Skip to contentSkip to conversation

drivers/net:new PMD using tun/tap host interface

PATCH Superseded Keith Wiles 10 participants 59 messages Sep 15, 2016 – Jan 20, 2017 Network drivers

lore
Share & tools
lore
#0 lore raw

PATCH 1/3 drivers/net:new PMD using tun/tap host interface

The rte_eth_tap.c PMD creates a device using TUN/TAP interfaces on the local host. The PMD allows for DPDK and the host to communicate using a raw device interface on the host and in the DPDK application. The device created is a Tap device with a L2 packet header.

---
 drivers/net/tap/Makefile                |  60 +++
 drivers/net/tap/rte_eth_tap.c           | 872 ++++++++++++++++++++++++++++++++
 drivers/net/tap/rte_pmd_tap_version.map |   4 +
 3 files changed, 936 insertions(+)
 create mode 100644 drivers/net/tap/Makefile
 create mode 100644 drivers/net/tap/rte_eth_tap.c
 create mode 100644 drivers/net/tap/rte_pmd_tap_version.map
drivers/net/tap/Makefile new file +60 -0
new file mode 100644
index 0000000..442a2fe
@@ -0,0 +1,60 @@
+#   BSD LICENSE
+#
+#   Copyright(c) 2014 John W. Linville <[email protected]>
+#   Copyright(c) 2010-2014 Intel Corporation. All rights reserved.
+#   Copyright(c) 2014 6WIND S.A.
+#   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_pmd_tap.a
+
+EXPORT_MAP := rte_pmd_tap_version.map
+
+LIBABIVER := 1
+
+CFLAGS += -O3
+CFLAGS += $(WERROR_FLAGS)
+
+#
+# all source are stored in SRCS-y
+#
+SRCS-$(CONFIG_RTE_LIBRTE_PMD_TAP) += rte_eth_tap.c
+
+# this lib depends upon:
+DEPDIRS-$(CONFIG_RTE_LIBRTE_PMD_TAP) += lib/librte_eal
+DEPDIRS-$(CONFIG_RTE_LIBRTE_PMD_TAP) += lib/librte_mbuf
+DEPDIRS-$(CONFIG_RTE_LIBRTE_PMD_TAP) += lib/librte_mempool
+DEPDIRS-$(CONFIG_RTE_LIBRTE_PMD_TAP) += lib/librte_ether
+DEPDIRS-$(CONFIG_RTE_LIBRTE_PMD_TAP) += lib/librte_kvargs
+
+include $(RTE_SDK)/mk/rte.lib.mk
drivers/net/tap/rte_eth_tap.c new file +872 -0

872 lines not on this page · raw message

drivers/net/tap/rte_pmd_tap_version.map new file +4 -0
new file mode 100644
index 0000000..61463bf
@@ -0,0 +1,4 @@
+DPDK_16.11 {
+
+	local: *;
+};
Signature · 1 line
2.8.0.GIT
#1 lore raw

PATCH 2/3 docs:tun/tap PMD information

---
 doc/guides/nics/tap.rst | 84 +++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 84 insertions(+)
 create mode 100644 doc/guides/nics/tap.rst
doc/guides/nics/tap.rst new file +84 -0
new file mode 100644
index 0000000..072def8
@@ -0,0 +1,84 @@
+..  BSD LICENSE
+    Copyright(c) 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.
+
+Tun/Tap Poll Mode Driver
+========================================
+
+The rte_eth_tap.c PMD creates a device using TUN/TAP interfaces on the local
+host. The PMD allows for DPDK and the host to communicate using a raw device
+interface on the host and in the DPDK application.
+
+The device created is a TAP device, which sends/receives packet in a raw format
+with a L2 header. The usage for a TAP PMD is for connectivity to the local host
+using a TAP interface. When the TAP PMD is initialized it will create a number
+of tap devices in the host accessed via 'ifconfig -a' or 'ip' command. The
+commands can be used to assign and query the virtual like device.
+
+These TAP interfaces can be used with wireshark or tcpdump or Pktgen-DPDK along
+with being able to be used as a network connection to the DPDK application. The
+method enable one or more interfaces is to use the --vdev=eth_tap option on the
+DPDK application  command line. Each --vdev=eth_tap option give will create an
+interface named dtap0, dtap1, ... and so forth.
+
+.. code-block:: console
+
+   The interfaced name can be changed by adding the iface=foo0
+   e.g. --vedv=eth_tap,iface=foo0 --vdev=eth_tap,iface=foo1, ...
+
+.. code-block:: console
+
+   Also the speed of the interface can be changed from 10G to whatever number
+   needed, but the interface does not enforce that speed.
+   e.g. --vdev=eth_tap,iface=foo0,speed=25000
+
+After the DPDK application is started you can send and receive packets on the
+interface using the standard rx_burst/tx_burst APIs in DPDK. From the host point
+of view you can use any host tool like tcpdump, wireshark, ping, Pktgen and
+others to communicate with the DPDK application. The DPDK application may not
+understand network protocols like IPv4/6, UDP or TCP unless the application has
+been written to understand these protocols.
+
+If you need the interface as a real network interface meaning running and has
+a valid IP address then you can do this with the following commands:
+
+.. code-block:: console
+
+   sudo ip link set dtap0 up; sudo ip addr add 192.168.0.250/24 dev dtap0
+   sudo ip link set dtap1 up; sudo ip addr add 192.168.1.250/24 dev dtap1
+
+Please change the IP addresses as you see fit.
+
+If routing is enabled on the host you can also communicate with the DPDK App
+over the internet via a standard socket layer application as long as you account
+for the protocol handing in the application.
+
+If you have a Network Stack in your DPDK application or something like it you
+can utilize that stack to handle the network protocols. Plus you would be able
+to address the interface using an IP address assigned to the internal interface.
Signature · 1 line
2.8.0.GIT
#2 lore raw

Re: [PATCH 2/3] docs:tun/tap PMD information

self Nak - just noticed the copyright notices are wrong.

Regards,
Keith

Quoted 101 lines · Keith Wiles

---
doc/guides/nics/tap.rst | 84 +++++++++++++++++++++++++++++++++++++++++++++++++

1 file changed, 84 insertions(+)

create mode 100644 doc/guides/nics/tap.rst

doc/guides/nics/tap.rst new file +84 -0
new file mode 100644
index 0000000..072def8
@@ -0,0 +1,84 @@
+..  BSD LICENSE
+    Copyright(c) 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.
+
+Tun/Tap Poll Mode Driver
+========================================
+
+The rte_eth_tap.c PMD creates a device using TUN/TAP interfaces on the local
+host. The PMD allows for DPDK and the host to communicate using a raw device
+interface on the host and in the DPDK application.
+
+The device created is a TAP device, which sends/receives packet in a raw format
+with a L2 header. The usage for a TAP PMD is for connectivity to the local host
+using a TAP interface. When the TAP PMD is initialized it will create a number
+of tap devices in the host accessed via 'ifconfig -a' or 'ip' command. The
+commands can be used to assign and query the virtual like device.
+
+These TAP interfaces can be used with wireshark or tcpdump or Pktgen-DPDK along
+with being able to be used as a network connection to the DPDK application. The
+method enable one or more interfaces is to use the --vdev=eth_tap option on the
+DPDK application  command line. Each --vdev=eth_tap option give will create an
+interface named dtap0, dtap1, ... and so forth.
+
+.. code-block:: console
+
+   The interfaced name can be changed by adding the iface=foo0
+   e.g. --vedv=eth_tap,iface=foo0 --vdev=eth_tap,iface=foo1, ...
+
+.. code-block:: console
+
+   Also the speed of the interface can be changed from 10G to whatever number
+   needed, but the interface does not enforce that speed.
+   e.g. --vdev=eth_tap,iface=foo0,speed=25000
+
+After the DPDK application is started you can send and receive packets on the
+interface using the standard rx_burst/tx_burst APIs in DPDK. From the host point
+of view you can use any host tool like tcpdump, wireshark, ping, Pktgen and
+others to communicate with the DPDK application. The DPDK application may not
+understand network protocols like IPv4/6, UDP or TCP unless the application has
+been written to understand these protocols.
+
+If you need the interface as a real network interface meaning running and has
+a valid IP address then you can do this with the following commands:
+
+.. code-block:: console
+
+   sudo ip link set dtap0 up; sudo ip addr add 192.168.0.250/24 dev dtap0
+   sudo ip link set dtap1 up; sudo ip addr add 192.168.1.250/24 dev dtap1
+
+Please change the IP addresses as you see fit.
+
+If routing is enabled on the host you can also communicate with the DPDK App
+over the internet via a standard socket layer application as long as you account
+for the protocol handing in the application.
+
+If you have a Network Stack in your DPDK application or something like it you
+can utilize that stack to handle the network protocols. Plus you would be able
+to address the interface using an IP address assigned to the internal interface.
Signature · 1 line
2.8.0.GIT
#3 lore raw

Re: [PATCH 2/3] docs:tun/tap PMD information

Regards,
Keith

Quoted 3 lines · Keith

self Nak - just noticed the copyright notices are wrong.

My mistake I was looking at the wrong file the headers appear correct.

Quoted 106 lines

Regards,
Keith

Quoted 101 lines · Keith Wiles

---
doc/guides/nics/tap.rst | 84 +++++++++++++++++++++++++++++++++++++++++++++++++

1 file changed, 84 insertions(+)

create mode 100644 doc/guides/nics/tap.rst

doc/guides/nics/tap.rst new file +84 -0
new file mode 100644
index 0000000..072def8
@@ -0,0 +1,84 @@
+..  BSD LICENSE
+    Copyright(c) 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.
+
+Tun/Tap Poll Mode Driver
+========================================
+
+The rte_eth_tap.c PMD creates a device using TUN/TAP interfaces on the local
+host. The PMD allows for DPDK and the host to communicate using a raw device
+interface on the host and in the DPDK application.
+
+The device created is a TAP device, which sends/receives packet in a raw format
+with a L2 header. The usage for a TAP PMD is for connectivity to the local host
+using a TAP interface. When the TAP PMD is initialized it will create a number
+of tap devices in the host accessed via 'ifconfig -a' or 'ip' command. The
+commands can be used to assign and query the virtual like device.
+
+These TAP interfaces can be used with wireshark or tcpdump or Pktgen-DPDK along
+with being able to be used as a network connection to the DPDK application. The
+method enable one or more interfaces is to use the --vdev=eth_tap option on the
+DPDK application  command line. Each --vdev=eth_tap option give will create an
+interface named dtap0, dtap1, ... and so forth.
+
+.. code-block:: console
+
+   The interfaced name can be changed by adding the iface=foo0
+   e.g. --vedv=eth_tap,iface=foo0 --vdev=eth_tap,iface=foo1, ...
+
+.. code-block:: console
+
+   Also the speed of the interface can be changed from 10G to whatever number
+   needed, but the interface does not enforce that speed.
+   e.g. --vdev=eth_tap,iface=foo0,speed=25000
+
+After the DPDK application is started you can send and receive packets on the
+interface using the standard rx_burst/tx_burst APIs in DPDK. From the host point
+of view you can use any host tool like tcpdump, wireshark, ping, Pktgen and
+others to communicate with the DPDK application. The DPDK application may not
+understand network protocols like IPv4/6, UDP or TCP unless the application has
+been written to understand these protocols.
+
+If you need the interface as a real network interface meaning running and has
+a valid IP address then you can do this with the following commands:
+
+.. code-block:: console
+
+   sudo ip link set dtap0 up; sudo ip addr add 192.168.0.250/24 dev dtap0
+   sudo ip link set dtap1 up; sudo ip addr add 192.168.1.250/24 dev dtap1
+
+Please change the IP addresses as you see fit.
+
+If routing is enabled on the host you can also communicate with the DPDK App
+over the internet via a standard socket layer application as long as you account
+for the protocol handing in the application.
+
+If you have a Network Stack in your DPDK application or something like it you
+can utilize that stack to handle the network protocols. Plus you would be able
+to address the interface using an IP address assigned to the internal interface.
Signature · 1 line
2.8.0.GIT
#4 lore raw

PATCH v3 1/1 drivers/net:new PMD using tun/tap host interface

In reply to Keith Wiles

The rte_eth_tap.c PMD creates a device using TUN/TAP interfaces on the local host. The PMD allows for DPDK and the host to communicate using a raw device interface on the host and in the DPDK application. The device created is a Tap device with a L2 packet header.

v3 - fix includes by removing ifdef for other type besides Linux.
Fix the copyright notice in the Makefile
v2 - merge all of the patches into one patch.
Fix a typo on naming the tap device.
Update the maintainers list

---
 MAINTAINERS                             |   5 +
 config/common_linuxapp                  |   2 +
 doc/guides/nics/tap.rst                 |  84 ++++
 drivers/net/Makefile                    |   1 +
 drivers/net/tap/Makefile                |  57 +++
 drivers/net/tap/rte_eth_tap.c           | 867 ++++++++++++++++++++++++++++++++
 drivers/net/tap/rte_pmd_tap_version.map |   4 +
 mk/rte.app.mk                           |   1 +
 8 files changed, 1021 insertions(+)
 create mode 100644 doc/guides/nics/tap.rst
 create mode 100644 drivers/net/tap/Makefile
 create mode 100644 drivers/net/tap/rte_eth_tap.c
 create mode 100644 drivers/net/tap/rte_pmd_tap_version.map
MAINTAINERS +5 -0
index 7c33ad4..fad74e4 100644
@@ -392,6 +392,11 @@ F: doc/guides/nics/pcap_ring.rst
 F: app/test/test_pmd_ring.c
 F: app/test/test_pmd_ring_perf.c
 
+Tap PMD
+M: Keith Wiles <[email protected]>
+F: drivers/net/tap
+F: doc/guides/nics/tap.rst
+
 Null Networking PMD
 M: Tetsuya Mukawa <[email protected]>
 F: drivers/net/null/
config/common_linuxapp +2 -0
index 2483dfa..59a2053 100644
@@ -44,3 +44,5 @@ CONFIG_RTE_LIBRTE_PMD_VHOST=y
 CONFIG_RTE_LIBRTE_PMD_AF_PACKET=y
 CONFIG_RTE_LIBRTE_POWER=y
 CONFIG_RTE_VIRTIO_USER=y
+CONFIG_RTE_LIBRTE_PMD_TAP=y
+CONFIG_RTE_PMD_TAP_MAX_QUEUES=32
doc/guides/nics/tap.rst new file +84 -0

84 lines not on this page · raw message

drivers/net/Makefile +1 -0
index bc93230..b4afa98 100644
@@ -55,6 +55,7 @@ DIRS-$(CONFIG_RTE_LIBRTE_THUNDERX_NICVF_PMD) += thunderx
 DIRS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD) += virtio
 DIRS-$(CONFIG_RTE_LIBRTE_VMXNET3_PMD) += vmxnet3
 DIRS-$(CONFIG_RTE_LIBRTE_PMD_XENVIRT) += xenvirt
+DIRS-$(CONFIG_RTE_LIBRTE_PMD_TAP) += tap
 
 ifeq ($(CONFIG_RTE_LIBRTE_VHOST),y)
 DIRS-$(CONFIG_RTE_LIBRTE_PMD_VHOST) += vhost
drivers/net/tap/Makefile new file +57 -0

57 lines not on this page · raw message

drivers/net/tap/rte_eth_tap.c new file +867 -0

867 lines not on this page · raw message

drivers/net/tap/rte_pmd_tap_version.map new file +4 -0
new file mode 100644
index 0000000..61463bf
@@ -0,0 +1,4 @@
+DPDK_16.11 {
+
+	local: *;
+};
mk/rte.app.mk +1 -0
index 1a0095b..bd1d10f 100644
@@ -129,6 +129,7 @@ ifeq ($(CONFIG_RTE_LIBRTE_VHOST),y)
 _LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_VHOST)      += -lrte_pmd_vhost
 endif # $(CONFIG_RTE_LIBRTE_VHOST)
 _LDLIBS-$(CONFIG_RTE_LIBRTE_VMXNET3_PMD)    += -lrte_pmd_vmxnet3_uio
+_LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_TAP)        += -lrte_pmd_tap
 
 ifeq ($(CONFIG_RTE_LIBRTE_CRYPTODEV),y)
 _LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_AESNI_MB)   += -lrte_pmd_aesni_mb
Signature · 1 line
2.8.0.GIT
#5 lore raw

PATCH v4 1/1 drivers/net:new PMD using tun/tap host interface

The rte_eth_tap.c PMD creates a device using TUN/TAP interfaces on the local host. The PMD allows for DPDK and the host to communicate using a raw device interface on the host and in the DPDK application. The device created is a Tap device with a L2 packet header.

v4 - merge with latest driver changes
v3 - fix includes by removing ifdef for other type besides Linux.
Fix the copyright notice in the Makefile
v2 - merge all of the patches into one patch.
Fix a typo on naming the tap device.
Update the maintainers list

---
 MAINTAINERS                             |   5 +
 config/common_linuxapp                  |   2 +
 doc/guides/nics/tap.rst                 |  84 ++++
 drivers/net/Makefile                    |   1 +
 drivers/net/tap/Makefile                |  57 +++
 drivers/net/tap/rte_eth_tap.c           | 866 ++++++++++++++++++++++++++++++++
 drivers/net/tap/rte_pmd_tap_version.map |   4 +
 mk/rte.app.mk                           |   1 +
 8 files changed, 1020 insertions(+)
 create mode 100644 doc/guides/nics/tap.rst
 create mode 100644 drivers/net/tap/Makefile
 create mode 100644 drivers/net/tap/rte_eth_tap.c
 create mode 100644 drivers/net/tap/rte_pmd_tap_version.map
MAINTAINERS +5 -0
index 7c33ad4..fad74e4 100644
@@ -392,6 +392,11 @@ F: doc/guides/nics/pcap_ring.rst
 F: app/test/test_pmd_ring.c
 F: app/test/test_pmd_ring_perf.c
 
+Tap PMD
+M: Keith Wiles <[email protected]>
+F: drivers/net/tap
+F: doc/guides/nics/tap.rst
+
 Null Networking PMD
 M: Tetsuya Mukawa <[email protected]>
 F: drivers/net/null/
config/common_linuxapp +2 -0
index 2483dfa..59a2053 100644
@@ -44,3 +44,5 @@ CONFIG_RTE_LIBRTE_PMD_VHOST=y
 CONFIG_RTE_LIBRTE_PMD_AF_PACKET=y
 CONFIG_RTE_LIBRTE_POWER=y
 CONFIG_RTE_VIRTIO_USER=y
+CONFIG_RTE_LIBRTE_PMD_TAP=y
+CONFIG_RTE_PMD_TAP_MAX_QUEUES=32
doc/guides/nics/tap.rst new file +84 -0

84 lines not on this page · raw message

drivers/net/Makefile +1 -0
index bc93230..b4afa98 100644
@@ -55,6 +55,7 @@ DIRS-$(CONFIG_RTE_LIBRTE_THUNDERX_NICVF_PMD) += thunderx
 DIRS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD) += virtio
 DIRS-$(CONFIG_RTE_LIBRTE_VMXNET3_PMD) += vmxnet3
 DIRS-$(CONFIG_RTE_LIBRTE_PMD_XENVIRT) += xenvirt
+DIRS-$(CONFIG_RTE_LIBRTE_PMD_TAP) += tap
 
 ifeq ($(CONFIG_RTE_LIBRTE_VHOST),y)
 DIRS-$(CONFIG_RTE_LIBRTE_PMD_VHOST) += vhost
drivers/net/tap/Makefile new file +57 -0

57 lines not on this page · raw message

drivers/net/tap/rte_eth_tap.c new file +866 -0

866 lines not on this page · raw message

drivers/net/tap/rte_pmd_tap_version.map new file +4 -0
new file mode 100644
index 0000000..61463bf
@@ -0,0 +1,4 @@
+DPDK_16.11 {
+
+	local: *;
+};
mk/rte.app.mk +1 -0
index 1a0095b..bd1d10f 100644
@@ -129,6 +129,7 @@ ifeq ($(CONFIG_RTE_LIBRTE_VHOST),y)
 _LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_VHOST)      += -lrte_pmd_vhost
 endif # $(CONFIG_RTE_LIBRTE_VHOST)
 _LDLIBS-$(CONFIG_RTE_LIBRTE_VMXNET3_PMD)    += -lrte_pmd_vmxnet3_uio
+_LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_TAP)        += -lrte_pmd_tap
 
 ifeq ($(CONFIG_RTE_LIBRTE_CRYPTODEV),y)
 _LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_AESNI_MB)   += -lrte_pmd_aesni_mb
Signature · 1 line
2.8.0.GIT
#6 lore raw

Hi Keith,

Quoted 14 lines · Keith Wiles

The rte_eth_tap.c PMD creates a device using TUN/TAP interfaces on the local host. The PMD allows for DPDK and the host to communicate using a raw device interface on the host and in the DPDK application. The device created is a Tap device with a L2 packet header.

v4 - merge with latest driver changes
v3 - fix includes by removing ifdef for other type besides Linux.
Fix the copyright notice in the Makefile
v2 - merge all of the patches into one patch.
Fix a typo on naming the tap device.
Update the maintainers list

Quoted 14 lines
---
 MAINTAINERS                             |   5 +
 config/common_linuxapp                  |   2 +
 doc/guides/nics/tap.rst                 |  84 ++++
 drivers/net/Makefile                    |   1 +
 drivers/net/tap/Makefile                |  57 +++
 drivers/net/tap/rte_eth_tap.c           | 866 ++++++++++++++++++++++++++++++++
 drivers/net/tap/rte_pmd_tap_version.map |   4 +
 mk/rte.app.mk                           |   1 +
 8 files changed, 1020 insertions(+)
 create mode 100644 doc/guides/nics/tap.rst
 create mode 100644 drivers/net/tap/Makefile
 create mode 100644 drivers/net/tap/rte_eth_tap.c
 create mode 100644 drivers/net/tap/rte_pmd_tap_version.map

This patch needs to be rebased on top of latest next-net, .init & .uninit are no more used.

Also patch gives a set of checkpatch warnings, fyi.

Thanks,
ferruh

#7 lore raw
In reply to Keith Wiles

2016-10-04 16:45 GMT+02:00, Keith Wiles <[email protected]>:

Quoted 5 lines

The rte_eth_tap.c PMD creates a device using TUN/TAP interfaces on the local host. The PMD allows for DPDK and the host to communicate using a raw device interface on the host and in the DPDK application. The device created is a Tap device with a L2 packet header.

[...]

Quoted 8 lines
drivers/net/tap/rte_eth_tap.c new file +8 -0
+static uint16_t
+pmd_rx_burst(void *queue, struct rte_mbuf **bufs, uint16_t nb_pkts)
+{
+	int len, n;
+	struct rte_mbuf *mbuf;
+	struct rx_queue *rxq = queue;
+	struct pollfd pfd;
+	uint16_t num_rx;

drivers/net/tap/rte_eth_tap.c line 207

Quoted 10 lines
drivers/net/tap/rte_eth_tap.c new file +10 -0
+	unsigned long num_rx_bytes = 0;
+
+	pfd.events = POLLIN;
+	pfd.fd = rxq->fd;
+	for (num_rx = 0; num_rx < nb_pkts; ) {
+		n = poll(&pfd, 1, 0);
+
+		if (n <= 0)
+			break;
+

Considering that syscalls are rather expensive, it would be cheaper to allocate an mbuf here and free it when read() returns -1 instead of calling poll() to check whether a packet is waiting. This way you save a syscall per packet and replace one syscall with one mbuf free per poll.

Best Regards,
Michał Mirosław

#8 lore raw

Regards,
Keith

Quoted 25 lines · Michał Mirosław

2016-10-04 16:45 GMT+02:00, Keith Wiles <[email protected]>:

Quoted 5 lines

The rte_eth_tap.c PMD creates a device using TUN/TAP interfaces on the local host. The PMD allows for DPDK and the host to communicate using a raw device interface on the host and in the DPDK application. The device created is a Tap device with a L2 packet header.

[...]

Quoted 16 lines
drivers/net/tap/rte_eth_tap.c new file +16 -0
+static uint16_t
+pmd_rx_burst(void *queue, struct rte_mbuf **bufs, uint16_t nb_pkts)
+{
+	int len, n;
+	struct rte_mbuf *mbuf;
+	struct rx_queue *rxq = queue;
+	struct pollfd pfd;
+	uint16_t num_rx;
+	unsigned long num_rx_bytes = 0;
+
+	pfd.events = POLLIN;
+	pfd.fd = rxq->fd;
+	for (num_rx = 0; num_rx < nb_pkts; ) {
+		n = poll(&pfd, 1, 0);
+
+		if (n <= 0)
Quoted 8 lines

Review

Quoted 2 lines
(quoted diff) +2 -0
+			break;
+

Considering that syscalls are rather expensive, it would be cheaper to allocate an mbuf here and free it when read() returns -1 instead of calling poll() to check whether a packet is waiting. This way you save a syscall per packet and replace one syscall with one mbuf free per poll.

I made this change, but saw no performance difference in the two methods. Removing poll seems reasonable as it is simpler. TAP is already so slow is why the performance did not change is my guess. Anyone wanting to use TAP as a high performance interface is going to be surprised. I believe the best use case for the TAP interface is for control or exception path.

Quoted 3 lines

Best Regards,
Michał Mirosław

#9 lore raw

2016-10-11 22:56 GMT+02:00 Wiles, Keith <[email protected]>:

Quoted 34 lines
Quoted 25 lines · Michał Mirosław

2016-10-04 16:45 GMT+02:00, Keith Wiles <[email protected]>:

Quoted 5 lines

The rte_eth_tap.c PMD creates a device using TUN/TAP interfaces on the local host. The PMD allows for DPDK and the host to communicate using a raw device interface on the host and in the DPDK application. The device created is a Tap device with a L2 packet header.

[...]

Quoted 16 lines
drivers/net/tap/rte_eth_tap.c new file +16 -0
+static uint16_t
+pmd_rx_burst(void *queue, struct rte_mbuf **bufs, uint16_t nb_pkts)
+{
+    int len, n;
+    struct rte_mbuf *mbuf;
+    struct rx_queue *rxq = queue;
+    struct pollfd pfd;
+    uint16_t num_rx;
+    unsigned long num_rx_bytes = 0;
+
+    pfd.events = POLLIN;
+    pfd.fd = rxq->fd;
+    for (num_rx = 0; num_rx < nb_pkts; ) {
+            n = poll(&pfd, 1, 0);
+
+            if (n <= 0)
Quoted 8 lines

Review

Quoted 2 lines
(quoted diff) +2 -0
+                    break;
+

Considering that syscalls are rather expensive, it would be cheaper to allocate an mbuf here and free it when read() returns -1 instead of calling poll() to check whether a packet is waiting. This way you save a syscall per packet and replace one syscall with one mbuf free per poll.

Quoted 1 line

I made this change, but saw no performance difference in the two methods. Removing poll seems reasonable as it is simpler. TAP is already so slow is why the performance did not change is my guess. Anyone wanting to use TAP as a high performance interface is going to be surprised. I believe the best use case for the TAP interface is for control or exception path.

Agreed, TAP does not look like designed for performance as a first goal.

You could do the same simplification for TX path, BTW.

Best Regards,
Michał Mirosław

#10 lore raw
In reply to Keith Wiles
Quoted 14 lines · Keith Wiles

The rte_eth_tap.c PMD creates a device using TUN/TAP interfaces on the local host. The PMD allows for DPDK and the host to communicate using a raw device interface on the host and in the DPDK application. The device created is a Tap device with a L2 packet header.

v4 - merge with latest driver changes
v3 - fix includes by removing ifdef for other type besides Linux.
Fix the copyright notice in the Makefile
v2 - merge all of the patches into one patch.
Fix a typo on naming the tap device.
Update the maintainers list

Quoted 15 lines
---
 MAINTAINERS                             |   5 +
 config/common_linuxapp                  |   2 +
 doc/guides/nics/tap.rst                 |  84 ++++
 drivers/net/Makefile                    |   1 +
 drivers/net/tap/Makefile                |  57 +++
 drivers/net/tap/rte_eth_tap.c           | 866 ++++++++++++++++++++++++++++++++
 drivers/net/tap/rte_pmd_tap_version.map |   4 +
 mk/rte.app.mk                           |   1 +
 8 files changed, 1020 insertions(+)
 create mode 100644 doc/guides/nics/tap.rst
 create mode 100644 drivers/net/tap/Makefile
 create mode 100644 drivers/net/tap/rte_eth_tap.c
 create mode 100644 drivers/net/tap/rte_pmd_tap_version.map

<>

config/common_linuxapp CONFIG_RTE_LIBRTE_PMD_VHOST=y · line 47

Quoted 9 lines
config/common_linuxapp +1 -0
index 2483dfa..59a2053 100644
@@ -44,3 +44,5 @@ CONFIG_RTE_LIBRTE_PMD_VHOST=y
 CONFIG_RTE_LIBRTE_PMD_AF_PACKET=y
 CONFIG_RTE_LIBRTE_POWER=y
 CONFIG_RTE_VIRTIO_USER=y
+CONFIG_RTE_LIBRTE_PMD_TAP=y

According existing config items, a default value of a config option should go to config/common_base, and environment specific config file overwrites it if required.
So this option needs to be added into config/common_base too as disabled by default.

config/common_linuxapp CONFIG_RTE_LIBRTE_PMD_VHOST=y · line 48

Quoted 1 line
config/common_linuxapp +1 -0
+CONFIG_RTE_PMD_TAP_MAX_QUEUES=32

Is the number of max queues really needs to be a config option, I assume in normal use case user won't need to update this and will use single queue, if that is true what about pushing this into source code to not make config file more complex?

Quoted 1 line
doc/guides/nics/tap.rst +0 -0

<...>

doc/guides/nics/tap.rst line 53

Quoted 4 lines

+.. code-block:: console

doc/guides/nics/tap.rst +3 -0
+
+   The interfaced name can be changed by adding the iface=foo0
+   e.g. --vedv=eth_tap,iface=foo0 --vdev=eth_tap,iface=foo1, ...

s/vedv/vdev
eth_tap needs to be net_tap as part of unifying device names work

<...>

drivers/net/Makefile DIRS-$(CONFIG_RTE_LIBRTE_THUNDERX_NICVF_PMD) += thunderx · line 58

Quoted 9 lines
drivers/net/Makefile +1 -0
index bc93230..b4afa98 100644
@@ -55,6 +55,7 @@ DIRS-$(CONFIG_RTE_LIBRTE_THUNDERX_NICVF_PMD) += thunderx
 DIRS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD) += virtio
 DIRS-$(CONFIG_RTE_LIBRTE_VMXNET3_PMD) += vmxnet3
 DIRS-$(CONFIG_RTE_LIBRTE_PMD_XENVIRT) += xenvirt
+DIRS-$(CONFIG_RTE_LIBRTE_PMD_TAP) += tap

Rest of the PMDs sorted alphabetically, please do same.

Quoted 3 lines
drivers/net/Makefile +0 -0
 ifeq ($(CONFIG_RTE_LIBRTE_VHOST),y)
 DIRS-$(CONFIG_RTE_LIBRTE_PMD_VHOST) += vhost

<...>

Quoted 1 line
drivers/net/tap/rte_eth_tap.c +0 -0

<...>

drivers/net/tap/rte_eth_tap.c line 68

Quoted 3 lines
drivers/net/tap/rte_eth_tap.c +3 -0
+
+static const char *drivername = "Tap PMD";
+static int tap_unit = 0;

No need to initialize to zero.

<...>

drivers/net/tap/rte_eth_tap.c line 106

Quoted 5 lines
drivers/net/tap/rte_eth_tap.c +5 -0
+
+struct pmd_internals {
+	char name[RTE_ETH_NAME_MAX_LEN];	/* Internal Tap device name */
+	uint16_t nb_queues;			/* Number of queues supported */
+	uint16_t pad0;

Why this padding? Is it reserved?

Quoted 7 lines
drivers/net/tap/rte_eth_tap.c +7 -0
+	struct ether_addr eth_addr;	/* Mac address of the device port */
+
+	int if_index;			/* IF_INDEX for the port */
+	int fds[RTE_PMD_TAP_MAX_QUEUES]; /* List of all file descriptors */
+
+	struct rx_queue rxq[RTE_PMD_TAP_MAX_QUEUES];	/* List of RX queues */
+	struct tx_queue txq[RTE_PMD_TAP_MAX_QUEUES];	/* List of TX queues */

drivers/net/tap/rte_eth_tap.c line 123

Quoted 10 lines
drivers/net/tap/rte_eth_tap.c +10 -0
+};
+
+/*
+ * Tun/Tap allocation routine
+ *
+ * name is the number of the interface to use, unless NULL to take the host
+ * supplied name.
+ */
+static int
+tun_alloc(char * name)

char *name

<...>

drivers/net/tap/rte_eth_tap.c line 167

Quoted 2 lines
drivers/net/tap/rte_eth_tap.c +2 -0
+
+	/* Always set the fiile descriptor to non-blocking */

s/fiile/file

drivers/net/tap/rte_eth_tap.c line 176

Quoted 9 lines
drivers/net/tap/rte_eth_tap.c +9 -0
+	if (fcntl(fd, F_SETFL, O_NONBLOCK) < 0) {
+		RTE_LOG(ERR, PMD, "Unable to set to nonblocking\n");
+		perror("F_SETFL, NONBLOCK");
+		goto error;
+	}
+
+	/* If the name is different that new name as default */
+	if (name && strcmp(name, ifr.ifr_name))
+		strcpy(name, ifr.ifr_name);

What about more secure copy?

Quoted 9 lines
drivers/net/tap/rte_eth_tap.c +9 -0
+
+	return fd;
+
+error:
+	if (fd > 0)
+		close(fd);
+	return -1;
+}
+

<...>

drivers/net/tap/rte_eth_tap.c line 334

Quoted 12 lines
drivers/net/tap/rte_eth_tap.c +12 -0
+
+static void
+tap_dev_info(struct rte_eth_dev *dev, struct rte_eth_dev_info *dev_info)
+{
+	struct pmd_internals *internals = dev->data->dev_private;
+
+	dev_info->driver_name = drivername;
+	dev_info->if_index = internals->if_index;
+	dev_info->max_mac_addrs = 1;
+	dev_info->max_rx_pktlen = (uint32_t)ETHER_MAX_VLAN_FRAME_LEN;
+	dev_info->max_rx_queues = (uint16_t)internals->nb_queues;
+	dev_info->max_tx_queues = (uint16_t)internals->nb_queues;

casting to uint16_t is not requires, it is already uint16_t.

drivers/net/tap/rte_eth_tap.c line 340

Quoted 6 lines
drivers/net/tap/rte_eth_tap.c +6 -0
+	dev_info->min_rx_bufsize = 0;
+	dev_info->pci_dev = NULL;
+}
+
+static void
+tap_stats_get(struct rte_eth_dev *dev, struct rte_eth_stats *igb_stats)

igb_stats?

Quoted 8 lines
drivers/net/tap/rte_eth_tap.c +8 -0
+{
+	unsigned i, imax;
+	unsigned long rx_total = 0, tx_total = 0, tx_err_total = 0;
+	unsigned long rx_bytes_total = 0, tx_bytes_total = 0;
+	const struct pmd_internals *internal = dev->data->dev_private;
+
+	imax = (internal->nb_queues < RTE_ETHDEV_QUEUE_STAT_CNTRS) ?
+		internal->nb_queues : RTE_ETHDEV_QUEUE_STAT_CNTRS;

drivers/net/tap/rte_eth_tap.c line 358

Quoted 10 lines
drivers/net/tap/rte_eth_tap.c +10 -0
+
+	for (i = 0; i < imax; i++) {
+		igb_stats->q_ipackets[i] = internal->rxq[i].stats.ipackets;
+		igb_stats->q_ibytes[i] = internal->rxq[i].stats.ibytes;
+		rx_total += igb_stats->q_ipackets[i];
+		rx_bytes_total += igb_stats->q_ibytes[i];
+	}
+
+	imax = (internal->nb_queues < RTE_ETHDEV_QUEUE_STAT_CNTRS) ?
+		internal->nb_queues : RTE_ETHDEV_QUEUE_STAT_CNTRS;

Do we need to duplicate imax calculation?

Quoted 17 lines
drivers/net/tap/rte_eth_tap.c +17 -0
+
+	for (i = 0; i < imax; i++) {
+		igb_stats->q_opackets[i] = internal->txq[i].stats.opackets;
+		igb_stats->q_errors[i] = internal->txq[i].stats.errs;
+		igb_stats->q_obytes[i] = internal->txq[i].stats.obytes;
+		tx_total += igb_stats->q_opackets[i];
+		tx_err_total += igb_stats->q_errors[i];
+		tx_bytes_total += igb_stats->q_obytes[i];
+	}
+
+	igb_stats->ipackets = rx_total;
+	igb_stats->ibytes = rx_bytes_total;
+	igb_stats->opackets = tx_total;
+	igb_stats->oerrors = tx_err_total;
+	igb_stats->obytes = tx_bytes_total;
+}
+

<...>

Quoted 18 lines
drivers/net/tap/rte_eth_tap.c +18 -0
+
+static int
+rte_eth_dev_create(const char *name,
+		   struct rte_eth_dev **eth_dev,
+		   const struct eth_dev_ops *dev_ops,
+		   void **internals, size_t internal_size,
+		   uint16_t flag)
+{
+	char buff[RTE_ETH_NAME_MAX_LEN];
+	int numa_node = rte_socket_id();
+	struct rte_eth_dev *dev = NULL;
+	struct rte_eth_dev_data *data = NULL;
+	void *priv = NULL;
+
+	if ((name == NULL) || (eth_dev == NULL) || (dev_ops == NULL) ||
+	    (internals == NULL) || (internal_size == 0)) {
+		RTE_PMD_DEBUG_TRACE("Paramters are invalid\n");
+		return -1;

drivers/net/tap/rte_eth_tap.c line 606

Quoted 10 lines
drivers/net/tap/rte_eth_tap.c +10 -0
+	}
+
+	dev = rte_eth_dev_allocate(name);
+	if (dev == NULL) {
+		RTE_PMD_DEBUG_TRACE("%s: rte_eth_dev_allocate failed for %s\n",
+				    name, buff);
+		goto error;
+	}
+
+	if (flag & RTE_USE_PRIVATE_DATA) {

You may need to save this flag value somewhere in internals, to decide how to free data later.

