Merge branch 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6

Conflicts:

	Documentation/kernel-parameters.txt
	arch/sh/include/asm/elf.h

6 files changed, 117 insertions, 155 deletions

diff --git a/drivers/firewire/fw-card.c b/drivers/firewire/fw-card.c
index bbd73a406e5..418c18f07e9 100644
--- a/drivers/firewire/fw-card.c
+++ b/drivers/firewire/fw-card.c
@@ -189,39 +189,16 @@ static const char gap_count_table[] = {
 	63, 5, 7, 8, 10, 13, 16, 18, 21, 24, 26, 29, 32, 35, 37, 40
 };
 
-struct bm_data {
-	struct fw_transaction t;
-	struct {
-		__be32 arg;
-		__be32 data;
-	} lock;
-	u32 old;
-	int rcode;
-	struct completion done;
-};
-
-static void
-complete_bm_lock(struct fw_card *card, int rcode,
-		 void *payload, size_t length, void *data)
-{
-	struct bm_data *bmd = data;
-
-	if (rcode == RCODE_COMPLETE)
-		bmd->old = be32_to_cpu(*(__be32 *) payload);
-	bmd->rcode = rcode;
-	complete(&bmd->done);
-}
-
 static void
 fw_card_bm_work(struct work_struct *work)
 {
 	struct fw_card *card = container_of(work, struct fw_card, work.work);
 	struct fw_device *root_device;
 	struct fw_node *root_node, *local_node;
-	struct bm_data bmd;
 	unsigned long flags;
-	int root_id, new_root_id, irm_id, gap_count, generation, grace;
+	int root_id, new_root_id, irm_id, gap_count, generation, grace, rcode;
 	bool do_reset = false;
+	__be32 lock_data[2];
 
 	spin_lock_irqsave(&card->lock, flags);
 	local_node = card->local_node;
@@ -263,33 +240,28 @@ fw_card_bm_work(struct work_struct *work)
 			goto pick_me;
 		}
 
-		bmd.lock.arg = cpu_to_be32(0x3f);
-		bmd.lock.data = cpu_to_be32(local_node->node_id);
+		lock_data[0] = cpu_to_be32(0x3f);
+		lock_data[1] = cpu_to_be32(local_node->node_id);
 
 		spin_unlock_irqrestore(&card->lock, flags);
 
-		init_completion(&bmd.done);
-		fw_send_request(card, &bmd.t, TCODE_LOCK_COMPARE_SWAP,
-				irm_id, generation,
-				SCODE_100, CSR_REGISTER_BASE + CSR_BUS_MANAGER_ID,
-				&bmd.lock, sizeof(bmd.lock),
-				complete_bm_lock, &bmd);
-		wait_for_completion(&bmd.done);
+		rcode = fw_run_transaction(card, TCODE_LOCK_COMPARE_SWAP,
+				irm_id, generation, SCODE_100,
+				CSR_REGISTER_BASE + CSR_BUS_MANAGER_ID,
+				lock_data, sizeof(lock_data));
 
-		if (bmd.rcode == RCODE_GENERATION) {
-			/*
-			 * Another bus reset happened. Just return,
-			 * the BM work has been rescheduled.
-			 */
+		if (rcode == RCODE_GENERATION)
+			/* Another bus reset, BM work has been rescheduled. */
 			goto out;
-		}
 
-		if (bmd.rcode == RCODE_COMPLETE && bmd.old != 0x3f)
+		if (rcode == RCODE_COMPLETE &&
+		    lock_data[0] != cpu_to_be32(0x3f))
 			/* Somebody else is BM, let them do the work. */
 			goto out;
 
 		spin_lock_irqsave(&card->lock, flags);
-		if (bmd.rcode != RCODE_COMPLETE) {
+
+		if (rcode != RCODE_COMPLETE) {
 			/*
 			 * The lock request failed, maybe the IRM
 			 * isn't really IRM capable after all. Let's
diff --git a/drivers/firewire/fw-cdev.c b/drivers/firewire/fw-cdev.c
index 2e6d5848d21..ed03234cbea 100644
--- a/drivers/firewire/fw-cdev.c
+++ b/drivers/firewire/fw-cdev.c
@@ -720,8 +720,8 @@ static int ioctl_create_iso_context(struct client *client, void *buffer)
 #define GET_PAYLOAD_LENGTH(v)	((v) & 0xffff)
 #define GET_INTERRUPT(v)	(((v) >> 16) & 0x01)
 #define GET_SKIP(v)		(((v) >> 17) & 0x01)
-#define GET_TAG(v)		(((v) >> 18) & 0x02)
-#define GET_SY(v)		(((v) >> 20) & 0x04)
+#define GET_TAG(v)		(((v) >> 18) & 0x03)
+#define GET_SY(v)		(((v) >> 20) & 0x0f)
 #define GET_HEADER_LENGTH(v)	(((v) >> 24) & 0xff)
 
 static int ioctl_queue_iso(struct client *client, void *buffer)
@@ -913,7 +913,7 @@ dispatch_ioctl(struct client *client, unsigned int cmd, void __user *arg)
 			return -EFAULT;
 	}
 
