This reverts commit c459c8278e
as the fix is not straightforwad, some host command codes in
private repos are expressed using C preprecessor which
breaks the assumption of this patch that all host commands
are expressed as four digit hex numbers.
Change-Id: I922de9ae8dbab6eef048463c5c09b1f338152083
Reviewed-on: https://chromium-review.googlesource.com/414492
Reviewed-by: Vadim Bendebury <vbendeb@chromium.org>
Commit-Queue: Vadim Bendebury <vbendeb@chromium.org>
Tested-by: Vadim Bendebury <vbendeb@chromium.org>
Use binary search in host command lookup dispatcher
BUG=chrome-os-partner:570895
TEST=manual testing on kevin
- Kevin boots
- ectool hello
make buildall -j
Verify *.smap hcmds section is sorted:
100bca94 R __hcmds
100bca94 R __host_cmd_0x0000
100bcaa0 R __host_cmd_0x0001
100bcaac R __host_cmd_0x0002
100bcab8 R __host_cmd_0x0003
100bcac4 R __host_cmd_0x0004
100bcad0 R __host_cmd_0x0005
100bcadc R __host_cmd_0x0006
100bcae8 R __host_cmd_0x0007
100bcaf4 R __host_cmd_0x0008
100bcb00 R __host_cmd_0x0009
100bcb0c R __host_cmd_0x000a
100bcb18 R __host_cmd_0x000b
100bcb24 R __host_cmd_0x000d
100bcb30 R __host_cmd_0x0010
100bcb3c R __host_cmd_0x0011
100bcb48 R __host_cmd_0x0012
100bcb54 R __host_cmd_0x0013
100bcb60 R __host_cmd_0x0015
100bcb6c R __host_cmd_0x0016
100bcb78 R __host_cmd_0x0017
100bcb84 R __host_cmd_0x0025
100bcb90 R __host_cmd_0x0026
100bcb9c R __host_cmd_0x0029
100bcba8 R __host_cmd_0x002a
100bcbb4 R __host_cmd_0x002b
100bcbc0 R __host_cmd_0x002c
100bcbcc R __host_cmd_0x0044
100bcbd8 R __host_cmd_0x0045
100bcbe4 R __host_cmd_0x0046
100bcbf0 R __host_cmd_0x0047
100bcbfc R __host_cmd_0x0061
100bcc08 R __host_cmd_0x0062
100bcc14 R __host_cmd_0x0064
100bcc20 R __host_cmd_0x0065
100bcc2c R __host_cmd_0x0067
100bcc38 R __host_cmd_0x0087
100bcc44 R __host_cmd_0x008c
100bcc50 R __host_cmd_0x008d
100bcc5c R __host_cmd_0x008f
100bcc68 R __host_cmd_0x0092
100bcc74 R __host_cmd_0x0093
100bcc80 R __host_cmd_0x0096
100bcc8c R __host_cmd_0x0097
100bcc98 R __host_cmd_0x0098
100bcca4 R __host_cmd_0x0099
100bccb0 R __host_cmd_0x009e
100bccbc R __host_cmd_0x00a0
100bccc8 R __host_cmd_0x00a1
100bccd4 R __host_cmd_0x00a8
100bcce0 R __host_cmd_0x00a9
100bccec R __host_cmd_0x00b6
100bccf8 R __host_cmd_0x00b7
100bcd04 R __host_cmd_0x00d2
100bcd10 R __host_cmd_0x00d3
100bcd1c R __host_cmd_0x00db
100bcd28 R __host_cmd_0x0101
100bcd34 R __host_cmd_0x0102
100bcd40 R __host_cmd_0x0103
100bcd4c R __host_cmd_0x0104
100bcd58 R __host_cmd_0x0110
100bcd64 R __host_cmd_0x0111
100bcd70 R __host_cmd_0x0112
100bcd7c R __host_cmd_0x0113
100bcd88 R __host_cmd_0x0114
100bcd94 R __host_cmd_0x0115
100bcda0 R __host_cmd_0x0116
100bcdac R __host_cmd_0x0117
100bcdb8 R __host_cmd_0x0118
100bcdc4 R __host_cmd_0x011a
100bcdd0 R __evt_src_EC_MKBP_EVENT_KEY_MATRIX
100bcdd0 R __hcmds_end
BRANCH=none
Change-Id: Ideb9951b318763f71915e2c4e5052f4b4bfab173
Reviewed-on: https://chromium-review.googlesource.com/405528
Commit-Ready: Sam Hurst <shurst@google.com>
Tested-by: Sam Hurst <shurst@google.com>
Reviewed-by: Randall Spangler <rspangler@chromium.org>
When "make buildall" has board which uses core/minute-ia,
CROSS_COMPILE?=i686-pc-linux-gnu- doesn't get set unless
CROSS_COMPILE is not defined; however, it's defined before
this line, and wrong toolchain is used.
