13 files changed, 713 insertions, 321 deletions

diff --git a/drivers/firewire/Kconfig b/drivers/firewire/Kconfig
index 25bdc2dd9ce1..fb4d391810b6 100644
--- a/drivers/firewire/Kconfig
+++ b/drivers/firewire/Kconfig
@@ -54,6 +54,11 @@ config FIREWIRE_OHCI
           directive, use "install modulename /bin/true" for the modules to be
           blacklisted.
 
+config FIREWIRE_OHCI_DEBUG
+        bool
+        depends on FIREWIRE_OHCI
+        default y
+
 config FIREWIRE_SBP2
         tristate "Support for storage devices (SBP-2 protocol driver)"
         depends on FIREWIRE && SCSI
diff --git a/drivers/firewire/fw-card.c b/drivers/firewire/fw-card.c
index a03462750b95..5b4c0d9f5173 100644
--- a/drivers/firewire/fw-card.c
+++ b/drivers/firewire/fw-card.c
@@ -167,7 +167,6 @@ fw_core_add_descriptor(struct fw_descriptor *desc)
 
         return 0;
 }
-EXPORT_SYMBOL(fw_core_add_descriptor);
 
 void
 fw_core_remove_descriptor(struct fw_descriptor *desc)
@@ -182,7 +181,6 @@ fw_core_remove_descriptor(struct fw_descriptor *desc)
 
         mutex_unlock(&card_mutex);
 }
-EXPORT_SYMBOL(fw_core_remove_descriptor);
 
 static const char gap_count_table[] = {
         63, 5, 7, 8, 10, 13, 16, 18, 21, 24, 26, 29, 32, 35, 37, 40
@@ -220,7 +218,7 @@ fw_card_bm_work(struct work_struct *work)
         struct bm_data bmd;
         unsigned long flags;
         int root_id, new_root_id, irm_id, gap_count, generation, grace;
-        int do_reset = 0;
+        bool do_reset = false;
 
         spin_lock_irqsave(&card->lock, flags);
         local_node = card->local_node;
@@ -331,7 +329,7 @@ fw_card_bm_work(struct work_struct *work)
                  */
                 spin_unlock_irqrestore(&card->lock, flags);
                 goto out;
-        } else if (root_device->config_rom[2] & BIB_CMC) {
+        } else if (root_device->cmc) {
                 /*
                  * FIXME: I suppose we should set the cmstr bit in the
                  * STATE_CLEAR register of this node, as described in
@@ -360,14 +358,14 @@ fw_card_bm_work(struct work_struct *work)
                 gap_count = 63;
 
         /*
-         * Finally, figure out if we should do a reset or not.  If we've
-         * done less that 5 resets with the same physical topology and we
+         * Finally, figure out if we should do a reset or not.  If we have
+         * done less than 5 resets with the same physical topology and we
          * have either a new root or a new gap count setting, let's do it.
          */
 
         if (card->bm_retries++ < 5 &&
             (card->gap_count != gap_count || new_root_id != root_id))
-                do_reset = 1;
+                do_reset = true;
 
         spin_unlock_irqrestore(&card->lock, flags);
 
@@ -398,7 +396,6 @@ fw_card_initialize(struct fw_card *card, const struct fw_card_driver *driver,
 {
         static atomic_t index = ATOMIC_INIT(-1);
 
-        kref_init(&card->kref);
         atomic_set(&card->device_count, 0);
         card->index = atomic_inc_return(&index);
         card->driver = driver;
@@ -429,12 +426,6 @@ fw_card_add(struct fw_card *card,
         card->link_speed = link_speed;
         card->guid = guid;
 
-        /*
-         * The subsystem grabs a reference when the card is added and
-         * drops it when the driver calls fw_core_remove_card.
-         */
-        fw_card_get(card);
-
         mutex_lock(&card_mutex);
         config_rom = generate_config_rom(card, &length);
         list_add_tail(&card->link, &card_list);
@@ -540,40 +531,9 @@ fw_core_remove_card(struct fw_card *card)
         cancel_delayed_work_sync(&card->work);
         fw_flush_transactions(card);
         del_timer_sync(&card->flush_timer);
-
-        fw_card_put(card);
 }
 EXPORT_SYMBOL(fw_core_remove_card);
 
-struct fw_card *
-fw_card_get(struct fw_card *card)
-{
-        kref_get(&card->kref);
-
-        return card;
-}
-EXPORT_SYMBOL(fw_card_get);
-
-static void
-release_card(struct kref *kref)
-{
-        struct fw_card *card = container_of(kref, struct fw_card, kref);
-
-        kfree(card);
-}
-
-/*
- * An assumption for fw_card_put() is that the card driver allocates
- * the fw_card struct with kalloc and that it has been shut down
- * before the last ref is dropped.
- */
-void
-fw_card_put(struct fw_card *card)
-{
-        kref_put(&card->kref, release_card);
-}
-EXPORT_SYMBOL(fw_card_put);
-
 int
 fw_core_initiate_bus_reset(struct fw_card *card, int short_reset)
 {
diff --git a/drivers/firewire/fw-cdev.c b/drivers/firewire/fw-cdev.c
index 46bc197a047f..4a541921a14a 100644
--- a/drivers/firewire/fw-cdev.c
+++ b/drivers/firewire/fw-cdev.c
@@ -269,21 +269,28 @@ static int ioctl_get_info(struct client *client, void *buffer)
 {
         struct fw_cdev_get_info *get_info = buffer;
         struct fw_cdev_event_bus_reset bus_reset;
+        unsigned long ret = 0;
 
         client->version = get_info->version;
         get_info->version = FW_CDEV_VERSION;
 
+        down_read(&fw_device_rwsem);
+
         if (get_info->rom != 0) {
                 void __user *uptr = u64_to_uptr(get_info->rom);
                 size_t want = get_info->rom_length;
                 size_t have = client->device->config_rom_length * 4;
 
-                if (copy_to_user(uptr, client->device->config_rom,
-                                 min(want, have)))
-                        return -EFAULT;
+                ret = copy_to_user(uptr, client->device->config_rom,
+                                   min(want, have));
         }
         get_info->rom_length = client->device->config_rom_length * 4;
 
+        up_read(&fw_device_rwsem);
+
+        if (ret != 0)
+                return -EFAULT;
+
         client->bus_reset_closure = get_info->bus_reset_closure;
         if (get_info->bus_reset != 0) {
                 void __user *uptr = u64_to_uptr(get_info->bus_reset);
diff --git a/drivers/firewire/fw-device.c b/drivers/firewire/fw-device.c
index 870125a3638e..2d01bc1b9752 100644
--- a/drivers/firewire/fw-device.c
+++ b/drivers/firewire/fw-device.c
@@ -25,7 +25,7 @@
 #include <linux/device.h>
 #include <linux/delay.h>
 #include <linux/idr.h>
-#include <linux/rwsem.h>
+#include <linux/string.h>
 #include <asm/semaphore.h>
 #include <asm/system.h>
 #include <linux/ctype.h>
@@ -160,9 +160,9 @@ static void fw_device_release(struct device *dev)
          * Take the card lock so we don't set this to NULL while a
          * FW_NODE_UPDATED callback is being handled.
          */
-        spin_lock_irqsave(&device->card->lock, flags);
+        spin_lock_irqsave(&card->lock, flags);
         device->node->data = NULL;
-        spin_unlock_irqrestore(&device->card->lock, flags);
+        spin_unlock_irqrestore(&card->lock, flags);
 
         fw_node_put(device->node);
         kfree(device->config_rom);
@@ -195,7 +195,9 @@ show_immediate(struct device *dev, struct device_attribute *dattr, char *buf)
                 container_of(dattr, struct config_rom_attribute, attr);
         struct fw_csr_iterator ci;
         u32 *dir;
-        int key, value;
+        int key, value, ret = -ENOENT;
+
+        down_read(&fw_device_rwsem);
 
         if (is_fw_unit(dev))
                 dir = fw_unit(dev)->directory;
@@ -204,11 +206,15 @@ show_immediate(struct device *dev, struct device_attribute *dattr, char *buf)
 
         fw_csr_iterator_init(&ci, dir);
         while (fw_csr_iterator_next(&ci, &key, &value))
-                if (attr->key == key)
-                        return snprintf(buf, buf ? PAGE_SIZE : 0,
-                                        "0x%06x\n", value);
+                if (attr->key == key) {
+                        ret = snprintf(buf, buf ? PAGE_SIZE : 0,
+                                       "0x%06x\n", value);
+                        break;
+                }
+
+        up_read(&fw_device_rwsem);
 
-        return -ENOENT;
+        return ret;
 }
 
 #define IMMEDIATE_ATTR(name, key)                               \
@@ -221,9 +227,11 @@ show_text_leaf(struct device *dev, struct device_attribute *dattr, char *buf)
                 container_of(dattr, struct config_rom_attribute, attr);
         struct fw_csr_iterator ci;
         u32 *dir, *block = NULL, *p, *end;
-        int length, key, value, last_key = 0;
+        int length, key, value, last_key = 0, ret = -ENOENT;
         char *b;
 
+        down_read(&fw_device_rwsem);
+
         if (is_fw_unit(dev))
                 dir = fw_unit(dev)->directory;
         else
@@ -238,18 +246,20 @@ show_text_leaf(struct device *dev, struct device_attribute *dattr, char *buf)
         }
 
         if (block == NULL)
-                return -ENOENT;
+                goto out;
 
         length = min(block[0] >> 16, 256U);
         if (length < 3)
-                return -ENOENT;
+                goto out;
 
         if (block[1] != 0 || block[2] != 0)
                 /* Unknown encoding. */
-                return -ENOENT;
+                goto out;
 
-        if (buf == NULL)
-                return length * 4;
+        if (buf == NULL) {
+                ret = length * 4;
+                goto out;
+        }
 
         b = buf;
         end = &block[length + 1];
@@ -259,8 +269,11 @@ show_text_leaf(struct device *dev, struct device_attribute *dattr, char *buf)
         /* Strip trailing whitespace and add newline. */
         while (b--, (isspace(*b) || *b == '\0') && b > buf);
         strcpy(b + 1, "\n");
+        ret = b + 2 - buf;
+ out:
+        up_read(&fw_device_rwsem);
 
-        return b + 2 - buf;
+        return ret;
 }
 
 #define TEXT_LEAF_ATTR(name, key)                               \
@@ -337,19 +350,28 @@ static ssize_t
 config_rom_show(struct device *dev, struct device_attribute *attr, char *buf)
 {
         struct fw_device *device = fw_device(dev);
+        size_t length;
 
-        memcpy(buf, device->config_rom, device->config_rom_length * 4);
+        down_read(&fw_device_rwsem);
+        length = device->config_rom_length * 4;
+        memcpy(buf, device->config_rom, length);
+        up_read(&fw_device_rwsem);
 
