[dpdk-dev] [PATCH] virtio: Update max RX packet length
Update max RX packet length since virtio PMD has the capability of receiving and transmitting jumbo frame. Signed-off-by: Changchun Ouyang Tested-by: Jingguo Fu --- lib/librte_pmd_virtio/virtio_ethdev.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/lib/librte_pmd_virtio/virtio_ethdev.h b/lib/librte_pmd_virtio/virtio_ethdev.h index d2e1eed..1da3c62 100644 --- a/lib/librte_pmd_virtio/virtio_ethdev.h +++ b/lib/librte_pmd_virtio/virtio_ethdev.h @@ -53,7 +53,7 @@ #define VIRTIO_MAX_TX_QUEUES 128 #define VIRTIO_MAX_MAC_ADDRS 1 #define VIRTIO_MIN_RX_BUFSIZE 64 -#define VIRTIO_MAX_RX_PKTLEN 1518 +#define VIRTIO_MAX_RX_PKTLEN 9728 /* Features desired/implemented by this driver. */ #define VTNET_FEATURES \ -- 1.8.4.2
[dpdk-dev] [PATCH] hash: added rte_hash_keys to extract all keys
Hello! If there's already an iterator then I guess this patch is superfluous. I tried searching for it among the e-mails but I can find it, can someone point me to it? If you still think this has some values I will take the given feedback and make corrections. Best regards, Tomas Vestelind ps. Stephen sorry for the extra mail, I forgot to cc the correct people in the first one. On 8/29/14, Stephen Hemminger wrote: > We implemented a more general hash iterator, thought the patch was already > submitted. >
[dpdk-dev] [PATCH 0/3] Transform the vhost example to two parts: vhost library and example.
This transform is to facilitate integration with DPDK accelerated ethernet switch. Huawei Xie (3): remove vhost example vhost library support to facilitate integration with DPDK accelerated vswitch This vhost example is based on vhost library. Besides, - This patch fixes hundreds of 80 character limitation issues. - Use structure assignment rather than memcpy in get_eth_conf according to Stephen's comments. config/common_linuxapp |7 + examples/vhost/Makefile | 10 +- examples/vhost/eventfd_link/Makefile | 39 - examples/vhost/eventfd_link/eventfd_link.c | 205 --- examples/vhost/eventfd_link/eventfd_link.h | 79 - examples/vhost/libvirt/qemu-wrap.py |5 +- examples/vhost/main.c| 2205 ++ examples/vhost/main.h| 85 +- examples/vhost/vhost-net-cdev.c | 367 - examples/vhost/vhost-net-cdev.h | 83 - examples/vhost/virtio-net.c | 1165 -- examples/vhost/virtio-net.h | 161 -- lib/Makefile |1 + lib/librte_vhost/Makefile| 48 + lib/librte_vhost/eventfd_link/Makefile | 39 + lib/librte_vhost/eventfd_link/eventfd_link.c | 196 +++ lib/librte_vhost/eventfd_link/eventfd_link.h | 40 + lib/librte_vhost/rte_virtio_net.h| 222 +++ lib/librte_vhost/vhost-net-cdev.c| 394 + lib/librte_vhost/vhost-net-cdev.h| 119 ++ lib/librte_vhost/vhost_rxtx.c| 316 lib/librte_vhost/virtio-net.c| 1113 + mk/rte.app.mk|5 + 23 files changed, 3371 insertions(+), 3533 deletions(-) delete mode 100644 examples/vhost/eventfd_link/Makefile delete mode 100644 examples/vhost/eventfd_link/eventfd_link.c delete mode 100644 examples/vhost/eventfd_link/eventfd_link.h mode change 100755 => 100644 examples/vhost/libvirt/qemu-wrap.py delete mode 100644 examples/vhost/vhost-net-cdev.c delete mode 100644 examples/vhost/vhost-net-cdev.h delete mode 100644 examples/vhost/virtio-net.c delete mode 100644 examples/vhost/virtio-net.h create mode 100644 lib/librte_vhost/Makefile create mode 100644 lib/librte_vhost/eventfd_link/Makefile create mode 100644 lib/librte_vhost/eventfd_link/eventfd_link.c create mode 100644 lib/librte_vhost/eventfd_link/eventfd_link.h create mode 100644 lib/librte_vhost/rte_virtio_net.h create mode 100644 lib/librte_vhost/vhost-net-cdev.c create mode 100644 lib/librte_vhost/vhost-net-cdev.h create mode 100644 lib/librte_vhost/vhost_rxtx.c create mode 100644 lib/librte_vhost/virtio-net.c -- 1.8.1.4
[dpdk-dev] [PATCH 1/3] examples/vhost: remove vhost example
Signed-off-by: Huawei Xie Acked-by: Konstantin Ananyev Acked-by: Thomos Long --- examples/vhost/Makefile| 60 - examples/vhost/eventfd_link/Makefile | 39 - examples/vhost/eventfd_link/eventfd_link.c | 205 -- examples/vhost/eventfd_link/eventfd_link.h | 79 - examples/vhost/libvirt/qemu-wrap.py| 367 --- examples/vhost/main.c | 3722 examples/vhost/main.h | 86 - examples/vhost/vhost-net-cdev.c| 367 --- examples/vhost/vhost-net-cdev.h| 83 - examples/vhost/virtio-net.c| 1165 - examples/vhost/virtio-net.h| 161 -- 11 files changed, 6334 deletions(-) delete mode 100644 examples/vhost/Makefile delete mode 100644 examples/vhost/eventfd_link/Makefile delete mode 100644 examples/vhost/eventfd_link/eventfd_link.c delete mode 100644 examples/vhost/eventfd_link/eventfd_link.h delete mode 100755 examples/vhost/libvirt/qemu-wrap.py delete mode 100644 examples/vhost/main.c delete mode 100644 examples/vhost/main.h delete mode 100644 examples/vhost/vhost-net-cdev.c delete mode 100644 examples/vhost/vhost-net-cdev.h delete mode 100644 examples/vhost/virtio-net.c delete mode 100644 examples/vhost/virtio-net.h diff --git a/examples/vhost/Makefile b/examples/vhost/Makefile deleted file mode 100644 index f45f83f..000 --- a/examples/vhost/Makefile +++ /dev/null @@ -1,60 +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. - -ifeq ($(RTE_SDK),) -$(error "Please define RTE_SDK environment variable") -endif - -# Default target, can be overriden by command line or environment -RTE_TARGET ?= x86_64-native-linuxapp-gcc - -include $(RTE_SDK)/mk/rte.vars.mk - -ifneq ($(CONFIG_RTE_EXEC_ENV),"linuxapp") -$(info This application can only operate in a linuxapp environment, \ -please change the definition of the RTE_TARGET environment variable) -all: -else - -# binary name -APP = vhost-switch - -# all source are stored in SRCS-y -#SRCS-y := cusedrv.c loopback-userspace.c -SRCS-y := main.c virtio-net.c vhost-net-cdev.c - -CFLAGS += -O2 -I/usr/local/include -D_FILE_OFFSET_BITS=64 -Wno-unused-parameter -CFLAGS += $(WERROR_FLAGS) -LDFLAGS += -lfuse - -include $(RTE_SDK)/mk/rte.extapp.mk - -endif diff --git a/examples/vhost/eventfd_link/Makefile b/examples/vhost/eventfd_link/Makefile deleted file mode 100644 index fc3927b..000 --- a/examples/vhost/eventfd_link/Makefile +++ /dev/null @@ -1,39 +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
[dpdk-dev] [PATCH 2/3] lib/librte_vhost: vhost library support to facilitate integration with DPDK accelerated vswitch
This library is turned off by default so that it doesn't break build on default system. Install fuse development library and turn it on. Signed-off-by: Huawei Xie Acked-by: Konstantin Ananyev Acked-by: Thomos Long --- config/common_linuxapp |7 + lib/Makefile |1 + lib/librte_vhost/Makefile| 48 ++ lib/librte_vhost/eventfd_link/Makefile | 39 + lib/librte_vhost/eventfd_link/eventfd_link.c | 196 + lib/librte_vhost/eventfd_link/eventfd_link.h | 40 + lib/librte_vhost/rte_virtio_net.h| 222 + lib/librte_vhost/vhost-net-cdev.c| 394 + lib/librte_vhost/vhost-net-cdev.h| 119 +++ lib/librte_vhost/vhost_rxtx.c| 316 lib/librte_vhost/virtio-net.c| 1113 ++ mk/rte.app.mk|5 + 12 files changed, 2500 insertions(+) create mode 100644 lib/librte_vhost/Makefile create mode 100644 lib/librte_vhost/eventfd_link/Makefile create mode 100644 lib/librte_vhost/eventfd_link/eventfd_link.c create mode 100644 lib/librte_vhost/eventfd_link/eventfd_link.h create mode 100644 lib/librte_vhost/rte_virtio_net.h create mode 100644 lib/librte_vhost/vhost-net-cdev.c create mode 100644 lib/librte_vhost/vhost-net-cdev.h create mode 100644 lib/librte_vhost/vhost_rxtx.c create mode 100644 lib/librte_vhost/virtio-net.c diff --git a/config/common_linuxapp b/config/common_linuxapp index 9047975..c7c1c83 100644 --- a/config/common_linuxapp +++ b/config/common_linuxapp @@ -390,6 +390,13 @@ CONFIG_RTE_KNI_VHOST_DEBUG_RX=n CONFIG_RTE_KNI_VHOST_DEBUG_TX=n # +# Compile vhost library +# fuse, fuse-devel, kernel-modules-extra packages are needed +# +CONFIG_RTE_LIBRTE_VHOST=n +CONFIG_RTE_LIBRTE_VHOST_DEBUG=n + +# #Compile Xen domain0 support # CONFIG_RTE_LIBRTE_XEN_DOM0=n diff --git a/lib/Makefile b/lib/Makefile index 10c5bb3..007c174 100644 --- a/lib/Makefile +++ b/lib/Makefile @@ -60,6 +60,7 @@ DIRS-$(CONFIG_RTE_LIBRTE_METER) += librte_meter DIRS-$(CONFIG_RTE_LIBRTE_SCHED) += librte_sched DIRS-$(CONFIG_RTE_LIBRTE_KVARGS) += librte_kvargs DIRS-$(CONFIG_RTE_LIBRTE_DISTRIBUTOR) += librte_distributor +DIRS-$(CONFIG_RTE_LIBRTE_VHOST) += librte_vhost DIRS-$(CONFIG_RTE_LIBRTE_PORT) += librte_port DIRS-$(CONFIG_RTE_LIBRTE_TABLE) += librte_table DIRS-$(CONFIG_RTE_LIBRTE_PIPELINE) += librte_pipeline diff --git a/lib/librte_vhost/Makefile b/lib/librte_vhost/Makefile new file mode 100644 index 000..6ad706d --- /dev/null +++ b/lib/librte_vhost/Makefile @@ -0,0 +1,48 @@ +# 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_vhost.a + +CFLAGS += $(WERROR_FLAGS) -I$(SRCDIR) -O3 -D_FILE_OFFSET_BITS=64 -lfuse +LDFLAGS += -lfuse +# all source are stored in SRCS-y +SRCS-$(CONFIG_RTE_LIBRTE_VHOST) := vhost-net-cdev.c virtio-net.c vhost_rxtx.c + +# install includes +SYMLINK-$(CONFIG_RTE_LIBRTE_VHOST)-include += rte_virtio_net.h + +# this lib needs eal +DEPDIRS-$(CONFIG_RTE_LIBRTE_VHOST) += lib/librte_eal lib/librte_mbuf + +include $(RTE_SDK)/mk/rte.lib.mk diff --git a/lib/librte_vhost/eventfd_link/Makefile b/lib/librte_vhost/eventfd_link/Makefile new file mode 100644 index 000..fc3927b --- /dev/null +++ b/lib/librte_vhost/e
[dpdk-dev] [PATCH 3/3] examples/vhost: vhost example based on vhost library.
