When many cpus compete for sending frames on a given qdisc, the qdisc
spinlock suffers from very high contention.
The cpu owning __QDISC_STATE_RUNNING bit has same priority to acquire
the lock, and cannot dequeue packets fast enough, since it must wait for
this lock for each dequeued packet.
One solution to this problem is to force all cpus spinning on a second
lock before trying to get the main lock, when/if they see
__QDISC_STATE_RUNNING already set.
The owning cpu then compete with at most one other cpu for the main
lock, allowing for higher dequeueing rate.
Based on a previous patch from Alexander Duyck. I added the heuristic to
avoid the atomic in fast path, and put the new lock far away from the
cache line used by the dequeue worker. Also try to release the busylock
lock as late as possible.
Tests with following script gave a boost from ~50.000 pps to ~600.000
pps on a dual quad core machine (E5450 @3.00GHz), tg3 driver.
(A single netperf flow can reach ~800.000 pps on this platform)
for j in `seq 0 3`; do
for i in `seq 0 7`; do
netperf -H 192.168.0.1 -t UDP_STREAM -l 60 -N -T $i -- -m 6 &
done
done
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Acked-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Define three helpers to manipulate QDISC_STATE_RUNNIG flag, that a
second patch will move on another location.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Currently such notifications are only generated when the device comes up or the
address changes. However one use case for these notifications is to enable
faster network recovery after a virtual machine migration (by causing switches
to relearn their MAC tables). A migration appears to the network stack as a
temporary loss of carrier and therefore does not trigger either of the current
conditions. Rather than adding carrier up as a trigger (which can cause issues
when interfaces a flapping) simply add an interface which the driver can use
to explicitly trigger the notification.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Cc: Stephen Hemminger <shemminger@linux-foundation.org>
Cc: Jeremy Fitzhardinge <jeremy@goop.org>
Cc: David S. Miller <davem@davemloft.net>
Cc: netdev@vger.kernel.org
Cc: stable@kernel.org
Signed-off-by: David S. Miller <davem@davemloft.net>
Up until now cls_cgroup has relied on fetching the classid out of
the current executing thread. This runs into trouble when a packet
processing is delayed in which case it may execute out of another
thread's context.
Furthermore, even when a packet is not delayed we may fail to
classify it if soft IRQs have been disabled, because this scenario
is indistinguishable from one where a packet unrelated to the
current thread is processed by a real soft IRQ.
In fact, the current semantics is inherently broken, as a single
skb may be constructed out of the writes of two different tasks.
A different manifestation of this problem is when the TCP stack
transmits in response of an incoming ACK. This is currently
unclassified.
As we already have a concept of packet ownership for accounting
purposes in the skb->sk pointer, this is a natural place to store
the classid in a persistent manner.
This patch adds the cls_cgroup classid in struct sock, filling up
an existing hole on 64-bit :)
The value is set at socket creation time. So all sockets created
via socket(2) automatically gains the ID of the thread creating it.
Whenever another process touches the socket by either reading or
writing to it, we will change the socket classid to that of the
process if it has a valid (non-zero) classid.
For sockets created on inbound connections through accept(2), we
inherit the classid of the original listening socket through
sk_clone, possibly preceding the actual accept(2) call.
In order to minimise risks, I have not made this the authoritative
classid. For now it is only used as a backup when we execute
with soft IRQs disabled. Once we're completely happy with its
semantics we can use it as the sole classid.
Footnote: I have rearranged the error path on cls_group module
creation. If we didn't do this, then there is a window where
someone could create a tc rule using cls_group before the cgroup
subsystem has been registered.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
Ben Pfaff reported a kernel oops and provided a test program to
reproduce it.
https://kerneltrap.org/mailarchive/linux-netdev/2010/5/21/6277805
tc_fill_qdisc() should not be called for builtin qdisc, or it
dereference a NULL pointer to get device ifindex.
Fix is to always use tc_qdisc_dump_ignore() before calling
tc_fill_qdisc().
Reported-by: Ben Pfaff <blp@nicira.com>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch removes from net/ (but not any netfilter files)
all the unnecessary return; statements that precede the
last closing brace of void functions.
It does not remove the returns that are immediately
preceded by a label as gcc doesn't like that.
Done via:
$ grep -rP --include=*.[ch] -l "return;\n}" net/ | \
xargs perl -i -e 'local $/ ; while (<>) { s/\n[ \t\n]+return;\n}/\n}/g; print; }'
Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
If the user has a bad classification configuration, and gets a packet
that goes through too many steps. Chances are more packets will arrive,
and the message spew will overrun syslog because it is not rate limited.
And because it is not tagged with appropriate priority it can't not be screened.
