Import the skein hashing algorithm, based on the threefish block cipher

Connect it to userland (libmd, libcrypt, sbin/md5) and kernel (crypto.ko)

Support for skein as a ZFS checksum algorithm was introduced in r289422
but is disconnected because FreeBSD lacked a Skein implementation.

A further commit will enable it in ZFS.

Reviewed by:	cem
Sponsored by:	ScaleEngine Inc.
Differential Revision:	https://reviews.freebsd.org/D6166

10 files changed, 3963 insertions, 0 deletions

diff --git a/sys/crypto/skein/amd64/skein_block_asm.s b/sys/crypto/skein/amd64/skein_block_asm.s
new file mode 100644
index 0000000..b2d0a83
--- /dev/null
+++ b/sys/crypto/skein/amd64/skein_block_asm.s
@@ -0,0 +1,1328 @@
+#
+#----------------------------------------------------------------
+# 64-bit x86 assembler code (gnu as) for Skein block functions
+#
+# Author: Doug Whiting, Hifn/Exar
+#
+# This code is released to the public domain.
+#----------------------------------------------------------------
+#
+    .text
+    .altmacro
+    .psize 0,128                            #list file has no page boundaries
+#
+_MASK_ALL_  =  (256+512+1024)               #all three algorithm bits
+_MAX_FRAME_ =  240
+#
+#################
+.ifndef SKEIN_USE_ASM
+_USE_ASM_         = _MASK_ALL_
+.else
+_USE_ASM_         = SKEIN_USE_ASM
+.endif
+#################
+.ifndef SKEIN_LOOP                          #configure loop unrolling
+_SKEIN_LOOP       =   2                     #default is fully unrolled for 256/512, twice for 1024
+.else
+_SKEIN_LOOP       = SKEIN_LOOP
+  .irp _NN_,%_SKEIN_LOOP                #only display loop unrolling if default changed on command line
+.print  "+++ SKEIN_LOOP = \_NN_"
+  .endr
+.endif
+# the unroll counts (0 --> fully unrolled)
+SKEIN_UNROLL_256  = (_SKEIN_LOOP / 100) % 10
+SKEIN_UNROLL_512  = (_SKEIN_LOOP /  10) % 10
+SKEIN_UNROLL_1024 = (_SKEIN_LOOP      ) % 10
+#
+SKEIN_ASM_UNROLL  = 0
+  .irp _NN_,256,512,1024
+    .if (SKEIN_UNROLL_\_NN_) == 0
+SKEIN_ASM_UNROLL  = SKEIN_ASM_UNROLL + \_NN_
+    .endif
+  .endr
+#################
+#
+.ifndef SKEIN_ROUNDS
+ROUNDS_256  =   72
+ROUNDS_512  =   72
+ROUNDS_1024 =   80
+.else
+ROUNDS_256  = 8*((((SKEIN_ROUNDS / 100) + 5) % 10) + 5)
+ROUNDS_512  = 8*((((SKEIN_ROUNDS /  10) + 5) % 10) + 5)
+ROUNDS_1024 = 8*((((SKEIN_ROUNDS      ) + 5) % 10) + 5)
+# only display rounds if default size is changed on command line
+.irp _NN_,256,512,1024
+  .if _USE_ASM_ && \_NN_
+    .irp _RR_,%(ROUNDS_\_NN_)
+      .if _NN_ < 1024
+.print  "+++ SKEIN_ROUNDS_\_NN_  = \_RR_"
+      .else
+.print  "+++ SKEIN_ROUNDS_\_NN_ = \_RR_"
+      .endif
+    .endr
+  .endif
+.endr
+.endif
+#################
+#
+.ifdef SKEIN_CODE_SIZE
+_SKEIN_CODE_SIZE = (1)
+.else
+.ifdef  SKEIN_PERF                           #use code size if SKEIN_PERF is defined
+_SKEIN_CODE_SIZE = (1)
+.else
+_SKEIN_CODE_SIZE = (0)
+.endif
+.endif
+#
+#################
+#
+.ifndef SKEIN_DEBUG
+_SKEIN_DEBUG      = 0
+.else
+_SKEIN_DEBUG      = 1
+.endif
+#################
+#
+# define offsets of fields in hash context structure
+#
+HASH_BITS   =   0                   #bits of hash output
+BCNT        =   8 + HASH_BITS       #number of bytes in BUFFER[]
+TWEAK       =   8 + BCNT            #tweak values[0..1]
+X_VARS      =  16 + TWEAK           #chaining vars
+#
+#(Note: buffer[] in context structure is NOT needed here :-)
+#
+KW_PARITY   =   0x1BD11BDAA9FC1A22  #overall parity of key schedule words
+FIRST_MASK  =   ~ (1 <<  6)
+FIRST_MASK64=   ~ (1 << 62)
+#
+# rotation constants for Skein
+#
+RC_256_0_0  = 14
+RC_256_0_1  = 16
+
+RC_256_1_0  = 52
+RC_256_1_1  = 57
+
+RC_256_2_0  = 23
+RC_256_2_1  = 40
+
+RC_256_3_0  =  5
+RC_256_3_1  = 37
+
+RC_256_4_0  = 25
+RC_256_4_1  = 33
+
+RC_256_5_0  = 46
+RC_256_5_1  = 12
+
+RC_256_6_0  = 58
+RC_256_6_1  = 22
+
+RC_256_7_0  = 32
+RC_256_7_1  = 32
+
+RC_512_0_0  = 46
+RC_512_0_1  = 36
+RC_512_0_2  = 19
+RC_512_0_3  = 37
+
+RC_512_1_0  = 33
+RC_512_1_1  = 27
+RC_512_1_2  = 14
+RC_512_1_3  = 42
+
+RC_512_2_0  = 17
+RC_512_2_1  = 49
+RC_512_2_2  = 36
+RC_512_2_3  = 39
+
+RC_512_3_0  = 44
+RC_512_3_1  =  9
+RC_512_3_2  = 54
+RC_512_3_3  = 56
+
+RC_512_4_0  = 39
+RC_512_4_1  = 30
+RC_512_4_2  = 34
+RC_512_4_3  = 24
+
+RC_512_5_0  = 13
+RC_512_5_1  = 50
+RC_512_5_2  = 10
+RC_512_5_3  = 17
+
+RC_512_6_0  = 25
+RC_512_6_1  = 29
+RC_512_6_2  = 39
+RC_512_6_3  = 43
+
+RC_512_7_0  =  8
+RC_512_7_1  = 35
+RC_512_7_2  = 56
+RC_512_7_3  = 22
+
+RC_1024_0_0 = 24
+RC_1024_0_1 = 13
+RC_1024_0_2 =  8
+RC_1024_0_3 = 47
+RC_1024_0_4 =  8
+RC_1024_0_5 = 17
+RC_1024_0_6 = 22
+RC_1024_0_7 = 37
+
+RC_1024_1_0 = 38
+RC_1024_1_1 = 19
+RC_1024_1_2 = 10
+RC_1024_1_3 = 55
+RC_1024_1_4 = 49
+RC_1024_1_5 = 18
+RC_1024_1_6 = 23
+RC_1024_1_7 = 52
+
+RC_1024_2_0 = 33
+RC_1024_2_1 =  4
+RC_1024_2_2 = 51
+RC_1024_2_3 = 13
+RC_1024_2_4 = 34
+RC_1024_2_5 = 41
+RC_1024_2_6 = 59
+RC_1024_2_7 = 17
+
+RC_1024_3_0 =  5
+RC_1024_3_1 = 20
+RC_1024_3_2 = 48
+RC_1024_3_3 = 41
+RC_1024_3_4 = 47
+RC_1024_3_5 = 28
+RC_1024_3_6 = 16
+RC_1024_3_7 = 25
+
+RC_1024_4_0 = 41
+RC_1024_4_1 =  9
+RC_1024_4_2 = 37
+RC_1024_4_3 = 31
+RC_1024_4_4 = 12
+RC_1024_4_5 = 47
+RC_1024_4_6 = 44
+RC_1024_4_7 = 30
+
+RC_1024_5_0 = 16
+RC_1024_5_1 = 34
+RC_1024_5_2 = 56
+RC_1024_5_3 = 51
+RC_1024_5_4 =  4
+RC_1024_5_5 = 53
+RC_1024_5_6 = 42
+RC_1024_5_7 = 41
+
+RC_1024_6_0 = 31
+RC_1024_6_1 = 44
+RC_1024_6_2 = 47
+RC_1024_6_3 = 46
+RC_1024_6_4 = 19
+RC_1024_6_5 = 42
+RC_1024_6_6 = 44
+RC_1024_6_7 = 25
+
+RC_1024_7_0 =  9
+RC_1024_7_1 = 48
+RC_1024_7_2 = 35
+RC_1024_7_3 = 52
+RC_1024_7_4 = 23
+RC_1024_7_5 = 31
+RC_1024_7_6 = 37
+RC_1024_7_7 = 20
+#
+#  Input:  reg
+# Output: <reg> <<< RC_BlkSize_roundNum_mixNum, BlkSize=256/512/1024
+#
+.macro RotL64   reg,BLK_SIZE,ROUND_NUM,MIX_NUM
+_RCNT_ = RC_\BLK_SIZE&_\ROUND_NUM&_\MIX_NUM
+  .if _RCNT_  #is there anything to do?
+    rolq    $_RCNT_,%\reg
+  .endif
+.endm
+#
+#----------------------------------------------------------------
+#
+# MACROS: define local vars and configure stack
+#
+#----------------------------------------------------------------
+# declare allocated space on the stack
+.macro StackVar localName,localSize
+\localName  =   _STK_OFFS_
+_STK_OFFS_  =   _STK_OFFS_+(\localSize)
+.endm #StackVar
+#
+#----------------------------------------------------------------
+#
+# MACRO: Configure stack frame, allocate local vars
+#
+.macro Setup_Stack BLK_BITS,KS_CNT,debugCnt
+    WCNT    =    (\BLK_BITS)/64
+#
+_PushCnt_   =   0                   #save nonvolatile regs on stack
+  .irp _reg_,rbp,rbx,r12,r13,r14,r15
+       pushq    %\_reg_
+_PushCnt_ = _PushCnt_ + 1           #track count to keep alignment
+  .endr
+#
+_STK_OFFS_  =   0                   #starting offset from rsp
+    #---- local  variables         #<-- rsp
+    StackVar    X_stk  ,8*(WCNT)    #local context vars
+    StackVar    ksTwk  ,8*3         #key schedule: tweak words
+    StackVar    ksKey  ,8*(WCNT)+8  #key schedule: key   words
+  .if (SKEIN_ASM_UNROLL && (\BLK_BITS)) == 0
+    StackVar    ksRot ,16*(\KS_CNT) #leave space for "rotation" to happen
+  .endif
+    StackVar    Wcopy  ,8*(WCNT)    #copy of input block    
+  .if _SKEIN_DEBUG
+  .if \debugCnt + 0                 #temp location for debug X[] info
+    StackVar    xDebug_\BLK_BITS ,8*(\debugCnt)
+  .endif
+  .endif
+  .if ((8*_PushCnt_ + _STK_OFFS_) % 8) == 0
+    StackVar    align16,8           #keep 16-byte aligned (adjust for retAddr?)
+tmpStk_\BLK_BITS = align16          #use this
+  .endif
+    #---- saved caller parameters (from regs rdi, rsi, rdx, rcx)
+    StackVar    ctxPtr ,8           #context ptr
+    StackVar    blkPtr ,8           #pointer to block data
+    StackVar    blkCnt ,8           #number of full blocks to process
+    StackVar    bitAdd ,8           #bit count to add to tweak
+LOCAL_SIZE  =   _STK_OFFS_          #size of "local" vars
+    #---- 
+    StackVar    savRegs,8*_PushCnt_ #saved registers
+    StackVar    retAddr,8           #return address
+    #---- caller's stack frame (aligned mod 16)
+#
+# set up the stack frame pointer (rbp)
+#
+FRAME_OFFS  =   ksTwk + 128         #allow short (negative) offset to ksTwk, kwKey
+  .if FRAME_OFFS > _STK_OFFS_       #keep rbp in the "locals" range
+FRAME_OFFS  =      _STK_OFFS_
+  .endif
+F_O         =   -FRAME_OFFS
+#
+  #put some useful defines in the .lst file (for grep)
+__STK_LCL_SIZE_\BLK_BITS = LOCAL_SIZE
+__STK_TOT_SIZE_\BLK_BITS = _STK_OFFS_
+__STK_FRM_OFFS_\BLK_BITS = FRAME_OFFS
+#
+# Notes on stack frame setup:
+#   * the most frequently used variable is X_stk[], based at [rsp+0]
+#   * the next most used is the key schedule arrays, ksKey and ksTwk
+#       so rbp is "centered" there, allowing short offsets to the key 
+#       schedule even in 1024-bit Skein case
+#   * the Wcopy variables are infrequently accessed, but they have long 
+#       offsets from both rsp and rbp only in the 1024-bit case.
+#   * all other local vars and calling parameters can be accessed 
+#       with short offsets, except in the 1024-bit case
+#
+    subq    $LOCAL_SIZE,%rsp        #make room for the locals
+    leaq    FRAME_OFFS(%rsp),%rbp   #maximize use of short offsets
+    movq    %rdi, ctxPtr+F_O(%rbp)  #save caller's parameters on the stack
+    movq    %rsi, blkPtr+F_O(%rbp)
+    movq    %rdx, blkCnt+F_O(%rbp)
+    movq    %rcx, bitAdd+F_O(%rbp)
+#
+.endm #Setup_Stack
+#
+#----------------------------------------------------------------
+#
+.macro Reset_Stack
+    addq    $LOCAL_SIZE,%rsp        #get rid of locals (wipe??)
+  .irp _reg_,r15,r14,r13,r12,rbx,rbp
+    popq    %\_reg_                 #restore caller's regs
+_PushCnt_ = _PushCnt_ - 1
+  .endr
+  .if _PushCnt_
+    .error  "Mismatched push/pops?"
+  .endif
+.endm # Reset_Stack
+#
+#----------------------------------------------------------------
+# macros to help debug internals
+#
+.if _SKEIN_DEBUG
+    .extern  Skein_Show_Block     #calls to C routines
+    .extern  Skein_Show_Round
+#
+SKEIN_RND_SPECIAL       =   1000
+SKEIN_RND_KEY_INITIAL   =   SKEIN_RND_SPECIAL+0
+SKEIN_RND_KEY_INJECT    =   SKEIN_RND_SPECIAL+1
+SKEIN_RND_FEED_FWD      =   SKEIN_RND_SPECIAL+2
+#
+.macro Skein_Debug_Block BLK_BITS
+#
+#void Skein_Show_Block(uint_t bits,const Skein_Ctxt_Hdr_t *h,const u64b_t *X,
+#                     const u08b_t *blkPtr, const u64b_t *wPtr, 
+#                     const u64b_t *ksPtr,const u64b_t *tsPtr)
+#
+_NN_ = 0
+  .irp _reg_,rax,rcx,rdx,rsi,rdi,r8,r9,r10,r11
+    pushq   %\_reg_                 #save all volatile regs on tack before the call
+_NN_ = _NN_ + 1
+  .endr
+    # get and push call parameters
+    movq    $\BLK_BITS      ,%rdi   #bits
+    movq    ctxPtr+F_O(%rbp),%rsi   #h (pointer)
+    leaq    X_VARS    (%rsi),%rdx   #X (pointer)
+    movq    blkPtr+F_O(%rbp),%rcx   #blkPtr
+    leaq    Wcopy +F_O(%rbp),%r8    #wPtr
+    leaq    ksKey +F_O(%rbp),%r9    #key pointer
+    leaq    ksTwk +F_O(%rbp),%rax   #tweak pointer
+    pushq   %rax                    #   (pass on the stack)
+    call    Skein_Show_Block        #call external debug handler
+    addq    $8*1,%rsp               #discard parameters on stack
+  .if (_NN_ % 2 ) == 0              #check stack alignment
+    .error "Stack misalignment problem in Skein_Debug_Block_\_BLK_BITS"
+  .endif
+  .irp _reg_,r11,r10,r9,r8,rdi,rsi,rdx,rcx,rax
+    popq    %\_reg_                 #restore regs
+_NN_ = _NN_ - 1
+  .endr
+  .if _NN_
+    .error "Push/pop mismatch problem in Skein_Debug_Block_\_BLK_BITS"
+  .endif
+.endm # Skein_Debug_Block
+#
+# the macro to "call" to debug a round
+#
+.macro Skein_Debug_Round BLK_BITS,R,RDI_OFFS,afterOp
+    # call the appropriate (local) debug "function"
+    pushq   %rdx                    #save rdx, so we can use it for round "number"
+  .if (SKEIN_ASM_UNROLL && \BLK_BITS) || (\R >= SKEIN_RND_SPECIAL)
+    movq    $\R,%rdx
+  .else                             #compute round number using edi
+_rOffs_ = \RDI_OFFS + 0
+   .if \BLK_BITS == 1024
+    movq    rIdx_offs+8(%rsp),%rdx  #get rIdx off the stack (adjust for pushq rdx above)
+    leaq    1+(((\R)-1) && 3)+_rOffs_(,%rdx,4),%rdx
+   .else
+    leaq    1+(((\R)-1) && 3)+_rOffs_(,%rdi,4),%rdx
+   .endif
+  .endif
+    call    Skein_Debug_Round_\BLK_BITS
+    popq    %rdx                    #restore origianl rdx value
+#
+    afterOp
+.endm  #  Skein_Debug_Round
+.else  #------- _SKEIN_DEBUG (dummy macros if debug not enabled)
+.macro Skein_Debug_Block BLK_BITS
+.endm
+#
+.macro Skein_Debug_Round BLK_BITS,R,RDI_OFFS,afterOp
+.endm
+#
+.endif # _SKEIN_DEBUG
+#
+#----------------------------------------------------------------
+#
+.macro  addReg dstReg,srcReg_A,srcReg_B,useAddOp,immOffs
+  .if \immOffs + 0
+       leaq    \immOffs(%\srcReg_A\srcReg_B,%\dstReg),%\dstReg
+  .elseif ((\useAddOp + 0) == 0)
+    .ifndef ASM_NO_LEA  #lea seems to be faster on Core 2 Duo CPUs!
+       leaq   (%\srcReg_A\srcReg_B,%\dstReg),%\dstReg
+    .else
+       addq    %\srcReg_A\srcReg_B,%\dstReg
+    .endif
+  .else
+       addq    %\srcReg_A\srcReg_B,%\dstReg
+  .endif
+.endm
+
+# keep Intel-style ordering here, to match addReg
+.macro  xorReg dstReg,srcReg_A,srcReg_B
+        xorq   %\srcReg_A\srcReg_B,%\dstReg
+.endm
+#
+#----------------------------------------------------------------
+#
+.macro C_label lName
+ \lName:        #use both "genders" to work across linkage conventions
+_\lName:
+    .global  \lName
+    .global _\lName
+.endm
+#
+#=================================== Skein_256 =============================================
+#
+.if _USE_ASM_ & 256
+#
+# void Skein_256_Process_Block(Skein_256_Ctxt_t *ctx,const u08b_t *blkPtr,size_t blkCnt,size_t bitcntAdd)#
+#
+#################
+#
+# code
+#
+C_label Skein_256_Process_Block
+    Setup_Stack 256,((ROUNDS_256/8)+1)
+    movq    TWEAK+8(%rdi),%r14
+    jmp     Skein_256_block_loop
+    .p2align 4
+    # main hash loop for Skein_256
+Skein_256_block_loop:
+    #
+    # general register usage:
+    #   RAX..RDX        = X0..X3    
+    #   R08..R12        = ks[0..4]
+    #   R13..R15        = ts[0..2]
+    #   RSP, RBP        = stack/frame pointers
+    #   RDI             = round counter or context pointer
+    #   RSI             = temp
+    #
+    movq    TWEAK+0(%rdi)     ,%r13
+    addq    bitAdd+F_O(%rbp)  ,%r13  #computed updated tweak value T0
+    movq    %r14              ,%r15
+    xorq    %r13              ,%r15  #now %r13.%r15 is set as the tweak 
+
+    movq    $KW_PARITY        ,%r12
+    movq       X_VARS+ 0(%rdi),%r8
+    movq       X_VARS+ 8(%rdi),%r9 
+    movq       X_VARS+16(%rdi),%r10
+    movq       X_VARS+24(%rdi),%r11
+    movq    %r13,TWEAK+0(%rdi)       #save updated tweak value ctx->h.T[0]
+    xorq    %r8               ,%r12  #start accumulating overall parity
+
+    movq    blkPtr +F_O(%rbp) ,%rsi  #esi --> input block
+    xorq    %r9               ,%r12
+    movq     0(%rsi)          ,%rax  #get X[0..3]
+    xorq    %r10              ,%r12
+    movq     8(%rsi)          ,%rbx
+    xorq    %r11              ,%r12
+    movq    16(%rsi)          ,%rcx
+    movq    24(%rsi)          ,%rdx
+
+    movq    %rax,Wcopy+ 0+F_O(%rbp)  #save copy of input block
+    movq    %rbx,Wcopy+ 8+F_O(%rbp)    
+    movq    %rcx,Wcopy+16+F_O(%rbp)    
+    movq    %rdx,Wcopy+24+F_O(%rbp)    
+
+    addq    %r8 ,%rax                #initial key injection
+    addq    %r9 ,%rbx 
+    addq    %r10,%rcx
+    addq    %r11,%rdx
+    addq    %r13,%rbx
+    addq    %r14,%rcx
+
+.if _SKEIN_DEBUG
+    movq    %r14,TWEAK+ 8(%rdi)      #save updated tweak T[1] (start bit cleared?)
