sdhci: handle hot-remove

Gracefully handle when the device is suddenly removed. Do a test read
and avoid any further access if that read returns -1.
Signed-off-by: Pierre Ossman <drzeus@drzeus.cx>

drivers/mmc/host/sdhci-pci.c

@@ -47,7 +47,7 @@ struct sdhci_pci_fixes {
 	int			(*probe)(struct sdhci_pci_chip*);
 
 	int			(*probe_slot)(struct sdhci_pci_slot*);
-	void			(*remove_slot)(struct sdhci_pci_slot*);
+	void			(*remove_slot)(struct sdhci_pci_slot*, int);
 
 	int			(*suspend)(struct sdhci_pci_chip*,
 					pm_message_t);
@@ -209,8 +209,11 @@ static int jmicron_probe_slot(struct sdhci_pci_slot *slot)
 	return 0;
 }
 
-static void jmicron_remove_slot(struct sdhci_pci_slot *slot)
+static void jmicron_remove_slot(struct sdhci_pci_slot *slot, int dead)
 {
+	if (dead)
+		return;
+
 	if (slot->chip->pdev->device == PCI_DEVICE_ID_JMICRON_JMB38X_MMC)
 		jmicron_enable_mmc(slot->host, 0);
 }
@@ -540,7 +543,7 @@ static struct sdhci_pci_slot * __devinit sdhci_pci_probe_slot(
 
 remove:
 	if (chip->fixes && chip->fixes->remove_slot)
-		chip->fixes->remove_slot(slot);
+		chip->fixes->remove_slot(slot, 0);
 
 unmap:
 	iounmap(host->ioaddr);
@@ -554,10 +557,18 @@ static struct sdhci_pci_slot * __devinit sdhci_pci_probe_slot(
 
 static void sdhci_pci_remove_slot(struct sdhci_pci_slot *slot)
 {
-	sdhci_remove_host(slot->host);
+	int dead;
+	u32 scratch;
+
+	dead = 0;
+	scratch = readl(slot->host->ioaddr + SDHCI_INT_STATUS);
+	if (scratch == (u32)-1)
+		dead = 1;
+
+	sdhci_remove_host(slot->host, dead);
 
 	if (slot->chip->fixes && slot->chip->fixes->remove_slot)
-		slot->chip->fixes->remove_slot(slot);
+		slot->chip->fixes->remove_slot(slot, dead);
 
 	pci_release_region(slot->chip->pdev, slot->pci_bar);
 

drivers/mmc/host/sdhci.c

@@ -712,7 +712,8 @@ static void sdhci_request(struct mmc_host *mmc, struct mmc_request *mrq)
 
 	host->mrq = mrq;
 
-	if (!(readl(host->ioaddr + SDHCI_PRESENT_STATE) & SDHCI_CARD_PRESENT)) {
+	if (!(readl(host->ioaddr + SDHCI_PRESENT_STATE) & SDHCI_CARD_PRESENT)
+		|| (host->flags & SDHCI_DEVICE_DEAD)) {
 		host->mrq->cmd->error = -ENOMEDIUM;
 		tasklet_schedule(&host->finish_tasklet);
 	} else
@@ -732,6 +733,9 @@ static void sdhci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
 
 	spin_lock_irqsave(&host->lock, flags);
 
+	if (host->flags & SDHCI_DEVICE_DEAD)
+		goto out;
+
 	/*
 	 * Reset the chip on each power off.
 	 * Should clear out any weird states.
@@ -770,6 +774,7 @@ static void sdhci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
 	if(host->quirks & SDHCI_QUIRK_RESET_CMD_DATA_ON_IOS)
 		sdhci_reset(host, SDHCI_RESET_CMD | SDHCI_RESET_DATA);
 
+out:
 	mmiowb();
 	spin_unlock_irqrestore(&host->lock, flags);
 }
@@ -784,6 +789,9 @@ static int sdhci_get_ro(struct mmc_host *mmc)
 
 	spin_lock_irqsave(&host->lock, flags);
 
