In preparation of introducing new update protocol version this patch
cleans up the existing implementation.
The receive and transmit PDU headers are separated into their own
structures and are now shared between the server and client sides.
Some comments have been added to better explain different protocol
versions' details.
BRANCH=none
BUG=chrome-os-partner:55789
TEST=verified that it is still possible to update RW_A and RW_B on a
Kevin CR50, works using both USB and SPI.
Change-Id: Ied65b2c2a7800bae045f1d2dd64e58dd5e793d27
Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/368989
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-by: Mary Ruthven <mruthven@chromium.org>
Enable the SHI interrupt only after we have received and begun
processing our host command. Disable the SHI interrupt once our
transaction is complete (with either success or error status). This will
prevent the SHI interrupt from being asserted at the same time as the CS
interrupt, which can lead to the SHI interrupt being serviced first.
Also, it avoids needless, non-useful SHI interrupts during error
transactions.
BUG=chrome-os-partner:55710,chrome-os-partner:55795,chrome-os-partner:56254
BRANCH=None
TEST=Manual on gru. Stress test flashrom w/ unpowered Donette attached
(for host command spam), verify no errors encountered after 100 minutes.
Also verify host command interface functions properly after sysjump.
Change-Id: I41e3deb382897cd4286e6ac96f4f3066bf7a94a7
Signed-off-by: Shawn Nematbakhsh <shawnn@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/371510
Reviewed-by: Aseda Aboagye <aaboagye@chromium.org>
Reviewed-by: Mulin Chao <mlchao@nuvoton.com>
Following steps need to be followed for SCI/SMI generation:
1. If automatic mode is being used, set SCIE/SMIE bit in HIPMIE register
during lpc_init. This step needs to be performed only once.
2. If SCI/SMI generation is required, set ST1/ST2 bit in HIPMST register
during update_host_event_status.
3. In lpc_generate_sci/smi, if automatic mode is being used, set
SCIB/SMIB bit in HIPMIC register.
BUG=chrome-os-partner:55489
BRANCH=None
TEST=lidclose behavior verified in kernel.
Change-Id: Ib6e0ee32bc06a8c51297010ac4843a7de45a99bd
Signed-off-by: Furquan Shaikh <furquan@google.com>
Reviewed-on: https://chromium-review.googlesource.com/371059
Commit-Ready: Furquan Shaikh <furquan@chromium.org>
Tested-by: Furquan Shaikh <furquan@chromium.org>
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Reviewed-by: Mulin Chao <mlchao@nuvoton.com>
Header version fields are instrumental when determining which of the
available images is started by the RO. Let's include the header
version when reporting the RW images' version as well as RO.
BRANCH=none
BUG=none
TEST=verified that RW header information is now included in the
version command output:
> vers
Chip: g cr50 B2
Board: 0
RO_A: * 0.0.8/8755904e
RO_B: -1.-1.-1/ffffffff
RW_A: 0.0.1/cr50_v1.1.5093-751a584+
RW_B: * 0.0.1/cr50_v1.1.5093-d27f65f
Build: 0.0.1/cr50_v1.1.5093-d27f65f
...
Change-Id: I675c473a277e272f55670324fafdab8a6e6edd78
Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/370939
Reviewed-by: Scott Collyer <scollyer@chromium.org>
This adds basic support for the stm32f446.
This consists of:
* New DMA model for stm32f4
* New clock domain support.
* MCO oscillator gpio export support.
* Flash support for irregular blocks.
BUG=chromium:608039
TEST=boots w/ correct clock, stm32f0 also boots.
BRANCH=None
Change-Id: I1c5cf6ddca09009c9dac60da8a3d0c5ceedfcf4d
Signed-off-by: Nick Sanders <nsanders@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/363992
Reviewed-by: Daisuke Nojiri <dnojiri@chromium.org>
ble_tx now resets values that could prevent its functioning.
Since there is no NRF51 shortcut from TX->RX, it also blocks until
packet is transmitted. This prevents calling RX before TX is completed,
specifically in advertising state.
Also added timeouts to prevent possibility of freezing in case of
unexpected state.
TEST=Used function before and after CL in a more fully implemented stack.
This improved reliability.
BUG=None
BRANCH=None
Change-Id: I6a5b0b6f36e37ac0102d254bbdc9dfcd29694bb8
Signed-off-by: Levi Oliver <levio@google.com>
Reviewed-on: https://chromium-review.googlesource.com/370597
Reviewed-by: Aseda Aboagye <aaboagye@chromium.org>
TEST=Used function before and after CL in a more fully implemented
stack. This improved reliability and lengthened connection times.
