David S. Miller [Mon, 15 Sep 2014 18:41:12 +0000 (14:41 -0400)]
Merge branch 'tcpflags'
Eric Dumazet says:
====================
tcp: no longer keep around headers in input path
Looking at tcp_try_coalesce() I was wondering why I did :
if (tcp_hdr(from)->fin)
return false;
The answer would be to allow the aggregation, if we simply OR the FIN and PSH
flags eventually present in @from to @to packet. (Note a change is also
needed in skb_try_coalesce() to avoid calling skb_put() with 0 len)
Then, looking at tcp_recvmsg(), I realized we access tcp_hdr(skb)->syn
(and maybe tcp_hdr(skb)->fin) for every packet we process from socket
receive queue.
We have to understand TCP flags are cold in cpu caches most of the time
(assuming TCP timestamps, and that application calls recvmsg() a long
time after incoming packet was processed), and bringing a whole
cache line only to access one bit is not very nice.
It would make sense to use in TCP input path TCP_SKB_CB(skb)->tcp_flags
as we do in output path.
This saves one cache line miss, and TCP tcp_collapse() can avoid dealing
with the headers.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Eric Dumazet [Mon, 15 Sep 2014 11:19:53 +0000 (04:19 -0700)]
tcp: do not copy headers in tcp_collapse()
tcp_collapse() wants to shrink skb so that the overhead is minimal.
Now we store tcp flags into TCP_SKB_CB(skb)->tcp_flags, we no longer
need to keep around full headers.
Whole available space is dedicated to the payload.
Signed-off-by: Eric Dumazet <edumazet@google.com> Acked-by: Neal Cardwell <ncardwell@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Eric Dumazet [Mon, 15 Sep 2014 11:19:51 +0000 (04:19 -0700)]
tcp: use TCP_SKB_CB(skb)->tcp_flags in input path
Input path of TCP do not currently uses TCP_SKB_CB(skb)->tcp_flags,
which is only used in output path.
tcp_recvmsg(), looks at tcp_hdr(skb)->syn for every skb found in receive queue,
and its unfortunate because this bit is located in a cache line right before
the payload.
We can simplify TCP by copying tcp flags into TCP_SKB_CB(skb)->tcp_flags.
This patch does so, and avoids the cache line miss in tcp_recvmsg()
Following patches will
- allow a segment with FIN being coalesced in tcp_try_coalesce()
- simplify tcp_collapse() by not copying the headers.
Signed-off-by: Eric Dumazet <edumazet@google.com> Acked-by: Neal Cardwell <ncardwell@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
net: bpf: correctly handle errors in sk_attach_filter()
Commit "net: bpf: make eBPF interpreter images read-only" has changed bpf_prog
to be vmalloc()ed but never handled some of the errors paths of the old code.
On error within sk_attach_filter (which userspace can easily trigger), we'd
kfree() the vmalloc()ed memory, and leak the internal bpf_work_struct.
Signed-off-by: Sasha Levin <sasha.levin@oracle.com> Acked-by: Daniel Borkmann <dborkman@redhat.com> Acked-by: Hannes Frederic Sowa <hannes@stressinduktion.org> Signed-off-by: David S. Miller <davem@davemloft.net>
Alexander Duyck [Sat, 13 Sep 2014 06:12:46 +0000 (23:12 -0700)]
netdevice: Support DSA tagging when DSA is built as a module
This change corrects an error seen when DSA tagging is built as a module.
Without this change it is not possible to get XDSA tagged frames as the
test for tagging is stripped by the #ifdef check.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com> Acked-by: Florian Fainelli <f.fainelli@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Bo Shen [Fri, 12 Sep 2014 23:57:49 +0000 (01:57 +0200)]
net/macb: Add hardware revision information during probe
Print the IP revision when probing.
Signed-off-by: Bo Shen <voice.shen@atmel.com> Signed-off-by: Nicolas Ferre <nicolas.ferre@atmel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Sat, 13 Sep 2014 21:32:29 +0000 (17:32 -0400)]
Merge branch 'fec-next'
Frank Li says:
====================
net: fec: imx6sx multiqueue support
These patches enable i.MX6SX multi queue support.
i.MX6SX support 3 queue and AVB feature.
Change from v3 to v4
- use "unsigned int" instead of "unsigned"
Change from v2 to v3
- fixed alignment requirement for ARM and NO-ARM platform
Change from v1 to v2.
- Change num_tx_queue to unsigned int
- Avoid block non-dt platform
- remove call netif_set_real_num_rx_queues
- seperate multi queue patch two part, one is tx and rx handle, with fixed queue 0
then other one is initilized multiqueue
- use two difference alignment for tx and rx path
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Frank Li [Fri, 12 Sep 2014 21:00:56 +0000 (05:00 +0800)]
ARM: Documentation: Update fec dts binding doc
This patch update fec devicetree binding doc that add Optional
properties "fsl,num-tx-queues" and "fsl,num-rx-queues".
