libata: handle AN interrupt

List:       linux-kernel
Subject:    [patch 3/3] libata: handle AN interrupt
From:       Kristen Carlson Accardi <kristen.c.accardi () intel ! com>
Date:       2007-03-28 23:44:57
Message-ID: 20070328164457.a675a7d2.kristen.c.accardi () intel ! com

[Download message RAW]

When we get an SDB FIS with the 'N' bit set, we should send an event to user space to indicate that there has been a media change. The ahci host controller will send the event via KOBJ_CHANGE uevent.

Signed-off-by Kristen Carlson Accardi <[email protected]>

Index: 2.6-mm/drivers/ata/ahci.c
===================================================================

drivers/ata/ahci.c +20 -0

@@ -1164,6 +1164,26 @@ static void ahci_host_intr(struct ata_po
 		return;
 	}
 
+	if (status & PORT_IRQ_SDB_FIS) {
+		/*
+		 * if this is an ATAPI device with AN turned on,
+		 * then we should interrogate the device to
+		 * determine the cause of the interrupt
+		 *
+		 * for AN - this we should check the SDB FIS
+		 * and find the I and N bits set
+		 */
+		const u32 *f = pp->rx_fis + RX_FIS_SDB;
+
+		/* check the 'N' bit in word 0 of the FIS */
+		if (f[0] & (1 << 15)) {
+			int port_addr =  ((f[0] & 0x00000f00) >> 8);
+			struct ata_device *adev = &ap->device[port_addr];
+			ata_port_printk(ap, KERN_INFO, "N bit set on SDB FIS!\n");
+			if (adev->flags & ATA_DFLAG_AN)
+				ata_async_notify(adev);
+		}
+	}
 	if (ap->sactive)
 		qc_active = readl(port_mmio + PORT_SCR_ACT);
 	else

Index: 2.6-mm/include/linux/libata.h
===================================================================

include/linux/libata.h +2 -0

@@ -492,6 +492,7 @@ struct ata_device {
 	/* ACPI objects info */
 	acpi_handle obj_handle;
 #endif
+	struct work_struct	async_notify;
 };
 
 /* Offset into struct ata_device.  Fields above it are maintained

@@ -826,6 +827,7 @@ extern void ata_scsi_slave_destroy(struc
 extern int ata_scsi_change_queue_depth(struct scsi_device *sdev,
 				       int queue_depth);
 extern struct ata_device *ata_dev_pair(struct ata_device *adev);
+extern void ata_async_notify(struct ata_device *atadev);
 extern int ata_do_set_mode(struct ata_port *ap, struct ata_device **r_failed_dev);
 extern u8 ata_irq_on(struct ata_port *ap);
 extern u8 ata_dummy_irq_on(struct ata_port *ap);

Index: 2.6-mm/drivers/ata/libata-core.c
===================================================================

drivers/ata/libata-core.c +22 -0

@@ -1576,6 +1576,26 @@ static void ata_dev_config_ncq(struct at
 		snprintf(desc, desc_sz, "NCQ (depth %d/%d)", hdepth, ddepth);
 }
 
+static void async_notify_thread(struct work_struct *work)
+{
+	struct ata_device *atadev =
+		container_of(work, struct ata_device, async_notify);
+
+	/*
+	 * TBD - who should send this event?  I couldn't find an
+	 * easy way to map an ata_device to a genhd device, so
+	 * decided maybe the ata host should send the event and
+	 * allow user space to figure out what happened?
+	 */
+	kobject_uevent(&atadev->ap->host->dev->kobj, KOBJ_CHANGE);
+}
+
+void ata_async_notify(struct ata_device *atadev)
+{
+	schedule_work(&atadev->async_notify);
+}
+
+
 /**
 *	ata_dev_configure - Configure the specified ATA/ATAPI device
 *	@dev: Target device to configure

@@ -1761,6 +1781,7 @@ int ata_dev_configure(struct ata_device
 				goto err_out_nosup;
 			}
 			dev->flags |= ATA_DFLAG_AN;
+			INIT_WORK(&dev->async_notify, async_notify_thread);
 		}
 
 		if (ata_id_cdb_intr(dev->id)) {

@@ -6650,6 +6671,7 @@ EXPORT_SYMBOL_GPL(ata_dummy_irq_on);
 EXPORT_SYMBOL_GPL(ata_irq_ack);
 EXPORT_SYMBOL_GPL(ata_dummy_irq_ack);
 EXPORT_SYMBOL_GPL(ata_dev_try_classify);
+EXPORT_SYMBOL_GPL(ata_async_notify);
 
 EXPORT_SYMBOL_GPL(ata_cable_40wire);
 EXPORT_SYMBOL_GPL(ata_cable_80wire);

For FC, we have several async events, and allow for LLDDs to send their own data or augment the generic transport event w/ additional LLDD-data. The infrastructure is implemented generically within the scsi midlayer.
We are using Netlink w/ broadcasts to deliver the events rather than kobject_uevent(). We considered using a variant of kobject events, but the general consensus was we didn't want to wrap transport events into the kobject infrastructure, and using netlink natively allowed for transport data to be sent with the event. Additionally, with the broadcast, we could support any number of tools concurrently listening for change events. To date, this has worked very well. You may want to consider this, especially if the types of events is expected to grow beyond the simple "change" notification.

Fair enough, though I definitely lean towards some use of sysfs / device model for AN-style events specifically. The media change events are generated by the device, not the transport, and we should definitely have an object in the device model that represents the device (if we don't already).

I could see transport-based events being delivered to a different endpoint, because transport events differ from device events.

Jeff