Quoted 4 lines
drivers/net/tap/rte_eth_tap.c +4 -0
+		/*
+		 * now do all data allocation - for eth_dev structure, dummy
+		 * pci driver and internal (private) data
+		 */

drivers/net/tap/rte_eth_tap.c line 620

Quoted 10 lines
drivers/net/tap/rte_eth_tap.c +10 -0
+		snprintf(buff, sizeof(buff), "D-%s-%d", name, numa_node);
+		data = rte_zmalloc_socket(buff, sizeof(struct rte_eth_dev_data),
+					  0, numa_node);
+		if (data == NULL) {
+			RTE_PMD_DEBUG_TRACE("%s: Unable to allocate memory\n",
+					    name);
+			goto error;
+		}
+		/* move the current state of the structure to the new one */
+		rte_memcpy(data, dev->data, sizeof(struct rte_eth_dev_data));

Why do we need to copy, trying to preserve which data?

Quoted 8 lines
drivers/net/tap/rte_eth_tap.c +8 -0
+		dev->data = data;	/* Override the current data pointer */
+	} else
+		data = dev->data;
+
+	snprintf(buff, sizeof(buff), "I-%s-%d", name, numa_node);
+	priv = rte_zmalloc_socket(buff, internal_size, 0, numa_node);
+	if (priv == NULL) {
+		RTE_PMD_DEBUG_TRACE("Unable to allocate internal memory %lu\n",

drivers/net/tap/rte_eth_tap.c line 637

Quoted 10 lines
drivers/net/tap/rte_eth_tap.c +9 -0
+				    internal_size);
+		goto error;
+	}
+
+	/* Setup some default values */
+	dev->dev_ops = dev_ops;
+	data->dev_private = priv;
 
+	data->port_id = dev->data->port_id;
+	memmove(data->name, dev->data->name, strlen(dev->data->name));

These two assignments are useless, needs to be done before "dev->data = data" assignment.

Quoted 29 lines
drivers/net/tap/rte_eth_tap.c +29 -0
+
+	dev->driver = NULL;
+	data->dev_flags = RTE_ETH_DEV_DETACHABLE;
+	data->kdrv = RTE_KDRV_NONE;
+	data->numa_node = numa_node;
+
+	*eth_dev = dev;
+	*internals = priv;
+
+	return 0;
+error:
+	rte_free(priv);
+
+	if (flag & RTE_USE_PRIVATE_DATA)
+		rte_free(data);
+
+	rte_eth_dev_release_port(dev);
+
+	return -1;
+}
+
+static int
+pmd_init_internals(const char *name, struct tap_info *tap,
+		   struct pmd_internals **internals,
+		   struct rte_eth_dev **eth_dev)
+{
+	struct rte_eth_dev *dev = NULL;
+	struct pmd_internals *internal = NULL;
+	struct rte_eth_dev_data *data = NULL;

drivers/net/tap/rte_eth_tap.c line 676

Quoted 10 lines
drivers/net/tap/rte_eth_tap.c +10 -0
+	int ret, i, fd = -1;
+
+	RTE_LOG(INFO, PMD,
+		"%s: Create TUN/TAP Ethernet device with %d queues on numa %u\n",
+		name, RTE_PMD_TAP_MAX_QUEUES, rte_socket_id());
+
+	pmd_link.link_speed = tap->speed;
+
+	ret = rte_eth_dev_create(tap->name, &dev, &ops,
+				 (void **)&internal, sizeof(struct pmd_internals),

Why rte_eth_dev_create() get "void **internals" which requires casting, but not "struct pmd_internals **internals" ?

drivers/net/tap/rte_eth_tap.c line 688

Quoted 12 lines
drivers/net/tap/rte_eth_tap.c +12 -0
+				 RTE_USE_PRIVATE_DATA);
+	if (ret < 0)
+		return -1;
+
+	strncpy(internal->name, tap->name, sizeof(internal->name));
+
+	internal->nb_queues = RTE_PMD_TAP_MAX_QUEUES;
+
+	/* Create the first Tap device */
+	if ((fd = tun_alloc(dev->data->name)) < 0) {
+		RTE_LOG(ERR, PMD, "tun_alloc(%s) failed\n", dev->data->name);
+		rte_free(internal);

rte_free(dev->data); ?
But needs to check RTE_USE_PRIVATE_DATA ..

Quoted 8 lines
drivers/net/tap/rte_eth_tap.c +8 -0
+		rte_eth_dev_release_port(dev);
+		return -1;
+	}
+
+	/* Presetup the fds to -1 as being not working */
+	for(i = 0; i < RTE_PMD_TAP_MAX_QUEUES; i++) {
+		internal->fds[i] = -1;
+		internal->rxq[i].fd = -1;

drivers/net/tap/rte_eth_tap.c line 706

Quoted 10 lines
drivers/net/tap/rte_eth_tap.c +10 -0
+		internal->txq[i].fd = -1;
+	}
+
+	/* Take the TUN/TAP fd and place in the first location */
+	internal->rxq[0].fd = fd;
+	internal->txq[0].fd = fd;
+	internal->fds[0] = fd;
+
+	if (pmd_mac_address(fd, dev, &internal->eth_addr) < 0) {
+		rte_free(internal);

rte_free(dev->data); ?

drivers/net/tap/rte_eth_tap.c line 717

Quoted 11 lines
drivers/net/tap/rte_eth_tap.c +11 -0
+		rte_eth_dev_release_port(dev);
+		return -1;
+	}
+
+	data = dev->data;
+
+	data->dev_link = pmd_link;
+	data->mac_addrs = &internal->eth_addr;
+
+	data->nb_rx_queues = (uint16_t)internal->nb_queues;
+	data->nb_tx_queues = (uint16_t)internal->nb_queues;

no cast required.

Quoted 8 lines
drivers/net/tap/rte_eth_tap.c +8 -0
+	data->drv_name = drivername;
+
+	*eth_dev = dev;
+	*internals = internal;
+
+	return 0;
+}
+

<...>

drivers/net/tap/rte_eth_tap.c line 759

Quoted 5 lines
drivers/net/tap/rte_eth_tap.c +5 -0
+
+static int
+set_interface_speed(const char *key __rte_unused,
+		    const char *value,
+		    void *extra_args __rte_unused)

need to drop __rte_unused for extra_args

Quoted 14 lines
drivers/net/tap/rte_eth_tap.c +14 -0
+{
+	struct tap_info *tap = (struct tap_info *)extra_args;
+
+	pmd_link.link_speed = (value) ? atoi(value) : ETH_SPEED_NUM_10G;
+	tap->speed = pmd_link.link_speed;
+
+	return 0;
+}
+
+/*
+ * Open a TAP interface device.
+ */
+static int
+rte_pmd_tap_devinit(const char *name, const char *params)

drivers/net/tap/rte_eth_tap.c line 783

Quoted 10 lines
drivers/net/tap/rte_eth_tap.c +10 -0
+{
+	int ret = 0;
+	struct rte_kvargs *kvlist;
+	struct tap_info tap_info;
+
+	/* Setup default values */
+	memset(&tap_info, 0, sizeof(tap_info));
+
+	tap_info.speed = ETH_SPEED_NUM_10G;
+	snprintf(tap_info.name, sizeof(tap_info.name), "dtap%d", tap_unit++);

What about extracting iface name "dtap" into a macro to make it more visible.

drivers/net/tap/rte_eth_tap.c line 788

Quoted 5 lines
drivers/net/tap/rte_eth_tap.c +5 -0
+
+	if ((params == NULL) || (params[0] == '\0')) {
+		RTE_LOG(INFO, PMD, "Initializing pmd_tap for %s\n", name);
+
+		ret = eth_dev_tap_create(name, &tap_info);

This "name" is not used at all (except from RTE_LOG), instead tap->name is used for interface name, so why carying this variable around?

Quoted 8 lines
drivers/net/tap/rte_eth_tap.c +8 -0
+		goto leave;
+	}
+
+	RTE_LOG(INFO, PMD, "Initialize %s with params (%s)\n", name, params);
+
+	kvlist = rte_kvargs_parse(params, valid_arguments);
+	if (!kvlist) {
+		ret = eth_dev_tap_create(name, &tap_info);

drivers/net/tap/rte_eth_tap.c line 806

Quoted 10 lines
drivers/net/tap/rte_eth_tap.c +10 -0
+		goto leave;
+	}
+
+	if (rte_kvargs_count(kvlist, ETH_TAP_SPEED_ARG) == 1) {
+		ret = rte_kvargs_process(kvlist, ETH_TAP_SPEED_ARG,
+					 &set_interface_speed, &tap_info);
+		if (ret < 0)
+			goto leave;
+	} else
+		set_interface_speed(NULL, NULL, &tap_info);

This call is redundant, tap_info already has default speed value set.

drivers/net/tap/rte_eth_tap.c line 814

Quoted 8 lines
drivers/net/tap/rte_eth_tap.c +8 -0
+
+	if (rte_kvargs_count(kvlist, ETH_TAP_IFACE_ARG) == 1) {
+		ret = rte_kvargs_process(kvlist, ETH_TAP_IFACE_ARG,
+					 &set_interface_name, &tap_info);
+		if (ret < 0)
+			goto leave;
+	} else
+		set_interface_name(NULL, NULL, (void *)&tap_info);

tap_info->name already set to default value (dtap%d), this call is not required.

Quoted 14 lines
drivers/net/tap/rte_eth_tap.c +14 -0
+
+	rte_kvargs_free(kvlist);
+
+leave:
+	if (ret == -1)
+		RTE_LOG(INFO, PMD, "Failed to create pmd_tap for %s\n", name);
+
+	return ret;
+}
+
+/*
+ * detach a TAP device.
+ */
+static int

drivers/net/tap/rte_eth_tap.c line 838

Quoted 10 lines
drivers/net/tap/rte_eth_tap.c +10 -0
+rte_pmd_tap_devuninit(const char *name)
+{
+	struct rte_eth_dev *eth_dev = NULL;
+	struct pmd_internals *internals;
+	int i;
+
+	RTE_LOG(INFO, PMD, "Closing TUN/TAP Ethernet device on numa %u\n",
+		rte_socket_id());
+
+	if (name == NULL)

This check is redundant, eal layer won't call this function with "name == NULL"

Quoted 4 lines
drivers/net/tap/rte_eth_tap.c +4 -0
+		return 0;
+
+	/* find the ethdev entry */
+	eth_dev = rte_eth_dev_allocated(name);

drivers/net/tap/rte_eth_tap.c line 852

Quoted 10 lines
drivers/net/tap/rte_eth_tap.c +10 -0
+	if (eth_dev == NULL)
+		return 0;
+
+	internals = eth_dev->data->dev_private;
+	for (i = 0; i < internals->nb_queues; i++)
+		if (internals->fds[i] != -1)
+			close(internals->fds[i]);
+
+	rte_free(eth_dev->data->dev_private);
+	rte_free(eth_dev->data);

data can be shared?
Don't we need a RTE_USE_PRIVATE_DATA flag check?

drivers/net/tap/rte_eth_tap.c line 864

Quoted 12 lines
drivers/net/tap/rte_eth_tap.c +12 -0
+
+	rte_eth_dev_release_port(eth_dev);
+
+	return 0;
+}
+
+static struct rte_vdev_driver pmd_tap_drv = {
+	.init = rte_pmd_tap_devinit,
+	.uninit = rte_pmd_tap_devuninit,
+};
+
+DRIVER_REGISTER_VDEV(eth_tap, pmd_tap_drv);

name convention is now: "net_tap"

Quoted 17 lines
drivers/net/tap/rte_eth_tap.c +2 -0
+DRIVER_REGISTER_PARAM_STRING(eth_tap,
+			     "iface=<string>,speed=N");
drivers/net/tap/rte_pmd_tap_version.map new file +4 -0
new file mode 100644
index 0000000..61463bf
@@ -0,0 +1,4 @@
+DPDK_16.11 {
+
+	local: *;
+};
mk/rte.app.mk +0 -0
index 1a0095b..bd1d10f 100644

mk/rte.app.mk ifeq ($(CONFIG_RTE_LIBRTE_VHOST),y) · line 132

Quoted 4 lines
mk/rte.app.mk +1 -0
@@ -129,6 +129,7 @@ ifeq ($(CONFIG_RTE_LIBRTE_VHOST),y)
 _LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_VHOST)      += -lrte_pmd_vhost
 endif # $(CONFIG_RTE_LIBRTE_VHOST)
 _LDLIBS-$(CONFIG_RTE_LIBRTE_VMXNET3_PMD)    += -lrte_pmd_vmxnet3_uio
+_LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_TAP)        += -lrte_pmd_tap

please put in alphebetical order

Quoted 4 lines
mk/rte.app.mk +0 -0
 ifeq ($(CONFIG_RTE_LIBRTE_CRYPTODEV),y)
 _LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_AESNI_MB)   += -lrte_pmd_aesni_mb
 
#11 lore raw

Regards,
Keith

Quoted 9 lines · Ferruh
Quoted 6 lines · Keith Wiles

The rte_eth_tap.c PMD creates a device using TUN/TAP interfaces on the local host. The PMD allows for DPDK and the host to communicate using a raw device interface on the host and in the DPDK application. The device created is a Tap device with a L2 packet header.

Will try to ship out a v5 soon.

Quoted 24 lines
Quoted 18 lines

v4 - merge with latest driver changes
v3 - fix includes by removing ifdef for other type besides Linux.
Fix the copyright notice in the Makefile
v2 - merge all of the patches into one patch.
Fix a typo on naming the tap device.
Update the maintainers list

---

MAINTAINERS                             |   5 +
config/common_linuxapp                  |   2 +
doc/guides/nics/tap.rst                 |  84 ++++
drivers/net/Makefile                    |   1 +
drivers/net/tap/Makefile                |  57 +++
drivers/net/tap/rte_eth_tap.c           | 866 ++++++++++++++++++++++++++++++++
drivers/net/tap/rte_pmd_tap_version.map |   4 +
mk/rte.app.mk                           |   1 +
8 files changed, 1020 insertions(+)
Quoted 5 lines

create mode 100644 doc/guides/nics/tap.rst
create mode 100644 drivers/net/tap/Makefile create mode 100644 drivers/net/tap/rte_eth_tap.c create mode 100644 drivers/net/tap/rte_pmd_tap_version.map

<>

Quoted 15 lines
Quoted 9 lines
config/common_linuxapp +0 -0
index 2483dfa..59a2053 100644
@@ -44,3 +44,5 @@ CONFIG_RTE_LIBRTE_PMD_VHOST=y

CONFIG_RTE_LIBRTE_PMD_AF_PACKET=y
CONFIG_RTE_LIBRTE_POWER=y
CONFIG_RTE_VIRTIO_USER=y
+CONFIG_RTE_LIBRTE_PMD_TAP=y

According existing config items, a default value of a config option should go to config/common_base, and environment specific config file overwrites it if required.
So this option needs to be added into config/common_base too as disabled by default.

Add the define to common_base as no, plus a comment for Linux only.

Quoted 2 lines
Quoted 2 lines

+CONFIG_RTE_PMD_TAP_MAX_QUEUES=32

Moved this to the .c file as a define.

Quoted 10 lines

Is the number of max queues really needs to be a config option, I assume in normal use case user won't need to update this and will use single queue, if that is true what about pushing this into source code to not make config file more complex?

Quoted 1 line
doc/guides/nics/tap.rst +0 -0

<...>

Quoted 7 lines

doc/guides/nics/tap.rst line 53

Quoted 4 lines

+.. code-block:: console

doc/guides/nics/tap.rst +3 -0
+
+   The interfaced name can be changed by adding the iface=foo0
+   e.g. --vedv=eth_tap,iface=foo0 --vdev=eth_tap,iface=foo1, ...

s/vedv/vdev
eth_tap needs to be net_tap as part of unifying device names work

Fixed.

Quoted 14 lines

<...>

Quoted 9 lines
drivers/net/Makefile +0 -0
index bc93230..b4afa98 100644
@@ -55,6 +55,7 @@ DIRS-$(CONFIG_RTE_LIBRTE_THUNDERX_NICVF_PMD) += thunderx

DIRS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD) += virtio
DIRS-$(CONFIG_RTE_LIBRTE_VMXNET3_PMD) += vmxnet3
DIRS-$(CONFIG_RTE_LIBRTE_PMD_XENVIRT) += xenvirt
+DIRS-$(CONFIG_RTE_LIBRTE_PMD_TAP) += tap

Rest of the PMDs sorted alphabetically, please do same.

Done.

Quoted 11 lines
Quoted 4 lines

ifeq ($(CONFIG_RTE_LIBRTE_VHOST),y)
DIRS-$(CONFIG_RTE_LIBRTE_PMD_VHOST) += vhost

<...>

Quoted 1 line
drivers/net/tap/rte_eth_tap.c +0 -0

<...>

Quoted 5 lines

drivers/net/tap/rte_eth_tap.c line 68

Quoted 3 lines
drivers/net/tap/rte_eth_tap.c +3 -0
+
+static const char *drivername = "Tap PMD";
+static int tap_unit = 0;

No need to initialize to zero.

Fixed

Quoted 10 lines

<...>

drivers/net/tap/rte_eth_tap.c line 106

Quoted 5 lines
drivers/net/tap/rte_eth_tap.c +5 -0
+
+struct pmd_internals {
+	char name[RTE_ETH_NAME_MAX_LEN];	/* Internal Tap device name */
+	uint16_t nb_queues;			/* Number of queues supported */
+	uint16_t pad0;

Why this padding? Is it reserved?

Removed pad0. I just like to know about gaps in the structures is the reason.

Quoted 12 lines
Quoted 12 lines
drivers/net/tap/rte_eth_tap.c +11 -0
+	struct ether_addr eth_addr;	/* Mac address of the device port */
+
+	int if_index;			/* IF_INDEX for the port */
+	int fds[RTE_PMD_TAP_MAX_QUEUES]; /* List of all file descriptors */
+
+	struct rx_queue rxq[RTE_PMD_TAP_MAX_QUEUES];	/* List of RX queues */
+	struct tx_queue txq[RTE_PMD_TAP_MAX_QUEUES];	/* List of TX queues */
+};
+
+/*
+ * Tun/Tap allocation routine
Quoted 8 lines

drivers/net/tap/rte_eth_tap.c line 123

Quoted 6 lines

+ *
+ * name is the number of the interface to use, unless NULL to take the host
+ * supplied name.
+ */

drivers/net/tap/rte_eth_tap.c +2 -0
+static int
+tun_alloc(char * name)

char *name

Fixed.

Quoted 8 lines

<...>

drivers/net/tap/rte_eth_tap.c line 167

Quoted 2 lines
drivers/net/tap/rte_eth_tap.c +2 -0
+
+	/* Always set the fiile descriptor to non-blocking */

s/fiile/file

Quoted 10 lines

drivers/net/tap/rte_eth_tap.c line 176

Quoted 9 lines
drivers/net/tap/rte_eth_tap.c +9 -0
+	if (fcntl(fd, F_SETFL, O_NONBLOCK) < 0) {
+		RTE_LOG(ERR, PMD, "Unable to set to nonblocking\n");
+		perror("F_SETFL, NONBLOCK");
+		goto error;
+	}
+
+	/* If the name is different that new name as default */
+	if (name && strcmp(name, ifr.ifr_name))
+		strcpy(name, ifr.ifr_name);

What about more secure copy?

Changed to be more secure.

Quoted 13 lines
Quoted 10 lines
drivers/net/tap/rte_eth_tap.c +9 -0
+
+	return fd;
+
+error:
+	if (fd > 0)
+		close(fd);
+	return -1;
+}
+

<...>

Quoted 13 lines

drivers/net/tap/rte_eth_tap.c line 334

Quoted 12 lines
drivers/net/tap/rte_eth_tap.c +12 -0
+
+static void
+tap_dev_info(struct rte_eth_dev *dev, struct rte_eth_dev_info *dev_info)
+{
+	struct pmd_internals *internals = dev->data->dev_private;
+
+	dev_info->driver_name = drivername;
+	dev_info->if_index = internals->if_index;
+	dev_info->max_mac_addrs = 1;
+	dev_info->max_rx_pktlen = (uint32_t)ETHER_MAX_VLAN_FRAME_LEN;
+	dev_info->max_rx_queues = (uint16_t)internals->nb_queues;
+	dev_info->max_tx_queues = (uint16_t)internals->nb_queues;

casting to uint16_t is not requires, it is already uint16_t.

Removed

Quoted 20 lines

drivers/net/tap/rte_eth_tap.c line 340

Quoted 7 lines
drivers/net/tap/rte_eth_tap.c +6 -0
+	dev_info->min_rx_bufsize = 0;
+	dev_info->pci_dev = NULL;
+}
+
+static void
+tap_stats_get(struct rte_eth_dev *dev, struct rte_eth_stats *igb_stats)

igb_stats?

Quoted 11 lines
drivers/net/tap/rte_eth_tap.c +11 -0
+{
+	unsigned i, imax;
+	unsigned long rx_total = 0, tx_total = 0, tx_err_total = 0;
+	unsigned long rx_bytes_total = 0, tx_bytes_total = 0;
+	const struct pmd_internals *internal = dev->data->dev_private;
+
+	imax = (internal->nb_queues < RTE_ETHDEV_QUEUE_STAT_CNTRS) ?
+		internal->nb_queues : RTE_ETHDEV_QUEUE_STAT_CNTRS;
+
+	for (i = 0; i < imax; i++) {
+		igb_stats->q_ipackets[i] = internal->rxq[i].stats.ipackets;
Quoted 8 lines

drivers/net/tap/rte_eth_tap.c line 358

Quoted 7 lines
drivers/net/tap/rte_eth_tap.c +7 -0
+		igb_stats->q_ibytes[i] = internal->rxq[i].stats.ibytes;
+		rx_total += igb_stats->q_ipackets[i];
+		rx_bytes_total += igb_stats->q_ibytes[i];
+	}
+
+	imax = (internal->nb_queues < RTE_ETHDEV_QUEUE_STAT_CNTRS) ?
+		internal->nb_queues : RTE_ETHDEV_QUEUE_STAT_CNTRS;

Do we need to duplicate imax calculation?

Removed

Quoted 45 lines
Quoted 19 lines
drivers/net/tap/rte_eth_tap.c +17 -0
+
+	for (i = 0; i < imax; i++) {
+		igb_stats->q_opackets[i] = internal->txq[i].stats.opackets;
+		igb_stats->q_errors[i] = internal->txq[i].stats.errs;
+		igb_stats->q_obytes[i] = internal->txq[i].stats.obytes;
+		tx_total += igb_stats->q_opackets[i];
+		tx_err_total += igb_stats->q_errors[i];
+		tx_bytes_total += igb_stats->q_obytes[i];
+	}
+
+	igb_stats->ipackets = rx_total;
+	igb_stats->ibytes = rx_bytes_total;
+	igb_stats->opackets = tx_total;
+	igb_stats->oerrors = tx_err_total;
+	igb_stats->obytes = tx_bytes_total;
+}
+

<...>

Quoted 23 lines
drivers/net/tap/rte_eth_tap.c +23 -0
+
+static int
+rte_eth_dev_create(const char *name,
+		   struct rte_eth_dev **eth_dev,
+		   const struct eth_dev_ops *dev_ops,
+		   void **internals, size_t internal_size,
+		   uint16_t flag)
+{
+	char buff[RTE_ETH_NAME_MAX_LEN];
+	int numa_node = rte_socket_id();
+	struct rte_eth_dev *dev = NULL;
+	struct rte_eth_dev_data *data = NULL;
+	void *priv = NULL;
+
+	if ((name == NULL) || (eth_dev == NULL) || (dev_ops == NULL) ||
+	    (internals == NULL) || (internal_size == 0)) {
+		RTE_PMD_DEBUG_TRACE("Paramters are invalid\n");
+		return -1;
+	}
+
+	dev = rte_eth_dev_allocate(name);
+	if (dev == NULL) {
+		RTE_PMD_DEBUG_TRACE("%s: rte_eth_dev_allocate failed for %s\n",
Quoted 8 lines

drivers/net/tap/rte_eth_tap.c line 606

Quoted 5 lines
drivers/net/tap/rte_eth_tap.c +5 -0
+				    name, buff);
+		goto error;
+	}
+
+	if (flag & RTE_USE_PRIVATE_DATA) {

You may need to save this flag value somewhere in internals, to decide how to free data later.

Let me look into this one more and see if it is required at all.

Quoted 33 lines
Quoted 5 lines
drivers/net/tap/rte_eth_tap.c +4 -0
+		/*
+		 * now do all data allocation - for eth_dev structure, dummy
+		 * pci driver and internal (private) data
+		 */

drivers/net/tap/rte_eth_tap.c line 620

Quoted 10 lines
drivers/net/tap/rte_eth_tap.c +10 -0
+		snprintf(buff, sizeof(buff), "D-%s-%d", name, numa_node);
+		data = rte_zmalloc_socket(buff, sizeof(struct rte_eth_dev_data),
+					  0, numa_node);
+		if (data == NULL) {
+			RTE_PMD_DEBUG_TRACE("%s: Unable to allocate memory\n",
+					    name);
+			goto error;
+		}
+		/* move the current state of the structure to the new one */
+		rte_memcpy(data, dev->data, sizeof(struct rte_eth_dev_data));

Why do we need to copy, trying to preserve which data?

Quoted 15 lines
drivers/net/tap/rte_eth_tap.c +15 -0
+		dev->data = data;	/* Override the current data pointer */
+	} else
+		data = dev->data;
+
+	snprintf(buff, sizeof(buff), "I-%s-%d", name, numa_node);
+	priv = rte_zmalloc_socket(buff, internal_size, 0, numa_node);
+	if (priv == NULL) {
+		RTE_PMD_DEBUG_TRACE("Unable to allocate internal memory %lu\n",
+				    internal_size);
+		goto error;
+	}
+
+	/* Setup some default values */
+	dev->dev_ops = dev_ops;
+	data->dev_private = priv;
Quoted 4 lines

drivers/net/tap/rte_eth_tap.c line 637

Quoted 2 lines
drivers/net/tap/rte_eth_tap.c +2 -0
+	data->port_id = dev->data->port_id;
+	memmove(data->name, dev->data->name, strlen(dev->data->name));

These two assignments are useless, needs to be done before "dev->data = data" assignment.

Reworked this code area to remove it.

Quoted 34 lines
Quoted 34 lines
drivers/net/tap/rte_eth_tap.c +33 -0
+
+	dev->driver = NULL;
+	data->dev_flags = RTE_ETH_DEV_DETACHABLE;
+	data->kdrv = RTE_KDRV_NONE;
+	data->numa_node = numa_node;
+
+	*eth_dev = dev;
+	*internals = priv;
+
+	return 0;
+error:
+	rte_free(priv);
+
+	if (flag & RTE_USE_PRIVATE_DATA)
+		rte_free(data);
+
+	rte_eth_dev_release_port(dev);
+
+	return -1;
+}
+
+static int
+pmd_init_internals(const char *name, struct tap_info *tap,
+		   struct pmd_internals **internals,
+		   struct rte_eth_dev **eth_dev)
+{
+	struct rte_eth_dev *dev = NULL;
+	struct pmd_internals *internal = NULL;
+	struct rte_eth_dev_data *data = NULL;
+	int ret, i, fd = -1;
+
+	RTE_LOG(INFO, PMD,
+		"%s: Create TUN/TAP Ethernet device with %d queues on numa %u\n",
Quoted 8 lines

drivers/net/tap/rte_eth_tap.c line 676

Quoted 6 lines
drivers/net/tap/rte_eth_tap.c +6 -0
+		name, RTE_PMD_TAP_MAX_QUEUES, rte_socket_id());
+
+	pmd_link.link_speed = tap->speed;
+
+	ret = rte_eth_dev_create(tap->name, &dev, &ops,
+				 (void **)&internal, sizeof(struct pmd_internals),

Why rte_eth_dev_create() get "void **internals" which requires casting, but not "struct pmd_internals **internals” ?

Fixed.

Quoted 15 lines

drivers/net/tap/rte_eth_tap.c line 688

Quoted 13 lines
drivers/net/tap/rte_eth_tap.c +12 -0
+				 RTE_USE_PRIVATE_DATA);
+	if (ret < 0)
+		return -1;
+
+	strncpy(internal->name, tap->name, sizeof(internal->name));
+
+	internal->nb_queues = RTE_PMD_TAP_MAX_QUEUES;
+
+	/* Create the first Tap device */
+	if ((fd = tun_alloc(dev->data->name)) < 0) {
+		RTE_LOG(ERR, PMD, "tun_alloc(%s) failed\n", dev->data->name);
+		rte_free(internal);

rte_free(dev->data); ?
But needs to check RTE_USE_PRIVATE_DATA ..

See above

Quoted 12 lines
Quoted 12 lines
drivers/net/tap/rte_eth_tap.c +11 -0
+		rte_eth_dev_release_port(dev);
+		return -1;
+	}
+
+	/* Presetup the fds to -1 as being not working */
+	for(i = 0; i < RTE_PMD_TAP_MAX_QUEUES; i++) {
+		internal->fds[i] = -1;
+		internal->rxq[i].fd = -1;
+		internal->txq[i].fd = -1;
+	}
+
Quoted 8 lines

drivers/net/tap/rte_eth_tap.c line 706

Quoted 7 lines
drivers/net/tap/rte_eth_tap.c +7 -0
+	/* Take the TUN/TAP fd and place in the first location */
+	internal->rxq[0].fd = fd;
+	internal->txq[0].fd = fd;
+	internal->fds[0] = fd;
+
+	if (pmd_mac_address(fd, dev, &internal->eth_addr) < 0) {
+		rte_free(internal);

rte_free(dev->data); ?

Yes Added.

Quoted 13 lines

drivers/net/tap/rte_eth_tap.c line 717

Quoted 12 lines
drivers/net/tap/rte_eth_tap.c +11 -0
+		rte_eth_dev_release_port(dev);
+		return -1;
+	}
+
+	data = dev->data;
+
+	data->dev_link = pmd_link;
+	data->mac_addrs = &internal->eth_addr;
+
+	data->nb_rx_queues = (uint16_t)internal->nb_queues;
+	data->nb_tx_queues = (uint16_t)internal->nb_queues;

no cast required.

Removed

Quoted 37 lines
Quoted 9 lines
drivers/net/tap/rte_eth_tap.c +8 -0
+	data->drv_name = drivername;
+
+	*eth_dev = dev;
+	*internals = internal;
+
+	return 0;
+}
+

<...>

drivers/net/tap/rte_eth_tap.c line 759

Quoted 5 lines
drivers/net/tap/rte_eth_tap.c +5 -0
+
+static int
+set_interface_speed(const char *key __rte_unused,
+		    const char *value,
+		    void *extra_args __rte_unused)

need to drop __rte_unused for extra_args

Quoted 18 lines
drivers/net/tap/rte_eth_tap.c +18 -0
+{
+	struct tap_info *tap = (struct tap_info *)extra_args;
+
+	pmd_link.link_speed = (value) ? atoi(value) : ETH_SPEED_NUM_10G;
+	tap->speed = pmd_link.link_speed;
+
+	return 0;
+}
+
+/*
+ * Open a TAP interface device.
+ */
+static int
+rte_pmd_tap_devinit(const char *name, const char *params)
+{
+	int ret = 0;
+	struct rte_kvargs *kvlist;
+	struct tap_info tap_info;
Quoted 8 lines

drivers/net/tap/rte_eth_tap.c line 783

Quoted 6 lines
drivers/net/tap/rte_eth_tap.c +6 -0
+
+	/* Setup default values */
+	memset(&tap_info, 0, sizeof(tap_info));
+
+	tap_info.speed = ETH_SPEED_NUM_10G;
+	snprintf(tap_info.name, sizeof(tap_info.name), "dtap%d", tap_unit++);

What about extracting iface name "dtap" into a macro to make it more visible.

Created macro for the default name.

Quoted 8 lines

drivers/net/tap/rte_eth_tap.c line 788

Quoted 6 lines
drivers/net/tap/rte_eth_tap.c +5 -0
+
+	if ((params == NULL) || (params[0] == '\0')) {
+		RTE_LOG(INFO, PMD, "Initializing pmd_tap for %s\n", name);
+
+		ret = eth_dev_tap_create(name, &tap_info);

This "name" is not used at all (except from RTE_LOG), instead tap->name is used for interface name, so why carying this variable around?

Fixed and changed the API.

Quoted 12 lines
Quoted 12 lines
drivers/net/tap/rte_eth_tap.c +11 -0
+		goto leave;
+	}
+
+	RTE_LOG(INFO, PMD, "Initialize %s with params (%s)\n", name, params);
+
+	kvlist = rte_kvargs_parse(params, valid_arguments);
+	if (!kvlist) {
+		ret = eth_dev_tap_create(name, &tap_info);
+		goto leave;
+	}
+
Quoted 8 lines

drivers/net/tap/rte_eth_tap.c line 806

Quoted 7 lines
drivers/net/tap/rte_eth_tap.c +7 -0
+	if (rte_kvargs_count(kvlist, ETH_TAP_SPEED_ARG) == 1) {
+		ret = rte_kvargs_process(kvlist, ETH_TAP_SPEED_ARG,
+					 &set_interface_speed, &tap_info);
+		if (ret < 0)
+			goto leave;
+	} else
+		set_interface_speed(NULL, NULL, &tap_info);

This call is redundant, tap_info already has default speed value set.

Removed

Quoted 11 lines

drivers/net/tap/rte_eth_tap.c line 814

Quoted 9 lines
drivers/net/tap/rte_eth_tap.c +8 -0
+
+	if (rte_kvargs_count(kvlist, ETH_TAP_IFACE_ARG) == 1) {
+		ret = rte_kvargs_process(kvlist, ETH_TAP_IFACE_ARG,
+					 &set_interface_name, &tap_info);
+		if (ret < 0)
+			goto leave;
+	} else
+		set_interface_name(NULL, NULL, (void *)&tap_info);

tap_info->name already set to default value (dtap%d), this call is not required.

Removed

Quoted 19 lines
Quoted 19 lines
drivers/net/tap/rte_eth_tap.c +18 -0
+
+	rte_kvargs_free(kvlist);
+
+leave:
+	if (ret == -1)
+		RTE_LOG(INFO, PMD, "Failed to create pmd_tap for %s\n", name);
+
+	return ret;
+}
+
+/*
+ * detach a TAP device.
+ */
+static int
+rte_pmd_tap_devuninit(const char *name)
+{
+	struct rte_eth_dev *eth_dev = NULL;
+	struct pmd_internals *internals;
Quoted 8 lines

drivers/net/tap/rte_eth_tap.c line 838

Quoted 6 lines
drivers/net/tap/rte_eth_tap.c +6 -0
+	int i;
+
+	RTE_LOG(INFO, PMD, "Closing TUN/TAP Ethernet device on numa %u\n",
+		rte_socket_id());
+
+	if (name == NULL)

This check is redundant, eal layer won't call this function with "name == NULL”

Removed

Quoted 18 lines
Quoted 5 lines
drivers/net/tap/rte_eth_tap.c +4 -0
+		return 0;
+
+	/* find the ethdev entry */
+	eth_dev = rte_eth_dev_allocated(name);

drivers/net/tap/rte_eth_tap.c line 852

Quoted 10 lines
drivers/net/tap/rte_eth_tap.c +10 -0
+	if (eth_dev == NULL)
+		return 0;
+
+	internals = eth_dev->data->dev_private;
+	for (i = 0; i < internals->nb_queues; i++)
+		if (internals->fds[i] != -1)
+			close(internals->fds[i]);
+
+	rte_free(eth_dev->data->dev_private);
+	rte_free(eth_dev->data);

data can be shared?
Don't we need a RTE_USE_PRIVATE_DATA flag check?

Quoted 13 lines

drivers/net/tap/rte_eth_tap.c line 864

Quoted 12 lines
drivers/net/tap/rte_eth_tap.c +12 -0
+
+	rte_eth_dev_release_port(eth_dev);
+
+	return 0;
+}
+
+static struct rte_vdev_driver pmd_tap_drv = {
+	.init = rte_pmd_tap_devinit,
+	.uninit = rte_pmd_tap_devuninit,
+};
+
+DRIVER_REGISTER_VDEV(eth_tap, pmd_tap_drv);

name convention is now: “net_tap"

Fixed

Quoted 15 lines
Quoted 15 lines
drivers/net/tap/rte_eth_tap.c +2 -0
+DRIVER_REGISTER_PARAM_STRING(eth_tap,
+			     "iface=<string>,speed=N");
drivers/net/tap/rte_pmd_tap_version.map new file +4 -0
new file mode 100644
index 0000000..61463bf
@@ -0,0 +1,4 @@
+DPDK_16.11 {
+
+	local: *;
+};
mk/rte.app.mk +0 -0
index 1a0095b..bd1d10f 100644
Quoted 8 lines
Quoted 7 lines
mk/rte.app.mk +0 -0
@@ -129,6 +129,7 @@ ifeq ($(CONFIG_RTE_LIBRTE_VHOST),y)

_LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_VHOST) += -lrte_pmd_vhost
endif # $(CONFIG_RTE_LIBRTE_VHOST)
_LDLIBS-$(CONFIG_RTE_LIBRTE_VMXNET3_PMD) += -lrte_pmd_vmxnet3_uio
+_LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_TAP) += -lrte_pmd_tap

please put in alphebetical order

Done

Quoted 4 lines
Quoted 4 lines

ifeq ($(CONFIG_RTE_LIBRTE_CRYPTODEV),y)
_LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_AESNI_MB) += -lrte_pmd_aesni_mb

#12 lore raw
In reply to Keith Wiles
Quoted 36 lines · Keith Wiles
drivers/net/tap/rte_eth_tap.c new file +36 -0
+/*
+ * Open a TAP interface device.
+ */
+static int
+rte_pmd_tap_devinit(const char *name, const char *params)
+{
+	int ret = 0;
+	struct rte_kvargs *kvlist;
+	struct tap_info tap_info;
+
+	/* Setup default values */
+	memset(&tap_info, 0, sizeof(tap_info));
+
+	tap_info.speed = ETH_SPEED_NUM_10G;
+	snprintf(tap_info.name, sizeof(tap_info.name), "dtap%d", tap_unit++);
+
+	if ((params == NULL) || (params[0] == '\0')) {
+		RTE_LOG(INFO, PMD, "Initializing pmd_tap for %s\n", name);
+
+		ret = eth_dev_tap_create(name, &tap_info);
+		goto leave;
+	}
+
+	RTE_LOG(INFO, PMD, "Initialize %s with params (%s)\n", name, params);
+
+	kvlist = rte_kvargs_parse(params, valid_arguments);
+	if (!kvlist) {
+		ret = eth_dev_tap_create(name, &tap_info);
+		goto leave;
+	}
+
+	if (rte_kvargs_count(kvlist, ETH_TAP_SPEED_ARG) == 1) {
+		ret = rte_kvargs_process(kvlist, ETH_TAP_SPEED_ARG,
+					 &set_interface_speed, &tap_info);
+		if (ret < 0)
+			goto leave;

drivers/net/tap/rte_eth_tap.c line 814

Quoted 10 lines
drivers/net/tap/rte_eth_tap.c new file +10 -0
+	} else
+		set_interface_speed(NULL, NULL, &tap_info);
+
+	if (rte_kvargs_count(kvlist, ETH_TAP_IFACE_ARG) == 1) {
+		ret = rte_kvargs_process(kvlist, ETH_TAP_IFACE_ARG,
+					 &set_interface_name, &tap_info);
+		if (ret < 0)
+			goto leave;
+	} else
+		set_interface_name(NULL, NULL, (void *)&tap_info);

Also there must be a eth_dev_tap_create() call after this point to use tap_info struct with custom values, right?
"--vdev eth_tap0,iface=foo0" parameter shouldn't be working with this code, right?

