feat: add string manipulation functions derived from musl libc

This commit introduces a variety of string manipulation functions, including:
- `memmem`: Searches for a substring in a memory area.
- `memmove`: Safely copies memory areas, handling overlaps.
- `mempcpy`: Copies memory and returns a pointer to the end of the destination.
- `memrchr`: Finds the last occurrence of a character in a memory area.
- `memset`: Fills a memory area with a specified value.
- `rindex`: Returns the last occurrence of a character in a string.
- `stpcpy` and `stpncpy`: Copies strings and returns a pointer to the end.
- `strcat`, `strchr`, `strcspn`, `strlcat`, `strlcpy`, `strlen`, `strncat`, `strncmp`, `strncpy`, `strnlen`, `strpbrk`, `strrchr`, `strstr`, `strtok`, and `strtok_r`: Various string manipulation functions for concatenation, searching, and tokenization.

Assembly implementations for `memcpy`, `memmove`, and `memset` are also included for optimized performance on x86_64 architectures.

56 files changed, 2624 insertions, 0 deletions

diff --git a/na/include/libc-shim/bits/alltypes.h b/na/include/libc-shim/bits/alltypes.h
new file mode 100644
index 0000000..cf21a5e
--- /dev/null
+++ b/na/include/libc-shim/bits/alltypes.h
@@ -0,0 +1,15 @@
+/* Shim: provide types musl string functions need */
+#ifndef _BITS_ALLTYPES_H
+#define _BITS_ALLTYPES_H
+
+#ifndef __cplusplus
+typedef unsigned long size_t;
+#endif
+
+typedef long ssize_t;
+typedef unsigned long uintptr_t;
+
+/* locale_t stub — unused in kernel, but declared by some headers */
+typedef void *locale_t;
+
+#endif
diff --git a/na/include/libc-shim/endian.h b/na/include/libc-shim/endian.h
new file mode 100644
index 0000000..8346386
--- /dev/null
+++ b/na/include/libc-shim/endian.h
@@ -0,0 +1,9 @@
+/* Shim: endianness for musl string functions */
+#ifndef _ENDIAN_H
+#define _ENDIAN_H
+
+#define __LITTLE_ENDIAN 1234
+#define __BIG_ENDIAN    4321
+#define __BYTE_ORDER    __LITTLE_ENDIAN  /* RISC-V is LE */
+
+#endif
diff --git a/na/include/libc-shim/features.h b/na/include/libc-shim/features.h
new file mode 100644
index 0000000..a3466c8
--- /dev/null
+++ b/na/include/libc-shim/features.h
@@ -0,0 +1,12 @@
+/* Shim: musl expects this header */
+#ifndef _FEATURES_H
+#define _FEATURES_H
+
+/* weak_alias: musl uses this for symbol aliasing (e.g. __memrchr -> memrchr) */
+#define weak_alias(old, new) \
+    extern __typeof(old) new __attribute__((__weak__, __alias__(#old)))
+
+/* hidden attribute used by some musl sources */
+#define hidden __attribute__((__visibility__("hidden")))
+
+#endif
diff --git a/na/include/libc-shim/limits.h b/na/include/libc-shim/limits.h
new file mode 100644
index 0000000..a1d546e
--- /dev/null
+++ b/na/include/libc-shim/limits.h
@@ -0,0 +1,17 @@
+/* Shim: limits for musl string functions (memmem uses SIZE_MAX) */
+#ifndef _LIMITS_H
+#define _LIMITS_H
+
+#define CHAR_BIT   8
+#define UCHAR_MAX  255
+#define SCHAR_MAX  127
+#define SCHAR_MIN  (-128)
+#define INT_MAX    2147483647
+#define INT_MIN    (-INT_MAX - 1)
+#define UINT_MAX   4294967295U
+#define LONG_MAX   9223372036854775807L
+#define LONG_MIN   (-LONG_MAX - 1L)
+#define ULONG_MAX  18446744073709551615UL
+#define SIZE_MAX   ULONG_MAX
+
+#endif
diff --git a/na/include/libc-shim/stdint.h b/na/include/libc-shim/stdint.h
new file mode 100644
index 0000000..650bdcf
--- /dev/null
+++ b/na/include/libc-shim/stdint.h
@@ -0,0 +1,21 @@
+/* Shim: stdint types for musl string functions */
+#ifndef _STDINT_H
+#define _STDINT_H
+
+typedef unsigned char      uint8_t;
+typedef unsigned short     uint16_t;
+typedef unsigned int       uint32_t;
+typedef unsigned long      uint64_t;
+typedef signed char        int8_t;
+typedef signed short       int16_t;
+typedef signed int         int32_t;
+typedef signed long        int64_t;
+typedef unsigned long      uintptr_t;
+typedef long               intptr_t;
+
+#define UINT8_MAX  0xFF
+#define UINT16_MAX 0xFFFF
+#define UINT32_MAX 0xFFFFFFFFU
+#define UINT64_MAX 0xFFFFFFFFFFFFFFFFUL
+
+#endif
diff --git a/na/include/libc-shim/string.h b/na/include/libc-shim/string.h
new file mode 100644
index 0000000..8a86304
--- /dev/null
+++ b/na/include/libc-shim/string.h
@@ -0,0 +1,110 @@
+#ifndef	_STRING_H
+#define	_STRING_H
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+#include <features.h>
+
+#if __cplusplus >= 201103L
+#define NULL nullptr
+#elif defined(__cplusplus)
+#define NULL 0L
+#else
+#define NULL ((void*)0)
+#endif
+
+#define __NEED_size_t
+#if defined(_POSIX_SOURCE) || defined(_POSIX_C_SOURCE) \
+ || defined(_XOPEN_SOURCE) || defined(_GNU_SOURCE) \
+ || defined(_BSD_SOURCE)
+#define __NEED_locale_t
+#endif
+
+#include <bits/alltypes.h>
+
+void *memcpy (void *__restrict, const void *__restrict, size_t);
+void *memmove (void *, const void *, size_t);
+void *memset (void *, int, size_t);
+int memcmp (const void *, const void *, size_t);
+void *memchr (const void *, int, size_t);
+
+char *strcpy (char *__restrict, const char *__restrict);
+char *strncpy (char *__restrict, const char *__restrict, size_t);
+
+char *strcat (char *__restrict, const char *__restrict);
+char *strncat (char *__restrict, const char *__restrict, size_t);
+
+int strcmp (const char *, const char *);
+int strncmp (const char *, const char *, size_t);
+
+int strcoll (const char *, const char *);
+size_t strxfrm (char *__restrict, const char *__restrict, size_t);
+
+char *strchr (const char *, int);
+char *strrchr (const char *, int);
+
+size_t strcspn (const char *, const char *);
+size_t strspn (const char *, const char *);
+char *strpbrk (const char *, const char *);
+char *strstr (const char *, const char *);
+char *strtok (char *__restrict, const char *__restrict);
+
+size_t strlen (const char *);
+
+char *strerror (int);
+
+#if defined(_BSD_SOURCE) || defined(_GNU_SOURCE)
+#include <strings.h>
+#endif
+
+#if defined(_POSIX_SOURCE) || defined(_POSIX_C_SOURCE) \
+ || defined(_XOPEN_SOURCE) || defined(_GNU_SOURCE) \
+ || defined(_BSD_SOURCE)
+char *strtok_r (char *__restrict, const char *__restrict, char **__restrict);
+int strerror_r (int, char *, size_t);
+char *stpcpy(char *__restrict, const char *__restrict);
+char *stpncpy(char *__restrict, const char *__restrict, size_t);
+size_t strnlen (const char *, size_t);
+char *strdup (const char *);
+char *strndup (const char *, size_t);
+char *strsignal(int);
+char *strerror_l (int, locale_t);
+int strcoll_l (const char *, const char *, locale_t);
+size_t strxfrm_l (char *__restrict, const char *__restrict, size_t, locale_t);
+void *memmem(const void *, size_t, const void *, size_t);
+#endif
+
+#if defined(_XOPEN_SOURCE) || defined(_GNU_SOURCE) \
+ || defined(_BSD_SOURCE)
+void *memccpy (void *__restrict, const void *__restrict, int, size_t);
+#endif
+
+#if defined(_GNU_SOURCE) || defined(_BSD_SOURCE)
+char *strsep(char **, const char *);
+size_t strlcat (char *, const char *, size_t);
+size_t strlcpy (char *, const char *, size_t);
+void explicit_bzero (void *, size_t);
+#endif
+
+#ifdef _GNU_SOURCE
+#define	strdupa(x)	strcpy(alloca(strlen(x)+1),x)
+int strverscmp (const char *, const char *);
+char *strchrnul(const char *, int);
+char *strcasestr(const char *, const char *);
+void *memrchr(const void *, int, size_t);
+void *mempcpy(void *, const void *, size_t);
+#endif
+
+/* musl internal symbols used across translation units */
+char *__strchrnul(const char *, int);
+void *__memrchr(const void *, int, size_t);
+char *__stpcpy(char *__restrict, const char *__restrict);
+char *__stpncpy(char *__restrict, const char *__restrict, size_t);
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif
diff --git a/na/include/libc-shim/strings.h b/na/include/libc-shim/strings.h
new file mode 100644
index 0000000..b7a5ea0
--- /dev/null
+++ b/na/include/libc-shim/strings.h
@@ -0,0 +1,40 @@
+#ifndef	_STRINGS_H
+#define	_STRINGS_H
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+#include <features.h>
+
+#define __NEED_size_t
+#define __NEED_locale_t
+#include <bits/alltypes.h>
+
+#if defined(_GNU_SOURCE) || defined(_BSD_SOURCE) || defined(_POSIX_SOURCE) \
+ || (defined(_POSIX_C_SOURCE) && _POSIX_C_SOURCE+0 < 200809L) \
+ || (defined(_XOPEN_SOURCE) && _XOPEN_SOURCE+0 < 700)
+int bcmp (const void *, const void *, size_t);
+void bcopy (const void *, void *, size_t);
+void bzero (void *, size_t);
+char *index (const char *, int);
+char *rindex (const char *, int);
+#endif
+
+#if defined(_XOPEN_SOURCE) || defined(_GNU_SOURCE)  || defined(_BSD_SOURCE)
+int ffs (int);
+int ffsl (long);
+int ffsll (long long);
+#endif
+
+int strcasecmp (const char *, const char *);
+int strncasecmp (const char *, const char *, size_t);
+
+int strcasecmp_l (const char *, const char *, locale_t);
+int strncasecmp_l (const char *, const char *, size_t, locale_t);
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif
diff --git a/na/lib/string/COPYRIGHT b/na/lib/string/COPYRIGHT
new file mode 100644
index 0000000..2f15edc
--- /dev/null
+++ b/na/lib/string/COPYRIGHT
@@ -0,0 +1,193 @@
+musl as a whole is licensed under the following standard MIT license:
+
+----------------------------------------------------------------------
+Copyright © 2005-2020 Rich Felker, et al.
+
+Permission is hereby granted, free of charge, to any person obtaining
+a copy of this software and associated documentation files (the
+"Software"), to deal in the Software without restriction, including
+without limitation the rights to use, copy, modify, merge, publish,
+distribute, sublicense, and/or sell copies of the Software, and to
+permit persons to whom the Software is furnished to do so, subject to
+the following conditions:
+
+The above copyright notice and this permission notice shall be
+included in all copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
+IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
+CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
+TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
+SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
+----------------------------------------------------------------------
+
+Authors/contributors include:
+
+A. Wilcox
+Ada Worcester
+Alex Dowad
+Alex Suykov
+Alexander Monakov
+Andre McCurdy
+Andrew Kelley
+Anthony G. Basile
+Aric Belsito
+Arvid Picciani
+Bartosz Brachaczek
+Benjamin Peterson
+Bobby Bingham
+Boris Brezillon
+Brent Cook
+Chris Spiegel
+Clément Vasseur
+Daniel Micay
+Daniel Sabogal
+Daurnimator
+David Carlier
+David Edelsohn
+Denys Vlasenko
+Dmitry Ivanov
+Dmitry V. Levin
+Drew DeVault
+Emil Renner Berthing
+Fangrui Song
+Felix Fietkau
+Felix Janda
+Gianluca Anzolin
+Hauke Mehrtens
+He X
+Hiltjo Posthuma
+Isaac Dunham
+Jaydeep Patil
+Jens Gustedt
+Jeremy Huntwork
+Jo-Philipp Wich
+Joakim Sindholt
+John Spencer
+Julien Ramseier
+Justin Cormack
+Kaarle Ritvanen
+Khem Raj
+Kylie McClain
+Leah Neukirchen
+Luca Barbato
+Luka Perkov
+Lynn Ochs
+M Farkas-Dyck (Strake)
+Mahesh Bodapati
+Markus Wichmann
+Masanori Ogino
+Michael Clark
+Michael Forney
+Mikhail Kremnyov
+Natanael Copa
+Nicholas J. Kain
+orc
+Pascal Cuoq
+Patrick Oppenlander
+Petr Hosek
+Petr Skocik
+Pierre Carrier
+Reini Urban
+Rich Felker
+Richard Pennington
+Ryan Fairfax
+Samuel Holland
+Segev Finer
+Shiz
+sin
+Solar Designer
+Stefan Kristiansson
+Stefan O'Rear
+Szabolcs Nagy
+Timo Teräs
+Trutz Behn
+Will Dietz
+William Haddon
+William Pitcock
+
+Portions of this software are derived from third-party works licensed
+under terms compatible with the above MIT license:
+
+The TRE regular expression implementation (src/regex/reg* and
+src/regex/tre*) is Copyright © 2001-2008 Ville Laurikari and licensed
+under a 2-clause BSD license (license text in the source files). The
+included version has been heavily modified by Rich Felker in 2012, in
+the interests of size, simplicity, and namespace cleanliness.