-	return 0;
+	return retval;
 }
 
 static long
diff --git a/drivers/firewire/fw-device.c b/drivers/firewire/fw-device.c
index 0855fb5568e..3fccdd48410 100644
--- a/drivers/firewire/fw-device.c
+++ b/drivers/firewire/fw-device.c
@@ -381,46 +381,21 @@ static struct device_attribute fw_device_attributes[] = {
 	__ATTR_NULL,
 };
 
-struct read_quadlet_callback_data {
-	struct completion done;
-	int rcode;
-	u32 data;
-};
-
-static void
-complete_transaction(struct fw_card *card, int rcode,
-		     void *payload, size_t length, void *data)
-{
-	struct read_quadlet_callback_data *callback_data = data;
-
-	if (rcode == RCODE_COMPLETE)
-		callback_data->data = be32_to_cpu(*(__be32 *)payload);
-	callback_data->rcode = rcode;
-	complete(&callback_data->done);
-}
-
 static int
 read_rom(struct fw_device *device, int generation, int index, u32 *data)
 {
-	struct read_quadlet_callback_data callback_data;
-	struct fw_transaction t;
-	u64 offset;
+	int rcode;
 
 	/* device->node_id, accessed below, must not be older than generation */
 	smp_rmb();
 
-	init_completion(&callback_data.done);
-
-	offset = (CSR_REGISTER_BASE | CSR_CONFIG_ROM) + index * 4;
-	fw_send_request(device->card, &t, TCODE_READ_QUADLET_REQUEST,
+	rcode = fw_run_transaction(device->card, TCODE_READ_QUADLET_REQUEST,
 			device->node_id, generation, device->max_speed,
-			offset, NULL, 4, complete_transaction, &callback_data);
-
-	wait_for_completion(&callback_data.done);
-
-	*data = callback_data.data;
+			(CSR_REGISTER_BASE | CSR_CONFIG_ROM) + index * 4,
+			data, 4);
+	be32_to_cpus(data);
 
-	return callback_data.rcode;
+	return rcode;
 }
 
 #define READ_BIB_ROM_SIZE	256
diff --git a/drivers/firewire/fw-sbp2.c b/drivers/firewire/fw-sbp2.c
index aaff50ebba1..ef0b9b419c2 100644
--- a/drivers/firewire/fw-sbp2.c
+++ b/drivers/firewire/fw-sbp2.c
@@ -29,6 +29,7 @@
  */
 
 #include <linux/blkdev.h>
+#include <linux/bug.h>
 #include <linux/delay.h>
 #include <linux/device.h>
 #include <linux/dma-mapping.h>
@@ -181,10 +182,16 @@ struct sbp2_target {
 #define SBP2_MAX_LOGIN_ORB_TIMEOUT	40000U	/* Timeout in ms */
 #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_CYCLE_LIMIT		(0xc8 << 12)	/* 200 125us cycles */
 
+/*
+ * The default maximum s/g segment size of a FireWire controller is
+ * usually 0x10000, but SBP-2 only allows 0xffff. Since buffers have to
+ * be quadlet-aligned, we set the length limit to 0xffff & ~3.
+ */
+#define SBP2_MAX_SEG_SIZE		0xfffc
+
 /* Unit directory keys */
 #define SBP2_CSR_UNIT_CHARACTERISTICS	0x3a
 #define SBP2_CSR_FIRMWARE_REVISION	0x3c
@@ -621,25 +628,15 @@ sbp2_send_management_orb(struct sbp2_logical_unit *lu, int node_id,
 	return retval;
 }
 
-static void
-complete_agent_reset_write(struct fw_card *card, int rcode,
-			   void *payload, size_t length, void *done)
-{
-	complete(done);
-}
-
 static void sbp2_agent_reset(struct sbp2_logical_unit *lu)
 {
 	struct fw_device *device = fw_device(lu->tgt->unit->device.parent);
-	DECLARE_COMPLETION_ONSTACK(done);
-	struct fw_transaction t;
-	static u32 z;
+	__be32 d = 0;
 