Quoted 9 lines
drivers/net/tap/rte_eth_tap.c new file +9 -0
+
+	rte_kvargs_free(kvlist);
+
+leave:
+	if (ret == -1)
+		RTE_LOG(INFO, PMD, "Failed to create pmd_tap for %s\n", name);
+
+	return ret;
+}
#13 lore raw

Regards,
Keith

Quoted 46 lines · Ferruh
Quoted 43 lines · Keith Wiles
drivers/net/tap/rte_eth_tap.c new file +43 -0
+/*
+ * Open a TAP interface device.
+ */
+static int
+rte_pmd_tap_devinit(const char *name, const char *params)
+{
+	int ret = 0;
+	struct rte_kvargs *kvlist;
+	struct tap_info tap_info;
+
+	/* Setup default values */
+	memset(&tap_info, 0, sizeof(tap_info));
+
+	tap_info.speed = ETH_SPEED_NUM_10G;
+	snprintf(tap_info.name, sizeof(tap_info.name), "dtap%d", tap_unit++);
+
+	if ((params == NULL) || (params[0] == '\0')) {
+		RTE_LOG(INFO, PMD, "Initializing pmd_tap for %s\n", name);
+
+		ret = eth_dev_tap_create(name, &tap_info);
+		goto leave;
+	}
+
+	RTE_LOG(INFO, PMD, "Initialize %s with params (%s)\n", name, params);
+
+	kvlist = rte_kvargs_parse(params, valid_arguments);
+	if (!kvlist) {
+		ret = eth_dev_tap_create(name, &tap_info);
+		goto leave;
+	}
+
+	if (rte_kvargs_count(kvlist, ETH_TAP_SPEED_ARG) == 1) {
+		ret = rte_kvargs_process(kvlist, ETH_TAP_SPEED_ARG,
+					 &set_interface_speed, &tap_info);
+		if (ret < 0)
+			goto leave;
+	} else
+		set_interface_speed(NULL, NULL, &tap_info);
+
+	if (rte_kvargs_count(kvlist, ETH_TAP_IFACE_ARG) == 1) {
+		ret = rte_kvargs_process(kvlist, ETH_TAP_IFACE_ARG,
+					 &set_interface_name, &tap_info);
+		if (ret < 0)
Quoted 8 lines

drivers/net/tap/rte_eth_tap.c line 814

Quoted 3 lines
drivers/net/tap/rte_eth_tap.c new file +3 -0
+			goto leave;
+	} else
+		set_interface_name(NULL, NULL, (void *)&tap_info);

Also there must be a eth_dev_tap_create() call after this point to use tap_info struct with custom values, right?
"--vdev eth_tap0,iface=foo0" parameter shouldn't be working with this code, right?

Removed the extra code.

Quoted 11 lines
Quoted 10 lines
drivers/net/tap/rte_eth_tap.c new file +9 -0
+
+	rte_kvargs_free(kvlist);
+
+leave:
+	if (ret == -1)
+		RTE_LOG(INFO, PMD, "Failed to create pmd_tap for %s\n", name);
+
+	return ret;
+}
#14 lore raw

PATCH v5 1/1 drivers/net:new PMD using tun/tap host interface

In reply to Keith Wiles

The rte_eth_tap.c PMD creates a device using TUN/TAP interfaces on the local host. The PMD allows for DPDK and the host to communicate using a raw device interface on the host and in the DPDK application. The device created is a Tap device with a L2 packet header.

v5 - merge in changes from list review see related emails.
fixed checkpatch issues and many minor edits
v4 - merge with latest driver changes
v3 - fix includes by removing ifdef for other type besides Linux.
Fix the copyright notice in the Makefile
v2 - merge all of the patches into one patch.
Fix a typo on naming the tap device.
Update the maintainers list

---
 MAINTAINERS                             |   5 +
 config/common_base                      |   5 +
 config/common_linuxapp                  |   1 +
 doc/guides/nics/tap.rst                 | 138 ++++++
 drivers/net/Makefile                    |   1 +
 drivers/net/tap/Makefile                |  57 +++
 drivers/net/tap/rte_eth_tap.c           | 767 ++++++++++++++++++++++++++++++++
 drivers/net/tap/rte_pmd_tap_version.map |   4 +
 mk/rte.app.mk                           |   1 +
 9 files changed, 979 insertions(+)
 create mode 100644 doc/guides/nics/tap.rst
 create mode 100644 drivers/net/tap/Makefile
 create mode 100644 drivers/net/tap/rte_eth_tap.c
 create mode 100644 drivers/net/tap/rte_pmd_tap_version.map
MAINTAINERS +5 -0
index cd8d167..f905709 100644
@@ -394,6 +394,11 @@ F: doc/guides/nics/pcap_ring.rst
 F: app/test/test_pmd_ring.c
 F: app/test/test_pmd_ring_perf.c
 
+Tap PMD
+M: Keith Wiles <[email protected]>
+F: drivers/net/tap
+F: doc/guides/nics/tap.rst
+
 Null Networking PMD
 M: Tetsuya Mukawa <[email protected]>
 F: drivers/net/null/
config/common_base +5 -0
index f5d2eff..356c631 100644
@@ -592,3 +592,8 @@ CONFIG_RTE_APP_TEST_RESOURCE_TAR=n
 CONFIG_RTE_TEST_PMD=y
 CONFIG_RTE_TEST_PMD_RECORD_CORE_CYCLES=n
 CONFIG_RTE_TEST_PMD_RECORD_BURST_STATS=n
+
+#
+# Set TAP PMD to 'n' as it is only supported in Linux for now.
+#
+CONFIG_RTE_LIBRTE_PMD_TAP=n
config/common_linuxapp +1 -0
index 2483dfa..782b503 100644
@@ -44,3 +44,4 @@ CONFIG_RTE_LIBRTE_PMD_VHOST=y
 CONFIG_RTE_LIBRTE_PMD_AF_PACKET=y
 CONFIG_RTE_LIBRTE_POWER=y
 CONFIG_RTE_VIRTIO_USER=y
+CONFIG_RTE_LIBRTE_PMD_TAP=y
doc/guides/nics/tap.rst new file +138 -0

138 lines not on this page · raw message

drivers/net/Makefile +1 -0
index bc93230..e366a85 100644
@@ -51,6 +51,7 @@ DIRS-$(CONFIG_RTE_LIBRTE_PMD_PCAP) += pcap
 DIRS-$(CONFIG_RTE_LIBRTE_QEDE_PMD) += qede
 DIRS-$(CONFIG_RTE_LIBRTE_PMD_RING) += ring
 DIRS-$(CONFIG_RTE_LIBRTE_PMD_SZEDATA2) += szedata2
+DIRS-$(CONFIG_RTE_LIBRTE_PMD_TAP) += tap
 DIRS-$(CONFIG_RTE_LIBRTE_THUNDERX_NICVF_PMD) += thunderx
 DIRS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD) += virtio
 DIRS-$(CONFIG_RTE_LIBRTE_VMXNET3_PMD) += vmxnet3
drivers/net/tap/Makefile new file +57 -0

57 lines not on this page · raw message

drivers/net/tap/rte_eth_tap.c new file +767 -0

767 lines not on this page · raw message

drivers/net/tap/rte_pmd_tap_version.map new file +4 -0
new file mode 100644
index 0000000..61463bf
@@ -0,0 +1,4 @@
+DPDK_16.11 {
+
+	local: *;
+};
mk/rte.app.mk +1 -0
index ac50a21..40d16f7 100644
@@ -123,6 +123,7 @@ _LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_PCAP)       += -lrte_pmd_pcap -lpcap
 _LDLIBS-$(CONFIG_RTE_LIBRTE_QEDE_PMD)       += -lrte_pmd_qede -lz
 _LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_RING)       += -lrte_pmd_ring
 _LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_SZEDATA2)   += -lrte_pmd_szedata2 -lsze2
+_LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_TAP)        += -lrte_pmd_tap
 _LDLIBS-$(CONFIG_RTE_LIBRTE_THUNDERX_NICVF_PMD) += -lrte_pmd_thunderx_nicvf -lm
 _LDLIBS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD)     += -lrte_pmd_virtio
 ifeq ($(CONFIG_RTE_LIBRTE_VHOST),y)
Signature · 1 line
2.8.0.GIT
#15 lore raw
Quoted 17 lines · Keith Wiles

The rte_eth_tap.c PMD creates a device using TUN/TAP interfaces on the local host. The PMD allows for DPDK and the host to communicate using a raw device interface on the host and in the DPDK application. The device created is a Tap device with a L2 packet header.

v5 - merge in changes from list review see related emails.
fixed checkpatch issues and many minor edits
v4 - merge with latest driver changes
v3 - fix includes by removing ifdef for other type besides Linux.
Fix the copyright notice in the Makefile
v2 - merge all of the patches into one patch.
Fix a typo on naming the tap device.
Update the maintainers list

---

<..>

config/common_base CONFIG_RTE_APP_TEST_RESOURCE_TAR=n · line 597

Quoted 11 lines
config/common_base +3 -0
index f5d2eff..356c631 100644
@@ -592,3 +592,8 @@ CONFIG_RTE_APP_TEST_RESOURCE_TAR=n
 CONFIG_RTE_TEST_PMD=y
 CONFIG_RTE_TEST_PMD_RECORD_CORE_CYCLES=n
 CONFIG_RTE_TEST_PMD_RECORD_BURST_STATS=n
+
+#
+# Set TAP PMD to 'n' as it is only supported in Linux for now.

This comments moved to final .config and creates confusion, can we remove it if you don't have a strong option to keep it?

<..>

Quoted 5 lines
doc/guides/nics/tap.rst new file +0 -0
new file mode 100644
index 0000000..eed81ec

<..>

doc/guides/nics/tap.rst line 53

Quoted 4 lines

+.. code-block:: console

doc/guides/nics/tap.rst new file +3 -0
+
+   The interfaced name can be changed by adding the iface=foo0
+   e.g. --vdev=eth_tap,iface=foo0 --vdev=eth_tap,iface=foo1, ...

For all file:
%s/eth_tap/net_tap/g, there are multiple lines with this usage

<..>

Quoted 5 lines
drivers/net/tap/rte_eth_tap.c new file +0 -0
new file mode 100644
index 0000000..c13aa1b

<..>

drivers/net/tap/rte_eth_tap.c line 81

Quoted 5 lines
drivers/net/tap/rte_eth_tap.c new file +5 -0
+
+struct tap_info {
+	char name[RTE_ETH_NAME_MAX_LEN]; /* Interface name supplied/given */
+	int speed;			 /* Speed of interface */
+};

This struct can go away, it is not used, since with the updated code rte_pmd_tap_probe() used tap_name and speed as seperate variables instead of struct.

<..>

drivers/net/tap/rte_eth_tap.c line 178

Quoted 4 lines
drivers/net/tap/rte_eth_tap.c new file +4 -0
+
+	/* If the name is different that new name as default */
+	if (name && strcmp(name, ifr.ifr_name))
+		snprintf(name, RTE_ETH_NAME_MAX_LEN-1, "%s", ifr.ifr_name);

syntax, space around "-"

<..>

drivers/net/tap/rte_eth_tap.c line 324

Quoted 5 lines
drivers/net/tap/rte_eth_tap.c new file +5 -0
+
+static void
+tap_stats_get(struct rte_eth_dev *dev, struct rte_eth_stats *tap_stats)
+{
+	unsigned i, imax;

checkpatch complain about not using "unsigned int"

<..>

Quoted 4 lines
drivers/net/tap/rte_eth_tap.c new file +4 -0
+static int
+tap_setup_queue(struct rte_eth_dev *dev,
+		struct pmd_internals *internals,
+		uint16_t qid)

drivers/net/tap/rte_eth_tap.c line 420

Quoted 10 lines
drivers/net/tap/rte_eth_tap.c new file +10 -0
+{
+	struct rx_queue *rx = &internals->rxq[qid];
+	struct tx_queue *tx = &internals->txq[qid];
+	int fd;
+
+	if ((fd = rx->fd) < 0)
+		if ((fd = tx->fd) < 0) {
+			RTE_LOG(INFO, PMD, "Add queue to TAP %s for qid %d\n",
+				dev->data->name, qid);
+			if ((fd = tun_alloc(dev->data->name)) < 0) {

checkpatch complain about assignment in the if condition

<..>

drivers/net/tap/rte_eth_tap.c line 457

Quoted 2 lines
drivers/net/tap/rte_eth_tap.c new file +2 -0
+	/* Now get the space available for data in the mbuf */
+	buf_size = (uint16_t) (rte_pktmbuf_data_room_size(mp) -

syntax, no space after cast

<..>

drivers/net/tap/rte_eth_tap.c line 609

Quoted 2 lines
drivers/net/tap/rte_eth_tap.c new file +2 -0
+	/* Create the first Tap device */
+	if ((fd = tun_alloc(tap_name)) < 0) {

checkpatch complains about assignment in if condition

drivers/net/tap/rte_eth_tap.c

Quoted 2 lines
drivers/net/tap/rte_eth_tap.c new file +2 -0
+		RTE_LOG(ERR, PMD, "tun_alloc(%s) failed\n", dev->data->name);
+		rte_free(pmd);

rte_free(data) or goto error_exit; ?

Quoted 8 lines
drivers/net/tap/rte_eth_tap.c new file +8 -0
+		rte_eth_dev_release_port(dev);
+		return -EINVAL;
+	}
+
+	/* Presetup the fds to -1 as being not working */
+	for(i = 0; i < RTE_PMD_TAP_MAX_QUEUES; i++) {
+		pmd->fds[i] = -1;
+		pmd->rxq[i].fd = -1;

drivers/net/tap/rte_eth_tap.c line 629

Quoted 10 lines
drivers/net/tap/rte_eth_tap.c new file +10 -0
+		pmd->txq[i].fd = -1;
+	}
+
+	/* Take the TUN/TAP fd and place in the first location */
+	pmd->rxq[0].fd = fd;
+	pmd->txq[0].fd = fd;
+	pmd->fds[0] = fd;
+
+	if (pmd_mac_address(fd, dev, &pmd->eth_addr) < 0) {
+		rte_free(pmd);

rte_free(data) or goto error_exit; ?

<..>

drivers/net/tap/rte_eth_tap.c line 655

Quoted 9 lines
drivers/net/tap/rte_eth_tap.c new file +9 -0
+static int
+set_interface_name(const char *key __rte_unused,
+		   const char *value,
+		   void *extra_args)
+{
+	char *name = (char *)extra_args;
+
+	if (value)
+		snprintf(name, RTE_ETH_NAME_MAX_LEN-1, "%s", value);

syntax, space around "-"

drivers/net/tap/rte_eth_tap.c line 657

Quoted 2 lines
drivers/net/tap/rte_eth_tap.c new file +2 -0
+	else
+		snprintf(name, RTE_ETH_NAME_MAX_LEN-1, "%s%d",

syntax, space around "-"

drivers/net/tap/rte_eth_tap.c line 658

Quoted 1 line
drivers/net/tap/rte_eth_tap.c new file +1 -0
+			 DEFAULT_TAP_NAME, (tap_unit-1));

syntax, space around "-"

drivers/net/tap/rte_eth_tap.c line 668

Quoted 10 lines
drivers/net/tap/rte_eth_tap.c new file +10 -0
+
+	return 0;
+}
+
+static int
+set_interface_speed(const char *key __rte_unused,
+		    const char *value,
+		    void *extra_args)
+{
+	*(int *)extra_args = (value)? atoi(value) : ETH_SPEED_NUM_10G;

syntax, space around "?"

<..>

drivers/net/tap/rte_eth_tap.c line 700

Quoted 7 lines
drivers/net/tap/rte_eth_tap.c new file +7 -0
+		kvlist = rte_kvargs_parse(params, valid_arguments);
+		if (kvlist) {
+			if (rte_kvargs_count(kvlist, ETH_TAP_SPEED_ARG) == 1) {
+				ret = rte_kvargs_process(kvlist,
+							 ETH_TAP_SPEED_ARG,
+							 &set_interface_speed,
+						         &speed);

whitespace, space and tab mixed

drivers/net/tap/rte_eth_tap.c line 709

Quoted 9 lines
drivers/net/tap/rte_eth_tap.c new file +9 -0
+				if (ret == -1)
+					goto leave;
+			}
+
+			if (rte_kvargs_count(kvlist, ETH_TAP_IFACE_ARG) == 1) {
+				ret = rte_kvargs_process(kvlist,
+							 ETH_TAP_IFACE_ARG,
+							 &set_interface_name,
+						         tap_name);

whitespace, space and tab mixed

<..>

drivers/net/tap/rte_eth_tap.c line 744

Quoted 12 lines
drivers/net/tap/rte_eth_tap.c new file +12 -0
+static int
+rte_pmd_tap_remove(const char *name)
+{
+	struct rte_eth_dev *eth_dev = NULL;
+	struct pmd_internals *internals;
+	int i;
+
+	RTE_LOG(INFO, PMD, "Closing TUN/TAP Ethernet device on numa %u\n",
+		rte_socket_id());
+
+	/* find the ethdev entry */
+	eth_dev = rte_eth_dev_allocated(name);

This may cause a problem. Device created by tap_name, but searching with name. I suspenct this will always return NULL.

<..>

Thanks,
ferru

#16 lore raw

Regards,
Keith

Quoted 23 lines · Ferruh
Quoted 17 lines · Keith Wiles

The rte_eth_tap.c PMD creates a device using TUN/TAP interfaces on the local host. The PMD allows for DPDK and the host to communicate using a raw device interface on the host and in the DPDK application. The device created is a Tap device with a L2 packet header.

v5 - merge in changes from list review see related emails.
fixed checkpatch issues and many minor edits
v4 - merge with latest driver changes
v3 - fix includes by removing ifdef for other type besides Linux.
Fix the copyright notice in the Makefile
v2 - merge all of the patches into one patch.
Fix a typo on naming the tap device.
Update the maintainers list

---

<..>

Quoted 14 lines

config/common_base CONFIG_RTE_APP_TEST_RESOURCE_TAR=n · line 597

Quoted 11 lines
config/common_base +0 -0
index f5d2eff..356c631 100644
@@ -592,3 +592,8 @@ CONFIG_RTE_APP_TEST_RESOURCE_TAR=n

CONFIG_RTE_TEST_PMD=y
CONFIG_RTE_TEST_PMD_RECORD_CORE_CYCLES=n
CONFIG_RTE_TEST_PMD_RECORD_BURST_STATS=n

config/common_base +3 -0
+
+#
+# Set TAP PMD to 'n' as it is only supported in Linux for now.

This comments moved to final .config and creates confusion, can we remove it if you don't have a strong option to keep it?

What do you mean, the statement is confusing or causes problems?

Quoted 11 lines

<..>

Quoted 5 lines
doc/guides/nics/tap.rst new file +0 -0
new file mode 100644
index 0000000..eed81ec

<..>

Quoted 7 lines

doc/guides/nics/tap.rst line 53

Quoted 4 lines

+.. code-block:: console

doc/guides/nics/tap.rst new file +3 -0
+
+   The interfaced name can be changed by adding the iface=foo0
+   e.g. --vdev=eth_tap,iface=foo0 --vdev=eth_tap,iface=foo1, ...

For all file:
%s/eth_tap/net_tap/g, there are multiple lines with this usage

Missed that one.

Quoted 12 lines

<..>

Quoted 5 lines
drivers/net/tap/rte_eth_tap.c new file +0 -0
new file mode 100644
index 0000000..c13aa1b

<..>

Quoted 10 lines

drivers/net/tap/rte_eth_tap.c line 81

Quoted 5 lines
drivers/net/tap/rte_eth_tap.c new file +5 -0
+
+struct tap_info {
+	char name[RTE_ETH_NAME_MAX_LEN]; /* Interface name supplied/given */
+	int speed;			 /* Speed of interface */
+};

This struct can go away, it is not used, since with the updated code rte_pmd_tap_probe() used tap_name and speed as seperate variables instead of struct.

OK.

Quoted 12 lines

<..>

drivers/net/tap/rte_eth_tap.c line 178

Quoted 4 lines
drivers/net/tap/rte_eth_tap.c new file +4 -0
+
+	/* If the name is different that new name as default */
+	if (name && strcmp(name, ifr.ifr_name))
+		snprintf(name, RTE_ETH_NAME_MAX_LEN-1, "%s", ifr.ifr_name);

syntax, space around "-"

<..>

Quoted 7 lines

drivers/net/tap/rte_eth_tap.c line 324

Quoted 5 lines
drivers/net/tap/rte_eth_tap.c new file +5 -0
+
+static void
+tap_stats_get(struct rte_eth_dev *dev, struct rte_eth_stats *tap_stats)
+{
+	unsigned i, imax;

checkpatch complain about not using "unsigned int”

I ran checkpatch on the patch and saw no errors reported via the scripts/checkpatch.sh. Which checkpatch are you using here?

Quoted 146 lines

<..>

Quoted 4 lines
drivers/net/tap/rte_eth_tap.c new file +4 -0
+static int
+tap_setup_queue(struct rte_eth_dev *dev,
+		struct pmd_internals *internals,
+		uint16_t qid)

drivers/net/tap/rte_eth_tap.c line 420

Quoted 10 lines
drivers/net/tap/rte_eth_tap.c new file +10 -0
+{
+	struct rx_queue *rx = &internals->rxq[qid];
+	struct tx_queue *tx = &internals->txq[qid];
+	int fd;
+
+	if ((fd = rx->fd) < 0)
+		if ((fd = tx->fd) < 0) {
+			RTE_LOG(INFO, PMD, "Add queue to TAP %s for qid %d\n",
+				dev->data->name, qid);
+			if ((fd = tun_alloc(dev->data->name)) < 0) {

checkpatch complain about assignment in the if condition

<..>

drivers/net/tap/rte_eth_tap.c line 457

Quoted 2 lines
drivers/net/tap/rte_eth_tap.c new file +2 -0
+	/* Now get the space available for data in the mbuf */
+	buf_size = (uint16_t) (rte_pktmbuf_data_room_size(mp) -

syntax, no space after cast

<..>

drivers/net/tap/rte_eth_tap.c line 609

Quoted 2 lines
drivers/net/tap/rte_eth_tap.c new file +2 -0
+	/* Create the first Tap device */
+	if ((fd = tun_alloc(tap_name)) < 0) {

checkpatch complains about assignment in if condition

drivers/net/tap/rte_eth_tap.c

Quoted 2 lines
drivers/net/tap/rte_eth_tap.c new file +2 -0
+		RTE_LOG(ERR, PMD, "tun_alloc(%s) failed\n", dev->data->name);
+		rte_free(pmd);

rte_free(data) or goto error_exit; ?

Quoted 8 lines
drivers/net/tap/rte_eth_tap.c new file +8 -0
+		rte_eth_dev_release_port(dev);
+		return -EINVAL;
+	}
+
+	/* Presetup the fds to -1 as being not working */
+	for(i = 0; i < RTE_PMD_TAP_MAX_QUEUES; i++) {
+		pmd->fds[i] = -1;
+		pmd->rxq[i].fd = -1;

drivers/net/tap/rte_eth_tap.c line 629

Quoted 10 lines
drivers/net/tap/rte_eth_tap.c new file +10 -0
+		pmd->txq[i].fd = -1;
+	}
+
+	/* Take the TUN/TAP fd and place in the first location */
+	pmd->rxq[0].fd = fd;
+	pmd->txq[0].fd = fd;
+	pmd->fds[0] = fd;
+
+	if (pmd_mac_address(fd, dev, &pmd->eth_addr) < 0) {
+		rte_free(pmd);

rte_free(data) or goto error_exit; ?

<..>

drivers/net/tap/rte_eth_tap.c line 655

Quoted 9 lines
drivers/net/tap/rte_eth_tap.c new file +9 -0
+static int
+set_interface_name(const char *key __rte_unused,
+		   const char *value,
+		   void *extra_args)
+{
+	char *name = (char *)extra_args;
+
+	if (value)
+		snprintf(name, RTE_ETH_NAME_MAX_LEN-1, "%s", value);

syntax, space around "-"

drivers/net/tap/rte_eth_tap.c line 657

Quoted 2 lines
drivers/net/tap/rte_eth_tap.c new file +2 -0
+	else
+		snprintf(name, RTE_ETH_NAME_MAX_LEN-1, "%s%d",

syntax, space around "-"

drivers/net/tap/rte_eth_tap.c line 658

Quoted 1 line
drivers/net/tap/rte_eth_tap.c new file +1 -0
+			 DEFAULT_TAP_NAME, (tap_unit-1));

syntax, space around "-"

drivers/net/tap/rte_eth_tap.c line 668

Quoted 10 lines
drivers/net/tap/rte_eth_tap.c new file +10 -0
+
+	return 0;
+}
+
+static int
+set_interface_speed(const char *key __rte_unused,
+		    const char *value,
+		    void *extra_args)
+{
+	*(int *)extra_args = (value)? atoi(value) : ETH_SPEED_NUM_10G;

syntax, space around "?"

<..>

drivers/net/tap/rte_eth_tap.c line 700

Quoted 7 lines
drivers/net/tap/rte_eth_tap.c new file +7 -0
+		kvlist = rte_kvargs_parse(params, valid_arguments);
+		if (kvlist) {
+			if (rte_kvargs_count(kvlist, ETH_TAP_SPEED_ARG) == 1) {
+				ret = rte_kvargs_process(kvlist,
+							 ETH_TAP_SPEED_ARG,
+							 &set_interface_speed,
+						         &speed);

whitespace, space and tab mixed

drivers/net/tap/rte_eth_tap.c line 709

Quoted 9 lines
drivers/net/tap/rte_eth_tap.c new file +9 -0
+				if (ret == -1)
+					goto leave;
+			}
+
+			if (rte_kvargs_count(kvlist, ETH_TAP_IFACE_ARG) == 1) {
+				ret = rte_kvargs_process(kvlist,
+							 ETH_TAP_IFACE_ARG,
+							 &set_interface_name,
+						         tap_name);

whitespace, space and tab mixed

<..>

drivers/net/tap/rte_eth_tap.c line 744

Quoted 12 lines
drivers/net/tap/rte_eth_tap.c new file +12 -0
+static int
+rte_pmd_tap_remove(const char *name)
+{
+	struct rte_eth_dev *eth_dev = NULL;
+	struct pmd_internals *internals;
+	int i;
+
+	RTE_LOG(INFO, PMD, "Closing TUN/TAP Ethernet device on numa %u\n",
+		rte_socket_id());
+
+	/* find the ethdev entry */
+	eth_dev = rte_eth_dev_allocated(name);

This may cause a problem. Device created by tap_name, but searching with name. I suspenct this will always return NULL.

<..>

Thanks,
ferru

#17 lore raw

Regards,
Keith

Quoted 16 lines · Keith

Regards,
Keith

Quoted 9 lines · Ferruh

drivers/net/tap/rte_eth_tap.c line 324

Quoted 6 lines
drivers/net/tap/rte_eth_tap.c new file +5 -0
+
+static void
+tap_stats_get(struct rte_eth_dev *dev, struct rte_eth_stats *tap_stats)
+{
+	unsigned i, imax;

checkpatch complain about not using "unsigned int”

Quoted 1 line

I ran checkpatch on the patch and saw no errors reported via the scripts/checkpatch.sh. Which checkpatch are you using here?

OK, the scripts/checkpatch.sh does not seem to report any failures, but running checkpatch.pl does report problems. Is the scripts/checkpatch.sh script suppose to display the errors or what?

It seems odd for the script file not to display warnings and errors, unless it is just for validating patch. I would expect the script show the problems normally or at least with an option.

#18 lore raw

PATCH v6 1/1 drivers/net:new PMD using tun/tap host interface

In reply to Keith Wiles

The rte_eth_tap.c PMD creates a device using TUN/TAP interfaces on the local host. The PMD allows for DPDK and the host to communicate using a raw device interface on the host and in the DPDK application. The device created is a Tap device with a L2 packet header.

v6 - fixed the checkpatch issues
v5 - merge in changes from list review see related emails
fixed many minor edits
v4 - merge with latest driver changes
v3 - fix includes by removing ifdef for other type besides Linux
Fix the copyright notice in the Makefile
v2 - merge all of the patches into one patch
Fix a typo on naming the tap device
Update the maintainers list

---
 MAINTAINERS                             |   5 +
 config/common_base                      |   5 +
 config/common_linuxapp                  |   1 +
 doc/guides/nics/tap.rst                 | 138 ++++++
 drivers/net/Makefile                    |   1 +
 drivers/net/tap/Makefile                |  57 +++
 drivers/net/tap/rte_eth_tap.c           | 759 ++++++++++++++++++++++++++++++++
 drivers/net/tap/rte_pmd_tap_version.map |   4 +
 mk/rte.app.mk                           |   1 +
 9 files changed, 971 insertions(+)
 create mode 100644 doc/guides/nics/tap.rst
 create mode 100644 drivers/net/tap/Makefile
 create mode 100644 drivers/net/tap/rte_eth_tap.c
 create mode 100644 drivers/net/tap/rte_pmd_tap_version.map
MAINTAINERS +5 -0
index cd8d167..f905709 100644
@@ -394,6 +394,11 @@ F: doc/guides/nics/pcap_ring.rst
 F: app/test/test_pmd_ring.c
 F: app/test/test_pmd_ring_perf.c
 
+Tap PMD
+M: Keith Wiles <[email protected]>
+F: drivers/net/tap
+F: doc/guides/nics/tap.rst
+
 Null Networking PMD
 M: Tetsuya Mukawa <[email protected]>
 F: drivers/net/null/
config/common_base +5 -0
index f5d2eff..356c631 100644
@@ -592,3 +592,8 @@ CONFIG_RTE_APP_TEST_RESOURCE_TAR=n
 CONFIG_RTE_TEST_PMD=y
 CONFIG_RTE_TEST_PMD_RECORD_CORE_CYCLES=n
 CONFIG_RTE_TEST_PMD_RECORD_BURST_STATS=n
+
+#
+# Set TAP PMD to 'n' as it is only supported in Linux for now.
+#
+CONFIG_RTE_LIBRTE_PMD_TAP=n
config/common_linuxapp +1 -0
index 2483dfa..782b503 100644
@@ -44,3 +44,4 @@ CONFIG_RTE_LIBRTE_PMD_VHOST=y
 CONFIG_RTE_LIBRTE_PMD_AF_PACKET=y
 CONFIG_RTE_LIBRTE_POWER=y
 CONFIG_RTE_VIRTIO_USER=y
+CONFIG_RTE_LIBRTE_PMD_TAP=y
doc/guides/nics/tap.rst new file +138 -0

138 lines not on this page · raw message

drivers/net/Makefile +1 -0
index bc93230..e366a85 100644
@@ -51,6 +51,7 @@ DIRS-$(CONFIG_RTE_LIBRTE_PMD_PCAP) += pcap
 DIRS-$(CONFIG_RTE_LIBRTE_QEDE_PMD) += qede
 DIRS-$(CONFIG_RTE_LIBRTE_PMD_RING) += ring
 DIRS-$(CONFIG_RTE_LIBRTE_PMD_SZEDATA2) += szedata2
+DIRS-$(CONFIG_RTE_LIBRTE_PMD_TAP) += tap
 DIRS-$(CONFIG_RTE_LIBRTE_THUNDERX_NICVF_PMD) += thunderx
 DIRS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD) += virtio
 DIRS-$(CONFIG_RTE_LIBRTE_VMXNET3_PMD) += vmxnet3
drivers/net/tap/Makefile new file +57 -0

57 lines not on this page · raw message

drivers/net/tap/rte_eth_tap.c new file +759 -0

759 lines not on this page · raw message

drivers/net/tap/rte_pmd_tap_version.map new file +4 -0
new file mode 100644
index 0000000..61463bf
@@ -0,0 +1,4 @@
+DPDK_16.11 {
+
+	local: *;
+};
mk/rte.app.mk +1 -0
index ac50a21..40d16f7 100644
@@ -123,6 +123,7 @@ _LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_PCAP)       += -lrte_pmd_pcap -lpcap
 _LDLIBS-$(CONFIG_RTE_LIBRTE_QEDE_PMD)       += -lrte_pmd_qede -lz
 _LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_RING)       += -lrte_pmd_ring
 _LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_SZEDATA2)   += -lrte_pmd_szedata2 -lsze2
+_LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_TAP)        += -lrte_pmd_tap
 _LDLIBS-$(CONFIG_RTE_LIBRTE_THUNDERX_NICVF_PMD) += -lrte_pmd_thunderx_nicvf -lm
 _LDLIBS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD)     += -lrte_pmd_virtio
 ifeq ($(CONFIG_RTE_LIBRTE_VHOST),y)
Signature · 1 line
2.8.0.GIT
#19 lore raw
Quoted 18 lines · Keith Wiles

The rte_eth_tap.c PMD creates a device using TUN/TAP interfaces on the local host. The PMD allows for DPDK and the host to communicate using a raw device interface on the host and in the DPDK application. The device created is a Tap device with a L2 packet header.

v6 - fixed the checkpatch issues
v5 - merge in changes from list review see related emails
fixed many minor edits
v4 - merge with latest driver changes
v3 - fix includes by removing ifdef for other type besides Linux
Fix the copyright notice in the Makefile
v2 - merge all of the patches into one patch
Fix a typo on naming the tap device
Update the maintainers list

---

<...>

Quoted 1 line
config/common_base +0 -0

...

config/common_base CONFIG_RTE_APP_TEST_RESOURCE_TAR=n · line 598

Quoted 4 lines
config/common_base +4 -0
+
+#
+# Set TAP PMD to 'n' as it is only supported in Linux for now.
+#

In final .config file, this comment says TAP PMD set to "n", but it is set to "y"
Also this behavior is not unique to this config option, many config items used as described in the comment, and they don't have same comment.

<...>

Quoted 1 line
doc/guides/nics/tap.rst +0 -0

...

doc/guides/nics/tap.rst line 59

Quoted 4 lines
doc/guides/nics/tap.rst +4 -0
+
+   Also the speed of the interface can be changed from 10G to whatever number
+   needed, but the interface does not enforce that speed.
+   e.g. --vdev=eth_tap,iface=foo0,speed=25000

Same comment with previous review, eth_tap should be net_tap, for all occurrences in this document

<...>

Quoted 1 line
drivers/net/tap/rte_eth_tap.c +0 -0

...

drivers/net/tap/rte_eth_tap.c line 81

Quoted 4 lines
drivers/net/tap/rte_eth_tap.c +4 -0
+struct tap_info {
+	char name[RTE_ETH_NAME_MAX_LEN]; /* Interface name supplied/given */
+	int speed;			 /* Speed of interface */
+};

Same comment with previous review, this struct is not used at all.

<...>

Quoted 2 lines
drivers/net/tap/rte_eth_tap.c +2 -0
+static int
+eth_dev_tap_create(const char *name, char *tap_name)

...

drivers/net/tap/rte_eth_tap.c line 606

Quoted 6 lines
drivers/net/tap/rte_eth_tap.c +6 -0
+	dev->data = data;
+	dev->dev_ops = &ops;
+	dev->driver = NULL;
+	dev->rx_pkt_burst = pmd_rx_burst;
+	dev->tx_pkt_burst = pmd_tx_burst;
+	snprintf(dev->data->name, sizeof(dev->data->name), "%s", tap_name);

[1]

<...>

Quoted 3 lines

+ */

drivers/net/tap/rte_eth_tap.c +2 -0
+static int
+rte_pmd_tap_remove(const char *name)

...

drivers/net/tap/rte_eth_tap.c line 736

Quoted 3 lines
drivers/net/tap/rte_eth_tap.c +3 -0
+
+	/* find the ethdev entry */
+	eth_dev = rte_eth_dev_allocated(name);

This still won't work, please see [1], dev->name overwritten by tap_name, and this function does simply a strcmp with name and dev->name.

