Merge branch 'sg' of git://git.kernel.dk/linux-2.6-block

* 'sg' of git://git.kernel.dk/linux-2.6-block:
  SG: work with the SCSI fixed maximum allocations.
  SG: Convert SCSI to use scatterlist helpers for sg chaining
  SG: Move functions to lib/scatterlist.c and add sg chaining allocator helpers
diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c
index 4cf902e..eb4911a 100644
--- a/drivers/scsi/scsi_lib.c
+++ b/drivers/scsi/scsi_lib.c
@@ -739,138 +739,43 @@
 	return index;
 }
 
-struct scatterlist *scsi_alloc_sgtable(struct scsi_cmnd *cmd, gfp_t gfp_mask)
+static void scsi_sg_free(struct scatterlist *sgl, unsigned int nents)
 {
 	struct scsi_host_sg_pool *sgp;
-	struct scatterlist *sgl, *prev, *ret;
-	unsigned int index;
-	int this, left;
+
+	sgp = scsi_sg_pools + scsi_sgtable_index(nents);
+	mempool_free(sgl, sgp->pool);
+}
+
+static struct scatterlist *scsi_sg_alloc(unsigned int nents, gfp_t gfp_mask)
+{
+	struct scsi_host_sg_pool *sgp;
+
+	sgp = scsi_sg_pools + scsi_sgtable_index(nents);
+	return mempool_alloc(sgp->pool, gfp_mask);
+}
+
+int scsi_alloc_sgtable(struct scsi_cmnd *cmd, gfp_t gfp_mask)
+{
+	int ret;
 
 	BUG_ON(!cmd->use_sg);
 
-	left = cmd->use_sg;
-	ret = prev = NULL;
-	do {
-		this = left;
-		if (this > SCSI_MAX_SG_SEGMENTS) {
-			this = SCSI_MAX_SG_SEGMENTS - 1;
-			index = SG_MEMPOOL_NR - 1;
-		} else
-			index = scsi_sgtable_index(this);
+	ret = __sg_alloc_table(&cmd->sg_table, cmd->use_sg,
+			       SCSI_MAX_SG_SEGMENTS, gfp_mask, scsi_sg_alloc);
+	if (unlikely(ret))
+		__sg_free_table(&cmd->sg_table, SCSI_MAX_SG_SEGMENTS,
+				scsi_sg_free);
 
-		left -= this;
-
-		sgp = scsi_sg_pools + index;
-
-		sgl = mempool_alloc(sgp->pool, gfp_mask);
-		if (unlikely(!sgl))
-			goto enomem;
-
-		sg_init_table(sgl, sgp->size);
-
-		/*
-		 * first loop through, set initial index and return value
-		 */
-		if (!ret)
-			ret = sgl;
-
-		/*
-		 * chain previous sglist, if any. we know the previous
-		 * sglist must be the biggest one, or we would not have
-		 * ended up doing another loop.
-		 */
-		if (prev)
-			sg_chain(prev, SCSI_MAX_SG_SEGMENTS, sgl);
-
-		/*
-		 * if we have nothing left, mark the last segment as
-		 * end-of-list
-		 */
-		if (!left)
-			sg_mark_end(&sgl[this - 1]);
-
-		/*
-		 * don't allow subsequent mempool allocs to sleep, it would
-		 * violate the mempool principle.
-		 */
-		gfp_mask &= ~__GFP_WAIT;
-		gfp_mask |= __GFP_HIGH;
-		prev = sgl;
-	} while (left);
-
-	/*
-	 * ->use_sg may get modified after dma mapping has potentially
-	 * shrunk the number of segments, so keep a copy of it for free.
-	 */
-	cmd->__use_sg = cmd->use_sg;
+	cmd->request_buffer = cmd->sg_table.sgl;
 	return ret;
-enomem:
-	if (ret) {
-		/*
-		 * Free entries chained off ret. Since we were trying to
-		 * allocate another sglist, we know that all entries are of
-		 * the max size.
-		 */
-		sgp = scsi_sg_pools + SG_MEMPOOL_NR - 1;
-		prev = ret;
-		ret = &ret[SCSI_MAX_SG_SEGMENTS - 1];
-
-		while ((sgl = sg_chain_ptr(ret)) != NULL) {
-			ret = &sgl[SCSI_MAX_SG_SEGMENTS - 1];
-			mempool_free(sgl, sgp->pool);
-		}
-
-		mempool_free(prev, sgp->pool);
-	}
-	return NULL;
 }
 