Signed-off-by: Fugang Duan <B38611@freescale.com> Signed-off-by: Frank Li <Frank.Li@freescale.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Fugang Duan <B38611@freescale.com> Signed-off-by: Frank Li <Frank.Li@freescale.com> Signed-off-by: David S. Miller <davem@davemloft.net>
net: fec: change FEC alignment according to i.mx6 sx requirement
i.MX6 SX change FEC alignment requirement.
i.MX6 SX change internal bus from AHB to AXI.
It require RX buffer must be 64 bytes alignment.
And remove TX buffer alignment requirement.
Signed-off-by: Fugang Duan <B38611@freescale.com> Signed-off-by: Frank Li <Frank.Li@freescale.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Add compatible string "fsl,imx6sx-fec" for i.MX6SX.
Signed-off-by: Fugang Duan <B38611@freescale.com> Signed-off-by: Frank Li <Frank.Li@freescale.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Frank Li [Fri, 12 Sep 2014 21:00:52 +0000 (05:00 +0800)]
net: fec: add enet-avb IP support
i.MX6SX Enet-AVB support 3 tx queues, 3 rx queues.
For tx queues: ring 0 -> best effort
ring 1 -> Class A
ring 2 -> Class B
For rx queues:
ring 0 -> best effort
ring 1 -> receive VLAN packet with classification match
ring 2 -> receive VLAN packet with classification match
Add enet-avb IP multiqueue support for the driver.
Signed-off-by: Fugang Duan <B38611@freescale.com> Signed-off-by: Frank Li <Frank.Li@freescale.com> Signed-off-by: David S. Miller <davem@davemloft.net>
net:fec: Disable enet-avb MAC instead of reset MAC
For i.MX6SX enet use AXI bus, reset MAC will make system bus dead
if ENET-AXI bus has pending access (AHB bus should not have such issue).
So, disable enet with AVB MAC instead of reset MAC itself.
Signed-off-by: Fugang Duan <B38611@freescale.com> Signed-off-by: Frank Li <Frank.Li@freescale.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Frank Li [Fri, 12 Sep 2014 21:00:50 +0000 (05:00 +0800)]
net: fec: init multi queue date structure
initilized all queues according to queue number get from DT file.
Signed-off-by: Frank Li <Frank.Li@freescale.com> Signed-off-by: Duan Fugang <B38611@freescale.com> Signed-off-by: David S. Miller <davem@davemloft.net>
By default, the tx/rx queue number is 1, user can config the queue number
at DTS file like this:
fsl,num-tx-queues=<3>;
fsl,num-rx-queues=<3>
Since i.MX6SX enet-AVB IP support multi queues, so use multi queues
interface to allocate and set up an Ethernet device.
Signed-off-by: Fugang Duan <B38611@freescale.com> Signed-off-by: Frank Li <Frank.Li@freescale.com> Signed-off-by: David S. Miller <davem@davemloft.net>
net: fec: change data structure to support multiqueue
This patch just change data structure to support multi-queue.
Only 1 queue enabled.
Ethernet multiqueue mechanism can improve performance in SMP system.
For single hw queue, multiqueue can balance cpu loading.
For multi hw queues, multiple cores can process network packets in parallel,
and refer the article for the detail advantage for multiqueue:
http://vger.kernel.org/~davem/davem_nyc09.pdf
Signed-off-by: Fugang Duan <B38611@freescale.com> Signed-off-by: Frank Li <frank.li@freescale.com> Signed-off-by: David S. Miller <davem@davemloft.net>
net:fec: add enet AVB feature macro define for imx6sx
Add enet AVB feature macro define for imx6sx.
Signed-off-by: Fugang Duan <B38611@freescale.com> Signed-off-by: Frank Li <Frank.Li@freescale.com> Signed-off-by: David S. Miller <davem@davemloft.net>
net:fec: add enet refrence clock for i.MX 6SX chip
i.MX6sx enet has below clocks for user config:
clk_ipg: ipg_clk_s, ipg_clk_mac0_s, 66Mhz
clk_ahb: enet system clock, it is enet AXI clock for imx6sx.
For imx6sx, it alos is the clock source of interrupt coalescing.
The clock range: 200Mhz ~ 266Mhz.
clk_ref: refrence clock for tx and rx. For imx6sx enet RGMII mode,
the refrence clock is 125Mhz coming from internal PLL or external.
In i.MX6sx-arm2 board, the clock is from internal PLL.
clk_ref is optional, depends on board.
clk_enet_out: The clock can be output from internal PLL. It can supply 50Mhz
clock for phy. clk_enet_out is optional, depends on chip and board.
clk_ptp: 1588 ts clock. It is optional, depends on chip.
The patch add clk_ref to distiguish the different clocks.
Signed-off-by: Fugang Duan <B38611@freescale.com> Signed-off-by: Frank Li <Frank.Li@freescale.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Andrew Lunn [Fri, 12 Sep 2014 21:58:44 +0000 (23:58 +0200)]
net: DSA: Marvell mv88e6171 switch driver
This is the Marvell driver with some cleanups by Claudio Leite
and myself.
Signed-off-by: Andrew Lunn <andrew@lunn.ch> Cc: Claudio Leite <leitec@staticky.com> Signed-off-by: Claudio Leite <leitec@staticky.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Kalesh AP [Fri, 12 Sep 2014 12:09:21 +0000 (17:39 +0530)]
be2net: enable PCIe error reporting on VFs too
Currently PCIe error reporting is enabled only on PFs. This patch enables
this feature on VFs too as Lancer VFs support it.
