firefly-linux-kernel-4.4.55.git
13 years agoadd hym8563 rtc
林辉辉 [Thu, 13 May 2010 01:47:43 +0000 (01:47 +0000)]
add hym8563 rtc

13 years agomod i2c
柯飞雄 [Thu, 13 May 2010 01:14:00 +0000 (01:14 +0000)]
mod i2c

13 years agomod i2c
柯飞雄 [Thu, 13 May 2010 01:11:20 +0000 (01:11 +0000)]
mod i2c

13 years agomod dm9000
刘益星 [Wed, 12 May 2010 15:34:37 +0000 (15:34 +0000)]
mod dm9000

13 years agomodify for dm9000
刘益星 [Wed, 12 May 2010 15:14:07 +0000 (15:14 +0000)]
modify for dm9000

13 years agomod i2c
柯飞雄 [Wed, 12 May 2010 10:06:12 +0000 (10:06 +0000)]
mod i2c

13 years agomod i2c
柯飞雄 [Wed, 12 May 2010 10:03:49 +0000 (10:03 +0000)]
mod i2c

13 years agomod i2c
柯飞雄 [Wed, 12 May 2010 09:59:41 +0000 (09:59 +0000)]
mod i2c

13 years agomod i2c
柯飞雄 [Wed, 12 May 2010 09:57:29 +0000 (09:57 +0000)]
mod i2c

13 years agodel i2c.h
柯飞雄 [Wed, 12 May 2010 09:51:53 +0000 (09:51 +0000)]
del i2c.h

13 years agoadd i2c
柯飞雄 [Wed, 12 May 2010 09:18:40 +0000 (09:18 +0000)]
add i2c

13 years agomod i2c
柯飞雄 [Wed, 12 May 2010 02:19:45 +0000 (02:19 +0000)]
mod i2c

13 years agoadd i2c
柯飞雄 [Tue, 11 May 2010 14:21:07 +0000 (14:21 +0000)]
add i2c

13 years agoadd i2c
柯飞雄 [Tue, 11 May 2010 14:19:49 +0000 (14:19 +0000)]
add i2c

13 years agoadd i2c
柯飞雄 [Tue, 11 May 2010 14:10:25 +0000 (14:10 +0000)]
add i2c

13 years agoadd i2c
柯飞雄 [Tue, 11 May 2010 14:07:00 +0000 (14:07 +0000)]
add i2c

13 years agoadd i2c
柯飞雄 [Tue, 11 May 2010 14:03:56 +0000 (14:03 +0000)]
add i2c

13 years agoadd i2c
柯飞雄 [Tue, 11 May 2010 14:03:38 +0000 (14:03 +0000)]
add i2c

13 years agoadd i2c
柯飞雄 [Tue, 11 May 2010 14:02:35 +0000 (14:02 +0000)]
add i2c

13 years agoadd spi driver
林辉辉 [Tue, 11 May 2010 07:34:46 +0000 (07:34 +0000)]
add spi driver

13 years agomap RK2818_SDRAMC_BASE
黄涛 [Tue, 11 May 2010 02:11:23 +0000 (02:11 +0000)]
map RK2818_SDRAMC_BASE

13 years agoadd scu.h
黄涛 [Tue, 11 May 2010 01:21:23 +0000 (01:21 +0000)]
add scu.h

13 years agomodify dwdma hardware connection
倪振宇 [Mon, 10 May 2010 13:52:39 +0000 (13:52 +0000)]
modify dwdma hardware connection

13 years agoenable TCM
黄涛 [Mon, 10 May 2010 08:46:32 +0000 (08:46 +0000)]
enable TCM

13 years agodisable dwdma testing switch
倪振宇 [Fri, 7 May 2010 11:58:00 +0000 (11:58 +0000)]
disable dwdma testing switch

13 years agomodify performance of dwdam interrupt method
倪振宇 [Fri, 7 May 2010 11:56:48 +0000 (11:56 +0000)]
modify performance of dwdam interrupt method

13 years agomodify method that get irq status
倪振宇 [Thu, 6 May 2010 08:21:12 +0000 (08:21 +0000)]
modify method that get irq status

13 years agomodify interrupt type
倪振宇 [Wed, 5 May 2010 03:30:46 +0000 (03:30 +0000)]
modify interrupt type

13 years agoupdate for gpio
沈睿汀 [Fri, 30 Apr 2010 02:52:30 +0000 (02:52 +0000)]
update for gpio

13 years agosupport clk_set_parent
黄涛 [Thu, 29 Apr 2010 14:19:17 +0000 (14:19 +0000)]
support clk_set_parent

13 years agoadd for gpio
沈睿汀 [Thu, 29 Apr 2010 09:23:23 +0000 (09:23 +0000)]
add for gpio

13 years agoupdate for gpio
沈睿汀 [Thu, 29 Apr 2010 09:19:35 +0000 (09:19 +0000)]
update for gpio

13 years agoupdate for gpio
沈睿汀 [Thu, 29 Apr 2010 09:12:25 +0000 (09:12 +0000)]
update for gpio

13 years agoupdate for gpio
沈睿汀 [Thu, 29 Apr 2010 09:10:42 +0000 (09:10 +0000)]
update for gpio

13 years agoupdate for gpio
沈睿汀 [Thu, 29 Apr 2010 09:07:44 +0000 (09:07 +0000)]
update for gpio