Signed-off-by: Huawei Xie Acked-by: Konstantin Ananyev --- examples/vhost/Makefile | 52 + examples/vhost/libvirt/qemu-wrap.py | 366 examples/vhost/main.c | 3145 +++ examples/vhost/main.h | 109 ++ 4 files changed, 3672 insertions(+) create mode 100644 examples/vhost/Makefile create mode 100644 examples/vhost/libvirt/qemu-wrap.py create mode 100644 examples/vhost/main.c create mode 100644 examples/vhost/main.h diff --git a/examples/vhost/Makefile b/examples/vhost/Makefile new file mode 100644 index 000..a4d4fb0 --- /dev/null +++ b/examples/vhost/Makefile @@ -0,0 +1,52 @@ +# 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. + +ifeq ($(RTE_SDK),) +$(error "Please define RTE_SDK environment variable") +endif + +# Default target, can be overriden by command line or environment +RTE_TARGET ?= x86_64-native-linuxapp-gcc + +include $(RTE_SDK)/mk/rte.vars.mk + +# binary name +APP = vhost-switch + +# all source are stored in SRCS-y +#SRCS-y := cusedrv.c loopback-userspace.c +SRCS-y := main.c + +CFLAGS += -O2 -I/usr/local/include -D_FILE_OFFSET_BITS=64 -Wno-unused-parameter +CFLAGS += $(WERROR_FLAGS) +LDFLAGS += -lfuse + +include $(RTE_SDK)/mk/rte.extapp.mk diff --git a/examples/vhost/libvirt/qemu-wrap.py b/examples/vhost/libvirt/qemu-wrap.py new file mode 100644 index 000..8d820be --- /dev/null +++ b/examples/vhost/libvirt/qemu-wrap.py @@ -0,0 +1,366 @@ +#!/usr/bin/python +#/* +# * 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. +# */ + +# +# This script is designed to modify the call to the
[dpdk-dev] [PATCH v2 4/6] bond: free mbufs if transmission fails in bonding tx_burst functions
> -Original Message- > From: Doherty, Declan > Sent: Monday, September 1, 2014 9:31 AM > To: dev at dpdk.org > Cc: thomas.monjalon at 6wind.com; rsanford at akamai.com; Doherty, Declan > Subject: [PATCH v2 4/6] bond: free mbufs if transmission fails in bonding > tx_burst > functions > > Fixing a number of corner cases that if transmission failed on slave devices > then > this > could lead to leaked mbufs > > V2 addresses behaviouraly issues in the first version and packets are no > longer > freed in > the bonding layer, except in the case of broadcast mode where in failures > happen > on > more than a single slave then mbufs will be freed in all slaves except the one > where > the least errors occured. Also contains new unit tests to test the > transmission > failure case in slaves for > round-robin, balance, and broadcast modes. > > > Signed-off-by: Declan Doherty > --- . > -- > 1.7.0.7 I've just noticed that I inserted an extra character into the patch when I was annotating this patch which will cause this a fail compilation. > + TEST_BCAST_SLAVE_TX_FAIL_BURST_SIZE -s > + TEST_BCAST_SLAVE_TX_FAIL_MIN_PACKETS_COUNT); should have been > + TEST_BCAST_SLAVE_TX_FAIL_BURST_SIZE - > + TEST_BCAST_SLAVE_TX_FAIL_MIN_PACKETS_COUNT); Thomas, do you want me to submit a V3 of this? Regards Declan
[dpdk-dev] [PATCH v2 4/6] bond: free mbufs if transmission fails in bonding tx_burst functions
Hi Declan, 2014-09-02 09:22, Doherty, Declan: > I've just noticed that I inserted an extra character into the patch when I > was annotating this > patch which will cause this a fail compilation. > > > + TEST_BCAST_SLAVE_TX_FAIL_BURST_SIZE -s > > + TEST_BCAST_SLAVE_TX_FAIL_MIN_PACKETS_COUNT); > > should have been > > + TEST_BCAST_SLAVE_TX_FAIL_BURST_SIZE - > > + TEST_BCAST_SLAVE_TX_FAIL_MIN_PACKETS_COUNT); > > Thomas, do you want me to submit a V3 of this? I'd prefer to have some comments first. Anyone to review this patchset? -- Thomas
[dpdk-dev] [PATCH 0/3] Transform the vhost example to two parts: vhost library and example.
Hi Huawei, 2014-09-02 16:55, Huawei Xie: > This transform is to facilitate integration with DPDK accelerated ethernet > switch. > > Huawei Xie (3): > remove vhost example > vhost library support to facilitate integration with DPDK accelerated > vswitch > This vhost example is based on vhost library. Besides, - > This patch fixes hundreds of 80 character limitation issues. > - Use structure assignment rather than memcpy in get_eth_conf > according to Stephen's comments. My previous comment was: " It would help if you made a first patch to move existing code, another patch to convert it into a lib, and a last one for the new example. So it would show how you transform the old example code and would be easier to review. " I mean your first patch would be some simple code moving without any change. Git handles move and shows it clearly, so it's really easy to review. Then your second patch would include the changes you made to make it a library. In this patchset we cannot see them because all is seen as new code. Thanks -- Thomas
[dpdk-dev] Wrong TCP checksum of packets sent by Linux guest (virtIO/vhost)
Hello, This is a repost of https://lists.01.org/pipermail/dpdk-ovs/2014-September/001788.html as the issue might be related to dpdk.org vhost code. I am using dpdk-ovs 1.1.0 (latest release) as follow : linux-guest (no DPKD) <--- virtIO ---> OVDK 1.1.0 (with latest DPDK [*]) < --- Niantic --- > linux-native [*] commit 8fd8bebc051704d7caecfed8d8a065a79c022329 Author: Adrien Mazarguil Date: Mon Sep 1 12:31:11 2014 +0200 UDP/ICMP connectivity is fine, but TCP checksum of packet sent by the guest are incorrect, as showed with tcpdump on linux-native. Is it supposed to work? If yes, what is the proper qemu version to use and its parameters? I am using qemu version provided by dpdk-ovs (same issue with PDDK 1.7.0 and latest one) and I followed this documentation: https://github.com/01org/dpdk-ovs/blob/development/docs/04_Sample_Configurations/02_Userspace-vHost.md Thanks! Franck This message and any attachments (the "message") are confidential, intended solely for the addressees. If you are not the intended recipient, please notify the sender immediately by e-mail and delete this message from your system. In this case, you are not authorized to use, copy this message and/or disclose the content to any other person. E-mails are susceptible to alteration. Neither Qosmos nor any of its subsidiaries or affiliates shall be liable for the message if altered, changed or falsified. Ce message et toutes ses pi?ces jointes (ci-apr?s le "message")sont confidentiels et ?tablis ? l'intention exclusive de ses destinataires. Si vous avez re?u ce message par erreur, merci d?en informer imm?diatement son ?metteur par courrier ?lectronique et d?effacer ce message de votre syst?me. Dans cette hypoth?se, vous n??tes pas autoris? ? utiliser, copier ce message et/ou en divulguer le contenu ? un tiers. Tout message ?lectronique est susceptible d'alt?ration. Qosmos et ses filiales d?clinent toute responsabilit? au titre de ce message s'il a ?t? alt?r?, d?form? ou falsifi?.