+    movq    %r8 ,ksKey+ 0+F_O(%rbp)  #save key schedule on stack for Skein_Debug_Block
+    movq    %r9 ,ksKey+ 8+F_O(%rbp)    
+    movq    %r10,ksKey+16+F_O(%rbp)    
+    movq    %r11,ksKey+24+F_O(%rbp)    
+    movq    %r12,ksKey+32+F_O(%rbp)    
+                                       
+    movq    %r13,ksTwk+ 0+F_O(%rbp)    
+    movq    %r14,ksTwk+ 8+F_O(%rbp)    
+    movq    %r15,ksTwk+16+F_O(%rbp)    
+                                       
+    movq    %rax,X_stk + 0(%rsp)     #save X[] on stack for Skein_Debug_Block
+    movq    %rbx,X_stk + 8(%rsp)       
+    movq    %rcx,X_stk +16(%rsp)       
+    movq    %rdx,X_stk +24(%rsp)       
+
+    Skein_Debug_Block 256            #debug dump
+    Skein_Debug_Round 256,SKEIN_RND_KEY_INITIAL
+.endif
+#
+.if ((SKEIN_ASM_UNROLL & 256) == 0)
+    movq    %r8 ,ksKey+40+F_O(%rbp)  #save key schedule on stack for looping code
+    movq    %r9 ,ksKey+ 8+F_O(%rbp)    
+    movq    %r10,ksKey+16+F_O(%rbp)    
+    movq    %r11,ksKey+24+F_O(%rbp)    
+    movq    %r12,ksKey+32+F_O(%rbp)    
+                                       
+    movq    %r13,ksTwk+24+F_O(%rbp)    
+    movq    %r14,ksTwk+ 8+F_O(%rbp)    
+    movq    %r15,ksTwk+16+F_O(%rbp)    
+.endif
+    addq    $WCNT*8,%rsi             #skip the block
+    movq    %rsi,blkPtr  +F_O(%rbp)  #update block pointer
+    #
+    # now the key schedule is computed. Start the rounds
+    #
+.if SKEIN_ASM_UNROLL & 256
+_UNROLL_CNT =   ROUNDS_256/8
+.else
+_UNROLL_CNT =   SKEIN_UNROLL_256
+  .if ((ROUNDS_256/8) % _UNROLL_CNT)
+    .error "Invalid SKEIN_UNROLL_256"
+  .endif
+    xorq    %rdi,%rdi                #rdi = iteration count
+Skein_256_round_loop:
+.endif
+_Rbase_ = 0
+.rept _UNROLL_CNT*2
+    # all X and ks vars in regs      # (ops to "rotate" ks vars, via mem, if not unrolled)
+    # round 4*_RBase_ + 0
+    addReg  rax, rbx
+    RotL64  rbx, 256,%((4*_Rbase_+0) % 8),0
+    addReg  rcx, rdx
+                .if (SKEIN_ASM_UNROLL & 256) == 0
+                    movq ksKey+8*1+F_O(%rbp,%rdi,8),%r8
+                .endif
+    xorReg  rbx, rax
+    RotL64  rdx, 256,%((4*_Rbase_+0) % 8),1
+    xorReg  rdx, rcx
+  .if SKEIN_ASM_UNROLL & 256
+    .irp _r0_,%( 8+(_Rbase_+3) % 5)
+    .irp _r1_,%(13+(_Rbase_+2) % 3)
+      leaq   (%r\_r0_,%r\_r1_),%rdi    #precompute key injection value for %rcx
+    .endr
+    .endr
+  .endif
+                .if (SKEIN_ASM_UNROLL & 256) == 0
+                    movq ksTwk+8*1+F_O(%rbp,%rdi,8),%r13
+                .endif
+    Skein_Debug_Round 256,%(4*_Rbase_+1)
+
+    # round 4*_Rbase_ + 1
+    addReg  rax, rdx
+    RotL64  rdx, 256,%((4*_Rbase_+1) % 8),0
+    xorReg  rdx, rax
+                .if (SKEIN_ASM_UNROLL & 256) == 0
+                    movq ksKey+8*2+F_O(%rbp,%rdi,8),%r9
+                .endif
+    addReg  rcx, rbx
+    RotL64  rbx, 256,%((4*_Rbase_+1) % 8),1
+    xorReg  rbx, rcx
+                .if (SKEIN_ASM_UNROLL & 256) == 0
+                    movq ksKey+8*4+F_O(%rbp,%rdi,8),%r11
+                .endif
+    Skein_Debug_Round 256,%(4*_Rbase_+2)
+ .if SKEIN_ASM_UNROLL & 256
+    .irp _r0_,%( 8+(_Rbase_+2) % 5)
+    .irp _r1_,%(13+(_Rbase_+1) % 3)
+      leaq   (%r\_r0_,%r\_r1_),%rsi     #precompute key injection value for %rbx
+    .endr
+    .endr
+ .endif
+    # round 4*_Rbase_ + 2
+    addReg  rax, rbx
+    RotL64  rbx, 256,%((4*_Rbase_+2) % 8),0
+    addReg  rcx, rdx
+                .if (SKEIN_ASM_UNROLL & 256) == 0
+                    movq ksKey+8*3+F_O(%rbp,%rdi,8),%r10
+                .endif
+    xorReg  rbx, rax
+    RotL64  rdx, 256,%((4*_Rbase_+2) % 8),1
+    xorReg  rdx, rcx
+                .if (SKEIN_ASM_UNROLL & 256) == 0
+                    movq %r8,ksKey+8*6+F_O(%rbp,%rdi,8)  #"rotate" the key
+                    leaq 1(%r11,%rdi),%r11               #precompute key + tweak
+                .endif
+    Skein_Debug_Round 256,%(4*_Rbase_+3)
+    # round 4*_Rbase_ + 3
+    addReg  rax, rdx
+    RotL64  rdx, 256,%((4*_Rbase_+3) % 8),0
+    addReg  rcx, rbx
+                .if (SKEIN_ASM_UNROLL & 256) == 0
+                    addq      ksTwk+8*2+F_O(%rbp,%rdi,8),%r10  #precompute key + tweak
+                    movq %r13,ksTwk+8*4+F_O(%rbp,%rdi,8)       #"rotate" the tweak
+                .endif
+    xorReg  rdx, rax
+    RotL64  rbx, 256,%((4*_Rbase_+3) % 8),1
+    xorReg  rbx, rcx
+    Skein_Debug_Round 256,%(4*_Rbase_+4)
+                .if (SKEIN_ASM_UNROLL & 256) == 0
+                    addReg r9 ,r13           #precompute key+tweak
+                .endif
+      #inject key schedule words
+_Rbase_ = _Rbase_+1
+  .if SKEIN_ASM_UNROLL & 256
+    addReg    rax,r,%(8+((_Rbase_+0) % 5))
+    addReg    rbx,rsi
+    addReg    rcx,rdi
+    addReg    rdx,r,%(8+((_Rbase_+3) % 5)),,_Rbase_
+  .else
+    incq      %rdi
+    addReg    rax,r8 
+    addReg    rcx,r10
+    addReg    rbx,r9 
+    addReg    rdx,r11
+  .endif
+    Skein_Debug_Round 256,SKEIN_RND_KEY_INJECT
+.endr #rept _UNROLL_CNT
+#
+.if (SKEIN_ASM_UNROLL & 256) == 0
+    cmpq    $2*(ROUNDS_256/8),%rdi
+    jb      Skein_256_round_loop
+.endif # (SKEIN_ASM_UNROLL & 256) == 0
+    movq    ctxPtr +F_O(%rbp),%rdi           #restore rdi --> context
+
+    #----------------------------
+    # feedforward:   ctx->X[i] = X[i] ^ w[i], {i=0..3}
+    movq    $FIRST_MASK64 ,%r14
+    xorq    Wcopy + 0+F_O (%rbp),%rax
+    xorq    Wcopy + 8+F_O (%rbp),%rbx
+    xorq    Wcopy +16+F_O (%rbp),%rcx
+    xorq    Wcopy +24+F_O (%rbp),%rdx
+    andq    TWEAK + 8     (%rdi),%r14
+    movq    %rax,X_VARS+ 0(%rdi)             #store final result
+    movq    %rbx,X_VARS+ 8(%rdi)        
+    movq    %rcx,X_VARS+16(%rdi)        
+    movq    %rdx,X_VARS+24(%rdi)        
+
+    Skein_Debug_Round 256,SKEIN_RND_FEED_FWD
+
+    # go back for more blocks, if needed
+    decq    blkCnt+F_O(%rbp)
+    jnz     Skein_256_block_loop
+    movq    %r14,TWEAK + 8(%rdi)
+    Reset_Stack
+    ret
+Skein_256_Process_Block_End:
+
+  .if _SKEIN_DEBUG
+Skein_Debug_Round_256:               #here with rdx == round "number" from macro
+    pushq   %rsi                     #save two regs for BLK_BITS-specific parms
+    pushq   %rdi
+    movq    24(%rsp),%rdi            #get back original rdx (pushed on stack in macro call) to rdi
+    movq    %rax,X_stk+ 0+F_O(%rbp)  #save X[] state on stack so debug routines can access it
+    movq    %rbx,X_stk+ 8+F_O(%rbp)  #(use FP_ since rsp has changed!)
+    movq    %rcx,X_stk+16+F_O(%rbp)
+    movq    %rdi,X_stk+24+F_O(%rbp)
+
+    movq    ctxPtr+F_O(%rbp),%rsi    #ctx_hdr_ptr
+    movq    $256,%rdi                #now <rdi,rsi,rdx> are set for the call
+    jmp     Skein_Debug_Round_Common
+  .endif
+#
+.if _SKEIN_CODE_SIZE
+C_label  Skein_256_Process_Block_CodeSize
+    movq    $(Skein_256_Process_Block_End-Skein_256_Process_Block),%rax
+    ret
+#
+C_label Skein_256_Unroll_Cnt
+  .if _UNROLL_CNT <> ROUNDS_256/8
+    movq    $_UNROLL_CNT,%rax
+  .else
+    xorq    %rax,%rax
+  .endif
+    ret
+.endif
+#
+.endif #_USE_ASM_ & 256
+#
+#=================================== Skein_512 =============================================
+#
+.if _USE_ASM_ & 512
+#
+# void Skein_512_Process_Block(Skein_512_Ctxt_t *ctx,const u08b_t *blkPtr,size_t blkCnt,size_t bitcntAdd)
+#
+# X[i] == %r[8+i]          #register assignments for X[] values during rounds (i=0..7)
+#
+#################
+# MACRO: one round for 512-bit blocks
+#
+.macro R_512_OneRound rn0,rn1,rn2,rn3,rn4,rn5,rn6,rn7,_Rn_,op1,op2,op3,op4
+#
+    addReg      r\rn0, r\rn1
+    RotL64      r\rn1, 512,%((_Rn_) % 8),0
+    xorReg      r\rn1, r\rn0
+            op1
+    addReg      r\rn2, r\rn3
+    RotL64      r\rn3, 512,%((_Rn_) % 8),1
+    xorReg      r\rn3, r\rn2
+            op2
+    addReg      r\rn4, r\rn5
+    RotL64      r\rn5, 512,%((_Rn_) % 8),2
+    xorReg      r\rn5, r\rn4
+            op3
+    addReg      r\rn6, r\rn7
+    RotL64      r\rn7, 512,%((_Rn_) % 8),3
+    xorReg      r\rn7, r\rn6
+            op4
+    Skein_Debug_Round 512,%(_Rn_+1),-4
+#
+.endm #R_512_OneRound
+#
+#################
+# MACRO: eight rounds for 512-bit blocks
+#
+.macro R_512_FourRounds _RR_    #RR = base round number (0 % 8)
+  .if (SKEIN_ASM_UNROLL && 512)
+    # here for fully unrolled case.
+    _II_ = ((_RR_)/4) + 1       #key injection counter
+    R_512_OneRound  8, 9,10,11,12,13,14,15,%((_RR_)+0),<movq ksKey+8*(((_II_)+3) % 9)+F_O(%rbp),%rax>,,<movq ksKey+8*(((_II_)+4) % 9)+F_O(%rbp),%rbx>
+    R_512_OneRound 10, 9,12,15,14,13, 8,11,%((_RR_)+1),<movq ksKey+8*(((_II_)+5) % 9)+F_O(%rbp),%rcx>,,<movq ksKey+8*(((_II_)+6) % 9)+F_O(%rbp),%rdx>
+    R_512_OneRound 12, 9,14,11, 8,13,10,15,%((_RR_)+2),<movq ksKey+8*(((_II_)+7) % 9)+F_O(%rbp),%rsi>,,<addq ksTwk+8*(((_II_)+0) % 3)+F_O(%rbp),%rcx>
+    R_512_OneRound 14, 9, 8,15,10,13,12,11,%((_RR_)+3),<addq ksTwk+8*(((_II_)+1) % 3)+F_O(%rbp),%rdx>,
+    # inject the key schedule
+    addq    ksKey+8*(((_II_)+0)%9)+F_O(%rbp),%r8
+    addReg   r11, rax
+    addq    ksKey+8*(((_II_)+1)%9)+F_O(%rbp),%r9
+    addReg   r12, rbx
+    addq    ksKey+8*(((_II_)+2)%9)+F_O(%rbp),%r10
+    addReg   r13, rcx
+    addReg   r14, rdx
+    addReg   r15, rsi,,,(_II_)
+  .else
+    # here for looping case                                                    #"rotate" key/tweak schedule (move up on stack)
+    incq    %rdi                 #bump key injection counter
+    R_512_OneRound  8, 9,10,11,12,13,14,15,%((_RR_)+0),<movq ksKey+8*6+F_O(%rbp,%rdi,8),%rdx>,<movq      ksTwk-8*1+F_O(%rbp,%rdi,8),%rax>,<movq      ksKey-8*1+F_O(%rbp,%rdi,8),%rsi>
+    R_512_OneRound 10, 9,12,15,14,13, 8,11,%((_RR_)+1),<movq ksKey+8*5+F_O(%rbp,%rdi,8),%rcx>,<movq %rax,ksTwk+8*2+F_O(%rbp,%rdi,8)     >,<movq %rsi,ksKey+8*8+F_O(%rbp,%rdi,8)>
+    R_512_OneRound 12, 9,14,11, 8,13,10,15,%((_RR_)+2),<movq ksKey+8*4+F_O(%rbp,%rdi,8),%rbx>,<addq      ksTwk+8*1+F_O(%rbp,%rdi,8),%rdx>,<movq      ksKey+8*7+F_O(%rbp,%rdi,8),%rsi>    
+    R_512_OneRound 14, 9, 8,15,10,13,12,11,%((_RR_)+3),<movq ksKey+8*3+F_O(%rbp,%rdi,8),%rax>,<addq      ksTwk+8*0+F_O(%rbp,%rdi,8),%rcx>
+    # inject the key schedule
+    addq    ksKey+8*0+F_O(%rbp,%rdi,8),%r8
+    addReg   r11, rax
+    addReg   r12, rbx
+    addq    ksKey+8*1+F_O(%rbp,%rdi,8),%r9
+    addReg   r13, rcx
+    addReg   r14, rdx
+    addq    ksKey+8*2+F_O(%rbp,%rdi,8),%r10
+    addReg   r15, rsi
+    addReg   r15, rdi              #inject the round number
+  .endif
+
+    #show the result of the key injection
+    Skein_Debug_Round 512,SKEIN_RND_KEY_INJECT
+.endm #R_512_EightRounds
+#
+#################
+# instantiated code
+#
+C_label Skein_512_Process_Block
+    Setup_Stack 512,ROUNDS_512/8
+    movq    TWEAK+ 8(%rdi),%rbx
+    jmp     Skein_512_block_loop
+    .p2align 4
+    # main hash loop for Skein_512
+Skein_512_block_loop:
+    # general register usage:
+    #   RAX..RDX       = temps for key schedule pre-loads
+    #   R8 ..R15       = X0..X7
+    #   RSP, RBP       = stack/frame pointers
+    #   RDI            = round counter or context pointer
+    #   RSI            = temp
+    #
+    movq    TWEAK +  0(%rdi),%rax
+    addq    bitAdd+F_O(%rbp),%rax     #computed updated tweak value T0
+    movq    %rbx,%rcx
+    xorq    %rax,%rcx                 #%rax/%rbx/%rcx = tweak schedule
+    movq    %rax,TWEAK+ 0    (%rdi)   #save updated tweak value ctx->h.T[0]
+    movq    %rax,ksTwk+ 0+F_O(%rbp)
+    movq    $KW_PARITY,%rdx
+    movq    blkPtr +F_O(%rbp),%rsi    #%rsi --> input block
+    movq    %rbx,ksTwk+ 8+F_O(%rbp)
+    movq    %rcx,ksTwk+16+F_O(%rbp)
+    .irp _Rn_,8,9,10,11,12,13,14,15
+      movq  X_VARS+8*(_Rn_-8)(%rdi),%r\_Rn_
+      xorq  %r\_Rn_,%rdx              #compute overall parity
+      movq  %r\_Rn_,ksKey+8*(_Rn_-8)+F_O(%rbp)
+    .endr                             #load state into %r8 ..%r15, compute parity
+      movq  %rdx,ksKey+8*(8)+F_O(%rbp)#save key schedule parity
+
+    addReg   r13,rax                  #precompute key injection for tweak
+    addReg   r14, rbx
+.if _SKEIN_DEBUG
+    movq    %rbx,TWEAK+ 8(%rdi)       #save updated tweak value ctx->h.T[1] for Skein_Debug_Block below
+.endif
+    movq     0(%rsi),%rax             #load input block
+    movq     8(%rsi),%rbx 
+    movq    16(%rsi),%rcx 
+    movq    24(%rsi),%rdx 
+    addReg   r8 , rax                 #do initial key injection
+    addReg   r9 , rbx
+    movq    %rax,Wcopy+ 0+F_O(%rbp)   #keep local copy for feedforward
+    movq    %rbx,Wcopy+ 8+F_O(%rbp)
+    addReg   r10, rcx
+    addReg   r11, rdx
+    movq    %rcx,Wcopy+16+F_O(%rbp)
+    movq    %rdx,Wcopy+24+F_O(%rbp)
+
+    movq    32(%rsi),%rax
+    movq    40(%rsi),%rbx 
+    movq    48(%rsi),%rcx 
+    movq    56(%rsi),%rdx
+    addReg   r12, rax
+    addReg   r13, rbx
+    addReg   r14, rcx
+    addReg   r15, rdx
+    movq    %rax,Wcopy+32+F_O(%rbp)    
+    movq    %rbx,Wcopy+40+F_O(%rbp)    
+    movq    %rcx,Wcopy+48+F_O(%rbp)    
+    movq    %rdx,Wcopy+56+F_O(%rbp)    
+
+.if _SKEIN_DEBUG
+    .irp _Rn_,8,9,10,11,12,13,14,15   #save values on stack for debug output
+      movq  %r\_Rn_,X_stk+8*(_Rn_-8)(%rsp)
+    .endr
+
+    Skein_Debug_Block 512             #debug dump
+    Skein_Debug_Round 512,SKEIN_RND_KEY_INITIAL
+.endif
+    addq    $8*WCNT,%rsi              #skip the block
+    movq    %rsi,blkPtr+F_O(%rbp)     #update block pointer
+    #
+    #################
+    # now the key schedule is computed. Start the rounds
+    #
+.if SKEIN_ASM_UNROLL & 512
+_UNROLL_CNT =   ROUNDS_512/8
+.else
+_UNROLL_CNT =   SKEIN_UNROLL_512
+  .if ((ROUNDS_512/8) % _UNROLL_CNT)
+    .err "Invalid SKEIN_UNROLL_512"
+  .endif
+    xorq    %rdi,%rdi                 #rdi = round counter
+Skein_512_round_loop:
+.endif
+#
+_Rbase_ = 0
+.rept _UNROLL_CNT*2
+      R_512_FourRounds %(4*_Rbase_+00)
+_Rbase_ = _Rbase_+1
+.endr #rept _UNROLL_CNT
+#
+.if (SKEIN_ASM_UNROLL & 512) == 0
+    cmpq    $2*(ROUNDS_512/8),%rdi
+    jb      Skein_512_round_loop
+    movq    ctxPtr +F_O(%rbp),%rdi           #restore rdi --> context
+.endif
+    # end of rounds
+    #################
+    # feedforward:   ctx->X[i] = X[i] ^ w[i], {i=0..7}
+    .irp _Rn_,8,9,10,11,12,13,14,15
+  .if (_Rn_ == 8)
+    movq    $FIRST_MASK64,%rbx
+  .endif
+      xorq  Wcopy+8*(_Rn_-8)+F_O(%rbp),%r\_Rn_  #feedforward XOR
+      movq  %r\_Rn_,X_VARS+8*(_Rn_-8)(%rdi)     #and store result
+  .if (_Rn_ == 14)
+    andq    TWEAK+ 8(%rdi),%rbx
+  .endif
+    .endr
+    Skein_Debug_Round 512,SKEIN_RND_FEED_FWD
+
+    # go back for more blocks, if needed
+    decq    blkCnt+F_O(%rbp)
+    jnz     Skein_512_block_loop
+    movq    %rbx,TWEAK + 8(%rdi)
+
+    Reset_Stack
+    ret
+Skein_512_Process_Block_End:
+#
+  .if _SKEIN_DEBUG
+# call here with rdx  = "round number"
+Skein_Debug_Round_512:
+    pushq   %rsi                     #save two regs for BLK_BITS-specific parms
+    pushq   %rdi
+  .irp _Rn_,8,9,10,11,12,13,14,15    #save X[] state on stack so debug routines can access it
+    movq    %r\_Rn_,X_stk+8*(_Rn_-8)+F_O(%rbp)
+  .endr
+    movq    ctxPtr+F_O(%rbp),%rsi    #ctx_hdr_ptr
+    movq    $512,%rdi                #now <rdi,rsi,rdx> are set for the call
+    jmp     Skein_Debug_Round_Common
+  .endif
+#
+.if _SKEIN_CODE_SIZE
+C_label Skein_512_Process_Block_CodeSize
+    movq    $(Skein_512_Process_Block_End-Skein_512_Process_Block),%rax
+    ret
+#
+C_label Skein_512_Unroll_Cnt
+  .if _UNROLL_CNT <> (ROUNDS_512/8)
+    movq    $_UNROLL_CNT,%rax
+  .else
+    xorq    %rax,%rax
+  .endif
+    ret
+.endif
+#
+.endif # _USE_ASM_ & 512
+#
+#=================================== Skein1024 =============================================
+.if _USE_ASM_ & 1024
+#
+# void Skein1024_Process_Block(Skein_1024_Ctxt_t *ctx,const u08b_t *blkPtr,size_t blkCnt,size_t bitcntAdd)#
+#
+#################
+# use details of permutation to make register assignments
+# 
+o1K_rdi =  0        #offsets in X[] associated with each register
+o1K_rsi =  1 
+o1K_rbp =  2 
+o1K_rax =  3 
+o1K_rcx =  4        #rcx is "shared" with X6, since X4/X6 alternate
+o1K_rbx =  5 
+o1K_rdx =  7 
+o1K_r8  =  8  
+o1K_r9  =  9  
+o1K_r10 = 10
+o1K_r11 = 11
+o1K_r12 = 12
+o1K_r13 = 13
+o1K_r14 = 14
+o1K_r15 = 15
+#
+rIdx_offs = tmpStk_1024
+#
+.macro r1024_Mix w0,w1,reg0,reg1,_RN0_,_Rn1_,op1
+    addReg      \reg0 , \reg1                      #perform the MIX
+    RotL64      \reg1 , 1024,%((_RN0_) % 8),_Rn1_
+    xorReg      \reg1 , \reg0
+.if ((_RN0_) && 3) == 3         #time to do key injection?
+ .if _SKEIN_DEBUG
+    movq       %\reg0 , xDebug_1024+8*w0(%rsp)     #save intermediate values for Debug_Round
+    movq       %\reg1 , xDebug_1024+8*w1(%rsp)     # (before inline key injection)
+ .endif
+_II_ = ((_RN0_)/4)+1            #injection count
+ .if SKEIN_ASM_UNROLL && 1024   #here to do fully unrolled key injection
+    addq        ksKey+ 8*((_II_+w0) % 17)(%rsp),%\reg0
+    addq        ksKey+ 8*((_II_+w1) % 17)(%rsp),%\reg1
+  .if     w1 == 13                                 #tweak injection
+    addq        ksTwk+ 8*((_II_+ 0) %  3)(%rsp),%\reg1
+  .elseif w0 == 14
+    addq        ksTwk+ 8*((_II_+ 1) %  3)(%rsp),%\reg0
+  .elseif w1 == 15
+    addq        $_II_, %\reg1                      #(injection counter)
+  .endif
+ .else                          #here to do looping  key injection
+  .if  (w0 == 0)
+    movq        %rdi, X_stk+8*w0(%rsp)             #if so, store N0 so we can use reg as index
+    movq         rIdx_offs(%rsp),%rdi              #get the injection counter index into rdi
+  .else
+    addq         ksKey+8+8*w0(%rsp,%rdi,8),%\reg0  #even key injection
+  .endif
+  .if     w1 == 13                                 #tweak injection
+    addq         ksTwk+8+8* 0(%rsp,%rdi,8),%\reg1  
+  .elseif w0 == 14
+    addq         ksTwk+8+8* 1(%rsp,%rdi,8),%\reg0  
+  .elseif w1 == 15
+    addReg      \reg1,rdi,,,1                      #(injection counter)
+  .endif
+    addq         ksKey+8+8*w1(%rsp,%rdi,8),%\reg1  #odd key injection
+ .endif
+.endif
+    # insert the op provided, .if any
+    op1
+.endm
+#################
+# MACRO: four rounds for 1024-bit blocks
+#
+.macro r1024_FourRounds _RR_    #RR = base round number (0 mod 4)
+    # should be here with X4 set properly, X6 stored on stack
+_Rn_ = (_RR_) + 0
+        r1024_Mix  0, 1,rdi,rsi,_Rn_,0
+        r1024_Mix  2, 3,rbp,rax,_Rn_,1
+        r1024_Mix  4, 5,rcx,rbx,_Rn_,2,<movq %rcx,X_stk+8*4(%rsp)>       #save X4  on  stack (x4/x6 alternate)
+        r1024_Mix  8, 9,r8 ,r9 ,_Rn_,4,<movq      X_stk+8*6(%rsp),%rcx>  #load X6 from stack 
+        r1024_Mix 10,11,r10,r11,_Rn_,5
+        r1024_Mix 12,13,r12,r13,_Rn_,6
+        r1024_Mix  6, 7,rcx,rdx,_Rn_,3
+        r1024_Mix 14,15,r14,r15,_Rn_,7
+    .if _SKEIN_DEBUG
+      Skein_Debug_Round 1024,%(_Rn_+1)
+    .endif
+_Rn_ = (_RR_) + 1
+        r1024_Mix  0, 9,rdi,r9 ,_Rn_,0
+        r1024_Mix  2,13,rbp,r13,_Rn_,1
+        r1024_Mix  6,11,rcx,r11,_Rn_,2,<movq %rcx,X_stk+8*6(%rsp)>       #save X6  on  stack (x4/x6 alternate)
+        r1024_Mix 10, 7,r10,rdx,_Rn_,4,<movq      X_stk+8*4(%rsp),%rcx>  #load X4 from stack 
+        r1024_Mix 12, 3,r12,rax,_Rn_,5
+        r1024_Mix 14, 5,r14,rbx,_Rn_,6
+        r1024_Mix  4,15,rcx,r15,_Rn_,3
+        r1024_Mix  8, 1,r8 ,rsi,_Rn_,7
+    .if _SKEIN_DEBUG
+      Skein_Debug_Round 1024,%(_Rn_+1)
+    .endif
+_Rn_ = (_RR_) + 2
+        r1024_Mix  0, 7,rdi,rdx,_Rn_,0
+        r1024_Mix  2, 5,rbp,rbx,_Rn_,1
+        r1024_Mix  4, 3,rcx,rax,_Rn_,2,<movq %rcx,X_stk+8*4(%rsp)>       #save X4  on  stack (x4/x6 alternate)
+        r1024_Mix 12,15,r12,r15,_Rn_,4,<movq      X_stk+8*6(%rsp),%rcx>  #load X6 from stack 
+        r1024_Mix 14,13,r14,r13,_Rn_,5
+        r1024_Mix  8,11,r8 ,r11,_Rn_,6
+        r1024_Mix  6, 1,rcx,rsi,_Rn_,3
+        r1024_Mix 10, 9,r10,r9 ,_Rn_,7
+    .if _SKEIN_DEBUG
+      Skein_Debug_Round 1024,%(_Rn_+1)
+    .endif
+_Rn_ = (_RR_) + 3
+        r1024_Mix  0,15,rdi,r15,_Rn_,0
+        r1024_Mix  2,11,rbp,r11,_Rn_,1
+        r1024_Mix  6,13,rcx,r13,_Rn_,2,<movq %rcx,X_stk+8*6(%rsp)>       #save X6  on  stack (x4/x6 alternate)
+        r1024_Mix 14, 1,r14,rsi,_Rn_,4,<movq      X_stk+8*4(%rsp),%rcx>  #load X4 from stack 
+        r1024_Mix  8, 5,r8 ,rbx,_Rn_,5
+        r1024_Mix 10, 3,r10,rax,_Rn_,6
+        r1024_Mix  4, 9,rcx,r9 ,_Rn_,3
+        r1024_Mix 12, 7,r12,rdx,_Rn_,7
+    .if _SKEIN_DEBUG
+      Skein_Debug_Round 1024,%(_Rn_+1)
+    .endif
+
+  .if (SKEIN_ASM_UNROLL && 1024) == 0           #here with rdi == rIdx, X0 on stack
+    #"rotate" the key schedule on the stack
+i8 = o1K_r8
+i0 = o1K_rdi
+    movq    %r8 , X_stk+8*i8(%rsp)              #free up a register (save it on the stack)
+    movq          ksKey+8* 0(%rsp,%rdi,8),%r8   #get  key  word
+    movq    %r8 , ksKey+8*17(%rsp,%rdi,8)       #rotate key (must do key first or tweak clobbers it!)
