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

* 'for-linus' of git://git.open-osd.org/linux-open-osd: (21 commits)
  ore: Enable RAID5 mounts
  exofs: Support for RAID5 read-4-write interface.
  ore: RAID5 Write
  ore: RAID5 read
  fs/Makefile: Always inspect exofs/
  ore: Make ore_calc_stripe_info EXPORT_SYMBOL
  ore/exofs: Change ore_check_io API
  ore/exofs: Define new ore_verify_layout
  ore: Support for partial component table
  ore: Support for short read/writes
  exofs: Support for short read/writes
  ore: Remove check for ios->kern_buff in _prepare_for_striping to later
  ore: cleanup: Embed an ore_striping_info inside ore_io_state
  ore: Only IO one group at a time (API change)
  ore/exofs: Change the type of the devices array (API change)
  ore: Make ore_striping_info and ore_calc_stripe_info public
  exofs: Remove unused data_map member from exofs_sb_info
  exofs: Rename struct ore_components comps => oc
  exofs/super.c: local functions should be static
  exofs/ore.c: local functions should be static
  ...

drivers/scsi/osd/Kconfig

@@ -11,10 +11,6 @@
 # it under the terms of the GNU General Public version 2 License as
 # published by the Free Software Foundation
 #
-# FIXME: SCSI_OSD_INITIATOR should select CONFIG (HMAC) SHA1 somehow.
-#        How is it done properly?
-#
-
 config SCSI_OSD_INITIATOR
 	tristate "OSD-Initiator library"
 	depends on SCSI

fs/Makefile

@@ -120,6 +120,6 @@ obj-$(CONFIG_DEBUG_FS)		+= debugfs/
 obj-$(CONFIG_OCFS2_FS)		+= ocfs2/
 obj-$(CONFIG_BTRFS_FS)		+= btrfs/
 obj-$(CONFIG_GFS2_FS)           += gfs2/
-obj-$(CONFIG_EXOFS_FS)          += exofs/
+obj-$(y)          		+= exofs/ # Multiple mods, used by nfs/objlayout
 obj-$(CONFIG_CEPH_FS)		+= ceph/
 obj-$(CONFIG_PSTORE)		+= pstore/

fs/exofs/Kbuild

@@ -13,7 +13,8 @@
 #
 
 # ore module library
-obj-$(CONFIG_ORE) += ore.o
+libore-y := ore.o ore_raid.o
+obj-$(CONFIG_ORE) += libore.o
 
 exofs-y := inode.o file.o symlink.o namei.o dir.o super.o
 obj-$(CONFIG_EXOFS_FS) += exofs.o

fs/exofs/Kconfig

+# Note ORE needs to "select ASYNC_XOR". So Not to force multiple selects
+# for every ORE user we do it like this. Any user should add itself here
+# at the "depends on EXOFS_FS || ..." with an ||. The dependencies are
+# selected here, and we default to "ON". So in effect it is like been
+# selected by any of the users.
 config ORE
 	tristate
+	depends on EXOFS_FS
+	select ASYNC_XOR
+	default SCSI_OSD_ULD
 
 config EXOFS_FS
 	tristate "exofs: OSD based file system support"
 	depends on SCSI_OSD_ULD
-	select ORE
 	help
 	  EXOFS is a file system that uses an OSD storage device,
 	  as its backing storage.

fs/exofs/exofs.h

@@ -53,6 +53,10 @@
 /* u64 has problems with printk this will cast it to unsigned long long */
 #define _LLU(x) (unsigned long long)(x)
 
+struct exofs_dev {
+	struct ore_dev ored;
+	unsigned did;
+};
 /*
  * our extension to the in-memory superblock
  */
@@ -66,13 +70,9 @@ struct exofs_sb_info {
 	u32		s_next_generation;	/* next gen # to use          */
 	atomic_t	s_curr_pending;		/* number of pending commands */
 
-	struct pnfs_osd_data_map data_map;	/* Default raid to use
-						 * FIXME: Needed ?
-						 */
 	struct ore_layout	layout;		/* Default files layout       */
 	struct ore_comp one_comp;		/* id & cred of partition id=0*/
-	struct ore_components comps;		/* comps for the partition    */
-	struct osd_dev	*_min_one_dev[1];	/* Place holder for one dev   */
+	struct ore_components oc;		/* comps for the partition    */
 };
 
 /*
@@ -86,7 +86,7 @@ struct exofs_i_info {
 	uint32_t       i_dir_start_lookup; /* which page to start lookup      */
 	uint64_t       i_commit_size;      /* the object's written length     */
 	struct ore_comp one_comp;	   /* same component for all devices  */
-	struct ore_components comps;	   /* inode view of the device table  */
+	struct ore_components oc;	   /* inode view of the device table  */
 };
 
 static inline osd_id exofs_oi_objno(struct exofs_i_info *oi)
@@ -207,7 +207,7 @@ extern const struct inode_operations exofs_fast_symlink_inode_operations;
  * bigger and that the device table repeats twice.
  * See: exofs_read_lookup_dev_table()
  */
-static inline void exofs_init_comps(struct ore_components *comps,
+static inline void exofs_init_comps(struct ore_components *oc,
 				    struct ore_comp *one_comp,
 				    struct exofs_sb_info *sbi, osd_id oid)
 {
@@ -217,13 +217,15 @@ static inline void exofs_init_comps(struct ore_components *comps,
 	one_comp->obj.id = oid;
 	exofs_make_credential(one_comp->cred, &one_comp->obj);
 
-	comps->numdevs = sbi->comps.numdevs;
-	comps->single_comp = EC_SINGLE_COMP;
-	comps->comps = one_comp;
+	oc->first_dev = 0;
+	oc->numdevs = sbi->layout.group_width * sbi->layout.mirrors_p1 *
+							sbi->layout.group_count;
+	oc->single_comp = EC_SINGLE_COMP;
+	oc->comps = one_comp;
 
 	/* Round robin device view of the table */
-	first_dev = (dev_mod * sbi->layout.mirrors_p1) % sbi->comps.numdevs;
-	comps->ods = sbi->comps.ods + first_dev;
+	first_dev = (dev_mod * sbi->layout.mirrors_p1) % sbi->oc.numdevs;
+	oc->ods = &sbi->oc.ods[first_dev];
 }
 
 #endif

fs/exofs/inode.c

@@ -37,11 +37,7 @@
 
 #define EXOFS_DBGMSG2(M...) do {} while (0)
 
-enum { BIO_MAX_PAGES_KMALLOC =
-		(PAGE_SIZE - sizeof(struct bio)) / sizeof(struct bio_vec),
-	MAX_PAGES_KMALLOC =
-		PAGE_SIZE / sizeof(struct page *),
-};
+enum {MAX_PAGES_KMALLOC = PAGE_SIZE / sizeof(struct page *), };
 
 unsigned exofs_max_io_pages(struct ore_layout *layout,
 			    unsigned expected_pages)
@@ -49,8 +45,7 @@ unsigned exofs_max_io_pages(struct ore_layout *layout,
 	unsigned pages = min_t(unsigned, expected_pages, MAX_PAGES_KMALLOC);
 
 	/* TODO: easily support bio chaining */
-	pages =  min_t(unsigned, pages,
-		       layout->group_width * BIO_MAX_PAGES_KMALLOC);
+	pages =  min_t(unsigned, pages, layout->max_io_length / PAGE_SIZE);
 	return pages;
 }
 
@@ -68,6 +63,7 @@ struct page_collect {
 	bool read_4_write; /* This means two things: that the read is sync
 			    * And the pages should not be unlocked.
 			    */
+	struct page *that_locked_page;
 };
 
 static void _pcol_init(struct page_collect *pcol, unsigned expected_pages,
@@ -86,6 +82,7 @@ static void _pcol_init(struct page_collect *pcol, unsigned expected_pages,
 	pcol->length = 0;
 	pcol->pg_first = -1;
 	pcol->read_4_write = false;
+	pcol->that_locked_page = NULL;
 }
 
 static void _pcol_reset(struct page_collect *pcol)
@@ -98,6 +95,7 @@ static void _pcol_reset(struct page_collect *pcol)
 	pcol->length = 0;
 	pcol->pg_first = -1;
 	pcol->ios = NULL;
+	pcol->that_locked_page = NULL;
 
 	/* this is probably the end of the loop but in writes
 	 * it might not end here. don't be left with nothing
@@ -149,14 +147,17 @@ static int pcol_add_page(struct page_collect *pcol, struct page *page,
 	return 0;
 }
 
+enum {PAGE_WAS_NOT_IN_IO = 17};
 static int update_read_page(struct page *page, int ret)
 {
-	if (ret == 0) {
+	switch (ret) {
+	case 0:
 		/* Everything is OK */
 		SetPageUptodate(page);
 		if (PageError(page))
 			ClearPageError(page);
-	} else if (ret == -EFAULT) {
+		break;
+	case -EFAULT:
 		/* In this case we were trying to read something that wasn't on
 		 * disk yet - return a page full of zeroes.  This should be OK,
 		 * because the object should be empty (if there was a write
@@ -167,16 +168,22 @@ static int update_read_page(struct page *page, int ret)
 		SetPageUptodate(page);
 		if (PageError(page))
 			ClearPageError(page);
-		ret = 0; /* recovered error */
 		EXOFS_DBGMSG("recovered read error\n");
-	} else /* Error */
+		/* fall through */
+	case PAGE_WAS_NOT_IN_IO:
+		ret = 0; /* recovered error */
+		break;
+	default:
 		SetPageError(page);
-
+	}
 	return ret;
 }
 
 static void update_write_page(struct page *page, int ret)
 {
+	if (unlikely(ret == PAGE_WAS_NOT_IN_IO))
+		return; /* don't pass start don't collect $200 */
+
 	if (ret) {
 		mapping_set_error(page->mapping, ret);
 		SetPageError(page);
@@ -190,15 +197,16 @@ static void update_write_page(struct page *page, int ret)
 static int __readpages_done(struct page_collect *pcol)
 {
 	int i;
-	u64 resid;
 	u64 good_bytes;
 	u64 length = 0;
-	int ret = ore_check_io(pcol->ios, &resid);
+	int ret = ore_check_io(pcol->ios, NULL);
 
-	if (likely(!ret))
+	if (likely(!ret)) {
 		good_bytes = pcol->length;
-	else
-		good_bytes = pcol->length - resid;
+		ret = PAGE_WAS_NOT_IN_IO;
+	} else {
+		good_bytes = 0;
+	}
 
 	EXOFS_DBGMSG2("readpages_done(0x%lx) good_bytes=0x%llx"
 		     " length=0x%lx nr_pages=%u\n",
@@ -259,6 +267,46 @@ static void _unlock_pcol_pages(struct page_collect *pcol, int ret, int rw)
 	}
 }
 
+static int _maybe_not_all_in_one_io(struct ore_io_state *ios,
+	struct page_collect *pcol_src, struct page_collect *pcol)
+{
+	/* length was wrong or offset was not page aligned */
+	BUG_ON(pcol_src->nr_pages < ios->nr_pages);
+
+	if (pcol_src->nr_pages > ios->nr_pages) {
+		struct page **src_page;
+		unsigned pages_less = pcol_src->nr_pages - ios->nr_pages;
+		unsigned long len_less = pcol_src->length - ios->length;
+		unsigned i;
+		int ret;
+
+		/* This IO was trimmed */
+		pcol_src->nr_pages = ios->nr_pages;
+		pcol_src->length = ios->length;
+
+		/* Left over pages are passed to the next io */
+		pcol->expected_pages += pages_less;
+		pcol->nr_pages = pages_less;
+		pcol->length = len_less;
+		src_page = pcol_src->pages + pcol_src->nr_pages;
+		pcol->pg_first = (*src_page)->index;
+
+		ret = pcol_try_alloc(pcol);
+		if (unlikely(ret))
+			return ret;
+
+		for (i = 0; i < pages_less; ++i)
+			pcol->pages[i] = *src_page++;
+
+		EXOFS_DBGMSG("Length was adjusted nr_pages=0x%x "
+			"pages_less=0x%x expected_pages=0x%x "
+			"next_offset=0x%llx next_len=0x%lx\n",
+			pcol_src->nr_pages, pages_less, pcol->expected_pages,
+			pcol->pg_first * PAGE_SIZE, pcol->length);
+	}
+	return 0;
+}
+
 static int read_exec(struct page_collect *pcol)
 {
 	struct exofs_i_info *oi = exofs_i(pcol->inode);
@@ -270,7 +318,7 @@ static int read_exec(struct page_collect *pcol)
 		return 0;
 
 	if (!pcol->ios) {
-		int ret = ore_get_rw_state(&pcol->sbi->layout, &oi->comps, true,
+		int ret = ore_get_rw_state(&pcol->sbi->layout, &oi->oc, true,
 					     pcol->pg_first << PAGE_CACHE_SHIFT,
 					     pcol->length, &pcol->ios);
 
@@ -280,7 +328,6 @@ static int read_exec(struct page_collect *pcol)
 
 	ios = pcol->ios;
 	ios->pages = pcol->pages;
-	ios->nr_pages = pcol->nr_pages;
 
 	if (pcol->read_4_write) {
 		ore_read(pcol->ios);
@@ -296,17 +343,23 @@ static int read_exec(struct page_collect *pcol)
 	*pcol_copy = *pcol;
 	ios->done = readpages_done;
 	ios->private = pcol_copy;
+
+	/* pages ownership was passed to pcol_copy */
+	_pcol_reset(pcol);
+
+	ret = _maybe_not_all_in_one_io(ios, pcol_copy, pcol);
+	if (unlikely(ret))
+		goto err;
+
+	EXOFS_DBGMSG2("read_exec(0x%lx) offset=0x%llx length=0x%llx\n",
+		pcol->inode->i_ino, _LLU(ios->offset), _LLU(ios->length));
+
 	ret = ore_read(ios);
 	if (unlikely(ret))
 		goto err;
 
 	atomic_inc(&pcol->sbi->s_curr_pending);
 
