From: Bhavana Nagendra <bnagendr@redhat.com>
Date: Mon, 10 Mar 2008 10:35:22 -0400
Subject: [usb] new iaa watchdog timer
Message-id: 20080310143521.8093.16553.sendpatchset@localhost.localdomain
O-Subject: [RHEL5.2 PATCH 1/2] Fix USB stress test issue
Bugzilla: 435670

This patch fixes the USB stress test issue discovered on SB600/700 systems.
It adds a new watchdog timer to the ehci-hcd driver to be used exclusively
for detecting lost or missing IAA notifications.

diff --git a/drivers/usb/host/ehci-dbg.c b/drivers/usb/host/ehci-dbg.c
index 65ac9fe..6557454 100644
--- a/drivers/usb/host/ehci-dbg.c
+++ b/drivers/usb/host/ehci-dbg.c
@@ -754,9 +754,7 @@ show_registers (struct class_device *class_dev, char *buf)
 	}
 
 	if (ehci->reclaim) {
-		temp = scnprintf (next, size, "reclaim qh %p%s\n",
-				ehci->reclaim,
-				ehci->reclaim_ready ? " ready" : "");
+		temp = scnprintf(next, size, "reclaim qh %p\n", ehci->reclaim);
 		size -= temp;
 		next += temp;
 	}
diff --git a/drivers/usb/host/ehci-hcd.c b/drivers/usb/host/ehci-hcd.c
index d63177a..cc44799 100644
--- a/drivers/usb/host/ehci-hcd.c
+++ b/drivers/usb/host/ehci-hcd.c
@@ -111,7 +111,7 @@ static const char	hcd_name [] = "ehci_hcd";
 #define	EHCI_TUNE_MULT_TT	1
 #define	EHCI_TUNE_FLS		2	/* (small) 256 frame schedule */
 
-#define EHCI_IAA_JIFFIES	(HZ/100)	/* arbitrary; ~10 msec */
+#define EHCI_IAA_MSECS         10              /* arbitrary */
 #define EHCI_IO_JIFFIES		(HZ/10)		/* io watchdog > irq_thresh */
 #define EHCI_ASYNC_JIFFIES	(HZ/20)		/* async idle timeout */
 #define EHCI_SHRINK_JIFFIES	(HZ/200)	/* async qh unlink delay */
@@ -254,6 +254,7 @@ static void ehci_quiesce (struct ehci_hcd *ehci)
 
 /*-------------------------------------------------------------------------*/
 
+static void end_unlink_async(struct ehci_hcd *ehci, struct pt_regs *regs);
 static void ehci_work(struct ehci_hcd *ehci, struct pt_regs *regs);
 
 #include "ehci-hub.c"
@@ -263,26 +264,41 @@ static void ehci_work(struct ehci_hcd *ehci, struct pt_regs *regs);
 
 /*-------------------------------------------------------------------------*/
 
-static void ehci_watchdog (unsigned long param)
+static void ehci_iaa_watchdog(unsigned long param)
 {
 	struct ehci_hcd		*ehci = (struct ehci_hcd *) param;
 	unsigned long		flags;
+	u32			status, cmd;
 
 	spin_lock_irqsave (&ehci->lock, flags);
+	WARN_ON(!ehci->reclaim);
 
-	/* lost IAA irqs wedge things badly; seen with a vt8235 */
-	if (ehci->reclaim) {
-		u32		status = readl (&ehci->regs->status);
+	status = readl(&ehci->regs->status);
+	cmd = readl(&ehci->regs->command);
+	ehci_dbg(ehci, "IAA watchdog: status %x cmd %x\n", status, cmd);
 
+	/* lost IAA irqs wedge things badly; seen first with a vt8235 */
+	if (ehci->reclaim) {
 		if (status & STS_IAA) {
 			ehci_vdbg (ehci, "lost IAA\n");
 			COUNT (ehci->stats.lost_iaa);
 			writel (STS_IAA, &ehci->regs->status);
-			ehci->reclaim_ready = 1;
 		}
+		writel(cmd & ~CMD_IAAD, &ehci->regs->command);
+		end_unlink_async(ehci, NULL);
 	}
 
