d/numam-dpdk
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity
numam-dpdk/examples/l3fwd/l3fwd_em_sse.h

99 lines
3.1 KiB
C
Raw Normal View History

examples/l3fwd: modularize The main problem with l3fwd is that it is too monolithic with everything being in one file, and the various options all controlled by compile time flags. This means that it's hard to read and understand, and when making any changes, you need to go to a lot of work to try and ensure you cover all the code paths, since a compile of the app will not touch large parts of the l3fwd codebase. Following changes were done to fix the issues mentioned above - Split out the various lpm and hash specific functionality into separate files, so that l3fwd code has one file for common code e.g. args processing, mempool creation, and then individual files for the various forwarding approaches. Following are new file lists main.c (Common code for args processing, memppol creation, etc) l3fwd_em.c (Hash/Exact match aka 'EM' functionality) l3fwd_em_sse.h (SSE4_1 buffer optimizated 'EM' code) l3fwd_lpm.c (Longest Prefix Match aka 'LPM' functionality) l3fwd_lpm_sse.h (SSE4_1 buffer optimizated 'LPM' code) l3fwd.h (Common include for 'EM' and 'LPM') - The choosing of the lpm/hash path should be done at runtime, not compile time, via a command-line argument. This will ensure that both code paths get compiled in a single go Following examples show runtime options provided Select 'LPM' or 'EM' based on run time selection f.e. > l3fwd -c 0x1 -n 1 -- -p 0x1 -E ... (EM) > l3fwd -c 0x1 -n 1 -- -p 0x1 -L ... (LPM) Options "E" and "L" are mutualy-exclusive. If none selected, "L" is default. Signed-off-by: Ravi Kerur <rkerur@gmail.com> Signed-off-by: Piotr Azarewicz <piotrx.t.azarewicz@intel.com> Tested-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
2016-02-25 11:24:24 +01:00
/*-
* BSD LICENSE
*
* Copyright(c) 2010-2016 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 __L3FWD_EM_SSE_H__
#define __L3FWD_EM_SSE_H__
examples/l3fwd: rework exact-match Current implementation of Exact-Match uses different execution path than for LPM. Unifying them allows to reuse big part of LPM code and sightly increase performance of Exact-Match. Main changes: ------------- * Packet classification stage is separated from the rest of path for both LPM and EM. * Packet processing, modifying and transmit part is the same for LPM and EM and mostly based on the current LPM implementation. * Shared code is moved to the common file "l3fwd_sse.h". * While sequential packet classification in EM path, seems to be faster than using multi hash lookup, used before, it is used by default. Old implementation is moved to the file l3fwd_em_hlm_sse.h and can be enabled with HASH_LOOKUP_MULTI global define in compilation time. Signed-off-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
2016-02-29 11:33:07 +01:00
#include "l3fwd_sse.h"
examples/l3fwd: modularize The main problem with l3fwd is that it is too monolithic with everything being in one file, and the various options all controlled by compile time flags. This means that it's hard to read and understand, and when making any changes, you need to go to a lot of work to try and ensure you cover all the code paths, since a compile of the app will not touch large parts of the l3fwd codebase. Following changes were done to fix the issues mentioned above - Split out the various lpm and hash specific functionality into separate files, so that l3fwd code has one file for common code e.g. args processing, mempool creation, and then individual files for the various forwarding approaches. Following are new file lists main.c (Common code for args processing, memppol creation, etc) l3fwd_em.c (Hash/Exact match aka 'EM' functionality) l3fwd_em_sse.h (SSE4_1 buffer optimizated 'EM' code) l3fwd_lpm.c (Longest Prefix Match aka 'LPM' functionality) l3fwd_lpm_sse.h (SSE4_1 buffer optimizated 'LPM' code) l3fwd.h (Common include for 'EM' and 'LPM') - The choosing of the lpm/hash path should be done at runtime, not compile time, via a command-line argument. This will ensure that both code paths get compiled in a single go Following examples show runtime options provided Select 'LPM' or 'EM' based on run time selection f.e. > l3fwd -c 0x1 -n 1 -- -p 0x1 -E ... (EM) > l3fwd -c 0x1 -n 1 -- -p 0x1 -L ... (LPM) Options "E" and "L" are mutualy-exclusive. If none selected, "L" is default. Signed-off-by: Ravi Kerur <rkerur@gmail.com> Signed-off-by: Piotr Azarewicz <piotrx.t.azarewicz@intel.com> Tested-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
2016-02-25 11:24:24 +01:00
examples/l3fwd: rework exact-match Current implementation of Exact-Match uses different execution path than for LPM. Unifying them allows to reuse big part of LPM code and sightly increase performance of Exact-Match. Main changes: ------------- * Packet classification stage is separated from the rest of path for both LPM and EM. * Packet processing, modifying and transmit part is the same for LPM and EM and mostly based on the current LPM implementation. * Shared code is moved to the common file "l3fwd_sse.h". * While sequential packet classification in EM path, seems to be faster than using multi hash lookup, used before, it is used by default. Old implementation is moved to the file l3fwd_em_hlm_sse.h and can be enabled with HASH_LOOKUP_MULTI global define in compilation time. Signed-off-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
2016-02-29 11:33:07 +01:00
static inline __attribute__((always_inline)) uint16_t
em_get_dst_port(const struct lcore_conf *qconf, struct rte_mbuf *pkt,
uint8_t portid)