-	EXOFS_DBGMSG2("read_exec obj=0x%llx start=0x%llx length=0x%lx\n",
-		  oi->one_comp.obj.id, _LLU(ios->offset), pcol->length);
-
-	/* pages ownership was passed to pcol_copy */
-	_pcol_reset(pcol);
 	return 0;
 
 err:
@@ -341,6 +394,8 @@ static int readpage_strip(void *data, struct page *page)
 		EXOFS_ERR("PageUptodate(0x%lx, 0x%lx)\n", pcol->inode->i_ino,
 			  page->index);
 
+	pcol->that_locked_page = page;
+
 	if (page->index < end_index)
 		len = PAGE_CACHE_SIZE;
 	else if (page->index == end_index)
@@ -429,6 +484,10 @@ static int exofs_readpages(struct file *file, struct address_space *mapping,
 		return ret;
 	}
 
+	ret = read_exec(&pcol);
+	if (unlikely(ret))
+		return ret;
+
 	return read_exec(&pcol);
 }
 
@@ -462,17 +521,18 @@ static void writepages_done(struct ore_io_state *ios, void *p)
 {
 	struct page_collect *pcol = p;
 	int i;
-	u64 resid;
 	u64  good_bytes;
 	u64  length = 0;
-	int ret = ore_check_io(ios, &resid);
+	int ret = ore_check_io(ios, NULL);
 
 	atomic_dec(&pcol->sbi->s_curr_pending);
 
-	if (likely(!ret))
+	if (likely(!ret)) {
 		good_bytes = pcol->length;
-	else
-		good_bytes = pcol->length - resid;
+		ret = PAGE_WAS_NOT_IN_IO;
+	} else {
+		good_bytes = 0;
+	}
 
 	EXOFS_DBGMSG2("writepages_done(0x%lx) good_bytes=0x%llx"
 		     " length=0x%lx nr_pages=%u\n",
@@ -505,6 +565,56 @@ static void writepages_done(struct ore_io_state *ios, void *p)
 	EXOFS_DBGMSG2("writepages_done END\n");
 }
 
+static struct page *__r4w_get_page(void *priv, u64 offset, bool *uptodate)
+{
+	struct page_collect *pcol = priv;
+	pgoff_t index = offset / PAGE_SIZE;
+
+	if (!pcol->that_locked_page ||
+	    (pcol->that_locked_page->index != index)) {
+		struct page *page = find_get_page(pcol->inode->i_mapping, index);
+
+		if (!page) {
+			page = find_or_create_page(pcol->inode->i_mapping,
+						   index, GFP_NOFS);
+			if (unlikely(!page)) {
+				EXOFS_DBGMSG("grab_cache_page Failed "
+					"index=0x%llx\n", _LLU(index));
+				return NULL;
+			}
+			unlock_page(page);
+		}
+		if (PageDirty(page) || PageWriteback(page))
+			*uptodate = true;
+		else
+			*uptodate = PageUptodate(page);
+		EXOFS_DBGMSG("index=0x%lx uptodate=%d\n", index, *uptodate);
+		return page;
+	} else {
+		EXOFS_DBGMSG("YES that_locked_page index=0x%lx\n",
+			     pcol->that_locked_page->index);
+		*uptodate = true;
+		return pcol->that_locked_page;
+	}
+}
+
+static void __r4w_put_page(void *priv, struct page *page)
+{
+	struct page_collect *pcol = priv;
+
+	if (pcol->that_locked_page != page) {
+		EXOFS_DBGMSG("index=0x%lx\n", page->index);
+		page_cache_release(page);
+		return;
+	}
+	EXOFS_DBGMSG("that_locked_page index=0x%lx\n", page->index);
+}
+
+static const struct _ore_r4w_op _r4w_op = {
+	.get_page = &__r4w_get_page,
+	.put_page = &__r4w_put_page,
+};
+
 static int write_exec(struct page_collect *pcol)
 {
 	struct exofs_i_info *oi = exofs_i(pcol->inode);
@@ -516,10 +626,9 @@ static int write_exec(struct page_collect *pcol)
 		return 0;
 
 	BUG_ON(pcol->ios);
-	ret = ore_get_rw_state(&pcol->sbi->layout, &oi->comps, false,
+	ret = ore_get_rw_state(&pcol->sbi->layout, &oi->oc, false,
 				 pcol->pg_first << PAGE_CACHE_SHIFT,
 				 pcol->length, &pcol->ios);
-
 	if (unlikely(ret))
 		goto err;
 
@@ -534,10 +643,20 @@ static int write_exec(struct page_collect *pcol)
 
 	ios = pcol->ios;
 	ios->pages = pcol_copy->pages;
-	ios->nr_pages = pcol_copy->nr_pages;
 	ios->done = writepages_done;
+	ios->r4w = &_r4w_op;
 	ios->private = pcol_copy;
 
+	/* pages ownership was passed to pcol_copy */
+	_pcol_reset(pcol);
+
+	ret = _maybe_not_all_in_one_io(ios, pcol_copy, pcol);
+	if (unlikely(ret))
+		goto err;
+
+	EXOFS_DBGMSG2("write_exec(0x%lx) offset=0x%llx length=0x%llx\n",
+		pcol->inode->i_ino, _LLU(ios->offset), _LLU(ios->length));
+
 	ret = ore_write(ios);
 	if (unlikely(ret)) {
 		EXOFS_ERR("write_exec: ore_write() Failed\n");
@@ -545,11 +664,6 @@ static int write_exec(struct page_collect *pcol)
 	}
 
 	atomic_inc(&pcol->sbi->s_curr_pending);
-	EXOFS_DBGMSG2("write_exec(0x%lx, 0x%llx) start=0x%llx length=0x%lx\n",
-		  pcol->inode->i_ino, pcol->pg_first, _LLU(ios->offset),
-		  pcol->length);
-	/* pages ownership was passed to pcol_copy */
-	_pcol_reset(pcol);
 	return 0;
 
 err:
@@ -689,14 +803,33 @@ static int exofs_writepages(struct address_space *mapping,
 	_pcol_init(&pcol, expected_pages, mapping->host);
 
 	ret = write_cache_pages(mapping, wbc, writepage_strip, &pcol);
-	if (ret) {
+	if (unlikely(ret)) {
 		EXOFS_ERR("write_cache_pages => %d\n", ret);
 		return ret;
 	}
 
-	return write_exec(&pcol);
+	ret = write_exec(&pcol);
+	if (unlikely(ret))
+		return ret;
+
+	if (wbc->sync_mode == WB_SYNC_ALL) {
+		return write_exec(&pcol); /* pump the last reminder */
+	} else if (pcol.nr_pages) {
+		/* not SYNC let the reminder join the next writeout */
+		unsigned i;
+
+		for (i = 0; i < pcol.nr_pages; i++) {
+			struct page *page = pcol.pages[i];
+
+			end_page_writeback(page);
+			set_page_dirty(page);
+			unlock_page(page);
+		}
+	}
+	return 0;
 }
 
+/*
 static int exofs_writepage(struct page *page, struct writeback_control *wbc)
 {
 	struct page_collect pcol;
@@ -712,7 +845,7 @@ static int exofs_writepage(struct page *page, struct writeback_control *wbc)
 
 	return write_exec(&pcol);
 }
-
+*/
 /* i_mutex held using inode->i_size directly */
 static void _write_failed(struct inode *inode, loff_t to)
 {
@@ -818,7 +951,7 @@ static void exofs_invalidatepage(struct page *page, unsigned long offset)
 const struct address_space_operations exofs_aops = {
 	.readpage	= exofs_readpage,
 	.readpages	= exofs_readpages,
-	.writepage	= exofs_writepage,
+	.writepage	= NULL,
 	.writepages	= exofs_writepages,
 	.write_begin	= exofs_write_begin_export,
 	.write_end	= exofs_write_end,
@@ -860,7 +993,7 @@ static int _do_truncate(struct inode *inode, loff_t newsize)
 
 	inode->i_mtime = inode->i_ctime = CURRENT_TIME;
 
-	ret = ore_truncate(&sbi->layout, &oi->comps, (u64)newsize);
+	ret = ore_truncate(&sbi->layout, &oi->oc, (u64)newsize);
 	if (likely(!ret))
 		truncate_setsize(inode, newsize);
 
@@ -927,14 +1060,14 @@ static int exofs_get_inode(struct super_block *sb, struct exofs_i_info *oi,
 	struct exofs_on_disk_inode_layout *layout;
 	int ret;
 
-	ret = ore_get_io_state(&sbi->layout, &oi->comps, &ios);
+	ret = ore_get_io_state(&sbi->layout, &oi->oc, &ios);
 	if (unlikely(ret)) {
 		EXOFS_ERR("%s: ore_get_io_state failed.\n", __func__);
 		return ret;
 	}
 
-	attrs[1].len = exofs_on_disk_inode_layout_size(sbi->comps.numdevs);
-	attrs[2].len = exofs_on_disk_inode_layout_size(sbi->comps.numdevs);
+	attrs[1].len = exofs_on_disk_inode_layout_size(sbi->oc.numdevs);
+	attrs[2].len = exofs_on_disk_inode_layout_size(sbi->oc.numdevs);
 
 	ios->in_attr = attrs;
 	ios->in_attr_len = ARRAY_SIZE(attrs);
@@ -1018,7 +1151,7 @@ struct inode *exofs_iget(struct super_block *sb, unsigned long ino)
 		return inode;
 	oi = exofs_i(inode);
 	__oi_init(oi);
-	exofs_init_comps(&oi->comps, &oi->one_comp, sb->s_fs_info,
+	exofs_init_comps(&oi->oc, &oi->one_comp, sb->s_fs_info,
 			 exofs_oi_objno(oi));
 
 	/* read the inode from the osd */
@@ -1172,13 +1305,13 @@ struct inode *exofs_new_inode(struct inode *dir, int mode)
 	spin_unlock(&sbi->s_next_gen_lock);
 	insert_inode_hash(inode);
 
-	exofs_init_comps(&oi->comps, &oi->one_comp, sb->s_fs_info,
+	exofs_init_comps(&oi->oc, &oi->one_comp, sb->s_fs_info,
 			 exofs_oi_objno(oi));
 	exofs_sbi_write_stats(sbi); /* Make sure new sbi->s_nextid is on disk */
 
 	mark_inode_dirty(inode);
 
-	ret = ore_get_io_state(&sbi->layout, &oi->comps, &ios);
+	ret = ore_get_io_state(&sbi->layout, &oi->oc, &ios);
 	if (unlikely(ret)) {
 		EXOFS_ERR("exofs_new_inode: ore_get_io_state failed\n");
 		return ERR_PTR(ret);
@@ -1267,7 +1400,7 @@ static int exofs_update_inode(struct inode *inode, int do_sync)
 	} else
 		memcpy(fcb->i_data, oi->i_data, sizeof(fcb->i_data));
 
-	ret = ore_get_io_state(&sbi->layout, &oi->comps, &ios);
+	ret = ore_get_io_state(&sbi->layout, &oi->oc, &ios);
 	if (unlikely(ret)) {
 		EXOFS_ERR("%s: ore_get_io_state failed.\n", __func__);
 		goto free_args;
@@ -1350,7 +1483,7 @@ void exofs_evict_inode(struct inode *inode)
 	/* ignore the error, attempt a remove anyway */
 
 	/* Now Remove the OSD objects */
-	ret = ore_get_io_state(&sbi->layout, &oi->comps, &ios);
+	ret = ore_get_io_state(&sbi->layout, &oi->oc, &ios);
 	if (unlikely(ret)) {
 		EXOFS_ERR("%s: ore_get_io_state failed\n", __func__);
 		return;

fs/exofs/ore.c

@@ -24,76 +24,287 @@
 
 #include <linux/slab.h>
 #include <asm/div64.h>
+#include <linux/lcm.h>
 
-#include <scsi/osd_ore.h>
+#include "ore_raid.h"
 
-#define ORE_ERR(fmt, a...) printk(KERN_ERR "ore: " fmt, ##a)
+MODULE_AUTHOR("Boaz Harrosh <bharrosh@panasas.com>");
+MODULE_DESCRIPTION("Objects Raid Engine ore.ko");
+MODULE_LICENSE("GPL");
+
+/* ore_verify_layout does a couple of things:
+ * 1. Given a minimum number of needed parameters fixes up the rest of the
+ *    members to be operatonals for the ore. The needed parameters are those
+ *    that are defined by the pnfs-objects layout STD.
+ * 2. Check to see if the current ore code actually supports these parameters
+ *    for example stripe_unit must be a multple of the system PAGE_SIZE,
+ *    and etc...
+ * 3. Cache some havily used calculations that will be needed by users.
+ */
+
+enum { BIO_MAX_PAGES_KMALLOC =
+		(PAGE_SIZE - sizeof(struct bio)) / sizeof(struct bio_vec),};
 
-#ifdef CONFIG_EXOFS_DEBUG
-#define ORE_DBGMSG(fmt, a...) \
-	printk(KERN_NOTICE "ore @%s:%d: " fmt, __func__, __LINE__, ##a)
-#else
-#define ORE_DBGMSG(fmt, a...) \
-	do { if (0) printk(fmt, ##a); } while (0)
-#endif
+int ore_verify_layout(unsigned total_comps, struct ore_layout *layout)
+{
+	u64 stripe_length;
+
+	switch (layout->raid_algorithm) {
+	case PNFS_OSD_RAID_0:
+		layout->parity = 0;
+		break;
+	case PNFS_OSD_RAID_5:
+		layout->parity = 1;
+		break;
+	case PNFS_OSD_RAID_PQ:
+	case PNFS_OSD_RAID_4:
+	default:
+		ORE_ERR("Only RAID_0/5 for now\n");
+		return -EINVAL;
+	}
+	if (0 != (layout->stripe_unit & ~PAGE_MASK)) {
+		ORE_ERR("Stripe Unit(0x%llx)"
+			  " must be Multples of PAGE_SIZE(0x%lx)\n",
+			  _LLU(layout->stripe_unit), PAGE_SIZE);
+		return -EINVAL;
+	}
+	if (layout->group_width) {
+		if (!layout->group_depth) {
+			ORE_ERR("group_depth == 0 && group_width != 0\n");
+			return -EINVAL;
+		}
+		if (total_comps < (layout->group_width * layout->mirrors_p1)) {
+			ORE_ERR("Data Map wrong, "
+				"numdevs=%d < group_width=%d * mirrors=%d\n",
+				total_comps, layout->group_width,
+				layout->mirrors_p1);
+			return -EINVAL;
+		}
+		layout->group_count = total_comps / layout->mirrors_p1 /
+						layout->group_width;
+	} else {
+		if (layout->group_depth) {
+			printk(KERN_NOTICE "Warning: group_depth ignored "
+				"group_width == 0 && group_depth == %lld\n",
+				_LLU(layout->group_depth));
+		}
+		layout->group_width = total_comps / layout->mirrors_p1;
+		layout->group_depth = -1;
+		layout->group_count = 1;
+	}
 
