* [dpdk-dev] [PATCH 1/2] doc: update testpmd guide about csum forward engine
2015-03-19 16:36 [dpdk-dev] [PATCH 0/2] doc: update offload API Olivier Matz
@ 2015-03-19 16:36 ` Olivier Matz
2015-03-27 16:20 ` De Lara Guarch, Pablo
2015-03-19 16:36 ` [dpdk-dev] [PATCH 2/2] doc: add a description of the offload API Olivier Matz
2015-03-31 0:28 ` [dpdk-dev] [PATCH 0/2] doc: update " Thomas Monjalon
2 siblings, 1 reply; 5+ messages in thread
From: Olivier Matz @ 2015-03-19 16:36 UTC (permalink / raw)
To: dev
Document the functions introduced by commit 64fc36064d.
Signed-off-by: Olivier Matz <olivier.matz@6wind.com>
---
app/test-pmd/cmdline.c | 2 +-
doc/guides/testpmd_app_ug/testpmd_funcs.rst | 53 +++++++++++++++++++++++------
2 files changed, 44 insertions(+), 11 deletions(-)
diff --git a/app/test-pmd/cmdline.c b/app/test-pmd/cmdline.c
index 604c3ec..e6107a0 100644
--- a/app/test-pmd/cmdline.c
+++ b/app/test-pmd/cmdline.c
@@ -317,7 +317,7 @@ static void cmd_help_long_parsed(void *parsed_result,
"csum set (ip|udp|tcp|sctp|outer-ip) (hw|sw) (port_id)\n"
" Select hardware or software calculation of the"
- " checksum with when transmitting a packet using the"
+ " checksum when transmitting a packet using the"
" csum forward engine.\n"
" ip|udp|tcp|sctp always concern the inner layer.\n"
" outer-ip concerns the outer IP layer in"
diff --git a/doc/guides/testpmd_app_ug/testpmd_funcs.rst b/doc/guides/testpmd_app_ug/testpmd_funcs.rst
index 4d430ff..b3abe9c 100644
--- a/doc/guides/testpmd_app_ug/testpmd_funcs.rst
+++ b/doc/guides/testpmd_app_ug/testpmd_funcs.rst
@@ -505,28 +505,60 @@ Disable hardware insertion of a VLAN header in packets sent on a port:
tx_vlan reset (port_id)
-tx_checksum set
-~~~~~~~~~~~~~~~
+csum set
+~~~~~~~~
Select hardware or software calculation of the checksum when
transmitting a packet using the csum forward engine:
-tx_cksum set (ip|udp|tcp|sctp|vxlan)
+csum set (ip|udp|tcp|sctp|outer-ip) (hw|sw) (port_id)
+
+- ip|udp|tcp|sctp always concern the inner layer.
-ip|udp|tcp|sctp always concern the inner layer.
-vxlan concerns the outer IP and UDP layer (in case the packet
-is recognized as a vxlan packet by the forward engine)
+- outer-ip concerns the outer IP layer in case the packet is recognized
+ as a tunnel packet by the forward engine (vxlan, gre and ipip are
+ supported). See "csum parse-tunnel" command.
.. note::
Check the NIC Datasheet for hardware limits.
-tx_checksum show
-~~~~~~~~~~~~~~~~
+csum parse-tunnel
+~~~~~~~~~~~~~~~~~
+
+Define how tunneled packets should be handled by the csum forward
+engine.
+
+csum parse-tunnel (on|off) (tx_port_id)
+
+If enabled, the csum forward engine will try to recognize supported
+tunnel headers (vxlan, gre, ipip).
+
+If disabled, treat tunnel packets as non-tunneled packets (a inner
+header is handled as a packet payload).
+
+.. note::
+
+ The port argument is the TX port like in the "csum set" command.
+
+Example:
+
+Consider a packet as following:
+"eth_out/ipv4_out/udp_out/vxlan/eth_in/ipv4_in/tcp_in"
+
+- If parse-tunnel is enabled, the ip|udp|tcp|sctp parameters of "csum
+ set" command are about inner headers (here ipv4_in and tcp_in), and the
+ outer-ip parameter is about outer headers (here ipv4_out).
+
+- If parse-tunnel is disabled, the ip|udp|tcp|sctp parameters of "csum
+ set" command are about outer headers, here ipv4_out and udp_out.
