By design, the imd library still provdes dynamic growth so that
feature is consistent. The imd-based cbmem packs small allocations
into a larger entry using a tiered imd. The following examples show
the reduced fragmentation and reduced memory usage.
Before with dynamic cbmem:
CBMEM ROOT 0. 023ff000 00001000
aaaabbbb 1. 023fe000 00001000
aaaabbbc 2. 023fd000 00001000
aaaabbbe 3. 023fc000 00001000
aaaacccc 4. 023fa000 00002000
aaaacccd 5. 023f9000 00001000
ROMSTAGE 6. 023f8000 00001000
CONSOLE 7. 023d8000 00020000
COREBOOT 8. 023d6000 00002000
After with tiered imd:
IMD ROOT 0. 023ff000 00001000
IMD SMALL 1. 023fe000 00001000
aaaacccc 2. 023fc000 00001060
aaaacccd 3. 023fb000 000007cf
CONSOLE 4. 023db000 00020000
COREBOOT 5. 023d9000 00002000
IMD small region:
IMD ROOT 0. 023fec00 00000400
aaaabbbb 1. 023febe0 00000020
aaaabbbc 2. 023feba0 00000040
aaaabbbe 3. 023feb20 00000080
ROMSTAGE 4. 023feb00 00000004
Side note: this CL provides a basis for what hoops one needs to
jump through when there are not writeable global variables on
a particular platform in the early stages.
Change-Id: If770246caa64b274819e45a26e100b62b9f8d2db
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/9169
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marc.jones@se-eng.com>
elog breaks the build if ELOG_FLASH_BASE isn't configured -
and CHROMEOS isn't enabled, since with Chrome OS builds, it
just uses fmap to find out the base.
So it makes sense to enable it on all Chrome OS builds - if
the code never uses it, the linker will drop it soon enough.
Change-Id: I7ee129fadf75caf15fb9bd32b0acf6f7d9d015d8
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Reviewed-on: http://review.coreboot.org/9965
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
Instead of switching off the LED ring for the normal boot path, turn
it on with the dedicated pattern, indicating that the device firmware
has started executing.
BRANCH=storm
BUG=chrome-os-partner:39044
TEST=program the new firmware on an SP5 device and verify that the new
'normal boot' pattern is displayed in all following scenarios:
- power up the device (the pattern shows within the first second)
- boot the device to bash prompt and restart.
- press the recovery button and reboot the device. Once the LED ring
shows the 'recovery button pressed' pattern, release the recovery
button.
Change-Id: Iedd66d2575ad587af77e35d23efb22a83a92858b
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: 092358bd17a323d0bd2fec555f43cb587486bbec
Original-Change-Id: I24de45326eab83b57bcf16b5598388f81c7d0f00
Original-Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/265536
Original-Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/9923
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
It became necessary to indicate the beginning of the normal boot
process. This patch adds a new pattern, a slow (over 2 seconds) fade
in into the 0, 87, 155 color.
BRANCH=storm
BUG=chrome-os-partner:39044
TEST=tested by the next patch.
Change-Id: Idd977688e5aa2cc55fc295072c0766526ae95016
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: 577c8bd6f8c69073cfdd7acd4a87e7ae603d48e6
Original-Change-Id: I9aff3f4558e733ff2e47206075533556e400f183
Original-Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/265535
Original-Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/9922
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
After testing on a final assembly the PD team adjusted the wipeout
request and recovery request modes' colors.
BRANCH=storm
BUG=none
TEST=verified new colors while booting an SP5 device in recovery mode
Change-Id: I9bd2dac63b99140573533c2cda8eaa9213478ab1
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: 41c34a619dc0317af67907f18ee844c71a73d623
Original-Change-Id: Iab84710ebdeed35ddd4a8a163bbb6b8ac9cdb799
Original-Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/262602
Original-Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/9890
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
Modify colors as suggested by product review folks. This is not final,
to make it easier to identify RGB locations in the hex dumps, express
their values in decimal as opposed to hex.
BRANCH=storm
BUG=chrome-os-partner:36059
TEST=verified new all three color schemes while pressing the recovery
button at boot for 20 seconds.
