On Tuesday, May 30, 2006 10:05 PM, James Bottomley wrote:
> +void sas_port_add_phy(struct sas_port *port, struct sas_phy *phy)
> +{
> + BUG_ON(!list_empty(&phy->port_siblings));
> + mutex_lock(&port->phy_list_mutex);
> + sas_port_create_link(port, phy);
> + list_add_tail(&phy->port_siblings, &port->phy_list);
> + port->num_phys++;
> + mutex_unlock(&port->phy_list_mutex);
> +}
> +EXPORT_SYMBOL(sas_port_add_phy);
> +
> +void sas_port_delete_phy(struct sas_port *port, struct sas_phy *phy)
> +{
> + mutex_lock(&port->phy_list_mutex);
> + sas_port_delete_link(port, phy);
> + list_del_init(&port->phy_list);
> + port->num_phys--;
> + mutex_unlock(&port->phy_list_mutex);
> +}
> +EXPORT_SYMBOL(sas_port_delete_phy);
> +
Just want to confirm whether these two API is for handling
adding/deleting phys to the sas_port object once the wide port is
formed?
During the cable push case, I will be reporting a narrow port
first, then later needing to populate the transport sas_port
object with the rest of my phys that form the wide port. It doesn't
all happen all at once, probably requiring about 2 seconds for
each link to come up, thus several calls to my hotplug_work thread.
If that is the case, I will begin adding your new API today into
the fusion driver. I've already added support in the fusion driver
for forming wide ports in sas_topology tree, that done in my internal
source a couple weeks back. I will provide you my sotd ( source of the
day) before I go on vacation.
Eric Moore
-
: send the line "unsubscribe linux-scsi" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