-        return device->config_rom_length * 4;
+        return length;
 }
 
 static ssize_t
 guid_show(struct device *dev, struct device_attribute *attr, char *buf)
 {
         struct fw_device *device = fw_device(dev);
+        int ret;
+
+        down_read(&fw_device_rwsem);
+        ret = snprintf(buf, PAGE_SIZE, "0x%08x%08x\n",
+                       device->config_rom[3], device->config_rom[4]);
+        up_read(&fw_device_rwsem);
 
-        return snprintf(buf, PAGE_SIZE, "0x%08x%08x\n",
-                        device->config_rom[3], device->config_rom[4]);
+        return ret;
 }
 
 static struct device_attribute fw_device_attributes[] = {
@@ -388,7 +410,7 @@ read_rom(struct fw_device *device, int generation, int index, u32 *data)
 
         init_completion(&callback_data.done);
 
-        offset = 0xfffff0000400ULL + index * 4;
+        offset = (CSR_REGISTER_BASE | CSR_CONFIG_ROM) + index * 4;
         fw_send_request(device->card, &t, TCODE_READ_QUADLET_REQUEST,
                         device->node_id, generation, device->max_speed,
                         offset, NULL, 4, complete_transaction, &callback_data);
@@ -400,6 +422,9 @@ read_rom(struct fw_device *device, int generation, int index, u32 *data)
         return callback_data.rcode;
 }
 
+#define READ_BIB_ROM_SIZE       256
+#define READ_BIB_STACK_SIZE     16
+
 /*
  * Read the bus info block, perform a speed probe, and read all of the rest of
  * the config ROM.  We do all this with a cached bus generation.  If the bus
@@ -409,16 +434,23 @@ read_rom(struct fw_device *device, int generation, int index, u32 *data)
  */
 static int read_bus_info_block(struct fw_device *device, int generation)
 {
-        static u32 rom[256];
-        u32 stack[16], sp, key;
-        int i, end, length;
+        u32 *rom, *stack, *old_rom, *new_rom;
+        u32 sp, key;
+        int i, end, length, ret = -1;
+
+        rom = kmalloc(sizeof(*rom) * READ_BIB_ROM_SIZE +
+                      sizeof(*stack) * READ_BIB_STACK_SIZE, GFP_KERNEL);
+        if (rom == NULL)
+                return -ENOMEM;
+
+        stack = &rom[READ_BIB_ROM_SIZE];
 
         device->max_speed = SCODE_100;
 
         /* First read the bus info block. */
         for (i = 0; i < 5; i++) {
                 if (read_rom(device, generation, i, &rom[i]) != RCODE_COMPLETE)
-                        return -1;
+                        goto out;
                 /*
                  * As per IEEE1212 7.2, during power-up, devices can
                  * reply with a 0 for the first quadlet of the config
@@ -428,7 +460,7 @@ static int read_bus_info_block(struct fw_device *device, int generation)
                  * retry mechanism will try again later.
                  */
                 if (i == 0 && rom[i] == 0)
-                        return -1;
+                        goto out;
         }
 
         device->max_speed = device->node->max_speed;
@@ -478,26 +510,26 @@ static int read_bus_info_block(struct fw_device *device, int generation)
                  */
                 key = stack[--sp];
                 i = key & 0xffffff;
-                if (i >= ARRAY_SIZE(rom))
+                if (i >= READ_BIB_ROM_SIZE)
                         /*
                          * The reference points outside the standard
                          * config rom area, something's fishy.
                          */
-                        return -1;
+                        goto out;
 
                 /* Read header quadlet for the block to get the length. */
                 if (read_rom(device, generation, i, &rom[i]) != RCODE_COMPLETE)
-                        return -1;
+                        goto out;
                 end = i + (rom[i] >> 16) + 1;
                 i++;
-                if (end > ARRAY_SIZE(rom))
+                if (end > READ_BIB_ROM_SIZE)
                         /*
                          * This block extends outside standard config
                          * area (and the array we're reading it
                          * into).  That's broken, so ignore this
                          * device.
                          */
-                        return -1;
+                        goto out;
 
                 /*
                  * Now read in the block.  If this is a directory
@@ -507,9 +539,9 @@ static int read_bus_info_block(struct fw_device *device, int generation)
                 while (i < end) {
                         if (read_rom(device, generation, i, &rom[i]) !=
                             RCODE_COMPLETE)
-                                return -1;
+                                goto out;
                         if ((key >> 30) == 3 && (rom[i] >> 30) > 1 &&
-                            sp < ARRAY_SIZE(stack))
+                            sp < READ_BIB_STACK_SIZE)
                                 stack[sp++] = i + rom[i];
                         i++;
                 }
@@ -517,13 +549,23 @@ static int read_bus_info_block(struct fw_device *device, int generation)
                         length = i;
         }
 
-        device->config_rom = kmalloc(length * 4, GFP_KERNEL);
-        if (device->config_rom == NULL)
-                return -1;
-        memcpy(device->config_rom, rom, length * 4);
+        old_rom = device->config_rom;
+        new_rom = kmemdup(rom, length * 4, GFP_KERNEL);
+        if (new_rom == NULL)
+                goto out;
+
+        down_write(&fw_device_rwsem);
+        device->config_rom = new_rom;
         device->config_rom_length = length;
+        up_write(&fw_device_rwsem);
 
-        return 0;
+        kfree(old_rom);
+        ret = 0;
+        device->cmc = rom[2] & 1 << 30;
+ out:
+        kfree(rom);
+
+        return ret;
 }
 
 static void fw_unit_release(struct device *dev)
@@ -592,7 +634,14 @@ static int shutdown_unit(struct device *device, void *data)
         return 0;
 }
 
-static DECLARE_RWSEM(idr_rwsem);
+/*
+ * fw_device_rwsem acts as dual purpose mutex:
+ *   - serializes accesses to fw_device_idr,
+ *   - serializes accesses to fw_device.config_rom/.config_rom_length and
+ *     fw_unit.directory, unless those accesses happen at safe occasions
+ */
+DECLARE_RWSEM(fw_device_rwsem);
+
 static DEFINE_IDR(fw_device_idr);
 int fw_cdev_major;
 
@@ -600,11 +649,11 @@ struct fw_device *fw_device_get_by_devt(dev_t devt)
 {
         struct fw_device *device;
 
-        down_read(&idr_rwsem);
+        down_read(&fw_device_rwsem);
         device = idr_find(&fw_device_idr, MINOR(devt));
         if (device)
                 fw_device_get(device);
-        up_read(&idr_rwsem);
+        up_read(&fw_device_rwsem);
 
         return device;
 }
@@ -619,9 +668,9 @@ static void fw_device_shutdown(struct work_struct *work)
         device_for_each_child(&device->device, NULL, shutdown_unit);
         device_unregister(&device->device);
 
-        down_write(&idr_rwsem);
+        down_write(&fw_device_rwsem);
         idr_remove(&fw_device_idr, minor);
-        up_write(&idr_rwsem);
+        up_write(&fw_device_rwsem);
         fw_device_put(device);
 }
 
@@ -674,10 +723,10 @@ static void fw_device_init(struct work_struct *work)
         err = -ENOMEM;
 
         fw_device_get(device);
-        down_write(&idr_rwsem);
+        down_write(&fw_device_rwsem);
         if (idr_pre_get(&fw_device_idr, GFP_KERNEL))
                 err = idr_get_new(&fw_device_idr, device, &minor);
-        up_write(&idr_rwsem);
+        up_write(&fw_device_rwsem);
 
         if (err < 0)
                 goto error;
@@ -711,7 +760,7 @@ static void fw_device_init(struct work_struct *work)
         if (atomic_cmpxchg(&device->state,
                     FW_DEVICE_INITIALIZING,
                     FW_DEVICE_RUNNING) == FW_DEVICE_SHUTDOWN) {
-                fw_device_shutdown(&device->work.work);
+                fw_device_shutdown(work);
         } else {
                 if (device->config_rom_retries)
                         fw_notify("created device %s: GUID %08x%08x, S%d00, "
@@ -725,6 +774,7 @@ static void fw_device_init(struct work_struct *work)
                                   device->device.bus_id,
                                   device->config_rom[3], device->config_rom[4],
                                   1 << device->max_speed);
+                device->config_rom_retries = 0;
         }
 
         /*
@@ -739,9 +789,9 @@ static void fw_device_init(struct work_struct *work)
         return;
 
  error_with_cdev:
-        down_write(&idr_rwsem);
+        down_write(&fw_device_rwsem);
         idr_remove(&fw_device_idr, minor);
-        up_write(&idr_rwsem);
+        up_write(&fw_device_rwsem);
  error:
         fw_device_put(device);          /* fw_device_idr's reference */
 
@@ -771,6 +821,106 @@ static void fw_device_update(struct work_struct *work)
         device_for_each_child(&device->device, NULL, update_unit);
 }
 
+enum {
+        REREAD_BIB_ERROR,
+        REREAD_BIB_GONE,
+        REREAD_BIB_UNCHANGED,
+        REREAD_BIB_CHANGED,
+};
+
+/* Reread and compare bus info block and header of root directory */
+static int reread_bus_info_block(struct fw_device *device, int generation)
+{
+        u32 q;
+        int i;
+
+        for (i = 0; i < 6; i++) {
+                if (read_rom(device, generation, i, &q) != RCODE_COMPLETE)
+                        return REREAD_BIB_ERROR;
+
+                if (i == 0 && q == 0)
+                        return REREAD_BIB_GONE;
+
+                if (i > device->config_rom_length || q != device->config_rom[i])
+                        return REREAD_BIB_CHANGED;
+        }
+
+        return REREAD_BIB_UNCHANGED;
+}
+
+static void fw_device_refresh(struct work_struct *work)
+{
+        struct fw_device *device =
+                container_of(work, struct fw_device, work.work);
+        struct fw_card *card = device->card;
+        int node_id = device->node_id;
+
+        switch (reread_bus_info_block(device, device->generation)) {
+        case REREAD_BIB_ERROR:
+                if (device->config_rom_retries < MAX_RETRIES / 2 &&
+                    atomic_read(&device->state) == FW_DEVICE_INITIALIZING) {
+                        device->config_rom_retries++;
+                        schedule_delayed_work(&device->work, RETRY_DELAY / 2);
+
+                        return;
+                }
+                goto give_up;
+
+        case REREAD_BIB_GONE:
+                goto gone;
+
+        case REREAD_BIB_UNCHANGED:
+                if (atomic_cmpxchg(&device->state,
+                            FW_DEVICE_INITIALIZING,
+                            FW_DEVICE_RUNNING) == FW_DEVICE_SHUTDOWN)
+                        goto gone;
+
+                fw_device_update(work);
+                device->config_rom_retries = 0;
+                goto out;
+
+        case REREAD_BIB_CHANGED:
+                break;
+        }
+
+        /*
+         * Something changed.  We keep things simple and don't investigate
+         * further.  We just destroy all previous units and create new ones.
+         */
+        device_for_each_child(&device->device, NULL, shutdown_unit);
+
+        if (read_bus_info_block(device, device->generation) < 0) {
+                if (device->config_rom_retries < MAX_RETRIES &&
+                    atomic_read(&device->state) == FW_DEVICE_INITIALIZING) {
+                        device->config_rom_retries++;
+                        schedule_delayed_work(&device->work, RETRY_DELAY);
+
+                        return;
+                }
+                goto give_up;
+        }
+
+        create_units(device);
+
+        if (atomic_cmpxchg(&device->state,
+                    FW_DEVICE_INITIALIZING,
+                    FW_DEVICE_RUNNING) == FW_DEVICE_SHUTDOWN)
+                goto gone;
+
+        fw_notify("refreshed device %s\n", device->device.bus_id);
+        device->config_rom_retries = 0;
+        goto out;
+
+ give_up:
+        fw_notify("giving up on refresh of device %s\n", device->device.bus_id);
+ gone:
+        atomic_set(&device->state, FW_DEVICE_SHUTDOWN);
+        fw_device_shutdown(work);
+ out:
+        if (node_id == card->root_node->node_id)
+                schedule_delayed_work(&card->work, 0);
+}
+
 void fw_node_event(struct fw_card *card, struct fw_node *node, int event)
 {
         struct fw_device *device;
@@ -780,7 +930,7 @@ void fw_node_event(struct fw_card *card, struct fw_node *node, int event)
         case FW_NODE_LINK_ON:
                 if (!node->link_on)
                         break;
-
+ create:
                 device = kzalloc(sizeof(*device), GFP_ATOMIC);
                 if (device == NULL)
                         break;
@@ -819,6 +969,23 @@ void fw_node_event(struct fw_card *card, struct fw_node *node, int event)
                 schedule_delayed_work(&device->work, INITIAL_DELAY);
                 break;
 
