Skip to content

Commit

Permalink
Squashed 'src/secp256k1/' changes from 8ab24e8da..c6b6b8f1b
Browse files Browse the repository at this point in the history
c6b6b8f1b Merge zcash#830: Rip out non-endomorphism code + dependencies
c582abade Consistency improvements to the comments
63c6b7161 Reorder comments/function around scalar_split_lambda
2edc514c9 WNAF of lambda_split output has max size 129
4232e5b7d Rip out non-endomorphism code
ebad8414b Check correctness of lambda split without -DVERIFY
fe7fc1fda Make lambda constant accessible
9d2f2b44d Add tests to exercise lambda split near bounds
9aca2f7f0 Add secp256k1_split_lambda_verify
acab934d2 Detailed comments for secp256k1_scalar_split_lambda
76ed922a5 Increase precision of g1 and g2
6173839c9 Switch to our own memcmp function
63150ab4d Merge zcash#827: Rename testrand functions to have test in name
c5257aed0 Merge zcash#821: travis: Explicitly set --with-valgrind
bb1f54280 Merge zcash#818: Add static assertion that uint32_t is unsigned int or wider
a45c1fa63 Rename testrand functions to have test in name
5006895bd Merge zcash#808: Exhaustive test improvements + exhaustive schnorrsig tests
4eecb4d6e travis: VALGRIND->RUN_VALGRIND to avoid confusion with WITH_VALGRIND
66a765c77 travis: Explicitly set --with-valgrind
d7838ba6a Merge zcash#813: Enable configuring Valgrind support
7ceb0b761 Merge zcash#819: Enable -Wundef warning
8b7dcdd95 Add exhaustive test for extrakeys and schnorrsig
08d7d8929 Make pubkey parsing test whether points are in the correct subgroup
87af00b51 Abstract out challenge computation in schnorrsig
63e1b2aa7 Disable output buffering in tests_exhaustive.c
39f67dd07 Support splitting exhaustive tests across cores
e99b26fcd Give exhaustive_tests count and seed cmdline inputs
49e6630bc refactor: move RNG seeding to testrand
b110c106f Change exhaustive test groups so they have a point with X=1
cec7b18a3 Select exhaustive lambda in function of order
78f6cdfaa Make the curve B constant a secp256k1_fe
d7f39ae4b Delete gej_is_valid_var: unused outside tests
8bcd78cd7 Make secp256k1_scalar_b32 detect overflow in scalar_low
c498366e5 Move exhaustive tests for recovery to module
be3179154 Make group order purely compile-time in exhaustive tests
e73ff3092 Enable -Wundef warning
c0041b5cf Add static assertion that uint32_t is unsigned int or wider
4ad408faf Merge zcash#782: Check if variable=yes instead of if var is set in travis.sh
412bf874d configure: Allow specifying --with[out]-valgrind explicitly
34debf7a6 Modify .travis.yml to explictly pass no in env vars instead of setting to nothing
a0e99fc12 Merge zcash#814: tests: Initialize random group elements fully
5738e8622 tests: Initialize random group elements fully
c9939ba55 Merge zcash#812: travis: run bench_schnorrsig
a51f2af62 travis: run bench_schnorrsig
ef37761fe Change travis.sh to check if variables are equal to yes instead of not-empty. Before this, setting `VALGRIND=wat` was considered as true, and to make it evaluate as false you had to unset the variable `VALGRIND=` but not it checks if `VALGRIND=yes` and if it's not `yes` then it's evaluated to false

