Merge git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6

Pull crypto update from Herbert Xu:
 - add multibuffer infrastructure (single_task_running scheduler helper,
   OKed by Peter on lkml.
 - add SHA1 multibuffer implementation for AVX2.
 - reenable "by8" AVX CTR optimisation after fixing counter overflow.
 - add APM X-Gene SoC RNG support.
 - SHA256/SHA512 now handles unaligned input correctly.
 - set lz4 decompressed length correctly.
 - fix algif socket buffer allocation failure for 64K page machines.
 - misc fixes

* git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6: (47 commits)
  crypto: sha - Handle unaligned input data in generic sha256 and sha512.
  Revert "crypto: aesni - disable "by8" AVX CTR optimization"
  crypto: aesni - remove unused defines in "by8" variant
  crypto: aesni - fix counter overflow handling in "by8" variant
  hwrng: printk replacement
  crypto: qat - Removed unneeded partial state
  crypto: qat - Fix typo in name of tasklet_struct
  crypto: caam - Dynamic allocation of addresses for various memory blocks in CAAM.
  crypto: mcryptd - Fix typos in CRYPTO_MCRYPTD description
  crypto: algif - avoid excessive use of socket buffer in skcipher
  arm64: dts: add random number generator dts node to APM X-Gene platform.
  Documentation: rng: Add X-Gene SoC RNG driver documentation
  hwrng: xgene - add support for APM X-Gene SoC RNG support
  crypto: mv_cesa - Add missing #define
  crypto: testmgr - add test for lz4 and lz4hc
  crypto: lz4,lz4hc - fix decompression
  crypto: qat - Use pci_enable_msix_exact() instead of pci_enable_msix()
  crypto: drbg - fix maximum value checks on 32 bit systems
  crypto: drbg - fix sparse warning for cpu_to_be[32|64]
  crypto: sha-mb - sha1_mb_alg_state can be static
  ...

Documentation/devicetree/bindings/rng/apm,rng.txt

+APM X-Gene SoC random number generator.
+
+Required properties:
+
+- compatible  : should be "apm,xgene-rng"
+- reg         : specifies base physical address and size of the registers map
+- clocks      : phandle to clock-controller plus clock-specifier pair
+- interrupts  : specify the fault interrupt for the RNG device
+
+Example:
+
+	rng: rng@10520000 {
+		compatible = "apm,xgene-rng";
+		reg = <0x0 0x10520000 0x0 0x100>;
+		interrupts =   <0x0 0x41 0x4>;
+		clocks = <&rngpkaclk 0>;
+	};

arch/arm64/boot/dts/apm-storm.dtsi

@@ -269,6 +269,19 @@ rtcclk: rtcclk@17000000 {
 				enable-mask = <0x2>;
 				clock-output-names = "rtcclk";
 			};
+
+			rngpkaclk: rngpkaclk@17000000 {
+				compatible = "apm,xgene-device-clock";
+				#clock-cells = <1>;
+				clocks = <&socplldiv2 0>;
+				reg = <0x0 0x17000000 0x0 0x2000>;
+				reg-names = "csr-reg";
+				csr-offset = <0xc>;
+				csr-mask = <0x10>;
+				enable-offset = <0x10>;
+				enable-mask = <0x10>;
+				clock-output-names = "rngpkaclk";
+			};
 		};
 
 		serial0: serial@1c020000 {
@@ -421,5 +434,13 @@ menetphy: menetphy@3 {
 
 			};
 		};
+
+		rng: rng@10520000 {
+			compatible = "apm,xgene-rng";
+			reg = <0x0 0x10520000 0x0 0x100>;
+			interrupts = <0x0 0x41 0x4>;
+			clocks = <&rngpkaclk 0>;
+		};
+
 	};
 };

arch/x86/crypto/Makefile

@@ -26,6 +26,7 @@ obj-$(CONFIG_CRYPTO_GHASH_CLMUL_NI_INTEL) += ghash-clmulni-intel.o
 
 obj-$(CONFIG_CRYPTO_CRC32C_INTEL) += crc32c-intel.o
 obj-$(CONFIG_CRYPTO_SHA1_SSSE3) += sha1-ssse3.o
+obj-$(CONFIG_CRYPTO_SHA1_MB) += sha-mb/
 obj-$(CONFIG_CRYPTO_CRC32_PCLMUL) += crc32-pclmul.o
 obj-$(CONFIG_CRYPTO_SHA256_SSSE3) += sha256-ssse3.o
 obj-$(CONFIG_CRYPTO_SHA512_SSSE3) += sha512-ssse3.o

arch/x86/crypto/aes_ctrby8_avx-x86_64.S

@@ -79,9 +79,6 @@
 #define xcounter	%xmm8
 #define xbyteswap	%xmm9
 #define xkey0		%xmm10
-#define xkey3		%xmm11
-#define xkey6		%xmm12
-#define xkey9		%xmm13
 #define xkey4		%xmm11
 #define xkey8		%xmm12
 #define xkey12		%xmm13
@@ -108,6 +105,10 @@
 
 byteswap_const:
 	.octa 0x000102030405060708090A0B0C0D0E0F
+ddq_low_msk:
+	.octa 0x0000000000000000FFFFFFFFFFFFFFFF
+ddq_high_add_1:
+	.octa 0x00000000000000010000000000000000
 ddq_add_1:
 	.octa 0x00000000000000000000000000000001
 ddq_add_2:
@@ -169,7 +170,12 @@ ddq_add_8:
 	.rept (by - 1)
 		club DDQ_DATA, i
 		club XDATA, i
-		vpaddd	var_ddq_add(%rip), xcounter, var_xdata
+		vpaddq	var_ddq_add(%rip), xcounter, var_xdata
+		vptest	ddq_low_msk(%rip), var_xdata
+		jnz 1f
+		vpaddq	ddq_high_add_1(%rip), var_xdata, var_xdata
+		vpaddq	ddq_high_add_1(%rip), xcounter, xcounter
+		1:
 		vpshufb	xbyteswap, var_xdata, var_xdata
 		.set i, (i +1)
 	.endr
@@ -178,7 +184,11 @@ ddq_add_8:
 
 	vpxor	xkey0, xdata0, xdata0
 	club DDQ_DATA, by
-	vpaddd	var_ddq_add(%rip), xcounter, xcounter
+	vpaddq	var_ddq_add(%rip), xcounter, xcounter
+	vptest	ddq_low_msk(%rip), xcounter
+	jnz	1f
+	vpaddq	ddq_high_add_1(%rip), xcounter, xcounter
+	1:
 
 	.set i, 1
 	.rept (by - 1)

arch/x86/crypto/aesni-intel_glue.c

@@ -481,7 +481,7 @@ static void ctr_crypt_final(struct crypto_aes_ctx *ctx,
 	crypto_inc(ctrblk, AES_BLOCK_SIZE);
 }
 
-#if 0	/* temporary disabled due to failing crypto tests */
+#ifdef CONFIG_AS_AVX
 static void aesni_ctr_enc_avx_tfm(struct crypto_aes_ctx *ctx, u8 *out,
 			      const u8 *in, unsigned int len, u8 *iv)
 {
@@ -1522,7 +1522,7 @@ static int __init aesni_init(void)
 		aesni_gcm_dec_tfm = aesni_gcm_dec;
 	}
 	aesni_ctr_enc_tfm = aesni_ctr_enc;
-#if 0	/* temporary disabled due to failing crypto tests */
+#ifdef CONFIG_AS_AVX
 	if (cpu_has_avx) {
 		/* optimize performance of ctr mode encryption transform */
 		aesni_ctr_enc_tfm = aesni_ctr_enc_avx_tfm;

arch/x86/crypto/sha-mb/Makefile

+#
+# Arch-specific CryptoAPI modules.
+#
+
+avx2_supported := $(call as-instr,vpgatherdd %ymm0$(comma)(%eax$(comma)%ymm1\
+                                $(comma)4)$(comma)%ymm2,yes,no)
+ifeq ($(avx2_supported),yes)
+	obj-$(CONFIG_CRYPTO_SHA1_MB) += sha1-mb.o
+	sha1-mb-y := sha1_mb.o sha1_mb_mgr_flush_avx2.o \
+	     sha1_mb_mgr_init_avx2.o sha1_mb_mgr_submit_avx2.o sha1_x8_avx2.o
+endif

