Adding upstream version 6.6.15.upstream/6.6.15

Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>

8 files changed, 2565 insertions, 0 deletions

diff --git a/arch/mips/crypto/.gitignore b/arch/mips/crypto/.gitignore
new file mode 100644
index 000000000..0d47d4f21
--- /dev/null
+++ b/arch/mips/crypto/.gitignore
@@ -0,0 +1,2 @@
+# SPDX-License-Identifier: GPL-2.0-only
+poly1305-core.S
diff --git a/arch/mips/crypto/Kconfig b/arch/mips/crypto/Kconfig
new file mode 100644
index 000000000..9003a5c1e
--- /dev/null
+++ b/arch/mips/crypto/Kconfig
@@ -0,0 +1,74 @@
+# SPDX-License-Identifier: GPL-2.0
+
+menu "Accelerated Cryptographic Algorithms for CPU (mips)"
+
+config CRYPTO_CRC32_MIPS
+	tristate "CRC32c and CRC32"
+	depends on MIPS_CRC_SUPPORT
+	select CRYPTO_HASH
+	help
+	  CRC32c and CRC32 CRC algorithms
+
+	  Architecture: mips
+
+config CRYPTO_POLY1305_MIPS
+	tristate "Hash functions: Poly1305"
+	depends on MIPS
+	select CRYPTO_ARCH_HAVE_LIB_POLY1305
+	help
+	  Poly1305 authenticator algorithm (RFC7539)
+
+	  Architecture: mips
+
+config CRYPTO_MD5_OCTEON
+	tristate "Digests: MD5 (OCTEON)"
+	depends on CPU_CAVIUM_OCTEON
+	select CRYPTO_MD5
+	select CRYPTO_HASH
+	help
+	  MD5 message digest algorithm (RFC1321)
+
+	  Architecture: mips OCTEON using crypto instructions, when available
+
+config CRYPTO_SHA1_OCTEON
+	tristate "Hash functions: SHA-1 (OCTEON)"
+	depends on CPU_CAVIUM_OCTEON
+	select CRYPTO_SHA1
+	select CRYPTO_HASH
+	help
+	  SHA-1 secure hash algorithm (FIPS 180)
+
+	  Architecture: mips OCTEON
+
+config CRYPTO_SHA256_OCTEON
+	tristate "Hash functions: SHA-224 and SHA-256 (OCTEON)"
+	depends on CPU_CAVIUM_OCTEON
+	select CRYPTO_SHA256
+	select CRYPTO_HASH
+	help
+	  SHA-224 and SHA-256 secure hash algorithms (FIPS 180)
+
+	  Architecture: mips OCTEON using crypto instructions, when available
+
+config CRYPTO_SHA512_OCTEON
+	tristate "Hash functions: SHA-384 and SHA-512 (OCTEON)"
+	depends on CPU_CAVIUM_OCTEON
+	select CRYPTO_SHA512
+	select CRYPTO_HASH
+	help
+	  SHA-384 and SHA-512 secure hash algorithms (FIPS 180)
+
+	  Architecture: mips OCTEON using crypto instructions, when available
+
+config CRYPTO_CHACHA_MIPS
+	tristate "Ciphers: ChaCha20, XChaCha20, XChaCha12 (MIPS32r2)"
+	depends on CPU_MIPS32_R2
+	select CRYPTO_SKCIPHER
+	select CRYPTO_ARCH_HAVE_LIB_CHACHA
+	help
+	  Length-preserving ciphers: ChaCha20, XChaCha20, and XChaCha12
+	  stream cipher algorithms
+
+	  Architecture: MIPS32r2
+
+endmenu
diff --git a/arch/mips/crypto/Makefile b/arch/mips/crypto/Makefile
new file mode 100644
index 000000000..5e4105ccc
--- /dev/null
+++ b/arch/mips/crypto/Makefile
@@ -0,0 +1,24 @@
+# SPDX-License-Identifier: GPL-2.0
+#
+# Makefile for MIPS crypto files..
+#
+
+obj-$(CONFIG_CRYPTO_CRC32_MIPS) += crc32-mips.o
+
+obj-$(CONFIG_CRYPTO_CHACHA_MIPS) += chacha-mips.o
+chacha-mips-y := chacha-core.o chacha-glue.o
+AFLAGS_chacha-core.o += -O2 # needed to fill branch delay slots
+
+obj-$(CONFIG_CRYPTO_POLY1305_MIPS) += poly1305-mips.o
+poly1305-mips-y := poly1305-core.o poly1305-glue.o
+
+perlasm-flavour-$(CONFIG_32BIT) := o32
+perlasm-flavour-$(CONFIG_64BIT) := 64
+
+quiet_cmd_perlasm = PERLASM $@
+      cmd_perlasm = $(PERL) $(<) $(perlasm-flavour-y) $(@)
+
+$(obj)/poly1305-core.S: $(src)/poly1305-mips.pl FORCE
+	$(call if_changed,perlasm)
+
+targets += poly1305-core.S
diff --git a/arch/mips/crypto/chacha-core.S b/arch/mips/crypto/chacha-core.S
new file mode 100644
index 000000000..5755f69cf
--- /dev/null
+++ b/arch/mips/crypto/chacha-core.S
@@ -0,0 +1,497 @@
+/* SPDX-License-Identifier: GPL-2.0 OR MIT */
+/*
+ * Copyright (C) 2016-2018 René van Dorst <opensource@vdorst.com>. All Rights Reserved.
+ * Copyright (C) 2015-2019 Jason A. Donenfeld <Jason@zx2c4.com>. All Rights Reserved.
+ */
+
+#define MASK_U32		0x3c
+#define CHACHA20_BLOCK_SIZE	64
+#define STACK_SIZE		32
+
+#define X0	$t0
+#define X1	$t1
+#define X2	$t2
+#define X3	$t3
+#define X4	$t4
+#define X5	$t5
+#define X6	$t6
+#define X7	$t7
+#define X8	$t8
+#define X9	$t9
+#define X10	$v1
+#define X11	$s6
+#define X12	$s5
+#define X13	$s4
+#define X14	$s3
+#define X15	$s2
+/* Use regs which are overwritten on exit for Tx so we don't leak clear data. */
+#define T0	$s1
+#define T1	$s0
+#define T(n)	T ## n
+#define X(n)	X ## n
+
+/* Input arguments */
+#define STATE		$a0
+#define OUT		$a1
+#define IN		$a2
+#define BYTES		$a3
+
+/* Output argument */
+/* NONCE[0] is kept in a register and not in memory.
+ * We don't want to touch original value in memory.
+ * Must be incremented every loop iteration.
+ */
+#define NONCE_0		$v0
+
+/* SAVED_X and SAVED_CA are set in the jump table.
+ * Use regs which are overwritten on exit else we don't leak clear data.
+ * They are used to handling the last bytes which are not multiple of 4.
+ */
+#define SAVED_X		X15
+#define SAVED_CA	$s7
+
+#define IS_UNALIGNED	$s7
+
+#if __BYTE_ORDER__ == __ORDER_BIG_ENDIAN__
+#define MSB 0
+#define LSB 3
+#define ROTx rotl
+#define ROTR(n) rotr n, 24
+#define	CPU_TO_LE32(n) \
+	wsbh	n; \
+	rotr	n, 16;
+#else
+#define MSB 3
+#define LSB 0
+#define ROTx rotr
+#define CPU_TO_LE32(n)
+#define ROTR(n)
+#endif
+
+#define FOR_EACH_WORD(x) \
+	x( 0); \
+	x( 1); \
+	x( 2); \
+	x( 3); \
+	x( 4); \
+	x( 5); \
+	x( 6); \
+	x( 7); \
+	x( 8); \
+	x( 9); \
+	x(10); \
+	x(11); \
+	x(12); \
+	x(13); \
+	x(14); \
+	x(15);
+
+#define FOR_EACH_WORD_REV(x) \
+	x(15); \
+	x(14); \
+	x(13); \
+	x(12); \
+	x(11); \
+	x(10); \
+	x( 9); \
+	x( 8); \
+	x( 7); \
+	x( 6); \
+	x( 5); \
+	x( 4); \
+	x( 3); \
+	x( 2); \
+	x( 1); \
+	x( 0);
+
+#define PLUS_ONE_0	 1
+#define PLUS_ONE_1	 2
+#define PLUS_ONE_2	 3
+#define PLUS_ONE_3	 4
+#define PLUS_ONE_4	 5
+#define PLUS_ONE_5	 6
+#define PLUS_ONE_6	 7
+#define PLUS_ONE_7	 8
+#define PLUS_ONE_8	 9
+#define PLUS_ONE_9	10
+#define PLUS_ONE_10	11
+#define PLUS_ONE_11	12
+#define PLUS_ONE_12	13
+#define PLUS_ONE_13	14
+#define PLUS_ONE_14	15
+#define PLUS_ONE_15	16
+#define PLUS_ONE(x)	PLUS_ONE_ ## x
+#define _CONCAT3(a,b,c)	a ## b ## c
+#define CONCAT3(a,b,c)	_CONCAT3(a,b,c)
+
+#define STORE_UNALIGNED(x) \
+CONCAT3(.Lchacha_mips_xor_unaligned_, PLUS_ONE(x), _b: ;) \
+	.if (x != 12); \
+		lw	T0, (x*4)(STATE); \
+	.endif; \
+	lwl	T1, (x*4)+MSB ## (IN); \
+	lwr	T1, (x*4)+LSB ## (IN); \
+	.if (x == 12); \
+		addu	X ## x, NONCE_0; \
+	.else; \
+		addu	X ## x, T0; \
+	.endif; \
+	CPU_TO_LE32(X ## x); \
+	xor	X ## x, T1; \
+	swl	X ## x, (x*4)+MSB ## (OUT); \
+	swr	X ## x, (x*4)+LSB ## (OUT);
+
+#define STORE_ALIGNED(x) \
+CONCAT3(.Lchacha_mips_xor_aligned_, PLUS_ONE(x), _b: ;) \
+	.if (x != 12); \
+		lw	T0, (x*4)(STATE); \
+	.endif; \
+	lw	T1, (x*4) ## (IN); \
+	.if (x == 12); \
+		addu	X ## x, NONCE_0; \
+	.else; \
+		addu	X ## x, T0; \
+	.endif; \
+	CPU_TO_LE32(X ## x); \
+	xor	X ## x, T1; \
+	sw	X ## x, (x*4) ## (OUT);
+
+/* Jump table macro.
+ * Used for setup and handling the last bytes, which are not multiple of 4.
+ * X15 is free to store Xn
+ * Every jumptable entry must be equal in size.
+ */
+#define JMPTBL_ALIGNED(x) \
+.Lchacha_mips_jmptbl_aligned_ ## x: ; \
+	.set	noreorder; \
+	b	.Lchacha_mips_xor_aligned_ ## x ## _b; \
+	.if (x == 12); \
+		addu	SAVED_X, X ## x, NONCE_0; \
+	.else; \
+		addu	SAVED_X, X ## x, SAVED_CA; \
+	.endif; \
+	.set	reorder
+
+#define JMPTBL_UNALIGNED(x) \
+.Lchacha_mips_jmptbl_unaligned_ ## x: ; \
+	.set	noreorder; \
+	b	.Lchacha_mips_xor_unaligned_ ## x ## _b; \
+	.if (x == 12); \
+		addu	SAVED_X, X ## x, NONCE_0; \
+	.else; \
+		addu	SAVED_X, X ## x, SAVED_CA; \
+	.endif; \
+	.set	reorder
+
+#define AXR(A, B, C, D,  K, L, M, N,  V, W, Y, Z,  S) \
+	addu	X(A), X(K); \
+	addu	X(B), X(L); \
+	addu	X(C), X(M); \
+	addu	X(D), X(N); \
+	xor	X(V), X(A); \
+	xor	X(W), X(B); \
+	xor	X(Y), X(C); \
+	xor	X(Z), X(D); \
+	rotl	X(V), S;    \
+	rotl	X(W), S;    \
+	rotl	X(Y), S;    \
+	rotl	X(Z), S;
+
+.text
+.set	reorder
+.set	noat
+.globl	chacha_crypt_arch
+.ent	chacha_crypt_arch
+chacha_crypt_arch:
+	.frame	$sp, STACK_SIZE, $ra
+
+	/* Load number of rounds */
+	lw	$at, 16($sp)
+
+	addiu	$sp, -STACK_SIZE
+
+	/* Return bytes = 0. */
+	beqz	BYTES, .Lchacha_mips_end
+
+	lw	NONCE_0, 48(STATE)
+
+	/* Save s0-s7 */
+	sw	$s0,  0($sp)
+	sw	$s1,  4($sp)
+	sw	$s2,  8($sp)
+	sw	$s3, 12($sp)
+	sw	$s4, 16($sp)
+	sw	$s5, 20($sp)
+	sw	$s6, 24($sp)
+	sw	$s7, 28($sp)
+
+	/* Test IN or OUT is unaligned.