+
+Much of the math library code (src/math/* and src/complex/*) is
+Copyright © 1993,2004 Sun Microsystems or
+Copyright © 2003-2011 David Schultz or
+Copyright © 2003-2009 Steven G. Kargl or
+Copyright © 2003-2009 Bruce D. Evans or
+Copyright © 2008 Stephen L. Moshier or
+Copyright © 2017-2018 Arm Limited
+and labelled as such in comments in the individual source files. All
+have been licensed under extremely permissive terms.
+
+The ARM memcpy code (src/string/arm/memcpy.S) is Copyright © 2008
+The Android Open Source Project and is licensed under a two-clause BSD
+license. It was taken from Bionic libc, used on Android.
+
+The AArch64 memcpy and memset code (src/string/aarch64/*) are
+Copyright © 1999-2019, Arm Limited.
+
+The implementation of DES for crypt (src/crypt/crypt_des.c) is
+Copyright © 1994 David Burren. It is licensed under a BSD license.
+
+The implementation of blowfish crypt (src/crypt/crypt_blowfish.c) was
+originally written by Solar Designer and placed into the public
+domain. The code also comes with a fallback permissive license for use
+in jurisdictions that may not recognize the public domain.
+
+The smoothsort implementation (src/stdlib/qsort.c) is Copyright © 2011
+Lynn Ochs and is licensed under an MIT-style license.
+
+The x86_64 port was written by Nicholas J. Kain and is licensed under
+the standard MIT terms.
+
+The mips and microblaze ports were originally written by Richard
+Pennington for use in the ellcc project. The original code was adapted
+by Rich Felker for build system and code conventions during upstream
+integration. It is licensed under the standard MIT terms.
+
+The mips64 port was contributed by Imagination Technologies and is
+licensed under the standard MIT terms.
+
+The powerpc port was also originally written by Richard Pennington,
+and later supplemented and integrated by John Spencer. It is licensed
+under the standard MIT terms.
+
+All other files which have no copyright comments are original works
+produced specifically for use as part of this library, written either
+by Rich Felker, the main author of the library, or by one or more
+contibutors listed above. Details on authorship of individual files
+can be found in the git version control history of the project. The
+omission of copyright and license comments in each file is in the
+interest of source tree size.
+
+In addition, permission is hereby granted for all public header files
+(include/* and arch/*/bits/*) and crt files intended to be linked into
+applications (crt/*, ldso/dlstart.c, and arch/*/crt_arch.h) to omit
+the copyright notice and permission notice otherwise required by the
+license, and to use these files without any requirement of
+attribution. These files include substantial contributions from:
+
+Bobby Bingham
+John Spencer
+Nicholas J. Kain
+Rich Felker
+Richard Pennington
+Stefan Kristiansson
+Szabolcs Nagy
+
+all of whom have explicitly granted such permission.
+
+This file previously contained text expressing a belief that most of
+the files covered by the above exception were sufficiently trivial not
+to be subject to copyright, resulting in confusion over whether it
+negated the permissions granted in the license. In the spirit of
+permissive licensing, and of not having licensing issues being an
+obstacle to adoption, that text has been removed.
diff --git a/na/lib/string/aarch64/memcpy.S b/na/lib/string/aarch64/memcpy.S
new file mode 100644
index 0000000..48bb8a8
--- /dev/null
+++ b/na/lib/string/aarch64/memcpy.S
@@ -0,0 +1,186 @@
+/*
+ * memcpy - copy memory area
+ *
+ * Copyright (c) 2012-2020, Arm Limited.
+ * SPDX-License-Identifier: MIT
+ */
+
+/* Assumptions:
+ *
+ * ARMv8-a, AArch64, unaligned accesses.
+ *
+ */
+
+#define dstin   x0
+#define src     x1
+#define count   x2
+#define dst     x3
+#define srcend  x4
+#define dstend  x5
+#define A_l     x6
+#define A_lw    w6
+#define A_h     x7
+#define B_l     x8
+#define B_lw    w8
+#define B_h     x9
+#define C_l     x10
+#define C_lw    w10
+#define C_h     x11
+#define D_l     x12
+#define D_h     x13
+#define E_l     x14
+#define E_h     x15
+#define F_l     x16
+#define F_h     x17
+#define G_l     count
+#define G_h     dst
+#define H_l     src
+#define H_h     srcend
+#define tmp1    x14
+
+/* This implementation of memcpy uses unaligned accesses and branchless
+   sequences to keep the code small, simple and improve performance.
+
+   Copies are split into 3 main cases: small copies of up to 32 bytes, medium
+   copies of up to 128 bytes, and large copies.  The overhead of the overlap
+   check is negligible since it is only required for large copies.
+
+   Large copies use a software pipelined loop processing 64 bytes per iteration.
+   The destination pointer is 16-byte aligned to minimize unaligned accesses.
+   The loop tail is handled by always copying 64 bytes from the end.
+*/
+
+.global memcpy
+.type memcpy,%function
+memcpy:
+	add     srcend, src, count
+	add     dstend, dstin, count
+	cmp     count, 128
+	b.hi    .Lcopy_long
+	cmp     count, 32
+	b.hi    .Lcopy32_128
+
+	/* Small copies: 0..32 bytes.  */
+	cmp     count, 16
+	b.lo    .Lcopy16
+	ldp     A_l, A_h, [src]
+	ldp     D_l, D_h, [srcend, -16]
+	stp     A_l, A_h, [dstin]
+	stp     D_l, D_h, [dstend, -16]
+	ret
+
+	/* Copy 8-15 bytes.  */
+.Lcopy16:
+	tbz     count, 3, .Lcopy8
+	ldr     A_l, [src]
+	ldr     A_h, [srcend, -8]
+	str     A_l, [dstin]
+	str     A_h, [dstend, -8]
+	ret
+
+	.p2align 3
+	/* Copy 4-7 bytes.  */
+.Lcopy8:
+	tbz     count, 2, .Lcopy4
+	ldr     A_lw, [src]
+	ldr     B_lw, [srcend, -4]
+	str     A_lw, [dstin]
+	str     B_lw, [dstend, -4]
+	ret
+
+	/* Copy 0..3 bytes using a branchless sequence.  */
+.Lcopy4:
+	cbz     count, .Lcopy0
+	lsr     tmp1, count, 1
+	ldrb    A_lw, [src]
+	ldrb    C_lw, [srcend, -1]
+	ldrb    B_lw, [src, tmp1]
+	strb    A_lw, [dstin]
+	strb    B_lw, [dstin, tmp1]
+	strb    C_lw, [dstend, -1]
+.Lcopy0:
+	ret
+
+	.p2align 4
+	/* Medium copies: 33..128 bytes.  */
+.Lcopy32_128:
+	ldp     A_l, A_h, [src]
+	ldp     B_l, B_h, [src, 16]
+	ldp     C_l, C_h, [srcend, -32]
+	ldp     D_l, D_h, [srcend, -16]
+	cmp     count, 64
+	b.hi    .Lcopy128
+	stp     A_l, A_h, [dstin]
+	stp     B_l, B_h, [dstin, 16]
+	stp     C_l, C_h, [dstend, -32]
+	stp     D_l, D_h, [dstend, -16]
+	ret
+
+	.p2align 4
+	/* Copy 65..128 bytes.  */
+.Lcopy128:
+	ldp     E_l, E_h, [src, 32]
+	ldp     F_l, F_h, [src, 48]
+	cmp     count, 96
+	b.ls    .Lcopy96
+	ldp     G_l, G_h, [srcend, -64]
+	ldp     H_l, H_h, [srcend, -48]
+	stp     G_l, G_h, [dstend, -64]
+	stp     H_l, H_h, [dstend, -48]
+.Lcopy96:
+	stp     A_l, A_h, [dstin]
+	stp     B_l, B_h, [dstin, 16]
+	stp     E_l, E_h, [dstin, 32]
+	stp     F_l, F_h, [dstin, 48]
+	stp     C_l, C_h, [dstend, -32]
+	stp     D_l, D_h, [dstend, -16]
+	ret
+
+	.p2align 4
+	/* Copy more than 128 bytes.  */
+.Lcopy_long:
+
+	/* Copy 16 bytes and then align dst to 16-byte alignment.  */
+
+	ldp     D_l, D_h, [src]
+	and     tmp1, dstin, 15
+	bic     dst, dstin, 15
+	sub     src, src, tmp1
+	add     count, count, tmp1      /* Count is now 16 too large.  */
+	ldp     A_l, A_h, [src, 16]
+	stp     D_l, D_h, [dstin]
+	ldp     B_l, B_h, [src, 32]
+	ldp     C_l, C_h, [src, 48]
+	ldp     D_l, D_h, [src, 64]!
+	subs    count, count, 128 + 16  /* Test and readjust count.  */
+	b.ls    .Lcopy64_from_end
+
+.Lloop64:
+	stp     A_l, A_h, [dst, 16]
+	ldp     A_l, A_h, [src, 16]
+	stp     B_l, B_h, [dst, 32]
+	ldp     B_l, B_h, [src, 32]
+	stp     C_l, C_h, [dst, 48]
+	ldp     C_l, C_h, [src, 48]
+	stp     D_l, D_h, [dst, 64]!
+	ldp     D_l, D_h, [src, 64]!
+	subs    count, count, 64
+	b.hi    .Lloop64
+
+	/* Write the last iteration and copy 64 bytes from the end.  */
+.Lcopy64_from_end:
+	ldp     E_l, E_h, [srcend, -64]
+	stp     A_l, A_h, [dst, 16]
+	ldp     A_l, A_h, [srcend, -48]
+	stp     B_l, B_h, [dst, 32]
+	ldp     B_l, B_h, [srcend, -32]
+	stp     C_l, C_h, [dst, 48]
+	ldp     C_l, C_h, [srcend, -16]
+	stp     D_l, D_h, [dst, 64]
+	stp     E_l, E_h, [dstend, -64]
+	stp     A_l, A_h, [dstend, -48]
+	stp     B_l, B_h, [dstend, -32]
+	stp     C_l, C_h, [dstend, -16]
+	ret
+
+.size memcpy,.-memcpy
diff --git a/na/lib/string/aarch64/memset.S b/na/lib/string/aarch64/memset.S
new file mode 100644
index 0000000..f0d29b7
--- /dev/null
+++ b/na/lib/string/aarch64/memset.S
@@ -0,0 +1,115 @@
+/*
+ * memset - fill memory with a constant byte
+ *
+ * Copyright (c) 2012-2020, Arm Limited.
+ * SPDX-License-Identifier: MIT
+ */
+
+/* Assumptions:
+ *
+ * ARMv8-a, AArch64, Advanced SIMD, unaligned accesses.
+ *
+ */
+
+#define dstin   x0
+#define val     x1
+#define valw    w1
+#define count   x2
+#define dst     x3
+#define dstend  x4
+#define zva_val x5
+
+.global memset
+.type memset,%function
+memset:
+
+	dup     v0.16B, valw
+	add     dstend, dstin, count
+
+	cmp     count, 96
+	b.hi    .Lset_long
+	cmp     count, 16
+	b.hs    .Lset_medium
+	mov     val, v0.D[0]
+
+	/* Set 0..15 bytes.  */
+	tbz     count, 3, 1f
+	str     val, [dstin]
+	str     val, [dstend, -8]
+	ret
+	nop
+1:      tbz     count, 2, 2f
+	str     valw, [dstin]
+	str     valw, [dstend, -4]
+	ret
+2:      cbz     count, 3f
+	strb    valw, [dstin]
+	tbz     count, 1, 3f
+	strh    valw, [dstend, -2]
+3:      ret
+
+	/* Set 17..96 bytes.  */
+.Lset_medium:
+	str     q0, [dstin]
+	tbnz    count, 6, .Lset96
+	str     q0, [dstend, -16]
+	tbz     count, 5, 1f
+	str     q0, [dstin, 16]
+	str     q0, [dstend, -32]
+1:      ret
+
+	.p2align 4
+	/* Set 64..96 bytes.  Write 64 bytes from the start and
+	   32 bytes from the end.  */
+.Lset96:
+	str     q0, [dstin, 16]
+	stp     q0, q0, [dstin, 32]
+	stp     q0, q0, [dstend, -32]
+	ret
+
+	.p2align 4
+.Lset_long:
+	and     valw, valw, 255
+	bic     dst, dstin, 15
+	str     q0, [dstin]
+	cmp     count, 160
+	ccmp    valw, 0, 0, hs
+	b.ne    .Lno_zva
+
+#ifndef SKIP_ZVA_CHECK
+	mrs     zva_val, dczid_el0
+	and     zva_val, zva_val, 31
+	cmp     zva_val, 4              /* ZVA size is 64 bytes.  */
+	b.ne    .Lno_zva
+#endif
+	str     q0, [dst, 16]
+	stp     q0, q0, [dst, 32]
+	bic     dst, dst, 63
+	sub     count, dstend, dst      /* Count is now 64 too large.  */
+	sub     count, count, 128       /* Adjust count and bias for loop.  */
+
+	.p2align 4
+.Lzva_loop:
+	add     dst, dst, 64
+	dc      zva, dst
+	subs    count, count, 64
+	b.hi    .Lzva_loop
+	stp     q0, q0, [dstend, -64]
+	stp     q0, q0, [dstend, -32]
+	ret
+
+.Lno_zva:
+	sub     count, dstend, dst      /* Count is 16 too large.  */
+	sub     dst, dst, 16            /* Dst is biased by -32.  */
+	sub     count, count, 64 + 16   /* Adjust count and bias for loop.  */
+.Lno_zva_loop:
+	stp     q0, q0, [dst, 32]
+	stp     q0, q0, [dst, 64]!
