAT_IDLE = (1 << 8),
+ NV_ADMA_STAT_LEGACY = (1 << 9),
+ NV_ADMA_STAT_STOPPED = (1 << 10),
+ NV_ADMA_STAT_DONE = (1 << 12),
+ NV_ADMA_STAT_ERR = (NV_ADMA_STAT_CPBERR | NV_ADMA_STAT_TIMEOUT),
+
+ // port flags
+ NV_ADMA_PORT_REGISTER_MODE = (1 << 0),
+
+};
+
+/* ADMA Physical Region Descriptor - one SG segment */
+struct nv_adma_prd {
+ __le64 addr;
+ __le32 len;
+ u8 flags;
+ u8 packet_len;
+ __le16 reserved;
+};
+
+enum nv_adma_regbits {
+ CMDEND = (1 << 15), /* end of command list */
+ WNB = (1 << 14), /* wait-not-BSY */
+ IGN = (1 << 13), /* ignore this entry */
+ CS1n = (1 << (4 + 8)), /* std. PATA signals follow... */
+ DA2 = (1 << (2 + 8)),
+ DA1 = (1 << (1 + 8)),
+ DA0 = (1 << (0 + 8)),
+};
+
+/* ADMA Command Parameter Block
+ The first 5 SG segments are stored inside the Command Parameter Block itself.
+ If there are more than 5 segments the remainder are stored in a separate
+ memory area indicated by next_aprd. */
+struct nv_adma_cpb {
+ u8 resp_flags; //0
+ u8 reserved1; //1
+ u8 ctl_flags; //2
+ // len is length of taskfile in 64 bit words
+ u8 len; //3
+ u8 tag; //4
+ u8 next_cpb_idx; //5
+ __le16 reserved2; //6-7
+ __le16 tf[12]; //8-31
+ struct nv_adma_prd aprd[5]; //32-111
+ __le64 next_aprd; //112-119
+ __le64 reserved3; //120-127
+};
+
+
+struct nv_adma_port_priv {
+ struct nv_adma_cpb *cpb;
+ dma_addr_t cpb_dma;
+ struct nv_adma_prd *aprd;
+ dma_addr_t aprd_dma;
+ u8 flags;
+ u32 notifier;
+ u32 notifier_error;
};
+#define NV_ADMA_CHECK_INTR(GCTL, PORT) ((GCTL) & ( 1 << (19 + (12 * (PORT)))))
+
static int nv_init_one (struct pci_dev *pdev, const struct pci_device_id *ent);
static void nv_ck804_host_stop(struct ata_host *host);
static irqreturn_t nv_generic_interrupt(int irq, void *dev_instance,
@@ -96,13 +228,27 @@ static void nv_nf2_thaw(struct ata_port
static void nv_ck804_freeze(struct ata_port *ap);
static void nv_ck804_thaw(struct ata_port *ap);
static void nv_error_handler(struct ata_port *ap);
+static void nv_adma_qc_prep(struct ata_queued_cmd *qc);
+static unsigned int nv_adma_qc_issue(struct ata_queued_cmd *qc);
+static irqreturn_t nv_adma_interrupt(int irq, void *dev_instance,
+ struct pt_regs *regs);
+static void nv_adma_irq_clear(struct ata_port *ap);
+static int nv_adma_port_start(struct ata_port *ap);
+static void nv_adma_port_stop(struct ata_port *ap);
+static void nv_adma_error_handler(struct ata_port *ap);
+static void nv_adma_host_stop(struct ata_host *host);
+static void nv_adma_bmdma_setup(struct ata_queued_cmd *qc);
+static void nv_adma_bmdma_start(struct ata_queued_cmd *qc);
+static void nv_adma_bmdma_stop(struct ata_queued_cmd *qc);
+static u8 nv_adma_bmdma_status(struct ata_port *ap);
enum nv_host_type
{
GENERIC,
NFORCE2,
NFORCE3 = NFORCE2, /* NF2 == NF3 as far as sata_nv is concerned */
- CK804
+ CK804,
+ ADMA
};
static const struct pci_device_id nv_pci_tbl[] = {
@@ -158,6 +304,25 @@ static struct scsi_host_template nv_sht
.bios_param = ata_std_bios_param,
};
+static struct scsi_host_template nv_adma_sht = {
+ .module = THIS_MODULE,
+ .name = DRV_NAME,
+ .ioctl = ata_scsi_ioctl,
+ .queuecommand = ata_scsi_queuecmd,
+ .can_queue = NV_ADMA_MAX_CPBS,
+ .this_id = ATA_SHT_THIS_ID,
+ .sg_tablesize = NV_ADMA_SGTBL_LEN + 5,
+ .max_sectors = ATA_MAX_SECTORS,
+ .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 const struct ata_port_operations nv_generic_ops = {
.port_disable = ata_port_disable,
.tf_load = ata_tf_load,
@@ -239,6 +404,33 @@ static const struct ata_port_operations
.host_stop = nv_ck804_host_stop,
};
+static const struct ata_port_operations nv_adma_ops = {
+ .port_disable = ata_port_disable,
+ .tf_load = ata_tf_load,
+ .tf_read = ata_tf_read,
+ .exec_command = ata_exec_command,
+ .check_status = ata_check_status,
+ .dev_select = ata_std_dev_select,
+ .bmdma_setup = nv_adma_bmdma_setup,
+ .bmdma_start = nv_adma_bmdma_start,
+ .bmdma_stop = nv_adma_bmdma_stop,
+ .bmdma_status = nv_adma_bmdma_status,
+ .qc_prep = nv_adma_qc_prep,
+ .qc_issue = nv_adma_qc_issue,
+ .freeze = nv_ck804_freeze,
+ .thaw = nv_ck804_thaw,
+ .error_handler = nv_adma_error_handler,
+ .post_internal_cmd = nv_adma_bmdma_stop,
+ .data_xfer = ata_mmio_data_xfer,
+ .irq_handler = nv_adma_interrupt,
+ .irq_clear = nv_adma_irq_clear,
+ .scr_read = nv_scr_read,
+ .scr_write = nv_scr_write,
+ .port_start = nv_adma_port_start,
+ .port_stop = nv_adma_port_stop,
+ .host_stop = nv_adma_host_stop,
+};
+
static struct ata_port_info nv_port_info[] = {
/* generic */
{
@@ -267,6 +459,16 @@ static struct ata_port_info nv_port_info
.udma_mask = NV_UDMA_MASK,
.port_ops = &nv_ck804_ops,
},
+ /* ADMA */
+ {
+ .sht = &nv_adma_sht,
+ .flags = ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY |
+ ATA_FLAG_MMIO | ATA_FLAG_NCQ,
+ .pio_mask = NV_PIO_MASK,
+ .mwdma_mask = NV_MWDMA_MASK,
+ .udma_mask = NV_UDMA_MASK,
+ .port_ops = &nv_adma_ops,
+ },
};
MODULE_AUTHOR("NVIDIA");
@@ -275,6 +477,605 @@ MODULE_LICENSE("GPL");
MODULE_DEVICE_TABLE(pci, nv_pci_tbl);
MODULE_VERSION(DRV_VERSION);
+static int adma_enabled = 0;
+
+static unsigned int nv_adma_tf_to_cpb(struct ata_taskfile *tf, u16 *cpb)
+{
+ unsigned int idx = 0;
+
+ cpb[idx++] = cpu_to_le16((ATA_REG_DEVICE << 8) | tf->device | WNB);
+
+ if ((tf->flags & ATA_TFLAG_LBA48) == 0) {
+ cpb[idx++] = cpu_to_le16(IGN);
+ cpb[idx++] = cpu_to_le16(IGN);
+ cpb[idx++] = cpu_to_le16(IGN);
+ cpb[idx++] = cpu_to_le16(IGN);
+ cpb[idx++] = cpu_to_le16(IGN);
+ }
+ else {
+ cpb[idx++] = cpu_to_le16((ATA_REG_ERR << 8) | tf->hob_feature);
+ cpb[idx++] = cpu_to_le16((ATA_REG_NSECT << 8) | tf->hob_nsect);
+ cpb[idx++] = cpu_to_le16((ATA_REG_LBAL << 8) | tf->hob_lbal);
+ cpb[idx++] = cpu_to_le16((ATA_REG_LBAM << 8) | tf->hob_lbam);
+ cpb[idx++] = cpu_to_le16((ATA_REG_LBAH << 8) | tf->hob_lbah);
+ }
+ cpb[idx++] = cpu_to_le16((ATA_REG_ERR << 8) | tf->feature);
+ cpb[idx++] = cpu_to_le16((ATA_REG_NSECT << 8) | tf->nsect);
+ cpb[idx++] = cpu_to_le16((ATA_REG_LBAL << 8) | tf->lbal);
+ cpb[idx++] = cpu_to_le16((ATA_REG_LBAM << 8) | tf->lbam);
+ cpb[idx++] = cpu_to_le16((ATA_REG_LBAH << 8) | tf->lbah);
+
+ cpb[idx++] = cpu_to_le16((ATA_REG_CMD << 8) | tf->command | CMDEND);
+
+ return idx;
+}
+
+static inline void __iomem *__nv_adma_ctl_block(void __iomem *mmio,
+ unsigned int port_no)
+{
+ mmio += NV_ADMA_PORT + port_no * NV_ADMA_PORT_SIZE;
+ return mmio;
+}
+
+static inline void __iomem *nv_adma_ctl_block(struct ata_port *ap)
+{
+ return __nv_adma_ctl_block(ap->host->mmio_base, ap->port_no);
+}
+
+static inline void __iomem *nv_adma_gen_block(struct ata_port *ap)
+{
+ return (ap->host->mmio_base + NV_ADMA_GEN);
+}
+
+static inline void __iomem *nv_adma_notifier_clear_block(struct ata_port *ap)
+{
+ return (nv_adma_gen_block(ap) + NV_ADMA_NOTIFIER_CLEAR + (4 * ap->port_no));
+}
+
+static void nv_adma_reset_channel(struct ata_port *ap)
+{
+ void __iomem *mmio = nv_adma_ctl_block(ap);
+ struct nv_adma_port_priv *pp = ap->private_data;
+ u16 tmp;
+ int i;
+
+ dev_printk(KERN_NOTICE, ap->dev, "Resetting port %u\n", ap->port_no);
+
+ /* clear CPB fetch count */
+ writew(0, mmio + NV_ADMA_CPB_COUNT);
+
+ /* clear GO */
+ tmp = readw(mmio + NV_ADMA_CTL);
+ writew(tmp & ~NV_ADMA_CTL_GO, mmio + NV_ADMA_CTL);
+
+ tmp = readw(mmio + NV_ADMA_CTL);
+ writew(tmp | NV_ADMA_CTL_CHANNEL_RESET, mmio + NV_ADMA_CTL);
+ readl( mmio + NV_ADMA_CTL ); /* flush posted write */
+ udelay(1);
+ writew(tmp & ~NV_ADMA_CTL_CHANNEL_RESET, mmio + NV_ADMA_CTL);
+ readl( mmio + NV_ADMA_CTL ); /* flush posted write */
+
+ /* Mark all of the CPBs as invalid to prevent them from being executed */
+ for( i=0;i<NV_ADMA_MAX_CPBS;i++)
+ pp->cpb[i].ctl_flags &= ~NV_CPB_CTL_CPB_VALID;
+
+ if(!(pp->flags & NV_ADMA_PORT_REGISTER_MODE)) {
+ tmp = readw(mmio + NV_ADMA_CTL);
+ writew(tmp | NV_ADMA_CTL_GO, mmio + NV_ADMA_CTL);
+ }
+}
+
+static int nv_adma_host_intr(struct ata_port *ap)
+{
+ void __iomem *mmio = nv_adma_ctl_block(ap);
+ struct nv_adma_port_priv *pp = ap->private_data;
+ u16 status;
+ u32 gen_ctl;
+ u16 flags;
+ int have_global_err = 0;
+
+ /* if in ATA register mode, use standard ata interrupt handler */
+ if (pp->flags & NV_ADMA_PORT_REGISTER_MODE) {
+ struct ata_queued_cmd *qc;
+ VPRINTK("in ATA register mode\n");
+ qc = ata_qc_from_tag(ap, ap->active_tag);
+ if (qc && (!(qc->tf.flags & ATA_TFLAG_POLLING)))
+ return ata_host_intr(ap, qc);
+ else {
+ // No request pending? Clear interrupt status
+ // anyway, in case there's one pending.
+ ap->ops->check_status(ap);
+ return 1;
+ }
+ }
+
+ gen_ctl = readl(nv_adma_gen_block(ap) + NV_ADMA_GEN_CTL);
+ if (!NV_ADMA_CHECK_INTR(gen_ctl, ap->port_no))
+ return 0;
+
+ status = readw(mmio + NV_ADMA_STAT);
+
+ if( !status && !pp->notifier && !pp->notifier_error)
+ /* Nothing to do */
+ return 0;
+
+ /* Clear status so that any CPB completions after this point in the handler will
+ raise another interrupt */
+ writew(status, mmio + NV_ADMA_STAT);
+ readl(mmio + NV_ADMA_STAT); /* flush posted write */
+ mb();
+
+ /* freeze if hotplugged */
+ if (unlikely(status & (NV_ADMA_STAT_HOTPLUG | NV_ADMA_STAT_HOTUNPLUG))) {
+ ata_port_freeze(ap);
+ return 1;
+ }
+
+ if (!pp->notifier && !pp->notifier_error) {
+ if (status) {
+ VPRINTK("XXX no notifier, but status 0x%x\n", status);
+ } else
+ return 0;
+ }
+ if (pp->notifier_error) {
+ dev_printk(KERN_ERR, ap->dev, "port %u notifier error, stat=0x%x, "
+ "notifier=0x%X, notifier_error=0x%X", ap->port_no, status, pp->notifier,
+ pp->notifier_error );
+ have_global_err = 1;
+ }
+
+ if (status & NV_ADMA_STAT_TIMEOUT) {
+ dev_printk(KERN_ERR, ap->dev, "port %u timeout, stat=0x%x\n", ap->port_no, status);
+ have_global_err = 1;
+ }
+ if (status & NV_ADMA_STAT_CPBERR) {
+ dev_printk(KERN_ERR, ap->dev, "port %u CPB error, stat=0x%x\n", ap->port_no, status);
+ have_global_err = 1;
+ }
+ if (status & NV_ADMA_STAT_STOPPED) {
+ VPRINTK("ADMA stopped, stat = 0x%x\n", status);
+ }
+ if (status & NV_ADMA_STAT_CMD_COMPLETE) {
+ VPRINTK("ADMA command complete, stat = 0x%x\n", status);
+ }
+ if ((status & NV_ADMA_STAT_DONE) || have_global_err) {
+ /** Check CPBs for completed commands */
+ int i;
+ for( i=0;i<NV_ADMA_MAX_CPBS;i++)
+ if( ap->sactive & (1 << i ) ||
+ ( ata_tag_valid(ap->active_tag) && ap->active_tag == i ) ) {
+ struct nv_adma_cpb *cpb = &pp->cpb[i];
+ int complete = 0, have_err = 0;
+ flags = cpb->resp_flags;
+ VPRINTK("CPB %d, stat=0x%x, flags=0x%x\n", i, status, flags);
+
+ if (flags & NV_CPB_RESP_DONE) {
+ VPRINTK("CPB flags done, flags=0x%x\n", flags);
+ complete = 1;
+ }
+ if (flags & NV_CPB_RESP_ATA_ERR) {
+ dev_printk(KERN_ERR, ap->dev, "port %u CPB flags ATA err, flags=0x%x\n", ap->port_no, flags);
+ have_err = 1;
+ complete = 1;
+ }
+ if (flags & NV_CPB_RESP_CMD_ERR) {
+ dev_printk(KERN_ERR, ap->dev, "port %u CPB flags CMD err, flags=0x%x\n", ap->port_no, flags);
+ have_err = 1;
+ complete = 1;
+ }
+ if (flags & NV_CPB_RESP_CPB_ERR) {
+ dev_printk(KERN_ERR, ap->dev, "port %u CPB flags CPB err, flags=0x%x\n", ap->port_no, flags);
+ have_err = 1;
+ complete = 1;
+ }
+ if(complete || have_global_err)
+ {
+ struct ata_queued_cmd *qc = ata_qc_from_tag(ap, i);
+ if(likely(qc)) {
+ u8 ata_status = 0;
+ /* Only use the ATA port status for non-NCQ commands.
+ For NCQ commands the current status may have nothing to do with
+ the command just completed. */
+ if(qc->tf.protocol != ATA_PROT_NCQ)
+ ata_status = readb(mmio + (ATA_REG_STATUS * 4));
+
+ if(have_err || have_global_err)
+ ata_status |= ATA_ERR;
+
+ qc->err_mask |= ac_err_mask(ata_status);
+ DPRINTK("Completing qc from tag %d with err_mask %u\n",i,qc->err_mask);
+ ata_qc_complete(qc);
+ }
+ }
+ }
+ }
+
+ return 1; /* irq handled if we got here */
+}
+
+static irqreturn_t nv_adma_interrupt(int irq, void *dev_instance,
+ struct pt_regs *regs)
+{
+ struct ata_host *host = dev_instance;
+ int i, handled = 0;
+
+ spin_lock(&host->lock);
+
+ for (i = 0; i < host->n_ports; i++) {
+ struct ata_port *ap = host->ports[i];
+
+ if (ap && !(ap->flags & ATA_FLAG_DISABLED)) {
+ struct nv_adma_port_priv *pp = ap->private_data;
+ void __iomem *mmio = nv_adma_ctl_block(ap);
+ // read notifiers
+ pp->notifier = readl(mmio + NV_ADMA_NOTIFIER);
+ pp->notifier_error = readl(mmio + NV_ADMA_NOTIFIER_ERROR);
+ handled += nv_adma_host_intr(ap);
+ }
+ }
+
+ /* Clear notifiers. Yes, this does appear to have to be done at the very end,
+ otherwise things grind to a halt quite quickly. */
+ if (handled)
+ for (i = 0; i < host->n_ports; i++) {
+ struct ata_port *ap = host->ports[i];
+ struct nv_adma_port_priv *pp = ap->private_data;
+ writel(pp->notifier | pp->notifier_error,
+ nv_adma_notifier_clear_block(ap));
+ }
+
+ spin_unlock(&host->lock);
+
+ return IRQ_RETVAL(handled);
+}
+
+static void nv_adma_irq_clear(struct ata_port *ap)
+{
+ void __iomem *mmio = nv_adma_ctl_block(ap);
+ u16 status = readw(mmio + NV_ADMA_STAT);
+ u32 notifier = readl(mmio + NV_ADMA_NOTIFIER);
+ u32 notifier_error = readl(mmio + NV_ADMA_NOTIFIER_ERROR);
+
+ /* clear ADMA status */
+ writew(status, mmio + NV_ADMA_STAT);
+ writel(notifier | notifier_error,
+ nv_adma_notifier_clear_block(ap));
+
+ /** clear legacy status */
+ ap->flags &= ~ATA_FLAG_MMIO;
+ ata_bmdma_irq_clear(ap);
+ ap->flags |= ATA_FLAG_MMIO;
+}
+
+static void nv_adma_bmdma_setup(struct ata_queued_cmd *qc)
+{
+ struct nv_adma_port_priv *pp = qc->ap->private_data;
+
+ if (!(pp->flags & NV_ADMA_PORT_REGISTER_MODE))
+ return;
+
+ qc->ap->flags &= ~ATA_FLAG_MMIO;
+ ata_bmdma_setup(qc);
+ qc->ap->flags |= ATA_FLAG_MMIO;
+}
+
+static void nv_adma_bmdma_start(struct ata_queued_cmd *qc)
+{
+ struct nv_adma_port_priv *pp = qc->ap->private_data;
+
+ if (!(pp->flags & NV_ADMA_PORT_REGISTER_MODE))
+ return;
+
+ qc->ap->flags &= ~ATA_FLAG_MMIO;
+ ata_bmdma_start(qc);
+ qc->ap->flags |= ATA_FLAG_MMIO;
+}
+
+static void nv_adma_bmdma_stop(struct ata_queued_cmd *qc)
+{
+ struct nv_adma_port_priv *pp = qc->ap->private_data;
+
+ if (!(pp->flags & NV_ADMA_PORT_REGISTER_MODE))
+ return;
+
+ qc->ap->flags &= ~ATA_FLAG_MMIO;
+ ata_bmdma_stop(qc);
+ qc->ap->flags |= ATA_FLAG_MMIO;
+}
+
+static u8 nv_adma_bmdma_status(struct ata_port *ap)
+{
+ u8 status;
+ ap->flags &= ~ATA_FLAG_MMIO;
+ status = ata_bmdma_status(ap);
+ ap->flags |= ATA_FLAG_MMIO;
+ return status;
+}
+
+static void nv_adma_register_mode(struct ata_port *ap)
+{
+ void __iomem *mmio = nv_adma_ctl_block(ap);
+ struct nv_adma_port_priv *pp = ap->private_data;
+ u16 tmp;
+
+ tmp = readw(mmio + NV_ADMA_CTL);
+ writew(tmp & ~NV_ADMA_CTL_GO, mmio + NV_ADMA_CTL);
+
+ pp->flags |= NV_ADMA_PORT_REGISTER_MODE;
+}
+
+static void nv_adma_mode(struct ata_port *ap)
+{
+ void __iomem *mmio = nv_adma_ctl_block(ap);
+ struct nv_adma_port_priv *pp = ap->private_data;
+ u16 tmp;
+
+ if (!(pp->flags & NV_ADMA_PORT_REGISTER_MODE))
+ return;
+
+ tmp = readw(mmio + NV_ADMA_CTL);
+ writew(tmp | NV_ADMA_CTL_GO, mmio + NV_ADMA_CTL);
+
+ pp->flags &= ~NV_ADMA_PORT_REGISTER_MODE;
+}
+
+static int nv_adma_port_start(struct ata_port *ap)
+{
+ struct device *dev = ap->host->dev;
+ struct nv_adma_port_priv *pp;
+ int rc;
+ void *mem;
+ dma_addr_t mem_dma;
+ void __iomem *mmio = nv_adma_ctl_block(ap);
+
+ VPRINTK("ENTER\n");
+
+ rc = ata_port_start(ap);
+ if (rc)
+ return rc;
+
+ pp = kzalloc(sizeof(*pp), GFP_KERNEL);
+ if (!pp) {
+ rc = -ENOMEM;
+ goto err_out;
+ }
+
+ mem = dma_alloc_coherent(dev, NV_ADMA_PORT_PRIV_DMA_SZ,
+ &mem_dma, GFP_KERNEL);
+
+ VPRINTK("dma memory: vaddr = 0x%p, paddr = 0x%p\n", mem, (void*)mem_dma);
+
+ if (!mem) {
+ rc = -ENOMEM;
+ goto err_out_kfree;
+ }
+ memset(mem, 0, NV_ADMA_PORT_PRIV_DMA_SZ);
+
+ /*
+ * First item in chunk of DMA memory:
+ * 128-byte command parameter block (CPB)
+ * one for each command tag
+ */
+ pp->cpb = mem;
+ pp->cpb_dma = mem_dma;
+
+ VPRINTK("cpb = 0x%p, cpb_dma = 0x%p\n", pp->cpb, (void*)pp->cpb_dma);
+
+ writel(mem_dma & 0xFFFFFFFF, mmio + NV_ADMA_CPB_BASE_LOW);
+ writel((mem_dma >> 16 ) >> 16, mmio + NV_ADMA_CPB_BASE_HIGH);
+
+ mem += NV_ADMA_MAX_CPBS * NV_ADMA_CPB_SZ;
+ mem_dma += NV_ADMA_MAX_CPBS * NV_ADMA_CPB_SZ;
+
+ /*
+ * Second item: block of ADMA_SGTBL_LEN s/g entries
+ */
+ pp->aprd = mem;
+ pp->aprd_dma = mem_dma;
+
+ VPRINTK("aprd = 0x%p, aprd_dma = 0x%p\n", pp->aprd, (void*)pp->aprd_dma);
+
+ ap->private_data = pp;
+
+ // clear any outstanding interrupt conditions
+ writew(0xffff, mmio + NV_ADMA_STAT);
+
+ // initialize port variables
+ pp->flags = NV_ADMA_PORT_REGISTER_MODE;
+
+ nv_adma_reset_channel(ap);
+
+ // make sure controller is in ATA register mode
+ nv_adma_register_mode(ap);
+
+ return 0;
+
+err_out_kfree:
+ kfree(pp);
+err_out:
+ ata_port_stop(ap);
+ return rc;
+}
+
+static void nv_adma_port_stop(struct ata_port *ap)
+{
+ struct device *dev = ap->host->dev;
+ struct nv_adma_port_priv *pp = ap->private_data;
+ void __iomem *mmio = nv_adma_ctl_block(ap);
+
+ VPRINTK("ENTER\n");
+
+ writew(0, mmio + NV_ADMA_CTL);
+
+ ap->private_data = NULL;
+ dma_free_coherent(dev, NV_ADMA_PORT_PRIV_DMA_SZ, pp->cpb, pp->cpb_dma);
+ kfree(pp);
+ ata_port_stop(ap);
+}
+
+
+static void nv_adma_setup_port(struct ata_probe_ent *probe_ent, unsigned int port)
+{
+ void __iomem *mmio = probe_ent->mmio_base;
+ struct ata_ioports *ioport = &probe_ent->port[port];
+
+ VPRINTK("ENTER\n");
+
+ mmio += NV_ADMA_PORT + port * NV_ADMA_PORT_SIZE;
+
+ ioport->cmd_addr = (unsigned long) mmio;
+ ioport->data_addr = (unsigned long) mmio + (ATA_REG_DATA * 4);
+ ioport->error_addr =
+ ioport->feature_addr = (unsigned long) mmio + (ATA_REG_ERR * 4);
+ ioport->nsect_addr = (unsigned long) mmio + (ATA_REG_NSECT * 4);
+ ioport->lbal_addr = (unsigned long) mmio + (ATA_REG_LBAL * 4);
+ ioport->lbam_addr = (unsigned long) mmio + (ATA_REG_LBAM * 4);
+ ioport->lbah_addr = (unsigned long) mmio + (ATA_REG_LBAH * 4);
+ ioport->device_addr = (unsigned long) mmio + (ATA_REG_DEVICE * 4);
+ ioport->status_addr =
+ ioport->command_addr = (unsigned long) mmio + (ATA_REG_STATUS * 4);
+ ioport->altstatus_addr =
+ ioport->ctl_addr = (unsigned long) mmio + 0x20;
+}
+
+static int nv_adma_host_init(struct ata_probe_ent *probe_ent)
+{
+ struct pci_dev *pdev = to_pci_dev(probe_ent->dev);
+ unsigned int i;
+ u32 tmp32;
+
+ VPRINTK("ENTER\n");
+
+ // enable ADMA on the ports
+ pci_read_config_dword(pdev, NV_MCP_SATA_CFG_20, &tmp32);
+ tmp32 |= NV_MCP_SATA_CFG_20_PORT0_EN |
+ NV_MCP_SATA_CFG_20_PORT0_PWB_EN |
+ NV_MCP_SATA_CFG_20_PORT1_EN |
+ NV_MCP_SATA_CFG_20_PORT1_PWB_EN;
+
+ pci_write_config_dword(pdev, NV_MCP_SATA_CFG_20, tmp32);
+
+ for (i = 0; i < probe_ent->n_ports; i++)
+ nv_adma_setup_port(probe_ent, i);
+
+ for (i = 0; i < probe_ent->n_ports; i++) {
+ void __iomem *mmio = __nv_adma_ctl_block(probe_ent->mmio_base, i);
+ u16 tmp;
+
+ /* enable interrupt, clear reset if not already clear */
+ tmp = readw(mmio + NV_ADMA_CTL);
+ writew(tmp | NV_ADMA_CTL_AIEN, mmio + NV_ADMA_CTL);
+ }
+
+ return 0;
+}
+
+
+static void nv_adma_fill_aprd(struct ata_queued_cmd *qc,
+ struct scatterlist *sg,
+ int idx,
+ struct nv_adma_prd *aprd)
+{
+ u32 flags;
+
+ memset(aprd, 0, sizeof(struct nv_adma_prd));
+
+ flags = 0;
+ if (qc->tf.flags & ATA_TFLAG_WRITE)
+ flags |= NV_APRD_WRITE;
+ if (idx == qc->n_elem - 1)
+ flags |= NV_APRD_END;
+ else if (idx != 4)
