Merge branch 'for-linus' of git://git.open-osd.org/linux-open-osd into next

Pull exofs raid6 support from Boaz Harrosh:
 "These simple patches will enable raid6 using the kernel's raid6_pq
  engine for support under exofs and pnfs-objects.

  There is nothing needed to do at exofs and pnfs-obj.  Just fire your
  mkfs.exofs with --raid=6 (that was already supported before) and off
  you go as usual.  The ORE will pick up the new map and will start
  writing two devices of redundancy bits.  The patches are so simple
  because most of the ORE was already for the general raid case, only a
  few bug fixes were needed and the actual wiring into the raid6_pq
  engine"

* 'for-linus' of git://git.open-osd.org/linux-open-osd:
  ore: Support for raid 6
  ore: Remove redundant dev_order(), more cleanups
  ore: (trivial) reformat some code

1 files changed, 64 insertions, 36 deletions

diff --git a/fs/exofs/ore.c b/fs/exofs/ore.c
index dae884694bd..cfc0205d62c 100644
--- a/fs/exofs/ore.c
+++ b/fs/exofs/ore.c
@@ -58,9 +58,12 @@ int ore_verify_layout(unsigned total_comps, struct ore_layout *layout)
 		layout->parity = 1;
 		break;
 	case PNFS_OSD_RAID_PQ:
+		layout->parity = 2;
+		break;
 	case PNFS_OSD_RAID_4:
 	default:
-		ORE_ERR("Only RAID_0/5 for now\n");
+		ORE_ERR("Only RAID_0/5/6 for now received-enum=%d\n",
+			layout->raid_algorithm);
 		return -EINVAL;
 	}
 	if (0 != (layout->stripe_unit & ~PAGE_MASK)) {
@@ -112,6 +115,8 @@ int ore_verify_layout(unsigned total_comps, struct ore_layout *layout)
 		layout->max_io_length /= stripe_length;
 		layout->max_io_length *= stripe_length;
 	}
+	ORE_DBGMSG("max_io_length=0x%lx\n", layout->max_io_length);
+
 	return 0;
 }
 EXPORT_SYMBOL(ore_verify_layout);
@@ -545,21 +550,24 @@ void ore_calc_stripe_info(struct ore_layout *layout, u64 file_offset,
 
 	/* "H - (N * U)" is just "H % U" so it's bound to u32 */
 	u32	C = (u32)(H - (N * U)) / stripe_unit + G * group_width;
+	u32 first_dev = C - C % group_width;
 
 	div_u64_rem(file_offset, stripe_unit, &si->unit_off);
 
 	si->obj_offset = si->unit_off + (N * stripe_unit) +
 				  (M * group_depth * stripe_unit);
+	si->cur_comp = C - first_dev;
+	si->cur_pg = si->unit_off / PAGE_SIZE;
 
 	if (parity) {
 		u32 LCMdP = lcm(group_width, parity) / parity;
 		/* R     = N % LCMdP; */
 		u32 RxP   = (N % LCMdP) * parity;
-		u32 first_dev = C - C % group_width;
 
 		si->par_dev = (group_width + group_width - parity - RxP) %
 			      group_width + first_dev;
-		si->dev = (group_width + C - RxP) % group_width + first_dev;
+		si->dev = (group_width + group_width + C - RxP) %
+			  group_width + first_dev;
 		si->bytes_in_stripe = U;
 		si->first_stripe_start = M * S + G * T + N * U;
 	} else {
@@ -649,6 +657,43 @@ out:	/* we fail the complete unit on an error eg don't advance
 	return ret;
 }
 
+static int _add_parity_units(struct ore_io_state *ios,
+			     struct ore_striping_info *si,
+			     unsigned dev, unsigned first_dev,
+			     unsigned mirrors_p1, unsigned devs_in_group,
+			     unsigned cur_len)
+{
+	unsigned do_parity;
+	int ret = 0;
+
+	for (do_parity = ios->layout->parity; do_parity; --do_parity) {
+		struct ore_per_dev_state *per_dev;
+
+		per_dev = &ios->per_dev[dev - first_dev];
+		if (!per_dev->length && !per_dev->offset) {
+			/* Only/always the parity unit of the first
+			 * stripe will be empty. So this is a chance to
+			 * initialize the per_dev info.
+			 */
+			per_dev->dev = dev;
+			per_dev->offset = si->obj_offset - si->unit_off;
+		}
+
+		ret = _ore_add_parity_unit(ios, si, per_dev, cur_len,
+					   do_parity == 1);
+		if (unlikely(ret))
+				break;
+
+		if (do_parity != 1) {
+			dev = ((dev + mirrors_p1) % devs_in_group) + first_dev;
+			si->cur_comp = (si->cur_comp + 1) %
+						       ios->layout->group_width;
+		}
+	}
+
+	return ret;
+}
+
 static int _prepare_for_striping(struct ore_io_state *ios)
 {
 	struct ore_striping_info *si = &ios->si;
@@ -658,7 +703,6 @@ static int _prepare_for_striping(struct ore_io_state *ios)
 	unsigned devs_in_group = group_width * mirrors_p1;
 	unsigned dev = si->dev;
 	unsigned first_dev = dev - (dev % devs_in_group);
-	unsigned dev_order;
 	unsigned cur_pg = ios->pages_consumed;
 	u64 length = ios->length;
 	int ret = 0;
@@ -670,16 +714,13 @@ static int _prepare_for_striping(struct ore_io_state *ios)
 