+
+csum show
+~~~~~~~~~
Display tx checksum offload configuration:
-tx_checksum show (port_id)
+csum show (port_id)
tso set
~~~~~~~
@@ -536,7 +568,8 @@ Enable TCP Segmentation Offload in csum forward engine:
tso set (segsize) (port_id)
.. note::
- Please check the NIC datasheet for HW limits
+
+ Check the NIC datasheet for hardware limits
tso show
~~~~~~~~
--
2.1.4
^ permalink raw reply [flat|nested] 5+ messages in thread* [dpdk-dev] [PATCH 2/2] doc: add a description of the offload API
2015-03-19 16:36 [dpdk-dev] [PATCH 0/2] doc: update offload API Olivier Matz
2015-03-19 16:36 ` [dpdk-dev] [PATCH 1/2] doc: update testpmd guide about csum forward engine Olivier Matz
@ 2015-03-19 16:36 ` Olivier Matz
2015-03-31 0:28 ` [dpdk-dev] [PATCH 0/2] doc: update " Thomas Monjalon
2 siblings, 0 replies; 5+ messages in thread
From: Olivier Matz @ 2015-03-19 16:36 UTC (permalink / raw)
To: dev
Signed-off-by: Olivier Matz <olivier.matz@6wind.com>
---
doc/guides/prog_guide/mbuf_lib.rst | 89 +++++++++++++++++++++++++++++++++
doc/guides/prog_guide/poll_mode_drv.rst | 24 +++++----
2 files changed, 104 insertions(+), 9 deletions(-)
diff --git a/doc/guides/prog_guide/mbuf_lib.rst b/doc/guides/prog_guide/mbuf_lib.rst
index 3d59e96..6e3b209 100644
--- a/doc/guides/prog_guide/mbuf_lib.rst
+++ b/doc/guides/prog_guide/mbuf_lib.rst
@@ -148,6 +148,95 @@ An mbuf also contains the input port (where it comes from), and the number of se
For chained buffers, only the first mbuf of the chain stores this meta information.
+For instance, this is the case on RX side for the IEEE1588 packet
+timestamp mechanism, the VLAN tagging and the IP checksum computation.
+
+On TX side, it is also possible for an application to delegate some
+processing to the hardware if it supports it. For instance, the
+PKT_TX_IP_CKSUM flag allows to offload the computation of the IPv4
+checksum.
+
+The following examples explain how to configure different TX offloads on
+a vxlan-encapsulated tcp packet:
+``out_eth/out_ip/out_udp/vxlan/in_eth/in_ip/in_tcp/payload``
+
+- calculate checksum of out_ip::
+
+ mb->l2_len = len(out_eth)
+ mb->l3_len = len(out_ip)
+ mb->ol_flags |= PKT_TX_IPV4 | PKT_TX_IP_CSUM
+ set out_ip checksum to 0 in the packet
+
+ This is supported on hardwares advertising DEV_TX_OFFLOAD_IPV4_CKSUM.
+
+- calculate checksum of out_ip and out_udp::
+
+ mb->l2_len = len(out_eth)
+ mb->l3_len = len(out_ip)
+ mb->ol_flags |= PKT_TX_IPV4 | PKT_TX_IP_CSUM | PKT_TX_UDP_CKSUM
+ set out_ip checksum to 0 in the packet
+ set out_udp checksum to pseudo header using rte_ipv4_phdr_cksum()
+
+ This is supported on hardwares advertising DEV_TX_OFFLOAD_IPV4_CKSUM
+ and DEV_TX_OFFLOAD_UDP_CKSUM.
+
+- calculate checksum of in_ip::
+
+ mb->l2_len = len(out_eth + out_ip + out_udp + vxlan + in_eth)
+ mb->l3_len = len(in_ip)
+ mb->ol_flags |= PKT_TX_IPV4 | PKT_TX_IP_CSUM
+ set in_ip checksum to 0 in the packet
+
+ This is similar to case 1), but l2_len is different. It is supported
+ on hardwares advertising DEV_TX_OFFLOAD_IPV4_CKSUM.
+ Note that it can only work if outer L4 checksum is 0.
+
+- calculate checksum of in_ip and in_tcp::
+
+ mb->l2_len = len(out_eth + out_ip + out_udp + vxlan + in_eth)
+ mb->l3_len = len(in_ip)
+ mb->ol_flags |= PKT_TX_IPV4 | PKT_TX_IP_CSUM | PKT_TX_TCP_CKSUM
+ set in_ip checksum to 0 in the packet
+ set in_tcp checksum to pseudo header using rte_ipv4_phdr_cksum()
+
+ This is similar to case 2), but l2_len is different. It is supported
+ on hardware advertising DEV_TX_OFFLOAD_IPV4_CKSUM and
+ DEV_TX_OFFLOAD_TCP_CKSUM.
+ Note that it can only work if outer L4 checksum is 0.
