Xoodoo-u1-armv6m-le-gcc.s

@
@ The eXtended Keccak Code Package (XKCP)
@ https://github.com/XKCP/XKCP
@
@ The Xoodoo permutation, designed by Joan Daemen, Seth Hoffert, Gilles Van Assche and Ronny Van Keer.
@
@ Implementation by Ronny Van Keer, hereby denoted as "the implementer".
@
@ For more information, feedback or questions, please refer to the Keccak Team website:
@ https://keccak.team/
@
@ To the extent possible under law, the implementer has waived all copyright
@ and related or neighboring rights to the source code in this file.
@ http://creativecommons.org/publicdomain/zero/1.0/
@

@ WARNING: These functions work only on little endian CPU with@ ARMv6m architecture (Cortex-M0, ...).


        .thumb
    .syntax unified
.text

@ ----------------------------------------------------------------------------
@
@  void Xoodoo_Initialize(void *state)
@
    .align  4
.global Xoodoo_Initialize
.type	Xoodoo_Initialize, %function;
Xoodoo_Initialize:
    movs    r1, #0
    movs    r2, #0
    movs    r3, #0
    stmia   r0!, { r1 - r3 }
    stmia   r0!, { r1 - r3 }
    stmia   r0!, { r1 - r3 }
    stmia   r0!, { r1 - r3 }
    bx      lr
    .align  4


@ ----------------------------------------------------------------------------
@
@  void Xoodoo_AddBytes(void *state, const unsigned char *data, unsigned int offset, unsigned int length)
@
.global Xoodoo_AddBytes
.type	Xoodoo_AddBytes, %function;
Xoodoo_AddBytes:
    push    {r4,lr}
    adds    r0, r0, r2                              @ state += offset
    subs    r3, r3, #4                              @ .if length >= 4
    bcc     Xoodoo_AddBytes_Bytes
    movs    r2, r0                                  @ and data pointer and offset both 32-bit    .align 8ed
    orrs    r2, r2, r1
    lsls    r2, #30
    bne     Xoodoo_AddBytes_Bytes
Xoodoo_AddBytes_LanesLoop:                           @ then, perform on words
    ldr     r2, [r0]
    ldmia   r1!, {r4}
    eors    r2, r2, r4
    stmia   r0!, {r2}
    subs    r3, r3, #4
    bcs     Xoodoo_AddBytes_LanesLoop
Xoodoo_AddBytes_Bytes:
    adds    r3, r3, #4
    beq     Xoodoo_AddBytes_Exit
    subs    r3, r3, #1
Xoodoo_AddBytes_BytesLoop:
    ldrb    r2, [r0, r3]
    ldrb    r4, [r1, r3]
    eors    r2, r2, r4
    strb    r2, [r0, r3]
    subs    r3, r3, #1
    bcs     Xoodoo_AddBytes_BytesLoop
Xoodoo_AddBytes_Exit:
    pop     {r4,pc}
    .align  4


@ ----------------------------------------------------------------------------
@
@  void Xoodoo_OverwriteBytes(void *state, const unsigned char *data, unsigned int offset, unsigned int length)
@
.global Xoodoo_OverwriteBytes
.type	Xoodoo_OverwriteBytes, %function;
Xoodoo_OverwriteBytes:
    adds    r0, r0, r2                              @ state += offset
    subs    r3, r3, #4                              @ .if length >= 4
    bcc     Xoodoo_OverwriteBytes_Bytes
    movs    r2, r0                                  @ and data pointer and offset both 32-bit    .align 8ed
    orrs    r2, r2, r1
    lsls    r2, #30
    bne     Xoodoo_OverwriteBytes_Bytes
Xoodoo_OverwriteBytes_LanesLoop:                 @ then, perform on words
    ldmia   r1!, {r2}
    stmia   r0!, {r2}
    subs    r3, r3, #4
    bcs     Xoodoo_OverwriteBytes_LanesLoop
Xoodoo_OverwriteBytes_Bytes:
    adds    r3, r3, #4
    beq     Xoodoo_OverwriteBytes_Exit
    subs    r3, r3, #1
Xoodoo_OverwriteBytes_BytesLoop:
    ldrb    r2, [r1, r3]
    strb    r2, [r0, r3]
    subs    r3, r3, #1
    bcs     Xoodoo_OverwriteBytes_BytesLoop
Xoodoo_OverwriteBytes_Exit:
    bx      lr
    .align  4


@ ----------------------------------------------------------------------------
@
@  void Xoodoo_OverwriteWithZeroes(void *state, unsigned int byteCount)
@
.global Xoodoo_OverwriteWithZeroes
.type	Xoodoo_OverwriteWithZeroes, %function;
Xoodoo_OverwriteWithZeroes:
    movs    r3, #0
    lsrs    r2, r1, #2
    beq     Xoodoo_OverwriteWithZeroes_Bytes
Xoodoo_OverwriteWithZeroes_LoopLanes:
    stm     r0!, { r3 }
    subs    r2, r2, #1
    bne     Xoodoo_OverwriteWithZeroes_LoopLanes
Xoodoo_OverwriteWithZeroes_Bytes:
    lsls    r1, r1, #32-2
    beq     Xoodoo_OverwriteWithZeroes_Exit
    lsrs    r1, r1, #32-2
Xoodoo_OverwriteWithZeroes_LoopBytes:
    subs    r1, r1, #1
    strb    r3, [r0, r1]
    bne     Xoodoo_OverwriteWithZeroes_LoopBytes
Xoodoo_OverwriteWithZeroes_Exit:
    bx      lr
    .align  4


