Device-specific headers belong in driver/ or chip/. The include/
directory should be for common interfaces.
Code should not normally need to include driver-specific headers. If
it does, it should use the full relative path from the EC project root
(for example, drivers/charger/bq24715.h).
Change-Id: Id23db37a431e2d802a74ec601db6f69b613352ba
Signed-off-by: Randall Spangler <rspangler@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/173746
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
The EC LPC implementation guarantees that the status byte will have at
least one zero bit, so there's no need to scan the parameter space as
well. Removing this unneeded check will slightly speed up ectool.
BUG=chrome-os-partner:10963
BRANCH=none
TEST=on an x86 chromebook (e.g. link), ectool hello still works
iotools io_read8 0x200 && iotools io_read8 0x204 -> not both 0xff
Change-Id: Ic02ca0ee686ab10e50093807717ec638aaa468c6
Signed-off-by: Randall Spangler <rspangler@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/174059
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
Some of the comments no longer apply. Others needed more info.
No code changes; just comment changes.
BUG=none
BRANCH=none
TEST=build all platforms
Change-Id: I1d52aa9a98427a78c9d9a8cf44934fb04c3c00c8
Signed-off-by: Randall Spangler <rspangler@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/174084
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
Changed the low power idle task to use the low speed clock in deep
sleep. The low power idle task is currently only enabled for Peppy,
Slippy, and Falco. This change decreases power consumption when
the AP is not running.
Note that the low speed clock is slow enough that the JTAG cannot be
used and the EC console UART cannot be used. To work around that,
this commit detects when the JTAG is in use and when the EC console
is in use, and will not use the low speed clock if either is in use.
The JTAG in use never clears after being set and the console in use
clears after a fixed timeout period.
BUG=None
BRANCH=None
TEST=Passes all unit tests.
Tested that the EC console works when in deep sleep.
Tested that it is possible to run flash_ec when in deep sleep and
using the low speed clock.
Change-Id: Ia65997eb8e607a5df9b2c7d68e4826bfb1e0194c
Signed-off-by: Alec Berg <alecaberg@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/173326
Reviewed-by: Vincent Palatin <vpalatin@chromium.org>
This console command is useful during bringup to interrogate
devices on the EC I2C interfaces.
Ported from STM32 into the common file. This command now
takes an additional argument for the port to be used.
BUG=chrome-os-partner:23449
BRANCH=samus
TEST=emerge-samus chromeos-ec, tested on samus
Change-Id: I8308fbc2f34e369a20051dca9c5d43872f239777
Signed-off-by: Duncan Laurie <dlaurie@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/173837
Reviewed-by: Randall Spangler <rspangler@chromium.org>
stm32l100 is mostly identical to stm32l151,
excepted that the RAM is smaller (10kB instead of 16kB for the RB SKU),
the EEPROM is smaller, there is no touch capability (but we are not
using those 2 features).
So, in the new stm32l100 variant configuration, we adjust the memory
size to 10kB and keep the regular UART RX buffer size (512 B) rather
than putting a 2kB buffer to fit in the new constraints.
Signed-off-by: Vincent Palatin <vpalatin@chromium.org>
BRANCH=none
BUG=none
TEST=run on discovery board reworked with stm32l100rct6 and Nyan
reworked with stm32l100rbt6.
Change-Id: Ifd78f59a102b3079f0f794af8058211dc724153d
Reviewed-on: https://chromium-review.googlesource.com/173632
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Reviewed-by: Vic Yang <victoryang@chromium.org>
Tested-by: Vincent Palatin <vpalatin@chromium.org>
Commit-Queue: Vincent Palatin <vpalatin@chromium.org>
Commit-Queue: David James <davidjames@chromium.org>
Add some missing descriptions in config.h and rename a few defines to
be more consistent.
No functional changes, just comments and symbol renaming.
BUG=chrome-os-partner:18343
BRANCH=none
TEST=build all platforms; pass unit tests
Change-Id: I05a9a2ed6fd7bc8b14a18a0dc57d7d22430de21a
Signed-off-by: Randall Spangler <rspangler@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/173111
Reviewed-by: Vic Yang <victoryang@chromium.org>
Rather than scan the entire GPIO table, stop as soon as all interrupt
bits have been handled. We hand-order the table so GPIOs with
interrupts are first, so this should reduce interrupt overhead.
