[PATCH] readahead fix

Changes the way in which the readahead code locates the readahead
setting for the underlying device.

- struct block_device and struct address_space gain a *pointer* to the
  current readahead tunable.

- The tunable lives in the request queue and is altered with the
  traditional ioctl.

- The value gets *copied* into the struct file at open() time.  So a
  fcntl() mode to modify it per-fd is simple.

- Filesystems which are not request_queue-backed get the address of the
  global `default_ra_pages'.  If we want, this can become a tunable.

- Filesystems are at liberty to alter address_space.ra_pages to point
  at some other fs-private default at new_inode/read_inode/alloc_inode
  time.

- The ra_pages pointer can become a structure pointer if, at some time
  in the future, high-level code needs more detailed information about
  device characteristics.

  In fact, it'll need to become a struct pointer for use by
  writeback: my current writeback code has the problem that multiple
  pdflush threads can get stuck on the same request queue.  That's a
  waste of resources.  I currently have a silly flag in the superblock
  to try to avoid this.

  The proper way to get this exclusion is for the high-level
  writeback code to be able to do a test-and-set against a
  per-request_queue flag.  That flag can live in a structure alongside
  ra_pages, conveniently accessible at the pagemap level.

One thing still to-be-done is going into all callers of blk_init_queue
and blk_queue_make_request and making sure that they're setting up a
sensible default.  ATA wants 248 sectors, and floppy drives don't want
128kbytes, I suspect.  Later.

3 files changed, 6 insertions, 4 deletions

diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h
index 5cb20016dd99..01497e05a70a 100644
--- a/include/linux/blkdev.h
+++ b/include/linux/blkdev.h
@@ -150,9 +150,9 @@ struct request_queue
 
 	/*
 	 * The VM-level readahead tunable for this device.  In
-	 * units of 512-byte sectors.
+	 * units of PAGE_CACHE_SIZE pages.
 	 */
-	unsigned ra_sectors;
+	unsigned long ra_pages;
 
 	/*
 	 * The queue owner gets to use this for whatever they like.
@@ -310,8 +310,7 @@ extern void blk_queue_hardsect_size(request_queue_t *q, unsigned short);
 extern void blk_queue_segment_boundary(request_queue_t *q, unsigned long);
 extern void blk_queue_assign_lock(request_queue_t *q, spinlock_t *);
 extern void blk_queue_prep_rq(request_queue_t *q, prep_rq_fn *pfn);
-extern int blk_set_readahead(struct block_device *bdev, unsigned sectors);
-extern unsigned blk_get_readahead(struct block_device *bdev);
+extern unsigned long *blk_get_ra_pages(kdev_t kdev);
 
 extern int blk_rq_map_sg(request_queue_t *, struct request *, struct scatterlist *);
 extern void blk_dump_rq_flags(struct request *, char *);
diff --git a/include/linux/fs.h b/include/linux/fs.h
index 31e255227683..f0d997aeecb4 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -398,6 +398,7 @@ struct address_space {
 	list_t			i_mmap_shared;	/* list of private mappings */
 	spinlock_t		i_shared_lock;  /* and spinlock protecting it */
 	int			gfp_mask;	/* how to allocate the pages */
+	unsigned long 		*ra_pages;	/* device readahead */
 };
 
 struct char_device {
@@ -513,6 +514,7 @@ struct file_ra_state {
 	unsigned long prev_page;	/* Cache last read() position */
 	unsigned long ahead_start;	/* Ahead window */
 	unsigned long ahead_size;
+	unsigned long ra_pages;		/* Maximum readahead window */
 };
 
 struct file {
diff --git a/include/linux/mm.h b/include/linux/mm.h
index 56b5f325e0df..5f1c731ddde1 100644
--- a/include/linux/mm.h
+++ b/include/linux/mm.h
@@ -504,6 +504,7 @@ void do_page_cache_readahead(struct file *file,
 void page_cache_readahead(struct file *file, unsigned long offset);
 void page_cache_readaround(struct file *file, unsigned long offset);
 void handle_ra_thrashing(struct file *file);
+extern unsigned long default_ra_pages;
 
 /* vma is the first one with  address < vma->vm_end,
  * and even  address < vma->vm_start. Have to extend vma. */