pandora-kernel.git
9 years agoALSA: hda - Always turn on pins for HDMI/DP
Takashi Iwai [Fri, 14 Dec 2012 09:22:35 +0000 (10:22 +0100)]
ALSA: hda - Always turn on pins for HDMI/DP

commit 6169b673618bf0b2518ce413b54925782a603f06 upstream.

We've seen the broken HDMI *video* output on some machines with GM965,
and the debugging session pointed that the culprit is the disabled
audio output pins.  Toggling these pins dynamically on demand caused
flickering of HDMI TV.

This patch changes the behavior to keep the pin ON constantly.

Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=51421

Signed-off-by: Takashi Iwai <tiwai@suse.de>
[bwh: Backported to 3.2: adjust context]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoMIPS: Fix poweroff failure when HOTPLUG_CPU configured.
Huacai Chen [Mon, 13 Aug 2012 12:52:24 +0000 (20:52 +0800)]
MIPS: Fix poweroff failure when HOTPLUG_CPU configured.

commit 8add1ecb81f541ef2fcb0b85a5470ad9ecfb4a84 upstream.

When poweroff machine, kernel_power_off() call disable_nonboot_cpus().
And if we have HOTPLUG_CPU configured, disable_nonboot_cpus() is not an
empty function but attempt to actually disable the nonboot cpus. Since
system state is SYSTEM_POWER_OFF, play_dead() won't be called and thus
disable_nonboot_cpus() hangs. Therefore, we make this patch to avoid
poweroff failure.

Signed-off-by: Huacai Chen <chenhc@lemote.com>
Signed-off-by: Hongliang Tao <taohl@lemote.com>
Signed-off-by: Hua Yan <yanh@lemote.com>
Cc: Yong Zhang <yong.zhang@windriver.com>
Cc: linux-mips@linux-mips.org
Cc: linux-kernel@vger.kernel.org
Cc: Fuxin Zhang <zhangfx@lemote.com>
Cc: Zhangjin Wu <wuzhangjin@gmail.com>
Patchwork: https://patchwork.linux-mips.org/patch/4211/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoALSA: hda - Fix pin configuration of HP Pavilion dv7
Takashi Iwai [Thu, 13 Dec 2012 13:33:42 +0000 (14:33 +0100)]
ALSA: hda - Fix pin configuration of HP Pavilion dv7

commit 8ae5865ec77c22462c736846a0679947a6953548 upstream.

Fix the quirk entry for HP Pavilion dv7 in order to make the bass
speaker working.

Reported-and-tested-by: Tomas Pospisek <tpo2@sourcepole.ch>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agomtd: nand: gpmi: reset BCH earlier, too, to avoid NAND startup problems
Wolfram Sang [Wed, 5 Dec 2012 20:46:02 +0000 (21:46 +0100)]
mtd: nand: gpmi: reset BCH earlier, too, to avoid NAND startup problems

commit 6f2a6a52560ad8d85710aabd92b7a3239b3a6b07 upstream.

It could happen (1 out of 100 times) that NAND did not start up
correctly after warm rebooting, so the kernel could not find the UBI or
DMA timed out due to a stalled BCH. When resetting BCH together with
GPMI, the issue could not be observed anymore (after 10000+ reboots). We
probably need the consistent state already before sending any command to
NAND, even when no ECC is needed. I chose to keep the extra reset for
BCH when changing the flash layout to be on the safe side.

Signed-off-by: Wolfram Sang <w.sang@pengutronix.de>
Acked-by: Huang Shijie <b32955@freescale.com>
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agotarget/tcm_fc: fix the lockdep warning due to inconsistent lock state
Yi Zou [Tue, 11 Dec 2012 01:04:00 +0000 (17:04 -0800)]
target/tcm_fc: fix the lockdep warning due to inconsistent lock state

commit 9f4ad44b264f8bb61ffdd607148215566568430d upstream.

The lockdep warning below is in theory correct but it will be in really weird
rare situation that ends up that deadlock since the tcm fc session is hashed
based the rport id. Nonetheless, the complaining below is about rcu callback
that does the transport_deregister_session() is happening in softirq, where
transport_register_session() that happens earlier is not. This triggers the
lockdep warning below. So, just fix this to make lockdep happy by disabling
the soft irq before calling transport_register_session() in ft_prli.

BTW, this was found in FCoE VN2VN over two VMs, couple of create and destroy
would get this triggered.

v1: was enforcing register to be in softirq context which was not righ. See,
http://www.spinics.net/lists/target-devel/msg03614.html

v2: following comments from Roland&Nick (thanks), it seems we don't have to
do transport_deregister_session() in rcu callback, so move it into ft_sess_free()
but still do kfree() of the corresponding ft_sess struct in rcu callback to
make sure the ft_sess is not freed till the rcu callback.

...
[ 1328.370592] scsi2 : FCoE Driver
[ 1328.383429] fcoe: No FDMI support.
[ 1328.384509] host2: libfc: Link up on port (000000)
[ 1328.934229] host2: Assigned Port ID 00a292
[ 1357.232132] host2: rport 00a393: Remove port
[ 1357.232568] host2: rport 00a393: Port sending LOGO from Ready state
[ 1357.233692] host2: rport 00a393: Delete port
[ 1357.234472] host2: rport 00a393: work event 3
[ 1357.234969] host2: rport 00a393: callback ev 3
[ 1357.235979] host2: rport 00a393: Received a LOGO response closed
[ 1357.236706] host2: rport 00a393: work delete
[ 1357.237481]
[ 1357.237631] =================================
[ 1357.238064] [ INFO: inconsistent lock state ]
[ 1357.238450] 3.7.0-rc7-yikvm+ #3 Tainted: G           O
[ 1357.238450] ---------------------------------
[ 1357.238450] inconsistent {SOFTIRQ-ON-W} -> {IN-SOFTIRQ-W} usage.
[ 1357.238450] ksoftirqd/0/3 [HC0[0]:SC1[1]:HE0:SE0] takes:
[ 1357.238450]  (&(&se_tpg->session_lock)->rlock){+.?...}, at: [<ffffffffa01eacd4>] transport_deregister_session+0x41/0x148 [target_core_mod]
[ 1357.238450] {SOFTIRQ-ON-W} state was registered at:
[ 1357.238450]   [<ffffffff810834f5>] mark_held_locks+0x6d/0x95
[ 1357.238450]   [<ffffffff8108364a>] trace_hardirqs_on_caller+0x12d/0x197
[ 1357.238450]   [<ffffffff810836c1>] trace_hardirqs_on+0xd/0xf
[ 1357.238450]   [<ffffffff8149caba>] _raw_spin_unlock_irq+0x2d/0x45
[ 1357.238450]   [<ffffffffa01e8d10>] __transport_register_session+0xb8/0x122 [target_core_mod]
[ 1357.238450]   [<ffffffffa01e8dbe>] transport_register_session+0x44/0x5a [target_core_mod]
[ 1357.238450]   [<ffffffffa018e32c>] ft_prli+0x1e3/0x275 [tcm_fc]
[ 1357.238450]   [<ffffffffa0160e8d>] fc_rport_recv_req+0x95e/0xdc5 [libfc]
[ 1357.238450]   [<ffffffffa015be88>] fc_lport_recv_els_req+0xc4/0xd5 [libfc]
[ 1357.238450]   [<ffffffffa015c778>] fc_lport_recv_req+0x12f/0x18f [libfc]
[ 1357.238450]   [<ffffffffa015a6d7>] fc_exch_recv+0x8ba/0x981 [libfc]
[ 1357.238450]   [<ffffffffa0176d7a>] fcoe_percpu_receive_thread+0x47a/0x4e2 [fcoe]
[ 1357.238450]   [<ffffffff810549f1>] kthread+0xb1/0xb9
[ 1357.238450]   [<ffffffff814a40ec>] ret_from_fork+0x7c/0xb0
[ 1357.238450] irq event stamp: 275411
[ 1357.238450] hardirqs last  enabled at (275410): [<ffffffff810bb6a0>] rcu_process_callbacks+0x229/0x42a
[ 1357.238450] hardirqs last disabled at (275411): [<ffffffff8149c2f7>] _raw_spin_lock_irqsave+0x22/0x8e
[ 1357.238450] softirqs last  enabled at (275394): [<ffffffff8103d669>] __do_softirq+0x246/0x26f
[ 1357.238450] softirqs last disabled at (275399): [<ffffffff8103d6bb>] run_ksoftirqd+0x29/0x62
[ 1357.238450]
[ 1357.238450] other info that might help us debug this:
[ 1357.238450]  Possible unsafe locking scenario:
[ 1357.238450]
[ 1357.238450]        CPU0
[ 1357.238450]        ----
[ 1357.238450]   lock(&(&se_tpg->session_lock)->rlock);
[ 1357.238450]   <Interrupt>
[ 1357.238450]     lock(&(&se_tpg->session_lock)->rlock);
[ 1357.238450]
[ 1357.238450]  *** DEADLOCK ***
[ 1357.238450]
[ 1357.238450] no locks held by ksoftirqd/0/3.
[ 1357.238450]
[ 1357.238450] stack backtrace:
[ 1357.238450] Pid: 3, comm: ksoftirqd/0 Tainted: G           O 3.7.0-rc7-yikvm+ #3
[ 1357.238450] Call Trace:
[ 1357.238450]  [<ffffffff8149399a>] print_usage_bug+0x1f5/0x206
[ 1357.238450]  [<ffffffff8100da59>] ? save_stack_trace+0x2c/0x49
[ 1357.238450]  [<ffffffff81082aae>] ? print_irq_inversion_bug.part.14+0x1ae/0x1ae
[ 1357.238450]  [<ffffffff81083336>] mark_lock+0x106/0x258
[ 1357.238450]  [<ffffffff81084e34>] __lock_acquire+0x2e7/0xe53
[ 1357.238450]  [<ffffffff8102903d>] ? pvclock_clocksource_read+0x48/0xb4
[ 1357.238450]  [<ffffffff810ba6a3>] ? rcu_process_gp_end+0xc0/0xc9
[ 1357.238450]  [<ffffffffa01eacd4>] ? transport_deregister_session+0x41/0x148 [target_core_mod]
[ 1357.238450]  [<ffffffff81085ef1>] lock_acquire+0x119/0x143
[ 1357.238450]  [<ffffffffa01eacd4>] ? transport_deregister_session+0x41/0x148 [target_core_mod]
[ 1357.238450]  [<ffffffff8149c329>] _raw_spin_lock_irqsave+0x54/0x8e
[ 1357.238450]  [<ffffffffa01eacd4>] ? transport_deregister_session+0x41/0x148 [target_core_mod]
[ 1357.238450]  [<ffffffffa01eacd4>] transport_deregister_session+0x41/0x148 [target_core_mod]
[ 1357.238450]  [<ffffffff810bb6a0>] ? rcu_process_callbacks+0x229/0x42a
[ 1357.238450]  [<ffffffffa018ddc5>] ft_sess_rcu_free+0x17/0x24 [tcm_fc]
[ 1357.238450]  [<ffffffffa018ddae>] ? ft_sess_free+0x1b/0x1b [tcm_fc]
[ 1357.238450]  [<ffffffff810bb6d7>] rcu_process_callbacks+0x260/0x42a
[ 1357.238450]  [<ffffffff8103d55d>] __do_softirq+0x13a/0x26f
[ 1357.238450]  [<ffffffff8149b34e>] ? __schedule+0x65f/0x68e
[ 1357.238450]  [<ffffffff8103d6bb>] run_ksoftirqd+0x29/0x62
[ 1357.238450]  [<ffffffff8105c83c>] smpboot_thread_fn+0x1a5/0x1aa
[ 1357.238450]  [<ffffffff8105c697>] ? smpboot_unregister_percpu_thread+0x47/0x47
[ 1357.238450]  [<ffffffff810549f1>] kthread+0xb1/0xb9
[ 1357.238450]  [<ffffffff8149b49d>] ? wait_for_common+0xbb/0x10a
[ 1357.238450]  [<ffffffff81054940>] ? __init_kthread_worker+0x59/0x59
[ 1357.238450]  [<ffffffff814a40ec>] ret_from_fork+0x7c/0xb0
[ 1357.238450]  [<ffffffff81054940>] ? __init_kthread_worker+0x59/0x59
[ 1417.440099]  rport-2:0-0: blocked FC remote port time out: removing rport

Signed-off-by: Yi Zou <yi.zou@intel.com>
Cc: Open-FCoE <devel@open-fcoe.org>
Cc: Nicholas A. Bellinger <nab@risingtidesystems.com>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoNFS: avoid NULL dereference in nfs_destroy_server
NeilBrown [Thu, 13 Dec 2012 04:14:36 +0000 (15:14 +1100)]
NFS: avoid NULL dereference in nfs_destroy_server

commit f259613a1e4b44a0cf85a5dafd931be96ee7c9e5 upstream.

In rare circumstances, nfs_clone_server() of a v2 or v3 server can get
an error between setting server->destory (to nfs_destroy_server), and
calling nfs_start_lockd (which will set server->nlm_host).

If this happens, nfs_clone_server will call nfs_free_server which
will call nfs_destroy_server and thence nlmclnt_done(NULL).  This
causes the NULL to be dereferenced.

So add a guard to only call nlmclnt_done() if ->nlm_host is not NULL.

The other guards there are irrelevant as nlm_host can only be non-NULL
if one of these flags are set - so remove those tests.  (Thanks to Trond
for this suggestion).

This is suitable for any stable kernel since 2.6.25.

Signed-off-by: NeilBrown <neilb@suse.de>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agodrm/radeon: fix amd afusion gpu setup aka sumo v2
Jerome Glisse [Tue, 11 Dec 2012 16:56:52 +0000 (11:56 -0500)]
drm/radeon: fix amd afusion gpu setup aka sumo v2

commit bd25f0783dc3fb72e1e2779c2b99b2d34b67fa8a upstream.

