aboutsummaryrefslogtreecommitdiff
path: root/src/Specific/CurveParameters
diff options
context:
space:
mode:
authorGravatar Jason Gross <jgross@mit.edu>2018-01-10 18:12:14 -0500
committerGravatar Andres Erbsen <andreser@mit.edu>2018-01-15 04:16:50 -0500
commitd50c368e0cf5c25520fdd9e57695b43640c77fe0 (patch)
tree786b6e1ef822a7b4cd42dd6f72ab8ad98d270e3b /src/Specific/CurveParameters
parent36e2e828be55cb58f985208de0ef7085cedecdd2 (diff)
Add x25519 donna versions with the new way of generating C code
Diffstat (limited to 'src/Specific/CurveParameters')
-rwxr-xr-xsrc/Specific/CurveParameters/remake_curves.sh2
-rw-r--r--src/Specific/CurveParameters/solinas32_2e255m19_10limbs_donna.json237
-rw-r--r--src/Specific/CurveParameters/solinas64_2e255m19_5limbs_donna.json80
3 files changed, 319 insertions, 0 deletions
diff --git a/src/Specific/CurveParameters/remake_curves.sh b/src/Specific/CurveParameters/remake_curves.sh
index bb51658cf..1a5c578b7 100755
--- a/src/Specific/CurveParameters/remake_curves.sh
+++ b/src/Specific/CurveParameters/remake_curves.sh
@@ -11,6 +11,8 @@ ${MAKE} "$@" x25519_c32.json ../X25519/C32/
${MAKE} "$@" x2448_c64_karatsuba.json ../X2448/Karatsuba/C64/
${MAKE} "$@" nistp256_amd128.json ../NISTP256/AMD128/
${MAKE} "$@" nistp256_amd64.json ../NISTP256/AMD64/
+${MAKE} "$@" solinas32_2e255m19_10limbs_donna.json ../solinas32_2e255m19_10limbs_donna/
+${MAKE} "$@" solinas64_2e255m19_5limbs_donna.json ../solinas64_2e255m19_5limbs_donna/
${MAKE} "$@" montgomery32_2e127m1_4limbs.json ../montgomery32_2e127m1_4limbs/
${MAKE} "$@" montgomery64_2e127m1_2limbs.json ../montgomery64_2e127m1_2limbs/
${MAKE} "$@" solinas32_2e127m1_5limbs.json ../solinas32_2e127m1_5limbs/
diff --git a/src/Specific/CurveParameters/solinas32_2e255m19_10limbs_donna.json b/src/Specific/CurveParameters/solinas32_2e255m19_10limbs_donna.json
new file mode 100644
index 000000000..9e80be3db
--- /dev/null
+++ b/src/Specific/CurveParameters/solinas32_2e255m19_10limbs_donna.json
@@ -0,0 +1,237 @@
+{
+ "base": "25.5",
+ "bitwidth": 32,
+ "carry_chains": "default",
+ "coef_div_modulus": "2",
+ "compiler": "clang -fbracket-depth=999999 -march=native -mtune=native -std=gnu11 -O3 -flto -fuse-ld=lld -fomit-frame-pointer -fwrapv -Wno-attributes -fno-strict-aliasing -Da24_hex='0x3039' -Da24_val='12345' -Da_minus_two_over_four_array='{0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x30,0x39}' -Dbitwidth='32' -Dlimb_weight_gaps_array='{26,25,26,25,26,25,26,25,26,25}' -Dmodulus_array='{0x7f,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xed}' -Dmodulus_bytes_val='32' -Dmodulus_limbs='10' -Dq_mpz='(1_mpz<<255) - 19'",
