DPDK patches and discussions
 help / color / mirror / Atom feed
From: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
To: dev@dpdk.org
Subject: [dpdk-dev] [v2 11/23] Packet Framework librte_table: Table API
Date: Wed,  4 Jun 2014 19:08:27 +0100	[thread overview]
Message-ID: <1401905319-8882-12-git-send-email-cristian.dumitrescu@intel.com> (raw)
In-Reply-To: <1401905319-8882-1-git-send-email-cristian.dumitrescu@intel.com>

This file defines the operations to be implemented by any Packet Framework table.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
---
 lib/librte_eal/common/include/rte_log.h |    1 +
 lib/librte_table/rte_table.h            |  202 +++++++++++++++++++++++++++++++
 2 files changed, 203 insertions(+), 0 deletions(-)
 create mode 100644 lib/librte_table/rte_table.h
 mode change 100755 => 100644 mk/rte.app.mk

diff --git a/lib/librte_eal/common/include/rte_log.h b/lib/librte_eal/common/include/rte_log.h
index 490dbc9..d119815 100644
--- a/lib/librte_eal/common/include/rte_log.h
+++ b/lib/librte_eal/common/include/rte_log.h
@@ -75,6 +75,7 @@ extern struct rte_logs rte_logs;
 #define RTE_LOGTYPE_METER   0x00000800 /**< Log related to QoS meter. */
 #define RTE_LOGTYPE_SCHED   0x00001000 /**< Log related to QoS port scheduler. */
 #define RTE_LOGTYPE_PORT    0x00002000 /**< Log related to port. */
+#define RTE_LOGTYPE_TABLE   0x00004000 /**< Log related to table. */
 