BUG=chrome-os-partner:23296
BRANCH=none
TEST=boot rambi
x86indebug -1
apshutdown
powerbtn
...That should print lots of 'x86 in' debug messages as pins
change state, showing that the interrupt handlers are still responding.
Change-Id: I7942cd51870ad51de068d90d68cf6634ff2fb1a0
Signed-off-by: Randall Spangler <rspangler@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/173031
Reviewed-by: Alec Berg <alecaberg@chromium.org>
Reviewed-by: Vic Yang <victoryang@chromium.org>
First implementation of a low power idle task for the LM4 chip. The
low power mode is selected by defining CONFIG_LOW_POWER_IDLE in a
board.h file. This commit turns it on for Peppy, Slippy, and Falco
only because those are the only boards tested.
When using the low power idle task, the chip goes in to deep sleep
when it can. Deep sleep disables clocks to most peripherals and puts
the onboard flash and RAM into a low power mode. The chip is woken
out of deep sleep using the RTC in the hibernate module. Increased
the idle task stack size to handle more involved idle task.
In board.c, the array of GPIO info can be used to select which GPIO
points can wake up the EC from deep sleep. Currenlty selected are
the power button, lid open, AC present, PCH_SLP_S3, and PCH_SLP_S5.
Additionally the port with the KB scan row GPIO point is also
enabled to wake up the EC from deep sleep.
Signed-off-by: Alec Berg <alecaberg@chromium.org>
BUG=None
BRANCH=none
TEST=Passes all unit tests. Runs on slippy, peppy, and falco with no
noticeable side affects. Verified that the power consumed by the EC
is lower when in S3, S5 and G3 by scoping the sense resistor
powering the chip.
Change-Id: I83fa9a159a4b79201b99f2c32678dc4fc8921726
Reviewed-on: https://chromium-review.googlesource.com/172183
Reviewed-by: Vincent Palatin <vpalatin@chromium.org>
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Commit-Queue: Alec Berg <alecaberg@chromium.org>
Tested-by: Alec Berg <alecaberg@chromium.org>
When the EC CPU is running at a decreased clock frequency, frequent
keyboard scans can starve other EC tasks of CPU and lead to dropped
data or watchdog timeouts.
Enforce a minimum number of EC clocks between keyboard scans to
prevent this from happening. The default chosen (16000 clocks) is
equal to the shortest post-scan delay (1 ms) of any current board when
the AP is in S0, so this should have no effect when the AP is in S0.
When the AP is in S3 or S5, we don't need to scan the keyboard as
frequently anyway. This can be overridden on a per-board basis for
future boards if needed.
BUG=chrome-os-partner:23247
BRANCH=pit
TEST=apshutdown, then hold down a key for 10 seconds. Should not see a
watchdog reset.
Change-Id: I228f53a32ad4769f6a137a9ab06903111bea115d
Signed-off-by: Randall Spangler <rspangler@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/172895
Reviewed-by: Vic Yang <victoryang@chromium.org>
Rambi's Silego chip inverts the pull as well as the signal. So it has
a pulldown for the signal instead of a pullup. The EC must drive the
signal push-pull, since open-drain only works for signals with pullups.
BUG=chrome-os-partner:23198
BRANCH=none
TEST=type on keyboard; C/D/E/3/F1-F4 should all work
Change-Id: Iaea4f540c523824571a568b70cca3e0fb467f79c
Signed-off-by: Randall Spangler <rspangler@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/172915
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
The Silego chip used on Rambi inverts column 2. So the EC should pull the signal low when NOT scanning column 2, and release it at all other times.
BUG=chrome-os-partner:23198
BRANCH=none
TEST=not yet; need to probe on scope
Change-Id: If6a784493533f11ae54d18f27591697e69aa2282
Signed-off-by: Randall Spangler <rspangler@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/172674
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Created a new function to enable or disable clocks to various
peripherals. This new function makes it easy to specify if you
want the clock enabled in run mode, sleep mode, and/or deep
sleep mode.
Added infrastructure to specify which GPIOs should interrupt the
EC from deep sleep.
BUG=none
BRANCH=none
TEST=Passes all unit tests. Ran on a peppy and verified that
the clock gate control registers in run mode (LM4_RCGC regs)
were the same before and after this change.
Change-Id: Ia5009ac8c837f61dca52fe86ebdeede2e1a7fe4d
Signed-off-by: Alec Berg <alecaberg@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/172454
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Rambi has a pair of LEDs which are attached to the PWM fan controller.