BUG=None
BRANCH=None
Change-Id: I60680c8855d6166e4e4a6a71639ee57464fa21ce
Signed-off-by: Levi Oliver <levio@google.com>
Reviewed-on: https://chromium-review.googlesource.com/370420
Reviewed-by: Aseda Aboagye <aaboagye@chromium.org>
Disable sleep for the first 3 minutes after initializing cr50 and
disable sleep for 3 minutes after a wakepin interrupt has been received.
BUG=none
BRANCH=none
TEST=manual
on a development board pull up dioa3 and dioa12 then verify cr50
does goes to sleep after 3 minutes.
on gru see sleep is delayed by 3 minutes when the wakeup
interrupt is called.
enable deep sleep and check cr50 does not go to sleep for 3
minutes after resuming.
Change-Id: I28ec3c2f5f86326b926d403ad52ffb4fc108e7ec
Signed-off-by: Mary Ruthven <mruthven@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/367880
Reviewed-by: Vadim Bendebury <vbendeb@chromium.org>
Adding a lowest priority hook init function, guaranteed to run after
all other initialization is completed, which is a good time to take
the EC out of reset.
Also moving pin hold release into the same function.
BRANCH=none
BUG=chrome-os-partner:55797
TEST=verified proper reset pulse generated on reef
verified that reef, kevin and gru all boot up as expected
Change-Id: Ic91e871f5040b9aa02a2f11cf76d2d596c48a04c
Signed-off-by: Timothy Chen <timothytim@google.com>
Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/367410
Reviewed-by: Marius Schilder <mschilder@chromium.org>
According the formula for computing prescaler of PWM, we needn't minus
one again in line 89 of pwm.c. On npcx's evb, we observe the deviation
of PWM 1 which freq is 10K is reduced from 0.6 kHz to 0.01 kHz. We also
change PWM 1's freq to 100 Hz with PWM_CONFIG_DSLEEP and the deviation
is reduced from 64.2 Hz to 9.2Hz.
Modified sources:
1. pwm.c: Modifed formula for computing prescaler of PWM.
BRANCH=none
BUG=chrome-os-partner:56052
TEST=make BOARD=npcx_evb; use "pwm 1 50" to observe deviation of freq
Change-Id: I9cd7fc0d807df62083d49a7e9ec6de1d9028be1b
Signed-off-by: Mulin Chao <mlchao@nuvoton.com>
Reviewed-on: https://chromium-review.googlesource.com/368560
Reviewed-by: Shawn N <shawnn@chromium.org>
Previously, timeout meant the number of attempts taken to receive. Now,
it means the number of microseconds before timing out.
TEST=printouts displaying time before and after rx attempt. Not included
in CL.
BUG=None
BRANCH=None
Change-Id: I00ccfc4bbf15f77c2777f35c911dceacaff98e4f
Signed-off-by: Levi Oliver <levio@google.com>
Reviewed-on: https://chromium-review.googlesource.com/368471
Reviewed-by: Aseda Aboagye <aaboagye@chromium.org>
In command_flash_spi_sel_lock(), and command_flash_tristate() a value
could be printed without being initialized first.
Only display the values in the paths where the variable gets read.
BUG=None
TEST=Build and boot Reef
BRANCH=None
Change-Id: I8ef86f966d017290491d6fe2b1486ce913cd09fb
Signed-off-by: Martin Roth <martinroth@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/368301
Reviewed-by: David Hendricks <dhendrix@chromium.org>
Re-wrote ble_radio_init to work with both data and
advertising packets. Updated all calls to refactored function.
RADIO_PCNF0_ADV renamed because it applies to advertisement and
data packet formats. Updated all references to this value.
TEST=make BOARD=hadoken
BUG=None
BRANCH=None
Change-Id: I0fdbe0eb146ce5cbc40e3ac67bf4d0e5465dcc2f
Signed-off-by: Levi Oliver <levio@google.com>
Reviewed-on: https://chromium-review.googlesource.com/367732
Commit-Ready: Myles Watson <mylesgw@chromium.org>
Reviewed-by: Aseda Aboagye <aaboagye@chromium.org>
Reviewed-by: Myles Watson <mylesgw@chromium.org>
Add espi driver for x86-based platform which support espi interface such
as skylake and so on.
Added source:
1. espi.c: Add drivers which supports the utilities of peripheral and
virtual-wire channels so far.
2. espi.h: Add espi virtual-wire declaration for power sequence FW.
Modified sources:
1. lpc.c: Add interrupts and initialization steps for espi.
2. gpio.c: Add interrupt handler of espi reset.