Change-Id: I7461acd7004e3d10cba6665a9bfe25ec8aa6f3ba
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: 7a075824a1954eb5d1b65ce887304924724a6d21
Original-Change-Id: I7f5968e361333572fd1f84aa11b7150194ad902a
Original-Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/261690
Original-Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/9880
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
Storm and whirlwind devices should leave developer mode as soon as
recovery is requested.
BRANCH=storm
BUG=chrome-os-partner:36059
TEST=with the rest of the patches applies observed desired behavior on
SP5
Change-Id: I3e8e481f85cd067eff4fe2049b8aa47e09f6d63e
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: c6dbafc16edb11ec687359b25098ce8a1b055b91
Original-Change-Id: Iac6fa62229556bc7a6960d7f1630d37570ba72c4
Original-Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/261621
Original-Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/9879
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
The whirlwind device is using a button instead of the switch to enable
dev mode.
BRANCH=storm
BUG=chrome-os-partner:36059
TEST=with the rest of the patches applied it is possible to enable and
disable dev mode on Whirlwind and the mode persists over reboots.
Change-Id: I6d31c8429ea0515bea2b7d707325d9092487048a
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: 495bd07e3a1d6c587a5f01efa2801d2c4e7edbe8
Original-Change-Id: I8879833ca13074e8275393c3b98df06d56a5361d
Original-Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/261416
Original-Reviewed-by: David Hendricks <dhendrix@chromium.org>
Reviewed-on: http://review.coreboot.org/9875
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
The originally loaded blinking program was written to allow gradual
change in LED brightness, which required controlling each LED with its
own engine. In fact there is no need in gradual brightness changes
when the firmware is controlling the ring. This allows to control all
LEDs by one engine, making the code simpler and more robust (no need
to synchronize the three engines any more).
BRANCH=storm
BUG=chrome-os-partner:36059
TEST=verified that recovery boot WW ring patterns work as expected.
Change-Id: I89d231fb61693f4e834d8d9323ae5a7ddd149525
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: 19809cf8120df8865da9b5b9e7b8e932334bf4b5
Original-Change-Id: I41038fd976dc9600f223dc0e9c9602331baf68f9
Original-Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/261026
Original-Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/9873
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
The two controllers on the ring are programmed independently, and if
the second controller is running the old pattern while the first one
was loaded with a new pattern, there is a window of when the two
unrelated patterns might interact.
To avoid this shut down execution on both controllers before starting
downloading the new pattern code.
BRANCH=storm
BUG=chrome-os-partner:36059
TEST=verified recovery/wipeout LED ring behavior did not change.
Change-Id: I163f2983d414fe839208054ae3e9025663a46aeb
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: 3502ca6b119c033855b45388e7b782d35cfdd82b
Original-Change-Id: I0f71f94a7e82f6c0e7f98d3aad1f93ece207248f
Original-Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/261200
Original-Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/9872
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
Add compiled lp55231 code snippets to allow display certain patterns
when booting the device with the recovery button pressed.
As soon as the press is detected, the low intensify solid white
pattern is enabled. Holding recovery button long enough causes the
device transition between the wipeout requested and recovery requested
states, with the appropriate changes in the displayed pattern.
The patch also includes the source code for the LED controller as well
as instructions on how to compile and modify the code to result in
different colors, intensities, blink periods and duty cycles.
BRANCH=storm
BUG=chrome-os-partner:36059
TEST=reboot an SP5 device with the LED ring attached, keep the
recovery button pressed, observe the changes in the LED display
pattern while the device progresses through the boot sequence.
Change-Id: Ic7d45fc7c313b6d21119d4ae6adaeb4f46f7d181
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: 0fd6a5c0067d705197816629f41640a931d2f7cd
Original-Change-Id: Ib5cc5188c2eeedbba128101bf4092a0b9a74e155
Original-Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/260670
Original-Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/9870
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
The patterns displayed on the LED ring while under the coreboot
control are not driven by the vboot, but by the board code instead,
The four distinct states of the LED display are:
- all off
- recovery button push detected, waiting for it to be released
- wipeout request pending - recovery button was pushed long enough
to trigger this request
- recovery request pending - recovery button was pushed long enough
to trigger this request.