+        case FW_NODE_INITIATED_RESET:
+                device = node->data;
+                if (device == NULL)
+                        goto create;
+
+                device->node_id = node->node_id;
+                smp_wmb();  /* update node_id before generation */
+                device->generation = card->generation;
+                if (atomic_cmpxchg(&device->state,
+                            FW_DEVICE_RUNNING,
+                            FW_DEVICE_INITIALIZING) == FW_DEVICE_RUNNING) {
+                        PREPARE_DELAYED_WORK(&device->work, fw_device_refresh);
+                        schedule_delayed_work(&device->work,
+                                node == card->local_node ? 0 : INITIAL_DELAY);
+                }
+                break;
+
         case FW_NODE_UPDATED:
                 if (!node->link_on || node->data == NULL)
                         break;
diff --git a/drivers/firewire/fw-device.h b/drivers/firewire/fw-device.h
index 78ecd3991b7f..5f131f5129da 100644
--- a/drivers/firewire/fw-device.h
+++ b/drivers/firewire/fw-device.h
@@ -21,6 +21,7 @@
 
 #include <linux/fs.h>
 #include <linux/cdev.h>
+#include <linux/rwsem.h>
 #include <asm/atomic.h>
 
 enum fw_device_state {
@@ -46,6 +47,11 @@ struct fw_attribute_group {
  * fw_device.node_id is guaranteed to be current too.
  *
  * The same applies to fw_device.card->node_id vs. fw_device.generation.
+ *
+ * fw_device.config_rom and fw_device.config_rom_length may be accessed during
+ * the lifetime of any fw_unit belonging to the fw_device, before device_del()
+ * was called on the last fw_unit.  Alternatively, they may be accessed while
+ * holding fw_device_rwsem.
  */
 struct fw_device {
         atomic_t state;
@@ -53,6 +59,7 @@ struct fw_device {
         int node_id;
         int generation;
         unsigned max_speed;
+        bool cmc;
         struct fw_card *card;
         struct device device;
         struct list_head link;
@@ -64,28 +71,24 @@ struct fw_device {
         struct fw_attribute_group attribute_group;
 };
 
-static inline struct fw_device *
-fw_device(struct device *dev)
+static inline struct fw_device *fw_device(struct device *dev)
 {
         return container_of(dev, struct fw_device, device);
 }
 
-static inline int
-fw_device_is_shutdown(struct fw_device *device)
+static inline int fw_device_is_shutdown(struct fw_device *device)
 {
         return atomic_read(&device->state) == FW_DEVICE_SHUTDOWN;
 }
 
-static inline struct fw_device *
-fw_device_get(struct fw_device *device)
+static inline struct fw_device *fw_device_get(struct fw_device *device)
 {
         get_device(&device->device);
 
         return device;
 }
 
-static inline void
-fw_device_put(struct fw_device *device)
+static inline void fw_device_put(struct fw_device *device)
 {
         put_device(&device->device);
 }
@@ -96,20 +99,35 @@ int fw_device_enable_phys_dma(struct fw_device *device);
 void fw_device_cdev_update(struct fw_device *device);
 void fw_device_cdev_remove(struct fw_device *device);
 
+extern struct rw_semaphore fw_device_rwsem;
 extern int fw_cdev_major;
 
+/*
+ * fw_unit.directory must not be accessed after device_del(&fw_unit.device).
+ */
 struct fw_unit {
         struct device device;
         u32 *directory;
         struct fw_attribute_group attribute_group;
 };
 
-static inline struct fw_unit *
-fw_unit(struct device *dev)
+static inline struct fw_unit *fw_unit(struct device *dev)
 {
         return container_of(dev, struct fw_unit, device);
 }
 
+static inline struct fw_unit *fw_unit_get(struct fw_unit *unit)
+{
+        get_device(&unit->device);
+
+        return unit;
+}
+
+static inline void fw_unit_put(struct fw_unit *unit)
+{
+        put_device(&unit->device);
+}
+
 #define CSR_OFFSET      0x40
 #define CSR_LEAF        0x80
 #define CSR_DIRECTORY   0xc0
diff --git a/drivers/firewire/fw-iso.c b/drivers/firewire/fw-iso.c
index 2b640e9be6de..bcbe794a3ea5 100644
--- a/drivers/firewire/fw-iso.c
+++ b/drivers/firewire/fw-iso.c
@@ -126,7 +126,6 @@ fw_iso_context_create(struct fw_card *card, int type,
 
         return ctx;
 }
-EXPORT_SYMBOL(fw_iso_context_create);
 
 void fw_iso_context_destroy(struct fw_iso_context *ctx)
 {
@@ -134,14 +133,12 @@ void fw_iso_context_destroy(struct fw_iso_context *ctx)
 
         card->driver->free_iso_context(ctx);
 }
-EXPORT_SYMBOL(fw_iso_context_destroy);
 
 int
 fw_iso_context_start(struct fw_iso_context *ctx, int cycle, int sync, int tags)
 {
         return ctx->card->driver->start_iso(ctx, cycle, sync, tags);
 }
-EXPORT_SYMBOL(fw_iso_context_start);
 
 int
 fw_iso_context_queue(struct fw_iso_context *ctx,
@@ -153,11 +150,9 @@ fw_iso_context_queue(struct fw_iso_context *ctx,
 
         return card->driver->queue_iso(ctx, packet, buffer, payload);
 }
-EXPORT_SYMBOL(fw_iso_context_queue);
 
 int
 fw_iso_context_stop(struct fw_iso_context *ctx)
 {
         return ctx->card->driver->stop_iso(ctx);
 }
-EXPORT_SYMBOL(fw_iso_context_stop);
diff --git a/drivers/firewire/fw-ohci.c b/drivers/firewire/fw-ohci.c
index ca6d51efd8bb..4f02c55f13e1 100644
--- a/drivers/firewire/fw-ohci.c
+++ b/drivers/firewire/fw-ohci.c
@@ -27,6 +27,7 @@
 #include <linux/kernel.h>
 #include <linux/mm.h>
 #include <linux/module.h>
+#include <linux/moduleparam.h>
 #include <linux/pci.h>
 #include <linux/spinlock.h>
 
@@ -177,9 +178,10 @@ struct fw_ohci {
         struct tasklet_struct bus_reset_tasklet;
         int node_id;
         int generation;
-        int request_generation;
+        int request_generation; /* for timestamping incoming requests */
         u32 bus_seconds;
         bool old_uninorth;
+        bool bus_reset_packet_quirk;
 
         /*
          * Spinlock for accessing fw_ohci data.  Never call out of
@@ -237,6 +239,196 @@ static inline struct fw_ohci *fw_ohci(struct fw_card *card)
 
 static char ohci_driver_name[] = KBUILD_MODNAME;
 