+	subs    count, count, 64
+	b.hi    .Lno_zva_loop
+	stp     q0, q0, [dstend, -64]
+	stp     q0, q0, [dstend, -32]
+	ret
+
+.size memset,.-memset
+
diff --git a/na/lib/string/arm/__aeabi_memcpy.s b/na/lib/string/arm/__aeabi_memcpy.s
new file mode 100644
index 0000000..3a527e4
--- /dev/null
+++ b/na/lib/string/arm/__aeabi_memcpy.s
@@ -0,0 +1,45 @@
+.syntax unified
+
+.global __aeabi_memcpy8
+.global __aeabi_memcpy4
+.global __aeabi_memcpy
+.global __aeabi_memmove8
+.global __aeabi_memmove4
+.global __aeabi_memmove
+
+.type __aeabi_memcpy8,%function
+.type __aeabi_memcpy4,%function
+.type __aeabi_memcpy,%function
+.type __aeabi_memmove8,%function
+.type __aeabi_memmove4,%function
+.type __aeabi_memmove,%function
+
+__aeabi_memmove8:
+__aeabi_memmove4:
+__aeabi_memmove:
+	cmp   r0, r1
+	bls   3f
+	cmp   r2, #0
+	beq   2f
+	adds  r0, r0, r2
+	adds  r2, r1, r2
+1:	subs  r2, r2, #1
+	ldrb  r3, [r2]
+	subs  r0, r0, #1
+	strb  r3, [r0]
+	cmp   r1, r2
+	bne   1b
+2:	bx    lr
+__aeabi_memcpy8:
+__aeabi_memcpy4:
+__aeabi_memcpy:
+3:	cmp   r2, #0
+	beq   2f
+	adds  r2, r1, r2
+1:	ldrb  r3, [r1]
+	adds  r1, r1, #1
+	strb  r3, [r0]
+	adds  r0, r0, #1
+	cmp   r1, r2
+	bne   1b
+2:	bx    lr
diff --git a/na/lib/string/arm/__aeabi_memset.s b/na/lib/string/arm/__aeabi_memset.s
new file mode 100644
index 0000000..f9f6058
--- /dev/null
+++ b/na/lib/string/arm/__aeabi_memset.s
@@ -0,0 +1,31 @@
+.syntax unified
+
+.global __aeabi_memclr8
+.global __aeabi_memclr4
+.global __aeabi_memclr
+.global __aeabi_memset8
+.global __aeabi_memset4
+.global __aeabi_memset
+
+.type __aeabi_memclr8,%function
+.type __aeabi_memclr4,%function
+.type __aeabi_memclr,%function
+.type __aeabi_memset8,%function
+.type __aeabi_memset4,%function
+.type __aeabi_memset,%function
+
+__aeabi_memclr8:
+__aeabi_memclr4:
+__aeabi_memclr:
+	movs  r2, #0
+__aeabi_memset8:
+__aeabi_memset4:
+__aeabi_memset:
+	cmp   r1, #0
+	beq   2f
+	adds  r1, r0, r1
+1:	strb  r2, [r0]
+	adds  r0, r0, #1
+	cmp   r1, r0
+	bne   1b
+2:	bx    lr
diff --git a/na/lib/string/arm/memcpy.S b/na/lib/string/arm/memcpy.S
new file mode 100644
index 0000000..869e344
--- /dev/null
+++ b/na/lib/string/arm/memcpy.S
@@ -0,0 +1,479 @@
+/*
+ * Copyright (C) 2008 The Android Open Source Project
+ * All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *  * Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ *  * Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in
+ *    the documentation and/or other materials provided with the
+ *    distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
+ * FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
+ * COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
+ * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
+ * BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS
+ * OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED
+ * AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
+ * OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT
+ * OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ */
+
+
+/*
+ * Optimized memcpy() for ARM.
+ *
+ * note that memcpy() always returns the destination pointer,
+ * so we have to preserve R0.
+  */
+
+/*
+ * This file has been modified from the original for use in musl libc.
+ * The main changes are: addition of .type memcpy,%function to make the
+ * code safely callable from thumb mode, adjusting the return
+ * instructions to be compatible with pre-thumb ARM cpus, removal of
+ * prefetch code that is not compatible with older cpus and support for
+ * building as thumb 2 and big-endian.
+ */
+
+.syntax unified
+
+.global memcpy
+.type memcpy,%function
+memcpy:
+	/* The stack must always be 64-bits aligned to be compliant with the
+	 * ARM ABI. Since we have to save R0, we might as well save R4
+	 * which we can use for better pipelining of the reads below
+	 */
+	.fnstart
+	.save       {r0, r4, lr}
+	stmfd       sp!, {r0, r4, lr}
+	/* Making room for r5-r11 which will be spilled later */
+	.pad        #28
+	sub         sp, sp, #28
+
+	/* it simplifies things to take care of len<4 early */
+	cmp     r2, #4
+	blo     copy_last_3_and_return
+
+	/* compute the offset to align the source
+	 * offset = (4-(src&3))&3 = -src & 3
+	 */
+	rsb     r3, r1, #0
+	ands    r3, r3, #3
+	beq     src_aligned
+
+	/* align source to 32 bits. We need to insert 2 instructions between
+	 * a ldr[b|h] and str[b|h] because byte and half-word instructions
+	 * stall 2 cycles.
+	 */
+	movs    r12, r3, lsl #31
+	sub     r2, r2, r3              /* we know that r3 <= r2 because r2 >= 4 */
+	ldrbmi r3, [r1], #1
+	ldrbcs r4, [r1], #1
+	ldrbcs r12,[r1], #1
+	strbmi r3, [r0], #1
+	strbcs r4, [r0], #1
+	strbcs r12,[r0], #1
+
+src_aligned:
+
+	/* see if src and dst are aligned together (congruent) */
+	eor     r12, r0, r1
+	tst     r12, #3
+	bne     non_congruent
+
+	/* Use post-incriment mode for stm to spill r5-r11 to reserved stack
+	 * frame. Don't update sp.
+	 */
+	stmea   sp, {r5-r11}
+
+	/* align the destination to a cache-line */
+	rsb     r3, r0, #0
+	ands    r3, r3, #0x1C
+	beq     congruent_aligned32
+	cmp     r3, r2
+	andhi   r3, r2, #0x1C
+
+	/* conditionnaly copies 0 to 7 words (length in r3) */
+	movs    r12, r3, lsl #28
+	ldmcs   r1!, {r4, r5, r6, r7}           /* 16 bytes */
+	ldmmi   r1!, {r8, r9}                   /*  8 bytes */
+	stmcs   r0!, {r4, r5, r6, r7}
+	stmmi   r0!, {r8, r9}
+	tst     r3, #0x4
+	ldrne   r10,[r1], #4                    /*  4 bytes */
+	strne   r10,[r0], #4
+	sub     r2, r2, r3
+
+congruent_aligned32:
+	/*
+	 * here source is aligned to 32 bytes.
+	 */
+
+cached_aligned32:
+	subs    r2, r2, #32
+	blo     less_than_32_left
+
+	/*
+	 * We preload a cache-line up to 64 bytes ahead. On the 926, this will
+	 * stall only until the requested world is fetched, but the linefill
+	 * continues in the the background.
+	 * While the linefill is going, we write our previous cache-line
+	 * into the write-buffer (which should have some free space).
+	 * When the linefill is done, the writebuffer will
+	 * start dumping its content into memory
+	 *
+	 * While all this is going, we then load a full cache line into
+	 * 8 registers, this cache line should be in the cache by now
+	 * (or partly in the cache).
+	 *
+	 * This code should work well regardless of the source/dest alignment.
+	 *
+	 */
+
+	/* Align the preload register to a cache-line because the cpu does
+	 * "critical word first" (the first word requested is loaded first).
+	 */
+	@ bic           r12, r1, #0x1F
+	@ add           r12, r12, #64
+
+1:      ldmia   r1!, { r4-r11 }
+	subs    r2, r2, #32
+
+	/* 
+	 * NOTE: if r12 is more than 64 ahead of r1, the following ldrhi
+	 * for ARM9 preload will not be safely guarded by the preceding subs.
+	 * When it is safely guarded the only possibility to have SIGSEGV here
+	 * is because the caller overstates the length.
+	 */
+	@ ldrhi         r3, [r12], #32      /* cheap ARM9 preload */
+	stmia   r0!, { r4-r11 }
+	bhs     1b
+
+	add     r2, r2, #32
+
+less_than_32_left:
+	/*
+	 * less than 32 bytes left at this point (length in r2)
+	 */
+
+	/* skip all this if there is nothing to do, which should
+	 * be a common case (if not executed the code below takes
+	 * about 16 cycles)
+	 */
+	tst     r2, #0x1F
+	beq     1f
+
+	/* conditionnaly copies 0 to 31 bytes */
+	movs    r12, r2, lsl #28
+	ldmcs   r1!, {r4, r5, r6, r7}           /* 16 bytes */
+	ldmmi   r1!, {r8, r9}                   /*  8 bytes */
+	stmcs   r0!, {r4, r5, r6, r7}
+	stmmi   r0!, {r8, r9}
+	movs    r12, r2, lsl #30
+	ldrcs   r3, [r1], #4                    /*  4 bytes */
+	ldrhmi r4, [r1], #2                     /*  2 bytes */
+	strcs   r3, [r0], #4
+	strhmi r4, [r0], #2
+	tst     r2, #0x1
+	ldrbne r3, [r1]                         /*  last byte  */
+	strbne r3, [r0]
+
+	/* we're done! restore everything and return */
+1:      ldmfd   sp!, {r5-r11}
+	ldmfd   sp!, {r0, r4, lr}
+	bx      lr
+
+	/********************************************************************/
+
+non_congruent:
+	/*
+	 * here source is aligned to 4 bytes
+	 * but destination is not.
+	 *
+	 * in the code below r2 is the number of bytes read
+	 * (the number of bytes written is always smaller, because we have
+	 * partial words in the shift queue)
+	 */
+	cmp     r2, #4
+	blo     copy_last_3_and_return
+
+	/* Use post-incriment mode for stm to spill r5-r11 to reserved stack
+	 * frame. Don't update sp.
+	 */
+	stmea   sp, {r5-r11}
+
+	/* compute shifts needed to align src to dest */
+	rsb     r5, r0, #0
+	and     r5, r5, #3                      /* r5 = # bytes in partial words */
+	mov     r12, r5, lsl #3         /* r12 = right */
+	rsb     lr, r12, #32            /* lr = left  */
+
+	/* read the first word */
+	ldr     r3, [r1], #4
+	sub     r2, r2, #4
+
+	/* write a partial word (0 to 3 bytes), such that destination
+	 * becomes aligned to 32 bits (r5 = nb of words to copy for alignment)
+	 */
+	movs    r5, r5, lsl #31
+
+#if __ARMEB__
+	movmi   r3, r3, ror #24
+	strbmi	r3, [r0], #1
+	movcs   r3, r3, ror #24
+	strbcs	r3, [r0], #1
+	movcs   r3, r3, ror #24
+	strbcs	r3, [r0], #1
+#else
+	strbmi r3, [r0], #1
+	movmi   r3, r3, lsr #8
+	strbcs r3, [r0], #1
+	movcs   r3, r3, lsr #8
+	strbcs r3, [r0], #1
+	movcs   r3, r3, lsr #8
+#endif
+
+	cmp     r2, #4
+	blo     partial_word_tail
+
+#if __ARMEB__
+	mov	r3, r3, lsr r12
+	mov	r3, r3, lsl r12
+#endif
+
+	/* Align destination to 32 bytes (cache line boundary) */
+1:      tst     r0, #0x1c
+	beq     2f
+	ldr     r5, [r1], #4
+	sub     r2, r2, #4
+#if __ARMEB__
+	mov     r4, r5,                 lsr lr
+	orr     r4, r4, r3
+	mov     r3, r5,                 lsl r12
+#else
+	mov     r4, r5,                 lsl lr
+	orr     r4, r4, r3
+	mov     r3, r5,                 lsr r12
+#endif
+	str     r4, [r0], #4
+	cmp     r2, #4
+	bhs     1b
+	blo     partial_word_tail
+
+	/* copy 32 bytes at a time */
+2:      subs    r2, r2, #32
+	blo     less_than_thirtytwo
+
+	/* Use immediate mode for the shifts, because there is an extra cycle
+	 * for register shifts, which could account for up to 50% of
+	 * performance hit.
