Squashed 'src/secp256k1/' changes from 6c527ec..7a49cac

7a49cac Merge #410: Add string.h include to ecmult_impl
0bbd5d4 Add string.h include to ecmult_impl
c5b32e1 Merge #405: Make secp256k1_fe_sqrt constant time
926836a Make secp256k1_fe_sqrt constant time
e2a8e92 Merge #404: Replace 3M + 4S doubling formula with 2M + 5S one
8ec49d8 Add note about 2M + 5S doubling formula
5a91bd7 Merge #400: A couple minor cleanups
ac01378 build: add -DSECP256K1_BUILD to benchmark_internal build flags
a6c6f99 Remove a bunch of unused stdlib #includes
65285a6 Merge #403: configure: add flag to disable OpenSSL tests
a9b2a5d configure: add flag to disable OpenSSL tests
b340123 Merge #402: Add support for testing quadratic residues
e6e9805 Add function for testing quadratic residue field/group elements.
efd953a Add Jacobi symbol test via GMP
fa36a0d Merge #401: ecmult_const: unify endomorphism and non-endomorphism skew cases
c6191fd ecmult_const: unify endomorphism and non-endomorphism skew cases
0b3e618 Merge #378: .gitignore build-aux cleanup
6042217 Merge #384: JNI: align shared files copyright/comments to bitcoinj's
24ad20f Merge #399: build: verify that the native compiler works for static precomp
b3be852 Merge #398: Test whether ECDH and Schnorr are enabled for JNI
aa0b1fd build: verify that the native compiler works for static precomp
eee808d Test whether ECDH and Schnorr are enabled for JNI
7b0fb18 Merge #366: ARM assembly implementation of field_10x26 inner (rebase of #173)
001f176 ARM assembly implementation of field_10x26 inner
0172be9 Merge #397: Small fixes for sha256
3f8b78e Fix undefs in hash_impl.h
2ab4695 Fix state size in sha256 struct
6875b01 Merge #386: Add some missing `VERIFY_CHECK(ctx != NULL)`
2c52b5d Merge #389: Cast pointers through uintptr_t under JNI
43097a4 Merge #390: Update bitcoin-core GitHub links
31c9c12 Merge #391: JNI: Only call ecdsa_verify if its inputs parsed correctly
1cb2302 Merge #392: Add testcase which hits additional branch in secp256k1_scalar_sqr
d2ee340 Merge #388: bench_ecdh: fix call to secp256k1_context_create
093a497 Add testcase which hits additional branch in secp256k1_scalar_sqr
a40c701 JNI: Only call ecdsa_verify if its inputs parsed correctly
faa2a11 Update bitcoin-core GitHub links
47b9e78 Cast pointers through uintptr_t under JNI
f36f9c6 bench_ecdh: fix call to secp256k1_context_create
bcc4881 Add some missing `VERIFY_CHECK(ctx != NULL)` for functions that use `ARG_CHECK`
6ceea2c align shared files copyright/comments to bitcoinj's
70141a8 Update .gitignore
7b549b1 Merge #373: build: fix x86_64 asm detection for some compilers
bc7c93c Merge #374: Add note about y=0 being possible on one of the sextic twists
e457018 Merge #364: JNI rebased
86e2d07 JNI library: cleanup, removed unimplemented code
3093576a JNI library
bd2895f Merge pull request #371
e72e93a Add note about y=0 being possible on one of the sextic twists
3f8fdfb build: fix x86_64 asm detection for some compilers
e5a9047 [Trivial] Remove double semicolons
c18b869 Merge pull request #360
3026daa Merge pull request #302
03d4611 Add sage verification script for the group laws
a965937 Merge pull request #361
83221ec Add experimental features to configure
5d4c5a3 Prevent damage_array in the signature test from going out of bounds.
419bf7f Merge pull request #356
03d84a4 Benchmark against OpenSSL verification

git-subtree-dir: src/secp256k1
git-subtree-split: 7a49cacd3937311fcb1cb36b6ba3336fca811991