[dpdk-dev] Wrong TCP checksum of packets sent by Linux guest (virtIO/vhost)
Hi Franck, 2014-09-02 13:20, Franck BAUDIN: > I am using dpdk-ovs 1.1.0 (latest release) as follow : > > linux-guest (no DPKD) <--- virtIO ---> OVDK 1.1.0 (with latest DPDK [*]) < > --- Niantic --- > linux-native > [...] > UDP/ICMP connectivity is fine, but TCP checksum of packet sent by the guest > are incorrect, as showed with tcpdump on linux-native. Thanks for reporting. Could you try virtio without vhost? > This message and any attachments (the "message") are confidential, intended > solely for the addressees. If you are not the intended recipient, please > notify the sender immediately by e-mail and delete this message from your > system. In this case, you are not authorized to use, copy this message and/or > disclose the content to any other person. E-mails are susceptible to > alteration. Neither Qosmos nor any of its subsidiaries or affiliates shall be > liable for the message if altered, changed or falsified. > > Ce message et toutes ses pi?ces jointes (ci-apr?s le "message")sont > confidentiels et ?tablis ? l'intention exclusive de ses destinataires. Si > vous avez re?u ce message par erreur, merci d?en informer imm?diatement son > ?metteur par courrier ?lectronique et d?effacer ce message de votre syst?me. > Dans cette hypoth?se, vous n??tes pas autoris? ? utiliser, copier ce message > et/ou en divulguer le contenu ? un tiers. Tout message ?lectronique est > susceptible d'alt?ration. Qosmos et ses filiales d?clinent toute > responsabilit? au titre de ce message s'il a ?t? alt?r?, d?form? ou falsifi?. This big disclaimer block has no sense on a public mailing list. Please try to remove it. Thanks -- Thomas
[dpdk-dev] [PATCH v2 0/6] link bonding
> -Original Message- > From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Declan Doherty > Sent: Monday, September 01, 2014 9:31 AM > To: dev at dpdk.org > Subject: [dpdk-dev] [PATCH v2 0/6] link bonding > > This patch set adds support for link status interrupt in the link bonding > pmd. It also contains some patches to tidy up the code structure and to > of the link bonding code and to fix bugs relating to transmission > failures in the under lying slave pmd which could lead to leaked mbufs. > > V2 addresses issues with the logic around the handling of fail transmissions. > In this version all modei behave in a manner similar to a standard PMD, > returning > the number of successfully transmitted mbufs and with the failing mbufs at > the end of bufs array for freeing / retansmission by the application software > > > Declan Doherty (6): > bond: link status interrupt support > bond: removing switch statement from rx burst method > bond: fix naming inconsistency in tx_burst_round_robin > bond: free mbufs if transmission fails in bonding tx_burst functions > test app: adding support for generating variable sized packets > testpmd: adding parameter to reconfig method to set socket_id when > adding new port to portlist > > app/test-pmd/cmdline.c |2 +- > app/test-pmd/testpmd.c |3 +- > app/test-pmd/testpmd.h |2 +- > app/test/packet_burst_generator.c | 22 +- > app/test/packet_burst_generator.h |6 +- > app/test/test_link_bonding.c | 620 > +--- > app/test/virtual_pmd.c | 80 - > app/test/virtual_pmd.h |7 + > lib/librte_pmd_bond/rte_eth_bond_api.c |4 + > lib/librte_pmd_bond/rte_eth_bond_pmd.c | 161 ++--- > 10 files changed, 779 insertions(+), 128 deletions(-) Acked-by Pablo de Lara
[dpdk-dev] [PATCHv5] librte_acl make it build/work for 'default' target
On Mon, Sep 01, 2014 at 04:28:44PM +0100, Konstantin Ananyev wrote: > Make ACL library to build/work on 'default' architecture: > - make rte_acl_classify_scalar really scalar > (make sure it wouldn't use sse4 instrincts through resolve_priority()). > - Provide two versions of rte_acl_classify code path: > rte_acl_classify_sse() - could be build and used only on systems with sse4.2 > and upper, return -ENOTSUP on lower arch. > rte_acl_classify_scalar() - a slower version, but could be build and used > on all systems. > - keep common code shared between these two codepaths. > > v2 chages: > run-time selection of most appropriate code-path for given ISA. > By default the highest supprted one is selected. > User can still override that selection by manually assigning new value to > the global function pointer rte_acl_default_classify. > rte_acl_classify() becomes a macro calling whatever rte_acl_default_classify > points to. > > V3 Changes > Updated classify pointer to be a function so as to better preserve ABI > REmoved macro definitions for match check functions to make them static > inline > > V4 Changes > Rewrote classification selection mechanim to use a function table, so that we > can just store the preferred alg in the rte_acl_ctx struct so that > multiprocess > access works. I understand that leaves us with an extra load instruction, > but I > think thats ok, because it also allows... > > Addition of a new function rte_acl_classify_alg. This function lets you > specify an enum value to override the acl contexts default algorith when > doing a > classification. This allows an application to specify a classification > algorithm without needing to pulicize each method. I know there was concern > over keeping those methods public, but we don't have a static ABI at the > moment, > so this seems to me a reasonable thing to do, as it gives us less of an ABI > surface to worry about. > > Fixed misc missed static declarations > Removed acl_match_check.h and moved match_check function to acl_run.h > typdeffed function pointer to match check. > > V5 Changes > Updated examples/l3fwd-acl to comply with latest changes. > Applied other code review comments (mostly style changes). > > Signed-off-by: Konstantin Ananyev Acked-by: Neil Horman Thanks Konstantin! Neil
[dpdk-dev] [PATCH v2 01/17] ixgbe: use the right debug macro
Hi David,
A couple of minor comments inline below. Looks good otherwise.
Jay
On Mon, Sep 1, 2014 at 5:24 AM, David Marchand
wrote:
> - We should not use DEBUGOUT*/DEBUGFUNC macros in non-shared code.
> These macros come as compat wrappers for shared code.
> - We should avoid calling RTE_LOG directly as pmd provides a wrapper for
> logs.
>
> Signed-off-by: David Marchand
> ---
> lib/librte_pmd_ixgbe/ixgbe_82599_bypass.c | 14
> lib/librte_pmd_ixgbe/ixgbe_bypass.c | 26 +++---
> lib/librte_pmd_ixgbe/ixgbe_ethdev.c | 27 +++
> lib/librte_pmd_ixgbe/ixgbe_pf.c |4 +--
> lib/librte_pmd_ixgbe/ixgbe_rxtx.c | 53
> +++--
> 5 files changed, 63 insertions(+), 61 deletions(-)
>
> diff --git a/lib/librte_pmd_ixgbe/ixgbe_82599_bypass.c
> b/lib/librte_pmd_ixgbe/ixgbe_82599_bypass.c
> index 0fc..2623419 100644
> --- a/lib/librte_pmd_ixgbe/ixgbe_82599_bypass.c
> +++ b/lib/librte_pmd_ixgbe/ixgbe_82599_bypass.c
> @@ -63,7 +63,7 @@ ixgbe_set_fiber_fixed_speed(struct ixgbe_hw *hw,
> ixgbe_link_speed speed)
> rs = IXGBE_SFF_SOFT_RS_SELECT_1G;
> break;
> default:
> - DEBUGOUT("Invalid fixed module speed\n");
> + PMD_DRV_LOG("Invalid fixed module speed");
> return;
> }
>
> @@ -72,7 +72,7 @@ ixgbe_set_fiber_fixed_speed(struct ixgbe_hw *hw,
> ixgbe_link_speed speed)
>IXGBE_I2C_EEPROM_DEV_ADDR2,
>&eeprom_data);
> if (status) {
> - DEBUGOUT("Failed to read Rx Rate Select RS0\n");
> + PMD_DRV_LOG("Failed to read Rx Rate Select RS0");
> goto out;
> }
>
> @@ -82,7 +82,7 @@ ixgbe_set_fiber_fixed_speed(struct ixgbe_hw *hw,
> ixgbe_link_speed speed)
> IXGBE_I2C_EEPROM_DEV_ADDR2,
> eeprom_data);
> if (status) {
> - DEBUGOUT("Failed to write Rx Rate Select RS0\n");
> + PMD_DRV_LOG("Failed to write Rx Rate Select RS0");
> goto out;
> }
>
> @@ -91,7 +91,7 @@ ixgbe_set_fiber_fixed_speed(struct ixgbe_hw *hw,
> ixgbe_link_speed speed)
>IXGBE_I2C_EEPROM_DEV_ADDR2,