Remove "?" to set correct CROSS_COMPILE.
BUG=none
BRANCH=none
TEST='make buildall -j' passed
Change-Id: Ied4a9f93a4d44714c012d3a3e50e4a34f41a7c1f
Signed-off-by: li feng <li1.feng@intel.com>
Reviewed-on: https://chromium-review.googlesource.com/410402
Commit-Ready: Li1 Feng <li1.feng@intel.com>
Tested-by: Li1 Feng <li1.feng@intel.com>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
In __wait_evt(), if a timer expiration occurs after we read event
status, before the timer is canceled, then TASK_EVENT_TIMER will be
propagated to the next task wait, likely leading to premature timeout.
Prevent this by clearing TASK_EVENT_TIMER after canceling our timer.
BUG=chrome-os-partner:58658
BRANCH=gru
TEST=Manual on gru, run 'pd # hard' for 12 hours with charger attached,
verify no TCPC I2C read errors occur.
Change-Id: Iac2f05a768b4ef29f82e7c3eb899f4c7dd5c3744
Signed-off-by: Shawn Nematbakhsh <shawnn@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/400968
Commit-Ready: Shawn N <shawnn@chromium.org>
Tested-by: Shawn N <shawnn@chromium.org>
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Reviewed-by: Vincent Palatin <vpalatin@chromium.org>
To reset the TPM task, we send it an event so that it will reset
only when it's not busy doing actual TPM stuff that might fiddle
with the stack or shared memory. But that means that we can't
always know when the task finally gets around to resetting
itself.
This CL adds a tpm_reset() function that blocks until the reset
actually occurs. Obviously it can't do that if it's being called
in interrupt context or from the TPM task itself, but otherwise
it does.
BUG=chrome-os-partner:52366
BRANCH=none
CQ-DEPEND=CL:361680
TEST=make buildall, test on Gru, manual tests
In addition to the normal rebooting, logging in/out, and so
forth. I added a temporary console command to call tpm_reset()
and scattered a bunch of ccprintfs around it. When called due to
SYS_RST_L, it didn't block. When invoked with the console
command, it did.
Change-Id: I51e8b1299dbdcd1a12273cf48a890e93ed32a8c8
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/388125
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Rather than doing a full cr50 reset when the system reset SYS_RST_L is
asserted, just reset the state of the TPM task and library. Re-clear
.bss for those modules, then re-initialize.
BRANCH=none
BUG=chrome-os-partner:52366
CQ-DEPEND=CL:366792
TEST=make buildall; test on Gru
Trigger a SYS_RST_L by using the AP's reboot command, power
off/on, log in/out/in.
See that the Cr50 does not reboot and the firmware and userspace
are still happy about the TPM.
Change-Id: I32cd2bb72316f68c74db77a20a8d09112b402d4b
Signed-off-by: Randall Spangler <rspangler@chromium.org>
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/361680
We found the api utility in ROM doesn't enable burst mode of GDMA. It
influences the performance of FW download a lot. The CL modified GDMA
for moving the code of the other region from flash to ram. And move a
function that kicks off GMDA transactions to suspend ram in case this
utility is erased by itself. This issue will be fixed in our next
generation.
Modified sources:
1. system.c: Implement GDMA bypass.
2. system_chip.h: Import flash addresses for GDMA bypass code.
3. registers.h: Add GDMA register definitions.
4. cortex-m/ec.lds.S: Add lowpower_ram2 section in linker script.
BRANCH=none
BUG=chrome-os-partner:56794
TEST=make BOARD=npcx_evb; test sysjump and measure download time
Change-Id: I8490f8f2e5a8cdcb6fd10511878c4a4af8073bbf
Signed-off-by: Mulin Chao <mlchao@nuvoton.com>
Reviewed-on: https://chromium-review.googlesource.com/381779
Commit-Ready: Shawn N <shawnn@chromium.org>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Add a hook to act when the a device is going in tablet mode and back.
BUG=chromium:606718
BRANCH=kevin
TEST=Test with evtest that an event is sent to the AP.
Change-Id: Ic9c3b158f1178504af41abff18b28de8e07fc7a7
Signed-off-by: Gwendal Grignou <gwendal@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/380412
Reviewed-by: Aseda Aboagye <aaboagye@chromium.org>
Even when CONFIG_RESTRICTED_CONSOLE_COMMANDS is enabled, there
are many commands that can't do anything dangerous. This marks
some of those commands as safe to use, even when restrictions are
enforced.
