In LP1 suspend, or during a failed LP0 suspend, the core power is
not turned off and the timer register is not reset. Saving the
timer value during suspend and adding it to the offset after resume
will cause the timer value to double for each suspend cycle.
To fix, when resuming subtract the current value of the timer
register from the offset. Also, use the rtc registers to add
the time that passed during suspend.
Change-Id: I9f7ee2089b98cf66af45271f8478fbd9eb2ee250
Signed-off-by: Colin Cross <ccross@android.com>
The 32k clocksource is never used, because it is always registered
at the same time as the us clocksource, which has a higher rating.
Using the 32k clocksource would cause a race condition with the
tegra_rtc driver on the RTC shadow registers. Remove the 32k
clocksource, add a replacement tegra_read_rtc_ms function for the
suspend code to track suspend time, and add a comment on
read_persistent_clock that it should never be called while the
tegra_rtc driver is running. It is currently only called from
timekeeping init, suspend, and resume.
Change-Id: Ic964578e9923b0f0338959f75de1ac170f850337
Signed-off-by: Colin Cross <ccross@android.com>
Turn on the cpuidle_device power_specified field
to enable usage of the defined cpuidle_state.power_usage fields
for each state.
Signed-off-by: James Wylder <james.wylder@motorola.com>
Change-Id: Ibca20a937bf4079f61e4b3ae64394d797b4eb287
Signed-off-by: Todd Poynor <toddpoynor@google.com>
Some peripheral clocks share enable bits. Refcount the enables so
that calling clk_disable on one clock will not turn off another
clock.
Change-Id: Iefb015bc83cac77b3faef8db86cfc42b25c0b6ae
Signed-off-by: Colin Cross <ccross@android.com>
New and improved dvfs:
Registered dynamically during init
Exports dvfs functions to control clocks that are not visible to
the clock subsystem
Supports multiple regulators per clock
Fix dvfs on disabled clocks
Adds /d/clock/dvfs to show current voltage requirements
Change-Id: I93794a7761dccc702566e8850bb79f344ff787a2
Signed-off-by: Colin Cross <ccross@android.com>
During cpu frequency changes, take an extra reference to pllx so
that it doesn't turn off and on while the cpu is on pllp.
Change-Id: I574d399a67aab21f99473296a29aa1eedc0eedb9
Signed-off-by: Colin Cross <ccross@android.com>
The PLL lock bits are not reliable, use per-PLL timeouts instead.
Change-Id: I2749e0d8bcfb0b60dbbcf6378b38307f77ff8d4d
Signed-off-by: Colin Cross <ccross@android.com>
Move dc init to the device init level so that all clocks are
initialized before the late init level. This depends on I2C
being initialized in the subsys init level.
Change-Id: Iafe759d943b6cc90d5c6718b69e872e14d6acd4e
Signed-off-by: Colin Cross <ccross@android.com>
Some drivers (like hdmi) need to communicate over i2c during probe.
Move the i2c bus init to the subsys init level so that i2c is ready
for the start of the device init level.
Change-Id: I81234d42aa26b657ffc619099e47df09e12083bb
Signed-off-by: Colin Cross <ccross@android.com>
Due to conflicting restrictions on the location of the framebuffer
between the bootloader and the protected aperture, the framebuffer
is likely to need to be moved during boot. This patch provides
tegra_move_framebuffer, which can handle move the framebuffer from
lowmem, highmem, or unmapped memory into unmapped memory.
Change-Id: Ic37e5e337cd3129065fe56fd7777a86d06ad69ac
This uses the patch by rmk to allow memblock_remove to be used to
remove areas of memory from the 1:1 mapping, allowing them to be
remapped later using iomap or, for Tegra, nvmap.
Also uses memblock_reserve to reserve the lp0 boot vector, so it
doesn't need to be copied later in tegra_init_suspend.
tegra_reserve should be called from the machine reserve callback
after any board-specific memory areas are reserved.
Change-Id: I26be8544a03b6da74fe66dc53a77681d431c303c
Signed-off-by: Colin Cross <ccross@android.com>
Based on patch by rmk on lkml at http://lkml.org/lkml/2010/10/11/179
Reverts changes to find_limits to fix crash when using memblock_remove
on the end of memory.
Original-author: Russell King <rmk+kernel@arm.linux.org.uk>
Signed-off-by: Colin Cross <ccross@android.com>
Change-Id: I6137a7939329381e4ed34bfcdc8b713dc50ebcc8
The framebuffer driver needs to be able to arbitrarily pin whatever
gets handed to it. Regardless of the interface used, functions need
to unpin as soon as they finish using the gart anyway.
Change-Id: Ida8aea2fb6eaca8bcbf3ae72f8dfa849dc198542
Signed-off-by: Rebecca Schultz Zavin <rebecca@android.com>
Support mono data formats such as DSP PCM Mode with 16 bit mono capture.
This patch also disables the in-driver downsampler.
Signed-off-by: Iliyan Malchev <malchev@google.com>
* 'upstream' of git://git.linux-mips.org/pub/scm/upstream-linus:
MIPS: O32 compat/N32: Fix to use compat syscall wrappers for AIO syscalls.
MAINTAINERS: Change list for ioc_serial to linux-serial.
SERIAL: ioc3_serial: Return -ENOMEM on memory allocation failure
MIPS: jz4740: Fix Kbuild Platform file.
MIPS: Repair Kbuild make clean breakage.
If the host is slow in reading data or doesn't read data at all,
blocking write calls not only blocked the program that called write()
but the entire guest itself.
To overcome this, let's not block till the host signals it has given
back the virtio ring element we passed it. Instead, send the buffer to
the host and return to userspace. This operation then becomes similar
to how non-blocking writes work, so let's use the existing code for this
path as well.
This code change also ensures blocking write calls do get blocked if
there's not enough room in the virtio ring as well as they don't return
-EAGAIN to userspace.
Signed-off-by: Amit Shah <amit.shah@redhat.com>
Acked-by: Hans de Goede <hdegoede@redhat.com>
CC: stable@kernel.org
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
