Florian Westphal [Mon, 4 Dec 2017 18:19:18 +0000 (19:19 +0100)]
rtnetlink: ipv6: convert remaining users to rtnl_register_module
convert remaining users of rtnl_register to rtnl_register_module
and un-export rtnl_register.
Requested-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Mon, 4 Dec 2017 18:18:54 +0000 (13:18 -0500)]
Merge tag 'linux-can-next-for-4.16-
20171201' of git://git./linux/kernel/git/mkl/linux-can-next
Marc Kleine-Budde says:
====================
pull-request: can-next 2017-12-01
this is a pull request of 10 patches for net-next/master.
The first two patches are by Arnd Bergmann, they convert the peak_usb
from using "struct timeval" to "ktime_t". The error handling in the
vxcan driver is clean up by Markus Elfring's patch. Bhumika Goyal
contributes a patch for the c_can_pci driver to make the pci data const.
The six patches by Pankaj Bansal for the flexcan driver add LS1021A
support by making the endianness of the driver configurable by the
device tree.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Mon, 4 Dec 2017 17:07:10 +0000 (12:07 -0500)]
Merge git://git./linux/kernel/git/bpf/bpf-next
Daniel Borkmann says:
====================
pull-request: bpf-next 2017-12-03
The following pull-request contains BPF updates for your *net-next* tree.
The main changes are:
1) Addition of a software model for BPF offloads in order to ease
testing code changes in that area and make semantics more clear.
This is implemented in a new driver called netdevsim, which can
later also be extended for other offloads. SR-IOV support is added
as well to netdevsim. BPF kernel selftests for offloading are
added so we can track basic functionality as well as exercising
all corner cases around BPF offloading, from Jakub.
2) Today drivers have to drop the reference on BPF progs they hold
due to XDP on device teardown themselves. Change this in order
to make XDP handling inside the drivers less error prone, and
move disabling XDP to the core instead, also from Jakub.
3) Misc set of BPF verifier improvements and cleanups as preparatory
work for upcoming BPF-to-BPF calls. Among others, this set also
improves liveness marking such that pruning can be slightly more
effective. Register and stack liveness information is now included
in the verifier log as well, from Alexei.
4) nfp JIT improvements in order to identify load/store sequences in
the BPF prog e.g. coming from memcpy lowering and optimizing them
through the NPU's command push pull (CPP) instruction, from Jiong.
5) Cleanups to test_cgrp2_attach2.c BPF sample code in oder to remove
bpf_prog_attach() magic values and replacing them with actual proper
attach flag instead, from David.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Mon, 4 Dec 2017 16:33:02 +0000 (11:33 -0500)]
Merge branch 'rtnetlink-rework-handler-registration'
Florian Westphal says:
====================
rtnetlink: rework handler (un)registering
Peter Zijlstra reported (referring to commit
019a316992ee0d983,
"rtnetlink: add reference counting to prevent module unload while dump is in progress"):
1) it not in fact a refcount, so using refcount_t is silly
2) there is a distinct lack of memory barriers, so we can easily
observe the decrement while the msg_handler is still in progress.
3) waiting with a schedule()/yield() loop is complete crap and subject
life-locks, imagine doing that rtnl_unregister_all() from a RT task.
In ancient times rtnetlink exposed a statically-sized table with
preset doit/dumpit handlers to be called for a protocol/type pair.
Later the rtnl_register interface was added and the table was allocated
on demand. Eventually these were also used by modules.
Problem is that nothing prevents module unload while a netlink dump
is in progress. netlink dumps can be span multiple recv calls and
netlink core saves the to-be-repeated dumper address for later invocation.
To prevent rmmod the netlink core expects callers to pass in the owning
module so a reference can be taken.
So far rtnetlink wasn't doing this, add new interface to pass THIS_MODULE.
Moreover, when converting parts of the rtnetlink handling to rcu this code
gained way too many READ_ONCE spots, remove them and the extra refcounting.
Take a module reference when running dumpit and doit callbacks
and never alter content of rtnl_link structures after they have been
published via rcu_assign_pointer.
Based partially on earlier patch from Peter.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Florian Westphal [Sat, 2 Dec 2017 20:44:08 +0000 (21:44 +0100)]
rtnetlink: remove __rtnl_register
This removes __rtnl_register and switches callers to either
rtnl_register or rtnl_register_module.
Also, rtnl_register() will now print an error if memory allocation
failed rather than panic the kernel.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
Florian Westphal [Sat, 2 Dec 2017 20:44:07 +0000 (21:44 +0100)]
net: use rtnl_register_module where needed
all of these can be compiled as a module, so use new
_module version to make sure module can no longer be removed
while callback/dump is in use.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
Florian Westphal [Sat, 2 Dec 2017 20:44:06 +0000 (21:44 +0100)]
rtnetlink: get reference on module before invoking handlers
Add yet another rtnl_register function. It will be used by modules
that can be removed.
The passed module struct is used to prevent module unload while
a netlink dump is in progress or when a DOIT_UNLOCKED doit callback
is called.
Cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
Florian Westphal [Sat, 2 Dec 2017 20:44:05 +0000 (21:44 +0100)]
net: rtnetlink: use rcu to free rtnl message handlers
rtnetlink is littered with READ_ONCE() because we can have read accesses
while another cpu can write to the structure we're reading by
(un)registering doit or dumpit handlers.
This patch changes this so that (un)registering cpu allocates a new
structure and then publishes it via rcu_assign_pointer, i.e. once
another cpu can see such pointer no modifications will occur anymore.
based on initial patch from Peter Zijlstra.
Cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
Heiner Kallweit [Sat, 2 Dec 2017 09:50:37 +0000 (10:50 +0100)]
net: phy: broadcom: re-add mistakenly removed config settings
Previous patch mistakenly removed three chip-specific config settings.
Add them again.
Fixes:
80274abafc60 "net: phy: remove generic settings for callbacks config_aneg and read_status from drivers"
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Mon, 4 Dec 2017 16:04:20 +0000 (11:04 -0500)]
Merge branch 'ipv6-gre-collect_md'
William Tu says:
====================
add ip6 gre and gretap collect_md mode
Similar to gre, vxlan, geneve, ipip tunnels, allow ip6gretap tunnels to
operate in collect metadata mode. The first patch adds the support to
ip6_gre.c. The second patch enables unsetting the csum for ipv6 tunnel,
when using bpf_skb_[gs]et_tunnel_key() helpers. Finally, the last patch
adds the ip6 gre and gretap tunnel test cases to BPF sample code.
The corresponding iproute2 patch:
https://marc.info/?l=linux-netdev&m=
151216943128087&w=2
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
William Tu [Fri, 1 Dec 2017 23:26:10 +0000 (15:26 -0800)]
samples/bpf: extend test_tunnel_bpf.sh with ip6gre
Extend existing tests for vxlan, gre, geneve, ipip, erspan,
to include ip6 gre and gretap tunnel.
Signed-off-by: William Tu <u9012063@gmail.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
William Tu [Fri, 1 Dec 2017 23:26:09 +0000 (15:26 -0800)]
bpf: allow disabling tunnel csum for ipv6
Before the patch, BPF_F_ZERO_CSUM_TX can be used only for ipv4 tunnel.
With introduction of ip6gretap collect_md mode, the flag should be also
supported for ipv6.
