Linux-libre 4.2-gnu

13 files changed, 2907 insertions, 1699 deletions

diff --git a/drivers/crypto/nx/Kconfig b/drivers/crypto/nx/Kconfig
index f82616621..e421c96c7 100644
--- a/drivers/crypto/nx/Kconfig
+++ b/drivers/crypto/nx/Kconfig
@@ -1,26 +1,55 @@
+
 config CRYPTO_DEV_NX_ENCRYPT
-	tristate "Encryption acceleration support"
-	depends on PPC64 && IBMVIO
+	tristate "Encryption acceleration support on pSeries platform"
+	depends on PPC_PSERIES && IBMVIO && !CPU_LITTLE_ENDIAN
 	default y
 	select CRYPTO_AES
-	select CRYPTO_CBC
-	select CRYPTO_ECB
 	select CRYPTO_CCM
-	select CRYPTO_GCM
-	select CRYPTO_AUTHENC
-	select CRYPTO_XCBC
-	select CRYPTO_SHA256
-	select CRYPTO_SHA512
 	help
-	  Support for Power7+ in-Nest encryption acceleration. This
-	  module supports acceleration for AES and SHA2 algorithms. If you
-	  choose 'M' here, this module will be called nx_crypto.
+	  Support for PowerPC Nest (NX) encryption acceleration. This
+	  module supports acceleration for AES and SHA2 algorithms on
+	  the pSeries platform.  If you choose 'M' here, this module
+	  will be called nx_crypto.
 
 config CRYPTO_DEV_NX_COMPRESS
 	tristate "Compression acceleration support"
-	depends on PPC64 && IBMVIO
 	default y
 	help
-	  Support for Power7+ in-Nest compression acceleration. This
-	  module supports acceleration for AES and SHA2 algorithms. If you
-	  choose 'M' here, this module will be called nx_compress.
+	  Support for PowerPC Nest (NX) compression acceleration. This
+	  module supports acceleration for compressing memory with the 842
+	  algorithm.  One of the platform drivers must be selected also.
+	  If you choose 'M' here, this module will be called nx_compress.
+
+if CRYPTO_DEV_NX_COMPRESS
+
+config CRYPTO_DEV_NX_COMPRESS_PSERIES
+	tristate "Compression acceleration support on pSeries platform"
+	depends on PPC_PSERIES && IBMVIO
+	default y
+	help
+	  Support for PowerPC Nest (NX) compression acceleration. This
+	  module supports acceleration for compressing memory with the 842
+	  algorithm.  This supports NX hardware on the pSeries platform.
+	  If you choose 'M' here, this module will be called nx_compress_pseries.
+
+config CRYPTO_DEV_NX_COMPRESS_POWERNV
+	tristate "Compression acceleration support on PowerNV platform"
+	depends on PPC_POWERNV
+	default y
+	help
+	  Support for PowerPC Nest (NX) compression acceleration. This
+	  module supports acceleration for compressing memory with the 842
+	  algorithm.  This supports NX hardware on the PowerNV platform.
+	  If you choose 'M' here, this module will be called nx_compress_powernv.
+
+config CRYPTO_DEV_NX_COMPRESS_CRYPTO
+	tristate "Compression acceleration cryptographic interface"
+	select CRYPTO_ALGAPI
+	select 842_DECOMPRESS
+	default y
+	help
+	  Support for PowerPC Nest (NX) accelerators using the cryptographic
+	  API.  If you choose 'M' here, this module will be called
+	  nx_compress_crypto.
+
+endif
diff --git a/drivers/crypto/nx/Makefile b/drivers/crypto/nx/Makefile
index bb770ea45..e1684f5ad 100644
--- a/drivers/crypto/nx/Makefile
+++ b/drivers/crypto/nx/Makefile
@@ -10,5 +10,12 @@ nx-crypto-objs := nx.o \
 		  nx-sha256.o \
 		  nx-sha512.o
 
