app-backup/tarsnap: backport strict aliasing/UB fix

Doing this in a new revision to allow easy testing compared w/ 1.0.40.

Signed-off-by: Sam James <sam@gentoo.org>

2 files changed, 215 insertions, 0 deletions

diff --git a/app-backup/tarsnap/files/tarsnap-1.0.40-strict-aliasing-fix.patch b/app-backup/tarsnap/files/tarsnap-1.0.40-strict-aliasing-fix.patch
new file mode 100644
index 000000000000..d168d99222e8
--- /dev/null
+++ b/app-backup/tarsnap/files/tarsnap-1.0.40-strict-aliasing-fix.patch
@@ -0,0 +1,158 @@
+https://github.com/Tarsnap/tarsnap/commit/ca40c06f290fb8298dc2e583303d45b58878f37b
+https://github.com/Tarsnap/tarsnap/commit/4af6d8350ab53d0f1f3104ce3d9072c2d5f9ef7a
+
+From 4af6d8350ab53d0f1f3104ce3d9072c2d5f9ef7a Mon Sep 17 00:00:00 2001
+From: Graham Percival <gperciva@tarsnap.com>
+Date: Fri, 1 Apr 2022 16:58:43 -0700
+Subject: [PATCH] scrypt: Fix strict aliasing
+
+The original scrypt code treated its data as blobs of bytes, accessing
+them in whatever manner was convenient from time to time:
+
+* as 32-bit words or vectors thereof for the purpose of the Salsa20/8 core
+* in machine-word-sized chunks (aka. size_t) for block copy and xor operations
+* as 32-bit words for the Integerify function.
+
+This worked fine at the time, but newer compilers apply strict aliasing rules
+which allow them to assume that e.g. data accessed as a uint32_t is not the
+same as data accessed as a size_t, resulting in miscompilation.
+
+Note that in recent versions of scrypt (after 2015-07-18; versions 1.2.0 and
+later) such miscompilation should be detected by the built-in runtime testing.
+
+To avoid aliasing problems, the generic scrypt code now operates on uint32_t
+throughout while the SSE2-enabled scrypt code operates on __m128i throughout.
+
+Experimentally, we found that memcpy() speeds up blkcpy() in the plain C
+case, but slowed it down in the _sse2.c case (probably because memcpy
+can make use of vector instructions internally, but cannot assume that
+it will always have a multiple of 16 bytes, as we do).
+--- a/lib/crypto/crypto_scrypt_smix.c
++++ b/lib/crypto/crypto_scrypt_smix.c
+@@ -27,39 +27,32 @@
+  * online backup system.
+  */
+ #include <stdint.h>
++#include <string.h>
+ 
+ #include "sysendian.h"
+ 
+ #include "crypto_scrypt_smix.h"
+ 
+-static void blkcpy(void *, const void *, size_t);
+-static void blkxor(void *, const void *, size_t);
++static void blkcpy(uint32_t *, const uint32_t *, size_t);
++static void blkxor(uint32_t *, const uint32_t *, size_t);
+ static void salsa20_8(uint32_t[16]);
+ static void blockmix_salsa8(const uint32_t *, uint32_t *, uint32_t *, size_t);
+-static uint64_t integerify(const void *, size_t);
++static uint64_t integerify(const uint32_t *, size_t);
+ 
+ static void
+-blkcpy(void * dest, const void * src, size_t len)
++blkcpy(uint32_t * dest, const uint32_t * src, size_t len)
+ {
+-	size_t * D = dest;
+-	const size_t * S = src;
+-	size_t L = len / sizeof(size_t);
+-	size_t i;
+ 
+-	for (i = 0; i < L; i++)
+-		D[i] = S[i];
++	memcpy(dest, src, len);
+ }
+ 
+ static void
+-blkxor(void * dest, const void * src, size_t len)
++blkxor(uint32_t * dest, const uint32_t * src, size_t len)
+ {
+-	size_t * D = dest;
+-	const size_t * S = src;
+-	size_t L = len / sizeof(size_t);
+ 	size_t i;
+ 
+-	for (i = 0; i < L; i++)
+-		D[i] ^= S[i];
++	for (i = 0; i < len / 4; i++)
++		dest[i] ^= src[i];
+ }
+ 
+ /**
+@@ -145,9 +138,9 @@ blockmix_salsa8(const uint32_t * Bin, uint32_t * Bout, uint32_t * X, size_t r)
+  * Return the result of parsing B_{2r-1} as a little-endian integer.
+  */
+ static uint64_t
