Scott Feldman [Sun, 10 May 2015 16:47:59 +0000 (09:47 -0700)]
bridge: restore br_setlink back to original
This is revert of:
commit 68e331c785b8 ("bridge: offload bridge port attributes to switch asic
if feature flag set")
Restore br_setlink back to original and don't call into SELF port driver.
rtnetlink.c:bridge_setlink() already does a call into port driver for SELF.
bridge set link cmd defaults to MASTER. From man page for bridge link set
cmd:
self link setting is configured on specified physical device
master link setting is configured on the software bridge (default)
The link setting has two values: the device-side value and the software
bridge-side value. These are independent and settable using the bridge
link set cmd by specifying some combination of [master] | [self].
Furthermore, the device-side and bridge-side settings have their own
initial value, viewable from bridge -d link show cmd.
Restoring br_setlink back to original makes rocker (the only in-kernel user
of SELF link settings) work as first implement: two-sided values.
It's true that when both MASTER and SELF are specified from the command,
two netlink notifications are generated, one for each side of the settings.
The user-space app can distiquish between the two notifications by
observing the MASTER or SELF flag.
Signed-off-by: Scott Feldman <sfeldma@gmail.com> Acked-by: Jiri Pirko <jiri@resnulli.us> Signed-off-by: David S. Miller <davem@davemloft.net>
Scott Feldman [Sun, 10 May 2015 16:47:56 +0000 (09:47 -0700)]
switchdev: add new switchdev bridge setlink
Add new switchdev_port_bridge_setlink that can be used by drivers
implementing .ndo_bridge_setlink to set switchdev bridge attributes.
Basically turn the raw rtnl_bridge_setlink netlink into switchdev attr
sets. Proper netlink attr policy checking is done on the protinfo part of
the netlink msg.
Currently, for protinfo, only bridge port attrs BR_LEARNING and
BR_LEARNING_SYNC are parsed and passed to port driver.
For afspec, VLAN objs are passed so switchdev driver can set VLANs assigned
to SELF. To illustrate with iproute2 cmd, we have:
bridge vlan add vid 10 dev sw1p1 self master
To add VLAN 10 to port sw1p1 for both the bridge (master) and the device
(self).
Signed-off-by: Scott Feldman <sfeldma@gmail.com> Acked-by: Jiri Pirko <jiri@resnulli.us> Signed-off-by: David S. Miller <davem@davemloft.net>
Scott Feldman [Sun, 10 May 2015 16:47:52 +0000 (09:47 -0700)]
switchdev: introduce switchdev add/del obj ops
Like switchdev attr get/set, add new switchdev obj add/del. switchdev objs
will be things like VLANs or FIB entries, so add/del fits better for
objects than get/set used for attributes.
Use same two-phase prepare-commit transaction model as in attr set.
Signed-off-by: Scott Feldman <sfeldma@gmail.com> Acked-by: Sridhar Samudrala <sridhar.samudrala@intel.com> Acked-by: Jiri Pirko <jiri@resnulli.us> Signed-off-by: David S. Miller <davem@davemloft.net>
Scott Feldman [Sun, 10 May 2015 16:47:51 +0000 (09:47 -0700)]
switchdev: convert STP update to switchdev attr set
STP update is just a settable port attribute, so convert
switchdev_port_stp_update to an attr set.
For DSA, the prepare phase is skipped and STP updates are only done in the
commit phase. This is because currently the DSA drivers don't need to
allocate any memory for STP updates and the STP update will not fail to HW
(unless something horrible goes wrong on the MDIO bus, in which case the
prepare phase wouldn't have been able to predict anyway).
Signed-off-by: Scott Feldman <sfeldma@gmail.com> Acked-by: Jiri Pirko <jiri@resnulli.us> Signed-off-by: David S. Miller <davem@davemloft.net>
Scott Feldman [Sun, 10 May 2015 16:47:50 +0000 (09:47 -0700)]
rocker: support prepare-commit transaction model
For rocker, support prepare-commit transaction model for setting attributes
(and for adding objects). This requires rocker to preallocate memory
needed for the commit up front in the prepare phase. Since rtnl_lock is
held between prepare-commit, store the allocated memory on a queue hanging
off of the rocker_port. Also, in prepare phase, do everything right up to
calling into HW. The same code paths are tranversed in the driver for both
prepare and commit phases. In some cases, any state modified in the
prepare phase must be reverted before returning so the commit phase makes
the same decisions.
As a consequence of holding rtnl_lock in process context for all attr sets
(and obj adds), all memory is GFP_KERNEL allocated and we don't need to
busy spin waiting for the device to complete the command. So the bulk of
this patch is simplifying the memory allocations to only use GFP_KERNEL and
to remove the nowait flag and busy spin loop.
Signed-off-by: Scott Feldman <sfeldma@gmail.com> Acked-by: Jiri Pirko <jiri@resnulli.us> Signed-off-by: David S. Miller <davem@davemloft.net>
Scott Feldman [Sun, 10 May 2015 16:47:49 +0000 (09:47 -0700)]
switchdev: convert parent_id_get to switchdev attr get
Switch ID is just a gettable port attribute. Convert switchdev op
switchdev_parent_id_get to a switchdev attr.
Note: for sysfs and netlink interfaces, SWITCHDEV_ATTR_PORT_PARENT_ID is
called with SWITCHDEV_F_NO_RECUSE to limit switch ID user-visiblity to only
port netdevs. So when a port is stacked under bond/bridge, the user can
only query switch id via the switch ports, but not via the upper devices
Signed-off-by: Scott Feldman <sfeldma@gmail.com> Acked-by: Jiri Pirko <jiri@resnulli.us> Signed-off-by: David S. Miller <davem@davemloft.net>
Scott Feldman [Sun, 10 May 2015 16:47:48 +0000 (09:47 -0700)]
switchdev: introduce get/set attrs ops
Add two new swdev ops for get/set switch port attributes. Most swdev
interactions on a port are gets or sets on port attributes, so rather than
adding ops for each attribute, let's define clean get/set ops for all
attributes, and then we can have clear, consistent rules on how attributes
propagate on stacked devs.
Add the basic algorithms for get/set attr ops. Use the same recusive algo
to walk lower devs we've used for STP updates, for example. For get,
compare attr value for each lower dev and only return success if attr
values match across all lower devs. For sets, set the same attr value for
all lower devs. We'll use a two-phase prepare-commit transaction model for
sets. In the first phase, the driver(s) are asked if attr set is OK. If
all OK, the commit attr set in second phase. A driver would NACK the
prepare phase if it can't set the attr due to lack of resources or support,
within it's control. RTNL lock must be held across both phases because
we'll recurse all lower devs first in prepare phase, and then recurse all
lower devs again in commit phase. If any lower dev fails the prepare
phase, we need to abort the transaction for all lower devs.
If lower dev recusion isn't desired, allow a flag SWITCHDEV_F_NO_RECURSE to
indicate get/set only work on port (lowest) device.
Signed-off-by: Scott Feldman <sfeldma@gmail.com> Acked-by: Jiri Pirko <jiri@resnulli.us> Signed-off-by: David S. Miller <davem@davemloft.net>
Jiri Pirko [Sun, 10 May 2015 16:47:47 +0000 (09:47 -0700)]
switchdev: s/swdev_/switchdev_/
Turned out that "switchdev" sticks. So just unify all related terms to use
this prefix.
Signed-off-by: Jiri Pirko <jiri@resnulli.us> Signed-off-by: Scott Feldman <sfeldma@gmail.com> Acked-by: Roopa Prabhu <roopa@cumulusnetworks.com> Acked-by: Andy Gospodarek <gospo@cumulusnetworks.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Jiri Pirko [Sun, 10 May 2015 16:47:46 +0000 (09:47 -0700)]
switchdev: s/netdev_switch_/switchdev_/ and s/NETDEV_SWITCH_/SWITCHDEV_/
Turned out that "switchdev" sticks. So just unify all related terms to use
this prefix.