arch/x86/crypto/sha-mb/sha1_mb_mgr_datastruct.S

+/*
+ * Header file for multi buffer SHA1 algorithm data structure
+ *
+ * This file is provided under a dual BSD/GPLv2 license.  When using or
+ * redistributing this file, you may do so under either license.
+ *
+ * GPL LICENSE SUMMARY
+ *
+ *  Copyright(c) 2014 Intel Corporation.
+ *
+ *  This program is free software; you can redistribute it and/or modify
+ *  it under the terms of version 2 of the GNU General Public License as
+ *  published by the Free Software Foundation.
+ *
+ *  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.
+ *
+ *  Contact Information:
+ *      James Guilford <james.guilford@intel.com>
+ *	Tim Chen <tim.c.chen@linux.intel.com>
+ *
+ *  BSD LICENSE
+ *
+ *  Copyright(c) 2014 Intel Corporation.
+ *
+ *  Redistribution and use in source and binary forms, with or without
+ *  modification, are permitted provided that the following conditions
+ *  are met:
+ *
+ *    * Redistributions of source code must retain the above copyright
+ *      notice, this list of conditions and the following disclaimer.
+ *    * Redistributions in binary form must reproduce the above copyright
+ *      notice, this list of conditions and the following disclaimer in
+ *      the documentation and/or other materials provided with the
+ *      distribution.
+ *    * Neither the name of Intel Corporation nor the names of its
+ *      contributors may be used to endorse or promote products derived
+ *      from this software without specific prior written permission.
+ *
+ *  THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ *  "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ *  LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+ *  A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+ *  OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ *  SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+ *  LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+ *  DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+ *  THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ *  (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+ * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+# Macros for defining data structures
+
+# Usage example
+
+#START_FIELDS	# JOB_AES
+###	name		size	align
+#FIELD	_plaintext,	8,	8	# pointer to plaintext
+#FIELD	_ciphertext,	8,	8	# pointer to ciphertext
+#FIELD	_IV,		16,	8	# IV
+#FIELD	_keys,		8,	8	# pointer to keys
+#FIELD	_len,		4,	4	# length in bytes
+#FIELD	_status,	4,	4	# status enumeration
+#FIELD	_user_data,	8,	8	# pointer to user data
+#UNION  _union,         size1,  align1, \
+#	                size2,  align2, \
+#	                size3,  align3, \
+#	                ...
+#END_FIELDS
+#%assign _JOB_AES_size	_FIELD_OFFSET
+#%assign _JOB_AES_align	_STRUCT_ALIGN
+
+#########################################################################
+
+# Alternate "struc-like" syntax:
+#	STRUCT job_aes2
+#	RES_Q	.plaintext,	1
+#	RES_Q	.ciphertext,	1
+#	RES_DQ	.IV,		1
+#	RES_B	.nested,	_JOB_AES_SIZE, _JOB_AES_ALIGN
+#	RES_U	.union,		size1, align1, \
+#				size2, align2, \
+#				...
+#	ENDSTRUCT
+#	# Following only needed if nesting
+#	%assign job_aes2_size	_FIELD_OFFSET
+#	%assign job_aes2_align	_STRUCT_ALIGN
+#
+# RES_* macros take a name, a count and an optional alignment.
+# The count in in terms of the base size of the macro, and the
+# default alignment is the base size.
+# The macros are:
+# Macro    Base size
+# RES_B	    1
+# RES_W	    2
+# RES_D     4
+# RES_Q     8
+# RES_DQ   16
+# RES_Y    32
+# RES_Z    64
+#
+# RES_U defines a union. It's arguments are a name and two or more
+# pairs of "size, alignment"
+#
+# The two assigns are only needed if this structure is being nested
+# within another. Even if the assigns are not done, one can still use
+# STRUCT_NAME_size as the size of the structure.
+#
+# Note that for nesting, you still need to assign to STRUCT_NAME_size.
+#
+# The differences between this and using "struc" directly are that each
+# type is implicitly aligned to its natural length (although this can be
+# over-ridden with an explicit third parameter), and that the structure
+# is padded at the end to its overall alignment.
+#
+
+#########################################################################
+
+#ifndef _SHA1_MB_MGR_DATASTRUCT_ASM_
+#define _SHA1_MB_MGR_DATASTRUCT_ASM_
+
+## START_FIELDS
+.macro START_FIELDS
+ _FIELD_OFFSET = 0
+ _STRUCT_ALIGN = 0
+.endm
+
+## FIELD name size align
+.macro FIELD name size align
+ _FIELD_OFFSET = (_FIELD_OFFSET + (\align) - 1) & (~ ((\align)-1))
+ \name	= _FIELD_OFFSET
+ _FIELD_OFFSET = _FIELD_OFFSET + (\size)
+.if (\align > _STRUCT_ALIGN)
+ _STRUCT_ALIGN = \align
+.endif
+.endm
+
+## END_FIELDS
+.macro END_FIELDS
+ _FIELD_OFFSET = (_FIELD_OFFSET + _STRUCT_ALIGN-1) & (~ (_STRUCT_ALIGN-1))
+.endm
+
+########################################################################
+
+.macro STRUCT p1
+START_FIELDS
+.struc \p1
+.endm
+
+.macro ENDSTRUCT
+ tmp = _FIELD_OFFSET
+ END_FIELDS
+ tmp = (_FIELD_OFFSET - %%tmp)
+.if (tmp > 0)
+	.lcomm	tmp
+.endif
+.endstruc
+.endm
+
+## RES_int name size align
+.macro RES_int p1 p2 p3
+ name = \p1
+ size = \p2
+ align = .\p3
+
+ _FIELD_OFFSET = (_FIELD_OFFSET + (align) - 1) & (~ ((align)-1))
+.align align
+.lcomm name size
+ _FIELD_OFFSET = _FIELD_OFFSET + (size)
+.if (align > _STRUCT_ALIGN)
+ _STRUCT_ALIGN = align
+.endif
+.endm
+
+
+
+# macro RES_B name, size [, align]
+.macro RES_B _name, _size, _align=1
+RES_int _name _size _align
+.endm
+
+# macro RES_W name, size [, align]
+.macro RES_W _name, _size, _align=2
+RES_int _name 2*(_size) _align
+.endm
+
+# macro RES_D name, size [, align]
+.macro RES_D _name, _size, _align=4
+RES_int _name 4*(_size) _align
+.endm
+
+# macro RES_Q name, size [, align]
+.macro RES_Q _name, _size, _align=8
+RES_int _name 8*(_size) _align
+.endm
+
+# macro RES_DQ name, size [, align]
+.macro RES_DQ _name, _size, _align=16
+RES_int _name 16*(_size) _align
+.endm
+
+# macro RES_Y name, size [, align]
+.macro RES_Y _name, _size, _align=32
+RES_int _name 32*(_size) _align
+.endm
+
+# macro RES_Z name, size [, align]
+.macro RES_Z _name, _size, _align=64
+RES_int _name 64*(_size) _align
+.endm
+
+
+#endif
+
+########################################################################
+#### Define constants
+########################################################################
+
+########################################################################
+#### Define SHA1 Out Of Order Data Structures
+########################################################################
+
+START_FIELDS    # LANE_DATA
+###     name            size    align
+FIELD   _job_in_lane,   8,      8       # pointer to job object
+END_FIELDS
+
+_LANE_DATA_size = _FIELD_OFFSET
+_LANE_DATA_align = _STRUCT_ALIGN
+
+########################################################################
+
+START_FIELDS    # SHA1_ARGS_X8
+###     name            size    align
+FIELD   _digest,        4*5*8,  16      # transposed digest
+FIELD   _data_ptr,      8*8,    8       # array of pointers to data
+END_FIELDS
+
+_SHA1_ARGS_X4_size =     _FIELD_OFFSET
+_SHA1_ARGS_X4_align =    _STRUCT_ALIGN
+_SHA1_ARGS_X8_size =     _FIELD_OFFSET
+_SHA1_ARGS_X8_align =    _STRUCT_ALIGN
+
+########################################################################
+
+START_FIELDS    # MB_MGR
+###     name            size    align
+FIELD   _args,          _SHA1_ARGS_X4_size, _SHA1_ARGS_X4_align
+FIELD   _lens,          4*8,    8
+FIELD   _unused_lanes,  8,      8
+FIELD   _ldata,         _LANE_DATA_size*8, _LANE_DATA_align
+END_FIELDS
+
+_MB_MGR_size =   _FIELD_OFFSET
+_MB_MGR_align =  _STRUCT_ALIGN
+
+_args_digest    =     _args + _digest
+_args_data_ptr  =     _args + _data_ptr
+
+
+########################################################################
+#### Define constants
+########################################################################
+
+#define STS_UNKNOWN             0
+#define STS_BEING_PROCESSED     1
+#define STS_COMPLETED           2
+
+########################################################################
+#### Define JOB_SHA1 structure
+########################################################################
+
+START_FIELDS    # JOB_SHA1
+
+###     name                            size    align
+FIELD   _buffer,                        8,      8       # pointer to buffer
+FIELD   _len,                           4,      4       # length in bytes
+FIELD   _result_digest,                 5*4,    32      # Digest (output)
+FIELD   _status,                        4,      4
+FIELD   _user_data,                     8,      8
+END_FIELDS
+
+_JOB_SHA1_size =  _FIELD_OFFSET
+_JOB_SHA1_align = _STRUCT_ALIGN