git-subtree-dir: src/secp256k1
git-subtree-split: c6b6b8f1bb044d7d1aa065ebb674adde98a36a8e
  • Loading branch information
str4d committed Oct 21, 2020
1 parent 2fa8a70 commit 52ac872
Show file tree
Hide file tree
Showing 39 changed files with 1,586 additions and 947 deletions.
20 changes: 8 additions & 12 deletions .travis.yml
Original file line number Diff line number Diff line change
Expand Up @@ -17,33 +17,29 @@ compiler:
- gcc
env:
global:
- WIDEMUL=auto BIGNUM=auto ENDOMORPHISM=no STATICPRECOMPUTATION=yes ECMULTGENPRECISION=auto ASM=no BUILD=check EXTRAFLAGS= HOST= ECDH=no RECOVERY=no SCHNORRSIG=no EXPERIMENTAL=no CTIMETEST=yes BENCH=yes ITERS=2
- WIDEMUL=auto BIGNUM=auto STATICPRECOMPUTATION=yes ECMULTGENPRECISION=auto ASM=no BUILD=check WITH_VALGRIND=yes RUN_VALGRIND=no EXTRAFLAGS= HOST= ECDH=no RECOVERY=no SCHNORRSIG=no EXPERIMENTAL=no CTIMETEST=yes BENCH=yes ITERS=2
matrix:
- WIDEMUL=int64 RECOVERY=yes
- WIDEMUL=int64 ECDH=yes EXPERIMENTAL=yes SCHNORRSIG=yes
- WIDEMUL=int64 ENDOMORPHISM=yes
- WIDEMUL=int128
- WIDEMUL=int128 RECOVERY=yes EXPERIMENTAL=yes SCHNORRSIG=yes
- WIDEMUL=int128 ENDOMORPHISM=yes
- WIDEMUL=int128 ENDOMORPHISM=yes ECDH=yes EXPERIMENTAL=yes SCHNORRSIG=yes
- WIDEMUL=int128 ECDH=yes EXPERIMENTAL=yes SCHNORRSIG=yes
- WIDEMUL=int128 ASM=x86_64
- WIDEMUL=int128 ENDOMORPHISM=yes ASM=x86_64
- BIGNUM=no
- BIGNUM=no ENDOMORPHISM=yes RECOVERY=yes EXPERIMENTAL=yes SCHNORRSIG=yes
- BIGNUM=no RECOVERY=yes EXPERIMENTAL=yes SCHNORRSIG=yes
- BIGNUM=no STATICPRECOMPUTATION=no
- BUILD=distcheck CTIMETEST= BENCH=
- BUILD=distcheck WITH_VALGRIND=no CTIMETEST=no BENCH=no
- CPPFLAGS=-DDETERMINISTIC
- CFLAGS=-O0 CTIMETEST=
- CFLAGS=-O0 CTIMETEST=no
- ECMULTGENPRECISION=2
- ECMULTGENPRECISION=8
- VALGRIND=yes ENDOMORPHISM=yes BIGNUM=no ASM=x86_64 EXPERIMENTAL=yes ECDH=yes RECOVERY=yes EXTRAFLAGS="--disable-openssl-tests" CPPFLAGS=-DVALGRIND BUILD=
- VALGRIND=yes BIGNUM=no ASM=x86_64 EXPERIMENTAL=yes ECDH=yes RECOVERY=yes EXTRAFLAGS="--disable-openssl-tests" CPPFLAGS=-DVALGRIND BUILD=
- RUN_VALGRIND=yes BIGNUM=no ASM=x86_64 EXPERIMENTAL=yes ECDH=yes RECOVERY=yes EXTRAFLAGS="--disable-openssl-tests" BUILD=
matrix:
fast_finish: true
include:
- compiler: clang
os: linux
env: HOST=i686-linux-gnu ENDOMORPHISM=yes
env: HOST=i686-linux-gnu
addons:
apt:
packages:
Expand All @@ -63,7 +59,7 @@ matrix:
- libtool-bin
- libc6-dbg:i386
- compiler: gcc
env: HOST=i686-linux-gnu ENDOMORPHISM=yes
env: HOST=i686-linux-gnu
os: linux
addons:
apt:
Expand Down
2 changes: 1 addition & 1 deletion README.md
Original file line number Diff line number Diff line change
Expand Up @@ -48,7 +48,7 @@ Implementation details
* Use wNAF notation for point multiplicands.
* Use a much larger window for multiples of G, using precomputed multiples.
* Use Shamir's trick to do the multiplication with the public key and the generator simultaneously.
* Optionally (off by default) use secp256k1's efficiently-computable endomorphism to split the P multiplicand into 2 half-sized ones.
* Use secp256k1's efficiently-computable endomorphism to split the P multiplicand into 2 half-sized ones.
* Point multiplication for signing
* Use a precomputed table of multiples of powers of 16 multiplied with the generator, so general multiplication becomes a series of additions.
* Intended to be completely free of timing sidechannels for secret-key operations (on reasonable hardware/toolchains)
Expand Down
31 changes: 17 additions & 14 deletions configure.ac
Original file line number Diff line number Diff line change
Expand Up @@ -67,7 +67,7 @@ esac