Signed-off-by: Kalesh AP <kalesh.purayil@emulex.com> Signed-off-by: Sathya Perla <sathya.perla@emulex.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Kalesh AP [Fri, 12 Sep 2014 12:09:20 +0000 (17:39 +0530)]
be2net: send a max of 8 EQs to be_cmd_modify_eqd() on Lancer
The MODIFY_EQ_DELAY FW cmd on Lancer is supported for a max of 8 EQs per cmd.
Signed-off-by: Kalesh AP <kalesh.purayil@emulex.com> Signed-off-by: Sathya Perla <sathya.perla@emulex.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Report the ethtool port-type/supported/advertising values based on the
cable_type for QSFP and SFP+ interfaces. The cable_type is parsed from
the transceiver data fetched from the FW.
Signed-off-by: Ravikumar Nelavelli <ravikumar.nelavelli@emulex.com> Signed-off-by: Suresh Reddy <Suresh.Reddy@emulex.com> Signed-off-by: Sathya Perla <sathya.perla@emulex.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Mark Leonard [Fri, 12 Sep 2014 12:09:18 +0000 (17:39 +0530)]
be2net: add ethtool "-m" option support
This patch adds support for the dump-module-eeprom and module-info
ethtool options.
Signed-off-by: Mark Leonard <mark.leonard@emulex.com> Signed-off-by: Suresh Reddy <Suresh.Reddy@emulex.com> Signed-off-by: Sathya Perla <sathya.perla@emulex.com> Signed-off-by: David S. Miller <davem@davemloft.net>
In some configurations the FW doesn't allow changing flow control settings
of a link. Unless a v1 version of the SET_FLOW_CONTROL cmd is used, the FW
doesn't report an error to the driver.
Signed-off-by: Suresh Reddy <Suresh.Reddy@emulex.com> Signed-off-by: Sathya Perla <sathya.perla@emulex.com> Signed-off-by: David S. Miller <davem@davemloft.net>
In the RX path, the driver currently consumes upto 64 (budget) packets in
one NAPI sweep. When the size of the packet received is larger than a
fragment size (2K), more than one fragment is consumed for each packet.
As the driver currently posts a max of 64 fragments, all the consumed
fragments may not be replenished. This can cause avoidable drops in RX path.
This patch fixes this by posting a max(consumed_frags, 64) frags. This is
done only when there are atleast 64 free slots in the RXQ.
Signed-off-by: Ajit Khaparde <ajit.khaparde@emulex.com> Signed-off-by: Kalesh AP <kalesh.purayil@emulex.com> Signed-off-by: Sathya Perla <sathya.perla@emulex.com> Signed-off-by: David S. Miller <davem@davemloft.net>
This patch fixes the following minor issues with log messages in be2net:
1) Period is not required at the end of log message.
2) Remove "Unknown grp5 event" logs to reduce noise. The driver can safely
ignore async events from FW it's not interested in.
3) Reword a log message for better readability to say that SRIOV
"is disabled" rather than "not supported".
Signed-off-by: Vasundhara Volam <vasundhara.volam@emulex.com> Signed-off-by: Sathya Perla <sathya.perla@emulex.com> Signed-off-by: David S. Miller <davem@davemloft.net>
net: filter: constify detection of pkt_type_offset
Currently we have 2 pkt_type_offset functions doing the same thing and
spread across the architecture files. Remove those and replace them
with a PKT_TYPE_OFFSET macro helper which gets the constant value from a
zero sized sk_buff member right in front of the bitfield with offsetof.
This new offset marker does not change size of struct sk_buff.
Cc: Eric Dumazet <eric.dumazet@gmail.com> Cc: Markos Chandras <markos.chandras@imgtec.com> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: Daniel Borkmann <dborkman@redhat.com> Cc: Alexei Starovoitov <alexei.starovoitov@gmail.com> Signed-off-by: Denis Kirjanov <kda@linux-powerpc.org> Signed-off-by: Hannes Frederic Sowa <hannes@stressinduktion.org> Acked-by: Alexei Starovoitov <ast@plumgrid.com> Acked-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Now that we introduced an additional multiplexing/demultiplexing layer
with commit 3e8a72d1dae37 ("net: dsa: reduce number of protocol hooks")
that lives within the DSA code, we no longer need to have a given switch
driver tag_protocol be an actual ethertype value, instead, we can
replace it with an enum: dsa_tag_protocol.
Do this replacement in the drivers, which allows us to get rid of the
cpu_to_be16()/htons() dance, and remove ETH_P_BRCMTAG since we do not
need it anymore.
Suggested-by: Alexander Duyck <alexander.duyck@gmail.com> Signed-off-by: Florian Fainelli <f.fainelli@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
net: stmmac: fix return value check in socfpga_dwmac_parse_data()
In case of error, the function devm_ioremap_resource() returns
ERR_PTR() and never returns NULL. The NULL test in the return
value check should be replaced with IS_ERR().