 EXPORT_SYMBOL(scsi_alloc_sgtable);
 
 void scsi_free_sgtable(struct scsi_cmnd *cmd)
 {
-	struct scatterlist *sgl = cmd->request_buffer;
-	struct scsi_host_sg_pool *sgp;
-
-	/*
-	 * if this is the biggest size sglist, check if we have
-	 * chained parts we need to free
-	 */
-	if (cmd->__use_sg > SCSI_MAX_SG_SEGMENTS) {
-		unsigned short this, left;
-		struct scatterlist *next;
-		unsigned int index;
-
-		left = cmd->__use_sg - (SCSI_MAX_SG_SEGMENTS - 1);
-		next = sg_chain_ptr(&sgl[SCSI_MAX_SG_SEGMENTS - 1]);
-		while (left && next) {
-			sgl = next;
-			this = left;
-			if (this > SCSI_MAX_SG_SEGMENTS) {
-				this = SCSI_MAX_SG_SEGMENTS - 1;
-				index = SG_MEMPOOL_NR - 1;
-			} else
-				index = scsi_sgtable_index(this);
-
-			left -= this;
-
-			sgp = scsi_sg_pools + index;
-
-			if (left)
-				next = sg_chain_ptr(&sgl[sgp->size - 1]);
-
-			mempool_free(sgl, sgp->pool);
-		}
-
-		/*
-		 * Restore original, will be freed below
-		 */
-		sgl = cmd->request_buffer;
-		sgp = scsi_sg_pools + SG_MEMPOOL_NR - 1;
-	} else
-		sgp = scsi_sg_pools + scsi_sgtable_index(cmd->__use_sg);
-
-	mempool_free(sgl, sgp->pool);
+	__sg_free_table(&cmd->sg_table, SCSI_MAX_SG_SEGMENTS, scsi_sg_free);
 }
 
 EXPORT_SYMBOL(scsi_free_sgtable);
@@ -1120,8 +1025,7 @@
 	/*
 	 * If sg table allocation fails, requeue request later.
 	 */
-	cmd->request_buffer = scsi_alloc_sgtable(cmd, GFP_ATOMIC);
-	if (unlikely(!cmd->request_buffer)) {
+	if (unlikely(scsi_alloc_sgtable(cmd, GFP_ATOMIC))) {
 		scsi_unprep_request(req);
 		return BLKPREP_DEFER;
 	}
diff --git a/drivers/scsi/scsi_tgt_lib.c b/drivers/scsi/scsi_tgt_lib.c
index 93ece8f..01e03f3 100644
--- a/drivers/scsi/scsi_tgt_lib.c
+++ b/drivers/scsi/scsi_tgt_lib.c
@@ -359,8 +359,7 @@
 	int count;
 
 	cmd->use_sg = rq->nr_phys_segments;
-	cmd->request_buffer = scsi_alloc_sgtable(cmd, gfp_mask);
-	if (!cmd->request_buffer)
+	if (scsi_alloc_sgtable(cmd, gfp_mask))
 		return -ENOMEM;
 
 	cmd->request_bufflen = rq->data_len;
diff --git a/include/linux/scatterlist.h b/include/linux/scatterlist.h
index e3ff21d..a3d567a 100644
--- a/include/linux/scatterlist.h
+++ b/include/linux/scatterlist.h
@@ -7,6 +7,12 @@
 #include <linux/string.h>
 #include <asm/io.h>
 
+struct sg_table {
+	struct scatterlist *sgl;	/* the list */
+	unsigned int nents;		/* number of mapped entries */
+	unsigned int orig_nents;	/* original size of list */
+};
+
 /*
  * Notes on SG table design.
  *
@@ -106,31 +112,6 @@
 	sg_set_page(sg, virt_to_page(buf), buflen, offset_in_page(buf));
 }
 