>&eeprom_data);
> if (status) {
> - DEBUGOUT("Failed to read Rx Rate Select RS1\n");
> + PMD_DRV_LOG("Failed to read Rx Rate Select RS1");
> goto out;
> }
>
> @@ -101,7 +101,7 @@ ixgbe_set_fiber_fixed_speed(struct ixgbe_hw *hw,
> ixgbe_link_speed speed)
> IXGBE_I2C_EEPROM_DEV_ADDR2,
> eeprom_data);
> if (status) {
> - DEBUGOUT("Failed to write Rx Rate Select RS1\n");
> + PMD_DRV_LOG("Failed to write Rx Rate Select RS1");
> goto out;
> }
> out:
> @@ -130,7 +130,7 @@ ixgbe_setup_mac_link_multispeed_fixed_fiber(struct
> ixgbe_hw *hw,
> bool link_up = false;
> bool negotiation;
>
> - DEBUGFUNC("");
> + PMD_INIT_FUNC_TRACE();
>
> /* Mask off requested but non-supported speeds */
> status = ixgbe_get_link_capabilities(hw, &link_speed,
> &negotiation);
> @@ -261,7 +261,7 @@ ixgbe_bypass_get_media_type(struct ixgbe_hw *hw)
> {
> enum ixgbe_media_type media_type;
>
> - DEBUGFUNC("");
> + PMD_INIT_FUNC_TRACE();
>
> if (hw->device_id == IXGBE_DEV_ID_82599_BYPASS) {
> media_type = ixgbe_media_type_fiber;
> diff --git a/lib/librte_pmd_ixgbe/ixgbe_bypass.c
> b/lib/librte_pmd_ixgbe/ixgbe_bypass.c
> index 1d21dc0..1a980b8 100644
> --- a/lib/librte_pmd_ixgbe/ixgbe_bypass.c
> +++ b/lib/librte_pmd_ixgbe/ixgbe_bypass.c
> @@ -40,20 +40,20 @@
> #defineBYPASS_STATUS_OFF_MASK 3
>
> /* Macros to check for invlaid function pointers. */
> -#defineFUNC_PTR_OR_ERR_RET(func, retval) do { \
> - if ((func) == NULL) { \
> - DEBUGOUT("%s:%d function not supported\n", \
> - __func__, __LINE__); \
> - return (retval); \
> - } \
> +#defineFUNC_PTR_OR_ERR_RET(func, retval) do { \
> + if ((func) == NULL) { \
> + PMD_DRV_LOG("%s:%d function not supported", \
> + __func__, __LINE__);\
> + return retval;\
>
Need to keep the parens around retval in your macro
> + } \
> } while(0)
>
> -#defineFUNC_PTR_OR_RET(func) do { \
>
[dpdk-dev] DPDK and custom memory
On Fri, Aug 29, 2014 at 06:40:08PM +, Saygin, Artur wrote: > Hello DPDK experts, > > Imagine a PMD for an FPGA-based NIC that is limited to accessing certain > memory regions . Is there a way to make DPDK use that exact > memory? > > Perhaps this is more of a hugetlbfs question than DPDK but I thought I'd > start here. > Theres no real custom memory need there. What you need access to is covered by interface like vfio, which dpdk uses fairly regularly within its other hardware pmds. Neil > Sincerely, > Artur Saygin > >
[dpdk-dev] [PATCH v2 01/17] ixgbe: use the right debug macro
Hello Jay,
On Tue, Sep 2, 2014 at 3:43 PM, Jay Rolette wrote:
>
> On Mon, Sep 1, 2014 at 5:24 AM, David Marchand
> wrote:
>
>>
>> diff --git a/lib/librte_pmd_ixgbe/ixgbe_bypass.c
>> b/lib/librte_pmd_ixgbe/ixgbe_bypass.c
>> index 1d21dc0..1a980b8 100644
>> --- a/lib/librte_pmd_ixgbe/ixgbe_bypass.c
>> +++ b/lib/librte_pmd_ixgbe/ixgbe_bypass.c
>> @@ -40,20 +40,20 @@
>> #defineBYPASS_STATUS_OFF_MASK 3
>>
>> /* Macros to check for invlaid function pointers. */
>> -#defineFUNC_PTR_OR_ERR_RET(func, retval) do { \
>> - if ((func) == NULL) { \
>> - DEBUGOUT("%s:%d function not supported\n", \
>> - __func__, __LINE__); \
>> - return (retval); \
>> - } \
>> +#defineFUNC_PTR_OR_ERR_RET(func, retval) do { \
>> + if ((func) == NULL) { \
>> + PMD_DRV_LOG("%s:%d function not supported", \
>> + __func__, __LINE__);\
>> + return retval;\
>>
> Need to keep the parens around retval in your macro
>
Actually, checkpatch complained about this.
So I can keep the parenthesis, but then I don't want Thomas to tell me my
patch does not pass checkpatch :-)
> diff --git a/lib/librte_pmd_ixgbe/ixgbe_rxtx.c
>> b/lib/librte_pmd_ixgbe/ixgbe_rxtx.c
>> index dfc2076..46962bc 100644
>> --- a/lib/librte_pmd_ixgbe/ixgbe_rxtx.c
>> +++ b/lib/librte_pmd_ixgbe/ixgbe_rxtx.c
>> @@ -1765,33 +1765,36 @@ ixgbe_dev_tx_queue_setup(struct rte_eth_dev *dev,
>> tx_free_thresh = (uint16_t)((tx_conf->tx_free_thresh) ?
>> tx_conf->tx_free_thresh : DEFAULT_TX_FREE_THRESH);
>> if (tx_rs_thresh >= (nb_desc - 2)) {
>> - RTE_LOG(ERR, PMD, "tx_rs_thresh must be less than the
>> number "
>> - "of TX descriptors minus 2. (tx_rs_thresh=%u
>> port=%d "
>> - "queue=%d)\n", (unsigned int)tx_rs_thresh,
>> - (int)dev->data->port_id, (int)queue_idx);
>> + PMD_INIT_LOG(ERR, "tx_rs_thresh must be less than the
>> number "
>> +"of TX descriptors minus 2. (tx_rs_thresh=%u
>> "
>> +"port=%d queue=%d)\n", (unsigned
>> int)tx_rs_thresh,
>>
> Embedded newline on this log message. I noticed you have \n still on all
> of the PMD_INIT_LOG() calls following. Intended?
>
Yep intended, a patch after this removes all \n (idem for the other
comments).
Thanks.
--
David Marchand
[dpdk-dev] [PATCH v2 01/17] ixgbe: use the right debug macro
2014-09-02 16:16, David Marchand:
> >> /* Macros to check for invlaid function pointers. */
Invlaid is an invalid word ;)
> >> -#defineFUNC_PTR_OR_ERR_RET(func, retval) do { \
> >> - if ((func) == NULL) { \
> >> - DEBUGOUT("%s:%d function not supported\n", \
> >> - __func__, __LINE__); \
> >> - return (retval); \
> >> - } \
> >> +#defineFUNC_PTR_OR_ERR_RET(func, retval) do { \
> >> + if ((func) == NULL) { \
> >> + PMD_DRV_LOG("%s:%d function not supported", \
> >> + __func__, __LINE__);\
> >> + return retval;\
> >>
> > Need to keep the parens around retval in your macro
>
> Actually, checkpatch complained about this.
> So I can keep the parenthesis, but then I don't want Thomas to tell me my
> patch does not pass checkpatch :-)
You're right, I care about checkpatch :)
I don't see a case where parens are needed with return. Please give an example.
--
Thomas
[dpdk-dev] Wrong TCP checksum of packets sent by Linux guest (virtIO/vhost)
Hi Thomas, On 09/02/14 15:29, Thomas Monjalon wrote: > UDP/ICMP connectivity is fine, but TCP checksum of packet sent by the guest > are incorrect, as showed with tcpdump on linux-native. > Thanks for reporting. > Could you try virtio without vhost? Nope, dpdk-ovs supports only vhost. Thanks! Franck
[dpdk-dev] [PATCH v2 03/17] ixgbe: clean log messages
Looks good
On Mon, Sep 1, 2014 at 5:24 AM, David Marchand
wrote:
> Clean log messages:
> - remove leading \n in some messages,
> - remove trailing \n in some messages,
> - split multi lines messages,
> - replace some PMD_INIT_LOG(DEBUG, "some_func") with PMD_INIT_FUNC_TRACE().
>
> Signed-off-by: David Marchand
> ---
> lib/librte_pmd_ixgbe/ixgbe_ethdev.c | 150
> +--
> lib/librte_pmd_ixgbe/ixgbe_fdir.c |6 +-
> lib/librte_pmd_ixgbe/ixgbe_rxtx.c | 93 +++---
> 3 files changed, 124 insertions(+), 125 deletions(-)
>
> diff --git a/lib/librte_pmd_ixgbe/ixgbe_ethdev.c
> b/lib/librte_pmd_ixgbe/ixgbe_ethdev.c
> index a8a7ed6..1419494 100644
> --- a/lib/librte_pmd_ixgbe/ixgbe_ethdev.c
> +++ b/lib/librte_pmd_ixgbe/ixgbe_ethdev.c
> @@ -547,12 +547,12 @@ ixgbe_dev_queue_stats_mapping_set(struct rte_eth_dev
> *eth_dev,
> if ((hw->mac.type != ixgbe_mac_82599EB) && (hw->mac.type !=
> ixgbe_mac_X540))
> return -ENOSYS;
>
> - PMD_INIT_LOG(INFO, "Setting port %d, %s queue_id %d to stat index
> %d\n",
> + PMD_INIT_LOG(INFO, "Setting port %d, %s queue_id %d to stat index
> %d",
> (int)(eth_dev->data->port_id), is_rx ? "RX" : "TX",
> queue_id, stat_idx);
>
> n = (uint8_t)(queue_id / NB_QMAP_FIELDS_PER_QSM_REG);
> if (n >= IXGBE_NB_STAT_MAPPING_REGS) {
> - PMD_INIT_LOG(ERR, "Nb of stat mapping registers
> exceeded\n");
> + PMD_INIT_LOG(ERR, "Nb of stat mapping registers exceeded");
> return -EIO;
> }
> offset = (uint8_t)(queue_id % NB_QMAP_FIELDS_PER_QSM_REG);
> @@ -572,19 +572,20 @@ ixgbe_dev_queue_stats_mapping_set(struct rte_eth_dev
> *eth_dev,
> else
> stat_mappings->rqsmr[n] |= qsmr_mask;
>
> - PMD_INIT_LOG(INFO, "Set port %d, %s queue_id %d to stat index %d\n"
> -"%s[%d] = 0x%08x\n",
> -(int)(eth_dev->data->port_id), is_rx ? "RX" : "TX",
> queue_id, stat_idx,
> -is_rx ? "RQSMR" : "TQSM",n, is_rx ?