BRANCH=none
BUG=chrome-os-partner:34346
TEST=make buildall -j; test nuvoton IC specific drivers
Change-Id: Ie80afe79d85aba47fc0b72898a8374c2898ec114
Signed-off-by: Mulin Chao <mlchao@nuvoton.com>
Reviewed-on: https://chromium-review.googlesource.com/366181
Reviewed-by: Randall Spangler <rspangler@chromium.org>
This fixes several errors caused by these being defined as static
while being used in the inline function i2c_handle_sda_irq().
From the C99 Draft 6.7.4.3:
An inline definition of a function with external linkage shall not
contain a definition of a modifiable object with static storage
duration, and shall not contain a reference to an identifier with
internal linkage.
TEST=Build with GCC 5.3 flash and boot Reef board
BUG=None
BRANCH=None
Change-Id: Ie487f17b92736c2a56280783267da5d3bb12b969
Signed-off-by: Martin Roth <martinroth@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/367486
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Certain tasks (eg. chipset) may be woken directly by other tasks /
unrelated interrupts. Add an explicit wake event for ADC conversion done
so that we're not mistakenly woken.
BUG=chrome-os-partner:54971
BRANCH=None
TEST=Manual on kevin rev5, run "reboot ap-off" then "sysjump rw", verify
console isn't spammed with GPIO warning message due to ADC failure to
read board version.
Signed-off-by: Shawn Nematbakhsh <shawnn@chromium.org>
Change-Id: I5477e11c2b434e4b350d81393f4463eea1a91e7c
Reviewed-on: https://chromium-review.googlesource.com/366943
Commit-Ready: Shawn N <shawnn@chromium.org>
Tested-by: Shawn N <shawnn@chromium.org>
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Sometimes a perfectly sane image enters rolling reboot mode in case
some data change triggered a bug which prevents the normal startup and
causes a reset.
The most likely task causing in in case of cr50 would be the tpm task.
Let's add another check of the restart counter: should it reach the
value of 50, do not start the TPM task.
BRANCH=none
BUG=chrome-os-partner:55708
TEST=with this code plus an unaligned access introduced in tpm
initialization sequence in both RW_A and RW_B, program the full
image on the dev board.
Observe the device reset 50 time is rapid succession and then
stop with the following message on the console:
Bldr |511709
retry|50
Himg =4F992103..408D193E
Hfss =384E4655..EE13EBD0
Hinf =44D21600..B70529BD
jump @00044000
--- UART initialized after reboot ---
[Reset cause: rtc-alarm]
[Image: RW, cr50_v1.1.5044-8d6f7a2+ private-cr51:v0.0.68-633229c ...
+ cryptoc:v0.0.4-5319e83 2016-08-07 19:37:16 vbendeb@kvasha]
[0.004130 Inits done]
[0.006919 Active NVram partition set to 0]
Console is enabled; type HELP for help.
> system_rolling_reboot_suspected: roling reboots suspected. Try \
powercycling to clear
this condition.
[0.010502 Task 2 (TPM) exited!]
Change-Id: I6b08c5c1a02da9edf9bdf394e57cc56d2e595ad1
Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/366892
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
The nvmem space defined in the ec code base for the cr50 board is used
by the TPM2 library, which has its own nvram size definition. The two
definitions must match.
On top of the fact that the definitions are not locked to each other,
there is a third completely unrelated nvram size definition in
board/cr50/board.c.
This patch unifies nvmem size definitions between cr50 and tpm2
repositories by adding a compile time check for the size to be the
same on both sides.
Also, it turns out that two certificates (RSA and ECC) together do not
quite fit into the cr50 TPM nvram. Hence the total allocated nvmem
space is being increased to 20K (note that the actual nvram size
available to the TPM is less than half of this).
BRANCH=none
BUG=chrome-os-partner:55898
TEST=tpm does not lock up any more when 'tpm_client --own' is ran on the
Kevin-tpm2 command line
CQ-DEPEND=CL:367010
Change-Id: I20b4f54118bd2fa12e5bd5741d6c58fbe91f65d1
Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/366796
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
Servo_micro sets usb config maxpower to 100mA.
Servo_v4 is set to self powered as it's powered by a
shared vbus and not be the bub it's connected to.
cr50 is self powered as no power is transmitted as part of CCD.
* Add CONFIG_USB_MAXPOWER_MA to define USB maximum power draw requested
per board.
* Add CONFIG_USB_SELF_POWERED to indicate that a device is not
powered by allocated USB power.