13 years agoupdate for gpio
沈睿汀 [Thu, 29 Apr 2010 09:06:03 +0000 (09:06 +0000)]
update for gpio

13 years agodelete macro test dma
倪振宇 [Wed, 28 Apr 2010 14:30:51 +0000 (14:30 +0000)]
delete macro test dma

13 years agoadd dwdma driver
倪振宇 [Wed, 28 Apr 2010 13:54:42 +0000 (13:54 +0000)]
add dwdma driver

13 years agoset zreladdr to 0x60008000
黄涛 [Tue, 27 Apr 2010 13:17:52 +0000 (13:17 +0000)]
set zreladdr to 0x60008000

13 years agoupdate debug code
黄涛 [Tue, 27 Apr 2010 12:48:00 +0000 (12:48 +0000)]
update debug code

13 years agoupdate .config can run ramdisk
林辉辉 [Tue, 27 Apr 2010 12:28:53 +0000 (12:28 +0000)]
update .config can run ramdisk

13 years agomore clock
黄涛 [Mon, 26 Apr 2010 13:13:44 +0000 (13:13 +0000)]
more clock

13 years agoadd iomux driver
林辉辉 [Mon, 26 Apr 2010 07:43:24 +0000 (07:43 +0000)]
add iomux driver

13 years agoupdate rk2818_serial
林辉辉 [Sat, 24 Apr 2010 06:33:04 +0000 (06:33 +0000)]
update rk2818_serial

13 years agofix clock name from uart_clk to uart, uartclk get use clk_get_rate
黄涛 [Sat, 24 Apr 2010 03:50:20 +0000 (03:50 +0000)]
fix clock name from uart_clk to uart, uartclk get use clk_get_rate

13 years agoadd clock support
黄涛 [Sat, 24 Apr 2010 03:48:44 +0000 (03:48 +0000)]
add clock support

13 years agoenable common clkdev
黄涛 [Fri, 23 Apr 2010 09:53:52 +0000 (09:53 +0000)]
enable common clkdev

13 years agoadd rk2818 serial and up iomap
林辉辉 [Fri, 23 Apr 2010 03:25:06 +0000 (03:25 +0000)]
add rk2818 serial and up iomap

13 years agork2818 for linux2.6.32
林辉辉 [Thu, 22 Apr 2010 04:26:22 +0000 (04:26 +0000)]
rk2818 for linux2.6.32

14 years agoImplement read(2) in ashmem driver
Bjorn Bringert [Thu, 15 Apr 2010 09:04:01 +0000 (10:04 +0100)]
Implement read(2) in ashmem driver

Bug: 2595601
Change-Id: I47c0016f594f9354fb8658ccb26e3d395bcb137b
Signed-off-by: Bjorn Bringert <bringert@android.com>
14 years agostaging: android: lowmemorykiller: Remove bitrotted codepath
San Mehat [Thu, 6 May 2010 22:43:46 +0000 (15:43 -0700)]
staging: android: lowmemorykiller: Remove bitrotted codepath

Now that we're murder-synchronous, this code path will never be
called (and if it does, it doesn't tell us anything useful other
than we killed a task that was already being killed by somebody
else but hadn't gotten its' signal yet)

Signed-off-by: San Mehat <san@google.com>
14 years agostaging: android: lowmemorykiller: Fix task_struct leak
San Mehat [Thu, 6 May 2010 22:40:07 +0000 (15:40 -0700)]
staging: android: lowmemorykiller: Fix task_struct leak

As it turns out, the CONFIG_PROFILING interfaces leak a
task struct if the notifier chain returns NOTIFY_OK.. doh.

This patch reworks lowmemkiller to use the new generic task
free notifier chain.

Signed-off-by: San Mehat <san@google.com>
14 years agosched: Add a generic notifier when a task struct is about to be freed
San Mehat [Thu, 6 May 2010 22:37:55 +0000 (15:37 -0700)]
sched: Add a generic notifier when a task struct is about to be freed

This patch adds a notifier which can be used by subsystems that may
be interested in when a task has completely died and is about to
have it's last resource freed.

  The Android lowmemory killer uses this to determine when a task
it has killed has finally given up its goods.

Signed-off-by: San Mehat <san@google.com>
14 years agobinder: Fix memory corruption via page aliasing
Christopher Lais [Sat, 1 May 2010 20:51:48 +0000 (15:51 -0500)]
binder: Fix memory corruption via page aliasing

binder_deferred_release was not unmapping the page from the buffer
before freeing it, causing memory corruption.  This only happened
when page(s) had not been freed by binder_update_page_range, which
properly unmaps the pages.

This only happens on architectures with VIPT aliasing.

To reproduce, create a program which opens, mmaps, munmaps, then closes
the binder very quickly.  This should leave a page allocated when the
binder is released.  When binder_deferrred_release is called on the
close, the page will remain mapped to the address in the linear
proc->buffer.  Later, we may map the same physical page to a different
virtual address that has different coloring, and this may cause
aliasing to occur.

PAGE_POISONING will greatly increase your chances of noticing any
problems.

Change-Id: I6941bf212881b8bf846bdfda43d3609c7ae4892e

Signed-off-by: Christopher Lais <chris+android@zenthought.org>
14 years agostaging: android: lowmemkiller: Substantially reduce overhead during reclaim
San Mehat [Wed, 5 May 2010 18:38:42 +0000 (11:38 -0700)]
staging: android: lowmemkiller: Substantially reduce overhead during reclaim

This patch optimizes lowmemkiller to not do any work when it has an outstanding
kill-request. This greatly reduces the pressure on the task_list lock
(improving interactivity), as well as improving the vmscan performance
when under heavy memory pressure (by up to 20x in tests).