-/* u64 has problems with printk this will cast it to unsigned long long */
-#define _LLU(x) (unsigned long long)(x)
+	stripe_length = (u64)layout->group_width * layout->stripe_unit;
+	if (stripe_length >= (1ULL << 32)) {
+		ORE_ERR("Stripe_length(0x%llx) >= 32bit is not supported\n",
+			_LLU(stripe_length));
+		return -EINVAL;
+	}
 
-#define ORE_DBGMSG2(M...) do {} while (0)
-/* #define ORE_DBGMSG2 ORE_DBGMSG */
+	layout->max_io_length =
+		(BIO_MAX_PAGES_KMALLOC * PAGE_SIZE - layout->stripe_unit) *
+							layout->group_width;
+	if (layout->parity) {
+		unsigned stripe_length =
+				(layout->group_width - layout->parity) *
+				layout->stripe_unit;
 
-MODULE_AUTHOR("Boaz Harrosh <bharrosh@panasas.com>");
-MODULE_DESCRIPTION("Objects Raid Engine ore.ko");
-MODULE_LICENSE("GPL");
+		layout->max_io_length /= stripe_length;
+		layout->max_io_length *= stripe_length;
+	}
+	return 0;
+}
+EXPORT_SYMBOL(ore_verify_layout);
 
 static u8 *_ios_cred(struct ore_io_state *ios, unsigned index)
 {
-	return ios->comps->comps[index & ios->comps->single_comp].cred;
+	return ios->oc->comps[index & ios->oc->single_comp].cred;
 }
 
 static struct osd_obj_id *_ios_obj(struct ore_io_state *ios, unsigned index)
 {
-	return &ios->comps->comps[index & ios->comps->single_comp].obj;
+	return &ios->oc->comps[index & ios->oc->single_comp].obj;
 }
 
 static struct osd_dev *_ios_od(struct ore_io_state *ios, unsigned index)
 {
-	return ios->comps->ods[index];
+	ORE_DBGMSG2("oc->first_dev=%d oc->numdevs=%d i=%d oc->ods=%p\n",
+		    ios->oc->first_dev, ios->oc->numdevs, index,
+		    ios->oc->ods);
+
+	return ore_comp_dev(ios->oc, index);
 }
 
-int  ore_get_rw_state(struct ore_layout *layout, struct ore_components *comps,
+int  _ore_get_io_state(struct ore_layout *layout,
+			struct ore_components *oc, unsigned numdevs,
+			unsigned sgs_per_dev, unsigned num_par_pages,
+			struct ore_io_state **pios)
+{
+	struct ore_io_state *ios;
+	struct page **pages;
+	struct osd_sg_entry *sgilist;
+	struct __alloc_all_io_state {
+		struct ore_io_state ios;
+		struct ore_per_dev_state per_dev[numdevs];
+		union {
+			struct osd_sg_entry sglist[sgs_per_dev * numdevs];
+			struct page *pages[num_par_pages];
+		};
+	} *_aios;
+
+	if (likely(sizeof(*_aios) <= PAGE_SIZE)) {
+		_aios = kzalloc(sizeof(*_aios), GFP_KERNEL);
+		if (unlikely(!_aios)) {
+			ORE_DBGMSG("Failed kzalloc bytes=%zd\n",
+				   sizeof(*_aios));
+			*pios = NULL;
+			return -ENOMEM;
+		}
+		pages = num_par_pages ? _aios->pages : NULL;
+		sgilist = sgs_per_dev ? _aios->sglist : NULL;
+		ios = &_aios->ios;
+	} else {
+		struct __alloc_small_io_state {
+			struct ore_io_state ios;
+			struct ore_per_dev_state per_dev[numdevs];
+		} *_aio_small;
+		union __extra_part {
+			struct osd_sg_entry sglist[sgs_per_dev * numdevs];
+			struct page *pages[num_par_pages];
+		} *extra_part;
+
+		_aio_small = kzalloc(sizeof(*_aio_small), GFP_KERNEL);
+		if (unlikely(!_aio_small)) {
+			ORE_DBGMSG("Failed alloc first part bytes=%zd\n",
+				   sizeof(*_aio_small));
+			*pios = NULL;
+			return -ENOMEM;
+		}
+		extra_part = kzalloc(sizeof(*extra_part), GFP_KERNEL);
+		if (unlikely(!extra_part)) {
+			ORE_DBGMSG("Failed alloc second part bytes=%zd\n",
+				   sizeof(*extra_part));
+			kfree(_aio_small);
+			*pios = NULL;
+			return -ENOMEM;
+		}
+
+		pages = num_par_pages ? extra_part->pages : NULL;
+		sgilist = sgs_per_dev ? extra_part->sglist : NULL;
+		/* In this case the per_dev[0].sgilist holds the pointer to
+		 * be freed
+		 */
+		ios = &_aio_small->ios;
+		ios->extra_part_alloc = true;
+	}
+
+	if (pages) {
+		ios->parity_pages = pages;
+		ios->max_par_pages = num_par_pages;
+	}
+	if (sgilist) {
+		unsigned d;
+
+		for (d = 0; d < numdevs; ++d) {
+			ios->per_dev[d].sglist = sgilist;
+			sgilist += sgs_per_dev;
+		}
+		ios->sgs_per_dev = sgs_per_dev;
+	}
+
+	ios->layout = layout;
+	ios->oc = oc;
+	*pios = ios;
+	return 0;
+}
+
+/* Allocate an io_state for only a single group of devices
+ *
+ * If a user needs to call ore_read/write() this version must be used becase it
+ * allocates extra stuff for striping and raid.
+ * The ore might decide to only IO less then @length bytes do to alignmets
+ * and constrains as follows:
+ * - The IO cannot cross group boundary.
+ * - In raid5/6 The end of the IO must align at end of a stripe eg.
+ *   (@offset + @length) % strip_size == 0. Or the complete range is within a
+ *   single stripe.
+ * - Memory condition only permitted a shorter IO. (A user can use @length=~0
+ *   And check the returned ios->length for max_io_size.)
+ *
+ * The caller must check returned ios->length (and/or ios->nr_pages) and
+ * re-issue these pages that fall outside of ios->length
+ */
+int  ore_get_rw_state(struct ore_layout *layout, struct ore_components *oc,
 		      bool is_reading, u64 offset, u64 length,
 		      struct ore_io_state **pios)
 {
 	struct ore_io_state *ios;
+	unsigned numdevs = layout->group_width * layout->mirrors_p1;
+	unsigned sgs_per_dev = 0, max_par_pages = 0;
+	int ret;
 
-	/*TODO: Maybe use kmem_cach per sbi of size
-	 * exofs_io_state_size(layout->s_numdevs)
-	 */
-	ios = kzalloc(ore_io_state_size(comps->numdevs), GFP_KERNEL);
-	if (unlikely(!ios)) {
-		ORE_DBGMSG("Failed kzalloc bytes=%d\n",
-			     ore_io_state_size(comps->numdevs));
-		*pios = NULL;
-		return -ENOMEM;
+	if (layout->parity && length) {
+		unsigned data_devs = layout->group_width - layout->parity;
+		unsigned stripe_size = layout->stripe_unit * data_devs;
+		unsigned pages_in_unit = layout->stripe_unit / PAGE_SIZE;
+		u32 remainder;
+		u64 num_stripes;
+		u64 num_raid_units;
+
+		num_stripes = div_u64_rem(length, stripe_size, &remainder);
+		if (remainder)
+			++num_stripes;
+
+		num_raid_units =  num_stripes * layout->parity;
+
+		if (is_reading) {
+			/* For reads add per_dev sglist array */
+			/* TODO: Raid 6 we need twice more. Actually:
+			*         num_stripes / LCMdP(W,P);
+			*         if (W%P != 0) num_stripes *= parity;
+			*/
+
+			/* first/last seg is split */
+			num_raid_units += layout->group_width;
+			sgs_per_dev = div_u64(num_raid_units, data_devs);
+		} else {
+			/* For Writes add parity pages array. */
+			max_par_pages = num_raid_units * pages_in_unit *
+						sizeof(struct page *);
+		}
 	}
 
-	ios->layout = layout;
-	ios->comps = comps;
-	ios->offset = offset;
-	ios->length = length;
+	ret = _ore_get_io_state(layout, oc, numdevs, sgs_per_dev, max_par_pages,
+				pios);
+	if (unlikely(ret))
+		return ret;
+
+	ios = *pios;
 	ios->reading = is_reading;
+	ios->offset = offset;
+
+	if (length) {
+		ore_calc_stripe_info(layout, offset, length, &ios->si);
+		ios->length = ios->si.length;
+		ios->nr_pages = (ios->length + PAGE_SIZE - 1) / PAGE_SIZE;
+		if (layout->parity)
+			_ore_post_alloc_raid_stuff(ios);
+	}
 
-	*pios = ios;
 	return 0;
 }
 EXPORT_SYMBOL(ore_get_rw_state);
 
-int  ore_get_io_state(struct ore_layout *layout, struct ore_components *comps,
-		      struct ore_io_state **ios)
+/* Allocate an io_state for all the devices in the comps array
+ *
+ * This version of io_state allocation is used mostly by create/remove
+ * and trunc where we currently need all the devices. The only wastful
+ * bit is the read/write_attributes with no IO. Those sites should
+ * be converted to use ore_get_rw_state() with length=0
+ */
+int  ore_get_io_state(struct ore_layout *layout, struct ore_components *oc,
+		      struct ore_io_state **pios)
 {
-	return ore_get_rw_state(layout, comps, true, 0, 0, ios);
+	return _ore_get_io_state(layout, oc, oc->numdevs, 0, 0, pios);
 }
 EXPORT_SYMBOL(ore_get_io_state);
 
@@ -111,6 +322,7 @@ void ore_put_io_state(struct ore_io_state *ios)
 				bio_put(per_dev->bio);
 		}
 
+		_ore_free_raid_stuff(ios);
 		kfree(ios);
 	}
 }
@@ -138,7 +350,7 @@ static void _done_io(struct osd_request *or, void *p)
 	kref_put(&ios->kref, _last_io);
 }
 
-static int ore_io_execute(struct ore_io_state *ios)
+int ore_io_execute(struct ore_io_state *ios)
 {
 	DECLARE_COMPLETION_ONSTACK(wait);
 	bool sync = (ios->done == NULL);
@@ -198,7 +410,7 @@ static void _clear_bio(struct bio *bio)
 	}
 }
 
-int ore_check_io(struct ore_io_state *ios, u64 *resid)
+int ore_check_io(struct ore_io_state *ios, ore_on_dev_error on_dev_error)
 {
 	enum osd_err_priority acumulated_osd_err = 0;
 	int acumulated_lin_err = 0;
@@ -206,7 +418,8 @@ int ore_check_io(struct ore_io_state *ios, u64 *resid)
 
 	for (i = 0; i < ios->numdevs; i++) {
 		struct osd_sense_info osi;
-		struct osd_request *or = ios->per_dev[i].or;
+		struct ore_per_dev_state *per_dev = &ios->per_dev[i];
+		struct osd_request *or = per_dev->or;
 		int ret;
 
 		if (unlikely(!or))
@@ -218,29 +431,31 @@ int ore_check_io(struct ore_io_state *ios, u64 *resid)
 
 		if (OSD_ERR_PRI_CLEAR_PAGES == osi.osd_err_pri) {
 			/* start read offset passed endof file */
-			_clear_bio(ios->per_dev[i].bio);
+			_clear_bio(per_dev->bio);
 			ORE_DBGMSG("start read offset passed end of file "
 				"offset=0x%llx, length=0x%llx\n",
-				_LLU(ios->per_dev[i].offset),
-				_LLU(ios->per_dev[i].length));
+				_LLU(per_dev->offset),
+				_LLU(per_dev->length));
 
 			continue; /* we recovered */
 		}
 
+		if (on_dev_error) {
+			u64 residual = ios->reading ?
+					or->in.residual : or->out.residual;
+			u64 offset = (ios->offset + ios->length) - residual;
+			struct ore_dev *od = ios->oc->ods[
+					per_dev->dev - ios->oc->first_dev];
+
+			on_dev_error(ios, od, per_dev->dev, osi.osd_err_pri,
+				     offset, residual);
+		}
 		if (osi.osd_err_pri >= acumulated_osd_err) {
 			acumulated_osd_err = osi.osd_err_pri;
 			acumulated_lin_err = ret;
 		}
 	}
 