Signed-off-by: William Tu <u9012063@gmail.com>
Cc: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
William Tu [Fri, 1 Dec 2017 23:26:08 +0000 (15:26 -0800)]
ip6_gre: add ip6 gre and gretap collect_md mode
Similar to gre, vxlan, geneve, ipip tunnels, allow ip6 gre and gretap
tunnels to operate in collect metadata mode. bpf_skb_[gs]et_tunnel_key()
helpers can make use of it right away. OVS can use it as well in the
future.
Signed-off-by: William Tu <u9012063@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Heiner Kallweit [Thu, 30 Nov 2017 22:57:00 +0000 (23:57 +0100)]
net: phy: core: don't disable device interrupts in phy_change
If state is not PHY_HALTED I see no need to temporarily disable
interrupts on the device. As long as the current interrupt isn't acked
on the device no new interrupt can happen anyway.
In addition remove a unneeded enabling of interrupts in the state
machine when handling state PHY_CHANGELINK.
Tested on a Odroid-C2 with RTL8211F phy in interrupt mode.
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Heiner Kallweit [Thu, 30 Nov 2017 22:55:15 +0000 (23:55 +0100)]
net: phy: core: remove now uneeded disabling of interrupts
After commits
c974bdbc3e "net: phy: Use threaded IRQ, to allow IRQ from
sleeping devices" and
664fcf123a30 "net: phy: Threaded interrupts allow
some simplification" all relevant code pieces run in process context
anyway and I don't think we need the disabling of interrupts any longer.
Interestingly enough, latter commit already removed the comment
explaining why interrupts need to be temporarily disabled.
On my system phy interrupt mode works fine with this patch.
However I may miss something, especially in the context of shared phy
interrupts, therefore I'd appreciate if more people could test this.
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Acked-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Sun, 3 Dec 2017 15:18:28 +0000 (10:18 -0500)]
Merge branch 'tcp-2nd-listener-hash'
Martin KaFai Lau says:
====================
tcp: Add a 2nd listener hashtable (port+addr)
This patch set adds a 2nd listener hashtable. It is to resolve
the performance issue when a process is listening at many IP
addresses with the same port (e.g. [IP1]:443, [IP2]:443... [IPN]:443)
v2:
- Move the new lhash2 and lhash2_mask before the existing
listening_hash to avoid adding another cacheline
to inet_hashinfo (Suggested by Eric Dumazet, Thanks!)
- I take this chance to plug an existing 4 bytes hole while
adding 'unsigned int lhash2_mask'.
- Add some comments about lhash2 in inet_hashtables.h
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Martin KaFai Lau [Fri, 1 Dec 2017 20:52:32 +0000 (12:52 -0800)]
tcp: Enable 2nd listener hashtable in TCP
Enable the second listener hashtable in TCP.
The scale is the same as UDP which is one slot per 2MB.
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Martin KaFai Lau [Fri, 1 Dec 2017 20:52:31 +0000 (12:52 -0800)]
inet: Add a 2nd listener hashtable (port+addr)
The current listener hashtable is hashed by port only.
When a process is listening at many IP addresses with the same port (e.g.
[IP1]:443, [IP2]:443... [IPN]:443), the inet[6]_lookup_listener()
performance is degraded to a link list. It is prone to syn attack.
UDP had a similar issue and a second hashtable was added to resolve it.
This patch adds a second hashtable for the listener's sockets.
The second hashtable is hashed by port and address.
It cannot reuse the existing skc_portaddr_node which is shared
with skc_bind_node. TCP listener needs to use skc_bind_node.
Instead, this patch adds a hlist_node 'icsk_listen_portaddr_node' to
the inet_connection_sock which the listener (like TCP) also belongs to.
The new portaddr hashtable may need two lookup (First by IP:PORT.
Second by INADDR_ANY:PORT if the IP:PORT is a not found). Hence,
it implements a similar cut off as UDP such that it will only consult the
new portaddr hashtable if the current port-only hashtable has >10
sk in the link-list.
lhash2 and lhash2_mask are added to 'struct inet_hashinfo'. I take
this chance to plug a 4 bytes hole. It is done by first moving
the existing bind_bucket_cachep up and then add the new
(int lhash2_mask, *lhash2) after the existing bhash_size.
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Martin KaFai Lau [Fri, 1 Dec 2017 20:52:30 +0000 (12:52 -0800)]
udp: Move udp[46]_portaddr_hash() to net/ip[v6].h
This patch moves the udp[46]_portaddr_hash()
to net/ip[v6].h. The function name is renamed to
ipv[46]_portaddr_hash().
It will be used by a later patch which adds a second listener
hashtable hashed by the address and port.
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Martin KaFai Lau [Fri, 1 Dec 2017 20:52:29 +0000 (12:52 -0800)]
inet: Add a count to struct inet_listen_hashbucket
This patch adds a count to the 'struct inet_listen_hashbucket'.
It counts how many sk is hashed to a bucket. It will be
used to decide if the (to-be-added) portaddr listener's hashtable
should be used during inet[6]_lookup_listener().
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Govindarajulu Varadarajan [Fri, 1 Dec 2017 18:21:40 +0000 (10:21 -0800)]
enic: add sw timestamp support
Add ethtool ops to advertise sw timestamping.
Call skb_tx_timestamp() just before ringing the wq doorbell.
Signed-off-by: Govindarajulu Varadarajan <gvaradar@cisco.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Sun, 3 Dec 2017 15:10:03 +0000 (10:10 -0500)]
Merge branch 'hv_netvsc-minor-optimizations'
Stephen Hemminger says:
====================
hv_netvsc: minor optimizations
These are a set of local optimizations the Hyper-V networking driver.
Also include a vmbus patch in this set, because it depends on the
netvsc that last used that function.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Stephen Hemminger [Fri, 1 Dec 2017 19:01:49 +0000 (11:01 -0800)]
vmbus: make hv_get_ringbuffer_availbytes local
The last use of hv_get_ringbuffer_availbytes in drivers is now
gone. Only used by the debug info routine so make it static. Also, add
READ_ONCE() to avoid any possible issues with potentially volatile
index values.
Signed-off-by: Stephen Hemminger <sthemmin@microsoft.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Stephen Hemminger [Fri, 1 Dec 2017 19:01:48 +0000 (11:01 -0800)]
hv_netvsc: optimize initialization of RNDIS header
The memset of the whole maximum possible RNDIS header is unnecessary.
For the main part of the header use a structure assignment.
No need to memset the whole per packet info. Instead rely on caller to
set what it wants. Also get rid of cast to void and signed/unsigned
conversion. Now return pointer to per packet data (rather than the
header) which simplifies use by code setting up the packet data.
Signed-off-by: Stephen Hemminger <sthemmin@microsoft.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Stephen Hemminger [Fri, 1 Dec 2017 19:01:47 +0000 (11:01 -0800)]
hv_netvsc: use reciprocal divide to speed up percent calculation
Every packet sent checks the available ring space. The calculation
can be sped up by using reciprocal divide which is multiplication.
Since ring_size can only be configured by module parameter, so it doesn't
have to be passed around everywhere. Also it should be unsigned
since it is number of pages.
Signed-off-by: Stephen Hemminger <sthemmin@microsoft.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Stephen Hemminger [Fri, 1 Dec 2017 19:01:46 +0000 (11:01 -0800)]
hv_netvsc: replace divide with mask when computing padding
Packet alignment is always a power of 2 therefore modulus can
be replaced with a faster and operation
Signed-off-by: Stephen Hemminger <sthemmin@microsoft.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Stephen Hemminger [Fri, 1 Dec 2017 19:01:45 +0000 (11:01 -0800)]
hv_netvsc: don't need local xmit_more
Since skb is always non-NULL in the copy portion of netvsc_send
do not need local variable.