+	 */
+
+	cmp     r12, #24
+	beq     loop24
+	cmp     r12, #8
+	beq     loop8
+
+loop16:
+	ldr     r12, [r1], #4
+1:      mov     r4, r12
+	ldmia   r1!, {   r5,r6,r7,  r8,r9,r10,r11}
+	subs    r2, r2, #32
+	ldrhs   r12, [r1], #4
+#if __ARMEB__
+	orr     r3, r3, r4, lsr #16
+	mov     r4, r4, lsl #16
+	orr     r4, r4, r5, lsr #16
+	mov     r5, r5, lsl #16
+	orr     r5, r5, r6, lsr #16
+	mov     r6, r6, lsl #16
+	orr     r6, r6, r7, lsr #16
+	mov     r7, r7, lsl #16
+	orr     r7, r7, r8, lsr #16
+	mov     r8, r8, lsl #16
+	orr     r8, r8, r9, lsr #16
+	mov     r9, r9, lsl #16
+	orr     r9, r9, r10, lsr #16
+	mov     r10, r10,               lsl #16
+	orr     r10, r10, r11, lsr #16
+	stmia   r0!, {r3,r4,r5,r6, r7,r8,r9,r10}
+	mov     r3, r11, lsl #16
+#else
+	orr     r3, r3, r4, lsl #16
+	mov     r4, r4, lsr #16
+	orr     r4, r4, r5, lsl #16
+	mov     r5, r5, lsr #16
+	orr     r5, r5, r6, lsl #16
+	mov     r6, r6, lsr #16
+	orr     r6, r6, r7, lsl #16
+	mov     r7, r7, lsr #16
+	orr     r7, r7, r8, lsl #16
+	mov     r8, r8, lsr #16
+	orr     r8, r8, r9, lsl #16
+	mov     r9, r9, lsr #16
+	orr     r9, r9, r10, lsl #16
+	mov     r10, r10,               lsr #16
+	orr     r10, r10, r11, lsl #16
+	stmia   r0!, {r3,r4,r5,r6, r7,r8,r9,r10}
+	mov     r3, r11, lsr #16
+#endif
+	bhs     1b
+	b       less_than_thirtytwo
+
+loop8:
+	ldr     r12, [r1], #4
+1:      mov     r4, r12
+	ldmia   r1!, {   r5,r6,r7,  r8,r9,r10,r11}
+	subs    r2, r2, #32
+	ldrhs   r12, [r1], #4
+#if __ARMEB__
+	orr     r3, r3, r4, lsr #24
+	mov     r4, r4, lsl #8
+	orr     r4, r4, r5, lsr #24
+	mov     r5, r5, lsl #8
+	orr     r5, r5, r6, lsr #24
+	mov     r6, r6,  lsl #8
+	orr     r6, r6, r7, lsr #24
+	mov     r7, r7,  lsl #8
+	orr     r7, r7, r8,             lsr #24
+	mov     r8, r8,  lsl #8
+	orr     r8, r8, r9,             lsr #24
+	mov     r9, r9,  lsl #8
+	orr     r9, r9, r10,    lsr #24
+	mov     r10, r10, lsl #8
+	orr     r10, r10, r11,  lsr #24
+	stmia   r0!, {r3,r4,r5,r6, r7,r8,r9,r10}
+	mov     r3, r11, lsl #8
+#else
+	orr     r3, r3, r4, lsl #24
+	mov     r4, r4, lsr #8
+	orr     r4, r4, r5, lsl #24
+	mov     r5, r5, lsr #8
+	orr     r5, r5, r6, lsl #24
+	mov     r6, r6,  lsr #8
+	orr     r6, r6, r7, lsl #24
+	mov     r7, r7,  lsr #8
+	orr     r7, r7, r8,             lsl #24
+	mov     r8, r8,  lsr #8
+	orr     r8, r8, r9,             lsl #24
+	mov     r9, r9,  lsr #8
+	orr     r9, r9, r10,    lsl #24
+	mov     r10, r10, lsr #8
+	orr     r10, r10, r11,  lsl #24
+	stmia   r0!, {r3,r4,r5,r6, r7,r8,r9,r10}
+	mov     r3, r11, lsr #8
+#endif
+	bhs     1b
+	b       less_than_thirtytwo
+
+loop24:
+	ldr     r12, [r1], #4
+1:      mov     r4, r12
+	ldmia   r1!, {   r5,r6,r7,  r8,r9,r10,r11}
+	subs    r2, r2, #32
+	ldrhs   r12, [r1], #4
+#if __ARMEB__
+	orr     r3, r3, r4, lsr #8
+	mov     r4, r4, lsl #24
+	orr     r4, r4, r5, lsr #8
+	mov     r5, r5, lsl #24
+	orr     r5, r5, r6, lsr #8
+	mov     r6, r6, lsl #24
+	orr     r6, r6, r7, lsr #8
+	mov     r7, r7, lsl #24
+	orr     r7, r7, r8, lsr #8
+	mov     r8, r8, lsl #24
+	orr     r8, r8, r9, lsr #8
+	mov     r9, r9, lsl #24
+	orr     r9, r9, r10, lsr #8
+	mov     r10, r10, lsl #24
+	orr     r10, r10, r11, lsr #8
+	stmia   r0!, {r3,r4,r5,r6, r7,r8,r9,r10}
+	mov     r3, r11, lsl #24
+#else
+	orr     r3, r3, r4, lsl #8
+	mov     r4, r4, lsr #24
+	orr     r4, r4, r5, lsl #8
+	mov     r5, r5, lsr #24
+	orr     r5, r5, r6, lsl #8
+	mov     r6, r6, lsr #24
+	orr     r6, r6, r7, lsl #8
+	mov     r7, r7, lsr #24
+	orr     r7, r7, r8, lsl #8
+	mov     r8, r8, lsr #24
+	orr     r8, r8, r9, lsl #8
+	mov     r9, r9, lsr #24
+	orr     r9, r9, r10, lsl #8
+	mov     r10, r10, lsr #24
+	orr     r10, r10, r11, lsl #8
+	stmia   r0!, {r3,r4,r5,r6, r7,r8,r9,r10}
+	mov     r3, r11, lsr #24
+#endif
+	bhs     1b
+
+less_than_thirtytwo:
+	/* copy the last 0 to 31 bytes of the source */
+	rsb     r12, lr, #32            /* we corrupted r12, recompute it  */
+	add     r2, r2, #32
+	cmp     r2, #4
+	blo     partial_word_tail
+
+1:      ldr     r5, [r1], #4
+	sub     r2, r2, #4
+#if __ARMEB__
+	mov     r4, r5,                 lsr lr
+	orr     r4, r4, r3
+	mov     r3,     r5,                     lsl r12
+#else
+	mov     r4, r5,                 lsl lr
+	orr     r4, r4, r3
+	mov     r3,     r5,                     lsr r12
+#endif
+	str     r4, [r0], #4
+	cmp     r2, #4
+	bhs     1b
+
+partial_word_tail:
+	/* we have a partial word in the input buffer */
+	movs    r5, lr, lsl #(31-3)
+#if __ARMEB__
+	movmi   r3, r3, ror #24
+	strbmi r3, [r0], #1
+	movcs   r3, r3, ror #24
+	strbcs r3, [r0], #1
+	movcs   r3, r3, ror #24
+	strbcs r3, [r0], #1
+#else
+	strbmi r3, [r0], #1
+	movmi   r3, r3, lsr #8
+	strbcs r3, [r0], #1
+	movcs   r3, r3, lsr #8
+	strbcs r3, [r0], #1
+#endif
+
+	/* Refill spilled registers from the stack. Don't update sp. */
+	ldmfd   sp, {r5-r11}
+
+copy_last_3_and_return:
+	movs    r2, r2, lsl #31 /* copy remaining 0, 1, 2 or 3 bytes */
+	ldrbmi r2, [r1], #1
+	ldrbcs r3, [r1], #1
+	ldrbcs r12,[r1]
+	strbmi r2, [r0], #1
+	strbcs r3, [r0], #1
+	strbcs r12,[r0]
+
+	/* we're done! restore sp and spilled registers and return */
+	add     sp,  sp, #28
+	ldmfd   sp!, {r0, r4, lr}
+	bx      lr
+
diff --git a/na/lib/string/bcmp.c b/na/lib/string/bcmp.c
new file mode 100644
index 0000000..d4c093f
--- /dev/null
+++ b/na/lib/string/bcmp.c
@@ -0,0 +1,11 @@
+/* SPDX-License-Identifier: MIT */
+/* Derived from musl libc (Copyright © 2005-2020 Rich Felker, et al.) */
+
+#define _BSD_SOURCE
+#include <string.h>
+#include <strings.h>
+
+int bcmp(const void *s1, const void *s2, size_t n)
+{
+	return memcmp(s1, s2, n);
+}
diff --git a/na/lib/string/bcopy.c b/na/lib/string/bcopy.c
new file mode 100644
index 0000000..25e98a2
--- /dev/null
+++ b/na/lib/string/bcopy.c
@@ -0,0 +1,11 @@
+/* SPDX-License-Identifier: MIT */
+/* Derived from musl libc (Copyright © 2005-2020 Rich Felker, et al.) */
+
+#define _BSD_SOURCE
+#include <string.h>
+#include <strings.h>
+
+void bcopy(const void *s1, void *s2, size_t n)
+{
+	memmove(s2, s1, n);
+}
diff --git a/na/lib/string/bzero.c b/na/lib/string/bzero.c
new file mode 100644
index 0000000..08f1398
--- /dev/null
+++ b/na/lib/string/bzero.c
@@ -0,0 +1,11 @@
+/* SPDX-License-Identifier: MIT */
+/* Derived from musl libc (Copyright © 2005-2020 Rich Felker, et al.) */
+
+#define _BSD_SOURCE
+#include <string.h>
+#include <strings.h>
+
+void bzero(void *s, size_t n)
+{
+	memset(s, 0, n);
+}
diff --git a/na/lib/string/explicit_bzero.c b/na/lib/string/explicit_bzero.c
new file mode 100644
index 0000000..c94f25f
--- /dev/null
+++ b/na/lib/string/explicit_bzero.c
@@ -0,0 +1,11 @@
+/* SPDX-License-Identifier: MIT */
+/* Derived from musl libc (Copyright © 2005-2020 Rich Felker, et al.) */
+
+#define _BSD_SOURCE
+#include <string.h>
+
+void explicit_bzero(void *d, size_t n)
+{
+	d = memset(d, 0, n);
+	__asm__ __volatile__ ("" : : "r"(d) : "memory");
+}
diff --git a/na/lib/string/i386/memcpy.s b/na/lib/string/i386/memcpy.s
new file mode 100644
index 0000000..0608dd8
--- /dev/null
+++ b/na/lib/string/i386/memcpy.s
@@ -0,0 +1,32 @@
+.global memcpy
+.global __memcpy_fwd
+.hidden __memcpy_fwd
+.type memcpy,@function
+memcpy:
+__memcpy_fwd:
+	push %esi
+	push %edi
+	mov 12(%esp),%edi
+	mov 16(%esp),%esi
+	mov 20(%esp),%ecx
+	mov %edi,%eax
+	cmp $4,%ecx
+	jc 1f
+	test $3,%edi
+	jz 1f
+2:	movsb
+	dec %ecx
+	test $3,%edi
+	jnz 2b
+1:	mov %ecx,%edx
+	shr $2,%ecx
+	rep
+	movsl
+	and $3,%edx
+	jz 1f
+2:	movsb
+	dec %edx
+	jnz 2b
+1:	pop %edi
+	pop %esi
+	ret
diff --git a/na/lib/string/i386/memmove.s b/na/lib/string/i386/memmove.s
new file mode 100644
index 0000000..2a6a504
--- /dev/null
+++ b/na/lib/string/i386/memmove.s
@@ -0,0 +1,22 @@
+.global memmove
+.type memmove,@function
+memmove:
+	mov 4(%esp),%eax
+	sub 8(%esp),%eax
+	cmp 12(%esp),%eax
+.hidden __memcpy_fwd
+	jae __memcpy_fwd
+	push %esi
+	push %edi
+	mov 12(%esp),%edi
+	mov 16(%esp),%esi
+	mov 20(%esp),%ecx
+	lea -1(%edi,%ecx),%edi
+	lea -1(%esi,%ecx),%esi
+	std
+	rep movsb
+	cld
+	lea 1(%edi),%eax
+	pop %edi
+	pop %esi
+	ret
diff --git a/na/lib/string/i386/memset.s b/na/lib/string/i386/memset.s
new file mode 100644
index 0000000..d00422c
--- /dev/null
+++ b/na/lib/string/i386/memset.s
@@ -0,0 +1,76 @@
+.global memset
+.type memset,@function
+memset:
+	mov 12(%esp),%ecx
+	cmp $62,%ecx
+	ja 2f
+
+	mov 8(%esp),%dl
+	mov 4(%esp),%eax
+	test %ecx,%ecx
+	jz 1f
+
+	mov %dl,%dh
+
+	mov %dl,(%eax)
+	mov %dl,-1(%eax,%ecx)
+	cmp $2,%ecx
+	jbe 1f
+
+	mov %dx,1(%eax)
+	mov %dx,(-1-2)(%eax,%ecx)
+	cmp $6,%ecx
+	jbe 1f
+
+	shl $16,%edx
+	mov 8(%esp),%dl
+	mov 8(%esp),%dh
+
+	mov %edx,(1+2)(%eax)
+	mov %edx,(-1-2-4)(%eax,%ecx)
+	cmp $14,%ecx
+	jbe 1f
+
+	mov %edx,(1+2+4)(%eax)
+	mov %edx,(1+2+4+4)(%eax)
+	mov %edx,(-1-2-4-8)(%eax,%ecx)
+	mov %edx,(-1-2-4-4)(%eax,%ecx)
+	cmp $30,%ecx