> stat_mappings->rqsmr[n] : stat_mappings->tqsm[n]);
> + PMD_INIT_LOG(INFO, "Set port %d, %s queue_id %d to stat index %d",
> +(int)(eth_dev->data->port_id), is_rx ? "RX" : "TX",
> +queue_id, stat_idx);
> + PMD_INIT_LOG(INFO, "%s[%d] = 0x%08x", is_rx ? "RQSMR" : "TQSM", n,
> +is_rx ? stat_mappings->rqsmr[n] :
> stat_mappings->tqsm[n]);
>
> /* Now write the mapping in the appropriate register */
> if (is_rx) {
> - PMD_INIT_LOG(INFO, "Write 0x%x to RX IXGBE stat mapping
> reg:%d\n",
> + PMD_INIT_LOG(INFO, "Write 0x%x to RX IXGBE stat mapping
> reg:%d",
> stat_mappings->rqsmr[n], n);
> IXGBE_WRITE_REG(hw, IXGBE_RQSMR(n),
> stat_mappings->rqsmr[n]);
> }
> else {
> - PMD_INIT_LOG(INFO, "Write 0x%x to TX IXGBE stat mapping
> reg:%d\n",
> + PMD_INIT_LOG(INFO, "Write 0x%x to TX IXGBE stat mapping
> reg:%d",
> stat_mappings->tqsm[n], n);
> IXGBE_WRITE_REG(hw, IXGBE_TQSM(n), stat_mappings->tqsm[n]);
> }
> @@ -790,12 +791,13 @@ eth_ixgbe_dev_init(__attribute__((unused)) struct
> eth_driver *eth_drv,
>
> if (diag == IXGBE_ERR_EEPROM_VERSION) {
> PMD_INIT_LOG(ERR, "This device is a pre-production
> adapter/"
> - "LOM. Please be aware there may be issues associated "
> - "with your hardware.\n If you are experiencing
> problems "
> + "LOM. Please be aware there may be issues associated "
> + "with your hardware.");
> + PMD_INIT_LOG(ERR, "If you are experiencing problems "
> "please contact your Intel or hardware representative "
> - "who provided you with this hardware.\n");
> + "who provided you with this hardware.");
> } else if (diag == IXGBE_ERR_SFP_NOT_SUPPORTED)
> - PMD_INIT_LOG(ERR, "Unsupported SFP+ Module\n");
> + PMD_INIT_LOG(ERR, "Unsupported SFP+ Module");
> if (diag) {
> PMD_INIT_LOG(ERR, "Hardware Initialization Failure: %d",
> diag);
> return -EIO;
> @@ -811,10 +813,9 @@ eth_ixgbe_dev_init(__attribute__((unused)) struct
> eth_driver *eth_drv,
> eth_dev->data->mac_addrs = rte_zmalloc("ixgbe", ETHER_ADDR_LEN *
> hw->mac.num_rar_entries, 0);
> if (eth_dev->data->mac_addrs == NULL) {
> - PMD_INIT_LOG(ERR,
> - "Failed to allocate %u bytes needed to store "
> - "MAC addresses",
> - ETHER_ADDR_LEN * hw->mac.num_rar_entries);
> + PMD_INIT_LOG(ERR, "Failed to allocate %u bytes needed to
> store "
> +"MAC
[dpdk-dev] [PATCH v2 01/17] ixgbe: use the right debug macro
On Tue, Sep 2, 2014 at 9:21 AM, Thomas Monjalon
wrote:
> >> -#defineFUNC_PTR_OR_ERR_RET(func, retval) do { \
> > >> - if ((func) == NULL) { \
> > >> - DEBUGOUT("%s:%d function not supported\n", \
> > >> - __func__, __LINE__); \
> > >> - return (retval); \
> > >> - } \
> > >> +#defineFUNC_PTR_OR_ERR_RET(func, retval) do { \
> > >> + if ((func) == NULL) { \
> > >> + PMD_DRV_LOG("%s:%d function not supported", \
> > >> + __func__, __LINE__);\
> > >> + return retval;\
> > >>
> > > Need to keep the parens around retval in your macro
> >
> > Actually, checkpatch complained about this.
> > So I can keep the parenthesis, but then I don't want Thomas to tell me my
> > patch does not pass checkpatch :-)
>
> You're right, I care about checkpatch :)
> I don't see a case where parens are needed with return. Please give an
> example.
Looking at it again, in this specific case you are correct. It is good
hygiene to always use parens around macro arguments, but this specific case
is ok without. It does add a small bit of danger if retval ends up getting
used elsewhere in the macro, but that's about it. Probably not worth
redoing the patch over that.
Jay
[dpdk-dev] [PATCH v2 0/6] link bonding
On Mon, 1 Sep 2014 09:31:23 +0100 Declan Doherty wrote: > This patch set adds support for link status interrupt in the link bonding > pmd. It also contains some patches to tidy up the code structure and to > of the link bonding code and to fix bugs relating to transmission > failures in the under lying slave pmd which could lead to leaked mbufs. Does this work when interfaces being bonded do not support link state like VF's and some other virtual devices (like vmxnet3?). I am concerned that the flag to say whether device support LSC or not is in drv_flags which is normally invariant. So either bond has to simulate link state if slave devices do not support (by polling) or provide another mechanism for application to determine availablity of link state.
[dpdk-dev] [PATCH v2 07/17] i40e: use the right debug macro
On Mon, Sep 1, 2014 at 5:24 AM, David Marchand
wrote:
> - Don't use DEBUGFUNC macro in non-shared code.
> - Don't use printf for logs.
> - We should avoid calling RTE_LOG directly as pmd provides a wrapper for
> logs.
>
> Signed-off-by: David Marchand
> ---
> lib/librte_pmd_i40e/i40e_ethdev.c| 146
> +-
> lib/librte_pmd_i40e/i40e_ethdev_vf.c |2 +-
> lib/librte_pmd_i40e/i40e_pf.c|6 +-
> lib/librte_pmd_i40e/i40e_rxtx.c | 64 +++
> 4 files changed, 110 insertions(+), 108 deletions(-)
>
> diff --git a/lib/librte_pmd_i40e/i40e_ethdev.c
> b/lib/librte_pmd_i40e/i40e_ethdev.c
> index 4e65ca4..352beb1 100644
> --- a/lib/librte_pmd_i40e/i40e_ethdev.c
> +++ b/lib/librte_pmd_i40e/i40e_ethdev.c
> @@ -1059,24 +1059,23 @@ i40e_update_vsi_stats(struct i40e_vsi *vsi)
> &oes->tx_errors, &nes->tx_errors);
> vsi->offset_loaded = true;
>
> -#ifdef RTE_LIBRTE_I40E_DEBUG_DRIVER
> - printf("* VSI[%u] stats start
> ***\n",
> -
> vsi->vsi_id);
> - printf("rx_bytes:%lu\n", nes->rx_bytes);
> - printf("rx_unicast: %lu\n", nes->rx_unicast);
> - printf("rx_multicast:%lu\n", nes->rx_multicast);
> - printf("rx_broadcast:%lu\n", nes->rx_broadcast);
> - printf("rx_discards: %lu\n", nes->rx_discards);
> - printf("rx_unknown_protocol: %lu\n", nes->rx_unknown_protocol);
> - printf("tx_bytes:%lu\n", nes->tx_bytes);
> - printf("tx_unicast: %lu\n", nes->tx_unicast);
> - printf("tx_multicast:%lu\n", nes->tx_multicast);
> - printf("tx_broadcast:%lu\n", nes->tx_broadcast);
> - printf("tx_discards: %lu\n", nes->tx_discards);
> - printf("tx_errors: %lu\n", nes->tx_errors);
> - printf("* VSI[%u] stats end ***\n",
> -
> vsi->vsi_id);
> -#endif /* RTE_LIBRTE_I40E_DEBUG_DRIVER */
> + PMD_DRV_LOG(DEBUG, "* VSI[%u] stats start
> ***\n",
> + vsi->vsi_id);
> + PMD_DRV_LOG(DEBUG, "rx_bytes:%lu\n", nes->rx_bytes);
> + PMD_DRV_LOG(DEBUG, "rx_unicast: %lu\n", nes->rx_unicast);
> + PMD_DRV_LOG(DEBUG, "rx_multicast:%lu\n",
> nes->rx_multicast);
> + PMD_DRV_LOG(DEBUG, "rx_broadcast:%lu\n",
> nes->rx_broadcast);