Signed-off-by: Wei Yongjun <yongjun_wei@trendmicro.com.cn> Signed-off-by: David S. Miller <davem@davemloft.net>
WANG Cong [Thu, 11 Sep 2014 22:07:16 +0000 (15:07 -0700)]
ipv6: exit early in addrconf_notify() if IPv6 is disabled
If IPv6 is explicitly disabled before the interface comes up,
it makes no sense to continue when it comes up, even just
print a message.
(I am not sure about other cases though, so I prefer not to touch)
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com> Acked-by: Hannes Frederic Sowa <hannes@stressinduktion.org> Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Sat, 13 Sep 2014 20:38:53 +0000 (16:38 -0400)]
Merge branch 'ipv6-cleanups'
Cong Wang says:
====================
ipv6: clean up locking code in anycast and mcast
This patchset cleans up the locking code in anycast.c and mcast.c
and makes the refcount code more readable.
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
v1 -> v2:
* refactor some code and make it in a separated patch
* update comments
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Sat, 13 Sep 2014 20:29:57 +0000 (16:29 -0400)]
Merge branch 'bonding-next'
Nikolay Aleksandrov says:
====================
bonding: get rid of curr_slave_lock
This is the second patch-set dealing with bond locking and the purpose here
is to convert curr_slave_lock into a spinlock called "mode_lock" which can
be used in the various modes for their specific needs. The first three
patches cleanup the use of curr_slave_lock and prepare it for the
conversion which is done in patch 4 and then the modes that were using
their own locks are converted to use the new "mode_lock" giving us the
opportunity to remove their locks.
This patch-set has been tested in each mode by running enslave/release of
slaves in parallel with traffic transmission and miimon=1 i.e. running
all the time. In fact this lead to the discovery of a subtle bug related to
RCU which will be fixed in -net.
Also did an allmodconfig test just in case :-)
v2: fix bond_3ad_state_machine_handler's use of mode_lock and
curr_slave_lock
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Now that locks have been removed, remove some unnecessary comments and
adjust others to reflect reality. Also add a comment to "mode_lock" to
describe its current users and give a brief summary why they need it.
Signed-off-by: Nikolay Aleksandrov <nikolay@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Now that we have bond->mode_lock, we can remove the state_machine_lock
and use it in its place. There're no fast paths requiring the per-port
spinlocks so it should be okay to consolidate them into mode_lock.
Also move it inside the unbinding function as we don't want to expose
mode_lock outside of the specific modes.
Suggested-by: Jay Vosburgh <jay.vosburgh@canonical.com> Signed-off-by: Nikolay Aleksandrov <nikolay@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
The ALB/TLB specific spinlocks are no longer necessary as we now have
bond->mode_lock for this purpose, so convert them and remove them from
struct alb_bond_info.
Also remove the unneeded lock/unlock functions and use spin_lock/unlock
directly.
Suggested-by: Jay Vosburgh <jay.vosburgh@canonical.com> Signed-off-by: Nikolay Aleksandrov <nikolay@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
bonding: convert curr_slave_lock to a spinlock and rename it
curr_slave_lock is now a misleading name, a much better name is
mode_lock as it'll be used for each mode's purposes and it's no longer
necessary to use a rwlock, a simple spinlock is enough.
Suggested-by: Jay Vosburgh <jay.vosburgh@canonical.com> Signed-off-by: Nikolay Aleksandrov <nikolay@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Mostly all users of curr_slave_lock already have RTNL as we've discussed
previously so there's no point in using it, the one case where the lock
must stay is the 3ad code, in fact it's the only one.
It's okay to remove it from bond_do_fail_over_mac() as it's called with
RTNL and drops the curr_slave_lock anyway.
bond_change_active_slave() is one of the main places where
curr_slave_lock was used, it's okay to remove it as all callers use RTNL
these days before calling it, that's why we move the ASSERT_RTNL() in
the beginning to catch any potential offenders to this rule.
The RTNL argument actually applies to all of the places where
curr_slave_lock has been removed from in this patch.
Also remove the unnecessary bond_deref_active_protected() macro and use
rtnl_dereference() instead.
Signed-off-by: Nikolay Aleksandrov <nikolay@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
First in rlb_teach_disabled_mac_on_primary() it's okay to remove
curr_slave_lock as all callers except bond_alb_monitor() already hold
RTNL, and in case bond_alb_monitor() is executing we can at most have a
period with bad throughput (very unlikely though).
In bond_alb_monitor() it's okay to remove the read_lock as the slave
list is walked with RCU and the worst that could happen is another
transmitter at the same time and thus for a period which currently is 10
seconds (bond_alb.h: BOND_ALB_LP_TICKS).
And bond_alb_handle_active_change() is okay because it's always called
with RTNL. Removed the ASSERT_RTNL() because it'll be inserted in the
parent function in a following patch.
Signed-off-by: Nikolay Aleksandrov <nikolay@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Remove the read_lock in bond_3ad_lacpdu_recv() since when the slave is
being released its rx_handler is removed before 3ad unbind, so even if
packets arrive, they won't see the slave in an inconsistent state.