Set the proper number of tile pipe that should be a multiple of
pipe depending on the number of se engine.

Fix:
https://bugs.freedesktop.org/show_bug.cgi?id=56405
https://bugs.freedesktop.org/show_bug.cgi?id=56720

v2: Don't change sumo2

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
[bwh: Backported to 3.2: don't define/use *_GB_ADDR_CONFIG_GOLDEN]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoALSA: hda - Add stereo-dmic fixup for Acer Aspire One 522
Takashi Iwai [Wed, 12 Dec 2012 11:10:01 +0000 (12:10 +0100)]
ALSA: hda - Add stereo-dmic fixup for Acer Aspire One 522

commit 63a077e27648b4043b1ca1b4e29f0c42d99616b6 upstream.

Acer Aspire One 522 has the infamous digital mic unit that needs the
phase inversion fixup for stereo.

Bugzilla: https://bugzilla.novell.com/show_bug.cgi?id=715737

Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agosolos-pci: fix double-free of TX skb in DMA mode
David Woodhouse [Tue, 11 Dec 2012 14:57:14 +0000 (14:57 +0000)]
solos-pci: fix double-free of TX skb in DMA mode

commit cae49ede00ec3d0cda290b03fee55b72b49efc11 upstream.

We weren't clearing card->tx_skb[port] when processing the TX done interrupt.
If there wasn't another skb ready to transmit immediately, this led to a
double-free because we'd free it *again* next time we did have a packet to
send.

Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agox86, 8042: Enable A20 using KBC to fix S3 resume on some MSI laptops
Ondrej Zary [Tue, 11 Dec 2012 21:18:05 +0000 (22:18 +0100)]
x86, 8042: Enable A20 using KBC to fix S3 resume on some MSI laptops

commit ad68652412276f68ad4fe3e1ecf5ee6880876783 upstream.

Some MSI laptop BIOSes are broken - INT 15h code uses port 92h to enable A20
line but resume code assumes that KBC was used.
The laptop will not resume from S3 otherwise but powers off after a while
and then powers on again stuck with a blank screen.

Fix it by enabling A20 using KBC in i8042_platform_init for x86.

Fixes https://bugzilla.kernel.org/show_bug.cgi?id=12878

Signed-off-by: Ondrej Zary <linux@rainbow-software.org>
Cc: Dmitry Torokhov <dmitry.torokhov@gmail.com>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Cc: Rafael J. Wysocki <rjw@sisk.pl>
Link: http://lkml.kernel.org/r/201212112218.06551.linux@rainbow-software.org
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agomm: dmapool: use provided gfp flags for all dma_alloc_coherent() calls
Marek Szyprowski [Wed, 7 Nov 2012 14:37:07 +0000 (15:37 +0100)]
mm: dmapool: use provided gfp flags for all dma_alloc_coherent() calls

commit 387870f2d6d679746020fa8e25ef786ff338dc98 upstream.

dmapool always calls dma_alloc_coherent() with GFP_ATOMIC flag,
regardless the flags provided by the caller. This causes excessive
pruning of emergency memory pools without any good reason. Additionaly,
on ARM architecture any driver which is using dmapools will sooner or
later  trigger the following error:
"ERROR: 256 KiB atomic DMA coherent pool is too small!
Please increase it with coherent_pool= kernel parameter!".
Increasing the coherent pool size usually doesn't help much and only
delays such error, because all GFP_ATOMIC DMA allocations are always
served from the special, very limited memory pool.

This patch changes the dmapool code to correctly use gfp flags provided
by the dmapool caller.

Reported-by: Soeren Moch <smoch@web.de>
Reported-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Tested-by: Andrew Lunn <andrew@lunn.ch>
Tested-by: Soeren Moch <smoch@web.de>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoiwlwifi: don't handle masked interrupt
Emmanuel Grumbach [Wed, 28 Nov 2012 08:51:34 +0000 (10:51 +0200)]
iwlwifi: don't handle masked interrupt

commit 25a172655f837bdb032e451f95441bb4acec51bb upstream.

This can lead to a panic if the driver isn't ready to
handle them. Since our interrupt line is shared, we can get
an interrupt at any time (and CONFIG_DEBUG_SHIRQ checks
that even when the interrupt is being freed).

If the op_mode has gone away, we musn't call it. To avoid
this the transport disables the interrupts when the hw is
stopped and the op_mode is leaving.
If there is an event that would cause an interrupt the INTA
register is updated regardless of the enablement of the
interrupts: even if the interrupts are disabled, the INTA
will be changed, but the device won't issue an interrupt.
But the ISR can be called at any time, so we ought ignore
the value in the INTA otherwise we can call the op_mode
after it was freed.

I found this bug when the op_mode_start failed, and called
iwl_trans_stop_hw(trans, true). Then I played with the
RFKILL button, and removed the module.
While removing the module, the IRQ is freed, and the ISR is
called (CONFIG_DEBUG_SHIRQ enabled). Panic.

Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Reviewed-by: Gregory Greenman <gregory.greenman@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
[bwh: Backported to 3.2:
 - Adjust context
 - Pass bus(trans), not trans, to iwl_{read,write}32()]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agonfsd: avoid permission checks on EXCLUSIVE_CREATE replay
Neil Brown [Fri, 7 Dec 2012 20:40:55 +0000 (15:40 -0500)]
nfsd: avoid permission checks on EXCLUSIVE_CREATE replay

commit 7007c90fb9fef593b4aeaeee57e6a6754276c97c upstream.

With NFSv4, if we create a file then open it we explicit avoid checking
the permissions on the file during the open because the fact that we
created it ensures we should be allow to open it (the create and the
open should appear to be a single operation).

However if the reply to an EXCLUSIVE create gets lots and the client
resends the create, the current code will perform the permission check -
because it doesn't realise that it did the open already..

This patch should fix this.

Note that I haven't actually seen this cause a problem.  I was just
looking at the code trying to figure out a different EXCLUSIVE open
related issue, and this looked wrong.

(Fix confirmed with pynfs 4.0 test OPEN4--bfields)

Signed-off-by: NeilBrown <neilb@suse.de>
[bfields: use OWNER_OVERRIDE and update for 4.1]
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
[bwh: Backported to 3.2:
 - Adjust context
 - Use current_fh as file handle in do_open_lookup()]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agodrm/i915: disable cpt phase pointer fdi rx workaround
Daniel Vetter [Sat, 8 Dec 2012 11:58:33 +0000 (12:58 +0100)]
drm/i915: disable cpt phase pointer fdi rx workaround

commit 539526b4137bc0e7a8806c38c8522f226814a0e6 upstream.

We've originally added this in

commit 291427f5fdadec6e4be2924172e83588880e1539
Author: Jesse Barnes <jbarnes@virtuousgeek.org>
Date:   Fri Jul 29 12:42:37 2011 -0700

    drm/i915: apply phase pointer override on SNB+ too

and then copy-pasted it over to ivb/ppt. The w/a was originally added
for ilk/ibx in

commit 5b2adf897146edeac6a1e438fb67b5a53dbbdf34
Author: Jesse Barnes <jbarnes@virtuousgeek.org>
Date:   Thu Oct 7 16:01:15 2010 -0700

    drm/i915: add Ironlake clock gating workaround for FDI link training

and fixed up a bit in

commit 6f06ce184c765fd8d50669a8d12fdd566c920859
Author: Jesse Barnes <jbarnes@virtuousgeek.org>
Date:   Tue Jan 4 15:09:38 2011 -0800

    drm/i915: set phase sync pointer override enable before setting phase sync pointer

It turns out that this w/a isn't actually required on cpt/ppt and
positively harmful on ivb/ppt when using fdi B/C links - it results in
a black screen occasionally, with seemingfully everything working as
it should. The only failure indication I've found in the hw is that
eventually (but not right after the modeset completes) a pipe underrun
is signalled.

Big thanks to Arthur Runyan for all the ideas for registers to check
and changes to test, otherwise I couldn't ever have tracked this down!

Cc: "Runyan, Arthur J" <arthur.j.runyan@intel.com>
Reviewed-by: Jesse Barnes <jbarnes@virtuousgeek.org>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
[bwh: Backported to 3.2: adjust context]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agodrm/i915: drop unnecessary check from fdi_link_train code
Daniel Vetter [Wed, 31 Oct 2012 21:52:28 +0000 (22:52 +0100)]
drm/i915: drop unnecessary check from fdi_link_train code

commit 8f5718a6d9a29bf1f51910c6b32bd0244f905303 upstream.

They are all written for a specific north disaplay->pch combination.
So stop pretending otherwise.

Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoath9k_hw: Fix signal strength / channel noise reporting
Felix Fietkau [Mon, 10 Dec 2012 13:03:17 +0000 (14:03 +0100)]
ath9k_hw: Fix signal strength / channel noise reporting

commit b7c0c238898d200e80487516e2b67aba2a522cc0 upstream.

While AR_PHY_CCA_NOM_VAL_* does contain the expected internal noise floor
for a chip measured in clean air, it refers to the lowest expected reading.

Depending on the frequency, this measurement can vary by about 6db, thus
causing a higher reported channel noise and signal strength.

Factor in the 6db offset when converting internal noisefloor to channel noise.

This patch makes the reported values more accurate for all chips without
affecting NF calibration behavior.

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agobcma: mips: fix clearing device IRQ
Rafał Miłecki [Mon, 10 Dec 2012 06:53:56 +0000 (07:53 +0100)]
bcma: mips: fix clearing device IRQ

commit cbbc0138efe1dcd5426b8fc5d87741f5057aee72 upstream.

We were using wrong IRQ number so clearing wasn't working at all.
Depending on a platform this could result in a one device having two
interrupts assigned. On BCM4706 this resulted in all IRQs being broken.

Cc: Hauke Mehrtens <hauke@hauke-m.de>
Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
Acked-by: Hauke Mehrtens <hauke@hauke-m.de>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoath9k: ar9003: fix OTP register offsets for AR9340
Gabor Juhos [Sun, 9 Dec 2012 22:57:09 +0000 (23:57 +0100)]
ath9k: ar9003: fix OTP register offsets for AR9340

commit b3cd8021379306c0be6932e4d3b4b01efc681769 upstream.

Trying to access the OTP memory on the AR9340
causes a data bus error like this:

  Data bus error, epc == 86e84164, ra == 86e84164
  Oops[#1]:
  Cpu 0
  $ 0   : 00000000 00000061 deadc0de 00000000
  $ 4   : b8115f18 00015f18 00000007 00000004
  $ 8   : 00000001 7c7c3c7c 7c7c7c7c 7c7c7c7c
  $12   : 7c7c3c7c 001f0041 00000000 7c7c7c3c
  $16   : 86ee0000 00015f18 00000000 00000007
  $20   : 00000004 00000064 00000004 86d71c44
  $24   : 00000000 86e6ca00
  $28   : 86d70000 86d71b20 86ece0c0 86e84164
  Hi    : 00000000
  Lo    : 00000064
  epc   : 86e84164 ath9k_hw_wait+0x58/0xb0 [ath9k_hw]
      Tainted: G           O
  ra    : 86e84164 ath9k_hw_wait+0x58/0xb0 [ath9k_hw]
  Status: 1100d403    KERNEL EXL IE
  Cause : 4080801c
  PrId  : 0001974c (MIPS 74Kc)
  Modules linked in: ath9k(O+) ath9k_common(O) ath9k_hw(O) ath(O) ar934x_nfc
  mac80211(O) usbcore usb_common scsi_mod nls_base nand nand_ecc nand_ids
  crc_ccitt cfg80211(O) compat(O) arc4 aes_generic crypto_blkcipher cryptomgr
  aead crypto_hash crypto_algapi ledtrig_timer ledtrig_default_on leds_gpio
  Process insmod (pid: 459, threadinfo=86d70000, task=87942140, tls=779ac440)
  Stack : 802fb500 000200da 804db150 804e0000 87816130 86ee0000 00010000 86d71b88
          86d71bc0 00000004 00000003 86e9fcd0 80305300 0002c0d0 86e74c50 800b4c20
          000003e8 00000001 00000000 86ee0000 000003ff 86e9fd64 80305300 80123938
          fffffffc 00000004 000058bc 00000000 86ea0000 86ee0000 000001ff 878d6000
          99999999 86e9fdc0 86ee0fcc 86e9e664 0000c0d0 86ee0000 0000700000007000
          ...
  Call Trace:
  [<86e84164>] ath9k_hw_wait+0x58/0xb0 [ath9k_hw]
  [<86e9fcd0>] ath9k_hw_setup_statusring+0x16b8/0x1c7c [ath9k_hw]

  Code: 0000a812  0040f809  00000000 <005310241054000b  24020001  0c05b5dc  2404000a  26520001

The cause of the error is that the OTP register
offsets are different on the AR9340 than the
actually used values.

Signed-off-by: Gabor Juhos <juhosg@openwrt.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoRevert "ath9k_hw: Update AR9003 high_power tx gain table"
Felix Fietkau [Thu, 6 Dec 2012 17:40:11 +0000 (18:40 +0100)]
Revert "ath9k_hw: Update AR9003 high_power tx gain table"

commit 9c170e068636deb3e3f96114034bb711675f0faa upstream.

This reverts commit f74b9d365ddd33a375802b064f96a5d0e99af7c0.

Turns out reverting commit a240dc7b3c7463bd60cf0a9b2a90f52f78aae0fd
"ath9k_hw: Updated AR9003 tx gain table for 5GHz" was not enough to
bring the tx power back to normal levels on devices like the
Buffalo WZR-HP-G450H, this one needs to be reverted as well.

This revert improves tx power by ~10 db on that device

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Cc: rmanohar@qca.qualcomm.com
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoregmap: debugfs: Avoid overflows for very small reads
Mark Brown [Mon, 10 Dec 2012 16:14:11 +0000 (01:14 +0900)]
regmap: debugfs: Avoid overflows for very small reads

commit db04328c167ff8e7c57f4a3532214aeada3a82fd upstream.