-/**
- * sg_next - return the next scatterlist entry in a list
- * @sg:		The current sg entry
- *
- * Description:
- *   Usually the next entry will be @sg@ + 1, but if this sg element is part
- *   of a chained scatterlist, it could jump to the start of a new
- *   scatterlist array.
- *
- **/
-static inline struct scatterlist *sg_next(struct scatterlist *sg)
-{
-#ifdef CONFIG_DEBUG_SG
-	BUG_ON(sg->sg_magic != SG_MAGIC);
-#endif
-	if (sg_is_last(sg))
-		return NULL;
-
-	sg++;
-	if (unlikely(sg_is_chain(sg)))
-		sg = sg_chain_ptr(sg);
-
-	return sg;
-}
-
 /*
  * Loop over each sg element, following the pointer to a new list if necessary
  */
@@ -138,40 +119,6 @@
 	for (__i = 0, sg = (sglist); __i < (nr); __i++, sg = sg_next(sg))
 
 /**
- * sg_last - return the last scatterlist entry in a list
- * @sgl:	First entry in the scatterlist
- * @nents:	Number of entries in the scatterlist
- *
- * Description:
- *   Should only be used casually, it (currently) scan the entire list
- *   to get the last entry.
- *
- *   Note that the @sgl@ pointer passed in need not be the first one,
- *   the important bit is that @nents@ denotes the number of entries that
- *   exist from @sgl@.
- *
- **/
-static inline struct scatterlist *sg_last(struct scatterlist *sgl,
-					  unsigned int nents)
-{
-#ifndef ARCH_HAS_SG_CHAIN
-	struct scatterlist *ret = &sgl[nents - 1];
-#else
-	struct scatterlist *sg, *ret = NULL;
-	unsigned int i;
-
-	for_each_sg(sgl, sg, nents, i)
-		ret = sg;
-
-#endif
-#ifdef CONFIG_DEBUG_SG
-	BUG_ON(sgl[0].sg_magic != SG_MAGIC);
-	BUG_ON(!sg_is_last(ret));
-#endif
-	return ret;
-}
-
-/**
  * sg_chain - Chain two sglists together
  * @prv:	First scatterlist
  * @prv_nents:	Number of entries in prv
@@ -223,47 +170,6 @@
 }
 
 /**
- * sg_init_table - Initialize SG table
- * @sgl:	   The SG table
- * @nents:	   Number of entries in table
- *
- * Notes:
- *   If this is part of a chained sg table, sg_mark_end() should be
- *   used only on the last table part.
- *
- **/
-static inline void sg_init_table(struct scatterlist *sgl, unsigned int nents)
-{
-	memset(sgl, 0, sizeof(*sgl) * nents);
-#ifdef CONFIG_DEBUG_SG
-	{
-		unsigned int i;
-		for (i = 0; i < nents; i++)
-			sgl[i].sg_magic = SG_MAGIC;
-	}
-#endif
-	sg_mark_end(&sgl[nents - 1]);
-}
-
-/**
- * sg_init_one - Initialize a single entry sg list
- * @sg:		 SG entry
- * @buf:	 Virtual address for IO
- * @buflen:	 IO length
- *
- * Notes:
- *   This should not be used on a single entry that is part of a larger
- *   table. Use sg_init_table() for that.
- *
- **/
-static inline void sg_init_one(struct scatterlist *sg, const void *buf,
-			       unsigned int buflen)
-{
-	sg_init_table(sg, 1);
-	sg_set_buf(sg, buf, buflen);
-}
-
-/**
  * sg_phys - Return physical address of an sg entry
  * @sg:	     SG entry
  *
@@ -293,4 +199,24 @@
 	return page_address(sg_page(sg)) + sg->offset;
 }
 
+struct scatterlist *sg_next(struct scatterlist *);
+struct scatterlist *sg_last(struct scatterlist *s, unsigned int);
+void sg_init_table(struct scatterlist *, unsigned int);
+void sg_init_one(struct scatterlist *, const void *, unsigned int);
+
+typedef struct scatterlist *(sg_alloc_fn)(unsigned int, gfp_t);
+typedef void (sg_free_fn)(struct scatterlist *, unsigned int);
+
+void __sg_free_table(struct sg_table *, unsigned int, sg_free_fn *);
+void sg_free_table(struct sg_table *);
+int __sg_alloc_table(struct sg_table *, unsigned int, unsigned int, gfp_t,
+		     sg_alloc_fn *);
+int sg_alloc_table(struct sg_table *, unsigned int, gfp_t);
+
+/*
+ * Maximum number of entries that will be allocated in one piece, if
+ * a list larger than this is required then chaining will be utilized.
+ */
+#define SG_MAX_SINGLE_ALLOC		(PAGE_SIZE / sizeof(struct scatterlist))
+
 #endif /* _LINUX_SCATTERLIST_H */
diff --git a/include/scsi/scsi_cmnd.h b/include/scsi/scsi_cmnd.h
index abd7479..a457fca 100644
--- a/include/scsi/scsi_cmnd.h
+++ b/include/scsi/scsi_cmnd.h
@@ -8,7 +8,6 @@
 #include <linux/scatterlist.h>
 
