linux-stable.git/drivers/net, branch v3.10.45

net/mlx4_core: Preserve pci_dev_data after __mlx4_remove_one()

2014-06-26T19:12:39+00:00

[ Upstream commit befdf8978accecac2e0739e6b5075afc62db37fe ]

pci_match_id() just match the static pci_device_id, which may return NULL if
someone binds the driver to a device manually using
/sys/bus/pci/drivers/.../new_id.

This patch wrap up a helper function __mlx4_remove_one() which does the tear
down function but preserve the drv_data. Functions like
mlx4_pci_err_detected() and mlx4_restart_one() will call this one with out
releasing drvdata.

Fixes: 97a5221 "net/mlx4_core: pass pci_device_id.driver_data to __mlx4_init_one during reset".

CC: Bjorn Helgaas 
CC: Amir Vadai 
CC: Jack Morgenstein 
CC: Or Gerlitz 
Signed-off-by: Wei Yang 
Acked-by: Jack Morgenstein 
Signed-off-by: David S. Miller 
Signed-off-by: Greg Kroah-Hartman

net/mlx4_core: pass pci_device_id.driver_data to __mlx4_init_one during reset

2014-06-26T19:12:39+00:00

[ No upstream commit, this is a cherry picked backport enabler. ]

The second parameter of __mlx4_init_one() is used to identify whether the
pci_dev is a PF or VF. Currently, when it is invoked in mlx4_pci_slot_reset()
this information is missed.

This patch match the pci_dev with mlx4_pci_table and passes the
pci_device_id.driver_data to __mlx4_init_one() in mlx4_pci_slot_reset().

Signed-off-by: Wei Yang 
Signed-off-by: David S. Miller 
Signed-off-by: Greg Kroah-Hartman

vxlan: use dev->needed_headroom instead of dev->hard_header_len

2014-06-26T19:12:39+00:00

[ Upstream commit 2853af6a2ea1a8ed09b09dd4fb578e7f435e8d34 ]

When we mirror packets from a vxlan tunnel to other device,
the mirror device should see the same packets (that is, without
outer header). Because vxlan tunnel sets dev->hard_header_len,
tcf_mirred() resets mac header back to outer mac, the mirror device
actually sees packets with outer headers

Vxlan tunnel should set dev->needed_headroom instead of
dev->hard_header_len, like what other ip tunnels do. This fixes
the above problem.

Cc: "David S. Miller" 
Cc: stephen hemminger 
Cc: Pravin B Shelar 
Signed-off-by: Cong Wang 
Signed-off-by: Cong Wang 
Signed-off-by: David S. Miller 
Signed-off-by: Greg Kroah-Hartman

net: force a list_del() in unregister_netdevice_many()

2014-06-26T19:12:38+00:00

[ Upstream commit 87757a917b0b3c0787e0563c679762152be81312 ]

unregister_netdevice_many() API is error prone and we had too
many bugs because of dangling LIST_HEAD on stacks.

See commit f87e6f47933e3e ("net: dont leave active on stack LIST_HEAD")

In fact, instead of making sure no caller leaves an active list_head,
just force a list_del() in the callee. No one seems to need to access
the list after unregister_netdevice_many()

Signed-off-by: Eric Dumazet 
Signed-off-by: David S. Miller 
Signed-off-by: Greg Kroah-Hartman

net: qmi_wwan: add Olivetti Olicard modems

2014-06-26T19:12:38+00:00

[ Upstream commit ba6de0f5304ccdc45ae260e7e0feb6e0ef2dd558 ]

Lars writes: "I'm only 99% sure that the net interfaces are qmi
interfaces, nothing to lose by adding them in my opinion."

And I tend to agree based on the similarity with the two Olicard
modems we already have here.

Reported-by: Lars Melin 
Signed-off-by: Bjørn Mork 
Signed-off-by: David S. Miller 
Signed-off-by: Greg Kroah-Hartman

team: fix mtu setting

2014-06-26T19:12:38+00:00

[ Upstream commit 9d0d68faea6962d62dd501cd6e71ce5cc8ed262b ]

Now it is not possible to set mtu to team device which has a port
enslaved to it. The reason is that when team_change_mtu() calls
dev_set_mtu() for port device, notificator for NETDEV_PRECHANGEMTU
event is called and team_device_event() returns NOTIFY_BAD forbidding
the change. So fix this by returning NOTIFY_DONE here in case team is
changing mtu in team_change_mtu().

Introduced-by: 3d249d4c "net: introduce ethernet teaming device"
Signed-off-by: Jiri Pirko 
Acked-by: Flavio Leitner 
Signed-off-by: David S. Miller 
Signed-off-by: Greg Kroah-Hartman

mlx4_en: don't use napi_synchronize inside mlx4_en_netpoll

2014-06-16T20:42:52+00:00

commit c98235cb8584a72e95786e17d695a8e5fafcd766 upstream.

