85 files changed, 3017 insertions, 1031 deletions
diff --git a/drivers/acpi/ec.c b/drivers/acpi/ec.c
index 9c52d87d6f04..cbdf031f3c09 100644
--- a/drivers/acpi/ec.c
+++ b/drivers/acpi/ec.c
@@ -424,7 +424,7 @@ static void acpi_ec_gpe_query(void *ec_cxt)
        snprintf(object_name, 8, "_Q%2.2X", value);
-        printk(KERN_INFO PREFIX "evaluating %s\n", object_name);
+        ACPI_DEBUG_PRINT((ACPI_DB_INFO, "Evaluating %s", object_name));
        acpi_evaluate_object(ec->handle, object_name, NULL, NULL);
 }
@@ -1016,8 +1016,8 @@ static int __init acpi_ec_set_intr_mode(char *str)
                acpi_ec_mode = EC_POLL;
        }
        acpi_ec_driver.ops.add = acpi_ec_add;
-        ACPI_DEBUG_PRINT((ACPI_DB_INFO, "EC %s mode.\n",
+        printk(KERN_NOTICE PREFIX "%s mode.\n",
-                          intr ? "interrupt" : "polling"));
+                          intr ? "interrupt" : "polling");
        return 1;
 }
diff --git a/drivers/acpi/processor_core.c b/drivers/acpi/processor_core.c
index 25718fed39f1..5f9496d59ed6 100644
--- a/drivers/acpi/processor_core.c
+++ b/drivers/acpi/processor_core.c
@@ -476,9 +476,6 @@ static int acpi_processor_get_info(struct acpi_processor *pr)
        if (cpu_index == -1) {
                if (ACPI_FAILURE
                    (acpi_processor_hotadd_init(pr->handle, &pr->id))) {
-                        printk(KERN_ERR PREFIX
-                                    "Getting cpuindex for acpiid 0x%x\n",
-                                    pr->acpi_id);
                        return -ENODEV;
                }
        }
diff --git a/drivers/acpi/processor_perflib.c b/drivers/acpi/processor_perflib.c
index 0e60382714bb..5207f9e4b443 100644
--- a/drivers/acpi/processor_perflib.c
+++ b/drivers/acpi/processor_perflib.c
@@ -736,10 +736,6 @@ int acpi_processor_preregister_performance(
        }
 err_ret:
-        if (retval) {
-                ACPI_DEBUG_PRINT((ACPI_DB_ERROR, "Error while parsing _PSD domain information. Assuming no coordination\n"));
-        }
        for_each_possible_cpu(i) {
                pr = processors[i];
                if (!pr || !pr->performance)
diff --git a/drivers/acpi/toshiba_acpi.c b/drivers/acpi/toshiba_acpi.c
index 88aeccbafaaf..d9b651ffcdc0 100644
--- a/drivers/acpi/toshiba_acpi.c
+++ b/drivers/acpi/toshiba_acpi.c
@@ -321,13 +321,16 @@ static int set_lcd_status(struct backlight_device *bd)
 static unsigned long write_lcd(const char *buffer, unsigned long count)
 {
        int value;
-        int ret = count;
+        int ret;
        if (sscanf(buffer, " brightness : %i", &value) == 1 &&
-            value >= 0 && value < HCI_LCD_BRIGHTNESS_LEVELS)
+            value >= 0 && value < HCI_LCD_BRIGHTNESS_LEVELS) {
                ret = set_lcd(value);
-        else
+                if (ret == 0)
+                        ret = count;
+        } else {
                ret = -EINVAL;
+        }
        return ret;
 }
diff --git a/drivers/ata/Kconfig b/drivers/ata/Kconfig
index b34e0a958d0f..da21552d2b1c 100644
--- a/drivers/ata/Kconfig
+++ b/drivers/ata/Kconfig
@@ -381,7 +381,7 @@ config PATA_OPTI
          If unsure, say N.
 config PATA_OPTIDMA
-        tristate "OPTI FireStar PATA support (Veyr Experimental)"
+        tristate "OPTI FireStar PATA support (Very Experimental)"
        depends on PCI && EXPERIMENTAL
        help
          This option enables DMA/PIO support for the later OPTi
diff --git a/drivers/ata/pata_hpt37x.c b/drivers/ata/pata_hpt37x.c
index 47082df7199e..dfb306057cf4 100644
--- a/drivers/ata/pata_hpt37x.c
+++ b/drivers/ata/pata_hpt37x.c
@@ -25,7 +25,7 @@
 #include <linux/libata.h>
 #define DRV_NAME        "pata_hpt37x"
-#define DRV_VERSION     "0.5.1"
+#define DRV_VERSION     "0.5.2"
 struct hpt_clock {
        u8      xfer_speed;
@@ -416,7 +416,7 @@ static const char *bad_ata100_5[] = {
 static unsigned long hpt370_filter(const struct ata_port *ap, struct ata_device *adev, unsigned long mask)
 {
-        if (adev->class != ATA_DEV_ATA) {
+        if (adev->class == ATA_DEV_ATA) {
                if (hpt_dma_blacklisted(adev, "UDMA", bad_ata33))
                        mask &= ~ATA_MASK_UDMA;
                if (hpt_dma_blacklisted(adev, "UDMA100", bad_ata100_5))
@@ -749,7 +749,7 @@ static void hpt37x_bmdma_stop(struct ata_queued_cmd *qc)
 {
        struct ata_port *ap = qc->ap;
        struct pci_dev *pdev = to_pci_dev(ap->host->dev);
-        int mscreg = 0x50 + 2 * ap->port_no;
+        int mscreg = 0x50 + 4 * ap->port_no;
        u8 bwsr_stat, msc_stat;
        pci_read_config_byte(pdev, 0x6A, &bwsr_stat);
diff --git a/drivers/block/pktcdvd.c b/drivers/block/pktcdvd.c
index 7c95c762950f..62462190e07e 100644
--- a/drivers/block/pktcdvd.c
+++ b/drivers/block/pktcdvd.c
@@ -765,47 +765,34 @@ static inline struct bio *pkt_get_list_first(struct bio **list_head, struct bio
 */
 static int pkt_generic_packet(struct pktcdvd_device *pd, struct packet_command *cgc)
 {
-        char sense[SCSI_SENSE_BUFFERSIZE];
+        request_queue_t *q = bdev_get_queue(pd->bdev);
-        request_queue_t *q;
        struct request *rq;
-        DECLARE_COMPLETION_ONSTACK(wait);
+        int ret = 0;
-        int err = 0;
-        q = bdev_get_queue(pd->bdev);
+        rq = blk_get_request(q, (cgc->data_direction == CGC_DATA_WRITE) ?
+                             WRITE : READ, __GFP_WAIT);
+        if (cgc->buflen) {
+                if (blk_rq_map_kern(q, rq, cgc->buffer, cgc->buflen, __GFP_WAIT))
+                        goto out;
+        }
+        rq->cmd_len = COMMAND_SIZE(rq->cmd[0]);
+        memcpy(rq->cmd, cgc->cmd, CDROM_PACKET_SIZE);
+        if (sizeof(rq->cmd) > CDROM_PACKET_SIZE)
+                memset(rq->cmd + CDROM_PACKET_SIZE, 0, sizeof(rq->cmd) - CDROM_PACKET_SIZE);
-        rq = blk_get_request(q, (cgc->data_direction == CGC_DATA_WRITE) ? WRITE : READ,
-                             __GFP_WAIT);
-        rq->errors = 0;
-        rq->rq_disk = pd->bdev->bd_disk;
-        rq->bio = NULL;
-        rq->buffer = NULL;
        rq->timeout = 60*HZ;
-        rq->data = cgc->buffer;
-        rq->data_len = cgc->buflen;
-        rq->sense = sense;
-        memset(sense, 0, sizeof(sense));
-        rq->sense_len = 0;
        rq->cmd_type = REQ_TYPE_BLOCK_PC;
        rq->cmd_flags |= REQ_HARDBARRIER;
        if (cgc->quiet)
                rq->cmd_flags |= REQ_QUIET;
-        memcpy(rq->cmd, cgc->cmd, CDROM_PACKET_SIZE);
-        if (sizeof(rq->cmd) > CDROM_PACKET_SIZE)
-                memset(rq->cmd + CDROM_PACKET_SIZE, 0, sizeof(rq->cmd) - CDROM_PACKET_SIZE);
-        rq->cmd_len = COMMAND_SIZE(rq->cmd[0]);
-        rq->ref_count++;
-        rq->end_io_data = &wait;
-        rq->end_io = blk_end_sync_rq;
-        elv_add_request(q, rq, ELEVATOR_INSERT_BACK, 1);
-        generic_unplug_device(q);
-        wait_for_completion(&wait);
-        if (rq->errors)
-                err = -EIO;
+        blk_execute_rq(rq->q, pd->bdev->bd_disk, rq, 0);
+        ret = rq->errors;
+out:
        blk_put_request(rq);
-        return err;
+        return ret;
 }
 /*
diff --git a/drivers/bluetooth/hci_usb.c b/drivers/bluetooth/hci_usb.c
index aeefec97fdee..6bdf593081d8 100644
--- a/drivers/bluetooth/hci_usb.c
+++ b/drivers/bluetooth/hci_usb.c
@@ -117,10 +117,17 @@ static struct usb_device_id blacklist_ids[] = {
        /* IBM/Lenovo ThinkPad with Broadcom chip */
        { USB_DEVICE(0x0a5c, 0x201e), .driver_info = HCI_WRONG_SCO_MTU },
+        { USB_DEVICE(0x0a5c, 0x2110), .driver_info = HCI_WRONG_SCO_MTU },
        /* ANYCOM Bluetooth USB-200 and USB-250 */
        { USB_DEVICE(0x0a5c, 0x2111), .driver_info = HCI_RESET },
+        /* HP laptop with Broadcom chip */
+        { USB_DEVICE(0x03f0, 0x171d), .driver_info = HCI_WRONG_SCO_MTU },
+        /* Dell laptop with Broadcom chip */
+        { USB_DEVICE(0x413c, 0x8126), .driver_info = HCI_WRONG_SCO_MTU },
        /* Microsoft Wireless Transceiver for Bluetooth 2.0 */
        { USB_DEVICE(0x045e, 0x009c), .driver_info = HCI_RESET },
diff --git a/drivers/char/ip2/i2ellis.h b/drivers/char/ip2/i2ellis.h
index 5eabe47b0bc8..433305062fb8 100644
--- a/drivers/char/ip2/i2ellis.h
+++ b/drivers/char/ip2/i2ellis.h
@@ -606,9 +606,9 @@ static int iiDownloadAll(i2eBordStrPtr, loadHdrStrPtr, int, int);
 // code and returning.
 //
 #define COMPLETE(pB,code) \
-        if(1){ \
+        do { \
                 pB->i2eError = code; \
                 return (code == I2EE_GOOD);\
-        }
+        } while (0)
 #endif   // I2ELLIS_H
diff --git a/drivers/connector/cn_proc.c b/drivers/connector/cn_proc.c
index 3ece69231343..5c9f67f98d10 100644
--- a/drivers/connector/cn_proc.c
+++ b/drivers/connector/cn_proc.c
@@ -28,6 +28,7 @@
 #include <linux/init.h>
 #include <linux/connector.h>
 #include <asm/atomic.h>
+#include <asm/unaligned.h>
 #include <linux/cn_proc.h>
@@ -60,7 +61,7 @@ void proc_fork_connector(struct task_struct *task)
        ev = (struct proc_event*)msg->data;
        get_seq(&msg->seq, &ev->cpu);
        ktime_get_ts(&ts); /* get high res monotonic timestamp */
-        ev->timestamp_ns = timespec_to_ns(&ts);
+        put_unaligned(timespec_to_ns(&ts), (__u64 *)&ev->timestamp_ns);
        ev->what = PROC_EVENT_FORK;
        ev->event_data.fork.parent_pid = task->real_parent->pid;
        ev->event_data.fork.parent_tgid = task->real_parent->tgid;
@@ -88,7 +89,7 @@ void proc_exec_connector(struct task_struct *task)
        ev = (struct proc_event*)msg->data;
        get_seq(&msg->seq, &ev->cpu);
        ktime_get_ts(&ts); /* get high res monotonic timestamp */
-        ev->timestamp_ns = timespec_to_ns(&ts);
+        put_unaligned(timespec_to_ns(&ts), (__u64 *)&ev->timestamp_ns);
        ev->what = PROC_EVENT_EXEC;
        ev->event_data.exec.process_pid = task->pid;
        ev->event_data.exec.process_tgid = task->tgid;
@@ -124,7 +125,7 @@ void proc_id_connector(struct task_struct *task, int which_id)
                return;
        get_seq(&msg->seq, &ev->cpu);
        ktime_get_ts(&ts); /* get high res monotonic timestamp */
-        ev->timestamp_ns = timespec_to_ns(&ts);
+        put_unaligned(timespec_to_ns(&ts), (__u64 *)&ev->timestamp_ns);
        memcpy(&msg->id, &cn_proc_event_id, sizeof(msg->id));
        msg->ack = 0; /* not used */
@@ -146,7 +147,7 @@ void proc_exit_connector(struct task_struct *task)
        ev = (struct proc_event*)msg->data;
        get_seq(&msg->seq, &ev->cpu);
        ktime_get_ts(&ts); /* get high res monotonic timestamp */
-        ev->timestamp_ns = timespec_to_ns(&ts);
+        put_unaligned(timespec_to_ns(&ts), (__u64 *)&ev->timestamp_ns);
        ev->what = PROC_EVENT_EXIT;
        ev->event_data.exit.process_pid = task->pid;
        ev->event_data.exit.process_tgid = task->tgid;
@@ -181,7 +182,7 @@ static void cn_proc_ack(int err, int rcvd_seq, int rcvd_ack)
        ev = (struct proc_event*)msg->data;
        msg->seq = rcvd_seq;
        ktime_get_ts(&ts); /* get high res monotonic timestamp */
-        ev->timestamp_ns = timespec_to_ns(&ts);
+        put_unaligned(timespec_to_ns(&ts), (__u64 *)&ev->timestamp_ns);
        ev->cpu = -1;
        ev->what = PROC_EVENT_NONE;
        ev->event_data.ack.err = err;
diff --git a/drivers/hid/Kconfig b/drivers/hid/Kconfig
index 96d4a0bb2203..ec796ad087df 100644
--- a/drivers/hid/Kconfig
+++ b/drivers/hid/Kconfig
@@ -6,13 +6,21 @@ menu "HID Devices"
 config HID
        tristate "Generic HID support"
+        depends on INPUT
        default y
        ---help---
-          Say Y here if you want generic HID support to connect keyboards,
+          A human interface device (HID) is a type of computer device that
-          mice, joysticks, graphic tablets, or any other HID based devices
+          interacts directly with and takes input from humans. The term "HID"
-          to your computer. You also need to select particular types of
+          most commonly used to refer to the USB-HID specification, but other
-          HID devices you want to compile support for, in the particular
+          devices (such as, but not strictly limited to, Bluetooth) are
-          driver menu (USB, Bluetooth)
+          designed using HID specification (this involves certain keyboards,
+          mice, tablets, etc). This option compiles into kernel the generic
+          HID layer code (parser, usages, etc.), which can then be used by
+          transport-specific HID implementation (like USB or Bluetooth).
+          For docs and specs, see http://www.usb.org/developers/hidpage/
+          If unsure, say Y
 endmenu
diff --git a/drivers/i2c/busses/Kconfig b/drivers/i2c/busses/Kconfig
index e1989f3a2684..9367c4cfe936 100644
--- a/drivers/i2c/busses/Kconfig
+++ b/drivers/i2c/busses/Kconfig
@@ -564,13 +564,4 @@ config I2C_PNX
          This driver can also be built as a module.  If so, the module
          will be called i2c-pnx.
-config I2C_PNX_EARLY
-        bool "Early initialization for I2C on PNXxxxx"
-        depends on I2C_PNX=y
-        help
-          Under certain circumstances one may need to make sure I2C on PNXxxxx
-          is initialized earlier than some other driver that depends on it
-          (for instance, that might be USB in case of PNX4008). With this
-          option turned on you can guarantee that.
 endmenu
diff --git a/drivers/i2c/busses/i2c-mv64xxx.c b/drivers/i2c/busses/i2c-mv64xxx.c
index bbc8e3a7ff55..490173611d6b 100644
--- a/drivers/i2c/busses/i2c-mv64xxx.c
+++ b/drivers/i2c/busses/i2c-mv64xxx.c
@@ -529,6 +529,8 @@ mv64xxx_i2c_probe(struct platform_device *pd)
        platform_set_drvdata(pd, drv_data);
        i2c_set_adapdata(&drv_data->adapter, drv_data);
+        mv64xxx_i2c_hw_init(drv_data);
        if (request_irq(drv_data->irq, mv64xxx_i2c_intr, 0,
                        MV64XXX_I2C_CTLR_NAME, drv_data)) {
                dev_err(&drv_data->adapter.dev,
@@ -542,8 +544,6 @@ mv64xxx_i2c_probe(struct platform_device *pd)
                goto exit_free_irq;
        }
-        mv64xxx_i2c_hw_init(drv_data);
        return 0;
        exit_free_irq:
diff --git a/drivers/i2c/busses/i2c-pnx.c b/drivers/i2c/busses/i2c-pnx.c
index de0bca77e926..17376feb1acc 100644
--- a/drivers/i2c/busses/i2c-pnx.c
+++ b/drivers/i2c/busses/i2c-pnx.c
@@ -305,8 +305,7 @@ static int i2c_pnx_master_rcv(struct i2c_adapter *adap)
        return 0;
 }
-static irqreturn_t
+static irqreturn_t i2c_pnx_interrupt(int irq, void *dev_id)
-i2c_pnx_interrupt(int irq, void *dev_id, struct pt_regs *regs)
 {
        u32 stat, ctl;
        struct i2c_adapter *adap = dev_id;
@@ -699,10 +698,6 @@ MODULE_AUTHOR("Vitaly Wool, Dennis Kovalev <source@mvista.com>");
 MODULE_DESCRIPTION("I2C driver for Philips IP3204-based I2C busses");
 MODULE_LICENSE("GPL");
-#ifdef CONFIG_I2C_PNX_EARLY
 /* We need to make sure I2C is initialized before USB */
 subsys_initcall(i2c_adap_pnx_init);
-#else
-mudule_init(i2c_adap_pnx_init);
-#endif
 module_exit(i2c_adap_pnx_exit);
diff --git a/drivers/i2c/chips/m41t00.c b/drivers/i2c/chips/m41t00.c
index 420377c86422..3fcb646e2073 100644
--- a/drivers/i2c/chips/m41t00.c
+++ b/drivers/i2c/chips/m41t00.c
@@ -209,6 +209,7 @@ m41t00_set(void *arg)
        buf[m41t00_chip->hour] = (buf[m41t00_chip->hour] & ~0x3f) | (hour& 0x3f);
        buf[m41t00_chip->day] = (buf[m41t00_chip->day] & ~0x3f) | (day & 0x3f);
        buf[m41t00_chip->mon] = (buf[m41t00_chip->mon] & ~0x1f) | (mon & 0x1f);
+        buf[m41t00_chip->year] = year;
        if (i2c_master_send(save_client, wbuf, 9) < 0)
                dev_err(&save_client->dev, "m41t00_set: Write error\n");
diff --git a/drivers/i2c/i2c-core.c b/drivers/i2c/i2c-core.c
index 3e31f1d265c9..b05378a3d673 100644
--- a/drivers/i2c/i2c-core.c
+++ b/drivers/i2c/i2c-core.c
@@ -95,16 +95,32 @@ struct device_driver i2c_adapter_driver = {
        .bus = &i2c_bus_type,
 };
+/* ------------------------------------------------------------------------- */
+/* I2C bus adapters -- one roots each I2C or SMBUS segment */
 static void i2c_adapter_class_dev_release(struct class_device *dev)
 {
        struct i2c_adapter *adap = class_dev_to_i2c_adapter(dev);
        complete(&adap->class_dev_released);
 }
+static ssize_t i2c_adapter_show_name(struct class_device *cdev, char *buf)
+{
+        struct i2c_adapter *adap = class_dev_to_i2c_adapter(cdev);
+        return sprintf(buf, "%s\n", adap->name);
+}
+static struct class_device_attribute i2c_adapter_attrs[] = {
+        __ATTR(name, S_IRUGO, i2c_adapter_show_name, NULL),
+        { },
+};
 struct class i2c_adapter_class = {
-        .owner =        THIS_MODULE,
+        .owner                  = THIS_MODULE,
-        .name =         "i2c-adapter",
+        .name                   = "i2c-adapter",
-        .release =      &i2c_adapter_class_dev_release,
+        .class_dev_attrs        = i2c_adapter_attrs,
+        .release                = &i2c_adapter_class_dev_release,
 };
 static ssize_t show_adapter_name(struct device *dev, struct device_attribute *attr, char *buf)
@@ -175,8 +191,12 @@ int i2c_add_adapter(struct i2c_adapter *adap)
         * If the parent pointer is not set up,
         * we add this adapter to the host bus.
         */
-        if (adap->dev.parent == NULL)
+        if (adap->dev.parent == NULL) {
                adap->dev.parent = &platform_bus;
+                printk(KERN_WARNING "**WARNING** I2C adapter driver [%s] "
+                       "forgot to specify physical device; fix it!\n",
+                       adap->name);
+        }
        sprintf(adap->dev.bus_id, "i2c-%d", adap->nr);
        adap->dev.driver = &i2c_adapter_driver;
        adap->dev.release = &i2c_adapter_dev_release;
diff --git a/drivers/ide/pci/atiixp.c b/drivers/ide/pci/atiixp.c
index ffdffb6379ef..524e65de4398 100644
--- a/drivers/ide/pci/atiixp.c
+++ b/drivers/ide/pci/atiixp.c
@@ -46,6 +46,8 @@ static atiixp_ide_timing mdma_timing[] = {
 static int save_mdma_mode[4];
+static DEFINE_SPINLOCK(atiixp_lock);
 /**
 *      atiixp_ratemask         -       compute rate mask for ATIIXP IDE
 *      @drive: IDE drive to compute for
@@ -105,7 +107,7 @@ static int atiixp_ide_dma_host_on(ide_drive_t *drive)
        unsigned long flags;
        u16 tmp16;
-        spin_lock_irqsave(&ide_lock, flags);
+        spin_lock_irqsave(&atiixp_lock, flags);
        pci_read_config_word(dev, ATIIXP_IDE_UDMA_CONTROL, &tmp16);
        if (save_mdma_mode[drive->dn])
@@ -114,7 +116,7 @@ static int atiixp_ide_dma_host_on(ide_drive_t *drive)
                tmp16 |= (1 << drive->dn);
        pci_write_config_word(dev, ATIIXP_IDE_UDMA_CONTROL, tmp16);
-        spin_unlock_irqrestore(&ide_lock, flags);
+        spin_unlock_irqrestore(&atiixp_lock, flags);
        return __ide_dma_host_on(drive);
 }
@@ -125,13 +127,13 @@ static int atiixp_ide_dma_host_off(ide_drive_t *drive)
        unsigned long flags;
        u16 tmp16;
-        spin_lock_irqsave(&ide_lock, flags);
+        spin_lock_irqsave(&atiixp_lock, flags);
        pci_read_config_word(dev, ATIIXP_IDE_UDMA_CONTROL, &tmp16);
        tmp16 &= ~(1 << drive->dn);
        pci_write_config_word(dev, ATIIXP_IDE_UDMA_CONTROL, tmp16);
-        spin_unlock_irqrestore(&ide_lock, flags);
+        spin_unlock_irqrestore(&atiixp_lock, flags);
        return __ide_dma_host_off(drive);
 }
@@ -152,7 +154,7 @@ static void atiixp_tuneproc(ide_drive_t *drive, u8 pio)
        u32 pio_timing_data;
        u16 pio_mode_data;
-        spin_lock_irqsave(&ide_lock, flags);
+        spin_lock_irqsave(&atiixp_lock, flags);
        pci_read_config_word(dev, ATIIXP_IDE_PIO_MODE, &pio_mode_data);
        pio_mode_data &= ~(0x07 << (drive->dn * 4));
@@ -165,7 +167,7 @@ static void atiixp_tuneproc(ide_drive_t *drive, u8 pio)
                 (pio_timing[pio].command_width << (timing_shift + 4));
        pci_write_config_dword(dev, ATIIXP_IDE_PIO_TIMING, pio_timing_data);
-        spin_unlock_irqrestore(&ide_lock, flags);
+        spin_unlock_irqrestore(&atiixp_lock, flags);
 }
 /**
@@ -189,7 +191,7 @@ static int atiixp_speedproc(ide_drive_t *drive, u8 xferspeed)
        speed = ide_rate_filter(atiixp_ratemask(drive), xferspeed);
-        spin_lock_irqsave(&ide_lock, flags);
+        spin_lock_irqsave(&atiixp_lock, flags);
        save_mdma_mode[drive->dn] = 0;
        if (speed >= XFER_UDMA_0) {
@@ -208,7 +210,7 @@ static int atiixp_speedproc(ide_drive_t *drive, u8 xferspeed)
                }
        }
-        spin_unlock_irqrestore(&ide_lock, flags);
+        spin_unlock_irqrestore(&atiixp_lock, flags);
        if (speed >= XFER_SW_DMA_0)
                pio = atiixp_dma_2_pio(speed);
diff --git a/drivers/ide/pci/via82cxxx.c b/drivers/ide/pci/via82cxxx.c
index 61f1a9665a7f..381cc6f101ce 100644
--- a/drivers/ide/pci/via82cxxx.c
+++ b/drivers/ide/pci/via82cxxx.c
@@ -123,7 +123,7 @@ struct via82cxxx_dev
 static void via_set_speed(ide_hwif_t *hwif, u8 dn, struct ide_timing *timing)
 {
        struct pci_dev *dev = hwif->pci_dev;
-        struct via82cxxx_dev *vdev = ide_get_hwifdata(hwif);
+        struct via82cxxx_dev *vdev = pci_get_drvdata(hwif->pci_dev);
        u8 t;
        if (~vdev->via_config->flags & VIA_BAD_AST) {
@@ -162,7 +162,7 @@ static void via_set_speed(ide_hwif_t *hwif, u8 dn, struct ide_timing *timing)
 static int via_set_drive(ide_drive_t *drive, u8 speed)
 {
        ide_drive_t *peer = HWIF(drive)->drives + (~drive->dn & 1);
-        struct via82cxxx_dev *vdev = ide_get_hwifdata(drive->hwif);
+        struct via82cxxx_dev *vdev = pci_get_drvdata(drive->hwif->pci_dev);
        struct ide_timing t, p;
        unsigned int T, UT;
@@ -225,7 +225,7 @@ static void via82cxxx_tune_drive(ide_drive_t *drive, u8 pio)
 static int via82cxxx_ide_dma_check (ide_drive_t *drive)
 {
        ide_hwif_t *hwif = HWIF(drive);
-        struct via82cxxx_dev *vdev = ide_get_hwifdata(hwif);
+        struct via82cxxx_dev *vdev = pci_get_drvdata(hwif->pci_dev);
        u16 w80 = hwif->udma_four;
        u16 speed = ide_find_best_mode(drive,
@@ -262,6 +262,53 @@ static struct via_isa_bridge *via_config_find(struct pci_dev **isa)
        return via_config;
 }
+/*
+ * Check and handle 80-wire cable presence
+ */
+static void __devinit via_cable_detect(struct via82cxxx_dev *vdev, u32 u)
+{
+        int i;
+        switch (vdev->via_config->flags & VIA_UDMA) {
+                case VIA_UDMA_66:
+                        for (i = 24; i >= 0; i -= 8)
+                                if (((u >> (i & 16)) & 8) &&
+                                    ((u >> i) & 0x20) &&
+                                     (((u >> i) & 7) < 2)) {
+                                        /*
+                                         * 2x PCI clock and
+                                         * UDMA w/ < 3T/cycle
+                                         */
+                                        vdev->via_80w |= (1 << (1 - (i >> 4)));
+                                }
+                        break;
+                case VIA_UDMA_100:
+                        for (i = 24; i >= 0; i -= 8)
+                                if (((u >> i) & 0x10) ||
+                                    (((u >> i) & 0x20) &&
+                                     (((u >> i) & 7) < 4))) {
+                                        /* BIOS 80-wire bit or
+                                         * UDMA w/ < 60ns/cycle
+                                         */
+                                        vdev->via_80w |= (1 << (1 - (i >> 4)));
+                                }
+                        break;
+                case VIA_UDMA_133:
+                        for (i = 24; i >= 0; i -= 8)
+                                if (((u >> i) & 0x10) ||
+                                    (((u >> i) & 0x20) &&
+                                     (((u >> i) & 7) < 6))) {
+                                        /* BIOS 80-wire bit or
+                                         * UDMA w/ < 60ns/cycle
+                                         */
+                                        vdev->via_80w |= (1 << (1 - (i >> 4)));
+                                }
+                        break;
+        }
+}
 /**
 *      init_chipset_via82cxxx  -       initialization handler
 *      @dev: PCI device
@@ -274,14 +321,22 @@ static struct via_isa_bridge *via_config_find(struct pci_dev **isa)
 static unsigned int __devinit init_chipset_via82cxxx(struct pci_dev *dev, const char *name)
 {
        struct pci_dev *isa = NULL;
+        struct via82cxxx_dev *vdev;
        struct via_isa_bridge *via_config;
        u8 t, v;
-        unsigned int u;
+        u32 u;
+        vdev = kzalloc(sizeof(*vdev), GFP_KERNEL);
+        if (!vdev) {
+                printk(KERN_ERR "VP_IDE: out of memory :(\n");
+                return -ENOMEM;
+        }
+        pci_set_drvdata(dev, vdev);
        /*
         * Find the ISA bridge to see how good the IDE is.
         */
-        via_config = via_config_find(&isa);
+        vdev->via_config = via_config = via_config_find(&isa);
        /* We checked this earlier so if it fails here deeep badness
           is involved */
@@ -289,16 +344,17 @@ static unsigned int __devinit init_chipset_via82cxxx(struct pci_dev *dev, const
        BUG_ON(!via_config->id);
        /*
-         * Setup or disable Clk66 if appropriate
+         * Detect cable and configure Clk66
         */
+        pci_read_config_dword(dev, VIA_UDMA_TIMING, &u);
+        via_cable_detect(vdev, u);
        if ((via_config->flags & VIA_UDMA) == VIA_UDMA_66) {
                /* Enable Clk66 */
-                pci_read_config_dword(dev, VIA_UDMA_TIMING, &u);
                pci_write_config_dword(dev, VIA_UDMA_TIMING, u|0x80008);
        } else if (via_config->flags & VIA_BAD_CLK66) {
                /* Would cause trouble on 596a and 686 */
-                pci_read_config_dword(dev, VIA_UDMA_TIMING, &u);
                pci_write_config_dword(dev, VIA_UDMA_TIMING, u & ~0x80008);
        }
@@ -367,75 +423,11 @@ static unsigned int __devinit init_chipset_via82cxxx(struct pci_dev *dev, const
        return 0;
 }
-/*
- * Check and handle 80-wire cable presence
- */
-static void __devinit via_cable_detect(struct pci_dev *dev, struct via82cxxx_dev *vdev)
-{
-        unsigned int u;
-        int i;
-        pci_read_config_dword(dev, VIA_UDMA_TIMING, &u);
-        switch (vdev->via_config->flags & VIA_UDMA) {
-                case VIA_UDMA_66:
-                        for (i = 24; i >= 0; i -= 8)
-                                if (((u >> (i & 16)) & 8) &&
-                                    ((u >> i) & 0x20) &&
-                                     (((u >> i) & 7) < 2)) {
-                                        /*
-                                         * 2x PCI clock and
-                                         * UDMA w/ < 3T/cycle
-                                         */
-                                        vdev->via_80w |= (1 << (1 - (i >> 4)));
-                                }
-                        break;
-                case VIA_UDMA_100:
-                        for (i = 24; i >= 0; i -= 8)
-                                if (((u >> i) & 0x10) ||
-                                    (((u >> i) & 0x20) &&
-                                     (((u >> i) & 7) < 4))) {
-                                        /* BIOS 80-wire bit or
-                                         * UDMA w/ < 60ns/cycle
-                                         */
-                                        vdev->via_80w |= (1 << (1 - (i >> 4)));
-                                }
-                        break;
-                case VIA_UDMA_133:
-                        for (i = 24; i >= 0; i -= 8)
-                                if (((u >> i) & 0x10) ||
-                                    (((u >> i) & 0x20) &&
-                                     (((u >> i) & 7) < 6))) {
-                                        /* BIOS 80-wire bit or
-                                         * UDMA w/ < 60ns/cycle
-                                         */
-                                        vdev->via_80w |= (1 << (1 - (i >> 4)));
-                                }
-                        break;
-        }
-}
 static void __devinit init_hwif_via82cxxx(ide_hwif_t *hwif)
 {
-        struct via82cxxx_dev *vdev = kmalloc(sizeof(struct via82cxxx_dev),
+        struct via82cxxx_dev *vdev = pci_get_drvdata(hwif->pci_dev);
-                GFP_KERNEL);
-        struct pci_dev *isa = NULL;
        int i;
-        if (vdev == NULL) {
-                printk(KERN_ERR "VP_IDE: out of memory :(\n");
-                return;
-        }
-        memset(vdev, 0, sizeof(struct via82cxxx_dev));
-        ide_set_hwifdata(hwif, vdev);
-        vdev->via_config = via_config_find(&isa);
-        via_cable_detect(hwif->pci_dev, vdev);
        hwif->autodma = 0;
        hwif->tuneproc = &via82cxxx_tune_drive;
diff --git a/drivers/infiniband/core/cma.c b/drivers/infiniband/core/cma.c
index 533193d4e5df..9e0ab048c878 100644
--- a/drivers/infiniband/core/cma.c
+++ b/drivers/infiniband/core/cma.c
@@ -1088,10 +1088,21 @@ static int cma_iw_handler(struct iw_cm_id *iw_id, struct iw_cm_event *iw_event)
                *sin = iw_event->local_addr;
                sin = (struct sockaddr_in *) &id_priv->id.route.addr.dst_addr;
                *sin = iw_event->remote_addr;
-                if (iw_event->status)
+                switch (iw_event->status) {
-                        event.event = RDMA_CM_EVENT_REJECTED;
+                case 0:
-                else
                        event.event = RDMA_CM_EVENT_ESTABLISHED;
+                        break;
+                case -ECONNRESET:
+                case -ECONNREFUSED:
+                        event.event = RDMA_CM_EVENT_REJECTED;
+                        break;
+                case -ETIMEDOUT:
+                        event.event = RDMA_CM_EVENT_UNREACHABLE;
+                        break;
+                default:
+                        event.event = RDMA_CM_EVENT_CONNECT_ERROR;
+                        break;
+                }
                break;
        case IW_CM_EVENT_ESTABLISHED:
                event.event = RDMA_CM_EVENT_ESTABLISHED;
diff --git a/drivers/infiniband/core/ucma.c b/drivers/infiniband/core/ucma.c
index 81a5cdc5733a..e2e8d329b443 100644
--- a/drivers/infiniband/core/ucma.c
+++ b/drivers/infiniband/core/ucma.c
@@ -209,10 +209,21 @@ static int ucma_event_handler(struct rdma_cm_id *cm_id,
        if (event->event == RDMA_CM_EVENT_CONNECT_REQUEST) {
                if (!ctx->backlog) {
                        ret = -EDQUOT;
+                        kfree(uevent);
                        goto out;
                }
                ctx->backlog--;
+        } else if (!ctx->uid) {
+                /*
+                 * We ignore events for new connections until userspace has set
+                 * their context.  This can only happen if an error occurs on a
+                 * new connection before the user accepts it.  This is okay,
+                 * since the accept will just fail later.
+                 */
+                kfree(uevent);
+                goto out;
        }
        list_add_tail(&uevent->list, &ctx->file->event_list);
        wake_up_interruptible(&ctx->file->poll_wait);
 out:
diff --git a/drivers/infiniband/hw/ehca/ehca_hca.c b/drivers/infiniband/hw/ehca/ehca_hca.c
index e1b618c5f685..b7be950ab47c 100644
--- a/drivers/infiniband/hw/ehca/ehca_hca.c
+++ b/drivers/infiniband/hw/ehca/ehca_hca.c
@@ -50,7 +50,7 @@ int ehca_query_device(struct ib_device *ibdev, struct ib_device_attr *props)
                                              ib_device);
        struct hipz_query_hca *rblock;
-        rblock = ehca_alloc_fw_ctrlblock();
+        rblock = ehca_alloc_fw_ctrlblock(GFP_KERNEL);
        if (!rblock) {
                ehca_err(&shca->ib_device, "Can't allocate rblock memory.");
                return -ENOMEM;
@@ -110,7 +110,7 @@ int ehca_query_port(struct ib_device *ibdev,
                                              ib_device);
        struct hipz_query_port *rblock;
-        rblock = ehca_alloc_fw_ctrlblock();
+        rblock = ehca_alloc_fw_ctrlblock(GFP_KERNEL);
        if (!rblock) {
                ehca_err(&shca->ib_device, "Can't allocate rblock memory.");
                return -ENOMEM;
@@ -179,7 +179,7 @@ int ehca_query_pkey(struct ib_device *ibdev, u8 port, u16 index, u16 *pkey)
                return -EINVAL;
        }
-        rblock = ehca_alloc_fw_ctrlblock();
+        rblock = ehca_alloc_fw_ctrlblock(GFP_KERNEL);
        if (!rblock) {
                ehca_err(&shca->ib_device,  "Can't allocate rblock memory.");
                return -ENOMEM;
@@ -212,7 +212,7 @@ int ehca_query_gid(struct ib_device *ibdev, u8 port,
                return -EINVAL;
        }
-        rblock = ehca_alloc_fw_ctrlblock();
+        rblock = ehca_alloc_fw_ctrlblock(GFP_KERNEL);
        if (!rblock) {
                ehca_err(&shca->ib_device, "Can't allocate rblock memory.");
                return -ENOMEM;
diff --git a/drivers/infiniband/hw/ehca/ehca_irq.c b/drivers/infiniband/hw/ehca/ehca_irq.c
index c3ea746e9045..e7209afb4250 100644
--- a/drivers/infiniband/hw/ehca/ehca_irq.c
+++ b/drivers/infiniband/hw/ehca/ehca_irq.c
@@ -138,7 +138,7 @@ int ehca_error_data(struct ehca_shca *shca, void *data,
        u64 *rblock;
        unsigned long block_count;
-        rblock = ehca_alloc_fw_ctrlblock();
+        rblock = ehca_alloc_fw_ctrlblock(GFP_ATOMIC);
        if (!rblock) {
                ehca_err(&shca->ib_device, "Cannot allocate rblock memory.");
                ret = -ENOMEM;
diff --git a/drivers/infiniband/hw/ehca/ehca_iverbs.h b/drivers/infiniband/hw/ehca/ehca_iverbs.h
index 3720e3032cce..cd7789f0d08e 100644
--- a/drivers/infiniband/hw/ehca/ehca_iverbs.h
+++ b/drivers/infiniband/hw/ehca/ehca_iverbs.h
@@ -180,10 +180,10 @@ int ehca_mmap_register(u64 physical,void **mapped,
 int ehca_munmap(unsigned long addr, size_t len);
 #ifdef CONFIG_PPC_64K_PAGES
-void *ehca_alloc_fw_ctrlblock(void);
+void *ehca_alloc_fw_ctrlblock(gfp_t flags);
 void ehca_free_fw_ctrlblock(void *ptr);
 #else
-#define ehca_alloc_fw_ctrlblock() ((void *) get_zeroed_page(GFP_KERNEL))
+#define ehca_alloc_fw_ctrlblock(flags) ((void *) get_zeroed_page(flags))
 #define ehca_free_fw_ctrlblock(ptr) free_page((unsigned long)(ptr))
 #endif
diff --git a/drivers/infiniband/hw/ehca/ehca_main.c b/drivers/infiniband/hw/ehca/ehca_main.c
index cc47e4c13a18..6574fbbaead5 100644
--- a/drivers/infiniband/hw/ehca/ehca_main.c
+++ b/drivers/infiniband/hw/ehca/ehca_main.c
@@ -106,9 +106,9 @@ static struct timer_list poll_eqs_timer;
 #ifdef CONFIG_PPC_64K_PAGES
 static struct kmem_cache *ctblk_cache = NULL;
-void *ehca_alloc_fw_ctrlblock(void)
+void *ehca_alloc_fw_ctrlblock(gfp_t flags)
 {
-        void *ret = kmem_cache_zalloc(ctblk_cache, GFP_KERNEL);
+        void *ret = kmem_cache_zalloc(ctblk_cache, flags);
        if (!ret)
                ehca_gen_err("Out of memory for ctblk");
        return ret;
@@ -206,7 +206,7 @@ int ehca_sense_attributes(struct ehca_shca *shca)
        u64 h_ret;
        struct hipz_query_hca *rblock;
-        rblock = ehca_alloc_fw_ctrlblock();
+        rblock = ehca_alloc_fw_ctrlblock(GFP_KERNEL);
        if (!rblock) {
                ehca_gen_err("Cannot allocate rblock memory.");
                return -ENOMEM;
@@ -258,7 +258,7 @@ static int init_node_guid(struct ehca_shca *shca)
        int ret = 0;
        struct hipz_query_hca *rblock;
-        rblock = ehca_alloc_fw_ctrlblock();
+        rblock = ehca_alloc_fw_ctrlblock(GFP_KERNEL);
        if (!rblock) {
                ehca_err(&shca->ib_device, "Can't allocate rblock memory.");
                return -ENOMEM;
@@ -469,7 +469,7 @@ static ssize_t  ehca_show_##name(struct device *dev,                       \
                                                                           \
        shca = dev->driver_data;                                           \
                                                                           \
-        rblock = ehca_alloc_fw_ctrlblock();                                \
+        rblock = ehca_alloc_fw_ctrlblock(GFP_KERNEL);                      \
        if (!rblock) {                                                     \
                dev_err(dev, "Can't allocate rblock memory.");             \
                return 0;                                                  \
diff --git a/drivers/infiniband/hw/ehca/ehca_mrmw.c b/drivers/infiniband/hw/ehca/ehca_mrmw.c
index 0a5e2214cc5f..cfb362a1029c 100644
--- a/drivers/infiniband/hw/ehca/ehca_mrmw.c
+++ b/drivers/infiniband/hw/ehca/ehca_mrmw.c
@@ -1013,7 +1013,7 @@ int ehca_reg_mr_rpages(struct ehca_shca *shca,
        u32 i;
        u64 *kpage;
-        kpage = ehca_alloc_fw_ctrlblock();
+        kpage = ehca_alloc_fw_ctrlblock(GFP_KERNEL);
        if (!kpage) {
                ehca_err(&shca->ib_device, "kpage alloc failed");
                ret = -ENOMEM;
@@ -1124,7 +1124,7 @@ inline int ehca_rereg_mr_rereg1(struct ehca_shca *shca,
        ehca_mrmw_map_acl(acl, &hipz_acl);
        ehca_mrmw_set_pgsize_hipz_acl(&hipz_acl);
-        kpage = ehca_alloc_fw_ctrlblock();
+        kpage = ehca_alloc_fw_ctrlblock(GFP_KERNEL);
        if (!kpage) {
                ehca_err(&shca->ib_device, "kpage alloc failed");
                ret = -ENOMEM;
diff --git a/drivers/infiniband/hw/ehca/ehca_qp.c b/drivers/infiniband/hw/ehca/ehca_qp.c
index c6c9cef203e3..34b85556d01e 100644
--- a/drivers/infiniband/hw/ehca/ehca_qp.c
+++ b/drivers/infiniband/hw/ehca/ehca_qp.c
@@ -807,7 +807,7 @@ static int internal_modify_qp(struct ib_qp *ibqp,
        unsigned long spl_flags = 0;
        /* do query_qp to obtain current attr values */
-        mqpcb = ehca_alloc_fw_ctrlblock();
+        mqpcb = ehca_alloc_fw_ctrlblock(GFP_KERNEL);
        if (!mqpcb) {
                ehca_err(ibqp->device, "Could not get zeroed page for mqpcb "
                         "ehca_qp=%p qp_num=%x ", my_qp, ibqp->qp_num);
@@ -1273,7 +1273,7 @@ int ehca_query_qp(struct ib_qp *qp,
                return -EINVAL;
        }
-        qpcb = ehca_alloc_fw_ctrlblock();
+        qpcb = ehca_alloc_fw_ctrlblock(GFP_KERNEL);
        if (!qpcb) {
                ehca_err(qp->device,"Out of memory for qpcb "
                         "ehca_qp=%p qp_num=%x", my_qp, qp->qp_num);
diff --git a/drivers/infiniband/hw/mthca/mthca_cq.c b/drivers/infiniband/hw/mthca/mthca_cq.c
index 283d50b76c3d..1159c8a0f2c5 100644
--- a/drivers/infiniband/hw/mthca/mthca_cq.c
+++ b/drivers/infiniband/hw/mthca/mthca_cq.c
@@ -54,6 +54,10 @@ enum {
        MTHCA_CQ_ENTRY_SIZE = 0x20
 };
+enum {
+        MTHCA_ATOMIC_BYTE_LEN = 8
+};
 /*
 * Must be packed because start is 64 bits but only aligned to 32 bits.