Signed-off-by: Nikolay Aleksandrov <nikolay@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Rusty Russell [Thu, 11 Sep 2014 00:47:38 +0000 (10:17 +0930)]
virtio_ring: unify direct/indirect code paths.
virtqueue_add() populates the virtqueue descriptor table from the sgs
given. If it uses an indirect descriptor table, then it puts a single
descriptor in the descriptor table pointing to the kmalloc'ed indirect
table where the sg is populated.
Previously vring_add_indirect() did the allocation and the simple
linear layout. We replace that with alloc_indirect() which allocates
the indirect table then chains it like the normal descriptor table so
we can reuse the core logic.
This slows down pktgen by less than 1/2 a percent (which uses direct
descriptors), as well as vring_bench, but it's far neater.
Rusty Russell [Thu, 11 Sep 2014 00:47:36 +0000 (10:17 +0930)]
virtio_net: pass well-formed sgs to virtqueue_add_*()
This is the only driver which doesn't hand virtqueue_add_inbuf and
virtqueue_add_outbuf a well-formed, well-terminated sg. Fix it,
so we can make virtio_add_* simpler.
David S. Miller [Sat, 13 Sep 2014 16:43:24 +0000 (12:43 -0400)]
Merge branch 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/jkirsher/net-next
Jeff Kirsher says:
====================
Intel Wired LAN Driver Updates 2014-09-12
This series contains updates to e1000, ixgbe and ixgbevf.
Mark provide two fixes to reduce compile warnings produce by ixgbe
and ixgbevf.
Alex provides two patches for ixgbe, first removes the receive buffer
allocation at the end of the ixgbe_clean_rx_irq(). The reason for
removing this is to avoid the extra latency introduced by the MMIO write.
Second patch addresses several issues in the current ixgbe implementation
of busy poll sockets. It was possible for frames to be delivered out of
order if they were held in GRO, so address this by flushing the GRO
buffers before releasing the q_vector back to the idle state. Also, we
were having to take a spinlock on changing the state to and from idle,
so to resolve this, replaced the state value with an atomic and use
atomic_cmpxchg to change the value from idle, and a simple atomic set
to restore it back to idle after we have acquired it. This allows us
to only use a locked operation on acquiring the vector without a need
for a locked operation to release it.
Florian Westphal provides several patches for e1000 which does some
cleanup and updating of the driver. Moved e1000_tbi_adjust_stats()
so that he could make the function static. Added a helper function
to deal with the tbi workaround that was located in 2 different
Rx clean functions. Added a e1000_rx_buffer struct for use on receive
since the transmit and receive have different requirements. Updates
e1000 to use napi_gro_frags API.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Sat, 13 Sep 2014 16:30:33 +0000 (12:30 -0400)]
Merge branch 'sched_rcu'
John Fastabend says:
====================
net/sched rcu classifiers and tcf
This series converts the tcf_proto usage to RCU.
This requires updating each classifier individually to handle the
new copy/update requirement and also to update the core list
traversals. This makes the assumption that updates to the tables
are infrequent in comparison to the packet per second being
classified. On a 10Gbps running near line rate we can easily
produce 12+ million packets per second so IMO this is a reasonable
assumption. The updates are serialized by RTNL.
I have done some basic testing on this series and do not see any
immediate splats or issues. The patch series has been running
on my dev systems for a month or so now and I've not seen any
issues. Although my configurations are not overly complicated.
My test cases at this point cover all the filters with a
tight loop to add/remove filters. Some basic estimator tests
where I add an estimator to the qdisc and verify the statistics
accurate using pktgen. And finally I have a small script to
exercise the 'tc actions' interface. Feel free to send me more
tests off list and I can run them.
This is prep work to drop the qdisc lock with the first
target being the ingress qdisc. To be done is making the
tc actions RCU safe and statistics per cpu. These patches
are in the works.
Comments:
- Checkpatch is still giving errors on some >80 char lines I know
about this. IMO the way to fix this is to restructure the sched
code to avoid being so heavily indented. But doing this here
bloats the patchset and anyways there are already lots of >80
chars in these files. I would prefer to keep the patches as is
but let me know if others think I should fix these and I will.
A follow up patch set could restructure the code and fix this
throughout the code blocks.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
John Fastabend [Sat, 13 Sep 2014 03:10:24 +0000 (20:10 -0700)]
net: sched: rcu'ify cls_bpf
This patch makes the cls_bpf classifier RCU safe. The tcf_lock
was being used to protect a list of cls_bpf_prog now this list
is RCU safe and updates occur with rcu_replace.
Signed-off-by: John Fastabend <john.r.fastabend@intel.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
John Fastabend [Sat, 13 Sep 2014 03:09:49 +0000 (20:09 -0700)]
net: sched: rcu'ify cls_rsvp
Signed-off-by: John Fastabend <john.r.fastabend@intel.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
John Fastabend [Sat, 13 Sep 2014 03:09:16 +0000 (20:09 -0700)]
net: sched: make cls_u32 lockless
Make cls_u32 classifier safe to run without holding lock. This patch
converts statistics that are kept in read section u32_classify into
per cpu counters.