-obj-$(CONFIG_CRYPTO_DEV_NX_COMPRESS) += nx-compress.o
+obj-$(CONFIG_CRYPTO_DEV_NX_COMPRESS) += nx-compress.o nx-compress-platform.o
+obj-$(CONFIG_CRYPTO_DEV_NX_COMPRESS_PSERIES) += nx-compress-pseries.o
+obj-$(CONFIG_CRYPTO_DEV_NX_COMPRESS_POWERNV) += nx-compress-powernv.o
+obj-$(CONFIG_CRYPTO_DEV_NX_COMPRESS_CRYPTO) += nx-compress-crypto.o
 nx-compress-objs := nx-842.o
+nx-compress-platform-objs := nx-842-platform.o
+nx-compress-pseries-objs := nx-842-pseries.o
+nx-compress-powernv-objs := nx-842-powernv.o
+nx-compress-crypto-objs := nx-842-crypto.o
diff --git a/drivers/crypto/nx/nx-842-crypto.c b/drivers/crypto/nx/nx-842-crypto.c
new file mode 100644
index 000000000..d53a1dcd7
--- /dev/null
+++ b/drivers/crypto/nx/nx-842-crypto.c
@@ -0,0 +1,580 @@
+/*
+ * Cryptographic API for the NX-842 hardware compression.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * Copyright (C) IBM Corporation, 2011-2015
+ *
+ * Original Authors: Robert Jennings <rcj@linux.vnet.ibm.com>
+ *                   Seth Jennings <sjenning@linux.vnet.ibm.com>
+ *
+ * Rewrite: Dan Streetman <ddstreet@ieee.org>
+ *
+ * This is an interface to the NX-842 compression hardware in PowerPC
+ * processors.  Most of the complexity of this drvier is due to the fact that
+ * the NX-842 compression hardware requires the input and output data buffers
+ * to be specifically aligned, to be a specific multiple in length, and within
+ * specific minimum and maximum lengths.  Those restrictions, provided by the
+ * nx-842 driver via nx842_constraints, mean this driver must use bounce
+ * buffers and headers to correct misaligned in or out buffers, and to split
+ * input buffers that are too large.
+ *
+ * This driver will fall back to software decompression if the hardware
+ * decompression fails, so this driver's decompression should never fail as
+ * long as the provided compressed buffer is valid.  Any compressed buffer
+ * created by this driver will have a header (except ones where the input
+ * perfectly matches the constraints); so users of this driver cannot simply
+ * pass a compressed buffer created by this driver over to the 842 software
+ * decompression library.  Instead, users must use this driver to decompress;
+ * if the hardware fails or is unavailable, the compressed buffer will be
+ * parsed and the header removed, and the raw 842 buffer(s) passed to the 842
+ * software decompression library.
+ *
+ * This does not fall back to software compression, however, since the caller
+ * of this function is specifically requesting hardware compression; if the
+ * hardware compression fails, the caller can fall back to software
+ * compression, and the raw 842 compressed buffer that the software compressor
+ * creates can be passed to this driver for hardware decompression; any
+ * buffer without our specific header magic is assumed to be a raw 842 buffer
+ * and passed directly to the hardware.  Note that the software compression
+ * library will produce a compressed buffer that is incompatible with the
+ * hardware decompressor if the original input buffer length is not a multiple
+ * of 8; if such a compressed buffer is passed to this driver for
+ * decompression, the hardware will reject it and this driver will then pass
+ * it over to the software library for decompression.
+ */
+
+#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
+
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/crypto.h>
+#include <linux/vmalloc.h>
+#include <linux/sw842.h>
+#include <linux/ratelimit.h>
+
+#include "nx-842.h"
+
+/* The first 5 bits of this magic are 0x1f, which is an invalid 842 5-bit
+ * template (see lib/842/842.h), so this magic number will never appear at
+ * the start of a raw 842 compressed buffer.  That is important, as any buffer
+ * passed to us without this magic is assumed to be a raw 842 compressed
+ * buffer, and passed directly to the hardware to decompress.
+ */
+#define NX842_CRYPTO_MAGIC	(0xf842)
+#define NX842_CRYPTO_GROUP_MAX	(0x20)
+#define NX842_CRYPTO_HEADER_SIZE(g)				\
+	(sizeof(struct nx842_crypto_header) +			\
+	 sizeof(struct nx842_crypto_header_group) * (g))
+#define NX842_CRYPTO_HEADER_MAX_SIZE				\
+	NX842_CRYPTO_HEADER_SIZE(NX842_CRYPTO_GROUP_MAX)
+
+/* bounce buffer size */
+#define BOUNCE_BUFFER_ORDER	(2)
+#define BOUNCE_BUFFER_SIZE					\
+	((unsigned int)(PAGE_SIZE << BOUNCE_BUFFER_ORDER))
+
+/* try longer on comp because we can fallback to sw decomp if hw is busy */
+#define COMP_BUSY_TIMEOUT	(250) /* ms */
+#define DECOMP_BUSY_TIMEOUT	(50) /* ms */
+
+struct nx842_crypto_header_group {
+	__be16 padding;			/* unused bytes at start of group */
+	__be32 compressed_length;	/* compressed bytes in group */
+	__be32 uncompressed_length;	/* bytes after decompression */
+} __packed;
+
+struct nx842_crypto_header {
+	__be16 magic;		/* NX842_CRYPTO_MAGIC */
+	__be16 ignore;		/* decompressed end bytes to ignore */
+	u8 groups;		/* total groups in this header */
+	struct nx842_crypto_header_group group[];
+} __packed;
+
+struct nx842_crypto_param {
+	u8 *in;
+	unsigned int iremain;
+	u8 *out;
+	unsigned int oremain;
+	unsigned int ototal;
+};
+
+static int update_param(struct nx842_crypto_param *p,
+			unsigned int slen, unsigned int dlen)
+{
+	if (p->iremain < slen)
+		return -EOVERFLOW;
+	if (p->oremain < dlen)
+		return -ENOSPC;
+
+	p->in += slen;
+	p->iremain -= slen;
+	p->out += dlen;
+	p->oremain -= dlen;
+	p->ototal += dlen;
+
+	return 0;
+}
+
+struct nx842_crypto_ctx {
+	u8 *wmem;
+	u8 *sbounce, *dbounce;
+
+	struct nx842_crypto_header header;
+	struct nx842_crypto_header_group group[NX842_CRYPTO_GROUP_MAX];
+};
+
+static int nx842_crypto_init(struct crypto_tfm *tfm)
+{
+	struct nx842_crypto_ctx *ctx = crypto_tfm_ctx(tfm);
+
+	ctx->wmem = kmalloc(nx842_workmem_size(), GFP_KERNEL);
+	ctx->sbounce = (u8 *)__get_free_pages(GFP_KERNEL, BOUNCE_BUFFER_ORDER);
+	ctx->dbounce = (u8 *)__get_free_pages(GFP_KERNEL, BOUNCE_BUFFER_ORDER);
+	if (!ctx->wmem || !ctx->sbounce || !ctx->dbounce) {
+		kfree(ctx->wmem);
+		free_page((unsigned long)ctx->sbounce);
+		free_page((unsigned long)ctx->dbounce);
+		return -ENOMEM;
+	}
+
+	return 0;
+}
+
+static void nx842_crypto_exit(struct crypto_tfm *tfm)
+{
+	struct nx842_crypto_ctx *ctx = crypto_tfm_ctx(tfm);
+
+	kfree(ctx->wmem);
+	free_page((unsigned long)ctx->sbounce);
+	free_page((unsigned long)ctx->dbounce);
+}
+
+static int read_constraints(struct nx842_constraints *c)
+{
+	int ret;
+
+	ret = nx842_constraints(c);
+	if (ret) {
+		pr_err_ratelimited("could not get nx842 constraints : %d\n",
+				   ret);
+		return ret;
+	}
+
+	/* limit maximum, to always have enough bounce buffer to decompress */
+	if (c->maximum > BOUNCE_BUFFER_SIZE) {
+		c->maximum = BOUNCE_BUFFER_SIZE;
+		pr_info_once("limiting nx842 maximum to %x\n", c->maximum);
+	}
+
+	return 0;
+}
+
+static int nx842_crypto_add_header(struct nx842_crypto_header *hdr, u8 *buf)
+{
+	int s = NX842_CRYPTO_HEADER_SIZE(hdr->groups);
+
+	/* compress should have added space for header */
+	if (s > be16_to_cpu(hdr->group[0].padding)) {
+		pr_err("Internal error: no space for header\n");
+		return -EINVAL;
+	}
+
+	memcpy(buf, hdr, s);
+
+	print_hex_dump_debug("header ", DUMP_PREFIX_OFFSET, 16, 1, buf, s, 0);
+
+	return 0;
+}
+
+static int compress(struct nx842_crypto_ctx *ctx,
+		    struct nx842_crypto_param *p,
+		    struct nx842_crypto_header_group *g,
+		    struct nx842_constraints *c,
+		    u16 *ignore,
+		    unsigned int hdrsize)
+{
+	unsigned int slen = p->iremain, dlen = p->oremain, tmplen;
+	unsigned int adj_slen = slen;
+	u8 *src = p->in, *dst = p->out;
+	int ret, dskip = 0;
+	ktime_t timeout;
+
+	if (p->iremain == 0)
+		return -EOVERFLOW;
+
+	if (p->oremain == 0 || hdrsize + c->minimum > dlen)
+		return -ENOSPC;
+
+	if (slen % c->multiple)
+		adj_slen = round_up(slen, c->multiple);
+	if (slen < c->minimum)
+		adj_slen = c->minimum;
+	if (slen > c->maximum)
+		adj_slen = slen = c->maximum;
+	if (adj_slen > slen || (u64)src % c->alignment) {
+		adj_slen = min(adj_slen, BOUNCE_BUFFER_SIZE);
+		slen = min(slen, BOUNCE_BUFFER_SIZE);
+		if (adj_slen > slen)
+			memset(ctx->sbounce + slen, 0, adj_slen - slen);
+		memcpy(ctx->sbounce, src, slen);
+		src = ctx->sbounce;
+		slen = adj_slen;
+		pr_debug("using comp sbounce buffer, len %x\n", slen);
+	}
+
+	dst += hdrsize;
+	dlen -= hdrsize;
+
+	if ((u64)dst % c->alignment) {
+		dskip = (int)(PTR_ALIGN(dst, c->alignment) - dst);
+		dst += dskip;
+		dlen -= dskip;
+	}
+	if (dlen % c->multiple)
+		dlen = round_down(dlen, c->multiple);
+	if (dlen < c->minimum) {
+nospc:
+		dst = ctx->dbounce;
+		dlen = min(p->oremain, BOUNCE_BUFFER_SIZE);
+		dlen = round_down(dlen, c->multiple);
+		dskip = 0;
+		pr_debug("using comp dbounce buffer, len %x\n", dlen);
+	}
+	if (dlen > c->maximum)
+		dlen = c->maximum;
+
+	tmplen = dlen;
+	timeout = ktime_add_ms(ktime_get(), COMP_BUSY_TIMEOUT);
+	do {
+		dlen = tmplen; /* reset dlen, if we're retrying */
+		ret = nx842_compress(src, slen, dst, &dlen, ctx->wmem);
+		/* possibly we should reduce the slen here, instead of
+		 * retrying with the dbounce buffer?
+		 */
+		if (ret == -ENOSPC && dst != ctx->dbounce)
+			goto nospc;
+	} while (ret == -EBUSY && ktime_before(ktime_get(), timeout));
+	if (ret)
+		return ret;
+
+	dskip += hdrsize;
+
+	if (dst == ctx->dbounce)
+		memcpy(p->out + dskip, dst, dlen);
+
+	g->padding = cpu_to_be16(dskip);
+	g->compressed_length = cpu_to_be32(dlen);
+	g->uncompressed_length = cpu_to_be32(slen);
+
+	if (p->iremain < slen) {
+		*ignore = slen - p->iremain;
+		slen = p->iremain;
+	}
+
+	pr_debug("compress slen %x ignore %x dlen %x padding %x\n",
+		 slen, *ignore, dlen, dskip);
+
+	return update_param(p, slen, dskip + dlen);
+}
+
+static int nx842_crypto_compress(struct crypto_tfm *tfm,
+				 const u8 *src, unsigned int slen,
+				 u8 *dst, unsigned int *dlen)
+{
+	struct nx842_crypto_ctx *ctx = crypto_tfm_ctx(tfm);
+	struct nx842_crypto_header *hdr = &ctx->header;
+	struct nx842_crypto_param p;
+	struct nx842_constraints c;
+	unsigned int groups, hdrsize, h;
+	int ret, n;
+	bool add_header;
+	u16 ignore = 0;
+
+	p.in = (u8 *)src;
+	p.iremain = slen;
+	p.out = dst;
+	p.oremain = *dlen;
+	p.ototal = 0;
+
+	*dlen = 0;
+
+	ret = read_constraints(&c);
+	if (ret)
+		return ret;
+
+	groups = min_t(unsigned int, NX842_CRYPTO_GROUP_MAX,
+		       DIV_ROUND_UP(p.iremain, c.maximum));
+	hdrsize = NX842_CRYPTO_HEADER_SIZE(groups);
+
+	/* skip adding header if the buffers meet all constraints */
+	add_header = (p.iremain % c.multiple	||
+		      p.iremain < c.minimum	||
+		      p.iremain > c.maximum	||
+		      (u64)p.in % c.alignment	||
+		      p.oremain % c.multiple	||
+		      p.oremain < c.minimum	||
+		      p.oremain > c.maximum	||
+		      (u64)p.out % c.alignment);
+
+	hdr->magic = cpu_to_be16(NX842_CRYPTO_MAGIC);
+	hdr->groups = 0;
+	hdr->ignore = 0;
+
+	while (p.iremain > 0) {
+		n = hdr->groups++;
+		if (hdr->groups > NX842_CRYPTO_GROUP_MAX)
+			return -ENOSPC;
+
+		/* header goes before first group */
+		h = !n && add_header ? hdrsize : 0;
+
+		if (ignore)
+			pr_warn("interal error, ignore is set %x\n", ignore);
+
+		ret = compress(ctx, &p, &hdr->group[n], &c, &ignore, h);
+		if (ret)
+			return ret;
+	}
+
+	if (!add_header && hdr->groups > 1) {
+		pr_err("Internal error: No header but multiple groups\n");
+		return -EINVAL;
+	}
+
+	/* ignore indicates the input stream needed to be padded */
+	hdr->ignore = cpu_to_be16(ignore);
+	if (ignore)
+		pr_debug("marked %d bytes as ignore\n", ignore);
+
+	if (add_header)
+		ret = nx842_crypto_add_header(hdr, dst);
+	if (ret)
+		return ret;
+
+	*dlen = p.ototal;
+
+	pr_debug("compress total slen %x dlen %x\n", slen, *dlen);
+
+	return 0;
+}
+
+static int decompress(struct nx842_crypto_ctx *ctx,
+		      struct nx842_crypto_param *p,
+		      struct nx842_crypto_header_group *g,
+		      struct nx842_constraints *c,
+		      u16 ignore,
+		      bool usehw)
+{
+	unsigned int slen = be32_to_cpu(g->compressed_length);
+	unsigned int required_len = be32_to_cpu(g->uncompressed_length);
+	unsigned int dlen = p->oremain, tmplen;
+	unsigned int adj_slen = slen;
+	u8 *src = p->in, *dst = p->out;
+	u16 padding = be16_to_cpu(g->padding);
+	int ret, spadding = 0, dpadding = 0;
+	ktime_t timeout;
+
+	if (!slen || !required_len)
+		return -EINVAL;
+
+	if (p->iremain <= 0 || padding + slen > p->iremain)
+		return -EOVERFLOW;
+
+	if (p->oremain <= 0 || required_len - ignore > p->oremain)
+		return -ENOSPC;
+
+	src += padding;
+
+	if (!usehw)
+		goto usesw;
+
+	if (slen % c->multiple)
+		adj_slen = round_up(slen, c->multiple);
+	if (slen < c->minimum)
+		adj_slen = c->minimum;
+	if (slen > c->maximum)
+		goto usesw;
+	if (slen < adj_slen || (u64)src % c->alignment) {
+		/* we can append padding bytes because the 842 format defines
+		 * an "end" template (see lib/842/842_decompress.c) and will
+		 * ignore any bytes following it.
+		 */
+		if (slen < adj_slen)
+			memset(ctx->sbounce + slen, 0, adj_slen - slen);
+		memcpy(ctx->sbounce, src, slen);
+		src = ctx->sbounce;
+		spadding = adj_slen - slen;
+		slen = adj_slen;
+		pr_debug("using decomp sbounce buffer, len %x\n", slen);
+	}
+
+	if (dlen % c->multiple)
+		dlen = round_down(dlen, c->multiple);
+	if (dlen < required_len || (u64)dst % c->alignment) {
+		dst = ctx->dbounce;
+		dlen = min(required_len, BOUNCE_BUFFER_SIZE);
+		pr_debug("using decomp dbounce buffer, len %x\n", dlen);
+	}
+	if (dlen < c->minimum)
+		goto usesw;
+	if (dlen > c->maximum)
+		dlen = c->maximum;
+
+	tmplen = dlen;
+	timeout = ktime_add_ms(ktime_get(), DECOMP_BUSY_TIMEOUT);
+	do {
+		dlen = tmplen; /* reset dlen, if we're retrying */
+		ret = nx842_decompress(src, slen, dst, &dlen, ctx->wmem);
+	} while (ret == -EBUSY && ktime_before(ktime_get(), timeout));
+	if (ret) {
+usesw:
+		/* reset everything, sw doesn't have constraints */
+		src = p->in + padding;
+		slen = be32_to_cpu(g->compressed_length);
+		spadding = 0;
+		dst = p->out;
+		dlen = p->oremain;
+		dpadding = 0;
+		if (dlen < required_len) { /* have ignore bytes */
+			dst = ctx->dbounce;
+			dlen = BOUNCE_BUFFER_SIZE;
+		}
+		pr_info_ratelimited("using software 842 decompression\n");
+		ret = sw842_decompress(src, slen, dst, &dlen);
+	}
+	if (ret)
+		return ret;
+
+	slen -= spadding;
+
+	dlen -= ignore;
+	if (ignore)
+		pr_debug("ignoring last %x bytes\n", ignore);
+
+	if (dst == ctx->dbounce)
+		memcpy(p->out, dst, dlen);
+
+	pr_debug("decompress slen %x padding %x dlen %x ignore %x\n",
+		 slen, padding, dlen, ignore);
+
+	return update_param(p, slen + padding, dlen);
+}
+
+static int nx842_crypto_decompress(struct crypto_tfm *tfm,
+				   const u8 *src, unsigned int slen,
+				   u8 *dst, unsigned int *dlen)
+{
+	struct nx842_crypto_ctx *ctx = crypto_tfm_ctx(tfm);
+	struct nx842_crypto_header *hdr;
+	struct nx842_crypto_param p;
+	struct nx842_constraints c;
+	int n, ret, hdr_len;
+	u16 ignore = 0;
+	bool usehw = true;
+
+	p.in = (u8 *)src;
+	p.iremain = slen;
+	p.out = dst;
+	p.oremain = *dlen;
+	p.ototal = 0;
+
+	*dlen = 0;
+
+	if (read_constraints(&c))
+		usehw = false;
+
+	hdr = (struct nx842_crypto_header *)src;
+
+	/* If it doesn't start with our header magic number, assume it's a raw
+	 * 842 compressed buffer and pass it directly to the hardware driver
+	 */
+	if (be16_to_cpu(hdr->magic) != NX842_CRYPTO_MAGIC) {
+		struct nx842_crypto_header_group g = {
+			.padding =		0,
+			.compressed_length =	cpu_to_be32(p.iremain),
+			.uncompressed_length =	cpu_to_be32(p.oremain),
+		};
+
+		ret = decompress(ctx, &p, &g, &c, 0, usehw);
+		if (ret)
+			return ret;
+
+		*dlen = p.ototal;
+
+		return 0;
+	}
+
+	if (!hdr->groups) {
+		pr_err("header has no groups\n");
+		return -EINVAL;
+	}
+	if (hdr->groups > NX842_CRYPTO_GROUP_MAX) {
+		pr_err("header has too many groups %x, max %x\n",
+		       hdr->groups, NX842_CRYPTO_GROUP_MAX);
+		return -EINVAL;
+	}
+
+	hdr_len = NX842_CRYPTO_HEADER_SIZE(hdr->groups);
+	if (hdr_len > slen)
+		return -EOVERFLOW;
+
+	memcpy(&ctx->header, src, hdr_len);
+	hdr = &ctx->header;
+
+	for (n = 0; n < hdr->groups; n++) {
+		/* ignore applies to last group */
+		if (n + 1 == hdr->groups)
+			ignore = be16_to_cpu(hdr->ignore);
+
+		ret = decompress(ctx, &p, &hdr->group[n], &c, ignore, usehw);
+		if (ret)
+			return ret;
+	}
+
+	*dlen = p.ototal;
+
+	pr_debug("decompress total slen %x dlen %x\n", slen, *dlen);
+
+	return 0;
+}
+
+static struct crypto_alg alg = {
+	.cra_name		= "842",
+	.cra_driver_name	= "842-nx",
+	.cra_priority		= 300,
+	.cra_flags		= CRYPTO_ALG_TYPE_COMPRESS,
+	.cra_ctxsize		= sizeof(struct nx842_crypto_ctx),
+	.cra_module		= THIS_MODULE,
+	.cra_init		= nx842_crypto_init,
+	.cra_exit		= nx842_crypto_exit,
+	.cra_u			= { .compress = {
+	.coa_compress		= nx842_crypto_compress,
+	.coa_decompress		= nx842_crypto_decompress } }
+};
+
+static int __init nx842_crypto_mod_init(void)
+{
+	return crypto_register_alg(&alg);
+}
+module_init(nx842_crypto_mod_init);
+
+static void __exit nx842_crypto_mod_exit(void)
+{
+	crypto_unregister_alg(&alg);
+}
+module_exit(nx842_crypto_mod_exit);
+
+MODULE_LICENSE("GPL");
+MODULE_DESCRIPTION("IBM PowerPC Nest (NX) 842 Hardware Compression Interface");
+MODULE_ALIAS_CRYPTO("842");
+MODULE_ALIAS_CRYPTO("842-nx");
+MODULE_AUTHOR("Dan Streetman <ddstreet@ieee.org>");
diff --git a/drivers/crypto/nx/nx-842-platform.c b/drivers/crypto/nx/nx-842-platform.c
new file mode 100644
index 000000000..664f13dd0
--- /dev/null
+++ b/drivers/crypto/nx/nx-842-platform.c
@@ -0,0 +1,84 @@
+
+#include "nx-842.h"
+
+/* this is needed, separate from the main nx-842.c driver, because that main
+ * driver loads the platform drivers during its init(), and it expects one
+ * (or none) of the platform drivers to set this pointer to its driver.
+ * That means this pointer can't be in the main nx-842 driver, because it
+ * wouldn't be accessible until after the main driver loaded, which wouldn't
+ * be possible as it's waiting for the platform driver to load.  So place it
+ * here.
+ */
+static struct nx842_driver *driver;
+static DEFINE_SPINLOCK(driver_lock);
+
+struct nx842_driver *nx842_platform_driver(void)
+{
+	return driver;
+}
+EXPORT_SYMBOL_GPL(nx842_platform_driver);
+
+bool nx842_platform_driver_set(struct nx842_driver *_driver)
+{
+	bool ret = false;
+
+	spin_lock(&driver_lock);
+
+	if (!driver) {
+		driver = _driver;
+		ret = true;
+	} else
+		WARN(1, "can't set platform driver, already set to %s\n",
+		     driver->name);
+
+	spin_unlock(&driver_lock);
+	return ret;
+}
+EXPORT_SYMBOL_GPL(nx842_platform_driver_set);
+
+/* only call this from the platform driver exit function */
+void nx842_platform_driver_unset(struct nx842_driver *_driver)
+{
+	spin_lock(&driver_lock);
+
+	if (driver == _driver)
+		driver = NULL;
+	else if (driver)
+		WARN(1, "can't unset platform driver %s, currently set to %s\n",
+		     _driver->name, driver->name);
+	else
+		WARN(1, "can't unset platform driver, already unset\n");
+
+	spin_unlock(&driver_lock);
+}
+EXPORT_SYMBOL_GPL(nx842_platform_driver_unset);
+
+bool nx842_platform_driver_get(void)
+{
+	bool ret = false;
+
+	spin_lock(&driver_lock);
+
+	if (driver)
+		ret = try_module_get(driver->owner);
+
+	spin_unlock(&driver_lock);
+
+	return ret;
+}
+EXPORT_SYMBOL_GPL(nx842_platform_driver_get);
+
+void nx842_platform_driver_put(void)
+{
+	spin_lock(&driver_lock);
+
+	if (driver)
+		module_put(driver->owner);
+
+	spin_unlock(&driver_lock);
+}
+EXPORT_SYMBOL_GPL(nx842_platform_driver_put);
+
+MODULE_LICENSE("GPL");
+MODULE_AUTHOR("Dan Streetman <ddstreet@ieee.org>");
+MODULE_DESCRIPTION("842 H/W Compression platform driver");
diff --git a/drivers/crypto/nx/nx-842-powernv.c b/drivers/crypto/nx/nx-842-powernv.c
new file mode 100644
index 000000000..33b3b0abf
--- /dev/null
+++ b/drivers/crypto/nx/nx-842-powernv.c
@@ -0,0 +1,637 @@
+/*
+ * Driver for IBM PowerNV 842 compression accelerator
+ *
+ * Copyright (C) 2015 Dan Streetman, IBM Corp
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
+
+#include "nx-842.h"
+
+#include <linux/timer.h>
+
+#include <asm/prom.h>
+#include <asm/icswx.h>
+
+MODULE_LICENSE("GPL");
+MODULE_AUTHOR("Dan Streetman <ddstreet@ieee.org>");
+MODULE_DESCRIPTION("842 H/W Compression driver for IBM PowerNV processors");
+
+#define WORKMEM_ALIGN	(CRB_ALIGN)
+#define CSB_WAIT_MAX	(5000) /* ms */
+
+struct nx842_workmem {
+	/* Below fields must be properly aligned */
+	struct coprocessor_request_block crb; /* CRB_ALIGN align */
+	struct data_descriptor_entry ddl_in[DDL_LEN_MAX]; /* DDE_ALIGN align */
+	struct data_descriptor_entry ddl_out[DDL_LEN_MAX]; /* DDE_ALIGN align */
+	/* Above fields must be properly aligned */
+
+	ktime_t start;
+
+	char padding[WORKMEM_ALIGN]; /* unused, to allow alignment */
+} __packed __aligned(WORKMEM_ALIGN);
+
+struct nx842_coproc {
+	unsigned int chip_id;
+	unsigned int ct;
+	unsigned int ci;
+	struct list_head list;
+};
+
+/* no cpu hotplug on powernv, so this list never changes after init */
+static LIST_HEAD(nx842_coprocs);
+static unsigned int nx842_ct;
+
+/**
+ * setup_indirect_dde - Setup an indirect DDE
+ *
+ * The DDE is setup with the the DDE count, byte count, and address of
+ * first direct DDE in the list.
+ */
+static void setup_indirect_dde(struct data_descriptor_entry *dde,
+			       struct data_descriptor_entry *ddl,
+			       unsigned int dde_count, unsigned int byte_count)
+{
+	dde->flags = 0;
+	dde->count = dde_count;
+	dde->index = 0;
+	dde->length = cpu_to_be32(byte_count);
+	dde->address = cpu_to_be64(nx842_get_pa(ddl));
+}
+
+/**
+ * setup_direct_dde - Setup single DDE from buffer
+ *
+ * The DDE is setup with the buffer and length.  The buffer must be properly
+ * aligned.  The used length is returned.
+ * Returns:
+ *   N    Successfully set up DDE with N bytes
+ */
+static unsigned int setup_direct_dde(struct data_descriptor_entry *dde,
+				     unsigned long pa, unsigned int len)
+{
+	unsigned int l = min_t(unsigned int, len, LEN_ON_PAGE(pa));
+
+	dde->flags = 0;
+	dde->count = 0;
+	dde->index = 0;
+	dde->length = cpu_to_be32(l);
+	dde->address = cpu_to_be64(pa);
+
+	return l;
+}
+
+/**
+ * setup_ddl - Setup DDL from buffer
+ *
+ * Returns:
+ *   0		Successfully set up DDL
+ */
+static int setup_ddl(struct data_descriptor_entry *dde,
+		     struct data_descriptor_entry *ddl,
+		     unsigned char *buf, unsigned int len,
+		     bool in)
+{
+	unsigned long pa = nx842_get_pa(buf);
+	int i, ret, total_len = len;
+
+	if (!IS_ALIGNED(pa, DDE_BUFFER_ALIGN)) {
+		pr_debug("%s buffer pa 0x%lx not 0x%x-byte aligned\n",
+			 in ? "input" : "output", pa, DDE_BUFFER_ALIGN);
+		return -EINVAL;
+	}
+
+	/* only need to check last mult; since buffer must be
+	 * DDE_BUFFER_ALIGN aligned, and that is a multiple of
+	 * DDE_BUFFER_SIZE_MULT, and pre-last page DDE buffers
+	 * are guaranteed a multiple of DDE_BUFFER_SIZE_MULT.
+	 */
+	if (len % DDE_BUFFER_LAST_MULT) {
+		pr_debug("%s buffer len 0x%x not a multiple of 0x%x\n",
+			 in ? "input" : "output", len, DDE_BUFFER_LAST_MULT);
+		if (in)
+			return -EINVAL;
+		len = round_down(len, DDE_BUFFER_LAST_MULT);
+	}
+
+	/* use a single direct DDE */
+	if (len <= LEN_ON_PAGE(pa)) {
+		ret = setup_direct_dde(dde, pa, len);
+		WARN_ON(ret < len);
+		return 0;
+	}
+
+	/* use the DDL */
+	for (i = 0; i < DDL_LEN_MAX && len > 0; i++) {
+		ret = setup_direct_dde(&ddl[i], pa, len);
+		buf += ret;
+		len -= ret;
+		pa = nx842_get_pa(buf);
+	}
+
+	if (len > 0) {
+		pr_debug("0x%x total %s bytes 0x%x too many for DDL.\n",
+			 total_len, in ? "input" : "output", len);
+		if (in)
+			return -EMSGSIZE;
+		total_len -= len;
+	}
+	setup_indirect_dde(dde, ddl, i, total_len);
+
+	return 0;
+}
+
+#define CSB_ERR(csb, msg, ...)					\
+	pr_err("ERROR: " msg " : %02x %02x %02x %02x %08x\n",	\
+	       ##__VA_ARGS__, (csb)->flags,			\
+	       (csb)->cs, (csb)->cc, (csb)->ce,			\
+	       be32_to_cpu((csb)->count))
+
+#define CSB_ERR_ADDR(csb, msg, ...)				\
+	CSB_ERR(csb, msg " at %lx", ##__VA_ARGS__,		\
+		(unsigned long)be64_to_cpu((csb)->address))
+
+/**
+ * wait_for_csb
+ */
+static int wait_for_csb(struct nx842_workmem *wmem,