31 files changed, 2651 insertions, 153 deletions

diff --git a/src/asm/field_10x26_arm.s b/src/asm/field_10x26_arm.s
new file mode 100644
index 0000000000..5df561f2fc
--- /dev/null
+++ b/src/asm/field_10x26_arm.s
@@ -0,0 +1,919 @@
+@ vim: set tabstop=8 softtabstop=8 shiftwidth=8 noexpandtab syntax=armasm:
+/**********************************************************************
+ * Copyright (c) 2014 Wladimir J. van der Laan                        *
+ * Distributed under the MIT software license, see the accompanying   *
+ * file COPYING or http://www.opensource.org/licenses/mit-license.php.*
+ **********************************************************************/
+/*
+ARM implementation of field_10x26 inner loops.
+
+Note:
+
+- To avoid unnecessary loads and make use of available registers, two
+  'passes' have every time been interleaved, with the odd passes accumulating c' and d' 
+  which will be added to c and d respectively in the the even passes
+
+*/
+
+	.syntax unified
+	.arch armv7-a
+	@ eabi attributes - see readelf -A
+	.eabi_attribute 8, 1  @ Tag_ARM_ISA_use = yes
+	.eabi_attribute 9, 0  @ Tag_Thumb_ISA_use = no
+	.eabi_attribute 10, 0 @ Tag_FP_arch = none
+	.eabi_attribute 24, 1 @ Tag_ABI_align_needed = 8-byte
+	.eabi_attribute 25, 1 @ Tag_ABI_align_preserved = 8-byte, except leaf SP
+	.eabi_attribute 30, 2 @ Tag_ABI_optimization_goals = Agressive Speed
+	.eabi_attribute 34, 1 @ Tag_CPU_unaligned_access = v6
+	.text
+
+	@ Field constants
+	.set field_R0, 0x3d10
+	.set field_R1, 0x400
+	.set field_not_M, 0xfc000000	@ ~M = ~0x3ffffff
+
+	.align	2
+	.global secp256k1_fe_mul_inner
+	.type	secp256k1_fe_mul_inner, %function
+	@ Arguments:
+	@  r0  r      Restrict: can overlap with a, not with b
+	@  r1  a
+	@  r2  b
+	@ Stack (total 4+10*4 = 44)
+	@  sp + #0        saved 'r' pointer
+	@  sp + #4 + 4*X  t0,t1,t2,t3,t4,t5,t6,t7,u8,t9
+secp256k1_fe_mul_inner:
+	stmfd	sp!, {r4, r5, r6, r7, r8, r9, r10, r11, r14}
+	sub	sp, sp, #48			@ frame=44 + alignment
+	str     r0, [sp, #0]			@ save result address, we need it only at the end
+
+	/******************************************
+	 * Main computation code.
+	 ******************************************
+
+	Allocation:
+	    r0,r14,r7,r8   scratch
+	    r1       a (pointer)
+	    r2       b (pointer)
+	    r3:r4    c
+	    r5:r6    d
+	    r11:r12  c'
+	    r9:r10   d'
+
+	Note: do not write to r[] here, it may overlap with a[]
+	*/
+
+	/* A - interleaved with B */
+	ldr	r7, [r1, #0*4]			@ a[0]
+	ldr	r8, [r2, #9*4]			@ b[9]
+	ldr	r0, [r1, #1*4]			@ a[1]
+	umull	r5, r6, r7, r8			@ d = a[0] * b[9]
+	ldr	r14, [r2, #8*4]			@ b[8]
+	umull	r9, r10, r0, r8			@ d' = a[1] * b[9]
+	ldr	r7, [r1, #2*4]			@ a[2]
+	umlal	r5, r6, r0, r14			@ d += a[1] * b[8]
+	ldr	r8, [r2, #7*4] 			@ b[7]
+	umlal	r9, r10, r7, r14		@ d' += a[2] * b[8]
+	ldr	r0, [r1, #3*4]   		@ a[3]
+	umlal	r5, r6, r7, r8   		@ d += a[2] * b[7]
+	ldr	r14, [r2, #6*4]   		@ b[6]
+	umlal	r9, r10, r0, r8  		@ d' += a[3] * b[7]
+	ldr	r7, [r1, #4*4]   		@ a[4]
+	umlal	r5, r6, r0, r14   		@ d += a[3] * b[6]
+	ldr	r8, [r2, #5*4]   		@ b[5]
+	umlal	r9, r10, r7, r14  		@ d' += a[4] * b[6]
+	ldr	r0, [r1, #5*4]   		@ a[5]
+	umlal	r5, r6, r7, r8   		@ d += a[4] * b[5]
+	ldr	r14, [r2, #4*4]   		@ b[4]
+	umlal	r9, r10, r0, r8  		@ d' += a[5] * b[5]
+	ldr	r7, [r1, #6*4]   		@ a[6]
+	umlal	r5, r6, r0, r14   		@ d += a[5] * b[4]
+	ldr	r8, [r2, #3*4]   		@ b[3]
+	umlal	r9, r10, r7, r14  		@ d' += a[6] * b[4]
+	ldr	r0, [r1, #7*4]   		@ a[7]
+	umlal	r5, r6, r7, r8   		@ d += a[6] * b[3]
+	ldr	r14, [r2, #2*4]   		@ b[2]
+	umlal	r9, r10, r0, r8  		@ d' += a[7] * b[3]
+	ldr	r7, [r1, #8*4]   		@ a[8]
+	umlal	r5, r6, r0, r14   		@ d += a[7] * b[2]
+	ldr	r8, [r2, #1*4]   		@ b[1]
+	umlal	r9, r10, r7, r14  		@ d' += a[8] * b[2]
+	ldr	r0, [r1, #9*4]   		@ a[9]
+	umlal	r5, r6, r7, r8   		@ d += a[8] * b[1]
+	ldr	r14, [r2, #0*4]   		@ b[0]
+	umlal	r9, r10, r0, r8  		@ d' += a[9] * b[1]
+	ldr	r7, [r1, #0*4]   		@ a[0]
+	umlal	r5, r6, r0, r14   		@ d += a[9] * b[0]
+	@ r7,r14 used in B
+
+	bic	r0, r5, field_not_M 		@ t9 = d & M
+	str     r0, [sp, #4 + 4*9]
+	mov	r5, r5, lsr #26     		@ d >>= 26 
+	orr	r5, r5, r6, asl #6
+	mov     r6, r6, lsr #26
+
+	/* B */
+	umull	r3, r4, r7, r14   		@ c = a[0] * b[0]
+	adds	r5, r5, r9       		@ d += d'
+	adc	r6, r6, r10
+
+	bic	r0, r5, field_not_M 		@ u0 = d & M
+	mov	r5, r5, lsr #26     		@ d >>= 26
+	orr	r5, r5, r6, asl #6
+	mov     r6, r6, lsr #26
+	movw    r14, field_R0			@ c += u0 * R0
+	umlal   r3, r4, r0, r14
+
+	bic	r14, r3, field_not_M 		@ t0 = c & M
+	str	r14, [sp, #4 + 0*4]
+	mov	r3, r3, lsr #26     		@ c >>= 26
+	orr	r3, r3, r4, asl #6
+	mov     r4, r4, lsr #26
+	mov     r14, field_R1			@ c += u0 * R1
+	umlal   r3, r4, r0, r14
+
+	/* C - interleaved with D */
+	ldr	r7, [r1, #0*4]   		@ a[0]
+	ldr	r8, [r2, #2*4]   		@ b[2]
+	ldr	r14, [r2, #1*4]   		@ b[1]
+	umull	r11, r12, r7, r8   		@ c' = a[0] * b[2]
+	ldr	r0, [r1, #1*4]   		@ a[1]
+	umlal   r3, r4, r7, r14   		@ c += a[0] * b[1]
+	ldr	r8, [r2, #0*4]   		@ b[0]
+	umlal   r11, r12, r0, r14   		@ c' += a[1] * b[1]
+	ldr	r7, [r1, #2*4]   		@ a[2]
+	umlal   r3, r4, r0, r8   		@ c += a[1] * b[0]
+	ldr	r14, [r2, #9*4]   		@ b[9]
+	umlal   r11, r12, r7, r8   		@ c' += a[2] * b[0]
+	ldr	r0, [r1, #3*4]   		@ a[3]
+	umlal	r5, r6, r7, r14   		@ d += a[2] * b[9]
+	ldr	r8, [r2, #8*4]   		@ b[8]
+	umull	r9, r10, r0, r14   		@ d' = a[3] * b[9]
+	ldr	r7, [r1, #4*4]   		@ a[4]
+	umlal	r5, r6, r0, r8   		@ d += a[3] * b[8]
+	ldr	r14, [r2, #7*4]   		@ b[7]
+	umlal	r9, r10, r7, r8   		@ d' += a[4] * b[8]
+	ldr	r0, [r1, #5*4]   		@ a[5]
+	umlal	r5, r6, r7, r14   		@ d += a[4] * b[7]
+	ldr	r8, [r2, #6*4]   		@ b[6]
+	umlal	r9, r10, r0, r14   		@ d' += a[5] * b[7]
+	ldr	r7, [r1, #6*4]   		@ a[6]
+	umlal	r5, r6, r0, r8   		@ d += a[5] * b[6]
+	ldr	r14, [r2, #5*4]   		@ b[5]
+	umlal	r9, r10, r7, r8   		@ d' += a[6] * b[6]
+	ldr	r0, [r1, #7*4]   		@ a[7]
+	umlal	r5, r6, r7, r14   		@ d += a[6] * b[5]
+	ldr	r8, [r2, #4*4]   		@ b[4]
+	umlal	r9, r10, r0, r14   		@ d' += a[7] * b[5]
+	ldr	r7, [r1, #8*4]   		@ a[8]
+	umlal	r5, r6, r0, r8   		@ d += a[7] * b[4]
+	ldr	r14, [r2, #3*4]   		@ b[3]
+	umlal	r9, r10, r7, r8   		@ d' += a[8] * b[4]
+	ldr	r0, [r1, #9*4]   		@ a[9]
+	umlal	r5, r6, r7, r14   		@ d += a[8] * b[3]
+	ldr	r8, [r2, #2*4]   		@ b[2]
+	umlal	r9, r10, r0, r14   		@ d' += a[9] * b[3]
+	umlal	r5, r6, r0, r8   		@ d += a[9] * b[2]
+
+	bic	r0, r5, field_not_M 		@ u1 = d & M
+	mov	r5, r5, lsr #26     		@ d >>= 26
+	orr	r5, r5, r6, asl #6
+	mov     r6, r6, lsr #26
+	movw    r14, field_R0			@ c += u1 * R0
+	umlal   r3, r4, r0, r14
+
+	bic	r14, r3, field_not_M 		@ t1 = c & M
+	str	r14, [sp, #4 + 1*4]
+	mov	r3, r3, lsr #26     		@ c >>= 26
+	orr	r3, r3, r4, asl #6
+	mov     r4, r4, lsr #26
+	mov     r14, field_R1			@ c += u1 * R1
+	umlal   r3, r4, r0, r14
+
+	/* D */
+	adds	r3, r3, r11			@ c += c'
+	adc	r4, r4, r12
+	adds	r5, r5, r9			@ d += d'
+	adc	r6, r6, r10
+
+	bic	r0, r5, field_not_M 		@ u2 = d & M
+	mov	r5, r5, lsr #26     		@ d >>= 26
+	orr	r5, r5, r6, asl #6
+	mov     r6, r6, lsr #26
+	movw    r14, field_R0			@ c += u2 * R0
+	umlal   r3, r4, r0, r14
+
+	bic	r14, r3, field_not_M 		@ t2 = c & M
+	str	r14, [sp, #4 + 2*4]
+	mov	r3, r3, lsr #26     		@ c >>= 26
+	orr	r3, r3, r4, asl #6
+	mov     r4, r4, lsr #26
+	mov     r14, field_R1			@ c += u2 * R1
+	umlal   r3, r4, r0, r14
+
+	/* E - interleaved with F */
+	ldr	r7, [r1, #0*4]   		@ a[0]
+	ldr	r8, [r2, #4*4]   		@ b[4]
+	umull	r11, r12, r7, r8   		@ c' = a[0] * b[4]
+	ldr	r8, [r2, #3*4]   		@ b[3]
+	umlal   r3, r4, r7, r8   		@ c += a[0] * b[3]
+	ldr	r7, [r1, #1*4]   		@ a[1]
+	umlal   r11, r12, r7, r8   		@ c' += a[1] * b[3]
+	ldr	r8, [r2, #2*4]   		@ b[2]
+	umlal   r3, r4, r7, r8   		@ c += a[1] * b[2]
+	ldr	r7, [r1, #2*4]   		@ a[2]
+	umlal   r11, r12, r7, r8   		@ c' += a[2] * b[2]
+	ldr	r8, [r2, #1*4]   		@ b[1]
+	umlal   r3, r4, r7, r8   		@ c += a[2] * b[1]
+	ldr	r7, [r1, #3*4]   		@ a[3]
+	umlal   r11, r12, r7, r8   		@ c' += a[3] * b[1]
+	ldr	r8, [r2, #0*4]   		@ b[0]
+	umlal   r3, r4, r7, r8   		@ c += a[3] * b[0]
+	ldr	r7, [r1, #4*4]   		@ a[4]
+	umlal   r11, r12, r7, r8   		@ c' += a[4] * b[0]
+	ldr	r8, [r2, #9*4]   		@ b[9]
+	umlal	r5, r6, r7, r8   		@ d += a[4] * b[9]
+	ldr	r7, [r1, #5*4]   		@ a[5]
+	umull	r9, r10, r7, r8   		@ d' = a[5] * b[9]
+	ldr	r8, [r2, #8*4]   		@ b[8]
+	umlal	r5, r6, r7, r8   		@ d += a[5] * b[8]
+	ldr	r7, [r1, #6*4]   		@ a[6]
+	umlal	r9, r10, r7, r8   		@ d' += a[6] * b[8]
+	ldr	r8, [r2, #7*4]   		@ b[7]
+	umlal	r5, r6, r7, r8   		@ d += a[6] * b[7]
+	ldr	r7, [r1, #7*4]   		@ a[7]
+	umlal	r9, r10, r7, r8   		@ d' += a[7] * b[7]
+	ldr	r8, [r2, #6*4]   		@ b[6]
+	umlal	r5, r6, r7, r8   		@ d += a[7] * b[6]
+	ldr	r7, [r1, #8*4]   		@ a[8]
+	umlal	r9, r10, r7, r8   		@ d' += a[8] * b[6]
+	ldr	r8, [r2, #5*4]   		@ b[5]
+	umlal	r5, r6, r7, r8   		@ d += a[8] * b[5]
+	ldr	r7, [r1, #9*4]   		@ a[9]
+	umlal	r9, r10, r7, r8   		@ d' += a[9] * b[5]
+	ldr	r8, [r2, #4*4]   		@ b[4]
+	umlal	r5, r6, r7, r8   		@ d += a[9] * b[4]
+
+	bic	r0, r5, field_not_M 		@ u3 = d & M
+	mov	r5, r5, lsr #26     		@ d >>= 26
+	orr	r5, r5, r6, asl #6
+	mov     r6, r6, lsr #26
+	movw    r14, field_R0			@ c += u3 * R0
+	umlal   r3, r4, r0, r14
+
+	bic	r14, r3, field_not_M 		@ t3 = c & M
+	str	r14, [sp, #4 + 3*4]
+	mov	r3, r3, lsr #26     		@ c >>= 26
+	orr	r3, r3, r4, asl #6
+	mov     r4, r4, lsr #26
+	mov     r14, field_R1			@ c += u3 * R1
+	umlal   r3, r4, r0, r14
+
+	/* F */
+	adds	r3, r3, r11			@ c += c'
+	adc	r4, r4, r12
+	adds	r5, r5, r9			@ d += d'
+	adc	r6, r6, r10
+
+	bic	r0, r5, field_not_M 		@ u4 = d & M
+	mov	r5, r5, lsr #26     		@ d >>= 26
+	orr	r5, r5, r6, asl #6
+	mov     r6, r6, lsr #26
+	movw    r14, field_R0			@ c += u4 * R0
+	umlal   r3, r4, r0, r14
+
+	bic	r14, r3, field_not_M 		@ t4 = c & M
+	str	r14, [sp, #4 + 4*4]
+	mov	r3, r3, lsr #26     		@ c >>= 26
+	orr	r3, r3, r4, asl #6
+	mov     r4, r4, lsr #26
+	mov     r14, field_R1			@ c += u4 * R1
+	umlal   r3, r4, r0, r14
+
+	/* G - interleaved with H */
+	ldr	r7, [r1, #0*4]   		@ a[0]
+	ldr	r8, [r2, #6*4]   		@ b[6]
+	ldr	r14, [r2, #5*4]   		@ b[5]
+	umull	r11, r12, r7, r8   		@ c' = a[0] * b[6]
+	ldr	r0, [r1, #1*4]   		@ a[1]
+	umlal   r3, r4, r7, r14   		@ c += a[0] * b[5]
+	ldr	r8, [r2, #4*4]   		@ b[4]
+	umlal   r11, r12, r0, r14   		@ c' += a[1] * b[5]
+	ldr	r7, [r1, #2*4]   		@ a[2]
+	umlal   r3, r4, r0, r8   		@ c += a[1] * b[4]
+	ldr	r14, [r2, #3*4]   		@ b[3]
+	umlal   r11, r12, r7, r8   		@ c' += a[2] * b[4]
+	ldr	r0, [r1, #3*4]   		@ a[3]
+	umlal   r3, r4, r7, r14   		@ c += a[2] * b[3]
+	ldr	r8, [r2, #2*4]   		@ b[2]
+	umlal   r11, r12, r0, r14   		@ c' += a[3] * b[3]
+	ldr	r7, [r1, #4*4]   		@ a[4]
+	umlal   r3, r4, r0, r8   		@ c += a[3] * b[2]
+	ldr	r14, [r2, #1*4]   		@ b[1]
+	umlal   r11, r12, r7, r8   		@ c' += a[4] * b[2]
+	ldr	r0, [r1, #5*4]   		@ a[5]
+	umlal   r3, r4, r7, r14   		@ c += a[4] * b[1]
+	ldr	r8, [r2, #0*4]   		@ b[0]
+	umlal   r11, r12, r0, r14   		@ c' += a[5] * b[1]
+	ldr	r7, [r1, #6*4]   		@ a[6]
+	umlal   r3, r4, r0, r8   		@ c += a[5] * b[0]
+	ldr	r14, [r2, #9*4]   		@ b[9]
+	umlal   r11, r12, r7, r8   		@ c' += a[6] * b[0]
+	ldr	r0, [r1, #7*4]   		@ a[7]
+	umlal	r5, r6, r7, r14   		@ d += a[6] * b[9]
+	ldr	r8, [r2, #8*4]   		@ b[8]
+	umull	r9, r10, r0, r14   		@ d' = a[7] * b[9]
+	ldr	r7, [r1, #8*4]   		@ a[8]
+	umlal	r5, r6, r0, r8   		@ d += a[7] * b[8]
+	ldr	r14, [r2, #7*4]   		@ b[7]
+	umlal	r9, r10, r7, r8   		@ d' += a[8] * b[8]
+	ldr	r0, [r1, #9*4]   		@ a[9]
+	umlal	r5, r6, r7, r14   		@ d += a[8] * b[7]
+	ldr	r8, [r2, #6*4]   		@ b[6]
+	umlal	r9, r10, r0, r14   		@ d' += a[9] * b[7]
+	umlal	r5, r6, r0, r8   		@ d += a[9] * b[6]
+
+	bic	r0, r5, field_not_M 		@ u5 = d & M
+	mov	r5, r5, lsr #26     		@ d >>= 26
+	orr	r5, r5, r6, asl #6
+	mov     r6, r6, lsr #26
+	movw    r14, field_R0			@ c += u5 * R0
+	umlal   r3, r4, r0, r14
+
+	bic	r14, r3, field_not_M 		@ t5 = c & M
+	str	r14, [sp, #4 + 5*4]
+	mov	r3, r3, lsr #26     		@ c >>= 26
+	orr	r3, r3, r4, asl #6
+	mov     r4, r4, lsr #26
+	mov     r14, field_R1			@ c += u5 * R1
+	umlal   r3, r4, r0, r14
+
+	/* H */
+	adds	r3, r3, r11			@ c += c'
+	adc	r4, r4, r12
+	adds	r5, r5, r9			@ d += d'
+	adc	r6, r6, r10
+
+	bic	r0, r5, field_not_M 		@ u6 = d & M
+	mov	r5, r5, lsr #26     		@ d >>= 26
+	orr	r5, r5, r6, asl #6
+	mov     r6, r6, lsr #26
+	movw    r14, field_R0			@ c += u6 * R0
+	umlal   r3, r4, r0, r14
+
+	bic	r14, r3, field_not_M 		@ t6 = c & M
+	str	r14, [sp, #4 + 6*4]
+	mov	r3, r3, lsr #26     		@ c >>= 26
+	orr	r3, r3, r4, asl #6
+	mov     r4, r4, lsr #26
+	mov     r14, field_R1			@ c += u6 * R1
+	umlal   r3, r4, r0, r14
+
+	/* I - interleaved with J */
+	ldr	r8, [r2, #8*4]   		@ b[8]
+	ldr	r7, [r1, #0*4]   		@ a[0]
+	ldr	r14, [r2, #7*4]   		@ b[7]
+	umull   r11, r12, r7, r8   		@ c' = a[0] * b[8]
+	ldr	r0, [r1, #1*4]   		@ a[1]
+	umlal   r3, r4, r7, r14   		@ c += a[0] * b[7]
+	ldr	r8, [r2, #6*4]   		@ b[6]
+	umlal   r11, r12, r0, r14   		@ c' += a[1] * b[7]
+	ldr	r7, [r1, #2*4]   		@ a[2]
+	umlal   r3, r4, r0, r8   		@ c += a[1] * b[6]
+	ldr	r14, [r2, #5*4]   		@ b[5]
+	umlal   r11, r12, r7, r8   		@ c' += a[2] * b[6]
+	ldr	r0, [r1, #3*4]   		@ a[3]
+	umlal   r3, r4, r7, r14   		@ c += a[2] * b[5]
+	ldr	r8, [r2, #4*4]   		@ b[4]
+	umlal   r11, r12, r0, r14   		@ c' += a[3] * b[5]
+	ldr	r7, [r1, #4*4]   		@ a[4]
+	umlal   r3, r4, r0, r8   		@ c += a[3] * b[4]
+	ldr	r14, [r2, #3*4]   		@ b[3]
+	umlal   r11, r12, r7, r8   		@ c' += a[4] * b[4]
+	ldr	r0, [r1, #5*4]   		@ a[5]
+	umlal   r3, r4, r7, r14   		@ c += a[4] * b[3]
+	ldr	r8, [r2, #2*4]   		@ b[2]
+	umlal   r11, r12, r0, r14   		@ c' += a[5] * b[3]
+	ldr	r7, [r1, #6*4]   		@ a[6]
+	umlal   r3, r4, r0, r8   		@ c += a[5] * b[2]
+	ldr	r14, [r2, #1*4]   		@ b[1]
+	umlal   r11, r12, r7, r8   		@ c' += a[6] * b[2]
+	ldr	r0, [r1, #7*4]   		@ a[7]
+	umlal   r3, r4, r7, r14   		@ c += a[6] * b[1]
+	ldr	r8, [r2, #0*4]   		@ b[0]
+	umlal   r11, r12, r0, r14   		@ c' += a[7] * b[1]
+	ldr	r7, [r1, #8*4]   		@ a[8]
+	umlal   r3, r4, r0, r8   		@ c += a[7] * b[0]
+	ldr	r14, [r2, #9*4]   		@ b[9]
+	umlal   r11, r12, r7, r8   		@ c' += a[8] * b[0]
+	ldr	r0, [r1, #9*4]   		@ a[9]
+	umlal	r5, r6, r7, r14   		@ d += a[8] * b[9]
+	ldr	r8, [r2, #8*4]   		@ b[8]
+	umull	r9, r10, r0, r14  		@ d' = a[9] * b[9]
+	umlal	r5, r6, r0, r8   		@ d += a[9] * b[8]
+
+	bic	r0, r5, field_not_M 		@ u7 = d & M
+	mov	r5, r5, lsr #26     		@ d >>= 26
+	orr	r5, r5, r6, asl #6
+	mov     r6, r6, lsr #26
+	movw    r14, field_R0			@ c += u7 * R0
+	umlal   r3, r4, r0, r14
+
+	bic	r14, r3, field_not_M 		@ t7 = c & M
+	str	r14, [sp, #4 + 7*4]
+	mov	r3, r3, lsr #26     		@ c >>= 26
+	orr	r3, r3, r4, asl #6
+	mov     r4, r4, lsr #26
+	mov     r14, field_R1			@ c += u7 * R1
+	umlal   r3, r4, r0, r14
+
+	/* J */
+	adds	r3, r3, r11			@ c += c'
+	adc	r4, r4, r12
+	adds	r5, r5, r9			@ d += d'
+	adc	r6, r6, r10
+
+	bic	r0, r5, field_not_M 		@ u8 = d & M
+	str	r0, [sp, #4 + 8*4]
+	mov	r5, r5, lsr #26     		@ d >>= 26
+	orr	r5, r5, r6, asl #6
+	mov     r6, r6, lsr #26
+	movw    r14, field_R0			@ c += u8 * R0
+	umlal   r3, r4, r0, r14
+
+	/******************************************
+	 * compute and write back result
+	 ******************************************
+	Allocation:
+	    r0    r
+	    r3:r4 c
+	    r5:r6 d
+	    r7    t0
+	    r8    t1
+	    r9    t2
+	    r11   u8
+	    r12   t9
+	    r1,r2,r10,r14 scratch
+
+	Note: do not read from a[] after here, it may overlap with r[]
+	*/
+	ldr	r0, [sp, #0]
+	add	r1, sp, #4 + 3*4		@ r[3..7] = t3..7, r11=u8, r12=t9
+	ldmia	r1, {r2,r7,r8,r9,r10,r11,r12}
+	add	r1, r0, #3*4
+	stmia	r1, {r2,r7,r8,r9,r10}
+
+	bic	r2, r3, field_not_M 		@ r[8] = c & M
+	str	r2, [r0, #8*4]
+	mov	r3, r3, lsr #26     		@ c >>= 26
+	orr	r3, r3, r4, asl #6
+	mov     r4, r4, lsr #26
+	mov     r14, field_R1			@ c += u8 * R1
+	umlal   r3, r4, r11, r14
+	movw    r14, field_R0			@ c += d * R0
+	umlal   r3, r4, r5, r14
+	adds	r3, r3, r12			@ c += t9
+	adc	r4, r4, #0
+
+	add	r1, sp, #4 + 0*4		@ r7,r8,r9 = t0,t1,t2
+	ldmia	r1, {r7,r8,r9}
+
+	ubfx	r2, r3, #0, #22     		@ r[9] = c & (M >> 4)
+	str	r2, [r0, #9*4]
+	mov	r3, r3, lsr #22     		@ c >>= 22
+	orr	r3, r3, r4, asl #10
+	mov     r4, r4, lsr #22
+	movw    r14, field_R1 << 4   		@ c += d * (R1 << 4)
+	umlal   r3, r4, r5, r14
+
+	movw    r14, field_R0 >> 4   		@ d = c * (R0 >> 4) + t0 (64x64 multiply+add)
+	umull	r5, r6, r3, r14			@ d = c.lo * (R0 >> 4)
+	adds	r5, r5, r7	    		@ d.lo += t0
+	mla	r6, r14, r4, r6			@ d.hi += c.hi * (R0 >> 4)
+	adc	r6, r6, 0	     		@ d.hi += carry
+
+	bic	r2, r5, field_not_M 		@ r[0] = d & M
+	str	r2, [r0, #0*4]
+
+	mov	r5, r5, lsr #26     		@ d >>= 26
+	orr	r5, r5, r6, asl #6
+	mov     r6, r6, lsr #26
+	
+	movw    r14, field_R1 >> 4   		@ d += c * (R1 >> 4) + t1 (64x64 multiply+add)
+	umull	r1, r2, r3, r14       		@ tmp = c.lo * (R1 >> 4)
+	adds	r5, r5, r8	    		@ d.lo += t1
+	adc	r6, r6, #0	    		@ d.hi += carry
+	adds	r5, r5, r1	    		@ d.lo += tmp.lo
+	mla	r2, r14, r4, r2      		@ tmp.hi += c.hi * (R1 >> 4)
+	adc	r6, r6, r2	   		@ d.hi += carry + tmp.hi
+
+	bic	r2, r5, field_not_M 		@ r[1] = d & M
+	str	r2, [r0, #1*4]
+	mov	r5, r5, lsr #26     		@ d >>= 26 (ignore hi)
+	orr	r5, r5, r6, asl #6
+
+	add	r5, r5, r9	  		@ d += t2
+	str	r5, [r0, #2*4]      		@ r[2] = d
+
+	add	sp, sp, #48
+	ldmfd	sp!, {r4, r5, r6, r7, r8, r9, r10, r11, pc}
+	.size	secp256k1_fe_mul_inner, .-secp256k1_fe_mul_inner
+
+	.align	2
+	.global secp256k1_fe_sqr_inner
+	.type	secp256k1_fe_sqr_inner, %function
+	@ Arguments:
+	@  r0  r	 Can overlap with a
+	@  r1  a
+	@ Stack (total 4+10*4 = 44)
+	@  sp + #0        saved 'r' pointer
+	@  sp + #4 + 4*X  t0,t1,t2,t3,t4,t5,t6,t7,u8,t9
+secp256k1_fe_sqr_inner:
+	stmfd	sp!, {r4, r5, r6, r7, r8, r9, r10, r11, r14}
+	sub	sp, sp, #48			@ frame=44 + alignment
+	str     r0, [sp, #0]			@ save result address, we need it only at the end
+	/******************************************
+	 * Main computation code.
+	 ******************************************
+
+	Allocation:
+	    r0,r14,r2,r7,r8   scratch
+	    r1       a (pointer)
+	    r3:r4    c
+	    r5:r6    d
+	    r11:r12  c'
+	    r9:r10   d'
+
+	Note: do not write to r[] here, it may overlap with a[]
+	*/
+	/* A interleaved with B */
+	ldr	r0, [r1, #1*4]			@ a[1]*2
+	ldr	r7, [r1, #0*4]			@ a[0]
+	mov	r0, r0, asl #1
+	ldr	r14, [r1, #9*4]			@ a[9]
+	umull	r3, r4, r7, r7			@ c = a[0] * a[0]
+	ldr	r8, [r1, #8*4]			@ a[8]
+	mov	r7, r7, asl #1
+	umull	r5, r6, r7, r14			@ d = a[0]*2 * a[9]
+	ldr	r7, [r1, #2*4]			@ a[2]*2
+	umull	r9, r10, r0, r14		@ d' = a[1]*2 * a[9]
+	ldr	r14, [r1, #7*4]			@ a[7]
+	umlal	r5, r6, r0, r8			@ d += a[1]*2 * a[8]
+	mov	r7, r7, asl #1
+	ldr	r0, [r1, #3*4]			@ a[3]*2
+	umlal	r9, r10, r7, r8			@ d' += a[2]*2 * a[8]
+	ldr	r8, [r1, #6*4]			@ a[6]
+	umlal	r5, r6, r7, r14			@ d += a[2]*2 * a[7]
+	mov	r0, r0, asl #1
+	ldr	r7, [r1, #4*4]			@ a[4]*2
+	umlal	r9, r10, r0, r14		@ d' += a[3]*2 * a[7]
+	ldr	r14, [r1, #5*4]			@ a[5]
+	mov	r7, r7, asl #1
+	umlal	r5, r6, r0, r8			@ d += a[3]*2 * a[6]
+	umlal	r9, r10, r7, r8			@ d' += a[4]*2 * a[6]
+	umlal	r5, r6, r7, r14			@ d += a[4]*2 * a[5]
+	umlal	r9, r10, r14, r14		@ d' += a[5] * a[5]
+
+	bic	r0, r5, field_not_M 		@ t9 = d & M
+	str     r0, [sp, #4 + 9*4]
+	mov	r5, r5, lsr #26     		@ d >>= 26 
+	orr	r5, r5, r6, asl #6
+	mov     r6, r6, lsr #26
+
+	/* B */
+	adds	r5, r5, r9			@ d += d'
+	adc	r6, r6, r10
+
+	bic	r0, r5, field_not_M 		@ u0 = d & M
+	mov	r5, r5, lsr #26     		@ d >>= 26
+	orr	r5, r5, r6, asl #6
+	mov     r6, r6, lsr #26
+	movw    r14, field_R0			@ c += u0 * R0
+	umlal   r3, r4, r0, r14
+	bic	r14, r3, field_not_M 		@ t0 = c & M
+	str	r14, [sp, #4 + 0*4]
+	mov	r3, r3, lsr #26     		@ c >>= 26
+	orr	r3, r3, r4, asl #6
+	mov     r4, r4, lsr #26
+	mov     r14, field_R1			@ c += u0 * R1
+	umlal   r3, r4, r0, r14
+
+	/* C interleaved with D */
+	ldr	r0, [r1, #0*4]			@ a[0]*2
+	ldr	r14, [r1, #1*4]			@ a[1]
+	mov	r0, r0, asl #1
+	ldr	r8, [r1, #2*4]			@ a[2]
+	umlal	r3, r4, r0, r14			@ c += a[0]*2 * a[1]
+	mov	r7, r8, asl #1                  @ a[2]*2
+	umull	r11, r12, r14, r14		@ c' = a[1] * a[1]
+	ldr	r14, [r1, #9*4]			@ a[9]
+	umlal	r11, r12, r0, r8		@ c' += a[0]*2 * a[2]
+	ldr	r0, [r1, #3*4]			@ a[3]*2
+	ldr	r8, [r1, #8*4]			@ a[8]
+	umlal	r5, r6, r7, r14			@ d += a[2]*2 * a[9]
+	mov	r0, r0, asl #1
+	ldr	r7, [r1, #4*4]			@ a[4]*2
+	umull	r9, r10, r0, r14		@ d' = a[3]*2 * a[9]
+	ldr	r14, [r1, #7*4]			@ a[7]
+	umlal	r5, r6, r0, r8			@ d += a[3]*2 * a[8]
+	mov	r7, r7, asl #1
+	ldr	r0, [r1, #5*4]			@ a[5]*2
+	umlal	r9, r10, r7, r8			@ d' += a[4]*2 * a[8]
+	ldr	r8, [r1, #6*4]			@ a[6]
+	mov	r0, r0, asl #1
+	umlal	r5, r6, r7, r14			@ d += a[4]*2 * a[7]
+	umlal	r9, r10, r0, r14		@ d' += a[5]*2 * a[7]
+	umlal	r5, r6, r0, r8			@ d += a[5]*2 * a[6]
+	umlal	r9, r10, r8, r8			@ d' += a[6] * a[6]
+
+	bic	r0, r5, field_not_M 		@ u1 = d & M
+	mov	r5, r5, lsr #26     		@ d >>= 26
+	orr	r5, r5, r6, asl #6
+	mov     r6, r6, lsr #26
+	movw    r14, field_R0			@ c += u1 * R0
+	umlal   r3, r4, r0, r14
+	bic	r14, r3, field_not_M 		@ t1 = c & M
+	str	r14, [sp, #4 + 1*4]
+	mov	r3, r3, lsr #26     		@ c >>= 26
+	orr	r3, r3, r4, asl #6
+	mov     r4, r4, lsr #26
+	mov     r14, field_R1			@ c += u1 * R1
+	umlal   r3, r4, r0, r14
+
+	/* D */
+	adds	r3, r3, r11			@ c += c'
+	adc	r4, r4, r12
+	adds	r5, r5, r9			@ d += d'
+	adc	r6, r6, r10
+
+	bic	r0, r5, field_not_M 		@ u2 = d & M
+	mov	r5, r5, lsr #26     		@ d >>= 26
+	orr	r5, r5, r6, asl #6
+	mov     r6, r6, lsr #26
+	movw    r14, field_R0			@ c += u2 * R0
+	umlal   r3, r4, r0, r14
+	bic	r14, r3, field_not_M 		@ t2 = c & M
+	str	r14, [sp, #4 + 2*4]
+	mov	r3, r3, lsr #26     		@ c >>= 26
+	orr	r3, r3, r4, asl #6
+	mov     r4, r4, lsr #26
+	mov     r14, field_R1			@ c += u2 * R1
+	umlal   r3, r4, r0, r14
+
+	/* E interleaved with F */
+	ldr	r7, [r1, #0*4]			@ a[0]*2
+	ldr	r0, [r1, #1*4]			@ a[1]*2
+	ldr	r14, [r1, #2*4]			@ a[2]
+	mov	r7, r7, asl #1
+	ldr	r8, [r1, #3*4]			@ a[3]
+	ldr	r2, [r1, #4*4]
+	umlal	r3, r4, r7, r8			@ c += a[0]*2 * a[3]
+	mov	r0, r0, asl #1
+	umull	r11, r12, r7, r2		@ c' = a[0]*2 * a[4]
+	mov	r2, r2, asl #1			@ a[4]*2
+	umlal	r11, r12, r0, r8		@ c' += a[1]*2 * a[3]
+	ldr	r8, [r1, #9*4]			@ a[9]
+	umlal	r3, r4, r0, r14			@ c += a[1]*2 * a[2]
+	ldr	r0, [r1, #5*4]			@ a[5]*2
+	umlal	r11, r12, r14, r14		@ c' += a[2] * a[2]
+	ldr	r14, [r1, #8*4]			@ a[8]
+	mov	r0, r0, asl #1
+	umlal	r5, r6, r2, r8			@ d += a[4]*2 * a[9]
+	ldr	r7, [r1, #6*4]			@ a[6]*2
+	umull	r9, r10, r0, r8			@ d' = a[5]*2 * a[9]
+	mov	r7, r7, asl #1
+	ldr	r8, [r1, #7*4]			@ a[7]
+	umlal	r5, r6, r0, r14			@ d += a[5]*2 * a[8]
+	umlal	r9, r10, r7, r14		@ d' += a[6]*2 * a[8]
+	umlal	r5, r6, r7, r8			@ d += a[6]*2 * a[7]
+	umlal	r9, r10, r8, r8			@ d' += a[7] * a[7]
+
+	bic	r0, r5, field_not_M 		@ u3 = d & M
+	mov	r5, r5, lsr #26     		@ d >>= 26
+	orr	r5, r5, r6, asl #6
+	mov     r6, r6, lsr #26
+	movw    r14, field_R0			@ c += u3 * R0
+	umlal   r3, r4, r0, r14
+	bic	r14, r3, field_not_M 		@ t3 = c & M
+	str	r14, [sp, #4 + 3*4]
+	mov	r3, r3, lsr #26     		@ c >>= 26
+	orr	r3, r3, r4, asl #6
+	mov     r4, r4, lsr #26
+	mov     r14, field_R1			@ c += u3 * R1
+	umlal   r3, r4, r0, r14
+
+	/* F */
+	adds	r3, r3, r11			@ c += c'
+	adc	r4, r4, r12
+	adds	r5, r5, r9			@ d += d'
+	adc	r6, r6, r10
+
+	bic	r0, r5, field_not_M 		@ u4 = d & M
+	mov	r5, r5, lsr #26     		@ d >>= 26
+	orr	r5, r5, r6, asl #6
+	mov     r6, r6, lsr #26
+	movw    r14, field_R0			@ c += u4 * R0
+	umlal   r3, r4, r0, r14
+	bic	r14, r3, field_not_M 		@ t4 = c & M
+	str	r14, [sp, #4 + 4*4]
+	mov	r3, r3, lsr #26     		@ c >>= 26
+	orr	r3, r3, r4, asl #6
+	mov     r4, r4, lsr #26
+	mov     r14, field_R1			@ c += u4 * R1
+	umlal   r3, r4, r0, r14
+
+	/* G interleaved with H */
+	ldr	r7, [r1, #0*4]			@ a[0]*2
+	ldr	r0, [r1, #1*4]			@ a[1]*2
+	mov	r7, r7, asl #1
+	ldr	r8, [r1, #5*4]			@ a[5]
+	ldr	r2, [r1, #6*4]			@ a[6]
+	umlal	r3, r4, r7, r8			@ c += a[0]*2 * a[5]
+	ldr	r14, [r1, #4*4]			@ a[4]
+	mov	r0, r0, asl #1
+	umull	r11, r12, r7, r2		@ c' = a[0]*2 * a[6]
+	ldr	r7, [r1, #2*4]			@ a[2]*2
+	umlal	r11, r12, r0, r8		@ c' += a[1]*2 * a[5]
+	mov	r7, r7, asl #1
+	ldr	r8, [r1, #3*4]			@ a[3]
+	umlal	r3, r4, r0, r14			@ c += a[1]*2 * a[4]
+	mov	r0, r2, asl #1			@ a[6]*2
+	umlal	r11, r12, r7, r14		@ c' += a[2]*2 * a[4]
+	ldr	r14, [r1, #9*4]			@ a[9]
+	umlal	r3, r4, r7, r8			@ c += a[2]*2 * a[3]
+	ldr	r7, [r1, #7*4]			@ a[7]*2
+	umlal	r11, r12, r8, r8		@ c' += a[3] * a[3]
+	mov	r7, r7, asl #1
+	ldr	r8, [r1, #8*4]			@ a[8]
+	umlal	r5, r6, r0, r14			@ d += a[6]*2 * a[9]
+	umull	r9, r10, r7, r14		@ d' = a[7]*2 * a[9]
+	umlal	r5, r6, r7, r8			@ d += a[7]*2 * a[8]
+	umlal	r9, r10, r8, r8			@ d' += a[8] * a[8]
+
+	bic	r0, r5, field_not_M 		@ u5 = d & M
+	mov	r5, r5, lsr #26     		@ d >>= 26
+	orr	r5, r5, r6, asl #6
+	mov     r6, r6, lsr #26
+	movw    r14, field_R0			@ c += u5 * R0
+	umlal   r3, r4, r0, r14
+	bic	r14, r3, field_not_M 		@ t5 = c & M
+	str	r14, [sp, #4 + 5*4]
+	mov	r3, r3, lsr #26     		@ c >>= 26
+	orr	r3, r3, r4, asl #6
+	mov     r4, r4, lsr #26
+	mov     r14, field_R1			@ c += u5 * R1
+	umlal   r3, r4, r0, r14
+
+	/* H */
+	adds	r3, r3, r11			@ c += c'
+	adc	r4, r4, r12
+	adds	r5, r5, r9			@ d += d'
+	adc	r6, r6, r10
+
+	bic	r0, r5, field_not_M 		@ u6 = d & M
+	mov	r5, r5, lsr #26     		@ d >>= 26
+	orr	r5, r5, r6, asl #6
+	mov     r6, r6, lsr #26
+	movw    r14, field_R0			@ c += u6 * R0
+	umlal   r3, r4, r0, r14
+	bic	r14, r3, field_not_M 		@ t6 = c & M
+	str	r14, [sp, #4 + 6*4]
+	mov	r3, r3, lsr #26     		@ c >>= 26
+	orr	r3, r3, r4, asl #6
+	mov     r4, r4, lsr #26
+	mov     r14, field_R1			@ c += u6 * R1
+	umlal   r3, r4, r0, r14
+
+	/* I interleaved with J */
+	ldr	r7, [r1, #0*4]			@ a[0]*2
+	ldr	r0, [r1, #1*4]			@ a[1]*2
+	mov	r7, r7, asl #1
+	ldr	r8, [r1, #7*4]			@ a[7]
+	ldr	r2, [r1, #8*4]			@ a[8]
+	umlal	r3, r4, r7, r8			@ c += a[0]*2 * a[7]
+	ldr	r14, [r1, #6*4]			@ a[6]
+	mov	r0, r0, asl #1
+	umull	r11, r12, r7, r2		@ c' = a[0]*2 * a[8]
+	ldr	r7, [r1, #2*4]			@ a[2]*2
+	umlal	r11, r12, r0, r8		@ c' += a[1]*2 * a[7]
+	ldr	r8, [r1, #5*4]			@ a[5]
+	umlal	r3, r4, r0, r14			@ c += a[1]*2 * a[6]
+	ldr	r0, [r1, #3*4]			@ a[3]*2
+	mov	r7, r7, asl #1
+	umlal	r11, r12, r7, r14		@ c' += a[2]*2 * a[6]
+	ldr	r14, [r1, #4*4]			@ a[4]
+	mov	r0, r0, asl #1
+	umlal	r3, r4, r7, r8			@ c += a[2]*2 * a[5]
+	mov	r2, r2, asl #1			@ a[8]*2
+	umlal	r11, r12, r0, r8		@ c' += a[3]*2 * a[5]
+	umlal	r3, r4, r0, r14			@ c += a[3]*2 * a[4]
+	umlal	r11, r12, r14, r14		@ c' += a[4] * a[4]
+	ldr	r8, [r1, #9*4]			@ a[9]
+	umlal	r5, r6, r2, r8			@ d += a[8]*2 * a[9]
+	@ r8 will be used in J
+
+	bic	r0, r5, field_not_M 		@ u7 = d & M
+	mov	r5, r5, lsr #26     		@ d >>= 26
+	orr	r5, r5, r6, asl #6
+	mov     r6, r6, lsr #26
+	movw    r14, field_R0			@ c += u7 * R0
+	umlal   r3, r4, r0, r14
+	bic	r14, r3, field_not_M 		@ t7 = c & M
+	str	r14, [sp, #4 + 7*4]
+	mov	r3, r3, lsr #26     		@ c >>= 26
+	orr	r3, r3, r4, asl #6
+	mov     r4, r4, lsr #26
+	mov     r14, field_R1			@ c += u7 * R1
+	umlal   r3, r4, r0, r14
+
+	/* J */
+	adds	r3, r3, r11			@ c += c'
+	adc	r4, r4, r12
+	umlal	r5, r6, r8, r8			@ d += a[9] * a[9]
+
+	bic	r0, r5, field_not_M 		@ u8 = d & M
+	str	r0, [sp, #4 + 8*4]
+	mov	r5, r5, lsr #26     		@ d >>= 26
+	orr	r5, r5, r6, asl #6
+	mov     r6, r6, lsr #26
+	movw    r14, field_R0			@ c += u8 * R0
+	umlal   r3, r4, r0, r14
+
+	/******************************************
+	 * compute and write back result
+	 ******************************************
+	Allocation:
+	    r0    r
+	    r3:r4 c
+	    r5:r6 d
+	    r7    t0
+	    r8    t1
+	    r9    t2
+	    r11   u8
+	    r12   t9
+	    r1,r2,r10,r14 scratch
+
+	Note: do not read from a[] after here, it may overlap with r[]
+	*/
+	ldr	r0, [sp, #0]
+	add	r1, sp, #4 + 3*4		@ r[3..7] = t3..7, r11=u8, r12=t9
+	ldmia	r1, {r2,r7,r8,r9,r10,r11,r12}
+	add	r1, r0, #3*4
+	stmia	r1, {r2,r7,r8,r9,r10}
+
+	bic	r2, r3, field_not_M 		@ r[8] = c & M
+	str	r2, [r0, #8*4]
+	mov	r3, r3, lsr #26     		@ c >>= 26
+	orr	r3, r3, r4, asl #6
+	mov     r4, r4, lsr #26
+	mov     r14, field_R1			@ c += u8 * R1
+	umlal   r3, r4, r11, r14
+	movw    r14, field_R0			@ c += d * R0
+	umlal   r3, r4, r5, r14
+	adds	r3, r3, r12			@ c += t9
+	adc	r4, r4, #0
+
+	add	r1, sp, #4 + 0*4		@ r7,r8,r9 = t0,t1,t2
+	ldmia	r1, {r7,r8,r9}
+
+	ubfx	r2, r3, #0, #22     		@ r[9] = c & (M >> 4)
+	str	r2, [r0, #9*4]
+	mov	r3, r3, lsr #22     		@ c >>= 22
+	orr	r3, r3, r4, asl #10
+	mov     r4, r4, lsr #22
+	movw    r14, field_R1 << 4   		@ c += d * (R1 << 4)
+	umlal   r3, r4, r5, r14
+
+	movw    r14, field_R0 >> 4   		@ d = c * (R0 >> 4) + t0 (64x64 multiply+add)
+	umull	r5, r6, r3, r14			@ d = c.lo * (R0 >> 4)
+	adds	r5, r5, r7	    		@ d.lo += t0
+	mla	r6, r14, r4, r6			@ d.hi += c.hi * (R0 >> 4)
+	adc	r6, r6, 0	     		@ d.hi += carry
+
+	bic	r2, r5, field_not_M 		@ r[0] = d & M
+	str	r2, [r0, #0*4]
+
+	mov	r5, r5, lsr #26     		@ d >>= 26
+	orr	r5, r5, r6, asl #6
+	mov     r6, r6, lsr #26
+	
+	movw    r14, field_R1 >> 4   		@ d += c * (R1 >> 4) + t1 (64x64 multiply+add)
+	umull	r1, r2, r3, r14       		@ tmp = c.lo * (R1 >> 4)
+	adds	r5, r5, r8	    		@ d.lo += t1
+	adc	r6, r6, #0	    		@ d.hi += carry
+	adds	r5, r5, r1	    		@ d.lo += tmp.lo
+	mla	r2, r14, r4, r2      		@ tmp.hi += c.hi * (R1 >> 4)
+	adc	r6, r6, r2	   		@ d.hi += carry + tmp.hi
+
+	bic	r2, r5, field_not_M 		@ r[1] = d & M
+	str	r2, [r0, #1*4]
+	mov	r5, r5, lsr #26     		@ d >>= 26 (ignore hi)
+	orr	r5, r5, r6, asl #6
+
+	add	r5, r5, r9	  		@ d += t2
+	str	r5, [r0, #2*4]      		@ r[2] = d
+
+	add	sp, sp, #48
+	ldmfd	sp!, {r4, r5, r6, r7, r8, r9, r10, r11, pc}
+	.size	secp256k1_fe_sqr_inner, .-secp256k1_fe_sqr_inner
+
diff --git a/src/bench_ecdh.c b/src/bench_ecdh.c
index 5a7c6376e0..cde5e2dbb4 100644
--- a/src/bench_ecdh.c
+++ b/src/bench_ecdh.c
@@ -28,7 +28,8 @@ static void bench_ecdh_setup(void* arg) {
         0xa2, 0xba, 0xd1, 0x84, 0xf8, 0x83, 0xc6, 0x9f
     };
 