@ ----------------------------------------------------------------------------
@
@  void Xoodoo_ExtractBytes(void *state, const unsigned char *data, unsigned int offset, unsigned int length)
@
.global Xoodoo_ExtractBytes
.type	Xoodoo_ExtractBytes, %function;
Xoodoo_ExtractBytes:
    adds    r0, r0, r2                              @ state += offset
    subs    r3, r3, #4                              @ .if length >= 4
    bcc     Xoodoo_ExtractBytes_Bytes
    movs    r2, r0                                  @ and data pointer and offset both 32-bit    .align 8ed
    orrs    r2, r2, r1
    lsls    r2, #30
    bne     Xoodoo_ExtractBytes_Bytes
Xoodoo_ExtractBytes_LanesLoop:                       @ then, perform on words
    ldmia   r0!, {r2}
    stmia   r1!, {r2}
    subs    r3, r3, #4
    bcs     Xoodoo_ExtractBytes_LanesLoop
Xoodoo_ExtractBytes_Bytes:
    adds    r3, r3, #4
    beq     Xoodoo_ExtractBytes_Exit
    subs    r3, r3, #1
Xoodoo_ExtractBytes_BytesLoop:
    ldrb    r2, [r0, r3]
    strb    r2, [r1, r3]
    subs    r3, r3, #1
    bcs     Xoodoo_ExtractBytes_BytesLoop
Xoodoo_ExtractBytes_Exit:
    bx      lr
    .align  4


@ ----------------------------------------------------------------------------
@
@  void Xoodoo_ExtractAndAddBytes(void *state, const unsigned char *input, unsigned char *output, unsigned int offset, unsigned int length)
@
.global Xoodoo_ExtractAndAddBytes
.type	Xoodoo_ExtractAndAddBytes, %function;
Xoodoo_ExtractAndAddBytes:
    push    {r4,r5}
    adds    r0, r0, r3                              @ state += offset (offset register no longer needed, reuse for length)
    ldr     r3, [sp, #8]                            @ get length argument from stack
    subs    r3, r3, #4                              @ .if length >= 4
    bcc     Xoodoo_ExtractAndAddBytes_Bytes
    movs    r5, r0                                  @ and input/output/state pointer all 32-bit    .align 8ed
    orrs    r5, r5, r1
    orrs    r5, r5, r2
    lsls    r5, #30
    bne     Xoodoo_ExtractAndAddBytes_Bytes
Xoodoo_ExtractAndAddBytes_LanesLoop:                 @ then, perform on words
    ldmia   r0!, {r5}
    ldmia   r1!, {r4}
    eors    r5, r5, r4
    stmia   r2!, {r5}
    subs    r3, r3, #4
    bcs     Xoodoo_ExtractAndAddBytes_LanesLoop
Xoodoo_ExtractAndAddBytes_Bytes:
    adds    r3, r3, #4
    beq     Xoodoo_ExtractAndAddBytes_Exit
    subs    r3, r3, #1
Xoodoo_ExtractAndAddBytes_BytesLoop:
    ldrb    r5, [r0, r3]
    ldrb    r4, [r1, r3]
    eors    r5, r5, r4
    strb    r5, [r2, r3]
    subs    r3, r3, #1
    bcs     Xoodoo_ExtractAndAddBytes_BytesLoop
Xoodoo_ExtractAndAddBytes_Exit:
    pop     {r4,r5}
    bx      lr
    .align  4


@ ----------------------------------------------------------------------------

@ offsets in RAM state
.equ _oA00      , 0*4
.equ _oA01      , 1*4
.equ _oA02      , 2*4
.equ _oA03      , 3*4
.equ _oA10      , 4*4
.equ _oA11      , 5*4
.equ _oA12      , 6*4
.equ _oA13      , 7*4
.equ _oA20      , 8*4
.equ _oA21      , 9*4
.equ _oA22      , 10*4
.equ _oA23      , 11*4

@ possible locations of state lanes
.equ locRegL    , 1
.equ locRegH    , 2
.equ locMem     , 3

@ ----------------------------------------------------------------------------

.equ _r0   , 5
.equ _r1   , 14
.equ _r2   , 1

.equ _w1   , 11

.equ _e0   , 2
.equ _e1   , 8

@ ----------------------------------------------------------------------------

.macro    mXor3   ro, a0, a1, a2, loc, tt
    mov     \ro, \a1
    eors    \ro, \ro, \a2
    .if \loc == locRegL
    eors    \ro, \ro, \a0
    .else
    .if \loc == locRegH
    mov     \tt, \a0
    .else
    ldr     \tt, [sp, #\a0]
    .endif
    eors    \ro, \ro, \tt
    .endif
    .endm