Added the qdisc to the message to try and give some more context when
the message does arrive.
Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Acked-by: Jamal Hadi Salim <hadi@cyberus.ca>
Signed-off-by: David S. Miller <davem@davemloft.net>
The previous patch encourage me to go look at all the messages in
the network scheduler and fix them. Many messages were missing
any severity level. Some serious ones that should never happen
were turned into WARN(), and the random noise messages that were
handled changed to pr_debug().
Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When attaching filters to a class pointing to a class higher up in the
hierarchy, classification may enter an endless loop. Currently this is
prevented for filters that are already resolved, but not for filters
resolved at runtime.
Only allow filters to point downwards in the hierarchy, similar to what
CBQ does.
Reported-by: Pawel Staszewski <pstaszewski@itcare.pl>
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
Several netem users use TBF for rate control. But every time the parameters
of TBF are changed it destroys the child qdisc, requiring reconfigation.
Better to just keep child qdisc and just notify it of changed limit.
Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Use low order bit of skb->_skb_dst to tell dst is not refcounted.
Change _skb_dst to _skb_refdst to make sure all uses are catched.
skb_dst() returns the dst, regardless of noref bit set or not, but
with a lockdep check to make sure a noref dst is not given if current
user is not rcu protected.
New skb_dst_set_noref() helper to set an notrefcounted dst on a skb.
(with lockdep check)
skb_dst_drop() drops a reference only if skb dst was refcounted.
skb_dst_force() helper is used to force a refcount on dst, when skb
is queued and not anymore RCU protected.
Use skb_dst_force() in __sk_add_backlog(), __dev_xmit_skb() if
!IFF_XMIT_DST_RELEASE or skb enqueued on qdisc queue, in
sock_queue_rcv_skb(), in __nf_queue().
Use skb_dst_force() in dev_requeue_skb().
Note: dst_use_noref() still dirties dst, we might transform it
later to do one dirtying per jiffies.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Per cpu variable softnet_data.total was shared between IRQ and SoftIRQ context
without any protection. And enqueue_to_backlog should update the netdev_rx_stat
of the target CPU.
This patch renames softnet_data.total to softnet_data.processed: the number of
packets processed in uppper levels(IP stacks).
softnet_stat data is moved into softnet_data.
Signed-off-by: Changli Gao <xiaosuo@gmail.com>
----
include/linux/netdevice.h | 17 +++++++----------
net/core/dev.c | 26 ++++++++++++--------------
net/sched/sch_generic.c | 2 +-
3 files changed, 20 insertions(+), 25 deletions(-)
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Sparse can help us find endianness bugs, but we need to make some
cleanups to be able to more easily spot real bugs.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
One of my test machine got a deadlock during "tc" sessions,
adding/deleting classes & filters, using traffic estimators.
After some analysis, I believe we have a potential use after free case
in est_timer() :
spin_lock(e->stats_lock); << HERE >>
read_lock(&est_lock);
if (e->bstats == NULL) << TEST >>
goto skip;
Test is done a bit late, because after estimator is killed, and before
rcu grace period elapsed, we might already have freed/reuse memory where
e->stats_locks points to (some qdisc->q.lock)
A possible fix is to respect a rcu grace period at Qdisc dismantle time.
On 64bit, sizeof(struct Qdisc) is exactly 192 bytes. Adding 16 bytes to
it (for struct rcu_head) is a problem because it might change
performance, given QDISC_ALIGNTO is 32 bytes.
This is why I also change QDISC_ALIGNTO to 64 bytes, to satisfy most
current alignment requirements.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
These changes were suggested by Alexey Dobriyan <adobriyan@gmail.com>:
- psched_show() does not use any private data so just pass NULL to
psched_open()
- remove unnecessary return statement
Signed-off-by: Tom Goff <thomas.goff@boeing.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
percpu.h is included by sched.h and module.h and thus ends up being
included when building most .c files. percpu.h includes slab.h which
in turn includes gfp.h making everything defined by the two files
universally available and complicating inclusion dependencies.
percpu.h -> slab.h dependency is about to be removed. Prepare for
this change by updating users of gfp and slab facilities include those
headers directly instead of assuming availability. As this conversion
needs to touch large number of source files, the following script is
used as the basis of conversion.
http://userweb.kernel.org/~tj/misc/slabh-sweep.py
The script does the followings.
* Scan files for gfp and slab usages and update includes such that
only the necessary includes are there. ie. if only gfp is used,
gfp.h, if slab is used, slab.h.
* When the script inserts a new include, it looks at the include
blocks and try to put the new include such that its order conforms
to its surrounding. It's put in the include block which contains
core kernel includes, in the same order that the rest are ordered -
alphabetical, Christmas tree, rev-Xmas-tree or at the end if there
doesn't seem to be any matching order.