+	jbe 1f
+
+	mov %edx,(1+2+4+8)(%eax)
+	mov %edx,(1+2+4+8+4)(%eax)
+	mov %edx,(1+2+4+8+8)(%eax)
+	mov %edx,(1+2+4+8+12)(%eax)
+	mov %edx,(-1-2-4-8-16)(%eax,%ecx)
+	mov %edx,(-1-2-4-8-12)(%eax,%ecx)
+	mov %edx,(-1-2-4-8-8)(%eax,%ecx)
+	mov %edx,(-1-2-4-8-4)(%eax,%ecx)
+
+1:	ret 	
+
+2:	movzbl 8(%esp),%eax
+	mov %edi,12(%esp)
+	imul $0x1010101,%eax
+	mov 4(%esp),%edi
+	test $15,%edi
+	mov %eax,-4(%edi,%ecx)
+	jnz 2f
+
+1:	shr $2, %ecx
+	rep
+	stosl
+	mov 4(%esp),%eax
+	mov 12(%esp),%edi
+	ret
+	
+2:	xor %edx,%edx
+	sub %edi,%edx
+	and $15,%edx
+	mov %eax,(%edi)
+	mov %eax,4(%edi)
+	mov %eax,8(%edi)
+	mov %eax,12(%edi)
+	sub %edx,%ecx
+	add %edx,%edi
+	jmp 1b
diff --git a/na/lib/string/index.c b/na/lib/string/index.c
new file mode 100644
index 0000000..8f6894c
--- /dev/null
+++ b/na/lib/string/index.c
@@ -0,0 +1,11 @@
+/* SPDX-License-Identifier: MIT */
+/* Derived from musl libc (Copyright © 2005-2020 Rich Felker, et al.) */
+
+#define _BSD_SOURCE
+#include <string.h>
+#include <strings.h>
+
+char *index(const char *s, int c)
+{
+	return strchr(s, c);
+}
diff --git a/na/lib/string/memccpy.c b/na/lib/string/memccpy.c
new file mode 100644
index 0000000..99f146f
--- /dev/null
+++ b/na/lib/string/memccpy.c
@@ -0,0 +1,37 @@
+/* SPDX-License-Identifier: MIT */
+/* Derived from musl libc (Copyright © 2005-2020 Rich Felker, et al.) */
+
+#include <string.h>
+#include <stdint.h>
+#include <limits.h>
+
+#define ALIGN (sizeof(size_t)-1)
+#define ONES ((size_t)-1/UCHAR_MAX)
+#define HIGHS (ONES * (UCHAR_MAX/2+1))
+#define HASZERO(x) ((x)-ONES & ~(x) & HIGHS)
+
+void *memccpy(void *restrict dest, const void *restrict src, int c, size_t n)
+{
+	unsigned char *d = dest;
+	const unsigned char *s = src;
+
+	c = (unsigned char)c;
+#ifdef __GNUC__
+	typedef size_t __attribute__((__may_alias__)) word;
+	word *wd;
+	const word *ws;
+	if (((uintptr_t)s & ALIGN) == ((uintptr_t)d & ALIGN)) {
+		for (; ((uintptr_t)s & ALIGN) && n && (*d=*s)!=c; n--, s++, d++);
+		if ((uintptr_t)s & ALIGN) goto tail;
+		size_t k = ONES * c;
+		wd=(void *)d; ws=(const void *)s;
+		for (; n>=sizeof(size_t) && !HASZERO(*ws^k);
+		       n-=sizeof(size_t), ws++, wd++) *wd = *ws;
+		d=(void *)wd; s=(const void *)ws;
+	}
+#endif
+	for (; n && (*d=*s)!=c; n--, s++, d++);
+tail:
+	if (n) return d+1;
+	return 0;
+}
diff --git a/na/lib/string/memchr.c b/na/lib/string/memchr.c
new file mode 100644
index 0000000..d0f670f
--- /dev/null
+++ b/na/lib/string/memchr.c
@@ -0,0 +1,30 @@
+/* SPDX-License-Identifier: MIT */
+/* Derived from musl libc (Copyright © 2005-2020 Rich Felker, et al.) */
+
+#include <string.h>
+#include <stdint.h>
+#include <limits.h>
+
+#define SS (sizeof(size_t))
+#define ALIGN (sizeof(size_t)-1)
+#define ONES ((size_t)-1/UCHAR_MAX)
+#define HIGHS (ONES * (UCHAR_MAX/2+1))
+#define HASZERO(x) ((x)-ONES & ~(x) & HIGHS)
+
+void *memchr(const void *src, int c, size_t n)
+{
+	const unsigned char *s = src;
+	c = (unsigned char)c;
+#ifdef __GNUC__
+	for (; ((uintptr_t)s & ALIGN) && n && *s != c; s++, n--);
+	if (n && *s != c) {
+		typedef size_t __attribute__((__may_alias__)) word;
+		const word *w;
+		size_t k = ONES * c;
+		for (w = (const void *)s; n>=SS && !HASZERO(*w^k); w++, n-=SS);
+		s = (const void *)w;
+	}
+#endif
+	for (; n && *s != c; s++, n--);
+	return n ? (void *)s : 0;
+}
diff --git a/na/lib/string/memcmp.c b/na/lib/string/memcmp.c
new file mode 100644
index 0000000..15606f4
--- /dev/null
+++ b/na/lib/string/memcmp.c
@@ -0,0 +1,11 @@
+/* SPDX-License-Identifier: MIT */
+/* Derived from musl libc (Copyright © 2005-2020 Rich Felker, et al.) */
+
+#include <string.h>
+
+int memcmp(const void *vl, const void *vr, size_t n)
+{
+	const unsigned char *l=vl, *r=vr;
+	for (; n && *l == *r; n--, l++, r++);
+	return n ? *l-*r : 0;
+}
diff --git a/na/lib/string/memcpy.c b/na/lib/string/memcpy.c
new file mode 100644
index 0000000..de26a00
--- /dev/null
+++ b/na/lib/string/memcpy.c
@@ -0,0 +1,127 @@
+/* SPDX-License-Identifier: MIT */
+/* Derived from musl libc (Copyright © 2005-2020 Rich Felker, et al.) */
+
+#include <string.h>
+#include <stdint.h>
+#include <endian.h>
+
+void *memcpy(void *restrict dest, const void *restrict src, size_t n)
+{
+	unsigned char *d = dest;
+	const unsigned char *s = src;
+
+#ifdef __GNUC__
+
+#if __BYTE_ORDER == __LITTLE_ENDIAN
+#define LS >>
+#define RS <<
+#else
+#define LS <<
+#define RS >>
+#endif
+
+	typedef uint32_t __attribute__((__may_alias__)) u32;
+	uint32_t w, x;
+
+	for (; (uintptr_t)s % 4 && n; n--) *d++ = *s++;
+
+	if ((uintptr_t)d % 4 == 0) {
+		for (; n>=16; s+=16, d+=16, n-=16) {
+			*(u32 *)(d+0) = *(u32 *)(s+0);
+			*(u32 *)(d+4) = *(u32 *)(s+4);
+			*(u32 *)(d+8) = *(u32 *)(s+8);
+			*(u32 *)(d+12) = *(u32 *)(s+12);
+		}
+		if (n&8) {
+			*(u32 *)(d+0) = *(u32 *)(s+0);
+			*(u32 *)(d+4) = *(u32 *)(s+4);
+			d += 8; s += 8;
+		}
+		if (n&4) {
+			*(u32 *)(d+0) = *(u32 *)(s+0);
+			d += 4; s += 4;
+		}
+		if (n&2) {
+			*d++ = *s++; *d++ = *s++;
+		}
+		if (n&1) {
+			*d = *s;
+		}
+		return dest;
+	}
+
+	if (n >= 32) switch ((uintptr_t)d % 4) {
+	case 1:
+		w = *(u32 *)s;
+		*d++ = *s++;
+		*d++ = *s++;
+		*d++ = *s++;
+		n -= 3;
+		for (; n>=17; s+=16, d+=16, n-=16) {
+			x = *(u32 *)(s+1);
+			*(u32 *)(d+0) = (w LS 24) | (x RS 8);
+			w = *(u32 *)(s+5);
+			*(u32 *)(d+4) = (x LS 24) | (w RS 8);
+			x = *(u32 *)(s+9);
+			*(u32 *)(d+8) = (w LS 24) | (x RS 8);
+			w = *(u32 *)(s+13);
+			*(u32 *)(d+12) = (x LS 24) | (w RS 8);
+		}
+		break;
+	case 2:
+		w = *(u32 *)s;
+		*d++ = *s++;
+		*d++ = *s++;
+		n -= 2;
+		for (; n>=18; s+=16, d+=16, n-=16) {
+			x = *(u32 *)(s+2);
+			*(u32 *)(d+0) = (w LS 16) | (x RS 16);
+			w = *(u32 *)(s+6);
+			*(u32 *)(d+4) = (x LS 16) | (w RS 16);
+			x = *(u32 *)(s+10);
+			*(u32 *)(d+8) = (w LS 16) | (x RS 16);
+			w = *(u32 *)(s+14);
+			*(u32 *)(d+12) = (x LS 16) | (w RS 16);
+		}
+		break;
+	case 3:
+		w = *(u32 *)s;
+		*d++ = *s++;
+		n -= 1;
+		for (; n>=19; s+=16, d+=16, n-=16) {
+			x = *(u32 *)(s+3);
+			*(u32 *)(d+0) = (w LS 8) | (x RS 24);
+			w = *(u32 *)(s+7);
+			*(u32 *)(d+4) = (x LS 8) | (w RS 24);
+			x = *(u32 *)(s+11);
+			*(u32 *)(d+8) = (w LS 8) | (x RS 24);
+			w = *(u32 *)(s+15);
+			*(u32 *)(d+12) = (x LS 8) | (w RS 24);
+		}
+		break;
+	}
+	if (n&16) {
+		*d++ = *s++; *d++ = *s++; *d++ = *s++; *d++ = *s++;
+		*d++ = *s++; *d++ = *s++; *d++ = *s++; *d++ = *s++;
+		*d++ = *s++; *d++ = *s++; *d++ = *s++; *d++ = *s++;
+		*d++ = *s++; *d++ = *s++; *d++ = *s++; *d++ = *s++;
+	}
+	if (n&8) {
+		*d++ = *s++; *d++ = *s++; *d++ = *s++; *d++ = *s++;
+		*d++ = *s++; *d++ = *s++; *d++ = *s++; *d++ = *s++;
+	}
+	if (n&4) {
+		*d++ = *s++; *d++ = *s++; *d++ = *s++; *d++ = *s++;
+	}
+	if (n&2) {
+		*d++ = *s++; *d++ = *s++;
+	}
+	if (n&1) {
+		*d = *s;
+	}
+	return dest;
+#endif
+
+	for (; n; n--) *d++ = *s++;
+	return dest;
+}
diff --git a/na/lib/string/memmem.c b/na/lib/string/memmem.c
new file mode 100644
index 0000000..cfd23bb
--- /dev/null
+++ b/na/lib/string/memmem.c
@@ -0,0 +1,152 @@
+/* SPDX-License-Identifier: MIT */
+/* Derived from musl libc (Copyright © 2005-2020 Rich Felker, et al.) */
+
+#define _GNU_SOURCE
+#include <string.h>
+#include <stdint.h>
+
+static char *twobyte_memmem(const unsigned char *h, size_t k, const unsigned char *n)
+{
+	uint16_t nw = n[0]<<8 | n[1], hw = h[0]<<8 | h[1];
+	for (h+=2, k-=2; k; k--, hw = hw<<8 | *h++)
+		if (hw == nw) return (char *)h-2;
+	return hw == nw ? (char *)h-2 : 0;
+}
+
+static char *threebyte_memmem(const unsigned char *h, size_t k, const unsigned char *n)
+{
+	uint32_t nw = (uint32_t)n[0]<<24 | n[1]<<16 | n[2]<<8;
+	uint32_t hw = (uint32_t)h[0]<<24 | h[1]<<16 | h[2]<<8;
+	for (h+=3, k-=3; k; k--, hw = (hw|*h++)<<8)
+		if (hw == nw) return (char *)h-3;
+	return hw == nw ? (char *)h-3 : 0;
+}
+
+static char *fourbyte_memmem(const unsigned char *h, size_t k, const unsigned char *n)
+{
+	uint32_t nw = (uint32_t)n[0]<<24 | n[1]<<16 | n[2]<<8 | n[3];
+	uint32_t hw = (uint32_t)h[0]<<24 | h[1]<<16 | h[2]<<8 | h[3];
+	for (h+=4, k-=4; k; k--, hw = hw<<8 | *h++)
+		if (hw == nw) return (char *)h-4;
+	return hw == nw ? (char *)h-4 : 0;
+}
+
+#define MAX(a,b) ((a)>(b)?(a):(b))
+#define MIN(a,b) ((a)<(b)?(a):(b))
+
+#define BITOP(a,b,op) \
+ ((a)[(size_t)(b)/(8*sizeof *(a))] op (size_t)1<<((size_t)(b)%(8*sizeof *(a))))
+
+static char *twoway_memmem(const unsigned char *h, const unsigned char *z, const unsigned char *n, size_t l)
+{
+	size_t i, ip, jp, k, p, ms, p0, mem, mem0;
+	size_t byteset[32 / sizeof(size_t)] = { 0 };
+	size_t shift[256];
+
+	/* Computing length of needle and fill shift table */
+	for (i=0; i<l; i++)
+		BITOP(byteset, n[i], |=), shift[n[i]] = i+1;
+
+	/* Compute maximal suffix */
+	ip = -1; jp = 0; k = p = 1;
+	while (jp+k<l) {
+		if (n[ip+k] == n[jp+k]) {
+			if (k == p) {
