diff options
author | Ido Barnea <ibarnea@cisco.com> | 2015-12-09 05:07:44 +0200 |
---|---|---|
committer | Ido Barnea <ibarnea@cisco.com> | 2015-12-27 08:50:12 +0200 |
commit | 509648b87434b9032d38b8ca5ad470ba3edcc036 (patch) | |
tree | 920548ce9e2e5aeed4c88c1b288290505e7d7987 /src/dpdk_lib18/librte_sched | |
parent | b161dc672544a913f7f1ddf3a086dd75f2f1134a (diff) |
Adding dpdk 2.2 instead of dpdk 1.8 and making changes to make compilation work.
40G and 10G filters do not work yet.
Diffstat (limited to 'src/dpdk_lib18/librte_sched')
-rwxr-xr-x | src/dpdk_lib18/librte_sched/Makefile | 56 | ||||
-rwxr-xr-x | src/dpdk_lib18/librte_sched/rte_approx.c | 196 | ||||
-rwxr-xr-x | src/dpdk_lib18/librte_sched/rte_approx.h | 75 | ||||
-rwxr-xr-x | src/dpdk_lib18/librte_sched/rte_bitmap.h | 563 | ||||
-rwxr-xr-x | src/dpdk_lib18/librte_sched/rte_red.c | 158 | ||||
-rwxr-xr-x | src/dpdk_lib18/librte_sched/rte_red.h | 453 | ||||
-rwxr-xr-x | src/dpdk_lib18/librte_sched/rte_sched.c | 2150 | ||||
-rwxr-xr-x | src/dpdk_lib18/librte_sched/rte_sched.h | 442 | ||||
-rwxr-xr-x | src/dpdk_lib18/librte_sched/rte_sched_common.h | 129 |
9 files changed, 0 insertions, 4222 deletions
diff --git a/src/dpdk_lib18/librte_sched/Makefile b/src/dpdk_lib18/librte_sched/Makefile deleted file mode 100755 index 1a25b211..00000000 --- a/src/dpdk_lib18/librte_sched/Makefile +++ /dev/null @@ -1,56 +0,0 @@ -# BSD LICENSE -# -# Copyright(c) 2010-2014 Intel Corporation. All rights reserved. -# All rights reserved. -# -# Redistribution and use in source and binary forms, with or without -# modification, are permitted provided that the following conditions -# are met: -# -# * Redistributions of source code must retain the above copyright -# notice, this list of conditions and the following disclaimer. -# * Redistributions in binary form must reproduce the above copyright -# notice, this list of conditions and the following disclaimer in -# the documentation and/or other materials provided with the -# distribution. -# * Neither the name of Intel Corporation nor the names of its -# contributors may be used to endorse or promote products derived -# from this software without specific prior written permission. -# -# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS -# "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT -# LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR -# A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT -# OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, -# SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT -# LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, -# DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY -# THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT -# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE -# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. - -include $(RTE_SDK)/mk/rte.vars.mk - -# -# library name -# -LIB = librte_sched.a - -CFLAGS += -O3 -CFLAGS += $(WERROR_FLAGS) - -CFLAGS_rte_red.o := -D_GNU_SOURCE - -# -# all source are stored in SRCS-y -# -SRCS-$(CONFIG_RTE_LIBRTE_SCHED) += rte_sched.c rte_red.c rte_approx.c - -# install includes -SYMLINK-$(CONFIG_RTE_LIBRTE_SCHED)-include := rte_sched.h rte_bitmap.h rte_sched_common.h rte_red.h rte_approx.h - -# this lib depends upon: -DEPDIRS-$(CONFIG_RTE_LIBRTE_SCHED) += lib/librte_mempool lib/librte_mbuf -DEPDIRS-$(CONFIG_RTE_LIBRTE_SCHED) += lib/librte_net lib/librte_timer - -include $(RTE_SDK)/mk/rte.lib.mk diff --git a/src/dpdk_lib18/librte_sched/rte_approx.c b/src/dpdk_lib18/librte_sched/rte_approx.c deleted file mode 100755 index 771c9518..00000000 --- a/src/dpdk_lib18/librte_sched/rte_approx.c +++ /dev/null @@ -1,196 +0,0 @@ -/*- - * BSD LICENSE - * - * Copyright(c) 2010-2014 Intel Corporation. All rights reserved. - * All rights reserved. - * - * Redistribution and use in source and binary forms, with or without - * modification, are permitted provided that the following conditions - * are met: - * - * * Redistributions of source code must retain the above copyright - * notice, this list of conditions and the following disclaimer. - * * Redistributions in binary form must reproduce the above copyright - * notice, this list of conditions and the following disclaimer in - * the documentation and/or other materials provided with the - * distribution. - * * Neither the name of Intel Corporation nor the names of its - * contributors may be used to endorse or promote products derived - * from this software without specific prior written permission. - * - * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS - * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT - * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR - * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT - * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, - * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT - * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, - * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY - * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT - * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE - * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. - */ - -#include <stdlib.h> - -#include "rte_approx.h" - -/* - * Based on paper "Approximating Rational Numbers by Fractions" by Michal - * Forisek forisek@dcs.fmph.uniba.sk - * - * Given a rational number alpha with 0 < alpha < 1 and a precision d, the goal - * is to find positive integers p, q such that alpha - d < p/q < alpha + d, and - * q is minimal. - * - * http://people.ksp.sk/~misof/publications/2007approx.pdf - */ - -/* fraction comparison: compare (a/b) and (c/d) */ -static inline uint32_t -less(uint32_t a, uint32_t b, uint32_t c, uint32_t d) -{ - return (a*d < b*c); -} - -static inline uint32_t -less_or_equal(uint32_t a, uint32_t b, uint32_t c, uint32_t d) -{ - return (a*d <= b*c); -} - -/* check whether a/b is a valid approximation */ -static inline uint32_t -matches(uint32_t a, uint32_t b, - uint32_t alpha_num, uint32_t d_num, uint32_t denum) -{ - if (less_or_equal(a, b, alpha_num - d_num, denum)) - return 0; - - if (less(a ,b, alpha_num + d_num, denum)) - return 1; - - return 0; -} - -static inline void -find_exact_solution_left(uint32_t p_a, uint32_t q_a, uint32_t p_b, uint32_t q_b, - uint32_t alpha_num, uint32_t d_num, uint32_t denum, uint32_t *p, uint32_t *q) -{ - uint32_t k_num = denum * p_b - (alpha_num + d_num) * q_b; - uint32_t k_denum = (alpha_num + d_num) * q_a - denum * p_a; - uint32_t k = (k_num / k_denum) + 1; - - *p = p_b + k * p_a; - *q = q_b + k * q_a; -} - -static inline void -find_exact_solution_right(uint32_t p_a, uint32_t q_a, uint32_t p_b, uint32_t q_b, - uint32_t alpha_num, uint32_t d_num, uint32_t denum, uint32_t *p, uint32_t *q) -{ - uint32_t k_num = - denum * p_b + (alpha_num - d_num) * q_b; - uint32_t k_denum = - (alpha_num - d_num) * q_a + denum * p_a; - uint32_t k = (k_num / k_denum) + 1; - - *p = p_b + k * p_a; - *q = q_b + k * q_a; -} - -static int -find_best_rational_approximation(uint32_t alpha_num, uint32_t d_num, uint32_t denum, uint32_t *p, uint32_t *q) -{ - uint32_t p_a, q_a, p_b, q_b; - - /* check assumptions on the inputs */ - if (!((0 < d_num) && (d_num < alpha_num) && (alpha_num < denum) && (d_num + alpha_num < denum))) { - return -1; - } - - /* set initial bounds for the search */ - p_a = 0; - q_a = 1; - p_b = 1; - q_b = 1; - - while (1) { - uint32_t new_p_a, new_q_a, new_p_b, new_q_b; - uint32_t x_num, x_denum, x; - int aa, bb; - - /* compute the number of steps to the left */ - x_num = denum * p_b - alpha_num * q_b; - x_denum = - denum * p_a + alpha_num * q_a; - x = (x_num + x_denum - 1) / x_denum; /* x = ceil(x_num / x_denum) */ - - /* check whether we have a valid approximation */ - aa = matches(p_b + x * p_a, q_b + x * q_a, alpha_num, d_num, denum); - bb = matches(p_b + (x-1) * p_a, q_b + (x - 1) * q_a, alpha_num, d_num, denum); - if (aa || bb) { - find_exact_solution_left(p_a, q_a, p_b, q_b, alpha_num, d_num, denum, p, q); - return 0; - } - - /* update the interval */ - new_p_a = p_b + (x - 1) * p_a ; - new_q_a = q_b + (x - 1) * q_a; - new_p_b = p_b + x * p_a ; - new_q_b = q_b + x * q_a; - - p_a = new_p_a ; - q_a = new_q_a; - p_b = new_p_b ; - q_b = new_q_b; - - /* compute the number of steps to the right */ - x_num = alpha_num * q_b - denum * p_b; - x_denum = - alpha_num * q_a + denum * p_a; - x = (x_num + x_denum - 1) / x_denum; /* x = ceil(x_num / x_denum) */ - - /* check whether we have a valid approximation */ - aa = matches(p_b + x * p_a, q_b + x * q_a, alpha_num, d_num, denum); - bb = matches(p_b + (x - 1) * p_a, q_b + (x - 1) * q_a, alpha_num, d_num, denum); - if (aa || bb) { - find_exact_solution_right(p_a, q_a, p_b, q_b, alpha_num, d_num, denum, p, q); - return 0; - } - - /* update the interval */ - new_p_a = p_b + (x - 1) * p_a; - new_q_a = q_b + (x - 1) * q_a; - new_p_b = p_b + x * p_a; - new_q_b = q_b + x * q_a; - - p_a = new_p_a; - q_a = new_q_a; - p_b = new_p_b; - q_b = new_q_b; - } -} - -int rte_approx(double alpha, double d, uint32_t *p, uint32_t *q) -{ - uint32_t alpha_num, d_num, denum; - - /* Check input arguments */ - if (!((0.0 < d) && (d < alpha) && (alpha < 1.0))) { - return -1; - } - - if ((p == NULL) || (q == NULL)) { - return -2; - } - - /* Compute alpha_num, d_num and denum */ - denum = 1; - while (d < 1) { - alpha *= 10; - d *= 10; - denum *= 10; - } - alpha_num = (uint32_t) alpha; - d_num = (uint32_t) d; - - /* Perform approximation */ - return find_best_rational_approximation(alpha_num, d_num, denum, p, q); -} diff --git a/src/dpdk_lib18/librte_sched/rte_approx.h b/src/dpdk_lib18/librte_sched/rte_approx.h deleted file mode 100755 index 09f30a87..00000000 --- a/src/dpdk_lib18/librte_sched/rte_approx.h +++ /dev/null @@ -1,75 +0,0 @@ -/*- - * BSD LICENSE - * - * Copyright(c) 2010-2014 Intel Corporation. All rights reserved. - * All rights reserved. - * - * Redistribution and use in source and binary forms, with or without - * modification, are permitted provided that the following conditions - * are met: - * - * * Redistributions of source code must retain the above copyright - * notice, this list of conditions and the following disclaimer. - * * Redistributions in binary form must reproduce the above copyright - * notice, this list of conditions and the following disclaimer in - * the documentation and/or other materials provided with the - * distribution. - * * Neither the name of Intel Corporation nor the names of its - * contributors may be used to endorse or promote products derived - * from this software without specific prior written permission. - * - * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS - * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT - * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR - * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT - * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, - * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT - * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, - * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY - * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT - * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE - * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. - */ - -#ifndef __INCLUDE_RTE_APPROX_H__ -#define __INCLUDE_RTE_APPROX_H__ - -#ifdef __cplusplus -extern "C" { -#endif - -/** - * @file - * RTE Rational Approximation - * - * Given a rational number alpha with 0 < alpha < 1 and a precision d, the goal - * is to find positive integers p, q such that alpha - d < p/q < alpha + d, and - * q is minimal. - * - ***/ - -#include <stdint.h> - -/** - * Find best rational approximation - * - * @param alpha - * Rational number to approximate - * @param d - * Precision for the rational approximation - * @param p - * Pointer to pre-allocated space where the numerator of the rational - * approximation will be stored when operation is successful - * @param q - * Pointer to pre-allocated space where the denominator of the rational - * approximation will be stored when operation is successful - * @return - * 0 upon success, error code otherwise - */ -int rte_approx(double alpha, double d, uint32_t *p, uint32_t *q); - -#ifdef __cplusplus -} -#endif - -#endif /* __INCLUDE_RTE_APPROX_H__ */ diff --git a/src/dpdk_lib18/librte_sched/rte_bitmap.h b/src/dpdk_lib18/librte_sched/rte_bitmap.h deleted file mode 100755 index 95f3c0d3..00000000 --- a/src/dpdk_lib18/librte_sched/rte_bitmap.h +++ /dev/null @@ -1,563 +0,0 @@ -/*- - * BSD LICENSE - * - * Copyright(c) 2010-2014 Intel Corporation. All rights reserved. - * All rights reserved. - * - * Redistribution and use in source and binary forms, with or without - * modification, are permitted provided that the following conditions - * are met: - * - * * Redistributions of source code must retain the above copyright - * notice, this list of conditions and the following disclaimer. - * * Redistributions in binary form must reproduce the above copyright - * notice, this list of conditions and the following disclaimer in - * the documentation and/or other materials provided with the - * distribution. - * * Neither the name of Intel Corporation nor the names of its - * contributors may be used to endorse or promote products derived - * from this software without specific prior written permission. - * - * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS - * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT - * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR - * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT - * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, - * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT - * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, - * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY - * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT - * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE - * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. - */ - -#ifndef __INCLUDE_RTE_BITMAP_H__ -#define __INCLUDE_RTE_BITMAP_H__ - -#ifdef __cplusplus -extern "C" { -#endif - -/** - * @file - * RTE Bitmap - * - * The bitmap component provides a mechanism to manage large arrays of bits - * through bit get/set/clear and bit array scan operations. - * - * The bitmap scan operation is optimized for 64-bit CPUs using 64-byte cache - * lines. The bitmap is hierarchically organized using two arrays (array1 and - * array2), with each bit in array1 being associated with a full cache line - * (512 bits) of bitmap bits, which are stored in array2: the bit in array1 is - * set only when there is at least one bit set within its associated array2 - * bits, otherwise the bit in array1 is cleared. The read and write operations - * for array1 and array2 are always done in slabs of 64 bits. - * - * This bitmap is not thread safe. For lock free operation on a specific bitmap - * instance, a single writer thread performing bit set/clear operations is - * allowed, only the writer thread can do bitmap scan operations, while there - * can be several reader threads performing bit get operations in parallel with - * the writer thread. When the use of locking primitives is acceptable, the - * serialization of the bit set/clear and bitmap scan operations needs to be - * enforced by the caller, while the bit get operation does not require locking - * the bitmap. - * - ***/ - -#include <rte_common.h> -#include <rte_debug.h> -#include <rte_memory.h> -#include <rte_branch_prediction.h> -#include <rte_prefetch.h> - -#ifndef RTE_BITMAP_OPTIMIZATIONS -#define RTE_BITMAP_OPTIMIZATIONS 1 -#endif -#if RTE_BITMAP_OPTIMIZATIONS -#include <tmmintrin.h> -#endif - -/* Slab */ -#define RTE_BITMAP_SLAB_BIT_SIZE 64 -#define RTE_BITMAP_SLAB_BIT_SIZE_LOG2 6 -#define RTE_BITMAP_SLAB_BIT_MASK (RTE_BITMAP_SLAB_BIT_SIZE - 1) - -/* Cache line (CL) */ -#define RTE_BITMAP_CL_BIT_SIZE (RTE_CACHE_LINE_SIZE * 8) -#define RTE_BITMAP_CL_BIT_SIZE_LOG2 9 -#define RTE_BITMAP_CL_BIT_MASK (RTE_BITMAP_CL_BIT_SIZE - 1) - -#define RTE_BITMAP_CL_SLAB_SIZE (RTE_BITMAP_CL_BIT_SIZE / RTE_BITMAP_SLAB_BIT_SIZE) -#define