 */
@@ -599,11 +603,11 @@ static inline int mthca_poll_one(struct mthca_dev *dev,
                        break;
                case MTHCA_OPCODE_ATOMIC_CS:
                        entry->opcode    = IB_WC_COMP_SWAP;
-                        entry->byte_len  = be32_to_cpu(cqe->byte_cnt);
+                        entry->byte_len  = MTHCA_ATOMIC_BYTE_LEN;
                        break;
                case MTHCA_OPCODE_ATOMIC_FA:
                        entry->opcode    = IB_WC_FETCH_ADD;
-                        entry->byte_len  = be32_to_cpu(cqe->byte_cnt);
+                        entry->byte_len  = MTHCA_ATOMIC_BYTE_LEN;
                        break;
                case MTHCA_OPCODE_BIND_MW:
                        entry->opcode    = IB_WC_BIND_MW;
diff --git a/drivers/infiniband/hw/mthca/mthca_memfree.c b/drivers/infiniband/hw/mthca/mthca_memfree.c
index 15cc2f6eb475..6b19645d946c 100644
--- a/drivers/infiniband/hw/mthca/mthca_memfree.c
+++ b/drivers/infiniband/hw/mthca/mthca_memfree.c
@@ -232,7 +232,7 @@ void *mthca_table_find(struct mthca_icm_table *table, int obj)
        list_for_each_entry(chunk, &icm->chunk_list, list) {
                for (i = 0; i < chunk->npages; ++i) {
-                        if (chunk->mem[i].length >= offset) {
+                        if (chunk->mem[i].length > offset) {
                                page = chunk->mem[i].page;
                                goto out;
                        }
diff --git a/drivers/infiniband/hw/mthca/mthca_qp.c b/drivers/infiniband/hw/mthca/mthca_qp.c
index d844a2569b47..5f5214c0337d 100644
--- a/drivers/infiniband/hw/mthca/mthca_qp.c
+++ b/drivers/infiniband/hw/mthca/mthca_qp.c
@@ -429,13 +429,18 @@ int mthca_query_qp(struct ib_qp *ibqp, struct ib_qp_attr *qp_attr, int qp_attr_m
 {
        struct mthca_dev *dev = to_mdev(ibqp->device);
        struct mthca_qp *qp = to_mqp(ibqp);
-        int err;
+        int err = 0;
-        struct mthca_mailbox *mailbox;
+        struct mthca_mailbox *mailbox = NULL;
        struct mthca_qp_param *qp_param;
        struct mthca_qp_context *context;
        int mthca_state;
        u8 status;
+        if (qp->state == IB_QPS_RESET) {
+                qp_attr->qp_state = IB_QPS_RESET;
+                goto done;
+        }
        mailbox = mthca_alloc_mailbox(dev, GFP_KERNEL);
        if (IS_ERR(mailbox))
                return PTR_ERR(mailbox);
@@ -454,7 +459,6 @@ int mthca_query_qp(struct ib_qp *ibqp, struct ib_qp_attr *qp_attr, int qp_attr_m
        mthca_state = be32_to_cpu(context->flags) >> 28;
        qp_attr->qp_state            = to_ib_qp_state(mthca_state);
-        qp_attr->cur_qp_state        = qp_attr->qp_state;
        qp_attr->path_mtu            = context->mtu_msgmax >> 5;
        qp_attr->path_mig_state      =
                to_ib_mig_state((be32_to_cpu(context->flags) >> 11) & 0x3);
@@ -464,11 +468,6 @@ int mthca_query_qp(struct ib_qp *ibqp, struct ib_qp_attr *qp_attr, int qp_attr_m
        qp_attr->dest_qp_num         = be32_to_cpu(context->remote_qpn) & 0xffffff;
        qp_attr->qp_access_flags     =
                to_ib_qp_access_flags(be32_to_cpu(context->params2));
-        qp_attr->cap.max_send_wr     = qp->sq.max;
-        qp_attr->cap.max_recv_wr     = qp->rq.max;
-        qp_attr->cap.max_send_sge    = qp->sq.max_gs;
-        qp_attr->cap.max_recv_sge    = qp->rq.max_gs;
-        qp_attr->cap.max_inline_data = qp->max_inline_data;
        if (qp->transport == RC || qp->transport == UC) {
                to_ib_ah_attr(dev, &qp_attr->ah_attr, &context->pri_path);
@@ -495,7 +494,16 @@ int mthca_query_qp(struct ib_qp *ibqp, struct ib_qp_attr *qp_attr, int qp_attr_m
        qp_attr->retry_cnt          = (be32_to_cpu(context->params1) >> 16) & 0x7;
        qp_attr->rnr_retry          = context->pri_path.rnr_retry >> 5;
        qp_attr->alt_timeout        = context->alt_path.ackto >> 3;
-        qp_init_attr->cap           = qp_attr->cap;
+done:
+        qp_attr->cur_qp_state        = qp_attr->qp_state;
+        qp_attr->cap.max_send_wr     = qp->sq.max;
+        qp_attr->cap.max_recv_wr     = qp->rq.max;
+        qp_attr->cap.max_send_sge    = qp->sq.max_gs;
+        qp_attr->cap.max_recv_sge    = qp->rq.max_gs;
+        qp_attr->cap.max_inline_data = qp->max_inline_data;
+        qp_init_attr->cap            = qp_attr->cap;
 out:
        mthca_free_mailbox(dev, mailbox);
diff --git a/drivers/infiniband/ulp/iser/iscsi_iser.c b/drivers/infiniband/ulp/iser/iscsi_iser.c
index 9b2041e25d59..dd221eda3ea6 100644
--- a/drivers/infiniband/ulp/iser/iscsi_iser.c
+++ b/drivers/infiniband/ulp/iser/iscsi_iser.c
@@ -177,7 +177,7 @@ iscsi_iser_mtask_xmit(struct iscsi_conn *conn,
         * - if yes, the mtask is recycled at iscsi_complete_pdu
         * - if no,  the mtask is recycled at iser_snd_completion
         */
-        if (error && error != -EAGAIN)
+        if (error && error != -ENOBUFS)
                iscsi_conn_failure(conn, ISCSI_ERR_CONN_FAILED);
        return error;
@@ -241,7 +241,7 @@ iscsi_iser_ctask_xmit(struct iscsi_conn *conn,
                error = iscsi_iser_ctask_xmit_unsol_data(conn, ctask);
 iscsi_iser_ctask_xmit_exit:
-        if (error && error != -EAGAIN)
+        if (error && error != -ENOBUFS)
                iscsi_conn_failure(conn, ISCSI_ERR_CONN_FAILED);
        return error;
 }
diff --git a/drivers/infiniband/ulp/iser/iser_initiator.c b/drivers/infiniband/ulp/iser/iser_initiator.c
index e73c87b9be43..0a7d1ab60e6d 100644
--- a/drivers/infiniband/ulp/iser/iser_initiator.c
+++ b/drivers/infiniband/ulp/iser/iser_initiator.c
@@ -304,18 +304,14 @@ int iser_conn_set_full_featured_mode(struct iscsi_conn *conn)
 static int
 iser_check_xmit(struct iscsi_conn *conn, void *task)
 {
-        int rc = 0;
        struct iscsi_iser_conn *iser_conn = conn->dd_data;
-        write_lock_bh(conn->recv_lock);
        if (atomic_read(&iser_conn->ib_conn->post_send_buf_count) ==
            ISER_QP_MAX_REQ_DTOS) {
-                iser_dbg("%ld can't xmit task %p, suspending tx\n",jiffies,task);
+                iser_dbg("%ld can't xmit task %p\n",jiffies,task);
-                set_bit(ISCSI_SUSPEND_BIT, &conn->suspend_tx);
+                return -ENOBUFS;
-                rc = -EAGAIN;
        }
-        write_unlock_bh(conn->recv_lock);
+        return 0;
-        return rc;
 }
@@ -340,7 +336,7 @@ int iser_send_command(struct iscsi_conn     *conn,
                return -EPERM;
        }
        if (iser_check_xmit(conn, ctask))
-                return -EAGAIN;
+                return -ENOBUFS;
        edtl = ntohl(hdr->data_length);
@@ -426,7 +422,7 @@ int iser_send_data_out(struct iscsi_conn     *conn,
        }
        if (iser_check_xmit(conn, ctask))
-                return -EAGAIN;
+                return -ENOBUFS;
        itt = ntohl(hdr->itt);
        data_seg_len = ntoh24(hdr->dlength);
@@ -498,7 +494,7 @@ int iser_send_control(struct iscsi_conn *conn,
        }
        if (iser_check_xmit(conn,mtask))
-                return -EAGAIN;
+                return -ENOBUFS;
        /* build the tx desc regd header and add it to the tx desc dto */
        mdesc->type = ISCSI_TX_CONTROL;
@@ -605,6 +601,7 @@ void iser_snd_completion(struct iser_desc *tx_desc)
        struct iscsi_iser_conn *iser_conn = ib_conn->iser_conn;
        struct iscsi_conn      *conn = iser_conn->iscsi_conn;
        struct iscsi_mgmt_task *mtask;
+        int resume_tx = 0;
        iser_dbg("Initiator, Data sent dto=0x%p\n", dto);
@@ -613,15 +610,16 @@ void iser_snd_completion(struct iser_desc *tx_desc)
        if (tx_desc->type == ISCSI_TX_DATAOUT)
                kmem_cache_free(ig.desc_cache, tx_desc);
+        if (atomic_read(&iser_conn->ib_conn->post_send_buf_count) ==
+            ISER_QP_MAX_REQ_DTOS)
+                resume_tx = 1;
        atomic_dec(&ib_conn->post_send_buf_count);
-        write_lock(conn->recv_lock);
+        if (resume_tx) {
-        if (conn->suspend_tx) {
                iser_dbg("%ld resuming tx\n",jiffies);
-                clear_bit(ISCSI_SUSPEND_BIT, &conn->suspend_tx);
                scsi_queue_work(conn->session->host, &conn->xmitwork);
        }
-        write_unlock(conn->recv_lock);
        if (tx_desc->type == ISCSI_TX_CONTROL) {
                /* this arithmetic is legal by libiscsi dd_data allocation */
diff --git a/drivers/kvm/kvm.h b/drivers/kvm/kvm.h
index 100df6f38d92..91e0c75aca8f 100644
--- a/drivers/kvm/kvm.h
+++ b/drivers/kvm/kvm.h
@@ -52,6 +52,8 @@
 #define KVM_MAX_VCPUS 1
 #define KVM_MEMORY_SLOTS 4
 #define KVM_NUM_MMU_PAGES 256
+#define KVM_MIN_FREE_MMU_PAGES 5
+#define KVM_REFILL_PAGES 25
 #define FX_IMAGE_SIZE 512
 #define FX_IMAGE_ALIGN 16
@@ -89,14 +91,54 @@ typedef unsigned long  hva_t;
 typedef u64            hpa_t;
 typedef unsigned long  hfn_t;
+#define NR_PTE_CHAIN_ENTRIES 5
+struct kvm_pte_chain {
+        u64 *parent_ptes[NR_PTE_CHAIN_ENTRIES];
+        struct hlist_node link;
+};
+/*
+ * kvm_mmu_page_role, below, is defined as:
+ *
+ *   bits 0:3 - total guest paging levels (2-4, or zero for real mode)
+ *   bits 4:7 - page table level for this shadow (1-4)
+ *   bits 8:9 - page table quadrant for 2-level guests
+ *   bit   16 - "metaphysical" - gfn is not a real page (huge page/real mode)
+ */
+union kvm_mmu_page_role {
+        unsigned word;
+        struct {
+                unsigned glevels : 4;
+                unsigned level : 4;
+                unsigned quadrant : 2;
+                unsigned pad_for_nice_hex_output : 6;
+                unsigned metaphysical : 1;
+        };
+};
 struct kvm_mmu_page {
        struct list_head link;
+        struct hlist_node hash_link;
+        /*
+         * The following two entries are used to key the shadow page in the
+         * hash table.
+         */
+        gfn_t gfn;
+        union kvm_mmu_page_role role;
        hpa_t page_hpa;
        unsigned long slot_bitmap; /* One bit set per slot which has memory
                                    * in this shadow page.
                                    */
        int global;              /* Set if all ptes in this page are global */
-        u64 *parent_pte;
+        int multimapped;         /* More than one parent_pte? */
+        int root_count;          /* Currently serving as active root */
+        union {
+                u64 *parent_pte;               /* !multimapped */
+                struct hlist_head parent_ptes; /* multimapped, kvm_pte_chain */
+        };
 };
 struct vmcs {
@@ -117,14 +159,26 @@ struct kvm_vcpu;
 struct kvm_mmu {
        void (*new_cr3)(struct kvm_vcpu *vcpu);
        int (*page_fault)(struct kvm_vcpu *vcpu, gva_t gva, u32 err);
-        void (*inval_page)(struct kvm_vcpu *vcpu, gva_t gva);
        void (*free)(struct kvm_vcpu *vcpu);
        gpa_t (*gva_to_gpa)(struct kvm_vcpu *vcpu, gva_t gva);
        hpa_t root_hpa;
        int root_level;
        int shadow_root_level;
+        u64 *pae_root;
+};
+#define KVM_NR_MEM_OBJS 20
+struct kvm_mmu_memory_cache {
+        int nobjs;
+        void *objects[KVM_NR_MEM_OBJS];
 };
+/*
+ * We don't want allocation failures within the mmu code, so we preallocate
+ * enough memory for a single page fault in a cache.
+ */
 struct kvm_guest_debug {
        int enabled;
        unsigned long bp[4];
@@ -173,6 +227,7 @@ struct kvm_vcpu {
        struct mutex mutex;
        int   cpu;
        int   launched;
+        int interrupt_window_open;
        unsigned long irq_summary; /* bit vector: 1 per word in irq_pending */
 #define NR_IRQ_WORDS KVM_IRQ_BITMAP_SIZE(unsigned long)
        unsigned long irq_pending[NR_IRQ_WORDS];
@@ -184,6 +239,7 @@ struct kvm_vcpu {
        unsigned long cr3;
        unsigned long cr4;
        unsigned long cr8;
+        u64 pdptrs[4]; /* pae */
        u64 shadow_efer;
        u64 apic_base;
        int nmsrs;
@@ -194,6 +250,12 @@ struct kvm_vcpu {
        struct kvm_mmu_page page_header_buf[KVM_NUM_MMU_PAGES];
        struct kvm_mmu mmu;
+        struct kvm_mmu_memory_cache mmu_pte_chain_cache;
+        struct kvm_mmu_memory_cache mmu_rmap_desc_cache;
+        gfn_t last_pt_write_gfn;
+        int   last_pt_write_count;
        struct kvm_guest_debug guest_debug;
        char fx_buf[FX_BUF_SIZE];
@@ -231,10 +293,16 @@ struct kvm {
        spinlock_t lock; /* protects everything except vcpus */
        int nmemslots;
        struct kvm_memory_slot memslots[KVM_MEMORY_SLOTS];
+        /*
+         * Hash table of struct kvm_mmu_page.
+         */
        struct list_head active_mmu_pages;
+        int n_free_mmu_pages;
+        struct hlist_head mmu_page_hash[KVM_NUM_MMU_PAGES];
        struct kvm_vcpu vcpus[KVM_MAX_VCPUS];
        int memory_config_version;
        int busy;
+        unsigned long rmap_overflow;
 };
 struct kvm_stat {
@@ -247,6 +315,9 @@ struct kvm_stat {
        u32 io_exits;
        u32 mmio_exits;
        u32 signal_exits;
+        u32 irq_window_exits;
+        u32 halt_exits;
+        u32 request_irq_exits;
        u32 irq_exits;
 };
@@ -279,6 +350,7 @@ struct kvm_arch_ops {
        void (*set_segment)(struct kvm_vcpu *vcpu,
                            struct kvm_segment *var, int seg);
        void (*get_cs_db_l_bits)(struct kvm_vcpu *vcpu, int *db, int *l);
+        void (*decache_cr0_cr4_guest_bits)(struct kvm_vcpu *vcpu);
        void (*set_cr0)(struct kvm_vcpu *vcpu, unsigned long cr0);
        void (*set_cr0_no_modeswitch)(struct kvm_vcpu *vcpu,
                                      unsigned long cr0);
@@ -323,7 +395,7 @@ int kvm_mmu_create(struct kvm_vcpu *vcpu);
 int kvm_mmu_setup(struct kvm_vcpu *vcpu);
 int kvm_mmu_reset_context(struct kvm_vcpu *vcpu);
-void kvm_mmu_slot_remove_write_access(struct kvm *kvm, int slot);
+void kvm_mmu_slot_remove_write_access(struct kvm_vcpu *vcpu, int slot);
 hpa_t gpa_to_hpa(struct kvm_vcpu *vcpu, gpa_t gpa);
 #define HPA_MSB ((sizeof(hpa_t) * 8) - 1)
@@ -396,6 +468,19 @@ int kvm_write_guest(struct kvm_vcpu *vcpu,
 unsigned long segment_base(u16 selector);
+void kvm_mmu_pre_write(struct kvm_vcpu *vcpu, gpa_t gpa, int bytes);
+void kvm_mmu_post_write(struct kvm_vcpu *vcpu, gpa_t gpa, int bytes);
+int kvm_mmu_unprotect_page_virt(struct kvm_vcpu *vcpu, gva_t gva);
+void kvm_mmu_free_some_pages(struct kvm_vcpu *vcpu);
+static inline int kvm_mmu_page_fault(struct kvm_vcpu *vcpu, gva_t gva,
+                                     u32 error_code)
+{
+        if (unlikely(vcpu->kvm->n_free_mmu_pages < KVM_MIN_FREE_MMU_PAGES))
+                kvm_mmu_free_some_pages(vcpu);
+        return vcpu->mmu.page_fault(vcpu, gva, error_code);
+}
 static inline struct page *_gfn_to_page(struct kvm *kvm, gfn_t gfn)
 {
        struct kvm_memory_slot *slot = gfn_to_memslot(kvm, gfn);
@@ -541,19 +626,4 @@ static inline u32 get_rdx_init_val(void)
 #define TSS_REDIRECTION_SIZE (256 / 8)
 #define RMODE_TSS_SIZE (TSS_BASE_SIZE + TSS_REDIRECTION_SIZE + TSS_IOPB_SIZE + 1)
-#ifdef CONFIG_X86_64
-/*
- * When emulating 32-bit mode, cr3 is only 32 bits even on x86_64.  Therefore
- * we need to allocate shadow page tables in the first 4GB of memory, which
- * happens to fit the DMA32 zone.
- */
-#define GFP_KVM_MMU (GFP_KERNEL | __GFP_DMA32)
-#else
-#define GFP_KVM_MMU GFP_KERNEL
-#endif
 #endif
diff --git a/drivers/kvm/kvm_main.c b/drivers/kvm/kvm_main.c
index ce7fe640f18d..67c1154960f0 100644
--- a/drivers/kvm/kvm_main.c
+++ b/drivers/kvm/kvm_main.c
@@ -58,6 +58,9 @@ static struct kvm_stats_debugfs_item {
        { "io_exits", &kvm_stat.io_exits },
        { "mmio_exits", &kvm_stat.mmio_exits },
        { "signal_exits", &kvm_stat.signal_exits },
+        { "irq_window", &kvm_stat.irq_window_exits },
+        { "halt_exits", &kvm_stat.halt_exits },
+        { "request_irq", &kvm_stat.request_irq_exits },
        { "irq_exits", &kvm_stat.irq_exits },
        { 0, 0 }
 };
@@ -227,6 +230,7 @@ static int kvm_dev_open(struct inode *inode, struct file *filp)
                struct kvm_vcpu *vcpu = &kvm->vcpus[i];
                mutex_init(&vcpu->mutex);
+                vcpu->kvm = kvm;
                vcpu->mmu.root_hpa = INVALID_PAGE;
                INIT_LIST_HEAD(&vcpu->free_pages);
        }
@@ -268,8 +272,8 @@ static void kvm_free_physmem(struct kvm *kvm)
 static void kvm_free_vcpu(struct kvm_vcpu *vcpu)
 {
-        kvm_arch_ops->vcpu_free(vcpu);
        kvm_mmu_destroy(vcpu);
+        kvm_arch_ops->vcpu_free(vcpu);
 }
 static void kvm_free_vcpus(struct kvm *kvm)
@@ -295,14 +299,17 @@ static void inject_gp(struct kvm_vcpu *vcpu)
        kvm_arch_ops->inject_gp(vcpu, 0);
 }
-static int pdptrs_have_reserved_bits_set(struct kvm_vcpu *vcpu,
+/*
-                                         unsigned long cr3)
+ * Load the pae pdptrs.  Return true is they are all valid.
+ */
+static int load_pdptrs(struct kvm_vcpu *vcpu, unsigned long cr3)
 {
        gfn_t pdpt_gfn = cr3 >> PAGE_SHIFT;
-        unsigned offset = (cr3 & (PAGE_SIZE-1)) >> 5;
+        unsigned offset = ((cr3 & (PAGE_SIZE-1)) >> 5) << 2;
        int i;
        u64 pdpte;
        u64 *pdpt;
+        int ret;
        struct kvm_memory_slot *memslot;
        spin_lock(&vcpu->kvm->lock);
@@ -310,16 +317,23 @@ static int pdptrs_have_reserved_bits_set(struct kvm_vcpu *vcpu,
        /* FIXME: !memslot - emulate? 0xff? */
        pdpt = kmap_atomic(gfn_to_page(memslot, pdpt_gfn), KM_USER0);
+        ret = 1;
        for (i = 0; i < 4; ++i) {
                pdpte = pdpt[offset + i];
-                if ((pdpte & 1) && (pdpte & 0xfffffff0000001e6ull))
+                if ((pdpte & 1) && (pdpte & 0xfffffff0000001e6ull)) {
-                        break;
+                        ret = 0;
+                        goto out;
+                }
        }
+        for (i = 0; i < 4; ++i)
+                vcpu->pdptrs[i] = pdpt[offset + i];
+out:
        kunmap_atomic(pdpt, KM_USER0);
        spin_unlock(&vcpu->kvm->lock);
-        return i != 4;
+        return ret;
 }
 void set_cr0(struct kvm_vcpu *vcpu, unsigned long cr0)
@@ -365,8 +379,7 @@ void set_cr0(struct kvm_vcpu *vcpu, unsigned long cr0)
                        }
                } else
 #endif
-                if (is_pae(vcpu) &&
+                if (is_pae(vcpu) && !load_pdptrs(vcpu, vcpu->cr3)) {
-                            pdptrs_have_reserved_bits_set(vcpu, vcpu->cr3)) {
                        printk(KERN_DEBUG "set_cr0: #GP, pdptrs "
                               "reserved bits\n");
                        inject_gp(vcpu);
@@ -387,6 +400,7 @@ EXPORT_SYMBOL_GPL(set_cr0);
 void lmsw(struct kvm_vcpu *vcpu, unsigned long msw)
 {
+        kvm_arch_ops->decache_cr0_cr4_guest_bits(vcpu);
        set_cr0(vcpu, (vcpu->cr0 & ~0x0ful) | (msw & 0x0f));
 }
 EXPORT_SYMBOL_GPL(lmsw);
@@ -407,7 +421,7 @@ void set_cr4(struct kvm_vcpu *vcpu, unsigned long cr4)
                        return;
                }
        } else if (is_paging(vcpu) && !is_pae(vcpu) && (cr4 & CR4_PAE_MASK)
-                   && pdptrs_have_reserved_bits_set(vcpu, vcpu->cr3)) {
+                   && !load_pdptrs(vcpu, vcpu->cr3)) {
                printk(KERN_DEBUG "set_cr4: #GP, pdptrs reserved bits\n");
                inject_gp(vcpu);
        }
@@ -439,7 +453,7 @@ void set_cr3(struct kvm_vcpu *vcpu, unsigned long cr3)
                        return;
                }
                if (is_paging(vcpu) && is_pae(vcpu) &&
-                    pdptrs_have_reserved_bits_set(vcpu, cr3)) {
+                    !load_pdptrs(vcpu, cr3)) {
                        printk(KERN_DEBUG "set_cr3: #GP, pdptrs "
                               "reserved bits\n");
                        inject_gp(vcpu);
@@ -449,7 +463,19 @@ void set_cr3(struct kvm_vcpu *vcpu, unsigned long cr3)
        vcpu->cr3 = cr3;
        spin_lock(&vcpu->kvm->lock);
-        vcpu->mmu.new_cr3(vcpu);
+        /*
+         * Does the new cr3 value map to physical memory? (Note, we
+         * catch an invalid cr3 even in real-mode, because it would
+         * cause trouble later on when we turn on paging anyway.)
+         *
+         * A real CPU would silently accept an invalid cr3 and would
+         * attempt to use it - with largely undefined (and often hard
+         * to debug) behavior on the guest side.
+         */
+        if (unlikely(!gfn_to_memslot(vcpu->kvm, cr3 >> PAGE_SHIFT)))
+                inject_gp(vcpu);
+        else
+                vcpu->mmu.new_cr3(vcpu);
        spin_unlock(&vcpu->kvm->lock);
 }
 EXPORT_SYMBOL_GPL(set_cr3);
@@ -517,7 +543,6 @@ static int kvm_dev_ioctl_create_vcpu(struct kvm *kvm, int n)
        vcpu->guest_fx_image = vcpu->host_fx_image + FX_IMAGE_SIZE;
        vcpu->cpu = -1;  /* First load will set up TR */
-        vcpu->kvm = kvm;
        r = kvm_arch_ops->vcpu_create(vcpu);
        if (r < 0)
                goto out_free_vcpus;
@@ -634,6 +659,7 @@ raced:
                                                     | __GFP_ZERO);
                        if (!new.phys_mem[i])
                                goto out_free;
+                        new.phys_mem[i]->private = 0;
                }
        }
@@ -688,6 +714,13 @@ out:
        return r;
 }
+static void do_remove_write_access(struct kvm_vcpu *vcpu, int slot)
+{
+        spin_lock(&vcpu->kvm->lock);
+        kvm_mmu_slot_remove_write_access(vcpu, slot);
+        spin_unlock(&vcpu->kvm->lock);
+}
 /*
 * Get (and clear) the dirty memory log for a memory slot.
 */
@@ -697,6 +730,7 @@ static int kvm_dev_ioctl_get_dirty_log(struct kvm *kvm,
        struct kvm_memory_slot *memslot;
        int r, i;
        int n;
+        int cleared;
        unsigned long any = 0;
        spin_lock(&kvm->lock);
@@ -727,15 +761,17 @@ static int kvm_dev_ioctl_get_dirty_log(struct kvm *kvm,
        if (any) {
-                spin_lock(&kvm->lock);
+                cleared = 0;
-                kvm_mmu_slot_remove_write_access(kvm, log->slot);
-                spin_unlock(&kvm->lock);
-                memset(memslot->dirty_bitmap, 0, n);
                for (i = 0; i < KVM_MAX_VCPUS; ++i) {
                        struct kvm_vcpu *vcpu = vcpu_load(kvm, i);
                        if (!vcpu)
                                continue;
+                        if (!cleared) {
+                                do_remove_write_access(vcpu, log->slot);
+                                memset(memslot->dirty_bitmap, 0, n);
+                                cleared = 1;
+                        }
                        kvm_arch_ops->tlb_flush(vcpu);
                        vcpu_put(vcpu);
                }
@@ -863,6 +899,27 @@ static int emulator_read_emulated(unsigned long addr,
        }
 }
+static int emulator_write_phys(struct kvm_vcpu *vcpu, gpa_t gpa,
+                               unsigned long val, int bytes)
+{
+        struct kvm_memory_slot *m;
+        struct page *page;
+        void *virt;
+        if (((gpa + bytes - 1) >> PAGE_SHIFT) != (gpa >> PAGE_SHIFT))
+                return 0;
+        m = gfn_to_memslot(vcpu->kvm, gpa >> PAGE_SHIFT);
+        if (!m)
+                return 0;
+        page = gfn_to_page(m, gpa >> PAGE_SHIFT);
+        kvm_mmu_pre_write(vcpu, gpa, bytes);
+        virt = kmap_atomic(page, KM_USER0);
+        memcpy(virt + offset_in_page(gpa), &val, bytes);
+        kunmap_atomic(virt, KM_USER0);
+        kvm_mmu_post_write(vcpu, gpa, bytes);
+        return 1;
+}
 static int emulator_write_emulated(unsigned long addr,
                                   unsigned long val,
                                   unsigned int bytes,
@@ -874,6 +931,9 @@ static int emulator_write_emulated(unsigned long addr,
        if (gpa == UNMAPPED_GVA)
                return X86EMUL_PROPAGATE_FAULT;
+        if (emulator_write_phys(vcpu, gpa, val, bytes))
+                return X86EMUL_CONTINUE;
        vcpu->mmio_needed = 1;
        vcpu->mmio_phys_addr = gpa;
        vcpu->mmio_size = bytes;
@@ -898,6 +958,30 @@ static int emulator_cmpxchg_emulated(unsigned long addr,
        return emulator_write_emulated(addr, new, bytes, ctxt);
 }
+#ifdef CONFIG_X86_32
+static int emulator_cmpxchg8b_emulated(unsigned long addr,
+                                       unsigned long old_lo,
+                                       unsigned long old_hi,
+                                       unsigned long new_lo,
+                                       unsigned long new_hi,
+                                       struct x86_emulate_ctxt *ctxt)
+{
+        static int reported;
+        int r;
+        if (!reported) {
+                reported = 1;
+                printk(KERN_WARNING "kvm: emulating exchange8b as write\n");
+        }
+        r = emulator_write_emulated(addr, new_lo, 4, ctxt);
+        if (r != X86EMUL_CONTINUE)
+                return r;
+        return emulator_write_emulated(addr+4, new_hi, 4, ctxt);
+}
+#endif
 static unsigned long get_segment_base(struct kvm_vcpu *vcpu, int seg)
 {
        return kvm_arch_ops->get_segment_base(vcpu, seg);
@@ -905,18 +989,15 @@ static unsigned long get_segment_base(struct kvm_vcpu *vcpu, int seg)
 int emulate_invlpg(struct kvm_vcpu *vcpu, gva_t address)
 {
-        spin_lock(&vcpu->kvm->lock);
-        vcpu->mmu.inval_page(vcpu, address);
-        spin_unlock(&vcpu->kvm->lock);
-        kvm_arch_ops->invlpg(vcpu, address);
        return X86EMUL_CONTINUE;
 }
 int emulate_clts(struct kvm_vcpu *vcpu)
 {
-        unsigned long cr0 = vcpu->cr0;
+        unsigned long cr0;
-        cr0 &= ~CR0_TS_MASK;
+        kvm_arch_ops->decache_cr0_cr4_guest_bits(vcpu);
+        cr0 = vcpu->cr0 & ~CR0_TS_MASK;
        kvm_arch_ops->set_cr0(vcpu, cr0);
        return X86EMUL_CONTINUE;
 }
@@ -975,6 +1056,9 @@ struct x86_emulate_ops emulate_ops = {
        .read_emulated       = emulator_read_emulated,
        .write_emulated      = emulator_write_emulated,
        .cmpxchg_emulated    = emulator_cmpxchg_emulated,
+#ifdef CONFIG_X86_32
+        .cmpxchg8b_emulated  = emulator_cmpxchg8b_emulated,
+#endif
 };
 int emulate_instruction(struct kvm_vcpu *vcpu,
@@ -1024,6 +1108,8 @@ int emulate_instruction(struct kvm_vcpu *vcpu,
        }
        if (r) {
+                if (kvm_mmu_unprotect_page_virt(vcpu, cr2))
+                        return EMULATE_DONE;
                if (!vcpu->mmio_needed) {
                        report_emulation_failure(&emulate_ctxt);
                        return EMULATE_FAIL;
@@ -1069,6 +1155,7 @@ void realmode_lmsw(struct kvm_vcpu *vcpu, unsigned long msw,
 unsigned long realmode_get_cr(struct kvm_vcpu *vcpu, int cr)
 {
+        kvm_arch_ops->decache_cr0_cr4_guest_bits(vcpu);
        switch (cr) {
        case 0:
                return vcpu->cr0;
@@ -1403,6 +1490,7 @@ static int kvm_dev_ioctl_get_sregs(struct kvm *kvm, struct kvm_sregs *sregs)
        sregs->gdt.limit = dt.limit;
        sregs->gdt.base = dt.base;
+        kvm_arch_ops->decache_cr0_cr4_guest_bits(vcpu);
        sregs->cr0 = vcpu->cr0;
        sregs->cr2 = vcpu->cr2;
        sregs->cr3 = vcpu->cr3;
@@ -1467,11 +1555,15 @@ static int kvm_dev_ioctl_set_sregs(struct kvm *kvm, struct kvm_sregs *sregs)
 #endif
        vcpu->apic_base = sregs->apic_base;
+        kvm_arch_ops->decache_cr0_cr4_guest_bits(vcpu);
        mmu_reset_needed |= vcpu->cr0 != sregs->cr0;
        kvm_arch_ops->set_cr0_no_modeswitch(vcpu, sregs->cr0);
        mmu_reset_needed |= vcpu->cr4 != sregs->cr4;
        kvm_arch_ops->set_cr4(vcpu, sregs->cr4);
+        if (!is_long_mode(vcpu) && is_pae(vcpu))
+                load_pdptrs(vcpu, vcpu->cr3);
        if (mmu_reset_needed)
                kvm_mmu_reset_context(vcpu);
@@ -1693,12 +1785,12 @@ static long kvm_dev_ioctl(struct file *filp,
                if (copy_from_user(&kvm_run, (void *)arg, sizeof kvm_run))
                        goto out;
                r = kvm_dev_ioctl_run(kvm, &kvm_run);
-                if (r < 0)
+                if (r < 0 &&  r != -EINTR)
                        goto out;
-                r = -EFAULT;
+                if (copy_to_user((void *)arg, &kvm_run, sizeof kvm_run)) {
-                if (copy_to_user((void *)arg, &kvm_run, sizeof kvm_run))
+                        r = -EFAULT;
                        goto out;
-                r = 0;
+                }
                break;
        }
        case KVM_GET_REGS: {
@@ -1842,6 +1934,7 @@ static long kvm_dev_ioctl(struct file *filp,
                                 num_msrs_to_save * sizeof(u32)))
                        goto out;
                r = 0;
+                break;
        }
        default:
                ;
@@ -1944,17 +2037,17 @@ int kvm_init_arch(struct kvm_arch_ops *ops, struct module *module)
                return -EEXIST;
        }
-        kvm_arch_ops = ops;
+        if (!ops->cpu_has_kvm_support()) {
-        if (!kvm_arch_ops->cpu_has_kvm_support()) {
                printk(KERN_ERR "kvm: no hardware support\n");
                return -EOPNOTSUPP;
        }
-        if (kvm_arch_ops->disabled_by_bios()) {
+        if (ops->disabled_by_bios()) {
                printk(KERN_ERR "kvm: disabled by bios\n");
                return -EOPNOTSUPP;
        }
+        kvm_arch_ops = ops;
        r = kvm_arch_ops->hardware_setup();
        if (r < 0)
            return r;
diff --git a/drivers/kvm/mmu.c b/drivers/kvm/mmu.c
index 790423c5f23d..c6f972914f08 100644
--- a/drivers/kvm/mmu.c
+++ b/drivers/kvm/mmu.c
@@ -26,7 +26,31 @@
 #include "vmx.h"
 #include "kvm.h"
+#undef MMU_DEBUG
+#undef AUDIT
+#ifdef AUDIT
+static void kvm_mmu_audit(struct kvm_vcpu *vcpu, const char *msg);
+#else
+static void kvm_mmu_audit(struct kvm_vcpu *vcpu, const char *msg) {}
+#endif
+#ifdef MMU_DEBUG
+#define pgprintk(x...) do { if (dbg) printk(x); } while (0)
+#define rmap_printk(x...) do { if (dbg) printk(x); } while (0)
+#else
 #define pgprintk(x...) do { } while (0)
+#define rmap_printk(x...) do { } while (0)
+#endif
+#if defined(MMU_DEBUG) || defined(AUDIT)
+static int dbg = 1;
+#endif
 #define ASSERT(x)                                                       \
        if (!(x)) {                                                     \
@@ -34,8 +58,10 @@
                       __FILE__, __LINE__, #x);                         \
        }
-#define PT64_ENT_PER_PAGE 512
+#define PT64_PT_BITS 9
-#define PT32_ENT_PER_PAGE 1024
+#define PT64_ENT_PER_PAGE (1 << PT64_PT_BITS)
+#define PT32_PT_BITS 10
+#define PT32_ENT_PER_PAGE (1 << PT32_PT_BITS)
 #define PT_WRITABLE_SHIFT 1
@@ -125,6 +151,13 @@
 #define PT_DIRECTORY_LEVEL 2
 #define PT_PAGE_TABLE_LEVEL 1
+#define RMAP_EXT 4
+struct kvm_rmap_desc {
+        u64 *shadow_ptes[RMAP_EXT];
+        struct kvm_rmap_desc *more;
+};
 static int is_write_protection(struct kvm_vcpu *vcpu)
 {
        return vcpu->cr0 & CR0_WP_MASK;
@@ -150,32 +183,272 @@ static int is_io_pte(unsigned long pte)
        return pte & PT_SHADOW_IO_MARK;
 }
+static int is_rmap_pte(u64 pte)
+{
+        return (pte & (PT_WRITABLE_MASK | PT_PRESENT_MASK))
+                == (PT_WRITABLE_MASK | PT_PRESENT_MASK);
+}
+static int mmu_topup_memory_cache(struct kvm_mmu_memory_cache *cache,
+                                  size_t objsize, int min)
+{
+        void *obj;
+        if (cache->nobjs >= min)
+                return 0;
+        while (cache->nobjs < ARRAY_SIZE(cache->objects)) {
+                obj = kzalloc(objsize, GFP_NOWAIT);
+                if (!obj)
+                        return -ENOMEM;
+                cache->objects[cache->nobjs++] = obj;
+        }
+        return 0;
+}
+static void mmu_free_memory_cache(struct kvm_mmu_memory_cache *mc)
+{
+        while (mc->nobjs)
+                kfree(mc->objects[--mc->nobjs]);
+}
+static int mmu_topup_memory_caches(struct kvm_vcpu *vcpu)
+{
+        int r;
+        r = mmu_topup_memory_cache(&vcpu->mmu_pte_chain_cache,
+                                   sizeof(struct kvm_pte_chain), 4);
+        if (r)
+                goto out;
+        r = mmu_topup_memory_cache(&vcpu->mmu_rmap_desc_cache,
+                                   sizeof(struct kvm_rmap_desc), 1);
+out:
+        return r;
+}
+static void mmu_free_memory_caches(struct kvm_vcpu *vcpu)
+{
+        mmu_free_memory_cache(&vcpu->mmu_pte_chain_cache);
+        mmu_free_memory_cache(&vcpu->mmu_rmap_desc_cache);
+}
+static void *mmu_memory_cache_alloc(struct kvm_mmu_memory_cache *mc,
+                                    size_t size)
+{
+        void *p;
+        BUG_ON(!mc->nobjs);
+        p = mc->objects[--mc->nobjs];
+        memset(p, 0, size);
+        return p;
+}
+static void mmu_memory_cache_free(struct kvm_mmu_memory_cache *mc, void *obj)
+{
+        if (mc->nobjs < KVM_NR_MEM_OBJS)
+                mc->objects[mc->nobjs++] = obj;
+        else
+                kfree(obj);
+}
+static struct kvm_pte_chain *mmu_alloc_pte_chain(struct kvm_vcpu *vcpu)
+{
+        return mmu_memory_cache_alloc(&vcpu->mmu_pte_chain_cache,
+                                      sizeof(struct kvm_pte_chain));
+}
+static void mmu_free_pte_chain(struct kvm_vcpu *vcpu,
+                               struct kvm_pte_chain *pc)
+{
+        mmu_memory_cache_free(&vcpu->mmu_pte_chain_cache, pc);
+}
+static struct kvm_rmap_desc *mmu_alloc_rmap_desc(struct kvm_vcpu *vcpu)
+{
+        return mmu_memory_cache_alloc(&vcpu->mmu_rmap_desc_cache,
+                                      sizeof(struct kvm_rmap_desc));
+}
+static void mmu_free_rmap_desc(struct kvm_vcpu *vcpu,
+                               struct kvm_rmap_desc *rd)
+{
+        mmu_memory_cache_free(&vcpu->mmu_rmap_desc_cache, rd);
+}
+/*
+ * Reverse mapping data structures:
+ *
+ * If page->private bit zero is zero, then page->private points to the
+ * shadow page table entry that points to page_address(page).
