| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
commit a9b3cd7f32 (rcu: convert uses of rcu_assign_pointer(x, NULL) to
RCU_INIT_POINTER) did a lot of incorrect changes, since it did a
complete conversion of rcu_assign_pointer(x, y) to RCU_INIT_POINTER(x,
y).
We miss needed barriers, even on x86, when y is not NULL.
Signed-off-by: Eric Dumazet <[email protected]>
CC: Stephen Hemminger <[email protected]>
CC: Paul E. McKenney <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
> net/core/sock.c: In function 'sk_update_clone':
> net/core/sock.c:1278:3: error: implicit declaration of function 'sock_update_memcg'
Reported-by: Randy Dunlap <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
dev_uc_sync() and dev_mc_sync() are acquiring netif_addr_lock for
destination device of synchronization. Since netif_addr_lock is
already held at the time for source device, this triggers lockdep
deadlock warning.
There's no way this deadlock can happen so use spin_lock_nested() to
silence the warning.
Signed-off-by: Jiri Pirko <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
Signed-off-by: Jiri Pirko <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
so move it there. Fixes build errors when CONFIG_INET is not defined:
In file included from include/linux/tcp.h:211:0,
from include/linux/ipv6.h:221,
from include/net/ipv6.h:16,
from include/linux/sunrpc/clnt.h:26,
from include/linux/nfs_fs.h:50,
from init/do_mounts.c:20:
include/net/sock.h: In function 'sk_update_clone':
include/net/sock.h:1109:3: error: implicit declaration of function 'sock_update_memcg' [-Werror=implicit-function-declaration]
Signed-off-by: Stephen Rothwell <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
In 882716604ec "pktgen: fix multiple queue warning" we added special
logic to handle the case where ntxq is zero. It's not clear to me that
ntxq can actually be zero. But if it were then we would set
->queue_map_min and ->queue_map_max to USHRT_MAX when probably we want
to set them to zero?
Signed-off-by: David S. Miller <[email protected]>
|
|
Sockets can also be created through sock_clone. Because it copies
all data in the sock structure, it also copies the memcg-related pointer,
and all should be fine. However, since we now use reference counts in
socket creation, we are left with some sockets that have no reference
counts. It matters when we destroy them, since it leads to a mismatch.
Signed-off-by: Glauber Costa <[email protected]>
CC: David S. Miller <[email protected]>
CC: Greg Thelen <[email protected]>
CC: Hiroyouki Kamezawa <[email protected]>
CC: Laurent Chavey <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
Once upon a time netlink was not sync and we had to get the effective
capabilities from the skb that was being received. Today we instead get
the capabilities from the current task. This has rendered the entire
purpose of the hook moot as it is now functionally equivalent to the
capable() call.
Signed-off-by: Eric Paris <[email protected]>
|
|
All implementations have been converted to implement set_rxnfc
instead.
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
Define special location values for RX NFC that request the driver to
select the actual rule location. This allows for implementation on
devices that use hash-based filter lookup, whereas currently the API is
more suited to devices with TCAM lookup or linear search.
In ethtool_set_rxnfc() and the compat wrapper ethtool_ioctl(), copy
the structure back to user-space after insertion so that the actual
location is returned.
Signed-off-by: Ben Hutchings <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
Add a routine that dumps memory-related values of a socket.
It's made as an array to make it possible to add more stuff
here later without breaking compatibility.
Since v1: The SK_MEMINFO_ constants are in userspace
visible part of sock_diag.h, the rest is under __KERNEL__.
Signed-off-by: Pavel Emelyanov <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
In order to perform a proper universal hash on a vector of integers,
we have to use different universal hashes on each vector element.
Which means we need 4 different hash randoms for ipv6.
Signed-off-by: David S. Miller <[email protected]>
|
|
Aim of this patch is to provide full range of rps_flow_cnt on 64bit arches.
Theorical limit on number of flows is 2^32
Fix some buggy RPS/RFS macros as well.
Signed-off-by: Eric Dumazet <[email protected]>
CC: Tom Herbert <[email protected]>
CC: Xi Wang <[email protected]>
CC: Laurent Chavey <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
Conflicts:
net/bluetooth/l2cap_core.c
Just two overlapping changes, one added an initialization of
a local variable, and another change added a new local variable.