RTE_BITMAP_CL_SLAB_SIZE_LOG2 3 -#define RTE_BITMAP_CL_SLAB_MASK (RTE_BITMAP_CL_SLAB_SIZE - 1) - -/** Bitmap data structure */ -struct rte_bitmap { - /* Context for array1 and array2 */ - uint64_t *array1; /**< Bitmap array1 */ - uint64_t *array2; /**< Bitmap array2 */ - uint32_t array1_size; /**< Number of 64-bit slabs in array1 that are actually used */ - uint32_t array2_size; /**< Number of 64-bit slabs in array2 */ - - /* Context for the "scan next" operation */ - uint32_t index1; /**< Bitmap scan: Index of current array1 slab */ - uint32_t offset1; /**< Bitmap scan: Offset of current bit within current array1 slab */ - uint32_t index2; /**< Bitmap scan: Index of current array2 slab */ - uint32_t go2; /**< Bitmap scan: Go/stop condition for current array2 cache line */ - - /* Storage space for array1 and array2 */ - uint8_t memory[0]; -}; - -static inline void -__rte_bitmap_index1_inc(struct rte_bitmap *bmp) -{ - bmp->index1 = (bmp->index1 + 1) & (bmp->array1_size - 1); -} - -static inline uint64_t -__rte_bitmap_mask1_get(struct rte_bitmap *bmp) -{ - return ((~1lu) << bmp->offset1); -} - -static inline void -__rte_bitmap_index2_set(struct rte_bitmap *bmp) -{ - bmp->index2 = (((bmp->index1 << RTE_BITMAP_SLAB_BIT_SIZE_LOG2) + bmp->offset1) << RTE_BITMAP_CL_SLAB_SIZE_LOG2); -} - -#if RTE_BITMAP_OPTIMIZATIONS - -static inline int -rte_bsf64(uint64_t slab, uint32_t *pos) -{ - if (likely(slab == 0)) { - return 0; - } - - *pos = __builtin_ctzll(slab); - return 1; -} - -#else - -static inline int -rte_bsf64(uint64_t slab, uint32_t *pos) -{ - uint64_t mask; - uint32_t i; - - if (likely(slab == 0)) { - return 0; - } - - for (i = 0, mask = 1; i < RTE_BITMAP_SLAB_BIT_SIZE; i ++, mask <<= 1) { - if (unlikely(slab & mask)) { - *pos = i; - return 1; - } - } - - return 0; -} - -#endif - -static inline uint32_t -__rte_bitmap_get_memory_footprint(uint32_t n_bits, - uint32_t *array1_byte_offset, uint32_t *array1_slabs, - uint32_t *array2_byte_offset, uint32_t *array2_slabs) -{ - uint32_t n_slabs_context, n_slabs_array1, n_cache_lines_context_and_array1; - uint32_t n_cache_lines_array2; - uint32_t n_bytes_total; - - n_cache_lines_array2 = (n_bits + RTE_BITMAP_CL_BIT_SIZE - 1) / RTE_BITMAP_CL_BIT_SIZE; - n_slabs_array1 = (n_cache_lines_array2 + RTE_BITMAP_SLAB_BIT_SIZE - 1) / RTE_BITMAP_SLAB_BIT_SIZE; - n_slabs_array1 = rte_align32pow2(n_slabs_array1); - n_slabs_context = (sizeof(struct rte_bitmap) + (RTE_BITMAP_SLAB_BIT_SIZE / 8) - 1) / (RTE_BITMAP_SLAB_BIT_SIZE / 8); - n_cache_lines_context_and_array1 = (n_slabs_context + n_slabs_array1 + RTE_BITMAP_CL_SLAB_SIZE - 1) / RTE_BITMAP_CL_SLAB_SIZE; - n_bytes_total = (n_cache_lines_context_and_array1 + n_cache_lines_array2) * RTE_CACHE_LINE_SIZE; - - if (array1_byte_offset) { - *array1_byte_offset = n_slabs_context * (RTE_BITMAP_SLAB_BIT_SIZE / 8); - } - if (array1_slabs) { - *array1_slabs = n_slabs_array1; - } - if (array2_byte_offset) { - *array2_byte_offset = n_cache_lines_context_and_array1 * RTE_CACHE_LINE_SIZE; - } - if (array2_slabs) { - *array2_slabs = n_cache_lines_array2 * RTE_BITMAP_CL_SLAB_SIZE; - } - - return n_bytes_total; -} - -static inline void -__rte_bitmap_scan_init(struct rte_bitmap *bmp) -{ - bmp->index1 = bmp->array1_size - 1; - bmp->offset1 = RTE_BITMAP_SLAB_BIT_SIZE - 1; - __rte_bitmap_index2_set(bmp); - bmp->index2 += RTE_BITMAP_CL_SLAB_SIZE; - - bmp->go2 = 0; -} - -/** - * Bitmap memory footprint calculation - * - * @param n_bits - * Number of bits in the bitmap - * @return - * Bitmap memory footprint measured in bytes on success, 0 on error - */ -static inline uint32_t -rte_bitmap_get_memory_footprint(uint32_t n_bits) { - /* Check input arguments */ - if (n_bits == 0) { - return 0; - } - - return __rte_bitmap_get_memory_footprint(n_bits, NULL, NULL, NULL, NULL); -} - -/** - * Bitmap initialization - * - * @param bmp - * Handle to bitmap instance - * @param array2 - * Base address of pre-allocated array2 - * @param n_bits - * Number of pre-allocated bits in array2. Must be non-zero and multiple of 512. - * @return - * 0 upon success, error code otherwise - */ -static inline struct rte_bitmap * -rte_bitmap_init(uint32_t n_bits, uint8_t *mem, uint32_t mem_size) -{ - struct rte_bitmap *bmp; - uint32_t array1_byte_offset, array1_slabs, array2_byte_offset, array2_slabs; - uint32_t size; - - /* Check input arguments */ - if (n_bits == 0) { - return NULL; - } - - if ((mem == NULL) || (((uintptr_t) mem) & RTE_CACHE_LINE_MASK)) { - return NULL; - } - - size = __rte_bitmap_get_memory_footprint(n_bits, - &array1_byte_offset, &array1_slabs, - &array2_byte_offset, &array2_slabs); - if (size < mem_size) { - return NULL; - } - - /* Setup bitmap */ - memset(mem, 0, size); - bmp = (struct rte_bitmap *) mem; - - bmp->array1 = (uint64_t *) &mem[array1_byte_offset]; - bmp->array1_size = array1_slabs; - bmp->array2 = (uint64_t *) &mem[array2_byte_offset]; - bmp->array2_size = array2_slabs; - - __rte_bitmap_scan_init(bmp); - - return bmp; -} - -/** - * Bitmap free - * - * @param bmp - * Handle to bitmap instance - * @return - * 0 upon success, error code otherwise - */ -static inline int -rte_bitmap_free(struct rte_bitmap *bmp) -{ - /* Check input arguments */ - if (bmp == NULL) { - return -1; - } - - return 0; -} - -/** - * Bitmap reset - * - * @param bmp - * Handle to bitmap instance - */ -static inline void -rte_bitmap_reset(struct rte_bitmap *bmp) -{ - memset(bmp->array1, 0, bmp->array1_size * sizeof(uint64_t)); - memset(bmp->array2, 0, bmp->array2_size * sizeof(uint64_t)); - __rte_bitmap_scan_init(bmp); -} - -/** - * Bitmap location prefetch into CPU L1 cache - * - * @param bmp - * Handle to bitmap instance - * @param pos - * Bit position - * @return - * 0 upon success, error code otherwise - */ -static inline void -rte_bitmap_prefetch0(struct rte_bitmap *bmp, uint32_t pos) -{ - uint64_t *slab2; - uint32_t index2; - - index2 = pos >> RTE_BITMAP_SLAB_BIT_SIZE_LOG2; - slab2 = bmp->array2 + index2; - rte_prefetch0((void *) slab2); -} - -/** - * Bitmap bit get - * - * @param bmp - * Handle to bitmap instance - * @param pos - * Bit position - * @return - * 0 when bit is cleared, non-zero when bit is set - */ -static inline uint64_t -rte_bitmap_get(struct rte_bitmap *bmp, uint32_t pos) -{ - uint64_t *slab2; - uint32_t index2, offset2; - - index2 = pos >> RTE_BITMAP_SLAB_BIT_SIZE_LOG2; - offset2 = pos & RTE_BITMAP_SLAB_BIT_MASK; - slab2 = bmp->array2 + index2; - return ((*slab2) & (1lu << offset2)); -} - -/** - * Bitmap bit set - * - * @param bmp - * Handle to bitmap instance - * @param pos - * Bit position - */ -static inline void -rte_bitmap_set(struct rte_bitmap *bmp, uint32_t pos) -{ - uint64_t *slab1, *slab2; - uint32_t index1, index2, offset1, offset2; - - /* Set bit in array2 slab and set bit in array1 slab */ - index2 = pos >> RTE_BITMAP_SLAB_BIT_SIZE_LOG2; - offset2 = pos & RTE_BITMAP_SLAB_BIT_MASK; - index1 = pos >> (RTE_BITMAP_SLAB_BIT_SIZE_LOG2 + RTE_BITMAP_CL_BIT_SIZE_LOG2); - offset1 = (pos >> RTE_BITMAP_CL_BIT_SIZE_LOG2) & RTE_BITMAP_SLAB_BIT_MASK; - slab2 = bmp->array2 + index2; - slab1 = bmp->array1 + index1; - - *slab2 |= 1lu << offset2; - *slab1 |= 1lu << offset1; -} - -/** - * Bitmap slab set - * - * @param bmp - * Handle to bitmap instance - * @param pos - * Bit position identifying the array2 slab - * @param slab - * Value to be assigned to the 64-bit slab in array2 - */ -static inline void -rte_bitmap_set_slab(struct rte_bitmap *bmp, uint32_t pos, uint64_t slab) -{ - uint64_t *slab1, *slab2; - uint32_t index1, index2, offset1; - - /* Set bits in array2 slab and set bit in array1 slab */ - index2 = pos >> RTE_BITMAP_SLAB_BIT_SIZE_LOG2; - index1 = pos >> (RTE_BITMAP_SLAB_BIT_SIZE_LOG2 + RTE_BITMAP_CL_BIT_SIZE_LOG2); - offset1 = (pos >> RTE_BITMAP_CL_BIT_SIZE_LOG2) & RTE_BITMAP_SLAB_BIT_MASK; - slab2 = bmp->array2 + index2; - slab1 = bmp->array1 + index1; - - *slab2 |= slab; - *slab1 |= 1lu << offset1; -} - -static inline uint64_t -__rte_bitmap_line_not_empty(uint64_t *slab2) -{ - uint64_t v1, v2, v3, v4; - - v1 = slab2[0] | slab2[1]; - v2 = slab2[2] | slab2[3]; - v3 = slab2[4] | slab2[5]; - v4 = slab2[6] | slab2[7]; - v1 |= v2; - v3 |= v4; - - return (v1 | v3); -} - -/** - * Bitmap bit clear - * - * @param bmp - * Handle to bitmap instance - * @param pos - * Bit position - */ -static inline void -rte_bitmap_clear(struct rte_bitmap *bmp, uint32_t pos) -{ - uint64_t *slab1, *slab2; - uint32_t index1, index2, offset1, offset2; - - /* Clear bit in array2 slab */ - index2 = pos >> RTE_BITMAP_SLAB_BIT_SIZE_LOG2; - offset2 = pos & RTE_BITMAP_SLAB_BIT_MASK; - slab2 = bmp->array2 + index2; - - /* Return if array2 slab is not all-zeros */ - *slab2 &= ~(1lu << offset2); - if (*slab2){ - return; - } - - /* Check the entire cache line of array2 for all-zeros */ - index2 &= ~ RTE_BITMAP_CL_SLAB_MASK; - slab2 = bmp->array2 + index2; - if (__rte_bitmap_line_not_empty(slab2)) { - return; - } - - /* The array2 cache line is all-zeros, so clear bit in array1 slab */ - index1 = pos >> (RTE_BITMAP_SLAB_BIT_SIZE_LOG2 + RTE_BITMAP_CL_BIT_SIZE_LOG2); - offset1 = (pos >> RTE_BITMAP_CL_BIT_SIZE_LOG2) & RTE_BITMAP_SLAB_BIT_MASK; - slab1 = bmp->array1 + index1; - *slab1 &= ~(1lu << offset1); - - return; -} - -static inline int -__rte_bitmap_scan_search(struct rte_bitmap *bmp) -{ - uint64_t value1; - uint32_t i; - - /* Check current array1 slab */ - value1 = bmp->array1[bmp->index1]; - value1 &= __rte_bitmap_mask1_get(bmp); - - if (rte_bsf64(value1, &bmp->offset1)) { - return 1; - } - - __rte_bitmap_index1_inc(bmp); - bmp->offset1 = 0; - - /* Look for another array1 slab */ - for (i = 0; i < bmp->array1_size; i ++, __rte_bitmap_index1_inc(bmp)) { - value1 = bmp->array1[bmp->index1]; - - if (rte_bsf64(value1, &bmp->offset1)) { - return 1; - } - } - - return 0; -} - -static inline void -__rte_bitmap_scan_read_init(struct rte_bitmap *bmp) -{ - __rte_bitmap_index2_set(bmp); - bmp->go2 = 1; - rte_prefetch1((void *)(bmp->array2 + bmp->index2 + 8)); -} - -static inline int -__rte_bitmap_scan_read(struct rte_bitmap *bmp, uint32_t *pos, uint64_t *slab) -{ - uint64_t *slab2; - - slab2 = bmp->array2 + bmp->index2; - for ( ; bmp->go2 ; bmp->index2 ++, slab2 ++, bmp->go2 = bmp->index2 & RTE_BITMAP_CL_SLAB_MASK) { - if (*slab2) { - *pos = bmp->index2 << RTE_BITMAP_SLAB_BIT_SIZE_LOG2; - *slab = *slab2; - - bmp->index2 ++; - slab2 ++; - bmp->go2 = bmp->index2 & RTE_BITMAP_CL_SLAB_MASK; - return 1; - } - } - - return 0; -} - -/** - * Bitmap scan (with automatic wrap-around) - * - * @param bmp - * Handle to bitmap instance - * @param pos - * When function call returns 1, pos contains the position of the next set - * bit, otherwise not modified - * @param slab - * When function call returns 1, slab contains the value of the entire 64-bit - * slab where the bit indicated by pos is located. Slabs are always 64-bit - * aligned, so the position of the first bit of the slab (this bit is not - * necessarily set) is pos / 64. Once a slab has been returned by the bitmap - * scan operation, the internal pointers of the bitmap are updated to point - * after this slab, so the same slab will not be returned again if it - * contains more than one bit which is set. When function call returns 0, - * slab is not modified. - * @return - * 0 if there is no bit set in the bitmap, 1 otherwise - */ -static inline int -rte_bitmap_scan(struct rte_bitmap *bmp, uint32_t *pos, uint64_t *slab) -{ - /* Return data from current array2 line if available */ - if (__rte_bitmap_scan_read(bmp, pos, slab)) { - return 1; - } - - /* Look for non-empty array2 line */ - if (__rte_bitmap_scan_search(bmp)) { - __rte_bitmap_scan_read_init(bmp); - __rte_bitmap_scan_read(bmp, pos, slab); - return 1; - } - - /* Empty bitmap */ - return 0; -} - -#ifdef __cplusplus -} -#endif - -#endif /* __INCLUDE_RTE_BITMAP_H__ */ diff --git a/src/dpdk_lib18/librte_sched/rte_red.c b/src/dpdk_lib18/librte_sched/rte_red.c deleted file mode 100755 index fdf40576..00000000 --- a/src/dpdk_lib18/librte_sched/rte_red.c +++ /dev/null @@ -1,158 +0,0 @@ -/*- - * BSD LICENSE - * - * Copyright(c) 2010-2014 Intel Corporation. All rights reserved. - * All rights reserved. - * - * Redistribution and use in source and binary forms, with or without - * modification, are permitted provided that the following conditions - * are met: - * - * * Redistributions of source code must retain the above copyright - * notice, this list of conditions and the following disclaimer. - * * Redistributions in binary form must reproduce the above copyright - * notice, this list of conditions and the following disclaimer in - * the documentation and/or other materials provided with the - * distribution. - * * Neither the name of Intel Corporation nor the names of its - * contributors may be used to endorse or promote products derived - * from this software without specific prior written permission. - * - * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS - * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT - * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR - * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT - * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, - * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT - * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, - * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY - * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT - * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE - * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. - */ - -#include <math.h> -#include "rte_red.h" -#include <rte_random.h> -#include <rte_common.h> - -#ifdef __INTEL_COMPILER -#pragma warning(disable:2259) /* conversion may lose significant bits */ -#endif - -static int rte_red_init_done = 0; /**< Flag to indicate that global initialisation is done */ -uint32_t rte_red_rand_val = 0; /**< Random value cache */ -uint32_t rte_red_rand_seed = 0; /**< Seed for random number generation */ - -/** - * table[i] = log2(1-Wq) * Scale * -1 - * Wq = 1/(2^i) - */ -uint16_t rte_red_log2_1_minus_Wq[RTE_RED_WQ_LOG2_NUM]; - -/** - * table[i] = 2^(i/16) * Scale - */ -uint16_t rte_red_pow2_frac_inv[16]; - -/** - * @brief Initialize tables used to compute average - * queue size when queue is empty. - */ -static void -__rte_red_init_tables(void) -{ - uint32_t i = 0; - double scale = 0.0; - double table_size = 0.0; - - scale = (double)(1 << RTE_RED_SCALING); - table_size = (double)(RTE_DIM(rte_red_pow2_frac_inv)); - - for (i = 0; i < RTE_DIM(rte_red_pow2_frac_inv); i++) { - double m = (double)i; - - rte_red_pow2_frac_inv[i] = (uint16_t) round(scale / pow(2, m / table_size)); - } - - scale = 1024.0; - - RTE_RED_ASSERT(RTE_RED_WQ_LOG2_NUM == RTE_DIM(rte_red_log2_1_minus_Wq)); - - for (i = RTE_RED_WQ_LOG2_MIN; i <= RTE_RED_WQ_LOG2_MAX; i++) { - double n = (double)i; - double Wq = pow(2, -n); - uint32_t index = i - RTE_RED_WQ_LOG2_MIN; - - rte_red_log2_1_minus_Wq[index] = (uint16_t) round(-1.0 * scale * log2(1.0 - Wq)); - /** - * Table entry of zero, corresponds to a Wq of zero - * which is not valid (avg would remain constant no - * matter how long the queue is empty). So we have - * to check for zero and round up to one. - */ - if (rte_red_log2_1_minus_Wq[index] == 0) { - rte_red_log2_1_minus_Wq[index] = 1; - } - } -} - -int -rte_red_rt_data_init(struct rte_red *red) -{ - if (red == NULL) - return -1; - - red->avg = 0; - red->count = 0; - red->q_time = 0; - return 0; -} - -int -rte_red_config_init(struct rte_red_config *red_cfg, - const uint16_t wq_log2, - const uint16_t min_th, - const uint16_t max_th, - const uint16_t maxp_inv) -{ - if (red_cfg == NULL) { - return -1; - } - if (max_th > RTE_RED_MAX_TH_MAX) { - return -2; - } - if (min_th >= max_th) { - return -3; - } - if (wq_log2 > RTE_RED_WQ_LOG2_MAX) { - return -4; - } - if (wq_log2 < RTE_RED_WQ_LOG2_MIN) { - return -5; - } - if (maxp_inv < RTE_RED_MAXP_INV_MIN) { - return -6; - } - if (maxp_inv > RTE_RED_MAXP_INV_MAX) { - return -7; - } - - /** - * Initialize the RED module if not already done - */ - if (!rte_red_init_done) { - rte_red_rand_seed = rte_rand(); - rte_red_rand_val = rte_fast_rand(); - __rte_red_init_tables(); - rte_red_init_done = 1; - } - - red_cfg->min_th = ((uint32_t) min_th) << (wq_log2 + RTE_RED_SCALING); - red_cfg->max_th = ((uint32_t) max_th) << (wq_log2 + RTE_RED_SCALING); - red_cfg->pa_const = (2 * (max_th - min_th) * maxp_inv) << RTE_RED_SCALING; - red_cfg->maxp_inv = maxp_inv; - red_cfg->wq_log2 = wq_log2; - - return 0; -} diff --git a/src/dpdk_lib18/librte_sched/rte_red.h b/src/dpdk_lib18/librte_sched/rte_red.h deleted file mode 100755 index 0d8412ff..00000000 --- a/src/dpdk_lib18/librte_sched/rte_red.h +++ /dev/null @@ -1,453 +0,0 @@ -/*- - * BSD LICENSE - * - * Copyright(c) 2010-2014 Intel Corporation. All rights reserved. - * All rights reserved. - * - * Redistribution and use in source and binary forms, with or without - * modification, are permitted provided that the following conditions - * are met: - * - * * Redistributions of source code must retain the above copyright - * notice, this list of conditions and the following disclaimer. - * * Redistributions in binary form must reproduce the above copyright - * notice, this list of conditions and the following disclaimer in - * the documentation and/or other materials provided with the - * distribution. - * * Neither the name of Intel Corporation nor the names of its - * contributors may be used to endorse or promote products derived - * from this software without specific prior written permission. - * - * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS - * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT - * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR - * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT - * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, - * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT - * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, - * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY - * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT - * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE - * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. - */ - -#ifndef __RTE_RED_H_INCLUDED__ -#define __RTE_RED_H_INCLUDED__ - -#ifdef __cplusplus -extern "C" { -#endif - -/** - * @file - * RTE Random Early Detection (RED) - * - * - ***/ - -#include <stdint.h> -#include <limits.h> -#include <rte_common.h> -#include <rte_debug.h> -#include <rte_cycles.h> -#include <rte_branch_prediction.h> - -#define RTE_RED_SCALING 10 /**< Fraction size for fixed-point */ -#define RTE_RED_S (1 << 22) /**< Packet size multiplied by number of leaf queues */ -#define RTE_RED_MAX_TH_MAX 1023 /**< Max threshold limit in fixed point format */ -#define RTE_RED_WQ_LOG2_MIN 1 /**< Min inverse filter weight value */ -#define RTE_RED_WQ_LOG2_MAX 12 /**< Max inverse filter weight value */ -#define RTE_RED_MAXP_INV_MIN 1 /**< Min inverse mark probability value */ -#define RTE_RED_MAXP_INV_MAX 255 /**< Max inverse mark probability value */ -#define RTE_RED_2POW16 (1<<16) /**< 2 power 16 */ -#define RTE_RED_INT16_NBITS (sizeof(uint16_t) * CHAR_BIT) -#define RTE_RED_WQ_LOG2_NUM (RTE_RED_WQ_LOG2_MAX - RTE_RED_WQ_LOG2_MIN + 1) - -#ifdef RTE_RED_DEBUG - -#define RTE_RED_ASSERT(exp) \ -if (!(exp)) { \ - rte_panic("line%d\tassert \"" #exp "\" failed\n", __LINE__); \ -} - -#else - -#define RTE_RED_ASSERT(exp) do { } while(0) - -#endif /* RTE_RED_DEBUG */ - -/** - * Externs - * - */ -extern uint32_t rte_red_rand_val; -extern uint32_t rte_red_rand_seed; -extern uint16_t rte_red_log2_1_minus_Wq[RTE_RED_WQ_LOG2_NUM]; -extern uint16_t rte_red_pow2_frac_inv[16]; - -/** - * RED configuration parameters passed by user - * - */ -struct rte_red_params { - uint16_t min_th; /**< Minimum threshold for queue (max_th) */ - uint16_t max_th; /**< Maximum threshold for queue (max_th) */ - uint16_t maxp_inv; /**< Inverse of packet marking probability maximum value (maxp = 1 / maxp_inv) */ - uint16_t wq_log2; /**< Negated log2 of queue weight (wq = 1 / (2 ^ wq_log2)) */ -}; - -/** - * RED configuration parameters - */ -struct rte_red_config { - uint32_t min_th; /**< min_th scaled in fixed-point format */ - uint32_t max_th; /**< max_th scaled in fixed-point format */ - uint32_t pa_const; /**< Precomputed constant value used for pa calculation (scaled in fixed-point format) */ - uint8_t maxp_inv; /**< maxp_inv */ - uint8_t wq_log2; /**< wq_log2 */ -}; - -/** - * RED run-time data - */ -struct rte_red { - uint32_t avg; /**< Average queue size (avg), scaled in fixed-point format */ - uint32_t count; /**< Number of packets since last marked packet (count) */ - uint64_t q_time; /**< Start of the queue idle time (q_time) */ -}; - -/** - * @brief Initialises run-time data - * - * @param [in,out] data pointer to RED runtime data - * - * @return Operation status - * @retval 0 success - * @retval !0 error - */ -int -rte_red_rt_data_init(struct rte_red *red); - -/** - * @brief Configures a single RED configuration parameter structure. - * - * @param [in,out] config pointer to a RED configuration parameter structure - * @param [in] wq_log2 log2 of the filter weight, valid range is: - * RTE_RED_WQ_LOG2_MIN <= wq_log2 <= RTE_RED_WQ_LOG2_MAX - * @param [in] min_th queue minimum threshold in number of packets - * @param [in] max_th queue maximum threshold in number of packets - * @param [in] maxp_inv inverse maximum mark probability - * - * @return Operation status - * @retval 0 success - * @retval !0 error - */ -int -rte_red_config_init(struct rte_red_config *red_cfg, - const uint16_t wq_log2, - const uint16_t min_th, - const uint16_t max_th, - const uint16_t maxp_inv); - -/** - * @brief Generate random number for RED - * - * Implemenetation based on: - * http://software.intel.com/en-us/articles/fast-random-number-generator-on-the-intel-pentiumr-4-processor/ - * - * 10 bit shift has been found through empirical tests (was 16). - * - * @return Random number between 0 and (2^22 - 1) - */ -static inline uint32_t -rte_fast_rand(void) -{ - rte_red_rand_seed = (214013 * rte_red_rand_seed) + 2531011; - return (rte_red_rand_seed >> 10); -} - -/** - * @brief calculate factor to scale average queue size when queue - * becomes empty - * - * @param [in] wq_log2, where EWMA filter weight wq = 1/(2 ^ wq_log2) - * @param [in] m exponent in the computed value (1 - wq) ^ m - * - * @return computed value - * @retval ((1 - wq) ^ m) scaled in fixed-point format - */ -static inline uint16_t -__rte_red_calc_qempty_factor(uint8_t wq_log2, uint16_t m) -{ - uint32_t n = 0; - uint32_t f = 0; - - /** - * Basic math tells us that: - * a^b = 2^(b * log2(a) ) - * - * in our case: - * a = (1-Wq) - * b = m - * Wq = 1/ (2^log2n) - * - * So we are computing this equation: - * factor = 2 ^ ( m * log2(1-Wq)) - * - * First we are computing: - * n = m * log2(1-Wq) - * - * To avoid dealing with signed numbers log2 values are positive - * but they should be negative because (1-Wq) is always < 1. - * Contents of log2 table values are also scaled for precision. - */ - - n = m * rte_red_log2_1_minus_Wq[wq_log2 - RTE_RED_WQ_LOG2_MIN]; - - /** - * The tricky part is computing 2^n, for this I split n into - * integer part and fraction part. - * f - is fraction part of n - * n - is integer part of original n - * - * Now using basic math we compute 2^n: - * 2^(f+n) = 2^f * 2^n - * 2^f - we use lookup table - * 2^n - can be replaced with bit shift right oeprations - */ - - f = (n >> 6) & 0xf; - n >>= 10; - - if (n < RTE_RED_SCALING) - return (uint16_t) ((rte_red_pow2_frac_inv[f] + (1 << (n - 1))) >> n); - - return 0; -} - -/** - * @brief Updates queue average in condition when queue is empty - * - * Note: packet is never dropped in this particular case. - * - * @param [in] config pointer to a RED configuration parameter structure - * @param [in,out] data pointer to RED runtime data - * @param [in] time current time stamp - * - * @return Operation status - * @retval 0 enqueue the packet - * @retval 1 drop the packet based on max threshold criterion - * @retval 2 drop the packet based on mark probability criterion - */ -static inline int -rte_red_enqueue_empty(const struct rte_red_config *red_cfg, - struct rte_red *red, - const uint64_t time) -{ - uint64_t time_diff = 0, m = 0; - - RTE_RED_ASSERT(red_cfg != NULL); - RTE_RED_ASSERT(red != NULL); - - red->count ++; - - /** - * We compute avg but we don't compare avg against - * min_th or max_th, nor calculate drop probability - */ - time_diff = time - red->q_time; - - /** - * m is the number of packets that might have arrived while the queue was empty. - * In this case we have time stamps provided by scheduler in byte units (bytes - * transmitted on network port). Such time stamp translates into time units as - * port speed is fixed but such approach simplifies the code. - */ - m = time_diff / RTE_RED_S; - - /** - * Check that m will fit into 16-bit unsigned integer - */ - if (m >= RTE_RED_2POW16) { - red->avg = 0; - } else { - red->avg = (red->avg >> RTE_RED_SCALING) * __rte_red_calc_qempty_factor(red_cfg->wq_log2, (uint16_t) m); - } - - return 0; -} - -/** - * Drop probability (Sally Floyd and Van Jacobson): - * - * pb = (1 / maxp_inv) * (avg - min_th) / (max_th - min_th) - * pa = pb / (2 - count * pb) - * - * - * (1 / maxp_inv) * (avg - min_th) - * --------------------------------- - * max_th - min_th - * pa = ----------------------------------------------- - * count * (1 / maxp_inv) * (avg - min_th) - * 2 - ----------------------------------------- - * max_th - min_th - * - * - * avg - min_th - * pa = ----------------------------------------------------------- - * 2 * (max_th - min_th) * maxp_inv - count * (avg - min_th) - * - * - * We define pa_const as: pa_const = 2 * (max_th - min_th) * maxp_inv. Then: - * - * - * avg - min_th - * pa = ----------------------------------- - * pa_const - count * (avg - min_th) - */ - -/** - * @brief make a decision to drop or enqueue a packet based on mark probability - * criteria - * - * @param [in] config pointer to structure defining RED parameters - * @param [in,out] data pointer to RED runtime data - * - * @return operation status - * @retval 0 enqueue the packet - * @retval 1 drop the packet - */ -static inline int -__rte_red_drop(const struct rte_red_config *red_cfg, struct rte_red *red) -{ - uint32_t pa_num = 0; /* numerator of drop-probability */ - uint32_t pa_den = 0; /* denominator of drop-probability */ - uint32_t pa_num_count = 0; - - pa_num = (red->avg - red_cfg->min_th) >> (red_cfg->wq_log2); - - pa_num_count = red->count * pa_num; - - if (red_cfg->pa_const <= pa_num_count) - return 1; - - pa_den = red_cfg->pa_const - pa_num_count; - - /* If drop, generate and save random number to be used next time */ - if (unlikely((rte_red_rand_val % pa_den) < pa_num)) { - rte_red_rand_val = rte_fast_rand(); - - return 1; - } - - /* No drop */ - return 0; -} - -/** - * @brief Decides if new packet should be enqeued or dropped in queue non-empty case - * - * @param [in] config pointer to a RED configuration parameter structure - * @param [in,out] data pointer to RED runtime data - * @param [in] q current queue size (measured in packets) - * - * @return Operation status - * @retval 0 enqueue the packet - * @retval 1 drop the packet based on max threshold criterion - * @retval 2 drop the packet based on mark probability criterion - */ -static inline int -rte_red_enqueue_nonempty(const struct rte_red_config *red_cfg, - struct rte_red *red, - const unsigned q) -{ - RTE_RED_ASSERT(red_cfg != NULL); - RTE_RED_ASSERT(red != NULL); - - /** - * EWMA filter (Sally Floyd and Van Jacobson): - * avg = (1 - wq) * avg + wq * q - * avg = avg + q * wq - avg * wq - * - * We select: wq = 2^(-n). Let scaled version of avg be: avg_s = avg * 2^(N+n). We get: - * avg_s = avg_s + q * 2^N - avg_s * 2^(-n) - * - * By using shift left/right operations, we get: - * avg_s = avg_s + (q << N) - (avg_s >> n) - * avg_s += (q << N) - (avg_s >> n) - */ - - /* avg update */ - red->avg += (q << RTE_RED_SCALING) - (red->avg >> red_cfg->wq_log2); - - /* avg < min_th: do not mark the packet */ - if (red->avg < red_cfg->min_th) { - red->count ++; - return 0; - } - - /* min_th <= avg < max_th: mark the packet with pa probability */ - if (red->avg < red_cfg->max_th) { - if (!__rte_red_drop(red_cfg, red)) { - red->count ++; - return 0; - } - - red->count = 0; - return 2; - } - - /* max_th <= avg: always mark the packet */ - red->count = 0; - return 1; -} - -/** - * @brief Decides if new packet should be enqeued or dropped - * Updates run time data based on new queue size value. - * Based on new queue average and RED configuration parameters - * gives verdict whether to enqueue or drop the packet. - * - * @param [in] config pointer to a RED configuration parameter structure - * @param [in,out] data pointer to RED runtime data - * @param [in] q updated queue size in packets - * @param [in] time current time stamp - * - * @return Operation status - * @retval 0 enqueue the packet - * @retval 1 drop the packet based on max threshold criteria - * @retval 2 drop the packet based on mark probability criteria - */ -static inline int -rte_red_enqueue(const struct rte_red_config *red_cfg, - struct rte_red *red, - const unsigned q, - const uint64_t time) -{ - RTE_RED_ASSERT(red_cfg != NULL); - RTE_RED_ASSERT(red != NULL); - - if (q != 0) { - return rte_red_enqueue_nonempty(red_cfg, red, q); - } else { - return rte_red_enqueue_empty(red_cfg, red, time); - } -} - -/** - * @brief Callback to records time that queue became empty - * - * @param [in,out] data pointer to RED runtime data - * @param [in] time current time stamp - */ -static inline void -rte_red_mark_queue_empty(struct rte_red *red, const uint64_t time) -{ - red->q_time = time; -} - -#ifdef __cplusplus -} -#endif - -#endif /* __RTE_RED_H_INCLUDED__ */ diff --git a/src/dpdk_lib18/librte_sched/rte_sched.c b/src/dpdk_lib18/librte_sched/rte_sched.c deleted file mode 100755 index 95dee273..00000000 --- a/src/dpdk_lib18/librte_sched/rte_sched.c +++ /dev/null @@ -1,2150 +0,0 @@ -/*- - * BSD LICENSE - * - * Copyright(c) 2010-2014 Intel Corporation. All rights reserved. - * All rights reserved. - * - * Redistribution and use in source and binary forms, with or without - * modification, are permitted provided that the following conditions - * are met: - * - * * Redistributions of source code must retain the above copyright - * notice, this list of conditions and the following disclaimer. - * * Redistributions in binary form must reproduce the above copyright - * notice, this list of conditions and the following disclaimer in - * the documentation and/or other materials provided with the - * distribution. - * * Neither the name of Intel Corporation nor the names of its - * contributors may be used to endorse or promote products derived - * from this software without specific prior written permission. - * - * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS - * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT - * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR - * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT - * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, - * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT - * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, - * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY - * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT - * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE - * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. - */ - -#include <stdio.h> -#include <string.h> - -#include <rte_common.h> -#include <rte_log.h> -#include <rte_memory.h> -#include <rte_malloc.h> -#include <rte_cycles.h> -#include <rte_prefetch.h> -#include <rte_branch_prediction.h> -#include <rte_mbuf.h> - -#include "rte_sched.h" -#include "rte_bitmap.h" -#include "rte_sched_common.h" -#include "rte_approx.h" - -#ifdef __INTEL_COMPILER -#pragma warning(disable:2259) /* conversion may lose significant bits */ -#endif - -#ifndef RTE_SCHED_DEBUG -#define RTE_SCHED_DEBUG 0 -#endif - -#ifndef RTE_SCHED_OPTIMIZATIONS -#define RTE_SCHED_OPTIMIZATIONS 0 -#endif - -#if RTE_SCHED_OPTIMIZATIONS -#include <immintrin.h> -#endif - -#define RTE_SCHED_ENQUEUE 1 - -#define RTE_SCHED_TS 1 - -#if RTE_SCHED_TS == 0 /* Infinite credits. Traffic shaping disabled. */ -#define RTE_SCHED_TS_CREDITS_UPDATE 0 -#define RTE_SCHED_TS_CREDITS_CHECK 0 -#else /* Real Credits. Full traffic shaping implemented. */ -#define RTE_SCHED_TS_CREDITS_UPDATE 1 -#define RTE_SCHED_TS_CREDITS_CHECK 1 -#endif - -#ifndef RTE_SCHED_TB_RATE_CONFIG_ERR -#define RTE_SCHED_TB_RATE_CONFIG_ERR (1e-7) -#endif - -#define RTE_SCHED_WRR 1 - -#ifndef RTE_SCHED_WRR_SHIFT -#define RTE_SCHED_WRR_SHIFT 3 -#endif - -#ifndef RTE_SCHED_PORT_N_GRINDERS -#define RTE_SCHED_PORT_N_GRINDERS 8 -#endif -#if (RTE_SCHED_PORT_N_GRINDERS == 0) || (RTE_SCHED_PORT_N_GRINDERS & (RTE_SCHED_PORT_N_GRINDERS - 1)) -#error Number of grinders must be non-zero and a power of 2 -#endif -#if (RTE_SCHED_OPTIMIZATIONS && (RTE_SCHED_PORT_N_GRINDERS != 8)) -#error Number of grinders must be 8 when RTE_SCHED_OPTIMIZATIONS is set -#endif - -#define RTE_SCHED_GRINDER_PCACHE_SIZE (64 / RTE_SCHED_QUEUES_PER_PIPE) - -#define RTE_SCHED_PIPE_INVALID UINT32_MAX - -#define RTE_SCHED_BMP_POS_INVALID UINT32_MAX - -struct rte_sched_subport { - /* Token bucket (TB) */ - uint64_t tb_time; /* time of last update */ - uint32_t tb_period; - uint32_t tb_credits_per_period; - uint32_t tb_size; - uint32_t tb_credits; - - /* Traffic classes (TCs) */ - uint64_t tc_time; /* time of next update */ - uint32_t tc_credits_per_period[RTE_SCHED_TRAFFIC_CLASSES_PER_PIPE]; - uint32_t tc_credits[RTE_SCHED_TRAFFIC_CLASSES_PER_PIPE]; - uint32_t tc_period; - - /* TC oversubscription */ - uint32_t tc_ov_wm; - uint32_t tc_ov_wm_min; - uint32_t tc_ov_wm_max; - uint8_t tc_ov_period_id; - uint8_t tc_ov; - uint32_t tc_ov_n; - double tc_ov_rate; - - /* Statistics */ - struct rte_sched_subport_stats stats; -}; - -struct rte_sched_pipe_profile { - /* Token bucket (TB) */ - uint32_t tb_period; - uint32_t tb_credits_per_period; - uint32_t tb_size; - - /* Pipe traffic classes */ - uint32_t tc_period; - uint32_t tc_credits_per_period[RTE_SCHED_TRAFFIC_CLASSES_PER_PIPE]; - uint8_t tc_ov_weight; - - /* Pipe queues */ - uint8_t wrr_cost[RTE_SCHED_QUEUES_PER_PIPE]; -}; - -struct rte_sched_pipe { - /* Token bucket (TB) */ - uint64_t tb_time; /* time of last update */ - uint32_t tb_credits; - - /* Pipe profile and flags */ - uint32_t profile; - - /* Traffic classes (TCs) */ - uint64_t tc_time; /* time of next update */ - uint32_t tc_credits[RTE_SCHED_TRAFFIC_CLASSES_PER_PIPE]; - - /* Weighted Round Robin (WRR) */ - uint8_t wrr_tokens[RTE_SCHED_QUEUES_PER_PIPE]; - - /* TC oversubscription */ - uint32_t tc_ov_credits; - uint8_t tc_ov_period_id; - uint8_t reserved[3]; -} __rte_cache_aligned; - -struct rte_sched_queue { - uint16_t qw; - uint16_t qr; -}; - -struct rte_sched_queue_extra { - struct rte_sched_queue_stats stats; -#ifdef RTE_SCHED_RED - struct rte_red red; -#endif -}; - -enum grinder_state { - e_GRINDER_PREFETCH_PIPE = 0, - e_GRINDER_PREFETCH_TC_QUEUE_ARRAYS, - e_GRINDER_PREFETCH_MBUF, - e_GRINDER_READ_MBUF -}; - -struct rte_sched_grinder { - /* Pipe cache */ - uint16_t pcache_qmask[RTE_SCHED_GRINDER_PCACHE_SIZE]; - uint32_t pcache_qindex[RTE_SCHED_GRINDER_PCACHE_SIZE]; - uint32_t pcache_w; - uint32_t pcache_r; - - /* Current pipe */ - enum grinder_state state; - uint32_t productive; - uint32_t pindex; - struct rte_sched_subport *subport; - struct rte_sched_pipe *pipe; - struct rte_sched_pipe_profile *pipe_params; - - /* TC cache */ - uint8_t tccache_qmask[4]; - uint32_t tccache_qindex[4]; - uint32_t tccache_w; - uint32_t tccache_r; - - /* Current TC */ - uint32_t tc_index; - struct rte_sched_queue *queue[RTE_SCHED_TRAFFIC_CLASSES_PER_PIPE]; - struct rte_mbuf **qbase[RTE_SCHED_TRAFFIC_CLASSES_PER_PIPE]; - uint32_t qindex[RTE_SCHED_TRAFFIC_CLASSES_PER_PIPE]; - uint16_t qsize; - uint32_t qmask; - uint32_t qpos; - struct rte_mbuf *pkt; - - /* WRR */ - uint16_t wrr_tokens[RTE_SCHED_QUEUES_PER_TRAFFIC_CLASS]; - uint16_t wrr_mask[RTE_SCHED_QUEUES_PER_TRAFFIC_CLASS]; - uint8_t wrr_cost[RTE_SCHED_QUEUES_PER_TRAFFIC_CLASS]; -}; - -struct rte_sched_port { - /* User parameters */ - uint32_t n_subports_per_port; - uint32_t n_pipes_per_subport; - uint32_t rate; - uint32_t mtu; - uint32_t frame_overhead; - uint16_t qsize[RTE_SCHED_TRAFFIC_CLASSES_PER_PIPE]; - uint32_t n_pipe_profiles; - uint32_t pipe_tc3_rate_max; -#ifdef RTE_SCHED_RED - struct rte_red_config red_config[RTE_SCHED_TRAFFIC_CLASSES_PER_PIPE][e_RTE_METER_COLORS]; -#endif - - /* Timing */ - uint64_t time_cpu_cycles; /* Current CPU time measured in CPU cyles */ - uint64_t time_cpu_bytes; /* Current CPU time measured in bytes */ - uint64_t time; /* Current NIC TX time measured in bytes */ - double cycles_per_byte; /* CPU cycles per byte */ - - /* Scheduling loop detection */ - uint32_t pipe_loop; - uint32_t pipe_exhaustion; - - /* Bitmap */ - struct rte_bitmap *bmp; - uint32_t grinder_base_bmp_pos[RTE_SCHED_PORT_N_GRINDERS] __rte_aligned_16; - - /* Grinders */ - struct rte_sched_grinder grinder[RTE_SCHED_PORT_N_GRINDERS]; - uint32_t busy_grinders; - struct rte_mbuf **pkts_out; - uint32_t n_pkts_out; - - /* Queue base calculation */ - uint32_t qsize_add[RTE_SCHED_QUEUES_PER_PIPE]; - uint32_t qsize_sum; - - /* Large data structures */ - struct rte_sched_subport *subport; - struct rte_sched_pipe *pipe; - struct rte_sched_queue *queue; - struct rte_sched_queue_extra *queue_extra; - struct rte_sched_pipe_profile *pipe_profiles; - uint8_t *bmp_array; - struct rte_mbuf **queue_array; - uint8_t memory[0] __rte_cache_aligned; -} __rte_cache_aligned; - -enum rte_sched_port_array { - e_RTE_SCHED_PORT_ARRAY_SUBPORT = 0, - e_RTE_SCHED_PORT_ARRAY_PIPE, - e_RTE_SCHED_PORT_ARRAY_QUEUE, - e_RTE_SCHED_PORT_ARRAY_QUEUE_EXTRA, - e_RTE_SCHED_PORT_ARRAY_PIPE_PROFILES, - e_RTE_SCHED_PORT_ARRAY_BMP_ARRAY, - e_RTE_SCHED_PORT_ARRAY_QUEUE_ARRAY, - e_RTE_SCHED_PORT_ARRAY_TOTAL, -}; - -#ifdef RTE_SCHED_COLLECT_STATS - -static inline uint32_t -rte_sched_port_queues_per_subport(struct rte_sched_port *port) -{ - return RTE_SCHED_QUEUES_PER_PIPE * port->n_pipes_per_subport; -} - -#endif - -static inline uint32_t -rte_sched_port_queues_per_port(struct rte_sched_port *port) -{ - return RTE_SCHED_QUEUES_PER_PIPE * port->n_pipes_per_subport * port->n_subports_per_port; -} - -static int -rte_sched_port_check_params(struct rte_sched_port_params *params) -{ - uint32_t i, j; - - if (params == NULL) { - return -1; - } - - /* socket */ - if ((params->socket < 0) || (params->socket >= RTE_MAX_NUMA_NODES)) { - return -3; - } - - /* rate */ - if (params->rate == 0) { - return -4; - } - - /* mtu */ - if (params->mtu == 0) { - return -5; - } - - /* n_subports_per_port: non-zero, power of 2 */ - if ((params->n_subports_per_port == 0) || (!rte_is_power_of_2(params->n_subports_per_port))) { - return -6; - } - - /* n_pipes_per_subport: non-zero, power of 2 */ - if ((params->n_pipes_per_subport == 0) || (!rte_is_power_of_2(params->n_pipes_per_subport))) { - return -7; - } - - /* qsize: non-zero, power of 2, no bigger than 32K (due to 16-bit read/write pointers) */ - for (i = 0; i < RTE_SCHED_TRAFFIC_CLASSES_PER_PIPE; i ++) { - uint16_t qsize = params->qsize[i]; - - if ((qsize == 0) || (!rte_is_power_of_2(qsize))) { - return -8; - } - } - - /* pipe_profiles and n_pipe_profiles */ - if ((params->pipe_profiles == NULL) || - (params->n_pipe_profiles == 0) || - (params->n_pipe_profiles > RTE_SCHED_PIPE_PROFILES_PER_PORT)) { - return -9; - } - - for (i = 0; i < params->n_pipe_profiles; i ++) { - struct rte_sched_pipe_params *p = params->pipe_profiles + i; - - /* TB rate: non-zero, not greater than port rate */ - if ((p->tb_rate == 0) || (p->tb_rate > params->rate)) { - return -10; - } - - /* TB size: non-zero */ - if (p->tb_size == 0) { - return -11; - } - - /* TC rate: non-zero, less than pipe rate */ - for (j = 0; j < RTE_SCHED_TRAFFIC_CLASSES_PER_PIPE; j ++) { - if ((p->tc_rate[j] == 0) || (p->tc_rate[j] > p->tb_rate)) { - return -12; - } - } - - /* TC period: non-zero */ - if (p->tc_period == 0) { - return -13; - } - -#ifdef RTE_SCHED_SUBPORT_TC_OV - /* TC3 oversubscription weight: non-zero */ - if (p->tc_ov_weight == 0) { - return -14; - } -#endif - - /* Queue WRR weights: non-zero */ - for (j = 0; j < RTE_SCHED_QUEUES_PER_PIPE; j ++) { - if (p->wrr_weights[j] == 0) { - return -15; - } - } - } - - return 0; -} - -static uint32_t -rte_sched_port_get_array_base(struct rte_sched_port_params *params, enum rte_sched_port_array array) -{ - uint32_t n_subports_per_port = params->n_subports_per_port; - uint32_t n_pipes_per_subport = params->n_pipes_per_subport; - uint32_t n_pipes_per_port = n_pipes_per_subport * n_subports_per_port; - uint32_t n_queues_per_port = RTE_SCHED_QUEUES_PER_PIPE * n_pipes_per_subport * n_subports_per_port; - - uint32_t size_subport = n_subports_per_port * sizeof(struct rte_sched_subport); - uint32_t size_pipe = n_pipes_per_port * sizeof(struct rte_sched_pipe); - uint32_t size_queue = n_queues_per_port * sizeof(struct rte_sched_queue); - uint32_t size_queue_extra = n_queues_per_port * sizeof(struct rte_sched_queue_extra); - uint32_t size_pipe_profiles = RTE_SCHED_PIPE_PROFILES_PER_PORT * sizeof(struct rte_sched_pipe_profile); - uint32_t size_bmp_array = rte_bitmap_get_memory_footprint(n_queues_per_port); - uint32_t size_per_pipe_queue_array, size_queue_array; - - uint32_t base, i; - - size_per_pipe_queue_array = 0; - for (i = 0; i < RTE_SCHED_TRAFFIC_CLASSES_PER_PIPE; i ++) { - size_per_pipe_queue_array += RTE_SCHED_QUEUES_PER_TRAFFIC_CLASS * params->qsize[i] * sizeof(struct rte_mbuf *); - } - size_queue_array = n_pipes_per_port * size_per_pipe_queue_array; - - base = 0; - - if (array == e_RTE_SCHED_PORT_ARRAY_SUBPORT) return base; - base += RTE_CACHE_LINE_ROUNDUP(size_subport); - - if (array == e_RTE_SCHED_PORT_ARRAY_PIPE) return base; - base += RTE_CACHE_LINE_ROUNDUP(size_pipe); - - if (array == e_RTE_SCHED_PORT_ARRAY_QUEUE) return base; - base += RTE_CACHE_LINE_ROUNDUP(size_queue); - - if (array == e_RTE_SCHED_PORT_ARRAY_QUEUE_EXTRA) return base; - base += RTE_CACHE_LINE_ROUNDUP(size_queue_extra); - - if (array == e_RTE_SCHED_PORT_ARRAY_PIPE_PROFILES) return base; - base += RTE_CACHE_LINE_ROUNDUP(size_pipe_profiles); - - if (array == e_RTE_SCHED_PORT_ARRAY_BMP_ARRAY) return base; - base += RTE_CACHE_LINE_ROUNDUP(size_bmp_array); - - if (array == e_RTE_SCHED_PORT_ARRAY_QUEUE_ARRAY) return base; - base += RTE_CACHE_LINE_ROUNDUP(size_queue_array); - - return base; -} - -uint32_t -rte_sched_port_get_memory_footprint(struct rte_sched_port_params *params) -{ - uint32_t size0, size1; - int status; - - status = rte_sched_port_check_params(params); - if (status != 0) { - RTE_LOG(INFO, SCHED, "Port scheduler params check failed (%d)\n", status); - - return 0; - } - - size0 = sizeof(struct rte_sched_port); - size1 = rte_sched_port_get_array_base(params, e_RTE_SCHED_PORT_ARRAY_TOTAL); - - return (size0 + size1); -} - -static void -rte_sched_port_config_qsize(struct rte_sched_port *port) -{ - /* TC 0 */ - port->qsize_add[0] = 0; - port->qsize_add[1] = port->qsize_add[0] + port->qsize[0]; - port->qsize_add[2] = port->qsize_add[1] + port->qsize[0]; - port->qsize_add[3] = port->qsize_add[2] + port->qsize[0]; - - /* TC 1 */ - port->qsize_add[4] = port->qsize_add[3] + port->qsize[0]; - port->qsize_add[5] = port->qsize_add[4] + port->qsize[1]; - port->qsize_add[6] = port->qsize_add[5] + port->qsize[1]; - port->qsize_add[7] = port->qsize_add[6] + port->qsize[1]; - - /* TC 2 */ - port->qsize_add[8] = port->qsize_add[7] + port->qsize[1]; - port->qsize_add[9] = port->qsize_add[8] + port->qsize[2]; - port->qsize_add[10] = port->qsize_add[9] + port->qsize[2]; - port->qsize_add[11] = port->qsize_add[10] + port->qsize[2]; - - /* TC 3 */ - port->qsize_add[12] = port->qsize_add[11] + port->qsize[2]; - port->qsize_add[13] = port->qsize_add[12] + port->qsize[3]; - port->qsize_add[14] = port->qsize_add[13] + port->qsize[3]; - port->qsize_add[15] = port->qsize_add[14] + port->qsize[3]; - - port->qsize_sum = port->qsize_add[15] + port->qsize[3]; -} - -static void -rte_sched_port_log_pipe_profile(struct rte_sched_port *port, uint32_t i) -{ - struct rte_sched_pipe_profile *p = port->pipe_profiles + i; - - RTE_LOG(INFO, SCHED, "Low level config for pipe profile %u:\n" - "\tToken bucket: period = %u, credits per period = %u, size = %u\n" - "\tTraffic classes: period = %u, credits per period = [%u, %u, %u, %u]\n" - "\tTraffic class 3 oversubscription: weight = %hhu\n" - "\tWRR cost: [%hhu, %hhu, %hhu, %hhu], [%hhu, %hhu, %hhu, %hhu], [%hhu, %hhu, %hhu, %hhu], [%hhu, %hhu, %hhu, %hhu]\n", - i, - - /* Token bucket */ - p->tb_period, - p->tb_credits_per_period, - p->tb_size, - - /* Traffic classes */ - p->tc_period, - p->tc_credits_per_period[0], - p->tc_credits_per_period[1], - p->tc_credits_per_period[2], - p->tc_credits_per_period[3], - - /* Traffic class 3 oversubscription */ - p->tc_ov_weight, - - /* WRR */ - p->wrr_cost[ 0], p->wrr_cost[ 1], p->wrr_cost[ 2], p->wrr_cost[ 3], - p->wrr_cost[ 4], p->wrr_cost[ 5], p->wrr_cost[ 6], p->wrr_cost[ 7], - p->wrr_cost[ 8], p->wrr_cost[ 9], p->wrr_cost[10], p->wrr_cost[11], - p->wrr_cost[12], p->wrr_cost[13], p->wrr_cost[14], p->wrr_cost[15]); -} - -static inline uint64_t -rte_sched_time_ms_to_bytes(uint32_t time_ms, uint32_t rate) -{ - uint64_t time = time_ms; - time = (time * rate) / 1000; - - return time; -} - -static void -rte_sched_port_config_pipe_profile_table(struct rte_sched_port *port, struct rte_sched_port_params *params) -{ - uint32_t i, j; - - for (i = 0; i < port->n_pipe_profiles; i ++) { - struct rte_sched_pipe_params *src = params->pipe_profiles + i; - struct rte_sched_pipe_profile *dst = port->pipe_profiles + i; - - /* Token Bucket */ - if (src->tb_rate == params->rate) { - dst->tb_credits_per_period = 1; - dst->tb_period = 1; - } else { - double tb_rate = ((double) src->tb_rate) / ((double) params->rate); - double d = RTE_SCHED_TB_RATE_CONFIG_ERR; - - rte_approx(tb_rate, d, &dst->tb_credits_per_period, &dst->tb_period); - } - dst->tb_size = src->tb_size; - - /* Traffic Classes */ - dst->tc_period = (uint32_t) rte_sched_time_ms_to_bytes(src->tc_period, params->rate); - for (j = 0; j < RTE_SCHED_TRAFFIC_CLASSES_PER_PIPE; j ++) { - dst->tc_credits_per_period[j] = (uint32_t) rte_sched_time_ms_to_bytes(src->tc_period, src->tc_rate[j]); - } -#ifdef RTE_SCHED_SUBPORT_TC_OV - dst->tc_ov_weight = src->tc_ov_weight; -#endif - - /* WRR */ - for (j = 0; j < RTE_SCHED_TRAFFIC_CLASSES_PER_PIPE; j ++) { - uint32_t wrr_cost[RTE_SCHED_QUEUES_PER_TRAFFIC_CLASS]; - uint32_t lcd, lcd1, lcd2; - uint32_t qindex; - - qindex = j * RTE_SCHED_QUEUES_PER_TRAFFIC_CLASS; - - wrr_cost[0] = src->wrr_weights[qindex]; - wrr_cost[1] = src->wrr_weights[qindex + 1]; - wrr_cost[2] = src->wrr_weights[qindex + 2]; - wrr_cost[3] = src->wrr_weights[qindex + 3]; - - lcd1 = rte_get_lcd(wrr_cost[0], wrr_cost[1]); - lcd2 = rte_get_lcd(wrr_cost[2], wrr_cost[3]); - lcd = rte_get_lcd(lcd1, lcd2); - - wrr_cost[0] = lcd / wrr_cost[0]; - wrr_cost[1] = lcd / wrr_cost[1]; - wrr_cost[2] = lcd / wrr_cost[2]; - wrr_cost[3] = lcd / wrr_cost[3]; - - dst->wrr_cost[qindex] = (uint8_t) wrr_cost[0]; - dst->wrr_cost[qindex + 1] = (uint8_t) wrr_cost[1]; - dst->wrr_cost[qindex + 2] = (uint8_t) wrr_cost[2]; - dst->wrr_cost[qindex + 3] = (uint8_t) wrr_cost[3]; - } - - rte_sched_port_log_pipe_profile(port, i); - } - - port->pipe_tc3_rate_max = 0; - for (i = 0; i < port->n_pipe_profiles; i ++) { - struct rte_sched_pipe_params *src = params->pipe_profiles + i; - uint32_t pipe_tc3_rate = src->tc_rate[3]; - - if (port->pipe_tc3_rate_max < pipe_tc3_rate) { - port->pipe_tc3_rate_max = pipe_tc3_rate; - } - } -} - -struct rte_sched_port * -rte_sched_port_config(struct rte_sched_port_params *params) -{ - struct rte_sched_port *port = NULL; - uint32_t mem_size, bmp_mem_size, n_queues_per_port, i; - - /* Check user parameters. Determine the amount of memory to allocate */ - mem_size = rte_sched_port_get_memory_footprint(params); - if (mem_size == 0) { - return NULL; - } - - /* Allocate memory to store the data structures */ - port = rte_zmalloc("qos_params", mem_size, RTE_CACHE_LINE_SIZE); - if (port == NULL) { - return NULL; - } - - /* User parameters */ - port->n_subports_per_port = params->n_subports_per_port; - port->n_pipes_per_subport = params->n_pipes_per_subport; - port->rate = params->rate; - port->mtu = params->mtu + params->frame_overhead; - port->frame_overhead = params->frame_overhead; - memcpy(port->qsize, params->qsize, sizeof(params->qsize)); - port->n_pipe_profiles = params->n_pipe_profiles; - -#ifdef RTE_SCHED_RED - for (i = 0; i < RTE_SCHED_TRAFFIC_CLASSES_PER_PIPE; i++) { - uint32_t j; - - for (j = 0; j < e_RTE_METER_COLORS; j++) { - if (rte_red_config_init(&port->red_config[i][j], - params->red_params[i][j].wq_log2, - params->red_params[i][j].min_th, - params->red_params[i][j].max_th, - params->red_params[i][j].maxp_inv) != 0) { - return NULL; - } - } - } -#endif - - /* Timing */ - port->time_cpu_cycles = rte_get_tsc_cycles(); - port->time_cpu_bytes = 0; - port->time = 0; - port->cycles_per_byte = ((double) rte_get_tsc_hz()) / ((double) params->rate); - - /* Scheduling loop detection */ - port->pipe_loop = RTE_SCHED_PIPE_INVALID; - port->pipe_exhaustion = 0; - - /* Grinders */ - port->busy_grinders = 0; - port->pkts_out = NULL; - port->n_pkts_out = 0; - - /* Queue base calculation */ - rte_sched_port_config_qsize(port); - - /* Large data structures */ - port->subport = (struct rte_sched_subport *) (port->memory + rte_sched_port_get_array_base(params, e_RTE_SCHED_PORT_ARRAY_SUBPORT)); - port->pipe = (struct rte_sched_pipe *) (port->memory + rte_sched_port_get_array_base(params, e_RTE_SCHED_PORT_ARRAY_PIPE)); - port->queue = (struct rte_sched_queue *) (port->memory + rte_sched_port_get_array_base(params, e_RTE_SCHED_PORT_ARRAY_QUEUE)); - port->queue_extra = (struct rte_sched_queue_extra *) (port->memory + rte_sched_port_get_array_base(params, e_RTE_SCHED_PORT_ARRAY_QUEUE_EXTRA)); - port->pipe_profiles = (struct rte_sched_pipe_profile *) (port->memory + rte_sched_port_get_array_base(params, e_RTE_SCHED_PORT_ARRAY_PIPE_PROFILES)); - port->bmp_array = port->memory + rte_sched_port_get_array_base(params, e_RTE_SCHED_PORT_ARRAY_BMP_ARRAY); - port->queue_array = (struct rte_mbuf **) (port->memory + rte_sched_port_get_array_base(params, e_RTE_SCHED_PORT_ARRAY_QUEUE_ARRAY)); - - /* Pipe profile table */ - rte_sched_port_config_pipe_profile_table(port, params); - - /* Bitmap */ - n_queues_per_port = rte_sched_port_queues_per_port(port); - bmp_mem_size = rte_bitmap_get_memory_footprint(n_queues_per_port); - port->bmp = rte_bitmap_init(n_queues_per_port, port->bmp_array, bmp_mem_size); - if (port->bmp == NULL) { - RTE_LOG(INFO, SCHED, "Bitmap init error\n"); - return NULL; - } - for (i = 0; i < RTE_SCHED_PORT_N_GRINDERS; i ++) { - port->grinder_base_bmp_pos[i] = RTE_SCHED_PIPE_INVALID; - } - - return port; -} - -void -rte_sched_port_free(struct rte_sched_port *port) -{ - /* Check user parameters */ - if (port == NULL){ - return; - } - - rte_bitmap_free(port->bmp); - rte_free(port); -} - -static void -rte_sched_port_log_subport_config(struct rte_sched_port *port, uint32_t i) -{ - struct rte_sched_subport *s = port->subport + i; - - RTE_LOG(INFO, SCHED, "Low level config for subport %u:\n" - "\tToken bucket: period = %u, credits per period = %u, size = %u\n" - "\tTraffic classes: period = %u, credits per period = [%u, %u, %u, %u]\n" - "\tTraffic class 3 oversubscription: wm min = %u, wm max = %u\n", - i, - - /* Token bucket */ - s->tb_period, - s->tb_credits_per_period, - s->tb_size, - - /* Traffic classes */ - s->tc_period, - s->tc_credits_per_period[0], - s->tc_credits_per_period[1], - s->tc_credits_per_period[2], - s->tc_credits_per_period[3], - - /* Traffic class 3 oversubscription */ - s->tc_ov_wm_min, - s->tc_ov_wm_max); -} - -int -rte_sched_subport_config(struct rte_sched_port *port, - uint32_t subport_id, - struct rte_sched_subport_params *params) -{ - struct rte_sched_subport *s; - uint32_t i; - - /* Check user parameters */ - if ((port == NULL) || - (subport_id >= port->n_subports_per_port) || - (params == NULL)) { - return -1; - } - - if ((params->tb_rate == 0) || (params->tb_rate > port->rate)) { - return -2; - } - - if (params->tb_size == 0) { - return -3; - } - - for (i = 0; i < RTE_SCHED_TRAFFIC_CLASSES_PER_PIPE; i ++) { - if ((params->tc_rate[i] == 0) || (params->tc_rate[i] > params->tb_rate)) { - return -4; - } - } - - if (params->tc_period == 0) { - return -5; - } - - s = port->subport + subport_id; - - /* Token Bucket (TB) */ - if (params->tb_rate == port->rate) { - s->tb_credits_per_period = 1; - s->tb_period = 1; - } else { - double tb_rate = ((double) params->tb_rate) / ((double) port->rate); - double d = RTE_SCHED_TB_RATE_CONFIG_ERR; - - rte_approx(tb_rate, d, &s->tb_credits_per_period, &s->tb_period); - } - s->tb_size = params->tb_size; - s->tb_time = port->time; - s->tb_credits = s->tb_size / 2; - - /* Traffic Classes (TCs) */ - s->tc_period = (uint32_t) rte_sched_time_ms_to_bytes(params->tc_period, port->rate); - for (i = 0; i < RTE_SCHED_TRAFFIC_CLASSES_PER_PIPE; i ++) { - s->tc_credits_per_period[i] = (uint32_t) rte_sched_time_ms_to_bytes(params->tc_period, params->tc_rate[i]); - } - s->tc_time = port->time + s->tc_period; - for (i = 0; i < RTE_SCHED_TRAFFIC_CLASSES_PER_PIPE; i ++) { - s->tc_credits[i] = s->tc_credits_per_period[i]; - } - -#ifdef RTE_SCHED_SUBPORT_TC_OV - /* TC oversubscription */ - s->tc_ov_wm_min = port->mtu; - s->tc_ov_wm_max = (uint32_t) rte_sched_time_ms_to_bytes(params->tc_period, port->pipe_tc3_rate_max); - s->tc_ov_wm = s->tc_ov_wm_max; - s->tc_ov_period_id = 0; - s->tc_ov = 0; - s->tc_ov_n = 0; - s->tc_ov_rate = 0; -#endif - - rte_sched_port_log_subport_config(port, subport_id); - - return 0; -} - -int -rte_sched_pipe_config(struct rte_sched_port *port, - uint32_t subport_id, - uint32_t pipe_id, - int32_t pipe_profile) -{ - struct rte_sched_subport *s; - struct rte_sched_pipe *p; - struct rte_sched_pipe_profile *params; - uint32_t deactivate, profile, i; - - /* Check user parameters */ - profile = (uint32_t) pipe_profile; - deactivate = (pipe_profile < 0); - if ((port == NULL) || - (subport_id >= port->n_subports_per_port) || - (pipe_id >= port->n_pipes_per_subport) || - ((!deactivate) && (profile >= port->n_pipe_profiles))) { - return -1; - } - - /* Check that subport configuration is valid */ - s = port->subport + subport_id; - if (s->tb_period == 0) { - return -2; - } - - p = port->pipe + (subport_id * port->n_pipes_per_subport + pipe_id); - - /* Handle the case when pipe already has a valid configuration */ - if (p->tb_time) { - params = port->pipe_profiles + p->profile; - -#ifdef RTE_SCHED_SUBPORT_TC_OV - double subport_tc3_rate = ((double) s->tc_credits_per_period[3]) / ((double) s->tc_period); - double pipe_tc3_rate = ((double) params->tc_credits_per_period[3]) / ((double) params->tc_period); - uint32_t tc3_ov = s->tc_ov; - - /* Unplug pipe from its subport */ - s->tc_ov_n -= params->tc_ov_weight; - s->tc_ov_rate -= pipe_tc3_rate; - s->tc_ov = s->tc_ov_rate > subport_tc3_rate; - - if (s->tc_ov != tc3_ov) { - RTE_LOG(INFO, SCHED, "Subport %u TC3 oversubscription is OFF (%.4lf >= %.4lf)\n", - subport_id, subport_tc3_rate, s->tc_ov_rate); - } -#endif - - /* Reset the pipe */ - memset(p, 0, sizeof(struct rte_sched_pipe)); - } - - if (deactivate) { - return 0; - } - - /* Apply the new pipe configuration */ - p->profile = profile; - params = port->pipe_profiles + p->profile; - - /* Token Bucket (TB) */ - p->tb_time = port->time; - p->tb_credits = params->tb_size / 2; - - /* Traffic Classes (TCs) */ - p->tc_time = port->time + params->tc_period; - for (i = 0; i < RTE_SCHED_TRAFFIC_CLASSES_PER_PIPE; i ++) { - p->tc_credits[i] = params->tc_credits_per_period[i]; - } - -#ifdef RTE_SCHED_SUBPORT_TC_OV - { - /* Subport TC3 oversubscription */ - double subport_tc3_rate = ((double) s->tc_credits_per_period[3]) / ((double) s->tc_period); - double pipe_tc3_rate = ((double) params->tc_credits_per_period[3]) / ((double) params->tc_period); - uint32_t tc3_ov = s->tc_ov; - - s->tc_ov_n += params->tc_ov_weight; - s->tc_ov_rate += pipe_tc3_rate; - s->tc_ov = s->tc_ov_rate > subport_tc3_rate; - - if (s->tc_ov != tc3_ov) { - RTE_LOG(INFO, SCHED, "Subport %u TC3 oversubscription is ON (%.4lf < %.4lf)\n", - subport_id, subport_tc3_rate, s->tc_ov_rate); - } - p->tc_ov_period_id = s->tc_ov_period_id; - p->tc_ov_credits = s->tc_ov_wm; - } -#endif - - return 0; -} - -int -rte_sched_subport_read_stats(struct rte_sched_port *port, - uint32_t subport_id, - struct rte_sched_subport_stats *stats, - uint32_t *tc_ov) -{ - struct rte_sched_subport *s; - - /* Check user parameters */ - if ((port == NULL) || - (subport_id >= port->n_subports_per_port) || - (stats == NULL) || - (tc_ov == NULL)) { - return -1; - } - s = port->subport + subport_id; - - /* Copy subport stats and clear */ - memcpy(stats, &s->stats, sizeof(struct rte_sched_subport_stats)); - memset(&s->stats, 0, sizeof(struct rte_sched_subport_stats)); - - /* Subport TC ovesubscription status */ - *tc_ov = s->tc_ov; - - return 0; -} - -int -rte_sched_queue_read_stats(struct rte_sched_port *port, - uint32_t queue_id, - struct rte_sched_queue_stats *stats, - uint16_t *qlen) -{ - struct rte_sched_queue *q; - struct rte_sched_queue_extra *qe; - - /* Check user parameters */ - if ((port == NULL) || - (queue_id >= rte_sched_port_queues_per_port(port)) || - (stats == NULL) || - (qlen == NULL)) { - return -1; - } - q = port->queue + queue_id; - qe = port->queue_extra + queue_id; - - /* Copy queue stats and clear */ - memcpy(stats, &qe->stats, sizeof(struct rte_sched_queue_stats)); - memset(&qe->stats, 0, sizeof(struct rte_sched_queue_stats)); - - /* Queue length */ - *qlen = q->qw - q->qr; - - return 0; -} - -static inline uint32_t -rte_sched_port_qindex(struct rte_sched_port *port, uint32_t subport, uint32_t pipe, uint32_t traffic_class, uint32_t queue) -{ - uint32_t result; - - result = subport * port->n_pipes_per_subport + pipe; - result = result * RTE_SCHED_TRAFFIC_CLASSES_PER_PIPE + traffic_class; - result = result * RTE_SCHED_QUEUES_PER_TRAFFIC_CLASS + queue; - - return result; -} - -static inline struct rte_mbuf ** -rte_sched_port_qbase(struct rte_sched_port *port, uint32_t qindex) -{ - uint32_t pindex = qindex >> 4; - uint32_t qpos = qindex & 0xF; - - return (port->queue_array + pindex * port->qsize_sum + port->qsize_add[qpos]); -} - -static inline uint16_t -rte_sched_port_qsize(struct rte_sched_port *port, uint32_t qindex) -{ - uint32_t tc = (qindex >> 2) & 0x3; - - return port->qsize[tc]; -} - -#if RTE_SCHED_DEBUG - -static inline int -rte_sched_port_queue_is_empty(struct rte_sched_port *port, uint32_t qindex) -{ - struct rte_sched_queue *queue = port->queue + qindex; - - return (queue->qr == queue->qw); -} - -static inline int -rte_sched_port_queue_is_full(struct rte_sched_port *port, uint32_t qindex) -{ - struct rte_sched_queue *queue = port->queue + qindex; - uint16_t qsize = rte_sched_port_qsize(port, qindex); - uint16_t qlen = queue->qw - queue->qr; - - return (qlen >= qsize); -} - -#endif /* RTE_SCHED_DEBUG */ - -#ifdef RTE_SCHED_COLLECT_STATS - -static inline void -rte_sched_port_update_subport_stats(struct rte_sched_port *port, uint32_t qindex, struct rte_mbuf *pkt) -{ - struct rte_sched_subport *s = port->subport + (qindex / rte_sched_port_queues_per_subport(port)); - uint32_t tc_index = (qindex >> 2) & 0x3; - uint32_t pkt_len = pkt->pkt_len; - - s->stats.n_pkts_tc[tc_index] += 1; - s->stats.n_bytes_tc[tc_index] += pkt_len; -} - -static inline void -rte_sched_port_update_subport_stats_on_drop(struct rte_sched_port *port, uint32_t qindex, struct rte_mbuf *pkt) -{ - struct rte_sched_subport *s = port->subport + (qindex / rte_sched_port_queues_per_subport(port)); - uint32_t tc_index = (qindex >> 2) & 0x3; - uint32_t pkt_len = pkt->pkt_len; - - s->stats.n_pkts_tc_dropped[tc_index] += 1; - s->stats.n_bytes_tc_dropped[tc_index] += pkt_len; -} - -static inline void -rte_sched_port_update_queue_stats(struct rte_sched_port *port, uint32_t qindex, struct rte_mbuf *pkt) -{ - struct rte_sched_queue_extra *qe = port->queue_extra + qindex; - uint32_t pkt_len = pkt->pkt_len; - - qe->stats.n_pkts += 1; - qe->stats.n_bytes += pkt_len; -} - -static inline void -rte_sched_port_update_queue_stats_on_drop(struct rte_sched_port *port, uint32_t qindex, struct rte_mbuf *pkt) -{ - struct rte_sched_queue_extra *qe = port->queue_extra + qindex; - uint32_t pkt_len = pkt->pkt_len; - - qe->stats.n_pkts_dropped += 1; - qe->stats.n_bytes_dropped += pkt_len; -} - -#endif /* RTE_SCHED_COLLECT_STATS */ - -#ifdef RTE_SCHED_RED - -static inline int -rte_sched_port_red_drop(struct rte_sched_port *port, struct rte_mbuf *pkt, uint32_t qindex, uint16_t qlen) -{ - struct rte_sched_queue_extra *qe; - struct rte_red_config *red_cfg; - struct rte_red *red; - uint32_t tc_index; - enum rte_meter_color color; - - tc_index = (qindex >> 2) & 0x3; - color = rte_sched_port_pkt_read_color(pkt); - red_cfg = &port->red_config[tc_index][color]; - - qe = port->queue_extra + qindex; - red = &qe->red; - - return rte_red_enqueue(red_cfg, red, qlen, port->time); -} - -static inline void -rte_sched_port_set_queue_empty_timestamp(struct rte_sched_port *port, uint32_t qindex) -{ - struct rte_sched_queue_extra *qe; - struct rte_red *red; - - qe = port->queue_extra + qindex; - red = &qe->red; - - rte_red_mark_queue_empty(red, port->time); -} - -#else - -#define rte_sched_port_red_drop(port, pkt, qindex, qlen) 0 - -#define rte_sched_port_set_queue_empty_timestamp(port, qindex) - -#endif /* RTE_SCHED_RED */ - -#if RTE_SCHED_DEBUG - -static inline int -debug_pipe_is_empty(struct rte_sched_port *port, uint32_t pindex) -{ - uint32_t qindex, i; - - qindex = pindex << 4; - - for (i = 0; i < 16; i ++){ - uint32_t queue_empty = rte_sched_port_queue_is_empty(port, qindex + i); - uint32_t bmp_bit_clear = (rte_bitmap_get(port->bmp, qindex + i) == 0); - - if (queue_empty != bmp_bit_clear){ - rte_panic("Queue status mismatch for queue %u of pipe %u\n", i, pindex); - } - - if (!queue_empty){ - return 0; - } - } - - return 1; -} - -static inline void -debug_check_queue_slab(struct rte_sched_port *port, uint32_t bmp_pos, uint64_t bmp_slab) -{ - uint64_t mask; - uint32_t i, panic; - - if (bmp_slab == 0){ - rte_panic("Empty slab at position %u\n", bmp_pos); - } - - panic = 0; - for (i = 0, mask = 1; i < 64; i ++, mask <<= 1) { - if (mask & bmp_slab){ - if (rte_sched_port_queue_is_empty(port, bmp_pos + i)) { - printf("Queue %u (slab offset %u) is empty\n", bmp_pos + i, i); - panic = 1; - } - } - } - - if (panic){ - rte_panic("Empty queues in slab 0x%" PRIx64 "starting at position %u\n", - bmp_slab, bmp_pos); - } -} - -#endif /* RTE_SCHED_DEBUG */ - -static inline uint32_t -rte_sched_port_enqueue_qptrs_prefetch0(struct rte_sched_port *port, struct rte_mbuf *pkt) -{ - struct rte_sched_queue *q; -#ifdef RTE_SCHED_COLLECT_STATS - struct rte_sched_queue_extra *qe; -#endif - uint32_t subport, pipe, traffic_class, queue, qindex; - - rte_sched_port_pkt_read_tree_path(pkt, &subport, &pipe, &traffic_class, &queue); - - qindex = rte_sched_port_qindex(port, subport, pipe, traffic_class, queue); - q = port->queue + qindex; - rte_prefetch0(q); -#ifdef RTE_SCHED_COLLECT_STATS - qe = port->queue_extra + qindex; - rte_prefetch0(qe); -#endif - - return qindex; -} - -static inline void -rte_sched_port_enqueue_qwa_prefetch0(struct rte_sched_port *port, uint32_t qindex, struct rte_mbuf **qbase) -{ - struct rte_sched_queue *q; - struct rte_mbuf **q_qw; - uint16_t qsize; - - q = port->queue + qindex; - qsize = rte_sched_port_qsize(port, qindex); - q_qw = qbase + (q->qw & (qsize - 1)); - - rte_prefetch0(q_qw); - rte_bitmap_prefetch0(port->bmp, qindex); -} - -static inline int -rte_sched_port_enqueue_qwa(struct rte_sched_port *port, uint32_t qindex, struct rte_mbuf **qbase, struct rte_mbuf *pkt) -{ - struct rte_sched_queue *q; - uint16_t qsize; - uint16_t qlen; - - q = port->queue + qindex; - qsize = rte_sched_port_qsize(port, qindex); - qlen = q->qw - q->qr; - - /* Drop the packet (and update drop stats) when queue is full */ - if (unlikely(rte_sched_port_red_drop(port, pkt, qindex, qlen) || (qlen >= qsize))) { - rte_pktmbuf_free(pkt); -#ifdef RTE_SCHED_COLLECT_STATS - rte_sched_port_update_subport_stats_on_drop(port, qindex, pkt); - rte_sched_port_update_queue_stats_on_drop(port, qindex, pkt); -#endif - return 0; - } - - /* Enqueue packet */ - qbase[q->qw & (qsize - 1)] = pkt; - q->qw ++; - - /* Activate queue in the port bitmap */ - rte_bitmap_set(port->bmp, qindex); - - /* Statistics */ -#ifdef RTE_SCHED_COLLECT_STATS - rte_sched_port_update_subport_stats(port, qindex, pkt); - rte_sched_port_update_queue_stats(port, qindex, pkt); -#endif - - return 1; -} - -#if RTE_SCHED_ENQUEUE == 0 - -int -rte_sched_port_enqueue(struct rte_sched_port *port, struct rte_mbuf **pkts, uint32_t n_pkts) -{ - uint32_t result, i; - - result = 0; - - for (i = 0; i < n_pkts; i ++) { - struct rte_mbuf *pkt; - struct rte_mbuf **q_base; - uint32_t subport, pipe, traffic_class, queue, qindex; - - pkt = pkts[i]; - - rte_sched_port_pkt_read_tree_path(pkt, &subport, &pipe, &traffic_class, &queue); - - qindex = rte_sched_port_qindex(port, subport, pipe, traffic_class, queue); - - q_base = rte_sched_port_qbase(port, qindex); - - result += rte_sched_port_enqueue_qwa(port, qindex, q_base, pkt); - } - - return result; -} - -#else - -/* The enqueue function implements a 4-level pipeline with each stage processing - * two different packets. The purpose of using a pipeline is to hide the latency - * of prefetching the data structures. The naming convention is presented in the - * diagram below: - * - * p00 _______ p10 _______ p20 _______ p30 _______ - * ----->| |----->| |----->| |----->| |-----> - * | 0 | | 1 | | 2 | | 3 | - * ----->|_______|----->|_______|----->|_______|----->|_______|-----> - * p01 p11 p21 p31 - * - ***/ -int -rte_sched_port_enqueue(struct rte_sched_port *port, struct rte_mbuf **pkts, uint32_t n_pkts) -{ - struct rte_mbuf *pkt00, *pkt01, *pkt10, *pkt11, *pkt20, *pkt21, *pkt30, *pkt31, *pkt_last; - struct rte_mbuf **q00_base, **q01_base, **q10_base, **q11_base, **q20_base, **q21_base, **q30_base, **q31_base, **q_last_base; - uint32_t q00, q01, q10, q11, q20, q21, q30, q31, q_last; - uint32_t r00, r01, r10, r11, r20, r21, r30, r31, r_last; - uint32_t result, i; - - result = 0; - - /* Less then 6 input packets available, which is not enough to feed the pipeline */ - if (unlikely(n_pkts < 6)) { - struct rte_mbuf **q_base[5]; - uint32_t q[5]; - - /* Prefetch the mbuf structure of each packet */ - for (i = 0; i < n_pkts; i ++) { - rte_prefetch0(pkts[i]); - } - - /* Prefetch the queue structure for each queue */ - for (i = 0; i < n_pkts; i ++) { - q[i] = rte_sched_port_enqueue_qptrs_prefetch0(port, pkts[i]); - } - - /* Prefetch the write pointer location of each queue */ - for (i = 0; i < n_pkts; i ++) { - q_base[i] = rte_sched_port_qbase(port, q[i]); - rte_sched_port_enqueue_qwa_prefetch0(port, q[i], q_base[i]); - } - - /* Write each packet to its queue */ - for (i = 0; i < n_pkts; i ++) { - result += rte_sched_port_enqueue_qwa(port, q[i], q_base[i], pkts[i]); - } - - return result; - } - - /* Feed the first 3 stages of the pipeline (6 packets needed) */ - pkt20 = pkts[0]; - pkt21 = pkts[1]; - rte_prefetch0(pkt20); - rte_prefetch0(pkt21); - - pkt10 = pkts[2]; - pkt11 = pkts[3]; - rte_prefetch0(pkt10); - rte_prefetch0(pkt11); - - q20 = rte_sched_port_enqueue_qptrs_prefetch0(port, pkt20); - q21 = rte_sched_port_enqueue_qptrs_prefetch0(port, pkt21); - - pkt00 = pkts[4]; - pkt01 = pkts[5]; - rte_prefetch0(pkt00); - rte_prefetch0(pkt01); - - q10 = rte_sched_port_enqueue_qptrs_prefetch0(port, pkt10); - q11 = rte_sched_port_enqueue_qptrs_prefetch0(port, pkt11); - - q20_base = rte_sched_port_qbase(port, q20); - q21_base = rte_sched_port_qbase(port, q21); - rte_sched_port_enqueue_qwa_prefetch0(port, q20, q20_base); - rte_sched_port_enqueue_qwa_prefetch0(port, q21, q21_base); - - /* Run the pipeline */ - for (i = 6; i < (n_pkts & (~1)); i += 2) { - /* Propagate stage inputs */ - pkt30 = pkt20; - pkt31 = pkt21; - pkt20 = pkt10; - pkt21 = pkt11; - pkt10 = pkt00; - pkt11 = pkt01; - q30 = q20; - q31 = q21; - q20 = q10; - q21 = q11; - q30_base = q20_base; - q31_base = q21_base; - - /* Stage 0: Get packets in */ - pkt00 = pkts[i]; - pkt01 = pkts[i + 1]; - rte_prefetch0(pkt00); - rte_prefetch0(pkt01); - - /* Stage 1: Prefetch queue structure storing queue pointers */ - q10 = rte_sched_port_enqueue_qptrs_prefetch0(port, pkt10); - q11 = rte_sched_port_enqueue_qptrs_prefetch0(port, pkt11); - - /* Stage 2: Prefetch queue write location */ - q20_base = rte_sched_port_qbase(port, q20); - q21_base = rte_sched_port_qbase(port, q21); - rte_sched_port_enqueue_qwa_prefetch0(port, q20, q20_base); - rte_sched_port_enqueue_qwa_prefetch0(port, q21, q21_base); - - /* Stage 3: Write packet to queue and activate queue */ - r30 = rte_sched_port_enqueue_qwa(port, q30, q30_base, pkt30); - r31 = rte_sched_port_enqueue_qwa(port, q31, q31_base, pkt31); - result += r30 + r31; - } - - /* Drain the pipeline (exactly 6 packets). Handle the last packet in the case - of an odd number of input packets. */ - pkt_last = pkts[n_pkts - 1]; - rte_prefetch0(pkt_last); - - q00 = rte_sched_port_enqueue_qptrs_prefetch0(port, pkt00); - q01 = rte_sched_port_enqueue_qptrs_prefetch0(port, pkt01); - - q10_base = rte_sched_port_qbase(port, q10); - q11_base = rte_sched_port_qbase(port, q11); - rte_sched_port_enqueue_qwa_prefetch0(port, q10, q10_base); - rte_sched_port_enqueue_qwa_prefetch0(port, q11, q11_base); - - r20 = rte_sched_port_enqueue_qwa(port, q20, q20_base, pkt20); - r21 = rte_sched_port_enqueue_qwa(port, q21, q21_base, pkt21); - result += r20 + r21; - - q_last = rte_sched_port_enqueue_qptrs_prefetch0(port, pkt_last); - - q00_base = rte_sched_port_qbase(port, q00); - q01_base = rte_sched_port_qbase(port, q01); - rte_sched_port_enqueue_qwa_prefetch0(port, q00, q00_base); - rte_sched_port_enqueue_qwa_prefetch0(port, q01, q01_base); - - r10 = rte_sched_port_enqueue_qwa(port, q10, q10_base, pkt10); - r11 = rte_sched_port_enqueue_qwa(port, q11, q11_base, pkt11); - result += r10 + r11; - - q_last_base = rte_sched_port_qbase(port, q_last); - rte_sched_port_enqueue_qwa_prefetch0(port, q_last, q_last_base); - - r00 = rte_sched_port_enqueue_qwa(port, q00, q00_base, pkt00); - r01 = rte_sched_port_enqueue_qwa(port, q01, q01_base, pkt01); - result += r00 + r01; - - if (n_pkts & 1) { - r_last = rte_sched_port_enqueue_qwa(port, q_last, q_last_base, pkt_last); - result += r_last; - } - - return result; -} - -#endif /* RTE_SCHED_ENQUEUE */ - -#if RTE_SCHED_TS_CREDITS_UPDATE == 0 - -#define grinder_credits_update(port, pos) - -#elif !defined(RTE_SCHED_SUBPORT_TC_OV) - -static inline void -grinder_credits_update(struct rte_sched_port *port, uint32_t pos) -{ - struct rte_sched_grinder *grinder = port->grinder + pos; - struct rte_sched_subport *subport = grinder->subport; - struct rte_sched_pipe *pipe = grinder->pipe; - struct rte_sched_pipe_profile *params = grinder->pipe_params; - uint64_t n_periods; - - /* Subport TB */ - n_periods = (port->time - subport->tb_time) / subport->tb_period; - subport->tb_credits += n_periods * subport->tb_credits_per_period; - subport->tb_credits = rte_sched_min_val_2_u32(subport->tb_credits, subport->tb_size); - subport->tb_time += n_periods * subport->tb_period; - - /* Pipe TB */ - n_periods = (port->time - pipe->tb_time) / params->tb_period; - pipe->tb_credits += n_periods * params->tb_credits_per_period; - pipe->tb_credits = rte_sched_min_val_2_u32(pipe->tb_credits, params->tb_size); - pipe->tb_time += n_periods * params->tb_period; - - /* Subport TCs */ - if (unlikely(port->time >= subport->tc_time)) { - subport->tc_credits[0] = subport->tc_credits_per_period[0]; - subport->tc_credits[1] = subport->tc_credits_per_period[1]; - subport->tc_credits[2] = subport->tc_credits_per_period[2]; - subport->tc_credits[3] = subport->tc_credits_per_period[3]; - subport->tc_time = port->time + subport->tc_period; - } - - /* Pipe TCs */ - if (unlikely(port->time >= pipe->tc_time)) { - pipe->tc_credits[0] = params->tc_credits_per_period[0]; - pipe->tc_credits[1] = params->tc_credits_per_period[1]; - pipe->tc_credits[2] = params->tc_credits_per_period[2]; - pipe->tc_credits[3] = params->tc_credits_per_period[3]; - pipe->tc_time = port->time + params->tc_period; - } -} - -#else - -static inline uint32_t -grinder_tc_ov_credits_update(struct rte_sched_port *port, uint32_t pos) -{ - struct rte_sched_grinder *grinder = port->grinder + pos; - struct rte_sched_subport *subport = grinder->subport; - uint32_t tc_ov_consumption[RTE_SCHED_TRAFFIC_CLASSES_PER_PIPE]; - uint32_t tc_ov_consumption_max; - uint32_t tc_ov_wm = subport->tc_ov_wm; - - if (subport->tc_ov == 0) { - return subport->tc_ov_wm_max; - } - - tc_ov_consumption[0] = subport->tc_credits_per_period[0] - subport->tc_credits[0]; - tc_ov_consumption[1] = subport->tc_credits_per_period[1] - subport->tc_credits[1]; - tc_ov_consumption[2] = subport->tc_credits_per_period[2] - subport->tc_credits[2]; - tc_ov_consumption[3] = subport->tc_credits_per_period[3] - subport->tc_credits[3]; - - tc_ov_consumption_max = subport->tc_credits_per_period[3] - - (tc_ov_consumption[0] + tc_ov_consumption[1] + tc_ov_consumption[2]); - - if (tc_ov_consumption[3] > (tc_ov_consumption_max - port->mtu)) { - tc_ov_wm -= tc_ov_wm >> 7; - if (tc_ov_wm < subport->tc_ov_wm_min) { - tc_ov_wm = subport->tc_ov_wm_min; - } - return tc_ov_wm; - } - - tc_ov_wm += (tc_ov_wm >> 7) + 1; - if (tc_ov_wm > subport->tc_ov_wm_max) { - tc_ov_wm = subport->tc_ov_wm_max; - } - return tc_ov_wm; -} - -static inline void -grinder_credits_update(struct rte_sched_port *port, uint32_t pos) -{ - struct rte_sched_grinder *grinder = port->grinder + pos; - struct rte_sched_subport *subport = grinder->subport; - struct rte_sched_pipe *pipe = grinder->pipe; - struct rte_sched_pipe_profile *params = grinder->pipe_params; - uint64_t n_periods; - - /* Subport TB */ - n_periods = (port->time - subport->tb_time) / subport->tb_period; - subport->tb_credits += n_periods * subport->tb_credits_per_period; - subport->tb_credits = rte_sched_min_val_2_u32(subport->tb_credits, subport->tb_size); - subport->tb_time += n_periods * subport->tb_period; - - /* Pipe TB */ - n_periods = (port->time - pipe->tb_time) / params->tb_period; - pipe->tb_credits += n_periods * params->tb_credits_per_period; - pipe->tb_credits = rte_sched_min_val_2_u32(pipe->tb_credits, params->tb_size); - pipe->tb_time += n_periods * params->tb_period; - - /* Subport TCs */ - if (unlikely(port->time >= subport->tc_time)) { - subport->tc_ov_wm = grinder_tc_ov_credits_update(port, pos); - - subport->tc_credits[0] = subport->tc_credits_per_period[0]; - subport->tc_credits[1] = subport->tc_credits_per_period[1]; - subport->tc_credits[2] = subport->tc_credits_per_period[2]; - subport->tc_credits[3] = subport->tc_credits_per_period[3]; - - subport->tc_time = port->time + subport->tc_period; - subport->tc_ov_period_id ++; - } - - /* Pipe TCs */ - if (unlikely(port->time >= pipe->tc_time)) { - pipe->tc_credits[0] = params->tc_credits_per_period[0]; - pipe->tc_credits[1] = params->tc_credits_per_period[1]; - pipe->tc_credits[2] = params->tc_credits_per_period[2]; - pipe->tc_credits[3] = params->tc_credits_per_period[3]; - pipe->tc_time = port->time + params->tc_period; - } - - /* Pipe TCs - Oversubscription */ - if (unlikely(pipe->tc_ov_period_id != subport->tc_ov_period_id)) { - pipe->tc_ov_credits = subport->tc_ov_wm * params->tc_ov_weight; - - pipe->tc_ov_period_id = subport->tc_ov_period_id; - } -} - -#endif /* RTE_SCHED_TS_CREDITS_UPDATE, RTE_SCHED_SUBPORT_TC_OV */ - -#if RTE_SCHED_TS_CREDITS_CHECK - -#ifndef RTE_SCHED_SUBPORT_TC_OV - -static inline int -grinder_credits_check(struct rte_sched_port *port, uint32_t pos) -{ - struct rte_sched_grinder *grinder = port->grinder + pos; - struct rte_sched_subport *subport = grinder->subport; - struct rte_sched_pipe *pipe = grinder->pipe; - struct rte_mbuf *pkt = grinder->pkt; - uint32_t tc_index = grinder->tc_index; - uint32_t pkt_len = pkt->pkt_len + port->frame_overhead; - uint32_t subport_tb_credits = subport->tb_credits; - uint32_t subport_tc_credits = subport->tc_credits[tc_index]; - uint32_t pipe_tb_credits = pipe->tb_credits; - uint32_t pipe_tc_credits = pipe->tc_credits[tc_index]; - int enough_credits; - - /* Check queue credits */ - enough_credits = (pkt_len <= subport_tb_credits) && - (pkt_len <= subport_tc_credits) && - (pkt_len <= pipe_tb_credits) && - (pkt_len <= pipe_tc_credits); - - if (!enough_credits) { - return 0; - } - - /* Update port credits */ - subport->tb_credits -= pkt_len; - subport->tc_credits[tc_index] -= pkt_len; - pipe->tb_credits -= pkt_len; - pipe->tc_credits[tc_index] -= pkt_len; - - return 1; -} - -#else - -static inline int -grinder_credits_check(struct rte_sched_port *port, uint32_t pos) -{ - struct rte_sched_grinder *grinder = port->grinder + pos; - struct rte_sched_subport *subport = grinder->subport; - struct rte_sched_pipe *pipe = grinder->pipe; - struct rte_mbuf *pkt = grinder->pkt; - uint32_t tc_index = grinder->tc_index; - uint32_t pkt_len = pkt->pkt_len + port->frame_overhead; - uint32_t subport_tb_credits = subport->tb_credits; - uint32_t subport_tc_credits = subport->tc_credits[tc_index]; - uint32_t pipe_tb_credits = pipe->tb_credits; - uint32_t pipe_tc_credits = pipe->tc_credits[tc_index]; - uint32_t pipe_tc_ov_mask1[] = {UINT32_MAX, UINT32_MAX, UINT32_MAX, pipe->tc_ov_credits}; - uint32_t pipe_tc_ov_mask2[] = {0, 0, 0, UINT32_MAX}; - uint32_t pipe_tc_ov_credits = pipe_tc_ov_mask1[tc_index]; - int enough_credits; - - /* Check pipe and subport credits */ - enough_credits = (pkt_len <= subport_tb_credits) && - (pkt_len <= subport_tc_credits) && - (pkt_len <= pipe_tb_credits) && - (pkt_len <= pipe_tc_credits) && - (pkt_len <= pipe_tc_ov_credits); - - if (!enough_credits) { - return 0; - } - - /* Update pipe and subport credits */ - subport->tb_credits -= pkt_len; - subport->tc_credits[tc_index] -= pkt_len; - pipe->tb_credits -= pkt_len; - pipe->tc_credits[tc_index] -= pkt_len; - pipe->tc_ov_credits -= pipe_tc_ov_mask2[tc_index] & pkt_len; - - return 1; -} - -#endif /* RTE_SCHED_SUBPORT_TC_OV */ - -#endif /* RTE_SCHED_TS_CREDITS_CHECK */ - -static inline int -grinder_schedule(struct rte_sched_port *port, uint32_t pos) -{ - struct rte_sched_grinder *grinder = port->grinder + pos; - struct rte_sched_queue *queue = grinder->queue[grinder->qpos]; - struct rte_mbuf *pkt = grinder->pkt; - uint32_t pkt_len = pkt->pkt_len + port->frame_overhead; - -#if RTE_SCHED_TS_CREDITS_CHECK - if (!grinder_credits_check(port, pos)) { - return 0; - } -#endif - - /* Advance port time */ - port->time += pkt_len; - - /* Send packet */ - port->pkts_out[port->n_pkts_out ++] = pkt; - queue->qr ++; - grinder->wrr_tokens[grinder->qpos] += pkt_len * grinder->wrr_cost[grinder->qpos]; - if (queue->qr == queue->qw) { - uint32_t qindex = grinder->qindex[grinder->qpos]; - - rte_bitmap_clear(port->bmp, qindex); - grinder->qmask &= ~(1 << grinder->qpos); - grinder->wrr_mask[grinder->qpos] = 0; - rte_sched_port_set_queue_empty_timestamp(port, qindex); - } - - /* Reset pipe loop detection */ - port->pipe_loop = RTE_SCHED_PIPE_INVALID; - grinder->productive = 1; - - return 1; -} - -#if RTE_SCHED_OPTIMIZATIONS - -static inline int -grinder_pipe_exists(struct rte_sched_port *port, uint32_t base_pipe) -{ - __m128i index = _mm_set1_epi32 (base_pipe); - __m128i pipes = _mm_load_si128((__m128i *)port->grinder_base_bmp_pos); - __m128i res = _mm_cmpeq_epi32(pipes, index); - pipes = _mm_load_si128((__m128i *)(port->grinder_base_bmp_pos + 4)); - pipes = _mm_cmpeq_epi32(pipes, index); - res = _mm_or_si128(res, pipes); - - if (_mm_testz_si128(res, res)) - return 0; - - return 1; -} - -#else - -static inline int -grinder_pipe_exists(struct rte_sched_port *port, uint32_t base_pipe) -{ - uint32_t i; - - for (i = 0; i < RTE_SCHED_PORT_N_GRINDERS; i ++) { - if (port->grinder_base_bmp_pos[i] == base_pipe) { - return 1; - } - } - - return 0; -} - -#endif /* RTE_SCHED_OPTIMIZATIONS */ - -static inline void -grinder_pcache_populate(struct rte_sched_port *port, uint32_t pos, uint32_t bmp_pos, uint64_t bmp_slab) -{ - struct rte_sched_grinder *grinder = port->grinder + pos; - uint16_t w[4]; - - grinder->pcache_w = 0; - grinder->pcache_r = 