CFLAGS="-W $CFLAGS"

warn_CFLAGS="-std=c89 -pedantic -Wall -Wextra -Wcast-align -Wnested-externs -Wshadow -Wstrict-prototypes -Wno-unused-function -Wno-long-long -Wno-overlength-strings"
warn_CFLAGS="-std=c89 -pedantic -Wall -Wextra -Wcast-align -Wnested-externs -Wshadow -Wstrict-prototypes -Wundef -Wno-unused-function -Wno-long-long -Wno-overlength-strings"
saved_CFLAGS="$CFLAGS"
CFLAGS="$warn_CFLAGS $CFLAGS"
AC_MSG_CHECKING([if ${CC} supports ${warn_CFLAGS}])
Expand Down Expand Up @@ -116,11 +116,6 @@ AC_ARG_ENABLE(exhaustive_tests,
[use_exhaustive_tests=$enableval],
[use_exhaustive_tests=yes])

AC_ARG_ENABLE(endomorphism,
AS_HELP_STRING([--enable-endomorphism],[enable endomorphism [default=no]]),
[use_endomorphism=$enableval],
[use_endomorphism=no])

AC_ARG_ENABLE(ecmult_static_precomputation,
AS_HELP_STRING([--enable-ecmult-static-precomputation],[enable precomputed ecmult table for signing [default=auto]]),
[use_ecmult_static_precomputation=$enableval],
Expand Down Expand Up @@ -164,8 +159,7 @@ AC_ARG_WITH([asm], [AS_HELP_STRING([--with-asm=x86_64|arm|no|auto],
AC_ARG_WITH([ecmult-window], [AS_HELP_STRING([--with-ecmult-window=SIZE|auto],
[window size for ecmult precomputation for verification, specified as integer in range [2..24].]
[Larger values result in possibly better performance at the cost of an exponentially larger precomputed table.]
[The table will store 2^(SIZE-2) * 64 bytes of data but can be larger in memory due to platform-specific padding and alignment.]
[If the endomorphism optimization is enabled, two tables of this size are used instead of only one.]
[The table will store 2^(SIZE-1) * 64 bytes of data but can be larger in memory due to platform-specific padding and alignment.]
["auto" is a reasonable setting for desktop machines (currently 15). [default=auto]]
)],
[req_ecmult_window=$withval], [req_ecmult_window=auto])
Expand All @@ -178,7 +172,21 @@ AC_ARG_WITH([ecmult-gen-precision], [AS_HELP_STRING([--with-ecmult-gen-precision
)],
[req_ecmult_gen_precision=$withval], [req_ecmult_gen_precision=auto])

AC_CHECK_HEADER([valgrind/memcheck.h], [enable_valgrind=yes], [enable_valgrind=no], [])
AC_ARG_WITH([valgrind], [AS_HELP_STRING([--with-valgrind=yes|no|auto],
[Build with extra checks for running inside Valgrind [default=auto]]
)],
[req_valgrind=$withval], [req_valgrind=auto])

if test x"$req_valgrind" = x"no"; then
enable_valgrind=no
else
AC_CHECK_HEADER([valgrind/memcheck.h], [enable_valgrind=yes], [
if test x"$req_valgrind" = x"yes"; then
AC_MSG_ERROR([Valgrind support explicitly requested but valgrind/memcheck.h header not available])
fi
enable_valgrind=no
], [])
fi
AM_CONDITIONAL([VALGRIND_ENABLED],[test "$enable_valgrind" = "yes"])

if test x"$enable_coverage" = x"yes"; then
Expand Down Expand Up @@ -415,10 +423,6 @@ if test x"$set_bignum" = x"gmp"; then
SECP_INCLUDES="$SECP_INCLUDES $GMP_CPPFLAGS"
fi