-	/* TODO: raid specific residual calculations */
-	if (resid) {
-		if (likely(!acumulated_lin_err))
-			*resid = 0;
-		else
-			*resid = ios->length;
-	}
-
 	return acumulated_lin_err;
 }
 EXPORT_SYMBOL(ore_check_io);
@@ -248,61 +463,65 @@ EXPORT_SYMBOL(ore_check_io);
 /*
  * L - logical offset into the file
  *
- * U - The number of bytes in a stripe within a group
+ * D - number of Data devices
+ *	D = group_width - parity
  *
- *	U = stripe_unit * group_width
+ * U - The number of bytes in a stripe within a group
+ *	U =  stripe_unit * D
  *
  * T - The number of bytes striped within a group of component objects
  *     (before advancing to the next group)
- *
- *	T = stripe_unit * group_width * group_depth
+ *	T = U * group_depth
  *
  * S - The number of bytes striped across all component objects
  *     before the pattern repeats
+ *	S = T * group_count
  *
- *	S = stripe_unit * group_width * group_depth * group_count
- *
- * M - The "major" (i.e., across all components) stripe number
- *
+ * M - The "major" (i.e., across all components) cycle number
  *	M = L / S
  *
- * G - Counts the groups from the beginning of the major stripe
- *
+ * G - Counts the groups from the beginning of the major cycle
  *	G = (L - (M * S)) / T	[or (L % S) / T]
  *
  * H - The byte offset within the group
- *
  *	H = (L - (M * S)) % T	[or (L % S) % T]
  *
  * N - The "minor" (i.e., across the group) stripe number
- *
  *	N = H / U
  *
  * C - The component index coresponding to L
  *
- *	C = (H - (N * U)) / stripe_unit + G * group_width
- *	[or (L % U) / stripe_unit + G * group_width]
+ *	C = (H - (N * U)) / stripe_unit + G * D
+ *	[or (L % U) / stripe_unit + G * D]
  *
  * O - The component offset coresponding to L
- *
  *	O = L % stripe_unit + N * stripe_unit + M * group_depth * stripe_unit
+ *
+ * LCMdP – Parity cycle: Lowest Common Multiple of group_width, parity
+ *          divide by parity
+ *	LCMdP = lcm(group_width, parity) / parity
+ *
+ * R - The parity Rotation stripe
+ *     (Note parity cycle always starts at a group's boundary)
+ *	R = N % LCMdP
+ *
+ * I = the first parity device index
+ *	I = (group_width + group_width - R*parity - parity) % group_width
+ *
+ * Craid - The component index Rotated
+ *	Craid = (group_width + C - R*parity) % group_width
+ *      (We add the group_width to avoid negative numbers modulo math)
  */
-struct _striping_info {
-	u64 obj_offset;
-	u64 group_length;
-	u64 M; /* for truncate */
-	unsigned dev;
-	unsigned unit_off;
-};
-
-static void _calc_stripe_info(struct ore_layout *layout, u64 file_offset,
-			      struct _striping_info *si)
+void ore_calc_stripe_info(struct ore_layout *layout, u64 file_offset,
+			  u64 length, struct ore_striping_info *si)
 {
 	u32	stripe_unit = layout->stripe_unit;
 	u32	group_width = layout->group_width;
 	u64	group_depth = layout->group_depth;
+	u32	parity      = layout->parity;
 
-	u32	U = stripe_unit * group_width;
+	u32	D = group_width - parity;
+	u32	U = D * stripe_unit;
 	u64	T = U * group_depth;
 	u64	S = T * layout->group_count;
 	u64	M = div64_u64(file_offset, S);
@@ -318,39 +537,65 @@ static void _calc_stripe_info(struct ore_layout *layout, u64 file_offset,
 	u32	N = div_u64(H, U);
 
 	/* "H - (N * U)" is just "H % U" so it's bound to u32 */
-	si->dev = (u32)(H - (N * U)) / stripe_unit + G * group_width;
-	si->dev *= layout->mirrors_p1;
+	u32	C = (u32)(H - (N * U)) / stripe_unit + G * 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->group_length = T - H;
+	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->bytes_in_stripe = U;
+		si->first_stripe_start = M * S + G * T + N * U;
+	} else {
+		/* Make the math correct see _prepare_one_group */
+		si->par_dev = group_width;
+		si->dev = C;
+	}
+
+	si->dev *= layout->mirrors_p1;
+	si->par_dev *= layout->mirrors_p1;
+	si->offset = file_offset;
+	si->length = T - H;
+	if (si->length > length)
+		si->length = length;
 	si->M = M;
 }
+EXPORT_SYMBOL(ore_calc_stripe_info);
 
-static int _add_stripe_unit(struct ore_io_state *ios,  unsigned *cur_pg,
-		unsigned pgbase, struct ore_per_dev_state *per_dev,
-		int cur_len)
+int _ore_add_stripe_unit(struct ore_io_state *ios,  unsigned *cur_pg,
+			 unsigned pgbase, struct page **pages,
+			 struct ore_per_dev_state *per_dev, int cur_len)
 {
 	unsigned pg = *cur_pg;
 	struct request_queue *q =
 			osd_request_queue(_ios_od(ios, per_dev->dev));
-
-	per_dev->length += cur_len;
+	unsigned len = cur_len;
+	int ret;
 
 	if (per_dev->bio == NULL) {
 		unsigned pages_in_stripe = ios->layout->group_width *
 					(ios->layout->stripe_unit / PAGE_SIZE);
-		unsigned bio_size = (ios->nr_pages + pages_in_stripe) /
-						ios->layout->group_width;
+		unsigned nr_pages = ios->nr_pages * ios->layout->group_width /
+					(ios->layout->group_width -
+					 ios->layout->parity);
+		unsigned bio_size = (nr_pages + pages_in_stripe) /
+					ios->layout->group_width;
 
 		per_dev->bio = bio_kmalloc(GFP_KERNEL, bio_size);
 		if (unlikely(!per_dev->bio)) {
 			ORE_DBGMSG("Failed to allocate BIO size=%u\n",
 				     bio_size);
-			return -ENOMEM;
+			ret = -ENOMEM;
+			goto out;
 		}
 	}
 
@@ -358,64 +603,90 @@ static int _add_stripe_unit(struct ore_io_state *ios,  unsigned *cur_pg,
 		unsigned pglen = min_t(unsigned, PAGE_SIZE - pgbase, cur_len);
 		unsigned added_len;
 
-		BUG_ON(ios->nr_pages <= pg);
 		cur_len -= pglen;
 
-		added_len = bio_add_pc_page(q, per_dev->bio, ios->pages[pg],
+		added_len = bio_add_pc_page(q, per_dev->bio, pages[pg],
 					    pglen, pgbase);
-		if (unlikely(pglen != added_len))
-			return -ENOMEM;
+		if (unlikely(pglen != added_len)) {
+			ORE_DBGMSG("Failed bio_add_pc_page bi_vcnt=%u\n",
+				   per_dev->bio->bi_vcnt);
+			ret = -ENOMEM;
+			goto out;
+		}
+		_add_stripe_page(ios->sp2d, &ios->si, pages[pg]);
+
 		pgbase = 0;
 		++pg;
 	}
 	BUG_ON(cur_len);
 
+	per_dev->length += len;
 	*cur_pg = pg;
-	return 0;
+	ret = 0;
+out:	/* we fail the complete unit on an error eg don't advance
+	 * per_dev->length and cur_pg. This means that we might have a bigger
+	 * bio than the CDB requested length (per_dev->length). That's fine
+	 * only the oposite is fatal.
+	 */
+	return ret;
 }
 
-static int _prepare_one_group(struct ore_io_state *ios, u64 length,
-			      struct _striping_info *si)
+static int _prepare_for_striping(struct ore_io_state *ios)
 {
+	struct ore_striping_info *si = &ios->si;
 	unsigned stripe_unit = ios->layout->stripe_unit;
 	unsigned mirrors_p1 = ios->layout->mirrors_p1;
-	unsigned devs_in_group = ios->layout->group_width * mirrors_p1;
+	unsigned group_width = ios->layout->group_width;
+	unsigned devs_in_group = group_width * mirrors_p1;
 	unsigned dev = si->dev;
 	unsigned first_dev = dev - (dev % devs_in_group);
-	unsigned max_comp = ios->numdevs ? ios->numdevs - mirrors_p1 : 0;
+	unsigned dev_order;
 	unsigned cur_pg = ios->pages_consumed;
+	u64 length = ios->length;
 	int ret = 0;
 
+	if (!ios->pages) {
+		ios->numdevs = ios->layout->mirrors_p1;
+		return 0;
+	}
+
+	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) {
-		struct ore_per_dev_state *per_dev = &ios->per_dev[dev];
+		unsigned comp = dev - first_dev;
+		struct ore_per_dev_state *per_dev = &ios->per_dev[comp];
 		unsigned cur_len, page_off = 0;
 
 		if (!per_dev->length) {
 			per_dev->dev = dev;
-			if (dev < si->dev) {
-				per_dev->offset = si->obj_offset + stripe_unit -
-								   si->unit_off;
-				cur_len = stripe_unit;
-			} else if (dev == si->dev) {
+			if (dev == si->dev) {
+				WARN_ON(dev == si->par_dev);
 				per_dev->offset = si->obj_offset;
 				cur_len = stripe_unit - si->unit_off;
 				page_off = si->unit_off & ~PAGE_MASK;
 				BUG_ON(page_off && (page_off != ios->pgbase));
-			} else { /* dev > si->dev */
-				per_dev->offset = si->obj_offset - si->unit_off;
+			} 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;
 				cur_len = stripe_unit;
 			}
-
-			if (max_comp < dev)
-				max_comp = dev;
 		} else {
 			cur_len = stripe_unit;
 		}
 		if (cur_len >= length)
 			cur_len = length;
 
-		ret = _add_stripe_unit(ios, &cur_pg, page_off , per_dev,
-				       cur_len);
+		ret = _ore_add_stripe_unit(ios, &cur_pg, page_off, ios->pages,
+					   per_dev, cur_len);
 		if (unlikely(ret))
 			goto out;
 
@@ -423,60 +694,60 @@ static int _prepare_one_group(struct ore_io_state *ios, u64 length,
 		dev = (dev % devs_in_group) + first_dev;
 
 		length -= cur_len;
-	}
-out:
-	ios->numdevs = max_comp + mirrors_p1;
-	ios->pages_consumed = cur_pg;
-	return ret;
-}
-
-static int _prepare_for_striping(struct ore_io_state *ios)
-{
-	u64 length = ios->length;
-	u64 offset = ios->offset;
-	struct _striping_info si;
-	int ret = 0;
 
-	if (!ios->pages) {
-		if (ios->kern_buff) {
-			struct ore_per_dev_state *per_dev = &ios->per_dev[0];
+		si->cur_comp = (si->cur_comp + 1) % group_width;
+		if (unlikely((dev == si->par_dev) || (!length && ios->sp2d))) {
+			if (!length && ios->sp2d) {
+				/* If we are writing and this is the very last
+				 * 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;
+			}
 
-			_calc_stripe_info(ios->layout, ios->offset, &si);
-			per_dev->offset = si.obj_offset;
-			per_dev->dev = si.dev;
+			ret = _ore_add_parity_unit(ios, si, per_dev, cur_len);
+			if (unlikely(ret))
+					goto out;
 
-			/* no cross device without page array */
-			BUG_ON((ios->layout->group_width > 1) &&
-			       (si.unit_off + ios->length >
-				ios->layout->stripe_unit));
+			/* Rotate next par_dev backwards with wraping */
+			si->par_dev = (devs_in_group + si->par_dev -
+				       ios->layout->parity * mirrors_p1) %
+				      devs_in_group + first_dev;
+			/* Next stripe, start fresh */
+			si->cur_comp = 0;
+			si->cur_pg = 0;
 		}
-		ios->numdevs = ios->layout->mirrors_p1;
-		return 0;
-	}
-
-	while (length) {
-		_calc_stripe_info(ios->layout, offset, &si);
-
-		if (length < si.group_length)
-			si.group_length = length;
-
-		ret = _prepare_one_group(ios, si.group_length, &si);
-		if (unlikely(ret))
-			goto out;
-
-		offset += si.group_length;
-		length -= si.group_length;
 	}
-
 out:
-	return ret;
+	ios->numdevs = devs_in_group;
+	ios->pages_consumed = cur_pg;
+	if (unlikely(ret)) {
+		if (length == ios->length)
+			return ret;
+		else
+			ios->length -= length;
+	}
+	return 0;
 }
 
 int ore_create(struct ore_io_state *ios)
 {
 	int i, ret;
 
-	for (i = 0; i < ios->comps->numdevs; i++) {
+	for (i = 0; i < ios->oc->numdevs; i++) {
 		struct osd_request *or;
 
 		or = osd_start_request(_ios_od(ios, i), GFP_KERNEL);
@@ -501,7 +772,7 @@ int ore_remove(struct ore_io_state *ios)
 {
 	int i, ret;
 
-	for (i = 0; i < ios->comps->numdevs; i++) {
+	for (i = 0; i < ios->oc->numdevs; i++) {
 		struct osd_request *or;
 
 		or = osd_start_request(_ios_od(ios, i), GFP_KERNEL);
@@ -543,7 +814,6 @@ static int _write_mirror(struct ore_io_state *ios, int cur_comp)
 			goto out;
 		}
 		per_dev->or = or;
-		per_dev->offset = master_dev->offset;
 
 		if (ios->pages) {
 			struct bio *bio;
@@ -562,6 +832,7 @@ static int _write_mirror(struct ore_io_state *ios, int cur_comp)
 				__bio_clone(bio, master_dev->bio);
 				bio->bi_bdev = NULL;
 				bio->bi_next = NULL;
+				per_dev->offset = master_dev->offset;
 				per_dev->length = master_dev->length;
 				per_dev->bio =  bio;
 				per_dev->dev = dev;
@@ -579,7 +850,15 @@ static int _write_mirror(struct ore_io_state *ios, int cur_comp)
 				     _LLU(per_dev->offset),
 				     _LLU(per_dev->length), dev);
 		} else if (ios->kern_buff) {
-			ret = osd_req_write_kern(or, _ios_obj(ios, dev),
+			per_dev->offset = ios->si.obj_offset;
+			per_dev->dev = ios->si.dev + dev;
+
+			/* no cross device without page array */
+			BUG_ON((ios->layout->group_width > 1) &&
+			       (ios->si.unit_off + ios->length >
+				ios->layout->stripe_unit));
+
+			ret = osd_req_write_kern(or, _ios_obj(ios, per_dev->dev),
 						 per_dev->offset,
 						 ios->kern_buff, ios->length);
 			if (unlikely(ret))
@@ -588,7 +867,7 @@ static int _write_mirror(struct ore_io_state *ios, int cur_comp)
 				      "length=0x%llx dev=%d\n",
 				     _LLU(_ios_obj(ios, dev)->id),
 				     _LLU(per_dev->offset),
-				     _LLU(ios->length), dev);
+				     _LLU(ios->length), per_dev->dev);
 		} else {
 			osd_req_set_attributes(or, _ios_obj(ios, dev));
 			ORE_DBGMSG2("obj(0x%llx) set_attributes=%d dev=%d\n",
@@ -614,6 +893,14 @@ int ore_write(struct ore_io_state *ios)
 	int i;
 	int ret;
 