Signed-off-by: David S. Miller <[email protected]>
|
|
skb->truesize might be big even for a small packet.
Its even bigger after commit 87fb4b7b533 (net: more accurate skb
truesize) and big MTU.
We should allow queueing at least one packet per receiver, even with a
low RCVBUF setting.
Reported-by: Michal Simek <[email protected]>
Signed-off-by: Eric Dumazet <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
Setting a large rps_flow_cnt like (1 << 30) on 32-bit platform will
cause a kernel oops due to insufficient bounds checking.
if (count > 1<<30) {
/* Enforce a limit to prevent overflow */
return -EINVAL;
}
count = roundup_pow_of_two(count);
table = vmalloc(RPS_DEV_FLOW_TABLE_SIZE(count));
Note that the macro RPS_DEV_FLOW_TABLE_SIZE(count) is defined as:
... + (count * sizeof(struct rps_dev_flow))
where sizeof(struct rps_dev_flow) is 8. (1 << 30) * 8 will overflow
32 bits.
This patch replaces the magic number (1 << 30) with a symbolic bound.
Suggested-by: Eric Dumazet <[email protected]>
Signed-off-by: Xi Wang <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
flow_cach_flush() might sleep but can be called from
atomic context via the xfrm garbage collector. So add
a flow_cache_flush_deferred() function and use this if
the xfrm garbage colector is invoked from within the
packet path.
Signed-off-by: Steffen Klassert <[email protected]>
Acked-by: Timo Teräs <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
This reverts commit 5c3ddec73d01a1fae9409c197078cb02c42238c3.
S390 qeth driver actually still uses the setup ops.
Reported-by: Frank Blaschka <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
Use IS_ENABLED(CONFIG_FOO)
instead of defined(CONFIG_FOO) || defined (CONFIG_FOO_MODULE)
Signed-off-by: Igor Maravić <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
We can't scan the proto_list to initialize sock cgroups, as it
holds a rwlock, and we also want to keep the code generic enough to
avoid calling the initialization functions of protocols directly,
Convert proto_list_lock into a mutex, so we can sleep and do the
necessary allocations. This lock is seldom taken, so there shouldn't
be any performance penalties associated with that
Signed-off-by: Glauber Costa <[email protected]>
CC: Hiroyouki Kamezawa <[email protected]>
CC: David S. Miller <[email protected]>
CC: Eric Dumazet <[email protected]>
CC: Stephen Rothwell <[email protected]>
CC: Randy Dunlap <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
All drivers that support modification of the RX flow hash indirection
table initialise it in the same way: RX rings are assigned to table
entries in rotation. Make that default policy explicit by having them
call a ethtool_rxfh_indir_default() function.
In the ethtool core, add support for a zero size value for
ETHTOOL_SRXFHINDIR, which resets the table to this default.
Partly-suggested-by: Matt Carlson <[email protected]>
Signed-off-by: Ben Hutchings <[email protected]>
Acked-by: Shreyas N Bhatewara <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
Add a new ethtool operation (get_rxfh_indir_size) to get the
indirectional table size. Use this to validate the user buffer size
before calling get_rxfh_indir or set_rxfh_indir. Use get_rxnfc to get
the number of RX rings, and validate the contents of the new
indirection table before calling set_rxfh_indir. Remove this
validation from drivers.
Signed-off-by: Ben Hutchings <[email protected]>
Acked-by: Dimitris Michailidis <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
The sk address is used as a cookie between dump/get_exact calls.
It will be required for unix socket sdumping, so move it from
inet_diag to sock_diag.
Signed-off-by: Pavel Emelyanov <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
I've made a mistake when fixing the sock_/inet_diag aliases :(
1. The sock_diag layer should request the family-based alias,
not just the IPPROTO_IP one;
2. The inet_diag layer should request for AF_INET+protocol alias,
not just the protocol one.
Thus fix this.
Signed-off-by: Pavel Emelyanov <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
Before adding a struct rtnl_link_ops into link_ops list, check it doesnt
clash with a prior one.
