linux-stable.git/net/ipv4/ipmr.c, branch linux-4.5.y

Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net

2015-12-04T02:09:12+00:00

Conflicts:
	drivers/net/ethernet/renesas/ravb_main.c
	kernel/bpf/syscall.c
	net/ipv4/ipmr.c

All three conflicts were cases of overlapping changes.

Signed-off-by: David S. Miller

net: ipmr: add mfc newroute/delroute netlink support

2015-11-30T20:26:23+00:00

This patch adds support to add and remove MFC entries. It uses the
same attributes like the already present dump support in order to be
consistent. There's one new entry - RTA_PREFSRC, it's used to denote an
MFC_PROXY entry (see MRT_ADD_MFC vs MRT_ADD_MFC_PROXY).
The already existing infrastructure is used to create and delete the
entries, the netlink message gets converted internally to a struct mfcctl
which is used with ipmr_mfc_add/delete.
The other used attributes are:
RTA_IIF - used for mfcc_parent (when adding it's required to be valid)
RTA_SRC - used for mfcc_origin
RTA_DST - used for mfcc_mcastgrp
RTA_TABLE - the MRT table id
RTA_MULTIPATH - the "oifs" ttl array

Signed-off-by: Nikolay Aleksandrov 
Signed-off-by: David S. Miller

net: ipmr: fix setsockopt error return

2015-11-30T20:26:22+00:00

We can have both errors and we'll return the second one, fix it to
return an error at a time as it's normal. I've overlooked this in my
previous set.

Signed-off-by: Nikolay Aleksandrov 
Signed-off-by: David S. Miller

net: ipmr: move pimsm_enabled to pim.h and rename

2015-11-30T20:26:22+00:00

Move the inline pimsm_enabled() to pim.h and rename it to
ipmr_pimsm_enabled to show it's for the ipv4 ipmr code since pim.h is
used by IPv6 too.

Signed-off-by: Nikolay Aleksandrov 
Signed-off-by: David S. Miller

net: ipmr: move struct mr_table and VIF_EXISTS to mroute.h

2015-11-30T20:26:22+00:00

Move the definitions of VIF_EXISTS() and struct mr_table to mroute.h

Signed-off-by: Nikolay Aleksandrov 
Signed-off-by: David S. Miller

net: ipmr: remove unused MFC_NOTIFY flag and make the flags enum

2015-11-30T20:26:22+00:00

MFC_NOTIFY was introduced in kernel 2.1.68 but afaik it hasn't been used
and I couldn't find any users currently so just remove it. Only
MFC_STATIC is left, so move it into an enum, add a description and use
BIT().

Signed-off-by: Nikolay Aleksandrov 
Signed-off-by: David S. Miller

net: ipmr, ip6mr: fix vif/tunnel failure race condition

2015-11-24T22:15:56+00:00

Since (at least) commit b17a7c179dd3 ("[NET]: Do sysfs registration as
part of register_netdevice."), netdev_run_todo() deals only with
unregistration, so we don't need to do the rtnl_unlock/lock cycle to
finish registration when failing pimreg or dvmrp device creation. In
fact that opens a race condition where someone can delete the device
while rtnl is unlocked because it's fully registered. The problem gets
worse when netlink support is introduced as there are more points of entry
that can cause it and it also makes reusing that code correctly impossible.

Signed-off-by: Nikolay Aleksandrov 
Reviewed-by: Cong Wang 
Signed-off-by: David S. Miller

net: ipmr: factor out common vif init code

2015-11-23T20:06:39+00:00

Factor out common vif init code used in both tunnel and pimreg
initialization and create ipmr_init_vif_indev() function.

Signed-off-by: Nikolay Aleksandrov 
Signed-off-by: David S. Miller

net: ipmr: rearrange and cleanup setsockopt

2015-11-23T20:06:39+00:00

Take rtnl in the beginning unconditionally as most options already need
it (one exception - MRT_DONE, see the comment inside), make the
lock/unlock places central and move out the switch() local variables.

Signed-off-by: Nikolay Aleksandrov 
Signed-off-by: David S. Miller

net: ipmr: drop ip_mr_init() mrt_cachep null check as we'll panic if it fails

2015-11-23T20:06:39+00:00

It's not necessary to check for null as SLAB_PANIC is used and we'll
panic if the alloc fails, so just drop it.

Signed-off-by: Nikolay Aleksandrov 
Signed-off-by: David S. Miller