I'm only marking commands that are used by the Cr50, since that's
the only board that has restrictions.
BUG=chrome-os-partner:55322
BRANCH=none
TEST=make buildall, test on Cr50 hardware
Change-Id: I6289d332830175b6adcb6b20cb4c21d01d27a25e
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/376188
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Since pretty much always, we've declared console commands to take
a "longhelp" argument with detailed explanations of what the
command does. But since almost as long, we've never actually used
that argument for anything - we just silently throw it away in
the macro. There's only one command (usbchargemode) that even
thinks it defines that argument.
We're never going to use this, let's just get rid of it.
BUG=none
BRANCH=none
CQ-DEPEND=CL:*279060
CQ-DEPEND=CL:*279158
CQ-DEPEND=CL:*279037
TEST=make buildall; tested on Cr50 hardware
Everything builds. Since we never used this arg anyway, there had
better not be any difference in the result.
Change-Id: Id3f71a53d02e3dc625cfcc12aa71ecb50e35eb9f
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/374163
Reviewed-by: Myles Watson <mylesgw@chromium.org>
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Add __keep attribute to irqprio entries to ensure they are not dropped
when CONFIG_LTO is enabled.
BUG=chrome-os-partner:55920
BRANCH=None
TEST=Manual on kevin. Check build/RO/ec.RO.map, verify that
.rodata.irqprio section is not empty.
Change-Id: I51ae23556d6f46b2cd7ba098f0e7a785292b2853
Signed-off-by: Shawn Nematbakhsh <shawnn@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/366571
Commit-Ready: Shawn N <shawnn@chromium.org>
Tested-by: Shawn N <shawnn@chromium.org>
Reviewed-by: Aseda Aboagye <aaboagye@chromium.org>
Currently, the matrix keyboard protocol does not have support for
handling non-matrixed keys. This commit adds support for buttons which
do not appear in the keyboard matrix as well as switches.
Additionally, the keyboard FIFO is now just a general MKBP events FIFO
which MKBP events are free to use. Now, buttons and switches wil join
the key matrix event.
BUG=chrome-os-partner:54988
BUG=chrome-os-partner:54976
BUG=chromium:626863
BRANCH=None
TEST=Flash kevin, and verify that keyboard is still functional.
TEST=make -j buildall
CQ-DEPEND=CL:358926
Change-Id: If4ada904cbd5d77823a0710d4671484b198c9d91
Signed-off-by: Aseda Aboagye <aaboagye@google.com>
Reviewed-on: https://chromium-review.googlesource.com/358633
Commit-Ready: Aseda Aboagye <aaboagye@chromium.org>
Tested-by: Aseda Aboagye <aaboagye@chromium.org>
Reviewed-by: Randall Spangler <rspangler@chromium.org>
cts.tasklist contains tasks run only for CTS. These tasks are added to the
tasks registered in ec.tasklist with higher priority. This design allows
board directories to be free from CTS stuff.
cts.tasklist can be placed in each suite directory (cts/suite/cts.tasklist).
If a suite does not define its own cts.tasklist, the common list is used
(i.e. cts/cts.tasklist).
BUG=chromium:624520
BRANCH=none
TEST=Ran the followings:
make buildall
make CTS_MODULE=gpio BOARD=nucleo-f072rb
make CTS_MODULE=gpio BOARD=stm32l476g-eval
Change-Id: Ibb242297ee10a397a8fcb6ff73d8cbc560daa885
Signed-off-by: Daisuke Nojiri <dnojiri@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/359445
Reviewed-by: Chris Chen <twothreecc@google.com>
Default setting is at 48MHz.
For PLL frequency at 24MHz:
1. USB module can't work, it requires 48MHz to work.
2. SSPI clock frequency is divide by two.
Signed-off-by: Dino Li <dino.li@ite.com.tw>
BRANCH=none
BUG=none
TEST=1. uart, i2c, timer, and pd modules are function normally
at different PLL frequency settings.
2. use 'flashrom' utility to flash EC binary with different
PLL settings.
Change-Id: Iabce4726baff493a6136136af18732b58df45d7f
Reviewed-on: https://chromium-review.googlesource.com/347551
Commit-Ready: Dino Li <Dino.Li@ite.com.tw>
Tested-by: Dino Li <Dino.Li@ite.com.tw>
Reviewed-by: Randall Spangler <rspangler@chromium.org>
LPC access interrupt only enabled when EC entering deep doze mode. This
will reduce interrupt of LPC access. Also, this interrupt is always
enabled for LPC platform to support "CONFIG_LOW_POWER_S0".
Signed-off-by: Dino Li <dino.li@ite.com.tw>
BRANCH=none
BUG=none
TEST=Tested ectool command 'version' x 10000.