Based on a previous patch from Alexander Smirnov
Signed-off-by: Eric Dumazet <[email protected]>
CC: Alexander Smirnov <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
It's simpler to just keep these things out until there is a real user
of them, so we can see what the needs actually are, rather than keep
these things around as useless overhead.
Signed-off-by: David S. Miller <[email protected]>
|
|
This patch introduces memory pressure controls for the tcp
protocol. It uses the generic socket memory pressure code
introduced in earlier patches, and fills in the
necessary data in cg_proto struct.
Signed-off-by: Glauber Costa <[email protected]>
Reviewed-by: KAMEZAWA Hiroyuki <[email protected]>
CC: Eric W. Biederman <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
The goal of this work is to move the memory pressure tcp
controls to a cgroup, instead of just relying on global
conditions.
To avoid excessive overhead in the network fast paths,
the code that accounts allocated memory to a cgroup is
hidden inside a static_branch(). This branch is patched out
until the first non-root cgroup is created. So when nobody
is using cgroups, even if it is mounted, no significant performance
penalty should be seen.
This patch handles the generic part of the code, and has nothing
tcp-specific.
Signed-off-by: Glauber Costa <[email protected]>
Reviewed-by: KAMEZAWA Hiroyuki <[email protected]>
CC: Kirill A. Shutemov <[email protected]>
CC: David S. Miller <[email protected]>
CC: Eric W. Biederman <[email protected]>
CC: Eric Dumazet <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
This patch replaces all uses of struct sock fields' memory_pressure,
memory_allocated, sockets_allocated, and sysctl_mem to acessor
macros. Those macros can either receive a socket argument, or a mem_cgroup
argument, depending on the context they live in.
Since we're only doing a macro wrapping here, no performance impact at all is
expected in the case where we don't have cgroups disabled.
Signed-off-by: Glauber Costa <[email protected]>
Reviewed-by: Hiroyouki Kamezawa <[email protected]>
CC: David S. Miller <[email protected]>
CC: Eric W. Biederman <[email protected]>
CC: Eric Dumazet <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
Instead of testing defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE)
Signed-off-by: Eric Dumazet <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
This reverts commit 865d9f9f748fdc1943679ea65d9ee1dc55e4a6ae.
This commit breaks the build with CONFIG_NETPRIO_CGROUP=y so
revert it. It does build as a module though. The SUBSYS macro
in the cgroup core code automatically defines a subsys structure
as extern. Long term we should fix the macro. And I need to
fully build test things.
Tested with CONFIG_NETPRIO_CGROUP={y|m|n} with and without
CONFIG_CGROUPS defined.
Signed-off-by: John Fastabend <[email protected]>
CC: Neil Horman <[email protected]>
Reported-By: Eric Dumazet <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
These tests are off by one because sock_diag_handlers[] only has AF_MAX
elements.
Signed-off-by: Dan Carpenter <[email protected]>
Acked-by: Pavel Emelyanov <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
net_prio_subsys can be made static this removes the sparse
warning it was throwing.
Signed-off-by: John Fastabend <[email protected]>
Acked-by: Neil Horman <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
|
|
On a CONFIG_NET=y build
net/core/secure_seq.c:22: warning: 'seq_scale' defined but not
used
Signed-off-by: Stephen Boyd <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
This patch moves the sock_ code from inet_diag.c to generic sock_diag.c
file and provides necessary request_module-s calls and a pointer on
inet_diag_compat dumping routine.
Signed-off-by: Pavel Emelyanov <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
Since commit c5ed63d66f24(tcp: fix three tcp sysctls tuning),
sysctl_max_syn_backlog is determined by tcp_hashinfo->ehash_mask,
and the minimal value is 128, and it will increase in proportion to the
memory of machine.
The original description for tcp_max_syn_backlog and sysctl_max_syn_backlog
are out of date.
Changelog:
V2: update description for sysctl_max_syn_backlog
Signed-off-by: Weiping Pan <[email protected]>
Reviewed-by: Shan Wei <[email protected]>
Acked-by: Neil Horman <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
netdev_queue_release() should be called even if CONFIG_XPS=n
to properly release device reference.
Signed-off-by: Eric Dumazet <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
To reflect the fact that a refrence is not obtained to the
resulting neighbour entry.
