rollback-raid5ch-to-55.patch

lustre 1.6.5 source code

--- linux-2.6.9-67/drivers/md/raid5.c	2007-11-21 21:12:50.000000000 -0700
+++ linux-2.6.9-55.0.12/drivers/md/raid5.c	2007-11-02 01:10:35.000000000 -0700
@@ -49,7 +49,7 @@
  * This macro is used to determine the 'next' bio in the list, given the sector
  * of the current stripe+device
  */
-#define r5_next_bio(bio, sect) ( ( (bio)->bi_sector + ((bio)->bi_size>>9) < sect + STRIPE_SECTORS) ? (bio)->bi_next : NULL)
+#define r5_next_bio(bio, sect) ( ( bio->bi_sector + (bio->bi_size>>9) < sect + STRIPE_SECTORS) ? bio->bi_next : NULL)
 /*
  * The following can be used to debug the driver
  */
@@ -232,7 +232,6 @@ static struct stripe_head *__find_stripe
 }
 
 static void unplug_slaves(mddev_t *mddev);
-static void raid5_unplug_device(request_queue_t *q);
 
 static struct stripe_head *get_active_stripe(raid5_conf_t *conf, sector_t sector,
 					     int pd_idx, int noblock) 
@@ -727,10 +726,6 @@ static void compute_parity(struct stripe
 				ptr[count++] = page_address(sh->dev[i].page);
 				chosen = sh->dev[i].towrite;
 				sh->dev[i].towrite = NULL;
-
-				if (test_and_clear_bit(R5_Overlap, &sh->dev[i].flags))
-					wake_up(&conf->wait_for_overlap);
-
 				if (sh->dev[i].written) BUG();
 				sh->dev[i].written = chosen;
 				check_xor();
@@ -743,10 +738,6 @@ static void compute_parity(struct stripe
 			if (i!=pd_idx && sh->dev[i].towrite) {
 				chosen = sh->dev[i].towrite;
 				sh->dev[i].towrite = NULL;
-
-				if (test_and_clear_bit(R5_Overlap, &sh->dev[i].flags))
-					wake_up(&conf->wait_for_overlap);
-
 				if (sh->dev[i].written) BUG();
 				sh->dev[i].written = chosen;
 			}
@@ -803,7 +794,7 @@ static void compute_parity(struct stripe
  * toread/towrite point to the first in a chain. 
  * The bi_next chain must be in order.
  */
-static int add_stripe_bio(struct stripe_head *sh, struct bio *bi, int dd_idx, int forwrite)
+static void add_stripe_bio (struct stripe_head *sh, struct bio *bi, int dd_idx, int forwrite)
 {
 	struct bio **bip;
 	raid5_conf_t *conf = sh->raid_conf;
@@ -820,13 +811,10 @@ static int add_stripe_bio(struct stripe_
 	else
 		bip = &sh->dev[dd_idx].toread;
 	while (*bip && (*bip)->bi_sector < bi->bi_sector) {
-		if ((*bip)->bi_sector + ((*bip)->bi_size >> 9) > bi->bi_sector)
-			goto overlap;
+		BUG_ON((*bip)->bi_sector + ((*bip)->bi_size >> 9) > bi->bi_sector);
 		bip = & (*bip)->bi_next;
 	}
-	if (*bip && (*bip)->bi_sector < bi->bi_sector + ((bi->bi_size)>>9))
-		goto overlap;
-
+/* FIXME do I need to worry about overlapping bion */
 	if (*bip && bi->bi_next && (*bip) != bi->bi_next)
 		BUG();
 	if (*bip)
@@ -841,7 +829,7 @@ static int add_stripe_bio(struct stripe_
 		(unsigned long long)sh->sector, dd_idx);
 
 	if (forwrite) {
-		/* check if page is covered */
+		/* check if page is coverred */
 		sector_t sector = sh->dev[dd_idx].sector;
 		for (bi=sh->dev[dd_idx].towrite;
 		     sector < sh->dev[dd_idx].sector + STRIPE_SECTORS &&
@@ -853,13 +841,6 @@ static int add_stripe_bio(struct stripe_
 		if (sector >= sh->dev[dd_idx].sector + STRIPE_SECTORS)
 			set_bit(R5_OVERWRITE, &sh->dev[dd_idx].flags);
 	}
-	return 1;
-
- overlap:
-	set_bit(R5_Overlap, &sh->dev[dd_idx].flags);
-	spin_unlock_irq(&conf->device_lock);
-	spin_unlock(&sh->lock);
-	return 0;
 }
 