+	if (unlikely(ios->sp2d && !ios->r4w)) {
+		/* A library is attempting a RAID-write without providing
+		 * a pages lock interface.
+		 */
+		WARN_ON_ONCE(1);
+		return -ENOTSUPP;
+	}
+
 	ret = _prepare_for_striping(ios);
 	if (unlikely(ret))
 		return ret;
@@ -629,7 +916,7 @@ int ore_write(struct ore_io_state *ios)
 }
 EXPORT_SYMBOL(ore_write);
 
-static int _read_mirror(struct ore_io_state *ios, unsigned cur_comp)
+int _ore_read_mirror(struct ore_io_state *ios, unsigned cur_comp)
 {
 	struct osd_request *or;
 	struct ore_per_dev_state *per_dev = &ios->per_dev[cur_comp];
@@ -648,22 +935,27 @@ static int _read_mirror(struct ore_io_state *ios, unsigned cur_comp)
 	per_dev->or = or;
 
 	if (ios->pages) {
-		osd_req_read(or, obj, per_dev->offset,
-				per_dev->bio, per_dev->length);
+		if (per_dev->cur_sg) {
+			/* finalize the last sg_entry */
+			_ore_add_sg_seg(per_dev, 0, false);
+			if (unlikely(!per_dev->cur_sg))
+				return 0; /* Skip parity only device */
+
+			osd_req_read_sg(or, obj, per_dev->bio,
+					per_dev->sglist, per_dev->cur_sg);
+		} else {
+			/* The no raid case */
+			osd_req_read(or, obj, per_dev->offset,
+				     per_dev->bio, per_dev->length);
+		}
+
 		ORE_DBGMSG("read(0x%llx) offset=0x%llx length=0x%llx"
-			     " dev=%d\n", _LLU(obj->id),
+			     " dev=%d sg_len=%d\n", _LLU(obj->id),
 			     _LLU(per_dev->offset), _LLU(per_dev->length),
-			     first_dev);
-	} else if (ios->kern_buff) {
-		int ret = osd_req_read_kern(or, obj, per_dev->offset,
-					    ios->kern_buff, ios->length);
-		ORE_DBGMSG2("read_kern(0x%llx) offset=0x%llx "
-			      "length=0x%llx dev=%d ret=>%d\n",
-			      _LLU(obj->id), _LLU(per_dev->offset),
-			      _LLU(ios->length), first_dev, ret);
-		if (unlikely(ret))
-			return ret;
+			     first_dev, per_dev->cur_sg);
 	} else {
+		BUG_ON(ios->kern_buff);
+
 		osd_req_get_attributes(or, obj);
 		ORE_DBGMSG2("obj(0x%llx) get_attributes=%d dev=%d\n",
 			      _LLU(obj->id),
@@ -688,7 +980,7 @@ int ore_read(struct ore_io_state *ios)
 		return ret;
 
 	for (i = 0; i < ios->numdevs; i += ios->layout->mirrors_p1) {
-		ret = _read_mirror(ios, i);
+		ret = _ore_read_mirror(ios, i);
 		if (unlikely(ret))
 			return ret;
 	}
@@ -744,31 +1036,29 @@ static int _truncate_mirrors(struct ore_io_state *ios, unsigned cur_comp,
 }
 
 struct _trunc_info {
-	struct _striping_info si;
+	struct ore_striping_info si;
 	u64 prev_group_obj_off;
 	u64 next_group_obj_off;
 
 	unsigned first_group_dev;
 	unsigned nex_group_dev;
-	unsigned max_devs;
 };
 
-void _calc_trunk_info(struct ore_layout *layout, u64 file_offset,
-		       struct _trunc_info *ti)
+static void _calc_trunk_info(struct ore_layout *layout, u64 file_offset,
+			     struct _trunc_info *ti)
 {
 	unsigned stripe_unit = layout->stripe_unit;
 
-	_calc_stripe_info(layout, file_offset, &ti->si);
+	ore_calc_stripe_info(layout, file_offset, 0, &ti->si);
 
 	ti->prev_group_obj_off = ti->si.M * stripe_unit;
 	ti->next_group_obj_off = ti->si.M ? (ti->si.M - 1) * stripe_unit : 0;
 
 	ti->first_group_dev = ti->si.dev - (ti->si.dev % layout->group_width);
 	ti->nex_group_dev = ti->first_group_dev + layout->group_width;
-	ti->max_devs = layout->group_width * layout->group_count;
 }
 
-int ore_truncate(struct ore_layout *layout, struct ore_components *comps,
+int ore_truncate(struct ore_layout *layout, struct ore_components *oc,
 		   u64 size)
 {
 	struct ore_io_state *ios;
@@ -779,22 +1069,22 @@ int ore_truncate(struct ore_layout *layout, struct ore_components *comps,
 	struct _trunc_info ti;
 	int i, ret;
 
-	ret = ore_get_io_state(layout, comps, &ios);
+	ret = ore_get_io_state(layout, oc, &ios);
 	if (unlikely(ret))
 		return ret;
 
 	_calc_trunk_info(ios->layout, size, &ti);
 
-	size_attrs = kcalloc(ti.max_devs, sizeof(*size_attrs),
+	size_attrs = kcalloc(ios->oc->numdevs, sizeof(*size_attrs),
 			     GFP_KERNEL);
 	if (unlikely(!size_attrs)) {
 		ret = -ENOMEM;
 		goto out;
 	}
 
-	ios->numdevs = ios->comps->numdevs;
+	ios->numdevs = ios->oc->numdevs;
 
-	for (i = 0; i < ti.max_devs; ++i) {
+	for (i = 0; i < ios->numdevs; ++i) {
 		struct exofs_trunc_attr *size_attr = &size_attrs[i];
 		u64 obj_size;
 
@@ -815,7 +1105,7 @@ int ore_truncate(struct ore_layout *layout, struct ore_components *comps,
 		size_attr->attr.val_ptr = &size_attr->newsize;
 
 		ORE_DBGMSG("trunc(0x%llx) obj_offset=0x%llx dev=%d\n",
-			     _LLU(comps->comps->obj.id), _LLU(obj_size), i);
+			     _LLU(oc->comps->obj.id), _LLU(obj_size), i);
 		ret = _truncate_mirrors(ios, i * ios->layout->mirrors_p1,
 					&size_attr->attr);
 		if (unlikely(ret))