BUG=chromium:631302
TEST=lsusb reports 100mA bMaxPower (micro), Self powered (v4)
BRANCH=None
Change-Id: I79b8ce46f32d94f16104a4a8080104e30dce7f2c
Signed-off-by: Nick Sanders <nsanders@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/363153
Reviewed-by: Vincent Palatin <vpalatin@chromium.org>
This change disables sleep only when a cable is actually detected.
Before it would disable sleep no matter what and if a debug cable was
not plugged in and then deteached then it would never enable it.
BUG=none
BRANCH=none
TEST=manual
update cr50
unplug suzyq
boot to kernel
verify cr50 is asleep
run 'powerd_dbus_suspend'
cr50 will wake up for like a second then go back to sleep
run 'apreset' on EC console
boot to kernel
verify cr50 is asleep
run 'powerd_dbus_suspend'
cr50 will wake up for like a second then go back to sleep
Change-Id: I8337f694853b7840ef932b38bd0fa6453d31cb3d
Signed-off-by: Mary Ruthven <mruthven@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/366861
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
Our UART interrupt must be able to preempt our SHI_CS interrupt,
otherwise console input may be lost. Adjust our relative IRQ priorities
to accommodate this.
BUG=chrome-os-partner:55920
BRANCH=None
TEST=Run `echo "kbpress 11 4 1" > /dev/pts/17` on kevin 200 times from
the recovery screen, verify that all input is received by the EC.
Change-Id: I36203511f5883272287ac22d0704098fbd933758
Signed-off-by: Shawn Nematbakhsh <shawnn@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/366622
Commit-Ready: Shawn N <shawnn@chromium.org>
Tested-by: Mulin Chao <mlchao@nuvoton.com>
Reviewed-by: Aseda Aboagye <aaboagye@chromium.org>
Reviewed-by: Mulin Chao <mlchao@nuvoton.com>
There are few reasons why the SoC may reboot which we haven't
been reporting (they just show up as "[Reset cause: other]").
This adds a bit of decoding to explain some of those "other"
reasons.
BUG=none
BRANCH=none
TEST=make buildall; try on Cr50
I tested one of the new reasons using "crash hang". It shows up
correctly as "{Reset cause: security]". I haven't specifically
tested all of the new reasons, but since this is basically just a
change to console message they should work too. I'll double-check
those cases once some blocking bugs are fixed.
Change-Id: I46daed29d7e37bda9034a3486127bed0ea25f803
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/366400
Reviewed-by: Vadim Bendebury <vbendeb@chromium.org>
This change fixes two issues. The first issue is that during rdd_init
debug_detect is set to the value of DEBUG_STATE_MAP. Debug_detect should
be the default map value of 0x420 which triggers an rdd interrupt when
cc1 and cc2 both have a value of 011 or 001. This will detect when a
debug cable is attached. ~debug_detect should detect when the cable is
detached.
The issue is after a soft reset DEBUG_STATE_MAP is not reset to the
default value so on the next rdd_init debug_detect was set to whatever
was DEBUG_STATE_MAP was last set to. If the debug cable is detected
before the soft reset then DEBUG_STATE_MAP will have a value of
~debug_detect and when rdd_init is called then debug_detect will be
initialized to that value of ~0x420. Now debug_detect does the opposite
of what we expect it to do. debug_detect detects when the cable is
detached and ~debug_detect detects when it is attached.
rdd_interrupt is called cr50 reads cc1 and cc2 and sees that a debug
accessory is attached so it sets DEBUG_STATE_MAP to ~debug_detect.
Because debug_detect does the opposite of what is expected then it will
trigger an interrupt when it detects a debug accessory. The debug
accessory is already attached so an interrupt is triggered. It reads cc1
and cc2 sees the debug_accessory attached resets the map to
~debug_detect. This cycle continues until the watchdog resets the
system. After this reset rdd_init reads the map to get debug_detect and
it is 0x420 because that is how it was last set and then rdd works
again.
This change fixes this issue by defining DEBUG_DETECT to be 0x420 and
not using the STATE_MAP value to do it.
It also looks like when the EC is off and we change the value of the
debug map, another RDD interrupt is triggered. Then it reads the debug
state and resets the map again another interrupt is triggered. It
continuously does this.
This change adds a check to see that the RDD was trying to detect the
state it currently senses before doing anything so the second interrupt
will be ignored.
BUG=chrome-os-partner:55793
BRANCH=none
TEST=manual
plug in a reworked suzyq
Flash the EC
verify a bunch of RDD interrupts aren't triggered and the
reworked suzyq can be used.
plug it in and out a couple of times to make sure everything
works.