Add support for them. Also add a generic 'pwmduty' command which can
be used to get/set the duty cycle for any PWM channel.
Also fix rounding errors in pwm module, so that set/get duty doesn't
keep rounding down.
BUG=chrome-os-partner:22895
BRANCH=none
TEST=Boot rambi. LEDs are off.
pwmduty -> both are 0%
pwmduty 0 10 -> green LED on dimly
pwmduty 1 10 -> red LED on dimly
pwmduty 0 99 -> green LED on brightly
pwmduty 1 100 -> red LED on brightly
pwmduty 1 0 -> red LED off
pwmduty 1 -1 -> red LED turns back on because fan controller is disabled
pwmduty -> channel 0 at 99%, channel 1 disabled
Build all platforms. Pass all unit tests.
Change-Id: Ib0a6289a757554e696a9a0153a85bdc34e2ee2ae
Signed-off-by: Randall Spangler <rspangler@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/172094
We only used I2C_PORTS_USED to iterate through the list of hardware ports
actually in use, but we defined it in board.h at the same place where we
matched particular I2C devices to the (possibly shared) buses they're on.
This CL makes I2C_PORTS_USED into a global constant, so it can be set
automatically where we initialize the ports, and doesn't have to be
related to the list of attached devices.
BUG=chrome-os-partner:18343
BRANCH=none
TEST=manual
Build everything, run all tests, should still work.
Change-Id: I65f22f5cadfc4b3afe51af48faa5fb369bc3aa09
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/171884
include/config.h should have the canonical list of all CONFIG_* macros used
everywhere else. This fixes some that weren't included, and some that had
been changed in one place but not in others.
BUG=chrome-os-partner:18343
BRANCH=none
TEST=manual
Build everything. It should still work.
cd src/plaform/ec
make runtests
for i in bds bolt daisy discovery falco kirby link mccroskey peppy pit puppy rambi samus slippy snow spring; do make BOARD=$i || touch died.$i; done
There shouldn't be any died.* files.
Change-Id: I0a1ec2d57668509c514dc5a521e547836a3e9894
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/171690
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Every board other than link does the same thing - filter out the power
button event, then set the WAKE_L gpio level based on the remaining
events. This code doesn't need to be duplicated 7 times, so make it common.
Link didn't filter out the power button wake signal, but works fine
with the common implementation. Like the other boards it gets a power
button wake event via the dedicated PCH PWRBTN# signal.
BUG=chrome-os-partner:18343
BRANCH=none
TEST=suspend link, then wake using power button press
compile all platforms; pass unit tests
Change-Id: Ib3a6d310d0f5e337374b3c331ab2872fe377bdf6
Signed-off-by: Randall Spangler <rspangler@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/171405
Reviewed-by: Alec Berg <alecaberg@chromium.org>
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-by: Duncan Laurie <dlaurie@chromium.org>
We don't use the EEPROM block on the LM4 chip, but if we ever do we'll
want this off-by-one error fixed.
BUG=chrome-os-partner:18343
BRANCH=none
TEST=Build bds, since that's the only board which even compiles this code.
Pass unit tests.
Change-Id: I598f043cf00ebc1eca75fa6e6b7815d85a61e353
Signed-off-by: Randall Spangler <rspangler@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/171141
For all GPIOs, the current values are recorded. A test can then change
the value of a GPIO input by gpio_set_level(). The changed value is
recorded and also interrupt is fired if the change fits the interrupt
flags defined in board/host/board.c.
BUG=chrome-os-partner:19235
TEST=Pass all tests
BRANCH=None
Change-Id: If8e547e5adf4a20dcb118f5fe2187293005d4ca3
Signed-off-by: Vic Yang <victoryang@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/170907
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Reviewed-by: Vincent Palatin <vpalatin@chromium.org>
This just does a copy/rename from Bolt. Tweaking for Samus' peculiarities
will come next.
BUG=chrome-os-partner:22870
BRANCH=none
TEST=manual
The only thing we can check is that it compiles:
cd src/platform/ec
make BOARD=samus
Change-Id: Ied95ebdd1137548b21334b4a65a298c68482c517
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/171081
Reviewed-by: Randall Spangler <rspangler@chromium.org>
This provides an interface to detach and re-attach fake I2C devices,
which can be used to test I2C connection failure.
