7 years agoclockevents: tTack broadcast device mode change in tick_broadcast_switch_to_oneshot()
Suresh Siddha [Mon, 9 Apr 2012 22:41:44 +0000]
clockevents: tTack broadcast device mode change in tick_broadcast_switch_to_oneshot()

In the commit 77b0d60c5adf39c74039e2142a1d3cd1e4d53799,
"clockevents: Leave the broadcast device in shutdown mode when not needed",
we were bailing out too quickly in tick_broadcast_switch_to_oneshot(),
with out tracking the broadcast device mode change to 'TICKDEV_MODE_ONESHOT'.

This breaks the platforms which need broadcast device oneshot services during
deep idle states. tick_broadcast_oneshot_control() thinks that it is
in periodic mode and fails to take proper decisions based on the
CLOCK_EVT_NOTIFY_BROADCAST_[ENTER, EXIT] notifications during deep
idle entry/exit.

Fix this by tracking the broadcast device mode as 'TICKDEV_MODE_ONESHOT',
before leaving the broadcast HW device in shutdown mode if there are no active
requests for the moment.

Reported-and-tested-by: Santosh Shilimkar <santosh.shilimkar@ti.com>
Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com>
Cc: johnstul@us.ibm.com
Link: http://lkml.kernel.org/r/1334011304.12400.81.camel@sbsiddha-desk.sc.intel.com
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>

7 years agoitimer: Use printk_once instead of WARN_ONCE
Thomas Gleixner [Tue, 10 Apr 2012 08:50:55 +0000]
itimer: Use printk_once instead of WARN_ONCE

David pointed out, that WARN_ONCE() to report usage of an deprecated
misfeature make folks unhappy. Use printk_once() instead.

Andrew told me to stop grumbling and to remove the silly typecast
while touching the file.

Reported-by: David Rientjes <rientjes@google.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>

7 years agonohz: Fix stale jiffies update in tick_nohz_restart()
Neal Cardwell [Tue, 27 Mar 2012 19:09:37 +0000]
nohz: Fix stale jiffies update in tick_nohz_restart()

Fix tick_nohz_restart() to not use a stale ktime_t "now" value when
calling tick_do_update_jiffies64(now).

If we reach this point in the loop it means that we crossed a tick
boundary since we grabbed the "now" timestamp, so at this point "now"
refers to a time in the old jiffy, so using the old value for "now" is
incorrect, and is likely to give us a stale jiffies value.

In particular, the first time through the loop the
tick_do_update_jiffies64(now) call is always a no-op, since the
caller, tick_nohz_restart_sched_tick(), will have already called
tick_do_update_jiffies64(now) with that "now" value.

Note that tick_nohz_stop_sched_tick() already uses the correct
approach: when we notice we cross a jiffy boundary, grab a new
timestamp with ktime_get(), and *then* update jiffies.

Signed-off-by: Neal Cardwell <ncardwell@google.com>
Cc: Ben Segall <bsegall@google.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: stable@vger.kernel.org
Link: http://lkml.kernel.org/r/1332875377-23014-1-git-send-email-ncardwell@google.com
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>

7 years agotick: Document TICK_ONESHOT config option
Thomas Gleixner [Sat, 31 Mar 2012 10:45:43 +0000]
tick: Document TICK_ONESHOT config option

This option has been selected from arch code as it was assumed that
it's necessary to support oneshot mode clockevent devices. But it's
just a core internal helper to compile tick-oneshot.c if NOHZ or
HIG_RES_TIMERS are selected.

Reported-by: Russell King <rmk+kernel@arm.linux.org.uk>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>

7 years agoproc: stats: Use arch_idle_time for idle and iowait times if available
Martin Schwidefsky [Fri, 30 Mar 2012 10:23:08 +0000]
proc: stats: Use arch_idle_time for idle and iowait times if available

Git commit a25cac5198d4ff28 "proc: Consider NO_HZ when printing idle and
iowait times" changes the code for /proc/stat to use get_cpu_idle_time_us
and get_cpu_iowait_time_us if the system is running with nohz enabled.
For architectures which define arch_idle_time (currently s390 only)
this is a change for the worse. The result of arch_idle_time is supposed
to be the exact sleep time of the target cpu and should be used instead
of the value kept by the scheduler.

Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Reviewed-by: Michal Hocko <mhocko@suse.cz>
Reviewed-by: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com>
Link: http://lkml.kernel.org/r/20120330122308.18720283@de.ibm.com
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>

7 years agoitimer: Schedule silent NULL pointer fixup in setitimer() for removal
Sasikantha babu [Wed, 18 Apr 2012 14:33:40 +0000]
itimer: Schedule silent NULL pointer fixup in setitimer() for removal

setitimer() should return -EFAULT if called with an invalid pointer
for value. The current code excludes a NULL pointer from this rule and
silently uses it to stop the timer. This violates the spec.

Warn about user space apps which rely on that feature and schedule it
for removal.

[ tglx: Massaged changelog, warn message and Doc entry ]

Signed-off-by: Sasikantha babu <sasikanth.v19@gmail.com>
Link: http://lkml.kernel.org/r/1332340854-26053-1-git-send-email-sasikanth.v19@gmail.com
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>

Conflicts:

Documentation/feature-removal-schedule.txt

Change-Id: Id0920441c09d32cf8375e58afb955add51325df9

7 years agox86: Use correct byte-sized register constraint in __add()
H. Peter Anvin [Fri, 6 Apr 2012 16:30:57 +0000]
x86: Use correct byte-sized register constraint in __add()

Similar to:

 2ca052a x86: Use correct byte-sized register constraint in __xchg_op()

... the __add() macro also needs to use a "q" constraint in the
byte-sized case, lest we try to generate an illegal register.

Link: http://lkml.kernel.org/r/4F7A3315.501@goop.org
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Cc: Jeremy Fitzhardinge <jeremy@goop.org>
Cc: Leigh Scott <leigh123linux@googlemail.com>
Cc: Thomas Reitmayr <treitmayr@devbase.at>
Cc: <stable@vger.kernel.org> v3.3

7 years agox86: Use correct byte-sized register constraint in __xchg_op()
Jeremy Fitzhardinge [Mon, 2 Apr 2012 23:15:33 +0000]
x86: Use correct byte-sized register constraint in __xchg_op()

x86-64 can access the low half of any register, but i386 can only do
it with a subset of registers.  'r' causes compilation failures on i386,
but 'q' expresses the constraint properly.

Signed-off-by: Jeremy Fitzhardinge <jeremy@goop.org>
Link: http://lkml.kernel.org/r/4F7A3315.501@goop.org
Reported-by: Leigh Scott <leigh123linux@googlemail.com>
Tested-by: Thomas Reitmayr <treitmayr@devbase.at>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Cc: <stable@vger.kernel.org> v3.3

7 years agox86: vsyscall: Use NULL instead 0 for a pointer argument
Emil Goode [Sun, 1 Apr 2012 18:48:04 +0000]
x86: vsyscall: Use NULL instead 0 for a pointer argument

This patch silences the following sparse warning:
arch/x86/kernel/vsyscall_64.c:250:34:
       warning: Using plain integer as NULL pointer

Signed-off-by: Emil Goode <emilgoode@gmail.com>
Acked-by: Andy Lutomirski <luto@amacapital.net>
Cc: john.stultz@linaro.org
Link: http://lkml.kernel.org/r/1333306084-3776-1-git-send-email-emilgoode@gmail.com
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>

7 years agopanic: fix stack dump print on direct call to panic()
Jason Wessel [Thu, 12 Apr 2012 19:49:17 +0000]
panic: fix stack dump print on direct call to panic()

Commit 6e6f0a1f0fa6 ("panic: don't print redundant backtraces on oops")
causes a regression where no stack trace will be printed at all for the
case where kernel code calls panic() directly while not processing an
oops, and of course there are 100's of instances of this type of call.

The original commit executed the check (!oops_in_progress), but this will
always be false because just before the dump_stack() there is a call to
bust_spinlocks(1), which does the following:

  void __attribute__((weak)) bust_spinlocks(int yes)
  {
if (yes) {
++oops_in_progress;

The proper way to resolve the problem that original commit tried to
solve is to avoid printing a stack dump from panic() when the either of
the following conditions is true:

  1) TAINT_DIE has been set (this is done by oops_end())
     This indicates and oops has already been printed.
  2) oops_in_progress > 1
     This guards against the rare case where panic() is invoked
     a second time, or in between oops_begin() and oops_end()

Signed-off-by: Jason Wessel <jason.wessel@windriver.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: <stable@vger.kernel.org> [3.3+]
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

7 years agodrivers/rtc/rtc-pl031.c: enable clock on all ST variants
Linus Walleij [Thu, 12 Apr 2012 19:49:16 +0000]
drivers/rtc/rtc-pl031.c: enable clock on all ST variants

The ST variants of the PL031 all require bit 26 in the control register
to be set before they work properly.  Discovered this when testing on
the Nomadik board where it would suprisingly just stand still.

Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Cc: Mian Yousaf Kaukab <mian.yousaf.kaukab@stericsson.com>
Cc: Alessandro Rubini <rubini@unipv.it>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

7 years agoRevert "mm: vmscan: fix misused nr_reclaimed in shrink_mem_cgroup_zone()"
Ying Han [Thu, 12 Apr 2012 19:49:16 +0000]
Revert "mm: vmscan: fix misused nr_reclaimed in shrink_mem_cgroup_zone()"

This reverts commit c38446cc65e1f2b3eb8630c53943b94c4f65f670.

