From: Bhavana Nagendra <bnagendr@redhat.com>
Date: Mon, 10 Mar 2008 10:35:31 -0400
Subject: [usb] fix iaa watchdog notifications
Message-id: 20080310143528.8093.69280.sendpatchset@localhost.localdomain
O-Subject: [RHEL5.2 PATCH 2/2] Fix USB stress test issue - fix IAA watchdog notifications
Bugzilla: 435670

The recent EHCI driver update to split the IAA watchdog timer out from
the other timers made several things work better, but not everything;
and it created a couple new issues in bugzilla.  Ergo this patch:

  - Handle a should-be-rare SMP race between the watchdog firing
    and (very late) IAA interrupts;

  - Remove a shouldn't-have-been-added WARN_ON() test;

  - Guard against one observed OOPS;

  - If this watchdog fires during clean HC shutdown, it should act
    as a NOP instead of interfering with the shutdown sequence;

  - Guard against silicon errata hypothesized by some vendors:
      * IAA status latch broken, but IAAD cleared OK;
      * IAAD wasn't cleared when IAA status got reported;

The WARN_ON is in bugzilla as 10168; the OOPS as 10078.

diff --git a/drivers/usb/host/ehci-hcd.c b/drivers/usb/host/ehci-hcd.c
index cc44799..a32a4d6 100644
--- a/drivers/usb/host/ehci-hcd.c
+++ b/drivers/usb/host/ehci-hcd.c
@@ -268,23 +268,44 @@ 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);
-
-	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");
+	/* Lost IAA irqs wedge things badly; seen first with a vt8235.
+	 * So we need this watchdog, but must protect it against both
+	 * (a) SMP races against real IAA firing and retriggering, and
+	 * (b) clean HC shutdown, when IAA watchdog was pending.
+	 */
+	if (ehci->reclaim
+			&& !timer_pending(&ehci->iaa_watchdog)
+			&& HC_IS_RUNNING(ehci_to_hcd(ehci)->state)) {
+		u32 cmd, status;
+
+		/* If we get here, IAA is *REALLY* late.  It's barely
+		 * conceivable that the system is so busy that CMD_IAAD
+		 * is still legitimately set, so let's be sure it's
+		 * clear before we read STS_IAA.  (The HC should clear
+		 * CMD_IAAD when it sets STS_IAA.)
+		 */
+		cmd = readl(&ehci->regs->command);
+		if (cmd & CMD_IAAD)
+			writel(cmd & ~CMD_IAAD,
+					&ehci->regs->command);
+
+		/* If IAA is set here it either legitimately triggered
+		 * before we cleared IAAD above (but _way_ late, so we'll
+		 * still count it as lost) ... or a silicon erratum:
+		 * - VIA seems to set IAA without triggering the IRQ;
+		 * - IAAD potentially cleared without setting IAA.
+		 */
+		status = readl(&ehci->regs->status);
+		if ((status & STS_IAA) || !(cmd & CMD_IAAD)) {
 			COUNT (ehci->stats.lost_iaa);
 			writel (STS_IAA, &ehci->regs->status);
 		}
-		writel(cmd & ~CMD_IAAD, &ehci->regs->command);
+
+		ehci_vdbg(ehci, "IAA watchdog: status %x cmd %x\n",
+				status, cmd);
 		end_unlink_async(ehci, NULL);
 	}
 
@@ -586,7 +607,7 @@ static int ehci_run (struct usb_hcd *hcd)
 static irqreturn_t ehci_irq (struct usb_hcd *hcd, struct pt_regs *regs)
 {
 	struct ehci_hcd		*ehci = hcd_to_ehci (hcd);
-	u32			status;
+	u32			status, cmd;
 	int			bh;
 
 	spin_lock (&ehci->lock);
@@ -607,7 +628,7 @@ static irqreturn_t ehci_irq (struct usb_hcd *hcd, struct pt_regs *regs)
 
 	/* clear (just) interrupts */
 	writel (status, &ehci->regs->status);
-	readl (&ehci->regs->command);	/* unblock posted write */
+	cmd = readl(&ehci->regs->command);
 	bh = 0;
 
 #ifdef	EHCI_VERBOSE_DEBUG
@@ -628,8 +649,17 @@ 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);
-		end_unlink_async(ehci, NULL);
+		/* guard against (alleged) silicon errata */
+		if (cmd & CMD_IAAD) {
+			writel(cmd & ~CMD_IAAD,
+					&ehci->regs->command);
+			ehci_dbg(ehci, "IAA with IAAD still set?\n");
+		}
+		if (ehci->reclaim) {
+			COUNT(ehci->stats.reclaim);
+			end_unlink_async(ehci, NULL);
+		} else
+			ehci_dbg(ehci, "IAA with nothing to reclaim?\n");
 	}
 
 	/* remote wakeup [4.3.1] */
diff --git a/drivers/usb/host/ehci-q.c b/drivers/usb/host/ehci-q.c
index d63dae9..d0b9d26 100644
--- a/drivers/usb/host/ehci-q.c
+++ b/drivers/usb/host/ehci-q.c
@@ -969,11 +969,6 @@ static void end_unlink_async (struct ehci_hcd *ehci, struct pt_regs *regs)
 
 	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;
 	qh->qh_next.qh = NULL;