Change-Id: I9053c4018b38a8a852c3c6254e1fcde625f3fa3a
Reviewed-on: https://chromium-review.googlesource.com/336112
Commit-Ready: Dino Li <dino0303@gmail.com>
Tested-by: Dino Li <dino0303@gmail.com>
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Previously the maximum number of deferred routines was specified by the
the default maximum number of deferred routines you had to override
this, and if you wanted fewer, you still payed the price of having the
defer_until array statically allocated to be the maximum size.
This change removes that define and instead creates the RAM state of
the deferred routine (the time to wait until to call the deferred) when
the deferred is declared.
Signed-off-by: Anton Staaf <robotboy@chromium.org>
BRANCH=None
BUG=None
TEST=make buildall -j
manually test on discovery-stm32f072
Change-Id: Id3db84ee1795226b7818c57f68c1f637567831dc
Reviewed-on: https://chromium-review.googlesource.com/335597
Commit-Ready: Anton Staaf <robotboy@chromium.org>
Tested-by: Anton Staaf <robotboy@chromium.org>
Reviewed-by: Aseda Aboagye <aaboagye@chromium.org>
If OUTDIR brings in a "@", the build breaks because that delimits the
path, leading to invalid file names.
This can happen (and happened) when building on a Jenkins CI instance
which uses jobname@number as path for parallel checkouts on a
single build node.
BRANCH=none
BUG=none
TEST=build with make out=foo@bar ... failed and works now.
Change-Id: Id0594f0d7312419110091443755ec11b5f8ee2d8
Signed-off-by: Patrick Georgi <pgeorgi@google.com>
Reviewed-on: https://chromium-review.googlesource.com/327110
Commit-Ready: Patrick Georgi <pgeorgi@chromium.org>
Tested-by: Patrick Georgi <pgeorgi@chromium.org>
Reviewed-by: Martin Roth <martinroth@chromium.org>
To get better power consumption in S0, we add FW support for
CONFIG_LOW_POWER_S0.
Before entering deep idle in S0, we must enable Host interrupt to wake up
EC if it needs to service LPC bus.
This version also add a new bit of sleep_mask (SLEEP_MASK_FAN) in system.h
to prevent EC enter deep idle if fan's duty isn't zero. Normally, the freq of
PWM fan is 25 kHz. It means we must select apb2 clock as the source clock of
PWM fan. Or fan would stop when ec enters deep idle because of no PWM signal.
In hwtimer.c, we reset the preload counter to maximum value in ITEI32's ISR
since preload counter is changed by __hw_clock_source_set all the time.
We also found there're no event set if it's deadline is over 32 bits but
current source clock isn't. To prevent ec doesn't wake-up in deep-idle even if
ITIM32 expires, FW set an event for ITIM32 after process_timers().
Modified sources:
1. wheatley/board.h: Add CONFIG_LOW_POWER_S0 definition.
2. clock.c: Enable Host interrupt for LPC.
3. clock.c: Disable LP_WK_CTL for better power consumption.
4. gpio.c: Add ISR for Host interrupt.
5. uart.c: Introduce bit 6 of USTAT to make sure transmitting is completed.
6. register.h: Add uart_clear_pending_wakeup function.
7. hwtimer.c: Fixed watchdog issue when ITIM32 is closed to overflow.
8. fan.c: Enable deep sleep if duty cycle is zero.
9. include/system.h: Add SLEEP_MASK_FAN for fan control loop.
10. core/cortex-m/task.c: Add "isb" to flash the garbage data in the
instruction pipeline.
BUG=chrome-os-partner:34346
TEST=make buildall -j; test nuvoton IC specific drivers
BRANCH=none
Change-Id: Ibe3630d0d68cf3f32206adb2afa1b5958916a2be
Signed-off-by: Mulin Chao <mlchao@nuvoton.com>
Reviewed-on: https://chromium-review.googlesource.com/324651
Reviewed-by: Shawn N <shawnn@chromium.org>
We have a limitation for EC interrupt vector registers.
System may read incorrect interrupt number in ISR so we need to add
a workaround to prevent it.
The following is a example that got incorrect interrupt number:
1. REG IVCTx = 0x10. (no interrupt pending)
2. EC INT6 interrupt occurs (IVCTx = 0x16) and jump to ISR.
3. Read interrupt vector register to determine interrupt number.
4. Higher priority interrupt of same interrupt group occurs
(for example: INT134, IVCTx = 0x96) while the system is reading the
interrupt vector register for EC INT6, we may end up with an incorrect
interrupt number between 0x16 and 0x96.