* If the script can't find a place to put a new include (mostly
because the file doesn't have fitting include block), it prints out
an error message indicating which .h file needs to be added to the
file.
The conversion was done in the following steps.
1. The initial automatic conversion of all .c files updated slightly
over 4000 files, deleting around 700 includes and adding ~480 gfp.h
and ~3000 slab.h inclusions. The script emitted errors for ~400
files.
2. Each error was manually checked. Some didn't need the inclusion,
some needed manual addition while adding it to implementation .h or
embedding .c file was more appropriate for others. This step added
inclusions to around 150 files.
3. The script was run again and the output was compared to the edits
from #2 to make sure no file was left behind.
4. Several build tests were done and a couple of problems were fixed.
e.g. lib/decompress_*.c used malloc/free() wrappers around slab
APIs requiring slab.h to be added manually.
5. The script was run on all .h files but without automatically
editing them as sprinkling gfp.h and slab.h inclusions around .h
files could easily lead to inclusion dependency hell. Most gfp.h
inclusion directives were ignored as stuff from gfp.h was usually
wildly available and often used in preprocessor macros. Each
slab.h inclusion directive was examined and added manually as
necessary.
6. percpu.h was updated not to include slab.h.
7. Build test were done on the following configurations and failures
were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my
distributed build env didn't work with gcov compiles) and a few
more options had to be turned off depending on archs to make things
build (like ipr on powerpc/64 which failed due to missing writeq).
* x86 and x86_64 UP and SMP allmodconfig and a custom test config.
* powerpc and powerpc64 SMP allmodconfig
* sparc and sparc64 SMP allmodconfig
* ia64 SMP allmodconfig
* s390 SMP allmodconfig
* alpha SMP allmodconfig
* um on x86_64 SMP allmodconfig
8. percpu.h modifications were reverted so that it could be applied as
a separate patch and serve as bisection point.
Given the fact that I had only a couple of failures from tests on step
6, I'm fairly confident about the coverage of this conversion patch.
If there is a breakage, it's likely to be something in one of the arch
headers which should be easily discoverable easily on most builds of
the specific arch.
Signed-off-by: Tejun Heo <tj@kernel.org>
Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
Allows the net_cls cgroup subsystem to be compiled as a module
This patch modifies net/sched/cls_cgroup.c to allow the net_cls subsystem
to be optionally compiled as a module instead of builtin. The
cgroup_subsys struct is moved around a bit to allow the subsys_id to be
either declared as a compile-time constant by the cgroup_subsys.h include
in cgroup.h, or, if it's a module, initialized within the struct by
cgroup_load_subsys.
Signed-off-by: Ben Blum <bblum@andrew.cmu.edu>
Acked-by: Li Zefan <lizf@cn.fujitsu.com>
Cc: Paul Menage <menage@google.com>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Lai Jiangshan <laijs@cn.fujitsu.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Mostly minor changes to add a net argument to various functions and
remove initial network namespace checks.
Make /proc/net/psched per network namespace.
Signed-off-by: Tom Goff <thomas.goff@boeing.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The action modules have been prefixed with 'act_', but the Kconfig
description was not changed.
Signed-off-by: Jan Luebbe <jluebbe@debian.org>
Acked-by: Jamal Hadi Salim <hadi@cyberus.ca>
Signed-off-by: David S. Miller <davem@davemloft.net>
This adds an additional queuing strategy, called pfifo_head_drop,
to remove the oldest skb in the case of an overflow within the queue -
the head element - instead of the last skb (tail). To remove the oldest
skb in congested situations is useful for sensor network environments
where newer packets reflect the superior information.
Reviewed-by: Florian Westphal <fw@strlen.de>
Acked-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: Hagen Paul Pfeifer <hagen@jauu.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
Not including net/atm/
Compiled tested x86 allyesconfig only
Added a > 80 column line or two, which I ignored.
Existing checkpatch plaints willfully, cheerfully ignored.
Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Generated with the following semantic patch
@@
struct net *n1;
struct net *n2;
@@
- n1 == n2
+ net_eq(n1, n2)
@@
struct net *n1;
struct net *n2;
@@
- n1 != n2
+ !net_eq(n1, n2)
applied over {include,net,drivers/net}.
Signed-off-by: Octavian Purdila <opurdila@ixiacom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This fix can probably wait 2.6.33, or should use another patch
if needed in 2.6.32 (no get_dev_by_index_rcu() before 2.6.33)
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
move checking if eaction is valid in tcf_mirred_init()
Signed-off-by: Changli Gao <xiaosuo@gmail.com>
Signed-off-by: Jamal Hadi Salim <hadi@cyberus.ca>
Signed-off-by: David S. Miller <davem@davemloft.net>
1. don't let go back using goto.