Change-Id: I6f4e7b15cf9c53b965533df7fe23fae5b340f70d
Signed-off-by: Mary Ruthven <mruthven@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/364724
Commit-Ready: Vadim Bendebury <vbendeb@chromium.org>
Reviewed-by: Vadim Bendebury <vbendeb@chromium.org>
This reverts commit 25f19f5bbd, which was
causing host command issues after sysjump. IBUFSTAT seems to behave
strangely. Additional investigation is required.
BUG=chrome-os-partner:55710,chrome-os-partner:55795
BRANCH=None
TEST=Verify host commands continue to work after sysjump.
Change-Id: Id94a9c4677bcae597b9353e081418e649e823564
Signed-off-by: Shawn Nematbakhsh <shawnn@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/366173
Reviewed-by: Randall Spangler <rspangler@chromium.org>
There is just one version of Cr50 firmware for all boards
that it's used on. However, on some boards the AP communicates
to the TPM via a SPI interface (i.e. Kevin) and on others, the
AP communicates via an I2C interface (i.e. Reef). In order to
dynamically discover which interface to configure, there are
strapping resistors added to the board which enables the Cr50
to detect which configuration to implement.
This CL is a first pass and is only looking at DIOA1 which is
pulled high for SPI and pulled low for I2C configurations.
The strapping resistor should be read when the AP is in reset
prior to it attempting to drive any of the lines used for
strapping. To ensure this condition is met, Cr50 will only
check the strapping options following a POR (power on reset).
Once the configuration type is discovered, a 'long_life'
register is used to hold the result so that the result can
always be available. The long_life register contents remain
unchanged until a subsequent power down event.
BRANCH=none
BUG=chrome-os-partner:50728
TEST=manual
Tested on Kevin and Reef. Verfifed by reading the stored value
that the SPI configuraiton is detected for Kevin and the I2C
interface is detected on Reef. In addition, verified on Kevin
that the Cr50 FW version is correctly reported to the AP which
means that TPM register reads via the slave SPI are functioning.
Change-Id: Ibd7624ad8e3b4126f6346dce0bc72f62a3cc6d18
Signed-off-by: Scott <scollyer@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/363014
Commit-Ready: Scott Collyer <scollyer@chromium.org>
Tested-by: Scott Collyer <scollyer@chromium.org>
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
Watchdog events are delivered as internal ARM interrupts, so we
can print a crash dump and then reboot. However, if interrupts
are disabled when the watchdog triggers, it just hangs forever.
This CL configures the watchdog and processor lockup events to
trigger a hard reboot through a security alert. This is the only
way to make these events non-maskable.
BUG=chrome-os-partner:52597
BRANCH=none
TEST=manual
I added this console command:
static int command_hang(int argc, char **argv)
{
interrupt_disable();
while (1)
;
return EC_ERROR_UNKNOWN; /* Not reached */
}
DECLARE_CONSOLE_COMMAND(hang, command_hang, NULL, "Hang", NULL);
Without this CL, that command locked the SoC up until it was
reset from outside. With this CL, it reboots after a couple of
seconds.
Change-Id: I773c0138fd2243cdbcdd86b2c7138520155d7920
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/365531
Reviewed-by: Vadim Bendebury <vbendeb@chromium.org>
RADIO_STATE is broken: remove it.
Build on the geneneric radio support to send and receive
Bluetooth LE packets.
Add macros in registers.h to configure PCNF0 and PCNF1.
BUG=None
BRANCH=None
TEST=Send advertisements with console commands
ble_adv type length [interval_us]
for example: ble_adv 2 8
Advertisements should be received by other devices
The Bluetooth Address has the form C5:A4:A3:A2:A1:A*
The device name is a substring of ABCDEFGH...
ABCDEFGH @ C5:A4:A3:A2:A1:A2 (name length is 8, type is 2)
ABCDEFGH @ C5:A4:A3:A2:A1:A2 (name length is 8, type is 2)
ABCDEF @ C5:A4:A3:A2:A1:A6 (name length is 6, type is 6)
TEST=Listen for advertisements with console commands
ble_adv_scan chan [num] [addr0]
for example: ble_scan 37
Example output:
BLE packet @ 20000448: type 2, len 33,
5c.f3.70.6b.65.d2 AdvA
20000454: 02 01 08 17 09 43 68 72
2000045c: 6f 6d 65 62 6f 78 20 66
20000464: 6f 72 20 4d 65 65 74 69
2000046c: 6e 46 16
02 01 08 = 2 bytes, Flags, LE and BR capable
17 09 43... = 23 bytes, Name, "Chromebox for Meetings"
Change-Id: I2bd3f1d87acb069da0b56c1d7878e7d4fd6a06f3
Signed-off-by: Myles Watson <mylesgw@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/361960
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Reviewed-by: Levi Oliver <levio@google.com>
Enable the SHI interrupt only after we have received and begun
processing our host command. Disable the SHI interrupt once our
transaction is complete (with either success or error status). This will
prevent the SHI interrupt from being asserted at the same time as the CS
interrupt, which can lead to the SHI interrupt being serviced first.