+ *
+ * If page->private bit zero is one, (then page->private & ~1) points
+ * to a struct kvm_rmap_desc containing more mappings.
+ */
+static void rmap_add(struct kvm_vcpu *vcpu, u64 *spte)
+{
+        struct page *page;
+        struct kvm_rmap_desc *desc;
+        int i;
+        if (!is_rmap_pte(*spte))
+                return;
+        page = pfn_to_page((*spte & PT64_BASE_ADDR_MASK) >> PAGE_SHIFT);
+        if (!page->private) {
+                rmap_printk("rmap_add: %p %llx 0->1\n", spte, *spte);
+                page->private = (unsigned long)spte;
+        } else if (!(page->private & 1)) {
+                rmap_printk("rmap_add: %p %llx 1->many\n", spte, *spte);
+                desc = mmu_alloc_rmap_desc(vcpu);
+                desc->shadow_ptes[0] = (u64 *)page->private;
+                desc->shadow_ptes[1] = spte;
+                page->private = (unsigned long)desc | 1;
+        } else {
+                rmap_printk("rmap_add: %p %llx many->many\n", spte, *spte);
+                desc = (struct kvm_rmap_desc *)(page->private & ~1ul);
+                while (desc->shadow_ptes[RMAP_EXT-1] && desc->more)
+                        desc = desc->more;
+                if (desc->shadow_ptes[RMAP_EXT-1]) {
+                        desc->more = mmu_alloc_rmap_desc(vcpu);
+                        desc = desc->more;
+                }
+                for (i = 0; desc->shadow_ptes[i]; ++i)
+                        ;
+                desc->shadow_ptes[i] = spte;
+        }
+}
+static void rmap_desc_remove_entry(struct kvm_vcpu *vcpu,
+                                   struct page *page,
+                                   struct kvm_rmap_desc *desc,
+                                   int i,
+                                   struct kvm_rmap_desc *prev_desc)
+{
+        int j;
+        for (j = RMAP_EXT - 1; !desc->shadow_ptes[j] && j > i; --j)
+                ;
+        desc->shadow_ptes[i] = desc->shadow_ptes[j];
+        desc->shadow_ptes[j] = 0;
+        if (j != 0)
+                return;
+        if (!prev_desc && !desc->more)
+                page->private = (unsigned long)desc->shadow_ptes[0];
+        else
+                if (prev_desc)
+                        prev_desc->more = desc->more;
+                else
+                        page->private = (unsigned long)desc->more | 1;
+        mmu_free_rmap_desc(vcpu, desc);
+}
+static void rmap_remove(struct kvm_vcpu *vcpu, u64 *spte)
+{
+        struct page *page;
+        struct kvm_rmap_desc *desc;
+        struct kvm_rmap_desc *prev_desc;
+        int i;
+        if (!is_rmap_pte(*spte))
+                return;
+        page = pfn_to_page((*spte & PT64_BASE_ADDR_MASK) >> PAGE_SHIFT);
+        if (!page->private) {
+                printk(KERN_ERR "rmap_remove: %p %llx 0->BUG\n", spte, *spte);
+                BUG();
+        } else if (!(page->private & 1)) {
+                rmap_printk("rmap_remove:  %p %llx 1->0\n", spte, *spte);
+                if ((u64 *)page->private != spte) {
+                        printk(KERN_ERR "rmap_remove:  %p %llx 1->BUG\n",
+                               spte, *spte);
+                        BUG();
+                }
+                page->private = 0;
+        } else {
+                rmap_printk("rmap_remove:  %p %llx many->many\n", spte, *spte);
+                desc = (struct kvm_rmap_desc *)(page->private & ~1ul);
+                prev_desc = NULL;
+                while (desc) {
+                        for (i = 0; i < RMAP_EXT && desc->shadow_ptes[i]; ++i)
+                                if (desc->shadow_ptes[i] == spte) {
+                                        rmap_desc_remove_entry(vcpu, page,
+                                                               desc, i,
+                                                               prev_desc);
+                                        return;
+                                }
+                        prev_desc = desc;
+                        desc = desc->more;
+                }
+                BUG();
+        }
+}
+static void rmap_write_protect(struct kvm_vcpu *vcpu, u64 gfn)
+{
+        struct kvm *kvm = vcpu->kvm;
+        struct page *page;
+        struct kvm_memory_slot *slot;
+        struct kvm_rmap_desc *desc;
+        u64 *spte;
+        slot = gfn_to_memslot(kvm, gfn);
+        BUG_ON(!slot);
+        page = gfn_to_page(slot, gfn);
+        while (page->private) {
+                if (!(page->private & 1))
+                        spte = (u64 *)page->private;
+                else {
+                        desc = (struct kvm_rmap_desc *)(page->private & ~1ul);
+                        spte = desc->shadow_ptes[0];
+                }
+                BUG_ON(!spte);
+                BUG_ON((*spte & PT64_BASE_ADDR_MASK) !=
+                       page_to_pfn(page) << PAGE_SHIFT);
+                BUG_ON(!(*spte & PT_PRESENT_MASK));
+                BUG_ON(!(*spte & PT_WRITABLE_MASK));
+                rmap_printk("rmap_write_protect: spte %p %llx\n", spte, *spte);
+                rmap_remove(vcpu, spte);
+                kvm_arch_ops->tlb_flush(vcpu);
+                *spte &= ~(u64)PT_WRITABLE_MASK;
+        }
+}
+static int is_empty_shadow_page(hpa_t page_hpa)
+{
+        u64 *pos;
+        u64 *end;
+        for (pos = __va(page_hpa), end = pos + PAGE_SIZE / sizeof(u64);
+                      pos != end; pos++)
+                if (*pos != 0) {
+                        printk(KERN_ERR "%s: %p %llx\n", __FUNCTION__,
+                               pos, *pos);
+                        return 0;
+                }
+        return 1;
+}
 static void kvm_mmu_free_page(struct kvm_vcpu *vcpu, hpa_t page_hpa)
 {
        struct kvm_mmu_page *page_head = page_header(page_hpa);
+        ASSERT(is_empty_shadow_page(page_hpa));
        list_del(&page_head->link);
        page_head->page_hpa = page_hpa;
        list_add(&page_head->link, &vcpu->free_pages);
+        ++vcpu->kvm->n_free_mmu_pages;
 }
-static int is_empty_shadow_page(hpa_t page_hpa)
+static unsigned kvm_page_table_hashfn(gfn_t gfn)
 {
-        u32 *pos;
+        return gfn;
-        u32 *end;
-        for (pos = __va(page_hpa), end = pos + PAGE_SIZE / sizeof(u32);
-                      pos != end; pos++)
-                if (*pos != 0)
-                        return 0;
-        return 1;
 }
-static hpa_t kvm_mmu_alloc_page(struct kvm_vcpu *vcpu, u64 *parent_pte)
+static struct kvm_mmu_page *kvm_mmu_alloc_page(struct kvm_vcpu *vcpu,
+                                               u64 *parent_pte)
 {
        struct kvm_mmu_page *page;
        if (list_empty(&vcpu->free_pages))
-                return INVALID_PAGE;
+                return NULL;
        page = list_entry(vcpu->free_pages.next, struct kvm_mmu_page, link);
        list_del(&page->link);
@@ -183,8 +456,239 @@ static hpa_t kvm_mmu_alloc_page(struct kvm_vcpu *vcpu, u64 *parent_pte)
        ASSERT(is_empty_shadow_page(page->page_hpa));
        page->slot_bitmap = 0;
        page->global = 1;
+        page->multimapped = 0;
        page->parent_pte = parent_pte;
-        return page->page_hpa;
+        --vcpu->kvm->n_free_mmu_pages;
+        return page;
+}
+static void mmu_page_add_parent_pte(struct kvm_vcpu *vcpu,
+                                    struct kvm_mmu_page *page, u64 *parent_pte)
+{
+        struct kvm_pte_chain *pte_chain;
+        struct hlist_node *node;
+        int i;
+        if (!parent_pte)
+                return;
+        if (!page->multimapped) {
+                u64 *old = page->parent_pte;
+                if (!old) {
+                        page->parent_pte = parent_pte;
+                        return;
+                }
+                page->multimapped = 1;
+                pte_chain = mmu_alloc_pte_chain(vcpu);
+                INIT_HLIST_HEAD(&page->parent_ptes);
+                hlist_add_head(&pte_chain->link, &page->parent_ptes);
+                pte_chain->parent_ptes[0] = old;
+        }
+        hlist_for_each_entry(pte_chain, node, &page->parent_ptes, link) {
+                if (pte_chain->parent_ptes[NR_PTE_CHAIN_ENTRIES-1])
+                        continue;
+                for (i = 0; i < NR_PTE_CHAIN_ENTRIES; ++i)
+                        if (!pte_chain->parent_ptes[i]) {
+                                pte_chain->parent_ptes[i] = parent_pte;
+                                return;
+                        }
+        }
+        pte_chain = mmu_alloc_pte_chain(vcpu);
+        BUG_ON(!pte_chain);
+        hlist_add_head(&pte_chain->link, &page->parent_ptes);
+        pte_chain->parent_ptes[0] = parent_pte;
+}
+static void mmu_page_remove_parent_pte(struct kvm_vcpu *vcpu,
+                                       struct kvm_mmu_page *page,
+                                       u64 *parent_pte)
+{
+        struct kvm_pte_chain *pte_chain;
+        struct hlist_node *node;
+        int i;
+        if (!page->multimapped) {
+                BUG_ON(page->parent_pte != parent_pte);
+                page->parent_pte = NULL;
+                return;
+        }
+        hlist_for_each_entry(pte_chain, node, &page->parent_ptes, link)
+                for (i = 0; i < NR_PTE_CHAIN_ENTRIES; ++i) {
+                        if (!pte_chain->parent_ptes[i])
+                                break;
+                        if (pte_chain->parent_ptes[i] != parent_pte)
+                                continue;
+                        while (i + 1 < NR_PTE_CHAIN_ENTRIES
+                                && pte_chain->parent_ptes[i + 1]) {
+                                pte_chain->parent_ptes[i]
+                                        = pte_chain->parent_ptes[i + 1];
+                                ++i;
+                        }
+                        pte_chain->parent_ptes[i] = NULL;
+                        if (i == 0) {
+                                hlist_del(&pte_chain->link);
+                                mmu_free_pte_chain(vcpu, pte_chain);
+                                if (hlist_empty(&page->parent_ptes)) {
+                                        page->multimapped = 0;
+                                        page->parent_pte = NULL;
+                                }
+                        }
+                        return;
+                }
+        BUG();
+}
+static struct kvm_mmu_page *kvm_mmu_lookup_page(struct kvm_vcpu *vcpu,
+                                                gfn_t gfn)
+{
+        unsigned index;
+        struct hlist_head *bucket;
+        struct kvm_mmu_page *page;
+        struct hlist_node *node;
+        pgprintk("%s: looking for gfn %lx\n", __FUNCTION__, gfn);
+        index = kvm_page_table_hashfn(gfn) % KVM_NUM_MMU_PAGES;
+        bucket = &vcpu->kvm->mmu_page_hash[index];
+        hlist_for_each_entry(page, node, bucket, hash_link)
+                if (page->gfn == gfn && !page->role.metaphysical) {
+                        pgprintk("%s: found role %x\n",
+                                 __FUNCTION__, page->role.word);
+                        return page;
+                }
+        return NULL;
+}
+static struct kvm_mmu_page *kvm_mmu_get_page(struct kvm_vcpu *vcpu,
+                                             gfn_t gfn,
+                                             gva_t gaddr,
+                                             unsigned level,
+                                             int metaphysical,
+                                             u64 *parent_pte)
+{
+        union kvm_mmu_page_role role;
+        unsigned index;
+        unsigned quadrant;
+        struct hlist_head *bucket;
+        struct kvm_mmu_page *page;
+        struct hlist_node *node;
+        role.word = 0;
+        role.glevels = vcpu->mmu.root_level;
+        role.level = level;
+        role.metaphysical = metaphysical;
+        if (vcpu->mmu.root_level <= PT32_ROOT_LEVEL) {
+                quadrant = gaddr >> (PAGE_SHIFT + (PT64_PT_BITS * level));
+                quadrant &= (1 << ((PT32_PT_BITS - PT64_PT_BITS) * level)) - 1;
+                role.quadrant = quadrant;
+        }
+        pgprintk("%s: looking gfn %lx role %x\n", __FUNCTION__,
+                 gfn, role.word);
+        index = kvm_page_table_hashfn(gfn) % KVM_NUM_MMU_PAGES;
+        bucket = &vcpu->kvm->mmu_page_hash[index];
+        hlist_for_each_entry(page, node, bucket, hash_link)
+                if (page->gfn == gfn && page->role.word == role.word) {
+                        mmu_page_add_parent_pte(vcpu, page, parent_pte);
+                        pgprintk("%s: found\n", __FUNCTION__);
+                        return page;
+                }
+        page = kvm_mmu_alloc_page(vcpu, parent_pte);
+        if (!page)
+                return page;
+        pgprintk("%s: adding gfn %lx role %x\n", __FUNCTION__, gfn, role.word);
+        page->gfn = gfn;
+        page->role = role;
+        hlist_add_head(&page->hash_link, bucket);
+        if (!metaphysical)
+                rmap_write_protect(vcpu, gfn);
+        return page;
+}
+static void kvm_mmu_page_unlink_children(struct kvm_vcpu *vcpu,
+                                         struct kvm_mmu_page *page)
+{
+        unsigned i;
+        u64 *pt;
+        u64 ent;
+        pt = __va(page->page_hpa);
+        if (page->role.level == PT_PAGE_TABLE_LEVEL) {
+                for (i = 0; i < PT64_ENT_PER_PAGE; ++i) {
+                        if (pt[i] & PT_PRESENT_MASK)
+                                rmap_remove(vcpu, &pt[i]);
+                        pt[i] = 0;
+                }
+                kvm_arch_ops->tlb_flush(vcpu);
+                return;
+        }
+        for (i = 0; i < PT64_ENT_PER_PAGE; ++i) {
+                ent = pt[i];
+                pt[i] = 0;
+                if (!(ent & PT_PRESENT_MASK))
+                        continue;
+                ent &= PT64_BASE_ADDR_MASK;
+                mmu_page_remove_parent_pte(vcpu, page_header(ent), &pt[i]);
+        }
+}
+static void kvm_mmu_put_page(struct kvm_vcpu *vcpu,
+                             struct kvm_mmu_page *page,
+                             u64 *parent_pte)
+{
+        mmu_page_remove_parent_pte(vcpu, page, parent_pte);
+}
+static void kvm_mmu_zap_page(struct kvm_vcpu *vcpu,
+                             struct kvm_mmu_page *page)
+{
+        u64 *parent_pte;
+        while (page->multimapped || page->parent_pte) {
+                if (!page->multimapped)
+                        parent_pte = page->parent_pte;
+                else {
+                        struct kvm_pte_chain *chain;
+                        chain = container_of(page->parent_ptes.first,
+                                             struct kvm_pte_chain, link);
+                        parent_pte = chain->parent_ptes[0];
+                }
+                BUG_ON(!parent_pte);
+                kvm_mmu_put_page(vcpu, page, parent_pte);
+                *parent_pte = 0;
+        }
+        kvm_mmu_page_unlink_children(vcpu, page);
+        if (!page->root_count) {
+                hlist_del(&page->hash_link);
+                kvm_mmu_free_page(vcpu, page->page_hpa);
+        } else {
+                list_del(&page->link);
+                list_add(&page->link, &vcpu->kvm->active_mmu_pages);
+        }
+}
+static int kvm_mmu_unprotect_page(struct kvm_vcpu *vcpu, gfn_t gfn)
+{
+        unsigned index;
+        struct hlist_head *bucket;
+        struct kvm_mmu_page *page;
+        struct hlist_node *node, *n;
+        int r;
+        pgprintk("%s: looking for gfn %lx\n", __FUNCTION__, gfn);
+        r = 0;
+        index = kvm_page_table_hashfn(gfn) % KVM_NUM_MMU_PAGES;
+        bucket = &vcpu->kvm->mmu_page_hash[index];
+        hlist_for_each_entry_safe(page, node, n, bucket, hash_link)
+                if (page->gfn == gfn && !page->role.metaphysical) {
+                        pgprintk("%s: gfn %lx role %x\n", __FUNCTION__, gfn,
+                                 page->role.word);
+                        kvm_mmu_zap_page(vcpu, page);
+                        r = 1;
+                }
+        return r;
 }
 static void page_header_update_slot(struct kvm *kvm, void *pte, gpa_t gpa)
@@ -225,35 +729,6 @@ hpa_t gva_to_hpa(struct kvm_vcpu *vcpu, gva_t gva)
        return gpa_to_hpa(vcpu, gpa);
 }
-static void release_pt_page_64(struct kvm_vcpu *vcpu, hpa_t page_hpa,
-                               int level)
-{
-        ASSERT(vcpu);
-        ASSERT(VALID_PAGE(page_hpa));
-        ASSERT(level <= PT64_ROOT_LEVEL && level > 0);
-        if (level == 1)
-                memset(__va(page_hpa), 0, PAGE_SIZE);
-        else {
-                u64 *pos;
-                u64 *end;
-                for (pos = __va(page_hpa), end = pos + PT64_ENT_PER_PAGE;
-                     pos != end; pos++) {
-                        u64 current_ent = *pos;
-                        *pos = 0;
-                        if (is_present_pte(current_ent))
-                                release_pt_page_64(vcpu,
-                                                  current_ent &
-                                                  PT64_BASE_ADDR_MASK,
-                                                  level - 1);
-                }
-        }
-        kvm_mmu_free_page(vcpu, page_hpa);
-}
 static void nonpaging_new_cr3(struct kvm_vcpu *vcpu)
 {
 }
@@ -266,52 +741,109 @@ static int nonpaging_map(struct kvm_vcpu *vcpu, gva_t v, hpa_t p)
        for (; ; level--) {
                u32 index = PT64_INDEX(v, level);
                u64 *table;
+                u64 pte;
                ASSERT(VALID_PAGE(table_addr));
                table = __va(table_addr);
                if (level == 1) {
+                        pte = table[index];
+                        if (is_present_pte(pte) && is_writeble_pte(pte))
+                                return 0;
                        mark_page_dirty(vcpu->kvm, v >> PAGE_SHIFT);
                        page_header_update_slot(vcpu->kvm, table, v);
                        table[index] = p | PT_PRESENT_MASK | PT_WRITABLE_MASK |
                                                                PT_USER_MASK;
+                        rmap_add(vcpu, &table[index]);
                        return 0;
                }
                if (table[index] == 0) {
-                        hpa_t new_table = kvm_mmu_alloc_page(vcpu,
+                        struct kvm_mmu_page *new_table;
-                                                             &table[index]);
+                        gfn_t pseudo_gfn;
-                        if (!VALID_PAGE(new_table)) {
+                        pseudo_gfn = (v & PT64_DIR_BASE_ADDR_MASK)
+                                >> PAGE_SHIFT;
+                        new_table = kvm_mmu_get_page(vcpu, pseudo_gfn,
+                                                     v, level - 1,
+                                                     1, &table[index]);
+                        if (!new_table) {
                                pgprintk("nonpaging_map: ENOMEM\n");
                                return -ENOMEM;
                        }
-                        if (level == PT32E_ROOT_LEVEL)
+                        table[index] = new_table->page_hpa | PT_PRESENT_MASK
-                                table[index] = new_table | PT_PRESENT_MASK;
+                                | PT_WRITABLE_MASK | PT_USER_MASK;
-                        else
-                                table[index] = new_table | PT_PRESENT_MASK |
-                                                PT_WRITABLE_MASK | PT_USER_MASK;
                }
                table_addr = table[index] & PT64_BASE_ADDR_MASK;
        }
 }
-static void nonpaging_flush(struct kvm_vcpu *vcpu)
+static void mmu_free_roots(struct kvm_vcpu *vcpu)
 {
-        hpa_t root = vcpu->mmu.root_hpa;
+        int i;
+        struct kvm_mmu_page *page;
-        ++kvm_stat.tlb_flush;
+#ifdef CONFIG_X86_64
-        pgprintk("nonpaging_flush\n");
+        if (vcpu->mmu.shadow_root_level == PT64_ROOT_LEVEL) {
-        ASSERT(VALID_PAGE(root));
+                hpa_t root = vcpu->mmu.root_hpa;
-        release_pt_page_64(vcpu, root, vcpu->mmu.shadow_root_level);
-        root = kvm_mmu_alloc_page(vcpu, NULL);
+                ASSERT(VALID_PAGE(root));
-        ASSERT(VALID_PAGE(root));
+                page = page_header(root);
-        vcpu->mmu.root_hpa = root;
+                --page->root_count;
-        if (is_paging(vcpu))
+                vcpu->mmu.root_hpa = INVALID_PAGE;
-                root |= (vcpu->cr3 & (CR3_PCD_MASK | CR3_WPT_MASK));
+                return;
-        kvm_arch_ops->set_cr3(vcpu, root);
+        }
-        kvm_arch_ops->tlb_flush(vcpu);
+#endif
+        for (i = 0; i < 4; ++i) {
+                hpa_t root = vcpu->mmu.pae_root[i];
+                ASSERT(VALID_PAGE(root));
+                root &= PT64_BASE_ADDR_MASK;
+                page = page_header(root);
+                --page->root_count;
+                vcpu->mmu.pae_root[i] = INVALID_PAGE;
+        }
+        vcpu->mmu.root_hpa = INVALID_PAGE;
+}
+static void mmu_alloc_roots(struct kvm_vcpu *vcpu)
+{
+        int i;
+        gfn_t root_gfn;
+        struct kvm_mmu_page *page;
+        root_gfn = vcpu->cr3 >> PAGE_SHIFT;
+#ifdef CONFIG_X86_64
+        if (vcpu->mmu.shadow_root_level == PT64_ROOT_LEVEL) {
+                hpa_t root = vcpu->mmu.root_hpa;
+                ASSERT(!VALID_PAGE(root));
+                page = kvm_mmu_get_page(vcpu, root_gfn, 0,
+                                        PT64_ROOT_LEVEL, 0, NULL);
+                root = page->page_hpa;
+                ++page->root_count;
+                vcpu->mmu.root_hpa = root;
+                return;
+        }
+#endif
+        for (i = 0; i < 4; ++i) {
+                hpa_t root = vcpu->mmu.pae_root[i];
+                ASSERT(!VALID_PAGE(root));
+                if (vcpu->mmu.root_level == PT32E_ROOT_LEVEL)
+                        root_gfn = vcpu->pdptrs[i] >> PAGE_SHIFT;
+                else if (vcpu->mmu.root_level == 0)
+                        root_gfn = 0;
+                page = kvm_mmu_get_page(vcpu, root_gfn, i << 30,
+                                        PT32_ROOT_LEVEL, !is_paging(vcpu),
+                                        NULL);
+                root = page->page_hpa;
+                ++page->root_count;
+                vcpu->mmu.pae_root[i] = root | PT_PRESENT_MASK;
+        }
+        vcpu->mmu.root_hpa = __pa(vcpu->mmu.pae_root);
 }
 static gpa_t nonpaging_gva_to_gpa(struct kvm_vcpu *vcpu, gva_t vaddr)
@@ -322,43 +854,29 @@ static gpa_t nonpaging_gva_to_gpa(struct kvm_vcpu *vcpu, gva_t vaddr)
 static int nonpaging_page_fault(struct kvm_vcpu *vcpu, gva_t gva,
                               u32 error_code)
 {
-        int ret;
        gpa_t addr = gva;
+        hpa_t paddr;
+        int r;
+        r = mmu_topup_memory_caches(vcpu);
+        if (r)
+                return r;
        ASSERT(vcpu);
        ASSERT(VALID_PAGE(vcpu->mmu.root_hpa));
-        for (;;) {
-             hpa_t paddr;
-             paddr = gpa_to_hpa(vcpu , addr & PT64_BASE_ADDR_MASK);
-             if (is_error_hpa(paddr))
+        paddr = gpa_to_hpa(vcpu , addr & PT64_BASE_ADDR_MASK);
-                     return 1;
-             ret = nonpaging_map(vcpu, addr & PAGE_MASK, paddr);
+        if (is_error_hpa(paddr))
-             if (ret) {
+                return 1;
-                     nonpaging_flush(vcpu);
-                     continue;
-             }
-             break;
-        }
-        return ret;
-}
-static void nonpaging_inval_page(struct kvm_vcpu *vcpu, gva_t addr)
+        return nonpaging_map(vcpu, addr & PAGE_MASK, paddr);
-{
 }
 static void nonpaging_free(struct kvm_vcpu *vcpu)
 {
-        hpa_t root;
+        mmu_free_roots(vcpu);
-        ASSERT(vcpu);
-        root = vcpu->mmu.root_hpa;
-        if (VALID_PAGE(root))
-                release_pt_page_64(vcpu, root, vcpu->mmu.shadow_root_level);
-        vcpu->mmu.root_hpa = INVALID_PAGE;
 }
 static int nonpaging_init_context(struct kvm_vcpu *vcpu)
@@ -367,40 +885,31 @@ static int nonpaging_init_context(struct kvm_vcpu *vcpu)
        context->new_cr3 = nonpaging_new_cr3;
        context->page_fault = nonpaging_page_fault;
-        context->inval_page = nonpaging_inval_page;
        context->gva_to_gpa = nonpaging_gva_to_gpa;
        context->free = nonpaging_free;
-        context->root_level = PT32E_ROOT_LEVEL;
+        context->root_level = 0;
        context->shadow_root_level = PT32E_ROOT_LEVEL;
-        context->root_hpa = kvm_mmu_alloc_page(vcpu, NULL);
+        mmu_alloc_roots(vcpu);
        ASSERT(VALID_PAGE(context->root_hpa));
        kvm_arch_ops->set_cr3(vcpu, context->root_hpa);
        return 0;
 }
 static void kvm_mmu_flush_tlb(struct kvm_vcpu *vcpu)
 {
-        struct kvm_mmu_page *page, *npage;
-        list_for_each_entry_safe(page, npage, &vcpu->kvm->active_mmu_pages,
-                                 link) {
-                if (page->global)
-                        continue;
-                if (!page->parent_pte)
-                        continue;
-                *page->parent_pte = 0;
-                release_pt_page_64(vcpu, page->page_hpa, 1);
-        }
        ++kvm_stat.tlb_flush;
        kvm_arch_ops->tlb_flush(vcpu);
 }
 static void paging_new_cr3(struct kvm_vcpu *vcpu)
 {
+        pgprintk("%s: cr3 %lx\n", __FUNCTION__, vcpu->cr3);
+        mmu_free_roots(vcpu);
+        if (unlikely(vcpu->kvm->n_free_mmu_pages < KVM_MIN_FREE_MMU_PAGES))
+                kvm_mmu_free_some_pages(vcpu);
+        mmu_alloc_roots(vcpu);
        kvm_mmu_flush_tlb(vcpu);
+        kvm_arch_ops->set_cr3(vcpu, vcpu->mmu.root_hpa);
 }
 static void mark_pagetable_nonglobal(void *shadow_pte)
@@ -412,7 +921,8 @@ static inline void set_pte_common(struct kvm_vcpu *vcpu,
                             u64 *shadow_pte,
                             gpa_t gaddr,
                             int dirty,
-                             u64 access_bits)
+                             u64 access_bits,
+                             gfn_t gfn)
 {
        hpa_t paddr;
@@ -420,13 +930,10 @@ static inline void set_pte_common(struct kvm_vcpu *vcpu,
        if (!dirty)
                access_bits &= ~PT_WRITABLE_MASK;
-        if (access_bits & PT_WRITABLE_MASK)
+        paddr = gpa_to_hpa(vcpu, gaddr & PT64_BASE_ADDR_MASK);
-                mark_page_dirty(vcpu->kvm, gaddr >> PAGE_SHIFT);
        *shadow_pte |= access_bits;
-        paddr = gpa_to_hpa(vcpu, gaddr & PT64_BASE_ADDR_MASK);
        if (!(*shadow_pte & PT_GLOBAL_MASK))
                mark_pagetable_nonglobal(shadow_pte);
@@ -434,10 +941,31 @@ static inline void set_pte_common(struct kvm_vcpu *vcpu,
                *shadow_pte |= gaddr;
                *shadow_pte |= PT_SHADOW_IO_MARK;
                *shadow_pte &= ~PT_PRESENT_MASK;
-        } else {
+                return;
-                *shadow_pte |= paddr;
+        }
-                page_header_update_slot(vcpu->kvm, shadow_pte, gaddr);
+        *shadow_pte |= paddr;
+        if (access_bits & PT_WRITABLE_MASK) {
+                struct kvm_mmu_page *shadow;
+                shadow = kvm_mmu_lookup_page(vcpu, gfn);
+                if (shadow) {
+                        pgprintk("%s: found shadow page for %lx, marking ro\n",
+                                 __FUNCTION__, gfn);
+                        access_bits &= ~PT_WRITABLE_MASK;
+                        if (is_writeble_pte(*shadow_pte)) {
+                                    *shadow_pte &= ~PT_WRITABLE_MASK;
+                                    kvm_arch_ops->tlb_flush(vcpu);
+                        }
+                }
        }
+        if (access_bits & PT_WRITABLE_MASK)
+                mark_page_dirty(vcpu->kvm, gaddr >> PAGE_SHIFT);
+        page_header_update_slot(vcpu->kvm, shadow_pte, gaddr);
+        rmap_add(vcpu, shadow_pte);
 }
 static void inject_page_fault(struct kvm_vcpu *vcpu,
@@ -474,41 +1002,6 @@ static int may_access(u64 pte, int write, int user)
        return 1;
 }
-/*
- * Remove a shadow pte.
- */
-static void paging_inval_page(struct kvm_vcpu *vcpu, gva_t addr)
-{
-        hpa_t page_addr = vcpu->mmu.root_hpa;
-        int level = vcpu->mmu.shadow_root_level;
-        ++kvm_stat.invlpg;
-        for (; ; level--) {
-                u32 index = PT64_INDEX(addr, level);
-                u64 *table = __va(page_addr);
-                if (level == PT_PAGE_TABLE_LEVEL ) {
-                        table[index] = 0;
-                        return;
-                }
-                if (!is_present_pte(table[index]))
-                        return;
-                page_addr = table[index] & PT64_BASE_ADDR_MASK;
-                if (level == PT_DIRECTORY_LEVEL &&
-                          (table[index] & PT_SHADOW_PS_MARK)) {
-                        table[index] = 0;
-                        release_pt_page_64(vcpu, page_addr, PT_PAGE_TABLE_LEVEL);
-                        kvm_arch_ops->tlb_flush(vcpu);
-                        return;
-                }
-        }
-}
 static void paging_free(struct kvm_vcpu *vcpu)
 {
        nonpaging_free(vcpu);
@@ -522,37 +1015,40 @@ static void paging_free(struct kvm_vcpu *vcpu)
 #include "paging_tmpl.h"
 #undef PTTYPE
-static int paging64_init_context(struct kvm_vcpu *vcpu)
+static int paging64_init_context_common(struct kvm_vcpu *vcpu, int level)
 {
        struct kvm_mmu *context = &vcpu->mmu;
        ASSERT(is_pae(vcpu));
        context->new_cr3 = paging_new_cr3;
        context->page_fault = paging64_page_fault;
-        context->inval_page = paging_inval_page;
        context->gva_to_gpa = paging64_gva_to_gpa;
        context->free = paging_free;
-        context->root_level = PT64_ROOT_LEVEL;
+        context->root_level = level;
-        context->shadow_root_level = PT64_ROOT_LEVEL;
+        context->shadow_root_level = level;
-        context->root_hpa = kvm_mmu_alloc_page(vcpu, NULL);
+        mmu_alloc_roots(vcpu);
        ASSERT(VALID_PAGE(context->root_hpa));
        kvm_arch_ops->set_cr3(vcpu, context->root_hpa |
                    (vcpu->cr3 & (CR3_PCD_MASK | CR3_WPT_MASK)));
        return 0;
 }
+static int paging64_init_context(struct kvm_vcpu *vcpu)
+{
+        return paging64_init_context_common(vcpu, PT64_ROOT_LEVEL);
+}
 static int paging32_init_context(struct kvm_vcpu *vcpu)
 {
        struct kvm_mmu *context = &vcpu->mmu;
        context->new_cr3 = paging_new_cr3;
        context->page_fault = paging32_page_fault;
-        context->inval_page = paging_inval_page;
        context->gva_to_gpa = paging32_gva_to_gpa;
        context->free = paging_free;
        context->root_level = PT32_ROOT_LEVEL;
        context->shadow_root_level = PT32E_ROOT_LEVEL;
-        context->root_hpa = kvm_mmu_alloc_page(vcpu, NULL);
+        mmu_alloc_roots(vcpu);
        ASSERT(VALID_PAGE(context->root_hpa));
        kvm_arch_ops->set_cr3(vcpu, context->root_hpa |
                    (vcpu->cr3 & (CR3_PCD_MASK | CR3_WPT_MASK)));
@@ -561,14 +1057,7 @@ static int paging32_init_context(struct kvm_vcpu *vcpu)
 static int paging32E_init_context(struct kvm_vcpu *vcpu)
 {
-        int ret;
+        return paging64_init_context_common(vcpu, PT32E_ROOT_LEVEL);
-        if ((ret = paging64_init_context(vcpu)))
-                return ret;
-        vcpu->mmu.root_level = PT32E_ROOT_LEVEL;
-        vcpu->mmu.shadow_root_level = PT32E_ROOT_LEVEL;
-        return 0;
 }
 static int init_kvm_mmu(struct kvm_vcpu *vcpu)
@@ -597,41 +1086,161 @@ static void destroy_kvm_mmu(struct kvm_vcpu *vcpu)
 int kvm_mmu_reset_context(struct kvm_vcpu *vcpu)
 {
+        int r;
        destroy_kvm_mmu(vcpu);
-        return init_kvm_mmu(vcpu);
+        r = init_kvm_mmu(vcpu);
+        if (r < 0)
+                goto out;
+        r = mmu_topup_memory_caches(vcpu);
+out:
+        return r;
 }
-static void free_mmu_pages(struct kvm_vcpu *vcpu)
+void kvm_mmu_pre_write(struct kvm_vcpu *vcpu, gpa_t gpa, int bytes)
 {
-        while (!list_empty(&vcpu->free_pages)) {
+        gfn_t gfn = gpa >> PAGE_SHIFT;
+        struct kvm_mmu_page *page;
+        struct kvm_mmu_page *child;
+        struct hlist_node *node, *n;
+        struct hlist_head *bucket;
+        unsigned index;
+        u64 *spte;
+        u64 pte;
+        unsigned offset = offset_in_page(gpa);
+        unsigned pte_size;
+        unsigned page_offset;
+        unsigned misaligned;
+        int level;
+        int flooded = 0;
+        pgprintk("%s: gpa %llx bytes %d\n", __FUNCTION__, gpa, bytes);
+        if (gfn == vcpu->last_pt_write_gfn) {
+                ++vcpu->last_pt_write_count;
+                if (vcpu->last_pt_write_count >= 3)
+                        flooded = 1;
+        } else {
+                vcpu->last_pt_write_gfn = gfn;
+                vcpu->last_pt_write_count = 1;
+        }
+        index = kvm_page_table_hashfn(gfn) % KVM_NUM_MMU_PAGES;
+        bucket = &vcpu->kvm->mmu_page_hash[index];
+        hlist_for_each_entry_safe(page, node, n, bucket, hash_link) {
+                if (page->gfn != gfn || page->role.metaphysical)
+                        continue;
+                pte_size = page->role.glevels == PT32_ROOT_LEVEL ? 4 : 8;
+                misaligned = (offset ^ (offset + bytes - 1)) & ~(pte_size - 1);
+                if (misaligned || flooded) {
+                        /*
+                         * Misaligned accesses are too much trouble to fix
+                         * up; also, they usually indicate a page is not used
+                         * as a page table.
+                         *
+                         * If we're seeing too many writes to a page,
+                         * it may no longer be a page table, or we may be
+                         * forking, in which case it is better to unmap the
+                         * page.
+                         */
+                        pgprintk("misaligned: gpa %llx bytes %d role %x\n",
+                                 gpa, bytes, page->role.word);
+                        kvm_mmu_zap_page(vcpu, page);
+                        continue;
+                }
+                page_offset = offset;
+                level = page->role.level;
+                if (page->role.glevels == PT32_ROOT_LEVEL) {
+                        page_offset <<= 1;          /* 32->64 */
+                        page_offset &= ~PAGE_MASK;
+                }
+                spte = __va(page->page_hpa);
+                spte += page_offset / sizeof(*spte);
+                pte = *spte;
+                if (is_present_pte(pte)) {
+                        if (level == PT_PAGE_TABLE_LEVEL)
+                                rmap_remove(vcpu, spte);
+                        else {
+                                child = page_header(pte & PT64_BASE_ADDR_MASK);
+                                mmu_page_remove_parent_pte(vcpu, child, spte);
+                        }
+                }
+                *spte = 0;
+        }
+}
+void kvm_mmu_post_write(struct kvm_vcpu *vcpu, gpa_t gpa, int bytes)
+{
+}
+int kvm_mmu_unprotect_page_virt(struct kvm_vcpu *vcpu, gva_t gva)
+{
+        gpa_t gpa = vcpu->mmu.gva_to_gpa(vcpu, gva);
+        return kvm_mmu_unprotect_page(vcpu, gpa >> PAGE_SHIFT);
+}
+void kvm_mmu_free_some_pages(struct kvm_vcpu *vcpu)
+{
+        while (vcpu->kvm->n_free_mmu_pages < KVM_REFILL_PAGES) {
                struct kvm_mmu_page *page;
+                page = container_of(vcpu->kvm->active_mmu_pages.prev,
+                                    struct kvm_mmu_page, link);
+                kvm_mmu_zap_page(vcpu, page);
+        }
+}
+EXPORT_SYMBOL_GPL(kvm_mmu_free_some_pages);
+static void free_mmu_pages(struct kvm_vcpu *vcpu)
+{
+        struct kvm_mmu_page *page;
+        while (!list_empty(&vcpu->kvm->active_mmu_pages)) {
+                page = container_of(vcpu->kvm->active_mmu_pages.next,
+                                    struct kvm_mmu_page, link);
+                kvm_mmu_zap_page(vcpu, page);
+        }
+        while (!list_empty(&vcpu->free_pages)) {
                page = list_entry(vcpu->free_pages.next,
                                  struct kvm_mmu_page, link);
                list_del(&page->link);
                __free_page(pfn_to_page(page->page_hpa >> PAGE_SHIFT));
                page->page_hpa = INVALID_PAGE;
        }
+        free_page((unsigned long)vcpu->mmu.pae_root);
 }
 static int alloc_mmu_pages(struct kvm_vcpu *vcpu)
 {
+        struct page *page;
        int i;
        ASSERT(vcpu);
        for (i = 0; i < KVM_NUM_MMU_PAGES; i++) {
-                struct page *page;
                struct kvm_mmu_page *page_header = &vcpu->page_header_buf[i];
                INIT_LIST_HEAD(&page_header->link);
-                if ((page = alloc_page(GFP_KVM_MMU)) == NULL)
+                if ((page = alloc_page(GFP_KERNEL)) == NULL)
                        goto error_1;
                page->private = (unsigned long)page_header;
                page_header->page_hpa = (hpa_t)page_to_pfn(page) << PAGE_SHIFT;
                memset(__va(page_header->page_hpa), 0, PAGE_SIZE);
                list_add(&page_header->link, &vcpu->free_pages);
+                ++vcpu->kvm->n_free_mmu_pages;
        }
+        /*
+         * When emulating 32-bit mode, cr3 is only 32 bits even on x86_64.
+         * Therefore we need to allocate shadow page tables in the first
+         * 4GB of memory, which happens to fit the DMA32 zone.