 /* these log types can be used in an application */
 #define RTE_LOGTYPE_USER1   0x01000000 /**< User-defined log type 1. */
diff --git a/lib/librte_table/rte_table.h b/lib/librte_table/rte_table.h
new file mode 100644
index 0000000..d57bc33
--- /dev/null
+++ b/lib/librte_table/rte_table.h
@@ -0,0 +1,202 @@
+/*-
+ *   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_TABLE_H__
+#define __INCLUDE_RTE_TABLE_H__
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/**
+ * @file
+ * RTE Table
+ *
+ * This tool is part of the Intel DPDK Packet Framework tool suite and provides
+ * a standard interface to implement different types of lookup tables for data
+ * plane processing.
+ *
+ * Virtually any search algorithm that can uniquely associate data to a lookup
+ * key can be fitted under this lookup table abstraction. For the flow table
+ * use-case, the lookup key is an n-tuple of packet fields that uniquely
+ * identifies a traffic flow, while data represents actions and action
+ * meta-data associated with the same traffic flow.
+ *
+ ***/
+
+#include <stdint.h>
+#include <rte_mbuf.h>
+#include <rte_port.h>
+
+/**
+ * Lookup table create
+ *
+ * @param params
+ *   Parameters for lookup table creation. The underlying data structure is
+ *   different for each lookup table type.
+ * @param socket_id
+ *   CPU socket ID (e.g. for memory allocation purpose)
+ * @param entry_size
+ *   Data size of each lookup table entry (measured in bytes)
+ * @return
+ *   Handle to lookup table instance
+ */
+typedef void* (*rte_table_op_create)(void *params, int socket_id,
+	uint32_t entry_size);
+
+/**
+ * Lookup table free
+ *
+ * @param table
+ *   Handle to lookup table instance
+ * @return
+ *   0 on success, error code otherwise
+ */
+typedef int (*rte_table_op_free)(void *table);
+
+/**
+ * Lookup table entry add
+ *
+ * @param table
+ *   Handle to lookup table instance
+ * @param key
+ *   Lookup key
+ * @param entry
+ *   Data to be associated with the current key. This parameter has to point to
+ *   a valid memory buffer where the first entry_size bytes (table create
+ *   parameter) are populated with the data.
+ * @param key_found
+ *   After successful invocation, *key_found is set to a value different than 0
+ *   if the current key is already present in the table and to 0 if not. This
+ *   pointer has to be set to a valid memory location before the table entry add
+ *   function is called.
+ * @param entry_ptr
+ *   After successful invocation, *entry_ptr stores the handle to the table
+ *   entry containing the data associated with the current key. This handle can
+ *   be used to perform further read-write accesses to this entry. This handle
+ *   is valid until the key is deleted from the table or the same key is
+ *   re-added to the table, typically to associate it with different data. This
+ *   pointer has to be set to a valid memory location before the function is
+ *   called.
+ * @return
+ *   0 on success, error code otherwise
+ */
+typedef int (*rte_table_op_entry_add)(
+	void *table,
+	void *key,
+	void *entry,
+	int *key_found,
+	void **entry_ptr);
+
+/**
+ * Lookup table entry delete
+ *
+ * @param table
+ *   Handle to lookup table instance
+ * @param key
+ *   Lookup key
+ * @param key_found
+ *   After successful invocation, *key_found is set to a value different than 0
+ *   if the current key was present in the table before the delete operation
+ *   was performed and to 0 if not. This pointer has to be set to a valid
+ *   memory location before the table entry delete function is called.
+ * @param entry
+ *   After successful invocation, if the key is found in the table (*key found
+ *   is different than 0 after function call is completed) and entry points to
+ *   a valid buffer (entry is set to a value different than NULL before the
+ *   function is called), then the first entry_size bytes (table create
+ *   parameter) in *entry store a copy of table entry that contained the data
+ *   associated with the current key before the key was deleted.
+ * @return
+ *   0 on success, error code otherwise
+ */
+typedef int (*rte_table_op_entry_delete)(
+	void *table,
+	void *key,
+	int *key_found,
+	void *entry);
+
+/**
+ * Lookup table lookup
+ *
+ * @param table
+ *   Handle to lookup table instance
+ * @param pkts
+ *   Burst of input packets specified as array of up to 64 pointers to struct
+ *   rte_mbuf
+ * @param pkts_mask
+ *   64-bit bitmask specifying which packets in the input burst are valid. When
+ *   pkts_mask bit n is set, then element n of pkts array is pointing to a
+ *   valid packet. Otherwise, element n of pkts array does not point to a valid
+ *   packet, therefore it will not be accessed.
+ * @param lookup_hit_mask
+ *   Once the table lookup operation is completed, this 64-bit bitmask
+ *   specifies which of the valid packets in the input burst resulted in lookup
+ *   hit. For each valid input packet (pkts_mask bit n is set), the following
+ *   are true on lookup hit: lookup_hit_mask bit n is set, element n of entries
+ *   array is valid and it points to the lookup table entry that was hit. For
+ *   each valid input packet (pkts_mask bit n is set), the following are true
+ *   on lookup miss: lookup_hit_mask bit n is not set and element n of entries
+ *   array is not valid.
+ * @param entries
+ *   Once the table lookup operation is completed, this array provides the
+ *   lookup table entries that were hit, as described above. It is required
+ *   that this array is always pre-allocated by the caller of this function
+ *   with exactly 64 elements. The implementation is allowed to speculatively
+ *   modify the elements of this array, so elements marked as invalid in
+ *   lookup_hit_mask once the table lookup operation is completed might have
+ *   been modified by this function.
+ * @return
+ *   0 on success, error code otherwise
+ */
+typedef int (*rte_table_op_lookup)(
+	void *table,
+	struct rte_mbuf **pkts,
+	uint64_t pkts_mask,
+	uint64_t *lookup_hit_mask,
+	void **entries);
+
+/** Lookup table interface defining the lookup table operation */
+struct rte_table_ops {
+	rte_table_op_create f_create;       /**< Create */
+	rte_table_op_free f_free;           /**< Free */
+	rte_table_op_entry_add f_add;       /**< Entry add */
+	rte_table_op_entry_delete f_delete; /**< Entry delete */
+	rte_table_op_lookup f_lookup;       /**< Lookup */
+};
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif
diff --git a/mk/rte.app.mk b/mk/rte.app.mk
old mode 100755
new mode 100644
-- 
1.7.7.6

  parent reply	other threads:[~2014-06-04 18:08 UTC|newest]

