Skip to content

Commit

Permalink
Make labels unique across new Edwards functions
Browse files Browse the repository at this point in the history 
using Torben Hansen's Python script listed in
#50
  • Loading branch information
@jargh
jargh committed Apr 21, 2023
1 parent b365f8e commit 8e80e0c
Show file tree
Hide file tree
Showing 6 changed files with 284 additions and 284 deletions.
100 changes: 50 additions & 50 deletions arm/curve25519/edwards25519_scalarmulbase.S
View file
Original file line number Diff line number Diff line change
Expand Up @@ -582,8 +582,8 @@ S2N_BN_SYMBOL(edwards25519_scalarmulbase):
// Initialize accumulator "acc" to either 0 or 2^251 * B depending on
// bit 251 of the (reduced) scalar. That leaves bits 0..250 to handle.

adr x10, edwards25519_0g
adr x11, edwards25519_251g
adr x10, edwards25519_scalarmulbase_edwards25519_0g
adr x11, edwards25519_scalarmulbase_edwards25519_251g
ldp x0, x1, [x10]
ldp x2, x3, [x11]
csel x0, x0, x2, eq
Expand Down Expand Up @@ -638,12 +638,12 @@ S2N_BN_SYMBOL(edwards25519_scalarmulbase):
// end because we made sure bit 251 is clear in the reduced scalar.

mov i, 0
adr tab, edwards25519_gtable
adr tab, edwards25519_scalarmulbase_edwards25519_gtable
mov bias, xzr

// Start of the main loop, repeated 63 times for i = 0, 4, 8, ..., 248

scalarloop:
edwards25519_scalarmulbase_scalarloop:

// Look at the next 4-bit field "bf", adding the previous bias as well.
// Choose the table index "ix" as bf when bf <= 8 and 16 - bf for bf >= 9,
Expand Down Expand Up @@ -926,7 +926,7 @@ scalarloop:

add i, i, 4
cmp i, 252
bcc scalarloop
bcc edwards25519_scalarmulbase_scalarloop

// Insert the optional negation of the projective X coordinate, and
// so by extension the final affine x coordinate x = X/Z and thus
Expand Down Expand Up @@ -959,7 +959,7 @@ scalarloop:
mov x0, 4
add x1, w_3
add x2, z_3
adr x3, p_25519
adr x3, edwards25519_scalarmulbase_p_25519
add x4, tmpspace