+         */
+        page = alloc_page(GFP_KERNEL | __GFP_DMA32);
+        if (!page)
+                goto error_1;
+        vcpu->mmu.pae_root = page_address(page);
+        for (i = 0; i < 4; ++i)
+                vcpu->mmu.pae_root[i] = INVALID_PAGE;
        return 0;
 error_1:
@@ -663,10 +1272,12 @@ void kvm_mmu_destroy(struct kvm_vcpu *vcpu)
        destroy_kvm_mmu(vcpu);
        free_mmu_pages(vcpu);
+        mmu_free_memory_caches(vcpu);
 }
-void kvm_mmu_slot_remove_write_access(struct kvm *kvm, int slot)
+void kvm_mmu_slot_remove_write_access(struct kvm_vcpu *vcpu, int slot)
 {
+        struct kvm *kvm = vcpu->kvm;
        struct kvm_mmu_page *page;
        list_for_each_entry(page, &kvm->active_mmu_pages, link) {
@@ -679,8 +1290,169 @@ void kvm_mmu_slot_remove_write_access(struct kvm *kvm, int slot)
                pt = __va(page->page_hpa);
                for (i = 0; i < PT64_ENT_PER_PAGE; ++i)
                        /* avoid RMW */
-                        if (pt[i] & PT_WRITABLE_MASK)
+                        if (pt[i] & PT_WRITABLE_MASK) {
+                                rmap_remove(vcpu, &pt[i]);
                                pt[i] &= ~PT_WRITABLE_MASK;
+                        }
+        }
+}
+#ifdef AUDIT
+static const char *audit_msg;
+static gva_t canonicalize(gva_t gva)
+{
+#ifdef CONFIG_X86_64
+        gva = (long long)(gva << 16) >> 16;
+#endif
+        return gva;
+}
+static void audit_mappings_page(struct kvm_vcpu *vcpu, u64 page_pte,
+                                gva_t va, int level)
+{
+        u64 *pt = __va(page_pte & PT64_BASE_ADDR_MASK);
+        int i;
+        gva_t va_delta = 1ul << (PAGE_SHIFT + 9 * (level - 1));
+        for (i = 0; i < PT64_ENT_PER_PAGE; ++i, va += va_delta) {
+                u64 ent = pt[i];
+                if (!ent & PT_PRESENT_MASK)
+                        continue;
+                va = canonicalize(va);
+                if (level > 1)
+                        audit_mappings_page(vcpu, ent, va, level - 1);
+                else {
+                        gpa_t gpa = vcpu->mmu.gva_to_gpa(vcpu, va);
+                        hpa_t hpa = gpa_to_hpa(vcpu, gpa);
+                        if ((ent & PT_PRESENT_MASK)
+                            && (ent & PT64_BASE_ADDR_MASK) != hpa)
+                                printk(KERN_ERR "audit error: (%s) levels %d"
+                                       " gva %lx gpa %llx hpa %llx ent %llx\n",
+                                       audit_msg, vcpu->mmu.root_level,
+                                       va, gpa, hpa, ent);
+                }
        }
 }
+static void audit_mappings(struct kvm_vcpu *vcpu)
+{
+        int i;
+        if (vcpu->mmu.root_level == 4)
+                audit_mappings_page(vcpu, vcpu->mmu.root_hpa, 0, 4);
+        else
+                for (i = 0; i < 4; ++i)
+                        if (vcpu->mmu.pae_root[i] & PT_PRESENT_MASK)
+                                audit_mappings_page(vcpu,
+                                                    vcpu->mmu.pae_root[i],
+                                                    i << 30,
+                                                    2);
+}
+static int count_rmaps(struct kvm_vcpu *vcpu)
+{
+        int nmaps = 0;
+        int i, j, k;
+        for (i = 0; i < KVM_MEMORY_SLOTS; ++i) {
+                struct kvm_memory_slot *m = &vcpu->kvm->memslots[i];
+                struct kvm_rmap_desc *d;
+                for (j = 0; j < m->npages; ++j) {
+                        struct page *page = m->phys_mem[j];
+                        if (!page->private)
+                                continue;
+                        if (!(page->private & 1)) {
+                                ++nmaps;
+                                continue;
+                        }
+                        d = (struct kvm_rmap_desc *)(page->private & ~1ul);
+                        while (d) {
+                                for (k = 0; k < RMAP_EXT; ++k)
+                                        if (d->shadow_ptes[k])
+                                                ++nmaps;
+                                        else
+                                                break;
+                                d = d->more;
+                        }
+                }
+        }
+        return nmaps;
+}
+static int count_writable_mappings(struct kvm_vcpu *vcpu)
+{
+        int nmaps = 0;
+        struct kvm_mmu_page *page;
+        int i;
+        list_for_each_entry(page, &vcpu->kvm->active_mmu_pages, link) {
+                u64 *pt = __va(page->page_hpa);
+                if (page->role.level != PT_PAGE_TABLE_LEVEL)
+                        continue;
+                for (i = 0; i < PT64_ENT_PER_PAGE; ++i) {
+                        u64 ent = pt[i];
+                        if (!(ent & PT_PRESENT_MASK))
+                                continue;
+                        if (!(ent & PT_WRITABLE_MASK))
+                                continue;
+                        ++nmaps;
+                }
+        }
+        return nmaps;
+}
+static void audit_rmap(struct kvm_vcpu *vcpu)
+{
+        int n_rmap = count_rmaps(vcpu);
+        int n_actual = count_writable_mappings(vcpu);
+        if (n_rmap != n_actual)
+                printk(KERN_ERR "%s: (%s) rmap %d actual %d\n",
+                       __FUNCTION__, audit_msg, n_rmap, n_actual);
+}
+static void audit_write_protection(struct kvm_vcpu *vcpu)
+{
+        struct kvm_mmu_page *page;
+        list_for_each_entry(page, &vcpu->kvm->active_mmu_pages, link) {
+                hfn_t hfn;
+                struct page *pg;
+                if (page->role.metaphysical)
+                        continue;
+                hfn = gpa_to_hpa(vcpu, (gpa_t)page->gfn << PAGE_SHIFT)
+                        >> PAGE_SHIFT;
+                pg = pfn_to_page(hfn);
+                if (pg->private)
+                        printk(KERN_ERR "%s: (%s) shadow page has writable"
+                               " mappings: gfn %lx role %x\n",
+                               __FUNCTION__, audit_msg, page->gfn,
+                               page->role.word);
+        }
+}
+static void kvm_mmu_audit(struct kvm_vcpu *vcpu, const char *msg)
+{
+        int olddbg = dbg;
+        dbg = 0;
+        audit_msg = msg;
+        audit_rmap(vcpu);
+        audit_write_protection(vcpu);
+        audit_mappings(vcpu);
+        dbg = olddbg;
+}
+#endif
diff --git a/drivers/kvm/paging_tmpl.h b/drivers/kvm/paging_tmpl.h
index 09bb9b4ed12d..2dbf4307ed9e 100644
--- a/drivers/kvm/paging_tmpl.h
+++ b/drivers/kvm/paging_tmpl.h
@@ -32,6 +32,11 @@
        #define SHADOW_PT_INDEX(addr, level) PT64_INDEX(addr, level)
        #define PT_LEVEL_MASK(level) PT64_LEVEL_MASK(level)
        #define PT_PTE_COPY_MASK PT64_PTE_COPY_MASK
+        #ifdef CONFIG_X86_64
+        #define PT_MAX_FULL_LEVELS 4
+        #else
+        #define PT_MAX_FULL_LEVELS 2
+        #endif
 #elif PTTYPE == 32
        #define pt_element_t u32
        #define guest_walker guest_walker32
@@ -42,6 +47,7 @@
        #define SHADOW_PT_INDEX(addr, level) PT64_INDEX(addr, level)
        #define PT_LEVEL_MASK(level) PT32_LEVEL_MASK(level)
        #define PT_PTE_COPY_MASK PT32_PTE_COPY_MASK
+        #define PT_MAX_FULL_LEVELS 2
 #else
        #error Invalid PTTYPE value
 #endif
@@ -52,93 +58,126 @@
 */
 struct guest_walker {
        int level;
+        gfn_t table_gfn[PT_MAX_FULL_LEVELS];
        pt_element_t *table;
+        pt_element_t *ptep;
        pt_element_t inherited_ar;
+        gfn_t gfn;
 };
-static void FNAME(init_walker)(struct guest_walker *walker,
+/*
-                               struct kvm_vcpu *vcpu)
+ * Fetch a guest pte for a guest virtual address
+ */
+static void FNAME(walk_addr)(struct guest_walker *walker,
+                             struct kvm_vcpu *vcpu, gva_t addr)
 {
        hpa_t hpa;
        struct kvm_memory_slot *slot;
+        pt_element_t *ptep;
+        pt_element_t root;
+        gfn_t table_gfn;
+        pgprintk("%s: addr %lx\n", __FUNCTION__, addr);
        walker->level = vcpu->mmu.root_level;
-        slot = gfn_to_memslot(vcpu->kvm,
+        walker->table = NULL;
-                              (vcpu->cr3 & PT64_BASE_ADDR_MASK) >> PAGE_SHIFT);
+        root = vcpu->cr3;
-        hpa = safe_gpa_to_hpa(vcpu, vcpu->cr3 & PT64_BASE_ADDR_MASK);
+#if PTTYPE == 64
+        if (!is_long_mode(vcpu)) {
+                walker->ptep = &vcpu->pdptrs[(addr >> 30) & 3];
+                root = *walker->ptep;
+                if (!(root & PT_PRESENT_MASK))
+                        return;
+                --walker->level;
+        }
+#endif
+        table_gfn = (root & PT64_BASE_ADDR_MASK) >> PAGE_SHIFT;
+        walker->table_gfn[walker->level - 1] = table_gfn;
+        pgprintk("%s: table_gfn[%d] %lx\n", __FUNCTION__,
+                 walker->level - 1, table_gfn);
+        slot = gfn_to_memslot(vcpu->kvm, table_gfn);
+        hpa = safe_gpa_to_hpa(vcpu, root & PT64_BASE_ADDR_MASK);
        walker->table = kmap_atomic(pfn_to_page(hpa >> PAGE_SHIFT), KM_USER0);
        ASSERT((!is_long_mode(vcpu) && is_pae(vcpu)) ||
               (vcpu->cr3 & ~(PAGE_MASK | CR3_FLAGS_MASK)) == 0);
-        walker->table = (pt_element_t *)( (unsigned long)walker->table |
-                (unsigned long)(vcpu->cr3 & ~(PAGE_MASK | CR3_FLAGS_MASK)) );
        walker->inherited_ar = PT_USER_MASK | PT_WRITABLE_MASK;
+        for (;;) {
+                int index = PT_INDEX(addr, walker->level);
+                hpa_t paddr;
+                ptep = &walker->table[index];
+                ASSERT(((unsigned long)walker->table & PAGE_MASK) ==
+                       ((unsigned long)ptep & PAGE_MASK));
+                if (is_present_pte(*ptep) && !(*ptep &  PT_ACCESSED_MASK))
+                        *ptep |= PT_ACCESSED_MASK;
+                if (!is_present_pte(*ptep))
+                        break;
+                if (walker->level == PT_PAGE_TABLE_LEVEL) {
+                        walker->gfn = (*ptep & PT_BASE_ADDR_MASK)
+                                >> PAGE_SHIFT;
+                        break;
+                }
+                if (walker->level == PT_DIRECTORY_LEVEL
+                    && (*ptep & PT_PAGE_SIZE_MASK)
+                    && (PTTYPE == 64 || is_pse(vcpu))) {
+                        walker->gfn = (*ptep & PT_DIR_BASE_ADDR_MASK)
+                                >> PAGE_SHIFT;
+                        walker->gfn += PT_INDEX(addr, PT_PAGE_TABLE_LEVEL);
+                        break;
+                }
+                if (walker->level != 3 || is_long_mode(vcpu))
+                        walker->inherited_ar &= walker->table[index];
+                table_gfn = (*ptep & PT_BASE_ADDR_MASK) >> PAGE_SHIFT;
+                paddr = safe_gpa_to_hpa(vcpu, *ptep & PT_BASE_ADDR_MASK);
+                kunmap_atomic(walker->table, KM_USER0);
+                walker->table = kmap_atomic(pfn_to_page(paddr >> PAGE_SHIFT),
+                                            KM_USER0);
+                --walker->level;
+                walker->table_gfn[walker->level - 1 ] = table_gfn;
+                pgprintk("%s: table_gfn[%d] %lx\n", __FUNCTION__,
+                         walker->level - 1, table_gfn);
+        }
+        walker->ptep = ptep;
+        pgprintk("%s: pte %llx\n", __FUNCTION__, (u64)*ptep);
 }
 static void FNAME(release_walker)(struct guest_walker *walker)
 {
-        kunmap_atomic(walker->table, KM_USER0);
+        if (walker->table)
+                kunmap_atomic(walker->table, KM_USER0);
 }
 static void FNAME(set_pte)(struct kvm_vcpu *vcpu, u64 guest_pte,
-                           u64 *shadow_pte, u64 access_bits)
+                           u64 *shadow_pte, u64 access_bits, gfn_t gfn)
 {
        ASSERT(*shadow_pte == 0);
        access_bits &= guest_pte;
        *shadow_pte = (guest_pte & PT_PTE_COPY_MASK);
        set_pte_common(vcpu, shadow_pte, guest_pte & PT_BASE_ADDR_MASK,
-                       guest_pte & PT_DIRTY_MASK, access_bits);
+                       guest_pte & PT_DIRTY_MASK, access_bits, gfn);
 }
 static void FNAME(set_pde)(struct kvm_vcpu *vcpu, u64 guest_pde,
-                           u64 *shadow_pte, u64 access_bits,
+                           u64 *shadow_pte, u64 access_bits, gfn_t gfn)
-                           int index)
 {
        gpa_t gaddr;
        ASSERT(*shadow_pte == 0);
        access_bits &= guest_pde;
-        gaddr = (guest_pde & PT_DIR_BASE_ADDR_MASK) + PAGE_SIZE * index;
+        gaddr = (gpa_t)gfn << PAGE_SHIFT;
        if (PTTYPE == 32 && is_cpuid_PSE36())
                gaddr |= (guest_pde & PT32_DIR_PSE36_MASK) <<
                        (32 - PT32_DIR_PSE36_SHIFT);
        *shadow_pte = guest_pde & PT_PTE_COPY_MASK;
        set_pte_common(vcpu, shadow_pte, gaddr,
-                       guest_pde & PT_DIRTY_MASK, access_bits);
+                       guest_pde & PT_DIRTY_MASK, access_bits, gfn);
-}
-/*
- * Fetch a guest pte from a specific level in the paging hierarchy.
- */
-static pt_element_t *FNAME(fetch_guest)(struct kvm_vcpu *vcpu,
-                                        struct guest_walker *walker,
-                                        int level,
-                                        gva_t addr)
-{
-        ASSERT(level > 0  && level <= walker->level);
-        for (;;) {
-                int index = PT_INDEX(addr, walker->level);
-                hpa_t paddr;
-                ASSERT(((unsigned long)walker->table & PAGE_MASK) ==
-                       ((unsigned long)&walker->table[index] & PAGE_MASK));
-                if (level == walker->level ||
-                    !is_present_pte(walker->table[index]) ||
-                    (walker->level == PT_DIRECTORY_LEVEL &&
-                     (walker->table[index] & PT_PAGE_SIZE_MASK) &&
-                     (PTTYPE == 64 || is_pse(vcpu))))
-                        return &walker->table[index];
-                if (walker->level != 3 || is_long_mode(vcpu))
-                        walker->inherited_ar &= walker->table[index];
-                paddr = safe_gpa_to_hpa(vcpu, walker->table[index] & PT_BASE_ADDR_MASK);
-                kunmap_atomic(walker->table, KM_USER0);
-                walker->table = kmap_atomic(pfn_to_page(paddr >> PAGE_SHIFT),
-                                            KM_USER0);
-                --walker->level;
-        }
 }
 /*
@@ -150,15 +189,26 @@ static u64 *FNAME(fetch)(struct kvm_vcpu *vcpu, gva_t addr,
        hpa_t shadow_addr;
        int level;
        u64 *prev_shadow_ent = NULL;
+        pt_element_t *guest_ent = walker->ptep;
+        if (!is_present_pte(*guest_ent))
+                return NULL;
        shadow_addr = vcpu->mmu.root_hpa;
        level = vcpu->mmu.shadow_root_level;
+        if (level == PT32E_ROOT_LEVEL) {
+                shadow_addr = vcpu->mmu.pae_root[(addr >> 30) & 3];
+                shadow_addr &= PT64_BASE_ADDR_MASK;
+                --level;
+        }
        for (; ; level--) {
                u32 index = SHADOW_PT_INDEX(addr, level);
                u64 *shadow_ent = ((u64 *)__va(shadow_addr)) + index;
-                pt_element_t *guest_ent;
+                struct kvm_mmu_page *shadow_page;
                u64 shadow_pte;
+                int metaphysical;
+                gfn_t table_gfn;
                if (is_present_pte(*shadow_ent) || is_io_pte(*shadow_ent)) {
                        if (level == PT_PAGE_TABLE_LEVEL)
@@ -168,21 +218,6 @@ static u64 *FNAME(fetch)(struct kvm_vcpu *vcpu, gva_t addr,
                        continue;
                }
-                if (PTTYPE == 32 && level > PT32_ROOT_LEVEL) {
-                        ASSERT(level == PT32E_ROOT_LEVEL);
-                        guest_ent = FNAME(fetch_guest)(vcpu, walker,
-                                                       PT32_ROOT_LEVEL, addr);
-                } else
-                        guest_ent = FNAME(fetch_guest)(vcpu, walker,
-                                                       level, addr);
-                if (!is_present_pte(*guest_ent))
-                        return NULL;
-                /* Don't set accessed bit on PAE PDPTRs */
-                if (vcpu->mmu.root_level != 3 || walker->level != 3)
-                        *guest_ent |= PT_ACCESSED_MASK;
                if (level == PT_PAGE_TABLE_LEVEL) {
                        if (walker->level == PT_DIRECTORY_LEVEL) {
@@ -190,21 +225,30 @@ static u64 *FNAME(fetch)(struct kvm_vcpu *vcpu, gva_t addr,
                                        *prev_shadow_ent |= PT_SHADOW_PS_MARK;
                                FNAME(set_pde)(vcpu, *guest_ent, shadow_ent,
                                               walker->inherited_ar,
-                                          PT_INDEX(addr, PT_PAGE_TABLE_LEVEL));
+                                               walker->gfn);
                        } else {
                                ASSERT(walker->level == PT_PAGE_TABLE_LEVEL);
-                                FNAME(set_pte)(vcpu, *guest_ent, shadow_ent, walker->inherited_ar);
+                                FNAME(set_pte)(vcpu, *guest_ent, shadow_ent,
+                                               walker->inherited_ar,
+                                               walker->gfn);
                        }
                        return shadow_ent;
                }
-                shadow_addr = kvm_mmu_alloc_page(vcpu, shadow_ent);
+                if (level - 1 == PT_PAGE_TABLE_LEVEL
-                if (!VALID_PAGE(shadow_addr))
+                    && walker->level == PT_DIRECTORY_LEVEL) {
-                        return ERR_PTR(-ENOMEM);
+                        metaphysical = 1;
-                shadow_pte = shadow_addr | PT_PRESENT_MASK;
+                        table_gfn = (*guest_ent & PT_BASE_ADDR_MASK)
-                if (vcpu->mmu.root_level > 3 || level != 3)
+                                >> PAGE_SHIFT;
-                        shadow_pte |= PT_ACCESSED_MASK
+                } else {
-                                | PT_WRITABLE_MASK | PT_USER_MASK;
+                        metaphysical = 0;
+                        table_gfn = walker->table_gfn[level - 2];
+                }
+                shadow_page = kvm_mmu_get_page(vcpu, table_gfn, addr, level-1,
+                                               metaphysical, shadow_ent);
+                shadow_addr = shadow_page->page_hpa;
+                shadow_pte = shadow_addr | PT_PRESENT_MASK | PT_ACCESSED_MASK
+                        | PT_WRITABLE_MASK | PT_USER_MASK;
                *shadow_ent = shadow_pte;
                prev_shadow_ent = shadow_ent;
        }
@@ -221,11 +265,13 @@ static int FNAME(fix_write_pf)(struct kvm_vcpu *vcpu,
                               u64 *shadow_ent,
                               struct guest_walker *walker,
                               gva_t addr,
-                               int user)
+                               int user,
+                               int *write_pt)
 {
        pt_element_t *guest_ent;
        int writable_shadow;
        gfn_t gfn;
+        struct kvm_mmu_page *page;
        if (is_writeble_pte(*shadow_ent))
                return 0;
@@ -250,17 +296,35 @@ static int FNAME(fix_write_pf)(struct kvm_vcpu *vcpu,
                        *shadow_ent &= ~PT_USER_MASK;
                }
-        guest_ent = FNAME(fetch_guest)(vcpu, walker, PT_PAGE_TABLE_LEVEL, addr);
+        guest_ent = walker->ptep;
        if (!is_present_pte(*guest_ent)) {
                *shadow_ent = 0;
                return 0;
        }
-        gfn = (*guest_ent & PT64_BASE_ADDR_MASK) >> PAGE_SHIFT;
+        gfn = walker->gfn;
+        if (user) {
+                /*
+                 * Usermode page faults won't be for page table updates.
+                 */
+                while ((page = kvm_mmu_lookup_page(vcpu, gfn)) != NULL) {
+                        pgprintk("%s: zap %lx %x\n",
+                                 __FUNCTION__, gfn, page->role.word);
+                        kvm_mmu_zap_page(vcpu, page);
+                }
+        } else if (kvm_mmu_lookup_page(vcpu, gfn)) {
+                pgprintk("%s: found shadow page for %lx, marking ro\n",
+                         __FUNCTION__, gfn);
+                *guest_ent |= PT_DIRTY_MASK;
+                *write_pt = 1;
+                return 0;
+        }
        mark_page_dirty(vcpu->kvm, gfn);
        *shadow_ent |= PT_WRITABLE_MASK;
        *guest_ent |= PT_DIRTY_MASK;
+        rmap_add(vcpu, shadow_ent);
        return 1;
 }
@@ -276,7 +340,8 @@ static int FNAME(fix_write_pf)(struct kvm_vcpu *vcpu,
 *   - normal guest page fault due to the guest pte marked not present, not
 *     writable, or not executable
 *
- *  Returns: 1 if we need to emulate the instruction, 0 otherwise
+ *  Returns: 1 if we need to emulate the instruction, 0 otherwise, or
+ *           a negative value on error.
 */
 static int FNAME(page_fault)(struct kvm_vcpu *vcpu, gva_t addr,
                               u32 error_code)
@@ -287,39 +352,47 @@ static int FNAME(page_fault)(struct kvm_vcpu *vcpu, gva_t addr,
        struct guest_walker walker;
        u64 *shadow_pte;
        int fixed;
+        int write_pt = 0;
+        int r;
+        pgprintk("%s: addr %lx err %x\n", __FUNCTION__, addr, error_code);
+        kvm_mmu_audit(vcpu, "pre page fault");
+        r = mmu_topup_memory_caches(vcpu);
+        if (r)
+                return r;
        /*
         * Look up the shadow pte for the faulting address.
         */
-        for (;;) {
+        FNAME(walk_addr)(&walker, vcpu, addr);
-                FNAME(init_walker)(&walker, vcpu);
+        shadow_pte = FNAME(fetch)(vcpu, addr, &walker);
-                shadow_pte = FNAME(fetch)(vcpu, addr, &walker);
-                if (IS_ERR(shadow_pte)) {  /* must be -ENOMEM */
-                        nonpaging_flush(vcpu);
-                        FNAME(release_walker)(&walker);
-                        continue;
-                }
-                break;
-        }
        /*
         * The page is not mapped by the guest.  Let the guest handle it.
         */
        if (!shadow_pte) {
+                pgprintk("%s: not mapped\n", __FUNCTION__);
                inject_page_fault(vcpu, addr, error_code);
                FNAME(release_walker)(&walker);
                return 0;
        }
+        pgprintk("%s: shadow pte %p %llx\n", __FUNCTION__,
+                 shadow_pte, *shadow_pte);
        /*
         * Update the shadow pte.
         */
        if (write_fault)
                fixed = FNAME(fix_write_pf)(vcpu, shadow_pte, &walker, addr,
-                                            user_fault);
+                                            user_fault, &write_pt);
        else
                fixed = fix_read_pf(shadow_pte);
+        pgprintk("%s: updated shadow pte %p %llx\n", __FUNCTION__,
+                 shadow_pte, *shadow_pte);
        FNAME(release_walker)(&walker);
        /*
@@ -331,20 +404,23 @@ static int FNAME(page_fault)(struct kvm_vcpu *vcpu, gva_t addr,
                pgprintk("%s: io work, no access\n", __FUNCTION__);
                inject_page_fault(vcpu, addr,
                                  error_code | PFERR_PRESENT_MASK);
+                kvm_mmu_audit(vcpu, "post page fault (io)");
                return 0;
        }
        /*
         * pte not present, guest page fault.
         */
-        if (pte_present && !fixed) {
+        if (pte_present && !fixed && !write_pt) {
                inject_page_fault(vcpu, addr, error_code);
+                kvm_mmu_audit(vcpu, "post page fault (guest)");
                return 0;
        }
        ++kvm_stat.pf_fixed;
+        kvm_mmu_audit(vcpu, "post page fault (fixed)");
-        return 0;
+        return write_pt;
 }
 static gpa_t FNAME(gva_to_gpa)(struct kvm_vcpu *vcpu, gva_t vaddr)
@@ -353,9 +429,8 @@ static gpa_t FNAME(gva_to_gpa)(struct kvm_vcpu *vcpu, gva_t vaddr)
        pt_element_t guest_pte;
        gpa_t gpa;
-        FNAME(init_walker)(&walker, vcpu);
+        FNAME(walk_addr)(&walker, vcpu, vaddr);
-        guest_pte = *FNAME(fetch_guest)(vcpu, &walker, PT_PAGE_TABLE_LEVEL,
+        guest_pte = *walker.ptep;
-                                        vaddr);
        FNAME(release_walker)(&walker);
        if (!is_present_pte(guest_pte))
@@ -389,3 +464,4 @@ static gpa_t FNAME(gva_to_gpa)(struct kvm_vcpu *vcpu, gva_t vaddr)
 #undef PT_PTE_COPY_MASK
 #undef PT_NON_PTE_COPY_MASK
 #undef PT_DIR_BASE_ADDR_MASK
+#undef PT_MAX_FULL_LEVELS
diff --git a/drivers/kvm/svm.c b/drivers/kvm/svm.c
index fa0428735717..ccc06b1b91b5 100644
--- a/drivers/kvm/svm.c
+++ b/drivers/kvm/svm.c
@@ -235,6 +235,8 @@ static void skip_emulated_instruction(struct kvm_vcpu *vcpu)
        vcpu->rip = vcpu->svm->vmcb->save.rip = vcpu->svm->next_rip;
        vcpu->svm->vmcb->control.int_state &= ~SVM_INTERRUPT_SHADOW_MASK;
+        vcpu->interrupt_window_open = 1;
 }
 static int has_svm(void)
@@ -495,7 +497,6 @@ static void init_vmcb(struct vmcb *vmcb)
                /*              (1ULL << INTERCEPT_SELECTIVE_CR0) | */
                                (1ULL << INTERCEPT_CPUID) |
                                (1ULL << INTERCEPT_HLT) |
-                                (1ULL << INTERCEPT_INVLPG) |
                                (1ULL << INTERCEPT_INVLPGA) |
                                (1ULL << INTERCEPT_IOIO_PROT) |
                                (1ULL << INTERCEPT_MSR_PROT) |
@@ -700,6 +701,10 @@ static void svm_set_gdt(struct kvm_vcpu *vcpu, struct descriptor_table *dt)
        vcpu->svm->vmcb->save.gdtr.base = dt->base ;
 }
+static void svm_decache_cr0_cr4_guest_bits(struct kvm_vcpu *vcpu)
+{
+}
 static void svm_set_cr0(struct kvm_vcpu *vcpu, unsigned long cr0)
 {
 #ifdef CONFIG_X86_64
@@ -847,6 +852,7 @@ static int pf_interception(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
        u64 fault_address;
        u32 error_code;
        enum emulation_result er;
+        int r;
        if (is_external_interrupt(exit_int_info))
                push_irq(vcpu, exit_int_info & SVM_EVTINJ_VEC_MASK);
@@ -855,7 +861,12 @@ static int pf_interception(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
        fault_address  = vcpu->svm->vmcb->control.exit_info_2;
        error_code = vcpu->svm->vmcb->control.exit_info_1;
-        if (!vcpu->mmu.page_fault(vcpu, fault_address, error_code)) {
+        r = kvm_mmu_page_fault(vcpu, fault_address, error_code);
+        if (r < 0) {
+                spin_unlock(&vcpu->kvm->lock);
+                return r;
+        }
+        if (!r) {
                spin_unlock(&vcpu->kvm->lock);
                return 1;
        }
@@ -1031,10 +1042,11 @@ static int halt_interception(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 {
        vcpu->svm->next_rip = vcpu->svm->vmcb->save.rip + 1;
        skip_emulated_instruction(vcpu);
-        if (vcpu->irq_summary && (vcpu->svm->vmcb->save.rflags & X86_EFLAGS_IF))
+        if (vcpu->irq_summary)
                return 1;
        kvm_run->exit_reason = KVM_EXIT_HLT;
+        ++kvm_stat.halt_exits;
        return 0;
 }
@@ -1186,6 +1198,23 @@ static int msr_interception(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
                return rdmsr_interception(vcpu, kvm_run);
 }
+static int interrupt_window_interception(struct kvm_vcpu *vcpu,
+                                   struct kvm_run *kvm_run)
+{
+        /*
+         * If the user space waits to inject interrupts, exit as soon as
+         * possible
+         */
+        if (kvm_run->request_interrupt_window &&
+            !vcpu->irq_summary) {
+                ++kvm_stat.irq_window_exits;
+                kvm_run->exit_reason = KVM_EXIT_IRQ_WINDOW_OPEN;
+                return 0;
+        }
+        return 1;
+}
 static int (*svm_exit_handlers[])(struct kvm_vcpu *vcpu,
                                      struct kvm_run *kvm_run) = {
        [SVM_EXIT_READ_CR0]                     = emulate_on_interception,
@@ -1210,6 +1239,7 @@ static int (*svm_exit_handlers[])(struct kvm_vcpu *vcpu,
        [SVM_EXIT_NMI]                          = nop_on_interception,
        [SVM_EXIT_SMI]                          = nop_on_interception,
        [SVM_EXIT_INIT]                         = nop_on_interception,
+        [SVM_EXIT_VINTR]                        = interrupt_window_interception,
        /* [SVM_EXIT_CR0_SEL_WRITE]             = emulate_on_interception, */
        [SVM_EXIT_CPUID]                        = cpuid_interception,
        [SVM_EXIT_HLT]                          = halt_interception,
@@ -1278,15 +1308,11 @@ static void pre_svm_run(struct kvm_vcpu *vcpu)
 }
-static inline void kvm_try_inject_irq(struct kvm_vcpu *vcpu)
+static inline void kvm_do_inject_irq(struct kvm_vcpu *vcpu)
 {
        struct vmcb_control_area *control;
-        if (!vcpu->irq_summary)
-                return;
        control = &vcpu->svm->vmcb->control;
        control->int_vector = pop_irq(vcpu);
        control->int_ctl &= ~V_INTR_PRIO_MASK;
        control->int_ctl |= V_IRQ_MASK |
@@ -1301,6 +1327,59 @@ static void kvm_reput_irq(struct kvm_vcpu *vcpu)
                control->int_ctl &= ~V_IRQ_MASK;
                push_irq(vcpu, control->int_vector);
        }
+        vcpu->interrupt_window_open =
+                !(control->int_state & SVM_INTERRUPT_SHADOW_MASK);
+}
+static void do_interrupt_requests(struct kvm_vcpu *vcpu,
+                                       struct kvm_run *kvm_run)
+{
+        struct vmcb_control_area *control = &vcpu->svm->vmcb->control;
+        vcpu->interrupt_window_open =
+                (!(control->int_state & SVM_INTERRUPT_SHADOW_MASK) &&
+                 (vcpu->svm->vmcb->save.rflags & X86_EFLAGS_IF));
+        if (vcpu->interrupt_window_open && vcpu->irq_summary)
+                /*
+                 * If interrupts enabled, and not blocked by sti or mov ss. Good.
+                 */
+                kvm_do_inject_irq(vcpu);
+        /*
+         * Interrupts blocked.  Wait for unblock.
+         */
+        if (!vcpu->interrupt_window_open &&
+            (vcpu->irq_summary || kvm_run->request_interrupt_window)) {
+                control->intercept |= 1ULL << INTERCEPT_VINTR;
+        } else
+                control->intercept &= ~(1ULL << INTERCEPT_VINTR);
+}
+static void post_kvm_run_save(struct kvm_vcpu *vcpu,
+                              struct kvm_run *kvm_run)
+{
+        kvm_run->ready_for_interrupt_injection = (vcpu->interrupt_window_open &&
+                                                  vcpu->irq_summary == 0);
+        kvm_run->if_flag = (vcpu->svm->vmcb->save.rflags & X86_EFLAGS_IF) != 0;
+        kvm_run->cr8 = vcpu->cr8;
+        kvm_run->apic_base = vcpu->apic_base;
+}
+/*
+ * Check if userspace requested an interrupt window, and that the
+ * interrupt window is open.
+ *
+ * No need to exit to userspace if we already have an interrupt queued.
+ */
+static int dm_request_for_irq_injection(struct kvm_vcpu *vcpu,
+                                          struct kvm_run *kvm_run)
+{
+        return (!vcpu->irq_summary &&
+                kvm_run->request_interrupt_window &&
+                vcpu->interrupt_window_open &&
+                (vcpu->svm->vmcb->save.rflags & X86_EFLAGS_IF));
 }
 static void save_db_regs(unsigned long *db_regs)
@@ -1324,9 +1403,10 @@ static int svm_vcpu_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
        u16 fs_selector;
        u16 gs_selector;
        u16 ldt_selector;
+        int r;
 again:
-        kvm_try_inject_irq(vcpu);
+        do_interrupt_requests(vcpu, kvm_run);
        clgi();
@@ -1487,18 +1567,28 @@ again:
        if (vcpu->svm->vmcb->control.exit_code == SVM_EXIT_ERR) {
                kvm_run->exit_type = KVM_EXIT_TYPE_FAIL_ENTRY;
                kvm_run->exit_reason = vcpu->svm->vmcb->control.exit_code;
+                post_kvm_run_save(vcpu, kvm_run);
                return 0;
        }
-        if (handle_exit(vcpu, kvm_run)) {
+        r = handle_exit(vcpu, kvm_run);
+        if (r > 0) {
                if (signal_pending(current)) {
                        ++kvm_stat.signal_exits;
+                        post_kvm_run_save(vcpu, kvm_run);
+                        return -EINTR;
+                }
+                if (dm_request_for_irq_injection(vcpu, kvm_run)) {
+                        ++kvm_stat.request_irq_exits;
+                        post_kvm_run_save(vcpu, kvm_run);
                        return -EINTR;
                }
                kvm_resched(vcpu);
                goto again;
        }
-        return 0;
+        post_kvm_run_save(vcpu, kvm_run);
+        return r;
 }
 static void svm_flush_tlb(struct kvm_vcpu *vcpu)
@@ -1565,6 +1655,7 @@ static struct kvm_arch_ops svm_arch_ops = {
        .get_segment = svm_get_segment,
        .set_segment = svm_set_segment,
        .get_cs_db_l_bits = svm_get_cs_db_l_bits,
+        .decache_cr0_cr4_guest_bits = svm_decache_cr0_cr4_guest_bits,
        .set_cr0 = svm_set_cr0,
        .set_cr0_no_modeswitch = svm_set_cr0,
        .set_cr3 = svm_set_cr3,
diff --git a/drivers/kvm/vmx.c b/drivers/kvm/vmx.c
index d0a2c2d5342a..d4701cb4c654 100644
--- a/drivers/kvm/vmx.c
+++ b/drivers/kvm/vmx.c
@@ -116,7 +116,7 @@ static void vmcs_clear(struct vmcs *vmcs)
 static void __vcpu_clear(void *arg)
 {
        struct kvm_vcpu *vcpu = arg;
-        int cpu = smp_processor_id();
+        int cpu = raw_smp_processor_id();
        if (vcpu->cpu == cpu)
                vmcs_clear(vcpu->vmcs);
@@ -152,15 +152,21 @@ static u64 vmcs_read64(unsigned long field)
 #endif
 }
+static noinline void vmwrite_error(unsigned long field, unsigned long value)
+{
+        printk(KERN_ERR "vmwrite error: reg %lx value %lx (err %d)\n",
+               field, value, vmcs_read32(VM_INSTRUCTION_ERROR));
+        dump_stack();
+}
 static void vmcs_writel(unsigned long field, unsigned long value)
 {
        u8 error;
        asm volatile (ASM_VMX_VMWRITE_RAX_RDX "; setna %0"
                       : "=q"(error) : "a"(value), "d"(field) : "cc" );
-        if (error)
+        if (unlikely(error))
-                printk(KERN_ERR "vmwrite error: reg %lx value %lx (err %d)\n",
+                vmwrite_error(field, value);
-                       field, value, vmcs_read32(VM_INSTRUCTION_ERROR));
 }
 static void vmcs_write16(unsigned long field, u16 value)
@@ -263,6 +269,7 @@ static void skip_emulated_instruction(struct kvm_vcpu *vcpu)
        if (interruptibility & 3)
                vmcs_write32(GUEST_INTERRUPTIBILITY_INFO,
                             interruptibility & ~3);
+        vcpu->interrupt_window_open = 1;
 }
 static void vmx_inject_gp(struct kvm_vcpu *vcpu, unsigned error_code)
@@ -541,7 +548,7 @@ static struct vmcs *alloc_vmcs_cpu(int cpu)
 static struct vmcs *alloc_vmcs(void)
 {
-        return alloc_vmcs_cpu(smp_processor_id());
+        return alloc_vmcs_cpu(raw_smp_processor_id());
 }
 static void free_vmcs(struct vmcs *vmcs)
@@ -736,6 +743,15 @@ static void exit_lmode(struct kvm_vcpu *vcpu)
 #endif
+static void vmx_decache_cr0_cr4_guest_bits(struct kvm_vcpu *vcpu)
+{
+        vcpu->cr0 &= KVM_GUEST_CR0_MASK;
+        vcpu->cr0 |= vmcs_readl(GUEST_CR0) & ~KVM_GUEST_CR0_MASK;
+        vcpu->cr4 &= KVM_GUEST_CR4_MASK;
+        vcpu->cr4 |= vmcs_readl(GUEST_CR4) & ~KVM_GUEST_CR4_MASK;
+}
 static void vmx_set_cr0(struct kvm_vcpu *vcpu, unsigned long cr0)
 {
        if (vcpu->rmode.active && (cr0 & CR0_PE_MASK))
@@ -1011,8 +1027,6 @@ static int vmx_vcpu_setup(struct kvm_vcpu *vcpu)
        vmcs_writel(GUEST_RIP, 0xfff0);
        vmcs_writel(GUEST_RSP, 0);
-        vmcs_writel(GUEST_CR3, 0);
        //todo: dr0 = dr1 = dr2 = dr3 = 0; dr6 = 0xffff0ff0
        vmcs_writel(GUEST_DR7, 0x400);
@@ -1049,7 +1063,6 @@ static int vmx_vcpu_setup(struct kvm_vcpu *vcpu)
                               | CPU_BASED_CR8_LOAD_EXITING    /* 20.6.2 */
                               | CPU_BASED_CR8_STORE_EXITING   /* 20.6.2 */
                               | CPU_BASED_UNCOND_IO_EXITING   /* 20.6.2 */
-                               | CPU_BASED_INVDPG_EXITING
                               | CPU_BASED_MOV_DR_EXITING
                               | CPU_BASED_USE_TSC_OFFSETING   /* 21.3 */
                        );
@@ -1094,14 +1107,6 @@ static int vmx_vcpu_setup(struct kvm_vcpu *vcpu)
        rdmsrl(MSR_IA32_SYSENTER_EIP, a);
        vmcs_writel(HOST_IA32_SYSENTER_EIP, a);   /* 22.2.3 */
-        ret = -ENOMEM;
-        vcpu->guest_msrs = kmalloc(PAGE_SIZE, GFP_KERNEL);
-        if (!vcpu->guest_msrs)
-                goto out;
-        vcpu->host_msrs = kmalloc(PAGE_SIZE, GFP_KERNEL);
-        if (!vcpu->host_msrs)
-                goto out_free_guest_msrs;
        for (i = 0; i < NR_VMX_MSR; ++i) {
                u32 index = vmx_msr_index[i];
                u32 data_low, data_high;
@@ -1155,8 +1160,6 @@ static int vmx_vcpu_setup(struct kvm_vcpu *vcpu)
        return 0;
-out_free_guest_msrs:
-        kfree(vcpu->guest_msrs);
 out:
        return ret;
 }
@@ -1224,21 +1227,34 @@ static void kvm_do_inject_irq(struct kvm_vcpu *vcpu)
                        irq | INTR_TYPE_EXT_INTR | INTR_INFO_VALID_MASK);
 }
-static void kvm_try_inject_irq(struct kvm_vcpu *vcpu)
+static void do_interrupt_requests(struct kvm_vcpu *vcpu,
+                                       struct kvm_run *kvm_run)
 {
-        if ((vmcs_readl(GUEST_RFLAGS) & X86_EFLAGS_IF)
+        u32 cpu_based_vm_exec_control;
-            && (vmcs_read32(GUEST_INTERRUPTIBILITY_INFO) & 3) == 0)
+        vcpu->interrupt_window_open =
+                ((vmcs_readl(GUEST_RFLAGS) & X86_EFLAGS_IF) &&
+                 (vmcs_read32(GUEST_INTERRUPTIBILITY_INFO) & 3) == 0);
+        if (vcpu->interrupt_window_open &&
+            vcpu->irq_summary &&
+            !(vmcs_read32(VM_ENTRY_INTR_INFO_FIELD) & INTR_INFO_VALID_MASK))
                /*
-                 * Interrupts enabled, and not blocked by sti or mov ss. Good.