0; - - w[0] = (uint16_t) bmp_slab; - w[1] = (uint16_t) (bmp_slab >> 16); - w[2] = (uint16_t) (bmp_slab >> 32); - w[3] = (uint16_t) (bmp_slab >> 48); - - grinder->pcache_qmask[grinder->pcache_w] = w[0]; - grinder->pcache_qindex[grinder->pcache_w] = bmp_pos; - grinder->pcache_w += (w[0] != 0); - - grinder->pcache_qmask[grinder->pcache_w] = w[1]; - grinder->pcache_qindex[grinder->pcache_w] = bmp_pos + 16; - grinder->pcache_w += (w[1] != 0); - - grinder->pcache_qmask[grinder->pcache_w] = w[2]; - grinder->pcache_qindex[grinder->pcache_w] = bmp_pos + 32; - grinder->pcache_w += (w[2] != 0); - - grinder->pcache_qmask[grinder->pcache_w] = w[3]; - grinder->pcache_qindex[grinder->pcache_w] = bmp_pos + 48; - grinder->pcache_w += (w[3] != 0); -} - -static inline void -grinder_tccache_populate(struct rte_sched_port *port, uint32_t pos, uint32_t qindex, uint16_t qmask) -{ - struct rte_sched_grinder *grinder = port->grinder + pos; - uint8_t b[4]; - - grinder->tccache_w = 0; - grinder->tccache_r = 0; - - b[0] = (uint8_t) (qmask & 0xF); - b[1] = (uint8_t) ((qmask >> 4) & 0xF); - b[2] = (uint8_t) ((qmask >> 8) & 0xF); - b[3] = (uint8_t) ((qmask >> 12) & 0xF); - - grinder->tccache_qmask[grinder->tccache_w] = b[0]; - grinder->tccache_qindex[grinder->tccache_w] = qindex; - grinder->tccache_w += (b[0] != 0); - - grinder->tccache_qmask[grinder->tccache_w] = b[1]; - grinder->tccache_qindex[grinder->tccache_w] = qindex + 4; - grinder->tccache_w += (b[1] != 0); - - grinder->tccache_qmask[grinder->tccache_w] = b[2]; - grinder->tccache_qindex[grinder->tccache_w] = qindex + 8; - grinder->tccache_w += (b[2] != 0); - - grinder->tccache_qmask[grinder->tccache_w] = b[3]; - grinder->tccache_qindex[grinder->tccache_w] = qindex + 12; - grinder->tccache_w += (b[3] != 0); -} - -static inline int -grinder_next_tc(struct rte_sched_port *port, uint32_t pos) -{ - struct rte_sched_grinder *grinder = port->grinder + pos; - struct rte_mbuf **qbase; - uint32_t qindex; - uint16_t qsize; - - if (grinder->tccache_r == grinder->tccache_w) { - return 0; - } - - qindex = grinder->tccache_qindex[grinder->tccache_r]; - qbase = rte_sched_port_qbase(port, qindex); - qsize = rte_sched_port_qsize(port, qindex); - - grinder->tc_index = (qindex >> 2) & 0x3; - grinder->qmask = grinder->tccache_qmask[grinder->tccache_r]; - grinder->qsize = qsize; - - grinder->qindex[0] = qindex; - grinder->qindex[1] = qindex + 1; - grinder->qindex[2] = qindex + 2; - grinder->qindex[3] = qindex + 3; - - grinder->queue[0] = port->queue + qindex; - grinder->queue[1] = port->queue + qindex + 1; - grinder->queue[2] = port->queue + qindex + 2; - grinder->queue[3] = port->queue + qindex + 3; - - grinder->qbase[0] = qbase; - grinder->qbase[1] = qbase + qsize; - grinder->qbase[2] = qbase + 2 * qsize; - grinder->qbase[3] = qbase + 3 * qsize; - - grinder->tccache_r ++; - return 1; -} - -static inline int -grinder_next_pipe(struct rte_sched_port *port, uint32_t pos) -{ - struct rte_sched_grinder *grinder = port->grinder + pos; - uint32_t pipe_qindex; - uint16_t pipe_qmask; - - if (grinder->pcache_r < grinder->pcache_w) { - pipe_qmask = grinder->pcache_qmask[grinder->pcache_r]; - pipe_qindex = grinder->pcache_qindex[grinder->pcache_r]; - grinder->pcache_r ++; - } else { - uint64_t bmp_slab = 0; - uint32_t bmp_pos = 0; - - /* Get another non-empty pipe group */ - if (unlikely(rte_bitmap_scan(port->bmp, &bmp_pos, &bmp_slab) <= 0)) { - return 0; - } - -#if RTE_SCHED_DEBUG - debug_check_queue_slab(port, bmp_pos, bmp_slab); -#endif - - /* Return if pipe group already in one of the other grinders */ - port->grinder_base_bmp_pos[pos] = RTE_SCHED_BMP_POS_INVALID; - if (unlikely(grinder_pipe_exists(port, bmp_pos))) { - return 0; - } - port->grinder_base_bmp_pos[pos] = bmp_pos; - - /* Install new pipe group into grinder's pipe cache */ - grinder_pcache_populate(port, pos, bmp_pos, bmp_slab); - - pipe_qmask = grinder->pcache_qmask[0]; - pipe_qindex = grinder->pcache_qindex[0]; - grinder->pcache_r = 1; - } - - /* Install new pipe in the grinder */ - grinder->pindex = pipe_qindex >> 4; - grinder->subport = port->subport + (grinder->pindex / port->n_pipes_per_subport); - grinder->pipe = port->pipe + grinder->pindex; - grinder->pipe_params = NULL; /* to be set after the pipe structure is prefetched */ - grinder->productive = 0; - - grinder_tccache_populate(port, pos, pipe_qindex, pipe_qmask); - grinder_next_tc(port, pos); - - /* Check for pipe exhaustion */ - if (grinder->pindex == port->pipe_loop) { - port->pipe_exhaustion = 1; - port->pipe_loop = RTE_SCHED_PIPE_INVALID; - } - - return 1; -} - -#if RTE_SCHED_WRR == 0 - -#define grinder_wrr_load(a,b) - -#define grinder_wrr_store(a,b) - -static inline void -grinder_wrr(struct rte_sched_port *port, uint32_t pos) -{ - struct rte_sched_grinder *grinder = port->grinder + pos; - uint64_t slab = grinder->qmask; - - if (rte_bsf64(slab, &grinder->qpos) == 0) { - rte_panic("grinder wrr\n"); - } -} - -#elif RTE_SCHED_WRR == 1 - -static inline void -grinder_wrr_load(struct rte_sched_port *port, uint32_t pos) -{ - struct rte_sched_grinder *grinder = port->grinder + pos; - struct rte_sched_pipe *pipe = grinder->pipe; - struct rte_sched_pipe_profile *pipe_params = grinder->pipe_params; - uint32_t tc_index = grinder->tc_index; - uint32_t qmask = grinder->qmask; - uint32_t qindex; - - qindex = tc_index * 4; - - grinder->wrr_tokens[0] = ((uint16_t) pipe->wrr_tokens[qindex]) << RTE_SCHED_WRR_SHIFT; - grinder->wrr_tokens[1] = ((uint16_t) pipe->wrr_tokens[qindex + 1]) << RTE_SCHED_WRR_SHIFT; - grinder->wrr_tokens[2] = ((uint16_t) pipe->wrr_tokens[qindex + 2]) << RTE_SCHED_WRR_SHIFT; - grinder->wrr_tokens[3] = ((uint16_t) pipe->wrr_tokens[qindex + 3]) << RTE_SCHED_WRR_SHIFT; - - grinder->wrr_mask[0] = (qmask & 0x1) * 0xFFFF; - grinder->wrr_mask[1] = ((qmask >> 1) & 0x1) * 0xFFFF; - grinder->wrr_mask[2] = ((qmask >> 2) & 0x1) * 0xFFFF; - grinder->wrr_mask[3] = ((qmask >> 3) & 0x1) * 0xFFFF; - - grinder->wrr_cost[0] = pipe_params->wrr_cost[qindex]; - grinder->wrr_cost[1] = pipe_params->wrr_cost[qindex + 1]; - grinder->wrr_cost[2] = pipe_params->wrr_cost[qindex + 2]; - grinder->wrr_cost[3] = pipe_params->wrr_cost[qindex + 3]; -} - -static inline void -grinder_wrr_store(struct rte_sched_port *port, uint32_t pos) -{ - struct rte_sched_grinder *grinder = port->grinder + pos; - struct rte_sched_pipe *pipe = grinder->pipe; - uint32_t tc_index = grinder->tc_index; - uint32_t qindex; - - qindex = tc_index * 4; - - pipe->wrr_tokens[qindex] = (uint8_t) ((grinder->wrr_tokens[0] & grinder->wrr_mask[0]) >> RTE_SCHED_WRR_SHIFT); - pipe->wrr_tokens[qindex + 1] = (uint8_t) ((grinder->wrr_tokens[1] & grinder->wrr_mask[1]) >> RTE_SCHED_WRR_SHIFT); - pipe->wrr_tokens[qindex + 2] = (uint8_t) ((grinder->wrr_tokens[2] & grinder->wrr_mask[2]) >> RTE_SCHED_WRR_SHIFT); - pipe->wrr_tokens[qindex + 3] = (uint8_t) ((grinder->wrr_tokens[3] & grinder->wrr_mask[3]) >> RTE_SCHED_WRR_SHIFT); -} - -static inline void -grinder_wrr(struct rte_sched_port *port, uint32_t pos) -{ - struct rte_sched_grinder *grinder = port->grinder + pos; - uint16_t wrr_tokens_min; - - grinder->wrr_tokens[0] |= ~grinder->wrr_mask[0]; - grinder->wrr_tokens[1] |= ~grinder->wrr_mask[1]; - grinder->wrr_tokens[2] |= ~grinder->wrr_mask[2]; - grinder->wrr_tokens[3] |= ~grinder->wrr_mask[3]; - - grinder->qpos = rte_min_pos_4_u16(grinder->wrr_tokens); - wrr_tokens_min = grinder->wrr_tokens[grinder->qpos]; - - grinder->wrr_tokens[0] -= wrr_tokens_min; - grinder->wrr_tokens[1] -= wrr_tokens_min; - grinder->wrr_tokens[2] -= wrr_tokens_min; - grinder->wrr_tokens[3] -= wrr_tokens_min; -} - -#else - -#error Invalid value for RTE_SCHED_WRR - -#endif /* RTE_SCHED_WRR */ - -#define grinder_evict(port, pos) - -static inline void -grinder_prefetch_pipe(struct rte_sched_port *port, uint32_t pos) -{ - struct rte_sched_grinder *grinder = port->grinder + pos; - - rte_prefetch0(grinder->pipe); - rte_prefetch0(grinder->queue[0]); -} - -static inline void -grinder_prefetch_tc_queue_arrays(struct rte_sched_port *port, uint32_t pos) -{ - struct rte_sched_grinder *grinder = port->grinder + pos; - uint16_t qsize, qr[4]; - - qsize = grinder->qsize; - qr[0] = grinder->queue[0]->qr & (qsize - 1); - qr[1] = grinder->queue[1]->qr & (qsize - 1); - qr[2] = grinder->queue[2]->qr & (qsize - 1); - qr[3] = grinder->queue[3]->qr & (qsize - 1); - - rte_prefetch0(grinder->qbase[0] + qr[0]); - rte_prefetch0(grinder->qbase[1] + qr[1]); - - grinder_wrr_load(port, pos); - grinder_wrr(port, pos); - - rte_prefetch0(grinder->qbase[2] + qr[2]); - rte_prefetch0(grinder->qbase[3] + qr[3]); -} - -static inline void -grinder_prefetch_mbuf(struct rte_sched_port *port, uint32_t pos) -{ - struct rte_sched_grinder *grinder = port->grinder + pos; - uint32_t qpos = grinder->qpos; - struct rte_mbuf **qbase = grinder->qbase[qpos]; - uint16_t qsize = grinder->qsize; - uint16_t qr = grinder->queue[qpos]->qr & (qsize - 1); - - grinder->pkt = qbase[qr]; - rte_prefetch0(grinder->pkt); - - if (unlikely((qr & 0x7) == 7)) { - uint16_t qr_next = (grinder->queue[qpos]->qr + 1) & (qsize - 1); - - rte_prefetch0(qbase + qr_next); - } -} - -static inline uint32_t -grinder_handle(struct rte_sched_port *port, uint32_t pos) -{ - struct rte_sched_grinder *grinder = port->grinder + pos; - - switch (grinder->state) { - case e_GRINDER_PREFETCH_PIPE: - { - if (grinder_next_pipe(port, pos)) { - grinder_prefetch_pipe(port, pos); - port->busy_grinders ++; - - grinder->state = e_GRINDER_PREFETCH_TC_QUEUE_ARRAYS; - return 0; - } - - return 0; - } - - case e_GRINDER_PREFETCH_TC_QUEUE_ARRAYS: - { - struct rte_sched_pipe *pipe = grinder->pipe; - - grinder->pipe_params = port->pipe_profiles + pipe->profile; - grinder_prefetch_tc_queue_arrays(port, pos); - grinder_credits_update(port, pos); - - grinder->state = e_GRINDER_PREFETCH_MBUF; - return 0; - } - - case e_GRINDER_PREFETCH_MBUF: - { - grinder_prefetch_mbuf(port, pos); - - grinder->state = e_GRINDER_READ_MBUF; - return 0; - } - - case e_GRINDER_READ_MBUF: - { - uint32_t result = 0; - - result = grinder_schedule(port, pos); - - /* Look for next packet within the same TC */ - if (result && grinder->qmask) { - grinder_wrr(port, pos); - grinder_prefetch_mbuf(port, pos); - - return 1; - } - grinder_wrr_store(port, pos); - - /* Look for another active TC within same pipe */ - if (grinder_next_tc(port, pos)) { - grinder_prefetch_tc_queue_arrays(port, pos); - - grinder->state = e_GRINDER_PREFETCH_MBUF; - return result; - } - if ((grinder->productive == 0) && (port->pipe_loop == RTE_SCHED_PIPE_INVALID)) { - port->pipe_loop = grinder->pindex; - } - grinder_evict(port, pos); - - /* Look for another active pipe */ - if (grinder_next_pipe(port, pos)) { - grinder_prefetch_pipe(port, pos); - - grinder->state = e_GRINDER_PREFETCH_TC_QUEUE_ARRAYS; - return result; - } - - /* No active pipe found */ - port->busy_grinders --; - - grinder->state = e_GRINDER_PREFETCH_PIPE; - return result; - } - - default: - rte_panic("Algorithmic error (invalid state)\n"); - return 0; - } -} - -static inline void -rte_sched_port_time_resync(struct rte_sched_port *port) -{ - uint64_t cycles = rte_get_tsc_cycles(); - uint64_t cycles_diff = cycles - port->time_cpu_cycles; - double bytes_diff = ((double) cycles_diff) / port->cycles_per_byte; - - /* Advance port time */ - port->time_cpu_cycles = cycles; - port->time_cpu_bytes += (uint64_t) bytes_diff; - if (port->time < port->time_cpu_bytes) { - port->time = port->time_cpu_bytes; - } - - /* Reset pipe loop detection */ - port->pipe_loop = RTE_SCHED_PIPE_INVALID; -} - -static inline int -rte_sched_port_exceptions(struct rte_sched_port *port, int second_pass) -{ - int exceptions; - - /* Check if any exception flag is set */ - exceptions = (second_pass && port->busy_grinders == 0) || - (port->pipe_exhaustion == 1); - - /* Clear exception flags */ - port->pipe_exhaustion = 0; - - return exceptions; -} - -int -rte_sched_port_dequeue(struct rte_sched_port *port, struct rte_mbuf **pkts, uint32_t n_pkts) -{ - uint32_t i, count; - - port->pkts_out = pkts; - port->n_pkts_out = 0; - - rte_sched_port_time_resync(port); - - /* Take each queue in the grinder one step further */ - for (i = 0, count = 0; ; i ++) { - count += grinder_handle(port, i & (RTE_SCHED_PORT_N_GRINDERS - 1)); - if ((count == n_pkts) || - rte_sched_port_exceptions(port, i >= RTE_SCHED_PORT_N_GRINDERS)) { - break; - } - } - - return count; -} diff --git a/src/dpdk_lib18/librte_sched/rte_sched.h b/src/dpdk_lib18/librte_sched/rte_sched.h deleted file mode 100755 index e6bba22e..00000000 --- a/src/dpdk_lib18/librte_sched/rte_sched.h +++ /dev/null @@ -1,442 +0,0 @@ -/*- - * BSD LICENSE - * - * Copyright(c) 2010-2014 Intel Corporation. All rights reserved. - * All rights reserved. - * - * Redistribution and use in source and binary forms, with or without - * modification, are permitted provided that the following conditions - * are met: - * - * * Redistributions of source code must retain the above copyright - * notice, this list of conditions and the following disclaimer. - * * Redistributions in binary form must reproduce the above copyright - * notice, this list of conditions and the following disclaimer in - * the documentation and/or other materials provided with the - * distribution. - * * Neither the name of Intel Corporation nor the names of its - * contributors may be used to endorse or promote products derived - * from this software without specific prior written permission. - * - * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS - * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT - * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR - * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT - * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, - * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT - * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, - * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY - * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT - * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE - * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. - */ - -#ifndef __INCLUDE_RTE_SCHED_H__ -#define __INCLUDE_RTE_SCHED_H__ - -#ifdef __cplusplus -extern "C" { -#endif - -/** - * @file - * RTE Hierarchical Scheduler - * - * The hierarchical scheduler prioritizes the transmission of packets from different - * users and traffic classes according to the Service Level Agreements (SLAs) defined - * for the current network node. - * - * The scheduler supports thousands of packet queues grouped under a 5-level hierarchy: - * 1. Port: - * - Typical usage: output Ethernet port; - * - Multiple ports are scheduled in round robin order with equal priority; - * 2. Subport: - * - Typical usage: group of users; - * - Traffic shaping using the token bucket algorithm (one bucket per subport); - * - Upper limit enforced per traffic class at subport level; - * - Lower priority traffic classes able to reuse subport bandwidth currently - * unused by higher priority traffic classes of the same subport; - * - When any subport traffic class is oversubscribed (configuration time - * event), the usage of subport member pipes with high demand for that - * traffic class pipes is truncated to a dynamically adjusted value with no - * impact to low demand pipes; - * 3. Pipe: - * - Typical usage: individual user/subscriber; - * - Traffic shaping using the token bucket algorithm (one bucket per pipe); - * 4. Traffic class: - * - Traffic classes of the same pipe handled in strict priority order; - * - Upper limit enforced per traffic class at the pipe level; - * - Lower priority traffic classes able to reuse pipe bandwidth currently - * unused by higher priority traffic classes of the same pipe; - * 5. Queue: - * - Typical usage: queue hosting packets from one or multiple connections - * of same traffic class belonging to the same user; - * - Weighted Round Robin (WRR) is used to service the queues within same - * pipe traffic class. - * - ***/ - -#include <sys/types.h> -#include <rte_mbuf.h> -#include <rte_meter.h> - -/** Random Early Detection (RED) */ -#ifdef RTE_SCHED_RED -#include "rte_red.h" -#endif - -/** Number of traffic classes per pipe (as well as subport). Cannot be changed. */ -#define RTE_SCHED_TRAFFIC_CLASSES_PER_PIPE 4 - -/** Number of queues per pipe traffic class. Cannot be changed. */ -#define RTE_SCHED_QUEUES_PER_TRAFFIC_CLASS 4 - -/** Number of queues per pipe. */ -#define RTE_SCHED_QUEUES_PER_PIPE \ - (RTE_SCHED_TRAFFIC_CLASSES_PER_PIPE * \ - RTE_SCHED_QUEUES_PER_TRAFFIC_CLASS) - -/** Maximum number of pipe profiles that can be defined per port. Compile-time configurable.