if test x"$use_endomorphism" = x"yes"; then
AC_DEFINE(USE_ENDOMORPHISM, 1, [Define this symbol to use endomorphism optimization])
fi

if test x"$set_precomp" = x"yes"; then
AC_DEFINE(USE_ECMULT_STATIC_PRECOMPUTATION, 1, [Define this symbol to use a statically generated ecmult table])
fi
Expand Down Expand Up @@ -500,7 +504,6 @@ AC_OUTPUT

echo
echo "Build Options:"
echo " with endomorphism = $use_endomorphism"
echo " with ecmult precomp = $set_precomp"
echo " with external callbacks = $use_external_default_callbacks"
echo " with benchmarks = $use_benchmark"
Expand Down
15 changes: 10 additions & 5 deletions contrib/travis.sh
Original file line number Diff line number Diff line change
Expand Up @@ -13,27 +13,28 @@ then
fi

./configure \
--enable-experimental="$EXPERIMENTAL" --enable-endomorphism="$ENDOMORPHISM" \
--enable-experimental="$EXPERIMENTAL" \
--with-test-override-wide-multiply="$WIDEMUL" --with-bignum="$BIGNUM" --with-asm="$ASM" \
--enable-ecmult-static-precomputation="$STATICPRECOMPUTATION" --with-ecmult-gen-precision="$ECMULTGENPRECISION" \
--enable-module-ecdh="$ECDH" --enable-module-recovery="$RECOVERY" \
--enable-module-schnorrsig="$SCHNORRSIG" \
--with-valgrind="$WITH_VALGRIND" \
--host="$HOST" $EXTRAFLAGS

if [ -n "$BUILD" ]
then
make -j2 "$BUILD"
fi
if [ -n "$VALGRIND" ]
if [ "$RUN_VALGRIND" = "yes" ]
then
make -j2
# the `--error-exitcode` is required to make the test fail if valgrind found errors, otherwise it'll return 0 (http://valgrind.org/docs/manual/manual-core.html)
valgrind --error-exitcode=42 ./tests 16
valgrind --error-exitcode=42 ./exhaustive_tests
fi
if [ -n "$BENCH" ]
if [ "$BENCH" = "yes" ]
then
if [ -n "$VALGRIND" ]
if [ "$RUN_VALGRIND" = "yes" ]
then
# Using the local `libtool` because on macOS the system's libtool has nothing to do with GNU libtool
EXEC='./libtool --mode=execute valgrind --error-exitcode=42'
Expand All @@ -56,8 +57,12 @@ then
then
$EXEC ./bench_ecdh >> bench.log 2>&1
fi
if [ "$SCHNORRSIG" = "yes" ]
then
$EXEC ./bench_schnorrsig >> bench.log 2>&1
fi
fi
if [ -n "$CTIMETEST" ]
if [ "$CTIMETEST" = "yes" ]
then
./libtool --mode=execute valgrind --error-exitcode=42 ./valgrind_ctime_test > valgrind_ctime_test.log 2>&1
fi
129 changes: 129 additions & 0 deletions sage/gen_exhaustive_groups.sage
Original file line number Diff line number Diff line change
@@ -0,0 +1,129 @@
# Define field size and field
P = 2^256 - 2^32 - 977
F = GF(P)
BETA = F(0x7ae96a2b657c07106e64479eac3434e99cf0497512f58995c1396c28719501ee)

assert(BETA != F(1) and BETA^3 == F(1))

orders_done = set()
results = {}
first = True
for b in range(1, P):
# There are only 6 curves (up to isomorphism) of the form y^2=x^3+B. Stop once we have tried all.
if len(orders_done) == 6:
break

E = EllipticCurve(F, [0, b])
print("Analyzing curve y^2 = x^3 + %i" % b)
n = E.order()
# Skip curves with an order we've already tried
if n in orders_done:
print("- Isomorphic to earlier curve")
continue
orders_done.add(n)
# Skip curves isomorphic to the real secp256k1
if n.is_pseudoprime():
print(" - Isomorphic to secp256k1")
continue

print("- Finding subgroups")

# Find what prime subgroups exist
for f, _ in n.factor():
print("- Analyzing subgroup of order %i" % f)
# Skip subgroups of order >1000
if f < 4 or f > 1000:
print(" - Bad size")
continue