+#ifdef CONFIG_FIREWIRE_OHCI_DEBUG
+
+#define OHCI_PARAM_DEBUG_AT_AR          1
+#define OHCI_PARAM_DEBUG_SELFIDS        2
+#define OHCI_PARAM_DEBUG_IRQS           4
+#define OHCI_PARAM_DEBUG_BUSRESETS      8 /* only effective before chip init */
+
+static int param_debug;
+module_param_named(debug, param_debug, int, 0644);
+MODULE_PARM_DESC(debug, "Verbose logging (default = 0"
+        ", AT/AR events = "     __stringify(OHCI_PARAM_DEBUG_AT_AR)
+        ", self-IDs = "         __stringify(OHCI_PARAM_DEBUG_SELFIDS)
+        ", IRQs = "             __stringify(OHCI_PARAM_DEBUG_IRQS)
+        ", busReset events = "  __stringify(OHCI_PARAM_DEBUG_BUSRESETS)
+        ", or a combination, or all = -1)");
+
+static void log_irqs(u32 evt)
+{
+        if (likely(!(param_debug &
+                        (OHCI_PARAM_DEBUG_IRQS | OHCI_PARAM_DEBUG_BUSRESETS))))
+                return;
+
+        if (!(param_debug & OHCI_PARAM_DEBUG_IRQS) &&
+            !(evt & OHCI1394_busReset))
+                return;
+
+        printk(KERN_DEBUG KBUILD_MODNAME ": IRQ "
+               "%08x%s%s%s%s%s%s%s%s%s%s%s%s%s\n",
+               evt,
+               evt & OHCI1394_selfIDComplete    ? " selfID"             : "",
+               evt & OHCI1394_RQPkt             ? " AR_req"             : "",
+               evt & OHCI1394_RSPkt             ? " AR_resp"            : "",
+               evt & OHCI1394_reqTxComplete     ? " AT_req"             : "",
+               evt & OHCI1394_respTxComplete    ? " AT_resp"            : "",
+               evt & OHCI1394_isochRx           ? " IR"                 : "",
+               evt & OHCI1394_isochTx           ? " IT"                 : "",
+               evt & OHCI1394_postedWriteErr    ? " postedWriteErr"     : "",
+               evt & OHCI1394_cycleTooLong      ? " cycleTooLong"       : "",
+               evt & OHCI1394_cycle64Seconds    ? " cycle64Seconds"     : "",
+               evt & OHCI1394_regAccessFail     ? " regAccessFail"      : "",
+               evt & OHCI1394_busReset          ? " busReset"           : "",
+               evt & ~(OHCI1394_selfIDComplete | OHCI1394_RQPkt |
+                       OHCI1394_RSPkt | OHCI1394_reqTxComplete |
+                       OHCI1394_respTxComplete | OHCI1394_isochRx |
+                       OHCI1394_isochTx | OHCI1394_postedWriteErr |
+                       OHCI1394_cycleTooLong | OHCI1394_cycle64Seconds |
+                       OHCI1394_regAccessFail | OHCI1394_busReset)
+                                                ? " ?"                  : "");
+}
+
+static const char *speed[] = {
+        [0] = "S100", [1] = "S200", [2] = "S400",    [3] = "beta",
+};
+static const char *power[] = {
+        [0] = "+0W",  [1] = "+15W", [2] = "+30W",    [3] = "+45W",
+        [4] = "-3W",  [5] = " ?W",  [6] = "-3..-6W", [7] = "-3..-10W",
+};
+static const char port[] = { '.', '-', 'p', 'c', };
+
+static char _p(u32 *s, int shift)
+{
+        return port[*s >> shift & 3];
+}
+
+static void log_selfids(int node_id, int generation, int self_id_count, u32 *s)
+{
+        if (likely(!(param_debug & OHCI_PARAM_DEBUG_SELFIDS)))
+                return;
+
+        printk(KERN_DEBUG KBUILD_MODNAME ": %d selfIDs, generation %d, "
+               "local node ID %04x\n", self_id_count, generation, node_id);
+
+        for (; self_id_count--; ++s)
+                if ((*s & 1 << 23) == 0)
+                        printk(KERN_DEBUG "selfID 0: %08x, phy %d [%c%c%c] "
+                               "%s gc=%d %s %s%s%s\n",
+                               *s, *s >> 24 & 63, _p(s, 6), _p(s, 4), _p(s, 2),
+                               speed[*s >> 14 & 3], *s >> 16 & 63,
+                               power[*s >> 8 & 7], *s >> 22 & 1 ? "L" : "",
+                               *s >> 11 & 1 ? "c" : "", *s & 2 ? "i" : "");
+                else
+                        printk(KERN_DEBUG "selfID n: %08x, phy %d "
+                               "[%c%c%c%c%c%c%c%c]\n",
+                               *s, *s >> 24 & 63,
+                               _p(s, 16), _p(s, 14), _p(s, 12), _p(s, 10),
+                               _p(s,  8), _p(s,  6), _p(s,  4), _p(s,  2));
+}
+
+static const char *evts[] = {
+        [0x00] = "evt_no_status",       [0x01] = "-reserved-",
+        [0x02] = "evt_long_packet",     [0x03] = "evt_missing_ack",
+        [0x04] = "evt_underrun",        [0x05] = "evt_overrun",
+        [0x06] = "evt_descriptor_read", [0x07] = "evt_data_read",
+        [0x08] = "evt_data_write",      [0x09] = "evt_bus_reset",
+        [0x0a] = "evt_timeout",         [0x0b] = "evt_tcode_err",
+        [0x0c] = "-reserved-",          [0x0d] = "-reserved-",
+        [0x0e] = "evt_unknown",         [0x0f] = "evt_flushed",
+        [0x10] = "-reserved-",          [0x11] = "ack_complete",
+        [0x12] = "ack_pending ",        [0x13] = "-reserved-",
+        [0x14] = "ack_busy_X",          [0x15] = "ack_busy_A",
+        [0x16] = "ack_busy_B",          [0x17] = "-reserved-",
+        [0x18] = "-reserved-",          [0x19] = "-reserved-",
+        [0x1a] = "-reserved-",          [0x1b] = "ack_tardy",
+        [0x1c] = "-reserved-",          [0x1d] = "ack_data_error",
+        [0x1e] = "ack_type_error",      [0x1f] = "-reserved-",
+        [0x20] = "pending/cancelled",
+};
+static const char *tcodes[] = {
+        [0x0] = "QW req",               [0x1] = "BW req",
+        [0x2] = "W resp",               [0x3] = "-reserved-",
+        [0x4] = "QR req",               [0x5] = "BR req",
+        [0x6] = "QR resp",              [0x7] = "BR resp",
+        [0x8] = "cycle start",          [0x9] = "Lk req",
+        [0xa] = "async stream packet",  [0xb] = "Lk resp",
+        [0xc] = "-reserved-",           [0xd] = "-reserved-",
+        [0xe] = "link internal",        [0xf] = "-reserved-",
+};
+static const char *phys[] = {
+        [0x0] = "phy config packet",    [0x1] = "link-on packet",
+        [0x2] = "self-id packet",       [0x3] = "-reserved-",
+};
+
+static void log_ar_at_event(char dir, int speed, u32 *header, int evt)
+{
+        int tcode = header[0] >> 4 & 0xf;
+        char specific[12];
+
+        if (likely(!(param_debug & OHCI_PARAM_DEBUG_AT_AR)))
+                return;
+
+        if (unlikely(evt >= ARRAY_SIZE(evts)))
+                        evt = 0x1f;
+
+        if (evt == OHCI1394_evt_bus_reset) {
+                printk(KERN_DEBUG "A%c evt_bus_reset, generation %d\n",
+                       dir, (header[2] >> 16) & 0xff);
+                return;
+        }
+
+        if (header[0] == ~header[1]) {
+                printk(KERN_DEBUG "A%c %s, %s, %08x\n",
+                       dir, evts[evt], phys[header[0] >> 30 & 0x3],
+                       header[0]);
+                return;
+        }
+
+        switch (tcode) {
+        case 0x0: case 0x6: case 0x8:
+                snprintf(specific, sizeof(specific), " = %08x",
+                         be32_to_cpu((__force __be32)header[3]));
+                break;
+        case 0x1: case 0x5: case 0x7: case 0x9: case 0xb:
+                snprintf(specific, sizeof(specific), " %x,%x",
+                         header[3] >> 16, header[3] & 0xffff);
+                break;
+        default:
+                specific[0] = '\0';
+        }
+
+        switch (tcode) {
+        case 0xe: case 0xa:
+                printk(KERN_DEBUG "A%c %s, %s\n",
+                       dir, evts[evt], tcodes[tcode]);
+                break;
+        case 0x0: case 0x1: case 0x4: case 0x5: case 0x9:
+                printk(KERN_DEBUG "A%c spd %x tl %02x, "
+                       "%04x -> %04x, %s, "
+                       "%s, %04x%08x%s\n",
+                       dir, speed, header[0] >> 10 & 0x3f,
+                       header[1] >> 16, header[0] >> 16, evts[evt],
+                       tcodes[tcode], header[1] & 0xffff, header[2], specific);
+                break;
+        default:
+                printk(KERN_DEBUG "A%c spd %x tl %02x, "
+                       "%04x -> %04x, %s, "
+                       "%s%s\n",
+                       dir, speed, header[0] >> 10 & 0x3f,
+                       header[1] >> 16, header[0] >> 16, evts[evt],
+                       tcodes[tcode], specific);
+        }
+}
+
+#else
+
+#define log_irqs(evt)
+#define log_selfids(node_id, generation, self_id_count, sid)
+#define log_ar_at_event(dir, speed, header, evt)
+
+#endif /* CONFIG_FIREWIRE_OHCI_DEBUG */
+
 static inline void reg_write(const struct fw_ohci *ohci, int offset, u32 data)
 {
         writel(data, ohci->registers + offset);
@@ -320,6 +512,7 @@ static __le32 *handle_ar_packet(struct ar_context *ctx, __le32 *buffer)
         struct fw_ohci *ohci = ctx->ohci;
         struct fw_packet p;
         u32 status, length, tcode;
+        int evt;
 
         p.header[0] = cond_le32_to_cpu(buffer[0]);
         p.header[1] = cond_le32_to_cpu(buffer[1]);
@@ -362,12 +555,15 @@ static __le32 *handle_ar_packet(struct ar_context *ctx, __le32 *buffer)
         /* FIXME: What to do about evt_* errors? */
         length = (p.header_length + p.payload_length + 3) / 4;
         status = cond_le32_to_cpu(buffer[length]);
+        evt    = (status >> 16) & 0x1f;
 
-        p.ack        = ((status >> 16) & 0x1f) - 16;
+        p.ack        = evt - 16;
         p.speed      = (status >> 21) & 0x7;
         p.timestamp  = status & 0xffff;
         p.generation = ohci->request_generation;
 
+        log_ar_at_event('R', p.speed, p.header, evt);
+
         /*
          * The OHCI bus reset handler synthesizes a phy packet with
          * the new generation number when a bus reset happens (see
@@ -376,14 +572,19 @@ static __le32 *handle_ar_packet(struct ar_context *ctx, __le32 *buffer)
          * generation.  We only need this for requests; for responses
          * we use the unique tlabel for finding the matching
          * request.
+         *
+         * Alas some chips sometimes emit bus reset packets with a
+         * wrong generation.  We set the correct generation for these
+         * at a slightly incorrect time (in bus_reset_tasklet).
          */
-
-        if (p.ack + 16 == 0x09)
-                ohci->request_generation = (p.header[2] >> 16) & 0xff;
-        else if (ctx == &ohci->ar_request_ctx)
+        if (evt == OHCI1394_evt_bus_reset) {
+                if (!ohci->bus_reset_packet_quirk)
+                        ohci->request_generation = (p.header[2] >> 16) & 0xff;
+        } else if (ctx == &ohci->ar_request_ctx) {
                 fw_core_handle_request(&ohci->card, &p);
-        else
+        } else {
                 fw_core_handle_response(&ohci->card, &p);
+        }
 
         return buffer + length + 1;
 }
@@ -770,8 +971,19 @@ at_context_queue_packet(struct context *ctx, struct fw_packet *packet)
                                      DESCRIPTOR_IRQ_ALWAYS |
                                      DESCRIPTOR_BRANCH_ALWAYS);
 
-        /* FIXME: Document how the locking works. */
-        if (ohci->generation != packet->generation) {
+        /*
+         * If the controller and packet generations don't match, we need to
+         * bail out and try again.  If IntEvent.busReset is set, the AT context
+         * is halted, so appending to the context and trying to run it is
+         * futile.  Most controllers do the right thing and just flush the AT
+         * queue (per section 7.2.3.2 of the OHCI 1.1 specification), but
+         * some controllers (like a JMicron JMB381 PCI-e) misbehave and wind
+         * up stalling out.  So we just bail out in software and try again
+         * later, and everyone is happy.
+         * FIXME: Document how the locking works.
+         */
+        if (ohci->generation != packet->generation ||
+            reg_read(ohci, OHCI1394_IntEventSet) & OHCI1394_busReset) {
                 if (packet->payload_length > 0)
                         dma_unmap_single(ohci->card.device, payload_bus,
                                          packet->payload_length, DMA_TO_DEVICE);
@@ -817,6 +1029,8 @@ static int handle_at_packet(struct context *context,
         evt = le16_to_cpu(last->transfer_status) & 0x1f;
         packet->timestamp = le16_to_cpu(last->res_count);
 
