Merge pull request #31309 from Civil/riscv

Initial support for risc-v
This commit is contained in:
alexey-milovidov 2021-11-12 03:41:57 +03:00 committed by GitHub
commit b1f177e2ae
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
24 changed files with 441 additions and 21 deletions

View File

@ -28,8 +28,8 @@
#define NO_INLINE __attribute__((__noinline__))
#define MAY_ALIAS __attribute__((__may_alias__))
#if !defined(__x86_64__) && !defined(__aarch64__) && !defined(__PPC__)
# error "The only supported platforms are x86_64 and AArch64, PowerPC (work in progress)"
#if !defined(__x86_64__) && !defined(__aarch64__) && !defined(__PPC__) && !(defined(__riscv) && (__riscv_xlen == 64))
# error "The only supported platforms are x86_64 and AArch64, PowerPC (work in progress) and RISC-V 64 (experimental)"
#endif
/// Check for presence of address sanitizer

View File

@ -16,3 +16,7 @@ endif ()
if (CMAKE_SYSTEM_PROCESSOR MATCHES "^(ppc64le.*|PPC64LE.*)")
set (ARCH_PPC64LE 1)
endif ()
if (CMAKE_SYSTEM_PROCESSOR MATCHES "riscv64")
set (ARCH_RISCV64 1)
endif ()

View File

@ -20,10 +20,10 @@ set (CMAKE_C_FLAGS_INIT "${CMAKE_C_FLAGS} --gcc-toolchain=${TOOLCHAIN_PATH}")
set (CMAKE_CXX_FLAGS_INIT "${CMAKE_CXX_FLAGS} --gcc-toolchain=${TOOLCHAIN_PATH}")
set (CMAKE_ASM_FLAGS_INIT "${CMAKE_ASM_FLAGS} --gcc-toolchain=${TOOLCHAIN_PATH}")
set (LINKER_NAME "ld.lld" CACHE STRING "" FORCE)
set (LINKER_NAME "riscv64-linux-gnu-ld.bfd" CACHE STRING "" FORCE)
set (CMAKE_EXE_LINKER_FLAGS_INIT "-fuse-ld=lld")
set (CMAKE_SHARED_LINKER_FLAGS_INIT "-fuse-ld=lld")
set (CMAKE_EXE_LINKER_FLAGS_INIT "-fuse-ld=bfd")
set (CMAKE_SHARED_LINKER_FLAGS_INIT "-fuse-ld=bfd")
set (HAS_PRE_1970_EXITCODE "0" CACHE STRING "Result from TRY_RUN" FORCE)
set (HAS_PRE_1970_EXITCODE__TRYRUN_OUTPUT "" CACHE STRING "Output from TRY_RUN" FORCE)

2
contrib/boost vendored

@ -1 +1 @@
Subproject commit 79358a3106aab6af464430ed67c7efafebf5cd6f
Subproject commit fcb058e1459ac273ecfe7cdf72791cb1479115af

View File

@ -196,6 +196,12 @@ if (NOT EXTERNAL_BOOST_FOUND)
"${LIBRARY_DIR}/libs/context/src/asm/make_ppc64_sysv_elf_gas.S"
"${LIBRARY_DIR}/libs/context/src/asm/ontop_ppc64_sysv_elf_gas.S"
)
elseif (ARCH_RISCV64)
set (SRCS_CONTEXT ${SRCS_CONTEXT}
"${LIBRARY_DIR}/libs/context/src/asm/jump_riscv64_sysv_elf_gas.S"
"${LIBRARY_DIR}/libs/context/src/asm/make_riscv64_sysv_elf_gas.S"
"${LIBRARY_DIR}/libs/context/src/asm/ontop_riscv64_sysv_elf_gas.S"
)
elseif(OS_DARWIN)
set (SRCS_CONTEXT ${SRCS_CONTEXT}
"${LIBRARY_DIR}/libs/context/src/asm/jump_x86_64_sysv_macho_gas.S"