If count is less than the size of a register then we may hit integer
wraparound when trying to move backwards to check if we're still in
the buffer. Instead move the position forwards to check if it's still
in the buffer, we are unlikely to be able to allocate a buffer
sufficiently big to overflow here.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
[bwh: Backported to 3.2: formatted length is tot_len]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agodrm/radeon: fix eDP clk and lane setup for scaled modes
Alex Deucher [Tue, 4 Dec 2012 21:50:28 +0000 (16:50 -0500)]
drm/radeon: fix eDP clk and lane setup for scaled modes

commit 93927f9c1db5f55085457e820f0631064c7bfa34 upstream.

Need to use the adjusted mode since we are sending native
timing and using the scaler for non-native modes.

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agodrm/radeon/dce32+: use fractional fb dividers for high clocks
Alex Deucher [Tue, 13 Nov 2012 23:03:41 +0000 (18:03 -0500)]
drm/radeon/dce32+: use fractional fb dividers for high clocks

commit a02dc74b317d78298cb0587b9b1f6f741fd5c139 upstream.

Fixes flickering with some high res montiors.

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
[bwh: Backported to 3.2: use pll->flags instead of radeon_crtc->pll_flags]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agodrm/radeon/kms: use frac fb div on APUs
Alex Deucher [Thu, 19 Apr 2012 14:48:38 +0000 (10:48 -0400)]
drm/radeon/kms: use frac fb div on APUs

commit 37d4174d2d252c37dcb3d88cafae488542087848 upstream.

Seems to be more stable on certain monitors.

Fixes:
https://bugs.freedesktop.org/show_bug.cgi?id=48880

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
[bwh: Backported to 3.2: drop the ASIC_IS_DCE61() case]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agopnpacpi: fix incorrect TEST_ALPHA() test
Alan Cox [Fri, 7 Dec 2012 22:11:14 +0000 (23:11 +0100)]
pnpacpi: fix incorrect TEST_ALPHA() test

commit cdc87c5a30f407ed1ce43d8a22261116873d5ef1 upstream.

TEST_ALPHA() is broken and always returns 0.

[akpm@linux-foundation.org: return false for '@' as well, per Bjorn]
Signed-off-by: Alan Cox <alan@lxorguk.ukuu.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agodrm/i915: Close race between processing unpin task and queueing the flip
Chris Wilson [Mon, 3 Dec 2012 11:36:30 +0000 (11:36 +0000)]
drm/i915: Close race between processing unpin task and queueing the flip

commit e7d841ca03b7ab668620045cd7b428eda9f41601 upstream.

Before queuing the flip but crucially after attaching the unpin-work to
the crtc, we continue to setup the unpin-work. However, should the
hardware fire early, we see the connected unpin-work and queue the task.
The task then promptly runs and unpins the fb before we finish taking
the required references or even pinning it... Havoc.

To close the race, we use the flip-pending atomic to indicate when the
flip is finally setup and enqueued. So during the flip-done processing,
we can check more accurately whether the flip was expected.

v2: Add the appropriate mb() to ensure that the writes to the page-flip
worker are complete prior to marking it active and emitting the MI_FLIP.
On the read side, the mb should be enforced by the spinlocks.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
[danvet: Review the barriers a bit, we need a write barrier both
before and after updating ->pending. Similarly we need a read barrier
in the interrupt handler both before and after reading ->pending. With
well-ordered irqs only one barrier in each place should be required,
but since this patch explicitly sets out to combat spurious interrupts
with is staged activation of the unpin work we need to go full-bore on
the barriers, too. Discussed with Chris Wilson on irc and changes
acked by him.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
[bwh: Backported to 3.2: adjust context]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agotarget/file: Fix 32-bit highmem breakage for SGL -> iovec mapping
Sebastian Andrzej Siewior [Wed, 5 Dec 2012 11:08:29 +0000 (12:08 +0100)]
target/file: Fix 32-bit highmem breakage for SGL -> iovec mapping

commit 40ff2c3b3da35dd3a00ac6722056a59b4b3f2caf upstream.

This patch changes vectored file I/O to use kmap + kunmap when mapping
incoming SGL memory -> struct iovec in order to properly support 32-bit
highmem configurations.  This is because an extra bounce buffer may be
required when processing scatterlist pages allocated with GFP_KERNEL.

Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
[bwh: Backported to 3.2: use task->task_sg{,_nents} for iteration]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agox86,AMD: Power driver support for AMD's family 16h processors
Boris Ostrovsky [Wed, 5 Dec 2012 11:12:42 +0000 (06:12 -0500)]
x86,AMD: Power driver support for AMD's family 16h processors

commit 22e32f4f57778ebc6e17812fa3008361c05d64f9 upstream.

Add family 16h PCI ID to AMD's power driver to allow it report
power consumption on these processors.

Signed-off-by: Boris Ostrovsky <boris.ostrovsky@amd.com>
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoACPI / video: ignore BIOS initial backlight value for HP Folio 13-2000
Zhang Rui [Tue, 4 Dec 2012 22:30:19 +0000 (23:30 +0100)]
ACPI / video: ignore BIOS initial backlight value for HP Folio 13-2000

commit 129ff8f8d58297b04f47b5d6fad81aa2d08404e1 upstream.

Or else the laptop will boot with a dimmed screen.

References: https://bugzilla.kernel.org/show_bug.cgi?id=51141
Tested-by: Stefan Nagy <public@stefan-nagy.at>
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoALSA: usb-audio: Fix missing autopm for MIDI input
Takashi Iwai [Mon, 3 Dec 2012 10:30:50 +0000 (11:30 +0100)]
ALSA: usb-audio: Fix missing autopm for MIDI input

commit f5f165418cabf2218eb466c0e94693b8b1aee88b upstream.

The commit [88a8516a: ALSA: usbaudio: implement USB autosuspend] added
the support of autopm for USB MIDI output, but it didn't take the MIDI
input into account.

This patch adds the following for fixing the autopm:
- Manage the URB start at the first MIDI input stream open, instead of
  the time of instance creation
- Move autopm code to the common substream_open()
- Make snd_usbmidi_input_start/_stop() more robust and add the running
  state check

Reviewd-by: Clemens Ladisch <clemens@ladisch.de>
Tested-by: Clemens Ladisch <clemens@ladisch.de>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoALSA: usb-audio: Avoid autopm calls after disconnection
Takashi Iwai [Mon, 3 Dec 2012 10:12:46 +0000 (11:12 +0100)]
ALSA: usb-audio: Avoid autopm calls after disconnection

commit 59866da9e4ae54819e3c4e0a8f426bdb0c2ef993 upstream.

Add a similar protection against the disconnection race and the
invalid use of usb instance after disconnection, as well as we've done
for the USB audio PCM.

Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=51201

Reviewd-by: Clemens Ladisch <clemens@ladisch.de>
Tested-by: Clemens Ladisch <clemens@ladisch.de>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoBluetooth: Add missing lock nesting notation
Gustavo Padovan [Wed, 21 Nov 2012 01:25:54 +0000 (23:25 -0200)]
Bluetooth: Add missing lock nesting notation

commit dc2a0e20fbc85a71c63aa4330b496fda33f6bf80 upstream.

This patch fixes the following report, it happens when accepting rfcomm
connections:

[  228.165378] =============================================
[  228.165378] [ INFO: possible recursive locking detected ]
[  228.165378] 3.7.0-rc1-00536-gc1d5dc4 #120 Tainted: G        W
[  228.165378] ---------------------------------------------
[  228.165378] bluetoothd/1341 is trying to acquire lock:
[  228.165378]  (sk_lock-AF_BLUETOOTH-BTPROTO_RFCOMM){+.+...}, at:
[<ffffffffa0000aa0>] bt_accept_dequeue+0xa0/0x180 [bluetooth]
[  228.165378]
[  228.165378] but task is already holding lock:
[  228.165378]  (sk_lock-AF_BLUETOOTH-BTPROTO_RFCOMM){+.+...}, at:
[<ffffffffa0205118>] rfcomm_sock_accept+0x58/0x2d0 [rfcomm]
[  228.165378]
[  228.165378] other info that might help us debug this:
[  228.165378]  Possible unsafe locking scenario:
[  228.165378]
[  228.165378]        CPU0
[  228.165378]        ----
[  228.165378]   lock(sk_lock-AF_BLUETOOTH-BTPROTO_RFCOMM);
[  228.165378]   lock(sk_lock-AF_BLUETOOTH-BTPROTO_RFCOMM);
[  228.165378]
[  228.165378]  *** DEADLOCK ***
[  228.165378]
[  228.165378]  May be due to missing lock nesting notation

Signed-off-by: Gustavo Padovan <gustavo.padovan@collabora.co.uk>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agos390/cio: fix pgid reserved check
Sebastian Ott [Fri, 30 Nov 2012 15:48:59 +0000 (16:48 +0100)]
s390/cio: fix pgid reserved check

commit d99e79ec5574fc556c988f613ed6175f6de66f4a upstream.

The check to whom a device is reserved is done by checking the path
state of the affected channel paths. If it turns out that one path is
flagged as reserved by someone else the whole device is marked as such.

However the meaning of the RESVD_ELSE bit is that the addressed device
is reserved to a different pathgroup (and not reserved to a different
LPAR). If we do this test on a path which is currently not a member of
the pathgroup we could erroneously mark the device as reserved to
someone else.

To fix this collect the reserved state for all potential members of the
pathgroup and only mark the device as reserved if all of those potential
members have the RESVD_ELSE bit set.

Acked-by: Peter Oberparleiter <peter.oberparleiter@de.ibm.com>
Signed-off-by: Sebastian Ott <sebott@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agolibata: set dma_mode to 0xff in reset
Aaron Lu [Mon, 3 Dec 2012 03:35:02 +0000 (11:35 +0800)]
libata: set dma_mode to 0xff in reset

commit 5416912af75de9cba5d1c75b99a7888b0bbbd2fb upstream.

ata_device->dma_mode's initial value is zero, which is not a valid dma
mode, but ata_dma_enabled will return true for this value. This patch
sets dma_mode to 0xff in reset function, so that ata_dma_enabled will
not return true for this case, or it will cause problem for pata_acpi.

The corrsponding bugzilla page is at:
https://bugzilla.kernel.org/show_bug.cgi?id=49151

Reported-by: Phillip Wood <phillip.wood@dunelm.org.uk>
Signed-off-by: Aaron Lu <aaron.lu@intel.com>
Tested-by: Szymon Janc <szymon@janc.net.pl>
Tested-by: Dutra Julio <dutra.julio@gmail.com>
Acked-by: Alan Cox <alan@linux.intel.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoHID: Add Apple wireless keyboard 2011 ANSI to special driver list
Ben Hutchings [Sun, 2 Dec 2012 14:38:23 +0000 (14:38 +0000)]
HID: Add Apple wireless keyboard 2011 ANSI to special driver list

commit f9af7b9edccb87d4d80b58687ab63e58f3b64c4c upstream.

Commit 0a97e1e9f9a6 ('HID: apple: Add Apple wireless keyboard 2011 ANSI PID')
did not update the special driver list in hid-core.c, so hid-generic may
still bind to this device.

Reported-by: Ari Pollak <ari@scvngr.com>
References: http://bugs.debian.org/694546
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
9 years agoHID: apple: Add Apple wireless keyboard 2011 ANSI PID
Alexey Kaminsky [Mon, 23 Apr 2012 16:02:18 +0000 (18:02 +0200)]
HID: apple: Add Apple wireless keyboard 2011 ANSI PID

commit 0a97e1e9f9a6765e6243030ac42b04694f3f3647 upstream.

Signed-off-by: Alexey Kaminsky <me@akaminsky.net>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
[bwh: Backported to 3.2: add the device ID to hid-ids.h]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agofirewire: net: Fix handling of fragmented multicast/broadcast packets.
Stephan Gatzka [Wed, 28 Nov 2012 19:04:32 +0000 (20:04 +0100)]
firewire: net: Fix handling of fragmented multicast/broadcast packets.

commit 9d2373420900a39f5212a3b289331aa3535b1000 upstream.

This patch fixes both the transmit and receive portion of sending
fragmented mutlicast and broadcast packets.

The transmit section was broken because the offset for INTFRAG and
LASTFRAG packets were just miscalculated by IEEE1394_GASP_HDR_SIZE (which
was reserved with skb_push() in fwnet_send_packet).

The receive section was broken because in fwnet_incoming_packet is a call
to fwnet_peer_find_by_node_id(). Called with generation == -1 it will
not find a peer and the partial datagrams are associated to a peer.

[Stefan R:  The fix to use context->card->generation is not perfect.
It relies on the IR tasklet which processes packets from the prior bus
generation to run before the self-ID-complete worklet which sets the
current card generation.  Alas, there is no simple way of a race-free
implementation.  Let's do it this way for now.]

Signed-off-by: Stephan Gatzka <stephan.gatzka@gmail.com>
Signed-off-by: Stefan Richter <stefanr@s5r6.in-berlin.de>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoring-buffer: Fix NULL pointer if rb_set_head_page() fails
Steven Rostedt [Fri, 30 Nov 2012 03:27:22 +0000 (22:27 -0500)]
ring-buffer: Fix NULL pointer if rb_set_head_page() fails

commit 54f7be5b831254199522523ccab4c3d954bbf576 upstream.

The function rb_set_head_page() searches the list of ring buffer
pages for a the page that has the HEAD page flag set. If it does
not find it, it will do a WARN_ON(), disable the ring buffer and
return NULL, as this should never happen.

But if this bug happens to happen, not all callers of this function
can handle a NULL pointer being returned from it. That needs to be
fixed.

Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoACPI / PNP: Do not crash due to stale pointer use during system resume
Rafael J. Wysocki [Fri, 30 Nov 2012 12:05:05 +0000 (13:05 +0100)]
ACPI / PNP: Do not crash due to stale pointer use during system resume

commit a6b5e88c0e42093b9057856f35770966c8c591e3 upstream.

During resume from system suspend the 'data' field of
struct pnp_dev in pnpacpi_set_resources() may be a stale pointer,
due to removal of the associated ACPI device node object in the
previous suspend-resume cycle.  This happens, for example, if a
dockable machine is booted in the docking station and then suspended
and resumed and suspended again.  If that happens,
pnpacpi_build_resource_template() called from pnpacpi_set_resources()
attempts to use that pointer and crashes.

However, pnpacpi_set_resources() actually checks the device's ACPI
handle, attempts to find the ACPI device node object attached to it
and returns an error code if that fails, so in fact it knows what the
correct value of dev->data should be.  Use this observation to update
dev->data with the correct value if necessary and dump a call trace
if that's the case (once).

We still need to fix the root cause of this issue, but preventing
systems from crashing because of it is an improvement too.

Reported-and-tested-by: Zdenek Kabelac <zdenek.kabelac@gmail.com>
References: https://bugzilla.kernel.org/show_bug.cgi?id=51071
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoACPI / video: Add "Asus UL30VT" to ACPI video detect blacklist
Lan Tianyu [Fri, 30 Nov 2012 12:02:50 +0000 (13:02 +0100)]
ACPI / video: Add "Asus UL30VT" to ACPI video detect blacklist

commit d0c2ce16bec0afa6013b4c5220ca4c9c67210215 upstream.

The ACPI video driver can't control backlight correctly on
Asus UL30VT.  Vendor driver (asus-laptop) can work.  This patch is to
add "Asus UL30VT" to ACPI video detect blacklist in order to use
asus-laptop for video control on the "Asus UL30VT" rather than ACPI
video driver.

References: https://bugzilla.kernel.org/show_bug.cgi?id=32592
Reported-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Lan Tianyu <tianyu.lan@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoacpi/video_detect: blacklist samsung x360
Corentin Chary [Wed, 13 Jun 2012 07:32:04 +0000 (09:32 +0200)]
acpi/video_detect: blacklist samsung x360

commit 084940d5b101e9ca91a689eb5048151b14076839 upstream.

On Samsung X360, the BIOS will set a flag (VDRV) if the generic
ACPI backlight device is used. This flag will definitively break
the backlight interface (even the vendor interface) untill next
reboot. It's why we should prevent video.ko from being used here
and we can't rely on a later call to acpi_video_unregister().

Signed-off-by: Corentin Chary <corentin.chary@gmail.com>
Signed-off-by: Matthew Garrett <mjg@redhat.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoACPI: do acpisleep dmi check when CONFIG_ACPI_SLEEP is set
Zhang Rui [Fri, 30 Nov 2012 11:57:03 +0000 (12:57 +0100)]
ACPI: do acpisleep dmi check when CONFIG_ACPI_SLEEP is set

commit 0ac1b1d7b7424cd6f129b5454b504b3cae746f0e upstream.

The current acpisleep DMI checks only run when CONFIG_SUSPEND is set.
And this may break hibernation on some platforms when CONFIG_SUSPEND
is cleared.

Move acpisleep DMI check into #ifdef CONFIG_ACPI_SLEEP instead.

[rjw: Added acpi_sleep_dmi_check() and rebased on top of earlier
 patches adding entries to acpisleep_dmi_table[].]
References: https://bugzilla.kernel.org/show_bug.cgi?id=45921
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoqla2xxx: Test and clear FCPORT_UPDATE_NEEDED atomically.
David Jeffery [Wed, 21 Nov 2012 07:39:54 +0000 (02:39 -0500)]
qla2xxx: Test and clear FCPORT_UPDATE_NEEDED atomically.

commit a394aac88506159e047630fc90dc2242568382d8 upstream.

When the qla2xxx driver loses access to multiple, remote ports, there is a race
condition which can occur which will keep the request stuck on a scsi request
queue indefinitely.

This bad state occurred do to a race condition with how the FCPORT_UPDATE_NEEDED
bit is set in qla2x00_schedule_rport_del(), and how it is cleared in
qla2x00_do_dpc().  The problem port has its drport pointer set, but it has never
been processed by the driver to inform the fc transport that the port has been
lost.  qla2x00_schedule_rport_del() sets drport, and then sets the
FCPORT_UPDATE_NEEDED bit.  In qla2x00_do_dpc(), the port lists are walked and
any drport pointer is handled and the fc transport informed of the port loss,
then the FCPORT_UPDATE_NEEDED bit is cleared.  This leaves a race where the
dpc thread is processing one port removal, another port removal is marked
with a call to qla2x00_schedule_rport_del(), and the dpc thread clears the
bit for both removals, even though only the first removal was actually
handled.  Until another event occurs to set FCPORT_UPDATE_NEEDED, the later
port removal is never finished and qla2xxx stays in a bad state which causes
requests to become stuck on request queues.

This patch updates the driver to test and clear FCPORT_UPDATE_NEEDED
atomically.  This ensures the port state changes are processed and not lost.

Signed-off-by: David Jeffery <djeffery@redhat.com>
Signed-off-by: Chad Dupuis <chad.dupuis@qlogic.com>
Signed-off-by: Saurav Kashyap <saurav.kashyap@qlogic.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agomvsas: fix undefined bit shift
Xi Wang [Fri, 16 Nov 2012 19:40:03 +0000 (14:40 -0500)]
mvsas: fix undefined bit shift

commit beecadea1b8d67f591b13f7099559f32f3fd601d upstream.

The macro bit(n) is defined as ((u32)1 << n), and thus it doesn't work
with n >= 32, such as in mvs_94xx_assign_reg_set():

if (i >= 32) {
mvi->sata_reg_set |= bit(i);
...
}

The shift ((u32)1 << n) with n >= 32 also leads to undefined behavior.
The result varies depending on the architecture.

This patch changes bit(n) to do a 64-bit shift.  It also simplifies
mv_ffc64() using __ffs64(), since invoking ffz() with ~0 is undefined.

Signed-off-by: Xi Wang <xi.wang@gmail.com>
Acked-by: Xiangliang Yu <yuxiangl@marvell.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoprevent stack buffer overflow in host_reset
Sasha Levin [Thu, 15 Nov 2012 20:51:46 +0000 (15:51 -0500)]
prevent stack buffer overflow in host_reset

commit 072f19b4bea31cdd482d79f805413f2f9ac9e233 upstream.

store_host_reset() has tried to re-invent the wheel to compare sysfs strings.
Unfortunately it did so poorly and never bothered to check the input from
userspace before overwriting stack with it, so something simple as:

echo "WoopsieWoopsie" >
/sys/devices/pseudo_0/adapter0/host0/scsi_host/host0/host_reset

would result in:

[  316.310101] Kernel panic - not syncing: stack-protector: Kernel stack is corrupted in: ffffffff81f5bac7
[  316.310101]
[  316.320051] Pid: 6655, comm: sh Tainted: G        W    3.7.0-rc5-next-20121114-sasha-00016-g5c9d68d-dirty #129
[  316.320051] Call Trace:
[  316.340058] pps pps0: PPS event at 1352918752.620355751
[  316.340062] pps pps0: capture assert seq #303
[  316.320051]  [<ffffffff83b3856b>] panic+0xcd/0x1f4
[  316.320051]  [<ffffffff81f5bac7>] ? store_host_reset+0xd7/0x100
[  316.320051]  [<ffffffff8110b996>] __stack_chk_fail+0x16/0x20
[  316.320051]  [<ffffffff81f5bac7>] store_host_reset+0xd7/0x100
[  316.320051]  [<ffffffff81e55bb3>] dev_attr_store+0x13/0x30
[  316.320051]  [<ffffffff812f7db1>] sysfs_write_file+0x101/0x170
[  316.320051]  [<ffffffff8127acc8>] vfs_write+0xb8/0x180
[  316.320051]  [<ffffffff8127ae80>] sys_write+0x50/0xa0
[  316.320051]  [<ffffffff83c03418>] tracesys+0xe1/0xe6

Fix this by uninventing whatever was going on there and just use sysfs_streq.

Bug introduced by 29443691 ("[SCSI] scsi: Added support for adapter and
firmware reset").

[jejb: added necessary const to prevent compile warnings]
Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agocan: Do not call dev_put if restart timer is running upon close
Alexander Stein [Tue, 27 Nov 2012 07:52:34 +0000 (08:52 +0100)]
can: Do not call dev_put if restart timer is running upon close

commit ab48b03ec9ae1840a1e427e2375bd0d9d554b4ed upstream.

If the restart timer is running due to BUS-OFF and the device is
disconnected an dev_put will decrease the usage counter to -1 thus
blocking the interface removal, resulting in the following dmesg
lines repeating every 10s:
can: notifier: receive list not found for dev can0
can: notifier: receive list not found for dev can0
can: notifier: receive list not found for dev can0
unregister_netdevice: waiting for can0 to become free. Usage count = -1

Signed-off-by: Alexander Stein <alexander.stein@systec-electronic.com>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoUSB: mark uas driver as BROKEN
Greg KH [Wed, 28 Nov 2012 18:19:16 +0000 (10:19 -0800)]
USB: mark uas driver as BROKEN

commit fb37ef98015f864d22be223a0e0d93547cd1d4ef upstream.

As reported https://bugzilla.kernel.org/show_bug.cgi?id=51031, the UAS
driver causes problems and has been asked to be not built into any of
the major distributions.  To prevent users from running into problems
with it, and for distros that were not notified, just mark the whole
thing as broken.

Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Acked-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoxhci: Add Lynx Point LP to list of Intel switchable hosts
Russell Webb [Fri, 9 Nov 2012 21:58:49 +0000 (13:58 -0800)]
xhci: Add Lynx Point LP to list of Intel switchable hosts

commit bb1e5dd7113d2fd178d3af9aca8f480ae0468edf upstream.

Like Lynx Point, Lynx Point LP is also switchable.  See
1c12443ab8eba71a658fae4572147e56d1f84f66 for more details.

This patch should be backported to stable kernels as old as 3.0,
that contain commit 69e848c2090aebba5698a1620604c7dccb448684
"Intel xhci: Support EHCI/xHCI port switching."

Signed-off-by: Russell Webb <russell.webb@linux.intel.com>
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agosata_promise: fix hardreset lockdep error
Mikael Pettersson [Sun, 16 Sep 2012 18:53:43 +0000 (20:53 +0200)]
sata_promise: fix hardreset lockdep error

commit 3100d49d3cd236443faae9d81137c81b22d36003 upstream.

sata_promise's pdc_hard_reset_port() needs to serialize because it
flips a port-specific bit in controller register that's shared by
all ports. The code takes the ata host lock for this, but that's
broken because an interrupt may arrive on our irq during the hard
reset sequence, and that too will take the ata host lock. With
lockdep enabled a big nasty warning is seen.

Fixed by adding private state to the ata host structure, containing
a second lock used only for serializing the hard reset sequences.
This eliminated the lockdep warnings both on my test rig and on
the original reporter's machine.

Signed-off-by: Mikael Pettersson <mikpe@it.uu.se>
Tested-by: Adko Branil <adkobranil@yahoo.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoiscsit: use GFP_ATOMIC under spin lock
Wei Yongjun [Fri, 23 Nov 2012 04:07:39 +0000 (12:07 +0800)]
iscsit: use GFP_ATOMIC under spin lock

commit 3c989d7603872bf878840f7ce3ea49b73bea4c6c upstream.

The function iscsit_build_conn_drop_async_message() is called
from iscsit_close_connection() with spin lock 'sess->conn_lock'
held, so we should use GFP_ATOMIC instead of GFP_KERNEL.

Signed-off-by: Wei Yongjun <yongjun_wei@trendmicro.com.cn>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoUSB: ftdi_sio: Add support for Newport AGILIS motor drivers
Martin Teichmann [Wed, 21 Nov 2012 15:45:07 +0000 (16:45 +0100)]
USB: ftdi_sio: Add support for Newport AGILIS motor drivers

commit d7e14b375b40c04cd735b115713043b69a2c68ac upstream.

The Newport AGILIS model AG-UC8 compact piezo motor controller
(http://search.newport.com/?q=*&x2=sku&q2=AG-UC8)
is yet another device using an FTDI USB-to-serial chip. It works
fine with the ftdi_sio driver when adding

  options ftdi-sio product=0x3000 vendor=0x104d

to modprobe.d. udevadm reports "Newport" as the manufacturer,
and "Agilis" as the product name.

Signed-off-by: Martin Teichmann <lkb.teichmann@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agousb: ftdi_sio: fixup BeagleBone A5+ quirk
Peter Korsgaard [Thu, 22 Nov 2012 15:30:46 +0000 (16:30 +0100)]
usb: ftdi_sio: fixup BeagleBone A5+ quirk

commit 1a88d5eee2ef2ad1d3c4e32043e9c4c5347d4fc1 upstream.

BeagleBone A5+ devices ended up getting shipped with the
'BeagleBone/XDS100V2' product string, and not XDS100 like it
was agreed, so adjust the quirk to match.

For details, see the thread on the beagle list:

https://groups.google.com/forum/#!msg/beagleboard/zrFPew9_Wvo/ibWr1-eE8JwJ

Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoUSB: cp210x: add Virtenio Preon32 device id
Markus Becker [Thu, 22 Nov 2012 08:41:23 +0000 (09:41 +0100)]
USB: cp210x: add Virtenio Preon32 device id

commit 356fe44f4b8ece867bdb9876b1854d7adbef9de2 upstream.

Signed-off-by: Markus Becker <mab@comnets.uni-bremen.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoUSB: option: blacklist network interface on Huawei E173
Bjørn Mork [Sun, 25 Nov 2012 16:05:10 +0000 (17:05 +0100)]
USB: option: blacklist network interface on Huawei E173

commit f36446cf9bbebaa03a80d95cfeeafbaf68218249 upstream.