Before the commit, the code makes senses to me but not after the commit.
The "nr_reclaimed" is the number of pages reclaimed by scanning through
the memcg's lru lists.  The "nr_to_reclaim" is the target value for the
whole function.  For example, we like to early break the reclaim if
reclaimed 32 pages under direct reclaim (not DEF_PRIORITY).

After the reverted commit, the target "nr_to_reclaim" is decremented each
time by "nr_reclaimed" but we still use it to compare the "nr_reclaimed".
It just doesn't make sense to me...

Signed-off-by: Ying Han <yinghan@google.com>
Acked-by: Hugh Dickins <hughd@google.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Hillf Danton <dhillf@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

7 years agohugetlb: fix race condition in hugetlb_fault()
Chris Metcalf [Thu, 12 Apr 2012 19:49:15 +0000]
hugetlb: fix race condition in hugetlb_fault()

The race is as follows:

Suppose a multi-threaded task forks a new process (on cpu A), thus
bumping up the ref count on all the pages.  While the fork is occurring
(and thus we have marked all the PTEs as read-only), another thread in
the original process (on cpu B) tries to write to a huge page, taking an
access violation from the write-protect and calling hugetlb_cow().  Now,
suppose the fork() fails.  It will undo the COW and decrement the ref
count on the pages, so the ref count on the huge page drops back to 1.
Meanwhile hugetlb_cow() also decrements the ref count by one on the
original page, since the original address space doesn't need it any
more, having copied a new page to replace the original page.  This
leaves the ref count at zero, and when we call unlock_page(), we panic.

fork on CPU A fault on CPU B
============= ==============
...
down_write(&parent->mmap_sem);
down_write_nested(&child->mmap_sem);
...
while duplicating vmas
if error
break;
...
up_write(&child->mmap_sem);
up_write(&parent->mmap_sem); ...
down_read(&parent->mmap_sem);
...
lock_page(page);
handle COW
page_mapcount(old_page) == 2
alloc and prepare new_page
...
handle error
page_remove_rmap(page);
put_page(page);
...
fold new_page into pte
page_remove_rmap(page);
put_page(page);
...
oops ==> unlock_page(page);
up_read(&parent->mmap_sem);

The solution is to take an extra reference to the page while we are
holding the lock on it.

Signed-off-by: Chris Metcalf <cmetcalf@tilera.com>
Cc: Hillf Danton <dhillf@gmail.com>
Cc: Michal Hocko <mhocko@suse.cz>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

7 years agodrivers/rtc/rtc-twl.c: use static register while reading time
Konstantin Shlyakhovoy [Thu, 12 Apr 2012 19:49:15 +0000]
drivers/rtc/rtc-twl.c: use static register while reading time

RTC stores time and date in several registers.  Due to the fact that
these registers can't be read instantaneously, there is a chance that
reading from counting registers gives an error of one minute, one hour,
one day, etc.

To address this issue, the RTC has hardware support to copy the RTC
counting registers to static shadowed registers.  The current
implementation does not use this feature, and in a stress test, we can
reproduce this error at a rate of around two times per 300000 readings.

Fix the implementation to ensure that the right snapshot of time is
captured.

Signed-off-by: Konstantin Shlyakhovoy <x0155534@ti.com>
Signed-off-by: Nishanth Menon <nm@ti.com>
Cc: Alessandro Zummo <a.zummo@towertech.it>
Cc: Benoit Cousson <b-cousson@ti.com>
Cc: linux-omap <linux-omap@vger.kernel.org>
Acked-by: Mykola Oleksiienko <x0174904@ti.com>
Acked-by: Oleksandr Dmytryshyn <oleksandr.dmytryshyn@ti.com>
Acked-by: Graeme Gregory <gg@slimlogic.co.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

7 years agodrivers/rtc/rtc-s3c.c: add placeholder for driver private data
Tushar Behera [Thu, 12 Apr 2012 19:49:14 +0000]
drivers/rtc/rtc-s3c.c: add placeholder for driver private data

Driver data field is a pointer, hence assigning that to an integer results
in compilation warnings.

Fixes following compilation warnings:

  drivers/rtc/rtc-s3c.c: In function `s3c_rtc_get_driver_data':
  drivers/rtc/rtc-s3c.c:452:3: warning: return makes integer from pointer without a cast [enabled by default]
  drivers/rtc/rtc-s3c.c: At top level:
  drivers/rtc/rtc-s3c.c:674:3: warning: initialization makes pointer from integer without a cast [enabled by default]
  drivers/rtc/rtc-s3c.c:674:3: warning: (near initialization for `s3c_rtc_dt_match[1].data') [enabled by default]
  drivers/rtc/rtc-s3c.c:677:3: warning: initialization makes pointer from integer without a cast [enabled by default]
  drivers/rtc/rtc-s3c.c:677:3: warning: (near initialization for `s3c_rtc_dt_match[2].data') [enabled by default]
  drivers/rtc/rtc-s3c.c:680:3: warning: initialization makes pointer from integer without a cast [enabled by default]
  drivers/rtc/rtc-s3c.c:680:3: warning: (near initialization for `s3c_rtc_dt_match[3].data') [enabled by default]

Signed-off-by: Tushar Behera <tushar.behera@linaro.org>
Cc: Heiko Stuebner <heiko@sntech.de>
Cc: Alessandro Zummo <a.zummo@towertech.it>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

7 years agodrivers/rtc/rtc-s3c.c: fix compilation error
Tushar Behera [Thu, 12 Apr 2012 19:49:14 +0000]
drivers/rtc/rtc-s3c.c: fix compilation error

Fix this error:

  drivers/rtc/rtc-s3c.c: At top level:
  drivers/rtc/rtc-s3c.c:671:3: error: request for member `data' in something not a structure or union
  drivers/rtc/rtc-s3c.c:674:3: error: request for member `data' in something not a structure or union
  drivers/rtc/rtc-s3c.c:677:3: error: request for member `data' in something not a structure or union
  drivers/rtc/rtc-s3c.c:680:3: error: request for member `data' in something not a structure or union

Signed-off-by: Tushar Behera <tushar.behera@linaro.org>
Cc: Heiko Stuebner <heiko@sntech.de>
Cc: Alessandro Zummo <a.zummo@towertech.it>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

7 years agoMAINTAINERS: add PCDP console maintainer
Khalid Aziz [Thu, 12 Apr 2012 19:49:13 +0000]
MAINTAINERS: add PCDP console maintainer

Add missing maintainer info for PCDP console code.

Signed-off-by: Khalid Aziz <khalid.aziz@hp.com>
Cc: Joe Perches <joe@perches.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

7 years agomemcg: do not open code accesses to res_counter members
Glauber Costa [Thu, 12 Apr 2012 19:49:13 +0000]
memcg: do not open code accesses to res_counter members

We should use the accessor res_counter_read_u64 for that.

Although a purely cosmetic change is sometimes better delayed, to avoid
conflicting with other people's work, we are starting to have people
touching this code as well, and reproducing the open code behavior
because that's the standard =)

Time to fix it, then.

Signed-off-by: Glauber Costa <glommer@parallels.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: Michal Hocko <mhocko@suse.cz>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

7 years agodrivers/rtc/rtc-efi.c: fix section mismatch warning
Jan Beulich [Thu, 12 Apr 2012 19:49:12 +0000]
drivers/rtc/rtc-efi.c: fix section mismatch warning

efi_rtc_init() uses platform_driver_probe(), so there's no need to also
set efi_rtc_driver's probe member (as it won't be used anyway).  This
fixes a modpost section mismatch warning (as efi_rtc_probe() validly is
__init).

Signed-off-by: Jan Beulich <jbeulich@suse.com>
Cc: Matthew Garrett <mjg@redhat.com>
Cc: Alessandro Zummo <a.zummo@towertech.it>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

7 years agodrivers/rtc/rtc-r9701.c: reset registers if invalid values are detected
Andreas Dumberger [Thu, 12 Apr 2012 19:49:12 +0000]
drivers/rtc/rtc-r9701.c: reset registers if invalid values are detected

hwclock refuses to set date/time if RTC registers contain invalid
values.  Check the date/time register values at probe time and
initialize them to make hwclock happy.

Signed-off-by: Andreas Dumberger <andreas.dumberger@tqs.de>
Signed-off-by: Anatolij Gustschin <agust@denx.de>
Cc: Alessandro Zummo <a.zummo@towertech.it>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

7 years agodrivers/char/random.c: fix boot id uniqueness race
Mathieu Desnoyers [Thu, 12 Apr 2012 19:49:12 +0000]
drivers/char/random.c: fix boot id uniqueness race

/proc/sys/kernel/random/boot_id can be read concurrently by userspace
processes.  If two (or more) user-space processes concurrently read
boot_id when sysctl_bootid is not yet assigned, a race can occur making
boot_id differ between the reads.  Because the whole point of the boot id
is to be unique across a kernel execution, fix this by protecting this
operation with a spinlock.

Given that this operation is not frequently used, hitting the spinlock
on each call should not be an issue.