+			struct coprocessor_status_block *csb)
+{
+	ktime_t start = wmem->start, now = ktime_get();
+	ktime_t timeout = ktime_add_ms(start, CSB_WAIT_MAX);
+
+	while (!(ACCESS_ONCE(csb->flags) & CSB_V)) {
+		cpu_relax();
+		now = ktime_get();
+		if (ktime_after(now, timeout))
+			break;
+	}
+
+	/* hw has updated csb and output buffer */
+	barrier();
+
+	/* check CSB flags */
+	if (!(csb->flags & CSB_V)) {
+		CSB_ERR(csb, "CSB still not valid after %ld us, giving up",
+			(long)ktime_us_delta(now, start));
+		return -ETIMEDOUT;
+	}
+	if (csb->flags & CSB_F) {
+		CSB_ERR(csb, "Invalid CSB format");
+		return -EPROTO;
+	}
+	if (csb->flags & CSB_CH) {
+		CSB_ERR(csb, "Invalid CSB chaining state");
+		return -EPROTO;
+	}
+
+	/* verify CSB completion sequence is 0 */
+	if (csb->cs) {
+		CSB_ERR(csb, "Invalid CSB completion sequence");
+		return -EPROTO;
+	}
+
+	/* check CSB Completion Code */
+	switch (csb->cc) {
+	/* no error */
+	case CSB_CC_SUCCESS:
+		break;
+	case CSB_CC_TPBC_GT_SPBC:
+		/* not an error, but the compressed data is
+		 * larger than the uncompressed data :(
+		 */
+		break;
+
+	/* input data errors */
+	case CSB_CC_OPERAND_OVERLAP:
+		/* input and output buffers overlap */
+		CSB_ERR(csb, "Operand Overlap error");
+		return -EINVAL;
+	case CSB_CC_INVALID_OPERAND:
+		CSB_ERR(csb, "Invalid operand");
+		return -EINVAL;
+	case CSB_CC_NOSPC:
+		/* output buffer too small */
+		return -ENOSPC;
+	case CSB_CC_ABORT:
+		CSB_ERR(csb, "Function aborted");
+		return -EINTR;
+	case CSB_CC_CRC_MISMATCH:
+		CSB_ERR(csb, "CRC mismatch");
+		return -EINVAL;
+	case CSB_CC_TEMPL_INVALID:
+		CSB_ERR(csb, "Compressed data template invalid");
+		return -EINVAL;
+	case CSB_CC_TEMPL_OVERFLOW:
+		CSB_ERR(csb, "Compressed data template shows data past end");
+		return -EINVAL;
+
+	/* these should not happen */
+	case CSB_CC_INVALID_ALIGN:
+		/* setup_ddl should have detected this */
+		CSB_ERR_ADDR(csb, "Invalid alignment");
+		return -EINVAL;
+	case CSB_CC_DATA_LENGTH:
+		/* setup_ddl should have detected this */
+		CSB_ERR(csb, "Invalid data length");
+		return -EINVAL;
+	case CSB_CC_WR_TRANSLATION:
+	case CSB_CC_TRANSLATION:
+	case CSB_CC_TRANSLATION_DUP1:
+	case CSB_CC_TRANSLATION_DUP2:
+	case CSB_CC_TRANSLATION_DUP3:
+	case CSB_CC_TRANSLATION_DUP4:
+	case CSB_CC_TRANSLATION_DUP5:
+	case CSB_CC_TRANSLATION_DUP6:
+		/* should not happen, we use physical addrs */
+		CSB_ERR_ADDR(csb, "Translation error");
+		return -EPROTO;
+	case CSB_CC_WR_PROTECTION:
+	case CSB_CC_PROTECTION:
+	case CSB_CC_PROTECTION_DUP1:
+	case CSB_CC_PROTECTION_DUP2:
+	case CSB_CC_PROTECTION_DUP3:
+	case CSB_CC_PROTECTION_DUP4:
+	case CSB_CC_PROTECTION_DUP5:
+	case CSB_CC_PROTECTION_DUP6:
+		/* should not happen, we use physical addrs */
+		CSB_ERR_ADDR(csb, "Protection error");
+		return -EPROTO;
+	case CSB_CC_PRIVILEGE:
+		/* shouldn't happen, we're in HYP mode */
+		CSB_ERR(csb, "Insufficient Privilege error");
+		return -EPROTO;
+	case CSB_CC_EXCESSIVE_DDE:
+		/* shouldn't happen, setup_ddl doesn't use many dde's */
+		CSB_ERR(csb, "Too many DDEs in DDL");
+		return -EINVAL;
+	case CSB_CC_TRANSPORT:
+		/* shouldn't happen, we setup CRB correctly */
+		CSB_ERR(csb, "Invalid CRB");
+		return -EINVAL;
+	case CSB_CC_SEGMENTED_DDL:
+		/* shouldn't happen, setup_ddl creates DDL right */
+		CSB_ERR(csb, "Segmented DDL error");
+		return -EINVAL;
+	case CSB_CC_DDE_OVERFLOW:
+		/* shouldn't happen, setup_ddl creates DDL right */
+		CSB_ERR(csb, "DDE overflow error");
+		return -EINVAL;
+	case CSB_CC_SESSION:
+		/* should not happen with ICSWX */
+		CSB_ERR(csb, "Session violation error");
+		return -EPROTO;
+	case CSB_CC_CHAIN:
+		/* should not happen, we don't use chained CRBs */
+		CSB_ERR(csb, "Chained CRB error");
+		return -EPROTO;
+	case CSB_CC_SEQUENCE:
+		/* should not happen, we don't use chained CRBs */
+		CSB_ERR(csb, "CRB seqeunce number error");
+		return -EPROTO;
+	case CSB_CC_UNKNOWN_CODE:
+		CSB_ERR(csb, "Unknown subfunction code");
+		return -EPROTO;
+
+	/* hardware errors */
+	case CSB_CC_RD_EXTERNAL:
+	case CSB_CC_RD_EXTERNAL_DUP1:
+	case CSB_CC_RD_EXTERNAL_DUP2:
+	case CSB_CC_RD_EXTERNAL_DUP3:
+		CSB_ERR_ADDR(csb, "Read error outside coprocessor");
+		return -EPROTO;
+	case CSB_CC_WR_EXTERNAL:
+		CSB_ERR_ADDR(csb, "Write error outside coprocessor");
+		return -EPROTO;
+	case CSB_CC_INTERNAL:
+		CSB_ERR(csb, "Internal error in coprocessor");
+		return -EPROTO;
+	case CSB_CC_PROVISION:
+		CSB_ERR(csb, "Storage provision error");
+		return -EPROTO;
+	case CSB_CC_HW:
+		CSB_ERR(csb, "Correctable hardware error");
+		return -EPROTO;
+
+	default:
+		CSB_ERR(csb, "Invalid CC %d", csb->cc);
+		return -EPROTO;
+	}
+
+	/* check Completion Extension state */
+	if (csb->ce & CSB_CE_TERMINATION) {
+		CSB_ERR(csb, "CSB request was terminated");
+		return -EPROTO;
+	}
+	if (csb->ce & CSB_CE_INCOMPLETE) {
+		CSB_ERR(csb, "CSB request not complete");
+		return -EPROTO;
+	}
+	if (!(csb->ce & CSB_CE_TPBC)) {
+		CSB_ERR(csb, "TPBC not provided, unknown target length");
+		return -EPROTO;
+	}
+
+	/* successful completion */
+	pr_debug_ratelimited("Processed %u bytes in %lu us\n", csb->count,
+			     (unsigned long)ktime_us_delta(now, start));
+
+	return 0;
+}
+
+/**
+ * nx842_powernv_function - compress/decompress data using the 842 algorithm
+ *
+ * (De)compression provided by the NX842 coprocessor on IBM PowerNV systems.
+ * This compresses or decompresses the provided input buffer into the provided
+ * output buffer.
+ *
+ * Upon return from this function @outlen contains the length of the
+ * output data.  If there is an error then @outlen will be 0 and an
+ * error will be specified by the return code from this function.
+ *
+ * The @workmem buffer should only be used by one function call at a time.
+ *
+ * @in: input buffer pointer
+ * @inlen: input buffer size
+ * @out: output buffer pointer
+ * @outlenp: output buffer size pointer
+ * @workmem: working memory buffer pointer, size determined by
+ *           nx842_powernv_driver.workmem_size
+ * @fc: function code, see CCW Function Codes in nx-842.h
+ *
+ * Returns:
+ *   0		Success, output of length @outlenp stored in the buffer at @out
+ *   -ENODEV	Hardware unavailable
+ *   -ENOSPC	Output buffer is to small
+ *   -EMSGSIZE	Input buffer too large
+ *   -EINVAL	buffer constraints do not fix nx842_constraints
+ *   -EPROTO	hardware error during operation
+ *   -ETIMEDOUT	hardware did not complete operation in reasonable time
+ *   -EINTR	operation was aborted
+ */
+static int nx842_powernv_function(const unsigned char *in, unsigned int inlen,
+				  unsigned char *out, unsigned int *outlenp,
+				  void *workmem, int fc)
+{
+	struct coprocessor_request_block *crb;
+	struct coprocessor_status_block *csb;
+	struct nx842_workmem *wmem;
+	int ret;
+	u64 csb_addr;
+	u32 ccw;
+	unsigned int outlen = *outlenp;
+
+	wmem = PTR_ALIGN(workmem, WORKMEM_ALIGN);
+
+	*outlenp = 0;
+
+	/* shoudn't happen, we don't load without a coproc */
+	if (!nx842_ct) {
+		pr_err_ratelimited("coprocessor CT is 0");
+		return -ENODEV;
+	}
+
+	crb = &wmem->crb;
+	csb = &crb->csb;
+
+	/* Clear any previous values */
+	memset(crb, 0, sizeof(*crb));
+
+	/* set up DDLs */
+	ret = setup_ddl(&crb->source, wmem->ddl_in,
+			(unsigned char *)in, inlen, true);
+	if (ret)
+		return ret;
+	ret = setup_ddl(&crb->target, wmem->ddl_out,
+			out, outlen, false);
+	if (ret)
+		return ret;
+
+	/* set up CCW */
+	ccw = 0;
+	ccw = SET_FIELD(ccw, CCW_CT, nx842_ct);
+	ccw = SET_FIELD(ccw, CCW_CI_842, 0); /* use 0 for hw auto-selection */
+	ccw = SET_FIELD(ccw, CCW_FC_842, fc);
+
+	/* set up CRB's CSB addr */
+	csb_addr = nx842_get_pa(csb) & CRB_CSB_ADDRESS;
+	csb_addr |= CRB_CSB_AT; /* Addrs are phys */
+	crb->csb_addr = cpu_to_be64(csb_addr);
+
+	wmem->start = ktime_get();
+
+	/* do ICSWX */
+	ret = icswx(cpu_to_be32(ccw), crb);
+
+	pr_debug_ratelimited("icswx CR %x ccw %x crb->ccw %x\n", ret,
+			     (unsigned int)ccw,
+			     (unsigned int)be32_to_cpu(crb->ccw));
+
+	switch (ret) {
+	case ICSWX_INITIATED:
+		ret = wait_for_csb(wmem, csb);
+		break;
+	case ICSWX_BUSY:
+		pr_debug_ratelimited("842 Coprocessor busy\n");
+		ret = -EBUSY;
+		break;
+	case ICSWX_REJECTED:
+		pr_err_ratelimited("ICSWX rejected\n");
+		ret = -EPROTO;
+		break;
+	default:
+		pr_err_ratelimited("Invalid ICSWX return code %x\n", ret);
+		ret = -EPROTO;
+		break;
+	}
+
+	if (!ret)
+		*outlenp = be32_to_cpu(csb->count);
+
+	return ret;
+}
+
+/**
+ * nx842_powernv_compress - Compress data using the 842 algorithm
+ *
+ * Compression provided by the NX842 coprocessor on IBM PowerNV systems.
+ * The input buffer is compressed and the result is stored in the
+ * provided output buffer.
+ *
+ * Upon return from this function @outlen contains the length of the
+ * compressed data.  If there is an error then @outlen will be 0 and an
+ * error will be specified by the return code from this function.
+ *
+ * @in: input buffer pointer
+ * @inlen: input buffer size
+ * @out: output buffer pointer
+ * @outlenp: output buffer size pointer
+ * @workmem: working memory buffer pointer, size determined by
+ *           nx842_powernv_driver.workmem_size
+ *
+ * Returns: see @nx842_powernv_function()
+ */
+static int nx842_powernv_compress(const unsigned char *in, unsigned int inlen,
+				  unsigned char *out, unsigned int *outlenp,
+				  void *wmem)
+{
+	return nx842_powernv_function(in, inlen, out, outlenp,
+				      wmem, CCW_FC_842_COMP_NOCRC);
+}
+
+/**
+ * nx842_powernv_decompress - Decompress data using the 842 algorithm
+ *
+ * Decompression provided by the NX842 coprocessor on IBM PowerNV systems.
+ * The input buffer is decompressed and the result is stored in the
+ * provided output buffer.
+ *
+ * Upon return from this function @outlen contains the length of the
+ * decompressed data.  If there is an error then @outlen will be 0 and an
+ * error will be specified by the return code from this function.
+ *
+ * @in: input buffer pointer
+ * @inlen: input buffer size
+ * @out: output buffer pointer
+ * @outlenp: output buffer size pointer
+ * @workmem: working memory buffer pointer, size determined by
+ *           nx842_powernv_driver.workmem_size
+ *
+ * Returns: see @nx842_powernv_function()
+ */
+static int nx842_powernv_decompress(const unsigned char *in, unsigned int inlen,
+				    unsigned char *out, unsigned int *outlenp,
+				    void *wmem)
+{
+	return nx842_powernv_function(in, inlen, out, outlenp,
+				      wmem, CCW_FC_842_DECOMP_NOCRC);
+}
+
+static int __init nx842_powernv_probe(struct device_node *dn)
+{
+	struct nx842_coproc *coproc;
+	struct property *ct_prop, *ci_prop;
+	unsigned int ct, ci;
+	int chip_id;
+
+	chip_id = of_get_ibm_chip_id(dn);
+	if (chip_id < 0) {
+		pr_err("ibm,chip-id missing\n");
+		return -EINVAL;
+	}
+	ct_prop = of_find_property(dn, "ibm,842-coprocessor-type", NULL);
+	if (!ct_prop) {
+		pr_err("ibm,842-coprocessor-type missing\n");
+		return -EINVAL;
+	}
+	ct = be32_to_cpu(*(unsigned int *)ct_prop->value);
+	ci_prop = of_find_property(dn, "ibm,842-coprocessor-instance", NULL);
+	if (!ci_prop) {
+		pr_err("ibm,842-coprocessor-instance missing\n");
+		return -EINVAL;
+	}
+	ci = be32_to_cpu(*(unsigned int *)ci_prop->value);
+
+	coproc = kmalloc(sizeof(*coproc), GFP_KERNEL);
+	if (!coproc)
+		return -ENOMEM;
+
+	coproc->chip_id = chip_id;
+	coproc->ct = ct;
+	coproc->ci = ci;
+	INIT_LIST_HEAD(&coproc->list);
+	list_add(&coproc->list, &nx842_coprocs);
+
+	pr_info("coprocessor found on chip %d, CT %d CI %d\n", chip_id, ct, ci);
+
+	if (!nx842_ct)
+		nx842_ct = ct;
+	else if (nx842_ct != ct)
+		pr_err("NX842 chip %d, CT %d != first found CT %d\n",
+		       chip_id, ct, nx842_ct);
+
+	return 0;
+}
+
+static struct nx842_constraints nx842_powernv_constraints = {
+	.alignment =	DDE_BUFFER_ALIGN,
+	.multiple =	DDE_BUFFER_LAST_MULT,
+	.minimum =	DDE_BUFFER_LAST_MULT,
+	.maximum =	(DDL_LEN_MAX - 1) * PAGE_SIZE,
+};
+
+static struct nx842_driver nx842_powernv_driver = {
+	.name =		KBUILD_MODNAME,
+	.owner =	THIS_MODULE,
+	.workmem_size =	sizeof(struct nx842_workmem),
+	.constraints =	&nx842_powernv_constraints,
+	.compress =	nx842_powernv_compress,
+	.decompress =	nx842_powernv_decompress,
+};
+
+static __init int nx842_powernv_init(void)
+{
+	struct device_node *dn;
+
+	/* verify workmem size/align restrictions */
+	BUILD_BUG_ON(WORKMEM_ALIGN % CRB_ALIGN);
+	BUILD_BUG_ON(CRB_ALIGN % DDE_ALIGN);
+	BUILD_BUG_ON(CRB_SIZE % DDE_ALIGN);
+	/* verify buffer size/align restrictions */
+	BUILD_BUG_ON(PAGE_SIZE % DDE_BUFFER_ALIGN);
+	BUILD_BUG_ON(DDE_BUFFER_ALIGN % DDE_BUFFER_SIZE_MULT);
+	BUILD_BUG_ON(DDE_BUFFER_SIZE_MULT % DDE_BUFFER_LAST_MULT);
+
+	pr_info("loading\n");
+
+	for_each_compatible_node(dn, NULL, "ibm,power-nx")
+		nx842_powernv_probe(dn);
+
+	if (!nx842_ct) {
+		pr_err("no coprocessors found\n");
+		return -ENODEV;
+	}
+
+	if (!nx842_platform_driver_set(&nx842_powernv_driver)) {
+		struct nx842_coproc *coproc, *n;
+
+		list_for_each_entry_safe(coproc, n, &nx842_coprocs, list) {
+			list_del(&coproc->list);
+			kfree(coproc);
+		}
+
+		return -EEXIST;
+	}
+
+	pr_info("loaded\n");
+
+	return 0;
+}
+module_init(nx842_powernv_init);
+
+static void __exit nx842_powernv_exit(void)
+{
+	struct nx842_coproc *coproc, *n;
+
+	nx842_platform_driver_unset(&nx842_powernv_driver);
+
+	list_for_each_entry_safe(coproc, n, &nx842_coprocs, list) {
+		list_del(&coproc->list);
+		kfree(coproc);
+	}
+
+	pr_info("unloaded\n");
+}
+module_exit(nx842_powernv_exit);
diff --git a/drivers/crypto/nx/nx-842-pseries.c b/drivers/crypto/nx/nx-842-pseries.c
new file mode 100644
index 000000000..3040a6091
--- /dev/null
+++ b/drivers/crypto/nx/nx-842-pseries.c
@@ -0,0 +1,1140 @@
+/*
+ * Driver for IBM Power 842 compression accelerator
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
+ *
+ * Copyright (C) IBM Corporation, 2012
+ *
+ * Authors: Robert Jennings <rcj@linux.vnet.ibm.com>
+ *          Seth Jennings <sjenning@linux.vnet.ibm.com>
+ */
+
+#include <asm/vio.h>
+
+#include "nx-842.h"
+#include "nx_csbcpb.h" /* struct nx_csbcpb */
+
+MODULE_LICENSE("GPL");
+MODULE_AUTHOR("Robert Jennings <rcj@linux.vnet.ibm.com>");
+MODULE_DESCRIPTION("842 H/W Compression driver for IBM Power processors");
+
+static struct nx842_constraints nx842_pseries_constraints = {
+	.alignment =	DDE_BUFFER_ALIGN,
+	.multiple =	DDE_BUFFER_LAST_MULT,
+	.minimum =	DDE_BUFFER_LAST_MULT,
+	.maximum =	PAGE_SIZE, /* dynamic, max_sync_size */
+};
+
+static int check_constraints(unsigned long buf, unsigned int *len, bool in)
+{
+	if (!IS_ALIGNED(buf, nx842_pseries_constraints.alignment)) {
+		pr_debug("%s buffer 0x%lx not aligned to 0x%x\n",
+			 in ? "input" : "output", buf,
+			 nx842_pseries_constraints.alignment);
+		return -EINVAL;
+	}
+	if (*len % nx842_pseries_constraints.multiple) {
+		pr_debug("%s buffer len 0x%x not multiple of 0x%x\n",
+			 in ? "input" : "output", *len,
+			 nx842_pseries_constraints.multiple);
+		if (in)
+			return -EINVAL;
+		*len = round_down(*len, nx842_pseries_constraints.multiple);
+	}
+	if (*len < nx842_pseries_constraints.minimum) {
+		pr_debug("%s buffer len 0x%x under minimum 0x%x\n",
+			 in ? "input" : "output", *len,
+			 nx842_pseries_constraints.minimum);
+		return -EINVAL;
+	}
+	if (*len > nx842_pseries_constraints.maximum) {
+		pr_debug("%s buffer len 0x%x over maximum 0x%x\n",
+			 in ? "input" : "output", *len,
+			 nx842_pseries_constraints.maximum);
+		if (in)
+			return -EINVAL;
+		*len = nx842_pseries_constraints.maximum;
+	}
+	return 0;
+}
+
+/* I assume we need to align the CSB? */
+#define WORKMEM_ALIGN	(256)
+
+struct nx842_workmem {
+	/* scatterlist */
+	char slin[4096];
+	char slout[4096];
+	/* coprocessor status/parameter block */
+	struct nx_csbcpb csbcpb;
+
+	char padding[WORKMEM_ALIGN];
+} __aligned(WORKMEM_ALIGN);
+
+/* Macros for fields within nx_csbcpb */
+/* Check the valid bit within the csbcpb valid field */
+#define NX842_CSBCBP_VALID_CHK(x) (x & BIT_MASK(7))
+
+/* CE macros operate on the completion_extension field bits in the csbcpb.
+ * CE0 0=full completion, 1=partial completion
+ * CE1 0=CE0 indicates completion, 1=termination (output may be modified)
+ * CE2 0=processed_bytes is source bytes, 1=processed_bytes is target bytes */
+#define NX842_CSBCPB_CE0(x)	(x & BIT_MASK(7))
+#define NX842_CSBCPB_CE1(x)	(x & BIT_MASK(6))
+#define NX842_CSBCPB_CE2(x)	(x & BIT_MASK(5))
+
+/* The NX unit accepts data only on 4K page boundaries */
+#define NX842_HW_PAGE_SIZE	(4096)
+#define NX842_HW_PAGE_MASK	(~(NX842_HW_PAGE_SIZE-1))
+
+enum nx842_status {
+	UNAVAILABLE,
+	AVAILABLE
+};
+
+struct ibm_nx842_counters {
+	atomic64_t comp_complete;
+	atomic64_t comp_failed;
+	atomic64_t decomp_complete;
+	atomic64_t decomp_failed;
+	atomic64_t swdecomp;
+	atomic64_t comp_times[32];
+	atomic64_t decomp_times[32];
+};
+
+static struct nx842_devdata {
+	struct vio_dev *vdev;
+	struct device *dev;
+	struct ibm_nx842_counters *counters;
+	unsigned int max_sg_len;
+	unsigned int max_sync_size;
+	unsigned int max_sync_sg;
+	enum nx842_status status;
+} __rcu *devdata;
+static DEFINE_SPINLOCK(devdata_mutex);
+
+#define NX842_COUNTER_INC(_x) \
+static inline void nx842_inc_##_x( \
+	const struct nx842_devdata *dev) { \
+	if (dev) \
+		atomic64_inc(&dev->counters->_x); \
+}
+NX842_COUNTER_INC(comp_complete);
+NX842_COUNTER_INC(comp_failed);
+NX842_COUNTER_INC(decomp_complete);
+NX842_COUNTER_INC(decomp_failed);
+NX842_COUNTER_INC(swdecomp);
+
+#define NX842_HIST_SLOTS 16
+
+static void ibm_nx842_incr_hist(atomic64_t *times, unsigned int time)
+{
+	int bucket = fls(time);
+
+	if (bucket)
+		bucket = min((NX842_HIST_SLOTS - 1), bucket - 1);
+
+	atomic64_inc(&times[bucket]);
+}
+
+/* NX unit operation flags */
+#define NX842_OP_COMPRESS	0x0
+#define NX842_OP_CRC		0x1
+#define NX842_OP_DECOMPRESS	0x2
+#define NX842_OP_COMPRESS_CRC   (NX842_OP_COMPRESS | NX842_OP_CRC)
+#define NX842_OP_DECOMPRESS_CRC (NX842_OP_DECOMPRESS | NX842_OP_CRC)
+#define NX842_OP_ASYNC		(1<<23)
+#define NX842_OP_NOTIFY		(1<<22)
+#define NX842_OP_NOTIFY_INT(x)	((x & 0xff)<<8)
+
+static unsigned long nx842_get_desired_dma(struct vio_dev *viodev)
+{
+	/* No use of DMA mappings within the driver. */
+	return 0;
+}
+
+struct nx842_slentry {
+	__be64 ptr; /* Real address (use __pa()) */
+	__be64 len;
+};
+
+/* pHyp scatterlist entry */
+struct nx842_scatterlist {
+	int entry_nr; /* number of slentries */
+	struct nx842_slentry *entries; /* ptr to array of slentries */
+};
+
+/* Does not include sizeof(entry_nr) in the size */
+static inline unsigned long nx842_get_scatterlist_size(
+				struct nx842_scatterlist *sl)
+{
+	return sl->entry_nr * sizeof(struct nx842_slentry);
+}
+
+static int nx842_build_scatterlist(unsigned long buf, int len,
+			struct nx842_scatterlist *sl)
+{
+	unsigned long entrylen;
+	struct nx842_slentry *entry;
+
+	sl->entry_nr = 0;
+
+	entry = sl->entries;
+	while (len) {
+		entry->ptr = cpu_to_be64(nx842_get_pa((void *)buf));
+		entrylen = min_t(int, len,
+				 LEN_ON_SIZE(buf, NX842_HW_PAGE_SIZE));
+		entry->len = cpu_to_be64(entrylen);
+
+		len -= entrylen;
+		buf += entrylen;
+
+		sl->entry_nr++;
+		entry++;
+	}
+
+	return 0;
+}
+
+static int nx842_validate_result(struct device *dev,
+	struct cop_status_block *csb)
+{
+	/* The csb must be valid after returning from vio_h_cop_sync */
+	if (!NX842_CSBCBP_VALID_CHK(csb->valid)) {
+		dev_err(dev, "%s: cspcbp not valid upon completion.\n",
+				__func__);
+		dev_dbg(dev, "valid:0x%02x cs:0x%02x cc:0x%02x ce:0x%02x\n",
+				csb->valid,
+				csb->crb_seq_number,
+				csb->completion_code,
+				csb->completion_extension);
+		dev_dbg(dev, "processed_bytes:%d address:0x%016lx\n",
+				be32_to_cpu(csb->processed_byte_count),
+				(unsigned long)be64_to_cpu(csb->address));
+		return -EIO;
+	}
+
+	/* Check return values from the hardware in the CSB */
+	switch (csb->completion_code) {
+	case 0:	/* Completed without error */
+		break;
+	case 64: /* Target bytes > Source bytes during compression */
+	case 13: /* Output buffer too small */
+		dev_dbg(dev, "%s: Compression output larger than input\n",
+					__func__);
+		return -ENOSPC;
+	case 66: /* Input data contains an illegal template field */
+	case 67: /* Template indicates data past the end of the input stream */
+		dev_dbg(dev, "%s: Bad data for decompression (code:%d)\n",
+					__func__, csb->completion_code);
+		return -EINVAL;
+	default:
+		dev_dbg(dev, "%s: Unspecified error (code:%d)\n",
+					__func__, csb->completion_code);
+		return -EIO;
+	}
+
+	/* Hardware sanity check */
+	if (!NX842_CSBCPB_CE2(csb->completion_extension)) {
+		dev_err(dev, "%s: No error returned by hardware, but "
+				"data returned is unusable, contact support.\n"
+				"(Additional info: csbcbp->processed bytes "
+				"does not specify processed bytes for the "
+				"target buffer.)\n", __func__);
+		return -EIO;
+	}
+
+	return 0;
+}
+
+/**
+ * nx842_pseries_compress - Compress data using the 842 algorithm
+ *
+ * Compression provide by the NX842 coprocessor on IBM Power systems.
+ * The input buffer is compressed and the result is stored in the
+ * provided output buffer.
+ *
+ * Upon return from this function @outlen contains the length of the
+ * compressed data.  If there is an error then @outlen will be 0 and an
+ * error will be specified by the return code from this function.
+ *
+ * @in: Pointer to input buffer
+ * @inlen: Length of input buffer
+ * @out: Pointer to output buffer
+ * @outlen: Length of output buffer
+ * @wrkmem: ptr to buffer for working memory, size determined by
+ *          nx842_pseries_driver.workmem_size
+ *
+ * Returns:
+ *   0		Success, output of length @outlen stored in the buffer at @out
+ *   -ENOMEM	Unable to allocate internal buffers
+ *   -ENOSPC	Output buffer is to small
+ *   -EIO	Internal error
+ *   -ENODEV	Hardware unavailable
+ */
+static int nx842_pseries_compress(const unsigned char *in, unsigned int inlen,
+				  unsigned char *out, unsigned int *outlen,
+				  void *wmem)
+{
+	struct nx842_devdata *local_devdata;
+	struct device *dev = NULL;
+	struct nx842_workmem *workmem;
+	struct nx842_scatterlist slin, slout;
+	struct nx_csbcpb *csbcpb;
+	int ret = 0, max_sync_size;
+	unsigned long inbuf, outbuf;
+	struct vio_pfo_op op = {
+		.done = NULL,
+		.handle = 0,
+		.timeout = 0,
+	};
+	unsigned long start = get_tb();
+
+	inbuf = (unsigned long)in;
+	if (check_constraints(inbuf, &inlen, true))
+		return -EINVAL;
+
+	outbuf = (unsigned long)out;
+	if (check_constraints(outbuf, outlen, false))
+		return -EINVAL;
+
+	rcu_read_lock();
+	local_devdata = rcu_dereference(devdata);
+	if (!local_devdata || !local_devdata->dev) {
+		rcu_read_unlock();
+		return -ENODEV;
+	}
+	max_sync_size = local_devdata->max_sync_size;
+	dev = local_devdata->dev;
+
+	/* Init scatterlist */
+	workmem = PTR_ALIGN(wmem, WORKMEM_ALIGN);
+	slin.entries = (struct nx842_slentry *)workmem->slin;
+	slout.entries = (struct nx842_slentry *)workmem->slout;
+
+	/* Init operation */
+	op.flags = NX842_OP_COMPRESS;
+	csbcpb = &workmem->csbcpb;
+	memset(csbcpb, 0, sizeof(*csbcpb));
+	op.csbcpb = nx842_get_pa(csbcpb);
+
+	if ((inbuf & NX842_HW_PAGE_MASK) ==
+	    ((inbuf + inlen - 1) & NX842_HW_PAGE_MASK)) {
+		/* Create direct DDE */
+		op.in = nx842_get_pa((void *)inbuf);
+		op.inlen = inlen;
+	} else {
+		/* Create indirect DDE (scatterlist) */
+		nx842_build_scatterlist(inbuf, inlen, &slin);
+		op.in = nx842_get_pa(slin.entries);
+		op.inlen = -nx842_get_scatterlist_size(&slin);
+	}
+
+	if ((outbuf & NX842_HW_PAGE_MASK) ==
+	    ((outbuf + *outlen - 1) & NX842_HW_PAGE_MASK)) {
+		/* Create direct DDE */
+		op.out = nx842_get_pa((void *)outbuf);
+		op.outlen = *outlen;
+	} else {
+		/* Create indirect DDE (scatterlist) */
+		nx842_build_scatterlist(outbuf, *outlen, &slout);
+		op.out = nx842_get_pa(slout.entries);
+		op.outlen = -nx842_get_scatterlist_size(&slout);
+	}
+
+	dev_dbg(dev, "%s: op.in %lx op.inlen %ld op.out %lx op.outlen %ld\n",
+		__func__, (unsigned long)op.in, (long)op.inlen,
+		(unsigned long)op.out, (long)op.outlen);
+
+	/* Send request to pHyp */
+	ret = vio_h_cop_sync(local_devdata->vdev, &op);
