DPDK patches and discussions
 help / color / mirror / Atom feed
From: Jia He <hejianet@gmail.com>
To: jerin.jacob@caviumnetworks.com, dev@dpdk.org, olivier.matz@6wind.com
Cc: konstantin.ananyev@intel.com, bruce.richardson@intel.com,
	jianbo.liu@arm.com, hemant.agrawal@nxp.com,
	Jia He <hejianet@gmail.com>, Jia He <jia.he@hxt-semitech.com>
Subject: [dpdk-dev] [PATCH v4 4/4] ring: introduce new header file to support C11 memory model
Date: Wed,  8 Nov 2017 09:54:41 +0000	[thread overview]
Message-ID: <1510134881-22987-5-git-send-email-hejianet@gmail.com> (raw)
In-Reply-To: <1510134881-22987-1-git-send-email-hejianet@gmail.com>

To fix the cpu reorder race condition in enque/deque, there are 2 options
suggested by Jerin:
1. use rte_smp_rmb
2. use load_acquire/store_release(refer to [1]).
for the 2nd option, CONFIG_RTE_RING_USE_C11_MEM_MODEL is provided, and by
default it is "y" only on arm64.

The reason why providing 2 options is due to the performance benchmark
difference in different arm machines, refer to [2].

We haven't tested on ppc64. If anyone verifies it, he can add
CONFIG_RTE_RING_USE_C11_MEM_MODEL=y to ppc64 config files.

[1] https://github.com/freebsd/freebsd/blob/master/sys/sys/buf_ring.h#L170
[2] http://dpdk.org/ml/archives/dev/2017-October/080861.html

Signed-off-by: Jia He <jia.he@hxt-semitech.com>
Suggested-by: Jerin Jacob <jerin.jacob@caviumnetworks.com>
---
 config/common_armv8a_linuxapp      |   2 +
 lib/librte_ring/Makefile           |   3 +-
 lib/librte_ring/rte_ring.h         |  14 ++-
 lib/librte_ring/rte_ring_c11_mem.h | 185 +++++++++++++++++++++++++++++++++++++
 4 files changed, 202 insertions(+), 2 deletions(-)
 create mode 100644 lib/librte_ring/rte_ring_c11_mem.h

diff --git a/config/common_armv8a_linuxapp b/config/common_armv8a_linuxapp
index 6732d1e..5b7b2eb 100644
--- a/config/common_armv8a_linuxapp
+++ b/config/common_armv8a_linuxapp
@@ -49,3 +49,5 @@ CONFIG_RTE_LIBRTE_SFC_EFX_PMD=n
 CONFIG_RTE_LIBRTE_AVP_PMD=n
 
 CONFIG_RTE_SCHED_VECTOR=n
+
+CONFIG_RTE_RING_USE_C11_MEM_MODEL=y
diff --git a/lib/librte_ring/Makefile b/lib/librte_ring/Makefile
index c959945..a2682e7 100644
--- a/lib/librte_ring/Makefile
+++ b/lib/librte_ring/Makefile
@@ -46,6 +46,7 @@ SRCS-$(CONFIG_RTE_LIBRTE_RING) := rte_ring.c
 
 # install includes
 SYMLINK-$(CONFIG_RTE_LIBRTE_RING)-include := rte_ring.h \
-					rte_ring_generic.h
+					rte_ring_generic.h \
+					rte_ring_c11_mem.h
 
 include $(RTE_SDK)/mk/rte.lib.mk
diff --git a/lib/librte_ring/rte_ring.h b/lib/librte_ring/rte_ring.h
index 519614c..3343eba 100644
--- a/lib/librte_ring/rte_ring.h
+++ b/lib/librte_ring/rte_ring.h
@@ -356,8 +356,20 @@ void rte_ring_dump(FILE *f, const struct rte_ring *r);
 	} \
 } while (0)
 