The Huawei E173 will normally appear as 12d1:1436 in Linux.  But
the modem has another mode with different device ID and a slightly
different set of descriptors. This is the mode used by Windows like
this:

  3Modem:      USB\VID_12D1&PID_140C&MI_00\6&3A1D2012&0&0000
  Networkcard: USB\VID_12D1&PID_140C&MI_01\6&3A1D2012&0&0001
  Appli.Inter: USB\VID_12D1&PID_140C&MI_02\6&3A1D2012&0&0002
  PC UI Inter: USB\VID_12D1&PID_140C&MI_03\6&3A1D2012&0&0003

All interfaces have the same ff/ff/ff class codes in this mode.
Blacklisting the network interface to allow it to be picked up by
the network driver.

Reported-by: Thomas Schäfer <tschaefer@t-online.de>
Signed-off-by: Bjørn Mork <bjorn@mork.no>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoUSB: OHCI: workaround for hardware bug: retired TDs not added to the Done Queue
Alan Stern [Mon, 26 Nov 2012 17:36:21 +0000 (12:36 -0500)]
USB: OHCI: workaround for hardware bug: retired TDs not added to the Done Queue

commit 50ce5c0683aa83eb161624ea89daa5a9eee0c2ce upstream.

This patch (as1636) is a partial workaround for a hardware bug
affecting OHCI controllers by NVIDIA at least, maybe others too.  When
the controller retires a Transfer Descriptor, it is supposed to add
the TD onto the Done Queue.  But sometimes this doesn't happen, with
the result that ohci-hcd never realizes the corresponding transfer has
finished.  Symptoms can vary; a typical result is that USB audio stops
working after a while.

The patch works around the problem by recognizing that TDs are always
processed in order.  Therefore, if a later TD is found on the Done
Queue than all the earlier TDs for the same endpoint must be finished
as well.

Unfortunately this won't solve the problem in cases where the missing
TD is the last one in the endpoint's queue.  A complete fix would
require a signficant amount of change to the driver.

Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Tested-by: Oliver Neukum <oneukum@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agonfsd: fix v4 reply caching
J. Bruce Fields [Fri, 16 Nov 2012 20:22:43 +0000 (15:22 -0500)]
nfsd: fix v4 reply caching

commit 57d276d71aef7d8305ff002a070cb98deb2edced upstream.

Very embarassing: 1091006c5eb15cba56785bd5b498a8d0b9546903 "nfsd: turn
on reply cache for NFSv4" missed a line, effectively leaving the reply
cache off in the v4 case.  I thought I'd tested that, but I guess not.

This time, wrote a pynfs test to confirm it works.

Signed-off-by: J. Bruce Fields <bfields@redhat.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoACPI / PM: Add Sony Vaio VPCEB1S1E to nonvs blacklist.
Lan Tianyu [Wed, 21 Nov 2012 22:12:12 +0000 (23:12 +0100)]
ACPI / PM: Add Sony Vaio VPCEB1S1E to nonvs blacklist.

commit 876ab79055019e248508cfd0dee7caa3c0c831ed upstream.

Sony Vaio VPCEB1S1E does not resume correctly without
acpi_sleep=nonvs, so add it to the ACPI sleep blacklist.

References: https://bugzilla.kernel.org/show_bug.cgi?id=48781
Reported-by: Sébastien Wilmet <swilmet@gnome.org>
Signed-off-by: Lan Tianyu <tianyu.lan@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoUSB: add new zte 3g-dongle's pid to option.c
li.rui27@zte.com.cn [Tue, 20 Nov 2012 06:31:47 +0000 (14:31 +0800)]
USB: add new zte 3g-dongle's pid to option.c

commit 31b6a1048b7292efff8b5b53ae3d9d29adde385e upstream.

Signed-off-by: Rui li <li.rui27@zte.com.cn>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agodrm/i915: make the panel fitter work on pipes B and C on IVB
Paulo Zanoni [Tue, 20 Nov 2012 15:27:41 +0000 (13:27 -0200)]
drm/i915: make the panel fitter work on pipes B and C on IVB

commit 13888d78c664a1f61d7b09d282f5916993827a40 upstream.

I actually found this problem on Haswell, but then discovered Ivy
Bridge also has it by reading the spec.

I don't have the hardware to test this.

Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Reviewed-by: Damien Lespiau <damien.lespiau@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoNFS: Add sequence_priviliged_ops for nfs4_proc_sequence()
Bryan Schumaker [Mon, 12 Nov 2012 21:55:38 +0000 (16:55 -0500)]
NFS: Add sequence_priviliged_ops for nfs4_proc_sequence()

commit 6bdb5f213c4344324f600dde885f25768fbd14db upstream.

If I mount an NFS v4.1 server to a single client multiple times and then
run xfstests over each mountpoint I usually get the client into a state
where recovery deadlocks.  The server informs the client of a
cb_path_down sequence error, the client then does a
bind_connection_to_session and checks the status of the lease.

I found that bind_connection_to_session sets the NFS4_SESSION_DRAINING
flag on the client, but this flag is never unset before
nfs4_check_lease() reaches nfs4_proc_sequence().  This causes the client
to deadlock, halting all NFS activity to the server.  nfs4_proc_sequence()
is only called by the state manager, so I can change it to run in privileged
mode to bypass the NFS4_SESSION_DRAINING check and avoid the deadlock.

Signed-off-by: Bryan Schumaker <bjschuma@netapp.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agomfd: Only unregister platform devices allocated by the mfd core
Charles Keepax [Fri, 9 Nov 2012 16:15:28 +0000 (16:15 +0000)]
mfd: Only unregister platform devices allocated by the mfd core

commit b9fbb62eb61452d728c39b2e5020739c575aac53 upstream.

mfd_remove_devices would iterate over all devices sharing a parent with
an mfd device regardless of whether they were allocated by the mfd core
or not. This especially caused problems when the device structure was
not contained within a platform_device, because to_platform_device is
used on each device pointer.

This patch defines a device_type for mfd devices and checks this is
present from mfd_remove_devices_fn before processing the device.

Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Tested-by: Peter Tyser <ptyser@xes-inc.com>
Reviewed-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agodrm: fix documentation for drm_crtc_set_mode()
Alex Deucher [Fri, 9 Nov 2012 17:26:32 +0000 (17:26 +0000)]
drm: fix documentation for drm_crtc_set_mode()

commit 4c9287c6009b37754c42e0ba73a4cc79de92d8f8 upstream.

x and y parameters are offsets, not width/height

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoregulator: wm831x: Set the new rather than old value for DVS VSEL
Mark Brown [Tue, 20 Nov 2012 01:02:06 +0000 (10:02 +0900)]
regulator: wm831x: Set the new rather than old value for DVS VSEL

commit 13ae633cf729b0ecb677b75b04886ff8fada8fad upstream.

Reported-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoBluetooth: Add support for BCM20702A0 [0b05, 17b5]
Jeff Cook [Fri, 9 Nov 2012 23:39:48 +0000 (16:39 -0700)]
Bluetooth: Add support for BCM20702A0 [0b05, 17b5]

commit 1ee3ff6110c16acfc915a79b1e3feb5013c41e75 upstream.

Vendor-specific ID for BCM20702A0.
Support for bluetooth over Asus Wi-Fi GO!, included with Asus P8Z77-V
Deluxe.

T:  Bus=07 Lev=02 Prnt=02 Port=00 Cnt=01 Dev#=  3 Spd=12  MxCh= 0
D:  Ver= 2.00 Cls=ff(vend.) Sub=01 Prot=01 MxPS=64 #Cfgs=  1
P:  Vendor=0b05 ProdID=17b5 Rev=01.12
S:  Manufacturer=Broadcom Corp
S:  Product=BCM20702A0
S:  SerialNumber=94DBC98AC113
C:  #Ifs= 4 Cfg#= 1 Atr=e0 MxPwr=0mA
I:  If#= 0 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=01 Prot=01 Driver=(none)
I:  If#= 1 Alt= 0 #EPs= 2 Cls=ff(vend.) Sub=01 Prot=01 Driver=(none)
I:  If#= 2 Alt= 0 #EPs= 2 Cls=ff(vend.) Sub=ff Prot=ff Driver=(none)
I:  If#= 3 Alt= 0 #EPs= 0 Cls=fe(app. ) Sub=01 Prot=01 Driver=(none)

Signed-off-by: Jeff Cook <jeff@deserettechnology.com>
Signed-off-by: Gustavo Padovan <gustavo.padovan@collabora.co.uk>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoBluetooth: ath3k: Add support for VAIO VPCEH [0489:e027]
Marcos Chaparro [Tue, 6 Nov 2012 19:19:11 +0000 (16:19 -0300)]
Bluetooth: ath3k: Add support for VAIO VPCEH [0489:e027]

commit acd9454433e28c1a365d8b069813c35c1c3a8ac3 upstream.

Added Atheros AR3011 internal bluetooth device found in Sony VAIO VPCEH to the
devices list.
Before this, the bluetooth module was identified as an Foxconn / Hai bluetooth
device [0489:e027], now it claims to be an AtherosAR3011 Bluetooth
[0cf3:3005].

T:  Bus=01 Lev=02 Prnt=02 Port=04 Cnt=02 Dev#=  4 Spd=12   MxCh= 0
D:  Ver= 1.10 Cls=e0(wlcon) Sub=01 Prot=01 MxPS=64 #Cfgs=  1
P:  Vendor=0489 ProdID=e027 Rev= 0.01
C:* #Ifs= 2 Cfg#= 1 Atr=e0 MxPwr=100mA
I:* If#= 0 Alt= 0 #EPs= 3 Cls=e0(wlcon) Sub=01 Prot=01 Driver=btusb
E:  Ad=81(I) Atr=03(Int.) MxPS=  16 Ivl=1ms
E:  Ad=82(I) Atr=02(Bulk) MxPS=  64 Ivl=0ms
E:  Ad=02(O) Atr=02(Bulk) MxPS=  64 Ivl=0ms
I:* If#= 1 Alt= 0 #EPs= 2 Cls=e0(wlcon) Sub=01 Prot=01 Driver=btusb
E:  Ad=83(I) Atr=01(Isoc) MxPS=   0 Ivl=1ms
E:  Ad=03(O) Atr=01(Isoc) MxPS=   0 Ivl=1ms
I:  If#= 1 Alt= 1 #EPs= 2 Cls=e0(wlcon) Sub=01 Prot=01 Driver=btusb
E:  Ad=83(I) Atr=01(Isoc) MxPS=   9 Ivl=1ms
E:  Ad=03(O) Atr=01(Isoc) MxPS=   9 Ivl=1ms
I:  If#= 1 Alt= 2 #EPs= 2 Cls=e0(wlcon) Sub=01 Prot=01 Driver=btusb
E:  Ad=83(I) Atr=01(Isoc) MxPS=  17 Ivl=1ms
E:  Ad=03(O) Atr=01(Isoc) MxPS=  17 Ivl=1ms
I:  If#= 1 Alt= 3 #EPs= 2 Cls=e0(wlcon) Sub=01 Prot=01 Driver=btusb
E:  Ad=83(I) Atr=01(Isoc) MxPS=  25 Ivl=1ms
E:  Ad=03(O) Atr=01(Isoc) MxPS=  25 Ivl=1ms
I:  If#= 1 Alt= 4 #EPs= 2 Cls=e0(wlcon) Sub=01 Prot=01 Driver=btusb
E:  Ad=83(I) Atr=01(Isoc) MxPS=  33 Ivl=1ms
E:  Ad=03(O) Atr=01(Isoc) MxPS=  33 Ivl=1ms
I:  If#= 1 Alt= 5 #EPs= 2 Cls=e0(wlcon) Sub=01 Prot=01 Driver=btusb
E:  Ad=83(I) Atr=01(Isoc) MxPS=  49 Ivl=1ms
E:  Ad=03(O) Atr=01(Isoc) MxPS=  49 Ivl=1ms

Signed-off-by: Marcos Chaparro <marcos@mrkindustries.com.ar>
Signed-off-by: Gustavo Padovan <gustavo.padovan@collabora.co.uk>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoBluetooth: Add support for BCM20702A0 [04ca, 2003]
Jaroslav Resler [Tue, 11 Sep 2012 09:25:32 +0000 (17:25 +0800)]
Bluetooth: Add support for BCM20702A0 [04ca, 2003]

commit 0c1abbd1aa0416258881c303a88e618cbca0759c upstream.