+        log_ar_at_event('T', packet->speed, packet->header, evt);
+
         switch (evt) {
         case OHCI1394_evt_timeout:
                 /* Async response transmit timed out. */
@@ -1019,20 +1233,30 @@ static void bus_reset_tasklet(unsigned long data)
         ohci->node_id = reg & (OHCI1394_NodeID_busNumber |
                                OHCI1394_NodeID_nodeNumber);
 
+        reg = reg_read(ohci, OHCI1394_SelfIDCount);
+        if (reg & OHCI1394_SelfIDCount_selfIDError) {
+                fw_notify("inconsistent self IDs\n");
+                return;
+        }
         /*
          * The count in the SelfIDCount register is the number of
          * bytes in the self ID receive buffer.  Since we also receive
          * the inverted quadlets and a header quadlet, we shift one
          * bit extra to get the actual number of self IDs.
          */
-
-        self_id_count = (reg_read(ohci, OHCI1394_SelfIDCount) >> 3) & 0x3ff;
+        self_id_count = (reg >> 3) & 0x3ff;
+        if (self_id_count == 0) {
+                fw_notify("inconsistent self IDs\n");
+                return;
+        }
         generation = (cond_le32_to_cpu(ohci->self_id_cpu[0]) >> 16) & 0xff;
         rmb();
 
         for (i = 1, j = 0; j < self_id_count; i += 2, j++) {
-                if (ohci->self_id_cpu[i] != ~ohci->self_id_cpu[i + 1])
-                        fw_error("inconsistent self IDs\n");
+                if (ohci->self_id_cpu[i] != ~ohci->self_id_cpu[i + 1]) {
+                        fw_notify("inconsistent self IDs\n");
+                        return;
+                }
                 ohci->self_id_buffer[j] =
                                 cond_le32_to_cpu(ohci->self_id_cpu[i]);
         }
@@ -1067,6 +1291,9 @@ static void bus_reset_tasklet(unsigned long data)
         context_stop(&ohci->at_response_ctx);
         reg_write(ohci, OHCI1394_IntEventClear, OHCI1394_busReset);
 
+        if (ohci->bus_reset_packet_quirk)
+                ohci->request_generation = generation;
+
         /*
          * This next bit is unrelated to the AT context stuff but we
          * have to do it under the spinlock also.  If a new config rom
@@ -1097,12 +1324,20 @@ static void bus_reset_tasklet(unsigned long data)
                 reg_write(ohci, OHCI1394_ConfigROMhdr, ohci->next_header);
         }
 
+#ifdef CONFIG_FIREWIRE_OHCI_REMOTE_DMA
+        reg_write(ohci, OHCI1394_PhyReqFilterHiSet, ~0);
+        reg_write(ohci, OHCI1394_PhyReqFilterLoSet, ~0);
+#endif
+
         spin_unlock_irqrestore(&ohci->lock, flags);
 
         if (free_rom)
                 dma_free_coherent(ohci->card.device, CONFIG_ROM_SIZE,
                                   free_rom, free_rom_bus);
 
+        log_selfids(ohci->node_id, generation,
+                    self_id_count, ohci->self_id_buffer);
+
         fw_core_handle_bus_reset(&ohci->card, ohci->node_id, generation,
                                  self_id_count, ohci->self_id_buffer);
 }
@@ -1118,7 +1353,9 @@ static irqreturn_t irq_handler(int irq, void *data)
         if (!event || !~event)
                 return IRQ_NONE;
 
-        reg_write(ohci, OHCI1394_IntEventClear, event);
+        /* busReset must not be cleared yet, see OHCI 1.1 clause 7.2.3.2 */
+        reg_write(ohci, OHCI1394_IntEventClear, event & ~OHCI1394_busReset);
+        log_irqs(event);
 
         if (event & OHCI1394_selfIDComplete)
                 tasklet_schedule(&ohci->bus_reset_tasklet);
@@ -1153,6 +1390,10 @@ static irqreturn_t irq_handler(int irq, void *data)
                 iso_event &= ~(1 << i);
         }
 
+        if (unlikely(event & OHCI1394_regAccessFail))
+                fw_error("Register access failure - "
+                         "please notify linux1394-devel@lists.sf.net\n");
+
         if (unlikely(event & OHCI1394_postedWriteErr))
                 fw_error("PCI posted write error\n");
 
@@ -1192,6 +1433,8 @@ static int ohci_enable(struct fw_card *card, u32 *config_rom, size_t length)
 {
         struct fw_ohci *ohci = fw_ohci(card);
         struct pci_dev *dev = to_pci_dev(card->device);
+        u32 lps;
+        int i;
 
         if (software_reset(ohci)) {
                 fw_error("Failed to reset ohci card.\n");
@@ -1203,13 +1446,24 @@ static int ohci_enable(struct fw_card *card, u32 *config_rom, size_t length)
          * most of the registers.  In fact, on some cards (ALI M5251),
          * accessing registers in the SClk domain without LPS enabled
          * will lock up the machine.  Wait 50msec to make sure we have
-         * full link enabled.
+         * full link enabled.  However, with some cards (well, at least
+         * a JMicron PCIe card), we have to try again sometimes.
          */
         reg_write(ohci, OHCI1394_HCControlSet,
                   OHCI1394_HCControl_LPS |
                   OHCI1394_HCControl_postedWriteEnable);
         flush_writes(ohci);
-        msleep(50);
+
+        for (lps = 0, i = 0; !lps && i < 3; i++) {
+                msleep(50);
+                lps = reg_read(ohci, OHCI1394_HCControlSet) &
+                      OHCI1394_HCControl_LPS;
+        }
+
+        if (!lps) {
+                fw_error("Failed to set Link Power Status\n");
+                return -EIO;
+        }
 
         reg_write(ohci, OHCI1394_HCControlClear,
                   OHCI1394_HCControl_noByteSwapData);
@@ -1237,7 +1491,10 @@ static int ohci_enable(struct fw_card *card, u32 *config_rom, size_t length)
                   OHCI1394_reqTxComplete | OHCI1394_respTxComplete |
                   OHCI1394_isochRx | OHCI1394_isochTx |
                   OHCI1394_postedWriteErr | OHCI1394_cycleTooLong |
-                  OHCI1394_cycle64Seconds | OHCI1394_masterIntEnable);
+                  OHCI1394_cycle64Seconds | OHCI1394_regAccessFail |
+                  OHCI1394_masterIntEnable);
+        if (param_debug & OHCI_PARAM_DEBUG_BUSRESETS)
+                reg_write(ohci, OHCI1394_IntMaskSet, OHCI1394_busReset);
 
         /* Activate link_on bit and contender bit in our self ID packets.*/
         if (ohci_update_phy_reg(card, 4, 0,
@@ -1421,6 +1678,7 @@ static int ohci_cancel_packet(struct fw_card *card, struct fw_packet *packet)
         if (packet->ack != 0)
                 goto out;
 
+        log_ar_at_event('T', packet->speed, packet->header, 0x20);
         driver_data->packet = NULL;
         packet->ack = RCODE_CANCELLED;
         packet->callback(packet, &ohci->card, packet->ack);
@@ -1435,6 +1693,9 @@ static int ohci_cancel_packet(struct fw_card *card, struct fw_packet *packet)
 static int
 ohci_enable_phys_dma(struct fw_card *card, int node_id, int generation)
 {
+#ifdef CONFIG_FIREWIRE_OHCI_REMOTE_DMA
+        return 0;
+#else
         struct fw_ohci *ohci = fw_ohci(card);
         unsigned long flags;
         int n, retval = 0;
@@ -1466,6 +1727,7 @@ ohci_enable_phys_dma(struct fw_card *card, int node_id, int generation)
  out:
         spin_unlock_irqrestore(&ohci->lock, flags);
         return retval;
+#endif /* CONFIG_FIREWIRE_OHCI_REMOTE_DMA */
 }
 
 static u64
@@ -2045,17 +2307,9 @@ static const struct fw_card_driver ohci_driver = {
         .stop_iso               = ohci_stop_iso,
 };
 
-static int __devinit
-pci_probe(struct pci_dev *dev, const struct pci_device_id *ent)
-{
-        struct fw_ohci *ohci;
-        u32 bus_options, max_receive, link_speed;
-        u64 guid;
-        int err;
-        size_t size;
-
 #ifdef CONFIG_PPC_PMAC
-        /* Necessary on some machines if fw-ohci was loaded/ unloaded before */
+static void ohci_pmac_on(struct pci_dev *dev)
+{
         if (machine_is(powermac)) {
                 struct device_node *ofn = pci_device_to_OF_node(dev);
 
@@ -2064,8 +2318,33 @@ pci_probe(struct pci_dev *dev, const struct pci_device_id *ent)
                         pmac_call_feature(PMAC_FTR_1394_ENABLE, ofn, 0, 1);
                 }
         }
+}
+
+static void ohci_pmac_off(struct pci_dev *dev)
+{
+        if (machine_is(powermac)) {
+                struct device_node *ofn = pci_device_to_OF_node(dev);
+
+                if (ofn) {
+                        pmac_call_feature(PMAC_FTR_1394_ENABLE, ofn, 0, 0);
+                        pmac_call_feature(PMAC_FTR_1394_CABLE_POWER, ofn, 0, 0);
+                }
+        }
+}
+#else
+#define ohci_pmac_on(dev)
+#define ohci_pmac_off(dev)
 #endif /* CONFIG_PPC_PMAC */
 
+static int __devinit
+pci_probe(struct pci_dev *dev, const struct pci_device_id *ent)
+{
+        struct fw_ohci *ohci;
+        u32 bus_options, max_receive, link_speed;
+        u64 guid;
+        int err;
+        size_t size;
+
         ohci = kzalloc(sizeof(*ohci), GFP_KERNEL);
         if (ohci == NULL) {
                 fw_error("Could not malloc fw_ohci data.\n");
@@ -2074,10 +2353,12 @@ pci_probe(struct pci_dev *dev, const struct pci_device_id *ent)
 
         fw_card_initialize(&ohci->card, &ohci_driver, &dev->dev);
 