+                 * If interrupts enabled, and not blocked by sti or mov ss. Good.
                 */
                kvm_do_inject_irq(vcpu);
-        else
+        cpu_based_vm_exec_control = vmcs_read32(CPU_BASED_VM_EXEC_CONTROL);
+        if (!vcpu->interrupt_window_open &&
+            (vcpu->irq_summary || kvm_run->request_interrupt_window))
                /*
                 * Interrupts blocked.  Wait for unblock.
                 */
-                vmcs_write32(CPU_BASED_VM_EXEC_CONTROL,
+                cpu_based_vm_exec_control |= CPU_BASED_VIRTUAL_INTR_PENDING;
-                             vmcs_read32(CPU_BASED_VM_EXEC_CONTROL)
+        else
-                             | CPU_BASED_VIRTUAL_INTR_PENDING);
+                cpu_based_vm_exec_control &= ~CPU_BASED_VIRTUAL_INTR_PENDING;
+        vmcs_write32(CPU_BASED_VM_EXEC_CONTROL, cpu_based_vm_exec_control);
 }
 static void kvm_guest_debug_pre(struct kvm_vcpu *vcpu)
@@ -1277,6 +1293,7 @@ static int handle_exception(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
        unsigned long cr2, rip;
        u32 vect_info;
        enum emulation_result er;
+        int r;
        vect_info = vmcs_read32(IDT_VECTORING_INFO_FIELD);
        intr_info = vmcs_read32(VM_EXIT_INTR_INFO);
@@ -1305,7 +1322,12 @@ static int handle_exception(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
                cr2 = vmcs_readl(EXIT_QUALIFICATION);
                spin_lock(&vcpu->kvm->lock);
-                if (!vcpu->mmu.page_fault(vcpu, cr2, error_code)) {
+                r = kvm_mmu_page_fault(vcpu, cr2, error_code);
+                if (r < 0) {
+                        spin_unlock(&vcpu->kvm->lock);
+                        return r;
+                }
+                if (!r) {
                        spin_unlock(&vcpu->kvm->lock);
                        return 1;
                }
@@ -1425,17 +1447,6 @@ static int handle_io(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
        return 0;
 }
-static int handle_invlpg(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
-{
-        u64 address = vmcs_read64(EXIT_QUALIFICATION);
-        int instruction_length = vmcs_read32(VM_EXIT_INSTRUCTION_LEN);
-        spin_lock(&vcpu->kvm->lock);
-        vcpu->mmu.inval_page(vcpu, address);
-        spin_unlock(&vcpu->kvm->lock);
-        vmcs_writel(GUEST_RIP, vmcs_readl(GUEST_RIP) + instruction_length);
-        return 1;
-}
 static int handle_cr(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 {
        u64 exit_qualification;
@@ -1575,23 +1586,40 @@ static int handle_wrmsr(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
        return 1;
 }
+static void post_kvm_run_save(struct kvm_vcpu *vcpu,
+                              struct kvm_run *kvm_run)
+{
+        kvm_run->if_flag = (vmcs_readl(GUEST_RFLAGS) & X86_EFLAGS_IF) != 0;
+        kvm_run->cr8 = vcpu->cr8;
+        kvm_run->apic_base = vcpu->apic_base;
+        kvm_run->ready_for_interrupt_injection = (vcpu->interrupt_window_open &&
+                                                  vcpu->irq_summary == 0);
+}
 static int handle_interrupt_window(struct kvm_vcpu *vcpu,
                                   struct kvm_run *kvm_run)
 {
-        /* Turn off interrupt window reporting. */
+        /*
-        vmcs_write32(CPU_BASED_VM_EXEC_CONTROL,
+         * If the user space waits to inject interrupts, exit as soon as
-                     vmcs_read32(CPU_BASED_VM_EXEC_CONTROL)
+         * possible
-                     & ~CPU_BASED_VIRTUAL_INTR_PENDING);
+         */
+        if (kvm_run->request_interrupt_window &&
+            !vcpu->irq_summary) {
+                kvm_run->exit_reason = KVM_EXIT_IRQ_WINDOW_OPEN;
+                ++kvm_stat.irq_window_exits;
+                return 0;
+        }
        return 1;
 }
 static int handle_halt(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 {
        skip_emulated_instruction(vcpu);
-        if (vcpu->irq_summary && (vmcs_readl(GUEST_RFLAGS) & X86_EFLAGS_IF))
+        if (vcpu->irq_summary)
                return 1;
        kvm_run->exit_reason = KVM_EXIT_HLT;
+        ++kvm_stat.halt_exits;
        return 0;
 }
@@ -1605,7 +1633,6 @@ static int (*kvm_vmx_exit_handlers[])(struct kvm_vcpu *vcpu,
        [EXIT_REASON_EXCEPTION_NMI]           = handle_exception,
        [EXIT_REASON_EXTERNAL_INTERRUPT]      = handle_external_interrupt,
        [EXIT_REASON_IO_INSTRUCTION]          = handle_io,
-        [EXIT_REASON_INVLPG]                  = handle_invlpg,
        [EXIT_REASON_CR_ACCESS]               = handle_cr,
        [EXIT_REASON_DR_ACCESS]               = handle_dr,
        [EXIT_REASON_CPUID]                   = handle_cpuid,
@@ -1642,11 +1669,27 @@ static int kvm_handle_exit(struct kvm_run *kvm_run, struct kvm_vcpu *vcpu)
        return 0;
 }
+/*
+ * Check if userspace requested an interrupt window, and that the
+ * interrupt window is open.
+ *
+ * No need to exit to userspace if we already have an interrupt queued.
+ */
+static int dm_request_for_irq_injection(struct kvm_vcpu *vcpu,
+                                          struct kvm_run *kvm_run)
+{
+        return (!vcpu->irq_summary &&
+                kvm_run->request_interrupt_window &&
+                vcpu->interrupt_window_open &&
+                (vmcs_readl(GUEST_RFLAGS) & X86_EFLAGS_IF));
+}
 static int vmx_vcpu_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 {
        u8 fail;
        u16 fs_sel, gs_sel, ldt_sel;
        int fs_gs_ldt_reload_needed;
+        int r;
 again:
        /*
@@ -1673,9 +1716,7 @@ again:
        vmcs_writel(HOST_GS_BASE, segment_base(gs_sel));
 #endif
-        if (vcpu->irq_summary &&
+        do_interrupt_requests(vcpu, kvm_run);
-            !(vmcs_read32(VM_ENTRY_INTR_INFO_FIELD) & INTR_INFO_VALID_MASK))
-                kvm_try_inject_irq(vcpu);
        if (vcpu->guest_debug.enabled)
                kvm_guest_debug_pre(vcpu);
@@ -1812,6 +1853,7 @@ again:
        fx_save(vcpu->guest_fx_image);
        fx_restore(vcpu->host_fx_image);
+        vcpu->interrupt_window_open = (vmcs_read32(GUEST_INTERRUPTIBILITY_INFO) & 3) == 0;
 #ifndef CONFIG_X86_64
        asm ("mov %0, %%ds; mov %0, %%es" : : "r"(__USER_DS));
@@ -1821,6 +1863,7 @@ again:
        if (fail) {
                kvm_run->exit_type = KVM_EXIT_TYPE_FAIL_ENTRY;
                kvm_run->exit_reason = vmcs_read32(VM_INSTRUCTION_ERROR);
+                r = 0;
        } else {
                if (fs_gs_ldt_reload_needed) {
                        load_ldt(ldt_sel);
@@ -1840,17 +1883,28 @@ again:
                }
                vcpu->launched = 1;
                kvm_run->exit_type = KVM_EXIT_TYPE_VM_EXIT;
-                if (kvm_handle_exit(kvm_run, vcpu)) {
+                r = kvm_handle_exit(kvm_run, vcpu);
+                if (r > 0) {
                        /* Give scheduler a change to reschedule. */
                        if (signal_pending(current)) {
                                ++kvm_stat.signal_exits;
+                                post_kvm_run_save(vcpu, kvm_run);
+                                return -EINTR;
+                        }
+                        if (dm_request_for_irq_injection(vcpu, kvm_run)) {
+                                ++kvm_stat.request_irq_exits;
+                                post_kvm_run_save(vcpu, kvm_run);
                                return -EINTR;
                        }
                        kvm_resched(vcpu);
                        goto again;
                }
        }
-        return 0;
+        post_kvm_run_save(vcpu, kvm_run);
+        return r;
 }
 static void vmx_flush_tlb(struct kvm_vcpu *vcpu)
@@ -1906,13 +1960,33 @@ static int vmx_create_vcpu(struct kvm_vcpu *vcpu)
 {
        struct vmcs *vmcs;
+        vcpu->guest_msrs = kmalloc(PAGE_SIZE, GFP_KERNEL);
+        if (!vcpu->guest_msrs)
+                return -ENOMEM;
+        vcpu->host_msrs = kmalloc(PAGE_SIZE, GFP_KERNEL);
+        if (!vcpu->host_msrs)
+                goto out_free_guest_msrs;
        vmcs = alloc_vmcs();
        if (!vmcs)
-                return -ENOMEM;
+                goto out_free_msrs;
        vmcs_clear(vmcs);
        vcpu->vmcs = vmcs;
        vcpu->launched = 0;
        return 0;
+out_free_msrs:
+        kfree(vcpu->host_msrs);
+        vcpu->host_msrs = NULL;
+out_free_guest_msrs:
+        kfree(vcpu->guest_msrs);
+        vcpu->guest_msrs = NULL;
+        return -ENOMEM;
 }
 static struct kvm_arch_ops vmx_arch_ops = {
@@ -1936,6 +2010,7 @@ static struct kvm_arch_ops vmx_arch_ops = {
        .get_segment = vmx_get_segment,
        .set_segment = vmx_set_segment,
        .get_cs_db_l_bits = vmx_get_cs_db_l_bits,
+        .decache_cr0_cr4_guest_bits = vmx_decache_cr0_cr4_guest_bits,
        .set_cr0 = vmx_set_cr0,
        .set_cr0_no_modeswitch = vmx_set_cr0_no_modeswitch,
        .set_cr3 = vmx_set_cr3,
diff --git a/drivers/kvm/x86_emulate.c b/drivers/kvm/x86_emulate.c
index 1bff3e925fda..be70795b4822 100644
--- a/drivers/kvm/x86_emulate.c
+++ b/drivers/kvm/x86_emulate.c
@@ -1323,7 +1323,7 @@ twobyte_special_insn:
                                                         ctxt)) != 0))
                                goto done;
                        if ((old_lo != _regs[VCPU_REGS_RAX])
-                            || (old_hi != _regs[VCPU_REGS_RDI])) {
+                            || (old_hi != _regs[VCPU_REGS_RDX])) {
                                _regs[VCPU_REGS_RAX] = old_lo;
                                _regs[VCPU_REGS_RDX] = old_hi;
                                _eflags &= ~EFLG_ZF;
diff --git a/drivers/leds/leds-s3c24xx.c b/drivers/leds/leds-s3c24xx.c
index fb1edc1c9edb..50914439d861 100644
--- a/drivers/leds/leds-s3c24xx.c
+++ b/drivers/leds/leds-s3c24xx.c
@@ -16,7 +16,7 @@
 #include <linux/platform_device.h>
 #include <linux/leds.h>
-#include <asm/arch/hardware.h>
+#include <asm/hardware.h>
 #include <asm/arch/regs-gpio.h>
 #include <asm/arch/leds-gpio.h>
diff --git a/drivers/macintosh/via-pmu.c b/drivers/macintosh/via-pmu.c
index c8558d4ed506..8ca75e52f637 100644
--- a/drivers/macintosh/via-pmu.c
+++ b/drivers/macintosh/via-pmu.c
@@ -44,6 +44,7 @@
 #include <linux/sysdev.h>
 #include <linux/freezer.h>
 #include <linux/syscalls.h>
+#include <linux/suspend.h>
 #include <linux/cpu.h>
 #include <asm/prom.h>
 #include <asm/machdep.h>
diff --git a/drivers/mmc/mmci.c b/drivers/mmc/mmci.c
index e9b80e920266..ccfe6561be24 100644
--- a/drivers/mmc/mmci.c
+++ b/drivers/mmc/mmci.c
@@ -42,6 +42,8 @@ mmci_request_end(struct mmci_host *host, struct mmc_request *mrq)
 {
        writel(0, host->base + MMCICOMMAND);
+        BUG_ON(host->data);
        host->mrq = NULL;
        host->cmd = NULL;
@@ -198,6 +200,8 @@ mmci_cmd_irq(struct mmci_host *host, struct mmc_command *cmd,
        }
        if (!cmd->data || cmd->error != MMC_ERR_NONE) {
+                if (host->data)
+                        mmci_stop_data(host);
                mmci_request_end(host, cmd->mrq);
        } else if (!(cmd->data->flags & MMC_DATA_READ)) {
                mmci_start_data(host, cmd->data);
diff --git a/drivers/net/Space.c b/drivers/net/Space.c
index 602ed31a5dd9..9305eb9b1b98 100644
--- a/drivers/net/Space.c
+++ b/drivers/net/Space.c
@@ -349,22 +349,11 @@ static void __init trif_probe2(int unit)
 #endif
-/*
- *      The loopback device is global so it can be directly referenced
- *      by the network code. Also, it must be first on device list.
- */
-extern int loopback_init(void);
 /*  Statically configured drivers -- order matters here. */
 static int __init net_olddevs_init(void)
 {
        int num;
-        if (loopback_init()) {
-                printk(KERN_ERR "Network loopback device setup failed\n");
-        }
 #ifdef CONFIG_SBNI
        for (num = 0; num < 8; ++num)
                sbni_probe(num);
diff --git a/drivers/net/bnx2.c b/drivers/net/bnx2.c
index ada5e9b9988c..ca5acc4736df 100644
--- a/drivers/net/bnx2.c
+++ b/drivers/net/bnx2.c
@@ -57,8 +57,8 @@
 #define DRV_MODULE_NAME         "bnx2"
 #define PFX DRV_MODULE_NAME     ": "
-#define DRV_MODULE_VERSION      "1.5.2"
+#define DRV_MODULE_VERSION      "1.5.3"
-#define DRV_MODULE_RELDATE      "December 13, 2006"
+#define DRV_MODULE_RELDATE      "January 8, 2007"
 #define RUN_AT(x) (jiffies + (x))
@@ -1345,8 +1345,6 @@ bnx2_init_copper_phy(struct bnx2 *bp)
 {
        u32 val;
-        bp->phy_flags |= PHY_CRC_FIX_FLAG;
        if (bp->phy_flags & PHY_CRC_FIX_FLAG) {
                bnx2_write_phy(bp, 0x18, 0x0c00);
                bnx2_write_phy(bp, 0x17, 0x000a);
@@ -3085,7 +3083,7 @@ bnx2_nvram_write(struct bnx2 *bp, u32 offset, u8 *data_buf,
                int buf_size)
 {
        u32 written, offset32, len32;
-        u8 *buf, start[4], end[4], *flash_buffer = NULL;
+        u8 *buf, start[4], end[4], *align_buf = NULL, *flash_buffer = NULL;
        int rc = 0;
        int align_start, align_end;
@@ -3113,16 +3111,17 @@ bnx2_nvram_write(struct bnx2 *bp, u32 offset, u8 *data_buf,
        }
        if (align_start || align_end) {
-                buf = kmalloc(len32, GFP_KERNEL);
+                align_buf = kmalloc(len32, GFP_KERNEL);
-                if (buf == NULL)
+                if (align_buf == NULL)
                        return -ENOMEM;
                if (align_start) {
-                        memcpy(buf, start, 4);
+                        memcpy(align_buf, start, 4);
                }
                if (align_end) {
-                        memcpy(buf + len32 - 4, end, 4);
+                        memcpy(align_buf + len32 - 4, end, 4);
                }
-                memcpy(buf + align_start, data_buf, buf_size);
+                memcpy(align_buf + align_start, data_buf, buf_size);
+                buf = align_buf;
        }
        if (bp->flash_info->buffered == 0) {
@@ -3256,11 +3255,8 @@ bnx2_nvram_write(struct bnx2 *bp, u32 offset, u8 *data_buf,
        }
 nvram_write_end:
-        if (bp->flash_info->buffered == 0)
+        kfree(flash_buffer);
-                kfree(flash_buffer);
+        kfree(align_buf);
-        if (align_start || align_end)
-                kfree(buf);
        return rc;
 }
@@ -5645,6 +5641,44 @@ poll_bnx2(struct net_device *dev)
 }
 #endif
+static void __devinit
+bnx2_get_5709_media(struct bnx2 *bp)
+{
+        u32 val = REG_RD(bp, BNX2_MISC_DUAL_MEDIA_CTRL);
+        u32 bond_id = val & BNX2_MISC_DUAL_MEDIA_CTRL_BOND_ID;
+        u32 strap;
+        if (bond_id == BNX2_MISC_DUAL_MEDIA_CTRL_BOND_ID_C)
+                return;
+        else if (bond_id == BNX2_MISC_DUAL_MEDIA_CTRL_BOND_ID_S) {
+                bp->phy_flags |= PHY_SERDES_FLAG;
+                return;
+        }
+        if (val & BNX2_MISC_DUAL_MEDIA_CTRL_STRAP_OVERRIDE)
+                strap = (val & BNX2_MISC_DUAL_MEDIA_CTRL_PHY_CTRL) >> 21;
+        else
+                strap = (val & BNX2_MISC_DUAL_MEDIA_CTRL_PHY_CTRL_STRAP) >> 8;
+        if (PCI_FUNC(bp->pdev->devfn) == 0) {
+                switch (strap) {
+                case 0x4:
+                case 0x5:
+                case 0x6:
+                        bp->phy_flags |= PHY_SERDES_FLAG;
+                        return;
+                }
+        } else {
+                switch (strap) {
+                case 0x1:
+                case 0x2:
+                case 0x4:
+                        bp->phy_flags |= PHY_SERDES_FLAG;
+                        return;
+                }
+        }
+}
 static int __devinit
 bnx2_init_board(struct pci_dev *pdev, struct net_device *dev)
 {
@@ -5865,10 +5899,9 @@ bnx2_init_board(struct pci_dev *pdev, struct net_device *dev)
        bp->phy_addr = 1;
        /* Disable WOL support if we are running on a SERDES chip. */
-        if (CHIP_NUM(bp) == CHIP_NUM_5709) {
+        if (CHIP_NUM(bp) == CHIP_NUM_5709)
-                if (CHIP_BOND_ID(bp) != BNX2_MISC_DUAL_MEDIA_CTRL_BOND_ID_C)
+                bnx2_get_5709_media(bp);
-                        bp->phy_flags |= PHY_SERDES_FLAG;
+        else if (CHIP_BOND_ID(bp) & CHIP_BOND_ID_SERDES_BIT)
-        } else if (CHIP_BOND_ID(bp) & CHIP_BOND_ID_SERDES_BIT)
                bp->phy_flags |= PHY_SERDES_FLAG;
        if (bp->phy_flags & PHY_SERDES_FLAG) {
@@ -5880,7 +5913,9 @@ bnx2_init_board(struct pci_dev *pdev, struct net_device *dev)
                        if (reg & BNX2_SHARED_HW_CFG_PHY_2_5G)
                                bp->phy_flags |= PHY_2_5G_CAPABLE_FLAG;
                }
-        }
+        } else if (CHIP_NUM(bp) == CHIP_NUM_5706 ||
+                   CHIP_NUM(bp) == CHIP_NUM_5708)
+                bp->phy_flags |= PHY_CRC_FIX_FLAG;
        if ((CHIP_ID(bp) == CHIP_ID_5708_A0) ||
            (CHIP_ID(bp) == CHIP_ID_5708_B0) ||
diff --git a/drivers/net/chelsio/my3126.c b/drivers/net/chelsio/my3126.c
index c7731b6f9de3..82fed1dd5005 100644
--- a/drivers/net/chelsio/my3126.c
+++ b/drivers/net/chelsio/my3126.c
@@ -170,9 +170,10 @@ static struct cphy *my3126_phy_create(adapter_t *adapter,
 {
        struct cphy *cphy = kzalloc(sizeof (*cphy), GFP_KERNEL);
-        if (cphy)
+        if (!cphy)
-                cphy_init(cphy, adapter, phy_addr, &my3126_ops, mdio_ops);
+                return NULL;
+        cphy_init(cphy, adapter, phy_addr, &my3126_ops, mdio_ops);
        INIT_DELAYED_WORK(&cphy->phy_update, my3216_poll);
        cphy->bmsr = 0;
diff --git a/drivers/net/e1000/e1000_main.c b/drivers/net/e1000/e1000_main.c
index 4c1ff752048c..c6259c7127f6 100644
--- a/drivers/net/e1000/e1000_main.c
+++ b/drivers/net/e1000/e1000_main.c
@@ -995,12 +995,6 @@ e1000_probe(struct pci_dev *pdev,
           (adapter->hw.mac_type != e1000_82547))
                netdev->features |= NETIF_F_TSO;
-#ifdef CONFIG_DEBUG_SLAB
-        /* 82544's work arounds do not play nicely with DEBUG SLAB */
-        if (adapter->hw.mac_type == e1000_82544)
-                netdev->features &= ~NETIF_F_TSO;
-#endif
 #ifdef NETIF_F_TSO6
        if (adapter->hw.mac_type > e1000_82547_rev_2)
                netdev->features |= NETIF_F_TSO6;
diff --git a/drivers/net/forcedeth.c b/drivers/net/forcedeth.c
index 2f48fe9a29a7..93f2b7a22160 100644
--- a/drivers/net/forcedeth.c
+++ b/drivers/net/forcedeth.c
@@ -234,6 +234,7 @@ enum {
 #define NVREG_XMITCTL_HOST_SEMA_MASK    0x0000f000
 #define NVREG_XMITCTL_HOST_SEMA_ACQ     0x0000f000
 #define NVREG_XMITCTL_HOST_LOADED       0x00004000
+#define NVREG_XMITCTL_TX_PATH_EN        0x01000000
        NvRegTransmitterStatus = 0x088,
 #define NVREG_XMITSTAT_BUSY     0x01
@@ -249,6 +250,7 @@ enum {
 #define NVREG_OFFLOAD_NORMAL    RX_NIC_BUFSIZE
        NvRegReceiverControl = 0x094,
 #define NVREG_RCVCTL_START      0x01
+#define NVREG_RCVCTL_RX_PATH_EN 0x01000000
        NvRegReceiverStatus = 0x98,
 #define NVREG_RCVSTAT_BUSY      0x01
@@ -1169,16 +1171,21 @@ static void nv_start_rx(struct net_device *dev)
 {
        struct fe_priv *np = netdev_priv(dev);
        u8 __iomem *base = get_hwbase(dev);
+        u32 rx_ctrl = readl(base + NvRegReceiverControl);
        dprintk(KERN_DEBUG "%s: nv_start_rx\n", dev->name);
        /* Already running? Stop it. */
-        if (readl(base + NvRegReceiverControl) & NVREG_RCVCTL_START) {
+        if ((readl(base + NvRegReceiverControl) & NVREG_RCVCTL_START) && !np->mac_in_use) {
-                writel(0, base + NvRegReceiverControl);
+                rx_ctrl &= ~NVREG_RCVCTL_START;
+                writel(rx_ctrl, base + NvRegReceiverControl);
                pci_push(base);
        }
        writel(np->linkspeed, base + NvRegLinkSpeed);
        pci_push(base);
-        writel(NVREG_RCVCTL_START, base + NvRegReceiverControl);
+        rx_ctrl |= NVREG_RCVCTL_START;
+        if (np->mac_in_use)
+                rx_ctrl &= ~NVREG_RCVCTL_RX_PATH_EN;
+        writel(rx_ctrl, base + NvRegReceiverControl);
        dprintk(KERN_DEBUG "%s: nv_start_rx to duplex %d, speed 0x%08x.\n",
                                dev->name, np->duplex, np->linkspeed);
        pci_push(base);
@@ -1186,39 +1193,59 @@ static void nv_start_rx(struct net_device *dev)
 static void nv_stop_rx(struct net_device *dev)
 {
+        struct fe_priv *np = netdev_priv(dev);
        u8 __iomem *base = get_hwbase(dev);
+        u32 rx_ctrl = readl(base + NvRegReceiverControl);
        dprintk(KERN_DEBUG "%s: nv_stop_rx\n", dev->name);
-        writel(0, base + NvRegReceiverControl);
+        if (!np->mac_in_use)
+                rx_ctrl &= ~NVREG_RCVCTL_START;
+        else
+                rx_ctrl |= NVREG_RCVCTL_RX_PATH_EN;
+        writel(rx_ctrl, base + NvRegReceiverControl);
        reg_delay(dev, NvRegReceiverStatus, NVREG_RCVSTAT_BUSY, 0,
                        NV_RXSTOP_DELAY1, NV_RXSTOP_DELAY1MAX,
                        KERN_INFO "nv_stop_rx: ReceiverStatus remained busy");
        udelay(NV_RXSTOP_DELAY2);
-        writel(0, base + NvRegLinkSpeed);
+        if (!np->mac_in_use)
+                writel(0, base + NvRegLinkSpeed);
 }
 static void nv_start_tx(struct net_device *dev)
 {
+        struct fe_priv *np = netdev_priv(dev);
        u8 __iomem *base = get_hwbase(dev);
+        u32 tx_ctrl = readl(base + NvRegTransmitterControl);
        dprintk(KERN_DEBUG "%s: nv_start_tx\n", dev->name);
-        writel(NVREG_XMITCTL_START, base + NvRegTransmitterControl);
+        tx_ctrl |= NVREG_XMITCTL_START;
+        if (np->mac_in_use)
+                tx_ctrl &= ~NVREG_XMITCTL_TX_PATH_EN;
+        writel(tx_ctrl, base + NvRegTransmitterControl);
        pci_push(base);
 }
 static void nv_stop_tx(struct net_device *dev)
 {
+        struct fe_priv *np = netdev_priv(dev);
        u8 __iomem *base = get_hwbase(dev);
+        u32 tx_ctrl = readl(base + NvRegTransmitterControl);
        dprintk(KERN_DEBUG "%s: nv_stop_tx\n", dev->name);
-        writel(0, base + NvRegTransmitterControl);
+        if (!np->mac_in_use)
+                tx_ctrl &= ~NVREG_XMITCTL_START;
+        else
+                tx_ctrl |= NVREG_XMITCTL_TX_PATH_EN;
+        writel(tx_ctrl, base + NvRegTransmitterControl);
        reg_delay(dev, NvRegTransmitterStatus, NVREG_XMITSTAT_BUSY, 0,
                        NV_TXSTOP_DELAY1, NV_TXSTOP_DELAY1MAX,
                        KERN_INFO "nv_stop_tx: TransmitterStatus remained busy");
        udelay(NV_TXSTOP_DELAY2);
-        writel(readl(base + NvRegTransmitPoll) & NVREG_TRANSMITPOLL_MAC_ADDR_REV, base + NvRegTransmitPoll);
+        if (!np->mac_in_use)
+                writel(readl(base + NvRegTransmitPoll) & NVREG_TRANSMITPOLL_MAC_ADDR_REV,
+                       base + NvRegTransmitPoll);
 }
 static void nv_txrx_reset(struct net_device *dev)
@@ -4148,20 +4175,6 @@ static int nv_mgmt_acquire_sema(struct net_device *dev)
        return 0;
 }
-/* Indicate to mgmt unit whether driver is loaded or not */
-static void nv_mgmt_driver_loaded(struct net_device *dev, int loaded)
-{
-        u8 __iomem *base = get_hwbase(dev);
-        u32 tx_ctrl;
-        tx_ctrl = readl(base + NvRegTransmitterControl);
-        if (loaded)
-                tx_ctrl |= NVREG_XMITCTL_HOST_LOADED;
-        else
-                tx_ctrl &= ~NVREG_XMITCTL_HOST_LOADED;
-        writel(tx_ctrl, base + NvRegTransmitterControl);
-}
 static int nv_open(struct net_device *dev)
 {
        struct fe_priv *np = netdev_priv(dev);
@@ -4659,33 +4672,24 @@ static int __devinit nv_probe(struct pci_dev *pci_dev, const struct pci_device_i
        writel(NVREG_MIISTAT_MASK, base + NvRegMIIStatus);
        if (id->driver_data & DEV_HAS_MGMT_UNIT) {
-                writel(0x1, base + 0x204); pci_push(base);
-                msleep(500);
                /* management unit running on the mac? */
-                np->mac_in_use = readl(base + NvRegTransmitterControl) & NVREG_XMITCTL_MGMT_ST;
+                if (readl(base + NvRegTransmitterControl) & NVREG_XMITCTL_SYNC_PHY_INIT) {
-                if (np->mac_in_use) {
+                        np->mac_in_use = readl(base + NvRegTransmitterControl) & NVREG_XMITCTL_MGMT_ST;
-                        u32 mgmt_sync;
+                        dprintk(KERN_INFO "%s: mgmt unit is running. mac in use %x.\n", pci_name(pci_dev), np->mac_in_use);
-                        /* management unit setup the phy already? */
+                        for (i = 0; i < 5000; i++) {
-                        mgmt_sync = readl(base + NvRegTransmitterControl) & NVREG_XMITCTL_SYNC_MASK;
+                                msleep(1);
-                        if (mgmt_sync == NVREG_XMITCTL_SYNC_NOT_READY) {
+                                if (nv_mgmt_acquire_sema(dev)) {
-                                if (!nv_mgmt_acquire_sema(dev)) {
+                                        /* management unit setup the phy already? */
-                                        for (i = 0; i < 5000; i++) {
+                                        if ((readl(base + NvRegTransmitterControl) & NVREG_XMITCTL_SYNC_MASK) ==
-                                                msleep(1);
+                                            NVREG_XMITCTL_SYNC_PHY_INIT) {
-                                                mgmt_sync = readl(base + NvRegTransmitterControl) & NVREG_XMITCTL_SYNC_MASK;
+                                                /* phy is inited by mgmt unit */
-                                                if (mgmt_sync == NVREG_XMITCTL_SYNC_NOT_READY)
+                                                phyinitialized = 1;
-                                                        continue;
+                                                dprintk(KERN_INFO "%s: Phy already initialized by mgmt unit.\n", pci_name(pci_dev));
-                                                if (mgmt_sync == NVREG_XMITCTL_SYNC_PHY_INIT)
+                                        } else {
-                                                        phyinitialized = 1;
+                                                /* we need to init the phy */
-                                                break;
                                        }
-                                } else {
+                                        break;
-                                        /* we need to init the phy */
                                }
-                        } else if (mgmt_sync == NVREG_XMITCTL_SYNC_PHY_INIT) {
-                                /* phy is inited by SMU */
-                                phyinitialized = 1;
-                        } else {
-                                /* we need to init the phy */
                        }
                }
        }
@@ -4724,10 +4728,12 @@ static int __devinit nv_probe(struct pci_dev *pci_dev, const struct pci_device_i
        if (!phyinitialized) {
                /* reset it */
                phy_init(dev);
-        }
+        } else {
+                /* see if it is a gigabit phy */
-        if (id->driver_data & DEV_HAS_MGMT_UNIT) {
+                u32 mii_status = mii_rw(dev, np->phyaddr, MII_BMSR, MII_READ);
-                nv_mgmt_driver_loaded(dev, 1);
+                if (mii_status & PHY_GIGABIT) {
+                        np->gigabit = PHY_GIGABIT;
+                }
        }
        /* set default link speed settings */
@@ -4749,8 +4755,6 @@ static int __devinit nv_probe(struct pci_dev *pci_dev, const struct pci_device_i
 out_error:
        if (phystate_orig)
                writel(phystate|NVREG_ADAPTCTL_RUNNING, base + NvRegAdapterControl);
-        if (np->mac_in_use)
-                nv_mgmt_driver_loaded(dev, 0);
        pci_set_drvdata(pci_dev, NULL);
 out_freering:
        free_rings(dev);
@@ -4780,9 +4784,6 @@ static void __devexit nv_remove(struct pci_dev *pci_dev)
        writel(np->orig_mac[0], base + NvRegMacAddrA);
        writel(np->orig_mac[1], base + NvRegMacAddrB);
-        if (np->mac_in_use)
-                nv_mgmt_driver_loaded(dev, 0);
        /* free all structures */
        free_rings(dev);
        iounmap(get_hwbase(dev));
diff --git a/drivers/net/ifb.c b/drivers/net/ifb.c
index c26a4b8e552a..ca2b21f9d444 100644
--- a/drivers/net/ifb.c
+++ b/drivers/net/ifb.c
@@ -154,8 +154,8 @@ static int ifb_xmit(struct sk_buff *skb, struct net_device *dev)
        int ret = 0;
        u32 from = G_TC_FROM(skb->tc_verd);
-        stats->tx_packets++;
+        stats->rx_packets++;
-        stats->tx_bytes+=skb->len;
+        stats->rx_bytes+=skb->len;
        if (!from || !skb->input_dev) {
 dropped:
diff --git a/drivers/net/ixgb/ixgb.h b/drivers/net/ixgb/ixgb.h
index 50ffe90488ff..f4aba4355b19 100644
--- a/drivers/net/ixgb/ixgb.h
+++ b/drivers/net/ixgb/ixgb.h
@@ -171,6 +171,7 @@ struct ixgb_adapter {
        /* TX */
        struct ixgb_desc_ring tx_ring ____cacheline_aligned_in_smp;
+        unsigned int restart_queue;
        unsigned long timeo_start;
        uint32_t tx_cmd_type;
        uint64_t hw_csum_tx_good;
diff --git a/drivers/net/ixgb/ixgb_ethtool.c b/drivers/net/ixgb/ixgb_ethtool.c
index cd22523fb035..82c044d6e08a 100644
--- a/drivers/net/ixgb/ixgb_ethtool.c
+++ b/drivers/net/ixgb/ixgb_ethtool.c
@@ -79,6 +79,7 @@ static struct ixgb_stats ixgb_gstrings_stats[] = {
        {"tx_window_errors", IXGB_STAT(net_stats.tx_window_errors)},
        {"tx_deferred_ok", IXGB_STAT(stats.dc)},
        {"tx_timeout_count", IXGB_STAT(tx_timeout_count) },
+        {"tx_restart_queue", IXGB_STAT(restart_queue) },
        {"rx_long_length_errors", IXGB_STAT(stats.roc)},
        {"rx_short_length_errors", IXGB_STAT(stats.ruc)},
 #ifdef NETIF_F_TSO
diff --git a/drivers/net/ixgb/ixgb_hw.c b/drivers/net/ixgb/ixgb_hw.c
index 02089b64e42c..ecbf45861c68 100644
--- a/drivers/net/ixgb/ixgb_hw.c
+++ b/drivers/net/ixgb/ixgb_hw.c
@@ -399,8 +399,9 @@ ixgb_init_rx_addrs(struct ixgb_hw *hw)
        /* Zero out the other 15 receive addresses. */
        DEBUGOUT("Clearing RAR[1-15]\n");
        for(i = 1; i < IXGB_RAR_ENTRIES; i++) {
-                IXGB_WRITE_REG_ARRAY(hw, RA, (i << 1), 0);
+                /* Write high reg first to disable the AV bit first */
                IXGB_WRITE_REG_ARRAY(hw, RA, ((i << 1) + 1), 0);
+                IXGB_WRITE_REG_ARRAY(hw, RA, (i << 1), 0);
        }
        return;
diff --git a/drivers/net/ixgb/ixgb_main.c b/drivers/net/ixgb/ixgb_main.c
index e628126c9c49..a083a9189230 100644
--- a/drivers/net/ixgb/ixgb_main.c
+++ b/drivers/net/ixgb/ixgb_main.c
@@ -36,7 +36,7 @@ static char ixgb_driver_string[] = "Intel(R) PRO/10GbE Network Driver";
 #else
 #define DRIVERNAPI "-NAPI"
 #endif
-#define DRV_VERSION             "1.0.117-k2"DRIVERNAPI
+#define DRV_VERSION             "1.0.126-k2"DRIVERNAPI
 char ixgb_driver_version[] = DRV_VERSION;
 static char ixgb_copyright[] = "Copyright (c) 1999-2006 Intel Corporation.";
@@ -1287,6 +1287,7 @@ ixgb_tx_map(struct ixgb_adapter *adapter, struct sk_buff *skb,
        struct ixgb_buffer *buffer_info;
        int len = skb->len;
        unsigned int offset = 0, size, count = 0, i;
+        unsigned int mss = skb_shinfo(skb)->gso_size;
        unsigned int nr_frags = skb_shinfo(skb)->nr_frags;
        unsigned int f;
@@ -1298,6 +1299,11 @@ ixgb_tx_map(struct ixgb_adapter *adapter, struct sk_buff *skb,
        while(len) {
                buffer_info = &tx_ring->buffer_info[i];
                size = min(len, IXGB_MAX_DATA_PER_TXD);
+                /* Workaround for premature desc write-backs
+                 * in TSO mode.  Append 4-byte sentinel desc */
+                if (unlikely(mss && !nr_frags && size == len && size > 8))
+                        size -= 4;
                buffer_info->length = size;
                WARN_ON(buffer_info->dma != 0);
                buffer_info->dma =
@@ -1324,6 +1330,13 @@ ixgb_tx_map(struct ixgb_adapter *adapter, struct sk_buff *skb,
                while(len) {
                        buffer_info = &tx_ring->buffer_info[i];
                        size = min(len, IXGB_MAX_DATA_PER_TXD);
+                        /* Workaround for premature desc write-backs
+                         * in TSO mode.  Append 4-byte sentinel desc */
+                        if (unlikely(mss && !nr_frags && size == len
+                                     && size > 8))
+                                size -= 4;
                        buffer_info->length = size;
                        buffer_info->dma =
                                pci_map_page(adapter->pdev,
@@ -1398,11 +1411,43 @@ ixgb_tx_queue(struct ixgb_adapter *adapter, int count, int vlan_id,int tx_flags)
        IXGB_WRITE_REG(&adapter->hw, TDT, i);
 }
+static int __ixgb_maybe_stop_tx(struct net_device *netdev, int size)
+{
+        struct ixgb_adapter *adapter = netdev_priv(netdev);
+        struct ixgb_desc_ring *tx_ring = &adapter->tx_ring;
+        netif_stop_queue(netdev);
+        /* Herbert's original patch had:
+         *  smp_mb__after_netif_stop_queue();
+         * but since that doesn't exist yet, just open code it. */
+        smp_mb();
+        /* We need to check again in a case another CPU has just
+         * made room available. */
+        if (likely(IXGB_DESC_UNUSED(tx_ring) < size))
+                return -EBUSY;
+        /* A reprieve! */
+        netif_start_queue(netdev);
+        ++adapter->restart_queue;
+        return 0;
+}
+static int ixgb_maybe_stop_tx(struct net_device *netdev,
+                              struct ixgb_desc_ring *tx_ring, int size)
+{
+        if (likely(IXGB_DESC_UNUSED(tx_ring) >= size))
+                return 0;
+        return __ixgb_maybe_stop_tx(netdev, size);
+}
 /* Tx Descriptors needed, worst case */
 #define TXD_USE_COUNT(S) (((S) >> IXGB_MAX_TXD_PWR) + \
                         (((S) & (IXGB_MAX_DATA_PER_TXD - 1)) ? 1 : 0))
-#define DESC_NEEDED TXD_USE_COUNT(IXGB_MAX_DATA_PER_TXD) + \
+#define DESC_NEEDED TXD_USE_COUNT(IXGB_MAX_DATA_PER_TXD) /* skb->date */ + \
-        MAX_SKB_FRAGS * TXD_USE_COUNT(PAGE_SIZE) + 1
+        MAX_SKB_FRAGS * TXD_USE_COUNT(PAGE_SIZE) + 1 /* for context */ \
+        + 1 /* one more needed for sentinel TSO workaround */
 static int
 ixgb_xmit_frame(struct sk_buff *skb, struct net_device *netdev)
@@ -1430,7 +1475,8 @@ ixgb_xmit_frame(struct sk_buff *skb, struct net_device *netdev)
        spin_lock_irqsave(&adapter->tx_lock, flags);
 #endif
-        if(unlikely(IXGB_DESC_UNUSED(&adapter->tx_ring) < DESC_NEEDED)) {
+        if (unlikely(ixgb_maybe_stop_tx(netdev, &adapter->tx_ring,
+                     DESC_NEEDED))) {
                netif_stop_queue(netdev);
                spin_unlock_irqrestore(&adapter->tx_lock, flags);
                return NETDEV_TX_BUSY;
@@ -1468,8 +1514,7 @@ ixgb_xmit_frame(struct sk_buff *skb, struct net_device *netdev)
 #ifdef NETIF_F_LLTX
        /* Make sure there is space in the ring for the next send. */
-        if(unlikely(IXGB_DESC_UNUSED(&adapter->tx_ring) < DESC_NEEDED))
+        ixgb_maybe_stop_tx(netdev, &adapter->tx_ring, DESC_NEEDED);
-                netif_stop_queue(netdev);
        spin_unlock_irqrestore(&adapter->tx_lock, flags);
diff --git a/drivers/net/loopback.c b/drivers/net/loopback.c
index 82c10dec1b5a..2b739fd584f1 100644
--- a/drivers/net/loopback.c
+++ b/drivers/net/loopback.c
@@ -229,9 +229,11 @@ struct net_device loopback_dev = {
 };
 /* Setup and register the loopback device. */
-int __init loopback_init(void)
+static int __init loopback_init(void)
 {
        return register_netdev(&loopback_dev);
 };
+module_init(loopback_init);
 EXPORT_SYMBOL(loopback_dev);
diff --git a/drivers/net/pcmcia/pcnet_cs.c b/drivers/net/pcmcia/pcnet_cs.c
index 2b1238e2dbdb..d88e9b2e93cf 100644
--- a/drivers/net/pcmcia/pcnet_cs.c
+++ b/drivers/net/pcmcia/pcnet_cs.c
@@ -1617,6 +1617,7 @@ static struct pcmcia_device_id pcnet_ids[] = {
        PCMCIA_DEVICE_PROD_ID12("corega K.K.", "corega FastEther PCC-TX", 0x5261440f, 0x485e85d9),
        PCMCIA_DEVICE_PROD_ID12("Corega,K.K.", "Ethernet LAN Card", 0x110d26d9, 0x9fd2f0a2),
        PCMCIA_DEVICE_PROD_ID12("corega,K.K.", "Ethernet LAN Card", 0x9791a90e, 0x9fd2f0a2),
+        PCMCIA_DEVICE_PROD_ID12("corega K.K.", "(CG-LAPCCTXD)", 0x5261440f, 0x73ec0d88),
        PCMCIA_DEVICE_PROD_ID12("CouplerlessPCMCIA", "100BASE", 0xee5af0ad, 0x7c2add04),
        PCMCIA_DEVICE_PROD_ID12("CyQ've", "ELA-010", 0x77008979, 0x9d8d445d),
        PCMCIA_DEVICE_PROD_ID12("CyQ've", "ELA-110E 10/100M LAN Card", 0x77008979, 0xfd184814),
@@ -1667,6 +1668,7 @@ static struct pcmcia_device_id pcnet_ids[] = {
        PCMCIA_DEVICE_PROD_ID12("Logitec", "LPM-LN100TX", 0x88fcdeda, 0x6d772737),
        PCMCIA_DEVICE_PROD_ID12("Logitec", "LPM-LN100TE", 0x88fcdeda, 0x0e714bee),
        PCMCIA_DEVICE_PROD_ID12("Logitec", "LPM-LN20T", 0x88fcdeda, 0x81090922),
+        PCMCIA_DEVICE_PROD_ID12("Logitec", "LPM-LN10TE", 0x88fcdeda, 0xc1e2521c),
        PCMCIA_DEVICE_PROD_ID12("LONGSHINE", "PCMCIA Ethernet Card", 0xf866b0b0, 0x6f6652e0),
        PCMCIA_DEVICE_PROD_ID12("MACNICA", "ME1-JEIDA", 0x20841b68, 0xaf8a3578),
        PCMCIA_DEVICE_PROD_ID12("Macsense", "MPC-10", 0xd830297f, 0xd265c307),
diff --git a/drivers/net/qla3xxx.c b/drivers/net/qla3xxx.c
index d79d141a601d..8844c20eac2d 100644
--- a/drivers/net/qla3xxx.c
+++ b/drivers/net/qla3xxx.c
@@ -208,6 +208,15 @@ static void ql_write_common_reg(struct ql3_adapter *qdev,
        return;
 }
+static void ql_write_nvram_reg(struct ql3_adapter *qdev,
+                                u32 __iomem *reg, u32 value)
+{
+        writel(value, reg);
+        readl(reg);
+        udelay(1);
+        return;
+}
 static void ql_write_page0_reg(struct ql3_adapter *qdev,
                               u32 __iomem *reg, u32 value)
 {
@@ -336,9 +345,9 @@ static void fm93c56a_select(struct ql3_adapter *qdev)
                        qdev->mem_map_registers;
        qdev->eeprom_cmd_data = AUBURN_EEPROM_CS_1;
-        ql_write_common_reg(qdev, &port_regs->CommonRegs.serialPortInterfaceReg,
+        ql_write_nvram_reg(qdev, &port_regs->CommonRegs.serialPortInterfaceReg,
                            ISP_NVRAM_MASK | qdev->eeprom_cmd_data);
-        ql_write_common_reg(qdev, &port_regs->CommonRegs.serialPortInterfaceReg,
+        ql_write_nvram_reg(qdev, &port_regs->CommonRegs.serialPortInterfaceReg,
                            ((ISP_NVRAM_MASK << 16) | qdev->eeprom_cmd_data));
 }
@@ -355,14 +364,14 @@ static void fm93c56a_cmd(struct ql3_adapter *qdev, u32 cmd, u32 eepromAddr)
                        qdev->mem_map_registers;
        /* Clock in a zero, then do the start bit */
-        ql_write_common_reg(qdev, &port_regs->CommonRegs.serialPortInterfaceReg,
+        ql_write_nvram_reg(qdev, &port_regs->CommonRegs.serialPortInterfaceReg,
                            ISP_NVRAM_MASK | qdev->eeprom_cmd_data |
                            AUBURN_EEPROM_DO_1);
-        ql_write_common_reg(qdev, &port_regs->CommonRegs.serialPortInterfaceReg,
+        ql_write_nvram_reg(qdev, &port_regs->CommonRegs.serialPortInterfaceReg,
                            ISP_NVRAM_MASK | qdev->
                            eeprom_cmd_data | AUBURN_EEPROM_DO_1 |
                            AUBURN_EEPROM_CLK_RISE);
-        ql_write_common_reg(qdev, &port_regs->CommonRegs.serialPortInterfaceReg,
+        ql_write_nvram_reg(qdev, &port_regs->CommonRegs.serialPortInterfaceReg,
                            ISP_NVRAM_MASK | qdev->
                            eeprom_cmd_data | AUBURN_EEPROM_DO_1 |
                            AUBURN_EEPROM_CLK_FALL);
@@ -378,20 +387,20 @@ static void fm93c56a_cmd(struct ql3_adapter *qdev, u32 cmd, u32 eepromAddr)
                         * If the bit changed, then change the DO state to
                         * match
                         */
-                        ql_write_common_reg(qdev,
+                        ql_write_nvram_reg(qdev,
                                            &port_regs->CommonRegs.
