On Sat, Jan 21, 2012 at 12:26 PM, Douglas Gilbert [off-list ref] wrote:

On 12-01-21 12:36 AM, Jack Wang wrote:

quoted

quoted

[Jack Wang]
Nice improvement.
Thanks.

quoted

It's difficult to determine which domain_device is triggering error

recovery,

quoted

so convert messages like:

  sas: ex 5001b4da000e703f phy08:T attached: 5001b4da000e7028
  sas: ex 5001b4da000e703f phy09:T attached: 5001b4da000e7029
  ...
  ata7: sas eh calling libata port error handler
  ata8: sas eh calling libata port error handler

...into:

  sas: ex 5001517e85cfefff phy05:T:9 attached: 5001517e85cfefe5 (stp)
  sas: ex 5001517e3b0af0bf phy11:T:8 attached: 5001517e3b0af0ab (stp)
  ...
  sas: ata7: end_device-21:1: dev error handler
  sas: ata8: end_device-20:0:5: dev error handler

which shows attached link rate, device type, and associates a
domain_device with its ata_port id to correlate messages emitted from
libata-eh.

A couple of comments. The "T" in the "phy08:T" stands for
an expander phy's routing attribute being "table" I presume.

Yes.

In smp_utils I'm thinking of changing that to "U" in the case
where that expander (SAS-2 or later) supports table-to-table
routing. The "U" indicates that phy can join other (sibling)
expander phys to become an "enclosure universal port" which
is nirvana for SAS expanders.

Ok, something like the following (please excuse mailer line wrapping):

diff --git a/drivers/scsi/libsas/sas_expander.c

b/drivers/scsi/libsas/sas_expander.c
index f8d941f..4b2ecd3 100644

--- a/drivers/scsi/libsas/sas_expander.c
+++ b/drivers/scsi/libsas/sas_expander.c

@@ -166,7 +166,22 @@ static inline void *alloc_smp_resp(int size)
        return kzalloc(size, GFP_KERNEL);
 }

-/* ---------- Expander configuration ---------- */
+static char sas_route_char(struct domain_device *dev, struct ex_phy *phy)
+{
+       switch (phy->routing_attr) {
+       case TABLE_ROUTING:
+               if (dev->ex_dev.t2t_supp)
+                       return 'U';
+               else
+                       return 'T';
+       case DIRECT_ROUTING:
+               return 'D';
+       case SUBTRACTIVE_ROUTING:
+               return 'S';
+       default:
+               return '?';
+       }
+}

 static void sas_set_ex_phy(struct domain_device *dev, int phy_id,
                           void *disc_resp)

@@ -257,10 +272,8 @@ static void sas_set_ex_phy(struct domain_device

*dev, int phy_id,

        SAS_DPRINTK("ex %016llx phy%02d:%c:%X attached: %016llx (%s)\n",
                    SAS_ADDR(dev->sas_addr), phy->phy_id,
-                   phy->routing_attr == TABLE_ROUTING ? 'T' :
-                   phy->routing_attr == DIRECT_ROUTING ? 'D' :
-                   phy->routing_attr == SUBTRACTIVE_ROUTING ? 'S' : '?',
-                   phy->linkrate, SAS_ADDR(phy->attached_sas_addr), type);
+                   sas_route_char(dev, phy), phy->linkrate,
+                   SAS_ADDR(phy->attached_sas_addr), type);
 }


...can also fix up the same in sas_print_parent_topology_bug

enum sas_dev_type::EDGE_DEV seems a pretty bad name and
should be changed to EXPANDER_DEV as that is more
accurate. Where fanout expanders ever made in SAS-1?
Not many I would expect. In SAS-2++ (SPL, SPL-2) the
"edge" and "fanout" distinction has been dropped and
the numerical value for fanout expanders (3) has been
marked as obsolete.

Ok, that's a cleanup for another patch, but I'll go ahead and
formalize the above routing char fixup and repost the patch as
"libsas: improve debug statements" rather than "libsas: improve ata
debug statements"

--
Dan