arch/x86/crypto/sha-mb/sha1_mb_mgr_flush_avx2.S

+/*
+ * Flush routine for SHA1 multibuffer
+ *
+ * This file is provided under a dual BSD/GPLv2 license.  When using or
+ * redistributing this file, you may do so under either license.
+ *
+ * GPL LICENSE SUMMARY
+ *
+ *  Copyright(c) 2014 Intel Corporation.
+ *
+ *  This program is free software; you can redistribute it and/or modify
+ *  it under the terms of version 2 of the GNU General Public License as
+ *  published by the Free Software Foundation.
+ *
+ *  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.
+ *
+ *  Contact Information:
+ *      James Guilford <james.guilford@intel.com>
+ *	Tim Chen <tim.c.chen@linux.intel.com>
+ *
+ *  BSD LICENSE
+ *
+ *  Copyright(c) 2014 Intel Corporation.
+ *
+ *  Redistribution and use in source and binary forms, with or without
+ *  modification, are permitted provided that the following conditions
+ *  are met:
+ *
+ *    * Redistributions of source code must retain the above copyright
+ *      notice, this list of conditions and the following disclaimer.
+ *    * Redistributions in binary form must reproduce the above copyright
+ *      notice, this list of conditions and the following disclaimer in
+ *      the documentation and/or other materials provided with the
+ *      distribution.
+ *    * Neither the name of Intel Corporation nor the names of its
+ *      contributors may be used to endorse or promote products derived
+ *      from this software without specific prior written permission.
+ *
+ *  THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ *  "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ *  LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+ *  A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+ *  OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ *  SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+ *  LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+ *  DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+ *  THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ *  (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+ * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+#include <linux/linkage.h>
+#include "sha1_mb_mgr_datastruct.S"
+
+
+.extern sha1_x8_avx2
+
+# LINUX register definitions
+#define arg1    %rdi
+#define arg2    %rsi
+
+# Common definitions
+#define state   arg1
+#define job     arg2
+#define len2    arg2
+
+# idx must be a register not clobbered by sha1_x8_avx2
+#define idx		%r8
+#define DWORD_idx	%r8d
+
+#define unused_lanes    %rbx
+#define lane_data       %rbx
+#define tmp2            %rbx
+#define tmp2_w		%ebx
+
+#define job_rax         %rax
+#define tmp1            %rax
+#define size_offset     %rax
+#define tmp             %rax
+#define start_offset    %rax
+
+#define tmp3            %arg1
+
+#define extra_blocks    %arg2
+#define p               %arg2
+
+
+# STACK_SPACE needs to be an odd multiple of 8
+_XMM_SAVE_SIZE  = 10*16
+_GPR_SAVE_SIZE  = 8*8
+_ALIGN_SIZE     = 8
+
+_XMM_SAVE       = 0
+_GPR_SAVE       = _XMM_SAVE + _XMM_SAVE_SIZE
+STACK_SPACE     = _GPR_SAVE + _GPR_SAVE_SIZE + _ALIGN_SIZE
+
+.macro LABEL prefix n
+\prefix\n\():
+.endm
+
+.macro JNE_SKIP i
+jne     skip_\i
+.endm
+
+.altmacro
+.macro SET_OFFSET _offset
+offset = \_offset
+.endm
+.noaltmacro
+
+# JOB* sha1_mb_mgr_flush_avx2(MB_MGR *state)
+# arg 1 : rcx : state
+ENTRY(sha1_mb_mgr_flush_avx2)
+	mov	%rsp, %r10
+	sub     $STACK_SPACE, %rsp
+	and     $~31, %rsp
+	mov     %rbx, _GPR_SAVE(%rsp)
+	mov     %r10, _GPR_SAVE+8*1(%rsp) #save rsp
+	mov	%rbp, _GPR_SAVE+8*3(%rsp)
+	mov	%r12, _GPR_SAVE+8*4(%rsp)
+	mov	%r13, _GPR_SAVE+8*5(%rsp)
+	mov	%r14, _GPR_SAVE+8*6(%rsp)
+	mov	%r15, _GPR_SAVE+8*7(%rsp)
+
+	# If bit (32+3) is set, then all lanes are empty
+	mov     _unused_lanes(state), unused_lanes
+	bt      $32+3, unused_lanes
+	jc      return_null
+
+	# find a lane with a non-null job
+	xor     idx, idx
+	offset = (_ldata + 1 * _LANE_DATA_size + _job_in_lane)
+	cmpq    $0, offset(state)
+	cmovne  one(%rip), idx
+	offset = (_ldata + 2 * _LANE_DATA_size + _job_in_lane)
+	cmpq    $0, offset(state)
+	cmovne  two(%rip), idx
+	offset = (_ldata + 3 * _LANE_DATA_size + _job_in_lane)
+	cmpq    $0, offset(state)
+	cmovne  three(%rip), idx
+	offset = (_ldata + 4 * _LANE_DATA_size + _job_in_lane)
+	cmpq    $0, offset(state)
+	cmovne  four(%rip), idx
+	offset = (_ldata + 5 * _LANE_DATA_size + _job_in_lane)
+	cmpq    $0, offset(state)
+	cmovne  five(%rip), idx
+	offset = (_ldata + 6 * _LANE_DATA_size + _job_in_lane)
+	cmpq    $0, offset(state)
+	cmovne  six(%rip), idx
+	offset = (_ldata + 7 * _LANE_DATA_size + _job_in_lane)
+	cmpq    $0, offset(state)
+	cmovne  seven(%rip), idx
+
+	# copy idx to empty lanes
+copy_lane_data:
+	offset =  (_args + _data_ptr)
+	mov     offset(state,idx,8), tmp
+
+	I = 0
+.rep 8
+	offset =  (_ldata + I * _LANE_DATA_size + _job_in_lane)
+	cmpq    $0, offset(state)
+.altmacro
+	JNE_SKIP %I
+	offset =  (_args + _data_ptr + 8*I)
+	mov     tmp, offset(state)
+	offset =  (_lens + 4*I)
+	movl    $0xFFFFFFFF, offset(state)
+LABEL skip_ %I
+	I = (I+1)
+.noaltmacro
+.endr
+
+	# Find min length
+	vmovdqa _lens+0*16(state), %xmm0
+	vmovdqa _lens+1*16(state), %xmm1
+
+	vpminud %xmm1, %xmm0, %xmm2     # xmm2 has {D,C,B,A}
+	vpalignr $8, %xmm2, %xmm3, %xmm3   # xmm3 has {x,x,D,C}
+	vpminud %xmm3, %xmm2, %xmm2        # xmm2 has {x,x,E,F}
+	vpalignr $4, %xmm2, %xmm3, %xmm3    # xmm3 has {x,x,x,E}
+	vpminud %xmm3, %xmm2, %xmm2        # xmm2 has min value in low dword
+
+	vmovd   %xmm2, DWORD_idx
+	mov	idx, len2
+	and	$0xF, idx
+	shr	$4, len2
+	jz	len_is_0
+
+	vpand   clear_low_nibble(%rip), %xmm2, %xmm2
+	vpshufd $0, %xmm2, %xmm2
+
+	vpsubd  %xmm2, %xmm0, %xmm0
+	vpsubd  %xmm2, %xmm1, %xmm1
+
+	vmovdqa %xmm0, _lens+0*16(state)
+	vmovdqa %xmm1, _lens+1*16(state)
+
+	# "state" and "args" are the same address, arg1
+	# len is arg2
+	call	sha1_x8_avx2
+	# state and idx are intact
+
+
+len_is_0:
+	# process completed job "idx"
+	imul    $_LANE_DATA_size, idx, lane_data
+	lea     _ldata(state, lane_data), lane_data
+
+	mov     _job_in_lane(lane_data), job_rax
+	movq    $0, _job_in_lane(lane_data)
+	movl    $STS_COMPLETED, _status(job_rax)
+	mov     _unused_lanes(state), unused_lanes
+	shl     $4, unused_lanes
+	or      idx, unused_lanes
+	mov     unused_lanes, _unused_lanes(state)
+
+	movl	$0xFFFFFFFF, _lens(state, idx, 4)
+
+	vmovd    _args_digest(state , idx, 4) , %xmm0
+	vpinsrd  $1, _args_digest+1*32(state, idx, 4), %xmm0, %xmm0
+	vpinsrd  $2, _args_digest+2*32(state, idx, 4), %xmm0, %xmm0
+	vpinsrd  $3, _args_digest+3*32(state, idx, 4), %xmm0, %xmm0
+	movl    _args_digest+4*32(state, idx, 4), tmp2_w
+
+	vmovdqu  %xmm0, _result_digest(job_rax)
+	offset =  (_result_digest + 1*16)
+	mov     tmp2_w, offset(job_rax)
+
+return:
+
+	mov     _GPR_SAVE(%rsp), %rbx
+	mov     _GPR_SAVE+8*1(%rsp), %r10 #saved rsp
+	mov	_GPR_SAVE+8*3(%rsp), %rbp
+	mov	_GPR_SAVE+8*4(%rsp), %r12
+	mov	_GPR_SAVE+8*5(%rsp), %r13
+	mov	_GPR_SAVE+8*6(%rsp), %r14
+	mov	_GPR_SAVE+8*7(%rsp), %r15
+	mov     %r10, %rsp
+
+	ret
+
+return_null:
+	xor     job_rax, job_rax
+	jmp     return
+ENDPROC(sha1_mb_mgr_flush_avx2)
+
+
+#################################################################
+
+.align 16
+ENTRY(sha1_mb_mgr_get_comp_job_avx2)
+	push    %rbx
+
+	## if bit 32+3 is set, then all lanes are empty
+	mov     _unused_lanes(state), unused_lanes
+	bt      $(32+3), unused_lanes
+	jc      .return_null
+
+	# Find min length
+	vmovdqa _lens(state), %xmm0
+	vmovdqa _lens+1*16(state), %xmm1
+
+	vpminud %xmm1, %xmm0, %xmm2        # xmm2 has {D,C,B,A}
+	vpalignr $8, %xmm2, %xmm3, %xmm3   # xmm3 has {x,x,D,C}
+	vpminud %xmm3, %xmm2, %xmm2        # xmm2 has {x,x,E,F}
+	vpalignr $4, %xmm2, %xmm3, %xmm3    # xmm3 has {x,x,x,E}
+	vpminud %xmm3, %xmm2, %xmm2        # xmm2 has min value in low dword
+
+	vmovd   %xmm2, DWORD_idx
+	test    $~0xF, idx
+	jnz     .return_null
+
+	# process completed job "idx"
+	imul    $_LANE_DATA_size, idx, lane_data
+	lea     _ldata(state, lane_data), lane_data
+
+	mov     _job_in_lane(lane_data), job_rax
+	movq    $0,  _job_in_lane(lane_data)
+	movl    $STS_COMPLETED, _status(job_rax)
+	mov     _unused_lanes(state), unused_lanes
+	shl     $4, unused_lanes
+	or      idx, unused_lanes
+	mov     unused_lanes, _unused_lanes(state)
+
+	movl    $0xFFFFFFFF, _lens(state,  idx, 4)
+
+	vmovd   _args_digest(state, idx, 4), %xmm0
+	vpinsrd $1, _args_digest+1*32(state, idx, 4), %xmm0, %xmm0
+	vpinsrd $2, _args_digest+2*32(state, idx, 4), %xmm0, %xmm0
+	vpinsrd $3, _args_digest+3*32(state, idx, 4), %xmm0, %xmm0
+	movl    _args_digest+4*32(state, idx, 4), tmp2_w
+
+	vmovdqu %xmm0, _result_digest(job_rax)
+	movl    tmp2_w, _result_digest+1*16(job_rax)
+
+	pop     %rbx
+
+	ret
+
+.return_null:
+	xor     job_rax, job_rax
+	pop     %rbx
+	ret
+ENDPROC(sha1_mb_mgr_get_comp_job_avx2)
+
+.data
+
+.align 16
+clear_low_nibble:
+.octa	0x000000000000000000000000FFFFFFF0
+one:
+.quad  1
+two:
+.quad  2
+three:
+.quad  3
+four:
+.quad  4
+five:
+.quad  5
+six:
+.quad  6
+seven:
+.quad  7