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Cc: "Theodore Ts'o" <tytso@mit.edu>
Cc: Matt Mackall <mpm@selenic.com>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Greg Kroah-Hartman <greg@kroah.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

7 years agomemcg: fix broken boolen expression
Kirill A. Shutemov [Thu, 12 Apr 2012 19:49:11 +0000]
memcg: fix broken boolen expression

action != CPU_DEAD || action != CPU_DEAD_FROZEN is always true.

Signed-off-by: Kirill A. Shutemov <kirill@shutemov.name>
Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Acked-by: Michal Hocko <mhocko@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

7 years agomemcg: fix up documentation on global LRU
Ying Han [Thu, 12 Apr 2012 19:49:10 +0000]
memcg: fix up documentation on global LRU

In v3.3-rc1, the global LRU was removed in commit 925b7673cce3 ("mm:
make per-memcg LRU lists exclusive").  The patch fixes up the memcg
docs.

I left the swap session to someone who has better understanding of
'memory+swap'.

Signed-off-by: Ying Han <yinghan@google.com>
Acked-by: Michal Hocko <mhocko@suse.cz>
Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

7 years agoNFC: Fix the LLCP Tx fragmentation loop
Samuel Ortiz [Tue, 10 Apr 2012 17:43:03 +0000]
NFC: Fix the LLCP Tx fragmentation loop

Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

7 years agortlwifi: Add missing DMA buffer unmapping for PCI drivers
Larry Finger [Mon, 26 Mar 2012 15:48:20 +0000]
rtlwifi: Add missing DMA buffer unmapping for PCI drivers

In https://bugzilla.kernel.org/show_bug.cgi?id=42976, a system with driver
rtl8192se used as an AP suffers from "Out of SW-IOMMU space" errors. These
are caused by the DMA buffers used for beacons never being unmapped.

This bug was also reported at
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/961618

Reported-and-Tested-by: Da Xue <da@lessconfused.com>
Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net>
Cc: Stable <stable@vger.kernel.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

7 years agortlwifi: Preallocate USB read buffers and eliminate kalloc in read routine
Larry Finger [Mon, 19 Mar 2012 20:44:31 +0000]
rtlwifi: Preallocate USB read buffers and eliminate kalloc in read routine

The current version of rtlwifi for USB operations uses kmalloc to
acquire a 32-bit buffer for each read of the device. When
_usb_read_sync() is called with the rcu_lock held, the result is
a "sleeping function called from invalid context" BUG. This is
reported for two cases in https://bugzilla.kernel.org/show_bug.cgi?id=42775.
The first case has the lock originating from within rtlwifi and could
be fixed by rearranging the locking; however, the second originates from
within mac80211. The kmalloc() call is removed from _usb_read_sync()
by creating a ring buffer pointer in the private area and
allocating the buffer data in the probe routine.

Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net>
Cc: Stable <stable@vger.kernel.org> [This version good for 3.3+ - different patch for 3.2 - 2.6.39]
Signed-off-by: John W. Linville <linville@tuxdriver.com>

7 years agoRevert "ath9k: fix going to full-sleep on PS idle"
Sujith Manoharan [Tue, 10 Apr 2012 06:56:11 +0000]
Revert "ath9k: fix going to full-sleep on PS idle"

This reverts commit c1afdaff90538ef085b756454f12b29575411214.

Users have reported connection failures in 3.3.1 and suspend/resume
failures in 3.4-rcX. Revert this commit for now - PS IDLE can be
fixed in a clean manner later on.

Cc: stable@vger.kernel.org
Signed-off-by: Sujith Manoharan <c_manoha@qca.qualcomm.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

7 years agort2x00: Fix rfkill_polling register function.
Chen, Chien-Chia [Thu, 29 Mar 2012 10:21:47 +0000]
rt2x00: Fix rfkill_polling register function.

  Move rt2x00rfkill_register(rt2x00dev) to rt2x00lib_probe_dev
  function. It fixes of starting rfkill_poll function at the
  right time if sets hard rfkill block and reboot. rt2x00mac_rfkill_poll
  should be starting before bringing up the wireless interface.

Signed-off-by: Chen, Chien-Chia <machen@suse.com>
Acked-by: Helmut Schaa <helmut.schaa@googlemail.com>
CC: Kevin Chou <kevin.chou@mediatek.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

7 years agobcma: fix build error on MIPS; implicit pcibios_enable_device
Paul Gortmaker [Mon, 26 Mar 2012 00:02:55 +0000]
bcma: fix build error on MIPS; implicit pcibios_enable_device

The following is seen during allmodconfig builds for MIPS:

drivers/bcma/driver_pci_host.c:518:2: error: implicit declaration
of function 'pcibios_enable_device' [-Werror=implicit-function-declaration]
cc1: some warnings being treated as errors
make[3]: *** [drivers/bcma/driver_pci_host.o] Error 1

Most likey introduced by commit 49dc9577155576b10ff79f0c1486c816b01f58bf

    "bcma: add PCIe host controller"

Add the header instead of implicitly assuming it will be present.
Sounds like a good idea, but that alone doesn't fix anything.

The real problem is that the Kconfig has settings related to whether
PCI is possible, i.e.

  config BCMA_HOST_PCI_POSSIBLE
        bool
        depends on BCMA && PCI = y
        default y

  config BCMA_HOST_PCI
        bool "Support for BCMA on PCI-host bus"
        depends on BCMA_HOST_PCI_POSSIBLE

...but what is missing is that BCMA_DRIVER_PCI_HOSTMODE doesn't
have any dependencies on the above.  Add one.

CC: Hauke Mehrtens <hauke@hauke-m.de>
CC: John W. Linville <linville@tuxdriver.com>
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Acked-by: Hauke Mehrtens <hauke@hauke-m.de>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

7 years agonet/wireless/wext-core.c: add missing kfree
Julia Lawall [Mon, 9 Apr 2012 09:01:09 +0000]
net/wireless/wext-core.c: add missing kfree

Free extra as done in the error-handling code just above.

Signed-off-by: Julia Lawall <Julia.Lawall@lip6.fr>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

7 years agortlwifi: Fix oops on rate-control failure
Larry Finger [Fri, 6 Apr 2012 21:54:17 +0000]
rtlwifi: Fix oops on rate-control failure

When the rate-control indexing is incorrectly set up, mac80211 issues
a warning and returns NULL from the call to ieee80211_get_tx_rate().
When this happens, avoid a NULL pointer dereference.

Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

7 years agomac80211: Convert WARN_ON to WARN_ON_ONCE
Larry Finger [Fri, 6 Apr 2012 21:35:53 +0000]
mac80211: Convert WARN_ON to WARN_ON_ONCE

When the control-rate tables are not set up correctly, it makes
little sense to spam the logs, thus change the WARN_ON to WARN_ON_ONCE.

Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

7 years agortlwifi: rtl8192de: Fix firmware initialization
Larry Finger [Tue, 3 Apr 2012 20:31:41 +0000]
rtlwifi: rtl8192de: Fix firmware initialization

Before the switch to asynchronous firmware loading (mainline commit b0302ab),
it was necessary to load firmware when initializing the first of the units
in a dual-mac system. After the change, it is necessary to load firmware in
both units.

Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

7 years agonl80211: ensure interface is up in various APIs
Johannes Berg [Wed, 18 Apr 2012 14:30:32 +0000]
nl80211: ensure interface is up in various APIs

The nl80211 handling code should ensure as much as
it can that the interface is in a valid state, it
can certainly ensure the interface is running.

Not doing so can cause calls through mac80211 into
the driver that result in warnings and unspecified
behaviour in the driver.

Change-Id: I60c836aac638b8be33011ffcf455d20e0e1bfc61
Cc: stable@vger.kernel.org
Reported-by: Ben Greear <greearb@candelatech.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

7 years agomac80211: fix association beacon wait timeout
Johannes Berg [Wed, 28 Mar 2012 19:02:46 +0000]
mac80211: fix association beacon wait timeout

The TU_TO_EXP_TIME() macro already includes the
"jiffies +" piece of the calculation, so don't
add jiffies again.

Reported-by: Oliver Hartkopp <socketcan@hartkopp.net>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Tested-by: Oliver Hartkopp <socketcan@hartkopp.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

7 years agoBluetooth: Fix userspace compatibility issue with mgmt interface
Gustavo Padovan [Thu, 29 Mar 2012 12:47:53 +0000]
Bluetooth: Fix userspace compatibility issue with mgmt interface

To ensure that old user space versions do not accidentally pick up and
try to use the management channel, use a different channel number.

Reported-by: Keith Packard <keithp@keithp.com>
Acked-by: Johan Hedberg <johan.hedberg@intel.com>
Acked-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Gustavo Padovan <gustavo@padovan.org>

7 years agoBluetooth: btusb: typo in Broadcom SoftSailing id
Don Zickus [Wed, 28 Mar 2012 20:41:11 +0000]
Bluetooth: btusb: typo in Broadcom SoftSailing id

I was trying to backport the following commit to RHEL-6

    From 0cea73465cd22373c5cd43a3edd25fbd4bb532ef Mon Sep 17 00:00:00 2001
    From: Oliver Neukum <oliver@neukum.org>
    Date: Wed, 21 Sep 2011 11:37:15 +0200
    Subject: [PATCH] btusb: add device entry for Broadcom SoftSailing

and noticed it wasn't working on an HP Elitebook.  Looking into the patch I
noticed a very subtle typo in the ids.  The patch has '0x05ac' instead of
'0x0a5c'.  A snippet of the lsusb -v output also shows this:

Bus 002 Device 003: ID 0a5c:21e1 Broadcom Corp.
Device Descriptor:
  bLength                18
  bDescriptorType         1
  bcdUSB               2.00
  bDeviceClass          255 Vendor Specific Class
  bDeviceSubClass         1
  bDeviceProtocol         1
  bMaxPacketSize0        64
  idVendor           0x0a5c Broadcom Corp.
  idProduct          0x21e1
  bcdDevice            1.12
  iManufacturer           1 Broadcom Corp
  iProduct                2 BCM20702A0
  iSerial                 3 60D819F0338C
  bNumConfigurations      1

Looking at other Broadcom ids, the fix matches them whereas the original patch
matches Apple's ids.