.macro    mXor    ro, ri, tt, loc
    .if \loc == locRegL
    eors    \ro, \ro, \ri
    .else
    .if \loc == locRegH
    mov     \tt, \ro
    eors    \tt, \tt, \ri
    mov     \ro, \tt
    .else
    ldr     \tt, [sp, #\ro]
    eors    \tt, \tt, \ri
    str     \tt, [sp, #\ro]
    .endif
    .endif
    .endm

.macro    mChi3   a0,a1,a2,r0,r1,a0s,loc
    mov     \r1, \a2
    mov     \r0, \a1
    bics    \r1, \r1, \r0
    eors    \a0, \a0, \r1
    .if \loc != locRegL
    .if \loc == locRegH
    mov     \a0s, \a0
    .else
    str     \a0, [sp, #\a0s]
    .endif
    .endif

    mov     \r0, \a0
    bics    \r0, \r0, \a2
    mov     \r1, \a1
    eors    \r1, \r1, \r0
    mov     \a1, \r1

    bics    \r1, \r1, \a0
    eors    \a2, \a2, \r1
    .endm

.macro    mRound  offsetRC, offsetA03

    @ Theta: Column Parity Mixer
    mXor3   r0, \offsetA03, lr, r7, locMem, r2
    mov     r1, r0
    movs    r2, #32-(_r1-_r0)
    rors    r1, r1, r2
    eors    r1, r1, r0
    movs    r2, #32-_r0
    rors    r1, r1, r2
    mXor3   r0, r3, r10, r4, locRegL, r2
    mXor    r3, r1, r2, locRegL
    mXor    r10, r1, r2, locRegH
    mXor    r4, r1, r2, locRegL

    mov     r1, r0
    movs    r2, #32-(_r1-_r0)
    rors    r1, r1, r2
    eors    r1, r1, r0
    movs    r2, #32-_r0
    rors    r1, r1, r2
    mXor3   r0, r8, r11, r5, locRegH, r2
    mXor    r8, r1, r2, locRegH
    mXor    r11, r1, r2, locRegH
    mXor    r5, r1, r2, locRegL

    mov     r1, r0
    movs    r2, #32-(_r1-_r0)
    rors    r1, r1, r2
    eors    r1, r1, r0
    movs    r2, #32-_r0
    rors    r1, r1, r2
    mXor3   r0, r9, r12, r6, locRegH, r2
    mXor    r9, r1, r2, locRegH
    mXor    r12, r1, r2, locRegH
    mXor    r6, r1, r2, locRegL

    mov     r1, r0
    movs    r2, #32-(_r1-_r0)
    rors    r1, r1, r2
    eors    r1, r1, r0
    movs    r2, #32-_r0
    rors    r1, r1, r2
    mXor    \offsetA03, r1, r2, locMem
    mXor    lr, r1, r2, locRegH
    mXor    r7, r1, r2, locRegL

    @ Rho-west: Plane shift
    movs    r0, #32-_w1
    rors    r4, r4, r0
    rors    r5, r5, r0
    rors    r6, r6, r0
    rors    r7, r7, r0
    mov     r0, lr
    mov     lr, r12
    mov     r12, r11
    mov     r11, r10
    mov     r10, r0

    @ Iota: round constant
    ldr     r0, [sp, #\offsetRC]
    ldmia   r0!, {r1}
    str     r0, [sp, #\offsetRC]
    eors    r3, r3, r1

    @ Chi: non linear step, on colums
    mChi3   r3, r10, r4, r0, r1, r3, locRegL
    mov     r2, r8
    mChi3   r2, r11, r5, r0, r1, r8, locRegH
    mov     r2, r9
    mChi3   r2, r12, r6, r0, r1, r9, locRegH
    ldr     r2, [sp, #\offsetA03]
    mChi3   r2, lr, r7, r0, r1, \offsetA03, locMem

    @ Rho-east: Plane shift
    movs    r0, #32-1
    mov     r1, r10
    rors    r1, r1, r0
    mov     r10, r1
    mov     r1, r11
    rors    r1, r1, r0
    mov     r11, r1
    mov     r1, r12
    rors    r1, r1, r0
    mov     r12, r1
    mov     r1, lr
    rors    r1, r1, r0
    mov     lr, r1

    movs    r0, #32-_e1
    rors    r4, r4, r0
    rors    r5, r5, r0
    rors    r6, r6, r0
    rors    r7, r7, r0

    mov     r0, r4
    mov     r4, r6
    mov     r6, r0
    mov     r0, r5
    mov     r5, r7
    mov     r7, r0

    .endm

@ ----------------------------------------------------------------------------
@
@ void Xoodoo_Permute_Nrounds(void *state, unsigned int nrounds)
@

@ offsets on stack
.equ Xoodoo_Permute_Nrounds_offsetA03   , 0
.equ Xoodoo_Permute_Nrounds_offsetRC    , 4
.equ Xoodoo_Permute_Nrounds_SAS         , 8
.equ Xoodoo_Permute_Nrounds_offsetState , Xoodoo_Permute_Nrounds_SAS

