5 files changed, 244 insertions, 65 deletions

diff --git a/block/elevator.c b/block/elevator.c
index 39dcccc82ada..c9f424d5399c 100644
--- a/block/elevator.c
+++ b/block/elevator.c
@@ -64,7 +64,7 @@ inline int elv_rq_merge_ok(struct request *rq, struct bio *bio)
 }
 EXPORT_SYMBOL(elv_rq_merge_ok);
 
-inline int elv_try_merge(struct request *__rq, struct bio *bio)
+static inline int elv_try_merge(struct request *__rq, struct bio *bio)
 {
         int ret = ELEVATOR_NO_MERGE;
 
@@ -80,7 +80,6 @@ inline int elv_try_merge(struct request *__rq, struct bio *bio)
 
         return ret;
 }
-EXPORT_SYMBOL(elv_try_merge);
 
 static struct elevator_type *elevator_find(const char *name)
 {
@@ -150,13 +149,20 @@ static void elevator_setup_default(void)
         if (!chosen_elevator[0])
                 strcpy(chosen_elevator, CONFIG_DEFAULT_IOSCHED);
 
+        /*
+         * Be backwards-compatible with previous kernels, so users
+         * won't get the wrong elevator.
+         */
+        if (!strcmp(chosen_elevator, "as"))
+                strcpy(chosen_elevator, "anticipatory");
+
         /*
-         * If the given scheduler is not available, fall back to no-op.
+         * If the given scheduler is not available, fall back to the default
          */
         if ((e = elevator_find(chosen_elevator)))
                 elevator_put(e);
         else
-                strcpy(chosen_elevator, "noop");
+                strcpy(chosen_elevator, CONFIG_DEFAULT_IOSCHED);
 }
 
 static int __init elevator_setup(char *str)
@@ -611,23 +617,23 @@ void elv_completed_request(request_queue_t *q, struct request *rq)
          * request is released from the driver, io must be done
          */
         if (blk_account_rq(rq)) {
-                struct request *first_rq = list_entry_rq(q->queue_head.next);
-
                 q->in_flight--;
+                if (blk_sorted_rq(rq) && e->ops->elevator_completed_req_fn)
+                        e->ops->elevator_completed_req_fn(q, rq);
+        }
 
-                /*
-                 * Check if the queue is waiting for fs requests to be
-                 * drained for flush sequence.
-                 */
-                if (q->ordseq && q->in_flight == 0 &&
+        /*
+         * Check if the queue is waiting for fs requests to be
+         * drained for flush sequence.
+         */
+        if (unlikely(q->ordseq)) {
+                struct request *first_rq = list_entry_rq(q->queue_head.next);
+                if (q->in_flight == 0 &&
                     blk_ordered_cur_seq(q) == QUEUE_ORDSEQ_DRAIN &&
                     blk_ordered_req_seq(first_rq) > QUEUE_ORDSEQ_DRAIN) {
                         blk_ordered_complete_seq(q, QUEUE_ORDSEQ_DRAIN, 0);
                         q->request_fn(q);
                 }
-
-                if (blk_sorted_rq(rq) && e->ops->elevator_completed_req_fn)
-                        e->ops->elevator_completed_req_fn(q, rq);
         }
 }
 
diff --git a/block/genhd.c b/block/genhd.c
index f1ed83f3f083..db57546a709d 100644
--- a/block/genhd.c
+++ b/block/genhd.c
@@ -38,34 +38,100 @@ static inline int major_to_index(int major)
         return major % MAX_PROBE_HASH;
 }
 
