By default the arch_fast_hash hashing function pointers are initialized
to jhash(2). If during boot-up a CPU with SSE4.2 is detected they get
updated to the CRC32 ones. This dispatching scheme incurs a function
pointer lookup and indirect call for every hashing operation.
rhashtable as a user of arch_fast_hash e.g. stores pointers to hashing
functions in its structure, too, causing two indirect branches per
hashing operation.
Using alternative_call we can get away with one of those indirect branches.
Acked-by: Daniel Borkmann <dborkman@redhat.com>
Cc: Thomas Graf <tgraf@suug.ch>
Signed-off-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Tom Lendacky says:
====================
amd-xgbe: AMD XGBE driver updates 2014-11-04
The following series of patches includes functional updates to the
driver as well as some trivial changes for function renaming and
spelling fixes.
- Move channel and ring structure allocation into the device open path
- Rename the pre_xmit function to dev_xmit
- Explicitly use the u32 data type for the device descriptors
- Use page allocation for the receive buffers
- Add support for split header/payload receive
- Add support for per DMA channel interrupts
- Add support for receive side scaling (RSS)
- Add support for ethtool receive side scaling commands
- Fix the spelling of descriptors
- After a PCS reset, sync the PCS and PHY modes
- Add dependency on HAS_IOMEM to both the amd-xgbe and amd-xgbe-phy
drivers
This patch series is based on net-next.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
The amd-xgbe-phy driver needs to perform ioremap calls, so add HAS_IOMEM
to its build dependency.
Signed-off-by: Tom Lendacky <thomas.lendacky@amd.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The amd-xgbe driver needs to perform ioremap calls, so add HAS_IOMEM
to its build dependency.
Signed-off-by: Tom Lendacky <thomas.lendacky@amd.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch adds support to sync the states of the PCS and the PHY
after a reset is performed. If the PCS and the PHY are not in the
same state after reset an extra mode change would be performed. This
extra mode change might not be needed if the PCS and the PHY are
synced up after reset.
Signed-off-by: Tom Lendacky <thomas.lendacky@amd.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch fixes the spelling of the word "descriptor" in a couple
of locations.
Signed-off-by: Tom Lendacky <thomas.lendacky@amd.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch adds support for ethtool receive side scaling (RSS) commands.
Support is added to get/set the RSS hash key and the RSS lookup table.
Signed-off-by: Tom Lendacky <thomas.lendacky@amd.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch provides support for receive side scaling (RSS). RSS allows
for spreading incoming network packets across the Rx queues. When used
in conjunction with the per DMA channel interrupt support, this allows
the receive processing to be spread across multiple processors.
Signed-off-by: Tom Lendacky <thomas.lendacky@amd.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch provides support for interrupts that are generated by the
Tx/Rx DMA channel pairs of the device. This allows for Tx and Rx
processing to run across multiple processsors.
Signed-off-by: Tom Lendacky <thomas.lendacky@amd.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Provide support for splitting IP packets so that the header and
payload can be sent to different DMA addresses. This will allow
the IP header to be put into the linear part of the skb while the
payload can be added as frags.
Signed-off-by: Tom Lendacky <thomas.lendacky@amd.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Use page allocations for Rx buffers instead of pre-allocating skbs
of a set size.
Signed-off-by: Tom Lendacky <thomas.lendacky@amd.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The Tx and Rx descriptors are unsigned 32 bit values. Use the u32
type, rather than unsigned int, to map these descriptors.
Signed-off-by: Tom Lendacky <thomas.lendacky@amd.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The pre_xmit function name implies that it performs operations prior
to transmitting the packet when in fact it is responsible for setting
up the descriptors and initiating the transmit. Rename this to
function from pre_xmit to dev_xmit, which is consistent with the name
used during receive processing - dev_read.
Signed-off-by: Tom Lendacky <thomas.lendacky@amd.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Move the channel and ring tracking structures allocation to device
open. This will allow for future support to vary the number of Tx/Rx
queues without unloading the module.
Signed-off-by: Tom Lendacky <thomas.lendacky@amd.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
It is only used in net/ipv6/inet6_hashtables.c.
Cc: David S. Miller <davem@davemloft.net>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This encapsulates all of the skb_copy_datagram_iovec() callers
with call argument signature "skb, offset, msghdr->msg_iov, length".
When we move to iov_iters in the networking, the iov_iter object will
sit in the msghdr.
