diff options
author | Daniel Baumann <daniel.baumann@progress-linux.org> | 2024-04-07 18:45:59 +0000 |
---|---|---|
committer | Daniel Baumann <daniel.baumann@progress-linux.org> | 2024-04-07 18:45:59 +0000 |
commit | 19fcec84d8d7d21e796c7624e521b60d28ee21ed (patch) | |
tree | 42d26aa27d1e3f7c0b8bd3fd14e7d7082f5008dc /src/spdk/dpdk/lib/librte_eal/include/generic | |
parent | Initial commit. (diff) | |
download | ceph-upstream.tar.xz ceph-upstream.zip |
Adding upstream version 16.2.11+ds.upstream/16.2.11+dsupstream
Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>
Diffstat (limited to 'src/spdk/dpdk/lib/librte_eal/include/generic')
13 files changed, 3433 insertions, 0 deletions
diff --git a/src/spdk/dpdk/lib/librte_eal/include/generic/rte_atomic.h b/src/spdk/dpdk/lib/librte_eal/include/generic/rte_atomic.h new file mode 100644 index 000000000..e6ab15a97 --- /dev/null +++ b/src/spdk/dpdk/lib/librte_eal/include/generic/rte_atomic.h @@ -0,0 +1,1150 @@ +/* SPDX-License-Identifier: BSD-3-Clause + * Copyright(c) 2010-2014 Intel Corporation + */ + +#ifndef _RTE_ATOMIC_H_ +#define _RTE_ATOMIC_H_ + +/** + * @file + * Atomic Operations + * + * This file defines a generic API for atomic operations. + */ + +#include <stdint.h> +#include <rte_common.h> + +#ifdef __DOXYGEN__ + +/** @name Memory Barrier + */ +///@{ +/** + * General memory barrier. + * + * Guarantees that the LOAD and STORE operations generated before the + * barrier occur before the LOAD and STORE operations generated after. + */ +static inline void rte_mb(void); + +/** + * Write memory barrier. + * + * Guarantees that the STORE operations generated before the barrier + * occur before the STORE operations generated after. + */ +static inline void rte_wmb(void); + +/** + * Read memory barrier. + * + * Guarantees that the LOAD operations generated before the barrier + * occur before the LOAD operations generated after. + */ +static inline void rte_rmb(void); +///@} + +/** @name SMP Memory Barrier + */ +///@{ +/** + * General memory barrier between lcores + * + * Guarantees that the LOAD and STORE operations that precede the + * rte_smp_mb() call are globally visible across the lcores + * before the LOAD and STORE operations that follows it. + */ +static inline void rte_smp_mb(void); + +/** + * Write memory barrier between lcores + * + * Guarantees that the STORE operations that precede the + * rte_smp_wmb() call are globally visible across the lcores + * before the STORE operations that follows it. + */ +static inline void rte_smp_wmb(void); + +/** + * Read memory barrier between lcores + * + * Guarantees that the LOAD operations that precede the + * rte_smp_rmb() call are globally visible across the lcores + * before the LOAD operations that follows it. + */ +static inline void rte_smp_rmb(void); +///@} + +/** @name I/O Memory Barrier + */ +///@{ +/** + * General memory barrier for I/O device + * + * Guarantees that the LOAD and STORE operations that precede the + * rte_io_mb() call are visible to I/O device or CPU before the + * LOAD and STORE operations that follow it. + */ +static inline void rte_io_mb(void); + +/** + * Write memory barrier for I/O device + * + * Guarantees that the STORE operations that precede the + * rte_io_wmb() call are visible to I/O device before the STORE + * operations that follow it. + */ +static inline void rte_io_wmb(void); + +/** + * Read memory barrier for IO device + * + * Guarantees that the LOAD operations on I/O device that precede the + * rte_io_rmb() call are visible to CPU before the LOAD + * operations that follow it. + */ +static inline void rte_io_rmb(void); +///@} + +/** @name Coherent I/O Memory Barrier + * + * Coherent I/O memory barrier is a lightweight version of I/O memory + * barriers which are system-wide data synchronization barriers. This + * is for only coherent memory domain between lcore and I/O device but + * it is same as the I/O memory barriers in most of architectures. + * However, some architecture provides even lighter barriers which are + * somewhere in between I/O memory barriers and SMP memory barriers. + * For example, in case of ARMv8, DMB(data memory barrier) instruction + * can have different shareability domains - inner-shareable and + * outer-shareable. And inner-shareable DMB fits for SMP memory + * barriers and outer-shareable DMB for coherent I/O memory barriers, + * which acts on coherent memory. + * + * In most cases, I/O memory barriers are safer but if operations are + * on coherent memory instead of incoherent MMIO region of a device, + * then coherent I/O memory barriers can be used and this could bring + * performance gain depending on architectures. + */ +///@{ +/** + * Write memory barrier for coherent memory between lcore and I/O device + * + * Guarantees that the STORE operations on coherent memory that + * precede the rte_cio_wmb() call are visible to I/O device before the + * STORE operations that follow it. + */ +static inline void rte_cio_wmb(void); + +/** + * Read memory barrier for coherent memory between lcore and I/O device + * + * Guarantees that the LOAD operations on coherent memory updated by + * I/O device that precede the rte_cio_rmb() call are visible to CPU + * before the LOAD operations that follow it. + */ +static inline void rte_cio_rmb(void); +///@} + +#endif /* __DOXYGEN__ */ + +/** + * Compiler barrier. + * + * Guarantees that operation reordering does not occur at compile time + * for operations directly before and after the barrier. + */ +#define rte_compiler_barrier() do { \ + asm volatile ("" : : : "memory"); \ +} while(0) + +/*------------------------- 16 bit atomic operations -------------------------*/ + +/** + * Atomic compare and set. + * + * (atomic) equivalent to: + * if (*dst == exp) + * *dst = src (all 16-bit words) + * + * @param dst + * The destination location into which the value will be written. + * @param exp + * The expected value. + * @param src + * The new value. + * @return + * Non-zero on success; 0 on failure. + */ +static inline int +rte_atomic16_cmpset(volatile uint16_t *dst, uint16_t exp, uint16_t src); + +#ifdef RTE_FORCE_INTRINSICS +static inline int +rte_atomic16_cmpset(volatile uint16_t *dst, uint16_t exp, uint16_t src) +{ + return __sync_bool_compare_and_swap(dst, exp, src); +} +#endif + +/** + * Atomic exchange. + * + * (atomic) equivalent to: + * ret = *dst + * *dst = val; + * return ret; + * + * @param dst + * The destination location into which the value will be written. + * @param val + * The new value. + * @return + * The original value at that location + */ +static inline uint16_t +rte_atomic16_exchange(volatile uint16_t *dst, uint16_t val); + +#ifdef RTE_FORCE_INTRINSICS +static inline uint16_t +rte_atomic16_exchange(volatile uint16_t *dst, uint16_t val) +{ +#if defined(__clang__) + return __atomic_exchange_n(dst, val, __ATOMIC_SEQ_CST); +#else + return __atomic_exchange_2(dst, val, __ATOMIC_SEQ_CST); +#endif +} +#endif + +/** + * The atomic counter structure. + */ +typedef struct { + volatile int16_t cnt; /**< An internal counter value. */ +} rte_atomic16_t; + +/** + * Static initializer for an atomic counter. + */ +#define RTE_ATOMIC16_INIT(val) { (val) } + +/** + * Initialize an atomic counter. + * + * @param v + * A pointer to the atomic counter. + */ +static inline void +rte_atomic16_init(rte_atomic16_t *v) +{ + v->cnt = 0; +} + +/** + * Atomically read a 16-bit value from a counter. + * + * @param v + * A pointer to the atomic counter. + * @return + * The value of the counter. + */ +static inline int16_t +rte_atomic16_read(const rte_atomic16_t *v) +{ + return v->cnt; +} + +/** + * Atomically set a counter to a 16-bit value. + * + * @param v + * A pointer to the atomic counter. + * @param new_value + * The new value for the counter. + */ +static inline void +rte_atomic16_set(rte_atomic16_t *v, int16_t new_value) +{ + v->cnt = new_value; +} + +/** + * Atomically add a 16-bit value to an atomic counter. + * + * @param v + * A pointer to the atomic counter. + * @param inc + * The value to be added to the counter. + */ +static inline void +rte_atomic16_add(rte_atomic16_t *v, int16_t inc) +{ + __sync_fetch_and_add(&v->cnt, inc); +} + +/** + * Atomically subtract a 16-bit value from an atomic counter. + * + * @param v + * A pointer to the atomic counter. + * @param dec + * The value to be subtracted from the counter. + */ +static inline void +rte_atomic16_sub(rte_atomic16_t *v, int16_t dec) +{ + __sync_fetch_and_sub(&v->cnt, dec); +} + +/** + * Atomically increment a counter by one. + * + * @param v + * A pointer to the atomic counter. + */ +static inline void +rte_atomic16_inc(rte_atomic16_t *v); + +#ifdef RTE_FORCE_INTRINSICS +static inline void +rte_atomic16_inc(rte_atomic16_t *v) +{ + rte_atomic16_add(v, 1); +} +#endif + +/** + * Atomically decrement a counter by one. + * + * @param v + * A pointer to the atomic counter. + */ +static inline void +rte_atomic16_dec(rte_atomic16_t *v); + +#ifdef RTE_FORCE_INTRINSICS +static inline void +rte_atomic16_dec(rte_atomic16_t *v) +{ + rte_atomic16_sub(v, 1); +} +#endif + +/** + * Atomically add a 16-bit value to a counter and return the result. + * + * Atomically adds the 16-bits value (inc) to the atomic counter (v) and + * returns the value of v after addition. + * + * @param v + * A pointer to the atomic counter. + * @param inc + * The value to be added to the counter. + * @return + * The value of v after the addition. + */ +static inline int16_t +rte_atomic16_add_return(rte_atomic16_t *v, int16_t inc) +{ + return __sync_add_and_fetch(&v->cnt, inc); +} + +/** + * Atomically subtract a 16-bit value from a counter and return + * the result. + * + * Atomically subtracts the 16-bit value (inc) from the atomic counter + * (v) and returns the value of v after the subtraction. + * + * @param v + * A pointer to the atomic counter. + * @param dec + * The value to be subtracted from the counter. + * @return + * The value of v after the subtraction. + */ +static inline int16_t +rte_atomic16_sub_return(rte_atomic16_t *v, int16_t dec) +{ + return __sync_sub_and_fetch(&v->cnt, dec); +} + +/** + * Atomically increment a 16-bit counter by one and test. + * + * Atomically increments the atomic counter (v) by one and returns true if + * the result is 0, or false in all other cases. + * + * @param v + * A pointer to the atomic counter. + * @return + * True if the result after the increment operation is 0; false otherwise. + */ +static inline int rte_atomic16_inc_and_test(rte_atomic16_t *v); + +#ifdef RTE_FORCE_INTRINSICS +static inline int rte_atomic16_inc_and_test(rte_atomic16_t *v) +{ + return __sync_add_and_fetch(&v->cnt, 1) == 0; +} +#endif + +/** + * Atomically decrement a 16-bit counter by one and test. + * + * Atomically decrements the atomic counter (v) by one and returns true if + * the result is 0, or false in all other cases. + * + * @param v + * A pointer to the atomic counter. + * @return + * True if the result after the decrement operation is 0; false otherwise. + */ +static inline int rte_atomic16_dec_and_test(rte_atomic16_t *v); + +#ifdef RTE_FORCE_INTRINSICS +static inline int