Add another vendor specific ID for BCM20702A0.

output of usb-devices:
T:  Bus=01 Lev=02 Prnt=02 Port=03 Cnt=02 Dev#=  4 Spd=12   MxCh= 0
D:  Ver= 2.00 Cls=ff(vend.) Sub=01 Prot=01 MxPS=64 #Cfgs=  1
P:  Vendor=04ca ProdID=2003 Rev= 1.12
S:  Manufacturer=Broadcom Corp
S:  Product=BCM20702A0
S:  SerialNumber=446D57861623
C:* #Ifs= 4 Cfg#= 1 Atr=e0 MxPwr=  0mA
I:* If#= 0 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=01 Prot=01 Driver=btusb
E:  Ad=81(I) Atr=03(Int.) MxPS=  16 Ivl=1ms
E:  Ad=82(I) Atr=02(Bulk) MxPS=  64 Ivl=0ms
E:  Ad=02(O) Atr=02(Bulk) MxPS=  64 Ivl=0ms
I:* If#= 1 Alt= 0 #EPs= 2 Cls=ff(vend.) Sub=01 Prot=01 Driver=btusb
E:  Ad=83(I) Atr=01(Isoc) MxPS=   0 Ivl=1ms
E:  Ad=03(O) Atr=01(Isoc) MxPS=   0 Ivl=1ms
I:  If#= 1 Alt= 1 #EPs= 2 Cls=ff(vend.) Sub=01 Prot=01 Driver=btusb
E:  Ad=83(I) Atr=01(Isoc) MxPS=   9 Ivl=1ms
E:  Ad=03(O) Atr=01(Isoc) MxPS=   9 Ivl=1ms
I:  If#= 1 Alt= 2 #EPs= 2 Cls=ff(vend.) Sub=01 Prot=01 Driver=btusb
E:  Ad=83(I) Atr=01(Isoc) MxPS=  17 Ivl=1ms
E:  Ad=03(O) Atr=01(Isoc) MxPS=  17 Ivl=1ms
I:  If#= 1 Alt= 3 #EPs= 2 Cls=ff(vend.) Sub=01 Prot=01 Driver=btusb
E:  Ad=83(I) Atr=01(Isoc) MxPS=  25 Ivl=1ms
E:  Ad=03(O) Atr=01(Isoc) MxPS=  25 Ivl=1ms
I:  If#= 1 Alt= 4 #EPs= 2 Cls=ff(vend.) Sub=01 Prot=01 Driver=btusb
E:  Ad=83(I) Atr=01(Isoc) MxPS=  33 Ivl=1ms
E:  Ad=03(O) Atr=01(Isoc) MxPS=  33 Ivl=1ms
I:  If#= 1 Alt= 5 #EPs= 2 Cls=ff(vend.) Sub=01 Prot=01 Driver=btusb
E:  Ad=83(I) Atr=01(Isoc) MxPS=  49 Ivl=1ms
E:  Ad=03(O) Atr=01(Isoc) MxPS=  49 Ivl=1ms
I:* If#= 2 Alt= 0 #EPs= 2 Cls=ff(vend.) Sub=ff Prot=ff Driver=(none)
E:  Ad=84(I) Atr=02(Bulk) MxPS=  32 Ivl=0ms
E:  Ad=04(O) Atr=02(Bulk) MxPS=  32 Ivl=0ms
I:* If#= 3 Alt= 0 #EPs= 0 Cls=fe(app. ) Sub=01 Prot=01 Driver=(none)

Signed-off-by: Cho, Yu-Chen <acho@suse.com>
Signed-off-by: Gustavo Padovan <gustavo.padovan@collabora.co.uk>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agocgroup: remove incorrect dget/dput() pair in cgroup_create_dir()
Tejun Heo [Mon, 19 Nov 2012 16:13:35 +0000 (08:13 -0800)]
cgroup: remove incorrect dget/dput() pair in cgroup_create_dir()

commit 175431635ec09b1d1bba04979b006b99e8305a83 upstream.

cgroup_create_dir() does weird dancing with dentry refcnt.  On
success, it gets and then puts it achieving nothing.  On failure, it
puts but there isn't no matching get anywhere leading to the following
oops if cgroup_create_file() fails for whatever reason.

  ------------[ cut here ]------------
  kernel BUG at /work/os/work/fs/dcache.c:552!
  invalid opcode: 0000 [#1] PREEMPT SMP DEBUG_PAGEALLOC
  Modules linked in:
  CPU 2
  Pid: 697, comm: mkdir Not tainted 3.7.0-rc4-work+ #3 Bochs Bochs
  RIP: 0010:[<ffffffff811d9c0c>]  [<ffffffff811d9c0c>] dput+0x1dc/0x1e0
  RSP: 0018:ffff88001a3ebef8  EFLAGS: 00010246
  RAX: 0000000000000000 RBX: ffff88000e5b1ef8 RCX: 0000000000000403
  RDX: 0000000000000303 RSI: 2000000000000000 RDI: ffff88000e5b1f58
  RBP: ffff88001a3ebf18 R08: ffffffff82c76960 R09: 0000000000000001
  R10: ffff880015022080 R11: ffd9bed70f48a041 R12: 00000000ffffffea
  R13: 0000000000000001 R14: ffff88000e5b1f58 R15: 00007fff57656d60
  FS:  00007ff05fcb3800(0000) GS:ffff88001fd00000(0000) knlGS:0000000000000000
  CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
  CR2: 00000000004046f0 CR3: 000000001315f000 CR4: 00000000000006e0
  DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
  DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
  Process mkdir (pid: 697, threadinfo ffff88001a3ea000, task ffff880015022080)
  Stack:
   ffff88001a3ebf48 00000000ffffffea 0000000000000001 0000000000000000
   ffff88001a3ebf38 ffffffff811cc889 0000000000000001 ffff88000e5b1ef8
   ffff88001a3ebf68 ffffffff811d1fc9 ffff8800198d7f18 ffff880019106ef8
  Call Trace:
   [<ffffffff811cc889>] done_path_create+0x19/0x50
   [<ffffffff811d1fc9>] sys_mkdirat+0x59/0x80
   [<ffffffff811d2009>] sys_mkdir+0x19/0x20
   [<ffffffff81be1e02>] system_call_fastpath+0x16/0x1b
  Code: 00 48 8d 90 18 01 00 00 48 89 93 c0 00 00 00 4c 89 a0 18 01 00 00 48 8b 83 a0 00 00 00 83 80 28 01 00 00 01 e8 e6 6f a0 00 eb 92 <0f> 0b 66 90 0f 1f 44 00 00 55 48 89 e5 41 57 41 56 49 89 fe 41
  RIP  [<ffffffff811d9c0c>] dput+0x1dc/0x1e0
   RSP <ffff88001a3ebef8>
  ---[ end trace 1277bcfd9561ddb0 ]---

Fix it by dropping the unnecessary dget/dput() pair.

Signed-off-by: Tejun Heo <tj@kernel.org>
Acked-by: Li Zefan <lizefan@huawei.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoACPI / battery: Correct battery capacity values on Thinkpads
Kamil Iskra [Fri, 16 Nov 2012 21:28:58 +0000 (22:28 +0100)]
ACPI / battery: Correct battery capacity values on Thinkpads

commit 4000e626156935dfb626321ce09cae2c833eabbb upstream.

Add a quirk to correctly report battery capacity on 2010 and 2011
Lenovo Thinkpad models.

The affected models that I tested (x201, t410, t410s, and x220)
exhibit a problem where, when battery capacity reporting unit is mAh,
the values being reported are wrong.  Pre-2010 and 2012 models appear
to always report in mWh and are thus unaffected.  Also, in mid-2012
Lenovo issued a BIOS update for the 2011 models that fixes the issue
(tested on x220 with a post-1.29 BIOS).  No such update is available
for the 2010 models, so those still need this patch.

Problem description: for some reason, the affected Thinkpads switch
the reporting unit between mAh and mWh; generally, mAh is used when a
laptop is plugged in and mWh when it's unplugged, although a
suspend/resume or rmmod/modprobe is needed for the switch to take
effect.  The values reported in mAh are *always* wrong.  This does
not appear to be a kernel regression; I believe that the values were
never reported correctly.  I tested back to kernel 2.6.34, with
multiple machines and BIOS versions.

Simply plugging a laptop into mains before turning it on is enough to
reproduce the problem.  Here's a sample /proc/acpi/battery/BAT0/info
from Thinkpad x220 (before a BIOS update) with a 4-cell battery:

present:                 yes
design capacity:         2886 mAh
last full capacity:      2909 mAh
battery technology:      rechargeable
design voltage:          14800 mV
design capacity warning: 145 mAh
design capacity low:     13 mAh
cycle count:              0
capacity granularity 1:  1 mAh
capacity granularity 2:  1 mAh
model number:            42T4899
serial number:           21064
battery type:            LION
OEM info:                SANYO

Once the laptop switches the unit to mWh (unplug from mains, suspend,
resume), the output changes to:

present:                 yes
design capacity:         28860 mWh
last full capacity:      29090 mWh
battery technology:      rechargeable
design voltage:          14800 mV
design capacity warning: 1454 mWh
design capacity low:     200 mWh
cycle count:              0
capacity granularity 1:  1 mWh
capacity granularity 2:  1 mWh
model number:            42T4899
serial number:           21064
battery type:            LION
OEM info:                SANYO

Can you see how the values for "design capacity", etc., differ by a
factor of 10 instead of 14.8 (the design voltage of this battery)?
On the battery itself it says: 14.8V, 1.95Ah, 29Wh, so clearly the
values reported in mWh are correct and the ones in mAh are not.

My guess is that this problem has been around ever since those
machines were released, but because the most common Thinkpad
batteries are rated at 10.8V, the error (8%) is small enough that it
simply hasn't been noticed or at least nobody could be bothered to
look into it.

My patch works around the problem by adjusting the incorrectly
reported mAh values by "10000 / design_voltage".  The patch also has
code to figure out if it should be activated or not.  It only
activates on Lenovo Thinkpads, only when the unit is mAh, and, as an
extra precaution, only when the battery capacity reported through
ACPI does not match what is reported through DMI (I've never
encountered a machine where the first two conditions would be true
but the last would not, but better safe than sorry).

I've been using this patch for close to a year on several systems
without any problems.

References: https://bugzilla.kernel.org/show_bug.cgi?id=41062
Acked-by: Henrique de Moraes Holschuh <hmh@hmh.eng.br>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoftrace: Clear bits properly in reset_iter_read()
Dan Carpenter [Sat, 9 Jun 2012 16:10:27 +0000 (19:10 +0300)]
ftrace: Clear bits properly in reset_iter_read()

commit 70f77b3f7ec010ff9624c1f2e39a81babc9e2429 upstream.

There is a typo here where '&' is used instead of '|' and it turns the
statement into a noop.  The original code is equivalent to:

iter->flags &= ~((1 << 2) & (1 << 4));

Link: http://lkml.kernel.org/r/20120609161027.GD6488@elgon.mountain
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agopowerpc: Fix CONFIG_RELOCATABLE=y CONFIG_CRASH_DUMP=n build
Anton Blanchard [Sun, 11 Nov 2012 19:01:05 +0000 (19:01 +0000)]
powerpc: Fix CONFIG_RELOCATABLE=y CONFIG_CRASH_DUMP=n build

commit 11ee7e99f35ecb15f59b21da6a82d96d2cd3fcc8 upstream.

If we build a kernel with CONFIG_RELOCATABLE=y CONFIG_CRASH_DUMP=n,
the kernel fails when we run at a non zero offset. It turns out
we were incorrectly wrapping some of the relocatable kernel code
with CONFIG_CRASH_DUMP.

Signed-off-by: Anton Blanchard <anton@samba.org>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoext4: init pagevec in ext4_da_block_invalidatepages
Eric Sandeen [Thu, 15 Nov 2012 03:22:05 +0000 (22:22 -0500)]
ext4: init pagevec in ext4_da_block_invalidatepages

commit 66bea92c69477a75a5d37b9bfed5773c92a3c4b4 upstream.

ext4_da_block_invalidatepages is missing a pagevec_init(),
which means that pvec->cold contains random garbage.

This affects whether the page goes to the front or
back of the LRU when ->cold makes it to
free_hot_cold_page()

Reviewed-by: Lukas Czerner <lczerner@redhat.com>
Reviewed-by: Carlos Maiolino <cmaiolino@redhat.com>
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
[bwh: Backported to 3.2: adjust context]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agorcu: Fix batch-limit size problem
Eric Dumazet [Thu, 18 Oct 2012 11:55:36 +0000 (04:55 -0700)]
rcu: Fix batch-limit size problem

commit 878d7439d0f45a95869e417576774673d1fa243f upstream.

Commit 29c00b4a1d9e27 (rcu: Add event-tracing for RCU callback
invocation) added a regression in rcu_do_batch()

Under stress, RCU is supposed to allow to process all items in queue,
instead of a batch of 10 items (blimit), but an integer overflow makes
the effective limit being 1.  So, unless there is frequent idle periods
(during which RCU ignores batch limits), RCU can be forced into a
state where it cannot keep up with the callback-generation rate,
eventually resulting in OOM.

This commit therefore converts a few variables in rcu_do_batch() from
int to long to fix this problem, along with the module parameters
controlling the batch limits.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
[bwh: Backported to 3.2:
 - Adjust context
 - Module parameters remain hidden from sysfs]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoStaging: bcm: Add two products and remove an existing product.
Kevin McKinney [Tue, 13 Nov 2012 03:20:30 +0000 (22:20 -0500)]
Staging: bcm: Add two products and remove an existing product.

commit 4f29ef050848245f7c180b95ccf67dfcd76b1fd8 upstream.

This patch adds two new products and modifies
the device id table to include them. In addition,
product of 0xbccd - BCM_USB_PRODUCT_ID_SM250 is
removed because Beceem, ZTE, Sprint use this id
for block devices.

Reported-by: Muhammad Minhazul Haque <mdminhazulhaque@gmail.com>
Signed-off-by: Kevin McKinney <klmckinney1@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[bwh: Backported to 3.2: adjust context]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoStaging: bcm: Create and initialize new device id in InterfaceInit
Kevin McKinney [Wed, 12 Sep 2012 02:19:06 +0000 (22:19 -0400)]
Staging: bcm: Create and initialize new device id in InterfaceInit

commit e66fc1fba248738d32f3b64508f9ef1176d9e767 upstream.

This patch create and initalizes a new device
id of 0x172 as reported by Rinat Camalov
<richman1000000d@gmail.com>. In addition, a
comment is added to the potential invalid
existing device id.

Reported-by: Rinat Camalov <richman1000000d@gmail.com>
Signed-off-by: Kevin McKinney <klmckinney1@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[bwh: Backported to 3.2: adjust context]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agousb: host: xhci: Stricter conditional for Z1 system models for Compliance Mode Patch
Alexis R. Cortes [Thu, 8 Nov 2012 22:59:27 +0000 (16:59 -0600)]
usb: host: xhci: Stricter conditional for Z1 system models for Compliance Mode Patch

commit b0e4e606ff6ff26da0f60826e75577b56ba4e463 upstream.