+    movq          ksTwk+8* 0(%rsp,%rdi,8),%r8   #get tweak word
+    movq    %r8 , ksTwk+8* 3(%rsp,%rdi,8)       #rotate tweak (onto the stack)
+    movq          X_stk+8*i8(%rsp)       ,%r8   #get the reg back
+    incq    %rdi                                #bump the index
+    movq    %rdi, rIdx_offs (%rsp)              #save rdi again
+    movq          ksKey+8*i0(%rsp,%rdi,8),%rdi  #get the key schedule word for X0 back
+    addq          X_stk+8*i0(%rsp)       ,%rdi  #perform the X0 key injection
+  .endif
+    #show the result of the key injection
+    Skein_Debug_Round 1024,SKEIN_RND_KEY_INJECT
+.endm #r1024_FourRounds
+#
+################
+# code
+#
+C_label Skein1024_Process_Block
+#
+    Setup_Stack 1024,ROUNDS_1024/8,WCNT
+    movq    TWEAK+ 8(%rdi),%r9
+    jmp     Skein1024_block_loop
+    # main hash loop for Skein1024
+    .p2align 4
+Skein1024_block_loop:
+    # general register usage:
+    #   RSP              = stack pointer
+    #   RAX..RDX,RSI,RDI = X1, X3..X7 (state words)
+    #   R8 ..R15         = X8..X15    (state words)
+    #   RBP              = temp (used for X0 and X2)
+    #
+  .if (SKEIN_ASM_UNROLL & 1024) == 0
+    xorq    %rax,%rax                      #init loop index on the stack
+    movq    %rax,rIdx_offs(%rsp)
+  .endif
+    movq         TWEAK+     0(%rdi),%r8
+    addq         bitAdd+  F_O(%rbp),%r8    #computed updated tweak value T0
+    movq    %r9 ,%r10 
+    xorq    %r8 ,%r10                      #%rax/%rbx/%rcx = tweak schedule
+    movq    %r8 ,TWEAK+     0(%rdi)        #save updated tweak value ctx->h.T[0]
+    movq    %r8 ,ksTwk+ 0+F_O(%rbp)
+    movq    %r9 ,ksTwk+ 8+F_O(%rbp)        #keep values in %r8 ,%r9  for initial tweak injection below
+    movq    %r10,ksTwk+16+F_O(%rbp)
+  .if _SKEIN_DEBUG
+    movq    %r9 ,TWEAK+     8(%rdi)        #save updated tweak value ctx->h.T[1] for Skein_Debug_Block
+  .endif
+    movq         blkPtr +F_O(%rbp),%rsi    # rsi --> input block
+    movq        $KW_PARITY        ,%rax    #overall key schedule parity
+
+    # the logic here assumes the set {rdi,rsi,rbp,rax} = X[0,1,2,3]
+    .irp _rN_,0,1,2,3,4,6                  #process the "initial" words, using r14/r15 as temps
+      movq       X_VARS+8*_rN_(%rdi),%r14  #get state word
+      movq              8*_rN_(%rsi),%r15  #get msg   word
+      xorq  %r14,%rax                      #update key schedule overall parity
+      movq  %r14,ksKey +8*_rN_+F_O(%rbp)   #save key schedule word on stack
+      movq  %r15,Wcopy +8*_rN_+F_O(%rbp)   #save local msg Wcopy 
+      addq  %r15,%r14                      #do the initial key injection
+      movq  %r14,X_stk +8*_rN_    (%rsp)   #save initial state var on stack
+    .endr
+    # now process the rest, using the "real" registers 
+    #     (MUST do it in reverse order to inject tweaks r8/r9 first)
+    .irp _rr_,r15,r14,r13,r12,r11,r10,r9,r8,rdx,rbx
+_oo_ = o1K_\_rr_                           #offset assocated with the register
+      movq  X_VARS+8*_oo_(%rdi),%\_rr_     #get key schedule word from context
+      movq         8*_oo_(%rsi),%rcx       #get next input msg word
+      movq  %\_rr_, ksKey +8*_oo_(%rsp)    #save key schedule on stack
+      xorq  %\_rr_, %rax                   #accumulate key schedule parity
+      movq  %rcx,Wcopy+8*_oo_+F_O(%rbp)    #save copy of msg word for feedforward
+      addq  %rcx,%\_rr_                    #do the initial  key  injection
+      .if    _oo_ == 13                    #do the initial tweak injection
+        addReg _rr_,r8                     #          (only in words 13/14)
+      .elseif _oo_ == 14
+        addReg _rr_,r9 
+      .endif
+    .endr
+    movq    %rax,ksKey+8*WCNT+F_O(%rbp)    #save key schedule parity
+.if _SKEIN_DEBUG
+    Skein_Debug_Block 1024                 #initial debug dump
+.endif
+    addq     $8*WCNT,%rsi                  #bump the msg ptr
+    movq     %rsi,blkPtr+F_O(%rbp)         #save bumped msg ptr
+    # re-load words 0..4 from stack, enter the main loop
+    .irp _rr_,rdi,rsi,rbp,rax,rcx          #(no need to re-load x6, already on stack)
+      movq  X_stk+8*o1K_\_rr_(%rsp),%\_rr_ #re-load state and get ready to go!
+    .endr
+.if _SKEIN_DEBUG
+    Skein_Debug_Round 1024,SKEIN_RND_KEY_INITIAL        #show state after initial key injection
+.endif
+    #
+    #################
+    # now the key schedule is computed. Start the rounds
+    #
+.if SKEIN_ASM_UNROLL & 1024
+_UNROLL_CNT =   ROUNDS_1024/8
+.else
+_UNROLL_CNT =   SKEIN_UNROLL_1024
+  .if ((ROUNDS_1024/8) % _UNROLL_CNT)
+    .error "Invalid SKEIN_UNROLL_1024"
+  .endif
+Skein1024_round_loop:
+.endif
+#
+_Rbase_ = 0
+.rept _UNROLL_CNT*2                        #implement the rounds, 4 at a time
+      r1024_FourRounds %(4*_Rbase_+00)
+_Rbase_ = _Rbase_+1
+.endr #rept _UNROLL_CNT
+#
+.if (SKEIN_ASM_UNROLL & 1024) == 0
+    cmpq    $2*(ROUNDS_1024/8),tmpStk_1024(%rsp) #see .if we are done
+    jb      Skein1024_round_loop    
+.endif
+    # end of rounds
+    #################
+    #
+    # feedforward:   ctx->X[i] = X[i] ^ w[i], {i=0..15}
+    movq    %rdx,X_stk+8*o1K_rdx(%rsp) #we need a register. x6 already on stack
+    movq       ctxPtr(%rsp),%rdx
+    
+    .irp _rr_,rdi,rsi,rbp,rax,rcx,rbx,r8,r9,r10,r11,r12,r13,r14,r15   #do all but x6,x7
+_oo_ = o1K_\_rr_
+      xorq  Wcopy +8*_oo_(%rsp),%\_rr_ #feedforward XOR
+      movq  %\_rr_,X_VARS+8*_oo_(%rdx) #save result into context
+      .if (_oo_ ==  9)
+        movq   $FIRST_MASK64 ,%r9
+      .endif
+      .if (_oo_ == 14)
+        andq   TWEAK+ 8(%rdx),%r9
+      .endif
+    .endr
+    # 
+    movq         X_stk +8*6(%rsp),%rax #now process x6,x7 (skipped in .irp above)
+    movq         X_stk +8*7(%rsp),%rbx
+    xorq         Wcopy +8*6(%rsp),%rax
+    xorq         Wcopy +8*7(%rsp),%rbx
+    movq    %rax,X_VARS+8*6(%rdx)
+    decq             blkCnt(%rsp)      #set zero flag iff done
+    movq    %rbx,X_VARS+8*7(%rdx)
+
+    Skein_Debug_Round 1024,SKEIN_RND_FEED_FWD,,<cmpq $0,blkCnt(%rsp)>
+    # go back for more blocks, if needed
+    movq             ctxPtr(%rsp),%rdi #don't muck with the flags here!
+    lea          FRAME_OFFS(%rsp),%rbp
+    jnz     Skein1024_block_loop
+    movq    %r9 ,TWEAK+   8(%rdx)
+    Reset_Stack
+    ret
+#
+Skein1024_Process_Block_End:
+#
+.if _SKEIN_DEBUG
+Skein_Debug_Round_1024:
+    # call here with rdx  = "round number",
+_SP_OFFS_ = 8*2                     #stack "offset" here: rdx, return addr
+    #
+  #save rest of X[] state on stack so debug routines can access it
+  .irp _rr_,rsi,rbp,rax,rbx,r8,r9,r10,r11,r12,r13,r14,r15
+    movq    %\_rr_,X_stk+8*o1K_\_rr_+_SP_OFFS_(%rsp)
+  .endr
+    # Figure out what to do with x0 (rdi).  When rdx == 0 mod 4, it's already on stack
+    cmpq    $SKEIN_RND_SPECIAL,%rdx #special rounds always save
+    jae     save_x0
+    testq   $3,%rdx                 #otherwise only if rdx != 0 mod 4
+    jz      save_x0_not
+save_x0:
+    movq    %rdi,X_stk+8*o1K_rdi+_SP_OFFS_(%rsp)
+save_x0_not:
+    #figure out the x4/x6 swapping state and save the correct one!
+    cmpq    $SKEIN_RND_SPECIAL,%rdx #special rounds always do x4
+    jae     save_x4
+    testq   $1,%rdx                  #and even ones have r4 as well
+    jz      save_x4
+    movq    %rcx,X_stk+8*6+_SP_OFFS_(%rsp)
+    jmp     debug_1024_go
+save_x4:
+    movq    %rcx,X_stk+8*4+_SP_OFFS_(%rsp)
+debug_1024_go:
+    #now all is saved in Xstk[] except for rdx
+    push    %rsi                    #save two regs for BLK_BITS-specific parms
+    push    %rdi
+_SP_OFFS_ = _SP_OFFS_ + 16          #adjust stack offset accordingly (now 32)
+
+    movq    _SP_OFFS_-8(%rsp),%rsi  #get back original %rdx (pushed on stack in macro call)
+    movq    %rsi,X_stk+8*o1K_rdx+_SP_OFFS_(%rsp) #and save it in its rightful place in X_stk[]
+
+    movq    ctxPtr+_SP_OFFS_(%rsp),%rsi  #rsi = ctx_hdr_ptr
+    movq    $1024,%rdi                   #rdi = block size
+    jmp     Skein_Debug_Round_Common
+.endif
+#
+.if _SKEIN_CODE_SIZE
+C_label Skein1024_Process_Block_CodeSize
+    movq    $(Skein1024_Process_Block_End-Skein1024_Process_Block),%rax
+    ret
+#
+C_label Skein1024_Unroll_Cnt
+  .if _UNROLL_CNT <> (ROUNDS_1024/8)
+    movq    $_UNROLL_CNT,%rax
+  .else
+    xorq    %rax,%rax
+  .endif
+    ret
+.endif
+#
+.endif # _USE_ASM_ and 1024
+#
+.if _SKEIN_DEBUG
+#----------------------------------------------------------------
+#local debug routine to set up for calls to:
+#  void Skein_Show_Round(uint_t bits,const Skein_Ctxt_Hdr_t *h,int r,const u64b_t *X)
+#                       [       rdi                        rsi   rdx              rcx]
+#
+# here with %rdx = round number
+#           %rsi = ctx_hdr_ptr
+#           %rdi = block size (256/512/1024)
+# on stack: saved rdi, saved rsi, retAddr, saved rdx  
+#
+Skein_Debug_Round_Common:
+_SP_OFFS_ = 32                        #account for four words on stack already
+  .irp _rr_,rax,rbx,rcx,rbp,r8,r9,r10,r11,r12,r13,r14,r15  #save the rest of the regs
+    pushq %\_rr_
+_SP_OFFS_ = _SP_OFFS_+8
+  .endr
+  .if (_SP_OFFS_ % 16)                # make sure stack is still 16-byte aligned here
+    .error  "Debug_Round_Common: stack alignment"
+  .endif
+    # compute %rcx  = ptr to the X[] array on the stack (final parameter to call)
+    leaq    X_stk+_SP_OFFS_(%rsp),%rcx #adjust for reg pushes, return address
+    cmpq    $SKEIN_RND_FEED_FWD,%rdx   #special handling for feedforward "round"?
+    jnz     _got_rcxA
+    leaq    X_VARS(%rsi),%rcx
+_got_rcxA:
+  .if _USE_ASM_ & 1024
+    # special handling for 1024-bit case
+    #    (for rounds right before with key injection: 
+    #        use xDebug_1024[] instead of X_stk[])
+    cmpq    $SKEIN_RND_SPECIAL,%rdx
+    jae     _got_rcxB               #must be a normal round
+    orq     %rdx,%rdx
+    jz      _got_rcxB               #just before key injection
+    test    $3,%rdx
+    jne     _got_rcxB
+    cmp     $1024,%rdi              #only 1024-bit(s) for now
+    jne     _got_rcxB
+    leaq    xDebug_1024+_SP_OFFS_(%rsp),%rcx
+_got_rcxB:
+  .endif
+    call    Skein_Show_Round        #call external debug handler
+
+  .irp _rr_,r15,r14,r13,r12,r11,r10,r9,r8,rbp,rcx,rbx,rax  #restore regs
+    popq  %\_rr_
+_SP_OFFS_ = _SP_OFFS_-8
+  .endr
+  .if _SP_OFFS_ - 32
+    .error   "Debug_Round_Common: push/pop misalignment!"
+  .endif    
+    popq    %rdi
+    popq    %rsi
+    ret
+.endif
+#----------------------------------------------------------------
+    .end
diff --git a/sys/crypto/skein/skein.c b/sys/crypto/skein/skein.c
new file mode 100644
index 0000000..b8d5905
--- /dev/null
+++ b/sys/crypto/skein/skein.c
@@ -0,0 +1,858 @@
+/***********************************************************************
+**
+** Implementation of the Skein hash function.
+**
+** Source code author: Doug Whiting, 2008.
+**
+** This algorithm and source code is released to the public domain.
+** 
+************************************************************************/
+
+#include <sys/cdefs.h>
+__FBSDID("$FreeBSD$");
+
+#include <sys/endian.h>
+#include <sys/types.h>
+
+/* get the memcpy/memset functions */
+#ifdef _KERNEL
+#include <sys/systm.h>
+#else
+#include <string.h>
+#endif
+
+#define  SKEIN_PORT_CODE /* instantiate any code in skein_port.h */
+
+#include "skein.h"       /* get the Skein API definitions   */
+#include "skein_iv.h"    /* get precomputed IVs */
+
+/*****************************************************************/
+/* External function to process blkCnt (nonzero) full block(s) of data. */
+void    Skein_256_Process_Block(Skein_256_Ctxt_t *ctx,const u08b_t *blkPtr,size_t blkCnt,size_t byteCntAdd);
+void    Skein_512_Process_Block(Skein_512_Ctxt_t *ctx,const u08b_t *blkPtr,size_t blkCnt,size_t byteCntAdd);
+void    Skein1024_Process_Block(Skein1024_Ctxt_t *ctx,const u08b_t *blkPtr,size_t blkCnt,size_t byteCntAdd);
+
+/*****************************************************************/
+/*     256-bit Skein                                             */
+/*****************************************************************/
+
+/*++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++*/
+/* init the context for a straight hashing operation  */
+int Skein_256_Init(Skein_256_Ctxt_t *ctx, size_t hashBitLen)
+    {
+    union
+        {
+        u08b_t  b[SKEIN_256_STATE_BYTES];
+        u64b_t  w[SKEIN_256_STATE_WORDS];
+        } cfg;                              /* config block */
+        
+    Skein_Assert(hashBitLen > 0,SKEIN_BAD_HASHLEN);
+    ctx->h.hashBitLen = hashBitLen;         /* output hash bit count */
+
+    switch (hashBitLen)
+        {             /* use pre-computed values, where available */
+#ifndef SKEIN_NO_PRECOMP
+        case  256: memcpy(ctx->X,SKEIN_256_IV_256,sizeof(ctx->X));  break;
+        case  224: memcpy(ctx->X,SKEIN_256_IV_224,sizeof(ctx->X));  break;
+        case  160: memcpy(ctx->X,SKEIN_256_IV_160,sizeof(ctx->X));  break;
+        case  128: memcpy(ctx->X,SKEIN_256_IV_128,sizeof(ctx->X));  break;
+#endif
+        default:
+            /* here if there is no precomputed IV value available */
+            /* build/process the config block, type == CONFIG (could be precomputed) */
+            Skein_Start_New_Type(ctx,CFG_FINAL);        /* set tweaks: T0=0; T1=CFG | FINAL */
+
+            cfg.w[0] = Skein_Swap64(SKEIN_SCHEMA_VER);  /* set the schema, version */
+            cfg.w[1] = Skein_Swap64(hashBitLen);        /* hash result length in bits */
+            cfg.w[2] = Skein_Swap64(SKEIN_CFG_TREE_INFO_SEQUENTIAL);
+            memset(&cfg.w[3],0,sizeof(cfg) - 3*sizeof(cfg.w[0])); /* zero pad config block */
+
+            /* compute the initial chaining values from config block */
+            memset(ctx->X,0,sizeof(ctx->X));            /* zero the chaining variables */
+            Skein_256_Process_Block(ctx,cfg.b,1,SKEIN_CFG_STR_LEN);
+            break;
+        }
+    /* The chaining vars ctx->X are now initialized for the given hashBitLen. */
+    /* Set up to process the data message portion of the hash (default) */
+    Skein_Start_New_Type(ctx,MSG);              /* T0=0, T1= MSG type */
+
+    return SKEIN_SUCCESS;
+    }
+
+/*++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++*/
+/* init the context for a MAC and/or tree hash operation */
+/* [identical to Skein_256_Init() when keyBytes == 0 && treeInfo == SKEIN_CFG_TREE_INFO_SEQUENTIAL] */
+int Skein_256_InitExt(Skein_256_Ctxt_t *ctx,size_t hashBitLen,u64b_t treeInfo, const u08b_t *key, size_t keyBytes)
+    {
+    union
+        {
+        u08b_t  b[SKEIN_256_STATE_BYTES];
+        u64b_t  w[SKEIN_256_STATE_WORDS];
+        } cfg;                              /* config block */
+        
+    Skein_Assert(hashBitLen > 0,SKEIN_BAD_HASHLEN);
+    Skein_Assert(keyBytes == 0 || key != NULL,SKEIN_FAIL);
+
+    /* compute the initial chaining values ctx->X[], based on key */
+    if (keyBytes == 0)                          /* is there a key? */
+        {                                   
+        memset(ctx->X,0,sizeof(ctx->X));        /* no key: use all zeroes as key for config block */
+        }
+    else                                        /* here to pre-process a key */
+        {
+        Skein_assert(sizeof(cfg.b) >= sizeof(ctx->X));
+        /* do a mini-Init right here */
+        ctx->h.hashBitLen=8*sizeof(ctx->X);     /* set output hash bit count = state size */
+        Skein_Start_New_Type(ctx,KEY);          /* set tweaks: T0 = 0; T1 = KEY type */
+        memset(ctx->X,0,sizeof(ctx->X));        /* zero the initial chaining variables */
+        Skein_256_Update(ctx,key,keyBytes);     /* hash the key */
+        Skein_256_Final_Pad(ctx,cfg.b);         /* put result into cfg.b[] */
+        memcpy(ctx->X,cfg.b,sizeof(cfg.b));     /* copy over into ctx->X[] */
+#if SKEIN_NEED_SWAP
+        {
+        uint_t i;
+        for (i=0;i<SKEIN_256_STATE_WORDS;i++)   /* convert key bytes to context words */
+            ctx->X[i] = Skein_Swap64(ctx->X[i]);
+        }
+#endif
+        }
+    /* build/process the config block, type == CONFIG (could be precomputed for each key) */
+    ctx->h.hashBitLen = hashBitLen;             /* output hash bit count */
+    Skein_Start_New_Type(ctx,CFG_FINAL);
+
+    memset(&cfg.w,0,sizeof(cfg.w));             /* pre-pad cfg.w[] with zeroes */
+    cfg.w[0] = Skein_Swap64(SKEIN_SCHEMA_VER);
+    cfg.w[1] = Skein_Swap64(hashBitLen);        /* hash result length in bits */
+    cfg.w[2] = Skein_Swap64(treeInfo);          /* tree hash config info (or SKEIN_CFG_TREE_INFO_SEQUENTIAL) */
+
+    Skein_Show_Key(256,&ctx->h,key,keyBytes);
+
+    /* compute the initial chaining values from config block */
+    Skein_256_Process_Block(ctx,cfg.b,1,SKEIN_CFG_STR_LEN);
+
+    /* The chaining vars ctx->X are now initialized */
+    /* Set up to process the data message portion of the hash (default) */
+    ctx->h.bCnt = 0;                            /* buffer b[] starts out empty */
+    Skein_Start_New_Type(ctx,MSG);
+    
+    return SKEIN_SUCCESS;
+    }
+
+/*++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++*/
+/* process the input bytes */
+int Skein_256_Update(Skein_256_Ctxt_t *ctx, const u08b_t *msg, size_t msgByteCnt)
+    {
+    size_t n;
+
+    Skein_Assert(ctx->h.bCnt <= SKEIN_256_BLOCK_BYTES,SKEIN_FAIL);    /* catch uninitialized context */
+
+    /* process full blocks, if any */
+    if (msgByteCnt + ctx->h.bCnt > SKEIN_256_BLOCK_BYTES)
+        {
+        if (ctx->h.bCnt)                              /* finish up any buffered message data */
+            {
+            n = SKEIN_256_BLOCK_BYTES - ctx->h.bCnt;  /* # bytes free in buffer b[] */
+            if (n)
+                {
+                Skein_assert(n < msgByteCnt);         /* check on our logic here */
+                memcpy(&ctx->b[ctx->h.bCnt],msg,n);
+                msgByteCnt  -= n;
+                msg         += n;
+                ctx->h.bCnt += n;
+                }
+            Skein_assert(ctx->h.bCnt == SKEIN_256_BLOCK_BYTES);
+            Skein_256_Process_Block(ctx,ctx->b,1,SKEIN_256_BLOCK_BYTES);
+            ctx->h.bCnt = 0;
+            }
+        /* now process any remaining full blocks, directly from input message data */
+        if (msgByteCnt > SKEIN_256_BLOCK_BYTES)
+            {
+            n = (msgByteCnt-1) / SKEIN_256_BLOCK_BYTES;   /* number of full blocks to process */
+            Skein_256_Process_Block(ctx,msg,n,SKEIN_256_BLOCK_BYTES);
+            msgByteCnt -= n * SKEIN_256_BLOCK_BYTES;
+            msg        += n * SKEIN_256_BLOCK_BYTES;
+            }
+        Skein_assert(ctx->h.bCnt == 0);
+        }
+
+    /* copy any remaining source message data bytes into b[] */
+    if (msgByteCnt)
+        {
+        Skein_assert(msgByteCnt + ctx->h.bCnt <= SKEIN_256_BLOCK_BYTES);
+        memcpy(&ctx->b[ctx->h.bCnt],msg,msgByteCnt);
+        ctx->h.bCnt += msgByteCnt;
+        }
+
+    return SKEIN_SUCCESS;
+    }
+   
+/*++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++*/
+/* finalize the hash computation and output the result */
+int Skein_256_Final(Skein_256_Ctxt_t *ctx, u08b_t *hashVal)
+    {
+    size_t i,n,byteCnt;
+    u64b_t X[SKEIN_256_STATE_WORDS];
+    Skein_Assert(ctx->h.bCnt <= SKEIN_256_BLOCK_BYTES,SKEIN_FAIL);    /* catch uninitialized context */
+
+    ctx->h.T[1] |= SKEIN_T1_FLAG_FINAL;                 /* tag as the final block */
+    if (ctx->h.bCnt < SKEIN_256_BLOCK_BYTES)            /* zero pad b[] if necessary */
+        memset(&ctx->b[ctx->h.bCnt],0,SKEIN_256_BLOCK_BYTES - ctx->h.bCnt);
+
+    Skein_256_Process_Block(ctx,ctx->b,1,ctx->h.bCnt);  /* process the final block */
+    
+    /* now output the result */
+    byteCnt = (ctx->h.hashBitLen + 7) >> 3;             /* total number of output bytes */
+
+    /* run Threefish in "counter mode" to generate output */
+    memset(ctx->b,0,sizeof(ctx->b));  /* zero out b[], so it can hold the counter */
+    memcpy(X,ctx->X,sizeof(X));       /* keep a local copy of counter mode "key" */
+    for (i=0;i*SKEIN_256_BLOCK_BYTES < byteCnt;i++)
+        {
+        ((u64b_t *)ctx->b)[0]= Skein_Swap64((u64b_t) i); /* build the counter block */
+        Skein_Start_New_Type(ctx,OUT_FINAL);
+        Skein_256_Process_Block(ctx,ctx->b,1,sizeof(u64b_t)); /* run "counter mode" */
+        n = byteCnt - i*SKEIN_256_BLOCK_BYTES;   /* number of output bytes left to go */
+        if (n >= SKEIN_256_BLOCK_BYTES)
+            n  = SKEIN_256_BLOCK_BYTES;
+        Skein_Put64_LSB_First(hashVal+i*SKEIN_256_BLOCK_BYTES,ctx->X,n);   /* "output" the ctr mode bytes */
+        Skein_Show_Final(256,&ctx->h,n,hashVal+i*SKEIN_256_BLOCK_BYTES);
+        memcpy(ctx->X,X,sizeof(X));   /* restore the counter mode key for next time */
+        }
+    return SKEIN_SUCCESS;
+    }
+
+#if defined(SKEIN_CODE_SIZE) || defined(SKEIN_PERF)
+size_t Skein_256_API_CodeSize(void)
+    {
+    return ((u08b_t *) Skein_256_API_CodeSize) -
+           ((u08b_t *) Skein_256_Init);
+    }
+#endif
+
+/*****************************************************************/
+/*     512-bit Skein                                             */
+/*****************************************************************/
+
+/*++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++*/
+/* init the context for a straight hashing operation  */
+int Skein_512_Init(Skein_512_Ctxt_t *ctx, size_t hashBitLen)
+    {
+    union
+        {
+        u08b_t  b[SKEIN_512_STATE_BYTES];
+        u64b_t  w[SKEIN_512_STATE_WORDS];
+        } cfg;                              /* config block */
+        
+    Skein_Assert(hashBitLen > 0,SKEIN_BAD_HASHLEN);
+    ctx->h.hashBitLen = hashBitLen;         /* output hash bit count */
+
+    switch (hashBitLen)
+        {             /* use pre-computed values, where available */
+#ifndef SKEIN_NO_PRECOMP
+        case  512: memcpy(ctx->X,SKEIN_512_IV_512,sizeof(ctx->X));  break;
+        case  384: memcpy(ctx->X,SKEIN_512_IV_384,sizeof(ctx->X));  break;
+        case  256: memcpy(ctx->X,SKEIN_512_IV_256,sizeof(ctx->X));  break;
+        case  224: memcpy(ctx->X,SKEIN_512_IV_224,sizeof(ctx->X));  break;
+#endif
+        default:
+            /* here if there is no precomputed IV value available */
+            /* build/process the config block, type == CONFIG (could be precomputed) */
+            Skein_Start_New_Type(ctx,CFG_FINAL);        /* set tweaks: T0=0; T1=CFG | FINAL */
+
+            cfg.w[0] = Skein_Swap64(SKEIN_SCHEMA_VER);  /* set the schema, version */
+            cfg.w[1] = Skein_Swap64(hashBitLen);        /* hash result length in bits */
+            cfg.w[2] = Skein_Swap64(SKEIN_CFG_TREE_INFO_SEQUENTIAL);
+            memset(&cfg.w[3],0,sizeof(cfg) - 3*sizeof(cfg.w[0])); /* zero pad config block */
+
+            /* compute the initial chaining values from config block */
+            memset(ctx->X,0,sizeof(ctx->X));            /* zero the chaining variables */
+            Skein_512_Process_Block(ctx,cfg.b,1,SKEIN_CFG_STR_LEN);
+            break;
+        }
+
+    /* The chaining vars ctx->X are now initialized for the given hashBitLen. */
+    /* Set up to process the data message portion of the hash (default) */
+    Skein_Start_New_Type(ctx,MSG);              /* T0=0, T1= MSG type */
+
+    return SKEIN_SUCCESS;
+    }
+
+/*++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++*/
+/* init the context for a MAC and/or tree hash operation */
+/* [identical to Skein_512_Init() when keyBytes == 0 && treeInfo == SKEIN_CFG_TREE_INFO_SEQUENTIAL] */
+int Skein_512_InitExt(Skein_512_Ctxt_t *ctx,size_t hashBitLen,u64b_t treeInfo, const u08b_t *key, size_t keyBytes)
+    {
+    union
+        {
+        u08b_t  b[SKEIN_512_STATE_BYTES];
+        u64b_t  w[SKEIN_512_STATE_WORDS];