Thread overview: 36+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-06-04 18:08 [dpdk-dev] [v2 00/23] Packet Framework Cristian Dumitrescu
2014-06-04 18:08 ` [dpdk-dev] [v2 01/23] librte_lpm: rule_is_present Cristian Dumitrescu
2014-06-04 18:08 ` [dpdk-dev] [v2 02/23] mbuf: meta-data Cristian Dumitrescu
2014-06-04 18:08 ` [dpdk-dev] [v2 03/23] Packet Framework librte_port: Port API Cristian Dumitrescu
2014-06-04 18:08 ` [dpdk-dev] [v2 04/23] Packet Framework librte_port: ethdev ports Cristian Dumitrescu
2014-06-04 18:08 ` [dpdk-dev] [v2 05/23] Packet Framework librte_port: ring ports Cristian Dumitrescu
2014-06-04 18:08 ` [dpdk-dev] [v2 06/23] Packet Framework librte_port: IPv4 frag port Cristian Dumitrescu
2014-06-04 18:08 ` [dpdk-dev] [v2 07/23] Packet Framework librte_port: IPv4 reassembly Cristian Dumitrescu
2014-06-04 18:08 ` [dpdk-dev] [v2 08/23] Packet Framework librte_port: hierarchical scheduler port Cristian Dumitrescu
2014-06-04 18:08 ` [dpdk-dev] [v2 09/23] Packet Framework librte_port: Source/Sink ports Cristian Dumitrescu
2014-06-04 18:08 ` [dpdk-dev] [v2 10/23] Packet Framework librte_port: Build infrastructure Cristian Dumitrescu
2014-06-04 18:08 ` Cristian Dumitrescu [this message]
2014-06-04 18:08 ` [dpdk-dev] [v2 12/23] Packet Framework librte_table: LPM IPv4 table Cristian Dumitrescu
2014-06-04 18:08 ` [dpdk-dev] [v2 13/23] Packet Framework librte_table: LPM IPv6 table Cristian Dumitrescu
2014-06-04 18:08 ` [dpdk-dev] [v2 14/23] Packet Framework librte_table: ACL table Cristian Dumitrescu
2014-06-04 18:08 ` [dpdk-dev] [v2 15/23] Packet Framework librte_table: Hash tables Cristian Dumitrescu
2014-06-04 18:08 ` [dpdk-dev] [v2 16/23] Packet Framework librte_table: array table Cristian Dumitrescu
2014-06-04 18:08 ` [dpdk-dev] [v2 17/23] Packet Framework librte_table: Stub table Cristian Dumitrescu
2014-06-04 18:08 ` [dpdk-dev] [v2 18/23] Packet Framework librte_table: Build infrastructure Cristian Dumitrescu
2014-06-04 18:08 ` [dpdk-dev] [v2 19/23] Packet Framework librte_pipeline: Pipeline Cristian Dumitrescu
2014-06-04 18:08 ` [dpdk-dev] [v2 20/23] librte_cfgfile: interpret config files Cristian Dumitrescu
2014-10-16 16:46   ` Thomas Monjalon
2014-10-17 18:16     ` Dumitrescu, Cristian
2014-10-17 18:50       ` Thomas Monjalon
2014-06-04 18:08 ` [dpdk-dev] [v2 21/23] Packet Framework performance application Cristian Dumitrescu
2014-06-04 18:08 ` [dpdk-dev] [v2 22/23] Packet Framework IPv4 pipeline sample app Cristian Dumitrescu
2014-06-09  9:11   ` Olivier MATZ
2014-06-09 10:49     ` Dumitrescu, Cristian
2014-06-09 12:13       ` Olivier MATZ
2014-06-09 13:25         ` Dumitrescu, Cristian
2014-06-09 15:51           ` Olivier MATZ
2014-06-04 18:08 ` [dpdk-dev] [v2 23/23] Packet Framework unit tests Cristian Dumitrescu
2014-06-05 11:01 ` [dpdk-dev] [v2 00/23] Packet Framework De Lara Guarch, Pablo
2014-06-05 11:43 ` Cao, Waterman
2014-06-05 14:40 ` Ivan Boule
2014-06-17  1:27   ` Thomas Monjalon

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=1401905319-8882-12-git-send-email-cristian.dumitrescu@intel.com \
    --to=cristian.dumitrescu@intel.com \
    --cc=dev@dpdk.org \
    /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).