<...>

Thanks,
ferruh

#20 lore raw

PATCH v7 1/1 drivers/net:new PMD using tun/tap host interface

In reply to Keith Wiles

The rte_eth_tap.c PMD creates a device using TUN/TAP interfaces on the local host. The PMD allows for DPDK and the host to communicate using a raw device interface on the host and in the DPDK application. The device created is a Tap device with a L2 packet header.

v7 - Reword the comment in common_base and fix the data->name issue v6 - fixed the checkpatch issues
v5 - merge in changes from list review see related emails
fixed many minor edits
v4 - merge with latest driver changes
v3 - fix includes by removing ifdef for other type besides Linux
Fix the copyright notice in the Makefile
v2 - merge all of the patches into one patch
Fix a typo on naming the tap device
Update the maintainers list

---
 MAINTAINERS                             |   5 +
 config/common_base                      |   9 +
 config/common_linuxapp                  |   1 +
 doc/guides/nics/tap.rst                 | 138 ++++++
 drivers/net/Makefile                    |   1 +
 drivers/net/tap/Makefile                |  57 +++
 drivers/net/tap/rte_eth_tap.c           | 754 ++++++++++++++++++++++++++++++++
 drivers/net/tap/rte_pmd_tap_version.map |   4 +
 mk/rte.app.mk                           |   1 +
 9 files changed, 970 insertions(+)
 create mode 100644 doc/guides/nics/tap.rst
 create mode 100644 drivers/net/tap/Makefile
 create mode 100644 drivers/net/tap/rte_eth_tap.c
 create mode 100644 drivers/net/tap/rte_pmd_tap_version.map
MAINTAINERS +5 -0
index cd8d167..f905709 100644
@@ -394,6 +394,11 @@ F: doc/guides/nics/pcap_ring.rst
 F: app/test/test_pmd_ring.c
 F: app/test/test_pmd_ring_perf.c
 
+Tap PMD
+M: Keith Wiles <[email protected]>
+F: drivers/net/tap
+F: doc/guides/nics/tap.rst
+
 Null Networking PMD
 M: Tetsuya Mukawa <[email protected]>
 F: drivers/net/null/
config/common_base +9 -0
index f5d2eff..47ef843 100644
@@ -592,3 +592,12 @@ CONFIG_RTE_APP_TEST_RESOURCE_TAR=n
 CONFIG_RTE_TEST_PMD=y
 CONFIG_RTE_TEST_PMD_RECORD_CORE_CYCLES=n
 CONFIG_RTE_TEST_PMD_RECORD_BURST_STATS=n
+
+#
+# Compile the TAP PMD
+#
+# The TAP PMD is currently only built for Linux and the
+# option is enabled by default in common_linuxapp file,
+# set to 'n' in the common_base file.
+#
+CONFIG_RTE_LIBRTE_PMD_TAP=n
config/common_linuxapp +1 -0
index 2483dfa..782b503 100644
@@ -44,3 +44,4 @@ CONFIG_RTE_LIBRTE_PMD_VHOST=y
 CONFIG_RTE_LIBRTE_PMD_AF_PACKET=y
 CONFIG_RTE_LIBRTE_POWER=y
 CONFIG_RTE_VIRTIO_USER=y
+CONFIG_RTE_LIBRTE_PMD_TAP=y
doc/guides/nics/tap.rst new file +138 -0

138 lines not on this page · raw message

drivers/net/Makefile +1 -0
index bc93230..e366a85 100644
@@ -51,6 +51,7 @@ DIRS-$(CONFIG_RTE_LIBRTE_PMD_PCAP) += pcap
 DIRS-$(CONFIG_RTE_LIBRTE_QEDE_PMD) += qede
 DIRS-$(CONFIG_RTE_LIBRTE_PMD_RING) += ring
 DIRS-$(CONFIG_RTE_LIBRTE_PMD_SZEDATA2) += szedata2
+DIRS-$(CONFIG_RTE_LIBRTE_PMD_TAP) += tap
 DIRS-$(CONFIG_RTE_LIBRTE_THUNDERX_NICVF_PMD) += thunderx
 DIRS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD) += virtio
 DIRS-$(CONFIG_RTE_LIBRTE_VMXNET3_PMD) += vmxnet3
drivers/net/tap/Makefile new file +57 -0

57 lines not on this page · raw message

drivers/net/tap/rte_eth_tap.c new file +754 -0

754 lines not on this page · raw message

drivers/net/tap/rte_pmd_tap_version.map new file +4 -0
new file mode 100644
index 0000000..61463bf
@@ -0,0 +1,4 @@
+DPDK_16.11 {
+
+	local: *;
+};
mk/rte.app.mk +1 -0
index ac50a21..40d16f7 100644
@@ -123,6 +123,7 @@ _LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_PCAP)       += -lrte_pmd_pcap -lpcap
 _LDLIBS-$(CONFIG_RTE_LIBRTE_QEDE_PMD)       += -lrte_pmd_qede -lz
 _LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_RING)       += -lrte_pmd_ring
 _LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_SZEDATA2)   += -lrte_pmd_szedata2 -lsze2
+_LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_TAP)        += -lrte_pmd_tap
 _LDLIBS-$(CONFIG_RTE_LIBRTE_THUNDERX_NICVF_PMD) += -lrte_pmd_thunderx_nicvf -lm
 _LDLIBS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD)     += -lrte_pmd_virtio
 ifeq ($(CONFIG_RTE_LIBRTE_VHOST),y)
Signature · 1 line
2.8.0.GIT
#21 lore raw

PATCH v8 1/1 drivers/net:new PMD using tun/tap host interface

In reply to Keith Wiles

The rte_eth_tap.c PMD creates a device using TUN/TAP interfaces on the local host. The PMD allows for DPDK and the host to communicate using a raw device interface on the host and in the DPDK application. The device created is a Tap device with a L2 packet header.

v8 - Fix issue with tap_tx_queue_setup() not return zero on success. v7 - Reword the comment in common_base and fix the data->name issue v6 - fixed the checkpatch issues
v5 - merge in changes from list review see related emails
fixed many minor edits
v4 - merge with latest driver changes
v3 - fix includes by removing ifdef for other type besides Linux
Fix the copyright notice in the Makefile
v2 - merge all of the patches into one patch
Fix a typo on naming the tap device
Update the maintainers list

---
 MAINTAINERS                             |   5 +
 config/common_base                      |   9 +
 config/common_linuxapp                  |   1 +
 doc/guides/nics/tap.rst                 | 138 ++++++
 drivers/net/Makefile                    |   1 +
 drivers/net/tap/Makefile                |  57 +++
 drivers/net/tap/rte_eth_tap.c           | 756 ++++++++++++++++++++++++++++++++
 drivers/net/tap/rte_pmd_tap_version.map |   4 +
 mk/rte.app.mk                           |   1 +
 9 files changed, 972 insertions(+)
 create mode 100644 doc/guides/nics/tap.rst
 create mode 100644 drivers/net/tap/Makefile
 create mode 100644 drivers/net/tap/rte_eth_tap.c
 create mode 100644 drivers/net/tap/rte_pmd_tap_version.map
MAINTAINERS +5 -0
index cd8d167..f905709 100644
@@ -394,6 +394,11 @@ F: doc/guides/nics/pcap_ring.rst
 F: app/test/test_pmd_ring.c
 F: app/test/test_pmd_ring_perf.c
 
+Tap PMD
+M: Keith Wiles <[email protected]>
+F: drivers/net/tap
+F: doc/guides/nics/tap.rst
+
 Null Networking PMD
 M: Tetsuya Mukawa <[email protected]>
 F: drivers/net/null/
config/common_base +9 -0
index f5d2eff..47ef843 100644
@@ -592,3 +592,12 @@ CONFIG_RTE_APP_TEST_RESOURCE_TAR=n
 CONFIG_RTE_TEST_PMD=y
 CONFIG_RTE_TEST_PMD_RECORD_CORE_CYCLES=n
 CONFIG_RTE_TEST_PMD_RECORD_BURST_STATS=n
+
+#
+# Compile the TAP PMD
+#
+# The TAP PMD is currently only built for Linux and the
+# option is enabled by default in common_linuxapp file,
+# set to 'n' in the common_base file.
+#
+CONFIG_RTE_LIBRTE_PMD_TAP=n
config/common_linuxapp +1 -0
index 2483dfa..782b503 100644
@@ -44,3 +44,4 @@ CONFIG_RTE_LIBRTE_PMD_VHOST=y
 CONFIG_RTE_LIBRTE_PMD_AF_PACKET=y
 CONFIG_RTE_LIBRTE_POWER=y
 CONFIG_RTE_VIRTIO_USER=y
+CONFIG_RTE_LIBRTE_PMD_TAP=y
doc/guides/nics/tap.rst new file +138 -0

138 lines not on this page · raw message

drivers/net/Makefile +1 -0
index bc93230..e366a85 100644
@@ -51,6 +51,7 @@ DIRS-$(CONFIG_RTE_LIBRTE_PMD_PCAP) += pcap
 DIRS-$(CONFIG_RTE_LIBRTE_QEDE_PMD) += qede
 DIRS-$(CONFIG_RTE_LIBRTE_PMD_RING) += ring
 DIRS-$(CONFIG_RTE_LIBRTE_PMD_SZEDATA2) += szedata2
+DIRS-$(CONFIG_RTE_LIBRTE_PMD_TAP) += tap
 DIRS-$(CONFIG_RTE_LIBRTE_THUNDERX_NICVF_PMD) += thunderx
 DIRS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD) += virtio
 DIRS-$(CONFIG_RTE_LIBRTE_VMXNET3_PMD) += vmxnet3
drivers/net/tap/Makefile new file +57 -0

57 lines not on this page · raw message

drivers/net/tap/rte_eth_tap.c new file +756 -0

756 lines not on this page · raw message

drivers/net/tap/rte_pmd_tap_version.map new file +4 -0
new file mode 100644
index 0000000..61463bf
@@ -0,0 +1,4 @@
+DPDK_16.11 {
+
+	local: *;
+};
mk/rte.app.mk +1 -0
index ac50a21..40d16f7 100644
@@ -123,6 +123,7 @@ _LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_PCAP)       += -lrte_pmd_pcap -lpcap
 _LDLIBS-$(CONFIG_RTE_LIBRTE_QEDE_PMD)       += -lrte_pmd_qede -lz
 _LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_RING)       += -lrte_pmd_ring
 _LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_SZEDATA2)   += -lrte_pmd_szedata2 -lsze2
+_LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_TAP)        += -lrte_pmd_tap
 _LDLIBS-$(CONFIG_RTE_LIBRTE_THUNDERX_NICVF_PMD) += -lrte_pmd_thunderx_nicvf -lm
 _LDLIBS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD)     += -lrte_pmd_virtio
 ifeq ($(CONFIG_RTE_LIBRTE_VHOST),y)
Signature · 1 line
2.8.0.GIT
#22 lore raw

Hi,

Some doc comments below. Apologies, for the late review, I didn't see the docs inline.

Quoted 27 lines

-----Original Message-----
From: dev [mailto:[email protected]] On Behalf Of Keith Wiles
Sent: Thursday, October 13, 2016 5:11 PM
To: [email protected]
Cc: [email protected]; [email protected]; Yigit, Ferruh <[email protected]>
Subject: [dpdk-dev] [PATCH v8] drivers/net:new PMD using tun/tap host interface

The rte_eth_tap.c PMD creates a device using TUN/TAP interfaces on the local host. The PMD allows for DPDK and the host to communicate using a raw device interface on the host and in the DPDK application. The device created is a Tap device with a L2 packet header.

v8 - Fix issue with tap_tx_queue_setup() not return zero on success. v7 - Reword the comment in common_base and fix the data->name issue v6 - fixed the checkpatch issues
v5 - merge in changes from list review see related emails
fixed many minor edits
v4 - merge with latest driver changes
v3 - fix includes by removing ifdef for other type besides Linux
Fix the copyright notice in the Makefile
v2 - merge all of the patches into one patch
Fix a typo on naming the tap device
Update the maintainers list

Quoted 8 lines
---
 MAINTAINERS                             |   5 +
 config/common_base                      |   9 +
 config/common_linuxapp                  |   1 +
 doc/guides/nics/tap.rst                 | 138 ++++++
 drivers/net/Makefile                    |   1 +
 drivers/net/tap/Makefile                |  57 +++
 drivers/net/tap/rte_eth_tap.c           | 756

You need to add "tap" to the doc/guides/nics/index.rst file to include the tap.rst in the docs.

Quoted 13 lines
(quoted diff) +13 -0
+
+Tun/Tap Poll Mode Driver
+========================================
+
+The rte_eth_tap.c PMD creates a device using TUN/TAP interfaces on the
+local host. The PMD allows for DPDK and the host to communicate using a
+raw device interface on the host and in the DPDK application.
+
+The device created is a TAP device, which sends/receives packet in a
+raw format with a L2 header. The usage for a TAP PMD is for
+connectivity to the local host using a TAP interface. When the TAP PMD
+is initialized it will create a number of tap devices in the host
+accessed via 'ifconfig -a' or 'ip' command. The commands can be used to
Quoted 1 line

assign and query the virtual like device.

The apostrophes around the commands should be replaced with `` to render them as fixed width.

Quoted 7 lines
(quoted diff) +6 -0
+
+These TAP interfaces can be used with wireshark or tcpdump or
+Pktgen-DPDK along with being able to be used as a network connection to
+the DPDK application. The method enable one or more interfaces is to
+use the --vdev=net_tap option on the DPDK application  command line.
+Each --vdev=net_tap option give will create an interface named dtap0,

dtap1, ... and so forth.

Same here, include any commands or variable names in backticks to render them as fixed width in the text: ``--vdev=net_tap``

doc/guides/nics/tap.rst line 53

Quoted 5 lines
doc/guides/nics/tap.rst new file +5 -0
+
+.. code-block:: console
+
+   The interfaced name can be changed by adding the iface=foo0
+   e.g. --vdev=net_tap,iface=foo0 --vdev=net_tap,iface=foo1, ...

This would be better formatted as follows:

The interfaced name can be changed by adding the ``iface=foo0``, for example::

   --vdev=eth_tap,iface=foo0 --vdev=eth_tap,iface=foo1, ...
Quoted 9 lines
(quoted diff) +3 -0
+If you have a Network Stack in your DPDK application or something like
+it you can utilize that stack to handle the network protocols. Plus you
+would be able to address the interface using an IP address assigned to

the internal interface.

doc/guides/nics/tap.rst new file +4 -0
+
+A very crude test you can do the following:
+
+Apply the patch below and make sure you have socat installed on your

system.

There is no patch below in the docs. ;-) Also, this would probably be better as a new section. Something like:

Example
-------

The following is a simple example of using the TUN/TAP PMD with the Pktgen packet generator. It requires that the ``socat`` utility is installed on the test system.

Build DPDK, then pull down Pktgen and build pktgen using the DPDK SDK/Target used to build the dpdk you pulled down.

...

Quoted 8 lines
doc/guides/nics/tap.rst new file +6 -0
+
+Build DPDK, then pull down Pktgen and build pktgen using the DPDK
+SDK/Target used to build the dpdk you pulled down.
+
+Run pktgen from the pktgen repo directory in an xterm:
+    Note: change the -b options to blacklist all of your physical ports.

The
+ following command line is all one line.

The RST syntax for Note and the indentation are wrong here. Also the note would be better after the example. Something like:

.. Note:

Change the ``-b`` options to blacklist all of your physical ports. The following command line is all one line.

Also, ``-f themes/black-yellow.theme`` is optional if the default colors work on your system configuration. See the Pktgen docs for more information.

Finally, if you want to include the TAP PMD in the NIC overview table:

http://dpdk.org/doc/guides/nics/overview.html

You need to include a feature .ini file in the following dir:

http://dpdk.org/browse/dpdk/tree/doc/guides/nics/features

Use the default as an example:

http://dpdk.org/browse/dpdk/tree/doc/guides/nics/features/default.ini

Apart from those small changes it is a nice introduction and example code.

John

#23 lore raw

PATCH v9 1/1 drivers/net:new PMD using tun/tap host interface

In reply to Keith Wiles

The rte_eth_tap.c PMD creates a device using TUN/TAP interfaces on the local host. The PMD allows for DPDK and the host to communicate using a raw device interface on the host and in the DPDK application. The device created is a Tap device with a L2 packet header.

v9 - Fix up the docs to use correct syntax
v8 - Fix issue with tap_tx_queue_setup() not return zero on success. v7 - Reword the comment in common_base and fix the data->name issue v6 - fixed the checkpatch issues
v5 - merge in changes from list review see related emails
fixed many minor edits
v4 - merge with latest driver changes
v3 - fix includes by removing ifdef for other type besides Linux
Fix the copyright notice in the Makefile
v2 - merge all of the patches into one patch
Fix a typo on naming the tap device
Update the maintainers list

---
 MAINTAINERS                             |   5 +
 config/common_base                      |   9 +
 config/common_linuxapp                  |   1 +
 doc/guides/nics/index.rst               |   1 +
 doc/guides/nics/tap.rst                 | 136 ++++++
 drivers/net/Makefile                    |   1 +
 drivers/net/tap/Makefile                |  57 +++
 drivers/net/tap/rte_eth_tap.c           | 756 ++++++++++++++++++++++++++++++++
 drivers/net/tap/rte_pmd_tap_version.map |   4 +
 mk/rte.app.mk                           |   1 +
 10 files changed, 971 insertions(+)
 create mode 100644 doc/guides/nics/tap.rst
 create mode 100644 drivers/net/tap/Makefile
 create mode 100644 drivers/net/tap/rte_eth_tap.c
 create mode 100644 drivers/net/tap/rte_pmd_tap_version.map
MAINTAINERS +5 -0
index 8f5fa82..433d402 100644
@@ -394,6 +394,11 @@ F: doc/guides/nics/pcap_ring.rst
 F: app/test/test_pmd_ring.c
 F: app/test/test_pmd_ring_perf.c
 
+Tap PMD
+M: Keith Wiles <[email protected]>
+F: drivers/net/tap
+F: doc/guides/nics/tap.rst
+
 Null Networking PMD
 M: Tetsuya Mukawa <[email protected]>
 F: drivers/net/null/
config/common_base +9 -0
index f5d2eff..47ef843 100644
@@ -592,3 +592,12 @@ CONFIG_RTE_APP_TEST_RESOURCE_TAR=n
 CONFIG_RTE_TEST_PMD=y
 CONFIG_RTE_TEST_PMD_RECORD_CORE_CYCLES=n
 CONFIG_RTE_TEST_PMD_RECORD_BURST_STATS=n
+
+#
+# Compile the TAP PMD
+#
+# The TAP PMD is currently only built for Linux and the
+# option is enabled by default in common_linuxapp file,
+# set to 'n' in the common_base file.
+#
+CONFIG_RTE_LIBRTE_PMD_TAP=n
config/common_linuxapp +1 -0
index 2483dfa..782b503 100644
@@ -44,3 +44,4 @@ CONFIG_RTE_LIBRTE_PMD_VHOST=y
 CONFIG_RTE_LIBRTE_PMD_AF_PACKET=y
 CONFIG_RTE_LIBRTE_POWER=y
 CONFIG_RTE_VIRTIO_USER=y
+CONFIG_RTE_LIBRTE_PMD_TAP=y
doc/guides/nics/index.rst +1 -0
index 92d56a5..f676a52 100644
@@ -56,6 +56,7 @@ Network Interface Controller Drivers
     vhost
     vmxnet3
     pcap_ring
+    tap
 
 **Figures**
 
doc/guides/nics/tap.rst new file +136 -0

136 lines not on this page · raw message

drivers/net/Makefile +1 -0
index bc93230..e366a85 100644
@@ -51,6 +51,7 @@ DIRS-$(CONFIG_RTE_LIBRTE_PMD_PCAP) += pcap
 DIRS-$(CONFIG_RTE_LIBRTE_QEDE_PMD) += qede
 DIRS-$(CONFIG_RTE_LIBRTE_PMD_RING) += ring
 DIRS-$(CONFIG_RTE_LIBRTE_PMD_SZEDATA2) += szedata2
+DIRS-$(CONFIG_RTE_LIBRTE_PMD_TAP) += tap
 DIRS-$(CONFIG_RTE_LIBRTE_THUNDERX_NICVF_PMD) += thunderx
 DIRS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD) += virtio
 DIRS-$(CONFIG_RTE_LIBRTE_VMXNET3_PMD) += vmxnet3
drivers/net/tap/Makefile new file +57 -0

57 lines not on this page · raw message

drivers/net/tap/rte_eth_tap.c new file +756 -0

756 lines not on this page · raw message

drivers/net/tap/rte_pmd_tap_version.map new file +4 -0
new file mode 100644
index 0000000..61463bf
@@ -0,0 +1,4 @@
+DPDK_16.11 {
+
+	local: *;
+};
mk/rte.app.mk +1 -0
index 72c2fe7..f20fc1a 100644
@@ -124,6 +124,7 @@ _LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_PCAP)       += -lrte_pmd_pcap -lpcap
 _LDLIBS-$(CONFIG_RTE_LIBRTE_QEDE_PMD)       += -lrte_pmd_qede -lz
 _LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_RING)       += -lrte_pmd_ring
 _LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_SZEDATA2)   += -lrte_pmd_szedata2 -lsze2
+_LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_TAP)        += -lrte_pmd_tap
 _LDLIBS-$(CONFIG_RTE_LIBRTE_THUNDERX_NICVF_PMD) += -lrte_pmd_thunderx_nicvf -lm
 _LDLIBS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD)     += -lrte_pmd_virtio
 ifeq ($(CONFIG_RTE_LIBRTE_VHOST),y)
Signature · 1 line
2.8.0.GIT
Acked-by on v9 1/1
#24 lore raw
Quoted 29 lines

-----Original Message-----
From: dev [mailto:[email protected]] On Behalf Of Keith Wiles
Sent: Thursday, October 13, 2016 11:04 PM
To: [email protected]
Cc: [email protected]; [email protected]; Yigit, Ferruh <[email protected]>
Subject: [dpdk-dev] [PATCH v9] drivers/net:new PMD using tun/tap host interface

The rte_eth_tap.c PMD creates a device using TUN/TAP interfaces on the local host. The PMD allows for DPDK and the host to communicate using a raw device interface on the host and in the DPDK application. The device created is a Tap device with a L2 packet header.

v9 - Fix up the docs to use correct syntax
v8 - Fix issue with tap_tx_queue_setup() not return zero on success. v7 - Reword the comment in common_base and fix the data->name issue v6 - fixed the checkpatch issues
v5 - merge in changes from list review see related emails
fixed many minor edits
v4 - merge with latest driver changes
v3 - fix includes by removing ifdef for other type besides Linux
Fix the copyright notice in the Makefile
v2 - merge all of the patches into one patch
Fix a typo on naming the tap device
Update the maintainers list

Quoted 1 line

For the doc part of the patch:

Reviewed-by on v9 1/1
#25 lore raw
In reply to Keith Wiles
Quoted 19 lines · Keith Wiles

The rte_eth_tap.c PMD creates a device using TUN/TAP interfaces on the local host. The PMD allows for DPDK and the host to communicate using a raw device interface on the host and in the DPDK application. The device created is a Tap device with a L2 packet header.

v9 - Fix up the docs to use correct syntax
v8 - Fix issue with tap_tx_queue_setup() not return zero on success. v7 - Reword the comment in common_base and fix the data->name issue v6 - fixed the checkpatch issues
v5 - merge in changes from list review see related emails
fixed many minor edits
v4 - merge with latest driver changes
v3 - fix includes by removing ifdef for other type besides Linux
Fix the copyright notice in the Makefile
v2 - merge all of the patches into one patch
Fix a typo on naming the tap device
Update the maintainers list

Quoted 2 lines

---

#26 lore raw
In reply to Keith Wiles
Quoted 19 lines · Keith Wiles

The rte_eth_tap.c PMD creates a device using TUN/TAP interfaces on the local host. The PMD allows for DPDK and the host to communicate using a raw device interface on the host and in the DPDK application. The device created is a Tap device with a L2 packet header.

v9 - Fix up the docs to use correct syntax
v8 - Fix issue with tap_tx_queue_setup() not return zero on success. v7 - Reword the comment in common_base and fix the data->name issue v6 - fixed the checkpatch issues
v5 - merge in changes from list review see related emails
fixed many minor edits
v4 - merge with latest driver changes
v3 - fix includes by removing ifdef for other type besides Linux
Fix the copyright notice in the Makefile
v2 - merge all of the patches into one patch
Fix a typo on naming the tap device
Update the maintainers list

Quoted 2 lines

---

Just a reminder, this is a new PMD and waiting for community review.

#27 lore raw

Keith,

This won't build when integrated with v16.11. The register macro prefix has been renamed. a v10 is needed.

drivers/net/tap/rte_eth_tap.c +3 -2
index 7f303db..297d4b6 100644
@@ -752,5 +752,6 @@ static struct rte_vdev_driver pmd_tap_drv = {
        .remove = rte_pmd_tap_remove,
 };
 
-DRIVER_REGISTER_VDEV(net_tap, pmd_tap_drv);
-DRIVER_REGISTER_PARAM_STRING(net_tap, "iface=<string>,speed=N");
+RTE_PMD_REGISTER_VDEV(net_tap, pmd_tap_drv);
+RTE_PMD_REGISTER_ALIAS(net_tap, eth_tap);
+RTE_PMD_REGISTER_PARAM_STRING(net_tap, "iface=<string>,speed=N");
Quoted 24 lines · Ferruh Yigit
Quoted 19 lines · Keith Wiles

The rte_eth_tap.c PMD creates a device using TUN/TAP interfaces on the local host. The PMD allows for DPDK and the host to communicate using a raw device interface on the host and in the DPDK application. The device created is a Tap device with a L2 packet header.

v9 - Fix up the docs to use correct syntax
v8 - Fix issue with tap_tx_queue_setup() not return zero on success. v7 - Reword the comment in common_base and fix the data->name issue v6 - fixed the checkpatch issues
v5 - merge in changes from list review see related emails
fixed many minor edits
v4 - merge with latest driver changes
v3 - fix includes by removing ifdef for other type besides Linux
Fix the copyright notice in the Makefile
v2 - merge all of the patches into one patch
Fix a typo on naming the tap device
Update the maintainers list

Quoted 2 lines

---

Just a reminder, this is a new PMD and waiting for community review.

#28 lore raw

I have verified that adding just a single tap device works with testpmd. But as soon as I try more than one tap device, I would get a coredump, e.g.:

root@
​localhost

:~# testpmd -c f -n 4 --socket-mem 512 --vdev=net_tap
​,iface=tap0​
--vdev=net_tap
​,iface=tap1​
-- -i
EAL: Detected 16 lcore(s)
EAL: Probing VFIO support...
EAL: VFIO support initialized
EAL: cannot open /proc/self/numa_maps, consider that all memory is in
socket_id 0
PMD: Initializing pmd_tap for net_tap as dtap0
PMD: net_tap: Create TAP Ethernet device with 32 queues on numa 0
PMD: Initializing pmd_tap for net_tap as dtap1
PMD: net_tap: Create TAP Ethernet device with 32 queues on numa 0
EAL: failed to initialize net_tap device
PANIC in rte_eal_init():
Cannot init pmd devices
6: [testpmd() [0x409149]]
5: [/lib/libc.so.6(__libc_start_main+0xf0) [0x7f3e65fa8740]]
4: [testpmd() [0x408b21]]
3: [/usr/lib/librte_eal.so.3.1(rte_eal_init+0xe09) [0x7f3e68ceaea9]]
2: [/usr/lib/librte_eal.so.3.1(__rte_panic+0xc0) [0x7f3e68ce9b5a]]
1: [/usr/lib/librte_eal.so.3.1(rte_dump_stack+0x18) [0x7f3e68cf2078]]
Aborted (core dumped)

root@
​localhost​
:~#

Quoted 46 lines · Aws Ismail

Keith,

This won't build when integrated with v16.11. The register macro prefix has been renamed. a v10 is needed.

drivers/net/tap/rte_eth_tap.c +3 -2
index 7f303db..297d4b6 100644
@@ -752,5 +752,6 @@ static struct rte_vdev_driver pmd_tap_drv = {
        .remove = rte_pmd_tap_remove,
 };
 
-DRIVER_REGISTER_VDEV(net_tap, pmd_tap_drv);
-DRIVER_REGISTER_PARAM_STRING(net_tap, "iface=<string>,speed=N");
+RTE_PMD_REGISTER_VDEV(net_tap, pmd_tap_drv);
+RTE_PMD_REGISTER_ALIAS(net_tap, eth_tap);
+RTE_PMD_REGISTER_PARAM_STRING(net_tap, "iface=<string>,speed=N");
Quoted 24 lines · Ferruh Yigit
Quoted 19 lines · Keith Wiles

The rte_eth_tap.c PMD creates a device using TUN/TAP interfaces on the local host. The PMD allows for DPDK and the host to communicate using a raw device interface on the host and in the DPDK application. The device created is a Tap device with a L2 packet header.

v9 - Fix up the docs to use correct syntax
v8 - Fix issue with tap_tx_queue_setup() not return zero on success. v7 - Reword the comment in common_base and fix the data->name issue v6 - fixed the checkpatch issues
v5 - merge in changes from list review see related emails
fixed many minor edits
v4 - merge with latest driver changes
v3 - fix includes by removing ifdef for other type besides Linux
Fix the copyright notice in the Makefile
v2 - merge all of the patches into one patch
Fix a typo on naming the tap device
Update the maintainers list

Quoted 2 lines

---

Just a reminder, this is a new PMD and waiting for community review.

#29 lore raw
Quoted 15 lines · Aws Ismail

I have verified that adding just a single tap device works with testpmd. But as soon as I try more than one tap device, I would get a coredump, e.g.:

root@​localhost:~# testpmd -c f -n 4 --socket-mem 512 --vdev=net_tap​,iface=tap0​ --vdev=net_tap​,iface=tap1​ -- -i
EAL: Detected 16 lcore(s)
EAL: Probing VFIO support...
EAL: VFIO support initialized
EAL: cannot open /proc/self/numa_maps, consider that all memory is in socket_id 0
PMD: Initializing pmd_tap for net_tap as dtap0
PMD: net_tap: Create TAP Ethernet device with 32 queues on numa 0
PMD: Initializing pmd_tap for net_tap as dtap1
PMD: net_tap: Create TAP Ethernet device with 32 queues on numa 0
EAL: failed to initialize net_tap device
Quoted 8 lines
PANIC in rte_eal_init():
Cannot init pmd devices
6: [testpmd() [0x409149]]
5: [/lib/libc.so.6(__libc_start_main+0xf0) [0x7f3e65fa8740]]
4: [testpmd() [0x408b21]]
3: [/usr/lib/librte_eal.so.3.1(rte_eal_init+0xe09) [0x7f3e68ceaea9]]
2: [/usr/lib/librte_eal.so.3.1(__rte_panic+0xc0) [0x7f3e68ce9b5a]]
1: [/usr/lib/librte_eal.so.3.1(rte_dump_stack+0x18) [0x7f3e68cf2078]]

It appears the call to rte_eth_dev_allocate() in the code is wrong. I did pass in the variable called tap_name created in the function, but I was told I needed to pass in ‘name’ that was passed into the function. The rte_eth_dev_allocate() needs a unique name for each call and name is all was the same.

Need to change that line to use tap_name instead or tell me the real way to handle this problem.

If you want a new patch I can try to get it done, but I am working on something else at this time and it could be a few days before I can get the patch out.

Quoted 46 lines

Aborted (core dumped)

root@​localhost​:~#

On Fri, Nov 25, 2016 at 2:38 PM, Aws Ismail <[email protected]> wrote:
Keith,

This won't build when integrated with v16.11. The register macro prefix has been renamed. a v10 is needed.

drivers/net/tap/rte_eth_tap.c +3 -2
index 7f303db..297d4b6 100644
@@ -752,5 +752,6 @@ static struct rte_vdev_driver pmd_tap_drv = {
        .remove = rte_pmd_tap_remove,
 };
 
-DRIVER_REGISTER_VDEV(net_tap, pmd_tap_drv);
-DRIVER_REGISTER_PARAM_STRING(net_tap, "iface=<string>,speed=N");
+RTE_PMD_REGISTER_VDEV(net_tap, pmd_tap_drv);
+RTE_PMD_REGISTER_ALIAS(net_tap, eth_tap);
+RTE_PMD_REGISTER_PARAM_STRING(net_tap, "iface=<string>,speed=N");
Quoted 20 lines · Ferruh Yigit
Quoted 19 lines · Keith Wiles

The rte_eth_tap.c PMD creates a device using TUN/TAP interfaces on the local host. The PMD allows for DPDK and the host to communicate using a raw device interface on the host and in the DPDK application. The device created is a Tap device with a L2 packet header.

v9 - Fix up the docs to use correct syntax
v8 - Fix issue with tap_tx_queue_setup() not return zero on success. v7 - Reword the comment in common_base and fix the data->name issue v6 - fixed the checkpatch issues
v5 - merge in changes from list review see related emails
fixed many minor edits
v4 - merge with latest driver changes
v3 - fix includes by removing ifdef for other type besides Linux
Fix the copyright notice in the Makefile
v2 - merge all of the patches into one patch
Fix a typo on naming the tap device
Update the maintainers list

Quoted 5 lines
Quoted 4 lines
Quoted 2 lines

---

Just a reminder, this is a new PMD and waiting for community review.

Regards,
Keith

#30 lore raw

PATCH v10 1/1 drivers/net:new TUN/TAP device PMD

In reply to Keith Wiles

The PMD allows for DPDK and the host to communicate using a raw device interface on the host and in the DPDK application. The device created is a Tap device with a L2 packet header.

v10- Change the string name used to allow for multiple devices. v9 - Fix up the docs to use correct syntax
v8 - Fix issue with tap_tx_queue_setup() not return zero on success. v7 - Reword the comment in common_base and fix the data->name issue v6 - fixed the checkpatch issues
v5 - merge in changes from list review see related emails
fixed many minor edits
v4 - merge with latest driver changes
v3 - fix includes by removing ifdef for other type besides Linux
Fix the copyright notice in the Makefile
v2 - merge all of the patches into one patch
Fix a typo on naming the tap device
Update the maintainers list

---
 MAINTAINERS                             |   5 +
 config/common_base                      |   9 +
 config/common_linuxapp                  |   1 +
 doc/guides/nics/tap.rst                 | 136 ++++++
 drivers/net/Makefile                    |   1 +
 drivers/net/tap/Makefile                |  57 +++
 drivers/net/tap/rte_eth_tap.c           | 765 ++++++++++++++++++++++++++++++++
 drivers/net/tap/rte_pmd_tap_version.map |   4 +
 mk/rte.app.mk                           |   1 +
 9 files changed, 979 insertions(+)
 create mode 100644 doc/guides/nics/tap.rst
 create mode 100644 drivers/net/tap/Makefile
 create mode 100644 drivers/net/tap/rte_eth_tap.c
 create mode 100644 drivers/net/tap/rte_pmd_tap_version.map
MAINTAINERS +5 -0
index 26d9590..842fb6d 100644
@@ -398,6 +398,11 @@ F: doc/guides/nics/pcap_ring.rst
 F: app/test/test_pmd_ring.c
 F: app/test/test_pmd_ring_perf.c
 
+Tap PMD
+M: Keith Wiles <[email protected]>
+F: drivers/net/tap
+F: doc/guides/nics/tap.rst
+
 Null Networking PMD
 M: Tetsuya Mukawa <[email protected]>
 F: drivers/net/null/
config/common_base +9 -0
index 652a839..eb51cdb 100644
@@ -590,3 +590,12 @@ CONFIG_RTE_APP_TEST_RESOURCE_TAR=n
 CONFIG_RTE_TEST_PMD=y
 CONFIG_RTE_TEST_PMD_RECORD_CORE_CYCLES=n
 CONFIG_RTE_TEST_PMD_RECORD_BURST_STATS=n
+
+#
+# Compile the TAP PMD
+#
+# The TAP PMD is currently only built for Linux and the
+# option is enabled by default in common_linuxapp file,
+# set to 'n' in the common_base file.
+#
+CONFIG_RTE_LIBRTE_PMD_TAP=n
config/common_linuxapp +1 -0
index 2483dfa..782b503 100644
@@ -44,3 +44,4 @@ CONFIG_RTE_LIBRTE_PMD_VHOST=y
 CONFIG_RTE_LIBRTE_PMD_AF_PACKET=y
 CONFIG_RTE_LIBRTE_POWER=y
 CONFIG_RTE_VIRTIO_USER=y
+CONFIG_RTE_LIBRTE_PMD_TAP=y
doc/guides/nics/tap.rst new file +136 -0

136 lines not on this page · raw message

drivers/net/Makefile +1 -0
index bc93230..e366a85 100644
@@ -51,6 +51,7 @@ DIRS-$(CONFIG_RTE_LIBRTE_PMD_PCAP) += pcap
 DIRS-$(CONFIG_RTE_LIBRTE_QEDE_PMD) += qede
 DIRS-$(CONFIG_RTE_LIBRTE_PMD_RING) += ring
 DIRS-$(CONFIG_RTE_LIBRTE_PMD_SZEDATA2) += szedata2
+DIRS-$(CONFIG_RTE_LIBRTE_PMD_TAP) += tap
 DIRS-$(CONFIG_RTE_LIBRTE_THUNDERX_NICVF_PMD) += thunderx
 DIRS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD) += virtio
 DIRS-$(CONFIG_RTE_LIBRTE_VMXNET3_PMD) += vmxnet3
drivers/net/tap/Makefile new file +57 -0

57 lines not on this page · raw message

drivers/net/tap/rte_eth_tap.c new file +765 -0

765 lines not on this page · raw message

drivers/net/tap/rte_pmd_tap_version.map new file +4 -0
new file mode 100644
index 0000000..61463bf
@@ -0,0 +1,4 @@
+DPDK_16.11 {
+
+	local: *;
+};
mk/rte.app.mk +1 -0
index f75f0e2..02c32ae 100644
@@ -124,6 +124,7 @@ _LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_PCAP)       += -lrte_pmd_pcap -lpcap
 _LDLIBS-$(CONFIG_RTE_LIBRTE_QEDE_PMD)       += -lrte_pmd_qede
 _LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_RING)       += -lrte_pmd_ring
 _LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_SZEDATA2)   += -lrte_pmd_szedata2 -lsze2
+_LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_TAP)        += -lrte_pmd_tap
 _LDLIBS-$(CONFIG_RTE_LIBRTE_THUNDERX_NICVF_PMD) += -lrte_pmd_thunderx_nicvf -lm
 _LDLIBS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD)     += -lrte_pmd_virtio
 ifeq ($(CONFIG_RTE_LIBRTE_VHOST),y)
Signature · 1 line
2.8.0.GIT
#32 lore raw

Re: [PATCH v10] drivers/net:new TUN/TAP device PMD

In reply to Keith Wiles
Quoted 18 lines · Keith Wiles

The PMD allows for DPDK and the host to communicate using a raw device interface on the host and in the DPDK application. The device created is a Tap device with a L2 packet header.

v10- Change the string name used to allow for multiple devices. v9 - Fix up the docs to use correct syntax
v8 - Fix issue with tap_tx_queue_setup() not return zero on success. v7 - Reword the comment in common_base and fix the data->name issue v6 - fixed the checkpatch issues
v5 - merge in changes from list review see related emails
fixed many minor edits
v4 - merge with latest driver changes
v3 - fix includes by removing ifdef for other type besides Linux
Fix the copyright notice in the Makefile
v2 - merge all of the patches into one patch
Fix a typo on naming the tap device
Update the maintainers list

Quoted 6 lines
---
 MAINTAINERS                             |   5 +
 config/common_base                      |   9 +
 config/common_linuxapp                  |   1 +
 doc/guides/nics/tap.rst                 | 136 ++++++

This is giving following warning [1], this file needs to be added into doc/guides/nics/index.rst

[1]
doc/guides/nics/tap.rst:: WARNING: document isn't included in any toctree

Quoted 11 lines
 drivers/net/Makefile                    |   1 +
 drivers/net/tap/Makefile                |  57 +++
 drivers/net/tap/rte_eth_tap.c           | 765 ++++++++++++++++++++++++++++++++
 drivers/net/tap/rte_pmd_tap_version.map |   4 +
 mk/rte.app.mk                           |   1 +
 9 files changed, 979 insertions(+)
 create mode 100644 doc/guides/nics/tap.rst
 create mode 100644 drivers/net/tap/Makefile
 create mode 100644 drivers/net/tap/rte_eth_tap.c
 create mode 100644 drivers/net/tap/rte_pmd_tap_version.map

<...>

drivers/net/tap/rte_pmd_tap_version.map line 1

Quoted 7 lines
drivers/net/tap/rte_pmd_tap_version.map new file +1 -0
new file mode 100644
index 0000000..61463bf
@@ -0,0 +1,4 @@
+DPDK_16.11 {

17.02

Quoted 3 lines
drivers/net/tap/rte_pmd_tap_version.map new file +3 -0
+
+	local: *;
+};

<...>

#33 lore raw

PATCH v11 1/1 drivers/net:new TUN/TAP device PMD

In reply to Keith Wiles

The PMD allows for DPDK and the host to communicate using a raw device interface on the host and in the DPDK application. The device created is a Tap device with a L2 packet header.