View File

@ -1,5 +1,5 @@
if (SANITIZE OR NOT (
((OS_LINUX OR OS_FREEBSD) AND (ARCH_AMD64 OR ARCH_ARM OR ARCH_PPC64LE)) OR
((OS_LINUX OR OS_FREEBSD) AND (ARCH_AMD64 OR ARCH_ARM OR ARCH_PPC64LE OR ARCH_RISCV64)) OR
(OS_DARWIN AND (CMAKE_BUILD_TYPE STREQUAL "RelWithDebInfo" OR CMAKE_BUILD_TYPE STREQUAL "Debug"))
))
if (ENABLE_JEMALLOC)
@ -112,6 +112,8 @@ elseif (ARCH_ARM)
set(JEMALLOC_INCLUDE_PREFIX "${JEMALLOC_INCLUDE_PREFIX}_aarch64")
elseif (ARCH_PPC64LE)
set(JEMALLOC_INCLUDE_PREFIX "${JEMALLOC_INCLUDE_PREFIX}_ppc64le")
elseif (ARCH_RISCV64)
set(JEMALLOC_INCLUDE_PREFIX "${JEMALLOC_INCLUDE_PREFIX}_riscv64")
else ()
message (FATAL_ERROR "internal jemalloc: This arch is not supported")
endif ()

View File

@ -0,0 +1,8 @@
Here are pre-generated files from jemalloc on Linux risc-v.
You can obtain these files by running ./autogen.sh inside jemalloc source directory.
Added #define GNU_SOURCE
Added JEMALLOC_OVERRIDE___POSIX_MEMALIGN because why not.
Removed JEMALLOC_HAVE_ATTR_FORMAT_GNU_PRINTF because it's non standard.
Removed JEMALLOC_PURGE_MADVISE_FREE because it's available only from Linux 4.5.
Added JEMALLOC_CONFIG_MALLOC_CONF substitution