+ "compilerxx": "clang++ -fbracket-depth=999999 -march=native -mtune=native -std=gnu++11 -O3 -flto -fuse-ld=lld -fomit-frame-pointer -fwrapv -Wno-attributes -fno-strict-aliasing -Da24_hex='0x3039' -Da24_val='12345' -Da_minus_two_over_four_array='{0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x30,0x39}' -Dbitwidth='32' -Dlimb_weight_gaps_array='{26,25,26,25,26,25,26,25,26,25}' -Dmodulus_array='{0x7f,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xed}' -Dmodulus_bytes_val='32' -Dmodulus_limbs='10' -Dq_mpz='(1_mpz<<255) - 19'",
+ "modulus": "2^255 - 19",
+ "operations": [
+ "femul",
+ "feadd",
+ "fesub",
+ "fesquare",
+ "fecarry",
+ "freeze"
+ ],
+ "sz": "10",
+ "mul_header" : "(* Micro-optimized form from curve25519-donna by Adam Langley (Google) and Daniel Bernstein. See <https://github.com/agl/curve25519-donna/blob/master/LICENSE.md>. *)",
+ "mul_code"
+ :
+ "
+ output[0] = ((limb) ((s32) in2[0])) * ((s32) in[0]);
+ output[1] = ((limb) ((s32) in2[0])) * ((s32) in[1]) +
+ ((limb) ((s32) in2[1])) * ((s32) in[0]);
+ output[2] = 2 * ((limb) ((s32) in2[1])) * ((s32) in[1]) +
+ ((limb) ((s32) in2[0])) * ((s32) in[2]) +
+ ((limb) ((s32) in2[2])) * ((s32) in[0]);
+ output[3] = ((limb) ((s32) in2[1])) * ((s32) in[2]) +
+ ((limb) ((s32) in2[2])) * ((s32) in[1]) +
+ ((limb) ((s32) in2[0])) * ((s32) in[3]) +
+ ((limb) ((s32) in2[3])) * ((s32) in[0]);
+ output[4] = ((limb) ((s32) in2[2])) * ((s32) in[2]) +
+ 2 * (((limb) ((s32) in2[1])) * ((s32) in[3]) +
+ ((limb) ((s32) in2[3])) * ((s32) in[1])) +
+ ((limb) ((s32) in2[0])) * ((s32) in[4]) +
+ ((limb) ((s32) in2[4])) * ((s32) in[0]);
+ output[5] = ((limb) ((s32) in2[2])) * ((s32) in[3]) +
+ ((limb) ((s32) in2[3])) * ((s32) in[2]) +
+ ((limb) ((s32) in2[1])) * ((s32) in[4]) +
+ ((limb) ((s32) in2[4])) * ((s32) in[1]) +
+ ((limb) ((s32) in2[0])) * ((s32) in[5]) +
+ ((limb) ((s32) in2[5])) * ((s32) in[0]);
+ output[6] = 2 * (((limb) ((s32) in2[3])) * ((s32) in[3]) +
+ ((limb) ((s32) in2[1])) * ((s32) in[5]) +
+ ((limb) ((s32) in2[5])) * ((s32) in[1])) +
+ ((limb) ((s32) in2[2])) * ((s32) in[4]) +
+ ((limb) ((s32) in2[4])) * ((s32) in[2]) +
+ ((limb) ((s32) in2[0])) * ((s32) in[6]) +
+ ((limb) ((s32) in2[6])) * ((s32) in[0]);
+ output[7] = ((limb) ((s32) in2[3])) * ((s32) in[4]) +
+ ((limb) ((s32) in2[4])) * ((s32) in[3]) +
+ ((limb) ((s32) in2[2])) * ((s32) in[5]) +
+ ((limb) ((s32) in2[5])) * ((s32) in[2]) +
+ ((limb) ((s32) in2[1])) * ((s32) in[6]) +
+ ((limb) ((s32) in2[6])) * ((s32) in[1]) +
+ ((limb) ((s32) in2[0])) * ((s32) in[7]) +
+ ((limb) ((s32) in2[7])) * ((s32) in[0]);
+ output[8] = ((limb) ((s32) in2[4])) * ((s32) in[4]) +
+ 2 * (((limb) ((s32) in2[3])) * ((s32) in[5]) +
+ ((limb) ((s32) in2[5])) * ((s32) in[3]) +
+ ((limb) ((s32) in2[1])) * ((s32) in[7]) +
+ ((limb) ((s32) in2[7])) * ((s32) in[1])) +