-    data->ctx = secp256k1_context_create(0);
+    /* create a context with no capabilities */
+    data->ctx = secp256k1_context_create(SECP256K1_FLAGS_TYPE_CONTEXT);
     for (i = 0; i < 32; i++) {
         data->scalar[i] = i + 1;
     }
diff --git a/src/bench_internal.c b/src/bench_internal.c
index 7809f5f8cf..0809f77bda 100644
--- a/src/bench_internal.c
+++ b/src/bench_internal.c
@@ -181,12 +181,12 @@ void bench_field_inverse_var(void* arg) {
     }
 }
 
-void bench_field_sqrt_var(void* arg) {
+void bench_field_sqrt(void* arg) {
     int i;
     bench_inv_t *data = (bench_inv_t*)arg;
 
     for (i = 0; i < 20000; i++) {
-        secp256k1_fe_sqrt_var(&data->fe_x, &data->fe_x);
+        secp256k1_fe_sqrt(&data->fe_x, &data->fe_x);
         secp256k1_fe_add(&data->fe_x, &data->fe_y);
     }
 }
@@ -227,6 +227,15 @@ void bench_group_add_affine_var(void* arg) {
     }
 }
 
+void bench_group_jacobi_var(void* arg) {
+    int i;
+    bench_inv_t *data = (bench_inv_t*)arg;
+
+    for (i = 0; i < 20000; i++) {
+        secp256k1_gej_has_quad_y_var(&data->gej_x);
+    }
+}
+
 void bench_ecmult_wnaf(void* arg) {
     int i;
     bench_inv_t *data = (bench_inv_t*)arg;
@@ -299,6 +308,21 @@ void bench_context_sign(void* arg) {
     }
 }
 
+#ifndef USE_NUM_NONE
+void bench_num_jacobi(void* arg) {
+    int i;
+    bench_inv_t *data = (bench_inv_t*)arg;
+    secp256k1_num nx, norder;
+
+    secp256k1_scalar_get_num(&nx, &data->scalar_x);
+    secp256k1_scalar_order_get_num(&norder);
+    secp256k1_scalar_get_num(&norder, &data->scalar_y);
+
+    for (i = 0; i < 200000; i++) {
+        secp256k1_num_jacobi(&nx, &norder);
+    }
+}
+#endif
 
 int have_flag(int argc, char** argv, char *flag) {
     char** argm = argv + argc;
@@ -333,12 +357,13 @@ int main(int argc, char **argv) {
     if (have_flag(argc, argv, "field") || have_flag(argc, argv, "mul")) run_benchmark("field_mul", bench_field_mul, bench_setup, NULL, &data, 10, 200000);
     if (have_flag(argc, argv, "field") || have_flag(argc, argv, "inverse")) run_benchmark("field_inverse", bench_field_inverse, bench_setup, NULL, &data, 10, 20000);
     if (have_flag(argc, argv, "field") || have_flag(argc, argv, "inverse")) run_benchmark("field_inverse_var", bench_field_inverse_var, bench_setup, NULL, &data, 10, 20000);
-    if (have_flag(argc, argv, "field") || have_flag(argc, argv, "sqrt")) run_benchmark("field_sqrt_var", bench_field_sqrt_var, bench_setup, NULL, &data, 10, 20000);
+    if (have_flag(argc, argv, "field") || have_flag(argc, argv, "sqrt")) run_benchmark("field_sqrt", bench_field_sqrt, bench_setup, NULL, &data, 10, 20000);
 
     if (have_flag(argc, argv, "group") || have_flag(argc, argv, "double")) run_benchmark("group_double_var", bench_group_double_var, bench_setup, NULL, &data, 10, 200000);
     if (have_flag(argc, argv, "group") || have_flag(argc, argv, "add")) run_benchmark("group_add_var", bench_group_add_var, bench_setup, NULL, &data, 10, 200000);
     if (have_flag(argc, argv, "group") || have_flag(argc, argv, "add")) run_benchmark("group_add_affine", bench_group_add_affine, bench_setup, NULL, &data, 10, 200000);
     if (have_flag(argc, argv, "group") || have_flag(argc, argv, "add")) run_benchmark("group_add_affine_var", bench_group_add_affine_var, bench_setup, NULL, &data, 10, 200000);
+    if (have_flag(argc, argv, "group") || have_flag(argc, argv, "jacobi")) run_benchmark("group_jacobi_var", bench_group_jacobi_var, bench_setup, NULL, &data, 10, 20000);
 
     if (have_flag(argc, argv, "ecmult") || have_flag(argc, argv, "wnaf")) run_benchmark("wnaf_const", bench_wnaf_const, bench_setup, NULL, &data, 10, 20000);
     if (have_flag(argc, argv, "ecmult") || have_flag(argc, argv, "wnaf")) run_benchmark("ecmult_wnaf", bench_ecmult_wnaf, bench_setup, NULL, &data, 10, 20000);
@@ -350,5 +375,8 @@ int main(int argc, char **argv) {
     if (have_flag(argc, argv, "context") || have_flag(argc, argv, "verify")) run_benchmark("context_verify", bench_context_verify, bench_setup, NULL, &data, 10, 20);
     if (have_flag(argc, argv, "context") || have_flag(argc, argv, "sign")) run_benchmark("context_sign", bench_context_sign, bench_setup, NULL, &data, 10, 200);
 
+#ifndef USE_NUM_NONE
+    if (have_flag(argc, argv, "num") || have_flag(argc, argv, "jacobi")) run_benchmark("num_jacobi", bench_num_jacobi, bench_setup, NULL, &data, 10, 200000);
+#endif
     return 0;
 }
diff --git a/src/bench_verify.c b/src/bench_verify.c
index 5718320cda..418defa0aa 100644
--- a/src/bench_verify.c
+++ b/src/bench_verify.c
@@ -11,6 +11,12 @@
 #include "util.h"
 #include "bench.h"
 
+#ifdef ENABLE_OPENSSL_TESTS
+#include <openssl/bn.h>
+#include <openssl/ecdsa.h>
+#include <openssl/obj_mac.h>
+#endif
+
 typedef struct {
     secp256k1_context *ctx;
     unsigned char msg[32];
@@ -19,6 +25,9 @@ typedef struct {
     size_t siglen;
     unsigned char pubkey[33];
     size_t pubkeylen;
+#ifdef ENABLE_OPENSSL_TESTS
+    EC_GROUP* ec_group;
+#endif
 } benchmark_verify_t;
 
 static void benchmark_verify(void* arg) {
@@ -40,6 +49,36 @@ static void benchmark_verify(void* arg) {
     }
 }
 
+#ifdef ENABLE_OPENSSL_TESTS
+static void benchmark_verify_openssl(void* arg) {
+    int i;
+    benchmark_verify_t* data = (benchmark_verify_t*)arg;
+
+    for (i = 0; i < 20000; i++) {
+        data->sig[data->siglen - 1] ^= (i & 0xFF);
+        data->sig[data->siglen - 2] ^= ((i >> 8) & 0xFF);
+        data->sig[data->siglen - 3] ^= ((i >> 16) & 0xFF);
+        {
+            EC_KEY *pkey = EC_KEY_new();
+            const unsigned char *pubkey = &data->pubkey[0];
+            int result;
+
+            CHECK(pkey != NULL);
+            result = EC_KEY_set_group(pkey, data->ec_group);
+            CHECK(result);
+            result = (o2i_ECPublicKey(&pkey, &pubkey, data->pubkeylen)) != NULL;
+            CHECK(result);
+            result = ECDSA_verify(0, &data->msg[0], sizeof(data->msg), &data->sig[0], data->siglen, pkey) == (i == 0);
+            CHECK(result);
+            EC_KEY_free(pkey);
+        }
+        data->sig[data->siglen - 1] ^= (i & 0xFF);
+        data->sig[data->siglen - 2] ^= ((i >> 8) & 0xFF);
+        data->sig[data->siglen - 3] ^= ((i >> 16) & 0xFF);
+    }
+}
+#endif
+
 int main(void) {
     int i;
     secp256k1_pubkey pubkey;
@@ -62,6 +101,11 @@ int main(void) {
     CHECK(secp256k1_ec_pubkey_serialize(data.ctx, data.pubkey, &data.pubkeylen, &pubkey, SECP256K1_EC_COMPRESSED) == 1);
 
     run_benchmark("ecdsa_verify", benchmark_verify, NULL, NULL, &data, 10, 20000);
+#ifdef ENABLE_OPENSSL_TESTS
+    data.ec_group = EC_GROUP_new_by_curve_name(NID_secp256k1);
+    run_benchmark("ecdsa_verify_openssl", benchmark_verify_openssl, NULL, NULL, &data, 10, 20000);
+    EC_GROUP_free(data.ec_group);
+#endif
 
     secp256k1_context_destroy(data.ctx);
     return 0;
diff --git a/src/ecmult_const_impl.h b/src/ecmult_const_impl.h
index 90ac94770e..7a6a25318c 100644
--- a/src/ecmult_const_impl.h
+++ b/src/ecmult_const_impl.h
@@ -58,22 +58,24 @@ static int secp256k1_wnaf_const(int *wnaf, secp256k1_scalar s, int w) {
     int global_sign;
     int skew = 0;
     int word = 0;
+
     /* 1 2 3 */
     int u_last;
     int u;
 
-#ifdef USE_ENDOMORPHISM
     int flip;
     int bit;
     secp256k1_scalar neg_s;
     int not_neg_one;
-    /* If we are using the endomorphism, we cannot handle even numbers by negating
-     * them, since we are working with 128-bit numbers whose negations would be 256
-     * bits, eliminating the performance advantage. Instead we use a technique from
+    /* Note that we cannot handle even numbers by negating them to be odd, as is
+     * done in other implementations, since if our scalars were specified to have
+     * width < 256 for performance reasons, their negations would have width 256
+     * and we'd lose any performance benefit. Instead, we use a technique from
      * Section 4.2 of the Okeya/Tagaki paper, which is to add either 1 (for even)
-     * or 2 (for odd) to the number we are encoding, then compensating after the
-     * multiplication. */
-    /* Negative 128-bit numbers will be negated, since otherwise they are 256-bit */
+     * or 2 (for odd) to the number we are encoding, returning a skew value indicating
+     * this, and having the caller compensate after doing the multiplication. */
+
+    /* Negative numbers will be negated to keep their bit representation below the maximum width */
     flip = secp256k1_scalar_is_high(&s);
     /* We add 1 to even numbers, 2 to odd ones, noting that negation flips parity */
     bit = flip ^ (s.d[0] & 1);
@@ -89,11 +91,6 @@ static int secp256k1_wnaf_const(int *wnaf, secp256k1_scalar s, int w) {
     global_sign = secp256k1_scalar_cond_negate(&s, flip);
     global_sign *= not_neg_one * 2 - 1;
     skew = 1 << bit;
-#else
-    /* Otherwise, we just negate to force oddness */
-    int is_even = secp256k1_scalar_is_even(&s);
-    global_sign = secp256k1_scalar_cond_negate(&s, is_even);
-#endif
 
     /* 4 */
     u_last = secp256k1_scalar_shr_int(&s, w);
@@ -127,15 +124,13 @@ static void secp256k1_ecmult_const(secp256k1_gej *r, const secp256k1_ge *a, cons
     secp256k1_ge tmpa;
     secp256k1_fe Z;
 
+    int skew_1;
+    int wnaf_1[1 + WNAF_SIZE(WINDOW_A - 1)];
 #ifdef USE_ENDOMORPHISM
     secp256k1_ge pre_a_lam[ECMULT_TABLE_SIZE(WINDOW_A)];
-    int wnaf_1[1 + WNAF_SIZE(WINDOW_A - 1)];
     int wnaf_lam[1 + WNAF_SIZE(WINDOW_A - 1)];
-    int skew_1;
     int skew_lam;
     secp256k1_scalar q_1, q_lam;
-#else
-    int wnaf[1 + WNAF_SIZE(WINDOW_A - 1)];
 #endif
 
     int i;
@@ -145,18 +140,10 @@ static void secp256k1_ecmult_const(secp256k1_gej *r, const secp256k1_ge *a, cons
 #ifdef USE_ENDOMORPHISM
     /* split q into q_1 and q_lam (where q = q_1 + q_lam*lambda, and q_1 and q_lam are ~128 bit) */
     secp256k1_scalar_split_lambda(&q_1, &q_lam, &sc);
-    /* no need for zero correction when using endomorphism since even
-     * numbers have one added to them anyway */
     skew_1   = secp256k1_wnaf_const(wnaf_1,   q_1,   WINDOW_A - 1);
     skew_lam = secp256k1_wnaf_const(wnaf_lam, q_lam, WINDOW_A - 1);
 #else
-    int is_zero = secp256k1_scalar_is_zero(scalar);
-    /* the wNAF ladder cannot handle zero, so bump this to one .. we will
-     * correct the result after the fact */
-    sc.d[0] += is_zero;
-    VERIFY_CHECK(!secp256k1_scalar_is_zero(&sc));
-
-    secp256k1_wnaf_const(wnaf, sc, WINDOW_A - 1);
+    skew_1   = secp256k1_wnaf_const(wnaf_1, sc, WINDOW_A - 1);
 #endif
 