+        ohci_pmac_on(dev);
+
         err = pci_enable_device(dev);
         if (err) {
                 fw_error("Failed to enable OHCI hardware.\n");
-                goto fail_put_card;
+                goto fail_free;
         }
 
         pci_set_master(dev);
@@ -2088,6 +2369,8 @@ pci_probe(struct pci_dev *dev, const struct pci_device_id *ent)
         ohci->old_uninorth = dev->vendor == PCI_VENDOR_ID_APPLE &&
                              dev->device == PCI_DEVICE_ID_APPLE_UNI_N_FW;
 #endif
+        ohci->bus_reset_packet_quirk = dev->vendor == PCI_VENDOR_ID_TI;
+
         spin_lock_init(&ohci->lock);
 
         tasklet_init(&ohci->bus_reset_tasklet,
@@ -2173,8 +2456,9 @@ pci_probe(struct pci_dev *dev, const struct pci_device_id *ent)
         pci_release_region(dev, 0);
  fail_disable:
         pci_disable_device(dev);
- fail_put_card:
-        fw_card_put(&ohci->card);
+ fail_free:
+        kfree(&ohci->card);
+        ohci_pmac_off(dev);
 
         return err;
 }
@@ -2202,72 +2486,42 @@ static void pci_remove(struct pci_dev *dev)
         pci_iounmap(dev, ohci->registers);
         pci_release_region(dev, 0);
         pci_disable_device(dev);
-        fw_card_put(&ohci->card);
-
-#ifdef CONFIG_PPC_PMAC
-        /* On UniNorth, power down the cable and turn off the chip clock
-         * to save power on laptops */
-        if (machine_is(powermac)) {
-                struct device_node *ofn = pci_device_to_OF_node(dev);
-
-                if (ofn) {
-                        pmac_call_feature(PMAC_FTR_1394_ENABLE, ofn, 0, 0);
-                        pmac_call_feature(PMAC_FTR_1394_CABLE_POWER, ofn, 0, 0);
-                }
-        }
-#endif /* CONFIG_PPC_PMAC */
+        kfree(&ohci->card);
+        ohci_pmac_off(dev);
 
         fw_notify("Removed fw-ohci device.\n");
 }
 
 #ifdef CONFIG_PM
-static int pci_suspend(struct pci_dev *pdev, pm_message_t state)
+static int pci_suspend(struct pci_dev *dev, pm_message_t state)
 {
-        struct fw_ohci *ohci = pci_get_drvdata(pdev);
+        struct fw_ohci *ohci = pci_get_drvdata(dev);
         int err;
 
         software_reset(ohci);
-        free_irq(pdev->irq, ohci);
-        err = pci_save_state(pdev);
+        free_irq(dev->irq, ohci);
+        err = pci_save_state(dev);
         if (err) {
                 fw_error("pci_save_state failed\n");
                 return err;
         }
-        err = pci_set_power_state(pdev, pci_choose_state(pdev, state));
+        err = pci_set_power_state(dev, pci_choose_state(dev, state));
         if (err)
                 fw_error("pci_set_power_state failed with %d\n", err);
-
-/* PowerMac suspend code comes last */
-#ifdef CONFIG_PPC_PMAC
-        if (machine_is(powermac)) {
-                struct device_node *ofn = pci_device_to_OF_node(pdev);
-
-                if (ofn)
-                        pmac_call_feature(PMAC_FTR_1394_ENABLE, ofn, 0, 0);
-        }
-#endif /* CONFIG_PPC_PMAC */
+        ohci_pmac_off(dev);
 
         return 0;
 }
 
-static int pci_resume(struct pci_dev *pdev)
+static int pci_resume(struct pci_dev *dev)
 {
-        struct fw_ohci *ohci = pci_get_drvdata(pdev);
+        struct fw_ohci *ohci = pci_get_drvdata(dev);
         int err;
 
-/* PowerMac resume code comes first */
-#ifdef CONFIG_PPC_PMAC
-        if (machine_is(powermac)) {
-                struct device_node *ofn = pci_device_to_OF_node(pdev);
-
-                if (ofn)
-                        pmac_call_feature(PMAC_FTR_1394_ENABLE, ofn, 0, 1);
-        }
-#endif /* CONFIG_PPC_PMAC */
-
-        pci_set_power_state(pdev, PCI_D0);
-        pci_restore_state(pdev);
-        err = pci_enable_device(pdev);
+        ohci_pmac_on(dev);
+        pci_set_power_state(dev, PCI_D0);
+        pci_restore_state(dev);
+        err = pci_enable_device(dev);
         if (err) {
                 fw_error("pci_enable_device failed\n");
                 return err;
diff --git a/drivers/firewire/fw-ohci.h b/drivers/firewire/fw-ohci.h
index dec4f04e6b24..a2fbb6240ca7 100644
--- a/drivers/firewire/fw-ohci.h
+++ b/drivers/firewire/fw-ohci.h
@@ -30,6 +30,7 @@
 #define  OHCI1394_HCControl_softReset           0x00010000
 #define OHCI1394_SelfIDBuffer                 0x064
 #define OHCI1394_SelfIDCount                  0x068
+#define  OHCI1394_SelfIDCount_selfIDError       0x80000000
 #define OHCI1394_IRMultiChanMaskHiSet         0x070
 #define OHCI1394_IRMultiChanMaskHiClear       0x074
 #define OHCI1394_IRMultiChanMaskLoSet         0x078
@@ -124,6 +125,7 @@
 #define OHCI1394_lockRespErr            0x00000200
 #define OHCI1394_selfIDComplete         0x00010000
 #define OHCI1394_busReset               0x00020000
+#define OHCI1394_regAccessFail          0x00040000
 #define OHCI1394_phy                    0x00080000
 #define OHCI1394_cycleSynch             0x00100000
 #define OHCI1394_cycle64Seconds         0x00200000
diff --git a/drivers/firewire/fw-sbp2.c b/drivers/firewire/fw-sbp2.c
index 62b4e47d0cc0..2a999373863e 100644
--- a/drivers/firewire/fw-sbp2.c
+++ b/drivers/firewire/fw-sbp2.c
@@ -153,6 +153,7 @@ struct sbp2_target {
         struct list_head lu_list;
 
         u64 management_agent_address;
+        u64 guid;
         int directory_id;
         int node_id;
         int address_high;
@@ -173,10 +174,8 @@ struct sbp2_target {
 #define SBP2_ORB_TIMEOUT                2000U   /* Timeout in ms */
 #define SBP2_ORB_NULL                   0x80000000
 #define SBP2_MAX_SG_ELEMENT_LENGTH      0xf000
-#define SBP2_RETRY_LIMIT                0xf     /* 15 retries */
-
-#define SBP2_DIRECTION_TO_MEDIA         0x0
-#define SBP2_DIRECTION_FROM_MEDIA       0x1
+#define SBP2_RETRY_LIMIT                0xf             /* 15 retries */
+#define SBP2_CYCLE_LIMIT                (0xc8 << 12)    /* 200 125us cycles */
 
 /* Unit directory keys */
 #define SBP2_CSR_UNIT_CHARACTERISTICS   0x3a
@@ -224,8 +223,8 @@ struct sbp2_status {
 };
 
 struct sbp2_pointer {
-        u32 high;
-        u32 low;
+        __be32 high;
+        __be32 low;
 };
 
 struct sbp2_orb {
@@ -253,8 +252,8 @@ struct sbp2_management_orb {
         struct {
                 struct sbp2_pointer password;
                 struct sbp2_pointer response;
-                u32 misc;
-                u32 length;
+                __be32 misc;
+                __be32 length;
                 struct sbp2_pointer status_fifo;
         } request;
         __be32 response[4];
@@ -263,20 +262,17 @@ struct sbp2_management_orb {
         struct sbp2_status status;
 };
 
-#define LOGIN_RESPONSE_GET_LOGIN_ID(v)  ((v).misc & 0xffff)
-#define LOGIN_RESPONSE_GET_LENGTH(v)    (((v).misc >> 16) & 0xffff)
-
 struct sbp2_login_response {
-        u32 misc;
+        __be32 misc;
         struct sbp2_pointer command_block_agent;
-        u32 reconnect_hold;
+        __be32 reconnect_hold;
 };
 #define COMMAND_ORB_DATA_SIZE(v)        ((v))
 #define COMMAND_ORB_PAGE_SIZE(v)        ((v) << 16)
 #define COMMAND_ORB_PAGE_TABLE_PRESENT  ((1) << 19)
 #define COMMAND_ORB_MAX_PAYLOAD(v)      ((v) << 20)
 #define COMMAND_ORB_SPEED(v)            ((v) << 24)
-#define COMMAND_ORB_DIRECTION(v)        ((v) << 27)
+#define COMMAND_ORB_DIRECTION           ((1) << 27)
 #define COMMAND_ORB_REQUEST_FORMAT(v)   ((v) << 29)
 #define COMMAND_ORB_NOTIFY              ((1) << 31)
 
@@ -285,7 +281,7 @@ struct sbp2_command_orb {
         struct {
                 struct sbp2_pointer next;
                 struct sbp2_pointer data_descriptor;
-                u32 misc;
+                __be32 misc;
                 u8 command_block[12];
         } request;
         struct scsi_cmnd *cmd;
@@ -459,8 +455,7 @@ sbp2_send_orb(struct sbp2_orb *orb, struct sbp2_logical_unit *lu,
         unsigned long flags;
 
         orb->pointer.high = 0;
-        orb->pointer.low = orb->request_bus;
-        fw_memcpy_to_be32(&orb->pointer, &orb->pointer, sizeof(orb->pointer));
+        orb->pointer.low = cpu_to_be32(orb->request_bus);
 
         spin_lock_irqsave(&device->card->lock, flags);
         list_add_tail(&orb->link, &lu->orb_list);
@@ -536,31 +531,31 @@ sbp2_send_management_orb(struct sbp2_logical_unit *lu, int node_id,
         if (dma_mapping_error(orb->response_bus))
                 goto fail_mapping_response;
 
-        orb->request.response.high    = 0;
-        orb->request.response.low     = orb->response_bus;
+        orb->request.response.high = 0;
+        orb->request.response.low  = cpu_to_be32(orb->response_bus);
 
-        orb->request.misc =
+        orb->request.misc = cpu_to_be32(
                 MANAGEMENT_ORB_NOTIFY |
                 MANAGEMENT_ORB_FUNCTION(function) |
-                MANAGEMENT_ORB_LUN(lun_or_login_id);
-        orb->request.length =
-                MANAGEMENT_ORB_RESPONSE_LENGTH(sizeof(orb->response));
+                MANAGEMENT_ORB_LUN(lun_or_login_id));
+        orb->request.length = cpu_to_be32(
+                MANAGEMENT_ORB_RESPONSE_LENGTH(sizeof(orb->response)));
 
-        orb->request.status_fifo.high = lu->address_handler.offset >> 32;
-        orb->request.status_fifo.low  = lu->address_handler.offset;
+        orb->request.status_fifo.high =
+                cpu_to_be32(lu->address_handler.offset >> 32);
+        orb->request.status_fifo.low  =
+                cpu_to_be32(lu->address_handler.offset);
 
         if (function == SBP2_LOGIN_REQUEST) {
                 /* Ask for 2^2 == 4 seconds reconnect grace period */
-                orb->request.misc |=
+                orb->request.misc |= cpu_to_be32(
                         MANAGEMENT_ORB_RECONNECT(2) |
-                        MANAGEMENT_ORB_EXCLUSIVE(sbp2_param_exclusive_login);
+                        MANAGEMENT_ORB_EXCLUSIVE(sbp2_param_exclusive_login));
                 timeout = lu->tgt->mgt_orb_timeout;
         } else {
                 timeout = SBP2_ORB_TIMEOUT;
         }
 