.global Xoodoo_Permute_Nrounds
.type	Xoodoo_Permute_Nrounds, %function;
Xoodoo_Permute_Nrounds:
    push    {r4-r6,lr}
    mov     r2, r8
    mov     r3, r9
    mov     r4, r10
    mov     r5, r11
    push    {r0,r2-r5,r7}

    sub     sp, #Xoodoo_Permute_Nrounds_SAS
    adr     r2, Xoodoo_Permute_RoundConstants12
    lsls    r1, r1, #2
    subs    r2, r2, r1
    str     r2, [sp, #Xoodoo_Permute_Nrounds_offsetRC]

    ldm     r0!, {r3,r5,r6,r7}
    mov     r8, r5
    mov     r9, r6
    str     r7, [sp, #Xoodoo_Permute_Nrounds_offsetA03]
    ldm     r0!, {r4,r5,r6,r7}
    mov     r10, r4
    mov     r11, r5
    mov     r12, r6
    mov     lr, r7
    ldm     r0!, {r4,r5,r6,r7}
Xoodoo_Permute_Nrouds_Loop:
    mRound  Xoodoo_Permute_Nrounds_offsetRC, Xoodoo_Permute_Nrounds_offsetA03
    ldr     r0, [sp, #Xoodoo_Permute_Nrounds_offsetRC]
    ldr     r0, [r0]
    cmp     r0, #0
    beq     Xoodoo_Permute_Nrouds_Done
    b       Xoodoo_Permute_Nrouds_Loop
Xoodoo_Permute_Nrouds_Done:
    ldr     r0, [sp, #Xoodoo_Permute_Nrounds_offsetState]

    stm     r0!, {r3}
    mov     r1, r8
    mov     r2, r9
    ldr     r3, [sp, #Xoodoo_Permute_Nrounds_offsetA03]
    stm     r0!, {r1,r2,r3}

    mov     r1, r10
    mov     r2, r11
    mov     r3, r12
    stm     r0!, {r1,r2,r3}

    mov     r1, lr
    stm     r0!, {r1,r4,r5,r6,r7}

    add     sp, #Xoodoo_Permute_Nrounds_SAS
    pop     {r0-r4,r7}
    mov     r8, r1
    mov     r9, r2
    mov     r10, r3
    mov     r11, r4
    pop     {r4-r6,pc}
    .align  4


Xoodoo_Permute_RoundConstants:
    .long      0x00000058
    .long      0x00000038
    .long      0x000003C0
    .long      0x000000D0
    .long      0x00000120
    .long      0x00000014
    .long      0x00000060
    .long      0x0000002C
    .long      0x00000380
    .long      0x000000F0
    .long      0x000001A0
    .long      0x00000012
Xoodoo_Permute_RoundConstants12:
    .long      0
    .align  4

@ ----------------------------------------------------------------------------
@
@  void Xoodoo_Permute_6rounds( void *state )
@
.global Xoodoo_Permute_6rounds
.type	Xoodoo_Permute_6rounds, %function;
Xoodoo_Permute_6rounds:
    movs    r1, #6
    b       Xoodoo_Permute_Nrounds
    .align  4



@ ----------------------------------------------------------------------------
@
@  void Xoodoo_Permute_12rounds( void *state )
@
.global Xoodoo_Permute_12rounds
.type	Xoodoo_Permute_12rounds, %function;
Xoodoo_Permute_12rounds:
    movs    r1, #12
    b       Xoodoo_Permute_Nrounds
    .align  4



.equ Xoofff_BlockSize       , 3*4*4

@ ----------------------------------------------------------------------------
@
@ void Xoofff_AddIs(BitSequence *output, const BitSequence *input, BitLength bitLen)
.global Xoofff_AddIs
.type	Xoofff_AddIs, %function;
Xoofff_AddIs:
    push    {r4-r6,lr}
    movs    r3, r0                                  @ check input and output pointer both 32-bit    .align 8ed
    orrs    r3, r3, r1
    lsls    r3, r3, #30
    bne     Xoofff_AddIs_Bytes
    subs    r2, r2, #16*8
    bcc     Xoofff_AddIs_LessThan16
Xoofff_AddIs_16Loop:
    ldr     r3, [r0, #0]
    ldr     r4, [r0, #4]
    ldmia   r1!, {r5,r6}
    eors    r3, r3, r5
    eors    r4, r4, r6
    stmia   r0!, {r3,r4}
    ldr     r3, [r0, #0]
    ldr     r4, [r0, #4]
    ldmia   r1!, {r5,r6}
    eors    r3, r3, r5
    eors    r4, r4, r6
    stmia   r0!, {r3,r4}
    subs    r2, r2, #16*8
    bcs     Xoofff_AddIs_16Loop
Xoofff_AddIs_LessThan16:
    adds    r2, r2, #16*8
    beq     Xoofff_AddIs_Return
    subs    r2, r2, #4*8
    bcc     Xoofff_AddIs_LessThan4
Xoofff_AddIs_4Loop:
    ldr     r3, [r0]
    ldmia   r1!, {r4}
    eors    r3, r3, r4
    stmia   r0!, {r3}
    subs    r2, r2, #4*8
    bcs     Xoofff_AddIs_4Loop
Xoofff_AddIs_LessThan4:
    adds    r2, r2, #4*8
    beq     Xoofff_AddIs_Return
Xoofff_AddIs_Bytes:
    subs    r2, r2, #8
    bcc     Xoofff_AddIs_LessThan1
Xoofff_AddIs_1Loop:
    ldrb    r3, [r0]
    ldrb    r4, [r1]
    adds    r1, r1, #1
    eors    r3, r3, r4
    strb    r3, [r0]
    adds    r0, r0, #1
    subs    r2, r2, #8
    bcs     Xoofff_AddIs_1Loop
Xoofff_AddIs_LessThan1:
    adds    r2, r2, #8
    beq     Xoofff_AddIs_Return
    ldrb    r3, [r0]
    ldrb    r4, [r1]
    movs    r1, #1
    eors    r3, r3, r4
    lsls    r1, r1, r2
    subs    r1, r1, #1
    ands    r3, r3, r1
    strb    r3, [r0]
Xoofff_AddIs_Return:
    pop     {r4-r6,pc}
    .align  4