                                            serialPortInterfaceReg,
                                            ISP_NVRAM_MASK | qdev->
                                            eeprom_cmd_data | dataBit);
                        previousBit = dataBit;
                }
-                ql_write_common_reg(qdev,
+                ql_write_nvram_reg(qdev,
                                    &port_regs->CommonRegs.
                                    serialPortInterfaceReg,
                                    ISP_NVRAM_MASK | qdev->
                                    eeprom_cmd_data | dataBit |
                                    AUBURN_EEPROM_CLK_RISE);
-                ql_write_common_reg(qdev,
+                ql_write_nvram_reg(qdev,
                                    &port_regs->CommonRegs.
                                    serialPortInterfaceReg,
                                    ISP_NVRAM_MASK | qdev->
@@ -412,20 +421,20 @@ static void fm93c56a_cmd(struct ql3_adapter *qdev, u32 cmd, u32 eepromAddr)
                         * If the bit changed, then change the DO state to
                         * match
                         */
-                        ql_write_common_reg(qdev,
+                        ql_write_nvram_reg(qdev,
                                            &port_regs->CommonRegs.
                                            serialPortInterfaceReg,
                                            ISP_NVRAM_MASK | qdev->
                                            eeprom_cmd_data | dataBit);
                        previousBit = dataBit;
                }
-                ql_write_common_reg(qdev,
+                ql_write_nvram_reg(qdev,
                                    &port_regs->CommonRegs.
                                    serialPortInterfaceReg,
                                    ISP_NVRAM_MASK | qdev->
                                    eeprom_cmd_data | dataBit |
                                    AUBURN_EEPROM_CLK_RISE);
-                ql_write_common_reg(qdev,
+                ql_write_nvram_reg(qdev,
                                    &port_regs->CommonRegs.
                                    serialPortInterfaceReg,
                                    ISP_NVRAM_MASK | qdev->
@@ -443,7 +452,7 @@ static void fm93c56a_deselect(struct ql3_adapter *qdev)
        struct ql3xxx_port_registers __iomem *port_regs =
                        qdev->mem_map_registers;
        qdev->eeprom_cmd_data = AUBURN_EEPROM_CS_0;
-        ql_write_common_reg(qdev, &port_regs->CommonRegs.serialPortInterfaceReg,
+        ql_write_nvram_reg(qdev, &port_regs->CommonRegs.serialPortInterfaceReg,
                            ISP_NVRAM_MASK | qdev->eeprom_cmd_data);
 }
@@ -461,12 +470,12 @@ static void fm93c56a_datain(struct ql3_adapter *qdev, unsigned short *value)
        /* Read the data bits */
        /* The first bit is a dummy.  Clock right over it. */
        for (i = 0; i < dataBits; i++) {
-                ql_write_common_reg(qdev,
+                ql_write_nvram_reg(qdev,
                                    &port_regs->CommonRegs.
                                    serialPortInterfaceReg,
                                    ISP_NVRAM_MASK | qdev->eeprom_cmd_data |
                                    AUBURN_EEPROM_CLK_RISE);
-                ql_write_common_reg(qdev,
+                ql_write_nvram_reg(qdev,
                                    &port_regs->CommonRegs.
                                    serialPortInterfaceReg,
                                    ISP_NVRAM_MASK | qdev->eeprom_cmd_data |
@@ -3370,7 +3379,6 @@ static int __devinit ql3xxx_probe(struct pci_dev *pdev,
        SET_MODULE_OWNER(ndev);
        SET_NETDEV_DEV(ndev, &pdev->dev);
-        ndev->features = NETIF_F_LLTX;
        if (pci_using_dac)
                ndev->features |= NETIF_F_HIGHDMA;
diff --git a/drivers/net/sungem.c b/drivers/net/sungem.c
index 785e4a535f9e..616be8d0fa85 100644
--- a/drivers/net/sungem.c
+++ b/drivers/net/sungem.c
@@ -90,7 +90,8 @@
 #define ADVERTISE_MASK  (SUPPORTED_10baseT_Half | SUPPORTED_10baseT_Full | \
                         SUPPORTED_100baseT_Half | SUPPORTED_100baseT_Full | \
-                         SUPPORTED_1000baseT_Half | SUPPORTED_1000baseT_Full)
+                         SUPPORTED_1000baseT_Half | SUPPORTED_1000baseT_Full | \
+                         SUPPORTED_Pause | SUPPORTED_Autoneg)
 #define DRV_NAME        "sungem"
 #define DRV_VERSION     "0.98"
diff --git a/drivers/net/sungem_phy.c b/drivers/net/sungem_phy.c
index 49800b25907d..d21991ee88c4 100644
--- a/drivers/net/sungem_phy.c
+++ b/drivers/net/sungem_phy.c
@@ -3,10 +3,9 @@
 *
 * This file could be shared with other drivers.
 *
- * (c) 2002, Benjamin Herrenscmidt (benh@kernel.crashing.org)
+ * (c) 2002-2007, Benjamin Herrenscmidt (benh@kernel.crashing.org)
 *
 * TODO:
- *  - Implement WOL
 *  - Add support for PHYs that provide an IRQ line
 *  - Eventually moved the entire polling state machine in
 *    there (out of the eth driver), so that it can easily be
@@ -152,6 +151,44 @@ static int bcm5221_suspend(struct mii_phy* phy)
        return 0;
 }
+static int bcm5241_init(struct mii_phy* phy)
+{
+        u16 data;
+        data = phy_read(phy, MII_BCM5221_TEST);
+        phy_write(phy, MII_BCM5221_TEST,
+                data | MII_BCM5221_TEST_ENABLE_SHADOWS);
+        data = phy_read(phy, MII_BCM5221_SHDOW_AUX_STAT2);
+        phy_write(phy, MII_BCM5221_SHDOW_AUX_STAT2,
+                data | MII_BCM5221_SHDOW_AUX_STAT2_APD);
+        data = phy_read(phy, MII_BCM5221_SHDOW_AUX_MODE4);
+        phy_write(phy, MII_BCM5221_SHDOW_AUX_MODE4,
+                data & ~MII_BCM5241_SHDOW_AUX_MODE4_STANDBYPWR);
+        data = phy_read(phy, MII_BCM5221_TEST);
+        phy_write(phy, MII_BCM5221_TEST,
+                data & ~MII_BCM5221_TEST_ENABLE_SHADOWS);
+        return 0;
+}
+static int bcm5241_suspend(struct mii_phy* phy)
+{
+        u16 data;
+        data = phy_read(phy, MII_BCM5221_TEST);
+        phy_write(phy, MII_BCM5221_TEST,
+                data | MII_BCM5221_TEST_ENABLE_SHADOWS);
+        data = phy_read(phy, MII_BCM5221_SHDOW_AUX_MODE4);
+        phy_write(phy, MII_BCM5221_SHDOW_AUX_MODE4,
+                  data | MII_BCM5241_SHDOW_AUX_MODE4_STANDBYPWR);
+        return 0;
+}
 static int bcm5400_init(struct mii_phy* phy)
 {
        u16 data;
@@ -373,6 +410,10 @@ static int bcm54xx_setup_aneg(struct mii_phy *phy, u32 advertise)
                adv |= ADVERTISE_100HALF;
        if (advertise & ADVERTISED_100baseT_Full)
                adv |= ADVERTISE_100FULL;
+        if (advertise & ADVERTISED_Pause)
+                adv |= ADVERTISE_PAUSE_CAP;
+        if (advertise & ADVERTISED_Asym_Pause)
+                adv |= ADVERTISE_PAUSE_ASYM;
        phy_write(phy, MII_ADVERTISE, adv);
        /* Setup 1000BT advertise */
@@ -436,12 +477,15 @@ static int bcm54xx_read_link(struct mii_phy *phy)
                val = phy_read(phy, MII_BCM5400_AUXSTATUS);
                link_mode = ((val & MII_BCM5400_AUXSTATUS_LINKMODE_MASK) >>
                             MII_BCM5400_AUXSTATUS_LINKMODE_SHIFT);
-                phy->duplex = phy_BCM5400_link_table[link_mode][0] ? DUPLEX_FULL : DUPLEX_HALF;
+                phy->duplex = phy_BCM5400_link_table[link_mode][0] ?
+                        DUPLEX_FULL : DUPLEX_HALF;
                phy->speed = phy_BCM5400_link_table[link_mode][2] ?
                                SPEED_1000 :
-                                (phy_BCM5400_link_table[link_mode][1] ? SPEED_100 : SPEED_10);
+                                (phy_BCM5400_link_table[link_mode][1] ?
+                                 SPEED_100 : SPEED_10);
                val = phy_read(phy, MII_LPA);
-                phy->pause = ((val & LPA_PAUSE) != 0);
+                phy->pause = (phy->duplex == DUPLEX_FULL) &&
+                        ((val & LPA_PAUSE) != 0);
        }
        /* On non-aneg, we assume what we put in BMCR is the speed,
         * though magic-aneg shouldn't prevent this case from occurring
@@ -450,6 +494,28 @@ static int bcm54xx_read_link(struct mii_phy *phy)
        return 0;
 }
+static int marvell88e1111_init(struct mii_phy* phy)
+{
+        u16 rev;
+        /* magic init sequence for rev 0 */
+        rev = phy_read(phy, MII_PHYSID2) & 0x000f;
+        if (rev == 0) {
+                phy_write(phy, 0x1d, 0x000a);
+                phy_write(phy, 0x1e, 0x0821);
+                phy_write(phy, 0x1d, 0x0006);
+                phy_write(phy, 0x1e, 0x8600);
+                phy_write(phy, 0x1d, 0x000b);
+                phy_write(phy, 0x1e, 0x0100);
+                phy_write(phy, 0x1d, 0x0004);
+                phy_write(phy, 0x1e, 0x4850);
+        }
+        return 0;
+}
 static int marvell_setup_aneg(struct mii_phy *phy, u32 advertise)
 {
        u16 ctl, adv;
@@ -471,6 +537,10 @@ static int marvell_setup_aneg(struct mii_phy *phy, u32 advertise)
                adv |= ADVERTISE_100HALF;
        if (advertise & ADVERTISED_100baseT_Full)
                adv |= ADVERTISE_100FULL;
+        if (advertise & ADVERTISED_Pause)
+                adv |= ADVERTISE_PAUSE_CAP;
+        if (advertise & ADVERTISED_Asym_Pause)
+                adv |= ADVERTISE_PAUSE_ASYM;
        phy_write(phy, MII_ADVERTISE, adv);
        /* Setup 1000BT advertise & enable crossover detect
@@ -549,7 +619,7 @@ static int marvell_setup_forced(struct mii_phy *phy, int speed, int fd)
 static int marvell_read_link(struct mii_phy *phy)
 {
-        u16 status;
+        u16 status, pmask;
        if (phy->autoneg) {
                status = phy_read(phy, MII_M1011_PHY_SPEC_STATUS);
@@ -565,7 +635,9 @@ static int marvell_read_link(struct mii_phy *phy)
                        phy->duplex = DUPLEX_FULL;
                else
                        phy->duplex = DUPLEX_HALF;
-                phy->pause = 0; /* XXX Check against spec ! */
+                pmask = MII_M1011_PHY_SPEC_STATUS_TX_PAUSE |
+                        MII_M1011_PHY_SPEC_STATUS_RX_PAUSE;
+                phy->pause = (status & pmask) == pmask;
        }
        /* On non-aneg, we assume what we put in BMCR is the speed,
         * though magic-aneg shouldn't prevent this case from occurring
@@ -595,6 +667,10 @@ static int genmii_setup_aneg(struct mii_phy *phy, u32 advertise)
                adv |= ADVERTISE_100HALF;
        if (advertise & ADVERTISED_100baseT_Full)
                adv |= ADVERTISE_100FULL;
+        if (advertise & ADVERTISED_Pause)
+                adv |= ADVERTISE_PAUSE_CAP;
+        if (advertise & ADVERTISED_Asym_Pause)
+                adv |= ADVERTISE_PAUSE_ASYM;
        phy_write(phy, MII_ADVERTISE, adv);
        /* Start/Restart aneg */
@@ -666,7 +742,8 @@ static int genmii_read_link(struct mii_phy *phy)
                        phy->speed = SPEED_100;
                else
                        phy->speed = SPEED_10;
-                phy->pause = 0;
+                phy->pause = (phy->duplex == DUPLEX_FULL) &&
+                        ((lpa & LPA_PAUSE) != 0);
        }
        /* On non-aneg, we assume what we put in BMCR is the speed,
         * though magic-aneg shouldn't prevent this case from occurring
@@ -676,11 +753,19 @@ static int genmii_read_link(struct mii_phy *phy)
 }
-#define MII_BASIC_FEATURES      (SUPPORTED_10baseT_Half | SUPPORTED_10baseT_Full | \
+#define MII_BASIC_FEATURES \
-                                 SUPPORTED_100baseT_Half | SUPPORTED_100baseT_Full | \
+        (SUPPORTED_10baseT_Half | SUPPORTED_10baseT_Full |      \
-                                 SUPPORTED_Autoneg | SUPPORTED_TP | SUPPORTED_MII)
+         SUPPORTED_100baseT_Half | SUPPORTED_100baseT_Full |    \
-#define MII_GBIT_FEATURES       (MII_BASIC_FEATURES | \
+         SUPPORTED_Autoneg | SUPPORTED_TP | SUPPORTED_MII |     \
-                                 SUPPORTED_1000baseT_Half | SUPPORTED_1000baseT_Full)
+         SUPPORTED_Pause)
+/* On gigabit capable PHYs, we advertise Pause support but not asym pause
+ * support for now as I'm not sure it's supported and Darwin doesn't do
+ * it neither. --BenH.
+ */
+#define MII_GBIT_FEATURES \
+        (MII_BASIC_FEATURES |   \
+         SUPPORTED_1000baseT_Half | SUPPORTED_1000baseT_Full)
 /* Broadcom BCM 5201 */
 static struct mii_phy_ops bcm5201_phy_ops = {
@@ -720,6 +805,24 @@ static struct mii_phy_def bcm5221_phy_def = {
        .ops            = &bcm5221_phy_ops
 };
+/* Broadcom BCM 5241 */
+static struct mii_phy_ops bcm5241_phy_ops = {
+        .suspend        = bcm5241_suspend,
+        .init           = bcm5241_init,
+        .setup_aneg     = genmii_setup_aneg,
+        .setup_forced   = genmii_setup_forced,
+        .poll_link      = genmii_poll_link,
+        .read_link      = genmii_read_link,
+};
+static struct mii_phy_def bcm5241_phy_def = {
+        .phy_id         = 0x0143bc30,
+        .phy_id_mask    = 0xfffffff0,
+        .name           = "BCM5241",
+        .features       = MII_BASIC_FEATURES,
+        .magic_aneg     = 1,
+        .ops            = &bcm5241_phy_ops
+};
 /* Broadcom BCM 5400 */
 static struct mii_phy_ops bcm5400_phy_ops = {
        .init           = bcm5400_init,
@@ -854,11 +957,8 @@ static struct mii_phy_def bcm5462V_phy_def = {
        .ops            = &bcm5462V_phy_ops
 };
-/* Marvell 88E1101 (Apple seem to deal with 2 different revs,
+/* Marvell 88E1101 amd 88E1111 */
- * I masked out the 8 last bits to get both, but some specs
+static struct mii_phy_ops marvell88e1101_phy_ops = {
- * would be useful here) --BenH.
- */
-static struct mii_phy_ops marvell_phy_ops = {
        .suspend        = generic_suspend,
        .setup_aneg     = marvell_setup_aneg,
        .setup_forced   = marvell_setup_forced,
@@ -866,13 +966,41 @@ static struct mii_phy_ops marvell_phy_ops = {
        .read_link      = marvell_read_link
 };
-static struct mii_phy_def marvell_phy_def = {
+static struct mii_phy_ops marvell88e1111_phy_ops = {
-        .phy_id         = 0x01410c00,
+        .init           = marvell88e1111_init,
-        .phy_id_mask    = 0xffffff00,
+        .suspend        = generic_suspend,
-        .name           = "Marvell 88E1101",
+        .setup_aneg     = marvell_setup_aneg,
+        .setup_forced   = marvell_setup_forced,
+        .poll_link      = genmii_poll_link,
+        .read_link      = marvell_read_link
+};
+/* two revs in darwin for the 88e1101 ... I could use a datasheet
+ * to get the proper names...
+ */
+static struct mii_phy_def marvell88e1101v1_phy_def = {
+        .phy_id         = 0x01410c20,
+        .phy_id_mask    = 0xfffffff0,
+        .name           = "Marvell 88E1101v1",
+        .features       = MII_GBIT_FEATURES,
+        .magic_aneg     = 1,
+        .ops            = &marvell88e1101_phy_ops
+};
+static struct mii_phy_def marvell88e1101v2_phy_def = {
+        .phy_id         = 0x01410c60,
+        .phy_id_mask    = 0xfffffff0,
+        .name           = "Marvell 88E1101v2",
+        .features       = MII_GBIT_FEATURES,
+        .magic_aneg     = 1,
+        .ops            = &marvell88e1101_phy_ops
+};
+static struct mii_phy_def marvell88e1111_phy_def = {
+        .phy_id         = 0x01410cc0,
+        .phy_id_mask    = 0xfffffff0,
+        .name           = "Marvell 88E1111",
        .features       = MII_GBIT_FEATURES,
        .magic_aneg     = 1,
-        .ops            = &marvell_phy_ops
+        .ops            = &marvell88e1111_phy_ops
 };
 /* Generic implementation for most 10/100 PHYs */
@@ -895,6 +1023,7 @@ static struct mii_phy_def genmii_phy_def = {
 static struct mii_phy_def* mii_phy_table[] = {
        &bcm5201_phy_def,
        &bcm5221_phy_def,
+        &bcm5241_phy_def,
        &bcm5400_phy_def,
        &bcm5401_phy_def,
        &bcm5411_phy_def,
@@ -902,7 +1031,9 @@ static struct mii_phy_def* mii_phy_table[] = {
        &bcm5421k2_phy_def,
        &bcm5461_phy_def,
        &bcm5462V_phy_def,
-        &marvell_phy_def,
+        &marvell88e1101v1_phy_def,
+        &marvell88e1101v2_phy_def,
+        &marvell88e1111_phy_def,
        &genmii_phy_def,
        NULL
 };
diff --git a/drivers/net/sungem_phy.h b/drivers/net/sungem_phy.h
index 8ee1ca0471cf..1d70ba6f9f10 100644
--- a/drivers/net/sungem_phy.h
+++ b/drivers/net/sungem_phy.h
@@ -30,7 +30,7 @@ struct mii_phy_def
 struct mii_phy
 {
        struct mii_phy_def*     def;
-        int                     advertising;
+        u32                     advertising;
        int                     mii_id;
        /* 1: autoneg enabled, 0: disabled */
@@ -85,6 +85,9 @@ extern int mii_phy_probe(struct mii_phy *phy, int mii_id);
 #define MII_BCM5221_SHDOW_AUX_MODE4_IDDQMODE    0x0001
 #define MII_BCM5221_SHDOW_AUX_MODE4_CLKLOPWR    0x0004
+/* MII BCM5241 Additional registers */
+#define MII_BCM5241_SHDOW_AUX_MODE4_STANDBYPWR  0x0008
 /* MII BCM5400 1000-BASET Control register */
 #define MII_BCM5400_GB_CONTROL                  0x09
 #define MII_BCM5400_GB_CONTROL_FULLDUPLEXCAP    0x0200
@@ -115,5 +118,7 @@ extern int mii_phy_probe(struct mii_phy *phy, int mii_id);
 #define MII_M1011_PHY_SPEC_STATUS_SPD_MASK      0xc000
 #define MII_M1011_PHY_SPEC_STATUS_FULLDUPLEX    0x2000
 #define MII_M1011_PHY_SPEC_STATUS_RESOLVED      0x0800
+#define MII_M1011_PHY_SPEC_STATUS_TX_PAUSE      0x0008
+#define MII_M1011_PHY_SPEC_STATUS_RX_PAUSE      0x0004
 #endif /* __SUNGEM_PHY_H__ */
diff --git a/drivers/net/tg3.c b/drivers/net/tg3.c
index 4056ba1ff3c7..f4bf62c2a7a5 100644
--- a/drivers/net/tg3.c
+++ b/drivers/net/tg3.c
@@ -68,8 +68,8 @@
 #define DRV_MODULE_NAME         "tg3"
 #define PFX DRV_MODULE_NAME     ": "
-#define DRV_MODULE_VERSION      "3.71"
+#define DRV_MODULE_VERSION      "3.72"
-#define DRV_MODULE_RELDATE      "December 15, 2006"
+#define DRV_MODULE_RELDATE      "January 8, 2007"
 #define TG3_DEF_MAC_MODE        0
 #define TG3_DEF_RX_MODE         0
@@ -1015,7 +1015,12 @@ out:
        else if (tp->tg3_flags2 & TG3_FLG2_PHY_JITTER_BUG) {
                tg3_writephy(tp, MII_TG3_AUX_CTRL, 0x0c00);
                tg3_writephy(tp, MII_TG3_DSP_ADDRESS, 0x000a);
-                tg3_writephy(tp, MII_TG3_DSP_RW_PORT, 0x010b);
+                if (tp->tg3_flags2 & TG3_FLG2_PHY_ADJUST_TRIM) {
+                        tg3_writephy(tp, MII_TG3_DSP_RW_PORT, 0x110b);
+                        tg3_writephy(tp, MII_TG3_TEST1,
+                                     MII_TG3_TEST1_TRIM_EN | 0x4);
+                } else
+                        tg3_writephy(tp, MII_TG3_DSP_RW_PORT, 0x010b);
                tg3_writephy(tp, MII_TG3_AUX_CTRL, 0x0400);
        }
        /* Set Extended packet length bit (bit 14) on all chips that */
@@ -10803,9 +10808,11 @@ static int __devinit tg3_get_invariants(struct tg3 *tp)
        if (tp->tg3_flags2 & TG3_FLG2_5705_PLUS) {
                if (GET_ASIC_REV(tp->pci_chip_rev_id) == ASIC_REV_5755 ||
-                    GET_ASIC_REV(tp->pci_chip_rev_id) == ASIC_REV_5787)
+                    GET_ASIC_REV(tp->pci_chip_rev_id) == ASIC_REV_5787) {
                        tp->tg3_flags2 |= TG3_FLG2_PHY_JITTER_BUG;
-                else if (GET_ASIC_REV(tp->pci_chip_rev_id) != ASIC_REV_5906)
+                        if (tp->pdev->device == PCI_DEVICE_ID_TIGON3_5755M)
+                                tp->tg3_flags2 |= TG3_FLG2_PHY_ADJUST_TRIM;
+                } else if (GET_ASIC_REV(tp->pci_chip_rev_id) != ASIC_REV_5906)
                        tp->tg3_flags2 |= TG3_FLG2_PHY_BER_BUG;
        }
diff --git a/drivers/net/tg3.h b/drivers/net/tg3.h
index cf78a7e5997b..80f59ac7ec58 100644
--- a/drivers/net/tg3.h
+++ b/drivers/net/tg3.h
@@ -1658,6 +1658,9 @@
 #define MII_TG3_EPHY_TEST               0x1f /* 5906 PHY register */
 #define MII_TG3_EPHY_SHADOW_EN          0x80
+#define MII_TG3_TEST1                   0x1e
+#define MII_TG3_TEST1_TRIM_EN           0x0010
 /* There are two ways to manage the TX descriptors on the tigon3.
 * Either the descriptors are in host DMA'able memory, or they
 * exist only in the cards on-chip SRAM.  All 16 send bds are under
@@ -2256,6 +2259,7 @@ struct tg3 {
 #define TG3_FLG2_1SHOT_MSI              0x10000000
 #define TG3_FLG2_PHY_JITTER_BUG         0x20000000
 #define TG3_FLG2_NO_FWARE_REPORTED      0x40000000
+#define TG3_FLG2_PHY_ADJUST_TRIM        0x80000000
        u32                             split_mode_max_reqs;
 #define SPLIT_MODE_5704_MAX_REQ         3
diff --git a/drivers/net/wireless/ipw2100.c b/drivers/net/wireless/ipw2100.c
index 0e94fbbf7a94..b85857a84870 100644
--- a/drivers/net/wireless/ipw2100.c
+++ b/drivers/net/wireless/ipw2100.c
@@ -2664,7 +2664,7 @@ static void __ipw2100_rx_process(struct ipw2100_priv *priv)
                                break;
                        }
 #endif
-                        if (stats.len < sizeof(u->rx_data.header))
+                        if (stats.len < sizeof(struct ieee80211_hdr_3addr))
                                break;
                        switch (WLAN_FC_GET_TYPE(u->rx_data.header.frame_ctl)) {
                        case IEEE80211_FTYPE_MGMT:
diff --git a/drivers/pci/Kconfig b/drivers/pci/Kconfig
index f1dd81a1d592..3cfb0a3575e6 100644
--- a/drivers/pci/Kconfig
+++ b/drivers/pci/Kconfig
@@ -19,7 +19,7 @@ config PCI_MSI
 config PCI_MULTITHREAD_PROBE
        bool "PCI Multi-threaded probe (EXPERIMENTAL)"
-        depends on PCI && EXPERIMENTAL
+        depends on PCI && EXPERIMENTAL && BROKEN
        help
          Say Y here if you want the PCI core to spawn a new thread for
          every PCI device that is probed.  This can cause a huge
diff --git a/drivers/pci/quirks.c b/drivers/pci/quirks.c
index 8f0322d6f3bf..0535efc4f184 100644
--- a/drivers/pci/quirks.c
+++ b/drivers/pci/quirks.c
@@ -1117,10 +1117,11 @@ DECLARE_PCI_FIXUP_RESUME(PCI_VENDOR_ID_INTEL,	PCI_DEVICE_ID_INTEL_ICH6_1,	asus_h
 static void quirk_sis_96x_smbus(struct pci_dev *dev)
 {
        u8 val = 0;
-        printk(KERN_INFO "Enabling SiS 96x SMBus.\n");
-        pci_read_config_byte(dev, 0x77, &val);
-        pci_write_config_byte(dev, 0x77, val & ~0x10);
        pci_read_config_byte(dev, 0x77, &val);
+        if (val & 0x10) {
+                printk(KERN_INFO "Enabling SiS 96x SMBus.\n");
+                pci_write_config_byte(dev, 0x77, val & ~0x10);
+        }
 }
 /*
@@ -1152,11 +1153,12 @@ static void quirk_sis_503(struct pci_dev *dev)
        printk(KERN_WARNING "Uncovering SIS%x that hid as a SIS503 (compatible=%d)\n", devid, sis_96x_compatible);
        /*
-         * Ok, it now shows up as a 96x.. The 96x quirks are after
+         * Ok, it now shows up as a 96x.. run the 96x quirk by
-         * the 503 quirk in the quirk table, so they'll automatically
+         * hand in case it has already been processed.
-         * run and enable things like the SMBus device
+         * (depends on link order, which is apparently not guaranteed)
         */
        dev->device = devid;
+        quirk_sis_96x_smbus(dev);
 }
 static void __init quirk_sis_96x_compatible(struct pci_dev *dev)
diff --git a/drivers/pci/search.c b/drivers/pci/search.c
index 45f2b20ef513..fab381ed853c 100644
--- a/drivers/pci/search.c
+++ b/drivers/pci/search.c
@@ -193,6 +193,18 @@ static struct pci_dev * pci_find_subsys(unsigned int vendor,
        struct pci_dev *dev;
        WARN_ON(in_interrupt());
+        /*
+         * pci_find_subsys() can be called on the ide_setup() path, super-early
+         * in boot.  But the down_read() will enable local interrupts, which
+         * can cause some machines to crash.  So here we detect and flag that
+         * situation and bail out early.
+         */
+        if (unlikely(list_empty(&pci_devices))) {
+                printk(KERN_INFO "pci_find_subsys() called while pci_devices "
+                                "is still empty\n");
+                return NULL;
+        }
        down_read(&pci_bus_sem);
        n = from ? from->global_list.next : pci_devices.next;
@@ -259,6 +271,18 @@ pci_get_subsys(unsigned int vendor, unsigned int device,
        struct pci_dev *dev;
        WARN_ON(in_interrupt());
+        /*
+         * pci_get_subsys() can potentially be called by drivers super-early
+         * in boot.  But the down_read() will enable local interrupts, which
+         * can cause some machines to crash.  So here we detect and flag that
+         * situation and bail out early.
+         */
+        if (unlikely(list_empty(&pci_devices))) {
+                printk(KERN_NOTICE "pci_get_subsys() called while pci_devices "
+                                "is still empty\n");
+                return NULL;
+        }
        down_read(&pci_bus_sem);
        n = from ? from->global_list.next : pci_devices.next;
diff --git a/drivers/rtc/rtc-at91rm9200.c b/drivers/rtc/rtc-at91rm9200.c
index 4f654c901c64..a724ab49a797 100644
--- a/drivers/rtc/rtc-at91rm9200.c
+++ b/drivers/rtc/rtc-at91rm9200.c
@@ -33,6 +33,8 @@
 #include <asm/mach/time.h>
+#include <asm/arch/at91_rtc.h>
 #define AT91_RTC_FREQ           1
 #define AT91_RTC_EPOCH          1900UL  /* just like arch/arm/common/rtctime.c */
diff --git a/drivers/rtc/rtc-rs5c372.c b/drivers/rtc/rtc-rs5c372.c
index 1460f6b769f2..e7851e3739ab 100644
--- a/drivers/rtc/rtc-rs5c372.c
+++ b/drivers/rtc/rtc-rs5c372.c
@@ -1,5 +1,5 @@
 /*
- * An I2C driver for the Ricoh RS5C372 RTC
+ * An I2C driver for Ricoh RS5C372 and RV5C38[67] RTCs
 *
 * Copyright (C) 2005 Pavel Mironchik <pmironchik@optifacio.net>
 * Copyright (C) 2006 Tower Technologies
@@ -13,7 +13,7 @@
 #include <linux/rtc.h>
 #include <linux/bcd.h>
-#define DRV_VERSION "0.3"
+#define DRV_VERSION "0.4"
 /* Addresses to scan */
 static unsigned short normal_i2c[] = { /* 0x32,*/ I2C_CLIENT_END };
@@ -21,6 +21,13 @@ static unsigned short normal_i2c[] = { /* 0x32,*/ I2C_CLIENT_END };
 /* Insmod parameters */
 I2C_CLIENT_INSMOD;
+/*
+ * Ricoh has a family of I2C based RTCs, which differ only slightly from
+ * each other.  Differences center on pinout (e.g. how many interrupts,
+ * output clock, etc) and how the control registers are used.  The '372
+ * is significant only because that's the one this driver first supported.
+ */
 #define RS5C372_REG_SECS        0
 #define RS5C372_REG_MINS        1
 #define RS5C372_REG_HOURS       2
@@ -29,59 +36,142 @@ I2C_CLIENT_INSMOD;
 #define RS5C372_REG_MONTH       5
 #define RS5C372_REG_YEAR        6
 #define RS5C372_REG_TRIM        7
+#       define RS5C372_TRIM_XSL         0x80
+#       define RS5C372_TRIM_MASK        0x7F
+#define RS5C_REG_ALARM_A_MIN    8                       /* or ALARM_W */
+#define RS5C_REG_ALARM_A_HOURS  9
+#define RS5C_REG_ALARM_A_WDAY   10
+#define RS5C_REG_ALARM_B_MIN    11                      /* or ALARM_D */
+#define RS5C_REG_ALARM_B_HOURS  12
+#define RS5C_REG_ALARM_B_WDAY   13                      /* (ALARM_B only) */
+#define RS5C_REG_CTRL1          14
+#       define RS5C_CTRL1_AALE          (1 << 7)        /* or WALE */
+#       define RS5C_CTRL1_BALE          (1 << 6)        /* or DALE */
+#       define RV5C387_CTRL1_24         (1 << 5)
+#       define RS5C372A_CTRL1_SL1       (1 << 5)
+#       define RS5C_CTRL1_CT_MASK       (7 << 0)
+#       define RS5C_CTRL1_CT0           (0 << 0)        /* no periodic irq */
+#       define RS5C_CTRL1_CT4           (4 << 0)        /* 1 Hz level irq */
+#define RS5C_REG_CTRL2          15
+#       define RS5C372_CTRL2_24         (1 << 5)
+#       define RS5C_CTRL2_XSTP          (1 << 4)
+#       define RS5C_CTRL2_CTFG          (1 << 2)
+#       define RS5C_CTRL2_AAFG          (1 << 1)        /* or WAFG */
+#       define RS5C_CTRL2_BAFG          (1 << 0)        /* or DAFG */
+/* to read (style 1) or write registers starting at R */
+#define RS5C_ADDR(R)            (((R) << 4) | 0)
+enum rtc_type {
+        rtc_undef = 0,
+        rtc_rs5c372a,
+        rtc_rs5c372b,
+        rtc_rv5c386,
+        rtc_rv5c387a,
+};
-#define RS5C372_TRIM_XSL        0x80
+/* REVISIT:  this assumes that:
-#define RS5C372_TRIM_MASK       0x7F
+ *  - we're in the 21st century, so it's safe to ignore the century
+ *    bit for rv5c38[67] (REG_MONTH bit 7);
+ *  - we should use ALARM_A not ALARM_B (may be wrong on some boards)
+ */
+struct rs5c372 {
+        struct i2c_client       *client;
+        struct rtc_device       *rtc;
+        enum rtc_type           type;
+        unsigned                time24:1;
+        unsigned                has_irq:1;
+        char                    buf[17];
+        char                    *regs;
+        /* on conversion to a "new style" i2c driver, this vanishes */
+        struct i2c_client       dev;
+};
-#define RS5C372_REG_BASE        0
+static int rs5c_get_regs(struct rs5c372 *rs5c)
+{
+        struct i2c_client       *client = rs5c->client;
+        struct i2c_msg          msgs[] = {
+                { client->addr, I2C_M_RD, sizeof rs5c->buf, rs5c->buf },
+        };
+        /* This implements the third reading method from the datasheet, using
+         * an internal address that's reset after each transaction (by STOP)
+         * to 0x0f ... so we read extra registers, and skip the first one.
+         *
+         * The first method doesn't work with the iop3xx adapter driver, on at
+         * least 80219 chips; this works around that bug.
