FLASH_DUAL_READ is a common configuration for all MediaTek SoCs, so we
move it to common folder and select it in SoCs' Kconfig.
As suggested in CB:58837, we also rename FLASH_DUAL_READ to
FLASH_DUAL_IO_READ to reduce confusion.
TEST=build pass
BUG=b:233720142
Signed-off-by: Bo-Chen Chen <rex-bc.chen@mediatek.com>
Change-Id: If267a332519412a7919c5b7817047fabe4a564c8
Reviewed-on: https://review.coreboot.org/c/coreboot/+/65620
Reviewed-by: Yidi Lin <yidilin@chromium.org>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
gpio_set_pull(), gpio_set_pull_pu_pd() and gpio_set_spec_pull_pupd()
can be reused for mt8192, mt8195 and mt8186, so move it to new file
"gpio_op.c" in common folder.
TEST=build pass
BUG=b:233720142
Signed-off-by: Bo-Chen Chen <rex-bc.chen@mediatek.com>
Change-Id: I81ab9b01ee20fccf3ef29c5902597b5045d3e36a
Reviewed-on: https://review.coreboot.org/c/coreboot/+/65641
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Reviewed-by: Yidi Lin <yidilin@chromium.org>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
timer_prepare() is the same for MT8195 and MT8186, so move it to
common folder.
TEST=build pass
BUG=b:233720142
Signed-off-by: Bo-Chen Chen <rex-bc.chen@mediatek.com>
Change-Id: I91a6f4ecc665a058cb7a0ba96c15b27d6dc97d13
Reviewed-on: https://review.coreboot.org/c/coreboot/+/65602
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Reviewed-by: Yidi Lin <yidilin@chromium.org>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Add watchdog support for MT8188.
This implementation is based on chapter 3.10.10 in MT8188 Functional
Specification.
TEST=build pass
BUG=b:233720142
Signed-off-by: Bo-Chen Chen <rex-bc.chen@mediatek.com>
Change-Id: Iaf56c78d89af53d0272583b463c050e69bbeb07a
Reviewed-on: https://review.coreboot.org/c/coreboot/+/65587
Reviewed-by: Yidi Lin <yidilin@chromium.org>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
There are more and more variables which are SoC-specific, so add
soc/wdt.h for each SoC and rename common/wdt.h to
common/wdt_common.h.
wdt_set_req() is almost the same for mt8192, mt8195 and mt8186, so
move it to a common file wdt_req.c.
TEST=build pass
BUG=b:233720142
Signed-off-by: Bo-Chen Chen <rex-bc.chen@mediatek.com>
Change-Id: I7a334b3e7cd4f24a848dd31aca546dc7236d5fb8
Reviewed-on: https://review.coreboot.org/c/coreboot/+/65636
Reviewed-by: Yidi Lin <yidilin@chromium.org>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Add new folder and basic drivers for Mediatek SoC 'MT8188'.
Difference of modules including in this patch between MT8188 and
existing SoCs:
Timer:
Similar to MT8195 and MT8186, MT8188 uses v2 timer.
EMI/PLL/SPI:
Different from existing SoCs.
The implementation is based on these files:
MT8188G_Application Processor Technical Brief_v0.4.pdf
MT8188G_Functional Specification v0.4.pdf
MT8188 Application Processor Registers-1.pdf
MT8188 Application Processor Registers-2.pdf
TEST=saw the coreboot uart log to bootblock
BUG=b:233720142
Signed-off-by: Bo-Chen Chen <rex-bc.chen@mediatek.com>
Change-Id: I3320f3d49a9b9ed781ceb812e4341e379db4ac20
Reviewed-on: https://review.coreboot.org/c/coreboot/+/65585
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yidi Lin <yidilin@chromium.org>
Most of the src/soc/Kconfig files are only there for AMD and Intel to
load the main SoC Kconfig files before any common files. That can be
done in src/Kconfig instead. Moving the loads to the lower level allows
the removal of all but the Intel soc/Kconfig file, which can be removed
in a follow-on patch.