+        } cfg;                              /* config block */
+        
+    Skein_Assert(hashBitLen > 0,SKEIN_BAD_HASHLEN);
+    Skein_Assert(keyBytes == 0 || key != NULL,SKEIN_FAIL);
+
+    /* compute the initial chaining values ctx->X[], based on key */
+    if (keyBytes == 0)                          /* is there a key? */
+        {                                   
+        memset(ctx->X,0,sizeof(ctx->X));        /* no key: use all zeroes as key for config block */
+        }
+    else                                        /* here to pre-process a key */
+        {
+        Skein_assert(sizeof(cfg.b) >= sizeof(ctx->X));
+        /* do a mini-Init right here */
+        ctx->h.hashBitLen=8*sizeof(ctx->X);     /* set output hash bit count = state size */
+        Skein_Start_New_Type(ctx,KEY);          /* set tweaks: T0 = 0; T1 = KEY type */
+        memset(ctx->X,0,sizeof(ctx->X));        /* zero the initial chaining variables */
+        Skein_512_Update(ctx,key,keyBytes);     /* hash the key */
+        Skein_512_Final_Pad(ctx,cfg.b);         /* put result into cfg.b[] */
+        memcpy(ctx->X,cfg.b,sizeof(cfg.b));     /* copy over into ctx->X[] */
+#if SKEIN_NEED_SWAP
+        {
+        uint_t i;
+        for (i=0;i<SKEIN_512_STATE_WORDS;i++)   /* convert key bytes to context words */
+            ctx->X[i] = Skein_Swap64(ctx->X[i]);
+        }
+#endif
+        }
+    /* build/process the config block, type == CONFIG (could be precomputed for each key) */
+    ctx->h.hashBitLen = hashBitLen;             /* output hash bit count */
+    Skein_Start_New_Type(ctx,CFG_FINAL);
+
+    memset(&cfg.w,0,sizeof(cfg.w));             /* pre-pad cfg.w[] with zeroes */
+    cfg.w[0] = Skein_Swap64(SKEIN_SCHEMA_VER);
+    cfg.w[1] = Skein_Swap64(hashBitLen);        /* hash result length in bits */
+    cfg.w[2] = Skein_Swap64(treeInfo);          /* tree hash config info (or SKEIN_CFG_TREE_INFO_SEQUENTIAL) */
+
+    Skein_Show_Key(512,&ctx->h,key,keyBytes);
+
+    /* compute the initial chaining values from config block */
+    Skein_512_Process_Block(ctx,cfg.b,1,SKEIN_CFG_STR_LEN);
+
+    /* The chaining vars ctx->X are now initialized */
+    /* Set up to process the data message portion of the hash (default) */
+    ctx->h.bCnt = 0;                            /* buffer b[] starts out empty */
+    Skein_Start_New_Type(ctx,MSG);
+    
+    return SKEIN_SUCCESS;
+    }
+
+/*++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++*/
+/* process the input bytes */
+int Skein_512_Update(Skein_512_Ctxt_t *ctx, const u08b_t *msg, size_t msgByteCnt)
+    {
+    size_t n;
+
+    Skein_Assert(ctx->h.bCnt <= SKEIN_512_BLOCK_BYTES,SKEIN_FAIL);    /* catch uninitialized context */
+
+    /* process full blocks, if any */
+    if (msgByteCnt + ctx->h.bCnt > SKEIN_512_BLOCK_BYTES)
+        {
+        if (ctx->h.bCnt)                              /* finish up any buffered message data */
+            {
+            n = SKEIN_512_BLOCK_BYTES - ctx->h.bCnt;  /* # bytes free in buffer b[] */
+            if (n)
+                {
+                Skein_assert(n < msgByteCnt);         /* check on our logic here */
+                memcpy(&ctx->b[ctx->h.bCnt],msg,n);
+                msgByteCnt  -= n;
+                msg         += n;
+                ctx->h.bCnt += n;
+                }
+            Skein_assert(ctx->h.bCnt == SKEIN_512_BLOCK_BYTES);
+            Skein_512_Process_Block(ctx,ctx->b,1,SKEIN_512_BLOCK_BYTES);
+            ctx->h.bCnt = 0;
+            }
+        /* now process any remaining full blocks, directly from input message data */
+        if (msgByteCnt > SKEIN_512_BLOCK_BYTES)
+            {
+            n = (msgByteCnt-1) / SKEIN_512_BLOCK_BYTES;   /* number of full blocks to process */
+            Skein_512_Process_Block(ctx,msg,n,SKEIN_512_BLOCK_BYTES);
+            msgByteCnt -= n * SKEIN_512_BLOCK_BYTES;
+            msg        += n * SKEIN_512_BLOCK_BYTES;
+            }
+        Skein_assert(ctx->h.bCnt == 0);
+        }
+
+    /* copy any remaining source message data bytes into b[] */
+    if (msgByteCnt)
+        {
+        Skein_assert(msgByteCnt + ctx->h.bCnt <= SKEIN_512_BLOCK_BYTES);
+        memcpy(&ctx->b[ctx->h.bCnt],msg,msgByteCnt);
+        ctx->h.bCnt += msgByteCnt;
+        }
+
+    return SKEIN_SUCCESS;
+    }
+   
+/*++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++*/
+/* finalize the hash computation and output the result */
+int Skein_512_Final(Skein_512_Ctxt_t *ctx, u08b_t *hashVal)
+    {
+    size_t i,n,byteCnt;
+    u64b_t X[SKEIN_512_STATE_WORDS];
+    Skein_Assert(ctx->h.bCnt <= SKEIN_512_BLOCK_BYTES,SKEIN_FAIL);    /* catch uninitialized context */
+
+    ctx->h.T[1] |= SKEIN_T1_FLAG_FINAL;                 /* tag as the final block */
+    if (ctx->h.bCnt < SKEIN_512_BLOCK_BYTES)            /* zero pad b[] if necessary */
+        memset(&ctx->b[ctx->h.bCnt],0,SKEIN_512_BLOCK_BYTES - ctx->h.bCnt);
+
+    Skein_512_Process_Block(ctx,ctx->b,1,ctx->h.bCnt);  /* process the final block */
+    
+    /* now output the result */
+    byteCnt = (ctx->h.hashBitLen + 7) >> 3;             /* total number of output bytes */
+
+    /* run Threefish in "counter mode" to generate output */
+    memset(ctx->b,0,sizeof(ctx->b));  /* zero out b[], so it can hold the counter */
+    memcpy(X,ctx->X,sizeof(X));       /* keep a local copy of counter mode "key" */
+    for (i=0;i*SKEIN_512_BLOCK_BYTES < byteCnt;i++)
+        {
+        ((u64b_t *)ctx->b)[0]= Skein_Swap64((u64b_t) i); /* build the counter block */
+        Skein_Start_New_Type(ctx,OUT_FINAL);
+        Skein_512_Process_Block(ctx,ctx->b,1,sizeof(u64b_t)); /* run "counter mode" */
+        n = byteCnt - i*SKEIN_512_BLOCK_BYTES;   /* number of output bytes left to go */
+        if (n >= SKEIN_512_BLOCK_BYTES)
+            n  = SKEIN_512_BLOCK_BYTES;
+        Skein_Put64_LSB_First(hashVal+i*SKEIN_512_BLOCK_BYTES,ctx->X,n);   /* "output" the ctr mode bytes */
+        Skein_Show_Final(512,&ctx->h,n,hashVal+i*SKEIN_512_BLOCK_BYTES);
+        memcpy(ctx->X,X,sizeof(X));   /* restore the counter mode key for next time */
+        }
+    return SKEIN_SUCCESS;
+    }
+
+#if defined(SKEIN_CODE_SIZE) || defined(SKEIN_PERF)
+size_t Skein_512_API_CodeSize(void)
+    {
+    return ((u08b_t *) Skein_512_API_CodeSize) -
+           ((u08b_t *) Skein_512_Init);
+    }
+#endif
+
+/*****************************************************************/
+/*    1024-bit Skein                                             */
+/*****************************************************************/
+
+/*++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++*/
+/* init the context for a straight hashing operation  */
+int Skein1024_Init(Skein1024_Ctxt_t *ctx, size_t hashBitLen)
+    {
+    union
+        {
+        u08b_t  b[SKEIN1024_STATE_BYTES];
+        u64b_t  w[SKEIN1024_STATE_WORDS];
+        } cfg;                              /* config block */
+        
+    Skein_Assert(hashBitLen > 0,SKEIN_BAD_HASHLEN);
+    ctx->h.hashBitLen = hashBitLen;         /* output hash bit count */
+
+    switch (hashBitLen)
+        {              /* use pre-computed values, where available */
+#ifndef SKEIN_NO_PRECOMP
+        case  512: memcpy(ctx->X,SKEIN1024_IV_512 ,sizeof(ctx->X)); break;
+        case  384: memcpy(ctx->X,SKEIN1024_IV_384 ,sizeof(ctx->X)); break;
+        case 1024: memcpy(ctx->X,SKEIN1024_IV_1024,sizeof(ctx->X)); break;
+#endif
+        default:
+            /* here if there is no precomputed IV value available */
+            /* build/process the config block, type == CONFIG (could be precomputed) */
+            Skein_Start_New_Type(ctx,CFG_FINAL);        /* set tweaks: T0=0; T1=CFG | FINAL */
+
+            cfg.w[0] = Skein_Swap64(SKEIN_SCHEMA_VER);  /* set the schema, version */
+            cfg.w[1] = Skein_Swap64(hashBitLen);        /* hash result length in bits */
+            cfg.w[2] = Skein_Swap64(SKEIN_CFG_TREE_INFO_SEQUENTIAL);
+            memset(&cfg.w[3],0,sizeof(cfg) - 3*sizeof(cfg.w[0])); /* zero pad config block */
+
+            /* compute the initial chaining values from config block */
+            memset(ctx->X,0,sizeof(ctx->X));            /* zero the chaining variables */
+            Skein1024_Process_Block(ctx,cfg.b,1,SKEIN_CFG_STR_LEN);
+            break;
+        }
+
+    /* The chaining vars ctx->X are now initialized for the given hashBitLen. */
+    /* Set up to process the data message portion of the hash (default) */
+    Skein_Start_New_Type(ctx,MSG);              /* T0=0, T1= MSG type */
+
+    return SKEIN_SUCCESS;
+    }
+
+/*++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++*/
+/* init the context for a MAC and/or tree hash operation */
+/* [identical to Skein1024_Init() when keyBytes == 0 && treeInfo == SKEIN_CFG_TREE_INFO_SEQUENTIAL] */
+int Skein1024_InitExt(Skein1024_Ctxt_t *ctx,size_t hashBitLen,u64b_t treeInfo, const u08b_t *key, size_t keyBytes)
+    {
+    union
+        {
+        u08b_t  b[SKEIN1024_STATE_BYTES];
+        u64b_t  w[SKEIN1024_STATE_WORDS];
+        } cfg;                              /* config block */
+        
+    Skein_Assert(hashBitLen > 0,SKEIN_BAD_HASHLEN);
+    Skein_Assert(keyBytes == 0 || key != NULL,SKEIN_FAIL);
+
+    /* compute the initial chaining values ctx->X[], based on key */
+    if (keyBytes == 0)                          /* is there a key? */
+        {                                   
+        memset(ctx->X,0,sizeof(ctx->X));        /* no key: use all zeroes as key for config block */
+        }
+    else                                        /* here to pre-process a key */
+        {
+        Skein_assert(sizeof(cfg.b) >= sizeof(ctx->X));
+        /* do a mini-Init right here */
+        ctx->h.hashBitLen=8*sizeof(ctx->X);     /* set output hash bit count = state size */
+        Skein_Start_New_Type(ctx,KEY);          /* set tweaks: T0 = 0; T1 = KEY type */
+        memset(ctx->X,0,sizeof(ctx->X));        /* zero the initial chaining variables */
+        Skein1024_Update(ctx,key,keyBytes);     /* hash the key */
+        Skein1024_Final_Pad(ctx,cfg.b);         /* put result into cfg.b[] */
+        memcpy(ctx->X,cfg.b,sizeof(cfg.b));     /* copy over into ctx->X[] */
+#if SKEIN_NEED_SWAP
+        {
+        uint_t i;
+        for (i=0;i<SKEIN1024_STATE_WORDS;i++)   /* convert key bytes to context words */
+            ctx->X[i] = Skein_Swap64(ctx->X[i]);
+        }
+#endif
+        }
+    /* build/process the config block, type == CONFIG (could be precomputed for each key) */
+    ctx->h.hashBitLen = hashBitLen;             /* output hash bit count */
+    Skein_Start_New_Type(ctx,CFG_FINAL);
+
+    memset(&cfg.w,0,sizeof(cfg.w));             /* pre-pad cfg.w[] with zeroes */
+    cfg.w[0] = Skein_Swap64(SKEIN_SCHEMA_VER);
+    cfg.w[1] = Skein_Swap64(hashBitLen);        /* hash result length in bits */
+    cfg.w[2] = Skein_Swap64(treeInfo);          /* tree hash config info (or SKEIN_CFG_TREE_INFO_SEQUENTIAL) */
+
+    Skein_Show_Key(1024,&ctx->h,key,keyBytes);
+
+    /* compute the initial chaining values from config block */
+    Skein1024_Process_Block(ctx,cfg.b,1,SKEIN_CFG_STR_LEN);
+
+    /* The chaining vars ctx->X are now initialized */
+    /* Set up to process the data message portion of the hash (default) */
+    ctx->h.bCnt = 0;                            /* buffer b[] starts out empty */
+    Skein_Start_New_Type(ctx,MSG);
+    
+    return SKEIN_SUCCESS;
+    }
+
+/*++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++*/
+/* process the input bytes */
+int Skein1024_Update(Skein1024_Ctxt_t *ctx, const u08b_t *msg, size_t msgByteCnt)
+    {
+    size_t n;
+
+    Skein_Assert(ctx->h.bCnt <= SKEIN1024_BLOCK_BYTES,SKEIN_FAIL);    /* catch uninitialized context */
+
+    /* process full blocks, if any */
+    if (msgByteCnt + ctx->h.bCnt > SKEIN1024_BLOCK_BYTES)
+        {
+        if (ctx->h.bCnt)                              /* finish up any buffered message data */
+            {
+            n = SKEIN1024_BLOCK_BYTES - ctx->h.bCnt;  /* # bytes free in buffer b[] */
+            if (n)
+                {
+                Skein_assert(n < msgByteCnt);         /* check on our logic here */
+                memcpy(&ctx->b[ctx->h.bCnt],msg,n);
+                msgByteCnt  -= n;
+                msg         += n;
+                ctx->h.bCnt += n;
+                }
+            Skein_assert(ctx->h.bCnt == SKEIN1024_BLOCK_BYTES);
+            Skein1024_Process_Block(ctx,ctx->b,1,SKEIN1024_BLOCK_BYTES);
+            ctx->h.bCnt = 0;
+            }
+        /* now process any remaining full blocks, directly from input message data */
+        if (msgByteCnt > SKEIN1024_BLOCK_BYTES)
+            {
+            n = (msgByteCnt-1) / SKEIN1024_BLOCK_BYTES;   /* number of full blocks to process */
+            Skein1024_Process_Block(ctx,msg,n,SKEIN1024_BLOCK_BYTES);
+            msgByteCnt -= n * SKEIN1024_BLOCK_BYTES;
+            msg        += n * SKEIN1024_BLOCK_BYTES;
+            }
+        Skein_assert(ctx->h.bCnt == 0);
+        }
+
+    /* copy any remaining source message data bytes into b[] */
+    if (msgByteCnt)
+        {
+        Skein_assert(msgByteCnt + ctx->h.bCnt <= SKEIN1024_BLOCK_BYTES);
+        memcpy(&ctx->b[ctx->h.bCnt],msg,msgByteCnt);
+        ctx->h.bCnt += msgByteCnt;
+        }
+
+    return SKEIN_SUCCESS;
+    }
+   
+/*++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++*/
+/* finalize the hash computation and output the result */
+int Skein1024_Final(Skein1024_Ctxt_t *ctx, u08b_t *hashVal)
+    {
+    size_t i,n,byteCnt;
+    u64b_t X[SKEIN1024_STATE_WORDS];
+    Skein_Assert(ctx->h.bCnt <= SKEIN1024_BLOCK_BYTES,SKEIN_FAIL);    /* catch uninitialized context */
+
+    ctx->h.T[1] |= SKEIN_T1_FLAG_FINAL;                 /* tag as the final block */
+    if (ctx->h.bCnt < SKEIN1024_BLOCK_BYTES)            /* zero pad b[] if necessary */
+        memset(&ctx->b[ctx->h.bCnt],0,SKEIN1024_BLOCK_BYTES - ctx->h.bCnt);
+
+    Skein1024_Process_Block(ctx,ctx->b,1,ctx->h.bCnt);  /* process the final block */
+    
+    /* now output the result */
+    byteCnt = (ctx->h.hashBitLen + 7) >> 3;             /* total number of output bytes */
+
+    /* run Threefish in "counter mode" to generate output */
+    memset(ctx->b,0,sizeof(ctx->b));  /* zero out b[], so it can hold the counter */
+    memcpy(X,ctx->X,sizeof(X));       /* keep a local copy of counter mode "key" */
+    for (i=0;i*SKEIN1024_BLOCK_BYTES < byteCnt;i++)
+        {
+        ((u64b_t *)ctx->b)[0]= Skein_Swap64((u64b_t) i); /* build the counter block */
+        Skein_Start_New_Type(ctx,OUT_FINAL);
+        Skein1024_Process_Block(ctx,ctx->b,1,sizeof(u64b_t)); /* run "counter mode" */
+        n = byteCnt - i*SKEIN1024_BLOCK_BYTES;   /* number of output bytes left to go */
+        if (n >= SKEIN1024_BLOCK_BYTES)
+            n  = SKEIN1024_BLOCK_BYTES;
+        Skein_Put64_LSB_First(hashVal+i*SKEIN1024_BLOCK_BYTES,ctx->X,n);   /* "output" the ctr mode bytes */
+        Skein_Show_Final(1024,&ctx->h,n,hashVal+i*SKEIN1024_BLOCK_BYTES);
+        memcpy(ctx->X,X,sizeof(X));   /* restore the counter mode key for next time */
+        }
+    return SKEIN_SUCCESS;
+    }
+
+#if defined(SKEIN_CODE_SIZE) || defined(SKEIN_PERF)
+size_t Skein1024_API_CodeSize(void)
+    {
+    return ((u08b_t *) Skein1024_API_CodeSize) -
+           ((u08b_t *) Skein1024_Init);
+    }
+#endif
+
+/**************** Functions to support MAC/tree hashing ***************/
+/*   (this code is identical for Optimized and Reference versions)    */
+
+/*++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++*/
+/* finalize the hash computation and output the block, no OUTPUT stage */
+int Skein_256_Final_Pad(Skein_256_Ctxt_t *ctx, u08b_t *hashVal)
+    {
+    Skein_Assert(ctx->h.bCnt <= SKEIN_256_BLOCK_BYTES,SKEIN_FAIL);    /* catch uninitialized context */
+
+    ctx->h.T[1] |= SKEIN_T1_FLAG_FINAL;        /* tag as the final block */
+    if (ctx->h.bCnt < SKEIN_256_BLOCK_BYTES)   /* zero pad b[] if necessary */
+        memset(&ctx->b[ctx->h.bCnt],0,SKEIN_256_BLOCK_BYTES - ctx->h.bCnt);
+    Skein_256_Process_Block(ctx,ctx->b,1,ctx->h.bCnt);    /* process the final block */
+    
+    Skein_Put64_LSB_First(hashVal,ctx->X,SKEIN_256_BLOCK_BYTES);   /* "output" the state bytes */
+    
+    return SKEIN_SUCCESS;
+    }
+
+/*++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++*/
+/* finalize the hash computation and output the block, no OUTPUT stage */
+int Skein_512_Final_Pad(Skein_512_Ctxt_t *ctx, u08b_t *hashVal)
+    {
+    Skein_Assert(ctx->h.bCnt <= SKEIN_512_BLOCK_BYTES,SKEIN_FAIL);    /* catch uninitialized context */
+
+    ctx->h.T[1] |= SKEIN_T1_FLAG_FINAL;        /* tag as the final block */
+    if (ctx->h.bCnt < SKEIN_512_BLOCK_BYTES)   /* zero pad b[] if necessary */
+        memset(&ctx->b[ctx->h.bCnt],0,SKEIN_512_BLOCK_BYTES - ctx->h.bCnt);
+    Skein_512_Process_Block(ctx,ctx->b,1,ctx->h.bCnt);    /* process the final block */
+    
+    Skein_Put64_LSB_First(hashVal,ctx->X,SKEIN_512_BLOCK_BYTES);   /* "output" the state bytes */
+    
+    return SKEIN_SUCCESS;
+    }
+
+/*++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++*/
+/* finalize the hash computation and output the block, no OUTPUT stage */
+int Skein1024_Final_Pad(Skein1024_Ctxt_t *ctx, u08b_t *hashVal)
+    {
+    Skein_Assert(ctx->h.bCnt <= SKEIN1024_BLOCK_BYTES,SKEIN_FAIL);    /* catch uninitialized context */
+
+    ctx->h.T[1] |= SKEIN_T1_FLAG_FINAL;        /* tag as the final block */
+    if (ctx->h.bCnt < SKEIN1024_BLOCK_BYTES)   /* zero pad b[] if necessary */
+        memset(&ctx->b[ctx->h.bCnt],0,SKEIN1024_BLOCK_BYTES - ctx->h.bCnt);
+    Skein1024_Process_Block(ctx,ctx->b,1,ctx->h.bCnt);    /* process the final block */
+    
+    Skein_Put64_LSB_First(hashVal,ctx->X,SKEIN1024_BLOCK_BYTES);   /* "output" the state bytes */
+    
+    return SKEIN_SUCCESS;
+    }
+
+#if SKEIN_TREE_HASH
+/*++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++*/
+/* just do the OUTPUT stage                                       */
+int Skein_256_Output(Skein_256_Ctxt_t *ctx, u08b_t *hashVal)
+    {
+    size_t i,n,byteCnt;
+    u64b_t X[SKEIN_256_STATE_WORDS];
+    Skein_Assert(ctx->h.bCnt <= SKEIN_256_BLOCK_BYTES,SKEIN_FAIL);    /* catch uninitialized context */
+
+    /* now output the result */
+    byteCnt = (ctx->h.hashBitLen + 7) >> 3;    /* total number of output bytes */
+
+    /* run Threefish in "counter mode" to generate output */
+    memset(ctx->b,0,sizeof(ctx->b));  /* zero out b[], so it can hold the counter */
+    memcpy(X,ctx->X,sizeof(X));       /* keep a local copy of counter mode "key" */
+    for (i=0;i*SKEIN_256_BLOCK_BYTES < byteCnt;i++)
+        {
+        ((u64b_t *)ctx->b)[0]= Skein_Swap64((u64b_t) i); /* build the counter block */
+        Skein_Start_New_Type(ctx,OUT_FINAL);
+        Skein_256_Process_Block(ctx,ctx->b,1,sizeof(u64b_t)); /* run "counter mode" */
+        n = byteCnt - i*SKEIN_256_BLOCK_BYTES;   /* number of output bytes left to go */
+        if (n >= SKEIN_256_BLOCK_BYTES)
+            n  = SKEIN_256_BLOCK_BYTES;
+        Skein_Put64_LSB_First(hashVal+i*SKEIN_256_BLOCK_BYTES,ctx->X,n);   /* "output" the ctr mode bytes */
+        Skein_Show_Final(256,&ctx->h,n,hashVal+i*SKEIN_256_BLOCK_BYTES);
+        memcpy(ctx->X,X,sizeof(X));   /* restore the counter mode key for next time */
+        }
+    return SKEIN_SUCCESS;
+    }
+
+/*++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++*/
+/* just do the OUTPUT stage                                       */
+int Skein_512_Output(Skein_512_Ctxt_t *ctx, u08b_t *hashVal)
+    {
+    size_t i,n,byteCnt;
+    u64b_t X[SKEIN_512_STATE_WORDS];
+    Skein_Assert(ctx->h.bCnt <= SKEIN_512_BLOCK_BYTES,SKEIN_FAIL);    /* catch uninitialized context */
+
+    /* now output the result */
+    byteCnt = (ctx->h.hashBitLen + 7) >> 3;    /* total number of output bytes */
+
+    /* run Threefish in "counter mode" to generate output */
+    memset(ctx->b,0,sizeof(ctx->b));  /* zero out b[], so it can hold the counter */
+    memcpy(X,ctx->X,sizeof(X));       /* keep a local copy of counter mode "key" */
+    for (i=0;i*SKEIN_512_BLOCK_BYTES < byteCnt;i++)
+        {
+        ((u64b_t *)ctx->b)[0]= Skein_Swap64((u64b_t) i); /* build the counter block */
+        Skein_Start_New_Type(ctx,OUT_FINAL);
+        Skein_512_Process_Block(ctx,ctx->b,1,sizeof(u64b_t)); /* run "counter mode" */
+        n = byteCnt - i*SKEIN_512_BLOCK_BYTES;   /* number of output bytes left to go */
+        if (n >= SKEIN_512_BLOCK_BYTES)
+            n  = SKEIN_512_BLOCK_BYTES;
+        Skein_Put64_LSB_First(hashVal+i*SKEIN_512_BLOCK_BYTES,ctx->X,n);   /* "output" the ctr mode bytes */
+        Skein_Show_Final(256,&ctx->h,n,hashVal+i*SKEIN_512_BLOCK_BYTES);
+        memcpy(ctx->X,X,sizeof(X));   /* restore the counter mode key for next time */
+        }
+    return SKEIN_SUCCESS;
+    }
+
+/*++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++*/
+/* just do the OUTPUT stage                                       */
+int Skein1024_Output(Skein1024_Ctxt_t *ctx, u08b_t *hashVal)
+    {
+    size_t i,n,byteCnt;
+    u64b_t X[SKEIN1024_STATE_WORDS];
+    Skein_Assert(ctx->h.bCnt <= SKEIN1024_BLOCK_BYTES,SKEIN_FAIL);    /* catch uninitialized context */
+
+    /* now output the result */
+    byteCnt = (ctx->h.hashBitLen + 7) >> 3;    /* total number of output bytes */
+
+    /* run Threefish in "counter mode" to generate output */
+    memset(ctx->b,0,sizeof(ctx->b));  /* zero out b[], so it can hold the counter */
+    memcpy(X,ctx->X,sizeof(X));       /* keep a local copy of counter mode "key" */
+    for (i=0;i*SKEIN1024_BLOCK_BYTES < byteCnt;i++)
+        {
+        ((u64b_t *)ctx->b)[0]= Skein_Swap64((u64b_t) i); /* build the counter block */
+        Skein_Start_New_Type(ctx,OUT_FINAL);
+        Skein1024_Process_Block(ctx,ctx->b,1,sizeof(u64b_t)); /* run "counter mode" */
+        n = byteCnt - i*SKEIN1024_BLOCK_BYTES;   /* number of output bytes left to go */
+        if (n >= SKEIN1024_BLOCK_BYTES)
+            n  = SKEIN1024_BLOCK_BYTES;
+        Skein_Put64_LSB_First(hashVal+i*SKEIN1024_BLOCK_BYTES,ctx->X,n);   /* "output" the ctr mode bytes */
+        Skein_Show_Final(256,&ctx->h,n,hashVal+i*SKEIN1024_BLOCK_BYTES);
+        memcpy(ctx->X,X,sizeof(X));   /* restore the counter mode key for next time */
+        }
+    return SKEIN_SUCCESS;
+    }
+
+
+/* Adapt the functions to match the prototype expected by libmd */
+void
+SKEIN256_Init(SKEIN256_CTX * ctx)
+{
+
+	Skein_256_Init(ctx, 256);
+}
+
+void
+SKEIN512_Init(SKEIN512_CTX * ctx)
+{
+
+	Skein_512_Init(ctx, 512);
+}
+
+void
+SKEIN1024_Init(SKEIN1024_CTX * ctx)
+{
+
+	Skein1024_Init(ctx, 1024);
+}
+
+void
+SKEIN256_Update(SKEIN256_CTX * ctx, const void *in, size_t len)
+{
+
+	Skein_256_Update(ctx, in, len);
+}
+
+void
+SKEIN512_Update(SKEIN512_CTX * ctx, const void *in, size_t len)
+{
+
+	Skein_512_Update(ctx, in, len);
+}
+
+void
+SKEIN1024_Update(SKEIN1024_CTX * ctx, const void *in, size_t len)
+{
+
+	Skein1024_Update(ctx, in, len);
+}
+
+void
+SKEIN256_Final(unsigned char digest[static SKEIN_256_BLOCK_BYTES], SKEIN256_CTX *ctx)
+{
+
+	Skein_256_Final(ctx, digest);
+}
+
+void
+SKEIN512_Final(unsigned char digest[static SKEIN_512_BLOCK_BYTES], SKEIN512_CTX *ctx)
+{
+
+	Skein_512_Final(ctx, digest);
+}
+
+void
+SKEIN1024_Final(unsigned char digest[static SKEIN1024_BLOCK_BYTES], SKEIN1024_CTX *ctx)
+{
+
+	Skein1024_Final(ctx, digest);
+}
+
+#ifdef WEAK_REFS
+/* When building libmd, provide weak references. Note: this is not
+   activated in the context of compiling these sources for internal
+   use in libcrypt.
