diff options
author | Or Gerlitz <ogerlitz@voltaire.com> | 2009-04-08 16:52:01 -0400 |
---|---|---|
committer | Roland Dreier <rolandd@cisco.com> | 2009-04-08 16:52:01 -0400 |
commit | 6a3335b43342b42dd6c69b4bbbde15d622cb49ca (patch) | |
tree | c2ec24376c3f3bac9b76c09373444bcea64a5062 /Documentation/infiniband | |
parent | edb5abb1e2a84fd8802a3577d95eac84fe1405ab (diff) |
IPoIB: Document newish features
Update the documentation to include connected mode, stateless offloads
and interrupt moderation, and add a reference to the connected mode RFC.
Signed-off-by: Or Gerlitz <ogerlitz@voltaire.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Diffstat (limited to 'Documentation/infiniband')
-rw-r--r-- | Documentation/infiniband/ipoib.txt | 45 |
1 files changed, 45 insertions, 0 deletions
diff --git a/Documentation/infiniband/ipoib.txt b/Documentation/infiniband/ipoib.txt index 864ff3283780..6d40f00b358c 100644 --- a/Documentation/infiniband/ipoib.txt +++ b/Documentation/infiniband/ipoib.txt | |||
@@ -24,6 +24,49 @@ Partitions and P_Keys | |||
24 | The P_Key for any interface is given by the "pkey" file, and the | 24 | The P_Key for any interface is given by the "pkey" file, and the |
25 | main interface for a subinterface is in "parent." | 25 | main interface for a subinterface is in "parent." |
26 | 26 | ||
27 | Datagram vs Connected modes | ||
28 | |||
29 | The IPoIB driver supports two modes of operation: datagram and | ||
30 | connected. The mode is set and read through an interface's | ||
31 | /sys/class/net/<intf name>/mode file. | ||
32 | |||
33 | In datagram mode, the IB UD (Unreliable Datagram) transport is used | ||
34 | and so the interface MTU has is equal to the IB L2 MTU minus the | ||
35 | IPoIB encapsulation header (4 bytes). For example, in a typical IB | ||
36 | fabric with a 2K MTU, the IPoIB MTU will be 2048 - 4 = 2044 bytes. | ||
37 | |||
38 | In connected mode, the IB RC (Reliable Connected) transport is used. | ||
39 | Connected mode is to takes advantage of the connected nature of the | ||
40 | IB transport and allows an MTU up to the maximal IP packet size of | ||
41 | 64K, which reduces the number of IP packets needed for handling | ||
42 | large UDP datagrams, TCP segments, etc and increases the performance | ||
43 | for large messages. | ||
44 | |||
45 | In connected mode, the interface's UD QP is still used for multicast | ||
46 | and communication with peers that don't support connected mode. In | ||
47 | this case, RX emulation of ICMP PMTU packets is used to cause the | ||
48 | networking stack to use the smaller UD MTU for these neighbours. | ||
49 | |||
50 | Stateless offloads | ||
51 | |||
52 | If the IB HW supports IPoIB stateless offloads, IPoIB advertises | ||
53 | TCP/IP checksum and/or Large Send (LSO) offloading capability to the | ||
54 | network stack. | ||
55 | |||
56 | Large Receive (LRO) offloading is also implemented and may be turned | ||
57 | on/off using ethtool calls. Currently LRO is supported only for | ||
58 | checksum offload capable devices. | ||
59 | |||
60 | Stateless offloads are supported only in datagram mode. | ||
61 | |||
62 | Interrupt moderation | ||
63 | |||
64 | If the underlying IB device supports CQ event moderation, one can | ||
65 | use ethtool to set interrupt mitigation parameters and thus reduce | ||
66 | the overhead incurred by handling interrupts. The main code path of | ||
67 | IPoIB doesn't use events for TX completion signaling so only RX | ||
68 | moderation is supported. | ||
69 | |||
27 | Debugging Information | 70 | Debugging Information |
28 | 71 | ||
29 | By compiling the IPoIB driver with CONFIG_INFINIBAND_IPOIB_DEBUG set | 72 | By compiling the IPoIB driver with CONFIG_INFINIBAND_IPOIB_DEBUG set |
@@ -55,3 +98,5 @@ References | |||
55 | http://ietf.org/rfc/rfc4391.txt | 98 | http://ietf.org/rfc/rfc4391.txt |
56 | IP over InfiniBand (IPoIB) Architecture (RFC 4392) | 99 | IP over InfiniBand (IPoIB) Architecture (RFC 4392) |
57 | http://ietf.org/rfc/rfc4392.txt | 100 | http://ietf.org/rfc/rfc4392.txt |
101 | IP over InfiniBand: Connected Mode (RFC 4755) | ||
102 | http://ietf.org/rfc/rfc4755.txt | ||