rte_atomic16_dec_and_test(rte_atomic16_t *v) +{ + return __sync_sub_and_fetch(&v->cnt, 1) == 0; +} +#endif + +/** + * Atomically test and set a 16-bit atomic counter. + * + * If the counter value is already set, return 0 (failed). Otherwise, set + * the counter value to 1 and return 1 (success). + * + * @param v + * A pointer to the atomic counter. + * @return + * 0 if failed; else 1, success. + */ +static inline int rte_atomic16_test_and_set(rte_atomic16_t *v); + +#ifdef RTE_FORCE_INTRINSICS +static inline int rte_atomic16_test_and_set(rte_atomic16_t *v) +{ + return rte_atomic16_cmpset((volatile uint16_t *)&v->cnt, 0, 1); +} +#endif + +/** + * Atomically set a 16-bit counter to 0. + * + * @param v + * A pointer to the atomic counter. + */ +static inline void rte_atomic16_clear(rte_atomic16_t *v) +{ + v->cnt = 0; +} + +/*------------------------- 32 bit atomic operations -------------------------*/ + +/** + * Atomic compare and set. + * + * (atomic) equivalent to: + * if (*dst == exp) + * *dst = src (all 32-bit words) + * + * @param dst + * The destination location into which the value will be written. + * @param exp + * The expected value. + * @param src + * The new value. + * @return + * Non-zero on success; 0 on failure. + */ +static inline int +rte_atomic32_cmpset(volatile uint32_t *dst, uint32_t exp, uint32_t src); + +#ifdef RTE_FORCE_INTRINSICS +static inline int +rte_atomic32_cmpset(volatile uint32_t *dst, uint32_t exp, uint32_t src) +{ + return __sync_bool_compare_and_swap(dst, exp, src); +} +#endif + +/** + * Atomic exchange. + * + * (atomic) equivalent to: + * ret = *dst + * *dst = val; + * return ret; + * + * @param dst + * The destination location into which the value will be written. + * @param val + * The new value. + * @return + * The original value at that location + */ +static inline uint32_t +rte_atomic32_exchange(volatile uint32_t *dst, uint32_t val); + +#ifdef RTE_FORCE_INTRINSICS +static inline uint32_t +rte_atomic32_exchange(volatile uint32_t *dst, uint32_t val) +{ +#if defined(__clang__) + return __atomic_exchange_n(dst, val, __ATOMIC_SEQ_CST); +#else + return __atomic_exchange_4(dst, val, __ATOMIC_SEQ_CST); +#endif +} +#endif + +/** + * The atomic counter structure. + */ +typedef struct { + volatile int32_t cnt; /**< An internal counter value. */ +} rte_atomic32_t; + +/** + * Static initializer for an atomic counter. + */ +#define RTE_ATOMIC32_INIT(val) { (val) } + +/** + * Initialize an atomic counter. + * + * @param v + * A pointer to the atomic counter. + */ +static inline void +rte_atomic32_init(rte_atomic32_t *v) +{ + v->cnt = 0; +} + +/** + * Atomically read a 32-bit value from a counter. + * + * @param v + * A pointer to the atomic counter. + * @return + * The value of the counter. + */ +static inline int32_t +rte_atomic32_read(const rte_atomic32_t *v) +{ + return v->cnt; +} + +/** + * Atomically set a counter to a 32-bit value. + * + * @param v + * A pointer to the atomic counter. + * @param new_value + * The new value for the counter. + */ +static inline void +rte_atomic32_set(rte_atomic32_t *v, int32_t new_value) +{ + v->cnt = new_value; +} + +/** + * Atomically add a 32-bit value to an atomic counter. + * + * @param v + * A pointer to the atomic counter. + * @param inc + * The value to be added to the counter. + */ +static inline void +rte_atomic32_add(rte_atomic32_t *v, int32_t inc) +{ + __sync_fetch_and_add(&v->cnt, inc); +} + +/** + * Atomically subtract a 32-bit value from an atomic counter. + * + * @param v + * A pointer to the atomic counter. + * @param dec + * The value to be subtracted from the counter. + */ +static inline void +rte_atomic32_sub(rte_atomic32_t *v, int32_t dec) +{ + __sync_fetch_and_sub(&v->cnt, dec); +} + +/** + * Atomically increment a counter by one. + * + * @param v + * A pointer to the atomic counter. + */ +static inline void +rte_atomic32_inc(rte_atomic32_t *v); + +#ifdef RTE_FORCE_INTRINSICS +static inline void +rte_atomic32_inc(rte_atomic32_t *v) +{ + rte_atomic32_add(v, 1); +} +#endif + +/** + * Atomically decrement a counter by one. + * + * @param v + * A pointer to the atomic counter. + */ +static inline void +rte_atomic32_dec(rte_atomic32_t *v); + +#ifdef RTE_FORCE_INTRINSICS +static inline void +rte_atomic32_dec(rte_atomic32_t *v) +{ + rte_atomic32_sub(v,1); +} +#endif + +/** + * Atomically add a 32-bit value to a counter and return the result. + * + * Atomically adds the 32-bits value (inc) to the atomic counter (v) and + * returns the value of v after addition. + * + * @param v + * A pointer to the atomic counter. + * @param inc + * The value to be added to the counter. + * @return + * The value of v after the addition. + */ +static inline int32_t +rte_atomic32_add_return(rte_atomic32_t *v, int32_t inc) +{ + return __sync_add_and_fetch(&v->cnt, inc); +} + +/** + * Atomically subtract a 32-bit value from a counter and return + * the result. + * + * Atomically subtracts the 32-bit value (inc) from the atomic counter + * (v) and returns the value of v after the subtraction. + * + * @param v + * A pointer to the atomic counter. + * @param dec + * The value to be subtracted from the counter. + * @return + * The value of v after the subtraction. + */ +static inline int32_t +rte_atomic32_sub_return(rte_atomic32_t *v, int32_t dec) +{ + return __sync_sub_and_fetch(&v->cnt, dec); +} + +/** + * Atomically increment a 32-bit counter by one and test. + * + * Atomically increments the atomic counter (v) by one and returns true if + * the result is 0, or false in all other cases. + * + * @param v + * A pointer to the atomic counter. + * @return + * True if the result after the increment operation is 0; false otherwise. + */ +static inline int rte_atomic32_inc_and_test(rte_atomic32_t *v); + +#ifdef RTE_FORCE_INTRINSICS +static inline int rte_atomic32_inc_and_test(rte_atomic32_t *v) +{ + return __sync_add_and_fetch(&v->cnt, 1) == 0; +} +#endif + +/** + * Atomically decrement a 32-bit counter by one and test. + * + * Atomically decrements the atomic counter (v) by one and returns true if + * the result is 0, or false in all other cases. + * + * @param v + * A pointer to the atomic counter. + * @return + * True if the result after the decrement operation is 0; false otherwise. + */ +static inline int rte_atomic32_dec_and_test(rte_atomic32_t *v); + +#ifdef RTE_FORCE_INTRINSICS +static inline int rte_atomic32_dec_and_test(rte_atomic32_t *v) +{ + return __sync_sub_and_fetch(&v->cnt, 1) == 0; +} +#endif + +/** + * Atomically test and set a 32-bit atomic counter. + * + * If the counter value is already set, return 0 (failed). Otherwise, set + * the counter value to 1 and return 1 (success). + * + * @param v + * A pointer to the atomic counter. + * @return + * 0 if failed; else 1, success. + */ +static inline int rte_atomic32_test_and_set(rte_atomic32_t *v); + +#ifdef RTE_FORCE_INTRINSICS +static inline int rte_atomic32_test_and_set(rte_atomic32_t *v) +{ + return rte_atomic32_cmpset((volatile uint32_t *)&v->cnt, 0, 1); +} +#endif + +/** + * Atomically set a 32-bit counter to 0. + * + * @param v + * A pointer to the atomic counter. + */ +static inline void rte_atomic32_clear(rte_atomic32_t *v) +{ + v->cnt = 0; +} + +/*------------------------- 64 bit atomic operations -------------------------*/ + +/** + * An atomic compare and set function used by the mutex functions. + * (atomic) equivalent to: + * if (*dst == exp) + * *dst = src (all 64-bit words) + * + * @param dst + * The destination into which the value will be written. + * @param exp + * The expected value. + * @param src + * The new value. + * @return + * Non-zero on success; 0 on failure. + */ +static inline int +rte_atomic64_cmpset(volatile uint64_t *dst, uint64_t exp, uint64_t src); + +#ifdef RTE_FORCE_INTRINSICS +static inline int +rte_atomic64_cmpset(volatile uint64_t *dst, uint64_t exp, uint64_t src) +{ + return __sync_bool_compare_and_swap(dst, exp, src); +} +#endif + +/** + * Atomic exchange. + * + * (atomic) equivalent to: + * ret = *dst + * *dst = val; + * return ret; + * + * @param dst + * The destination location into which the value will be written. + * @param val + * The new value. + * @return + * The original value at that location + */ +static inline uint64_t +rte_atomic64_exchange(volatile uint64_t *dst, uint64_t val); + +#ifdef RTE_FORCE_INTRINSICS +static inline uint64_t +rte_atomic64_exchange(volatile uint64_t *dst, uint64_t val) +{ +#if defined(__clang__) + return __atomic_exchange_n(dst, val, __ATOMIC_SEQ_CST); +#else + return __atomic_exchange_8(dst, val, __ATOMIC_SEQ_CST); +#endif +} +#endif + +/** + * The atomic counter structure. + */ +typedef struct { + volatile int64_t cnt; /**< Internal counter value. */ +} rte_atomic64_t; + +/** + * Static initializer for an atomic counter. + */ +#define RTE_ATOMIC64_INIT(val) { (val) } + +/** + * Initialize the atomic counter. + * + * @param v + * A pointer to the atomic counter. + */ +static inline void +rte_atomic64_init(rte_atomic64_t *v); + +#ifdef RTE_FORCE_INTRINSICS +static inline void +rte_atomic64_init(rte_atomic64_t *v) +{ +#ifdef __LP64__ + v->cnt = 0; +#else + int success = 0; + uint64_t tmp; + + while (success == 0) { + tmp = v->cnt; + success = rte_atomic64_cmpset((volatile uint64_t *)&v->cnt, + tmp, 0); + } +#endif +} +#endif + +/** + * Atomically read a 64-bit counter. + * + * @param v + * A pointer to the atomic counter. + * @return + * The value of the counter. + */ +static inline int64_t +rte_atomic64_read(rte_atomic64_t *v); + +#ifdef RTE_FORCE_INTRINSICS +static inline int64_t +rte_atomic64_read(rte_atomic64_t *v) +{ +#ifdef __LP64__ + return v->cnt; +#else + int success = 0; + uint64_t tmp; + + while (success == 0) { + tmp = v->cnt; + /* replace the value by itself */ + success = rte_atomic64_cmpset((volatile uint64_t *)&v->cnt, + tmp, tmp); + } + return tmp; +#endif +} +#endif + +/** + * Atomically set a 64-bit counter. + * + * @param v + * A pointer to the atomic counter. + * @param new_value + * The new value of the counter. + */ +static inline void +rte_atomic64_set(rte_atomic64_t *v, int64_t new_value); + +#ifdef RTE_FORCE_INTRINSICS +static inline void +rte_atomic64_set(rte_atomic64_t *v, int64_t new_value) +{ +#ifdef __LP64__ + v->cnt = new_value; +#else + int success = 0; + uint64_t tmp; + + while (success == 0) { + tmp = v->cnt; + success = rte_atomic64_cmpset((volatile uint64_t *)&v->cnt, + tmp, new_value); + } +#endif +} +#endif + +/** + * Atomically add a 64-bit value to a counter. + * + * @param v + * A pointer to the atomic counter. + * @param inc + * The value to be added to the counter. + */ +static inline void +rte_atomic64_add(rte_atomic64_t *v, int64_t inc); + +#ifdef RTE_FORCE_INTRINSICS +static inline void +rte_atomic64_add(rte_atomic64_t *v, int64_t inc) +{ + __sync_fetch_and_add(&v->cnt, inc); +} +#endif + +/** + * Atomically subtract a 64-bit value from a counter. + * + * @param v + * A pointer to the atomic counter. + * @param dec + * The value to be subtracted from the counter. + */ +static inline void +rte_atomic64_sub(rte_atomic64_t *v, int64_t dec); + +#ifdef RTE_FORCE_INTRINSICS +static inline void +rte_atomic64_sub(rte_atomic64_t *v, int64_t dec) +{ + __sync_fetch_and_sub(&v->cnt, dec); +} +#endif + +/** + * Atomically increment a 64-bit counter by one and test. + * + * @param v + * A pointer to the atomic counter. + */ +static inline void +rte_atomic64_inc(rte_atomic64_t *v); + +#ifdef RTE_FORCE_INTRINSICS +static inline void +rte_atomic64_inc(rte_atomic64_t *v) +{ + rte_atomic64_add(v, 1); +} +#endif + +/** + * Atomically decrement a 64-bit counter by one and test. + * + * @param v + * A pointer to the atomic counter. + */ +static inline void +rte_atomic64_dec(rte_atomic64_t *v); + +#ifdef RTE_FORCE_INTRINSICS +static inline void +rte_atomic64_dec(rte_atomic64_t *v) +{ + rte_atomic64_sub(v, 1); +} +#endif + +/** + * Add a 64-bit value to an atomic counter and return the result. + * + * Atomically adds the 64-bit value (inc) to the atomic counter (v) and + * returns the value of v after the addition. + * + * @param v + * A pointer to the atomic counter. + * @param inc + * The value to be added to the counter. + * @return + * The value of v after the addition. + */ +static inline int64_t +rte_atomic64_add_return(rte_atomic64_t *v, int64_t inc); + +#ifdef RTE_FORCE_INTRINSICS +static inline int64_t +rte_atomic64_add_return(rte_atomic64_t *v, int64_t inc) +{ + return __sync_add_and_fetch(&v->cnt, inc); +} +#endif + +/** + * Subtract a 64-bit value from an atomic counter and return the result. + * + * Atomically subtracts the 64-bit value (dec) from the atomic counter (v) + * and returns the value of v after the subtraction. + * + * @param v + * A pointer to the atomic counter. + * @param dec + * The value to be subtracted from the counter. + * @return + * The value of v after the subtraction. + */ +static inline int64_t +rte_atomic64_sub_return(rte_atomic64_t *v, int64_t dec); + +#ifdef RTE_FORCE_INTRINSICS +static inline int64_t +rte_atomic64_sub_return(rte_atomic64_t *v, int64_t dec) +{ + return __sync_sub_and_fetch(&v->cnt, dec); +} +#endif + +/** + * Atomically increment a 64-bit counter by one and test. + * + * Atomically increments the atomic counter (v) by one and returns + * true if the result is 0, or false in all other cases. + * + * @param v + * A pointer to the atomic counter. + * @return + * True if the result after the addition is 0; false otherwise. + */ +static inline int rte_atomic64_inc_and_test(rte_atomic64_t *v); + +#ifdef RTE_FORCE_INTRINSICS +static inline int rte_atomic64_inc_and_test(rte_atomic64_t *v) +{ + return rte_atomic64_add_return(v, 1) == 0; +} +#endif + +/** + * Atomically decrement a 64-bit counter by one and test. + * + * Atomically decrements the atomic counter (v) by one and returns true if + * the result is 0, or false in all other cases. + * + * @param v + * A pointer to the atomic counter. + * @return + * True if the result after subtraction is 0; false otherwise. + */ +static inline int rte_atomic64_dec_and_test(rte_atomic64_t *v); + +#ifdef RTE_FORCE_INTRINSICS +static inline int rte_atomic64_dec_and_test(rte_atomic64_t *v) +{ + return rte_atomic64_sub_return(v, 1) == 0; +} +#endif + +/** + * Atomically test and set a 64-bit atomic counter. + * + * If the counter value is already set, return 0 (failed). Otherwise, set + * the counter value to 1 and return 1 (success). + * + * @param v + * A pointer to the atomic counter. + * @return + * 0 if failed; else 1, success. + */ +static inline int rte_atomic64_test_and_set(rte_atomic64_t *v); + +#ifdef RTE_FORCE_INTRINSICS +static inline int rte_atomic64_test_and_set(rte_atomic64_t *v) +{ + return rte_atomic64_cmpset((volatile uint64_t *)&v->cnt, 0, 1); +} +#endif + +/** + * Atomically set a 64-bit counter to 0. + * + * @param v + * A pointer to the atomic counter. + */ +static inline void rte_atomic64_clear(rte_atomic64_t *v); + +#ifdef RTE_FORCE_INTRINSICS +static inline void rte_atomic64_clear(rte_atomic64_t *v) +{ + rte_atomic64_set(v, 0); +} +#endif + +/*------------------------ 128 bit atomic operations -------------------------*/ + +/** + * 128-bit integer structure. + */ +RTE_STD_C11 +typedef struct { + RTE_STD_C11 + union { + uint64_t val[2]; +#ifdef RTE_ARCH_64 + __extension__ __int128 int128; +#endif + }; +} __rte_aligned(16) rte_int128_t; + +#ifdef __DOXYGEN__ + +/** + * An atomic compare and set function used by the mutex functions. + * (Atomically) Equivalent to: + * @code + * if (*dst == *exp) + * *dst = *src + * else + * *exp = *dst + * @endcode + * + * @note This function is currently available for the x86-64 and aarch64 + * platforms. + * + * @note The success and failure arguments must be one of the __ATOMIC_* values + * defined in the C++11 standard. For details on their behavior, refer to the + * standard. + * + * @param dst + * The destination into which the value will be written. + * @param exp + * Pointer to the expected value. If the operation fails, this memory is + * updated with the actual value. + * @param src + * Pointer to the new value. + * @param weak + * A value of true allows the comparison to spuriously fail and allows the + * 'exp' update to occur non-atomically (i.e. a torn read may occur). + * Implementations may ignore this argument and only implement the strong + * variant. + * @param success + * If successful, the operation's memory behavior conforms to this (or a + * stronger) model. + * @param failure + * If unsuccessful, the operation's memory behavior conforms to this (or a + * stronger) model. This argument cannot be __ATOMIC_RELEASE, + * __ATOMIC_ACQ_REL, or a stronger model than success. + * @return + * Non-zero on success; 0 on failure. + */ +__rte_experimental +static inline int +rte_atomic128_cmp_exchange(rte_int128_t *dst, + rte_int128_t *exp, + const rte_int128_t *src, + unsigned int weak, + int success, + int failure); + +#endif /* __DOXYGEN__ */ + +#endif /* _RTE_ATOMIC_H_ */ diff --git a/src/spdk/dpdk/lib/librte_eal/include/generic/rte_byteorder.h b/src/spdk/dpdk/lib/librte_eal/include/generic/rte_byteorder.h new file mode 100644 index 000000000..9ca960932 --- /dev/null +++ b/src/spdk/dpdk/lib/librte_eal/include/generic/rte_byteorder.h @@ -0,0 +1,247 @@ +/* SPDX-License-Identifier: BSD-3-Clause + * Copyright(c) 2010-2014 Intel Corporation + */ + +#ifndef _RTE_BYTEORDER_H_ +#define _RTE_BYTEORDER_H_ + +/** + * @file + * + * Byte Swap Operations + * + * This file defines a generic API for byte swap operations. Part of + * the implementation is architecture-specific. + */ + +#include <stdint.h> +#ifdef RTE_EXEC_ENV_FREEBSD +#include <sys/endian.h> +#else +#include <endian.h> +#endif + +#include <rte_common.h> +#include <rte_config.h> + +/* + * Compile-time endianness detection + */ +#define RTE_BIG_ENDIAN 1 +#define RTE_LITTLE_ENDIAN 2 +#if defined __BYTE_ORDER__ +#if __BYTE_ORDER__ == __ORDER_BIG_ENDIAN__ +#define RTE_BYTE_ORDER RTE_BIG_ENDIAN +#elif __BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__ +#define RTE_BYTE_ORDER RTE_LITTLE_ENDIAN +#endif /* __BYTE_ORDER__ */ +#elif defined __BYTE_ORDER +#if __BYTE_ORDER == __BIG_ENDIAN +#define RTE_BYTE_ORDER RTE_BIG_ENDIAN +#elif __BYTE_ORDER == __LITTLE_ENDIAN +#define RTE_BYTE_ORDER RTE_LITTLE_ENDIAN +#endif /* __BYTE_ORDER */ +#elif defined __BIG_ENDIAN__ +#define RTE_BYTE_ORDER RTE_BIG_ENDIAN +#elif defined __LITTLE_ENDIAN__ +#define RTE_BYTE_ORDER RTE_LITTLE_ENDIAN +#endif +#if !defined(RTE_BYTE_ORDER) +#error Unknown endianness. +#endif + +#define RTE_STATIC_BSWAP16(v) \ + ((((uint16_t)(v) & UINT16_C(0x00ff)) << 8) | \ + (((uint16_t)(v) & UINT16_C(0xff00)) >> 8)) + +#define RTE_STATIC_BSWAP32(v) \ + ((((uint32_t)(v) & UINT32_C(0x000000ff)) << 24) | \ + (((uint32_t)(v) & UINT32_C(0x0000ff00)) << 8) | \ + (((uint32_t)(v) & UINT32_C(0x00ff0000)) >> 8) | \ + (((uint32_t)(v) & UINT32_C(0xff000000)) >> 24)) + +#define RTE_STATIC_BSWAP64(v) \ + ((((uint64_t)(v) & UINT64_C(0x00000000000000ff)) << 56) | \ + (((uint64_t)(v) & UINT64_C(0x000000000000ff00)) << 40) | \ + (((uint64_t)(v) & UINT64_C(0x0000000000ff0000)) << 24) | \ + (((uint64_t)(v) & UINT64_C(0x00000000ff000000)) << 8) | \ + (((uint64_t)(v) & UINT64_C(0x000000ff00000000)) >> 8) | \ + (((uint64_t)(v) & UINT64_C(0x0000ff0000000000)) >> 24) | \ + (((uint64_t)(v) & UINT64_C(0x00ff000000000000)) >> 40) | \ + (((uint64_t)(v) & UINT64_C(0xff00000000000000)) >> 56)) + +/* + * These macros are functionally similar to rte_cpu_to_(be|le)(16|32|64)(), + * they take values in host CPU order and return them converted to the + * intended endianness. + * + * They resolve at compilation time to integer constants which can safely be + * used with static initializers, since those cannot involve function calls. + * + * On the other hand, they are not as optimized as their rte_cpu_to_*() + * counterparts, therefore applications should refrain from using them on + * variable values, particularly inside performance-sensitive code. + */ +#if RTE_BYTE_ORDER == RTE_BIG_ENDIAN +#define RTE_BE16(v) (rte_be16_t)(v) +#define RTE_BE32(v) (rte_be32_t)(v) +#define RTE_BE64(v) (rte_be64_t)(v) +#define RTE_LE16(v) (rte_le16_t)(RTE_STATIC_BSWAP16(v)) +#define RTE_LE32(v) (rte_le32_t)(RTE_STATIC_BSWAP32(v)) +#define RTE_LE64(v) (rte_le64_t)(RTE_STATIC_BSWAP64(v)) +#elif RTE_BYTE_ORDER == RTE_LITTLE_ENDIAN +#define RTE_BE16(v) (rte_be16_t)(RTE_STATIC_BSWAP16(v)) +#define RTE_BE32(v) (rte_be32_t)(RTE_STATIC_BSWAP32(v)) +#define RTE_BE64(v) (rte_be64_t)(RTE_STATIC_BSWAP64(v)) +#define RTE_LE16(v) (rte_le16_t)(v) +#define RTE_LE32(v) (rte_le32_t)(v) +#define RTE_LE64(v) (rte_le64_t)(v) +#else +#error Unsupported endianness. +#endif + +/* + * The following types should be used when handling values according to a + * specific byte ordering, which may differ from that of the host CPU. + * + * Libraries, public APIs and applications are encouraged to use them for + * documentation purposes. + */ +typedef uint16_t rte_be16_t; /**< 16-bit big-endian value. */ +typedef uint32_t rte_be32_t; /**< 32-bit big-endian value. */ +typedef uint64_t rte_be64_t; /**< 64-bit big-endian value. */ +typedef uint16_t rte_le16_t; /**< 16-bit little-endian value. */ +typedef uint32_t rte_le32_t; /**< 32-bit little-endian value. */ +typedef uint64_t rte_le64_t; /**< 64-bit little-endian value. */ + +/* + * An internal function to swap bytes in a 16-bit value. + * + * It is used by rte_bswap16() when the value is constant. Do not use + * this function directly; rte_bswap16() is preferred. + */ +static inline uint16_t +rte_constant_bswap16(uint16_t x) +{ + return (uint16_t)RTE_STATIC_BSWAP16(x); +} + +/* + * An internal function to swap bytes in a 32-bit value. + * + * It is used by rte_bswap32() when the value is constant. Do not use + * this function directly; rte_bswap32() is preferred. + */ +static inline uint32_t +rte_constant_bswap32(uint32_t x) +{ + return (uint32_t)RTE_STATIC_BSWAP32(x); +} + +/* + * An internal function to swap bytes of a 64-bit value. + * + * It is used by rte_bswap64() when the value is constant. Do not use + * this function directly; rte_bswap64() is preferred. + */ +static inline uint64_t +rte_constant_bswap64(uint64_t x) +{ + return (uint64_t)RTE_STATIC_BSWAP64(x); +} + + +#ifdef __DOXYGEN__ + +/** + * Swap bytes in a 16-bit value. + */ +static