+
+	/* Check for pHyp error */
+	if (ret) {
+		dev_dbg(dev, "%s: vio_h_cop_sync error (ret=%d, hret=%ld)\n",
+			__func__, ret, op.hcall_err);
+		ret = -EIO;
+		goto unlock;
+	}
+
+	/* Check for hardware error */
+	ret = nx842_validate_result(dev, &csbcpb->csb);
+	if (ret)
+		goto unlock;
+
+	*outlen = be32_to_cpu(csbcpb->csb.processed_byte_count);
+	dev_dbg(dev, "%s: processed_bytes=%d\n", __func__, *outlen);
+
+unlock:
+	if (ret)
+		nx842_inc_comp_failed(local_devdata);
+	else {
+		nx842_inc_comp_complete(local_devdata);
+		ibm_nx842_incr_hist(local_devdata->counters->comp_times,
+			(get_tb() - start) / tb_ticks_per_usec);
+	}
+	rcu_read_unlock();
+	return ret;
+}
+
+/**
+ * nx842_pseries_decompress - Decompress data using the 842 algorithm
+ *
+ * Decompression provide by the NX842 coprocessor on IBM Power systems.
+ * The input buffer is decompressed and the result is stored in the
+ * provided output buffer.  The size allocated to the output buffer is
+ * provided by the caller of this function in @outlen.  Upon return from
+ * this function @outlen contains the length of the decompressed data.
+ * If there is an error then @outlen will be 0 and an error will be
+ * specified by the return code from this function.
+ *
+ * @in: Pointer to input buffer
+ * @inlen: Length of input buffer
+ * @out: Pointer to output buffer
+ * @outlen: Length of output buffer
+ * @wrkmem: ptr to buffer for working memory, size determined by
+ *          nx842_pseries_driver.workmem_size
+ *
+ * Returns:
+ *   0		Success, output of length @outlen stored in the buffer at @out
+ *   -ENODEV	Hardware decompression device is unavailable
+ *   -ENOMEM	Unable to allocate internal buffers
+ *   -ENOSPC	Output buffer is to small
+ *   -EINVAL	Bad input data encountered when attempting decompress
+ *   -EIO	Internal error
+ */
+static int nx842_pseries_decompress(const unsigned char *in, unsigned int inlen,
+				    unsigned char *out, unsigned int *outlen,
+				    void *wmem)
+{
+	struct nx842_devdata *local_devdata;
+	struct device *dev = NULL;
+	struct nx842_workmem *workmem;
+	struct nx842_scatterlist slin, slout;
+	struct nx_csbcpb *csbcpb;
+	int ret = 0, max_sync_size;
+	unsigned long inbuf, outbuf;
+	struct vio_pfo_op op = {
+		.done = NULL,
+		.handle = 0,
+		.timeout = 0,
+	};
+	unsigned long start = get_tb();
+
+	/* Ensure page alignment and size */
+	inbuf = (unsigned long)in;
+	if (check_constraints(inbuf, &inlen, true))
+		return -EINVAL;
+
+	outbuf = (unsigned long)out;
+	if (check_constraints(outbuf, outlen, false))
+		return -EINVAL;
+
+	rcu_read_lock();
+	local_devdata = rcu_dereference(devdata);
+	if (!local_devdata || !local_devdata->dev) {
+		rcu_read_unlock();
+		return -ENODEV;
+	}
+	max_sync_size = local_devdata->max_sync_size;
+	dev = local_devdata->dev;
+
+	workmem = PTR_ALIGN(wmem, WORKMEM_ALIGN);
+
+	/* Init scatterlist */
+	slin.entries = (struct nx842_slentry *)workmem->slin;
+	slout.entries = (struct nx842_slentry *)workmem->slout;
+
+	/* Init operation */
+	op.flags = NX842_OP_DECOMPRESS;
+	csbcpb = &workmem->csbcpb;
+	memset(csbcpb, 0, sizeof(*csbcpb));
+	op.csbcpb = nx842_get_pa(csbcpb);
+
+	if ((inbuf & NX842_HW_PAGE_MASK) ==
+	    ((inbuf + inlen - 1) & NX842_HW_PAGE_MASK)) {
+		/* Create direct DDE */
+		op.in = nx842_get_pa((void *)inbuf);
+		op.inlen = inlen;
+	} else {
+		/* Create indirect DDE (scatterlist) */
+		nx842_build_scatterlist(inbuf, inlen, &slin);
+		op.in = nx842_get_pa(slin.entries);
+		op.inlen = -nx842_get_scatterlist_size(&slin);
+	}
+
+	if ((outbuf & NX842_HW_PAGE_MASK) ==
+	    ((outbuf + *outlen - 1) & NX842_HW_PAGE_MASK)) {
+		/* Create direct DDE */
+		op.out = nx842_get_pa((void *)outbuf);
+		op.outlen = *outlen;
+	} else {
+		/* Create indirect DDE (scatterlist) */
+		nx842_build_scatterlist(outbuf, *outlen, &slout);
+		op.out = nx842_get_pa(slout.entries);
+		op.outlen = -nx842_get_scatterlist_size(&slout);
+	}
+
+	dev_dbg(dev, "%s: op.in %lx op.inlen %ld op.out %lx op.outlen %ld\n",
+		__func__, (unsigned long)op.in, (long)op.inlen,
+		(unsigned long)op.out, (long)op.outlen);
+
+	/* Send request to pHyp */
+	ret = vio_h_cop_sync(local_devdata->vdev, &op);
+
+	/* Check for pHyp error */
+	if (ret) {
+		dev_dbg(dev, "%s: vio_h_cop_sync error (ret=%d, hret=%ld)\n",
+			__func__, ret, op.hcall_err);
+		goto unlock;
+	}
+
+	/* Check for hardware error */
+	ret = nx842_validate_result(dev, &csbcpb->csb);
+	if (ret)
+		goto unlock;
+
+	*outlen = be32_to_cpu(csbcpb->csb.processed_byte_count);
+
+unlock:
+	if (ret)
+		/* decompress fail */
+		nx842_inc_decomp_failed(local_devdata);
+	else {
+		nx842_inc_decomp_complete(local_devdata);
+		ibm_nx842_incr_hist(local_devdata->counters->decomp_times,
+			(get_tb() - start) / tb_ticks_per_usec);
+	}
+
+	rcu_read_unlock();
+	return ret;
+}
+
+/**
+ * nx842_OF_set_defaults -- Set default (disabled) values for devdata
+ *
+ * @devdata - struct nx842_devdata to update
+ *
+ * Returns:
+ *  0 on success
+ *  -ENOENT if @devdata ptr is NULL
+ */
+static int nx842_OF_set_defaults(struct nx842_devdata *devdata)
+{
+	if (devdata) {
+		devdata->max_sync_size = 0;
+		devdata->max_sync_sg = 0;
+		devdata->max_sg_len = 0;
+		devdata->status = UNAVAILABLE;
+		return 0;
+	} else
+		return -ENOENT;
+}
+
+/**
+ * nx842_OF_upd_status -- Update the device info from OF status prop
+ *
+ * The status property indicates if the accelerator is enabled.  If the
+ * device is in the OF tree it indicates that the hardware is present.
+ * The status field indicates if the device is enabled when the status
+ * is 'okay'.  Otherwise the device driver will be disabled.
+ *
+ * @devdata - struct nx842_devdata to update
+ * @prop - struct property point containing the maxsyncop for the update
+ *
+ * Returns:
+ *  0 - Device is available
+ *  -EINVAL - Device is not available
+ */
+static int nx842_OF_upd_status(struct nx842_devdata *devdata,
+					struct property *prop) {
+	int ret = 0;
+	const char *status = (const char *)prop->value;
+
+	if (!strncmp(status, "okay", (size_t)prop->length)) {
+		devdata->status = AVAILABLE;
+	} else {
+		dev_info(devdata->dev, "%s: status '%s' is not 'okay'\n",
+				__func__, status);
+		devdata->status = UNAVAILABLE;
+	}
+
+	return ret;
+}
+
+/**
+ * nx842_OF_upd_maxsglen -- Update the device info from OF maxsglen prop
+ *
+ * Definition of the 'ibm,max-sg-len' OF property:
+ *  This field indicates the maximum byte length of a scatter list
+ *  for the platform facility. It is a single cell encoded as with encode-int.
+ *
+ * Example:
+ *  # od -x ibm,max-sg-len
+ *  0000000 0000 0ff0
+ *
+ *  In this example, the maximum byte length of a scatter list is
+ *  0x0ff0 (4,080).
+ *
+ * @devdata - struct nx842_devdata to update
+ * @prop - struct property point containing the maxsyncop for the update
+ *
+ * Returns:
+ *  0 on success
+ *  -EINVAL on failure
+ */
+static int nx842_OF_upd_maxsglen(struct nx842_devdata *devdata,
+					struct property *prop) {
+	int ret = 0;
+	const unsigned int maxsglen = of_read_number(prop->value, 1);
+
+	if (prop->length != sizeof(maxsglen)) {
+		dev_err(devdata->dev, "%s: unexpected format for ibm,max-sg-len property\n", __func__);
+		dev_dbg(devdata->dev, "%s: ibm,max-sg-len is %d bytes long, expected %lu bytes\n", __func__,
+				prop->length, sizeof(maxsglen));
+		ret = -EINVAL;
+	} else {
+		devdata->max_sg_len = min_t(unsigned int,
+					    maxsglen, NX842_HW_PAGE_SIZE);
+	}
+
+	return ret;
+}
+
+/**
+ * nx842_OF_upd_maxsyncop -- Update the device info from OF maxsyncop prop
+ *
+ * Definition of the 'ibm,max-sync-cop' OF property:
+ *  Two series of cells.  The first series of cells represents the maximums
+ *  that can be synchronously compressed. The second series of cells
+ *  represents the maximums that can be synchronously decompressed.
+ *  1. The first cell in each series contains the count of the number of
+ *     data length, scatter list elements pairs that follow – each being
+ *     of the form
+ *    a. One cell data byte length
+ *    b. One cell total number of scatter list elements
+ *
+ * Example:
+ *  # od -x ibm,max-sync-cop
+ *  0000000 0000 0001 0000 1000 0000 01fe 0000 0001
+ *  0000020 0000 1000 0000 01fe
+ *
+ *  In this example, compression supports 0x1000 (4,096) data byte length
+ *  and 0x1fe (510) total scatter list elements.  Decompression supports
+ *  0x1000 (4,096) data byte length and 0x1f3 (510) total scatter list
+ *  elements.
+ *
+ * @devdata - struct nx842_devdata to update
+ * @prop - struct property point containing the maxsyncop for the update
+ *
+ * Returns:
+ *  0 on success
+ *  -EINVAL on failure
+ */
+static int nx842_OF_upd_maxsyncop(struct nx842_devdata *devdata,
+					struct property *prop) {
+	int ret = 0;
+	unsigned int comp_data_limit, decomp_data_limit;
+	unsigned int comp_sg_limit, decomp_sg_limit;
+	const struct maxsynccop_t {
+		__be32 comp_elements;
+		__be32 comp_data_limit;
+		__be32 comp_sg_limit;
+		__be32 decomp_elements;
+		__be32 decomp_data_limit;
+		__be32 decomp_sg_limit;
+	} *maxsynccop;
+
+	if (prop->length != sizeof(*maxsynccop)) {
+		dev_err(devdata->dev, "%s: unexpected format for ibm,max-sync-cop property\n", __func__);
+		dev_dbg(devdata->dev, "%s: ibm,max-sync-cop is %d bytes long, expected %lu bytes\n", __func__, prop->length,
+				sizeof(*maxsynccop));
+		ret = -EINVAL;
+		goto out;
+	}
+
+	maxsynccop = (const struct maxsynccop_t *)prop->value;
+	comp_data_limit = be32_to_cpu(maxsynccop->comp_data_limit);
+	comp_sg_limit = be32_to_cpu(maxsynccop->comp_sg_limit);
+	decomp_data_limit = be32_to_cpu(maxsynccop->decomp_data_limit);
+	decomp_sg_limit = be32_to_cpu(maxsynccop->decomp_sg_limit);
+
+	/* Use one limit rather than separate limits for compression and
+	 * decompression. Set a maximum for this so as not to exceed the
+	 * size that the header can support and round the value down to
+	 * the hardware page size (4K) */
+	devdata->max_sync_size = min(comp_data_limit, decomp_data_limit);
+
+	devdata->max_sync_size = min_t(unsigned int, devdata->max_sync_size,
+					65536);
+
+	if (devdata->max_sync_size < 4096) {
+		dev_err(devdata->dev, "%s: hardware max data size (%u) is "
+				"less than the driver minimum, unable to use "
+				"the hardware device\n",
+				__func__, devdata->max_sync_size);
+		ret = -EINVAL;
+		goto out;
+	}
+
+	nx842_pseries_constraints.maximum = devdata->max_sync_size;
+
+	devdata->max_sync_sg = min(comp_sg_limit, decomp_sg_limit);
+	if (devdata->max_sync_sg < 1) {
+		dev_err(devdata->dev, "%s: hardware max sg size (%u) is "
+				"less than the driver minimum, unable to use "
+				"the hardware device\n",
+				__func__, devdata->max_sync_sg);
+		ret = -EINVAL;
+		goto out;
+	}
+
+out:
+	return ret;
+}
+
+/**
+ *
+ * nx842_OF_upd -- Handle OF properties updates for the device.
+ *
+ * Set all properties from the OF tree.  Optionally, a new property
+ * can be provided by the @new_prop pointer to overwrite an existing value.
+ * The device will remain disabled until all values are valid, this function
+ * will return an error for updates unless all values are valid.
+ *
+ * @new_prop: If not NULL, this property is being updated.  If NULL, update
+ *  all properties from the current values in the OF tree.
+ *
+ * Returns:
+ *  0 - Success
+ *  -ENOMEM - Could not allocate memory for new devdata structure
+ *  -EINVAL - property value not found, new_prop is not a recognized
+ *	property for the device or property value is not valid.
+ *  -ENODEV - Device is not available
+ */
+static int nx842_OF_upd(struct property *new_prop)
+{
+	struct nx842_devdata *old_devdata = NULL;
+	struct nx842_devdata *new_devdata = NULL;
+	struct device_node *of_node = NULL;
+	struct property *status = NULL;
+	struct property *maxsglen = NULL;
+	struct property *maxsyncop = NULL;
+	int ret = 0;
+	unsigned long flags;
+
+	spin_lock_irqsave(&devdata_mutex, flags);
+	old_devdata = rcu_dereference_check(devdata,
+			lockdep_is_held(&devdata_mutex));
+	if (old_devdata)
+		of_node = old_devdata->dev->of_node;
+
+	if (!old_devdata || !of_node) {
+		pr_err("%s: device is not available\n", __func__);
+		spin_unlock_irqrestore(&devdata_mutex, flags);
+		return -ENODEV;
+	}
+
+	new_devdata = kzalloc(sizeof(*new_devdata), GFP_NOFS);
+	if (!new_devdata) {
+		dev_err(old_devdata->dev, "%s: Could not allocate memory for device data\n", __func__);
+		ret = -ENOMEM;
+		goto error_out;
+	}
+
+	memcpy(new_devdata, old_devdata, sizeof(*old_devdata));
+	new_devdata->counters = old_devdata->counters;
+
+	/* Set ptrs for existing properties */
+	status = of_find_property(of_node, "status", NULL);
+	maxsglen = of_find_property(of_node, "ibm,max-sg-len", NULL);
+	maxsyncop = of_find_property(of_node, "ibm,max-sync-cop", NULL);
+	if (!status || !maxsglen || !maxsyncop) {
+		dev_err(old_devdata->dev, "%s: Could not locate device properties\n", __func__);
+		ret = -EINVAL;
+		goto error_out;
+	}
+
+	/*
+	 * If this is a property update, there are only certain properties that
+	 * we care about. Bail if it isn't in the below list
+	 */
+	if (new_prop && (strncmp(new_prop->name, "status", new_prop->length) ||
+		         strncmp(new_prop->name, "ibm,max-sg-len", new_prop->length) ||
+		         strncmp(new_prop->name, "ibm,max-sync-cop", new_prop->length)))
+		goto out;
+
+	/* Perform property updates */
+	ret = nx842_OF_upd_status(new_devdata, status);
+	if (ret)
+		goto error_out;
+
+	ret = nx842_OF_upd_maxsglen(new_devdata, maxsglen);
+	if (ret)
+		goto error_out;
+
+	ret = nx842_OF_upd_maxsyncop(new_devdata, maxsyncop);
+	if (ret)
+		goto error_out;
+
+out:
+	dev_info(old_devdata->dev, "%s: max_sync_size new:%u old:%u\n",
+			__func__, new_devdata->max_sync_size,
+			old_devdata->max_sync_size);
+	dev_info(old_devdata->dev, "%s: max_sync_sg new:%u old:%u\n",
+			__func__, new_devdata->max_sync_sg,
+			old_devdata->max_sync_sg);
+	dev_info(old_devdata->dev, "%s: max_sg_len new:%u old:%u\n",
+			__func__, new_devdata->max_sg_len,
+			old_devdata->max_sg_len);
+
+	rcu_assign_pointer(devdata, new_devdata);
+	spin_unlock_irqrestore(&devdata_mutex, flags);
+	synchronize_rcu();
+	dev_set_drvdata(new_devdata->dev, new_devdata);
+	kfree(old_devdata);
+	return 0;
+
+error_out:
+	if (new_devdata) {
+		dev_info(old_devdata->dev, "%s: device disabled\n", __func__);
+		nx842_OF_set_defaults(new_devdata);
+		rcu_assign_pointer(devdata, new_devdata);
+		spin_unlock_irqrestore(&devdata_mutex, flags);
+		synchronize_rcu();
+		dev_set_drvdata(new_devdata->dev, new_devdata);
+		kfree(old_devdata);
+	} else {
+		dev_err(old_devdata->dev, "%s: could not update driver from hardware\n", __func__);
+		spin_unlock_irqrestore(&devdata_mutex, flags);
+	}
+
+	if (!ret)
+		ret = -EINVAL;
+	return ret;
+}
+
+/**
+ * nx842_OF_notifier - Process updates to OF properties for the device
+ *
+ * @np: notifier block
+ * @action: notifier action
+ * @update: struct pSeries_reconfig_prop_update pointer if action is
+ *	PSERIES_UPDATE_PROPERTY
+ *
+ * Returns:
+ *	NOTIFY_OK on success
+ *	NOTIFY_BAD encoded with error number on failure, use
+ *		notifier_to_errno() to decode this value
+ */
+static int nx842_OF_notifier(struct notifier_block *np, unsigned long action,
+			     void *data)
+{
+	struct of_reconfig_data *upd = data;
+	struct nx842_devdata *local_devdata;
+	struct device_node *node = NULL;
+
+	rcu_read_lock();
+	local_devdata = rcu_dereference(devdata);
+	if (local_devdata)
+		node = local_devdata->dev->of_node;
+
+	if (local_devdata &&
+			action == OF_RECONFIG_UPDATE_PROPERTY &&
+			!strcmp(upd->dn->name, node->name)) {
+		rcu_read_unlock();
+		nx842_OF_upd(upd->prop);
+	} else
+		rcu_read_unlock();
+
+	return NOTIFY_OK;
+}
+
+static struct notifier_block nx842_of_nb = {
+	.notifier_call = nx842_OF_notifier,
+};
+
+#define nx842_counter_read(_name)					\
+static ssize_t nx842_##_name##_show(struct device *dev,		\
+		struct device_attribute *attr,				\
+		char *buf) {						\
+	struct nx842_devdata *local_devdata;			\
+	int p = 0;							\
+	rcu_read_lock();						\
+	local_devdata = rcu_dereference(devdata);			\
+	if (local_devdata)						\
+		p = snprintf(buf, PAGE_SIZE, "%ld\n",			\
+		       atomic64_read(&local_devdata->counters->_name));	\
+	rcu_read_unlock();						\
+	return p;							\
+}
+
+#define NX842DEV_COUNTER_ATTR_RO(_name)					\
+	nx842_counter_read(_name);					\
+	static struct device_attribute dev_attr_##_name = __ATTR(_name,	\
+						0444,			\
+						nx842_##_name##_show,\
+						NULL);
+
+NX842DEV_COUNTER_ATTR_RO(comp_complete);
+NX842DEV_COUNTER_ATTR_RO(comp_failed);
+NX842DEV_COUNTER_ATTR_RO(decomp_complete);
+NX842DEV_COUNTER_ATTR_RO(decomp_failed);
+NX842DEV_COUNTER_ATTR_RO(swdecomp);
+
+static ssize_t nx842_timehist_show(struct device *,
+		struct device_attribute *, char *);
+
+static struct device_attribute dev_attr_comp_times = __ATTR(comp_times, 0444,
+		nx842_timehist_show, NULL);
+static struct device_attribute dev_attr_decomp_times = __ATTR(decomp_times,
+		0444, nx842_timehist_show, NULL);
+
+static ssize_t nx842_timehist_show(struct device *dev,
+		struct device_attribute *attr, char *buf) {
+	char *p = buf;
+	struct nx842_devdata *local_devdata;
+	atomic64_t *times;
+	int bytes_remain = PAGE_SIZE;
+	int bytes;
+	int i;
+
+	rcu_read_lock();
+	local_devdata = rcu_dereference(devdata);
+	if (!local_devdata) {
+		rcu_read_unlock();
+		return 0;
+	}
+
+	if (attr == &dev_attr_comp_times)
+		times = local_devdata->counters->comp_times;
+	else if (attr == &dev_attr_decomp_times)
+		times = local_devdata->counters->decomp_times;
+	else {
+		rcu_read_unlock();
+		return 0;
+	}
+
+	for (i = 0; i < (NX842_HIST_SLOTS - 2); i++) {
+		bytes = snprintf(p, bytes_remain, "%u-%uus:\t%ld\n",
+			       i ? (2<<(i-1)) : 0, (2<<i)-1,
+			       atomic64_read(&times[i]));
+		bytes_remain -= bytes;
+		p += bytes;
+	}
+	/* The last bucket holds everything over
+	 * 2<<(NX842_HIST_SLOTS - 2) us */
+	bytes = snprintf(p, bytes_remain, "%uus - :\t%ld\n",
+			2<<(NX842_HIST_SLOTS - 2),
+			atomic64_read(&times[(NX842_HIST_SLOTS - 1)]));
+	p += bytes;
+
+	rcu_read_unlock();
+	return p - buf;
+}
+
+static struct attribute *nx842_sysfs_entries[] = {
+	&dev_attr_comp_complete.attr,
+	&dev_attr_comp_failed.attr,
+	&dev_attr_decomp_complete.attr,
+	&dev_attr_decomp_failed.attr,
+	&dev_attr_swdecomp.attr,
+	&dev_attr_comp_times.attr,
+	&dev_attr_decomp_times.attr,
+	NULL,
+};
+
+static struct attribute_group nx842_attribute_group = {
+	.name = NULL,		/* put in device directory */
+	.attrs = nx842_sysfs_entries,
+};
+
+static struct nx842_driver nx842_pseries_driver = {
+	.name =		KBUILD_MODNAME,
+	.owner =	THIS_MODULE,
+	.workmem_size =	sizeof(struct nx842_workmem),
+	.constraints =	&nx842_pseries_constraints,
+	.compress =	nx842_pseries_compress,
+	.decompress =	nx842_pseries_decompress,
+};
+
+static int __init nx842_probe(struct vio_dev *viodev,
+				  const struct vio_device_id *id)
+{
+	struct nx842_devdata *old_devdata, *new_devdata = NULL;
+	unsigned long flags;
+	int ret = 0;
+
+	spin_lock_irqsave(&devdata_mutex, flags);
+	old_devdata = rcu_dereference_check(devdata,
+			lockdep_is_held(&devdata_mutex));
+
+	if (old_devdata && old_devdata->vdev != NULL) {
+		dev_err(&viodev->dev, "%s: Attempt to register more than one instance of the hardware\n", __func__);
+		ret = -1;
+		goto error_unlock;
+	}
+
+	dev_set_drvdata(&viodev->dev, NULL);
+
+	new_devdata = kzalloc(sizeof(*new_devdata), GFP_NOFS);
+	if (!new_devdata) {
+		dev_err(&viodev->dev, "%s: Could not allocate memory for device data\n", __func__);
+		ret = -ENOMEM;
+		goto error_unlock;
+	}
+
+	new_devdata->counters = kzalloc(sizeof(*new_devdata->counters),
+			GFP_NOFS);
+	if (!new_devdata->counters) {
+		dev_err(&viodev->dev, "%s: Could not allocate memory for performance counters\n", __func__);
+		ret = -ENOMEM;
+		goto error_unlock;
+	}
+
+	new_devdata->vdev = viodev;
+	new_devdata->dev = &viodev->dev;
+	nx842_OF_set_defaults(new_devdata);
+
+	rcu_assign_pointer(devdata, new_devdata);
+	spin_unlock_irqrestore(&devdata_mutex, flags);
+	synchronize_rcu();
+	kfree(old_devdata);
+
+	of_reconfig_notifier_register(&nx842_of_nb);
+
+	ret = nx842_OF_upd(NULL);
+	if (ret && ret != -ENODEV) {
+		dev_err(&viodev->dev, "could not parse device tree. %d\n", ret);
+		ret = -1;
+		goto error;
+	}
+
+	rcu_read_lock();
+	dev_set_drvdata(&viodev->dev, rcu_dereference(devdata));
+	rcu_read_unlock();
+
+	if (sysfs_create_group(&viodev->dev.kobj, &nx842_attribute_group)) {
+		dev_err(&viodev->dev, "could not create sysfs device attributes\n");
+		ret = -1;
+		goto error;
+	}
+
+	return 0;
+
+error_unlock:
+	spin_unlock_irqrestore(&devdata_mutex, flags);
+	if (new_devdata)
+		kfree(new_devdata->counters);
+	kfree(new_devdata);
+error:
+	return ret;
+}
+
+static int __exit nx842_remove(struct vio_dev *viodev)
+{
+	struct nx842_devdata *old_devdata;
+	unsigned long flags;
+
+	pr_info("Removing IBM Power 842 compression device\n");
+	sysfs_remove_group(&viodev->dev.kobj, &nx842_attribute_group);
+
+	spin_lock_irqsave(&devdata_mutex, flags);
+	old_devdata = rcu_dereference_check(devdata,
+			lockdep_is_held(&devdata_mutex));
+	of_reconfig_notifier_unregister(&nx842_of_nb);
+	RCU_INIT_POINTER(devdata, NULL);
+	spin_unlock_irqrestore(&devdata_mutex, flags);
+	synchronize_rcu();
+	dev_set_drvdata(&viodev->dev, NULL);
+	if (old_devdata)
+		kfree(old_devdata->counters);
+	kfree(old_devdata);
+
+	return 0;
+}
+
+static struct vio_device_id nx842_vio_driver_ids[] = {
+	{"ibm,compression-v1", "ibm,compression"},
+	{"", ""},
+};
+
+static struct vio_driver nx842_vio_driver = {
+	.name = KBUILD_MODNAME,
+	.probe = nx842_probe,
+	.remove = __exit_p(nx842_remove),
+	.get_desired_dma = nx842_get_desired_dma,
+	.id_table = nx842_vio_driver_ids,
+};
+
+static int __init nx842_init(void)
+{
+	struct nx842_devdata *new_devdata;
+	int ret;
+
+	pr_info("Registering IBM Power 842 compression driver\n");
+
+	if (!of_find_compatible_node(NULL, NULL, "ibm,compression"))
+		return -ENODEV;
+
+	RCU_INIT_POINTER(devdata, NULL);
+	new_devdata = kzalloc(sizeof(*new_devdata), GFP_KERNEL);
+	if (!new_devdata) {
+		pr_err("Could not allocate memory for device data\n");
+		return -ENOMEM;
+	}
+	new_devdata->status = UNAVAILABLE;
+	RCU_INIT_POINTER(devdata, new_devdata);
+
+	ret = vio_register_driver(&nx842_vio_driver);
+	if (ret) {
+		pr_err("Could not register VIO driver %d\n", ret);
+
+		kfree(new_devdata);
+		return ret;
+	}
+
+	if (!nx842_platform_driver_set(&nx842_pseries_driver)) {
+		vio_unregister_driver(&nx842_vio_driver);
+		kfree(new_devdata);
+		return -EEXIST;
+	}
+
+	return 0;
+}
+
+module_init(nx842_init);
+
+static void __exit nx842_exit(void)
+{
+	struct nx842_devdata *old_devdata;
+	unsigned long flags;
+
+	pr_info("Exiting IBM Power 842 compression driver\n");
+	nx842_platform_driver_unset(&nx842_pseries_driver);
+	spin_lock_irqsave(&devdata_mutex, flags);
+	old_devdata = rcu_dereference_check(devdata,
+			lockdep_is_held(&devdata_mutex));
+	RCU_INIT_POINTER(devdata, NULL);
+	spin_unlock_irqrestore(&devdata_mutex, flags);
+	synchronize_rcu();
+	if (old_devdata && old_devdata->dev)
+		dev_set_drvdata(old_devdata->dev, NULL);
+	kfree(old_devdata);
+	vio_unregister_driver(&nx842_vio_driver);
+}
+
+module_exit(nx842_exit);
+
diff --git a/drivers/crypto/nx/nx-842.c b/drivers/crypto/nx/nx-842.c
index 887196e9b..6e5e0d60d 100644
--- a/drivers/crypto/nx/nx-842.c
+++ b/drivers/crypto/nx/nx-842.c
@@ -1,5 +1,10 @@
 /*
- * Driver for IBM Power 842 compression accelerator
+ * Driver frontend for IBM Power 842 compression accelerator
+ *
+ * Copyright (C) 2015 Dan Streetman, IBM Corp
+ *
+ * Designer of the Power data compression engine:
+ *   Bulent Abali <abali@us.ibm.com>
  *
  * This program is free software; you can redistribute it and/or modify
  * it under the terms of the GNU General Public License as published by
@@ -10,1594 +15,89 @@
  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
- *
- * Copyright (C) IBM Corporation, 2012
- *
- * Authors: Robert Jennings <rcj@linux.vnet.ibm.com>
- *          Seth Jennings <sjenning@linux.vnet.ibm.com>
  */
 