Signed-off-by: Dino Li <dino.li@ite.com.tw>
BRANCH=none
BUG=none
TEST=1. EC interrupts work normally: WUI (GPIO interrupt), timer, uart,
LPC, I2C, and PECI.
2. Console command 'taskinfo'.
Change-Id: I54e61f417ad506eb3b4cd5d0652f64eed9a28a17
Reviewed-on: https://chromium-review.googlesource.com/322097
Commit-Ready: Dino Li <dino.li@ite.com.tw>
Tested-by: Dino Li <dino.li@ite.com.tw>
Reviewed-by: Randall Spangler <rspangler@chromium.org>
There are hooks for chipset power sequencing but not one to indicate
that the system has reset at runtime. Add a hook for this and
implement for lm4 and mec1322. The hook is notified on any platform
reset, including those that happen on the way into S3/S5 state.
There is a new config variable added because the hook is notified in
the interrupt handler and needs a deferrable function that needs to
be added to every board.
BUG=chrome-os-partner:46049
BRANCH=none
TEST=tested on glados and samus
Change-Id: I3be639414e18586344e0ec84632a50dfc1df586b
Signed-off-by: Duncan Laurie <dlaurie@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/315221
Commit-Ready: Aaron Durbin <adurbin@chromium.org>
Tested-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-by: Shawn N <shawnn@chromium.org>
[task]
1. Copy task_wait_event_mask() function of cortex-m0.
[system]
2. Include host_command.h for host_command_pd_request_hibernate().
[i2c]
3. Update i2c handler to use task_wait_event_mask.
Signed-off-by: Dino Li <dino.li@ite.com.tw>
BRANCH=none
BUG=chromium:435611,chromium:435612
TEST=1. console commands: i2cscan, i2cxfer, and battery.
Change-Id: If5bb4407460d28c0b021ab133ca4b635ff7bc3c9
Reviewed-on: https://chromium-review.googlesource.com/320440
Commit-Ready: Dino Li <dino.li@ite.com.tw>
Tested-by: Dino Li <dino.li@ite.com.tw>
Reviewed-by: Randall Spangler <rspangler@chromium.org>
[symptom]
The 'vboot_hash_start()' always return busy error and variable 'in_progress'
got a strange value(should be 0 or 1).
'start_irq_handler()' causes scratchpad overflow in first context switch.
It must be called after SP switch to system stack in ISR.
NOTE:
The scratchpad is still also need more size even if
'start_irq_handler()' is using system stack. following is detail:
1. uint32_t scratchpad[19] 0x81d34 ~ 0x81d7f
[__task_start:]
2. /* put the dummy stack pointer at the top of the stack in scratchpad */
addi $sp, $r3, 4 * 18
-> SP 0x81d7c
3. syscall push return address (-4)
-> SP 0x81d78
[ISR:]
4. push r15, fp, lp, and sp (-0x10)
-> SP 0x81d68
5. push r0 ~ r5 (-0x18)
-> SP 0x81d50
[__switch_task:]
6. /* save ipsw, ipc, r6, r7, r8, r9, r10 on the current process stack */ (-0x1C)
-> SP 0x81d34
Signed-off-by: Dino Li <dino.li@ite.com.tw>
BRANCH=none
BUG=none
TEST=1. hash done.
2. console command 'taskinfo'.
3. the scratchpad does not overflow after first context switch.
Change-Id: If5d89ff5c945a777010492fcfb54bf41f434ed69
Reviewed-on: https://chromium-review.googlesource.com/317468
Commit-Ready: Dino Li <dino.li@ite.com.tw>
Tested-by: Dino Li <dino.li@ite.com.tw>
Reviewed-by: Randall Spangler <rspangler@chromium.org>
A typical EC image includes two similar in their functionality
subsections, RO and RW. CR50 has a small RO subsection, all it does -
detects a proper RW image to run and starts it up. To provide for
reliable firmware updates, the CR50 image needs to include two RW
sections, while the code is running from one RW subsection, the other
one can be upgraded.
This patch adds the ability to generate two identical RW sections,
mapped half flash size apart, and include them into the resulting EC
image.
To keep things simple the previously existing RW section's name is not
being changed, while the new (identical) RW section is named RW_B.
Two configuration options need to be defined to enable building of the
new image type: CONFIG_RW_B to enable the feature and
CONFIG_RW_B_MEM_OFF to define where RW_B should be mapped into the
flash.
A new rule added to Makefile.rules allows to generate a different lds
file from the same source (core/cortex-m/ec.lds.S) by defining a
compile time variable to pick a different base address for the
rewritable section, when RW_B is built.