View File

@ -0,0 +1,367 @@
/* include/jemalloc/internal/jemalloc_internal_defs.h. Generated from jemalloc_internal_defs.h.in by configure. */
#ifndef JEMALLOC_INTERNAL_DEFS_H_
#define JEMALLOC_INTERNAL_DEFS_H_
/*
* If JEMALLOC_PREFIX is defined via --with-jemalloc-prefix, it will cause all
* public APIs to be prefixed. This makes it possible, with some care, to use
* multiple allocators simultaneously.
*/
/* #undef JEMALLOC_PREFIX */
/* #undef JEMALLOC_CPREFIX */
/*
* Define overrides for non-standard allocator-related functions if they are
* present on the system.
*/
#define JEMALLOC_OVERRIDE___LIBC_CALLOC
#define JEMALLOC_OVERRIDE___LIBC_FREE
#define JEMALLOC_OVERRIDE___LIBC_MALLOC
#define JEMALLOC_OVERRIDE___LIBC_MEMALIGN
#define JEMALLOC_OVERRIDE___LIBC_REALLOC
#define JEMALLOC_OVERRIDE___LIBC_VALLOC
/* #undef JEMALLOC_OVERRIDE___POSIX_MEMALIGN */
/*
* JEMALLOC_PRIVATE_NAMESPACE is used as a prefix for all library-private APIs.
* For shared libraries, symbol visibility mechanisms prevent these symbols
* from being exported, but for static libraries, naming collisions are a real
* possibility.
*/
#define JEMALLOC_PRIVATE_NAMESPACE je_
/*
* Hyper-threaded CPUs may need a special instruction inside spin loops in
* order to yield to another virtual CPU.
*/
#define CPU_SPINWAIT
/* 1 if CPU_SPINWAIT is defined, 0 otherwise. */
#define HAVE_CPU_SPINWAIT 0
/*
* Number of significant bits in virtual addresses. This may be less than the
* total number of bits in a pointer, e.g. on x64, for which the uppermost 16
* bits are the same as bit 47.
*/
#define LG_VADDR 48
/* Defined if C11 atomics are available. */
#define JEMALLOC_C11_ATOMICS 1
/* Defined if GCC __atomic atomics are available. */
#define JEMALLOC_GCC_ATOMIC_ATOMICS 1
/* and the 8-bit variant support. */
#define JEMALLOC_GCC_U8_ATOMIC_ATOMICS 1
/* Defined if GCC __sync atomics are available. */
#define JEMALLOC_GCC_SYNC_ATOMICS 1
/* and the 8-bit variant support. */
#define JEMALLOC_GCC_U8_SYNC_ATOMICS 1
/*
* Defined if __builtin_clz() and __builtin_clzl() are available.
*/
#define JEMALLOC_HAVE_BUILTIN_CLZ
/*
* Defined if os_unfair_lock_*() functions are available, as provided by Darwin.
*/
/* #undef JEMALLOC_OS_UNFAIR_LOCK */
/* Defined if syscall(2) is usable. */
#define JEMALLOC_USE_SYSCALL
/*
* Defined if secure_getenv(3) is available.
*/
// #define JEMALLOC_HAVE_SECURE_GETENV
/*
* Defined if issetugid(2) is available.
*/
/* #undef JEMALLOC_HAVE_ISSETUGID */
/* Defined if pthread_atfork(3) is available. */
#define JEMALLOC_HAVE_PTHREAD_ATFORK
/* Defined if pthread_setname_np(3) is available. */
#define JEMALLOC_HAVE_PTHREAD_SETNAME_NP
/*
* Defined if clock_gettime(CLOCK_MONOTONIC_COARSE, ...) is available.
*/
#define JEMALLOC_HAVE_CLOCK_MONOTONIC_COARSE 1
/*
* Defined if clock_gettime(CLOCK_MONOTONIC, ...) is available.
*/
#define JEMALLOC_HAVE_CLOCK_MONOTONIC 1
/*
* Defined if mach_absolute_time() is available.
*/
/* #undef JEMALLOC_HAVE_MACH_ABSOLUTE_TIME */
/*
* Defined if _malloc_thread_cleanup() exists. At least in the case of
* FreeBSD, pthread_key_create() allocates, which if used during malloc
* bootstrapping will cause recursion into the pthreads library. Therefore, if
* _malloc_thread_cleanup() exists, use it as the basis for thread cleanup in
* malloc_tsd.
*/
/* #undef JEMALLOC_MALLOC_THREAD_CLEANUP */
/*
* Defined if threaded initialization is known to be safe on this platform.
* Among other things, it must be possible to initialize a mutex without
* triggering allocation in order for threaded allocation to be safe.
*/
#define JEMALLOC_THREADED_INIT
/*
* Defined if the pthreads implementation defines
* _pthread_mutex_init_calloc_cb(), in which case the function is used in order
* to avoid recursive allocation during mutex initialization.
*/
/* #undef JEMALLOC_MUTEX_INIT_CB */
/* Non-empty if the tls_model attribute is supported. */
#define JEMALLOC_TLS_MODEL __attribute__((tls_model("initial-exec")))
/*
* JEMALLOC_DEBUG enables assertions and other sanity checks, and disables
* inline functions.
*/
/* #undef JEMALLOC_DEBUG */
/* JEMALLOC_STATS enables statistics calculation. */
#define JEMALLOC_STATS
/* JEMALLOC_EXPERIMENTAL_SMALLOCX_API enables experimental smallocx API. */
/* #undef JEMALLOC_EXPERIMENTAL_SMALLOCX_API */
/* JEMALLOC_PROF enables allocation profiling. */
/* #undef JEMALLOC_PROF */
/* Use libunwind for profile backtracing if defined. */
/* #undef JEMALLOC_PROF_LIBUNWIND */
/* Use libgcc for profile backtracing if defined. */
/* #undef JEMALLOC_PROF_LIBGCC */
/* Use gcc intrinsics for profile backtracing if defined. */
/* #undef JEMALLOC_PROF_GCC */
/*
* JEMALLOC_DSS enables use of sbrk(2) to allocate extents from the data storage
* segment (DSS).
*/
#define JEMALLOC_DSS
/* Support memory filling (junk/zero). */
#define JEMALLOC_FILL
/* Support utrace(2)-based tracing. */
/* #undef JEMALLOC_UTRACE */
/* Support optional abort() on OOM. */
/* #undef JEMALLOC_XMALLOC */
/* Support lazy locking (avoid locking unless a second thread is launched). */
/* #undef JEMALLOC_LAZY_LOCK */
/*
* Minimum allocation alignment is 2^LG_QUANTUM bytes (ignoring tiny size
* classes).
*/
/* #undef LG_QUANTUM */
/* One page is 2^LG_PAGE bytes. */
#define LG_PAGE 16
/*
* One huge page is 2^LG_HUGEPAGE bytes. Note that this is defined even if the