> + PMD_DRV_LOG(DEBUG, "rx_discards: %lu\n", nes->rx_discards);
> + PMD_DRV_LOG(DEBUG, "rx_unknown_protocol: %lu\n",
> + nes->rx_unknown_protocol);
> + PMD_DRV_LOG(DEBUG, "tx_bytes:%lu\n", nes->tx_bytes);
> + PMD_DRV_LOG(DEBUG, "tx_unicast: %lu\n", nes->tx_unicast);
> + PMD_DRV_LOG(DEBUG, "tx_multicast:%lu\n",
> nes->tx_multicast);
> + PMD_DRV_LOG(DEBUG, "tx_broadcast:%lu\n",
> nes->tx_broadcast);
> + PMD_DRV_LOG(DEBUG, "tx_discards: %lu\n", nes->tx_discards);
> + PMD_DRV_LOG(DEBUG, "tx_errors: %lu\n", nes->tx_errors);
> + PMD_DRV_LOG(DEBUG, "* VSI[%u] stats end
> ***\n",
> + vsi->vsi_id);
> }
>
> /* Get all statistics of a port */
> @@ -1277,69 +1276,74 @@ i40e_dev_stats_get(struct rte_eth_dev *dev, struct
> rte_eth_stats *stats)
> if (pf->main_vsi)
> i40e_update_vsi_stats(pf->main_vsi);
>
> -#ifdef RTE_LIBRTE_I40E_DEBUG_DRIVER
> - printf("* PF stats start ***\n");
> - printf("rx_bytes:%lu\n", ns->eth.rx_bytes);
> - printf("rx_unicast: %lu\n", ns->eth.rx_unicast);
> - printf("rx_multicast:%lu\n", ns->eth.rx_multicast);
> - printf("rx_broadcast:%lu\n", ns->eth.rx_broadcast);
> - printf("rx_discards: %lu\n", ns->eth.rx_discards);
> - printf("rx_unknown_protocol: %lu\n", ns->eth.rx_unknown_protocol);
> - printf("tx_bytes:%lu\n", ns->eth.tx_bytes);
> - printf("tx_unicast: %lu\n", ns->eth.tx_unicast);
> - printf("tx_multicast:%lu\n", ns->eth.tx_multicast);
> - printf("tx_broadcast:%lu\n", ns->eth.tx_broadcast);
> - printf("tx_discards: %lu\n", ns->eth.tx_discards);
> - printf("tx_errors: %lu\n", ns->eth.tx_errors);
> -
> - printf("tx_dropped_link_down: %lu\n",
> ns->tx_dropped_link_down);
> - printf("crc_errors: %lu\n", ns->crc_errors);
> - printf("illegal_bytes:%lu\n", ns->illegal_bytes);
> - printf("error_bytes: %lu\n", ns->error_bytes);
> - printf("mac_local_faults: %lu\n", ns->mac_local_faults);
> - printf("mac_remote_faults:%lu\n", ns->mac_remote_faults);
> - printf("rx_length_errors: %lu\n", ns->rx_length_errors);
> - printf("link_xon_rx: %lu\n",
[dpdk-dev] [PATCH v2 12/17] e1000: use the right debug macro
On Mon, Sep 1, 2014 at 5:24 AM, David Marchand
wrote:
> - We should not use DEBUGOUT* / DEBUGFUNC macros in non-shared code.
> These macros come as compat wrappers for shared code.
> - We should avoid calling RTE_LOG directly as pmd provides a wrapper for
> logs.
>
> Signed-off-by: David Marchand
> ---
> lib/librte_pmd_e1000/em_rxtx.c| 32 ++--
> lib/librte_pmd_e1000/igb_ethdev.c |9 +
> lib/librte_pmd_e1000/igb_pf.c |5 +++--
> lib/librte_pmd_e1000/igb_rxtx.c | 16 +++-
> 4 files changed, 33 insertions(+), 29 deletions(-)
>
> diff --git a/lib/librte_pmd_e1000/em_rxtx.c
> b/lib/librte_pmd_e1000/em_rxtx.c
> index f254858..01efa50 100644
> --- a/lib/librte_pmd_e1000/em_rxtx.c
> +++ b/lib/librte_pmd_e1000/em_rxtx.c
> @@ -1211,18 +1211,21 @@ eth_em_tx_queue_setup(struct rte_eth_dev *dev,
> DEFAULT_TX_RS_THRESH);
>
> if (tx_free_thresh >= (nb_desc - 3)) {
> - RTE_LOG(ERR, PMD, "tx_free_thresh must be less than the "
> - "number of TX descriptors minus 3.
> (tx_free_thresh=%u "
> - "port=%d queue=%d)\n", (unsigned
> int)tx_free_thresh,
> - (int)dev->data->port_id, (int)queue_idx);
> + PMD_INIT_LOG(ERR, "tx_free_thresh must be less than the "
> +"number of TX descriptors minus 3. "
> +"(tx_free_thresh=%u port=%d queue=%d)\n",
> +(unsigned int)tx_free_thresh,
> +(int)dev->data->port_id, (int)queue_idx);
> return -(EINVAL);
> }
> if (tx_rs_thresh > tx_free_thresh) {
> - RTE_LOG(ERR, PMD, "tx_rs_thresh must be less than or equal
> to "
> - "tx_free_thresh. (tx_free_thresh=%u
> tx_rs_thresh=%u "
> - "port=%d queue=%d)\n", (unsigned
> int)tx_free_thresh,
> - (unsigned int)tx_rs_thresh,
> (int)dev->data->port_id,
> - (int)queue_idx);
> + PMD_INIT_LOG(ERR, "tx_rs_thresh must be less than or equal
> to "
> +"tx_free_thresh. (tx_free_thresh=%u "
> +"tx_rs_thresh=%u port=%d queue=%d)\n",
> +(unsigned int)tx_free_thresh,
> +(unsigned int)tx_rs_thresh,
> +(int)dev->data->port_id,
> +(int)queue_idx);
> return -(EINVAL);
> }
>
> @@ -1233,10 +1236,10 @@ eth_em_tx_queue_setup(struct rte_eth_dev *dev,
> * accumulates WTHRESH descriptors.
> */
> if (tx_conf->tx_thresh.wthresh != 0 && tx_rs_thresh != 1) {
> - RTE_LOG(ERR, PMD, "TX WTHRESH must be set to 0 if "
> - "tx_rs_thresh is greater than 1. (tx_rs_thresh=%u "
> - "port=%d queue=%d)\n", (unsigned int)tx_rs_thresh,
> - (int)dev->data->port_id, (int)queue_idx);
> + PMD_INIT_LOG(ERR, "TX WTHRESH must be set to 0 if "
> +"tx_rs_thresh is greater than 1.
> (tx_rs_thresh=%u "
> +"port=%d queue=%d)\n", (unsigned
> int)tx_rs_thresh,
> +(int)dev->data->port_id, (int)queue_idx);
> return -(EINVAL);
> }
>
> @@ -1366,7 +1369,8 @@ eth_em_rx_queue_setup(struct rte_eth_dev *dev,
> * EM devices don't support drop_en functionality
> */
> if (rx_conf->rx_drop_en) {
> - RTE_LOG(ERR, PMD, "drop_en functionality not supported by
> device\n");
> + PMD_INIT_LOG(ERR, "drop_en functionality not supported by "
> +"device\n");
> return (-EINVAL);
> }
>
> diff --git a/lib/librte_pmd_e1000/igb_ethdev.c
> b/lib/librte_pmd_e1000/igb_ethdev.c
> index 3187d92..b45eb24 100644
> --- a/lib/librte_pmd_e1000/igb_ethdev.c
> +++ b/lib/librte_pmd_e1000/igb_ethdev.c
> @@ -400,7 +400,7 @@ igb_reset_swfw_lock(struct e1000_hw *hw)
> * So force the release of the faulty lock.
> */
> if (e1000_get_hw_semaphore_generic(hw) < 0) {
> - DEBUGOUT("SMBI lock released");
> + PMD_DRV_LOG(DEBUG, "SMBI lock released");
> }
> e1000_put_hw_semaphore_generic(hw);
>
> @@ -416,7 +416,8 @@ igb_reset_swfw_lock(struct e1000_hw *hw)
> if (hw->bus.func > E1000_FUNC_1)
> mask <<= 2;
> if (hw->mac.ops.acquire_swfw_sync(hw, mask) < 0) {
> - DEBUGOUT1("SWFW phy%d lock released",
> hw->bus.func);
> + PMD_DRV_LOG(DEBUG, "SWFW phy%d lock released",
> + hw->bus.func);
> }
> hw->mac.ops.rel
[dpdk-dev] [PATCH v2 14/17] e1000: clean log messages
Couple of minor nits noted. The bigger feedback would be that if you could
keep the "whitespace / line-wrapping / formatting" only changes in separate
patches, these reviews would be easier to get through.
Jay
On Mon, Sep 1, 2014 at 5:24 AM, David Marchand
wrote:
> Clean log messages:
> - remove leading \n in some messages,
> - remove trailing \n in some messages,
> - split multi lines messages,
> - replace some PMD_INIT_LOG(DEBUG, "some_func") with PMD_INIT_FUNC_TRACE().