+ flags |= NV_APRD_CONT;
+
+ aprd->addr = cpu_to_le64(((u64)sg_dma_address(sg)));
+ aprd->len = cpu_to_le32(((u32)sg_dma_len(sg))); /* len in bytes */
+ aprd->flags = cpu_to_le32(flags);
+}
+
+static void nv_adma_fill_sg(struct ata_queued_cmd *qc, struct nv_adma_cpb *cpb)
+{
+ struct nv_adma_port_priv *pp = qc->ap->private_data;
+ unsigned int idx;
+ struct nv_adma_prd *aprd;
+ struct scatterlist *sg;
+
+ VPRINTK("ENTER\n");
+
+ idx = 0;
+
+ ata_for_each_sg(sg, qc) {
+ aprd = (idx < 5) ? &cpb->aprd[idx] : &pp->aprd[NV_ADMA_SGTBL_LEN * qc->tag + (idx-5)];
+ nv_adma_fill_aprd(qc, sg, idx, aprd);
+ idx++;
+ }
+ if (idx > 5)
+ cpb->next_aprd = cpu_to_le64(((u64)(pp->aprd_dma + NV_ADMA_SGTBL_SZ * qc->tag)));
+}
+
+static void nv_adma_qc_prep(struct ata_queued_cmd *qc)
+{
+ struct nv_adma_port_priv *pp = qc->ap->private_data;
+ struct nv_adma_cpb *cpb = &pp->cpb[qc->tag];
+ u8 ctl_flags = NV_CPB_CTL_CPB_VALID |
+ NV_CPB_CTL_APRD_VALID |
+ NV_CPB_CTL_IEN;
+
+ VPRINTK("ENTER\n");
+
+ VPRINTK("qc->flags = 0x%x\n", (u32)qc->flags);
+
+ if (!(qc->flags & ATA_QCFLAG_DMAMAP)) {
+ ata_qc_prep(qc);
+ return;
+ }
+
+ memset(cpb, 0, sizeof(struct nv_adma_cpb));
+
+ cpb->len = 3;
+ cpb->tag = qc->tag;
+ cpb->next_cpb_idx = 0;
+
+ // turn on NCQ flags for NCQ commands
+ if (qc->tf.protocol == ATA_PROT_NCQ)
+ ctl_flags |= NV_CPB_CTL_QUEUE | NV_CPB_CTL_FPDMA;
+
+ nv_adma_tf_to_cpb(&qc->tf, cpb->tf);
+
+ nv_adma_fill_sg(qc, cpb);
+
+ /* Be paranoid and don't let the device see NV_CPB_CTL_CPB_VALID until we are
+ finished filling in all of the contents */
+ wmb();
+ cpb->ctl_flags = ctl_flags;
+}
+
+static unsigned int nv_adma_qc_issue(struct ata_queued_cmd *qc)
+{
+ void __iomem *mmio = nv_adma_ctl_block(qc->ap);
+
+ VPRINTK("ENTER\n");
+
+ if (!(qc->flags & ATA_QCFLAG_DMAMAP)) {
+ VPRINTK("no dmamap, using ATA register mode: 0x%x\n", (u32)qc->flags);
+ // use ATA register mode
+ nv_adma_register_mode(qc->ap);
+ return ata_qc_issue_prot(qc);
+ } else
+ nv_adma_mode(qc->ap);
+
+ //
+ // write append register, command tag in lower 8 bits
+ // and (number of cpbs to append -1) in top 8 bits
+ //
+ wmb();
+ writew(qc->tag, mmio + NV_ADMA_APPEND);
+
+ DPRINTK("Issued tag %u\n",qc->tag);
+
+ return 0;
+}
+
static irqreturn_t nv_generic_interrupt(int irq, void *dev_instance,
struct pt_regs *regs)
{
@@ -461,12 +1262,47 @@ static int nv_hardreset(struct ata_port
return sata_std_hardreset(ap, &dummy);
}
+static int nv_adma_hardreset(struct ata_port *ap, unsigned int *class)
+{
+ unsigned int dummy;
+
+ nv_adma_reset_channel(ap);
+
+ /* SATA hardreset fails to retrieve proper device signature on
+ * some controllers. Don't classify on hardreset. For more
+ * info, see http://bugme.osdl.org/show_bug.cgi?id=3352
+ */
+ return sata_std_hardreset(ap, &dummy);
+}
+
static void nv_error_handler(struct ata_port *ap)
{
ata_bmdma_drive_eh(ap, ata_std_prereset, ata_std_softreset,
nv_hardreset, ata_std_postreset);
}
+static void nv_adma_error_handler(struct ata_port *ap)
+{
+ struct nv_adma_port_priv *pp = ap->private_data;
+ void __iomem *mmio = nv_adma_ctl_block(ap);
+ int i;
+ u32 notifier = readl(mmio + NV_ADMA_NOTIFIER);
+ u32 notifier_error = readl(mmio + NV_ADMA_NOTIFIER_ERROR);
+ u32 gen_ctl = readl(nv_adma_gen_block(ap) + NV_ADMA_GEN_CTL);
+ u32 status = readw(mmio + NV_ADMA_STAT);
+
+ dev_printk(KERN_ERR, ap->dev, "Error handling port %u, notifier 0x%X, notifier_error 0x%X, gen_ctl 0x%X, int pending %d, status 0x%X, dumping CPB states\n",ap->port_no,
+ notifier, notifier_error, gen_ctl, NV_ADMA_CHECK_INTR(gen_ctl, ap->port_no), status);
+ for( i=0;i<NV_ADMA_MAX_CPBS;i++) {
+ struct nv_adma_cpb *cpb = &pp->cpb[i];
+ if( cpb->ctl_flags || cpb->resp_flags )
+ dev_printk(KERN_ERR, ap->dev, "CPB %d: ctl_flags 0x%x, resp_flags 0x%x\n", i, cpb->ctl_flags, cpb->resp_flags);
+ }
+
+ ata_bmdma_drive_eh(ap, ata_std_prereset, ata_std_softreset,
+ nv_adma_hardreset, ata_std_postreset);
+}
+
static int nv_init_one (struct pci_dev *pdev, const struct pci_device_id *ent)
{
static int printed_version = 0;
@@ -476,6 +1312,8 @@ static int nv_init_one (struct pci_dev *
int rc;
u32 bar;
unsigned long base;
+ unsigned long type = ent->driver_data;
+ u64 dma_mask = ATA_DMA_MASK;
// Make sure this is a SATA controller by counting the number of bars
// (NVIDIA SATA controllers will always have six bars). Otherwise,
@@ -486,6 +1324,12 @@ static int nv_init_one (struct pci_dev *
if (!printed_version++)
dev_printk(KERN_DEBUG, &pdev->dev, "version " DRV_VERSION "\n");
+
+ if(type >= CK804 && adma_enabled) {
+ dev_printk(KERN_NOTICE, &pdev->dev, "Using ADMA mode\n");
+ type = ADMA;
+ dma_mask = DMA_64BIT_MASK;
+ }
rc = pci_enable_device(pdev);
if (rc)
@@ -497,16 +1341,16 @@ static int nv_init_one (struct pci_dev *
goto err_out_disable;
}
- rc = pci_set_dma_mask(pdev, ATA_DMA_MASK);
+ rc = pci_set_dma_mask(pdev, dma_mask);
if (rc)
goto err_out_regions;
- rc = pci_set_consistent_dma_mask(pdev, ATA_DMA_MASK);
+ rc = pci_set_consistent_dma_mask(pdev, dma_mask);
if (rc)
goto err_out_regions;
rc = -ENOMEM;
- ppi[0] = ppi[1] = &nv_port_info[ent->driver_data];
+ ppi[0] = ppi[1] = &nv_port_info[type];
probe_ent = ata_pci_init_native_mode(pdev, ppi, ATA_PORT_PRIMARY | ATA_PORT_SECONDARY);
if (!probe_ent)
goto err_out_regions;
@@ -523,7 +1367,7 @@ static int nv_init_one (struct pci_dev *
probe_ent->port[1].scr_addr = base + NV_PORT1_SCR_REG_OFFSET;
/* enable SATA space for CK804 */
- if (ent->driver_data == CK804) {
+ if (type >= CK804) {
u8 regval;
pci_read_config_byte(pdev, NV_MCP_SATA_CFG_20, ®val);
@@ -533,6 +1377,12 @@ static int nv_init_one (struct pci_dev *
pci_set_master(pdev);
+ if (type == ADMA) {
+ rc = nv_adma_host_init(probe_ent);
+ if (rc)
+ goto err_out_iounmap;
+ }
+
rc = ata_device_add(probe_ent);
if (rc != NV_PORTS)
goto err_out_iounmap;
@@ -567,6 +1417,33 @@ static void nv_ck804_host_stop(struct at
ata_pci_host_stop(host);
}
+static void nv_adma_host_stop(struct ata_host *host)
+{
+ struct pci_dev *pdev = to_pci_dev(host->dev);
+ int i;
+ u32 tmp32;
+
+ for (i = 0; i < host->n_ports; i++) {
+ void __iomem *mmio = __nv_adma_ctl_block(host->mmio_base, i);
+ u16 tmp;
+
+ /* disable interrupt */
+ tmp = readw(mmio + NV_ADMA_CTL);
+ writew(tmp & ~NV_ADMA_CTL_AIEN, mmio + NV_ADMA_CTL);
+ }
+
+ // disable ADMA on the ports
+ pci_read_config_dword(pdev, NV_MCP_SATA_CFG_20, &tmp32);
+ tmp32 &= ~(NV_MCP_SATA_CFG_20_PORT0_EN |
+ NV_MCP_SATA_CFG_20_PORT0_PWB_EN |
+ NV_MCP_SATA_CFG_20_PORT1_EN |
+ NV_MCP_SATA_CFG_20_PORT1_PWB_EN);
+
+ pci_write_config_dword(pdev, NV_MCP_SATA_CFG_20, tmp32);
+
+ nv_ck804_host_stop(host);
+}
+
static int __init nv_init(void)
{
return pci_register_driver(&nv_pci_driver);
@@ -579,3 +1456,5 @@ static void __exit nv_exit(void)
module_init(nv_init);
module_exit(nv_exit);
+module_param_named(adma, adma_enabled, bool, 0444);
+MODULE_PARM_DESC(adma, "Enable use of ADMA (Default: false)");
^ permalink raw reply [flat|nested] 23+ messages in thread
* Re: [RFC PATCH] nForce4 ADMA with NCQ: It's aliiiive..
2006-10-07 8:08 [RFC PATCH] nForce4 ADMA with NCQ: It's aliiiive Robert Hancock
@ 2006-10-07 9:42 ` Prakash Punnoor
2006-10-07 15:54 ` Robert Hancock
2006-10-07 15:28 ` Jeff Garzik
1 sibling, 1 reply; 23+ messages in thread
From: Prakash Punnoor @ 2006-10-07 9:42 UTC (permalink / raw)
To: Robert Hancock; +Cc: linux-kernel, linux-ide, jeff
[-- Attachment #1: Type: text/plain, Size: 618 bytes --]
Am Samstag 07 Oktober 2006 10:08 schrieb Robert Hancock:
> I've been working on the patch for sata_nv ADMA support for nForce4 that
Nice!
> make the default 1). I only enabled ADMA on those chipsets and not
> MCP51, MCP55 or MCP61 since that was all that the original NVIDIA
> version did. I assume there was a reason for this, though maybe not.
Unfortunately it doesn't work for me on MCP51 if I change GENERIC to ADMA. So
I wonder whether MCP51 has ADMA mode or what needs to be done to get NCQ
working. :-(
Cheers,
--
(°= =°)
//\ Prakash Punnoor /\\
V_/ \_V
[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]
^ permalink raw reply [flat|nested] 23+ messages in thread
* Re: [RFC PATCH] nForce4 ADMA with NCQ: It's aliiiive..
2006-10-07 8:08 [RFC PATCH] nForce4 ADMA with NCQ: It's aliiiive Robert Hancock
2006-10-07 9:42 ` Prakash Punnoor
@ 2006-10-07 15:28 ` Jeff Garzik
2006-10-07 23:49 ` Robert Hancock
2006-10-10 6:52 ` Allen Martin
1 sibling, 2 replies; 23+ messages in thread
From: Jeff Garzik @ 2006-10-07 15:28 UTC (permalink / raw)
To: Robert Hancock; +Cc: linux-kernel, linux-ide, prakash
Robert Hancock wrote:
> I've been working on the patch for sata_nv ADMA support for nForce4 that
> Jeff Garzik has in a git branch. I've gotten it into a state where the
> ADMA and NCQ features appear to be working with no obvious problems.
> I've attached a patch against 2.6.18-mm2.
>
> The code was mostly in a working state for the non-NCQ case but there
> were a number of heinous bugs that prevented NCQ from working, like in
> the sg list to APRD conversion code and in the interrupt handler.
>
> This is still in quite an experimental state. It has survived system
> boots into Fedora Core 5 and Bonnie++ benchmark runs without blowing up,
> but there could still be bugs that could corrupt data, etc. so test with
> caution.
>
> There is a module parameter adma_enabled which has to be set to 1 to
> enable ADMA on CK804/MCP04 chipsets (either that or hack the code to
> make the default 1). I only enabled ADMA on those chipsets and not
> MCP51, MCP55 or MCP61 since that was all that the original NVIDIA
> version did. I assume there was a reason for this, though maybe not.
> Someone with one of these chipsets should probably try it out (replacing
> the GENERIC type with CK804 in the PCI device table may be all it takes).
Nice! Good work.
> A few outstanding issues:
>
> -Error handling likely needs work. EH works well enough to get past
> drive detection but that's likely about all. When I ran into errors
> while debugging, it usually locked up the machine when trying to do a
> soft reset.
>
> -Error handling is also noisy at the moment (it dumps a bunch of
> controller state information).
>
> -Jeff will probably cringe at how I implemented the
> bmdma_stop/start/status/setup functions. This kludge of toggling
> ATA_FLAG_MMIO off for the call into libata was needed since this
> controller is almost what libata calls ATA_FLAG_MMIO, but not quite (the
> ATA taskfile registers are MMIO but the BMDMA registers are PIO). This
> is also why I needed the patch to libata-sff.c to use the adapter's
> bmdma_status function rather than hardcoded ata_bmdma_status.
*shrug* I don't cringe if that's the most expedient way to do something.
But I really don't think that is necessary. I will take a look at docs
and see how things match up, when I am much more awake. Most likely you
need to be using another set of registers, and be all MMIO, all the time.
Jeff
^ permalink raw reply [flat|nested] 23+ messages in thread
* Re: [RFC PATCH] nForce4 ADMA with NCQ: It's aliiiive..
2006-10-07 9:42 ` Prakash Punnoor
@ 2006-10-07 15:54 ` Robert Hancock
2006-10-10 6:44 ` Allen Martin
0 siblings, 1 reply; 23+ messages in thread
From: Robert Hancock @ 2006-10-07 15:54 UTC (permalink / raw)
To: Prakash Punnoor; +Cc: linux-kernel, linux-ide, jeff
Prakash Punnoor wrote:
> Am Samstag 07 Oktober 2006 10:08 schrieb Robert Hancock:
>> I've been working on the patch for sata_nv ADMA support for nForce4 that
>
> Nice!
>
>> make the default 1). I only enabled ADMA on those chipsets and not
>> MCP51, MCP55 or MCP61 since that was all that the original NVIDIA
>> version did. I assume there was a reason for this, though maybe not.
>
> Unfortunately it doesn't work for me on MCP51 if I change GENERIC to ADMA. So
> I wonder whether MCP51 has ADMA mode or what needs to be done to get NCQ
> working. :-(
What happened when you tried it? It would be useful if you could change
the #undef in these lines:
53 #undef ATA_DEBUG /* debugging output */
54 #undef ATA_VERBOSE_DEBUG /* yet more debugging output */
in include/linux/libata.h to #define and rebuild and try it then, that
will spew out a bunch more output and I can see if any reasonable
looking values are showing up at all. I was capturing this output the
crude way, booting with vga=6 to get a smaller font and taking a picture
of the screen :-) Also, maybe post the lspci -v output from the SATA
controller..
If that doesn't provide any insight, maybe the docs Jeff has provide the
answer for whether or not the MCP5x/MCP61 controllers have the same
interface as the CK804/MCP04..
--
Robert Hancock Saskatoon, SK, Canada
To email, remove "nospam" from hancockr@nospamshaw.ca
Home Page: http://www.roberthancock.com/
^ permalink raw reply [flat|nested] 23+ messages in thread
* Re: [RFC PATCH] nForce4 ADMA with NCQ: It's aliiiive..
2006-10-07 15:28 ` Jeff Garzik
@ 2006-10-07 23:49 ` Robert Hancock
2006-10-10 6:52 ` Allen Martin
1 sibling, 0 replies; 23+ messages in thread
From: Robert Hancock @ 2006-10-07 23:49 UTC (permalink / raw)
To: Jeff Garzik; +Cc: linux-kernel, linux-ide, prakash
[-- Attachment #1: Type: text/plain, Size: 1545 bytes --]
Jeff Garzik wrote:
>> -Jeff will probably cringe at how I implemented the
>> bmdma_stop/start/status/setup functions. This kludge of toggling
>> ATA_FLAG_MMIO off for the call into libata was needed since this
>> controller is almost what libata calls ATA_FLAG_MMIO, but not quite
>> (the ATA taskfile registers are MMIO but the BMDMA registers are PIO).
>> This is also why I needed the patch to libata-sff.c to use the
>> adapter's bmdma_status function rather than hardcoded ata_bmdma_status.
>
> *shrug* I don't cringe if that's the most expedient way to do something.
>
> But I really don't think that is necessary. I will take a look at docs
> and see how things match up, when I am much more awake. Most likely you
> need to be using another set of registers, and be all MMIO, all the time.
>
> Jeff
Hopefully that is the case, it would clean things up a bit..
I've cut another patch, this one against 2.6.18-mm3. I've fixed a couple
of issues:
-Error handling now pushes the port back into port-register mode so that
the reset will actually work properly.
-It seems there was a problem with the checks for deciding whether to
scan the CPB list in the interrupt handler that occasionally caused
interrupts to get lost and commands to time out (which unintentionally
tested the EH fix above). I think this is fixed now.
Signed-off-by: Robert Hancock <hancockr@shaw.ca>
--
Robert Hancock Saskatoon, SK, Canada
To email, remove "nospam" from hancockr@nospamshaw.ca
Home Page: http://www.roberthancock.com/
[-- Attachment #2: sata_nv-adma-ncq-v2.patch --]
[-- Type: text/plain, Size: 30352 bytes --]
diff -rup --exclude='*.o' --exclude='*mod.c' --exclude='*.cmd' --exclude='*.ko' linux-2.6.18-mm3/drivers/ata/libata-sff.c linux-2.6.18-mm3-adma/drivers/ata/libata-sff.c
--- linux-2.6.18-mm3/drivers/ata/libata-sff.c 2006-10-07 12:11:11.000000000 -0600
+++ linux-2.6.18-mm3-adma/drivers/ata/libata-sff.c 2006-10-07 12:12:57.000000000 -0600
@@ -732,7 +732,7 @@ void ata_bmdma_drive_eh(struct ata_port
qc->tf.protocol == ATA_PROT_ATAPI_DMA)) {
u8 host_stat;
- host_stat = ata_bmdma_status(ap);
+ host_stat = ap->ops->bmdma_status(ap);
ata_ehi_push_desc(&ehc->i, "BMDMA stat 0x%x", host_stat);
diff -rup --exclude='*.o' --exclude='*mod.c' --exclude='*.cmd' --exclude='*.ko' linux-2.6.18-mm3/drivers/ata/sata_nv.c linux-2.6.18-mm3-adma/drivers/ata/sata_nv.c
--- linux-2.6.18-mm3/drivers/ata/sata_nv.c 2006-10-07 12:11:11.000000000 -0600
+++ linux-2.6.18-mm3-adma/drivers/ata/sata_nv.c 2006-10-07 16:46:49.000000000 -0600
@@ -29,6 +29,11 @@
* NV-specific details such as register offsets, SATA phy location,
* hotplug info, etc.
*
+ * CK804 (and later?) controllers support an alternate programming interface
+ * similar to the ADMA specification (with some modifications).
+ * This allows the use of NCQ. Currently non-DMA-mapped ATA commands are still
+ * sent through the legacy interface.
+ *
*/
#include <linux/kernel.h>
@@ -43,7 +48,7 @@
#include <linux/libata.h>
#define DRV_NAME "sata_nv"
-#define DRV_VERSION "2.0"
+#define DRV_VERSION "3.0"
enum {
NV_PORTS = 2,
@@ -78,8 +83,135 @@ enum {
// For PCI config register 20
NV_MCP_SATA_CFG_20 = 0x50,
NV_MCP_SATA_CFG_20_SATA_SPACE_EN = 0x04,
+ NV_MCP_SATA_CFG_20_PORT0_EN = (1 << 17),
+ NV_MCP_SATA_CFG_20_PORT1_EN = (1 << 16),
+ NV_MCP_SATA_CFG_20_PORT0_PWB_EN = (1 << 14),
+ NV_MCP_SATA_CFG_20_PORT1_PWB_EN = (1 << 12),
+
+ NV_ADMA_MAX_CPBS = 32,
+ NV_ADMA_CPB_SZ = 128,
+ NV_ADMA_APRD_SZ = 16,
+ NV_ADMA_SGTBL_LEN = (1024 - NV_ADMA_CPB_SZ) / NV_ADMA_APRD_SZ,
+ NV_ADMA_SGTBL_SZ = NV_ADMA_SGTBL_LEN * NV_ADMA_APRD_SZ,
+ NV_ADMA_PORT_PRIV_DMA_SZ = NV_ADMA_MAX_CPBS * (NV_ADMA_CPB_SZ + NV_ADMA_SGTBL_SZ),
+
+ // BAR5 offset to ADMA general registers
+ NV_ADMA_GEN = 0x400,
+ NV_ADMA_GEN_CTL = 0x00,
+ NV_ADMA_NOTIFIER_CLEAR = 0x30,
+
+ // BAR5 offset to ADMA ports
+ NV_ADMA_PORT = 0x480,
+
+ // size of ADMA port register space
+ NV_ADMA_PORT_SIZE = 0x100,
+
+ // ADMA port registers
+ NV_ADMA_CTL = 0x40,
+ NV_ADMA_CPB_COUNT = 0x42,
+ NV_ADMA_NEXT_CPB_IDX = 0x43,
+ NV_ADMA_STAT = 0x44,
+ NV_ADMA_CPB_BASE_LOW = 0x48,
+ NV_ADMA_CPB_BASE_HIGH = 0x4C,
+ NV_ADMA_APPEND = 0x50,
+ NV_ADMA_NOTIFIER = 0x68,
+ NV_ADMA_NOTIFIER_ERROR = 0x6C,
+
+ // NV_ADMA_CTL register bits
+ NV_ADMA_CTL_HOTPLUG_IEN = (1 << 0),
+ NV_ADMA_CTL_CHANNEL_RESET = (1 << 5),
+ NV_ADMA_CTL_GO = (1 << 7),
+ NV_ADMA_CTL_AIEN = (1 << 8),
+ NV_ADMA_CTL_READ_NON_COHERENT = (1 << 11),
+ NV_ADMA_CTL_WRITE_NON_COHERENT = (1 << 12),
+
+ // CPB response flag bits
+ NV_CPB_RESP_DONE = (1 << 0),
+ NV_CPB_RESP_ATA_ERR = (1 << 3),
+ NV_CPB_RESP_CMD_ERR = (1 << 4),
+ NV_CPB_RESP_CPB_ERR = (1 << 7),
+
+ // CPB control flag bits
+ NV_CPB_CTL_CPB_VALID = (1 << 0),
+ NV_CPB_CTL_QUEUE = (1 << 1),
+ NV_CPB_CTL_APRD_VALID = (1 << 2),
+ NV_CPB_CTL_IEN = (1 << 3),
+ NV_CPB_CTL_FPDMA = (1 << 4),
+
+ // APRD flags
+ NV_APRD_WRITE = (1 << 1),
+ NV_APRD_END = (1 << 2),
+ NV_APRD_CONT = (1 << 3),
+
+ // NV_ADMA_STAT flags
+ NV_ADMA_STAT_TIMEOUT = (1 << 0),
+ NV_ADMA_STAT_HOTUNPLUG = (1 << 1),
+ NV_ADMA_STAT_HOTPLUG = (1 << 2),
+ NV_ADMA_STAT_CPBERR = (1 << 4),
+ NV_ADMA_STAT_SERROR = (1 << 5),
+ NV_ADMA_STAT_CMD_COMPLETE = (1 << 6),
+ NV_ADMA_STAT_IDLE = (1 << 8),
+ NV_ADMA_STAT_LEGACY = (1 << 9),
+ NV_ADMA_STAT_STOPPED = (1 << 10),
+ NV_ADMA_STAT_DONE = (1 << 12),
+ NV_ADMA_STAT_ERR = (NV_ADMA_STAT_CPBERR | NV_ADMA_STAT_TIMEOUT),
+
+ // port flags
+ NV_ADMA_PORT_REGISTER_MODE = (1 << 0),
+
+};
+
+/* ADMA Physical Region Descriptor - one SG segment */
+struct nv_adma_prd {
+ __le64 addr;
+ __le32 len;
+ u8 flags;
+ u8 packet_len;
+ __le16 reserved;
+};
+
+enum nv_adma_regbits {
+ CMDEND = (1 << 15), /* end of command list */
+ WNB = (1 << 14), /* wait-not-BSY */
+ IGN = (1 << 13), /* ignore this entry */
+ CS1n = (1 << (4 + 8)), /* std. PATA signals follow... */
+ DA2 = (1 << (2 + 8)),
+ DA1 = (1 << (1 + 8)),
+ DA0 = (1 << (0 + 8)),
};
+/* ADMA Command Parameter Block
+ The first 5 SG segments are stored inside the Command Parameter Block itself.