+				jp += p;
+				k = 1;
+			} else k++;
+		} else if (n[ip+k] > n[jp+k]) {
+			jp += k;
+			k = 1;
+			p = jp - ip;
+		} else {
+			ip = jp++;
+			k = p = 1;
+		}
+	}
+	ms = ip;
+	p0 = p;
+
+	/* And with the opposite comparison */
+	ip = -1; jp = 0; k = p = 1;
+	while (jp+k<l) {
+		if (n[ip+k] == n[jp+k]) {
+			if (k == p) {
+				jp += p;
+				k = 1;
+			} else k++;
+		} else if (n[ip+k] < n[jp+k]) {
+			jp += k;
+			k = 1;
+			p = jp - ip;
+		} else {
+			ip = jp++;
+			k = p = 1;
+		}
+	}
+	if (ip+1 > ms+1) ms = ip;
+	else p = p0;
+
+	/* Periodic needle? */
+	if (memcmp(n, n+p, ms+1)) {
+		mem0 = 0;
+		p = MAX(ms, l-ms-1) + 1;
+	} else mem0 = l-p;
+	mem = 0;
+
+	/* Search loop */
+	for (;;) {
+		/* If remainder of haystack is shorter than needle, done */
+		if (z-h < l) return 0;
+
+		/* Check last byte first; advance by shift on mismatch */
+		if (BITOP(byteset, h[l-1], &)) {
+			k = l-shift[h[l-1]];
+			if (k) {
+				if (k < mem) k = mem;
+				h += k;
+				mem = 0;
+				continue;
+			}
+		} else {
+			h += l;
+			mem = 0;
+			continue;
+		}
+
+		/* Compare right half */
+		for (k=MAX(ms+1,mem); k<l && n[k] == h[k]; k++);
+		if (k < l) {
+			h += k-ms;
+			mem = 0;
+			continue;
+		}
+		/* Compare left half */
+		for (k=ms+1; k>mem && n[k-1] == h[k-1]; k--);
+		if (k <= mem) return (char *)h;
+		h += p;
+		mem = mem0;
+	}
+}
+
+void *memmem(const void *h0, size_t k, const void *n0, size_t l)
+{
+	const unsigned char *h = h0, *n = n0;
+
+	/* Return immediately on empty needle */
+	if (!l) return (void *)h;
+
+	/* Return immediately when needle is longer than haystack */
+	if (k<l) return 0;
+
+	/* Use faster algorithms for short needles */
+	h = memchr(h0, *n, k);
+	if (!h || l==1) return (void *)h;
+	k -= h - (const unsigned char *)h0;
+	if (k<l) return 0;
+	if (l==2) return twobyte_memmem(h, k, n);
+	if (l==3) return threebyte_memmem(h, k, n);
+	if (l==4) return fourbyte_memmem(h, k, n);
+
+	return twoway_memmem(h, h+k, n, l);
+}
diff --git a/na/lib/string/memmove.c b/na/lib/string/memmove.c
new file mode 100644
index 0000000..56e1a81
--- /dev/null
+++ b/na/lib/string/memmove.c
@@ -0,0 +1,45 @@
+/* SPDX-License-Identifier: MIT */
+/* Derived from musl libc (Copyright © 2005-2020 Rich Felker, et al.) */
+
+#include <string.h>
+#include <stdint.h>
+
+#ifdef __GNUC__
+typedef __attribute__((__may_alias__)) size_t WT;
+#define WS (sizeof(WT))
+#endif
+
+void *memmove(void *dest, const void *src, size_t n)
+{
+	char *d = dest;
+	const char *s = src;
+
+	if (d==s) return d;
+	if ((uintptr_t)s-(uintptr_t)d-n <= -2*n) return memcpy(d, s, n);
+
+	if (d<s) {
+#ifdef __GNUC__
+		if ((uintptr_t)s % WS == (uintptr_t)d % WS) {
+			while ((uintptr_t)d % WS) {
+				if (!n--) return dest;
+				*d++ = *s++;
+			}
+			for (; n>=WS; n-=WS, d+=WS, s+=WS) *(WT *)d = *(WT *)s;
+		}
+#endif
+		for (; n; n--) *d++ = *s++;
+	} else {
+#ifdef __GNUC__
+		if ((uintptr_t)s % WS == (uintptr_t)d % WS) {
+			while ((uintptr_t)(d+n) % WS) {
+				if (!n--) return dest;
+				d[n] = s[n];
+			}
+			while (n>=WS) n-=WS, *(WT *)(d+n) = *(WT *)(s+n);
+		}
+#endif
+		while (n) n--, d[n] = s[n];
+	}
+
+	return dest;
+}
diff --git a/na/lib/string/mempcpy.c b/na/lib/string/mempcpy.c
new file mode 100644
index 0000000..1f33e45
--- /dev/null
+++ b/na/lib/string/mempcpy.c
@@ -0,0 +1,10 @@
+/* SPDX-License-Identifier: MIT */
+/* Derived from musl libc (Copyright © 2005-2020 Rich Felker, et al.) */
+
+#define _GNU_SOURCE
+#include <string.h>
+
+void *mempcpy(void *dest, const void *src, size_t n)
+{
+	return (char *)memcpy(dest, src, n) + n;
+}
diff --git a/na/lib/string/memrchr.c b/na/lib/string/memrchr.c
new file mode 100644
index 0000000..8d5c742
--- /dev/null
+++ b/na/lib/string/memrchr.c
@@ -0,0 +1,14 @@
+/* SPDX-License-Identifier: MIT */
+/* Derived from musl libc (Copyright © 2005-2020 Rich Felker, et al.) */
+
+#include <string.h>
+
+void *__memrchr(const void *m, int c, size_t n)
+{
+	const unsigned char *s = m;
+	c = (unsigned char)c;
+	while (n--) if (s[n]==c) return (void *)(s+n);
+	return 0;
+}
+
+weak_alias(__memrchr, memrchr);
diff --git a/na/lib/string/memset.c b/na/lib/string/memset.c
new file mode 100644
index 0000000..e780c71
--- /dev/null
+++ b/na/lib/string/memset.c
@@ -0,0 +1,93 @@
+/* SPDX-License-Identifier: MIT */
+/* Derived from musl libc (Copyright © 2005-2020 Rich Felker, et al.) */
+
+#include <string.h>
+#include <stdint.h>
+
+void *memset(void *dest, int c, size_t n)
+{
+	unsigned char *s = dest;
+	size_t k;
+
+	/* Fill head and tail with minimal branching. Each
+	 * conditional ensures that all the subsequently used
+	 * offsets are well-defined and in the dest region. */
+
+	if (!n) return dest;
+	s[0] = c;
+	s[n-1] = c;
+	if (n <= 2) return dest;
+	s[1] = c;
+	s[2] = c;
+	s[n-2] = c;
+	s[n-3] = c;
+	if (n <= 6) return dest;
+	s[3] = c;
+	s[n-4] = c;
+	if (n <= 8) return dest;
+
+	/* Advance pointer to align it at a 4-byte boundary,
+	 * and truncate n to a multiple of 4. The previous code
+	 * already took care of any head/tail that get cut off
+	 * by the alignment. */
+
+	k = -(uintptr_t)s & 3;
+	s += k;
+	n -= k;
+	n &= -4;
+
+#ifdef __GNUC__
+	typedef uint32_t __attribute__((__may_alias__)) u32;
+	typedef uint64_t __attribute__((__may_alias__)) u64;
+
+	u32 c32 = ((u32)-1)/255 * (unsigned char)c;
+
+	/* In preparation to copy 32 bytes at a time, aligned on
+	 * an 8-byte bounary, fill head/tail up to 28 bytes each.
+	 * As in the initial byte-based head/tail fill, each
+	 * conditional below ensures that the subsequent offsets
+	 * are valid (e.g. !(n<=24) implies n>=28). */
+
+	*(u32 *)(s+0) = c32;
+	*(u32 *)(s+n-4) = c32;
+	if (n <= 8) return dest;
+	*(u32 *)(s+4) = c32;
+	*(u32 *)(s+8) = c32;
+	*(u32 *)(s+n-12) = c32;
+	*(u32 *)(s+n-8) = c32;
+	if (n <= 24) return dest;
+	*(u32 *)(s+12) = c32;
+	*(u32 *)(s+16) = c32;
+	*(u32 *)(s+20) = c32;
+	*(u32 *)(s+24) = c32;
+	*(u32 *)(s+n-28) = c32;
+	*(u32 *)(s+n-24) = c32;
+	*(u32 *)(s+n-20) = c32;
+	*(u32 *)(s+n-16) = c32;
+
+	/* Align to a multiple of 8 so we can fill 64 bits at a time,
+	 * and avoid writing the same bytes twice as much as is
+	 * practical without introducing additional branching. */
+
+	k = 24 + ((uintptr_t)s & 4);
+	s += k;
+	n -= k;
+
+	/* If this loop is reached, 28 tail bytes have already been
+	 * filled, so any remainder when n drops below 32 can be
+	 * safely ignored. */
+
+	u64 c64 = c32 | ((u64)c32 << 32);
+	for (; n >= 32; n-=32, s+=32) {
+		*(u64 *)(s+0) = c64;
+		*(u64 *)(s+8) = c64;
+		*(u64 *)(s+16) = c64;
+		*(u64 *)(s+24) = c64;
+	}
+#else
+	/* Pure C fallback with no aliasing violations. */
+	for (; n; n--, s++) *s = c;
+#endif
+
+	return dest;
+}
diff --git a/na/lib/string/rindex.c b/na/lib/string/rindex.c
new file mode 100644
index 0000000..897e19e
--- /dev/null
+++ b/na/lib/string/rindex.c
@@ -0,0 +1,11 @@
+/* SPDX-License-Identifier: MIT */
+/* Derived from musl libc (Copyright © 2005-2020 Rich Felker, et al.) */
+
+#define _BSD_SOURCE
+#include <string.h>
+#include <strings.h>
+
+char *rindex(const char *s, int c)
+{
+	return strrchr(s, c);
+}
diff --git a/na/lib/string/stpcpy.c b/na/lib/string/stpcpy.c
new file mode 100644
index 0000000..0c1f1b4
--- /dev/null
+++ b/na/lib/string/stpcpy.c
@@ -0,0 +1,32 @@
+/* SPDX-License-Identifier: MIT */
+/* Derived from musl libc (Copyright © 2005-2020 Rich Felker, et al.) */
+
+#include <string.h>
+#include <stdint.h>
+#include <limits.h>
+
+#define ALIGN (sizeof(size_t))
+#define ONES ((size_t)-1/UCHAR_MAX)
+#define HIGHS (ONES * (UCHAR_MAX/2+1))
+#define HASZERO(x) ((x)-ONES & ~(x) & HIGHS)
+
+char *__stpcpy(char *restrict d, const char *restrict s)
+{
+#ifdef __GNUC__
+	typedef size_t __attribute__((__may_alias__)) word;
+	word *wd;
+	const word *ws;
+	if ((uintptr_t)s % ALIGN == (uintptr_t)d % ALIGN) {
+		for (; (uintptr_t)s % ALIGN; s++, d++)
+			if (!(*d=*s)) return d;
+		wd=(void *)d; ws=(const void *)s;
+		for (; !HASZERO(*ws); *wd++ = *ws++);
+		d=(void *)wd; s=(const void *)ws;
+	}
+#endif
+	for (; (*d=*s); s++, d++);
+
+	return d;
+}
+
+weak_alias(__stpcpy, stpcpy);
diff --git a/na/lib/string/stpncpy.c b/na/lib/string/stpncpy.c
new file mode 100644
index 0000000..ca2ef78
--- /dev/null
+++ b/na/lib/string/stpncpy.c
@@ -0,0 +1,35 @@
+/* SPDX-License-Identifier: MIT */
+/* Derived from musl libc (Copyright © 2005-2020 Rich Felker, et al.) */
+
+#include <string.h>
+#include <stdint.h>
+#include <limits.h>
+
+#define ALIGN (sizeof(size_t)-1)
+#define ONES ((size_t)-1/UCHAR_MAX)
+#define HIGHS (ONES * (UCHAR_MAX/2+1))
+#define HASZERO(x) ((x)-ONES & ~(x) & HIGHS)
+
+char *__stpncpy(char *restrict d, const char *restrict s, size_t n)
+{
+#ifdef __GNUC__
+	typedef size_t __attribute__((__may_alias__)) word;
+	word *wd;
+	const word *ws;
+	if (((uintptr_t)s & ALIGN) == ((uintptr_t)d & ALIGN)) {