 struct request;
-struct scatterlist;
 struct Scsi_Host;
 struct scsi_device;
 
@@ -68,8 +67,8 @@
 	void *request_buffer;		/* Actual requested buffer */
 
 	/* These elements define the operation we ultimately want to perform */
+	struct sg_table sg_table;
 	unsigned short use_sg;	/* Number of pieces of scatter-gather */
-	unsigned short __use_sg;
 
 	unsigned underflow;	/* Return error if less than
 				   this amount is transferred */
@@ -128,14 +127,14 @@
 				 size_t *offset, size_t *len);
 extern void scsi_kunmap_atomic_sg(void *virt);
 
-extern struct scatterlist *scsi_alloc_sgtable(struct scsi_cmnd *, gfp_t);
+extern int scsi_alloc_sgtable(struct scsi_cmnd *, gfp_t);
 extern void scsi_free_sgtable(struct scsi_cmnd *);
 
 extern int scsi_dma_map(struct scsi_cmnd *cmd);
 extern void scsi_dma_unmap(struct scsi_cmnd *cmd);
 
 #define scsi_sg_count(cmd) ((cmd)->use_sg)
-#define scsi_sglist(cmd) ((struct scatterlist *)(cmd)->request_buffer)
+#define scsi_sglist(cmd) ((cmd)->sg_table.sgl)
 #define scsi_bufflen(cmd) ((cmd)->request_bufflen)
 
 static inline void scsi_set_resid(struct scsi_cmnd *cmd, int resid)
diff --git a/lib/Makefile b/lib/Makefile
index b6793ed..89841dc 100644
--- a/lib/Makefile
+++ b/lib/Makefile
@@ -6,7 +6,7 @@
 	 rbtree.o radix-tree.o dump_stack.o \
 	 idr.o int_sqrt.o extable.o prio_tree.o \
 	 sha1.o irq_regs.o reciprocal_div.o argv_split.o \
-	 proportions.o prio_heap.o
+	 proportions.o prio_heap.o scatterlist.o
 