Signed-off-by: Martin Roth <gaumless@gmail.com>
Change-Id: I5061191fe23e0b7c745e90874bd7b390806bbcfa
Reviewed-on: https://review.coreboot.org/c/coreboot/+/65327
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Felix Singer <felixsinger@posteo.net>
Reviewed-by: Elyes Haouas <ehaouas@noos.fr>
Clean up Makefile.inc by sorting entries and moving common entries to
all-y. In this way it is more clear to know what drivers have been
involved in each stage and the hardware differences between each SoC.
BUG=none
TEST=emerge-corsola coreboot
TEST=emerge-asurada coreboot
TEST=emerge-cherry coreboot
Change-Id: Idfc7de36ebf36650f7c6bd1584ef77e2a540cde9
Signed-off-by: Yidi Lin <yidilin@chromium.org>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/65315
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Reviewed-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
There is a lot of going back-and-forth with the KiB arguments, start
the work to migrate away from this.
Change-Id: I329864d36137e9a99b5640f4f504c45a02060a40
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/64658
Reviewed-by: Werner Zeh <werner.zeh@siemens.com>
Reviewed-by: Arthur Heymans <arthur@aheymans.xyz>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Some eMMCs (for example, Kingston-EMMC64G-TX29-HP) may enter the ready
state by sending CMD1 twice. If it is in the ready state, then the
payload (for example, depthcharge) will not send CMD1, but the access
mode is only available from the response of CMD1.
Therefore, we need to pass the access mode to the payload by defining
the following types:
- MMC_STATUS_CMD1_READY: in ready state and access mode is byte mode.
- MMC_STATUS_CMD1_READY_HCS: in ready state and access mode is sector
mode.
BUG=b:234672726
BRANCH=cherry
TEST=boot ok
Signed-off-by: Wenbin Mei <wenbin.mei@mediatek.com>
Change-Id: Iad905781d8ba0105911cf87a6b845cd8df57521e
Reviewed-on: https://review.coreboot.org/c/coreboot/+/65054
Reviewed-by: Yidi Lin <yidilin@chromium.org>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
To simplify the calling sequence for mtk_wdt_preinit() and we always
adjust request setting in mtk_wdt_preinit(), we rename
mtk_wdt_preinit() to mtk_wdt_set_req() and call it in mtk_wdt_init().
From this modification, we can also enable thermal hardware reset
feature (CB:64676, CB:64675) in MT8192 and MT8195.
BUG=none
TEST=build pass
Signed-off-by: Bo-Chen Chen <rex-bc.chen@mediatek.com>
Change-Id: I1904ff9387f7677a077068f2c3df923bd642ea3d
Reviewed-on: https://review.coreboot.org/c/coreboot/+/64861
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Under the current watchdog setting, the system will not reboot when the
temperature is too high. To enable thermal hardware reset, we need to
enable thermal control request and set it to reboot mode.
Note that because thermal throttle (by lowering cpu frequency) is
currently enabled, the thermal hardware reset shouldn't be triggered
under normal circumstances.
This feature is only for new hardware structure for thermal. Therefore,
we only need to apply it on MT8192/MT8195/MT8186.
This setting is based on thermal and watchdog section of MT8192
Function Specification.
BUG=none
TEST=build pass
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: I98b062c2070384527624c3bcf0dfded25a2c8ce4
Reviewed-on: https://review.coreboot.org/c/coreboot/+/64676
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Under the current watchdog setting, the system will not reboot when the
temperature is too high. To enable thermal hardware reset, we need to
enable thermal control request and set it to reboot mode.
Note that because thermal throttle (by lowering cpu frequency) is
currently enabled, the thermal hardware reset shouldn't be triggered
under normal circumstances.
This feature is only for new hardware structure for thermal. Therefore,
we only need to apply it on MT8192/MT8195/MT8186.
This setting is based on thermal and watchdog section of MT8195
Function Specification.
