[PATCH] sched: TASK_NONINTERACTIVE
[pandora-kernel.git] / fs / buffer.c
index 0befa72..6cbfcea 100644 (file)
@@ -40,6 +40,7 @@
 #include <linux/cpu.h>
 #include <linux/bitops.h>
 #include <linux/mpage.h>
+#include <linux/bit_spinlock.h>
 
 static int fsync_buffers_list(spinlock_t *lock, struct list_head *list);
 static void invalidate_bh_lrus(void);
@@ -278,7 +279,7 @@ EXPORT_SYMBOL(thaw_bdev);
  */
 static void do_sync(unsigned long wait)
 {
-       wakeup_bdflush(0);
+       wakeup_pdflush(0);
        sync_inodes(0);         /* All mappings, inodes and their blockdevs */
        DQUOT_SYNC(NULL);
        sync_supers();          /* Write the superblocks */
@@ -331,7 +332,7 @@ int file_fsync(struct file *filp, struct dentry *dentry, int datasync)
        return ret;
 }
 
-asmlinkage long sys_fsync(unsigned int fd)
+static long do_fsync(unsigned int fd, int datasync)
 {
        struct file * file;
        struct address_space *mapping;
@@ -342,14 +343,14 @@ asmlinkage long sys_fsync(unsigned int fd)
        if (!file)
                goto out;
 
-       mapping = file->f_mapping;
-
        ret = -EINVAL;
        if (!file->f_op || !file->f_op->fsync) {
                /* Why?  We can still call filemap_fdatawrite */
                goto out_putf;
        }
 
+       mapping = file->f_mapping;
+
        current->flags |= PF_SYNCWRITE;
        ret = filemap_fdatawrite(mapping);
 
@@ -358,7 +359,7 @@ asmlinkage long sys_fsync(unsigned int fd)
         * which could cause livelocks in fsync_buffers_list
         */
        down(&mapping->host->i_sem);
-       err = file->f_op->fsync(file, file->f_dentry, 0);
+       err = file->f_op->fsync(file, file->f_dentry, datasync);
        if (!ret)
                ret = err;
        up(&mapping->host->i_sem);
@@ -373,39 +374,14 @@ out:
        return ret;
 }
 
-asmlinkage long sys_fdatasync(unsigned int fd)
+asmlinkage long sys_fsync(unsigned int fd)
 {
-       struct file * file;
-       struct address_space *mapping;
-       int ret, err;
-
-       ret = -EBADF;
-       file = fget(fd);
-       if (!file)
-               goto out;
-
-       ret = -EINVAL;
-       if (!file->f_op || !file->f_op->fsync)
-               goto out_putf;
-
-       mapping = file->f_mapping;
-
-       current->flags |= PF_SYNCWRITE;
-       ret = filemap_fdatawrite(mapping);
-       down(&mapping->host->i_sem);
-       err = file->f_op->fsync(file, file->f_dentry, 1);
-       if (!ret)
-               ret = err;
-       up(&mapping->host->i_sem);
-       err = filemap_fdatawait(mapping);
-       if (!ret)
-               ret = err;
-       current->flags &= ~PF_SYNCWRITE;
+       return do_fsync(fd, 0);
+}
 
-out_putf:
-       fput(file);
-out:
-       return ret;
+asmlinkage long sys_fdatasync(unsigned int fd)
+{
+       return do_fsync(fd, 1);
 }
 
 /*
@@ -522,7 +498,7 @@ static void free_more_memory(void)
        struct zone **zones;
        pg_data_t *pgdat;
 