examples/l3fwd: modularize The main problem with l3fwd is that it is too monolithic with everything being in one file, and the various options all controlled by compile time flags. This means that it's hard to read and understand, and when making any changes, you need to go to a lot of work to try and ensure you cover all the code paths, since a compile of the app will not touch large parts of the l3fwd codebase. Following changes were done to fix the issues mentioned above - Split out the various lpm and hash specific functionality into separate files, so that l3fwd code has one file for common code e.g. args processing, mempool creation, and then individual files for the various forwarding approaches. Following are new file lists main.c (Common code for args processing, memppol creation, etc) l3fwd_em.c (Hash/Exact match aka 'EM' functionality) l3fwd_em_sse.h (SSE4_1 buffer optimizated 'EM' code) l3fwd_lpm.c (Longest Prefix Match aka 'LPM' functionality) l3fwd_lpm_sse.h (SSE4_1 buffer optimizated 'LPM' code) l3fwd.h (Common include for 'EM' and 'LPM') - The choosing of the lpm/hash path should be done at runtime, not compile time, via a command-line argument. This will ensure that both code paths get compiled in a single go Following examples show runtime options provided Select 'LPM' or 'EM' based on run time selection f.e. > l3fwd -c 0x1 -n 1 -- -p 0x1 -E ... (EM) > l3fwd -c 0x1 -n 1 -- -p 0x1 -L ... (LPM) Options "E" and "L" are mutualy-exclusive. If none selected, "L" is default. Signed-off-by: Ravi Kerur <rkerur@gmail.com> Signed-off-by: Piotr Azarewicz <piotrx.t.azarewicz@intel.com> Tested-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
2016-02-25 11:24:24 +01:00
{
examples/l3fwd: rework exact-match Current implementation of Exact-Match uses different execution path than for LPM. Unifying them allows to reuse big part of LPM code and sightly increase performance of Exact-Match. Main changes: ------------- * Packet classification stage is separated from the rest of path for both LPM and EM. * Packet processing, modifying and transmit part is the same for LPM and EM and mostly based on the current LPM implementation. * Shared code is moved to the common file "l3fwd_sse.h". * While sequential packet classification in EM path, seems to be faster than using multi hash lookup, used before, it is used by default. Old implementation is moved to the file l3fwd_em_hlm_sse.h and can be enabled with HASH_LOOKUP_MULTI global define in compilation time. Signed-off-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
2016-02-29 11:33:07 +01:00
uint8_t next_hop;
struct ipv4_hdr *ipv4_hdr;
struct ipv6_hdr *ipv6_hdr;
examples/l3fwd: modularize The main problem with l3fwd is that it is too monolithic with everything being in one file, and the various options all controlled by compile time flags. This means that it's hard to read and understand, and when making any changes, you need to go to a lot of work to try and ensure you cover all the code paths, since a compile of the app will not touch large parts of the l3fwd codebase. Following changes were done to fix the issues mentioned above - Split out the various lpm and hash specific functionality into separate files, so that l3fwd code has one file for common code e.g. args processing, mempool creation, and then individual files for the various forwarding approaches. Following are new file lists main.c (Common code for args processing, memppol creation, etc) l3fwd_em.c (Hash/Exact match aka 'EM' functionality) l3fwd_em_sse.h (SSE4_1 buffer optimizated 'EM' code) l3fwd_lpm.c (Longest Prefix Match aka 'LPM' functionality) l3fwd_lpm_sse.h (SSE4_1 buffer optimizated 'LPM' code) l3fwd.h (Common include for 'EM' and 'LPM') - The choosing of the lpm/hash path should be done at runtime, not compile time, via a command-line argument. This will ensure that both code paths get compiled in a single go Following examples show runtime options provided Select 'LPM' or 'EM' based on run time selection f.e. > l3fwd -c 0x1 -n 1 -- -p 0x1 -E ... (EM) > l3fwd -c 0x1 -n 1 -- -p 0x1 -L ... (LPM) Options "E" and "L" are mutualy-exclusive. If none selected, "L" is default. Signed-off-by: Ravi Kerur <rkerur@gmail.com> Signed-off-by: Piotr Azarewicz <piotrx.t.azarewicz@intel.com> Tested-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
2016-02-25 11:24:24 +01:00
examples/l3fwd: rework exact-match Current implementation of Exact-Match uses different execution path than for LPM. Unifying them allows to reuse big part of LPM code and sightly increase performance of Exact-Match. Main changes: ------------- * Packet classification stage is separated from the rest of path for both LPM and EM. * Packet processing, modifying and transmit part is the same for LPM and EM and mostly based on the current LPM implementation. * Shared code is moved to the common file "l3fwd_sse.h". * While sequential packet classification in EM path, seems to be faster than using multi hash lookup, used before, it is used by default. Old implementation is moved to the file l3fwd_em_hlm_sse.h and can be enabled with HASH_LOOKUP_MULTI global define in compilation time. Signed-off-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
2016-02-29 11:33:07 +01:00
if (RTE_ETH_IS_IPV4_HDR(pkt->packet_type)) {
examples/l3fwd: modularize The main problem with l3fwd is that it is too monolithic with everything being in one file, and the various options all controlled by compile time flags. This means that it's hard to read and understand, and when making any changes, you need to go to a lot of work to try and ensure you cover all the code paths, since a compile of the app will not touch large parts of the l3fwd codebase. Following changes were done to fix the issues mentioned above - Split out the various lpm and hash specific functionality into separate files, so that l3fwd code has one file for common code e.g. args processing, mempool creation, and then individual files for the various forwarding approaches. Following are new file lists main.c (Common code for args processing, memppol creation, etc) l3fwd_em.c (Hash/Exact match aka 'EM' functionality) l3fwd_em_sse.h (SSE4_1 buffer optimizated 'EM' code) l3fwd_lpm.c (Longest Prefix Match aka 'LPM' functionality) l3fwd_lpm_sse.h (SSE4_1 buffer optimizated 'LPM' code) l3fwd.h (Common include for 'EM' and 'LPM') - The choosing of the lpm/hash path should be done at runtime, not compile time, via a command-line argument. This will ensure that both code paths get compiled in a single go Following examples show runtime options provided Select 'LPM' or 'EM' based on run time selection f.e. > l3fwd -c 0x1 -n 1 -- -p 0x1 -E ... (EM) > l3fwd -c 0x1 -n 1 -- -p 0x1 -L ... (LPM) Options "E" and "L" are mutualy-exclusive. If none selected, "L" is default. Signed-off-by: Ravi Kerur <rkerur@gmail.com> Signed-off-by: Piotr Azarewicz <piotrx.t.azarewicz@intel.com> Tested-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