Note: For this enhancement to work, you need CONFIG_PROFILING

Signed-off-by: San Mehat <san@google.com>
14 years ago[ARM] Do not call flush_cache_user_range with mmap_sem held
Dima Zavin [Wed, 28 Apr 2010 03:57:04 +0000 (20:57 -0700)]
[ARM] Do not call flush_cache_user_range with mmap_sem held

We can't be holding the mmap_sem while calling flush_cache_user_range
because the flush can fault. If we fault on a user address, the
page fault handler will try to take mmap_sem again. Since both places
acquire the read lock, most of the time it succeeds. However, if another
thread tries to acquire the write lock on the mmap_sem (e.g. mmap) in
between the call to flush_cache_user_range and the fault, the down_read
in do_page_fault will deadlock.

Also, since we really can't be holding the mmap_sem while calling
flush_cache_user_range AND vma is actually unused by the flush itself,
get rid of vma as an argument.

Change-Id: If55409bde41ad1060fa4fe7cbd4ac530d4d9a106
Signed-off-by: Dima Zavin <dima@android.com>
14 years agoUSB: android gadget: mass storage: Fix format issue for Vista Host
Velempati Chiranjeevi [Wed, 21 Oct 2009 04:23:10 +0000 (09:53 +0530)]
USB: android gadget: mass storage: Fix format issue for Vista Host

As part of formating the mass storage device, Host sends the INQUIRY
scsi command. As per the standard, the command length for this
command should be 6 bytes, whereas the Vista host sends 12 bytes.

When the command length of the command is not equal to the standard
length, the device sends a phase error as part of the status phase.
When the host receives a phase error, it re-enumerates, hence the
error.

If the command is INQUIRY, and the command length is 12 bytes,
treating this as a good command and not sending the phase error
to the host fixes this issue.

Signed-off-by: Velempati Chiranjeevi <c_cvelem@qualcomm.com>
Signed-off-by: Mike Lockwood <lockwood@android.com>
14 years agolowmemorykiller: Don't try to kill the same pid over and over
San Mehat [Mon, 26 Apr 2010 22:11:04 +0000 (15:11 -0700)]
lowmemorykiller: Don't try to kill the same pid over and over

  Under certain circumstances, a process can take awhile to
handle a sig-kill (especially if it's in a scheduler group with
a very low share ratio). When this occurs, lowmemkiller returns
to vmscan indicating the process memory has been freed - even
though the process is still waiting to die. Since the memory
hasn't actually freed, lowmemkiller is called again shortly after,
and picks the same process to die; regardless of the fact that
it has already been 'scheduled' to die and the memory has already
been reported to vmscan as having been freed.

  Solution is to check fatal_signal_pending() on the selected
task, and if it's already pending destruction return; indicating
to vmscan that no resources were freed on this pass.

Signed-off-by: San Mehat <san@google.com>
14 years agoStaging: android: binder: Create dedicated workqueue for binder deferred work
Arve Hjønnevåg [Thu, 22 Apr 2010 22:53:23 +0000 (15:53 -0700)]
Staging: android: binder: Create dedicated workqueue for binder deferred work

Some drivers flush the global workqueue when closed. This would deadlock if
the last reference to the file was released from the binder.

Change-Id: Ifdabc0b383fecb20836d1bbb9786c632402a14e1
Signed-off-by: Arve Hjønnevåg <arve@android.com>
14 years agoproc: pagemap: Hold mmap_sem during page walk
KAMEZAWA Hiroyuki [Fri, 2 Apr 2010 00:11:29 +0000 (09:11 +0900)]
proc: pagemap: Hold mmap_sem during page walk

In initial design, walk_page_range() was designed just for walking page
table and it didn't require mmap_sem.  Now, find_vma() etc..  are used
in walk_page_range() and we need mmap_sem around it.

This patch adds mmap_sem around walk_page_range().

Because /proc/<pid>/pagemap's callback routine use put_user(), we have
to get rid of it to do sane fix.

Changelog: 2010/Apr/2
 - fixed start_vaddr and end overflow
Changelog: 2010/Apr/1
 - fixed start_vaddr calculation
 - removed unnecessary cast.
 - removed unnecessary change in smaps.
 - use GFP_TEMPORARY instead of GFP_KERNEL

Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Matt Mackall <mpm@selenic.com>
Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Brian Swetland <swetland@google.com>
Cc: Dave Hansen <haveblue@us.ibm.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: San Mehat <san@google.com>
[ Fixed kmalloc failure return code as per Matt ]
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
14 years agoRevert "proc: pagemap: Hold mmap_sem during page walk"
San Mehat [Mon, 19 Apr 2010 14:45:33 +0000 (07:45 -0700)]
Revert "proc: pagemap: Hold mmap_sem during page walk"

This reverts commit 0bf551684025a5f0f3c9aac694cd5cf73706cb70.

Reverting for proper up-stream fix

Signed-off-by: San Mehat <san@google.com>
14 years agoStaging: android: timed_gpio: Properly discard invalid timeout values.
Mike Lockwood [Sat, 17 Apr 2010 16:01:35 +0000 (12:01 -0400)]
Staging: android: timed_gpio: Properly discard invalid timeout values.