// Inline copy of bignum_modinv, identical except for stripping out the
Expand All @@ -971,7 +971,7 @@ scalarloop:
add x21, x4, x10
add x22, x21, x10
mov x10, xzr
copyloop:
edwards25519_scalarmulbase_copyloop:
ldr x11, [x2, x10, lsl #3]
ldr x12, [x3, x10, lsl #3]
str x11, [x21, x10, lsl #3]
Expand All @@ -980,7 +980,7 @@ copyloop:
str xzr, [x1, x10, lsl #3]
add x10, x10, #0x1
cmp x10, x0
b.cc copyloop
b.cc edwards25519_scalarmulbase_copyloop
ldr x11, [x4]
sub x12, x11, #0x1
str x12, [x4]
Expand All @@ -997,7 +997,7 @@ copyloop:
madd x20, x12, x20, x20
madd x20, x11, x20, x20
lsl x2, x0, #7
outerloop:
edwards25519_scalarmulbase_outerloop:
add x10, x2, #0x3f
lsr x5, x10, #6
cmp x5, x0
Expand All @@ -1008,7 +1008,7 @@ outerloop:
mov x16, xzr
mov x19, xzr
mov x10, xzr
toploop:
edwards25519_scalarmulbase_toploop:
ldr x11, [x21, x10, lsl #3]
ldr x12, [x22, x10, lsl #3]
orr x17, x11, x12
Expand All @@ -1022,7 +1022,7 @@ toploop:
csetm x19, ne
add x10, x10, #0x1
cmp x10, x5
b.cc toploop
b.cc edwards25519_scalarmulbase_toploop
orr x11, x13, x14
clz x12, x11
negs x17, x12
Expand All @@ -1042,7 +1042,7 @@ toploop:
mov x9, #0x1
mov x10, #0x3a
tst x15, #0x1
innerloop:
edwards25519_scalarmulbase_innerloop:
csel x11, x14, xzr, ne
csel x12, x16, xzr, ne
csel x17, x8, xzr, ne
Expand All @@ -1064,13 +1064,13 @@ innerloop:
add x8, x8, x8
add x9, x9, x9
sub x10, x10, #0x1
cbnz x10, innerloop
cbnz x10, edwards25519_scalarmulbase_innerloop
mov x13, xzr
mov x14, xzr
mov x17, xzr
mov x19, xzr
mov x10, xzr
congloop:
edwards25519_scalarmulbase_congloop:
ldr x11, [x4, x10, lsl #3]
ldr x12, [x1, x10, lsl #3]
mul x15, x6, x11
Expand All @@ -1097,7 +1097,7 @@ congloop:
adc x14, x14, x15
add x10, x10, #0x1
cmp x10, x0
b.cc congloop
b.cc edwards25519_scalarmulbase_congloop
extr x13, x13, x17, #58
extr x14, x14, x19, #58
ldr x11, [x4]
Expand All @@ -1108,8 +1108,8 @@ congloop:
adds x11, x11, x15
mov x10, #0x1
sub x11, x0, #0x1
cbz x11, wmontend
wmontloop:
cbz x11, edwards25519_scalarmulbase_wmontend
edwards25519_scalarmulbase_wmontloop:
ldr x11, [x3, x10, lsl #3]
ldr x12, [x4, x10, lsl #3]
mul x15, x17, x11
Expand All @@ -1121,32 +1121,32 @@ wmontloop:
str x12, [x4, x15, lsl #3]
add x10, x10, #0x1
sub x11, x10, x0
cbnz x11, wmontloop
wmontend:
cbnz x11, edwards25519_scalarmulbase_wmontloop
edwards25519_scalarmulbase_wmontend:
adcs x16, x16, x13
adc x13, xzr, xzr
sub x15, x10, #0x1
str x16, [x4, x15, lsl #3]
negs x10, xzr
wcmploop:
edwards25519_scalarmulbase_wcmploop:
ldr x11, [x4, x10, lsl #3]
ldr x12, [x3, x10, lsl #3]
sbcs xzr, x11, x12
add x10, x10, #0x1
sub x11, x10, x0
cbnz x11, wcmploop
cbnz x11, edwards25519_scalarmulbase_wcmploop
sbcs xzr, x13, xzr
csetm x13, cs
negs x10, xzr
wcorrloop:
edwards25519_scalarmulbase_wcorrloop:
ldr x11, [x4, x10, lsl #3]
ldr x12, [x3, x10, lsl #3]
and x12, x12, x13
sbcs x11, x11, x12
str x11, [x4, x10, lsl #3]
add x10, x10, #0x1
sub x11, x10, x0
cbnz x11, wcorrloop
cbnz x11, edwards25519_scalarmulbase_wcorrloop
ldr x11, [x1]
mul x17, x11, x20
ldr x12, [x3]
Expand All @@ -1155,8 +1155,8 @@ wcorrloop:
adds x11, x11, x15
mov x10, #0x1
sub x11, x0, #0x1
cbz x11, zmontend
zmontloop:
cbz x11, edwards25519_scalarmulbase_zmontend
edwards25519_scalarmulbase_zmontloop:
ldr x11, [x3, x10, lsl #3]
ldr x12, [x1, x10, lsl #3]
mul x15, x17, x11
Expand All @@ -1168,38 +1168,38 @@ zmontloop:
str x12, [x1, x15, lsl #3]
add x10, x10, #0x1
sub x11, x10, x0
cbnz x11, zmontloop
zmontend:
cbnz x11, edwards25519_scalarmulbase_zmontloop
edwards25519_scalarmulbase_zmontend:
adcs x16, x16, x14
adc x14, xzr, xzr
sub x15, x10, #0x1
str x16, [x1, x15, lsl #3]
negs x10, xzr
zcmploop:
edwards25519_scalarmulbase_zcmploop:
ldr x11, [x1, x10, lsl #3]
ldr x12, [x3, x10, lsl #3]
sbcs xzr, x11, x12
add x10, x10, #0x1
sub x11, x10, x0
cbnz x11, zcmploop
cbnz x11, edwards25519_scalarmulbase_zcmploop
sbcs xzr, x14, xzr
csetm x14, cs
negs x10, xzr
zcorrloop:
edwards25519_scalarmulbase_zcorrloop:
ldr x11, [x1, x10, lsl #3]
ldr x12, [x3, x10, lsl #3]
and x12, x12, x14
sbcs x11, x11, x12
str x11, [x1, x10, lsl #3]
add x10, x10, #0x1
sub x11, x10, x0
cbnz x11, zcorrloop
cbnz x11, edwards25519_scalarmulbase_zcorrloop
mov x13, xzr
mov x14, xzr
mov x17, xzr
mov x19, xzr
mov x10, xzr
crossloop:
edwards25519_scalarmulbase_crossloop:
ldr x11, [x21, x10, lsl #3]
ldr x12, [x22, x10, lsl #3]
mul x15, x6, x11
Expand All @@ -1226,13 +1226,13 @@ crossloop:
csetm x19, cc
add x10, x10, #0x1
cmp x10, x5
b.cc crossloop
b.cc edwards25519_scalarmulbase_crossloop
cmn x17, x17
ldr x15, [x21]
mov x10, xzr
sub x6, x5, #0x1
cbz x6, negskip1
negloop1:
cbz x6, edwards25519_scalarmulbase_negskip1
edwards25519_scalarmulbase_negloop1:
add x11, x10, #0x8
ldr x12, [x21, x11]
extr x15, x12, x15, #58
Expand All @@ -1242,8 +1242,8 @@ negloop1:
mov x15, x12
add x10, x10, #0x8
sub x6, x6, #0x1
cbnz x6, negloop1
negskip1:
cbnz x6, edwards25519_scalarmulbase_negloop1
edwards25519_scalarmulbase_negskip1:
extr x15, x13, x15, #58
eor x15, x15, x17
adcs x15, x15, xzr
Expand All @@ -1252,8 +1252,8 @@ negskip1:
ldr x15, [x22]
mov x10, xzr
sub x6, x5, #0x1
cbz x6, negskip2
negloop2:
cbz x6, edwards25519_scalarmulbase_negskip2
edwards25519_scalarmulbase_negloop2:
add x11, x10, #0x8
ldr x12, [x22, x11]
extr x15, x12, x15, #58
Expand All @@ -1263,15 +1263,15 @@ negloop2:
mov x15, x12
add x10, x10, #0x8
sub x6, x6, #0x1
cbnz x6, negloop2
negskip2:
cbnz x6, edwards25519_scalarmulbase_negloop2
edwards25519_scalarmulbase_negskip2:
extr x15, x14, x15, #58
eor x15, x15, x19
adcs x15, x15, xzr
str x15, [x22, x10]
mov x10, xzr
cmn x17, x17
wfliploop:
edwards25519_scalarmulbase_wfliploop:
ldr x11, [x3, x10, lsl #3]
ldr x12, [x4, x10, lsl #3]
and x11, x11, x17
Expand All @@ -1280,11 +1280,11 @@ wfliploop:
str x11, [x4, x10, lsl #3]
add x10, x10, #0x1
sub x11, x10, x0
cbnz x11, wfliploop
cbnz x11, edwards25519_scalarmulbase_wfliploop
mvn x19, x19
mov x10, xzr
cmn x19, x19
zfliploop:
edwards25519_scalarmulbase_zfliploop:
ldr x11, [x3, x10, lsl #3]
ldr x12, [x1, x10, lsl #3]
and x11, x11, x19
Expand All @@ -1293,9 +1293,9 @@ zfliploop:
str x11, [x1, x10, lsl #3]
add x10, x10, #0x1
sub x11, x10, x0
cbnz x11, zfliploop
cbnz x11, edwards25519_scalarmulbase_zfliploop
subs x2, x2, #0x3a
b.hi outerloop
b.hi edwards25519_scalarmulbase_outerloop