+         */
+        if ((i2c_transfer(client->adapter, msgs, 1)) != 1) {
+                pr_debug("%s: can't read registers\n", rs5c->rtc->name);
+                return -EIO;
+        }
-static int rs5c372_attach(struct i2c_adapter *adapter);
+        dev_dbg(&client->dev,
-static int rs5c372_detach(struct i2c_client *client);
+                "%02x %02x %02x (%02x) %02x %02x %02x (%02x), "
-static int rs5c372_probe(struct i2c_adapter *adapter, int address, int kind);
+                "%02x %02x %02x, %02x %02x %02x; %02x %02x\n",
+                rs5c->regs[0],  rs5c->regs[1],  rs5c->regs[2],  rs5c->regs[3],
+                rs5c->regs[4],  rs5c->regs[5],  rs5c->regs[6],  rs5c->regs[7],
+                rs5c->regs[8],  rs5c->regs[9],  rs5c->regs[10], rs5c->regs[11],
+                rs5c->regs[12], rs5c->regs[13], rs5c->regs[14], rs5c->regs[15]);
-struct rs5c372 {
+        return 0;
-        u8 reg_addr;
+}
-        u8 regs[17];
-        struct i2c_msg msg[1];
-        struct i2c_client client;
-        struct rtc_device *rtc;
-};
-static struct i2c_driver rs5c372_driver = {
+static unsigned rs5c_reg2hr(struct rs5c372 *rs5c, unsigned reg)
-        .driver         = {
+{
-                .name   = "rs5c372",
+        unsigned        hour;
-        },
-        .attach_adapter = &rs5c372_attach,
-        .detach_client  = &rs5c372_detach,
-};
-static int rs5c372_get_datetime(struct i2c_client *client, struct rtc_time *tm)
+        if (rs5c->time24)
+                return BCD2BIN(reg & 0x3f);
+        hour = BCD2BIN(reg & 0x1f);
+        if (hour == 12)
+                hour = 0;
+        if (reg & 0x20)
+                hour += 12;
+        return hour;
+}
+static unsigned rs5c_hr2reg(struct rs5c372 *rs5c, unsigned hour)
 {
+        if (rs5c->time24)
+                return BIN2BCD(hour);
+        if (hour > 12)
+                return 0x20 | BIN2BCD(hour - 12);
+        if (hour == 12)
+                return 0x20 | BIN2BCD(12);
+        if (hour == 0)
+                return BIN2BCD(12);
+        return BIN2BCD(hour);
+}
-        struct rs5c372 *rs5c372 = i2c_get_clientdata(client);
+static int rs5c372_get_datetime(struct i2c_client *client, struct rtc_time *tm)
-        u8 *buf = &(rs5c372->regs[1]);
+{
+        struct rs5c372  *rs5c = i2c_get_clientdata(client);
+        int             status = rs5c_get_regs(rs5c);
-        /* this implements the 3rd reading method, according
+        if (status < 0)
-         * to the datasheet. rs5c372 defaults to internal
+                return status;
-         * address 0xF, so 0x0 is in regs[1]
-         */
-        if ((i2c_transfer(client->adapter, rs5c372->msg, 1)) != 1) {
+        tm->tm_sec = BCD2BIN(rs5c->regs[RS5C372_REG_SECS] & 0x7f);
-                dev_err(&client->dev, "%s: read error\n", __FUNCTION__);
+        tm->tm_min = BCD2BIN(rs5c->regs[RS5C372_REG_MINS] & 0x7f);
-                return -EIO;
+        tm->tm_hour = rs5c_reg2hr(rs5c, rs5c->regs[RS5C372_REG_HOURS]);
-        }
-        tm->tm_sec = BCD2BIN(buf[RS5C372_REG_SECS] & 0x7f);
+        tm->tm_wday = BCD2BIN(rs5c->regs[RS5C372_REG_WDAY] & 0x07);
-        tm->tm_min = BCD2BIN(buf[RS5C372_REG_MINS] & 0x7f);
+        tm->tm_mday = BCD2BIN(rs5c->regs[RS5C372_REG_DAY] & 0x3f);
-        tm->tm_hour = BCD2BIN(buf[RS5C372_REG_HOURS] & 0x3f);
-        tm->tm_wday = BCD2BIN(buf[RS5C372_REG_WDAY] & 0x07);
-        tm->tm_mday = BCD2BIN(buf[RS5C372_REG_DAY] & 0x3f);
        /* tm->tm_mon is zero-based */
-        tm->tm_mon = BCD2BIN(buf[RS5C372_REG_MONTH] & 0x1f) - 1;
+        tm->tm_mon = BCD2BIN(rs5c->regs[RS5C372_REG_MONTH] & 0x1f) - 1;
        /* year is 1900 + tm->tm_year */
-        tm->tm_year = BCD2BIN(buf[RS5C372_REG_YEAR]) + 100;
+        tm->tm_year = BCD2BIN(rs5c->regs[RS5C372_REG_YEAR]) + 100;
        dev_dbg(&client->dev, "%s: tm is secs=%d, mins=%d, hours=%d, "
                "mday=%d, mon=%d, year=%d, wday=%d\n",
@@ -89,22 +179,25 @@ static int rs5c372_get_datetime(struct i2c_client *client, struct rtc_time *tm)
                tm->tm_sec, tm->tm_min, tm->tm_hour,
                tm->tm_mday, tm->tm_mon, tm->tm_year, tm->tm_wday);
-        return 0;
+        /* rtc might need initialization */
+        return rtc_valid_tm(tm);
 }
 static int rs5c372_set_datetime(struct i2c_client *client, struct rtc_time *tm)
 {
-        unsigned char buf[8] = { RS5C372_REG_BASE };
+        struct rs5c372  *rs5c = i2c_get_clientdata(client);
+        unsigned char   buf[8];
-        dev_dbg(&client->dev,
+        dev_dbg(&client->dev, "%s: tm is secs=%d, mins=%d, hours=%d "
-                "%s: secs=%d, mins=%d, hours=%d "
                "mday=%d, mon=%d, year=%d, wday=%d\n",
-                __FUNCTION__, tm->tm_sec, tm->tm_min, tm->tm_hour,
+                __FUNCTION__,
+                tm->tm_sec, tm->tm_min, tm->tm_hour,
                tm->tm_mday, tm->tm_mon, tm->tm_year, tm->tm_wday);
+        buf[0] = RS5C_ADDR(RS5C372_REG_SECS);
        buf[1] = BIN2BCD(tm->tm_sec);
        buf[2] = BIN2BCD(tm->tm_min);
-        buf[3] = BIN2BCD(tm->tm_hour);
+        buf[3] = rs5c_hr2reg(rs5c, tm->tm_hour);
        buf[4] = BIN2BCD(tm->tm_wday);
        buf[5] = BIN2BCD(tm->tm_mday);
        buf[6] = BIN2BCD(tm->tm_mon + 1);
@@ -118,21 +211,43 @@ static int rs5c372_set_datetime(struct i2c_client *client, struct rtc_time *tm)
        return 0;
 }
+#if defined(CONFIG_RTC_INTF_PROC) || defined(CONFIG_RTC_INTF_PROC_MODULE)
+#define NEED_TRIM
+#endif
+#if defined(CONFIG_RTC_INTF_SYSFS) || defined(CONFIG_RTC_INTF_SYSFS_MODULE)
+#define NEED_TRIM
+#endif
+#ifdef  NEED_TRIM
 static int rs5c372_get_trim(struct i2c_client *client, int *osc, int *trim)
 {
        struct rs5c372 *rs5c372 = i2c_get_clientdata(client);
-        u8 tmp = rs5c372->regs[RS5C372_REG_TRIM + 1];
+        u8 tmp = rs5c372->regs[RS5C372_REG_TRIM];
        if (osc)
                *osc = (tmp & RS5C372_TRIM_XSL) ? 32000 : 32768;
        if (trim) {
-                *trim = tmp & RS5C372_TRIM_MASK;
+                dev_dbg(&client->dev, "%s: raw trim=%x\n", __FUNCTION__, tmp);
-                dev_dbg(&client->dev, "%s: raw trim=%x\n", __FUNCTION__, *trim);
+                tmp &= RS5C372_TRIM_MASK;
+                if (tmp & 0x3e) {
+                        int t = tmp & 0x3f;
+                        if (tmp & 0x40)
+                                t = (~t | (s8)0xc0) + 1;
+                        else
+                                t = t - 1;
+                        tmp = t * 2;
+                } else
+                        tmp = 0;
+                *trim = tmp;
        }
        return 0;
 }
+#endif
 static int rs5c372_rtc_read_time(struct device *dev, struct rtc_time *tm)
 {
@@ -144,25 +259,190 @@ static int rs5c372_rtc_set_time(struct device *dev, struct rtc_time *tm)
        return rs5c372_set_datetime(to_i2c_client(dev), tm);
 }
+#if defined(CONFIG_RTC_INTF_DEV) || defined(CONFIG_RTC_INTF_DEV_MODULE)
+static int
+rs5c_rtc_ioctl(struct device *dev, unsigned int cmd, unsigned long arg)
+{
+        struct i2c_client       *client = to_i2c_client(dev);
+        struct rs5c372          *rs5c = i2c_get_clientdata(client);
+        unsigned char           buf[2];
+        int                     status;
+        buf[1] = rs5c->regs[RS5C_REG_CTRL1];
+        switch (cmd) {
+        case RTC_UIE_OFF:
+        case RTC_UIE_ON:
+                /* some 327a modes use a different IRQ pin for 1Hz irqs */
+                if (rs5c->type == rtc_rs5c372a
+                                && (buf[1] & RS5C372A_CTRL1_SL1))
+                        return -ENOIOCTLCMD;
+        case RTC_AIE_OFF:
+        case RTC_AIE_ON:
+                /* these irq management calls only make sense for chips
+                 * which are wired up to an IRQ.
+                 */
+                if (!rs5c->has_irq)
+                        return -ENOIOCTLCMD;
+                break;
+        default:
+                return -ENOIOCTLCMD;
+        }
+        status = rs5c_get_regs(rs5c);
+        if (status < 0)
+                return status;
+        buf[0] = RS5C_ADDR(RS5C_REG_CTRL1);
+        switch (cmd) {
+        case RTC_AIE_OFF:       /* alarm off */
+                buf[1] &= ~RS5C_CTRL1_AALE;
+                break;
+        case RTC_AIE_ON:        /* alarm on */
+                buf[1] |= RS5C_CTRL1_AALE;
+                break;
+        case RTC_UIE_OFF:       /* update off */
+                buf[1] &= ~RS5C_CTRL1_CT_MASK;
+                break;
+        case RTC_UIE_ON:        /* update on */
+                buf[1] &= ~RS5C_CTRL1_CT_MASK;
+                buf[1] |= RS5C_CTRL1_CT4;
+                break;
+        }
+        if ((i2c_master_send(client, buf, 2)) != 2) {
+                printk(KERN_WARNING "%s: can't update alarm\n",
+                        rs5c->rtc->name);
+                status = -EIO;
+        } else
+                rs5c->regs[RS5C_REG_CTRL1] = buf[1];
+        return status;
+}
+#else
+#define rs5c_rtc_ioctl  NULL
+#endif
+/* NOTE:  Since RTC_WKALM_{RD,SET} were originally defined for EFI,
+ * which only exposes a polled programming interface; and since
+ * these calls map directly to those EFI requests; we don't demand
+ * we have an IRQ for this chip when we go through this API.
+ *
+ * The older x86_pc derived RTC_ALM_{READ,SET} calls require irqs
+ * though, managed through RTC_AIE_{ON,OFF} requests.
+ */
+static int rs5c_read_alarm(struct device *dev, struct rtc_wkalrm *t)
+{
+        struct i2c_client       *client = to_i2c_client(dev);
+        struct rs5c372          *rs5c = i2c_get_clientdata(client);
+        int                     status;
+        status = rs5c_get_regs(rs5c);
+        if (status < 0)
+                return status;
+        /* report alarm time */
+        t->time.tm_sec = 0;
+        t->time.tm_min = BCD2BIN(rs5c->regs[RS5C_REG_ALARM_A_MIN] & 0x7f);
+        t->time.tm_hour = rs5c_reg2hr(rs5c, rs5c->regs[RS5C_REG_ALARM_A_HOURS]);
+        t->time.tm_mday = -1;
+        t->time.tm_mon = -1;
+        t->time.tm_year = -1;
+        t->time.tm_wday = -1;
+        t->time.tm_yday = -1;
+        t->time.tm_isdst = -1;
+        /* ... and status */
+        t->enabled = !!(rs5c->regs[RS5C_REG_CTRL1] & RS5C_CTRL1_AALE);
+        t->pending = !!(rs5c->regs[RS5C_REG_CTRL2] & RS5C_CTRL2_AAFG);
+        return 0;
+}
+static int rs5c_set_alarm(struct device *dev, struct rtc_wkalrm *t)
+{
+        struct i2c_client       *client = to_i2c_client(dev);
+        struct rs5c372          *rs5c = i2c_get_clientdata(client);
+        int                     status;
+        unsigned char           buf[4];
+        /* only handle up to 24 hours in the future, like RTC_ALM_SET */
+        if (t->time.tm_mday != -1
+                        || t->time.tm_mon != -1
+                        || t->time.tm_year != -1)
+                return -EINVAL;
+        /* REVISIT: round up tm_sec */
+        /* if needed, disable irq (clears pending status) */
+        status = rs5c_get_regs(rs5c);
+        if (status < 0)
+                return status;
+        if (rs5c->regs[RS5C_REG_CTRL1] & RS5C_CTRL1_AALE) {
+                buf[0] = RS5C_ADDR(RS5C_REG_CTRL1);
+                buf[1] = rs5c->regs[RS5C_REG_CTRL1] & ~RS5C_CTRL1_AALE;
+                if (i2c_master_send(client, buf, 2) != 2) {
+                        pr_debug("%s: can't disable alarm\n", rs5c->rtc->name);
+                        return -EIO;
+                }
+                rs5c->regs[RS5C_REG_CTRL1] = buf[1];
+        }
+        /* set alarm */
+        buf[0] = RS5C_ADDR(RS5C_REG_ALARM_A_MIN);
+        buf[1] = BIN2BCD(t->time.tm_min);
+        buf[2] = rs5c_hr2reg(rs5c, t->time.tm_hour);
+        buf[3] = 0x7f;  /* any/all days */
+        if ((i2c_master_send(client, buf, 4)) != 4) {
+                pr_debug("%s: can't set alarm time\n", rs5c->rtc->name);
+                return -EIO;
+        }
+        /* ... and maybe enable its irq */
+        if (t->enabled) {
+                buf[0] = RS5C_ADDR(RS5C_REG_CTRL1);
+                buf[1] = rs5c->regs[RS5C_REG_CTRL1] | RS5C_CTRL1_AALE;
+                if ((i2c_master_send(client, buf, 2)) != 2)
+                        printk(KERN_WARNING "%s: can't enable alarm\n",
+                                rs5c->rtc->name);
+                rs5c->regs[RS5C_REG_CTRL1] = buf[1];
+        }
+        return 0;
+}
+#if defined(CONFIG_RTC_INTF_PROC) || defined(CONFIG_RTC_INTF_PROC_MODULE)
 static int rs5c372_rtc_proc(struct device *dev, struct seq_file *seq)
 {
        int err, osc, trim;
        err = rs5c372_get_trim(to_i2c_client(dev), &osc, &trim);
        if (err == 0) {
-                seq_printf(seq, "%d.%03d KHz\n", osc / 1000, osc % 1000);
+                seq_printf(seq, "crystal\t\t: %d.%03d KHz\n",
-                seq_printf(seq, "trim\t: %d\n", trim);
+                                osc / 1000, osc % 1000);
+                seq_printf(seq, "trim\t\t: %d\n", trim);
        }
        return 0;
 }
+#else
+#define rs5c372_rtc_proc        NULL
+#endif
 static const struct rtc_class_ops rs5c372_rtc_ops = {
        .proc           = rs5c372_rtc_proc,
+        .ioctl          = rs5c_rtc_ioctl,
        .read_time      = rs5c372_rtc_read_time,
        .set_time       = rs5c372_rtc_set_time,
+        .read_alarm     = rs5c_read_alarm,
+        .set_alarm      = rs5c_set_alarm,
 };
+#if defined(CONFIG_RTC_INTF_SYSFS) || defined(CONFIG_RTC_INTF_SYSFS_MODULE)
 static ssize_t rs5c372_sysfs_show_trim(struct device *dev,
                                struct device_attribute *attr, char *buf)
 {
@@ -172,7 +452,7 @@ static ssize_t rs5c372_sysfs_show_trim(struct device *dev,
        if (err)
                return err;
-        return sprintf(buf, "0x%2x\n", trim);
+        return sprintf(buf, "%d\n", trim);
 }
 static DEVICE_ATTR(trim, S_IRUGO, rs5c372_sysfs_show_trim, NULL);
@@ -189,16 +469,35 @@ static ssize_t rs5c372_sysfs_show_osc(struct device *dev,
 }
 static DEVICE_ATTR(osc, S_IRUGO, rs5c372_sysfs_show_osc, NULL);
-static int rs5c372_attach(struct i2c_adapter *adapter)
+static int rs5c_sysfs_register(struct device *dev)
 {
-        return i2c_probe(adapter, &addr_data, rs5c372_probe);
+        int err;
+        err = device_create_file(dev, &dev_attr_trim);
+        if (err)
+                return err;
+        err = device_create_file(dev, &dev_attr_osc);
+        if (err)
+                device_remove_file(dev, &dev_attr_trim);
+        return err;
+}
+#else
+static int rs5c_sysfs_register(struct device *dev)
+{
+        return 0;
 }
+#endif  /* SYSFS */
+static struct i2c_driver rs5c372_driver;
 static int rs5c372_probe(struct i2c_adapter *adapter, int address, int kind)
 {
        int err = 0;
        struct i2c_client *client;
        struct rs5c372 *rs5c372;
+        struct rtc_time tm;
        dev_dbg(adapter->class_dev.dev, "%s\n", __FUNCTION__);
@@ -211,7 +510,15 @@ static int rs5c372_probe(struct i2c_adapter *adapter, int address, int kind)
                err = -ENOMEM;
                goto exit;
        }
-        client = &rs5c372->client;
+        /* we read registers 0x0f then 0x00-0x0f; skip the first one */
+        rs5c372->regs=&rs5c372->buf[1];
+        /* On conversion to a "new style" i2c driver, we'll be handed
+         * the i2c_client (we won't create it)
+         */
+        client = &rs5c372->dev;
+        rs5c372->client = client;
        /* I2C client */
        client->addr = address;
@@ -222,16 +529,99 @@ static int rs5c372_probe(struct i2c_adapter *adapter, int address, int kind)
        i2c_set_clientdata(client, rs5c372);
-        rs5c372->msg[0].addr = address;
-        rs5c372->msg[0].flags = I2C_M_RD;
-        rs5c372->msg[0].len = sizeof(rs5c372->regs);
-        rs5c372->msg[0].buf = rs5c372->regs;
        /* Inform the i2c layer */
        if ((err = i2c_attach_client(client)))
                goto exit_kfree;
-        dev_info(&client->dev, "chip found, driver version " DRV_VERSION "\n");
+        err = rs5c_get_regs(rs5c372);
+        if (err < 0)
+                goto exit_detach;
+        /* For "new style" drivers, irq is in i2c_client and chip type
+         * info comes from i2c_client.dev.platform_data.  Meanwhile:
+         *
+         * STICK BOARD-SPECIFIC SETUP CODE RIGHT HERE
+         */
+        if (rs5c372->type == rtc_undef) {
+                rs5c372->type = rtc_rs5c372b;
+                dev_warn(&client->dev, "assuming rs5c372b\n");
+        }
+        /* clock may be set for am/pm or 24 hr time */
+        switch (rs5c372->type) {
+        case rtc_rs5c372a:
+        case rtc_rs5c372b:
+                /* alarm uses ALARM_A; and nINTRA on 372a, nINTR on 372b.
+                 * so does periodic irq, except some 327a modes.
+                 */
+                if (rs5c372->regs[RS5C_REG_CTRL2] & RS5C372_CTRL2_24)
+                        rs5c372->time24 = 1;
+                break;
+        case rtc_rv5c386:
+        case rtc_rv5c387a:
+                if (rs5c372->regs[RS5C_REG_CTRL1] & RV5C387_CTRL1_24)
+                        rs5c372->time24 = 1;
+                /* alarm uses ALARM_W; and nINTRB for alarm and periodic
+                 * irq, on both 386 and 387
+                 */
+                break;
+        default:
+                dev_err(&client->dev, "unknown RTC type\n");
+                goto exit_detach;
+        }
+        /* if the oscillator lost power and no other software (like
+         * the bootloader) set it up, do it here.
+         */
+        if (rs5c372->regs[RS5C_REG_CTRL2] & RS5C_CTRL2_XSTP) {
+                unsigned char buf[3];
+                rs5c372->regs[RS5C_REG_CTRL2] &= ~RS5C_CTRL2_XSTP;
+                buf[0] = RS5C_ADDR(RS5C_REG_CTRL1);
+                buf[1] = rs5c372->regs[RS5C_REG_CTRL1];
+                buf[2] = rs5c372->regs[RS5C_REG_CTRL2];
+                /* use 24hr mode */
+                switch (rs5c372->type) {
+                case rtc_rs5c372a:
+                case rtc_rs5c372b:
+                        buf[2] |= RS5C372_CTRL2_24;
+                        rs5c372->time24 = 1;
+                        break;
+                case rtc_rv5c386:
+                case rtc_rv5c387a:
+                        buf[1] |= RV5C387_CTRL1_24;
+                        rs5c372->time24 = 1;
+                        break;
+                default:
+                        /* impossible */
+                        break;
+                }
+                if ((i2c_master_send(client, buf, 3)) != 3) {
+                        dev_err(&client->dev, "setup error\n");
+                        goto exit_detach;
+                }
+                rs5c372->regs[RS5C_REG_CTRL1] = buf[1];
+                rs5c372->regs[RS5C_REG_CTRL2] = buf[2];
+        }
+        if (rs5c372_get_datetime(client, &tm) < 0)
+                dev_warn(&client->dev, "clock needs to be set\n");
+        dev_info(&client->dev, "%s found, %s, driver version " DRV_VERSION "\n",
+                        ({ char *s; switch (rs5c372->type) {
+                        case rtc_rs5c372a:      s = "rs5c372a"; break;
+                        case rtc_rs5c372b:      s = "rs5c372b"; break;
+                        case rtc_rv5c386:       s = "rv5c386"; break;
+                        case rtc_rv5c387a:      s = "rv5c387a"; break;
+                        default:                s = "chip"; break;
+                        }; s;}),
+                        rs5c372->time24 ? "24hr" : "am/pm"
+                        );
+        /* FIXME when client->irq exists, use it to register alarm irq */
        rs5c372->rtc = rtc_device_register(rs5c372_driver.driver.name,
                                &client->dev, &rs5c372_rtc_ops, THIS_MODULE);
@@ -241,18 +631,12 @@ static int rs5c372_probe(struct i2c_adapter *adapter, int address, int kind)
                goto exit_detach;
        }
-        err = device_create_file(&client->dev, &dev_attr_trim);
+        err = rs5c_sysfs_register(&client->dev);
        if (err)
                goto exit_devreg;
-        err = device_create_file(&client->dev, &dev_attr_osc);
-        if (err)
-                goto exit_trim;
        return 0;
-exit_trim:
-        device_remove_file(&client->dev, &dev_attr_trim);
 exit_devreg:
        rtc_device_unregister(rs5c372->rtc);
@@ -266,6 +650,11 @@ exit:
        return err;
 }
+static int rs5c372_attach(struct i2c_adapter *adapter)
+{
+        return i2c_probe(adapter, &addr_data, rs5c372_probe);
+}
 static int rs5c372_detach(struct i2c_client *client)
 {
        int err;
@@ -274,6 +663,8 @@ static int rs5c372_detach(struct i2c_client *client)
        if (rs5c372->rtc)
                rtc_device_unregister(rs5c372->rtc);
+        /* REVISIT properly destroy the sysfs files ... */
        if ((err = i2c_detach_client(client)))
                return err;
@@ -281,6 +672,14 @@ static int rs5c372_detach(struct i2c_client *client)
        return 0;
 }
+static struct i2c_driver rs5c372_driver = {
+        .driver         = {
+                .name   = "rtc-rs5c372",
+        },
+        .attach_adapter = &rs5c372_attach,
+        .detach_client  = &rs5c372_detach,
+};
 static __init int rs5c372_init(void)
 {
        return i2c_add_driver(&rs5c372_driver);
diff --git a/drivers/s390/char/vmcp.c b/drivers/s390/char/vmcp.c
index 1678b6c757ec..a420cd099041 100644
--- a/drivers/s390/char/vmcp.c
+++ b/drivers/s390/char/vmcp.c
@@ -117,7 +117,7 @@ vmcp_write(struct file *file, const char __user * buff, size_t count,
                return -ENOMEM;
        }
        debug_text_event(vmcp_debug, 1, cmd);
-        session->resp_size = __cpcmd(cmd, session->response,
+        session->resp_size = cpcmd(cmd, session->response,
                                     session->bufsize,
                                     &session->resp_code);
        up(&session->mutex);
diff --git a/drivers/s390/cio/cio.c b/drivers/s390/cio/cio.c
index b471ac4a1bf6..ae1bf231d089 100644
--- a/drivers/s390/cio/cio.c
+++ b/drivers/s390/cio/cio.c
@@ -880,19 +880,15 @@ static void cio_reset_pgm_check_handler(void)
 static int stsch_reset(struct subchannel_id schid, volatile struct schib *addr)
 {
        int rc;
-        register struct subchannel_id reg1 asm ("1") = schid;
        pgm_check_occured = 0;
        s390_reset_pgm_handler = cio_reset_pgm_check_handler;
+        rc = stsch(schid, addr);
+        s390_reset_pgm_handler = NULL;
-        asm volatile(
+        /* The program check handler could have changed pgm_check_occured */
-                "       stsch   0(%2)\n"
+        barrier();
-                "       ipm     %0\n"
-                "       srl     %0,28"
-                : "=d" (rc)
-                : "d" (reg1), "a" (addr), "m" (*addr) : "memory", "cc");
-        s390_reset_pgm_handler = NULL;
        if (pgm_check_occured)
                return -EIO;
        else
diff --git a/drivers/s390/net/Kconfig b/drivers/s390/net/Kconfig
index 1a93fa684e9f..52625153a4f0 100644
--- a/drivers/s390/net/Kconfig
+++ b/drivers/s390/net/Kconfig
@@ -27,10 +27,7 @@ config IUCV
        help
          Select this option if you want to use inter-user communication
          under VM or VIF. If unsure, say "Y" to enable a fast communication
-          link between VM guests. At boot time the user ID of the guest needs
+          link between VM guests.
-          to be passed to the kernel. Note that both kernels need to be
-          compiled with this option and both need to be booted with the user ID
-          of the other VM guest.
 config NETIUCV
        tristate "IUCV network device support (VM only)"
diff --git a/drivers/s390/net/qeth.h b/drivers/s390/net/qeth.h
index 53c358c7d368..e95c281f1e36 100644
--- a/drivers/s390/net/qeth.h
+++ b/drivers/s390/net/qeth.h
@@ -710,7 +710,7 @@ struct qeth_reply {
        int (*callback)(struct qeth_card *,struct qeth_reply *,unsigned long);
        u32 seqno;
        unsigned long offset;
-        int received;
+        atomic_t received;
        int rc;
        void *param;
        struct qeth_card *card;
diff --git a/drivers/s390/net/qeth_main.c b/drivers/s390/net/qeth_main.c
index 2bde4f1fb9c2..d2efa5ff125d 100644
--- a/drivers/s390/net/qeth_main.c
+++ b/drivers/s390/net/qeth_main.c
@@ -471,7 +471,7 @@ qeth_irq(struct ccw_device *cdev, unsigned long intparm, struct irb *irb)
            channel->state == CH_STATE_UP)
                qeth_issue_next_read(card);
-        tasklet_schedule(&channel->irq_tasklet);
+        qeth_irq_tasklet((unsigned long)channel);
        return;
 out:
        wake_up(&card->wait_q);
@@ -951,40 +951,6 @@ qeth_do_run_thread(struct qeth_card *card, unsigned long thread)
 }
 static int
-qeth_register_ip_addresses(void *ptr)
-{
-        struct qeth_card *card;
-        card = (struct qeth_card *) ptr;
-        daemonize("qeth_reg_ip");
-        QETH_DBF_TEXT(trace,4,"regipth1");
-        if (!qeth_do_run_thread(card, QETH_SET_IP_THREAD))
-                return 0;
-        QETH_DBF_TEXT(trace,4,"regipth2");
-        qeth_set_ip_addr_list(card);
-        qeth_clear_thread_running_bit(card, QETH_SET_IP_THREAD);
-        return 0;
-}
-/*
- * Drive the SET_PROMISC_MODE thread
- */
-static int
-qeth_set_promisc_mode(void *ptr)
-{
-        struct qeth_card *card = (struct qeth_card *) ptr;
-        daemonize("qeth_setprm");
-        QETH_DBF_TEXT(trace,4,"setprm1");
-        if (!qeth_do_run_thread(card, QETH_SET_PROMISC_MODE_THREAD))
-                return 0;
-        QETH_DBF_TEXT(trace,4,"setprm2");
-        qeth_setadp_promisc_mode(card);
-        qeth_clear_thread_running_bit(card, QETH_SET_PROMISC_MODE_THREAD);
-        return 0;
-}
-static int
 qeth_recover(void *ptr)
 {
        struct qeth_card *card;
@@ -1047,11 +1013,6 @@ qeth_start_kernel_thread(struct work_struct *work)
        if (card->read.state != CH_STATE_UP &&
            card->write.state != CH_STATE_UP)
                return;
-        if (qeth_do_start_thread(card, QETH_SET_IP_THREAD))
-                kernel_thread(qeth_register_ip_addresses, (void *)card,SIGCHLD);
-        if (qeth_do_start_thread(card, QETH_SET_PROMISC_MODE_THREAD))
-                kernel_thread(qeth_set_promisc_mode, (void *)card, SIGCHLD);
        if (qeth_do_start_thread(card, QETH_RECOVER_THREAD))
                kernel_thread(qeth_recover, (void *) card, SIGCHLD);
 }
@@ -1074,7 +1035,7 @@ qeth_set_intial_options(struct qeth_card *card)
                card->options.layer2 = 1;
        else
                card->options.layer2 = 0;
-        card->options.performance_stats = 1;
+        card->options.performance_stats = 0;
 }
 /**
@@ -1613,8 +1574,6 @@ qeth_issue_next_read(struct qeth_card *card)
                return -ENOMEM;
        }
        qeth_setup_ccw(&card->read, iob->data, QETH_BUFSIZE);
-        wait_event(card->wait_q,
-                   atomic_cmpxchg(&card->read.irq_pending, 0, 1) == 0);
        QETH_DBF_TEXT(trace, 6, "noirqpnd");
        rc = ccw_device_start(card->read.ccwdev, &card->read.ccw,
                              (addr_t) iob, 0, 0);
@@ -1635,6 +1594,7 @@ qeth_alloc_reply(struct qeth_card *card)
        reply = kzalloc(sizeof(struct qeth_reply), GFP_ATOMIC);
        if (reply){
                atomic_set(&reply->refcnt, 1);
+                atomic_set(&reply->received, 0);
                reply->card = card;
        };
        return reply;
@@ -1655,31 +1615,6 @@ qeth_put_reply(struct qeth_reply *reply)
                kfree(reply);
 }
-static void
-qeth_cmd_timeout(unsigned long data)
-{
-        struct qeth_reply *reply, *list_reply, *r;
-        unsigned long flags;
-        reply = (struct qeth_reply *) data;
-        spin_lock_irqsave(&reply->card->lock, flags);
-        list_for_each_entry_safe(list_reply, r,
-                                 &reply->card->cmd_waiter_list, list) {
-                if (reply == list_reply){
-                        qeth_get_reply(reply);
-                        list_del_init(&reply->list);
-                        spin_unlock_irqrestore(&reply->card->lock, flags);
-                        reply->rc = -ETIME;
-                        reply->received = 1;
-                        wake_up(&reply->wait_q);
-                        qeth_put_reply(reply);
-                        return;
-                }
-        }
-        spin_unlock_irqrestore(&reply->card->lock, flags);
-}
 static struct qeth_ipa_cmd *
 qeth_check_ipa_data(struct qeth_card *card, struct qeth_cmd_buffer *iob)
 {
@@ -1745,7 +1680,7 @@ qeth_clear_ipacmd_list(struct qeth_card *card)
        list_for_each_entry_safe(reply, r, &card->cmd_waiter_list, list) {
                qeth_get_reply(reply);
                reply->rc = -EIO;
-                reply->received = 1;
+                atomic_inc(&reply->received);
                list_del_init(&reply->list);
                wake_up(&reply->wait_q);
                qeth_put_reply(reply);
@@ -1814,7 +1749,7 @@ qeth_send_control_data_cb(struct qeth_channel *channel,
                                              &card->cmd_waiter_list);
                                spin_unlock_irqrestore(&card->lock, flags);
                        } else {
-                                reply->received = 1;
+                                atomic_inc(&reply->received);
                                wake_up(&reply->wait_q);
                        }
                        qeth_put_reply(reply);
@@ -1858,7 +1793,7 @@ qeth_send_control_data(struct qeth_card *card, int len,
        int rc;
        unsigned long flags;
        struct qeth_reply *reply = NULL;
-        struct timer_list timer;
+        unsigned long timeout;
        QETH_DBF_TEXT(trace, 2, "sendctl");
@@ -1873,21 +1808,20 @@ qeth_send_control_data(struct qeth_card *card, int len,
                reply->seqno = QETH_IDX_COMMAND_SEQNO;
        else
                reply->seqno = card->seqno.ipa++;
-        init_timer(&timer);
-        timer.function = qeth_cmd_timeout;
-        timer.data = (unsigned long) reply;
        init_waitqueue_head(&reply->wait_q);
        spin_lock_irqsave(&card->lock, flags);
        list_add_tail(&reply->list, &card->cmd_waiter_list);
        spin_unlock_irqrestore(&card->lock, flags);
        QETH_DBF_HEX(control, 2, iob->data, QETH_DBF_CONTROL_LEN);
-        wait_event(card->wait_q,
-                   atomic_cmpxchg(&card->write.irq_pending, 0, 1) == 0);
+        while (atomic_cmpxchg(&card->write.irq_pending, 0, 1)) ;
        qeth_prepare_control_data(card, len, iob);
        if (IS_IPA(iob->data))
-                timer.expires = jiffies + QETH_IPA_TIMEOUT;
+                timeout = jiffies + QETH_IPA_TIMEOUT;
        else
-                timer.expires = jiffies + QETH_TIMEOUT;
+                timeout = jiffies + QETH_TIMEOUT;
        QETH_DBF_TEXT(trace, 6, "noirqpnd");
        spin_lock_irqsave(get_ccwdev_lock(card->write.ccwdev), flags);
        rc = ccw_device_start(card->write.ccwdev, &card->write.ccw,
@@ -1906,9 +1840,16 @@ qeth_send_control_data(struct qeth_card *card, int len,
                wake_up(&card->wait_q);
                return rc;
        }
-        add_timer(&timer);
+        while (!atomic_read(&reply->received)) {
-        wait_event(reply->wait_q, reply->received);
+                if (time_after(jiffies, timeout)) {
-        del_timer_sync(&timer);
+                        spin_lock_irqsave(&reply->card->lock, flags);
+                        list_del_init(&reply->list);
+                        spin_unlock_irqrestore(&reply->card->lock, flags);
+                        reply->rc = -ETIME;
+                        atomic_inc(&reply->received);
+                        wake_up(&reply->wait_q);
+                }
+        };
        rc = reply->rc;
        qeth_put_reply(reply);
        return rc;
@@ -2466,32 +2407,17 @@ qeth_rebuild_skb_fake_ll(struct qeth_card *card, struct sk_buff *skb,
                qeth_rebuild_skb_fake_ll_eth(card, skb, hdr);
 }
-static inline __u16
+static inline void
 qeth_layer2_rebuild_skb(struct qeth_card *card, struct sk_buff *skb,
                        struct qeth_hdr *hdr)
 {
-        unsigned short vlan_id = 0;
-#ifdef CONFIG_QETH_VLAN
-        struct vlan_hdr *vhdr;
-#endif
        skb->pkt_type = PACKET_HOST;
        skb->protocol = qeth_type_trans(skb, skb->dev);
        if (card->options.checksum_type == NO_CHECKSUMMING)
                skb->ip_summed = CHECKSUM_UNNECESSARY;
        else
                skb->ip_summed = CHECKSUM_NONE;
-#ifdef CONFIG_QETH_VLAN
-        if (hdr->hdr.l2.flags[2] & (QETH_LAYER2_FLAG_VLAN)) {
-                vhdr = (struct vlan_hdr *) skb->data;
-                skb->protocol =
-                        __constant_htons(vhdr->h_vlan_encapsulated_proto);
-                vlan_id = hdr->hdr.l2.vlan_id;
-                skb_pull(skb, VLAN_HLEN);
-        }
-#endif
        *((__u32 *)skb->cb) = ++card->seqno.pkt_seqno;
-        return vlan_id;
 }
 static inline __u16
@@ -2560,7 +2486,6 @@ qeth_process_inbound_buffer(struct qeth_card *card,
        int offset;
        int rxrc;
        __u16 vlan_tag = 0;
-        __u16 *vlan_addr;
        /* get first element of current buffer */
        element = (struct qdio_buffer_element *)&buf->buffer->element[0];
@@ -2571,7 +2496,7 @@ qeth_process_inbound_buffer(struct qeth_card *card,
                                       &offset, &hdr))) {
                skb->dev = card->dev;
                if (hdr->hdr.l2.id == QETH_HEADER_TYPE_LAYER2)
-                        vlan_tag = qeth_layer2_rebuild_skb(card, skb, hdr);
+                        qeth_layer2_rebuild_skb(card, skb, hdr);
                else if (hdr->hdr.l3.id == QETH_HEADER_TYPE_LAYER3)
                        vlan_tag = qeth_rebuild_skb(card, skb, hdr);
                else { /*in case of OSN*/
@@ -3968,13 +3893,22 @@ static inline struct sk_buff *
 qeth_prepare_skb(struct qeth_card *card, struct sk_buff *skb,
                 struct qeth_hdr **hdr, int ipv)
 {
-        struct sk_buff *new_skb;
+        struct sk_buff *new_skb, *new_skb2;
        
        QETH_DBF_TEXT(trace, 6, "prepskb");
+        new_skb = skb;
-        new_skb = qeth_realloc_headroom(card, skb, sizeof(struct qeth_hdr));
+        new_skb = qeth_pskb_unshare(skb, GFP_ATOMIC);
-        if (new_skb == NULL)
+        if (!new_skb)
+                return NULL;
+        new_skb2 = qeth_realloc_headroom(card, new_skb,
+                                         sizeof(struct qeth_hdr));
+        if (!new_skb2) {
+                __qeth_free_new_skb(skb, new_skb);
                return NULL;
+        }
+        if (new_skb != skb)
+                __qeth_free_new_skb(new_skb2, new_skb);
+        new_skb = new_skb2;
        *hdr = __qeth_prepare_skb(card, new_skb, ipv);
        if (*hdr == NULL) {
                __qeth_free_new_skb(skb, new_skb);
@@ -4844,9 +4778,11 @@ qeth_arp_query(struct qeth_card *card, char __user *udata)
                           "(0x%x/%d)\n",
                           QETH_CARD_IFNAME(card), qeth_arp_get_error_cause(&rc),
                           tmp, tmp);
-                copy_to_user(udata, qinfo.udata, 4);
+                if (copy_to_user(udata, qinfo.udata, 4))
+                        rc = -EFAULT;
        } else {
-                copy_to_user(udata, qinfo.udata, qinfo.udata_len);
+                if (copy_to_user(udata, qinfo.udata, qinfo.udata_len))
+                        rc = -EFAULT;
        }
        kfree(qinfo.udata);
        return rc;
@@ -4992,8 +4928,10 @@ qeth_snmp_command(struct qeth_card *card, char __user *udata)
        if (rc)
                PRINT_WARN("SNMP command failed on %s: (0x%x)\n",
                           QETH_CARD_IFNAME(card), rc);
-         else
+        else {
-                copy_to_user(udata, qinfo.udata, qinfo.udata_len);
+                if (copy_to_user(udata, qinfo.udata, qinfo.udata_len))
+                        rc = -EFAULT;
+        }
        kfree(ureq);
        kfree(qinfo.udata);
@@ -5544,12 +5482,10 @@ qeth_set_multicast_list(struct net_device *dev)
        qeth_add_multicast_ipv6(card);
 #endif
 out:
-        if (qeth_set_thread_start_bit(card, QETH_SET_IP_THREAD) == 0)
+        qeth_set_ip_addr_list(card);
-                schedule_work(&card->kernel_thread_starter);
        if (!qeth_adp_supported(card, IPA_SETADP_SET_PROMISC_MODE))
                return;
-        if (qeth_set_thread_start_bit(card, QETH_SET_PROMISC_MODE_THREAD)==0)
+        qeth_setadp_promisc_mode(card);
-                schedule_work(&card->kernel_thread_starter);
 }
 static int
@@ -6351,6 +6287,42 @@ static struct ethtool_ops qeth_ethtool_ops = {
 };
 static int
+qeth_hard_header_parse(struct sk_buff *skb, unsigned char *haddr)
+{
+        struct qeth_card *card;
+        struct ethhdr *eth;
+        card = qeth_get_card_from_dev(skb->dev);
+        if (card->options.layer2)
+                goto haveheader;
+#ifdef CONFIG_QETH_IPV6
+        /* cause of the manipulated arp constructor and the ARP
+           flag for OSAE devices we have some nasty exceptions */
+        if (card->info.type == QETH_CARD_TYPE_OSAE) {
+                if (!card->options.fake_ll) {
+                        if ((skb->pkt_type==PACKET_OUTGOING) &&
+                            (skb->protocol==ETH_P_IPV6))
+                                goto haveheader;
+                        else
+                                return 0;
+                } else {
+                        if ((skb->pkt_type==PACKET_OUTGOING) &&
+                            (skb->protocol==ETH_P_IP))
+                                return 0;
+                        else
+                                goto haveheader;
+                }
+        }
+#endif
+        if (!card->options.fake_ll)
+                return 0;
+haveheader:
+        eth = eth_hdr(skb);
+        memcpy(haddr, eth->h_source, ETH_ALEN);
+        return ETH_ALEN;
+}
+static int
 qeth_netdev_init(struct net_device *dev)
 {
        struct qeth_card *card;
@@ -6388,7 +6360,10 @@ qeth_netdev_init(struct net_device *dev)
        if (card->options.fake_ll &&
                (qeth_get_netdev_flags(card) & IFF_NOARP))
                        dev->hard_header = qeth_fake_header;
-        dev->hard_header_parse = NULL;
+        if (dev->type == ARPHRD_IEEE802_TR)
+                dev->hard_header_parse = NULL;
+        else
+                dev->hard_header_parse = qeth_hard_header_parse;
        dev->set_mac_address = qeth_layer2_set_mac_address;
        dev->flags |= qeth_get_netdev_flags(card);
        if ((card->options.fake_broadcast) ||
@@ -8235,8 +8210,7 @@ qeth_add_vipa(struct qeth_card *card, enum qeth_prot_versions proto,
        }
        if (!qeth_add_ip(card, ipaddr))
                kfree(ipaddr);
-        if (qeth_set_thread_start_bit(card, QETH_SET_IP_THREAD) == 0)
+        qeth_set_ip_addr_list(card);
-                schedule_work(&card->kernel_thread_starter);
        return rc;
 }
@@ -8264,8 +8238,7 @@ qeth_del_vipa(struct qeth_card *card, enum qeth_prot_versions proto,
                return;
        if (!qeth_delete_ip(card, ipaddr))
                kfree(ipaddr);
-        if (qeth_set_thread_start_bit(card, QETH_SET_IP_THREAD) == 0)
+        qeth_set_ip_addr_list(card);
-                schedule_work(&card->kernel_thread_starter);
 }
 /*
@@ -8308,8 +8281,7 @@ qeth_add_rxip(struct qeth_card *card, enum qeth_prot_versions proto,
        }
        if (!qeth_add_ip(card, ipaddr))
                kfree(ipaddr);
-        if (qeth_set_thread_start_bit(card, QETH_SET_IP_THREAD) == 0)
+        qeth_set_ip_addr_list(card);
-                schedule_work(&card->kernel_thread_starter);
        return 0;
 }
@@ -8337,8 +8309,7 @@ qeth_del_rxip(struct qeth_card *card, enum qeth_prot_versions proto,
                return;
        if (!qeth_delete_ip(card, ipaddr))
                kfree(ipaddr);
-        if (qeth_set_thread_start_bit(card, QETH_SET_IP_THREAD) == 0)
+        qeth_set_ip_addr_list(card);
-                schedule_work(&card->kernel_thread_starter);
 }
 /**
@@ -8380,8 +8351,7 @@ qeth_ip_event(struct notifier_block *this,
        default:
                break;
        }
-        if (qeth_set_thread_start_bit(card, QETH_SET_IP_THREAD) == 0)
+        qeth_set_ip_addr_list(card);
-                schedule_work(&card->kernel_thread_starter);
 out:
        return NOTIFY_DONE;
 }
@@ -8433,8 +8403,7 @@ qeth_ip6_event(struct notifier_block *this,
        default:
                break;
        }
-        if (qeth_set_thread_start_bit(card, QETH_SET_IP_THREAD) == 0)
+        qeth_set_ip_addr_list(card);
-                schedule_work(&card->kernel_thread_starter);
 out:
        return NOTIFY_DONE;
 }
diff --git a/drivers/serial/mpc52xx_uart.c b/drivers/serial/mpc52xx_uart.c
index 9d11a75663e6..3c4b6c243712 100644
--- a/drivers/serial/mpc52xx_uart.c
+++ b/drivers/serial/mpc52xx_uart.c
@@ -789,7 +789,9 @@ static struct console mpc52xx_console = {
 static int __init
 mpc52xx_console_init(void)
 {
+#if defined(CONFIG_PPC_MERGE)
        mpc52xx_uart_of_enumerate();
+#endif
        register_console(&mpc52xx_console);
        return 0;
 }
diff --git a/drivers/usb/class/usblp.c b/drivers/usb/class/usblp.c
index 24ee8be359f5..6377db1b446d 100644
--- a/drivers/usb/class/usblp.c
+++ b/drivers/usb/class/usblp.c
@@ -217,6 +217,7 @@ static const struct quirk_printer_struct quirk_printers[] = {
        { 0x0409, 0xbef4, USBLP_QUIRK_BIDIR }, /* NEC Picty760 (HP OEM) */
        { 0x0409, 0xf0be, USBLP_QUIRK_BIDIR }, /* NEC Picty920 (HP OEM) */
        { 0x0409, 0xf1be, USBLP_QUIRK_BIDIR }, /* NEC Picty800 (HP OEM) */
+        { 0x0482, 0x0010, USBLP_QUIRK_BIDIR }, /* Kyocera Mita FS 820, by zut <kernel@zut.de> */
        { 0, 0 }
 };
diff --git a/drivers/usb/core/endpoint.c b/drivers/usb/core/endpoint.c
index c505b767cee1..5e628ae3aec7 100644
--- a/drivers/usb/core/endpoint.c
+++ b/drivers/usb/core/endpoint.c
@@ -268,6 +268,7 @@ static void ep_device_release(struct device *dev)
        struct ep_device *ep_dev = to_ep_device(dev);
        dev_dbg(dev, "%s called for %s\n", __FUNCTION__, dev->bus_id);
+        endpoint_free_minor(ep_dev);
        kfree(ep_dev);
 }
@@ -349,7 +350,6 @@ void usb_remove_ep_files(struct usb_host_endpoint *endpoint)
                sprintf(name, "ep_%02x", endpoint->desc.bEndpointAddress);
                sysfs_remove_link(&ep_dev->dev.parent->kobj, name);
                sysfs_remove_group(&ep_dev->dev.kobj, &ep_dev_attr_grp);
-                endpoint_free_minor(ep_dev);
                device_unregister(&ep_dev->dev);
                endpoint->ep_dev = NULL;
                destroy_endpoint_class();
diff --git a/drivers/usb/gadget/omap_udc.c b/drivers/usb/gadget/omap_udc.c
index 15d77c307930..cdcfd42843d4 100644
--- a/drivers/usb/gadget/omap_udc.c
+++ b/drivers/usb/gadget/omap_udc.c
@@ -42,6 +42,7 @@
 #include <linux/usb_gadget.h>
 #include <linux/usb/otg.h>
 #include <linux/dma-mapping.h>
+#include <linux/clk.h>
 #include <asm/byteorder.h>
 #include <asm/io.h>
@@ -60,6 +61,11 @@
 /* bulk DMA seems to be behaving for both IN and OUT */
 #define USE_DMA
+/* FIXME: OMAP2 currently has some problem in DMA mode */
+#ifdef CONFIG_ARCH_OMAP2
+#undef USE_DMA
+#endif
 /* ISO too */
 #define USE_ISO
@@ -99,7 +105,7 @@ static unsigned fifo_mode = 0;
 * boot parameter "omap_udc:fifo_mode=42"
 */
 module_param (fifo_mode, uint, 0);
-MODULE_PARM_DESC (fifo_mode, "endpoint setup (0 == default)");
+MODULE_PARM_DESC (fifo_mode, "endpoint configuration");
 #ifdef  USE_DMA
 static unsigned use_dma = 1;
@@ -122,7 +128,7 @@ static const char driver_desc [] = DRIVER_DESC;
 /*-------------------------------------------------------------------------*/
 /* there's a notion of "current endpoint" for modifying endpoint
- * state, and PIO access to its FIFO.  