This patch was tested with a tight u32 filter add/delete loop while
generating traffic with pktgen. By running pktgen on vlan devices
created on top of a physical device we can hit the qdisc layer
correctly. For ingress qdisc's a loopback cable was used.
for i in {1..100}; do
q=`echo $i%8|bc`;
echo -n "u32 tos: iteration $i on queue $q";
tc filter add dev p3p2 parent $p prio $i u32 match ip tos 0x10 0xff \
action skbedit queue_mapping $q;
sleep 1;
tc filter del dev p3p2 prio $i;
echo -n "u32 tos hash table: iteration $i on queue $q";
tc filter add dev p3p2 parent $p protocol ip prio $i handle 628: u32 divisor 1
tc filter add dev p3p2 parent $p protocol ip prio $i u32 \
match ip protocol 17 0xff link 628: offset at 0 mask 0xf00 shift 6 plus 0
tc filter add dev p3p2 parent $p protocol ip prio $i u32 \
ht 628:0 match ip tos 0x10 0xff action skbedit queue_mapping $q
sleep 2;
tc filter del dev p3p2 prio $i
sleep 1;
done
Signed-off-by: John Fastabend <john.r.fastabend@intel.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
John Fastabend [Sat, 13 Sep 2014 03:08:47 +0000 (20:08 -0700)]
net: sched: make cls_u32 per cpu
This uses per cpu counters in cls_u32 in preparation
to convert over to rcu.
Signed-off-by: John Fastabend <john.r.fastabend@intel.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
John Fastabend [Sat, 13 Sep 2014 03:08:20 +0000 (20:08 -0700)]
net: sched: RCU cls_tcindex
Make cls_tcindex RCU safe.
This patch addds a new RCU routine rcu_dereference_bh_rtnl() to check
caller either holds the rcu read lock or RTNL. This is needed to
handle the case where tcindex_lookup() is being called in both cases.
Signed-off-by: John Fastabend <john.r.fastabend@intel.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
John Fastabend [Sat, 13 Sep 2014 03:07:50 +0000 (20:07 -0700)]
net: sched: RCU cls_route
RCUify the route classifier. For now however spinlock's are used to
protect fastmap cache.
The issue here is the fastmap may be read by one CPU while the
cache is being updated by another. An array of pointers could be
one possible solution.
Signed-off-by: John Fastabend <john.r.fastabend@intel.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
John Fastabend [Sat, 13 Sep 2014 03:07:22 +0000 (20:07 -0700)]
net: sched: fw use RCU
RCU'ify fw classifier.
Signed-off-by: John Fastabend <john.r.fastabend@intel.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
John Fastabend [Sat, 13 Sep 2014 03:06:55 +0000 (20:06 -0700)]
net: sched: cls_flow use RCU
Signed-off-by: John Fastabend <john.r.fastabend@intel.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
John Fastabend [Sat, 13 Sep 2014 03:06:26 +0000 (20:06 -0700)]
net: sched: cls_cgroup use RCU
Make cgroup classifier safe for RCU.
Also drops the calls in the classify routine that were doing a
rcu_read_lock()/rcu_read_unlock(). If the rcu_read_lock() isn't held
entering this routine we have issues with deleting the classifier
chain so remove the unnecessary rcu_read_lock()/rcu_read_unlock()
pair noting all paths AFAIK hold rcu_read_lock.
If there is a case where classify is called without the rcu read lock
then an rcu splat will occur and we can correct it.
Signed-off-by: John Fastabend <john.r.fastabend@intel.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
John Fastabend [Sat, 13 Sep 2014 03:05:59 +0000 (20:05 -0700)]
net: sched: cls_basic use RCU
Enable basic classifier for RCU.
Dereferencing tp->root may look a bit strange here but it is needed
by my accounting because it is allocated at init time and needs to
be kfree'd at destroy time. However because it may be referenced in
the classify() path we must wait an RCU grace period before free'ing
it. We use kfree_rcu() and rcu_ APIs to enforce this. This pattern
is used in all the classifiers.
Also the hgenerator can be incremented without concern because it
is always incremented under RTNL.
Signed-off-by: John Fastabend <john.r.fastabend@intel.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
John Fastabend [Sat, 13 Sep 2014 03:05:27 +0000 (20:05 -0700)]
net: rcu-ify tcf_proto
rcu'ify tcf_proto this allows calling tc_classify() without holding
any locks. Updaters are protected by RTNL.
This patch prepares the core net_sched infrastracture for running
the classifier/action chains without holding the qdisc lock however
it does nothing to ensure cls_xxx and act_xxx types also work without
locking. Additional patches are required to address the fall out.
Signed-off-by: John Fastabend <john.r.fastabend@intel.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
John Fastabend [Sat, 13 Sep 2014 03:04:52 +0000 (20:04 -0700)]
net: qdisc: use rcu prefix and silence sparse warnings
Add __rcu notation to qdisc handling by doing this we can make
smatch output more legible. And anyways some of the cases should
be using rcu_dereference() see qdisc_all_tx_empty(),
qdisc_tx_chainging(), and so on.
Also *wake_queue() API is commonly called from driver timer routines
without rcu lock or rtnl lock. So I added rcu_read_lock() blocks
around netif_wake_subqueue and netif_tx_wake_queue.