// The final result is x = X * inv(Z), y = Y * inv(Z).
// These are the only operations in the whole computation that
Expand Down Expand Up @@ -1324,7 +1324,7 @@ zfliploop:

// The modulus p_25519 = 2^255 - 19, for the modular inverse

p_25519:
edwards25519_scalarmulbase_p_25519:
.quad 0xffffffffffffffed
.quad 0xffffffffffffffff
.quad 0xffffffffffffffff
Expand All @@ -1333,7 +1333,7 @@ p_25519:
// 0 * B = 0 and 2^251 * B in extended-projective coordinates
// but with Z = 1 assumed and hence left out, so they are (X,Y,T) only.

edwards25519_0g:
edwards25519_scalarmulbase_edwards25519_0g:

.quad 0x0000000000000000
.quad 0x0000000000000000
Expand All @@ -1350,7 +1350,7 @@ edwards25519_0g:
.quad 0x0000000000000000
.quad 0x0000000000000000

edwards25519_251g:
edwards25519_scalarmulbase_edwards25519_251g:

.quad 0x525f946d7c7220e7
.quad 0x4636b0b2f1e35444
Expand All @@ -1368,7 +1368,7 @@ edwards25519_251g:
// Precomputed table of multiples of generator for edwards25519
// all in precomputed extended-projective (y-x,x+y,2*d*x*y) triples.

edwards25519_gtable:
edwards25519_scalarmulbase_edwards25519_gtable:

// 2^0 * 1 * G

Expand Down
Loading

0 comments on commit 8e80e0c

Please sign in to comment.