arch/x86/crypto/sha-mb/sha1_mb_mgr_init_avx2.c

+/*
+ * Initialization code for multi buffer SHA1 algorithm for AVX2
+ *
+ * This file is provided under a dual BSD/GPLv2 license.  When using or
+ * redistributing this file, you may do so under either license.
+ *
+ * GPL LICENSE SUMMARY
+ *
+ *  Copyright(c) 2014 Intel Corporation.
+ *
+ *  This program is free software; you can redistribute it and/or modify
+ *  it under the terms of version 2 of the GNU General Public License as
+ *  published by the Free Software Foundation.
+ *
+ *  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.
+ *
+ *  Contact Information:
+ *	Tim Chen <tim.c.chen@linux.intel.com>
+ *
+ *  BSD LICENSE
+ *
+ *  Copyright(c) 2014 Intel Corporation.
+ *
+ *  Redistribution and use in source and binary forms, with or without
+ *  modification, are permitted provided that the following conditions
+ *  are met:
+ *
+ *    * Redistributions of source code must retain the above copyright
+ *      notice, this list of conditions and the following disclaimer.
+ *    * Redistributions in binary form must reproduce the above copyright
+ *      notice, this list of conditions and the following disclaimer in
+ *      the documentation and/or other materials provided with the
+ *      distribution.
+ *    * Neither the name of Intel Corporation nor the names of its
+ *      contributors may be used to endorse or promote products derived
+ *      from this software without specific prior written permission.
+ *
+ *  THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ *  "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ *  LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+ *  A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+ *  OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ *  SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+ *  LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+ *  DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+ *  THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ *  (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+ * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#include "sha_mb_mgr.h"
+
+void sha1_mb_mgr_init_avx2(struct sha1_mb_mgr *state)
+{
+	unsigned int j;
+	state->unused_lanes = 0xF76543210;
+	for (j = 0; j < 8; j++) {
+		state->lens[j] = 0xFFFFFFFF;
+		state->ldata[j].job_in_lane = NULL;
+	}
+}

