[PATCH 2.6.19-rc6] sata_promise updates

[PATCH 2.6.19-rc6] sata_promise updates

[Mikael Pettersson]

This patch updates the sata_promise driver as follows:
- Correct typo in definition of PDC_TBG_MODE: it's at 0x41C not 0x41
  in first-generation chips. This error caused PCI access alignment
  exceptions on SPARC64, and on all platforms it disabled the expected
  initialisation of TBG mode.
- Prevent TBG mode and SLEW rate initialisation in second-generation chips.
  These two registers have moved, TBG mode has been redefined, and
  Promise's ulsata2 driver no longer attempts to initialise them.
- Correct PCI device table so devices 0x3570, 0x3571, and 0x3d73 are
  marked as 2057x (2nd gen) not 2037x (1st gen).
- Correct PCI device table so device 0x3d17 is marked as 40518
  (2nd gen 4 ports) not 20319 (1st gen 4 ports).
- Correct pdc_ata_init_one() to treat 20771 as a second-generation chip.
  (The 20771 classification is redundant as it is equivalent to 2057x.)

Tested on 0x3d75 (2nd gen), 0x3d73 (2nd gen), and 0x3373 (1st gen) chips.
The information comes from the newly uploaded Promise SATA HW specs,
Promise's ultra and ulsata2 drivers, and debugging on 3d75/3d73/3373 chips.

Signed-off-by: Mikael Pettersson <mikpe@it.uu.se>

diff -rupN linux-2.6.19-rc6/drivers/ata/sata_promise.c linux-2.6.19-rc6.sata_promise-mikpe1/drivers/ata/sata_promise.c
--- linux-2.6.19-rc6/drivers/ata/sata_promise.c	2006-11-21 21:14:22.000000000 +0100
+++ linux-2.6.19-rc6.sata_promise-mikpe1/drivers/ata/sata_promise.c	2006-11-21 21:19:38.000000000 +0100
@@ -52,14 +52,14 @@
 enum {
 	PDC_PKT_SUBMIT		= 0x40, /* Command packet pointer addr */
 	PDC_INT_SEQMASK		= 0x40,	/* Mask of asserted SEQ INTs */
-	PDC_TBG_MODE		= 0x41,	/* TBG mode */
 	PDC_FLASH_CTL		= 0x44, /* Flash control register */
 	PDC_PCI_CTL		= 0x48, /* PCI control and status register */
 	PDC_GLOBAL_CTL		= 0x48, /* Global control/status (per port) */
 	PDC_CTLSTAT		= 0x60,	/* IDE control and status (per port) */
 	PDC_SATA_PLUG_CSR	= 0x6C, /* SATA Plug control/status reg */
 	PDC2_SATA_PLUG_CSR	= 0x60, /* SATAII Plug control/status reg */
-	PDC_SLEW_CTL		= 0x470, /* slew rate control reg */
+	PDC_TBG_MODE		= 0x41C, /* TBG mode (not SATAII) */
+	PDC_SLEW_CTL		= 0x470, /* slew rate control reg (not SATAII) */
 
 	PDC_ERR_MASK		= (1<<19) | (1<<20) | (1<<21) | (1<<22) |
 				  (1<<8) | (1<<9) | (1<<10),
@@ -88,6 +88,7 @@ struct pdc_port_priv {
 
 struct pdc_host_priv {
 	int			hotplug_offset;
+	int			no_tbg_slew_init;
 };
 
 static u32 pdc_sata_scr_read (struct ata_port *ap, unsigned int sc_reg);
@@ -235,20 +236,20 @@ static const struct ata_port_info pdc_po
 
 static const struct pci_device_id pdc_ata_pci_tbl[] = {
 	{ PCI_VDEVICE(PROMISE, 0x3371), board_2037x },
-	{ PCI_VDEVICE(PROMISE, 0x3570), board_2037x },
-	{ PCI_VDEVICE(PROMISE, 0x3571), board_2037x },
 	{ PCI_VDEVICE(PROMISE, 0x3373), board_2037x },
 	{ PCI_VDEVICE(PROMISE, 0x3375), board_2037x },
 	{ PCI_VDEVICE(PROMISE, 0x3376), board_2037x },
+	{ PCI_VDEVICE(PROMISE, 0x3570), board_2057x },
+	{ PCI_VDEVICE(PROMISE, 0x3571), board_2057x },
 	{ PCI_VDEVICE(PROMISE, 0x3574), board_2057x },
+	{ PCI_VDEVICE(PROMISE, 0x3d73), board_2057x },
 	{ PCI_VDEVICE(PROMISE, 0x3d75), board_2057x },
-	{ PCI_VDEVICE(PROMISE, 0x3d73), board_2037x },
 