+	 * IS_UNALIGNED = ( IN | OUT ) & 0x00000003
+	 */
+	or	IS_UNALIGNED, IN, OUT
+	andi	IS_UNALIGNED, 0x3
+
+	b	.Lchacha_rounds_start
+
+.align 4
+.Loop_chacha_rounds:
+	addiu	IN,  CHACHA20_BLOCK_SIZE
+	addiu	OUT, CHACHA20_BLOCK_SIZE
+	addiu	NONCE_0, 1
+
+.Lchacha_rounds_start:
+	lw	X0,  0(STATE)
+	lw	X1,  4(STATE)
+	lw	X2,  8(STATE)
+	lw	X3,  12(STATE)
+
+	lw	X4,  16(STATE)
+	lw	X5,  20(STATE)
+	lw	X6,  24(STATE)
+	lw	X7,  28(STATE)
+	lw	X8,  32(STATE)
+	lw	X9,  36(STATE)
+	lw	X10, 40(STATE)
+	lw	X11, 44(STATE)
+
+	move	X12, NONCE_0
+	lw	X13, 52(STATE)
+	lw	X14, 56(STATE)
+	lw	X15, 60(STATE)
+
+.Loop_chacha_xor_rounds:
+	addiu	$at, -2
+	AXR( 0, 1, 2, 3,  4, 5, 6, 7, 12,13,14,15, 16);
+	AXR( 8, 9,10,11, 12,13,14,15,  4, 5, 6, 7, 12);
+	AXR( 0, 1, 2, 3,  4, 5, 6, 7, 12,13,14,15,  8);
+	AXR( 8, 9,10,11, 12,13,14,15,  4, 5, 6, 7,  7);
+	AXR( 0, 1, 2, 3,  5, 6, 7, 4, 15,12,13,14, 16);
+	AXR(10,11, 8, 9, 15,12,13,14,  5, 6, 7, 4, 12);
+	AXR( 0, 1, 2, 3,  5, 6, 7, 4, 15,12,13,14,  8);
+	AXR(10,11, 8, 9, 15,12,13,14,  5, 6, 7, 4,  7);
+	bnez	$at, .Loop_chacha_xor_rounds
+
+	addiu	BYTES, -(CHACHA20_BLOCK_SIZE)
+
+	/* Is data src/dst unaligned? Jump */
+	bnez	IS_UNALIGNED, .Loop_chacha_unaligned
+
+	/* Set number rounds here to fill delayslot. */
+	lw	$at, (STACK_SIZE+16)($sp)
+
+	/* BYTES < 0, it has no full block. */
+	bltz	BYTES, .Lchacha_mips_no_full_block_aligned
+
+	FOR_EACH_WORD_REV(STORE_ALIGNED)
+
+	/* BYTES > 0? Loop again. */
+	bgtz	BYTES, .Loop_chacha_rounds
+
+	/* Place this here to fill delay slot */
+	addiu	NONCE_0, 1
+
+	/* BYTES < 0? Handle last bytes */
+	bltz	BYTES, .Lchacha_mips_xor_bytes
+
+.Lchacha_mips_xor_done:
+	/* Restore used registers */
+	lw	$s0,  0($sp)
+	lw	$s1,  4($sp)
+	lw	$s2,  8($sp)
+	lw	$s3, 12($sp)
+	lw	$s4, 16($sp)
+	lw	$s5, 20($sp)
+	lw	$s6, 24($sp)
+	lw	$s7, 28($sp)
+
+	/* Write NONCE_0 back to right location in state */
+	sw	NONCE_0, 48(STATE)
+
+.Lchacha_mips_end:
+	addiu	$sp, STACK_SIZE
+	jr	$ra
+
+.Lchacha_mips_no_full_block_aligned:
+	/* Restore the offset on BYTES */
+	addiu	BYTES, CHACHA20_BLOCK_SIZE
+
+	/* Get number of full WORDS */
+	andi	$at, BYTES, MASK_U32
+
+	/* Load upper half of jump table addr */
+	lui	T0, %hi(.Lchacha_mips_jmptbl_aligned_0)
+
+	/* Calculate lower half jump table offset */
+	ins	T0, $at, 1, 6
+
+	/* Add offset to STATE */
+	addu	T1, STATE, $at
+
+	/* Add lower half jump table addr */
+	addiu	T0, %lo(.Lchacha_mips_jmptbl_aligned_0)
+
+	/* Read value from STATE */
+	lw	SAVED_CA, 0(T1)
+
+	/* Store remaining bytecounter as negative value */
+	subu	BYTES, $at, BYTES
+
+	jr	T0
+
+	/* Jump table */
+	FOR_EACH_WORD(JMPTBL_ALIGNED)
+
+
+.Loop_chacha_unaligned:
+	/* Set number rounds here to fill delayslot. */
+	lw	$at, (STACK_SIZE+16)($sp)
+
+	/* BYTES > 0, it has no full block. */
+	bltz	BYTES, .Lchacha_mips_no_full_block_unaligned
+
+	FOR_EACH_WORD_REV(STORE_UNALIGNED)
+
+	/* BYTES > 0? Loop again. */
+	bgtz	BYTES, .Loop_chacha_rounds
+
+	/* Write NONCE_0 back to right location in state */
+	sw	NONCE_0, 48(STATE)
+
+	.set noreorder
+	/* Fall through to byte handling */
+	bgez	BYTES, .Lchacha_mips_xor_done
+.Lchacha_mips_xor_unaligned_0_b:
+.Lchacha_mips_xor_aligned_0_b:
+	/* Place this here to fill delay slot */
+	addiu	NONCE_0, 1
+	.set reorder
+
+.Lchacha_mips_xor_bytes:
+	addu	IN, $at
+	addu	OUT, $at
+	/* First byte */
+	lbu	T1, 0(IN)
+	addiu	$at, BYTES, 1
+	CPU_TO_LE32(SAVED_X)
+	ROTR(SAVED_X)
+	xor	T1, SAVED_X
+	sb	T1, 0(OUT)
+	beqz	$at, .Lchacha_mips_xor_done
+	/* Second byte */
+	lbu	T1, 1(IN)
+	addiu	$at, BYTES, 2
+	ROTx	SAVED_X, 8
+	xor	T1, SAVED_X
+	sb	T1, 1(OUT)
+	beqz	$at, .Lchacha_mips_xor_done
+	/* Third byte */
+	lbu	T1, 2(IN)
+	ROTx	SAVED_X, 8
+	xor	T1, SAVED_X
+	sb	T1, 2(OUT)
+	b	.Lchacha_mips_xor_done
+
+.Lchacha_mips_no_full_block_unaligned:
+	/* Restore the offset on BYTES */
+	addiu	BYTES, CHACHA20_BLOCK_SIZE
+
+	/* Get number of full WORDS */
+	andi	$at, BYTES, MASK_U32
+
+	/* Load upper half of jump table addr */
+	lui	T0, %hi(.Lchacha_mips_jmptbl_unaligned_0)
+
+	/* Calculate lower half jump table offset */
+	ins	T0, $at, 1, 6
+
+	/* Add offset to STATE */
+	addu	T1, STATE, $at
+
+	/* Add lower half jump table addr */
+	addiu	T0, %lo(.Lchacha_mips_jmptbl_unaligned_0)
+
+	/* Read value from STATE */
+	lw	SAVED_CA, 0(T1)
+
+	/* Store remaining bytecounter as negative value */
+	subu	BYTES, $at, BYTES
+
+	jr	T0
+
+	/* Jump table */
+	FOR_EACH_WORD(JMPTBL_UNALIGNED)
+.end chacha_crypt_arch
+.set at
+
+/* Input arguments
+ * STATE	$a0
+ * OUT		$a1
+ * NROUND	$a2
+ */
+
+#undef X12
+#undef X13
+#undef X14
+#undef X15
+
+#define X12	$a3
+#define X13	$at
+#define X14	$v0
+#define X15	STATE
+
+.set noat
+.globl	hchacha_block_arch
+.ent	hchacha_block_arch
+hchacha_block_arch:
+	.frame	$sp, STACK_SIZE, $ra
+
+	addiu	$sp, -STACK_SIZE
+
+	/* Save X11(s6) */
+	sw	X11, 0($sp)
+
+	lw	X0,  0(STATE)
+	lw	X1,  4(STATE)
+	lw	X2,  8(STATE)
+	lw	X3,  12(STATE)
+	lw	X4,  16(STATE)
+	lw	X5,  20(STATE)
+	lw	X6,  24(STATE)
+	lw	X7,  28(STATE)
+	lw	X8,  32(STATE)
+	lw	X9,  36(STATE)
+	lw	X10, 40(STATE)
+	lw	X11, 44(STATE)
+	lw	X12, 48(STATE)
+	lw	X13, 52(STATE)
+	lw	X14, 56(STATE)
+	lw	X15, 60(STATE)
+
+.Loop_hchacha_xor_rounds:
+	addiu	$a2, -2
+	AXR( 0, 1, 2, 3,  4, 5, 6, 7, 12,13,14,15, 16);
+	AXR( 8, 9,10,11, 12,13,14,15,  4, 5, 6, 7, 12);
+	AXR( 0, 1, 2, 3,  4, 5, 6, 7, 12,13,14,15,  8);
+	AXR( 8, 9,10,11, 12,13,14,15,  4, 5, 6, 7,  7);
+	AXR( 0, 1, 2, 3,  5, 6, 7, 4, 15,12,13,14, 16);
+	AXR(10,11, 8, 9, 15,12,13,14,  5, 6, 7, 4, 12);
+	AXR( 0, 1, 2, 3,  5, 6, 7, 4, 15,12,13,14,  8);
+	AXR(10,11, 8, 9, 15,12,13,14,  5, 6, 7, 4,  7);
+	bnez	$a2, .Loop_hchacha_xor_rounds
+
+	/* Restore used register */
+	lw	X11, 0($sp)
+
+	sw	X0,  0(OUT)
+	sw	X1,  4(OUT)
+	sw	X2,  8(OUT)
+	sw	X3,  12(OUT)
+	sw	X12, 16(OUT)
+	sw	X13, 20(OUT)
+	sw	X14, 24(OUT)
+	sw	X15, 28(OUT)
+
+	addiu	$sp, STACK_SIZE
+	jr	$ra
+.end hchacha_block_arch
+.set at
diff --git a/arch/mips/crypto/chacha-glue.c b/arch/mips/crypto/chacha-glue.c
new file mode 100644
index 000000000..d1fd23e6e
--- /dev/null
+++ b/arch/mips/crypto/chacha-glue.c
@@ -0,0 +1,152 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * MIPS accelerated ChaCha and XChaCha stream ciphers,
+ * including ChaCha20 (RFC7539)
+ *
+ * Copyright (C) 2019 Linaro, Ltd. <ard.biesheuvel@linaro.org>
+ */
+
+#include <asm/byteorder.h>
+#include <crypto/algapi.h>
+#include <crypto/internal/chacha.h>
+#include <crypto/internal/skcipher.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+
+asmlinkage void chacha_crypt_arch(u32 *state, u8 *dst, const u8 *src,
+				  unsigned int bytes, int nrounds);
+EXPORT_SYMBOL(chacha_crypt_arch);
+
+asmlinkage void hchacha_block_arch(const u32 *state, u32 *stream, int nrounds);
+EXPORT_SYMBOL(hchacha_block_arch);
+
+void chacha_init_arch(u32 *state, const u32 *key, const u8 *iv)
+{
+	chacha_init_generic(state, key, iv);
+}
+EXPORT_SYMBOL(chacha_init_arch);
+
+static int chacha_mips_stream_xor(struct skcipher_request *req,
+				  const struct chacha_ctx *ctx, const u8 *iv)
+{
+	struct skcipher_walk walk;
+	u32 state[16];
+	int err;
+
+	err = skcipher_walk_virt(&walk, req, false);
+
+	chacha_init_generic(state, ctx->key, iv);
+
+	while (walk.nbytes > 0) {
+		unsigned int nbytes = walk.nbytes;
+
+		if (nbytes < walk.total)
+			nbytes = round_down(nbytes, walk.stride);
+
+		chacha_crypt(state, walk.dst.virt.addr, walk.src.virt.addr,
+			     nbytes, ctx->nrounds);
+		err = skcipher_walk_done(&walk, walk.nbytes - nbytes);
+	}
+
+	return err;
+}
+
+static int chacha_mips(struct skcipher_request *req)
+{
+	struct crypto_skcipher *tfm = crypto_skcipher_reqtfm(req);
+	struct chacha_ctx *ctx = crypto_skcipher_ctx(tfm);
+
+	return chacha_mips_stream_xor(req, ctx, req->iv);
+}
+
+static int xchacha_mips(struct skcipher_request *req)
+{
+	struct crypto_skcipher *tfm = crypto_skcipher_reqtfm(req);
+	struct chacha_ctx *ctx = crypto_skcipher_ctx(tfm);
+	struct chacha_ctx subctx;
+	u32 state[16];
+	u8 real_iv[16];
+
+	chacha_init_generic(state, ctx->key, req->iv);
+
+	hchacha_block(state, subctx.key, ctx->nrounds);
+	subctx.nrounds = ctx->nrounds;
+
+	memcpy(&real_iv[0], req->iv + 24, 8);
+	memcpy(&real_iv[8], req->iv + 16, 8);
+	return chacha_mips_stream_xor(req, &subctx, real_iv);
+}
+
+static struct skcipher_alg algs[] = {
+	{
+		.base.cra_name		= "chacha20",
+		.base.cra_driver_name	= "chacha20-mips",
+		.base.cra_priority	= 200,
+		.base.cra_blocksize	= 1,
+		.base.cra_ctxsize	= sizeof(struct chacha_ctx),
+		.base.cra_module	= THIS_MODULE,
+
+		.min_keysize		= CHACHA_KEY_SIZE,
+		.max_keysize		= CHACHA_KEY_SIZE,
+		.ivsize			= CHACHA_IV_SIZE,
+		.chunksize		= CHACHA_BLOCK_SIZE,
+		.setkey			= chacha20_setkey,
+		.encrypt		= chacha_mips,
+		.decrypt		= chacha_mips,
+	}, {
+		.base.cra_name		= "xchacha20",
+		.base.cra_driver_name	= "xchacha20-mips",
+		.base.cra_priority	= 200,
+		.base.cra_blocksize	= 1,
+		.base.cra_ctxsize	= sizeof(struct chacha_ctx),
+		.base.cra_module	= THIS_MODULE,
+
+		.min_keysize		= CHACHA_KEY_SIZE,
+		.max_keysize		= CHACHA_KEY_SIZE,
+		.ivsize			= XCHACHA_IV_SIZE,
+		.chunksize		= CHACHA_BLOCK_SIZE,
+		.setkey			= chacha20_setkey,
+		.encrypt		= xchacha_mips,
+		.decrypt		= xchacha_mips,
+	}, {
+		.base.cra_name		= "xchacha12",
+		.base.cra_driver_name	= "xchacha12-mips",
+		.base.cra_priority	= 200,
+		.base.cra_blocksize	= 1,
+		.base.cra_ctxsize	= sizeof(struct chacha_ctx),
+		.base.cra_module	= THIS_MODULE,
+
+		.min_keysize		= CHACHA_KEY_SIZE,
+		.max_keysize		= CHACHA_KEY_SIZE,
+		.ivsize			= XCHACHA_IV_SIZE,
+		.chunksize		= CHACHA_BLOCK_SIZE,
+		.setkey			= chacha12_setkey,
+		.encrypt		= xchacha_mips,
+		.decrypt		= xchacha_mips,
+	}
+};
+
+static int __init chacha_simd_mod_init(void)
+{
+	return IS_REACHABLE(CONFIG_CRYPTO_SKCIPHER) ?