BUG=chrome-os-partner:19235
TEST=Pass sbs_charging test along with the next CL
BRANCH=None
Change-Id: Ibfee79b13d45e62377d894aa28547e77bef2189e
Signed-off-by: Vic Yang <victoryang@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/170752
Reviewed-by: Randall Spangler <rspangler@chromium.org>
With the PWM interface refactoring, only the CPU fan uses the fan.h
interface. All other PWM channels (keyboard backlight, etc.) use the
pwm.h interface. Remove the unused constants, and rename FAN_CH_CPU
to CONFIG_FAN_CH_CPU so it fits with the other fan config options.
No functional changes; just renaming things.
BUG=chrome-os-partner:18343
BRANCH=none
TEST=build all targets; pass unit tests
Change-Id: I391fbeaf54afcc29a11c2799a4520b7ad8784796
Signed-off-by: Randall Spangler <rspangler@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/170534
Reviewed-by: Vic Yang <victoryang@chromium.org>
This should have all the correct GPIO mappings.
Chipset and charger tasks are currently disabled, until we bring up
the voltage rails and I2C communication.
BUG=chrome-os-partner:22895
BRANCH=none
TEST=compiles; everything else needs to wait until we get hardware
Change-Id: Iea49fe7ab8bd17f61c8cc6c71f236a503418ee28
Signed-off-by: Randall Spangler <rspangler@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/170540
Reviewed-by: Vic Yang <victoryang@chromium.org>
This moves the mock function from common layer down to physical layer to
complete the test of common layer.
Also disable flash test for hardware tests, as this is only testing
common layer.
BUG=chrome-os-partner:19236
TEST=util/make_all.sh
BRANCH=None
Change-Id: Idd1c2c44591952894486f84d428872cfbf2cfdad
Signed-off-by: Vic Yang <victoryang@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/170297
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Currently, it doesn't compile unless CONFIG_FAN is defined.
BUG=chrome-os-partner:22803
BRANCH=none
TEST=temporarily undefine CONFIG_FAN in board/link/board.h; code compiles
and all unit tests pass
Change-Id: I251d670ccd299f7a50b1455364a817e07fad4cb1
Signed-off-by: Randall Spangler <rspangler@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/170106
Currently, CONFIG_UART_HOST must be defined for all LM4-based
platforms. Future platforms may not need host UART support, so make
it optional.
BUG=chrome-os-partner:22802
BRANCH=none
TEST=Temporarily comment out CONFIG_UART_HOST in board/link/board.h;
code still compiles.
Change-Id: I8775d9daaf8f6e45f1f374c9be7f10316611447f
Signed-off-by: Randall Spangler <rspangler@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/170104
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
To test drivers, we need a way to fake I2C periphrals. With this CL, a
fake peripheral can be done by declaring its own I2C read/write
functions. The fake I2C peripherals may return EC_ERROR_INVAL to
indicate it's not responding. The emulator I2C read/write call scans
through all registered I2C peripherals and uses the first response.
BUG=chrome-os-partner:19235
TEST=Pass sbs_charging test with the next CL.
BRANCH=None
Change-Id: I9380dc40e147781b42e09eb6979c864bbd9f2ac4
Signed-off-by: Vic Yang <victoryang@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/169511
Reviewed-by: Randall Spangler <rspangler@chromium.org>
STM32 has a single-byte mailbox for UART I/O. When the core clock
runs at 16Mhz we can service interrupts fast enough to handle 115200
baud input, but when we drop to 1MHz we drop characters. Using DMA to
receive input solves this problem.
The STM32 DMA engine can only generate interrupts when the transfer is
half-done / all-done, so we need to poll the DMA receive-head-pointer
to see if individual characters have been received. Do this in the
tick task (every 250ms). When a character is received, poll more
quickly for a bit (5 times before the next tick) so the input console
is more responsive to typing.
BUG=chrome-os-partner:20485
BRANCH=none
TEST=Console is responsive to debug commands. For example, help -> prints help
apshutdown -> shuts down AP
arrow keys -> move cursor and scroll through command history
Ctrl+Q, help, wait a second, Ctrl+S -> help output printed after Ctrl+S
Then in chip/stm32/config_chip.h, comment out #define CONFIG_UART_RX_DMA
and rebuild/reflash the EC. When the AP is up, the console works normally
but after 'apshutdown', the EC drops to 1MHz core clock, and the arrow
keys don't work. (This step confirms that adding DMA support did not
change the behavior of systems where CONFIG_UART_RX_DMA is not defined.)