The timed output device never previously checked the return value of sscanf,
resulting in an uninitialized int being passed to enable() if input value
was invalid.

Signed-off-by: Mike Lockwood <lockwood@android.com>
14 years agoUSB: gadget: f_mass_storage: Defer handling interface changes until it is safe.
Mike Lockwood [Thu, 15 Apr 2010 19:04:07 +0000 (15:04 -0400)]
USB: gadget: f_mass_storage: Defer handling interface changes until it is safe.

Pulling in some code from file_storage.c, we now handle interface changes
in do_set_config(), which is now not called until any pending requests have
been successfully completed or cancelled.

This fixes a race condition that resulted in usb_ep_free_request() being called
while the request is still busy.

Signed-off-by: Mike Lockwood <lockwood@android.com>
14 years agofs: partitions: Fix warnings in fs/partitions/check.c
Colin Cross [Wed, 7 Apr 2010 19:08:27 +0000 (12:08 -0700)]
fs: partitions: Fix warnings in fs/partitions/check.c

Change-Id: I4398ace0c55d4833b1fcbb7a4e71ab8f0b1b044a
Signed-off-by: Colin Cross <ccross@android.com>
14 years agoBluetooth: Hack: Don't dereference null pointer.
Nick Pelly [Thu, 8 Apr 2010 23:23:32 +0000 (16:23 -0700)]
Bluetooth: Hack: Don't dereference null pointer.

This avoids the S305 panic during incoming connection.

S305 sends PSM 25 L2CAP connection request before the L2CAP info response.
When we receive that info response we crash on null pointer here.

Bug: 2127637
Change-Id: Ib637516251f46fa9a9c87ac015dc2f27df5a27fd
Signed-off-by: Nick Pelly <npelly@google.com>
14 years agoUSB: gadget: android: mass_storage: Use spin_lock_irqsave/spin_unlock_irqrestore
Arve Hjønnevåg [Thu, 11 Dec 2008 04:01:15 +0000 (20:01 -0800)]
USB: gadget: android: mass_storage: Use spin_lock_irqsave/spin_unlock_irqrestore

The old code did not allways disable interrupts when called from thread
context, but tried to lock the same spinlock from interrupt context.

This was merged from a change to drivers/usb/function/mass_storage.c
in the android-msm-2.6.29 branch.

Signed-off-by: Arve Hjønnevåg <arve@android.com>
Signed-off-by: Mike Lockwood <lockwood@android.com>
14 years agoproc: pagemap: Hold mmap_sem during page walk
San Mehat [Tue, 30 Mar 2010 23:43:18 +0000 (16:43 -0700)]
proc: pagemap: Hold mmap_sem during page walk

If the mmap_sem is not held while we walk_page_range(), then
it is possible for find_vma() to race with a remove_vma_list()
caused by do_munmap() (or others).

Unable to handle kernel paging request at virtual address 6b6b6b5b
Internal error: Oops: 5 [#1] PREEMPT
CPU: 0    Not tainted  (2.6.32.9-27154-ge3e6e27 #1)
PC is at find_vma+0x40/0x7c
LR is at walk_page_range+0x70/0x230
pc : [<c00aa3ac>]    lr : [<c00b298c>]    psr: 20000013
sp : c6aa9eb8  ip : 6b6b6b53  fp : c6a58f60
r10: c7e1d1b8  r9 : 0001bca0  r8 : 47000000
r7 : c6aa9f80  r6 : c6aa8000  r5 : 46fbd000  r4 : 6b6b6b6b
r3 : c7ca4820  r2 : 6b6b6b6b  r1 : 46fbd000  r0 : c70e3e40
Flags: nzCv  IRQs on  FIQs on  Mode SVC_32  ISA ARM  Segment user
Control: 10c5787d  Table: 26574019  DAC: 00000015

[<c00aa3ac>] (find_vma+0x40/0x7c) from [<c00b298c>] (walk_page_range+0x70/0x230)
[<c00b298c>] (walk_page_range+0x70/0x230) from [<c00f5d3c>] (pagemap_read+0x1a4/0x278)
[<c00f5d3c>] (pagemap_read+0x1a4/0x278) from [<c00bac40>] (vfs_read+0xa8/0x150)
[<c00bac40>] (vfs_read+0xa8/0x150) from [<c00bad94>] (sys_read+0x3c/0x68)
[<c00bad94>] (sys_read+0x3c/0x68) from [<c0026f00>] (ret_fast_syscall+0x0/0x2c)
Code: 98bd8010 e5932004 e3a00000 ea000008 (e5124010)

Signed-off-by: San Mehat <san@google.com>
CC: Brian Swetland <swetland@google.com>
CC: Matt Mackall <mpm@selenic.com>
CC: Andrew Morton <akpm@linux-foundation.org>
CC: Linus Torvalds <torvalds@linux-foundation.org>
14 years agowlan: Extract generic wlan platform data from tiwlan specific header
Dmitry Shmidt [Thu, 18 Mar 2010 23:04:18 +0000 (16:04 -0700)]
wlan: Extract generic wlan platform data from tiwlan specific header

Signed-off-by: Dmitry Shmidt <dimitrysh@google.com>
14 years agoUSB: gadget: android: check for null _android_dev in android_register_function()
Mike Lockwood [Wed, 17 Mar 2010 19:42:29 +0000 (15:42 -0400)]
USB: gadget: android: check for null _android_dev in android_register_function()

This fixes a load ordering issue that occurred if a function driver loads before
the android gadget driver is initialized.