- 	/* stop async processing after it's idled a bit */
+	spin_unlock_irqrestore(&ehci->lock, flags);
+}
+
+static void ehci_watchdog(unsigned long param)
+{
+	struct ehci_hcd         *ehci = (struct ehci_hcd *) param;
+	unsigned long           flags;
+
+	spin_lock_irqsave(&ehci->lock, flags);
+
+	/* stop async processing after it's idled a bit */
 	if (test_bit (TIMER_ASYNC_OFF, &ehci->actions))
  		start_unlink_async (ehci, ehci->async);
 
@@ -334,8 +350,6 @@ static void ehci_port_power (struct ehci_hcd *ehci, int is_on)
 static void ehci_work (struct ehci_hcd *ehci, struct pt_regs *regs)
 {
 	timer_action_done (ehci, TIMER_IO_WATCHDOG);
-	if (ehci->reclaim_ready)
-		end_unlink_async (ehci, regs);
 
 	/* another CPU may drop ehci->lock during a schedule scan while
 	 * it reports urb completions.  this flag guards against bogus
@@ -370,6 +384,7 @@ static void ehci_stop (struct usb_hcd *hcd)
 
 	/* no more interrupts ... */
 	del_timer_sync (&ehci->watchdog);
+	del_timer_sync(&ehci->iaa_watchdog);
 
 	spin_lock_irq(&ehci->lock);
 	if (HC_IS_RUNNING (hcd->state))
@@ -417,6 +432,10 @@ static int ehci_init(struct usb_hcd *hcd)
 	ehci->watchdog.function = ehci_watchdog;
 	ehci->watchdog.data = (unsigned long) ehci;
 
+	init_timer(&ehci->iaa_watchdog);
+	ehci->iaa_watchdog.function = ehci_iaa_watchdog;
+	ehci->iaa_watchdog.data = (unsigned long) ehci;
+
 	/*
 	 * hw default: 1K periodic list heads, one per frame.
 	 * periodic_size can shrink by USBCMD update if hcc_params allows.
@@ -433,7 +452,6 @@ static int ehci_init(struct usb_hcd *hcd)
 		ehci->i_thresh = 2 + HCC_ISOC_THRES(hcc_params);
 
 	ehci->reclaim = NULL;
-	ehci->reclaim_ready = 0;
 	ehci->next_uframe = -1;
 
 	/*
@@ -611,8 +629,7 @@ static irqreturn_t ehci_irq (struct usb_hcd *hcd, struct pt_regs *regs)
 	/* complete the unlinking of some qh [4.15.2.3] */
 	if (status & STS_IAA) {
 		COUNT (ehci->stats.reclaim);
-		ehci->reclaim_ready = 1;
-		bh = 1;
+		end_unlink_async(ehci, NULL);
 	}
 
 	/* remote wakeup [4.3.1] */
@@ -715,10 +732,16 @@ static int ehci_urb_enqueue (
 
 static void unlink_async (struct ehci_hcd *ehci, struct ehci_qh *qh)
 {
-	/* if we need to use IAA and it's busy, defer */
-	if (qh->qh_state == QH_STATE_LINKED
-			&& ehci->reclaim
-			&& HC_IS_RUNNING (ehci_to_hcd(ehci)->state)) {
+	/* failfast */
+	if (!HC_IS_RUNNING(ehci_to_hcd(ehci)->state))
+		end_unlink_async(ehci, NULL);
+
+	/* if it's not linked then there's nothing to do */
+	if (qh->qh_state != QH_STATE_LINKED)
+		;
+
+	/* defer till later if busy */
+	else if (ehci->reclaim) {
 		struct ehci_qh		*last;
 
 		for (last = ehci->reclaim;
@@ -728,12 +751,8 @@ static void unlink_async (struct ehci_hcd *ehci, struct ehci_qh *qh)
 		qh->qh_state = QH_STATE_UNLINK_WAIT;
 		last->reclaim = qh;
 