arch/x86/crypto/sha-mb/sha1_mb_mgr_submit_avx2.S

+/*
+ * Buffer submit code for multi buffer SHA1 algorithm
+ *
+ * This file is provided under a dual BSD/GPLv2 license.  When using or
+ * redistributing this file, you may do so under either license.
+ *
+ * GPL LICENSE SUMMARY
+ *
+ *  Copyright(c) 2014 Intel Corporation.
+ *
+ *  This program is free software; you can redistribute it and/or modify
+ *  it under the terms of version 2 of the GNU General Public License as
+ *  published by the Free Software Foundation.
+ *
+ *  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.
+ *
+ *  Contact Information:
+ *      James Guilford <james.guilford@intel.com>
+ *	Tim Chen <tim.c.chen@linux.intel.com>
+ *
+ *  BSD LICENSE
+ *
+ *  Copyright(c) 2014 Intel Corporation.
+ *
+ *  Redistribution and use in source and binary forms, with or without
+ *  modification, are permitted provided that the following conditions
+ *  are met:
+ *
+ *    * Redistributions of source code must retain the above copyright
+ *      notice, this list of conditions and the following disclaimer.
+ *    * Redistributions in binary form must reproduce the above copyright
+ *      notice, this list of conditions and the following disclaimer in
+ *      the documentation and/or other materials provided with the
+ *      distribution.
+ *    * Neither the name of Intel Corporation nor the names of its
+ *      contributors may be used to endorse or promote products derived
+ *      from this software without specific prior written permission.
+ *
+ *  THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ *  "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ *  LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+ *  A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+ *  OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ *  SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+ *  LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+ *  DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+ *  THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ *  (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+ * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#include <linux/linkage.h>
+#include "sha1_mb_mgr_datastruct.S"
+
+
+.extern sha1_x8_avx
+
+# LINUX register definitions
+arg1    = %rdi
+arg2    = %rsi
+size_offset	= %rcx
+tmp2		= %rcx
+extra_blocks	= %rdx
+
+# Common definitions
+#define state   arg1
+#define job     %rsi
+#define len2    arg2
+#define p2      arg2
+
+# idx must be a register not clobberred by sha1_x8_avx2
+idx		= %r8
+DWORD_idx	= %r8d
+last_len	= %r8
+
+p               = %r11
+start_offset    = %r11
+
+unused_lanes    = %rbx
+BYTE_unused_lanes = %bl
+
+job_rax         = %rax
+len             = %rax
+DWORD_len	= %eax
+
+lane            = %rbp
+tmp3            = %rbp
+
+tmp             = %r9
+DWORD_tmp	= %r9d
+
+lane_data       = %r10
+
+# STACK_SPACE needs to be an odd multiple of 8
+STACK_SPACE     = 8*8 + 16*10 + 8
+
+# JOB* submit_mb_mgr_submit_avx2(MB_MGR *state, job_sha1 *job)
+# arg 1 : rcx : state
+# arg 2 : rdx : job
+ENTRY(sha1_mb_mgr_submit_avx2)
+
+	mov	%rsp, %r10
+	sub     $STACK_SPACE, %rsp
+	and	$~31, %rsp
+
+	mov     %rbx, (%rsp)
+	mov	%r10, 8*2(%rsp)	#save old rsp
+	mov     %rbp, 8*3(%rsp)
+	mov	%r12, 8*4(%rsp)
+	mov	%r13, 8*5(%rsp)
+	mov	%r14, 8*6(%rsp)
+	mov	%r15, 8*7(%rsp)
+
+	mov     _unused_lanes(state), unused_lanes
+	mov	unused_lanes, lane
+	and	$0xF, lane
+	shr     $4, unused_lanes
+	imul    $_LANE_DATA_size, lane, lane_data
+	movl    $STS_BEING_PROCESSED, _status(job)
+	lea     _ldata(state, lane_data), lane_data
+	mov     unused_lanes, _unused_lanes(state)
+	movl    _len(job),  DWORD_len
+
+	mov	job, _job_in_lane(lane_data)
+	shl	$4, len
+	or	lane, len
+
+	movl    DWORD_len,  _lens(state , lane, 4)
+
+	# Load digest words from result_digest
+	vmovdqu	_result_digest(job), %xmm0
+	mov	_result_digest+1*16(job), DWORD_tmp
+	vmovd    %xmm0, _args_digest(state, lane, 4)
+	vpextrd  $1, %xmm0, _args_digest+1*32(state , lane, 4)
+	vpextrd  $2, %xmm0, _args_digest+2*32(state , lane, 4)
+	vpextrd  $3, %xmm0, _args_digest+3*32(state , lane, 4)
+	movl    DWORD_tmp, _args_digest+4*32(state , lane, 4)
+
+	mov     _buffer(job), p
+	mov     p, _args_data_ptr(state, lane, 8)
+
+	cmp     $0xF, unused_lanes
+	jne     return_null
+
+start_loop:
+	# Find min length
+	vmovdqa _lens(state), %xmm0
+	vmovdqa _lens+1*16(state), %xmm1
+
+	vpminud %xmm1, %xmm0, %xmm2        # xmm2 has {D,C,B,A}
+	vpalignr $8, %xmm2, %xmm3, %xmm3   # xmm3 has {x,x,D,C}
+	vpminud %xmm3, %xmm2, %xmm2        # xmm2 has {x,x,E,F}
+	vpalignr $4, %xmm2, %xmm3, %xmm3   # xmm3 has {x,x,x,E}
+	vpminud %xmm3, %xmm2, %xmm2        # xmm2 has min value in low dword
+
+	vmovd   %xmm2, DWORD_idx
+	mov    idx, len2
+	and    $0xF, idx
+	shr    $4, len2
+	jz     len_is_0
+
+	vpand   clear_low_nibble(%rip), %xmm2, %xmm2
+	vpshufd $0, %xmm2, %xmm2
+
+	vpsubd  %xmm2, %xmm0, %xmm0
+	vpsubd  %xmm2, %xmm1, %xmm1
+
+	vmovdqa %xmm0, _lens + 0*16(state)
+	vmovdqa %xmm1, _lens + 1*16(state)
+
+
+	# "state" and "args" are the same address, arg1
+	# len is arg2
+	call    sha1_x8_avx2
+
+	# state and idx are intact
+
+len_is_0:
+	# process completed job "idx"
+	imul    $_LANE_DATA_size, idx, lane_data
+	lea     _ldata(state, lane_data), lane_data
+
+	mov     _job_in_lane(lane_data), job_rax
+	mov     _unused_lanes(state), unused_lanes
+	movq    $0, _job_in_lane(lane_data)
+	movl    $STS_COMPLETED, _status(job_rax)
+	shl     $4, unused_lanes
+	or      idx, unused_lanes
+	mov     unused_lanes, _unused_lanes(state)
+
+	movl	$0xFFFFFFFF, _lens(state, idx, 4)
+
+	vmovd    _args_digest(state, idx, 4), %xmm0
+	vpinsrd  $1, _args_digest+1*32(state , idx, 4), %xmm0, %xmm0
+	vpinsrd  $2, _args_digest+2*32(state , idx, 4), %xmm0, %xmm0
+	vpinsrd  $3, _args_digest+3*32(state , idx, 4), %xmm0, %xmm0
+	movl    4*32(state, idx, 4), DWORD_tmp
+
+	vmovdqu  %xmm0, _result_digest(job_rax)
+	movl    DWORD_tmp, _result_digest+1*16(job_rax)
+
+return:
+
+	mov     (%rsp), %rbx
+	mov	8*2(%rsp), %r10	#save old rsp
+	mov     8*3(%rsp), %rbp
+	mov	8*4(%rsp), %r12
+	mov	8*5(%rsp), %r13
+	mov	8*6(%rsp), %r14
+	mov	8*7(%rsp), %r15
+	mov     %r10, %rsp
+
+	ret
+
+return_null:
+	xor     job_rax, job_rax
+	jmp     return
+
+ENDPROC(sha1_mb_mgr_submit_avx2)
+
+.data
+
+.align 16
+clear_low_nibble:
+	.octa	0x000000000000000000000000FFFFFFF0

arch/x86/crypto/sha-mb/sha_mb_ctx.h

+/*
+ * Header file for multi buffer SHA context
+ *
+ * This file is provided under a dual BSD/GPLv2 license.  When using or
+ * redistributing this file, you may do so under either license.
+ *
+ * GPL LICENSE SUMMARY
+ *
+ *  Copyright(c) 2014 Intel Corporation.
+ *
+ *  This program is free software; you can redistribute it and/or modify
+ *  it under the terms of version 2 of the GNU General Public License as
+ *  published by the Free Software Foundation.
+ *
+ *  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.
+ *
+ *  Contact Information:
+ *	Tim Chen <tim.c.chen@linux.intel.com>
+ *
+ *  BSD LICENSE
+ *
+ *  Copyright(c) 2014 Intel Corporation.
+ *
+ *  Redistribution and use in source and binary forms, with or without
+ *  modification, are permitted provided that the following conditions
+ *  are met:
+ *
+ *    * Redistributions of source code must retain the above copyright
+ *      notice, this list of conditions and the following disclaimer.
+ *    * Redistributions in binary form must reproduce the above copyright
+ *      notice, this list of conditions and the following disclaimer in
+ *      the documentation and/or other materials provided with the
+ *      distribution.
+ *    * Neither the name of Intel Corporation nor the names of its
+ *      contributors may be used to endorse or promote products derived
+ *      from this software without specific prior written permission.
+ *
+ *  THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ *  "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ *  LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+ *  A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+ *  OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ *  SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+ *  LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+ *  DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+ *  THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ *  (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+ * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#ifndef _SHA_MB_CTX_INTERNAL_H
+#define _SHA_MB_CTX_INTERNAL_H
+
+#include "sha_mb_mgr.h"
+
+#define HASH_UPDATE          0x00
+#define HASH_FIRST           0x01
+#define HASH_LAST            0x02
+#define HASH_ENTIRE          0x03
+#define HASH_DONE	     0x04
+#define HASH_FINAL	     0x08
+
+#define HASH_CTX_STS_IDLE       0x00
+#define HASH_CTX_STS_PROCESSING 0x01
+#define HASH_CTX_STS_LAST       0x02
+#define HASH_CTX_STS_COMPLETE   0x04
+
+enum hash_ctx_error {
+	HASH_CTX_ERROR_NONE               =  0,
+	HASH_CTX_ERROR_INVALID_FLAGS      = -1,
+	HASH_CTX_ERROR_ALREADY_PROCESSING = -2,
+	HASH_CTX_ERROR_ALREADY_COMPLETED  = -3,
+
+#ifdef HASH_CTX_DEBUG
+	HASH_CTX_ERROR_DEBUG_DIGEST_MISMATCH = -4,
+#endif
+};
+
+
+#define hash_ctx_user_data(ctx)  ((ctx)->user_data)
+#define hash_ctx_digest(ctx)     ((ctx)->job.result_digest)
+#define hash_ctx_processing(ctx) ((ctx)->status & HASH_CTX_STS_PROCESSING)
+#define hash_ctx_complete(ctx)   ((ctx)->status == HASH_CTX_STS_COMPLETE)
+#define hash_ctx_status(ctx)     ((ctx)->status)
+#define hash_ctx_error(ctx)      ((ctx)->error)
+#define hash_ctx_init(ctx) \
+	do { \
+		(ctx)->error = HASH_CTX_ERROR_NONE; \
+		(ctx)->status = HASH_CTX_STS_COMPLETE; \
+	} while (0)
+
+
+/* Hash Constants and Typedefs */
+#define SHA1_DIGEST_LENGTH          5
+#define SHA1_LOG2_BLOCK_SIZE        6
+
+#define SHA1_PADLENGTHFIELD_SIZE    8
+
+#ifdef SHA_MB_DEBUG
+#define assert(expr) \
+do { \
+	if (unlikely(!(expr))) { \
+		printk(KERN_ERR "Assertion failed! %s,%s,%s,line=%d\n", \
+		#expr, __FILE__, __func__, __LINE__); \
+	} \
+} while (0)
+#else
+#define assert(expr) do {} while (0)
+#endif
+
+struct sha1_ctx_mgr {
+	struct sha1_mb_mgr mgr;
+};
+
+/* typedef struct sha1_ctx_mgr sha1_ctx_mgr; */
+
+struct sha1_hash_ctx {
+	/* Must be at struct offset 0 */
+	struct job_sha1       job;
+	/* status flag */
+	int status;
+	/* error flag */
+	int error;
+
+	uint32_t	total_length;
+	const void	*incoming_buffer;
+	uint32_t	incoming_buffer_length;
+	uint8_t		partial_block_buffer[SHA1_BLOCK_SIZE * 2];
+	uint32_t	partial_block_buffer_length;
+	void		*user_data;
+};
+
+#endif