Signed-off-by: Mike Lockwood <lockwood@android.com>
14 years agoapanic: make APANIC_LABEL depend on APANIC
Iliyan Malchev [Wed, 17 Mar 2010 19:02:26 +0000 (12:02 -0700)]
apanic: make APANIC_LABEL depend on APANIC

Signed-off-by: Iliyan Malchev <malchev@google.com>
14 years agoUSB: gadget: f_mass_storage: Flush writes after every 4 MB.
Mike Lockwood [Fri, 12 Mar 2010 16:01:05 +0000 (11:01 -0500)]
USB: gadget: f_mass_storage: Flush writes after every 4 MB.

This avoids excessive caching at the block level layer when copying large
files to the storage device.

Signed-off-by: Mike Lockwood <lockwood@android.com>
14 years agoMerge commit 'v2.6.32.9' into android-2.6.32
Arve Hjønnevåg [Thu, 11 Mar 2010 00:38:33 +0000 (16:38 -0800)]
Merge commit 'v2.6.32.9' into android-2.6.32

14 years agoRevert "sysfs: Fix sysfs owner and group."
Arve Hjønnevåg [Thu, 11 Mar 2010 00:38:13 +0000 (16:38 -0800)]
Revert "sysfs: Fix sysfs owner and group."

This reverts commit f1f8e7cd6f796b15ed011361f09b57db796d9b4e.

14 years agoUSB: gadget: f_rndis: Add platform data for RNDIS vendor ID and MAC address.
Mike Lockwood [Wed, 10 Mar 2010 22:05:03 +0000 (17:05 -0500)]
USB: gadget: f_rndis: Add platform data for RNDIS vendor ID and MAC address.

This fixes a kernel panic in rndis.c when receiving the
OID_GEN_VENDOR_DESCRIPTION command.

Signed-off-by: Mike Lockwood <lockwood@android.com>
14 years agoBluetooth: Use non-flushable pb flag by default for ACL data on capable chipsets.
Nick Pelly [Wed, 9 Dec 2009 03:42:21 +0000 (19:42 -0800)]
Bluetooth: Use non-flushable pb flag by default for ACL data on capable chipsets.

With Bluetooth 2.1 ACL packets can be flushable or non-flushable. This commit
makes ACL data packets non-flushable by default on compatible chipsets, and
adds the L2CAP_LM_FLUSHABLE socket option to explicitly request flushable ACL
data packets for a given L2CAP socket. This is useful for A2DP data which can
be safely discarded if it can not be delivered within a short time (while
other ACL data should not be discarded).

Note that making ACL data flushable has no effect unless the automatic flush
timeout for that ACL link is changed from its default of 0 (infinite).

Change-Id: Ie3d4befdeaefb8c979de7ae603ff5ec462b3483c
Signed-off-by: Nick Pelly <npelly@google.com>
14 years agoRevert "Bluetooth: Introduce L2CAP_LM_FLUSHABLE to allow flushing of ACL packets."
Nick Pelly [Tue, 9 Mar 2010 20:08:00 +0000 (12:08 -0800)]
Revert "Bluetooth: Introduce L2CAP_LM_FLUSHABLE to allow flushing of ACL packets."

This reverts commit d7897fd1e9fb3a5df0740dc2dc45ec94ca0965f2.

Change-Id: I3401550b6dc97b683104e9fdac30a617a2db8c8e
Signed-off-by: Nick Pelly <npelly@google.com>
14 years agopmem: Add cache flush ioctl for pmem buffers
Dima Zavin [Tue, 2 Mar 2010 23:47:57 +0000 (15:47 -0800)]
pmem: Add cache flush ioctl for pmem buffers

Change-Id: I9156bad829e8c65087f122b48cc57638902fab12
Signed-off-by: Dima Zavin <dima@android.com>
14 years agoUSB: gadget: composite: Don't increment interface number for alt settings.
Mike Lockwood [Fri, 26 Feb 2010 14:34:19 +0000 (09:34 -0500)]
USB: gadget: composite: Don't increment interface number for alt settings.

Signed-off-by: Mike Lockwood <lockwood@android.com>
14 years agoUSB: gadget: composite: Don't call set_alt() on functions that are hidden.
Mike Lockwood [Fri, 26 Feb 2010 14:30:01 +0000 (09:30 -0500)]
USB: gadget: composite: Don't call set_alt() on functions that are hidden.

Signed-off-by: Mike Lockwood <lockwood@android.com>
14 years agoBluetooth: Allow SCO/eSCO packet type selection for outgoing SCO connections.
Nick Pelly [Thu, 11 Feb 2010 19:54:28 +0000 (11:54 -0800)]
Bluetooth: Allow SCO/eSCO packet type selection for outgoing SCO connections.

__u16 sco_pkt_type is introduced to struct sockaddr_sco. It allows bitwise
selection of SCO/eSCO packet types. Currently those bits are:

0x0001 HV1 may be used.
0x0002 HV2 may be used.
0x0004 HV3 may be used.
0x0008 EV3 may be used.
0x0010 EV4 may be used.
0x0020 EV5 may be used.
0x0040 2-EV3 may be used.
0x0080 3-EV3 may be used.
0x0100 2-EV5 may be used.
0x0200 3-EV5 may be used.