Having a helper like this means there will be less places to touch
during that transformation.
Based upon descriptions and patch from Al Viro.
Signed-off-by: David S. Miller <davem@davemloft.net>
Tom Herbert says:
====================
gue: Remote checksum offload
This patch set implements remote checksum offload for
GUE, which is a mechanism that provides checksum offload of
encapsulated packets using rudimentary offload capabilities found in
most Network Interface Card (NIC) devices. The outer header checksum
for UDP is enabled in packets and, with some additional meta
information in the GUE header, a receiver is able to deduce the
checksum to be set for an inner encapsulated packet. Effectively this
offloads the computation of the inner checksum. Enabling the outer
checksum in encapsulation has the additional advantage that it covers
more of the packet than the inner checksum including the encapsulation
headers.
Remote checksum offload is described in:
http://tools.ietf.org/html/draft-herbert-remotecsumoffload-01
The GUE transmit and receive paths are modified to support the
remote checksum offload option. The option contains a checksum
offset and checksum start which are directly derived from values
set in stack when doing CHECKSUM_PARTIAL. On receipt of the option, the
operation is to calculate the packet checksum from "start" to end of
the packet (normally derived for checksum complete), and then set
the resultant value at checksum "offset" (the checksum field has
already been primed with the pseudo header). This emulates a NIC
that implements NETIF_F_HW_CSUM.
The primary purpose of this feature is to eliminate cost of performing
checksum calculation over a packet when encpasulating.
In this patch set:
- Move fou_build_header into fou.c and split it into a couple of
functions
- Enable offloading of outer UDP checksum in encapsulation
- Change udp_offload to support remote checksum offload, includes
new GSO type and ensuring encapsulated layers (TCP) doesn't try to
set a checksum covered by RCO
- TX support for RCO with GUE. This is configured through ip_tunnel
and set the option on transmit when packet being encapsulated is
CHECKSUM_PARTIAL
- RX support for RCO with GUE for normal and GRO paths. Includes
resolving the offloaded checksum
v2:
Address comments from davem: Move accounting for private option
field in gue_encap_hlen to patch in which we add the remote checksum
offload option.
Testing:
I ran performance numbers using netperf TCP_STREAM and TCP_RR with 200
streams, comparing GUE with and without remote checksum offload (doing
checksum-unnecessary to complete conversion in both cases). These
were run on mlnx4 and bnx2x. Some mlnx4 results are below.
GRE/GUE
TCP_STREAM
IPv4, with remote checksum offload
9.71% TX CPU utilization
7.42% RX CPU utilization
36380 Mbps
IPv4, without remote checksum offload
12.40% TX CPU utilization
7.36% RX CPU utilization
36591 Mbps
TCP_RR
IPv4, with remote checksum offload
77.79% CPU utilization
91/144/216 90/95/99% latencies
1.95127e+06 tps
IPv4, without remote checksum offload
78.70% CPU utilization
89/152/297 90/95/99% latencies
1.95458e+06 tps
IPIP/GUE
TCP_STREAM
With remote checksum offload
10.30% TX CPU utilization
7.43% RX CPU utilization
36486 Mbps
Without remote checksum offload
12.47% TX CPU utilization
7.49% RX CPU utilization
36694 Mbps
TCP_RR
With remote checksum offload
77.80% CPU utilization
87/153/270 90/95/99% latencies
1.98735e+06 tps
Without remote checksum offload
77.98% CPU utilization
87/150/287 90/95/99% latencies
1.98737e+06 tps
SIT/GUE
TCP_STREAM
With remote checksum offload
9.68% TX CPU utilization
7.36% RX CPU utilization
35971 Mbps
Without remote checksum offload
12.95% TX CPU utilization
8.04% RX CPU utilization
36177 Mbps
TCP_RR
With remote checksum offload
79.32% CPU utilization
94/158/295 90/95/99% latencies
1.88842e+06 tps
Without remote checksum offload
80.23% CPU utilization
94/149/226 90/95/99% latencies
1.90338e+06 tps
VXLAN
TCP_STREAM
35.03% TX CPU utilization
20.85% RX CPU utilization
36230 Mbps
TCP_RR
77.36% CPU utilization
84/146/270 90/95/99% latencies
2.08063e+06 tps
We can also look at CPU time in csum_partial using perf (with bnx2x
setup). For GRE with TCP_STREAM I see:
With remote checksum offload
0.33% TX
1.81% RX
Without remote checksum offload
6.00% TX
0.51% RX
I suspect the fact that time in csum_partial noticably increases
with remote checksum offload for RX is due to taking the cache miss on
the encapsulated header in that function. By similar reasoning, if on
the TX side the packet were not in cache (say we did a splice from a
file whose data was never touched by the CPU) the CPU savings for TX
would probably be more pronounced.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Add processing of the remote checksum offload option in both the normal
path as well as the GRO path. The implements patching the affected
checksum to derive the offloaded checksum.
Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add if_tunnel flag TUNNEL_ENCAP_FLAG_REMCSUM to configure
remote checksum offload on an IP tunnel. Add logic in gue_build_header
to insert remote checksum offload option.
Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Define a private flag for remote checksun offload as well as a length
for the option.
Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add a new GSO type, SKB_GSO_TUNNEL_REMCSUM, which indicates remote
checksum offload being done (in this case inner checksum must not
be offloaded to the NIC).
Added logic in __skb_udp_tunnel_segment to handle remote checksum
offload case.
Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add functions and basic definitions for processing standard flags,
private flags, and control messages. This includes definitions
to compute length of optional fields corresponding to a set of flags.
Flag validation is in validate_gue_flags function. This checks for
unknown flags, and that length of optional fields is <= length
in guehdr hlen.
Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In __skb_udp_tunnel_segment if outer UDP checksums are enabled and
ip_summed is not already CHECKSUM_PARTIAL, set up checksum offload
if device features allow it.
Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Move fou_build_header out of ip_tunnel.c and into fou.c splitting
it up into fou_build_header, gue_build_header, and fou_build_udp.
This allows for other users for TX of FOU or GUE. Change ip_tunnel_encap
to call fou_build_header or gue_build_header based on the tunnel
encapsulation type. Similarly, added fou_encap_hlen and gue_encap_hlen
functions which are called by ip_encap_hlen. New net/fou.h has
prototypes and defines for this.
Added NET_FOU_IP_TUNNELS configuration. When this is set, IP tunnels
can use FOU/GUE and fou module is also selected.
Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Giuseppe Cavallaro says:
====================
stmmac: review driver Koptions
Recently many Koption options have been added to have new glue logic on several
platforms.
The main goal behind this work is to guarantee that the driver built
fine on all the branches where it is present independently of which
glue logic is selected.
IMHO, it is better to remove all the not necessary Koption(s) that can hide
build problems when something changes in the driver and especially when
the DT compatibility allows us to manage all the platform data.
I compiled the driver w/o any issue on net-next Git for:
x86, arm and sh4.
In case of there are build problems on some repos now it will be
easy to catch them and cherry-pick patches from mainstream.
For sure, do not hesitate to contact me in case of issue.
Also this set removes STMMAC_DEBUG_FS and BUS_MODE_DA. The latter is useless
and the former can be replaced by DEBUG_FS (always to make safe the build).
V2: patch-set re-based on top of the latest updates for net-next
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
This is a very old and often unused option to configure
a bit in a register inside the DMA. This support should
not stay under Koption and should be extended for new chips too.
This will be do later maybe via device-tree parameters.
Also no performance impact when remove this setting on STi platforms.
Signed-off-by: Giuseppe Cavallaro <peppe.cavallaro@st.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
the STMMAC_DEBUG_FS Koption is now removed from the
driver configuration and this support will be built
by default when DEBUG_FS is present. This can also be
useful on building driver verification.
Signed-off-by: Giuseppe Cavallaro <peppe.cavallaro@st.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch removes all the Koptions added to build the glue-logic files
for all different architectures: DWMAC_MESON, DWMAC_SUNXI, DWMAC_STI ...
Nowadays the stmmac needs to be compiled on several platforms; in some
case it very convenient to guarantee that its build is always completed
with success on all the branches where the driver is present.
Signed-off-by: Giuseppe Cavallaro <peppe.cavallaro@st.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When a sysfs binary file is asked to be removed, it is found by
attribute name, so strictly speaking this change is not a fix, but
just in case when attribute name is changed in the driver or sysfs
internals are changed, it might be better to remove the previously
created file using right the same binary attribute.
Signed-off-by: Vladimir Zapolskiy <vz@mleia.com>
Cc: Florian Fainelli <f.fainelli@gmail.com>
Cc: David S. Miller <davem@davemloft.net>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The "else" block is on several lines and use bracket.