 lib-$(CONFIG_MMU) += ioremap.o
 lib-$(CONFIG_SMP) += cpumask.o
diff --git a/lib/scatterlist.c b/lib/scatterlist.c
new file mode 100644
index 0000000..acca490
--- /dev/null
+++ b/lib/scatterlist.c
@@ -0,0 +1,294 @@
+/*
+ * Copyright (C) 2007 Jens Axboe <jens.axboe@oracle.com>
+ *
+ * Scatterlist handling helpers.
+ *
+ * This source code is licensed under the GNU General Public License,
+ * Version 2. See the file COPYING for more details.
+ */
+#include <linux/module.h>
+#include <linux/scatterlist.h>
+
+/**
+ * sg_next - return the next scatterlist entry in a list
+ * @sg:		The current sg entry
+ *
+ * Description:
+ *   Usually the next entry will be @sg@ + 1, but if this sg element is part
+ *   of a chained scatterlist, it could jump to the start of a new
+ *   scatterlist array.
+ *
+ **/
+struct scatterlist *sg_next(struct scatterlist *sg)
+{
+#ifdef CONFIG_DEBUG_SG
+	BUG_ON(sg->sg_magic != SG_MAGIC);
+#endif
+	if (sg_is_last(sg))
+		return NULL;
+
+	sg++;
+	if (unlikely(sg_is_chain(sg)))
+		sg = sg_chain_ptr(sg);
+
+	return sg;
+}
+EXPORT_SYMBOL(sg_next);
+
+/**
+ * sg_last - return the last scatterlist entry in a list
+ * @sgl:	First entry in the scatterlist
+ * @nents:	Number of entries in the scatterlist
+ *
+ * Description:
+ *   Should only be used casually, it (currently) scans the entire list
+ *   to get the last entry.
+ *
+ *   Note that the @sgl@ pointer passed in need not be the first one,
+ *   the important bit is that @nents@ denotes the number of entries that
+ *   exist from @sgl@.
+ *
+ **/
+struct scatterlist *sg_last(struct scatterlist *sgl, unsigned int nents)
+{
+#ifndef ARCH_HAS_SG_CHAIN
+	struct scatterlist *ret = &sgl[nents - 1];
+#else
+	struct scatterlist *sg, *ret = NULL;
+	unsigned int i;
+
+	for_each_sg(sgl, sg, nents, i)
+		ret = sg;
+
+#endif
+#ifdef CONFIG_DEBUG_SG
+	BUG_ON(sgl[0].sg_magic != SG_MAGIC);
+	BUG_ON(!sg_is_last(ret));
+#endif
+	return ret;
+}
+EXPORT_SYMBOL(sg_last);
+
+/**
+ * sg_init_table - Initialize SG table
+ * @sgl:	   The SG table
+ * @nents:	   Number of entries in table
+ *
+ * Notes:
+ *   If this is part of a chained sg table, sg_mark_end() should be
+ *   used only on the last table part.
+ *
+ **/
+void sg_init_table(struct scatterlist *sgl, unsigned int nents)
+{
+	memset(sgl, 0, sizeof(*sgl) * nents);
+#ifdef CONFIG_DEBUG_SG
+	{
+		unsigned int i;
+		for (i = 0; i < nents; i++)
+			sgl[i].sg_magic = SG_MAGIC;
+	}
+#endif
+	sg_mark_end(&sgl[nents - 1]);
+}
+EXPORT_SYMBOL(sg_init_table);
+
+/**
+ * sg_init_one - Initialize a single entry sg list
+ * @sg:		 SG entry
+ * @buf:	 Virtual address for IO
+ * @buflen:	 IO length
+ *
+ **/
+void sg_init_one(struct scatterlist *sg, const void *buf, unsigned int buflen)
+{
+	sg_init_table(sg, 1);
+	sg_set_buf(sg, buf, buflen);
+}
+EXPORT_SYMBOL(sg_init_one);
+
+/*
+ * The default behaviour of sg_alloc_table() is to use these kmalloc/kfree
+ * helpers.
+ */
+static struct scatterlist *sg_kmalloc(unsigned int nents, gfp_t gfp_mask)
+{
+	if (nents == SG_MAX_SINGLE_ALLOC)
+		return (struct scatterlist *) __get_free_page(gfp_mask);
+	else
+		return kmalloc(nents * sizeof(struct scatterlist), gfp_mask);
+}
+
+static void sg_kfree(struct scatterlist *sg, unsigned int nents)
+{
+	if (nents == SG_MAX_SINGLE_ALLOC)
+		free_page((unsigned long) sg);
+	else
+		kfree(sg);
+}
+
+/**
+ * __sg_free_table - Free a previously mapped sg table
+ * @table:	The sg table header to use
+ * @max_ents:	The maximum number of entries per single scatterlist
+ * @free_fn:	Free function
+ *
+ *  Description:
+ *    Free an sg table previously allocated and setup with
+ *    __sg_alloc_table().  The @max_ents value must be identical to
+ *    that previously used with __sg_alloc_table().
+ *
+ **/
+void __sg_free_table(struct sg_table *table, unsigned int max_ents,