2016-02-25 11:24:24 +01:00
examples/l3fwd: rework exact-match Current implementation of Exact-Match uses different execution path than for LPM. Unifying them allows to reuse big part of LPM code and sightly increase performance of Exact-Match. Main changes: ------------- * Packet classification stage is separated from the rest of path for both LPM and EM. * Packet processing, modifying and transmit part is the same for LPM and EM and mostly based on the current LPM implementation. * Shared code is moved to the common file "l3fwd_sse.h". * While sequential packet classification in EM path, seems to be faster than using multi hash lookup, used before, it is used by default. Old implementation is moved to the file l3fwd_em_hlm_sse.h and can be enabled with HASH_LOOKUP_MULTI global define in compilation time. Signed-off-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
2016-02-29 11:33:07 +01:00
/* Handle IPv4 headers.*/
ipv4_hdr = rte_pktmbuf_mtod_offset(pkt, struct ipv4_hdr *,
sizeof(struct ether_hdr));
examples/l3fwd: modularize The main problem with l3fwd is that it is too monolithic with everything being in one file, and the various options all controlled by compile time flags. This means that it's hard to read and understand, and when making any changes, you need to go to a lot of work to try and ensure you cover all the code paths, since a compile of the app will not touch large parts of the l3fwd codebase. Following changes were done to fix the issues mentioned above - Split out the various lpm and hash specific functionality into separate files, so that l3fwd code has one file for common code e.g. args processing, mempool creation, and then individual files for the various forwarding approaches. Following are new file lists main.c (Common code for args processing, memppol creation, etc) l3fwd_em.c (Hash/Exact match aka 'EM' functionality) l3fwd_em_sse.h (SSE4_1 buffer optimizated 'EM' code) l3fwd_lpm.c (Longest Prefix Match aka 'LPM' functionality) l3fwd_lpm_sse.h (SSE4_1 buffer optimizated 'LPM' code) l3fwd.h (Common include for 'EM' and 'LPM') - The choosing of the lpm/hash path should be done at runtime, not compile time, via a command-line argument. This will ensure that both code paths get compiled in a single go Following examples show runtime options provided Select 'LPM' or 'EM' based on run time selection f.e. > l3fwd -c 0x1 -n 1 -- -p 0x1 -E ... (EM) > l3fwd -c 0x1 -n 1 -- -p 0x1 -L ... (LPM) Options "E" and "L" are mutualy-exclusive. If none selected, "L" is default. Signed-off-by: Ravi Kerur <rkerur@gmail.com> Signed-off-by: Piotr Azarewicz <piotrx.t.azarewicz@intel.com> Tested-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
2016-02-25 11:24:24 +01:00
examples/l3fwd: rework exact-match Current implementation of Exact-Match uses different execution path than for LPM. Unifying them allows to reuse big part of LPM code and sightly increase performance of Exact-Match. Main changes: ------------- * Packet classification stage is separated from the rest of path for both LPM and EM. * Packet processing, modifying and transmit part is the same for LPM and EM and mostly based on the current LPM implementation. * Shared code is moved to the common file "l3fwd_sse.h". * While sequential packet classification in EM path, seems to be faster than using multi hash lookup, used before, it is used by default. Old implementation is moved to the file l3fwd_em_hlm_sse.h and can be enabled with HASH_LOOKUP_MULTI global define in compilation time. Signed-off-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
2016-02-29 11:33:07 +01:00
next_hop = em_get_ipv4_dst_port(ipv4_hdr, portid,
qconf->ipv4_lookup_struct);
examples/l3fwd: modularize The main problem with l3fwd is that it is too monolithic with everything being in one file, and the various options all controlled by compile time flags. This means that it's hard to read and understand, and when making any changes, you need to go to a lot of work to try and ensure you cover all the code paths, since a compile of the app will not touch large parts of the l3fwd codebase. Following changes were done to fix the issues mentioned above - Split out the various lpm and hash specific functionality into separate files, so that l3fwd code has one file for common code e.g. args processing, mempool creation, and then individual files for the various forwarding approaches. Following are new file lists main.c (Common code for args processing, memppol creation, etc) l3fwd_em.c (Hash/Exact match aka 'EM' functionality) l3fwd_em_sse.h (SSE4_1 buffer optimizated 'EM' code) l3fwd_lpm.c (Longest Prefix Match aka 'LPM' functionality) l3fwd_lpm_sse.h (SSE4_1 buffer optimizated 'LPM' code) l3fwd.h (Common include for 'EM' and 'LPM') - The choosing of the lpm/hash path should be done at runtime, not compile time, via a command-line argument. This will ensure that both code paths get compiled in a single go Following examples show runtime options provided Select 'LPM' or 'EM' based on run time selection f.e. > l3fwd -c 0x1 -n 1 -- -p 0x1 -E ... (EM) > l3fwd -c 0x1 -n 1 -- -p 0x1 -L ... (LPM) Options "E" and "L" are mutualy-exclusive. If none selected, "L" is default. Signed-off-by: Ravi Kerur <rkerur@gmail.com> Signed-off-by: Piotr Azarewicz <piotrx.t.azarewicz@intel.com> Tested-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