-       wakeup_bdflush(1024);
+       wakeup_pdflush(1024);
        yield();
 
        for_each_pgdat(pgdat) {
@@ -538,8 +514,8 @@ static void free_more_memory(void)
  */
 static void end_buffer_async_read(struct buffer_head *bh, int uptodate)
 {
-       static DEFINE_SPINLOCK(page_uptodate_lock);
        unsigned long flags;
+       struct buffer_head *first;
        struct buffer_head *tmp;
        struct page *page;
        int page_uptodate = 1;
@@ -561,7 +537,9 @@ static void end_buffer_async_read(struct buffer_head *bh, int uptodate)
         * two buffer heads end IO at almost the same time and both
         * decide that the page is now completely done.
         */
-       spin_lock_irqsave(&page_uptodate_lock, flags);
+       first = page_buffers(page);
+       local_irq_save(flags);
+       bit_spin_lock(BH_Uptodate_Lock, &first->b_state);
        clear_buffer_async_read(bh);
        unlock_buffer(bh);
        tmp = bh;
@@ -574,7 +552,8 @@ static void end_buffer_async_read(struct buffer_head *bh, int uptodate)
                }
                tmp = tmp->b_this_page;
        } while (tmp != bh);
-       spin_unlock_irqrestore(&page_uptodate_lock, flags);
+       bit_spin_unlock(BH_Uptodate_Lock, &first->b_state);
+       local_irq_restore(flags);
 
        /*
         * If none of the buffers had errors and they are all
@@ -586,7 +565,8 @@ static void end_buffer_async_read(struct buffer_head *bh, int uptodate)
        return;
 
 still_busy:
-       spin_unlock_irqrestore(&page_uptodate_lock, flags);
+       bit_spin_unlock(BH_Uptodate_Lock, &first->b_state);
+       local_irq_restore(flags);
        return;
 }
 
@@ -597,8 +577,8 @@ still_busy:
 void end_buffer_async_write(struct buffer_head *bh, int uptodate)
 {
        char b[BDEVNAME_SIZE];
-       static DEFINE_SPINLOCK(page_uptodate_lock);
        unsigned long flags;
+       struct buffer_head *first;
        struct buffer_head *tmp;
        struct page *page;
 
@@ -619,7 +599,10 @@ void end_buffer_async_write(struct buffer_head *bh, int uptodate)
                SetPageError(page);
        }
 
-       spin_lock_irqsave(&page_uptodate_lock, flags);
+       first = page_buffers(page);
+       local_irq_save(flags);
+       bit_spin_lock(BH_Uptodate_Lock, &first->b_state);
+
        clear_buffer_async_write(bh);
        unlock_buffer(bh);
        tmp = bh->b_this_page;
@@ -630,12 +613,14 @@ void end_buffer_async_write(struct buffer_head *bh, int uptodate)
                }
                tmp = tmp->b_this_page;
        }
-       spin_unlock_irqrestore(&page_uptodate_lock, flags);
+       bit_spin_unlock(BH_Uptodate_Lock, &first->b_state);
+       local_irq_restore(flags);
        end_page_writeback(page);
        return;
 
 still_busy:
-       spin_unlock_irqrestore(&page_uptodate_lock, flags);
+       bit_spin_unlock(BH_Uptodate_Lock, &first->b_state);
+       local_irq_restore(flags);
        return;
 }
 
@@ -933,8 +918,7 @@ static int fsync_buffers_list(spinlock_t *lock, struct list_head *list)
                                 * contents - it is a noop if I/O is still in
                                 * flight on potentially older contents.
                                 */
-                               wait_on_buffer(bh);
-                               ll_rw_block(WRITE, 1, &bh);
+                               ll_rw_block(SWRITE, 1, &bh);
                                brelse(bh);
                                spin_lock(lock);
                        }
@@ -1951,7 +1935,6 @@ static int __block_prepare_write(struct inode *inode, struct page *page,
                        if (err)
                                break;
                        if (buffer_new(bh)) {
-                               clear_buffer_new(bh);
                                unmap_underlying_metadata(bh->b_bdev,
                                                        bh->b_blocknr);
                                if (PageUptodate(page)) {
@@ -1993,9 +1976,14 @@ static int __block_prepare_write(struct inode *inode, struct page *page,
                if (!buffer_uptodate(*wait_bh))
                        err = -EIO;
        }
-       if (!err)
-               return err;
-
+       if (!err) {
+               bh = head;
+               do {
+                       if (buffer_new(bh))
+                               clear_buffer_new(bh);
+               } while ((bh = bh->b_this_page) != head);
+               return 0;
+       }
        /* Error case: */
        /*
         * Zero out any newly allocated blocks to avoid exposing stale
@@ -2805,21 +2793,22 @@ int submit_bh(int rw, struct buffer_head * bh)
 