Also, it avoids needless, non-useful SHI interrupts during error
transactions.
BUG=chrome-os-partner:55710,chrome-os-partner:55795
BRANCH=None
TEST=Manual on gru. Stress test flashrom w/ unpowered Donette attached
(for host command spam), verify no errors encountered after 100 minutes.
Change-Id: I0ab20b0202ebcfe15c04b272ec67001a6a358dad
Signed-off-by: Shawn Nematbakhsh <shawnn@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/364698
Commit-Ready: Shawn N <shawnn@chromium.org>
Tested-by: Shawn N <shawnn@chromium.org>
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Reviewed-by: Mulin Chao <mlchao@nuvoton.com>
Add functions to initialize and disable the radio.
Add packet definitions.
Update the spelling of the FICR_OVERRIDEEN register.
BUG=None
BRANCH=None
TEST=make buildall -j
Change-Id: I3a9e500d0f177b6ce77a3b6ed6a42acd4f49eb7e
Signed-off-by: Myles Watson <mylesgw@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/362175
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Reviewed-by: Levi Oliver <levio@google.com>
Cr50 cant retain the TPM state in deep sleep so it wont be enabled until
it knows that the AP is off. If the AP is off it wont be asserting
SPS_CS_L, but it may be low because the AP isn't pulling it up.
This change disables it as a wake source in deep sleep.
BUG=chrome-os-partner:54796
BRANCH=none
TEST=run 'idle d'. Make sure cr50 goes into deep sleep and only resumes
due to a rdd event or when sys_rst_l is asserted.
Change-Id: Idf3ded6b439b71a27ac7eb4682a65dcdd6342cb9
Signed-off-by: Mary Ruthven <mruthven@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/364864
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
Increasing the USB console TX buffer size allows to see pretty much
all early startup messages generated before USB console is
initialized.
There is still plenty of room left in SRAM, 23K on cr50, much more on
all other g based boards.
BRANCH=none
BUG=none
TEST=observed better USB console output on cr50 restarts.
Change-Id: I82f37ee7f3aecd8b7e95f3d421789c11375b2fd4
Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/364811
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-by: Nadim Taha <ntaha@chromium.org>
The only place where two separate buffers for the RO version strings
is required is the tpm_registers.c:set_version_string() function.
In preparation of reporting the build string along with the version
string, let's rearrange the function not to require separate buffers
for the RO versions.
BRANCH=none
BUG=chrome-os-partner:55558
TEST=verified that version reported by the TPM driver on Kevin is
still correct:
localhost ~ # grep cr50 /sys/firmware/log
Firmware version: RO_A: 0.0.1/84e2dde7 RO_B:* 0.0.2/13eda43f RW_A:*...
Change-Id: I8924ac48bd838851670f0d659e95aa92a8524665
Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/364587
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
As a temp measure until a proper solution is implemented, reset the
restart counter when the PCR_Read command is issued by the host.
This is a good indication that Chrome OS is through the boot process,
as PCR value is used to determine the boot mode.
BRANCH=none
BUG=chrome-os-partner:55667
TEST=installed the new image on a Kevin cr50 and rebooted it in normal
and recovery modes, observed on the cr50 console the message like
> system_process_retry_counter:retry counter 1
Change-Id: Ib55e161d5edbf8f6e2d387fd756b94aa53c20ed8
Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/364311
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
- Pass-thru to IBF handler code in case both IBHF and IBF interrupts are
pending, in order to properly keep track our Tx byte count.
- Don't disable the SHI IRQ in our host command handler callback since
system-wide interrupts are already disabled.
BUG=chrome-os-partner:55711,chrome-os-partner:55721
BRANCH=None
TEST=Manual on gru with subsequent commit. Verify `flashrom -p ec -r
file.bin` passes 100x with no errors or warnings.
Signed-off-by: Shawn Nematbakhsh <shawnn@chromium.org>
Change-Id: I6225ffde1fe0127c7484933fe4a151d22f42415c
Reviewed-on: https://chromium-review.googlesource.com/364234
Commit-Ready: Shawn N <shawnn@chromium.org>
Tested-by: Mulin Chao <mlchao@nuvoton.com>
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Reviewed-by: Mulin Chao <mlchao@nuvoton.com>
Whether the bootrom locks the bootloader or not is deteremined by
fuses and/or flags in the bootloader's signed header. This CL
locks the active bootloader, just case those aren't configured to
do so.