arch/x86/crypto/sha-mb/sha_mb_mgr.h

+/*
+ * Header file for multi buffer SHA1 algorithm manager
+ *
+ * This file is provided under a dual BSD/GPLv2 license.  When using or
+ * redistributing this file, you may do so under either license.
+ *
+ * GPL LICENSE SUMMARY
+ *
+ *  Copyright(c) 2014 Intel Corporation.
+ *
+ *  This program is free software; you can redistribute it and/or modify
+ *  it under the terms of version 2 of the GNU General Public License as
+ *  published by the Free Software Foundation.
+ *
+ *  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.
+ *
+ *  Contact Information:
+ *      James Guilford <james.guilford@intel.com>
+ *	Tim Chen <tim.c.chen@linux.intel.com>
+ *
+ *  BSD LICENSE
+ *
+ *  Copyright(c) 2014 Intel Corporation.
+ *
+ *  Redistribution and use in source and binary forms, with or without
+ *  modification, are permitted provided that the following conditions
+ *  are met:
+ *
+ *    * Redistributions of source code must retain the above copyright
+ *      notice, this list of conditions and the following disclaimer.
+ *    * Redistributions in binary form must reproduce the above copyright
+ *      notice, this list of conditions and the following disclaimer in
+ *      the documentation and/or other materials provided with the
+ *      distribution.
+ *    * Neither the name of Intel Corporation nor the names of its
+ *      contributors may be used to endorse or promote products derived
+ *      from this software without specific prior written permission.
+ *
+ *  THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ *  "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ *  LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+ *  A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+ *  OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ *  SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+ *  LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+ *  DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+ *  THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ *  (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+ * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+#ifndef __SHA_MB_MGR_H
+#define __SHA_MB_MGR_H
+
+
+#include <linux/types.h>
+
+#define NUM_SHA1_DIGEST_WORDS 5
+
+enum job_sts {	STS_UNKNOWN = 0,
+		STS_BEING_PROCESSED = 1,
+		STS_COMPLETED = 2,
+		STS_INTERNAL_ERROR = 3,
+		STS_ERROR = 4
+};
+
+struct job_sha1 {
+	u8	*buffer;
+	u32	len;
+	u32	result_digest[NUM_SHA1_DIGEST_WORDS] __aligned(32);
+	enum	job_sts status;
+	void	*user_data;
+};
+
+/* SHA1 out-of-order scheduler */
+
+/* typedef uint32_t sha1_digest_array[5][8]; */
+
+struct sha1_args_x8 {
+	uint32_t	digest[5][8];
+	uint8_t		*data_ptr[8];
+};
+
+struct sha1_lane_data {
+	struct job_sha1 *job_in_lane;
+};
+
+struct sha1_mb_mgr {
+	struct sha1_args_x8 args;
+
+	uint32_t lens[8];
+
+	/* each byte is index (0...7) of unused lanes */
+	uint64_t unused_lanes;
+	/* byte 4 is set to FF as a flag */
+	struct sha1_lane_data ldata[8];
+};
+
+
+#define SHA1_MB_MGR_NUM_LANES_AVX2 8
+
+void sha1_mb_mgr_init_avx2(struct sha1_mb_mgr *state);
+struct job_sha1 *sha1_mb_mgr_submit_avx2(struct sha1_mb_mgr *state,
+					 struct job_sha1 *job);
+struct job_sha1 *sha1_mb_mgr_flush_avx2(struct sha1_mb_mgr *state);
+struct job_sha1 *sha1_mb_mgr_get_comp_job_avx2(struct sha1_mb_mgr *state);
+
+#endif

crypto/Kconfig

@@ -158,6 +158,20 @@ config CRYPTO_CRYPTD
 	  converts an arbitrary synchronous software crypto algorithm
 	  into an asynchronous algorithm that executes in a kernel thread.
 
+config CRYPTO_MCRYPTD
+	tristate "Software async multi-buffer crypto daemon"
+	select CRYPTO_BLKCIPHER
+	select CRYPTO_HASH
+	select CRYPTO_MANAGER
+	select CRYPTO_WORKQUEUE
+	help
+	  This is a generic software asynchronous crypto daemon that
+	  provides the kernel thread to assist multi-buffer crypto
+	  algorithms for submitting jobs and flushing jobs in multi-buffer
+	  crypto algorithms.  Multi-buffer crypto algorithms are executed
+	  in the context of this kernel thread and drivers can post
+	  their crypto request asynchronously to be processed by this daemon.
+
 config CRYPTO_AUTHENC
 	tristate "Authenc support"
 	select CRYPTO_AEAD
@@ -559,6 +573,22 @@ config CRYPTO_SHA1_PPC
 	  This is the powerpc hardware accelerated implementation of the
 	  SHA-1 secure hash standard (FIPS 180-1/DFIPS 180-2).
 
+config CRYPTO_SHA1_MB
+	tristate "SHA1 digest algorithm (x86_64 Multi-Buffer, Experimental)"
+	depends on X86 && 64BIT
+	select CRYPTO_SHA1
+	select CRYPTO_HASH
+	select CRYPTO_MCRYPTD
+	help
+	  SHA-1 secure hash standard (FIPS 180-1/DFIPS 180-2) implemented
+	  using multi-buffer technique.  This algorithm computes on
+	  multiple data lanes concurrently with SIMD instructions for
+	  better throughput.  It should not be enabled by default but
+	  used when there is significant amount of work to keep the keep
+	  the data lanes filled to get performance benefit.  If the data
+	  lanes remain unfilled, a flush operation will be initiated to
+	  process the crypto jobs, adding a slight latency.
+
 config CRYPTO_SHA256
 	tristate "SHA224 and SHA256 digest algorithm"
 	select CRYPTO_HASH

crypto/Makefile

@@ -60,6 +60,7 @@ obj-$(CONFIG_CRYPTO_GCM) += gcm.o
 obj-$(CONFIG_CRYPTO_CCM) += ccm.o
 obj-$(CONFIG_CRYPTO_PCRYPT) += pcrypt.o
 obj-$(CONFIG_CRYPTO_CRYPTD) += cryptd.o
+obj-$(CONFIG_CRYPTO_MCRYPTD) += mcryptd.o
 obj-$(CONFIG_CRYPTO_DES) += des_generic.o
 obj-$(CONFIG_CRYPTO_FCRYPT) += fcrypt.o
 obj-$(CONFIG_CRYPTO_BLOWFISH) += blowfish_generic.o

crypto/ahash.c

@@ -131,8 +131,10 @@ int crypto_hash_walk_first(struct ahash_request *req,
 {
 	walk->total = req->nbytes;
 
-	if (!walk->total)
+	if (!walk->total) {
+		walk->entrylen = 0;
 		return 0;
+	}
 
 	walk->alignmask = crypto_ahash_alignmask(crypto_ahash_reqtfm(req));
 	walk->sg = req->src;
@@ -147,8 +149,10 @@ int crypto_ahash_walk_first(struct ahash_request *req,
 {
 	walk->total = req->nbytes;
 
-	if (!walk->total)
+	if (!walk->total) {
+		walk->entrylen = 0;
 		return 0;
+	}
 
 	walk->alignmask = crypto_ahash_alignmask(crypto_ahash_reqtfm(req));
 	walk->sg = req->src;
@@ -167,8 +171,10 @@ int crypto_hash_walk_first_compat(struct hash_desc *hdesc,
 {
 	walk->total = len;
 
-	if (!walk->total)
+	if (!walk->total) {
+		walk->entrylen = 0;
 		return 0;
+	}
 
 	walk->alignmask = crypto_hash_alignmask(hdesc->tfm);
 	walk->sg = sg;

crypto/algif_skcipher.c

@@ -49,7 +49,7 @@ struct skcipher_ctx {
 	struct ablkcipher_request req;
 };
 