-#ifdef CONFIG_PROC_FS
-/* get block device names in somewhat random order */
-int get_blkdev_list(char *p, int used)
+struct blkdev_info {
+        int index;
+        struct blk_major_name *bd;
+};
+
+/*
+ * iterate over a list of blkdev_info structures.  allows
+ * the major_names array to be iterated over from outside this file
+ * must be called with the block_subsys_sem held
+ */
+void *get_next_blkdev(void *dev)
+{
+        struct blkdev_info *info;
+
+        if (dev == NULL) {
+                info = kmalloc(sizeof(*info), GFP_KERNEL);
+                if (!info)
+                        goto out;
+                info->index=0;
+                info->bd = major_names[info->index];
+                if (info->bd)
+                        goto out;
+        } else {
+                info = dev;
+        }
+
+        while (info->index < ARRAY_SIZE(major_names)) {
+                if (info->bd)
+                        info->bd = info->bd->next;
+                if (info->bd)
+                        goto out;
+                /*
+                 * No devices on this chain, move to the next
+                 */
+                info->index++;
+                info->bd = (info->index < ARRAY_SIZE(major_names)) ?
+                        major_names[info->index] : NULL;
+                if (info->bd)
+                        goto out;
+        }
+
+out:
+        return info;
+}
+
+void *acquire_blkdev_list(void)
+{
+        down(&block_subsys_sem);
+        return get_next_blkdev(NULL);
+}
+
+void release_blkdev_list(void *dev)
+{
+        up(&block_subsys_sem);
+        kfree(dev);
+}
+
+
+/*
+ * Count the number of records in the blkdev_list.
+ * must be called with the block_subsys_sem held
+ */
+int count_blkdev_list(void)
 {
         struct blk_major_name *n;
-        int i, len;
+        int i, count;
 
-        len = snprintf(p, (PAGE_SIZE-used), "\nBlock devices:\n");
+        count = 0;
 
-        down(&block_subsys_sem);
         for (i = 0; i < ARRAY_SIZE(major_names); i++) {
-                for (n = major_names[i]; n; n = n->next) {
-                        /*
-                         * If the curent string plus the 5 extra characters
-                         * in the line would run us off the page, then we're done
-                         */
-                        if ((len + used + strlen(n->name) + 5) >= PAGE_SIZE)
-                                goto page_full;
-                        len += sprintf(p+len, "%3d %s\n",
-                                       n->major, n->name);
-                }
+                for (n = major_names[i]; n; n = n->next)
+                                count++;
         }
-page_full:
-        up(&block_subsys_sem);
 
-        return len;
+        return count;
 }
-#endif
+
+/*
+ * extract the major and name values from a blkdev_info struct
+ * passed in as a void to *dev.  Must be called with
+ * block_subsys_sem held
+ */
+int get_blkdev_info(void *dev, int *major, char **name)
+{
+        struct blkdev_info *info = dev;
+
+        if (info->bd == NULL)
+                return 1;
+
+        *major = info->bd->major;
+        *name = info->bd->name;
+        return 0;
+}
+
 
 int register_blkdev(unsigned int major, const char *name)
 {
diff --git a/block/ioctl.c b/block/ioctl.c
index 6e278474f9a8..e1109491c234 100644
--- a/block/ioctl.c
+++ b/block/ioctl.c
@@ -1,6 +1,7 @@
-#include <linux/sched.h>                /* for capable() */
+#include <linux/capability.h>
 #include <linux/blkdev.h>
 #include <linux/blkpg.h>
+#include <linux/hdreg.h>
 #include <linux/backing-dev.h>
 #include <linux/buffer_head.h>
 #include <linux/smp_lock.h>
@@ -245,6 +246,27 @@ int blkdev_ioctl(struct inode *inode, struct file *file, unsigned cmd,
                 set_device_ro(bdev, n);
                 unlock_kernel();
                 return 0;
+        case HDIO_GETGEO: {
+                struct hd_geometry geo;
+
+                if (!arg)
+                        return -EINVAL;
+                if (!disk->fops->getgeo)
+                        return -ENOTTY;
+
+                /*
+                 * We need to set the startsect first, the driver may
+                 * want to override it.
+                 */
+                geo.start = get_start_sect(bdev);
+                ret = disk->fops->getgeo(bdev, &geo);
+                if (ret)
+                        return ret;
+                if (copy_to_user((struct hd_geometry __user *)arg, &geo,
+                                        sizeof(geo)))
+                        return -EFAULT;
+                return 0;
+        }
         }
 
         lock_kernel();
diff --git a/block/ll_rw_blk.c b/block/ll_rw_blk.c
index 91d3b4828c49..8e27d0ab0d7c 100644
--- a/block/ll_rw_blk.c
+++ b/block/ll_rw_blk.c
@@ -26,7 +26,8 @@
 #include <linux/slab.h>
 #include <linux/swap.h>
 #include <linux/writeback.h>
-#include <linux/blkdev.h>
+#include <linux/interrupt.h>
+#include <linux/cpu.h>
 