Signed-off-by: Florent Fourcot <florent.fourcot@enst-bretagne.fr>
Signed-off-by: David S. Miller <davem@davemloft.net>
Florian Westphal says:
====================
net: allow setting ecn via routing table
Here is v4 of the patchset, its exactly the same as v3 except in patch3/3
where I added the missing 'const' qualifier to a function argument that
Eric spotted during review.
I preserved Erics Acks so that he doesn't have to resend them.
v3 cover letter:
When using syn cookies, then do not simply trust that the echoed timestamp
was not modified to make sure that ecn is not turned on magically when it
is disabled on the host.
The first two patches, which were not part of earlier series, prepare
the cookie code for the ecn route metrics change by allowing is to
more easily use the existing dst object for ecn validation.
The 3rd patch adds the ecn route metric feature support.
It is almost the same as in v2, except that we'll now also test the
dst_features when decoding a syn cookie timestamp that indicates ecn support.
These three patches then allow turning on explicit congestion notification
based on the destination network.
For example, assuming the default tcp_ecn sysctl '2', the following will
enable ecn (tcp_ecn=1 behaviour, i.e. request ecn to be enabled for a
tcp connection) for all connections to hosts inside the 192.168.2/24 network:
ip route change 192.168.2.0/24 dev eth0 features ecn
Having a more fine-grained per-route setting can be beneficial for
various reasons, for example 1) within data centers, or 2) local ISPs
may deploy ECN support for their own video/streaming services [1], etc.
Joint work with Daniel Borkmann, feature suggested by Hannes Frederic Sowa.
The patch to enable this in iproute2 will be posted shortly, it is currently
also available here:
http://git.breakpoint.cc/cgit/fw/iproute2.git/commit/?h=iproute_features&id=8843d2d8973fb81c78a7efe6d42e3a17d739003e
[1] http://www.ietf.org/proceedings/89/slides/slides-89-tsvarea-1.pdf, p.15
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch allows to set ECN on a per-route basis in case the sysctl
tcp_ecn is not set to 1. In other words, when ECN is set for specific
routes, it provides a tcp_ecn=1 behaviour for that route while the rest
of the stack acts according to the global settings.
One can use 'ip route change dev $dev $net features ecn' to toggle this.
Having a more fine-grained per-route setting can be beneficial for various
reasons, for example, 1) within data centers, or 2) local ISPs may deploy
ECN support for their own video/streaming services [1], etc.
There was a recent measurement study/paper [2] which scanned the Alexa's
publicly available top million websites list from a vantage point in US,
Europe and Asia:
Half of the Alexa list will now happily use ECN (tcp_ecn=2, most likely
blamed to commit 255cac91c3 ("tcp: extend ECN sysctl to allow server-side
only ECN") ;)); the break in connectivity on-path was found is about
1 in 10,000 cases. Timeouts rather than receiving back RSTs were much
more common in the negotiation phase (and mostly seen in the Alexa
middle band, ranks around 50k-150k): from 12-thousand hosts on which
there _may_ be ECN-linked connection failures, only 79 failed with RST
when _not_ failing with RST when ECN is not requested.
It's unclear though, how much equipment in the wild actually marks CE
when buffers start to fill up.
We thought about a fallback to non-ECN for retransmitted SYNs as another
global option (which could perhaps one day be made default), but as Eric
points out, there's much more work needed to detect broken middleboxes.
Two examples Eric mentioned are buggy firewalls that accept only a single
SYN per flow, and middleboxes that successfully let an ECN flow establish,
but later mark CE for all packets (so cwnd converges to 1).
[1] http://www.ietf.org/proceedings/89/slides/slides-89-tsvarea-1.pdf, p.15
[2] http://ecn.ethz.ch/
Joint work with Daniel Borkmann.
Reference: http://thread.gmane.org/gmane.linux.network/335797
Suggested-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Acked-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Daniel Borkmann <dborkman@redhat.com>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
The function cookie_check_timestamp(), both called from IPv4/6 context,
is being used to decode the echoed timestamp from the SYN/ACK into TCP
options used for follow-up communication with the peer.
We can remove ECN handling from that function, split it into a separate
one, and simply rename the original function into cookie_decode_options().
cookie_decode_options() just fills in tcp_option struct based on the
echoed timestamp received from the peer. Anything that fails in this
function will actually discard the request socket.