The mlx4 driver is triggering schedules while atomic inside
mlx4_en_netpoll:

	spin_lock_irqsave(&cq->lock, flags);
	napi_synchronize(&cq->napi);
		^^^^^ msleep here
	mlx4_en_process_rx_cq(dev, cq, 0);
	spin_unlock_irqrestore(&cq->lock, flags);

This was part of a patch by Alexander Guller from Mellanox in 2011,
but it still isn't upstream.

Signed-off-by: Chris Mason 
Acked-By: Amir Vadai 
Signed-off-by: David S. Miller 
Signed-off-by: Jiri Slaby 
Cc: Masoud Sharbiani 
Signed-off-by: Greg Kroah-Hartman

ath9k: protect tid->sched check

2014-06-07T20:25:40+00:00

[ Upstream commit 21f8aaee0c62708654988ce092838aa7df4d25d8 ]

We check tid->sched without a lock taken on ath_tx_aggr_sleep(). That
is race condition which can result of doing list_del(&tid->list) twice
(second time with poisoned list node) and cause crash like shown below:

[424271.637220] BUG: unable to handle kernel paging request at 00100104
[424271.637328] IP: [] ath_tx_aggr_sleep+0x62/0xe0 [ath9k]
...
[424271.639953] Call Trace:
[424271.639998]  [] ? ath9k_get_survey+0x110/0x110 [ath9k]
[424271.640083]  [] ath9k_sta_notify+0x42/0x50 [ath9k]
[424271.640177]  [] sta_ps_start+0x8f/0x1c0 [mac80211]
[424271.640258]  [] ? free_compound_page+0x2e/0x40
[424271.640346]  [] ieee80211_rx_handlers+0x9d5/0x2340 [mac80211]
[424271.640437]  [] ? kmem_cache_free+0x1d8/0x1f0
[424271.640510]  [] ? kfree_skbmem+0x34/0x90
[424271.640578]  [] ? put_page+0x2c/0x40
[424271.640640]  [] ? kfree_skbmem+0x34/0x90
[424271.640706]  [] ? kfree_skbmem+0x34/0x90
[424271.640787]  [] ? ieee80211_rx_handlers_result+0x73/0x1d0 [mac80211]
[424271.640897]  [] ieee80211_prepare_and_rx_handle+0x520/0xad0 [mac80211]
[424271.641009]  [] ? ieee80211_rx_handlers+0x2ed/0x2340 [mac80211]
[424271.641104]  [] ? ip_output+0x7e/0xd0
[424271.641182]  [] ieee80211_rx+0x307/0x7c0 [mac80211]
[424271.641266]  [] ath_rx_tasklet+0x88e/0xf70 [ath9k]
[424271.641358]  [] ? ieee80211_rx+0x1dc/0x7c0 [mac80211]
[424271.641445]  [] ath9k_tasklet+0xcb/0x130 [ath9k]

Bug report:
https://bugzilla.kernel.org/show_bug.cgi?id=70551

Reported-and-tested-by: Max Sydorenko 
Signed-off-by: Stanislaw Gruszka 
Signed-off-by: John W. Linville 
[ xl: backported to 3.10: adjusted context ]
Signed-off-by: Xiangyu Lu 
Signed-off-by: Greg Kroah-Hartman

brcmsmac: fix deadlock on missing firmware

2014-06-07T20:25:34+00:00

commit 8fc1e8c240aab968db658b2d8d079b4391207a36 upstream.

When brcm80211 firmware is not installed networking hangs.
A deadlock happens because we call ieee80211_unregister_hw()
from the .start callback of struct ieee80211_ops. When .start
is called we are under rtnl lock and ieee80211_unregister_hw()
tries to take it again.

Function call stack:

dev_change_flags()
	__dev_change_flags()
		__dev_open()
			ASSERT_RTNL() <-- Assert rtnl lock
			ops->ndo_open()

.ndo_open = ieee80211_open,

ieee80211_open()
	ieee80211_do_open()
		drv_start()
			local->ops->start()

.start = brcms_ops_start,

brcms_ops_start()
	brcms_remove()
		ieee80211_unregister_hw()
			rtnl_lock() <-- Here we deadlock

Introduced by:
commit 25b5632fb35ca61b8ae3eee235edcdc2883f7a5e
("brcmsmac: request firmware in .start() callback")

This patch fixes the bug by removing the call to brcms_remove()
and moves the brcms_request_fw() call to the top of the .start
callback to not initiate anything unless firmware is installed.

Signed-off-by: Emil Goode 
Signed-off-by: John W. Linville 
Signed-off-by: Greg Kroah-Hartman

rtl8192cu: Fix unbalanced irq enable in error path of rtl92cu_hw_init()

2014-06-07T20:25:34+00:00

commit 3234f5b06fc3094176a86772cc64baf3decc98fc upstream.

Fixes: a53268be0cb9 ('rtlwifi: rtl8192cu: Fix too long disable of IRQs')
Signed-off-by: Ben Hutchings 
Signed-off-by: John W. Linville 
Signed-off-by: Greg Kroah-Hartman