 	{ PCI_VDEVICE(PROMISE, 0x3318), board_20319 },
 	{ PCI_VDEVICE(PROMISE, 0x3319), board_20319 },
 	{ PCI_VDEVICE(PROMISE, 0x3515), board_20319 },
 	{ PCI_VDEVICE(PROMISE, 0x3519), board_20319 },
-	{ PCI_VDEVICE(PROMISE, 0x3d17), board_20319 },
+	{ PCI_VDEVICE(PROMISE, 0x3d17), board_40518 },
 	{ PCI_VDEVICE(PROMISE, 0x3d18), board_40518 },
 
 	{ PCI_VDEVICE(PROMISE, 0x6629), board_20619 },
@@ -653,6 +654,9 @@ static void pdc_host_init(unsigned int c
 	tmp = readl(mmio + hotplug_offset);
 	writel(tmp | 0xff0000, mmio + hotplug_offset);
 
+	if (hp->no_tbg_slew_init)
+		return;
+
 	/* reduce TBG clock to 133 Mhz. */
 	tmp = readl(mmio + PDC_TBG_MODE);
 	tmp &= ~0x30000; /* clear bit 17, 16*/
@@ -746,6 +750,8 @@ static int pdc_ata_init_one (struct pci_
 	/* notice 4-port boards */
 	switch (board_idx) {
 	case board_40518:
+		/* TBG/SLEW have changed and don't need init */
+		hp->no_tbg_slew_init = 1;
 		/* Override hotplug offset for SATAII150 */
 		hp->hotplug_offset = PDC2_SATA_PLUG_CSR;
 		/* Fall through */
@@ -759,15 +765,15 @@ static int pdc_ata_init_one (struct pci_
 		probe_ent->port[3].scr_addr = base + 0x700;
 		break;
 	case board_2057x:
+	case board_20771:
+		/* TBG/SLEW have changed and don't need init */
+		hp->no_tbg_slew_init = 1;
 		/* Override hotplug offset for SATAII150 */
 		hp->hotplug_offset = PDC2_SATA_PLUG_CSR;
 		/* Fall through */
 	case board_2037x:
 		probe_ent->n_ports = 2;
 		break;
-	case board_20771:
-		probe_ent->n_ports = 2;
-		break;
 	case board_20619:
 		probe_ent->n_ports = 4;
 

Re: [PATCH 2.6.19-rc6] sata_promise updates

[Jeff Garzik]

Mikael Pettersson wrote:
> This patch updates the sata_promise driver as follows:
> - Correct typo in definition of PDC_TBG_MODE: it's at 0x41C not 0x41
>   in first-generation chips. This error caused PCI access alignment
>   exceptions on SPARC64, and on all platforms it disabled the expected
>   initialisation of TBG mode.
> - Prevent TBG mode and SLEW rate initialisation in second-generation chips.
>   These two registers have moved, TBG mode has been redefined, and
>   Promise's ulsata2 driver no longer attempts to initialise them.
> - Correct PCI device table so devices 0x3570, 0x3571, and 0x3d73 are
>   marked as 2057x (2nd gen) not 2037x (1st gen).
> - Correct PCI device table so device 0x3d17 is marked as 40518
>   (2nd gen 4 ports) not 20319 (1st gen 4 ports).
> - Correct pdc_ata_init_one() to treat 20771 as a second-generation chip.
>   (The 20771 classification is redundant as it is equivalent to 2057x.)
> 
> Tested on 0x3d75 (2nd gen), 0x3d73 (2nd gen), and 0x3373 (1st gen) chips.
> The information comes from the newly uploaded Promise SATA HW specs,
> Promise's ultra and ulsata2 drivers, and debugging on 3d75/3d73/3373 chips.
> 
> Signed-off-by: Mikael Pettersson <mikpe@it.uu.se>

Looks pretty decent to me.  Two small nits:

1) no_tbg_slew_init should be a bit flag ("1 << 0") inside a 'flags' 
variable in struct pdc_host_priv.

2) Check pdc_ulsata2 again, I think the flash control register is 
programmed with a different value on SATAI versus SATAII.