BRANCH=storm
BUG=chrome-os-partner:36059
TEST=no functional changes
Change-Id: I38d9a3028013b902a7a67ccd4eb1c5d533bf071c
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: bdfff0e646283da6a2faaacf33e0179d2fea221c
Original-Change-Id: Ie279151b6060a2888268a2e9a0d4dc22ecaba460
Original-Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/260649
Original-Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/9868
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
When in development environment, some SP5 devices might not have the
LED ring attached. They are still fully functional, but when booting
up are generating massive amount of i2c error messages. This patch
prevents accesses to non-existing lp55321 devices.
When loading the program into the device the vendor recommends 1 ms
delay when accessing the program control register. This patch
separates these accesses into a function and add a delay after every
access.
Another fix - advance the program address when loading multipage
programs.
Set the global variable register 3c, not used by coreboot programs, to
a fixed value. This will allow depthcharge to avoid re-initializing
the controller when not necessary.
BRANCH=storm
BUG=chrome-os-partner:36059
TEST=booted firmware on an SP5 with no LED ring attached, no excessive
error messages are generated, saw the default pattern displayed
when the recovery button is pressed during reset.
Change-Id: I6a2a27968684c40dae15317540a16405b1419e30
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: 5e0b4c84aca27460db594da1faf627ddee56f399
Original-Change-Id: I10f1f53cefb866d11ecf76ea48f74131d8b0ce77
Original-Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/260648
Original-Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/9867
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
The recovery switch on storm is overloaded: it needs to be pressed for
a certain duration at startup to signal different requests:
- keeping it pressed for 8 to 16 seconds after startup signals the need for
factory reset (wipeout);
- keeping it pressed for longer than 16 seconds signals the need for Chrome
OS recovery.
This patch adds a function to report the wipeout request status and
enables the new feature on Storm.
BRANCH=storm
BUG=chrome-os-partner:37219
TEST=verified that keeping the recovery button pressed between 8 and
16 seconds at startup results in the wipeout request generated
(crossystem 'wipeout_request' returns 1). Keeping the button
pressed for more than 16 seconds triggers recovery mode.
Change-Id: I17131593e12833866a22837271feb0e6989e6750
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: 3c503ec13c2b096d4a21fb299c0dd0396f1d01e9
Original-Change-Id: Ic3678217906e56307d47378fa8a6defeb314084e
Original-Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/259844
Reviewed-on: http://review.coreboot.org/9863
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
The latest whirlwind requirements call for the need to indicate
different device states while it is still in coreboot (it could be
waiting for recovery or for factory reset trigger).
Initialize the LED ring when running on the SP5 hardware (which is the
first proper Whirlwind device).
BRANCH=storm
BUG=chrome-os-partner:36059
TEST=when the device starts the LED ring gets shut down
Change-Id: I9dd0bca4849a2a8500322c84c7351aeef00d862e
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: 24e4da56d4c43d03f235d0cfd5995ef235e6a2c5
Original-Change-Id: Ica37301aa27f35897d2bf467ae319fb5e68adc1d
Original-Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/258271
Original-Reviewed-by: David Hendricks <dhendrix@chromium.org>
Reviewed-on: http://review.coreboot.org/9859
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
This is a copy of the depthcharge ww ring driver implementation ported
into coreboot. The main differences are:
- direct use of the i2c driver instead of using the callback driver
description
- no dynamic memory allocation for the controller structures
BRANCH=storm
BUG=chrome-os-partner:36059
TEST=with the rest of the patches applied the LED ring gets
initialized to the default pattern at coreboot start.
Change-Id: I6902c8b76fc173ad2ec28b8cc94695e892df338a
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: eda24b78f8aff311dd6296d458bdfecf26c3d65a
Original-Change-Id: I5660dc3f255aab8fbe3a87041c72916a645c193b
Original-Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/257730
Original-Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/9858
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
abuild -x (we're running out of letters) builds with CHROMEOS enabled.