2016-02-25 11:24:24 +01:00
examples/l3fwd: rework exact-match Current implementation of Exact-Match uses different execution path than for LPM. Unifying them allows to reuse big part of LPM code and sightly increase performance of Exact-Match. Main changes: ------------- * Packet classification stage is separated from the rest of path for both LPM and EM. * Packet processing, modifying and transmit part is the same for LPM and EM and mostly based on the current LPM implementation. * Shared code is moved to the common file "l3fwd_sse.h". * While sequential packet classification in EM path, seems to be faster than using multi hash lookup, used before, it is used by default. Old implementation is moved to the file l3fwd_em_hlm_sse.h and can be enabled with HASH_LOOKUP_MULTI global define in compilation time. Signed-off-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
2016-02-29 11:33:07 +01:00
if (next_hop >= RTE_MAX_ETHPORTS ||
(enabled_port_mask & 1 << next_hop) == 0)
next_hop = portid;
examples/l3fwd: modularize The main problem with l3fwd is that it is too monolithic with everything being in one file, and the various options all controlled by compile time flags. This means that it's hard to read and understand, and when making any changes, you need to go to a lot of work to try and ensure you cover all the code paths, since a compile of the app will not touch large parts of the l3fwd codebase. Following changes were done to fix the issues mentioned above - Split out the various lpm and hash specific functionality into separate files, so that l3fwd code has one file for common code e.g. args processing, mempool creation, and then individual files for the various forwarding approaches. Following are new file lists main.c (Common code for args processing, memppol creation, etc) l3fwd_em.c (Hash/Exact match aka 'EM' functionality) l3fwd_em_sse.h (SSE4_1 buffer optimizated 'EM' code) l3fwd_lpm.c (Longest Prefix Match aka 'LPM' functionality) l3fwd_lpm_sse.h (SSE4_1 buffer optimizated 'LPM' code) l3fwd.h (Common include for 'EM' and 'LPM') - The choosing of the lpm/hash path should be done at runtime, not compile time, via a command-line argument. This will ensure that both code paths get compiled in a single go Following examples show runtime options provided Select 'LPM' or 'EM' based on run time selection f.e. > l3fwd -c 0x1 -n 1 -- -p 0x1 -E ... (EM) > l3fwd -c 0x1 -n 1 -- -p 0x1 -L ... (LPM) Options "E" and "L" are mutualy-exclusive. If none selected, "L" is default. Signed-off-by: Ravi Kerur <rkerur@gmail.com> Signed-off-by: Piotr Azarewicz <piotrx.t.azarewicz@intel.com> Tested-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
2016-02-25 11:24:24 +01:00
examples/l3fwd: rework exact-match Current implementation of Exact-Match uses different execution path than for LPM. Unifying them allows to reuse big part of LPM code and sightly increase performance of Exact-Match. Main changes: ------------- * Packet classification stage is separated from the rest of path for both LPM and EM. * Packet processing, modifying and transmit part is the same for LPM and EM and mostly based on the current LPM implementation. * Shared code is moved to the common file "l3fwd_sse.h". * While sequential packet classification in EM path, seems to be faster than using multi hash lookup, used before, it is used by default. Old implementation is moved to the file l3fwd_em_hlm_sse.h and can be enabled with HASH_LOOKUP_MULTI global define in compilation time. Signed-off-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
2016-02-29 11:33:07 +01:00
return next_hop;
examples/l3fwd: modularize The main problem with l3fwd is that it is too monolithic with everything being in one file, and the various options all controlled by compile time flags. This means that it's hard to read and understand, and when making any changes, you need to go to a lot of work to try and ensure you cover all the code paths, since a compile of the app will not touch large parts of the l3fwd codebase. Following changes were done to fix the issues mentioned above - Split out the various lpm and hash specific functionality into separate files, so that l3fwd code has one file for common code e.g. args processing, mempool creation, and then individual files for the various forwarding approaches. Following are new file lists main.c (Common code for args processing, memppol creation, etc) l3fwd_em.c (Hash/Exact match aka 'EM' functionality) l3fwd_em_sse.h (SSE4_1 buffer optimizated 'EM' code) l3fwd_lpm.c (Longest Prefix Match aka 'LPM' functionality) l3fwd_lpm_sse.h (SSE4_1 buffer optimizated 'LPM' code) l3fwd.h (Common include for 'EM' and 'LPM') - The choosing of the lpm/hash path should be done at runtime, not compile time, via a command-line argument. This will ensure that both code paths get compiled in a single go Following examples show runtime options provided Select 'LPM' or 'EM' based on run time selection f.e. > l3fwd -c 0x1 -n 1 -- -p 0x1 -E ... (EM) > l3fwd -c 0x1 -n 1 -- -p 0x1 -L ... (LPM) Options "E" and "L" are mutualy-exclusive. If none selected, "L" is default. Signed-off-by: Ravi Kerur <rkerur@gmail.com> Signed-off-by: Piotr Azarewicz <piotrx.t.azarewicz@intel.com> Tested-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
2016-02-25 11:24:24 +01:00
examples/l3fwd: rework exact-match Current implementation of Exact-Match uses different execution path than for LPM. Unifying them allows to reuse big part of LPM code and sightly increase performance of Exact-Match. Main changes: ------------- * Packet classification stage is separated from the rest of path for both LPM and EM. * Packet processing, modifying and transmit part is the same for LPM and EM and mostly based on the current LPM implementation. * Shared code is moved to the common file "l3fwd_sse.h". * While sequential packet classification in EM path, seems to be faster than using multi hash lookup, used before, it is used by default. Old implementation is moved to the file l3fwd_em_hlm_sse.h and can be enabled with HASH_LOOKUP_MULTI global define in compilation time. Signed-off-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