+ If there are more than 5 segments the remainder are stored in a separate
+ memory area indicated by next_aprd. */
+struct nv_adma_cpb {
+ u8 resp_flags; //0
+ u8 reserved1; //1
+ u8 ctl_flags; //2
+ // len is length of taskfile in 64 bit words
+ u8 len; //3
+ u8 tag; //4
+ u8 next_cpb_idx; //5
+ __le16 reserved2; //6-7
+ __le16 tf[12]; //8-31
+ struct nv_adma_prd aprd[5]; //32-111
+ __le64 next_aprd; //112-119
+ __le64 reserved3; //120-127
+};
+
+
+struct nv_adma_port_priv {
+ struct nv_adma_cpb *cpb;
+ dma_addr_t cpb_dma;
+ struct nv_adma_prd *aprd;
+ dma_addr_t aprd_dma;
+ u8 flags;
+ u32 notifier;
+ u32 notifier_error;
+};
+
+#define NV_ADMA_CHECK_INTR(GCTL, PORT) ((GCTL) & ( 1 << (19 + (12 * (PORT)))))
+
static int nv_init_one (struct pci_dev *pdev, const struct pci_device_id *ent);
static void nv_ck804_host_stop(struct ata_host *host);
static irqreturn_t nv_generic_interrupt(int irq, void *dev_instance,
@@ -96,13 +228,27 @@ static void nv_nf2_thaw(struct ata_port
static void nv_ck804_freeze(struct ata_port *ap);
static void nv_ck804_thaw(struct ata_port *ap);
static void nv_error_handler(struct ata_port *ap);
+static void nv_adma_qc_prep(struct ata_queued_cmd *qc);
+static unsigned int nv_adma_qc_issue(struct ata_queued_cmd *qc);
+static irqreturn_t nv_adma_interrupt(int irq, void *dev_instance,
+ struct pt_regs *regs);
+static void nv_adma_irq_clear(struct ata_port *ap);
+static int nv_adma_port_start(struct ata_port *ap);
+static void nv_adma_port_stop(struct ata_port *ap);
+static void nv_adma_error_handler(struct ata_port *ap);
+static void nv_adma_host_stop(struct ata_host *host);
+static void nv_adma_bmdma_setup(struct ata_queued_cmd *qc);
+static void nv_adma_bmdma_start(struct ata_queued_cmd *qc);
+static void nv_adma_bmdma_stop(struct ata_queued_cmd *qc);
+static u8 nv_adma_bmdma_status(struct ata_port *ap);
enum nv_host_type
{
GENERIC,
NFORCE2,
NFORCE3 = NFORCE2, /* NF2 == NF3 as far as sata_nv is concerned */
- CK804
+ CK804,
+ ADMA
};
static const struct pci_device_id nv_pci_tbl[] = {
@@ -159,6 +305,25 @@ static struct scsi_host_template nv_sht
.bios_param = ata_std_bios_param,
};
+static struct scsi_host_template nv_adma_sht = {
+ .module = THIS_MODULE,
+ .name = DRV_NAME,
+ .ioctl = ata_scsi_ioctl,
+ .queuecommand = ata_scsi_queuecmd,
+ .can_queue = NV_ADMA_MAX_CPBS,
+ .this_id = ATA_SHT_THIS_ID,
+ .sg_tablesize = NV_ADMA_SGTBL_LEN + 5,
+ .max_sectors = ATA_MAX_SECTORS,
+ .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 const struct ata_port_operations nv_generic_ops = {
.port_disable = ata_port_disable,
.tf_load = ata_tf_load,
@@ -240,6 +405,33 @@ static const struct ata_port_operations
.host_stop = nv_ck804_host_stop,
};
+static const struct ata_port_operations nv_adma_ops = {
+ .port_disable = ata_port_disable,
+ .tf_load = ata_tf_load,
+ .tf_read = ata_tf_read,
+ .exec_command = ata_exec_command,
+ .check_status = ata_check_status,
+ .dev_select = ata_std_dev_select,
+ .bmdma_setup = nv_adma_bmdma_setup,
+ .bmdma_start = nv_adma_bmdma_start,
+ .bmdma_stop = nv_adma_bmdma_stop,
+ .bmdma_status = nv_adma_bmdma_status,
+ .qc_prep = nv_adma_qc_prep,
+ .qc_issue = nv_adma_qc_issue,
+ .freeze = nv_ck804_freeze,
+ .thaw = nv_ck804_thaw,
+ .error_handler = nv_adma_error_handler,
+ .post_internal_cmd = nv_adma_bmdma_stop,
+ .data_xfer = ata_mmio_data_xfer,
+ .irq_handler = nv_adma_interrupt,
+ .irq_clear = nv_adma_irq_clear,
+ .scr_read = nv_scr_read,
+ .scr_write = nv_scr_write,
+ .port_start = nv_adma_port_start,
+ .port_stop = nv_adma_port_stop,
+ .host_stop = nv_adma_host_stop,
+};
+
static struct ata_port_info nv_port_info[] = {
/* generic */
{
@@ -268,6 +460,16 @@ static struct ata_port_info nv_port_info
.udma_mask = NV_UDMA_MASK,
.port_ops = &nv_ck804_ops,
},
+ /* ADMA */
+ {
+ .sht = &nv_adma_sht,
+ .flags = ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY |
+ ATA_FLAG_MMIO | ATA_FLAG_NCQ,
+ .pio_mask = NV_PIO_MASK,
+ .mwdma_mask = NV_MWDMA_MASK,
+ .udma_mask = NV_UDMA_MASK,
+ .port_ops = &nv_adma_ops,
+ },
};
MODULE_AUTHOR("NVIDIA");
@@ -276,6 +478,579 @@ MODULE_LICENSE("GPL");
MODULE_DEVICE_TABLE(pci, nv_pci_tbl);
MODULE_VERSION(DRV_VERSION);
+static int adma_enabled;
+
+static unsigned int nv_adma_tf_to_cpb(struct ata_taskfile *tf, u16 *cpb)
+{
+ unsigned int idx = 0;
+
+ cpb[idx++] = cpu_to_le16((ATA_REG_DEVICE << 8) | tf->device | WNB);
+
+ if ((tf->flags & ATA_TFLAG_LBA48) == 0) {
+ cpb[idx++] = cpu_to_le16(IGN);
+ cpb[idx++] = cpu_to_le16(IGN);
+ cpb[idx++] = cpu_to_le16(IGN);
+ cpb[idx++] = cpu_to_le16(IGN);
+ cpb[idx++] = cpu_to_le16(IGN);
+ }
+ else {
+ cpb[idx++] = cpu_to_le16((ATA_REG_ERR << 8) | tf->hob_feature);
+ cpb[idx++] = cpu_to_le16((ATA_REG_NSECT << 8) | tf->hob_nsect);
+ cpb[idx++] = cpu_to_le16((ATA_REG_LBAL << 8) | tf->hob_lbal);
+ cpb[idx++] = cpu_to_le16((ATA_REG_LBAM << 8) | tf->hob_lbam);
+ cpb[idx++] = cpu_to_le16((ATA_REG_LBAH << 8) | tf->hob_lbah);
+ }
+ cpb[idx++] = cpu_to_le16((ATA_REG_ERR << 8) | tf->feature);
+ cpb[idx++] = cpu_to_le16((ATA_REG_NSECT << 8) | tf->nsect);
+ cpb[idx++] = cpu_to_le16((ATA_REG_LBAL << 8) | tf->lbal);
+ cpb[idx++] = cpu_to_le16((ATA_REG_LBAM << 8) | tf->lbam);
+ cpb[idx++] = cpu_to_le16((ATA_REG_LBAH << 8) | tf->lbah);
+
+ cpb[idx++] = cpu_to_le16((ATA_REG_CMD << 8) | tf->command | CMDEND);
+
+ return idx;
+}
+
+static inline void __iomem *__nv_adma_ctl_block(void __iomem *mmio,
+ unsigned int port_no)
+{
+ mmio += NV_ADMA_PORT + port_no * NV_ADMA_PORT_SIZE;
+ return mmio;
+}
+
+static inline void __iomem *nv_adma_ctl_block(struct ata_port *ap)
+{
+ return __nv_adma_ctl_block(ap->host->mmio_base, ap->port_no);
+}
+
+static inline void __iomem *nv_adma_gen_block(struct ata_port *ap)
+{
+ return (ap->host->mmio_base + NV_ADMA_GEN);
+}
+
+static inline void __iomem *nv_adma_notifier_clear_block(struct ata_port *ap)
+{
+ return (nv_adma_gen_block(ap) + NV_ADMA_NOTIFIER_CLEAR + (4 * ap->port_no));
+}
+
+static int nv_adma_host_intr(struct ata_port *ap)
+{
+ void __iomem *mmio = nv_adma_ctl_block(ap);
+ struct nv_adma_port_priv *pp = ap->private_data;
+ u16 status;
+ u32 gen_ctl;
+ u16 flags;
+ int have_global_err = 0;
+
+ /* if in ATA register mode, use standard ata interrupt handler */
+ if (pp->flags & NV_ADMA_PORT_REGISTER_MODE) {
+ struct ata_queued_cmd *qc;
+ VPRINTK("in ATA register mode\n");
+ qc = ata_qc_from_tag(ap, ap->active_tag);
+ if (qc && (!(qc->tf.flags & ATA_TFLAG_POLLING)))
+ return ata_host_intr(ap, qc);
+ else {
+ // No request pending? Clear interrupt status
+ // anyway, in case there's one pending.
+ ap->ops->check_status(ap);
+ return 1;
+ }
+ }
+
+ gen_ctl = readl(nv_adma_gen_block(ap) + NV_ADMA_GEN_CTL);
+
+ if( !NV_ADMA_CHECK_INTR(gen_ctl, ap->port_no) && !pp->notifier && !pp->notifier_error)
+ /* Nothing to do */
+ return 0;
+
+ status = readw(mmio + NV_ADMA_STAT);
+
+ /* Clear status so that any CPB completions after this point in the handler will
+ raise another interrupt */
+ writew(status, mmio + NV_ADMA_STAT);
+ readl(mmio + NV_ADMA_STAT); /* flush posted write */
+ mb();
+
+ /* freeze if hotplugged */
+ if (unlikely(status & (NV_ADMA_STAT_HOTPLUG | NV_ADMA_STAT_HOTUNPLUG))) {
+ dev_printk(KERN_NOTICE, ap->dev, "Hotplug event on port %u, freezing\n",
+ ap->port_no);
+ ata_port_freeze(ap);
+ return 1;
+ }
+
+ if (pp->notifier_error) {
+ dev_printk(KERN_ERR, ap->dev, "port %u notifier error, stat=0x%x, "
+ "notifier=0x%X, notifier_error=0x%X", ap->port_no, status, pp->notifier,
+ pp->notifier_error );
+ have_global_err = 1;
+ }
+
+ if (status & NV_ADMA_STAT_TIMEOUT) {
+ dev_printk(KERN_ERR, ap->dev, "port %u timeout, stat=0x%x\n", ap->port_no, status);
+ have_global_err = 1;
+ }
+ if (status & NV_ADMA_STAT_CPBERR) {
+ dev_printk(KERN_ERR, ap->dev, "port %u CPB error, stat=0x%x\n", ap->port_no, status);
+ have_global_err = 1;
+ }
+ if (status & NV_ADMA_STAT_STOPPED) {
+ VPRINTK("ADMA stopped, stat = 0x%x\n", status);
+ }
+ if (status & NV_ADMA_STAT_CMD_COMPLETE) {
+ VPRINTK("ADMA command complete, stat = 0x%x\n", status);
+ }
+ if ((status & NV_ADMA_STAT_DONE) || have_global_err) {
+ /** Check CPBs for completed commands */
+ int i;
+ for( i=0;i<NV_ADMA_MAX_CPBS;i++)
+ if( ap->sactive & (1 << i ) ||
+ ( ata_tag_valid(ap->active_tag) && ap->active_tag == i ) ) {
+ struct nv_adma_cpb *cpb = &pp->cpb[i];
+ int complete = 0, have_err = 0;
+ flags = cpb->resp_flags;
+ VPRINTK("CPB %d, stat=0x%x, flags=0x%x\n", i, status, flags);
+
+ if (flags & NV_CPB_RESP_DONE) {
+ VPRINTK("CPB flags done, flags=0x%x\n", flags);
+ complete = 1;
+ }
+ if (flags & NV_CPB_RESP_ATA_ERR) {
+ dev_printk(KERN_ERR, ap->dev, "port %u CPB flags ATA err, flags=0x%x\n", ap->port_no, flags);
+ have_err = 1;
+ complete = 1;
+ }
+ if (flags & NV_CPB_RESP_CMD_ERR) {
+ dev_printk(KERN_ERR, ap->dev, "port %u CPB flags CMD err, flags=0x%x\n", ap->port_no, flags);
+ have_err = 1;
+ complete = 1;
+ }
+ if (flags & NV_CPB_RESP_CPB_ERR) {
+ dev_printk(KERN_ERR, ap->dev, "port %u CPB flags CPB err, flags=0x%x\n", ap->port_no, flags);
+ have_err = 1;
+ complete = 1;
+ }
+ if(complete || have_global_err)
+ {
+ struct ata_queued_cmd *qc = ata_qc_from_tag(ap, i);
+ if(likely(qc)) {
+ u8 ata_status = 0;
+ /* Only use the ATA port status for non-NCQ commands.
+ For NCQ commands the current status may have nothing to do with
+ the command just completed. */
+ if(qc->tf.protocol != ATA_PROT_NCQ)
+ ata_status = readb(mmio + (ATA_REG_STATUS * 4));
+
+ if(have_err || have_global_err)
+ ata_status |= ATA_ERR;
+
+ qc->err_mask |= ac_err_mask(ata_status);
+ DPRINTK("Completing qc from tag %d with err_mask %u\n",i,qc->err_mask);
+ ata_qc_complete(qc);
+ }
+ }
+ }
+ }
+
+ return 1; /* irq handled if we got here */
+}
+
+static irqreturn_t nv_adma_interrupt(int irq, void *dev_instance,
+ struct pt_regs *regs)
+{
+ struct ata_host *host = dev_instance;
+ int i, handled = 0;
+
+ spin_lock(&host->lock);
+
+ for (i = 0; i < host->n_ports; i++) {
+ struct ata_port *ap = host->ports[i];
+
+ if (ap && !(ap->flags & ATA_FLAG_DISABLED)) {
+ struct nv_adma_port_priv *pp = ap->private_data;
+ void __iomem *mmio = nv_adma_ctl_block(ap);
+ // read notifiers
+ pp->notifier = readl(mmio + NV_ADMA_NOTIFIER);
+ pp->notifier_error = readl(mmio + NV_ADMA_NOTIFIER_ERROR);
+ handled += nv_adma_host_intr(ap);
+ }
+ }
+
+ /* Clear notifiers. Yes, this does appear to have to be done at the very end,
+ otherwise things grind to a halt quite quickly. */
+ if (handled)
+ for (i = 0; i < host->n_ports; i++) {
+ struct ata_port *ap = host->ports[i];
+ struct nv_adma_port_priv *pp = ap->private_data;
+ writel(pp->notifier | pp->notifier_error,
+ nv_adma_notifier_clear_block(ap));
+ }
+
+ spin_unlock(&host->lock);
+
+ return IRQ_RETVAL(handled);
+}
+
+static void nv_adma_irq_clear(struct ata_port *ap)
+{
+ void __iomem *mmio = nv_adma_ctl_block(ap);
+ u16 status = readw(mmio + NV_ADMA_STAT);
+ u32 notifier = readl(mmio + NV_ADMA_NOTIFIER);
+ u32 notifier_error = readl(mmio + NV_ADMA_NOTIFIER_ERROR);
+
+ /* clear ADMA status */
+ writew(status, mmio + NV_ADMA_STAT);
+ writel(notifier | notifier_error,
+ nv_adma_notifier_clear_block(ap));
+
+ /** clear legacy status */
+ ap->flags &= ~ATA_FLAG_MMIO;
+ ata_bmdma_irq_clear(ap);
+ ap->flags |= ATA_FLAG_MMIO;
+}
+
+static void nv_adma_bmdma_setup(struct ata_queued_cmd *qc)
+{
+ struct nv_adma_port_priv *pp = qc->ap->private_data;
+
+ if (!(pp->flags & NV_ADMA_PORT_REGISTER_MODE))
+ return;
+
+ qc->ap->flags &= ~ATA_FLAG_MMIO;
+ ata_bmdma_setup(qc);
+ qc->ap->flags |= ATA_FLAG_MMIO;
+}
+
+static void nv_adma_bmdma_start(struct ata_queued_cmd *qc)
+{
+ struct nv_adma_port_priv *pp = qc->ap->private_data;
+
+ if (!(pp->flags & NV_ADMA_PORT_REGISTER_MODE))
+ return;
+
+ qc->ap->flags &= ~ATA_FLAG_MMIO;
+ ata_bmdma_start(qc);
+ qc->ap->flags |= ATA_FLAG_MMIO;
+}
+
+static void nv_adma_bmdma_stop(struct ata_queued_cmd *qc)
+{
+ struct nv_adma_port_priv *pp = qc->ap->private_data;
+
+ if (!(pp->flags & NV_ADMA_PORT_REGISTER_MODE))
+ return;
+
+ qc->ap->flags &= ~ATA_FLAG_MMIO;
+ ata_bmdma_stop(qc);
+ qc->ap->flags |= ATA_FLAG_MMIO;
+}
+
+static u8 nv_adma_bmdma_status(struct ata_port *ap)
+{
+ u8 status;
+ ap->flags &= ~ATA_FLAG_MMIO;
+ status = ata_bmdma_status(ap);
+ ap->flags |= ATA_FLAG_MMIO;
+ return status;
+}
+
+static void nv_adma_register_mode(struct ata_port *ap)
+{
+ void __iomem *mmio = nv_adma_ctl_block(ap);
+ struct nv_adma_port_priv *pp = ap->private_data;
+ u16 tmp;
+
+ if (pp->flags & NV_ADMA_PORT_REGISTER_MODE)
+ return;
+
+ tmp = readw(mmio + NV_ADMA_CTL);
+ writew(tmp & ~NV_ADMA_CTL_GO, mmio + NV_ADMA_CTL);
+
+ pp->flags |= NV_ADMA_PORT_REGISTER_MODE;
+}
+
+static void nv_adma_mode(struct ata_port *ap)
+{
+ void __iomem *mmio = nv_adma_ctl_block(ap);
+ struct nv_adma_port_priv *pp = ap->private_data;
+ u16 tmp;
+
+ if (!(pp->flags & NV_ADMA_PORT_REGISTER_MODE))
+ return;
+
+ tmp = readw(mmio + NV_ADMA_CTL);
+ writew(tmp | NV_ADMA_CTL_GO, mmio + NV_ADMA_CTL);
+
+ pp->flags &= ~NV_ADMA_PORT_REGISTER_MODE;
+}
+
+static int nv_adma_port_start(struct ata_port *ap)
+{
+ struct device *dev = ap->host->dev;
+ struct nv_adma_port_priv *pp;
+ int rc;
+ void *mem;
+ dma_addr_t mem_dma;
+ void __iomem *mmio = nv_adma_ctl_block(ap);
+ u16 tmp;
+
+ VPRINTK("ENTER\n");
+
+ rc = ata_port_start(ap);
+ if (rc)
+ return rc;
+
+ pp = kzalloc(sizeof(*pp), GFP_KERNEL);
+ if (!pp) {
+ rc = -ENOMEM;
+ goto err_out;
+ }
+
+ mem = dma_alloc_coherent(dev, NV_ADMA_PORT_PRIV_DMA_SZ,
+ &mem_dma, GFP_KERNEL);
+
+ VPRINTK("dma memory: vaddr = 0x%p, paddr = 0x%p\n", mem, (void*)mem_dma);
+
+ if (!mem) {
+ rc = -ENOMEM;
+ goto err_out_kfree;
+ }
+ memset(mem, 0, NV_ADMA_PORT_PRIV_DMA_SZ);
+
+ /*
+ * First item in chunk of DMA memory:
+ * 128-byte command parameter block (CPB)
+ * one for each command tag
+ */
+ pp->cpb = mem;
+ pp->cpb_dma = mem_dma;
+
+ VPRINTK("cpb = 0x%p, cpb_dma = 0x%p\n", pp->cpb, (void*)pp->cpb_dma);
+
+ writel(mem_dma & 0xFFFFFFFF, mmio + NV_ADMA_CPB_BASE_LOW);
+ writel((mem_dma >> 16 ) >> 16, mmio + NV_ADMA_CPB_BASE_HIGH);
+
+ mem += NV_ADMA_MAX_CPBS * NV_ADMA_CPB_SZ;
+ mem_dma += NV_ADMA_MAX_CPBS * NV_ADMA_CPB_SZ;
+
+ /*
+ * Second item: block of ADMA_SGTBL_LEN s/g entries
+ */
+ pp->aprd = mem;
+ pp->aprd_dma = mem_dma;
+
+ VPRINTK("aprd = 0x%p, aprd_dma = 0x%p\n", pp->aprd, (void*)pp->aprd_dma);
+
+ ap->private_data = pp;
+
+ // clear any outstanding interrupt conditions
+ writew(0xffff, mmio + NV_ADMA_STAT);
+
+ // initialize port variables
+ pp->flags = NV_ADMA_PORT_REGISTER_MODE;
+
+ /* clear CPB fetch count */
+ writew(0, mmio + NV_ADMA_CPB_COUNT);
+
+ /* clear GO for register mode */
+ tmp = readw(mmio + NV_ADMA_CTL);
+ writew(tmp & ~NV_ADMA_CTL_GO, mmio + NV_ADMA_CTL);
+
+ tmp = readw(mmio + NV_ADMA_CTL);
+ writew(tmp | NV_ADMA_CTL_CHANNEL_RESET, mmio + NV_ADMA_CTL);
+ readl( mmio + NV_ADMA_CTL ); /* flush posted write */
+ udelay(1);
+ writew(tmp & ~NV_ADMA_CTL_CHANNEL_RESET, mmio + NV_ADMA_CTL);
+ readl( mmio + NV_ADMA_CTL ); /* flush posted write */
+
+ return 0;
+
+err_out_kfree:
+ kfree(pp);
+err_out:
+ ata_port_stop(ap);
+ return rc;
+}
+
+static void nv_adma_port_stop(struct ata_port *ap)
+{
+ struct device *dev = ap->host->dev;
+ struct nv_adma_port_priv *pp = ap->private_data;
+ void __iomem *mmio = nv_adma_ctl_block(ap);
+
+ VPRINTK("ENTER\n");
+
+ writew(0, mmio + NV_ADMA_CTL);
+
+ ap->private_data = NULL;
+ dma_free_coherent(dev, NV_ADMA_PORT_PRIV_DMA_SZ, pp->cpb, pp->cpb_dma);
+ kfree(pp);
+ ata_port_stop(ap);
+}
+
+
+static void nv_adma_setup_port(struct ata_probe_ent *probe_ent, unsigned int port)
+{
+ void __iomem *mmio = probe_ent->mmio_base;
+ struct ata_ioports *ioport = &probe_ent->port[port];
+
+ VPRINTK("ENTER\n");
+
+ mmio += NV_ADMA_PORT + port * NV_ADMA_PORT_SIZE;
+
+ ioport->cmd_addr = (unsigned long) mmio;
+ ioport->data_addr = (unsigned long) mmio + (ATA_REG_DATA * 4);
+ ioport->error_addr =
+ ioport->feature_addr = (unsigned long) mmio + (ATA_REG_ERR * 4);
+ ioport->nsect_addr = (unsigned long) mmio + (ATA_REG_NSECT * 4);
+ ioport->lbal_addr = (unsigned long) mmio + (ATA_REG_LBAL * 4);
+ ioport->lbam_addr = (unsigned long) mmio + (ATA_REG_LBAM * 4);
+ ioport->lbah_addr = (unsigned long) mmio + (ATA_REG_LBAH * 4);
+ ioport->device_addr = (unsigned long) mmio + (ATA_REG_DEVICE * 4);
+ ioport->status_addr =
+ ioport->command_addr = (unsigned long) mmio + (ATA_REG_STATUS * 4);
+ ioport->altstatus_addr =
+ ioport->ctl_addr = (unsigned long) mmio + 0x20;
+}
+
+static int nv_adma_host_init(struct ata_probe_ent *probe_ent)
+{
+ struct pci_dev *pdev = to_pci_dev(probe_ent->dev);
+ unsigned int i;
+ u32 tmp32;
+
+ VPRINTK("ENTER\n");
+
+ // enable ADMA on the ports
+ pci_read_config_dword(pdev, NV_MCP_SATA_CFG_20, &tmp32);
+ tmp32 |= NV_MCP_SATA_CFG_20_PORT0_EN |
+ NV_MCP_SATA_CFG_20_PORT0_PWB_EN |
+ NV_MCP_SATA_CFG_20_PORT1_EN |
+ NV_MCP_SATA_CFG_20_PORT1_PWB_EN;
+
+ pci_write_config_dword(pdev, NV_MCP_SATA_CFG_20, tmp32);
+
+ for (i = 0; i < probe_ent->n_ports; i++)
+ nv_adma_setup_port(probe_ent, i);
+
+ for (i = 0; i < probe_ent->n_ports; i++) {
+ void __iomem *mmio = __nv_adma_ctl_block(probe_ent->mmio_base, i);
+ u16 tmp;
+
+ /* enable interrupt, clear reset if not already clear */
+ tmp = readw(mmio + NV_ADMA_CTL);
+ writew(tmp | NV_ADMA_CTL_AIEN, mmio + NV_ADMA_CTL);
+ }
+
+ return 0;
+}
+
+
+static void nv_adma_fill_aprd(struct ata_queued_cmd *qc,
+ struct scatterlist *sg,
+ int idx,
+ struct nv_adma_prd *aprd)
+{
+ u32 flags;
+
+ memset(aprd, 0, sizeof(struct nv_adma_prd));
+
+ flags = 0;
+ if (qc->tf.flags & ATA_TFLAG_WRITE)
+ flags |= NV_APRD_WRITE;
+ if (idx == qc->n_elem - 1)
+ flags |= NV_APRD_END;
+ else if (idx != 4)
+ flags |= NV_APRD_CONT;
+
+ aprd->addr = cpu_to_le64(((u64)sg_dma_address(sg)));
+ aprd->len = cpu_to_le32(((u32)sg_dma_len(sg))); /* len in bytes */
+ aprd->flags = cpu_to_le32(flags);
+}
+
+static void nv_adma_fill_sg(struct ata_queued_cmd *qc, struct nv_adma_cpb *cpb)
+{
+ struct nv_adma_port_priv *pp = qc->ap->private_data;
+ unsigned int idx;
+ struct nv_adma_prd *aprd;
+ struct scatterlist *sg;
+
+ VPRINTK("ENTER\n");
+
+ idx = 0;
+
+ ata_for_each_sg(sg, qc) {
+ aprd = (idx < 5) ? &cpb->aprd[idx] : &pp->aprd[NV_ADMA_SGTBL_LEN * qc->tag + (idx-5)];
+ nv_adma_fill_aprd(qc, sg, idx, aprd);
+ idx++;
+ }
+ if (idx > 5)
+ cpb->next_aprd = cpu_to_le64(((u64)(pp->aprd_dma + NV_ADMA_SGTBL_SZ * qc->tag)));