Change-Id: Ie9abd8aa999dd339aab113ff28c16671b2a17845
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Reviewed-on: http://review.coreboot.org/9966
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
abuild only created compile.status for successful builds,
but sometimes it's helpful to easily identify all failed
builds of a full run:
$ grep -l failed coreboot-builds/*/compile.status
Change-Id: Ic90280fb2e8cff1f8f558a2e67ffad741beddbdf
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Reviewed-on: http://review.coreboot.org/9964
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Some compilers and linkers require a strict order or fail to find
all symbols.
Change-Id: I3f44bec1f0e21e7313a751fbc99c61c1aa9b7cf1
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Reviewed-on: http://review.coreboot.org/9962
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
For boards with MAINBOARD_HAS_CHROMEOS, we should also
state what kind of storage is available for vboot's
non-volatile data.
The flags are taken from the chromium repository and
have no effect with CHROMEOS disabled.
Change-Id: I1747ad26c8c7f6d4076740ec2800dbd52c5d6b3d
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Reviewed-on: http://review.coreboot.org/9952
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Since CHROMEOS_VBNV_* are selected by mainboards, they
may be active without CHROMEOS being selected. In this
case, they should be a no-op.
Change-Id: I3b84e2a919ffaa809d713e72e5e4df7a7575e6b9
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Reviewed-on: http://review.coreboot.org/9954
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
Many chipsets were using a stage cache for reference code
or when using a relocatable ramstage. Provide a common
API for the chipsets to use while reducing code duplication.
Change-Id: Ia36efa169fe6bd8a3dbe07bf57a9729c7edbdd46
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/8625
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marc.jones@se-eng.com>
A tiered imd allows for both small and large allocations. The
small allocations are packed into a large region. Utilizing a
tiered imd reduces internal fragmentation within the imd.
Change-Id: I0bcd6473aacbc714844815b24d77cb5c542abdd0
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/8623
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marc.jones@se-eng.com>
The imd (internal memory database) library provides a way to
track memory regions by assigning ids to each region. The implementation
is a direct descendant of dynamic cbmem. The intent is to replace
the existing mechanisms which do similar things: dynamic cbmem, stage
cache, etc.
Differences between dynamic cbmem and imd:
- All structures/objects are relative to one another. There
are no absolute pointers serialized to memory.
- Allow limiting the size of the idm. i.e. provide a maximum
memory usage.
- Allow setting the size of the root structure which allows
control of the number of allocations to track.
Change-Id: Id7438cff80d396a594d6a7330d09b45bb4fedf2e
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/8621
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marc.jones@se-eng.com>
This fixes some compilation issues observed with CONFIG_CHROMEOS.
Nothing within the vbootX subdirectories is functional yet, but
a partial compilation within the chromeos direction works now.
Notable fixes: duplicate definitions and missing prototypes.
Change-Id: I53c7b6dcf06b8bcf41a8555094b48968c0740026
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/9936
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
They were keyed to VBOOT_VERIFY_FIRMWARE which made them invisible
under some circumstances.
Change-Id: I61c56b4d245351fae0ec14f80bcd17ba93184651
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Reviewed-on: http://review.coreboot.org/9956
Tested-by: build bot (Jenkins)
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
When RTC is not selected, return all 0.
Change-Id: I892a9489fc1d82fb8e61cf02666f797dc6412e05
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Reviewed-on: http://review.coreboot.org/9955
Tested-by: build bot (Jenkins)
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
RTC drivers now select RTC, so that code which depends on them
can implement fallback behavior for systems that lack the
hardware or driver.
Change-Id: I0f5a15d643b0c45c511f1151a98e071b4155fb5a
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Reviewed-on: http://review.coreboot.org/9953
Tested-by: build bot (Jenkins)
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
They were already moved to src/lib/bootmode.c in
commit 5687fc9 Declare recovery and developer modes outside ChromeOS
Change-Id: Ia27a0c79baa364ce3779a8a699e9246d26d02ecb
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Reviewed-on: http://review.coreboot.org/9951
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Some recently upstreamed boards are missing this flag
Change-Id: I89d73970f23eed6ea127e620c38f9687b2f5b048
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Reviewed-on: http://review.coreboot.org/9949
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
The preprocessor flags that are manipulated in that line are
managed exclusively in CPPFLAGS since commit 58f73a69.