uint16_t rte_bswap16(uint16_t _x); + +/** + * Swap bytes in a 32-bit value. + */ +static uint32_t rte_bswap32(uint32_t x); + +/** + * Swap bytes in a 64-bit value. + */ +static uint64_t rte_bswap64(uint64_t x); + +/** + * Convert a 16-bit value from CPU order to little endian. + */ +static rte_le16_t rte_cpu_to_le_16(uint16_t x); + +/** + * Convert a 32-bit value from CPU order to little endian. + */ +static rte_le32_t rte_cpu_to_le_32(uint32_t x); + +/** + * Convert a 64-bit value from CPU order to little endian. + */ +static rte_le64_t rte_cpu_to_le_64(uint64_t x); + + +/** + * Convert a 16-bit value from CPU order to big endian. + */ +static rte_be16_t rte_cpu_to_be_16(uint16_t x); + +/** + * Convert a 32-bit value from CPU order to big endian. + */ +static rte_be32_t rte_cpu_to_be_32(uint32_t x); + +/** + * Convert a 64-bit value from CPU order to big endian. + */ +static rte_be64_t rte_cpu_to_be_64(uint64_t x); + + +/** + * Convert a 16-bit value from little endian to CPU order. + */ +static uint16_t rte_le_to_cpu_16(rte_le16_t x); + +/** + * Convert a 32-bit value from little endian to CPU order. + */ +static uint32_t rte_le_to_cpu_32(rte_le32_t x); + +/** + * Convert a 64-bit value from little endian to CPU order. + */ +static uint64_t rte_le_to_cpu_64(rte_le64_t x); + + +/** + * Convert a 16-bit value from big endian to CPU order. + */ +static uint16_t rte_be_to_cpu_16(rte_be16_t x); + +/** + * Convert a 32-bit value from big endian to CPU order. + */ +static uint32_t rte_be_to_cpu_32(rte_be32_t x); + +/** + * Convert a 64-bit value from big endian to CPU order. + */ +static uint64_t rte_be_to_cpu_64(rte_be64_t x); + +#endif /* __DOXYGEN__ */ + +#ifdef RTE_FORCE_INTRINSICS +#if __GNUC__ > 4 || (__GNUC__ == 4 && __GNUC_MINOR__ >= 8) +#define rte_bswap16(x) __builtin_bswap16(x) +#endif + +#define rte_bswap32(x) __builtin_bswap32(x) + +#define rte_bswap64(x) __builtin_bswap64(x) + +#endif + +#endif /* _RTE_BYTEORDER_H_ */ diff --git a/src/spdk/dpdk/lib/librte_eal/include/generic/rte_cpuflags.h b/src/spdk/dpdk/lib/librte_eal/include/generic/rte_cpuflags.h new file mode 100644 index 000000000..872f0ebe3 --- /dev/null +++ b/src/spdk/dpdk/lib/librte_eal/include/generic/rte_cpuflags.h @@ -0,0 +1,79 @@ +/* SPDX-License-Identifier: BSD-3-Clause + * Copyright(c) 2010-2014 Intel Corporation + */ + +#ifndef _RTE_CPUFLAGS_H_ +#define _RTE_CPUFLAGS_H_ + +/** + * @file + * Architecture specific API to determine available CPU features at runtime. + */ + +#include "rte_common.h" +#include <errno.h> + +/** + * Enumeration of all CPU features supported + */ +__extension__ +enum rte_cpu_flag_t; + +/** + * Get name of CPU flag + * + * @param feature + * CPU flag ID + * @return + * flag name + * NULL if flag ID is invalid + */ +__extension__ +const char * +rte_cpu_get_flag_name(enum rte_cpu_flag_t feature); + +/** + * Function for checking a CPU flag availability + * + * @param feature + * CPU flag to query CPU for + * @return + * 1 if flag is available + * 0 if flag is not available + * -ENOENT if flag is invalid + */ +__extension__ +int +rte_cpu_get_flag_enabled(enum rte_cpu_flag_t feature); + +/** + * This function checks that the currently used CPU supports the CPU features + * that were specified at compile time. It is called automatically within the + * EAL, so does not need to be used by applications. This version returns a + * result so that decisions may be made (for instance, graceful shutdowns). + */ +int +rte_cpu_is_supported(void); + +/** + * This function attempts to retrieve a value from the auxiliary vector. + * If it is unsuccessful, the result will be 0, and errno will be set. + * + * @return A value from the auxiliary vector. When the value is 0, check + * errno to determine if an error occurred. + */ +unsigned long +rte_cpu_getauxval(unsigned long type); + +/** + * This function retrieves a value from the auxiliary vector, and compares it + * as a string against the value retrieved. + * + * @return The result of calling strcmp() against the value retrieved from + * the auxiliary vector. When the value is 0 (meaning a match is found), + * check errno to determine if an error occurred. + */ +int +rte_cpu_strcmp_auxval(unsigned long type, const char *str); + +#endif /* _RTE_CPUFLAGS_H_ */ diff --git a/src/spdk/dpdk/lib/librte_eal/include/generic/rte_cycles.h b/src/spdk/dpdk/lib/librte_eal/include/generic/rte_cycles.h new file mode 100644 index 000000000..73d1fa7b9 --- /dev/null +++ b/src/spdk/dpdk/lib/librte_eal/include/generic/rte_cycles.h @@ -0,0 +1,181 @@ +/* SPDX-License-Identifier: BSD-3-Clause + * Copyright(c) 2010-2014 Intel Corporation. + * Copyright(c) 2013 6WIND S.A. + */ + +#ifndef _RTE_CYCLES_H_ +#define _RTE_CYCLES_H_ + +/** + * @file + * + * Simple Time Reference Functions (Cycles and HPET). + */ + +#include <stdint.h> +#include <rte_compat.h> +#include <rte_debug.h> +#include <rte_atomic.h> + +#define MS_PER_S 1000 +#define US_PER_S 1000000 +#define NS_PER_S 1000000000 + +enum timer_source { + EAL_TIMER_TSC = 0, + EAL_TIMER_HPET +}; +extern enum timer_source eal_timer_source; + +/** + * Get the measured frequency of the RDTSC counter + * + * @return + * The TSC frequency for this lcore + */ +uint64_t +rte_get_tsc_hz(void); + +/** + * Return the number of TSC cycles since boot + * + * @return + * the number of cycles + */ +static inline uint64_t +rte_get_tsc_cycles(void); + +#ifdef RTE_LIBEAL_USE_HPET +/** + * Return the number of HPET cycles since boot + * + * This counter is global for all execution units. The number of + * cycles in one second can be retrieved using rte_get_hpet_hz(). + * + * @return + * the number of cycles + */ +uint64_t +rte_get_hpet_cycles(void); + +/** + * Get the number of HPET cycles in one second. + * + * @return + * The number of cycles in one second. + */ +uint64_t +rte_get_hpet_hz(void); + +/** + * Initialise the HPET for use. This must be called before the rte_get_hpet_hz + * and rte_get_hpet_cycles APIs are called. If this function does not succeed, + * then the HPET functions are unavailable and should not be called. + * + * @param make_default + * If set, the hpet timer becomes the default timer whose values are + * returned by the rte_get_timer_hz/cycles API calls + * + * @return + * 0 on success, + * -1 on error, and the make_default parameter is ignored. + */ +int rte_eal_hpet_init(int make_default); + +#endif + +/** + * Get the number of cycles since boot from the default timer. + * + * @return + * The number of cycles + */ +static inline uint64_t +rte_get_timer_cycles(void) +{ +#ifdef RTE_LIBEAL_USE_HPET + switch(eal_timer_source) { + case EAL_TIMER_TSC: +#endif + return rte_get_tsc_cycles(); +#ifdef RTE_LIBEAL_USE_HPET + case EAL_TIMER_HPET: + return rte_get_hpet_cycles(); + default: rte_panic("Invalid timer source specified\n"); + } +#endif +} + +/** + * Get the number of cycles in one second for the default timer. + * + * @return + * The number of cycles in one second. + */ +static inline uint64_t +rte_get_timer_hz(void) +{ +#ifdef RTE_LIBEAL_USE_HPET + switch(eal_timer_source) { + case EAL_TIMER_TSC: +#endif + return rte_get_tsc_hz(); +#ifdef RTE_LIBEAL_USE_HPET + case EAL_TIMER_HPET: + return rte_get_hpet_hz(); + default: rte_panic("Invalid timer source specified\n"); + } +#endif +} +/** + * Wait at least us microseconds. + * This function can be replaced with user-defined function. + * @see rte_delay_us_callback_register + * + * @param us + * The number of microseconds to wait. + */ +extern void +(*rte_delay_us)(unsigned int us); + +/** + * Wait at least ms milliseconds. + * + * @param ms + * The number of milliseconds to wait. + */ +static inline void +rte_delay_ms(unsigned ms) +{ + rte_delay_us(ms * 1000); +} + +/** + * Blocking delay function. + * + * @param us + * Number of microseconds to wait. + */ +void rte_delay_us_block(unsigned int us); + +/** + * Delay function that uses system sleep. + * Does not block the CPU core. + * + * @param us + * Number of microseconds to wait. + */ +__rte_experimental +void +rte_delay_us_sleep(unsigned int us); + +/** + * Replace rte_delay_us with user defined function. + * + * @param userfunc + * User function which replaces rte_delay_us. rte_delay_us_block restores + * builtin block delay function. + */ +void rte_delay_us_callback_register(void(*userfunc)(unsigned int)); + +#endif /* _RTE_CYCLES_H_ */ diff --git a/src/spdk/dpdk/lib/librte_eal/include/generic/rte_io.h b/src/spdk/dpdk/lib/librte_eal/include/generic/rte_io.h new file mode 100644 index 000000000..da457f7f7 --- /dev/null +++ b/src/spdk/dpdk/lib/librte_eal/include/generic/rte_io.h @@ -0,0 +1,350 @@ +/* SPDX-License-Identifier: BSD-3-Clause + * Copyright(c) 2016 Cavium, Inc + */ + +#ifndef _RTE_IO_H_ +#define _RTE_IO_H_ + +/** + * @file + * I/O device memory operations + * + * This file defines the generic API for I/O device memory read/write operations + */ + +#include <stdint.h> +#include <rte_common.h> +#include <rte_atomic.h> + +#ifdef __DOXYGEN__ + +/** + * Read a 8-bit value from I/O device memory address *addr*. + * + * The relaxed version does not have additional I/O memory barrier, useful in + * accessing the device registers of integrated controllers which implicitly + * strongly ordered with respect to memory access. + * + * @param addr + * I/O memory address to read the value from + * @return + * read value + */ +static inline uint8_t +rte_read8_relaxed(const volatile void *addr); + +/** + * Read a 16-bit value from I/O device memory address *addr*. + * + * The relaxed version does not have additional I/O memory barrier, useful in + * accessing the device registers of integrated controllers which implicitly + * strongly ordered with respect to memory access. + * + * @param addr + * I/O memory address to read the value from + * @return + * read value + */ +static inline uint16_t +rte_read16_relaxed(const volatile void *addr); + +/** + * Read a 32-bit value from I/O device memory address *addr*. + * + * The relaxed version does not have additional I/O memory barrier, useful in + * accessing the device registers of integrated controllers which implicitly + * strongly ordered with respect to memory access. + * + * @param addr + * I/O memory address to read the value from + * @return + * read value + */ +static inline uint32_t +rte_read32_relaxed(const volatile void *addr); + +/** + * Read a 64-bit value from I/O device memory address *addr*. + * + * The relaxed version does not have additional I/O memory barrier, useful in + * accessing the device registers of integrated controllers which implicitly + * strongly ordered with respect to memory access. + * + * @param addr + * I/O memory address to read the value from + * @return + * read value + */ +static inline uint64_t +rte_read64_relaxed(const volatile void *addr); + +/** + * Write a 8-bit value to I/O device memory address *addr*. + * + * The relaxed version does not have additional I/O memory barrier, useful in + * accessing the device registers of integrated controllers which implicitly + * strongly ordered with respect to memory access. + * + * @param value + * Value to write + * @param addr + * I/O memory address