+ */
+#undef SKEIN256_Init
+__weak_reference(_libmd_SKEIN256_Init, SKEIN256_Init);
+#undef SKEIN256_Update
+__weak_reference(_libmd_SKEIN256_Update, SKEIN256_Update);
+#undef SKEIN256_Final
+__weak_reference(_libmd_SKEIN256_Final, SKEIN256_Final);
+
+#undef SKEIN512_Init
+__weak_reference(_libmd_SKEIN512_Init, SKEIN512_Init);
+#undef SKEIN512_Update
+__weak_reference(_libmd_SKEIN512_Update, SKEIN512_Update);
+#undef SKEIN512_Final
+__weak_reference(_libmd_SKEIN512_Final, SKEIN512_Final);
+
+#undef SKEIN1024_Init
+__weak_reference(_libmd_SKEIN1024_Init, SKEIN1024_Init);
+#undef SKEIN1024_Update
+__weak_reference(_libmd_SKEIN1024_Update, SKEIN1024_Update);
+#undef SKEIN1024_Final
+__weak_reference(_libmd_SKEIN1024_Final, SKEIN1024_Final);
+#endif
+
+#endif
diff --git a/sys/crypto/skein/skein.h b/sys/crypto/skein/skein.h
new file mode 100644
index 0000000..ca919ac
--- /dev/null
+++ b/sys/crypto/skein/skein.h
@@ -0,0 +1,333 @@
+/*	$FreeBSD$	*/
+#ifndef _SKEIN_H_
+#define _SKEIN_H_     1
+/**************************************************************************
+**
+** Interface declarations and internal definitions for Skein hashing.
+**
+** Source code author: Doug Whiting, 2008.
+**
+** This algorithm and source code is released to the public domain.
+**
+***************************************************************************
+** 
+** The following compile-time switches may be defined to control some
+** tradeoffs between speed, code size, error checking, and security.
+**
+** The "default" note explains what happens when the switch is not defined.
+**
+**  SKEIN_DEBUG            -- make callouts from inside Skein code
+**                            to examine/display intermediate values.
+**                            [default: no callouts (no overhead)]
+**
+**  SKEIN_ERR_CHECK        -- how error checking is handled inside Skein
+**                            code. If not defined, most error checking 
+**                            is disabled (for performance). Otherwise, 
+**                            the switch value is interpreted as:
+**                                0: use assert()      to flag errors
+**                                1: return SKEIN_FAIL to flag errors
+**
+***************************************************************************/
+#ifdef __cplusplus
+extern "C"
+{
+#endif
+
+#ifndef _KERNEL
+#include <stddef.h>                          /* get size_t definition */
+#endif
+#include "skein_port.h"                      /* get platform-specific definitions */
+
+enum
+    {
+    SKEIN_SUCCESS         =      0,          /* return codes from Skein calls */
+    SKEIN_FAIL            =      1,
+    SKEIN_BAD_HASHLEN     =      2
+    };
+
+#define  SKEIN_MODIFIER_WORDS  ( 2)          /* number of modifier (tweak) words */
+
+#define  SKEIN_256_STATE_WORDS ( 4)
+#define  SKEIN_512_STATE_WORDS ( 8)
+#define  SKEIN1024_STATE_WORDS (16)
+#define  SKEIN_MAX_STATE_WORDS (16)
+
+#define  SKEIN_256_STATE_BYTES ( 8*SKEIN_256_STATE_WORDS)
+#define  SKEIN_512_STATE_BYTES ( 8*SKEIN_512_STATE_WORDS)
+#define  SKEIN1024_STATE_BYTES ( 8*SKEIN1024_STATE_WORDS)
+
+#define  SKEIN_256_STATE_BITS  (64*SKEIN_256_STATE_WORDS)
+#define  SKEIN_512_STATE_BITS  (64*SKEIN_512_STATE_WORDS)
+#define  SKEIN1024_STATE_BITS  (64*SKEIN1024_STATE_WORDS)
+
+#define  SKEIN_256_BLOCK_BYTES ( 8*SKEIN_256_STATE_WORDS)
+#define  SKEIN_512_BLOCK_BYTES ( 8*SKEIN_512_STATE_WORDS)
+#define  SKEIN1024_BLOCK_BYTES ( 8*SKEIN1024_STATE_WORDS)
+
+typedef struct
+    {
+    size_t  hashBitLen;                      /* size of hash result, in bits */
+    size_t  bCnt;                            /* current byte count in buffer b[] */
+    u64b_t  T[SKEIN_MODIFIER_WORDS];         /* tweak words: T[0]=byte cnt, T[1]=flags */
+    } Skein_Ctxt_Hdr_t;
+
+typedef struct                               /*  256-bit Skein hash context structure */
+    {
+    Skein_Ctxt_Hdr_t h;                      /* common header context variables */
+    u64b_t  X[SKEIN_256_STATE_WORDS];        /* chaining variables */
+    u08b_t  b[SKEIN_256_BLOCK_BYTES];        /* partial block buffer (8-byte aligned) */
+    } Skein_256_Ctxt_t;
+
+typedef struct                               /*  512-bit Skein hash context structure */
+    {
+    Skein_Ctxt_Hdr_t h;                      /* common header context variables */
+    u64b_t  X[SKEIN_512_STATE_WORDS];        /* chaining variables */
+    u08b_t  b[SKEIN_512_BLOCK_BYTES];        /* partial block buffer (8-byte aligned) */
+    } Skein_512_Ctxt_t;
+
+typedef struct                               /* 1024-bit Skein hash context structure */
+    {
+    Skein_Ctxt_Hdr_t h;                      /* common header context variables */
+    u64b_t  X[SKEIN1024_STATE_WORDS];        /* chaining variables */
+    u08b_t  b[SKEIN1024_BLOCK_BYTES];        /* partial block buffer (8-byte aligned) */
+    } Skein1024_Ctxt_t;
+
+/*   Skein APIs for (incremental) "straight hashing" */
+int  Skein_256_Init  (Skein_256_Ctxt_t *ctx, size_t hashBitLen);
+int  Skein_512_Init  (Skein_512_Ctxt_t *ctx, size_t hashBitLen);
+int  Skein1024_Init  (Skein1024_Ctxt_t *ctx, size_t hashBitLen);
+
+int  Skein_256_Update(Skein_256_Ctxt_t *ctx, const u08b_t *msg, size_t msgByteCnt);
+int  Skein_512_Update(Skein_512_Ctxt_t *ctx, const u08b_t *msg, size_t msgByteCnt);
+int  Skein1024_Update(Skein1024_Ctxt_t *ctx, const u08b_t *msg, size_t msgByteCnt);
+
+int  Skein_256_Final (Skein_256_Ctxt_t *ctx, u08b_t * hashVal);
+int  Skein_512_Final (Skein_512_Ctxt_t *ctx, u08b_t * hashVal);
+int  Skein1024_Final (Skein1024_Ctxt_t *ctx, u08b_t * hashVal);
+
+/*
+**   Skein APIs for "extended" initialization: MAC keys, tree hashing.
+**   After an InitExt() call, just use Update/Final calls as with Init().
+**
+**   Notes: Same parameters as _Init() calls, plus treeInfo/key/keyBytes.
+**          When keyBytes == 0 and treeInfo == SKEIN_SEQUENTIAL, 
+**              the results of InitExt() are identical to calling Init().
+**          The function Init() may be called once to "precompute" the IV for
+**              a given hashBitLen value, then by saving a copy of the context
+**              the IV computation may be avoided in later calls.
+**          Similarly, the function InitExt() may be called once per MAC key 
+**              to precompute the MAC IV, then a copy of the context saved and
+**              reused for each new MAC computation.
+**/
+int  Skein_256_InitExt(Skein_256_Ctxt_t *ctx, size_t hashBitLen, u64b_t treeInfo, const u08b_t *key, size_t keyBytes);
+int  Skein_512_InitExt(Skein_512_Ctxt_t *ctx, size_t hashBitLen, u64b_t treeInfo, const u08b_t *key, size_t keyBytes);
+int  Skein1024_InitExt(Skein1024_Ctxt_t *ctx, size_t hashBitLen, u64b_t treeInfo, const u08b_t *key, size_t keyBytes);
+
+/*
+**   Skein APIs for MAC and tree hash:
+**      Final_Pad:  pad, do final block, but no OUTPUT type
+**      Output:     do just the output stage
+*/
+int  Skein_256_Final_Pad(Skein_256_Ctxt_t *ctx, u08b_t * hashVal);
+int  Skein_512_Final_Pad(Skein_512_Ctxt_t *ctx, u08b_t * hashVal);
+int  Skein1024_Final_Pad(Skein1024_Ctxt_t *ctx, u08b_t * hashVal);
+
+#ifndef SKEIN_TREE_HASH
+#define SKEIN_TREE_HASH (1)
+#endif
+#if  SKEIN_TREE_HASH
+int  Skein_256_Output   (Skein_256_Ctxt_t *ctx, u08b_t * hashVal);
+int  Skein_512_Output   (Skein_512_Ctxt_t *ctx, u08b_t * hashVal);
+int  Skein1024_Output   (Skein1024_Ctxt_t *ctx, u08b_t * hashVal);
+#endif
+
+/*****************************************************************
+** "Internal" Skein definitions
+**    -- not needed for sequential hashing API, but will be 
+**           helpful for other uses of Skein (e.g., tree hash mode).
+**    -- included here so that they can be shared between
+**           reference and optimized code.
+******************************************************************/
+
+/* tweak word T[1]: bit field starting positions */
+#define SKEIN_T1_BIT(BIT)       ((BIT) - 64)            /* offset 64 because it's the second word  */
+                                
+#define SKEIN_T1_POS_TREE_LVL   SKEIN_T1_BIT(112)       /* bits 112..118: level in hash tree       */
+#define SKEIN_T1_POS_BIT_PAD    SKEIN_T1_BIT(119)       /* bit  119     : partial final input byte */
+#define SKEIN_T1_POS_BLK_TYPE   SKEIN_T1_BIT(120)       /* bits 120..125: type field               */
+#define SKEIN_T1_POS_FIRST      SKEIN_T1_BIT(126)       /* bits 126     : first block flag         */
+#define SKEIN_T1_POS_FINAL      SKEIN_T1_BIT(127)       /* bit  127     : final block flag         */
+                                
+/* tweak word T[1]: flag bit definition(s) */
+#define SKEIN_T1_FLAG_FIRST     (((u64b_t)  1 ) << SKEIN_T1_POS_FIRST)
+#define SKEIN_T1_FLAG_FINAL     (((u64b_t)  1 ) << SKEIN_T1_POS_FINAL)
+#define SKEIN_T1_FLAG_BIT_PAD   (((u64b_t)  1 ) << SKEIN_T1_POS_BIT_PAD)
+                                
+/* tweak word T[1]: tree level bit field mask */
+#define SKEIN_T1_TREE_LVL_MASK  (((u64b_t)0x7F) << SKEIN_T1_POS_TREE_LVL)
+#define SKEIN_T1_TREE_LEVEL(n)  (((u64b_t) (n)) << SKEIN_T1_POS_TREE_LVL)
+
+/* tweak word T[1]: block type field */
+#define SKEIN_BLK_TYPE_KEY      ( 0)                    /* key, for MAC and KDF */
+#define SKEIN_BLK_TYPE_CFG      ( 4)                    /* configuration block */
+#define SKEIN_BLK_TYPE_PERS     ( 8)                    /* personalization string */
+#define SKEIN_BLK_TYPE_PK       (12)                    /* public key (for digital signature hashing) */
+#define SKEIN_BLK_TYPE_KDF      (16)                    /* key identifier for KDF */
+#define SKEIN_BLK_TYPE_NONCE    (20)                    /* nonce for PRNG */
+#define SKEIN_BLK_TYPE_MSG      (48)                    /* message processing */
+#define SKEIN_BLK_TYPE_OUT      (63)                    /* output stage */
+#define SKEIN_BLK_TYPE_MASK     (63)                    /* bit field mask */
+
+#define SKEIN_T1_BLK_TYPE(T)   (((u64b_t) (SKEIN_BLK_TYPE_##T)) << SKEIN_T1_POS_BLK_TYPE)
+#define SKEIN_T1_BLK_TYPE_KEY   SKEIN_T1_BLK_TYPE(KEY)  /* key, for MAC and KDF */
+#define SKEIN_T1_BLK_TYPE_CFG   SKEIN_T1_BLK_TYPE(CFG)  /* configuration block */
+#define SKEIN_T1_BLK_TYPE_PERS  SKEIN_T1_BLK_TYPE(PERS) /* personalization string */
+#define SKEIN_T1_BLK_TYPE_PK    SKEIN_T1_BLK_TYPE(PK)   /* public key (for digital signature hashing) */
+#define SKEIN_T1_BLK_TYPE_KDF   SKEIN_T1_BLK_TYPE(KDF)  /* key identifier for KDF */
+#define SKEIN_T1_BLK_TYPE_NONCE SKEIN_T1_BLK_TYPE(NONCE)/* nonce for PRNG */
+#define SKEIN_T1_BLK_TYPE_MSG   SKEIN_T1_BLK_TYPE(MSG)  /* message processing */
+#define SKEIN_T1_BLK_TYPE_OUT   SKEIN_T1_BLK_TYPE(OUT)  /* output stage */
+#define SKEIN_T1_BLK_TYPE_MASK  SKEIN_T1_BLK_TYPE(MASK) /* field bit mask */
+
+#define SKEIN_T1_BLK_TYPE_CFG_FINAL       (SKEIN_T1_BLK_TYPE_CFG | SKEIN_T1_FLAG_FINAL)
+#define SKEIN_T1_BLK_TYPE_OUT_FINAL       (SKEIN_T1_BLK_TYPE_OUT | SKEIN_T1_FLAG_FINAL)
+
+#define SKEIN_VERSION           (1)
+
+#ifndef SKEIN_ID_STRING_LE      /* allow compile-time personalization */
+#define SKEIN_ID_STRING_LE      (0x33414853)            /* "SHA3" (little-endian)*/
+#endif
+
+#define SKEIN_MK_64(hi32,lo32)  ((lo32) + (((u64b_t) (hi32)) << 32))
+#define SKEIN_SCHEMA_VER        SKEIN_MK_64(SKEIN_VERSION,SKEIN_ID_STRING_LE)
+#define SKEIN_KS_PARITY         SKEIN_MK_64(0x1BD11BDA,0xA9FC1A22)
+
+#define SKEIN_CFG_STR_LEN       (4*8)
+
+/* bit field definitions in config block treeInfo word */
+#define SKEIN_CFG_TREE_LEAF_SIZE_POS  ( 0)
+#define SKEIN_CFG_TREE_NODE_SIZE_POS  ( 8)
+#define SKEIN_CFG_TREE_MAX_LEVEL_POS  (16)
+
+#define SKEIN_CFG_TREE_LEAF_SIZE_MSK  (((u64b_t) 0xFF) << SKEIN_CFG_TREE_LEAF_SIZE_POS)
+#define SKEIN_CFG_TREE_NODE_SIZE_MSK  (((u64b_t) 0xFF) << SKEIN_CFG_TREE_NODE_SIZE_POS)
+#define SKEIN_CFG_TREE_MAX_LEVEL_MSK  (((u64b_t) 0xFF) << SKEIN_CFG_TREE_MAX_LEVEL_POS)
+
+#define SKEIN_CFG_TREE_INFO(leaf,node,maxLvl)                   \
+    ( (((u64b_t)(leaf  )) << SKEIN_CFG_TREE_LEAF_SIZE_POS) |    \
+      (((u64b_t)(node  )) << SKEIN_CFG_TREE_NODE_SIZE_POS) |    \
+      (((u64b_t)(maxLvl)) << SKEIN_CFG_TREE_MAX_LEVEL_POS) )
+
+#define SKEIN_CFG_TREE_INFO_SEQUENTIAL SKEIN_CFG_TREE_INFO(0,0,0) /* use as treeInfo in InitExt() call for sequential processing */
+
+/*
+**   Skein macros for getting/setting tweak words, etc.
+**   These are useful for partial input bytes, hash tree init/update, etc.