Change-Id: I2263401a292b4f7435659b24cf4f695a927015ef
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Reviewed-on: http://review.coreboot.org/9948
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
When using qemu-armv7 to load coreboot.rom with the -kernel
flag the rom is offset by 0x10000. Therefore only allow
mappings within 0x10000 and 0x10000 + CONFIG_ROM_SIZE.
TEST= QEMU_AUDIO_DRV=none qemu-system-arm -M vexpress-a9 \
-m 1024M -nographic \
-kernel coreboot-builds/emulation_qemu-armv7/coreboot.rom
Change-Id: Ifec5761a7d54685f664c54efaa31949b8cc94bad
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/9935
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
Instead of always loading romstage from cbfs provide a
way, similar to ramstage and payload, for other
program loaders to intervene. For now, only the cbfs
loader is consulted.
TEST=Booted to end of ramstage on qemu-armv7
Change-Id: I87c3e2e566d7a0723e775aa427de58af745ecdd5
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/9934
Tested-by: build bot (Jenkins)
Reviewed-by: Edward O'Callaghan <edward.ocallaghan@koparo.com>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
It is necessary to trigger console buffer contents dump on reset.
Let's make sure all vboot resets are routed through the same function.
BRANCH=none
BUG=chromium:475347
TEST=built and booted storm
Change-Id: I0d8580fb65417ba4b06dfae763dd6455afc8fc26
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: 9788e2043cb1bd5df7e30574f7df4de4f25caa0d
Original-Change-Id: Iafca416700c51a0546249438ca583a415a1ca944
Original-Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/265292
Original-Reviewed-by: Julius Werner <jwerner@chromium.org>
Reviewed-on: http://review.coreboot.org/9931
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
Enable auto entry and auto exit self-refresh.
Configure entry idle time to 16x long count sequences.
Where a long count sequence is 1024 cycles.
The idle entry configuration is based on 32x of the DLL lock time (512 cycles).
A conservative setting to help minimize self-refresh enter/exit thrashing.
BUG=chrome-os-partner:36456
BRANCH=broadcom-firmware
TEST=When enable configuration CYGNUS_SDRAM_TEST_DDR,
print on console:
sdram initialization is completed.
test ddr start from 0x60000000 to 0x80000000
...
test ddr end: fail=0
Translation table is @ 02004000
Mapping address range [0x00000000:0x00000000) as uncached
Change-Id: Ibad220429fd52ead2933db03bec1a555f9385e53
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: 3768f82ca268fb854f8c4753916518a1efdf887d
Original-Reviewed-on: https://chrome-internal-review.googlesource.com/212125
Original-Reviewed-by: Scott Branden <sbranden@broadcom.com>
Original-Reviewed-by: Daisuke Nojiri <dnojiri@google.com>
Original-Commit-Queue: Daisuke Nojiri <dnojiri@google.com>
Original-Tested-by: Daisuke Nojiri <dnojiri@google.com>
Original-Signed-off-by: Icarus Chau <ichau@broadcom.com>
Original-Change-Id: Icac1e12745d048b32e1804a546f6b49c8b5953c0
Original-Reviewed-on: https://chromium-review.googlesource.com/265862
Original-Reviewed-by: Daisuke Nojiri <dnojiri@chromium.org>
Original-Commit-Queue: Daisuke Nojiri <dnojiri@chromium.org>
Original-Trybot-Ready: Daisuke Nojiri <dnojiri@chromium.org>
Original-Tested-by: Daisuke Nojiri <dnojiri@chromium.org>
Reviewed-on: http://review.coreboot.org/9930
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
Currently (EVT) this is a pullup resistor on the board (i.e. always in
dev mode). Future builds it will be pull down and require servo or HW
modification to control. Either way, this change means the FW should
acknowledge it.