Change-Id: I199448354824bd747c7b290ea7fd5ccf354c11bb
Signed-off-by: Randall Spangler <rspangler@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/169406
Reviewed-by: Simon Glass <sjg@chromium.org>
All calls to it did
if (uart_tx_stopped())
uart_tx_start();
And that was the only use of uart_tx_stopped(). Merge the functions.
BUG=chrome-os-partner:20485
BRANCH=none
TEST=EC debug console still prints output and accepts commands.
Ctrl+Q pauses output and Ctrl+S resumes it.
Change-Id: I113c64f5fdfc6b02b63034a74b1a3c6c6a76c351
Signed-off-by: Randall Spangler <rspangler@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/169329
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
This reduces the number of UART interrupts by a factor of 12, and
reduces the overall interrupt rate on STM32 by a factor of 2.
BUG=chrome-os-partner:20485
BRANCH=none (not required for pit branch)
TEST=Boot pit. Ctrl+Q pauses debug output; Ctrl+S resumes it.
'crash divzero' still prints a full crash dump.
And util/makeall.sh passes builds all platforms and passes tests.
Change-Id: I86993e14b436150298dcb2c6d29086cc3c9db418
Signed-off-by: Randall Spangler <rspangler@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/168814
This is a precursor to DMA-based UART transfers, which require
different processing for DMA vs PIO output types.
BUG=chrome-os-partner:20485
BRANCH=pit
TEST=Boot pit; verify EC console still works.
Change-Id: I6d6f55561eeebe9bd2928b2bfb25278c86f689d1
Signed-off-by: Randall Spangler <rspangler@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/168811
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
The definition of GPIO interface allows passing in multi-bit mask, and
this is what's done by gpio_config_module(). Fix STM32L's function so
that it doesn't accidentally set incorrect GPIO register values.
BUG=chrome-os-partner:22605
TEST=On Kirby, do 'led r 0' and check the value of 0x40020800 is
0x01540000.
BRANCH=None
Change-Id: I9a1c8074aab7345485a590ecf138bf99d0742997
Signed-off-by: Vic Yang <victoryang@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/168739
Reviewed-by: Vincent Palatin <vpalatin@chromium.org>
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Tested-by: Randall Spangler <rspangler@chromium.org>
This is in preparation for enabling DMA-based UART transfers, to
improve UART performance on STM32.
BUG=chrome-os-partner:20485
BRANCH=none
TEST=Boot pit. Host commands should still be received; this verifies DMA
is still operational.
Change-Id: Ibc3b2e2cd187547eb61b85e4a086704accd7f2fb
Signed-off-by: Randall Spangler <rspangler@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/168810
BQ27541 is not a smart battery IC, and thus we cannot use existing smart
battery driver. Let's add a driver that implements a smart-battery-like
interface.
The 'battery' console command is also moved to battery.c so that it can
be reused by different battery driver.
BUG=chrome-os-partner:22048
TEST=Type 'battery' and check the reported values are sane.
TEST=Check 'battery' command works fine on Spring.
BRANCH=None
Change-Id: I5d1eaeb3f801478f3b9473fd43c1f2a2eda75859
Signed-off-by: Vic Yang <victoryang@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/66340
ADC module on STM32L is clocked by HSI oscillator, and thus we need to
switch to HSI if using MSI. After the conversion, if the system is not
in S0, clock is switched back to MSI again.
There are several register bits that can only be written when ADC is
powered down. For now, let's just power down the ADC after each
conversion.
Currently ADC watchdog is not working and is disabled on STM32L.
BUG=chrome-os-partner:22242
TEST=Try multiple all-channel and single-channel reads in S0 and S5.
BRANCH=None
Change-Id: I769dda8a9c69ac9de1eb22d6d259034eef8c1ac4
Signed-off-by: Vic Yang <victoryang@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/167454
If we reboots the emulator without flushing test coverage information,
the test coverage report will be incorrect. Let's fix this by flushing
it before every reboot.
BUG=chrome-os-partner:19235
TEST=Generate test coverage report and check correctness.