 /**
  * ll_rw_block: low-level access to block devices (DEPRECATED)
- * @rw: whether to %READ or %WRITE or maybe %READA (readahead)
+ * @rw: whether to %READ or %WRITE or %SWRITE or maybe %READA (readahead)
  * @nr: number of &struct buffer_heads in the array
  * @bhs: array of pointers to &struct buffer_head
  *
- * ll_rw_block() takes an array of pointers to &struct buffer_heads,
- * and requests an I/O operation on them, either a %READ or a %WRITE.
- * The third %READA option is described in the documentation for
- * generic_make_request() which ll_rw_block() calls.
+ * ll_rw_block() takes an array of pointers to &struct buffer_heads, and
+ * requests an I/O operation on them, either a %READ or a %WRITE.  The third
+ * %SWRITE is like %WRITE only we make sure that the *current* data in buffers
+ * are sent to disk. The fourth %READA option is described in the documentation
+ * for generic_make_request() which ll_rw_block() calls.
  *
  * This function drops any buffer that it cannot get a lock on (with the
- * BH_Lock state bit), any buffer that appears to be clean when doing a
- * write request, and any buffer that appears to be up-to-date when doing
- * read request.  Further it marks as clean buffers that are processed for
- * writing (the buffer cache won't assume that they are actually clean until
- * the buffer gets unlocked).
+ * BH_Lock state bit) unless SWRITE is required, any buffer that appears to be
+ * clean when doing a write request, and any buffer that appears to be
+ * up-to-date when doing read request.  Further it marks as clean buffers that
+ * are processed for writing (the buffer cache won't assume that they are
+ * actually clean until the buffer gets unlocked).
  *
  * ll_rw_block sets b_end_io to simple completion handler that marks
  * the buffer up-to-date (if approriate), unlocks the buffer and wakes
@@ -2835,11 +2824,13 @@ void ll_rw_block(int rw, int nr, struct buffer_head *bhs[])
        for (i = 0; i < nr; i++) {
                struct buffer_head *bh = bhs[i];
 
-               if (test_set_buffer_locked(bh))
+               if (rw == SWRITE)
+                       lock_buffer(bh);
+               else if (test_set_buffer_locked(bh))
                        continue;
 
                get_bh(bh);
-               if (rw == WRITE) {
+               if (rw == WRITE || rw == SWRITE) {
                        if (test_clear_buffer_dirty(bh)) {
                                bh->b_end_io = end_buffer_write_sync;
                                submit_bh(WRITE, bh);
@@ -3058,10 +3049,9 @@ struct buffer_head *alloc_buffer_head(unsigned int __nocast gfp_flags)
 {
        struct buffer_head *ret = kmem_cache_alloc(bh_cachep, gfp_flags);
        if (ret) {
-               preempt_disable();
-               __get_cpu_var(bh_accounting).nr++;
+               get_cpu_var(bh_accounting).nr++;
                recalc_bh_state();
-               preempt_enable();
+               put_cpu_var(bh_accounting);
        }
        return ret;
 }
@@ -3071,10 +3061,9 @@ void free_buffer_head(struct buffer_head *bh)
 {
        BUG_ON(!list_empty(&bh->b_assoc_buffers));
        kmem_cache_free(bh_cachep, bh);
-       preempt_disable();
-       __get_cpu_var(bh_accounting).nr--;
+       get_cpu_var(bh_accounting).nr--;
        recalc_bh_state();
-       preempt_enable();
+       put_cpu_var(bh_accounting);
 }
 EXPORT_SYMBOL(free_buffer_head);