This minor patch creates a more stricter conditional for the Z1 sytems for applying
the Compliance Mode Patch, this to avoid the quirk to be applied to models that
contain a "Z1" in their dmi product string but are different from Z1 systems.

This patch should be backported to stable kernels as old as 3.2, that
contain the commit 71c731a296f1b08a3724bd1b514b64f1bda87a23 "usb: host:
xhci: Fix Compliance Mode on SN65LVPE502CP Hardware"

Signed-off-by: Alexis R. Cortes <alexis.cortes@ti.com>
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoxhci: Extend Fresco Logic MSI quirk.
Sarah Sharp [Wed, 17 Oct 2012 20:44:06 +0000 (13:44 -0700)]
xhci: Extend Fresco Logic MSI quirk.

commit bba18e33f25072ebf70fd8f7f0cdbf8cdb59a746 upstream.

Ali reports that plugging a device into the Fresco Logic xHCI host with
PCI device ID 1400 produces an IRQ error:

 do_IRQ: 3.176 No irq handler for vector (irq -1)

Other early Fresco Logic host revisions don't support MSI, even though
their PCI config space claims they do.  Extend the quirk to disabling
MSI to this chipset revision.  Also enable the short transfer quirk,
since it's likely this revision also has that quirk, and it should be
harmless to enable.

04:00.0 0c03: 1b73:1400 (rev 01) (prog-if 30 [XHCI])
        Subsystem: 1d5c:1000
        Physical Slot: 3
        Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx+
        Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
        Latency: 0, Cache Line Size: 64 bytes
        Interrupt: pin A routed to IRQ 51
        Region 0: Memory at d4600000 (32-bit, non-prefetchable) [size=64K]
        Capabilities: [50] Power Management version 3
                Flags: PMEClk- DSI- D1- D2- AuxCurrent=0mA PME(D0+,D1-,D2-,D3hot+,D3cold-)
                Status: D0 NoSoftRst- PME-Enable- DSel=0 DScale=0 PME-
        Capabilities: [68] MSI: Enable+ Count=1/1 Maskable- 64bit+
                Address: 00000000feeff00c  Data: 41b1
        Capabilities: [80] Express (v1) Endpoint, MSI 00
                DevCap: MaxPayload 128 bytes, PhantFunc 0, Latency L0s <2us, L1 <32us
                        ExtTag- AttnBtn- AttnInd- PwrInd- RBE+ FLReset-
                DevCtl: Report errors: Correctable- Non-Fatal- Fatal- Unsupported-
                        RlxdOrd+ ExtTag- PhantFunc- AuxPwr- NoSnoop+
                        MaxPayload 128 bytes, MaxReadReq 512 bytes
                DevSta: CorrErr- UncorrErr- FatalErr- UnsuppReq- AuxPwr- TransPend-
                LnkCap: Port #0, Speed 2.5GT/s, Width x1, ASPM L0s L1, Latency L0 unlimited, L1 unlimited
                        ClockPM- Surprise- LLActRep- BwNot-
                LnkCtl: ASPM Disabled; RCB 64 bytes Disabled- Retrain- CommClk+
                        ExtSynch- ClockPM- AutWidDis- BWInt- AutBWInt-
                LnkSta: Speed 2.5GT/s, Width x1, TrErr- Train- SlotClk+ DLActive- BWMgmt- ABWMgmt-
        Kernel driver in use: xhci_hcd

This patch should be backported to stable kernels as old as 2.6.36, that
contain the commit f5182b4155b9d686c5540a6822486400e34ddd98 "xhci:
Disable MSI for some Fresco Logic hosts."

Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Reported-by: A Sh <smr.ash1991@gmail.com>
Tested-by: A Sh <smr.ash1991@gmail.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoxhci: fix null-pointer dereference when destroying half-built segment rings
Julius Werner [Thu, 1 Nov 2012 19:47:59 +0000 (12:47 -0700)]
xhci: fix null-pointer dereference when destroying half-built segment rings

commit 68e5254adb88bede68285f11fb442a4d34fb550c upstream.

xhci_alloc_segments_for_ring() builds a list of xhci_segments and links
the tail to head at the end (forming a ring). When it bails out for OOM
reasons half-way through, it tries to destroy its half-built list with
xhci_free_segments_for_ring(), even though it is not a ring yet. This
causes a null-pointer dereference upon hitting the last element.

Furthermore, one of its callers (xhci_ring_alloc()) mistakenly believes
the output parameters to be valid upon this kind of OOM failure, and
calls xhci_ring_free() on them. Since the (incomplete) list/ring should
already be destroyed in that case, this would lead to a use after free.

This patch fixes those issues by having xhci_alloc_segments_for_ring()
destroy its half-built, non-circular list manually and destroying the
invalid struct xhci_ring in xhci_ring_alloc() with a plain kfree().

This patch should be backported to kernels as old as 2.6.31, that
contains the commit 0ebbab37422315a5d0cb29792271085bafdf38c0 "USB: xhci:
Ring allocation and initialization."

A separate patch will need to be developed for kernels older than 3.4,
since the ring allocation code was refactored in that kernel.

Signed-off-by: Julius Werner <jwerner@chromium.org>
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
[bwh: Backported to 3.2:
 - Adjust context
 - Since segment allocation is done directly in xhci_ring_alloc(), walk
   the list starting from ring->first_seg when freeing]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoxHCI: Fix TD Size calculation on 1.0 hosts.
Sarah Sharp [Thu, 25 Oct 2012 22:56:40 +0000 (15:56 -0700)]
xHCI: Fix TD Size calculation on 1.0 hosts.

commit 4525c0a10dff7ad3669763c28016c7daffc3900e upstream.

The xHCI 1.0 specification made a change to the TD Size field in TRBs.
The value is now the number of packets that remain to be sent in the TD,
not including this TRB.  The TD Size value for the last TRB in a TD must
always be zero.

The xHCI function xhci_v1_0_td_remainder() attempts to calculate this,
but it gets it wrong.  First, it erroneously reuses the old
xhci_td_remainder function, which will right shift the value by 10.  The
xHCI 1.0 spec as of June 2011 says nothing about right shifting by 10.
Second, it does not set the TD size for the last TRB in a TD to zero.

Third, it uses roundup instead of DIV_ROUND_UP.  The total packet count
is supposed to be the total number of bytes in this TD, divided by the
max packet size, rounded up.  DIV_ROUND_UP is the right function to use
in that case.

With the old code, a TD on an endpoint with max packet size 1024 would
be set up like so:
TRB 1, TRB length = 600 bytes, TD size = 0
TRB 1, TRB length = 200 bytes, TD size = 0
TRB 1, TRB length = 100 bytes, TD size = 0

With the new code, the TD would be set up like this:
TRB 1, TRB length = 600 bytes, TD size = 1
TRB 1, TRB length = 200 bytes, TD size = 1
TRB 1, TRB length = 100 bytes, TD size = 0

This commit should be backported to kernels as old as 3.0, that contain
the commit 4da6e6f247a2601ab9f1e63424e4d944ed4124f3 "xhci 1.0: Update TD
size field format."

Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Reported-by: Chintan Mehta <chintan.mehta@sibridgetech.com>
Reported-by: Shimmer Huang <shimmering.h@gmail.com>
Tested-by: Bhavik Kothari <bhavik.kothari@sibridgetech.com>
Tested-by: Shimmer Huang <shimmering.h@gmail.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoxhci: Fix conditional check in bandwidth calculation.
Sarah Sharp [Thu, 25 Oct 2012 20:44:12 +0000 (13:44 -0700)]
xhci: Fix conditional check in bandwidth calculation.

commit 392a07ae3316f2b90b39ce41e66d6f6b5c95de90 upstream.

David reports that at drivers/usb/host/xhci.c:2257:

static bool xhci_is_sync_in_ep(unsigned int ep_type)
{
    return (ep_type == ISOC_IN_EP || ep_type != INT_IN_EP);
}

The static analyser cppcheck says

[linux-3.7-rc2/drivers/usb/host/xhci.c:2257]: (style) Redundant condition: If ep_type == 5, the comparison ep_type != 7 is always true.

Maybe the original programmer intention was something like

static bool xhci_is_sync_in_ep(unsigned int ep_type)
{
    return (ep_type == ISOC_IN_EP || ep_type == INT_IN_EP);
}

Fix this.

This patch should be backported to stable kernels as old as 3.2, that
contain the commit 2b69899934c63b7b9432568584fb4c4a2924f40c "xhci: USB
3.0 BW checking."

Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Reported-by: David Binderman <dcb314@hotmail.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agogenirq: Always force thread affinity
Thomas Gleixner [Sat, 3 Nov 2012 10:52:09 +0000 (11:52 +0100)]
genirq: Always force thread affinity

commit 04aa530ec04f61875b99c12721162e2964e3318c upstream.

Sankara reported that the genirq core code fails to adjust the
affinity of an interrupt thread in several cases:

 1) On request/setup_irq() the call to setup_affinity() happens before
    the new action is registered, so the new thread is not notified.

 2) For secondary shared interrupts nothing notifies the new thread to
    change its affinity.

 3) Interrupts which have the IRQ_NO_BALANCE flag set are not moving
    the thread either.

Fix this by setting the thread affinity flag right on thread creation
time. This ensures that under all circumstances the thread moves to
the right place. Requires a check in irq_thread_check_affinity for an
existing affinity mask (CONFIG_CPU_MASK_OFFSTACK=y)

Reported-and-tested-by: Sankara Muthukrishnan <sankara.m@gmail.com>
Link: http://lkml.kernel.org/r/alpine.LFD.2.02.1209041738200.2754@ionos
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoUSB: EHCI: bugfix: urb->hcpriv should not be NULL
Alan Stern [Thu, 8 Nov 2012 15:17:01 +0000 (10:17 -0500)]
USB: EHCI: bugfix: urb->hcpriv should not be NULL

commit 2656a9abcf1ec8dd5fee6a75d6997a0f2fa0094e upstream.

This patch (as1632b) fixes a bug in ehci-hcd.  The USB core uses
urb->hcpriv to determine whether or not an URB is active; host
controller drivers are supposed to set this pointer to a non-NULL
value when an URB is queued.  However ehci-hcd sets it to NULL for
isochronous URBs, which defeats the check in usbcore.

In itself this isn't a big deal.  But people have recently found that
certain sequences of actions will cause the snd-usb-audio driver to
reuse URBs without waiting for them to complete.  In the absence of
proper checking by usbcore, the URBs get added to their endpoint list
twice.  This leads to list corruption and a system freeze.

The patch makes ehci-hcd assign a meaningful value to urb->hcpriv for
isochronous URBs.  Improving robustness always helps.

Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Reported-by: Artem S. Tashkinov <t.artem@lycos.com>
Reported-by: Christof Meerwald <cmeerw@cmeerw.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[bwh: Backported to 3.2:
 - Adjust context
 - Also use usb_pipetype() to work out whether we should call qh_put()]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoUSB: fix endpoint-disabling for failed config changes
Alan Stern [Wed, 7 Nov 2012 15:31:30 +0000 (10:31 -0500)]
USB: fix endpoint-disabling for failed config changes

commit 36caff5d795429c572443894e8789c2150dd796b upstream.

This patch (as1631) fixes a bug that shows up when a config change
fails for a device under an xHCI controller.  The controller needs to
be told to disable the endpoints that have been enabled for the new
config.  The existing code does this, but before storing the
information about which endpoints were enabled!  As a result, any
second attempt to install the new config is doomed to fail because
xhci-hcd will refuse to enable an endpoint that is already enabled.

The patch optimistically initializes the new endpoints' device
structures before asking the device to switch to the new config.  If
the request fails then the endpoint information is already stored, so
we can use usb_hcd_alloc_bandwidth() to disable the endpoints with no
trouble.  The rest of the error path is slightly more complex now; we
have to disable the new interfaces and call put_device() rather than
simply deallocating them.

Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Reported-and-tested-by: Matthias Schniedermeyer <ms@citd.de>
CC: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[bwh: Backported to 3.2: adjust context]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoARM: mm: use pteval_t to represent page protection values
Will Deacon [Tue, 18 Sep 2012 18:18:35 +0000 (19:18 +0100)]
ARM: mm: use pteval_t to represent page protection values

commit 864aa04cd02979c2c755cb28b5f4fe56039171c0 upstream.

When updating the page protection map after calculating the user_pgprot
value, the base protection map is temporarily stored in an unsigned long
type, causing truncation of the protection bits when LPAE is enabled.
This effectively means that calls to mprotect() will corrupt the upper
page attributes, clearing the XN bit unconditionally.

This patch uses pteval_t to store the intermediate protection values,
preserving the upper bits for 64-bit descriptors.

Acked-by: Nicolas Pitre <nico@linaro.org>
Acked-by: Catalin Marinas <catalin.marinas@arm.com>
Signed-off-by: Will Deacon <will.deacon@arm.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoext4: fix memory leak in ext4_xattr_set_acl()'s error path
Eugene Shatokhin [Thu, 8 Nov 2012 20:11:11 +0000 (15:11 -0500)]
ext4: fix memory leak in ext4_xattr_set_acl()'s error path

commit 24ec19b0ae83a385ad9c55520716da671274b96c upstream.

In ext4_xattr_set_acl(), if ext4_journal_start() returns an error,
posix_acl_release() will not be called for 'acl' which may result in a
memory leak.

This patch fixes that.

Reviewed-by: Lukas Czerner <lczerner@redhat.com>
Signed-off-by: Eugene Shatokhin <eugene.shatokhin@rosalab.ru>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoiscsi-target: Always send a response before terminating iSCSI connection
Roland Dreier [Tue, 6 Nov 2012 02:02:42 +0000 (18:02 -0800)]
iscsi-target: Always send a response before terminating iSCSI connection

commit 1c5c12c666fda27c7c494b34934a0a0631a48130 upstream.