-#define MAX_SGL_ENTS ((PAGE_SIZE - sizeof(struct skcipher_sg_list)) / \
+#define MAX_SGL_ENTS ((4096 - sizeof(struct skcipher_sg_list)) / \
 		      sizeof(struct scatterlist) - 1)
 
 static inline int skcipher_sndbuf(struct sock *sk)

crypto/drbg.c

@@ -117,27 +117,18 @@ static const struct drbg_core drbg_cores[] = {
 	{
 		.flags = DRBG_CTR | DRBG_STRENGTH128,
 		.statelen = 32, /* 256 bits as defined in 10.2.1 */
-		.max_addtllen = 35,
-		.max_bits = 19,
-		.max_req = 48,
 		.blocklen_bytes = 16,
 		.cra_name = "ctr_aes128",
 		.backend_cra_name = "ecb(aes)",
 	}, {
 		.flags = DRBG_CTR | DRBG_STRENGTH192,
 		.statelen = 40, /* 320 bits as defined in 10.2.1 */
-		.max_addtllen = 35,
-		.max_bits = 19,
-		.max_req = 48,
 		.blocklen_bytes = 16,
 		.cra_name = "ctr_aes192",
 		.backend_cra_name = "ecb(aes)",
 	}, {
 		.flags = DRBG_CTR | DRBG_STRENGTH256,
 		.statelen = 48, /* 384 bits as defined in 10.2.1 */
-		.max_addtllen = 35,
-		.max_bits = 19,
-		.max_req = 48,
 		.blocklen_bytes = 16,
 		.cra_name = "ctr_aes256",
 		.backend_cra_name = "ecb(aes)",
@@ -147,36 +138,24 @@ static const struct drbg_core drbg_cores[] = {
 	{
 		.flags = DRBG_HASH | DRBG_STRENGTH128,
 		.statelen = 55, /* 440 bits */
-		.max_addtllen = 35,
-		.max_bits = 19,
-		.max_req = 48,
 		.blocklen_bytes = 20,
 		.cra_name = "sha1",
 		.backend_cra_name = "sha1",
 	}, {
 		.flags = DRBG_HASH | DRBG_STRENGTH256,
 		.statelen = 111, /* 888 bits */
-		.max_addtllen = 35,
-		.max_bits = 19,
-		.max_req = 48,
 		.blocklen_bytes = 48,
 		.cra_name = "sha384",
 		.backend_cra_name = "sha384",
 	}, {
 		.flags = DRBG_HASH | DRBG_STRENGTH256,
 		.statelen = 111, /* 888 bits */
-		.max_addtllen = 35,
-		.max_bits = 19,
-		.max_req = 48,
 		.blocklen_bytes = 64,
 		.cra_name = "sha512",
 		.backend_cra_name = "sha512",
 	}, {
 		.flags = DRBG_HASH | DRBG_STRENGTH256,
 		.statelen = 55, /* 440 bits */
-		.max_addtllen = 35,
-		.max_bits = 19,
-		.max_req = 48,
 		.blocklen_bytes = 32,
 		.cra_name = "sha256",
 		.backend_cra_name = "sha256",
@@ -186,36 +165,24 @@ static const struct drbg_core drbg_cores[] = {
 	{
 		.flags = DRBG_HMAC | DRBG_STRENGTH128,
 		.statelen = 20, /* block length of cipher */
-		.max_addtllen = 35,
-		.max_bits = 19,
-		.max_req = 48,
 		.blocklen_bytes = 20,
 		.cra_name = "hmac_sha1",
 		.backend_cra_name = "hmac(sha1)",
 	}, {
 		.flags = DRBG_HMAC | DRBG_STRENGTH256,
 		.statelen = 48, /* block length of cipher */
-		.max_addtllen = 35,
-		.max_bits = 19,
-		.max_req = 48,
 		.blocklen_bytes = 48,
 		.cra_name = "hmac_sha384",
 		.backend_cra_name = "hmac(sha384)",
 	}, {
 		.flags = DRBG_HMAC | DRBG_STRENGTH256,
 		.statelen = 64, /* block length of cipher */
-		.max_addtllen = 35,
-		.max_bits = 19,
-		.max_req = 48,
 		.blocklen_bytes = 64,
 		.cra_name = "hmac_sha512",
 		.backend_cra_name = "hmac(sha512)",
 	}, {
 		.flags = DRBG_HMAC | DRBG_STRENGTH256,
 		.statelen = 32, /* block length of cipher */
-		.max_addtllen = 35,
-		.max_bits = 19,
-		.max_req = 48,
 		.blocklen_bytes = 32,
 		.cra_name = "hmac_sha256",
 		.backend_cra_name = "hmac(sha256)",
@@ -302,20 +269,19 @@ static bool drbg_fips_continuous_test(struct drbg_state *drbg,
  * Convert an integer into a byte representation of this integer.
  * The byte representation is big-endian
  *
- * @buf buffer holding the converted integer
  * @val value to be converted
- * @buflen length of buffer
+ * @buf buffer holding the converted integer -- caller must ensure that
+ *      buffer size is at least 32 bit
  */
 #if (defined(CONFIG_CRYPTO_DRBG_HASH) || defined(CONFIG_CRYPTO_DRBG_CTR))
-static inline void drbg_int2byte(unsigned char *buf, uint64_t val,
-				 size_t buflen)
+static inline void drbg_cpu_to_be32(__u32 val, unsigned char *buf)
 {
-	unsigned char *byte;
-	uint64_t i;
+	struct s {
+		__be32 conv;
+	};
+	struct s *conversion = (struct s *) buf;
 
-	byte = buf + (buflen - 1);
-	for (i = 0; i < buflen; i++)
-		*(byte--) = val >> (i * 8) & 0xff;
+	conversion->conv = cpu_to_be32(val);
 }
 
 /*
@@ -483,10 +449,10 @@ static int drbg_ctr_df(struct drbg_state *drbg,
 	/* 10.4.2 step 2 -- calculate the entire length of all input data */
 	list_for_each_entry(seed, seedlist, list)
 		inputlen += seed->len;
-	drbg_int2byte(&L_N[0], inputlen, 4);
+	drbg_cpu_to_be32(inputlen, &L_N[0]);
 
 	/* 10.4.2 step 3 */
-	drbg_int2byte(&L_N[4], bytes_to_return, 4);
+	drbg_cpu_to_be32(bytes_to_return, &L_N[4]);
 
 	/* 10.4.2 step 5: length is L_N, input_string, one byte, padding */
 	padlen = (inputlen + sizeof(L_N) + 1) % (drbg_blocklen(drbg));
@@ -517,7 +483,7 @@ static int drbg_ctr_df(struct drbg_state *drbg,
 		 * holds zeros after allocation -- even the increment of i
 		 * is irrelevant as the increment remains within length of i
 		 */
-		drbg_int2byte(iv, i, 4);
+		drbg_cpu_to_be32(i, iv);
 		/* 10.4.2 step 9.2 -- BCC and concatenation with temp */
 		ret = drbg_ctr_bcc(drbg, temp + templen, K, &bcc_list);
 		if (ret)
@@ -729,11 +695,9 @@ static int drbg_hmac_update(struct drbg_state *drbg, struct list_head *seed,
 	LIST_HEAD(seedlist);
 	LIST_HEAD(vdatalist);
 
-	if (!reseed) {
-		/* 10.1.2.3 step 2 */
-		memset(drbg->C, 0, drbg_statelen(drbg));
+	if (!reseed)
+		/* 10.1.2.3 step 2 -- memset(0) of C is implicit with kzalloc */
 		memset(drbg->V, 1, drbg_statelen(drbg));
-	}
 
 	drbg_string_fill(&seed1, drbg->V, drbg_statelen(drbg));
 	list_add_tail(&seed1.list, &seedlist);
@@ -862,7 +826,7 @@ static int drbg_hash_df(struct drbg_state *drbg,
 
 	/* 10.4.1 step 3 */
 	input[0] = 1;
-	drbg_int2byte(&input[1], (outlen * 8), 4);
+	drbg_cpu_to_be32((outlen * 8), &input[1]);
 
 	/* 10.4.1 step 4.1 -- concatenation of data for input into hash */
 	drbg_string_fill(&data, input, 5);
@@ -1023,7 +987,10 @@ static int drbg_hash_generate(struct drbg_state *drbg,
 {
 	int len = 0;
 	int ret = 0;
-	unsigned char req[8];
+	union {
+		unsigned char req[8];
+		__be64 req_int;
+	} u;
 	unsigned char prefix = DRBG_PREFIX3;
 	struct drbg_string data1, data2;
 	LIST_HEAD(datalist);
@@ -1053,8 +1020,8 @@ static int drbg_hash_generate(struct drbg_state *drbg,
 		     drbg->scratchpad, drbg_blocklen(drbg));
 	drbg_add_buf(drbg->V, drbg_statelen(drbg),
 		     drbg->C, drbg_statelen(drbg));
-	drbg_int2byte(req, drbg->reseed_ctr, sizeof(req));
-	drbg_add_buf(drbg->V, drbg_statelen(drbg), req, 8);
+	u.req_int = cpu_to_be64(drbg->reseed_ctr);
+	drbg_add_buf(drbg->V, drbg_statelen(drbg), u.req, 8);
 
 out:
 	memset(drbg->scratchpad, 0, drbg_blocklen(drbg));
@@ -1142,6 +1109,11 @@ static int drbg_seed(struct drbg_state *drbg, struct drbg_string *pers,
 		pr_devel("DRBG: using personalization string\n");
 	}
 