This is similar to the Packet Type parameter in the HCI Setup Synchronous
Connection Command, except that we are not reversing the logic on the EDR bits.
This makes the use of sco_pkt_tpye forward portable for the use case of
white-listing packet types, which we expect will be the primary use case.

If sco_pkt_type is zero, or userspace uses the old struct sockaddr_sco,
then the default behavior is to allow all packet types.

Packet type selection is just a request made to the Bluetooth chipset, and
it is up to the link manager on the chipset to negiotiate and decide on the
actual packet types used. Furthermore, when a SCO/eSCO connection is eventually
made there is no way for the host stack to determine which packet type was used
(however it is possible to get the link type of SCO or eSCO).

sco_pkt_type is ignored for incoming SCO connections. It is possible
to add this in the future as a parameter to the Accept Synchronous Connection
Command, however its a little trickier because the kernel does not
currently preserve sockaddr_sco data between userspace calls to accept().

The most common use for sco_pkt_type will be to white-list only SCO packets,
which can be done with the hci.h constant SCO_ESCO_MASK.

This patch is motivated by broken Bluetooth carkits such as the Motorolo
HF850 (it claims to support eSCO, but will actually reject eSCO connections
after 5 seconds) and the 2007/2008 Infiniti G35/37 (fails to route audio
if a 2-EV5 packet type is negiotiated). With this patch userspace can maintain
a list of compatible packet types to workaround remote devices such as these.

Based on a patch by Marcel Holtmann.

Change-Id: I304d8fda5b4145254820a3003820163bf53de5a5
Signed-off-by: Nick Pelly <npelly@google.com>
14 years agoRevert "Bluetooth: Change RFCOMM to use BT_CONNECT2 for BT_DEFER_SETUP"
Nick Pelly [Wed, 24 Feb 2010 17:29:45 +0000 (09:29 -0800)]
Revert "Bluetooth: Change RFCOMM to use BT_CONNECT2 for BT_DEFER_SETUP"

This reverts commit 8bf4794174659b06d43cc5e290cd384757374613.

Change-Id: Ieaeed6866996446e0392e387b77b446361f23d46
Signed-off-by: Nick Pelly <npelly@google.com>
14 years agoRevert "Bluetooth: Fix removing of RFCOMM DLC timer with DEFER_SETUP"
Nick Pelly [Wed, 24 Feb 2010 17:29:10 +0000 (09:29 -0800)]
Revert "Bluetooth: Fix removing of RFCOMM DLC timer with DEFER_SETUP"

This reverts commit e2139b32726e5dd184974c785ea3f62026590801.

Change-Id: Id948d5860840aaf85eca7d8894a7bb3192f0bb47
Signed-off-by: Nick Pelly <npelly@google.com>
14 years agoUSB: gadget: android: Disable UMS when RNDIS ethernet is active.
Mike Lockwood [Wed, 24 Feb 2010 15:20:59 +0000 (10:20 -0500)]
USB: gadget: android: Disable UMS when RNDIS ethernet is active.

Signed-off-by: Mike Lockwood <lockwood@android.com>
14 years agoandroid: logger: Add new system log for framework/system log messages
San Mehat [Wed, 24 Feb 2010 00:09:47 +0000 (16:09 -0800)]
android: logger: Add new system log for framework/system log messages

Signed-off-by: San Mehat <san@google.com>
14 years agoLinux 2.6.32.9
Greg Kroah-Hartman [Tue, 23 Feb 2010 15:38:51 +0000 (07:38 -0800)]
Linux 2.6.32.9

14 years agodm log: userspace fix overhead_size calcuations
Jonathan Brassow [Tue, 16 Feb 2010 18:42:53 +0000 (18:42 +0000)]
dm log: userspace fix overhead_size calcuations

commit ebfd32bba9b518d684009d9d21a56742337ca1b3 upstream.

This patch fixes two bugs that revolve around the miscalculation and
misuse of the variable 'overhead_size'.  'overhead_size' is the size of
the various header structures used during communication.

The first bug is the use of 'sizeof' with the pointer of a structure
instead of the structure itself - resulting in the wrong size being
computed.  This is then used in a check to see if the payload
(data_size) would be to large for the preallocated structure.  Since the
bug produces a smaller value for the overhead, it was possible for the
structure to be breached.  (Although the current users of the code do
not currently send enough data to trigger this bug.)

The second bug is that the 'overhead_size' value is used to compute how
much of the preallocated space should be cleared before populating it
with fresh data.  This should have simply been 'sizeof(struct cn_msg)'
not overhead_size.  The fact that 'overhead_size' was computed
incorrectly made this problem "less bad" - leaving only a pointer's
worth of space at the end uncleared.  Thus, this bug was never producing
a bad result, but still needs to be fixed - especially now that the
value is computed correctly.

Signed-off-by: Jonathan Brassow <jbrassow@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agodm stripe: avoid divide by zero with invalid stripe count
Nikanth Karthikesan [Tue, 16 Feb 2010 18:42:47 +0000 (18:42 +0000)]
dm stripe: avoid divide by zero with invalid stripe count

commit 781248c1b50c776a9ef4be1130f84ced1cba42fe upstream.

If a table containing zero as stripe count is passed into stripe_ctr
the code attempts to divide by zero.

This patch changes DM_TABLE_LOAD to return -EINVAL if the stripe count
is zero.