to write the value to + */ + +static inline void +rte_write8_relaxed(uint8_t value, volatile void *addr); + +/** + * Write a 16-bit value to I/O device memory address *addr*. + * + * The relaxed version does not have additional I/O memory barrier, useful in + * accessing the device registers of integrated controllers which implicitly + * strongly ordered with respect to memory access. + * + * @param value + * Value to write + * @param addr + * I/O memory address to write the value to + */ +static inline void +rte_write16_relaxed(uint16_t value, volatile void *addr); + +/** + * Write a 32-bit value to I/O device memory address *addr*. + * + * The relaxed version does not have additional I/O memory barrier, useful in + * accessing the device registers of integrated controllers which implicitly + * strongly ordered with respect to memory access. + * + * @param value + * Value to write + * @param addr + * I/O memory address to write the value to + */ +static inline void +rte_write32_relaxed(uint32_t value, volatile void *addr); + +/** + * Write a 64-bit value to I/O device memory address *addr*. + * + * The relaxed version does not have additional I/O memory barrier, useful in + * accessing the device registers of integrated controllers which implicitly + * strongly ordered with respect to memory access. + * + * @param value + * Value to write + * @param addr + * I/O memory address to write the value to + */ +static inline void +rte_write64_relaxed(uint64_t value, volatile void *addr); + +/** + * Read a 8-bit value from I/O device memory address *addr*. + * + * @param addr + * I/O memory address to read the value from + * @return + * read value + */ +static inline uint8_t +rte_read8(const volatile void *addr); + +/** + * Read a 16-bit value from I/O device memory address *addr*. + * + * + * @param addr + * I/O memory address to read the value from + * @return + * read value + */ +static inline uint16_t +rte_read16(const volatile void *addr); + +/** + * Read a 32-bit value from I/O device memory address *addr*. + * + * @param addr + * I/O memory address to read the value from + * @return + * read value + */ +static inline uint32_t +rte_read32(const volatile void *addr); + +/** + * Read a 64-bit value from I/O device memory address *addr*. + * + * @param addr + * I/O memory address to read the value from + * @return + * read value + */ +static inline uint64_t +rte_read64(const volatile void *addr); + +/** + * Write a 8-bit value to I/O device memory address *addr*. + * + * @param value + * Value to write + * @param addr + * I/O memory address to write the value to + */ + +static inline void +rte_write8(uint8_t value, volatile void *addr); + +/** + * Write a 16-bit value to I/O device memory address *addr*. + * + * @param value + * Value to write + * @param addr + * I/O memory address to write the value to + */ +static inline void +rte_write16(uint16_t value, volatile void *addr); + +/** + * Write a 32-bit value to I/O device memory address *addr*. + * + * @param value + * Value to write + * @param addr + * I/O memory address to write the value to + */ +static inline void +rte_write32(uint32_t value, volatile void *addr); + +/** + * Write a 64-bit value to I/O device memory address *addr*. + * + * @param value + * Value to write + * @param addr + * I/O memory address to write the value to + */ +static inline void +rte_write64(uint64_t value, volatile void *addr); + +#endif /* __DOXYGEN__ */ + +#ifndef RTE_OVERRIDE_IO_H + +static __rte_always_inline uint8_t +rte_read8_relaxed(const volatile void *addr) +{ + return *(const volatile uint8_t *)addr; +} + +static __rte_always_inline uint16_t +rte_read16_relaxed(const volatile void *addr) +{ + return *(const volatile uint16_t *)addr; +} + +static __rte_always_inline uint32_t +rte_read32_relaxed(const volatile void *addr) +{ + return *(const volatile uint32_t *)addr; +} + +static __rte_always_inline uint64_t +rte_read64_relaxed(const volatile void *addr) +{ + return *(const volatile uint64_t *)addr; +} + +static __rte_always_inline void +rte_write8_relaxed(uint8_t value, volatile void *addr) +{ + *(volatile uint8_t *)addr = value; +} + +static __rte_always_inline void +rte_write16_relaxed(uint16_t value, volatile void *addr) +{ + *(volatile uint16_t *)addr = value; +} + +static __rte_always_inline void +rte_write32_relaxed(uint32_t value, volatile void *addr) +{ + *(volatile uint32_t *)addr = value; +} + +static __rte_always_inline void +rte_write64_relaxed(uint64_t value, volatile void *addr) +{ + *(volatile uint64_t *)addr = value; +} + +static __rte_always_inline uint8_t +rte_read8(const volatile void *addr) +{ + uint8_t val; + val = rte_read8_relaxed(addr); + rte_io_rmb(); + return val; +} + +static __rte_always_inline uint16_t +rte_read16(const volatile void *addr) +{ + uint16_t val; + val = rte_read16_relaxed(addr); + rte_io_rmb(); + return val; +} + +static __rte_always_inline uint32_t +rte_read32(const volatile void *addr) +{ + uint32_t val; + val = rte_read32_relaxed(addr); + rte_io_rmb(); + return val; +} + +static __rte_always_inline uint64_t +rte_read64(const volatile void *addr) +{ + uint64_t val; + val = rte_read64_relaxed(addr); + rte_io_rmb(); + return val; +} + +static __rte_always_inline void +rte_write8(uint8_t value, volatile void *addr) +{ + rte_io_wmb(); + rte_write8_relaxed(value, addr); +} + +static __rte_always_inline void +rte_write16(uint16_t value, volatile void *addr) +{ + rte_io_wmb(); + rte_write16_relaxed(value, addr); +} + +static __rte_always_inline void +rte_write32(uint32_t value, volatile void *addr) +{ + rte_io_wmb(); + rte_write32_relaxed(value, addr); +} + +static __rte_always_inline void +rte_write64(uint64_t value, volatile void *addr) +{ + rte_io_wmb(); + rte_write64_relaxed(value, addr); +} + +#endif /* RTE_OVERRIDE_IO_H */ + +#endif /* _RTE_IO_H_ */ diff --git a/src/spdk/dpdk/lib/librte_eal/include/generic/rte_mcslock.h b/src/spdk/dpdk/lib/librte_eal/include/generic/rte_mcslock.h new file mode 100644 index 000000000..2bef28351 --- /dev/null +++ b/src/spdk/dpdk/lib/librte_eal/include/generic/rte_mcslock.h @@ -0,0 +1,179 @@ +/* SPDX-License-Identifier: BSD-3-Clause + * Copyright(c) 2019 Arm Limited + */ + +#ifndef _RTE_MCSLOCK_H_ +#define _RTE_MCSLOCK_H_ + +/** + * @file + * + * RTE MCS lock + * + * This file defines the main data structure and APIs for MCS queued lock. + * + * The MCS lock (proposed by John M. Mellor-Crummey and Michael L. Scott) + * provides scalability by spinning on a CPU/thread local variable which + * avoids expensive cache bouncings. It provides fairness by maintaining + * a list of acquirers and passing the lock to each CPU/thread in the order + * they acquired the lock. + */ + +#include <rte_lcore.h> +#include <rte_common.h> +#include <rte_pause.h> + +/** + * The rte_mcslock_t type. + */ +typedef struct rte_mcslock { + struct rte_mcslock *next; + int locked; /* 1 if the queue locked, 0 otherwise */ +} rte_mcslock_t; + +/** + * @warning + * @b EXPERIMENTAL: This API may change without prior notice + * + * Take the MCS lock. + * + * @param msl + * A pointer to the pointer of a MCS lock. + * When the lock is initialized or declared, the msl pointer should be + * set to NULL. + * @param me + * A pointer to a new node of MCS lock. Each CPU/thread acquiring the + * lock should use its 'own node'. + */ +__rte_experimental +static inline void +rte_mcslock_lock(rte_mcslock_t **msl, rte_mcslock_t *me) +{ + rte_mcslock_t *prev; + + /* Init me node */ + __atomic_store_n(&me->locked, 1, __ATOMIC_RELAXED); + __atomic_store_n(&me->next, NULL, __ATOMIC_RELAXED); + + /* If the queue is empty, the exchange operation is enough to acquire + * the lock. Hence, the exchange operation requires acquire semantics. + * The store to me->next above should complete before the node is + * visible to other CPUs/threads. Hence, the exchange operation requires + * release semantics as well. + */ + prev = __atomic_exchange_n(msl, me, __ATOMIC_ACQ_REL); + if (likely(prev == NULL)) { + /* Queue was empty, no further action required, + * proceed with lock taken. + */ + return; + } + __atomic_store_n(&prev->next, me, __ATOMIC_RELAXED); + + /* The while-load of me->locked should not move above the previous + * store to prev->next. Otherwise it will cause a deadlock. Need a + * store-load barrier. + */ + __atomic_thread_fence(__ATOMIC_ACQ_REL); + /* If the lock has already been acquired, it first atomically + * places the node at the end of the queue and then proceeds + * to spin on me->locked until the previous lock holder resets + * the me->locked using mcslock_unlock(). + */ + while (__atomic_load_n(&me->locked, __ATOMIC_ACQUIRE)) + rte_pause(); +} + +/** + * @warning + * @b EXPERIMENTAL: This API may change without prior notice + * + * Release the MCS lock. + * + * @param msl + * A pointer to the pointer of a MCS lock. + * @param me + * A pointer to the node of MCS lock passed in rte_mcslock_lock. + */ +__rte_experimental +static inline void +rte_mcslock_unlock(rte_mcslock_t **msl, rte_mcslock_t *me) +{ + /* Check if there are more nodes in the queue. */ + if (likely(__atomic_load_n(&me->next, __ATOMIC_RELAXED) == NULL)) { + /* No, last member in the queue. */ + rte_mcslock_t *save_me = __atomic_load_n(&me, __ATOMIC_RELAXED); + + /* Release the lock by setting it to NULL */ + if (likely(__atomic_compare_exchange_n(msl, &save_me, NULL, 0, + __ATOMIC_RELEASE, __ATOMIC_RELAXED))) + return; + + /* Speculative execution would be allowed to read in the + * while-loop first. This has the potential to cause a + * deadlock. Need a load barrier. + */ + __atomic_thread_fence(__ATOMIC_ACQUIRE); + /* More nodes added to the queue by other CPUs. + * Wait until the next pointer is set. + */ + while (__atomic_load_n(&me->next, __ATOMIC_RELAXED) == NULL) + rte_pause(); + } + + /* Pass lock to next waiter. */ + __atomic_store_n(&me->next->locked, 0, __ATOMIC_RELEASE); +} + +/** + * @warning + * @b EXPERIMENTAL: This API may change without prior notice + * + * Try to take the lock. + * + * @param msl + * A pointer to the pointer of a MCS lock. + * @param me + * A pointer to a new node of MCS lock. + * @return + * 1 if the lock is successfully taken; 0 otherwise. + */ +__rte_experimental +static inline int +rte_mcslock_trylock(rte_mcslock_t **msl, rte_mcslock_t *me) +{ + /* Init me node */ + __atomic_store_n(&me->next, NULL, __ATOMIC_RELAXED); + + /* Try to lock */ + rte_mcslock_t *expected = NULL; + + /* The lock can be taken only when the queue is empty. Hence, + * the compare-exchange operation requires acquire semantics. + * The store to me->next above should complete before the node + * is visible to other CPUs/threads. Hence, the compare-exchange + * operation requires release semantics as well. + */ + return __atomic_compare_exchange_n(msl, &expected, me, 0, + __ATOMIC_ACQ_REL, __ATOMIC_RELAXED); +} + +/** + * @warning + * @b EXPERIMENTAL: This API may change without prior notice + * + * Test if the lock is taken. + * + * @param msl + * A pointer to a MCS lock node. + * @return + * 1 if the lock is currently taken; 0 otherwise. + */ +__rte_experimental +static inline int +rte_mcslock_is_locked(rte_mcslock_t *msl) +{ + return (__atomic_load_n(&msl, __ATOMIC_RELAXED) != NULL); +} + +#endif /* _RTE_MCSLOCK_H_ */ diff --git a/src/spdk/dpdk/lib/librte_eal/include/generic/rte_memcpy.h b/src/spdk/dpdk/lib/librte_eal/include/generic/rte_memcpy.h new file mode 100644 index 000000000..701e550c3 --- /dev/null +++ b/src/spdk/dpdk/lib/librte_eal/include/generic/rte_memcpy.h @@ -0,0 +1,112 @@ +/* SPDX-License-Identifier: BSD-3-Clause + * Copyright(c) 2010-2014 Intel Corporation + */ + +#ifndef _RTE_MEMCPY_H_ +#define _RTE_MEMCPY_H_ + +/** + * @file + * + * Functions for vectorised implementation of memcpy(). + */ + +/** + * Copy 16 bytes from one location to another using optimised + * instructions. The locations should not overlap. + * + * @param dst + * Pointer to the destination of the data. + * @param src + * Pointer to the source data. + */ +static inline void +rte_mov16(uint8_t *dst, const uint8_t *src); + +/** + * Copy 32 bytes from one location to another using optimised + * instructions. The locations should not overlap. + * + * @param dst + * Pointer to the destination of the data. + * @param src + * Pointer to the source data. + */ +static inline void +rte_mov32(uint8_t *dst, const uint8_t *src); + +#ifdef __DOXYGEN__ + +/** + * Copy 48 bytes from one location to another using optimised + * instructions. The locations should not overlap. + * + * @param dst + * Pointer to the destination of the data. + * @param src + * Pointer to the source data. + */ +static inline void +rte_mov48(uint8_t *dst, const uint8_t *src); + +#endif /* __DOXYGEN__ */ + +/** + * Copy 64 bytes from one location to another using optimised + * instructions. The locations should not overlap. + * + * @param dst + * Pointer to the destination of the data. + * @param src + * Pointer to the source data. + */ +static inline void +rte_mov64(uint8_t *dst, const uint8_t *src); + +/** + * Copy 128 bytes from one location to another using optimised + * instructions. The locations should not overlap. + * + * @param dst + * Pointer to the destination of the data. + * @param src + * Pointer to the source data. + */ +static inline void +rte_mov128(uint8_t *dst, const uint8_t *src); + +/** + * Copy 256 bytes from one location to another using optimised + * instructions. The locations should not overlap. + * + * @param dst + * Pointer to the destination of the data. + * @param src + * Pointer to the source data. + */ +static inline void +rte_mov256(uint8_t *dst, const uint8_t *src); + +#ifdef __DOXYGEN__ + +/** + * Copy bytes from one location to another. The locations must not overlap. + * + * @note This is implemented as a macro, so it's address should not be taken + * and care is needed as parameter expressions may be evaluated multiple times. + * + * @param dst + * Pointer to the destination of the data. + * @param src + * Pointer to the source data. + * @param n + * Number of bytes to copy. + * @return + * Pointer to the destination data. + */ +static void * +rte_memcpy(void *dst, const void *src, size_t n); + +#endif /* __DOXYGEN__ */ + +#endif /* _RTE_MEMCPY_H_ */ diff --git a/src/spdk/dpdk/lib/librte_eal/include/generic/rte_pause.h b/src/spdk/dpdk/lib/librte_eal/include/generic/rte_pause.h new file mode 100644 index 000000000..7422785f1 --- /dev/null +++ b/src/spdk/dpdk/lib/librte_eal/include/generic/rte_pause.h @@ -0,0 +1,128 @@ +/* SPDX-License-Identifier: BSD-3-Clause + * Copyright(c) 2017 Cavium, Inc + * Copyright(c) 2019 Arm Limited + */ + +#ifndef _RTE_PAUSE_H_ +#define _RTE_PAUSE_H_ + +/** + * @file + * + * CPU pause operation. + * + */ + +#include <stdint.h> +#include <assert.h> +#include <rte_common.h> +#include <rte_atomic.h> +#include <rte_compat.h> + +/** + * Pause CPU execution for a short while + * + * This call is intended for tight loops which poll a shared resource or wait + * for an event. A short pause within the loop may reduce the power consumption. + */ +static inline void rte_pause(void); + +/** + * @warning + * @b EXPERIMENTAL: this API may change, or be removed, without prior notice + * + * Wait for *addr to be updated with a 16-bit expected value, with a relaxed + * memory ordering model meaning the loads around this API can be reordered. + * + * @param addr + * A pointer to the memory location. + * @param expected + * A 16-bit expected value to be in the memory location. + * @param memorder + * Two different memory orders that can be specified: + * __ATOMIC_ACQUIRE and __ATOMIC_RELAXED. These map to + * C++11 memory orders with the same names, see the C++11 standard or + * the GCC wiki on atomic synchronization for detailed definition. + */ +__rte_experimental +static __rte_always_inline void +rte_wait_until_equal_16(volatile uint16_t *addr, uint16_t expected, + int memorder); + +/** + * @warning + * @b EXPERIMENTAL: this API may change, or be removed, without prior notice + * + * Wait for *addr to be updated with a 32-bit expected value, with a relaxed + * memory ordering model meaning the loads around this API can be reordered. + * + * @param addr + * A pointer to the memory location. + * @param expected + * A 32-bit expected value to be in the memory location. + * @param memorder + * Two different memory orders that can be specified: + * __ATOMIC_ACQUIRE and __ATOMIC_RELAXED. These map to + * C++11 memory orders with the same names, see the C++11 standard or + * the GCC wiki on atomic synchronization for detailed definition. + */ +__rte_experimental +static __rte_always_inline void +rte_wait_until_equal_32(volatile uint32_t *addr, uint32_t expected, + int memorder); + +/** + * @warning + * @b EXPERIMENTAL: this API may change, or be removed, without prior notice + * + * Wait for *addr to be updated with a 64-bit expected value, with a relaxed + * memory ordering model meaning the loads around this API can be reordered. + * + * @param addr + * A pointer to the memory location. + * @param expected + * A 64-bit expected value to be in the memory location. + * @param memorder + * Two different memory orders that can be specified: + * __ATOMIC_ACQUIRE and __ATOMIC_RELAXED. These map to + * C++11 memory orders with the same names, see the C++11 standard or + * the GCC wiki on atomic synchronization for detailed definition. + */ +__rte_experimental +static __rte_always_inline void +rte_wait_until_equal_64(volatile uint64_t *addr, uint64_t expected, + int memorder); + +#ifndef RTE_WAIT_UNTIL_EQUAL_ARCH_DEFINED +static __rte_always_inline void +rte_wait_until_equal_16(volatile uint16_t *addr, uint16_t expected, + int memorder) +{ + assert(memorder == __ATOMIC_ACQUIRE || memorder == __ATOMIC_RELAXED); + + while (__atomic_load_n(addr, memorder) != expected) + rte_pause(); +} + +static __rte_always_inline void +rte_wait_until_equal_32(volatile uint32_t *addr, uint32_t expected, + int memorder) +{ + assert(memorder == __ATOMIC_ACQUIRE || memorder == __ATOMIC_RELAXED); + + while (__atomic_load_n(addr, memorder) != expected) + rte_pause(); +} + +static __rte_always_inline void +rte_wait_until_equal_64(volatile uint64_t *addr, uint64_t expected, + int memorder) +{ + assert(memorder == __ATOMIC_ACQUIRE || memorder == __ATOMIC_RELAXED); + + while (__atomic_load_n(addr, memorder) != expected) + rte_pause(); +} +#endif + +#endif /* _RTE_PAUSE_H_ */ diff --git a/src/spdk/dpdk/lib/librte_eal/include/generic/rte_prefetch.h b/src/spdk/dpdk/lib/librte_eal/include/generic/rte_prefetch.h new file mode 100644 index 000000000..6e47bdfba --- /dev/null +++ b/src/spdk/dpdk/lib/librte_eal/include/generic/rte_prefetch.h @@ -0,0 +1,54 @@ +/* SPDX-License-Identifier: BSD-3-Clause + * Copyright(c) 2010-2015 Intel Corporation + */ + +#ifndef _RTE_PREFETCH_H_ +#define _RTE_PREFETCH_H_ + +/** + * @file + * + * Prefetch operations. + * + * This file defines an API for prefetch macros / inline-functions, + * which are architecture-dependent. Prefetching occurs when a + * processor requests an instruction or data from memory to cache + * before it is actually needed, potentially speeding up the execution of the + * program. + */ + +/** + * Prefetch a cache line into all cache levels. + * @param p + * Address to prefetch + */ +static inline void rte_prefetch0(const volatile void *p); + +/** + * Prefetch a cache line into all cache levels except the 0th cache level. + * @param p + * Address to prefetch + */ +static inline void rte_prefetch1(const volatile void *p); + +/** + * Prefetch a cache line into all cache levels except the 0th and 1th cache + * levels. + * @param p + * Address to prefetch + */ +static inline void rte_prefetch2(const volatile void *p); + +/** + * Prefetch a cache line into all cache levels (non-temporal/transient version) + * + * The non-temporal prefetch is intended as a prefetch hint that processor will + * use the prefetched data only once or short period, unlike the + * rte_prefetch0() function which imply that prefetched data to use repeatedly. + * + * @param p + * Address to prefetch + */ +static inline void rte_prefetch_non_temporal(const volatile void *p); + +#endif /* _RTE_PREFETCH_H_ */ diff --git a/src/spdk/dpdk/lib/librte_eal/include/generic/rte_rwlock.h b/src/spdk/dpdk/lib/librte_eal/include/generic/rte_rwlock.h new file mode 100644 index 000000000..da9bc3e9c --- /dev/null +++ b/src/spdk/dpdk/lib/librte_eal/include/generic/rte_rwlock.h @@ -0,0 +1,239 @@ +/* SPDX-License-Identifier: BSD-3-Clause + * Copyright(c) 2010-2014 Intel Corporation + */ + +#ifndef _RTE_RWLOCK_H_ +#define _RTE_RWLOCK_H_ + +/** + * @file + * + * RTE Read-Write Locks + * + * This file defines an API for read-write locks. The lock is used to + * protect data that allows multiple readers in parallel, but only + * one writer. All readers are blocked until the writer is finished + * writing. + * + */ + +#ifdef __cplusplus +extern "C" { +#endif + +#include <rte_common.h> +#include <rte_atomic.h> +#include <rte_pause.h> + +/** + * The rte_rwlock_t type. + * + * cnt is -1 when write lock is held, and > 0 when read locks are held. + */ +typedef struct { + volatile int32_t cnt; /**< -1 when W lock held, > 0 when R locks held. */ +} rte_rwlock_t; + +/** + * A static rwlock initializer. + */ +#define RTE_RWLOCK_INITIALIZER { 0 } + +/** + * Initialize the rwlock to an unlocked state. + * + * @param rwl + * A pointer to the rwlock structure. + */ +static inline void +rte_rwlock_init(rte_rwlock_t *rwl) +{ + rwl->cnt = 0; +} + +/** + * Take a read lock. Loop until the lock is held. + * + * @param rwl + * A pointer to a rwlock structure. + */ +static inline void +rte_rwlock_read_lock(rte_rwlock_t *rwl) +{ + int32_t x; + int success = 0; + + while (success == 0) { + x = __atomic_load_n(&rwl->cnt, __ATOMIC_RELAXED); + /* write lock is held */ + if (x < 0) { + rte_pause(); + continue; + } + success = __atomic_compare_exchange_n(&rwl->cnt, &x, x + 1, 1, + __ATOMIC_ACQUIRE, __ATOMIC_RELAXED); + } +} + +/** + * @warning + * @b EXPERIMENTAL: this API may change without prior notice. + * + * try to take a read lock. + * + * @param rwl + * A pointer to a rwlock structure. + * @return + * - zero if the lock is successfully taken + * - -EBUSY if lock could not be acquired for reading