BRANCH=None
Change-Id: I1eb060e419b767f382325bed841366c491ba56b7
Signed-off-by: Vic Yang <victoryang@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/167770
Reviewed-by: Vincent Palatin <vpalatin@chromium.org>
GPIO_INPUT is defined as 0, and any GPIO flag cannot be examined against
GPIO_INPUT. Change GPIO_INPUT to non-zero value to avoid this.
BUG=chrome-os-partner:22275
TEST=On Kirby, set a GPIO to output and pull it low, and then set it back to
input. Check it can be pull high externally.
TEST=Build all boards.
TEST=Boot link and spring.
BRANCH=None (unless this bug hits some other boards.)
Change-Id: I84b9936c24af538ac59c36129fda27ca879bf9d1
Signed-off-by: Vic Yang <victoryang@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/167190
Reviewed-by: Vincent Palatin <vpalatin@chromium.org>
and when the bus seems wedged at i2c_init().
BUG=chrome-os-partner:19286
TEST=Manual test on peach pit. Tried the following wedged cases:
(1) Bit bang a transaction but only read part of the response.
(Refer to https://chromium-review.googlesource.com/#/c/66389)
Command to wedge: i2cwedge 0x90 0 2 2
(2) Bit bang a transaction to do a "write" and stop while the other side is
acking. (Refer to https://chromium-review.googlesource.com/#/c/66389)
Command to wedge: i2cwedge 0x90 0 1
(3) Same as (1) but do a reboot instead of returning and see
that the unwedge works at init time w/ no cancelled transactions.
(Refer to https://chromium-review.googlesource.com/#/c/66389)
Command to wedge: i2cwedge 0x90 0 6 2
(4) Same as (2) but do a reboot instead of returning and see
that the unwedge works at init time w/ no cancelled transactions.
(Refer to https://chromium-review.googlesource.com/#/c/66389)
Command to wedge: i2cwedge 0x90 0 5
(5) Manually pull down on SCL.
(Refer to https://chromium-review.googlesource.com/#/c/66063)
All five cases successfully wedged the bus and were recovered by this change.
BRANCH=pit
[dianders: made sure we don't change SCL after SCL high, misc other bits]
Change-Id: I23f16fcaa2a76ea37025f8204ab1cdb27e9ef6d1
Signed-off-by: Hung-ying Tyan <tyanh@chromium.org>
Signed-off-by: Doug Anderson <dianders@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/66915
Previously we could get into some pretty bad situtation around
frequency changes. Specifically:
* If the freq change code exectued during a transaction then it would
get all messed up.
* If the freq change notifier executed during a transaction then we'd
reset the bus midway through transaction. Badness.
BRANCH=pit
BUG=chrome-os-partner:22093
TEST=With all patches together:
- on AP: suspend_stress_test
- on EC: battery 10000 50
Change-Id: I24be29d459fe229a6278829a7c03c1e102358e8c
Signed-off-by: Doug Anderson <dianders@chromium.org>
Previous-Reviewed-on: https://chromium-review.googlesource.com/167102
(cherry picked from commit deab2433c13bc3b3bb75cd33c12dad633ba6a42a)
Reviewed-on: https://chromium-review.googlesource.com/167149
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Before this change drivers had no way of knowing that a frequency
change was coming. This could cause problems for some drivers (like
i2c) that need to make sure that a transaction isn't happening while a
frequency change is happening.
The PRE_FREQ_CHANGE archiecture is very simple here and we don't allow
any way to cancel it.
At the moment, we guarantee:
- We won't call PRE_FREQ_CHANGE with interrupts disabled, so acquiring
locks / sleeping is OK.
- We'll call the actual HOOK_FREQ_CHANGE after the PRE_FREQ_CHANGE.
PRE_FREQ_CHANGE and HOOK_FREQ_CHANGE should not use deferred function
calls.
BRANCH=pit
BUG=chrome-os-partner:22093
TEST=With all patches together:
- on AP: suspend_stress_test
- on EC: battery 10000 50
Change-Id: I2731a3e85d41e749fa571fdb74b5c9b12043cda6
Signed-off-by: Doug Anderson <dianders@chromium.org>
Previous-Reviewed-on: https://chromium-review.googlesource.com/167101
(cherry picked from commit d84c0dbbf7c5a72917a820e292ecfdfa698d0fb9)
Reviewed-on: https://chromium-review.googlesource.com/167148
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Currently, the STM32L I2C driver queues the stop condition, but
doesn't actually wait for it to take effect before returning. If
another back-to-back transfer is started, this may attempt to send a
start condition before the stop condition completes. If this happens
after a slave read, this can look to the slave like just another pulse
on SCL, causing it to clock out another bit - potentially driving SDA
low and hanging the bus.