+		     sg_free_fn *free_fn)
+{
+	struct scatterlist *sgl, *next;
+
+	if (unlikely(!table->sgl))
+		return;
+
+	sgl = table->sgl;
+	while (table->orig_nents) {
+		unsigned int alloc_size = table->orig_nents;
+		unsigned int sg_size;
+
+		/*
+		 * If we have more than max_ents segments left,
+		 * then assign 'next' to the sg table after the current one.
+		 * sg_size is then one less than alloc size, since the last
+		 * element is the chain pointer.
+		 */
+		if (alloc_size > max_ents) {
+			next = sg_chain_ptr(&sgl[max_ents - 1]);
+			alloc_size = max_ents;
+			sg_size = alloc_size - 1;
+		} else {
+			sg_size = alloc_size;
+			next = NULL;
+		}
+
+		table->orig_nents -= sg_size;
+		free_fn(sgl, alloc_size);
+		sgl = next;
+	}
+
+	table->sgl = NULL;
+}
+EXPORT_SYMBOL(__sg_free_table);
+
+/**
+ * sg_free_table - Free a previously allocated sg table
+ * @table:	The mapped sg table header
+ *
+ **/
+void sg_free_table(struct sg_table *table)
+{
+	__sg_free_table(table, SG_MAX_SINGLE_ALLOC, sg_kfree);
+}
+EXPORT_SYMBOL(sg_free_table);
+
+/**
+ * __sg_alloc_table - Allocate and initialize an sg table with given allocator
+ * @table:	The sg table header to use
+ * @nents:	Number of entries in sg list
+ * @max_ents:	The maximum number of entries the allocator returns per call
+ * @gfp_mask:	GFP allocation mask
+ * @alloc_fn:	Allocator to use
+ *
+ * Description:
+ *   This function returns a @table @nents long. The allocator is
+ *   defined to return scatterlist chunks of maximum size @max_ents.
+ *   Thus if @nents is bigger than @max_ents, the scatterlists will be
+ *   chained in units of @max_ents.
+ *
+ * Notes:
+ *   If this function returns non-0 (eg failure), the caller must call
+ *   __sg_free_table() to cleanup any leftover allocations.
+ *
+ **/
+int __sg_alloc_table(struct sg_table *table, unsigned int nents,
+		     unsigned int max_ents, gfp_t gfp_mask,
+		     sg_alloc_fn *alloc_fn)
+{
+	struct scatterlist *sg, *prv;
+	unsigned int left;
+
+#ifndef ARCH_HAS_SG_CHAIN
+	BUG_ON(nents > max_ents);
+#endif
+
+	memset(table, 0, sizeof(*table));
+
+	left = nents;
+	prv = NULL;
+	do {
+		unsigned int sg_size, alloc_size = left;
+
+		if (alloc_size > max_ents) {
+			alloc_size = max_ents;
+			sg_size = alloc_size - 1;
+		} else
+			sg_size = alloc_size;
+
+		left -= sg_size;
+
+		sg = alloc_fn(alloc_size, gfp_mask);
+		if (unlikely(!sg))
+			return -ENOMEM;
+
+		sg_init_table(sg, alloc_size);
+		table->nents = table->orig_nents += sg_size;
+
+		/*
+		 * If this is the first mapping, assign the sg table header.
+		 * If this is not the first mapping, chain previous part.
+		 */
+		if (prv)
+			sg_chain(prv, max_ents, sg);
+		else
+			table->sgl = sg;
+
+		/*
+		 * If no more entries after this one, mark the end
+		 */
+		if (!left)
+			sg_mark_end(&sg[sg_size - 1]);
+
+		/*
+		 * only really needed for mempool backed sg allocations (like
+		 * SCSI), a possible improvement here would be to pass the
+		 * table pointer into the allocator and let that clear these
+		 * flags
+		 */
+		gfp_mask &= ~__GFP_WAIT;
+		gfp_mask |= __GFP_HIGH;
+		prv = sg;
+	} while (left);
+
+	return 0;
+}
+EXPORT_SYMBOL(__sg_alloc_table);
+
+/**
+ * sg_alloc_table - Allocate and initialize an sg table
+ * @table:	The sg table header to use
+ * @nents:	Number of entries in sg list
+ * @gfp_mask:	GFP allocation mask
+ *
+ *  Description:
+ *    Allocate and initialize an sg table. If @nents@ is larger than
+ *    SG_MAX_SINGLE_ALLOC a chained sg table will be setup.
+ *
+ **/
+int sg_alloc_table(struct sg_table *table, unsigned int nents, gfp_t gfp_mask)
+{
+	int ret;
+
+	ret = __sg_alloc_table(table, nents, SG_MAX_SINGLE_ALLOC,
+			       gfp_mask, sg_kmalloc);
+	if (unlikely(ret))
+		__sg_free_table(table, SG_MAX_SINGLE_ALLOC, sg_kfree);
+
+	return ret;
+}
+EXPORT_SYMBOL(sg_alloc_table);