+ * state, and PIO access to its FIFO.
 */
 static void use_ep(struct omap_ep *ep, u16 select)
@@ -391,7 +397,7 @@ done(struct omap_ep *ep, struct omap_req *req, int status)
 #define FIFO_EMPTY      (UDC_NON_ISO_FIFO_EMPTY | UDC_ISO_FIFO_EMPTY)
 #define FIFO_UNREADABLE (UDC_EP_HALTED | FIFO_EMPTY)
-static inline int 
+static inline int
 write_packet(u8 *buf, struct omap_req *req, unsigned max)
 {
        unsigned        len;
@@ -456,7 +462,7 @@ static int write_fifo(struct omap_ep *ep, struct omap_req *req)
        return is_last;
 }
-static inline int 
+static inline int
 read_packet(u8 *buf, struct omap_req *req, unsigned avail)
 {
        unsigned        len;
@@ -542,9 +548,9 @@ static inline dma_addr_t dma_csac(unsigned lch)
        /* omap 3.2/3.3 erratum: sometimes 0 is returned if CSAC/CDAC is
         * read before the DMA controller finished disabling the channel.
         */
-        csac = omap_readw(OMAP_DMA_CSAC(lch));
+        csac = OMAP_DMA_CSAC_REG(lch);
        if (csac == 0)
-                csac = omap_readw(OMAP_DMA_CSAC(lch));
+                csac = OMAP_DMA_CSAC_REG(lch);
        return csac;
 }
@@ -555,9 +561,9 @@ static inline dma_addr_t dma_cdac(unsigned lch)
        /* omap 3.2/3.3 erratum: sometimes 0 is returned if CSAC/CDAC is
         * read before the DMA controller finished disabling the channel.
         */
-        cdac = omap_readw(OMAP_DMA_CDAC(lch));
+        cdac = OMAP_DMA_CDAC_REG(lch);
        if (cdac == 0)
-                cdac = omap_readw(OMAP_DMA_CDAC(lch));
+                cdac = OMAP_DMA_CDAC_REG(lch);
        return cdac;
 }
@@ -582,7 +588,7 @@ static u16 dma_src_len(struct omap_ep *ep, dma_addr_t start)
 }
 #define DMA_DEST_LAST(x) (cpu_is_omap15xx() \
-                ? omap_readw(OMAP_DMA_CSAC(x)) /* really: CPC */ \
+                ? OMAP_DMA_CSAC_REG(x) /* really: CPC */ \
                : dma_cdac(x))
 static u16 dma_dest_len(struct omap_ep *ep, dma_addr_t start)
@@ -620,17 +626,19 @@ static void next_in_dma(struct omap_ep *ep, struct omap_req *req)
                        || (cpu_is_omap15xx() && length < ep->maxpacket)) {
                txdma_ctrl = UDC_TXN_EOT | length;
                omap_set_dma_transfer_params(ep->lch, OMAP_DMA_DATA_TYPE_S8,
-                                length, 1, sync_mode);
+                                length, 1, sync_mode, 0, 0);
        } else {
                length = min(length / ep->maxpacket,
                                (unsigned) UDC_TXN_TSC + 1);
-                txdma_ctrl = length;
+                txdma_ctrl = length;
                omap_set_dma_transfer_params(ep->lch, OMAP_DMA_DATA_TYPE_S16,
-                                ep->ep.maxpacket >> 1, length, sync_mode);
+                                ep->ep.maxpacket >> 1, length, sync_mode,
+                                0, 0);
                length *= ep->maxpacket;
        }
        omap_set_dma_src_params(ep->lch, OMAP_DMA_PORT_EMIFF,
-                OMAP_DMA_AMODE_POST_INC, req->req.dma + req->req.actual);
+                OMAP_DMA_AMODE_POST_INC, req->req.dma + req->req.actual,
+                0, 0);
        omap_start_dma(ep->lch);
        ep->dma_counter = dma_csac(ep->lch);
@@ -675,9 +683,11 @@ static void next_out_dma(struct omap_ep *ep, struct omap_req *req)
        req->dma_bytes = packets * ep->ep.maxpacket;
        omap_set_dma_transfer_params(ep->lch, OMAP_DMA_DATA_TYPE_S16,
                        ep->ep.maxpacket >> 1, packets,
-                        OMAP_DMA_SYNC_ELEMENT);
+                        OMAP_DMA_SYNC_ELEMENT,
+                        0, 0);
        omap_set_dma_dest_params(ep->lch, OMAP_DMA_PORT_EMIFF,
-                OMAP_DMA_AMODE_POST_INC, req->req.dma + req->req.actual);
+                OMAP_DMA_AMODE_POST_INC, req->req.dma + req->req.actual,
+                0, 0);
        ep->dma_counter = DMA_DEST_LAST(ep->lch);
        UDC_RXDMA_REG(ep->dma_channel) = UDC_RXN_STOP | (packets - 1);
@@ -820,7 +830,8 @@ static void dma_channel_claim(struct omap_ep *ep, unsigned channel)
                        omap_set_dma_dest_params(ep->lch,
                                OMAP_DMA_PORT_TIPB,
                                OMAP_DMA_AMODE_CONSTANT,
-                                (unsigned long) io_v2p((u32)&UDC_DATA_DMA_REG));
+                                (unsigned long) io_v2p((u32)&UDC_DATA_DMA_REG),
+                                0, 0);
                }
        } else {
                status = omap_request_dma(OMAP_DMA_USB_W2FC_RX0 - 1 + channel,
@@ -831,7 +842,8 @@ static void dma_channel_claim(struct omap_ep *ep, unsigned channel)
                        omap_set_dma_src_params(ep->lch,
                                OMAP_DMA_PORT_TIPB,
                                OMAP_DMA_AMODE_CONSTANT,
-                                (unsigned long) io_v2p((u32)&UDC_DATA_DMA_REG));
+                                (unsigned long) io_v2p((u32)&UDC_DATA_DMA_REG),
+                                0, 0);
                        /* EMIFF */
                        omap_set_dma_dest_burst_mode(ep->lch,
                                                OMAP_DMA_DATA_BURST_4);
@@ -846,7 +858,7 @@ static void dma_channel_claim(struct omap_ep *ep, unsigned channel)
                /* channel type P: hw synch (fifo) */
                if (!cpu_is_omap15xx())
-                        omap_writew(2, OMAP_DMA_LCH_CTRL(ep->lch));
+                        OMAP1_DMA_LCH_CTRL_REG(ep->lch) = 2;
        }
 just_restart:
@@ -893,7 +905,7 @@ static void dma_channel_release(struct omap_ep *ep)
        else
                req = NULL;
-        active = ((1 << 7) & omap_readl(OMAP_DMA_CCR(ep->lch))) != 0;
+        active = ((1 << 7) & OMAP_DMA_CCR_REG(ep->lch)) != 0;
        DBG("%s release %s %cxdma%d %p\n", ep->ep.name,
                        active ? "active" : "idle",
@@ -1117,7 +1129,7 @@ static int omap_ep_dequeue(struct usb_ep *_ep, struct usb_request *_req)
                 */
                dma_channel_release(ep);
                dma_channel_claim(ep, channel);
-        } else 
+        } else
                done(ep, req, -ECONNRESET);
        spin_unlock_irqrestore(&ep->udc->lock, flags);
        return 0;
@@ -1153,7 +1165,7 @@ static int omap_ep_set_halt(struct usb_ep *_ep, int value)
                /* IN endpoints must already be idle */
                if ((ep->bEndpointAddress & USB_DIR_IN)
-                                && !list_empty(&ep->queue)) { 
+                                && !list_empty(&ep->queue)) {
                        status = -EAGAIN;
                        goto done;
                }
@@ -1298,6 +1310,23 @@ static void pullup_disable(struct omap_udc *udc)
        UDC_SYSCON1_REG &= ~UDC_PULLUP_EN;
 }
+static struct omap_udc *udc;
+static void omap_udc_enable_clock(int enable)
+{
+        if (udc == NULL || udc->dc_clk == NULL || udc->hhc_clk == NULL)
+                return;
+        if (enable) {
+                clk_enable(udc->dc_clk);
+                clk_enable(udc->hhc_clk);
+                udelay(100);
+        } else {
+                clk_disable(udc->hhc_clk);
+                clk_disable(udc->dc_clk);
+        }
+}
 /*
 * Called by whatever detects VBUS sessions:  external transceiver
 * driver, or maybe GPIO0 VBUS IRQ.  May request 48 MHz clock.
@@ -1318,10 +1347,22 @@ static int omap_vbus_session(struct usb_gadget *gadget, int is_active)
                else
                        FUNC_MUX_CTRL_0_REG &= ~VBUS_CTRL_1510;
        }
+        if (udc->dc_clk != NULL && is_active) {
+                if (!udc->clk_requested) {
+                        omap_udc_enable_clock(1);
+                        udc->clk_requested = 1;
+                }
+        }
        if (can_pullup(udc))
                pullup_enable(udc);
        else
                pullup_disable(udc);
+        if (udc->dc_clk != NULL && !is_active) {
+                if (udc->clk_requested) {
+                        omap_udc_enable_clock(0);
+                        udc->clk_requested = 0;
+                }
+        }
        spin_unlock_irqrestore(&udc->lock, flags);
        return 0;
 }
@@ -1441,7 +1482,7 @@ static void ep0_irq(struct omap_udc *udc, u16 irq_src)
                }
        }
-        /* IN/OUT packets mean we're in the DATA or STATUS stage.  
+        /* IN/OUT packets mean we're in the DATA or STATUS stage.
         * This driver uses only uses protocol stalls (ep0 never halts),
         * and if we got this far the gadget driver already had a
         * chance to stall.  Tries to be forgiving of host oddities.
@@ -1509,7 +1550,7 @@ static void ep0_irq(struct omap_udc *udc, u16 irq_src)
                                } else if (stat == 0)
                                        UDC_CTRL_REG = UDC_SET_FIFO_EN;
                                UDC_EP_NUM_REG = 0;
-                                
                                /* activate status stage */
                                if (stat == 1) {
                                        done(ep0, req, 0);
@@ -1866,7 +1907,7 @@ static void pio_out_timer(unsigned long _ep)
        spin_lock_irqsave(&ep->udc->lock, flags);
        if (!list_empty(&ep->queue) && ep->ackwait) {
-                use_ep(ep, 0);
+                use_ep(ep, UDC_EP_SEL);
                stat_flg = UDC_STAT_FLG_REG;
                if ((stat_flg & UDC_ACK) && (!(stat_flg & UDC_FIFO_EN)
@@ -1876,12 +1917,12 @@ static void pio_out_timer(unsigned long _ep)
                        VDBG("%s: lose, %04x\n", ep->ep.name, stat_flg);
                        req = container_of(ep->queue.next,
                                        struct omap_req, queue);
-                        UDC_EP_NUM_REG = ep->bEndpointAddress | UDC_EP_SEL;
                        (void) read_fifo(ep, req);
                        UDC_EP_NUM_REG = ep->bEndpointAddress;
                        UDC_CTRL_REG = UDC_SET_FIFO_EN;
                        ep->ackwait = 1 + ep->double_buf;
-                }
+                } else
+                        deselect_ep();
        }
        mod_timer(&ep->timer, PIO_OUT_TIMEOUT);
        spin_unlock_irqrestore(&ep->udc->lock, flags);
@@ -2028,7 +2069,17 @@ static irqreturn_t omap_udc_iso_irq(int irq, void *_dev)
 /*-------------------------------------------------------------------------*/
-static struct omap_udc *udc;
+static inline int machine_needs_vbus_session(void)
+{
+        return (machine_is_omap_innovator()
+                || machine_is_omap_osk()
+                || machine_is_omap_apollon()
+#ifndef CONFIG_MACH_OMAP_H4_OTG
+                || machine_is_omap_h4()
+#endif
+                || machine_is_sx1()
+                );
+}
 int usb_gadget_register_driver (struct usb_gadget_driver *driver)
 {
@@ -2070,6 +2121,9 @@ int usb_gadget_register_driver (struct usb_gadget_driver *driver)
        udc->gadget.dev.driver = &driver->driver;
        spin_unlock_irqrestore(&udc->lock, flags);
+        if (udc->dc_clk != NULL)
+                omap_udc_enable_clock(1);
        status = driver->bind (&udc->gadget);
        if (status) {
                DBG("bind to %s --> %d\n", driver->driver.name, status);
@@ -2103,10 +2157,12 @@ int usb_gadget_register_driver (struct usb_gadget_driver *driver)
        /* boards that don't have VBUS sensing can't autogate 48MHz;
         * can't enter deep sleep while a gadget driver is active.
         */
-        if (machine_is_omap_innovator() || machine_is_omap_osk())
+        if (machine_needs_vbus_session())
                omap_vbus_session(&udc->gadget, 1);
 done:
+        if (udc->dc_clk != NULL)
+                omap_udc_enable_clock(0);
        return status;
 }
 EXPORT_SYMBOL(usb_gadget_register_driver);
@@ -2121,7 +2177,10 @@ int usb_gadget_unregister_driver (struct usb_gadget_driver *driver)
        if (!driver || driver != udc->driver || !driver->unbind)
                return -EINVAL;
-        if (machine_is_omap_innovator() || machine_is_omap_osk())
+        if (udc->dc_clk != NULL)
+                omap_udc_enable_clock(1);
+        if (machine_needs_vbus_session())
                omap_vbus_session(&udc->gadget, 0);
        if (udc->transceiver)
@@ -2137,6 +2196,8 @@ int usb_gadget_unregister_driver (struct usb_gadget_driver *driver)
        udc->gadget.dev.driver = NULL;
        udc->driver = NULL;
+        if (udc->dc_clk != NULL)
+                omap_udc_enable_clock(0);
        DBG("unregistered driver '%s'\n", driver->driver.name);
        return status;
 }
@@ -2219,7 +2280,7 @@ static char *trx_mode(unsigned m, int enabled)
        case 0:         return enabled ? "*6wire" : "unused";
        case 1:         return "4wire";
        case 2:         return "3wire";
-        case 3:         return "6wire";
+        case 3:         return "6wire";
        default:        return "unknown";
        }
 }
@@ -2228,11 +2289,18 @@ static int proc_otg_show(struct seq_file *s)
 {
        u32             tmp;
        u32             trans;
+        char            *ctrl_name;
        tmp = OTG_REV_REG;
-        trans = USB_TRANSCEIVER_CTRL_REG;
+        if (cpu_is_omap24xx()) {
-        seq_printf(s, "\nOTG rev %d.%d, transceiver_ctrl %05x\n",
+                ctrl_name = "control_devconf";
-                tmp >> 4, tmp & 0xf, trans);
+                trans = CONTROL_DEVCONF_REG;
+        } else {
+                ctrl_name = "tranceiver_ctrl";
+                trans = USB_TRANSCEIVER_CTRL_REG;
+        }
+        seq_printf(s, "\nOTG rev %d.%d, %s %05x\n",
+                tmp >> 4, tmp & 0xf, ctrl_name, trans);
        tmp = OTG_SYSCON_1_REG;
        seq_printf(s, "otg_syscon1 %08x usb2 %s, usb1 %s, usb0 %s,"
                        FOURBITS "\n", tmp,
@@ -2307,7 +2375,7 @@ static int proc_udc_show(struct seq_file *s, void *_)
                driver_desc,
                use_dma ?  " (dma)" : "");
-        tmp = UDC_REV_REG & 0xff; 
+        tmp = UDC_REV_REG & 0xff;
        seq_printf(s,
                "UDC rev %d.%d, fifo mode %d, gadget %s\n"
                "hmc %d, transceiver %s\n",
@@ -2315,11 +2383,16 @@ static int proc_udc_show(struct seq_file *s, void *_)
                fifo_mode,
                udc->driver ? udc->driver->driver.name : "(none)",
                HMC,
-                udc->transceiver ? udc->transceiver->label : "(none)");
+                udc->transceiver
-        seq_printf(s, "ULPD control %04x req %04x status %04x\n",
+                        ? udc->transceiver->label
-                __REG16(ULPD_CLOCK_CTRL),
+                        : ((cpu_is_omap1710() || cpu_is_omap24xx())
-                __REG16(ULPD_SOFT_REQ),
+                                ? "external" : "(none)"));
-                __REG16(ULPD_STATUS_REQ));
+        if (cpu_class_is_omap1()) {
+                seq_printf(s, "ULPD control %04x req %04x status %04x\n",
+                        __REG16(ULPD_CLOCK_CTRL),
+                        __REG16(ULPD_SOFT_REQ),
+                        __REG16(ULPD_STATUS_REQ));
+        }
        /* OTG controller registers */
        if (!cpu_is_omap15xx())
@@ -2504,9 +2577,10 @@ omap_ep_setup(char *name, u8 addr, u8 type,
                dbuf = 1;
        } else {
                /* double-buffering "not supported" on 15xx,
-                 * and ignored for PIO-IN on 16xx
+                 * and ignored for PIO-IN on newer chips
+                 * (for more reliable behavior)
                 */
-                if (!use_dma || cpu_is_omap15xx())
+                if (!use_dma || cpu_is_omap15xx() || cpu_is_omap24xx())
                        dbuf = 0;
                switch (maxp) {
@@ -2549,7 +2623,7 @@ omap_ep_setup(char *name, u8 addr, u8 type,
        ep->bEndpointAddress = addr;
        ep->bmAttributes = type;
        ep->double_buf = dbuf;
-        ep->udc = udc; 
+        ep->udc = udc;
        ep->ep.name = ep->name;
        ep->ep.ops = &omap_ep_ops;
@@ -2709,15 +2783,37 @@ static int __init omap_udc_probe(struct platform_device *pdev)
        struct otg_transceiver  *xceiv = NULL;
        const char              *type = NULL;
        struct omap_usb_config  *config = pdev->dev.platform_data;
+        struct clk              *dc_clk;
+        struct clk              *hhc_clk;
        /* NOTE:  "knows" the order of the resources! */
-        if (!request_mem_region(pdev->resource[0].start, 
+        if (!request_mem_region(pdev->resource[0].start,
                        pdev->resource[0].end - pdev->resource[0].start + 1,
                        driver_name)) {
                DBG("request_mem_region failed\n");
                return -EBUSY;
        }
+        if (cpu_is_omap16xx()) {
+                dc_clk = clk_get(&pdev->dev, "usb_dc_ck");
+                hhc_clk = clk_get(&pdev->dev, "usb_hhc_ck");
+                BUG_ON(IS_ERR(dc_clk) || IS_ERR(hhc_clk));
+                /* can't use omap_udc_enable_clock yet */
+                clk_enable(dc_clk);
+                clk_enable(hhc_clk);
+                udelay(100);
+        }
+        if (cpu_is_omap24xx()) {
+                dc_clk = clk_get(&pdev->dev, "usb_fck");
+                hhc_clk = clk_get(&pdev->dev, "usb_l4_ick");
+                BUG_ON(IS_ERR(dc_clk) || IS_ERR(hhc_clk));
+                /* can't use omap_udc_enable_clock yet */
+                clk_enable(dc_clk);
+                clk_enable(hhc_clk);
+                udelay(100);
+        }
        INFO("OMAP UDC rev %d.%d%s\n",
                UDC_REV_REG >> 4, UDC_REV_REG & 0xf,
                config->otg ? ", Mini-AB" : "");
@@ -2727,7 +2823,7 @@ static int __init omap_udc_probe(struct platform_device *pdev)
                hmc = HMC_1510;
                type = "(unknown)";
-                if (machine_is_omap_innovator()) {
+                if (machine_is_omap_innovator() || machine_is_sx1()) {
                        /* just set up software VBUS detect, and then
                         * later rig it so we always report VBUS.
                         * FIXME without really sensing VBUS, we can't
@@ -2756,6 +2852,15 @@ static int __init omap_udc_probe(struct platform_device *pdev)
                }
                hmc = HMC_1610;
+                if (cpu_is_omap24xx()) {
+                        /* this could be transceiverless in one of the
+                         * "we don't need to know" modes.
+                         */
+                        type = "external";
+                        goto known;
+                }
                switch (hmc) {
                case 0:                 /* POWERUP DEFAULT == 0 */
                case 4:
@@ -2794,6 +2899,7 @@ bad_on_1710:
                        goto cleanup0;
                }
        }
+known:
        INFO("hmc mode %d, %s transceiver\n", hmc, type);
        /* a "gadget" abstracts/virtualizes the controller */
@@ -2818,8 +2924,8 @@ bad_on_1710:
        status = request_irq(pdev->resource[1].start, omap_udc_irq,
                        IRQF_SAMPLE_RANDOM, driver_name, udc);
        if (status != 0) {
-                ERR( "can't get irq %ld, err %d\n",
+                ERR("can't get irq %d, err %d\n",
-                        pdev->resource[1].start, status);
+                        (int) pdev->resource[1].start, status);
                goto cleanup1;
        }
@@ -2827,24 +2933,41 @@ bad_on_1710:
        status = request_irq(pdev->resource[2].start, omap_udc_pio_irq,
                        IRQF_SAMPLE_RANDOM, "omap_udc pio", udc);
        if (status != 0) {
-                ERR( "can't get irq %ld, err %d\n",
+                ERR("can't get irq %d, err %d\n",
-                        pdev->resource[2].start, status);
+                        (int) pdev->resource[2].start, status);
                goto cleanup2;
        }
 #ifdef  USE_ISO
        status = request_irq(pdev->resource[3].start, omap_udc_iso_irq,
                        IRQF_DISABLED, "omap_udc iso", udc);
        if (status != 0) {
-                ERR("can't get irq %ld, err %d\n",
+                ERR("can't get irq %d, err %d\n",
-                        pdev->resource[3].start, status);
+                        (int) pdev->resource[3].start, status);
                goto cleanup3;
        }
 #endif
+        if (cpu_is_omap16xx()) {
+                udc->dc_clk = dc_clk;
+                udc->hhc_clk = hhc_clk;
+                clk_disable(hhc_clk);
+                clk_disable(dc_clk);
+        }
+        if (cpu_is_omap24xx()) {
+                udc->dc_clk = dc_clk;
+                udc->hhc_clk = hhc_clk;
+                /* FIXME OMAP2 don't release hhc & dc clock */
+#if 0
+                clk_disable(hhc_clk);
+                clk_disable(dc_clk);
+#endif
+        }
        create_proc_file();
-        device_add(&udc->gadget.dev);
+        status = device_add(&udc->gadget.dev);
-        return 0;
+        if (!status)
+                return status;
+        /* If fail, fall through */
 #ifdef  USE_ISO
 cleanup3:
        free_irq(pdev->resource[2].start, udc);
@@ -2860,8 +2983,17 @@ cleanup1:
 cleanup0:
        if (xceiv)
                put_device(xceiv->dev);
+        if (cpu_is_omap16xx() || cpu_is_omap24xx()) {
+                clk_disable(hhc_clk);
+                clk_disable(dc_clk);
+                clk_put(hhc_clk);
+                clk_put(dc_clk);
+        }
        release_mem_region(pdev->resource[0].start,
                        pdev->resource[0].end - pdev->resource[0].start + 1);
        return status;
 }
@@ -2891,6 +3023,13 @@ static int __exit omap_udc_remove(struct platform_device *pdev)
        free_irq(pdev->resource[2].start, udc);
        free_irq(pdev->resource[1].start, udc);
+        if (udc->dc_clk) {
+                if (udc->clk_requested)
+                        omap_udc_enable_clock(0);
+                clk_put(udc->hhc_clk);
+                clk_put(udc->dc_clk);
+        }
        release_mem_region(pdev->resource[0].start,
                        pdev->resource[0].end - pdev->resource[0].start + 1);
diff --git a/drivers/usb/gadget/omap_udc.h b/drivers/usb/gadget/omap_udc.h
index 652ee4627344..1dc398bb9ab2 100644
--- a/drivers/usb/gadget/omap_udc.h
+++ b/drivers/usb/gadget/omap_udc.h
@@ -175,6 +175,9 @@ struct omap_udc {
        unsigned                        ep0_reset_config:1;
        unsigned                        ep0_setup:1;
        struct completion               *done;
+        struct clk                      *dc_clk;
+        struct clk                      *hhc_clk;
+        unsigned                        clk_requested:1;
 };
 /*-------------------------------------------------------------------------*/
diff --git a/drivers/usb/host/uhci-hcd.c b/drivers/usb/host/uhci-hcd.c
index acd101caeeeb..e0d4c2358b39 100644
--- a/drivers/usb/host/uhci-hcd.c
+++ b/drivers/usb/host/uhci-hcd.c
@@ -209,24 +209,16 @@ static int resume_detect_interrupts_are_broken(struct uhci_hcd *uhci)
 static int remote_wakeup_is_broken(struct uhci_hcd *uhci)
 {
-        static struct dmi_system_id broken_wakeup_table[] = {
-                {
-                        .ident = "Asus A7V8X",
-                        .matches = {
-                                DMI_MATCH(DMI_BOARD_VENDOR, "ASUSTeK"),
-                                DMI_MATCH(DMI_BOARD_NAME, "A7V8X"),
-                                DMI_MATCH(DMI_BOARD_VERSION, "REV 1.xx"),
-                        }
-                },
-                { }
-        };
        int port;
+        char *sys_info;
+        static char bad_Asus_board[] = "A7V8X";
        /* One of Asus's motherboards has a bug which causes it to
         * wake up immediately from suspend-to-RAM if any of the ports
         * are connected.  In such cases we will not set EGSM.
         */
-        if (dmi_check_system(broken_wakeup_table)) {
+        sys_info = dmi_get_system_info(DMI_BOARD_NAME);
+        if (sys_info && !strcmp(sys_info, bad_Asus_board)) {
                for (port = 0; port < uhci->rh_numports; ++port) {
                        if (inw(uhci->io_addr + USBPORTSC1 + port * 2) &
                                        USBPORTSC_CCS)
@@ -265,7 +257,9 @@ __acquires(uhci->lock)
        int_enable = USBINTR_RESUME;
        if (remote_wakeup_is_broken(uhci))
                egsm_enable = 0;
-        if (resume_detect_interrupts_are_broken(uhci) || !egsm_enable)
+        if (resume_detect_interrupts_are_broken(uhci) || !egsm_enable ||
+                        !device_may_wakeup(
+                                &uhci_to_hcd(uhci)->self.root_hub->dev))
                uhci->working_RD = int_enable = 0;
        outw(int_enable, uhci->io_addr + USBINTR);
diff --git a/drivers/usb/input/Kconfig b/drivers/usb/input/Kconfig
index f877cd4f317a..258a5d09d3dc 100644
--- a/drivers/usb/input/Kconfig
+++ b/drivers/usb/input/Kconfig
@@ -12,10 +12,8 @@ config USB_HID
        ---help---
          Say Y here if you want full HID support to connect USB keyboards,
          mice, joysticks, graphic tablets, or any other HID based devices
-          to your computer via USB. You also need to select HID Input layer
+          to your computer via USB, as well as Uninterruptible Power Supply
-          support (below) if you want to use keyboards, mice, joysticks and
+          (UPS) and monitor control devices.
-          the like ... as well as Uninterruptible Power Supply (UPS) and
-          monitor control devices.
          You can't use this driver and the HIDBP (Boot Protocol) keyboard
          and mouse drivers at the same time. More information is available:
diff --git a/drivers/usb/misc/sisusbvga/sisusb_con.c b/drivers/usb/misc/sisusbvga/sisusb_con.c
index bf26c3c56990..9148694627d5 100644
--- a/drivers/usb/misc/sisusbvga/sisusb_con.c
+++ b/drivers/usb/misc/sisusbvga/sisusb_con.c
@@ -403,7 +403,7 @@ sisusbcon_putc(struct vc_data *c, int ch, int y, int x)
        sisusb_copy_memory(sisusb, (char *)SISUSB_VADDR(x, y),
-                                (u32)SISUSB_HADDR(x, y), 2, &written);
+                                (long)SISUSB_HADDR(x, y), 2, &written);
        mutex_unlock(&sisusb->lock);
 }
@@ -438,7 +438,7 @@ sisusbcon_putcs(struct vc_data *c, const unsigned short *s,
        }
        sisusb_copy_memory(sisusb, (char *)SISUSB_VADDR(x, y),
-                                (u32)SISUSB_HADDR(x, y), count * 2, &written);
+                                (long)SISUSB_HADDR(x, y), count * 2, &written);
        mutex_unlock(&sisusb->lock);
 }
@@ -492,7 +492,7 @@ sisusbcon_clear(struct vc_data *c, int y, int x, int height, int width)
        sisusb_copy_memory(sisusb, (unsigned char *)SISUSB_VADDR(x, y),
-                                (u32)SISUSB_HADDR(x, y), length, &written);
+                                (long)SISUSB_HADDR(x, y), length, &written);
        mutex_unlock(&sisusb->lock);
 }
@@ -564,7 +564,7 @@ sisusbcon_bmove(struct vc_data *c, int sy, int sx,
        sisusb_copy_memory(sisusb, (unsigned char *)SISUSB_VADDR(dx, dy),
-                                (u32)SISUSB_HADDR(dx, dy), length, &written);
+                                (long)SISUSB_HADDR(dx, dy), length, &written);
        mutex_unlock(&sisusb->lock);
 }
@@ -612,7 +612,7 @@ sisusbcon_switch(struct vc_data *c)
                                                                length);
        sisusb_copy_memory(sisusb, (unsigned char *)c->vc_origin,
-                                (u32)SISUSB_HADDR(0, 0),
+                                (long)SISUSB_HADDR(0, 0),
                                length, &written);
        mutex_unlock(&sisusb->lock);
@@ -939,7 +939,7 @@ sisusbcon_scroll_area(struct vc_data *c, struct sisusb_usb_data *sisusb,
        }
        sisusb_copy_memory(sisusb, (char *)SISUSB_VADDR(0, t),
-                                (u32)SISUSB_HADDR(0, t), length, &written);
+                                (long)SISUSB_HADDR(0, t), length, &written);
        mutex_unlock(&sisusb->lock);
diff --git a/drivers/usb/net/asix.c b/drivers/usb/net/asix.c
index 95e682e2c9d6..f538013965b0 100644
--- a/drivers/usb/net/asix.c
+++ b/drivers/usb/net/asix.c
@@ -920,7 +920,7 @@ static int ax88772_bind(struct usbnet *dev, struct usb_interface *intf)
                goto out2;
        if ((ret = asix_write_cmd(dev, AX_CMD_SW_PHY_SELECT,
-                                0x0000, 0, 0, buf)) < 0) {
+                                1, 0, 0, buf)) < 0) {
                dbg("Select PHY #1 failed: %d", ret);
                goto out2;
        }
diff --git a/drivers/usb/serial/Kconfig b/drivers/usb/serial/Kconfig
index 2f4d303ee36f..c8999ae58652 100644
--- a/drivers/usb/serial/Kconfig
+++ b/drivers/usb/serial/Kconfig
@@ -170,7 +170,7 @@ config USB_SERIAL_FTDI_SIO
 config USB_SERIAL_FUNSOFT
        tristate "USB Fundamental Software Dongle Driver"
-        depends on USB_SERIAL
+        depends on USB_SERIAL && !(SPARC || SPARC64)
        ---help---
          Say Y here if you want to use the Fundamental Software dongle.
diff --git a/drivers/usb/serial/option.c b/drivers/usb/serial/option.c
index 819266b7e2f8..5ca04e82ea19 100644
--- a/drivers/usb/serial/option.c
+++ b/drivers/usb/serial/option.c
@@ -625,6 +625,9 @@ static int option_send_setup(struct usb_serial_port *port)
        dbg("%s", __FUNCTION__);
+        if (port->number != 0)
+                return 0;
        portdata = usb_get_serial_port_data(port);
        if (port->tty) {
diff --git a/drivers/usb/storage/unusual_devs.h b/drivers/usb/storage/unusual_devs.h
index 5fe7ff441a09..cddef3efba0a 100644
--- a/drivers/usb/storage/unusual_devs.h
+++ b/drivers/usb/storage/unusual_devs.h
@@ -728,7 +728,7 @@ UNUSUAL_DEV( 0x05ac, 0x1204, 0x0000, 0x9999,
                "Apple",
                "iPod",
                US_SC_DEVICE, US_PR_DEVICE, NULL,
-                US_FL_FIX_CAPACITY ),
+                US_FL_FIX_CAPACITY | US_FL_NOT_LOCKABLE ),
 UNUSUAL_DEV( 0x05ac, 0x1205, 0x0000, 0x9999,
                "Apple",
@@ -1358,6 +1358,21 @@ UNUSUAL_DEV(  0x1370, 0x6828, 0x0110, 0x0110,
                US_SC_DEVICE, US_PR_DEVICE, NULL,
                US_FL_IGNORE_RESIDUE ),
+/* Reported by Francesco Foresti <frafore@tiscali.it> */
+UNUSUAL_DEV(  0x14cd, 0x6600, 0x0201, 0x0201,
+                "Super Top",
+                "IDE DEVICE",
+                US_SC_DEVICE, US_PR_DEVICE, NULL,
+                US_FL_IGNORE_RESIDUE ),
+/* Reported by Robert Schedel <r.schedel@yahoo.de>
+ * Note: this is a 'super top' device like the above 14cd/6600 device */
+UNUSUAL_DEV(  0x1652, 0x6600, 0x0201, 0x0201,
+                "Teac",
+                "HD-35PUK-B",
+                US_SC_DEVICE, US_PR_DEVICE, NULL,
+                US_FL_IGNORE_RESIDUE ),
 /* patch submitted by Davide Perini <perini.davide@dpsoftware.org>
 * and Renato Perini <rperini@email.it>
 */
diff --git a/drivers/video/backlight/corgi_bl.c b/drivers/video/backlight/corgi_bl.c
index 61587ca2cdbb..fde1d9518123 100644
--- a/drivers/video/backlight/corgi_bl.c
+++ b/drivers/video/backlight/corgi_bl.c
@@ -121,7 +121,7 @@ static int corgibl_probe(struct platform_device *pdev)
                machinfo->limit_mask = -1;
        corgi_backlight_device = backlight_device_register ("corgi-bl",
-                NULL, &corgibl_data);
+                &pdev->dev, NULL, &corgibl_data);
        if (IS_ERR (corgi_backlight_device))
                return PTR_ERR (corgi_backlight_device);
diff --git a/drivers/video/backlight/hp680_bl.c b/drivers/video/backlight/hp680_bl.c
index 1c569fb543ae..c07d8207fb54 100644
--- a/drivers/video/backlight/hp680_bl.c
+++ b/drivers/video/backlight/hp680_bl.c
@@ -105,7 +105,7 @@ static struct backlight_properties hp680bl_data = {
 static int __init hp680bl_probe(struct platform_device *dev)
 {
        hp680_backlight_device = backlight_device_register ("hp680-bl",
-                NULL, &hp680bl_data);
+                &dev->dev, NULL, &hp680bl_data);
        if (IS_ERR (hp680_backlight_device))
                return PTR_ERR (hp680_backlight_device);
diff --git a/drivers/video/backlight/locomolcd.c b/drivers/video/backlight/locomolcd.c
index 2d7905410b2a..fc812d96c31d 100644
--- a/drivers/video/backlight/locomolcd.c
+++ b/drivers/video/backlight/locomolcd.c
@@ -184,7 +184,7 @@ static int locomolcd_probe(struct locomo_dev *ldev)
        local_irq_restore(flags);
-        locomolcd_bl_device = backlight_device_register("locomo-bl", NULL, &locomobl_data);
+        locomolcd_bl_device = backlight_device_register("locomo-bl", &ldev->dev, NULL, &locomobl_data);
        if (IS_ERR (locomolcd_bl_device))
                return PTR_ERR (locomolcd_bl_device);