BUG=none
TEST=build pass
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: Ia6489bb953d148a43af173454d6f2b3e2a1dfcf9
Reviewed-on: https://review.coreboot.org/c/coreboot/+/64675
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Under the current watchdog setting, the system will not reboot when the
temperature is too high. To enable thermal hardware reset, we need to
enable thermal control request and set it to reboot mode.
Note that because thermal throttle (by lowering cpu frequency) is
currently enabled, the thermal hardware reset shouldn't be triggered
under normal circumstances.
This feature is only for new hardware structure for thermal. Therefore,
we only need to apply it on MT8192/MT8195/MT8186.
This setting is based on thermal and watchdog section of MT8186
Function Specification.
BUG=none
TEST=emerge-corsola coreboot
TEST=thermal hardware reset is working.
Signed-off-by: Runyang Chen <runyang.chen@mediatek.corp-partner.google.com>
Change-Id: Id2ed55e6d4f4eec450bf7c849f726a389eeb6694
Reviewed-on: https://review.coreboot.org/c/coreboot/+/64659
Reviewed-by: Bo-Chen Chen <rex-bc.chen@mediatek.com>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
SCP core 2 is enabled for MT8195 camera feature. It requires the same
register access permission as SCP core 1. Therefore, we configure the
same domain ID for both cores.
BRANCH=cherry
BUG=b:193814857
TEST=cherry boot ok
Signed-off-by: Tinghan Shen <tinghan.shen@mediatek.com>
Change-Id: Idf335593936b12c083c926a252fa99c3b76cda6a
Reviewed-on: https://review.coreboot.org/c/coreboot/+/64575
Reviewed-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Reviewed-by: Yidi Lin <yidilin@chromium.org>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
There are efforts to have bootflows that do not follow a traditional
bootblock-romstage-postcar-ramstage model. As part of that CBMEM
initialisation hooks will need to move from romstage to bootblock.
The interface towards platforms and drivers will change to use one of
CBMEM_CREATION_HOOK() or CBMEM_READY_HOOK(). Former will only be called
in the first stage with CBMEM available.
Change-Id: Ie24bf4e818ca69f539196c3a814f3c52d4103d7e
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/63375
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Arthur Heymans <arthur@aheymans.xyz>
Reviewed-by: Julius Werner <jwerner@chromium.org>
Reviewed-by: Raul Rangel <rrangel@chromium.org>
In order to pass PCIe base address to payloads, implement pcie_fill_lb()
to fill coreboot table with PCIe info.
TEST=Build pass and boot up to kernel successfully via SSD on Dojo
board, here is the SSD information in boot log:
== NVME IDENTIFY CONTROLLER DATA ==
PCI VID : 0x15b7
PCI SSVID : 0x15b7
SN : 21517J440114
MN : WDC PC SN530 SDBPTPZ-256G-1006
RAB : 0x4
AERL : 0x7
SQES : 0x66
CQES : 0x44
NN : 0x1
Identified NVMe model WDC PC SN530 SDBPTPZ-256G-1006
BUG=b:178565024
BRANCH=cherry
Signed-off-by: Jianjun Wang <jianjun.wang@mediatek.com>
Change-Id: Ib2988694f60aac9cbfc09ef9a26d47e01c004406
Reviewed-on: https://review.coreboot.org/c/coreboot/+/63252
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
It's expected that the mismatch logs will be shown when doing
calibration for spmi clock. If it is failed to do calibration for spmi
clock for all data, the system will enter "die". Therefore, we adjust
the log level from BIOS_ERR to BIOS_INFO.
BUG=b:231531254
TEST=emerge-cherry coreboot
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: I148b4aeaaeb10e1c269a8eccbb19e8d8e17e40ff
Reviewed-on: https://review.coreboot.org/c/coreboot/+/64090
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
PMIC_CPSDSA4[4:0] controls the power-down at the specified time slot.
Setting it to 0xA would cause an extra delay of 20ms compared to 0xF.