Tested on an HP Elitebook 8760w.  The btusb binds and the userspace stuff loads
correctly.

Cc: Oliver Neukum <oliver@neukum.org>
Signed-off-by: Don Zickus <dzickus@redhat.com>
Acked-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>

7 years agoBluetooth: mgmt: Fix timeout type
Andrei Emeltchenko [Mon, 19 Mar 2012 07:42:31 +0000]
Bluetooth: mgmt: Fix timeout type

Silence sparse warnings:
net/bluetooth/mgmt.c:865:19: warning: cast to restricted __le16

Signed-off-by: Andrei Emeltchenko <andrei.emeltchenko@intel.com>
Acked-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Gustavo Padovan <gustavo@padovan.org>

7 years agoBluetooth: mgmt: Add missing endian conversion
Andrei Emeltchenko [Mon, 12 Mar 2012 10:13:11 +0000]
Bluetooth: mgmt: Add missing endian conversion

Add missing endian conversion for page scan interval and window.

Signed-off-by: Andrei Emeltchenko <andrei.emeltchenko@intel.com>
Acked-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>

7 years agoBluetooth: mgmt: Fix corruption of device_connected pkt
Brian Gix [Fri, 9 Mar 2012 22:07:03 +0000]
Bluetooth: mgmt: Fix corruption of device_connected pkt

Incorrect pointer passed to eir_append_data made mgmt_device_connected
event unparsable by mgmt user space entity.

Signed-off-by: Brian Gix <bgix@codeaurora.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>

7 years agoBluetooth: Add support for Atheros [04ca:3005]
AceLan Kao [Wed, 28 Mar 2012 02:25:36 +0000]
Bluetooth: Add support for Atheros [04ca:3005]

Add another vendor specific ID for Atheros AR3012 device.
This chip is wrapped by Lite-On Technology Corp.

output of usb-devices:
T:  Bus=04 Lev=01 Prnt=01 Port=03 Cnt=01 Dev#=  2 Spd=12  MxCh= 0
D:  Ver= 1.10 Cls=e0(wlcon) Sub=01 Prot=01 MxPS=64 #Cfgs=  1
P:  Vendor=04ca ProdID=3005 Rev=00.02
S:  Manufacturer=Atheros Communications
S:  Product=Bluetooth USB Host Controller
S:  SerialNumber=Alaska Day 2006
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
I:  If#= 1 Alt= 0 #EPs= 2 Cls=e0(wlcon) Sub=01 Prot=01 Driver=btusb

Signed-off-by: AceLan Kao <acelan.kao@canonical.com>
Signed-off-by: Gustavo Padovan <gustavo@padovan.org>

7 years agoBluetooth: Check for minimum data length in eir_has_data_type()
Johan Hedberg [Mon, 26 Mar 2012 11:21:42 +0000]
Bluetooth: Check for minimum data length in eir_has_data_type()

If passed 0 as data_length the (parsed < data_length - 1) test will be
true and cause a buffer overflow. In practice we need at least two bytes
for the element length and type so add a test for it to the very
beginning of the function.

Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
Acked-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Gustavo Padovan <gustavo@padovan.org>

7 years agoBluetooth: Don't increment twice in eir_has_data_type()
Johan Hedberg [Mon, 26 Mar 2012 11:21:41 +0000]
Bluetooth: Don't increment twice in eir_has_data_type()

The parsed variable is already incremented inside the for-loop so there
no need to increment it again (not to mention that the code was
incrementing it the wrong amount).

Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
Acked-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Gustavo Padovan <gustavo@padovan.org>

7 years agoBluetooth: Fix memory leaks due to chan refcnt
Andrei Emeltchenko [Fri, 23 Mar 2012 07:42:15 +0000]
Bluetooth: Fix memory leaks due to chan refcnt

When we queue delayed work we hold(chan) and delayed work
shall put(chan) after execution.

Signed-off-by: Andrei Emeltchenko <andrei.emeltchenko@intel.com>
Signed-off-by: Gustavo Padovan <gustavo@padovan.org>

7 years agoBluetooth: hci_core: fix NULL-pointer dereference at unregister
Johan Hovold [Thu, 15 Mar 2012 13:48:41 +0000]
Bluetooth: hci_core: fix NULL-pointer dereference at unregister

Make sure hci_dev_open returns immediately if hci_dev_unregister has
been called.

This fixes a race between hci_dev_open and hci_dev_unregister which can
lead to a NULL-pointer dereference.

Bug is 100% reproducible using hciattach and a disconnected serial port:

0. # hciattach -n /dev/ttyO1 any noflow

1. hci_dev_open called from hci_power_on grabs req lock
2. hci_init_req executes but device fails to initialise (times out
   eventually)
3. hci_dev_open is called from hci_sock_ioctl and sleeps on req lock
4. hci_uart_tty_close calls hci_dev_unregister and sleeps on req lock in
   hci_dev_do_close
5. hci_dev_open (1) releases req lock
6. hci_dev_do_close grabs req lock and returns as device is not up
7. hci_dev_unregister sleeps in destroy_workqueue
8. hci_dev_open (3) grabs req lock, calls hci_init_req and eventually sleeps
9. hci_dev_unregister finishes, while hci_dev_open is still running...

[   79.627136] INFO: trying to register non-static key.
[   79.632354] the code is fine but needs lockdep annotation.
[   79.638122] turning off the locking correctness validator.
[   79.643920] [<c00188bc>] (unwind_backtrace+0x0/0xf8) from [<c00729c4>] (__lock_acquire+0x1590/0x1ab0)
[   79.653594] [<c00729c4>] (__lock_acquire+0x1590/0x1ab0) from [<c00733f8>] (lock_acquire+0x9c/0x128)
[   79.663085] [<c00733f8>] (lock_acquire+0x9c/0x128) from [<c0040a88>] (run_timer_softirq+0x150/0x3ac)
[   79.672668] [<c0040a88>] (run_timer_softirq+0x150/0x3ac) from [<c003a3b8>] (__do_softirq+0xd4/0x22c)
[   79.682281] [<c003a3b8>] (__do_softirq+0xd4/0x22c) from [<c003a924>] (irq_exit+0x8c/0x94)
[   79.690856] [<c003a924>] (irq_exit+0x8c/0x94) from [<c0013a50>] (handle_IRQ+0x34/0x84)
[   79.699157] [<c0013a50>] (handle_IRQ+0x34/0x84) from [<c0008530>] (omap3_intc_handle_irq+0x48/0x4c)
[   79.708648] [<c0008530>] (omap3_intc_handle_irq+0x48/0x4c) from [<c037499c>] (__irq_usr+0x3c/0x60)
[   79.718048] Exception stack(0xcf281fb0 to 0xcf281ff8)
[   79.723358] 1fa0:                                     0001e6a0 be8dab00 0001e698 00036698
[   79.731933] 1fc0: 0002df98 0002df38 0000001f 00000000 b6f234d0 00000000 00000004 00000000
[   79.740509] 1fe0: 0001e6f8 be8d6aa0 be8dac50 0000aab8 80000010 ffffffff
[   79.747497] Unable to handle kernel NULL pointer dereference at virtual address 00000000
[   79.756011] pgd = cf3b4000
[   79.758850] [00000000] *pgd=8f0c7831, *pte=00000000, *ppte=00000000
[   79.765502] Internal error: Oops: 80000007 [#1]
[   79.770294] Modules linked in:
[   79.773529] CPU: 0    Tainted: G        W     (3.3.0-rc6-00002-gb5d5c87 #421)
[   79.781066] PC is at 0x0
[   79.783721] LR is at run_timer_softirq+0x16c/0x3ac
[   79.788787] pc : [<00000000>]    lr : [<c0040aa4>]    psr: 60000113
[   79.788787] sp : cf281ee0  ip : 00000000  fp : cf280000
[   79.800903] r10: 00000004  r9 : 00000100  r8 : b6f234d0
[   79.806427] r7 : c0519c28  r6 : cf093488  r5 : c0561a00  r4 : 00000000
[   79.813323] r3 : 00000000  r2 : c054eee0  r1 : 00000001  r0 : 00000000
[   79.820190] Flags: nZCv  IRQs on  FIQs on  Mode SVC_32  ISA ARM  Segment user
[   79.827728] Control: 10c5387d  Table: 8f3b4019  DAC: 00000015
[   79.833801] Process gpsd (pid: 1265, stack limit = 0xcf2802e8)
[   79.839965] Stack: (0xcf281ee0 to 0xcf282000)
[   79.844573] 1ee0: 00000002 00000000 c0040a24 00000000 00000002 cf281f08 00200200 00000000
[   79.853210] 1f00: 00000000 cf281f18 cf281f08 00000000 00000000 00000000 cf281f18 cf281f18
[   79.861816] 1f20: 00000000 00000001 c056184c 00000000 00000001 b6f234d0 c0561848 00000004
[   79.870452] 1f40: cf280000 c003a3b8 c051e79c 00000001 00000000 00000100 3fa9e7b8 0000000a
[   79.879089] 1f60: 00000025 cf280000 00000025 00000000 00000000 b6f234d0 00000000 00000004
[   79.887756] 1f80: 00000000 c003a924 c053ad38 c0013a50 fa200000 cf281fb0 ffffffff c0008530
[   79.896362] 1fa0: 0001e6a0 0000aab8 80000010 c037499c 0001e6a0 be8dab00 0001e698 00036698
[   79.904998] 1fc0: 0002df98 0002df38 0000001f 00000000 b6f234d0 00000000 00000004 00000000
[   79.913665] 1fe0: 0001e6f8 be8d6aa0 be8dac50 0000aab8 80000010 ffffffff 00fbf700 04ffff00
[   79.922302] [<c0040aa4>] (run_timer_softirq+0x16c/0x3ac) from [<c003a3b8>] (__do_softirq+0xd4/0x22c)
[   79.931945] [<c003a3b8>] (__do_softirq+0xd4/0x22c) from [<c003a924>] (irq_exit+0x8c/0x94)
[   79.940582] [<c003a924>] (irq_exit+0x8c/0x94) from [<c0013a50>] (handle_IRQ+0x34/0x84)
[   79.948913] [<c0013a50>] (handle_IRQ+0x34/0x84) from [<c0008530>] (omap3_intc_handle_irq+0x48/0x4c)
[   79.958404] [<c0008530>] (omap3_intc_handle_irq+0x48/0x4c) from [<c037499c>] (__irq_usr+0x3c/0x60)
[   79.967773] Exception stack(0xcf281fb0 to 0xcf281ff8)
[   79.973083] 1fa0:                                     0001e6a0 be8dab00 0001e698 00036698
[   79.981658] 1fc0: 0002df98 0002df38 0000001f 00000000 b6f234d0 00000000 00000004 00000000
[   79.990234] 1fe0: 0001e6f8 be8d6aa0 be8dac50 0000aab8 80000010 ffffffff
[   79.997161] Code: bad PC value
[   80.000396] ---[ end trace 6f6739840475f9ee ]---
[   80.005279] Kernel panic - not syncing: Fatal exception in interrupt