-#include <linux/kernel.h>
-#include <linux/module.h>
-#include <linux/nx842.h>
-#include <linux/of.h>
-#include <linux/slab.h>
-
-#include <asm/page.h>
-#include <asm/vio.h>
+#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
 
-#include "nx_csbcpb.h" /* struct nx_csbcpb */
+#include "nx-842.h"
 
-#define MODULE_NAME "nx-compress"
 MODULE_LICENSE("GPL");
-MODULE_AUTHOR("Robert Jennings <rcj@linux.vnet.ibm.com>");
+MODULE_AUTHOR("Dan Streetman <ddstreet@ieee.org>");
 MODULE_DESCRIPTION("842 H/W Compression driver for IBM Power processors");
 
-#define SHIFT_4K 12
-#define SHIFT_64K 16
-#define SIZE_4K (1UL << SHIFT_4K)
-#define SIZE_64K (1UL << SHIFT_64K)
-
-/* IO buffer must be 128 byte aligned */
-#define IO_BUFFER_ALIGN 128
-
-struct nx842_header {
-	int blocks_nr; /* number of compressed blocks */
-	int offset; /* offset of the first block (from beginning of header) */
-	int sizes[0]; /* size of compressed blocks */
-};
-
-static inline int nx842_header_size(const struct nx842_header *hdr)
-{
-	return sizeof(struct nx842_header) +
-			hdr->blocks_nr * sizeof(hdr->sizes[0]);
-}
-
-/* Macros for fields within nx_csbcpb */
-/* Check the valid bit within the csbcpb valid field */
-#define NX842_CSBCBP_VALID_CHK(x) (x & BIT_MASK(7))
-
-/* CE macros operate on the completion_extension field bits in the csbcpb.
- * CE0 0=full completion, 1=partial completion
- * CE1 0=CE0 indicates completion, 1=termination (output may be modified)
- * CE2 0=processed_bytes is source bytes, 1=processed_bytes is target bytes */
-#define NX842_CSBCPB_CE0(x)	(x & BIT_MASK(7))
-#define NX842_CSBCPB_CE1(x)	(x & BIT_MASK(6))
-#define NX842_CSBCPB_CE2(x)	(x & BIT_MASK(5))
-
-/* The NX unit accepts data only on 4K page boundaries */
-#define NX842_HW_PAGE_SHIFT	SHIFT_4K
-#define NX842_HW_PAGE_SIZE	(ASM_CONST(1) << NX842_HW_PAGE_SHIFT)
-#define NX842_HW_PAGE_MASK	(~(NX842_HW_PAGE_SIZE-1))
-
-enum nx842_status {
-	UNAVAILABLE,
-	AVAILABLE
-};
-
-struct ibm_nx842_counters {
-	atomic64_t comp_complete;
-	atomic64_t comp_failed;
-	atomic64_t decomp_complete;
-	atomic64_t decomp_failed;
-	atomic64_t swdecomp;
-	atomic64_t comp_times[32];
-	atomic64_t decomp_times[32];
-};
-
-static struct nx842_devdata {
-	struct vio_dev *vdev;
-	struct device *dev;
-	struct ibm_nx842_counters *counters;
-	unsigned int max_sg_len;
-	unsigned int max_sync_size;
-	unsigned int max_sync_sg;
-	enum nx842_status status;
-} __rcu *devdata;
-static DEFINE_SPINLOCK(devdata_mutex);
-
-#define NX842_COUNTER_INC(_x) \
-static inline void nx842_inc_##_x( \
-	const struct nx842_devdata *dev) { \
-	if (dev) \
-		atomic64_inc(&dev->counters->_x); \
-}
-NX842_COUNTER_INC(comp_complete);
-NX842_COUNTER_INC(comp_failed);
-NX842_COUNTER_INC(decomp_complete);
-NX842_COUNTER_INC(decomp_failed);
-NX842_COUNTER_INC(swdecomp);
-
-#define NX842_HIST_SLOTS 16
-
-static void ibm_nx842_incr_hist(atomic64_t *times, unsigned int time)
-{
-	int bucket = fls(time);
-
-	if (bucket)
-		bucket = min((NX842_HIST_SLOTS - 1), bucket - 1);
-
-	atomic64_inc(&times[bucket]);
-}
-
-/* NX unit operation flags */
-#define NX842_OP_COMPRESS	0x0
-#define NX842_OP_CRC		0x1
-#define NX842_OP_DECOMPRESS	0x2
-#define NX842_OP_COMPRESS_CRC   (NX842_OP_COMPRESS | NX842_OP_CRC)
-#define NX842_OP_DECOMPRESS_CRC (NX842_OP_DECOMPRESS | NX842_OP_CRC)
-#define NX842_OP_ASYNC		(1<<23)
-#define NX842_OP_NOTIFY		(1<<22)
-#define NX842_OP_NOTIFY_INT(x)	((x & 0xff)<<8)
-
-static unsigned long nx842_get_desired_dma(struct vio_dev *viodev)
-{
-	/* No use of DMA mappings within the driver. */
-	return 0;
-}
-
-struct nx842_slentry {
-	unsigned long ptr; /* Real address (use __pa()) */
-	unsigned long len;
-};
-
-/* pHyp scatterlist entry */
-struct nx842_scatterlist {
-	int entry_nr; /* number of slentries */
-	struct nx842_slentry *entries; /* ptr to array of slentries */
-};
-
-/* Does not include sizeof(entry_nr) in the size */
-static inline unsigned long nx842_get_scatterlist_size(
-				struct nx842_scatterlist *sl)
-{
-	return sl->entry_nr * sizeof(struct nx842_slentry);
-}
-
-static inline unsigned long nx842_get_pa(void *addr)
-{
-	if (is_vmalloc_addr(addr))
-		return page_to_phys(vmalloc_to_page(addr))
-		       + offset_in_page(addr);
-	else
-		return __pa(addr);
-}
-
-static int nx842_build_scatterlist(unsigned long buf, int len,
-			struct nx842_scatterlist *sl)
-{
-	unsigned long nextpage;
-	struct nx842_slentry *entry;
-
-	sl->entry_nr = 0;
-
-	entry = sl->entries;
-	while (len) {
-		entry->ptr = nx842_get_pa((void *)buf);
-		nextpage = ALIGN(buf + 1, NX842_HW_PAGE_SIZE);
-		if (nextpage < buf + len) {
-			/* we aren't at the end yet */
-			if (IS_ALIGNED(buf, NX842_HW_PAGE_SIZE))
-				/* we are in the middle (or beginning) */
-				entry->len = NX842_HW_PAGE_SIZE;
-			else
-				/* we are at the beginning */
-				entry->len = nextpage - buf;
-		} else {
-			/* at the end */
-			entry->len = len;
-		}
-
-		len -= entry->len;
-		buf += entry->len;
-		sl->entry_nr++;
-		entry++;
-	}
-
-	return 0;
-}
-
-/*
- * Working memory for software decompression
- */
-struct sw842_fifo {
-	union {
-		char f8[256][8];
-		char f4[512][4];
-	};
-	char f2[256][2];
-	unsigned char f84_full;
-	unsigned char f2_full;
-	unsigned char f8_count;
-	unsigned char f2_count;
-	unsigned int f4_count;
-};
-
-/*
- * Working memory for crypto API
+/**
+ * nx842_constraints
+ *
+ * This provides the driver's constraints.  Different nx842 implementations
+ * may have varying requirements.  The constraints are:
+ *   @alignment:	All buffers should be aligned to this
+ *   @multiple:		All buffer lengths should be a multiple of this
+ *   @minimum:		Buffer lengths must not be less than this amount
+ *   @maximum:		Buffer lengths must not be more than this amount
+ *
+ * The constraints apply to all buffers and lengths, both input and output,
+ * for both compression and decompression, except for the minimum which
+ * only applies to compression input and decompression output; the
+ * compressed data can be less than the minimum constraint.  It can be
+ * assumed that compressed data will always adhere to the multiple
+ * constraint.
+ *
+ * The driver may succeed even if these constraints are violated;
+ * however the driver can return failure or suffer reduced performance
+ * if any constraint is not met.
  */
-struct nx842_workmem {
-	char bounce[PAGE_SIZE]; /* bounce buffer for decompression input */
-	union {
-		/* hardware working memory */
-		struct {
-			/* scatterlist */
-			char slin[SIZE_4K];
-			char slout[SIZE_4K];
-			/* coprocessor status/parameter block */
-			struct nx_csbcpb csbcpb;
-		};
-		/* software working memory */
-		struct sw842_fifo swfifo; /* software decompression fifo */
-	};
-};
-
-int nx842_get_workmem_size(void)
-{
-	return sizeof(struct nx842_workmem) + NX842_HW_PAGE_SIZE;
-}
-EXPORT_SYMBOL_GPL(nx842_get_workmem_size);
-
-int nx842_get_workmem_size_aligned(void)
-{
-	return sizeof(struct nx842_workmem);
-}
-EXPORT_SYMBOL_GPL(nx842_get_workmem_size_aligned);
-
-static int nx842_validate_result(struct device *dev,
-	struct cop_status_block *csb)
+int nx842_constraints(struct nx842_constraints *c)
 {
-	/* The csb must be valid after returning from vio_h_cop_sync */
-	if (!NX842_CSBCBP_VALID_CHK(csb->valid)) {
-		dev_err(dev, "%s: cspcbp not valid upon completion.\n",
-				__func__);
-		dev_dbg(dev, "valid:0x%02x cs:0x%02x cc:0x%02x ce:0x%02x\n",
-				csb->valid,
-				csb->crb_seq_number,
-				csb->completion_code,
-				csb->completion_extension);
-		dev_dbg(dev, "processed_bytes:%d address:0x%016lx\n",
-				csb->processed_byte_count,
-				(unsigned long)csb->address);
-		return -EIO;
-	}
-
-	/* Check return values from the hardware in the CSB */
-	switch (csb->completion_code) {
-	case 0:	/* Completed without error */
-		break;
-	case 64: /* Target bytes > Source bytes during compression */
-	case 13: /* Output buffer too small */
-		dev_dbg(dev, "%s: Compression output larger than input\n",
-					__func__);
-		return -ENOSPC;
-	case 66: /* Input data contains an illegal template field */
-	case 67: /* Template indicates data past the end of the input stream */
-		dev_dbg(dev, "%s: Bad data for decompression (code:%d)\n",
-					__func__, csb->completion_code);
-		return -EINVAL;
-	default:
-		dev_dbg(dev, "%s: Unspecified error (code:%d)\n",
-					__func__, csb->completion_code);
-		return -EIO;
-	}
-
-	/* Hardware sanity check */
-	if (!NX842_CSBCPB_CE2(csb->completion_extension)) {
-		dev_err(dev, "%s: No error returned by hardware, but "
-				"data returned is unusable, contact support.\n"
-				"(Additional info: csbcbp->processed bytes "
-				"does not specify processed bytes for the "
-				"target buffer.)\n", __func__);
-		return -EIO;
-	}
-
+	memcpy(c, nx842_platform_driver()->constraints, sizeof(*c));
 	return 0;
 }
+EXPORT_SYMBOL_GPL(nx842_constraints);
 
 /**
- * nx842_compress - Compress data using the 842 algorithm
- *
- * Compression provide by the NX842 coprocessor on IBM Power systems.
- * The input buffer is compressed and the result is stored in the
- * provided output buffer.
- *
- * Upon return from this function @outlen contains the length of the
- * compressed data.  If there is an error then @outlen will be 0 and an
- * error will be specified by the return code from this function.
- *
- * @in: Pointer to input buffer, must be page aligned
- * @inlen: Length of input buffer, must be PAGE_SIZE
- * @out: Pointer to output buffer
- * @outlen: Length of output buffer
- * @wrkmem: ptr to buffer for working memory, size determined by
- *          nx842_get_workmem_size()
+ * nx842_workmem_size
  *
- * Returns:
- *   0		Success, output of length @outlen stored in the buffer at @out
- *   -ENOMEM	Unable to allocate internal buffers
- *   -ENOSPC	Output buffer is to small
- *   -EMSGSIZE	XXX Difficult to describe this limitation
- *   -EIO	Internal error
- *   -ENODEV	Hardware unavailable
+ * Get the amount of working memory the driver requires.
  */
-int nx842_compress(const unsigned char *in, unsigned int inlen,
-		       unsigned char *out, unsigned int *outlen, void *wmem)
+size_t nx842_workmem_size(void)
 {
-	struct nx842_header *hdr;
-	struct nx842_devdata *local_devdata;
-	struct device *dev = NULL;
-	struct nx842_workmem *workmem;
-	struct nx842_scatterlist slin, slout;
-	struct nx_csbcpb *csbcpb;
-	int ret = 0, max_sync_size, i, bytesleft, size, hdrsize;
-	unsigned long inbuf, outbuf, padding;
-	struct vio_pfo_op op = {
-		.done = NULL,
-		.handle = 0,
-		.timeout = 0,
-	};
-	unsigned long start_time = get_tb();
-
-	/*
-	 * Make sure input buffer is 64k page aligned.  This is assumed since
-	 * this driver is designed for page compression only (for now).  This
-	 * is very nice since we can now use direct DDE(s) for the input and
-	 * the alignment is guaranteed.
-	*/
-	inbuf = (unsigned long)in;
-	if (!IS_ALIGNED(inbuf, PAGE_SIZE) || inlen != PAGE_SIZE)
-		return -EINVAL;
-
-	rcu_read_lock();
-	local_devdata = rcu_dereference(devdata);
-	if (!local_devdata || !local_devdata->dev) {
-		rcu_read_unlock();
-		return -ENODEV;
-	}
-	max_sync_size = local_devdata->max_sync_size;
-	dev = local_devdata->dev;
-
-	/* Create the header */
-	hdr = (struct nx842_header *)out;
-	hdr->blocks_nr = PAGE_SIZE / max_sync_size;
-	hdrsize = nx842_header_size(hdr);
-	outbuf = (unsigned long)out + hdrsize;
-	bytesleft = *outlen - hdrsize;
-
-	/* Init scatterlist */
-	workmem = (struct nx842_workmem *)ALIGN((unsigned long)wmem,
-		NX842_HW_PAGE_SIZE);
-	slin.entries = (struct nx842_slentry *)workmem->slin;
-	slout.entries = (struct nx842_slentry *)workmem->slout;
-
-	/* Init operation */
-	op.flags = NX842_OP_COMPRESS;
-	csbcpb = &workmem->csbcpb;
-	memset(csbcpb, 0, sizeof(*csbcpb));
-	op.csbcpb = nx842_get_pa(csbcpb);
-	op.out = nx842_get_pa(slout.entries);
-
-	for (i = 0; i < hdr->blocks_nr; i++) {
-		/*
-		 * Aligning the output blocks to 128 bytes does waste space,
-		 * but it prevents the need for bounce buffers and memory
-		 * copies.  It also simplifies the code a lot.  In the worst
-		 * case (64k page, 4k max_sync_size), you lose up to
-		 * (128*16)/64k = ~3% the compression factor. For 64k
-		 * max_sync_size, the loss would be at most 128/64k = ~0.2%.
-		 */
-		padding = ALIGN(outbuf, IO_BUFFER_ALIGN) - outbuf;
-		outbuf += padding;
-		bytesleft -= padding;
-		if (i == 0)
-			/* save offset into first block in header */
-			hdr->offset = padding + hdrsize;
-
-		if (bytesleft <= 0) {
-			ret = -ENOSPC;
-			goto unlock;
-		}
-
-		/*
-		 * NOTE: If the default max_sync_size is changed from 4k
-		 * to 64k, remove the "likely" case below, since a
-		 * scatterlist will always be needed.
-		 */
-		if (likely(max_sync_size == NX842_HW_PAGE_SIZE)) {
-			/* Create direct DDE */
-			op.in = nx842_get_pa((void *)inbuf);
-			op.inlen = max_sync_size;
-
-		} else {
-			/* Create indirect DDE (scatterlist) */
-			nx842_build_scatterlist(inbuf, max_sync_size, &slin);
-			op.in = nx842_get_pa(slin.entries);
-			op.inlen = -nx842_get_scatterlist_size(&slin);
-		}
-
-		/*
-		 * If max_sync_size != NX842_HW_PAGE_SIZE, an indirect
-		 * DDE is required for the outbuf.
-		 * If max_sync_size == NX842_HW_PAGE_SIZE, outbuf must
-		 * also be page aligned (1 in 128/4k=32 chance) in order
-		 * to use a direct DDE.
-		 * This is unlikely, just use an indirect DDE always.
-		 */
-		nx842_build_scatterlist(outbuf,
-			min(bytesleft, max_sync_size), &slout);
-		/* op.out set before loop */
-		op.outlen = -nx842_get_scatterlist_size(&slout);
-
-		/* Send request to pHyp */
-		ret = vio_h_cop_sync(local_devdata->vdev, &op);
-
-		/* Check for pHyp error */
-		if (ret) {
-			dev_dbg(dev, "%s: vio_h_cop_sync error (ret=%d, hret=%ld)\n",
-				__func__, ret, op.hcall_err);
-			ret = -EIO;
-			goto unlock;
-		}
-
-		/* Check for hardware error */
-		ret = nx842_validate_result(dev, &csbcpb->csb);
-		if (ret && ret != -ENOSPC)
-			goto unlock;
-
-		/* Handle incompressible data */
-		if (unlikely(ret == -ENOSPC)) {
-			if (bytesleft < max_sync_size) {
-				/*
-				 * Not enough space left in the output buffer
-				 * to store uncompressed block
-				 */
-				goto unlock;
-			} else {
-				/* Store incompressible block */
-				memcpy((void *)outbuf, (void *)inbuf,
-					max_sync_size);
-				hdr->sizes[i] = -max_sync_size;
-				outbuf += max_sync_size;
-				bytesleft -= max_sync_size;
-				/* Reset ret, incompressible data handled */
-				ret = 0;
-			}
-		} else {
-			/* Normal case, compression was successful */
-			size = csbcpb->csb.processed_byte_count;
-			dev_dbg(dev, "%s: processed_bytes=%d\n",
-				__func__, size);
-			hdr->sizes[i] = size;
-			outbuf += size;
-			bytesleft -= size;
-		}
-
-		inbuf += max_sync_size;
-	}
-
-	*outlen = (unsigned int)(outbuf - (unsigned long)out);
-
-unlock:
-	if (ret)
-		nx842_inc_comp_failed(local_devdata);
-	else {
-		nx842_inc_comp_complete(local_devdata);
-		ibm_nx842_incr_hist(local_devdata->counters->comp_times,
-			(get_tb() - start_time) / tb_ticks_per_usec);
-	}
-	rcu_read_unlock();
-	return ret;
+	return nx842_platform_driver()->workmem_size;
 }
-EXPORT_SYMBOL_GPL(nx842_compress);
-
-static int sw842_decompress(const unsigned char *, int, unsigned char *, int *,
-			const void *);
+EXPORT_SYMBOL_GPL(nx842_workmem_size);
 
-/**
- * nx842_decompress - Decompress data using the 842 algorithm
- *
- * Decompression provide by the NX842 coprocessor on IBM Power systems.
- * The input buffer is decompressed and the result is stored in the
- * provided output buffer.  The size allocated to the output buffer is
- * provided by the caller of this function in @outlen.  Upon return from
- * this function @outlen contains the length of the decompressed data.
- * If there is an error then @outlen will be 0 and an error will be
- * specified by the return code from this function.
- *
- * @in: Pointer to input buffer, will use bounce buffer if not 128 byte
- *      aligned
- * @inlen: Length of input buffer
- * @out: Pointer to output buffer, must be page aligned
- * @outlen: Length of output buffer, must be PAGE_SIZE
- * @wrkmem: ptr to buffer for working memory, size determined by
- *          nx842_get_workmem_size()
- *
- * Returns:
- *   0		Success, output of length @outlen stored in the buffer at @out
- *   -ENODEV	Hardware decompression device is unavailable
- *   -ENOMEM	Unable to allocate internal buffers
- *   -ENOSPC	Output buffer is to small
- *   -EINVAL	Bad input data encountered when attempting decompress
- *   -EIO	Internal error
- */
-int nx842_decompress(const unsigned char *in, unsigned int inlen,
-			 unsigned char *out, unsigned int *outlen, void *wmem)
+int nx842_compress(const unsigned char *in, unsigned int ilen,
+		   unsigned char *out, unsigned int *olen, void *wmem)
 {
-	struct nx842_header *hdr;
-	struct nx842_devdata *local_devdata;
-	struct device *dev = NULL;
-	struct nx842_workmem *workmem;
-	struct nx842_scatterlist slin, slout;
-	struct nx_csbcpb *csbcpb;
-	int ret = 0, i, size, max_sync_size;
-	unsigned long inbuf, outbuf;
-	struct vio_pfo_op op = {
-		.done = NULL,
-		.handle = 0,
-		.timeout = 0,
-	};
-	unsigned long start_time = get_tb();
-
-	/* Ensure page alignment and size */
-	outbuf = (unsigned long)out;
-	if (!IS_ALIGNED(outbuf, PAGE_SIZE) || *outlen != PAGE_SIZE)
-		return -EINVAL;
-
-	rcu_read_lock();
-	local_devdata = rcu_dereference(devdata);
-	if (local_devdata)
-		dev = local_devdata->dev;
-
-	/* Get header */
-	hdr = (struct nx842_header *)in;
-
-	workmem = (struct nx842_workmem *)ALIGN((unsigned long)wmem,
-		NX842_HW_PAGE_SIZE);
-
-	inbuf = (unsigned long)in + hdr->offset;
-	if (likely(!IS_ALIGNED(inbuf, IO_BUFFER_ALIGN))) {
-		/* Copy block(s) into bounce buffer for alignment */
-		memcpy(workmem->bounce, in + hdr->offset, inlen - hdr->offset);
-		inbuf = (unsigned long)workmem->bounce;
-	}
-
-	/* Init scatterlist */
-	slin.entries = (struct nx842_slentry *)workmem->slin;
-	slout.entries = (struct nx842_slentry *)workmem->slout;
-
-	/* Init operation */
-	op.flags = NX842_OP_DECOMPRESS;
-	csbcpb = &workmem->csbcpb;
-	memset(csbcpb, 0, sizeof(*csbcpb));
-	op.csbcpb = nx842_get_pa(csbcpb);
-
-	/*
-	 * max_sync_size may have changed since compression,
-	 * so we can't read it from the device info. We need
-	 * to derive it from hdr->blocks_nr.
-	 */
-	max_sync_size = PAGE_SIZE / hdr->blocks_nr;
-
-	for (i = 0; i < hdr->blocks_nr; i++) {
-		/* Skip padding */
-		inbuf = ALIGN(inbuf, IO_BUFFER_ALIGN);
-
-		if (hdr->sizes[i] < 0) {
-			/* Negative sizes indicate uncompressed data blocks */
-			size = abs(hdr->sizes[i]);
-			memcpy((void *)outbuf, (void *)inbuf, size);
-			outbuf += size;
-			inbuf += size;
-			continue;
-		}
-
-		if (!dev)
-			goto sw;
-
-		/*
-		 * The better the compression, the more likely the "likely"
-		 * case becomes.
-		 */
-		if (likely((inbuf & NX842_HW_PAGE_MASK) ==
-			((inbuf + hdr->sizes[i] - 1) & NX842_HW_PAGE_MASK))) {
-			/* Create direct DDE */
-			op.in = nx842_get_pa((void *)inbuf);
-			op.inlen = hdr->sizes[i];
-		} else {
-			/* Create indirect DDE (scatterlist) */
-			nx842_build_scatterlist(inbuf, hdr->sizes[i] , &slin);
-			op.in = nx842_get_pa(slin.entries);
-			op.inlen = -nx842_get_scatterlist_size(&slin);
-		}
-
-		/*
-		 * NOTE: If the default max_sync_size is changed from 4k
-		 * to 64k, remove the "likely" case below, since a
-		 * scatterlist will always be needed.
-		 */
-		if (likely(max_sync_size == NX842_HW_PAGE_SIZE)) {
-			/* Create direct DDE */
-			op.out = nx842_get_pa((void *)outbuf);
-			op.outlen = max_sync_size;
-		} else {
-			/* Create indirect DDE (scatterlist) */
-			nx842_build_scatterlist(outbuf, max_sync_size, &slout);
-			op.out = nx842_get_pa(slout.entries);
-			op.outlen = -nx842_get_scatterlist_size(&slout);
-		}
-
-		/* Send request to pHyp */
-		ret = vio_h_cop_sync(local_devdata->vdev, &op);
-
-		/* Check for pHyp error */
-		if (ret) {
-			dev_dbg(dev, "%s: vio_h_cop_sync error (ret=%d, hret=%ld)\n",
-				__func__, ret, op.hcall_err);
-			dev = NULL;
-			goto sw;
-		}
-
-		/* Check for hardware error */
-		ret = nx842_validate_result(dev, &csbcpb->csb);
-		if (ret) {
-			dev = NULL;
-			goto sw;
-		}
-
-		/* HW decompression success */
-		inbuf += hdr->sizes[i];
-		outbuf += csbcpb->csb.processed_byte_count;
-		continue;
-
-sw:
-		/* software decompression */
-		size = max_sync_size;
-		ret = sw842_decompress(
-			(unsigned char *)inbuf, hdr->sizes[i],
-			(unsigned char *)outbuf, &size, wmem);
-		if (ret)
-			pr_debug("%s: sw842_decompress failed with %d\n",
-				__func__, ret);
-
-		if (ret) {
-			if (ret != -ENOSPC && ret != -EINVAL &&
-					ret != -EMSGSIZE)
-				ret = -EIO;
-			goto unlock;
-		}
-
-		/* SW decompression success */
-		inbuf += hdr->sizes[i];
-		outbuf += size;
-	}
-
-	*outlen = (unsigned int)(outbuf - (unsigned long)out);
-
-unlock:
-	if (ret)
-		/* decompress fail */
-		nx842_inc_decomp_failed(local_devdata);
-	else {
-		if (!dev)
-			/* software decompress */
-			nx842_inc_swdecomp(local_devdata);
-		nx842_inc_decomp_complete(local_devdata);
-		ibm_nx842_incr_hist(local_devdata->counters->decomp_times,
-			(get_tb() - start_time) / tb_ticks_per_usec);
-	}
-
-	rcu_read_unlock();
-	return ret;
+	return nx842_platform_driver()->compress(in, ilen, out, olen, wmem);
 }
-EXPORT_SYMBOL_GPL(nx842_decompress);
+EXPORT_SYMBOL_GPL(nx842_compress);
 