+
+- segment inner TCP::
+
+ mb->l2_len = len(out_eth + out_ip + out_udp + vxlan + in_eth)
+ mb->l3_len = len(in_ip)
+ mb->l4_len = len(in_tcp)
+ mb->ol_flags |= PKT_TX_IPV4 | PKT_TX_IP_CKSUM | PKT_TX_TCP_CKSUM |
+ PKT_TX_TCP_SEG;
+ set in_ip checksum to 0 in the packet
+ set in_tcp checksum to pseudo header without including the IP
+ payload length using rte_ipv4_phdr_cksum()
+
+ This is supported on hardware advertising DEV_TX_OFFLOAD_TCP_TSO.
+ Note that it can only work if outer L4 checksum is 0.
+
+- calculate checksum of out_ip, in_ip, in_tcp::
+
+ mb->outer_l2_len = len(out_eth)
+ mb->outer_l3_len = len(out_ip)
+ mb->l2_len = len(out_udp + vxlan + in_eth)
+ mb->l3_len = len(in_ip)
+ mb->ol_flags |= PKT_TX_OUTER_IPV4 | PKT_TX_OUTER_IP_CKSUM | \
+ PKT_TX_IP_CKSUM | PKT_TX_TCP_CKSUM;
+ set out_ip checksum to 0 in the packet
+ set in_ip checksum to 0 in the packet
+ set in_tcp checksum to pseudo header using rte_ipv4_phdr_cksum()
+
+ This is supported on hardware advertising DEV_TX_OFFLOAD_IPV4_CKSUM,
+ DEV_TX_OFFLOAD_UDP_CKSUM and DEV_TX_OFFLOAD_OUTER_IPV4_CKSUM.
+
+The list of flags and their precise meaning is described in the mbuf API
+documentation (rte_mbuf.h). Also refer to the testpmd source code
+(specifically the csumonly.c file) for details.
+
Direct and Indirect Buffers
---------------------------
diff --git a/doc/guides/prog_guide/poll_mode_drv.rst b/doc/guides/prog_guide/poll_mode_drv.rst
index e9e1132..7b9e7ca 100755
--- a/doc/guides/prog_guide/poll_mode_drv.rst
+++ b/doc/guides/prog_guide/poll_mode_drv.rst
@@ -198,15 +198,7 @@ the Intel® 82599 10 Gigabit Ethernet Controller controllers in the testpmd appl
Other features such as the L3/L4 5-Tuple packet filtering feature of a port can be configured in the same way.
Ethernet* flow control (pause frame) can be configured on the individual port.
Refer to the testpmd source code for details.
-Also, L4 (UDP/TCP/ SCTP) checksum offload by the NIC can be enabled for an individual packet as long as the packet mbuf is set up correctly.
-In terms of UDP tunneling packet, the PKT_TX_UDP_TUNNEL_PKT flag must be set to enable tunneling packet TX checksum offload for both outer layer and inner layer.
-Refer to the testpmd source code (specifically the csumonly.c file) for details.
-
-That being said, the support of some offload features implies the addition of dedicated status bit(s) and value field(s) into the rte_mbuf
-data structure, along with their appropriate handling by the receive/transmit functions exported by each PMD.
-
-For instance, this is the case for the IEEE1588 packet timestamp mechanism, the VLAN tagging and the IP checksum computation, as described in
-the Section 7.6 "Meta Information".
+Also, L4 (UDP/TCP/ SCTP) checksum offload by the NIC can be enabled for an individual packet as long as the packet mbuf is set up correctly. See `Hardware Offload`_ for details.
Configuration of Transmit and Receive Queues
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
@@ -257,6 +249,20 @@ One descriptor in the TX ring is used as a sentinel to avoid a hardware race con
When configuring for DCB operation, at port initialization, both the number of transmit queues and the number of receive queues must be set to 128.
+Hardware Offload
+~~~~~~~~~~~~~~~~
+
+Depending on driver capabilities advertised by
+``rte_eth_dev_info_get()``, the PMD may support hardware offloading
+feature like checksumming, TCP segmentation or VLAN insertion.
+
+The support of these offload features implies the addition of dedicated
+status bit(s) and value field(s) into the rte_mbuf data structure, along
+with their appropriate handling by the receive/transmit functions
+exported by each PMD. The list of flags and their precise meaning is
+described in the mbuf API documentation and in the in :ref:`Mbuf Library
+<Mbuf_Library>`, section "Meta Information".
+
Poll Mode Driver API
--------------------
--
2.1.4
^ permalink raw reply [flat|nested] 5+ messages in thread