v11- Add the tap.rst to the nic/index.rst file
v10- Change the string name used to allow for multiple devices. v9 - Fix up the docs to use correct syntax
v8 - Fix issue with tap_tx_queue_setup() not return zero on success. v7 - Reword the comment in common_base and fix the data->name issue v6 - fixed the checkpatch issues
v5 - merge in changes from list review see related emails
fixed many minor edits
v4 - merge with latest driver changes
v3 - fix includes by removing ifdef for other type besides Linux
Fix the copyright notice in the Makefile
v2 - merge all of the patches into one patch
Fix a typo on naming the tap device
Update the maintainers list

---
 MAINTAINERS                             |   5 +
 config/common_base                      |   9 +
 config/common_linuxapp                  |   1 +
 doc/guides/nics/index.rst               |   1 +
 doc/guides/nics/tap.rst                 | 136 ++++++
 drivers/net/Makefile                    |   1 +
 drivers/net/tap/Makefile                |  57 +++
 drivers/net/tap/rte_eth_tap.c           | 765 ++++++++++++++++++++++++++++++++
 drivers/net/tap/rte_pmd_tap_version.map |   4 +
 mk/rte.app.mk                           |   1 +
 10 files changed, 980 insertions(+)
 create mode 100644 doc/guides/nics/tap.rst
 create mode 100644 drivers/net/tap/Makefile
 create mode 100644 drivers/net/tap/rte_eth_tap.c
 create mode 100644 drivers/net/tap/rte_pmd_tap_version.map
MAINTAINERS +5 -0
index 26d9590..842fb6d 100644
@@ -398,6 +398,11 @@ F: doc/guides/nics/pcap_ring.rst
 F: app/test/test_pmd_ring.c
 F: app/test/test_pmd_ring_perf.c
 
+Tap PMD
+M: Keith Wiles <[email protected]>
+F: drivers/net/tap
+F: doc/guides/nics/tap.rst
+
 Null Networking PMD
 M: Tetsuya Mukawa <[email protected]>
 F: drivers/net/null/
config/common_base +9 -0
index 652a839..eb51cdb 100644
@@ -590,3 +590,12 @@ CONFIG_RTE_APP_TEST_RESOURCE_TAR=n
 CONFIG_RTE_TEST_PMD=y
 CONFIG_RTE_TEST_PMD_RECORD_CORE_CYCLES=n
 CONFIG_RTE_TEST_PMD_RECORD_BURST_STATS=n
+
+#
+# Compile the TAP PMD
+#
+# The TAP PMD is currently only built for Linux and the
+# option is enabled by default in common_linuxapp file,
+# set to 'n' in the common_base file.
+#
+CONFIG_RTE_LIBRTE_PMD_TAP=n
config/common_linuxapp +1 -0
index 2483dfa..782b503 100644
@@ -44,3 +44,4 @@ CONFIG_RTE_LIBRTE_PMD_VHOST=y
 CONFIG_RTE_LIBRTE_PMD_AF_PACKET=y
 CONFIG_RTE_LIBRTE_POWER=y
 CONFIG_RTE_VIRTIO_USER=y
+CONFIG_RTE_LIBRTE_PMD_TAP=y
doc/guides/nics/index.rst +1 -0
index 92d56a5..af92529 100644
@@ -51,6 +51,7 @@ Network Interface Controller Drivers
     nfp
     qede
     szedata2
+    tap
     thunderx
     virtio
     vhost
doc/guides/nics/tap.rst new file +136 -0

136 lines not on this page · raw message

drivers/net/Makefile +1 -0
index bc93230..e366a85 100644
@@ -51,6 +51,7 @@ DIRS-$(CONFIG_RTE_LIBRTE_PMD_PCAP) += pcap
 DIRS-$(CONFIG_RTE_LIBRTE_QEDE_PMD) += qede
 DIRS-$(CONFIG_RTE_LIBRTE_PMD_RING) += ring
 DIRS-$(CONFIG_RTE_LIBRTE_PMD_SZEDATA2) += szedata2
+DIRS-$(CONFIG_RTE_LIBRTE_PMD_TAP) += tap
 DIRS-$(CONFIG_RTE_LIBRTE_THUNDERX_NICVF_PMD) += thunderx
 DIRS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD) += virtio
 DIRS-$(CONFIG_RTE_LIBRTE_VMXNET3_PMD) += vmxnet3
drivers/net/tap/Makefile new file +57 -0

57 lines not on this page · raw message

drivers/net/tap/rte_eth_tap.c new file +765 -0

765 lines not on this page · raw message

drivers/net/tap/rte_pmd_tap_version.map new file +4 -0
new file mode 100644
index 0000000..61463bf
@@ -0,0 +1,4 @@
+DPDK_16.11 {
+
+	local: *;
+};
mk/rte.app.mk +1 -0
index f75f0e2..02c32ae 100644
@@ -124,6 +124,7 @@ _LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_PCAP)       += -lrte_pmd_pcap -lpcap
 _LDLIBS-$(CONFIG_RTE_LIBRTE_QEDE_PMD)       += -lrte_pmd_qede
 _LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_RING)       += -lrte_pmd_ring
 _LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_SZEDATA2)   += -lrte_pmd_szedata2 -lsze2
+_LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_TAP)        += -lrte_pmd_tap
 _LDLIBS-$(CONFIG_RTE_LIBRTE_THUNDERX_NICVF_PMD) += -lrte_pmd_thunderx_nicvf -lm
 _LDLIBS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD)     += -lrte_pmd_virtio
 ifeq ($(CONFIG_RTE_LIBRTE_VHOST),y)
Signature · 1 line
2.8.0.GIT
Tested-by on v11 1/1
#34 lore raw

Re: [PATCH v11] drivers/net:new TUN/TAP device PMD

Quoted 30 lines

-----Original Message-----
From: dev [mailto:[email protected]] On Behalf Of Keith Wiles
Sent: Friday, December 09, 2016 21:05
To: [email protected]
Cc: [email protected]; [email protected]; [email protected]; [email protected]
Subject: [dpdk-dev] [PATCH v11] drivers/net:new TUN/TAP device PMD

The PMD allows for DPDK and the host to communicate using a raw device interface on the host and in the DPDK application. The device created is a Tap device with a L2 packet header.

v11- Add the tap.rst to the nic/index.rst file
v10- Change the string name used to allow for multiple devices. v9 - Fix up the docs to use correct syntax
v8 - Fix issue with tap_tx_queue_setup() not return zero on success. v7 - Reword the comment in common_base and fix the data->name issue v6 - fixed the checkpatch issues
v5 - merge in changes from list review see related emails
fixed many minor edits
v4 - merge with latest driver changes
v3 - fix includes by removing ifdef for other type besides Linux
Fix the copyright notice in the Makefile
v2 - merge all of the patches into one patch
Fix a typo on naming the tap device
Update the maintainers list

Quoted 13 lines
---
 MAINTAINERS                             |   5 +
 config/common_base                      |   9 +
 config/common_linuxapp                  |   1 +
 doc/guides/nics/index.rst               |   1 +
 doc/guides/nics/tap.rst                 | 136 ++++++
 drivers/net/Makefile                    |   1 +
 drivers/net/tap/Makefile                |  57 +++
 drivers/net/tap/rte_eth_tap.c           | 765
++++++++++++++++++++++++++++++++
 drivers/net/tap/rte_pmd_tap_version.map |   4 +
 mk/rte.app.mk                           |   1 +
 10 files changed, 980 insertions(+)
#35 lore raw

PATCH v12 1/1 net/tap: new TUN/TAP device PMD

In reply to Keith Wiles

The PMD allows for DPDK and the host to communicate using a raw device interface on the host and in the DPDK application. The device created is a Tap device with a L2 packet header.

v12- Fixup minor changes for driver_name and version number v11- Add the tap.rst to the nic/index.rst file
v10- Change the string name used to allow for multiple devices. v9 - Fix up the docs to use correct syntax
v8 - Fix issue with tap_tx_queue_setup() not return zero on success. v7 - Reword the comment in common_base and fix the data->name issue v6 - fixed the checkpatch issues
v5 - merge in changes from list review see related emails
fixed many minor edits
v4 - merge with latest driver changes
v3 - fix includes by removing ifdef for other type besides Linux
Fix the copyright notice in the Makefile
v2 - merge all of the patches into one patch
Fix a typo on naming the tap device
Update the maintainers list

---
 MAINTAINERS                             |   5 +
 config/common_base                      |   9 +
 config/common_linuxapp                  |   1 +
 doc/guides/nics/index.rst               |   1 +
 doc/guides/nics/tap.rst                 | 136 ++++++
 drivers/net/Makefile                    |   1 +
 drivers/net/tap/Makefile                |  57 +++
 drivers/net/tap/rte_eth_tap.c           | 765 ++++++++++++++++++++++++++++++++
 drivers/net/tap/rte_pmd_tap_version.map |   4 +
 mk/rte.app.mk                           |   1 +
 10 files changed, 980 insertions(+)
 create mode 100644 doc/guides/nics/tap.rst
 create mode 100644 drivers/net/tap/Makefile
 create mode 100644 drivers/net/tap/rte_eth_tap.c
 create mode 100644 drivers/net/tap/rte_pmd_tap_version.map
MAINTAINERS +5 -0
index 26d9590..842fb6d 100644
@@ -398,6 +398,11 @@ F: doc/guides/nics/pcap_ring.rst
 F: app/test/test_pmd_ring.c
 F: app/test/test_pmd_ring_perf.c
 
+Tap PMD
+M: Keith Wiles <[email protected]>
+F: drivers/net/tap
+F: doc/guides/nics/tap.rst
+
 Null Networking PMD
 M: Tetsuya Mukawa <[email protected]>
 F: drivers/net/null/
config/common_base +9 -0
index 652a839..eb51cdb 100644
@@ -590,3 +590,12 @@ CONFIG_RTE_APP_TEST_RESOURCE_TAR=n
 CONFIG_RTE_TEST_PMD=y
 CONFIG_RTE_TEST_PMD_RECORD_CORE_CYCLES=n
 CONFIG_RTE_TEST_PMD_RECORD_BURST_STATS=n
+
+#
+# Compile the TAP PMD
+#
+# The TAP PMD is currently only built for Linux and the
+# option is enabled by default in common_linuxapp file,
+# set to 'n' in the common_base file.
+#
+CONFIG_RTE_LIBRTE_PMD_TAP=n
config/common_linuxapp +1 -0
index 2483dfa..782b503 100644
@@ -44,3 +44,4 @@ CONFIG_RTE_LIBRTE_PMD_VHOST=y
 CONFIG_RTE_LIBRTE_PMD_AF_PACKET=y
 CONFIG_RTE_LIBRTE_POWER=y
 CONFIG_RTE_VIRTIO_USER=y
+CONFIG_RTE_LIBRTE_PMD_TAP=y
doc/guides/nics/index.rst +1 -0
index 92d56a5..af92529 100644
@@ -51,6 +51,7 @@ Network Interface Controller Drivers
     nfp
     qede
     szedata2
+    tap
     thunderx
     virtio
     vhost
doc/guides/nics/tap.rst new file +136 -0

136 lines not on this page · raw message

drivers/net/Makefile +1 -0
index bc93230..e366a85 100644
@@ -51,6 +51,7 @@ DIRS-$(CONFIG_RTE_LIBRTE_PMD_PCAP) += pcap
 DIRS-$(CONFIG_RTE_LIBRTE_QEDE_PMD) += qede
 DIRS-$(CONFIG_RTE_LIBRTE_PMD_RING) += ring
 DIRS-$(CONFIG_RTE_LIBRTE_PMD_SZEDATA2) += szedata2
+DIRS-$(CONFIG_RTE_LIBRTE_PMD_TAP) += tap
 DIRS-$(CONFIG_RTE_LIBRTE_THUNDERX_NICVF_PMD) += thunderx
 DIRS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD) += virtio
 DIRS-$(CONFIG_RTE_LIBRTE_VMXNET3_PMD) += vmxnet3
drivers/net/tap/Makefile new file +57 -0

57 lines not on this page · raw message

drivers/net/tap/rte_eth_tap.c new file +765 -0

765 lines not on this page · raw message

drivers/net/tap/rte_pmd_tap_version.map new file +4 -0
new file mode 100644
index 0000000..61463bf
@@ -0,0 +1,4 @@
+DPDK_16.11 {
+
+	local: *;
+};
mk/rte.app.mk +1 -0
index f75f0e2..02c32ae 100644
@@ -124,6 +124,7 @@ _LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_PCAP)       += -lrte_pmd_pcap -lpcap
 _LDLIBS-$(CONFIG_RTE_LIBRTE_QEDE_PMD)       += -lrte_pmd_qede
 _LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_RING)       += -lrte_pmd_ring
 _LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_SZEDATA2)   += -lrte_pmd_szedata2 -lsze2
+_LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_TAP)        += -lrte_pmd_tap
 _LDLIBS-$(CONFIG_RTE_LIBRTE_THUNDERX_NICVF_PMD) += -lrte_pmd_thunderx_nicvf -lm
 _LDLIBS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD)     += -lrte_pmd_virtio
 ifeq ($(CONFIG_RTE_LIBRTE_VHOST),y)
Signature · 1 line
2.8.0.GIT
#36 lore raw

PATCH v12 1/1 net/tap: new TUN/TAP device PMD

In reply to Keith Wiles

The PMD allows for DPDK and the host to communicate using a raw device interface on the host and in the DPDK application. The device created is a Tap device with a L2 packet header.

v12- Fixup minor changes for driver_name and version number v11- Add the tap.rst to the nic/index.rst file
v10- Change the string name used to allow for multiple devices. v9 - Fix up the docs to use correct syntax
v8 - Fix issue with tap_tx_queue_setup() not return zero on success. v7 - Reword the comment in common_base and fix the data->name issue v6 - fixed the checkpatch issues
v5 - merge in changes from list review see related emails
fixed many minor edits
v4 - merge with latest driver changes
v3 - fix includes by removing ifdef for other type besides Linux
Fix the copyright notice in the Makefile
v2 - merge all of the patches into one patch
Fix a typo on naming the tap device
Update the maintainers list

---
 MAINTAINERS                             |   5 +
 config/common_base                      |   9 +
 config/common_linuxapp                  |   1 +
 doc/guides/nics/index.rst               |   1 +
 doc/guides/nics/tap.rst                 | 136 ++++++
 drivers/net/Makefile                    |   1 +
 drivers/net/tap/Makefile                |  57 +++
 drivers/net/tap/rte_eth_tap.c           | 765 ++++++++++++++++++++++++++++++++
 drivers/net/tap/rte_pmd_tap_version.map |   4 +
 mk/rte.app.mk                           |   1 +
 10 files changed, 980 insertions(+)
 create mode 100644 doc/guides/nics/tap.rst
 create mode 100644 drivers/net/tap/Makefile
 create mode 100644 drivers/net/tap/rte_eth_tap.c
 create mode 100644 drivers/net/tap/rte_pmd_tap_version.map
MAINTAINERS +5 -0
index 26d9590..842fb6d 100644
@@ -398,6 +398,11 @@ F: doc/guides/nics/pcap_ring.rst
 F: app/test/test_pmd_ring.c
 F: app/test/test_pmd_ring_perf.c
 
+Tap PMD
+M: Keith Wiles <[email protected]>
+F: drivers/net/tap
+F: doc/guides/nics/tap.rst
+
 Null Networking PMD
 M: Tetsuya Mukawa <[email protected]>
 F: drivers/net/null/
config/common_base +9 -0
index 652a839..eb51cdb 100644
@@ -590,3 +590,12 @@ CONFIG_RTE_APP_TEST_RESOURCE_TAR=n
 CONFIG_RTE_TEST_PMD=y
 CONFIG_RTE_TEST_PMD_RECORD_CORE_CYCLES=n
 CONFIG_RTE_TEST_PMD_RECORD_BURST_STATS=n
+
+#
+# Compile the TAP PMD
+#
+# The TAP PMD is currently only built for Linux and the
+# option is enabled by default in common_linuxapp file,
+# set to 'n' in the common_base file.
+#
+CONFIG_RTE_LIBRTE_PMD_TAP=n
config/common_linuxapp +1 -0
index 2483dfa..782b503 100644
@@ -44,3 +44,4 @@ CONFIG_RTE_LIBRTE_PMD_VHOST=y
 CONFIG_RTE_LIBRTE_PMD_AF_PACKET=y
 CONFIG_RTE_LIBRTE_POWER=y
 CONFIG_RTE_VIRTIO_USER=y
+CONFIG_RTE_LIBRTE_PMD_TAP=y
doc/guides/nics/index.rst +1 -0
index 92d56a5..af92529 100644
@@ -51,6 +51,7 @@ Network Interface Controller Drivers
     nfp
     qede
     szedata2
+    tap
     thunderx
     virtio
     vhost
doc/guides/nics/tap.rst new file +136 -0

136 lines not on this page · raw message

drivers/net/Makefile +1 -0
index bc93230..e366a85 100644
@@ -51,6 +51,7 @@ DIRS-$(CONFIG_RTE_LIBRTE_PMD_PCAP) += pcap
 DIRS-$(CONFIG_RTE_LIBRTE_QEDE_PMD) += qede
 DIRS-$(CONFIG_RTE_LIBRTE_PMD_RING) += ring
 DIRS-$(CONFIG_RTE_LIBRTE_PMD_SZEDATA2) += szedata2
+DIRS-$(CONFIG_RTE_LIBRTE_PMD_TAP) += tap
 DIRS-$(CONFIG_RTE_LIBRTE_THUNDERX_NICVF_PMD) += thunderx
 DIRS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD) += virtio
 DIRS-$(CONFIG_RTE_LIBRTE_VMXNET3_PMD) += vmxnet3
drivers/net/tap/Makefile new file +57 -0

57 lines not on this page · raw message

drivers/net/tap/rte_eth_tap.c new file +765 -0

765 lines not on this page · raw message

drivers/net/tap/rte_pmd_tap_version.map new file +4 -0
new file mode 100644
index 0000000..31eca32
@@ -0,0 +1,4 @@
+DPDK_17.02 {
+
+	local: *;
+};
mk/rte.app.mk +1 -0
index f75f0e2..02c32ae 100644
@@ -124,6 +124,7 @@ _LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_PCAP)       += -lrte_pmd_pcap -lpcap
 _LDLIBS-$(CONFIG_RTE_LIBRTE_QEDE_PMD)       += -lrte_pmd_qede
 _LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_RING)       += -lrte_pmd_ring
 _LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_SZEDATA2)   += -lrte_pmd_szedata2 -lsze2
+_LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_TAP)        += -lrte_pmd_tap
 _LDLIBS-$(CONFIG_RTE_LIBRTE_THUNDERX_NICVF_PMD) += -lrte_pmd_thunderx_nicvf -lm
 _LDLIBS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD)     += -lrte_pmd_virtio
 ifeq ($(CONFIG_RTE_LIBRTE_VHOST),y)
Signature · 1 line
2.8.0.GIT
#37 lore raw

Re: [PATCH v12] net/tap: new TUN/TAP device PMD

Keith,

A bit late, but two very high level questions. Do you have performance numbers compared to KNI? Did you consider using AF_PACKET PACKET_MMAP which could potentially reduce the number of syscalls to 1 for RX and TX of a burst?

Marc

Quoted 1,156 lines · Keith Wiles

The PMD allows for DPDK and the host to communicate using a raw device interface on the host and in the DPDK application. The device created is a Tap device with a L2 packet header.

v12- Fixup minor changes for driver_name and version number v11- Add the tap.rst to the nic/index.rst file
v10- Change the string name used to allow for multiple devices. v9 - Fix up the docs to use correct syntax
v8 - Fix issue with tap_tx_queue_setup() not return zero on success. v7 - Reword the comment in common_base and fix the data->name issue v6 - fixed the checkpatch issues
v5 - merge in changes from list review see related emails
fixed many minor edits
v4 - merge with latest driver changes
v3 - fix includes by removing ifdef for other type besides Linux
Fix the copyright notice in the Makefile
v2 - merge all of the patches into one patch
Fix a typo on naming the tap device
Update the maintainers list

---
 MAINTAINERS                             |   5 +
 config/common_base                      |   9 +
 config/common_linuxapp                  |   1 +
 doc/guides/nics/index.rst               |   1 +
 doc/guides/nics/tap.rst                 | 136 ++++++
 drivers/net/Makefile                    |   1 +
 drivers/net/tap/Makefile                |  57 +++
 drivers/net/tap/rte_eth_tap.c           | 765
++++++++++++++++++++++++++++++++
 drivers/net/tap/rte_pmd_tap_version.map |   4 +
 mk/rte.app.mk                           |   1 +
 10 files changed, 980 insertions(+)
 create mode 100644 doc/guides/nics/tap.rst
 create mode 100644 drivers/net/tap/Makefile
 create mode 100644 drivers/net/tap/rte_eth_tap.c
 create mode 100644 drivers/net/tap/rte_pmd_tap_version.map
MAINTAINERS +5 -0
index 26d9590..842fb6d 100644
@@ -398,6 +398,11 @@ F: doc/guides/nics/pcap_ring.rst
 F: app/test/test_pmd_ring.c
 F: app/test/test_pmd_ring_perf.c
 
+Tap PMD
+M: Keith Wiles <[email protected]>
+F: drivers/net/tap
+F: doc/guides/nics/tap.rst
+
 Null Networking PMD
 M: Tetsuya Mukawa <[email protected]>
 F: drivers/net/null/
config/common_base +9 -0
index 652a839..eb51cdb 100644
@@ -590,3 +590,12 @@ CONFIG_RTE_APP_TEST_RESOURCE_TAR=n
 CONFIG_RTE_TEST_PMD=y
 CONFIG_RTE_TEST_PMD_RECORD_CORE_CYCLES=n
 CONFIG_RTE_TEST_PMD_RECORD_BURST_STATS=n
+
+#
+# Compile the TAP PMD
+#
+# The TAP PMD is currently only built for Linux and the
+# option is enabled by default in common_linuxapp file,
+# set to 'n' in the common_base file.
+#
+CONFIG_RTE_LIBRTE_PMD_TAP=n
config/common_linuxapp +1 -0
index 2483dfa..782b503 100644
@@ -44,3 +44,4 @@ CONFIG_RTE_LIBRTE_PMD_VHOST=y
 CONFIG_RTE_LIBRTE_PMD_AF_PACKET=y
 CONFIG_RTE_LIBRTE_POWER=y
 CONFIG_RTE_VIRTIO_USER=y
+CONFIG_RTE_LIBRTE_PMD_TAP=y
doc/guides/nics/index.rst +1 -0
index 92d56a5..af92529 100644
@@ -51,6 +51,7 @@ Network Interface Controller Drivers
     nfp
     qede
     szedata2
+    tap
     thunderx
     virtio
     vhost
doc/guides/nics/tap.rst new file +35 -0
new file mode 100644
index 0000000..622b9e7
@@ -0,0 +1,136 @@
+..  BSD LICENSE
+    Copyright(c) 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.
+
+Tun/Tap Poll Mode Driver
+========================
+
+The ``rte_eth_tap.c`` PMD creates a device using TUN/TAP interfaces on the
+local host. The PMD allows for DPDK and the host to communicate using a

raw

doc/guides/nics/tap.rst new file +4 -0
+device interface on the host and in the DPDK application.
+
+The device created is a TAP device, which sends/receives packet in a raw
+format with a L2 header. The usage for a TAP PMD is for connectivity to

the

doc/guides/nics/tap.rst new file +3 -0
+local host using a TAP interface. When the TAP PMD is initialized it will
+create a number of tap devices in the host accessed via ``ifconfig -a`` or
+``ip`` command. The commands can be used to assign and query the virtual

like

doc/guides/nics/tap.rst new file +7 -0
+device.
+
+These TAP interfaces can be used with Wireshark or tcpdump or Pktgen-DPDK
+along with being able to be used as a network connection to the DPDK
+application. The method enable one or more interfaces is to use the
+``--vdev=net_tap`` option on the DPDK application command line. Each
+``--vdev=net_tap`` option give will create an interface named dtap0,

dtap1,

doc/guides/nics/tap.rst new file +3 -0
+and so on.
+
+The interfaced name can be changed by adding the ``iface=foo0``, for

example::

doc/guides/nics/tap.rst new file +9 -0
+
+   --vdev=net_tap,iface=foo0 --vdev=net_tap,iface=foo1, ...
+
+Also the speed of the interface can be changed from 10G to whatever number
+needed, but the interface does not enforce that speed, for example::
+
+   --vdev=net_tap,iface=foo0,speed=25000
+
+After the DPDK application is started you can send and receive packets on

the

doc/guides/nics/tap.rst new file +2 -0
+interface using the standard rx_burst/tx_burst APIs in DPDK. From the host
+point of view you can use any host tool like tcpdump, Wireshark, ping,

Pktgen
+and others to communicate with the DPDK application. The DPDK application may

doc/guides/nics/tap.rst new file +4 -0
+not understand network protocols like IPv4/6, UDP or TCP unless the
+application has been written to understand these protocols.
+
+If you need the interface as a real network interface meaning running and

has

doc/guides/nics/tap.rst new file +8 -0
+a valid IP address then you can do this with the following commands::
+
+   sudo ip link set dtap0 up; sudo ip addr add 192.168.0.250/24 dev dtap0
+   sudo ip link set dtap1 up; sudo ip addr add 192.168.1.250/24 dev dtap1
+
+Please change the IP addresses as you see fit.
+
+If routing is enabled on the host you can also communicate with the DPDK

App

doc/guides/nics/tap.rst new file +4 -0
+over the internet via a standard socket layer application as long as you
+account for the protocol handing in the application.
+
+If you have a Network Stack in your DPDK application or something like it

you
+can utilize that stack to handle the network protocols. Plus you would be able

doc/guides/nics/tap.rst new file +8 -0
+to address the interface using an IP address assigned to the internal
+interface.
+
+Example
+-------
+
+The following is a simple example of using the TUN/TAP PMD with the Pktgen
+packet generator. It requires that the ``socat`` utility is installed on

the

doc/guides/nics/tap.rst new file +3 -0
+test system.
+
+Build DPDK, then pull down Pktgen and build pktgen using the DPDK

SDK/Target

doc/guides/nics/tap.rst new file +3 -0
+used to build the dpdk you pulled down.
+
+Run pktgen from the pktgen directory in a terminal with a commandline

like the

doc/guides/nics/tap.rst new file +4 -0
+following::
+
+    sudo ./app/app/x86_64-native-linuxapp-gcc/app/pktgen -l 1-5 -n 4
    \
+     --proc-type auto --log-level 8 --socket-mem 512,512 --file-prefix

pg \

doc/guides/nics/tap.rst new file +12 -0
+     --vdev=net_tap --vdev=net_tap -b 05:00.0 -b 05:00.1
  \
+     -b 04:00.0 -b 04:00.1 -b 04:00.2 -b 04:00.3
  \
+     -b 81:00.0 -b 81:00.1 -b 81:00.2 -b 81:00.3
  \
+     -b 82:00.0 -b 83:00.0 -- -T -P -m [2:3].0 -m [4:5].1
   \
+     -f themes/black-yellow.theme
+
+.. Note:
+
+   Change the ``-b`` options to blacklist all of your physical ports. The
+   following command line is all one line.
+
+   Also, ``-f themes/black-yellow.theme`` is optional if the default

colors

doc/guides/nics/tap.rst new file +4 -0
+   work on your system configuration. See the Pktgen docs for more
+   information.
+
+Verify with ``ifconfig -a`` command in a different xterm window, should

have a
+``dtap0`` and ``dtap1`` interfaces created.

doc/guides/nics/tap.rst new file +4 -0
+
+Next set the links for the two interfaces to up via the commands below::
+
+    sudo ip link set dtap0 up; sudo ip addr add 192.168.0.250/24 dev

dtap0
+ sudo ip link set dtap1 up; sudo ip addr add 192.168.1.250/24 dev
dtap1

doc/guides/nics/tap.rst new file +6 -0
+
+Then use socat to create a loopback for the two interfaces::
+
+    sudo socat interface:dtap0 interface:dtap1
+
+Then on the Pktgen command line interface you can start sending packets

using

doc/guides/nics/tap.rst new file +2 -0
+the commands ``start 0`` and ``start 1`` or you can start both at the same
+time with ``start all``. The command ``str`` is an alias for ``start

all`` and
+``stp`` is an alias for ``stop all``.

doc/guides/nics/tap.rst new file +3 -0
+
+While running you should see the 64 byte counters increasing to verify the
+traffic is being looped back. You can use ``set all size XXX`` to change

the

doc/guides/nics/tap.rst new file +2 -0
+size of the packets after you stop the traffic. Use the pktgen ``help``
+command to see a list of all commands. You can also use the ``-f`` option

to

doc/guides/nics/tap.rst new file +1 -0
+load commands at startup.
drivers/net/Makefile +1 -0
index bc93230..e366a85 100644
@@ -51,6 +51,7 @@ DIRS-$(CONFIG_RTE_LIBRTE_PMD_PCAP) += pcap
 DIRS-$(CONFIG_RTE_LIBRTE_QEDE_PMD) += qede
 DIRS-$(CONFIG_RTE_LIBRTE_PMD_RING) += ring
 DIRS-$(CONFIG_RTE_LIBRTE_PMD_SZEDATA2) += szedata2
+DIRS-$(CONFIG_RTE_LIBRTE_PMD_TAP) += tap
 DIRS-$(CONFIG_RTE_LIBRTE_THUNDERX_NICVF_PMD) += thunderx
 DIRS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD) += virtio
 DIRS-$(CONFIG_RTE_LIBRTE_VMXNET3_PMD) += vmxnet3
drivers/net/tap/Makefile new file +57 -0
new file mode 100644
index 0000000..e18f30c
@@ -0,0 +1,57 @@
+#   BSD LICENSE
+#
+#   Copyright(c) 2016 Intel Corporation. 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_pmd_tap.a
+
+EXPORT_MAP := rte_pmd_tap_version.map
+
+LIBABIVER := 1
+
+CFLAGS += -O3
+CFLAGS += $(WERROR_FLAGS)
+
+#
+# all source are stored in SRCS-y
+#
+SRCS-$(CONFIG_RTE_LIBRTE_PMD_TAP) += rte_eth_tap.c
+
+# this lib depends upon:
+DEPDIRS-$(CONFIG_RTE_LIBRTE_PMD_TAP) += lib/librte_eal
+DEPDIRS-$(CONFIG_RTE_LIBRTE_PMD_TAP) += lib/librte_mbuf
+DEPDIRS-$(CONFIG_RTE_LIBRTE_PMD_TAP) += lib/librte_mempool
+DEPDIRS-$(CONFIG_RTE_LIBRTE_PMD_TAP) += lib/librte_ether
+DEPDIRS-$(CONFIG_RTE_LIBRTE_PMD_TAP) += lib/librte_kvargs
+
+include $(RTE_SDK)/mk/rte.lib.mk
drivers/net/tap/rte_eth_tap.c new file +101 -0
new file mode 100644
index 0000000..976f2d9
@@ -0,0 +1,765 @@
+/*-
+ *   BSD LICENSE
+ *
+ *   Copyright(c) 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.
+ */
+
+#include <rte_mbuf.h>
+#include <rte_ethdev.h>
+#include <rte_malloc.h>
+#include <rte_vdev.h>
+#include <rte_kvargs.h>
+
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <sys/socket.h>
+#include <sys/ioctl.h>
+#include <sys/mman.h>
+#include <unistd.h>
+#include <poll.h>
+#include <arpa/inet.h>
+#include <linux/if.h>
+#include <linux/if_tun.h>
+#include <linux/if_ether.h>
+#include <fcntl.h>
+
+/* Linux based path to the TUN device */
+#define TUN_TAP_DEV_PATH        "/dev/net/tun"
+#define DEFAULT_TAP_NAME        "dtap"
+
+#define ETH_TAP_IFACE_ARG       "iface"
+#define ETH_TAP_SPEED_ARG       "speed"
+
+#define RTE_PMD_TAP_MAX_QUEUES 16
+
+static struct rte_vdev_driver pmd_tap_drv;
+
+static const char *valid_arguments[] = {
+       ETH_TAP_IFACE_ARG,
+       ETH_TAP_SPEED_ARG,
+       NULL
+};
+
+static int tap_unit;
+
+static struct rte_eth_link pmd_link = {
+       .link_speed = ETH_SPEED_NUM_10G,
+       .link_duplex = ETH_LINK_FULL_DUPLEX,
+       .link_status = ETH_LINK_DOWN,
+       .link_autoneg = ETH_LINK_SPEED_AUTONEG
+};
+
+struct pkt_stats {
+       uint64_t opackets;              /* Number of output packets */
+       uint64_t ipackets;              /* Number of input packets */
+       uint64_t obytes;                /* Number of bytes on output */
+       uint64_t ibytes;                /* Number of bytes on input */
+       uint64_t errs;                  /* Number of error packets */
+};
+
+struct rx_queue {
+       struct rte_mempool *mp;         /* Mempool for RX packets */
+       uint16_t in_port;               /* Port ID */
+       int fd;
+
+       struct pkt_stats stats;         /* Stats for this RX queue */
+};
+
+struct tx_queue {
+       int fd;
+       struct pkt_stats stats;         /* Stats for this TX queue */
+};
+
+struct pmd_internals {
+       char name[RTE_ETH_NAME_MAX_LEN];        /* Internal Tap device

name */

drivers/net/tap/rte_eth_tap.c new file +2 -0
+       uint16_t nb_queues;             /* Number of queues supported */
+       struct ether_addr eth_addr;     /* Mac address of the device port

*/

drivers/net/tap/rte_eth_tap.c new file +5 -0
+
+       int if_index;                   /* IF_INDEX for the port */
+       int fds[RTE_PMD_TAP_MAX_QUEUES]; /* List of all file descriptors */
+
+       struct rx_queue rxq[RTE_PMD_TAP_MAX_QUEUES];    /* List of RX

queues */
+ struct tx_queue txq[RTE_PMD_TAP_MAX_QUEUES]; /* List of TX
queues */

drivers/net/tap/rte_eth_tap.c new file +5 -0
+};
+
+/* Tun/Tap allocation routine
+ *
+ * name is the number of the interface to use, unless NULL to take the

host
+ * supplied name.
+ */

drivers/net/tap/rte_eth_tap.c new file +27 -0
+static int
+tun_alloc(char *name)
+{
+       struct ifreq ifr;
+       unsigned int features;
+       int fd;
+
+       memset(&ifr, 0, sizeof(struct ifreq));
+
+       ifr.ifr_flags = IFF_TAP | IFF_NO_PI;
+       if (name && name[0])
+               strncpy(ifr.ifr_name, name, IFNAMSIZ);
+
+       fd = open(TUN_TAP_DEV_PATH, O_RDWR);
+       if (fd < 0) {
+               RTE_LOG(ERR, PMD, "Unable to create TAP interface");
+               goto error;
+       }
+
+       /* Grab the TUN features to verify we can work */
+       if (ioctl(fd, TUNGETFEATURES, &features) < 0) {
+               RTE_LOG(ERR, PMD, "Unable to get TUN/TAP features\n");
+               goto error;
+       }
+       RTE_LOG(DEBUG, PMD, "TUN/TAP Features %08x\n", features);
+
+       if (!(features & IFF_MULTI_QUEUE) && (RTE_PMD_TAP_MAX_QUEUES > 1))