Signed-off-by: Jiri Pirko <jiri@resnulli.us> Signed-off-by: Scott Feldman <sfeldma@gmail.com> Acked-by: Roopa Prabhu <roopa@cumulusnetworks.com> Acked-by: Andy Gospodarek <gospo@cumulusnetworks.com> Signed-off-by: David S. Miller <davem@davemloft.net>
David Ward [Sun, 10 May 2015 02:01:46 +0000 (22:01 -0400)]
net_sched: gred: add TCA_GRED_LIMIT attribute
In a GRED qdisc, if the default "virtual queue" (VQ) does not have drop
parameters configured, then packets for the default VQ are not subjected
to RED and are only dropped if the queue is larger than the net_device's
tx_queue_len. This behavior is useful for WRED mode, since these packets
will still influence the calculated average queue length and (therefore)
the drop probability for all of the other VQs. However, for some drivers
tx_queue_len is zero. In other cases the user may wish to make the limit
the same for all VQs (including the default VQ with no drop parameters).
This change adds a TCA_GRED_LIMIT attribute to set the GRED queue limit,
in bytes, during qdisc setup. (This limit is in bytes to be consistent
with the drop parameters.) The default limit is the same as for a bfifo
queue (tx_queue_len * psched_mtu). If the drop parameters of any VQ are
configured with a smaller limit than the GRED queue limit, that VQ will
still observe the smaller limit instead.
Signed-off-by: David Ward <david.ward@ll.mit.edu> Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Tue, 12 May 2015 14:39:27 +0000 (10:39 -0400)]
Merge branch 'netdev_page_frags'
Alexander Duyck says:
====================
Refactor netdev page frags and move them into mm/
This patch series addresses several things.
First I found an issue in the performance of the pfmemalloc check from
build_skb. To work around it I have provided a cached copy of pfmemalloc
to be used in __netdev_alloc_skb and __napi_alloc_skb.
Second I moved the page fragment allocation logic into the mm tree and
added functionality for freeing page fragments. I had to fix igb before I
could do this as it was using a reference to NETDEV_FRAG_PAGE_MAX_SIZE
incorrectly.
Finally I went through and replaced all of the duplicate code that was
calling put_page and replaced it with calls to skb_free_frag.
With these changes in place a simple receive and drop test increased from a
packet rate of 8.9Mpps to 9.8Mpps. The gains breakdown as follows:
8.9Mpps Before 9.8Mpps After
------------------------ ------------------------
7.8% put_compound_page 9.1% __free_page_frag
3.9% skb_free_head
1.1% put_page
Alexander Duyck [Thu, 7 May 2015 04:12:20 +0000 (21:12 -0700)]
e1000: Replace e1000_free_frag with skb_free_frag
Signed-off-by: Alexander Duyck <alexander.h.duyck@redhat.com> Acked-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Alexander Duyck [Thu, 7 May 2015 04:12:03 +0000 (21:12 -0700)]
net: Add skb_free_frag to replace use of put_page in freeing skb->head
This change adds a function called skb_free_frag which is meant to
compliment the function netdev_alloc_frag. The general idea is to enable a
more lightweight version of page freeing since we don't actually need all
the overhead of a put_page, and we don't quite fit the model of __free_pages.
Signed-off-by: Alexander Duyck <alexander.h.duyck@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Alexander Duyck [Thu, 7 May 2015 04:11:57 +0000 (21:11 -0700)]
mm/net: Rename and move page fragment handling from net/ to mm/
This change moves the __alloc_page_frag functionality out of the networking
stack and into the page allocation portion of mm. The idea it so help make
this maintainable by placing it with other page allocation functions.
Since we are moving it from skbuff.c to page_alloc.c I have also renamed
the basic defines and structure from netdev_alloc_cache to page_frag_cache
to reflect that this is now part of a different kernel subsystem.
I have also added a simple __free_page_frag function which can handle
freeing the frags based on the skb->head pointer. The model for this is
based off of __free_pages since we don't actually need to deal with all of
the cases that put_page handles. I incorporated the virt_to_head_page call
and compound_order into the function as it actually allows for a signficant
size reduction by reducing code duplication.
Signed-off-by: Alexander Duyck <alexander.h.duyck@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Alexander Duyck [Thu, 7 May 2015 04:11:51 +0000 (21:11 -0700)]
net: Store virtual address instead of page in netdev_alloc_cache
This change makes it so that we store the virtual address of the page
in the netdev_alloc_cache instead of the page pointer. The idea behind
this is to avoid multiple calls to page_address since the virtual address
is required for every access, but the page pointer is only needed at
allocation or reset of the page.
While I was at it I also reordered the netdev_alloc_cache structure a bit
so that the size is always 16 bytes by dropping size in the case where
PAGE_SIZE is greater than or equal to 32KB.
Signed-off-by: Alexander Duyck <alexander.h.duyck@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Alexander Duyck [Thu, 7 May 2015 04:11:45 +0000 (21:11 -0700)]
igb: Don't use NETDEV_FRAG_PAGE_MAX_SIZE in descriptor calculation
This change updates igb so that it will correctly perform the descriptor
count calculation. Previously it was taking NETDEV_FRAG_PAGE_MAX_SIZE
into account with isn't really correct since a different value is used to
determine the size of the pages used for TCP. That is actually determined
by SKB_FRAG_PAGE_ORDER.
Signed-off-by: Alexander Duyck <alexander.h.duyck@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Alexander Duyck [Thu, 7 May 2015 04:11:40 +0000 (21:11 -0700)]
net: Use cached copy of pfmemalloc to avoid accessing page
While testing I found that the testing for pfmemalloc in build_skb was
rather expensive. I found the issue to be two-fold. First we have to get
from the virtual address to the head page and that comes at the cost of
something like 11 cycles. Then there is the cost for reading pfmemalloc out
of the head page which can be cache cold due to the fact that
put_page_testzero is likely invalidating the cache-line on one or more
CPUs as the fragments can be shared.
To avoid this extra expense I have added a pfmemalloc member to the
netdev_alloc_cache. I then pushed pieces of __alloc_rx_skb into
__napi_alloc_skb and __netdev_alloc_skb so that I could rewrite them to
make use of the cached pfmemalloc value. The result is that my perf traces
show a reduction from 9.28% overhead to 3.7% for the code covered by
build_skb, __alloc_rx_skb, and __napi_alloc_skb when performing a test with
the packet being dropped instead of being handed to napi_gro_receive.
Signed-off-by: Alexander Duyck <alexander.h.duyck@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Daniel Borkmann [Mon, 11 May 2015 17:28:49 +0000 (19:28 +0200)]
net: sched: fix typo in net_device ifdef
This should have been #ifdef not #if.
Reported-by: Fengguang Wu <fengguang.wu@intel.com> Fixes: d2788d34885d ("net: sched: further simplify handle_ing") Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Signed-off-by: David S. Miller <davem@davemloft.net>
Daniel Borkmann [Sat, 9 May 2015 20:51:32 +0000 (22:51 +0200)]
net: sched: further simplify handle_ing
Ingress qdisc has no other purpose than calling into tc_classify()
that executes attached classifier(s) and action(s).
It has a 1:1 relationship to dev->ingress_queue. After having commit 087c1a601ad7 ("net: sched: run ingress qdisc without locks") removed
the central ingress lock, one major contention point is gone.
The extra indirection layers however, are not necessary for calling
into ingress qdisc. pktgen calling locally into netif_receive_skb()
with a dummy u32, single CPU result on a Supermicro X10SLM-F, Xeon
E3-1240: before ~21,1 Mpps, after patch ~22,9 Mpps.
We can redirect the private classifier list to the netdev directly,
without changing any classifier API bits (!) and execute on that from
handle_ing() side. The __QDISC_STATE_DEACTIVATE test can be removed,
ingress qdisc doesn't have a queue and thus dev_deactivate_queue()
is also not applicable, ingress_cl_list provides similar behaviour.