.macro    mLdu    rv, ri, tt
    ldrb    \rv, [\ri, #3]
    lsls    \rv, \rv, #8
    ldrb    \tt, [\ri, #2]
    orrs    \rv, \rv, \tt
    lsls    \rv, \rv, #8
    ldrb    \tt, [\ri, #1]
    orrs    \rv, \rv, \tt
    lsls    \rv, \rv, #8
    ldrb    \tt, [\ri, #0]
    orrs    \rv, \rv, \tt
    adds    \ri, \ri, #4
    .endm

@ ----------------------------------------------------------------------------
@
@ size_t Xoofff_CompressFastLoop(unsigned char *kRoll, unsigned char *xAccu, const unsigned char *input, size_t length)
@

@ offsets on stack
.equ Xoofff_CompressFastLoop_offsetA03  , 0
.equ Xoofff_CompressFastLoop_offsetRC   , 4
.equ Xoofff_CompressFastLoop_SAS        , 8
.equ Xoofff_CompressFastLoop_kRoll      , Xoofff_CompressFastLoop_SAS+0
.equ Xoofff_CompressFastLoop_input      , Xoofff_CompressFastLoop_SAS+4
.equ Xoofff_CompressFastLoop_xAccu      , Xoofff_CompressFastLoop_SAS+8+16
.equ Xoofff_CompressFastLoop_iInput     , Xoofff_CompressFastLoop_SAS+12+16
.equ Xoofff_CompressFastLoop_length     , Xoofff_CompressFastLoop_SAS+16+16

.global Xoofff_CompressFastLoop
.type	Xoofff_CompressFastLoop, %function;
Xoofff_CompressFastLoop:
    subs    r3, #Xoofff_BlockSize       @ length must be greater than block size
    push    {r1-r7,lr}
    mov     r4, r8
    mov     r5, r9
    mov     r6, r10
    mov     r7, r11
    push    {r0,r2,r4-r7}
    sub     sp, #Xoofff_CompressFastLoop_SAS
    ldm     r0!, {r3,r5,r6,r7}      @ get initial kRoll
    mov     r8, r5
    mov     r9, r6
    str     r7, [sp, #Xoofff_CompressFastLoop_offsetA03]
    ldm     r0!, {r4,r5,r6,r7}
    mov     r10, r4
    mov     r11, r5
    mov     r12, r6
    mov     lr, r7
    ldm     r0!, {r4,r5,r6,r7}
Xoofff_CompressFastLoop_Loop:
    adr     r1, Xoofff_CompressFastLoop_RoundConstants6
    str     r1, [sp, #Xoofff_CompressFastLoop_offsetRC]

    ldr     r0, [sp, #Xoofff_CompressFastLoop_input]    @ add input
    lsls    r1, r0, #30
    bne     Xoofff_CompressFastLoop_Unaligned

Xoofff_CompressFastLoop_Aligned:
    ldmia   r0!, {r1}
    eors    r3, r3, r1
    ldmia   r0!, {r1}
    mov     r2, r8
    eors    r2, r2, r1
    mov     r8, r2
    ldmia   r0!, {r1}
    mov     r2, r9
    eors    r2, r2, r1
    mov     r9, r2
    ldmia   r0!, {r1}
    ldr     r2, [sp, #Xoofff_CompressFastLoop_offsetA03]
    eors    r2, r2, r1
    str     r2, [sp, #Xoofff_CompressFastLoop_offsetA03]

    ldmia   r0!, {r1}
    mov     r2, r10
    eors    r2, r2, r1
    mov     r10, r2
    ldmia   r0!, {r1}
    mov     r2, r11
    eors    r2, r2, r1
    mov     r11, r2
    ldmia   r0!, {r1}
    mov     r2, r12
    eors    r2, r2, r1
    mov     r12, r2
    ldmia   r0!, {r1}
    mov     r2, lr
    eors    r2, r2, r1
    mov     lr, r2

    ldmia   r0!, {r1,r2}
    eors    r4, r4, r1
    eors    r5, r5, r2
    ldmia   r0!, {r1,r2}
    eors    r6, r6, r1
    eors    r7, r7, r2

    b       Xoofff_CompressFastLoop_Permute
    .align  4
Xoofff_CompressFastLoop_RoundConstants6:
    .long      0x00000060
    .long      0x0000002C
    .long      0x00000380
    .long      0x000000F0
    .long      0x000001A0
    .long      0x00000012
    .long      0