{

drivers/net/tap/rte_eth_tap.c new file +30 -0
+               RTE_LOG(DEBUG, PMD, "TUN/TAP device only one queue\n");
+               goto error;
+       } else if ((features & IFF_ONE_QUEUE) &&
+                       (RTE_PMD_TAP_MAX_QUEUES == 1)) {
+               ifr.ifr_flags |= IFF_ONE_QUEUE;
+               RTE_LOG(DEBUG, PMD, "Single queue only support\n");
+       } else {
+               ifr.ifr_flags |= IFF_MULTI_QUEUE;
+               RTE_LOG(DEBUG, PMD, "Multi-queue support for %d queues\n",
+                       RTE_PMD_TAP_MAX_QUEUES);
+       }
+
+       /* Set the TUN/TAP configuration and get the name if needed */
+       if (ioctl(fd, TUNSETIFF, (void *)&ifr) < 0) {
+               RTE_LOG(ERR, PMD, "Unable to set TUNSETIFF for %s\n",
+                       ifr.ifr_name);
+               perror("TUNSETIFF");
+               goto error;
+       }
+
+       /* Always set the file descriptor to non-blocking */
+       if (fcntl(fd, F_SETFL, O_NONBLOCK) < 0) {
+               RTE_LOG(ERR, PMD, "Unable to set to nonblocking\n");
+               perror("F_SETFL, NONBLOCK");
+               goto error;
+       }
+
+       /* If the name is different that new name as default */
+       if (name && strcmp(name, ifr.ifr_name))
+               snprintf(name, RTE_ETH_NAME_MAX_LEN - 1, "%s",

ifr.ifr_name);

drivers/net/tap/rte_eth_tap.c new file +238 -0
+
+       return fd;
+
+error:
+       if (fd > 0)
+               close(fd);
+       return -1;
+}
+
+/* Callback to handle the rx burst of packets to the correct interface and
+ * file descriptor(s) in a multi-queue setup.
+ */
+static uint16_t
+pmd_rx_burst(void *queue, struct rte_mbuf **bufs, uint16_t nb_pkts)
+{
+       int len;
+       struct rte_mbuf *mbuf;
+       struct rx_queue *rxq = queue;
+       uint16_t num_rx;
+       unsigned long num_rx_bytes = 0;
+
+       for (num_rx = 0; num_rx < nb_pkts; ) {
+               /* allocate the next mbuf */
+               mbuf = rte_pktmbuf_alloc(rxq->mp);
+               if (unlikely(!mbuf)) {
+                       RTE_LOG(WARNING, PMD, "Unable to allocate mbuf\n");
+                       break;
+               }
+
+               len = read(rxq->fd, rte_pktmbuf_mtod(mbuf, char *),
+                          rte_pktmbuf_tailroom(mbuf));
+               if (len <= 0) {
+                       rte_pktmbuf_free(mbuf);
+                       break;
+               }
+
+               mbuf->data_len = len;
+               mbuf->pkt_len = len;
+               mbuf->port = rxq->in_port;
+
+               /* account for the receive frame */
+               bufs[num_rx++] = mbuf;
+               num_rx_bytes += mbuf->pkt_len;
+       }
+       rxq->stats.ipackets += num_rx;
+       rxq->stats.ibytes += num_rx_bytes;
+
+       return num_rx;
+}
+
+/* Callback to handle sending packets from the tap interface
+ */
+static uint16_t
+pmd_tx_burst(void *queue, struct rte_mbuf **bufs, uint16_t nb_pkts)
+{
+       struct rte_mbuf *mbuf;
+       struct tx_queue *txq = queue;
+       struct pollfd pfd;
+       uint16_t num_tx = 0;
+       unsigned long num_tx_bytes = 0;
+       int i, n;
+
+       if (unlikely(nb_pkts == 0))
+               return 0;
+
+       pfd.events = POLLOUT;
+       pfd.fd = txq->fd;
+       for (i = 0; i < nb_pkts; i++) {
+               n = poll(&pfd, 1, 0);
+
+               if (n <= 0)
+                       break;
+
+               if (pfd.revents & POLLOUT) {
+                       /* copy the tx frame data */
+                       mbuf = bufs[num_tx];
+                       n = write(pfd.fd, rte_pktmbuf_mtod(mbuf, void*),
+                                 rte_pktmbuf_pkt_len(mbuf));
+                       if (n <= 0)
+                               break;
+
+                       num_tx++;
+                       num_tx_bytes += mbuf->pkt_len;
+                       rte_pktmbuf_free(mbuf);
+               }
+       }
+
+       txq->stats.opackets += num_tx;
+       txq->stats.errs += nb_pkts - num_tx;
+       txq->stats.obytes += num_tx_bytes;
+
+       return num_tx;
+}
+
+static int
+tap_dev_start(struct rte_eth_dev *dev)
+{
+       /* Force the Link up */
+       dev->data->dev_link.link_status = ETH_LINK_UP;
+
+       return 0;
+}
+
+/* This function gets called when the current port gets stopped.
+ */
+static void
+tap_dev_stop(struct rte_eth_dev *dev)
+{
+       int i;
+       struct pmd_internals *internals = dev->data->dev_private;
+
+       for (i = 0; i < internals->nb_queues; i++)
+               if (internals->fds[i] != -1)
+                       close(internals->fds[i]);
+
+       dev->data->dev_link.link_status = ETH_LINK_DOWN;
+}
+
+static int
+tap_dev_configure(struct rte_eth_dev *dev __rte_unused)
+{
+       return 0;
+}
+
+static void
+tap_dev_info(struct rte_eth_dev *dev, struct rte_eth_dev_info *dev_info)
+{
+       struct pmd_internals *internals = dev->data->dev_private;
+
+       dev_info->if_index = internals->if_index;
+       dev_info->max_mac_addrs = 1;
+       dev_info->max_rx_pktlen = (uint32_t)ETHER_MAX_VLAN_FRAME_LEN;
+       dev_info->max_rx_queues = internals->nb_queues;
+       dev_info->max_tx_queues = internals->nb_queues;
+       dev_info->min_rx_bufsize = 0;
+       dev_info->pci_dev = NULL;
+}
+
+static void
+tap_stats_get(struct rte_eth_dev *dev, struct rte_eth_stats *tap_stats)
+{
+       unsigned int i, imax;
+       unsigned long rx_total = 0, tx_total = 0, tx_err_total = 0;
+       unsigned long rx_bytes_total = 0, tx_bytes_total = 0;
+       const struct pmd_internals *pmd = dev->data->dev_private;
+
+       imax = (pmd->nb_queues < RTE_ETHDEV_QUEUE_STAT_CNTRS) ?
+               pmd->nb_queues : RTE_ETHDEV_QUEUE_STAT_CNTRS;
+
+       for (i = 0; i < imax; i++) {
+               tap_stats->q_ipackets[i] = pmd->rxq[i].stats.ipackets;
+               tap_stats->q_ibytes[i] = pmd->rxq[i].stats.ibytes;
+               rx_total += tap_stats->q_ipackets[i];
+               rx_bytes_total += tap_stats->q_ibytes[i];
+       }
+
+       for (i = 0; i < imax; i++) {
+               tap_stats->q_opackets[i] = pmd->txq[i].stats.opackets;
+               tap_stats->q_errors[i] = pmd->txq[i].stats.errs;
+               tap_stats->q_obytes[i] = pmd->txq[i].stats.obytes;
+               tx_total += tap_stats->q_opackets[i];
+               tx_err_total += tap_stats->q_errors[i];
+               tx_bytes_total += tap_stats->q_obytes[i];
+       }
+
+       tap_stats->ipackets = rx_total;
+       tap_stats->ibytes = rx_bytes_total;
+       tap_stats->opackets = tx_total;
+       tap_stats->oerrors = tx_err_total;
+       tap_stats->obytes = tx_bytes_total;
+}
+
+static void
+tap_stats_reset(struct rte_eth_dev *dev)
+{
+       int i;
+       struct pmd_internals *pmd = dev->data->dev_private;
+
+       for (i = 0; i < pmd->nb_queues; i++) {
+               pmd->rxq[i].stats.ipackets = 0;
+               pmd->rxq[i].stats.ibytes = 0;
+       }
+
+       for (i = 0; i < pmd->nb_queues; i++) {
+               pmd->txq[i].stats.opackets = 0;
+               pmd->txq[i].stats.errs = 0;
+               pmd->txq[i].stats.obytes = 0;
+       }
+}
+
+static void
+tap_dev_close(struct rte_eth_dev *dev __rte_unused)
+{
+}
+
+static void
+tap_rx_queue_release(void *queue)
+{
+       struct rx_queue *rxq = queue;
+
+       if (rxq && (rxq->fd > 0)) {
+               close(rxq->fd);
+               rxq->fd = -1;
+       }
+}
+
+static void
+tap_tx_queue_release(void *queue)
+{
+       struct tx_queue *txq = queue;
+
+       if (txq && (txq->fd > 0)) {
+               close(txq->fd);
+               txq->fd = -1;
+       }
+}
+
+static int
+tap_link_update(struct rte_eth_dev *dev __rte_unused,
+               int wait_to_complete __rte_unused)
+{
+       return 0;
+}
+
+static int
+tap_setup_queue(struct rte_eth_dev *dev,
+               struct pmd_internals *internals,
+               uint16_t qid)
+{
+       struct rx_queue *rx = &internals->rxq[qid];
+       struct tx_queue *tx = &internals->txq[qid];
+       int fd;
+
+       fd = rx->fd;
+       if (fd < 0) {
+               fd = tx->fd;
+               if (fd < 0) {
+                       RTE_LOG(INFO, PMD, "Add queue to TAP %s for qid

%d\n",

drivers/net/tap/rte_eth_tap.c new file +57 -0
+                               dev->data->name, qid);
+                       fd = tun_alloc(dev->data->name);
+                       if (fd < 0) {
+                               RTE_LOG(ERR, PMD, "tun_alloc(%s) failed\n",
+                                       dev->data->name);
+                               return -1;
+                       }
+               }
+       }
+       dev->data->rx_queues[qid] = rx;
+       dev->data->tx_queues[qid] = tx;
+
+       rx->fd = fd;
+       tx->fd = fd;
+
+       return fd;
+}
+
+static int
+tap_rx_queue_setup(struct rte_eth_dev *dev,
+                  uint16_t rx_queue_id,
+                  uint16_t nb_rx_desc __rte_unused,
+                  unsigned int socket_id __rte_unused,
+                  const struct rte_eth_rxconf *rx_conf __rte_unused,
+                  struct rte_mempool *mp)
+{
+       struct pmd_internals *internals = dev->data->dev_private;
+       uint16_t buf_size;
+       int fd;
+
+       if ((rx_queue_id >= internals->nb_queues) || !mp) {
+               RTE_LOG(ERR, PMD, "nb_queues %d mp %p\n",
+                       internals->nb_queues, mp);
+               return -1;
+       }
+
+       internals->rxq[rx_queue_id].mp = mp;
+       internals->rxq[rx_queue_id].in_port = dev->data->port_id;
+
+       /* Now get the space available for data in the mbuf */
+       buf_size = (uint16_t)(rte_pktmbuf_data_room_size(mp) -
+                               RTE_PKTMBUF_HEADROOM);
+
+       if (buf_size < ETH_FRAME_LEN) {
+               RTE_LOG(ERR, PMD,
+                       "%s: %d bytes will not fit in mbuf (%d bytes)\n",
+                       dev->data->name, ETH_FRAME_LEN, buf_size);
+               return -ENOMEM;
+       }
+
+       fd = tap_setup_queue(dev, internals, rx_queue_id);
+       if (fd == -1)
+               return -1;
+
+       internals->fds[rx_queue_id] = fd;
+       RTE_LOG(INFO, PMD, "RX TAP device name %s, qid %d on fd %d\n",
+               dev->data->name, rx_queue_id, internals->rxq[rx_queue_id].

fd);

drivers/net/tap/rte_eth_tap.c new file +23 -0
+
+       return 0;
+}
+
+static int
+tap_tx_queue_setup(struct rte_eth_dev *dev,
+                  uint16_t tx_queue_id,
+                  uint16_t nb_tx_desc __rte_unused,
+                  unsigned int socket_id __rte_unused,
+                  const struct rte_eth_txconf *tx_conf __rte_unused)
+{
+       struct pmd_internals *internals = dev->data->dev_private;
+       int ret;
+
+       if (tx_queue_id >= internals->nb_queues)
+               return -1;
+
+       ret = tap_setup_queue(dev, internals, tx_queue_id);
+       if (ret == -1)
+               return -1;
+
+       RTE_LOG(INFO, PMD, "TX TAP device name %s, qid %d on fd %d\n",
+               dev->data->name, tx_queue_id, internals->txq[tx_queue_id].

fd);

drivers/net/tap/rte_eth_tap.c new file +44 -0
+
+       return 0;
+}
+
+static const struct eth_dev_ops ops = {
+       .dev_start              = tap_dev_start,
+       .dev_stop               = tap_dev_stop,
+       .dev_close              = tap_dev_close,
+       .dev_configure          = tap_dev_configure,
+       .dev_infos_get          = tap_dev_info,
+       .rx_queue_setup         = tap_rx_queue_setup,
+       .tx_queue_setup         = tap_tx_queue_setup,
+       .rx_queue_release       = tap_rx_queue_release,
+       .tx_queue_release       = tap_tx_queue_release,
+       .link_update            = tap_link_update,
+       .stats_get              = tap_stats_get,
+       .stats_reset            = tap_stats_reset,
+};
+
+static int
+pmd_mac_address(int fd, struct rte_eth_dev *dev, struct ether_addr *addr)
+{
+       struct ifreq ifr;
+
+       if ((fd <= 0) || !dev || !addr)
+               return -1;
+
+       memset(&ifr, 0, sizeof(ifr));
+
+       if (ioctl(fd, SIOCGIFHWADDR, &ifr) == -1) {
+               RTE_LOG(ERR, PMD, "ioctl failed (SIOCGIFHWADDR) (%s)\n",
+                       ifr.ifr_name);
+               return -1;
+       }
+
+       /* Set the host based MAC address to this special MAC format */
+       ifr.ifr_hwaddr.sa_data[0] = 'T';
+       ifr.ifr_hwaddr.sa_data[1] = 'a';
+       ifr.ifr_hwaddr.sa_data[2] = 'p';
+       ifr.ifr_hwaddr.sa_data[3] = '-';
+       ifr.ifr_hwaddr.sa_data[4] = dev->data->port_id;
+       ifr.ifr_hwaddr.sa_data[5] = dev->data->numa_node;
+       if (ioctl(fd, SIOCSIFHWADDR, &ifr) == -1) {
+               RTE_LOG(ERR, PMD, "%s: ioctl failed (SIOCSIFHWADDR)

(%s)\n",

drivers/net/tap/rte_eth_tap.c new file +5 -0
+                       dev->data->name, ifr.ifr_name);
+               return -1;
+       }
+
+       /* Set the local application MAC address, needs to be different

then

drivers/net/tap/rte_eth_tap.c new file +24 -0
+        * the host based MAC address.
+        */
+       ifr.ifr_hwaddr.sa_data[0] = 'd';
+       ifr.ifr_hwaddr.sa_data[1] = 'n';
+       ifr.ifr_hwaddr.sa_data[2] = 'e';
+       ifr.ifr_hwaddr.sa_data[3] = 't';
+       ifr.ifr_hwaddr.sa_data[4] = dev->data->port_id;
+       ifr.ifr_hwaddr.sa_data[5] = dev->data->numa_node;
+       rte_memcpy(addr, ifr.ifr_hwaddr.sa_data, ETH_ALEN);
+
+       return 0;
+}
+
+static int
+eth_dev_tap_create(const char *name, char *tap_name)
+{
+       int numa_node = rte_socket_id();
+       struct rte_eth_dev *dev = NULL;
+       struct pmd_internals *pmd = NULL;
+       struct rte_eth_dev_data *data = NULL;
+       int i, fd = -1;
+
+       RTE_LOG(INFO, PMD,
+               "%s: Create TAP Ethernet device with %d queues on numa

%u\n",

drivers/net/tap/rte_eth_tap.c new file +131 -0
+                name, RTE_PMD_TAP_MAX_QUEUES, rte_socket_id());
+
+       data = rte_zmalloc_socket(tap_name, sizeof(*data), 0, numa_node);
+       if (!data) {
+               RTE_LOG(INFO, PMD, "Failed to allocate data\n");
+               goto error_exit;
+       }
+
+       pmd = rte_zmalloc_socket(tap_name, sizeof(*pmd), 0, numa_node);
+       if (!pmd) {
+               RTE_LOG(INFO, PMD, "Unable to allocate internal struct\n");
+               goto error_exit;
+       }
+
+       /* Use the name and not the tap_name */
+       dev = rte_eth_dev_allocate(tap_name);
+       if (!dev) {
+               RTE_LOG(INFO, PMD, "Unable to allocate device struct\n");
+               goto error_exit;
+       }
+
+       snprintf(pmd->name, sizeof(pmd->name), "%s", tap_name);
+
+       pmd->nb_queues = RTE_PMD_TAP_MAX_QUEUES;
+
+       /* Setup some default values */
+       data->dev_private = pmd;
+       data->port_id = dev->data->port_id;
+       data->dev_flags = RTE_ETH_DEV_DETACHABLE;
+       data->kdrv = RTE_KDRV_NONE;
+       data->drv_name = pmd_tap_drv.driver.name;
+       data->numa_node = numa_node;
+
+       data->dev_link = pmd_link;
+       data->mac_addrs = &pmd->eth_addr;
+       data->nb_rx_queues = pmd->nb_queues;
+       data->nb_tx_queues = pmd->nb_queues;
+
+       dev->data = data;
+       dev->dev_ops = &ops;
+       dev->driver = NULL;
+       dev->rx_pkt_burst = pmd_rx_burst;
+       dev->tx_pkt_burst = pmd_tx_burst;
+       snprintf(dev->data->name, sizeof(dev->data->name), "%s", name);
+
+       /* Create the first Tap device */
+       fd = tun_alloc(tap_name);
+       if (fd < 0) {
+               RTE_LOG(INFO, PMD, "tun_alloc() failed\n");
+               goto error_exit;
+       }
+
+       /* Presetup the fds to -1 as being not working */
+       for (i = 0; i < RTE_PMD_TAP_MAX_QUEUES; i++) {
+               pmd->fds[i] = -1;
+               pmd->rxq[i].fd = -1;
+               pmd->txq[i].fd = -1;
+       }
+
+       /* Take the TUN/TAP fd and place in the first location */
+       pmd->rxq[0].fd = fd;
+       pmd->txq[0].fd = fd;
+       pmd->fds[0] = fd;
+
+       if (pmd_mac_address(fd, dev, &pmd->eth_addr) < 0) {
+               RTE_LOG(INFO, PMD, "Unable to get MAC address\n");
+               goto error_exit;
+       }
+
+       return 0;
+
+error_exit:
+       RTE_PMD_DEBUG_TRACE("Unable to initialize %s\n", name);
+
+       rte_free(data);
+       rte_free(pmd);
+
+       rte_eth_dev_release_port(dev);
+
+       return -EINVAL;
+}
+
+static int
+set_interface_name(const char *key __rte_unused,
+                  const char *value,
+                  void *extra_args)
+{
+       char *name = (char *)extra_args;
+
+       if (value)
+               snprintf(name, RTE_ETH_NAME_MAX_LEN - 1, "%s", value);
+       else
+               snprintf(name, RTE_ETH_NAME_MAX_LEN - 1, "%s%d",
+                        DEFAULT_TAP_NAME, (tap_unit - 1));
+
+       return 0;
+}
+
+static int
+set_interface_speed(const char *key __rte_unused,
+                   const char *value,
+                   void *extra_args)
+{
+       *(int *)extra_args = (value) ? atoi(value) : ETH_SPEED_NUM_10G;
+
+       return 0;
+}
+
+/* Open a TAP interface device.
+ */
+static int
+rte_pmd_tap_probe(const char *name, const char *params)
+{
+       int ret;
+       struct rte_kvargs *kvlist = NULL;
+       int speed;
+       char tap_name[RTE_ETH_NAME_MAX_LEN];
+
+       speed = ETH_SPEED_NUM_10G;
+       snprintf(tap_name, sizeof(tap_name), "%s%d",
+                DEFAULT_TAP_NAME, tap_unit++);
+
+       RTE_LOG(INFO, PMD, "Initializing pmd_tap for %s as %s\n",
+               name, tap_name);
+
+       if (params && (params[0] != '\0')) {
+               RTE_LOG(INFO, PMD, "paramaters (%s)\n", params);
+
+               kvlist = rte_kvargs_parse(params, valid_arguments);
+               if (kvlist) {
+                       if (rte_kvargs_count(kvlist, ETH_TAP_SPEED_ARG) ==

1) {

drivers/net/tap/rte_eth_tap.c new file +3 -0
+                               ret = rte_kvargs_process(kvlist,
+                                                        ETH_TAP_SPEED_ARG,
+

&set_interface_speed,

drivers/net/tap/rte_eth_tap.c new file +6 -0
+                                                        &speed);
+                               if (ret == -1)
+                                       goto leave;
+                       }
+
+                       if (rte_kvargs_count(kvlist, ETH_TAP_IFACE_ARG) ==

1) {

drivers/net/tap/rte_eth_tap.c new file +3 -0
+                               ret = rte_kvargs_process(kvlist,
+                                                        ETH_TAP_IFACE_ARG,
+

&set_interface_name,

drivers/net/tap/rte_eth_tap.c new file +58 -0
+                                                        tap_name);
+                               if (ret == -1)
+                                       goto leave;
+                       }
+               }
+       }
+       pmd_link.link_speed = speed;
+
+       ret = eth_dev_tap_create(name, tap_name);
+
+leave:
+       if (ret == -1) {
+               RTE_LOG(INFO, PMD, "Failed to create pmd for %s as %s\n",
+                       name, tap_name);
+               tap_unit--;             /* Restore the unit number */
+       }
+       rte_kvargs_free(kvlist);
+
+       return ret;
+}
+
+/* detach a TAP device.
+ */
+static int
+rte_pmd_tap_remove(const char *name)
+{
+       struct rte_eth_dev *eth_dev = NULL;
+       struct pmd_internals *internals;
+       int i;
+
+       RTE_LOG(INFO, PMD, "Closing TUN/TAP Ethernet device on numa %u\n",
+               rte_socket_id());
+
+       /* find the ethdev entry */
+       eth_dev = rte_eth_dev_allocated(name);
+       if (!eth_dev)
+               return 0;
+
+       internals = eth_dev->data->dev_private;
+       for (i = 0; i < internals->nb_queues; i++)
+               if (internals->fds[i] != -1)
+                       close(internals->fds[i]);
+
+       rte_free(eth_dev->data->dev_private);
+       rte_free(eth_dev->data);
+
+       rte_eth_dev_release_port(eth_dev);
+
+       return 0;
+}
+
+static struct rte_vdev_driver pmd_tap_drv = {
+       .probe = rte_pmd_tap_probe,
+       .remove = rte_pmd_tap_remove,
+};
+RTE_PMD_REGISTER_VDEV(net_tap, pmd_tap_drv);
+RTE_PMD_REGISTER_ALIAS(net_tap, eth_tap);
+RTE_PMD_REGISTER_PARAM_STRING(net_tap, "iface=<string>,speed=N");
drivers/net/tap/rte_pmd_tap_version.map new file +4 -0
new file mode 100644
index 0000000..31eca32
@@ -0,0 +1,4 @@
+DPDK_17.02 {
+
+       local: *;
+};
mk/rte.app.mk +1 -3
index f75f0e2..02c32ae 100644
@@ -124,6 +124,7 @@ _LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_PCAP)       +=
-lrte_pmd_pcap -lpcap
 _LDLIBS-$(CONFIG_RTE_LIBRTE_QEDE_PMD)       += -lrte_pmd_qede
 _LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_RING)       += -lrte_pmd_ring
 _LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_SZEDATA2)   += -lrte_pmd_szedata2 -lsze2
+_LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_TAP)        += -lrte_pmd_tap
 _LDLIBS-$(CONFIG_RTE_LIBRTE_THUNDERX_NICVF_PMD) +=
-lrte_pmd_thunderx_nicvf -lm
 _LDLIBS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD)     += -lrte_pmd_virtio
 ifeq ($(CONFIG_RTE_LIBRTE_VHOST),y)
--

2.8.0.GIT

#38 lore raw

Re: [PATCH v12] net/tap: new TUN/TAP device PMD

Quoted 6 lines · Marc

Keith,

A bit late, but two very high level questions. Do you have performance numbers compared to KNI? Did you consider using AF_PACKET PACKET_MMAP which could potentially reduce the number of syscalls to 1 for RX and TX of a burst?

Hi Marc,

I was not trying to create a high performance interface, just a Tap interface to use standard applications and calls to send/receive traffic to the DPDK application. I did not expect other then some management like interface in the application would use the Tap PMD.

Quoted 1,107 lines

Marc

On 12 December 2016 at 15:38, Keith Wiles <[email protected]> wrote:
The PMD allows for DPDK and the host to communicate using a raw device interface on the host and in the DPDK application. The device created is a Tap device with a L2 packet header.

v12- Fixup minor changes for driver_name and version number v11- Add the tap.rst to the nic/index.rst file
v10- Change the string name used to allow for multiple devices. v9 - Fix up the docs to use correct syntax
v8 - Fix issue with tap_tx_queue_setup() not return zero on success. v7 - Reword the comment in common_base and fix the data->name issue v6 - fixed the checkpatch issues
v5 - merge in changes from list review see related emails
fixed many minor edits
v4 - merge with latest driver changes
v3 - fix includes by removing ifdef for other type besides Linux
Fix the copyright notice in the Makefile
v2 - merge all of the patches into one patch
Fix a typo on naming the tap device
Update the maintainers list

---
 MAINTAINERS                             |   5 +
 config/common_base                      |   9 +
 config/common_linuxapp                  |   1 +
 doc/guides/nics/index.rst               |   1 +
 doc/guides/nics/tap.rst                 | 136 ++++++
 drivers/net/Makefile                    |   1 +
 drivers/net/tap/Makefile                |  57 +++
 drivers/net/tap/rte_eth_tap.c           | 765 ++++++++++++++++++++++++++++++++
 drivers/net/tap/rte_pmd_tap_version.map |   4 +
 mk/rte.app.mk                           |   1 +
 10 files changed, 980 insertions(+)
 create mode 100644 doc/guides/nics/tap.rst
 create mode 100644 drivers/net/tap/Makefile
 create mode 100644 drivers/net/tap/rte_eth_tap.c
 create mode 100644 drivers/net/tap/rte_pmd_tap_version.map
MAINTAINERS +5 -0
index 26d9590..842fb6d 100644
@@ -398,6 +398,11 @@ F: doc/guides/nics/pcap_ring.rst
 F: app/test/test_pmd_ring.c
 F: app/test/test_pmd_ring_perf.c
 