In other words, ingress qdisc acts like TCQ_F_BUILTIN qdisc.
One next possible step is the removal of the dev's ingress (dummy)
netdev_queue, and to only have the list member in the netdevice
itself.
Note, the filter chain is RCU protected and individual filter elements
are being kfree'd by sched subsystem after RCU grace period. RCU read
lock is being held by __netif_receive_skb_core().
Joint work with Alexei Starovoitov.
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Signed-off-by: Alexei Starovoitov <ast@plumgrid.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Daniel Borkmann [Sat, 9 May 2015 20:51:31 +0000 (22:51 +0200)]
net: sched: consolidate handle_ing and ing_filter
Given quite some code has been removed from ing_filter(), we can just
consolidate that function into handle_ing() and get rid of a few
instructions at the same time.
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Acked-by: Alexei Starovoitov <ast@plumgrid.com> Signed-off-by: David S. Miller <davem@davemloft.net>
This is useful to ensure other JITs won't have a similar bug.
Link: https://lkml.org/lkml/2015/5/8/458 Cc: Alexei Starovoitov <ast@plumgrid.com> Cc: Will Deacon <will.deacon@arm.com> Signed-off-by: Xi Wang <xi.wang@gmail.com> Acked-by: Alexei Starovoitov <ast@plumgrid.com> Acked-by: Daniel Borkmann <daniel@iogearbox.net> Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Mon, 11 May 2015 14:59:32 +0000 (10:59 -0400)]
Merge branch 'bonding_netlink_lacp'
Jonathan Toppins says:
====================
add netlink support for new lacp bonding parameters
This is a resubmit of Mahesh's last 3 bonding patches from this series
(http://marc.info/?l=linux-netdev&m=142432864626179&w=2) with one
additional kernel patch which adds the netlink bits. I have noted any
modifications I did to the original patches just above my signoff line.
Patch 5 is the iproute2 support for these bonding options. All patches
were coded against the net-next branch of their respective projects.
v2:
* rebased
* only send these new parameters via netlink when bond is in mode 4
* fixed ad_actor_sys_prio to be 0xFFFF by default even when the bond
is initially created in mode 0 and switched to mode 4
v3:
* reverted changes to bond_option_ad_actor_system_set() from v1 in Mahesh's
patch "bonding: Allow userspace to set actors' macaddr in an AD-system."
Instead implementing all setting in the option specific set function as
Nik suggested.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Andy Gospodarek [Sat, 9 May 2015 07:01:58 +0000 (00:01 -0700)]
bonding: add netlink support for sys prio, actor sys mac, and port key
Adds netlink support for the following bonding options:
* BOND_OPT_AD_ACTOR_SYS_PRIO
* BOND_OPT_AD_ACTOR_SYSTEM
* BOND_OPT_AD_USER_PORT_KEY
When setting the actor system mac address we assume the netlink message
contains a binary mac and not a string representation of a mac.
Signed-off-by: Andy Gospodarek <gospo@cumulusnetworks.com>
[jt: completed the setting side of the netlink attributes] Signed-off-by: Jonathan Toppins <jtoppins@cumulusnetworks.com> Signed-off-by: Nikolay Aleksandrov <razor@blackwall.org> Signed-off-by: David S. Miller <davem@davemloft.net>
Mahesh Bandewar [Sat, 9 May 2015 07:01:57 +0000 (00:01 -0700)]
bonding: Implement user key part of port_key in an AD system.
The port key has three components - user-key, speed-part, and duplex-part.
The LSBit is for the duplex-part, next 5 bits are for the speed while the
remaining 10 bits are the user defined key bits. Get these 10 bits
from the user-space (through the SysFs interface) and use it to form the
admin port-key. Allowed range for the user-key is 0 - 1023 (10 bits). If
it is not provided then use zero for the user-key-bits (default).
It can set using following example code -
# modprobe bonding mode=4
# usr_port_key=$(( RANDOM & 0x3FF ))
# echo $usr_port_key > /sys/class/net/bond0/bonding/ad_user_port_key
# echo +eth1 > /sys/class/net/bond0/bonding/slaves
...
# ip link set bond0 up
Signed-off-by: Mahesh Bandewar <maheshb@google.com> Reviewed-by: Nikolay Aleksandrov <nikolay@redhat.com>
[jt: * fixed up style issues reported by checkpatch
* fixed up context from change in ad_actor_sys_prio patch] Signed-off-by: Jonathan Toppins <jtoppins@cumulusnetworks.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Mahesh Bandewar [Sat, 9 May 2015 07:01:56 +0000 (00:01 -0700)]
bonding: Allow userspace to set actors' macaddr in an AD-system.
In an AD system, the communication between actor and partner is the
business between these two entities. In the current setup anyone on the
same L2 can "guess" the LACPDU contents and then possibly send the
spoofed LACPDUs and trick the partner causing connectivity issues for
the AD system. This patch allows to use a random mac-address obscuring
it's identity making it harder for someone in the L2 is do the same thing.
This patch allows user-space to choose the mac-address for the AD-system.
This mac-address can not be NULL or a Multicast. If the mac-address is set
from user-space; kernel will honor it and will not overwrite it. In the
absence (value from user space); the logic will default to using the
masters' mac as the mac-address for the AD-system.
It can be set using example code below -
# modprobe bonding mode=4
# sys_mac_addr=$(printf '%02x:%02x:%02x:%02x:%02x:%02x' \
$(( (RANDOM & 0xFE) | 0x02 )) \
$(( RANDOM & 0xFF )) \
$(( RANDOM & 0xFF )) \
$(( RANDOM & 0xFF )) \
$(( RANDOM & 0xFF )) \
$(( RANDOM & 0xFF )))
# echo $sys_mac_addr > /sys/class/net/bond0/bonding/ad_actor_system
# echo +eth1 > /sys/class/net/bond0/bonding/slaves
...
# ip link set bond0 up
Signed-off-by: Mahesh Bandewar <maheshb@google.com> Reviewed-by: Nikolay Aleksandrov <nikolay@redhat.com>
[jt: fixed up style issues reported by checkpatch] Signed-off-by: Jonathan Toppins <jtoppins@cumulusnetworks.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Mahesh Bandewar [Sat, 9 May 2015 07:01:55 +0000 (00:01 -0700)]
bonding: Allow userspace to set actors' system_priority in AD system
This patch allows user to randomize the system-priority in an ad-system.
The allowed range is 1 - 0xFFFF while default value is 0xFFFF. If user
does not specify this value, the system defaults to 0xFFFF, which is
what it was before this patch.
Following example code could set the value -
# modprobe bonding mode=4
# sys_prio=$(( 1 + RANDOM + RANDOM ))
# echo $sys_prio > /sys/class/net/bond0/bonding/ad_actor_sys_prio
# echo +eth1 > /sys/class/net/bond0/bonding/slaves
...
# ip link set bond0 up
Signed-off-by: Mahesh Bandewar <maheshb@google.com> Reviewed-by: Nikolay Aleksandrov <nikolay@redhat.com>
[jt: * fixed up style issues reported by checkpatch
* changed how the default value is set in bond_check_params(), this
makes the default consistent between what gets set for a new bond
and what the default is claimed to be in the bonding options.] Signed-off-by: Jonathan Toppins <jtoppins@cumulusnetworks.com> Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Mon, 11 May 2015 14:50:19 +0000 (10:50 -0400)]
Merge branch 'kernel_socket_netns'
Eric W. Biederman says:
====================
Cleanup the kernel sockets.
Right now the situtation for allocating kernel sockets is a mess.
- sock_create_kern does not take a namespace parameter.
- kernel sockets must not reference count a network namespace and keep
it alive or else we will have a reference counting loop.
- The way we avoid the reference counting loop with sk_change_net
and sk_release_kernel are major hacks.
This patchset addresses this mess by fixing sock_create_kern to do
everything necessary to create a kernel socket. None of the current
users of kernel sockets need the network namespace reference counted.