Signed-off-by: Stephen Hemminger <sthemmin@microsoft.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Stephen Hemminger [Fri, 1 Dec 2017 19:01:44 +0000 (11:01 -0800)]
hv_netvsc: drop unused macros
Signed-off-by: Stephen Hemminger <sthemmin@microsoft.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Gao Feng [Fri, 1 Dec 2017 08:33:03 +0000 (16:33 +0800)]
ipvlan: Add new func ipvlan_is_valid_dev instead of duplicated codes
There are multiple duplicated condition checks in the current codes, so
I add the new func ipvlan_is_valid_dev instead of the duplicated codes to
check if the netdev is real ipvlan dev.
Signed-off-by: Gao Feng <gfree.wind@vip.163.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Sun, 3 Dec 2017 14:38:17 +0000 (09:38 -0500)]
Merge branch 'realtek-phy-improvements'
Martin Blumenstingl says:
====================
Realtek Ethernet PHY driver improvements
This series provides some small improvements and cleanups for the
Realtek Ethernet PHY driver.
None of the patches in this series should change any functionality.
The goal is to make the code a bit easier to read by:
- re-using the BIT and GENMASK macros (which makes it easier to compare
the #defines in the kernel with the values from the datasheets)
- rename a #define from a generic name to a PHY-specific name since it's
only used for one specific PHY
- logically group the register #defines and their register bit #defines
together
- indentation cleanups
- removed some code duplicating for reading/writing registers on a
Realtek specific "page"
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Martin Blumenstingl [Sat, 2 Dec 2017 21:51:28 +0000 (22:51 +0100)]
net: phy: realtek: add utility functions to read/write page addresses
Realtek PHYs implement the concept of so-called "extension pages". The
reason for this is probably because these PHYs expose more registers
than available in the standard address range.
After all read/write operations on such a page are done the driver
should switch back to page 0 where the standard MII registers (such as
MII_BMCR) are available.
When referring to such a register the datasheets of RTL8211E and
RTL8211F always specify:
- the page / "ext. page" which has to be written to RTL821x_PAGE_SELECT
- an address (sometimes also called reg)
These new utility functions make the existing code easier to read since
it removes some duplication (switching back to page 0 is done within the
new helpers for example).
No functional changes are intended.
Signed-off-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
Martin Blumenstingl [Sat, 2 Dec 2017 21:51:27 +0000 (22:51 +0100)]
net: phy: realtek: use the same indentation for all #defines
This simply makes the code easier to read. No functional changes.
Signed-off-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
Martin Blumenstingl [Sat, 2 Dec 2017 21:51:26 +0000 (22:51 +0100)]
net: phy: realtek: group all register bit #defines for RTL821x_INER
This simply moves all register bit #defines which describe the (PHY
specific) bits in the RTL821x_INER right below the RTL821x_INER register
definition. This makes it easier to spot which registers and bits belong
together.
No functional changes.
Signed-off-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
Martin Blumenstingl [Sat, 2 Dec 2017 21:51:25 +0000 (22:51 +0100)]
net: phy: realtek: rename RTL821x_INER_INIT to RTL8211B_INER_INIT
This macro is only used by the RTL8211B code. RTL8211E and RTL8211F both
use other bits to initialize the RTL821x_INER register.
No functional changes.
Signed-off-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
Martin Blumenstingl [Sat, 2 Dec 2017 21:51:24 +0000 (22:51 +0100)]
net: phy: realtek: use the BIT and GENMASK macros
This makes it easier to compare the #defines with the datasheets.
No functional changes.
Signed-off-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Sun, 3 Dec 2017 02:21:29 +0000 (21:21 -0500)]
Merge branch 'dsa-cross-chip-FDB-support'
Vivien Didelot says:
====================
net: dsa: cross-chip FDB support
DSA can have interconnected switches. For instance, the ZII Dev Rev B
board described in arch/arm/boot/dts/vf610-zii-dev-rev-b.dts has a
switch fabric composed of 3 switch devices like this:
lan4 lan6
CPU (eth1) | lan5 | lan7
| | | | |
[0 1 2 3 4 6 5]---[6 0 1 2 3 4 5]---[9 0 1 2 3 4 5 6 7 8]
| | | | | | |
lan0 | lan2 lan3 lan8 | optical4
lan1 optical3
One current issue with DSA is cross-chip FDB. If we add a static MAC
address on lan3, only its parent switch 1 (the one in the middle) will
be programmed. That is not correct in a cross-chip environment, because
the DSA ports connecting to switch 1 of adjacent switch 0 (on the left)
and switch 2 (on the right) must be programmed too.
Without this patchset, a dump of the hardware FDB of switches 0, 1 and 2
after programming a MAC address on lan3 looks like this (*):
# bridge fdb add 11:22:33:44:55:66 dev lan3
# cat /sys/kernel/debug/mv88e6xxx/sw*/atu/0 | grep -v FID
0 ff:ff:ff:ff:ff:ff MC_STATIC n 0 1 2 3 4 5 6
0 11:22:33:44:55:66 MC_STATIC_MGMT_PO n 0 - - - - - -
0 ff:ff:ff:ff:ff:ff MC_STATIC n 0 1 2 3 4 5 6
0 ff:ff:ff:ff:ff:ff MC_STATIC n 0 1 2 3 4 5 6 7 8 9
With this patchset applied, adjacent DSA ports get programmed too:
# bridge fdb add 11:22:33:44:55:66 dev lan3
# cat /sys/kernel/debug/mv88e6xxx/sw*/atu/0 | grep -v FID
0 11:22:33:44:55:66 MC_STATIC_MGMT_PO n - - - - - 5 -
0 ff:ff:ff:ff:ff:ff MC_STATIC n 0 1 2 3 4 5 6
0 11:22:33:44:55:66 MC_STATIC_MGMT_PO n 0 - - - - - -
0 ff:ff:ff:ff:ff:ff MC_STATIC n 0 1 2 3 4 5 6
0 11:22:33:44:55:66 MC_STATIC_MGMT_PO n - - - - - - - - - 9
0 ff:ff:ff:ff:ff:ff MC_STATIC n 0 1 2 3 4 5 6 7 8 9
In order to do that, the first commit introduces a dsa_towards_port()
helper which returns the local port of a switch which must be used to
reach an arbitrary switch port (local or from an adjacent switch.)
The second patch uses this helper to configure the port reaching the
target port for every switches of the fabric.
(*) a patch for squashed debugfs interface which applies on top of this
patchset is available here:
https://github.com/vivien/linux/commit/
f8e6ba34c68a72d3bf42f4dea79abacb2e61a3cc.patch
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Vivien Didelot [Thu, 30 Nov 2017 17:56:43 +0000 (12:56 -0500)]
net: dsa: support cross-chip FDB operations
When a MAC address is added to or removed from a switch port in the
fabric, the target switch must program its port and adjacent switches
must program their local DSA port used to reach the target switch.
For this purpose, use the dsa_towards_port() helper to identify the
local switch port which must be programmed.
Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Vivien Didelot [Thu, 30 Nov 2017 17:56:42 +0000 (12:56 -0500)]
net: dsa: introduce dsa_towards_port helper
Add a new helper returning the local port used to reach an arbitrary
switch port in the fabric.
Its only user at the moment is the dsa_upstream_port helper, which
returns the local port reaching the dedicated CPU port, but it will be
used in cross-chip FDB operations.
Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Sun, 3 Dec 2017 02:18:57 +0000 (21:18 -0500)]
Merge branch 'dsa-simplify-switchdev-prepare-phase'
Vivien Didelot says:
====================
net: dsa: simplify switchdev prepare phase
This patch series brings no functional changes.
It removes the unused switchdev_trans arguments from the dsa_switch_ops
for both MDB and VLAN operations, and provides functions to prepare and
add these objects for a given bitmap of ports.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Vivien Didelot [Thu, 30 Nov 2017 16:24:00 +0000 (11:24 -0500)]
net: dsa: add switch mdb bitmap functions
This patch brings no functional changes.
It moves out the MDB code iterating on a multicast group into new
dsa_switch_mdb_{prepare,add}_bitmap() functions.
This gives us a better isolation of the two switchdev phases.
Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Vivien Didelot [Thu, 30 Nov 2017 16:23:59 +0000 (11:23 -0500)]
net: dsa: add switch vlan bitmap functions
This patch brings no functional changes.
It moves out the VLAN code iterating on a list of VLAN members into new
dsa_switch_vlan_{prepare,add}_bitmap() functions.
This gives us a better isolation of the two switchdev phases.
Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Vivien Didelot [Thu, 30 Nov 2017 16:23:58 +0000 (11:23 -0500)]
net: dsa: remove trans argument from mdb ops
The DSA switch MDB ops pass the switchdev_trans structure down to the
drivers, but no one is using them and they aren't supposed to anyway.
Remove the trans argument from MDB prepare and add operations.
Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Vivien Didelot [Thu, 30 Nov 2017 16:23:57 +0000 (11:23 -0500)]
net: dsa: remove trans argument from vlan ops
The DSA switch VLAN ops pass the switchdev_trans structure down to the
drivers, but no one is using them and they aren't supposed to anyway.
Remove the trans argument from VLAN prepare and add operations.
At the same time, fix the following checkpatch warning:
WARNING: line over 80 characters
#74: FILE: drivers/net/dsa/dsa_loop.c:177:
+ const struct switchdev_obj_port_vlan *vlan)
Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Paolo Abeni [Thu, 30 Nov 2017 14:35:33 +0000 (15:35 +0100)]
openvswitch: do not propagate headroom updates to internal port
After commit
3a927bc7cf9d ("ovs: propagate per dp max headroom to
all vports") the need_headroom for the internal vport is updated
accordingly to the max needed headroom in its datapath.
That avoids the pskb_expand_head() costs when sending/forwarding
packets towards tunnel devices, at least for some scenarios.
We still require such copy when using the ovs-preferred configuration
for vxlan tunnels:
br_int
/ \
tap vxlan
(remote_ip:X)
br_phy
\
NIC
where the route towards the IP 'X' is via 'br_phy'.
When forwarding traffic from the tap towards the vxlan device, we
will call pskb_expand_head() in vxlan_build_skb() because
br-phy->needed_headroom is equal to tun->needed_headroom.
With this change we avoid updating the internal vport needed_headroom,
so that in the above scenario no head copy is needed, giving 5%
performance improvement in UDP throughput test.
As a trade-off, packets sent from the internal port towards a tunnel
device will now experience the head copy overhead. The rationale is
that the latter use-case is less relevant performance-wise.
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Acked-by: Pravin B Shelar <pshelar@ovn.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Daniel Borkmann [Sat, 2 Dec 2017 23:27:58 +0000 (00:27 +0100)]
Merge branch 'bpf-xdp-stack-uninit-and-offload-tests'
Jakub Kicinski says:
====================
The purpose of this series is to add a software model of BPF offloads
to make it easier for everyone to test them and make some of the more
arcane rules and assumptions more clear.
The series starts with 3 patches aiming to make XDP handling in the
drivers less error prone. Currently driver authors have to remember
to free XDP programs if XDP is active during unregister. With this
series the core will disable XDP on its own. It will take place
after close, drivers are not expected to perform reconfiguration
when disabling XDP on a downed device.
Next two patches add the software netdev driver, followed by a python
test which exercises all the corner cases which came to my mind.
Test needs to be run as root. It will print basic information to
stdout, but can also create a more detailed log of all commands
when --log option is passed. Log is in Emacs Org-mode format.
./tools/testing/selftests/bpf/test_offload.py --log /tmp/log
Last two patches replace the SR-IOV API implementation of dummy.
v3:
- move the freeing of vfs to release (Phil).
v2:
- free device from the release function;
- use bus-based name generatin instead of netdev name.
v1:
- replace the SR-IOV API implementation of dummy;
- make the dev_xdp_uninstall() also handle the XDP generic (Daniel).
====================
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Jakub Kicinski [Fri, 1 Dec 2017 23:09:02 +0000 (15:09 -0800)]
net: dummy: remove fake SR-IOV functionality
netdevsim driver seems like a better place for fake SR-IOV
functionality. Remove the code previously added to dummy.
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Quentin Monnet <quentin.monnet@netronome.com>
Acked-by: Phil Sutter <phil@nwl.cc>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Jakub Kicinski [Fri, 1 Dec 2017 23:09:01 +0000 (15:09 -0800)]
netdevsim: add SR-IOV functionality
dummy driver was extended with VF-related netdev APIs for testing
SR-IOV-related software. netdevsim did not exist back then.
Implement SR-IOV functionality in netdevsim. Notable difference
is that since netdevsim has no module parameters, we will actually
create a device with sriov_numvfs attribute for each netdev.
The zero MAC address is accepted as some HW use it to mean any
address is allowed. Link state is also now validated.
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Quentin Monnet <quentin.monnet@netronome.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Jakub Kicinski [Fri, 1 Dec 2017 23:09:00 +0000 (15:09 -0800)]
selftests/bpf: add offload test based on netdevsim
Add a test of BPF offload control path interfaces based on
just-added netdevsim driver. Perform various checks of both
the stack and the expected driver behaviour.
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Simon Horman <simon.horman@netronome.com>
Reviewed-by: Quentin Monnet <quentin.monnet@netronome.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Jakub Kicinski [Fri, 1 Dec 2017 23:08:59 +0000 (15:08 -0800)]
netdevsim: add bpf offload support
Add support for loading programs for netdevsim devices and
expose the related information via DebugFS. Both offload
of XDP and cls_bpf programs is supported.
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Simon Horman <simon.horman@netronome.com>
Reviewed-by: Quentin Monnet <quentin.monnet@netronome.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Jakub Kicinski [Fri, 1 Dec 2017 23:08:58 +0000 (15:08 -0800)]
netdevsim: add software driver for testing offloads
To be able to run selftests without any hardware required we
need a software model. The model can also serve as an example
implementation for those implementing actual HW offloads.
The dummy driver have previously been extended to test SR-IOV,
but the general consensus seems to be against adding further
features to it.
Add a new driver for purposes of software modelling only.
eBPF and SR-IOV will be added here shortly, others are invited
to further extend the driver with their offload models.
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Simon Horman <simon.horman@netronome.com>
Reviewed-by: Quentin Monnet <quentin.monnet@netronome.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Jakub Kicinski [Fri, 1 Dec 2017 23:08:57 +0000 (15:08 -0800)]
net: xdp: make the stack take care of the tear down
Since day one of XDP drivers had to remember to free the program
on the remove path. This leads to code duplication and is error
prone. Make the stack query the installed programs on unregister
and if something is installed, remove the program. Freeing of
program attached to XDP generic is moved from free_netdev() as well.