BUG=chrome-os-partner:55261
BRANCH=none
TEST=manual
On an unlocked bootloader, I see this after booting:
> rw 0x40090100
read 0x40090100 = 0x00000001
With this CL applied, I see this instead:
> rw 0x40090100
read 0x40090100 = 0x00000000
Change-Id: I2e1396b7d7e71c8633d97d3cb573e9468eeb51e7
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/364280
Reviewed-by: Vadim Bendebury <vbendeb@chromium.org>
It has rare chance for FW to get a unexpected value when reading
IBUFSTAT. This is because the clock source of SHI and CPU are
asynchronous. The reading value is invalid if IBUFSTAT is during
transition state. Use two consecutive equal reading can make sure
the value is valid.
BUG=chrome-os-partner:34346
TEST=run "while true; do ectool version; done" on gru, verify each
failure happens about 50000 host commands
BRANCH=none
Change-Id: Ie246561d201dd87d89cb2424c23d016dcdcd47c9
Signed-off-by: CHLin <CHLIN56@nuvoton.com>
Reviewed-on: https://chromium-review.googlesource.com/362734
Commit-Ready: Randall Spangler <rspangler@chromium.org>
Tested-by: CH Lin <chlin56@nuvoton.com>
Tested-by: Mulin Chao <mlchao@nuvoton.com>
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Reviewed-by: Mulin Chao <mlchao@nuvoton.com>
We need to ignore sys_rst_l right now when we use the usb spi endpoint
to update the AP or EC. We hold the EC and AP in reset and this causes
sys_rst_l to be asserted at the start of updating the AP and when the EC
comes out of reset.
Using the USB SPI endpoint may require doing a bunch of transactions
back to back. Cr50 should not reset itself between each one.
This change postpones the reset until we're done using the usb spi
endpoint. Once sys_rst_l just resets the TPM we can remove all of this.
BUG=chrome-os-partner:52366
BUG=chrome-os-partner:54982
BRANCH=none
TEST=manual
verify 'util/flash_ec --board=kevin --raiden' updates the EC
'sudo flashrom -p raiden_debug_spi:target=AP -w $IMG' updates
the AP
The AP and cr50 reset after usb_spi is disabled.
Change-Id: I68a76012bc7bf6d3abd073a70f0b90e440d72c49
Signed-off-by: Mary Ruthven <mruthven@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/364051
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
It is useful to be able to see which pins are set as wake pins and what
type they are. This change adds prints to show_pinmux to describe the
wake pins.
BUG=none
BRANCH=none
TEST='pinmux' should show DIOA12 as a wake_low source.
Change-Id: I2a0ccdbf9b07abb627c3d52c7dd28433a2beff3c
Signed-off-by: Mary Ruthven <mruthven@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/363494
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
Cr50 was not waking up long enough after SPS_CS_L was asserted for the
spi slave transactions to start and disable sleep. It also was not
handling SYS_RST_L properly when it was asleep.
This change sets SPS_CS_L to be an edge triggered wake up source instead
of level triggered, because cr50 should just wake up on the edge and
disable sleep until the spi transaction is done.
It also adds sys_rst_l as a wakeup source. The sys_rst_asserted
interrupt cannot be triggered while cr50 is asleep, so the
pmu_wakeup_interrupt will call sys_rst_asserted if SYS_RST_L is low at
resume. This change relies on the EC extending the delay in
chipset_reset to be long enough for SYS_RST_L to still be asserted when
cr50 resumes.
BUG=chrome-os-partner:54331
BRANCH=none
TEST=manual
make sure suzyq is disconnected.
verify ap boots up to the kernel after running
'gpioset SYS_RST_L 0' then 'gpioset SYS_RST_L 1' on the ec
console.
Check that cr50 goes to sleep when the AP is not trying to use
the TPM.
When cr50 is asleep pwrbtn + refresh still resets the system.
Disable SYS_RST_L_IN as a wake source and verify the system
verification fails and requests a recovery image.
Change-Id: I807b1918842d96c9d2922aa33404d87ab28b9906
Signed-off-by: Mary Ruthven <mruthven@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/363606
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
Reduce size of UART Tx buffer to 1024 bytes on all npcx platforms and
increase size of code memory by 6K bytes on Kevin.