Either kernel sockets are network namespace aware (and using the current
hacks) or kernel sockets are limited to the initial network namespace
in which case it does not matter.
This patchset starts by addressing tun which should be using normal
userspace sockets like macvtap.
Then sock_create_kern is fixed to take a network namespace.
Then the in kernel status of sockets are passed through to sk_alloc.
Then sk_alloc is fixed to not reference count the network namespace
of kernel sockets.
Then the callers of sock_create_kern are fixed up to stop using hacks.
Then netlink which uses it's own flavor of sock_create_kern is fixed.
Finally the hacks that are sk_change_net and sk_release_kernel are removed.
When it is all done the code is easier to follow, easier to use, easier
to maintain and shorter by about 70 lines.
====================
Reported-by: Ying Xue <ying.xue@windriver.com> Signed-off-by: David S. Miller <davem@davemloft.net>
net: Modify sk_alloc to not reference count the netns of kernel sockets.
Now that sk_alloc knows when a kernel socket is being allocated modify
it to not reference count the network namespace of kernel sockets.
Keep track of if a socket needs reference counting by adding a flag to
struct sock called sk_net_refcnt.
Update all of the callers of sock_create_kern to stop using
sk_change_net and sk_release_kernel as those hacks are no longer
needed, to avoid reference counting a kernel socket.
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com> Signed-off-by: David S. Miller <davem@davemloft.net>
net: Pass kern from net_proto_family.create to sk_alloc
In preparation for changing how struct net is refcounted
on kernel sockets pass the knowledge that we are creating
a kernel socket from sock_create_kern through to sk_alloc.
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com> Signed-off-by: David S. Miller <davem@davemloft.net>
tun: Utilize the normal socket network namespace refcounting.
There is no need for tun to do the weird network namespace refcounting.
The existing network namespace refcounting in tfile has almost exactly
the same lifetime. So rewrite the code to use the struct sock network
namespace refcounting and remove the unnecessary hand rolled network
namespace refcounting and the unncesary tfile->net.
This change allows the tun code to directly call sock_put bypassing
sock_release and making SOCK_EXTERNALLY_ALLOCATED unnecessary.
Remove the now unncessary tun_release so that if anything tries to use
the sock_release code path the kernel will oops, and let us know about
the bug.
The macvtap code already uses it's internal socket this way.
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com> Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Sun, 10 May 2015 02:26:06 +0000 (22:26 -0400)]
Merge branch 'pktgen-next'
Jesper Dangaard Brouer says:
====================
The following series introduce some pktgen changes
Patch01:
Cleanup my own work when I introduced NO_TIMESTAMP.
Patch02:
Took over patch from Alexei, and addressed my own concerns, as Alexie
is too busy with other work, and this will provide an easy tool for
measuring ingress path performance, which is a hot topic ATM.
Changes were primarily user interface related. Introduced a separate
"xmit_mode" setting, instead of stealing one of the dev flags like
Alexei did.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Introduce xmit_mode 'netif_receive' for pktgen which generates the
packets using familiar pktgen commands, but feeds them into
netif_receive_skb() instead of ndo_start_xmit().
Default mode is called 'start_xmit'.
It is designed to test netif_receive_skb and ingress qdisc
performace only. Make sure to understand how it works before
using it for other rx benchmarking.
Sample script 'pktgen.sh':
\#!/bin/bash
function pgset() {
local result
echo $1 > $PGDEV
result=`cat $PGDEV | fgrep "Result: OK:"`
if [ "$result" = "" ]; then
cat $PGDEV | fgrep Result:
fi
}
Raw netif_receive_skb speed should be ~43 million packet
per second on 3.7Ghz x86 and 'perf report' should look like:
37.69% kpktgend_0 [kernel.vmlinux] [k] __netif_receive_skb_core
25.81% kpktgend_0 [kernel.vmlinux] [k] kfree_skb
7.22% kpktgend_0 [kernel.vmlinux] [k] ip_rcv
5.68% kpktgend_0 [pktgen] [k] pktgen_thread_worker
If fib_table_lookup is seen on top, it means skb was processed
by the stack. To benchmark netif_receive_skb only make sure
that 'dst_mac' of your pktgen script is different from
receiving device mac and it will be dropped by ip_rcv
Signed-off-by: Alexei Starovoitov <ast@plumgrid.com> Signed-off-by: Jesper Dangaard Brouer <brouer@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
pktgen: adjust flag NO_TIMESTAMP to be more pktgen compliant
Allow flag NO_TIMESTAMP to turn timestamping on again, like other flags,
with a negation of the flag like !NO_TIMESTAMP.
Also document the option flag NO_TIMESTAMP.
Fixes: afb84b626184 ("pktgen: add flag NO_TIMESTAMP to disable timestamping") Signed-off-by: Jesper Dangaard Brouer <brouer@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Sun, 10 May 2015 02:15:31 +0000 (22:15 -0400)]
Merge branch 'netns-scalability'
Nicolas Dichtel says:
====================
netns: ease netlink use with a lot of netns
This idea was informally discussed in Ottawa / netdev0.1. The goal is to
ease the use/scalability of netns, from a userland point of view.
Today, users need to open one netlink socket per family and per netns.
Thus, when the number of netns inscreases (for example 5K or more), the
number of sockets needed to manage them grows a lot.
The goal of this series is to be able to monitor netlink events, for a
specified family, for a set of netns, with only one netlink socket. For
this purpose, a netlink socket option is added: NETLINK_LISTEN_ALL_NSID.
When this option is set on a netlink socket, this socket will receive
netlink notifications from all netns that have a nsid assigned into the
netns where the socket has been opened.
The nsid is sent to userland via an anscillary data.
Here is an example with a patched iproute2. vxlan10 is created in the
current netns (netns0, nsid 0) and then moved to another netns (netns1,
nsid 1):
$ ip netns exec netns0 ip monitor all-nsid label
[nsid 0][NSID]nsid 1 (iproute2 netns name: netns1)
[nsid 0][NEIGH]??? lladdr 00:00:00:00:00:00 REACHABLE,PERMANENT
[nsid 0][LINK]5: vxlan10@NONE: <BROADCAST,MULTICAST> mtu 1450 qdisc noop state DOWN group default
link/ether 92:33:17:e6:e7:1d brd ff:ff:ff:ff:ff:ff
[nsid 0][LINK]Deleted 5: vxlan10@NONE: <BROADCAST,MULTICAST> mtu 1450 qdisc noop state DOWN group default
link/ether 92:33:17:e6:e7:1d brd ff:ff:ff:ff:ff:ff
[nsid 1][NSID]nsid 0 (iproute2 netns name: netns0)
[nsid 1][LINK]5: vxlan10@NONE: <BROADCAST,MULTICAST> mtu 1450 qdisc noop state DOWN group default
link/ether 92:33:17:e6:e7:1d brd ff:ff:ff:ff:ff:ff link-netnsid 0
[nsid 1][ADDR]5: vxlan10 inet 192.168.0.249/24 brd 192.168.0.255 scope global vxlan10
valid_lft forever preferred_lft forever
[nsid 1][ROUTE]local 192.168.0.249 dev vxlan10 table local proto kernel scope host src 192.168.0.249
[nsid 1][ROUTE]ff00::/8 dev vxlan10 table local metric 256 pref medium
[nsid 1][ROUTE]2001:123::/64 dev vxlan10 proto kernel metric 256 pref medium
[nsid 1][LINK]5: vxlan10@NONE: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1450 qdisc noqueue state UNKNOWN group default
link/ether 92:33:17:e6:e7:1d brd ff:ff:ff:ff:ff:ff link-netnsid 0
[nsid 1][ROUTE]broadcast 192.168.0.255 dev vxlan10 table local proto kernel scope link src 192.168.0.249
[nsid 1][ROUTE]192.168.0.0/24 dev vxlan10 proto kernel scope link src 192.168.0.249
[nsid 1][ROUTE]broadcast 192.168.0.0 dev vxlan10 table local proto kernel scope link src 192.168.0.249
[nsid 1][ROUTE]fe80::/64 dev vxlan10 proto kernel metric 256 pref medium
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Nicolas Dichtel [Thu, 7 May 2015 09:02:53 +0000 (11:02 +0200)]
netlink: allow to listen "all" netns
More accurately, listen all netns that have a nsid assigned into the netns
where the netlink socket is opened.