+}
+
+static void nv_adma_qc_prep(struct ata_queued_cmd *qc)
+{
+ struct nv_adma_port_priv *pp = qc->ap->private_data;
+ struct nv_adma_cpb *cpb = &pp->cpb[qc->tag];
+ u8 ctl_flags = NV_CPB_CTL_CPB_VALID |
+ NV_CPB_CTL_APRD_VALID |
+ NV_CPB_CTL_IEN;
+
+ VPRINTK("ENTER\n");
+
+ VPRINTK("qc->flags = 0x%x\n", (u32)qc->flags);
+
+ if (!(qc->flags & ATA_QCFLAG_DMAMAP)) {
+ ata_qc_prep(qc);
+ return;
+ }
+
+ memset(cpb, 0, sizeof(struct nv_adma_cpb));
+
+ cpb->len = 3;
+ cpb->tag = qc->tag;
+ cpb->next_cpb_idx = 0;
+
+ // turn on NCQ flags for NCQ commands
+ if (qc->tf.protocol == ATA_PROT_NCQ)
+ ctl_flags |= NV_CPB_CTL_QUEUE | NV_CPB_CTL_FPDMA;
+
+ nv_adma_tf_to_cpb(&qc->tf, cpb->tf);
+
+ nv_adma_fill_sg(qc, cpb);
+
+ /* Be paranoid and don't let the device see NV_CPB_CTL_CPB_VALID until we are
+ finished filling in all of the contents */
+ wmb();
+ cpb->ctl_flags = ctl_flags;
+}
+
+static unsigned int nv_adma_qc_issue(struct ata_queued_cmd *qc)
+{
+ void __iomem *mmio = nv_adma_ctl_block(qc->ap);
+
+ VPRINTK("ENTER\n");
+
+ if (!(qc->flags & ATA_QCFLAG_DMAMAP)) {
+ VPRINTK("no dmamap, using ATA register mode: 0x%x\n", (u32)qc->flags);
+ // use ATA register mode
+ nv_adma_register_mode(qc->ap);
+ return ata_qc_issue_prot(qc);
+ } else
+ nv_adma_mode(qc->ap);
+
+ //
+ // write append register, command tag in lower 8 bits
+ // and (number of cpbs to append -1) in top 8 bits
+ //
+ wmb();
+ writew(qc->tag, mmio + NV_ADMA_APPEND);
+
+ DPRINTK("Issued tag %u\n",qc->tag);
+
+ return 0;
+}
+
static irqreturn_t nv_generic_interrupt(int irq, void *dev_instance,
struct pt_regs *regs)
{
@@ -468,6 +1243,56 @@ static void nv_error_handler(struct ata_
nv_hardreset, ata_std_postreset);
}
+static void nv_adma_error_handler(struct ata_port *ap)
+{
+ struct nv_adma_port_priv *pp = ap->private_data;
+ if(!(pp->flags & NV_ADMA_PORT_REGISTER_MODE)) {
+ void __iomem *mmio = nv_adma_ctl_block(ap);
+ int i;
+ u16 tmp;
+
+ u32 notifier = readl(mmio + NV_ADMA_NOTIFIER);
+ u32 notifier_error = readl(mmio + NV_ADMA_NOTIFIER_ERROR);
+ u32 gen_ctl = readl(nv_adma_gen_block(ap) + NV_ADMA_GEN_CTL);
+ u32 status = readw(mmio + NV_ADMA_STAT);
+
+ dev_printk(KERN_ERR, ap->dev, "EH port %u in ADMA, notifier 0x%X "
+ "notifier_error 0x%X gen_ctl 0x%X status 0x%X\n",ap->port_no,
+ notifier, notifier_error, gen_ctl, status);
+
+ for( i=0;i<NV_ADMA_MAX_CPBS;i++) {
+ struct nv_adma_cpb *cpb = &pp->cpb[i];
+ if( cpb->ctl_flags || cpb->resp_flags )
+ dev_printk(KERN_ERR, ap->dev,
+ "CPB %d: ctl_flags 0x%x, resp_flags 0x%x\n",
+ i, cpb->ctl_flags, cpb->resp_flags);
+ }
+
+ /* Push us back into port register mode for error handling. */
+ nv_adma_register_mode(ap);
+
+ dev_printk(KERN_NOTICE, ap->dev, "Resetting port %u\n", ap->port_no);
+
+ /* Mark all of the CPBs as invalid to prevent them from being executed */
+ for( i=0;i<NV_ADMA_MAX_CPBS;i++)
+ pp->cpb[i].ctl_flags &= ~NV_CPB_CTL_CPB_VALID;
+
+ /* clear CPB fetch count */
+ writew(0, mmio + NV_ADMA_CPB_COUNT);
+
+ /* Reset channel */
+ tmp = readw(mmio + NV_ADMA_CTL);
+ writew(tmp | NV_ADMA_CTL_CHANNEL_RESET, mmio + NV_ADMA_CTL);
+ readl( mmio + NV_ADMA_CTL ); /* flush posted write */
+ udelay(1);
+ writew(tmp & ~NV_ADMA_CTL_CHANNEL_RESET, mmio + NV_ADMA_CTL);
+ readl( mmio + NV_ADMA_CTL ); /* flush posted write */
+ }
+
+ ata_bmdma_drive_eh(ap, ata_std_prereset, ata_std_softreset,
+ nv_hardreset, ata_std_postreset);
+}
+
static int nv_init_one (struct pci_dev *pdev, const struct pci_device_id *ent)
{
static int printed_version = 0;
@@ -477,6 +1302,8 @@ static int nv_init_one (struct pci_dev *
int rc;
u32 bar;
unsigned long base;
+ unsigned long type = ent->driver_data;
+ u64 dma_mask = ATA_DMA_MASK;
// Make sure this is a SATA controller by counting the number of bars
// (NVIDIA SATA controllers will always have six bars). Otherwise,
@@ -487,6 +1314,12 @@ static int nv_init_one (struct pci_dev *
if (!printed_version++)
dev_printk(KERN_DEBUG, &pdev->dev, "version " DRV_VERSION "\n");
+
+ if(type >= CK804 && adma_enabled) {
+ dev_printk(KERN_NOTICE, &pdev->dev, "Using ADMA mode\n");
+ type = ADMA;
+ dma_mask = DMA_64BIT_MASK;
+ }
rc = pci_enable_device(pdev);
if (rc)
@@ -498,16 +1331,16 @@ static int nv_init_one (struct pci_dev *
goto err_out_disable;
}
- rc = pci_set_dma_mask(pdev, ATA_DMA_MASK);
+ rc = pci_set_dma_mask(pdev, dma_mask);
if (rc)
goto err_out_regions;
- rc = pci_set_consistent_dma_mask(pdev, ATA_DMA_MASK);
+ rc = pci_set_consistent_dma_mask(pdev, dma_mask);
if (rc)
goto err_out_regions;
rc = -ENOMEM;
- ppi[0] = ppi[1] = &nv_port_info[ent->driver_data];
+ ppi[0] = ppi[1] = &nv_port_info[type];
probe_ent = ata_pci_init_native_mode(pdev, ppi, ATA_PORT_PRIMARY | ATA_PORT_SECONDARY);
if (!probe_ent)
goto err_out_regions;
@@ -524,7 +1357,7 @@ static int nv_init_one (struct pci_dev *
probe_ent->port[1].scr_addr = base + NV_PORT1_SCR_REG_OFFSET;
/* enable SATA space for CK804 */
- if (ent->driver_data == CK804) {
+ if (type >= CK804) {
u8 regval;
pci_read_config_byte(pdev, NV_MCP_SATA_CFG_20, ®val);
@@ -534,6 +1367,12 @@ static int nv_init_one (struct pci_dev *
pci_set_master(pdev);
+ if (type == ADMA) {
+ rc = nv_adma_host_init(probe_ent);
+ if (rc)
+ goto err_out_iounmap;
+ }
+
rc = ata_device_add(probe_ent);
if (rc != NV_PORTS)
goto err_out_iounmap;
@@ -568,6 +1407,33 @@ static void nv_ck804_host_stop(struct at
ata_pci_host_stop(host);
}
+static void nv_adma_host_stop(struct ata_host *host)
+{
+ struct pci_dev *pdev = to_pci_dev(host->dev);
+ int i;
+ u32 tmp32;
+
+ for (i = 0; i < host->n_ports; i++) {
+ void __iomem *mmio = __nv_adma_ctl_block(host->mmio_base, i);
+ u16 tmp;
+
+ /* disable interrupt */
+ tmp = readw(mmio + NV_ADMA_CTL);
+ writew(tmp & ~NV_ADMA_CTL_AIEN, mmio + NV_ADMA_CTL);
+ }
+
+ // disable ADMA on the ports
+ pci_read_config_dword(pdev, NV_MCP_SATA_CFG_20, &tmp32);
+ tmp32 &= ~(NV_MCP_SATA_CFG_20_PORT0_EN |
+ NV_MCP_SATA_CFG_20_PORT0_PWB_EN |
+ NV_MCP_SATA_CFG_20_PORT1_EN |
+ NV_MCP_SATA_CFG_20_PORT1_PWB_EN);
+
+ pci_write_config_dword(pdev, NV_MCP_SATA_CFG_20, tmp32);
+
+ nv_ck804_host_stop(host);
+}
+
static int __init nv_init(void)
{
return pci_register_driver(&nv_pci_driver);
@@ -580,3 +1446,5 @@ static void __exit nv_exit(void)
module_init(nv_init);
module_exit(nv_exit);
+module_param_named(adma, adma_enabled, bool, 0444);
+MODULE_PARM_DESC(adma, "Enable use of ADMA (Default: false)");
Only in linux-2.6.18-mm3-adma/drivers/ata: sata_nv.c.orig
^ permalink raw reply [flat|nested] 23+ messages in thread
* RE: [RFC PATCH] nForce4 ADMA with NCQ: It's aliiiive..
2006-10-07 15:54 ` Robert Hancock
@ 2006-10-10 6:44 ` Allen Martin
2006-10-10 7:52 ` Prakash Punnoor
2006-11-30 16:46 ` Prakash Punnoor
0 siblings, 2 replies; 23+ messages in thread
From: Allen Martin @ 2006-10-10 6:44 UTC (permalink / raw)
To: Robert Hancock, Prakash Punnoor; +Cc: linux-kernel, linux-ide, jeff
> > Unfortunately it doesn't work for me on MCP51 if I change
> GENERIC to
> > ADMA. So I wonder whether MCP51 has ADMA mode or what needs
> to be done
> > to get NCQ working. :-(
>
> What happened when you tried it? It would be useful if you
> could change the #undef in these lines:
>
> 53 #undef ATA_DEBUG /* debugging output */
> 54 #undef ATA_VERBOSE_DEBUG /* yet more debugging output */
>
> in include/linux/libata.h to #define and rebuild and try it
> then, that will spew out a bunch more output and I can see if
> any reasonable looking values are showing up at all. I was
> capturing this output the crude way, booting with vga=6 to
> get a smaller font and taking a picture of the screen :-)
> Also, maybe post the lspci -v output from the SATA controller..
>
> If that doesn't provide any insight, maybe the docs Jeff has
> provide the answer for whether or not the MCP5x/MCP61
> controllers have the same interface as the CK804/MCP04..
No, only CK804 and MCP04 support ADMA. We'll be publishing some patches
for NCQ support for MCP55/MCP61 soon.
-----------------------------------------------------------------------------------
This email message is for the sole use of the intended recipient(s) and may contain
confidential information. Any unauthorized review, use, disclosure or distribution
is prohibited. If you are not the intended recipient, please contact the sender by
reply email and destroy all copies of the original message.
-----------------------------------------------------------------------------------
^ permalink raw reply [flat|nested] 23+ messages in thread
* RE: [RFC PATCH] nForce4 ADMA with NCQ: It's aliiiive..
2006-10-07 15:28 ` Jeff Garzik
2006-10-07 23:49 ` Robert Hancock
@ 2006-10-10 6:52 ` Allen Martin
2006-10-11 5:07 ` Robert Hancock
1 sibling, 1 reply; 23+ messages in thread
From: Allen Martin @ 2006-10-10 6:52 UTC (permalink / raw)
To: Jeff Garzik, Robert Hancock; +Cc: linux-kernel, linux-ide, prakash
> > -Jeff will probably cringe at how I implemented the
> > bmdma_stop/start/status/setup functions. This kludge of toggling
> > ATA_FLAG_MMIO off for the call into libata was needed since this
> > controller is almost what libata calls ATA_FLAG_MMIO, but not quite
> > (the ATA taskfile registers are MMIO but the BMDMA
> registers are PIO).
> > This is also why I needed the patch to libata-sff.c to use the
> > adapter's bmdma_status function rather than hardcoded
> ata_bmdma_status.
>
> *shrug* I don't cringe if that's the most expedient way to
> do something.
>
> But I really don't think that is necessary. I will take a
> look at docs and see how things match up, when I am much more
> awake. Most likely you need to be using another set of
> registers, and be all MMIO, all the time.
You shouldn't be touching BM registers when ADMA is enabled, it can
cause bad things to happen.
You should be using BM registers when doing ATAPI protocol though, as it
doesn't work through ADMA. So I wouldn't say you should be using MMIO
all the time.
-Allen
-----------------------------------------------------------------------------------
This email message is for the sole use of the intended recipient(s) and may contain
confidential information. Any unauthorized review, use, disclosure or distribution
is prohibited. If you are not the intended recipient, please contact the sender by
reply email and destroy all copies of the original message.
-----------------------------------------------------------------------------------
^ permalink raw reply [flat|nested] 23+ messages in thread
* Re: [RFC PATCH] nForce4 ADMA with NCQ: It's aliiiive..
2006-10-10 6:44 ` Allen Martin
@ 2006-10-10 7:52 ` Prakash Punnoor
2006-11-30 16:46 ` Prakash Punnoor
1 sibling, 0 replies; 23+ messages in thread
From: Prakash Punnoor @ 2006-10-10 7:52 UTC (permalink / raw)
To: Allen Martin; +Cc: Robert Hancock, linux-kernel, linux-ide, jeff
[-- Attachment #1: Type: text/plain, Size: 764 bytes --]
Am Dienstag 10 Oktober 2006 08:44 schrieb Allen Martin:
> > > Unfortunately it doesn't work for me on MCP51 if I change
> >
> > GENERIC to
> >
> > > ADMA. So I wonder whether MCP51 has ADMA mode or what needs
> >
> > to be done
> >
> > > to get NCQ working. :-(
> >
> > If that doesn't provide any insight, maybe the docs Jeff has
> > provide the answer for whether or not the MCP5x/MCP61
> > controllers have the same interface as the CK804/MCP04..
>
> No, only CK804 and MCP04 support ADMA. We'll be publishing some patches
> for NCQ support for MCP55/MCP61 soon.
That's great news! I hope you don't forget the MCP51, as well, as you didn't
mention it. :-)
--
(°= =°)
//\ Prakash Punnoor /\\
V_/ \_V
[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]
^ permalink raw reply [flat|nested] 23+ messages in thread
* Re: [RFC PATCH] nForce4 ADMA with NCQ: It's aliiiive..
2006-10-10 6:52 ` Allen Martin
@ 2006-10-11 5:07 ` Robert Hancock
2006-10-11 10:30 ` Jens Axboe
0 siblings, 1 reply; 23+ messages in thread
From: Robert Hancock @ 2006-10-11 5:07 UTC (permalink / raw)
To: Allen Martin; +Cc: Jeff Garzik, linux-kernel, linux-ide, prakash
[-- Attachment #1: Type: text/plain, Size: 1861 bytes --]
Allen Martin wrote:
>> But I really don't think that is necessary. I will take a
>> look at docs and see how things match up, when I am much more
>> awake. Most likely you need to be using another set of
>> registers, and be all MMIO, all the time.
>
> You shouldn't be touching BM registers when ADMA is enabled, it can
> cause bad things to happen.
>
> You should be using BM registers when doing ATAPI protocol though, as it
> doesn't work through ADMA. So I wouldn't say you should be using MMIO
> all the time.
>
> -Allen
OK, I've updated the code to take this into account, an updated patch is
attached. However, this does raise an issue. If we have to fall back to
legacy mode to do ATAPI DMA, this means that we can't do 64-bit DMA for
such transfers. Since by the time the driver gets a request the SGs have
already been created based on the set DMA mask, the only way I can see
to handle this is to either allow ATAPI DMA or 64-bit DMA, not both.
I've chosen to default to 64-bit DMA in this version, but there is a
module parameter which allows overriding this if you care more about
using ATAPI devices than efficiency with over 4GB of RAM. I'm open to
suggestions on a better way to handle this..
I also updated the comments to clarify which chipsets are supported by
ADMA as per Allen's previous comment.
This patch is now against 2.6.19-rc1-mm1. If there are no more
suggestions, are there any objections to merging this into libata-dev or
the -mm tree? I think it is ready for some more exposure, and since ADMA
is currently disabled by default there should be little risk of breakage
for those who don't choose to enable it.
Signed-off-by: Robert Hancock <hancockr@shaw.ca>
--
Robert Hancock Saskatoon, SK, Canada
To email, remove "nospam" from hancockr@nospamshaw.ca
Home Page: http://www.roberthancock.com/
[-- Attachment #2: sata_nv-adma-ncq-v3.patch --]
[-- Type: text/plain, Size: 31580 bytes --]
diff -rup --exclude='*.o' --exclude='*mod.c' --exclude='*.cmd' --exclude='*.ko' linux-2.6.19-rc1-mm1/drivers/ata/libata-sff.c linux-2.6.19-rc1-mm1-adma/drivers/ata/libata-sff.c
--- linux-2.6.19-rc1-mm1/drivers/ata/libata-sff.c 2006-10-10 20:50:42.000000000 -0600
+++ linux-2.6.19-rc1-mm1-adma/drivers/ata/libata-sff.c 2006-10-10 20:54:58.000000000 -0600
@@ -732,7 +732,7 @@ void ata_bmdma_drive_eh(struct ata_port
qc->tf.protocol == ATA_PROT_ATAPI_DMA)) {
u8 host_stat;
- host_stat = ata_bmdma_status(ap);
+ host_stat = ap->ops->bmdma_status(ap);
ata_ehi_push_desc(&ehc->i, "BMDMA stat 0x%x", host_stat);
diff -rup --exclude='*.o' --exclude='*mod.c' --exclude='*.cmd' --exclude='*.ko' linux-2.6.19-rc1-mm1/drivers/ata/sata_nv.c linux-2.6.19-rc1-mm1-adma/drivers/ata/sata_nv.c
--- linux-2.6.19-rc1-mm1/drivers/ata/sata_nv.c 2006-10-10 20:51:10.000000000 -0600
+++ linux-2.6.19-rc1-mm1-adma/drivers/ata/sata_nv.c 2006-10-10 22:00:10.000000000 -0600
@@ -29,6 +29,11 @@
* NV-specific details such as register offsets, SATA phy location,
* hotplug info, etc.
*
+ * CK804/MCP04 controllers support an alternate programming interface
+ * similar to the ADMA specification (with some modifications).
+ * This allows the use of NCQ. Non-DMA-mapped ATA commands are still
+ * sent through the legacy interface.
+ *
*/
#include <linux/kernel.h>
@@ -43,7 +48,7 @@
#include <linux/libata.h>
#define DRV_NAME "sata_nv"
-#define DRV_VERSION "2.0"
+#define DRV_VERSION "3.0"
enum {
NV_PORTS = 2,
@@ -78,8 +83,135 @@ enum {
// For PCI config register 20
NV_MCP_SATA_CFG_20 = 0x50,
NV_MCP_SATA_CFG_20_SATA_SPACE_EN = 0x04,
+ NV_MCP_SATA_CFG_20_PORT0_EN = (1 << 17),
+ NV_MCP_SATA_CFG_20_PORT1_EN = (1 << 16),
+ NV_MCP_SATA_CFG_20_PORT0_PWB_EN = (1 << 14),
+ NV_MCP_SATA_CFG_20_PORT1_PWB_EN = (1 << 12),
+
+ NV_ADMA_MAX_CPBS = 32,
+ NV_ADMA_CPB_SZ = 128,
+ NV_ADMA_APRD_SZ = 16,
+ NV_ADMA_SGTBL_LEN = (1024 - NV_ADMA_CPB_SZ) / NV_ADMA_APRD_SZ,
+ NV_ADMA_SGTBL_SZ = NV_ADMA_SGTBL_LEN * NV_ADMA_APRD_SZ,
+ NV_ADMA_PORT_PRIV_DMA_SZ = NV_ADMA_MAX_CPBS * (NV_ADMA_CPB_SZ + NV_ADMA_SGTBL_SZ),
+
+ // BAR5 offset to ADMA general registers
+ NV_ADMA_GEN = 0x400,
+ NV_ADMA_GEN_CTL = 0x00,
+ NV_ADMA_NOTIFIER_CLEAR = 0x30,
+
+ // BAR5 offset to ADMA ports
+ NV_ADMA_PORT = 0x480,
+
+ // size of ADMA port register space
+ NV_ADMA_PORT_SIZE = 0x100,
+
+ // ADMA port registers
+ NV_ADMA_CTL = 0x40,
+ NV_ADMA_CPB_COUNT = 0x42,
+ NV_ADMA_NEXT_CPB_IDX = 0x43,
+ NV_ADMA_STAT = 0x44,
+ NV_ADMA_CPB_BASE_LOW = 0x48,
+ NV_ADMA_CPB_BASE_HIGH = 0x4C,
+ NV_ADMA_APPEND = 0x50,
+ NV_ADMA_NOTIFIER = 0x68,
+ NV_ADMA_NOTIFIER_ERROR = 0x6C,
+
+ // NV_ADMA_CTL register bits
+ NV_ADMA_CTL_HOTPLUG_IEN = (1 << 0),
+ NV_ADMA_CTL_CHANNEL_RESET = (1 << 5),
+ NV_ADMA_CTL_GO = (1 << 7),
+ NV_ADMA_CTL_AIEN = (1 << 8),
+ NV_ADMA_CTL_READ_NON_COHERENT = (1 << 11),
+ NV_ADMA_CTL_WRITE_NON_COHERENT = (1 << 12),
+
+ // CPB response flag bits
+ NV_CPB_RESP_DONE = (1 << 0),
+ NV_CPB_RESP_ATA_ERR = (1 << 3),
+ NV_CPB_RESP_CMD_ERR = (1 << 4),
+ NV_CPB_RESP_CPB_ERR = (1 << 7),
+
+ // CPB control flag bits
+ NV_CPB_CTL_CPB_VALID = (1 << 0),
+ NV_CPB_CTL_QUEUE = (1 << 1),
+ NV_CPB_CTL_APRD_VALID = (1 << 2),
+ NV_CPB_CTL_IEN = (1 << 3),
+ NV_CPB_CTL_FPDMA = (1 << 4),
+
+ // APRD flags
+ NV_APRD_WRITE = (1 << 1),
+ NV_APRD_END = (1 << 2),
+ NV_APRD_CONT = (1 << 3),
+
+ // NV_ADMA_STAT flags
+ NV_ADMA_STAT_TIMEOUT = (1 << 0),
+ NV_ADMA_STAT_HOTUNPLUG = (1 << 1),
+ NV_ADMA_STAT_HOTPLUG = (1 << 2),
+ NV_ADMA_STAT_CPBERR = (1 << 4),
+ NV_ADMA_STAT_SERROR = (1 << 5),
+ NV_ADMA_STAT_CMD_COMPLETE = (1 << 6),
+ NV_ADMA_STAT_IDLE = (1 << 8),
+ NV_ADMA_STAT_LEGACY = (1 << 9),
+ NV_ADMA_STAT_STOPPED = (1 << 10),
+ NV_ADMA_STAT_DONE = (1 << 12),
+ NV_ADMA_STAT_ERR = (NV_ADMA_STAT_CPBERR | NV_ADMA_STAT_TIMEOUT),
+
+ // port flags
+ NV_ADMA_PORT_REGISTER_MODE = (1 << 0),
+
+};
+
+/* ADMA Physical Region Descriptor - one SG segment */
+struct nv_adma_prd {
+ __le64 addr;
+ __le32 len;
+ u8 flags;
+ u8 packet_len;
+ __le16 reserved;
+};
+
+enum nv_adma_regbits {
+ CMDEND = (1 << 15), /* end of command list */
+ WNB = (1 << 14), /* wait-not-BSY */
+ IGN = (1 << 13), /* ignore this entry */
+ CS1n = (1 << (4 + 8)), /* std. PATA signals follow... */
+ DA2 = (1 << (2 + 8)),
+ DA1 = (1 << (1 + 8)),
+ DA0 = (1 << (0 + 8)),
+};
+
+/* ADMA Command Parameter Block
+ The first 5 SG segments are stored inside the Command Parameter Block itself.