Because the remove will now be called before notifiers are
invoked, BPF offload state of the program will not get destroyed
before uninstall.
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Simon Horman <simon.horman@netronome.com>
Reviewed-by: Quentin Monnet <quentin.monnet@netronome.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Jakub Kicinski [Fri, 1 Dec 2017 23:08:56 +0000 (15:08 -0800)]
net: xdp: report flags program was installed with on query
Some drivers enforce that flags on program replacement and
removal must match the flags passed on install. This leaves
the possibility open to enable simultaneous loading
of XDP programs both to HW and DRV.
Allow such drivers to report the flags back to the stack.
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Simon Horman <simon.horman@netronome.com>
Reviewed-by: Quentin Monnet <quentin.monnet@netronome.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Jakub Kicinski [Fri, 1 Dec 2017 23:08:55 +0000 (15:08 -0800)]
net: xdp: avoid output parameters when querying XDP prog
The output parameters will get unwieldy if we want to add more
information about the program. Simply pass the entire
struct netdev_bpf in.
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Simon Horman <simon.horman@netronome.com>
Reviewed-by: Quentin Monnet <quentin.monnet@netronome.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
David S. Miller [Fri, 1 Dec 2017 21:36:33 +0000 (16:36 -0500)]
Merge branch 'cpsw-ale-cleanups'
Grygorii Strashko says:
====================
net: ethernet: ti: cpsw/ale clean up and optimization
This is set of non critical clean ups and optimizations for TI
CPSW and ALE drivers.
Rebased on top on net-next.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Grygorii Strashko [Fri, 1 Dec 2017 00:21:20 +0000 (18:21 -0600)]
net: ethernet: ti: ale: fix port check in cpsw_ale_control_set/get
ALE ports number includes the Host port and ext Ports, and
ALE ports numbering starts from 0, so correct corresponding port
checks in cpsw_ale_control_set/get().
Signed-off-by: Grygorii Strashko <grygorii.strashko@ti.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Grygorii Strashko [Fri, 1 Dec 2017 00:21:19 +0000 (18:21 -0600)]
net: ethernet: ti: ale: use devm_kzalloc in cpsw_ale_create()
Use cpsw_ale_create in cpsw_ale_create(). This also makes
cpsw_ale_destroy() function nop, so remove it.
Signed-off-by: Grygorii Strashko <grygorii.strashko@ti.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Grygorii Strashko [Fri, 1 Dec 2017 00:21:18 +0000 (18:21 -0600)]
net: ethernet: ti: ale: move static initialization in cpsw_ale_create()
Move static initialization from cpsw_ale_start() to cpsw_ale_create() as it
does not make much sence to perform static initializtion in
cpsw_ale_start() which is called everytime netif[s] is opened.
Signed-off-by: Grygorii Strashko <grygorii.strashko@ti.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Grygorii Strashko [Fri, 1 Dec 2017 00:21:17 +0000 (18:21 -0600)]
net: ethernet: ti: ale: optimize ale entry mask bits configuartion
The ale->params.ale_ports parameter can be used to deriver values for all
ale entry mask bits: port_mask_bits, port_mask_bits, port_num_bits.
Hence, calculate above values and drop all hardcoded values. For
port_num_bits calcualtion use order_base_2() API.
Signed-off-by: Grygorii Strashko <grygorii.strashko@ti.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Grygorii Strashko [Fri, 1 Dec 2017 00:21:16 +0000 (18:21 -0600)]
net: ethernet: ti: ale: disable ale from stop()
ALE is enabled from cpsw_ale_start() now, but disabled only from
cpsw_ale_destroy() which introduces inconsitance as cpsw_ale_start() is
called when netif[s] is opened, but cpsw_ale_destroy() is called when
driver is removed. Hence, move ALE disabling in cpsw_ale_stop().
Signed-off-by: Grygorii Strashko <grygorii.strashko@ti.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Grygorii Strashko [Fri, 1 Dec 2017 00:21:15 +0000 (18:21 -0600)]
net: ethernet: ti: ale: use proper io apis
Switch to use writel_relaxed/readl_relaxed() IO API instead of raw version
as it is recommended.
Signed-off-by: Grygorii Strashko <grygorii.strashko@ti.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Grygorii Strashko [Fri, 1 Dec 2017 00:21:14 +0000 (18:21 -0600)]
net: ethernet: ti: cpsw: fix ale port numbers
TI OMAP/Sitara SoCs have fixed number of ALE ports 3, which includes Host
port also.
Hence, use fixed value instead of value calcualted from DT, which can be
set by user and might not reflect actual HW configuration.
Signed-off-by: Grygorii Strashko <grygorii.strashko@ti.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Grygorii Strashko [Fri, 1 Dec 2017 00:21:13 +0000 (18:21 -0600)]
net: ethernet: ti: cpsw: move mac_hi/lo defines in cpsw.h
Move mac_hi/lo defines in common header cpsw.h and re-use
them for netcp_ethss.c.
Signed-off-by: Grygorii Strashko <grygorii.strashko@ti.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Grygorii Strashko [Fri, 1 Dec 2017 00:21:12 +0000 (18:21 -0600)]
net: ethernet: ti: cpsw: move platform data struct to .c file
CPSW platform data struct cpsw_platform_data and struct cpsw_slave_data are
used only incide cpsw.c module, so move these definitions there.
Signed-off-by: Grygorii Strashko <grygorii.strashko@ti.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Grygorii Strashko [Fri, 1 Dec 2017 00:21:11 +0000 (18:21 -0600)]
net: ethernet: ti: cpsw: use proper io apis
Switch to use writel_relaxed/readl_relaxed() IO API instead of raw version
as it is recommended.
Signed-off-by: Grygorii Strashko <grygorii.strashko@ti.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Grygorii Strashko [Fri, 1 Dec 2017 00:21:10 +0000 (18:21 -0600)]
net: ethernet: ti: cpsw: drop unused var poll from cpsw_update_channels_res
Drop unused variable "poll" from cpsw_update_channels_res().
Signed-off-by: Grygorii Strashko <grygorii.strashko@ti.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Heiner Kallweit [Thu, 30 Nov 2017 22:47:52 +0000 (23:47 +0100)]
net: phy: remove generic settings for callbacks config_aneg and read_status from drivers
Remove generic settings for callbacks config_aneg and read_status
from drivers.
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Heiner Kallweit [Thu, 30 Nov 2017 22:46:19 +0000 (23:46 +0100)]
net: phy: core: use genphy version of callbacks read_status and config_aneg per default
read_status and config_aneg are the only mandatory callbacks and most
of the time the generic implementation is used by drivers.
So make the core fall back to the generic version if a driver doesn't
implement the respective callback.
Also currently the core doesn't seem to verify that drivers implement
the mandatory calls. If a driver doesn't do so we'd just get a NPE.
With this patch this potential issue doesn't exit any longer.
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Fri, 1 Dec 2017 20:33:27 +0000 (15:33 -0500)]
Merge branch 'ip6_gre-add-erspan-native-tunnel-for-ipv6'
William Tu says:
====================
ip6_gre: add erspan native tunnel for ipv6
The patch series add support for ERSPAN tunnel over ipv6. The first patch
refectors the existing ipv4 gre implementation and the second refactors the
ipv6 gre's xmit code. Finally the last patch introduces erspan protocol.
change in v5:
- add cover-letter description
change in v4:
- rebase on top of net-next
- use log_ecn_error in ip6_tnl_rcv
change in v3:
- add inline for functions in header
- rebase on top of net-next
change in v2:
- remove inline
- fix some indent
- fix errors reports by clang and scan-build
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
William Tu [Thu, 30 Nov 2017 19:51:29 +0000 (11:51 -0800)]
ip6_gre: Add ERSPAN native tunnel support
The patch adds support for ERSPAN tunnel over ipv6.