+		crypto_register_skciphers(algs, ARRAY_SIZE(algs)) : 0;
+}
+
+static void __exit chacha_simd_mod_fini(void)
+{
+	if (IS_REACHABLE(CONFIG_CRYPTO_SKCIPHER))
+		crypto_unregister_skciphers(algs, ARRAY_SIZE(algs));
+}
+
+module_init(chacha_simd_mod_init);
+module_exit(chacha_simd_mod_fini);
+
+MODULE_DESCRIPTION("ChaCha and XChaCha stream ciphers (MIPS accelerated)");
+MODULE_AUTHOR("Ard Biesheuvel <ard.biesheuvel@linaro.org>");
+MODULE_LICENSE("GPL v2");
+MODULE_ALIAS_CRYPTO("chacha20");
+MODULE_ALIAS_CRYPTO("chacha20-mips");
+MODULE_ALIAS_CRYPTO("xchacha20");
+MODULE_ALIAS_CRYPTO("xchacha20-mips");
+MODULE_ALIAS_CRYPTO("xchacha12");
+MODULE_ALIAS_CRYPTO("xchacha12-mips");
diff --git a/arch/mips/crypto/crc32-mips.c b/arch/mips/crypto/crc32-mips.c
new file mode 100644
index 000000000..3e4f5ba10
--- /dev/null
+++ b/arch/mips/crypto/crc32-mips.c
@@ -0,0 +1,352 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * crc32-mips.c - CRC32 and CRC32C using optional MIPSr6 instructions
+ *
+ * Module based on arm64/crypto/crc32-arm.c
+ *
+ * Copyright (C) 2014 Linaro Ltd <yazen.ghannam@linaro.org>
+ * Copyright (C) 2018 MIPS Tech, LLC
+ */
+
+#include <linux/cpufeature.h>
+#include <linux/init.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/string.h>
+#include <asm/mipsregs.h>
+#include <asm/unaligned.h>
+
+#include <crypto/internal/hash.h>
+
+enum crc_op_size {
+	b, h, w, d,
+};
+
+enum crc_type {
+	crc32,
+	crc32c,
+};
+
+#ifndef TOOLCHAIN_SUPPORTS_CRC
+#define _ASM_SET_CRC(OP, SZ, TYPE)					  \
+_ASM_MACRO_3R(OP, rt, rs, rt2,						  \
+	".ifnc	\\rt, \\rt2\n\t"					  \
+	".error	\"invalid operands \\\"" #OP " \\rt,\\rs,\\rt2\\\"\"\n\t" \
+	".endif\n\t"							  \
+	_ASM_INSN_IF_MIPS(0x7c00000f | (__rt << 16) | (__rs << 21) |	  \
+			  ((SZ) <<  6) | ((TYPE) << 8))			  \
+	_ASM_INSN32_IF_MM(0x00000030 | (__rs << 16) | (__rt << 21) |	  \
+			  ((SZ) << 14) | ((TYPE) << 3)))
+#define _ASM_UNSET_CRC(op, SZ, TYPE) ".purgem " #op "\n\t"
+#else /* !TOOLCHAIN_SUPPORTS_CRC */
+#define _ASM_SET_CRC(op, SZ, TYPE) ".set\tcrc\n\t"
+#define _ASM_UNSET_CRC(op, SZ, TYPE)
+#endif
+
+#define __CRC32(crc, value, op, SZ, TYPE)		\
+do {							\
+	__asm__ __volatile__(				\
+		".set	push\n\t"			\
+		_ASM_SET_CRC(op, SZ, TYPE)		\
+		#op "	%0, %1, %0\n\t"			\
+		_ASM_UNSET_CRC(op, SZ, TYPE)		\
+		".set	pop"				\
+		: "+r" (crc)				\
+		: "r" (value));				\
+} while (0)
+
+#define _CRC32_crc32b(crc, value)	__CRC32(crc, value, crc32b, 0, 0)
+#define _CRC32_crc32h(crc, value)	__CRC32(crc, value, crc32h, 1, 0)
+#define _CRC32_crc32w(crc, value)	__CRC32(crc, value, crc32w, 2, 0)
+#define _CRC32_crc32d(crc, value)	__CRC32(crc, value, crc32d, 3, 0)
+#define _CRC32_crc32cb(crc, value)	__CRC32(crc, value, crc32cb, 0, 1)
+#define _CRC32_crc32ch(crc, value)	__CRC32(crc, value, crc32ch, 1, 1)
+#define _CRC32_crc32cw(crc, value)	__CRC32(crc, value, crc32cw, 2, 1)
+#define _CRC32_crc32cd(crc, value)	__CRC32(crc, value, crc32cd, 3, 1)
+
+#define _CRC32(crc, value, size, op) \
+	_CRC32_##op##size(crc, value)
+
+#define CRC32(crc, value, size) \
+	_CRC32(crc, value, size, crc32)
+
+#define CRC32C(crc, value, size) \
+	_CRC32(crc, value, size, crc32c)
+
+static u32 crc32_mips_le_hw(u32 crc_, const u8 *p, unsigned int len)
+{
+	u32 crc = crc_;
+
+#ifdef CONFIG_64BIT
+	while (len >= sizeof(u64)) {
+		u64 value = get_unaligned_le64(p);
+
+		CRC32(crc, value, d);
+		p += sizeof(u64);
+		len -= sizeof(u64);
+	}
+
+	if (len & sizeof(u32)) {
+#else /* !CONFIG_64BIT */
+	while (len >= sizeof(u32)) {
+#endif
+		u32 value = get_unaligned_le32(p);
+
+		CRC32(crc, value, w);
+		p += sizeof(u32);
+		len -= sizeof(u32);
+	}
+
+	if (len & sizeof(u16)) {
+		u16 value = get_unaligned_le16(p);
+
+		CRC32(crc, value, h);
+		p += sizeof(u16);
+	}
+
+	if (len & sizeof(u8)) {
+		u8 value = *p++;
+
+		CRC32(crc, value, b);
+	}
+
+	return crc;
+}
+
+static u32 crc32c_mips_le_hw(u32 crc_, const u8 *p, unsigned int len)
+{
+	u32 crc = crc_;
+
+#ifdef CONFIG_64BIT
+	while (len >= sizeof(u64)) {
+		u64 value = get_unaligned_le64(p);
+
+		CRC32C(crc, value, d);
+		p += sizeof(u64);
+		len -= sizeof(u64);
+	}
+
+	if (len & sizeof(u32)) {
+#else /* !CONFIG_64BIT */
+	while (len >= sizeof(u32)) {
+#endif
+		u32 value = get_unaligned_le32(p);
+
+		CRC32C(crc, value, w);
+		p += sizeof(u32);
+		len -= sizeof(u32);
+	}
+
+	if (len & sizeof(u16)) {
+		u16 value = get_unaligned_le16(p);
+
+		CRC32C(crc, value, h);
+		p += sizeof(u16);
+	}
+
+	if (len & sizeof(u8)) {
+		u8 value = *p++;
+
+		CRC32C(crc, value, b);
+	}
+	return crc;
+}
+
+#define CHKSUM_BLOCK_SIZE	1
+#define CHKSUM_DIGEST_SIZE	4
+
+struct chksum_ctx {
+	u32 key;
+};
+
+struct chksum_desc_ctx {
+	u32 crc;
+};
+
+static int chksum_init(struct shash_desc *desc)
+{
+	struct chksum_ctx *mctx = crypto_shash_ctx(desc->tfm);
+	struct chksum_desc_ctx *ctx = shash_desc_ctx(desc);
+
+	ctx->crc = mctx->key;
+
+	return 0;
+}
+
+/*
+ * Setting the seed allows arbitrary accumulators and flexible XOR policy
+ * If your algorithm starts with ~0, then XOR with ~0 before you set
+ * the seed.