* system does not explicitly support huge pages; system calls that require
* explicit huge page support are separately configured.
*/
#define LG_HUGEPAGE 29
/*
* If defined, adjacent virtual memory mappings with identical attributes
* automatically coalesce, and they fragment when changes are made to subranges.
* This is the normal order of things for mmap()/munmap(), but on Windows
* VirtualAlloc()/VirtualFree() operations must be precisely matched, i.e.
* mappings do *not* coalesce/fragment.
*/
#define JEMALLOC_MAPS_COALESCE
/*
* If defined, retain memory for later reuse by default rather than using e.g.
* munmap() to unmap freed extents. This is enabled on 64-bit Linux because
* common sequences of mmap()/munmap() calls will cause virtual memory map
* holes.
*/
#define JEMALLOC_RETAIN
/* TLS is used to map arenas and magazine caches to threads. */
#define JEMALLOC_TLS
/*
* Used to mark unreachable code to quiet "end of non-void" compiler warnings.
* Don't use this directly; instead use unreachable() from util.h
*/
#define JEMALLOC_INTERNAL_UNREACHABLE __builtin_unreachable
/*
* ffs*() functions to use for bitmapping. Don't use these directly; instead,
* use ffs_*() from util.h.
*/
#define JEMALLOC_INTERNAL_FFSLL __builtin_ffsll
#define JEMALLOC_INTERNAL_FFSL __builtin_ffsl
#define JEMALLOC_INTERNAL_FFS __builtin_ffs
/*
* popcount*() functions to use for bitmapping.
*/
#define JEMALLOC_INTERNAL_POPCOUNTL __builtin_popcountl
#define JEMALLOC_INTERNAL_POPCOUNT __builtin_popcount
/*
* If defined, explicitly attempt to more uniformly distribute large allocation
* pointer alignments across all cache indices.
*/
#define JEMALLOC_CACHE_OBLIVIOUS
/*
* If defined, enable logging facilities. We make this a configure option to
* avoid taking extra branches everywhere.
*/
/* #undef JEMALLOC_LOG */
/*
* If defined, use readlinkat() (instead of readlink()) to follow
* /etc/malloc_conf.
*/
/* #undef JEMALLOC_READLINKAT */
/*
* Darwin (OS X) uses zones to work around Mach-O symbol override shortcomings.
*/
/* #undef JEMALLOC_ZONE */
/*
* Methods for determining whether the OS overcommits.
* JEMALLOC_PROC_SYS_VM_OVERCOMMIT_MEMORY: Linux's
* /proc/sys/vm.overcommit_memory file.
* JEMALLOC_SYSCTL_VM_OVERCOMMIT: FreeBSD's vm.overcommit sysctl.
*/
/* #undef JEMALLOC_SYSCTL_VM_OVERCOMMIT */
#define JEMALLOC_PROC_SYS_VM_OVERCOMMIT_MEMORY
/* Defined if madvise(2) is available. */
#define JEMALLOC_HAVE_MADVISE
/*
* Defined if transparent huge pages are supported via the MADV_[NO]HUGEPAGE
* arguments to madvise(2).
*/
#define JEMALLOC_HAVE_MADVISE_HUGE
/*
* Methods for purging unused pages differ between operating systems.
*
* madvise(..., MADV_FREE) : This marks pages as being unused, such that they
* will be discarded rather than swapped out.
* madvise(..., MADV_DONTNEED) : If JEMALLOC_PURGE_MADVISE_DONTNEED_ZEROS is
* defined, this immediately discards pages,
* such that new pages will be demand-zeroed if
* the address region is later touched;
* otherwise this behaves similarly to
* MADV_FREE, though typically with higher
* system overhead.
*/
#define JEMALLOC_PURGE_MADVISE_FREE
#define JEMALLOC_PURGE_MADVISE_DONTNEED
#define JEMALLOC_PURGE_MADVISE_DONTNEED_ZEROS
/* Defined if madvise(2) is available but MADV_FREE is not (x86 Linux only). */
/* #undef JEMALLOC_DEFINE_MADVISE_FREE */
/*
* Defined if MADV_DO[NT]DUMP is supported as an argument to madvise.
*/
#define JEMALLOC_MADVISE_DONTDUMP
/*
* Defined if transparent huge pages (THPs) are supported via the
* MADV_[NO]HUGEPAGE arguments to madvise(2), and THP support is enabled.
*/
/* #undef JEMALLOC_THP */
/* Define if operating system has alloca.h header. */
#define JEMALLOC_HAS_ALLOCA_H 1
/* C99 restrict keyword supported. */
#define JEMALLOC_HAS_RESTRICT 1
/* For use by hash code. */
/* #undef JEMALLOC_BIG_ENDIAN */
/* sizeof(int) == 2^LG_SIZEOF_INT. */
#define LG_SIZEOF_INT 2
/* sizeof(long) == 2^LG_SIZEOF_LONG. */
#define LG_SIZEOF_LONG 3
/* sizeof(long long) == 2^LG_SIZEOF_LONG_LONG. */
#define LG_SIZEOF_LONG_LONG 3
/* sizeof(intmax_t) == 2^LG_SIZEOF_INTMAX_T. */
#define LG_SIZEOF_INTMAX_T 3
/* glibc malloc hooks (__malloc_hook, __realloc_hook, __free_hook). */
#define JEMALLOC_GLIBC_MALLOC_HOOK
/* glibc memalign hook. */
#define JEMALLOC_GLIBC_MEMALIGN_HOOK
/* pthread support */
#define JEMALLOC_HAVE_PTHREAD
/* dlsym() support */
#define JEMALLOC_HAVE_DLSYM
/* Adaptive mutex support in pthreads. */
#define JEMALLOC_HAVE_PTHREAD_MUTEX_ADAPTIVE_NP
/* GNU specific sched_getcpu support */
#define JEMALLOC_HAVE_SCHED_GETCPU
/* GNU specific sched_setaffinity support */
#define JEMALLOC_HAVE_SCHED_SETAFFINITY
/*
* If defined, all the features necessary for background threads are present.
*/
#define JEMALLOC_BACKGROUND_THREAD 1
/*
* If defined, jemalloc symbols are not exported (doesn't work when
* JEMALLOC_PREFIX is not defined).
*/
/* #undef JEMALLOC_EXPORT */
/* config.malloc_conf options string. */
#define JEMALLOC_CONFIG_MALLOC_CONF "@JEMALLOC_CONFIG_MALLOC_CONF@"
/* If defined, jemalloc takes the malloc/free/etc. symbol names. */
#define JEMALLOC_IS_MALLOC 1
/*
* Defined if strerror_r returns char * if _GNU_SOURCE is defined.
*/
#define JEMALLOC_STRERROR_R_RETURNS_CHAR_WITH_GNU_SOURCE
/* Performs additional safety checks when defined. */
/* #undef JEMALLOC_OPT_SAFETY_CHECKS */
#endif /* JEMALLOC_INTERNAL_DEFS_H_ */