BUG=chrome-os-partner:38663
TEST=Manually verified that servo devmode switch toggles this GPIO. requires FW signing to verify GPIO is observed and dev mode active.
BRANCH=none
Change-Id: Ib05216992abc5f6175fe7395471bd379f185b61f
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: 702c8d222a3d19d6b8db89d122dcdf594c85da99
Original-Change-Id: I1d0f31819b9f7a1ab63deac52bcaf0b996499b0c
Original-Reviewed-on: https://chromium-review.googlesource.com/263529
Original-Reviewed-by: Hung-Te Lin <hungte@chromium.org>
Original-Tested-by: Jonathan Dixon <joth@chromium.org>
Original-Commit-Queue: Jonathan Dixon <joth@chromium.org>
Reviewed-on: http://review.coreboot.org/9928
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
Disable and enable GIC before switching off a CPU and after bringing
it up back respectively.
BUG=None
BRANCH=None
TEST=Compiles successfully and psci commands work for ryu.
Change-Id: Ib43af60e994e3d072e897a59595775d0b2dcef83
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: d5271d731f0a569583c2b32ef6726dadbfa846d3
Original-Change-Id: I672945fcb0ff416008a1aad5ed625cfa91bb9cbd
Original-Signed-off-by: Furquan Shaikh <furquan@google.com>
Original-Reviewed-on: https://chromium-review.googlesource.com/265623
Original-Trybot-Ready: Furquan Shaikh <furquan@chromium.org>
Original-Tested-by: Furquan Shaikh <furquan@chromium.org>
Original-Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Original-Commit-Queue: Furquan Shaikh <furquan@chromium.org>
Reviewed-on: http://review.coreboot.org/9926
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
Tested-by: build bot (Jenkins)
PSCI_CPU_OFF is SMC32 call, there is not SMC64 version. Register SMC32
and SMC64 types of PSCI calls.
BUG=None
BRANCH=None
TEST=Compiles successfully and CPU off works fine with PSCI command.
Change-Id: I8df2eabfff52924625426b3607720c5219d38b58
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: 9228c07f9d9a4dd6325afb1f64b41b9b8711b146
Original-Change-Id: I2f387291893c1acf40bb6aa26f3d2ee8d5d843ea
Original-Signed-off-by: Furquan Shaikh <furquan@google.com>
Original-Reviewed-on: https://chromium-review.googlesource.com/265622
Original-Trybot-Ready: Furquan Shaikh <furquan@chromium.org>
Original-Tested-by: Furquan Shaikh <furquan@chromium.org>
Original-Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Original-Commit-Queue: Furquan Shaikh <furquan@chromium.org>
Reviewed-on: http://review.coreboot.org/9925
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
We support SMC32 calls from AARCH64, however we do not support SMC32
calls from AARCH32. Reflect this policy in the code by using
appropriate names for exception type check in SMC handler.
BUG=None
BRANCH=None
TEST=Compiles successfully and able to turn CPU1 on and off using psci commands.
Change-Id: Ifc3c9e2fe0c4e6e395f2647769a2d07f5f41f57f
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: cbaf712c2c45273a9eb0b0808a0d4d0630023fdd
Original-Change-Id: I133b2c0bbc4968401a028382532bd051d6298802
Original-Signed-off-by: Furquan Shaikh <furquan@google.com>
Original-Reviewed-on: https://chromium-review.googlesource.com/265621
Original-Trybot-Ready: Furquan Shaikh <furquan@chromium.org>
Original-Tested-by: Furquan Shaikh <furquan@chromium.org>
Original-Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Original-Commit-Queue: Furquan Shaikh <furquan@chromium.org>
Reviewed-on: http://review.coreboot.org/9924
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
CNTFRQ_EL0 can only be set in highest implemented exception level.
Save and restore CNTFRQ_EL0 for secondary cpus in coreboot.
This patch fix the error below:
SANITY CHECK: Unexpected variation in cntfrq. Boot CPU:
0x00000000c65d40, CPU1: 0x00000000000000
BRANCH=none
BUG=none
TEST=boot to kernel on oak board and check secondary cpu's cntfrq.
confirmed cpu1's cntfrq is same as boot cpu's.