We now get the following error messages:
  device-mapper: table: 253:0: striped: Invalid stripe count
  device-mapper: ioctl: error adding target to table

Signed-off-by: Nikanth Karthikesan <knikanth@suse.de>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoiwlwifi: Fix to set correct ht configuration
Wey-Yi Guy [Fri, 5 Feb 2010 19:40:00 +0000 (11:40 -0800)]
iwlwifi: Fix to set correct ht configuration

commit 098dfded5b1b09927995e89c6d689f85a0f53384 upstream.

iwl_set_rxon_ht() only get called in iwl_post_associate(); which cause
possible incorrect ht configuration. Adding the call in iwl_mac_config() if
IEEE80211_CONF_CHANGE_CHANNEL flag is set to re-configure and send rxon
command.

Fixes
http://bugzilla.intellinuxwireless.org/show_bug.cgi?id=2146

Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>
Signed-off-by: Reinette Chatre <reinette.chatre@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agomac80211: Fix probe request filtering in IBSS mode
Benoit Papillault [Fri, 5 Feb 2010 00:21:03 +0000 (01:21 +0100)]
mac80211: Fix probe request filtering in IBSS mode

commit 0da780c269957783d341fc3559e6b4c9912af7b4 upstream.

We only reply to probe request if either the requested SSID is the
broadcast SSID or if the requested SSID matches our own SSID. This
latter case was not properly handled since we were replying to different
SSID with the same length as our own SSID.

Signed-off-by: Benoit Papillault <benoit.papillault@free.fr>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoath9k: Fix sequence numbers for PAE frames
Sujith [Tue, 9 Feb 2010 04:37:00 +0000 (10:07 +0530)]
ath9k: Fix sequence numbers for PAE frames

commit 6c8afef551fef87a3bf24f8a74c69a7f2f72fc82 upstream.

Currently, PAE frames are not assigned proper sequence numbers.
Since sending PAE frames as part of aggregates breaks
crupto with several APs, they are sent as normal MPDUs.
Fix the seqeuence number issue by updating the frame with the
internal sequence number.

Tested-by: Felix Fietkau <nbd@openwrt.org>
Signed-off-by: Sujith <Sujith.Manoharan@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agob43: Fix throughput regression
Larry Finger [Tue, 2 Feb 2010 16:08:19 +0000 (10:08 -0600)]
b43: Fix throughput regression

commit b6c3f5be7c6ac3375f44de4545c1ffe216b34022 upstream.

Commit c7ab5ef9bcd281135c21b4732c9be779585181be entitled "b43: implement
short slot and basic rate handling" reduced the transmit throughput for
my BCM4311 device from 18 Mb/s to 0.7 Mb/s. The basic rate handling
portion is OK, the problem is in the short slot handling.

Prior to this change, the short slot enable/disable routines were never
called. Experimentation showed that the critical part was changing the
value at offset 0x0010 in the shared memory. This is supposed to contain
the 802.11 Slot Time in usec, but if it is changed from its initial value
of zero, performance is destroyed. On the other hand, changing the value
in the MMIO register corresponding to the Interframe Slot Time increased
performance from 18 to 22 Mb/s. A BCM4306/3 also shows dramatic
improvement of the transmit rate from 5.3 to 19.0 Mb/s.

Other changes in the patch include removal of the magic number for the
MMIO register, and allowing the slot time to be set for any PHY operating
in the 2.4 GHz band. Previously, the routine was executed only for G PHYs.

Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoeCryptfs: Add getattr function
Tyler Hicks [Wed, 4 Nov 2009 08:48:01 +0000 (02:48 -0600)]
eCryptfs: Add getattr function

commit f8f484d1b6677dd5cd5e7e605db747e8c30bbd47 upstream.

The i_blocks field of an eCryptfs inode cannot be trusted, but
generic_fillattr() uses it to instantiate the blocks field of a stat()
syscall when a filesystem doesn't implement its own getattr().  Users
have noticed that the output of du is incorrect on newly created files.

This patch creates ecryptfs_getattr() which calls into the lower
filesystem's getattr() so that eCryptfs can use its kstat.blocks value
after calling generic_fillattr().  It is important to note that the
block count includes the eCryptfs metadata stored in the beginning of
the lower file plus any padding used to fill an extent before
encryption.

https://bugs.launchpad.net/ecryptfs/+bug/390833

Reported-by: Dominic Sacré <dominic.sacre@gmx.de>
Signed-off-by: Tyler Hicks <tyhicks@linux.vnet.ibm.com>
Cc: Tim Gardner <timg@tpi.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoNFS: Too many GETATTR and ACCESS calls after direct I/O
Chuck Lever [Mon, 15 Feb 2010 17:19:53 +0000 (12:19 -0500)]
NFS: Too many GETATTR and ACCESS calls after direct I/O

commit 65d269538a1129495ac45a14a777cd11cfe881d8 upstream.

The cached read and write paths initialize fattr->time_start in their
setup procedures.  The value of fattr->time_start is propagated to
read_cache_jiffies by nfs_update_inode().  Subsequent calls to
nfs_attribute_timeout() will then use a good time stamp when
computing the attribute cache timeout, and squelch unneeded GETATTR
calls.

Since the direct I/O paths erroneously leave the inode's
fattr->time_start field set to zero, read_cache_jiffies for that inode
is set to zero after any direct read or write operation.  This
triggers an otw GETATTR or ACCESS call to update the file's attribute
and access caches properly, even when the NFS READ or WRITE replies
have usable post-op attributes.

Make sure the direct read and write setup code performs the same fattr
initialization as the cached I/O paths to prevent unnecessary GETATTR
calls.