-	fw_send_request(device->card, &t, TCODE_WRITE_QUADLET_REQUEST,
-			lu->tgt->node_id, lu->generation, device->max_speed,
-			lu->command_block_agent_address + SBP2_AGENT_RESET,
-			&z, sizeof(z), complete_agent_reset_write, &done);
-	wait_for_completion(&done);
+	fw_run_transaction(device->card, TCODE_WRITE_QUADLET_REQUEST,
+			   lu->tgt->node_id, lu->generation, device->max_speed,
+			   lu->command_block_agent_address + SBP2_AGENT_RESET,
+			   &d, sizeof(d));
 }
 
 static void
@@ -653,7 +650,7 @@ static void sbp2_agent_reset_no_wait(struct sbp2_logical_unit *lu)
 {
 	struct fw_device *device = fw_device(lu->tgt->unit->device.parent);
 	struct fw_transaction *t;
-	static u32 z;
+	static __be32 d;
 
 	t = kmalloc(sizeof(*t), GFP_ATOMIC);
 	if (t == NULL)
@@ -662,7 +659,7 @@ static void sbp2_agent_reset_no_wait(struct sbp2_logical_unit *lu)
 	fw_send_request(device->card, t, TCODE_WRITE_QUADLET_REQUEST,
 			lu->tgt->node_id, lu->generation, device->max_speed,
 			lu->command_block_agent_address + SBP2_AGENT_RESET,
-			&z, sizeof(z), complete_agent_reset_write_no_wait, t);
+			&d, sizeof(d), complete_agent_reset_write_no_wait, t);
 }
 
 static void sbp2_set_generation(struct sbp2_logical_unit *lu, int generation)
@@ -823,13 +820,6 @@ static void sbp2_target_put(struct sbp2_target *tgt)
 	kref_put(&tgt->kref, sbp2_release_target);
 }
 
-static void
-complete_set_busy_timeout(struct fw_card *card, int rcode,
-			  void *payload, size_t length, void *done)
-{
-	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
@@ -849,17 +839,12 @@ complete_set_busy_timeout(struct fw_card *card, int rcode,
 static void sbp2_set_busy_timeout(struct sbp2_logical_unit *lu)
 {
 	struct fw_device *device = fw_device(lu->tgt->unit->device.parent);
-	DECLARE_COMPLETION_ONSTACK(done);
-	struct fw_transaction t;
-	static __be32 busy_timeout;
-
-	busy_timeout = cpu_to_be32(SBP2_CYCLE_LIMIT | SBP2_RETRY_LIMIT);
+	__be32 d = 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,
-			CSR_REGISTER_BASE + CSR_BUSY_TIMEOUT, &busy_timeout,
-			sizeof(busy_timeout), complete_set_busy_timeout, &done);
-	wait_for_completion(&done);
+	fw_run_transaction(device->card, TCODE_WRITE_QUADLET_REQUEST,
+			   lu->tgt->node_id, lu->generation, device->max_speed,
+			   CSR_REGISTER_BASE + CSR_BUSY_TIMEOUT,
+			   &d, sizeof(d));
 }
 
 static void sbp2_reconnect(struct work_struct *work);
@@ -1121,6 +1106,10 @@ static int sbp2_probe(struct device *dev)
 	struct Scsi_Host *shost;
 	u32 model, firmware_revision;
 
+	if (dma_get_max_seg_size(device->card->device) > SBP2_MAX_SEG_SIZE)
+		BUG_ON(dma_set_max_seg_size(device->card->device,
+					    SBP2_MAX_SEG_SIZE));
+
 	shost = scsi_host_alloc(&scsi_driver_template, sizeof(*tgt));
 	if (shost == NULL)
 		return -ENOMEM;
@@ -1369,14 +1358,12 @@ static int
 sbp2_map_scatterlist(struct sbp2_command_orb *orb, struct fw_device *device,
 		     struct sbp2_logical_unit *lu)
 {
-	struct scatterlist *sg;
-	int sg_len, l, i, j, count;
-	dma_addr_t sg_addr;
-
-	sg = scsi_sglist(orb->cmd);
-	count = dma_map_sg(device->card->device, sg, scsi_sg_count(orb->cmd),
-			   orb->cmd->sc_data_direction);
-	if (count == 0)
+	struct scatterlist *sg = scsi_sglist(orb->cmd);
+	int i, n;
+
+	n = dma_map_sg(device->card->device, sg, scsi_sg_count(orb->cmd),
+		       orb->cmd->sc_data_direction);
+	if (n == 0)
 		goto fail;
 