     /* Calculate odd multiples of a.
@@ -179,21 +166,15 @@ static void secp256k1_ecmult_const(secp256k1_gej *r, const secp256k1_ge *a, cons
     /* first loop iteration (separated out so we can directly set r, rather
      * than having it start at infinity, get doubled several times, then have
      * its new value added to it) */
-#ifdef USE_ENDOMORPHISM
     i = wnaf_1[WNAF_SIZE(WINDOW_A - 1)];
     VERIFY_CHECK(i != 0);
     ECMULT_CONST_TABLE_GET_GE(&tmpa, pre_a, i, WINDOW_A);
     secp256k1_gej_set_ge(r, &tmpa);
-
+#ifdef USE_ENDOMORPHISM
     i = wnaf_lam[WNAF_SIZE(WINDOW_A - 1)];
     VERIFY_CHECK(i != 0);
     ECMULT_CONST_TABLE_GET_GE(&tmpa, pre_a_lam, i, WINDOW_A);
     secp256k1_gej_add_ge(r, r, &tmpa);
-#else
-    i = wnaf[WNAF_SIZE(WINDOW_A - 1)];
-    VERIFY_CHECK(i != 0);
-    ECMULT_CONST_TABLE_GET_GE(&tmpa, pre_a, i, WINDOW_A);
-    secp256k1_gej_set_ge(r, &tmpa);
 #endif
     /* remaining loop iterations */
     for (i = WNAF_SIZE(WINDOW_A - 1) - 1; i >= 0; i--) {
@@ -202,59 +183,57 @@ static void secp256k1_ecmult_const(secp256k1_gej *r, const secp256k1_ge *a, cons
         for (j = 0; j < WINDOW_A - 1; ++j) {
             secp256k1_gej_double_nonzero(r, r, NULL);
         }
-#ifdef USE_ENDOMORPHISM
+
         n = wnaf_1[i];
         ECMULT_CONST_TABLE_GET_GE(&tmpa, pre_a, n, WINDOW_A);
         VERIFY_CHECK(n != 0);
         secp256k1_gej_add_ge(r, r, &tmpa);
-
+#ifdef USE_ENDOMORPHISM
         n = wnaf_lam[i];
         ECMULT_CONST_TABLE_GET_GE(&tmpa, pre_a_lam, n, WINDOW_A);
         VERIFY_CHECK(n != 0);
         secp256k1_gej_add_ge(r, r, &tmpa);
-#else
-        n = wnaf[i];
-        VERIFY_CHECK(n != 0);
-        ECMULT_CONST_TABLE_GET_GE(&tmpa, pre_a, n, WINDOW_A);
-        secp256k1_gej_add_ge(r, r, &tmpa);
 #endif
     }
 
     secp256k1_fe_mul(&r->z, &r->z, &Z);
 
-#ifdef USE_ENDOMORPHISM
     {
         /* Correct for wNAF skew */
         secp256k1_ge correction = *a;
         secp256k1_ge_storage correction_1_stor;
+#ifdef USE_ENDOMORPHISM
         secp256k1_ge_storage correction_lam_stor;
+#endif
         secp256k1_ge_storage a2_stor;
         secp256k1_gej tmpj;
         secp256k1_gej_set_ge(&tmpj, &correction);
         secp256k1_gej_double_var(&tmpj, &tmpj, NULL);
         secp256k1_ge_set_gej(&correction, &tmpj);
         secp256k1_ge_to_storage(&correction_1_stor, a);
+#ifdef USE_ENDOMORPHISM
         secp256k1_ge_to_storage(&correction_lam_stor, a);
+#endif
         secp256k1_ge_to_storage(&a2_stor, &correction);
 
         /* For odd numbers this is 2a (so replace it), for even ones a (so no-op) */
         secp256k1_ge_storage_cmov(&correction_1_stor, &a2_stor, skew_1 == 2);
+#ifdef USE_ENDOMORPHISM
         secp256k1_ge_storage_cmov(&correction_lam_stor, &a2_stor, skew_lam == 2);
+#endif
 
         /* Apply the correction */
         secp256k1_ge_from_storage(&correction, &correction_1_stor);
         secp256k1_ge_neg(&correction, &correction);
         secp256k1_gej_add_ge(r, r, &correction);
 
+#ifdef USE_ENDOMORPHISM
         secp256k1_ge_from_storage(&correction, &correction_lam_stor);
         secp256k1_ge_neg(&correction, &correction);
         secp256k1_ge_mul_lambda(&correction, &correction);
         secp256k1_gej_add_ge(r, r, &correction);
-    }
-#else
-    /* correct for zero */
-    r->infinity |= is_zero;
 #endif
+    }
 }
 
 #endif
diff --git a/src/ecmult_impl.h b/src/ecmult_impl.h
index e6e5f47188..81ae08e100 100644
--- a/src/ecmult_impl.h
+++ b/src/ecmult_impl.h
@@ -11,6 +11,8 @@
 #include "scalar.h"
 #include "ecmult.h"
 
+#include <string.h>
+
 /* optimal for 128-bit and 256-bit exponents. */
 #define WINDOW_A 5
 
diff --git a/src/field.h b/src/field.h
index 2d52af5e36..c5ba074244 100644
--- a/src/field.h
+++ b/src/field.h
@@ -57,6 +57,9 @@ static int secp256k1_fe_is_zero(const secp256k1_fe *a);
 static int secp256k1_fe_is_odd(const secp256k1_fe *a);
 
 /** Compare two field elements. Requires magnitude-1 inputs. */
+static int secp256k1_fe_equal(const secp256k1_fe *a, const secp256k1_fe *b);
+
+/** Same as secp256k1_fe_equal, but may be variable time. */
 static int secp256k1_fe_equal_var(const secp256k1_fe *a, const secp256k1_fe *b);
 
 /** Compare two field elements. Requires both inputs to be normalized */
@@ -92,7 +95,10 @@ static void secp256k1_fe_sqr(secp256k1_fe *r, const secp256k1_fe *a);
  *  The input's magnitude can be at most 8. The output magnitude is 1 (but not
  *  guaranteed to be normalized). The result in r will always be a square
  *  itself. */
-static int secp256k1_fe_sqrt_var(secp256k1_fe *r, const secp256k1_fe *a);
+static int secp256k1_fe_sqrt(secp256k1_fe *r, const secp256k1_fe *a);
+
+/** Checks whether a field element is a quadratic residue. */
+static int secp256k1_fe_is_quad_var(const secp256k1_fe *a);
 
 /** Sets a field element to be the (modular) inverse of another. Requires the input's magnitude to be
  *  at most 8. The output magnitude is 1 (but not guaranteed to be normalized). */
diff --git a/src/field_10x26_impl.h b/src/field_10x26_impl.h
index 212cc5396a..7b8c079608 100644
--- a/src/field_10x26_impl.h
+++ b/src/field_10x26_impl.h
@@ -7,8 +7,6 @@
 #ifndef _SECP256K1_FIELD_REPR_IMPL_H_
 #define _SECP256K1_FIELD_REPR_IMPL_H_
 
-#include <stdio.h>
-#include <string.h>
 #include "util.h"
 #include "num.h"
 #include "field.h"
@@ -429,6 +427,14 @@ SECP256K1_INLINE static void secp256k1_fe_add(secp256k1_fe *r, const secp256k1_f
 #endif
 }
 
+#if defined(USE_EXTERNAL_ASM)
+
+/* External assembler implementation */
+void secp256k1_fe_mul_inner(uint32_t *r, const uint32_t *a, const uint32_t * SECP256K1_RESTRICT b);
+void secp256k1_fe_sqr_inner(uint32_t *r, const uint32_t *a);
+
+#else
+
 #ifdef VERIFY
 #define VERIFY_BITS(x, n) VERIFY_CHECK(((x) >> (n)) == 0)
 #else
@@ -1037,7 +1043,7 @@ SECP256K1_INLINE static void secp256k1_fe_sqr_inner(uint32_t *r, const uint32_t
     VERIFY_BITS(r[2], 27);
     /* [r9 r8 r7 r6 r5 r4 r3 r2 r1 r0] = [p18 p17 p16 p15 p14 p13 p12 p11 p10 p9 p8 p7 p6 p5 p4 p3 p2 p1 p0] */
 }
-
+#endif
 
 static void secp256k1_fe_mul(secp256k1_fe *r, const secp256k1_fe *a, const secp256k1_fe * SECP256K1_RESTRICT b) {
 #ifdef VERIFY
diff --git a/src/field_5x52_impl.h b/src/field_5x52_impl.h
index b31e24ab81..7a99eb21ec 100644
--- a/src/field_5x52_impl.h
+++ b/src/field_5x52_impl.h
@@ -11,7 +11,6 @@
 #include "libsecp256k1-config.h"
 #endif
 
-#include <string.h>
 #include "util.h"
 #include "num.h"
 #include "field.h"
diff --git a/src/field_5x52_int128_impl.h b/src/field_5x52_int128_impl.h
index 9280bb5ea2..0bf22bdd3e 100644
--- a/src/field_5x52_int128_impl.h
+++ b/src/field_5x52_int128_impl.h
@@ -137,7 +137,7 @@ SECP256K1_INLINE static void secp256k1_fe_mul_inner(uint64_t *r, const uint64_t
     VERIFY_BITS(r[2], 52);
     VERIFY_BITS(c, 63);
     /* [d 0 0 0 t4 t3+c r2 r1 r0] = [p8 p7 p6 p5 p4 p3 p2 p1 p0] */
-    c   += d * R + t3;;
+    c   += d * R + t3;
     VERIFY_BITS(c, 100);
     /* [t4 c r2 r1 r0] = [p8 p7 p6 p5 p4 p3 p2 p1 p0] */
     r[3] = c & M; c >>= 52;
@@ -259,7 +259,7 @@ SECP256K1_INLINE static void secp256k1_fe_sqr_inner(uint64_t *r, const uint64_t
     VERIFY_BITS(c, 63);
     /* [d 0 0 0 t4 t3+c r2 r1 r0] = [p8 p7 p6 p5 p4 p3 p2 p1 p0] */
 
-    c   += d * R + t3;;
+    c   += d * R + t3;
     VERIFY_BITS(c, 100);
     /* [t4 c r2 r1 r0] = [p8 p7 p6 p5 p4 p3 p2 p1 p0] */
     r[3] = c & M; c >>= 52;
diff --git a/src/field_impl.h b/src/field_impl.h
index 77f4aae2f9..52cd902eb3 100644
--- a/src/field_impl.h
+++ b/src/field_impl.h
@@ -21,6 +21,13 @@
 #error "Please select field implementation"
 #endif
 
+SECP256K1_INLINE static int secp256k1_fe_equal(const secp256k1_fe *a, const secp256k1_fe *b) {
+    secp256k1_fe na;
+    secp256k1_fe_negate(&na, a, 1);
+    secp256k1_fe_add(&na, b);
+    return secp256k1_fe_normalizes_to_zero(&na);
+}
+
 SECP256K1_INLINE static int secp256k1_fe_equal_var(const secp256k1_fe *a, const secp256k1_fe *b) {
     secp256k1_fe na;
     secp256k1_fe_negate(&na, a, 1);
@@ -28,7 +35,7 @@ SECP256K1_INLINE static int secp256k1_fe_equal_var(const secp256k1_fe *a, const
     return secp256k1_fe_normalizes_to_zero_var(&na);
 }
 
-static int secp256k1_fe_sqrt_var(secp256k1_fe *r, const secp256k1_fe *a) {
+static int secp256k1_fe_sqrt(secp256k1_fe *r, const secp256k1_fe *a) {
     /** Given that p is congruent to 3 mod 4, we can compute the square root of
      *  a mod p as the (p+1)/4'th power of a.
      *
@@ -123,7 +130,7 @@ static int secp256k1_fe_sqrt_var(secp256k1_fe *r, const secp256k1_fe *a) {
     /* Check that a square root was actually calculated */
 
     secp256k1_fe_sqr(&t1, r);
-    return secp256k1_fe_equal_var(&t1, a);
+    return secp256k1_fe_equal(&t1, a);
 }
 
 static void secp256k1_fe_inv(secp256k1_fe *r, const secp256k1_fe *a) {
@@ -280,4 +287,29 @@ static void secp256k1_fe_inv_all_var(size_t len, secp256k1_fe *r, const secp256k
     r[0] = u;
 }
 
+static int secp256k1_fe_is_quad_var(const secp256k1_fe *a) {
+#ifndef USE_NUM_NONE
+    unsigned char b[32];
+    secp256k1_num n;
+    secp256k1_num m;
+    /* secp256k1 field prime, value p defined in "Standards for Efficient Cryptography" (SEC2) 2.7.1. */
+    static const unsigned char prime[32] = {
+        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,0xFE,0xFF,0xFF,0xFC,0x2F
+    };
+
+    secp256k1_fe c = *a;
+    secp256k1_fe_normalize_var(&c);
+    secp256k1_fe_get_b32(b, &c);
+    secp256k1_num_set_bin(&n, b, 32);
+    secp256k1_num_set_bin(&m, prime, 32);
+    return secp256k1_num_jacobi(&n, &m) >= 0;
+#else
+    secp256k1_fe r;
+    return secp256k1_fe_sqrt(&r, a);
+#endif
+}
+
 #endif
diff --git a/src/group.h b/src/group.h
index ebfe1ca70c..d515716744 100644
--- a/src/group.h
+++ b/src/group.h
@@ -47,7 +47,7 @@ static void secp256k1_ge_set_xy(secp256k1_ge *r, const secp256k1_fe *x, const se
  *  and a Y coordinate that is a quadratic residue modulo p. The return value
  *  is true iff a coordinate with the given X coordinate exists.
  */
-static int secp256k1_ge_set_xquad_var(secp256k1_ge *r, const secp256k1_fe *x);
+static int secp256k1_ge_set_xquad(secp256k1_ge *r, const secp256k1_fe *x);
 
 /** Set a group element (affine) equal to the point with the given X coordinate, and given oddness
  *  for Y. Return value indicates whether the result is valid. */
@@ -94,6 +94,9 @@ static void secp256k1_gej_neg(secp256k1_gej *r, const secp256k1_gej *a);
 /** Check whether a group element is the point at infinity. */
 static int secp256k1_gej_is_infinity(const secp256k1_gej *a);
 
+/** Check whether a group element's y coordinate is a quadratic residue. */
+static int secp256k1_gej_has_quad_y_var(const secp256k1_gej *a);
+
 /** Set r equal to the double of a. If rzr is not-NULL, r->z = a->z * *rzr (where infinity means an implicit z = 0).
  * a may not be zero. Constant time. */
 static void secp256k1_gej_double_nonzero(secp256k1_gej *r, const secp256k1_gej *a, secp256k1_fe *rzr);
diff --git a/src/group_impl.h b/src/group_impl.h
index 42e2f6e6eb..3e9c4c410d 100644
--- a/src/group_impl.h
+++ b/src/group_impl.h
@@ -7,8 +7,6 @@
 #ifndef _SECP256K1_GROUP_IMPL_H_
 #define _SECP256K1_GROUP_IMPL_H_
 
-#include <string.h>
-
 #include "num.h"
 #include "field.h"
 #include "group.h"
@@ -165,7 +163,7 @@ static void secp256k1_ge_clear(secp256k1_ge *r) {
     secp256k1_fe_clear(&r->y);
 }
 
-static int secp256k1_ge_set_xquad_var(secp256k1_ge *r, const secp256k1_fe *x) {
+static int secp256k1_ge_set_xquad(secp256k1_ge *r, const secp256k1_fe *x) {
     secp256k1_fe x2, x3, c;
     r->x = *x;
     secp256k1_fe_sqr(&x2, x);
@@ -173,11 +171,11 @@ static int secp256k1_ge_set_xquad_var(secp256k1_ge *r, const secp256k1_fe *x) {
     r->infinity = 0;
     secp256k1_fe_set_int(&c, 7);
     secp256k1_fe_add(&c, &x3);
-    return secp256k1_fe_sqrt_var(&r->y, &c);
+    return secp256k1_fe_sqrt(&r->y, &c);
 }
 
 static int secp256k1_ge_set_xo_var(secp256k1_ge *r, const secp256k1_fe *x, int odd) {
-    if (!secp256k1_ge_set_xquad_var(r, x)) {
+    if (!secp256k1_ge_set_xquad(r, x)) {
         return 0;
     }
     secp256k1_fe_normalize_var(&r->y);
@@ -251,11 +249,23 @@ static int secp256k1_ge_is_valid_var(const secp256k1_ge *a) {
 }
 
 static void secp256k1_gej_double_var(secp256k1_gej *r, const secp256k1_gej *a, secp256k1_fe *rzr) {
-    /* Operations: 3 mul, 4 sqr, 0 normalize, 12 mul_int/add/negate */
+    /* Operations: 3 mul, 4 sqr, 0 normalize, 12 mul_int/add/negate.
+     *
+     * Note that there is an implementation described at
+     *     https://hyperelliptic.org/EFD/g1p/auto-shortw-jacobian-0.html#doubling-dbl-2009-l
+     * which trades a multiply for a square, but in practice this is actually slower,
+     * mainly because it requires more normalizations.
+     */
     secp256k1_fe t1,t2,t3,t4;
     /** For secp256k1, 2Q is infinity if and only if Q is infinity. This is because if 2Q = infinity,
      *  Q must equal -Q, or that Q.y == -(Q.y), or Q.y is 0. For a point on y^2 = x^3 + 7 to have
      *  y=0, x^3 must be -7 mod p. However, -7 has no cube root mod p.
+     *  
+     *  Having said this, if this function receives a point on a sextic twist, e.g. by
+     *  a fault attack, it is possible for y to be 0. This happens for y^2 = x^3 + 6,
+     *  since -6 does have a cube root mod p. For this point, this function will not set
+     *  the infinity flag even though the point doubles to infinity, and the result
+     *  point will be gibberish (z = 0 but infinity = 0).
      */
     r->infinity = a->infinity;
     if (r->infinity) {
@@ -623,4 +633,18 @@ static void secp256k1_ge_mul_lambda(secp256k1_ge *r, const secp256k1_ge *a) {
 }
 #endif
 
+static int secp256k1_gej_has_quad_y_var(const secp256k1_gej *a) {
+    secp256k1_fe yz;
+
+    if (a->infinity) {
+        return 0;
+    }
+
+    /* We rely on the fact that the Jacobi symbol of 1 / a->z^3 is the same as
+     * that of a->z. Thus a->y / a->z^3 is a quadratic residue iff a->y * a->z
+       is */
+    secp256k1_fe_mul(&yz, &a->y, &a->z);
+    return secp256k1_fe_is_quad_var(&yz);
+}
+
 #endif
diff --git a/src/hash.h b/src/hash.h
index 0ff01e63fa..fca98cab9f 100644
--- a/src/hash.h
+++ b/src/hash.h
@@ -11,7 +11,7 @@
 #include <stdint.h>
 
 typedef struct {
-    uint32_t s[32];
+    uint32_t s[8];
     uint32_t buf[16]; /* In big endian */
     size_t bytes;
 } secp256k1_sha256_t;
diff --git a/src/hash_impl.h b/src/hash_impl.h
index ae55df6d8a..b47e65f830 100644
--- a/src/hash_impl.h
+++ b/src/hash_impl.h
@@ -269,15 +269,13 @@ static void secp256k1_rfc6979_hmac_sha256_finalize(secp256k1_rfc6979_hmac_sha256
     rng->retry = 0;
 }
 
-
+#undef BE32
 #undef Round
-#undef sigma0
 #undef sigma1
-#undef Sigma0
+#undef sigma0
 #undef Sigma1
-#undef Ch
+#undef Sigma0
 #undef Maj
-#undef ReadBE32
-#undef WriteBE32
+#undef Ch
 
 #endif
diff --git a/src/java/org/bitcoin/NativeSecp256k1.java b/src/java/org/bitcoin/NativeSecp256k1.java
index 90a498eaa2..be67048fbe 100644
--- a/src/java/org/bitcoin/NativeSecp256k1.java
+++ b/src/java/org/bitcoin/NativeSecp256k1.java
@@ -1,60 +1,478 @@
+/*
+ * Copyright 2013 Google Inc.
+ * Copyright 2014-2016 the libsecp256k1 contributors
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ *    http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
 package org.bitcoin;
 
 import java.nio.ByteBuffer;
 import java.nio.ByteOrder;
 
+import java.math.BigInteger;
 import com.google.common.base.Preconditions;
-
+import java.util.concurrent.locks.Lock;
+import java.util.concurrent.locks.ReentrantReadWriteLock;
+import static org.bitcoin.NativeSecp256k1Util.*;
 
 /**
- * This class holds native methods to handle ECDSA verification.
- * You can find an example library that can be used for this at
- * https://github.com/sipa/secp256k1
+ * <p>This class holds native methods to handle ECDSA verification.</p>
+ *
+ * <p>You can find an example library that can be used for this at https://github.com/bitcoin/secp256k1</p>
+ *
+ * <p>To build secp256k1 for use with bitcoinj, run
+ * `./configure --enable-jni --enable-experimental --enable-module-schnorr --enable-module-ecdh`
+ * and `make` then copy `.libs/libsecp256k1.so` to your system library path
+ * or point the JVM to the folder containing it with -Djava.library.path
+ * </p>
  */
 public class NativeSecp256k1 {
-    public static final boolean enabled;
-    static {
-        boolean isEnabled = true;
-        try {
-            System.loadLibrary("javasecp256k1");
-        } catch (UnsatisfiedLinkError e) {
-            isEnabled = false;
-        }
-        enabled = isEnabled;
-    }
-    
+
+    private static final ReentrantReadWriteLock rwl = new ReentrantReadWriteLock();
+    private static final Lock r = rwl.readLock();
+    private static final Lock w = rwl.writeLock();
     private static ThreadLocal<ByteBuffer> nativeECDSABuffer = new ThreadLocal<ByteBuffer>();
     /**
      * Verifies the given secp256k1 signature in native code.
      * Calling when enabled == false is undefined (probably library not loaded)
-     * 
+     *
      * @param data The data which was signed, must be exactly 32 bytes
      * @param signature The signature
      * @param pub The public key which did the signing
      */
-    public static boolean verify(byte[] data, byte[] signature, byte[] pub) {
+    public static boolean verify(byte[] data, byte[] signature, byte[] pub) throws AssertFailException{
         Preconditions.checkArgument(data.length == 32 && signature.length <= 520 && pub.length <= 520);
 