Instead, wait for the bus to go idle, then wait another clock period
(10 us) to give the slaves plenty of time to detect bus-idle before
the next start condition.
BUG=chrome-os-partner:22093
BRANCH=pit
TEST=repeatedly run the battery i2c command from the EC console while
running 'ectool i2cxfer 0 0x48 1' from a root shell. Should not hang
the I2C bus.
Change-Id: I5e65ee242537dbc801fba4ae57847a5af5104186
Signed-off-by: Randall Spangler <rspangler@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/66997
Reviewed-by: Douglas Anderson <dianders@chromium.org>
Commit-Queue: Douglas Anderson <dianders@chromium.org>
Dumping before the ADDR bit is cleared also has the effect of clearing the ADDR
bit.
BUG=chrome-os-partner:22235
TEST=Manual test on peach pit.
Keep executing the "battery" command on the EC console while busy running
flashrom on the host to read back the EC. See that there's no error produced.
BRANCH=pit
Signed-off-by: Hung-ying Tyan <tyanh@chromium.org>
Change-Id: I10e88a8512f4e9eb90267ef1aca1df5dd214318d
Reviewed-on: https://chromium-review.googlesource.com/66930
Reviewed-by: Douglas Anderson <dianders@chromium.org>
Commit-Queue: Hung-ying Tyan <tyanh@chromium.org>
Tested-by: Hung-ying Tyan <tyanh@chromium.org>
Reviewed-by: Randall Spangler <rspangler@chromium.org>
This unifies the PWM module interface for LM4 and STM32. Now PWM
channels are defined in board.h/board.c. Instead of calling functions
named pwm_set_fan_duty(x), one can now use pwm_set_duty(PWM_CH_FAN, x),
which prevents additional functions added when we have a new PWM
channel.
BUG=chrome-os-partner:18343
TEST=Limit input current on Spring.
TEST=Check power LED in S0/S3/S5 on Snow.
TEST=Check keyboard backlight functionality on Link.
TEST=Check fan speed control/detecting on Link.
BRANCH=None
Change-Id: Ibac4d79f72e65c94776d503558a7592f7db859dc
Signed-off-by: Vic Yang <victoryang@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/64450
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Problems with existing thermal control loop:
* Not multi-board friendly. thermal.c only supports Link and needs
refactoring. Temp thresholds and fan speeds are hard-coded.
* Only the PECI temp is used to determine the fan speed. Other temp sensors
are ignored.
* Has confusing data structures. Values in the CPU temp thresholds array mix
ACPI thresholds with fan step values.
With this change, the thermal task monitors all temp sensors in order to
perform two completely independent functions:
Function one: Determine if the host needs to be throttled by or informed of
any thermal events.
For thermal events, each temp sensor will have three threshold levels.
TEMP_HOST_WARN
* When any sensor goes above this level, host_throttle_cpu(1) will be called
to ask the CPU to slow itself down.
* When all sensors drop below this level, host_throttle_cpu(0) will be called.
* Exactly AT this level, nothing happens (this provides hysteresis).
TEMP_HOST_HIGH
* When any sensor goes above this level, chipset_throttle_cpu(1) will be
called to slow the CPU down whether it wants to or not.
* When all sensors drop below this level, chipset_throttle_cpu(0) will be
called.
* Exactly AT this level, nothing happens (this provides hysteresis).
TEMP_HOST_SHUTDOWN
* When any sensor is above this level, chipset_force_shutdown() will be
called to halt the CPU.
* Nothing turns the CPU back on again - the user just has to wait for things
to cool off. Pressing the power button too soon will just trigger shutdown
again as soon as the EC can read the host temp.
Function two: Determine the amount of fan cooling needed
For fan cooling, each temp sensor will have two levels.
TEMP_FAN_OFF
* At or below this temperature, no active cooling is needed.
TEMP_FAN_MAX
* At or above this temperature, active cooling should be running at maximum.
The highest level of all temp sensors will be used to request the amount of
active cooling needed. The function pwm_fan_percent_to_rpm() is invoked to
convert the amount of cooling to the target fan RPM.