+ ((limb) ((s32) in2[2])) * ((s32) in[6]) +
+ ((limb) ((s32) in2[6])) * ((s32) in[2]) +
+ ((limb) ((s32) in2[0])) * ((s32) in[8]) +
+ ((limb) ((s32) in2[8])) * ((s32) in[0]);
+ output[9] = ((limb) ((s32) in2[4])) * ((s32) in[5]) +
+ ((limb) ((s32) in2[5])) * ((s32) in[4]) +
+ ((limb) ((s32) in2[3])) * ((s32) in[6]) +
+ ((limb) ((s32) in2[6])) * ((s32) in[3]) +
+ ((limb) ((s32) in2[2])) * ((s32) in[7]) +
+ ((limb) ((s32) in2[7])) * ((s32) in[2]) +
+ ((limb) ((s32) in2[1])) * ((s32) in[8]) +
+ ((limb) ((s32) in2[8])) * ((s32) in[1]) +
+ ((limb) ((s32) in2[0])) * ((s32) in[9]) +
+ ((limb) ((s32) in2[9])) * ((s32) in[0]);
+ output[10] = 2 * (((limb) ((s32) in2[5])) * ((s32) in[5]) +
+ ((limb) ((s32) in2[3])) * ((s32) in[7]) +
+ ((limb) ((s32) in2[7])) * ((s32) in[3]) +
+ ((limb) ((s32) in2[1])) * ((s32) in[9]) +
+ ((limb) ((s32) in2[9])) * ((s32) in[1])) +
+ ((limb) ((s32) in2[4])) * ((s32) in[6]) +
+ ((limb) ((s32) in2[6])) * ((s32) in[4]) +
+ ((limb) ((s32) in2[2])) * ((s32) in[8]) +
+ ((limb) ((s32) in2[8])) * ((s32) in[2]);
+ output[11] = ((limb) ((s32) in2[5])) * ((s32) in[6]) +
+ ((limb) ((s32) in2[6])) * ((s32) in[5]) +
+ ((limb) ((s32) in2[4])) * ((s32) in[7]) +
+ ((limb) ((s32) in2[7])) * ((s32) in[4]) +
+ ((limb) ((s32) in2[3])) * ((s32) in[8]) +
+ ((limb) ((s32) in2[8])) * ((s32) in[3]) +
+ ((limb) ((s32) in2[2])) * ((s32) in[9]) +
+ ((limb) ((s32) in2[9])) * ((s32) in[2]);
+ output[12] = ((limb) ((s32) in2[6])) * ((s32) in[6]) +
+ 2 * (((limb) ((s32) in2[5])) * ((s32) in[7]) +
+ ((limb) ((s32) in2[7])) * ((s32) in[5]) +
+ ((limb) ((s32) in2[3])) * ((s32) in[9]) +
+ ((limb) ((s32) in2[9])) * ((s32) in[3])) +
+ ((limb) ((s32) in2[4])) * ((s32) in[8]) +
+ ((limb) ((s32) in2[8])) * ((s32) in[4]);
+ output[13] = ((limb) ((s32) in2[6])) * ((s32) in[7]) +
+ ((limb) ((s32) in2[7])) * ((s32) in[6]) +
+ ((limb) ((s32) in2[5])) * ((s32) in[8]) +
+ ((limb) ((s32) in2[8])) * ((s32) in[5]) +
+ ((limb) ((s32) in2[4])) * ((s32) in[9]) +
+ ((limb) ((s32) in2[9])) * ((s32) in[4]);
+ output[14] = 2 * (((limb) ((s32) in2[7])) * ((s32) in[7]) +
+ ((limb) ((s32) in2[5])) * ((s32) in[9]) +
+ ((limb) ((s32) in2[9])) * ((s32) in[5])) +
+ ((limb) ((s32) in2[6])) * ((s32) in[8]) +
+ ((limb) ((s32) in2[8])) * ((s32) in[6]);
+ output[15] = ((limb) ((s32) in2[7])) * ((s32) in[8]) +
+ ((limb) ((s32) in2[8])) * ((s32) in[7]) +
+ ((limb) ((s32) in2[6])) * ((s32) in[9]) +
+ ((limb) ((s32) in2[9])) * ((s32) in[6]);
+ output[16] = ((limb) ((s32) in2[8])) * ((s32) in[8]) +
+ 2 * (((limb) ((s32) in2[7])) * ((s32) in[9]) +
+ ((limb) ((s32) in2[9])) * ((s32) in[7]));
+ output[17] = ((limb) ((s32) in2[8])) * ((s32) in[9]) +
+ ((limb) ((s32) in2[9])) * ((s32) in[8]);
+ output[18] = 2 * ((limb) ((s32) in2[9])) * ((s32) in[9]);