+**/
+#define Skein_Get_Tweak(ctxPtr,TWK_NUM)         ((ctxPtr)->h.T[TWK_NUM])
+#define Skein_Set_Tweak(ctxPtr,TWK_NUM,tVal)    {(ctxPtr)->h.T[TWK_NUM] = (tVal);}
+
+#define Skein_Get_T0(ctxPtr)    Skein_Get_Tweak(ctxPtr,0)
+#define Skein_Get_T1(ctxPtr)    Skein_Get_Tweak(ctxPtr,1)
+#define Skein_Set_T0(ctxPtr,T0) Skein_Set_Tweak(ctxPtr,0,T0)
+#define Skein_Set_T1(ctxPtr,T1) Skein_Set_Tweak(ctxPtr,1,T1)
+
+/* set both tweak words at once */
+#define Skein_Set_T0_T1(ctxPtr,T0,T1)           \
+    {                                           \
+    Skein_Set_T0(ctxPtr,(T0));                  \
+    Skein_Set_T1(ctxPtr,(T1));                  \
+    }
+
+#define Skein_Set_Type(ctxPtr,BLK_TYPE)         \
+    Skein_Set_T1(ctxPtr,SKEIN_T1_BLK_TYPE_##BLK_TYPE)
+
+/* set up for starting with a new type: h.T[0]=0; h.T[1] = NEW_TYPE; h.bCnt=0; */
+#define Skein_Start_New_Type(ctxPtr,BLK_TYPE)   \
+    { Skein_Set_T0_T1(ctxPtr,0,SKEIN_T1_FLAG_FIRST | SKEIN_T1_BLK_TYPE_##BLK_TYPE); (ctxPtr)->h.bCnt=0; }
+
+#define Skein_Clear_First_Flag(hdr)      { (hdr).T[1] &= ~SKEIN_T1_FLAG_FIRST;       }
+#define Skein_Set_Bit_Pad_Flag(hdr)      { (hdr).T[1] |=  SKEIN_T1_FLAG_BIT_PAD;     }
+
+#define Skein_Set_Tree_Level(hdr,height) { (hdr).T[1] |= SKEIN_T1_TREE_LEVEL(height);}
+
+/*****************************************************************
+** "Internal" Skein definitions for debugging and error checking
+******************************************************************/
+#ifdef  SKEIN_DEBUG             /* examine/display intermediate values? */
+#include "skein_debug.h"
+#else                           /* default is no callouts */
+#define Skein_Show_Block(bits,ctx,X,blkPtr,wPtr,ksEvenPtr,ksOddPtr)
+#define Skein_Show_Round(bits,ctx,r,X)
+#define Skein_Show_R_Ptr(bits,ctx,r,X_ptr)
+#define Skein_Show_Final(bits,ctx,cnt,outPtr)
+#define Skein_Show_Key(bits,ctx,key,keyBytes)
+#endif
+
+#ifndef SKEIN_ERR_CHECK        /* run-time checks (e.g., bad params, uninitialized context)? */
+#define Skein_Assert(x,retCode)/* default: ignore all Asserts, for performance */
+#define Skein_assert(x)
+#elif   defined(SKEIN_ASSERT)
+#include <assert.h>     
+#define Skein_Assert(x,retCode) assert(x) 
+#define Skein_assert(x)         assert(x) 
+#else
+#include <assert.h>     
+#define Skein_Assert(x,retCode) { if (!(x)) return retCode; } /*  caller  error */
+#define Skein_assert(x)         assert(x)                     /* internal error */
+#endif
+
+/*****************************************************************
+** Skein block function constants (shared across Ref and Opt code)
+******************************************************************/
+enum    
+    {   
+        /* Skein_256 round rotation constants */
+    R_256_0_0=14, R_256_0_1=16,
+    R_256_1_0=52, R_256_1_1=57,
+    R_256_2_0=23, R_256_2_1=40,
+    R_256_3_0= 5, R_256_3_1=37,
+    R_256_4_0=25, R_256_4_1=33,
+    R_256_5_0=46, R_256_5_1=12,
+    R_256_6_0=58, R_256_6_1=22,
+    R_256_7_0=32, R_256_7_1=32,
+
+        /* Skein_512 round rotation constants */
+    R_512_0_0=46, R_512_0_1=36, R_512_0_2=19, R_512_0_3=37,
+    R_512_1_0=33, R_512_1_1=27, R_512_1_2=14, R_512_1_3=42,
+    R_512_2_0=17, R_512_2_1=49, R_512_2_2=36, R_512_2_3=39,
+    R_512_3_0=44, R_512_3_1= 9, R_512_3_2=54, R_512_3_3=56,
+    R_512_4_0=39, R_512_4_1=30, R_512_4_2=34, R_512_4_3=24,
+    R_512_5_0=13, R_512_5_1=50, R_512_5_2=10, R_512_5_3=17,
+    R_512_6_0=25, R_512_6_1=29, R_512_6_2=39, R_512_6_3=43,
+    R_512_7_0= 8, R_512_7_1=35, R_512_7_2=56, R_512_7_3=22,
+
+        /* Skein1024 round rotation constants */
+    R1024_0_0=24, R1024_0_1=13, R1024_0_2= 8, R1024_0_3=47, R1024_0_4= 8, R1024_0_5=17, R1024_0_6=22, R1024_0_7=37,
+    R1024_1_0=38, R1024_1_1=19, R1024_1_2=10, R1024_1_3=55, R1024_1_4=49, R1024_1_5=18, R1024_1_6=23, R1024_1_7=52,
+    R1024_2_0=33, R1024_2_1= 4, R1024_2_2=51, R1024_2_3=13, R1024_2_4=34, R1024_2_5=41, R1024_2_6=59, R1024_2_7=17,
+    R1024_3_0= 5, R1024_3_1=20, R1024_3_2=48, R1024_3_3=41, R1024_3_4=47, R1024_3_5=28, R1024_3_6=16, R1024_3_7=25,
+    R1024_4_0=41, R1024_4_1= 9, R1024_4_2=37, R1024_4_3=31, R1024_4_4=12, R1024_4_5=47, R1024_4_6=44, R1024_4_7=30,
+    R1024_5_0=16, R1024_5_1=34, R1024_5_2=56, R1024_5_3=51, R1024_5_4= 4, R1024_5_5=53, R1024_5_6=42, R1024_5_7=41,
+    R1024_6_0=31, R1024_6_1=44, R1024_6_2=47, R1024_6_3=46, R1024_6_4=19, R1024_6_5=42, R1024_6_6=44, R1024_6_7=25,
+    R1024_7_0= 9, R1024_7_1=48, R1024_7_2=35, R1024_7_3=52, R1024_7_4=23, R1024_7_5=31, R1024_7_6=37, R1024_7_7=20
+    };
+
+#ifndef SKEIN_ROUNDS
+#define SKEIN_256_ROUNDS_TOTAL (72)          /* number of rounds for the different block sizes */
+#define SKEIN_512_ROUNDS_TOTAL (72)
+#define SKEIN1024_ROUNDS_TOTAL (80)
+#else                                        /* allow command-line define in range 8*(5..14)   */
+#define SKEIN_256_ROUNDS_TOTAL (8*((((SKEIN_ROUNDS/100) + 5) % 10) + 5))
+#define SKEIN_512_ROUNDS_TOTAL (8*((((SKEIN_ROUNDS/ 10) + 5) % 10) + 5))
+#define SKEIN1024_ROUNDS_TOTAL (8*((((SKEIN_ROUNDS    ) + 5) % 10) + 5))
+#endif
+
+#ifdef __cplusplus
+}
+#endif
+
+/* Pull in FreeBSD specific shims */
+#include "skein_freebsd.h"
+
+#endif  /* ifndef _SKEIN_H_ */
diff --git a/sys/crypto/skein/skein_block.c b/sys/crypto/skein/skein_block.c
new file mode 100644
index 0000000..c924550
--- /dev/null
+++ b/sys/crypto/skein/skein_block.c
@@ -0,0 +1,706 @@
+/***********************************************************************
+**
+** Implementation of the Skein block functions.
+**
+** Source code author: Doug Whiting, 2008.
+**
+** This algorithm and source code is released to the public domain.
+**
+** Compile-time switches:
+**
+**  SKEIN_USE_ASM  -- set bits (256/512/1024) to select which
+**                    versions use ASM code for block processing
+**                    [default: use C for all block sizes]
+**
+************************************************************************/
+
+#include <sys/cdefs.h>
+__FBSDID("$FreeBSD$");
+
+#include <sys/endian.h>
+#include <sys/types.h>
+
+#ifdef _KERNEL
+#include <sys/systm.h>
+#else
+#include <string.h>
+#endif
+
+#include "skein.h"
+
+#ifndef SKEIN_USE_ASM
+#define SKEIN_USE_ASM   (0)                     /* default is all C code (no ASM) */
+#endif
+
+#ifndef SKEIN_LOOP
+#define SKEIN_LOOP 001                          /* default: unroll 256 and 512, but not 1024 */
+#endif
+
+#define BLK_BITS        (WCNT*64)               /* some useful definitions for code here */
+#define KW_TWK_BASE     (0)
+#define KW_KEY_BASE     (3)
+#define ks              (kw + KW_KEY_BASE)                
+#define ts              (kw + KW_TWK_BASE)
+
+#ifdef SKEIN_DEBUG
+#define DebugSaveTweak(ctx) { ctx->h.T[0] = ts[0]; ctx->h.T[1] = ts[1]; }
+#else
+#define DebugSaveTweak(ctx)
+#endif
+
+/*****************************************************************/
+/* functions to process blkCnt (nonzero) full block(s) of data. */
+void    Skein_256_Process_Block(Skein_256_Ctxt_t *ctx,const u08b_t *blkPtr,size_t blkCnt,size_t byteCntAdd);
+void    Skein_512_Process_Block(Skein_512_Ctxt_t *ctx,const u08b_t *blkPtr,size_t blkCnt,size_t byteCntAdd);
+void    Skein1024_Process_Block(Skein1024_Ctxt_t *ctx,const u08b_t *blkPtr,size_t blkCnt,size_t byteCntAdd);
+
+/*****************************  Skein_256 ******************************/
+#if !(SKEIN_USE_ASM & 256)
+void Skein_256_Process_Block(Skein_256_Ctxt_t *ctx,const u08b_t *blkPtr,size_t blkCnt,size_t byteCntAdd)
+    { /* do it in C */
+    enum
+        {
+        WCNT = SKEIN_256_STATE_WORDS
+        };
+#undef  RCNT
+#define RCNT  (SKEIN_256_ROUNDS_TOTAL/8)
+
+#ifdef  SKEIN_LOOP                              /* configure how much to unroll the loop */
+#define SKEIN_UNROLL_256 (((SKEIN_LOOP)/100)%10)
+#else
+#define SKEIN_UNROLL_256 (0)
+#endif
+
+#if SKEIN_UNROLL_256
+#if (RCNT % SKEIN_UNROLL_256)
+#error "Invalid SKEIN_UNROLL_256"               /* sanity check on unroll count */
+#endif
+    size_t  r;
+    u64b_t  kw[WCNT+4+RCNT*2];                  /* key schedule words : chaining vars + tweak + "rotation"*/
+#else
+    u64b_t  kw[WCNT+4];                         /* key schedule words : chaining vars + tweak */
+#endif
+    u64b_t  X0,X1,X2,X3;                        /* local copy of context vars, for speed */
+    u64b_t  w [WCNT];                           /* local copy of input block */
+#ifdef SKEIN_DEBUG
+    const u64b_t *Xptr[4];                      /* use for debugging (help compiler put Xn in registers) */
+    Xptr[0] = &X0;  Xptr[1] = &X1;  Xptr[2] = &X2;  Xptr[3] = &X3;
+#endif
+    Skein_assert(blkCnt != 0);                  /* never call with blkCnt == 0! */
+    ts[0] = ctx->h.T[0];
+    ts[1] = ctx->h.T[1];
+    do  {
+        /* this implementation only supports 2**64 input bytes (no carry out here) */
+        ts[0] += byteCntAdd;                    /* update processed length */
+
+        /* precompute the key schedule for this block */
+        ks[0] = ctx->X[0];     
+        ks[1] = ctx->X[1];
+        ks[2] = ctx->X[2];
+        ks[3] = ctx->X[3];
+        ks[4] = ks[0] ^ ks[1] ^ ks[2] ^ ks[3] ^ SKEIN_KS_PARITY;
+
+        ts[2] = ts[0] ^ ts[1];
+
+        Skein_Get64_LSB_First(w,blkPtr,WCNT);   /* get input block in little-endian format */
+        DebugSaveTweak(ctx);
+        Skein_Show_Block(BLK_BITS,&ctx->h,ctx->X,blkPtr,w,ks,ts);
+
+        X0 = w[0] + ks[0];                      /* do the first full key injection */
+        X1 = w[1] + ks[1] + ts[0];
+        X2 = w[2] + ks[2] + ts[1];
+        X3 = w[3] + ks[3];
+
+        Skein_Show_R_Ptr(BLK_BITS,&ctx->h,SKEIN_RND_KEY_INITIAL,Xptr);    /* show starting state values */
+
+        blkPtr += SKEIN_256_BLOCK_BYTES;
+
+        /* run the rounds */
+
+#define Round256(p0,p1,p2,p3,ROT,rNum)                              \
+    X##p0 += X##p1; X##p1 = RotL_64(X##p1,ROT##_0); X##p1 ^= X##p0; \
+    X##p2 += X##p3; X##p3 = RotL_64(X##p3,ROT##_1); X##p3 ^= X##p2; \
+
+#if SKEIN_UNROLL_256 == 0                       
+#define R256(p0,p1,p2,p3,ROT,rNum)           /* fully unrolled */   \
+    Round256(p0,p1,p2,p3,ROT,rNum)                                  \
+    Skein_Show_R_Ptr(BLK_BITS,&ctx->h,rNum,Xptr);
+
+#define I256(R)                                                     \
+    X0   += ks[((R)+1) % 5];    /* inject the key schedule value */ \
+    X1   += ks[((R)+2) % 5] + ts[((R)+1) % 3];                      \
+    X2   += ks[((R)+3) % 5] + ts[((R)+2) % 3];                      \
+    X3   += ks[((R)+4) % 5] +     (R)+1;                            \
+    Skein_Show_R_Ptr(BLK_BITS,&ctx->h,SKEIN_RND_KEY_INJECT,Xptr);
+#else                                       /* looping version */
+#define R256(p0,p1,p2,p3,ROT,rNum)                                  \
+    Round256(p0,p1,p2,p3,ROT,rNum)                                  \
+    Skein_Show_R_Ptr(BLK_BITS,&ctx->h,4*(r-1)+rNum,Xptr);
+
+#define I256(R)                                                     \
+    X0   += ks[r+(R)+0];        /* inject the key schedule value */ \
+    X1   += ks[r+(R)+1] + ts[r+(R)+0];                              \
+    X2   += ks[r+(R)+2] + ts[r+(R)+1];                              \
+    X3   += ks[r+(R)+3] +    r+(R)   ;                              \
+    ks[r + (R)+4    ]   = ks[r+(R)-1];     /* rotate key schedule */\
+    ts[r + (R)+2    ]   = ts[r+(R)-1];                              \
+    Skein_Show_R_Ptr(BLK_BITS,&ctx->h,SKEIN_RND_KEY_INJECT,Xptr);
+
+    for (r=1;r < 2*RCNT;r+=2*SKEIN_UNROLL_256)  /* loop thru it */
+#endif  
+        {    
+#define R256_8_rounds(R)                  \
+        R256(0,1,2,3,R_256_0,8*(R) + 1);  \
+        R256(0,3,2,1,R_256_1,8*(R) + 2);  \
+        R256(0,1,2,3,R_256_2,8*(R) + 3);  \
+        R256(0,3,2,1,R_256_3,8*(R) + 4);  \
+        I256(2*(R));                      \
+        R256(0,1,2,3,R_256_4,8*(R) + 5);  \
+        R256(0,3,2,1,R_256_5,8*(R) + 6);  \
+        R256(0,1,2,3,R_256_6,8*(R) + 7);  \
+        R256(0,3,2,1,R_256_7,8*(R) + 8);  \
+        I256(2*(R)+1);
+
+        R256_8_rounds( 0);
+
+#define R256_Unroll_R(NN) ((SKEIN_UNROLL_256 == 0 && SKEIN_256_ROUNDS_TOTAL/8 > (NN)) || (SKEIN_UNROLL_256 > (NN)))
+
+  #if   R256_Unroll_R( 1)
+        R256_8_rounds( 1);
+  #endif
+  #if   R256_Unroll_R( 2)
+        R256_8_rounds( 2);
+  #endif
+  #if   R256_Unroll_R( 3)
+        R256_8_rounds( 3);
+  #endif
+  #if   R256_Unroll_R( 4)
+        R256_8_rounds( 4);
+  #endif
+  #if   R256_Unroll_R( 5)
+        R256_8_rounds( 5);
+  #endif
+  #if   R256_Unroll_R( 6)
+        R256_8_rounds( 6);
+  #endif
+  #if   R256_Unroll_R( 7)
+        R256_8_rounds( 7);
+  #endif
+  #if   R256_Unroll_R( 8)
+        R256_8_rounds( 8);
+  #endif
+  #if   R256_Unroll_R( 9)
+        R256_8_rounds( 9);
+  #endif
+  #if   R256_Unroll_R(10)
+        R256_8_rounds(10);
+  #endif
+  #if   R256_Unroll_R(11)
+        R256_8_rounds(11);
+  #endif
+  #if   R256_Unroll_R(12)
+        R256_8_rounds(12);
+  #endif
+  #if   R256_Unroll_R(13)
+        R256_8_rounds(13);
+  #endif
+  #if   R256_Unroll_R(14)
+        R256_8_rounds(14);
+  #endif
+  #if  (SKEIN_UNROLL_256 > 14)
+#error  "need more unrolling in Skein_256_Process_Block"
+  #endif
+        }
+        /* do the final "feedforward" xor, update context chaining vars */
+        ctx->X[0] = X0 ^ w[0];
+        ctx->X[1] = X1 ^ w[1];
+        ctx->X[2] = X2 ^ w[2];
+        ctx->X[3] = X3 ^ w[3];
+
+        Skein_Show_Round(BLK_BITS,&ctx->h,SKEIN_RND_FEED_FWD,ctx->X);
+
+        ts[1] &= ~SKEIN_T1_FLAG_FIRST;
+        }
+    while (--blkCnt);
+    ctx->h.T[0] = ts[0];
+    ctx->h.T[1] = ts[1];
+    }
+
+#if defined(SKEIN_CODE_SIZE) || defined(SKEIN_PERF)
+size_t Skein_256_Process_Block_CodeSize(void)
+    {
+    return ((u08b_t *) Skein_256_Process_Block_CodeSize) -
+           ((u08b_t *) Skein_256_Process_Block);
+    }
+uint_t Skein_256_Unroll_Cnt(void)
+    {
+    return SKEIN_UNROLL_256;
+    }
+#endif
+#endif
+
+/*****************************  Skein_512 ******************************/
+#if !(SKEIN_USE_ASM & 512)
+void Skein_512_Process_Block(Skein_512_Ctxt_t *ctx,const u08b_t *blkPtr,size_t blkCnt,size_t byteCntAdd)
+    { /* do it in C */
+    enum
+        {
+        WCNT = SKEIN_512_STATE_WORDS
+        };
+#undef  RCNT
+#define RCNT  (SKEIN_512_ROUNDS_TOTAL/8)
+
+#ifdef  SKEIN_LOOP                              /* configure how much to unroll the loop */
+#define SKEIN_UNROLL_512 (((SKEIN_LOOP)/10)%10)
+#else
+#define SKEIN_UNROLL_512 (0)
+#endif
+
+#if SKEIN_UNROLL_512
+#if (RCNT % SKEIN_UNROLL_512)
+#error "Invalid SKEIN_UNROLL_512"               /* sanity check on unroll count */
+#endif
+    size_t  r;
+    u64b_t  kw[WCNT+4+RCNT*2];                  /* key schedule words : chaining vars + tweak + "rotation"*/
+#else
+    u64b_t  kw[WCNT+4];                         /* key schedule words : chaining vars + tweak */
+#endif
+    u64b_t  X0,X1,X2,X3,X4,X5,X6,X7;            /* local copy of vars, for speed */
+    u64b_t  w [WCNT];                           /* local copy of input block */
+#ifdef SKEIN_DEBUG
+    const u64b_t *Xptr[8];                      /* use for debugging (help compiler put Xn in registers) */
+    Xptr[0] = &X0;  Xptr[1] = &X1;  Xptr[2] = &X2;  Xptr[3] = &X3;
+    Xptr[4] = &X4;  Xptr[5] = &X5;  Xptr[6] = &X6;  Xptr[7] = &X7;
+#endif
+
+    Skein_assert(blkCnt != 0);                  /* never call with blkCnt == 0! */
+    ts[0] = ctx->h.T[0];
+    ts[1] = ctx->h.T[1];
+    do  {
+        /* this implementation only supports 2**64 input bytes (no carry out here) */
+        ts[0] += byteCntAdd;                    /* update processed length */
+
+        /* precompute the key schedule for this block */
+        ks[0] = ctx->X[0];
+        ks[1] = ctx->X[1];
+        ks[2] = ctx->X[2];
+        ks[3] = ctx->X[3];
+        ks[4] = ctx->X[4];
+        ks[5] = ctx->X[5];
+        ks[6] = ctx->X[6];
+        ks[7] = ctx->X[7];
+        ks[8] = ks[0] ^ ks[1] ^ ks[2] ^ ks[3] ^ 
+                ks[4] ^ ks[5] ^ ks[6] ^ ks[7] ^ SKEIN_KS_PARITY;
+
+        ts[2] = ts[0] ^ ts[1];
+
+        Skein_Get64_LSB_First(w,blkPtr,WCNT); /* get input block in little-endian format */
+        DebugSaveTweak(ctx);
+        Skein_Show_Block(BLK_BITS,&ctx->h,ctx->X,blkPtr,w,ks,ts);
+
+        X0   = w[0] + ks[0];                    /* do the first full key injection */
+        X1   = w[1] + ks[1];
+        X2   = w[2] + ks[2];
+        X3   = w[3] + ks[3];
+        X4   = w[4] + ks[4];
+        X5   = w[5] + ks[5] + ts[0];
+        X6   = w[6] + ks[6] + ts[1];
+        X7   = w[7] + ks[7];
+
+        blkPtr += SKEIN_512_BLOCK_BYTES;
+
+        Skein_Show_R_Ptr(BLK_BITS,&ctx->h,SKEIN_RND_KEY_INITIAL,Xptr);
+        /* run the rounds */
+#define Round512(p0,p1,p2,p3,p4,p5,p6,p7,ROT,rNum)                  \
+    X##p0 += X##p1; X##p1 = RotL_64(X##p1,ROT##_0); X##p1 ^= X##p0; \
+    X##p2 += X##p3; X##p3 = RotL_64(X##p3,ROT##_1); X##p3 ^= X##p2; \
+    X##p4 += X##p5; X##p5 = RotL_64(X##p5,ROT##_2); X##p5 ^= X##p4; \
+    X##p6 += X##p7; X##p7 = RotL_64(X##p7,ROT##_3); X##p7 ^= X##p6; \
+
+#if SKEIN_UNROLL_512 == 0                       
+#define R512(p0,p1,p2,p3,p4,p5,p6,p7,ROT,rNum)      /* unrolled */  \
+    Round512(p0,p1,p2,p3,p4,p5,p6,p7,ROT,rNum)                      \
+    Skein_Show_R_Ptr(BLK_BITS,&ctx->h,rNum,Xptr);
+
+#define I512(R)                                                     \
+    X0   += ks[((R)+1) % 9];   /* inject the key schedule value */  \
+    X1   += ks[((R)+2) % 9];                                        \
+    X2   += ks[((R)+3) % 9];                                        \
+    X3   += ks[((R)+4) % 9];                                        \
+    X4   += ks[((R)+5) % 9];                                        \
+    X5   += ks[((R)+6) % 9] + ts[((R)+1) % 3];                      \
+    X6   += ks[((R)+7) % 9] + ts[((R)+2) % 3];                      \
+    X7   += ks[((R)+8) % 9] +     (R)+1;                            \
+    Skein_Show_R_Ptr(BLK_BITS,&ctx->h,SKEIN_RND_KEY_INJECT,Xptr);
+#else                                       /* looping version */
+#define R512(p0,p1,p2,p3,p4,p5,p6,p7,ROT,rNum)                      \
+    Round512(p0,p1,p2,p3,p4,p5,p6,p7,ROT,rNum)                      \
+    Skein_Show_R_Ptr(BLK_BITS,&ctx->h,4*(r-1)+rNum,Xptr);
+
+#define I512(R)                                                     \
+    X0   += ks[r+(R)+0];        /* inject the key schedule value */ \
+    X1   += ks[r+(R)+1];                                            \
+    X2   += ks[r+(R)+2];                                            \
+    X3   += ks[r+(R)+3];                                            \
+    X4   += ks[r+(R)+4];                                            \
+    X5   += ks[r+(R)+5] + ts[r+(R)+0];                              \
+    X6   += ks[r+(R)+6] + ts[r+(R)+1];                              \
+    X7   += ks[r+(R)+7] +    r+(R)   ;                              \
+    ks[r +       (R)+8] = ks[r+(R)-1];  /* rotate key schedule */   \
+    ts[r +       (R)+2] = ts[r+(R)-1];                              \
+    Skein_Show_R_Ptr(BLK_BITS,&ctx->h,SKEIN_RND_KEY_INJECT,Xptr);
+
+    for (r=1;r < 2*RCNT;r+=2*SKEIN_UNROLL_512)   /* loop thru it */
+#endif                         /* end of looped code definitions */
+        {
+#define R512_8_rounds(R)  /* do 8 full rounds */  \
+        R512(0,1,2,3,4,5,6,7,R_512_0,8*(R)+ 1);   \
+        R512(2,1,4,7,6,5,0,3,R_512_1,8*(R)+ 2);   \
+        R512(4,1,6,3,0,5,2,7,R_512_2,8*(R)+ 3);   \
+        R512(6,1,0,7,2,5,4,3,R_512_3,8*(R)+ 4);   \
+        I512(2*(R));                              \
+        R512(0,1,2,3,4,5,6,7,R_512_4,8*(R)+ 5);   \
+        R512(2,1,4,7,6,5,0,3,R_512_5,8*(R)+ 6);   \
+        R512(4,1,6,3,0,5,2,7,R_512_6,8*(R)+ 7);   \
+        R512(6,1,0,7,2,5,4,3,R_512_7,8*(R)+ 8);   \
+        I512(2*(R)+1);        /* and key injection */
+
+        R512_8_rounds( 0);
+
+#define R512_Unroll_R(NN) ((SKEIN_UNROLL_512 == 0 && SKEIN_512_ROUNDS_TOTAL/8 > (NN)) || (SKEIN_UNROLL_512 > (NN)))
+
+  #if   R512_Unroll_R( 1)
+        R512_8_rounds( 1);
+  #endif
+  #if   R512_Unroll_R( 2)
+        R512_8_rounds( 2);
+  #endif
+  #if   R512_Unroll_R( 3)
+        R512_8_rounds( 3);
+  #endif
+  #if   R512_Unroll_R( 4)
+        R512_8_rounds( 4);
+  #endif
+  #if   R512_Unroll_R( 5)
+        R512_8_rounds( 5);
+  #endif
+  #if   R512_Unroll_R( 6)
+        R512_8_rounds( 6);
+  #endif
+  #if   R512_Unroll_R( 7)
+        R512_8_rounds( 7);
+  #endif
+  #if   R512_Unroll_R( 8)
+        R512_8_rounds( 8);
+  #endif
+  #if   R512_Unroll_R( 9)
+        R512_8_rounds( 9);
+  #endif
+  #if   R512_Unroll_R(10)
+        R512_8_rounds(10);
+  #endif
+  #if   R512_Unroll_R(11)
+        R512_8_rounds(11);
+  #endif
+  #if   R512_Unroll_R(12)
+        R512_8_rounds(12);
+  #endif
+  #if   R512_Unroll_R(13)
+        R512_8_rounds(13);
+  #endif
+  #if   R512_Unroll_R(14)
+        R512_8_rounds(14);
+  #endif
+  #if  (SKEIN_UNROLL_512 > 14)
+#error  "need more unrolling in Skein_512_Process_Block"
+  #endif
+        }
+
+        /* do the final "feedforward" xor, update context chaining vars */
+        ctx->X[0] = X0 ^ w[0];
+        ctx->X[1] = X1 ^ w[1];
+        ctx->X[2] = X2 ^ w[2];
+        ctx->X[3] = X3 ^ w[3];
+        ctx->X[4] = X4 ^ w[4];
+        ctx->X[5] = X5 ^ w[5];
+        ctx->X[6] = X6 ^ w[6];
+        ctx->X[7] = X7 ^ w[7];
+        Skein_Show_Round(BLK_BITS,&ctx->h,SKEIN_RND_FEED_FWD,ctx->X);
+
+        ts[1] &= ~SKEIN_T1_FLAG_FIRST;
+        }