BRANCH=none
BUG=chromium:43025
TEST=as follows:
- make buildall -j still succeeds
- verified that regular CR50 image starts successfully
- modified chip/g/loader/main.c to launch RW_B first, re-built and
re-run the image, observed on the console:
vvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvv
cr50 bootloader, 20151118_11218@80881, no USB, full crypto
Valid image found at 0x00084000, jumping
--- UART initialized after reboot ---
[Reset cause: power-on]
[Image: unknown, cr50_v1.1.4160-4c8a789-dirty 2015-12-07 18:54:27 vbendeb@eskimo.mtv.corp.google.com]
[0.001148 Inits done]
This FPGA image has no USB support
Console is enabled; type HELP for help.
> [0.002212 task 2 waiting for events...]
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(note that the image base address is 0x840000, which is RW_B).
Change-Id: Ia2f90d5e5b7a9f252ea3ecf3ff5babfad8a97444
Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/316703
Commit-Ready: Nicolas Boichat <drinkcat@chromium.org>
Reviewed-by: Aseda Aboagye <aaboagye@chromium.org>
Reviewed-by: Randall Spangler <rspangler@chromium.org>
On mutex contention, call task_wait_event_mask(), which will wait for
a mutex event without clearing other pending events.
BUG=chrome-os-partner:47918,chromium:435611,chromium:435612
BRANCH=None
TEST=Manual on snoball. Verify samus can successfully negotiate PD power
contract when attached to snoball.
Signed-off-by: Shawn Nematbakhsh <shawnn@chromium.org>
Change-Id: I85cd32f2670246da9e4787025390aba2c93f9c36
Reviewed-on: https://chromium-review.googlesource.com/314492
Commit-Ready: Shawn N <shawnn@chromium.org>
Tested-by: Shawn N <shawnn@chromium.org>
Reviewed-by: Alec Berg <alecaberg@chromium.org>
This patch introduces a facility which would allow to compile in
callbacks for arbitrary commands passed over various communication
protocols.
Typically this will be used for testing, when various test commands
are multiplexed over an existing protocol.
The callbacks are associated with 16 bit command codes. On input the
callback receives a buffer, containing the command's argument, the
size of the command argument and the maximum size of the buffer. On
output the callback stores processing result in the same buffer and
updates the size to the actual amount of returned data.
Callback descriptors are stored in a dedicated read only section which
is scanned by extension_route_command() to find a callback associated
with a certain command code.
A console channel is also being introduced to allow controlling
console output generated by extension commands handlers.
BRANCH=none
BUG=chrome-os-partner:47524
TEST=none yet
Change-Id: I8ae16a78ca7d72176a5e7f74dd7a232078e7c06c
Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/312586
Reviewed-by: Randall Spangler <rspangler@chromium.org>
The .bss.slow input section needs to be included before __bss_end,
otherwise it won't get cleared to zero.
BUG=None
BRANCH=None
TEST=Build jerry. Verify that __bss_end is at the end of the .bss.slow
section.
TEST=make -j buildall tests
Change-Id: I41d028ee166d05b34a889499cdcb0254341be1b6
Signed-off-by: Aseda Aboagye <aaboagye@google.com>
Reviewed-on: https://chromium-review.googlesource.com/312404
Commit-Ready: Aseda Aboagye <aaboagye@chromium.org>
Tested-by: Aseda Aboagye <aaboagye@chromium.org>
Reviewed-by: Randall Spangler <rspangler@chromium.org>
With the proper RO in place, RW must be signed in the same manner, as
RO. This patch makes sure that there is room in the RW header for the
signature.
BRANCH=none
BUG=chrome-os-partner:43025
TEST=with the rest of the patches applies the RO successfully boots up
the RW.
Change-Id: I1538195e0181c23c874ddd300887cf5da8c5a867
Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/311421
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
The point at which we reloaded the loader was too early. When items are
placed into .bss.slow via CONFIG_REPLACE_LOADER_WITH_BSS_SLOW, other
tasks could still access their variables that may have been in that
region after we had replaced those contents with the loader. This
commit moves the reloading of the loader to as late as possible once all
tasks have done their HOOK_SYSJUMP work.
Also, fixed a bug with the .bss.slow section. If a board is not using
the config option but items are placed in that section, that part of RAM
would not be cleared out.
BUG=chrome-os-partner:46056
BRANCH=None
TEST=Enable config option on GLaDOS and add a few variables to the
.bss.slow section. 'sysjump' between RO and RW and verify that no data
bus error is encountered.