+ output[8] += output[18] << 4;
+ output[8] += output[18] << 1;
+ output[8] += output[18];
+ output[7] += output[17] << 4;
+ output[7] += output[17] << 1;
+ output[7] += output[17];
+ output[6] += output[16] << 4;
+ output[6] += output[16] << 1;
+ output[6] += output[16];
+ output[5] += output[15] << 4;
+ output[5] += output[15] << 1;
+ output[5] += output[15];
+ output[4] += output[14] << 4;
+ output[4] += output[14] << 1;
+ output[4] += output[14];
+ output[3] += output[13] << 4;
+ output[3] += output[13] << 1;
+ output[3] += output[13];
+ output[2] += output[12] << 4;
+ output[2] += output[12] << 1;
+ output[2] += output[12];
+ output[1] += output[11] << 4;
+ output[1] += output[11] << 1;
+ output[1] += output[11];
+ output[0] += output[10] << 4;
+ output[0] += output[10] << 1;
+ output[0] += output[10];
+",
+ "square_header" : "(* Micro-optimized form from curve25519-donna by Adam Langley (Google) and Daniel Bernstein. See <https://github.com/agl/curve25519-donna/blob/master/LICENSE.md>. *)",
+ "square_code"
+ :
+ "
+ output[0] = ((limb) ((s32) in[0])) * ((s32) in[0]);
+ output[1] = 2 * ((limb) ((s32) in[0])) * ((s32) in[1]);
+ output[2] = 2 * (((limb) ((s32) in[1])) * ((s32) in[1]) +
+ ((limb) ((s32) in[0])) * ((s32) in[2]));
+ output[3] = 2 * (((limb) ((s32) in[1])) * ((s32) in[2]) +
+ ((limb) ((s32) in[0])) * ((s32) in[3]));
+ output[4] = ((limb) ((s32) in[2])) * ((s32) in[2]) +
+ 4 * ((limb) ((s32) in[1])) * ((s32) in[3]) +
+ 2 * ((limb) ((s32) in[0])) * ((s32) in[4]);
+ output[5] = 2 * (((limb) ((s32) in[2])) * ((s32) in[3]) +
+ ((limb) ((s32) in[1])) * ((s32) in[4]) +
+ ((limb) ((s32) in[0])) * ((s32) in[5]));
+ output[6] = 2 * (((limb) ((s32) in[3])) * ((s32) in[3]) +
+ ((limb) ((s32) in[2])) * ((s32) in[4]) +
+ ((limb) ((s32) in[0])) * ((s32) in[6]) +
+ 2 * ((limb) ((s32) in[1])) * ((s32) in[5]));
+ output[7] = 2 * (((limb) ((s32) in[3])) * ((s32) in[4]) +
+ ((limb) ((s32) in[2])) * ((s32) in[5]) +
+ ((limb) ((s32) in[1])) * ((s32) in[6]) +
+ ((limb) ((s32) in[0])) * ((s32) in[7]));
+ output[8] = ((limb) ((s32) in[4])) * ((s32) in[4]) +
+ 2 * (((limb) ((s32) in[2])) * ((s32) in[6]) +
+ ((limb) ((s32) in[0])) * ((s32) in[8]) +
+ 2 * (((limb) ((s32) in[1])) * ((s32) in[7]) +
+ ((limb) ((s32) in[3])) * ((s32) in[5])));
+ output[9] = 2 * (((limb) ((s32) in[4])) * ((s32) in[5]) +
+ ((limb) ((s32) in[3])) * ((s32) in[6]) +
+ ((limb) ((s32) in[2])) * ((s32) in[7]) +
+ ((limb) ((s32) in[1])) * ((s32) in[8]) +
+ ((limb) ((s32) in[0])) * ((s32) in[9]));
+ output[10] = 2 * (((limb) ((s32) in[5])) * ((s32) in[5]) +
+ ((limb) ((s32) in[4])) * ((s32) in[6]) +
+ ((limb) ((s32) in[2])) * ((s32) in[8]) +
+ 2 * (((limb) ((s32) in[3])) * ((s32) in[7]) +
+ ((limb) ((s32) in[1])) * ((s32) in[9])));