Signed-off-by: John Fastabend <john.r.fastabend@intel.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
John Fastabend [Sat, 13 Sep 2014 03:10:24 +0000 (20:10 -0700)]
net: sched: rcu'ify cls_bpf
This patch makes the cls_bpf classifier RCU safe. The tcf_lock
was being used to protect a list of cls_bpf_prog now this list
is RCU safe and updates occur with rcu_replace.
Signed-off-by: John Fastabend <john.r.fastabend@intel.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
John Fastabend [Sat, 13 Sep 2014 03:09:49 +0000 (20:09 -0700)]
net: sched: rcu'ify cls_rsvp
Signed-off-by: John Fastabend <john.r.fastabend@intel.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
John Fastabend [Sat, 13 Sep 2014 03:09:16 +0000 (20:09 -0700)]
net: sched: make cls_u32 lockless
Make cls_u32 classifier safe to run without holding lock. This patch
converts statistics that are kept in read section u32_classify into
per cpu counters.
This patch was tested with a tight u32 filter add/delete loop while
generating traffic with pktgen. By running pktgen on vlan devices
created on top of a physical device we can hit the qdisc layer
correctly. For ingress qdisc's a loopback cable was used.
for i in {1..100}; do
q=`echo $i%8|bc`;
echo -n "u32 tos: iteration $i on queue $q";
tc filter add dev p3p2 parent $p prio $i u32 match ip tos 0x10 0xff \
action skbedit queue_mapping $q;
sleep 1;
tc filter del dev p3p2 prio $i;
echo -n "u32 tos hash table: iteration $i on queue $q";
tc filter add dev p3p2 parent $p protocol ip prio $i handle 628: u32 divisor 1
tc filter add dev p3p2 parent $p protocol ip prio $i u32 \
match ip protocol 17 0xff link 628: offset at 0 mask 0xf00 shift 6 plus 0
tc filter add dev p3p2 parent $p protocol ip prio $i u32 \
ht 628:0 match ip tos 0x10 0xff action skbedit queue_mapping $q
sleep 2;
tc filter del dev p3p2 prio $i
sleep 1;
done
Signed-off-by: John Fastabend <john.r.fastabend@intel.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
John Fastabend [Sat, 13 Sep 2014 03:08:47 +0000 (20:08 -0700)]
net: sched: make cls_u32 per cpu
This uses per cpu counters in cls_u32 in preparation
to convert over to rcu.
Signed-off-by: John Fastabend <john.r.fastabend@intel.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
John Fastabend [Sat, 13 Sep 2014 03:08:20 +0000 (20:08 -0700)]
net: sched: RCU cls_tcindex
Make cls_tcindex RCU safe.
This patch addds a new RCU routine rcu_dereference_bh_rtnl() to check
caller either holds the rcu read lock or RTNL. This is needed to
handle the case where tcindex_lookup() is being called in both cases.
Signed-off-by: John Fastabend <john.r.fastabend@intel.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
John Fastabend [Sat, 13 Sep 2014 03:07:50 +0000 (20:07 -0700)]
net: sched: RCU cls_route
RCUify the route classifier. For now however spinlock's are used to
protect fastmap cache.
The issue here is the fastmap may be read by one CPU while the
cache is being updated by another. An array of pointers could be
one possible solution.
Signed-off-by: John Fastabend <john.r.fastabend@intel.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
John Fastabend [Sat, 13 Sep 2014 03:07:22 +0000 (20:07 -0700)]
net: sched: fw use RCU
RCU'ify fw classifier.
Signed-off-by: John Fastabend <john.r.fastabend@intel.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
John Fastabend [Sat, 13 Sep 2014 03:06:55 +0000 (20:06 -0700)]
net: sched: cls_flow use RCU
Signed-off-by: John Fastabend <john.r.fastabend@intel.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
John Fastabend [Sat, 13 Sep 2014 03:06:26 +0000 (20:06 -0700)]
net: sched: cls_cgroup use RCU
Make cgroup classifier safe for RCU.
Also drops the calls in the classify routine that were doing a
rcu_read_lock()/rcu_read_unlock(). If the rcu_read_lock() isn't held
entering this routine we have issues with deleting the classifier
chain so remove the unnecessary rcu_read_lock()/rcu_read_unlock()
pair noting all paths AFAIK hold rcu_read_lock.
If there is a case where classify is called without the rcu read lock
then an rcu splat will occur and we can correct it.
Signed-off-by: John Fastabend <john.r.fastabend@intel.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
John Fastabend [Sat, 13 Sep 2014 03:05:59 +0000 (20:05 -0700)]
net: sched: cls_basic use RCU
Enable basic classifier for RCU.
Dereferencing tp->root may look a bit strange here but it is needed
by my accounting because it is allocated at init time and needs to
be kfree'd at destroy time. However because it may be referenced in
the classify() path we must wait an RCU grace period before free'ing
it. We use kfree_rcu() and rcu_ APIs to enforce this. This pattern
is used in all the classifiers.
Also the hgenerator can be incremented without concern because it
is always incremented under RTNL.