         ByteBuffer byteBuff = nativeECDSABuffer.get();
-        if (byteBuff == null) {
-            byteBuff = ByteBuffer.allocateDirect(32 + 8 + 520 + 520);
+        if (byteBuff == null || byteBuff.capacity() < 520) {
+            byteBuff = ByteBuffer.allocateDirect(520);
             byteBuff.order(ByteOrder.nativeOrder());
             nativeECDSABuffer.set(byteBuff);
         }
         byteBuff.rewind();
         byteBuff.put(data);
-        byteBuff.putInt(signature.length);
-        byteBuff.putInt(pub.length);
         byteBuff.put(signature);
         byteBuff.put(pub);
-        return secp256k1_ecdsa_verify(byteBuff) == 1;
+
+        byte[][] retByteArray;
+
+        r.lock();
+        try {
+          return secp256k1_ecdsa_verify(byteBuff, Secp256k1Context.getContext(), signature.length, pub.length) == 1;
+        } finally {
+          r.unlock();
+        }
+    }
+
+    /**
+     * libsecp256k1 Create an ECDSA signature.
+     *
+     * @param data Message hash, 32 bytes
+     * @param key Secret key, 32 bytes
+     *
+     * Return values
+     * @param sig byte array of signature
+     */
+    public static byte[] sign(byte[] data, byte[] sec) throws AssertFailException{
+        Preconditions.checkArgument(data.length == 32 && sec.length <= 32);
+
+        ByteBuffer byteBuff = nativeECDSABuffer.get();
+        if (byteBuff == null || byteBuff.capacity() < 32 + 32) {
+            byteBuff = ByteBuffer.allocateDirect(32 + 32);
+            byteBuff.order(ByteOrder.nativeOrder());
+            nativeECDSABuffer.set(byteBuff);
+        }
+        byteBuff.rewind();
+        byteBuff.put(data);
+        byteBuff.put(sec);
+
+        byte[][] retByteArray;
+
+        r.lock();
+        try {
+          retByteArray = secp256k1_ecdsa_sign(byteBuff, Secp256k1Context.getContext());
+        } finally {
+          r.unlock();
+        }
+
+        byte[] sigArr = retByteArray[0];
+        int sigLen = new BigInteger(new byte[] { retByteArray[1][0] }).intValue();
+        int retVal = new BigInteger(new byte[] { retByteArray[1][1] }).intValue();
+
+        assertEquals(sigArr.length, sigLen, "Got bad signature length.");
+
+        return retVal == 0 ? new byte[0] : sigArr;
+    }
+
+    /**
+     * libsecp256k1 Seckey Verify - returns 1 if valid, 0 if invalid
+     *
+     * @param seckey ECDSA Secret key, 32 bytes
+     */
+    public static boolean secKeyVerify(byte[] seckey) {
+        Preconditions.checkArgument(seckey.length == 32);
+
+        ByteBuffer byteBuff = nativeECDSABuffer.get();
+        if (byteBuff == null || byteBuff.capacity() < seckey.length) {
+            byteBuff = ByteBuffer.allocateDirect(seckey.length);
+            byteBuff.order(ByteOrder.nativeOrder());
+            nativeECDSABuffer.set(byteBuff);
+        }
+        byteBuff.rewind();
+        byteBuff.put(seckey);
+
+        r.lock();
+        try {
+          return secp256k1_ec_seckey_verify(byteBuff,Secp256k1Context.getContext()) == 1;
+        } finally {
+          r.unlock();
+        }
+    }
+
+
+    /**
+     * libsecp256k1 Compute Pubkey - computes public key from secret key
+     *
+     * @param seckey ECDSA Secret key, 32 bytes
+     *
+     * Return values
+     * @param pubkey ECDSA Public key, 33 or 65 bytes
+     */
+    //TODO add a 'compressed' arg
+    public static byte[] computePubkey(byte[] seckey) throws AssertFailException{
+        Preconditions.checkArgument(seckey.length == 32);
+
+        ByteBuffer byteBuff = nativeECDSABuffer.get();
+        if (byteBuff == null || byteBuff.capacity() < seckey.length) {
+            byteBuff = ByteBuffer.allocateDirect(seckey.length);
+            byteBuff.order(ByteOrder.nativeOrder());
+            nativeECDSABuffer.set(byteBuff);
+        }
+        byteBuff.rewind();
+        byteBuff.put(seckey);
+
+        byte[][] retByteArray;
+
+        r.lock();
+        try {
+          retByteArray = secp256k1_ec_pubkey_create(byteBuff, Secp256k1Context.getContext());
+        } finally {
+          r.unlock();
+        }
+
+        byte[] pubArr = retByteArray[0];
+        int pubLen = new BigInteger(new byte[] { retByteArray[1][0] }).intValue();
+        int retVal = new BigInteger(new byte[] { retByteArray[1][1] }).intValue();
+
+        assertEquals(pubArr.length, pubLen, "Got bad pubkey length.");
+
+        return retVal == 0 ? new byte[0]: pubArr;
+    }
+
+    /**
+     * libsecp256k1 Cleanup - This destroys the secp256k1 context object
+     * This should be called at the end of the program for proper cleanup of the context.
+     */
+    public static synchronized void cleanup() {
+        w.lock();
+        try {
+          secp256k1_destroy_context(Secp256k1Context.getContext());
+        } finally {
+          w.unlock();
+        }
+    }
+
+    public static long cloneContext() {
+       r.lock();
+       try {
+        return secp256k1_ctx_clone(Secp256k1Context.getContext());
+       } finally { r.unlock(); }
+    }
+
+    /**
+     * libsecp256k1 PrivKey Tweak-Mul - Tweak privkey by multiplying to it
+     *
+     * @param tweak some bytes to tweak with
+     * @param seckey 32-byte seckey
+     */
+    public static byte[] privKeyTweakMul(byte[] privkey, byte[] tweak) throws AssertFailException{
+        Preconditions.checkArgument(privkey.length == 32);
+
+        ByteBuffer byteBuff = nativeECDSABuffer.get();
+        if (byteBuff == null || byteBuff.capacity() < privkey.length + tweak.length) {
+            byteBuff = ByteBuffer.allocateDirect(privkey.length + tweak.length);
+            byteBuff.order(ByteOrder.nativeOrder());
+            nativeECDSABuffer.set(byteBuff);
+        }
+        byteBuff.rewind();
+        byteBuff.put(privkey);
+        byteBuff.put(tweak);
+
+        byte[][] retByteArray;
+        r.lock();
+        try {
+          retByteArray = secp256k1_privkey_tweak_mul(byteBuff,Secp256k1Context.getContext());
+        } finally {
+          r.unlock();
+        }
+
+        byte[] privArr = retByteArray[0];
+
+        int privLen = (byte) new BigInteger(new byte[] { retByteArray[1][0] }).intValue() & 0xFF;
+        int retVal = new BigInteger(new byte[] { retByteArray[1][1] }).intValue();
+
+        assertEquals(privArr.length, privLen, "Got bad pubkey length.");
+
+        assertEquals(retVal, 1, "Failed return value check.");
+
+        return privArr;
     }
 
     /**
-     * @param byteBuff signature format is byte[32] data,
-     *        native-endian int signatureLength, native-endian int pubkeyLength,
-     *        byte[signatureLength] signature, byte[pubkeyLength] pub
-     * @returns 1 for valid signature, anything else for invalid
+     * libsecp256k1 PrivKey Tweak-Add - Tweak privkey by adding to it
+     *
+     * @param tweak some bytes to tweak with
+     * @param seckey 32-byte seckey
      */
-    private static native int secp256k1_ecdsa_verify(ByteBuffer byteBuff);
+    public static byte[] privKeyTweakAdd(byte[] privkey, byte[] tweak) throws AssertFailException{
+        Preconditions.checkArgument(privkey.length == 32);
+
+        ByteBuffer byteBuff = nativeECDSABuffer.get();
+        if (byteBuff == null || byteBuff.capacity() < privkey.length + tweak.length) {
+            byteBuff = ByteBuffer.allocateDirect(privkey.length + tweak.length);
+            byteBuff.order(ByteOrder.nativeOrder());
+            nativeECDSABuffer.set(byteBuff);
+        }
+        byteBuff.rewind();
+        byteBuff.put(privkey);
+        byteBuff.put(tweak);
+
+        byte[][] retByteArray;
+        r.lock();
+        try {
+          retByteArray = secp256k1_privkey_tweak_add(byteBuff,Secp256k1Context.getContext());
+        } finally {
+          r.unlock();
+        }
+
+        byte[] privArr = retByteArray[0];
+
+        int privLen = (byte) new BigInteger(new byte[] { retByteArray[1][0] }).intValue() & 0xFF;
+        int retVal = new BigInteger(new byte[] { retByteArray[1][1] }).intValue();
+
+        assertEquals(privArr.length, privLen, "Got bad pubkey length.");
+
+        assertEquals(retVal, 1, "Failed return value check.");
+
+        return privArr;
+    }
+
+    /**
+     * libsecp256k1 PubKey Tweak-Add - Tweak pubkey by adding to it
+     *
+     * @param tweak some bytes to tweak with
+     * @param pubkey 32-byte seckey
+     */
+    public static byte[] pubKeyTweakAdd(byte[] pubkey, byte[] tweak) throws AssertFailException{
+        Preconditions.checkArgument(pubkey.length == 33 || pubkey.length == 65);
+
+        ByteBuffer byteBuff = nativeECDSABuffer.get();
+        if (byteBuff == null || byteBuff.capacity() < pubkey.length + tweak.length) {
+            byteBuff = ByteBuffer.allocateDirect(pubkey.length + tweak.length);
+            byteBuff.order(ByteOrder.nativeOrder());
+            nativeECDSABuffer.set(byteBuff);
+        }
+        byteBuff.rewind();
+        byteBuff.put(pubkey);
+        byteBuff.put(tweak);
+
+        byte[][] retByteArray;
+        r.lock();
+        try {
+          retByteArray = secp256k1_pubkey_tweak_add(byteBuff,Secp256k1Context.getContext(), pubkey.length);
+        } finally {
+          r.unlock();
+        }
+
+        byte[] pubArr = retByteArray[0];
+
+        int pubLen = (byte) new BigInteger(new byte[] { retByteArray[1][0] }).intValue() & 0xFF;
+        int retVal = new BigInteger(new byte[] { retByteArray[1][1] }).intValue();
+
+        assertEquals(pubArr.length, pubLen, "Got bad pubkey length.");
+
+        assertEquals(retVal, 1, "Failed return value check.");
+
+        return pubArr;
+    }
+
+    /**
+     * libsecp256k1 PubKey Tweak-Mul - Tweak pubkey by multiplying to it
+     *
+     * @param tweak some bytes to tweak with
+     * @param pubkey 32-byte seckey
+     */
+    public static byte[] pubKeyTweakMul(byte[] pubkey, byte[] tweak) throws AssertFailException{
+        Preconditions.checkArgument(pubkey.length == 33 || pubkey.length == 65);
+
+        ByteBuffer byteBuff = nativeECDSABuffer.get();
+        if (byteBuff == null || byteBuff.capacity() < pubkey.length + tweak.length) {
+            byteBuff = ByteBuffer.allocateDirect(pubkey.length + tweak.length);
+            byteBuff.order(ByteOrder.nativeOrder());
+            nativeECDSABuffer.set(byteBuff);
+        }
+        byteBuff.rewind();
+        byteBuff.put(pubkey);
+        byteBuff.put(tweak);
+
+        byte[][] retByteArray;
+        r.lock();
+        try {
+          retByteArray = secp256k1_pubkey_tweak_mul(byteBuff,Secp256k1Context.getContext(), pubkey.length);
+        } finally {
+          r.unlock();
+        }
+
+        byte[] pubArr = retByteArray[0];
+
+        int pubLen = (byte) new BigInteger(new byte[] { retByteArray[1][0] }).intValue() & 0xFF;
+        int retVal = new BigInteger(new byte[] { retByteArray[1][1] }).intValue();
+
+        assertEquals(pubArr.length, pubLen, "Got bad pubkey length.");
+
+        assertEquals(retVal, 1, "Failed return value check.");
+
+        return pubArr;
+    }
+
+    /**
+     * libsecp256k1 create ECDH secret - constant time ECDH calculation
+     *
+     * @param seckey byte array of secret key used in exponentiaion
+     * @param pubkey byte array of public key used in exponentiaion
+     */
+    public static byte[] createECDHSecret(byte[] seckey, byte[] pubkey) throws AssertFailException{
+        Preconditions.checkArgument(seckey.length <= 32 && pubkey.length <= 65);
+
+        ByteBuffer byteBuff = nativeECDSABuffer.get();
+        if (byteBuff == null || byteBuff.capacity() < 32 + pubkey.length) {
+            byteBuff = ByteBuffer.allocateDirect(32 + pubkey.length);
+            byteBuff.order(ByteOrder.nativeOrder());
+            nativeECDSABuffer.set(byteBuff);
+        }
+        byteBuff.rewind();
+        byteBuff.put(seckey);
+        byteBuff.put(pubkey);
+
+        byte[][] retByteArray;
+        r.lock();
+        try {
+          retByteArray = secp256k1_ecdh(byteBuff, Secp256k1Context.getContext(), pubkey.length);
+        } finally {
+          r.unlock();
+        }
+
+        byte[] resArr = retByteArray[0];
+        int retVal = new BigInteger(new byte[] { retByteArray[1][0] }).intValue();
+
+        assertEquals(resArr.length, 32, "Got bad result length.");
+        assertEquals(retVal, 1, "Failed return value check.");
+
+        return resArr;
+    }
+
+    /**
+     * libsecp256k1 randomize - updates the context randomization
+     *
+     * @param seed 32-byte random seed
+     */
+    public static synchronized boolean randomize(byte[] seed) throws AssertFailException{
+        Preconditions.checkArgument(seed.length == 32 || seed == null);
+
+        ByteBuffer byteBuff = nativeECDSABuffer.get();
+        if (byteBuff == null || byteBuff.capacity() < seed.length) {
+            byteBuff = ByteBuffer.allocateDirect(seed.length);
+            byteBuff.order(ByteOrder.nativeOrder());
+            nativeECDSABuffer.set(byteBuff);
+        }
+        byteBuff.rewind();
+        byteBuff.put(seed);
+
+        w.lock();
+        try {
+          return secp256k1_context_randomize(byteBuff, Secp256k1Context.getContext()) == 1;
+        } finally {
+          w.unlock();
+        }
+    }
+
+    public static byte[] schnorrSign(byte[] data, byte[] sec) throws AssertFailException {
+        Preconditions.checkArgument(data.length == 32 && sec.length <= 32);
+
+        ByteBuffer byteBuff = nativeECDSABuffer.get();
+        if (byteBuff == null) {
+            byteBuff = ByteBuffer.allocateDirect(32 + 32);
+            byteBuff.order(ByteOrder.nativeOrder());
+            nativeECDSABuffer.set(byteBuff);
+        }
+        byteBuff.rewind();
+        byteBuff.put(data);
+        byteBuff.put(sec);
+
+        byte[][] retByteArray;
+
+        r.lock();
+        try {
+          retByteArray = secp256k1_schnorr_sign(byteBuff, Secp256k1Context.getContext());
+        } finally {
+          r.unlock();
+        }
+
+        byte[] sigArr = retByteArray[0];
+        int retVal = new BigInteger(new byte[] { retByteArray[1][0] }).intValue();
+
+        assertEquals(sigArr.length, 64, "Got bad signature length.");
+
+        return retVal == 0 ? new byte[0] : sigArr;
+    }
+
+    private static native long secp256k1_ctx_clone(long context);
+
+    private static native int secp256k1_context_randomize(ByteBuffer byteBuff, long context);
+
+    private static native byte[][] secp256k1_privkey_tweak_add(ByteBuffer byteBuff, long context);
+
+    private static native byte[][] secp256k1_privkey_tweak_mul(ByteBuffer byteBuff, long context);
+
+    private static native byte[][] secp256k1_pubkey_tweak_add(ByteBuffer byteBuff, long context, int pubLen);
+
+    private static native byte[][] secp256k1_pubkey_tweak_mul(ByteBuffer byteBuff, long context, int pubLen);
+
+    private static native void secp256k1_destroy_context(long context);
+
+    private static native int secp256k1_ecdsa_verify(ByteBuffer byteBuff, long context, int sigLen, int pubLen);
+
+    private static native byte[][] secp256k1_ecdsa_sign(ByteBuffer byteBuff, long context);
+
+    private static native int secp256k1_ec_seckey_verify(ByteBuffer byteBuff, long context);
+
+    private static native byte[][] secp256k1_ec_pubkey_create(ByteBuffer byteBuff, long context);
+
+    private static native byte[][] secp256k1_ec_pubkey_parse(ByteBuffer byteBuff, long context, int inputLen);
+
+    private static native byte[][] secp256k1_schnorr_sign(ByteBuffer byteBuff, long context);
+
+    private static native byte[][] secp256k1_ecdh(ByteBuffer byteBuff, long context, int inputLen);
+
 }
diff --git a/src/java/org/bitcoin/NativeSecp256k1Test.java b/src/java/org/bitcoin/NativeSecp256k1Test.java
new file mode 100644
index 0000000000..f18ce95810
--- /dev/null
+++ b/src/java/org/bitcoin/NativeSecp256k1Test.java
@@ -0,0 +1,247 @@
+package org.bitcoin;
+
+import com.google.common.io.BaseEncoding;
+import java.util.Arrays;
+import java.math.BigInteger;
+import javax.xml.bind.DatatypeConverter;
+import static org.bitcoin.NativeSecp256k1Util.*;
+
+/**
+ * This class holds test cases defined for testing this library.
+ */
+public class NativeSecp256k1Test {
+
+    //TODO improve comments/add more tests
+    /**
+      * This tests verify() for a valid signature
+      */
+    public static void testVerifyPos() throws AssertFailException{
+        boolean result = false;
+        byte[] data = BaseEncoding.base16().lowerCase().decode("CF80CD8AED482D5D1527D7DC72FCEFF84E6326592848447D2DC0B0E87DFC9A90".toLowerCase()); //sha256hash of "testing"
+        byte[] sig = BaseEncoding.base16().lowerCase().decode("3044022079BE667EF9DCBBAC55A06295CE870B07029BFCDB2DCE28D959F2815B16F817980220294F14E883B3F525B5367756C2A11EF6CF84B730B36C17CB0C56F0AAB2C98589".toLowerCase());
+        byte[] pub = BaseEncoding.base16().lowerCase().decode("040A629506E1B65CD9D2E0BA9C75DF9C4FED0DB16DC9625ED14397F0AFC836FAE595DC53F8B0EFE61E703075BD9B143BAC75EC0E19F82A2208CAEB32BE53414C40".toLowerCase());
+
+        result = NativeSecp256k1.verify( data, sig, pub);
+        assertEquals( result, true , "testVerifyPos");
+    }
+
+    /**
+      * This tests verify() for a non-valid signature
+      */
+    public static void testVerifyNeg() throws AssertFailException{
+        boolean result = false;
+        byte[] data = BaseEncoding.base16().lowerCase().decode("CF80CD8AED482D5D1527D7DC72FCEFF84E6326592848447D2DC0B0E87DFC9A91".toLowerCase()); //sha256hash of "testing"
+        byte[] sig = BaseEncoding.base16().lowerCase().decode("3044022079BE667EF9DCBBAC55A06295CE870B07029BFCDB2DCE28D959F2815B16F817980220294F14E883B3F525B5367756C2A11EF6CF84B730B36C17CB0C56F0AAB2C98589".toLowerCase());
+        byte[] pub = BaseEncoding.base16().lowerCase().decode("040A629506E1B65CD9D2E0BA9C75DF9C4FED0DB16DC9625ED14397F0AFC836FAE595DC53F8B0EFE61E703075BD9B143BAC75EC0E19F82A2208CAEB32BE53414C40".toLowerCase());
+
+        result = NativeSecp256k1.verify( data, sig, pub);
+        //System.out.println(" TEST " + new BigInteger(1, resultbytes).toString(16));
+        assertEquals( result, false , "testVerifyNeg");
+    }
+
+    /**
+      * This tests secret key verify() for a valid secretkey
+      */
+    public static void testSecKeyVerifyPos() throws AssertFailException{
+        boolean result = false;
+        byte[] sec = BaseEncoding.base16().lowerCase().decode("67E56582298859DDAE725F972992A07C6C4FB9F62A8FFF58CE3CA926A1063530".toLowerCase());
+
+        result = NativeSecp256k1.secKeyVerify( sec );
+        //System.out.println(" TEST " + new BigInteger(1, resultbytes).toString(16));
+        assertEquals( result, true , "testSecKeyVerifyPos");
+    }
+
+    /**
+      * This tests secret key verify() for a invalid secretkey
+      */
+    public static void testSecKeyVerifyNeg() throws AssertFailException{
+        boolean result = false;
+        byte[] sec = BaseEncoding.base16().lowerCase().decode("FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF".toLowerCase());
+
+        result = NativeSecp256k1.secKeyVerify( sec );
+        //System.out.println(" TEST " + new BigInteger(1, resultbytes).toString(16));
+        assertEquals( result, false , "testSecKeyVerifyNeg");
+    }
+
+    /**
+      * This tests public key create() for a valid secretkey
+      */
+    public static void testPubKeyCreatePos() throws AssertFailException{
+        byte[] sec = BaseEncoding.base16().lowerCase().decode("67E56582298859DDAE725F972992A07C6C4FB9F62A8FFF58CE3CA926A1063530".toLowerCase());
+
+        byte[] resultArr = NativeSecp256k1.computePubkey( sec);
+        String pubkeyString = javax.xml.bind.DatatypeConverter.printHexBinary(resultArr);
+        assertEquals( pubkeyString , "04C591A8FF19AC9C4E4E5793673B83123437E975285E7B442F4EE2654DFFCA5E2D2103ED494718C697AC9AEBCFD19612E224DB46661011863ED2FC54E71861E2A6" , "testPubKeyCreatePos");
+    }
+
+    /**
+      * This tests public key create() for a invalid secretkey
+      */
+    public static void testPubKeyCreateNeg() throws AssertFailException{
+       byte[] sec = BaseEncoding.base16().lowerCase().decode("FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF".toLowerCase());
+
+       byte[] resultArr = NativeSecp256k1.computePubkey( sec);
+       String pubkeyString = javax.xml.bind.DatatypeConverter.printHexBinary(resultArr);
+       assertEquals( pubkeyString, "" , "testPubKeyCreateNeg");
+    }
+
+    /**
+      * This tests sign() for a valid secretkey
+      */
+    public static void testSignPos() throws AssertFailException{
+
+        byte[] data = BaseEncoding.base16().lowerCase().decode("CF80CD8AED482D5D1527D7DC72FCEFF84E6326592848447D2DC0B0E87DFC9A90".toLowerCase()); //sha256hash of "testing"
+        byte[] sec = BaseEncoding.base16().lowerCase().decode("67E56582298859DDAE725F972992A07C6C4FB9F62A8FFF58CE3CA926A1063530".toLowerCase());
+
+        byte[] resultArr = NativeSecp256k1.sign(data, sec);
+        String sigString = javax.xml.bind.DatatypeConverter.printHexBinary(resultArr);
+        assertEquals( sigString, "30440220182A108E1448DC8F1FB467D06A0F3BB8EA0533584CB954EF8DA112F1D60E39A202201C66F36DA211C087F3AF88B50EDF4F9BDAA6CF5FD6817E74DCA34DB12390C6E9" , "testSignPos");
+    }
+
+    /**
+      * This tests sign() for a invalid secretkey
+      */
+    public static void testSignNeg() throws AssertFailException{
+        byte[] data = BaseEncoding.base16().lowerCase().decode("CF80CD8AED482D5D1527D7DC72FCEFF84E6326592848447D2DC0B0E87DFC9A90".toLowerCase()); //sha256hash of "testing"
+        byte[] sec = BaseEncoding.base16().lowerCase().decode("FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF".toLowerCase());
+
+        byte[] resultArr = NativeSecp256k1.sign(data, sec);
+        String sigString = javax.xml.bind.DatatypeConverter.printHexBinary(resultArr);
+        assertEquals( sigString, "" , "testSignNeg");
+    }
+
+    /**
+      * This tests private key tweak-add
+      */
+    public static void testPrivKeyTweakAdd_1() throws AssertFailException {
+        byte[] sec = BaseEncoding.base16().lowerCase().decode("67E56582298859DDAE725F972992A07C6C4FB9F62A8FFF58CE3CA926A1063530".toLowerCase());
+        byte[] data = BaseEncoding.base16().lowerCase().decode("3982F19BEF1615BCCFBB05E321C10E1D4CBA3DF0E841C2E41EEB6016347653C3".toLowerCase()); //sha256hash of "tweak"
+
+        byte[] resultArr = NativeSecp256k1.privKeyTweakAdd( sec , data );
+        String sigString = javax.xml.bind.DatatypeConverter.printHexBinary(resultArr);
+        assertEquals( sigString , "A168571E189E6F9A7E2D657A4B53AE99B909F7E712D1C23CED28093CD57C88F3" , "testPrivKeyAdd_1");
+    }
+
+    /**
+      * This tests private key tweak-mul
+      */
+    public static void testPrivKeyTweakMul_1() throws AssertFailException {
+        byte[] sec = BaseEncoding.base16().lowerCase().decode("67E56582298859DDAE725F972992A07C6C4FB9F62A8FFF58CE3CA926A1063530".toLowerCase());
+        byte[] data = BaseEncoding.base16().lowerCase().decode("3982F19BEF1615BCCFBB05E321C10E1D4CBA3DF0E841C2E41EEB6016347653C3".toLowerCase()); //sha256hash of "tweak"
+
+        byte[] resultArr = NativeSecp256k1.privKeyTweakMul( sec , data );
+        String sigString = javax.xml.bind.DatatypeConverter.printHexBinary(resultArr);
+        assertEquals( sigString , "97F8184235F101550F3C71C927507651BD3F1CDB4A5A33B8986ACF0DEE20FFFC" , "testPrivKeyMul_1");
+    }
+
+    /**
+      * This tests private key tweak-add uncompressed
+      */
+    public static void testPrivKeyTweakAdd_2() throws AssertFailException {
+        byte[] pub = BaseEncoding.base16().lowerCase().decode("040A629506E1B65CD9D2E0BA9C75DF9C4FED0DB16DC9625ED14397F0AFC836FAE595DC53F8B0EFE61E703075BD9B143BAC75EC0E19F82A2208CAEB32BE53414C40".toLowerCase());
+        byte[] data = BaseEncoding.base16().lowerCase().decode("3982F19BEF1615BCCFBB05E321C10E1D4CBA3DF0E841C2E41EEB6016347653C3".toLowerCase()); //sha256hash of "tweak"
+
+        byte[] resultArr = NativeSecp256k1.pubKeyTweakAdd( pub , data );
+        String sigString = javax.xml.bind.DatatypeConverter.printHexBinary(resultArr);
+        assertEquals( sigString , "0411C6790F4B663CCE607BAAE08C43557EDC1A4D11D88DFCB3D841D0C6A941AF525A268E2A863C148555C48FB5FBA368E88718A46E205FABC3DBA2CCFFAB0796EF" , "testPrivKeyAdd_2");
+    }
+
+    /**
+      * This tests private key tweak-mul uncompressed
+      */
+    public static void testPrivKeyTweakMul_2() throws AssertFailException {
+        byte[] pub = BaseEncoding.base16().lowerCase().decode("040A629506E1B65CD9D2E0BA9C75DF9C4FED0DB16DC9625ED14397F0AFC836FAE595DC53F8B0EFE61E703075BD9B143BAC75EC0E19F82A2208CAEB32BE53414C40".toLowerCase());
+        byte[] data = BaseEncoding.base16().lowerCase().decode("3982F19BEF1615BCCFBB05E321C10E1D4CBA3DF0E841C2E41EEB6016347653C3".toLowerCase()); //sha256hash of "tweak"
+
+        byte[] resultArr = NativeSecp256k1.pubKeyTweakMul( pub , data );
+        String sigString = javax.xml.bind.DatatypeConverter.printHexBinary(resultArr);
+        assertEquals( sigString , "04E0FE6FE55EBCA626B98A807F6CAF654139E14E5E3698F01A9A658E21DC1D2791EC060D4F412A794D5370F672BC94B722640B5F76914151CFCA6E712CA48CC589" , "testPrivKeyMul_2");
+    }
+
+    /**
+      * This tests seed randomization
+      */
+    public static void testRandomize() throws AssertFailException {
+        byte[] seed = BaseEncoding.base16().lowerCase().decode("A441B15FE9A3CF56661190A0B93B9DEC7D04127288CC87250967CF3B52894D11".toLowerCase()); //sha256hash of "random"
+        boolean result = NativeSecp256k1.randomize(seed);
+        assertEquals( result, true, "testRandomize");
+    }
+
+    /**
+      * This tests signSchnorr() for a valid secretkey
+      */
+    public static void testSchnorrSign() throws AssertFailException{
+
+        byte[] data = BaseEncoding.base16().lowerCase().decode("CF80CD8AED482D5D1527D7DC72FCEFF84E6326592848447D2DC0B0E87DFC9A90".toLowerCase()); //sha256hash of "testing"
+        byte[] sec = BaseEncoding.base16().lowerCase().decode("67E56582298859DDAE725F972992A07C6C4FB9F62A8FFF58CE3CA926A1063530".toLowerCase());
+
+        byte[] resultArr = NativeSecp256k1.schnorrSign(data, sec);
+        String sigString = javax.xml.bind.DatatypeConverter.printHexBinary(resultArr);
+        assertEquals( sigString, "C5E929AA058B982048760422D3B563749B7D0E50C5EBD8CD2FFC23214BD6A2F1B072C13880997EBA847CF20F2F90FCE07C1CA33A890A4127095A351127F8D95F" , "testSchnorrSign");
+    }
+
+    /**
+      * This tests signSchnorr() for a valid secretkey
+      */
+    public static void testCreateECDHSecret() throws AssertFailException{
+
+        byte[] sec = BaseEncoding.base16().lowerCase().decode("67E56582298859DDAE725F972992A07C6C4FB9F62A8FFF58CE3CA926A1063530".toLowerCase());
+        byte[] pub = BaseEncoding.base16().lowerCase().decode("040A629506E1B65CD9D2E0BA9C75DF9C4FED0DB16DC9625ED14397F0AFC836FAE595DC53F8B0EFE61E703075BD9B143BAC75EC0E19F82A2208CAEB32BE53414C40".toLowerCase());
+
+        byte[] resultArr = NativeSecp256k1.createECDHSecret(sec, pub);
+        String ecdhString = javax.xml.bind.DatatypeConverter.printHexBinary(resultArr);
+        assertEquals( ecdhString, "2A2A67007A926E6594AF3EB564FC74005B37A9C8AEF2033C4552051B5C87F043" , "testCreateECDHSecret");
+    }
+
+    public static void main(String[] args) throws AssertFailException{
+
+
+        System.out.println("\n libsecp256k1 enabled: " + Secp256k1Context.isEnabled() + "\n");
+
+        assertEquals( Secp256k1Context.isEnabled(), true, "isEnabled" );
+
+        //Test verify() success/fail
+        testVerifyPos();
+        testVerifyNeg();
+
+        //Test secKeyVerify() success/fail
+        testSecKeyVerifyPos();
+        testSecKeyVerifyNeg();
+
+        //Test computePubkey() success/fail
+        testPubKeyCreatePos();
+        testPubKeyCreateNeg();
+
+        //Test sign() success/fail
+        testSignPos();
+        testSignNeg();
+
+        //Test Schnorr (partial support) //TODO
+        testSchnorrSign();
+        //testSchnorrVerify
+        //testSchnorrRecovery
+
+        //Test privKeyTweakAdd() 1
+        testPrivKeyTweakAdd_1();
+
+        //Test privKeyTweakMul() 2
+        testPrivKeyTweakMul_1();
+
+        //Test privKeyTweakAdd() 3
+        testPrivKeyTweakAdd_2();
+
+        //Test privKeyTweakMul() 4
+        testPrivKeyTweakMul_2();
+
+        //Test randomize()
+        testRandomize();
+
+        //Test ECDH
+        testCreateECDHSecret();
+
+        NativeSecp256k1.cleanup();
+
+        System.out.println(" All tests passed." );
+
+    }
+}
diff --git a/src/java/org/bitcoin/NativeSecp256k1Util.java b/src/java/org/bitcoin/NativeSecp256k1Util.java
new file mode 100644
index 0000000000..04732ba044
--- /dev/null
+++ b/src/java/org/bitcoin/NativeSecp256k1Util.java
@@ -0,0 +1,45 @@
+/*
+ * Copyright 2014-2016 the libsecp256k1 contributors
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ *    http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package org.bitcoin;
+
+public class NativeSecp256k1Util{
+
+    public static void assertEquals( int val, int val2, String message ) throws AssertFailException{
+      if( val != val2 )
+        throw new AssertFailException("FAIL: " + message);
+    }
+
+    public static void assertEquals( boolean val, boolean val2, String message ) throws AssertFailException{
+      if( val != val2 )
+        throw new AssertFailException("FAIL: " + message);
+      else
+        System.out.println("PASS: " + message);
+    }
+
+    public static void assertEquals( String val, String val2, String message ) throws AssertFailException{
+      if( !val.equals(val2) )
+        throw new AssertFailException("FAIL: " + message);
+      else
+        System.out.println("PASS: " + message);
+    }
+
+    public static class AssertFailException extends Exception {
+      public AssertFailException(String message) {
+        super( message );
+      }
+    }
+}
diff --git a/src/java/org/bitcoin/Secp256k1Context.java b/src/java/org/bitcoin/Secp256k1Context.java
new file mode 100644
index 0000000000..216c986a8b
--- /dev/null
+++ b/src/java/org/bitcoin/Secp256k1Context.java
@@ -0,0 +1,51 @@
+/*
+ * Copyright 2014-2016 the libsecp256k1 contributors
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ *    http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package org.bitcoin;
+
+/**
+ * This class holds the context reference used in native methods 
+ * to handle ECDSA operations.
+ */
+public class Secp256k1Context {
+  private static final boolean enabled; //true if the library is loaded
+  private static final long context; //ref to pointer to context obj
+
+  static { //static initializer
+      boolean isEnabled = true;
+      long contextRef = -1;
+      try {
+          System.loadLibrary("secp256k1");
+          contextRef = secp256k1_init_context();
+      } catch (UnsatisfiedLinkError e) {
+          System.out.println("UnsatisfiedLinkError: " + e.toString());
+          isEnabled = false;
+      }
+      enabled = isEnabled;
+      context = contextRef;
+  }
+
+  public static boolean isEnabled() {
+     return enabled;
+  }
+
+  public static long getContext() {
+     if(!enabled) return -1; //sanity check
+     return context;
+  }
+
+  private static native long secp256k1_init_context();
+}
diff --git a/src/java/org_bitcoin_NativeSecp256k1.c b/src/java/org_bitcoin_NativeSecp256k1.c
index bb4cd70728..dba9524dd4 100644
--- a/src/java/org_bitcoin_NativeSecp256k1.c
+++ b/src/java/org_bitcoin_NativeSecp256k1.c
@@ -1,23 +1,411 @@
+#include <stdlib.h>
+#include <stdint.h>
+#include <string.h>
 #include "org_bitcoin_NativeSecp256k1.h"
 #include "include/secp256k1.h"
+#include "include/secp256k1_ecdh.h"
+#include "include/secp256k1_recovery.h"
+#include "include/secp256k1_schnorr.h"
 