+-integerify(const void * B, size_t r)
++integerify(const uint32_t * B, size_t r)
+ {
+-	const uint32_t * X = (const void *)((uintptr_t)(B) + (2 * r - 1) * 64);
++	const uint32_t * X = B + (2 * r - 1) * 16;
+ 
+ 	return (((uint64_t)(X[1]) << 32) + X[0]);
+ }
+--- a/lib/crypto/crypto_scrypt_smix_sse2.c
++++ b/lib/crypto/crypto_scrypt_smix_sse2.c
+@@ -36,34 +36,30 @@
+ 
+ #include "crypto_scrypt_smix_sse2.h"
+ 
+-static void blkcpy(void *, const void *, size_t);
+-static void blkxor(void *, const void *, size_t);
++static void blkcpy(__m128i *, const __m128i *, size_t);
++static void blkxor(__m128i *, const __m128i *, size_t);
+ static void salsa20_8(__m128i[4]);
+ static void blockmix_salsa8(const __m128i *, __m128i *, __m128i *, size_t);
+-static uint64_t integerify(const void *, size_t);
++static uint64_t integerify(const __m128i *, size_t);
+ 
+ static void
+-blkcpy(void * dest, const void * src, size_t len)
++blkcpy(__m128i * dest, const __m128i * src, size_t len)
+ {
+-	__m128i * D = dest;
+-	const __m128i * S = src;
+ 	size_t L = len / 16;
+ 	size_t i;
+ 
+ 	for (i = 0; i < L; i++)
+-		D[i] = S[i];
++		dest[i] = src[i];
+ }
+ 
+ static void
+-blkxor(void * dest, const void * src, size_t len)
++blkxor(__m128i * dest, const __m128i * src, size_t len)
+ {
+-	__m128i * D = dest;
+-	const __m128i * S = src;
+ 	size_t L = len / 16;
+ 	size_t i;
+ 
+ 	for (i = 0; i < L; i++)
+-		D[i] = _mm_xor_si128(D[i], S[i]);
++		dest[i] = _mm_xor_si128(dest[i], src[i]);
+ }
+ 
+ /**
+@@ -168,11 +164,18 @@ blockmix_salsa8(const __m128i * Bin, __m128i * Bout, __m128i * X, size_t r)
+  * Note that B's layout is permuted compared to the generic implementation.
+  */
+ static uint64_t
+-integerify(const void * B, size_t r)
++integerify(const __m128i * B, size_t r)
+ {
+-	const uint32_t * X = (const void *)((uintptr_t)(B) + (2 * r - 1) * 64);
++	const __m128i * X = B + (2*r - 1) * 4;
++	uint32_t X0, X13;
+ 
+-	return (((uint64_t)(X[13]) << 32) + X[0]);
++	/* Get the first 32-bit element in X[0]. */
++	X0 = (uint32_t)_mm_cvtsi128_si32(X[0]);
++
++	/* Get the second 32-bit element in X[3]. */
++	X13 = (uint32_t)_mm_cvtsi128_si32(_mm_srli_si128(X[3], 4));
++
++	return (((uint64_t)(X13) << 32) + X0);
+ }
+ 
+ /**
diff --git a/app-backup/tarsnap/tarsnap-1.0.40-r1.ebuild b/app-backup/tarsnap/tarsnap-1.0.40-r1.ebuild
new file mode 100644
index 000000000000..3adbe02afcf0
--- /dev/null
+++ b/app-backup/tarsnap/tarsnap-1.0.40-r1.ebuild
@@ -0,0 +1,57 @@
+# Copyright 1999-2023 Gentoo Authors
+# Distributed under the terms of the GNU General Public License v2
+
+EAPI=8
+
+inherit bash-completion-r1 toolchain-funcs
+
+DESCRIPTION="Online backups for the truly paranoid"
+HOMEPAGE="https://www.tarsnap.com/"
+SRC_URI="https://www.tarsnap.com/download/${PN}-autoconf-${PV}.tgz"
+S="${WORKDIR}"/${PN}-autoconf-${PV}
+
+LICENSE="tarsnap"
+SLOT="0"
+KEYWORDS="~amd64 ~x86"
+IUSE="acl bzip2 lzma xattr"
+
+RDEPEND="
+	dev-libs/openssl:=
+	sys-fs/e2fsprogs
+	sys-libs/zlib
+	acl? ( sys-apps/acl )
+	bzip2? ( app-arch/bzip2 )
+	lzma? ( app-arch/xz-utils )
+	xattr? ( sys-apps/attr )
+"
+# Required for "magic.h"
+DEPEND="
+	${RDEPEND}
+	virtual/os-headers
+"
+
+PATCHES=(
+	"${FILESDIR}"/${PN}-1.0.39-respect-AR.patch
+	"${FILESDIR}"/${P}-strict-aliasing-fix.patch
+)
+
+src_configure() {
+	local myeconfargs=(
+		$(use_enable xattr)
+		$(use_enable acl)
+		$(use_with bzip2 bz2lib)
+		--without-lzmadec
+		$(use_with lzma)
+	)
+
+	econf "${myeconfargs[@]}"
+}
+
+src_compile() {
+	emake AR="$(tc-getAR)"
+}
+
+src_install() {
+	default
+	dobashcomp misc/bash_completion.d/*
+}