-/**
- * nx842_OF_set_defaults -- Set default (disabled) values for devdata
- *
- * @devdata - struct nx842_devdata to update
- *
- * Returns:
- *  0 on success
- *  -ENOENT if @devdata ptr is NULL
- */
-static int nx842_OF_set_defaults(struct nx842_devdata *devdata)
+int nx842_decompress(const unsigned char *in, unsigned int ilen,
+		     unsigned char *out, unsigned int *olen, void *wmem)
 {
-	if (devdata) {
-		devdata->max_sync_size = 0;
-		devdata->max_sync_sg = 0;
-		devdata->max_sg_len = 0;
-		devdata->status = UNAVAILABLE;
-		return 0;
-	} else
-		return -ENOENT;
-}
-
-/**
- * nx842_OF_upd_status -- Update the device info from OF status prop
- *
- * The status property indicates if the accelerator is enabled.  If the
- * device is in the OF tree it indicates that the hardware is present.
- * The status field indicates if the device is enabled when the status
- * is 'okay'.  Otherwise the device driver will be disabled.
- *
- * @devdata - struct nx842_devdata to update
- * @prop - struct property point containing the maxsyncop for the update
- *
- * Returns:
- *  0 - Device is available
- *  -EINVAL - Device is not available
- */
-static int nx842_OF_upd_status(struct nx842_devdata *devdata,
-					struct property *prop) {
-	int ret = 0;
-	const char *status = (const char *)prop->value;
-
-	if (!strncmp(status, "okay", (size_t)prop->length)) {
-		devdata->status = AVAILABLE;
-	} else {
-		dev_info(devdata->dev, "%s: status '%s' is not 'okay'\n",
-				__func__, status);
-		devdata->status = UNAVAILABLE;
-	}
-
-	return ret;
-}
-
-/**
- * nx842_OF_upd_maxsglen -- Update the device info from OF maxsglen prop
- *
- * Definition of the 'ibm,max-sg-len' OF property:
- *  This field indicates the maximum byte length of a scatter list
- *  for the platform facility. It is a single cell encoded as with encode-int.
- *
- * Example:
- *  # od -x ibm,max-sg-len
- *  0000000 0000 0ff0
- *
- *  In this example, the maximum byte length of a scatter list is
- *  0x0ff0 (4,080).
- *
- * @devdata - struct nx842_devdata to update
- * @prop - struct property point containing the maxsyncop for the update
- *
- * Returns:
- *  0 on success
- *  -EINVAL on failure
- */
-static int nx842_OF_upd_maxsglen(struct nx842_devdata *devdata,
-					struct property *prop) {
-	int ret = 0;
-	const int *maxsglen = prop->value;
-
-	if (prop->length != sizeof(*maxsglen)) {
-		dev_err(devdata->dev, "%s: unexpected format for ibm,max-sg-len property\n", __func__);
-		dev_dbg(devdata->dev, "%s: ibm,max-sg-len is %d bytes long, expected %lu bytes\n", __func__,
-				prop->length, sizeof(*maxsglen));
-		ret = -EINVAL;
-	} else {
-		devdata->max_sg_len = (unsigned int)min(*maxsglen,
-				(int)NX842_HW_PAGE_SIZE);
-	}
-
-	return ret;
-}
-
-/**
- * nx842_OF_upd_maxsyncop -- Update the device info from OF maxsyncop prop
- *
- * Definition of the 'ibm,max-sync-cop' OF property:
- *  Two series of cells.  The first series of cells represents the maximums
- *  that can be synchronously compressed. The second series of cells
- *  represents the maximums that can be synchronously decompressed.
- *  1. The first cell in each series contains the count of the number of
- *     data length, scatter list elements pairs that follow – each being
- *     of the form
- *    a. One cell data byte length
- *    b. One cell total number of scatter list elements
- *
- * Example:
- *  # od -x ibm,max-sync-cop
- *  0000000 0000 0001 0000 1000 0000 01fe 0000 0001
- *  0000020 0000 1000 0000 01fe
- *
- *  In this example, compression supports 0x1000 (4,096) data byte length
- *  and 0x1fe (510) total scatter list elements.  Decompression supports
- *  0x1000 (4,096) data byte length and 0x1f3 (510) total scatter list
- *  elements.
- *
- * @devdata - struct nx842_devdata to update
- * @prop - struct property point containing the maxsyncop for the update
- *
- * Returns:
- *  0 on success
- *  -EINVAL on failure
- */
-static int nx842_OF_upd_maxsyncop(struct nx842_devdata *devdata,
-					struct property *prop) {
-	int ret = 0;
-	const struct maxsynccop_t {
-		int comp_elements;
-		int comp_data_limit;
-		int comp_sg_limit;
-		int decomp_elements;
-		int decomp_data_limit;
-		int decomp_sg_limit;
-	} *maxsynccop;
-
-	if (prop->length != sizeof(*maxsynccop)) {
-		dev_err(devdata->dev, "%s: unexpected format for ibm,max-sync-cop property\n", __func__);
-		dev_dbg(devdata->dev, "%s: ibm,max-sync-cop is %d bytes long, expected %lu bytes\n", __func__, prop->length,
-				sizeof(*maxsynccop));
-		ret = -EINVAL;
-		goto out;
-	}
-
-	maxsynccop = (const struct maxsynccop_t *)prop->value;
-
-	/* Use one limit rather than separate limits for compression and
-	 * decompression. Set a maximum for this so as not to exceed the
-	 * size that the header can support and round the value down to
-	 * the hardware page size (4K) */
-	devdata->max_sync_size =
-			(unsigned int)min(maxsynccop->comp_data_limit,
-					maxsynccop->decomp_data_limit);
-
-	devdata->max_sync_size = min_t(unsigned int, devdata->max_sync_size,
-					SIZE_64K);
-
-	if (devdata->max_sync_size < SIZE_4K) {
-		dev_err(devdata->dev, "%s: hardware max data size (%u) is "
-				"less than the driver minimum, unable to use "
-				"the hardware device\n",
-				__func__, devdata->max_sync_size);
-		ret = -EINVAL;
-		goto out;
-	}
-
-	devdata->max_sync_sg = (unsigned int)min(maxsynccop->comp_sg_limit,
-						maxsynccop->decomp_sg_limit);
-	if (devdata->max_sync_sg < 1) {
-		dev_err(devdata->dev, "%s: hardware max sg size (%u) is "
-				"less than the driver minimum, unable to use "
-				"the hardware device\n",
-				__func__, devdata->max_sync_sg);
-		ret = -EINVAL;
-		goto out;
-	}
-
-out:
-	return ret;
+	return nx842_platform_driver()->decompress(in, ilen, out, olen, wmem);
 }
+EXPORT_SYMBOL_GPL(nx842_decompress);
 
-/**
- *
- * nx842_OF_upd -- Handle OF properties updates for the device.
- *
- * Set all properties from the OF tree.  Optionally, a new property
- * can be provided by the @new_prop pointer to overwrite an existing value.
- * The device will remain disabled until all values are valid, this function
- * will return an error for updates unless all values are valid.
- *
- * @new_prop: If not NULL, this property is being updated.  If NULL, update
- *  all properties from the current values in the OF tree.
- *
- * Returns:
- *  0 - Success
- *  -ENOMEM - Could not allocate memory for new devdata structure
- *  -EINVAL - property value not found, new_prop is not a recognized
- *	property for the device or property value is not valid.
- *  -ENODEV - Device is not available
- */
-static int nx842_OF_upd(struct property *new_prop)
+static __init int nx842_init(void)
 {
-	struct nx842_devdata *old_devdata = NULL;
-	struct nx842_devdata *new_devdata = NULL;
-	struct device_node *of_node = NULL;
-	struct property *status = NULL;
-	struct property *maxsglen = NULL;
-	struct property *maxsyncop = NULL;
-	int ret = 0;
-	unsigned long flags;
-
-	spin_lock_irqsave(&devdata_mutex, flags);
-	old_devdata = rcu_dereference_check(devdata,
-			lockdep_is_held(&devdata_mutex));
-	if (old_devdata)
-		of_node = old_devdata->dev->of_node;
+	request_module("nx-compress-powernv");
+	request_module("nx-compress-pseries");
 
-	if (!old_devdata || !of_node) {
-		pr_err("%s: device is not available\n", __func__);
-		spin_unlock_irqrestore(&devdata_mutex, flags);
-		return -ENODEV;
-	}
-
-	new_devdata = kzalloc(sizeof(*new_devdata), GFP_NOFS);
-	if (!new_devdata) {
-		dev_err(old_devdata->dev, "%s: Could not allocate memory for device data\n", __func__);
-		ret = -ENOMEM;
-		goto error_out;
-	}
-
-	memcpy(new_devdata, old_devdata, sizeof(*old_devdata));
-	new_devdata->counters = old_devdata->counters;
-
-	/* Set ptrs for existing properties */
-	status = of_find_property(of_node, "status", NULL);
-	maxsglen = of_find_property(of_node, "ibm,max-sg-len", NULL);
-	maxsyncop = of_find_property(of_node, "ibm,max-sync-cop", NULL);
-	if (!status || !maxsglen || !maxsyncop) {
-		dev_err(old_devdata->dev, "%s: Could not locate device properties\n", __func__);
-		ret = -EINVAL;
-		goto error_out;
-	}
-
-	/*
-	 * If this is a property update, there are only certain properties that
-	 * we care about. Bail if it isn't in the below list
+	/* we prevent loading if there's no platform driver, and we get the
+	 * module that set it so it won't unload, so we don't need to check
+	 * if it's set in any of the above functions
 	 */
-	if (new_prop && (strncmp(new_prop->name, "status", new_prop->length) ||
-		         strncmp(new_prop->name, "ibm,max-sg-len", new_prop->length) ||
-		         strncmp(new_prop->name, "ibm,max-sync-cop", new_prop->length)))
-		goto out;
-
-	/* Perform property updates */
-	ret = nx842_OF_upd_status(new_devdata, status);
-	if (ret)
-		goto error_out;
-
-	ret = nx842_OF_upd_maxsglen(new_devdata, maxsglen);
-	if (ret)
-		goto error_out;
-
-	ret = nx842_OF_upd_maxsyncop(new_devdata, maxsyncop);
-	if (ret)
-		goto error_out;
-
-out:
-	dev_info(old_devdata->dev, "%s: max_sync_size new:%u old:%u\n",
-			__func__, new_devdata->max_sync_size,
-			old_devdata->max_sync_size);
-	dev_info(old_devdata->dev, "%s: max_sync_sg new:%u old:%u\n",
-			__func__, new_devdata->max_sync_sg,
-			old_devdata->max_sync_sg);
-	dev_info(old_devdata->dev, "%s: max_sg_len new:%u old:%u\n",
-			__func__, new_devdata->max_sg_len,
-			old_devdata->max_sg_len);
-
-	rcu_assign_pointer(devdata, new_devdata);
-	spin_unlock_irqrestore(&devdata_mutex, flags);
-	synchronize_rcu();
-	dev_set_drvdata(new_devdata->dev, new_devdata);
-	kfree(old_devdata);
-	return 0;
-
-error_out:
-	if (new_devdata) {
-		dev_info(old_devdata->dev, "%s: device disabled\n", __func__);
-		nx842_OF_set_defaults(new_devdata);
-		rcu_assign_pointer(devdata, new_devdata);
-		spin_unlock_irqrestore(&devdata_mutex, flags);
-		synchronize_rcu();
-		dev_set_drvdata(new_devdata->dev, new_devdata);
-		kfree(old_devdata);
-	} else {
-		dev_err(old_devdata->dev, "%s: could not update driver from hardware\n", __func__);
-		spin_unlock_irqrestore(&devdata_mutex, flags);
-	}
-
-	if (!ret)
-		ret = -EINVAL;
-	return ret;
-}
-
-/**
- * nx842_OF_notifier - Process updates to OF properties for the device
- *
- * @np: notifier block
- * @action: notifier action
- * @update: struct pSeries_reconfig_prop_update pointer if action is
- *	PSERIES_UPDATE_PROPERTY
- *
- * Returns:
- *	NOTIFY_OK on success
- *	NOTIFY_BAD encoded with error number on failure, use
- *		notifier_to_errno() to decode this value
- */
-static int nx842_OF_notifier(struct notifier_block *np, unsigned long action,
-			     void *data)
-{
-	struct of_reconfig_data *upd = data;
-	struct nx842_devdata *local_devdata;
-	struct device_node *node = NULL;
-
-	rcu_read_lock();
-	local_devdata = rcu_dereference(devdata);
-	if (local_devdata)
-		node = local_devdata->dev->of_node;
-
-	if (local_devdata &&
-			action == OF_RECONFIG_UPDATE_PROPERTY &&
-			!strcmp(upd->dn->name, node->name)) {
-		rcu_read_unlock();
-		nx842_OF_upd(upd->prop);
-	} else
-		rcu_read_unlock();
-
-	return NOTIFY_OK;
-}
-
-static struct notifier_block nx842_of_nb = {
-	.notifier_call = nx842_OF_notifier,
-};
-
-#define nx842_counter_read(_name)					\
-static ssize_t nx842_##_name##_show(struct device *dev,		\
-		struct device_attribute *attr,				\
-		char *buf) {						\
-	struct nx842_devdata *local_devdata;			\
-	int p = 0;							\
-	rcu_read_lock();						\
-	local_devdata = rcu_dereference(devdata);			\
-	if (local_devdata)						\
-		p = snprintf(buf, PAGE_SIZE, "%ld\n",			\
-		       atomic64_read(&local_devdata->counters->_name));	\
-	rcu_read_unlock();						\
-	return p;							\
-}
-
-#define NX842DEV_COUNTER_ATTR_RO(_name)					\
-	nx842_counter_read(_name);					\
-	static struct device_attribute dev_attr_##_name = __ATTR(_name,	\
-						0444,			\
-						nx842_##_name##_show,\
-						NULL);
-
-NX842DEV_COUNTER_ATTR_RO(comp_complete);
-NX842DEV_COUNTER_ATTR_RO(comp_failed);
-NX842DEV_COUNTER_ATTR_RO(decomp_complete);
-NX842DEV_COUNTER_ATTR_RO(decomp_failed);
-NX842DEV_COUNTER_ATTR_RO(swdecomp);
-
-static ssize_t nx842_timehist_show(struct device *,
-		struct device_attribute *, char *);
-
-static struct device_attribute dev_attr_comp_times = __ATTR(comp_times, 0444,
-		nx842_timehist_show, NULL);
-static struct device_attribute dev_attr_decomp_times = __ATTR(decomp_times,
-		0444, nx842_timehist_show, NULL);
-
-static ssize_t nx842_timehist_show(struct device *dev,
-		struct device_attribute *attr, char *buf) {
-	char *p = buf;
-	struct nx842_devdata *local_devdata;
-	atomic64_t *times;
-	int bytes_remain = PAGE_SIZE;
-	int bytes;
-	int i;
-
-	rcu_read_lock();
-	local_devdata = rcu_dereference(devdata);
-	if (!local_devdata) {
-		rcu_read_unlock();
-		return 0;
-	}
-
-	if (attr == &dev_attr_comp_times)
-		times = local_devdata->counters->comp_times;
-	else if (attr == &dev_attr_decomp_times)
-		times = local_devdata->counters->decomp_times;
-	else {
-		rcu_read_unlock();
-		return 0;
-	}
-
-	for (i = 0; i < (NX842_HIST_SLOTS - 2); i++) {
-		bytes = snprintf(p, bytes_remain, "%u-%uus:\t%ld\n",
-			       i ? (2<<(i-1)) : 0, (2<<i)-1,
-			       atomic64_read(&times[i]));
-		bytes_remain -= bytes;
-		p += bytes;
-	}
-	/* The last bucket holds everything over
-	 * 2<<(NX842_HIST_SLOTS - 2) us */
-	bytes = snprintf(p, bytes_remain, "%uus - :\t%ld\n",
-			2<<(NX842_HIST_SLOTS - 2),
-			atomic64_read(&times[(NX842_HIST_SLOTS - 1)]));
-	p += bytes;
-
-	rcu_read_unlock();
-	return p - buf;
-}
-
-static struct attribute *nx842_sysfs_entries[] = {
-	&dev_attr_comp_complete.attr,
-	&dev_attr_comp_failed.attr,
-	&dev_attr_decomp_complete.attr,
-	&dev_attr_decomp_failed.attr,
-	&dev_attr_swdecomp.attr,
-	&dev_attr_comp_times.attr,
-	&dev_attr_decomp_times.attr,
-	NULL,
-};
-
-static struct attribute_group nx842_attribute_group = {
-	.name = NULL,		/* put in device directory */
-	.attrs = nx842_sysfs_entries,
-};
-
-static int __init nx842_probe(struct vio_dev *viodev,
-				  const struct vio_device_id *id)
-{
-	struct nx842_devdata *old_devdata, *new_devdata = NULL;
-	unsigned long flags;
-	int ret = 0;
-
-	spin_lock_irqsave(&devdata_mutex, flags);
-	old_devdata = rcu_dereference_check(devdata,
-			lockdep_is_held(&devdata_mutex));
-
-	if (old_devdata && old_devdata->vdev != NULL) {
-		dev_err(&viodev->dev, "%s: Attempt to register more than one instance of the hardware\n", __func__);
-		ret = -1;
-		goto error_unlock;
-	}
-
-	dev_set_drvdata(&viodev->dev, NULL);
-
-	new_devdata = kzalloc(sizeof(*new_devdata), GFP_NOFS);
-	if (!new_devdata) {
-		dev_err(&viodev->dev, "%s: Could not allocate memory for device data\n", __func__);
-		ret = -ENOMEM;
-		goto error_unlock;
-	}
-
-	new_devdata->counters = kzalloc(sizeof(*new_devdata->counters),
-			GFP_NOFS);
-	if (!new_devdata->counters) {
-		dev_err(&viodev->dev, "%s: Could not allocate memory for performance counters\n", __func__);
-		ret = -ENOMEM;
-		goto error_unlock;
-	}
-
-	new_devdata->vdev = viodev;
-	new_devdata->dev = &viodev->dev;
-	nx842_OF_set_defaults(new_devdata);
-
-	rcu_assign_pointer(devdata, new_devdata);
-	spin_unlock_irqrestore(&devdata_mutex, flags);
-	synchronize_rcu();
-	kfree(old_devdata);
-
-	of_reconfig_notifier_register(&nx842_of_nb);
-
-	ret = nx842_OF_upd(NULL);
-	if (ret && ret != -ENODEV) {
-		dev_err(&viodev->dev, "could not parse device tree. %d\n", ret);
-		ret = -1;
-		goto error;
-	}
-
-	rcu_read_lock();
-	dev_set_drvdata(&viodev->dev, rcu_dereference(devdata));
-	rcu_read_unlock();
-
-	if (sysfs_create_group(&viodev->dev.kobj, &nx842_attribute_group)) {
-		dev_err(&viodev->dev, "could not create sysfs device attributes\n");
-		ret = -1;
-		goto error;
+	if (!nx842_platform_driver_get()) {
+		pr_err("no nx842 driver found.\n");
+		return -ENODEV;
 	}
 
 	return 0;
-
-error_unlock:
-	spin_unlock_irqrestore(&devdata_mutex, flags);
-	if (new_devdata)
-		kfree(new_devdata->counters);
-	kfree(new_devdata);
-error:
-	return ret;
-}
-
-static int __exit nx842_remove(struct vio_dev *viodev)
-{
-	struct nx842_devdata *old_devdata;
-	unsigned long flags;
-
-	pr_info("Removing IBM Power 842 compression device\n");
-	sysfs_remove_group(&viodev->dev.kobj, &nx842_attribute_group);
-
-	spin_lock_irqsave(&devdata_mutex, flags);
-	old_devdata = rcu_dereference_check(devdata,
-			lockdep_is_held(&devdata_mutex));
-	of_reconfig_notifier_unregister(&nx842_of_nb);
-	RCU_INIT_POINTER(devdata, NULL);
-	spin_unlock_irqrestore(&devdata_mutex, flags);
-	synchronize_rcu();
-	dev_set_drvdata(&viodev->dev, NULL);
-	if (old_devdata)
-		kfree(old_devdata->counters);
-	kfree(old_devdata);
-	return 0;
-}
-
-static struct vio_device_id nx842_driver_ids[] = {
-	{"ibm,compression-v1", "ibm,compression"},
-	{"", ""},
-};
-
-static struct vio_driver nx842_driver = {
-	.name = MODULE_NAME,
-	.probe = nx842_probe,
-	.remove = __exit_p(nx842_remove),
-	.get_desired_dma = nx842_get_desired_dma,
-	.id_table = nx842_driver_ids,
-};
-
-static int __init nx842_init(void)
-{
-	struct nx842_devdata *new_devdata;
-	pr_info("Registering IBM Power 842 compression driver\n");
-
-	RCU_INIT_POINTER(devdata, NULL);
-	new_devdata = kzalloc(sizeof(*new_devdata), GFP_KERNEL);
-	if (!new_devdata) {
-		pr_err("Could not allocate memory for device data\n");
-		return -ENOMEM;
-	}
-	new_devdata->status = UNAVAILABLE;
-	RCU_INIT_POINTER(devdata, new_devdata);
-
-	return vio_register_driver(&nx842_driver);
 }
-
 module_init(nx842_init);
 