There are some cases, for example when the initiator sends an
out-of-bounds ErrorRecoveryLevel value, where the iSCSI target
terminates the connection without sending back any error.  Audit the
login path and add appropriate iscsit_tx_login_rsp() calls to make
sure this doesn't happen.

Signed-off-by: Roland Dreier <roland@purestorage.com>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
[bwh: Backported to 3.2: adjust context]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agonfs: fix wrong object type in lockowner_slab
Yanchuan Nian [Wed, 24 Oct 2012 06:44:19 +0000 (14:44 +0800)]
nfs: fix wrong object type in lockowner_slab

commit 3c40794b2dd0f355ef4e6bf8d85af5dcd7da7ece upstream.

The object type in the cache of lockowner_slab is wrong, and it is
better to fix it.

Signed-off-by: Yanchuan Nian <ycnian@gmail.com>
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agousb: musb: cppi_dma: export cppi_interrupt()
Sergei Shtylyov [Mon, 5 Nov 2012 19:26:40 +0000 (22:26 +0300)]
usb: musb: cppi_dma: export cppi_interrupt()

commit 8b416b0b25d5d8ddb3a91c1d20e1373582c50405 upstream.

Now that DaVinci glue layer can be modular, we must export cppi_interrupt()
that it may call...

Signed-off-by: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Signed-off-by: Felipe Balbi <balbi@ti.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agox86: hpet: Fix masking of MSI interrupts
Jan Beulich [Fri, 2 Nov 2012 14:02:40 +0000 (14:02 +0000)]
x86: hpet: Fix masking of MSI interrupts

commit 6acf5a8c931da9d26c8dd77d784daaf07fa2bff0 upstream.

HPET_TN_FSB is not a proper mask bit; it merely toggles between MSI and
legacy interrupt delivery. The proper mask bit is HPET_TN_ENABLE, so
use both bits when (un)masking the interrupt.

Signed-off-by: Jan Beulich <jbeulich@suse.com>
Link: http://lkml.kernel.org/r/5093E09002000078000A60E6@nat28.tlf.novell.com
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agousb: gadget: uvc: fix error path in uvc_function_bind()
Sebastian Andrzej Siewior [Mon, 22 Oct 2012 20:15:05 +0000 (22:15 +0200)]
usb: gadget: uvc: fix error path in uvc_function_bind()

commit 0f9df939385527049c8062a099fbfa1479fe7ce0 upstream.

The "video->minor = -1" assigment is done in V4L2 by
video_register_device() so it is removed here.
Now. uvc_function_bind() calls in error case uvc_function_unbind() for
cleanup. The problem is that uvc_function_unbind() frees the uvc struct
and uvc_bind_config() does as well in error case of usb_add_function().
Removing kfree() in usb_add_function() would make the patch smaller but
it would look odd because the new allocated memory is not cleaned up.
However it is not guaranteed that if we call usb_add_function() we also
get to the bind function.
Therefore the patch extracts the conditional cleanup from
uvc_function_unbind() applies to uvc_function_bind().
uvc_function_unbind() now contains only the complete cleanup which is
required once everything has been registrated.

Cc: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Cc: Bhupesh Sharma <bhupesh.sharma@st.com>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Felipe Balbi <balbi@ti.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agousb: gadget: phonet: free requests in pn_bind()'s error path
Sebastian Andrzej Siewior [Mon, 22 Oct 2012 20:15:04 +0000 (22:15 +0200)]
usb: gadget: phonet: free requests in pn_bind()'s error path

commit d0eca719dd11ad0619e8dd6a1f3eceb95b0216dd upstream.

Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Felipe Balbi <balbi@ti.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agousb: gadget: midi: free hs descriptors
Sebastian Andrzej Siewior [Mon, 22 Oct 2012 20:15:02 +0000 (22:15 +0200)]
usb: gadget: midi: free hs descriptors

commit d185039f7982eb82cf8d03b6fb6689587ca5af24 upstream.

The HS descriptors are only created if HS is supported by the UDC but we
never free them.

Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Felipe Balbi <balbi@ti.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agousb: gadget: network: fix bind() error path
Sebastian Andrzej Siewior [Mon, 22 Oct 2012 20:15:00 +0000 (22:15 +0200)]
usb: gadget: network: fix bind() error path

commit e79cc615a9bb44da72c499ccfa2c9c4bbea3aa84 upstream.

I think this is wrong since 72c973dd ("usb: gadget: add
usb_endpoint_descriptor to struct usb_ep"). If we fail to allocate an ep
or bail out early we shouldn't check for the descriptor which is
assigned at ep_enable() time.

Cc: Tatyana Brokhman <tlinder@codeaurora.org>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Felipe Balbi <balbi@ti.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoKVM: PPC: 44x: fix DCR read/write
Alexander Graf [Sat, 6 Oct 2012 01:56:35 +0000 (03:56 +0200)]
KVM: PPC: 44x: fix DCR read/write

commit e43a028752fed049e4bd94ef895542f96d79fa74 upstream.

When remembering the direction of a DCR transaction, we should write
to the same variable that we interpret on later when doing vcpu_run
again.

Signed-off-by: Alexander Graf <agraf@suse.de>
[bwh: Backported to 3.2: adjust context, indentation]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agoath9k_hw: Enable hw PLL power save for AR9462
Rajkumar Manoharan [Thu, 25 Oct 2012 11:41:31 +0000 (17:11 +0530)]
ath9k_hw: Enable hw PLL power save for AR9462

commit 1680260226a8fd2aab590319da83ad8e610da9bd upstream.

This reduced the power consumption to half in full and network sleep.

Cc: Paul Stewart <pstew@chromium.org>
Signed-off-by: Rajkumar Manoharan <rmanohar@qca.qualcomm.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
[bwh: Backported to 3.2:
 - INIT_INI_ARRAY macro requires an explicit size argument
 - Remove the now-redundant macro PCIE_PLL_ON_CREQ_DIS_L1_2P0]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agovirtio: force vring descriptors to be allocated from lowmem
Will Deacon [Fri, 19 Oct 2012 13:03:33 +0000 (14:03 +0100)]
virtio: force vring descriptors to be allocated from lowmem

commit b92b1b89a33c172c075edccf6afb0edc41d851fd upstream.

Virtio devices may attempt to add descriptors to a virtqueue from atomic
context using GFP_ATOMIC allocation. This is problematic because such
allocations can fall outside of the lowmem mapping, causing virt_to_phys
to report bogus physical addresses which are subsequently passed to
userspace via the buffers for the virtual device.

This patch masks out __GFP_HIGH and __GFP_HIGHMEM from the requested
flags when allocating descriptors for a virtqueue. If an atomic
allocation is requested and later fails, we will return -ENOSPC which
will be handled by the driver.

Cc: Sasha Levin <levinsasha928@gmail.com>
Signed-off-by: Will Deacon <will.deacon@arm.com>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agovirtio: 9p: correctly pass physical address to userspace for high pages
Will Deacon [Fri, 19 Oct 2012 13:03:32 +0000 (14:03 +0100)]
virtio: 9p: correctly pass physical address to userspace for high pages

commit b9cdc88df8e63e81c723b82c286fc97f5d0dc325 upstream.

When using a virtio transport, the 9p net device may pass the physical
address of a kernel buffer to userspace via a scatterlist inside a
virtqueue. If the kernel buffer is mapped outside of the linear mapping
(e.g. highmem), then virt_to_page will return a bogus value and we will
populate the scatterlist with junk.

This patch uses kmap_to_page when populating the page array for a kernel
buffer.

Cc: Sasha Levin <levinsasha928@gmail.com>
Signed-off-by: Will Deacon <will.deacon@arm.com>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agomm: highmem: export kmap_to_page for modules
Will Deacon [Fri, 19 Oct 2012 13:03:31 +0000 (14:03 +0100)]
mm: highmem: export kmap_to_page for modules

commit f0263d2d222e9e25f2587e51a9dc58c6fb2a9352 upstream.

Some virtio device drivers (9p) need to translate high virtual addresses
to physical addresses, which are inserted into the virtqueue for
processing by userspace.

This patch exports the kmap_to_page symbol, so that the affected drivers
can be compiled as modules.

Signed-off-by: Will Deacon <will.deacon@arm.com>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agomm: add kmap_to_page()
Ben Hutchings [Tue, 31 Jul 2012 23:45:02 +0000 (16:45 -0700)]
mm: add kmap_to_page()

This is extracted from Mel Gorman's commit 5a178119b0fb ('mm: add
support for direct_IO to highmem pages') upstream.

Required to backport commit b9cdc88df8e6 ('virtio: 9p: correctly pass
physical address to userspace for high pages').

Cc: Mel Gorman <mgorman@suse.de>
Cc: Rik van Riel <riel@redhat.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agofreezer: add missing mb's to freezer_count() and freezer_should_skip()
Tejun Heo [Tue, 16 Oct 2012 22:03:14 +0000 (15:03 -0700)]
freezer: add missing mb's to freezer_count() and freezer_should_skip()

commit dd67d32dbc5de299d70cc9e10c6c1e29ffa56b92 upstream.

A task is considered frozen enough between freezer_do_not_count() and
freezer_count() and freezers use freezer_should_skip() to test this
condition.  This supposedly works because freezer_count() always calls
try_to_freezer() after clearing %PF_FREEZER_SKIP.

However, there currently is nothing which guarantees that
freezer_count() sees %true freezing() after clearing %PF_FREEZER_SKIP
when freezing is in progress, and vice-versa.  A task can escape the
freezing condition in effect by freezer_count() seeing !freezing() and
freezer_should_skip() seeing %PF_FREEZER_SKIP.

This patch adds smp_mb()'s to freezer_count() and
freezer_should_skip() such that either %true freezing() is visible to
freezer_count() or !PF_FREEZER_SKIP is visible to
freezer_should_skip().

Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Rafael J. Wysocki <rjw@sisk.pl>
[bwh: Backported to 3.2:
 - Adjust context and indentation
 - freezer_do_not_count() and freezer_count() are no-ops for kernel tasks]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agocgroup: cgroup_subsys->fork() should be called after the task is added to css_set
Tejun Heo [Tue, 16 Oct 2012 22:03:14 +0000 (15:03 -0700)]
cgroup: cgroup_subsys->fork() should be called after the task is added to css_set

commit 5edee61edeaaebafe584f8fb7074c1ef4658596b upstream.

cgroup core has a bug which violates a basic rule about event
notifications - when a new entity needs to be added, you add that to
the notification list first and then make the new entity conform to
the current state.  If done in the reverse order, an event happening
inbetween will be lost.

cgroup_subsys->fork() is invoked way before the new task is added to
the css_set.  Currently, cgroup_freezer is the only user of ->fork()
and uses it to make new tasks conform to the current state of the
freezer.  If FROZEN state is requested while fork is in progress
between cgroup_fork_callbacks() and cgroup_post_fork(), the child
could escape freezing - the cgroup isn't frozen when ->fork() is
called and the freezer couldn't see the new task on the css_set.

This patch moves cgroup_subsys->fork() invocation to
cgroup_post_fork() after the new task is added to the css_set.
cgroup_fork_callbacks() is removed.

Because now a task may be migrated during cgroup_subsys->fork(),
freezer_fork() is updated so that it adheres to the usual RCU locking
and the rather pointless comment on why locking can be different there
is removed (if it doesn't make anything simpler, why even bother?).

Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Rafael J. Wysocki <rjw@sisk.pl>
[bwh: Backported to 3.2:
 - Adjust context
 - Iterate over first CGROUP_BUILTIN_SUBSYS_COUNT elements of subsys
 - cgroup_subsys::fork takes cgroup_subsys pointer as first parameter]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
9 years agos390/kvm: dont announce RRBM support
Christian Borntraeger [Tue, 2 Oct 2012 14:25:38 +0000 (16:25 +0200)]
s390/kvm: dont announce RRBM support

commit 87cac8f879a5ecd7109dbe688087e8810b3364eb upstream.

Newer kernels (linux-next with the transparent huge page patches)
use rrbm if the feature is announced via feature bit 66.
RRBM will cause intercepts, so KVM does not handle it right now,
causing an illegal instruction in the guest.
The  easy solution is to disable the feature bit for the guest.

This fixes bugs like:
Kernel BUG at 0000000000124c2a [verbose debug info unavailable]
illegal operation: 0001 [#1] SMP
Modules linked in: virtio_balloon virtio_net ipv6 autofs4
CPU: 0 Not tainted 3.5.4 #1
Process fmempig (pid: 659, task: 000000007b712fd0, ksp: 000000007bed3670)
Krnl PSW : 0704d00180000000 0000000000124c2a (pmdp_clear_flush_young+0x5e/0x80)
     R:0 T:1 IO:1 EX:1 Key:0 M:1 W:0 P:0 AS:3 CC:1 PM:0 EA:3
     00000000003cc000 0000000000000004 0000000000000000 0000000079800000
     0000000000040000 0000000000000000 000000007bed3918 000000007cf40000
     0000000000000001 000003fff7f00000 000003d281a94000 000000007bed383c
     000000007bed3918 00000000005ecbf8 00000000002314a6 000000007bed36e0
 Krnl Code:>0000000000124c2ab9810025          ogr     %r2,%r5
           0000000000124c2e41343000           la      %r3,0(%r4,%r3)
           0000000000124c32a716fffa           brct    %r1,124c26
           0000000000124c36b9010022           lngr    %r2,%r2
           0000000000124c3ae3d0f0800004       lg      %r13,128(%r15)
           0000000000124c40eb22003f000c       srlg    %r2,%r2,63
[ 2150.713198] Call Trace:
[ 2150.713223] ([<00000000002312c4>] page_referenced_one+0x6c/0x27c)
[ 2150.713749]  [<0000000000233812>] page_referenced+0x32a/0x410
[...]

CC: Alex Graf <agraf@suse.de>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>