 /*
  * for max sense size
@@ -62,13 +63,15 @@ static wait_queue_head_t congestion_wqh[2] = {
 /*
  * Controlling structure to kblockd
  */
-static struct workqueue_struct *kblockd_workqueue; 
+static struct workqueue_struct *kblockd_workqueue;
 
 unsigned long blk_max_low_pfn, blk_max_pfn;
 
 EXPORT_SYMBOL(blk_max_low_pfn);
 EXPORT_SYMBOL(blk_max_pfn);
 
+static DEFINE_PER_CPU(struct list_head, blk_cpu_done);
+
 /* Amount of time in which a process may batch requests */
 #define BLK_BATCH_TIME  (HZ/50UL)
 
@@ -207,6 +210,13 @@ void blk_queue_merge_bvec(request_queue_t *q, merge_bvec_fn *mbfn)
 
 EXPORT_SYMBOL(blk_queue_merge_bvec);
 
+void blk_queue_softirq_done(request_queue_t *q, softirq_done_fn *fn)
+{
+        q->softirq_done_fn = fn;
+}
+
+EXPORT_SYMBOL(blk_queue_softirq_done);
+
 /**
  * blk_queue_make_request - define an alternate make_request function for a device
  * @q:  the request queue for the device to be affected
@@ -270,6 +280,7 @@ EXPORT_SYMBOL(blk_queue_make_request);
 static inline void rq_init(request_queue_t *q, struct request *rq)
 {
         INIT_LIST_HEAD(&rq->queuelist);
+        INIT_LIST_HEAD(&rq->donelist);
 
         rq->errors = 0;
         rq->rq_status = RQ_ACTIVE;
@@ -286,6 +297,7 @@ static inline void rq_init(request_queue_t *q, struct request *rq)
         rq->sense = NULL;
         rq->end_io = NULL;
         rq->end_io_data = NULL;
+        rq->completion_data = NULL;
 }
 
 /**
@@ -2735,30 +2747,6 @@ static inline int attempt_front_merge(request_queue_t *q, struct request *rq)
         return 0;
 }
 
-/**
- * blk_attempt_remerge  - attempt to remerge active head with next request
- * @q:    The &request_queue_t belonging to the device
- * @rq:   The head request (usually)
- *
- * Description:
- *    For head-active devices, the queue can easily be unplugged so quickly
- *    that proper merging is not done on the front request. This may hurt
- *    performance greatly for some devices. The block layer cannot safely
- *    do merging on that first request for these queues, but the driver can
- *    call this function and make it happen any way. Only the driver knows
- *    when it is safe to do so.
- **/
-void blk_attempt_remerge(request_queue_t *q, struct request *rq)
-{
-        unsigned long flags;
-
-        spin_lock_irqsave(q->queue_lock, flags);
-        attempt_back_merge(q, rq);
-        spin_unlock_irqrestore(q->queue_lock, flags);
-}
-
-EXPORT_SYMBOL(blk_attempt_remerge);
-
 static void init_request_from_bio(struct request *req, struct bio *bio)
 {
         req->flags |= REQ_CMD;
@@ -3287,6 +3275,87 @@ int end_that_request_chunk(struct request *req, int uptodate, int nr_bytes)
 EXPORT_SYMBOL(end_that_request_chunk);
 