+	if (!reseed) {
+		memset(drbg->V, 0, drbg_statelen(drbg));
+		memset(drbg->C, 0, drbg_statelen(drbg));
+	}
+
 	ret = drbg->d_ops->update(drbg, &seedlist, reseed);
 	if (ret)
 		goto out;
@@ -1151,8 +1123,7 @@ static int drbg_seed(struct drbg_state *drbg, struct drbg_string *pers,
 	drbg->reseed_ctr = 1;
 
 out:
-	if (entropy)
-		kzfree(entropy);
+	kzfree(entropy);
 	return ret;
 }
 
@@ -1161,19 +1132,15 @@ static inline void drbg_dealloc_state(struct drbg_state *drbg)
 {
 	if (!drbg)
 		return;
-	if (drbg->V)
-		kzfree(drbg->V);
+	kzfree(drbg->V);
 	drbg->V = NULL;
-	if (drbg->C)
-		kzfree(drbg->C);
+	kzfree(drbg->C);
 	drbg->C = NULL;
-	if (drbg->scratchpad)
-		kzfree(drbg->scratchpad);
+	kzfree(drbg->scratchpad);
 	drbg->scratchpad = NULL;
 	drbg->reseed_ctr = 0;
 #ifdef CONFIG_CRYPTO_FIPS
-	if (drbg->prev)
-		kzfree(drbg->prev);
+	kzfree(drbg->prev);
 	drbg->prev = NULL;
 	drbg->fips_primed = false;
 #endif
@@ -1188,17 +1155,14 @@ static inline int drbg_alloc_state(struct drbg_state *drbg)
 	int ret = -ENOMEM;
 	unsigned int sb_size = 0;
 
-	if (!drbg)
-		return -EINVAL;
-
-	drbg->V = kzalloc(drbg_statelen(drbg), GFP_KERNEL);
+	drbg->V = kmalloc(drbg_statelen(drbg), GFP_KERNEL);
 	if (!drbg->V)
 		goto err;
-	drbg->C = kzalloc(drbg_statelen(drbg), GFP_KERNEL);
+	drbg->C = kmalloc(drbg_statelen(drbg), GFP_KERNEL);
 	if (!drbg->C)
 		goto err;
 #ifdef CONFIG_CRYPTO_FIPS
-	drbg->prev = kzalloc(drbg_blocklen(drbg), GFP_KERNEL);
+	drbg->prev = kmalloc(drbg_blocklen(drbg), GFP_KERNEL);
 	if (!drbg->prev)
 		goto err;
 	drbg->fips_primed = false;
@@ -1263,15 +1227,6 @@ static int drbg_make_shadow(struct drbg_state *drbg, struct drbg_state **shadow)
 	int ret = -ENOMEM;
 	struct drbg_state *tmp = NULL;
 
-	if (!drbg || !drbg->core || !drbg->V || !drbg->C) {
-		pr_devel("DRBG: attempt to generate shadow copy for "
-			 "uninitialized DRBG state rejected\n");
-		return -EINVAL;
-	}
-	/* HMAC does not have a scratchpad */
-	if (!(drbg->core->flags & DRBG_HMAC) && NULL == drbg->scratchpad)
-		return -EINVAL;
-
 	tmp = kzalloc(sizeof(struct drbg_state), GFP_KERNEL);
 	if (!tmp)
 		return -ENOMEM;
@@ -1293,8 +1248,7 @@ static int drbg_make_shadow(struct drbg_state *drbg, struct drbg_state **shadow)
 	return 0;
 
 err:
-	if (tmp)
-		kzfree(tmp);
+	kzfree(tmp);
 	return ret;
 }
 
@@ -1385,11 +1339,9 @@ static int drbg_generate(struct drbg_state *drbg,
 		shadow->seeded = false;
 
 	/* allocate cipher handle */
-	if (shadow->d_ops->crypto_init) {
-		len = shadow->d_ops->crypto_init(shadow);
-		if (len)
-			goto err;
-	}
+	len = shadow->d_ops->crypto_init(shadow);
+	if (len)
+		goto err;
 
 	if (shadow->pr || !shadow->seeded) {
 		pr_devel("DRBG: reseeding before generation (prediction "
@@ -1471,8 +1423,7 @@ static int drbg_generate(struct drbg_state *drbg,
 #endif
 
 err:
-	if (shadow->d_ops->crypto_fini)
-		shadow->d_ops->crypto_fini(shadow);
+	shadow->d_ops->crypto_fini(shadow);
 	drbg_restore_shadow(drbg, &shadow);
 	return len;
 }
@@ -1566,11 +1517,10 @@ static int drbg_instantiate(struct drbg_state *drbg, struct drbg_string *pers,
 		return ret;
 
 	ret = -EFAULT;
-	if (drbg->d_ops->crypto_init && drbg->d_ops->crypto_init(drbg))
+	if (drbg->d_ops->crypto_init(drbg))
 		goto err;
 	ret = drbg_seed(drbg, pers, false);
-	if (drbg->d_ops->crypto_fini)
-		drbg->d_ops->crypto_fini(drbg);
+	drbg->d_ops->crypto_fini(drbg);
 	if (ret)
 		goto err;
 

crypto/lz4.c

@@ -68,7 +68,7 @@ static int lz4_decompress_crypto(struct crypto_tfm *tfm, const u8 *src,
 	size_t tmp_len = *dlen;
 	size_t __slen = slen;
 
-	err = lz4_decompress(src, &__slen, dst, tmp_len);
+	err = lz4_decompress_unknownoutputsize(src, __slen, dst, &tmp_len);
 	if (err < 0)
 		return -EINVAL;
 

crypto/lz4hc.c

@@ -68,7 +68,7 @@ static int lz4hc_decompress_crypto(struct crypto_tfm *tfm, const u8 *src,
 	size_t tmp_len = *dlen;
 	size_t __slen = slen;
 
-	err = lz4_decompress(src, &__slen, dst, tmp_len);
+	err = lz4_decompress_unknownoutputsize(src, __slen, dst, &tmp_len);
 	if (err < 0)
 		return -EINVAL;
 

crypto/sha256_generic.c

@@ -24,6 +24,7 @@
 #include <linux/types.h>
 #include <crypto/sha.h>
 #include <asm/byteorder.h>
+#include <asm/unaligned.h>
 
 static inline u32 Ch(u32 x, u32 y, u32 z)
 {
@@ -42,7 +43,7 @@ static inline u32 Maj(u32 x, u32 y, u32 z)
 
 static inline void LOAD_OP(int I, u32 *W, const u8 *input)
 {
-	W[I] = __be32_to_cpu( ((__be32*)(input))[I] );
+	W[I] = get_unaligned_be32((__u32 *)input + I);
 }
 
 static inline void BLEND_OP(int I, u32 *W)

crypto/sha512_generic.c

@@ -20,6 +20,7 @@
 #include <crypto/sha.h>
 #include <linux/percpu.h>
 #include <asm/byteorder.h>
+#include <asm/unaligned.h>
 
 static inline u64 Ch(u64 x, u64 y, u64 z)
 {
@@ -68,7 +69,7 @@ static const u64 sha512_K[80] = {
 
 static inline void LOAD_OP(int I, u64 *W, const u8 *input)
 {
-	W[I] = __be64_to_cpu( ((__be64*)(input))[I] );
+	W[I] = get_unaligned_be64((__u64 *)input + I);
 }
 
 static inline void BLEND_OP(int I, u64 *W)

drivers/char/hw_random/Kconfig

@@ -333,6 +333,19 @@ config HW_RANDOM_MSM
 
 	  If unsure, say Y.
 
+config HW_RANDOM_XGENE
+	tristate "APM X-Gene True Random Number Generator (TRNG) support"
+	depends on HW_RANDOM && ARCH_XGENE
+	default HW_RANDOM
+	---help---
+	  This driver provides kernel-side support for the Random Number
+	  Generator hardware found on APM X-Gene SoC.
+
+	  To compile this driver as a module, choose M here: the
+	  module will be called xgene_rng.
+
+	  If unsure, say Y.
+
 endif # HW_RANDOM
 
 config UML_RANDOM

drivers/char/hw_random/Makefile

@@ -29,3 +29,4 @@ obj-$(CONFIG_HW_RANDOM_EXYNOS)	+= exynos-rng.o
 obj-$(CONFIG_HW_RANDOM_TPM) += tpm-rng.o
 obj-$(CONFIG_HW_RANDOM_BCM2835) += bcm2835-rng.o
 obj-$(CONFIG_HW_RANDOM_MSM) += msm-rng.o
+obj-$(CONFIG_HW_RANDOM_XGENE) += xgene-rng.o

drivers/char/hw_random/amd-rng.c

@@ -142,10 +142,10 @@ static int __init mod_init(void)
 	amd_rng.priv = (unsigned long)pmbase;
 	amd_pdev = pdev;
 
-	printk(KERN_INFO "AMD768 RNG detected\n");
+	pr_info("AMD768 RNG detected\n");
 	err = hwrng_register(&amd_rng);
 	if (err) {
-		printk(KERN_ERR PFX "RNG registering failed (%d)\n",
+		pr_err(PFX "RNG registering failed (%d)\n",
 		       err);
 		release_region(pmbase + 0xF0, 8);
 		goto out;

drivers/char/hw_random/pasemi-rng.c

@@ -113,7 +113,7 @@ static int rng_probe(struct platform_device *ofdev)
 
 	pasemi_rng.priv = (unsigned long)rng_regs;
 
-	printk(KERN_INFO "Registering PA Semi RNG\n");
+	pr_info("Registering PA Semi RNG\n");
 
 	err = hwrng_register(&pasemi_rng);