This was likely introduced by commit 0e574af1 in 2.6.15, which appears
to add new nfs_fattr_init() call sites in the cached read and write
paths, but not in the equivalent places in fs/nfs/direct.c.  A
subsequent commit in the same series, 33801147, introduces the
fattr->time_start field.

Interestingly, the direct write reschedule path already has a call to
nfs_fattr_init() in the right place.

Reported-by: Quentin Barnes <qbarnes@yahoo-inc.com>
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agodrm/radeon/kms: use udelay for short delays
Dave Airlie [Sat, 30 Jan 2010 21:07:14 +0000 (07:07 +1000)]
drm/radeon/kms: use udelay for short delays

commit 01d4503968f471f876fb44335800d2cf8dc5a2ce upstream.

For usec delays use udelay instead of scheduling, this should
allow reclocking to happen faster. This also was the cause
of reported 33s delays at bootup on certain systems.

fixes: freedesktop.org bug 25506

Signed-off-by: Dave Airlie <airlied@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoACPI: fix High cpu temperature with 2.6.32
Arjan van de Ven [Wed, 27 Jan 2010 23:25:39 +0000 (15:25 -0800)]
ACPI: fix High cpu temperature with 2.6.32

commit 370d5cd88509b93b76eb2f5f97efbd71c25061cb upstream.

Since the rewrite of the CPU idle governor in 2.6.32, two laptops have
surfaced where the BIOS advertises a C2 power state, but for some reason
this state is not functioning (as verified in both cases by powertop
before the patch in .32).

The old governor had the accidental behavior that if a non-working state
was chosen too many times, it would end up falling back to C1.  The new
governor works differently and this accidental behavior is no longer
there; the result is a high temperature on these two machines.

This patch adds these 2 machines to the DMI table for C state anomalies;
by just not using C2 both these machines are better off (the TSC can be
used instead of the pm timer, giving a performance boost for example).

Addresses http://bugzilla.kernel.org/show_bug.cgi?id=14742

Signed-off-by: Arjan van de Ven <arjan@linux.intel.com>
Reported-by: <akwatts@ymail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Len Brown <len.brown@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoACPI: Add NULL pointer check in acpi_bus_start
Thomas Renninger [Fri, 29 Jan 2010 16:48:51 +0000 (17:48 +0100)]
ACPI: Add NULL pointer check in acpi_bus_start

commit d2f6650a950dadd20667a04a9dc785f240d43695 upstream.

If acpi_bus_add does not return a device and it's passed
to acpi_bus_start, bad things will happen:

BUG: unable to handle kernel NULL pointer dereference at 0000000000000008
IP: [<ffffffff8128402d>] acpi_bus_start+0x14/0x24
...
[<ffffffffa008977a>] acpiphp_bus_add+0xba/0x130 [acpiphp]
[<ffffffffa008aa72>] enable_device+0x132/0x2ff [acpiphp]
[<ffffffffa0089b68>] acpiphp_enable_slot+0xb8/0x130 [acpiphp]
[<ffffffffa0089df7>] handle_hotplug_event_func+0x87/0x190 [acpiphp]

Next patch would make this NULL pointer check obsolete, but
better having one more than one missing...

Signed-off-by: Thomas Renninger <trenn@suse.de>
Acked-by: Bjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: Len Brown <len.brown@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agortl8187: Add new device ID
Jelle Martijn Kok [Wed, 10 Feb 2010 15:34:09 +0000 (09:34 -0600)]
rtl8187: Add new device ID

commit 174b24963eaf96dc5e093502ee09639aed13eb2f upstream.

Add new RTL8187B device.

Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoUSB: usbfs: properly clean up the as structure on error paths
Linus Torvalds [Tue, 16 Feb 2010 20:35:07 +0000 (12:35 -0800)]
USB: usbfs: properly clean up the as structure on error paths

commit ddeee0b2eec2a51b0712b04de4b39e7bec892a53 upstream.

I notice that the processcompl_compat() function seems to be leaking the
'struct async *as' in the error paths.

I think that the calling convention is fundamentally buggered. The
caller is the one that did the "reap_as()" to get the as thing, the
caller should be the one to free it too.

Freeing it in the caller also means that it very clearly always gets
freed, and avoids the need for any "free in the error case too".

From: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Alan Stern <stern@rowland.harvard.edu>
Cc: Marcus Meissner <meissner@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoUSB: usbfs: only copy the actual data received
Greg KH [Mon, 15 Feb 2010 17:37:46 +0000 (09:37 -0800)]
USB: usbfs: only copy the actual data received

commit d4a4683ca054ed9917dfc9e3ff0f7ecf74ad90d6 upstream.

We need to only copy the data received by the device to userspace, not
the whole kernel buffer, which can contain "stale" data.

Thanks to Marcus Meissner for pointing this out and testing the fix.

Reported-by: Marcus Meissner <meissner@suse.de>
Tested-by: Marcus Meissner <meissner@suse.de>
Cc: Alan Stern <stern@rowland.harvard.edu>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoclass: Free the class private data in class_release
Laurent Pinchart [Wed, 10 Feb 2010 12:32:49 +0000 (13:32 +0100)]
class: Free the class private data in class_release

commit 18d19c96457d172d913510c083bc7411ed40cb10 upstream.

Fix a memory leak by freeing the memory allocated in __class_register
for the class private data.

Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Acked-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>