+ output[11] = 2 * (((limb) ((s32) in[5])) * ((s32) in[6]) +
+ ((limb) ((s32) in[4])) * ((s32) in[7]) +
+ ((limb) ((s32) in[3])) * ((s32) in[8]) +
+ ((limb) ((s32) in[2])) * ((s32) in[9]));
+ output[12] = ((limb) ((s32) in[6])) * ((s32) in[6]) +
+ 2 * (((limb) ((s32) in[4])) * ((s32) in[8]) +
+ 2 * (((limb) ((s32) in[5])) * ((s32) in[7]) +
+ ((limb) ((s32) in[3])) * ((s32) in[9])));
+ output[13] = 2 * (((limb) ((s32) in[6])) * ((s32) in[7]) +
+ ((limb) ((s32) in[5])) * ((s32) in[8]) +
+ ((limb) ((s32) in[4])) * ((s32) in[9]));
+ output[14] = 2 * (((limb) ((s32) in[7])) * ((s32) in[7]) +
+ ((limb) ((s32) in[6])) * ((s32) in[8]) +
+ 2 * ((limb) ((s32) in[5])) * ((s32) in[9]));
+ output[15] = 2 * (((limb) ((s32) in[7])) * ((s32) in[8]) +
+ ((limb) ((s32) in[6])) * ((s32) in[9]));
+ output[16] = ((limb) ((s32) in[8])) * ((s32) in[8]) +
+ 4 * ((limb) ((s32) in[7])) * ((s32) in[9]);
+ output[17] = 2 * ((limb) ((s32) in[8])) * ((s32) in[9]);
+ output[18] = 2 * ((limb) ((s32) in[9])) * ((s32) in[9]);
+ output[8] += output[18] << 4;
+ output[8] += output[18] << 1;
+ output[8] += output[18];
+ output[7] += output[17] << 4;
+ output[7] += output[17] << 1;
+ output[7] += output[17];
+ output[6] += output[16] << 4;
+ output[6] += output[16] << 1;
+ output[6] += output[16];
+ output[5] += output[15] << 4;
+ output[5] += output[15] << 1;
+ output[5] += output[15];
+ output[4] += output[14] << 4;
+ output[4] += output[14] << 1;
+ output[4] += output[14];
+ output[3] += output[13] << 4;
+ output[3] += output[13] << 1;
+ output[3] += output[13];
+ output[2] += output[12] << 4;
+ output[2] += output[12] << 1;
+ output[2] += output[12];
+ output[1] += output[11] << 4;
+ output[1] += output[11] << 1;
+ output[1] += output[11];
+ output[0] += output[10] << 4;
+ output[0] += output[10] << 1;
+ output[0] += output[10];
+"
+}
diff --git a/src/Specific/CurveParameters/solinas64_2e255m19_5limbs_donna.json b/src/Specific/CurveParameters/solinas64_2e255m19_5limbs_donna.json
new file mode 100644
index 000000000..68a2032c2
--- /dev/null
+++ b/src/Specific/CurveParameters/solinas64_2e255m19_5limbs_donna.json
@@ -0,0 +1,80 @@
+{
+ "base": "51",
+ "bitwidth": 64,
+ "carry_chains": "default",
+ "coef_div_modulus": "2",
+ "compiler": "clang -fbracket-depth=999999 -march=native -mtune=native -std=gnu11 -O3 -flto -fuse-ld=lld -fomit-frame-pointer -fwrapv -Wno-attributes -fno-strict-aliasing -Da24_hex='0x3039' -Da24_val='12345' -Da_minus_two_over_four_array='{0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x30,0x39}' -Dbitwidth='64' -Dlimb_weight_gaps_array='{51,51,51,51,51}' -Dmodulus_array='{0x7f,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xed}' -Dmodulus_bytes_val='32' -Dmodulus_limbs='5' -Dq_mpz='(1_mpz<<255) - 19'",