# Iterate over X coordinates until we find one that is on the curve, has order f,
# and for which curve isomorphism exists that maps it to X coordinate 1.
for x in range(1, P):
# Skip X coordinates not on the curve, and construct the full point otherwise.
if not E.is_x_coord(x):
continue
G = E.lift_x(F(x))

print(" - Analyzing (multiples of) point with X=%i" % x)

# Skip points whose order is not a multiple of f. Project the point to have
# order f otherwise.
if (G.order() % f):
print(" - Bad order")
continue
G = G * (G.order() // f)

# Find lambda for endomorphism. Skip if none can be found.
lam = None
for l in Integers(f)(1).nth_root(3, all=True):
if int(l)*G == E(BETA*G[0], G[1]):
lam = int(l)
break
if lam is None:
print(" - No endomorphism for this subgroup")
break

# Now look for an isomorphism of the curve that gives this point an X
# coordinate equal to 1.
# If (x,y) is on y^2 = x^3 + b, then (a^2*x, a^3*y) is on y^2 = x^3 + a^6*b.
# So look for m=a^2=1/x.
m = F(1)/G[0]
if not m.is_square():
print(" - No curve isomorphism maps it to a point with X=1")
continue
a = m.sqrt()
rb = a^6*b
RE = EllipticCurve(F, [0, rb])

# Use as generator twice the image of G under the above isormorphism.
# This means that generator*(1/2 mod f) will have X coordinate 1.
RG = RE(1, a^3*G[1]) * 2
# And even Y coordinate.
if int(RG[1]) % 2:
RG = -RG
assert(RG.order() == f)
assert(lam*RG == RE(BETA*RG[0], RG[1]))

# We have found curve RE:y^2=x^3+rb with generator RG of order f. Remember it
results[f] = {"b": rb, "G": RG, "lambda": lam}
print(" - Found solution")
break

print("")

print("")
print("")
print("/* To be put in src/group_impl.h: */")
first = True
for f in sorted(results.keys()):
b = results[f]["b"]
G = results[f]["G"]
print("# %s EXHAUSTIVE_TEST_ORDER == %i" % ("if" if first else "elif", f))
first = False
print("static const secp256k1_ge secp256k1_ge_const_g = SECP256K1_GE_CONST(")
print(" 0x%08x, 0x%08x, 0x%08x, 0x%08x," % tuple((int(G[0]) >> (32 * (7 - i))) & 0xffffffff for i in range(4)))
print(" 0x%08x, 0x%08x, 0x%08x, 0x%08x," % tuple((int(G[0]) >> (32 * (7 - i))) & 0xffffffff for i in range(4, 8)))
print(" 0x%08x, 0x%08x, 0x%08x, 0x%08x," % tuple((int(G[1]) >> (32 * (7 - i))) & 0xffffffff for i in range(4)))
print(" 0x%08x, 0x%08x, 0x%08x, 0x%08x" % tuple((int(G[1]) >> (32 * (7 - i))) & 0xffffffff for i in range(4, 8)))
print(");")
print("static const secp256k1_fe secp256k1_fe_const_b = SECP256K1_FE_CONST(")
print(" 0x%08x, 0x%08x, 0x%08x, 0x%08x," % tuple((int(b) >> (32 * (7 - i))) & 0xffffffff for i in range(4)))
print(" 0x%08x, 0x%08x, 0x%08x, 0x%08x" % tuple((int(b) >> (32 * (7 - i))) & 0xffffffff for i in range(4, 8)))
print(");")
print("# else")
print("# error No known generator for the specified exhaustive test group order.")
print("# endif")

print("")
print("")
print("/* To be put in src/scalar_impl.h: */")
first = True
for f in sorted(results.keys()):
lam = results[f]["lambda"]
print("# %s EXHAUSTIVE_TEST_ORDER == %i" % ("if" if first else "elif", f))
first = False
print("# define EXHAUSTIVE_TEST_LAMBDA %i" % lam)
print("# else")
print("# error No known lambda for the specified exhaustive test group order.")
print("# endif")
print("")
8 changes: 7 additions & 1 deletion src/assumptions.h
Original file line number Diff line number Diff line change
Expand Up @@ -7,6 +7,8 @@
#ifndef SECP256K1_ASSUMPTIONS_H
#define SECP256K1_ASSUMPTIONS_H