>
> Signed-off-by: David Marchand
> ---
> lib/librte_pmd_e1000/e1000_logs.h |4 +-
> lib/librte_pmd_e1000/em_ethdev.c | 64 ++
> lib/librte_pmd_e1000/em_rxtx.c| 109
> ++---
> lib/librte_pmd_e1000/igb_ethdev.c | 91 +++
> lib/librte_pmd_e1000/igb_pf.c |4 +-
> lib/librte_pmd_e1000/igb_rxtx.c | 45 +++
> 6 files changed, 153 insertions(+), 164 deletions(-)
>
> diff --git a/lib/librte_pmd_e1000/e1000_logs.h
> b/lib/librte_pmd_e1000/e1000_logs.h
> index fe6e023..4dd7208 100644
> --- a/lib/librte_pmd_e1000/e1000_logs.h
> +++ b/lib/librte_pmd_e1000/e1000_logs.h
> @@ -37,8 +37,10 @@
> #ifdef RTE_LIBRTE_E1000_DEBUG_INIT
> #define PMD_INIT_LOG(level, fmt, args...) \
> RTE_LOG(level, PMD, "%s(): " fmt "\n", __func__, ## args)
> +#define PMD_INIT_FUNC_TRACE() PMD_INIT_LOG(DEBUG, " >>")
> #else
> -#define PMD_INIT_LOG(level, fmt, args...) do { } while(0)
> +#define PMD_INIT_LOG(level, fmt, args...) do { } while (0)
> +#define PMD_INIT_FUNC_TRACE() do { } while (0)
> #endif
>
> #ifdef RTE_LIBRTE_E1000_DEBUG_RX
> diff --git a/lib/librte_pmd_e1000/em_ethdev.c
> b/lib/librte_pmd_e1000/em_ethdev.c
> index 4555294..fd36b37 100644
> --- a/lib/librte_pmd_e1000/em_ethdev.c
> +++ b/lib/librte_pmd_e1000/em_ethdev.c
> @@ -249,9 +249,9 @@ eth_em_dev_init(__attribute__((unused)) struct
> eth_driver *eth_drv,
> if (e1000_setup_init_funcs(hw, TRUE) != E1000_SUCCESS ||
> em_hw_init(hw) != 0) {
> PMD_INIT_LOG(ERR, "port_id %d vendorID=0x%x deviceID=0x%x:
> "
> - "failed to init HW",
> - eth_dev->data->port_id, pci_dev->id.vendor_id,
> - pci_dev->id.device_id);
> +"failed to init HW",
> +eth_dev->data->port_id, pci_dev->id.vendor_id,
> +pci_dev->id.device_id);
> return -(ENODEV);
> }
>
> @@ -260,8 +260,8 @@ eth_em_dev_init(__attribute__((unused)) struct
> eth_driver *eth_drv,
> hw->mac.rar_entry_count, 0);
> if (eth_dev->data->mac_addrs == NULL) {
> PMD_INIT_LOG(ERR, "Failed to allocate %d bytes needed to "
> - "store MAC addresses",
> - ETHER_ADDR_LEN * hw->mac.rar_entry_count);
> +"store MAC addresses",
> +ETHER_ADDR_LEN * hw->mac.rar_entry_count);
> return -(ENOMEM);
> }
>
> @@ -272,9 +272,9 @@ eth_em_dev_init(__attribute__((unused)) struct
> eth_driver *eth_drv,
> /* initialize the vfta */
> memset(shadow_vfta, 0, sizeof(*shadow_vfta));
>
> - PMD_INIT_LOG(INFO, "port_id %d vendorID=0x%x deviceID=0x%x\n",
> - eth_dev->data->port_id, pci_dev->id.vendor_id,
> - pci_dev->id.device_id);
> + PMD_INIT_LOG(INFO, "port_id %d vendorID=0x%x deviceID=0x%x",
> +eth_dev->data->port_id, pci_dev->id.vendor_id,
> +pci_dev->id.device_id);
>
> rte_intr_callback_register(&(pci_dev->intr_handle),
> eth_em_interrupt_handler, (void *)eth_dev);
> @@ -306,17 +306,17 @@ em_hw_init(struct e1000_hw *hw)
>
> diag = hw->mac.ops.init_params(hw);
> if (diag != 0) {
> - PMD_INIT_LOG(ERR, "MAC Initialization Error\n");
> + PMD_INIT_LOG(ERR, "MAC Initialization Error");
> return diag;
> }
> diag = hw->nvm.ops.init_params(hw);
> if (diag != 0) {
> - PMD_INIT_LOG(ERR, "NVM Initialization Error\n");
> + PMD_INIT_LOG(ERR, "NVM Initialization Error");
> return diag;
> }
> diag = hw->phy.ops.init_params(hw);
> if (diag != 0) {
> - PMD_INIT_LOG(ERR, "PHY Initialization Error\n");
> + PMD_INIT_LOG(ERR, "PHY Initialization Error");
> return diag;
> }
> (void) e1000_get_bus_info(hw);
> @@ -375,7 +375,7 @@ em_hw_init(struct e1000_hw *hw)
> diag = e1000_check_reset_block(hw);
> if (diag < 0) {
> PMD_INIT_LOG(ERR, "PHY reset is blocked due to "
> - "SOL/IDER session");
> +"SOL/IDER session");
> }
> return (0);
>
> @@ -390,11 +390,10 @@ eth_em_configure
[dpdk-dev] [PATCH v2 14/17] e1000: clean log messages
On Tue, Sep 2, 2014 at 9:05 PM, Jay Rolette wrote:
> Couple of minor nits noted. The bigger feedback would be that if you could
> keep the "whitespace / line-wrapping / formatting" only changes in separate
> patches, these reviews would be easier to get through.
>
Hum, indeed, I should have done this.
Will keep in mind.
/* Free memory prior to re-allocation if needed */
>> if (dev->data->tx_queues[queue_idx] != NULL) {
>> @@ -1271,7 +1271,7 @@ eth_igb_tx_queue_setup(struct rte_eth_dev *dev,
>> igb_tx_queue_release(txq);
>> return (-ENOMEM);
>> }
>> - PMD_INIT_LOG(DEBUG, "sw_ring=%p hw_ring=%p
>> dma_addr=0x%"PRIx64"\n",
>> + PMD_INIT_LOG(DEBUG, "sw_ring=%p hw_ring=%p dma_addr=0x%"PRIx64"",
>>
> The empty string after PRIx64 is extraneous
>
>
>> txq->sw_ring, txq->tx_ring, txq->tx_ring_phys_addr);
>>
>> igb_reset_tx_queue(txq, dev);
>> @@ -1409,7 +1409,7 @@ eth_igb_rx_queue_setup(struct rte_eth_dev *dev,
>> igb_rx_queue_release(rxq);
>> return (-ENOMEM);
>> }
>> - PMD_INIT_LOG(DEBUG, "sw_ring=%p hw_ring=%p
>> dma_addr=0x%"PRIx64"\n",
>> + PMD_INIT_LOG(DEBUG, "sw_ring=%p hw_ring=%p dma_addr=0x%"PRIx64"",
>>
> The empty string after PRIx64 is extraneous
>
>
Ok, I will update.
Thanks.
--
David Marchand
[dpdk-dev] [PATCH v2 16/17] e1000: add a message when forcing scatter mode
Self nack, forgot to remove the \n ...