	Jeff

Re: [PATCH 2.6.19-rc6] sata_promise updates

[Jeff Garzik]

Jeff Garzik wrote:
> 1) no_tbg_slew_init should be a bit flag ("1 << 0") inside a 'flags' 
> variable in struct pdc_host_priv.


Thinking some more, a 'is_gen_II' bit flag inside a 'flags' struct 
member would more accurately describe what the boolean flag indicates.

Once NCQ support is added, we will probably have to test this flag in 
code unrelated to TBG/slew registers.

	Jeff

Re: [PATCH 2.6.19-rc6] sata_promise updates

[Jeff Garzik]

And......  is there any chance you could be talked into working on these 
sata_promise to-dos:

* conversion to new error handling system (grep for 'freeze', 'thaw', 
'error_handler', replaces 'phy_reset' and 'eng_timeout')

* support for ATAPI devices (just need to set up ATAPI commands via the 
packet format described in the public docs)

* support for command queueing

Command queueing is the most involved task, the others should be 
reasonably straightforward.

Cheers!

	Jeff

Re: [PATCH 2.6.19-rc6] sata_promise updates

[Mikael Pettersson]

On Tue, 21 Nov 2006 16:07:17 -0500, Jeff Garzik wrote:
>Mikael Pettersson wrote:
>> This patch updates the sata_promise driver as follows:
...
>Looks pretty decent to me.  Two small nits:
>
>1) no_tbg_slew_init should be a bit flag ("1 << 0") inside a 'flags' 
>variable in struct pdc_host_priv.
>
>2) Check pdc_ulsata2 again, I think the flash control register is 
>programmed with a different value on SATAI versus SATAII.

Done. Updated patch below.

This patch updates the sata_promise driver as follows:
- Correct typo in definition of PDC_TBG_MODE: it's at 0x41C not 0x41
  in first-generation chips. This error caused PCI access alignment
  exceptions on SPARC64, and on all platforms it disabled the expected
  initialisation of TBG mode.
- Add flags field to struct pdc_host_priv. Define PDC_FLAG_GEN_II
  and use it to distinguish first- and second-generation chips.
- Prevent the FLASH_CTL FIFO_SHD bit from being set to 1 on second-
  generation chips. This matches Promises' ulsata2 driver.
- Prevent TBG mode and SLEW rate initialisation in second-generation chips.
  These two registers have moved, TBG mode has been redefined, and
  Promise's ulsata2 driver no longer attempts to initialise them.
- Correct PCI device table so devices 0x3570, 0x3571, and 0x3d73 are
  marked as 2057x (2nd gen) not 2037x (1st gen).
- Correct PCI device table so device 0x3d17 is marked as 40518
  (2nd gen 4 ports) not 20319 (1st gen 4 ports).
- Correct pdc_ata_init_one() to treat 20771 as a second-generation chip.

Tested on 0x3d75 (2nd gen), 0x3d73 (2nd gen), and 0x3373 (1st gen) chips.
The information comes from the newly uploaded Promise SATA HW specs,
Promise's ultra and ulsata2 drivers, and debugging on 3d75/3d73/3373 chips.

hp->hotplug_offset could now be removed and its value recomputed
in pdc_host_init() using hp->flags, but that would be a cleanup
not a functional change, so I'm ignoring it for now.

Signed-off-by: Mikael Pettersson <mikpe@it.uu.se>

diff -rupN linux-2.6.19-rc6/drivers/ata/sata_promise.c linux-2.6.19-rc6.sata_promise-mikpe2/drivers/ata/sata_promise.c
--- linux-2.6.19-rc6/drivers/ata/sata_promise.c	2006-11-16 19:04:30.000000000 +0100
+++ linux-2.6.19-rc6.sata_promise-mikpe2/drivers/ata/sata_promise.c	2006-11-22 11:30:18.000000000 +0100
@@ -52,14 +52,14 @@
 enum {
 	PDC_PKT_SUBMIT		= 0x40, /* Command packet pointer addr */
 	PDC_INT_SEQMASK		= 0x40,	/* Mask of asserted SEQ INTs */
-	PDC_TBG_MODE		= 0x41,	/* TBG mode */
 	PDC_FLASH_CTL		= 0x44, /* Flash control register */
 	PDC_PCI_CTL		= 0x48, /* PCI control and status register */
 	PDC_GLOBAL_CTL		= 0x48, /* Global control/status (per port) */
 	PDC_CTLSTAT		= 0x60,	/* IDE control and status (per port) */
 	PDC_SATA_PLUG_CSR	= 0x6C, /* SATA Plug control/status reg */
 	PDC2_SATA_PLUG_CSR	= 0x60, /* SATAII Plug control/status reg */
-	PDC_SLEW_CTL		= 0x470, /* slew rate control reg */
+	PDC_TBG_MODE		= 0x41C, /* TBG mode (not SATAII) */
+	PDC_SLEW_CTL		= 0x470, /* slew rate control reg (not SATAII) */
 