 
@@ -920,8 +901,6 @@ static void handle_stripe(struct stripe_
 			spin_lock_irq(&conf->device_lock);
 			rbi = dev->toread;
 			dev->toread = NULL;
-			if (test_and_clear_bit(R5_Overlap, &dev->flags))
-				wake_up(&conf->wait_for_overlap);
 			spin_unlock_irq(&conf->device_lock);
 			while (rbi && rbi->bi_sector < dev->sector + STRIPE_SECTORS) {
 				copy_data(0, rbi, dev->page, dev->sector);
@@ -969,9 +948,6 @@ static void handle_stripe(struct stripe_
 			sh->dev[i].towrite = NULL;
 			if (bi) to_write--;
 
-			if (test_and_clear_bit(R5_Overlap, &sh->dev[i].flags))
-				wake_up(&conf->wait_for_overlap);
-
 			while (bi && bi->bi_sector < sh->dev[i].sector + STRIPE_SECTORS){
 				struct bio *nextbi = r5_next_bio(bi, sh->dev[i].sector);
 				clear_bit(BIO_UPTODATE, &bi->bi_flags);
@@ -1000,8 +976,6 @@ static void handle_stripe(struct stripe_
 			if (!test_bit(R5_Insync, &sh->dev[i].flags)) {
 				bi = sh->dev[i].toread;
 				sh->dev[i].toread = NULL;
-				if (test_and_clear_bit(R5_Overlap, &sh->dev[i].flags))
-					wake_up(&conf->wait_for_overlap);
 				if (bi) to_read--;
 				while (bi && bi->bi_sector < sh->dev[i].sector + STRIPE_SECTORS){
 					struct bio *nextbi = r5_next_bio(bi, sh->dev[i].sector);
@@ -1441,7 +1415,6 @@ static int make_request (request_queue_t
 	if ( bio_data_dir(bi) == WRITE )
 		md_write_start(mddev);
 	for (;logical_sector < last_sector; logical_sector += STRIPE_SECTORS) {
-		DEFINE_WAIT(w);
 		
 		new_sector = raid5_compute_sector(logical_sector,
 						  raid_disks, data_disks, &dd_idx, &pd_idx, conf);
@@ -1450,28 +1423,17 @@ static int make_request (request_queue_t
 			(unsigned long long)new_sector, 
 			(unsigned long long)logical_sector);
 
-	retry:
-		prepare_to_wait(&conf->wait_for_overlap, &w, TASK_UNINTERRUPTIBLE);
 		sh = get_active_stripe(conf, new_sector, pd_idx, (bi->bi_rw&RWA_MASK));
 		if (sh) {
-			if (!add_stripe_bio(sh, bi, dd_idx, (bi->bi_rw&RW_MASK))) {
-				/* Add failed due to overlap.  Flush everything
-				 * and wait a while
-				 */
-				raid5_unplug_device(mddev->queue);
-				release_stripe(sh);
-				schedule();
-				goto retry;
-			}
-			finish_wait(&conf->wait_for_overlap, &w);
+
+			add_stripe_bio(sh, bi, dd_idx, (bi->bi_rw&RW_MASK));
+
 			raid5_plug_device(conf);
 			handle_stripe(sh);
 			release_stripe(sh);
-
 		} else {
 			/* cannot get stripe for read-ahead, just give-up */
 			clear_bit(BIO_UPTODATE, &bi->bi_flags);
-			finish_wait(&conf->wait_for_overlap, &w);
 			break;
 		}
 			
@@ -1619,7 +1581,6 @@ static int run (mddev_t *mddev)
 
 	conf->device_lock = SPIN_LOCK_UNLOCKED;
 	init_waitqueue_head(&conf->wait_for_stripe);
-	init_waitqueue_head(&conf->wait_for_overlap);
 	INIT_LIST_HEAD(&conf->handle_list);
 	INIT_LIST_HEAD(&conf->delayed_list);
 	INIT_LIST_HEAD(&conf->inactive_list);
--- linux-2.6.9-67/include/linux/raid/raid5.h	2007-11-21 21:12:41.000000000 -0700
+++ linux-2.6.9-55.0.12/include/linux/raid/raid5.h	2004-10-18 15:54:55.000000000 -0600
@@ -152,7 +152,6 @@ struct stripe_head {
 #define	R5_Wantread	4	/* want to schedule a read */
 #define	R5_Wantwrite	5
 #define	R5_Syncio	6	/* this io need to be accounted as resync io */
-#define	R5_Overlap	7	/* There is a pending overlapping request on this block */
 