For this purpose, a netlink socket option is added:
NETLINK_LISTEN_ALL_NSID. When this option is set on a netlink socket, this
socket will receive netlink notifications from all netns that have a nsid
assigned into the netns where the socket has been opened. The nsid is sent
to userland via an anscillary data.
With this patch, a daemon needs only one socket to listen many netns. This
is useful when the number of netns is high.
Because 0 is a valid value for a nsid, the field nsid_is_set indicates if
the field nsid is valid or not. skb->cb is initialized to 0 on skb
allocation, thus we are sure that we will never send a nsid 0 by error to
the userland.
Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Acked-by: Thomas Graf <tgraf@suug.ch> Signed-off-by: David S. Miller <davem@davemloft.net>
Nicolas Dichtel [Thu, 7 May 2015 09:02:52 +0000 (11:02 +0200)]
netlink: rename private flags and states
These flags and states have the same prefix (NETLINK_) that netlink socket
options. To avoid confusion and to be able to name a flag like a socket
option, let's use an other prefix: NETLINK_[S|F]_.
Note: a comment has been fixed, it was talking about
NETLINK_RECV_NO_ENOBUFS socket option instead of NETLINK_NO_ENOBUFS.
Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Acked-by: Thomas Graf <tgraf@suug.ch> Signed-off-by: David S. Miller <davem@davemloft.net>
Nicolas Dichtel [Thu, 7 May 2015 09:02:51 +0000 (11:02 +0200)]
netns: use a spin_lock to protect nsid management
Before this patch, nsid were protected by the rtnl lock. The goal of this
patch is to be able to find a nsid without needing to hold the rtnl lock.
The next patch will introduce a netlink socket option to listen to all
netns that have a nsid assigned into the netns where the socket is opened.
Thus, it's important to call rtnl_net_notifyid() outside the spinlock, to
avoid a recursive lock (nsid are notified via rtnl). This was the main
reason of the previous patch.
Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Nicolas Dichtel [Thu, 7 May 2015 09:02:50 +0000 (11:02 +0200)]
netns: notify new nsid outside __peernet2id()
There is no functional change with this patch. It will ease the refactoring
of the locking system that protects nsids and the support of the netlink
socket option NETLINK_LISTEN_ALL_NSID.
Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Acked-by: Thomas Graf <tgraf@suug.ch> Signed-off-by: David S. Miller <davem@davemloft.net>
Nicolas Dichtel [Thu, 7 May 2015 09:02:49 +0000 (11:02 +0200)]
netns: rename peernet2id() to peernet2id_alloc()
In a following commit, a new function will be introduced to only lookup for
a nsid (no allocation if the nsid doesn't exist). To avoid confusion, the
existing function is renamed.
Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Acked-by: Thomas Graf <tgraf@suug.ch> Signed-off-by: David S. Miller <davem@davemloft.net>
Nicolas Dichtel [Thu, 7 May 2015 09:02:48 +0000 (11:02 +0200)]
netns: always provide the id to rtnl_net_fill()
The goal of this commit is to prepare the rework of the locking of nsnid
protection.
After this patch, rtnl_net_notifyid() will not call anymore __peernet2id(),
ie no idr_* operation into this function.
Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Acked-by: Thomas Graf <tgraf@suug.ch> Signed-off-by: David S. Miller <davem@davemloft.net>
Nicolas Dichtel [Thu, 7 May 2015 09:02:47 +0000 (11:02 +0200)]
netns: returns always an id in __peernet2id()
All callers of this function expect a nsid, not an error.
Thus, returns NETNSA_NSID_NOT_ASSIGNED in case of error so that callers
don't have to convert the error to NETNSA_NSID_NOT_ASSIGNED.
Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Acked-by: Thomas Graf <tgraf@suug.ch> Signed-off-by: David S. Miller <davem@davemloft.net>
this is a pull request of a seven patches for net-next/master.
Andreas Gröger contributes two patches for the janz-ican3 driver. In
the first patch, the documentation for already existing sysfs entries
is added, the second patch adds support for another module/firmware
variant. A patch by Shawn Landden makes the padding in the struct
can_frame explicit. The next 4 patches target the flexcan driver, the
first one is by David Jander adding some documentation, the reaming
three by me add more documentation and two small code cleanups.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Harini Katakam [Wed, 6 May 2015 16:57:18 +0000 (22:27 +0530)]
net: macb: Add change_mtu callback with jumbo support
Add macb_change_mtu callback; if jumbo frame support is present allow
mtu size changes upto (jumbo max length allowed - headers).
Signed-off-by: Harini Katakam <harinik@xilinx.com> Reviewed-by: Punnaiah Choudary Kalluri <punnaia@xilinx.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Harini Katakam [Wed, 6 May 2015 16:57:17 +0000 (22:27 +0530)]
net: macb: Add support for jumbo frames
Enable jumbo frame support for Zynq Ultrascale+ MPSoC.
Update the NWCFG register and descriptor length masks accordingly.
Jumbo max length register should be set according to support in SoC; it is
set to 10240 for Zynq Ultrascale+ MPSoC.
Signed-off-by: Harini Katakam <harinik@xilinx.com> Reviewed-by: Punnaiah Choudary Kalluri <punnaia@xilinx.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Harini Katakam [Wed, 6 May 2015 16:57:16 +0000 (22:27 +0530)]
net: macb: Add compatible string for Zynq Ultrascale+ MPSoC
Add compatible string and config structure for Zynq Ultrascale+ MPSoC
Signed-off-by: Harini Katakam <harinik@xilinx.com> Reviewed-by: Punnaiah Choudary Kalluri <punnaia@xilinx.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Harini Katakam [Wed, 6 May 2015 16:57:15 +0000 (22:27 +0530)]
devicetree: Add compatible string for Zynq Ultrascale+ MPSoC
Add "cdns,zynqmp-gem" to be used for Zynq Ultrascale+ MPSoC.
Signed-off-by: Harini Katakam <harinik@xilinx.com> Reviewed-by: Punnaiah Choudary Kalluri <punnaia@xilinx.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Jason Baron [Wed, 6 May 2015 15:52:23 +0000 (15:52 +0000)]
tcp: set SOCK_NOSPACE under memory pressure
Under tcp memory pressure, calling epoll_wait() in edge triggered
mode after -EAGAIN, can result in an indefinite hang in epoll_wait(),
even when there is sufficient memory available to continue making
progress. The problem is that when __sk_mem_schedule() returns 0
under memory pressure, we do not set the SOCK_NOSPACE flag in the
tcp write paths (tcp_sendmsg() or do_tcp_sendpages()). Then, since
SOCK_NOSPACE is used to trigger wakeups when incoming acks create
sufficient new space in the write queue, all outstanding packets
are acked, but we never wake up with the the EPOLLOUT that we are
expecting from epoll_wait().
This issue is currently limited to epoll() when used in edge trigger
mode, since 'tcp_poll()', does in fact currently set SOCK_NOSPACE.
This is sufficient for poll()/select() and epoll() in level trigger
mode. However, in edge trigger mode, epoll() is relying on the write
path to set SOCK_NOSPACE. EPOLL(7) says that in edge-trigger mode we
can only call epoll_wait() after read/write return -EAGAIN. Thus, in
the case of the socket write, we are relying on the fact that
tcp_sendmsg()/network write paths are going to issue a wakeup for
us at some point in the future when we get -EAGAIN.