Signed-off-by: David S. Miller <[email protected]>
Acked-by: Roland Dreier <[email protected]>
|
|
We discovered that TCP stack could retransmit misaligned skbs if a
malicious peer acknowledged sub MSS frame. This currently can happen
only if output interface is non SG enabled : If SG is enabled, tcp
builds headless skbs (all payload is included in fragments), so the tcp
trimming process only removes parts of skb fragments, header stay
aligned.
Some arches cant handle misalignments, so force a head reallocation and
shrink headroom to MAX_TCP_HEADER.
Dont care about misaligments on x86 and PPC (or other arches setting
NET_IP_ALIGN to 0)
This patch introduces __pskb_copy() which can specify the headroom of
new head, and pskb_copy() becomes a wrapper on top of __pskb_copy()
Signed-off-by: Eric Dumazet <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
|
|
commit b00055aacdb ([NET] core: add RFC2863 operstate) changed
net_device flags from unsigned short to unsigned int.
Some core functions still assume its an unsigned short.
Signed-off-by: Eric Dumazet <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
Within nested statements, the break statement terminates only the
do, for, switch, or while statement that immediately encloses it,
So replace the break with goto.
Signed-off-by: RongQing.Li <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
If the neigh entry has device private state, it will need
constructor/destructor ops.
Signed-off-by: David S. Miller <[email protected]>
|
|
netdev->neigh_priv_len records the private area length.
This will trigger for neigh_table objects which set tbl->entry_size
to zero, and the first instances of this will be forthcoming.
Signed-off-by: David S. Miller <[email protected]>
|
|
We are going to alloc for device specific private areas for
neighbour entries, and in order to do that we have to move
away from the fixed allocation size enforced by using
neigh_table->kmem_cachep
As a nice side effect we can now use kfree_rcu().
Signed-off-by: David S. Miller <[email protected]>
|
|
Change function rcu_dereference to rcu_dereference_bh to avoid warning
[ INFO: suspicious RCU usage. ]
-------------------------------
net/core/dev.c:2459 suspicious rcu_dereference_check() usage!
because we are locking with
rcu_read_lock_bh();
in function dev_queue_xmit(struct sk_buff *skb)
Signed-off-by: Igor Maravic <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
Le lundi 28 novembre 2011 à 19:06 -0500, David Miller a écrit :
> From: Dimitris Michailidis <[email protected]>
> Date: Mon, 28 Nov 2011 08:25:39 -0800
>
> >> +bool skb_flow_dissect(const struct sk_buff *skb, struct flow_keys
> >> *flow)
> >> +{
> >> + int poff, nhoff = skb_network_offset(skb);
> >> + u8 ip_proto;
> >> + u16 proto = skb->protocol;
> >
> > __be16 instead of u16 for proto?
>
> I'll take care of this when I apply these patches.
( CC trimmed )
Thanks David !
Here is a small patch to use one 64bit load/store on x86_64 instead of
two 32bit load/stores.
[PATCH net-next] flow_dissector: use a 64bit load/store
gcc compiler is smart enough to use a single load/store if we
memcpy(dptr, sptr, 8) on x86_64, regardless of
CONFIG_CC_OPTIMIZE_FOR_SIZE
In IP header, daddr immediately follows saddr, this wont change in the
future. We only need to make sure our flow_keys (src,dst) fields wont
break the rule.
Signed-off-by: Eric Dumazet <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
Networking stack support for byte queue limits, uses dynamic queue
limits library. Byte queue limits are maintained per transmit queue,
and a dql structure has been added to netdev_queue structure for this
purpose.
Configuration of bql is in the tx-<n> sysfs directory for the queue
under the byte_queue_limits directory. Configuration includes:
limit_min, bql minimum limit
limit_max, bql maximum limit
hold_time, bql slack hold time
Also under the directory are:
limit, current byte limit
inflight, current number of bytes on the queue
Signed-off-by: Tom Herbert <[email protected]>
Acked-by: Eric Dumazet <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
This patch moves the xps specific parts in netdev_queue_release into
its own function which netdev_queue_release can call. This allows
netdev_queue_release to be more generic (for adding new attributes
to tx queues).
Signed-off-by: Tom Herbert <[email protected]>
Acked-by: Eric Dumazet <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|