 /*
+ * splice the completion data to a local structure and hand off to
+ * process_completion_queue() to complete the requests
+ */
+static void blk_done_softirq(struct softirq_action *h)
+{
+        struct list_head *cpu_list;
+        LIST_HEAD(local_list);
+
+        local_irq_disable();
+        cpu_list = &__get_cpu_var(blk_cpu_done);
+        list_splice_init(cpu_list, &local_list);
+        local_irq_enable();
+
+        while (!list_empty(&local_list)) {
+                struct request *rq = list_entry(local_list.next, struct request, donelist);
+
+                list_del_init(&rq->donelist);
+                rq->q->softirq_done_fn(rq);
+        }
+}
+
+#ifdef CONFIG_HOTPLUG_CPU
+
+static int blk_cpu_notify(struct notifier_block *self, unsigned long action,
+                          void *hcpu)
+{
+        /*
+         * If a CPU goes away, splice its entries to the current CPU
+         * and trigger a run of the softirq
+         */
+        if (action == CPU_DEAD) {
+                int cpu = (unsigned long) hcpu;
+
+                local_irq_disable();
+                list_splice_init(&per_cpu(blk_cpu_done, cpu),
+                                 &__get_cpu_var(blk_cpu_done));
+                raise_softirq_irqoff(BLOCK_SOFTIRQ);
+                local_irq_enable();
+        }
+
+        return NOTIFY_OK;
+}
+
+
+static struct notifier_block __devinitdata blk_cpu_notifier = {
+        .notifier_call  = blk_cpu_notify,
+};
+
+#endif /* CONFIG_HOTPLUG_CPU */
+
+/**
+ * blk_complete_request - end I/O on a request
+ * @req:      the request being processed
+ *
+ * Description:
+ *     Ends all I/O on a request. It does not handle partial completions,
+ *     unless the driver actually implements this in its completionc callback
+ *     through requeueing. Theh actual completion happens out-of-order,
+ *     through a softirq handler. The user must have registered a completion
+ *     callback through blk_queue_softirq_done().
+ **/
+
+void blk_complete_request(struct request *req)
+{
+        struct list_head *cpu_list;
+        unsigned long flags;
+
+        BUG_ON(!req->q->softirq_done_fn);
+                
+        local_irq_save(flags);
+
+        cpu_list = &__get_cpu_var(blk_cpu_done);
+        list_add_tail(&req->donelist, cpu_list);
+        raise_softirq_irqoff(BLOCK_SOFTIRQ);
+
+        local_irq_restore(flags);
+}
+
+EXPORT_SYMBOL(blk_complete_request);
+        
+/*
  * queue lock must be held
  */
 void end_that_request_last(struct request *req, int uptodate)
@@ -3364,6 +3433,8 @@ EXPORT_SYMBOL(kblockd_flush);
 
 int __init blk_dev_init(void)
 {
+        int i;
+
         kblockd_workqueue = create_workqueue("kblockd");
         if (!kblockd_workqueue)
                 panic("Failed to create kblockd\n");
@@ -3377,6 +3448,14 @@ int __init blk_dev_init(void)
         iocontext_cachep = kmem_cache_create("blkdev_ioc",
                         sizeof(struct io_context), 0, SLAB_PANIC, NULL, NULL);
 
+        for (i = 0; i < NR_CPUS; i++)
+                INIT_LIST_HEAD(&per_cpu(blk_cpu_done, i));
+
+        open_softirq(BLOCK_SOFTIRQ, blk_done_softirq, NULL);
+#ifdef CONFIG_HOTPLUG_CPU
+        register_cpu_notifier(&blk_cpu_notifier);
+#endif
+
         blk_max_low_pfn = max_low_pfn;
         blk_max_pfn = max_pfn;
 
diff --git a/block/scsi_ioctl.c b/block/scsi_ioctl.c
index c2ac36dfe4f3..cc72210687eb 100644
--- a/block/scsi_ioctl.c
+++ b/block/scsi_ioctl.c
@@ -21,6 +21,7 @@
 #include <linux/string.h>
 #include <linux/module.h>
 #include <linux/blkdev.h>
+#include <linux/capability.h>
 #include <linux/completion.h>
 #include <linux/cdrom.h>
 #include <linux/slab.h>
@@ -190,16 +191,21 @@ static int verify_command(struct file *file, unsigned char *cmd)
                 safe_for_write(GPCMD_SET_STREAMING),
         };
         unsigned char type = cmd_type[cmd[0]];
+        int has_write_perm = 0;
 
         /* Anybody who can open the device can do a read-safe command */
         if (type & CMD_READ_SAFE)
                 return 0;
 
+        /*
+         * file can be NULL from ioctl_by_bdev()...
+         */
+        if (file)
+                has_write_perm = file->f_mode & FMODE_WRITE;
+
         /* Write-safe commands just require a writable open.. */
-        if (type & CMD_WRITE_SAFE) {
-                if (file->f_mode & FMODE_WRITE)
-                        return 0;
-        }
+        if ((type & CMD_WRITE_SAFE) && has_write_perm)
+                return 0;
 
         /* And root can do any command.. */
         if (capable(CAP_SYS_RAWIO))