2016-02-29 11:33:07 +01:00
} else if (RTE_ETH_IS_IPV6_HDR(pkt->packet_type)) {
examples/l3fwd: modularize The main problem with l3fwd is that it is too monolithic with everything being in one file, and the various options all controlled by compile time flags. This means that it's hard to read and understand, and when making any changes, you need to go to a lot of work to try and ensure you cover all the code paths, since a compile of the app will not touch large parts of the l3fwd codebase. Following changes were done to fix the issues mentioned above - Split out the various lpm and hash specific functionality into separate files, so that l3fwd code has one file for common code e.g. args processing, mempool creation, and then individual files for the various forwarding approaches. Following are new file lists main.c (Common code for args processing, memppol creation, etc) l3fwd_em.c (Hash/Exact match aka 'EM' functionality) l3fwd_em_sse.h (SSE4_1 buffer optimizated 'EM' code) l3fwd_lpm.c (Longest Prefix Match aka 'LPM' functionality) l3fwd_lpm_sse.h (SSE4_1 buffer optimizated 'LPM' code) l3fwd.h (Common include for 'EM' and 'LPM') - The choosing of the lpm/hash path should be done at runtime, not compile time, via a command-line argument. This will ensure that both code paths get compiled in a single go Following examples show runtime options provided Select 'LPM' or 'EM' based on run time selection f.e. > l3fwd -c 0x1 -n 1 -- -p 0x1 -E ... (EM) > l3fwd -c 0x1 -n 1 -- -p 0x1 -L ... (LPM) Options "E" and "L" are mutualy-exclusive. If none selected, "L" is default. Signed-off-by: Ravi Kerur <rkerur@gmail.com> Signed-off-by: Piotr Azarewicz <piotrx.t.azarewicz@intel.com> Tested-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
2016-02-25 11:24:24 +01:00
examples/l3fwd: rework exact-match Current implementation of Exact-Match uses different execution path than for LPM. Unifying them allows to reuse big part of LPM code and sightly increase performance of Exact-Match. Main changes: ------------- * Packet classification stage is separated from the rest of path for both LPM and EM. * Packet processing, modifying and transmit part is the same for LPM and EM and mostly based on the current LPM implementation. * Shared code is moved to the common file "l3fwd_sse.h". * While sequential packet classification in EM path, seems to be faster than using multi hash lookup, used before, it is used by default. Old implementation is moved to the file l3fwd_em_hlm_sse.h and can be enabled with HASH_LOOKUP_MULTI global define in compilation time. Signed-off-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
2016-02-29 11:33:07 +01:00
/* Handle IPv6 headers.*/
ipv6_hdr = rte_pktmbuf_mtod_offset(pkt, struct ipv6_hdr *,
sizeof(struct ether_hdr));
examples/l3fwd: modularize The main problem with l3fwd is that it is too monolithic with everything being in one file, and the various options all controlled by compile time flags. This means that it's hard to read and understand, and when making any changes, you need to go to a lot of work to try and ensure you cover all the code paths, since a compile of the app will not touch large parts of the l3fwd codebase. Following changes were done to fix the issues mentioned above - Split out the various lpm and hash specific functionality into separate files, so that l3fwd code has one file for common code e.g. args processing, mempool creation, and then individual files for the various forwarding approaches. Following are new file lists main.c (Common code for args processing, memppol creation, etc) l3fwd_em.c (Hash/Exact match aka 'EM' functionality) l3fwd_em_sse.h (SSE4_1 buffer optimizated 'EM' code) l3fwd_lpm.c (Longest Prefix Match aka 'LPM' functionality) l3fwd_lpm_sse.h (SSE4_1 buffer optimizated 'LPM' code) l3fwd.h (Common include for 'EM' and 'LPM') - The choosing of the lpm/hash path should be done at runtime, not compile time, via a command-line argument. This will ensure that both code paths get compiled in a single go Following examples show runtime options provided Select 'LPM' or 'EM' based on run time selection f.e. > l3fwd -c 0x1 -n 1 -- -p 0x1 -E ... (EM) > l3fwd -c 0x1 -n 1 -- -p 0x1 -L ... (LPM) Options "E" and "L" are mutualy-exclusive. If none selected, "L" is default. Signed-off-by: Ravi Kerur <rkerur@gmail.com> Signed-off-by: Piotr Azarewicz <piotrx.t.azarewicz@intel.com> Tested-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
2016-02-25 11:24:24 +01:00
examples/l3fwd: rework exact-match Current implementation of Exact-Match uses different execution path than for LPM. Unifying them allows to reuse big part of LPM code and sightly increase performance of Exact-Match. Main changes: ------------- * Packet classification stage is separated from the rest of path for both LPM and EM. * Packet processing, modifying and transmit part is the same for LPM and EM and mostly based on the current LPM implementation. * Shared code is moved to the common file "l3fwd_sse.h". * While sequential packet classification in EM path, seems to be faster than using multi hash lookup, used before, it is used by default. Old implementation is moved to the file l3fwd_em_hlm_sse.h and can be enabled with HASH_LOOKUP_MULTI global define in compilation time. Signed-off-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
2016-02-29 11:33:07 +01:00
next_hop = em_get_ipv6_dst_port(ipv6_hdr, portid,
qconf->ipv6_lookup_struct);