Signed-off-by: John Fastabend <john.r.fastabend@intel.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
John Fastabend [Sat, 13 Sep 2014 03:05:27 +0000 (20:05 -0700)]
net: rcu-ify tcf_proto
rcu'ify tcf_proto this allows calling tc_classify() without holding
any locks. Updaters are protected by RTNL.
This patch prepares the core net_sched infrastracture for running
the classifier/action chains without holding the qdisc lock however
it does nothing to ensure cls_xxx and act_xxx types also work without
locking. Additional patches are required to address the fall out.
Signed-off-by: John Fastabend <john.r.fastabend@intel.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
John Fastabend [Sat, 13 Sep 2014 03:04:52 +0000 (20:04 -0700)]
net: qdisc: use rcu prefix and silence sparse warnings
Add __rcu notation to qdisc handling by doing this we can make
smatch output more legible. And anyways some of the cases should
be using rcu_dereference() see qdisc_all_tx_empty(),
qdisc_tx_chainging(), and so on.
Also *wake_queue() API is commonly called from driver timer routines
without rcu lock or rtnl lock. So I added rcu_read_lock() blocks
around netif_wake_subqueue and netif_tx_wake_queue.
Signed-off-by: John Fastabend <john.r.fastabend@intel.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
When sending out a burst of packets across multiple descriptors,
it is sufficient to send one LDC "start" trigger for
the first descriptor, so do not send an LDC "start" for every
pass through vnet_start_xmit. Similarly, it is sufficient to send
one "DRING_STOPPED" trigger for the last dring (and if that
fails, hold off and send the trigger later).
Optimizations to the number of LDC messages helps avoid
filling up the LDC channel with superfluous LDC messages
that risk triggering flow-control on the channel,
and also boosts performance.
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com> Acked-by: Raghuram Kothakota <raghuram.kothakota@oracle.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Scott Wood [Thu, 11 Sep 2014 02:23:18 +0000 (21:23 -0500)]
udp: Fix inverted NAPI_GRO_CB(skb)->flush test
Commit 2abb7cdc0d ("udp: Add support for doing checksum unnecessary
conversion") caused napi_gro_cb structs with the "flush" field zero to
take the "udp_gro_receive" path rather than the "set flush to 1" path
that they would previously take. As a result I saw booting from an NFS
root hang shortly after starting userspace, with "server not
responding" messages.
This change to the handling of "flush == 0" packets appears to be
incidental to the goal of adding new code in the case where
skb_gro_checksum_validate_zero_check() returns zero. Based on that and
the fact that it breaks things, I'm assuming that it is unintentional.
Fixes: 2abb7cdc0d ("udp: Add support for doing checksum unnecessary conversion") Cc: Tom Herbert <therbert@google.com> Signed-off-by: Scott Wood <scottwood@freescale.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Fri, 12 Sep 2014 21:51:32 +0000 (17:51 -0400)]
Merge branch 'sock_queue_err_skb'
Alexander Duyck says:
====================
Address reference counting issues with sock_queue_err_skb
After looking over the code for skb_clone_sk after some comments made by
Eric Dumazet I have come to the conclusion that skb_clone_sk is taking the
correct approach in how to handle the sk_refcnt when creating a buffer that
is eventually meant to be returned to the socket via the sock_queue_err_skb
function.
However upon review of other callers I found what I believe to be a
possible reference count issue in the path for handling "wifi ack" packets.
To address this I have applied the same logic that is currently in place so
that the sk_refcnt will be forced to stay at least 1, or we will not
provide an skb to return in the sk_error_queue.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Alexander Duyck [Wed, 10 Sep 2014 22:05:42 +0000 (18:05 -0400)]
mac80211: Resolve sk_refcnt/sk_wmem_alloc issue in wifi ack path
There is a possible issue with the use, or lack thereof of sk_refcnt and
sk_wmem_alloc in the wifi ack status functionality.
Specifically if a socket were to request acknowledgements, and the socket
were to have sk_refcnt drop to 0 resulting in it waiting on sk_wmem_alloc
to reach 0 it would be possible to have sock_queue_err_skb orphan the last
buffer, resulting in __sk_free being called on the socket. After this the
buffer is enqueued on sk_error_queue, however the queue has already been
flushed resulting in at least a memory leak, if not a data corruption.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com> Acked-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: David S. Miller <davem@davemloft.net>
Revert "ipv4: Clarify in docs that accept_local requires rp_filter."
This reverts commit c801e3cc1925 ("ipv4: Clarify in docs that accept_local requires rp_filter.").
It is not needed anymore since commit 1dced6a85482 ("ipv4: Restore accept_local behaviour in fib_validate_source()").
Suggested-by: Julian Anastasov <ja@ssi.bg> Cc: Gregory Detal <gregory.detal@uclouvain.be> Cc: Christoph Paasch <christoph.paasch@uclouvain.be> Cc: Hannes Frederic Sowa <hannes@redhat.com> Cc: Sergei Shtylyov <sergei.shtylyov@cogentembedded.com> Signed-off-by: Sébastien Barré <sebastien.barre@uclouvain.be> Signed-off-by: David S. Miller <davem@davemloft.net>
projects / linux-beck.git / log