fs/exofs/ore_raid.c

+/*
+ * Copyright (C) 2011
+ * Boaz Harrosh <bharrosh@panasas.com>
+ *
+ * This file is part of the objects raid engine (ore).
+ *
+ * It is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as published
+ * by the Free Software Foundation.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with "ore". If not, write to the Free Software Foundation, Inc:
+ *	"Free Software Foundation <info@fsf.org>"
+ */
+
+#include <linux/gfp.h>
+#include <linux/async_tx.h>
+
+#include "ore_raid.h"
+
+#undef ORE_DBGMSG2
+#define ORE_DBGMSG2 ORE_DBGMSG
+
+struct page *_raid_page_alloc(void)
+{
+	return alloc_page(GFP_KERNEL);
+}
+
+void _raid_page_free(struct page *p)
+{
+	__free_page(p);
+}
+
+/* This struct is forward declare in ore_io_state, but is private to here.
+ * It is put on ios->sp2d for RAID5/6 writes only. See _gen_xor_unit.
+ *
+ * __stripe_pages_2d is a 2d array of pages, and it is also a corner turn.
+ * Ascending page index access is sp2d(p-minor, c-major). But storage is
+ * sp2d[p-minor][c-major], so it can be properlly presented to the async-xor
+ * API.
+ */
+struct __stripe_pages_2d {
+	/* Cache some hot path repeated calculations */
+	unsigned parity;
+	unsigned data_devs;
+	unsigned pages_in_unit;
+
+	bool needed ;
+
+	/* Array size is pages_in_unit (layout->stripe_unit / PAGE_SIZE) */
+	struct __1_page_stripe {
+		bool alloc;
+		unsigned write_count;
+		struct async_submit_ctl submit;
+		struct dma_async_tx_descriptor *tx;
+
+		/* The size of this array is data_devs + parity */
+		struct page **pages;
+		struct page **scribble;
+		/* bool array, size of this array is data_devs */
+		char *page_is_read;
+	} _1p_stripes[];
+};
+
+/* This can get bigger then a page. So support multiple page allocations
+ * _sp2d_free should be called even if _sp2d_alloc fails (by returning
+ * none-zero).
+ */
+static int _sp2d_alloc(unsigned pages_in_unit, unsigned group_width,
+		       unsigned parity, struct __stripe_pages_2d **psp2d)
+{
+	struct __stripe_pages_2d *sp2d;
+	unsigned data_devs = group_width - parity;
+	struct _alloc_all_bytes {
+		struct __alloc_stripe_pages_2d {
+			struct __stripe_pages_2d sp2d;
+			struct __1_page_stripe _1p_stripes[pages_in_unit];
+		} __asp2d;
+		struct __alloc_1p_arrays {
+			struct page *pages[group_width];
+			struct page *scribble[group_width];
+			char page_is_read[data_devs];
+		} __a1pa[pages_in_unit];
+	} *_aab;
+	struct __alloc_1p_arrays *__a1pa;
+	struct __alloc_1p_arrays *__a1pa_end;
+	const unsigned sizeof__a1pa = sizeof(_aab->__a1pa[0]);
+	unsigned num_a1pa, alloc_size, i;
+
+	/* FIXME: check these numbers in ore_verify_layout */
+	BUG_ON(sizeof(_aab->__asp2d) > PAGE_SIZE);
+	BUG_ON(sizeof__a1pa > PAGE_SIZE);
+
+	if (sizeof(*_aab) > PAGE_SIZE) {
+		num_a1pa = (PAGE_SIZE - sizeof(_aab->__asp2d)) / sizeof__a1pa;
+		alloc_size = sizeof(_aab->__asp2d) + sizeof__a1pa * num_a1pa;
+	} else {
+		num_a1pa = pages_in_unit;
+		alloc_size = sizeof(*_aab);
+	}
+
+	_aab = kzalloc(alloc_size, GFP_KERNEL);
+	if (unlikely(!_aab)) {
+		ORE_DBGMSG("!! Failed to alloc sp2d size=%d\n", alloc_size);
+		return -ENOMEM;
+	}
+
+	sp2d = &_aab->__asp2d.sp2d;
+	*psp2d = sp2d; /* From here Just call _sp2d_free */
+
+	__a1pa = _aab->__a1pa;
+	__a1pa_end = __a1pa + num_a1pa;
+
+	for (i = 0; i < pages_in_unit; ++i) {
+		if (unlikely(__a1pa >= __a1pa_end)) {
+			num_a1pa = min_t(unsigned, PAGE_SIZE / sizeof__a1pa,
+							pages_in_unit - i);
+
+			__a1pa = kzalloc(num_a1pa * sizeof__a1pa, GFP_KERNEL);
+			if (unlikely(!__a1pa)) {
+				ORE_DBGMSG("!! Failed to _alloc_1p_arrays=%d\n",
+					   num_a1pa);
+				return -ENOMEM;
+			}
+			__a1pa_end = __a1pa + num_a1pa;
+			/* First *pages is marked for kfree of the buffer */
+			sp2d->_1p_stripes[i].alloc = true;
+		}
+
+		sp2d->_1p_stripes[i].pages = __a1pa->pages;
+		sp2d->_1p_stripes[i].scribble = __a1pa->scribble ;
+		sp2d->_1p_stripes[i].page_is_read = __a1pa->page_is_read;
+		++__a1pa;
+	}
+
+	sp2d->parity = parity;
+	sp2d->data_devs = data_devs;
+	sp2d->pages_in_unit = pages_in_unit;
+	return 0;
+}
+
+static void _sp2d_reset(struct __stripe_pages_2d *sp2d,
+			const struct _ore_r4w_op *r4w, void *priv)
+{
+	unsigned data_devs = sp2d->data_devs;
+	unsigned group_width = data_devs + sp2d->parity;
+	unsigned p;
+
+	if (!sp2d->needed)
+		return;
+
+	for (p = 0; p < sp2d->pages_in_unit; p++) {
+		struct __1_page_stripe *_1ps = &sp2d->_1p_stripes[p];
+
+		if (_1ps->write_count < group_width) {
+			unsigned c;
+
+			for (c = 0; c < data_devs; c++)
+				if (_1ps->page_is_read[c]) {
+					struct page *page = _1ps->pages[c];
+
+					r4w->put_page(priv, page);
+					_1ps->page_is_read[c] = false;
+				}
+		}
+
+		memset(_1ps->pages, 0, group_width * sizeof(*_1ps->pages));
+		_1ps->write_count = 0;
+		_1ps->tx = NULL;
+	}
+
+	sp2d->needed = false;
+}
+
+static void _sp2d_free(struct __stripe_pages_2d *sp2d)
+{
+	unsigned i;
+
+	if (!sp2d)
+		return;
+
+	for (i = 0; i < sp2d->pages_in_unit; ++i) {
+		if (sp2d->_1p_stripes[i].alloc)
+			kfree(sp2d->_1p_stripes[i].pages);
+	}
+
+	kfree(sp2d);
+}
+
+static unsigned _sp2d_min_pg(struct __stripe_pages_2d *sp2d)
+{
+	unsigned p;
+
+	for (p = 0; p < sp2d->pages_in_unit; p++) {
+		struct __1_page_stripe *_1ps = &sp2d->_1p_stripes[p];
+
+		if (_1ps->write_count)
+			return p;
+	}
+
+	return ~0;
+}
+
+static unsigned _sp2d_max_pg(struct __stripe_pages_2d *sp2d)
+{
+	unsigned p;
+
+	for (p = sp2d->pages_in_unit - 1; p >= 0; --p) {
+		struct __1_page_stripe *_1ps = &sp2d->_1p_stripes[p];
+
+		if (_1ps->write_count)
+			return p;
+	}
+
+	return ~0;
+}
+
+static void _gen_xor_unit(struct __stripe_pages_2d *sp2d)
+{
+	unsigned p;
+	for (p = 0; p < sp2d->pages_in_unit; p++) {
+		struct __1_page_stripe *_1ps = &sp2d->_1p_stripes[p];
+
+		if (!_1ps->write_count)
+			continue;
+
+		init_async_submit(&_1ps->submit,
+			ASYNC_TX_XOR_ZERO_DST | ASYNC_TX_ACK,
+			NULL,
+			NULL, NULL,
+			(addr_conv_t *)_1ps->scribble);
+
+		/* TODO: raid6 */
+		_1ps->tx = async_xor(_1ps->pages[sp2d->data_devs], _1ps->pages,
+				     0, sp2d->data_devs, PAGE_SIZE,
+				     &_1ps->submit);
+	}
+
+	for (p = 0; p < sp2d->pages_in_unit; p++) {
+		struct __1_page_stripe *_1ps = &sp2d->_1p_stripes[p];
+		/* NOTE: We wait for HW synchronously (I don't have such HW
+		 * to test with.) Is parallelism needed with today's multi
+		 * cores?
+		 */
+		async_tx_issue_pending(_1ps->tx);
+	}
+}
+
+void _ore_add_stripe_page(struct __stripe_pages_2d *sp2d,
+		       struct ore_striping_info *si, struct page *page)
+{
+	struct __1_page_stripe *_1ps;
+
+	sp2d->needed = true;
+
+	_1ps = &sp2d->_1p_stripes[si->cur_pg];
+	_1ps->pages[si->cur_comp] = page;
+	++_1ps->write_count;
+
+	si->cur_pg = (si->cur_pg + 1) % sp2d->pages_in_unit;
+	/* si->cur_comp is advanced outside at main loop */
+}
+
+void _ore_add_sg_seg(struct ore_per_dev_state *per_dev, unsigned cur_len,
+		     bool not_last)
+{
+	struct osd_sg_entry *sge;
+
+	ORE_DBGMSG("dev=%d cur_len=0x%x not_last=%d cur_sg=%d "
+		     "offset=0x%llx length=0x%x last_sgs_total=0x%x\n",
+		     per_dev->dev, cur_len, not_last, per_dev->cur_sg,
+		     _LLU(per_dev->offset), per_dev->length,
+		     per_dev->last_sgs_total);
+
+	if (!per_dev->cur_sg) {
+		sge = per_dev->sglist;
+
+		/* First time we prepare two entries */
+		if (per_dev->length) {
+			++per_dev->cur_sg;
+			sge->offset = per_dev->offset;
+			sge->len = per_dev->length;
+		} else {
+			/* Here the parity is the first unit of this object.
+			 * This happens every time we reach a parity device on
+			 * the same stripe as the per_dev->offset. We need to
+			 * just skip this unit.
+			 */
+			per_dev->offset += cur_len;
+			return;
+		}
+	} else {
+		/* finalize the last one */
+		sge = &per_dev->sglist[per_dev->cur_sg - 1];
+		sge->len = per_dev->length - per_dev->last_sgs_total;
+	}
+
+	if (not_last) {
+		/* Partly prepare the next one */
+		struct osd_sg_entry *next_sge = sge + 1;
+
+		++per_dev->cur_sg;
+		next_sge->offset = sge->offset + sge->len + cur_len;
+		/* Save cur len so we know how mutch was added next time */
+		per_dev->last_sgs_total = per_dev->length;
+		next_sge->len = 0;
+	} else if (!sge->len) {
+		/* Optimize for when the last unit is a parity */
+		--per_dev->cur_sg;
+	}
+}
+
+static int _alloc_read_4_write(struct ore_io_state *ios)
+{
+	struct ore_layout *layout = ios->layout;
+	int ret;
+	/* We want to only read those pages not in cache so worst case
+	 * is a stripe populated with every other page
+	 */
+	unsigned sgs_per_dev = ios->sp2d->pages_in_unit + 2;
+
+	ret = _ore_get_io_state(layout, ios->oc,
+				layout->group_width * layout->mirrors_p1,
+				sgs_per_dev, 0, &ios->ios_read_4_write);
+	return ret;
+}
+
+/* @si contains info of the to-be-inserted page. Update of @si should be
+ * maintained by caller. Specificaly si->dev, si->obj_offset, ...
+ */
+static int _add_to_read_4_write(struct ore_io_state *ios,
+				struct ore_striping_info *si, struct page *page)
+{
+	struct request_queue *q;
+	struct ore_per_dev_state *per_dev;
+	struct ore_io_state *read_ios;
+	unsigned first_dev = si->dev - (si->dev %
+			  (ios->layout->group_width * ios->layout->mirrors_p1));
+	unsigned comp = si->dev - first_dev;
+	unsigned added_len;
+
+	if (!ios->ios_read_4_write) {
+		int ret = _alloc_read_4_write(ios);
+
+		if (unlikely(ret))
+			return ret;
+	}
+
+	read_ios = ios->ios_read_4_write;
+	read_ios->numdevs = ios->layout->group_width * ios->layout->mirrors_p1;
+
+	per_dev = &read_ios->per_dev[comp];
+	if (!per_dev->length) {
+		per_dev->bio = bio_kmalloc(GFP_KERNEL,
+					   ios->sp2d->pages_in_unit);
+		if (unlikely(!per_dev->bio)) {
+			ORE_DBGMSG("Failed to allocate BIO size=%u\n",
+				     ios->sp2d->pages_in_unit);
+			return -ENOMEM;
+		}
+		per_dev->offset = si->obj_offset;
+		per_dev->dev = si->dev;
+	} else if (si->obj_offset != (per_dev->offset + per_dev->length)) {
+		u64 gap = si->obj_offset - (per_dev->offset + per_dev->length);
+
+		_ore_add_sg_seg(per_dev, gap, true);
+	}
+	q = osd_request_queue(ore_comp_dev(read_ios->oc, per_dev->dev));
+	added_len = bio_add_pc_page(q, per_dev->bio, page, PAGE_SIZE, 0);
+	if (unlikely(added_len != PAGE_SIZE)) {
+		ORE_DBGMSG("Failed to bio_add_pc_page bi_vcnt=%d\n",
+			      per_dev->bio->bi_vcnt);
+		return -ENOMEM;
+	}
+
+	per_dev->length += PAGE_SIZE;
+	return 0;
+}
+
+static void _mark_read4write_pages_uptodate(struct ore_io_state *ios, int ret)
+{
+	struct bio_vec *bv;
+	unsigned i, d;
+
+	/* loop on all devices all pages */
+	for (d = 0; d < ios->numdevs; d++) {
+		struct bio *bio = ios->per_dev[d].bio;
+
+		if (!bio)
+			continue;
+
+		__bio_for_each_segment(bv, bio, i, 0) {
+			struct page *page = bv->bv_page;
+
+			SetPageUptodate(page);
+			if (PageError(page))
+				ClearPageError(page);
+		}
+	}
+}
+
+/* read_4_write is hacked to read the start of the first stripe and/or
+ * the end of the last stripe. If needed, with an sg-gap at each device/page.
+ * It is assumed to be called after the to_be_written pages of the first stripe
+ * are populating ios->sp2d[][]
+ *
+ * NOTE: We call ios->r4w->lock_fn for all pages needed for parity calculations
+ * These pages are held at sp2d[p].pages[c] but with
+ * sp2d[p].page_is_read[c] = true. At _sp2d_reset these pages are
+ * ios->r4w->lock_fn(). The ios->r4w->lock_fn might signal that the page is
+ * @uptodate=true, so we don't need to read it, only unlock, after IO.
+ *
+ * TODO: The read_4_write should calc a need_to_read_pages_count, if bigger then
+ * to-be-written count, we should consider the xor-in-place mode.
+ * need_to_read_pages_count is the actual number of pages not present in cache.
+ * maybe "devs_in_group - ios->sp2d[p].write_count" is a good enough
+ * approximation? In this mode the read pages are put in the empty places of
+ * ios->sp2d[p][*], xor is calculated the same way. These pages are
+ * allocated/freed and don't go through cache
+ */
+static int _read_4_write(struct ore_io_state *ios)
+{
+	struct ore_io_state *ios_read;
+	struct ore_striping_info read_si;
+	struct __stripe_pages_2d *sp2d = ios->sp2d;
+	u64 offset = ios->si.first_stripe_start;
+	u64 last_stripe_end;
+	unsigned bytes_in_stripe = ios->si.bytes_in_stripe;
+	unsigned i, c, p, min_p = sp2d->pages_in_unit, max_p = -1;
+	int ret;
+
+	if (offset == ios->offset) /* Go to start collect $200 */
+		goto read_last_stripe;
+
+	min_p = _sp2d_min_pg(sp2d);
+	max_p = _sp2d_max_pg(sp2d);
+
+	for (c = 0; ; c++) {
+		ore_calc_stripe_info(ios->layout, offset, 0, &read_si);
+		read_si.obj_offset += min_p * PAGE_SIZE;
+		offset += min_p * PAGE_SIZE;
+		for (p = min_p; p <= max_p; p++) {
+			struct __1_page_stripe *_1ps = &sp2d->_1p_stripes[p];
+			struct page **pp = &_1ps->pages[c];
+			bool uptodate;
+
+			if (*pp)
+				/* to-be-written pages start here */
+				goto read_last_stripe;
+
+			*pp = ios->r4w->get_page(ios->private, offset,
+						 &uptodate);
+			if (unlikely(!*pp))
+				return -ENOMEM;
+
+			if (!uptodate)
+				_add_to_read_4_write(ios, &read_si, *pp);
+
+			/* Mark read-pages to be cache_released */
+			_1ps->page_is_read[c] = true;
+			read_si.obj_offset += PAGE_SIZE;
+			offset += PAGE_SIZE;
+		}
+		offset += (sp2d->pages_in_unit - p) * PAGE_SIZE;
+	}
+
+read_last_stripe:
+	offset = ios->offset + (ios->length + PAGE_SIZE - 1) /
+				PAGE_SIZE * PAGE_SIZE;
+	last_stripe_end = div_u64(offset + bytes_in_stripe - 1, bytes_in_stripe)
+				 * bytes_in_stripe;
+	if (offset == last_stripe_end) /* Optimize for the aligned case */
+		goto read_it;
+
+	ore_calc_stripe_info(ios->layout, offset, 0, &read_si);
+	p = read_si.unit_off / PAGE_SIZE;
+	c = _dev_order(ios->layout->group_width * ios->layout->mirrors_p1,
+		       ios->layout->mirrors_p1, read_si.par_dev, read_si.dev);
+
+	BUG_ON(ios->si.first_stripe_start + bytes_in_stripe != last_stripe_end);
+	/* unaligned IO must be within a single stripe */
+
+	if (min_p == sp2d->pages_in_unit) {
+		/* Didn't do it yet */
+		min_p = _sp2d_min_pg(sp2d);
+		max_p = _sp2d_max_pg(sp2d);
+	}
+
+	while (offset < last_stripe_end) {
+		struct __1_page_stripe *_1ps = &sp2d->_1p_stripes[p];
+
+		if ((min_p <= p) && (p <= max_p)) {
+			struct page *page;
+			bool uptodate;
+
+			BUG_ON(_1ps->pages[c]);
+			page = ios->r4w->get_page(ios->private, offset,
+						  &uptodate);
+			if (unlikely(!page))
+				return -ENOMEM;
+
+			_1ps->pages[c] = page;
+			/* Mark read-pages to be cache_released */
+			_1ps->page_is_read[c] = true;
+			if (!uptodate)
+				_add_to_read_4_write(ios, &read_si, page);
+		}
+
+		offset += PAGE_SIZE;
+		if (p == (sp2d->pages_in_unit - 1)) {
+			++c;
+			p = 0;
+			ore_calc_stripe_info(ios->layout, offset, 0, &read_si);
+		} else {
+			read_si.obj_offset += PAGE_SIZE;
+			++p;
+		}
+	}
+
+read_it:
+	ios_read = ios->ios_read_4_write;
+	if (!ios_read)
+		return 0;
+
+	/* FIXME: Ugly to signal _sbi_read_mirror that we have bio(s). Change
+	 * to check for per_dev->bio
+	 */
+	ios_read->pages = ios->pages;
+
+	/* Now read these devices */
+	for (i = 0; i < ios_read->numdevs; i += ios_read->layout->mirrors_p1) {
+		ret = _ore_read_mirror(ios_read, i);
+		if (unlikely(ret))
+			return ret;
+	}
+
+	ret = ore_io_execute(ios_read); /* Synchronus execution */
+	if (unlikely(ret)) {
+		ORE_DBGMSG("!! ore_io_execute => %d\n", ret);
+		return ret;
+	}
+
+	_mark_read4write_pages_uptodate(ios_read, ret);
+	return 0;
+}
+
+/* In writes @cur_len means length left. .i.e cur_len==0 is the last parity U */
+int _ore_add_parity_unit(struct ore_io_state *ios,
+			    struct ore_striping_info *si,
+			    struct ore_per_dev_state *per_dev,
+			    unsigned cur_len)
+{
+	if (ios->reading) {
+		BUG_ON(per_dev->cur_sg >= ios->sgs_per_dev);
+		_ore_add_sg_seg(per_dev, cur_len, true);
+	} else {
+		struct __stripe_pages_2d *sp2d = ios->sp2d;
+		struct page **pages = ios->parity_pages + ios->cur_par_page;
+		unsigned num_pages;
+		unsigned array_start = 0;
+		unsigned i;
+		int ret;
+
+		si->cur_pg = _sp2d_min_pg(sp2d);
+		num_pages  = _sp2d_max_pg(sp2d) + 1 - si->cur_pg;
+
+		if (!cur_len) /* If last stripe operate on parity comp */
+			si->cur_comp = sp2d->data_devs;
+
+		if (!per_dev->length) {
+			per_dev->offset += si->cur_pg * PAGE_SIZE;
+			/* If first stripe, Read in all read4write pages
+			 * (if needed) before we calculate the first parity.
+			 */
+			_read_4_write(ios);
+		}
+
+		for (i = 0; i < num_pages; i++) {
+			pages[i] = _raid_page_alloc();
+			if (unlikely(!pages[i]))
+				return -ENOMEM;
+
+			++(ios->cur_par_page);
+		}
+
+		BUG_ON(si->cur_comp != sp2d->data_devs);
+		BUG_ON(si->cur_pg + num_pages > sp2d->pages_in_unit);
+
+		ret = _ore_add_stripe_unit(ios,  &array_start, 0, pages,
+					   per_dev, num_pages * PAGE_SIZE);
+		if (unlikely(ret))
+			return ret;
+
+		/* TODO: raid6 if (last_parity_dev) */
+		_gen_xor_unit(sp2d);
+		_sp2d_reset(sp2d, ios->r4w, ios->private);
+	}
+	return 0;
+}
+
+int _ore_post_alloc_raid_stuff(struct ore_io_state *ios)
+{
+	struct ore_layout *layout = ios->layout;
+
+	if (ios->parity_pages) {
+		unsigned pages_in_unit = layout->stripe_unit / PAGE_SIZE;
+		unsigned stripe_size = ios->si.bytes_in_stripe;
+		u64 last_stripe, first_stripe;
+
+		if (_sp2d_alloc(pages_in_unit, layout->group_width,
+				layout->parity, &ios->sp2d)) {
+			return -ENOMEM;
+		}
+
+		BUG_ON(ios->offset % PAGE_SIZE);
+
+		/* Round io down to last full strip */
+		first_stripe = div_u64(ios->offset, stripe_size);
+		last_stripe = div_u64(ios->offset + ios->length, stripe_size);
+
+		/* If an IO spans more then a single stripe it must end at
+		 * a stripe boundary. The reminder at the end is pushed into the
+		 * next IO.
+		 */
+		if (last_stripe != first_stripe) {
+			ios->length = last_stripe * stripe_size - ios->offset;
+
+			BUG_ON(!ios->length);
+			ios->nr_pages = (ios->length + PAGE_SIZE - 1) /
+					PAGE_SIZE;
+			ios->si.length = ios->length; /*make it consistent */
+		}
+	}
+	return 0;
+}
+
+void _ore_free_raid_stuff(struct ore_io_state *ios)
+{
+	if (ios->sp2d) { /* writing and raid */
+		unsigned i;
+
+		for (i = 0; i < ios->cur_par_page; i++) {
+			struct page *page = ios->parity_pages[i];
+
+			if (page)
+				_raid_page_free(page);
+		}
+		if (ios->extra_part_alloc)
+			kfree(ios->parity_pages);
+		/* If IO returned an error pages might need unlocking */
+		_sp2d_reset(ios->sp2d, ios->r4w, ios->private);
+		_sp2d_free(ios->sp2d);
+	} else {
+		/* Will only be set if raid reading && sglist is big */
+		if (ios->extra_part_alloc)
+			kfree(ios->per_dev[0].sglist);
+	}
+	if (ios->ios_read_4_write)
+		ore_put_io_state(ios->ios_read_4_write);
+}