-	/* bypass IAA if the hc can't care */
-	} else if (!HC_IS_RUNNING (ehci_to_hcd(ehci)->state) && ehci->reclaim)
-		end_unlink_async (ehci, NULL);
-
-	/* something else might have unlinked the qh by now */
-	if (qh->qh_state == QH_STATE_LINKED)
+	/* start IAA cycle */
+	} else
 		start_unlink_async (ehci, qh);
 }
 
@@ -755,7 +774,19 @@ static int ehci_urb_dequeue (struct usb_hcd *hcd, struct urb *urb)
 		qh = (struct ehci_qh *) urb->hcpriv;
 		if (!qh)
 			break;
-		unlink_async (ehci, qh);
+		switch (qh->qh_state) {
+		case QH_STATE_LINKED:
+		case QH_STATE_COMPLETING:
+			unlink_async(ehci, qh);
+			break;
+		case QH_STATE_UNLINK:
+		case QH_STATE_UNLINK_WAIT:
+			/* already started */
+			break;
+		case QH_STATE_IDLE:
+			WARN_ON(1);
+			break;
+		}
 		break;
 
 	case PIPE_INTERRUPT:
@@ -847,6 +878,7 @@ rescan:
 		unlink_async (ehci, qh);
 		/* FALL THROUGH */
 	case QH_STATE_UNLINK:		/* wait for hw to finish? */
+	case QH_STATE_UNLINK_WAIT:
 idle_timeout:
 		spin_unlock_irqrestore (&ehci->lock, flags);
 		schedule_timeout_uninterruptible(1);
diff --git a/drivers/usb/host/ehci-hub.c b/drivers/usb/host/ehci-hub.c
index d03e3ca..66b66d5 100644
--- a/drivers/usb/host/ehci-hub.c
+++ b/drivers/usb/host/ehci-hub.c
@@ -37,6 +37,8 @@ static int ehci_bus_suspend (struct usb_hcd *hcd)
 
 	if (time_before (jiffies, ehci->next_statechange))
 		msleep(5);
+	del_timer_sync(&ehci->watchdog);
+	del_timer_sync(&ehci->iaa_watchdog);
 
 	port = HCS_N_PORTS (ehci->hcs_params);
 	spin_lock_irq (&ehci->lock);
@@ -48,7 +50,7 @@ static int ehci_bus_suspend (struct usb_hcd *hcd)
 	}
 	ehci->command = readl (&ehci->regs->command);
 	if (ehci->reclaim)
-		ehci->reclaim_ready = 1;
+		end_unlink_async(ehci, NULL);
 	ehci_work(ehci, NULL);
 
 	/* suspend any active/unsuspended ports, maybe allow wakeup */
@@ -72,7 +74,6 @@ static int ehci_bus_suspend (struct usb_hcd *hcd)
 	}
 
 	/* turn off now-idle HC */
-	del_timer_sync (&ehci->watchdog);
 	ehci_halt (ehci);
 	hcd->state = HC_STATE_SUSPENDED;
 
diff --git a/drivers/usb/host/ehci-pci.c b/drivers/usb/host/ehci-pci.c
index cadffac..f1e0315 100644
--- a/drivers/usb/host/ehci-pci.c
+++ b/drivers/usb/host/ehci-pci.c
@@ -297,7 +297,7 @@ restart:
 	/* emptying the schedule aborts any urbs */
 	spin_lock_irq(&ehci->lock);
 	if (ehci->reclaim)
-		ehci->reclaim_ready = 1;
+		end_unlink_async(ehci, NULL);
 	ehci_work(ehci, NULL);
 	spin_unlock_irq(&ehci->lock);
 
diff --git a/drivers/usb/host/ehci-q.c b/drivers/usb/host/ehci-q.c
index e469221..d63dae9 100644
--- a/drivers/usb/host/ehci-q.c
+++ b/drivers/usb/host/ehci-q.c
@@ -967,7 +967,12 @@ static void end_unlink_async (struct ehci_hcd *ehci, struct pt_regs *regs)
 	struct ehci_qh		*qh = ehci->reclaim;
 	struct ehci_qh		*next;
 