examples/l3fwd: modularize The main problem with l3fwd is that it is too monolithic with everything being in one file, and the various options all controlled by compile time flags. This means that it's hard to read and understand, and when making any changes, you need to go to a lot of work to try and ensure you cover all the code paths, since a compile of the app will not touch large parts of the l3fwd codebase. Following changes were done to fix the issues mentioned above - Split out the various lpm and hash specific functionality into separate files, so that l3fwd code has one file for common code e.g. args processing, mempool creation, and then individual files for the various forwarding approaches. Following are new file lists main.c (Common code for args processing, memppol creation, etc) l3fwd_em.c (Hash/Exact match aka 'EM' functionality) l3fwd_em_sse.h (SSE4_1 buffer optimizated 'EM' code) l3fwd_lpm.c (Longest Prefix Match aka 'LPM' functionality) l3fwd_lpm_sse.h (SSE4_1 buffer optimizated 'LPM' code) l3fwd.h (Common include for 'EM' and 'LPM') - The choosing of the lpm/hash path should be done at runtime, not compile time, via a command-line argument. This will ensure that both code paths get compiled in a single go Following examples show runtime options provided Select 'LPM' or 'EM' based on run time selection f.e. > l3fwd -c 0x1 -n 1 -- -p 0x1 -E ... (EM) > l3fwd -c 0x1 -n 1 -- -p 0x1 -L ... (LPM) Options "E" and "L" are mutualy-exclusive. If none selected, "L" is default. Signed-off-by: Ravi Kerur <rkerur@gmail.com> Signed-off-by: Piotr Azarewicz <piotrx.t.azarewicz@intel.com> Tested-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
2016-02-25 11:24:24 +01:00
examples/l3fwd: rework exact-match Current implementation of Exact-Match uses different execution path than for LPM. Unifying them allows to reuse big part of LPM code and sightly increase performance of Exact-Match. Main changes: ------------- * Packet classification stage is separated from the rest of path for both LPM and EM. * Packet processing, modifying and transmit part is the same for LPM and EM and mostly based on the current LPM implementation. * Shared code is moved to the common file "l3fwd_sse.h". * While sequential packet classification in EM path, seems to be faster than using multi hash lookup, used before, it is used by default. Old implementation is moved to the file l3fwd_em_hlm_sse.h and can be enabled with HASH_LOOKUP_MULTI global define in compilation time. Signed-off-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
2016-02-29 11:33:07 +01:00
if (next_hop >= RTE_MAX_ETHPORTS ||
(enabled_port_mask & 1 << next_hop) == 0)
next_hop = portid;
examples/l3fwd: modularize The main problem with l3fwd is that it is too monolithic with everything being in one file, and the various options all controlled by compile time flags. This means that it's hard to read and understand, and when making any changes, you need to go to a lot of work to try and ensure you cover all the code paths, since a compile of the app will not touch large parts of the l3fwd codebase. Following changes were done to fix the issues mentioned above - Split out the various lpm and hash specific functionality into separate files, so that l3fwd code has one file for common code e.g. args processing, mempool creation, and then individual files for the various forwarding approaches. Following are new file lists main.c (Common code for args processing, memppol creation, etc) l3fwd_em.c (Hash/Exact match aka 'EM' functionality) l3fwd_em_sse.h (SSE4_1 buffer optimizated 'EM' code) l3fwd_lpm.c (Longest Prefix Match aka 'LPM' functionality) l3fwd_lpm_sse.h (SSE4_1 buffer optimizated 'LPM' code) l3fwd.h (Common include for 'EM' and 'LPM') - The choosing of the lpm/hash path should be done at runtime, not compile time, via a command-line argument. This will ensure that both code paths get compiled in a single go Following examples show runtime options provided Select 'LPM' or 'EM' based on run time selection f.e. > l3fwd -c 0x1 -n 1 -- -p 0x1 -E ... (EM) > l3fwd -c 0x1 -n 1 -- -p 0x1 -L ... (LPM) Options "E" and "L" are mutualy-exclusive. If none selected, "L" is default. Signed-off-by: Ravi Kerur <rkerur@gmail.com> Signed-off-by: Piotr Azarewicz <piotrx.t.azarewicz@intel.com> Tested-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
2016-02-25 11:24:24 +01:00
examples/l3fwd: rework exact-match Current implementation of Exact-Match uses different execution path than for LPM. Unifying them allows to reuse big part of LPM code and sightly increase performance of Exact-Match. Main changes: ------------- * Packet classification stage is separated from the rest of path for both LPM and EM. * Packet processing, modifying and transmit part is the same for LPM and EM and mostly based on the current LPM implementation. * Shared code is moved to the common file "l3fwd_sse.h". * While sequential packet classification in EM path, seems to be faster than using multi hash lookup, used before, it is used by default. Old implementation is moved to the file l3fwd_em_hlm_sse.h and can be enabled with HASH_LOOKUP_MULTI global define in compilation time. Signed-off-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
2016-02-29 11:33:07 +01:00
return next_hop;
examples/l3fwd: modularize The main problem with l3fwd is that it is too monolithic with everything being in one file, and the various options all controlled by compile time flags. This means that it's hard to read and understand, and when making any changes, you need to go to a lot of work to try and ensure you cover all the code paths, since a compile of the app will not touch large parts of the l3fwd codebase. Following changes were done to fix the issues mentioned above - Split out the various lpm and hash specific functionality into separate files, so that l3fwd code has one file for common code e.g. args processing, mempool creation, and then individual files for the various forwarding approaches. Following are new file lists main.c (Common code for args processing, memppol creation, etc) l3fwd_em.c (Hash/Exact match aka 'EM' functionality) l3fwd_em_sse.h (SSE4_1 buffer optimizated 'EM' code) l3fwd_lpm.c (Longest Prefix Match aka 'LPM' functionality) l3fwd_lpm_sse.h (SSE4_1 buffer optimizated 'LPM' code) l3fwd.h (Common include for 'EM' and 'LPM') - The choosing of the lpm/hash path should be done at runtime, not compile time, via a command-line argument. This will ensure that both code paths get compiled in a single go Following examples show runtime options provided Select 'LPM' or 'EM' based on run time selection f.e. > l3fwd -c 0x1 -n 1 -- -p 0x1 -E ... (EM) > l3fwd -c 0x1 -n 1 -- -p 0x1 -L ... (LPM) Options "E" and "L" are mutualy-exclusive. If none selected, "L" is default. Signed-off-by: Ravi Kerur <rkerur@gmail.com> Signed-off-by: Piotr Azarewicz <piotrx.t.azarewicz@intel.com> Tested-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