TEST=make -j buildall tests
Change-Id: I3084700b9d5c144e86e2e408b72d2e3075a67413
Signed-off-by: Aseda Aboagye <aaboagye@google.com>
Reviewed-on: https://chromium-review.googlesource.com/306173
Commit-Ready: Aseda Aboagye <aaboagye@chromium.org>
Tested-by: Aseda Aboagye <aaboagye@chromium.org>
Reviewed-by: Randall Spangler <rspangler@chromium.org>
1. In combinational mode and clock source is 8MHz,
if timer 3 counter register always equals to 7, then timer 4 will be a
32-bit MHz free-running counter.
2. Fix TIMER_32P768K_CNT_TO_US(), each count should be 30.5175 us,
not 32.768us.
3. Fix TIMER_CNT_8M_32P768K().
4. Make sure LPC wake up interrupt is enabled before entering doze /
deep doze mode.
Signed-off-by: Dino Li <dino.li@ite.com.tw>
BRANCH=none
BUG=none
TEST=1. Console commands: 'gettime', 'timerinfo', 'waitms', and 'forcetime'.
2. Enabled Hook debug, no warning message received (48hrs).
3. Tested ectool command 'version' x 2000.
Change-Id: I796d985361d3c18bc5813c58705b41923e28c5b1
Reviewed-on: https://chromium-review.googlesource.com/310039
Commit-Ready: Dino Li <dino.li@ite.com.tw>
Tested-by: Dino Li <dino.li@ite.com.tw>
Reviewed-by: Randall Spangler <rspangler@chromium.org>
1. To enable INT_PRI (hardware interrupt priority level 0~3)
register, bit0@INT_CTRL = 0.
2. GIE need to be enabled before UART is initialized.
[symptom]
To define CONFIG_RWSIG / CONFIG_RSA / CONFIG_SHA256, then power on:
after RW image is verified, firmware stuck in uart_flush_output().
Signed-off-by: Dino Li <dino.li@ite.com.tw>
BRANCH=none
BUG=none
TEST=1. bit0@INT_CTRL = 0.
2. The RW image is verified and jump to image RW.
Change-Id: I393a3d5f87ea257885b872c91bfce43aecbaea8b
Reviewed-on: https://chromium-review.googlesource.com/309400
Commit-Ready: Dino Li <dino.li@ite.com.tw>
Tested-by: Dino Li <dino.li@ite.com.tw>
Reviewed-by: Randall Spangler <rspangler@chromium.org>
1. Implement deep doze mode for CONFIG_LOW_POWER_IDLE.
Signed-off-by: Dino Li <dino.li@ite.com.tw>
BRANCH=none
BUG=none
TEST=test the following items in deep doze mode.
1. WUI interrupts wake-up OK. (For example, power button, lid,
uart rx, keyboard ksi, and so on)
2. LPC access interrupt wake-up OK.
3. Enabled Hook debug, no warning message received (48hrs).
Change-Id: I8702a112632cb6c1c0fa75d682badf272130a7d4
Reviewed-on: https://chromium-review.googlesource.com/307060
Commit-Ready: Dino Li <dino.li@ite.com.tw>
Tested-by: Dino Li <dino.li@ite.com.tw>
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Certain ECs which cannot load both the RO and RW images into memory
simultaneously have a little firmware (LFW) loader that is loaded into
the beginning of the program memory. This loader is responsible for
loading the RO/RW images, however once we begin executing code from
main(), the loader has already completed it's purpose. We can reuse
this space by allocating portions of the .bss section here. This
currently saves us 3k for mec1322. This section is .bss.slow.
To use this feature simply define the following config option in
board.h.
#define CONFIG_REPLACE_LOADER_WITH_BSS_SLOW
In some cases, such as the mec1322, this region of RAM is optimized for
code and not data. Therefore, ideally infrequently used data should go
into this region.
BUG=chrome-os-partner:46056
BUG=chrome-os-partner:46063
BRANCH=None
TEST=Flash GLaDOS; sysjump rw; Verify that we are in rw; sysjump ro;
Verify we are back in ro.
TEST=Build GLaDOS, check output memory map for LDR_REGION and .bss.slow
section.
TEST=make -j buildall tests
Change-Id: I5b95cacc211278cf780a857b5a2ecfc1a0a28c45
Signed-off-by: Aseda Aboagye <aaboagye@google.com>
Reviewed-on: https://chromium-review.googlesource.com/305254
Commit-Ready: Aseda Aboagye <aaboagye@chromium.org>
Tested-by: Aseda Aboagye <aaboagye@chromium.org>
Reviewed-by: Andrey Petrov <andrey.petrov@intel.com>
Reviewed-by: Shawn N <shawnn@chromium.org>
Code for hard iron calibration: Every seconds (or faster if enough
samples), find a sphere that fit the compass data.
Based on Android code.
BRANCH=smaug
BUG=chrome-os-partner:39900
TEST=Check hard-iron bias is removed. Works better outside.