fs/exofs/ore_raid.h

+/*
+ * Copyright (C) from 2011
+ * Boaz Harrosh <bharrosh@panasas.com>
+ *
+ * This file is part of the objects raid engine (ore).
+ *
+ * It is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as published
+ * by the Free Software Foundation.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with "ore". If not, write to the Free Software Foundation, Inc:
+ *	"Free Software Foundation <info@fsf.org>"
+ */
+
+#include <scsi/osd_ore.h>
+
+#define ORE_ERR(fmt, a...) printk(KERN_ERR "ore: " fmt, ##a)
+
+#ifdef CONFIG_EXOFS_DEBUG
+#define ORE_DBGMSG(fmt, a...) \
+	printk(KERN_NOTICE "ore @%s:%d: " fmt, __func__, __LINE__, ##a)
+#else
+#define ORE_DBGMSG(fmt, a...) \
+	do { if (0) printk(fmt, ##a); } while (0)
+#endif
+
+/* u64 has problems with printk this will cast it to unsigned long long */
+#define _LLU(x) (unsigned long long)(x)
+
+#define ORE_DBGMSG2(M...) do {} while (0)
+/* #define ORE_DBGMSG2 ORE_DBGMSG */
+
+/* Calculate the component order in a stripe. eg the logical data unit
+ * address within the stripe of @dev given the @par_dev of this stripe.
+ */
+static inline unsigned _dev_order(unsigned devs_in_group, unsigned mirrors_p1,
+				  unsigned par_dev, unsigned dev)
+{
+	unsigned first_dev = dev - dev % devs_in_group;
+
+	dev -= first_dev;
+	par_dev -= first_dev;
+
+	if (devs_in_group == par_dev) /* The raid 0 case */
+		return dev / mirrors_p1;
+	/* raid4/5/6 case */
+	return ((devs_in_group + dev - par_dev - mirrors_p1) % devs_in_group) /
+	       mirrors_p1;
+}
+
+/* ios_raid.c stuff needed by ios.c */
+int _ore_post_alloc_raid_stuff(struct ore_io_state *ios);
+void _ore_free_raid_stuff(struct ore_io_state *ios);
+
+void _ore_add_sg_seg(struct ore_per_dev_state *per_dev, unsigned cur_len,
+		 bool not_last);
+int _ore_add_parity_unit(struct ore_io_state *ios, struct ore_striping_info *si,
+		     struct ore_per_dev_state *per_dev, unsigned cur_len);
+void _ore_add_stripe_page(struct __stripe_pages_2d *sp2d,
+		       struct ore_striping_info *si, struct page *page);
+static inline void _add_stripe_page(struct __stripe_pages_2d *sp2d,
+				struct ore_striping_info *si, struct page *page)
+{
+	if (!sp2d) /* Inline the fast path */
+		return; /* Hay no raid stuff */
+	_ore_add_stripe_page(sp2d, si, page);
+}
+
+/* ios.c stuff needed by ios_raid.c */
+int  _ore_get_io_state(struct ore_layout *layout,
+			struct ore_components *oc, unsigned numdevs,
+			unsigned sgs_per_dev, unsigned num_par_pages,
+			struct ore_io_state **pios);
+int _ore_add_stripe_unit(struct ore_io_state *ios,  unsigned *cur_pg,
+		unsigned pgbase, struct page **pages,
+		struct ore_per_dev_state *per_dev, int cur_len);
+int _ore_read_mirror(struct ore_io_state *ios, unsigned cur_comp);
+int ore_io_execute(struct ore_io_state *ios);

fs/exofs/super.c

@@ -266,7 +266,7 @@ static int __sbi_read_stats(struct exofs_sb_info *sbi)
 	struct ore_io_state *ios;
 	int ret;
 
-	ret = ore_get_io_state(&sbi->layout, &sbi->comps, &ios);
+	ret = ore_get_io_state(&sbi->layout, &sbi->oc, &ios);
 	if (unlikely(ret)) {
 		EXOFS_ERR("%s: ore_get_io_state failed.\n", __func__);
 		return ret;
@@ -321,7 +321,7 @@ int exofs_sbi_write_stats(struct exofs_sb_info *sbi)
 	struct ore_io_state *ios;
 	int ret;
 
-	ret = ore_get_io_state(&sbi->layout, &sbi->comps, &ios);
+	ret = ore_get_io_state(&sbi->layout, &sbi->oc, &ios);
 	if (unlikely(ret)) {
 		EXOFS_ERR("%s: ore_get_io_state failed.\n", __func__);
 		return ret;
@@ -355,12 +355,12 @@ static const struct export_operations exofs_export_ops;
 /*
  * Write the superblock to the OSD
  */
-int exofs_sync_fs(struct super_block *sb, int wait)
+static int exofs_sync_fs(struct super_block *sb, int wait)
 {
 	struct exofs_sb_info *sbi;
 	struct exofs_fscb *fscb;
 	struct ore_comp one_comp;
-	struct ore_components comps;
+	struct ore_components oc;
 	struct ore_io_state *ios;
 	int ret = -ENOMEM;
 
@@ -378,9 +378,9 @@ int exofs_sync_fs(struct super_block *sb, int wait)
 	 * the writeable info is set in exofs_sbi_write_stats() above.
 	 */
 
-	exofs_init_comps(&comps, &one_comp, sbi, EXOFS_SUPER_ID);
+	exofs_init_comps(&oc, &one_comp, sbi, EXOFS_SUPER_ID);
 
-	ret = ore_get_io_state(&sbi->layout, &comps, &ios);
+	ret = ore_get_io_state(&sbi->layout, &oc, &ios);
 	if (unlikely(ret))
 		goto out;
 
@@ -429,19 +429,20 @@ static void _exofs_print_device(const char *msg, const char *dev_path,
 		msg, dev_path ?: "", odi->osdname, _LLU(pid));
 }
 
-void exofs_free_sbi(struct exofs_sb_info *sbi)
+static void exofs_free_sbi(struct exofs_sb_info *sbi)
 {
-	while (sbi->comps.numdevs) {
-		int i = --sbi->comps.numdevs;
-		struct osd_dev *od = sbi->comps.ods[i];
+	unsigned numdevs = sbi->oc.numdevs;
+
+	while (numdevs) {
+		unsigned i = --numdevs;
+		struct osd_dev *od = ore_comp_dev(&sbi->oc, i);
 
 		if (od) {
-			sbi->comps.ods[i] = NULL;
+			ore_comp_set_dev(&sbi->oc, i, NULL);
 			osduld_put_device(od);
 		}
 	}
-	if (sbi->comps.ods != sbi->_min_one_dev)
-		kfree(sbi->comps.ods);
+	kfree(sbi->oc.ods);
 	kfree(sbi);
 }
 
@@ -468,7 +469,7 @@ static void exofs_put_super(struct super_block *sb)
 				  msecs_to_jiffies(100));
 	}
 
-	_exofs_print_device("Unmounting", NULL, sbi->comps.ods[0],
+	_exofs_print_device("Unmounting", NULL, ore_comp_dev(&sbi->oc, 0),
 			    sbi->one_comp.obj.partition);
 
 	bdi_destroy(&sbi->bdi);
@@ -479,76 +480,20 @@ static void exofs_put_super(struct super_block *sb)
 static int _read_and_match_data_map(struct exofs_sb_info *sbi, unsigned numdevs,
 				    struct exofs_device_table *dt)
 {
-	u64 stripe_length;
+	int ret;
 
-	sbi->data_map.odm_num_comps   =
-				le32_to_cpu(dt->dt_data_map.cb_num_comps);
-	sbi->data_map.odm_stripe_unit =
+	sbi->layout.stripe_unit =
 				le64_to_cpu(dt->dt_data_map.cb_stripe_unit);
-	sbi->data_map.odm_group_width =
+	sbi->layout.group_width =
 				le32_to_cpu(dt->dt_data_map.cb_group_width);
-	sbi->data_map.odm_group_depth =
+	sbi->layout.group_depth =
 				le32_to_cpu(dt->dt_data_map.cb_group_depth);
-	sbi->data_map.odm_mirror_cnt  =
-				le32_to_cpu(dt->dt_data_map.cb_mirror_cnt);
-	sbi->data_map.odm_raid_algorithm  =
+	sbi->layout.mirrors_p1  =
+				le32_to_cpu(dt->dt_data_map.cb_mirror_cnt) + 1;
+	sbi->layout.raid_algorithm  =
 				le32_to_cpu(dt->dt_data_map.cb_raid_algorithm);
 
-/* FIXME: Only raid0 for now. if not so, do not mount */
-	if (sbi->data_map.odm_num_comps != numdevs) {
-		EXOFS_ERR("odm_num_comps(%u) != numdevs(%u)\n",
-			  sbi->data_map.odm_num_comps, numdevs);
-		return -EINVAL;
-	}
-	if (sbi->data_map.odm_raid_algorithm != PNFS_OSD_RAID_0) {
-		EXOFS_ERR("Only RAID_0 for now\n");
-		return -EINVAL;
-	}
-	if (0 != (numdevs % (sbi->data_map.odm_mirror_cnt + 1))) {
-		EXOFS_ERR("Data Map wrong, numdevs=%d mirrors=%d\n",
-			  numdevs, sbi->data_map.odm_mirror_cnt);
-		return -EINVAL;
-	}
-
-	if (0 != (sbi->data_map.odm_stripe_unit & ~PAGE_MASK)) {
-		EXOFS_ERR("Stripe Unit(0x%llx)"
-			  " must be Multples of PAGE_SIZE(0x%lx)\n",
-			  _LLU(sbi->data_map.odm_stripe_unit), PAGE_SIZE);
-		return -EINVAL;
-	}
-
-	sbi->layout.stripe_unit = sbi->data_map.odm_stripe_unit;
-	sbi->layout.mirrors_p1 = sbi->data_map.odm_mirror_cnt + 1;
-
-	if (sbi->data_map.odm_group_width) {
-		sbi->layout.group_width = sbi->data_map.odm_group_width;
-		sbi->layout.group_depth = sbi->data_map.odm_group_depth;
-		if (!sbi->layout.group_depth) {
-			EXOFS_ERR("group_depth == 0 && group_width != 0\n");
-			return -EINVAL;
-		}
-		sbi->layout.group_count = sbi->data_map.odm_num_comps /
-						sbi->layout.mirrors_p1 /
-						sbi->data_map.odm_group_width;
-	} else {
-		if (sbi->data_map.odm_group_depth) {
-			printk(KERN_NOTICE "Warning: group_depth ignored "
-				"group_width == 0 && group_depth == %d\n",
-				sbi->data_map.odm_group_depth);
-			sbi->data_map.odm_group_depth = 0;
-		}
-		sbi->layout.group_width = sbi->data_map.odm_num_comps /
-							sbi->layout.mirrors_p1;
-		sbi->layout.group_depth = -1;
-		sbi->layout.group_count = 1;
-	}
-
-	stripe_length = (u64)sbi->layout.group_width * sbi->layout.stripe_unit;
-	if (stripe_length >= (1ULL << 32)) {
-		EXOFS_ERR("Total Stripe length(0x%llx)"
-			  " >= 32bit is not supported\n", _LLU(stripe_length));
-		return -EINVAL;
-	}
+	ret = ore_verify_layout(numdevs, &sbi->layout);
 