-JNIEXPORT jint JNICALL Java_org_bitcoin_NativeSecp256k1_secp256k1_1ecdsa_1verify
-  (JNIEnv* env, jclass classObject, jobject byteBufferObject)
+
+SECP256K1_API jlong JNICALL Java_org_bitcoin_NativeSecp256k1_secp256k1_1ctx_1clone
+  (JNIEnv* env, jclass classObject, jlong ctx_l)
+{
+  const secp256k1_context *ctx = (secp256k1_context*)(uintptr_t)ctx_l;
+
+  jlong ctx_clone_l = (uintptr_t) secp256k1_context_clone(ctx);
+
+  (void)classObject;(void)env;
+
+  return ctx_clone_l;
+
+}
+
+SECP256K1_API jint JNICALL Java_org_bitcoin_NativeSecp256k1_secp256k1_1context_1randomize
+  (JNIEnv* env, jclass classObject, jobject byteBufferObject, jlong ctx_l)
+{
+  secp256k1_context *ctx = (secp256k1_context*)(uintptr_t)ctx_l;
+
+  const unsigned char* seed = (unsigned char*) (*env)->GetDirectBufferAddress(env, byteBufferObject);
+
+  (void)classObject;
+
+  return secp256k1_context_randomize(ctx, seed);
+
+}
+
+SECP256K1_API void JNICALL Java_org_bitcoin_NativeSecp256k1_secp256k1_1destroy_1context
+  (JNIEnv* env, jclass classObject, jlong ctx_l)
+{
+  secp256k1_context *ctx = (secp256k1_context*)(uintptr_t)ctx_l;
+
+  secp256k1_context_destroy(ctx);
+
+  (void)classObject;(void)env;
+}
+
+SECP256K1_API jint JNICALL Java_org_bitcoin_NativeSecp256k1_secp256k1_1ecdsa_1verify
+  (JNIEnv* env, jclass classObject, jobject byteBufferObject, jlong ctx_l, jint siglen, jint publen)
+{
+  secp256k1_context *ctx = (secp256k1_context*)(uintptr_t)ctx_l;
+
+  unsigned char* data = (unsigned char*) (*env)->GetDirectBufferAddress(env, byteBufferObject);
+  const unsigned char* sigdata = {  (unsigned char*) (data + 32) };
+  const unsigned char* pubdata = { (unsigned char*) (data + siglen + 32) };
+
+  secp256k1_ecdsa_signature sig;
+  secp256k1_pubkey pubkey;
+
+  int ret = secp256k1_ecdsa_signature_parse_der(ctx, &sig, sigdata, siglen);
+
+  if( ret ) {
+    ret = secp256k1_ec_pubkey_parse(ctx, &pubkey, pubdata, publen);
+
+    if( ret ) {
+      ret = secp256k1_ecdsa_verify(ctx, &sig, data, &pubkey);
+    }
+  }
+
+  (void)classObject;
+
+  return ret;
+}
+
+SECP256K1_API jobjectArray JNICALL Java_org_bitcoin_NativeSecp256k1_secp256k1_1ecdsa_1sign
+  (JNIEnv* env, jclass classObject, jobject byteBufferObject, jlong ctx_l)
+{
+  secp256k1_context *ctx = (secp256k1_context*)(uintptr_t)ctx_l;
+  unsigned char* data = (unsigned char*) (*env)->GetDirectBufferAddress(env, byteBufferObject);
+  unsigned char* secKey = (unsigned char*) (data + 32);
+
+  jobjectArray retArray;
+  jbyteArray sigArray, intsByteArray;
+  unsigned char intsarray[2];
+
+  secp256k1_ecdsa_signature sig[72];
+
+  int ret = secp256k1_ecdsa_sign(ctx, sig, data, secKey, NULL, NULL );
+
+  unsigned char outputSer[72];
+  size_t outputLen = 72;
+
+  if( ret ) {
+    int ret2 = secp256k1_ecdsa_signature_serialize_der(ctx,outputSer, &outputLen, sig ); (void)ret2;
+  }
+
+  intsarray[0] = outputLen;
+  intsarray[1] = ret;
+
+  retArray = (*env)->NewObjectArray(env, 2,
+    (*env)->FindClass(env, "[B"),
+    (*env)->NewByteArray(env, 1));
+
+  sigArray = (*env)->NewByteArray(env, outputLen);
+  (*env)->SetByteArrayRegion(env, sigArray, 0, outputLen, (jbyte*)outputSer);
+  (*env)->SetObjectArrayElement(env, retArray, 0, sigArray);
+
+  intsByteArray = (*env)->NewByteArray(env, 2);
+  (*env)->SetByteArrayRegion(env, intsByteArray, 0, 2, (jbyte*)intsarray);
+  (*env)->SetObjectArrayElement(env, retArray, 1, intsByteArray);
+
+  (void)classObject;
+
+  return retArray;
+}
+
+SECP256K1_API jint JNICALL Java_org_bitcoin_NativeSecp256k1_secp256k1_1ec_1seckey_1verify
+  (JNIEnv* env, jclass classObject, jobject byteBufferObject, jlong ctx_l)
+{
+  secp256k1_context *ctx = (secp256k1_context*)(uintptr_t)ctx_l;
+  unsigned char* secKey = (unsigned char*) (*env)->GetDirectBufferAddress(env, byteBufferObject);
+
+  (void)classObject;
+
+  return secp256k1_ec_seckey_verify(ctx, secKey);
+}
+
+SECP256K1_API jobjectArray JNICALL Java_org_bitcoin_NativeSecp256k1_secp256k1_1ec_1pubkey_1create
+  (JNIEnv* env, jclass classObject, jobject byteBufferObject, jlong ctx_l)
+{
+  secp256k1_context *ctx = (secp256k1_context*)(uintptr_t)ctx_l;
+  const unsigned char* secKey = (unsigned char*) (*env)->GetDirectBufferAddress(env, byteBufferObject);
+
+  secp256k1_pubkey pubkey;
+
+  jobjectArray retArray;
+  jbyteArray pubkeyArray, intsByteArray;
+  unsigned char intsarray[2];
+
+  int ret = secp256k1_ec_pubkey_create(ctx, &pubkey, secKey);
+
+  unsigned char outputSer[65];
+  size_t outputLen = 65;
+
+  if( ret ) {
+    int ret2 = secp256k1_ec_pubkey_serialize(ctx,outputSer, &outputLen, &pubkey,SECP256K1_EC_UNCOMPRESSED );(void)ret2;
+  }
+
+  intsarray[0] = outputLen;
+  intsarray[1] = ret;
+
+  retArray = (*env)->NewObjectArray(env, 2,
+    (*env)->FindClass(env, "[B"),
+    (*env)->NewByteArray(env, 1));
+
+  pubkeyArray = (*env)->NewByteArray(env, outputLen);
+  (*env)->SetByteArrayRegion(env, pubkeyArray, 0, outputLen, (jbyte*)outputSer);
+  (*env)->SetObjectArrayElement(env, retArray, 0, pubkeyArray);
+
+  intsByteArray = (*env)->NewByteArray(env, 2);
+  (*env)->SetByteArrayRegion(env, intsByteArray, 0, 2, (jbyte*)intsarray);
+  (*env)->SetObjectArrayElement(env, retArray, 1, intsByteArray);
+
+  (void)classObject;
+
+  return retArray;
+
+}
+
+SECP256K1_API jobjectArray JNICALL Java_org_bitcoin_NativeSecp256k1_secp256k1_1privkey_1tweak_1add
+  (JNIEnv* env, jclass classObject, jobject byteBufferObject, jlong ctx_l)
+{
+  secp256k1_context *ctx = (secp256k1_context*)(uintptr_t)ctx_l;
+  unsigned char* privkey = (unsigned char*) (*env)->GetDirectBufferAddress(env, byteBufferObject);
+  const unsigned char* tweak = (unsigned char*) (privkey + 32);
+
+  jobjectArray retArray;
+  jbyteArray privArray, intsByteArray;
+  unsigned char intsarray[2];
+
+  int privkeylen = 32;
+
+  int ret = secp256k1_ec_privkey_tweak_add(ctx, privkey, tweak);
+
+  intsarray[0] = privkeylen;
+  intsarray[1] = ret;
+
+  retArray = (*env)->NewObjectArray(env, 2,
+    (*env)->FindClass(env, "[B"),
+    (*env)->NewByteArray(env, 1));
+
+  privArray = (*env)->NewByteArray(env, privkeylen);
+  (*env)->SetByteArrayRegion(env, privArray, 0, privkeylen, (jbyte*)privkey);
+  (*env)->SetObjectArrayElement(env, retArray, 0, privArray);
+
+  intsByteArray = (*env)->NewByteArray(env, 2);
+  (*env)->SetByteArrayRegion(env, intsByteArray, 0, 2, (jbyte*)intsarray);
+  (*env)->SetObjectArrayElement(env, retArray, 1, intsByteArray);
+
+  (void)classObject;
+
+  return retArray;
+}
+
+SECP256K1_API jobjectArray JNICALL Java_org_bitcoin_NativeSecp256k1_secp256k1_1privkey_1tweak_1mul
+  (JNIEnv* env, jclass classObject, jobject byteBufferObject, jlong ctx_l)
+{
+  secp256k1_context *ctx = (secp256k1_context*)(uintptr_t)ctx_l;
+  unsigned char* privkey = (unsigned char*) (*env)->GetDirectBufferAddress(env, byteBufferObject);
+  const unsigned char* tweak = (unsigned char*) (privkey + 32);
+
+  jobjectArray retArray;
+  jbyteArray privArray, intsByteArray;
+  unsigned char intsarray[2];
+
+  int privkeylen = 32;
+
+  int ret = secp256k1_ec_privkey_tweak_mul(ctx, privkey, tweak);
+
+  intsarray[0] = privkeylen;
+  intsarray[1] = ret;
+
+  retArray = (*env)->NewObjectArray(env, 2,
+    (*env)->FindClass(env, "[B"),
+    (*env)->NewByteArray(env, 1));
+
+  privArray = (*env)->NewByteArray(env, privkeylen);
+  (*env)->SetByteArrayRegion(env, privArray, 0, privkeylen, (jbyte*)privkey);
+  (*env)->SetObjectArrayElement(env, retArray, 0, privArray);
+
+  intsByteArray = (*env)->NewByteArray(env, 2);
+  (*env)->SetByteArrayRegion(env, intsByteArray, 0, 2, (jbyte*)intsarray);
+  (*env)->SetObjectArrayElement(env, retArray, 1, intsByteArray);
+
+  (void)classObject;
+
+  return retArray;
+}
+
+SECP256K1_API jobjectArray JNICALL Java_org_bitcoin_NativeSecp256k1_secp256k1_1pubkey_1tweak_1add
+  (JNIEnv* env, jclass classObject, jobject byteBufferObject, jlong ctx_l, jint publen)
+{
+  secp256k1_context *ctx = (secp256k1_context*)(uintptr_t)ctx_l;
+/*  secp256k1_pubkey* pubkey = (secp256k1_pubkey*) (*env)->GetDirectBufferAddress(env, byteBufferObject);*/
+  unsigned char* pkey = (*env)->GetDirectBufferAddress(env, byteBufferObject);
+  const unsigned char* tweak = (unsigned char*) (pkey + publen);
+
+  jobjectArray retArray;
+  jbyteArray pubArray, intsByteArray;
+  unsigned char intsarray[2];
+  unsigned char outputSer[65];
+  size_t outputLen = 65;
+
+  secp256k1_pubkey pubkey;
+  int ret = secp256k1_ec_pubkey_parse(ctx, &pubkey, pkey, publen);
+
+  if( ret ) {
+    ret = secp256k1_ec_pubkey_tweak_add(ctx, &pubkey, tweak);
+  }
+
+  if( ret ) {
+    int ret2 = secp256k1_ec_pubkey_serialize(ctx,outputSer, &outputLen, &pubkey,SECP256K1_EC_UNCOMPRESSED );(void)ret2;
+  }
+
+  intsarray[0] = outputLen;
+  intsarray[1] = ret;
+
+  retArray = (*env)->NewObjectArray(env, 2,
+    (*env)->FindClass(env, "[B"),
+    (*env)->NewByteArray(env, 1));
+
+  pubArray = (*env)->NewByteArray(env, outputLen);
+  (*env)->SetByteArrayRegion(env, pubArray, 0, outputLen, (jbyte*)outputSer);
+  (*env)->SetObjectArrayElement(env, retArray, 0, pubArray);
+
+  intsByteArray = (*env)->NewByteArray(env, 2);
+  (*env)->SetByteArrayRegion(env, intsByteArray, 0, 2, (jbyte*)intsarray);
+  (*env)->SetObjectArrayElement(env, retArray, 1, intsByteArray);
+
+  (void)classObject;
+
+  return retArray;
+}
+
+SECP256K1_API jobjectArray JNICALL Java_org_bitcoin_NativeSecp256k1_secp256k1_1pubkey_1tweak_1mul
+  (JNIEnv* env, jclass classObject, jobject byteBufferObject, jlong ctx_l, jint publen)
 {
-	unsigned char* data = (unsigned char*) (*env)->GetDirectBufferAddress(env, byteBufferObject);
-	int sigLen = *((int*)(data + 32));
-	int pubLen = *((int*)(data + 32 + 4));
+  secp256k1_context *ctx = (secp256k1_context*)(uintptr_t)ctx_l;
+  unsigned char* pkey = (*env)->GetDirectBufferAddress(env, byteBufferObject);
+  const unsigned char* tweak = (unsigned char*) (pkey + publen);
+
+  jobjectArray retArray;
+  jbyteArray pubArray, intsByteArray;
+  unsigned char intsarray[2];
+  unsigned char outputSer[65];
+  size_t outputLen = 65;
+
+  secp256k1_pubkey pubkey;
+  int ret = secp256k1_ec_pubkey_parse(ctx, &pubkey, pkey, publen);
+
+  if ( ret ) {
+    ret = secp256k1_ec_pubkey_tweak_mul(ctx, &pubkey, tweak);
+  }
+
+  if( ret ) {
+    int ret2 = secp256k1_ec_pubkey_serialize(ctx,outputSer, &outputLen, &pubkey,SECP256K1_EC_UNCOMPRESSED );(void)ret2;
+  }
+
+  intsarray[0] = outputLen;
+  intsarray[1] = ret;
+
+  retArray = (*env)->NewObjectArray(env, 2,
+    (*env)->FindClass(env, "[B"),
+    (*env)->NewByteArray(env, 1));
+
+  pubArray = (*env)->NewByteArray(env, outputLen);
+  (*env)->SetByteArrayRegion(env, pubArray, 0, outputLen, (jbyte*)outputSer);
+  (*env)->SetObjectArrayElement(env, retArray, 0, pubArray);
 