-        fw_memcpy_to_be32(&orb->request, &orb->request, sizeof(orb->request));
-
         init_completion(&orb->done);
         orb->base.callback = complete_management_orb;
 
@@ -605,8 +600,7 @@ sbp2_send_management_orb(struct sbp2_logical_unit *lu, int node_id,
                          sizeof(orb->response), DMA_FROM_DEVICE);
  fail_mapping_response:
         if (response)
-                fw_memcpy_from_be32(response,
-                                    orb->response, sizeof(orb->response));
+                memcpy(response, orb->response, sizeof(orb->response));
         kref_put(&orb->base.kref, free_orb);
 
         return retval;
@@ -701,10 +695,8 @@ static void sbp2_conditionally_block(struct sbp2_logical_unit *lu)
         if (!tgt->dont_block && !lu->blocked &&
             lu->generation != card->generation) {
                 lu->blocked = true;
-                if (++tgt->blocked == 1) {
+                if (++tgt->blocked == 1)
                         scsi_block_requests(shost);
-                        fw_notify("blocked %s\n", lu->tgt->bus_id);
-                }
         }
         spin_unlock_irqrestore(&card->lock, flags);
 }
@@ -731,10 +723,8 @@ static void sbp2_conditionally_unblock(struct sbp2_logical_unit *lu)
         }
         spin_unlock_irqrestore(&card->lock, flags);
 
-        if (unblock) {
+        if (unblock)
                 scsi_unblock_requests(shost);
-                fw_notify("unblocked %s\n", lu->tgt->bus_id);
-        }
 }
 
 /*
@@ -796,7 +786,7 @@ static void sbp2_release_target(struct kref *kref)
         scsi_remove_host(shost);
         fw_notify("released %s\n", tgt->bus_id);
 
-        put_device(&tgt->unit->device);
+        fw_unit_put(tgt->unit);
         scsi_host_put(shost);
         fw_device_put(device);
 }
@@ -825,6 +815,22 @@ complete_set_busy_timeout(struct fw_card *card, int rcode,
         complete(done);
 }
 
+/*
+ * Write retransmit retry values into the BUSY_TIMEOUT register.
+ * - The single-phase retry protocol is supported by all SBP-2 devices, but the
+ *   default retry_limit value is 0 (i.e. never retry transmission). We write a
+ *   saner value after logging into the device.
+ * - The dual-phase retry protocol is optional to implement, and if not
+ *   supported, writes to the dual-phase portion of the register will be
+ *   ignored. We try to write the original 1394-1995 default here.
+ * - In the case of devices that are also SBP-3-compliant, all writes are
+ *   ignored, as the register is read-only, but contains single-phase retry of
+ *   15, which is what we're trying to set for all SBP-2 device anyway, so this
+ *   write attempt is safe and yields more consistent behavior for all devices.
+ *
+ * See section 8.3.2.3.5 of the 1394-1995 spec, section 6.2 of the SBP-2 spec,
+ * and section 6.4 of the SBP-3 spec for further details.
+ */
 static void sbp2_set_busy_timeout(struct sbp2_logical_unit *lu)
 {
         struct fw_device *device = fw_device(lu->tgt->unit->device.parent);
@@ -832,8 +838,7 @@ static void sbp2_set_busy_timeout(struct sbp2_logical_unit *lu)
         struct fw_transaction t;
         static __be32 busy_timeout;
 
-        /* FIXME: we should try to set dual-phase cycle_limit too */
-        busy_timeout = cpu_to_be32(SBP2_RETRY_LIMIT);
+        busy_timeout = cpu_to_be32(SBP2_CYCLE_LIMIT | SBP2_RETRY_LIMIT);
 
         fw_send_request(device->card, &t, TCODE_WRITE_QUADLET_REQUEST,
                         lu->tgt->node_id, lu->generation, device->max_speed,
@@ -885,11 +890,10 @@ static void sbp2_login(struct work_struct *work)
         tgt->address_high = local_node_id << 16;
         sbp2_set_generation(lu, generation);
 
-        /* Get command block agent offset and login id. */
         lu->command_block_agent_address =
-                ((u64) (response.command_block_agent.high & 0xffff) << 32) |
-                response.command_block_agent.low;
-        lu->login_id = LOGIN_RESPONSE_GET_LOGIN_ID(response);
+                ((u64)(be32_to_cpu(response.command_block_agent.high) & 0xffff)
+                      << 32) | be32_to_cpu(response.command_block_agent.low);
+        lu->login_id = be32_to_cpu(response.misc) & 0xffff;
 
         fw_notify("%s: logged in to LUN %04x (%d retries)\n",
                   tgt->bus_id, lu->lun, lu->retries);
@@ -1111,6 +1115,7 @@ static int sbp2_probe(struct device *dev)
         kref_init(&tgt->kref);
         INIT_LIST_HEAD(&tgt->lu_list);
         tgt->bus_id = unit->device.bus_id;
+        tgt->guid = (u64)device->config_rom[3] << 32 | device->config_rom[4];
 
         if (fw_device_enable_phys_dma(device) < 0)
                 goto fail_shost_put;
@@ -1119,6 +1124,7 @@ static int sbp2_probe(struct device *dev)
                 goto fail_shost_put;
 
         fw_device_get(device);
+        fw_unit_get(unit);
 
         /* Initialize to values that won't match anything in our table. */
         firmware_revision = 0xff000000;
@@ -1134,8 +1140,6 @@ static int sbp2_probe(struct device *dev)
 
         sbp2_init_workarounds(tgt, model, firmware_revision);
 
-        get_device(&unit->device);
-
         /* Do the login in a workqueue so we can easily reschedule retries. */
         list_for_each_entry(lu, &tgt->lu_list, link)
                 sbp2_queue_work(lu, 0);
@@ -1367,9 +1371,12 @@ sbp2_map_scatterlist(struct sbp2_command_orb *orb, struct fw_device *device,
          * tables.
          */
         if (count == 1 && sg_dma_len(sg) < SBP2_MAX_SG_ELEMENT_LENGTH) {
-                orb->request.data_descriptor.high = lu->tgt->address_high;
-                orb->request.data_descriptor.low  = sg_dma_address(sg);
-                orb->request.misc |= COMMAND_ORB_DATA_SIZE(sg_dma_len(sg));
+                orb->request.data_descriptor.high =
+                        cpu_to_be32(lu->tgt->address_high);
+                orb->request.data_descriptor.low  =
+                        cpu_to_be32(sg_dma_address(sg));
+                orb->request.misc |=
+                        cpu_to_be32(COMMAND_ORB_DATA_SIZE(sg_dma_len(sg)));
                 return 0;
         }
 
@@ -1390,16 +1397,14 @@ sbp2_map_scatterlist(struct sbp2_command_orb *orb, struct fw_device *device,
                                 goto fail_page_table;
                         }
                         l = min(sg_len, SBP2_MAX_SG_ELEMENT_LENGTH);
-                        orb->page_table[j].low = sg_addr;
-                        orb->page_table[j].high = (l << 16);
+                        orb->page_table[j].low = cpu_to_be32(sg_addr);
+                        orb->page_table[j].high = cpu_to_be32(l << 16);
                         sg_addr += l;
                         sg_len -= l;
                         j++;
                 }
         }
 
-        fw_memcpy_to_be32(orb->page_table, orb->page_table,
-                          sizeof(orb->page_table[0]) * j);
         orb->page_table_bus =
                 dma_map_single(device->card->device, orb->page_table,
                                sizeof(orb->page_table), DMA_TO_DEVICE);
@@ -1413,11 +1418,10 @@ sbp2_map_scatterlist(struct sbp2_command_orb *orb, struct fw_device *device,
          * initiator (i.e. us), but data_descriptor can refer to data
          * on other nodes so we need to put our ID in descriptor.high.
          */
-        orb->request.data_descriptor.high = lu->tgt->address_high;
-        orb->request.data_descriptor.low  = orb->page_table_bus;
-        orb->request.misc |=
-                COMMAND_ORB_PAGE_TABLE_PRESENT |
-                COMMAND_ORB_DATA_SIZE(j);
+        orb->request.data_descriptor.high = cpu_to_be32(lu->tgt->address_high);
+        orb->request.data_descriptor.low  = cpu_to_be32(orb->page_table_bus);
+        orb->request.misc |= cpu_to_be32(COMMAND_ORB_PAGE_TABLE_PRESENT |
+                                         COMMAND_ORB_DATA_SIZE(j));
 
         return 0;
 
@@ -1463,8 +1467,7 @@ static int sbp2_scsi_queuecommand(struct scsi_cmnd *cmd, scsi_done_fn_t done)
         orb->done = done;
         orb->cmd  = cmd;
 
-        orb->request.next.high   = SBP2_ORB_NULL;
-        orb->request.next.low    = 0x0;
+        orb->request.next.high   = cpu_to_be32(SBP2_ORB_NULL);
         /*
          * At speed 100 we can do 512 bytes per packet, at speed 200,
          * 1024 bytes per packet etc.  The SBP-2 max_payload field
@@ -1473,25 +1476,17 @@ static int sbp2_scsi_queuecommand(struct scsi_cmnd *cmd, scsi_done_fn_t done)
          */
         max_payload = min(device->max_speed + 7,
                           device->card->max_receive - 1);
-        orb->request.misc =
+        orb->request.misc = cpu_to_be32(
                 COMMAND_ORB_MAX_PAYLOAD(max_payload) |
                 COMMAND_ORB_SPEED(device->max_speed) |
-                COMMAND_ORB_NOTIFY;
+                COMMAND_ORB_NOTIFY);
 
         if (cmd->sc_data_direction == DMA_FROM_DEVICE)
-                orb->request.misc |=
-                        COMMAND_ORB_DIRECTION(SBP2_DIRECTION_FROM_MEDIA);
-        else if (cmd->sc_data_direction == DMA_TO_DEVICE)
-                orb->request.misc |=
-                        COMMAND_ORB_DIRECTION(SBP2_DIRECTION_TO_MEDIA);
+                orb->request.misc |= cpu_to_be32(COMMAND_ORB_DIRECTION);
 
         if (scsi_sg_count(cmd) && sbp2_map_scatterlist(orb, device, lu) < 0)
                 goto out;
 
-        fw_memcpy_to_be32(&orb->request, &orb->request, sizeof(orb->request));
-
-        memset(orb->request.command_block,
-               0, sizeof(orb->request.command_block));
         memcpy(orb->request.command_block, cmd->cmnd, COMMAND_SIZE(*cmd->cmnd));
 
         orb->base.callback = complete_command_orb;
@@ -1519,11 +1514,8 @@ static int sbp2_scsi_slave_alloc(struct scsi_device *sdev)
 
         sdev->allow_restart = 1;
 