Cc: stable <stable@vger.kernel.org>
Signed-off-by: Johan Hovold <jhovold@gmail.com>
Acked-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>

7 years agoBluetooth: Add Atheros maryann PIDVID support
Cho, Yu-Chen [Wed, 14 Mar 2012 20:01:21 +0000]
Bluetooth: Add Atheros maryann PIDVID support

Add Atheros maryann 0cf3:311d PIDVID support
This module is AR3012 Series.

Include /sys/kernel/debug/usb/devices output here for reference

before:
T:  Bus=04 Lev=01 Prnt=01 Port=01 Cnt=01 Dev#=  2 Spd=12   MxCh= 0
D:  Ver= 1.10 Cls=e0(wlcon) Sub=01 Prot=01 MxPS=64 #Cfgs=  1
P:  Vendor=0cf3 ProdID=311d Rev= 0.01
S:  Manufacturer=Atheros Communications
S:  Product=Bluetooth USB Host Controller
S:  SerialNumber=Alaska Day 2006
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

after:
T:  Bus=04 Lev=01 Prnt=01 Port=01 Cnt=01 Dev#=  2 Spd=12   MxCh= 0
D:  Ver= 1.10 Cls=e0(wlcon) Sub=01 Prot=01 MxPS=64 #Cfgs=  1
P:  Vendor=0cf3 ProdID=311d Rev= 0.02
S:  Manufacturer=Atheros Communications
S:  Product=Bluetooth USB Host Controller
S:  SerialNumber=Alaska Day 2006
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: Cho, Yu-Chen <acho@suse.com>
cked-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>

7 years agoBluetooth: btusb: Add USB device ID "0a5c 21e8"
João Paulo Rechi Vita [Wed, 14 Mar 2012 19:45:16 +0000]
Bluetooth: btusb: Add USB device ID "0a5c 21e8"

One more vendor-specific ID for BCM20702A0.

T:  Bus=01 Lev=03 Prnt=05 Port=02 Cnt=01 Dev#=  9 Spd=12  MxCh= 0
D:  Ver= 2.00 Cls=ff(vend.) Sub=01 Prot=01 MxPS=64 #Cfgs=  1
P:  Vendor=0a5c ProdID=21e8 Rev=01.12
S:  Manufacturer=Broadcom Corp
S:  Product=BCM20702A0
S:  SerialNumber=00027221F4E2
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: João Paulo Rechi Vita <jprvita@openbossa.org>
Acked-by: Gustavo F. Padovan <padovan@profusion.mobi>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>

7 years agoBluetooth: Fix Endian Bug.
Santosh Nayak [Thu, 1 Mar 2012 17:16:36 +0000]
Bluetooth: Fix Endian Bug.

Fix network to host endian conversion for L2CAP chan id.

Signed-off-by: Santosh Nayak <santoshprasadnayak@gmail.com>
Acked-by: Andrei Emeltchenko <andrei.emeltchenko@intel.com>
Signed-off-by: Gustavo F. Padovan <padovan@profusion.mobi>

7 years agoBluetooth: Use correct flags for checking HCI_SSP_ENABLED bit
Hemant Gupta [Wed, 28 Mar 2012 11:39:09 +0000]
Bluetooth: Use correct flags for checking HCI_SSP_ENABLED bit

This patch uses the correct flags for checking the HCI_SSP_ENABLED bit.
Without this authentication request was not being initiated.

Signed-off-by: Hemant Gupta <hemant.gupta@stericsson.com>
Acked-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>

7 years agoMAINTAINERS: update Bluetooth tree locations
Marcel Holtmann [Tue, 27 Mar 2012 16:21:00 +0000]
MAINTAINERS: update Bluetooth tree locations

Make use of a shared tree setup to limit the confusions on which
tree is current.

Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Gustavo Padovan <gustavo@padovan.org>

7 years agotcp: avoid order-1 allocations on wifi and tx path
Eric Dumazet [Tue, 10 Apr 2012 20:30:48 +0000]
tcp: avoid order-1 allocations on wifi and tx path

Marc Merlin reported many order-1 allocations failures in TX path on its
wireless setup, that dont make any sense with MTU=1500 network, and non
SG capable hardware.

After investigation, it turns out TCP uses sk_stream_alloc_skb() and
used as a convention skb_tailroom(skb) to know how many bytes of data
payload could be put in this skb (for non SG capable devices)

Note : these skb used kmalloc-4096 (MTU=1500 + MAX_HEADER +
sizeof(struct skb_shared_info) being above 2048)

Later, mac80211 layer need to add some bytes at the tail of skb
(IEEE80211_ENCRYPT_TAILROOM = 18 bytes) and since no more tailroom is
available has to call pskb_expand_head() and request order-1
allocations.

This patch changes sk_stream_alloc_skb() so that only
sk->sk_prot->max_header bytes of headroom are reserved, and use a new
skb field, avail_size to hold the data payload limit.

This way, order-0 allocations done by TCP stack can leave more than 2 KB
of tailroom and no more allocation is performed in mac80211 layer (or
any layer needing some tailroom)

avail_size is unioned with mark/dropcount, since mark will be set later
in IP stack for output packets. Therefore, skb size is unchanged.

Reported-by: Marc MERLIN <marc@merlins.org>
Tested-by: Marc MERLIN <marc@merlins.org>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

7 years agonet: allow pskb_expand_head() to get maximum tailroom
Eric Dumazet [Tue, 10 Apr 2012 20:08:39 +0000]
net: allow pskb_expand_head() to get maximum tailroom

Marc Merlin reported many order-1 allocations failures in TX path on its
wireless setup, that dont make any sense with MTU=1500 network, and non
SG capable hardware.

Turns out part of the problem comes from pskb_expand_head() not using
ksize() to get exact head size given by kmalloc(). Doing the same thing
than __alloc_skb() allows more tailroom in skb and can prevent future
reallocations.

As a bonus, struct skb_shared_info becomes cache line aligned.

Reported-by: Marc MERLIN <marc@merlins.org>
Tested-by: Marc MERLIN <marc@merlins.org>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

7 years agobridge: Do not send queries on multicast group leaves
Herbert Xu [Wed, 4 Apr 2012 01:01:20 +0000]
bridge: Do not send queries on multicast group leaves

As it stands the bridge IGMP snooping system will respond to
group leave messages with queries for remaining membership.
This is both unnecessary and undesirable.  First of all any
multicast routers present should be doing this rather than us.
What's more the queries that we send may end up upsetting other
multicast snooping swithces in the system that are buggy.

In fact, we can simply remove the code that send these queries
because the existing membership expiry mechanism doesn't rely
on them anyway.

So this patch simply removes all code associated with group
queries in response to group leave messages.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>

7 years agoMAINTAINERS: Mark NATSEMI driver as orphan'd.
David S. Miller [Wed, 11 Apr 2012 01:10:43 +0000]
MAINTAINERS: Mark NATSEMI driver as orphan'd.