 /*
  * Write method
@@ -220,7 +219,6 @@ struct raid5_private_data {
 	atomic_t		active_stripes;
 	struct list_head	inactive_list;
 	wait_queue_head_t	wait_for_stripe;
-	wait_queue_head_t	wait_for_overlap;
 	int			inactive_blocked;	/* release of inactive stripes blocked,
 							 * waiting for 25% to be free
 							 */        
--- linux-2.6.9-67/drivers/md/raid6main.c	2007-11-21 21:12:41.000000000 -0700
+++ linux-2.6.9-55.0.12/drivers/md/raid6main.c	2007-11-02 01:10:20.000000000 -0700
@@ -54,7 +54,7 @@
  * This macro is used to determine the 'next' bio in the list, given the sector
  * of the current stripe+device
  */
-#define r5_next_bio(bio, sect) ( ( (bio)->bi_sector + ((bio)->bi_size>>9) < sect + STRIPE_SECTORS) ? (bio)->bi_next : NULL)
+#define r5_next_bio(bio, sect) ( ( bio->bi_sector + (bio->bi_size>>9) < sect + STRIPE_SECTORS) ? bio->bi_next : NULL)
 /*
  * The following can be used to debug the driver
  */
@@ -740,10 +740,6 @@ static void compute_parity(struct stripe
 			if ( i != pd_idx && i != qd_idx && sh->dev[i].towrite ) {
 				chosen = sh->dev[i].towrite;
 				sh->dev[i].towrite = NULL;
-
-				if (test_and_clear_bit(R5_Overlap, &sh->dev[i].flags))
-					wake_up(&conf->wait_for_overlap);
-
 				if (sh->dev[i].written) BUG();
 				sh->dev[i].written = chosen;
 			}
@@ -902,7 +898,7 @@ static void compute_block_2(struct strip
  * toread/towrite point to the first in a chain.
  * The bi_next chain must be in order.
  */
-static int add_stripe_bio(struct stripe_head *sh, struct bio *bi, int dd_idx, int forwrite)
+static void add_stripe_bio (struct stripe_head *sh, struct bio *bi, int dd_idx, int forwrite)
 {
 	struct bio **bip;
 	raid6_conf_t *conf = sh->raid_conf;
@@ -919,13 +915,10 @@ static int add_stripe_bio(struct stripe_
 	else
 		bip = &sh->dev[dd_idx].toread;
 	while (*bip && (*bip)->bi_sector < bi->bi_sector) {
-		if ((*bip)->bi_sector + ((*bip)->bi_size >> 9) > bi->bi_sector)
-			goto overlap;
-		bip = &(*bip)->bi_next;
+		BUG_ON((*bip)->bi_sector + ((*bip)->bi_size >> 9) > bi->bi_sector);
+		bip = & (*bip)->bi_next;
 	}
-	if (*bip && (*bip)->bi_sector < bi->bi_sector + ((bi->bi_size)>>9))
-		goto overlap;
-
+/* FIXME do I need to worry about overlapping bion */
 	if (*bip && bi->bi_next && (*bip) != bi->bi_next)
 		BUG();
 	if (*bip)
@@ -940,7 +933,7 @@ static int add_stripe_bio(struct stripe_
 		(unsigned long long)sh->sector, dd_idx);
 
 	if (forwrite) {
-		/* check if page is covered */
+		/* check if page is coverred */
 		sector_t sector = sh->dev[dd_idx].sector;
 		for (bi=sh->dev[dd_idx].towrite;
 		     sector < sh->dev[dd_idx].sector + STRIPE_SECTORS &&
@@ -952,13 +945,6 @@ static int add_stripe_bio(struct stripe_
 		if (sector >= sh->dev[dd_idx].sector + STRIPE_SECTORS)
 			set_bit(R5_OVERWRITE, &sh->dev[dd_idx].flags);
 	}
-	return 1;
-
- overlap:
-	set_bit(R5_Overlap, &sh->dev[dd_idx].flags);
-	spin_unlock_irq(&conf->device_lock);
-	spin_unlock(&sh->lock);
-	return 0;
 }
 