+	if (host->flags & SDHCI_DEVICE_DEAD)
+		present = 0;
+	else
 		present = readl(host->ioaddr + SDHCI_PRESENT_STATE);
 
 	spin_unlock_irqrestore(&host->lock, flags);
@@ -801,6 +809,9 @@ static void sdhci_enable_sdio_irq(struct mmc_host *mmc, int enable)
 
 	spin_lock_irqsave(&host->lock, flags);
 
+	if (host->flags & SDHCI_DEVICE_DEAD)
+		goto out;
+
 	ier = readl(host->ioaddr + SDHCI_INT_ENABLE);
 
 	ier &= ~SDHCI_INT_CARD_INT;
@@ -810,6 +821,7 @@ static void sdhci_enable_sdio_irq(struct mmc_host *mmc, int enable)
 	writel(ier, host->ioaddr + SDHCI_INT_ENABLE);
 	writel(ier, host->ioaddr + SDHCI_SIGNAL_ENABLE);
 
+out:
 	mmiowb();
 
 	spin_unlock_irqrestore(&host->lock, flags);
@@ -875,10 +887,11 @@ static void sdhci_tasklet_finish(unsigned long param)
 	 * The controller needs a reset of internal state machines
 	 * upon error conditions.
 	 */
-	if (mrq->cmd->error ||
+	if (!(host->flags & SDHCI_DEVICE_DEAD) &&
+		(mrq->cmd->error ||
 		 (mrq->data && (mrq->data->error ||
 		  (mrq->data->stop && mrq->data->stop->error))) ||
-		(host->quirks & SDHCI_QUIRK_RESET_AFTER_REQUEST)) {
+		   (host->quirks & SDHCI_QUIRK_RESET_AFTER_REQUEST))) {
 
 		/* Some controllers need this kick or reset won't work here */
 		if (host->quirks & SDHCI_QUIRK_CLOCK_BEFORE_RESET) {
@@ -1378,14 +1391,33 @@ int sdhci_add_host(struct sdhci_host *host)
 
 EXPORT_SYMBOL_GPL(sdhci_add_host);
 
-void sdhci_remove_host(struct sdhci_host *host)
+void sdhci_remove_host(struct sdhci_host *host, int dead)
 {
+	unsigned long flags;
+
+	if (dead) {
+		spin_lock_irqsave(&host->lock, flags);
+
+		host->flags |= SDHCI_DEVICE_DEAD;
+
+		if (host->mrq) {
+			printk(KERN_ERR "%s: Controller removed during "
+				" transfer!\n", mmc_hostname(host->mmc));
+
+			host->mrq->cmd->error = -ENOMEDIUM;
+			tasklet_schedule(&host->finish_tasklet);
+		}
+
+		spin_unlock_irqrestore(&host->lock, flags);
+	}
+
 	mmc_remove_host(host->mmc);
 
 #ifdef CONFIG_LEDS_CLASS
 	led_classdev_unregister(&host->led);
 #endif
 
+	if (!dead)
 		sdhci_reset(host, SDHCI_RESET_ALL);
 
 	free_irq(host->irq, host);

drivers/mmc/host/sdhci.h

@@ -198,6 +198,7 @@ struct sdhci_host {
 	int			flags;		/* Host attributes */
 #define SDHCI_USE_DMA		(1<<0)		/* Host is DMA capable */
 #define SDHCI_REQ_USE_DMA	(1<<1)		/* Use DMA for this req. */
+#define SDHCI_DEVICE_DEAD	(1<<2)		/* Device unresponsive */
 
 	unsigned int		max_clk;	/* Max possible freq (MHz) */
 	unsigned int		timeout_clk;	/* Timeout freq (KHz) */
@@ -239,7 +240,7 @@ static inline void *sdhci_priv(struct sdhci_host *host)
 }
 
 extern int sdhci_add_host(struct sdhci_host *host);
-extern void sdhci_remove_host(struct sdhci_host *host);
+extern void sdhci_remove_host(struct sdhci_host *host, int dead);
 
 #ifdef CONFIG_PM
 extern int sdhci_suspend_host(struct sdhci_host *host, pm_message_t state);