+ */
+static int chksum_setkey(struct crypto_shash *tfm, const u8 *key,
+			 unsigned int keylen)
+{
+	struct chksum_ctx *mctx = crypto_shash_ctx(tfm);
+
+	if (keylen != sizeof(mctx->key))
+		return -EINVAL;
+	mctx->key = get_unaligned_le32(key);
+	return 0;
+}
+
+static int chksum_update(struct shash_desc *desc, const u8 *data,
+			 unsigned int length)
+{
+	struct chksum_desc_ctx *ctx = shash_desc_ctx(desc);
+
+	ctx->crc = crc32_mips_le_hw(ctx->crc, data, length);
+	return 0;
+}
+
+static int chksumc_update(struct shash_desc *desc, const u8 *data,
+			 unsigned int length)
+{
+	struct chksum_desc_ctx *ctx = shash_desc_ctx(desc);
+
+	ctx->crc = crc32c_mips_le_hw(ctx->crc, data, length);
+	return 0;
+}
+
+static int chksum_final(struct shash_desc *desc, u8 *out)
+{
+	struct chksum_desc_ctx *ctx = shash_desc_ctx(desc);
+
+	put_unaligned_le32(ctx->crc, out);
+	return 0;
+}
+
+static int chksumc_final(struct shash_desc *desc, u8 *out)
+{
+	struct chksum_desc_ctx *ctx = shash_desc_ctx(desc);
+
+	put_unaligned_le32(~ctx->crc, out);
+	return 0;
+}
+
+static int __chksum_finup(u32 crc, const u8 *data, unsigned int len, u8 *out)
+{
+	put_unaligned_le32(crc32_mips_le_hw(crc, data, len), out);
+	return 0;
+}
+
+static int __chksumc_finup(u32 crc, const u8 *data, unsigned int len, u8 *out)
+{
+	put_unaligned_le32(~crc32c_mips_le_hw(crc, data, len), out);
+	return 0;
+}
+
+static int chksum_finup(struct shash_desc *desc, const u8 *data,
+			unsigned int len, u8 *out)
+{
+	struct chksum_desc_ctx *ctx = shash_desc_ctx(desc);
+
+	return __chksum_finup(ctx->crc, data, len, out);
+}
+
+static int chksumc_finup(struct shash_desc *desc, const u8 *data,
+			unsigned int len, u8 *out)
+{
+	struct chksum_desc_ctx *ctx = shash_desc_ctx(desc);
+
+	return __chksumc_finup(ctx->crc, data, len, out);
+}
+
+static int chksum_digest(struct shash_desc *desc, const u8 *data,
+			 unsigned int length, u8 *out)
+{
+	struct chksum_ctx *mctx = crypto_shash_ctx(desc->tfm);
+
+	return __chksum_finup(mctx->key, data, length, out);
+}
+
+static int chksumc_digest(struct shash_desc *desc, const u8 *data,
+			 unsigned int length, u8 *out)
+{
+	struct chksum_ctx *mctx = crypto_shash_ctx(desc->tfm);
+
+	return __chksumc_finup(mctx->key, data, length, out);
+}
+
+static int chksum_cra_init(struct crypto_tfm *tfm)
+{
+	struct chksum_ctx *mctx = crypto_tfm_ctx(tfm);
+
+	mctx->key = ~0;
+	return 0;
+}
+
+static struct shash_alg crc32_alg = {
+	.digestsize		=	CHKSUM_DIGEST_SIZE,
+	.setkey			=	chksum_setkey,
+	.init			=	chksum_init,
+	.update			=	chksum_update,
+	.final			=	chksum_final,
+	.finup			=	chksum_finup,
+	.digest			=	chksum_digest,
+	.descsize		=	sizeof(struct chksum_desc_ctx),
+	.base			=	{
+		.cra_name		=	"crc32",
+		.cra_driver_name	=	"crc32-mips-hw",
+		.cra_priority		=	300,
+		.cra_flags		=	CRYPTO_ALG_OPTIONAL_KEY,
+		.cra_blocksize		=	CHKSUM_BLOCK_SIZE,
+		.cra_alignmask		=	0,
+		.cra_ctxsize		=	sizeof(struct chksum_ctx),
+		.cra_module		=	THIS_MODULE,
+		.cra_init		=	chksum_cra_init,
+	}
+};
+
+static struct shash_alg crc32c_alg = {
+	.digestsize		=	CHKSUM_DIGEST_SIZE,
+	.setkey			=	chksum_setkey,
+	.init			=	chksum_init,
+	.update			=	chksumc_update,
+	.final			=	chksumc_final,
+	.finup			=	chksumc_finup,
+	.digest			=	chksumc_digest,
+	.descsize		=	sizeof(struct chksum_desc_ctx),
+	.base			=	{
+		.cra_name		=	"crc32c",
+		.cra_driver_name	=	"crc32c-mips-hw",
+		.cra_priority		=	300,
+		.cra_flags		=	CRYPTO_ALG_OPTIONAL_KEY,
+		.cra_blocksize		=	CHKSUM_BLOCK_SIZE,
+		.cra_alignmask		=	0,
+		.cra_ctxsize		=	sizeof(struct chksum_ctx),
+		.cra_module		=	THIS_MODULE,
+		.cra_init		=	chksum_cra_init,
+	}
+};
+
+static int __init crc32_mod_init(void)
+{
+	int err;
+
+	err = crypto_register_shash(&crc32_alg);
+
+	if (err)
+		return err;
+
+	err = crypto_register_shash(&crc32c_alg);
+
+	if (err) {
+		crypto_unregister_shash(&crc32_alg);
+		return err;
+	}
+
+	return 0;
+}
+
+static void __exit crc32_mod_exit(void)
+{
+	crypto_unregister_shash(&crc32_alg);
+	crypto_unregister_shash(&crc32c_alg);
+}
+
+MODULE_AUTHOR("Marcin Nowakowski <marcin.nowakowski@mips.com");
+MODULE_DESCRIPTION("CRC32 and CRC32C using optional MIPS instructions");
+MODULE_LICENSE("GPL v2");
+
+module_cpu_feature_match(MIPS_CRC32, crc32_mod_init);
+module_exit(crc32_mod_exit);
diff --git a/arch/mips/crypto/poly1305-glue.c b/arch/mips/crypto/poly1305-glue.c
new file mode 100644
index 000000000..bc6110fb9
--- /dev/null
+++ b/arch/mips/crypto/poly1305-glue.c
@@ -0,0 +1,191 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * OpenSSL/Cryptogams accelerated Poly1305 transform for MIPS
+ *
+ * Copyright (C) 2019 Linaro Ltd. <ard.biesheuvel@linaro.org>
+ */
+
+#include <asm/unaligned.h>
+#include <crypto/algapi.h>
+#include <crypto/internal/hash.h>
+#include <crypto/internal/poly1305.h>
+#include <linux/cpufeature.h>
+#include <linux/crypto.h>
+#include <linux/module.h>
+
+asmlinkage void poly1305_init_mips(void *state, const u8 *key);
+asmlinkage void poly1305_blocks_mips(void *state, const u8 *src, u32 len, u32 hibit);
+asmlinkage void poly1305_emit_mips(void *state, u8 *digest, const u32 *nonce);
+
+void poly1305_init_arch(struct poly1305_desc_ctx *dctx, const u8 key[POLY1305_KEY_SIZE])
+{
+	poly1305_init_mips(&dctx->h, key);
+	dctx->s[0] = get_unaligned_le32(key + 16);
+	dctx->s[1] = get_unaligned_le32(key + 20);
+	dctx->s[2] = get_unaligned_le32(key + 24);
+	dctx->s[3] = get_unaligned_le32(key + 28);
+	dctx->buflen = 0;
+}
+EXPORT_SYMBOL(poly1305_init_arch);
+
+static int mips_poly1305_init(struct shash_desc *desc)
+{
+	struct poly1305_desc_ctx *dctx = shash_desc_ctx(desc);
+
+	dctx->buflen = 0;
+	dctx->rset = 0;
+	dctx->sset = false;
+
+	return 0;
+}
+
+static void mips_poly1305_blocks(struct poly1305_desc_ctx *dctx, const u8 *src,
+				 u32 len, u32 hibit)
+{
+	if (unlikely(!dctx->sset)) {
+		if (!dctx->rset) {
+			poly1305_init_mips(&dctx->h, src);
+			src += POLY1305_BLOCK_SIZE;
+			len -= POLY1305_BLOCK_SIZE;
+			dctx->rset = 1;
+		}
+		if (len >= POLY1305_BLOCK_SIZE) {
+			dctx->s[0] = get_unaligned_le32(src +  0);
+			dctx->s[1] = get_unaligned_le32(src +  4);
+			dctx->s[2] = get_unaligned_le32(src +  8);
+			dctx->s[3] = get_unaligned_le32(src + 12);
+			src += POLY1305_BLOCK_SIZE;
+			len -= POLY1305_BLOCK_SIZE;
+			dctx->sset = true;
+		}
+		if (len < POLY1305_BLOCK_SIZE)
+			return;
+	}
+
+	len &= ~(POLY1305_BLOCK_SIZE - 1);
+
+	poly1305_blocks_mips(&dctx->h, src, len, hibit);
+}
+
+static int mips_poly1305_update(struct shash_desc *desc, const u8 *src,
+				unsigned int len)
+{
+	struct poly1305_desc_ctx *dctx = shash_desc_ctx(desc);
+
+	if (unlikely(dctx->buflen)) {
+		u32 bytes = min(len, POLY1305_BLOCK_SIZE - dctx->buflen);
+
+		memcpy(dctx->buf + dctx->buflen, src, bytes);
+		src += bytes;
+		len -= bytes;
+		dctx->buflen += bytes;
+
+		if (dctx->buflen == POLY1305_BLOCK_SIZE) {
+			mips_poly1305_blocks(dctx, dctx->buf, POLY1305_BLOCK_SIZE, 1);
+			dctx->buflen = 0;
+		}
+	}
+
+	if (likely(len >= POLY1305_BLOCK_SIZE)) {
+		mips_poly1305_blocks(dctx, src, len, 1);
+		src += round_down(len, POLY1305_BLOCK_SIZE);
+		len %= POLY1305_BLOCK_SIZE;
+	}
+
+	if (unlikely(len)) {
+		dctx->buflen = len;
+		memcpy(dctx->buf, src, len);
+	}
+	return 0;
+}
+
+void poly1305_update_arch(struct poly1305_desc_ctx *dctx, const u8 *src,
+			  unsigned int nbytes)
+{
+	if (unlikely(dctx->buflen)) {
+		u32 bytes = min(nbytes, POLY1305_BLOCK_SIZE - dctx->buflen);
+
+		memcpy(dctx->buf + dctx->buflen, src, bytes);
+		src += bytes;
+		nbytes -= bytes;
+		dctx->buflen += bytes;
+
+		if (dctx->buflen == POLY1305_BLOCK_SIZE) {
+			poly1305_blocks_mips(&dctx->h, dctx->buf,
+					     POLY1305_BLOCK_SIZE, 1);
+			dctx->buflen = 0;
+		}
+	}
+
+	if (likely(nbytes >= POLY1305_BLOCK_SIZE)) {
+		unsigned int len = round_down(nbytes, POLY1305_BLOCK_SIZE);
+
+		poly1305_blocks_mips(&dctx->h, src, len, 1);
+		src += len;
+		nbytes %= POLY1305_BLOCK_SIZE;
+	}
+
+	if (unlikely(nbytes)) {
+		dctx->buflen = nbytes;
+		memcpy(dctx->buf, src, nbytes);
+	}
+}
+EXPORT_SYMBOL(poly1305_update_arch);
+
+void poly1305_final_arch(struct poly1305_desc_ctx *dctx, u8 *dst)
+{
+	if (unlikely(dctx->buflen)) {
+		dctx->buf[dctx->buflen++] = 1;
+		memset(dctx->buf + dctx->buflen, 0,
+		       POLY1305_BLOCK_SIZE - dctx->buflen);
+		poly1305_blocks_mips(&dctx->h, dctx->buf, POLY1305_BLOCK_SIZE, 0);
+	}
+
+	poly1305_emit_mips(&dctx->h, dst, dctx->s);
+	*dctx = (struct poly1305_desc_ctx){};
+}
+EXPORT_SYMBOL(poly1305_final_arch);
+
+static int mips_poly1305_final(struct shash_desc *desc, u8 *dst)
+{
+	struct poly1305_desc_ctx *dctx = shash_desc_ctx(desc);
+
+	if (unlikely(!dctx->sset))
+		return -ENOKEY;
+
+	poly1305_final_arch(dctx, dst);
+	return 0;
+}
+
+static struct shash_alg mips_poly1305_alg = {
+	.init			= mips_poly1305_init,
+	.update			= mips_poly1305_update,
+	.final			= mips_poly1305_final,
+	.digestsize		= POLY1305_DIGEST_SIZE,
+	.descsize		= sizeof(struct poly1305_desc_ctx),
+
+	.base.cra_name		= "poly1305",
+	.base.cra_driver_name	= "poly1305-mips",
+	.base.cra_priority	= 200,
+	.base.cra_blocksize	= POLY1305_BLOCK_SIZE,
+	.base.cra_module	= THIS_MODULE,
+};
+
+static int __init mips_poly1305_mod_init(void)
+{
+	return IS_REACHABLE(CONFIG_CRYPTO_HASH) ?