Xoofff_CompressFastLoop_Unaligned:
    mLdu    r1, r0, r2
    eors    r3, r3, r1
    mLdu    r1, r0, r2
    mov     r2, r8
    eors    r2, r2, r1
    mov     r8, r2
    mLdu    r1, r0, r2
    mov     r2, r9
    eors    r2, r2, r1
    mov     r9, r2
    mLdu    r1, r0, r2
    ldr     r2, [sp, #Xoofff_CompressFastLoop_offsetA03]
    eors    r2, r2, r1
    str     r2, [sp, #Xoofff_CompressFastLoop_offsetA03]

    mLdu    r1, r0, r2
    mov     r2, r10
    eors    r2, r2, r1
    mov     r10, r2
    mLdu    r1, r0, r2
    mov     r2, r11
    eors    r2, r2, r1
    mov     r11, r2
    mLdu    r1, r0, r2
    mov     r2, r12
    eors    r2, r2, r1
    mov     r12, r2
    mLdu    r1, r0, r2
    mov     r2, lr
    eors    r2, r2, r1
    mov     lr, r2

    mLdu    r1, r0, r2
    eors    r4, r4, r1
    mLdu    r1, r0, r2
    eors    r5, r5, r1
    mLdu    r1, r0, r2
    eors    r6, r6, r1
    mLdu    r1, r0, r2
    eors    r7, r7, r1

Xoofff_CompressFastLoop_Permute:
    str     r0, [sp, #Xoofff_CompressFastLoop_input]
Xoofff_CompressFastLoop_PermuteLoop:
    mRound  Xoofff_CompressFastLoop_offsetRC, Xoofff_CompressFastLoop_offsetA03
    ldr     r0, [sp, #Xoofff_CompressFastLoop_offsetRC]
    ldr     r0, [r0]
    cmp     r0, #0
    beq     Xoofff_CompressFastLoop_PermuteDone
    b       Xoofff_CompressFastLoop_PermuteLoop
Xoofff_CompressFastLoop_PermuteDone:

    @ Extract and add into xAccu
    ldr     r0, [sp, #Xoofff_CompressFastLoop_xAccu]

    ldr     r1, [r0]
    eors    r1, r1, r3
    stmia   r0!, {r1}

    ldr     r1, [r0]
    mov     r2, r8
    eors    r1, r1, r2
    stmia   r0!, {r1}

    ldr     r1, [r0]
    mov     r2, r9
    eors    r1, r1, r2
    stmia   r0!, {r1}

    ldr     r1, [r0]
    ldr     r2, [sp, #Xoofff_CompressFastLoop_offsetA03]
    eors    r1, r1, r2
    stmia   r0!, {r1}


    ldr     r1, [r0]
    mov     r2, r10
    eors    r1, r1, r2
    stmia   r0!, {r1}
    ldr     r1, [r0]
    mov     r2, r11
    eors    r1, r1, r2
    stmia   r0!, {r1}
    ldr     r1, [r0]
    mov     r2, r12
    eors    r1, r1, r2
    stmia   r0!, {r1}
    ldr     r1, [r0]
    mov     r2, lr
    eors    r1, r1, r2
    stmia   r0!, {r1}

    ldr     r1, [r0, #0]
    ldr     r2, [r0, #4]
    ldr     r3, [r0, #8]
    eors    r1, r1, r4
    ldr     r4, [r0, #12]
    eors    r2, r2, r5
    eors    r3, r3, r6
    eors    r4, r4, r7
    stm     r0!, {r1,r2,r3,r4}

    @roll kRoll-c
    ldr     r0, [sp, #Xoofff_CompressFastLoop_kRoll]
    ldmia   r0!, {r7}
    ldmia   r0!, {r4-r6}
    ldmia   r0!, {r3}
    ldmia   r0!, {r1,r2}
    mov     r8, r1
    mov     r9, r2
    ldmia   r0!, {r1,r2}
    str     r1, [sp, #Xoofff_CompressFastLoop_offsetA03]
    mov     r10, r2
    ldmia   r0!, {r1,r2}
    mov     r11, r1
    mov     r12, r2
    ldmia   r0!, {r1}
    mov     lr, r1

    lsls    r1, r7, #13
    eors    r7, r7, r1
    mov     r1, r3
    movs    r2, #32-3
    rors    r1, r1, r2
    eors    r7, r7, r1

    subs    r0, r0, #Xoofff_BlockSize
    stmia   r0!, {r3}
    mov     r1, r8
    mov     r2, r9
    stmia   r0!, {r1,r2}
    ldr     r1, [sp, #Xoofff_CompressFastLoop_offsetA03]
    mov     r2, r10
    stmia   r0!, {r1,r2}
    mov     r1, r11
    mov     r2, r12
    stmia   r0!, {r1,r2}
    mov     r1, lr
    stmia   r0!, {r1,r4-r7}

    @ loop management
    ldr     r0, [sp, #Xoofff_CompressFastLoop_length]
    subs    r0, #Xoofff_BlockSize
    str     r0, [sp, #Xoofff_CompressFastLoop_length]
    bcc     Xoofff_CompressFastLoop_Done
    b       Xoofff_CompressFastLoop_Loop
Xoofff_CompressFastLoop_Done:
    @ return number of bytes processed
    ldr     r0, [sp, #Xoofff_CompressFastLoop_input]
    ldr     r1, [sp, #Xoofff_CompressFastLoop_iInput]
    subs    r0, r0, r1
    add     sp, #Xoofff_CompressFastLoop_SAS+8
    pop     {r4-r7}
    mov     r8, r4
    mov     r9, r5
    mov     r10, r6
    mov     r11, r7
    pop     {r1-r7,pc}
    .align  4


.macro    mStu    rv, ro
    strb    \rv, [\ro, #0]
    lsrs    \rv, \rv, #8
    strb    \rv, [\ro, #1]
    lsrs    \rv, \rv, #8
    strb    \rv, [\ro, #2]
    lsrs    \rv, \rv, #8
    strb    \rv, [\ro, #3]
    adds    \ro, \ro, #4
    .endm

@ ----------------------------------------------------------------------------
@
@ size_t Xoofff_ExpandFastLoop(unsigned char *yAccu, const unsigned char *kRoll, unsigned char *output, size_t length)
@

@ offsets on stack
.equ Xoofff_ExpandFastLoop_offsetA03, 0
.equ Xoofff_ExpandFastLoop_offsetRC , 4
.equ Xoofff_ExpandFastLoop_SAS      , 8
.equ Xoofff_ExpandFastLoop_yAccu    , Xoofff_ExpandFastLoop_SAS+0
.equ Xoofff_ExpandFastLoop_output   , Xoofff_ExpandFastLoop_SAS+4
.equ Xoofff_ExpandFastLoop_kRoll    , Xoofff_ExpandFastLoop_SAS+8+16
.equ Xoofff_ExpandFastLoop_iOutput  , Xoofff_ExpandFastLoop_SAS+12+16
.equ Xoofff_ExpandFastLoop_length   , Xoofff_ExpandFastLoop_SAS+16+16

.global Xoofff_ExpandFastLoop
.type	Xoofff_ExpandFastLoop, %function;
Xoofff_ExpandFastLoop:
    subs    r3, #Xoofff_BlockSize                       @ length must be greater than block size
    push    {r1-r7,lr}
    mov     r4, r8
    mov     r5, r9
    mov     r6, r10
    mov     r7, r11
    push    {r0,r2,r4-r7}
    sub     sp, #Xoofff_ExpandFastLoop_SAS

    ldm     r0!, {r3,r5,r6,r7}                      @ get initial yAccu
    mov     r8, r5
    mov     r9, r6
    str     r7, [sp, #Xoofff_ExpandFastLoop_offsetA03]
    ldm     r0!, {r4,r5,r6,r7}
    mov     r10, r4
    mov     r11, r5
    mov     r12, r6
    mov     lr, r7
    ldm     r0!, {r4,r5,r6,r7}
Xoofff_ExpandFastLoop_Loop:
    adr     r1, Xoofff_ExpandFastLoop_RoundConstants6
    str     r1, [sp, #Xoofff_ExpandFastLoop_offsetRC]
Xoofff_ExpandFastLoop_PermuteLoop:
    mRound  Xoofff_ExpandFastLoop_offsetRC, Xoofff_ExpandFastLoop_offsetA03
    ldr     r0, [sp, #Xoofff_ExpandFastLoop_offsetRC]
    ldr     r0, [r0]
    cmp     r0, #0
    beq     Xoofff_ExpandFastLoop_PermuteDone
    b       Xoofff_ExpandFastLoop_PermuteLoop
Xoofff_ExpandFastLoop_RoundConstants6:
    .long      0x00000060
    .long      0x0000002C
    .long      0x00000380
    .long      0x000000F0
    .long      0x000001A0
    .long      0x00000012
    .long      0
Xoofff_ExpandFastLoop_PermuteDone:
    @ Add k and extract
    ldr     r0, [sp, #Xoofff_ExpandFastLoop_kRoll]
    ldr     r1, [sp, #Xoofff_ExpandFastLoop_output]    @ add input
    lsls    r2, r1, #30
    bne     Xoofff_ExpandFastLoop_Unaligned
Xoofff_ExpandFastLoop_Aligned:
    ldmia   r0!, {r2}
    eors    r2, r2, r3
    stmia   r1!, {r2}
    ldmia   r0!, {r2}
    mov     r3, r8
    eors    r2, r2, r3
    stmia   r1!, {r2}
    ldmia   r0!, {r2}
    mov     r3, r9
    eors    r2, r2, r3
    stmia   r1!, {r2}
    ldmia   r0!, {r2}
    ldr     r3, [sp, #Xoofff_ExpandFastLoop_offsetA03]
    eors    r2, r2, r3
    stmia   r1!, {r2}

    ldmia   r0!, {r2}
    mov     r3, r10
    eors    r2, r2, r3
    stmia   r1!, {r2}
    ldmia   r0!, {r2}
    mov     r3, r11
    eors    r2, r2, r3
    stmia   r1!, {r2}
    ldmia   r0!, {r2}
    mov     r3, r12
    eors    r2, r2, r3
    stmia   r1!, {r2}
    ldmia   r0!, {r2}
    mov     r3, lr
    eors    r2, r2, r3
    stmia   r1!, {r2}

    ldmia   r0!, {r2,r3}
    eors    r2, r2, r4
    eors    r3, r3, r5
    stmia   r1!, {r2,r3}
    ldmia   r0!, {r2,r3}
    eors    r2, r2, r6
    eors    r3, r3, r7
    stmia   r1!, {r2,r3}
    b       Xoofff_ExpandFastLoop_ExtractDone

Xoofff_ExpandFastLoop_Unaligned:
    ldmia   r0!, {r2}
    eors    r2, r2, r3
    mStu    r2, r1
    ldmia   r0!, {r2}
    mov     r3, r8
    eors    r2, r2, r3
    mStu    r2, r1
    ldmia   r0!, {r2}
    mov     r3, r9
    eors    r2, r2, r3
    mStu    r2, r1
    ldmia   r0!, {r2}
    ldr     r3, [sp, #Xoofff_ExpandFastLoop_offsetA03]
    eors    r2, r2, r3
    mStu    r2, r1

    ldmia   r0!, {r2}
    mov     r3, r10
    eors    r2, r2, r3
    mStu    r2, r1
    ldmia   r0!, {r2}
    mov     r3, r11
    eors    r2, r2, r3
    mStu    r2, r1
    ldmia   r0!, {r2}
    mov     r3, r12
    eors    r2, r2, r3
    mStu    r2, r1
    ldmia   r0!, {r2}
    mov     r3, lr
    eors    r2, r2, r3
    mStu    r2, r1

    ldmia   r0!, {r2,r3}
    eors    r2, r2, r4
    mStu    r2, r1
    eors    r3, r3, r5
    mStu    r3, r1
    ldmia   r0!, {r2,r3}
    eors    r2, r2, r6
    mStu    r2, r1
    eors    r3, r3, r7
    mStu    r3, r1

Xoofff_ExpandFastLoop_ExtractDone:
    str     r1, [sp, #Xoofff_ExpandFastLoop_output]

    @ roll-e yAccu
    ldr     r0, [sp, #Xoofff_ExpandFastLoop_yAccu]
    ldmia   r0!, {r7}
    ldmia   r0!, {r4-r6}
    ldmia   r0!, {r3}
    ldmia   r0!, {r1,r2}
    mov     r8, r1
    mov     r9, r2
    ldmia   r0!, {r1,r2}
    str     r1, [sp, #Xoofff_ExpandFastLoop_offsetA03]
    mov     r10, r2
    ldmia   r0!, {r1,r2}
    mov     r11, r1
    mov     r12, r2
    ldmia   r0!, {r1}
    mov     lr, r1

    mov     r1, r10
    ands    r1, r1, r3
    movs    r2, #32-5
    rors    r7, r7, r2
    eors    r7, r7, r1
    movs    r2, #32-13
    mov     r1, r3
    rors    r1, r1, r2
    eors    r7, r7, r1
    movs    r1, #7
    eors    r7, r7, r1

    subs    r0, r0, #Xoofff_BlockSize
    stmia   r0!, {r3}
    mov     r1, r8
    mov     r2, r9
    stmia   r0!, {r1,r2}
    ldr     r1, [sp, #Xoofff_ExpandFastLoop_offsetA03]
    mov     r2, r10
    stmia   r0!, {r1,r2}
    mov     r1, r11
    mov     r2, r12
    stmia   r0!, {r1,r2}
    mov     r1, lr
    stmia   r0!, {r1,r4-r7}

    @ loop management
    ldr     r0, [sp, #Xoofff_ExpandFastLoop_length]
    subs    r0, #Xoofff_BlockSize
    str     r0, [sp, #Xoofff_ExpandFastLoop_length]
    bcc     Xoofff_ExpandFastLoop_Done
    b       Xoofff_ExpandFastLoop_Loop
Xoofff_ExpandFastLoop_Done:
    @ return number of bytes processed
    ldr     r0, [sp, #Xoofff_ExpandFastLoop_output]
    ldr     r1, [sp, #Xoofff_ExpandFastLoop_iOutput]
    subs    r0, r0, r1
    add     sp, #Xoofff_ExpandFastLoop_SAS+8
    pop     {r4-r7}
    mov     r8, r4
    mov     r9, r5
    mov     r10, r6
    mov     r11, r7
    pop     {r1-r7,pc}
    .align  4