The value of time slot is from 0x0 to 0x1F which represents the delay
when reset occurs.
To avoid the delay, change the value from 0xA to 0xF.
This modification is based on chapter 3.7 in the MT8186 functional
specification.
BUG=b:218630683, b:218630684
TEST=the power-off waveform is correct.
Signed-off-by: zhiyong tao <zhiyong.tao@mediatek.corp-partner.google.com>
Change-Id: I537fe87740f0f8c25b923d7d536e81503b71762b
Reviewed-on: https://review.coreboot.org/c/coreboot/+/64038
Reviewed-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
The per-file hash for CBFS_VERIFICATION, stored as a CBFS file
attribute, would increase the total RO metadata size by 75% (3796->6656
for corsola). Therefore, in order to make RO metadata cache fit into
CBFS_MCACHE, enlarge it from 8K to 16K.
Adjust the memlayout by decreasing the DRAM_INIT_CODE from 196K to 184K
(only 160K needed for now), and moving VBOOT2_WORK region to L2C. Also
shuffle the regions in SRAM with better comments.
BUG=b:229670703
TEST=emerge-corsola coreboot
TEST=Enabled CBFS_VERIFICATION and booted kingler into kernel
BRANCH=none
Change-Id: I8e07eb9fae1644a0fbfbdc599ca0a0e11bbe54b5
Signed-off-by: Yu-Ping Wu <yupingso@chromium.org>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/63924
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Hung-Te Lin <hungte@chromium.org>
Reviewed-by: Julius Werner <jwerner@chromium.org>
The MT8186 platform fails to suspend due to premature wakeup by USB.
In MT8186, we use low level latch to keep USB wakeup signal. However,
hardware could latch a wrong signal if it debounces more than one time.
As a result, it would enable wakeup function too early.
To prevent this issue, we do the following modification:
- Delay about 100 us to enable wakeup function in kernel drivers [1].
- To guarantee 100 us is enough, we need to disable the USB debounce by
default in coreboot.
According to section register 0x404 and 0x420 in
"(CODA) MT8169_PERICFG_REG.xls" which is only for MediaTek internal use:
The current default value of debounce register for MT8186 USB IP0 and
IP1 is incorrect. The reason we add in coreboot is that the default
value should be correct when SoC is booting up.
This modification is only for MT8186. The subsequent SoCs will adjust
the wakeup function to correct register value by default.
[1]: 0d8cfeeef3f5 (usb: xhci-mtk: fix random remote wakeup)
TEST=after stress test, not found premature wakeup by USB
BUG=b:228773975
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: I296c4491c5959670a39fa8bd6ef987557bbc459f
Reviewed-on: https://review.coreboot.org/c/coreboot/+/63858
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Reviewed-by: Paul Menzel <paulepanter@mailbox.org>
Reviewed-by: Hung-Te Lin <hungte@chromium.org>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
To save the power consumption, we disable the unused power of
optional components in coreboot.
BUG=none
TEST=the value of power consumption is as expected.
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: Ic0c7c2d1b6a4c26980a3029b60051ab1406406ea
Reviewed-on: https://review.coreboot.org/c/coreboot/+/63247
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Fix return type of early_init_get_elapsed_time_us() to comply with the
data type of return value.
Also replace memset() with struct initializer.
Signed-off-by: Jianjun Wang <jianjun.wang@mediatek.com>
Fixes: commit 41faa22 (soc/mediatek: Add early_init for passing data
across stages)
Change-Id: I7c361828362c2dfec91358ad8a420f5360243da0
Reviewed-on: https://review.coreboot.org/c/coreboot/+/63190
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Reviewed-by: Felix Held <felix-coreboot@felixheld.de>
According to the PCIe CEM specification, the deassertion of PERST#
should occur at least 100ms after the assertion. To ensure the 100ms
delay requirement is met, calculate the elapsed time since assertion. If
it is smaller than 100ms, do an extra delay.