+		for (; ((uintptr_t)s & ALIGN) && n && (*d=*s); n--, s++, d++);
+		if (!n || !*s) goto tail;
+		wd=(void *)d; ws=(const void *)s;
+		for (; n>=sizeof(size_t) && !HASZERO(*ws);
+		       n-=sizeof(size_t), ws++, wd++) *wd = *ws;
+		d=(void *)wd; s=(const void *)ws;
+	}
+#endif
+	for (; n && (*d=*s); n--, s++, d++);
+tail:
+	memset(d, 0, n);
+	return d;
+}
+
+weak_alias(__stpncpy, stpncpy);
+
diff --git a/na/lib/string/strcat.c b/na/lib/string/strcat.c
new file mode 100644
index 0000000..723c2ff
--- /dev/null
+++ b/na/lib/string/strcat.c
@@ -0,0 +1,10 @@
+/* SPDX-License-Identifier: MIT */
+/* Derived from musl libc (Copyright © 2005-2020 Rich Felker, et al.) */
+
+#include <string.h>
+
+char *strcat(char *restrict dest, const char *restrict src)
+{
+	strcpy(dest + strlen(dest), src);
+	return dest;
+}
diff --git a/na/lib/string/strchr.c b/na/lib/string/strchr.c
new file mode 100644
index 0000000..ddeca01
--- /dev/null
+++ b/na/lib/string/strchr.c
@@ -0,0 +1,10 @@
+/* SPDX-License-Identifier: MIT */
+/* Derived from musl libc (Copyright © 2005-2020 Rich Felker, et al.) */
+
+#include <string.h>
+
+char *strchr(const char *s, int c)
+{
+	char *r = __strchrnul(s, c);
+	return *(unsigned char *)r == (unsigned char)c ? r : 0;
+}
diff --git a/na/lib/string/strchrnul.c b/na/lib/string/strchrnul.c
new file mode 100644
index 0000000..6fc86d1
--- /dev/null
+++ b/na/lib/string/strchrnul.c
@@ -0,0 +1,31 @@
+/* SPDX-License-Identifier: MIT */
+/* Derived from musl libc (Copyright © 2005-2020 Rich Felker, et al.) */
+
+#include <string.h>
+#include <stdint.h>
+#include <limits.h>
+
+#define ALIGN (sizeof(size_t))
+#define ONES ((size_t)-1/UCHAR_MAX)
+#define HIGHS (ONES * (UCHAR_MAX/2+1))
+#define HASZERO(x) ((x)-ONES & ~(x) & HIGHS)
+
+char *__strchrnul(const char *s, int c)
+{
+	c = (unsigned char)c;
+	if (!c) return (char *)s + strlen(s);
+
+#ifdef __GNUC__
+	typedef size_t __attribute__((__may_alias__)) word;
+	const word *w;
+	for (; (uintptr_t)s % ALIGN; s++)
+		if (!*s || *(unsigned char *)s == c) return (char *)s;
+	size_t k = ONES * c;
+	for (w = (void *)s; !HASZERO(*w) && !HASZERO(*w^k); w++);
+	s = (void *)w;
+#endif
+	for (; *s && *(unsigned char *)s != c; s++);
+	return (char *)s;
+}
+
+weak_alias(__strchrnul, strchrnul);
diff --git a/na/lib/string/strcmp.c b/na/lib/string/strcmp.c
new file mode 100644
index 0000000..afcb1b7
--- /dev/null
+++ b/na/lib/string/strcmp.c
@@ -0,0 +1,10 @@
+/* SPDX-License-Identifier: MIT */
+/* Derived from musl libc (Copyright © 2005-2020 Rich Felker, et al.) */
+
+#include <string.h>
+
+int strcmp(const char *l, const char *r)
+{
+	for (; *l==*r && *l; l++, r++);
+	return *(unsigned char *)l - *(unsigned char *)r;
+}
diff --git a/na/lib/string/strcpy.c b/na/lib/string/strcpy.c
new file mode 100644
index 0000000..a761577
--- /dev/null
+++ b/na/lib/string/strcpy.c
@@ -0,0 +1,10 @@
+/* SPDX-License-Identifier: MIT */
+/* Derived from musl libc (Copyright © 2005-2020 Rich Felker, et al.) */
+
+#include <string.h>
+
+char *strcpy(char *restrict dest, const char *restrict src)
+{
+	__stpcpy(dest, src);
+	return dest;
+}
diff --git a/na/lib/string/strcspn.c b/na/lib/string/strcspn.c
new file mode 100644
index 0000000..ba55fe6
--- /dev/null
+++ b/na/lib/string/strcspn.c
@@ -0,0 +1,20 @@
+/* SPDX-License-Identifier: MIT */
+/* Derived from musl libc (Copyright © 2005-2020 Rich Felker, et al.) */
+
+#include <string.h>
+
+#define BITOP(a,b,op) \
+ ((a)[(size_t)(b)/(8*sizeof *(a))] op (size_t)1<<((size_t)(b)%(8*sizeof *(a))))
+
+size_t strcspn(const char *s, const char *c)
+{
+	const char *a = s;
+	size_t byteset[32/sizeof(size_t)];
+
+	if (!c[0] || !c[1]) return __strchrnul(s, *c)-a;
+
+	memset(byteset, 0, sizeof byteset);
+	for (; *c && BITOP(byteset, *(unsigned char *)c, |=); c++);
+	for (; *s && !BITOP(byteset, *(unsigned char *)s, &); s++);
+	return s-a;
+}
diff --git a/na/lib/string/strlcat.c b/na/lib/string/strlcat.c
new file mode 100644
index 0000000..a61821d
--- /dev/null
+++ b/na/lib/string/strlcat.c
@@ -0,0 +1,12 @@
+/* SPDX-License-Identifier: MIT */
+/* Derived from musl libc (Copyright © 2005-2020 Rich Felker, et al.) */
+
+#define _BSD_SOURCE
+#include <string.h>
+
+size_t strlcat(char *d, const char *s, size_t n)
+{
+	size_t l = strnlen(d, n);
+	if (l == n) return l + strlen(s);
+	return l + strlcpy(d+l, s, n-l);
+}
diff --git a/na/lib/string/strlcpy.c b/na/lib/string/strlcpy.c
new file mode 100644
index 0000000..c2a5142
--- /dev/null
+++ b/na/lib/string/strlcpy.c
@@ -0,0 +1,37 @@
+/* SPDX-License-Identifier: MIT */
+/* Derived from musl libc (Copyright © 2005-2020 Rich Felker, et al.) */
+
+#define _BSD_SOURCE
+#include <string.h>
+#include <stdint.h>
+#include <limits.h>
+
+#define ALIGN (sizeof(size_t)-1)
+#define ONES ((size_t)-1/UCHAR_MAX)
+#define HIGHS (ONES * (UCHAR_MAX/2+1))
+#define HASZERO(x) ((x)-ONES & ~(x) & HIGHS)
+
+size_t strlcpy(char *d, const char *s, size_t n)
+{
+	char *d0 = d;
+	size_t *wd;
+
+	if (!n--) goto finish;
+#ifdef __GNUC__
+	typedef size_t __attribute__((__may_alias__)) word;
+	const word *ws;
+	if (((uintptr_t)s & ALIGN) == ((uintptr_t)d & ALIGN)) {
+		for (; ((uintptr_t)s & ALIGN) && n && (*d=*s); n--, s++, d++);
+		if (n && *s) {
+			wd=(void *)d; ws=(const void *)s;
+			for (; n>=sizeof(size_t) && !HASZERO(*ws);
+			       n-=sizeof(size_t), ws++, wd++) *wd = *ws;
+			d=(void *)wd; s=(const void *)ws;
+		}
+	}
+#endif
+	for (; n && (*d=*s); n--, s++, d++);
+	*d = 0;
+finish:
+	return d-d0 + strlen(s);
+}
diff --git a/na/lib/string/strlen.c b/na/lib/string/strlen.c
new file mode 100644
index 0000000..76e14f2
--- /dev/null
+++ b/na/lib/string/strlen.c
@@ -0,0 +1,25 @@
+/* SPDX-License-Identifier: MIT */
+/* Derived from musl libc (Copyright © 2005-2020 Rich Felker, et al.) */
+
+#include <string.h>
+#include <stdint.h>
+#include <limits.h>
+
+#define ALIGN (sizeof(size_t))
+#define ONES ((size_t)-1/UCHAR_MAX)
+#define HIGHS (ONES * (UCHAR_MAX/2+1))
+#define HASZERO(x) ((x)-ONES & ~(x) & HIGHS)
+
+size_t strlen(const char *s)
+{
+	const char *a = s;
+#ifdef __GNUC__
+	typedef size_t __attribute__((__may_alias__)) word;
+	const word *w;
+	for (; (uintptr_t)s % ALIGN; s++) if (!*s) return s-a;
+	for (w = (const void *)s; !HASZERO(*w); w++);
+	s = (const void *)w;
+#endif
+	for (; *s; s++);
+	return s-a;
+}
diff --git a/na/lib/string/strncat.c b/na/lib/string/strncat.c
new file mode 100644
index 0000000..a144e37
--- /dev/null
+++ b/na/lib/string/strncat.c
@@ -0,0 +1,13 @@
+/* SPDX-License-Identifier: MIT */
+/* Derived from musl libc (Copyright © 2005-2020 Rich Felker, et al.) */
+
+#include <string.h>
+
+char *strncat(char *restrict d, const char *restrict s, size_t n)
+{
+	char *a = d;
+	d += strlen(d);
+	while (n && *s) n--, *d++ = *s++;
+	*d++ = 0;
+	return a;
+}
diff --git a/na/lib/string/strncmp.c b/na/lib/string/strncmp.c
new file mode 100644
index 0000000..b98473b
--- /dev/null
+++ b/na/lib/string/strncmp.c
@@ -0,0 +1,12 @@
+/* SPDX-License-Identifier: MIT */
+/* Derived from musl libc (Copyright © 2005-2020 Rich Felker, et al.) */
+
+#include <string.h>
+
+int strncmp(const char *_l, const char *_r, size_t n)
+{
+	const unsigned char *l=(void *)_l, *r=(void *)_r;
+	if (!n--) return 0;
+	for (; *l && *r && n && *l == *r ; l++, r++, n--);
+	return *l - *r;
+}
diff --git a/na/lib/string/strncpy.c b/na/lib/string/strncpy.c
new file mode 100644
index 0000000..a828465
--- /dev/null
+++ b/na/lib/string/strncpy.c
@@ -0,0 +1,10 @@
+/* SPDX-License-Identifier: MIT */
+/* Derived from musl libc (Copyright © 2005-2020 Rich Felker, et al.) */
+
+#include <string.h>
+
+char *strncpy(char *restrict d, const char *restrict s, size_t n)
+{
+	__stpncpy(d, s, n);
+	return d;
+}
diff --git a/na/lib/string/strnlen.c b/na/lib/string/strnlen.c
new file mode 100644
index 0000000..8931c52
--- /dev/null
+++ b/na/lib/string/strnlen.c
@@ -0,0 +1,10 @@
+/* SPDX-License-Identifier: MIT */
+/* Derived from musl libc (Copyright © 2005-2020 Rich Felker, et al.) */
+
+#include <string.h>
+
+size_t strnlen(const char *s, size_t n)
+{
+	const char *p = memchr(s, 0, n);
+	return p ? p-s : n;
+}
diff --git a/na/lib/string/strpbrk.c b/na/lib/string/strpbrk.c
new file mode 100644
index 0000000..82fbc92
--- /dev/null
+++ b/na/lib/string/strpbrk.c
@@ -0,0 +1,10 @@
+/* SPDX-License-Identifier: MIT */
+/* Derived from musl libc (Copyright © 2005-2020 Rich Felker, et al.) */
+
+#include <string.h>
+
+char *strpbrk(const char *s, const char *b)
+{
+	s += strcspn(s, b);
+	return *s ? (char *)s : 0;
+}
diff --git a/na/lib/string/strrchr.c b/na/lib/string/strrchr.c
new file mode 100644
index 0000000..a45b459
--- /dev/null
+++ b/na/lib/string/strrchr.c
@@ -0,0 +1,9 @@
+/* SPDX-License-Identifier: MIT */
+/* Derived from musl libc (Copyright © 2005-2020 Rich Felker, et al.) */