#include <limits.h>

#include "util.h"

/* This library, like most software, relies on a number of compiler implementation defined (but not undefined)
Expand All @@ -19,7 +21,11 @@ struct secp256k1_assumption_checker {
allowed. */
int dummy_array[(
/* Bytes are 8 bits. */
CHAR_BIT == 8 &&
(CHAR_BIT == 8) &&

/* No integer promotion for uint32_t. This ensures that we can multiply uintXX_t values where XX >= 32
without signed overflow, which would be undefined behaviour. */
(UINT_MAX <= UINT32_MAX) &&

/* Conversions from unsigned to signed outside of the bounds of the signed type are
implementation-defined. Verify that they function as reinterpreting the lower
Expand Down
1 change: 0 additions & 1 deletion src/basic-config.h
Original file line number Diff line number Diff line change
Expand Up @@ -11,7 +11,6 @@

#undef USE_ASM_X86_64
#undef USE_ECMULT_STATIC_PRECOMPUTATION
#undef USE_ENDOMORPHISM
#undef USE_EXTERNAL_ASM
#undef USE_EXTERNAL_DEFAULT_CALLBACKS
#undef USE_FIELD_INV_BUILTIN
Expand Down
4 changes: 0 additions & 4 deletions src/bench_internal.c
Original file line number Diff line number Diff line change
Expand Up @@ -117,7 +117,6 @@ void bench_scalar_mul(void* arg, int iters) {
}
}

#ifdef USE_ENDOMORPHISM
void bench_scalar_split(void* arg, int iters) {
int i, j = 0;
bench_inv *data = (bench_inv*)arg;
Expand All @@ -128,7 +127,6 @@ void bench_scalar_split(void* arg, int iters) {
}
CHECK(j <= iters);
}
#endif

void bench_scalar_inverse(void* arg, int iters) {
int i, j = 0;
Expand Down Expand Up @@ -397,9 +395,7 @@ int main(int argc, char **argv) {
if (have_flag(argc, argv, "scalar") || have_flag(argc, argv, "negate")) run_benchmark("scalar_negate", bench_scalar_negate, bench_setup, NULL, &data, 10, iters*100);
if (have_flag(argc, argv, "scalar") || have_flag(argc, argv, "sqr")) run_benchmark("scalar_sqr", bench_scalar_sqr, bench_setup, NULL, &data, 10, iters*10);
if (have_flag(argc, argv, "scalar") || have_flag(argc, argv, "mul")) run_benchmark("scalar_mul", bench_scalar_mul, bench_setup, NULL, &data, 10, iters*10);
#ifdef USE_ENDOMORPHISM
if (have_flag(argc, argv, "scalar") || have_flag(argc, argv, "split")) run_benchmark("scalar_split", bench_scalar_split, bench_setup, NULL, &data, 10, iters);
#endif
if (have_flag(argc, argv, "scalar") || have_flag(argc, argv, "inverse")) run_benchmark("scalar_inverse", bench_scalar_inverse, bench_setup, NULL, &data, 10, 2000);
if (have_flag(argc, argv, "scalar") || have_flag(argc, argv, "inverse")) run_benchmark("scalar_inverse_var", bench_scalar_inverse_var, bench_setup, NULL, &data, 10, 2000);

Expand Down
2 changes: 0 additions & 2 deletions src/ecmult.h
Original file line number Diff line number Diff line change
Expand Up @@ -15,9 +15,7 @@
typedef struct {
/* For accelerating the computation of a*P + b*G: */
secp256k1_ge_storage (*pre_g)[]; /* odd multiples of the generator */
#ifdef USE_ENDOMORPHISM
secp256k1_ge_storage (*pre_g_128)[]; /* odd multiples of 2^128*generator */
#endif
} secp256k1_ecmult_context;

static const size_t SECP256K1_ECMULT_CONTEXT_PREALLOCATED_SIZE;
Expand Down
Loading

0 comments on commit 52ac872

Please sign in to comment.