[PATCH] libata: Cable strategy

[PATCH] libata: Cable strategy

[Alan Cox]

This isolates the cable strategy into one routine and implements some
changes that together seem to sort the failure reports we are still
getting

- If the controller says it knows, believe it
- Otherwise if *any* drive sees 80wire believe it

The obvious approach of taking the worst drive view doesn't work. Lots of
cases in bugzilla show up as only the slave detecting the cable.

Signed-off-by: Alan Cox <alan@redhat.com>

diff -u --new-file --recursive --exclude-from /usr/src/exclude linux.vanilla-2.6.24-mm1/drivers/ata/libata-core.c linux-2.6.24-mm1/drivers/ata/libata-core.c
--- linux.vanilla-2.6.24-mm1/drivers/ata/libata-core.c	2008-02-06 14:14:39.000000000 +0000
+++ linux-2.6.24-mm1/drivers/ata/libata-core.c	2008-02-06 14:31:28.000000000 +0000
@@ -4258,6 +4258,49 @@
 }
 
 /**
+ *	cable_is_40wire		-	40/80/SATA decider
+ *	@ap: port to consider
+ *
+ *	This function encapsulates the policy for speed management
+ *	in one place. At the moment we don't cache the result but
+ *	there is a good case for setting ap->cbl to the result when
+ *	we are called with unknown cables (and figuring out if it
+ *	impacts hotplug at all).
+ *
+ *	Return 1 if the cable appears to be 40 wire.
+ */
+
+static int cable_is_40wire(struct ata_port *ap)
+{
+	struct ata_link *link;
+	struct ata_device *dev;
+
+	/* If the controller thinks we are 40 wire, we are */
+	if (ap->cbl == ATA_CBL_PATA40)
+		return 1;
+	/* If the controller thinks we are 80 wire, we are */
+	if (ap->cbl == ATA_CBL_PATA80 || ap->cbl == ATA_CBL_SATA)
+		return 0;
+	/* If the controller doesn't know we scan
+	
+	   - Note: We look for all 40 wire detects at this point.
+	     Any 80 wire detect is taken to be 80 wire cable
+	     because
+	     - In many setups only the one drive (slave if present)
+               will give a valid detect
+             - If you have a non detect capable drive you don't
+               want it to colour the choice
+        */
+	ata_port_for_each_link(link, ap) {
+		ata_link_for_each_dev(dev, link) {
+			if (!ata_is_40wire(dev))
+				return 1;
+		}
+	}
+	return 1;
+}
+
+/**
  *	ata_dev_xfermask - Compute supported xfermask of the given device
  *	@dev: Device to compute xfermask for
  *
@@ -4325,10 +4368,7 @@
 	 */
 	if (xfer_mask & (0xF8 << ATA_SHIFT_UDMA))
 		/* UDMA/44 or higher would be available */
-		if ((ap->cbl == ATA_CBL_PATA40) ||
-		    (ata_is_40wire(dev) &&
-		    (ap->cbl == ATA_CBL_PATA_UNK ||
-		     ap->cbl == ATA_CBL_PATA80))) {
+		if (cable_is_40wire(ap)) {
 			ata_dev_printk(dev, KERN_WARNING,
 				 "limited to UDMA/33 due to 40-wire cable\n");
 			xfer_mask &= ~(0xF8 << ATA_SHIFT_UDMA);

Re: [PATCH] libata: Cable strategy

[Chuck Ebbert]

On 02/08/2008 10:21 AM, Alan Cox wrote:
> +		ata_link_for_each_dev(dev, link) {
> +			if (!ata_is_40wire(dev))
> +				return 1;
> +		}

				return 0;  ???

Re: [PATCH] libata: Cable strategy

[Alan Cox]

On Fri, 08 Feb 2008 10:45:21 -0500
Chuck Ebbert <cebbert@redhat.com> wrote:

> On 02/08/2008 10:21 AM, Alan Cox wrote:
> > +		ata_link_for_each_dev(dev, link) {
> > +			if (!ata_is_40wire(dev))
> > +				return 1;
> > +		}
> 
> 				return 0;  ???

Err yes. That'll teach me to clean up the patch and not re-test all the
cases. I'll post a replacement diff