Forward port of ashmem to 2.6.27.
Signed-off-by: Robert Love <rlove@google.com>
ashmem: Don't install fault handler for private mmaps.
Ashmem is used to create named private heaps. If this heap is backed
by a tmpfs file it will allocate two pages for every page touched.
In 2.6.27, the extra page would later be freed, but 2.6.29 does not
scan anonymous pages when running without swap so the memory is not
freed while the file is referenced. This change changes the behavior
of private ashmem mmaps to match /dev/zero instead tmpfs.
Signed-off-by: Arve Hjønnevåg <arve@android.com>
ashmem: Add common prefix to name reported in /proc/pid/maps
Signed-off-by: Arve Hjønnevåg <arve@android.com>
ashmem: don't require a page aligned size
This makes ashmem more similar to shmem and mmap, by
not requiring the specified size to be page aligned,
instead rounding it internally as needed.
Signed-off-by: Marco Nelissen <marcone@android.com>
Signed-off-by: Rebecca Schultz <rschultz@google.com>
pmem: Use the thread group leader insted of the current thread.
Instead of keeping track of the current thread, use the thread group leader
Signed-off-by: Rebecca Schultz <rschultz@google.com>
pmem: Add some apis to reference and flush pmem files by file struct
The api to refer to pmem files by fd should be depricated, it can
cause problems if a processes fd table changes while the kernel is processing
data in a pmem file. This change adds the safer api.
Signed-off-by: Rebecca Schultz Zavin <rebecca@android.com>
pmem: Remove unused depricated fd api to pmem.
Signed-off-by: Rebecca Schultz Zavin <rebecca@android.com>
pmem: Remove error message when calling get_pmem_addr
This call is used from the mdp driver to determine if the memory
is in pmem or in the fb. We will encounter this case during normal operation
so this error message should be removed.
Signed-off-by: Rebecca Schultz Zavin <rebecca@android.com>
pmem: Add include sched.h to fix compile errors
Signed-off-by: Rebecca Schultz Zavin <rebecca@android.com>
pmem: remove HW3D_* ioctls
Signed-off-by: Dima Zavin <dima@android.com>
pmem: Expose is_pmem_file to the in-kernel users.
Signed-off-by: Dima Zavin <dima@android.com>
pmem: Make the exposed functions be noops if CONFIG_ANDROID_PMEM is not set.
Signed-off-by: Dima Zavin <dima@android.com>
misc: pmem: don't flush if file was opened with O_SYNC
Change-Id: I067218658a0d7f7ecc1fe73e9ff6b0c3b3054653
Signed-off-by: Dima Zavin <dima@android.com>
When connecting usb or the charger the device would often go back to sleep
before the charge led and screen turned on.
Signed-off-by: Arve Hjønnevåg <arve@android.com>
Since wall time can jump backwards, it cannot be used to determine if one
event occured before another or for how long a key was pressed.
Signed-off-by: Arve Hjønnevåg <arve@android.com>
When we resume we only know how many whole seconds has elapsed.
These errors would accumulate in delta. We now only set the delta
if it would change by more than two seconds. If we drift back by
by more than a second add one in resume.
Signed-off-by: Arve Hjønnevåg <arve@android.com>
Drivers can now create alarms that will use an hrtimer while the
system is running and the rtc to wake up from suspend.
Signed-off-by: Arve Hjønnevåg <arve@android.com>
Rather than signaling a full update of the display from userspace via a
console switch, this patch introduces 2 files int /sys/power,
wait_for_fb_sleep and wait_for_fb_wake. Reading these files will block
until the requested state has been entered. When a read from
wait_for_fb_sleep returns userspace should stop drawing. When
wait_for_fb_wake returns, it should do a full update. If either are called
when the fb driver is already in the requested state, they will return
immediately.
Signed-off-by: Rebecca Schultz <rschultz@google.com>
Signed-off-by: Arve Hjønnevåg <arve@android.com>
This adds /sys/power/wake_lock and /sys/power/wake_unlock.
Writing a string to wake_lock creates a wake lock the
first time is sees a string and locks it. Optionally, the
string can be followed by a timeout.
To unlock the wake lock, write the same string to wake_unlock.
If EARLYSUSPEND is enabled then writes to /sys/power/state no longer
blocks, and the kernel will try to enter the requested state every
time no wakelocks are held. Write "on" to resume normal operation.
PM: wakelock: Replace expire work with a timer
The expire work function did not work in the normal case.
Signed-off-by: Arve Hjønnevåg <arve@android.com>
The search/pullup/add/remove device attributes were 0666 which would
allow arbitrary users to affect the 1 wire bus. Change to 0664 to
prevent that.
Signed-off-by: Brian Swetland <swetland@google.com>
If CACHE_FLUSH_RANGE_LIMIT is defined, then the entire dcache will
be flushed if the requested range is larger than this limit.
Signed-off-by: Arve Hjønnevåg <arve@android.com>
This improves futex performance until our user-space code is fixed to use
FUTEX_PRIVATE_FLAG for non-shared futexes.
Signed-off-by: Arve Hjønnevåg <arve@android.com>
This fixes a problem where the first pageblock got marked MIGRATE_RESERVE even
though it only had a few free pages. This in turn caused no contiguous memory
to be reserved and frequent kswapd wakeups that emptied the caches to get more
contiguous memory.
By default the kernel tries to keep half as much memory free at each
order as it does for one order below. This can be too agressive when
running without swap.
Signed-off-by: Arve Hjønnevåg <arve@android.com>
First call to sysfs_setattr ignored the requested owner and group.
Change-Id: I8343f818c65c54f6dbb55533cd34b31eeba5b554
Signed-off-by: Arve Hjønnevåg <arve@android.com>
83f9ac removed a call to effective_prio() in wake_up_new_task(), which
leads to tasks running at MAX_PRIO.
This is caused by the idle thread being set to MAX_PRIO before forking
off init. O(1) used that to make sure idle was always preempted, CFS
uses check_preempt_curr_idle() for that so we can savely remove this bit
of legacy code.
Reported-by: Mike Galbraith <efault@gmx.de>
Tested-by: Mike Galbraith <efault@gmx.de>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1259754383.4003.610.camel@laptop>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
commit 1152dcc28c upstream
Similar to 6000 and 1000 series, RTS/CTS is the recommended protection
mechanism for 5000 series in HT mode based on the HW design.
Using RTS/CTS will better protect the inner exchange from interference,
especially in highly-congested environment, it also prevent uCode encounter
TX FIFO underrun and other HT mode related performance issues.
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>
upstream in 2.6.33-rc: 5d76b6f6c1
Refreshed here for 2.6.32.y, applies w/ offset back to 2.6.29.y.
Linux has always ignored ACPI BIOS C2 with exit latency > 100 usec,
and the ACPI spec is clear that is correct FADT-supplied C2.
However, the ACPI spec explicitly states that _CST-supplied C-states
have no latency limits.
So move the 100usec C2 test out of the code shared
by FADT and _CST code-paths, and into the FADT-specific path.
This bug has not been visible until Nehalem, which advertises
a CPU-C2 worst case exit latency on servers of 205usec.
That (incorrect) figure is being used by BIOS writers
on mobile Nehalem systems for the AC configuration.
Thus, Linux ignores C2 leaving just C1, which is
saves less power, and also impacts performance
by preventing the use of turbo mode.
http://bugzilla.kernel.org/show_bug.cgi?id=15064
Tested-by: Alex Chiang <achiang@hp.com>
Signed-off-by: Len Brown <len.brown@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