2016-02-25 11:24:24 +01:00
examples/l3fwd: rework exact-match Current implementation of Exact-Match uses different execution path than for LPM. Unifying them allows to reuse big part of LPM code and sightly increase performance of Exact-Match. Main changes: ------------- * Packet classification stage is separated from the rest of path for both LPM and EM. * Packet processing, modifying and transmit part is the same for LPM and EM and mostly based on the current LPM implementation. * Shared code is moved to the common file "l3fwd_sse.h". * While sequential packet classification in EM path, seems to be faster than using multi hash lookup, used before, it is used by default. Old implementation is moved to the file l3fwd_em_hlm_sse.h and can be enabled with HASH_LOOKUP_MULTI global define in compilation time. Signed-off-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
2016-02-29 11:33:07 +01:00
}
examples/l3fwd: modularize The main problem with l3fwd is that it is too monolithic with everything being in one file, and the various options all controlled by compile time flags. This means that it's hard to read and understand, and when making any changes, you need to go to a lot of work to try and ensure you cover all the code paths, since a compile of the app will not touch large parts of the l3fwd codebase. Following changes were done to fix the issues mentioned above - Split out the various lpm and hash specific functionality into separate files, so that l3fwd code has one file for common code e.g. args processing, mempool creation, and then individual files for the various forwarding approaches. Following are new file lists main.c (Common code for args processing, memppol creation, etc) l3fwd_em.c (Hash/Exact match aka 'EM' functionality) l3fwd_em_sse.h (SSE4_1 buffer optimizated 'EM' code) l3fwd_lpm.c (Longest Prefix Match aka 'LPM' functionality) l3fwd_lpm_sse.h (SSE4_1 buffer optimizated 'LPM' code) l3fwd.h (Common include for 'EM' and 'LPM') - The choosing of the lpm/hash path should be done at runtime, not compile time, via a command-line argument. This will ensure that both code paths get compiled in a single go Following examples show runtime options provided Select 'LPM' or 'EM' based on run time selection f.e. > l3fwd -c 0x1 -n 1 -- -p 0x1 -E ... (EM) > l3fwd -c 0x1 -n 1 -- -p 0x1 -L ... (LPM) Options "E" and "L" are mutualy-exclusive. If none selected, "L" is default. Signed-off-by: Ravi Kerur <rkerur@gmail.com> Signed-off-by: Piotr Azarewicz <piotrx.t.azarewicz@intel.com> Tested-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
2016-02-25 11:24:24 +01:00
examples/l3fwd: rework exact-match Current implementation of Exact-Match uses different execution path than for LPM. Unifying them allows to reuse big part of LPM code and sightly increase performance of Exact-Match. Main changes: ------------- * Packet classification stage is separated from the rest of path for both LPM and EM. * Packet processing, modifying and transmit part is the same for LPM and EM and mostly based on the current LPM implementation. * Shared code is moved to the common file "l3fwd_sse.h". * While sequential packet classification in EM path, seems to be faster than using multi hash lookup, used before, it is used by default. Old implementation is moved to the file l3fwd_em_hlm_sse.h and can be enabled with HASH_LOOKUP_MULTI global define in compilation time. Signed-off-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
2016-02-29 11:33:07 +01:00
return portid;
examples/l3fwd: modularize The main problem with l3fwd is that it is too monolithic with everything being in one file, and the various options all controlled by compile time flags. This means that it's hard to read and understand, and when making any changes, you need to go to a lot of work to try and ensure you cover all the code paths, since a compile of the app will not touch large parts of the l3fwd codebase. Following changes were done to fix the issues mentioned above - Split out the various lpm and hash specific functionality into separate files, so that l3fwd code has one file for common code e.g. args processing, mempool creation, and then individual files for the various forwarding approaches. Following are new file lists main.c (Common code for args processing, memppol creation, etc) l3fwd_em.c (Hash/Exact match aka 'EM' functionality) l3fwd_em_sse.h (SSE4_1 buffer optimizated 'EM' code) l3fwd_lpm.c (Longest Prefix Match aka 'LPM' functionality) l3fwd_lpm_sse.h (SSE4_1 buffer optimizated 'LPM' code) l3fwd.h (Common include for 'EM' and 'LPM') - The choosing of the lpm/hash path should be done at runtime, not compile time, via a command-line argument. This will ensure that both code paths get compiled in a single go Following examples show runtime options provided Select 'LPM' or 'EM' based on run time selection f.e. > l3fwd -c 0x1 -n 1 -- -p 0x1 -E ... (EM) > l3fwd -c 0x1 -n 1 -- -p 0x1 -L ... (LPM) Options "E" and "L" are mutualy-exclusive. If none selected, "L" is default. Signed-off-by: Ravi Kerur <rkerur@gmail.com> Signed-off-by: Piotr Azarewicz <piotrx.t.azarewicz@intel.com> Tested-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
2016-02-25 11:24:24 +01:00
}
/*
* Buffer optimized handling of packets, invoked
* from main_loop.