+		crypto_register_shash(&mips_poly1305_alg) : 0;
+}
+
+static void __exit mips_poly1305_mod_exit(void)
+{
+	if (IS_REACHABLE(CONFIG_CRYPTO_HASH))
+		crypto_unregister_shash(&mips_poly1305_alg);
+}
+
+module_init(mips_poly1305_mod_init);
+module_exit(mips_poly1305_mod_exit);
+
+MODULE_LICENSE("GPL v2");
+MODULE_ALIAS_CRYPTO("poly1305");
+MODULE_ALIAS_CRYPTO("poly1305-mips");
diff --git a/arch/mips/crypto/poly1305-mips.pl b/arch/mips/crypto/poly1305-mips.pl
new file mode 100644
index 000000000..b05bab884
--- /dev/null
+++ b/arch/mips/crypto/poly1305-mips.pl
@@ -0,0 +1,1273 @@
+#!/usr/bin/env perl
+# SPDX-License-Identifier: GPL-1.0+ OR BSD-3-Clause
+#
+# ====================================================================
+# Written by Andy Polyakov, @dot-asm, originally for the OpenSSL
+# project.
+# ====================================================================
+
+# Poly1305 hash for MIPS.
+#
+# May 2016
+#
+# Numbers are cycles per processed byte with poly1305_blocks alone.
+#
+#		IALU/gcc
+# R1x000	~5.5/+130%	(big-endian)
+# Octeon II	2.50/+70%	(little-endian)
+#
+# March 2019
+#
+# Add 32-bit code path.
+#
+# October 2019
+#
+# Modulo-scheduling reduction allows to omit dependency chain at the
+# end of inner loop and improve performance. Also optimize MIPS32R2
+# code path for MIPS 1004K core. Per René von Dorst's suggestions.
+#
+#		IALU/gcc
+# R1x000	~9.8/?		(big-endian)
+# Octeon II	3.65/+140%	(little-endian)
+# MT7621/1004K	4.75/?		(little-endian)
+#
+######################################################################
+# There is a number of MIPS ABI in use, O32 and N32/64 are most
+# widely used. Then there is a new contender: NUBI. It appears that if
+# one picks the latter, it's possible to arrange code in ABI neutral
+# manner. Therefore let's stick to NUBI register layout:
+#
+($zero,$at,$t0,$t1,$t2)=map("\$$_",(0..2,24,25));
+($a0,$a1,$a2,$a3,$a4,$a5,$a6,$a7)=map("\$$_",(4..11));
+($s0,$s1,$s2,$s3,$s4,$s5,$s6,$s7,$s8,$s9,$s10,$s11)=map("\$$_",(12..23));
+($gp,$tp,$sp,$fp,$ra)=map("\$$_",(3,28..31));
+#
+# The return value is placed in $a0. Following coding rules facilitate
+# interoperability:
+#
+# - never ever touch $tp, "thread pointer", former $gp [o32 can be
+#   excluded from the rule, because it's specified volatile];
+# - copy return value to $t0, former $v0 [or to $a0 if you're adapting
+#   old code];
+# - on O32 populate $a4-$a7 with 'lw $aN,4*N($sp)' if necessary;
+#
+# For reference here is register layout for N32/64 MIPS ABIs:
+#
+# ($zero,$at,$v0,$v1)=map("\$$_",(0..3));
+# ($a0,$a1,$a2,$a3,$a4,$a5,$a6,$a7)=map("\$$_",(4..11));
+# ($t0,$t1,$t2,$t3,$t8,$t9)=map("\$$_",(12..15,24,25));
+# ($s0,$s1,$s2,$s3,$s4,$s5,$s6,$s7)=map("\$$_",(16..23));
+# ($gp,$sp,$fp,$ra)=map("\$$_",(28..31));
+#
+# <appro@openssl.org>
+#
+######################################################################
+
+$flavour = shift || "64"; # supported flavours are o32,n32,64,nubi32,nubi64
+
+$v0 = ($flavour =~ /nubi/i) ? $a0 : $t0;
+
+if ($flavour =~ /64|n32/i) {{{
+######################################################################
+# 64-bit code path
+#
+
+my ($ctx,$inp,$len,$padbit) = ($a0,$a1,$a2,$a3);
+my ($in0,$in1,$tmp0,$tmp1,$tmp2,$tmp3,$tmp4) = ($a4,$a5,$a6,$a7,$at,$t0,$t1);
+
+$code.=<<___;
+#if (defined(_MIPS_ARCH_MIPS64R3) || defined(_MIPS_ARCH_MIPS64R5) || \\
+     defined(_MIPS_ARCH_MIPS64R6)) \\
+     && !defined(_MIPS_ARCH_MIPS64R2)
+# define _MIPS_ARCH_MIPS64R2
+#endif
+
+#if defined(_MIPS_ARCH_MIPS64R6)
+# define dmultu(rs,rt)
+# define mflo(rd,rs,rt)	dmulu	rd,rs,rt
+# define mfhi(rd,rs,rt)	dmuhu	rd,rs,rt
+#else
+# define dmultu(rs,rt)		dmultu	rs,rt
+# define mflo(rd,rs,rt)	mflo	rd
+# define mfhi(rd,rs,rt)	mfhi	rd
+#endif
+
+#ifdef	__KERNEL__
+# define poly1305_init   poly1305_init_mips
+# define poly1305_blocks poly1305_blocks_mips
+# define poly1305_emit   poly1305_emit_mips
+#endif
+
+#if defined(__MIPSEB__) && !defined(MIPSEB)
+# define MIPSEB
+#endif
+
+#ifdef MIPSEB
+# define MSB 0
+# define LSB 7
+#else
+# define MSB 7
+# define LSB 0
+#endif
+
+.text
+.set	noat
+.set	noreorder
+
+.align	5
+.globl	poly1305_init
+.ent	poly1305_init
+poly1305_init:
+	.frame	$sp,0,$ra
+	.set	reorder
+
+	sd	$zero,0($ctx)
+	sd	$zero,8($ctx)
+	sd	$zero,16($ctx)
+
+	beqz	$inp,.Lno_key
+
+#if defined(_MIPS_ARCH_MIPS64R6)
+	andi	$tmp0,$inp,7		# $inp % 8
+	dsubu	$inp,$inp,$tmp0		# align $inp
+	sll	$tmp0,$tmp0,3		# byte to bit offset
+	ld	$in0,0($inp)
+	ld	$in1,8($inp)
+	beqz	$tmp0,.Laligned_key
+	ld	$tmp2,16($inp)
+
+	subu	$tmp1,$zero,$tmp0
+# ifdef	MIPSEB
+	dsllv	$in0,$in0,$tmp0
+	dsrlv	$tmp3,$in1,$tmp1
+	dsllv	$in1,$in1,$tmp0
+	dsrlv	$tmp2,$tmp2,$tmp1
+# else
+	dsrlv	$in0,$in0,$tmp0
+	dsllv	$tmp3,$in1,$tmp1
+	dsrlv	$in1,$in1,$tmp0
+	dsllv	$tmp2,$tmp2,$tmp1
+# endif
+	or	$in0,$in0,$tmp3
+	or	$in1,$in1,$tmp2
+.Laligned_key:
+#else
+	ldl	$in0,0+MSB($inp)
+	ldl	$in1,8+MSB($inp)
+	ldr	$in0,0+LSB($inp)
+	ldr	$in1,8+LSB($inp)
+#endif
+#ifdef	MIPSEB
+# if defined(_MIPS_ARCH_MIPS64R2)
+	dsbh	$in0,$in0		# byte swap
+	 dsbh	$in1,$in1
+	dshd	$in0,$in0
+	 dshd	$in1,$in1
+# else
+	ori	$tmp0,$zero,0xFF
+	dsll	$tmp2,$tmp0,32
+	or	$tmp0,$tmp2		# 0x000000FF000000FF
+
+	and	$tmp1,$in0,$tmp0	# byte swap
+	 and	$tmp3,$in1,$tmp0
+	dsrl	$tmp2,$in0,24
+	 dsrl	$tmp4,$in1,24
+	dsll	$tmp1,24
+	 dsll	$tmp3,24
+	and	$tmp2,$tmp0
+	 and	$tmp4,$tmp0
+	dsll	$tmp0,8			# 0x0000FF000000FF00
+	or	$tmp1,$tmp2
+	 or	$tmp3,$tmp4
+	and	$tmp2,$in0,$tmp0
+	 and	$tmp4,$in1,$tmp0
+	dsrl	$in0,8
+	 dsrl	$in1,8
+	dsll	$tmp2,8
+	 dsll	$tmp4,8
+	and	$in0,$tmp0
+	 and	$in1,$tmp0
+	or	$tmp1,$tmp2
+	 or	$tmp3,$tmp4
+	or	$in0,$tmp1
+	 or	$in1,$tmp3
+	dsrl	$tmp1,$in0,32
+	 dsrl	$tmp3,$in1,32
+	dsll	$in0,32
+	 dsll	$in1,32
+	or	$in0,$tmp1
+	 or	$in1,$tmp3
+# endif
+#endif
+	li	$tmp0,1
+	dsll	$tmp0,32		# 0x0000000100000000
+	daddiu	$tmp0,-63		# 0x00000000ffffffc1
+	dsll	$tmp0,28		# 0x0ffffffc10000000
+	daddiu	$tmp0,-1		# 0x0ffffffc0fffffff
+
+	and	$in0,$tmp0
+	daddiu	$tmp0,-3		# 0x0ffffffc0ffffffc
+	and	$in1,$tmp0
+
+	sd	$in0,24($ctx)
+	dsrl	$tmp0,$in1,2
+	sd	$in1,32($ctx)
+	daddu	$tmp0,$in1		# s1 = r1 + (r1 >> 2)
+	sd	$tmp0,40($ctx)
+
+.Lno_key:
+	li	$v0,0			# return 0
+	jr	$ra
+.end	poly1305_init
+___
+{
+my $SAVED_REGS_MASK = ($flavour =~ /nubi/i) ? "0x0003f000" : "0x00030000";
+
+my ($h0,$h1,$h2,$r0,$r1,$rs1,$d0,$d1,$d2) =
+   ($s0,$s1,$s2,$s3,$s4,$s5,$in0,$in1,$t2);
+my ($shr,$shl) = ($s6,$s7);		# used on R6
+
+$code.=<<___;
+.align	5
+.globl	poly1305_blocks
+.ent	poly1305_blocks
+poly1305_blocks:
+	.set	noreorder
+	dsrl	$len,4			# number of complete blocks
+	bnez	$len,poly1305_blocks_internal
+	nop
+	jr	$ra