After discussion with Tim Hockin.

Signed-off-by: David S. Miller <davem@davemloft.net>

7 years agotcp: fix tcp_rcv_rtt_update() use of an unscaled RTT sample
Neal Cardwell [Tue, 10 Apr 2012 07:59:20 +0000]
tcp: fix tcp_rcv_rtt_update() use of an unscaled RTT sample

Fix a code path in tcp_rcv_rtt_update() that was comparing scaled and
unscaled RTT samples.

The intent in the code was to only use the 'm' measurement if it was a
new minimum.  However, since 'm' had not yet been shifted left 3 bits
but 'new_sample' had, this comparison would nearly always succeed,
leading us to erroneously set our receive-side RTT estimate to the 'm'
sample when that sample could be nearly 8x too high to use.

The overall effect is to often cause the receive-side RTT estimate to
be significantly too large (up to 40% too large for brief periods in
my tests).

Signed-off-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

7 years agotcp: restore correct limit
Eric Dumazet [Tue, 10 Apr 2012 00:56:42 +0000]
tcp: restore correct limit

Commit c43b874d5d714f (tcp: properly initialize tcp memory limits) tried
to fix a regression added in commits 4acb4190 & 3dc43e3,
but still get it wrong.

Result is machines with low amount of memory have too small tcp_rmem[2]
value and slow tcp receives : Per socket limit being 1/1024 of memory
instead of 1/128 in old kernels, so rcv window is capped to small
values.

Fix this to match comment and previous behavior.

Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Jason Wang <jasowang@redhat.com>
Cc: Glauber Costa <glommer@parallels.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

7 years agonetfilter: nf_conntrack: fix incorrect logic in nf_conntrack_init_net
Gao feng [Sat, 7 Apr 2012 16:08:28 +0000]
netfilter: nf_conntrack: fix incorrect logic in nf_conntrack_init_net

in function nf_conntrack_init_net,when nf_conntrack_timeout_init falied,
we should call nf_conntrack_ecache_fini to do rollback.
but the current code calls nf_conntrack_timeout_fini.

Signed-off-by: Gao feng <gaofeng@cn.fujitsu.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

7 years agonetfilter: nf_ct_ipv4: packets with wrong ihl are invalid
Jozsef Kadlecsik [Tue, 3 Apr 2012 20:02:01 +0000]
netfilter: nf_ct_ipv4: packets with wrong ihl are invalid

It was reported that the Linux kernel sometimes logs:

klogd: [2629147.402413] kernel BUG at net / netfilter /
nf_conntrack_proto_tcp.c: 447!
klogd: [1072212.887368] kernel BUG at net / netfilter /
nf_conntrack_proto_tcp.c: 392

ipv4_get_l4proto() in nf_conntrack_l3proto_ipv4.c and tcp_error() in
nf_conntrack_proto_tcp.c should catch malformed packets, so the errors
at the indicated lines - TCP options parsing - should not happen.
However, tcp_error() relies on the "dataoff" offset to the TCP header,
calculated by ipv4_get_l4proto().  But ipv4_get_l4proto() does not check
bogus ihl values in IPv4 packets, which then can slip through tcp_error()
and get caught at the TCP options parsing routines.

The patch fixes ipv4_get_l4proto() by invalidating packets with bogus
ihl value.

The patch closes netfilter bugzilla id 771.

Signed-off-by: Jozsef Kadlecsik <kadlec@blackhole.kfki.hu>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

7 years agonetfilter: nf_ct_ipv4: handle invalid IPv4 and IPv6 packets consistently
Jozsef Kadlecsik [Mon, 9 Apr 2012 14:32:16 +0000]
netfilter: nf_ct_ipv4: handle invalid IPv4 and IPv6 packets consistently

IPv6 conntrack marked invalid packets as INVALID and let the user
drop those by an explicit rule, while IPv4 conntrack dropped such
packets itself.

IPv4 conntrack is changed so that it marks INVALID packets and let
the user to drop them.

Signed-off-by: Jozsef Kadlecsik <kadlec@blackhole.kfki.hu>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

7 years agonetfilter: ip6_tables: ip6t_ext_hdr is now static inline
Pablo Neira Ayuso [Fri, 6 Apr 2012 16:12:54 +0000]
netfilter: ip6_tables: ip6t_ext_hdr is now static inline

We may hit this in xt_LOG:

net/built-in.o:xt_LOG.c:function dump_ipv6_packet:
error: undefined reference to 'ip6t_ext_hdr'

happens with these config options:

CONFIG_NETFILTER_XT_TARGET_LOG=y
CONFIG_IP6_NF_IPTABLES=m

ip6t_ext_hdr is fairly small and it is called in the packet path.
Make it static inline.

Reported-by: Simon Kirby <sim@netnation.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

7 years agonetfilter: nf_ct_tcp: don't scale the size of the window up twice
Changli Gao [Sun, 1 Apr 2012 17:25:06 +0000]
netfilter: nf_ct_tcp: don't scale the size of the window up twice

For a picked up connection, the window win is scaled twice: one is by the
initialization code, and the other is by the sender updating code.

I use the temporary variable swin instead of modifying the variable win.

Signed-off-by: Changli Gao <xiaosuo@gmail.com>
Acked-by: Jozsef Kadlecsik <kadlec@blackhole.kfki.hu>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

7 years agodrm/exynos: fixed exynos broken ioctl
Inki Dae [Thu, 12 Apr 2012 07:42:54 +0000]
drm/exynos: fixed exynos broken ioctl

this patch removes the pointer of uint64_t *edid. it should be just
a uint64_t.

Signed-off-by: Inki Dae <inki.dae@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>

7 years agodrm/exynos: fix to pointer manager member of struct exynos_drm_subdrv
Joonyoung Shim [Thu, 5 Apr 2012 11:49:27 +0000]
drm/exynos: fix to pointer manager member of struct exynos_drm_subdrv

The struct exynos_drm_manager has to exist for exynos drm sub driver
using encoder and connector. If it isn't NULL to member of struct
exynos_drm_subdrv, will create encoder and connector else will not. And
the is_local member also doesn't need.

Signed-off-by: Joonyoung Shim <jy0922.shim@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>

7 years agodrm/exynos: fix struct for operation callback functions to driver name
Joonyoung Shim [Thu, 5 Apr 2012 11:49:26 +0000]
drm/exynos: fix struct for operation callback functions to driver name

The mixer driver and hdmi driver have each operation callback functions
and they is registered to hdmi common driver. Their struct names in hdmi
common driver include display, manager and overlay. It confuses to
appear whose operation and two driver cannot register same operation
callback functions at the same time. Use their struct names to driver
name.

Signed-off-by: Joonyoung Shim <jy0922.shim@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>

7 years agodrm/exynos: use define instead of default_win member in struct mixer_context
Joonyoung Shim [Thu, 5 Apr 2012 11:49:25 +0000]
drm/exynos: use define instead of default_win member in struct mixer_context

The default_win member in struct mixer_context isn't change its value
after initialized to 0, so it's better using to define.

Signed-off-by: Joonyoung Shim <jy0922.shim@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>

7 years agodrm/exynos: rename s/HDMI_OVERLAY_NUMBER/MIXER_WIN_NR
Joonyoung Shim [Thu, 5 Apr 2012 11:49:24 +0000]
drm/exynos: rename s/HDMI_OVERLAY_NUMBER/MIXER_WIN_NR

HDMI_OVERLAY_NUMBER is specific of mixer driver and be used "windows
layer" term in exynos user manaual, so rename it.

Signed-off-by: Joonyoung Shim <jy0922.shim@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>

7 years agodrm/exynos: remove unused codes in hdmi and mixer
Joonyoung Shim [Thu, 5 Apr 2012 11:49:23 +0000]
drm/exynos: remove unused codes in hdmi and mixer

Some members in struct mixer_context aren't used and the define
HDMI_OVERLAY_NUMBER is unused in hdmi driver, remove them.

Signed-off-by: Joonyoung Shim <jy0922.shim@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>

7 years agodrm/exynos: remove unnecessary type conversion of hdmi and mixer
Joonyoung Shim [Thu, 5 Apr 2012 11:49:22 +0000]
drm/exynos: remove unnecessary type conversion of hdmi and mixer

When the void pointer type variable is assigned to the specific pointer
type variable, don't need to do type conversion.

Signed-off-by: Joonyoung Shim <jy0922.shim@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>

7 years agodrm/exynos: add format list of plane
Seung-Woo Kim [Thu, 5 Apr 2012 02:21:09 +0000]
drm/exynos: add format list of plane

NV12, NV12M and NV12MT are added to format list of plane to use these
formats for hdmi vp layer.

Signed-off-by: Seung-Woo Kim <sw0312.kim@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>

7 years agodrm/exynos: fixed duplicated page allocation bug.
Inki Dae [Tue, 3 Apr 2012 12:49:15 +0000]
drm/exynos: fixed duplicated page allocation bug.

this patch fixes that buf->pages is allocated two times when it allocates
physically continuous memory region and removes unnecessary codes.

Signed-off-by: Inki Dae <inki.dae@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>

7 years agodrm/exynos: fixed page align and code clean.
Inki Dae [Tue, 3 Apr 2012 12:27:58 +0000]
drm/exynos: fixed page align and code clean.

1M section, 64k page count also should be rounded up so this patch
rounds up them and caculates page count of them properly and also
checks memory flags from user.