TEST=Build pass and boot up to kernel successfully via SSD on Dojo
board, here is the measured PERST# time:
[DEBUG] mtk_pcie_domain_enable: 432517 us elapsed since assert PERST#
[INFO ] mtk_pcie_domain_enable: PCIe link up success (17 tries)
And the SSD information in boot log is as follows:
== NVME IDENTIFY CONTROLLER DATA ==
PCI VID : 0x15b7
PCI SSVID : 0x15b7
SN : 21517J440114
MN : WDC PC SN530 SDBPTPZ-256G-1006
RAB : 0x4
AERL : 0x7
SQES : 0x66
CQES : 0x44
NN : 0x1
Identified NVMe model WDC PC SN530 SDBPTPZ-256G-1006
BUG=b:178565024
BRANCH=cherry
Signed-off-by: Jianjun Wang <jianjun.wang@mediatek.com>
Change-Id: Ie2b7b6174abdf951af5796ab5ed141c45f32fc71
Reviewed-on: https://review.coreboot.org/c/coreboot/+/62933
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Described in PCIe CEM specification sections 2.2 (PERST# Signal) and
2.2.1 (Initial Power-Up (G3 to S0)). The deassertion of PERST# should be
delayed 100ms (TPVPERL) for the power and clock to become stable.
Instead of asserting PERST# right before PCIe initialization and waiting
for 100ms, which is currently the only function of 'mtk_pcie_pre_init',
so that the extra 100ms delay in ramstage is avoided.
TEST=Build pass and boot up to kernel successfully via SSD on Dojo
board, here is the SSD information in boot log:
== NVME IDENTIFY CONTROLLER DATA ==
PCI VID : 0x15b7
PCI SSVID : 0x15b7
SN : 21517J440114
MN : WDC PC SN530 SDBPTPZ-256G-1006
RAB : 0x4
AERL : 0x7
SQES : 0x66
CQES : 0x44
NN : 0x1
Identified NVMe model WDC PC SN530 SDBPTPZ-256G-1006
BUG=b:178565024
BRANCH=cherry
Signed-off-by: Jianjun Wang <jianjun.wang@mediatek.com>
Change-Id: Id5b9369e6f8599f93415588ea585c952a41c5e7d
Reviewed-on: https://review.coreboot.org/c/coreboot/+/62359
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Paul Menzel <paulepanter@mailbox.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Add early init support for MT8195 platform.
TEST=Build pass and boot up to kernel successfully via SSD on Dojo
board, here is the SSD information in boot log:
== NVME IDENTIFY CONTROLLER DATA ==
PCI VID : 0x15b7
PCI SSVID : 0x15b7
SN : 21517J440114
MN : WDC PC SN530 SDBPTPZ-256G-1006
RAB : 0x4
AERL : 0x7
SQES : 0x66
CQES : 0x44
NN : 0x1
Identified NVMe model WDC PC SN530 SDBPTPZ-256G-1006
BUG=b:178565024
BRANCH=cherry
Signed-off-by: Jianjun Wang <jianjun.wang@mediatek.com>
Change-Id: I4eb7da53ff76c385cab18bbf84970e96b61662ac
Reviewed-on: https://review.coreboot.org/c/coreboot/+/63020
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Add support for "early_init_data" region, which can be used to store
data initialized in an early stage (such as bootblock), and retrieve it
in later stages (such as ramstage).
TEST=Build pass and boot up to kernel successfully via SSD on Dojo
board, here is the SSD information in boot log:
== NVME IDENTIFY CONTROLLER DATA ==
PCI VID : 0x15b7
PCI SSVID : 0x15b7
SN : 21517J440114
MN : WDC PC SN530 SDBPTPZ-256G-1006
RAB : 0x4
AERL : 0x7
SQES : 0x66
CQES : 0x44
NN : 0x1
Identified NVMe model WDC PC SN530 SDBPTPZ-256G-1006
BUG=b:178565024
BRANCH=cherry
Signed-off-by: Jianjun Wang <jianjun.wang@mediatek.com>
Change-Id: I01f91b7fe2cbe4f73b5c616bb7aae778dee27d9a
Reviewed-on: https://review.coreboot.org/c/coreboot/+/63019
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
The feature "USE_CBMEM_DRAM_INFO" is supported in MT8186.