+	nop
+.end	poly1305_blocks
+
+.align	5
+.ent	poly1305_blocks_internal
+poly1305_blocks_internal:
+	.set	noreorder
+#if defined(_MIPS_ARCH_MIPS64R6)
+	.frame	$sp,8*8,$ra
+	.mask	$SAVED_REGS_MASK|0x000c0000,-8
+	dsubu	$sp,8*8
+	sd	$s7,56($sp)
+	sd	$s6,48($sp)
+#else
+	.frame	$sp,6*8,$ra
+	.mask	$SAVED_REGS_MASK,-8
+	dsubu	$sp,6*8
+#endif
+	sd	$s5,40($sp)
+	sd	$s4,32($sp)
+___
+$code.=<<___ if ($flavour =~ /nubi/i);	# optimize non-nubi prologue
+	sd	$s3,24($sp)
+	sd	$s2,16($sp)
+	sd	$s1,8($sp)
+	sd	$s0,0($sp)
+___
+$code.=<<___;
+	.set	reorder
+
+#if defined(_MIPS_ARCH_MIPS64R6)
+	andi	$shr,$inp,7
+	dsubu	$inp,$inp,$shr		# align $inp
+	sll	$shr,$shr,3		# byte to bit offset
+	subu	$shl,$zero,$shr
+#endif
+
+	ld	$h0,0($ctx)		# load hash value
+	ld	$h1,8($ctx)
+	ld	$h2,16($ctx)
+
+	ld	$r0,24($ctx)		# load key
+	ld	$r1,32($ctx)
+	ld	$rs1,40($ctx)
+
+	dsll	$len,4
+	daddu	$len,$inp		# end of buffer
+	b	.Loop
+
+.align	4
+.Loop:
+#if defined(_MIPS_ARCH_MIPS64R6)
+	ld	$in0,0($inp)		# load input
+	ld	$in1,8($inp)
+	beqz	$shr,.Laligned_inp
+
+	ld	$tmp2,16($inp)
+# ifdef	MIPSEB
+	dsllv	$in0,$in0,$shr
+	dsrlv	$tmp3,$in1,$shl
+	dsllv	$in1,$in1,$shr
+	dsrlv	$tmp2,$tmp2,$shl
+# else
+	dsrlv	$in0,$in0,$shr
+	dsllv	$tmp3,$in1,$shl
+	dsrlv	$in1,$in1,$shr
+	dsllv	$tmp2,$tmp2,$shl
+# endif
+	or	$in0,$in0,$tmp3
+	or	$in1,$in1,$tmp2
+.Laligned_inp:
+#else
+	ldl	$in0,0+MSB($inp)	# load input
+	ldl	$in1,8+MSB($inp)
+	ldr	$in0,0+LSB($inp)
+	ldr	$in1,8+LSB($inp)
+#endif
+	daddiu	$inp,16
+#ifdef	MIPSEB
+# if defined(_MIPS_ARCH_MIPS64R2)
+	dsbh	$in0,$in0		# byte swap
+	 dsbh	$in1,$in1
+	dshd	$in0,$in0
+	 dshd	$in1,$in1
+# else
+	ori	$tmp0,$zero,0xFF
+	dsll	$tmp2,$tmp0,32
+	or	$tmp0,$tmp2		# 0x000000FF000000FF
+
+	and	$tmp1,$in0,$tmp0	# byte swap
+	 and	$tmp3,$in1,$tmp0
+	dsrl	$tmp2,$in0,24
+	 dsrl	$tmp4,$in1,24
+	dsll	$tmp1,24
+	 dsll	$tmp3,24
+	and	$tmp2,$tmp0
+	 and	$tmp4,$tmp0
+	dsll	$tmp0,8			# 0x0000FF000000FF00
+	or	$tmp1,$tmp2
+	 or	$tmp3,$tmp4
+	and	$tmp2,$in0,$tmp0
+	 and	$tmp4,$in1,$tmp0
+	dsrl	$in0,8
+	 dsrl	$in1,8
+	dsll	$tmp2,8
+	 dsll	$tmp4,8
+	and	$in0,$tmp0
+	 and	$in1,$tmp0
+	or	$tmp1,$tmp2
+	 or	$tmp3,$tmp4
+	or	$in0,$tmp1
+	 or	$in1,$tmp3
+	dsrl	$tmp1,$in0,32
+	 dsrl	$tmp3,$in1,32
+	dsll	$in0,32
+	 dsll	$in1,32
+	or	$in0,$tmp1
+	 or	$in1,$tmp3
+# endif
+#endif
+	dsrl	$tmp1,$h2,2		# modulo-scheduled reduction
+	andi	$h2,$h2,3
+	dsll	$tmp0,$tmp1,2
+
+	daddu	$d0,$h0,$in0		# accumulate input
+	 daddu	$tmp1,$tmp0
+	sltu	$tmp0,$d0,$h0
+	daddu	$d0,$d0,$tmp1		# ... and residue
+	sltu	$tmp1,$d0,$tmp1
+	daddu	$d1,$h1,$in1
+	daddu	$tmp0,$tmp1
+	sltu	$tmp1,$d1,$h1
+	daddu	$d1,$tmp0
+
+	dmultu	($r0,$d0)		# h0*r0
+	 daddu	$d2,$h2,$padbit
+	 sltu	$tmp0,$d1,$tmp0
+	mflo	($h0,$r0,$d0)
+	mfhi	($h1,$r0,$d0)
+
+	dmultu	($rs1,$d1)		# h1*5*r1
+	 daddu	$d2,$tmp1
+	 daddu	$d2,$tmp0
+	mflo	($tmp0,$rs1,$d1)
+	mfhi	($tmp1,$rs1,$d1)
+
+	dmultu	($r1,$d0)		# h0*r1
+	mflo	($tmp2,$r1,$d0)
+	mfhi	($h2,$r1,$d0)
+	 daddu	$h0,$tmp0
+	 daddu	$h1,$tmp1
+	 sltu	$tmp0,$h0,$tmp0
+
+	dmultu	($r0,$d1)		# h1*r0
+	 daddu	$h1,$tmp0
+	 daddu	$h1,$tmp2
+	mflo	($tmp0,$r0,$d1)
+	mfhi	($tmp1,$r0,$d1)
+
+	dmultu	($rs1,$d2)		# h2*5*r1
+	 sltu	$tmp2,$h1,$tmp2
+	 daddu	$h2,$tmp2
+	mflo	($tmp2,$rs1,$d2)
+
+	dmultu	($r0,$d2)		# h2*r0
+	 daddu	$h1,$tmp0
+	 daddu	$h2,$tmp1
+	mflo	($tmp3,$r0,$d2)
+	 sltu	$tmp0,$h1,$tmp0
+	 daddu	$h2,$tmp0
+
+	daddu	$h1,$tmp2
+	sltu	$tmp2,$h1,$tmp2
+	daddu	$h2,$tmp2
+	daddu	$h2,$tmp3
+
+	bne	$inp,$len,.Loop
+
+	sd	$h0,0($ctx)		# store hash value
+	sd	$h1,8($ctx)
+	sd	$h2,16($ctx)
+
+	.set	noreorder
+#if defined(_MIPS_ARCH_MIPS64R6)
+	ld	$s7,56($sp)
+	ld	$s6,48($sp)
+#endif
+	ld	$s5,40($sp)		# epilogue
+	ld	$s4,32($sp)
+___
+$code.=<<___ if ($flavour =~ /nubi/i);	# optimize non-nubi epilogue
+	ld	$s3,24($sp)
+	ld	$s2,16($sp)
+	ld	$s1,8($sp)
+	ld	$s0,0($sp)
+___
+$code.=<<___;
+	jr	$ra
+#if defined(_MIPS_ARCH_MIPS64R6)
+	daddu	$sp,8*8
+#else
+	daddu	$sp,6*8
+#endif
+.end	poly1305_blocks_internal
+___
+}
+{
+my ($ctx,$mac,$nonce) = ($a0,$a1,$a2);
+
+$code.=<<___;
+.align	5
+.globl	poly1305_emit
+.ent	poly1305_emit
+poly1305_emit:
+	.frame	$sp,0,$ra
+	.set	reorder
+
+	ld	$tmp2,16($ctx)
+	ld	$tmp0,0($ctx)
+	ld	$tmp1,8($ctx)
+
+	li	$in0,-4			# final reduction
+	dsrl	$in1,$tmp2,2
+	and	$in0,$tmp2
+	andi	$tmp2,$tmp2,3
+	daddu	$in0,$in1
+
+	daddu	$tmp0,$tmp0,$in0
+	sltu	$in1,$tmp0,$in0
+	 daddiu	$in0,$tmp0,5		# compare to modulus
+	daddu	$tmp1,$tmp1,$in1
+	 sltiu	$tmp3,$in0,5
+	sltu	$tmp4,$tmp1,$in1
+	 daddu	$in1,$tmp1,$tmp3
+	daddu	$tmp2,$tmp2,$tmp4
+	 sltu	$tmp3,$in1,$tmp3
+	 daddu	$tmp2,$tmp2,$tmp3
+
+	dsrl	$tmp2,2			# see if it carried/borrowed
+	dsubu	$tmp2,$zero,$tmp2
+
+	xor	$in0,$tmp0
+	xor	$in1,$tmp1
+	and	$in0,$tmp2
+	and	$in1,$tmp2
+	xor	$in0,$tmp0
+	xor	$in1,$tmp1
+
+	lwu	$tmp0,0($nonce)		# load nonce
+	lwu	$tmp1,4($nonce)
+	lwu	$tmp2,8($nonce)
+	lwu	$tmp3,12($nonce)
+	dsll	$tmp1,32
+	dsll	$tmp3,32
+	or	$tmp0,$tmp1
+	or	$tmp2,$tmp3
+
+	daddu	$in0,$tmp0		# accumulate nonce
+	daddu	$in1,$tmp2
+	sltu	$tmp0,$in0,$tmp0
+	daddu	$in1,$tmp0
+
+	dsrl	$tmp0,$in0,8		# write mac value
+	dsrl	$tmp1,$in0,16
+	dsrl	$tmp2,$in0,24
+	sb	$in0,0($mac)
+	dsrl	$tmp3,$in0,32
+	sb	$tmp0,1($mac)
+	dsrl	$tmp0,$in0,40
+	sb	$tmp1,2($mac)
+	dsrl	$tmp1,$in0,48
+	sb	$tmp2,3($mac)
+	dsrl	$tmp2,$in0,56
+	sb	$tmp3,4($mac)
+	dsrl	$tmp3,$in1,8
+	sb	$tmp0,5($mac)
+	dsrl	$tmp0,$in1,16
+	sb	$tmp1,6($mac)
+	dsrl	$tmp1,$in1,24
+	sb	$tmp2,7($mac)
+
+	sb	$in1,8($mac)
+	dsrl	$tmp2,$in1,32
+	sb	$tmp3,9($mac)
+	dsrl	$tmp3,$in1,40
+	sb	$tmp0,10($mac)
+	dsrl	$tmp0,$in1,48
+	sb	$tmp1,11($mac)
+	dsrl	$tmp1,$in1,56
+	sb	$tmp2,12($mac)
+	sb	$tmp3,13($mac)
+	sb	$tmp0,14($mac)
+	sb	$tmp1,15($mac)
+
+	jr	$ra
+.end	poly1305_emit
+.rdata
+.asciiz	"Poly1305 for MIPS64, CRYPTOGAMS by \@dot-asm"
+.align	2
+___
+}
+}}} else {{{
+######################################################################
+# 32-bit code path
+#
+
+my ($ctx,$inp,$len,$padbit) = ($a0,$a1,$a2,$a3);
+my ($in0,$in1,$in2,$in3,$tmp0,$tmp1,$tmp2,$tmp3) =
+   ($a4,$a5,$a6,$a7,$at,$t0,$t1,$t2);
+
+$code.=<<___;
+#if (defined(_MIPS_ARCH_MIPS32R3) || defined(_MIPS_ARCH_MIPS32R5) || \\
+     defined(_MIPS_ARCH_MIPS32R6)) \\
+     && !defined(_MIPS_ARCH_MIPS32R2)
+# define _MIPS_ARCH_MIPS32R2
+#endif
+
+#if defined(_MIPS_ARCH_MIPS32R6)
+# define multu(rs,rt)
+# define mflo(rd,rs,rt)	mulu	rd,rs,rt
+# define mfhi(rd,rs,rt)	muhu	rd,rs,rt
+#else
+# define multu(rs,rt)	multu	rs,rt
+# define mflo(rd,rs,rt)	mflo	rd
+# define mfhi(rd,rs,rt)	mfhi	rd
+#endif
+
+#ifdef	__KERNEL__
+# define poly1305_init   poly1305_init_mips
+# define poly1305_blocks poly1305_blocks_mips
+# define poly1305_emit   poly1305_emit_mips