Normally, epoll() edge trigger works fine when we've exceeded the
sk->sndbuf because in that case we do set SOCK_NOSPACE. However, when
we return -EAGAIN from the write path b/c we are over the tcp memory
limits and not b/c we are over the sndbuf, we are never going to get
another wakeup.
I can reproduce this issue, using SO_SNDBUF, since __sk_mem_schedule()
will return 0, or failure more readily with SO_SNDBUF:
1) create socket and set SO_SNDBUF to N
2) add socket as edge trigger
3) write to socket and block in epoll on -EAGAIN
4) cause tcp mem pressure via: echo "<small val>" > net.ipv4.tcp_mem
The fix here is simply to set SOCK_NOSPACE in sk_stream_wait_memory()
when the socket is non-blocking. Note that SOCK_NOSPACE, in addition
to waking up outstanding waiters is also used to expand the size of
the sk->sndbuf. However, we will not expand it by setting it in this
case because tcp_should_expand_sndbuf(), ensures that no expansion
occurs when we are under tcp memory pressure.
Note that we could still hang if sk->sk_wmem_queue is 0, when we get
the -EAGAIN. In this case the SOCK_NOSPACE bit will not help, since we
are waiting for and event that will never happen. I believe
that this case is harder to hit (and did not hit in my testing),
in that over the tcp 'soft' memory limits, we continue to guarantee a
minimum write buffer size. Perhaps, we could return -ENOSPC in this
case, or maybe we simply issue a wakeup in this case, such that we
keep retrying the write. Note that this case is not specific to
epoll() ET, but rather would affect blocking sockets as well. So I
view this patch as bringing epoll() edge-trigger into sync with the
current poll()/select()/epoll() level trigger and blocking sockets
behavior.
Signed-off-by: Jason Baron <jbaron@akamai.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Claudiu Manoil [Wed, 6 May 2015 15:07:30 +0000 (18:07 +0300)]
gianfar: Enable changing mac addr when if up
Use device flag IFF_LIVE_ADDR_CHANGE to signal that
the device supports changing the hardware address when
the device is running.
This allows eth_mac_addr() to change the mac address
also when the network device's interface is open.
This capability is required by certain applications,
like bonding mode 6 (Adaptive Load Balancing).
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Claudiu Manoil [Wed, 6 May 2015 15:07:29 +0000 (18:07 +0300)]
gianfar: Move TxFIFO underrun handling to reset path
Handle TxFIFO underrun exceptions outside the fast path.
A controller reset is more reliable in this exceptional
case, as opposed to re-enabling on-the-fly the Tx DMA.
As the controller reset is handled outside the fast path
by the reset_gfar() workqueue handler, the locking
scheme on the Tx path is significantly simplified.
Because the Tx processing (xmit queues and tx napi) is
disabled during controller reset, tstat access from xmit
does not require locking. So the scope of the txlock on
the processing path is now reduced to num_txbdfree, which
is shared only between process context (xmit) and softirq
(clean_tx_ring). As a result, the txlock must not guard
against interrupt context, and the spin_lock_irqsave()
from xmit can be replaced by spin_lock_bh(). Likewise,
the locking has been downgraded for clean_tx_ring().
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com> Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Sat, 9 May 2015 21:35:05 +0000 (17:35 -0400)]
Merge branch 'bpf_seccomp'
Daniel Borkmann says:
====================
BPF updates
This set gets rid of BPF special handling in seccomp filter preparation
and provides generic infrastructure from BPF side, which eventually also
allows for classic BPF JITs to add support for seccomp filters.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Daniel Borkmann [Wed, 6 May 2015 14:12:30 +0000 (16:12 +0200)]
seccomp, filter: add and use bpf_prog_create_from_user from seccomp
Seccomp has always been a special candidate when it comes to preparation
of its filters in seccomp_prepare_filter(). Due to the extra checks and
filter rewrite it partially duplicates code and has BPF internals exposed.
This patch adds a generic API inside the BPF code code that seccomp can use
and thus keep it's filter preparation code minimal and better maintainable.
The other side-effect is that now classic JITs can add seccomp support as
well by only providing a BPF_LDX | BPF_W | BPF_ABS translation.
Tested with seccomp and BPF test suites.
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Cc: Nicolas Schichan <nschichan@freebox.fr> Cc: Alexei Starovoitov <ast@plumgrid.com> Cc: Kees Cook <keescook@chromium.org> Acked-by: Alexei Starovoitov <ast@plumgrid.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Daniel Borkmann [Wed, 6 May 2015 14:12:29 +0000 (16:12 +0200)]
net: filter: add __GFP_NOWARN flag for larger kmem allocs
When seccomp BPF was added, it was discussed to add __GFP_NOWARN
flag for their configuration path as f.e. up to 32K allocations are
more prone to fail under stress. As we're going to reuse BPF API,
add __GFP_NOWARN flags where larger kmalloc() and friends allocations
could fail.
It doesn't make much sense to pass around __GFP_NOWARN everywhere as
an extra argument only for seccomp while we just as well could run
into similar issues for socket filters, where it's not desired to
have a user application throw a WARN() due to allocation failure.
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Cc: Nicolas Schichan <nschichan@freebox.fr> Cc: Alexei Starovoitov <ast@plumgrid.com> Cc: Kees Cook <keescook@chromium.org> Acked-by: Alexei Starovoitov <ast@plumgrid.com> Signed-off-by: David S. Miller <davem@davemloft.net>
seccomp: simplify seccomp_prepare_filter and reuse bpf_prepare_filter
Remove the calls to bpf_check_classic(), bpf_convert_filter() and
bpf_migrate_runtime() and let bpf_prepare_filter() take care of that
instead.
seccomp_check_filter() is passed to bpf_prepare_filter() so that it
gets called from there, after bpf_check_classic().
We can now remove exposure of two internal classic BPF functions
previously used by seccomp. The export of bpf_check_classic() symbol,
previously known as sk_chk_filter(), was there since pre git times,
and no in-tree module was using it, therefore remove it.
Joint work with Daniel Borkmann.
Signed-off-by: Nicolas Schichan <nschichan@freebox.fr> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Cc: Alexei Starovoitov <ast@plumgrid.com> Cc: Kees Cook <keescook@chromium.org> Acked-by: Alexei Starovoitov <ast@plumgrid.com> Signed-off-by: David S. Miller <davem@davemloft.net>
net: filter: add a callback to allow classic post-verifier transformations
This is in preparation for use by the seccomp code, the rationale is
not to duplicate additional code within the seccomp layer, but instead,
have it abstracted and hidden within the classic BPF API.
As an interim step, this now also makes bpf_prepare_filter() visible
(not as exported symbol though), so that seccomp can reuse that code
path instead of reimplementing it.
Joint work with Daniel Borkmann.
Signed-off-by: Nicolas Schichan <nschichan@freebox.fr> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Cc: Alexei Starovoitov <ast@plumgrid.com> Cc: Kees Cook <keescook@chromium.org> Acked-by: Alexei Starovoitov <ast@plumgrid.com> Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Sat, 9 May 2015 21:27:25 +0000 (17:27 -0400)]
Merge tag 'mac80211-next-for-davem-2015-05-06' of git://git.kernel.org/pub/scm/linux/kernel/git/jberg/mac80211-next
Johannes Berg says:
====================
Lots of updates for net-next for this cycle. As usual, we have
a lot of small fixes and cleanups, the bigger items are:
* proper mac80211 rate control locking, to fix some random crashes
(this required changing other locking as well)
* mac80211 "fast-xmit", a mechanism to reduce, in most cases, the
amount of code we execute while going from ndo_start_xmit() to
the driver
* this also clears the way for properly supporting S/G and checksum
and segmentation offloads
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Eric Dumazet [Wed, 6 May 2015 21:26:25 +0000 (14:26 -0700)]
tcp: add TCPWinProbe and TCPKeepAlive SNMP counters
Diagnosing problems related to Window Probes has been hard because
we lack a counter.