 static void __exit nx842_exit(void)
 {
-	struct nx842_devdata *old_devdata;
-	unsigned long flags;
-
-	pr_info("Exiting IBM Power 842 compression driver\n");
-	spin_lock_irqsave(&devdata_mutex, flags);
-	old_devdata = rcu_dereference_check(devdata,
-			lockdep_is_held(&devdata_mutex));
-	RCU_INIT_POINTER(devdata, NULL);
-	spin_unlock_irqrestore(&devdata_mutex, flags);
-	synchronize_rcu();
-	if (old_devdata)
-		dev_set_drvdata(old_devdata->dev, NULL);
-	kfree(old_devdata);
-	vio_unregister_driver(&nx842_driver);
+	nx842_platform_driver_put();
 }
-
 module_exit(nx842_exit);
-
-/*********************************
- * 842 software decompressor
-*********************************/
-typedef int (*sw842_template_op)(const char **, int *, unsigned char **,
-						struct sw842_fifo *);
-
-static int sw842_data8(const char **, int *, unsigned char **,
-						struct sw842_fifo *);
-static int sw842_data4(const char **, int *, unsigned char **,
-						struct sw842_fifo *);
-static int sw842_data2(const char **, int *, unsigned char **,
-						struct sw842_fifo *);
-static int sw842_ptr8(const char **, int *, unsigned char **,
-						struct sw842_fifo *);
-static int sw842_ptr4(const char **, int *, unsigned char **,
-						struct sw842_fifo *);
-static int sw842_ptr2(const char **, int *, unsigned char **,
-						struct sw842_fifo *);
-
-/* special templates */
-#define SW842_TMPL_REPEAT 0x1B
-#define SW842_TMPL_ZEROS 0x1C
-#define SW842_TMPL_EOF 0x1E
-
-static sw842_template_op sw842_tmpl_ops[26][4] = {
-	{ sw842_data8, NULL}, /* 0 (00000) */
-	{ sw842_data4, sw842_data2, sw842_ptr2,  NULL},
-	{ sw842_data4, sw842_ptr2,  sw842_data2, NULL},
-	{ sw842_data4, sw842_ptr2,  sw842_ptr2,  NULL},
-	{ sw842_data4, sw842_ptr4,  NULL},
-	{ sw842_data2, sw842_ptr2,  sw842_data4, NULL},
-	{ sw842_data2, sw842_ptr2,  sw842_data2, sw842_ptr2},
-	{ sw842_data2, sw842_ptr2,  sw842_ptr2,  sw842_data2},
-	{ sw842_data2, sw842_ptr2,  sw842_ptr2,  sw842_ptr2,},
-	{ sw842_data2, sw842_ptr2,  sw842_ptr4,  NULL},
-	{ sw842_ptr2,  sw842_data2, sw842_data4, NULL}, /* 10 (01010) */
-	{ sw842_ptr2,  sw842_data4, sw842_ptr2,  NULL},
-	{ sw842_ptr2,  sw842_data2, sw842_ptr2,  sw842_data2},
-	{ sw842_ptr2,  sw842_data2, sw842_ptr2,  sw842_ptr2},
-	{ sw842_ptr2,  sw842_data2, sw842_ptr4,  NULL},
-	{ sw842_ptr2,  sw842_ptr2,  sw842_data4, NULL},
-	{ sw842_ptr2,  sw842_ptr2,  sw842_data2, sw842_ptr2},
-	{ sw842_ptr2,  sw842_ptr2,  sw842_ptr2,  sw842_data2},
-	{ sw842_ptr2,  sw842_ptr2,  sw842_ptr2,  sw842_ptr2},
-	{ sw842_ptr2,  sw842_ptr2,  sw842_ptr4,  NULL},
-	{ sw842_ptr4,  sw842_data4, NULL}, /* 20 (10100) */
-	{ sw842_ptr4,  sw842_data2, sw842_ptr2,  NULL},
-	{ sw842_ptr4,  sw842_ptr2,  sw842_data2, NULL},
-	{ sw842_ptr4,  sw842_ptr2,  sw842_ptr2,  NULL},
-	{ sw842_ptr4,  sw842_ptr4,  NULL},
-	{ sw842_ptr8,  NULL}
-};
-
-/* Software decompress helpers */
-
-static uint8_t sw842_get_byte(const char *buf, int bit)
-{
-	uint8_t tmpl;
-	uint16_t tmp;
-	tmp = htons(*(uint16_t *)(buf));
-	tmp = (uint16_t)(tmp << bit);
-	tmp = ntohs(tmp);
-	memcpy(&tmpl, &tmp, 1);
-	return tmpl;
-}
-
-static uint8_t sw842_get_template(const char **buf, int *bit)
-{
-	uint8_t byte;
-	byte = sw842_get_byte(*buf, *bit);
-	byte = byte >> 3;
-	byte &= 0x1F;
-	*buf += (*bit + 5) / 8;
-	*bit = (*bit + 5) % 8;
-	return byte;
-}
-
-/* repeat_count happens to be 5-bit too (like the template) */
-static uint8_t sw842_get_repeat_count(const char **buf, int *bit)
-{
-	uint8_t byte;
-	byte = sw842_get_byte(*buf, *bit);
-	byte = byte >> 2;
-	byte &= 0x3F;
-	*buf += (*bit + 6) / 8;
-	*bit = (*bit + 6) % 8;
-	return byte;
-}
-
-static uint8_t sw842_get_ptr2(const char **buf, int *bit)
-{
-	uint8_t ptr;
-	ptr = sw842_get_byte(*buf, *bit);
-	(*buf)++;
-	return ptr;
-}
-
-static uint16_t sw842_get_ptr4(const char **buf, int *bit,
-		struct sw842_fifo *fifo)
-{
-	uint16_t ptr;
-	ptr = htons(*(uint16_t *)(*buf));
-	ptr = (uint16_t)(ptr << *bit);
-	ptr = ptr >> 7;
-	ptr &= 0x01FF;
-	*buf += (*bit + 9) / 8;
-	*bit = (*bit + 9) % 8;
-	return ptr;
-}
-
-static uint8_t sw842_get_ptr8(const char **buf, int *bit,
-		struct sw842_fifo *fifo)
-{
-	return sw842_get_ptr2(buf, bit);
-}
-
-/* Software decompress template ops */
-
-static int sw842_data8(const char **inbuf, int *inbit,
-		unsigned char **outbuf, struct sw842_fifo *fifo)
-{
-	int ret;
-
-	ret = sw842_data4(inbuf, inbit, outbuf, fifo);
-	if (ret)
-		return ret;
-	ret = sw842_data4(inbuf, inbit, outbuf, fifo);
-	return ret;
-}
-
-static int sw842_data4(const char **inbuf, int *inbit,
-		unsigned char **outbuf, struct sw842_fifo *fifo)
-{
-	int ret;
-
-	ret = sw842_data2(inbuf, inbit, outbuf, fifo);
-	if (ret)
-		return ret;
-	ret = sw842_data2(inbuf, inbit, outbuf, fifo);
-	return ret;
-}
-
-static int sw842_data2(const char **inbuf, int *inbit,
-		unsigned char **outbuf, struct sw842_fifo *fifo)
-{
-	**outbuf = sw842_get_byte(*inbuf, *inbit);
-	(*inbuf)++;
-	(*outbuf)++;
-	**outbuf = sw842_get_byte(*inbuf, *inbit);
-	(*inbuf)++;
-	(*outbuf)++;
-	return 0;
-}
-
-static int sw842_ptr8(const char **inbuf, int *inbit,
-		unsigned char **outbuf, struct sw842_fifo *fifo)
-{
-	uint8_t ptr;
-	ptr = sw842_get_ptr8(inbuf, inbit, fifo);
-	if (!fifo->f84_full && (ptr >= fifo->f8_count))
-		return 1;
-	memcpy(*outbuf, fifo->f8[ptr], 8);
-	*outbuf += 8;
-	return 0;
-}
-
-static int sw842_ptr4(const char **inbuf, int *inbit,
-		unsigned char **outbuf, struct sw842_fifo *fifo)
-{
-	uint16_t ptr;
-	ptr = sw842_get_ptr4(inbuf, inbit, fifo);
-	if (!fifo->f84_full && (ptr >= fifo->f4_count))
-		return 1;
-	memcpy(*outbuf, fifo->f4[ptr], 4);
-	*outbuf += 4;
-	return 0;
-}
-
-static int sw842_ptr2(const char **inbuf, int *inbit,
-		unsigned char **outbuf, struct sw842_fifo *fifo)
-{
-	uint8_t ptr;
-	ptr = sw842_get_ptr2(inbuf, inbit);
-	if (!fifo->f2_full && (ptr >= fifo->f2_count))
-		return 1;
-	memcpy(*outbuf, fifo->f2[ptr], 2);
-	*outbuf += 2;
-	return 0;
-}
-
-static void sw842_copy_to_fifo(const char *buf, struct sw842_fifo *fifo)
-{
-	unsigned char initial_f2count = fifo->f2_count;
-
-	memcpy(fifo->f8[fifo->f8_count], buf, 8);
-	fifo->f4_count += 2;
-	fifo->f8_count += 1;
-
-	if (!fifo->f84_full && fifo->f4_count >= 512) {
-		fifo->f84_full = 1;
-		fifo->f4_count /= 512;
-	}
-
-	memcpy(fifo->f2[fifo->f2_count++], buf, 2);
-	memcpy(fifo->f2[fifo->f2_count++], buf + 2, 2);
-	memcpy(fifo->f2[fifo->f2_count++], buf + 4, 2);
-	memcpy(fifo->f2[fifo->f2_count++], buf + 6, 2);
-	if (fifo->f2_count < initial_f2count)
-		fifo->f2_full = 1;
-}
-
-static int sw842_decompress(const unsigned char *src, int srclen,
-			unsigned char *dst, int *destlen,
-			const void *wrkmem)
-{
-	uint8_t tmpl;
-	const char *inbuf;
-	int inbit = 0;
-	unsigned char *outbuf, *outbuf_end, *origbuf, *prevbuf;
-	const char *inbuf_end;
-	sw842_template_op op;
-	int opindex;
-	int i, repeat_count;
-	struct sw842_fifo *fifo;
-	int ret = 0;
-
-	fifo = &((struct nx842_workmem *)(wrkmem))->swfifo;
-	memset(fifo, 0, sizeof(*fifo));
-
-	origbuf = NULL;
-	inbuf = src;
-	inbuf_end = src + srclen;
-	outbuf = dst;
-	outbuf_end = dst + *destlen;
-
-	while ((tmpl = sw842_get_template(&inbuf, &inbit)) != SW842_TMPL_EOF) {
-		if (inbuf >= inbuf_end) {
-			ret = -EINVAL;
-			goto out;
-		}
-
-		opindex = 0;
-		prevbuf = origbuf;
-		origbuf = outbuf;
-		switch (tmpl) {
-		case SW842_TMPL_REPEAT:
-			if (prevbuf == NULL) {
-				ret = -EINVAL;
-				goto out;
-			}
-
-			repeat_count = sw842_get_repeat_count(&inbuf,
-								&inbit) + 1;
-
-			/* Did the repeat count advance past the end of input */
-			if (inbuf > inbuf_end) {
-				ret = -EINVAL;
-				goto out;
-			}
-
-			for (i = 0; i < repeat_count; i++) {
-				/* Would this overflow the output buffer */
-				if ((outbuf + 8) > outbuf_end) {
-					ret = -ENOSPC;
-					goto out;
-				}
-
-				memcpy(outbuf, prevbuf, 8);
-				sw842_copy_to_fifo(outbuf, fifo);
-				outbuf += 8;
-			}
-			break;
-
-		case SW842_TMPL_ZEROS:
-			/* Would this overflow the output buffer */
-			if ((outbuf + 8) > outbuf_end) {
-				ret = -ENOSPC;
-				goto out;
-			}
-
-			memset(outbuf, 0, 8);
-			sw842_copy_to_fifo(outbuf, fifo);
-			outbuf += 8;
-			break;
-
-		default:
-			if (tmpl > 25) {
-				ret = -EINVAL;
-				goto out;
-			}
-
-			/* Does this go past the end of the input buffer */
-			if ((inbuf + 2) > inbuf_end) {
-				ret = -EINVAL;
-				goto out;
-			}
-
-			/* Would this overflow the output buffer */
-			if ((outbuf + 8) > outbuf_end) {
-				ret = -ENOSPC;
-				goto out;
-			}
-
-			while (opindex < 4 &&
-				(op = sw842_tmpl_ops[tmpl][opindex++])
-					!= NULL) {
-				ret = (*op)(&inbuf, &inbit, &outbuf, fifo);
-				if (ret) {
-					ret = -EINVAL;
-					goto out;
-				}
-				sw842_copy_to_fifo(origbuf, fifo);
-			}
-		}
-	}
-
-out:
-	if (!ret)
-		*destlen = (unsigned int)(outbuf - dst);
-	else
-		*destlen = 0;
-
-	return ret;
-}
diff --git a/drivers/crypto/nx/nx-842.h b/drivers/crypto/nx/nx-842.h
new file mode 100644
index 000000000..ac0ea79d0
--- /dev/null
+++ b/drivers/crypto/nx/nx-842.h
@@ -0,0 +1,144 @@
+
+#ifndef __NX_842_H__
+#define __NX_842_H__
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/sw842.h>
+#include <linux/of.h>
+#include <linux/slab.h>
+#include <linux/io.h>
+#include <linux/mm.h>
+#include <linux/ratelimit.h>
+
+/* Restrictions on Data Descriptor List (DDL) and Entry (DDE) buffers
+ *
+ * From NX P8 workbook, sec 4.9.1 "842 details"
+ *   Each DDE buffer is 128 byte aligned
+ *   Each DDE buffer size is a multiple of 32 bytes (except the last)
+ *   The last DDE buffer size is a multiple of 8 bytes
+ */
+#define DDE_BUFFER_ALIGN	(128)
+#define DDE_BUFFER_SIZE_MULT	(32)
+#define DDE_BUFFER_LAST_MULT	(8)
+
+/* Arbitrary DDL length limit
+ * Allows max buffer size of MAX-1 to MAX pages
+ * (depending on alignment)
+ */
+#define DDL_LEN_MAX		(17)
+
+/* CCW 842 CI/FC masks
+ * NX P8 workbook, section 4.3.1, figure 4-6
+ * "CI/FC Boundary by NX CT type"
+ */
+#define CCW_CI_842		(0x00003ff8)
+#define CCW_FC_842		(0x00000007)
+
+/* CCW Function Codes (FC) for 842
+ * NX P8 workbook, section 4.9, table 4-28
+ * "Function Code Definitions for 842 Memory Compression"
+ */
+#define CCW_FC_842_COMP_NOCRC	(0)
+#define CCW_FC_842_COMP_CRC	(1)
+#define CCW_FC_842_DECOMP_NOCRC	(2)
+#define CCW_FC_842_DECOMP_CRC	(3)
+#define CCW_FC_842_MOVE		(4)
+
+/* CSB CC Error Types for 842
+ * NX P8 workbook, section 4.10.3, table 4-30
+ * "Reported Error Types Summary Table"
+ */
+/* These are all duplicates of existing codes defined in icswx.h. */
+#define CSB_CC_TRANSLATION_DUP1	(80)
+#define CSB_CC_TRANSLATION_DUP2	(82)
+#define CSB_CC_TRANSLATION_DUP3	(84)
+#define CSB_CC_TRANSLATION_DUP4	(86)
+#define CSB_CC_TRANSLATION_DUP5	(92)
+#define CSB_CC_TRANSLATION_DUP6	(94)
+#define CSB_CC_PROTECTION_DUP1	(81)
+#define CSB_CC_PROTECTION_DUP2	(83)
+#define CSB_CC_PROTECTION_DUP3	(85)
+#define CSB_CC_PROTECTION_DUP4	(87)
+#define CSB_CC_PROTECTION_DUP5	(93)
+#define CSB_CC_PROTECTION_DUP6	(95)
+#define CSB_CC_RD_EXTERNAL_DUP1	(89)
+#define CSB_CC_RD_EXTERNAL_DUP2	(90)
+#define CSB_CC_RD_EXTERNAL_DUP3	(91)
+/* These are specific to NX */
+/* 842 codes */
+#define CSB_CC_TPBC_GT_SPBC	(64) /* no error, but >1 comp ratio */
+#define CSB_CC_CRC_MISMATCH	(65) /* decomp crc mismatch */
+#define CSB_CC_TEMPL_INVALID	(66) /* decomp invalid template value */
+#define CSB_CC_TEMPL_OVERFLOW	(67) /* decomp template shows data after end */
+/* sym crypt codes */
+#define CSB_CC_DECRYPT_OVERFLOW	(64)
+/* asym crypt codes */
+#define CSB_CC_MINV_OVERFLOW	(128)
+/* These are reserved for hypervisor use */
+#define CSB_CC_HYP_RESERVE_START	(240)
+#define CSB_CC_HYP_RESERVE_END		(253)
+#define CSB_CC_HYP_NO_HW		(254)
+#define CSB_CC_HYP_HANG_ABORTED		(255)
+
+/* CCB Completion Modes (CM) for 842
+ * NX P8 workbook, section 4.3, figure 4-5
+ * "CRB Details - Normal Cop_Req (CL=00, C=1)"
+ */
+#define CCB_CM_EXTRA_WRITE	(CCB_CM0_ALL_COMPLETIONS & CCB_CM12_STORE)
+#define CCB_CM_INTERRUPT	(CCB_CM0_ALL_COMPLETIONS & CCB_CM12_INTERRUPT)
+
+#define LEN_ON_SIZE(pa, size)	((size) - ((pa) & ((size) - 1)))
+#define LEN_ON_PAGE(pa)		LEN_ON_SIZE(pa, PAGE_SIZE)
+
+static inline unsigned long nx842_get_pa(void *addr)
+{
+	if (!is_vmalloc_addr(addr))
+		return __pa(addr);
+
+	return page_to_phys(vmalloc_to_page(addr)) + offset_in_page(addr);
+}
+
+/* Get/Set bit fields */
+#define MASK_LSH(m)		(__builtin_ffsl(m) - 1)
+#define GET_FIELD(v, m)		(((v) & (m)) >> MASK_LSH(m))
+#define SET_FIELD(v, m, val)	(((v) & ~(m)) | (((val) << MASK_LSH(m)) & (m)))
+
+struct nx842_constraints {
+	int alignment;
+	int multiple;
+	int minimum;
+	int maximum;
+};
+
+struct nx842_driver {
+	char *name;
+	struct module *owner;
+	size_t workmem_size;
+
+	struct nx842_constraints *constraints;
+
+	int (*compress)(const unsigned char *in, unsigned int in_len,
+			unsigned char *out, unsigned int *out_len,
+			void *wrkmem);
+	int (*decompress)(const unsigned char *in, unsigned int in_len,
+			  unsigned char *out, unsigned int *out_len,
+			  void *wrkmem);
+};
+
+struct nx842_driver *nx842_platform_driver(void);
+bool nx842_platform_driver_set(struct nx842_driver *driver);
+void nx842_platform_driver_unset(struct nx842_driver *driver);
+bool nx842_platform_driver_get(void);
+void nx842_platform_driver_put(void);
+
+size_t nx842_workmem_size(void);
+
+int nx842_constraints(struct nx842_constraints *constraints);
+
+int nx842_compress(const unsigned char *in, unsigned int in_len,
+		   unsigned char *out, unsigned int *out_len, void *wrkmem);
+int nx842_decompress(const unsigned char *in, unsigned int in_len,
+		     unsigned char *out, unsigned int *out_len, void *wrkmem);
+
+#endif /* __NX_842_H__ */
diff --git a/drivers/crypto/nx/nx-aes-gcm.c b/drivers/crypto/nx/nx-aes-gcm.c
index c6ebeb644..92c993f08 100644
--- a/drivers/crypto/nx/nx-aes-gcm.c
+++ b/drivers/crypto/nx/nx-aes-gcm.c
@@ -93,17 +93,6 @@ out:
 	return rc;
 }
 
-static int gcm_aes_nx_setauthsize(struct crypto_aead *tfm,
-				  unsigned int authsize)
-{
-	if (authsize > crypto_aead_alg(tfm)->maxauthsize)
-		return -EINVAL;
-
-	crypto_aead_crt(tfm)->authsize = authsize;
-
-	return 0;
-}
-
 static int gcm4106_aes_nx_setauthsize(struct crypto_aead *tfm,
 				      unsigned int authsize)
 {
@@ -116,8 +105,6 @@ static int gcm4106_aes_nx_setauthsize(struct crypto_aead *tfm,
 		return -EINVAL;
 	}
 
-	crypto_aead_crt(tfm)->authsize = authsize;
-
 	return 0;
 }
 
@@ -134,7 +121,7 @@ static int nx_gca(struct nx_crypto_ctx  *nx_ctx,
 	unsigned int max_sg_len;
 
 	if (nbytes <= AES_BLOCK_SIZE) {
-		scatterwalk_start(&walk, req->assoc);
+		scatterwalk_start(&walk, req->src);
 		scatterwalk_copychunks(out, &walk, nbytes, SCATTERWALK_FROM_SG);
 		scatterwalk_done(&walk, SCATTERWALK_FROM_SG, 0);
 		return 0;
@@ -159,7 +146,7 @@ static int nx_gca(struct nx_crypto_ctx  *nx_ctx,
 				   NX_PAGE_SIZE * (max_sg_len - 1));
 
 		nx_sg = nx_walk_and_build(nx_ctx->in_sg, max_sg_len,
-					  req->assoc, processed, &to_process);
+					  req->src, processed, &to_process);
 
 		if ((to_process + processed) < nbytes)
 			NX_CPB_FDM(csbcpb_aead) |= NX_FDM_INTERMEDIATE;
@@ -225,7 +212,7 @@ static int gmac(struct aead_request *req, struct blkcipher_desc *desc)
 				   NX_PAGE_SIZE * (max_sg_len - 1));
 
 		nx_sg = nx_walk_and_build(nx_ctx->in_sg, max_sg_len,
-					  req->assoc, processed, &to_process);
+					  req->src, processed, &to_process);
 
 		if ((to_process + processed) < nbytes)
 			NX_CPB_FDM(csbcpb) |= NX_FDM_INTERMEDIATE;
@@ -378,7 +365,8 @@ static int gcm_aes_nx_crypt(struct aead_request *req, int enc)
 		csbcpb->cpb.aes_gcm.bit_length_data = nbytes * 8;
 		desc.tfm = (struct crypto_blkcipher *) req->base.tfm;
 		rc = nx_build_sg_lists(nx_ctx, &desc, req->dst,
-				       req->src, &to_process, processed,
+				       req->src, &to_process,
+				       processed + req->assoclen,
 				       csbcpb->cpb.aes_gcm.iv_or_cnt);
 
 		if (rc)
@@ -413,17 +401,19 @@ static int gcm_aes_nx_crypt(struct aead_request *req, int enc)
 mac:
 	if (enc) {
 		/* copy out the auth tag */
-		scatterwalk_map_and_copy(csbcpb->cpb.aes_gcm.out_pat_or_mac,
-				 req->dst, nbytes,
-				 crypto_aead_authsize(crypto_aead_reqtfm(req)),
-				 SCATTERWALK_TO_SG);
+		scatterwalk_map_and_copy(
+			csbcpb->cpb.aes_gcm.out_pat_or_mac,
+			req->dst, req->assoclen + nbytes,
+			crypto_aead_authsize(crypto_aead_reqtfm(req)),
+			SCATTERWALK_TO_SG);
 	} else {
 		u8 *itag = nx_ctx->priv.gcm.iauth_tag;
 		u8 *otag = csbcpb->cpb.aes_gcm.out_pat_or_mac;
 
-		scatterwalk_map_and_copy(itag, req->src, nbytes,
-				 crypto_aead_authsize(crypto_aead_reqtfm(req)),
-				 SCATTERWALK_FROM_SG);
+		scatterwalk_map_and_copy(
+			itag, req->src, req->assoclen + nbytes,
+			crypto_aead_authsize(crypto_aead_reqtfm(req)),
+			SCATTERWALK_FROM_SG);
 		rc = memcmp(itag, otag,
 			    crypto_aead_authsize(crypto_aead_reqtfm(req))) ?
 		     -EBADMSG : 0;
@@ -484,45 +474,39 @@ static int gcm4106_aes_nx_decrypt(struct aead_request *req)
  * during encrypt/decrypt doesn't solve this problem, because it calls
  * blkcipher_walk_done under the covers, which doesn't use walk->blocksize,
  * but instead uses this tfm->blocksize. */
-struct crypto_alg nx_gcm_aes_alg = {
-	.cra_name        = "gcm(aes)",
-	.cra_driver_name = "gcm-aes-nx",
-	.cra_priority    = 300,
-	.cra_flags       = CRYPTO_ALG_TYPE_AEAD,
-	.cra_blocksize   = 1,
-	.cra_ctxsize     = sizeof(struct nx_crypto_ctx),
-	.cra_type        = &crypto_aead_type,
-	.cra_module      = THIS_MODULE,
-	.cra_init        = nx_crypto_ctx_aes_gcm_init,
-	.cra_exit        = nx_crypto_ctx_exit,
-	.cra_aead = {
-		.ivsize      = AES_BLOCK_SIZE,
-		.maxauthsize = AES_BLOCK_SIZE,
-		.setkey      = gcm_aes_nx_set_key,
-		.setauthsize = gcm_aes_nx_setauthsize,
-		.encrypt     = gcm_aes_nx_encrypt,
-		.decrypt     = gcm_aes_nx_decrypt,
-	}
+struct aead_alg nx_gcm_aes_alg = {
+	.base = {
+		.cra_name        = "gcm(aes)",
+		.cra_driver_name = "gcm-aes-nx",
+		.cra_priority    = 300,
+		.cra_blocksize   = 1,
+		.cra_ctxsize     = sizeof(struct nx_crypto_ctx),
+		.cra_module      = THIS_MODULE,
+	},
+	.init        = nx_crypto_ctx_aes_gcm_init,
+	.exit        = nx_crypto_ctx_aead_exit,
+	.ivsize      = 12,
+	.maxauthsize = AES_BLOCK_SIZE,
+	.setkey      = gcm_aes_nx_set_key,
+	.encrypt     = gcm_aes_nx_encrypt,
+	.decrypt     = gcm_aes_nx_decrypt,
 };
 