+Tap PMD
+M: Keith Wiles <[email protected]>
+F: drivers/net/tap
+F: doc/guides/nics/tap.rst
+
 Null Networking PMD
 M: Tetsuya Mukawa <[email protected]>
 F: drivers/net/null/
config/common_base +9 -0
index 652a839..eb51cdb 100644
@@ -590,3 +590,12 @@ CONFIG_RTE_APP_TEST_RESOURCE_TAR=n
 CONFIG_RTE_TEST_PMD=y
 CONFIG_RTE_TEST_PMD_RECORD_CORE_CYCLES=n
 CONFIG_RTE_TEST_PMD_RECORD_BURST_STATS=n
+
+#
+# Compile the TAP PMD
+#
+# The TAP PMD is currently only built for Linux and the
+# option is enabled by default in common_linuxapp file,
+# set to 'n' in the common_base file.
+#
+CONFIG_RTE_LIBRTE_PMD_TAP=n
config/common_linuxapp +1 -0
index 2483dfa..782b503 100644
@@ -44,3 +44,4 @@ CONFIG_RTE_LIBRTE_PMD_VHOST=y
 CONFIG_RTE_LIBRTE_PMD_AF_PACKET=y
 CONFIG_RTE_LIBRTE_POWER=y
 CONFIG_RTE_VIRTIO_USER=y
+CONFIG_RTE_LIBRTE_PMD_TAP=y
doc/guides/nics/index.rst +1 -0
index 92d56a5..af92529 100644
@@ -51,6 +51,7 @@ Network Interface Controller Drivers
     nfp
     qede
     szedata2
+    tap
     thunderx
     virtio
     vhost
doc/guides/nics/tap.rst new file +136 -0
new file mode 100644
index 0000000..622b9e7
@@ -0,0 +1,136 @@
+..  BSD LICENSE
+    Copyright(c) 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.
+
+Tun/Tap Poll Mode Driver
+========================
+
+The ``rte_eth_tap.c`` PMD creates a device using TUN/TAP interfaces on the
+local host. The PMD allows for DPDK and the host to communicate using a raw
+device interface on the host and in the DPDK application.
+
+The device created is a TAP device, which sends/receives packet in a raw
+format with a L2 header. The usage for a TAP PMD is for connectivity to the
+local host using a TAP interface. When the TAP PMD is initialized it will
+create a number of tap devices in the host accessed via ``ifconfig -a`` or
+``ip`` command. The commands can be used to assign and query the virtual like
+device.
+
+These TAP interfaces can be used with Wireshark or tcpdump or Pktgen-DPDK
+along with being able to be used as a network connection to the DPDK
+application. The method enable one or more interfaces is to use the
+``--vdev=net_tap`` option on the DPDK application command line. Each
+``--vdev=net_tap`` option give will create an interface named dtap0, dtap1,
+and so on.
+
+The interfaced name can be changed by adding the ``iface=foo0``, for example::
+
+   --vdev=net_tap,iface=foo0 --vdev=net_tap,iface=foo1, ...
+
+Also the speed of the interface can be changed from 10G to whatever number
+needed, but the interface does not enforce that speed, for example::
+
+   --vdev=net_tap,iface=foo0,speed=25000
+
+After the DPDK application is started you can send and receive packets on the
+interface using the standard rx_burst/tx_burst APIs in DPDK. From the host
+point of view you can use any host tool like tcpdump, Wireshark, ping, Pktgen
+and others to communicate with the DPDK application. The DPDK application may
+not understand network protocols like IPv4/6, UDP or TCP unless the
+application has been written to understand these protocols.
+
+If you need the interface as a real network interface meaning running and has
+a valid IP address then you can do this with the following commands::
+
+   sudo ip link set dtap0 up; sudo ip addr add 192.168.0.250/24 dev dtap0
+   sudo ip link set dtap1 up; sudo ip addr add 192.168.1.250/24 dev dtap1
+
+Please change the IP addresses as you see fit.
+
+If routing is enabled on the host you can also communicate with the DPDK App
+over the internet via a standard socket layer application as long as you
+account for the protocol handing in the application.
+
+If you have a Network Stack in your DPDK application or something like it you
+can utilize that stack to handle the network protocols. Plus you would be able
+to address the interface using an IP address assigned to the internal
+interface.
+
+Example
+-------
+
+The following is a simple example of using the TUN/TAP PMD with the Pktgen
+packet generator. It requires that the ``socat`` utility is installed on the
+test system.
+
+Build DPDK, then pull down Pktgen and build pktgen using the DPDK SDK/Target
+used to build the dpdk you pulled down.
+
+Run pktgen from the pktgen directory in a terminal with a commandline like the
+following::
+
+    sudo ./app/app/x86_64-native-linuxapp-gcc/app/pktgen -l 1-5 -n 4        \
+     --proc-type auto --log-level 8 --socket-mem 512,512 --file-prefix pg   \
+     --vdev=net_tap --vdev=net_tap -b 05:00.0 -b 05:00.1                    \
+     -b 04:00.0 -b 04:00.1 -b 04:00.2 -b 04:00.3                            \
+     -b 81:00.0 -b 81:00.1 -b 81:00.2 -b 81:00.3                            \
+     -b 82:00.0 -b 83:00.0 -- -T -P -m [2:3].0 -m [4:5].1                   \
+     -f themes/black-yellow.theme
+
+.. Note:
+
+   Change the ``-b`` options to blacklist all of your physical ports. The
+   following command line is all one line.
+
+   Also, ``-f themes/black-yellow.theme`` is optional if the default colors
+   work on your system configuration. See the Pktgen docs for more
+   information.
+
+Verify with ``ifconfig -a`` command in a different xterm window, should have a
+``dtap0`` and ``dtap1`` interfaces created.
+
+Next set the links for the two interfaces to up via the commands below::
+
+    sudo ip link set dtap0 up; sudo ip addr add 192.168.0.250/24 dev dtap0
+    sudo ip link set dtap1 up; sudo ip addr add 192.168.1.250/24 dev dtap1
+
+Then use socat to create a loopback for the two interfaces::
+
+    sudo socat interface:dtap0 interface:dtap1
+
+Then on the Pktgen command line interface you can start sending packets using
+the commands ``start 0`` and ``start 1`` or you can start both at the same
+time with ``start all``. The command ``str`` is an alias for ``start all`` and
+``stp`` is an alias for ``stop all``.
+
+While running you should see the 64 byte counters increasing to verify the
+traffic is being looped back. You can use ``set all size XXX`` to change the
+size of the packets after you stop the traffic. Use the pktgen ``help``
+command to see a list of all commands. You can also use the ``-f`` option to
+load commands at startup.
drivers/net/Makefile +1 -0
index bc93230..e366a85 100644
@@ -51,6 +51,7 @@ DIRS-$(CONFIG_RTE_LIBRTE_PMD_PCAP) += pcap
 DIRS-$(CONFIG_RTE_LIBRTE_QEDE_PMD) += qede
 DIRS-$(CONFIG_RTE_LIBRTE_PMD_RING) += ring
 DIRS-$(CONFIG_RTE_LIBRTE_PMD_SZEDATA2) += szedata2
+DIRS-$(CONFIG_RTE_LIBRTE_PMD_TAP) += tap
 DIRS-$(CONFIG_RTE_LIBRTE_THUNDERX_NICVF_PMD) += thunderx
 DIRS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD) += virtio
 DIRS-$(CONFIG_RTE_LIBRTE_VMXNET3_PMD) += vmxnet3
drivers/net/tap/Makefile new file +57 -0
new file mode 100644
index 0000000..e18f30c
@@ -0,0 +1,57 @@
+#   BSD LICENSE
+#
+#   Copyright(c) 2016 Intel Corporation. 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_pmd_tap.a
+
+EXPORT_MAP := rte_pmd_tap_version.map
+
+LIBABIVER := 1
+
+CFLAGS += -O3
+CFLAGS += $(WERROR_FLAGS)
+
+#
+# all source are stored in SRCS-y
+#
+SRCS-$(CONFIG_RTE_LIBRTE_PMD_TAP) += rte_eth_tap.c
+
+# this lib depends upon:
+DEPDIRS-$(CONFIG_RTE_LIBRTE_PMD_TAP) += lib/librte_eal
+DEPDIRS-$(CONFIG_RTE_LIBRTE_PMD_TAP) += lib/librte_mbuf
+DEPDIRS-$(CONFIG_RTE_LIBRTE_PMD_TAP) += lib/librte_mempool
+DEPDIRS-$(CONFIG_RTE_LIBRTE_PMD_TAP) += lib/librte_ether
+DEPDIRS-$(CONFIG_RTE_LIBRTE_PMD_TAP) += lib/librte_kvargs
+
+include $(RTE_SDK)/mk/rte.lib.mk
drivers/net/tap/rte_eth_tap.c new file +765 -0
new file mode 100644
index 0000000..976f2d9
@@ -0,0 +1,765 @@
+/*-
+ *   BSD LICENSE
+ *
+ *   Copyright(c) 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.
+ */
+
+#include <rte_mbuf.h>
+#include <rte_ethdev.h>
+#include <rte_malloc.h>
+#include <rte_vdev.h>
+#include <rte_kvargs.h>
+
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <sys/socket.h>
+#include <sys/ioctl.h>
+#include <sys/mman.h>
+#include <unistd.h>
+#include <poll.h>
+#include <arpa/inet.h>
+#include <linux/if.h>
+#include <linux/if_tun.h>
+#include <linux/if_ether.h>
+#include <fcntl.h>
+
+/* Linux based path to the TUN device */
+#define TUN_TAP_DEV_PATH        "/dev/net/tun"
+#define DEFAULT_TAP_NAME        "dtap"
+
+#define ETH_TAP_IFACE_ARG       "iface"
+#define ETH_TAP_SPEED_ARG       "speed"
+
+#define RTE_PMD_TAP_MAX_QUEUES 16
+
+static struct rte_vdev_driver pmd_tap_drv;
+
+static const char *valid_arguments[] = {
+       ETH_TAP_IFACE_ARG,
+       ETH_TAP_SPEED_ARG,
+       NULL
+};
+
+static int tap_unit;
+
+static struct rte_eth_link pmd_link = {
+       .link_speed = ETH_SPEED_NUM_10G,
+       .link_duplex = ETH_LINK_FULL_DUPLEX,
+       .link_status = ETH_LINK_DOWN,
+       .link_autoneg = ETH_LINK_SPEED_AUTONEG
+};
+
+struct pkt_stats {
+       uint64_t opackets;              /* Number of output packets */
+       uint64_t ipackets;              /* Number of input packets */
+       uint64_t obytes;                /* Number of bytes on output */
+       uint64_t ibytes;                /* Number of bytes on input */
+       uint64_t errs;                  /* Number of error packets */
+};
+
+struct rx_queue {
+       struct rte_mempool *mp;         /* Mempool for RX packets */
+       uint16_t in_port;               /* Port ID */
+       int fd;
+
+       struct pkt_stats stats;         /* Stats for this RX queue */
+};
+
+struct tx_queue {
+       int fd;
+       struct pkt_stats stats;         /* Stats for this TX queue */
+};
+
+struct pmd_internals {
+       char name[RTE_ETH_NAME_MAX_LEN];        /* Internal Tap device name */
+       uint16_t nb_queues;             /* Number of queues supported */
+       struct ether_addr eth_addr;     /* Mac address of the device port */
+
+       int if_index;                   /* IF_INDEX for the port */
+       int fds[RTE_PMD_TAP_MAX_QUEUES]; /* List of all file descriptors */
+
+       struct rx_queue rxq[RTE_PMD_TAP_MAX_QUEUES];    /* List of RX queues */
+       struct tx_queue txq[RTE_PMD_TAP_MAX_QUEUES];    /* List of TX queues */
+};
+
+/* Tun/Tap allocation routine
+ *
+ * name is the number of the interface to use, unless NULL to take the host
+ * supplied name.
+ */
+static int
+tun_alloc(char *name)
+{
+       struct ifreq ifr;
+       unsigned int features;
+       int fd;
+
+       memset(&ifr, 0, sizeof(struct ifreq));
+
+       ifr.ifr_flags = IFF_TAP | IFF_NO_PI;
+       if (name && name[0])
+               strncpy(ifr.ifr_name, name, IFNAMSIZ);
+
+       fd = open(TUN_TAP_DEV_PATH, O_RDWR);
+       if (fd < 0) {
+               RTE_LOG(ERR, PMD, "Unable to create TAP interface");
+               goto error;
+       }
+
+       /* Grab the TUN features to verify we can work */
+       if (ioctl(fd, TUNGETFEATURES, &features) < 0) {
+               RTE_LOG(ERR, PMD, "Unable to get TUN/TAP features\n");
+               goto error;
+       }
+       RTE_LOG(DEBUG, PMD, "TUN/TAP Features %08x\n", features);
+
+       if (!(features & IFF_MULTI_QUEUE) && (RTE_PMD_TAP_MAX_QUEUES > 1)) {
+               RTE_LOG(DEBUG, PMD, "TUN/TAP device only one queue\n");
+               goto error;
+       } else if ((features & IFF_ONE_QUEUE) &&
+                       (RTE_PMD_TAP_MAX_QUEUES == 1)) {
+               ifr.ifr_flags |= IFF_ONE_QUEUE;
+               RTE_LOG(DEBUG, PMD, "Single queue only support\n");
+       } else {
+               ifr.ifr_flags |= IFF_MULTI_QUEUE;
+               RTE_LOG(DEBUG, PMD, "Multi-queue support for %d queues\n",
+                       RTE_PMD_TAP_MAX_QUEUES);
+       }
+
+       /* Set the TUN/TAP configuration and get the name if needed */
+       if (ioctl(fd, TUNSETIFF, (void *)&ifr) < 0) {
+               RTE_LOG(ERR, PMD, "Unable to set TUNSETIFF for %s\n",
+                       ifr.ifr_name);
+               perror("TUNSETIFF");
+               goto error;
+       }
+
+       /* Always set the file descriptor to non-blocking */
+       if (fcntl(fd, F_SETFL, O_NONBLOCK) < 0) {
+               RTE_LOG(ERR, PMD, "Unable to set to nonblocking\n");
+               perror("F_SETFL, NONBLOCK");
+               goto error;
+       }
+
+       /* If the name is different that new name as default */
+       if (name && strcmp(name, ifr.ifr_name))
+               snprintf(name, RTE_ETH_NAME_MAX_LEN - 1, "%s", ifr.ifr_name);
+
+       return fd;
+
+error:
+       if (fd > 0)
+               close(fd);
+       return -1;
+}
+
+/* Callback to handle the rx burst of packets to the correct interface and
+ * file descriptor(s) in a multi-queue setup.
+ */
+static uint16_t
+pmd_rx_burst(void *queue, struct rte_mbuf **bufs, uint16_t nb_pkts)
+{
+       int len;
+       struct rte_mbuf *mbuf;
+       struct rx_queue *rxq = queue;
+       uint16_t num_rx;
+       unsigned long num_rx_bytes = 0;
+
+       for (num_rx = 0; num_rx < nb_pkts; ) {
+               /* allocate the next mbuf */
+               mbuf = rte_pktmbuf_alloc(rxq->mp);
+               if (unlikely(!mbuf)) {
+                       RTE_LOG(WARNING, PMD, "Unable to allocate mbuf\n");
+                       break;
+               }
+
+               len = read(rxq->fd, rte_pktmbuf_mtod(mbuf, char *),
+                          rte_pktmbuf_tailroom(mbuf));
+               if (len <= 0) {
+                       rte_pktmbuf_free(mbuf);
+                       break;
+               }
+
+               mbuf->data_len = len;
+               mbuf->pkt_len = len;
+               mbuf->port = rxq->in_port;
+
+               /* account for the receive frame */
+               bufs[num_rx++] = mbuf;
+               num_rx_bytes += mbuf->pkt_len;
+       }
+       rxq->stats.ipackets += num_rx;
+       rxq->stats.ibytes += num_rx_bytes;
+
+       return num_rx;
+}
+
+/* Callback to handle sending packets from the tap interface
+ */
+static uint16_t
+pmd_tx_burst(void *queue, struct rte_mbuf **bufs, uint16_t nb_pkts)
+{
+       struct rte_mbuf *mbuf;
+       struct tx_queue *txq = queue;
+       struct pollfd pfd;
+       uint16_t num_tx = 0;
+       unsigned long num_tx_bytes = 0;
+       int i, n;
+
+       if (unlikely(nb_pkts == 0))
+               return 0;
+
+       pfd.events = POLLOUT;
+       pfd.fd = txq->fd;
+       for (i = 0; i < nb_pkts; i++) {
+               n = poll(&pfd, 1, 0);
+
+               if (n <= 0)
+                       break;
+
+               if (pfd.revents & POLLOUT) {
+                       /* copy the tx frame data */
+                       mbuf = bufs[num_tx];
+                       n = write(pfd.fd, rte_pktmbuf_mtod(mbuf, void*),
+                                 rte_pktmbuf_pkt_len(mbuf));
+                       if (n <= 0)
+                               break;
+
+                       num_tx++;
+                       num_tx_bytes += mbuf->pkt_len;
+                       rte_pktmbuf_free(mbuf);
+               }
+       }
+
+       txq->stats.opackets += num_tx;
+       txq->stats.errs += nb_pkts - num_tx;
+       txq->stats.obytes += num_tx_bytes;
+
+       return num_tx;
+}
+
+static int
+tap_dev_start(struct rte_eth_dev *dev)
+{
+       /* Force the Link up */
+       dev->data->dev_link.link_status = ETH_LINK_UP;
+
+       return 0;
+}
+
+/* This function gets called when the current port gets stopped.
+ */
+static void
+tap_dev_stop(struct rte_eth_dev *dev)
+{
+       int i;
+       struct pmd_internals *internals = dev->data->dev_private;
+
+       for (i = 0; i < internals->nb_queues; i++)
+               if (internals->fds[i] != -1)
+                       close(internals->fds[i]);
+
+       dev->data->dev_link.link_status = ETH_LINK_DOWN;
+}
+
+static int
+tap_dev_configure(struct rte_eth_dev *dev __rte_unused)
+{
+       return 0;
+}
+
+static void
+tap_dev_info(struct rte_eth_dev *dev, struct rte_eth_dev_info *dev_info)
+{
+       struct pmd_internals *internals = dev->data->dev_private;
+
+       dev_info->if_index = internals->if_index;
+       dev_info->max_mac_addrs = 1;
+       dev_info->max_rx_pktlen = (uint32_t)ETHER_MAX_VLAN_FRAME_LEN;
+       dev_info->max_rx_queues = internals->nb_queues;
+       dev_info->max_tx_queues = internals->nb_queues;
+       dev_info->min_rx_bufsize = 0;
+       dev_info->pci_dev = NULL;
+}
+
+static void
+tap_stats_get(struct rte_eth_dev *dev, struct rte_eth_stats *tap_stats)
+{
+       unsigned int i, imax;
+       unsigned long rx_total = 0, tx_total = 0, tx_err_total = 0;
+       unsigned long rx_bytes_total = 0, tx_bytes_total = 0;
+       const struct pmd_internals *pmd = dev->data->dev_private;
+
+       imax = (pmd->nb_queues < RTE_ETHDEV_QUEUE_STAT_CNTRS) ?
+               pmd->nb_queues : RTE_ETHDEV_QUEUE_STAT_CNTRS;
+
+       for (i = 0; i < imax; i++) {
+               tap_stats->q_ipackets[i] = pmd->rxq[i].stats.ipackets;
+               tap_stats->q_ibytes[i] = pmd->rxq[i].stats.ibytes;
+               rx_total += tap_stats->q_ipackets[i];
+               rx_bytes_total += tap_stats->q_ibytes[i];
+       }
+
+       for (i = 0; i < imax; i++) {
+               tap_stats->q_opackets[i] = pmd->txq[i].stats.opackets;
+               tap_stats->q_errors[i] = pmd->txq[i].stats.errs;
+               tap_stats->q_obytes[i] = pmd->txq[i].stats.obytes;
+               tx_total += tap_stats->q_opackets[i];
+               tx_err_total += tap_stats->q_errors[i];
+               tx_bytes_total += tap_stats->q_obytes[i];
+       }
+
+       tap_stats->ipackets = rx_total;
+       tap_stats->ibytes = rx_bytes_total;
+       tap_stats->opackets = tx_total;
+       tap_stats->oerrors = tx_err_total;
+       tap_stats->obytes = tx_bytes_total;
+}
+
+static void
+tap_stats_reset(struct rte_eth_dev *dev)
+{
+       int i;
+       struct pmd_internals *pmd = dev->data->dev_private;
+
+       for (i = 0; i < pmd->nb_queues; i++) {
+               pmd->rxq[i].stats.ipackets = 0;
+               pmd->rxq[i].stats.ibytes = 0;
+       }
+
+       for (i = 0; i < pmd->nb_queues; i++) {
+               pmd->txq[i].stats.opackets = 0;
+               pmd->txq[i].stats.errs = 0;
+               pmd->txq[i].stats.obytes = 0;
+       }
+}
+
+static void
+tap_dev_close(struct rte_eth_dev *dev __rte_unused)
+{
+}
+
+static void
+tap_rx_queue_release(void *queue)
+{
+       struct rx_queue *rxq = queue;
+
+       if (rxq && (rxq->fd > 0)) {
+               close(rxq->fd);
+               rxq->fd = -1;
+       }
+}
+
+static void
+tap_tx_queue_release(void *queue)
+{
+       struct tx_queue *txq = queue;
+
+       if (txq && (txq->fd > 0)) {
+               close(txq->fd);
+               txq->fd = -1;
+       }
+}
+
+static int
+tap_link_update(struct rte_eth_dev *dev __rte_unused,
+               int wait_to_complete __rte_unused)
+{
+       return 0;
+}
+
+static int
+tap_setup_queue(struct rte_eth_dev *dev,
+               struct pmd_internals *internals,
+               uint16_t qid)
+{
+       struct rx_queue *rx = &internals->rxq[qid];
+       struct tx_queue *tx = &internals->txq[qid];
+       int fd;
+
+       fd = rx->fd;
+       if (fd < 0) {
+               fd = tx->fd;
+               if (fd < 0) {
+                       RTE_LOG(INFO, PMD, "Add queue to TAP %s for qid %d\n",
+                               dev->data->name, qid);
+                       fd = tun_alloc(dev->data->name);
+                       if (fd < 0) {
+                               RTE_LOG(ERR, PMD, "tun_alloc(%s) failed\n",
+                                       dev->data->name);
+                               return -1;
+                       }
+               }
+       }
+       dev->data->rx_queues[qid] = rx;
+       dev->data->tx_queues[qid] = tx;
+
+       rx->fd = fd;
+       tx->fd = fd;
+
+       return fd;
+}
+
+static int
+tap_rx_queue_setup(struct rte_eth_dev *dev,
+                  uint16_t rx_queue_id,
+                  uint16_t nb_rx_desc __rte_unused,
+                  unsigned int socket_id __rte_unused,
+                  const struct rte_eth_rxconf *rx_conf __rte_unused,
+                  struct rte_mempool *mp)
+{
+       struct pmd_internals *internals = dev->data->dev_private;
+       uint16_t buf_size;
+       int fd;
+
+       if ((rx_queue_id >= internals->nb_queues) || !mp) {
+               RTE_LOG(ERR, PMD, "nb_queues %d mp %p\n",
+                       internals->nb_queues, mp);
+               return -1;
+       }
+
+       internals->rxq[rx_queue_id].mp = mp;
+       internals->rxq[rx_queue_id].in_port = dev->data->port_id;
+
+       /* Now get the space available for data in the mbuf */
+       buf_size = (uint16_t)(rte_pktmbuf_data_room_size(mp) -
+                               RTE_PKTMBUF_HEADROOM);
+
+       if (buf_size < ETH_FRAME_LEN) {
+               RTE_LOG(ERR, PMD,
+                       "%s: %d bytes will not fit in mbuf (%d bytes)\n",
+                       dev->data->name, ETH_FRAME_LEN, buf_size);
+               return -ENOMEM;
+       }
+
+       fd = tap_setup_queue(dev, internals, rx_queue_id);
+       if (fd == -1)
+               return -1;
+
+       internals->fds[rx_queue_id] = fd;
+       RTE_LOG(INFO, PMD, "RX TAP device name %s, qid %d on fd %d\n",
+               dev->data->name, rx_queue_id, internals->rxq[rx_queue_id].fd);
+
+       return 0;
+}
+
+static int
+tap_tx_queue_setup(struct rte_eth_dev *dev,
+                  uint16_t tx_queue_id,
+                  uint16_t nb_tx_desc __rte_unused,
+                  unsigned int socket_id __rte_unused,
+                  const struct rte_eth_txconf *tx_conf __rte_unused)
+{
+       struct pmd_internals *internals = dev->data->dev_private;
+       int ret;
+
+       if (tx_queue_id >= internals->nb_queues)
+               return -1;
+
+       ret = tap_setup_queue(dev, internals, tx_queue_id);
+       if (ret == -1)
+               return -1;
+
+       RTE_LOG(INFO, PMD, "TX TAP device name %s, qid %d on fd %d\n",
+               dev->data->name, tx_queue_id, internals->txq[tx_queue_id].fd);
+
+       return 0;
+}
+
+static const struct eth_dev_ops ops = {
+       .dev_start              = tap_dev_start,
+       .dev_stop               = tap_dev_stop,
+       .dev_close              = tap_dev_close,
+       .dev_configure          = tap_dev_configure,
+       .dev_infos_get          = tap_dev_info,
+       .rx_queue_setup         = tap_rx_queue_setup,
+       .tx_queue_setup         = tap_tx_queue_setup,
+       .rx_queue_release       = tap_rx_queue_release,
+       .tx_queue_release       = tap_tx_queue_release,
+       .link_update            = tap_link_update,
+       .stats_get              = tap_stats_get,
+       .stats_reset            = tap_stats_reset,
+};
+
+static int
+pmd_mac_address(int fd, struct rte_eth_dev *dev, struct ether_addr *addr)
+{
+       struct ifreq ifr;
+
+       if ((fd <= 0) || !dev || !addr)
+               return -1;
+
+       memset(&ifr, 0, sizeof(ifr));
+
+       if (ioctl(fd, SIOCGIFHWADDR, &ifr) == -1) {
+               RTE_LOG(ERR, PMD, "ioctl failed (SIOCGIFHWADDR) (%s)\n",
+                       ifr.ifr_name);
+               return -1;
+       }
+
+       /* Set the host based MAC address to this special MAC format */
+       ifr.ifr_hwaddr.sa_data[0] = 'T';
+       ifr.ifr_hwaddr.sa_data[1] = 'a';
+       ifr.ifr_hwaddr.sa_data[2] = 'p';
+       ifr.ifr_hwaddr.sa_data[3] = '-';
+       ifr.ifr_hwaddr.sa_data[4] = dev->data->port_id;
+       ifr.ifr_hwaddr.sa_data[5] = dev->data->numa_node;
+       if (ioctl(fd, SIOCSIFHWADDR, &ifr) == -1) {
+               RTE_LOG(ERR, PMD, "%s: ioctl failed (SIOCSIFHWADDR) (%s)\n",
+                       dev->data->name, ifr.ifr_name);
+               return -1;
+       }
+
+       /* Set the local application MAC address, needs to be different then
+        * the host based MAC address.
+        */
+       ifr.ifr_hwaddr.sa_data[0] = 'd';
+       ifr.ifr_hwaddr.sa_data[1] = 'n';
+       ifr.ifr_hwaddr.sa_data[2] = 'e';
+       ifr.ifr_hwaddr.sa_data[3] = 't';
+       ifr.ifr_hwaddr.sa_data[4] = dev->data->port_id;
+       ifr.ifr_hwaddr.sa_data[5] = dev->data->numa_node;
+       rte_memcpy(addr, ifr.ifr_hwaddr.sa_data, ETH_ALEN);
+
+       return 0;
+}
+
+static int
+eth_dev_tap_create(const char *name, char *tap_name)
+{
+       int numa_node = rte_socket_id();
+       struct rte_eth_dev *dev = NULL;
+       struct pmd_internals *pmd = NULL;
+       struct rte_eth_dev_data *data = NULL;
+       int i, fd = -1;
+
+       RTE_LOG(INFO, PMD,
+               "%s: Create TAP Ethernet device with %d queues on numa %u\n",
+                name, RTE_PMD_TAP_MAX_QUEUES, rte_socket_id());
+
+       data = rte_zmalloc_socket(tap_name, sizeof(*data), 0, numa_node);
+       if (!data) {
+               RTE_LOG(INFO, PMD, "Failed to allocate data\n");
+               goto error_exit;
+       }
+
+       pmd = rte_zmalloc_socket(tap_name, sizeof(*pmd), 0, numa_node);
+       if (!pmd) {
+               RTE_LOG(INFO, PMD, "Unable to allocate internal struct\n");
+               goto error_exit;
+       }
+
+       /* Use the name and not the tap_name */
+       dev = rte_eth_dev_allocate(tap_name);
+       if (!dev) {
+               RTE_LOG(INFO, PMD, "Unable to allocate device struct\n");
+               goto error_exit;
+       }
+
+       snprintf(pmd->name, sizeof(pmd->name), "%s", tap_name);
+
+       pmd->nb_queues = RTE_PMD_TAP_MAX_QUEUES;
+
+       /* Setup some default values */
+       data->dev_private = pmd;
+       data->port_id = dev->data->port_id;
+       data->dev_flags = RTE_ETH_DEV_DETACHABLE;
+       data->kdrv = RTE_KDRV_NONE;
+       data->drv_name = pmd_tap_drv.driver.name;
+       data->numa_node = numa_node;
+
+       data->dev_link = pmd_link;
+       data->mac_addrs = &pmd->eth_addr;
+       data->nb_rx_queues = pmd->nb_queues;
+       data->nb_tx_queues = pmd->nb_queues;
+
+       dev->data = data;
+       dev->dev_ops = &ops;
+       dev->driver = NULL;
+       dev->rx_pkt_burst = pmd_rx_burst;
+       dev->tx_pkt_burst = pmd_tx_burst;
+       snprintf(dev->data->name, sizeof(dev->data->name), "%s", name);
+
+       /* Create the first Tap device */
+       fd = tun_alloc(tap_name);
+       if (fd < 0) {
+               RTE_LOG(INFO, PMD, "tun_alloc() failed\n");
+               goto error_exit;
+       }
+
+       /* Presetup the fds to -1 as being not working */
+       for (i = 0; i < RTE_PMD_TAP_MAX_QUEUES; i++) {
+               pmd->fds[i] = -1;
+               pmd->rxq[i].fd = -1;
+               pmd->txq[i].fd = -1;
+       }
+
+       /* Take the TUN/TAP fd and place in the first location */
+       pmd->rxq[0].fd = fd;
+       pmd->txq[0].fd = fd;
+       pmd->fds[0] = fd;
+
+       if (pmd_mac_address(fd, dev, &pmd->eth_addr) < 0) {
+               RTE_LOG(INFO, PMD, "Unable to get MAC address\n");
+               goto error_exit;
+       }
+
+       return 0;
+
+error_exit:
+       RTE_PMD_DEBUG_TRACE("Unable to initialize %s\n", name);
+
+       rte_free(data);
+       rte_free(pmd);
+
+       rte_eth_dev_release_port(dev);
+
+       return -EINVAL;
+}
+
+static int
+set_interface_name(const char *key __rte_unused,
+                  const char *value,
+                  void *extra_args)
+{
+       char *name = (char *)extra_args;
+
+       if (value)
+               snprintf(name, RTE_ETH_NAME_MAX_LEN - 1, "%s", value);
+       else
+               snprintf(name, RTE_ETH_NAME_MAX_LEN - 1, "%s%d",
+                        DEFAULT_TAP_NAME, (tap_unit - 1));
+
+       return 0;
+}
+
+static int
+set_interface_speed(const char *key __rte_unused,
+                   const char *value,
+                   void *extra_args)
+{
+       *(int *)extra_args = (value) ? atoi(value) : ETH_SPEED_NUM_10G;
+
+       return 0;
+}
+
+/* Open a TAP interface device.
+ */
+static int
+rte_pmd_tap_probe(const char *name, const char *params)
+{
+       int ret;
+       struct rte_kvargs *kvlist = NULL;
+       int speed;
+       char tap_name[RTE_ETH_NAME_MAX_LEN];
+
+       speed = ETH_SPEED_NUM_10G;
+       snprintf(tap_name, sizeof(tap_name), "%s%d",
+                DEFAULT_TAP_NAME, tap_unit++);
+
+       RTE_LOG(INFO, PMD, "Initializing pmd_tap for %s as %s\n",
+               name, tap_name);
+
+       if (params && (params[0] != '\0')) {
+               RTE_LOG(INFO, PMD, "paramaters (%s)\n", params);
+
+               kvlist = rte_kvargs_parse(params, valid_arguments);
+               if (kvlist) {
+                       if (rte_kvargs_count(kvlist, ETH_TAP_SPEED_ARG) == 1) {
+                               ret = rte_kvargs_process(kvlist,
+                                                        ETH_TAP_SPEED_ARG,
+                                                        &set_interface_speed,
+                                                        &speed);
+                               if (ret == -1)
+                                       goto leave;
+                       }
+
+                       if (rte_kvargs_count(kvlist, ETH_TAP_IFACE_ARG) == 1) {
+                               ret = rte_kvargs_process(kvlist,
+                                                        ETH_TAP_IFACE_ARG,
+                                                        &set_interface_name,
+                                                        tap_name);
+                               if (ret == -1)
+                                       goto leave;
+                       }
+               }
+       }
+       pmd_link.link_speed = speed;
+
+       ret = eth_dev_tap_create(name, tap_name);
+
+leave:
+       if (ret == -1) {
+               RTE_LOG(INFO, PMD, "Failed to create pmd for %s as %s\n",
+                       name, tap_name);
+               tap_unit--;             /* Restore the unit number */
+       }
+       rte_kvargs_free(kvlist);
+
+       return ret;
+}
+
+/* detach a TAP device.
+ */
+static int
+rte_pmd_tap_remove(const char *name)
+{
+       struct rte_eth_dev *eth_dev = NULL;
+       struct pmd_internals *internals;
+       int i;
+
+       RTE_LOG(INFO, PMD, "Closing TUN/TAP Ethernet device on numa %u\n",
+               rte_socket_id());
+
+       /* find the ethdev entry */
+       eth_dev = rte_eth_dev_allocated(name);
+       if (!eth_dev)
+               return 0;
+
+       internals = eth_dev->data->dev_private;
+       for (i = 0; i < internals->nb_queues; i++)
+               if (internals->fds[i] != -1)
+                       close(internals->fds[i]);
+
+       rte_free(eth_dev->data->dev_private);
+       rte_free(eth_dev->data);
+
+       rte_eth_dev_release_port(eth_dev);
+
+       return 0;
+}
+
+static struct rte_vdev_driver pmd_tap_drv = {
+       .probe = rte_pmd_tap_probe,
+       .remove = rte_pmd_tap_remove,
+};
+RTE_PMD_REGISTER_VDEV(net_tap, pmd_tap_drv);
+RTE_PMD_REGISTER_ALIAS(net_tap, eth_tap);
+RTE_PMD_REGISTER_PARAM_STRING(net_tap, "iface=<string>,speed=N");
drivers/net/tap/rte_pmd_tap_version.map new file +4 -0
new file mode 100644
index 0000000..31eca32
@@ -0,0 +1,4 @@
+DPDK_17.02 {
+
+       local: *;
+};
mk/rte.app.mk +1 -1
index f75f0e2..02c32ae 100644
@@ -124,6 +124,7 @@ _LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_PCAP)       += -lrte_pmd_pcap -lpcap
 _LDLIBS-$(CONFIG_RTE_LIBRTE_QEDE_PMD)       += -lrte_pmd_qede
 _LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_RING)       += -lrte_pmd_ring
 _LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_SZEDATA2)   += -lrte_pmd_szedata2 -lsze2
+_LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_TAP)        += -lrte_pmd_tap
 _LDLIBS-$(CONFIG_RTE_LIBRTE_THUNDERX_NICVF_PMD) += -lrte_pmd_thunderx_nicvf -lm
 _LDLIBS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD)     += -lrte_pmd_virtio
 ifeq ($(CONFIG_RTE_LIBRTE_VHOST),y)
--
Quoted 3 lines

2.8.0.GIT

Regards,
Keith

Reviewed-by Tested-by Aws Ismail Tested-by Vasily Philipov on v12 1/1 Applied
#39 lore raw

Re: [PATCH v12] net/tap: new TUN/TAP device PMD

In reply to Keith Wiles
Quoted 20 lines · Keith Wiles

The PMD allows for DPDK and the host to communicate using a raw device interface on the host and in the DPDK application. The device created is a Tap device with a L2 packet header.

v12- Fixup minor changes for driver_name and version number v11- Add the tap.rst to the nic/index.rst file
v10- Change the string name used to allow for multiple devices. v9 - Fix up the docs to use correct syntax
v8 - Fix issue with tap_tx_queue_setup() not return zero on success. v7 - Reword the comment in common_base and fix the data->name issue v6 - fixed the checkpatch issues
v5 - merge in changes from list review see related emails
fixed many minor edits
v4 - merge with latest driver changes
v3 - fix includes by removing ifdef for other type besides Linux
Fix the copyright notice in the Makefile
v2 - merge all of the patches into one patch
Fix a typo on naming the tap device
Update the maintainers list

Quoted 2 lines

---

Applied to dpdk-next-net/master, thanks.

#40 lore raw

Re: [PATCH v12] net/tap: new TUN/TAP device PMD

Quoted 21 lines · Ferruh Yigit
Quoted 20 lines · Keith Wiles

The PMD allows for DPDK and the host to communicate using a raw device interface on the host and in the DPDK application. The device created is a Tap device with a L2 packet header.

v12- Fixup minor changes for driver_name and version number v11- Add the tap.rst to the nic/index.rst file
v10- Change the string name used to allow for multiple devices. v9 - Fix up the docs to use correct syntax
v8 - Fix issue with tap_tx_queue_setup() not return zero on success. v7 - Reword the comment in common_base and fix the data->name issue v6 - fixed the checkpatch issues
v5 - merge in changes from list review see related emails
fixed many minor edits
v4 - merge with latest driver changes
v3 - fix includes by removing ifdef for other type besides Linux
Fix the copyright notice in the Makefile
v2 - merge all of the patches into one patch
Fix a typo on naming the tap device
Update the maintainers list

Quoted 9 lines
Quoted 2 lines

---

Applied to dpdk-next-net/master, thanks.

Hi Keith,

test-report [1] shows build error [2] for the PMD, I guess there is a kernel version dependency, FYI.

Issue has been caught after integration.

[1]
http://dpdk.org/ml/archives/test-report/2017-January/009616.html

[2]
DPDK/drivers/net/tap/rte_eth_tap.c: In function ‘tun_alloc’:
DPDK/drivers/net/tap/rte_eth_tap.c:143:19: error: ‘IFF_MULTI_QUEUE’ undeclared (first use in this function)
compilation terminated due to -Wfatal-errors.

#41 lore raw

PATCH 1/1 net/tap: fix IFF_MULTI_QUEUE in older kernels not found

---
 drivers/net/tap/rte_eth_tap.c | 10 ++++++++++
 1 file changed, 10 insertions(+)
drivers/net/tap/rte_eth_tap.c +10 -0
index d5e2fc3..5cf9831 100644
@@ -139,6 +139,7 @@ tun_alloc(char *name)
 	}
 	RTE_LOG(DEBUG, PMD, "TUN/TAP Features %08x\n", features);
 
+#ifdef IFF_MULTI_QUEUE
 	if (!(features & IFF_MULTI_QUEUE) && (RTE_PMD_TAP_MAX_QUEUES > 1)) {
 		RTE_LOG(DEBUG, PMD, "TUN/TAP device only one queue\n");
 		goto error;
@@ -151,6 +152,15 @@ tun_alloc(char *name)
 		RTE_LOG(DEBUG, PMD, "Multi-queue support for %d queues\n",
 			RTE_PMD_TAP_MAX_QUEUES);
 	}
+#else
+    if (RTE_PMD_TAP_MAX_QUEUES > 1) {
+        RTE_LOG(DEBUG, PMD, "TUN/TAP device only one queue\n");
+        goto error;
+    } else {
+        ifr.ifr_flags |= IFF_ONE_QUEUE;
+        RTE_LOG(DEBUG, PMD, "Single queue only support\n");
+    }
+#endif
 
 	/* Set the TUN/TAP configuration and get the name if needed */
 	if (ioctl(fd, TUNSETIFF, (void *)&ifr) < 0) {
Signature · 1 line
2.8.0.GIT
#42 lore raw

PATCH 1/1 net/tap: fix IFF_MULTI_QUEUE in older kernels not found

In reply to Ferruh Yigit
---
 drivers/net/tap/rte_eth_tap.c | 10 ++++++++++
 1 file changed, 10 insertions(+)
drivers/net/tap/rte_eth_tap.c +10 -0
index d5e2fc3..cb96e99 100644
@@ -139,6 +139,7 @@ tun_alloc(char *name)
 	}
 	RTE_LOG(DEBUG, PMD, "TUN/TAP Features %08x\n", features);
 
+#ifdef IFF_MULTI_QUEUE
 	if (!(features & IFF_MULTI_QUEUE) && (RTE_PMD_TAP_MAX_QUEUES > 1)) {
 		RTE_LOG(DEBUG, PMD, "TUN/TAP device only one queue\n");
 		goto error;
@@ -151,6 +152,15 @@ tun_alloc(char *name)
 		RTE_LOG(DEBUG, PMD, "Multi-queue support for %d queues\n",
 			RTE_PMD_TAP_MAX_QUEUES);
 	}
+#else
+	if (RTE_PMD_TAP_MAX_QUEUES > 1) {
+		RTE_LOG(DEBUG, PMD, "TUN/TAP device only one queue\n");
+		goto error;
+	} else {
+		ifr.ifr_flags |= IFF_ONE_QUEUE;
+		RTE_LOG(DEBUG, PMD, "Single queue only support\n");
+	}
+#endif
 
 	/* Set the TUN/TAP configuration and get the name if needed */
 	if (ioctl(fd, TUNSETIFF, (void *)&ifr) < 0) {
Signature · 1 line
2.8.0.GIT
#44 lore raw

Re: [PATCH] net/tap: fix IFF_MULTI_QUEUE in older kernels not found

2017-01-20 16:11, Ferruh Yigit:

Quoted 9 lines · Keith Wiles
Quoted 1 line

Hi Thomas,

Although this is a driver patch, would you mind getting it to the main tree to fix build error asap (instead of waiting next integration)?

Applied, thanks

#45 lore raw

PATCH 3/3 drivers/net:build support for new tap device driver

In reply to Keith Wiles
---
 config/common_linuxapp | 3 +++
 drivers/net/Makefile   | 1 +
 mk/rte.app.mk          | 1 +
 3 files changed, 5 insertions(+)
config/common_linuxapp +3 -0
index 2483dfa..704c01c 100644
@@ -44,3 +44,6 @@ CONFIG_RTE_LIBRTE_PMD_VHOST=y
 CONFIG_RTE_LIBRTE_PMD_AF_PACKET=y
 CONFIG_RTE_LIBRTE_POWER=y
 CONFIG_RTE_VIRTIO_USER=y
+CONFIG_RTE_LIBRTE_PMD_TAP=y
+CONFIG_RTE_PMD_TAP_MAX_QUEUES=32
+
drivers/net/Makefile +1 -0
index bc93230..b4afa98 100644
@@ -55,6 +55,7 @@ DIRS-$(CONFIG_RTE_LIBRTE_THUNDERX_NICVF_PMD) += thunderx
 DIRS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD) += virtio
 DIRS-$(CONFIG_RTE_LIBRTE_VMXNET3_PMD) += vmxnet3
 DIRS-$(CONFIG_RTE_LIBRTE_PMD_XENVIRT) += xenvirt
+DIRS-$(CONFIG_RTE_LIBRTE_PMD_TAP) += tap
 
 ifeq ($(CONFIG_RTE_LIBRTE_VHOST),y)
 DIRS-$(CONFIG_RTE_LIBRTE_PMD_VHOST) += vhost
mk/rte.app.mk +1 -0
index 1a0095b..bd1d10f 100644
@@ -129,6 +129,7 @@ ifeq ($(CONFIG_RTE_LIBRTE_VHOST),y)
 _LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_VHOST)      += -lrte_pmd_vhost
 endif # $(CONFIG_RTE_LIBRTE_VHOST)
 _LDLIBS-$(CONFIG_RTE_LIBRTE_VMXNET3_PMD)    += -lrte_pmd_vmxnet3_uio
+_LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_TAP)        += -lrte_pmd_tap
 
 ifeq ($(CONFIG_RTE_LIBRTE_CRYPTODEV),y)
 _LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_AESNI_MB)   += -lrte_pmd_aesni_mb
Signature · 1 line
2.8.0.GIT
#46 lore raw

Re: [PATCH 3/3] drivers/net:build support for new tap device driver

Quoted 36 lines · Keith Wiles
---
 config/common_linuxapp | 3 +++
 drivers/net/Makefile   | 1 +
 mk/rte.app.mk          | 1 +
 3 files changed, 5 insertions(+)
config/common_linuxapp +3 -0
index 2483dfa..704c01c 100644
@@ -44,3 +44,6 @@ CONFIG_RTE_LIBRTE_PMD_VHOST=y
 CONFIG_RTE_LIBRTE_PMD_AF_PACKET=y
 CONFIG_RTE_LIBRTE_POWER=y
 CONFIG_RTE_VIRTIO_USER=y
+CONFIG_RTE_LIBRTE_PMD_TAP=y
+CONFIG_RTE_PMD_TAP_MAX_QUEUES=32
+
drivers/net/Makefile +1 -0
index bc93230..b4afa98 100644
@@ -55,6 +55,7 @@ DIRS-$(CONFIG_RTE_LIBRTE_THUNDERX_NICVF_PMD) += thunderx
 DIRS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD) += virtio
 DIRS-$(CONFIG_RTE_LIBRTE_VMXNET3_PMD) += vmxnet3
 DIRS-$(CONFIG_RTE_LIBRTE_PMD_XENVIRT) += xenvirt
+DIRS-$(CONFIG_RTE_LIBRTE_PMD_TAP) += tap
 
 ifeq ($(CONFIG_RTE_LIBRTE_VHOST),y)
 DIRS-$(CONFIG_RTE_LIBRTE_PMD_VHOST) += vhost
mk/rte.app.mk +0 -0
index 1a0095b..bd1d10f 100644

mk/rte.app.mk ifeq ($(CONFIG_RTE_LIBRTE_VHOST),y) · line 135

Quoted 7 lines
mk/rte.app.mk +1 -0
@@ -129,6 +129,7 @@ ifeq ($(CONFIG_RTE_LIBRTE_VHOST),y)
 _LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_VHOST)      += -lrte_pmd_vhost
 endif # $(CONFIG_RTE_LIBRTE_VHOST)
 _LDLIBS-$(CONFIG_RTE_LIBRTE_VMXNET3_PMD)    += -lrte_pmd_vmxnet3_uio
+_LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_TAP)        += -lrte_pmd_tap
 
 ifeq ($(CONFIG_RTE_LIBRTE_CRYPTODEV),y)
 _LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_AESNI_MB)   += -lrte_pmd_aesni_mb

Splitting the Makefile and config changes into a separate patch makes no sense at all in this case. Just do it in the patch introducing the driver.

And actually, ditto for documentation.

- Panu -

#47 lore raw

Re: [PATCH 3/3] drivers/net:build support for new tap device driver

Regards,
Keith

Quoted 43 lines · Panu Matilainen
Quoted 40 lines · Keith Wiles

---

config/common_linuxapp | 3 +++
drivers/net/Makefile   | 1 +
mk/rte.app.mk          | 1 +
3 files changed, 5 insertions(+)
config/common_linuxapp +0 -0
index 2483dfa..704c01c 100644
@@ -44,3 +44,6 @@ CONFIG_RTE_LIBRTE_PMD_VHOST=y

CONFIG_RTE_LIBRTE_PMD_AF_PACKET=y
CONFIG_RTE_LIBRTE_POWER=y
CONFIG_RTE_VIRTIO_USER=y

config/common_linuxapp +3 -0
+CONFIG_RTE_LIBRTE_PMD_TAP=y
+CONFIG_RTE_PMD_TAP_MAX_QUEUES=32
+
drivers/net/Makefile +0 -0
index bc93230..b4afa98 100644
@@ -55,6 +55,7 @@ DIRS-$(CONFIG_RTE_LIBRTE_THUNDERX_NICVF_PMD) += thunderx

DIRS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD) += virtio
DIRS-$(CONFIG_RTE_LIBRTE_VMXNET3_PMD) += vmxnet3
DIRS-$(CONFIG_RTE_LIBRTE_PMD_XENVIRT) += xenvirt
+DIRS-$(CONFIG_RTE_LIBRTE_PMD_TAP) += tap

ifeq ($(CONFIG_RTE_LIBRTE_VHOST),y)
DIRS-$(CONFIG_RTE_LIBRTE_PMD_VHOST) += vhost

mk/rte.app.mk +0 -0
index 1a0095b..bd1d10f 100644
@@ -129,6 +129,7 @@ ifeq ($(CONFIG_RTE_LIBRTE_VHOST),y)

_LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_VHOST) += -lrte_pmd_vhost
endif # $(CONFIG_RTE_LIBRTE_VHOST)
_LDLIBS-$(CONFIG_RTE_LIBRTE_VMXNET3_PMD) += -lrte_pmd_vmxnet3_uio
+_LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_TAP) += -lrte_pmd_tap

Quoted 7 lines
Quoted 3 lines

ifeq ($(CONFIG_RTE_LIBRTE_CRYPTODEV),y)
_LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_AESNI_MB) += -lrte_pmd_aesni_mb

Splitting the Makefile and config changes into a separate patch makes no sense at all in this case. Just do it in the patch introducing the driver.

And actually, ditto for documentation.

OK, will do, but I thought other patches followed something like this as I had thought I was following the normally process. Without some place defining the real process it is hard to decide the real way to send patches. :-(

I will send a new single patch later today with one little typo I fixed.

Quoted 2 lines

- Panu -

#48 lore raw

PATCH v2 1/1 drivers/net:new PMD using tun/tap host interface

In reply to Keith Wiles

The rte_eth_tap.c PMD creates a device using TUN/TAP interfaces on the local host. The PMD allows for DPDK and the host to communicate using a raw device interface on the host and in the DPDK application. The device created is a Tap device with a L2 packet header.

v2 - merge all of the patches into one patch.
Fix a typo on naming the tap device.
Update the maintainers list

---
 MAINTAINERS                             |   5 +
 config/common_linuxapp                  |   2 +
 doc/guides/nics/tap.rst                 |  84 +++
 drivers/net/Makefile                    |   1 +
 drivers/net/tap/Makefile                |  60 +++
 drivers/net/tap/rte_eth_tap.c           | 872 ++++++++++++++++++++++++++++++++
 drivers/net/tap/rte_pmd_tap_version.map |   4 +
 mk/rte.app.mk                           |   1 +
 8 files changed, 1029 insertions(+)
 create mode 100644 doc/guides/nics/tap.rst
 create mode 100644 drivers/net/tap/Makefile
 create mode 100644 drivers/net/tap/rte_eth_tap.c
 create mode 100644 drivers/net/tap/rte_pmd_tap_version.map
MAINTAINERS +5 -0
index 7c33ad4..fad74e4 100644
@@ -392,6 +392,11 @@ F: doc/guides/nics/pcap_ring.rst
 F: app/test/test_pmd_ring.c
 F: app/test/test_pmd_ring_perf.c
 
+Tap PMD
+M: Keith Wiles <[email protected]>
+F: drivers/net/tap
+F: doc/guides/nics/tap.rst
+
 Null Networking PMD
 M: Tetsuya Mukawa <[email protected]>
 F: drivers/net/null/
config/common_linuxapp +2 -0
index 2483dfa..59a2053 100644
@@ -44,3 +44,5 @@ CONFIG_RTE_LIBRTE_PMD_VHOST=y
 CONFIG_RTE_LIBRTE_PMD_AF_PACKET=y
 CONFIG_RTE_LIBRTE_POWER=y
 CONFIG_RTE_VIRTIO_USER=y
+CONFIG_RTE_LIBRTE_PMD_TAP=y
+CONFIG_RTE_PMD_TAP_MAX_QUEUES=32
doc/guides/nics/tap.rst new file +84 -0

84 lines not on this page · raw message

drivers/net/Makefile +1 -0
index bc93230..b4afa98 100644
@@ -55,6 +55,7 @@ DIRS-$(CONFIG_RTE_LIBRTE_THUNDERX_NICVF_PMD) += thunderx
 DIRS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD) += virtio
 DIRS-$(CONFIG_RTE_LIBRTE_VMXNET3_PMD) += vmxnet3
 DIRS-$(CONFIG_RTE_LIBRTE_PMD_XENVIRT) += xenvirt
+DIRS-$(CONFIG_RTE_LIBRTE_PMD_TAP) += tap
 
 ifeq ($(CONFIG_RTE_LIBRTE_VHOST),y)
 DIRS-$(CONFIG_RTE_LIBRTE_PMD_VHOST) += vhost
drivers/net/tap/Makefile new file +60 -0

60 lines not on this page · raw message

drivers/net/tap/rte_eth_tap.c new file +872 -0

872 lines not on this page · raw message

drivers/net/tap/rte_pmd_tap_version.map new file +4 -0
new file mode 100644
index 0000000..61463bf
@@ -0,0 +1,4 @@
+DPDK_16.11 {
+
+	local: *;
+};
mk/rte.app.mk +1 -0
index 1a0095b..bd1d10f 100644
@@ -129,6 +129,7 @@ ifeq ($(CONFIG_RTE_LIBRTE_VHOST),y)
 _LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_VHOST)      += -lrte_pmd_vhost
 endif # $(CONFIG_RTE_LIBRTE_VHOST)
 _LDLIBS-$(CONFIG_RTE_LIBRTE_VMXNET3_PMD)    += -lrte_pmd_vmxnet3_uio
+_LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_TAP)        += -lrte_pmd_tap
 
 ifeq ($(CONFIG_RTE_LIBRTE_CRYPTODEV),y)
 _LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_AESNI_MB)   += -lrte_pmd_aesni_mb
Signature · 1 line
2.8.0.GIT
#49 lore raw
Quoted 11 lines · Keith Wiles

The rte_eth_tap.c PMD creates a device using TUN/TAP interfaces on the local host. The PMD allows for DPDK and the host to communicate using a raw device interface on the host and in the DPDK application. The device created is a Tap device with a L2 packet header.

v2 - merge all of the patches into one patch.
Fix a typo on naming the tap device.
Update the maintainers list

Hi,

FYI, my robot caught some build errors with this patch applied.