-	return secp256k1_ecdsa_verify(data, 32, data+32+8, sigLen, data+32+8+sigLen, pubLen);
+  intsByteArray = (*env)->NewByteArray(env, 2);
+  (*env)->SetByteArrayRegion(env, intsByteArray, 0, 2, (jbyte*)intsarray);
+  (*env)->SetObjectArrayElement(env, retArray, 1, intsByteArray);
+
+  (void)classObject;
+
+  return retArray;
 }
 
-static void __javasecp256k1_attach(void) __attribute__((constructor));
-static void __javasecp256k1_detach(void) __attribute__((destructor));
+SECP256K1_API jlong JNICALL Java_org_bitcoin_NativeSecp256k1_secp256k1_1ecdsa_1pubkey_1combine
+  (JNIEnv * env, jclass classObject, jobject byteBufferObject, jlong ctx_l, jint numkeys)
+{
+  (void)classObject;(void)env;(void)byteBufferObject;(void)ctx_l;(void)numkeys;
 
-static void __javasecp256k1_attach(void) {
-	secp256k1_start(SECP256K1_START_VERIFY);
+  return 0;
 }
 
-static void __javasecp256k1_detach(void) {
-	secp256k1_stop();
+SECP256K1_API jobjectArray JNICALL Java_org_bitcoin_NativeSecp256k1_secp256k1_1schnorr_1sign
+  (JNIEnv* env, jclass classObject, jobject byteBufferObject, jlong ctx_l)
+{
+  secp256k1_context *ctx = (secp256k1_context*)(uintptr_t)ctx_l;
+  unsigned char* data = (unsigned char*) (*env)->GetDirectBufferAddress(env, byteBufferObject);
+  unsigned char* secKey = (unsigned char*) (data + 32);
+
+  jobjectArray retArray;
+  jbyteArray sigArray, intsByteArray;
+  unsigned char intsarray[1];
+  unsigned char sig[64];
+
+  int ret = secp256k1_schnorr_sign(ctx, sig, data, secKey, NULL, NULL);
+
+  intsarray[0] = ret;
+
+  retArray = (*env)->NewObjectArray(env, 2,
+    (*env)->FindClass(env, "[B"),
+    (*env)->NewByteArray(env, 1));
+
+  sigArray = (*env)->NewByteArray(env, 64);
+  (*env)->SetByteArrayRegion(env, sigArray, 0, 64, (jbyte*)sig);
+  (*env)->SetObjectArrayElement(env, retArray, 0, sigArray);
+
+  intsByteArray = (*env)->NewByteArray(env, 1);
+  (*env)->SetByteArrayRegion(env, intsByteArray, 0, 1, (jbyte*)intsarray);
+  (*env)->SetObjectArrayElement(env, retArray, 1, intsByteArray);
+
+  (void)classObject;
+
+  return retArray;
+}
+
+SECP256K1_API jobjectArray JNICALL Java_org_bitcoin_NativeSecp256k1_secp256k1_1ecdh
+  (JNIEnv* env, jclass classObject, jobject byteBufferObject, jlong ctx_l, jint publen)
+{
+  secp256k1_context *ctx = (secp256k1_context*)(uintptr_t)ctx_l;
+  const unsigned char* secdata = (*env)->GetDirectBufferAddress(env, byteBufferObject);
+  const unsigned char* pubdata = (const unsigned char*) (secdata + 32);
+
+  jobjectArray retArray;
+  jbyteArray outArray, intsByteArray;
+  unsigned char intsarray[1];
+  secp256k1_pubkey pubkey;
+  unsigned char nonce_res[32];
+  size_t outputLen = 32;
+
+  int ret = secp256k1_ec_pubkey_parse(ctx, &pubkey, pubdata, publen);
+
+  if (ret) {
+    ret = secp256k1_ecdh(
+      ctx,
+      nonce_res,
+      &pubkey,
+      secdata
+    );
+  }
+
+  intsarray[0] = ret;
+
+  retArray = (*env)->NewObjectArray(env, 2,
+    (*env)->FindClass(env, "[B"),
+    (*env)->NewByteArray(env, 1));
+
+  outArray = (*env)->NewByteArray(env, outputLen);
+  (*env)->SetByteArrayRegion(env, outArray, 0, 32, (jbyte*)nonce_res);
+  (*env)->SetObjectArrayElement(env, retArray, 0, outArray);
+
+  intsByteArray = (*env)->NewByteArray(env, 1);
+  (*env)->SetByteArrayRegion(env, intsByteArray, 0, 1, (jbyte*)intsarray);
+  (*env)->SetObjectArrayElement(env, retArray, 1, intsByteArray);
+
+  (void)classObject;
+
+  return retArray;
 }
diff --git a/src/java/org_bitcoin_NativeSecp256k1.h b/src/java/org_bitcoin_NativeSecp256k1.h
index d7fb004fa8..4125a1f523 100644
--- a/src/java/org_bitcoin_NativeSecp256k1.h
+++ b/src/java/org_bitcoin_NativeSecp256k1.h
@@ -1,5 +1,6 @@
 /* DO NOT EDIT THIS FILE - it is machine generated */
 #include <jni.h>
+#include "include/secp256k1.h"
 /* Header for class org_bitcoin_NativeSecp256k1 */
 
 #ifndef _Included_org_bitcoin_NativeSecp256k1
@@ -9,11 +10,116 @@ extern "C" {
 #endif
 /*
  * Class:     org_bitcoin_NativeSecp256k1
+ * Method:    secp256k1_ctx_clone
+ * Signature: (J)J
+ */
+SECP256K1_API jlong JNICALL Java_org_bitcoin_NativeSecp256k1_secp256k1_1ctx_1clone
+  (JNIEnv *, jclass, jlong);
+
+/*
+ * Class:     org_bitcoin_NativeSecp256k1
+ * Method:    secp256k1_context_randomize
+ * Signature: (Ljava/nio/ByteBuffer;J)I
+ */
+SECP256K1_API jint JNICALL Java_org_bitcoin_NativeSecp256k1_secp256k1_1context_1randomize
+  (JNIEnv *, jclass, jobject, jlong);
+
+/*
+ * Class:     org_bitcoin_NativeSecp256k1
+ * Method:    secp256k1_privkey_tweak_add
+ * Signature: (Ljava/nio/ByteBuffer;J)[[B
+ */
+SECP256K1_API jobjectArray JNICALL Java_org_bitcoin_NativeSecp256k1_secp256k1_1privkey_1tweak_1add
+  (JNIEnv *, jclass, jobject, jlong);
+
+/*
+ * Class:     org_bitcoin_NativeSecp256k1
+ * Method:    secp256k1_privkey_tweak_mul
+ * Signature: (Ljava/nio/ByteBuffer;J)[[B
+ */
+SECP256K1_API jobjectArray JNICALL Java_org_bitcoin_NativeSecp256k1_secp256k1_1privkey_1tweak_1mul
+  (JNIEnv *, jclass, jobject, jlong);
+
+/*
+ * Class:     org_bitcoin_NativeSecp256k1
+ * Method:    secp256k1_pubkey_tweak_add
+ * Signature: (Ljava/nio/ByteBuffer;JI)[[B
+ */
+SECP256K1_API jobjectArray JNICALL Java_org_bitcoin_NativeSecp256k1_secp256k1_1pubkey_1tweak_1add
+  (JNIEnv *, jclass, jobject, jlong, jint);
+
+/*
+ * Class:     org_bitcoin_NativeSecp256k1
+ * Method:    secp256k1_pubkey_tweak_mul
+ * Signature: (Ljava/nio/ByteBuffer;JI)[[B
+ */
+SECP256K1_API jobjectArray JNICALL Java_org_bitcoin_NativeSecp256k1_secp256k1_1pubkey_1tweak_1mul
+  (JNIEnv *, jclass, jobject, jlong, jint);
+
+/*
+ * Class:     org_bitcoin_NativeSecp256k1
+ * Method:    secp256k1_destroy_context
+ * Signature: (J)V
+ */
+SECP256K1_API void JNICALL Java_org_bitcoin_NativeSecp256k1_secp256k1_1destroy_1context
+  (JNIEnv *, jclass, jlong);
+
+/*
+ * Class:     org_bitcoin_NativeSecp256k1
  * Method:    secp256k1_ecdsa_verify
- * Signature: (Ljava/nio/ByteBuffer;)I
+ * Signature: (Ljava/nio/ByteBuffer;JII)I
+ */
+SECP256K1_API jint JNICALL Java_org_bitcoin_NativeSecp256k1_secp256k1_1ecdsa_1verify
+  (JNIEnv *, jclass, jobject, jlong, jint, jint);
+
+/*
+ * Class:     org_bitcoin_NativeSecp256k1
+ * Method:    secp256k1_ecdsa_sign
+ * Signature: (Ljava/nio/ByteBuffer;J)[[B
+ */
+SECP256K1_API jobjectArray JNICALL Java_org_bitcoin_NativeSecp256k1_secp256k1_1ecdsa_1sign
+  (JNIEnv *, jclass, jobject, jlong);
+
+/*
+ * Class:     org_bitcoin_NativeSecp256k1
+ * Method:    secp256k1_ec_seckey_verify
+ * Signature: (Ljava/nio/ByteBuffer;J)I
+ */
+SECP256K1_API jint JNICALL Java_org_bitcoin_NativeSecp256k1_secp256k1_1ec_1seckey_1verify
+  (JNIEnv *, jclass, jobject, jlong);
+
+/*
+ * Class:     org_bitcoin_NativeSecp256k1
+ * Method:    secp256k1_ec_pubkey_create
+ * Signature: (Ljava/nio/ByteBuffer;J)[[B
  */
-JNIEXPORT jint JNICALL Java_org_bitcoin_NativeSecp256k1_secp256k1_1ecdsa_1verify
-  (JNIEnv *, jclass, jobject);
+SECP256K1_API jobjectArray JNICALL Java_org_bitcoin_NativeSecp256k1_secp256k1_1ec_1pubkey_1create
+  (JNIEnv *, jclass, jobject, jlong);
+
+/*
+ * Class:     org_bitcoin_NativeSecp256k1
+ * Method:    secp256k1_ec_pubkey_parse
+ * Signature: (Ljava/nio/ByteBuffer;JI)[[B
+ */
+SECP256K1_API jobjectArray JNICALL Java_org_bitcoin_NativeSecp256k1_secp256k1_1ec_1pubkey_1parse
+  (JNIEnv *, jclass, jobject, jlong, jint);
+
+/*
+ * Class:     org_bitcoin_NativeSecp256k1
+ * Method:    secp256k1_schnorr_sign
+ * Signature: (Ljava/nio/ByteBuffer;JI)[[B
+ */
+SECP256K1_API jobjectArray JNICALL Java_org_bitcoin_NativeSecp256k1_secp256k1_1schnorr_1sign
+  (JNIEnv* env, jclass classObject, jobject byteBufferObject, jlong ctx_l);
+
+/*
+ * Class:     org_bitcoin_NativeSecp256k1
+ * Method:    secp256k1_ecdh
+ * Signature: (Ljava/nio/ByteBuffer;JI)[[B
+ */
+SECP256K1_API jobjectArray JNICALL Java_org_bitcoin_NativeSecp256k1_secp256k1_1ecdh
+  (JNIEnv* env, jclass classObject, jobject byteBufferObject, jlong ctx_l, jint publen);
+
 
 #ifdef __cplusplus
 }
diff --git a/src/java/org_bitcoin_Secp256k1Context.c b/src/java/org_bitcoin_Secp256k1Context.c
new file mode 100644
index 0000000000..a52939e7e7
--- /dev/null
+++ b/src/java/org_bitcoin_Secp256k1Context.c
@@ -0,0 +1,15 @@
+#include <stdlib.h>
+#include <stdint.h>
+#include "org_bitcoin_Secp256k1Context.h"
+#include "include/secp256k1.h"
+
+SECP256K1_API jlong JNICALL Java_org_bitcoin_Secp256k1Context_secp256k1_1init_1context
+  (JNIEnv* env, jclass classObject)
+{
+  secp256k1_context *ctx = secp256k1_context_create(SECP256K1_CONTEXT_SIGN | SECP256K1_CONTEXT_VERIFY);
+
+  (void)classObject;(void)env;
+
+  return (uintptr_t)ctx;
+}
+
diff --git a/src/java/org_bitcoin_Secp256k1Context.h b/src/java/org_bitcoin_Secp256k1Context.h
new file mode 100644
index 0000000000..0d2bc84b7f
--- /dev/null
+++ b/src/java/org_bitcoin_Secp256k1Context.h
@@ -0,0 +1,22 @@
+/* DO NOT EDIT THIS FILE - it is machine generated */
+#include <jni.h>
+#include "include/secp256k1.h"
+/* Header for class org_bitcoin_Secp256k1Context */
+
+#ifndef _Included_org_bitcoin_Secp256k1Context
+#define _Included_org_bitcoin_Secp256k1Context
+#ifdef __cplusplus
+extern "C" {
+#endif
+/*
+ * Class:     org_bitcoin_Secp256k1Context
+ * Method:    secp256k1_init_context
+ * Signature: ()J
+ */
+SECP256K1_API jlong JNICALL Java_org_bitcoin_Secp256k1Context_secp256k1_1init_1context
+  (JNIEnv *, jclass);
+
+#ifdef __cplusplus
+}
+#endif
+#endif
diff --git a/src/modules/ecdh/Makefile.am.include b/src/modules/ecdh/Makefile.am.include
index 670b9c1152..e3088b4697 100644
--- a/src/modules/ecdh/Makefile.am.include
+++ b/src/modules/ecdh/Makefile.am.include
@@ -4,5 +4,5 @@ noinst_HEADERS += src/modules/ecdh/tests_impl.h
 if USE_BENCHMARK
 noinst_PROGRAMS += bench_ecdh
 bench_ecdh_SOURCES = src/bench_ecdh.c
-bench_ecdh_LDADD = libsecp256k1.la $(SECP_LIBS)
+bench_ecdh_LDADD = libsecp256k1.la $(SECP_LIBS) $(COMMON_LIB)
 endif
diff --git a/src/modules/recovery/Makefile.am.include b/src/modules/recovery/Makefile.am.include
index 5de3ea33ea..bf23c26e71 100644
--- a/src/modules/recovery/Makefile.am.include
+++ b/src/modules/recovery/Makefile.am.include
@@ -4,5 +4,5 @@ noinst_HEADERS += src/modules/recovery/tests_impl.h
 if USE_BENCHMARK
 noinst_PROGRAMS += bench_recover
 bench_recover_SOURCES = src/bench_recover.c
-bench_recover_LDADD = libsecp256k1.la $(SECP_LIBS)
+bench_recover_LDADD = libsecp256k1.la $(SECP_LIBS) $(COMMON_LIB)
 endif
diff --git a/src/modules/schnorr/Makefile.am.include b/src/modules/schnorr/Makefile.am.include
index b3bfa7d5cc..f1af8e8325 100644
--- a/src/modules/schnorr/Makefile.am.include
+++ b/src/modules/schnorr/Makefile.am.include
@@ -6,5 +6,5 @@ noinst_HEADERS += src/modules/schnorr/tests_impl.h
 if USE_BENCHMARK
 noinst_PROGRAMS += bench_schnorr_verify
 bench_schnorr_verify_SOURCES = src/bench_schnorr_verify.c
-bench_schnorr_verify_LDADD = libsecp256k1.la $(SECP_LIBS)
+bench_schnorr_verify_LDADD = libsecp256k1.la $(SECP_LIBS) $(COMMON_LIB)
 endif
diff --git a/src/num.h b/src/num.h
index ebfa71eb44..7bb9c5be8c 100644
--- a/src/num.h
+++ b/src/num.h
@@ -32,6 +32,9 @@ static void secp256k1_num_set_bin(secp256k1_num *r, const unsigned char *a, unsi
 /** Compute a modular inverse. The input must be less than the modulus. */
 static void secp256k1_num_mod_inverse(secp256k1_num *r, const secp256k1_num *a, const secp256k1_num *m);
 
+/** Compute the jacobi symbol (a|b). b must be positive and odd. */
+static int secp256k1_num_jacobi(const secp256k1_num *a, const secp256k1_num *b);
+
 /** Compare the absolute value of two numbers. */
 static int secp256k1_num_cmp(const secp256k1_num *a, const secp256k1_num *b);
 
@@ -57,6 +60,9 @@ static void secp256k1_num_shift(secp256k1_num *r, int bits);
 /** Check whether a number is zero. */
 static int secp256k1_num_is_zero(const secp256k1_num *a);
 
+/** Check whether a number is one. */
+static int secp256k1_num_is_one(const secp256k1_num *a);
+
 /** Check whether a number is strictly negative. */
 static int secp256k1_num_is_neg(const secp256k1_num *a);
 
diff --git a/src/num_gmp_impl.h b/src/num_gmp_impl.h
index 7b6a89719a..3a46495eea 100644
--- a/src/num_gmp_impl.h
+++ b/src/num_gmp_impl.h
@@ -144,6 +144,32 @@ static void secp256k1_num_mod_inverse(secp256k1_num *r, const secp256k1_num *a,
     memset(v, 0, sizeof(v));
 }
 
+static int secp256k1_num_jacobi(const secp256k1_num *a, const secp256k1_num *b) {
+    int ret;
+    mpz_t ga, gb;
+    secp256k1_num_sanity(a);
+    secp256k1_num_sanity(b);
+    VERIFY_CHECK(!b->neg && (b->limbs > 0) && (b->data[0] & 1));
+
+    mpz_inits(ga, gb, NULL);
+
+    mpz_import(gb, b->limbs, -1, sizeof(mp_limb_t), 0, 0, b->data);
+    mpz_import(ga, a->limbs, -1, sizeof(mp_limb_t), 0, 0, a->data);
+    if (a->neg) {
+        mpz_neg(ga, ga);
+    }
+
+    ret = mpz_jacobi(ga, gb);
+
+    mpz_clears(ga, gb, NULL);
+
+    return ret;
+}
+
+static int secp256k1_num_is_one(const secp256k1_num *a) {
+    return (a->limbs == 1 && a->data[0] == 1);
+}
+
 static int secp256k1_num_is_zero(const secp256k1_num *a) {
     return (a->limbs == 1 && a->data[0] == 0);
 }
diff --git a/src/scalar_impl.h b/src/scalar_impl.h
index 88ea97de86..c5baf4df41 100644
--- a/src/scalar_impl.h
+++ b/src/scalar_impl.h
@@ -7,8 +7,6 @@
 #ifndef _SECP256K1_SCALAR_IMPL_H_
 #define _SECP256K1_SCALAR_IMPL_H_
 