BUG=chrome-os-partner:52876
BRANCH=None
TEST=`make buildall -j` with subsequent commit.
Signed-off-by: Shawn Nematbakhsh <shawnn@chromium.org>
Change-Id: Ib9e52a4406f84cfc434984f8819d7ef02b70beb4
Reviewed-on: https://chromium-review.googlesource.com/363591
Commit-Ready: Shawn N <shawnn@chromium.org>
Tested-by: Shawn N <shawnn@chromium.org>
Reviewed-by: Aseda Aboagye <aaboagye@chromium.org>
A recent cr50 loader modification introduced a counter in a scratch
register which is incremented on every startup. The idea is that valid
RW would decrement the counter, signaling that the start was
successful.
Should the counter exceed the value of 5, the loader assumes that the
RW being started is not fit to run, and picks the older RW to run, if
available.
This patch adds a function to process the startup retry counter.
First of all the counter is zeroed, as this function is supposed to be
called only once the RW run is considered successful and reliable.
Then the current situation is examined. If the counter value read from
the scratch register exceeds 5 AND running image is not the newer of A
and B, it is considered an indication of a fallback from a bad newer
image.
To prevent the newer image from being considered a contender on the
following startups, its header is corrupted.
BRANCH=none
BUG=chrome-os-partner:55151, chrome-os-partner:55667
TEST=modified code for testing purposes, by adding a call to
system_process_retry_counter() to tpm_task() after line 534, which
would cause the new function to be called soon after boot.
built a new image and installed it on the debug board. Then
modified the image to throw an exception early in the boot up
sequence, and installed it as a newer image on the debug board.
Observed the debug board restart the new image several time and
then fall back to the older image, printing the following on the
console:
system_process_retry_counter:retry counter 7
corrupt_other_header: RW fallback must have happened, magic at 44000 before: ffffffff
corrupt_other_header: magic after: 0
The following restarts start the older image without trying to run
the failing newer image.
Change-Id: Ia7497401e38fe2c3957af910cf745e45da985245
Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/362776
Reviewed-by: Randall Spangler <rspangler@chromium.org>
The SoC looks for two RO images at reset, and is typically
configured for two RW images as well. This CL reports version
strings for all those images, as well as identifying the active
RO and RW copies.
Since the RO image doesn't contain a version string, we create
one using the epoch_, major_, minor_, and img_chk_ members of its
signed header.
BUG=chrome-os-partner:55558
BRANCH=none
TEST=make buildall; run on Cr50 hardware
The "version" command now includes information like this:
RO_A: * 0.0.2/a3c3d5ea
RO_B: 0.0.2/8895c9eb
RW_A: cr50_v1.1.4965-a6c1c73-dirty
RW_B: * cr50_v1.1.4959-2f49d5c
The '*' indicates the active image.
The test/tpm_test/tpmtest.py program has been updated to request
the version information at startup, and it also now reports
similar information, just all on one line:
RO_A:* 0.0.2/a3c3d5ea RO_B: 0.0.2/8895c9eb RW_A: cr50_v1.1 ...
The active images are marked with a '*' following the ':', so
that the same regexp can match either format:
($ro, $rw) = m/RO_[AB]:\s*\*\s+(\S+).*RW_[AB]:\s*\*\s+(\S+)/s;
Change-Id: Ic27e295d9122045b2ec5a638933924b65ecc8e43
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/362861
Reviewed-by: Vadim Bendebury <vbendeb@chromium.org>
There are two g chip versions in circulation currently, B1 and B2.
Make the 'version' command properly report it.
BRANCH=none
BUG=none
TEST=verified that both B1 and B2 report versions properly
Change-Id: I1c5b9f0da0170cda2c636b857e92b9d3de165422
Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/362643
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
This change implements logic for installing
endorsement certificates in the RW section.
The endorsement certificates are initially
provisioned in a fixed RO flash region and
are copied in the RW TPM data region (once
this region has been initialized).
Also add code for reading from the info bank,
which is where the endorsement seed is
initially stored.
BRANCH=none
BUG=chrome-os-partner:43025,chrome-os-partner:47524
BUG=chrome-os-partner:50115
TEST=TCG tests running
Change-Id: Id8c16d399202eee4ac0c4e397bdd29641ff9d2f3
Signed-off-by: nagendra modadugu <ngm@google.com>
Reviewed-on: https://chromium-review.googlesource.com/362402
Reviewed-by: Vadim Bendebury <vbendeb@chromium.org>
Commit-Queue: Vadim Bendebury <vbendeb@chromium.org>
Tested-by: Vadim Bendebury <vbendeb@chromium.org>