Therefore, we select this configuration to enable it.
BUG=none
TEST=build pass
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: Ieaaf57aaee79c9dce69cc1acaa092207f0f906de
Reviewed-on: https://review.coreboot.org/c/coreboot/+/63114
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
The memory initialization reference code didn't support returning
DRAM information in the old platforms, for example MT8192 and MT8195.
So we have to add a new configuration USE_CBMEM_DRAM_INFO to make
sure the common code will try to get DRAM information on new
platforms supporting that.
BUG=none
TEST=build pass
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: Iebe9ea0c1d01890b09fdf586813d85adde9702e1
Reviewed-on: https://review.coreboot.org/c/coreboot/+/63109
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
The current pmif register setting for low power mode is incorrect,
which is causing suspend failure. The issue of suspend failure is that
SRCLKENA0 will not be pulled down. EC will not be informed AP is
suspending now becuase of this. Therefore, add pmif_spmi_set_lp_mode()
to correct the setting.
This implementation is based on chapter 3.7 in MT8186 Functional
Specification.
BUG=b:215639203
TEST=test of suspend and resume pass.
Signed-off-by: Zhiyong Tao <zhiyong.tao@mediatek.com>
Change-Id: I2d02198f19f9cb052fba612c02404a6af1a10adb
Reviewed-on: https://review.coreboot.org/c/coreboot/+/63089
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Reviewed-by: Paul Menzel <paulepanter@mailbox.org>
The Cherry follower projects may choose Max98390 for audio output
so we have to add a new config CHERRY_USE_MAX98390. Also, the
'dojo' device is the first one to use it.
BUG=b:204391159
BRANCH=cherry
TEST=emerge-cherry coreboot
TEST=Verify beep function through CLI in depthcharge successfully
Signed-off-by: Trevor Wu <trevor.wu@mediatek.com>
Change-Id: I9b6bc5a5520292dd502b0389217f5062479b4490
Reviewed-on: https://review.coreboot.org/c/coreboot/+/63083
Reviewed-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Reviewed-by: Hung-Te Lin <hungte@chromium.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
To control I2S in MT8195 for dojo project, we need to enable adsp
power before audio power. Therefore, we need to update bus protection
steps to correct the setting.
TEST=build pass
BUG=b:204391159
BRANCH=cherry
Signed-off-by: Chun-Jie Chen <chun-jie.chen@mediatek.com>
Change-Id: I0bcf1ddeebf0d3df0a1d6b22273123be1aaf85a8
Reviewed-on: https://review.coreboot.org/c/coreboot/+/63106
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Originally, dram size is hard-coded to 4GB by default. To support
different dram size, calculate it from the mem chip info stored
in CBMEM.
BUG=b:206014043
TEST=Output "dram size: 0x100000000" on Kingler
Signed-off-by: Xi Chen <xixi.chen@mediatek.corp-partner.google.com>
Change-Id: I017e9d1a2d6e26f1fc21b67b5962dfb5c6ade8a5
Reviewed-on: https://review.coreboot.org/c/coreboot/+/62065
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Store dram info in cbmem for ramstage or payloads to use.
BUG=b:206014043
TEST=Build pass on Kingler
Signed-off-by: Xi Chen <xixi.chen@mediatek.corp-partner.google.com>
Change-Id: I195187c0c757a43bb6d2c57c8f303249f2a7995a
Reviewed-on: https://review.coreboot.org/c/coreboot/+/61334
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
In MT8186, we need to disable spm_thermal to prevent it from
influencing other wdt status.
There are two hardware pathes which are used for asserting watchdog
from thermal. We can disable status of path 1 because status of
path 2 is used.
