Macbook Pro SATA Drives not seen in 2.6.25

I have a couple of Macbook Pros. The newest one -- based on the Penryn Core 2 Duo with ICH8 -- will not boot with any 2.6.25 kernel (I have tried 2.6.25 and 2.6.25.3).

It boots fine with 2.6.24.7 as well as with
2.6.26-rc2. It also boots fine with 2.6.22. I have specific reasons why I need to run 2.6.25 so I would appreciate any help here. It seems the SATA drives are detected, but then for each partition during bootup I see:

ata3: SATA link down (SStatus 0 SControl 0)

The same 2.6.25 kernels boot fine on my older Macbook Pro with ICH7 and I do not see this error.

I would be happy to run a git bisect to help identify the issue, but if this problem is already well understood (hey, it's fixed in 2.6.26), I can think of better ways to spend my time.

Please copy me personally on any replies. I often subscribe to the list but it's too much mail right now. Thanks in advance.

Quoted 2 lines · On Tue, 13 May 2008 23:27:21 -0400 AndrewL733 <[email protected]> wrote:

I have a couple of Macbook Pros. The newest one -- based on the Penryn Core 2 Duo with ICH8 -- will not boot with any 2.6.25 kernel (I have tried 2.6.25 and 2.6.25.3).

Let's cc linux-ide.

Quoted 4 lines

It boots fine with 2.6.24.7 as well as with 2.6.26-rc2. It also boots fine with 2.6.22. I have specific reasons why I need to run 2.6.25 so I would appreciate any help here. It seems the SATA drives are detected, but then for each partition during bootup I see:

ata3: SATA link down (SStatus 0 SControl 0)

Quoted 4 lines

The same 2.6.25 kernels boot fine on my older Macbook Pro with ICH7 and I do not see this error.

I would be happy to run a git bisect to help identify the issue, but if this problem is already well understood (hey, it's fixed in 2.6.26), I can think of better ways to spend my time.

The below went into 2.6.25.1 (or will do so). It looks hopful. Can you test it please?

From: Tejun Heo <[email protected]>

commit cb6716c879ecf49e2af344926c6a476821812061 upstream

On certain configurations (certain macbooks), even though all the conditions for SIDPR access described in the datasheet are met, actually reading those registers just returns 0 and have no effect on write. Verify SIDPR is actually working before enabling it.

This is reported by Ryan Roth in bz#10512.

Signed-off-by Tejun Heo <[email protected]>

Cc Ryan Roth <[email protected]>

Signed-off-by Jeff Garzik <[email protected]>

Signed-off-by Greg Kroah-Hartman <[email protected]>

---
 drivers/ata/ata_piix.c |   25 +++++++++++++++++++++++++
 1 file changed, 25 insertions(+)

drivers/ata/ata_piix.c +25 -0

@@ -1531,6 +1531,8 @@ static void __devinit piix_init_sidpr(st
 {
    struct pci_dev *pdev = to_pci_dev(host->dev);
    struct piix_host_priv *hpriv = host->private_data;
+   struct ata_device *dev0 = &host->ports[0]->link.device[0];
+   u32 scontrol;
    int i;
 
    /* check for availability */

@@ -1549,6 +1551,29 @@ static void __devinit piix_init_sidpr(st
        return;
 
    hpriv->sidpr = pcim_iomap_table(pdev)[PIIX_SIDPR_BAR];
+
+   /* SCR access via SIDPR doesn't work on some configurations.
+    * Give it a test drive by inhibiting power save modes which
+    * we'll do anyway.
+    */
+   scontrol = piix_sidpr_read(dev0, SCR_CONTROL);
+
+   /* if IPM is already 3, SCR access is probably working.  Don't
+    * un-inhibit power save modes as BIOS might have inhibited
+    * them for a reason.
+    */
+   if ((scontrol & 0xf00) != 0x300) {
+       scontrol |= 0x300;
+       piix_sidpr_write(dev0, SCR_CONTROL, scontrol);
+       scontrol = piix_sidpr_read(dev0, SCR_CONTROL);
+
+       if ((scontrol & 0xf00) != 0x300) {
+           dev_printk(KERN_INFO, host->dev, "SCR access via "
+                  "SIDPR is available but doesn't work\n");
+           return;
+       }
+   }
+
    host->ports[0]->ops = &piix_sidpr_sata_ops;
    host->ports[1]->ops = &piix_sidpr_sata_ops;
 }

Quoted 3 lines

The below went into 2.6.25.1 (or will do so). It looks hopful. Can you test it please?