Change-Id: I9fbc3c82c2544f0b59ec34b1d631dadf4b9d40eb
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: b47e4e649efc7f79f016522c7d8a240f98225598
Original-Signed-off-by: Jimmy Huang <jimmy.huang@mediatek.com>
Original-Change-Id: I2d71b0ccfe42e8a30cd1367d10b0f8993431ef8c
Original-Reviewed-on: https://chromium-review.googlesource.com/264914
Original-Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/9921
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
This patch allows a board without a secdata storage (typically TPM) to pass
the verification stage if recovery path is taken. It's useful for bringup
when the actual board is not ready.
BUG=none
BRANCH=none
TEST=booted the kernel from a usb stick on a cygnus reference board
Change-Id: I5ab97d1198057d102a1708338d71c606fe106c75
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: 5d45acee31fd5b7bfe7444f12e3622bae49fc329
Original-Signed-off-by: Daisuke Nojiri <dnojiri@chromium.org>
Original-Reviewed-on: https://chrome-internal-review.googlesource.com/212418
Original-Reviewed-by: Daisuke Nojiri <dnojiri@google.com>
Original-Commit-Queue: Daisuke Nojiri <dnojiri@google.com>
Original-Tested-by: Daisuke Nojiri <dnojiri@google.com>
Original-Change-Id: Iddd9af19a2b6428704254af0c17b642e7a976fb8
Original-Reviewed-on: https://chromium-review.googlesource.com/265046
Reviewed-on: http://review.coreboot.org/9919
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
When the PHY is compiled to run in HDR(half data rate),
then either NOBUB or FXDAT must be set to 1 in the DDR
system general configuration register. NOBUB specifies
that reads should be returned to the controller with
no bubbles and this is felt preferable to the fixed
latency option (FXDAT). Both of them inrease read
latency.
BRANCH=none
BUG=chrome-os-partner:37087
TEST=tested on Pistachio bring up board -> DDR initialized
properly and ramstage executed correctly
Change-Id: Iee530ba5bb0acc889fba447dc2ee5cb965ba6926
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: e7944b4af45d9504098f8b4af44d0f5abafea42c
Original-Change-Id: I9ced76bd670fc4efa7441d57e15f97871b046ae9
Original-Signed-off-by: Ionela Voinescu <ionela.voinescu@imgtec.com>
Original-Reviewed-on: https://chromium-review.googlesource.com/264341
Original-Reviewed-by: James Hartley <james.hartley@imgtec.com>
Original-Reviewed-by: David Hendricks <dhendrix@chromium.org>
Reviewed-on: http://review.coreboot.org/9917
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
The DRAM configuration register, apart from holding the
device density and width also has a rudimentary address
mapping scheme. Currently this is set to the default
Bank/Row/Column. This means that the memory is segmented
into 8 chunks, each with a page detector. If all the
activity is in one section of memory then the other 7
page detectors could be idle.
Changing this to Row/Bank/Column would concatenate the
page detectors meaning that all 8 could be used by a
single initiator. This may not gain anything in a
synthetic bandwidth test but could yield extra performance
in a real world application or benchmark.
BRANCH=none
BUG=chrome-os-partner:37087
TEST=tested on Pistachio bring up board -> DDR initialized
properly; all access to DDR works properly in
Coreboot ramstage, Depthcharge and Linux;
no performance tests were ran so far.
Change-Id: I22d86bf3b679ed63884d7436d9d7bbaf1726f640
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: e852ed42afcdc2062a0037144bab723227cb1f1f
Original-Change-Id: If90b0cf5ce86db5e3d6d362873d22d4269e3a49f
Original-Signed-off-by: Ionela Voinescu <ionela.voinescu@imgtec.com>
Original-Reviewed-on: https://chromium-review.googlesource.com/264340
Original-Reviewed-by: James Hartley <james.hartley@imgtec.com>
Original-Reviewed-by: David Hendricks <dhendrix@chromium.org>
Reviewed-on: http://review.coreboot.org/9916
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
Aaron Durbin