[SCSI] aacraid: initialization timeout

Received from Mark Salyzyn from Adaptec.

In the rare instances where the adapter, or the motherboard, is
misbehaving; driver initialization or shutdown becomes problematic. By
introducing a 3 minute timeout on the first interrupt driven command
during initialization, or the issuance of the adapter shutdown command
during driver unload, we can resolve the lockup problems induced by
common (but rare) hardware misbehaviors.

The timeout during initialization, should it occur, is accompanied by a
message presented to the console and the logs indicating that the user
should inspect and resolve problems with interrupt routing.

Signed-off-by: Mark Haverkamp <markh@osdl.org>
Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>

3 files changed, 31 insertions, 3 deletions

diff --git a/drivers/scsi/aacraid/aachba.c b/drivers/scsi/aacraid/aachba.c
index 85d133c40bd..a0735a247e5 100644
--- a/drivers/scsi/aacraid/aachba.c
+++ b/drivers/scsi/aacraid/aachba.c
@@ -692,7 +692,7 @@ int aac_get_adapter_info(struct aac_dev* dev)
 			 fibptr, 
 			 sizeof(*info),
 			 FsaNormal, 
-			 1, 1, 
+			 -1, 1, /* First `interrupt' command uses special wait */
 			 NULL, 
 			 NULL);
 
diff --git a/drivers/scsi/aacraid/comminit.c b/drivers/scsi/aacraid/comminit.c
index 7f11c8540ea..9e054a509b4 100644
--- a/drivers/scsi/aacraid/comminit.c
+++ b/drivers/scsi/aacraid/comminit.c
@@ -195,7 +195,7 @@ int aac_send_shutdown(struct aac_dev * dev)
 			  fibctx,
 			  sizeof(struct aac_close),
 			  FsaNormal,
-			  1, 1,
+			  -2 /* Timeout silently */, 1,
 			  NULL, NULL);
 
 	if (status == 0)
diff --git a/drivers/scsi/aacraid/commsup.c b/drivers/scsi/aacraid/commsup.c
index 69985b08a27..3b983f3ed96 100644
--- a/drivers/scsi/aacraid/commsup.c
+++ b/drivers/scsi/aacraid/commsup.c
@@ -41,6 +41,7 @@
 #include <scsi/scsi_host.h>
 #include <scsi/scsi_device.h>
 #include <asm/semaphore.h>
+#include <asm/delay.h>
 
 #include "aacraid.h"
 
@@ -541,7 +542,34 @@ int fib_send(u16 command, struct fib * fibptr, unsigned long size,  int priority
     
 	if (wait) {
 		spin_unlock_irqrestore(&fibptr->event_lock, flags);
-		down(&fibptr->event_wait);
+		/* Only set for first known interruptable command */
+		if (wait < 0) {
+			/*
+			 * *VERY* Dangerous to time out a command, the
+			 * assumption is made that we have no hope of
+			 * functioning because an interrupt routing or other
+			 * hardware failure has occurred.
+			 */
+			unsigned long count = 36000000L; /* 3 minutes */
+			unsigned long qflags;
+			while (down_trylock(&fibptr->event_wait)) {
+				if (--count == 0) {
+					spin_lock_irqsave(q->lock, qflags);
+					q->numpending--;
+					list_del(&fibptr->queue);
+					spin_unlock_irqrestore(q->lock, qflags);
+					if (wait == -1) {
+	        				printk(KERN_ERR "aacraid: fib_send: first asynchronous command timed out.\n"
+						  "Usually a result of a PCI interrupt routing problem;\n"
+						  "update mother board BIOS or consider utilizing one of\n"
+						  "the SAFE mode kernel options (acpi, apic etc)\n");
+					}
+					return -ETIMEDOUT;
+				}
+				udelay(5);
+			}
+		} else
+			down(&fibptr->event_wait);
 		if(fibptr->done == 0)
 			BUG();