+#endif
+
+#if defined(__MIPSEB__) && !defined(MIPSEB)
+# define MIPSEB
+#endif
+
+#ifdef MIPSEB
+# define MSB 0
+# define LSB 3
+#else
+# define MSB 3
+# define LSB 0
+#endif
+
+.text
+.set	noat
+.set	noreorder
+
+.align	5
+.globl	poly1305_init
+.ent	poly1305_init
+poly1305_init:
+	.frame	$sp,0,$ra
+	.set	reorder
+
+	sw	$zero,0($ctx)
+	sw	$zero,4($ctx)
+	sw	$zero,8($ctx)
+	sw	$zero,12($ctx)
+	sw	$zero,16($ctx)
+
+	beqz	$inp,.Lno_key
+
+#if defined(_MIPS_ARCH_MIPS32R6)
+	andi	$tmp0,$inp,3		# $inp % 4
+	subu	$inp,$inp,$tmp0		# align $inp
+	sll	$tmp0,$tmp0,3		# byte to bit offset
+	lw	$in0,0($inp)
+	lw	$in1,4($inp)
+	lw	$in2,8($inp)
+	lw	$in3,12($inp)
+	beqz	$tmp0,.Laligned_key
+
+	lw	$tmp2,16($inp)
+	subu	$tmp1,$zero,$tmp0
+# ifdef	MIPSEB
+	sllv	$in0,$in0,$tmp0
+	srlv	$tmp3,$in1,$tmp1
+	sllv	$in1,$in1,$tmp0
+	or	$in0,$in0,$tmp3
+	srlv	$tmp3,$in2,$tmp1
+	sllv	$in2,$in2,$tmp0
+	or	$in1,$in1,$tmp3
+	srlv	$tmp3,$in3,$tmp1
+	sllv	$in3,$in3,$tmp0
+	or	$in2,$in2,$tmp3
+	srlv	$tmp2,$tmp2,$tmp1
+	or	$in3,$in3,$tmp2
+# else
+	srlv	$in0,$in0,$tmp0
+	sllv	$tmp3,$in1,$tmp1
+	srlv	$in1,$in1,$tmp0
+	or	$in0,$in0,$tmp3
+	sllv	$tmp3,$in2,$tmp1
+	srlv	$in2,$in2,$tmp0
+	or	$in1,$in1,$tmp3
+	sllv	$tmp3,$in3,$tmp1
+	srlv	$in3,$in3,$tmp0
+	or	$in2,$in2,$tmp3
+	sllv	$tmp2,$tmp2,$tmp1
+	or	$in3,$in3,$tmp2
+# endif
+.Laligned_key:
+#else
+	lwl	$in0,0+MSB($inp)
+	lwl	$in1,4+MSB($inp)
+	lwl	$in2,8+MSB($inp)
+	lwl	$in3,12+MSB($inp)
+	lwr	$in0,0+LSB($inp)
+	lwr	$in1,4+LSB($inp)
+	lwr	$in2,8+LSB($inp)
+	lwr	$in3,12+LSB($inp)
+#endif
+#ifdef	MIPSEB
+# if defined(_MIPS_ARCH_MIPS32R2)
+	wsbh	$in0,$in0		# byte swap
+	wsbh	$in1,$in1
+	wsbh	$in2,$in2
+	wsbh	$in3,$in3
+	rotr	$in0,$in0,16
+	rotr	$in1,$in1,16
+	rotr	$in2,$in2,16
+	rotr	$in3,$in3,16
+# else
+	srl	$tmp0,$in0,24		# byte swap
+	srl	$tmp1,$in0,8
+	andi	$tmp2,$in0,0xFF00
+	sll	$in0,$in0,24
+	andi	$tmp1,0xFF00
+	sll	$tmp2,$tmp2,8
+	or	$in0,$tmp0
+	 srl	$tmp0,$in1,24
+	or	$tmp1,$tmp2
+	 srl	$tmp2,$in1,8
+	or	$in0,$tmp1
+	 andi	$tmp1,$in1,0xFF00
+	 sll	$in1,$in1,24
+	 andi	$tmp2,0xFF00
+	 sll	$tmp1,$tmp1,8
+	 or	$in1,$tmp0
+	srl	$tmp0,$in2,24
+	 or	$tmp2,$tmp1
+	srl	$tmp1,$in2,8
+	 or	$in1,$tmp2
+	andi	$tmp2,$in2,0xFF00
+	sll	$in2,$in2,24
+	andi	$tmp1,0xFF00
+	sll	$tmp2,$tmp2,8
+	or	$in2,$tmp0
+	 srl	$tmp0,$in3,24
+	or	$tmp1,$tmp2
+	 srl	$tmp2,$in3,8
+	or	$in2,$tmp1
+	 andi	$tmp1,$in3,0xFF00
+	 sll	$in3,$in3,24
+	 andi	$tmp2,0xFF00
+	 sll	$tmp1,$tmp1,8
+	 or	$in3,$tmp0
+	 or	$tmp2,$tmp1
+	 or	$in3,$tmp2
+# endif
+#endif
+	lui	$tmp0,0x0fff
+	ori	$tmp0,0xffff		# 0x0fffffff
+	and	$in0,$in0,$tmp0
+	subu	$tmp0,3			# 0x0ffffffc
+	and	$in1,$in1,$tmp0
+	and	$in2,$in2,$tmp0
+	and	$in3,$in3,$tmp0
+
+	sw	$in0,20($ctx)
+	sw	$in1,24($ctx)
+	sw	$in2,28($ctx)
+	sw	$in3,32($ctx)
+
+	srl	$tmp1,$in1,2
+	srl	$tmp2,$in2,2
+	srl	$tmp3,$in3,2
+	addu	$in1,$in1,$tmp1		# s1 = r1 + (r1 >> 2)
+	addu	$in2,$in2,$tmp2
+	addu	$in3,$in3,$tmp3
+	sw	$in1,36($ctx)
+	sw	$in2,40($ctx)
+	sw	$in3,44($ctx)
+.Lno_key:
+	li	$v0,0
+	jr	$ra
+.end	poly1305_init
+___
+{
+my $SAVED_REGS_MASK = ($flavour =~ /nubi/i) ? "0x00fff000" : "0x00ff0000";
+
+my ($h0,$h1,$h2,$h3,$h4, $r0,$r1,$r2,$r3, $rs1,$rs2,$rs3) =
+   ($s0,$s1,$s2,$s3,$s4, $s5,$s6,$s7,$s8, $s9,$s10,$s11);
+my ($d0,$d1,$d2,$d3) =
+   ($a4,$a5,$a6,$a7);
+my $shr = $t2;		# used on R6
+my $one = $t2;		# used on R2
+
+$code.=<<___;
+.globl	poly1305_blocks
+.align	5
+.ent	poly1305_blocks
+poly1305_blocks:
+	.frame	$sp,16*4,$ra
+	.mask	$SAVED_REGS_MASK,-4
+	.set	noreorder
+	subu	$sp, $sp,4*12
+	sw	$s11,4*11($sp)
+	sw	$s10,4*10($sp)
+	sw	$s9, 4*9($sp)
+	sw	$s8, 4*8($sp)
+	sw	$s7, 4*7($sp)
+	sw	$s6, 4*6($sp)
+	sw	$s5, 4*5($sp)
+	sw	$s4, 4*4($sp)
+___
+$code.=<<___ if ($flavour =~ /nubi/i);	# optimize non-nubi prologue
+	sw	$s3, 4*3($sp)
+	sw	$s2, 4*2($sp)
+	sw	$s1, 4*1($sp)
+	sw	$s0, 4*0($sp)
+___
+$code.=<<___;
+	.set	reorder
+
+	srl	$len,4			# number of complete blocks
+	li	$one,1
+	beqz	$len,.Labort
+
+#if defined(_MIPS_ARCH_MIPS32R6)
+	andi	$shr,$inp,3
+	subu	$inp,$inp,$shr		# align $inp
+	sll	$shr,$shr,3		# byte to bit offset
+#endif
+
+	lw	$h0,0($ctx)		# load hash value
+	lw	$h1,4($ctx)
+	lw	$h2,8($ctx)
+	lw	$h3,12($ctx)
+	lw	$h4,16($ctx)
+
+	lw	$r0,20($ctx)		# load key
+	lw	$r1,24($ctx)
+	lw	$r2,28($ctx)
+	lw	$r3,32($ctx)
+	lw	$rs1,36($ctx)
+	lw	$rs2,40($ctx)
+	lw	$rs3,44($ctx)
+
+	sll	$len,4
+	addu	$len,$len,$inp		# end of buffer
+	b	.Loop
+
+.align	4
+.Loop:
+#if defined(_MIPS_ARCH_MIPS32R6)
+	lw	$d0,0($inp)		# load input
+	lw	$d1,4($inp)
+	lw	$d2,8($inp)
+	lw	$d3,12($inp)
+	beqz	$shr,.Laligned_inp
+
+	lw	$t0,16($inp)
+	subu	$t1,$zero,$shr
+# ifdef	MIPSEB
+	sllv	$d0,$d0,$shr
+	srlv	$at,$d1,$t1
+	sllv	$d1,$d1,$shr
+	or	$d0,$d0,$at
+	srlv	$at,$d2,$t1
+	sllv	$d2,$d2,$shr
+	or	$d1,$d1,$at
+	srlv	$at,$d3,$t1
+	sllv	$d3,$d3,$shr
+	or	$d2,$d2,$at
+	srlv	$t0,$t0,$t1
+	or	$d3,$d3,$t0
+# else
+	srlv	$d0,$d0,$shr
+	sllv	$at,$d1,$t1
+	srlv	$d1,$d1,$shr
+	or	$d0,$d0,$at
+	sllv	$at,$d2,$t1
+	srlv	$d2,$d2,$shr
+	or	$d1,$d1,$at
+	sllv	$at,$d3,$t1
+	srlv	$d3,$d3,$shr
+	or	$d2,$d2,$at
+	sllv	$t0,$t0,$t1
+	or	$d3,$d3,$t0
+# endif
+.Laligned_inp:
+#else
+	lwl	$d0,0+MSB($inp)		# load input
+	lwl	$d1,4+MSB($inp)
+	lwl	$d2,8+MSB($inp)
+	lwl	$d3,12+MSB($inp)
+	lwr	$d0,0+LSB($inp)
+	lwr	$d1,4+LSB($inp)
+	lwr	$d2,8+LSB($inp)
+	lwr	$d3,12+LSB($inp)
+#endif
+#ifdef	MIPSEB
+# if defined(_MIPS_ARCH_MIPS32R2)
+	wsbh	$d0,$d0			# byte swap
+	wsbh	$d1,$d1
+	wsbh	$d2,$d2
+	wsbh	$d3,$d3
+	rotr	$d0,$d0,16
+	rotr	$d1,$d1,16
+	rotr	$d2,$d2,16
+	rotr	$d3,$d3,16
+# else
+	srl	$at,$d0,24		# byte swap
+	srl	$t0,$d0,8
+	andi	$t1,$d0,0xFF00
+	sll	$d0,$d0,24
+	andi	$t0,0xFF00
+	sll	$t1,$t1,8
+	or	$d0,$at
+	 srl	$at,$d1,24
+	or	$t0,$t1
+	 srl	$t1,$d1,8
+	or	$d0,$t0
+	 andi	$t0,$d1,0xFF00
+	 sll	$d1,$d1,24
+	 andi	$t1,0xFF00
+	 sll	$t0,$t0,8
+	 or	$d1,$at
+	srl	$at,$d2,24
+	 or	$t1,$t0
+	srl	$t0,$d2,8
+	 or	$d1,$t1
+	andi	$t1,$d2,0xFF00
+	sll	$d2,$d2,24
+	andi	$t0,0xFF00
+	sll	$t1,$t1,8
+	or	$d2,$at
+	 srl	$at,$d3,24
+	or	$t0,$t1
+	 srl	$t1,$d3,8
+	or	$d2,$t0
+	 andi	$t0,$d3,0xFF00
+	 sll	$d3,$d3,24
+	 andi	$t1,0xFF00
+	 sll	$t0,$t0,8
+	 or	$d3,$at
+	 or	$t1,$t0
+	 or	$d3,$t1
+# endif
+#endif
+	srl	$t0,$h4,2		# modulo-scheduled reduction
+	andi	$h4,$h4,3
+	sll	$at,$t0,2
+
+	addu	$d0,$d0,$h0		# accumulate input
+	 addu	$t0,$t0,$at
+	sltu	$h0,$d0,$h0
+	addu	$d0,$d0,$t0		# ... and residue
+	sltu	$at,$d0,$t0
+
+	addu	$d1,$d1,$h1
+	 addu	$h0,$h0,$at		# carry
+	sltu	$h1,$d1,$h1
+	addu	$d1,$d1,$h0
+	sltu	$h0,$d1,$h0
+
+	addu	$d2,$d2,$h2
+	 addu	$h1,$h1,$h0		# carry
+	sltu	$h2,$d2,$h2
+	addu	$d2,$d2,$h1
+	sltu	$h1,$d2,$h1
+
+	addu	$d3,$d3,$h3
+	 addu	$h2,$h2,$h1		# carry
+	sltu	$h3,$d3,$h3