Signed-off-by: William Tu <u9012063@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
William Tu [Thu, 30 Nov 2017 19:51:28 +0000 (11:51 -0800)]
ip6_gre: Refactor ip6gre xmit codes
This patch refactors the ip6gre_xmit_{ipv4, ipv6}.
It is a prep work to add the ip6erspan tunnel.
Signed-off-by: William Tu <u9012063@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
William Tu [Thu, 30 Nov 2017 19:51:27 +0000 (11:51 -0800)]
ip_gre: Refector the erpsan tunnel code.
Move two erspan functions to header file, erspan.h, so ipv6
erspan implementation can use it.
Signed-off-by: William Tu <u9012063@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Fri, 1 Dec 2017 20:29:40 +0000 (15:29 -0500)]
Merge branch 'ethtool-reset-AP'
Scott Branden says:
====================
net: ethtool: add support for ETH_RESET_AP
Add support to reset appplication processors inside SmartNICs by
defining new ETH_RESET_AP bit.
And use new ETH_RESET_AP bit in bnxt ethernet driver.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Scott Branden [Thu, 30 Nov 2017 19:36:00 +0000 (11:36 -0800)]
bnxt_en: Add ETH_RESET_AP support
Add ETH_RESET_AP support handling to reset the internal
Application Processor(s) of the SmartNIC card.
Signed-off-by: Scott Branden <scott.branden@broadcom.com>
Acked-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Scott Branden [Thu, 30 Nov 2017 19:35:59 +0000 (11:35 -0800)]
net: ethtool: add support for reset of AP inside NIC interface.
Add ETH_RESET_AP to reset the application processor(s) inside the NIC
interface.
Current ETH_RESET_MGMT supports a management processor inside this NIC.
This is typically used for remote NIC management purposes.
Application processors exist inside some SmartNICs to run various
applications inside the NIC processor - be it a simple algorithm without
an OS to as complex as hosting multiple VMs.
Signed-off-by: Scott Branden <scott.branden@broadcom.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Fri, 1 Dec 2017 20:25:15 +0000 (15:25 -0500)]
Merge branch 'rds-tcp-netns-delete-related-fixes'
Sowmini Varadhan says:
====================
rds-tcp netns delete related fixes
Patchset contains cleanup and bug fixes. Patch 1 is the removal
of some redundant code/functions. Patch 2 and 3 are fixes for
corner cases identified by syzkaller. I've not been able to
reproduce the actual use-after-free race flagged in the syzkaller
reports, thus these fixes are based on code inspection plus
manual testing to make sure the modified code paths are executed
without problems in the commonly encountered timing cases.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Sowmini Varadhan [Thu, 30 Nov 2017 19:11:29 +0000 (11:11 -0800)]
rds: tcp: atomically purge entries from rds_tcp_conn_list during netns delete
The rds_tcp_kill_sock() function parses the rds_tcp_conn_list
to find the rds_connection entries marked for deletion as part
of the netns deletion under the protection of the rds_tcp_conn_lock.
Since the rds_tcp_conn_list tracks rds_tcp_connections (which
have a 1:1 mapping with rds_conn_path), multiple tc entries in
the rds_tcp_conn_list will map to a single rds_connection, and will
be deleted as part of the rds_conn_destroy() operation that is
done outside the rds_tcp_conn_lock.
The rds_tcp_conn_list traversal done under the protection of
rds_tcp_conn_lock should not leave any doomed tc entries in
the list after the rds_tcp_conn_lock is released, else another
concurrently executiong netns delete (for a differnt netns) thread
may trip on these entries.
Reported-by: syzbot <syzkaller@googlegroups.com>
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Acked-by: Santosh Shilimkar <santosh.shilimkar@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Sowmini Varadhan [Thu, 30 Nov 2017 19:11:28 +0000 (11:11 -0800)]
rds: tcp: correctly sequence cleanup on netns deletion.
Commit
8edc3affc077 ("rds: tcp: Take explicit refcounts on struct net")
introduces a regression in rds-tcp netns cleanup. The cleanup_net(),
(and thus rds_tcp_dev_event notification) is only called from put_net()
when all netns refcounts go to 0, but this cannot happen if the
rds_connection itself is holding a c_net ref that it expects to
release in rds_tcp_kill_sock.
Instead, the rds_tcp_kill_sock callback should make sure to
tear down state carefully, ensuring that the socket teardown
is only done after all data-structures and workqs that depend
on it are quiesced.
The original motivation for commit
8edc3affc077 ("rds: tcp: Take explicit
refcounts on struct net") was to resolve a race condition reported by
syzkaller where workqs for tx/rx/connect were triggered after the
namespace was deleted. Those worker threads should have been
cancelled/flushed before socket tear-down and indeed,
rds_conn_path_destroy() does try to sequence this by doing
/* cancel cp_send_w */
/* cancel cp_recv_w */
/* flush cp_down_w */
/* free data structures */
Here the "flush cp_down_w" will trigger rds_conn_shutdown and thus
invoke rds_tcp_conn_path_shutdown() to close the tcp socket, so that
we ought to have satisfied the requirement that "socket-close is
done after all other dependent state is quiesced". However,
rds_conn_shutdown has a bug in that it *always* triggers the reconnect
workq (and if connection is successful, we always restart tx/rx
workqs so with the right timing, we risk the race conditions reported
by syzkaller).
Netns deletion is like module teardown- no need to restart a
reconnect in this case. We can use the c_destroy_in_prog bit
to avoid restarting the reconnect.
Fixes:
8edc3affc077 ("rds: tcp: Take explicit refcounts on struct net")
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Acked-by: Santosh Shilimkar <santosh.shilimkar@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Sowmini Varadhan [Thu, 30 Nov 2017 19:11:27 +0000 (11:11 -0800)]
rds: tcp: remove redundant function rds_tcp_conn_paths_destroy()
A side-effect of Commit
c14b0366813a ("rds: tcp: set linger to 1
when unloading a rds-tcp") is that we always send a RST on the tcp
connection for rds_conn_destroy(), so rds_tcp_conn_paths_destroy()
is not needed any more and is removed in this patch.
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Acked-by: Santosh Shilimkar <santosh.shilimkar@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Jon Maloy [Thu, 30 Nov 2017 15:47:25 +0000 (16:47 +0100)]
tipc: fall back to smaller MTU if allocation of local send skb fails
When sending node local messages the code is using an 'mtu' of 66060
bytes to avoid unnecessary fragmentation. During situations of low
memory tipc_msg_build() may sometimes fail to allocate such large
buffers, resulting in unnecessary send failures. This can easily be
remedied by falling back to a smaller MTU, and then reassemble the
buffer chain as if the message were arriving from a remote node.
At the same time, we change the initial MTU setting of the broadcast
link to a lower value, so that large messages always are fragmented
into smaller buffers even when we run in single node mode. Apart from
obtaining the same advantage as for the 'fallback' solution above, this
turns out to give a significant performance improvement. This can
probably be explained with the __pskb_copy() operation performed on the
buffer for each recipient during reception. We found the optimal value
for this, considering the most relevant skb pool, to be 3744 bytes.