-#include <string.h>
-
 #include "group.h"
 #include "scalar.h"
 
diff --git a/src/secp256k1.c b/src/secp256k1.c
index 62d192baeb..7973d60c36 100644
--- a/src/secp256k1.c
+++ b/src/secp256k1.c
@@ -4,8 +4,6 @@
  * file COPYING or http://www.opensource.org/licenses/mit-license.php.*
  **********************************************************************/
 
-#define SECP256K1_BUILD (1)
-
 #include "include/secp256k1.h"
 
 #include "util.h"
@@ -152,7 +150,6 @@ static void secp256k1_pubkey_save(secp256k1_pubkey* pubkey, secp256k1_ge* ge) {
 int secp256k1_ec_pubkey_parse(const secp256k1_context* ctx, secp256k1_pubkey* pubkey, const unsigned char *input, size_t inputlen) {
     secp256k1_ge Q;
 
-    (void)ctx;
     VERIFY_CHECK(ctx != NULL);
     ARG_CHECK(pubkey != NULL);
     memset(pubkey, 0, sizeof(*pubkey));
@@ -170,7 +167,6 @@ int secp256k1_ec_pubkey_serialize(const secp256k1_context* ctx, unsigned char *o
     size_t len;
     int ret = 0;
 
-    (void)ctx;
     VERIFY_CHECK(ctx != NULL);
     ARG_CHECK(outputlen != NULL);
     ARG_CHECK(*outputlen >= ((flags & SECP256K1_FLAGS_BIT_COMPRESSION) ? 33 : 65));
@@ -216,7 +212,7 @@ static void secp256k1_ecdsa_signature_save(secp256k1_ecdsa_signature* sig, const
 int secp256k1_ecdsa_signature_parse_der(const secp256k1_context* ctx, secp256k1_ecdsa_signature* sig, const unsigned char *input, size_t inputlen) {
     secp256k1_scalar r, s;
 
-    (void)ctx;
+    VERIFY_CHECK(ctx != NULL);
     ARG_CHECK(sig != NULL);
     ARG_CHECK(input != NULL);
 
@@ -234,7 +230,7 @@ int secp256k1_ecdsa_signature_parse_compact(const secp256k1_context* ctx, secp25
     int ret = 1;
     int overflow = 0;
 
-    (void)ctx;
+    VERIFY_CHECK(ctx != NULL);
     ARG_CHECK(sig != NULL);
     ARG_CHECK(input64 != NULL);
 
@@ -253,7 +249,7 @@ int secp256k1_ecdsa_signature_parse_compact(const secp256k1_context* ctx, secp25
 int secp256k1_ecdsa_signature_serialize_der(const secp256k1_context* ctx, unsigned char *output, size_t *outputlen, const secp256k1_ecdsa_signature* sig) {
     secp256k1_scalar r, s;
 
-    (void)ctx;
+    VERIFY_CHECK(ctx != NULL);
     ARG_CHECK(output != NULL);
     ARG_CHECK(outputlen != NULL);
     ARG_CHECK(sig != NULL);
@@ -265,7 +261,7 @@ int secp256k1_ecdsa_signature_serialize_der(const secp256k1_context* ctx, unsign
 int secp256k1_ecdsa_signature_serialize_compact(const secp256k1_context* ctx, unsigned char *output64, const secp256k1_ecdsa_signature* sig) {
     secp256k1_scalar r, s;
 
-    (void)ctx;
+    VERIFY_CHECK(ctx != NULL);
     ARG_CHECK(output64 != NULL);
     ARG_CHECK(sig != NULL);
 
@@ -398,7 +394,6 @@ int secp256k1_ec_seckey_verify(const secp256k1_context* ctx, const unsigned char
     int overflow;
     VERIFY_CHECK(ctx != NULL);
     ARG_CHECK(seckey != NULL);
-    (void)ctx;
 
     secp256k1_scalar_set_b32(&sec, seckey, &overflow);
     ret = !overflow && !secp256k1_scalar_is_zero(&sec);
@@ -437,7 +432,6 @@ int secp256k1_ec_privkey_tweak_add(const secp256k1_context* ctx, unsigned char *
     VERIFY_CHECK(ctx != NULL);
     ARG_CHECK(seckey != NULL);
     ARG_CHECK(tweak != NULL);
-    (void)ctx;
 
     secp256k1_scalar_set_b32(&term, tweak, &overflow);
     secp256k1_scalar_set_b32(&sec, seckey, NULL);
@@ -485,7 +479,6 @@ int secp256k1_ec_privkey_tweak_mul(const secp256k1_context* ctx, unsigned char *
     VERIFY_CHECK(ctx != NULL);
     ARG_CHECK(seckey != NULL);
     ARG_CHECK(tweak != NULL);
-    (void)ctx;
 
     secp256k1_scalar_set_b32(&factor, tweak, &overflow);
     secp256k1_scalar_set_b32(&sec, seckey, NULL);
diff --git a/src/tests.c b/src/tests.c
index 687a5f2fdd..b32cb90813 100644
--- a/src/tests.c
+++ b/src/tests.c
@@ -473,6 +473,8 @@ void test_num_negate(void) {
 }
 
 void test_num_add_sub(void) {
+    int i;
+    secp256k1_scalar s;
     secp256k1_num n1;
     secp256k1_num n2;
     secp256k1_num n1p2, n2p1, n1m2, n2m1;
@@ -498,6 +500,110 @@ void test_num_add_sub(void) {
     CHECK(!secp256k1_num_eq(&n2p1, &n1));
     secp256k1_num_sub(&n2p1, &n2p1, &n2); /* n2p1 = R2 + R1 - R2 = R1 */
     CHECK(secp256k1_num_eq(&n2p1, &n1));
+
+    /* check is_one */
+    secp256k1_scalar_set_int(&s, 1);
+    secp256k1_scalar_get_num(&n1, &s);
+    CHECK(secp256k1_num_is_one(&n1));
+    /* check that 2^n + 1 is never 1 */
+    secp256k1_scalar_get_num(&n2, &s);
+    for (i = 0; i < 250; ++i) {
+        secp256k1_num_add(&n1, &n1, &n1);    /* n1 *= 2 */
+        secp256k1_num_add(&n1p2, &n1, &n2);  /* n1p2 = n1 + 1 */
+        CHECK(!secp256k1_num_is_one(&n1p2));
+    }
+}
+
+void test_num_mod(void) {
+    int i;
+    secp256k1_scalar s;
+    secp256k1_num order, n;
+
+    /* check that 0 mod anything is 0 */
+    random_scalar_order_test(&s); 
+    secp256k1_scalar_get_num(&order, &s);
+    secp256k1_scalar_set_int(&s, 0);
+    secp256k1_scalar_get_num(&n, &s);
+    secp256k1_num_mod(&n, &order);
+    CHECK(secp256k1_num_is_zero(&n));
+
+    /* check that anything mod 1 is 0 */
+    secp256k1_scalar_set_int(&s, 1);
+    secp256k1_scalar_get_num(&order, &s);
+    secp256k1_scalar_get_num(&n, &s);
+    secp256k1_num_mod(&n, &order);
+    CHECK(secp256k1_num_is_zero(&n));
+
+    /* check that increasing the number past 2^256 does not break this */
+    random_scalar_order_test(&s); 
+    secp256k1_scalar_get_num(&n, &s);
+    /* multiply by 2^8, which'll test this case with high probability */
+    for (i = 0; i < 8; ++i) {
+        secp256k1_num_add(&n, &n, &n);
+    }
+    secp256k1_num_mod(&n, &order);
+    CHECK(secp256k1_num_is_zero(&n));
+}
+
+void test_num_jacobi(void) {
+    secp256k1_scalar sqr;
+    secp256k1_scalar small;
+    secp256k1_scalar five;  /* five is not a quadratic residue */
+    secp256k1_num order, n;
+    int i;
+    /* squares mod 5 are 1, 4 */
+    const int jacobi5[10] = { 0, 1, -1, -1, 1, 0, 1, -1, -1, 1 };
+
+    /* check some small values with 5 as the order */
+    secp256k1_scalar_set_int(&five, 5);
+    secp256k1_scalar_get_num(&order, &five);
+    for (i = 0; i < 10; ++i) {
+        secp256k1_scalar_set_int(&small, i);
+        secp256k1_scalar_get_num(&n, &small);
+        CHECK(secp256k1_num_jacobi(&n, &order) == jacobi5[i]);
+    }
+
+    /** test large values with 5 as group order */
+    secp256k1_scalar_get_num(&order, &five);
+    /* we first need a scalar which is not a multiple of 5 */
+    do {
+        secp256k1_num fiven;
+        random_scalar_order_test(&sqr); 
+        secp256k1_scalar_get_num(&fiven, &five);
+        secp256k1_scalar_get_num(&n, &sqr);
+        secp256k1_num_mod(&n, &fiven);
+    } while (secp256k1_num_is_zero(&n));
+    /* next force it to be a residue. 2 is a nonresidue mod 5 so we can
+     * just multiply by two, i.e. add the number to itself */
+    if (secp256k1_num_jacobi(&n, &order) == -1) {
+        secp256k1_num_add(&n, &n, &n);
+    }
+
+    /* test residue */
+    CHECK(secp256k1_num_jacobi(&n, &order) == 1);
+    /* test nonresidue */
+    secp256k1_num_add(&n, &n, &n);
+    CHECK(secp256k1_num_jacobi(&n, &order) == -1);
+
+    /** test with secp group order as order */
+    secp256k1_scalar_order_get_num(&order);
+    random_scalar_order_test(&sqr); 
+    secp256k1_scalar_sqr(&sqr, &sqr);
+    /* test residue */
+    secp256k1_scalar_get_num(&n, &sqr);
+    CHECK(secp256k1_num_jacobi(&n, &order) == 1);
+    /* test nonresidue */
+    secp256k1_scalar_mul(&sqr, &sqr, &five);
+    secp256k1_scalar_get_num(&n, &sqr);
+    CHECK(secp256k1_num_jacobi(&n, &order) == -1);
+    /* test multiple of the order*/
+    CHECK(secp256k1_num_jacobi(&order, &order) == 0);
+
+    /* check one less than the order */
+    secp256k1_scalar_set_int(&small, 1);
+    secp256k1_scalar_get_num(&n, &small);
+    secp256k1_num_sub(&n, &order, &n);
+    CHECK(secp256k1_num_jacobi(&n, &order) == 1);  /* sage confirms this is 1 */
 }
 
 void run_num_smalltests(void) {
@@ -505,6 +611,8 @@ void run_num_smalltests(void) {
     for (i = 0; i < 100*count; i++) {
         test_num_negate();
         test_num_add_sub();
+        test_num_mod();
+        test_num_jacobi();
     }
 }
 #endif
@@ -689,6 +797,10 @@ void scalar_test(void) {
             secp256k1_scalar_inverse(&inv, &inv);
             /* Inverting one must result in one. */
             CHECK(secp256k1_scalar_is_one(&inv));
+#ifndef USE_NUM_NONE
+            secp256k1_scalar_get_num(&invnum, &inv);
+            CHECK(secp256k1_num_is_one(&invnum));
+#endif
         }
     }
 
@@ -855,7 +967,7 @@ void run_scalar_tests(void) {
         secp256k1_scalar zzv;
 #endif
         int overflow;
-        unsigned char chal[32][2][32] = {
+        unsigned char chal[33][2][32] = {
             {{0xff, 0xff, 0x03, 0x07, 0x00, 0x00, 0x00, 0x00,
               0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0x03,
               0x00, 0x00, 0x00, 0x00, 0x00, 0xf8, 0xff, 0xff,
@@ -1111,9 +1223,17 @@ void run_scalar_tests(void) {
              {0xff, 0x01, 0x00, 0x00, 0x00, 0x00, 0xf8, 0xff,
               0xff, 0xff, 0xff, 0xff, 0xff, 0x1f, 0x00, 0x00,
               0xf8, 0x07, 0x00, 0x80, 0xff, 0xff, 0xff, 0xff,
-              0xff, 0xc7, 0xff, 0xff, 0xe0, 0xff, 0xff, 0xff}}
+              0xff, 0xc7, 0xff, 0xff, 0xe0, 0xff, 0xff, 0xff}},
+            {{0x82, 0xc9, 0xfa, 0xb0, 0x68, 0x04, 0xa0, 0x00,
+              0x82, 0xc9, 0xfa, 0xb0, 0x68, 0x04, 0xa0, 0x00,
+              0xff, 0xff, 0xff, 0xff, 0xff, 0x6f, 0x03, 0xfb,
+              0xfa, 0x8a, 0x7d, 0xdf, 0x13, 0x86, 0xe2, 0x03},
+             {0x82, 0xc9, 0xfa, 0xb0, 0x68, 0x04, 0xa0, 0x00,
+              0x82, 0xc9, 0xfa, 0xb0, 0x68, 0x04, 0xa0, 0x00,
+              0xff, 0xff, 0xff, 0xff, 0xff, 0x6f, 0x03, 0xfb,
+              0xfa, 0x8a, 0x7d, 0xdf, 0x13, 0x86, 0xe2, 0x03}}
         };
-        unsigned char res[32][2][32] = {
+        unsigned char res[33][2][32] = {
             {{0x0c, 0x3b, 0x0a, 0xca, 0x8d, 0x1a, 0x2f, 0xb9,
               0x8a, 0x7b, 0x53, 0x5a, 0x1f, 0xc5, 0x22, 0xa1,
               0x07, 0x2a, 0x48, 0xea, 0x02, 0xeb, 0xb3, 0xd6,
@@ -1369,10 +1489,18 @@ void run_scalar_tests(void) {
              {0xe4, 0xf1, 0x23, 0x84, 0xe1, 0xb5, 0x9d, 0xf2,
               0xb8, 0x73, 0x8b, 0x45, 0x2b, 0x35, 0x46, 0x38,
               0x10, 0x2b, 0x50, 0xf8, 0x8b, 0x35, 0xcd, 0x34,
-              0xc8, 0x0e, 0xf6, 0xdb, 0x09, 0x35, 0xf0, 0xda}}
+              0xc8, 0x0e, 0xf6, 0xdb, 0x09, 0x35, 0xf0, 0xda}},
+            {{0xdb, 0x21, 0x5c, 0x8d, 0x83, 0x1d, 0xb3, 0x34,
+              0xc7, 0x0e, 0x43, 0xa1, 0x58, 0x79, 0x67, 0x13,
+              0x1e, 0x86, 0x5d, 0x89, 0x63, 0xe6, 0x0a, 0x46,
+              0x5c, 0x02, 0x97, 0x1b, 0x62, 0x43, 0x86, 0xf5},
+             {0xdb, 0x21, 0x5c, 0x8d, 0x83, 0x1d, 0xb3, 0x34,
+              0xc7, 0x0e, 0x43, 0xa1, 0x58, 0x79, 0x67, 0x13,
+              0x1e, 0x86, 0x5d, 0x89, 0x63, 0xe6, 0x0a, 0x46,
+              0x5c, 0x02, 0x97, 0x1b, 0x62, 0x43, 0x86, 0xf5}}
         };
         secp256k1_scalar_set_int(&one, 1);
-        for (i = 0; i < 32; i++) {
+        for (i = 0; i < 33; i++) {
             secp256k1_scalar_set_b32(&x, chal[i][0], &overflow);
             CHECK(!overflow);
             secp256k1_scalar_set_b32(&y, chal[i][1], &overflow);
@@ -1446,7 +1574,7 @@ void random_fe_non_zero(secp256k1_fe *nz) {
 void random_fe_non_square(secp256k1_fe *ns) {
     secp256k1_fe r;
     random_fe_non_zero(ns);
-    if (secp256k1_fe_sqrt_var(&r, ns)) {
+    if (secp256k1_fe_sqrt(&r, ns)) {
         secp256k1_fe_negate(ns, ns, 1);
     }
 }
@@ -1641,7 +1769,7 @@ void run_sqr(void) {
 
 void test_sqrt(const secp256k1_fe *a, const secp256k1_fe *k) {
     secp256k1_fe r1, r2;
-    int v = secp256k1_fe_sqrt_var(&r1, a);
+    int v = secp256k1_fe_sqrt(&r1, a);
     CHECK((v == 0) == (k == NULL));
 
     if (k != NULL) {
@@ -1951,8 +2079,8 @@ void test_add_neg_y_diff_x(void) {
      * of the sum to be wrong (since infinity has no xy coordinates).
      * HOWEVER, if the x-coordinates are different, infinity is the
      * wrong answer, and such degeneracies are exposed. This is the
-     * root of https://github.com/bitcoin/secp256k1/issues/257 which
-     * this test is a regression test for.
+     * root of https://github.com/bitcoin-core/secp256k1/issues/257
+     * which this test is a regression test for.
      *
      * These points were generated in sage as
      * # secp256k1 params
@@ -2051,15 +2179,16 @@ void run_ec_combine(void) {
 void test_group_decompress(const secp256k1_fe* x) {
     /* The input itself, normalized. */
     secp256k1_fe fex = *x;
-    secp256k1_fe tmp;
+    secp256k1_fe fez;
     /* Results of set_xquad_var, set_xo_var(..., 0), set_xo_var(..., 1). */
     secp256k1_ge ge_quad, ge_even, ge_odd;
+    secp256k1_gej gej_quad;
     /* Return values of the above calls. */
     int res_quad, res_even, res_odd;
 
     secp256k1_fe_normalize_var(&fex);
 
-    res_quad = secp256k1_ge_set_xquad_var(&ge_quad, &fex);
+    res_quad = secp256k1_ge_set_xquad(&ge_quad, &fex);
     res_even = secp256k1_ge_set_xo_var(&ge_even, &fex, 0);
     res_odd = secp256k1_ge_set_xo_var(&ge_odd, &fex, 1);
 
@@ -2085,13 +2214,29 @@ void test_group_decompress(const secp256k1_fe* x) {
         CHECK(secp256k1_fe_equal_var(&ge_odd.x, x));
 
         /* Check that the Y coordinate result in ge_quad is a square. */
-        CHECK(secp256k1_fe_sqrt_var(&tmp, &ge_quad.y));
-        secp256k1_fe_sqr(&tmp, &tmp);
-        CHECK(secp256k1_fe_equal_var(&tmp, &ge_quad.y));
+        CHECK(secp256k1_fe_is_quad_var(&ge_quad.y));
 
         /* Check odd/even Y in ge_odd, ge_even. */
         CHECK(secp256k1_fe_is_odd(&ge_odd.y));
         CHECK(!secp256k1_fe_is_odd(&ge_even.y));
+
+        /* Check secp256k1_gej_has_quad_y_var. */
+        secp256k1_gej_set_ge(&gej_quad, &ge_quad);
+        CHECK(secp256k1_gej_has_quad_y_var(&gej_quad));
+        do {
+            random_fe_test(&fez);
+        } while (secp256k1_fe_is_zero(&fez));
+        secp256k1_gej_rescale(&gej_quad, &fez);
+        CHECK(secp256k1_gej_has_quad_y_var(&gej_quad));
+        secp256k1_gej_neg(&gej_quad, &gej_quad);
+        CHECK(!secp256k1_gej_has_quad_y_var(&gej_quad));
+        do {
+            random_fe_test(&fez);
+        } while (secp256k1_fe_is_zero(&fez));
+        secp256k1_gej_rescale(&gej_quad, &fez);
+        CHECK(!secp256k1_gej_has_quad_y_var(&gej_quad));
+        secp256k1_gej_neg(&gej_quad, &gej_quad);
+        CHECK(secp256k1_gej_has_quad_y_var(&gej_quad));
     }
 }
 
@@ -2383,9 +2528,7 @@ void test_constant_wnaf(const secp256k1_scalar *number, int w) {
     secp256k1_scalar x, shift;
     int wnaf[256] = {0};
     int i;
-#ifdef USE_ENDOMORPHISM
     int skew;
-#endif
     secp256k1_scalar num = *number;
 
     secp256k1_scalar_set_int(&x, 0);
@@ -2395,10 +2538,8 @@ void test_constant_wnaf(const secp256k1_scalar *number, int w) {
     for (i = 0; i < 16; ++i) {
         secp256k1_scalar_shr_int(&num, 8);
     }
-    skew = secp256k1_wnaf_const(wnaf, num, w);
-#else
-    secp256k1_wnaf_const(wnaf, num, w);
 #endif
+    skew = secp256k1_wnaf_const(wnaf, num, w);
 
     for (i = WNAF_SIZE(w); i >= 0; --i) {
         secp256k1_scalar t;
@@ -2417,10 +2558,8 @@ void test_constant_wnaf(const secp256k1_scalar *number, int w) {
         }
         secp256k1_scalar_add(&x, &x, &t);
     }
-#ifdef USE_ENDOMORPHISM
-    /* Skew num because when encoding 128-bit numbers as odd we use an offset */
+    /* Skew num because when encoding numbers as odd we use an offset */
     secp256k1_scalar_cadd_bit(&num, skew == 2, 1);
-#endif
     CHECK(secp256k1_scalar_eq(&x, &num));
 }
 
@@ -3484,12 +3623,14 @@ void run_ecdsa_end_to_end(void) {
 
 int test_ecdsa_der_parse(const unsigned char *sig, size_t siglen, int certainly_der, int certainly_not_der) {
     static const unsigned char zeroes[32] = {0};
+#ifdef ENABLE_OPENSSL_TESTS
     static const unsigned char max_scalar[32] = {
         0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff,
         0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xfe,
         0xba, 0xae, 0xdc, 0xe6, 0xaf, 0x48, 0xa0, 0x3b,
         0xbf, 0xd2, 0x5e, 0x8c, 0xd0, 0x36, 0x41, 0x40
     };
+#endif
 
     int ret = 0;
 
@@ -3607,13 +3748,13 @@ static void assign_big_endian(unsigned char *ptr, size_t ptrlen, uint32_t val) {
 static void damage_array(unsigned char *sig, size_t *len) {
     int pos;
     int action = secp256k1_rand_bits(3);
-    if (action < 1) {
+    if (action < 1 && *len > 3) {
         /* Delete a byte. */
         pos = secp256k1_rand_int(*len);
         memmove(sig + pos, sig + pos + 1, *len - pos - 1);
         (*len)--;
         return;
-    } else if (action < 2) {
+    } else if (action < 2 && *len < 2048) {
         /* Insert a byte. */
         pos = secp256k1_rand_int(1 + *len);
         memmove(sig + pos + 1, sig + pos, *len - pos);
@@ -3785,6 +3926,7 @@ void run_ecdsa_der_parse(void) {
         int certainly_der = 0;
         int certainly_not_der = 0;
         random_ber_signature(buffer, &buflen, &certainly_der, &certainly_not_der);
+        CHECK(buflen <= 2048);
         for (j = 0; j < 16; j++) {
             int ret = 0;
             if (j > 0) {