-	timer_action_done (ehci, TIMER_IAA_WATCHDOG);
+	iaa_watchdog_done(ehci);
+
+	if (!qh) {
+		WARN_ON(1);
+		return;
+	}
 
 	// qh->hw_next = cpu_to_le32 (qh->qh_dma);
 	qh->qh_state = QH_STATE_IDLE;
@@ -977,7 +982,6 @@ static void end_unlink_async (struct ehci_hcd *ehci, struct pt_regs *regs)
 	/* other unlink(s) may be pending (in QH_STATE_UNLINK_WAIT) */
 	next = qh->reclaim;
 	ehci->reclaim = next;
-	ehci->reclaim_ready = 0;
 	qh->reclaim = NULL;
 
 	qh_completions (ehci, qh, regs);
@@ -1052,11 +1056,10 @@ static void start_unlink_async (struct ehci_hcd *ehci, struct ehci_qh *qh)
 		return;
 	}
 
-	ehci->reclaim_ready = 0;
 	cmd |= CMD_IAAD;
 	writel (cmd, &ehci->regs->command);
 	(void) readl (&ehci->regs->command);
-	timer_action (ehci, TIMER_IAA_WATCHDOG);
+	iaa_watchdog_start(ehci);
 }
 
 /*-------------------------------------------------------------------------*/
diff --git a/drivers/usb/host/ehci.h b/drivers/usb/host/ehci.h
index 679c1cd..60b011f 100644
--- a/drivers/usb/host/ehci.h
+++ b/drivers/usb/host/ehci.h
@@ -58,7 +58,6 @@ struct ehci_hcd {			/* one per controller */
 	/* async schedule support */
 	struct ehci_qh		*async;
 	struct ehci_qh		*reclaim;
-	unsigned		reclaim_ready : 1;
 	unsigned		scanning : 1;
 
 	/* periodic schedule support */
@@ -81,6 +80,7 @@ struct ehci_hcd {			/* one per controller */
 	struct dma_pool		*itd_pool;	/* itd per iso urb */
 	struct dma_pool		*sitd_pool;	/* sitd per split iso urb */
 
+	struct timer_list       iaa_watchdog;
 	struct timer_list	watchdog;
 	struct notifier_block	reboot_notifier;
 	unsigned long		actions;
@@ -115,9 +115,21 @@ static inline struct usb_hcd *ehci_to_hcd (struct ehci_hcd *ehci)
 }
 
 
+static inline void
+iaa_watchdog_start(struct ehci_hcd *ehci)
+{
+	WARN_ON(timer_pending(&ehci->iaa_watchdog));
+	mod_timer(&ehci->iaa_watchdog,
+			jiffies + msecs_to_jiffies(EHCI_IAA_MSECS));
+}
+
+static inline void iaa_watchdog_done(struct ehci_hcd *ehci)
+{
+	del_timer(&ehci->iaa_watchdog);
+}
+
 enum ehci_timer_action {
 	TIMER_IO_WATCHDOG,
-	TIMER_IAA_WATCHDOG,
 	TIMER_ASYNC_SHRINK,
 	TIMER_ASYNC_OFF,
 };
@@ -135,9 +147,6 @@ timer_action (struct ehci_hcd *ehci, enum ehci_timer_action action)
 		unsigned long t;
 
 		switch (action) {
-		case TIMER_IAA_WATCHDOG:
-			t = EHCI_IAA_JIFFIES;
-			break;
 		case TIMER_IO_WATCHDOG:
 			t = EHCI_IO_JIFFIES;
 			break;
@@ -154,8 +163,7 @@ timer_action (struct ehci_hcd *ehci, enum ehci_timer_action action)
 		// async queue SHRINK often precedes IAA.  while it's ready
 		// to go OFF neither can matter, and afterwards the IO
 		// watchdog stops unless there's still periodic traffic.
-		if (action != TIMER_IAA_WATCHDOG
-				&& t > ehci->watchdog.expires
+		if (time_before_eq(t, ehci->watchdog.expires)
 				&& timer_pending (&ehci->watchdog))
 			return;
 		mod_timer (&ehci->watchdog, t);