-/* Move common functions to generic file */
+/* Between load and load. there might be cpu reorder in weak model
+ * (powerpc/arm).
+ * There are 2 choices for the users
+ * 1.use rmb() memory barrier
+ * 2.use one-direcion load_acquire/store_release barrier,defined by
+ * CONFIG_RTE_RING_USE_C11_MEM_MODEL=y
+ * It depends on performance test results.
+ * By default, move common functions to rte_ring_generic.h
+ */
+#ifdef RTE_RING_USE_C11_MEM_MODEL
+#include "rte_ring_c11_mem.h"
+#else
 #include "rte_ring_generic.h"
+#endif
 
 /**
  * @internal Enqueue several objects on the ring
diff --git a/lib/librte_ring/rte_ring_c11_mem.h b/lib/librte_ring/rte_ring_c11_mem.h
new file mode 100644
index 0000000..c167b46
--- /dev/null
+++ b/lib/librte_ring/rte_ring_c11_mem.h
@@ -0,0 +1,185 @@
+/*-
+ *   BSD LICENSE
+ *
+ *   Copyright(c) 2017 hxt-semitech. 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 hxt-semitech 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_RING_C11_MEM_H_
+#define _RTE_RING_C11_MEM_H_
+
+static __rte_always_inline void
+update_tail(struct rte_ring_headtail *ht, uint32_t old_val, uint32_t new_val,
+		uint32_t single, uint32_t enqueue)
+{
+	RTE_SET_USED(enqueue);
+
+	/*
+	 * If there are other enqueues/dequeues in progress that preceded us,
+	 * we need to wait for them to complete
+	 */
+	if (!single)
+		while (unlikely(ht->tail != old_val))
+			rte_pause();
+
+	__atomic_store_n(&ht->tail, new_val, __ATOMIC_RELEASE);
+}
+
+/**
+ * @internal This function updates the producer head for enqueue
+ *
+ * @param r
+ *   A pointer to the ring structure
+ * @param is_sp
+ *   Indicates whether multi-producer path is needed or not
+ * @param n
+ *   The number of elements we will want to enqueue, i.e. how far should the
+ *   head be moved
+ * @param behavior
+ *   RTE_RING_QUEUE_FIXED:    Enqueue a fixed number of items from a ring
+ *   RTE_RING_QUEUE_VARIABLE: Enqueue as many items as possible from ring
+ * @param old_head
+ *   Returns head value as it was before the move, i.e. where enqueue starts
+ * @param new_head
+ *   Returns the current/new head value i.e. where enqueue finishes
+ * @param free_entries
+ *   Returns the amount of free space in the ring BEFORE head was moved
+ * @return
+ *   Actual number of objects enqueued.
+ *   If behavior == RTE_RING_QUEUE_FIXED, this will be 0 or n only.
+ */
+static __rte_always_inline unsigned int
+__rte_ring_move_prod_head(struct rte_ring *r, int is_sp,
+		unsigned int n, enum rte_ring_queue_behavior behavior,
+		uint32_t *old_head, uint32_t *new_head,
+		uint32_t *free_entries)
+{
+	const uint32_t capacity = r->capacity;
+	unsigned int max = n;
+	int success;
+
+	do {
+		/* Reset n to the initial burst count */
+		n = max;
+
+		*old_head = __atomic_load_n(&r->prod.head,
+					__ATOMIC_ACQUIRE);
+		const uint32_t cons_tail = r->cons.tail;
+		/*