-        /*
-         * Update the dma alignment (minimum alignment requirements for
-         * start and end of DMA transfers) to be a sector
-         */
-        blk_queue_update_dma_alignment(sdev->request_queue, 511);
+        /* SBP-2 requires quadlet alignment of the data buffers. */
+        blk_queue_update_dma_alignment(sdev->request_queue, 4 - 1);
 
         if (lu->tgt->workarounds & SBP2_WORKAROUND_INQUIRY_36)
                 sdev->inquiry_len = 36;
@@ -1581,16 +1573,14 @@ sbp2_sysfs_ieee1394_id_show(struct device *dev, struct device_attribute *attr,
 {
         struct scsi_device *sdev = to_scsi_device(dev);
         struct sbp2_logical_unit *lu;
-        struct fw_device *device;
 
         if (!sdev)
                 return 0;
 
         lu = sdev->hostdata;
-        device = fw_device(lu->tgt->unit->device.parent);
 
-        return sprintf(buf, "%08x%08x:%06x:%04x\n",
-                        device->config_rom[3], device->config_rom[4],
+        return sprintf(buf, "%016llx:%06x:%04x\n",
+                        (unsigned long long)lu->tgt->guid,
                         lu->tgt->directory_id, lu->lun);
 }
 
diff --git a/drivers/firewire/fw-topology.c b/drivers/firewire/fw-topology.c
index d2c7a3d7e1cb..213b0ff8f3d6 100644
--- a/drivers/firewire/fw-topology.c
+++ b/drivers/firewire/fw-topology.c
@@ -108,6 +108,7 @@ static struct fw_node *fw_node_create(u32 sid, int port_count, int color)
         node->node_id = LOCAL_BUS | SELF_ID_PHY_ID(sid);
         node->link_on = SELF_ID_LINK_ON(sid);
         node->phy_speed = SELF_ID_PHY_SPEED(sid);
+        node->initiated_reset = SELF_ID_PHY_INITIATOR(sid);
         node->port_count = port_count;
 
         atomic_set(&node->ref_count, 1);
@@ -289,12 +290,11 @@ static struct fw_node *build_tree(struct fw_card *card,
                         beta_repeaters_present = true;
 
                 /*
-                 * If all PHYs does not report the same gap count
-                 * setting, we fall back to 63 which will force a gap
-                 * count reconfiguration and a reset.
+                 * If PHYs report different gap counts, set an invalid count
+                 * which will force a gap count reconfiguration and a reset.
                  */
                 if (SELF_ID_GAP_COUNT(q) != gap_count)
-                        gap_count = 63;
+                        gap_count = 0;
 
                 update_hop_count(node);
 
@@ -431,6 +431,8 @@ update_tree(struct fw_card *card, struct fw_node *root)
                         event = FW_NODE_LINK_OFF;
                 else if (!node0->link_on && node1->link_on)
                         event = FW_NODE_LINK_ON;
+                else if (node1->initiated_reset && node1->link_on)
+                        event = FW_NODE_INITIATED_RESET;
                 else
                         event = FW_NODE_UPDATED;
 
diff --git a/drivers/firewire/fw-topology.h b/drivers/firewire/fw-topology.h
index cedc1ec906e9..addb9f8ea776 100644
--- a/drivers/firewire/fw-topology.h
+++ b/drivers/firewire/fw-topology.h
@@ -20,11 +20,12 @@
 #define __fw_topology_h
 
 enum {
-        FW_NODE_CREATED =   0x00,
-        FW_NODE_UPDATED =   0x01,
-        FW_NODE_DESTROYED = 0x02,
-        FW_NODE_LINK_ON =   0x03,
-        FW_NODE_LINK_OFF =  0x04,
+        FW_NODE_CREATED,
+        FW_NODE_UPDATED,
+        FW_NODE_DESTROYED,
+        FW_NODE_LINK_ON,
+        FW_NODE_LINK_OFF,
+        FW_NODE_INITIATED_RESET,
 };
 
 struct fw_node {
diff --git a/drivers/firewire/fw-transaction.c b/drivers/firewire/fw-transaction.c
index e6f1bda38940..3a59e9b783b0 100644
--- a/drivers/firewire/fw-transaction.c
+++ b/drivers/firewire/fw-transaction.c
@@ -18,6 +18,7 @@
  * Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
  */
 
+#include <linux/completion.h>
 #include <linux/kernel.h>
 #include <linux/module.h>
 #include <linux/init.h>
@@ -294,42 +295,40 @@ fw_send_request(struct fw_card *card, struct fw_transaction *t,
 }
 EXPORT_SYMBOL(fw_send_request);
 
+struct fw_phy_packet {
+        struct fw_packet packet;
+        struct completion done;
+};
+
 static void
 transmit_phy_packet_callback(struct fw_packet *packet,
                              struct fw_card *card, int status)
 {
-        kfree(packet);
-}
-
-static void send_phy_packet(struct fw_card *card, u32 data, int generation)
-{
-        struct fw_packet *packet;
-
-        packet = kzalloc(sizeof(*packet), GFP_ATOMIC);
-        if (packet == NULL)
-                return;
-
-        packet->header[0] = data;
-        packet->header[1] = ~data;
-        packet->header_length = 8;
-        packet->payload_length = 0;
-        packet->speed = SCODE_100;
-        packet->generation = generation;
-        packet->callback = transmit_phy_packet_callback;
+        struct fw_phy_packet *p =
+                        container_of(packet, struct fw_phy_packet, packet);
 
-        card->driver->send_request(card, packet);
+        complete(&p->done);
 }
 
 void fw_send_phy_config(struct fw_card *card,
                         int node_id, int generation, int gap_count)
 {
-        u32 q;
-
-        q = PHY_IDENTIFIER(PHY_PACKET_CONFIG) |
-                PHY_CONFIG_ROOT_ID(node_id) |
-                PHY_CONFIG_GAP_COUNT(gap_count);
-
-        send_phy_packet(card, q, generation);
+        struct fw_phy_packet p;
+        u32 data = PHY_IDENTIFIER(PHY_PACKET_CONFIG) |
+                   PHY_CONFIG_ROOT_ID(node_id) |
+                   PHY_CONFIG_GAP_COUNT(gap_count);
+
+        p.packet.header[0] = data;
+        p.packet.header[1] = ~data;
+        p.packet.header_length = 8;
+        p.packet.payload_length = 0;
+        p.packet.speed = SCODE_100;
+        p.packet.generation = generation;
+        p.packet.callback = transmit_phy_packet_callback;
+        init_completion(&p.done);
+
+        card->driver->send_request(card, &p.packet);
+        wait_for_completion(&p.done);
 }
 
 void fw_flush_transactions(struct fw_card *card)
@@ -389,21 +388,21 @@ lookup_enclosing_address_handler(struct list_head *list,
 static DEFINE_SPINLOCK(address_handler_lock);
 static LIST_HEAD(address_handler_list);
 
-const struct fw_address_region fw_low_memory_region =
-        { .start = 0x000000000000ULL, .end = 0x000100000000ULL,  };
 const struct fw_address_region fw_high_memory_region =
         { .start = 0x000100000000ULL, .end = 0xffffe0000000ULL,  };
+EXPORT_SYMBOL(fw_high_memory_region);
+
+#if 0
+const struct fw_address_region fw_low_memory_region =
+        { .start = 0x000000000000ULL, .end = 0x000100000000ULL,  };
 const struct fw_address_region fw_private_region =
         { .start = 0xffffe0000000ULL, .end = 0xfffff0000000ULL,  };
 const struct fw_address_region fw_csr_region =
-        { .start = 0xfffff0000000ULL, .end = 0xfffff0000800ULL,  };
+        { .start = CSR_REGISTER_BASE,
+          .end   = CSR_REGISTER_BASE | CSR_CONFIG_ROM_END,  };
 const struct fw_address_region fw_unit_space_region =
         { .start = 0xfffff0000900ULL, .end = 0x1000000000000ULL, };
-EXPORT_SYMBOL(fw_low_memory_region);
-EXPORT_SYMBOL(fw_high_memory_region);
-EXPORT_SYMBOL(fw_private_region);
-EXPORT_SYMBOL(fw_csr_region);
-EXPORT_SYMBOL(fw_unit_space_region);
+#endif  /*  0  */
 
 /**
  * Allocate a range of addresses in the node space of the OHCI
@@ -747,7 +746,8 @@ fw_core_handle_response(struct fw_card *card, struct fw_packet *p)
 EXPORT_SYMBOL(fw_core_handle_response);
 
 static const struct fw_address_region topology_map_region =
-        { .start = 0xfffff0001000ull, .end = 0xfffff0001400ull, };
+        { .start = CSR_REGISTER_BASE | CSR_TOPOLOGY_MAP,
+          .end   = CSR_REGISTER_BASE | CSR_TOPOLOGY_MAP_END, };
 
 static void
 handle_topology_map(struct fw_card *card, struct fw_request *request,
@@ -785,7 +785,8 @@ static struct fw_address_handler topology_map = {
 };
 
 static const struct fw_address_region registers_region =
-        { .start = 0xfffff0000000ull, .end = 0xfffff0000400ull, };
+        { .start = CSR_REGISTER_BASE,
+          .end   = CSR_REGISTER_BASE | CSR_CONFIG_ROM, };
 
 static void
 handle_registers(struct fw_card *card, struct fw_request *request,
@@ -794,7 +795,7 @@ handle_registers(struct fw_card *card, struct fw_request *request,
                  unsigned long long offset,
                  void *payload, size_t length, void *callback_data)
 {
-        int reg = offset - CSR_REGISTER_BASE;
+        int reg = offset & ~CSR_REGISTER_BASE;
         unsigned long long bus_time;
         __be32 *data = payload;
 
diff --git a/drivers/firewire/fw-transaction.h b/drivers/firewire/fw-transaction.h
index a43bb22912f9..04d3854f6560 100644
--- a/drivers/firewire/fw-transaction.h
+++ b/drivers/firewire/fw-transaction.h
@@ -201,11 +201,7 @@ struct fw_address_region {
         u64 end;
 };
 
-extern const struct fw_address_region fw_low_memory_region;
 extern const struct fw_address_region fw_high_memory_region;
-extern const struct fw_address_region fw_private_region;
-extern const struct fw_address_region fw_csr_region;
-extern const struct fw_address_region fw_unit_space_region;
 
 int fw_core_add_address_handler(struct fw_address_handler *handler,
                                 const struct fw_address_region *region);
@@ -221,12 +217,9 @@ struct fw_card {
         const struct fw_card_driver *driver;
         struct device *device;
         atomic_t device_count;
-        struct kref kref;
 
         int node_id;
         int generation;
-        /* This is the generation used for timestamping incoming requests. */
-        int request_generation;
         int current_tlabel, tlabel_mask;
         struct list_head transaction_list;
         struct timer_list flush_timer;
@@ -263,9 +256,6 @@ struct fw_card {
         int bm_generation;
 };
 
-struct fw_card *fw_card_get(struct fw_card *card);
-void fw_card_put(struct fw_card *card);
-
 /*
  * The iso packet format allows for an immediate header/payload part
  * stored in 'header' immediately after the packet info plus an