The default pwm_fan_percent_to_rpm() function converts smoothly between the
configured CONFIG_PWM_FAN_RPM_MIN and CONFIG_PWM_FAN_RPM_MAX for percentages
between 1 and 100. 0% means "off".
The default function probably provide the smoothest and quietest behavior,
but individual boards can provide their own pwm_fan_percent_to_rpm() to
implement whatever curves, hysteresis, feedback, or other hackery they wish.
BUG=chrome-os-partner:20805
BRANCH=none
TEST=manual
Compile-time test with
make BOARD=falco runtests
On the EC console, the existing fan commands should work correctly:
faninfo - display the fan state
fanduty NUM - force the fan PWM to the specified percentage (0-100)
fanset RPM - force the fan to the specified RPM
fanset NUM% - force the fan to the specified percentage (0-100) between
its configured minimum and maximum speeds from board.h
(CONFIG_PWM_FAN_RPM_MIN and CONFIG_PWM_FAN_RPM_MAX)
fanauto - let the EC control the fan automatically
You can test the default pwm_fan_percent_to_rpm() with
fanset 1%
faninfo
The fan should be turning at CONFIG_PWM_FAN_RPM_MIN. Let the EC control it
automatically again with
fanauto
Also on the EC console, the thermal settings can be examined or changed:
> temps
PECI : 327 K = 54 C
ECInternal : 320 K = 47 C
G781Internal : 319 K = 46 C
G781External : 318 K = 45 C
>
> thermalget
sensor warn high shutdown fan_off fan_max name
0 373 387 383 333 363 PECI
1 0 0 0 0 0 ECInternal
2 0 0 0 0 0 G781Internal
3 0 0 0 0 0 G781External
>
> help thermalset
Usage: thermalset sensor warn [high [shutdown [fan_off [fan_max]]]]
set thermal parameters (-1 to skip)
>
> thermalset 2 -1 -1 999
sensor warn high shutdown fan_off fan_max name
0 373 387 383 333 363 PECI
1 0 0 0 0 0 ECInternal
2 0 0 999 0 0 G781Internal
3 0 0 0 0 0 G781External
>
From the host, ectool can be used to get and set these parameters with
nearly identical commands:
ectool thermalget
ectool thermalset 2 -1 -1 999
Change-Id: Idb27977278f766826045fb7d41929953ec6b1cca
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/66688
Reviewed-by: Randall Spangler <rspangler@chromium.org>
The BOOTCFG register configures a couple of important things: whether to
allow jumping into the builtin ROM bootloader at reset, and whether or not
to allow JTAG access for programing and debugging.
The default is "no" and "yes". But the BOOTCFG register can be locked so
that it can't be changed again, which means that if the wrong values are put
into it, the system is pretty much bricked.
On Link, we wrote a BOOTCFG value that allowed a GPIO to be used as a bypass
to optionally trigger the ROM bootloader, but on Slippy and its derivatives
that GPIO is not pulled up. If you program the Link values into BOOTCFG on a
Slippy, the system is stuck in the ROM bootloader more or less forever.
This change disables that GPIO, keeps JTAG enabled, and locks those settings
for all LM4 chips (it's a chip config now, not a board config). We've never
actually used the GPIO to invoke the ROM bootloader, but we have managed to
brick a number of systems just by having it enabled, so we're going to lock
it into a safe configuration now.
BUG=chrome-os-partner:19247
BRANCH=falco,peppy
TEST=manual
Reflash, boot, power cycle (actually unplug the EC from AC and battery) a
few times. It should continue to work.
Change-Id: Iaf1a81d6814104421a56425490e3d5164ea9b617
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/66538
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-by: Duncan Laurie <dlaurie@chromium.org>
Reviewed-by: Randall Spangler <rspangler@chromium.org>
This tests that command history is as expected. Also fix a bug that some
checks in console_edit test are skipped.
BUG=chrome-os-partner:19236
TEST=Pass console_edit test.
BRANCH=None
Change-Id: Ifbd3d1690f25b35bf5efe523e656b013aa534d26
Signed-off-by: Vic Yang <victoryang@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/64837
BUG=chrome-os-partner:21847
BRANCH=peppy
TEST=Manual. Check state of GPIO_P5000_FAN_EN with lid open
and lid closed. Can also check with meter via TP109.
Change-Id: I8a64c14d53dd84a5d586c0abb04ccb71de0e78b3
Signed-off-by: Dave Parker <dparker@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/65674