 	PDC_ERR_MASK		= (1<<19) | (1<<20) | (1<<21) | (1<<22) |
 				  (1<<8) | (1<<9) | (1<<10),
@@ -78,6 +78,9 @@ enum {
 	PDC_COMMON_FLAGS	= ATA_FLAG_NO_LEGACY | ATA_FLAG_SRST |
 				  ATA_FLAG_MMIO | ATA_FLAG_NO_ATAPI |
 				  ATA_FLAG_PIO_POLLING,
+
+	/* hp->flags bits */
+	PDC_FLAG_GEN_II		= (1 << 0),
 };
 
 
@@ -87,6 +90,7 @@ struct pdc_port_priv {
 };
 
 struct pdc_host_priv {
+	unsigned long		flags;
 	int			hotplug_offset;
 };
 
@@ -235,20 +239,20 @@ static const struct ata_port_info pdc_po
 
 static const struct pci_device_id pdc_ata_pci_tbl[] = {
 	{ PCI_VDEVICE(PROMISE, 0x3371), board_2037x },
-	{ PCI_VDEVICE(PROMISE, 0x3570), board_2037x },
-	{ PCI_VDEVICE(PROMISE, 0x3571), board_2037x },
 	{ PCI_VDEVICE(PROMISE, 0x3373), board_2037x },
 	{ PCI_VDEVICE(PROMISE, 0x3375), board_2037x },
 	{ PCI_VDEVICE(PROMISE, 0x3376), board_2037x },
+	{ PCI_VDEVICE(PROMISE, 0x3570), board_2057x },
+	{ PCI_VDEVICE(PROMISE, 0x3571), board_2057x },
 	{ PCI_VDEVICE(PROMISE, 0x3574), board_2057x },
+	{ PCI_VDEVICE(PROMISE, 0x3d73), board_2057x },
 	{ PCI_VDEVICE(PROMISE, 0x3d75), board_2057x },
-	{ PCI_VDEVICE(PROMISE, 0x3d73), board_2037x },
 
 	{ PCI_VDEVICE(PROMISE, 0x3318), board_20319 },
 	{ PCI_VDEVICE(PROMISE, 0x3319), board_20319 },
 	{ PCI_VDEVICE(PROMISE, 0x3515), board_20319 },
 	{ PCI_VDEVICE(PROMISE, 0x3519), board_20319 },
-	{ PCI_VDEVICE(PROMISE, 0x3d17), board_20319 },
+	{ PCI_VDEVICE(PROMISE, 0x3d17), board_40518 },
 	{ PCI_VDEVICE(PROMISE, 0x3d18), board_40518 },
 
 	{ PCI_VDEVICE(PROMISE, 0x6629), board_20619 },
@@ -640,9 +644,11 @@ static void pdc_host_init(unsigned int c
 	 * "TODO: figure out why we do this"
 	 */
 
-	/* change FIFO_SHD to 8 dwords, enable BMR_BURST */
+	/* enable BMR_BURST, maybe change FIFO_SHD to 8 dwords */
 	tmp = readl(mmio + PDC_FLASH_CTL);
-	tmp |= 0x12000;	/* bit 16 (fifo 8 dw) and 13 (bmr burst?) */
+	tmp |= 0x02000;	/* bit 13 (enable bmr burst) */
+	if (!(hp->flags & PDC_FLAG_GEN_II))
+		tmp |= 0x10000;	/* bit 16 (fifo threshold at 8 dw) */
 	writel(tmp, mmio + PDC_FLASH_CTL);
 
 	/* clear plug/unplug flags for all ports */
@@ -653,6 +659,10 @@ static void pdc_host_init(unsigned int c
 	tmp = readl(mmio + hotplug_offset);
 	writel(tmp | 0xff0000, mmio + hotplug_offset);
 