-struct crypto_alg nx_gcm4106_aes_alg = {
-	.cra_name        = "rfc4106(gcm(aes))",
-	.cra_driver_name = "rfc4106-gcm-aes-nx",
-	.cra_priority    = 300,
-	.cra_flags       = CRYPTO_ALG_TYPE_AEAD,
-	.cra_blocksize   = 1,
-	.cra_ctxsize     = sizeof(struct nx_crypto_ctx),
-	.cra_type        = &crypto_nivaead_type,
-	.cra_module      = THIS_MODULE,
-	.cra_init        = nx_crypto_ctx_aes_gcm_init,
-	.cra_exit        = nx_crypto_ctx_exit,
-	.cra_aead = {
-		.ivsize      = 8,
-		.maxauthsize = AES_BLOCK_SIZE,
-		.geniv       = "seqiv",
-		.setkey      = gcm4106_aes_nx_set_key,
-		.setauthsize = gcm4106_aes_nx_setauthsize,
-		.encrypt     = gcm4106_aes_nx_encrypt,
-		.decrypt     = gcm4106_aes_nx_decrypt,
-	}
+struct aead_alg nx_gcm4106_aes_alg = {
+	.base = {
+		.cra_name        = "rfc4106(gcm(aes))",
+		.cra_driver_name = "rfc4106-gcm-aes-nx",
+		.cra_priority    = 300,
+		.cra_blocksize   = 1,
+		.cra_ctxsize     = sizeof(struct nx_crypto_ctx),
+		.cra_module      = THIS_MODULE,
+	},
+	.init        = nx_crypto_ctx_aes_gcm_init,
+	.exit        = nx_crypto_ctx_aead_exit,
+	.ivsize      = 8,
+	.maxauthsize = AES_BLOCK_SIZE,
+	.setkey      = gcm4106_aes_nx_set_key,
+	.setauthsize = gcm4106_aes_nx_setauthsize,
+	.encrypt     = gcm4106_aes_nx_encrypt,
+	.decrypt     = gcm4106_aes_nx_decrypt,
 };
diff --git a/drivers/crypto/nx/nx-sha256.c b/drivers/crypto/nx/nx-sha256.c
index 08f8d5cd6..becb738c8 100644
--- a/drivers/crypto/nx/nx-sha256.c
+++ b/drivers/crypto/nx/nx-sha256.c
@@ -71,7 +71,6 @@ static int nx_sha256_update(struct shash_desc *desc, const u8 *data,
 	struct sha256_state *sctx = shash_desc_ctx(desc);
 	struct nx_crypto_ctx *nx_ctx = crypto_tfm_ctx(&desc->tfm->base);
 	struct nx_csbcpb *csbcpb = (struct nx_csbcpb *)nx_ctx->csbcpb;
-	struct nx_sg *in_sg;
 	struct nx_sg *out_sg;
 	u64 to_process = 0, leftover, total;
 	unsigned long irq_flags;
@@ -97,7 +96,6 @@ static int nx_sha256_update(struct shash_desc *desc, const u8 *data,
 	NX_CPB_FDM(csbcpb) |= NX_FDM_INTERMEDIATE;
 	NX_CPB_FDM(csbcpb) |= NX_FDM_CONTINUATION;
 
-	in_sg = nx_ctx->in_sg;
 	max_sg_len = min_t(u64, nx_ctx->ap->sglen,
 			nx_driver.of.max_sg_len/sizeof(struct nx_sg));
 	max_sg_len = min_t(u64, max_sg_len,
@@ -114,17 +112,12 @@ static int nx_sha256_update(struct shash_desc *desc, const u8 *data,
 	}
 
 	do {
-		/*
-		 * to_process: the SHA256_BLOCK_SIZE data chunk to process in
-		 * this update. This value is also restricted by the sg list
-		 * limits.
-		 */
-		to_process = total - to_process;
-		to_process = to_process & ~(SHA256_BLOCK_SIZE - 1);
+		int used_sgs = 0;
+		struct nx_sg *in_sg = nx_ctx->in_sg;
 
 		if (buf_len) {
 			data_len = buf_len;
-			in_sg = nx_build_sg_list(nx_ctx->in_sg,
+			in_sg = nx_build_sg_list(in_sg,
 						 (u8 *) sctx->buf,
 						 &data_len,
 						 max_sg_len);
@@ -133,15 +126,27 @@ static int nx_sha256_update(struct shash_desc *desc, const u8 *data,
 				rc = -EINVAL;
 				goto out;
 			}
+			used_sgs = in_sg - nx_ctx->in_sg;
 		}
 
+		/* to_process: SHA256_BLOCK_SIZE aligned chunk to be
+		 * processed in this iteration. This value is restricted
+		 * by sg list limits and number of sgs we already used
+		 * for leftover data. (see above)
+		 * In ideal case, we could allow NX_PAGE_SIZE * max_sg_len,
+		 * but because data may not be aligned, we need to account
+		 * for that too. */
+		to_process = min_t(u64, total,
+			(max_sg_len - 1 - used_sgs) * NX_PAGE_SIZE);
+		to_process = to_process & ~(SHA256_BLOCK_SIZE - 1);
+
 		data_len = to_process - buf_len;
 		in_sg = nx_build_sg_list(in_sg, (u8 *) data,
 					 &data_len, max_sg_len);
 
 		nx_ctx->op.inlen = (nx_ctx->in_sg - in_sg) * sizeof(struct nx_sg);
 
-		to_process = (data_len + buf_len);
+		to_process = data_len + buf_len;
 		leftover = total - to_process;
 
 		/*
diff --git a/drivers/crypto/nx/nx-sha512.c b/drivers/crypto/nx/nx-sha512.c
index aff0fe58e..b6e183d58 100644
--- a/drivers/crypto/nx/nx-sha512.c
+++ b/drivers/crypto/nx/nx-sha512.c
@@ -71,7 +71,6 @@ static int nx_sha512_update(struct shash_desc *desc, const u8 *data,
 	struct sha512_state *sctx = shash_desc_ctx(desc);
 	struct nx_crypto_ctx *nx_ctx = crypto_tfm_ctx(&desc->tfm->base);
 	struct nx_csbcpb *csbcpb = (struct nx_csbcpb *)nx_ctx->csbcpb;
-	struct nx_sg *in_sg;
 	struct nx_sg *out_sg;
 	u64 to_process, leftover = 0, total;
 	unsigned long irq_flags;
@@ -97,7 +96,6 @@ static int nx_sha512_update(struct shash_desc *desc, const u8 *data,
 	NX_CPB_FDM(csbcpb) |= NX_FDM_INTERMEDIATE;
 	NX_CPB_FDM(csbcpb) |= NX_FDM_CONTINUATION;
 
-	in_sg = nx_ctx->in_sg;
 	max_sg_len = min_t(u64, nx_ctx->ap->sglen,
 			nx_driver.of.max_sg_len/sizeof(struct nx_sg));
 	max_sg_len = min_t(u64, max_sg_len,
@@ -114,18 +112,12 @@ static int nx_sha512_update(struct shash_desc *desc, const u8 *data,
 	}
 
 	do {
-		/*
-		 * to_process: the SHA512_BLOCK_SIZE data chunk to process in
-		 * this update. This value is also restricted by the sg list
-		 * limits.
-		 */
-		to_process = total - leftover;
-		to_process = to_process & ~(SHA512_BLOCK_SIZE - 1);
-		leftover = total - to_process;
+		int used_sgs = 0;
+		struct nx_sg *in_sg = nx_ctx->in_sg;
 
 		if (buf_len) {
 			data_len = buf_len;
-			in_sg = nx_build_sg_list(nx_ctx->in_sg,
+			in_sg = nx_build_sg_list(in_sg,
 						 (u8 *) sctx->buf,
 						 &data_len, max_sg_len);
 
@@ -133,8 +125,20 @@ static int nx_sha512_update(struct shash_desc *desc, const u8 *data,
 				rc = -EINVAL;
 				goto out;
 			}
+			used_sgs = in_sg - nx_ctx->in_sg;
 		}
 
+		/* to_process: SHA512_BLOCK_SIZE aligned chunk to be
+		 * processed in this iteration. This value is restricted
+		 * by sg list limits and number of sgs we already used
+		 * for leftover data. (see above)
+		 * In ideal case, we could allow NX_PAGE_SIZE * max_sg_len,
+		 * but because data may not be aligned, we need to account
+		 * for that too. */
+		to_process = min_t(u64, total,
+			(max_sg_len - 1 - used_sgs) * NX_PAGE_SIZE);
+		to_process = to_process & ~(SHA512_BLOCK_SIZE - 1);
+
 		data_len = to_process - buf_len;
 		in_sg = nx_build_sg_list(in_sg, (u8 *) data,
 					 &data_len, max_sg_len);
@@ -146,7 +150,7 @@ static int nx_sha512_update(struct shash_desc *desc, const u8 *data,
 			goto out;
 		}
 
-		to_process = (data_len + buf_len);
+		to_process = data_len + buf_len;
 		leftover = total - to_process;
 
 		/*
diff --git a/drivers/crypto/nx/nx.c b/drivers/crypto/nx/nx.c
index 737d33dc5..436971343 100644
--- a/drivers/crypto/nx/nx.c
+++ b/drivers/crypto/nx/nx.c
@@ -19,8 +19,8 @@
  * Author: Kent Yoder <yoder1@us.ibm.com>
  */
 
+#include <crypto/internal/aead.h>
 #include <crypto/internal/hash.h>
-#include <crypto/hash.h>
 #include <crypto/aes.h>
 #include <crypto/sha.h>
 #include <crypto/algapi.h>
@@ -29,10 +29,10 @@
 #include <linux/moduleparam.h>
 #include <linux/types.h>
 #include <linux/mm.h>
-#include <linux/crypto.h>
 #include <linux/scatterlist.h>
 #include <linux/device.h>
 #include <linux/of.h>
+#include <linux/types.h>
 #include <asm/hvcall.h>
 #include <asm/vio.h>
 
@@ -399,6 +399,13 @@ static void nx_of_update_msc(struct device   *dev,
 				goto next_loop;
 			}
 
+			if (!trip->sglen || trip->databytelen < NX_PAGE_SIZE) {
+				dev_warn(dev, "bogus sglen/databytelen: "
+					 "%u/%u (ignored)\n", trip->sglen,
+					 trip->databytelen);
+				goto next_loop;
+			}
+
 			switch (trip->keybitlen) {
 			case 128:
 			case 160:
@@ -491,6 +498,72 @@ static void nx_of_init(struct device *dev, struct nx_of *props)
 		nx_of_update_msc(dev, p, props);
 }
 
+static bool nx_check_prop(struct device *dev, u32 fc, u32 mode, int slot)
+{
+	struct alg_props *props = &nx_driver.of.ap[fc][mode][slot];
+
+	if (!props->sglen || props->databytelen < NX_PAGE_SIZE) {
+		if (dev)
+			dev_warn(dev, "bogus sglen/databytelen for %u/%u/%u: "
+				 "%u/%u (ignored)\n", fc, mode, slot,
+				 props->sglen, props->databytelen);
+		return false;
+	}
+
+	return true;
+}
+
+static bool nx_check_props(struct device *dev, u32 fc, u32 mode)
+{
+	int i;
+
+	for (i = 0; i < 3; i++)
+		if (!nx_check_prop(dev, fc, mode, i))
+			return false;
+
+	return true;
+}
+
+static int nx_register_alg(struct crypto_alg *alg, u32 fc, u32 mode)
+{
+	return nx_check_props(&nx_driver.viodev->dev, fc, mode) ?
+	       crypto_register_alg(alg) : 0;
+}
+
+static int nx_register_aead(struct aead_alg *alg, u32 fc, u32 mode)
+{
+	return nx_check_props(&nx_driver.viodev->dev, fc, mode) ?
+	       crypto_register_aead(alg) : 0;
+}
+
+static int nx_register_shash(struct shash_alg *alg, u32 fc, u32 mode, int slot)
+{
+	return (slot >= 0 ? nx_check_prop(&nx_driver.viodev->dev,
+					  fc, mode, slot) :
+			    nx_check_props(&nx_driver.viodev->dev, fc, mode)) ?
+	       crypto_register_shash(alg) : 0;
+}
+
+static void nx_unregister_alg(struct crypto_alg *alg, u32 fc, u32 mode)
+{
+	if (nx_check_props(NULL, fc, mode))
+		crypto_unregister_alg(alg);
+}
+
+static void nx_unregister_aead(struct aead_alg *alg, u32 fc, u32 mode)
+{
+	if (nx_check_props(NULL, fc, mode))
+		crypto_unregister_aead(alg);
+}
+
+static void nx_unregister_shash(struct shash_alg *alg, u32 fc, u32 mode,
+				int slot)
+{
+	if (slot >= 0 ? nx_check_prop(NULL, fc, mode, slot) :
+			nx_check_props(NULL, fc, mode))
+		crypto_unregister_shash(alg);
+}
+
 /**
  * nx_register_algs - register algorithms with the crypto API
  *
@@ -515,72 +588,77 @@ static int nx_register_algs(void)
 
 	nx_driver.of.status = NX_OKAY;
 
-	rc = crypto_register_alg(&nx_ecb_aes_alg);
+	rc = nx_register_alg(&nx_ecb_aes_alg, NX_FC_AES, NX_MODE_AES_ECB);
 	if (rc)
 		goto out;
 
-	rc = crypto_register_alg(&nx_cbc_aes_alg);
+	rc = nx_register_alg(&nx_cbc_aes_alg, NX_FC_AES, NX_MODE_AES_CBC);
 	if (rc)
 		goto out_unreg_ecb;
 
-	rc = crypto_register_alg(&nx_ctr_aes_alg);
+	rc = nx_register_alg(&nx_ctr_aes_alg, NX_FC_AES, NX_MODE_AES_CTR);
 	if (rc)
 		goto out_unreg_cbc;
 
-	rc = crypto_register_alg(&nx_ctr3686_aes_alg);
+	rc = nx_register_alg(&nx_ctr3686_aes_alg, NX_FC_AES, NX_MODE_AES_CTR);
 	if (rc)
 		goto out_unreg_ctr;
 
-	rc = crypto_register_alg(&nx_gcm_aes_alg);
+	rc = nx_register_aead(&nx_gcm_aes_alg, NX_FC_AES, NX_MODE_AES_GCM);
 	if (rc)
 		goto out_unreg_ctr3686;
 
-	rc = crypto_register_alg(&nx_gcm4106_aes_alg);
+	rc = nx_register_aead(&nx_gcm4106_aes_alg, NX_FC_AES, NX_MODE_AES_GCM);
 	if (rc)
 		goto out_unreg_gcm;
 
-	rc = crypto_register_alg(&nx_ccm_aes_alg);
+	rc = nx_register_alg(&nx_ccm_aes_alg, NX_FC_AES, NX_MODE_AES_CCM);
 	if (rc)
 		goto out_unreg_gcm4106;
 
-	rc = crypto_register_alg(&nx_ccm4309_aes_alg);
+	rc = nx_register_alg(&nx_ccm4309_aes_alg, NX_FC_AES, NX_MODE_AES_CCM);
 	if (rc)
 		goto out_unreg_ccm;
 
-	rc = crypto_register_shash(&nx_shash_sha256_alg);
+	rc = nx_register_shash(&nx_shash_sha256_alg, NX_FC_SHA, NX_MODE_SHA,
+			       NX_PROPS_SHA256);
 	if (rc)
 		goto out_unreg_ccm4309;
 
-	rc = crypto_register_shash(&nx_shash_sha512_alg);
+	rc = nx_register_shash(&nx_shash_sha512_alg, NX_FC_SHA, NX_MODE_SHA,
+			       NX_PROPS_SHA512);
 	if (rc)
 		goto out_unreg_s256;
 
-	rc = crypto_register_shash(&nx_shash_aes_xcbc_alg);
+	rc = nx_register_shash(&nx_shash_aes_xcbc_alg,
+			       NX_FC_AES, NX_MODE_AES_XCBC_MAC, -1);
 	if (rc)
 		goto out_unreg_s512;
 
 	goto out;
 
 out_unreg_s512:
-	crypto_unregister_shash(&nx_shash_sha512_alg);
+	nx_unregister_shash(&nx_shash_sha512_alg, NX_FC_SHA, NX_MODE_SHA,
+			    NX_PROPS_SHA512);
 out_unreg_s256:
-	crypto_unregister_shash(&nx_shash_sha256_alg);
+	nx_unregister_shash(&nx_shash_sha256_alg, NX_FC_SHA, NX_MODE_SHA,
+			    NX_PROPS_SHA256);
 out_unreg_ccm4309:
-	crypto_unregister_alg(&nx_ccm4309_aes_alg);
+	nx_unregister_alg(&nx_ccm4309_aes_alg, NX_FC_AES, NX_MODE_AES_CCM);
 out_unreg_ccm:
-	crypto_unregister_alg(&nx_ccm_aes_alg);
+	nx_unregister_alg(&nx_ccm_aes_alg, NX_FC_AES, NX_MODE_AES_CCM);
 out_unreg_gcm4106:
-	crypto_unregister_alg(&nx_gcm4106_aes_alg);
+	nx_unregister_aead(&nx_gcm4106_aes_alg, NX_FC_AES, NX_MODE_AES_GCM);
 out_unreg_gcm:
-	crypto_unregister_alg(&nx_gcm_aes_alg);
+	nx_unregister_aead(&nx_gcm_aes_alg, NX_FC_AES, NX_MODE_AES_GCM);
 out_unreg_ctr3686:
-	crypto_unregister_alg(&nx_ctr3686_aes_alg);
+	nx_unregister_alg(&nx_ctr3686_aes_alg, NX_FC_AES, NX_MODE_AES_CTR);
 out_unreg_ctr:
-	crypto_unregister_alg(&nx_ctr_aes_alg);
+	nx_unregister_alg(&nx_ctr_aes_alg, NX_FC_AES, NX_MODE_AES_CTR);
 out_unreg_cbc:
-	crypto_unregister_alg(&nx_cbc_aes_alg);
+	nx_unregister_alg(&nx_cbc_aes_alg, NX_FC_AES, NX_MODE_AES_CBC);
 out_unreg_ecb:
-	crypto_unregister_alg(&nx_ecb_aes_alg);
+	nx_unregister_alg(&nx_ecb_aes_alg, NX_FC_AES, NX_MODE_AES_ECB);
 out:
 	return rc;
 }
@@ -635,15 +713,16 @@ static int nx_crypto_ctx_init(struct nx_crypto_ctx *nx_ctx, u32 fc, u32 mode)
 /* entry points from the crypto tfm initializers */
 int nx_crypto_ctx_aes_ccm_init(struct crypto_tfm *tfm)
 {
-	tfm->crt_aead.reqsize = sizeof(struct nx_ccm_rctx);
+	crypto_aead_set_reqsize(__crypto_aead_cast(tfm),
+				sizeof(struct nx_ccm_rctx));
 	return nx_crypto_ctx_init(crypto_tfm_ctx(tfm), NX_FC_AES,
 				  NX_MODE_AES_CCM);
 }
 
-int nx_crypto_ctx_aes_gcm_init(struct crypto_tfm *tfm)
+int nx_crypto_ctx_aes_gcm_init(struct crypto_aead *tfm)
 {
-	tfm->crt_aead.reqsize = sizeof(struct nx_gcm_rctx);
-	return nx_crypto_ctx_init(crypto_tfm_ctx(tfm), NX_FC_AES,
+	crypto_aead_set_reqsize(tfm, sizeof(struct nx_gcm_rctx));
+	return nx_crypto_ctx_init(crypto_aead_ctx(tfm), NX_FC_AES,
 				  NX_MODE_AES_GCM);
 }
 
@@ -695,6 +774,13 @@ void nx_crypto_ctx_exit(struct crypto_tfm *tfm)
 	nx_ctx->out_sg = NULL;
 }
 
+void nx_crypto_ctx_aead_exit(struct crypto_aead *tfm)
+{
+	struct nx_crypto_ctx *nx_ctx = crypto_aead_ctx(tfm);
+
+	kzfree(nx_ctx->kmem);
+}
+
 static int nx_probe(struct vio_dev *viodev, const struct vio_device_id *id)
 {
 	dev_dbg(&viodev->dev, "driver probed: %s resource id: 0x%x\n",
@@ -721,17 +807,24 @@ static int nx_remove(struct vio_dev *viodev)
 	if (nx_driver.of.status == NX_OKAY) {
 		NX_DEBUGFS_FINI(&nx_driver);
 
-		crypto_unregister_alg(&nx_ccm_aes_alg);
-		crypto_unregister_alg(&nx_ccm4309_aes_alg);
-		crypto_unregister_alg(&nx_gcm_aes_alg);
-		crypto_unregister_alg(&nx_gcm4106_aes_alg);
-		crypto_unregister_alg(&nx_ctr_aes_alg);
-		crypto_unregister_alg(&nx_ctr3686_aes_alg);
-		crypto_unregister_alg(&nx_cbc_aes_alg);
-		crypto_unregister_alg(&nx_ecb_aes_alg);
-		crypto_unregister_shash(&nx_shash_sha256_alg);
-		crypto_unregister_shash(&nx_shash_sha512_alg);
-		crypto_unregister_shash(&nx_shash_aes_xcbc_alg);
+		nx_unregister_shash(&nx_shash_aes_xcbc_alg,
+				    NX_FC_AES, NX_MODE_AES_XCBC_MAC, -1);
+		nx_unregister_shash(&nx_shash_sha512_alg,
+				    NX_FC_SHA, NX_MODE_SHA, NX_PROPS_SHA256);
+		nx_unregister_shash(&nx_shash_sha256_alg,
+				    NX_FC_SHA, NX_MODE_SHA, NX_PROPS_SHA512);
+		nx_unregister_alg(&nx_ccm4309_aes_alg,
+				  NX_FC_AES, NX_MODE_AES_CCM);
+		nx_unregister_alg(&nx_ccm_aes_alg, NX_FC_AES, NX_MODE_AES_CCM);
+		nx_unregister_aead(&nx_gcm4106_aes_alg,
+				   NX_FC_AES, NX_MODE_AES_GCM);
+		nx_unregister_aead(&nx_gcm_aes_alg,
+				   NX_FC_AES, NX_MODE_AES_GCM);
+		nx_unregister_alg(&nx_ctr3686_aes_alg,
+				  NX_FC_AES, NX_MODE_AES_CTR);
+		nx_unregister_alg(&nx_ctr_aes_alg, NX_FC_AES, NX_MODE_AES_CTR);
+		nx_unregister_alg(&nx_cbc_aes_alg, NX_FC_AES, NX_MODE_AES_CBC);
+		nx_unregister_alg(&nx_ecb_aes_alg, NX_FC_AES, NX_MODE_AES_ECB);
 	}
 
 	return 0;
diff --git a/drivers/crypto/nx/nx.h b/drivers/crypto/nx/nx.h
index c3ed83764..cdff03a42 100644
--- a/drivers/crypto/nx/nx.h
+++ b/drivers/crypto/nx/nx.h
@@ -151,13 +151,14 @@ struct nx_crypto_ctx {
 
 /* prototypes */
 int nx_crypto_ctx_aes_ccm_init(struct crypto_tfm *tfm);
-int nx_crypto_ctx_aes_gcm_init(struct crypto_tfm *tfm);
+int nx_crypto_ctx_aes_gcm_init(struct crypto_aead *tfm);
 int nx_crypto_ctx_aes_xcbc_init(struct crypto_tfm *tfm);
 int nx_crypto_ctx_aes_ctr_init(struct crypto_tfm *tfm);
 int nx_crypto_ctx_aes_cbc_init(struct crypto_tfm *tfm);
 int nx_crypto_ctx_aes_ecb_init(struct crypto_tfm *tfm);
 int nx_crypto_ctx_sha_init(struct crypto_tfm *tfm);
 void nx_crypto_ctx_exit(struct crypto_tfm *tfm);
+void nx_crypto_ctx_aead_exit(struct crypto_aead *tfm);
 void nx_ctx_init(struct nx_crypto_ctx *nx_ctx, unsigned int function);
 int nx_hcall_sync(struct nx_crypto_ctx *ctx, struct vio_pfo_op *op,
 		  u32 may_sleep);
@@ -184,8 +185,8 @@ void nx_debugfs_fini(struct nx_crypto_driver *);
 
 extern struct crypto_alg nx_cbc_aes_alg;
 extern struct crypto_alg nx_ecb_aes_alg;
-extern struct crypto_alg nx_gcm_aes_alg;
-extern struct crypto_alg nx_gcm4106_aes_alg;
+extern struct aead_alg nx_gcm_aes_alg;
+extern struct aead_alg nx_gcm4106_aes_alg;
 extern struct crypto_alg nx_ctr_aes_alg;
 extern struct crypto_alg nx_ctr3686_aes_alg;
 extern struct crypto_alg nx_ccm_aes_alg;