Signed-off-by: Inki Dae <inki.dae@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>

7 years agodrm/i915: clear fencing tracking state when retiring requests
Daniel Vetter [Wed, 11 Apr 2012 23:27:57 +0000]
drm/i915: clear fencing tracking state when retiring requests

This fixes a resume regression introduced in

commit 7dd4906586274f3945f2aeaaa5a33b451c3b4bba
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date:   Wed Mar 21 10:48:18 2012 +0000

    drm/i915: Mark untiled BLT commands as fenced on gen2/3

which fixed fencing tracking for untiled blt commands.

A side effect of that patch was that now also untiled objects have a
non-zero obj->last_fenced_seqno to track when a fence can be set up
after a pipelined tiling change. Unfortunately this was only cleared
by the fence setup and teardown code, resulting in tons of untiled but
inactive objects with non-zero last_fenced_seqno.

Now after resume we completely reset the seqno tracking, both on the
driver side (by setting dev_priv->next_seqno = 1) and on the hw side
(by allocating a new hws page, which contains the seqnos). Hilarity
and indefinite waits ensued from the stale seqnos in
obj->last_fenced_seqno from before the suspend.

The fix is to properly clear the fencing tracking state like we
already do for the normal gpu rendering while moving objects off the
active list.

Reported-and-tested-by: "Rafael J. Wysocki" <rjw@sisk.pl>
Cc: Jiri Slaby <jslaby@suse.cz>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

7 years agodrm/i915: make rc6 module parameter read-only
Jesse Barnes [Wed, 11 Apr 2012 16:39:02 +0000]
drm/i915: make rc6 module parameter read-only

People have been getting confused and thinking this is a runtime control.

Cc: stable@vger.kernel.org
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

7 years agodrm/i915: implement ColorBlt w/a
Daniel Vetter [Sat, 31 Mar 2012 09:21:57 +0000]
drm/i915: implement ColorBlt w/a

According to an internal workaround master list, we need to set bit 5
of register 9400 to avoid issues with color blits.

Testing shows that this seems to fix the blitter hangs when fbc is
enabled on snb, thanks to Chris Wilson for figuring this out.

Tested-by: Chris Wilson <chris@chris-wilson.co.uk>
Tested-by: Michael "brot" Groh <michael.groh@minad.de>
Acked-by: Ben Widawsky <ben@bwidawsk.net>
Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

7 years agodrm/i915/ringbuffer: Exclude last 2 cachlines of ring on 845g
Chris Wilson [Mon, 9 Apr 2012 12:59:46 +0000]
drm/i915/ringbuffer: Exclude last 2 cachlines of ring on 845g

The 845g shares the errata with i830 whereby executing a command
within 2 cachelines of the end of the ringbuffer may cause a GPU hang.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: stable@kernel.org
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

7 years agoRevert "drm/i915: reenable gmbus on gen3+ again"
Daniel Vetter [Mon, 9 Apr 2012 19:10:38 +0000]
Revert "drm/i915: reenable gmbus on gen3+ again"

This reverts commit c3dfefa0a6d235bd465309e12f4c56ea16e71111.

gmbus in 3.4 has simply too many known issues:
- gmbus is too noisy, we need to rework the logging:
  https://bugs.freedesktop.org/show_bug.cgi?id=48248
- zero-length writes cause an OOPS, and they are
  userspace-triggerable:
  https://lkml.org/lkml/2012/3/30/176
- same for zero-length reads:
  https://bugs.freedesktop.org/show_bug.cgi?id=48269

We can try again for 3.5.

Acked-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

7 years agodrm/i915: properly compute dp dithering for user-created modes
Daniel Vetter [Tue, 10 Apr 2012 08:42:36 +0000]
drm/i915: properly compute dp dithering for user-created modes

We've only computed whether we need to fall back to 6bpc due to dp
link bandwidth constrains in mode_valid, but not mode_fixup. Under
various circumstances X likes to create new modes which then lack
proper 6bpc flags (if required), resulting in mode_fixup failures and
ultimately black screens.

Chris Wilson pointed out that we still get things wrong for bpp > 24,
but that should be fixed in another patch (and it'll be easier because
this patch consolidates the logic).

The likely culprit for this regression is

commit 3d794f87238f74d80e78a7611c7fbde8a54c85c2
Author: Keith Packard <keithp@keithp.com>
Date:   Wed Jan 25 08:16:25 2012 -0800

    drm/i915: Force explicit bpp selection for intel_dp_link_required

v2: Fix indentation and tune down the too bold claim that this should
fix the world. Both noticed by Chris Wilson.

v3: Try to really git add things.

Reported-and-tested-by: Brice Goglin <Brice.Goglin@ens-lyon.org>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=48170
Cc: stable@kernel.org
Reviewed-by: Adam Jackson <ajax@redhat.com>
Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

7 years agodrm/i915: Finish any pending operations on the framebuffer before disabling
Chris Wilson [Tue, 3 Apr 2012 16:58:35 +0000]
drm/i915: Finish any pending operations on the framebuffer before disabling

Similar to the case where we are changing from one framebuffer to
another, we need to be sure that there are no pending WAIT_FOR_EVENTs on
the pipe for the current framebuffer before switching. If we disable the
pipe, and then try to execute a WAIT_FOR_EVENT it will block
indefinitely and cause a GPU hang.

We attempted to fix this in commit 85345517fe6d4de27b0d6ca19fef9d28ac947c4a
(drm/i915: Retire any pending operations on the old scanout when switching)
for the case of mode switching, but this leaves the condition where we
are switching off the pipe vulnerable.

There still remains the race condition were a display may be unplugged,
switched off by the core, a uevent sent to notify the DDX and the DDX
may issue a WAIT_FOR_EVENT before it processes the uevent. This window
does not exist if the pipe is only switched off in response to the
uevent. Time to make sure that is so...

Reported-by: Francis Leblanc <Francis.Leblanc-Lebeau@verint.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=36515
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=45413
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Eugeni Dodonov <eugeni.dodonov@intel.com>
[danvet: fixup spelling in comment, noticed by Eugeni.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

7 years agodrm/i915: Removed IVB forced enable of sprite dest key.
acreese [Wed, 4 Apr 2012 23:22:32 +0000]
drm/i915: Removed IVB forced enable of sprite dest key.

The destination color key is always enabled for IVB.  Removed
the line that does this.

Signed-off-by: Armin Reese <armin.c.reese@intel.com>
Acked-by: Jesse Barnes <jbarnes@virtuousgeek.org>
Cc: stable@kernel.org
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

7 years agodrm/radeon: only add the mm i2c bus if the hw_i2c module param is set
Alex Deucher [Tue, 10 Apr 2012 16:14:27 +0000]
drm/radeon: only add the mm i2c bus if the hw_i2c module param is set

It seems it can corrupt the monitor EDID in certain cases on certain
boards when running sensors detect.  It's rarely used anyway outside
of AIW boards.

http://lists.lm-sensors.org/pipermail/lm-sensors/2012-April/035847.html
http://lists.freedesktop.org/archives/xorg/2011-January/052239.html

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
Acked-by: Jean Delvare <khali@linux-fr.org>
Signed-off-by: Dave Airlie <airlied@redhat.com>

7 years agovgaarb.h: fix build warnings
Randy Dunlap [Mon, 30 Jan 2012 19:43:52 +0000]
vgaarb.h: fix build warnings

Fix build warnings by providing a struct stub since no fields of
the struct are used:

include/linux/vgaarb.h:66:9: warning: 'struct pci_dev' declared inside parameter list
include/linux/vgaarb.h:66:9: warning: its scope is only this definition or declaration, which is probably not what you want
include/linux/vgaarb.h:99:34: warning: 'struct pci_dev' declared inside parameter list
include/linux/vgaarb.h:109:6: warning: 'struct pci_dev' declared inside parameter list
include/linux/vgaarb.h:121:8: warning: 'struct pci_dev' declared inside parameter list
include/linux/vgaarb.h:140:37: warning: 'struct pci_dev' declared inside parameter list

Signed-off-by: Randy Dunlap <rdunlap@xenotime.net>
Signed-off-by: Dave Airlie <airlied@redhat.com>

7 years agodrm/radeon/kms: fix DVO setup on some r4xx chips
Alex Deucher [Tue, 3 Apr 2012 21:05:41 +0000]
drm/radeon/kms: fix DVO setup on some r4xx chips

Some r4xx chips have the wrong frev in the
DVOEncoderControl table.  It should always be 1
on r4xx.  Fixes modesetting on DVO on r4xx chips
with the bad frev.

Reported by twied on #radeon.

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
Signed-off-by: Dave Airlie <airlied@redhat.com>

7 years agodrm/savage: fix integer overflows in savage_bci_cmdbuf()
Xi Wang [Fri, 6 Apr 2012 21:38:24 +0000]
drm/savage: fix integer overflows in savage_bci_cmdbuf()

Since cmdbuf->size and cmdbuf->nbox are from userspace, a large value
would overflow the allocation size, leading to out-of-bounds access.

Signed-off-by: Xi Wang <xi.wang@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

7 years agodrm/radeon: replace udelay with mdelay for long timeouts
Arnd Bergmann [Thu, 5 Apr 2012 18:58:22 +0000]
drm/radeon: replace udelay with mdelay for long timeouts

Some architectures require that delays longer than a few
miliseconds are called through mdelay. This was triggered
on ARM randconfig builds.

Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
Signed-off-by: Dave Airlie <airlied@redhat.com>

7 years agomd/bitmap: prevent bitmap_daemon_work running while initialising bitmap
NeilBrown [Thu, 12 Apr 2012 06:05:06 +0000]
md/bitmap: prevent bitmap_daemon_work running while initialising bitmap

If a bitmap is added while the array is active, it is possible
for bitmap_daemon_work to run while the bitmap is being
initialised.
This is particularly a problem if bitmap_daemon_work sees
bitmap->filemap as non-NULL before it has been filled in properly.
So hold bitmap_info.mutex while filling in ->filemap
to prevent problems.

This patch is suitable for any -stable kernel, though it might not
apply cleanly before about 3.1.

Cc: stable@vger.kernel.org
Signed-off-by: NeilBrown <neilb@suse.de>

7 years agomd/raid1,raid10: Fix calculation of 'vcnt' when processing error recovery.
majianpeng [Thu, 12 Apr 2012 06:04:47 +0000]
md/raid1,raid10: Fix calculation of 'vcnt' when processing error recovery.

If r1bio->sectors % 8 != 0,then the memcmp and a later
memcpy will omit the last bio_vec.

This is suitable for any stable kernel since 3.1 when bad-block
management was introduced.

Cc: stable@vger.kernel.org
Signed-off-by: majianpeng <majianpeng@gmail.com>
Signed-off-by: NeilBrown <neilb@suse.de>

7 years agoMD: Bitmap version cleanup.
Andrei Warkentin [Thu, 12 Apr 2012 05:55:21 +0000]
MD: Bitmap version cleanup.

bitmap_new_disk_sb() would still create V3 bitmap superblock
with host-endian layout.

Perhaps I'm confused, but shouldn't bitmap_new_disk_sb() be
creating a V4 bitmap superblock instead, that is portable,
as per comment in bitmap.h?

Signed-off-by: Andrei Warkentin <andrey.warkentin@gmail.com>
Signed-off-by: NeilBrown <neilb@suse.de>

7 years agoirq_domain: Move irq_virq_count into NOMAP revmap
Grant Likely [Wed, 15 Feb 2012 22:06:08 +0000]
irq_domain: Move irq_virq_count into NOMAP revmap

This patch replaces the old global setting of irq_virq_count that is only
used by the NOMAP mapping and instead uses a revmap_data property so that
the maximum NOMAP allocation can be set per NOMAP irq_domain.

There is exactly one user of irq_virq_count in-tree right now: PS3.
Also, irq_virq_count is only useful for the NOMAP mapping.  So,
instead of having a single global irq_virq_count values, this change
drops it entirely and added a max_irq argument to irq_domain_add_nomap().
That makes it a property of an individual nomap irq domain instead of
a global system settting.

Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
Tested-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Milton Miller <miltonm@bga.com>

7 years agoirqdomain: Fix debugfs formatting
Grant Likely [Wed, 11 Apr 2012 06:26:25 +0000]
irqdomain: Fix debugfs formatting

This patch fixes the irq_domain_mapping debugfs output to pad pointer
values with leading zeros so that pointer values are displayed
correctly.  Otherwise you get output similar to "0x 5e0000000000000".
Also, when the irq_domain is set to 'null'

Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
Cc: David Daney <david.daney@cavium.com>
Cc: Mika Westerberg <mika.westerberg@linux.intel.com>

7 years agoirq_domain: correct the debugfs file name
Mika Westerberg [Tue, 10 Apr 2012 12:25:42 +0000]
irq_domain: correct the debugfs file name

The actual name of the irq_domain mapping debugfs file is
"irq_domain_mapping" not "virq_mapping".

Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: Grant Likely <grant.likely@secretlab.ca>

7 years agoirq: Kill pointless irqd_to_hw export
Grant Likely [Tue, 3 Apr 2012 13:11:04 +0000]
irq: Kill pointless irqd_to_hw export

It makes no sense to export this trivial function.  Make it a static inline
instead.

This patch also drops virq_to_hw from arch/c6x since it is unused by that
architecture.

v2: Move irq_hw_number_t into types.h to fix ARM build failure

Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
Acked-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>

7 years agoirq/irq_domain: Quit ignoring error returns from irq_alloc_desc_from().
David Daney [Thu, 5 Apr 2012 23:52:13 +0000]
irq/irq_domain: Quit ignoring error returns from irq_alloc_desc_from().

In commit 4bbdd45a (irq_domain/powerpc: eliminate irq_map; use
irq_alloc_desc() instead) code was added that ignores error returns
from irq_alloc_desc_from() by (silently) casting the return value to
unsigned.  The negitive value error return now suddenly looks like a
valid irq number.

Commits cc79ca69 (irq_domain: Move irq_domain code from powerpc to
kernel/irq) and 1bc04f2c (irq_domain: Add support for base irq and
hwirq in legacy mappings) move this code to its current location in
irqdomain.c

The result of all of this is a null pointer dereference OOPS if one of
the error cases is hit.

The fix: Don't cast away the negativeness of the return value and then
check for errors.

Signed-off-by: David Daney <david.daney@cavium.com>
Acked-by: Rob Herring <rob.herring@calxeda.com>
[grant.likely: dropped addition of new 'irq' variable]
Signed-off-by: Grant Likely <grant.likely@secretlab.ca>

7 years agoInput: da9052 - fix memory leak in da9052_onkey_probe()
Jesper Juhl [Thu, 12 Apr 2012 03:55:18 +0000]
Input: da9052 - fix memory leak in da9052_onkey_probe()

If, in drivers/input/misc/da9052_onkey.c::da9052_onkey_probe(), the
call to either kzalloc() or input_allocate_device() fails then we will
return -ENOMEM from the function without freeing the other allocation
that may have succeeded, thus we leak either the memory allocated for
'onkey' or the memory allocated for 'input_dev' if one succeeds and
the other fails.
Fix that by jumping to the 'err_free_mem' label at the end of the
function that properly cleans up rather than returning directly.

Signed-off-by: Jesper Juhl <jj@chaosbits.net>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>

7 years agoInput: gpio_mouse - use linux/gpio.h rather than asm/gpio.h
Mark Brown [Tue, 10 Apr 2012 16:03:03 +0000]
Input: gpio_mouse - use linux/gpio.h rather than asm/gpio.h

Direct usage of the asm include has long been deprecated by the
introduction of gpiolib.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>

7 years agoInput: trackpoint - use psmouse_fmt() for messages
JJ Ding [Tue, 10 Apr 2012 07:25:01 +0000]
Input: trackpoint - use psmouse_fmt() for messages

Use psmouse_*() macros introduced in commit b5d21704361ee.

Signed-off-by: JJ Ding <dgdunix@gmail.com>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>

7 years agoInput: elantech - v4 is a clickpad, with only one button
JJ Ding [Tue, 10 Apr 2012 07:30:12 +0000]
Input: elantech - v4 is a clickpad, with only one button

Add pointer and buttonpad properties for v4 hardware.

Also, Jachiet reported that on Asus UX31, right button has no effect.
It turns out v4 has only one button, the right-button effect is
implemented with software when Windows driver is installed, or in
firmware when touchpad is in relative mode. So remove BTN_RIGHT
while at it.

Reported-by: Jachiet Louis <louis@jachiet.com>
Signed-off-by: JJ Ding <jj_ding@emc.com.tw>
Reviewed-by: Chase Douglas <chase.douglas@canonical.com>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>

7 years agoInput: elantech - reset touchpad before configuring it
JJ Ding [Tue, 10 Apr 2012 07:29:12 +0000]
Input: elantech - reset touchpad before configuring it

Acer VH40 has a Fn key toggling the touchpad on and off, but it's
implemented in system firmware, and the EC chip has to receive
reset command to activate this function. Also when this machine
wakes up after resume, psmouse_reset is necessary to bring the
touchpad back on.

Signed-off-by: JJ Ding <jj_ding@emc.com.tw>
Reviewed-by: Chase Douglas <chase.douglas@canonical.com>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>

7 years agoInput: sentelic - filter taps in absolute mode
Oskari Saarenmaa [Tue, 3 Apr 2012 16:46:32 +0000]
Input: sentelic - filter taps in absolute mode

Taps in absolute positioning single-finger mode are currently reported
as physical clicks by the driver. This should be handled by userspace,
not the kernel.

When a tap occurs, the FSP_PB0_LBTN bit is set, but the FSP_PB0_PHY_BTN
is not. We use this to filter out physical clicks from taps.

Signed-off-by: Oskari Saarenmaa <os@ohmu.fi>
Reviewed-by: Tai-hwa Liang <avatar@sentelic.com>
Reviewed-by: Chase Douglas <chase.douglas@canonical.com>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>

7 years agoInput: tps6507x-ts - fix MODULE_ALIAS to match driver name
Uwe Kleine-König [Tue, 3 Apr 2012 16:33:58 +0000]
Input: tps6507x-ts - fix MODULE_ALIAS to match driver name

This is needed to make module auto loading work.

[dtor@mail.ru: remove file name from comment]
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>

7 years agotypo fix in Documentation/filesystems/vfs.txt
Al Viro [Mon, 2 Apr 2012 23:02:48 +0000]
typo fix in Documentation/filesystems/vfs.txt

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>