+    while (--blkCnt);
+    ctx->h.T[0] = ts[0];
+    ctx->h.T[1] = ts[1];
+    }
+
+#if defined(SKEIN_CODE_SIZE) || defined(SKEIN_PERF)
+size_t Skein_512_Process_Block_CodeSize(void)
+    {
+    return ((u08b_t *) Skein_512_Process_Block_CodeSize) -
+           ((u08b_t *) Skein_512_Process_Block);
+    }
+uint_t Skein_512_Unroll_Cnt(void)
+    {
+    return SKEIN_UNROLL_512;
+    }
+#endif
+#endif
+
+/*****************************  Skein1024 ******************************/
+#if !(SKEIN_USE_ASM & 1024)
+void Skein1024_Process_Block(Skein1024_Ctxt_t *ctx,const u08b_t *blkPtr,size_t blkCnt,size_t byteCntAdd)
+    { /* do it in C, always looping (unrolled is bigger AND slower!) */
+    enum
+        {
+        WCNT = SKEIN1024_STATE_WORDS
+        };
+#undef  RCNT
+#define RCNT  (SKEIN1024_ROUNDS_TOTAL/8)
+
+#ifdef  SKEIN_LOOP                              /* configure how much to unroll the loop */
+#define SKEIN_UNROLL_1024 ((SKEIN_LOOP)%10)
+#else
+#define SKEIN_UNROLL_1024 (0)
+#endif
+
+#if (SKEIN_UNROLL_1024 != 0)
+#if (RCNT % SKEIN_UNROLL_1024)
+#error "Invalid SKEIN_UNROLL_1024"              /* sanity check on unroll count */
+#endif
+    size_t  r;
+    u64b_t  kw[WCNT+4+RCNT*2];                  /* key schedule words : chaining vars + tweak + "rotation"*/
+#else
+    u64b_t  kw[WCNT+4];                         /* key schedule words : chaining vars + tweak */
+#endif
+
+    u64b_t  X00,X01,X02,X03,X04,X05,X06,X07,    /* local copy of vars, for speed */
+            X08,X09,X10,X11,X12,X13,X14,X15;
+    u64b_t  w [WCNT];                           /* local copy of input block */
+#ifdef SKEIN_DEBUG
+    const u64b_t *Xptr[16];                     /* use for debugging (help compiler put Xn in registers) */
+    Xptr[ 0] = &X00;  Xptr[ 1] = &X01;  Xptr[ 2] = &X02;  Xptr[ 3] = &X03;
+    Xptr[ 4] = &X04;  Xptr[ 5] = &X05;  Xptr[ 6] = &X06;  Xptr[ 7] = &X07;
+    Xptr[ 8] = &X08;  Xptr[ 9] = &X09;  Xptr[10] = &X10;  Xptr[11] = &X11;
+    Xptr[12] = &X12;  Xptr[13] = &X13;  Xptr[14] = &X14;  Xptr[15] = &X15;
+#endif
+
+    Skein_assert(blkCnt != 0);                  /* never call with blkCnt == 0! */
+    ts[0] = ctx->h.T[0];
+    ts[1] = ctx->h.T[1];
+    do  {
+        /* this implementation only supports 2**64 input bytes (no carry out here) */
+        ts[0] += byteCntAdd;                    /* update processed length */
+
+        /* precompute the key schedule for this block */
+        ks[ 0] = ctx->X[ 0];
+        ks[ 1] = ctx->X[ 1];
+        ks[ 2] = ctx->X[ 2];
+        ks[ 3] = ctx->X[ 3];
+        ks[ 4] = ctx->X[ 4];
+        ks[ 5] = ctx->X[ 5];
+        ks[ 6] = ctx->X[ 6];
+        ks[ 7] = ctx->X[ 7];
+        ks[ 8] = ctx->X[ 8];
+        ks[ 9] = ctx->X[ 9];
+        ks[10] = ctx->X[10];
+        ks[11] = ctx->X[11];
+        ks[12] = ctx->X[12];
+        ks[13] = ctx->X[13];
+        ks[14] = ctx->X[14];
+        ks[15] = ctx->X[15];
+        ks[16] = ks[ 0] ^ ks[ 1] ^ ks[ 2] ^ ks[ 3] ^
+                 ks[ 4] ^ ks[ 5] ^ ks[ 6] ^ ks[ 7] ^
+                 ks[ 8] ^ ks[ 9] ^ ks[10] ^ ks[11] ^
+                 ks[12] ^ ks[13] ^ ks[14] ^ ks[15] ^ SKEIN_KS_PARITY;
+
+        ts[2]  = ts[0] ^ ts[1];
+
+        Skein_Get64_LSB_First(w,blkPtr,WCNT); /* get input block in little-endian format */
+        DebugSaveTweak(ctx);
+        Skein_Show_Block(BLK_BITS,&ctx->h,ctx->X,blkPtr,w,ks,ts);
+
+        X00    = w[ 0] + ks[ 0];                 /* do the first full key injection */
+        X01    = w[ 1] + ks[ 1];
+        X02    = w[ 2] + ks[ 2];
+        X03    = w[ 3] + ks[ 3];
+        X04    = w[ 4] + ks[ 4];
+        X05    = w[ 5] + ks[ 5];
+        X06    = w[ 6] + ks[ 6];
+        X07    = w[ 7] + ks[ 7];
+        X08    = w[ 8] + ks[ 8];
+        X09    = w[ 9] + ks[ 9];
+        X10    = w[10] + ks[10];
+        X11    = w[11] + ks[11];
+        X12    = w[12] + ks[12];
+        X13    = w[13] + ks[13] + ts[0];
+        X14    = w[14] + ks[14] + ts[1];
+        X15    = w[15] + ks[15];
+
+        Skein_Show_R_Ptr(BLK_BITS,&ctx->h,SKEIN_RND_KEY_INITIAL,Xptr);
+
+#define Round1024(p0,p1,p2,p3,p4,p5,p6,p7,p8,p9,pA,pB,pC,pD,pE,pF,ROT,rNum) \
+    X##p0 += X##p1; X##p1 = RotL_64(X##p1,ROT##_0); X##p1 ^= X##p0;   \
+    X##p2 += X##p3; X##p3 = RotL_64(X##p3,ROT##_1); X##p3 ^= X##p2;   \
+    X##p4 += X##p5; X##p5 = RotL_64(X##p5,ROT##_2); X##p5 ^= X##p4;   \
+    X##p6 += X##p7; X##p7 = RotL_64(X##p7,ROT##_3); X##p7 ^= X##p6;   \
+    X##p8 += X##p9; X##p9 = RotL_64(X##p9,ROT##_4); X##p9 ^= X##p8;   \
+    X##pA += X##pB; X##pB = RotL_64(X##pB,ROT##_5); X##pB ^= X##pA;   \
+    X##pC += X##pD; X##pD = RotL_64(X##pD,ROT##_6); X##pD ^= X##pC;   \
+    X##pE += X##pF; X##pF = RotL_64(X##pF,ROT##_7); X##pF ^= X##pE;   \
+
+#if SKEIN_UNROLL_1024 == 0                      
+#define R1024(p0,p1,p2,p3,p4,p5,p6,p7,p8,p9,pA,pB,pC,pD,pE,pF,ROT,rn) \
+    Round1024(p0,p1,p2,p3,p4,p5,p6,p7,p8,p9,pA,pB,pC,pD,pE,pF,ROT,rn) \
+    Skein_Show_R_Ptr(BLK_BITS,&ctx->h,rn,Xptr);
+
+#define I1024(R)                                                      \
+    X00   += ks[((R)+ 1) % 17]; /* inject the key schedule value */   \
+    X01   += ks[((R)+ 2) % 17];                                       \
+    X02   += ks[((R)+ 3) % 17];                                       \
+    X03   += ks[((R)+ 4) % 17];                                       \
+    X04   += ks[((R)+ 5) % 17];                                       \
+    X05   += ks[((R)+ 6) % 17];                                       \
+    X06   += ks[((R)+ 7) % 17];                                       \
+    X07   += ks[((R)+ 8) % 17];                                       \
+    X08   += ks[((R)+ 9) % 17];                                       \
+    X09   += ks[((R)+10) % 17];                                       \
+    X10   += ks[((R)+11) % 17];                                       \
+    X11   += ks[((R)+12) % 17];                                       \
+    X12   += ks[((R)+13) % 17];                                       \
+    X13   += ks[((R)+14) % 17] + ts[((R)+1) % 3];                     \
+    X14   += ks[((R)+15) % 17] + ts[((R)+2) % 3];                     \
+    X15   += ks[((R)+16) % 17] +     (R)+1;                           \
+    Skein_Show_R_Ptr(BLK_BITS,&ctx->h,SKEIN_RND_KEY_INJECT,Xptr); 
+#else                                       /* looping version */
+#define R1024(p0,p1,p2,p3,p4,p5,p6,p7,p8,p9,pA,pB,pC,pD,pE,pF,ROT,rn) \
+    Round1024(p0,p1,p2,p3,p4,p5,p6,p7,p8,p9,pA,pB,pC,pD,pE,pF,ROT,rn) \
+    Skein_Show_R_Ptr(BLK_BITS,&ctx->h,4*(r-1)+rn,Xptr);
+
+#define I1024(R)                                                      \
+    X00   += ks[r+(R)+ 0];    /* inject the key schedule value */     \
+    X01   += ks[r+(R)+ 1];                                            \
+    X02   += ks[r+(R)+ 2];                                            \
+    X03   += ks[r+(R)+ 3];                                            \
+    X04   += ks[r+(R)+ 4];                                            \
+    X05   += ks[r+(R)+ 5];                                            \
+    X06   += ks[r+(R)+ 6];                                            \
+    X07   += ks[r+(R)+ 7];                                            \
+    X08   += ks[r+(R)+ 8];                                            \
+    X09   += ks[r+(R)+ 9];                                            \
+    X10   += ks[r+(R)+10];                                            \
+    X11   += ks[r+(R)+11];                                            \
+    X12   += ks[r+(R)+12];                                            \
+    X13   += ks[r+(R)+13] + ts[r+(R)+0];                              \
+    X14   += ks[r+(R)+14] + ts[r+(R)+1];                              \
+    X15   += ks[r+(R)+15] +    r+(R)   ;                              \
+    ks[r  +       (R)+16] = ks[r+(R)-1];  /* rotate key schedule */   \
+    ts[r  +       (R)+ 2] = ts[r+(R)-1];                              \
+    Skein_Show_R_Ptr(BLK_BITS,&ctx->h,SKEIN_RND_KEY_INJECT,Xptr);
+
+    for (r=1;r <= 2*RCNT;r+=2*SKEIN_UNROLL_1024)    /* loop thru it */
+#endif  
+        {
+#define R1024_8_rounds(R)    /* do 8 full rounds */                               \
+        R1024(00,01,02,03,04,05,06,07,08,09,10,11,12,13,14,15,R1024_0,8*(R) + 1); \
+        R1024(00,09,02,13,06,11,04,15,10,07,12,03,14,05,08,01,R1024_1,8*(R) + 2); \
+        R1024(00,07,02,05,04,03,06,01,12,15,14,13,08,11,10,09,R1024_2,8*(R) + 3); \
+        R1024(00,15,02,11,06,13,04,09,14,01,08,05,10,03,12,07,R1024_3,8*(R) + 4); \
+        I1024(2*(R));                                                             \
+        R1024(00,01,02,03,04,05,06,07,08,09,10,11,12,13,14,15,R1024_4,8*(R) + 5); \
+        R1024(00,09,02,13,06,11,04,15,10,07,12,03,14,05,08,01,R1024_5,8*(R) + 6); \
+        R1024(00,07,02,05,04,03,06,01,12,15,14,13,08,11,10,09,R1024_6,8*(R) + 7); \
+        R1024(00,15,02,11,06,13,04,09,14,01,08,05,10,03,12,07,R1024_7,8*(R) + 8); \
+        I1024(2*(R)+1);
+
+        R1024_8_rounds( 0);
+
+#define R1024_Unroll_R(NN) ((SKEIN_UNROLL_1024 == 0 && SKEIN1024_ROUNDS_TOTAL/8 > (NN)) || (SKEIN_UNROLL_1024 > (NN)))
+
+  #if   R1024_Unroll_R( 1)
+        R1024_8_rounds( 1);
+  #endif
+  #if   R1024_Unroll_R( 2)
+        R1024_8_rounds( 2);
+  #endif
+  #if   R1024_Unroll_R( 3)
+        R1024_8_rounds( 3);
+  #endif
+  #if   R1024_Unroll_R( 4)
+        R1024_8_rounds( 4);
+  #endif
+  #if   R1024_Unroll_R( 5)
+        R1024_8_rounds( 5);
+  #endif
+  #if   R1024_Unroll_R( 6)
+        R1024_8_rounds( 6);
+  #endif
+  #if   R1024_Unroll_R( 7)
+        R1024_8_rounds( 7);
+  #endif
+  #if   R1024_Unroll_R( 8)
+        R1024_8_rounds( 8);
+  #endif
+  #if   R1024_Unroll_R( 9)
+        R1024_8_rounds( 9);
+  #endif
+  #if   R1024_Unroll_R(10)
+        R1024_8_rounds(10);
+  #endif
+  #if   R1024_Unroll_R(11)
+        R1024_8_rounds(11);
+  #endif
+  #if   R1024_Unroll_R(12)
+        R1024_8_rounds(12);
+  #endif
+  #if   R1024_Unroll_R(13)
+        R1024_8_rounds(13);
+  #endif
+  #if   R1024_Unroll_R(14)
+        R1024_8_rounds(14);
+  #endif
+  #if  (SKEIN_UNROLL_1024 > 14)
+#error  "need more unrolling in Skein_1024_Process_Block"
+  #endif
+        }
+        /* do the final "feedforward" xor, update context chaining vars */
+
+        ctx->X[ 0] = X00 ^ w[ 0];
+        ctx->X[ 1] = X01 ^ w[ 1];
+        ctx->X[ 2] = X02 ^ w[ 2];
+        ctx->X[ 3] = X03 ^ w[ 3];
+        ctx->X[ 4] = X04 ^ w[ 4];
+        ctx->X[ 5] = X05 ^ w[ 5];
+        ctx->X[ 6] = X06 ^ w[ 6];
+        ctx->X[ 7] = X07 ^ w[ 7];
+        ctx->X[ 8] = X08 ^ w[ 8];
+        ctx->X[ 9] = X09 ^ w[ 9];
+        ctx->X[10] = X10 ^ w[10];
+        ctx->X[11] = X11 ^ w[11];
+        ctx->X[12] = X12 ^ w[12];
+        ctx->X[13] = X13 ^ w[13];
+        ctx->X[14] = X14 ^ w[14];
+        ctx->X[15] = X15 ^ w[15];
+
+        Skein_Show_Round(BLK_BITS,&ctx->h,SKEIN_RND_FEED_FWD,ctx->X);
+        
+        ts[1] &= ~SKEIN_T1_FLAG_FIRST;
+        blkPtr += SKEIN1024_BLOCK_BYTES;
+        }
+    while (--blkCnt);
+    ctx->h.T[0] = ts[0];
+    ctx->h.T[1] = ts[1];
+    }
+
+#if defined(SKEIN_CODE_SIZE) || defined(SKEIN_PERF)
+size_t Skein1024_Process_Block_CodeSize(void)
+    {
+    return ((u08b_t *) Skein1024_Process_Block_CodeSize) -
+           ((u08b_t *) Skein1024_Process_Block);
+    }
+uint_t Skein1024_Unroll_Cnt(void)
+    {
+    return SKEIN_UNROLL_1024;
+    }
+#endif
+#endif
diff --git a/sys/crypto/skein/skein_debug.c b/sys/crypto/skein/skein_debug.c
new file mode 100644
index 0000000..fac5038
--- /dev/null
+++ b/sys/crypto/skein/skein_debug.c
@@ -0,0 +1,247 @@
+/***********************************************************************
+**
+** Debug output functions for Skein hashing.
+**
+** Source code author: Doug Whiting, 2008.
+**
+** This algorithm and source code is released to the public domain.
+**
+************************************************************************/
+#include <stdio.h>
+
+#ifdef SKEIN_DEBUG  /* only instantiate this code if SKEIN_DEBUG is on */
+#include "skein.h"
+
+static const char INDENT[] =  "    ";  /* how much to indent on new line */
+
+uint_t skein_DebugFlag = 0;  /* off by default. Must be set externally */
+
+static void Show64_step(size_t cnt,const u64b_t *X,size_t step)
+    {
+    size_t i,j;
+    for (i=j=0;i < cnt;i++,j+=step)
+        {
+        if (i % 4 ==  0) printf(INDENT);
+        printf(" %08X.%08X ",(uint_32t)(X[j] >> 32),(uint_32t)X[j]);
+        if (i % 4 ==  3 || i==cnt-1) printf("\n");
+        fflush(stdout);
+        }
+    }
+
+#define Show64(cnt,X) Show64_step(cnt,X,1)
+
+static void Show64_flag(size_t cnt,const u64b_t *X)
+    {
+    size_t xptr = (size_t) X;
+    size_t step = (xptr & 1) ? 2 : 1;
+    if (step != 1)
+        {
+        X = (const u64b_t *) (xptr & ~1);
+        }
+    Show64_step(cnt,X,step);
+    }
+
+static void Show08(size_t cnt,const u08b_t *b)
+    {
+    size_t i;
+    for (i=0;i < cnt;i++)
+        {
+        if (i %16 ==  0) printf(INDENT);
+        else if (i % 4 == 0) printf(" ");
+        printf(" %02X",b[i]);
+        if (i %16 == 15 || i==cnt-1) printf("\n");
+        fflush(stdout);
+        }
+    }
+
+static const char *AlgoHeader(uint_t bits)
+    {
+    if (skein_DebugFlag & SKEIN_DEBUG_THREEFISH)
+        switch (bits)
+            {
+            case  256:  return ":Threefish-256: ";
+            case  512:  return ":Threefish-512: ";
+            case 1024:  return ":Threefish-1024:";
+            }
+    else
+        switch (bits)
+            {
+            case  256:  return ":Skein-256: ";
+            case  512:  return ":Skein-512: ";
+            case 1024:  return ":Skein-1024:";
+            }
+    return NULL;
+    }
+
+void Skein_Show_Final(uint_t bits,const Skein_Ctxt_Hdr_t *h,size_t cnt,const u08b_t *outPtr)
+    {
+    if (skein_DebugFlag & SKEIN_DEBUG_CONFIG || ((h->T[1] & SKEIN_T1_BLK_TYPE_MASK) != SKEIN_T1_BLK_TYPE_CFG))
+    if (skein_DebugFlag & SKEIN_DEBUG_FINAL)
+        {
+        printf("\n%s Final output=\n",AlgoHeader(bits));
+        Show08(cnt,outPtr);
+        printf("    ++++++++++\n");
+        fflush(stdout);
+        }
+    }
+
+/* show state after a round (or "pseudo-round") */
+void Skein_Show_Round(uint_t bits,const Skein_Ctxt_Hdr_t *h,size_t r,const u64b_t *X)
+    {
+    static uint_t injectNum=0;  /* not multi-thread safe! */
+
+    if (skein_DebugFlag & SKEIN_DEBUG_CONFIG || ((h->T[1] & SKEIN_T1_BLK_TYPE_MASK) != SKEIN_T1_BLK_TYPE_CFG))
+    if (skein_DebugFlag)
+        {
+        if (r >= SKEIN_RND_SPECIAL) 
+            {       /* a key injection (or feedforward) point */
+            injectNum = (r == SKEIN_RND_KEY_INITIAL) ? 0 : injectNum+1;
+            if (  skein_DebugFlag & SKEIN_DEBUG_INJECT ||
+                ((skein_DebugFlag & SKEIN_DEBUG_FINAL) && r == SKEIN_RND_FEED_FWD))
+                {
+                printf("\n%s",AlgoHeader(bits));
+                switch (r)
+                    {
+                    case SKEIN_RND_KEY_INITIAL:
+                        printf(" [state after initial key injection]");
+                        break;
+                    case SKEIN_RND_KEY_INJECT:
+                        printf(" [state after key injection #%02d]",injectNum);
+                        break;
+                    case SKEIN_RND_FEED_FWD:
+                        printf(" [state after plaintext feedforward]");
+                        injectNum = 0;
+                        break;
+                    }
+                printf("=\n");
+                Show64(bits/64,X);
+                if (r== SKEIN_RND_FEED_FWD)
+                    printf("    ----------\n");
+                }
+            }
+        else if (skein_DebugFlag & SKEIN_DEBUG_ROUNDS)
+            {
+            uint_t j;
+            u64b_t p[SKEIN_MAX_STATE_WORDS];
+            const u08b_t *perm;
+            const static u08b_t PERM_256 [4][ 4] = { { 0,1,2,3 }, { 0,3,2,1 }, { 0,1,2,3 }, { 0,3,2,1 } };
+            const static u08b_t PERM_512 [4][ 8] = { { 0,1,2,3,4,5,6,7 },
+                                                     { 2,1,4,7,6,5,0,3 },
+                                                     { 4,1,6,3,0,5,2,7 },
+                                                     { 6,1,0,7,2,5,4,3 }
+                                                   };
+            const static u08b_t PERM_1024[4][16] = { { 0, 1, 2, 3, 4, 5, 6, 7, 8, 9,10,11,12,13,14,15 },
+                                                     { 0, 9, 2,13, 6,11, 4,15,10, 7,12, 3,14, 5, 8, 1 },
+                                                     { 0, 7, 2, 5, 4, 3, 6, 1,12,15,14,13, 8,11,10, 9 },
+                                                     { 0,15, 2,11, 6,13, 4, 9,14, 1, 8, 5,10, 3,12, 7 }
+                                                   };
+                    
+            if ((skein_DebugFlag & SKEIN_DEBUG_PERMUTE) && (r & 3))
+                {
+                printf("\n%s [state after round %2d (permuted)]=\n",AlgoHeader(bits),(int)r);
+                switch (bits)
+                    {
+                    case  256: perm = PERM_256 [r&3];   break;
+                    case  512: perm = PERM_512 [r&3];   break;
+                    default:   perm = PERM_1024[r&3];   break;
+                    }
+                for (j=0;j<bits/64;j++)
+                    p[j] = X[perm[j]];
+                Show64(bits/64,p);
+                }
+            else
+                {
+                printf("\n%s [state after round %2d]=\n",AlgoHeader(bits),(int)r);
+                Show64(bits/64,X);
+                }
+            }
+        }
+    }
+
+/* show state after a round (or "pseudo-round"), given a list of pointers */
+void Skein_Show_R_Ptr(uint_t bits,const Skein_Ctxt_Hdr_t *h,size_t r,const u64b_t *X_ptr[])
+    {
+    uint_t i;
+    u64b_t X[SKEIN_MAX_STATE_WORDS];
+
+    for (i=0;i<bits/64;i++)     /* copy over the words */ 
+        X[i] = X_ptr[i][0];
+    Skein_Show_Round(bits,h,r,X);
+    }
+
+
+/* show the state at the start of a block */
+void Skein_Show_Block(uint_t bits,const Skein_Ctxt_Hdr_t *h,const u64b_t *X,const u08b_t *blkPtr,
+                      const u64b_t *wPtr, const u64b_t *ksPtr, const u64b_t *tsPtr)
+    {
+    uint_t n;
+    if (skein_DebugFlag & SKEIN_DEBUG_CONFIG || ((h->T[1] & SKEIN_T1_BLK_TYPE_MASK) != SKEIN_T1_BLK_TYPE_CFG))
+    if (skein_DebugFlag)
+        {
+        if (skein_DebugFlag & SKEIN_DEBUG_HDR)
+            {
+            printf("\n%s Block: outBits=%4d. T0=%06X.",AlgoHeader(bits),(uint_t) h->hashBitLen,(uint_t)h->T[0]);
+            printf(" Type=");
+            n = (uint_t) ((h->T[1] & SKEIN_T1_BLK_TYPE_MASK) >> SKEIN_T1_POS_BLK_TYPE);
+            switch (n)
+                {
+                case SKEIN_BLK_TYPE_KEY:  printf("KEY. ");  break;
+                case SKEIN_BLK_TYPE_CFG:  printf("CFG. ");  break;
+                case SKEIN_BLK_TYPE_PERS: printf("PERS.");  break;
+                case SKEIN_BLK_TYPE_PK :  printf("PK.  ");  break;
+                case SKEIN_BLK_TYPE_KDF:  printf("KDF. ");  break;
+                case SKEIN_BLK_TYPE_MSG:  printf("MSG. ");  break;
+                case SKEIN_BLK_TYPE_OUT:  printf("OUT. ");  break;
+                default:    printf("0x%02X.",n); break;
+                }
+            printf(" Flags=");
+            printf((h->T[1] & SKEIN_T1_FLAG_FIRST)   ? " First":"      ");
+            printf((h->T[1] & SKEIN_T1_FLAG_FINAL)   ? " Final":"      ");
+            printf((h->T[1] & SKEIN_T1_FLAG_BIT_PAD) ? " Pad"  :"    ");
+            n = (uint_t) ((h->T[1] & SKEIN_T1_TREE_LVL_MASK) >> SKEIN_T1_POS_TREE_LVL);
+            if (n)
+                printf("  TreeLevel = %02X",n);
+            printf("\n");
+            fflush(stdout);
+            }
+        if (skein_DebugFlag & SKEIN_DEBUG_TWEAK)
+            {
+            printf("  Tweak:\n");
+            Show64(2,h->T);
+            }
+        if (skein_DebugFlag & SKEIN_DEBUG_STATE)
+            {
+            printf("  %s words:\n",(skein_DebugFlag & SKEIN_DEBUG_THREEFISH)?"Key":"State");
+            Show64(bits/64,X);
+            }
+        if (skein_DebugFlag & SKEIN_DEBUG_KEYSCHED)
+            {
+            printf("  Tweak schedule:\n");
+            Show64_flag(3,tsPtr);
+            printf("  Key   schedule:\n");
+            Show64_flag((bits/64)+1,ksPtr);
+            }
+        if (skein_DebugFlag & SKEIN_DEBUG_INPUT_64)
+            {
+            printf("  Input block (words):\n");
+            Show64(bits/64,wPtr);
+            }
+        if (skein_DebugFlag & SKEIN_DEBUG_INPUT_08)
+            {
+            printf("  Input block (bytes):\n");
+            Show08(bits/8,blkPtr);
+            }
+        }
+    }
+
+void Skein_Show_Key(uint_t bits,const Skein_Ctxt_Hdr_t *h,const u08b_t *key,size_t keyBytes)
+    {
+    if (keyBytes)
+    if (skein_DebugFlag & SKEIN_DEBUG_CONFIG || ((h->T[1] & SKEIN_T1_BLK_TYPE_MASK) != SKEIN_T1_BLK_TYPE_CFG))
+    if (skein_DebugFlag & SKEIN_DEBUG_KEY)
+        {
+        printf("\n%s MAC key = %4u bytes\n",AlgoHeader(bits),(unsigned) keyBytes);
+        Show08(keyBytes,key);
+        }
+    }
+#endif
diff --git a/sys/crypto/skein/skein_debug.h b/sys/crypto/skein/skein_debug.h
new file mode 100644
index 0000000..7775c016
--- /dev/null
+++ b/sys/crypto/skein/skein_debug.h
@@ -0,0 +1,48 @@
+#ifndef _SKEIN_DEBUG_H_
+#define _SKEIN_DEBUG_H_
+/***********************************************************************
+**
+** Interface definitions for Skein hashing debug output.