+ If there are more than 5 segments the remainder are stored in a separate
+ memory area indicated by next_aprd. */
+struct nv_adma_cpb {
+ u8 resp_flags; //0
+ u8 reserved1; //1
+ u8 ctl_flags; //2
+ // len is length of taskfile in 64 bit words
+ u8 len; //3
+ u8 tag; //4
+ u8 next_cpb_idx; //5
+ __le16 reserved2; //6-7
+ __le16 tf[12]; //8-31
+ struct nv_adma_prd aprd[5]; //32-111
+ __le64 next_aprd; //112-119
+ __le64 reserved3; //120-127
+};
+
+
+struct nv_adma_port_priv {
+ struct nv_adma_cpb *cpb;
+ dma_addr_t cpb_dma;
+ struct nv_adma_prd *aprd;
+ dma_addr_t aprd_dma;
+ u8 flags;
+ u32 notifier;
+ u32 notifier_error;
};
+#define NV_ADMA_CHECK_INTR(GCTL, PORT) ((GCTL) & ( 1 << (19 + (12 * (PORT)))))
+
static int nv_init_one (struct pci_dev *pdev, const struct pci_device_id *ent);
static void nv_ck804_host_stop(struct ata_host *host);
static irqreturn_t nv_generic_interrupt(int irq, void *dev_instance);
@@ -93,13 +225,27 @@ static void nv_nf2_thaw(struct ata_port
static void nv_ck804_freeze(struct ata_port *ap);
static void nv_ck804_thaw(struct ata_port *ap);
static void nv_error_handler(struct ata_port *ap);
+static void nv_adma_qc_prep(struct ata_queued_cmd *qc);
+static unsigned int nv_adma_qc_issue(struct ata_queued_cmd *qc);
+static irqreturn_t nv_adma_interrupt(int irq, void *dev_instance);
+static void nv_adma_irq_clear(struct ata_port *ap);
+static int nv_adma_port_start(struct ata_port *ap);
+static void nv_adma_port_stop(struct ata_port *ap);
+static void nv_adma_error_handler(struct ata_port *ap);
+static void nv_adma_host_stop(struct ata_host *host);
+static void nv_adma_bmdma_setup(struct ata_queued_cmd *qc);
+static void nv_adma_bmdma_start(struct ata_queued_cmd *qc);
+static void nv_adma_bmdma_stop(struct ata_queued_cmd *qc);
+static u8 nv_adma_bmdma_status(struct ata_port *ap);
+static int nv_adma_check_atapi_dma(struct ata_queued_cmd *qc);
enum nv_host_type
{
GENERIC,
NFORCE2,
NFORCE3 = NFORCE2, /* NF2 == NF3 as far as sata_nv is concerned */
- CK804
+ CK804,
+ ADMA
};
static const struct pci_device_id nv_pci_tbl[] = {
@@ -156,6 +302,25 @@ static struct scsi_host_template nv_sht
.bios_param = ata_std_bios_param,
};
+static struct scsi_host_template nv_adma_sht = {
+ .module = THIS_MODULE,
+ .name = DRV_NAME,
+ .ioctl = ata_scsi_ioctl,
+ .queuecommand = ata_scsi_queuecmd,
+ .can_queue = NV_ADMA_MAX_CPBS,
+ .this_id = ATA_SHT_THIS_ID,
+ .sg_tablesize = NV_ADMA_SGTBL_LEN + 5,
+ .max_sectors = ATA_MAX_SECTORS,
+ .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 const struct ata_port_operations nv_generic_ops = {
.port_disable = ata_port_disable,
.tf_load = ata_tf_load,
@@ -237,6 +402,34 @@ static const struct ata_port_operations
.host_stop = nv_ck804_host_stop,
};
+static const struct ata_port_operations nv_adma_ops = {
+ .port_disable = ata_port_disable,
+ .tf_load = ata_tf_load,
+ .tf_read = ata_tf_read,
+ .exec_command = ata_exec_command,
+ .check_atapi_dma = nv_adma_check_atapi_dma,
+ .check_status = ata_check_status,
+ .dev_select = ata_std_dev_select,
+ .bmdma_setup = nv_adma_bmdma_setup,
+ .bmdma_start = nv_adma_bmdma_start,
+ .bmdma_stop = nv_adma_bmdma_stop,
+ .bmdma_status = nv_adma_bmdma_status,
+ .qc_prep = nv_adma_qc_prep,
+ .qc_issue = nv_adma_qc_issue,
+ .freeze = nv_ck804_freeze,
+ .thaw = nv_ck804_thaw,
+ .error_handler = nv_adma_error_handler,
+ .post_internal_cmd = nv_adma_bmdma_stop,
+ .data_xfer = ata_mmio_data_xfer,
+ .irq_handler = nv_adma_interrupt,
+ .irq_clear = nv_adma_irq_clear,
+ .scr_read = nv_scr_read,
+ .scr_write = nv_scr_write,
+ .port_start = nv_adma_port_start,
+ .port_stop = nv_adma_port_stop,
+ .host_stop = nv_adma_host_stop,
+};
+
static struct ata_port_info nv_port_info[] = {
/* generic */
{
@@ -265,6 +458,16 @@ static struct ata_port_info nv_port_info
.udma_mask = NV_UDMA_MASK,
.port_ops = &nv_ck804_ops,
},
+ /* ADMA */
+ {
+ .sht = &nv_adma_sht,
+ .flags = ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY |
+ ATA_FLAG_MMIO | ATA_FLAG_NCQ,
+ .pio_mask = NV_PIO_MASK,
+ .mwdma_mask = NV_MWDMA_MASK,
+ .udma_mask = NV_UDMA_MASK,
+ .port_ops = &nv_adma_ops,
+ },
};
MODULE_AUTHOR("NVIDIA");
@@ -273,6 +476,593 @@ MODULE_LICENSE("GPL");
MODULE_DEVICE_TABLE(pci, nv_pci_tbl);
MODULE_VERSION(DRV_VERSION);
+static int adma_enabled = 1;
+static int adma_enable_atapi_dma;
+
+static unsigned int nv_adma_tf_to_cpb(struct ata_taskfile *tf, u16 *cpb)
+{
+ unsigned int idx = 0;
+
+ cpb[idx++] = cpu_to_le16((ATA_REG_DEVICE << 8) | tf->device | WNB);
+
+ if ((tf->flags & ATA_TFLAG_LBA48) == 0) {
+ cpb[idx++] = cpu_to_le16(IGN);
+ cpb[idx++] = cpu_to_le16(IGN);
+ cpb[idx++] = cpu_to_le16(IGN);
+ cpb[idx++] = cpu_to_le16(IGN);
+ cpb[idx++] = cpu_to_le16(IGN);
+ }
+ else {
+ cpb[idx++] = cpu_to_le16((ATA_REG_ERR << 8) | tf->hob_feature);
+ cpb[idx++] = cpu_to_le16((ATA_REG_NSECT << 8) | tf->hob_nsect);
+ cpb[idx++] = cpu_to_le16((ATA_REG_LBAL << 8) | tf->hob_lbal);
+ cpb[idx++] = cpu_to_le16((ATA_REG_LBAM << 8) | tf->hob_lbam);
+ cpb[idx++] = cpu_to_le16((ATA_REG_LBAH << 8) | tf->hob_lbah);
+ }
+ cpb[idx++] = cpu_to_le16((ATA_REG_ERR << 8) | tf->feature);
+ cpb[idx++] = cpu_to_le16((ATA_REG_NSECT << 8) | tf->nsect);
+ cpb[idx++] = cpu_to_le16((ATA_REG_LBAL << 8) | tf->lbal);
+ cpb[idx++] = cpu_to_le16((ATA_REG_LBAM << 8) | tf->lbam);
+ cpb[idx++] = cpu_to_le16((ATA_REG_LBAH << 8) | tf->lbah);
+
+ cpb[idx++] = cpu_to_le16((ATA_REG_CMD << 8) | tf->command | CMDEND);
+
+ return idx;
+}
+
+static inline void __iomem *__nv_adma_ctl_block(void __iomem *mmio,
+ unsigned int port_no)
+{
+ mmio += NV_ADMA_PORT + port_no * NV_ADMA_PORT_SIZE;
+ return mmio;
+}
+
+static inline void __iomem *nv_adma_ctl_block(struct ata_port *ap)
+{
+ return __nv_adma_ctl_block(ap->host->mmio_base, ap->port_no);
+}
+
+static inline void __iomem *nv_adma_gen_block(struct ata_port *ap)
+{
+ return (ap->host->mmio_base + NV_ADMA_GEN);
+}
+
+static inline void __iomem *nv_adma_notifier_clear_block(struct ata_port *ap)
+{
+ return (nv_adma_gen_block(ap) + NV_ADMA_NOTIFIER_CLEAR + (4 * ap->port_no));
+}
+
+static int nv_adma_host_intr(struct ata_port *ap)
+{
+ void __iomem *mmio = nv_adma_ctl_block(ap);
+ struct nv_adma_port_priv *pp = ap->private_data;
+ u16 status;
+ u32 gen_ctl;
+ u16 flags;
+ int have_global_err = 0;
+
+ /* if in ATA register mode, use standard ata interrupt handler */
+ if (pp->flags & NV_ADMA_PORT_REGISTER_MODE) {
+ struct ata_queued_cmd *qc;
+ VPRINTK("in ATA register mode\n");
+ qc = ata_qc_from_tag(ap, ap->active_tag);
+ if (qc && (!(qc->tf.flags & ATA_TFLAG_POLLING)))
+ return ata_host_intr(ap, qc);
+ else {
+ // No request pending? Clear interrupt status
+ // anyway, in case there's one pending.
+ ap->ops->check_status(ap);
+ return 1;
+ }
+ }
+
+ gen_ctl = readl(nv_adma_gen_block(ap) + NV_ADMA_GEN_CTL);
+
+ if( !NV_ADMA_CHECK_INTR(gen_ctl, ap->port_no) && !pp->notifier && !pp->notifier_error)
+ /* Nothing to do */
+ return 0;
+
+ status = readw(mmio + NV_ADMA_STAT);
+
+ /* Clear status so that any CPB completions after this point in the handler will
+ raise another interrupt */
+ writew(status, mmio + NV_ADMA_STAT);
+ readl(mmio + NV_ADMA_STAT); /* flush posted write */
+ mb();
+
+ /* freeze if hotplugged */
+ if (unlikely(status & (NV_ADMA_STAT_HOTPLUG | NV_ADMA_STAT_HOTUNPLUG))) {
+ dev_printk(KERN_NOTICE, ap->dev, "Hotplug event on port %u, freezing\n",
+ ap->port_no);
+ ata_port_freeze(ap);
+ return 1;
+ }
+
+ if (pp->notifier_error) {
+ dev_printk(KERN_ERR, ap->dev, "port %u notifier error, stat=0x%x, "
+ "notifier=0x%X, notifier_error=0x%X", ap->port_no, status, pp->notifier,
+ pp->notifier_error );
+ have_global_err = 1;
+ }
+
+ if (status & NV_ADMA_STAT_TIMEOUT) {
+ dev_printk(KERN_ERR, ap->dev, "port %u timeout, stat=0x%x\n", ap->port_no, status);
+ have_global_err = 1;
+ }
+ if (status & NV_ADMA_STAT_CPBERR) {
+ dev_printk(KERN_ERR, ap->dev, "port %u CPB error, stat=0x%x\n", ap->port_no, status);
+ have_global_err = 1;
+ }
+ if (status & NV_ADMA_STAT_STOPPED) {
+ VPRINTK("ADMA stopped, stat = 0x%x\n", status);
+ }
+ if (status & NV_ADMA_STAT_CMD_COMPLETE) {
+ VPRINTK("ADMA command complete, stat = 0x%x\n", status);
+ }
+ if ((status & NV_ADMA_STAT_DONE) || have_global_err) {
+ /** Check CPBs for completed commands */
+ int i;
+ for( i=0;i<NV_ADMA_MAX_CPBS;i++)
+ if( ap->sactive & (1 << i ) ||
+ ( ata_tag_valid(ap->active_tag) && ap->active_tag == i ) ) {
+ struct nv_adma_cpb *cpb = &pp->cpb[i];
+ int complete = 0, have_err = 0;
+ flags = cpb->resp_flags;
+ VPRINTK("CPB %d, stat=0x%x, flags=0x%x\n", i, status, flags);
+
+ if (flags & NV_CPB_RESP_DONE) {
+ VPRINTK("CPB flags done, flags=0x%x\n", flags);
+ complete = 1;
+ }
+ if (flags & NV_CPB_RESP_ATA_ERR) {
+ dev_printk(KERN_ERR, ap->dev, "port %u CPB flags ATA err, flags=0x%x\n", ap->port_no, flags);
+ have_err = 1;
+ complete = 1;
+ }
+ if (flags & NV_CPB_RESP_CMD_ERR) {
+ dev_printk(KERN_ERR, ap->dev, "port %u CPB flags CMD err, flags=0x%x\n", ap->port_no, flags);
+ have_err = 1;
+ complete = 1;
+ }
+ if (flags & NV_CPB_RESP_CPB_ERR) {
+ dev_printk(KERN_ERR, ap->dev, "port %u CPB flags CPB err, flags=0x%x\n", ap->port_no, flags);
+ have_err = 1;
+ complete = 1;
+ }
+ if(complete || have_global_err)
+ {
+ struct ata_queued_cmd *qc = ata_qc_from_tag(ap, i);
+ if(likely(qc)) {
+ u8 ata_status = 0;
+ /* Only use the ATA port status for non-NCQ commands.
+ For NCQ commands the current status may have nothing to do with
+ the command just completed. */
+ if(qc->tf.protocol != ATA_PROT_NCQ)
+ ata_status = readb(mmio + (ATA_REG_STATUS * 4));
+
+ if(have_err || have_global_err)
+ ata_status |= ATA_ERR;
+
+ qc->err_mask |= ac_err_mask(ata_status);
+ DPRINTK("Completing qc from tag %d with err_mask %u\n",i,qc->err_mask);
+ ata_qc_complete(qc);
+ }
+ }
+ }
+ }
+
+ return 1; /* irq handled if we got here */
+}
+
+static irqreturn_t nv_adma_interrupt(int irq, void *dev_instance)
+{
+ struct ata_host *host = dev_instance;
+ int i, handled = 0;
+
+ spin_lock(&host->lock);
+
+ for (i = 0; i < host->n_ports; i++) {
+ struct ata_port *ap = host->ports[i];
+
+ if (ap && !(ap->flags & ATA_FLAG_DISABLED)) {
+ struct nv_adma_port_priv *pp = ap->private_data;
+ void __iomem *mmio = nv_adma_ctl_block(ap);
+ // read notifiers
+ pp->notifier = readl(mmio + NV_ADMA_NOTIFIER);
+ pp->notifier_error = readl(mmio + NV_ADMA_NOTIFIER_ERROR);
+ handled += nv_adma_host_intr(ap);
+ }
+ }
+
+ /* Clear notifiers. Yes, this does appear to have to be done at the very end,
+ otherwise things grind to a halt quite quickly. */
+ if (handled)
+ for (i = 0; i < host->n_ports; i++) {
+ struct ata_port *ap = host->ports[i];
+ struct nv_adma_port_priv *pp = ap->private_data;
+ writel(pp->notifier | pp->notifier_error,
+ nv_adma_notifier_clear_block(ap));
+ }
+
+ spin_unlock(&host->lock);
+
+ return IRQ_RETVAL(handled);
+}
+
+static void nv_adma_irq_clear(struct ata_port *ap)
+{
+ void __iomem *mmio = nv_adma_ctl_block(ap);
+ u16 status = readw(mmio + NV_ADMA_STAT);
+ u32 notifier = readl(mmio + NV_ADMA_NOTIFIER);
+ u32 notifier_error = readl(mmio + NV_ADMA_NOTIFIER_ERROR);
+
+ /* clear ADMA status */
+ writew(status, mmio + NV_ADMA_STAT);
+ writel(notifier | notifier_error,
+ nv_adma_notifier_clear_block(ap));
+
+ /** clear legacy status */
+ ap->flags &= ~ATA_FLAG_MMIO;
+ ata_bmdma_irq_clear(ap);
+ ap->flags |= ATA_FLAG_MMIO;
+}
+
+static void nv_adma_bmdma_setup(struct ata_queued_cmd *qc)
+{
+ struct nv_adma_port_priv *pp = qc->ap->private_data;
+
+ if(pp->flags & NV_ADMA_PORT_REGISTER_MODE) {
+ WARN_ON(1);
+ return;
+ }
+
+ qc->ap->flags &= ~ATA_FLAG_MMIO;
+ ata_bmdma_setup(qc);
+ qc->ap->flags |= ATA_FLAG_MMIO;
+}
+
+static void nv_adma_bmdma_start(struct ata_queued_cmd *qc)
+{
+ struct nv_adma_port_priv *pp = qc->ap->private_data;
+
+ if(pp->flags & NV_ADMA_PORT_REGISTER_MODE) {
+ WARN_ON(1);
+ return;
+ }
+
+ qc->ap->flags &= ~ATA_FLAG_MMIO;
+ ata_bmdma_start(qc);
+ qc->ap->flags |= ATA_FLAG_MMIO;
+}
+
+static void nv_adma_bmdma_stop(struct ata_queued_cmd *qc)
+{
+ struct nv_adma_port_priv *pp = qc->ap->private_data;
+
+ if(pp->flags & NV_ADMA_PORT_REGISTER_MODE)
+ return;
+
+ qc->ap->flags &= ~ATA_FLAG_MMIO;
+ ata_bmdma_stop(qc);
+ qc->ap->flags |= ATA_FLAG_MMIO;
+}
+
+static u8 nv_adma_bmdma_status(struct ata_port *ap)
+{
+ u8 status;
+ struct nv_adma_port_priv *pp = ap->private_data;
+
+ WARN_ON(pp->flags & NV_ADMA_PORT_REGISTER_MODE);
+
+ ap->flags &= ~ATA_FLAG_MMIO;
+ status = ata_bmdma_status(ap);
+ ap->flags |= ATA_FLAG_MMIO;
+ return status;
+}
+
+static void nv_adma_register_mode(struct ata_port *ap)
+{
+ void __iomem *mmio = nv_adma_ctl_block(ap);
+ struct nv_adma_port_priv *pp = ap->private_data;
+ u16 tmp;
+
+ if (pp->flags & NV_ADMA_PORT_REGISTER_MODE)
+ return;
+
+ tmp = readw(mmio + NV_ADMA_CTL);
+ writew(tmp & ~NV_ADMA_CTL_GO, mmio + NV_ADMA_CTL);
+
+ pp->flags |= NV_ADMA_PORT_REGISTER_MODE;
+}
+
+static void nv_adma_mode(struct ata_port *ap)
+{
+ void __iomem *mmio = nv_adma_ctl_block(ap);
+ struct nv_adma_port_priv *pp = ap->private_data;
+ u16 tmp;
+
+ if (!(pp->flags & NV_ADMA_PORT_REGISTER_MODE))
+ return;
+
+ tmp = readw(mmio + NV_ADMA_CTL);
+ writew(tmp | NV_ADMA_CTL_GO, mmio + NV_ADMA_CTL);
+
+ pp->flags &= ~NV_ADMA_PORT_REGISTER_MODE;
+}
+
+static int nv_adma_port_start(struct ata_port *ap)
+{
+ struct device *dev = ap->host->dev;
+ struct nv_adma_port_priv *pp;
+ int rc;
+ void *mem;
+ dma_addr_t mem_dma;
+ void __iomem *mmio = nv_adma_ctl_block(ap);
+ u16 tmp;
+
+ VPRINTK("ENTER\n");
+
+ rc = ata_port_start(ap);
+ if (rc)
+ return rc;
+
+ pp = kzalloc(sizeof(*pp), GFP_KERNEL);
+ if (!pp) {
+ rc = -ENOMEM;
+ goto err_out;
+ }
+
+ mem = dma_alloc_coherent(dev, NV_ADMA_PORT_PRIV_DMA_SZ,
+ &mem_dma, GFP_KERNEL);
+
+ VPRINTK("dma memory: vaddr = 0x%p, paddr = 0x%p\n", mem, (void*)mem_dma);
+
+ if (!mem) {
+ rc = -ENOMEM;
+ goto err_out_kfree;
+ }
+ memset(mem, 0, NV_ADMA_PORT_PRIV_DMA_SZ);
+
+ /*
+ * First item in chunk of DMA memory:
+ * 128-byte command parameter block (CPB)
+ * one for each command tag
+ */
+ pp->cpb = mem;
+ pp->cpb_dma = mem_dma;
+
+ VPRINTK("cpb = 0x%p, cpb_dma = 0x%p\n", pp->cpb, (void*)pp->cpb_dma);
+
+ writel(mem_dma & 0xFFFFFFFF, mmio + NV_ADMA_CPB_BASE_LOW);
+ writel((mem_dma >> 16 ) >> 16, mmio + NV_ADMA_CPB_BASE_HIGH);
+
+ mem += NV_ADMA_MAX_CPBS * NV_ADMA_CPB_SZ;
+ mem_dma += NV_ADMA_MAX_CPBS * NV_ADMA_CPB_SZ;
+
+ /*
+ * Second item: block of ADMA_SGTBL_LEN s/g entries
+ */
+ pp->aprd = mem;
+ pp->aprd_dma = mem_dma;
+
+ VPRINTK("aprd = 0x%p, aprd_dma = 0x%p\n", pp->aprd, (void*)pp->aprd_dma);
+
+ ap->private_data = pp;
+
+ // clear any outstanding interrupt conditions
+ writew(0xffff, mmio + NV_ADMA_STAT);
+
+ // initialize port variables
+ pp->flags = NV_ADMA_PORT_REGISTER_MODE;
+
+ /* clear CPB fetch count */
+ writew(0, mmio + NV_ADMA_CPB_COUNT);
+
+ /* clear GO for register mode */
+ tmp = readw(mmio + NV_ADMA_CTL);
+ writew(tmp & ~NV_ADMA_CTL_GO, mmio + NV_ADMA_CTL);
+
+ tmp = readw(mmio + NV_ADMA_CTL);
+ writew(tmp | NV_ADMA_CTL_CHANNEL_RESET, mmio + NV_ADMA_CTL);
+ readl( mmio + NV_ADMA_CTL ); /* flush posted write */
+ udelay(1);
+ writew(tmp & ~NV_ADMA_CTL_CHANNEL_RESET, mmio + NV_ADMA_CTL);
+ readl( mmio + NV_ADMA_CTL ); /* flush posted write */
+
+ return 0;
+
+err_out_kfree:
+ kfree(pp);
+err_out:
+ ata_port_stop(ap);
+ return rc;
+}
+
+static void nv_adma_port_stop(struct ata_port *ap)
+{
+ struct device *dev = ap->host->dev;
+ struct nv_adma_port_priv *pp = ap->private_data;
+ void __iomem *mmio = nv_adma_ctl_block(ap);
+
+ VPRINTK("ENTER\n");
+
+ writew(0, mmio + NV_ADMA_CTL);
+
+ ap->private_data = NULL;
+ dma_free_coherent(dev, NV_ADMA_PORT_PRIV_DMA_SZ, pp->cpb, pp->cpb_dma);
+ kfree(pp);
+ ata_port_stop(ap);
+}
+
+
+static void nv_adma_setup_port(struct ata_probe_ent *probe_ent, unsigned int port)
+{
+ void __iomem *mmio = probe_ent->mmio_base;
+ struct ata_ioports *ioport = &probe_ent->port[port];
+
+ VPRINTK("ENTER\n");
+
+ mmio += NV_ADMA_PORT + port * NV_ADMA_PORT_SIZE;
+
+ ioport->cmd_addr = (unsigned long) mmio;
+ ioport->data_addr = (unsigned long) mmio + (ATA_REG_DATA * 4);
+ ioport->error_addr =
+ ioport->feature_addr = (unsigned long) mmio + (ATA_REG_ERR * 4);
+ ioport->nsect_addr = (unsigned long) mmio + (ATA_REG_NSECT * 4);
+ ioport->lbal_addr = (unsigned long) mmio + (ATA_REG_LBAL * 4);
+ ioport->lbam_addr = (unsigned long) mmio + (ATA_REG_LBAM * 4);
+ ioport->lbah_addr = (unsigned long) mmio + (ATA_REG_LBAH * 4);
+ ioport->device_addr = (unsigned long) mmio + (ATA_REG_DEVICE * 4);
+ ioport->status_addr =
+ ioport->command_addr = (unsigned long) mmio + (ATA_REG_STATUS * 4);
+ ioport->altstatus_addr =
+ ioport->ctl_addr = (unsigned long) mmio + 0x20;
+}
+
+static int nv_adma_host_init(struct ata_probe_ent *probe_ent)
+{
+ struct pci_dev *pdev = to_pci_dev(probe_ent->dev);
+ unsigned int i;
+ u32 tmp32;
+
+ VPRINTK("ENTER\n");
+
+ // enable ADMA on the ports
+ pci_read_config_dword(pdev, NV_MCP_SATA_CFG_20, &tmp32);
+ tmp32 |= NV_MCP_SATA_CFG_20_PORT0_EN |
+ NV_MCP_SATA_CFG_20_PORT0_PWB_EN |
+ NV_MCP_SATA_CFG_20_PORT1_EN |
+ NV_MCP_SATA_CFG_20_PORT1_PWB_EN;
+
+ pci_write_config_dword(pdev, NV_MCP_SATA_CFG_20, tmp32);
+
+ for (i = 0; i < probe_ent->n_ports; i++)
+ nv_adma_setup_port(probe_ent, i);
+
+ for (i = 0; i < probe_ent->n_ports; i++) {
+ void __iomem *mmio = __nv_adma_ctl_block(probe_ent->mmio_base, i);
+ u16 tmp;
+
+ /* enable interrupt, clear reset if not already clear */
+ tmp = readw(mmio + NV_ADMA_CTL);
+ writew(tmp | NV_ADMA_CTL_AIEN, mmio + NV_ADMA_CTL);
+ }
+
+ return 0;
+}
+
+static int nv_adma_check_atapi_dma(struct ata_queued_cmd *qc)
+{
+ return !adma_enable_atapi_dma;
+}
+
+static void nv_adma_fill_aprd(struct ata_queued_cmd *qc,
+ struct scatterlist *sg,
+ int idx,
+ struct nv_adma_prd *aprd)
+{
+ u32 flags;
+
+ memset(aprd, 0, sizeof(struct nv_adma_prd));
+
+ flags = 0;
+ if (qc->tf.flags & ATA_TFLAG_WRITE)
+ flags |= NV_APRD_WRITE;
+ if (idx == qc->n_elem - 1)
+ flags |= NV_APRD_END;
+ else if (idx != 4)
+ flags |= NV_APRD_CONT;
+
+ aprd->addr = cpu_to_le64(((u64)sg_dma_address(sg)));
+ aprd->len = cpu_to_le32(((u32)sg_dma_len(sg))); /* len in bytes */
+ aprd->flags = cpu_to_le32(flags);
+}
+
+static void nv_adma_fill_sg(struct ata_queued_cmd *qc, struct nv_adma_cpb *cpb)
+{
+ struct nv_adma_port_priv *pp = qc->ap->private_data;
+ unsigned int idx;
+ struct nv_adma_prd *aprd;
+ struct scatterlist *sg;
+
+ VPRINTK("ENTER\n");
+
+ idx = 0;
+
+ ata_for_each_sg(sg, qc) {
+ aprd = (idx < 5) ? &cpb->aprd[idx] : &pp->aprd[NV_ADMA_SGTBL_LEN * qc->tag + (idx-5)];
+ nv_adma_fill_aprd(qc, sg, idx, aprd);
+ idx++;
+ }
+ if (idx > 5)
+ cpb->next_aprd = cpu_to_le64(((u64)(pp->aprd_dma + NV_ADMA_SGTBL_SZ * qc->tag)));
+}
+
+static void nv_adma_qc_prep(struct ata_queued_cmd *qc)
+{
+ struct nv_adma_port_priv *pp = qc->ap->private_data;
+ struct nv_adma_cpb *cpb = &pp->cpb[qc->tag];
+ u8 ctl_flags = NV_CPB_CTL_CPB_VALID |
+ NV_CPB_CTL_APRD_VALID |
+ NV_CPB_CTL_IEN;
+
+ VPRINTK("ENTER\n");
+
+ VPRINTK("qc->flags = 0x%x\n", (u32)qc->flags);
+
+ if (!(qc->flags & ATA_QCFLAG_DMAMAP) ||
+ (qc->tf.protocol == ATA_PROT_ATAPI_DMA && !adma_enable_atapi_dma)) {
+ ata_qc_prep(qc);
+ return;
+ }
+
+ memset(cpb, 0, sizeof(struct nv_adma_cpb));
+
+ cpb->len = 3;
+ cpb->tag = qc->tag;
+ cpb->next_cpb_idx = 0;
+
+ // turn on NCQ flags for NCQ commands
+ if (qc->tf.protocol == ATA_PROT_NCQ)
+ ctl_flags |= NV_CPB_CTL_QUEUE | NV_CPB_CTL_FPDMA;
+
+ nv_adma_tf_to_cpb(&qc->tf, cpb->tf);
+
+ nv_adma_fill_sg(qc, cpb);
+
+ /* Be paranoid and don't let the device see NV_CPB_CTL_CPB_VALID until we are
+ finished filling in all of the contents */
+ wmb();
+ cpb->ctl_flags = ctl_flags;
+}
+
+static unsigned int nv_adma_qc_issue(struct ata_queued_cmd *qc)
+{
+ void __iomem *mmio = nv_adma_ctl_block(qc->ap);
+
+ VPRINTK("ENTER\n");
+
+ if (!(qc->flags & ATA_QCFLAG_DMAMAP) ||
+ (qc->tf.protocol == ATA_PROT_ATAPI_DMA && !adma_enable_atapi_dma)) {
+ VPRINTK("no dmamap, using ATA register mode: 0x%x\n", (u32)qc->flags);
+ // use ATA register mode
+ nv_adma_register_mode(qc->ap);
+ return ata_qc_issue_prot(qc);
+ } else
+ nv_adma_mode(qc->ap);
+
+ //
+ // write append register, command tag in lower 8 bits
+ // and (number of cpbs to append -1) in top 8 bits
+ //
+ wmb();
+ writew(qc->tag, mmio + NV_ADMA_APPEND);
+
+ DPRINTK("Issued tag %u\n",qc->tag);
+
+ return 0;
+}
+
static irqreturn_t nv_generic_interrupt(int irq, void *dev_instance)
{
struct ata_host *host = dev_instance;
@@ -462,6 +1252,56 @@ static void nv_error_handler(struct ata_
nv_hardreset, ata_std_postreset);
}
+static void nv_adma_error_handler(struct ata_port *ap)
+{
+ struct nv_adma_port_priv *pp = ap->private_data;
+ if(!(pp->flags & NV_ADMA_PORT_REGISTER_MODE)) {
+ void __iomem *mmio = nv_adma_ctl_block(ap);
+ int i;
+ u16 tmp;
+
+ u32 notifier = readl(mmio + NV_ADMA_NOTIFIER);
+ u32 notifier_error = readl(mmio + NV_ADMA_NOTIFIER_ERROR);
+ u32 gen_ctl = readl(nv_adma_gen_block(ap) + NV_ADMA_GEN_CTL);
+ u32 status = readw(mmio + NV_ADMA_STAT);
+
+ dev_printk(KERN_ERR, ap->dev, "EH port %u in ADMA, notifier 0x%X "
+ "notifier_error 0x%X gen_ctl 0x%X status 0x%X\n",ap->port_no,
+ notifier, notifier_error, gen_ctl, status);
+
+ for( i=0;i<NV_ADMA_MAX_CPBS;i++) {
+ struct nv_adma_cpb *cpb = &pp->cpb[i];
+ if( cpb->ctl_flags || cpb->resp_flags )
+ dev_printk(KERN_ERR, ap->dev,
+ "CPB %d: ctl_flags 0x%x, resp_flags 0x%x\n",
+ i, cpb->ctl_flags, cpb->resp_flags);
+ }
+
+ /* Push us back into port register mode for error handling. */
+ nv_adma_register_mode(ap);
+
+ dev_printk(KERN_NOTICE, ap->dev, "Resetting port %u\n", ap->port_no);
+
+ /* Mark all of the CPBs as invalid to prevent them from being executed */
+ for( i=0;i<NV_ADMA_MAX_CPBS;i++)
+ pp->cpb[i].ctl_flags &= ~NV_CPB_CTL_CPB_VALID;
+
+ /* clear CPB fetch count */
+ writew(0, mmio + NV_ADMA_CPB_COUNT);
+
+ /* Reset channel */
+ tmp = readw(mmio + NV_ADMA_CTL);
+ writew(tmp | NV_ADMA_CTL_CHANNEL_RESET, mmio + NV_ADMA_CTL);
+ readl( mmio + NV_ADMA_CTL ); /* flush posted write */
+ udelay(1);
+ writew(tmp & ~NV_ADMA_CTL_CHANNEL_RESET, mmio + NV_ADMA_CTL);
+ readl( mmio + NV_ADMA_CTL ); /* flush posted write */
+ }
+
+ ata_bmdma_drive_eh(ap, ata_std_prereset, ata_std_softreset,
+ nv_hardreset, ata_std_postreset);
+}
+
static int nv_init_one (struct pci_dev *pdev, const struct pci_device_id *ent)
{
static int printed_version = 0;
@@ -471,6 +1311,8 @@ static int nv_init_one (struct pci_dev *
int rc;
u32 bar;
unsigned long base;
+ unsigned long type = ent->driver_data;
+ u64 dma_mask = ATA_DMA_MASK;
// Make sure this is a SATA controller by counting the number of bars
// (NVIDIA SATA controllers will always have six bars). Otherwise,
@@ -481,6 +1323,26 @@ static int nv_init_one (struct pci_dev *
if (!printed_version++)
dev_printk(KERN_DEBUG, &pdev->dev, "version " DRV_VERSION "\n");
+
+ if(type >= CK804 && adma_enabled) {
+ dev_printk(KERN_NOTICE, &pdev->dev, "Using ADMA mode\n");
+ type = ADMA;
+
+ /*
+ * NVIDIA reports that ADMA mode does not support ATAPI commands.