Do you mean 2.6.25.4 ? As I said above, I already tried 2.6.25.3 yesterday and it doesn't boot.

Quoted 2 lines · On Tue, 13 May 2008 23:27:21 -0400 AndrewL733 <[email protected]> wrote:

I have a couple of Macbook Pros. The newest one -- based on the Penryn Core 2 Duo with ICH8 -- will not boot with any 2.6.25 kernel (I have tried 2.6.25 and 2.6.25.3).

Let's cc linux-ide.

Quoted 8 lines

It boots fine with 2.6.24.7 as well as with 2.6.26-rc2. It also boots fine with 2.6.22. I have specific reasons why I need to run 2.6.25 so I would appreciate any help here. It seems the SATA drives are detected, but then for each partition during bootup I see:

ata3: SATA link down (SStatus 0 SControl 0)

The same 2.6.25 kernels boot fine on my older Macbook Pro with ICH7 and I do not see this error.

I would be happy to run a git bisect to help identify the issue, but if this problem is already well understood (hey, it's fixed in 2.6.26), I can think of better ways to spend my time.

The below went into 2.6.25.1 (or will do so). It looks hopful. Can you test it please?

Quoted 2 lines

Please copy me personally on any replies. I often subscribe to the list but it's too much mail right now. Thanks in advance.

We do that as a matter of course. Or we should...

From: Tejun Heo <[email protected]>

commit cb6716c879ecf49e2af344926c6a476821812061 upstream

On certain configurations (certain macbooks), even though all the conditions for SIDPR access described in the datasheet are met, actually reading those registers just returns 0 and have no effect on write. Verify SIDPR is actually working before enabling it.

This is reported by Ryan Roth in bz#10512.

Signed-off-by Tejun Heo <[email protected]>

Cc Ryan Roth <[email protected]>

Signed-off-by Jeff Garzik <[email protected]>

Signed-off-by Greg Kroah-Hartman <[email protected]>

---
 drivers/ata/ata_piix.c |   25 +++++++++++++++++++++++++
 1 file changed, 25 insertions(+)

drivers/ata/ata_piix.c +25 -0

@@ -1531,6 +1531,8 @@ static void __devinit piix_init_sidpr(st
 {
    struct pci_dev *pdev = to_pci_dev(host->dev);
    struct piix_host_priv *hpriv = host->private_data;
+   struct ata_device *dev0 = &host->ports[0]->link.device[0];
+   u32 scontrol;
    int i;
 
    /* check for availability */

@@ -1549,6 +1551,29 @@ static void __devinit piix_init_sidpr(st
        return;
 
    hpriv->sidpr = pcim_iomap_table(pdev)[PIIX_SIDPR_BAR];
+
+   /* SCR access via SIDPR doesn't work on some configurations.
+    * Give it a test drive by inhibiting power save modes which
+    * we'll do anyway.
+    */
+   scontrol = piix_sidpr_read(dev0, SCR_CONTROL);
+
+   /* if IPM is already 3, SCR access is probably working.  Don't
+    * un-inhibit power save modes as BIOS might have inhibited
+    * them for a reason.
+    */
+   if ((scontrol & 0xf00) != 0x300) {
+       scontrol |= 0x300;
+       piix_sidpr_write(dev0, SCR_CONTROL, scontrol);
+       scontrol = piix_sidpr_read(dev0, SCR_CONTROL);
+
+       if ((scontrol & 0xf00) != 0x300) {
+           dev_printk(KERN_INFO, host->dev, "SCR access via "
+                  "SIDPR is available but doesn't work\n");
+           return;
+       }
+   }
+
    host->ports[0]->ops = &piix_sidpr_sata_ops;
    host->ports[1]->ops = &piix_sidpr_sata_ops;
 }