[PATCH v2 3/4] brd: enable discard

[Mikulas Patocka <mpatocka@xxxxxxxxxx>]

This patch implements discard in the brd driver. We use RCU to free the
page, so that if there are any concurrent readers or writes, they won't
touch the page after it is freed.

Signed-off-by: Mikulas Patocka <mpatocka@xxxxxxxxxx>

---
 drivers/block/brd.c |   64 +++++++++++++++++++++++++++++++++++++++++++++++++++-
 1 file changed, 63 insertions(+), 1 deletion(-)

Index: linux-2.6/drivers/block/brd.c
===================================================================
--- linux-2.6.orig/drivers/block/brd.c
+++ linux-2.6/drivers/block/brd.c
@@ -48,6 +48,8 @@ struct brd_device {
 	u64			brd_nr_pages;
 };
 
+static bool discard;
+
 /*
  * Look up and return a brd's page for a given sector.
  * This must be called with the rcu lock held.
@@ -107,6 +109,25 @@ static bool brd_insert_page(struct brd_d
 	return true;
 }
 
+static void brd_free_page_rcu(struct rcu_head *head)
+{
+	struct page *page = container_of(head, struct page, rcu_head);
+	__free_page(page);
+}
+
+static void brd_free_page(struct brd_device *brd, sector_t sector)
+{
+	struct page *page;
+	pgoff_t idx;
+
+	spin_lock(&brd->brd_lock);
+	idx = sector >> PAGE_SECTORS_SHIFT;
+	page = radix_tree_delete(&brd->brd_pages, idx);
+	spin_unlock(&brd->brd_lock);
+	if (page)
+		call_rcu(&page->rcu_head, brd_free_page_rcu);
+}
+
 /*
  * Free all backing store pages and radix tree. This must only be called when
  * there are no other users of the device.
@@ -277,13 +298,44 @@ out:
 	return err;
 }
 
+void brd_do_discard(struct brd_device *brd, struct bio *bio)
+{
+	sector_t sector, len, front_pad;
+
+	if (unlikely(!discard)) {
+		bio->bi_status = BLK_STS_NOTSUPP;
+		return;
+	}
+
+	sector = bio->bi_iter.bi_sector;
+	len = bio_sectors(bio);
+	front_pad = -sector & (PAGE_SECTORS - 1);
+	sector += front_pad;
+	if (unlikely(len <= front_pad))
+		return;
+	len -= front_pad;
+	len = round_down(len, PAGE_SECTORS);
+	while (len) {
+		brd_free_page(brd, sector);
+		sector += PAGE_SECTORS;
+		len -= PAGE_SECTORS;
+		cond_resched();
+	}
+}
+
 static void brd_submit_bio(struct bio *bio)
 {
 	struct brd_device *brd = bio->bi_bdev->bd_disk->private_data;
-	sector_t sector = bio->bi_iter.bi_sector;
+	sector_t sector;
 	struct bio_vec bvec;
 	struct bvec_iter iter;
 
+	if (bio_op(bio) == REQ_OP_DISCARD) {
+		brd_do_discard(brd, bio);
+		goto endio;
+	}
+
+	sector = bio->bi_iter.bi_sector;
 	bio_for_each_segment(bvec, bio, iter) {
 		unsigned int len = bvec.bv_len;
 		int err;
@@ -301,6 +353,7 @@ static void brd_submit_bio(struct bio *b
 		sector += len >> SECTOR_SHIFT;
 	}
 
+endio:
 	bio_endio(bio);
 }
 
@@ -338,6 +391,10 @@ static int max_part = 1;
 module_param(max_part, int, 0444);
 MODULE_PARM_DESC(max_part, "Num Minors to reserve between devices");
 
+static bool discard = false;
+module_param(discard, bool, 0444);
+MODULE_PARM_DESC(discard, "Support discard");
+
 MODULE_LICENSE("GPL");
 MODULE_ALIAS_BLOCKDEV_MAJOR(RAMDISK_MAJOR);
 MODULE_ALIAS("rd");
@@ -404,6 +461,11 @@ static int brd_alloc(int i)
 	 */
 	blk_queue_physical_block_size(disk->queue, PAGE_SIZE);
 
+	if (discard) {
+		disk->queue->limits.discard_granularity = PAGE_SIZE;
+		blk_queue_max_discard_sectors(disk->queue, UINT_MAX);
+	}
+
 	/* Tell the block layer that this is not a rotational device */
 	blk_queue_flag_set(QUEUE_FLAG_NONROT, disk->queue);
 	blk_queue_flag_clear(QUEUE_FLAG_ADD_RANDOM, disk->queue);