because a + * writer holds the lock + */ +__rte_experimental +static inline int +rte_rwlock_read_trylock(rte_rwlock_t *rwl) +{ + int32_t x; + int success = 0; + + while (success == 0) { + x = __atomic_load_n(&rwl->cnt, __ATOMIC_RELAXED); + /* write lock is held */ + if (x < 0) + return -EBUSY; + success = __atomic_compare_exchange_n(&rwl->cnt, &x, x + 1, 1, + __ATOMIC_ACQUIRE, __ATOMIC_RELAXED); + } + + return 0; +} + +/** + * Release a read lock. + * + * @param rwl + * A pointer to the rwlock structure. + */ +static inline void +rte_rwlock_read_unlock(rte_rwlock_t *rwl) +{ + __atomic_fetch_sub(&rwl->cnt, 1, __ATOMIC_RELEASE); +} + +/** + * @warning + * @b EXPERIMENTAL: this API may change without prior notice. + * + * try to take a write lock. + * + * @param rwl + * A pointer to a rwlock structure. + * @return + * - zero if the lock is successfully taken + * - -EBUSY if lock could not be acquired for writing because + * it was already locked for reading or writing + */ +__rte_experimental +static inline int +rte_rwlock_write_trylock(rte_rwlock_t *rwl) +{ + int32_t x; + + x = __atomic_load_n(&rwl->cnt, __ATOMIC_RELAXED); + if (x != 0 || __atomic_compare_exchange_n(&rwl->cnt, &x, -1, 1, + __ATOMIC_ACQUIRE, __ATOMIC_RELAXED) == 0) + return -EBUSY; + + return 0; +} + +/** + * Take a write lock. Loop until the lock is held. + * + * @param rwl + * A pointer to a rwlock structure. + */ +static inline void +rte_rwlock_write_lock(rte_rwlock_t *rwl) +{ + int32_t x; + int success = 0; + + while (success == 0) { + x = __atomic_load_n(&rwl->cnt, __ATOMIC_RELAXED); + /* a lock is held */ + if (x != 0) { + rte_pause(); + continue; + } + success = __atomic_compare_exchange_n(&rwl->cnt, &x, -1, 1, + __ATOMIC_ACQUIRE, __ATOMIC_RELAXED); + } +} + +/** + * Release a write lock. + * + * @param rwl + * A pointer to a rwlock structure. + */ +static inline void +rte_rwlock_write_unlock(rte_rwlock_t *rwl) +{ + __atomic_store_n(&rwl->cnt, 0, __ATOMIC_RELEASE); +} + +/** + * Try to execute critical section in a hardware memory transaction, if it + * fails or not available take a read lock + * + * NOTE: An attempt to perform a HW I/O operation inside a hardware memory + * transaction always aborts the transaction since the CPU is not able to + * roll-back should the transaction fail. Therefore, hardware transactional + * locks are not advised to be used around rte_eth_rx_burst() and + * rte_eth_tx_burst() calls. + * + * @param rwl + * A pointer to a rwlock structure. + */ +static inline void +rte_rwlock_read_lock_tm(rte_rwlock_t *rwl); + +/** + * Commit hardware memory transaction or release the read lock if the lock is used as a fall-back + * + * @param rwl + * A pointer to the rwlock structure. + */ +static inline void +rte_rwlock_read_unlock_tm(rte_rwlock_t *rwl); + +/** + * Try to execute critical section in a hardware memory transaction, if it + * fails or not available take a write lock + * + * NOTE: An attempt to perform a HW I/O operation inside a hardware memory + * transaction always aborts the transaction since the CPU is not able to + * roll-back should the transaction fail. Therefore, hardware transactional + * locks are not advised to be used around rte_eth_rx_burst() and + * rte_eth_tx_burst() calls. + * + * @param rwl + * A pointer to a rwlock structure. + */ +static inline void +rte_rwlock_write_lock_tm(rte_rwlock_t *rwl); + +/** + * Commit hardware memory transaction or release the write lock if the lock is used as a fall-back + * + * @param rwl + * A pointer to a rwlock structure. + */ +static inline void +rte_rwlock_write_unlock_tm(rte_rwlock_t *rwl); + +#ifdef __cplusplus +} +#endif + +#endif /* _RTE_RWLOCK_H_ */ diff --git a/src/spdk/dpdk/lib/librte_eal/include/generic/rte_spinlock.h b/src/spdk/dpdk/lib/librte_eal/include/generic/rte_spinlock.h new file mode 100644 index 000000000..87ae7a4f1 --- /dev/null +++ b/src/spdk/dpdk/lib/librte_eal/include/generic/rte_spinlock.h @@ -0,0 +1,305 @@ +/* SPDX-License-Identifier: BSD-3-Clause + * Copyright(c) 2010-2014 Intel Corporation + */ + +#ifndef _RTE_SPINLOCK_H_ +#define _RTE_SPINLOCK_H_ + +/** + * @file + * + * RTE Spinlocks + * + * This file defines an API for read-write locks, which are implemented + * in an architecture-specific way. This kind of lock simply waits in + * a loop repeatedly checking until the lock becomes available. + * + * All locks must be initialised before use, and only initialised once. + * + */ + +#include <rte_lcore.h> +#ifdef RTE_FORCE_INTRINSICS +#include <rte_common.h> +#endif +#include <rte_pause.h> + +/** + * The rte_spinlock_t type. + */ +typedef struct { + volatile int locked; /**< lock status 0 = unlocked, 1 = locked */ +} rte_spinlock_t; + +/** + * A static spinlock initializer. + */ +#define RTE_SPINLOCK_INITIALIZER { 0 } + +/** + * Initialize the spinlock to an unlocked state. + * + * @param sl + * A pointer to the spinlock. + */ +static inline void +rte_spinlock_init(rte_spinlock_t *sl) +{ + sl->locked = 0; +} + +/** + * Take the spinlock. + * + * @param sl + * A pointer to the spinlock. + */ +static inline void +rte_spinlock_lock(rte_spinlock_t *sl); + +#ifdef RTE_FORCE_INTRINSICS +static inline void +rte_spinlock_lock(rte_spinlock_t *sl) +{ + int exp = 0; + + while (!__atomic_compare_exchange_n(&sl->locked, &exp, 1, 0, + __ATOMIC_ACQUIRE, __ATOMIC_RELAXED)) { + while (__atomic_load_n(&sl->locked, __ATOMIC_RELAXED)) + rte_pause(); + exp = 0; + } +} +#endif + +/** + * Release the spinlock. + * + * @param sl + * A pointer to the spinlock. + */ +static inline void +rte_spinlock_unlock (rte_spinlock_t *sl); + +#ifdef RTE_FORCE_INTRINSICS +static inline void +rte_spinlock_unlock (rte_spinlock_t *sl) +{ + __atomic_store_n(&sl->locked, 0, __ATOMIC_RELEASE); +} +#endif + +/** + * Try to take the lock. + * + * @param sl + * A pointer to the spinlock. + * @return + * 1 if the lock is successfully taken; 0 otherwise. + */ +static inline int +rte_spinlock_trylock (rte_spinlock_t *sl); + +#ifdef RTE_FORCE_INTRINSICS +static inline int +rte_spinlock_trylock (rte_spinlock_t *sl) +{ + int exp = 0; + return __atomic_compare_exchange_n(&sl->locked, &exp, 1, + 0, /* disallow spurious failure */ + __ATOMIC_ACQUIRE, __ATOMIC_RELAXED); +} +#endif + +/** + * Test if the lock is taken. + * + * @param sl + * A pointer to the spinlock. + * @return + * 1 if the lock is currently taken; 0 otherwise. + */ +static inline int rte_spinlock_is_locked (rte_spinlock_t *sl) +{ + return __atomic_load_n(&sl->locked, __ATOMIC_ACQUIRE); +} + +/** + * Test if hardware transactional memory (lock elision) is supported + * + * @return + * 1 if the hardware transactional memory is supported; 0 otherwise. + */ +static inline int rte_tm_supported(void); + +/** + * Try to execute critical section in a hardware memory transaction, + * if it fails or not available take the spinlock. + * + * NOTE: An attempt to perform a HW I/O operation inside a hardware memory + * transaction always aborts the transaction since the CPU is not able to + * roll-back should the transaction fail. Therefore, hardware transactional + * locks are not advised to be used around rte_eth_rx_burst() and + * rte_eth_tx_burst() calls. + * + * @param sl + * A pointer to the spinlock. + */ +static inline void +rte_spinlock_lock_tm(rte_spinlock_t *sl); + +/** + * Commit hardware memory transaction or release the spinlock if + * the spinlock is used as a fall-back + * + * @param sl + * A pointer to the spinlock. + */ +static inline void +rte_spinlock_unlock_tm(rte_spinlock_t *sl); + +/** + * Try to execute critical section in a hardware memory transaction, + * if it fails or not available try to take the lock. + * + * NOTE: An attempt to perform a HW I/O operation inside a hardware memory + * transaction always aborts the transaction since the CPU is not able to + * roll-back should the transaction fail. Therefore, hardware transactional + * locks are not advised to be used around rte_eth_rx_burst() and + * rte_eth_tx_burst() calls. + * + * @param sl + * A pointer to the spinlock. + * @return + * 1 if the hardware memory transaction is successfully started + * or lock is successfully taken; 0 otherwise. + */ +static inline int +rte_spinlock_trylock_tm(rte_spinlock_t *sl); + +/** + * The rte_spinlock_recursive_t type. + */ +typedef struct { + rte_spinlock_t sl; /**< the actual spinlock */ + volatile int user; /**< core id using lock, -1 for unused */ + volatile int count; /**< count of time this lock has been called */ +} rte_spinlock_recursive_t; + +/** + * A static recursive spinlock initializer. + */ +#define RTE_SPINLOCK_RECURSIVE_INITIALIZER {RTE_SPINLOCK_INITIALIZER, -1, 0} + +/** + * Initialize the recursive spinlock to an unlocked state. + * + * @param slr + * A pointer to the recursive spinlock. + */ +static inline void rte_spinlock_recursive_init(rte_spinlock_recursive_t *slr) +{ + rte_spinlock_init(&slr->sl); + slr->user = -1; + slr->count = 0; +} + +/** + * Take the recursive spinlock. + * + * @param slr + * A pointer to the recursive spinlock. + */ +static inline void rte_spinlock_recursive_lock(rte_spinlock_recursive_t *slr) +{ + int id = rte_gettid(); + + if (slr->user != id) { + rte_spinlock_lock(&slr->sl); + slr->user = id; + } + slr->count++; +} +/** + * Release the recursive spinlock. + * + * @param slr + * A pointer to the recursive spinlock. + */ +static inline void rte_spinlock_recursive_unlock(rte_spinlock_recursive_t *slr) +{ + if (--(slr->count) == 0) { + slr->user = -1; + rte_spinlock_unlock(&slr->sl); + } + +} + +/** + * Try to take the recursive lock. + * + * @param slr + * A pointer to the recursive spinlock. + * @return + * 1 if the lock is successfully taken; 0 otherwise. + */ +static inline int rte_spinlock_recursive_trylock(rte_spinlock_recursive_t *slr) +{ + int id = rte_gettid(); + + if (slr->user != id) { + if (rte_spinlock_trylock(&slr->sl) == 0) + return 0; + slr->user = id; + } + slr->count++; + return 1; +} + + +/** + * Try to execute critical section in a hardware memory transaction, + * if it fails or not available take the recursive spinlocks + * + * NOTE: An attempt to perform a HW I/O operation inside a hardware memory + * transaction always aborts the transaction since the CPU is not able to + * roll-back should the transaction fail. Therefore, hardware transactional + * locks are not advised to be used around rte_eth_rx_burst() and + * rte_eth_tx_burst() calls. + * + * @param slr + * A pointer to the recursive spinlock. + */ +static inline void rte_spinlock_recursive_lock_tm( + rte_spinlock_recursive_t *slr); + +/** + * Commit hardware memory transaction or release the recursive spinlock + * if the recursive spinlock is used as a fall-back + * + * @param slr + * A pointer to the recursive spinlock. + */ +static inline void rte_spinlock_recursive_unlock_tm( + rte_spinlock_recursive_t *slr); + +/** + * Try to execute critical section in a hardware memory transaction, + * if it fails or not available try to take the recursive lock + * + * NOTE: An attempt to perform a HW I/O operation inside a hardware memory + * transaction always aborts the transaction since the CPU is not able to + * roll-back should