libata: ignore EH scheduling during initialization

PATCH Applied Tejun Heo 2 participants 2 messages May 1, 2007 – May 11, 2007 libata

Share & tools

Tejun Heo

author

2007-05-01 09:51 UTC #0 lore raw

PATCH 1/1 libata: ignore EH scheduling during initialization

libata enables SCSI host during ATA host activation which happens after IRQ handler is registered and IRQ is enabled. All ATA ports are in frozen state when IRQ is enabled but frozen ports may raise limited number of IRQs after being frozen - IOW, ->freeze() is not responsible for clearing pending IRQs. During normal operation, the IRQ handler is responsible for clearing spurious IRQs on frozen ports and it usually doesn't require any extra code.

Unfortunately, during host initialization, the IRQ handler can end up scheduling EH for a port whose SCSI host isn't initialized yet. This results in OOPS in the SCSI midlayer. This is relatively short window and scheduling EH for probing is the first thing libata does after initialization, so ignoring EH scheduling until initialization is complete solves the problem nicely.

This problem was spotted by Berck E. Nash in the following thread.

http://thread.gmane.org/gmane.linux.kernel/519412

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

Cc Berck E. Nash <[email protected]>

drivers/ata/libata-core.c +2 -0

index 6d0a946..a7219aa 100644
@@ -5897,6 +5897,7 @@ struct ata_port *ata_port_alloc(struct a
 	if (!ap)
 		return NULL;
 
+	ap->pflags |= ATA_PFLAG_INITIALIZING;
 	ap->lock = &host->lock;
 	ap->flags = ATA_FLAG_DISABLED;
 	ap->print_id = -1;

@@ -6265,6 +6266,7 @@ int ata_host_register(struct ata_host *h
 			ehi->action |= ATA_EH_SOFTRESET;
 			ehi->flags |= ATA_EHI_NO_AUTOPSY | ATA_EHI_QUIET;
 
+			ap->pflags &= ~ATA_PFLAG_INITIALIZING;
 			ap->pflags |= ATA_PFLAG_LOADING;
 			ata_port_schedule_eh(ap);

drivers/ata/libata-eh.c +3 -0

index 0dbee55..26260ce 100644
@@ -546,6 +546,9 @@ void ata_port_schedule_eh(struct ata_por
 {
 	WARN_ON(!ap->ops->error_handler);
 
+	if (ap->pflags & ATA_PFLAG_INITIALIZING)
+		return;
+
 	ap->pflags |= ATA_PFLAG_EH_PENDING;
 	scsi_schedule_eh(ap->scsi_host);

include/linux/libata.h +1 -0

index 73b86dd..cdb23a7 100644
@@ -191,6 +191,7 @@ enum {
 	ATA_PFLAG_LOADING	= (1 << 4), /* boot/loading probe */
 	ATA_PFLAG_UNLOADING	= (1 << 5), /* module is unloading */
 	ATA_PFLAG_SCSI_HOTPLUG	= (1 << 6), /* SCSI hotplug scheduled */
+	ATA_PFLAG_INITIALIZING	= (1 << 7), /* being initialized, don't touch */
 
 	ATA_PFLAG_FLUSH_PORT_TASK = (1 << 16), /* flush port task */
 	ATA_PFLAG_SUSPENDED	= (1 << 17), /* port is suspended (power) */

Jeff Garzik

Applied

2007-05-11 22:10 UTC #1 lore raw

Quoted 15 lines · Tejun Heo

libata enables SCSI host during ATA host activation which happens after IRQ handler is registered and IRQ is enabled. All ATA ports are in frozen state when IRQ is enabled but frozen ports may raise limited number of IRQs after being frozen - IOW, ->freeze() is not responsible for clearing pending IRQs. During normal operation, the IRQ handler is responsible for clearing spurious IRQs on frozen ports and it usually doesn't require any extra code.
Unfortunately, during host initialization, the IRQ handler can end up scheduling EH for a port whose SCSI host isn't initialized yet. This results in OOPS in the SCSI midlayer. This is relatively short window and scheduling EH for probing is the first thing libata does after initialization, so ignoring EH scheduling until initialization is complete solves the problem nicely.

Quoted 6 lines

This problem was spotted by Berck E. Nash in the following thread.
http://thread.gmane.org/gmane.linux.kernel/519412
Signed-off-by Tejun Heo <[email protected]>
Cc Berck E. Nash <[email protected]>

applied