View File

@ -66,7 +66,7 @@
#cmakedefine WITH_SASL_OAUTHBEARER 1
#cmakedefine WITH_SASL_CYRUS 1
// crc32chw
#if !defined(__PPC__) && (!defined(__aarch64__) || defined(__ARM_FEATURE_CRC32)) && !(defined(__aarch64__) && defined(__APPLE__))
#if !defined(__PPC__) && !defined(__riscv) && (!defined(__aarch64__) || defined(__ARM_FEATURE_CRC32)) && !(defined(__aarch64__) && defined(__APPLE__))
#define WITH_CRC32C_HW 1
#endif
// regex

2
contrib/poco vendored

@ -1 +1 @@
Subproject commit 39fd359765a3a77b46d94ec3c5def3c7802a920f
Subproject commit 173fb31717837d366152c508619b09dcf11786da

2
contrib/rocksdb vendored

@ -1 +1 @@
Subproject commit 296c1b8b95fd448b8097a1b2cc9f704ff4a73a2c
Subproject commit e7c2b2f7bcf3b4b33892a1a6d25c32a93edfbdb9

@ -1 +1 @@
Subproject commit 94644e92f0a3ff14bd35ed902a8622a2d15f7be4
Subproject commit f431047ac8da13179c488018dddf1c0d0771a997

2
contrib/sysroot vendored

@ -1 +1 @@
Subproject commit 6172893931e19b028f9cabb7095a44361be863df
Subproject commit 1a64956aa7c280448be6526251bb2b8e6d380ab1

View File

@ -1,5 +1,5 @@
---
toc_priority: 71
toc_priority: 72
toc_title: Source Code Browser
---