+ * Therefore ATAPI commands are sent through the legacy interface.
+ * However, the legacy interface only supports 32-bit DMA.
+ * Therefore, if we need to do DMA through the legacy interface, we must
+ * use only 32-bit DMA as we cannot guarantee that we can satisfy DMA requests
+ * above 4GB.
+ * If the user requests enabling ATAPI DMA, restrict DMA to 32-bit.
+ */
+ if(!adma_enable_atapi_dma)
+ dma_mask = DMA_64BIT_MASK;
+ else
+ dev_printk(KERN_NOTICE, &pdev->dev,
+ "ADMA ATAPI DMA enabled, DMA restricted to below 4GB\n");
+ }
rc = pci_enable_device(pdev);
if (rc)
@@ -492,16 +1354,16 @@ static int nv_init_one (struct pci_dev *
goto err_out_disable;
}
- rc = pci_set_dma_mask(pdev, ATA_DMA_MASK);
+ rc = pci_set_dma_mask(pdev, dma_mask);
if (rc)
goto err_out_regions;
- rc = pci_set_consistent_dma_mask(pdev, ATA_DMA_MASK);
+ rc = pci_set_consistent_dma_mask(pdev, dma_mask);
if (rc)
goto err_out_regions;
rc = -ENOMEM;
- ppi[0] = ppi[1] = &nv_port_info[ent->driver_data];
+ ppi[0] = ppi[1] = &nv_port_info[type];
probe_ent = ata_pci_init_native_mode(pdev, ppi, ATA_PORT_PRIMARY | ATA_PORT_SECONDARY);
if (!probe_ent)
goto err_out_regions;
@@ -518,7 +1380,7 @@ static int nv_init_one (struct pci_dev *
probe_ent->port[1].scr_addr = base + NV_PORT1_SCR_REG_OFFSET;
/* enable SATA space for CK804 */
- if (ent->driver_data == CK804) {
+ if (type >= CK804) {
u8 regval;
pci_read_config_byte(pdev, NV_MCP_SATA_CFG_20, ®val);
@@ -528,6 +1390,12 @@ static int nv_init_one (struct pci_dev *
pci_set_master(pdev);
+ if (type == ADMA) {
+ rc = nv_adma_host_init(probe_ent);
+ if (rc)
+ goto err_out_iounmap;
+ }
+
rc = ata_device_add(probe_ent);
if (rc != NV_PORTS)
goto err_out_iounmap;
@@ -562,6 +1430,33 @@ static void nv_ck804_host_stop(struct at
ata_pci_host_stop(host);
}
+static void nv_adma_host_stop(struct ata_host *host)
+{
+ struct pci_dev *pdev = to_pci_dev(host->dev);
+ int i;
+ u32 tmp32;
+
+ for (i = 0; i < host->n_ports; i++) {
+ void __iomem *mmio = __nv_adma_ctl_block(host->mmio_base, i);
+ u16 tmp;
+
+ /* disable interrupt */
+ tmp = readw(mmio + NV_ADMA_CTL);
+ writew(tmp & ~NV_ADMA_CTL_AIEN, mmio + NV_ADMA_CTL);
+ }
+
+ // disable ADMA on the ports
+ pci_read_config_dword(pdev, NV_MCP_SATA_CFG_20, &tmp32);
+ tmp32 &= ~(NV_MCP_SATA_CFG_20_PORT0_EN |
+ NV_MCP_SATA_CFG_20_PORT0_PWB_EN |
+ NV_MCP_SATA_CFG_20_PORT1_EN |
+ NV_MCP_SATA_CFG_20_PORT1_PWB_EN);
+
+ pci_write_config_dword(pdev, NV_MCP_SATA_CFG_20, tmp32);
+
+ nv_ck804_host_stop(host);
+}
+
static int __init nv_init(void)
{
return pci_register_driver(&nv_pci_driver);
@@ -574,3 +1469,7 @@ static void __exit nv_exit(void)
module_init(nv_init);
module_exit(nv_exit);
+module_param_named(adma, adma_enabled, bool, 0444);
+MODULE_PARM_DESC(adma, "Enable use of ADMA (Default: false)");
+module_param_named(adma_atapi_dma, adma_enable_atapi_dma, bool, 0444);
+MODULE_PARM_DESC(adma, "Enable ATAPI DMA in ADMA mode, restricts DMA to under 4GB (Default: false)");
^ permalink raw reply [flat|nested] 23+ messages in thread
* Re: [RFC PATCH] nForce4 ADMA with NCQ: It's aliiiive..
2006-10-11 5:07 ` Robert Hancock
@ 2006-10-11 10:30 ` Jens Axboe
2006-10-13 3:17 ` Robert Hancock
0 siblings, 1 reply; 23+ messages in thread
From: Jens Axboe @ 2006-10-11 10:30 UTC (permalink / raw)
To: Robert Hancock
Cc: Allen Martin, Jeff Garzik, linux-kernel, linux-ide, prakash
On Tue, Oct 10 2006, Robert Hancock wrote:
> Allen Martin wrote:
> >>But I really don't think that is necessary. I will take a
> >>look at docs and see how things match up, when I am much more
> >>awake. Most likely you need to be using another set of
> >>registers, and be all MMIO, all the time.
> >
> >You shouldn't be touching BM registers when ADMA is enabled, it can
> >cause bad things to happen.
> >
> >You should be using BM registers when doing ATAPI protocol though, as it
> >doesn't work through ADMA. So I wouldn't say you should be using MMIO
> >all the time.
> >
> >-Allen
>
> OK, I've updated the code to take this into account, an updated patch is
> attached. However, this does raise an issue. If we have to fall back to
> legacy mode to do ATAPI DMA, this means that we can't do 64-bit DMA for
> such transfers. Since by the time the driver gets a request the SGs have
> already been created based on the set DMA mask, the only way I can see
> to handle this is to either allow ATAPI DMA or 64-bit DMA, not both.
> I've chosen to default to 64-bit DMA in this version, but there is a
> module parameter which allows overriding this if you care more about
> using ATAPI devices than efficiency with over 4GB of RAM. I'm open to
> suggestions on a better way to handle this..
Should be easily fixable - in general, set 64-bit dma mask. Then when
you detect an atapi device, lower the dma mask settings to 32-bit dma
for that device only. So the pci device in question gets a full 64-bit
dma mask, the attached scsi devices can have lower masks if necessary.
I'd suggest doing this off slave config.
--
Jens Axboe
^ permalink raw reply [flat|nested] 23+ messages in thread
* Re: [RFC PATCH] nForce4 ADMA with NCQ: It's aliiiive..
2006-10-11 10:30 ` Jens Axboe
@ 2006-10-13 3:17 ` Robert Hancock
2006-10-13 8:04 ` Jens Axboe
2006-10-16 16:17 ` [RFC PATCH] nForce4 ADMA with NCQ: It's aliiiive Mark Lord
0 siblings, 2 replies; 23+ messages in thread
From: Robert Hancock @ 2006-10-13 3:17 UTC (permalink / raw)
To: Jens Axboe; +Cc: Allen Martin, Jeff Garzik, linux-kernel, linux-ide, prakash
Jens Axboe wrote:
> On Tue, Oct 10 2006, Robert Hancock wrote:
>> Allen Martin wrote:
>>>> But I really don't think that is necessary. I will take a
>>>> look at docs and see how things match up, when I am much more
>>>> awake. Most likely you need to be using another set of
>>>> registers, and be all MMIO, all the time.
>>> You shouldn't be touching BM registers when ADMA is enabled, it can
>>> cause bad things to happen.
>>>
>>> You should be using BM registers when doing ATAPI protocol though, as it
>>> doesn't work through ADMA. So I wouldn't say you should be using MMIO
>>> all the time.
>>>
>>> -Allen
>> OK, I've updated the code to take this into account, an updated patch is
>> attached. However, this does raise an issue. If we have to fall back to
>> legacy mode to do ATAPI DMA, this means that we can't do 64-bit DMA for
>> such transfers. Since by the time the driver gets a request the SGs have
>> already been created based on the set DMA mask, the only way I can see
>> to handle this is to either allow ATAPI DMA or 64-bit DMA, not both.
>> I've chosen to default to 64-bit DMA in this version, but there is a
>> module parameter which allows overriding this if you care more about
>> using ATAPI devices than efficiency with over 4GB of RAM. I'm open to
>> suggestions on a better way to handle this..
>
> Should be easily fixable - in general, set 64-bit dma mask. Then when
> you detect an atapi device, lower the dma mask settings to 32-bit dma
> for that device only. So the pci device in question gets a full 64-bit
> dma mask, the attached scsi devices can have lower masks if necessary.
> I'd suggest doing this off slave config.
>
I think that should be feasible.. However, one problem is that
slave_config only has access to the struct scsi_device and the
ata_scsi_find_dev function to turn that into a struct ata_device isn't
exported, which it would need to be in order to do anything useful
inside the driver for slave_config. We could export it, or I suppose the
other place we could do this handling would be postreset, as at that
point we should know what kind of device is attached.. any comments?
Also, how is the driver supposed to be setting the DMA mask for the SCSI
device? I suppose blk_queue_bounce_limit would work, but it seems a bit
odd to use block layer calls at the libata driver level.
I also noticed that I'm still using the default 64KB libata dma_boundary
value, this should be 4GB for ADMA mode (but fixed up back to the
default if an ATAPI device is connected, same as with the DMA mask).
--
Robert Hancock Saskatoon, SK, Canada
To email, remove "nospam" from hancockr@nospamshaw.ca
Home Page: http://www.roberthancock.com/
^ permalink raw reply [flat|nested] 23+ messages in thread
* Re: [RFC PATCH] nForce4 ADMA with NCQ: It's aliiiive..
2006-10-13 3:17 ` Robert Hancock
@ 2006-10-13 8:04 ` Jens Axboe
2006-10-17 3:34 ` [PATCH] sata_nv ADMA/NCQ support for nForce4 Robert Hancock
2006-10-16 16:17 ` [RFC PATCH] nForce4 ADMA with NCQ: It's aliiiive Mark Lord
1 sibling, 1 reply; 23+ messages in thread
From: Jens Axboe @ 2006-10-13 8:04 UTC (permalink / raw)
To: Robert Hancock
Cc: Allen Martin, Jeff Garzik, linux-kernel, linux-ide, prakash
On Thu, Oct 12 2006, Robert Hancock wrote:
> Jens Axboe wrote:
> >On Tue, Oct 10 2006, Robert Hancock wrote:
> >>Allen Martin wrote:
> >>>>But I really don't think that is necessary. I will take a
> >>>>look at docs and see how things match up, when I am much more
> >>>>awake. Most likely you need to be using another set of
> >>>>registers, and be all MMIO, all the time.
> >>>You shouldn't be touching BM registers when ADMA is enabled, it can
> >>>cause bad things to happen.
> >>>
> >>>You should be using BM registers when doing ATAPI protocol though, as it
> >>>doesn't work through ADMA. So I wouldn't say you should be using MMIO
> >>>all the time.
> >>>
> >>>-Allen
> >>OK, I've updated the code to take this into account, an updated patch is
> >>attached. However, this does raise an issue. If we have to fall back to
> >>legacy mode to do ATAPI DMA, this means that we can't do 64-bit DMA for
> >>such transfers. Since by the time the driver gets a request the SGs have
> >>already been created based on the set DMA mask, the only way I can see
> >>to handle this is to either allow ATAPI DMA or 64-bit DMA, not both.
> >>I've chosen to default to 64-bit DMA in this version, but there is a
> >>module parameter which allows overriding this if you care more about
> >>using ATAPI devices than efficiency with over 4GB of RAM. I'm open to
> >>suggestions on a better way to handle this..
> >
> >Should be easily fixable - in general, set 64-bit dma mask. Then when
> >you detect an atapi device, lower the dma mask settings to 32-bit dma
> >for that device only. So the pci device in question gets a full 64-bit
> >dma mask, the attached scsi devices can have lower masks if necessary.
> >I'd suggest doing this off slave config.
> >
>
> I think that should be feasible.. However, one problem is that
> slave_config only has access to the struct scsi_device and the
> ata_scsi_find_dev function to turn that into a struct ata_device isn't
> exported, which it would need to be in order to do anything useful
> inside the driver for slave_config. We could export it, or I suppose the
> other place we could do this handling would be postreset, as at that
> point we should know what kind of device is attached.. any comments?
What else do you need? From the scsi device, call the
blk_queue_bounce_limit() on the queue tored in there. That shold be it.
> Also, how is the driver supposed to be setting the DMA mask for the SCSI
> device? I suppose blk_queue_bounce_limit would work, but it seems a bit
> odd to use block layer calls at the libata driver level.
That is the right thing to do.
> I also noticed that I'm still using the default 64KB libata dma_boundary
> value, this should be 4GB for ADMA mode (but fixed up back to the
> default if an ATAPI device is connected, same as with the DMA mask).
You can set that from the same location as the bounce limit.
--
Jens Axboe
^ permalink raw reply [flat|nested] 23+ messages in thread
* Re: [RFC PATCH] nForce4 ADMA with NCQ: It's aliiiive..
2006-10-13 3:17 ` Robert Hancock
2006-10-13 8:04 ` Jens Axboe
@ 2006-10-16 16:17 ` Mark Lord
2006-10-16 23:40 ` Robert Hancock
1 sibling, 1 reply; 23+ messages in thread
From: Mark Lord @ 2006-10-16 16:17 UTC (permalink / raw)
To: Robert Hancock
Cc: Jens Axboe, Allen Martin, Jeff Garzik, linux-kernel, linux-ide,
prakash
Robert Hancock wrote:
>
> I also noticed that I'm still using the default 64KB libata dma_boundary
> value, this should be 4GB for ADMA mode (but fixed up back to the
> default if an ATAPI device is connected, same as with the DMA mask).
Be careful of that. The original PDC hardware for ADMA still had
the "don't cross a 64KB boundary" requirement.
Cheers
^ permalink raw reply [flat|nested] 23+ messages in thread
* Re: [RFC PATCH] nForce4 ADMA with NCQ: It's aliiiive..
2006-10-16 16:17 ` [RFC PATCH] nForce4 ADMA with NCQ: It's aliiiive Mark Lord
@ 2006-10-16 23:40 ` Robert Hancock
2006-10-17 0:04 ` Mark Lord
0 siblings, 1 reply; 23+ messages in thread
From: Robert Hancock @ 2006-10-16 23:40 UTC (permalink / raw)
To: Mark Lord
Cc: Jens Axboe, Allen Martin, Jeff Garzik, linux-kernel, linux-ide,
prakash
Mark Lord wrote:
> Robert Hancock wrote:
>>
>> I also noticed that I'm still using the default 64KB libata
>> dma_boundary value, this should be 4GB for ADMA mode (but fixed up
>> back to the default if an ATAPI device is connected, same as with the
>> DMA mask).
>
> Be careful of that. The original PDC hardware for ADMA still had
> the "don't cross a 64KB boundary" requirement.
>
> Cheers
That is part of the ADMA spec - but in that case, how come the
pdc_adma.c driver sets the dma_boundary in the SCSI host template to
4GB? That seems wrong.
In any case, however, I really doubt the NVIDIA ADMA controller shares
this limitation - they provide 32 bits for the region length (standard
ADMA only has 16 bits), which wouldn't be very useful if you couldn't
cross a 64KB boundary..
--
Robert Hancock Saskatoon, SK, Canada
To email, remove "nospam" from hancockr@nospamshaw.ca
Home Page: http://www.roberthancock.com/
^ permalink raw reply [flat|nested] 23+ messages in thread
* Re: [RFC PATCH] nForce4 ADMA with NCQ: It's aliiiive..
2006-10-16 23:40 ` Robert Hancock
@ 2006-10-17 0:04 ` Mark Lord
0 siblings, 0 replies; 23+ messages in thread
From: Mark Lord @ 2006-10-17 0:04 UTC (permalink / raw)
To: Robert Hancock
Cc: Mark Lord, Jens Axboe, Allen Martin, Jeff Garzik, linux-kernel,
linux-ide, prakash
Robert Hancock wrote:
> Mark Lord wrote:
>> Robert Hancock wrote:
..
>> Be careful of that. The original PDC hardware for ADMA still had
>> the "don't cross a 64KB boundary" requirement.
..
> That is part of the ADMA spec - but in that case, how come the
> pdc_adma.c driver sets the dma_boundary in the SCSI host template to
> 4GB? That seems wrong.
Ah, you are correct. My memory is fading of such things.
The PDC parts are good for 2^34 transfer sizes (bytes).
Cheers
^ permalink raw reply [flat|nested] 23+ messages in thread
* [PATCH] sata_nv ADMA/NCQ support for nForce4
2006-10-13 8:04 ` Jens Axboe
@ 2006-10-17 3:34 ` Robert Hancock
2006-10-17 3:37 ` Mark Lord
0 siblings, 1 reply; 23+ messages in thread
From: Robert Hancock @ 2006-10-17 3:34 UTC (permalink / raw)
To: Jens Axboe; +Cc: Allen Martin, Jeff Garzik, linux-kernel, linux-ide, prakash
[-- Attachment #1: Type: text/plain, Size: 2371 bytes --]
Jens Axboe wrote:
>> I think that should be feasible.. However, one problem is that
>> slave_config only has access to the struct scsi_device and the
>> ata_scsi_find_dev function to turn that into a struct ata_device isn't
>> exported, which it would need to be in order to do anything useful
>> inside the driver for slave_config. We could export it, or I suppose the
>> other place we could do this handling would be postreset, as at that
>> point we should know what kind of device is attached.. any comments?
>
> What else do you need? From the scsi device, call the
> blk_queue_bounce_limit() on the queue tored in there. That shold be it.
>
>> Also, how is the driver supposed to be setting the DMA mask for the SCSI
>> device? I suppose blk_queue_bounce_limit would work, but it seems a bit
>> odd to use block layer calls at the libata driver level.
>
> That is the right thing to do.
>
>> I also noticed that I'm still using the default 64KB libata dma_boundary
>> value, this should be 4GB for ADMA mode (but fixed up back to the
>> default if an ATAPI device is connected, same as with the DMA mask).
>
> You can set that from the same location as the bounce limit.
OK, I've cut another patch (v4) which incorporates the auto-setting of
the lower bounce limit and segment boundary when an ATAPI device is
connected to the port. ATAPI was not actually tested as I don't have any
ATAPI SATA devices. I also cleaned up a few other things such as
handling failure to set the 64-bit DMA mask as well as using
ata_port_printk where appropriate.
I think this resolves all of the outstanding comments and known issues..
as before, commenters, testers, and mergers are welcome :-)
This patch is against 2.6.19-rc2.
---
This patch adds support for ADMA mode on NVIDIA nForce4 (CK804/MCP04)
SATA controllers to the sata_nv driver. Benefits of ADMA mode include:
-NCQ support
-Reduced CPU overhead (controller DMAs command information from memory
instead of them being pushed in by the CPU)
-Full 64-bit DMA support
ADMA mode is disabled by default in this version. To enable ADMA on
supported controllers, set the module parameter adma_enabled=1.
Signed-off-by: Robert Hancock <hancockr@shaw.ca>
--
Robert Hancock Saskatoon, SK, Canada
To email, remove "nospam" from hancockr@nospamshaw.ca
Home Page: http://www.roberthancock.com/
[-- Attachment #2: sata_nv-adma-ncq-v4.patch --]
[-- Type: text/plain, Size: 32000 bytes --]
diff -rup --exclude='*.o' --exclude='*mod.c' --exclude='*.cmd' --exclude='*.ko' linux-2.6.19-rc2/drivers/ata/libata-sff.c linux-2.6.19-rc2-adma/drivers/ata/libata-sff.c
--- linux-2.6.19-rc2/drivers/ata/libata-sff.c 2006-10-16 18:04:44.000000000 -0600
+++ linux-2.6.19-rc2-adma/drivers/ata/libata-sff.c 2006-10-16 18:02:34.000000000 -0600
@@ -732,7 +732,7 @@ void ata_bmdma_drive_eh(struct ata_port
qc->tf.protocol == ATA_PROT_ATAPI_DMA)) {
u8 host_stat;
- host_stat = ata_bmdma_status(ap);
+ host_stat = ap->ops->bmdma_status(ap);
ata_ehi_push_desc(&ehc->i, "BMDMA stat 0x%x", host_stat);
diff -rup --exclude='*.o' --exclude='*mod.c' --exclude='*.cmd' --exclude='*.ko' linux-2.6.19-rc2/drivers/ata/sata_nv.c linux-2.6.19-rc2-adma/drivers/ata/sata_nv.c
--- linux-2.6.19-rc2/drivers/ata/sata_nv.c 2006-10-16 18:04:44.000000000 -0600
+++ linux-2.6.19-rc2-adma/drivers/ata/sata_nv.c 2006-10-16 20:50:45.000000000 -0600
@@ -29,6 +29,11 @@
* NV-specific details such as register offsets, SATA phy location,
* hotplug info, etc.
*
+ * CK804/MCP04 controllers support an alternate programming interface
+ * similar to the ADMA specification (with some modifications).
+ * This allows the use of NCQ. Non-DMA-mapped ATA commands are still
+ * sent through the legacy interface.