 	/*
@@ -1386,7 +1373,7 @@ sbp2_map_scatterlist(struct sbp2_command_orb *orb, struct fw_device *device,
 	 * as the second generation iPod which doesn't support page
 	 * tables.
 	 */
-	if (count == 1 && sg_dma_len(sg) < SBP2_MAX_SG_ELEMENT_LENGTH) {
+	if (n == 1) {
 		orb->request.data_descriptor.high =
 			cpu_to_be32(lu->tgt->address_high);
 		orb->request.data_descriptor.low  =
@@ -1396,29 +1383,9 @@ sbp2_map_scatterlist(struct sbp2_command_orb *orb, struct fw_device *device,
 		return 0;
 	}
 
-	/*
-	 * Convert the scatterlist to an sbp2 page table.  If any
-	 * scatterlist entries are too big for sbp2, we split them as we
-	 * go.  Even if we ask the block I/O layer to not give us sg
-	 * elements larger than 65535 bytes, some IOMMUs may merge sg elements
-	 * during DMA mapping, and Linux currently doesn't prevent this.
-	 */
-	for (i = 0, j = 0; i < count; i++, sg = sg_next(sg)) {
-		sg_len = sg_dma_len(sg);
-		sg_addr = sg_dma_address(sg);
-		while (sg_len) {
-			/* FIXME: This won't get us out of the pinch. */
-			if (unlikely(j >= ARRAY_SIZE(orb->page_table))) {
-				fw_error("page table overflow\n");
-				goto fail_page_table;
-			}
-			l = min(sg_len, SBP2_MAX_SG_ELEMENT_LENGTH);
-			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++;
-		}
+	for_each_sg(sg, sg, n, i) {
+		orb->page_table[i].high = cpu_to_be32(sg_dma_len(sg) << 16);
+		orb->page_table[i].low = cpu_to_be32(sg_dma_address(sg));
 	}
 
 	orb->page_table_bus =
@@ -1437,13 +1404,13 @@ sbp2_map_scatterlist(struct sbp2_command_orb *orb, struct fw_device *device,
 	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));
+					 COMMAND_ORB_DATA_SIZE(n));
 
 	return 0;
 
  fail_page_table:
-	dma_unmap_sg(device->card->device, sg, scsi_sg_count(orb->cmd),
-		     orb->cmd->sc_data_direction);
+	dma_unmap_sg(device->card->device, scsi_sglist(orb->cmd),
+		     scsi_sg_count(orb->cmd), orb->cmd->sc_data_direction);
  fail:
 	return -ENOMEM;
 }
@@ -1456,7 +1423,7 @@ static int sbp2_scsi_queuecommand(struct scsi_cmnd *cmd, scsi_done_fn_t done)
 	struct fw_device *device = fw_device(lu->tgt->unit->device.parent);
 	struct sbp2_command_orb *orb;
 	unsigned int max_payload;
-	int retval = SCSI_MLQUEUE_HOST_BUSY;
+	int generation, retval = SCSI_MLQUEUE_HOST_BUSY;
 
 	/*
 	 * Bidirectional commands are not yet implemented, and unknown
@@ -1500,6 +1467,9 @@ static int sbp2_scsi_queuecommand(struct scsi_cmnd *cmd, scsi_done_fn_t done)
 	if (cmd->sc_data_direction == DMA_FROM_DEVICE)
 		orb->request.misc |= cpu_to_be32(COMMAND_ORB_DIRECTION);
 
+	generation = device->generation;
+	smp_rmb();    /* sbp2_map_scatterlist looks at tgt->address_high */
+
 	if (scsi_sg_count(cmd) && sbp2_map_scatterlist(orb, device, lu) < 0)
 		goto out;
 
@@ -1512,7 +1482,7 @@ static int sbp2_scsi_queuecommand(struct scsi_cmnd *cmd, scsi_done_fn_t done)
 	if (dma_mapping_error(device->card->device, orb->base.request_bus))
 		goto out;
 