Acked-by: Ying Xue <ying.xue@ericsson.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Daniel Borkmann [Fri, 1 Dec 2017 19:59:21 +0000 (20:59 +0100)]
Merge branch 'bpf-nfp-jmp-memcpy-improvements'
Jiong Wang says:
====================
Currently, compiler will lower memcpy function call in XDP/eBPF C program
into a sequence of eBPF load/store pairs for some scenarios.
Compiler is thinking this "inline" optimiation is beneficial as it could
avoid function call and also increase code locality.
However, Netronome NPU is not an tranditional load/store architecture that
doing a sequence of individual load/store actions are not efficient.
This patch set tries to identify the load/store sequences composed of
load/store pairs that comes from memcpy lowering, then accelerates them
through NPU's Command Push Pull (CPP) instruction.
This patch set registered an new optimization pass before doing the actual
JIT work, it traverse through eBPF IR, once found candidate sequence then
record the memory copy source, destination and length information in the
first load instruction starting the sequence and marks all remaining
instructions in the sequence into skipable status. Later, when JITing the
first load instructoin, optimal instructions will be generated using those
record information.
For this safety of this transformation:
- jump into the middle of the sequence will cancel the optimization.
- overlapped memory access will cancel the optimization.
- the load destination register still contains the same value as before
the transformation.
====================
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Jiong Wang [Fri, 1 Dec 2017 05:33:00 +0000 (21:33 -0800)]
nfp: bpf: detect load/store sequences lowered from memory copy
This patch add the optimization frontend, but adding a new eBPF IR scan
pass "nfp_bpf_opt_ldst_gather".
The pass will traverse the IR to recognize the load/store pairs sequences
that come from lowering of memory copy builtins.
The gathered memory copy information will be kept in the meta info
structure of the first load instruction in the sequence and will be
consumed by the optimization backend added in the previous patches.
NOTE: a sequence with cross memory access doesn't qualify this
optimization, i.e. if one load in the sequence will load from place that
has been written by previous store. This is because when we turn the
sequence into single CPP operation, we are reading all contents at once
into NFP transfer registers, then write them out as a whole. This is not
identical with what the original load/store sequence is doing.
Detecting cross memory access for two random pointers will be difficult,
fortunately under XDP/eBPF's restrictied runtime environment, the copy
normally happen among map, packet data and stack, they do not overlap with
each other.
And for cases supported by NFP, cross memory access will only happen on
PTR_TO_PACKET. Fortunately for this, there is ID information that we could
do accurate memory alias check.
Signed-off-by: Jiong Wang <jiong.wang@netronome.com>
Reviewed-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Jiong Wang [Fri, 1 Dec 2017 05:32:59 +0000 (21:32 -0800)]
nfp: bpf: implement memory bulk copy for length bigger than 32-bytes
When the gathered copy length is bigger than 32-bytes and within 128-bytes
(the maximum length a single CPP Pull/Push request can finish), the
strategy of read/write are changeed into:
* Read.
- use direct reference mode when length is within 32-bytes.
- use indirect mode when length is bigger than 32-bytes.
* Write.
- length <= 8-bytes
use write8 (direct_ref).
- length <= 32-byte and 4-bytes aligned
use write32 (direct_ref).
- length <= 32-bytes but not 4-bytes aligned
use write8 (indirect_ref).
- length > 32-bytes and 4-bytes aligned
use write32 (indirect_ref).
- length > 32-bytes and not 4-bytes aligned and <= 40-bytes
use write32 (direct_ref) to finish the first 32-bytes.
use write8 (direct_ref) to finish all remaining hanging part.
- length > 32-bytes and not 4-bytes aligned
use write32 (indirect_ref) to finish those 4-byte aligned parts.
use write8 (direct_ref) to finish all remaining hanging part.
Signed-off-by: Jiong Wang <jiong.wang@netronome.com>
Reviewed-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Jiong Wang [Fri, 1 Dec 2017 05:32:58 +0000 (21:32 -0800)]
nfp: bpf: implement memory bulk copy for length within 32-bytes
For NFP, we want to re-group a sequence of load/store pairs lowered from
memcpy/memmove into single memory bulk operation which then could be
accelerated using NFP CPP bus.
This patch extends the existing load/store auxiliary information by adding
two new fields:
struct bpf_insn *paired_st;
s16 ldst_gather_len;
Both fields are supposed to be carried by the the load instruction at the
head of the sequence. "paired_st" is the corresponding store instruction at
the head and "ldst_gather_len" is the gathered length.
If "ldst_gather_len" is negative, then the sequence is doing memory
load/store in descending order, otherwise it is in ascending order. We need
this information to detect overlapped memory access.
This patch then optimize memory bulk copy when the copy length is within
32-bytes.
The strategy of read/write used is:
* Read.
Use read32 (direct_ref), always.
* Write.
- length <= 8-bytes
write8 (direct_ref).
- length <= 32-bytes and is 4-byte aligned
write32 (direct_ref).
- length <= 32-bytes but is not 4-byte aligned
write8 (indirect_ref).
NOTE: the optimization should not change program semantics. The destination
register of the last load instruction should contain the same value before
and after this optimization.
Signed-off-by: Jiong Wang <jiong.wang@netronome.com>
Reviewed-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Jiong Wang [Fri, 1 Dec 2017 05:32:57 +0000 (21:32 -0800)]
nfp: bpf: factor out is_mbpf_load & is_mbpf_store
It is usual that we need to check if one BPF insn is for loading/storeing
data from/to memory.
Therefore, it makes sense to factor out related code to become common
helper functions.
Signed-off-by: Jiong Wang <jiong.wang@netronome.com>
Reviewed-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Jakub Kicinski [Fri, 1 Dec 2017 05:32:56 +0000 (21:32 -0800)]
nfp: bpf: encode indirect commands
Add support for emitting commands with field overwrites.
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: Jiong Wang <jiong.wang@netronome.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Jiong Wang [Fri, 1 Dec 2017 05:32:55 +0000 (21:32 -0800)]
nfp: bpf: correct the encoding for No-Dest immed
When immed is used with No-Dest, the emitter should use reg.dst instead of
reg.areg for the destination, using the latter will actually encode
register zero.
Signed-off-by: Jiong Wang <jiong.wang@netronome.com>
Reviewed-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Jiong Wang [Fri, 1 Dec 2017 05:32:54 +0000 (21:32 -0800)]
nfp: bpf: relax source operands check
The NFP normally requires the source operands to be difference addressing
modes, but we should rule out the very special NN_REG_NONE type.
There are instruction that ignores both A/B operands, for example:
local_csr_rd
For these instructions, we might pass the same operand type, NN_REG_NONE,
for both A/B operands.
NOTE: in current NFP ISA, it is only possible for instructions with
unrestricted operands to take none operands, but in case there is new and
similar instructoin in restricted form, they would follow similar rules,
so swreg_to_restricted is updated as well.
Signed-off-by: Jiong Wang <jiong.wang@netronome.com>
Reviewed-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Jiong Wang [Fri, 1 Dec 2017 05:32:53 +0000 (21:32 -0800)]
nfp: bpf: don't do ld/shifts combination if shifts are jump destination
If any of the shift insns in the ld/shift sequence is jump destination,
don't do combination.
Signed-off-by: Jiong Wang <jiong.wang@netronome.com>
Reviewed-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Jiong Wang [Fri, 1 Dec 2017 05:32:52 +0000 (21:32 -0800)]
nfp: bpf: don't do ld/mask combination if mask is jump destination
If the mask insn in the ld/mask pair is jump destination, then don't do
combination.