+ *
*/
#include <linux/kernel.h>
@@ -40,10 +45,13 @@
#include <linux/interrupt.h>
#include <linux/device.h>
#include <scsi/scsi_host.h>
+#include <scsi/scsi_device.h>
#include <linux/libata.h>
#define DRV_NAME "sata_nv"
-#define DRV_VERSION "2.0"
+#define DRV_VERSION "3.0"
+
+#define NV_ADMA_DMA_BOUNDARY 0xffffffffUL
enum {
NV_PORTS = 2,
@@ -78,8 +86,135 @@ enum {
// For PCI config register 20
NV_MCP_SATA_CFG_20 = 0x50,
NV_MCP_SATA_CFG_20_SATA_SPACE_EN = 0x04,
+ NV_MCP_SATA_CFG_20_PORT0_EN = (1 << 17),
+ NV_MCP_SATA_CFG_20_PORT1_EN = (1 << 16),
+ NV_MCP_SATA_CFG_20_PORT0_PWB_EN = (1 << 14),
+ NV_MCP_SATA_CFG_20_PORT1_PWB_EN = (1 << 12),
+
+ NV_ADMA_MAX_CPBS = 32,
+ NV_ADMA_CPB_SZ = 128,
+ NV_ADMA_APRD_SZ = 16,
+ NV_ADMA_SGTBL_LEN = (1024 - NV_ADMA_CPB_SZ) / NV_ADMA_APRD_SZ,
+ NV_ADMA_SGTBL_SZ = NV_ADMA_SGTBL_LEN * NV_ADMA_APRD_SZ,
+ NV_ADMA_PORT_PRIV_DMA_SZ = NV_ADMA_MAX_CPBS * (NV_ADMA_CPB_SZ + NV_ADMA_SGTBL_SZ),
+
+ // BAR5 offset to ADMA general registers
+ NV_ADMA_GEN = 0x400,
+ NV_ADMA_GEN_CTL = 0x00,
+ NV_ADMA_NOTIFIER_CLEAR = 0x30,
+
+ // BAR5 offset to ADMA ports
+ NV_ADMA_PORT = 0x480,
+
+ // size of ADMA port register space
+ NV_ADMA_PORT_SIZE = 0x100,
+
+ // ADMA port registers
+ NV_ADMA_CTL = 0x40,
+ NV_ADMA_CPB_COUNT = 0x42,
+ NV_ADMA_NEXT_CPB_IDX = 0x43,
+ NV_ADMA_STAT = 0x44,
+ NV_ADMA_CPB_BASE_LOW = 0x48,
+ NV_ADMA_CPB_BASE_HIGH = 0x4C,
+ NV_ADMA_APPEND = 0x50,
+ NV_ADMA_NOTIFIER = 0x68,
+ NV_ADMA_NOTIFIER_ERROR = 0x6C,
+
+ // NV_ADMA_CTL register bits
+ NV_ADMA_CTL_HOTPLUG_IEN = (1 << 0),
+ NV_ADMA_CTL_CHANNEL_RESET = (1 << 5),
+ NV_ADMA_CTL_GO = (1 << 7),
+ NV_ADMA_CTL_AIEN = (1 << 8),
+ NV_ADMA_CTL_READ_NON_COHERENT = (1 << 11),
+ NV_ADMA_CTL_WRITE_NON_COHERENT = (1 << 12),
+
+ // CPB response flag bits
+ NV_CPB_RESP_DONE = (1 << 0),
+ NV_CPB_RESP_ATA_ERR = (1 << 3),
+ NV_CPB_RESP_CMD_ERR = (1 << 4),
+ NV_CPB_RESP_CPB_ERR = (1 << 7),
+
+ // CPB control flag bits
+ NV_CPB_CTL_CPB_VALID = (1 << 0),
+ NV_CPB_CTL_QUEUE = (1 << 1),
+ NV_CPB_CTL_APRD_VALID = (1 << 2),
+ NV_CPB_CTL_IEN = (1 << 3),
+ NV_CPB_CTL_FPDMA = (1 << 4),
+
+ // APRD flags
+ NV_APRD_WRITE = (1 << 1),
+ NV_APRD_END = (1 << 2),
+ NV_APRD_CONT = (1 << 3),
+
+ // NV_ADMA_STAT flags
+ NV_ADMA_STAT_TIMEOUT = (1 << 0),
+ NV_ADMA_STAT_HOTUNPLUG = (1 << 1),
+ NV_ADMA_STAT_HOTPLUG = (1 << 2),
+ NV_ADMA_STAT_CPBERR = (1 << 4),
+ NV_ADMA_STAT_SERROR = (1 << 5),
+ NV_ADMA_STAT_CMD_COMPLETE = (1 << 6),
+ NV_ADMA_STAT_IDLE = (1 << 8),
+ NV_ADMA_STAT_LEGACY = (1 << 9),
+ NV_ADMA_STAT_STOPPED = (1 << 10),
+ NV_ADMA_STAT_DONE = (1 << 12),
+ NV_ADMA_STAT_ERR = (NV_ADMA_STAT_CPBERR | NV_ADMA_STAT_TIMEOUT),
+
+ // port flags
+ NV_ADMA_PORT_REGISTER_MODE = (1 << 0),
+
+};
+
+/* ADMA Physical Region Descriptor - one SG segment */
+struct nv_adma_prd {
+ __le64 addr;
+ __le32 len;
+ u8 flags;
+ u8 packet_len;
+ __le16 reserved;
+};
+
+enum nv_adma_regbits {
+ CMDEND = (1 << 15), /* end of command list */
+ WNB = (1 << 14), /* wait-not-BSY */
+ IGN = (1 << 13), /* ignore this entry */
+ CS1n = (1 << (4 + 8)), /* std. PATA signals follow... */
+ DA2 = (1 << (2 + 8)),
+ DA1 = (1 << (1 + 8)),
+ DA0 = (1 << (0 + 8)),
};
+/* ADMA Command Parameter Block
+ The first 5 SG segments are stored inside the Command Parameter Block itself.
+ If there are more than 5 segments the remainder are stored in a separate
+ memory area indicated by next_aprd. */
+struct nv_adma_cpb {
+ u8 resp_flags; //0
+ u8 reserved1; //1
+ u8 ctl_flags; //2
+ // len is length of taskfile in 64 bit words
+ u8 len; //3
+ u8 tag; //4
+ u8 next_cpb_idx; //5
+ __le16 reserved2; //6-7
+ __le16 tf[12]; //8-31
+ struct nv_adma_prd aprd[5]; //32-111
+ __le64 next_aprd; //112-119
+ __le64 reserved3; //120-127
+};
+
+
+struct nv_adma_port_priv {
+ struct nv_adma_cpb *cpb;
+ dma_addr_t cpb_dma;
+ struct nv_adma_prd *aprd;
+ dma_addr_t aprd_dma;
+ u8 flags;
+ u32 notifier;
+ u32 notifier_error;
+};
+
+#define NV_ADMA_CHECK_INTR(GCTL, PORT) ((GCTL) & ( 1 << (19 + (12 * (PORT)))))
+
static int nv_init_one (struct pci_dev *pdev, const struct pci_device_id *ent);
static void nv_ck804_host_stop(struct ata_host *host);
static irqreturn_t nv_generic_interrupt(int irq, void *dev_instance);
@@ -93,13 +228,27 @@ static void nv_nf2_thaw(struct ata_port
static void nv_ck804_freeze(struct ata_port *ap);
static void nv_ck804_thaw(struct ata_port *ap);
static void nv_error_handler(struct ata_port *ap);
+static int nv_adma_slave_config(struct scsi_device *sdev);
+static void nv_adma_qc_prep(struct ata_queued_cmd *qc);
+static unsigned int nv_adma_qc_issue(struct ata_queued_cmd *qc);
+static irqreturn_t nv_adma_interrupt(int irq, void *dev_instance);
+static void nv_adma_irq_clear(struct ata_port *ap);
+static int nv_adma_port_start(struct ata_port *ap);
+static void nv_adma_port_stop(struct ata_port *ap);
+static void nv_adma_error_handler(struct ata_port *ap);
+static void nv_adma_host_stop(struct ata_host *host);
+static void nv_adma_bmdma_setup(struct ata_queued_cmd *qc);
+static void nv_adma_bmdma_start(struct ata_queued_cmd *qc);
+static void nv_adma_bmdma_stop(struct ata_queued_cmd *qc);
+static u8 nv_adma_bmdma_status(struct ata_port *ap);
enum nv_host_type
{
GENERIC,
NFORCE2,
NFORCE3 = NFORCE2, /* NF2 == NF3 as far as sata_nv is concerned */
- CK804
+ CK804,
+ ADMA
};
static const struct pci_device_id nv_pci_tbl[] = {
@@ -156,6 +305,25 @@ static struct scsi_host_template nv_sht
.bios_param = ata_std_bios_param,
};
+static struct scsi_host_template nv_adma_sht = {
+ .module = THIS_MODULE,
+ .name = DRV_NAME,
+ .ioctl = ata_scsi_ioctl,
+ .queuecommand = ata_scsi_queuecmd,
+ .can_queue = NV_ADMA_MAX_CPBS,
+ .this_id = ATA_SHT_THIS_ID,
+ .sg_tablesize = NV_ADMA_SGTBL_LEN + 5,
+ .max_sectors = ATA_MAX_SECTORS,
+ .cmd_per_lun = ATA_SHT_CMD_PER_LUN,
+ .emulated = ATA_SHT_EMULATED,
+ .use_clustering = ATA_SHT_USE_CLUSTERING,
+ .proc_name = DRV_NAME,
+ .dma_boundary = NV_ADMA_DMA_BOUNDARY,
+ .slave_configure = nv_adma_slave_config,
+ .slave_destroy = ata_scsi_slave_destroy,
+ .bios_param = ata_std_bios_param,
+};
+
static const struct ata_port_operations nv_generic_ops = {
.port_disable = ata_port_disable,
.tf_load = ata_tf_load,
@@ -237,6 +405,33 @@ static const struct ata_port_operations
.host_stop = nv_ck804_host_stop,
};
+static const struct ata_port_operations nv_adma_ops = {
+ .port_disable = ata_port_disable,
+ .tf_load = ata_tf_load,
+ .tf_read = ata_tf_read,
+ .exec_command = ata_exec_command,
+ .check_status = ata_check_status,
+ .dev_select = ata_std_dev_select,
+ .bmdma_setup = nv_adma_bmdma_setup,
+ .bmdma_start = nv_adma_bmdma_start,
+ .bmdma_stop = nv_adma_bmdma_stop,
+ .bmdma_status = nv_adma_bmdma_status,
+ .qc_prep = nv_adma_qc_prep,
+ .qc_issue = nv_adma_qc_issue,
+ .freeze = nv_ck804_freeze,
+ .thaw = nv_ck804_thaw,
+ .error_handler = nv_adma_error_handler,
+ .post_internal_cmd = nv_adma_bmdma_stop,
+ .data_xfer = ata_mmio_data_xfer,
+ .irq_handler = nv_adma_interrupt,
+ .irq_clear = nv_adma_irq_clear,
+ .scr_read = nv_scr_read,
+ .scr_write = nv_scr_write,
+ .port_start = nv_adma_port_start,
+ .port_stop = nv_adma_port_stop,
+ .host_stop = nv_adma_host_stop,
+};
+
static struct ata_port_info nv_port_info[] = {
/* generic */
{
@@ -265,6 +460,16 @@ static struct ata_port_info nv_port_info
.udma_mask = NV_UDMA_MASK,
.port_ops = &nv_ck804_ops,
},
+ /* ADMA */
+ {
+ .sht = &nv_adma_sht,
+ .flags = ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY |
+ ATA_FLAG_MMIO | ATA_FLAG_NCQ,
+ .pio_mask = NV_PIO_MASK,
+ .mwdma_mask = NV_MWDMA_MASK,
+ .udma_mask = NV_UDMA_MASK,
+ .port_ops = &nv_adma_ops,
+ },
};
MODULE_AUTHOR("NVIDIA");
@@ -273,6 +478,621 @@ MODULE_LICENSE("GPL");
MODULE_DEVICE_TABLE(pci, nv_pci_tbl);
MODULE_VERSION(DRV_VERSION);
+static int adma_enabled;
+
+static int nv_adma_slave_config(struct scsi_device *sdev)
+{
+ struct ata_port *ap = ata_shost_to_port(sdev->host);
+ u64 bounce_limit;
+ unsigned long segment_boundary;
+ int rc;
+
+ rc = ata_scsi_slave_config(sdev);
+
+ if (sdev->id >= ATA_MAX_DEVICES || sdev->channel || sdev->lun)
+ /* Not a proper libata device, ignore */
+ return rc;
+
+ if (ap->device[sdev->id].class == ATA_DEV_ATAPI) {
+ /*
+ * NVIDIA reports that ADMA mode does not support ATAPI commands.
+ * Therefore ATAPI commands are sent through the legacy interface.
+ * However, the legacy interface only supports 32-bit DMA.
+ * Restrict DMA to 32-bit and DMA boundary to 64KB as required by the
+ * legacy interface when an ATAPI device is connected.
+ */
+ bounce_limit = ATA_DMA_MASK;
+ segment_boundary = ATA_DMA_BOUNDARY;
+ }
+ else {
+ bounce_limit = *ap->dev->dma_mask;
+ segment_boundary = NV_ADMA_DMA_BOUNDARY;
+ }
+
+ blk_queue_bounce_limit(sdev->request_queue, bounce_limit);
+ blk_queue_segment_boundary(sdev->request_queue, segment_boundary);
+ ata_port_printk(ap, KERN_INFO, "bounce limit 0x%llX, segment boundary 0x%lX\n",
+ (unsigned long long)bounce_limit, segment_boundary);
+ return rc;
+}
+
+static unsigned int nv_adma_tf_to_cpb(struct ata_taskfile *tf, u16 *cpb)
+{
+ unsigned int idx = 0;
+
+ cpb[idx++] = cpu_to_le16((ATA_REG_DEVICE << 8) | tf->device | WNB);
+
+ if ((tf->flags & ATA_TFLAG_LBA48) == 0) {
+ cpb[idx++] = cpu_to_le16(IGN);
+ cpb[idx++] = cpu_to_le16(IGN);
+ cpb[idx++] = cpu_to_le16(IGN);
+ cpb[idx++] = cpu_to_le16(IGN);
+ cpb[idx++] = cpu_to_le16(IGN);
+ }
+ else {
+ cpb[idx++] = cpu_to_le16((ATA_REG_ERR << 8) | tf->hob_feature);
+ cpb[idx++] = cpu_to_le16((ATA_REG_NSECT << 8) | tf->hob_nsect);
+ cpb[idx++] = cpu_to_le16((ATA_REG_LBAL << 8) | tf->hob_lbal);
+ cpb[idx++] = cpu_to_le16((ATA_REG_LBAM << 8) | tf->hob_lbam);
+ cpb[idx++] = cpu_to_le16((ATA_REG_LBAH << 8) | tf->hob_lbah);
+ }
+ cpb[idx++] = cpu_to_le16((ATA_REG_ERR << 8) | tf->feature);
+ cpb[idx++] = cpu_to_le16((ATA_REG_NSECT << 8) | tf->nsect);
+ cpb[idx++] = cpu_to_le16((ATA_REG_LBAL << 8) | tf->lbal);
+ cpb[idx++] = cpu_to_le16((ATA_REG_LBAM << 8) | tf->lbam);
+ cpb[idx++] = cpu_to_le16((ATA_REG_LBAH << 8) | tf->lbah);
+
+ cpb[idx++] = cpu_to_le16((ATA_REG_CMD << 8) | tf->command | CMDEND);
+
+ return idx;
+}
+
+static inline void __iomem *__nv_adma_ctl_block(void __iomem *mmio,
+ unsigned int port_no)
+{
+ mmio += NV_ADMA_PORT + port_no * NV_ADMA_PORT_SIZE;
+ return mmio;
+}
+
+static inline void __iomem *nv_adma_ctl_block(struct ata_port *ap)
+{
+ return __nv_adma_ctl_block(ap->host->mmio_base, ap->port_no);
+}
+
+static inline void __iomem *nv_adma_gen_block(struct ata_port *ap)
+{
+ return (ap->host->mmio_base + NV_ADMA_GEN);
+}
+
+static inline void __iomem *nv_adma_notifier_clear_block(struct ata_port *ap)
+{
+ return (nv_adma_gen_block(ap) + NV_ADMA_NOTIFIER_CLEAR + (4 * ap->port_no));
+}
+
+static int nv_adma_host_intr(struct ata_port *ap)
+{
+ void __iomem *mmio = nv_adma_ctl_block(ap);
+ struct nv_adma_port_priv *pp = ap->private_data;
+ u16 status;
+ u32 gen_ctl;
+ u16 flags;
+ int have_global_err = 0;
+
+ /* if in ATA register mode, use standard ata interrupt handler */
+ if (pp->flags & NV_ADMA_PORT_REGISTER_MODE) {
+ struct ata_queued_cmd *qc;
+ VPRINTK("in ATA register mode\n");
+ qc = ata_qc_from_tag(ap, ap->active_tag);
+ if (qc && (!(qc->tf.flags & ATA_TFLAG_POLLING)))
+ return ata_host_intr(ap, qc);
+ else {
+ // No request pending? Clear interrupt status
+ // anyway, in case there's one pending.
+ ap->ops->check_status(ap);
+ return 1;
+ }
+ }
+
+ gen_ctl = readl(nv_adma_gen_block(ap) + NV_ADMA_GEN_CTL);
+
+ if( !NV_ADMA_CHECK_INTR(gen_ctl, ap->port_no) && !pp->notifier && !pp->notifier_error)
+ /* Nothing to do */
+ return 0;
+
+ status = readw(mmio + NV_ADMA_STAT);
+
+ /* Clear status so that any CPB completions after this point in the handler will
+ raise another interrupt */
+ writew(status, mmio + NV_ADMA_STAT);
+ readl(mmio + NV_ADMA_STAT); /* flush posted write */
+ mb();
+
+ /* freeze if hotplugged */
+ if (unlikely(status & (NV_ADMA_STAT_HOTPLUG | NV_ADMA_STAT_HOTUNPLUG))) {
+ ata_port_printk(ap, KERN_NOTICE, "Hotplug event on port %u, freezing\n",
+ ap->port_no);
+ ata_port_freeze(ap);
+ return 1;
+ }
+
+ if (pp->notifier_error) {
+ ata_port_printk(ap, KERN_ERR, "port %u notifier error, stat=0x%x, "
+ "notifier=0x%X, notifier_error=0x%X", ap->port_no, status, pp->notifier,
+ pp->notifier_error );
+ have_global_err = 1;
+ }
+
+ if (status & NV_ADMA_STAT_TIMEOUT) {
+ ata_port_printk(ap, KERN_ERR, "port %u timeout, stat=0x%x\n", ap->port_no, status);
+ have_global_err = 1;
+ }
+ if (status & NV_ADMA_STAT_CPBERR) {
+ ata_port_printk(ap, KERN_ERR, "port %u CPB error, stat=0x%x\n", ap->port_no, status);
+ have_global_err = 1;
+ }
+ if (status & NV_ADMA_STAT_STOPPED) {
+ VPRINTK("ADMA stopped, stat = 0x%x\n", status);
+ }
+ if (status & NV_ADMA_STAT_CMD_COMPLETE) {
+ VPRINTK("ADMA command complete, stat = 0x%x\n", status);
+ }
+ if ((status & NV_ADMA_STAT_DONE) || have_global_err) {
+ /** Check CPBs for completed commands */
+ int i;
+ for( i=0;i<NV_ADMA_MAX_CPBS;i++)
+ if( ap->sactive & (1 << i ) ||
+ ( ata_tag_valid(ap->active_tag) && ap->active_tag == i ) ) {
+ struct nv_adma_cpb *cpb = &pp->cpb[i];
+ int complete = 0, have_err = 0;
+ flags = cpb->resp_flags;
+ VPRINTK("CPB %d, stat=0x%x, flags=0x%x\n", i, status, flags);
+
+ if (flags & NV_CPB_RESP_DONE) {
+ VPRINTK("CPB flags done, flags=0x%x\n", flags);
+ complete = 1;
+ }
+ if (flags & NV_CPB_RESP_ATA_ERR) {
+ ata_port_printk(ap, KERN_ERR, "port %u CPB flags ATA err, flags=0x%x\n", ap->port_no, flags);
+ have_err = 1;
+ complete = 1;
+ }
+ if (flags & NV_CPB_RESP_CMD_ERR) {
+ ata_port_printk(ap, KERN_ERR, "port %u CPB flags CMD err, flags=0x%x\n", ap->port_no, flags);
+ have_err = 1;
+ complete = 1;
+ }
+ if (flags & NV_CPB_RESP_CPB_ERR) {
+ ata_port_printk(ap, KERN_ERR, "port %u CPB flags CPB err, flags=0x%x\n", ap->port_no, flags);
+ have_err = 1;
+ complete = 1;
+ }
+ if(complete || have_global_err)
+ {
+ struct ata_queued_cmd *qc = ata_qc_from_tag(ap, i);
+ if(likely(qc)) {
+ u8 ata_status = 0;
+ /* Only use the ATA port status for non-NCQ commands.