View File

@ -0,0 +1,30 @@
---
toc_priority: 68
toc_title: Build on Linux for RISC-V 64
---
# How to Build ClickHouse on Linux for RISC-V 64 Architecture {#how-to-build-clickhouse-on-linux-for-risc-v-64-architecture}
As of writing (11.11.2021) building for risc-v considered to be highly experimental. Not all features can be enabled.
This is for the case when you have Linux machine and want to use it to build `clickhouse` binary that will run on another Linux machine with RISC-V 64 CPU architecture. This is intended for continuous integration checks that run on Linux servers.
The cross-build for RISC-V 64 is based on the [Build instructions](../development/build.md), follow them first.
## Install Clang-13
Follow the instructions from https://apt.llvm.org/ for your Ubuntu or Debian setup or do
```
sudo bash -c "$(wget -O - https://apt.llvm.org/llvm.sh)"
```
## Build ClickHouse {#build-clickhouse}
``` bash
cd ClickHouse
mkdir build-riscv64
CC=clang-13 CXX=clang++-13 cmake . -Bbuild-riscv64 -G Ninja -DCMAKE_TOOLCHAIN_FILE=cmake/linux/toolchain-riscv64.cmake DGLIBC_COMPATIBILITY=OFF -DENABLE_LDAP=OFF -DOPENSSL_NO_ASM=ON -DENABLE_JEMALLOC=ON -DENABLE_PARQUET=OFF -DUSE_INTERNAL_PARQUET_LIBRARY=OFF -DENABLE_ORC=OFF -DUSE_INTERNAL_ORC_LIBRARY=OFF -DUSE_UNWIND=OFF -DUSE_INTERNAL_PROTOBUF_LIBRARY=ON -DENABLE_GRPC=OFF -DUSE_INTERNAL_GRPC_LIBRARY=OFF -DENABLE_HDFS=OFF -DUSE_INTERNAL_HDFS3_LIBRARY=OFF -DENABLE_MYSQL=OFF -DUSE_INTERNAL_MYSQL_LIBRARY=OFF
ninja -C build-riscv64
```
The resulting binary will run only on Linux with the RISC-V 64 CPU architecture.

View File

@ -1,5 +1,5 @@
---
toc_priority: 70
toc_priority: 71
toc_title: Third-Party Libraries Used
---

View File

@ -1,5 +1,5 @@
---
toc_priority: 68
toc_priority: 69
toc_title: C++ Guide
---

View File

@ -1,5 +1,5 @@
---
toc_priority: 69
toc_priority: 70
toc_title: Testing
---

View File

@ -1,5 +1,5 @@
---
toc_priority: 71
toc_priority: 72
toc_title: "Навигация по коду ClickHouse"
---

View File

@ -0,0 +1 @@
../../en/development/build-cross-riscv.md

View File

@ -1,5 +1,5 @@
---
toc_priority: 70
toc_priority: 71
toc_title: "Используемые сторонние библиотеки"
---

View File

@ -1,5 +1,5 @@
---
toc_priority: 68
toc_priority: 69
toc_title: "Как писать код на C++"
---

View File

@ -190,6 +190,8 @@ static void * getCallerAddress(const ucontext_t & context)
return reinterpret_cast<void *>(context.uc_mcontext.pc);
#elif defined(__powerpc64__)
return reinterpret_cast<void *>(context.uc_mcontext.gp_regs[PT_NIP]);
#elif defined(__riscv)
return reinterpret_cast<void *>(context.uc_mcontext.__gregs[REG_PC]);
#else
return nullptr;
#endif

View File

@ -70,7 +70,7 @@ inline void copyOverlap8(UInt8 * op, const UInt8 *& match, size_t offset)
}
#if defined(__x86_64__) || defined(__PPC__)
#if defined(__x86_64__) || defined(__PPC__) || defined(__riscv)
/** We use 'xmm' (128bit SSE) registers here to shuffle 16 bytes.
*
@ -261,7 +261,7 @@ inline void copyOverlap16(UInt8 * op, const UInt8 *& match, const size_t offset)
}
#if defined(__x86_64__) || defined(__PPC__)
#if defined(__x86_64__) || defined(__PPC__) || defined (__riscv)
inline void copyOverlap16Shuffle(UInt8 * op, const UInt8 *& match, const size_t offset)
{