Change-Id: Iab479d5113b6560b4f01b0fd87373d2eecdb9b54
Signed-off-by: Gwendal Grignou <gwendal@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/299583
Reviewed-by: Anton Staaf <robotboy@chromium.org>
[chip config]
1. No hardware specific udelay().
2. Enable watchdog.
[watchdog]
3. Watchdog period is "CONFIG_WATCHDOG_PERIOD_MS" of config.h.
4. Watchdog auxiliary timer period is "CONFIG_AUX_TIMER_PERIOD_MS".
[task and irq]
5. Write 1 to clear interrupt pending status, no |.
6. A global variable for store interrupt number of software interrupt.
[uart]
7. Always reset UART module before config it.
[hwtimer]
8. Use more external timers for HW timer module.
[task]
9. Fix task profiling.
Signed-off-by: Dino Li <dino.li@ite.com.tw>
BRANCH=none
BUG=none
TEST=[watchdog]
1. console "waitms 1100", only pre-watchdog warning message.
2. console "waitms 1600", warning message and watchdog reset.
[hwtimer]
3. console commands "gettime", "timerinfo", and "forcetime".
4. enable hook debug and there is no delayed by more than 10%
warning message over 48 hours.
5. There is no watchdog reset too.
[task]
6. console 'taskinfo'
Task Ready Name Events Time (s) StkUsed
0 R << idle >> 00000000 32.927724 308/512
1 HOOKS 00000000 0.034267 372/768
2 R CONSOLE 00000000 0.116763 468/768
3 HOSTCMD 00000000 0.000641 372/512
4 KEYPROTO 00000000 0.000042 212/512
5 KEYSCAN 00000000 0.000908 356/512
IRQ counts by type:
38 2932
155 1
158 261
160 67
Service calls: 87
Total exceptions: 3348
Task switches: 167
Task switching started: 0.001999 s
Time in tasks: 33.282819 s
Time in exceptions: 0.164717 s
Change-Id: I234085cec231cd855d2a5e639ea1b0966c61d796
Reviewed-on: https://chromium-review.googlesource.com/296939
Commit-Ready: Dino Li <dino.li@ite.com.tw>
Tested-by: Dino Li <dino.li@ite.com.tw>
Reviewed-by: Randall Spangler <rspangler@chromium.org>
For signed EC RW images (CONFIG_RWSIG), there's no point in
embedding the public key or signature into the image itself since
it will just be replaced by the signer (either as the next step
in the build process, or after the fact for MP releases). This
takes that out and just points to where the pubkey and signature
will be placed.
BUG=none
BRANCH=none
TEST=make buildall
I also checked the signatures with
futility show -t build/*/ec.bin
They still look good, and the one signed image I booted (Cr50)
works as before.
Change-Id: Ib39b7c508914851f81a1bebb2450e08ef0def76c
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/302630
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Since CONFIG_RO_SIZE and CONFIG_RW_SIZE now exist (which may
theoretically be different sizes), it is no longer useful to globally
define the size of an image.
BUG=chromium:535027
BRANCH=None
TEST=`make buildall -j`. Also, verify glados / glados_pd continue to
function as expected.
Signed-off-by: Shawn Nematbakhsh <shawnn@chromium.org>
Change-Id: Ie29959923bc5d02b4d7d6d507ff2191bcb7d24c8
Reviewed-on: https://chromium-review.googlesource.com/301743
Commit-Ready: Shawn N <shawnn@chromium.org>
Tested-by: Shawn N <shawnn@chromium.org>
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
The board-specific board.h defines things like this:
#define CONFIG_RO_SIZE 1000
#define CONFIG_RW_SIZE 1000
#define CONFIG_SHAREDLIB_SIZE 200
And in the linker scripts we define some preprocessor macros like
so:
#define FW_SIZE_(section) CONFIG_##section##_SIZE
#define FW_SIZE(section) FW_SIZE_(section)
So that we can say things like this:
FLASH (rx) : ORIGIN = FW_OFF(SECTION), LENGTH = FW_SIZE(SECTION)
Note that we have to use FW_SIZE, not FW_SIZE_
The difference is only noticeable when SECTION is #defined. If
${CFLAGS} has
-DSECTION=RW
Then the expansion is this:
FW_SIZE_(SECTION) => CONFIG_SECTION_SIZE
FW_SIZE_(RW) => 1000
There's no difference in the output for this particular CL, but
we should use the correct macro anyway to avoid confusion.
BUG=none
BRANCH=none
TEST=make buildall
Change-Id: I61edc76a1aaeb1c675b384371700f37dda39ed47
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/302150
Reviewed-by: Aseda Aboagye <aaboagye@chromium.org>