 
@@ -1022,8 +1008,6 @@ static void handle_stripe(struct stripe_
 			spin_lock_irq(&conf->device_lock);
 			rbi = dev->toread;
 			dev->toread = NULL;
-			if (test_and_clear_bit(R5_Overlap, &dev->flags))
-				wake_up(&conf->wait_for_overlap);
 			spin_unlock_irq(&conf->device_lock);
 			while (rbi && rbi->bi_sector < dev->sector + STRIPE_SECTORS) {
 				copy_data(0, rbi, dev->page, dev->sector);
@@ -1073,9 +1057,6 @@ static void handle_stripe(struct stripe_
 			sh->dev[i].towrite = NULL;
 			if (bi) to_write--;
 
-			if (test_and_clear_bit(R5_Overlap, &sh->dev[i].flags))
-				wake_up(&conf->wait_for_overlap);
-
 			while (bi && bi->bi_sector < sh->dev[i].sector + STRIPE_SECTORS){
 				struct bio *nextbi = r5_next_bio(bi, sh->dev[i].sector);
 				clear_bit(BIO_UPTODATE, &bi->bi_flags);
@@ -1104,8 +1085,6 @@ static void handle_stripe(struct stripe_
 			if (!test_bit(R5_Insync, &sh->dev[i].flags)) {
 				bi = sh->dev[i].toread;
 				sh->dev[i].toread = NULL;
-				if (test_and_clear_bit(R5_Overlap, &sh->dev[i].flags))
-					wake_up(&conf->wait_for_overlap);
 				if (bi) to_read--;
 				while (bi && bi->bi_sector < sh->dev[i].sector + STRIPE_SECTORS){
 					struct bio *nextbi = r5_next_bio(bi, sh->dev[i].sector);
@@ -1597,7 +1576,6 @@ static int make_request (request_queue_t
 	if ( bio_data_dir(bi) == WRITE )
 		md_write_start(mddev);
 	for (;logical_sector < last_sector; logical_sector += STRIPE_SECTORS) {
-		DEFINE_WAIT(w);
 
 		new_sector = raid6_compute_sector(logical_sector,
 						  raid_disks, data_disks, &dd_idx, &pd_idx, conf);
@@ -1606,27 +1584,17 @@ static int make_request (request_queue_t
 		       (unsigned long long)new_sector,
 		       (unsigned long long)logical_sector);
 
-	retry:
-		prepare_to_wait(&conf->wait_for_overlap, &w, TASK_UNINTERRUPTIBLE);
 		sh = get_active_stripe(conf, new_sector, pd_idx, (bi->bi_rw&RWA_MASK));
 		if (sh) {
-			if (!add_stripe_bio(sh, bi, dd_idx, (bi->bi_rw&RW_MASK))) {
-				/* Add failed due to overlap.  Flush everything
-				 * and wait a while
-				 */
-				raid6_unplug_device(mddev->queue);
-				release_stripe(sh);
-				schedule();
-				goto retry;
-			}
-			finish_wait(&conf->wait_for_overlap, &w);
+
+			add_stripe_bio(sh, bi, dd_idx, (bi->bi_rw&RW_MASK));
+
 			raid6_plug_device(conf);
 			handle_stripe(sh);
 			release_stripe(sh);
 		} else {
 			/* cannot get stripe for read-ahead, just give-up */
 			clear_bit(BIO_UPTODATE, &bi->bi_flags);
-			finish_wait(&conf->wait_for_overlap, &w);
 			break;
 		}
 
@@ -1774,7 +1742,6 @@ static int run (mddev_t *mddev)
 
 	conf->device_lock = SPIN_LOCK_UNLOCKED;
 	init_waitqueue_head(&conf->wait_for_stripe);
-	init_waitqueue_head(&conf->wait_for_overlap);
 	INIT_LIST_HEAD(&conf->handle_list);
 	INIT_LIST_HEAD(&conf->delayed_list);
 	INIT_LIST_HEAD(&conf->inactive_list);