 	EXOFS_DBGMSG("exofs: layout: "
 		"num_comps=%u stripe_unit=0x%x group_width=%u "
@@ -558,8 +503,8 @@ static int _read_and_match_data_map(struct exofs_sb_info *sbi, unsigned numdevs,
 		sbi->layout.group_width,
 		_LLU(sbi->layout.group_depth),
 		sbi->layout.mirrors_p1,
-		sbi->data_map.odm_raid_algorithm);
-	return 0;
+		sbi->layout.raid_algorithm);
+	return ret;
 }
 
 static unsigned __ra_pages(struct ore_layout *layout)
@@ -605,12 +550,40 @@ static int exofs_devs_2_odi(struct exofs_dt_device_info *dt_dev,
 	return !(odi->systemid_len || odi->osdname_len);
 }
 
+int __alloc_dev_table(struct exofs_sb_info *sbi, unsigned numdevs,
+		      struct exofs_dev **peds)
+{
+	struct __alloc_ore_devs_and_exofs_devs {
+		/* Twice bigger table: See exofs_init_comps() and comment at
+		 * exofs_read_lookup_dev_table()
+		 */
+		struct ore_dev *oreds[numdevs * 2 - 1];
+		struct exofs_dev eds[numdevs];
+	} *aoded;
+	struct exofs_dev *eds;
+	unsigned i;
+
+	aoded = kzalloc(sizeof(*aoded), GFP_KERNEL);
+	if (unlikely(!aoded)) {
+		EXOFS_ERR("ERROR: faild allocating Device array[%d]\n",
+			  numdevs);
+		return -ENOMEM;
+	}
+
+	sbi->oc.ods = aoded->oreds;
+	*peds = eds = aoded->eds;
+	for (i = 0; i < numdevs; ++i)
+		aoded->oreds[i] = &eds[i].ored;
+	return 0;
+}
+
 static int exofs_read_lookup_dev_table(struct exofs_sb_info *sbi,
 				       struct osd_dev *fscb_od,
 				       unsigned table_count)
 {
 	struct ore_comp comp;
 	struct exofs_device_table *dt;
+	struct exofs_dev *eds;
 	unsigned table_bytes = table_count * sizeof(dt->dt_dev_table[0]) +
 					     sizeof(*dt);
 	unsigned numdevs, i;
@@ -623,7 +596,7 @@ static int exofs_read_lookup_dev_table(struct exofs_sb_info *sbi,
 		return -ENOMEM;
 	}
 
-	sbi->comps.numdevs = 0;
+	sbi->oc.numdevs = 0;
 
 	comp.obj.partition = sbi->one_comp.obj.partition;
 	comp.obj.id = EXOFS_DEVTABLE_ID;
@@ -647,20 +620,16 @@ static int exofs_read_lookup_dev_table(struct exofs_sb_info *sbi,
 	if (unlikely(ret))
 		goto out;
 
-	if (likely(numdevs > 1)) {
-		unsigned size = numdevs * sizeof(sbi->comps.ods[0]);
-
-		/* Twice bigger table: See exofs_init_comps() and below
-		 * comment
-		 */
-		sbi->comps.ods = kzalloc(size + size - 1, GFP_KERNEL);
-		if (unlikely(!sbi->comps.ods)) {
-			EXOFS_ERR("ERROR: faild allocating Device array[%d]\n",
-				  numdevs);
-			ret = -ENOMEM;
-			goto out;
-		}
-	}
+	ret = __alloc_dev_table(sbi, numdevs, &eds);
+	if (unlikely(ret))
+		goto out;
+	/* exofs round-robins the device table view according to inode
+	 * number. We hold a: twice bigger table hence inodes can point
+	 * to any device and have a sequential view of the table
+	 * starting at this device. See exofs_init_comps()
+	 */
+	memcpy(&sbi->oc.ods[numdevs], &sbi->oc.ods[0],
+		(numdevs - 1) * sizeof(sbi->oc.ods[0]));
 
 	for (i = 0; i < numdevs; i++) {
 		struct exofs_fscb fscb;
@@ -676,13 +645,16 @@ static int exofs_read_lookup_dev_table(struct exofs_sb_info *sbi,
 		printk(KERN_NOTICE "Add device[%d]: osd_name-%s\n",
 		       i, odi.osdname);
 
+		/* the exofs id is currently the table index */
+		eds[i].did = i;
+
 		/* On all devices the device table is identical. The user can
 		 * specify any one of the participating devices on the command
 		 * line. We always keep them in device-table order.
 		 */
 		if (fscb_od && osduld_device_same(fscb_od, &odi)) {
-			sbi->comps.ods[i] = fscb_od;
-			++sbi->comps.numdevs;
+			eds[i].ored.od = fscb_od;
+			++sbi->oc.numdevs;
 			fscb_od = NULL;
 			continue;
 		}
@@ -695,8 +667,8 @@ static int exofs_read_lookup_dev_table(struct exofs_sb_info *sbi,
 			goto out;
 		}
 
-		sbi->comps.ods[i] = od;
-		++sbi->comps.numdevs;
+		eds[i].ored.od = od;
+		++sbi->oc.numdevs;
 
 		/* Read the fscb of the other devices to make sure the FS
 		 * partition is there.
@@ -718,21 +690,10 @@ static int exofs_read_lookup_dev_table(struct exofs_sb_info *sbi,
 
 out:
 	kfree(dt);
-	if (likely(!ret)) {
-		unsigned numdevs = sbi->comps.numdevs;
-
-		if (unlikely(fscb_od)) {
+	if (unlikely(fscb_od && !ret)) {
 			EXOFS_ERR("ERROR: Bad device-table container device not present\n");
 			osduld_put_device(fscb_od);
 			return -EINVAL;
-		}
-		/* exofs round-robins the device table view according to inode
-		 * number. We hold a: twice bigger table hence inodes can point
-		 * to any device and have a sequential view of the table
-		 * starting at this device. See exofs_init_comps()
-		 */
-		for (i = 0; i < numdevs - 1; ++i)
-			sbi->comps.ods[i + numdevs] = sbi->comps.ods[i];
 	}
 	return ret;
 }
@@ -783,10 +744,9 @@ static int exofs_fill_super(struct super_block *sb, void *data, int silent)
 	sbi->one_comp.obj.partition = opts->pid;
 	sbi->one_comp.obj.id = 0;
 	exofs_make_credential(sbi->one_comp.cred, &sbi->one_comp.obj);
-	sbi->comps.numdevs = 1;
-	sbi->comps.single_comp = EC_SINGLE_COMP;
-	sbi->comps.comps = &sbi->one_comp;
-	sbi->comps.ods = sbi->_min_one_dev;
+	sbi->oc.numdevs = 1;
+	sbi->oc.single_comp = EC_SINGLE_COMP;
+	sbi->oc.comps = &sbi->one_comp;
 
 	/* fill in some other data by hand */
 	memset(sb->s_id, 0, sizeof(sb->s_id));
@@ -835,7 +795,13 @@ static int exofs_fill_super(struct super_block *sb, void *data, int silent)
 		if (unlikely(ret))
 			goto free_sbi;
 	} else {
-		sbi->comps.ods[0] = od;
+		struct exofs_dev *eds;
+
+		ret = __alloc_dev_table(sbi, 1, &eds);
+		if (unlikely(ret))
+			goto free_sbi;
+
+		ore_comp_set_dev(&sbi->oc, 0, od);
 	}
 
 	__sbi_read_stats(sbi);
@@ -875,7 +841,8 @@ static int exofs_fill_super(struct super_block *sb, void *data, int silent)
 		goto free_sbi;
 	}
 
-	_exofs_print_device("Mounting", opts->dev_name, sbi->comps.ods[0],
+	_exofs_print_device("Mounting", opts->dev_name,
+			    ore_comp_dev(&sbi->oc, 0),
 			    sbi->one_comp.obj.partition);
 	return 0;
 
@@ -924,7 +891,7 @@ static int exofs_statfs(struct dentry *dentry, struct kstatfs *buf)
 	uint64_t used = ULLONG_MAX;
 	int ret;
 
-	ret = ore_get_io_state(&sbi->layout, &sbi->comps, &ios);
+	ret = ore_get_io_state(&sbi->layout, &sbi->oc, &ios);
 	if (ret) {
 		EXOFS_DBGMSG("ore_get_io_state failed.\n");
 		return ret;
@@ -981,7 +948,7 @@ static const struct super_operations exofs_sops = {
  * EXPORT OPERATIONS
  *****************************************************************************/
 
-struct dentry *exofs_get_parent(struct dentry *child)
+static struct dentry *exofs_get_parent(struct dentry *child)
 {
 	unsigned long ino = exofs_parent_ino(child);
 

include/scsi/osd_ore.h

@@ -34,15 +34,30 @@ struct ore_comp {
 
 struct ore_layout {
 	/* Our way of looking at the data_map */
+	enum pnfs_osd_raid_algorithm4
+		 raid_algorithm;
 	unsigned stripe_unit;
 	unsigned mirrors_p1;
 
 	unsigned group_width;
+	unsigned parity;
 	u64	 group_depth;
 	unsigned group_count;
+
+	/* Cached often needed calculations filled in by
+	 * ore_verify_layout
+	 */
+	unsigned long max_io_length;	/* Max length that should be passed to
+					 * ore_get_rw_state
+					 */
+};
+
+struct ore_dev {
+	struct osd_dev *od;
 };
 
 struct ore_components {
+	unsigned	first_dev;		/* First logical device no    */
 	unsigned	numdevs;		/* Num of devices in array    */
 	/* If @single_comp == EC_SINGLE_COMP, @comps points to a single
 	 * component. else there are @numdevs components
@@ -51,20 +66,60 @@ struct ore_components {
 		EC_SINGLE_COMP = 0, EC_MULTPLE_COMPS = 0xffffffff
 	}		single_comp;
 	struct ore_comp	*comps;
-	struct osd_dev	**ods;			/* osd_dev array              */
+
+	/* Array of pointers to ore_dev-* . User will usually have these pointed
+	 * too a bigger struct which contain an "ore_dev ored" member and use
+	 * container_of(oc->ods[i], struct foo_dev, ored) to access the bigger
+	 * structure.
+	 */
+	struct ore_dev	**ods;
+};
+
+/* ore_comp_dev Recievies a logical device index */
+static inline struct osd_dev *ore_comp_dev(
+	const struct ore_components *oc, unsigned i)
+{
+	BUG_ON((i < oc->first_dev) || (oc->first_dev + oc->numdevs <= i));
+	return oc->ods[i - oc->first_dev]->od;
+}
+
+static inline void ore_comp_set_dev(
+	struct ore_components *oc, unsigned i, struct osd_dev *od)
+{
+	oc->ods[i - oc->first_dev]->od = od;
+}
+
+struct ore_striping_info {
+	u64 offset;
+	u64 obj_offset;
+	u64 length;
+	u64 first_stripe_start; /* only used in raid writes */
+	u64 M; /* for truncate */
+	unsigned bytes_in_stripe;
+	unsigned dev;
+	unsigned par_dev;
+	unsigned unit_off;
+	unsigned cur_pg;
+	unsigned cur_comp;
 };
 
 struct ore_io_state;
 typedef void (*ore_io_done_fn)(struct ore_io_state *ios, void *private);
+struct _ore_r4w_op {
+	/* @Priv given here is passed ios->private */
+	struct page * (*get_page)(void *priv, u64 page_index, bool *uptodate);
+	void (*put_page)(void *priv, struct page *page);
+};
 
 struct ore_io_state {
 	struct kref		kref;
+	struct ore_striping_info si;
 
 	void			*private;
 	ore_io_done_fn	done;
 
 	struct ore_layout	*layout;
-	struct ore_components	*comps;
+	struct ore_components	*oc;
 
 	/* Global read/write IO*/
 	loff_t			offset;
@@ -84,6 +139,16 @@ struct ore_io_state {
 
 	bool			reading;
 
+	/* House keeping of Parity pages */
+	bool			extra_part_alloc;
+	struct page		**parity_pages;
+	unsigned		max_par_pages;
+	unsigned		cur_par_page;
+	unsigned		sgs_per_dev;
+	struct __stripe_pages_2d *sp2d;
+	struct ore_io_state	 *ios_read_4_write;
+	const struct _ore_r4w_op *r4w;
+
 	/* Variable array of size numdevs */
 	unsigned numdevs;
 	struct ore_per_dev_state {
@@ -91,7 +156,10 @@ struct ore_io_state {
 		struct bio *bio;
 		loff_t offset;
 		unsigned length;
+		unsigned last_sgs_total;
 		unsigned dev;
+		struct osd_sg_entry *sglist;
+		unsigned cur_sg;
 	} per_dev[];
 };
 
@@ -102,6 +170,9 @@ static inline unsigned ore_io_state_size(unsigned numdevs)
 }
 
 /* ore.c */
+int ore_verify_layout(unsigned total_comps, struct ore_layout *layout);
+void ore_calc_stripe_info(struct ore_layout *layout, u64 file_offset,
+			  u64 length, struct ore_striping_info *si);
 int ore_get_rw_state(struct ore_layout *layout, struct ore_components *comps,
 		     bool is_reading, u64 offset, u64 length,
 		     struct ore_io_state **ios);
@@ -109,7 +180,10 @@ int ore_get_io_state(struct ore_layout *layout, struct ore_components *comps,
 		     struct ore_io_state **ios);
 void ore_put_io_state(struct ore_io_state *ios);
 
-int ore_check_io(struct ore_io_state *ios, u64 *resid);
+typedef void (*ore_on_dev_error)(struct ore_io_state *ios, struct ore_dev *od,
+	unsigned dev_index, enum osd_err_priority oep,
+	u64 dev_offset, u64  dev_len);
+int ore_check_io(struct ore_io_state *ios, ore_on_dev_error rep);
 
 int ore_create(struct ore_io_state *ios);
 int ore_remove(struct ore_io_state *ios);