 	BUG_ON(length > si->length);
 
-	dev_order = _dev_order(devs_in_group, mirrors_p1, si->par_dev, dev);
-	si->cur_comp = dev_order;
-	si->cur_pg = si->unit_off / PAGE_SIZE;
-
 	while (length) {
-		unsigned comp = dev - first_dev;
-		struct ore_per_dev_state *per_dev = &ios->per_dev[comp];
+		struct ore_per_dev_state *per_dev =
+						&ios->per_dev[dev - first_dev];
 		unsigned cur_len, page_off = 0;
 
-		if (!per_dev->length) {
+		if (!per_dev->length && !per_dev->offset) {
+			/* First time initialize the per_dev info. */
 			per_dev->dev = dev;
 			if (dev == si->dev) {
 				WARN_ON(dev == si->par_dev);
@@ -688,13 +729,7 @@ static int _prepare_for_striping(struct ore_io_state *ios)
 				page_off = si->unit_off & ~PAGE_MASK;
 				BUG_ON(page_off && (page_off != ios->pgbase));
 			} else {
-				if (si->cur_comp > dev_order)
-					per_dev->offset =
-						si->obj_offset - si->unit_off;
-				else /* si->cur_comp < dev_order */
-					per_dev->offset =
-						si->obj_offset + stripe_unit -
-								   si->unit_off;
+				per_dev->offset = si->obj_offset - si->unit_off;
 				cur_len = stripe_unit;
 			}
 		} else {
@@ -708,11 +743,9 @@ static int _prepare_for_striping(struct ore_io_state *ios)
 		if (unlikely(ret))
 			goto out;
 
-		dev += mirrors_p1;
-		dev = (dev % devs_in_group) + first_dev;
-
 		length -= cur_len;
 
+		dev = ((dev + mirrors_p1) % devs_in_group) + first_dev;
 		si->cur_comp = (si->cur_comp + 1) % group_width;
 		if (unlikely((dev == si->par_dev) || (!length && ios->sp2d))) {
 			if (!length && ios->sp2d) {
@@ -720,23 +753,16 @@ static int _prepare_for_striping(struct ore_io_state *ios)
 				 * stripe. then operate on parity dev.
 				 */
 				dev = si->par_dev;
-			}
-			if (ios->sp2d)
-				/* In writes cur_len just means if it's the
-				 * last one. See _ore_add_parity_unit.
-				 */
-				cur_len = length;
-			per_dev = &ios->per_dev[dev - first_dev];
-			if (!per_dev->length) {
-				/* Only/always the parity unit of the first
-				 * stripe will be empty. So this is a chance to
-				 * initialize the per_dev info.
-				 */
-				per_dev->dev = dev;
-				per_dev->offset = si->obj_offset - si->unit_off;
+				/* If last stripe operate on parity comp */
+				si->cur_comp = group_width - ios->layout->parity;
 			}
 
-			ret = _ore_add_parity_unit(ios, si, per_dev, cur_len);
+			/* In writes cur_len just means if it's the
+			 * last one. See _ore_add_parity_unit.
+			 */
+			ret = _add_parity_units(ios, si, dev, first_dev,
+						mirrors_p1, devs_in_group,
+						ios->sp2d ? length : cur_len);
 			if (unlikely(ret))
 					goto out;
 
@@ -747,6 +773,8 @@ static int _prepare_for_striping(struct ore_io_state *ios)
 			/* Next stripe, start fresh */
 			si->cur_comp = 0;
 			si->cur_pg = 0;
+			si->obj_offset += cur_len;
+			si->unit_off = 0;
 		}
 	}
 out: