SHA-256 Merkle tree: Difference between revisions

add task to aarch64 assembly raspberry pi
(add task to arm assembly raspberry pi)
(add task to aarch64 assembly raspberry pi)
Line 5:
 
Implement this algorithm in your language; you can use the code from the [[SHA-256]] task for the actual hash computations. For better manageability and portability, build the tree using a smaller block size of only 1024 bytes, and demonstrate it on [https://{{SERVERNAME}}/mw/title.png the RosettaCode title image] with that block size. The final result should be the hexadecimal digest value <tt style="font-weight:bold">a4f902cf9d51fe51eda156a6792e1445dff65edf3a217a1f3334cc9cf1495c2c</tt>.
=={{header|AArch64 Assembly}}==
{{works with|as|Raspberry Pi 3B version Buster 64 bits <br> or android 64 bits with application Termux }}
<lang AArch64 Assembly>
/* ARM assembly AARCH64 Raspberry PI 3B or android 64 bits */
/* program merkleRoot64.s */
 
/*******************************************/
/* Constantes file */
/*******************************************/
/* for this file see task include a file in language AArch64 assembly*/
.include "../includeConstantesARM64.inc"
 
.equ O_RDWR, 0x0002 // open for reading and writing
 
.equ BUFFERSIZE, 65535 // file buffer
.equ LGBLOCK, 1024 // block length
 
.equ LGHASH, 32 // hash length
.equ NBELEMENTS, 40 // array size
.equ LGZONETRAV, 2048 // process array size
 
/*******************************************/
/* Structures */
/********************************************/
/* structure variables hash compute */
.struct 0
var_a: // a
.struct var_a + 4
var_b: // b
.struct var_b + 4
var_c: // c
.struct var_c + 4
var_d: // d
.struct var_d + 4
var_e: // e
.struct var_e + 4
var_f: // f
.struct var_f + 4
var_g: // g
.struct var_g + 4
var_h: // h
.struct var_h + 4
/***********************************/
/* Initialized data */
/***********************************/
.data
szFileName: .asciz "title.png"
szCarriageReturn: .asciz "\n"
szMessErreur: .asciz "Error detected.\n"
szMessNbHash: .asciz "Start hash number : @ \n"
szMessTest: .asciz "Rosetta code"
.align 4
/* array constantes Hi */
tbConstHi: .int 0x6A09E667 // H0
.int 0xBB67AE85 // H1
.int 0x3C6EF372 // H2
.int 0xA54FF53A // H3
.int 0x510E527F // H4
.int 0x9B05688C // H5
.int 0x1F83D9AB // H6
.int 0x5BE0CD19 // H7
/* array 64 constantes Kt */
tbConstKt:
.int 0x428a2f98, 0x71374491, 0xb5c0fbcf, 0xe9b5dba5, 0x3956c25b, 0x59f111f1, 0x923f82a4, 0xab1c5ed5
.int 0xd807aa98, 0x12835b01, 0x243185be, 0x550c7dc3, 0x72be5d74, 0x80deb1fe, 0x9bdc06a7, 0xc19bf174
.int 0xe49b69c1, 0xefbe4786, 0x0fc19dc6, 0x240ca1cc, 0x2de92c6f, 0x4a7484aa, 0x5cb0a9dc, 0x76f988da
.int 0x983e5152, 0xa831c66d, 0xb00327c8, 0xbf597fc7, 0xc6e00bf3, 0xd5a79147, 0x06ca6351, 0x14292967
.int 0x27b70a85, 0x2e1b2138, 0x4d2c6dfc, 0x53380d13, 0x650a7354, 0x766a0abb, 0x81c2c92e, 0x92722c85
.int 0xa2bfe8a1, 0xa81a664b, 0xc24b8b70, 0xc76c51a3, 0xd192e819, 0xd6990624, 0xf40e3585, 0x106aa070
.int 0x19a4c116, 0x1e376c08, 0x2748774c, 0x34b0bcb5, 0x391c0cb3, 0x4ed8aa4a, 0x5b9cca4f, 0x682e6ff3
.int 0x748f82ee, 0x78a5636f, 0x84c87814, 0x8cc70208, 0x90befffa, 0xa4506ceb, 0xbef9a3f7, 0xc67178f2
/***********************************/
/* UnInitialized data */
/***********************************/
.bss
sBuffer: .skip BUFFERSIZE // file buffer
.align 8
qNbBlocs: .skip 8
qNbHash: .skip 8
sZoneConv: .skip 24
sZoneTrav: .skip LGZONETRAV
HashResult: .skip LGHASH + 8
HashProcess: .skip LGHASH * 2
tbListHash: .skip LGHASH * NBELEMENTS
tbH: .skip 4 * 8 // 8 variables H
tbabcdefgh: .skip 4 * 8
tbW: .skip 4 * 64 // 64 words W
 
/***********************************/
/* code section */
/***********************************/
.text
.global main
main:
mov x0,AT_FDCWD // current directory
ldr x1,qAdrszFileName // File name
mov x2,#O_RDWR // flags
mov x3,#0 // mode
mov x8,#OPEN // open file
svc #0
cmp x0,#0 // error ?
ble error
mov x19,x0 // save fd
ldr x1,qAdrsBuffer
mov x2,#BUFFERSIZE
mov x8,#READ // call system read file
svc 0
cmp x0,#0 // error read ?
ble error
mov x7,x0 // number of read characters
ldr x6,qAdrsBuffer
mov x5,#0 // counter characters block
1:
add x0,x6,x5 // start address of each block
ldr x1,qAdrHashResult
mov x2,#LGBLOCK
bl computeSHA256LG
bl storeHash // store hash in start array
cmp x0,#-1
beq error
add x5,x5,#LGBLOCK // new buffer offset
sub x0,x7,x5 //
cmp x0,#LGBLOCK // last block
bge 1b // and loop
sub x2,x7,x5 // length last block
add x0,x6,x5 // address last block
ldr x1,qAdrHashResult
bl computeSHA256LG
bl storeHash
cmp x0,#-1
beq error
ldr x0,qAdrqNbHash
ldr x0,[x0]
ldr x1,qAdrsZoneConv
bl conversion10
ldr x0,qAdrszMessNbHash
ldr x1,qAdrsZoneConv
bl strInsertAtCharInc
bl affichageMess
ldr x0,qAdrtbListHash
ldr x1,qAdrHashResult
bl calculerMerkleRoot
ldr x0,qAdrHashResult
bl displaySHA256 // display résult
end:
mov x0,x19 // FD
mov x8, #CLOSE // call system close file
svc #0
cmp x0,#0
blt error
mov x0,#0 // return code
b 100f
error:
ldr x0,qAdrszMessErreur // error message
bl affichageMess
mov x0,#1 // return error code
100: // standard end of the program
mov x8, #EXIT // request to exit program
svc 0 // perform system call
qAdrsBuffer: .quad sBuffer
qAdrszFileName: .quad szFileName
qAdrszMessErreur: .quad szMessErreur
qAdrszCarriageReturn: .quad szCarriageReturn
qAdrHashResult: .quad HashResult
qAdrszMessNbHash: .quad szMessNbHash
qAdrszMessTest: .quad szMessTest
/******************************************************************/
/* store hash in start array */
/******************************************************************/
/* x0 hash address */
storeHash:
stp x1,lr,[sp,-16]! // save registers
stp x2,x3,[sp,-16]! // save registers
stp x4,x5,[sp,-16]! // save registers
ldr x2,qAdrqNbHash // number element counter
ldr x3,[x2]
ldr x1,qAdrtbListHash // address array hash
mov x4,#LGHASH // hash length
madd x5,x4,x3,x1 // compute store address
mov x1,#0 // counter
1:
ldr w4,[x0,x1] // load four bytes
str w4,[x5,x1] // store four bytes
add x1,x1,#4
cmp x1,#LGHASH // 32 bytes ?
blt 1b // no -> loop
add x3,x3,#1
cmp x3,#NBELEMENTS
bge 99f // error
str x3,[x2] // store new counter hash
b 100f
99:
mov x0,#-1 // error ?
100:
ldp x4,x5,[sp],16 // restaur 2 registers
ldp x2,x3,[sp],16 // restaur 2 registers
ldp x1,lr,[sp],16 // restaur 2 registers
ret // return to address lr x30
qAdrqNbHash: .quad qNbHash
qAdrtbListHash: .quad tbListHash
/******************************************************************/
/* compute hash root Merkle */
/******************************************************************/
// x0 start array hash address
// x1 result array address (32 bytes)
calculerMerkleRoot:
stp x1,lr,[sp,-16]! // save registers
stp x2,x3,[sp,-16]! // save registers
mov x10,x1
mov x12,sp // save stack adresse
ldr x3,qAdrqNbHash
ldr x3,[x3]
cmp x3,#0 // 0 hash ?
beq 99f // error
mov x4,#LGHASH * 2
mul x1,x3,x4 // compute hash size * blocks number * 2
sub sp,sp,x1 // reserve array
mov fp,sp // address previousTreeLayer
lsr x1,x1,#1 // reserve size / 2
add x7,fp,x1 // address TreeLayer
 
mov x2,#0 // counter
mov x4,#LGHASH
1:
mul x1,x2,x4
add x6,x0,x1
add x8,fp,x1
mov x5,#0
2: // loop copying 32 octets hash
ldr w9,[x6,x5]
str w9,[x8,x5]
add x5,x5,#4 // count
cmp x5,#LGHASH
blt 2b
add x2,x2,#1 // next hash block
cmp x2,x3 // maxi ?
blt 1b
mov x0,fp
mov x11,#0 // indice TreeLayer
mov x5,#0 // indice layer
3: // loop
cmp x3,#1 // one hash ?
beq 12f // yes -> end
sub x3,x3,#1
mov x4,#LGHASH
madd x8,x11,x4,fp // address hash 1
mov x9,x7 // raz TreeLayer
4:
cmp x11,x3
bgt 11f // end loop ?
blt 5f
mov x0,x8 // last odd hash
add x11,x11,#1 // no concatenation
mov x1,#0
41: // but loop copy odd hash in treelayer
ldr w4,[x0,x1]
str w4,[x9,x1]
add x1,x1,#4
cmp x1,#LGHASH
blt 41b
b 9f
5: // other hashes
add x6,x8,#LGHASH // address hash N + 1
6:
add x11,x11,#1
ldr x1,qAdrHashProcess // address array hash concatenation
mov x0,#0
7: // loop copy element N
ldr w4,[x8,x0]
rev w4,w4 // inversion byte in word
str w4,[x1,x0]
add x0,x0,#4
cmp x0,#LGHASH
blt 7b
add x1,x1,#LGHASH
mov x0,#0
8: // loop copy element N + 1
ldr w4,[x6,x0]
rev w4,w4 // inversion byte in word
str w4,[x1,x0]
add x0,x0,#4
cmp x0,#LGHASH
blt 8b
ldr x0,qAdrHashProcess
mov x1,x9
mov x2,#LGHASH * 2
bl computeSHA256LG // calcul du hash complet
9:
add x11,x11,#1 // incremente counter previousTreeLayer
add x5,x5,#1 // incremente counter treeLayer
add x9,x9,#LGHASH // next address treeLayer
add x8,x6,#LGHASH // maj element N avec N + 2
b 4b
11:
mov x0,fp
mov fp,x7 // treelayer in previous
mov x7,x0
mov x3,x5 // counter previous = counter treelayer
mov x5,#0 // raz treelayer
mov x11,#0 // raz previousTreeLayer
b 3b // and loop
12: // end process
mov x1,fp
mov x2,#0
13: // loop copy result
ldr w3,[x1,x2]
str w3,[x10,x2]
add x2,x2,#4
cmp x2,#LGHASH
blt 13b
mov x0,x10 // return address result
b 100f
99: // error
mov x0,#-1
100:
mov sp,x12 // restaur stack
ldp x2,x3,[sp],16 // restaur 2 registers
ldp x1,lr,[sp],16 // restaur 2 registers
ret
qAdrHashProcess: .quad HashProcess
 
/******************************************************************/
/* compute SHA256 avec longueur */
/******************************************************************/
/* x0 contains the string address */
/* x1 contains the return area address */
/* x2 contains string length */
computeSHA256LG:
stp x1,lr,[sp,-16]! // save registers
stp x2,x3,[sp,-16]! // save registers
stp x4,x5,[sp,-16]! // save registers
stp x6,x7,[sp,-16]! // save registers
stp x8,x9,[sp,-16]! // save registers
stp x10,x11,[sp,-16]! // save registers
stp x12,x13,[sp,-16]! // save registers
stp x19,x20,[sp,-16]! // save registers
mov x19,x1
ldr x1,qAdrtbH
mov x3,0
1: // modif 12/11/2021 boucle raz des 3 zones
str xzr,[x1,x3,lsl 3]
add x3,x3,1
cmp x3,4+4+32
blt 1b
ldr x1,qAdrsZoneTrav
mov x3,0
2: // modif 12/11/2021 boucle raz zone de travail
str xzr,[x1,x3,lsl 3]
add x3,x3,1
cmp x3,LGZONETRAV/8
blt 2b
mov x5,x2
mov x2,#0 // counter length
debCopy: // copy string in work area
ldrb w3,[x0,x2]
strb w3,[x1,x2]
cmp x2,x5 // modification 21/11/2021
add x4,x2,1
csel x2,x4,x2,ne
bne debCopy
//affmemtit copieSH x1 3
//affregtit copie 0
lsl x6,x2,#3 // initial message length in bits
mov x3,#0b10000000 // add bit 1 at end of string
strb w3,[x1,x2]
//affmemtit copieSH1 x1 3
add x2,x2,#1 // length in bytes
lsl x4,x2,#3 // length in bits
mov x3,#0
addZeroes:
lsr x5,x2,#6
lsl x5,x5,#6
sub x5,x2,x5
cmp x5,#56
beq storeLength // yes -> end add
strb w3,[x1,x2] // add zero at message end
add x2,x2,#1 // increment lenght bytes
add x4,x4,#8 // increment length in bits
b addZeroes
storeLength:
add x2,x2,#4 // add four bytes
rev w6,w6 // inversion bits initials message length
str w6,[x1,x2] // and store at end
 
ldr x7,qAdrtbConstHi // constantes H address
ldr x4,qAdrtbH // start area H
mov x5,#0
loopConst: // init array H with start constantes
ldr w6,[x7,x5,lsl #2] // load constante
str w6,[x4,x5,lsl #2] // and store
add x5,x5,#1
cmp x5,#8
blt loopConst
// split into block of 64 bytes
add x2,x2,#4 //
lsr x4,x2,#6 // blocks number
ldr x0,qAdrqNbBlocs
str x4,[x0] // save block maxi
mov x7,#0 // n° de block et x1 contient l adresse zone de travail
loopBlock: // begin loop of each block of 64 bytes
mov x0,x7
bl inversion // inversion each word because little indian
ldr x3,qAdrtbW // working area W address
mov x6,#0 // indice t
/* x2 address begin each block */
ldr x1,qAdrsZoneTrav
add x2,x1,x7,lsl #6 // compute block begin indice * 4 * 16
//mov x0,x2
//affmemtit verifBloc x0 10
loopPrep: // loop for expand 80 words
cmp x6,#15 //
bgt expand1
ldr w0,[x2,x6,lsl #2] // load word message
str w0,[x3,x6,lsl #2] // store in first 16 block
b expandEnd
 
expand1:
sub x8,x6,#2
ldr w9,[x3,x8,lsl #2]
ror w10,w9,#17 // fonction e1 (256)
ror w11,w9,#19
eor w10,w10,w11
lsr w11,w9,#10
eor w10,w10,w11
sub x8,x6,#7
ldr w9,[x3,x8,lsl #2]
add w9,w9,w10 // + w - 7
sub x8,x6,#15
ldr w10,[x3,x8,lsl #2]
ror w11,w10,#7 // fonction e0 (256)
ror w12,w10,#18
eor w11,w11,w12
lsr w12,w10,#3
eor w10,w11,w12
add w9,w9,w10
sub x8,x6,#16
ldr w11,[x3,x8,lsl #2]
add w9,w9,w11
 
str w9,[x3,x6,lsl #2]
expandEnd:
add x6,x6,#1
cmp x6,#64 // 64 words ?
blt loopPrep // and loop
 
 
/* COMPUTING THE MESSAGE DIGEST */
/* x1 area H constantes address */
/* x3 working area W address */
/* x5 address constantes K */
/* x6 counter t */
/* x7 block counter */
/* x8 addresse variables a b c d e f g h */
//ldr x0,qAdrtbW
//affmemtit verifW80 x0 20
// init variable a b c d e f g h
ldr x0,qAdrtbH
//affmemtit veriftbh x0 20
ldr x8,qAdrtbabcdefgh
mov x1,#0
loopInita:
ldr w9,[x0,x1,lsl #2]
str w9,[x8,x1,lsl #2]
add x1,x1,#1
cmp x1,#8
blt loopInita
 
ldr x1,qAdrtbConstHi
ldr x5,qAdrtbConstKt
mov x6,#0
loop64T: // begin loop 64 t
ldr w9,[x8,#var_h]
ldr w10,[x8,#var_e] // calcul T1
ror w11,w10,#6 // fonction sigma 1
ror w12,w10,#11
eor w11,w11,w12
ror w12,w10,#25
eor w11,w11,w12
add w9,w9,w11 // h + sigma1 (e)
ldr w0,[x8,#var_f] // fonction ch x and y xor (non x and z)
ldr w4,[x8,#var_g]
and w11,w10,w0
mvn w12,w10
and w12,w12,w4
eor w11,w11,w12
add w9,w9,w11 // h + sigma1 (e) + ch (e,f,g)
ldr w0,[x5,x6,lsl #2] // load constantes k0
add w9,w9,w0
ldr w0,[x3,x6,lsl #2] // Wt
add w9,w9,w0
// calcul T2
ldr w10,[x8,#var_a] // fonction sigma 0
ror w11,w10,#2
ror w12,w10,#13
eor w11,w11,w12
ror w12,w10,#22
eor w11,w11,w12
ldr w2,[x8,#var_b]
ldr w4,[x8,#var_c]
// fonction maj x and y xor x and z xor y and z
and w12,w10,w2
and w0,w10,w4
eor w12,w12,w0
and w0,w2,w4
eor w12,w12,w0 //
add w12,w12,w11 // T2
// compute variables
ldr w4,[x8,#var_g]
str w4,[x8,#var_h]
ldr w4,[x8,#var_f]
str w4,[x8,#var_g]
ldr w4,[x8,#var_e]
str w4,[x8,#var_f]
ldr w4,[x8,#var_d]
add w4,w4,w9 // add T1
str w4,[x8,#var_e]
ldr w4,[x8,#var_c]
str w4,[x8,#var_d]
ldr w4,[x8,#var_b]
str w4,[x8,#var_c]
ldr w4,[x8,#var_a]
str w4,[x8,#var_b]
add w4,w9,w12 // add T1 T2
str w4,[x8,#var_a]
 
add x6,x6,#1 // increment t
cmp x6,#64
blt loop64T
// End block
ldr x0,qAdrtbH // start area H
mov x10,#0
loopStoreH:
ldr w9,[x8,x10,lsl #2]
ldr w3,[x0,x10,lsl #2]
add w3,w3,w9
str w3,[x0,x10,lsl #2] // store variables in H0
add x10,x10,#1
cmp x10,#8
blt loopStoreH
// other bloc
add x7,x7,#1 // increment block
ldr x0,qAdrqNbBlocs
ldr x4,[x0] // restaur maxi block
cmp x7,x4 // maxi ?
 
blt loopBlock // loop other block
ldr x0,qAdrtbH // adresse resultat
//affmemtit resultat x0 3
mov x3,0
loopRes: // boucle de copie du hash
ldr w1,[x0,x3] // dans zone de retour
str w1,[x19,x3]
add x3,x3,4
cmp x3,28
ble loopRes
mov x0,x19
100:
ldp x19,x20,[sp],16 // restaur 2 registers
ldp x12,x13,[sp],16 // restaur 2 registers
ldp x10,x11,[sp],16 // restaur 2 registers
ldp x8,x9,[sp],16 // restaur 2 registers
ldp x6,x7,[sp],16 // restaur 2 registers
ldp x4,x5,[sp],16 // restaur 2 registers
ldp x2,x3,[sp],16 // restaur 2 registers
ldp x1,lr,[sp],16 // restaur 2 registers
ret // return to address lr x30
qAdrsZoneTrav: .quad sZoneTrav
qAdrtbConstHi: .quad tbConstHi
qAdrtbConstKt: .quad tbConstKt
qAdrtbH: .quad tbH
qAdrtbW: .quad tbW
qAdrtbabcdefgh: .quad tbabcdefgh
qAdrqNbBlocs: .quad qNbBlocs
/******************************************************************/
/* inversion des mots de 32 bits d un bloc */
/******************************************************************/
/* x0 contains N° block */
inversion:
stp x1,lr,[sp,-16]! // save registers
stp x2,x3,[sp,-16]! // save registers
ldr x1,qAdrsZoneTrav
add x1,x1,x0,lsl #6 // debut du bloc
mov x2,#0
1: // start loop
ldr w3,[x1,x2,lsl #2]
rev w3,w3
str w3,[x1,x2,lsl #2]
add x2,x2,#1
cmp x2,#16
blt 1b
100:
ldp x2,x3,[sp],16 // restaur 2 registers
ldp x1,lr,[sp],16 // restaur 2 registers
ret // return to address lr x30
/******************************************************************/
/* display hash SHA1 */
/******************************************************************/
/* x0 contains the address of hash */
/* x1 contains the address of recept zone */
conversionSHA256:
stp x1,lr,[sp,-16]! // save registers
stp x2,x3,[sp,-16]! // save registers
mov x3,x0
mov x2,#0
1:
ldr w0,[x3,x2,lsl #2] // load 4 bytes
//rev x0,x0 // reverse bytes
bl conversion16_4W // conversion hexa
add x1,x1,8
add x2,x2,#1
cmp x2,#LGHASH / 4
blt 1b // and loop
 
100:
ldp x2,x3,[sp],16 // restaur 2 registers
ldp x1,lr,[sp],16 // restaur 2 registers
ret // return to address lr x30
/******************************************************************/
/* display hash SHA1 */
/******************************************************************/
/* x0 contains the address of hash */
displaySHA256:
stp x1,lr,[sp,-16]! // save registers
stp x2,x3,[sp,-16]! // save registers
mov x3,x0
mov x2,#0
1:
ldr w0,[x3,x2,lsl #2] // load 4 bytes
//rev x0,x0 // reverse bytes
ldr x1,qAdrsZoneConv
bl conversion16_4W // conversion hexa
ldr x0,qAdrsZoneConv
bl affichageMess
add x2,x2,#1
cmp x2,#LGHASH / 4
blt 1b // and loop
ldr x0,qAdrszCarriageReturn
bl affichageMess // display message
100:
ldp x2,x3,[sp],16 // restaur 2 registers
ldp x1,lr,[sp],16 // restaur 2 registers
ret // return to address lr x30
qAdrsZoneConv: .quad sZoneConv
/******************************************************************/
/* conversion hexadecimal register 32 bits */
/******************************************************************/
/* x0 contains value and x1 address zone receptrice */
conversion16_4W:
stp x0,lr,[sp,-48]! // save registres
stp x1,x2,[sp,32] // save registres
stp x3,x4,[sp,16] // save registres
mov x2,#28 // start bit position
mov x4,#0xF0000000 // mask
mov x3,x0 // save entry value
1: // start loop
and x0,x3,x4 // value register and mask
lsr x0,x0,x2 // right shift
cmp x0,#10 // >= 10 ?
bge 2f // yes
add x0,x0,#48 // no is digit
b 3f
2:
add x0,x0,#55 // else is a letter A-F
3:
strb w0,[x1],#1 // load result and + 1 in address
lsr x4,x4,#4 // shift mask 4 bits left
subs x2,x2,#4 // decrement counter 4 bits <= zero ?
bge 1b // no -> loop
 
100: // fin standard de la fonction
ldp x3,x4,[sp,16] // restaur des 2 registres
ldp x1,x2,[sp,32] // restaur des 2 registres
ldp x0,lr,[sp],48 // restaur des 2 registres
ret
 
/********************************************************/
/* File Include fonctions */
/********************************************************/
/* for this file see task include a file in language AArch64 assembly */
.include "../includeARM64.inc"
 
</lang>
<pre>
Start hash number : 22
A4F902CF9D51FE51EDA156A6792E1445DFF65EDF3A217A1F3334CC9CF1495C2C
</pre>
=={{header|ARM Assembly}}==
{{works with|as|Raspberry Pi}}