	--yliu

---
In file included from /yeti/vm/ubuntu-initrd-16.04-x86_64-build/dpdk/drivers/net/tap/rte_eth_tap.c:50:
/usr/include/linux/if.h:71:2: error: redefinition of enumerator 'IFF_UP'
IFF_UP = 1<<0, /* sysfs */
^
/usr/include/net/if.h:45:17: note: expanded from macro 'IFF_UP'

# define IFF_UP IFF_UP
                ^
/usr/include/net/if.h:44:5: note: previous definition is here
    IFF_UP = 0x1,               /* Interface is up.  */
    ^
In file included from /yeti/vm/ubuntu-initrd-16.04-x86_64-build/dpdk/drivers/net/tap/rte_eth_tap.c:50:
/usr/include/linux/if.h:72:2: error: redefinition of enumerator 'IFF_BROADCAST'
        IFF_BROADCAST                   = 1<<1,  /* __volatile__ */
        ^
/usr/include/net/if.h:47:24: note: expanded from macro 'IFF_BROADCAST'
# define IFF_BROADCAST  IFF_BROADCAST
                        ^
/usr/include/net/if.h:46:5: note: previous definition is here
    IFF_BROADCAST = 0x2,        /* Broadcast address valid.  */
    ^
In file included from /yeti/vm/ubuntu-initrd-16.04-x86_64-build/dpdk/drivers/net/tap/rte_eth_tap.c:50:
/usr/include/linux/if.h:73:2: error: redefinition of enumerator 'IFF_DEBUG'
        IFF_DEBUG                       = 1<<2,  /* sysfs */
        ^
/usr/include/net/if.h:49:20: note: expanded from macro 'IFF_DEBUG'
# define IFF_DEBUG      IFF_DEBUG
                        ^
/usr/include/net/if.h:48:5: note: previous definition is here
    IFF_DEBUG = 0x4,            /* Turn on debugging.  */
    ^
In file included from /yeti/vm/ubuntu-initrd-16.04-x86_64-build/dpdk/drivers/net/tap/rte_eth_tap.c:50:
/usr/include/linux/if.h:74:2: error: redefinition of enumerator 'IFF_LOOPBACK'
        IFF_LOOPBACK                    = 1<<3,  /* __volatile__ */
        ^
/usr/include/net/if.h:51:23: note: expanded from macro 'IFF_LOOPBACK'
# define IFF_LOOPBACK   IFF_LOOPBACK
                        ^
/usr/include/net/if.h:50:5: note: previous definition is here
    IFF_LOOPBACK = 0x8,         /* Is a loopback net.  */
    ^
In file included from /yeti/vm/ubuntu-initrd-16.04-x86_64-build/dpdk/drivers/net/tap/rte_eth_tap.c:50:
/usr/include/linux/if.h:75:2: error: redefinition of enumerator 'IFF_POINTOPOINT'
        IFF_POINTOPOINT                 = 1<<4,  /* __volatile__ */
        ^
/usr/include/net/if.h:53:26: note: expanded from macro 'IFF_POINTOPOINT'
# define IFF_POINTOPOINT IFF_POINTOPOINT
                         ^
/usr/include/net/if.h:52:5: note: previous definition is here
    IFF_POINTOPOINT = 0x10,     /* Interface is point-to-point link.  */
    ^
In file included from /yeti/vm/ubuntu-initrd-16.04-x86_64-build/dpdk/drivers/net/tap/rte_eth_tap.c:50:
/usr/include/linux/if.h:76:2: error: redefinition of enumerator 'IFF_NOTRAILERS'
        IFF_NOTRAILERS                  = 1<<5,  /* sysfs */
        ^
/usr/include/net/if.h:55:25: note: expanded from macro 'IFF_NOTRAILERS'
# define IFF_NOTRAILERS IFF_NOTRAILERS
                        ^
/usr/include/net/if.h:54:5: note: previous definition is here
    IFF_NOTRAILERS = 0x20,      /* Avoid use of trailers.  */
    ^
In file included from /yeti/vm/ubuntu-initrd-16.04-x86_64-build/dpdk/drivers/net/tap/rte_eth_tap.c:50:
/usr/include/linux/if.h:77:2: error: redefinition of enumerator 'IFF_RUNNING'
        IFF_RUNNING                     = 1<<6,  /* __volatile__ */
        ^
/usr/include/net/if.h:57:22: note: expanded from macro 'IFF_RUNNING'
# define IFF_RUNNING    IFF_RUNNING
                        ^
/usr/include/net/if.h:56:5: note: previous definition is here
    IFF_RUNNING = 0x40,         /* Resources allocated.  */
    ^
In file included from /yeti/vm/ubuntu-initrd-16.04-x86_64-build/dpdk/drivers/net/tap/rte_eth_tap.c:50:
/usr/include/linux/if.h:78:2: error: redefinition of enumerator 'IFF_NOARP'
        IFF_NOARP                       = 1<<7,  /* sysfs */
        ^
/usr/include/net/if.h:59:20: note: expanded from macro 'IFF_NOARP'
# define IFF_NOARP      IFF_NOARP
                        ^
/usr/include/net/if.h:58:5: note: previous definition is here
    IFF_NOARP = 0x80,           /* No address resolution protocol.  */
    ^
In file included from /yeti/vm/ubuntu-initrd-16.04-x86_64-build/dpdk/drivers/net/tap/rte_eth_tap.c:50:
/usr/include/linux/if.h:79:2: error: redefinition of enumerator 'IFF_PROMISC'
        IFF_PROMISC                     = 1<<8,  /* sysfs */
        ^
/usr/include/net/if.h:61:22: note: expanded from macro 'IFF_PROMISC'
# define IFF_PROMISC    IFF_PROMISC
                        ^
/usr/include/net/if.h:60:5: note: previous definition is here
    IFF_PROMISC = 0x100,        /* Receive all packets.  */
    ^
In file included from /yeti/vm/ubuntu-initrd-16.04-x86_64-build/dpdk/drivers/net/tap/rte_eth_tap.c:50:
/usr/include/linux/if.h:80:2: error: redefinition of enumerator 'IFF_ALLMULTI'
        IFF_ALLMULTI                    = 1<<9,  /* sysfs */
        ^
/usr/include/net/if.h:65:23: note: expanded from macro 'IFF_ALLMULTI'
# define IFF_ALLMULTI   IFF_ALLMULTI
                        ^
/usr/include/net/if.h:64:5: note: previous definition is here
    IFF_ALLMULTI = 0x200,       /* Receive all multicast packets.  */
    ^
In file included from /yeti/vm/ubuntu-initrd-16.04-x86_64-build/dpdk/drivers/net/tap/rte_eth_tap.c:50:
/usr/include/linux/if.h:81:2: error: redefinition of enumerator 'IFF_MASTER'
        IFF_MASTER                      = 1<<10, /* __volatile__ */
        ^
/usr/include/net/if.h:68:21: note: expanded from macro 'IFF_MASTER'
# define IFF_MASTER     IFF_MASTER
                        ^
/usr/include/net/if.h:67:5: note: previous definition is here
    IFF_MASTER = 0x400,         /* Master of a load balancer.  */
    ^
In file included from /yeti/vm/ubuntu-initrd-16.04-x86_64-build/dpdk/drivers/net/tap/rte_eth_tap.c:50:
/usr/include/linux/if.h:82:2: error: redefinition of enumerator 'IFF_SLAVE'
        IFF_SLAVE                       = 1<<11, /* __volatile__ */
        ^
/usr/include/net/if.h:70:20: note: expanded from macro 'IFF_SLAVE'
# define IFF_SLAVE      IFF_SLAVE
                        ^
/usr/include/net/if.h:69:5: note: previous definition is here
    IFF_SLAVE = 0x800,          /* Slave of a load balancer.  */
    ^
In file included from /yeti/vm/ubuntu-initrd-16.04-x86_64-build/dpdk/drivers/net/tap/rte_eth_tap.c:50:
/usr/include/linux/if.h:83:2: error: redefinition of enumerator 'IFF_MULTICAST'
        IFF_MULTICAST                   = 1<<12, /* sysfs */
        ^
/usr/include/net/if.h:73:24: note: expanded from macro 'IFF_MULTICAST'
# define IFF_MULTICAST  IFF_MULTICAST
                        ^
/usr/include/net/if.h:72:5: note: previous definition is here
    IFF_MULTICAST = 0x1000,     /* Supports multicast.  */
    ^
In file included from /yeti/vm/ubuntu-initrd-16.04-x86_64-build/dpdk/drivers/net/tap/rte_eth_tap.c:50:
/usr/include/linux/if.h:84:2: error: redefinition of enumerator 'IFF_PORTSEL'
        IFF_PORTSEL                     = 1<<13, /* sysfs */
        ^
/usr/include/net/if.h:76:22: note: expanded from macro 'IFF_PORTSEL'
# define IFF_PORTSEL    IFF_PORTSEL
                        ^
/usr/include/net/if.h:75:5: note: previous definition is here
    IFF_PORTSEL = 0x2000,       /* Can set media type.  */
    ^
In file included from /yeti/vm/ubuntu-initrd-16.04-x86_64-build/dpdk/drivers/net/tap/rte_eth_tap.c:50:
/usr/include/linux/if.h:85:2: error: redefinition of enumerator 'IFF_AUTOMEDIA'
        IFF_AUTOMEDIA                   = 1<<14, /* sysfs */
        ^
/usr/include/net/if.h:78:24: note: expanded from macro 'IFF_AUTOMEDIA'
# define IFF_AUTOMEDIA  IFF_AUTOMEDIA
                        ^
/usr/include/net/if.h:77:5: note: previous definition is here
    IFF_AUTOMEDIA = 0x4000,     /* Auto media select active.  */
    ^
In file included from /yeti/vm/ubuntu-initrd-16.04-x86_64-build/dpdk/drivers/net/tap/rte_eth_tap.c:50:
/usr/include/linux/if.h:86:2: error: redefinition of enumerator 'IFF_DYNAMIC'
        IFF_DYNAMIC                     = 1<<15, /* sysfs */
        ^
/usr/include/net/if.h:80:22: note: expanded from macro 'IFF_DYNAMIC'
# define IFF_DYNAMIC    IFF_DYNAMIC
                        ^
/usr/include/net/if.h:79:5: note: previous definition is here
    IFF_DYNAMIC = 0x8000        /* Dialup device with changing addresses.  */
    ^
In file included from /yeti/vm/ubuntu-initrd-16.04-x86_64-build/dpdk/drivers/net/tap/rte_eth_tap.c:50:
/usr/include/linux/if.h:169:8: error: redefinition of 'ifmap'
struct ifmap {
       ^
/usr/include/net/if.h:111:8: note: previous definition is here
struct ifmap
       ^
In file included from /yeti/vm/ubuntu-initrd-16.04-x86_64-build/dpdk/drivers/net/tap/rte_eth_tap.c:50:
/usr/include/linux/if.h:203:8: error: redefinition of 'ifreq'
struct ifreq {
       ^
/usr/include/net/if.h:126:8: note: previous definition is here
struct ifreq
       ^
In file included from /yeti/vm/ubuntu-initrd-16.04-x86_64-build/dpdk/drivers/net/tap/rte_eth_tap.c:50:
/usr/include/linux/if.h:252:8: error: redefinition of 'ifconf'
struct ifconf  {
       ^
/usr/include/net/if.h:176:8: note: previous definition is here
struct ifconf
       ^
19 errors generated.
make[6]: *** [rte_eth_tap.o] Error 1
make[5]: *** [tap] Error 2
make[5]: *** Waiting for unfinished jobs....
make[4]: *** [net] Error 2
make[3]: *** [drivers] Error 2
make[2]: *** [all] Error 2
make[1]: *** [pre_install] Error 2
make: *** [install] Error 2
error: build failed
#50 lore raw

Regards,
Keith

Quoted 13 lines · Yuanhan Liu
Quoted 10 lines · Keith Wiles

The rte_eth_tap.c PMD creates a device using TUN/TAP interfaces on the local host. The PMD allows for DPDK and the host to communicate using a raw device interface on the host and in the DPDK application. The device created is a Tap device with a L2 packet header.

v2 - merge all of the patches into one patch.
Fix a typo on naming the tap device.
Update the maintainers list

Quoted 5 lines
Quoted 1 line

Hi,

FYI, my robot caught some build errors with this patch applied.

All of the below errors are from Linux header files and not the Tap driver. How are you building DPDK, which config file are you using, OS version, .... Some details would be nice to understand how these errors are being generated and how I can recreate them.

Quoted 195 lines
	--yliu

---
In file included from /yeti/vm/ubuntu-initrd-16.04-x86_64-build/dpdk/drivers/net/tap/rte_eth_tap.c:50:
/usr/include/linux/if.h:71:2: error: redefinition of enumerator 'IFF_UP'
IFF_UP = 1<<0, /* sysfs */
^
/usr/include/net/if.h:45:17: note: expanded from macro 'IFF_UP'

# define IFF_UP IFF_UP
               ^
/usr/include/net/if.h:44:5: note: previous definition is here
   IFF_UP = 0x1,               /* Interface is up.  */
   ^
In file included from /yeti/vm/ubuntu-initrd-16.04-x86_64-build/dpdk/drivers/net/tap/rte_eth_tap.c:50:
/usr/include/linux/if.h:72:2: error: redefinition of enumerator 'IFF_BROADCAST'
       IFF_BROADCAST                   = 1<<1,  /* __volatile__ */
       ^
/usr/include/net/if.h:47:24: note: expanded from macro 'IFF_BROADCAST'
# define IFF_BROADCAST  IFF_BROADCAST
                       ^
/usr/include/net/if.h:46:5: note: previous definition is here
   IFF_BROADCAST = 0x2,        /* Broadcast address valid.  */
   ^
In file included from /yeti/vm/ubuntu-initrd-16.04-x86_64-build/dpdk/drivers/net/tap/rte_eth_tap.c:50:
/usr/include/linux/if.h:73:2: error: redefinition of enumerator 'IFF_DEBUG'
       IFF_DEBUG                       = 1<<2,  /* sysfs */
       ^
/usr/include/net/if.h:49:20: note: expanded from macro 'IFF_DEBUG'
# define IFF_DEBUG      IFF_DEBUG
                       ^
/usr/include/net/if.h:48:5: note: previous definition is here
   IFF_DEBUG = 0x4,            /* Turn on debugging.  */
   ^
In file included from /yeti/vm/ubuntu-initrd-16.04-x86_64-build/dpdk/drivers/net/tap/rte_eth_tap.c:50:
/usr/include/linux/if.h:74:2: error: redefinition of enumerator 'IFF_LOOPBACK'
       IFF_LOOPBACK                    = 1<<3,  /* __volatile__ */
       ^
/usr/include/net/if.h:51:23: note: expanded from macro 'IFF_LOOPBACK'
# define IFF_LOOPBACK   IFF_LOOPBACK
                       ^
/usr/include/net/if.h:50:5: note: previous definition is here
   IFF_LOOPBACK = 0x8,         /* Is a loopback net.  */
   ^
In file included from /yeti/vm/ubuntu-initrd-16.04-x86_64-build/dpdk/drivers/net/tap/rte_eth_tap.c:50:
/usr/include/linux/if.h:75:2: error: redefinition of enumerator 'IFF_POINTOPOINT'
       IFF_POINTOPOINT                 = 1<<4,  /* __volatile__ */
       ^
/usr/include/net/if.h:53:26: note: expanded from macro 'IFF_POINTOPOINT'
# define IFF_POINTOPOINT IFF_POINTOPOINT
                        ^
/usr/include/net/if.h:52:5: note: previous definition is here
   IFF_POINTOPOINT = 0x10,     /* Interface is point-to-point link.  */
   ^
In file included from /yeti/vm/ubuntu-initrd-16.04-x86_64-build/dpdk/drivers/net/tap/rte_eth_tap.c:50:
/usr/include/linux/if.h:76:2: error: redefinition of enumerator 'IFF_NOTRAILERS'
       IFF_NOTRAILERS                  = 1<<5,  /* sysfs */
       ^
/usr/include/net/if.h:55:25: note: expanded from macro 'IFF_NOTRAILERS'
# define IFF_NOTRAILERS IFF_NOTRAILERS
                       ^
/usr/include/net/if.h:54:5: note: previous definition is here
   IFF_NOTRAILERS = 0x20,      /* Avoid use of trailers.  */
   ^
In file included from /yeti/vm/ubuntu-initrd-16.04-x86_64-build/dpdk/drivers/net/tap/rte_eth_tap.c:50:
/usr/include/linux/if.h:77:2: error: redefinition of enumerator 'IFF_RUNNING'
       IFF_RUNNING                     = 1<<6,  /* __volatile__ */
       ^
/usr/include/net/if.h:57:22: note: expanded from macro 'IFF_RUNNING'
# define IFF_RUNNING    IFF_RUNNING
                       ^
/usr/include/net/if.h:56:5: note: previous definition is here
   IFF_RUNNING = 0x40,         /* Resources allocated.  */
   ^
In file included from /yeti/vm/ubuntu-initrd-16.04-x86_64-build/dpdk/drivers/net/tap/rte_eth_tap.c:50:
/usr/include/linux/if.h:78:2: error: redefinition of enumerator 'IFF_NOARP'
       IFF_NOARP                       = 1<<7,  /* sysfs */
       ^
/usr/include/net/if.h:59:20: note: expanded from macro 'IFF_NOARP'
# define IFF_NOARP      IFF_NOARP
                       ^
/usr/include/net/if.h:58:5: note: previous definition is here
   IFF_NOARP = 0x80,           /* No address resolution protocol.  */
   ^
In file included from /yeti/vm/ubuntu-initrd-16.04-x86_64-build/dpdk/drivers/net/tap/rte_eth_tap.c:50:
/usr/include/linux/if.h:79:2: error: redefinition of enumerator 'IFF_PROMISC'
       IFF_PROMISC                     = 1<<8,  /* sysfs */
       ^
/usr/include/net/if.h:61:22: note: expanded from macro 'IFF_PROMISC'
# define IFF_PROMISC    IFF_PROMISC
                       ^
/usr/include/net/if.h:60:5: note: previous definition is here
   IFF_PROMISC = 0x100,        /* Receive all packets.  */
   ^
In file included from /yeti/vm/ubuntu-initrd-16.04-x86_64-build/dpdk/drivers/net/tap/rte_eth_tap.c:50:
/usr/include/linux/if.h:80:2: error: redefinition of enumerator 'IFF_ALLMULTI'
       IFF_ALLMULTI                    = 1<<9,  /* sysfs */
       ^
/usr/include/net/if.h:65:23: note: expanded from macro 'IFF_ALLMULTI'
# define IFF_ALLMULTI   IFF_ALLMULTI
                       ^
/usr/include/net/if.h:64:5: note: previous definition is here
   IFF_ALLMULTI = 0x200,       /* Receive all multicast packets.  */
   ^
In file included from /yeti/vm/ubuntu-initrd-16.04-x86_64-build/dpdk/drivers/net/tap/rte_eth_tap.c:50:
/usr/include/linux/if.h:81:2: error: redefinition of enumerator 'IFF_MASTER'
       IFF_MASTER                      = 1<<10, /* __volatile__ */
       ^
/usr/include/net/if.h:68:21: note: expanded from macro 'IFF_MASTER'
# define IFF_MASTER     IFF_MASTER
                       ^
/usr/include/net/if.h:67:5: note: previous definition is here
   IFF_MASTER = 0x400,         /* Master of a load balancer.  */
   ^
In file included from /yeti/vm/ubuntu-initrd-16.04-x86_64-build/dpdk/drivers/net/tap/rte_eth_tap.c:50:
/usr/include/linux/if.h:82:2: error: redefinition of enumerator 'IFF_SLAVE'
       IFF_SLAVE                       = 1<<11, /* __volatile__ */
       ^
/usr/include/net/if.h:70:20: note: expanded from macro 'IFF_SLAVE'
# define IFF_SLAVE      IFF_SLAVE
                       ^
/usr/include/net/if.h:69:5: note: previous definition is here
   IFF_SLAVE = 0x800,          /* Slave of a load balancer.  */
   ^
In file included from /yeti/vm/ubuntu-initrd-16.04-x86_64-build/dpdk/drivers/net/tap/rte_eth_tap.c:50:
/usr/include/linux/if.h:83:2: error: redefinition of enumerator 'IFF_MULTICAST'
       IFF_MULTICAST                   = 1<<12, /* sysfs */
       ^
/usr/include/net/if.h:73:24: note: expanded from macro 'IFF_MULTICAST'
# define IFF_MULTICAST  IFF_MULTICAST
                       ^
/usr/include/net/if.h:72:5: note: previous definition is here
   IFF_MULTICAST = 0x1000,     /* Supports multicast.  */
   ^
In file included from /yeti/vm/ubuntu-initrd-16.04-x86_64-build/dpdk/drivers/net/tap/rte_eth_tap.c:50:
/usr/include/linux/if.h:84:2: error: redefinition of enumerator 'IFF_PORTSEL'
       IFF_PORTSEL                     = 1<<13, /* sysfs */
       ^
/usr/include/net/if.h:76:22: note: expanded from macro 'IFF_PORTSEL'
# define IFF_PORTSEL    IFF_PORTSEL
                       ^
/usr/include/net/if.h:75:5: note: previous definition is here
   IFF_PORTSEL = 0x2000,       /* Can set media type.  */
   ^
In file included from /yeti/vm/ubuntu-initrd-16.04-x86_64-build/dpdk/drivers/net/tap/rte_eth_tap.c:50:
/usr/include/linux/if.h:85:2: error: redefinition of enumerator 'IFF_AUTOMEDIA'
       IFF_AUTOMEDIA                   = 1<<14, /* sysfs */
       ^
/usr/include/net/if.h:78:24: note: expanded from macro 'IFF_AUTOMEDIA'
# define IFF_AUTOMEDIA  IFF_AUTOMEDIA
                       ^
/usr/include/net/if.h:77:5: note: previous definition is here
   IFF_AUTOMEDIA = 0x4000,     /* Auto media select active.  */
   ^
In file included from /yeti/vm/ubuntu-initrd-16.04-x86_64-build/dpdk/drivers/net/tap/rte_eth_tap.c:50:
/usr/include/linux/if.h:86:2: error: redefinition of enumerator 'IFF_DYNAMIC'
       IFF_DYNAMIC                     = 1<<15, /* sysfs */
       ^
/usr/include/net/if.h:80:22: note: expanded from macro 'IFF_DYNAMIC'
# define IFF_DYNAMIC    IFF_DYNAMIC
                       ^
/usr/include/net/if.h:79:5: note: previous definition is here
   IFF_DYNAMIC = 0x8000        /* Dialup device with changing addresses.  */
   ^
In file included from /yeti/vm/ubuntu-initrd-16.04-x86_64-build/dpdk/drivers/net/tap/rte_eth_tap.c:50:
/usr/include/linux/if.h:169:8: error: redefinition of 'ifmap'
struct ifmap {
      ^
/usr/include/net/if.h:111:8: note: previous definition is here
struct ifmap
      ^
In file included from /yeti/vm/ubuntu-initrd-16.04-x86_64-build/dpdk/drivers/net/tap/rte_eth_tap.c:50:
/usr/include/linux/if.h:203:8: error: redefinition of 'ifreq'
struct ifreq {
      ^
/usr/include/net/if.h:126:8: note: previous definition is here
struct ifreq
      ^
In file included from /yeti/vm/ubuntu-initrd-16.04-x86_64-build/dpdk/drivers/net/tap/rte_eth_tap.c:50:
/usr/include/linux/if.h:252:8: error: redefinition of 'ifconf'
struct ifconf  {
      ^
/usr/include/net/if.h:176:8: note: previous definition is here
struct ifconf
      ^
19 errors generated.
make[6]: *** [rte_eth_tap.o] Error 1
make[5]: *** [tap] Error 2
make[5]: *** Waiting for unfinished jobs....
make[4]: *** [net] Error 2
make[3]: *** [drivers] Error 2
make[2]: *** [all] Error 2
make[1]: *** [pre_install] Error 2
make: *** [install] Error 2
error: build failed
#51 lore raw
Quoted 17 lines · Keith

Regards,
Keith

Quoted 13 lines · Yuanhan Liu
Quoted 10 lines · Keith Wiles

The rte_eth_tap.c PMD creates a device using TUN/TAP interfaces on the local host. The PMD allows for DPDK and the host to communicate using a raw device interface on the host and in the DPDK application. The device created is a Tap device with a L2 packet header.

v2 - merge all of the patches into one patch.
Fix a typo on naming the tap device.
Update the maintainers list

Quoted 7 lines
Quoted 5 lines
Quoted 1 line

Hi,

FYI, my robot caught some build errors with this patch applied.

All of the below errors are from Linux header files and not the Tap driver.

Yes, but you are referencing them, so ...

Quoted 1 line

How are you building DPDK, which config file are you using, OS version, .... Some details would be nice to understand how these errors are being generated and how I can recreate them.

My bad. I should have included all of them: my robot is still at a very rough stage: those info should be included automatically in future (badly, I don't even have time to enhance it).

However, for this issue, I think it can be triggered on most (if not any) linux platforms: I got the report from ubuntu-16.04, and I have just tried it on my dev box, which is fedora 20: I got the same error.

	--yliu
Quoted 8 lines
Quoted 8 lines

---
In file included from /yeti/vm/ubuntu-initrd-16.04-x86_64-build/dpdk/drivers/net/tap/rte_eth_tap.c:50:
/usr/include/linux/if.h:71:2: error: redefinition of enumerator 'IFF_UP'
IFF_UP = 1<<0, /* sysfs */
^
/usr/include/net/if.h:45:17: note: expanded from macro 'IFF_UP'

# define IFF_UP IFF_UP
               ^
#52 lore raw

Sent from my iPhone

On Sep 18, 2016, at 7:28 PM, Yuanhan Liu <[email protected]<mailto:[email protected]>> wrote:

On Sun, Sep 18, 2016 at 04:20:12PM +0000, Wiles, Keith wrote:

Regards,
Keith

On Sep 18, 2016, at 8:25 AM, Yuanhan Liu <[email protected]<mailto:[email protected]>> wrote:

On Fri, Sep 16, 2016 at 11:22:13AM -0500, Keith Wiles wrote:
The rte_eth_tap.c PMD creates a device using TUN/TAP interfaces on the local host. The PMD allows for DPDK and the host to communicate using a raw device interface on the host and in the DPDK application. The device created is a Tap device with a L2 packet header.

v2 - merge all of the patches into one patch.
Fix a typo on naming the tap device.
Update the maintainers list

Hi,

FYI, my robot caught some build errors with this patch applied.

All of the below errors are from Linux header files and not the Tap driver.

Yes, but you are referencing them, so ...

Yes I am referencing them, but still they are not interacting with the tap driver. It would seem like you could delete all of the code in the tap driver and still get these errors. Which you could try if you want, just ifdef the code and see if it still happens.

On my ubuntu 16.04 machine I do not get these errors. I need to know how to reproduce the failure to fix it.

How are you building DPDK, which config file are you using, OS version, .... Some details would be nice to understand how these errors are being generated and how I can recreate them.

My bad. I should have included all of them: my robot is still at a very rough stage: those info should be included automatically in future (badly, I don't even have time to enhance it).

However, for this issue, I think it can be triggered on most (if not any) linux platforms: I got the report from ubuntu-16.04, and I have just tried it on my dev box, which is fedora 20: I got the same error.

--yliu

---
In file included from /yeti/vm/ubuntu-initrd-16.04-x86_64-build/dpdk/drivers/net/tap/rte_eth_tap.c:50:
/usr/include/linux/if.h:71:2: error: redefinition of enumerator 'IFF_UP'
IFF_UP = 1<<0, /* sysfs */
^
/usr/include/net/if.h:45:17: note: expanded from macro 'IFF_UP'

# define IFF_UP IFF_UP
              ^
#53 lore raw
Quoted 9 lines · Keith

All of the below errors are from Linux header files and not the Tap driver.

Yes, but you are referencing them, so ...

Yes I am referencing them, but still they are not interacting with the tap driver.

Yes, they are not. But the tap driver code references them. I mean, you might have done that wrongly. e.g., the build passes if I removed few includes for linux header files:

    diff --git a/drivers/net/tap/rte_eth_tap.c b/drivers/net/tap/rte_eth_tap.c
    index 4621c55..2a6bcd0 100644
    --- a/drivers/net/tap/rte_eth_tap.c
    +++ b/drivers/net/tap/rte_eth_tap.c
    @@ -47,9 +47,7 @@
     #include <arpa/inet.h>
     #include <net/if.h>
     #ifdef __linux__
    -#include <linux/if.h>
     #include <linux/if_tun.h>
    -#include <linux/if_ether.h>
     #else
     #include <netinet/if_ether.h>
     #endif
Quoted 6 lines

It would seem like you could delete all of the code in the tap driver and still get these errors. Which you could try if you want, just ifdef the code and see if it still happens.

On my ubuntu 16.04 machine I do not get these errors. I need to know how to reproduce the failure to fix it.

Interesting! I really did no magic to reproduce it. As said, I can reproduce it pretty easily with ubuntu 16.04 and fedora 20, just with default build:

    $ make install T=$RTE_SDK

Anyway, the error is straightforward after all: both files (net/if.h and linux/if.h) define same structures and macros. Including both of them would result errors like I reported.

	--yliu
#54 lore raw
In reply to Keith Wiles

[ just got few more comments after some fiddling with the build issue ]

drivers/net/tap/Makefile line 5

Quoted 11 lines · Keith Wiles
drivers/net/tap/Makefile new file +5 -0
new file mode 100644
index 0000000..442a2fe
@@ -0,0 +1,60 @@
+#   BSD LICENSE
+#
+#   Copyright(c) 2014 John W. Linville <[email protected]>
+#   Copyright(c) 2010-2014 Intel Corporation. All rights reserved.
+#   Copyright(c) 2014 6WIND S.A.

Just wondering, is it on purpose, or just yet another copy-paste error?

drivers/net/tap/Makefile line 8

Quoted 3 lines
drivers/net/tap/Makefile new file +3 -0
+#include <arpa/inet.h>
+#include <net/if.h>
+#ifdef __linux__

So, you meant to add other OS support, say BSD?

drivers/net/tap/Makefile line 20

Quoted 12 lines
drivers/net/tap/Makefile new file +12 -0
+#include <linux/if.h>
+#include <linux/if_tun.h>
+#include <linux/if_ether.h>
+#else
+#include <netinet/if_ether.h>
+#endif
+#include <fcntl.h>
+
+#include <poll.h>
+
+/* Linux based path to the TUN device */
+#define TUN_TAP_DEV_PATH        "/dev/net/tun"

However, you hardcoded a linux only path here. While checking the code from qemu, I saw that the path is actually different from different UNIX variants, even for FreeBSD and NetBSD.

    [yliu@yliu-dev ~/qemu]$ grep -E "/dev/.*(tap|tun)" net/tap*.c
    net/tap-bsd.c:            snprintf(dname, sizeof dname, "/dev/tun%d", i);
    net/tap-bsd.c:            snprintf(dname, sizeof dname, "/dev/tap%d", i);
    net/tap-bsd.c:#define PATH_NET_TAP "/dev/tap"
    net/tap-linux.c:#define PATH_NET_TUN "/dev/net/tun"
    net/tap-solaris.c:    TFR(tap_fd = open("/dev/tap", O_RDWR, 0));
    ...

	--yliu
#55 lore raw

Regards,
Keith

Quoted 5 lines · Yuanhan Liu

[ just got few more comments after some fiddling with the build issue ]

On Fri, Sep 16, 2016 at 11:22:13AM -0500, Keith Wiles wrote:

Quoted 13 lines

drivers/net/tap/Makefile line 5

Quoted 11 lines
drivers/net/tap/Makefile new file +5 -0
new file mode 100644
index 0000000..442a2fe
@@ -0,0 +1,60 @@
+#   BSD LICENSE
+#
+#   Copyright(c) 2014 John W. Linville <[email protected]>
+#   Copyright(c) 2010-2014 Intel Corporation. All rights reserved.
+#   Copyright(c) 2014 6WIND S.A.

Just wondering, is it on purpose, or just yet another copy-paste error?

Yes, I will send another patch to fix this one.

Quoted 6 lines

drivers/net/tap/Makefile line 8

Quoted 4 lines
drivers/net/tap/Makefile new file +3 -0
+#include <arpa/inet.h>
+#include <net/if.h>
+#ifdef __linux__

So, you meant to add other OS support, say BSD?

I did not mean to support BSD, just Linux and I will remove the ifdef to reflect it. If you look at the configuration file I only added it to the common_linux file and not BSD.

The concern I am having is on my standard Ubuntu 16.04 system these errors do not appear. I would like to understand why they appeared on your system. The ifdef __linux__ must be enabled as the defines in the linux/if_tun.h file do not give an error. I did play with the headers already and removed the ifdef, but as I could not reproduce your build failure it did not trigger anything new.

I will look at it some more, but it does not make sense and I do want to make sure it works.

Quoted 17 lines

drivers/net/tap/Makefile line 20

Quoted 13 lines
drivers/net/tap/Makefile new file +12 -0
+#include <linux/if.h>
+#include <linux/if_tun.h>
+#include <linux/if_ether.h>
+#else
+#include <netinet/if_ether.h>
+#endif
+#include <fcntl.h>
+
+#include <poll.h>
+
+/* Linux based path to the TUN device */
+#define TUN_TAP_DEV_PATH        "/dev/net/tun"

However, you hardcoded a linux only path here. While checking the code from qemu, I saw that the path is actually different from different UNIX variants, even for FreeBSD and NetBSD.

I only assumed this to work for Linux and not FreeBSD/NetBSD as the handling of the opens on the tun device are different then linux. I also only added it to the common_linux configuration file.

Quoted 10 lines
   [yliu@yliu-dev ~/qemu]$ grep -E "/dev/.*(tap|tun)" net/tap*.c
   net/tap-bsd.c:            snprintf(dname, sizeof dname, "/dev/tun%d", i);
   net/tap-bsd.c:            snprintf(dname, sizeof dname, "/dev/tap%d", i);
   net/tap-bsd.c:#define PATH_NET_TAP "/dev/tap"
   net/tap-linux.c:#define PATH_NET_TUN "/dev/net/tun"
   net/tap-solaris.c:    TFR(tap_fd = open("/dev/tap", O_RDWR, 0));
   ...
	--yliu
#56 lore raw
Quoted 1 line · Keith

The concern I am having is on my standard Ubuntu 16.04 system these errors do not appear.

Yes, that's really weird. To me, it's such a solid error that could be triggered on all linux platforms.

Anyway, I saw you removed the <net/if.h> including in v3, which makes my robot happy :)

Quoted 22 lines

I would like to understand why they appeared on your system. The ifdef __linux__ must be enabled as the defines in the linux/if_tun.h file do not give an error. I did play with the headers already and removed the ifdef, but as I could not reproduce your build failure it did not trigger anything new.

I will look at it some more, but it does not make sense and I do want to make sure it works.

Quoted 17 lines

drivers/net/tap/rte_eth_tap.c line 61

Quoted 13 lines
drivers/net/tap/rte_eth_tap.c new file +12 -0
+#include <linux/if.h>
+#include <linux/if_tun.h>
+#include <linux/if_ether.h>
+#else
+#include <netinet/if_ether.h>
+#endif
+#include <fcntl.h>
+
+#include <poll.h>
+
+/* Linux based path to the TUN device */
+#define TUN_TAP_DEV_PATH        "/dev/net/tun"

However, you hardcoded a linux only path here. While checking the code from qemu, I saw that the path is actually different from different UNIX variants, even for FreeBSD and NetBSD.

Quoted 1 line

I only assumed this to work for Linux and not FreeBSD/NetBSD as the handling of the opens on the tun device are different then linux. I also only added it to the common_linux configuration file.

I didn't notice the common_linux file, but I have guessed so (that you meant to enable linux only): the "ifdef __linux" stuff just confuses me a bit.

	--yliu
Quoted 11 lines
Quoted 11 lines

   [yliu@yliu-dev ~/qemu]$ grep -E "/dev/.*(tap|tun)" net/tap*.c
   net/tap-bsd.c:            snprintf(dname, sizeof dname, "/dev/tun%d", i);
   net/tap-bsd.c:            snprintf(dname, sizeof dname, "/dev/tap%d", i);
   net/tap-bsd.c:#define PATH_NET_TAP "/dev/tap"
   net/tap-linux.c:#define PATH_NET_TUN "/dev/net/tun"
   net/tap-solaris.c:    TFR(tap_fd = open("/dev/tap", O_RDWR, 0));
   ...
	--yliu
#57 lore raw

2016-09-21 10:13, Yuanhan Liu:

Quoted 6 lines · Keith
Quoted 1 line

I only assumed this to work for Linux and not FreeBSD/NetBSD as the handling of the opens on the tun device are different then linux. I also only added it to the common_linux configuration file.

I didn't notice the common_linux file, but I have guessed so (that you meant to enable linux only): the "ifdef __linux" stuff just confuses me a bit.

By the way, we have a macro for Linux environment: RTE_EXEC_ENV_LINUXAPP Please prefer it over __linux

#58 lore raw

Regards,
Keith

Quoted 12 lines · Thomas Monjalon

2016-09-21 10:13, Yuanhan Liu:

Quoted 6 lines · Keith
Quoted 1 line

I only assumed this to work for Linux and not FreeBSD/NetBSD as the handling of the opens on the tun device are different then linux. I also only added it to the common_linux configuration file.

I didn't notice the common_linux file, but I have guessed so (that you meant to enable linux only): the "ifdef __linux" stuff just confuses me a bit.

By the way, we have a macro for Linux environment: RTE_EXEC_ENV_LINUXAPP Please prefer it over __linux

I removed the __linux__ ifdef in v3, BTW I think we need to grep the code for the __linux__ as I started with one of the other drivers. I will try to remember to grep the drivers when I am back home.