TCPWinProbe counts the number of ACK packets a sender has to send
at regular intervals to make sure a reverse ACK packet opening back
a window had not been lost.
TCPKeepAlive counts the number of ACK packets sent to keep TCP
flows alive (SO_KEEPALIVE)
Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Yuchung Cheng <ycheng@google.com> Acked-by: Neal Cardwell <ncardwell@google.com> Acked-by: Nandita Dukkipati <nanditad@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Eric Dumazet [Wed, 6 May 2015 21:26:24 +0000 (14:26 -0700)]
tcp: adjust window probe timers to safer values
With the advent of small rto timers in datacenter TCP,
(ip route ... rto_min x), the following can happen :
1) Qdisc is full, transmit fails.
TCP sets a timer based on icsk_rto to retry the transmit, without
exponential backoff.
With low icsk_rto, and lot of sockets, all cpus are servicing timer
interrupts like crazy.
Intent of the code was to retry with a timer between 200 (TCP_RTO_MIN)
and 500ms (TCP_RESOURCE_PROBE_INTERVAL)
2) Receivers can send zero windows if they don't drain their receive queue.
TCP sends zero window probes, based on icsk_rto current value, with
exponential backoff.
With /proc/sys/net/ipv4/tcp_retries2 being 15 (or even smaller in
some cases), sender can abort in less than one or two minutes !
If receiver stops the sender, it obviously doesn't care of very tight
rto. Probability of dropping the ACK reopening the window is not
worth the risk.
Lets change the base timer to be at least 200ms (TCP_RTO_MIN) for these
events (but not normal RTO based retransmits)
A followup patch adds a new SNMP counter, as it would have helped a lot
diagnosing this issue.
Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Yuchung Cheng <ycheng@google.com> Acked-by: Neal Cardwell <ncardwell@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Richard Alpe [Wed, 6 May 2015 11:58:56 +0000 (13:58 +0200)]
tipc: send explicit not supported error in nl compat
The legacy netlink API treated EPERM (permission denied) as
"operation not supported".
Reported-by: Tomi Ollila <tomi.ollila@iki.fi> Signed-off-by: Richard Alpe <richard.alpe@ericsson.com> Reviewed-by: Erik Hugne <erik.hugne@ericsson.com> Reviewed-by: Ying Xue <ying.xue@windriver.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Richard Alpe [Wed, 6 May 2015 11:58:55 +0000 (13:58 +0200)]
tipc: add broadcast link window set/get to nl api
Add the ability to get or set the broadcast link window through the
new netlink API. The functionality was unintentionally missing from
the new netlink API. Adding this means that we also fix the breakage
in the old API when coming through the compat layer.
Fixes: 37e2d4843f9e (tipc: convert legacy nl link prop set to nl compat) Reported-by: Tomi Ollila <tomi.ollila@iki.fi> Signed-off-by: Richard Alpe <richard.alpe@ericsson.com> Reviewed-by: Erik Hugne <erik.hugne@ericsson.com> Reviewed-by: Ying Xue <ying.xue@windriver.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Richard Alpe [Wed, 6 May 2015 11:58:54 +0000 (13:58 +0200)]
tipc: fix default link prop regression in nl compat
Default link properties can be set for media or bearer. This
functionality was missed when introducing the NL compatibility layer.
This patch implements this functionality in the compat netlink
layer. It works the same way as it did in the old API. We search for
media and bearers matching the "link name". If we find a matching
media or bearer the link tolerance, priority or window is used as
default for new links on that media or bearer.
Fixes: 37e2d4843f9e (tipc: convert legacy nl link prop set to nl compat) Reported-by: Tomi Ollila <tomi.ollila@iki.fi> Signed-off-by: Richard Alpe <richard.alpe@ericsson.com> Reviewed-by: Erik Hugne <erik.hugne@ericsson.com> Reviewed-by: Ying Xue <ying.xue@windriver.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Implements t4_init_rss_mode() to initialize the rss_mode for all the ports. If
Tunnel All Lookup isn't specified in the global RSS Configuration, then we need
to specify a default Ingress Queue for any ingress packets which aren't hashed.
We'll use our first ingress queue.
Signed-off-by: Hariprasad Shenai <hariprasad@chelsio.com> Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Sat, 9 May 2015 20:27:04 +0000 (16:27 -0400)]
Merge branch 'be2net'
Sathya Perla says:
====================
be2net: patch-set
The following patch-set has two new feature additions, and a few
minor fixes and cleanups.
Pls consider applying to the net-next tree. Thanks.
v2 changes:
a) dropped the "don't enable pause by default" patch
b) described how the "spoof check" works in patch 1's commit log
c) I had to update our email addresses from "@emulex" to
"@avagotech". I'll send a separate patch updating the
maintainers
Patch 1 adds support for the "spoofchk" knob for VFs.
When it is enabled, "spoof checking" is done for both MAC-address
and VLAN. For each VF, the HW ensures that the source MAC address
(or vlan) of every outgoing packet of the VF exists in the MAC-list
(or vlan-list) configured for RX filtering for that VF.
If not, the packet is dropped and an error is reported to the driver
in the TX completion.
Patch 2 improves interrupt moderation on Skyhawk-R chip by using
the EQ-DB mechanism to set a "re-arm to interrupt" delay. Currently
interrupt moderation is adjusted by calculating and configuring an
EQ-delay every second. This is done via a FW-cmd. This patch uses
the EQ_DB facility to calculate and set the interrupt delay every 1ms.
This helps moderating interrupts better when the traffic is bursty.
Patch 3 adds L3/L4 error accounting to BE3 VFs, by passing L3/4 error
packets to the network stack.
Patch 4 adds an extra FW-cmd error value check in the driver to identify
an "out of vlan filters" scenario.
Patch 5 stops enabling pause by default as this setting fails in
some HW-configs where priority pause is enabled in FW. If the user
tries to do the same, an appropriate error is returned via ethtool.
Patch 5 posts the full RXQ in be_open() to prevent packet drops due to
bursty traffic when the interface is enabled.
Patch 6 refactors the be_check_ufi_compatibility() routine, that checks
to see if a UFI file meant for a lower rev of a chip is being flashed
on a higher rev, to make it simpler.
Patch 7 replaces the usage of !be_physfn() macro with be_virtfn()
that is already avialble in the driver.
Patch 8 updates the year in the copyright text to 2015.
Path 9 bumps up the driver version to 10.6.02.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Vasundhara Volam <vasundhara.volam@avagotech.com> Signed-off-by: Sathya Perla <sathya.perla@avagotech.com> Signed-off-by: David S. Miller <davem@davemloft.net>
The code in be_check_ufi_compatibility() checks to see if a UFI file meant
for a lower rev of a chip is being flashed on a higher rev, which is
disallowed. This patch re-writes the code needed for this check in a much
simpler manner.
Signed-off-by: Vasundhara Volam <vasundhara.volam@avagotech.com> Signed-off-by: Sathya Perla <sathya.perla@avagotech.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Suresh Reddy [Wed, 6 May 2015 09:30:36 +0000 (05:30 -0400)]
be2net: post full RXQ on interface enable
When an RXQ is created in be_open(), the driver currently posts only
64 buffers. This sometimes results in packet drops when there is a traffic
burst as soon as the interface is enabled.
This patch fixes this problem by posting the full RXQ on interface enable.
Signed-off-by: Suresh Reddy <Suresh.Reddy@avagotech.com> Signed-off-by: Sathya Perla <sathya.perla@avagotech.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Kalesh AP [Wed, 6 May 2015 09:30:35 +0000 (05:30 -0400)]
be2net: check for INSUFFICIENT_VLANS error
When the FW runs out of vlan filters it can either return an
INSUFFICIENT_RESOURCES error or an INSUFFICIENT_VLANS error.
The driver currently checks only for the former error value.
This patch adds a check for the latter value too.