+
+#include <string.h>
+
+char *strrchr(const char *s, int c)
+{
+	return __memrchr(s, c, strlen(s) + 1);
+}
diff --git a/na/lib/string/strsep.c b/na/lib/string/strsep.c
new file mode 100644
index 0000000..2d513f4
--- /dev/null
+++ b/na/lib/string/strsep.c
@@ -0,0 +1,16 @@
+/* SPDX-License-Identifier: MIT */
+/* Derived from musl libc (Copyright © 2005-2020 Rich Felker, et al.) */
+
+#define _GNU_SOURCE
+#include <string.h>
+
+char *strsep(char **str, const char *sep)
+{
+	char *s = *str, *end;
+	if (!s) return NULL;
+	end = s + strcspn(s, sep);
+	if (*end) *end++ = 0;
+	else end = 0;
+	*str = end;
+	return s;
+}
diff --git a/na/lib/string/strspn.c b/na/lib/string/strspn.c
new file mode 100644
index 0000000..f27710d
--- /dev/null
+++ b/na/lib/string/strspn.c
@@ -0,0 +1,23 @@
+/* SPDX-License-Identifier: MIT */
+/* Derived from musl libc (Copyright © 2005-2020 Rich Felker, et al.) */
+
+#include <string.h>
+
+#define BITOP(a,b,op) \
+ ((a)[(size_t)(b)/(8*sizeof *(a))] op (size_t)1<<((size_t)(b)%(8*sizeof *(a))))
+
+size_t strspn(const char *s, const char *c)
+{
+	const char *a = s;
+	size_t byteset[32/sizeof(size_t)] = { 0 };
+
+	if (!c[0]) return 0;
+	if (!c[1]) {
+		for (; *s == *c; s++);
+		return s-a;
+	}
+
+	for (; *c && BITOP(byteset, *(unsigned char *)c, |=); c++);
+	for (; *s && BITOP(byteset, *(unsigned char *)s, &); s++);
+	return s-a;
+}
diff --git a/na/lib/string/strstr.c b/na/lib/string/strstr.c
new file mode 100644
index 0000000..07e2982
--- /dev/null
+++ b/na/lib/string/strstr.c
@@ -0,0 +1,157 @@
+/* SPDX-License-Identifier: MIT */
+/* Derived from musl libc (Copyright © 2005-2020 Rich Felker, et al.) */
+
+#include <string.h>
+#include <stdint.h>
+
+static char *twobyte_strstr(const unsigned char *h, const unsigned char *n)
+{
+	uint16_t nw = n[0]<<8 | n[1], hw = h[0]<<8 | h[1];
+	for (h++; *h && hw != nw; hw = hw<<8 | *++h);
+	return *h ? (char *)h-1 : 0;
+}
+
+static char *threebyte_strstr(const unsigned char *h, const unsigned char *n)
+{
+	uint32_t nw = (uint32_t)n[0]<<24 | n[1]<<16 | n[2]<<8;
+	uint32_t hw = (uint32_t)h[0]<<24 | h[1]<<16 | h[2]<<8;
+	for (h+=2; *h && hw != nw; hw = (hw|*++h)<<8);
+	return *h ? (char *)h-2 : 0;
+}
+
+static char *fourbyte_strstr(const unsigned char *h, const unsigned char *n)
+{
+	uint32_t nw = (uint32_t)n[0]<<24 | n[1]<<16 | n[2]<<8 | n[3];
+	uint32_t hw = (uint32_t)h[0]<<24 | h[1]<<16 | h[2]<<8 | h[3];
+	for (h+=3; *h && hw != nw; hw = hw<<8 | *++h);
+	return *h ? (char *)h-3 : 0;
+}
+
+#define MAX(a,b) ((a)>(b)?(a):(b))
+#define MIN(a,b) ((a)<(b)?(a):(b))
+
+#define BITOP(a,b,op) \
+ ((a)[(size_t)(b)/(8*sizeof *(a))] op (size_t)1<<((size_t)(b)%(8*sizeof *(a))))
+
+static char *twoway_strstr(const unsigned char *h, const unsigned char *n)
+{
+	const unsigned char *z;
+	size_t l, ip, jp, k, p, ms, p0, mem, mem0;
+	size_t byteset[32 / sizeof(size_t)] = { 0 };
+	size_t shift[256];
+
+	/* Computing length of needle and fill shift table */
+	for (l=0; n[l] && h[l]; l++)
+		BITOP(byteset, n[l], |=), shift[n[l]] = l+1;
+	if (n[l]) return 0; /* hit the end of h */
+
+	/* Compute maximal suffix */
+	ip = -1; jp = 0; k = p = 1;
+	while (jp+k<l) {
+		if (n[ip+k] == n[jp+k]) {
+			if (k == p) {
+				jp += p;
+				k = 1;
+			} else k++;
+		} else if (n[ip+k] > n[jp+k]) {
+			jp += k;
+			k = 1;
+			p = jp - ip;
+		} else {
+			ip = jp++;
+			k = p = 1;
+		}
+	}
+	ms = ip;
+	p0 = p;
+
+	/* And with the opposite comparison */
+	ip = -1; jp = 0; k = p = 1;
+	while (jp+k<l) {
+		if (n[ip+k] == n[jp+k]) {
+			if (k == p) {
+				jp += p;
+				k = 1;
+			} else k++;
+		} else if (n[ip+k] < n[jp+k]) {
+			jp += k;
+			k = 1;
+			p = jp - ip;
+		} else {
+			ip = jp++;
+			k = p = 1;
+		}
+	}
+	if (ip+1 > ms+1) ms = ip;
+	else p = p0;
+
+	/* Periodic needle? */
+	if (memcmp(n, n+p, ms+1)) {
+		mem0 = 0;
+		p = MAX(ms, l-ms-1) + 1;
+	} else mem0 = l-p;
+	mem = 0;
+
+	/* Initialize incremental end-of-haystack pointer */
+	z = h;
+
+	/* Search loop */
+	for (;;) {
+		/* Update incremental end-of-haystack pointer */
+		if (z-h < l) {
+			/* Fast estimate for MAX(l,63) */
+			size_t grow = l | 63;
+			const unsigned char *z2 = memchr(z, 0, grow);
+			if (z2) {
+				z = z2;
+				if (z-h < l) return 0;
+			} else z += grow;
+		}
+
+		/* Check last byte first; advance by shift on mismatch */
+		if (BITOP(byteset, h[l-1], &)) {
+			k = l-shift[h[l-1]];
+			if (k) {
+				if (k < mem) k = mem;
+				h += k;
+				mem = 0;
+				continue;
+			}
+		} else {
+			h += l;
+			mem = 0;
+			continue;
+		}
+
+		/* Compare right half */
+		for (k=MAX(ms+1,mem); n[k] && n[k] == h[k]; k++);
+		if (n[k]) {
+			h += k-ms;
+			mem = 0;
+			continue;
+		}
+		/* Compare left half */
+		for (k=ms+1; k>mem && n[k-1] == h[k-1]; k--);
+		if (k <= mem) return (char *)h;
+		h += p;
+		mem = mem0;
+	}
+}
+
+char *strstr(const char *h, const char *n)
+{
+	/* Return immediately on empty needle */
+	if (!n[0]) return (char *)h;
+
+	/* Use faster algorithms for short needles */
+	h = strchr(h, *n);
+	if (!h || !n[1]) return (char *)h;
+	if (!h[1]) return 0;
+	if (!n[2]) return twobyte_strstr((void *)h, (void *)n);
+	if (!h[2]) return 0;
+	if (!n[3]) return threebyte_strstr((void *)h, (void *)n);
+	if (!h[3]) return 0;
+	if (!n[4]) return fourbyte_strstr((void *)h, (void *)n);
+
+	return twoway_strstr((void *)h, (void *)n);
+}
diff --git a/na/lib/string/strtok.c b/na/lib/string/strtok.c
new file mode 100644
index 0000000..86d9c97
--- /dev/null
+++ b/na/lib/string/strtok.c
@@ -0,0 +1,16 @@
+/* SPDX-License-Identifier: MIT */
+/* Derived from musl libc (Copyright © 2005-2020 Rich Felker, et al.) */
+
+#include <string.h>
+
+char *strtok(char *restrict s, const char *restrict sep)
+{
+	static char *p;
+	if (!s && !(s = p)) return NULL;
+	s += strspn(s, sep);
+	if (!*s) return p = 0;
+	p = s + strcspn(s, sep);
+	if (*p) *p++ = 0;
+	else p = 0;
+	return s;
+}
diff --git a/na/lib/string/strtok_r.c b/na/lib/string/strtok_r.c
new file mode 100644
index 0000000..a702703
--- /dev/null
+++ b/na/lib/string/strtok_r.c
@@ -0,0 +1,15 @@
+/* SPDX-License-Identifier: MIT */
+/* Derived from musl libc (Copyright © 2005-2020 Rich Felker, et al.) */
+
+#include <string.h>
+
+char *strtok_r(char *restrict s, const char *restrict sep, char **restrict p)
+{
+	if (!s && !(s = *p)) return NULL;
+	s += strspn(s, sep);
+	if (!*s) return *p = 0;
+	*p = s + strcspn(s, sep);
+	if (**p) *(*p)++ = 0;
+	else *p = 0;
+	return s;
+}
diff --git a/na/lib/string/x86_64/memcpy.s b/na/lib/string/x86_64/memcpy.s
new file mode 100644
index 0000000..3d960ef
--- /dev/null
+++ b/na/lib/string/x86_64/memcpy.s
@@ -0,0 +1,25 @@
+.global memcpy
+.global __memcpy_fwd
+.hidden __memcpy_fwd
+.type memcpy,@function
+memcpy:
+__memcpy_fwd:
+	mov %rdi,%rax
+	cmp $8,%rdx
+	jc 1f
+	test $7,%edi
+	jz 1f
+2:	movsb
+	dec %rdx
+	test $7,%edi
+	jnz 2b
+1:	mov %rdx,%rcx
+	shr $3,%rcx
+	rep
+	movsq
+	and $7,%edx
+	jz 1f
+2:	movsb
+	dec %edx
+	jnz 2b
+1:	ret
diff --git a/na/lib/string/x86_64/memmove.s b/na/lib/string/x86_64/memmove.s
new file mode 100644
index 0000000..172c025
--- /dev/null
+++ b/na/lib/string/x86_64/memmove.s
@@ -0,0 +1,16 @@
+.global memmove
+.type memmove,@function
+memmove:
+	mov %rdi,%rax
+	sub %rsi,%rax
+	cmp %rdx,%rax
+.hidden __memcpy_fwd
+	jae __memcpy_fwd
+	mov %rdx,%rcx
+	lea -1(%rdi,%rdx),%rdi
+	lea -1(%rsi,%rdx),%rsi
+	std
+	rep movsb
+	cld
+	lea 1(%rdi),%rax
+	ret
diff --git a/na/lib/string/x86_64/memset.s b/na/lib/string/x86_64/memset.s
new file mode 100644
index 0000000..2d3f5e5
--- /dev/null
+++ b/na/lib/string/x86_64/memset.s
@@ -0,0 +1,72 @@
+.global memset
+.type memset,@function
+memset:
+	movzbq %sil,%rax
+	mov $0x101010101010101,%r8
+	imul %r8,%rax
+
+	cmp $126,%rdx
+	ja 2f
+
+	test %edx,%edx
+	jz 1f
+
+	mov %sil,(%rdi)
+	mov %sil,-1(%rdi,%rdx)
+	cmp $2,%edx
+	jbe 1f
+
+	mov %ax,1(%rdi)
+	mov %ax,(-1-2)(%rdi,%rdx)
+	cmp $6,%edx
+	jbe 1f
+
+	mov %eax,(1+2)(%rdi)
+	mov %eax,(-1-2-4)(%rdi,%rdx)
+	cmp $14,%edx
+	jbe 1f
+
+	mov %rax,(1+2+4)(%rdi)
+	mov %rax,(-1-2-4-8)(%rdi,%rdx)
+	cmp $30,%edx
+	jbe 1f
+
+	mov %rax,(1+2+4+8)(%rdi)
+	mov %rax,(1+2+4+8+8)(%rdi)
+	mov %rax,(-1-2-4-8-16)(%rdi,%rdx)
+	mov %rax,(-1-2-4-8-8)(%rdi,%rdx)
+	cmp $62,%edx
+	jbe 1f
+
+	mov %rax,(1+2+4+8+16)(%rdi)
+	mov %rax,(1+2+4+8+16+8)(%rdi)
+	mov %rax,(1+2+4+8+16+16)(%rdi)
+	mov %rax,(1+2+4+8+16+24)(%rdi)
+	mov %rax,(-1-2-4-8-16-32)(%rdi,%rdx)
+	mov %rax,(-1-2-4-8-16-24)(%rdi,%rdx)
+	mov %rax,(-1-2-4-8-16-16)(%rdi,%rdx)
+	mov %rax,(-1-2-4-8-16-8)(%rdi,%rdx)
+
+1:	mov %rdi,%rax
+	ret
+
+2:	test $15,%edi
+	mov %rdi,%r8
+	mov %rax,-8(%rdi,%rdx)
+	mov %rdx,%rcx
+	jnz 2f
+
+1:	shr $3,%rcx
+	rep
+	stosq
+	mov %r8,%rax
+	ret
+
+2:	xor %edx,%edx
+	sub %edi,%edx
+	and $15,%edx
+	mov %rax,(%rdi)
+	mov %rax,8(%rdi)
+	sub %rdx,%rcx
+	add %rdx,%rdi
+	jmp 1b