+ For NCQ commands the current status may have nothing to do with
+ the command just completed. */
+ if(qc->tf.protocol != ATA_PROT_NCQ)
+ ata_status = readb(mmio + (ATA_REG_STATUS * 4));
+
+ if(have_err || have_global_err)
+ ata_status |= ATA_ERR;
+
+ qc->err_mask |= ac_err_mask(ata_status);
+ DPRINTK("Completing qc from tag %d with err_mask %u\n",i,qc->err_mask);
+ ata_qc_complete(qc);
+ }
+ }
+ }
+ }
+
+ return 1; /* irq handled if we got here */
+}
+
+static irqreturn_t nv_adma_interrupt(int irq, void *dev_instance)
+{
+ struct ata_host *host = dev_instance;
+ int i, handled = 0;
+
+ spin_lock(&host->lock);
+
+ for (i = 0; i < host->n_ports; i++) {
+ struct ata_port *ap = host->ports[i];
+
+ if (ap && !(ap->flags & ATA_FLAG_DISABLED)) {
+ struct nv_adma_port_priv *pp = ap->private_data;
+ void __iomem *mmio = nv_adma_ctl_block(ap);
+ // read notifiers
+ pp->notifier = readl(mmio + NV_ADMA_NOTIFIER);
+ pp->notifier_error = readl(mmio + NV_ADMA_NOTIFIER_ERROR);
+ handled += nv_adma_host_intr(ap);
+ }
+ }
+
+ /* Clear notifiers. Yes, this does appear to have to be done at the very end,
+ otherwise things grind to a halt quite quickly. */
+ if (handled)
+ for (i = 0; i < host->n_ports; i++) {
+ struct ata_port *ap = host->ports[i];
+ struct nv_adma_port_priv *pp = ap->private_data;
+ writel(pp->notifier | pp->notifier_error,
+ nv_adma_notifier_clear_block(ap));
+ }
+
+ spin_unlock(&host->lock);
+
+ return IRQ_RETVAL(handled);
+}
+
+static void nv_adma_irq_clear(struct ata_port *ap)
+{
+ void __iomem *mmio = nv_adma_ctl_block(ap);
+ u16 status = readw(mmio + NV_ADMA_STAT);
+ u32 notifier = readl(mmio + NV_ADMA_NOTIFIER);
+ u32 notifier_error = readl(mmio + NV_ADMA_NOTIFIER_ERROR);
+
+ /* clear ADMA status */
+ writew(status, mmio + NV_ADMA_STAT);
+ writel(notifier | notifier_error,
+ nv_adma_notifier_clear_block(ap));
+
+ /** clear legacy status */
+ ap->flags &= ~ATA_FLAG_MMIO;
+ ata_bmdma_irq_clear(ap);
+ ap->flags |= ATA_FLAG_MMIO;
+}
+
+static void nv_adma_bmdma_setup(struct ata_queued_cmd *qc)
+{
+ struct nv_adma_port_priv *pp = qc->ap->private_data;
+
+ if(pp->flags & NV_ADMA_PORT_REGISTER_MODE) {
+ WARN_ON(1);
+ return;
+ }
+
+ qc->ap->flags &= ~ATA_FLAG_MMIO;
+ ata_bmdma_setup(qc);
+ qc->ap->flags |= ATA_FLAG_MMIO;
+}
+
+static void nv_adma_bmdma_start(struct ata_queued_cmd *qc)
+{
+ struct nv_adma_port_priv *pp = qc->ap->private_data;
+
+ if(pp->flags & NV_ADMA_PORT_REGISTER_MODE) {
+ WARN_ON(1);
+ return;
+ }
+
+ qc->ap->flags &= ~ATA_FLAG_MMIO;
+ ata_bmdma_start(qc);
+ qc->ap->flags |= ATA_FLAG_MMIO;
+}
+
+static void nv_adma_bmdma_stop(struct ata_queued_cmd *qc)
+{
+ struct nv_adma_port_priv *pp = qc->ap->private_data;
+
+ if(pp->flags & NV_ADMA_PORT_REGISTER_MODE)
+ return;
+
+ qc->ap->flags &= ~ATA_FLAG_MMIO;
+ ata_bmdma_stop(qc);
+ qc->ap->flags |= ATA_FLAG_MMIO;
+}
+
+static u8 nv_adma_bmdma_status(struct ata_port *ap)
+{
+ u8 status;
+ struct nv_adma_port_priv *pp = ap->private_data;
+
+ WARN_ON(pp->flags & NV_ADMA_PORT_REGISTER_MODE);
+
+ ap->flags &= ~ATA_FLAG_MMIO;
+ status = ata_bmdma_status(ap);
+ ap->flags |= ATA_FLAG_MMIO;
+ return status;
+}
+
+static void nv_adma_register_mode(struct ata_port *ap)
+{
+ void __iomem *mmio = nv_adma_ctl_block(ap);
+ struct nv_adma_port_priv *pp = ap->private_data;
+ u16 tmp;
+
+ if (pp->flags & NV_ADMA_PORT_REGISTER_MODE)
+ return;
+
+ tmp = readw(mmio + NV_ADMA_CTL);
+ writew(tmp & ~NV_ADMA_CTL_GO, mmio + NV_ADMA_CTL);
+
+ pp->flags |= NV_ADMA_PORT_REGISTER_MODE;
+}
+
+static void nv_adma_mode(struct ata_port *ap)
+{
+ void __iomem *mmio = nv_adma_ctl_block(ap);
+ struct nv_adma_port_priv *pp = ap->private_data;
+ u16 tmp;
+
+ if (!(pp->flags & NV_ADMA_PORT_REGISTER_MODE))
+ return;
+
+ tmp = readw(mmio + NV_ADMA_CTL);
+ writew(tmp | NV_ADMA_CTL_GO, mmio + NV_ADMA_CTL);
+
+ pp->flags &= ~NV_ADMA_PORT_REGISTER_MODE;
+}
+
+static int nv_adma_port_start(struct ata_port *ap)
+{
+ struct device *dev = ap->host->dev;
+ struct nv_adma_port_priv *pp;
+ int rc;
+ void *mem;
+ dma_addr_t mem_dma;
+ void __iomem *mmio = nv_adma_ctl_block(ap);
+ u16 tmp;
+
+ VPRINTK("ENTER\n");
+
+ rc = ata_port_start(ap);
+ if (rc)
+ return rc;
+
+ pp = kzalloc(sizeof(*pp), GFP_KERNEL);
+ if (!pp) {
+ rc = -ENOMEM;
+ goto err_out;
+ }
+
+ mem = dma_alloc_coherent(dev, NV_ADMA_PORT_PRIV_DMA_SZ,
+ &mem_dma, GFP_KERNEL);
+
+ VPRINTK("dma memory: vaddr = 0x%p, paddr = 0x%p\n", mem, (void*)mem_dma);
+
+ if (!mem) {
+ rc = -ENOMEM;
+ goto err_out_kfree;
+ }
+ memset(mem, 0, NV_ADMA_PORT_PRIV_DMA_SZ);
+
+ /*
+ * First item in chunk of DMA memory:
+ * 128-byte command parameter block (CPB)
+ * one for each command tag
+ */
+ pp->cpb = mem;
+ pp->cpb_dma = mem_dma;
+
+ VPRINTK("cpb = 0x%p, cpb_dma = 0x%p\n", pp->cpb, (void*)pp->cpb_dma);
+
+ writel(mem_dma & 0xFFFFFFFF, mmio + NV_ADMA_CPB_BASE_LOW);
+ writel((mem_dma >> 16 ) >> 16, mmio + NV_ADMA_CPB_BASE_HIGH);
+
+ mem += NV_ADMA_MAX_CPBS * NV_ADMA_CPB_SZ;
+ mem_dma += NV_ADMA_MAX_CPBS * NV_ADMA_CPB_SZ;
+
+ /*
+ * Second item: block of ADMA_SGTBL_LEN s/g entries
+ */
+ pp->aprd = mem;
+ pp->aprd_dma = mem_dma;
+
+ VPRINTK("aprd = 0x%p, aprd_dma = 0x%p\n", pp->aprd, (void*)pp->aprd_dma);
+
+ ap->private_data = pp;
+
+ // clear any outstanding interrupt conditions
+ writew(0xffff, mmio + NV_ADMA_STAT);
+
+ // initialize port variables
+ pp->flags = NV_ADMA_PORT_REGISTER_MODE;
+
+ /* clear CPB fetch count */
+ writew(0, mmio + NV_ADMA_CPB_COUNT);
+
+ /* clear GO for register mode */
+ tmp = readw(mmio + NV_ADMA_CTL);
+ writew(tmp & ~NV_ADMA_CTL_GO, mmio + NV_ADMA_CTL);
+
+ tmp = readw(mmio + NV_ADMA_CTL);
+ writew(tmp | NV_ADMA_CTL_CHANNEL_RESET, mmio + NV_ADMA_CTL);
+ readl( mmio + NV_ADMA_CTL ); /* flush posted write */
+ udelay(1);
+ writew(tmp & ~NV_ADMA_CTL_CHANNEL_RESET, mmio + NV_ADMA_CTL);
+ readl( mmio + NV_ADMA_CTL ); /* flush posted write */
+
+ return 0;
+
+err_out_kfree:
+ kfree(pp);
+err_out:
+ ata_port_stop(ap);
+ return rc;
+}
+
+static void nv_adma_port_stop(struct ata_port *ap)
+{
+ struct device *dev = ap->host->dev;
+ struct nv_adma_port_priv *pp = ap->private_data;
+ void __iomem *mmio = nv_adma_ctl_block(ap);
+
+ VPRINTK("ENTER\n");
+
+ writew(0, mmio + NV_ADMA_CTL);
+
+ ap->private_data = NULL;
+ dma_free_coherent(dev, NV_ADMA_PORT_PRIV_DMA_SZ, pp->cpb, pp->cpb_dma);
+ kfree(pp);
+ ata_port_stop(ap);
+}
+
+
+static void nv_adma_setup_port(struct ata_probe_ent *probe_ent, unsigned int port)
+{
+ void __iomem *mmio = probe_ent->mmio_base;
+ struct ata_ioports *ioport = &probe_ent->port[port];
+
+ VPRINTK("ENTER\n");
+
+ mmio += NV_ADMA_PORT + port * NV_ADMA_PORT_SIZE;
+
+ ioport->cmd_addr = (unsigned long) mmio;
+ ioport->data_addr = (unsigned long) mmio + (ATA_REG_DATA * 4);
+ ioport->error_addr =
+ ioport->feature_addr = (unsigned long) mmio + (ATA_REG_ERR * 4);
+ ioport->nsect_addr = (unsigned long) mmio + (ATA_REG_NSECT * 4);
+ ioport->lbal_addr = (unsigned long) mmio + (ATA_REG_LBAL * 4);
+ ioport->lbam_addr = (unsigned long) mmio + (ATA_REG_LBAM * 4);
+ ioport->lbah_addr = (unsigned long) mmio + (ATA_REG_LBAH * 4);
+ ioport->device_addr = (unsigned long) mmio + (ATA_REG_DEVICE * 4);
+ ioport->status_addr =
+ ioport->command_addr = (unsigned long) mmio + (ATA_REG_STATUS * 4);
+ ioport->altstatus_addr =
+ ioport->ctl_addr = (unsigned long) mmio + 0x20;
+}
+
+static int nv_adma_host_init(struct ata_probe_ent *probe_ent)
+{
+ struct pci_dev *pdev = to_pci_dev(probe_ent->dev);
+ unsigned int i;
+ u32 tmp32;
+
+ VPRINTK("ENTER\n");
+
+ // enable ADMA on the ports
+ pci_read_config_dword(pdev, NV_MCP_SATA_CFG_20, &tmp32);
+ tmp32 |= NV_MCP_SATA_CFG_20_PORT0_EN |
+ NV_MCP_SATA_CFG_20_PORT0_PWB_EN |
+ NV_MCP_SATA_CFG_20_PORT1_EN |
+ NV_MCP_SATA_CFG_20_PORT1_PWB_EN;
+
+ pci_write_config_dword(pdev, NV_MCP_SATA_CFG_20, tmp32);
+
+ for (i = 0; i < probe_ent->n_ports; i++)
+ nv_adma_setup_port(probe_ent, i);
+
+ for (i = 0; i < probe_ent->n_ports; i++) {
+ void __iomem *mmio = __nv_adma_ctl_block(probe_ent->mmio_base, i);
+ u16 tmp;
+
+ /* enable interrupt, clear reset if not already clear */
+ tmp = readw(mmio + NV_ADMA_CTL);
+ writew(tmp | NV_ADMA_CTL_AIEN, mmio + NV_ADMA_CTL);
+ }
+
+ return 0;
+}
+
+static void nv_adma_fill_aprd(struct ata_queued_cmd *qc,
+ struct scatterlist *sg,
+ int idx,
+ struct nv_adma_prd *aprd)
+{
+ u32 flags;
+
+ memset(aprd, 0, sizeof(struct nv_adma_prd));
+
+ flags = 0;
+ if (qc->tf.flags & ATA_TFLAG_WRITE)
+ flags |= NV_APRD_WRITE;
+ if (idx == qc->n_elem - 1)
+ flags |= NV_APRD_END;
+ else if (idx != 4)
+ flags |= NV_APRD_CONT;
+
+ aprd->addr = cpu_to_le64(((u64)sg_dma_address(sg)));
+ aprd->len = cpu_to_le32(((u32)sg_dma_len(sg))); /* len in bytes */
+ aprd->flags = cpu_to_le32(flags);
+}
+
+static void nv_adma_fill_sg(struct ata_queued_cmd *qc, struct nv_adma_cpb *cpb)
+{
+ struct nv_adma_port_priv *pp = qc->ap->private_data;
+ unsigned int idx;
+ struct nv_adma_prd *aprd;
+ struct scatterlist *sg;
+
+ VPRINTK("ENTER\n");
+
+ idx = 0;
+
+ ata_for_each_sg(sg, qc) {
+ aprd = (idx < 5) ? &cpb->aprd[idx] : &pp->aprd[NV_ADMA_SGTBL_LEN * qc->tag + (idx-5)];
+ nv_adma_fill_aprd(qc, sg, idx, aprd);
+ idx++;
+ }
+ if (idx > 5)
+ cpb->next_aprd = cpu_to_le64(((u64)(pp->aprd_dma + NV_ADMA_SGTBL_SZ * qc->tag)));
+}
+
+static void nv_adma_qc_prep(struct ata_queued_cmd *qc)
+{
+ struct nv_adma_port_priv *pp = qc->ap->private_data;
+ struct nv_adma_cpb *cpb = &pp->cpb[qc->tag];
+ u8 ctl_flags = NV_CPB_CTL_CPB_VALID |
+ NV_CPB_CTL_APRD_VALID |
+ NV_CPB_CTL_IEN;
+
+ VPRINTK("ENTER\n");
+
+ VPRINTK("qc->flags = 0x%x\n", (u32)qc->flags);
+
+ if (!(qc->flags & ATA_QCFLAG_DMAMAP)) {
+ ata_qc_prep(qc);
+ return;
+ }
+
+ memset(cpb, 0, sizeof(struct nv_adma_cpb));
+
+ cpb->len = 3;
+ cpb->tag = qc->tag;
+ cpb->next_cpb_idx = 0;
+
+ // turn on NCQ flags for NCQ commands
+ if (qc->tf.protocol == ATA_PROT_NCQ)
+ ctl_flags |= NV_CPB_CTL_QUEUE | NV_CPB_CTL_FPDMA;
+
+ nv_adma_tf_to_cpb(&qc->tf, cpb->tf);
+
+ nv_adma_fill_sg(qc, cpb);
+
+ /* Be paranoid and don't let the device see NV_CPB_CTL_CPB_VALID until we are
+ finished filling in all of the contents */
+ wmb();
+ cpb->ctl_flags = ctl_flags;
+}
+
+static unsigned int nv_adma_qc_issue(struct ata_queued_cmd *qc)
+{
+ void __iomem *mmio = nv_adma_ctl_block(qc->ap);
+
+ VPRINTK("ENTER\n");
+
+ if (!(qc->flags & ATA_QCFLAG_DMAMAP)) {
+ VPRINTK("no dmamap, using ATA register mode: 0x%x\n", (u32)qc->flags);
+ // use ATA register mode
+ nv_adma_register_mode(qc->ap);
+ return ata_qc_issue_prot(qc);
+ } else
+ nv_adma_mode(qc->ap);
+
+ //
+ // write append register, command tag in lower 8 bits
+ // and (number of cpbs to append -1) in top 8 bits
+ //
+ wmb();
+ writew(qc->tag, mmio + NV_ADMA_APPEND);
+
+ DPRINTK("Issued tag %u\n",qc->tag);
+
+ return 0;
+}
+
static irqreturn_t nv_generic_interrupt(int irq, void *dev_instance)
{
struct ata_host *host = dev_instance;
@@ -462,6 +1282,56 @@ static void nv_error_handler(struct ata_
nv_hardreset, ata_std_postreset);
}
+static void nv_adma_error_handler(struct ata_port *ap)
+{
+ struct nv_adma_port_priv *pp = ap->private_data;
+ if(!(pp->flags & NV_ADMA_PORT_REGISTER_MODE)) {
+ void __iomem *mmio = nv_adma_ctl_block(ap);
+ int i;
+ u16 tmp;
+
+ u32 notifier = readl(mmio + NV_ADMA_NOTIFIER);
+ u32 notifier_error = readl(mmio + NV_ADMA_NOTIFIER_ERROR);
+ u32 gen_ctl = readl(nv_adma_gen_block(ap) + NV_ADMA_GEN_CTL);
+ u32 status = readw(mmio + NV_ADMA_STAT);
+
+ ata_port_printk(ap, KERN_ERR, "EH port %u in ADMA, notifier 0x%X "
+ "notifier_error 0x%X gen_ctl 0x%X status 0x%X\n",ap->port_no,
+ notifier, notifier_error, gen_ctl, status);
+
+ for( i=0;i<NV_ADMA_MAX_CPBS;i++) {
+ struct nv_adma_cpb *cpb = &pp->cpb[i];
+ if( cpb->ctl_flags || cpb->resp_flags )
+ ata_port_printk(ap, KERN_ERR,
+ "CPB %d: ctl_flags 0x%x, resp_flags 0x%x\n",
+ i, cpb->ctl_flags, cpb->resp_flags);
+ }
+
+ /* Push us back into port register mode for error handling. */
+ nv_adma_register_mode(ap);
+
+ ata_port_printk(ap, KERN_ERR, "Resetting port %u\n", ap->port_no);
+
+ /* Mark all of the CPBs as invalid to prevent them from being executed */
+ for( i=0;i<NV_ADMA_MAX_CPBS;i++)
+ pp->cpb[i].ctl_flags &= ~NV_CPB_CTL_CPB_VALID;
+
+ /* clear CPB fetch count */
+ writew(0, mmio + NV_ADMA_CPB_COUNT);
+
+ /* Reset channel */
+ tmp = readw(mmio + NV_ADMA_CTL);
+ writew(tmp | NV_ADMA_CTL_CHANNEL_RESET, mmio + NV_ADMA_CTL);
+ readl( mmio + NV_ADMA_CTL ); /* flush posted write */
+ udelay(1);
+ writew(tmp & ~NV_ADMA_CTL_CHANNEL_RESET, mmio + NV_ADMA_CTL);
+ readl( mmio + NV_ADMA_CTL ); /* flush posted write */
+ }
+
+ ata_bmdma_drive_eh(ap, ata_std_prereset, ata_std_softreset,
+ nv_hardreset, ata_std_postreset);
+}
+
static int nv_init_one (struct pci_dev *pdev, const struct pci_device_id *ent)
{
static int printed_version = 0;
@@ -471,6 +1341,8 @@ static int nv_init_one (struct pci_dev *
int rc;
u32 bar;
unsigned long base;
+ unsigned long type = ent->driver_data;
+ int mask_set = 0;
// Make sure this is a SATA controller by counting the number of bars
// (NVIDIA SATA controllers will always have six bars). Otherwise,
@@ -481,7 +1353,7 @@ static int nv_init_one (struct pci_dev *
if (!printed_version++)
dev_printk(KERN_DEBUG, &pdev->dev, "version " DRV_VERSION "\n");
-
+
rc = pci_enable_device(pdev);
if (rc)
goto err_out;
@@ -492,16 +1364,26 @@ static int nv_init_one (struct pci_dev *
goto err_out_disable;
}
- rc = pci_set_dma_mask(pdev, ATA_DMA_MASK);
- if (rc)
- goto err_out_regions;
- rc = pci_set_consistent_dma_mask(pdev, ATA_DMA_MASK);
- if (rc)
- goto err_out_regions;
+ if(type >= CK804 && adma_enabled) {
+ dev_printk(KERN_NOTICE, &pdev->dev, "Using ADMA mode\n");
+ type = ADMA;
+ if(!pci_set_dma_mask(pdev, DMA_64BIT_MASK) &&
+ !pci_set_consistent_dma_mask(pdev, DMA_64BIT_MASK))
+ mask_set = 1;
+ }
+
+ if(!mask_set) {
+ rc = pci_set_dma_mask(pdev, ATA_DMA_MASK);
+ if (rc)
+ goto err_out_regions;
+ rc = pci_set_consistent_dma_mask(pdev, ATA_DMA_MASK);
+ if (rc)
+ goto err_out_regions;
+ }
rc = -ENOMEM;
- ppi[0] = ppi[1] = &nv_port_info[ent->driver_data];
+ ppi[0] = ppi[1] = &nv_port_info[type];
probe_ent = ata_pci_init_native_mode(pdev, ppi, ATA_PORT_PRIMARY | ATA_PORT_SECONDARY);
if (!probe_ent)
goto err_out_regions;
@@ -518,7 +1400,7 @@ static int nv_init_one (struct pci_dev *
probe_ent->port[1].scr_addr = base + NV_PORT1_SCR_REG_OFFSET;
/* enable SATA space for CK804 */
- if (ent->driver_data == CK804) {
+ if (type >= CK804) {
u8 regval;
pci_read_config_byte(pdev, NV_MCP_SATA_CFG_20, ®val);
@@ -528,6 +1410,12 @@ static int nv_init_one (struct pci_dev *
pci_set_master(pdev);
+ if (type == ADMA) {
+ rc = nv_adma_host_init(probe_ent);
+ if (rc)
+ goto err_out_iounmap;
+ }
+
rc = ata_device_add(probe_ent);
if (rc != NV_PORTS)
goto err_out_iounmap;
@@ -562,6 +1450,33 @@ static void nv_ck804_host_stop(struct at
ata_pci_host_stop(host);
}
+static void nv_adma_host_stop(struct ata_host *host)
+{
+ struct pci_dev *pdev = to_pci_dev(host->dev);
+ int i;
+ u32 tmp32;
+
+ for (i = 0; i < host->n_ports; i++) {
+ void __iomem *mmio = __nv_adma_ctl_block(host->mmio_base, i);
+ u16 tmp;
+
+ /* disable interrupt */
+ tmp = readw(mmio + NV_ADMA_CTL);
+ writew(tmp & ~NV_ADMA_CTL_AIEN, mmio + NV_ADMA_CTL);
+ }
+
+ // disable ADMA on the ports
+ pci_read_config_dword(pdev, NV_MCP_SATA_CFG_20, &tmp32);
+ tmp32 &= ~(NV_MCP_SATA_CFG_20_PORT0_EN |
+ NV_MCP_SATA_CFG_20_PORT0_PWB_EN |
+ NV_MCP_SATA_CFG_20_PORT1_EN |
+ NV_MCP_SATA_CFG_20_PORT1_PWB_EN);
+
+ pci_write_config_dword(pdev, NV_MCP_SATA_CFG_20, tmp32);
+
+ nv_ck804_host_stop(host);
+}
+
static int __init nv_init(void)
{
return pci_register_driver(&nv_pci_driver);
@@ -574,3 +1489,5 @@ static void __exit nv_exit(void)
module_init(nv_init);
module_exit(nv_exit);
+module_param_named(adma, adma_enabled, bool, 0444);
+MODULE_PARM_DESC(adma, "Enable use of ADMA (Default: false)");
^ permalink raw reply [flat|nested] 23+ messages in thread
* Re: [PATCH] sata_nv ADMA/NCQ support for nForce4
2006-10-17 3:34 ` [PATCH] sata_nv ADMA/NCQ support for nForce4 Robert Hancock
@ 2006-10-17 3:37 ` Mark Lord
2006-10-17 4:24 ` Robert Hancock
0 siblings, 1 reply; 23+ messages in thread
From: Mark Lord @ 2006-10-17 3:37 UTC (permalink / raw)
To: Robert Hancock
Cc: Jens Axboe, Allen Martin, Jeff Garzik, linux-kernel, linux-ide,
prakash
Robert Hancock wrote:
>
> +/* ADMA Physical Region Descriptor - one SG segment */
> +struct nv_adma_prd {
> + __le64 addr;
> + __le32 len;
> + u8 flags;
> + u8 packet_len;
> + __le16 reserved;
> +};
..
> +/* ADMA Command Parameter Block
> + The first 5 SG segments are stored inside the Command Parameter Block itself.
> + If there are more than 5 segments the remainder are stored in a separate
> + memory area indicated by next_aprd. */
> +struct nv_adma_cpb {
> + u8 resp_flags; //0
> + u8 reserved1; //1
> + u8 ctl_flags; //2
> + // len is length of taskfile in 64 bit words
> + u8 len; //3
> + u8 tag; //4
> + u8 next_cpb_idx; //5
> + __le16 reserved2; //6-7
> + __le16 tf[12]; //8-31
> + struct nv_adma_prd aprd[5]; //32-111
> + __le64 next_aprd; //112-119
> + __le64 reserved3; //120-127
> +};
Are those CPB / PRD structs endian-safe when using a big-endian CPU?
Cheers
^ permalink raw reply [flat|nested] 23+ messages in thread
* Re: [PATCH] sata_nv ADMA/NCQ support for nForce4
2006-10-17 3:37 ` Mark Lord
@ 2006-10-17 4:24 ` Robert Hancock
2006-10-17 16:22 ` Mark Lord
0 siblings, 1 reply; 23+ messages in thread
From: Robert Hancock @ 2006-10-17 4:24 UTC (permalink / raw)
To: Mark Lord
Cc: Jens Axboe, Allen Martin, Jeff Garzik, linux-kernel, linux-ide,
prakash
Mark Lord wrote:
> Robert Hancock wrote:
>>
>> +/* ADMA Physical Region Descriptor - one SG segment */
>> +struct nv_adma_prd {
>> + __le64 addr;
>> + __le32 len;
>> + u8 flags;
>> + u8 packet_len;
>> + __le16 reserved;
>> +};
> ..
>> +/* ADMA Command Parameter Block
>> + The first 5 SG segments are stored inside the Command Parameter
>> Block itself.
>> + If there are more than 5 segments the remainder are stored in a
>> separate
>> + memory area indicated by next_aprd. */
>> +struct nv_adma_cpb {
>> + u8 resp_flags; //0
>> + u8 reserved1; //1
>> + u8 ctl_flags; //2
>> + // len is length of taskfile in 64 bit words
>> + u8 len; //3 + u8
>> tag; //4
>> + u8 next_cpb_idx; //5
>> + __le16 reserved2; //6-7
>> + __le16 tf[12]; //8-31
>> + struct nv_adma_prd aprd[5]; //32-111
>> + __le64 next_aprd; //112-119
>> + __le64 reserved3; //120-127
>> +};
>
>
> Are those CPB / PRD structs endian-safe when using a big-endian CPU?
>
> Cheers
They should be, I believe cpu_to_leXX is used whenever the multi-byte
elements are being written. Not that anyone would likely install a
big-endian CPU on an nForce4 chipset ;-)
--
Robert Hancock Saskatoon, SK, Canada
To email, remove "nospam" from hancockr@nospamshaw.ca
Home Page: http://www.roberthancock.com/
^ permalink raw reply [flat|nested] 23+ messages in thread
* Re: [PATCH] sata_nv ADMA/NCQ support for nForce4
2006-10-17 4:24 ` Robert Hancock
@ 2006-10-17 16:22 ` Mark Lord
2006-10-17 17:58 ` Roland Dreier
0 siblings, 1 reply; 23+ messages in thread
From: Mark Lord @ 2006-10-17 16:22 UTC (permalink / raw)
To: Robert Hancock
Cc: Jens Axboe, Allen Martin, Jeff Garzik, linux-kernel, linux-ide,
prakash
Robert Hancock wrote:
> Mark Lord wrote:
>> Robert Hancock wrote:
>>>
>>> +/* ADMA Physical Region Descriptor - one SG segment */
>>> +struct nv_adma_prd {
>>> + __le64 addr;
>>> + __le32 len;
>>> + u8 flags;
>>> + u8 packet_len;
..
>>> +struct nv_adma_cpb {
>>> + u8 resp_flags; //0
>>> + u8 reserved1; //1
>>> + u8 ctl_flags; //2
>>> + // len is length of taskfile in 64 bit words
>>> + u8 len; //3 + u8
>>> tag; //4
>>> + u8 next_cpb_idx; //5
>..
>> Are those CPB / PRD structs endian-safe when using a big-endian CPU?
>>
>> Cheers
>
> They should be, I believe cpu_to_leXX is used whenever the multi-byte
> elements are being written.
I was thinking more about the non wordsized fields,
such as the various u8 bytes that gcc will lay out differently
depending upon endianess.
Cheers
^ permalink raw reply [flat|nested] 23+ messages in thread
* Re: [PATCH] sata_nv ADMA/NCQ support for nForce4
2006-10-17 16:22 ` Mark Lord
@ 2006-10-17 17:58 ` Roland Dreier
2006-10-17 20:54 ` Mark Lord
0 siblings, 1 reply; 23+ messages in thread
From: Roland Dreier @ 2006-10-17 17:58 UTC (permalink / raw)
To: Mark Lord
Cc: Robert Hancock, Jens Axboe, Allen Martin, Jeff Garzik,
linux-kernel, linux-ide, prakash
Mark> I was thinking more about the non wordsized fields, such as
Mark> the various u8 bytes that gcc will lay out differently
Mark> depending upon endianess.
I don't know of any gcc version that changes the order of struct
fields. You might be thinking of bitfields, which are laid out in an
endian-dependent way.
- R.
^ permalink raw reply [flat|nested] 23+ messages in thread
* Re: [PATCH] sata_nv ADMA/NCQ support for nForce4
2006-10-17 17:58 ` Roland Dreier
@ 2006-10-17 20:54 ` Mark Lord
0 siblings, 0 replies; 23+ messages in thread
From: Mark Lord @ 2006-10-17 20:54 UTC (permalink / raw)
To: Roland Dreier
Cc: Robert Hancock, Jens Axboe, Allen Martin, Jeff Garzik,
linux-kernel, linux-ide, prakash
Roland Dreier wrote:
> Mark> I was thinking more about the non wordsized fields, such as
> Mark> the various u8 bytes that gcc will lay out differently
> Mark> depending upon endianess.
>
> I don't know of any gcc version that changes the order of struct
> fields. You might be thinking of bitfields, which are laid out in an
> endian-dependent way.
Ack. And also thinking about other compilers I work with,
which apparently are not so friendly.
You are correct -- I just verified it (again) for myself on x86 & PPC,
and the layout of u8 doesn't change.
Cheers!
^ permalink raw reply [flat|nested] 23+ messages in thread
* Re: [RFC PATCH] nForce4 ADMA with NCQ: It's aliiiive..
2006-10-10 6:44 ` Allen Martin
2006-10-10 7:52 ` Prakash Punnoor
@ 2006-11-30 16:46 ` Prakash Punnoor
2006-12-01 0:07 ` Robert Hancock
1 sibling, 1 reply; 23+ messages in thread
From: Prakash Punnoor @ 2006-11-30 16:46 UTC (permalink / raw)
To: Allen Martin; +Cc: Robert Hancock, linux-kernel, linux-ide, jeff
[-- Attachment #1: Type: text/plain, Size: 326 bytes --]
Am Dienstag 10 Oktober 2006 08:44 schrieb Allen Martin:
> No, only CK804 and MCP04 support ADMA. We'll be publishing some patches
> for NCQ support for MCP55/MCP61 soon.
Just wondering whether I missed them? It is now 1,5 month later...
--
(°= =°)
//\ Prakash Punnoor /\\
V_/ \_V
[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]
^ permalink raw reply [flat|nested] 23+ messages in thread
* Re: [RFC PATCH] nForce4 ADMA with NCQ: It's aliiiive..
2006-11-30 16:46 ` Prakash Punnoor
@ 2006-12-01 0:07 ` Robert Hancock
0 siblings, 0 replies; 23+ messages in thread
From: Robert Hancock @ 2006-12-01 0:07 UTC (permalink / raw)
To: Prakash Punnoor; +Cc: Allen Martin, linux-kernel, linux-ide, jeff
Prakash Punnoor wrote:
> Am Dienstag 10 Oktober 2006 08:44 schrieb Allen Martin:
>
>> No, only CK804 and MCP04 support ADMA. We'll be publishing some patches
>> for NCQ support for MCP55/MCP61 soon.
>
> Just wondering whether I missed them? It is now 1,5 month later...
I don't think you missed them.. It would be nice to get this support in,
those are the last NVIDIA chipsets which have NCQ support (apparently)
that don't support it under Linux. Even just the hardware docs would be
useful..
---
Robert Hancock Saskatoon, SK, Canada
To email, remove "nospam" from hancockr@nospamshaw.ca
Home Page: http://www.roberthancock.com/
^ permalink raw reply [flat|nested] 23+ messages in thread
end of thread, other threads:[~2006-12-01 0:09 UTC | newest]
Thread overview: 23+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2006-10-07 8:08 [RFC PATCH] nForce4 ADMA with NCQ: It's aliiiive Robert Hancock
2006-10-07 9:42 ` Prakash Punnoor
2006-10-07 15:54 ` Robert Hancock
2006-10-10 6:44 ` Allen Martin
2006-10-10 7:52 ` Prakash Punnoor
2006-11-30 16:46 ` Prakash Punnoor
2006-12-01 0:07 ` Robert Hancock
2006-10-07 15:28 ` Jeff Garzik
2006-10-07 23:49 ` Robert Hancock
2006-10-10 6:52 ` Allen Martin
2006-10-11 5:07 ` Robert Hancock
2006-10-11 10:30 ` Jens Axboe
2006-10-13 3:17 ` Robert Hancock
2006-10-13 8:04 ` Jens Axboe
2006-10-17 3:34 ` [PATCH] sata_nv ADMA/NCQ support for nForce4 Robert Hancock
2006-10-17 3:37 ` Mark Lord
2006-10-17 4:24 ` Robert Hancock
2006-10-17 16:22 ` Mark Lord
2006-10-17 17:58 ` Roland Dreier
2006-10-17 20:54 ` Mark Lord
2006-10-16 16:17 ` [RFC PATCH] nForce4 ADMA with NCQ: It's aliiiive Mark Lord
2006-10-16 23:40 ` Robert Hancock
2006-10-17 0:04 ` Mark Lord
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).