Signed-off-by: Kalesh AP <kalesh.purayil@emulex.com> Signed-off-by: Sathya Perla <sathya.perla@avagotech.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Somnath Kotur [Wed, 6 May 2015 09:30:34 +0000 (05:30 -0400)]
be2net: receive pkts with L3, L4 errors on VFs
Currently pkts with L3 or L4 errors received on PFs are not dropped
by the adapter, but instead sent to the stack. This helps the network stack
to better reflect error statistics. This was not being done on BE3 VFs.
This patch fixes this for BE3 VFs.
Signed-off-by: Somnath Kotur <somnath.kotur@avagotech.com> Signed-off-by: Sathya Perla <sathya.perla@avagotech.com> Signed-off-by: David S. Miller <davem@davemloft.net>
be2net: set interrupt moderation for Skyhawk-R using EQ-DB
Currently adaptive interrupt moderation is set by calculating
and configuring an EQ-delay every second. This is done via
a FW-cmd. But, on Skyhawk-R a "re-arm to interrupt" delay
can be set while ringing the EQ-DB. This patch uses this
facility to calculate and set the interrupt delay every 1ms.
This helps moderating interrupts better when the traffic
is bursty.
Signed-off-by: Padmanabh Ratnakar <padmanabh.ratnakar@avagotech.com> Signed-off-by: Sathya Perla <sathya.perla@avagotech.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Kalesh AP [Wed, 6 May 2015 09:30:32 +0000 (05:30 -0400)]
be2net: add support for spoofchk setting
This patch adds support for spoofchk configuration for VFs.
When it is enabled, "spoof checking" is done for both MAC-address and VLAN.
For each VF, the HW ensures that the source MAC address (or vlan) of
every outgoing packet exists in the MAC-list (or vlan-list) configured
for RX filtering for that VF. If not, the packet is dropped and an error
is reported to the driver in the TX completion; this is reflected in the
"tx_spoof_check_err" ethtool counter.
This feature is supported in Skyhawk FW version 10.6.31.0 and above.
Signed-off-by: Kalesh AP <kalesh.purayil@emulex.com> Signed-off-by: Sathya Perla <sathya.perla@avagotech.com> Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Sat, 9 May 2015 20:16:49 +0000 (16:16 -0400)]
Merge branch 'sfc-next'
Shradha Shah says:
====================
sfc: Enabling EF10 Vf's, set up vswitching and bind the SFC driver to the VF's
This set of patches makes way for the implementation of EF10
SR-IOV driver starting with some cleanup code.
NIC specific SR-IOV functions are moved to their own header
and netdev_ops are made generic instead of being NIC specific
Next in line comes the patch to enable VF's using sriov_configure.
VEB vswitching hierarchy is set up next followed by patches to
prepare sfc driver to bind to enabled VF's
This is followed by patch to support use of shared RSS contexts
which makes VF's use shared RSS contexts in all cases.
Patch series ends with a patch to bind the sfc driver to the
enabled VF's which creates network interfaces corresponding to
the VF's.
Coming up soon are the patches to set_vf_mac, set_vf_config,
set_vf_vlan, vf_spoofcheck, etc.
These patches have been tested with and without CONFIG_SFC_SRIOV.
In the case of CONFIG_SFC_SRIOV=y enabling of VF's using
sriov_configure is also tested. The enabled VF's bind to the
installed sfc driver succesfully to create network interfaces.
In the case of CONFIG_SFC_SRIOV=n enabling of VF's using
sriov_configure returns the correct error message:
"Function not implemented".
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Shradha Shah [Wed, 6 May 2015 00:00:07 +0000 (01:00 +0100)]
sfc: Bind the sfc driver to any available VF's
Add the device ID of the VF to the PCI device ID table.
Added a boolean flag is_vf in efx_nic_type to differentiate
between a VF and PF at probe time. This flag is useful in later
patches while setting MAC address specially in the
PCI-passthrough case.
Signed-off-by: Shradha Shah <sshah@solarflare.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Jon Cooper [Tue, 5 May 2015 23:59:38 +0000 (00:59 +0100)]
sfc: Add use of shared RSS contexts.
Allow PFs to allocate shared RSS contexts if we exhaust our
exclusive RSS contexts. Make VFs use shared RSS contexts in
all cases.
Spruce up error handling so that the shadow copy of the RSS
table is updated after successful update, rather than in all
cases, so that we report the actual contents of the RSS table
after a failure to set it, rather than what we'd like it to be.
Populate context_size parameter when vacuously allocating RSS
context of size 1.
Signed-off-by: Shradha Shah <sshah@solarflare.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Edward Cree [Tue, 5 May 2015 23:59:18 +0000 (00:59 +0100)]
sfc: Cope with permissions enforcement added to firmware for SR-IOV
* Accept EPERM in some simple cases, the following cases are handled:
1) efx_mcdi_read_assertion()
Unprivileged PCI functions aren't allowed to GET_ASSERTS.
We return success as it's up to the primary PF to deal with asserts.
2) efx_mcdi_mon_probe() in efx_ef10_probe()
Unprivileged PCI functions aren't allowed to read sensor info, and
worrying about sensor data is the primary PF's job.
3) phy_op->reconfigure() in efx_init_port() and efx_reset_up()
Unprivileged functions aren't allowed to MC_CMD_SET_LINK, they just have
to accept the settings (including flow-control, which is what
efx_init_port() is worried about) they've been given.
4) Fallback to GET_WORKAROUNDS in efx_ef10_probe()
Unprivileged PCI functions aren't allowed to set workarounds. So if
efx_mcdi_set_workaround() fails EPERM, use efx_mcdi_get_workarounds()
to find out if workaround_35388 is enabled.
5) If DRV_ATTACH gets EPERM, try without specifying fw-variant
Unprivileged PCI functions have to use a FIRMWARE_ID of 0xffffffff
(MC_CMD_FW_DONT_CARE).
6) Don't try to exit_assertion unless one had fired
Previously we called efx_mcdi_exit_assertion even if
efx_mcdi_read_assertion had received MC_CMD_GET_ASSERTS_FLAGS_NO_FAILS.
This is unnecessary, and the resulting MC_CMD_REBOOT, even if the
AFTER_ASSERTION flag made it a no-op, would fail EPERM for unprivileged
PCI functions.
So make efx_mcdi_read_assertion return whether an assert happened, and only
call efx_mcdi_exit_assertion if it has.
Signed-off-by: Shradha Shah <sshah@solarflare.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Shradha Shah [Tue, 5 May 2015 23:58:54 +0000 (00:58 +0100)]
sfc: manually allocate and free vadaptors
To be able to use MC_CMD_VADAPTOR_SET_MAC, vadaptors must be
manually allocated and freed as automatic vadaptors will disappear
when their reference_count reaches zero, which must happen before
the MAC address is changed.
Vadaptors are allocated and freed in the vswitching_probe/remove
functions for PFs and VFs, and this means that vadaptors are restored
correctly following an MC reboot or other reset when required.
Signed-off-by: Shradha Shah <sshah@solarflare.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Shradha Shah [Tue, 5 May 2015 23:58:31 +0000 (00:58 +0100)]
sfc: create vports for VFs and assign random MAC addresses
The parent PF creates vports for all its child VFs and adds MAC
addresses to these. When the VF driver loads, it can make an MCDI
call to get the MAC address that the parent PF assigned it.
The parent PF also assigns a mac address to its own vport because
implicit creation of a vAdaptor will only work on evb ports with
MAC addresses assigned.
The vport MAC address needs to be stored in the PF's nic_data
struct as it can later be changed on the vadaptor (and its net_dev
struct). When removing a vport the original MAC address must be
deleted.
A new flag is needed in the VF data structure to identify whether
a vport has been assigned to the VF. This is to determine whether
it needs to be un-assigned before freeing the vport. Also,
attempting to un-assign a vport which is not assigned will result
in an EALREADY error.
Signed-off-by: Shradha Shah <sshah@solarflare.com> Signed-off-by: David S. Miller <davem@davemloft.net>
projects / linux-beck.git / log