+**
+** Source code author: Doug Whiting, 2008.
+**
+** This algorithm and source code is released to the public domain.
+**
+************************************************************************/
+
+#ifdef  SKEIN_DEBUG
+/* callout functions used inside Skein code */
+void    Skein_Show_Block(uint_t bits,const Skein_Ctxt_Hdr_t *h,const u64b_t *X,const u08b_t *blkPtr,
+                         const u64b_t *wPtr,const u64b_t *ksPtr,const u64b_t *tsPtr);
+void    Skein_Show_Round(uint_t bits,const Skein_Ctxt_Hdr_t *h,size_t r,const u64b_t *X);
+void    Skein_Show_R_Ptr(uint_t bits,const Skein_Ctxt_Hdr_t *h,size_t r,const u64b_t *X_ptr[]);
+void    Skein_Show_Final(uint_t bits,const Skein_Ctxt_Hdr_t *h,size_t cnt,const u08b_t *outPtr);
+void    Skein_Show_Key  (uint_t bits,const Skein_Ctxt_Hdr_t *h,const u08b_t *key,size_t keyBytes);
+
+extern  uint_t skein_DebugFlag;            /* flags to control debug output (0 --> none) */
+
+#define SKEIN_RND_SPECIAL       (1000u)
+#define SKEIN_RND_KEY_INITIAL   (SKEIN_RND_SPECIAL+0u)
+#define SKEIN_RND_KEY_INJECT    (SKEIN_RND_SPECIAL+1u)
+#define SKEIN_RND_FEED_FWD      (SKEIN_RND_SPECIAL+2u)
+
+/* flag bits:  skein_DebugFlag */
+#define SKEIN_DEBUG_KEY         (1u << 1)  /* show MAC key */
+#define SKEIN_DEBUG_CONFIG      (1u << 2)  /* show config block processing */
+#define SKEIN_DEBUG_STATE       (1u << 3)  /* show input state during Show_Block() */
+#define SKEIN_DEBUG_TWEAK       (1u << 4)  /* show input state during Show_Block() */
+#define SKEIN_DEBUG_KEYSCHED    (1u << 5)  /* show expanded key schedule */
+#define SKEIN_DEBUG_INPUT_64    (1u << 6)  /* show input block as 64-bit words */
+#define SKEIN_DEBUG_INPUT_08    (1u << 7)  /* show input block as  8-bit bytes */
+#define SKEIN_DEBUG_INJECT      (1u << 8)  /* show state after key injection & feedforward points */
+#define SKEIN_DEBUG_ROUNDS      (1u << 9)  /* show state after all rounds */
+#define SKEIN_DEBUG_FINAL       (1u <<10)  /* show final output of Skein */
+#define SKEIN_DEBUG_HDR         (1u <<11)  /* show block header */
+#define SKEIN_DEBUG_THREEFISH   (1u <<12)  /* use Threefish name instead of Skein */
+#define SKEIN_DEBUG_PERMUTE     (1u <<13)  /* use word permutations */
+#define SKEIN_DEBUG_ALL         ((~0u) & ~(SKEIN_DEBUG_THREEFISH | SKEIN_DEBUG_PERMUTE))
+#define THREEFISH_DEBUG_ALL     (SKEIN_DEBUG_ALL | SKEIN_DEBUG_THREEFISH)
+
+#endif /*  SKEIN_DEBUG    */
+
+#endif /* _SKEIN_DEBUG_H_ */
diff --git a/sys/crypto/skein/skein_freebsd.h b/sys/crypto/skein/skein_freebsd.h
new file mode 100644
index 0000000..935fa09
--- /dev/null
+++ b/sys/crypto/skein/skein_freebsd.h
@@ -0,0 +1,79 @@
+/*-
+ * Copyright 2016 Allan Jude
+ * All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. 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.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE AUTHOR 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 AUTHOR 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.
+ *
+ * $FreeBSD$
+ */
+
+#ifndef _SKEIN_FREEBSD_H_
+#define _SKEIN_FREEBSD_H_
+
+#define  SKEIN_256_BLOCK_BYTES ( 8*SKEIN_256_STATE_WORDS)
+#define  SKEIN_512_BLOCK_BYTES ( 8*SKEIN_512_STATE_WORDS)
+#define  SKEIN1024_BLOCK_BYTES ( 8*SKEIN1024_STATE_WORDS)
+
+#define SKEIN256_BLOCK_LENGTH		SKEIN_256_BLOCK_BYTES
+#define SKEIN256_DIGEST_LENGTH		32
+#define SKEIN256_DIGEST_STRING_LENGTH	(SKEIN256_DIGEST_LENGTH * 2 + 1)
+#define SKEIN512_BLOCK_LENGTH		SKEIN_512_BLOCK_BYTES
+#define SKEIN512_DIGEST_LENGTH		64
+#define SKEIN512_DIGEST_STRING_LENGTH	(SKEIN512_DIGEST_LENGTH * 2 + 1)
+#define SKEIN1024_BLOCK_LENGTH		SKEIN1024_BLOCK_BYTES
+#define SKEIN1024_DIGEST_LENGTH		128
+#define SKEIN1024_DIGEST_STRING_LENGTH	(SKEIN1024_DIGEST_LENGTH * 2 + 1)
+
+/* Make the context types look like the other hashes on FreeBSD */
+typedef Skein_256_Ctxt_t    SKEIN256_CTX;
+typedef Skein_512_Ctxt_t    SKEIN512_CTX;
+typedef Skein1024_Ctxt_t    SKEIN1024_CTX;
+
+/* Make the prototypes look like the other hashes */
+void SKEIN256_Init  (SKEIN256_CTX *ctx);
+void SKEIN512_Init  (SKEIN512_CTX *ctx);
+void SKEIN1024_Init  (SKEIN1024_CTX *ctx);
+
+void SKEIN256_Update(SKEIN256_CTX *ctx, const void *in, size_t len);
+void SKEIN512_Update(SKEIN512_CTX *ctx, const void *in, size_t len);
+void SKEIN1024_Update(SKEIN1024_CTX *ctx, const void *in, size_t len);
+
+void SKEIN256_Final(unsigned char digest[static SKEIN256_DIGEST_LENGTH], SKEIN256_CTX *ctx);
+void SKEIN512_Final(unsigned char digest[static SKEIN512_DIGEST_LENGTH], SKEIN512_CTX *ctx);
+void SKEIN1024_Final(unsigned char digest[static SKEIN1024_DIGEST_LENGTH], SKEIN1024_CTX *ctx);
+
+#ifndef _KERNEL
+char   *SKEIN256_End(SKEIN256_CTX *, char *);
+char   *SKEIN512_End(SKEIN512_CTX *, char *);
+char   *SKEIN1024_End(SKEIN1024_CTX *, char *);
+char   *SKEIN256_Data(const void *, unsigned int, char *);
+char   *SKEIN512_Data(const void *, unsigned int, char *);
+char   *SKEIN1024_Data(const void *, unsigned int, char *);
+char   *SKEIN256_File(const char *, char *);
+char   *SKEIN512_File(const char *, char *);
+char   *SKEIN1024_File(const char *, char *);
+char   *SKEIN256_FileChunk(const char *, char *, off_t, off_t);
+char   *SKEIN512_FileChunk(const char *, char *, off_t, off_t);
+char   *SKEIN1024_FileChunk(const char *, char *, off_t, off_t);
+#endif
+
+#endif	/* ifndef _SKEIN_FREEBSD_H_ */
diff --git a/sys/crypto/skein/skein_iv.h b/sys/crypto/skein/skein_iv.h
new file mode 100644
index 0000000..6e66166
--- /dev/null
+++ b/sys/crypto/skein/skein_iv.h
@@ -0,0 +1,200 @@
+/*	$FreeBSD$	*/
+#ifndef _SKEIN_IV_H_
+#define _SKEIN_IV_H_
+
+#include "skein.h"    /* get Skein macros and types */
+
+/*
+***************** Pre-computed Skein IVs *******************
+**
+** NOTE: these values are not "magic" constants, but
+** are generated using the Threefish block function.
+** They are pre-computed here only for speed; i.e., to
+** avoid the need for a Threefish call during Init().
+**
+** The IV for any fixed hash length may be pre-computed.
+** Only the most common values are included here.
+**
+************************************************************
+**/
+
+#define MK_64 SKEIN_MK_64
+
+/* blkSize =  256 bits. hashSize =  128 bits */
+const u64b_t SKEIN_256_IV_128[] =
+    {
+    MK_64(0xE1111906,0x964D7260),
+    MK_64(0x883DAAA7,0x7C8D811C),
+    MK_64(0x10080DF4,0x91960F7A),
+    MK_64(0xCCF7DDE5,0xB45BC1C2)
+    };
+
+/* blkSize =  256 bits. hashSize =  160 bits */
+const u64b_t SKEIN_256_IV_160[] =
+    {
+    MK_64(0x14202314,0x72825E98),
+    MK_64(0x2AC4E9A2,0x5A77E590),
+    MK_64(0xD47A5856,0x8838D63E),
+    MK_64(0x2DD2E496,0x8586AB7D)
+    };
+
+/* blkSize =  256 bits. hashSize =  224 bits */
+const u64b_t SKEIN_256_IV_224[] =
+    {
+    MK_64(0xC6098A8C,0x9AE5EA0B),
+    MK_64(0x876D5686,0x08C5191C),
+    MK_64(0x99CB88D7,0xD7F53884),
+    MK_64(0x384BDDB1,0xAEDDB5DE)
+    };
+
+/* blkSize =  256 bits. hashSize =  256 bits */
+const u64b_t SKEIN_256_IV_256[] =
+    {
+    MK_64(0xFC9DA860,0xD048B449),
+    MK_64(0x2FCA6647,0x9FA7D833),
+    MK_64(0xB33BC389,0x6656840F),
+    MK_64(0x6A54E920,0xFDE8DA69)
+    };
+
+/* blkSize =  512 bits. hashSize =  128 bits */
+const u64b_t SKEIN_512_IV_128[] =
+    {
+    MK_64(0xA8BC7BF3,0x6FBF9F52),
+    MK_64(0x1E9872CE,0xBD1AF0AA),
+    MK_64(0x309B1790,0xB32190D3),
+    MK_64(0xBCFBB854,0x3F94805C),
+    MK_64(0x0DA61BCD,0x6E31B11B),
+    MK_64(0x1A18EBEA,0xD46A32E3),
+    MK_64(0xA2CC5B18,0xCE84AA82),
+    MK_64(0x6982AB28,0x9D46982D)
+    };
+
+/* blkSize =  512 bits. hashSize =  160 bits */
+const u64b_t SKEIN_512_IV_160[] =
+    {
+    MK_64(0x28B81A2A,0xE013BD91),
+    MK_64(0xC2F11668,0xB5BDF78F),
+    MK_64(0x1760D8F3,0xF6A56F12),
+    MK_64(0x4FB74758,0x8239904F),
+    MK_64(0x21EDE07F,0x7EAF5056),
+    MK_64(0xD908922E,0x63ED70B8),
+    MK_64(0xB8EC76FF,0xECCB52FA),
+    MK_64(0x01A47BB8,0xA3F27A6E)
+    };
+
+/* blkSize =  512 bits. hashSize =  224 bits */
+const u64b_t SKEIN_512_IV_224[] =
+    {
+    MK_64(0xCCD06162,0x48677224),
+    MK_64(0xCBA65CF3,0xA92339EF),
+    MK_64(0x8CCD69D6,0x52FF4B64),
+    MK_64(0x398AED7B,0x3AB890B4),
+    MK_64(0x0F59D1B1,0x457D2BD0),
+    MK_64(0x6776FE65,0x75D4EB3D),
+    MK_64(0x99FBC70E,0x997413E9),
+    MK_64(0x9E2CFCCF,0xE1C41EF7)
+    };
+
+/* blkSize =  512 bits. hashSize =  256 bits */
+const u64b_t SKEIN_512_IV_256[] =
+    {
+    MK_64(0xCCD044A1,0x2FDB3E13),
+    MK_64(0xE8359030,0x1A79A9EB),
+    MK_64(0x55AEA061,0x4F816E6F),
+    MK_64(0x2A2767A4,0xAE9B94DB),
+    MK_64(0xEC06025E,0x74DD7683),
+    MK_64(0xE7A436CD,0xC4746251),
+    MK_64(0xC36FBAF9,0x393AD185),
+    MK_64(0x3EEDBA18,0x33EDFC13)
+    };
+
+/* blkSize =  512 bits. hashSize =  384 bits */
+const u64b_t SKEIN_512_IV_384[] =
+    {
+    MK_64(0xA3F6C6BF,0x3A75EF5F),
+    MK_64(0xB0FEF9CC,0xFD84FAA4),
+    MK_64(0x9D77DD66,0x3D770CFE),
+    MK_64(0xD798CBF3,0xB468FDDA),
+    MK_64(0x1BC4A666,0x8A0E4465),
+    MK_64(0x7ED7D434,0xE5807407),
+    MK_64(0x548FC1AC,0xD4EC44D6),
+    MK_64(0x266E1754,0x6AA18FF8)
+    };
+
+/* blkSize =  512 bits. hashSize =  512 bits */
+const u64b_t SKEIN_512_IV_512[] =
+    {
+    MK_64(0x4903ADFF,0x749C51CE),
+    MK_64(0x0D95DE39,0x9746DF03),
+    MK_64(0x8FD19341,0x27C79BCE),
+    MK_64(0x9A255629,0xFF352CB1),
+    MK_64(0x5DB62599,0xDF6CA7B0),
+    MK_64(0xEABE394C,0xA9D5C3F4),
+    MK_64(0x991112C7,0x1A75B523),
+    MK_64(0xAE18A40B,0x660FCC33)
+    };
+
+/* blkSize = 1024 bits. hashSize =  384 bits */
+const u64b_t SKEIN1024_IV_384[] =
+    {
+    MK_64(0x5102B6B8,0xC1894A35),
+    MK_64(0xFEEBC9E3,0xFE8AF11A),
+    MK_64(0x0C807F06,0xE32BED71),
+    MK_64(0x60C13A52,0xB41A91F6),
+    MK_64(0x9716D35D,0xD4917C38),
+    MK_64(0xE780DF12,0x6FD31D3A),
+    MK_64(0x797846B6,0xC898303A),
+    MK_64(0xB172C2A8,0xB3572A3B),
+    MK_64(0xC9BC8203,0xA6104A6C),
+    MK_64(0x65909338,0xD75624F4),
+    MK_64(0x94BCC568,0x4B3F81A0),
+    MK_64(0x3EBBF51E,0x10ECFD46),
+    MK_64(0x2DF50F0B,0xEEB08542),
+    MK_64(0x3B5A6530,0x0DBC6516),
+    MK_64(0x484B9CD2,0x167BBCE1),
+    MK_64(0x2D136947,0xD4CBAFEA)
+    };
+
+/* blkSize = 1024 bits. hashSize =  512 bits */
+const u64b_t SKEIN1024_IV_512[] =
+    {
+    MK_64(0xCAEC0E5D,0x7C1B1B18),
+    MK_64(0xA01B0E04,0x5F03E802),
+    MK_64(0x33840451,0xED912885),
+    MK_64(0x374AFB04,0xEAEC2E1C),
+    MK_64(0xDF25A0E2,0x813581F7),
+    MK_64(0xE4004093,0x8B12F9D2),
+    MK_64(0xA662D539,0xC2ED39B6),
+    MK_64(0xFA8B85CF,0x45D8C75A),
+    MK_64(0x8316ED8E,0x29EDE796),
+    MK_64(0x053289C0,0x2E9F91B8),
+    MK_64(0xC3F8EF1D,0x6D518B73),
+    MK_64(0xBDCEC3C4,0xD5EF332E),
+    MK_64(0x549A7E52,0x22974487),
+    MK_64(0x67070872,0x5B749816),
+    MK_64(0xB9CD28FB,0xF0581BD1),
+    MK_64(0x0E2940B8,0x15804974)
+    };
+
+/* blkSize = 1024 bits. hashSize = 1024 bits */
+const u64b_t SKEIN1024_IV_1024[] =
+    {
+    MK_64(0xD593DA07,0x41E72355),
+    MK_64(0x15B5E511,0xAC73E00C),
+    MK_64(0x5180E5AE,0xBAF2C4F0),
+    MK_64(0x03BD41D3,0xFCBCAFAF),
+    MK_64(0x1CAEC6FD,0x1983A898),
+    MK_64(0x6E510B8B,0xCDD0589F),
+    MK_64(0x77E2BDFD,0xC6394ADA),
+    MK_64(0xC11E1DB5,0x24DCB0A3),
+    MK_64(0xD6D14AF9,0xC6329AB5),
+    MK_64(0x6A9B0BFC,0x6EB67E0D),
+    MK_64(0x9243C60D,0xCCFF1332),
+    MK_64(0x1A1F1DDE,0x743F02D4),
+    MK_64(0x0996753C,0x10ED0BB8),
+    MK_64(0x6572DD22,0xF2B4969A),
+    MK_64(0x61FD3062,0xD00A579A),
+    MK_64(0x1DE0536E,0x8682E539)
+    };
+
+#endif /* _SKEIN_IV_H_ */
diff --git a/sys/crypto/skein/skein_port.h b/sys/crypto/skein/skein_port.h
new file mode 100644
index 0000000..65f3405
--- /dev/null
+++ b/sys/crypto/skein/skein_port.h
@@ -0,0 +1,156 @@
+/*	$FreeBSD$	*/
+#ifndef _SKEIN_PORT_H_
+#define _SKEIN_PORT_H_
+/*******************************************************************
+**
+** Platform-specific definitions for Skein hash function.
+**
+** Source code author: Doug Whiting, 2008.
+**
+** This algorithm and source code is released to the public domain.
+**
+** Many thanks to Brian Gladman for his portable header files.
+**
+** To port Skein to an "unsupported" platform, change the definitions
+** in this file appropriately.
+** 
+********************************************************************/
+
+#include <sys/endian.h>
+#include <sys/types.h>
+
+typedef unsigned int    uint_t;             /* native unsigned integer */
+typedef u_int8_t        u08b_t;             /*  8-bit unsigned integer */
+typedef u_int32_t       uint_32t;           /* 32-bit unsigned integer */
+typedef u_int64_t       u64b_t;             /* 64-bit unsigned integer */
+
+#ifndef RotL_64
+#define RotL_64(x,N)    (((x) << (N)) | ((x) >> (64-(N))))
+#endif
+
+__BEGIN_DECLS
+
+/*
+ * Skein is "natively" little-endian (unlike SHA-xxx), for optimal
+ * performance on x86 CPUs.  The Skein code requires the following
+ * definitions for dealing with endianness:
+ *
+ *    SKEIN_NEED_SWAP:  0 for little-endian, 1 for big-endian
+ *    Skein_Put64_LSB_First
+ *    Skein_Get64_LSB_First
+ *    Skein_Swap64
+ *
+ * If SKEIN_NEED_SWAP is defined at compile time, it is used here
+ * along with the portable versions of Put64/Get64/Swap64, which 
+ * are slow in general.
+ *
+ * Otherwise, an "auto-detect" of endianness is attempted below.
+ * If the default handling doesn't work well, the user may insert
+ * platform-specific code instead (e.g., for big-endian CPUs).
+ *
+ */
+#ifndef SKEIN_NEED_SWAP /* compile-time "override" for endianness? */
+
+#if BYTE_ORDER == BIG_ENDIAN
+    /* here for big-endian CPUs */
+#define SKEIN_NEED_SWAP   (1)
+#ifdef  SKEIN_PORT_CODE
+void    Skein_Put64_LSB_First(u08b_t *dst,const u64b_t *src,size_t bCnt);
+void    Skein_Get64_LSB_First(u64b_t *dst,const u08b_t *src,size_t wCnt);
+#endif /* ifdef SKEIN_PORT_CODE */
+#elif BYTE_ORDER == LITTLE_ENDIAN
+    /* here for x86 and x86-64 CPUs (and other detected little-endian CPUs) */
+#define SKEIN_NEED_SWAP   (0)
+#define Skein_Put64_LSB_First(dst08,src64,bCnt) memcpy(dst08,src64,bCnt)
+#define Skein_Get64_LSB_First(dst64,src08,wCnt) memcpy(dst64,src08,8*(wCnt))
+#else
+#error "Skein needs endianness setting!"
+#endif
+
+#endif /* ifndef SKEIN_NEED_SWAP */
+
+/*
+ ******************************************************************
+ *      Provide any definitions still needed.
+ ******************************************************************
+ */
+#ifndef Skein_Swap64  /* swap for big-endian, nop for little-endian */
+#if     SKEIN_NEED_SWAP
+#define Skein_Swap64(w64)  bswap64(w64)
+#else
+#define Skein_Swap64(w64)  (w64)
+#endif
+#endif  /* ifndef Skein_Swap64 */
+
+
+#ifndef Skein_Put64_LSB_First
+void    Skein_Put64_LSB_First(u08b_t *dst,const u64b_t *src,size_t bCnt)
+#ifdef  SKEIN_PORT_CODE /* instantiate the function code here? */
+{
+    size_t n;
+
+    for (n = 0; n < bCnt / 8; n++)
+        le64enc(dst + n * 8, src[n]);
+}
+#else
+; /* output only the function prototype */
+#endif
+#endif   /* ifndef Skein_Put64_LSB_First */
+
+
+#ifndef Skein_Get64_LSB_First
+void    Skein_Get64_LSB_First(u64b_t *dst,const u08b_t *src,size_t wCnt)
+#ifdef  SKEIN_PORT_CODE /* instantiate the function code here? */
+{
+    size_t n;
+
+    for (n = 0; n < wCnt; n++)
+        dst[n] = le64dec(src + n * 8);
+}
+#else
+; /* output only the function prototype */
+#endif
+#endif   /* ifndef Skein_Get64_LSB_First */
+
+/* Start FreeBSD libmd shims */
+
+/* Ensure libmd symbols do not clash with libcrypto */
+#ifndef SKEIN256_Init
+#define SKEIN256_Init		_libmd_SKEIN256_Init
+#define SKEIN512_Init		_libmd_SKEIN512_Init
+#define SKEIN1024_Init		_libmd_SKEIN1024_Init
+#endif
+#ifndef SKEIN256_Update
+#define SKEIN256_Update		_libmd_SKEIN256_Update
+#define SKEIN512_Update		_libmd_SKEIN512_Update
+#define SKEIN1024_Update	_libmd_SKEIN1024_Update
+#endif
+#ifndef SKEIN256_Final
+#define SKEIN256_Final		_libmd_SKEIN256_Final
+#define SKEIN512_Final		_libmd_SKEIN512_Final
+#define SKEIN1024_Final		_libmd_SKEIN1024_Final
+#endif
+#ifndef SKEIN256_End
+#define SKEIN256_End		_libmd_SKEIN256_End
+#define SKEIN512_End		_libmd_SKEIN512_End
+#define SKEIN1024_End		_libmd_SKEIN1024_End
+#endif
+#ifndef SKEIN256_File
+#define SKEIN256_File		_libmd_SKEIN256_File
+#define SKEIN512_File		_libmd_SKEIN512_File
+#define SKEIN1024_File		_libmd_SKEIN1024_File
+#endif
+#ifndef SKEIN256_FileChunk
+#define SKEIN256_FileChunk	_libmd_SKEIN256_FileChunk
+#define SKEIN512_FileChunk	_libmd_SKEIN512_FileChunk
+#define SKEIN1024_FileChunk	_libmd_SKEIN1024_FileChunk
+#endif
+#ifndef SKEIN256_Data
+#define SKEIN256_Data		_libmd_SKEIN256_Data
+#define SKEIN512_Data		_libmd_SKEIN512_Data
+#define SKEIN1024_Data		_libmd_SKEIN1024_Data
+#endif
+
+__END_DECLS
+
+#endif   /* ifndef _SKEIN_PORT_H_ */
diff --git a/sys/modules/crypto/Makefile b/sys/modules/crypto/Makefile
index b476571..35e533b 100644
--- a/sys/modules/crypto/Makefile
+++ b/sys/modules/crypto/Makefile
@@ -8,6 +8,7 @@
 .PATH:	${.CURDIR}/../../crypto/rijndael
 .PATH:	${.CURDIR}/../../crypto/sha2
 .PATH:	${.CURDIR}/../../crypto/siphash
+.PATH:	${.CURDIR}/../../crypto/skein
 
 KMOD	= crypto
 SRCS	= crypto.c cryptodev_if.c
@@ -17,6 +18,13 @@ SRCS	+= skipjack.c bf_enc.c bf_ecb.c bf_skey.c
 SRCS	+= camellia.c camellia-api.c
 SRCS	+= des_ecb.c des_enc.c des_setkey.c
 SRCS	+= sha1.c sha256c.c sha512c.c
+SRCS	+= skein.c skein_block.c
+.if exists(${MACHINE_ARCH}/skein_block_asm.s)
+.PATH:	${.CURDIR}/../../crypto/skein/${MACHINE_ARCH}
+SRCS	+= skein_block_asm.s
+CFLAGS	+= -DSKEIN_ASM -DSKEIN_USE_ASM=1792 # list of block functions to replace with assembly: 256+512+1024 = 1792
+ACFLAGS	+= -DELF -Wa,--noexecstack
+.endif
 SRCS	+= siphash.c
 SRCS	+= gmac.c gfmult.c
 SRCS	+= opt_param.h cryptodev_if.h bus_if.h device_if.h