the transaction fail. Therefore, hardware transactional + * locks are not advised to be used around rte_eth_rx_burst() and + * rte_eth_tx_burst() calls. + * + * @param slr + * A pointer to the recursive spinlock. + * @return + * 1 if the hardware memory transaction is successfully started + * or lock is successfully taken; 0 otherwise. + */ +static inline int rte_spinlock_recursive_trylock_tm( + rte_spinlock_recursive_t *slr); + +#endif /* _RTE_SPINLOCK_H_ */ diff --git a/src/spdk/dpdk/lib/librte_eal/include/generic/rte_ticketlock.h b/src/spdk/dpdk/lib/librte_eal/include/generic/rte_ticketlock.h new file mode 100644 index 000000000..c295ae7f7 --- /dev/null +++ b/src/spdk/dpdk/lib/librte_eal/include/generic/rte_ticketlock.h @@ -0,0 +1,223 @@ +/* SPDX-License-Identifier: BSD-3-Clause + * Copyright(c) 2019 Arm Limited + */ + +#ifndef _RTE_TICKETLOCK_H_ +#define _RTE_TICKETLOCK_H_ + +/** + * @file + * + * RTE ticket locks + * + * This file defines an API for ticket locks, which give each waiting + * thread a ticket and take the lock one by one, first come, first + * serviced. + * + * All locks must be initialised before use, and only initialised once. + * + */ + +#ifdef __cplusplus +extern "C" { +#endif + +#include <rte_common.h> +#include <rte_lcore.h> +#include <rte_pause.h> + +/** + * The rte_ticketlock_t type. + */ +typedef union { + uint32_t tickets; + struct { + uint16_t current; + uint16_t next; + } s; +} rte_ticketlock_t; + +/** + * A static ticketlock initializer. + */ +#define RTE_TICKETLOCK_INITIALIZER { 0 } + +/** + * Initialize the ticketlock to an unlocked state. + * + * @param tl + * A pointer to the ticketlock. + */ +__rte_experimental +static inline void +rte_ticketlock_init(rte_ticketlock_t *tl) +{ + __atomic_store_n(&tl->tickets, 0, __ATOMIC_RELAXED); +} + +/** + * Take the ticketlock. + * + * @param tl + * A pointer to the ticketlock. + */ +__rte_experimental +static inline void +rte_ticketlock_lock(rte_ticketlock_t *tl) +{ + uint16_t me = __atomic_fetch_add(&tl->s.next, 1, __ATOMIC_RELAXED); + rte_wait_until_equal_16(&tl->s.current, me, __ATOMIC_ACQUIRE); +} + +/** + * Release the ticketlock. + * + * @param tl + * A pointer to the ticketlock. + */ +__rte_experimental +static inline void +rte_ticketlock_unlock(rte_ticketlock_t *tl) +{ + uint16_t i = __atomic_load_n(&tl->s.current, __ATOMIC_RELAXED); + __atomic_store_n(&tl->s.current, i + 1, __ATOMIC_RELEASE); +} + +/** + * Try to take the lock. + * + * @param tl + * A pointer to the ticketlock. + * @return + * 1 if the lock is successfully taken; 0 otherwise. + */ +__rte_experimental +static inline int +rte_ticketlock_trylock(rte_ticketlock_t *tl) +{ + rte_ticketlock_t old, new; + old.tickets = __atomic_load_n(&tl->tickets, __ATOMIC_RELAXED); + new.tickets = old.tickets; + new.s.next++; + if (old.s.next == old.s.current) { + if (__atomic_compare_exchange_n(&tl->tickets, &old.tickets, + new.tickets, 0, __ATOMIC_ACQUIRE, __ATOMIC_RELAXED)) + return 1; + } + + return 0; +} + +/** + * Test if the lock is taken. + * + * @param tl + * A pointer to the ticketlock. + * @return + * 1 if the lock is currently taken; 0 otherwise. + */ +__rte_experimental +static inline int +rte_ticketlock_is_locked(rte_ticketlock_t *tl) +{ + rte_ticketlock_t tic; + tic.tickets = __atomic_load_n(&tl->tickets, __ATOMIC_ACQUIRE); + return (tic.s.current != tic.s.next); +} + +/** + * The rte_ticketlock_recursive_t type. + */ +#define TICKET_LOCK_INVALID_ID -1 + +typedef struct { + rte_ticketlock_t tl; /**< the actual ticketlock */ + int user; /**< core id using lock, TICKET_LOCK_INVALID_ID for unused */ + unsigned int count; /**< count of time this lock has been called */ +} rte_ticketlock_recursive_t; + +/** + * A static recursive ticketlock initializer. + */ +#define RTE_TICKETLOCK_RECURSIVE_INITIALIZER {RTE_TICKETLOCK_INITIALIZER, \ + TICKET_LOCK_INVALID_ID, 0} + +/** + * Initialize the recursive ticketlock to an unlocked state. + * + * @param tlr + * A pointer to the recursive ticketlock. + */ +__rte_experimental +static inline void +rte_ticketlock_recursive_init(rte_ticketlock_recursive_t *tlr) +{ + rte_ticketlock_init(&tlr->tl); + __atomic_store_n(&tlr->user, TICKET_LOCK_INVALID_ID, __ATOMIC_RELAXED); + tlr->count = 0; +} + +/** + * Take the recursive ticketlock. + * + * @param tlr + * A pointer to the recursive ticketlock. + */ +__rte_experimental +static inline void +rte_ticketlock_recursive_lock(rte_ticketlock_recursive_t *tlr) +{ + int id = rte_gettid(); + + if (__atomic_load_n(&tlr->user, __ATOMIC_RELAXED) != id) { + rte_ticketlock_lock(&tlr->tl); + __atomic_store_n(&tlr->user, id, __ATOMIC_RELAXED); + } + tlr->count++; +} + +/** + * Release the recursive ticketlock. + * + * @param tlr + * A pointer to the recursive ticketlock. + */ +__rte_experimental +static inline void +rte_ticketlock_recursive_unlock(rte_ticketlock_recursive_t *tlr) +{ + if (--(tlr->count) == 0) { + __atomic_store_n(&tlr->user, TICKET_LOCK_INVALID_ID, + __ATOMIC_RELAXED); + rte_ticketlock_unlock(&tlr->tl); + } +} + +/** + * Try to take the recursive lock. + * + * @param tlr + * A pointer to the recursive ticketlock. + * @return + * 1 if the lock is successfully taken; 0 otherwise. + */ +__rte_experimental +static inline int +rte_ticketlock_recursive_trylock(rte_ticketlock_recursive_t *tlr) +{ + int id = rte_gettid(); + + if (__atomic_load_n(&tlr->user, __ATOMIC_RELAXED) != id) { + if (rte_ticketlock_trylock(&tlr->tl) == 0) + return 0; + __atomic_store_n(&tlr->user, id, __ATOMIC_RELAXED); + } + tlr->count++; + return 1; +} + +#ifdef __cplusplus +} +#endif + +#endif /* _RTE_TICKETLOCK_H_ */ diff --git a/src/spdk/dpdk/lib/librte_eal/include/generic/rte_vect.h b/src/spdk/dpdk/lib/librte_eal/include/generic/rte_vect.h new file mode 100644 index 000000000..3fc47979f --- /dev/null +++ b/src/spdk/dpdk/lib/librte_eal/include/generic/rte_vect.h @@ -0,0 +1,186 @@ +/* SPDX-License-Identifier: BSD-3-Clause + * Copyright 2016 6WIND S.A. + */ + +#ifndef _RTE_VECT_H_ +#define _RTE_VECT_H_ + +/** + * @file + * SIMD vector types + * + * This file defines types to use vector instructions with generic C code. + */ + +#include <stdint.h> + +/* Unsigned vector types */ + +/** + * 64 bits vector size to use with unsigned 8 bits elements. + * + * a = (rte_v64u8_t){ a0, a1, a2, a3, a4, a5, a6, a7 } + */ +typedef uint8_t rte_v64u8_t __attribute__((vector_size(8), aligned(8))); + +/** + * 64 bits vector size to use with unsigned 16 bits elements. + * + * a = (rte_v64u16_t){ a0, a1, a2, a3 } + */ +typedef uint16_t rte_v64u16_t __attribute__((vector_size(8), aligned(8))); + +/** + * 64 bits vector size to use with unsigned 32 bits elements. + * + * a = (rte_v64u32_t){ a0, a1 } + */ +typedef uint32_t rte_v64u32_t __attribute__((vector_size(8), aligned(8))); + +/** + * 128 bits vector size to use with unsigned 8 bits elements. + * + * a = (rte_v128u8_t){ a00, a01, a02, a03, a04, a05, a06, a07, + * a08, a09, a10, a11, a12, a13, a14, a15 } + */ +typedef uint8_t rte_v128u8_t __attribute__((vector_size(16), aligned(16))); + +/** + * 128 bits vector size to use with unsigned 16 bits elements. + * + * a = (rte_v128u16_t){ a0, a1, a2, a3, a4, a5, a6, a7 } + */ +typedef uint16_t rte_v128u16_t __attribute__((vector_size(16), aligned(16))); + +/** + * 128 bits vector size to use with unsigned 32 bits elements. + * + * a = (rte_v128u32_t){ a0, a1, a2, a3 } + */ +typedef uint32_t rte_v128u32_t __attribute__((vector_size(16), aligned(16))); + +/** + * 128 bits vector size to use with unsigned 64 bits elements. + * + * a = (rte_v128u64_t){ a0, a1 } + */ +typedef uint64_t rte_v128u64_t __attribute__((vector_size(16), aligned(16))); + +/** + * 256 bits vector size to use with unsigned 8 bits elements. + * + * a = (rte_v256u8_t){ a00, a01, a02, a03, a04, a05, a06, a07, + * a08, a09, a10, a11, a12, a13, a14, a15, + * a16, a17, a18, a19, a20, a21, a22, a23, + * a24, a25, a26, a27, a28, a29, a30, a31 } + */ +typedef uint8_t rte_v256u8_t __attribute__((vector_size(32), aligned(32))); + +/** + * 256 bits vector size to use with unsigned 16 bits elements. + * + * a = (rte_v256u16_t){ a00, a01, a02, a03, a04, a05, a06, a07, + * a08, a09, a10, a11, a12, a13, a14, a15 } + */ +typedef uint16_t rte_v256u16_t __attribute__((vector_size(32), aligned(32))); + +/** + * 256 bits vector size to use with unsigned 32 bits elements. + * + * a = (rte_v256u32_t){ a0, a1, a2, a3, a4, a5, a6, a7 } + */ +typedef uint32_t rte_v256u32_t __attribute__((vector_size(32), aligned(32))); + +/** + * 256 bits vector size to use with unsigned 64 bits elements. + * + * a = (rte_v256u64_t){ a0, a1, a2, a3 } + */ +typedef uint64_t rte_v256u64_t __attribute__((vector_size(32), aligned(32))); + + +/* Signed vector types */ + +/** + * 64 bits vector size to use with 8 bits elements. + * + * a = (rte_v64s8_t){ a0, a1, a2, a3, a4, a5, a6, a7 } + */ +typedef int8_t rte_v64s8_t __attribute__((vector_size(8), aligned(8))); + +/** + * 64 bits vector size to use with 16 bits elements. + * + * a = (rte_v64s16_t){ a0, a1, a2, a3 } + */ +typedef int16_t rte_v64s16_t __attribute__((vector_size(8), aligned(8))); + +/** + * 64 bits vector size to use with 32 bits elements. + * + * a = (rte_v64s32_t){ a0, a1 } + */ +typedef int32_t rte_v64s32_t __attribute__((vector_size(8), aligned(8))); + +/** + * 128 bits vector size to use with 8 bits elements. + * + * a = (rte_v128s8_t){ a00, a01, a02, a03, a04, a05, a06, a07, + * a08, a09, a10, a11, a12, a13, a14, a15 } + */ +typedef int8_t rte_v128s8_t __attribute__((vector_size(16), aligned(16))); + +/** + * 128 bits vector size to use with 16 bits elements. + * + * a = (rte_v128s16_t){ a0, a1, a2, a3, a4, a5, a6, a7 } + */ +typedef int16_t rte_v128s16_t __attribute__((vector_size(16), aligned(16))); + +/** + * 128 bits vector size to use with 32 bits elements. + * + * a = (rte_v128s32_t){ a0, a1, a2, a3 } + */ +typedef int32_t rte_v128s32_t __attribute__((vector_size(16), aligned(16))); + +/** + * 128 bits vector size to use with 64 bits elements. + * + * a = (rte_v128s64_t){ a1, a2 } + */ +typedef int64_t rte_v128s64_t __attribute__((vector_size(16), aligned(16))); + +/** + * 256 bits vector size to use with 8 bits elements. + * + * a = (rte_v256s8_t){ a00, a01, a02, a03, a04, a05, a06, a07, + * a08, a09, a10, a11, a12, a13, a14, a15, + * a16, a17, a18, a19, a20, a21, a22, a23, + * a24, a25, a26, a27, a28, a29, a30, a31 } + */ +typedef int8_t rte_v256s8_t __attribute__((vector_size(32), aligned(32))); + +/** + * 256 bits vector size to use with 16 bits elements. + * + * a = (rte_v256s16_t){ a00, a01, a02, a03, a04, a05, a06, a07, + * a08, a09, a10, a11, a12, a13, a14, a15 } + */ +typedef int16_t rte_v256s16_t __attribute__((vector_size(32), aligned(32))); + +/** + * 256 bits vector size to use with 32 bits elements. + * + * a = (rte_v256s32_t){ a0, a1, a2, a3, a4, a5, a6, a7 } + */ +typedef int32_t rte_v256s32_t __attribute__((vector_size(32), aligned(32))); + +/** + * 256 bits vector size to use with 64 bits elements. + * + * a = (rte_v256s64_t){ a0, a1, a2, a3 } + */ +typedef int64_t rte_v256s64_t __attribute__((vector_size(32), aligned(32))); + +#endif /* _RTE_VECT_H_ */ |