While this is the natural place for decoding options such as ECN which
commit 172d69e63c ("syncookies: add support for ECN") added, we argue
that in particular for ECN handling, it can be checked at a later point
in time as the request sock would actually not need to be dropped from
this, but just ECN support turned off.
Therefore, we split this functionality into cookie_ecn_ok(), which tells
us if the timestamp indicates ECN support AND the tcp_ecn sysctl is enabled.
This prepares for per-route ECN support: just looking at the tcp_ecn sysctl
won't be enough anymore at that point; if the timestamp indicates ECN
and sysctl tcp_ecn == 0, we will also need to check the ECN dst metric.
This would mean adding a route lookup to cookie_check_timestamp(), which
we definitely want to avoid. As we already do a route lookup at a later
point in cookie_{v4,v6}_check(), we can simply make use of that as well
for the new cookie_ecn_ok() function w/o any additional cost.
Joint work with Daniel Borkmann.
Acked-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Daniel Borkmann <dborkman@redhat.com>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
Was a bit more difficult to read than needed due to magic shifts;
add defines and document the used encoding scheme.
Joint work with Daniel Borkmann.
Acked-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Daniel Borkmann <dborkman@redhat.com>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
else is unnecessary after return 0 in __udp4_lib_rcv()
Signed-off-by: Fabian Frederick <fabf@skynet.be>
Signed-off-by: David S. Miller <davem@davemloft.net>
remove __inline__ / inline and let compiler decide what to do
with static functions
Inspired-by: "David S. Miller" <davem@davemloft.net>
Signed-off-by: Fabian Frederick <fabf@skynet.be>
Signed-off-by: David S. Miller <davem@davemloft.net>
Yaogong replaces TCP out of order receive queue by an RB tree.
As netem already does a private skb->{next/prev/tstamp} union
with a 'struct rb_node', lets do this in a cleaner way.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Yaogong Wang <wygivan@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Jeff Kirsher says:
====================
Intel Wired LAN Driver Updates 2014-11-03
This series contains updates to i40e and i40evf.
Akeem adds a check for i40e so that flow director flush and reinit are
not done when flow director is not enabled.
Mitch fixes the i40evf driver to properly handle multiple admin queue
messages, by reinit the msg_size field each time we go through the loop.
Without this, we may receive truncated messages due to the firmware
thinking we have insufficient buffer size. Also fixes the link checking
logic to only check the carrier state if the interface is actually
open, which allows link changes to be reported correctly without spamming
the VFs. Updates i40e to inset the VSI ID in the QTX_CTL register
when configuring queues for VMDq VSIs.
Paul adds support for 10G-base-T in i40evf.
Jesse fixes i40e where the call to irq_dynamic_disable() was turning off
the interrupt completely when trying to set ITR to 0 (for lowest
moderation).
Shannon removes debugfs dump stats function, since it was not being
kept up-to-date and was redundant with the ethtool output. Also, scales
back the LAN MSIx usage to force queue/vector sharing and leave some
vectors for Flow Director, VMDq, etc. when there are more cores than
vectors available to the PF. Cleans up the error reporting for
get_lump() resource tracking errors. Also adds a check for the
debug module parameter earlier to be able to catch the early configuration
phase admin queue messages.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
this is check for dev is unnecessary, as we are already checking dev
after allocating it via alloc_netdev, and jumping to label: out
if it is NULL.
Signed-off-by: Sudip Mukherjee <sudip@vectorindia.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Or Gerlitz says:
====================
Mellanox ethernet driver update Oct-30-2014
The 1st patch from Saeed fixes a bug in the last net-next batch where
a VF could get access to set port configuration, the next patch from Amir
fixes a race in the port VPI logic. Next are two performance patches from Ido.
The patch to add checksum complete status on GRE and such packets was
preceded with a patch that converted the driver to only use napi_gro_receive
vs. the current code which goes through napi_gro_frags on it's usual track.
Eric D. has some thoughts and suggestions on that change for which we
want to take the time and consider, so for the time being dropped that
patch and the ones that depend on it.
Changes from V0:
- have the caller to provide the __GFP_COLD hint to the service function
- dropped the patch that changes the GRO logic and the subsequent dependent
patches.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Hannes Frederic Sowa
David S. Miller
Lendacky, Thomas
WANG Cong
Tom Herbert
Giuseppe CAVALLARO
Vladimir Zapolskiy
Fabian Frederick
Florent Fourcot
Florian Westphal
Eric Dumazet
Sudip Mukherjee
Denis Kirjanov