-	sbp2_send_orb(&orb->base, lu, lu->tgt->node_id, lu->generation,
+	sbp2_send_orb(&orb->base, lu, lu->tgt->node_id, generation,
 		      lu->command_block_agent_address + SBP2_ORB_POINTER);
 	retval = 0;
  out:
@@ -1564,6 +1534,8 @@ static int sbp2_scsi_slave_configure(struct scsi_device *sdev)
 	if (lu->tgt->workarounds & SBP2_WORKAROUND_128K_MAX_TRANS)
 		blk_queue_max_sectors(sdev->request_queue, 128 * 1024 / 512);
 
+	blk_queue_max_segment_size(sdev->request_queue, SBP2_MAX_SEG_SIZE);
+
 	return 0;
 }
 
diff --git a/drivers/firewire/fw-transaction.c b/drivers/firewire/fw-transaction.c
index e5d1a0b64fc..022ac4fabb6 100644
--- a/drivers/firewire/fw-transaction.c
+++ b/drivers/firewire/fw-transaction.c
@@ -247,7 +247,7 @@ fw_fill_request(struct fw_packet *packet, int tcode, int tlabel,
  */
 void
 fw_send_request(struct fw_card *card, struct fw_transaction *t,
-		int tcode, int node_id, int generation, int speed,
+		int tcode, int destination_id, int generation, int speed,
 		unsigned long long offset,
 		void *payload, size_t length,
 		fw_transaction_callback_t callback, void *callback_data)
@@ -279,13 +279,14 @@ fw_send_request(struct fw_card *card, struct fw_transaction *t,
 	card->current_tlabel = (card->current_tlabel + 1) & 0x1f;
 	card->tlabel_mask |= (1 << tlabel);
 
-	t->node_id = node_id;
+	t->node_id = destination_id;
 	t->tlabel = tlabel;
 	t->callback = callback;
 	t->callback_data = callback_data;
 
-	fw_fill_request(&t->packet, tcode, t->tlabel, node_id, card->node_id,
-			generation, speed, offset, payload, length);
+	fw_fill_request(&t->packet, tcode, t->tlabel,
+			destination_id, card->node_id, generation,
+			speed, offset, payload, length);
 	t->packet.callback = transmit_complete_callback;
 
 	list_add_tail(&t->link, &card->transaction_list);
@@ -296,6 +297,45 @@ fw_send_request(struct fw_card *card, struct fw_transaction *t,
 }
 EXPORT_SYMBOL(fw_send_request);
 
+struct transaction_callback_data {
+	struct completion done;
+	void *payload;
+	int rcode;
+};
+
+static void transaction_callback(struct fw_card *card, int rcode,
+				 void *payload, size_t length, void *data)
+{
+	struct transaction_callback_data *d = data;
+
+	if (rcode == RCODE_COMPLETE)
+		memcpy(d->payload, payload, length);
+	d->rcode = rcode;
+	complete(&d->done);
+}
+
+/**
+ * fw_run_transaction - send request and sleep until transaction is completed
+ *
+ * Returns the RCODE.
+ */
+int fw_run_transaction(struct fw_card *card, int tcode, int destination_id,
+		int generation, int speed, unsigned long long offset,
+		void *data, size_t length)
+{
+	struct transaction_callback_data d;
+	struct fw_transaction t;
+
+	init_completion(&d.done);
+	d.payload = data;
+	fw_send_request(card, &t, tcode, destination_id, generation, speed,
+			offset, data, length, transaction_callback, &d);
+	wait_for_completion(&d.done);
+
+	return d.rcode;
+}
+EXPORT_SYMBOL(fw_run_transaction);
+
 static DEFINE_MUTEX(phy_config_mutex);
 static DECLARE_COMPLETION(phy_config_done);
 
diff --git a/drivers/firewire/fw-transaction.h b/drivers/firewire/fw-transaction.h
index 2ae1b0d6cb7..027f58ce81a 100644
--- a/drivers/firewire/fw-transaction.h
+++ b/drivers/firewire/fw-transaction.h
@@ -426,11 +426,14 @@ fw_core_initiate_bus_reset(struct fw_card *card, int short_reset);
 
 void
 fw_send_request(struct fw_card *card, struct fw_transaction *t,
-		int tcode, int node_id, int generation, int speed,
-		unsigned long long offset,
-		void *data, size_t length,
+		int tcode, int destination_id, int generation, int speed,
+		unsigned long long offset, void *data, size_t length,
 		fw_transaction_callback_t callback, void *callback_data);
 
+int fw_run_transaction(struct fw_card *card, int tcode, int destination_id,
+		       int generation, int speed, unsigned long long offset,
+		       void *data, size_t length);
+
 int fw_cancel_transaction(struct fw_card *card,
 			  struct fw_transaction *transaction);