#ifndef __NET_SCHED_CODEL_IMPL_H #define __NET_SCHED_CODEL_IMPL_H /* * Codel - The Controlled-Delay Active Queue Management algorithm * * Copyright (C) 2011-2012 Kathleen Nichols * Copyright (C) 2011-2012 Van Jacobson * Copyright (C) 2012 Michael D. Taht * Copyright (C) 2012,2015 Eric Dumazet * * Redistribution and use in source and binary forms, with or without * modification, are permitted provided that the following conditions * are met: * 1. Redistributions of source code must retain the above copyright * notice, this list of conditions, and the following disclaimer, * without modification. * 2. 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. * 3. The names of the authors may not be used to endorse or promote products * derived from this software without specific prior written permission. * * Alternatively, provided that this notice is retained in full, this * software may be distributed under the terms of the GNU General * Public License ("GPL") version 2, in which case the provisions of the * GPL apply INSTEAD OF those given above. * * 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. * */ /* Controlling Queue Delay (CoDel) algorithm * ========================================= * Source : Kathleen Nichols and Van Jacobson * http://queue.acm.org/detail.cfm?id=2209336 * * Implemented on linux by Dave Taht and Eric Dumazet */ static void codel_params_init(struct codel_params *params) { params->interval = MS2TIME(100); params->target = MS2TIME(5); params->ce_threshold = CODEL_DISABLED_THRESHOLD; params->ecn = false; } static void codel_vars_init(struct codel_vars *vars) { memset(vars, 0, sizeof(*vars)); } static void codel_stats_init(struct codel_stats *stats) { stats->maxpacket = 0; } /* * http://en.wikipedia.org/wiki/Methods_of_computing_square_roots#Iterative_methods_for_reciprocal_square_roots * new_invsqrt = (invsqrt / 2) * (3 - count * invsqrt^2) * * Here, invsqrt is a fixed point number (< 1.0), 32bit mantissa, aka Q0.32 */ static void codel_Newton_step(struct codel_vars *vars) { u32 invsqrt = ((u32)vars->rec_inv_sqrt) << REC_INV_SQRT_SHIFT; u32 invsqrt2 = ((u64)invsqrt * invsqrt) >> 32; u64 val = (3LL << 32) - ((u64)vars->count * invsqrt2); val >>= 2; /* avoid overflow in following multiply */ val = (val * invsqrt) >> (32 - 2 + 1); vars->rec_inv_sqrt = val >> REC_INV_SQRT_SHIFT; } /* * CoDel control_law is t + interval/sqrt(count) * We maintain in rec_inv_sqrt the reciprocal value of sqrt(count) to avoid * both sqrt() and divide operation. */ static codel_time_t codel_control_law(codel_time_t t, codel_time_t interval, u32 rec_inv_sqrt) { return t + reciprocal_scale(interval, rec_inv_sqrt << REC_INV_SQRT_SHIFT); } static bool codel_should_drop(const struct sk_buff *skb, void *ctx, struct codel_vars *vars, struct codel_params *params, struct codel_stats *stats, codel_skb_len_t skb_len_func, codel_skb_time_t skb_time_func, u32 *backlog, codel_time_t now) { bool ok_to_drop; u32 skb_len; if (!skb) { vars->first_above_time = 0; return false; } skb_len = skb_len_func(skb); vars->ldelay = now - skb_time_func(skb); if (unlikely(skb_len > stats->maxpacket)) stats->maxpacket = skb_len; if (codel_time_before(vars->ldelay, params->target) || *backlog <= params->mtu) { /* went below - stay below for at least interval */ vars->first_above_time = 0; return false; } ok_to_drop = false; if (vars->first_above_time == 0) { /* just went above from below. If we stay above * for at least interval we'll say it's ok to drop */ vars->first_above_time = now + params->interval; } else if (codel_time_after(now, vars->first_above_time)) { ok_to_drop = true; } return ok_to_drop; } static struct sk_buff *codel_dequeue(void *ctx, u32 *backlog, struct codel_params *params, struct codel_vars *vars, struct codel_stats *stats, codel_skb_len_t skb_len_func, codel_skb_time_t skb_time_func, codel_skb_drop_t drop_func, codel_skb_dequeue_t dequeue_func) { struct sk_buff *skb = dequeue_func(vars, ctx); codel_time_t now; bool drop; if (!skb) { vars->dropping = false; return skb; } now = codel_get_time(); drop = codel_should_drop(skb, ctx, vars, params, stats, skb_len_func, skb_time_func, backlog, now); if (vars->dropping) { if (!drop) { /* sojourn time below target - leave dropping state */ vars->dropping = false; } else if (codel_time_after_eq(now, vars->drop_next)) { /* It's time for the next drop. Drop the current * packet and dequeue the next. The dequeue might * take us out of dropping state. * If not, schedule the next drop. * A large backlog might result in drop rates so high * that the next drop should happen now, * hence the while loop. */ while (vars->dropping && codel_time_after_eq(now, vars->drop_next)) { vars->count++; /* dont care of possible wrap * since there is no more divide */ codel_Newton_step(vars); if (params->ecn && INET_ECN_set_ce(skb)) { stats->ecn_mark++; vars->drop_next = codel_control_law(vars->drop_next, params->interval, vars->rec_inv_sqrt); goto end; } stats->drop_len += skb_len_func(skb); drop_func(skb, ctx); stats->drop_count++; skb = dequeue_func(vars, ctx); if (!codel_should_drop(skb, ctx, vars, params, stats, skb_len_func, skb_time_func, backlog, now)) { /* leave dropping state */ vars->dropping = false; } else { /* and schedule the next drop */ vars->drop_next = codel_control_law(vars->drop_next, params->interval, vars->rec_inv_sqrt); } } } } else if (drop) { u32 delta; if (params->ecn && INET_ECN_set_ce(skb)) { stats->ecn_mark++; } else { stats->drop_len += skb_len_func(skb); drop_func(skb, ctx); stats->drop_count++; skb = dequeue_func(vars, ctx); drop = codel_should_drop(skb, ctx, vars, params, stats, skb_len_func, skb_time_func, backlog, now); } vars->dropping = true; /* if min went above target close to when we last went below it * assume that the drop rate that controlled the queue on the * last cycle is a good starting point to control it now. */ delta = vars->count - vars->lastcount; if (delta > 1 && codel_time_before(now - vars->drop_next, 16 * params->interval)) { vars->count = delta; /* we dont care if rec_inv_sqrt approximation * is not very precise : * Next Newton steps will correct it quadratically. */ codel_Newton_step(vars); } else { vars->count = 1; vars->rec_inv_sqrt = ~0U >> REC_INV_SQRT_SHIFT; } vars->lastcount = vars->count; vars->drop_next = codel_control_law(now, params->interval, vars->rec_inv_sqrt); } end: if (skb && codel_time_after(vars->ldelay, params->ce_threshold) && INET_ECN_set_ce(skb)) stats->ce_mark++; return skb; } #endif some generic thing that doesn't make the odd semantics explicit. So this introduces the new architecture primitive clear_bit_unlock_is_negative_byte(); and adds the trivial implementation for x86. We have a generic non-optimized fallback (that just does a "clear_bit()"+"test_bit(7)" combination) which can be overridden by any architecture that can do better. According to Nick, Power has the same hickup x86 has, for example, but some other architectures may not even care. All these optimizations mean that my page locking stress-test (which is just executing a lot of small short-lived shell scripts: "make test" in the git source tree) no longer makes our page locking look horribly bad. Before all these optimizations, just the unlock_page() costs were just over 3% of all CPU overhead on "make test". After this, it's down to 0.66%, so just a quarter of the cost it used to be. (The difference on NUMA is bigger, but there this micro-optimization is likely less noticeable, since the big issue on NUMA was not the accesses to 'struct page', but the waitqueue accesses that were already removed by Nick's earlier commit). Acked-by: Nick Piggin <npiggin@gmail.com> Cc: Dave Hansen <dave.hansen@linux.intel.com> Cc: Bob Peterson <rpeterso@redhat.com> Cc: Steven Whitehouse <swhiteho@redhat.com> Cc: Andrew Lutomirski <luto@kernel.org> Cc: Andreas Gruenbacher <agruenba@redhat.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Mel Gorman <mgorman@techsingularity.net> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Diffstat (limited to 'net/ipv4/inet_connection_sock.c')