*/
static inline void
l3fwd_em_send_packets(int nb_rx, struct rte_mbuf **pkts_burst,
uint8_t portid, struct lcore_conf *qconf)
{
int32_t j;
lpm: extend IPv4 next hop field This patch extend next_hop field from 8-bits to 24-bits in LPM library for IPv4. Added versioning symbols to functions and updated library and applications that have a dependency on LPM library. Signed-off-by: Michal Kobylinski <michalx.kobylinski@intel.com> Acked-by: David Hunt <david.hunt@intel.com>
2016-03-09 17:57:15 +01:00
uint32_t dst_port[MAX_PKT_BURST];
examples/l3fwd: modularize The main problem with l3fwd is that it is too monolithic with everything being in one file, and the various options all controlled by compile time flags. This means that it's hard to read and understand, and when making any changes, you need to go to a lot of work to try and ensure you cover all the code paths, since a compile of the app will not touch large parts of the l3fwd codebase. Following changes were done to fix the issues mentioned above - Split out the various lpm and hash specific functionality into separate files, so that l3fwd code has one file for common code e.g. args processing, mempool creation, and then individual files for the various forwarding approaches. Following are new file lists main.c (Common code for args processing, memppol creation, etc) l3fwd_em.c (Hash/Exact match aka 'EM' functionality) l3fwd_em_sse.h (SSE4_1 buffer optimizated 'EM' code) l3fwd_lpm.c (Longest Prefix Match aka 'LPM' functionality) l3fwd_lpm_sse.h (SSE4_1 buffer optimizated 'LPM' code) l3fwd.h (Common include for 'EM' and 'LPM') - The choosing of the lpm/hash path should be done at runtime, not compile time, via a command-line argument. This will ensure that both code paths get compiled in a single go Following examples show runtime options provided Select 'LPM' or 'EM' based on run time selection f.e. > l3fwd -c 0x1 -n 1 -- -p 0x1 -E ... (EM) > l3fwd -c 0x1 -n 1 -- -p 0x1 -L ... (LPM) Options "E" and "L" are mutualy-exclusive. If none selected, "L" is default. Signed-off-by: Ravi Kerur <rkerur@gmail.com> Signed-off-by: Piotr Azarewicz <piotrx.t.azarewicz@intel.com> Tested-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
2016-02-25 11:24:24 +01:00
examples/l3fwd: rework exact-match Current implementation of Exact-Match uses different execution path than for LPM. Unifying them allows to reuse big part of LPM code and sightly increase performance of Exact-Match. Main changes: ------------- * Packet classification stage is separated from the rest of path for both LPM and EM. * Packet processing, modifying and transmit part is the same for LPM and EM and mostly based on the current LPM implementation. * Shared code is moved to the common file "l3fwd_sse.h". * While sequential packet classification in EM path, seems to be faster than using multi hash lookup, used before, it is used by default. Old implementation is moved to the file l3fwd_em_hlm_sse.h and can be enabled with HASH_LOOKUP_MULTI global define in compilation time. Signed-off-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
2016-02-29 11:33:07 +01:00
for (j = 0; j < nb_rx; j++)
dst_port[j] = em_get_dst_port(qconf, pkts_burst[j], portid);
examples/l3fwd: modularize The main problem with l3fwd is that it is too monolithic with everything being in one file, and the various options all controlled by compile time flags. This means that it's hard to read and understand, and when making any changes, you need to go to a lot of work to try and ensure you cover all the code paths, since a compile of the app will not touch large parts of the l3fwd codebase. Following changes were done to fix the issues mentioned above - Split out the various lpm and hash specific functionality into separate files, so that l3fwd code has one file for common code e.g. args processing, mempool creation, and then individual files for the various forwarding approaches. Following are new file lists main.c (Common code for args processing, memppol creation, etc) l3fwd_em.c (Hash/Exact match aka 'EM' functionality) l3fwd_em_sse.h (SSE4_1 buffer optimizated 'EM' code) l3fwd_lpm.c (Longest Prefix Match aka 'LPM' functionality) l3fwd_lpm_sse.h (SSE4_1 buffer optimizated 'LPM' code) l3fwd.h (Common include for 'EM' and 'LPM') - The choosing of the lpm/hash path should be done at runtime, not compile time, via a command-line argument. This will ensure that both code paths get compiled in a single go Following examples show runtime options provided Select 'LPM' or 'EM' based on run time selection f.e. > l3fwd -c 0x1 -n 1 -- -p 0x1 -E ... (EM) > l3fwd -c 0x1 -n 1 -- -p 0x1 -L ... (LPM) Options "E" and "L" are mutualy-exclusive. If none selected, "L" is default. Signed-off-by: Ravi Kerur <rkerur@gmail.com> Signed-off-by: Piotr Azarewicz <piotrx.t.azarewicz@intel.com> Tested-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
2016-02-25 11:24:24 +01:00
examples/l3fwd: rework exact-match Current implementation of Exact-Match uses different execution path than for LPM. Unifying them allows to reuse big part of LPM code and sightly increase performance of Exact-Match. Main changes: ------------- * Packet classification stage is separated from the rest of path for both LPM and EM. * Packet processing, modifying and transmit part is the same for LPM and EM and mostly based on the current LPM implementation. * Shared code is moved to the common file "l3fwd_sse.h". * While sequential packet classification in EM path, seems to be faster than using multi hash lookup, used before, it is used by default. Old implementation is moved to the file l3fwd_em_hlm_sse.h and can be enabled with HASH_LOOKUP_MULTI global define in compilation time. Signed-off-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
2016-02-29 11:33:07 +01:00
send_packets_multi(qconf, pkts_burst, dst_port, nb_rx);
examples/l3fwd: modularize The main problem with l3fwd is that it is too monolithic with everything being in one file, and the various options all controlled by compile time flags. This means that it's hard to read and understand, and when making any changes, you need to go to a lot of work to try and ensure you cover all the code paths, since a compile of the app will not touch large parts of the l3fwd codebase. Following changes were done to fix the issues mentioned above - Split out the various lpm and hash specific functionality into separate files, so that l3fwd code has one file for common code e.g. args processing, mempool creation, and then individual files for the various forwarding approaches. Following are new file lists main.c (Common code for args processing, memppol creation, etc) l3fwd_em.c (Hash/Exact match aka 'EM' functionality) l3fwd_em_sse.h (SSE4_1 buffer optimizated 'EM' code) l3fwd_lpm.c (Longest Prefix Match aka 'LPM' functionality) l3fwd_lpm_sse.h (SSE4_1 buffer optimizated 'LPM' code) l3fwd.h (Common include for 'EM' and 'LPM') - The choosing of the lpm/hash path should be done at runtime, not compile time, via a command-line argument. This will ensure that both code paths get compiled in a single go Following examples show runtime options provided Select 'LPM' or 'EM' based on run time selection f.e. > l3fwd -c 0x1 -n 1 -- -p 0x1 -E ... (EM) > l3fwd -c 0x1 -n 1 -- -p 0x1 -L ... (LPM) Options "E" and "L" are mutualy-exclusive. If none selected, "L" is default. Signed-off-by: Ravi Kerur <rkerur@gmail.com> Signed-off-by: Piotr Azarewicz <piotrx.t.azarewicz@intel.com> Tested-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
2016-02-25 11:24:24 +01:00
}
#endif /* __L3FWD_EM_SSE_H__ */
Reference in New Issue Copy Permalink