+	addu	$d3,$d3,$h2
+
+#if defined(_MIPS_ARCH_MIPS32R2) && !defined(_MIPS_ARCH_MIPS32R6)
+	multu	$r0,$d0			# d0*r0
+	 sltu	$h2,$d3,$h2
+	maddu	$rs3,$d1		# d1*s3
+	 addu	$h3,$h3,$h2		# carry
+	maddu	$rs2,$d2		# d2*s2
+	 addu	$h4,$h4,$padbit
+	maddu	$rs1,$d3		# d3*s1
+	 addu	$h4,$h4,$h3
+	mfhi	$at
+	mflo	$h0
+
+	multu	$r1,$d0			# d0*r1
+	maddu	$r0,$d1			# d1*r0
+	maddu	$rs3,$d2		# d2*s3
+	maddu	$rs2,$d3		# d3*s2
+	maddu	$rs1,$h4		# h4*s1
+	maddu	$at,$one		# hi*1
+	mfhi	$at
+	mflo	$h1
+
+	multu	$r2,$d0			# d0*r2
+	maddu	$r1,$d1			# d1*r1
+	maddu	$r0,$d2			# d2*r0
+	maddu	$rs3,$d3		# d3*s3
+	maddu	$rs2,$h4		# h4*s2
+	maddu	$at,$one		# hi*1
+	mfhi	$at
+	mflo	$h2
+
+	mul	$t0,$r0,$h4		# h4*r0
+
+	multu	$r3,$d0			# d0*r3
+	maddu	$r2,$d1			# d1*r2
+	maddu	$r1,$d2			# d2*r1
+	maddu	$r0,$d3			# d3*r0
+	maddu	$rs3,$h4		# h4*s3
+	maddu	$at,$one		# hi*1
+	mfhi	$at
+	mflo	$h3
+
+	 addiu	$inp,$inp,16
+
+	addu	$h4,$t0,$at
+#else
+	multu	($r0,$d0)		# d0*r0
+	mflo	($h0,$r0,$d0)
+	mfhi	($h1,$r0,$d0)
+
+	 sltu	$h2,$d3,$h2
+	 addu	$h3,$h3,$h2		# carry
+
+	multu	($rs3,$d1)		# d1*s3
+	mflo	($at,$rs3,$d1)
+	mfhi	($t0,$rs3,$d1)
+
+	 addu	$h4,$h4,$padbit
+	 addiu	$inp,$inp,16
+	 addu	$h4,$h4,$h3
+
+	multu	($rs2,$d2)		# d2*s2
+	mflo	($a3,$rs2,$d2)
+	mfhi	($t1,$rs2,$d2)
+	 addu	$h0,$h0,$at
+	 addu	$h1,$h1,$t0
+	multu	($rs1,$d3)		# d3*s1
+	 sltu	$at,$h0,$at
+	 addu	$h1,$h1,$at
+
+	mflo	($at,$rs1,$d3)
+	mfhi	($t0,$rs1,$d3)
+	 addu	$h0,$h0,$a3
+	 addu	$h1,$h1,$t1
+	multu	($r1,$d0)		# d0*r1
+	 sltu	$a3,$h0,$a3
+	 addu	$h1,$h1,$a3
+
+
+	mflo	($a3,$r1,$d0)
+	mfhi	($h2,$r1,$d0)
+	 addu	$h0,$h0,$at
+	 addu	$h1,$h1,$t0
+	multu	($r0,$d1)		# d1*r0
+	 sltu	$at,$h0,$at
+	 addu	$h1,$h1,$at
+
+	mflo	($at,$r0,$d1)
+	mfhi	($t0,$r0,$d1)
+	 addu	$h1,$h1,$a3
+	 sltu	$a3,$h1,$a3
+	multu	($rs3,$d2)		# d2*s3
+	 addu	$h2,$h2,$a3
+
+	mflo	($a3,$rs3,$d2)
+	mfhi	($t1,$rs3,$d2)
+	 addu	$h1,$h1,$at
+	 addu	$h2,$h2,$t0
+	multu	($rs2,$d3)		# d3*s2
+	 sltu	$at,$h1,$at
+	 addu	$h2,$h2,$at
+
+	mflo	($at,$rs2,$d3)
+	mfhi	($t0,$rs2,$d3)
+	 addu	$h1,$h1,$a3
+	 addu	$h2,$h2,$t1
+	multu	($rs1,$h4)		# h4*s1
+	 sltu	$a3,$h1,$a3
+	 addu	$h2,$h2,$a3
+
+	mflo	($a3,$rs1,$h4)
+	 addu	$h1,$h1,$at
+	 addu	$h2,$h2,$t0
+	multu	($r2,$d0)		# d0*r2
+	 sltu	$at,$h1,$at
+	 addu	$h2,$h2,$at
+
+
+	mflo	($at,$r2,$d0)
+	mfhi	($h3,$r2,$d0)
+	 addu	$h1,$h1,$a3
+	 sltu	$a3,$h1,$a3
+	multu	($r1,$d1)		# d1*r1
+	 addu	$h2,$h2,$a3
+
+	mflo	($a3,$r1,$d1)
+	mfhi	($t1,$r1,$d1)
+	 addu	$h2,$h2,$at
+	 sltu	$at,$h2,$at
+	multu	($r0,$d2)		# d2*r0
+	 addu	$h3,$h3,$at
+
+	mflo	($at,$r0,$d2)
+	mfhi	($t0,$r0,$d2)
+	 addu	$h2,$h2,$a3
+	 addu	$h3,$h3,$t1
+	multu	($rs3,$d3)		# d3*s3
+	 sltu	$a3,$h2,$a3
+	 addu	$h3,$h3,$a3
+
+	mflo	($a3,$rs3,$d3)
+	mfhi	($t1,$rs3,$d3)
+	 addu	$h2,$h2,$at
+	 addu	$h3,$h3,$t0
+	multu	($rs2,$h4)		# h4*s2
+	 sltu	$at,$h2,$at
+	 addu	$h3,$h3,$at
+
+	mflo	($at,$rs2,$h4)
+	 addu	$h2,$h2,$a3
+	 addu	$h3,$h3,$t1
+	multu	($r3,$d0)		# d0*r3
+	 sltu	$a3,$h2,$a3
+	 addu	$h3,$h3,$a3
+
+
+	mflo	($a3,$r3,$d0)
+	mfhi	($t1,$r3,$d0)
+	 addu	$h2,$h2,$at
+	 sltu	$at,$h2,$at
+	multu	($r2,$d1)		# d1*r2
+	 addu	$h3,$h3,$at
+
+	mflo	($at,$r2,$d1)
+	mfhi	($t0,$r2,$d1)
+	 addu	$h3,$h3,$a3
+	 sltu	$a3,$h3,$a3
+	multu	($r0,$d3)		# d3*r0
+	 addu	$t1,$t1,$a3
+
+	mflo	($a3,$r0,$d3)
+	mfhi	($d3,$r0,$d3)
+	 addu	$h3,$h3,$at
+	 addu	$t1,$t1,$t0
+	multu	($r1,$d2)		# d2*r1
+	 sltu	$at,$h3,$at
+	 addu	$t1,$t1,$at
+
+	mflo	($at,$r1,$d2)
+	mfhi	($t0,$r1,$d2)
+	 addu	$h3,$h3,$a3
+	 addu	$t1,$t1,$d3
+	multu	($rs3,$h4)		# h4*s3
+	 sltu	$a3,$h3,$a3
+	 addu	$t1,$t1,$a3
+
+	mflo	($a3,$rs3,$h4)
+	 addu	$h3,$h3,$at
+	 addu	$t1,$t1,$t0
+	multu	($r0,$h4)		# h4*r0
+	 sltu	$at,$h3,$at
+	 addu	$t1,$t1,$at
+
+
+	mflo	($h4,$r0,$h4)
+	 addu	$h3,$h3,$a3
+	 sltu	$a3,$h3,$a3
+	 addu	$t1,$t1,$a3
+	addu	$h4,$h4,$t1
+
+	li	$padbit,1		# if we loop, padbit is 1
+#endif
+	bne	$inp,$len,.Loop
+
+	sw	$h0,0($ctx)		# store hash value
+	sw	$h1,4($ctx)
+	sw	$h2,8($ctx)
+	sw	$h3,12($ctx)
+	sw	$h4,16($ctx)
+
+	.set	noreorder
+.Labort:
+	lw	$s11,4*11($sp)
+	lw	$s10,4*10($sp)
+	lw	$s9, 4*9($sp)
+	lw	$s8, 4*8($sp)
+	lw	$s7, 4*7($sp)
+	lw	$s6, 4*6($sp)
+	lw	$s5, 4*5($sp)
+	lw	$s4, 4*4($sp)
+___
+$code.=<<___ if ($flavour =~ /nubi/i);	# optimize non-nubi prologue
+	lw	$s3, 4*3($sp)
+	lw	$s2, 4*2($sp)
+	lw	$s1, 4*1($sp)
+	lw	$s0, 4*0($sp)
+___
+$code.=<<___;
+	jr	$ra
+	addu	$sp,$sp,4*12
+.end	poly1305_blocks
+___
+}
+{
+my ($ctx,$mac,$nonce,$tmp4) = ($a0,$a1,$a2,$a3);
+
+$code.=<<___;
+.align	5
+.globl	poly1305_emit
+.ent	poly1305_emit
+poly1305_emit:
+	.frame	$sp,0,$ra
+	.set	reorder
+
+	lw	$tmp4,16($ctx)
+	lw	$tmp0,0($ctx)
+	lw	$tmp1,4($ctx)
+	lw	$tmp2,8($ctx)
+	lw	$tmp3,12($ctx)
+
+	li	$in0,-4			# final reduction
+	srl	$ctx,$tmp4,2
+	and	$in0,$in0,$tmp4
+	andi	$tmp4,$tmp4,3
+	addu	$ctx,$ctx,$in0
+
+	addu	$tmp0,$tmp0,$ctx
+	sltu	$ctx,$tmp0,$ctx
+	 addiu	$in0,$tmp0,5		# compare to modulus
+	addu	$tmp1,$tmp1,$ctx
+	 sltiu	$in1,$in0,5
+	sltu	$ctx,$tmp1,$ctx
+	 addu	$in1,$in1,$tmp1
+	addu	$tmp2,$tmp2,$ctx
+	 sltu	$in2,$in1,$tmp1
+	sltu	$ctx,$tmp2,$ctx
+	 addu	$in2,$in2,$tmp2
+	addu	$tmp3,$tmp3,$ctx
+	 sltu	$in3,$in2,$tmp2
+	sltu	$ctx,$tmp3,$ctx
+	 addu	$in3,$in3,$tmp3
+	addu	$tmp4,$tmp4,$ctx
+	 sltu	$ctx,$in3,$tmp3
+	 addu	$ctx,$tmp4
+
+	srl	$ctx,2			# see if it carried/borrowed
+	subu	$ctx,$zero,$ctx
+
+	xor	$in0,$tmp0
+	xor	$in1,$tmp1
+	xor	$in2,$tmp2
+	xor	$in3,$tmp3
+	and	$in0,$ctx
+	and	$in1,$ctx
+	and	$in2,$ctx
+	and	$in3,$ctx
+	xor	$in0,$tmp0
+	xor	$in1,$tmp1
+	xor	$in2,$tmp2
+	xor	$in3,$tmp3
+
+	lw	$tmp0,0($nonce)		# load nonce
+	lw	$tmp1,4($nonce)
+	lw	$tmp2,8($nonce)
+	lw	$tmp3,12($nonce)
+
+	addu	$in0,$tmp0		# accumulate nonce
+	sltu	$ctx,$in0,$tmp0
+
+	addu	$in1,$tmp1
+	sltu	$tmp1,$in1,$tmp1
+	addu	$in1,$ctx
+	sltu	$ctx,$in1,$ctx
+	addu	$ctx,$tmp1
+
+	addu	$in2,$tmp2
+	sltu	$tmp2,$in2,$tmp2
+	addu	$in2,$ctx
+	sltu	$ctx,$in2,$ctx
+	addu	$ctx,$tmp2
+
+	addu	$in3,$tmp3
+	addu	$in3,$ctx
+
+	srl	$tmp0,$in0,8		# write mac value
+	srl	$tmp1,$in0,16
+	srl	$tmp2,$in0,24
+	sb	$in0, 0($mac)
+	sb	$tmp0,1($mac)
+	srl	$tmp0,$in1,8
+	sb	$tmp1,2($mac)
+	srl	$tmp1,$in1,16
+	sb	$tmp2,3($mac)
+	srl	$tmp2,$in1,24
+	sb	$in1, 4($mac)
+	sb	$tmp0,5($mac)
+	srl	$tmp0,$in2,8
+	sb	$tmp1,6($mac)
+	srl	$tmp1,$in2,16
+	sb	$tmp2,7($mac)
+	srl	$tmp2,$in2,24
+	sb	$in2, 8($mac)
+	sb	$tmp0,9($mac)
+	srl	$tmp0,$in3,8
+	sb	$tmp1,10($mac)
+	srl	$tmp1,$in3,16
+	sb	$tmp2,11($mac)
+	srl	$tmp2,$in3,24
+	sb	$in3, 12($mac)
+	sb	$tmp0,13($mac)
+	sb	$tmp1,14($mac)
+	sb	$tmp2,15($mac)
+
+	jr	$ra
+.end	poly1305_emit
+.rdata
+.asciiz	"Poly1305 for MIPS32, CRYPTOGAMS by \@dot-asm"
+.align	2
+___
+}
+}}}
+
+$output=pop and open STDOUT,">$output";
+print $code;
+close STDOUT;