--
David Marchand
On Mon, Sep 1, 2014 at 12:24 PM, David Marchand
wrote:
> Signed-off-by: David Marchand
> ---
> lib/librte_pmd_e1000/em_rxtx.c |4
> lib/librte_pmd_e1000/igb_rxtx.c | 14 ++
> 2 files changed, 18 insertions(+)
>
> diff --git a/lib/librte_pmd_e1000/em_rxtx.c
> b/lib/librte_pmd_e1000/em_rxtx.c
> index 83ecb33..a6cea32 100644
> --- a/lib/librte_pmd_e1000/em_rxtx.c
> +++ b/lib/librte_pmd_e1000/em_rxtx.c
> @@ -1707,6 +1707,8 @@ eth_em_rx_init(struct rte_eth_dev *dev)
> */
> if (dev->data->dev_conf.rxmode.jumbo_frame ||
> rctl_bsize < ETHER_MAX_LEN) {
> + if (!dev->data->scattered_rx)
> + PMD_INIT_LOG(DEBUG, "forcing scatter
> mode\n");
> dev->rx_pkt_burst =
> (eth_rx_burst_t)eth_em_recv_scattered_pkts;
> dev->data->scattered_rx = 1;
> @@ -1714,6 +1716,8 @@ eth_em_rx_init(struct rte_eth_dev *dev)
> }
>
> if (dev->data->dev_conf.rxmode.enable_scatter) {
> + if (!dev->data->scattered_rx)
> + PMD_INIT_LOG(DEBUG, "forcing scatter mode\n");
> dev->rx_pkt_burst = eth_em_recv_scattered_pkts;
> dev->data->scattered_rx = 1;
> }
> diff --git a/lib/librte_pmd_e1000/igb_rxtx.c
> b/lib/librte_pmd_e1000/igb_rxtx.c
> index 5ca06c9..1d8f74a 100644
> --- a/lib/librte_pmd_e1000/igb_rxtx.c
> +++ b/lib/librte_pmd_e1000/igb_rxtx.c
> @@ -1978,6 +1978,9 @@ eth_igb_rx_init(struct rte_eth_dev *dev)
> /* It adds dual VLAN length for supporting dual
> VLAN */
> if ((dev->data->dev_conf.rxmode.max_rx_pkt_len +
> 2 * VLAN_TAG_SIZE) >
> buf_size){
> + if (!dev->data->scattered_rx)
> + PMD_INIT_LOG(DEBUG,
> +"forcing scatter
> mode\n");
> dev->rx_pkt_burst =
> eth_igb_recv_scattered_pkts;
> dev->data->scattered_rx = 1;
> }
> @@ -1987,6 +1990,8 @@ eth_igb_rx_init(struct rte_eth_dev *dev)
> */
> if ((rctl_bsize == 0) || (rctl_bsize > buf_size))
> rctl_bsize = buf_size;
> + if (!dev->data->scattered_rx)
> + PMD_INIT_LOG(DEBUG, "forcing scatter
> mode\n");
> dev->rx_pkt_burst = eth_igb_recv_scattered_pkts;
> dev->data->scattered_rx = 1;
> }
> @@ -2008,6 +2013,8 @@ eth_igb_rx_init(struct rte_eth_dev *dev)
> }
>
> if (dev->data->dev_conf.rxmode.enable_scatter) {
> + if (!dev->data->scattered_rx)
> + PMD_INIT_LOG(DEBUG, "forcing scatter mode\n");
> dev->rx_pkt_burst = eth_igb_recv_scattered_pkts;
> dev->data->scattered_rx = 1;
> }
> @@ -2242,6 +2249,9 @@ eth_igbvf_rx_init(struct rte_eth_dev *dev)
> /* It adds dual VLAN length for supporting dual
> VLAN */
> if ((dev->data->dev_conf.rxmode.max_rx_pkt_len +
> 2 * VLAN_TAG_SIZE) >
> buf_size){
> + if (!dev->data->scattered_rx)
> + PMD_INIT_LOG(DEBUG,
> +"forcing scatter
> mode\n");
> dev->rx_pkt_burst =
> eth_igb_recv_scattered_pkts;
> dev->data->scattered_rx = 1;
> }
> @@ -2251,6 +2261,8 @@ eth_igbvf_rx_init(struct rte_eth_dev *dev)
> */
> if ((rctl_bsize == 0) || (rctl_bsize > buf_size))
> rctl_bsize = buf_size;
> + if (!dev->data->scattered_rx)
> + PMD_INIT_LOG(DEBUG, "forcing scatter
> mode\n");
> dev->rx_pkt_burst = eth_igb_recv_scattered_pkts;
> dev->data->scattered_rx = 1;
> }
> @@ -2282,6 +2294,8 @@ eth_igbvf_rx_init(struct rte_eth_dev *dev)
> }
>
> if (dev->data->dev_conf.rxmode.enable_scatter) {
> + if (!dev->data->scattered_rx)
> + PMD_INIT_LOG(DEBUG, "forcing scatter mode\n");
> dev->rx_pkt_burst = eth_igb_recv_scattered_pkts;
> dev->data->scattered_rx = 1;
> }
> --
> 1.7.10.4
>
>
[dpdk-dev] [PATCH v2 09/17] i40e: clean log messages
On Mon, Sep 1, 2014 at 5:24 AM, David Marchand
wrote:
> Clean log messages:
> - remove leading \n in some messages,
> - remove trailing \n in some messages,
> - split multi lines messages,
> - replace some PMD_INIT_LOG(DEBUG, "some_func") with PMD_INIT_FUNC_TRACE().
>
> Signed-off-by: David Marchand
> ---
> lib/librte_pmd_i40e/i40e_ethdev.c| 418
> +-
> lib/librte_pmd_i40e/i40e_ethdev_vf.c | 166 +++---
> lib/librte_pmd_i40e/i40e_pf.c| 75 +++---
> lib/librte_pmd_i40e/i40e_rxtx.c | 118 +-
> 4 files changed, 385 insertions(+), 392 deletions(-)
>
> diff --git a/lib/librte_pmd_i40e/i40e_ethdev.c
> b/lib/librte_pmd_i40e/i40e_ethdev.c
> index 352beb1..1a5b55d 100644
> --- a/lib/librte_pmd_i40e/i40e_ethdev.c
> +++ b/lib/librte_pmd_i40e/i40e_ethdev.c
> @@ -371,7 +371,7 @@ eth_i40e_dev_init(__rte_unused struct eth_driver
> *eth_drv,
> hw->hw_addr = (uint8_t *)(pci_dev->mem_resource[0].addr);
> if (!hw->hw_addr) {
> PMD_INIT_LOG(ERR, "Hardware is not available, "
> - "as address is NULL\n");
> +"as address is NULL");
> return -ENODEV;
> }
>
> @@ -395,7 +395,8 @@ eth_i40e_dev_init(__rte_unused struct eth_driver
> *eth_drv,
> /* Initialize the shared code (base driver) */
> ret = i40e_init_shared_code(hw);
> if (ret) {
> - PMD_INIT_LOG(ERR, "Failed to init shared code (base
> driver): %d", ret);
> + PMD_INIT_LOG(ERR, "Failed to init shared code (base
> driver):"
> +"%d", ret);
> return ret;
> }
>
> @@ -406,8 +407,7 @@ eth_i40e_dev_init(__rte_unused struct eth_driver
> *eth_drv,
> PMD_INIT_LOG(ERR, "Failed to init adminq: %d", ret);
> return -EIO;
> }
> - PMD_INIT_LOG(INFO, "FW %d.%d API %d.%d NVM "
> - "%02d.%02d.%02d eetrack %04x\n",
> + PMD_INIT_LOG(INFO, "FW %d.%d API %d.%d NVM %02d.%02d.%02d eetrack
> %04x",
> hw->aq.fw_maj_ver, hw->aq.fw_min_ver,
> hw->aq.api_maj_ver, hw->aq.api_min_ver,
> ((hw->nvm.version >> 12) & 0xf),
> @@ -417,7 +417,7 @@ eth_i40e_dev_init(__rte_unused struct eth_driver
> *eth_drv,
> /* Disable LLDP */
> ret = i40e_aq_stop_lldp(hw, true, NULL);
> if (ret != I40E_SUCCESS) /* Its failure can be ignored */
> - PMD_INIT_LOG(INFO, "Failed to stop lldp\n");
> + PMD_INIT_LOG(INFO, "Failed to stop lldp");
>
> /* Clear PXE mode */
> i40e_clear_pxe_mode(hw);
> @@ -439,13 +439,13 @@ eth_i40e_dev_init(__rte_unused struct eth_driver
> *eth_drv,
> /* Initialize the queue management */
> ret = i40e_res_pool_init(&pf->qp_pool, 0, hw->func_caps.num_tx_qp);
> if (ret < 0) {
> - PMD_INIT_LOG(ERR, "Failed to init queue pool\n");
> + PMD_INIT_LOG(ERR, "Failed to init queue pool");
> goto err_qp_pool_init;
> }
> ret = i40e_res_pool_init(&pf->msix_pool, 1,
> hw->func_caps.num_msix_vectors - 1);
> if (ret < 0) {
> - PMD_INIT_LOG(ERR, "Failed to init MSIX pool\n");
> + PMD_INIT_LOG(ERR, "Failed to init MSIX pool");
> goto err_msix_pool_init;
> }
>
> @@ -499,8 +499,8 @@ eth_i40e_dev_init(__rte_unused struct eth_driver
> *eth_drv,
> /* Should be after VSI initialized */
> dev->data->mac_addrs = rte_zmalloc("i40e", len, 0);
> if (!dev->data->mac_addrs) {
> - PMD_INIT_LOG(ERR, "Failed to allocated memory "
> - "for storing mac address");
> + PMD_INIT_LOG(ERR, "Failed to allocated memory for storing "
> +"mac address");
> goto err_get_mac_addr;
> }
> ether_addr_copy((struct ether_addr *)hw->mac.perm_addr,
> @@ -723,9 +723,9 @@ i40e_phy_conf_link(struct i40e_hw *hw, uint8_t
> abilities, uint8_t force_speed)
> phy_conf.eeer = phy_ab.eeer_val;
> phy_conf.low_power_ctrl = phy_ab.d3_lpan;
>
> - PMD_DRV_LOG(DEBUG, "\n\tCurrent: abilities %x, link_speed %x\n"
> - "\tConfig: abilities %x, link_speed %x",
> - phy_ab.abilities, phy_ab.link_speed,
> + PMD_DRV_LOG(DEBUG, "\tCurrent: abilities %x, link_speed %x",
> + phy_ab.abilities, phy_ab.link_speed);
> + PMD_DRV_LOG(DEBUG, "\tConfig: abilities %x, link_speed %x",
> phy_conf.abilities, phy_conf.link_speed);
>
> status = i40e_aq_set_phy_config(hw, &phy_conf, NULL);
> @@ -763,7 +763,7 @@ i40e_dev_start(struct rte_eth_dev *dev)
>
> if ((dev->data->dev_conf.link_duplex != ETH_LINK_AUTONEG_DUPLEX) &&
> (dev->data->dev_co
[dpdk-dev] [PATCHv5] librte_acl make it build/work for 'default' target
On Wed, Sep 03, 2014 at 03:29:16AM +0200, Thomas Monjalon wrote: > > > Make ACL library to build/work on 'default' architecture: Upon reading all the steps taken to implement the new multi-arch version of the code, I had a funny feeling where each time I asked, "But... what if X???" the next step addressed X, until I had gone through all my questions. Now I am looking forward to using that code in the future. Matthew.
[dpdk-dev] [PATCH 2/3] lib/librte_vhost: vhost library support to facilitate integration with DPDK accelerated vswitch
On Wed, 3 Sep 2014 05:39:24 + "Xie, Huawei" wrote: > Thanks Tetsuya: > Some of them are due to 80 character limitation. Is it ok to break the > limitation for better indentation? Many of these comments could just be removed or shortened. Often removing comments makes code clearer and less buggy, because there is no risk that comment does not match code.