+		 *  The subtraction is done between two unsigned 32bits value
+		 * (the result is always modulo 32 bits even if we have
+		 * *old_head > cons_tail). So 'free_entries' is always between 0
+		 * and capacity (which is < size).
+		 */
+		*free_entries = (capacity + cons_tail - *old_head);
+
+		/* check that we have enough room in ring */
+		if (unlikely(n > *free_entries))
+			n = (behavior == RTE_RING_QUEUE_FIXED) ?
+					0 : *free_entries;
+
+		if (n == 0)
+			return 0;
+
+		*new_head = *old_head + n;
+		if (is_sp)
+			r->prod.head = *new_head, success = 1;
+		else
+			success = __atomic_compare_exchange_n(&r->prod.head,
+					old_head, *new_head,
+					0, __ATOMIC_ACQUIRE,
+					__ATOMIC_RELAXED);
+	} while (unlikely(success == 0));
+	return n;
+}
+
+/**
+ * @internal This function updates the consumer head for dequeue
+ *
+ * @param r
+ *   A pointer to the ring structure
+ * @param is_sc
+ *   Indicates whether multi-consumer path is needed or not
+ * @param n
+ *   The number of elements we will want to enqueue, i.e. how far should the
+ *   head be moved
+ * @param behavior
+ *   RTE_RING_QUEUE_FIXED:    Dequeue a fixed number of items from a ring
+ *   RTE_RING_QUEUE_VARIABLE: Dequeue as many items as possible from ring
+ * @param old_head
+ *   Returns head value as it was before the move, i.e. where dequeue starts
+ * @param new_head
+ *   Returns the current/new head value i.e. where dequeue finishes
+ * @param entries
+ *   Returns the number of entries in the ring BEFORE head was moved
+ * @return
+ *   - Actual number of objects dequeued.
+ *     If behavior == RTE_RING_QUEUE_FIXED, this will be 0 or n only.
+ */
+static __rte_always_inline unsigned int
+__rte_ring_move_cons_head(struct rte_ring *r, int is_sc,
+		unsigned int n, enum rte_ring_queue_behavior behavior,
+		uint32_t *old_head, uint32_t *new_head,
+		uint32_t *entries)
+{
+	unsigned int max = n;
+	int success;
+
+	/* move cons.head atomically */
+	do {
+		/* Restore n as it may change every loop */
+		n = max;
+		*old_head = __atomic_load_n(&r->cons.head,
+					__ATOMIC_ACQUIRE);
+		const uint32_t prod_tail = r->prod.tail;
+		/* The subtraction is done between two unsigned 32bits value
+		 * (the result is always modulo 32 bits even if we have
+		 * cons_head > prod_tail). So 'entries' is always between 0
+		 * and size(ring)-1. */
+		*entries = (prod_tail - *old_head);
+
+		/* Set the actual entries for dequeue */
+		if (n > *entries)
+			n = (behavior == RTE_RING_QUEUE_FIXED) ? 0 : *entries;
+
+		if (unlikely(n == 0))
+			return 0;
+
+		*new_head = *old_head + n;
+		if (is_sc)
+			r->cons.head = *new_head, success = 1;
+		else
+			success = __atomic_compare_exchange_n(&r->cons.head,
+							old_head, *new_head,
+							0, __ATOMIC_ACQUIRE,
+							__ATOMIC_RELAXED);
+	} while (unlikely(success == 0));
+	return n;
+}
+
+#endif /* _RTE_RING_C11_MEM_H_ */
-- 
2.7.4