+ "compilerxx": "clang++ -fbracket-depth=999999 -march=native -mtune=native -std=gnu++11 -O3 -flto -fuse-ld=lld -fomit-frame-pointer -fwrapv -Wno-attributes -fno-strict-aliasing -Da24_hex='0x3039' -Da24_val='12345' -Da_minus_two_over_four_array='{0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x30,0x39}' -Dbitwidth='64' -Dlimb_weight_gaps_array='{51,51,51,51,51}' -Dmodulus_array='{0x7f,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xed}' -Dmodulus_bytes_val='32' -Dmodulus_limbs='5' -Dq_mpz='(1_mpz<<255) - 19'",
+ "modulus": "2^255 - 19",
+ "operations": [
+ "femul",
+ "feadd",
+ "fesub",
+ "fesquare",
+ "fecarry",
+ "freeze"
+ ],
+ "sz": "5",
+ "mul_header" : "(* Micro-optimized form from curve25519-donna-c64 by Adam Langley (Google) and Daniel Bernstein. See <https://github.com/agl/curve25519-donna/blob/master/LICENSE.md>. *)",
+ "mul_code"
+ :
+ "
+ uint128_t t[5];
+ limb r0,r1,r2,r3,r4,s0,s1,s2,s3,s4,c;
+
+ r0 = in[0];
+ r1 = in[1];
+ r2 = in[2];
+ r3 = in[3];
+ r4 = in[4];
+
+ s0 = in2[0];
+ s1 = in2[1];
+ s2 = in2[2];
+ s3 = in2[3];
+ s4 = in2[4];
+
+ t[0] = ((uint128_t) r0) * s0;
+ t[1] = ((uint128_t) r0) * s1 + ((uint128_t) r1) * s0;
+ t[2] = ((uint128_t) r0) * s2 + ((uint128_t) r2) * s0 + ((uint128_t) r1) * s1;
+ t[3] = ((uint128_t) r0) * s3 + ((uint128_t) r3) * s0 + ((uint128_t) r1) * s2 + ((uint128_t) r2) * s1;
+ t[4] = ((uint128_t) r0) * s4 + ((uint128_t) r4) * s0 + ((uint128_t) r3) * s1 + ((uint128_t) r1) * s3 + ((uint128_t) r2) * s2;
+
+ r4 *= 19;
+ r1 *= 19;
+ r2 *= 19;
+ r3 *= 19;
+
+ t[0] += ((uint128_t) r4) * s1 + ((uint128_t) r1) * s4 + ((uint128_t) r2) * s3 + ((uint128_t) r3) * s2;
+ t[1] += ((uint128_t) r4) * s2 + ((uint128_t) r2) * s4 + ((uint128_t) r3) * s3;
+ t[2] += ((uint128_t) r4) * s3 + ((uint128_t) r3) * s4;
+ t[3] += ((uint128_t) r4) * s4;
+",
+ "square_header" : "(* Micro-optimized form from curve25519-donna-c64 by Adam Langley (Google) and Daniel Bernstein. See <https://github.com/agl/curve25519-donna/blob/master/LICENSE.md>. *)",
+ "square_code"
+ :
+ "
+ uint128_t t[5];
+ limb r0,r1,r2,r3,r4,c;
+ limb d0,d1,d2,d4,d419;
+
+ r0 = in[0];
+ r1 = in[1];
+ r2 = in[2];
+ r3 = in[3];
+ r4 = in[4];
+
+ do {
+ d0 = r0 * 2;
+ d1 = r1 * 2;
+ d2 = r2 * 2 * 19;
+ d419 = r4 * 19;
+ d4 = d419 * 2;
+
+ t[0] = ((uint128_t) r0) * r0 + ((uint128_t) d4) * r1 + (((uint128_t) d2) * (r3 ));
+ t[1] = ((uint128_t) d0) * r1 + ((uint128_t) d4) * r2 + (((uint128_t) r3) * (r3 * 19));
+ t[2] = ((uint128_t) d0) * r2 + ((uint128_t) r1) * r1 + (((uint128_t) d4) * (r3 ));
+ t[3] = ((uint128_t) d0) * r3 + ((uint128_t) d1) * r2 + (((uint128_t) r4) * (d419 ));
+ t[4] = ((uint128_t) d0) * r4 + ((uint128_t) d1) * r3 + (((uint128_t) r2) * (r2 ));
+"
+}