*/ -#ifndef RTE_SCHED_PIPE_PROFILES_PER_PORT -#define RTE_SCHED_PIPE_PROFILES_PER_PORT 256 -#endif - -/** Ethernet framing overhead. Overhead fields per Ethernet frame: - 1. Preamble: 7 bytes; - 2. Start of Frame Delimiter (SFD): 1 byte; - 3. Frame Check Sequence (FCS): 4 bytes; - 4. Inter Frame Gap (IFG): 12 bytes. -The FCS is considered overhead only if not included in the packet length (field pkt_len -of struct rte_mbuf). */ -#ifndef RTE_SCHED_FRAME_OVERHEAD_DEFAULT -#define RTE_SCHED_FRAME_OVERHEAD_DEFAULT 24 -#endif - -/** Subport configuration parameters. The period and credits_per_period parameters are measured -in bytes, with one byte meaning the time duration associated with the transmission of one byte -on the physical medium of the output port, with pipe or pipe traffic class rate (measured as -percentage of output port rate) determined as credits_per_period divided by period. One credit -represents one byte. */ -struct rte_sched_subport_params { - /* Subport token bucket */ - uint32_t tb_rate; /**< Subport token bucket rate (measured in bytes per second) */ - uint32_t tb_size; /**< Subport token bucket size (measured in credits) */ - - /* Subport traffic classes */ - uint32_t tc_rate[RTE_SCHED_TRAFFIC_CLASSES_PER_PIPE]; /**< Subport traffic class rates (measured in bytes per second) */ - uint32_t tc_period; /**< Enforcement period for traffic class rates (measured in milliseconds) */ -}; - -/** Subport statistics */ -struct rte_sched_subport_stats { - /* Packets */ - uint32_t n_pkts_tc[RTE_SCHED_TRAFFIC_CLASSES_PER_PIPE]; /**< Number of packets successfully written to current - subport for each traffic class */ - uint32_t n_pkts_tc_dropped[RTE_SCHED_TRAFFIC_CLASSES_PER_PIPE]; /**< Number of packets dropped by the current - subport for each traffic class due to subport queues being full or congested*/ - - /* Bytes */ - uint32_t n_bytes_tc[RTE_SCHED_TRAFFIC_CLASSES_PER_PIPE]; /**< Number of bytes successfully written to current - subport for each traffic class*/ - uint32_t n_bytes_tc_dropped[RTE_SCHED_TRAFFIC_CLASSES_PER_PIPE]; /**< Number of bytes dropped by the current - subport for each traffic class due to subport queues being full or congested */ -}; - -/** Pipe configuration parameters. The period and credits_per_period parameters are measured -in bytes, with one byte meaning the time duration associated with the transmission of one byte -on the physical medium of the output port, with pipe or pipe traffic class rate (measured as -percentage of output port rate) determined as credits_per_period divided by period. One credit -represents one byte. */ -struct rte_sched_pipe_params { - /* Pipe token bucket */ - uint32_t tb_rate; /**< Pipe token bucket rate (measured in bytes per second) */ - uint32_t tb_size; /**< Pipe token bucket size (measured in credits) */ - - /* Pipe traffic classes */ - uint32_t tc_rate[RTE_SCHED_TRAFFIC_CLASSES_PER_PIPE]; /**< Pipe traffic class rates (measured in bytes per second) */ - uint32_t tc_period; /**< Enforcement period for pipe traffic class rates (measured in milliseconds) */ -#ifdef RTE_SCHED_SUBPORT_TC_OV - uint8_t tc_ov_weight; /**< Weight for the current pipe in the event of subport traffic class 3 oversubscription */ -#endif - - /* Pipe queues */ - uint8_t wrr_weights[RTE_SCHED_QUEUES_PER_PIPE]; /**< WRR weights for the queues of the current pipe */ -}; - -/** Queue statistics */ -struct rte_sched_queue_stats { - /* Packets */ - uint32_t n_pkts; /**< Number of packets successfully written to current queue */ - uint32_t n_pkts_dropped; /**< Number of packets dropped due to current queue being full or congested */ - - /* Bytes */ - uint32_t n_bytes; /**< Number of bytes successfully written to current queue */ - uint32_t n_bytes_dropped; /**< Number of bytes dropped due to current queue being full or congested */ -}; - -/** Port configuration parameters. */ -struct rte_sched_port_params { - const char *name; /**< Literal string to be associated to the current port scheduler instance */ - int socket; /**< CPU socket ID where the memory for port scheduler should be allocated */ - uint32_t rate; /**< Output port rate (measured in bytes per second) */ - uint32_t mtu; /**< Maximum Ethernet frame size (measured in bytes). Should not include the framing overhead. */ - uint32_t frame_overhead; /**< Framing overhead per packet (measured in bytes) */ - uint32_t n_subports_per_port; /**< Number of subports for the current port scheduler instance*/ - uint32_t n_pipes_per_subport; /**< Number of pipes for each port scheduler subport */ - uint16_t qsize[RTE_SCHED_TRAFFIC_CLASSES_PER_PIPE]; /**< Packet queue size for each traffic class. All queues - within the same pipe traffic class have the same size. Queues from - different pipes serving the same traffic class have the same size. */ - struct rte_sched_pipe_params *pipe_profiles; /**< Pipe profile table defined for current port scheduler instance. - Every pipe of the current port scheduler is configured using one of the - profiles from this table. */ - uint32_t n_pipe_profiles; /**< Number of profiles in the pipe profile table */ -#ifdef RTE_SCHED_RED - struct rte_red_params red_params[RTE_SCHED_TRAFFIC_CLASSES_PER_PIPE][e_RTE_METER_COLORS]; /**< RED parameters */ -#endif -}; - -/** Path through the scheduler hierarchy used by the scheduler enqueue operation to -identify the destination queue for the current packet. Stored in the field hash.sched -of struct rte_mbuf of each packet, typically written by the classification stage and read by -scheduler enqueue.*/ -struct rte_sched_port_hierarchy { - uint32_t queue:2; /**< Queue ID (0 .. 3) */ - uint32_t traffic_class:2; /**< Traffic class ID (0 .. 3)*/ - uint32_t pipe:20; /**< Pipe ID */ - uint32_t subport:6; /**< Subport ID */ - uint32_t color:2; /**< Color */ -}; - -/* - * Configuration - * - ***/ - -/** - * Hierarchical scheduler port configuration - * - * @param params - * Port scheduler configuration parameter structure - * @return - * Handle to port scheduler instance upon success or NULL otherwise. - */ -struct rte_sched_port * -rte_sched_port_config(struct rte_sched_port_params *params); - -/** - * Hierarchical scheduler port free - * - * @param port - * Handle to port scheduler instance - */ -void -rte_sched_port_free(struct rte_sched_port *port); - -/** - * Hierarchical scheduler subport configuration - * - * @param port - * Handle to port scheduler instance - * @param subport_id - * Subport ID - * @param params - * Subport configuration parameters - * @return - * 0 upon success, error code otherwise - */ -int -rte_sched_subport_config(struct rte_sched_port *port, - uint32_t subport_id, - struct rte_sched_subport_params *params); - -/** - * Hierarchical scheduler pipe configuration - * - * @param port - * Handle to port scheduler instance - * @param subport_id - * Subport ID - * @param pipe_id - * Pipe ID within subport - * @param pipe_profile - * ID of port-level pre-configured pipe profile - * @return - * 0 upon success, error code otherwise - */ -int -rte_sched_pipe_config(struct rte_sched_port *port, - uint32_t subport_id, - uint32_t pipe_id, - int32_t pipe_profile); - -/** - * Hierarchical scheduler memory footprint size per port - * - * @param params - * Port scheduler configuration parameter structure - * @return - * Memory footprint size in bytes upon success, 0 otherwise - */ -uint32_t -rte_sched_port_get_memory_footprint(struct rte_sched_port_params *params); - -/* - * Statistics - * - ***/ - -/** - * Hierarchical scheduler subport statistics read - * - * @param port - * Handle to port scheduler instance - * @param subport_id - * Subport ID - * @param stats - * Pointer to pre-allocated subport statistics structure where the statistics - * counters should be stored - * @param tc_ov - * Pointer to pre-allocated 4-entry array where the oversubscription status for - * each of the 4 subport traffic classes should be stored. - * @return - * 0 upon success, error code otherwise - */ -int -rte_sched_subport_read_stats(struct rte_sched_port *port, - uint32_t subport_id, - struct rte_sched_subport_stats *stats, - uint32_t *tc_ov); - -/** - * Hierarchical scheduler queue statistics read - * - * @param port - * Handle to port scheduler instance - * @param queue_id - * Queue ID within port scheduler - * @param stats - * Pointer to pre-allocated subport statistics structure where the statistics - * counters should be stored - * @param qlen - * Pointer to pre-allocated variable where the current queue length should be stored. - * @return - * 0 upon success, error code otherwise - */ -int -rte_sched_queue_read_stats(struct rte_sched_port *port, - uint32_t queue_id, - struct rte_sched_queue_stats *stats, - uint16_t *qlen); - -/* - * Run-time - * - ***/ - -/** - * Scheduler hierarchy path write to packet descriptor. Typically called by the - * packet classification stage. - * - * @param pkt - * Packet descriptor handle - * @param subport - * Subport ID - * @param pipe - * Pipe ID within subport - * @param traffic_class - * Traffic class ID within pipe (0 .. 3) - * @param queue - * Queue ID within pipe traffic class (0 .. 3) - */ -static inline void -rte_sched_port_pkt_write(struct rte_mbuf *pkt, - uint32_t subport, uint32_t pipe, uint32_t traffic_class, uint32_t queue, enum rte_meter_color color) -{ - struct rte_sched_port_hierarchy *sched = (struct rte_sched_port_hierarchy *) &pkt->hash.sched; - - sched->color = (uint32_t) color; - sched->subport = subport; - sched->pipe = pipe; - sched->traffic_class = traffic_class; - sched->queue = queue; -} - -/** - * Scheduler hierarchy path read from packet descriptor (struct rte_mbuf). Typically - * called as part of the hierarchical scheduler enqueue operation. The subport, - * pipe, traffic class and queue parameters need to be pre-allocated by the caller. - * - * @param pkt - * Packet descriptor handle - * @param subport - * Subport ID - * @param pipe - * Pipe ID within subport - * @param traffic_class - * Traffic class ID within pipe (0 .. 3) - * @param queue - * Queue ID within pipe traffic class (0 .. 3) - * - */ -static inline void -rte_sched_port_pkt_read_tree_path(struct rte_mbuf *pkt, uint32_t *subport, uint32_t *pipe, uint32_t *traffic_class, uint32_t *queue) -{ - struct rte_sched_port_hierarchy *sched = (struct rte_sched_port_hierarchy *) &pkt->hash.sched; - - *subport = sched->subport; - *pipe = sched->pipe; - *traffic_class = sched->traffic_class; - *queue = sched->queue; -} - -static inline enum rte_meter_color -rte_sched_port_pkt_read_color(struct rte_mbuf *pkt) -{ - struct rte_sched_port_hierarchy *sched = (struct rte_sched_port_hierarchy *) &pkt->hash.sched; - - return (enum rte_meter_color) sched->color; -} - -/** - * Hierarchical scheduler port enqueue. Writes up to n_pkts to port scheduler and - * returns the number of packets actually written. For each packet, the port scheduler - * queue to write the packet to is identified by reading the hierarchy path from the - * packet descriptor; if the queue is full or congested and the packet is not written - * to the queue, then the packet is automatically dropped without any action required - * from the caller. - * - * @param port - * Handle to port scheduler instance - * @param pkts - * Array storing the packet descriptor handles - * @param n_pkts - * Number of packets to enqueue from the pkts array into the port scheduler - * @return - * Number of packets successfully enqueued - */ -int -rte_sched_port_enqueue(struct rte_sched_port *port, struct rte_mbuf **pkts, uint32_t n_pkts); - -/** - * Hierarchical scheduler port dequeue. Reads up to n_pkts from the port scheduler - * and stores them in the pkts array and returns the number of packets actually read. - * The pkts array needs to be pre-allocated by the caller with at least n_pkts entries. - * - * @param port - * Handle to port scheduler instance - * @param pkts - * Pre-allocated packet descriptor array where the packets dequeued from the port - * scheduler should be stored - * @param n_pkts - * Number of packets to dequeue from the port scheduler - * @return - * Number of packets successfully dequeued and placed in the pkts array - */ -int -rte_sched_port_dequeue(struct rte_sched_port *port, struct rte_mbuf **pkts, uint32_t n_pkts); - -#ifdef __cplusplus -} -#endif - -#endif /* __INCLUDE_RTE_SCHED_H__ */ diff --git a/src/dpdk_lib18/librte_sched/rte_sched_common.h b/src/dpdk_lib18/librte_sched/rte_sched_common.h deleted file mode 100755 index 8920adec..00000000 --- a/src/dpdk_lib18/librte_sched/rte_sched_common.h +++ /dev/null @@ -1,129 +0,0 @@ -/*- - * BSD LICENSE - * - * Copyright(c) 2010-2014 Intel Corporation. All rights reserved. - * All rights reserved. - * - * Redistribution and use in source and binary forms, with or without - * modification, are permitted provided that the following conditions - * are met: - * - * * Redistributions of source code must retain the above copyright - * notice, this list of conditions and the following disclaimer. - * * Redistributions in binary form must reproduce the above copyright - * notice, this list of conditions and the following disclaimer in - * the documentation and/or other materials provided with the - * distribution. - * * Neither the name of Intel Corporation nor the names of its - * contributors may be used to endorse or promote products derived - * from this software without specific prior written permission. - * - * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS - * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT - * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR - * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT - * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, - * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT - * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, - * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY - * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT - * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE - * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. - */ - -#ifndef __INCLUDE_RTE_SCHED_COMMON_H__ -#define __INCLUDE_RTE_SCHED_COMMON_H__ - -#ifdef __cplusplus -extern "C" { -#endif - -#include <sys/types.h> - -#define __rte_aligned_16 __attribute__((__aligned__(16))) - -static inline uint32_t -rte_sched_min_val_2_u32(uint32_t x, uint32_t y) -{ - return (x < y)? x : y; -} - -#if 0 -static inline uint32_t -rte_min_pos_4_u16(uint16_t *x) -{ - uint32_t pos0, pos1; - - pos0 = (x[0] <= x[1])? 0 : 1; - pos1 = (x[2] <= x[3])? 2 : 3; - - return (x[pos0] <= x[pos1])? pos0 : pos1; -} - -#else - -/* simplified version to remove branches with CMOV instruction */ -static inline uint32_t -rte_min_pos_4_u16(uint16_t *x) -{ - uint32_t pos0 = 0; - uint32_t pos1 = 2; - - if (x[1] <= x[0]) pos0 = 1; - if (x[3] <= x[2]) pos1 = 3; - if (x[pos1] <= x[pos0]) pos0 = pos1; - - return pos0; -} - -#endif - -/* - * Compute the Greatest Common Divisor (GCD) of two numbers. - * This implementation uses Euclid's algorithm: - * gcd(a, 0) = a - * gcd(a, b) = gcd(b, a mod b) - * - */ -static inline uint32_t -rte_get_gcd(uint32_t a, uint32_t b) -{ - uint32_t c; - - if (a == 0) - return b; - if (b == 0) - return a; - - if (a < b) { - c = a; - a = b; - b = c; - } - - while (b != 0) { - c = a % b; - a = b; - b = c; - } - - return a; -} - -/* - * Compute the Lowest Common Denominator (LCD) of two numbers. - * This implementation computes GCD first: - * LCD(a, b) = (a * b) / GCD(a, b) - * - */ -static inline uint32_t -rte_get_lcd(uint32_t a, uint32_t b) -{ - return (a * b) / rte_get_gcd(a, b); -} - -#ifdef __cplusplus -} -#endif - -#endif /* __INCLUDE_RTE_SCHED_COMMON_H__ */ |