  parent reply	other threads:[~2017-11-08 10:47 UTC|newest]

Thread overview: 39+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
     [not found] <1510118764-29697-1-git-send-email-hejianet@gmail.com>
2017-11-08  9:54 ` [dpdk-dev] [PATCH v4 0/4] fix race condition in enqueue/dequeue because of cpu reorder Jia He
2017-11-08  9:54   ` [dpdk-dev] [PATCH v4 1/4] eal/arm64: remove the braces {} for dmb() and dsb() Jia He
2017-11-08  9:54   ` [dpdk-dev] [PATCH v4 2/4] ring: guarantee load/load order in enqueue and dequeue Jia He
2017-11-08  9:54   ` [dpdk-dev] [PATCH v4 3/4] ring: introduce new header file to include common functions Jia He
2017-11-08  9:54   ` Jia He [this message]
2017-11-08 12:15   ` [dpdk-dev] [PATCH v4 0/4] fix race condition in enqueue/dequeue because of cpu reorder Bruce Richardson
2017-11-08 15:11     ` Jia He
2017-11-08 16:29       ` Jerin Jacob
2017-11-08 18:36       ` Ananyev, Konstantin
     [not found]       ` <2459a535-920e-9ac5-2f46-1d1dd00e275b@gmail.com>
2017-11-24  9:24         ` Bruce Richardson
2017-11-10  1:51 ` [dpdk-dev] [PATCH v5 0/1] " Jia He
2017-11-10  1:51   ` [dpdk-dev] [PATCH v5 1/1] ring: guarantee load/load order in enqueue and dequeue Jia He
2017-11-10  2:46     ` Jerin Jacob
2017-11-10  3:12       ` Jianbo Liu
2017-11-10  9:59     ` Ananyev, Konstantin
2017-11-10  3:30   ` [dpdk-dev] [PATCH v6] " Jia He
2017-11-10  3:30     ` [dpdk-dev] [PATCH v6] ring: " Jia He
2017-11-12 17:51       ` [dpdk-dev] [dpdk-stable] " Thomas Monjalon
2017-11-10  5:23   ` [dpdk-dev] [PATCH v5 0/3] support c11 memory model barrier in librte_ring Jia He
2017-11-10  5:23     ` [dpdk-dev] [PATCH v5 1/3] eal/arm64: remove the braces {} for dmb() and dsb() Jia He
2017-11-10  5:23     ` [dpdk-dev] [PATCH v5 2/3] ring: introduce new header file to include common functions Jia He
2017-11-10  5:23     ` [dpdk-dev] [PATCH v6 3/3] ring: introduce new header file to support C11 memory model Jia He
2017-11-27  2:00     ` [dpdk-dev] [PATCH V6 0/3] support c11 memory model barrier in librte_ring Jia He
2017-11-27  2:00       ` [dpdk-dev] [PATCH V6 1/3] eal/arm64: remove the braces {} for dmb() and dsb() Jia He
2017-12-03 11:11         ` Jerin Jacob
2017-11-27  2:00       ` [dpdk-dev] [PATCH V6 2/3] ring: introduce new header file to include common functions Jia He
2017-12-03 12:13         ` Jerin Jacob
2017-11-27  2:00       ` [dpdk-dev] [PATCH V6 3/3] ring: introduce new header file to support C11 memory model Jia He
2017-12-03 12:14         ` Jerin Jacob
2017-12-04  1:50       ` [dpdk-dev] [PATCH V7 0/3] support c11 memory model barrier in librte_ring Jia He
2017-12-04  1:50         ` [dpdk-dev] [PATCH V7 1/3] eal/arm64: remove the braces {} for dmb() and dsb() Jia He
2017-12-04  1:50         ` [dpdk-dev] [PATCH V7 2/3] ring: introduce new header file to include common functions Jia He
2018-01-12 17:09           ` Thomas Monjalon
2018-01-16  2:06             ` Jia He
2018-01-16 15:19           ` Olivier Matz
2017-12-04  1:50         ` [dpdk-dev] [PATCH V7 3/3] ring: introduce new header file to support C11 memory model Jia He
2017-12-04  8:05           ` Jianbo Liu
2018-01-12 17:18           ` Thomas Monjalon
2018-01-16 15:18           ` Olivier Matz

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1510134881-22987-5-git-send-email-hejianet@gmail.com \
    --to=hejianet@gmail.com \
    --cc=bruce.richardson@intel.com \
    --cc=dev@dpdk.org \
    --cc=hemant.agrawal@nxp.com \
    --cc=jerin.jacob@caviumnetworks.com \
    --cc=jia.he@hxt-semitech.com \
    --cc=jianbo.liu@arm.com \
    --cc=konstantin.ananyev@intel.com \
    --cc=olivier.matz@6wind.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).