Signed-off-by: Jiong Wang <jiong.wang@netronome.com>
Reviewed-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Jiong Wang [Fri, 1 Dec 2017 05:32:51 +0000 (21:32 -0800)]
nfp: bpf: flag jump destination to guide insn combine optimizations
NFP eBPF offload JIT engine is doing some instruction combine based
optimizations which however must not be safe if the combined sequences
are across basic block boarders.
Currently, there are post checks during fixing jump destinations. If the
jump destination is found to be eBPF insn that has been combined into
another one, then JIT engine will raise error and abort.
This is not optimal. The JIT engine ought to disable the optimization on
such cross-bb-border sequences instead of abort.
As there is no control flow information in eBPF infrastructure that we
can't do basic block based optimizations, this patch extends the existing
jump destination record pass to also flag the jump destination, then in
instruction combine passes we could skip the optimizations if insns in the
sequence are jump targets.
Suggested-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: Jiong Wang <jiong.wang@netronome.com>
Reviewed-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Jiong Wang [Fri, 1 Dec 2017 05:32:50 +0000 (21:32 -0800)]
nfp: bpf: record jump destination to simplify jump fixup
eBPF insns are internally organized as dual-list inside NFP offload JIT.
Random access to an insn needs to be done by either forward or backward
traversal along the list.
One place we need to do such traversal is at nfp_fixup_branches where one
traversal is needed for each jump insn to find the destination. Such
traversals could be avoided if jump destinations are collected through a
single travesal in a pre-scan pass, and such information could also be
useful in other places where jump destination info are needed.
This patch adds such jump destination collection in nfp_prog_prepare.
Suggested-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: Jiong Wang <jiong.wang@netronome.com>
Reviewed-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Jiong Wang [Fri, 1 Dec 2017 05:32:49 +0000 (21:32 -0800)]
nfp: bpf: support backward jump
This patch adds support for backward jump on NFP.
- restrictions on backward jump in various functions have been removed.
- nfp_fixup_branches now supports backward jump.
There is one thing to note, currently an input eBPF JMP insn may generate
several NFP insns, for example,
NFP imm move insn A \
NFP compare insn B --> 3 NFP insn jited from eBPF JMP insn M
NFP branch insn C /
---
NFP insn X --> 1 NFP insn jited from eBPF insn N
---
...
therefore, we are doing sanity check to make sure the last jited insn from
an eBPF JMP is a NFP branch instruction.
Once backward jump is allowed, it is possible an eBPF JMP insn is at the
end of the program. This is however causing trouble for the sanity check.
Because the sanity check requires the end index of the NFP insns jited from
one eBPF insn while only the start index is recorded before this patch that
we can only get the end index by:
start_index_of_the_next_eBPF_insn - 1
or for the above example:
start_index_of_eBPF_insn_N (which is the index of NFP insn X) - 1
nfp_fixup_branches was using nfp_for_each_insn_walk2 to expose *next* insn
to each iteration during the traversal so the last index could be
calculated from which. Now, it needs some extra code to handle the last
insn. Meanwhile, the use of walk2 is actually unnecessary, we could simply
use generic single instruction walk to do this, the next insn could be
easily calculated using list_next_entry.
So, this patch migrates the jump fixup traversal method to
*list_for_each_entry*, this simplifies the code logic a little bit.
The other thing to note is a new state variable "last_bpf_off" is
introduced to track the index of the last jited NFP insn. This is necessary
because NFP is generating special purposes epilogue sequences, so the index
of the last jited NFP insn is *not* always nfp_prog->prog_len - 1.
Suggested-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: Jiong Wang <jiong.wang@netronome.com>
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Jakub Kicinski [Fri, 1 Dec 2017 05:32:48 +0000 (21:32 -0800)]
nfp: fix old kdoc issues
Since commit
3a025e1d1c2e ("Add optional check for bad kernel-doc
comments") when built with W=1 build will complain about kdoc errors.
Fix the kdoc issues we have. kdoc is still confused by defines in
nfp_net_ctrl.h but those are not really errors.
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Daniel Borkmann [Fri, 1 Dec 2017 10:25:11 +0000 (11:25 +0100)]
Merge branch 'bpf-verifier-misc-improvements'
Alexei Starovoitov says:
====================
Small set of verifier improvements and cleanups which is
necessary for bigger patch set of bpf-to-bpf calls coming later.
See individual patches for details.
Tested on x86 and arm64 hw.
====================
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Alexei Starovoitov [Fri, 1 Dec 2017 05:31:41 +0000 (21:31 -0800)]
selftests/bpf: adjust test_align expected output
since verifier started to print liveness state of the registers
adjust expected output of test_align.
Now this test checks for both proper alignment handling by verifier
and correctness of liveness marks.
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Alexei Starovoitov [Fri, 1 Dec 2017 05:31:40 +0000 (21:31 -0800)]
bpf: cleanup register_is_null()
don't pass large struct bpf_reg_state by value.
Instead pass it by pointer.
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: John Fastabend <john.fastabend@gmail.com>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Alexei Starovoitov [Fri, 1 Dec 2017 05:31:39 +0000 (21:31 -0800)]
bpf: improve JEQ/JNE path walking
verifier knows how to trim paths that are known not to be
taken at run-time when register containing run-time constant
is compared with another constant.
It was done only for JEQ comparison.
Extend it to include JNE as well.
More cases can be added in the future.
before after
bpf_lb-DLB_L3.o 2270 2051
bpf_lb-DLB_L4.o 3682 3287
bpf_lb-DUNKNOWN.o 1110 1080
bpf_lxc-DDROP_ALL.o 27876 24980
bpf_lxc-DUNKNOWN.o 38780 34308
bpf_netdev.o 16937 15404
bpf_overlay.o 7929 7191
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: John Fastabend <john.fastabend@gmail.com>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Alexei Starovoitov [Fri, 1 Dec 2017 05:31:38 +0000 (21:31 -0800)]
bpf: improve verifier liveness marks
registers with pointers filled from stack were missing live_written marks
which caused liveness propagation to unnecessary mark more registers as
live_read and miss state pruning opportunities later on.
before after
bpf_lb-DLB_L3.o 2285 2270
bpf_lb-DLB_L4.o 3723 3682
bpf_lb-DUNKNOWN.o 1110 1110
bpf_lxc-DDROP_ALL.o 27954 27876
bpf_lxc-DUNKNOWN.o 38954 38780
bpf_netdev.o 16943 16937
bpf_overlay.o 7929 7929
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: John Fastabend <john.fastabend@gmail.com>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Alexei Starovoitov [Fri, 1 Dec 2017 05:31:37 +0000 (21:31 -0800)]
bpf: don't mark FP reg as uninit
when verifier hits an internal bug don't mark register R10==FP as uninit,
since it's read only register and it's not technically correct to let
verifier run further, since it may assume that R10 has valid auxiliary state.
While developing subsequent patches this issue was discovered,
though the code eventually changed that aux reg state doesn't have
pointers any more it is still safer to avoid clearing readonly register.
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: John Fastabend <john.fastabend@gmail.com>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Alexei Starovoitov [Fri, 1 Dec 2017 05:31:36 +0000 (21:31 -0800)]
bpf: print liveness info to verifier log
let verifier print register and stack liveness information
into verifier log
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: John Fastabend <john.fastabend@gmail.com>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>