+	/* don't initialise TBG or SLEW on 2nd generation chips */
+	if (hp->flags & PDC_FLAG_GEN_II)
+		return;
+
 	/* reduce TBG clock to 133 Mhz. */
 	tmp = readl(mmio + PDC_TBG_MODE);
 	tmp &= ~0x30000; /* clear bit 17, 16*/
@@ -746,6 +756,7 @@ static int pdc_ata_init_one (struct pci_
 	/* notice 4-port boards */
 	switch (board_idx) {
 	case board_40518:
+		hp->flags |= PDC_FLAG_GEN_II;
 		/* Override hotplug offset for SATAII150 */
 		hp->hotplug_offset = PDC2_SATA_PLUG_CSR;
 		/* Fall through */
@@ -759,15 +770,14 @@ static int pdc_ata_init_one (struct pci_
 		probe_ent->port[3].scr_addr = base + 0x700;
 		break;
 	case board_2057x:
+	case board_20771:
+		hp->flags |= PDC_FLAG_GEN_II;
 		/* Override hotplug offset for SATAII150 */
 		hp->hotplug_offset = PDC2_SATA_PLUG_CSR;
 		/* Fall through */
 	case board_2037x:
 		probe_ent->n_ports = 2;
 		break;
-	case board_20771:
-		probe_ent->n_ports = 2;
-		break;
 	case board_20619:
 		probe_ent->n_ports = 4;
 

Re: [PATCH 2.6.19-rc6] sata_promise updates

[Jeff Garzik]

Mikael Pettersson wrote:
> On Tue, 21 Nov 2006 16:07:17 -0500, Jeff Garzik wrote:
>> Mikael Pettersson wrote:
>>> This patch updates the sata_promise driver as follows:
> ...
>> Looks pretty decent to me.  Two small nits:
>>
>> 1) no_tbg_slew_init should be a bit flag ("1 << 0") inside a 'flags' 
>> variable in struct pdc_host_priv.
>>
>> 2) Check pdc_ulsata2 again, I think the flash control register is 
>> programmed with a different value on SATAI versus SATAII.
> 
> Done. Updated patch below.
> 
> This patch updates the sata_promise driver as follows:
> - Correct typo in definition of PDC_TBG_MODE: it's at 0x41C not 0x41
>   in first-generation chips. This error caused PCI access alignment
>   exceptions on SPARC64, and on all platforms it disabled the expected
>   initialisation of TBG mode.
> - Add flags field to struct pdc_host_priv. Define PDC_FLAG_GEN_II
>   and use it to distinguish first- and second-generation chips.
> - Prevent the FLASH_CTL FIFO_SHD bit from being set to 1 on second-
>   generation chips. This matches Promises' ulsata2 driver.
> - Prevent TBG mode and SLEW rate initialisation in second-generation chips.
>   These two registers have moved, TBG mode has been redefined, and
>   Promise's ulsata2 driver no longer attempts to initialise them.
> - Correct PCI device table so devices 0x3570, 0x3571, and 0x3d73 are
>   marked as 2057x (2nd gen) not 2037x (1st gen).
> - Correct PCI device table so device 0x3d17 is marked as 40518
>   (2nd gen 4 ports) not 20319 (1st gen 4 ports).
> - Correct pdc_ata_init_one() to treat 20771 as a second-generation chip.
> 
> Tested on 0x3d75 (2nd gen), 0x3d73 (2nd gen), and 0x3373 (1st gen) chips.
> The information comes from the newly uploaded Promise SATA HW specs,
> Promise's ultra and ulsata2 drivers, and debugging on 3d75/3d73/3373 chips.
> 
> hp->hotplug_offset could now be removed and its value recomputed
> in pdc_host_init() using hp->flags, but that would be a cleanup
> not a functional change, so I'm ignoring it for now.
> 
> Signed-off-by: Mikael Pettersson <mikpe@it.uu.se>

applied to libata-dev.git#upstream (queued for 2.6.20)

Re: [PATCH 2.6.19-rc6] sata_promise updates

[Mikael Pettersson]

On Tue, 21 Nov 2006 16:17:18 -0500, Jeff Garzik wrote:
>And......  is there any chance you could be talked into working on these 
>sata_promise to-dos:
>
>* conversion to new error handling system (grep for 'freeze', 'thaw', 
>'error_handler', replaces 'phy_reset' and 'eng_timeout')
>
>* support for ATAPI devices (just need to set up ATAPI commands via the 
>packet format described in the public docs)
>
>* support for command queueing
>
>Command queueing is the most involved task, the others should be 
>reasonably straightforward.

I can try, but be warned that I have near-zero experience with ATA
programming. I'll look into new_eh first, and then maybe ATAPI.
NCQ most likely requires someone who understands ATA better than me.

/Mikael