2. don't call skb_act_clone() until it is necessary.
3. one exit of the critical context.
Signed-off-by: Changli Gao <xiaosuo@gmail.com>
Signed-off-by: Jamal Hadi Salim <hadi@cyberus.ca>
Signed-off-by: David S. Miller <davem@davemloft.net>
Recent changes in the TX error propagation require additional checking
and masking of values returned from hard_start_xmit(), mainly to
separate cases where skb was consumed. This aim can be simplified by
changing the order of NETDEV_TX and NET_XMIT codes, because the latter
are treated similarly to negative (ERRNO) values.
After this change much simpler dev_xmit_complete() is also used in
sch_direct_xmit(), so it is moved to netdevice.h.
Additionally NET_RX definitions in netdevice.h are moved up from
between TX codes to avoid confusion while reading the TX comment.
Signed-off-by: Jarek Poplawski <jarkao2@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Currently the ->ndo_hard_start_xmit() callbacks are only permitted to return
one of the NETDEV_TX codes. This prevents any kind of error propagation for
virtual devices, like queue congestion of the underlying device in case of
layered devices, or unreachability in case of tunnels.
This patches changes the NET_XMIT codes to avoid clashes with the NETDEV_TX
codes and changes the two callers of dev_hard_start_xmit() to expect either
errno codes, NET_XMIT codes or NETDEV_TX codes as return value.
In case of qdisc_restart(), all non NETDEV_TX codes are mapped to NETDEV_TX_OK
since no error propagation is possible when using qdiscs. In case of
dev_queue_xmit(), the error is propagated upwards.
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
Use RCU to walk list of network devices in qdisc dump.
This could be optimized for large number of devices.
Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Acked-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
something-bility is spelled as something-blity
so a grep for 'blit' would find these lines
this is so trivial that I didn't split it by subsystem / copy
additional maintainers - all changes are to comments
The only purpose is to get fewer false positives when grepping
around the kernel sources.
Signed-off-by: Dirk Hohndel <hohndel@infradead.org>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
We hold RTNL in tc_dump_tfilter(), we can avoid dev_hold()/dev_put()
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Another rcu conversion to avoid one dev_hold()/dev_put() pair
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This probably deserves to go into -stable.
Pedit will reject a policy that is large because it
uses the wrong structure in the policy validation.
This fixes it.
Signed-off-by: Jamal Hadi Salim <hadi@cyberus.ca>
Signed-off-by: David S. Miller <davem@davemloft.net>
Commit 9ef1d4c7c7 ("[NETLINK]: Missing
initializations in dumped data") introduced a typo in
initialization. This patch fixes this.
Signed-off-by: Jiri Pirko <jpirko@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Jarek Poplawski a écrit :
>
>
> Hmm... So you made me to do some "real" work here, and guess what?:
> there is one serious checkpatch warning! ;-) Plus, this new parameter
> should be added to the function description. Otherwise:
> Signed-off-by: Jarek Poplawski <jarkao2@gmail.com>
>
> Thanks,
> Jarek P.
>
> PS: I guess full "Don't" would show we really mean it...
Okay :) Here is the last round, before the night !
Thanks again
[RFC] pkt_sched: gen_estimator: Don't report fake rate estimators
We currently send TCA_STATS_RATE_EST elements to netlink users, even if no estimator
is running.
# tc -s -d qdisc
qdisc pfifo_fast 0: dev eth0 root bands 3 priomap 1 2 2 2 1 2 0 0 1 1 1 1 1 1 1 1
Sent 112833764978 bytes 1495081739 pkt (dropped 0, overlimits 0 requeues 0)
rate 0bit 0pps backlog 0b 0p requeues 0
User has no way to tell if the "rate 0bit 0pps" is a real estimation, or a fake
one (because no estimator is active)
After this patch, tc command output is :
$ tc -s -d qdisc
qdisc pfifo_fast 0: dev eth0 root bands 3 priomap 1 2 2 2 1 2 0 0 1 1 1 1 1 1 1 1
Sent 561075 bytes 1196 pkt (dropped 0, overlimits 0 requeues 0)
backlog 0b 0p requeues 0
We add a parameter to gnet_stats_copy_rate_est() function so that
it can use gen_estimator_active(bstats, r), as suggested by Jarek.
This parameter can be NULL if check is not necessary, (htb for
example has a mandatory rate estimator)
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: Jarek Poplawski <jarkao2@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Some classful qdiscs miss qstats.qlen updating with q.qlen of their
child qdiscs in dump_stats methods.
Signed-off-by: Jarek Poplawski <jarkao2@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