1. Thermal -> SPM -> WDT
2. Thermal -> WDT
Spm_thermal (path 1) is a flexible option for software control, and
the hardware designer suggests that we should disable it if we don't
use it.
BUG=none
TEST=build pass
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: I0ffde6bad3000a64e3b5782edaa72c62da034302
Reviewed-on: https://review.coreboot.org/c/coreboot/+/62890
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Because we close external signal in kernel driver since MT8195, it's
more reasonable to trigger sw reset with exteranl signal again
whenever the wdt status is not equal to 0.
BUG=none
TEST=build pass
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: Ic6128df7eadaebcf7ff8d4c5492e3e0cfbab6e36
Reviewed-on: https://review.coreboot.org/c/coreboot/+/62797
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Without setting the set_resources field for soc_ops, we will get an
error during device initialization:
[ERROR] CPU_CLUSTER: 0 missing set_resources
Because the set_resources field is considered mandatory, explicitly set
it as no-op noop_set_resources.
BUG=b:224419346
TEST=emerge-corsola coreboot
TEST=Did not see the error on krabby
BRANCH=none
Change-Id: Ic82b86f0482a9de09e942c1674be5f0ac615851f
Signed-off-by: Yu-Ping Wu <yupingso@chromium.org>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/62785
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Paul Menzel <paulepanter@mailbox.org>
Reviewed-by: Hung-Te Lin <hungte@chromium.org>
Remove global variable and use 'pcidev_path_on_root()' to get the base
address of PCIe controller.
TEST=Build pass and boot up to kernel successfully via SSD on Dojo
board, here is the SSD information in boot log:
== NVME IDENTIFY CONTROLLER DATA ==
PCI VID : 0x15b7
PCI SSVID : 0x15b7
SN : 21517J440114
MN : WDC PC SN530 SDBPTPZ-256G-1006
RAB : 0x4
AERL : 0x7
SQES : 0x66
CQES : 0x44
NN : 0x1
Identified NVMe model WDC PC SN530 SDBPTPZ-256G-1006
BUG=b:178565024
Signed-off-by: Jianjun Wang <jianjun.wang@mediatek.com>
Change-Id: Ia41c82a7aa5d6e9d936e242550851cef83afeae9
Reviewed-on: https://review.coreboot.org/c/coreboot/+/62800
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Add chip config for setting PCIe config.
TEST=Build pass and boot up to kernel successfully via SSD on Dojo
board, here is the SSD information in boot log:
== NVME IDENTIFY CONTROLLER DATA ==
PCI VID : 0x15b7
PCI SSVID : 0x15b7
SN : 21517J440114
MN : WDC PC SN530 SDBPTPZ-256G-1006
RAB : 0x4
AERL : 0x7
SQES : 0x66
CQES : 0x44
NN : 0x1
Identified NVMe model WDC PC SN530 SDBPTPZ-256G-1006
BUG=b:178565024
Signed-off-by: Jianjun Wang <jianjun.wang@mediatek.com>
Change-Id: Icff83f2a9f76862065987a74cfcc7e511be80a20
Reviewed-on: https://review.coreboot.org/c/coreboot/+/62791
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
This patch introduces CONFIG_I2C_TRANSFER_TIMEOUT_US,
which controls how long to wait for an I2C devices to
produce/accept all the data bytes in a single transfer.
(The device can delay transfer by stretching the clock of
the ack bit.)
The default value of this new setting is 500ms. Existing
code had timeouts anywhere from tens of milliseconds to a
full second beween various drivers. Drivers can still have
their own shorter timeouts for setup/communication with the
I2C host controller (as opposed to transactions with I2C
devices on the bus.)
In general, the timeout is not meant to be reached except in
situations where there is already serious problem with the
boot, and serves to make sure that some useful diagnostic
output is produced on the console.
Change-Id: I6423122f32aad1dbcee0bfe240cdaa8cb512791f
Signed-off-by: Jes B. Klinke <jbk@chromium.org>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/62278
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Julius Werner <jwerner@chromium.org>
The pmic hwcid dumping should not be a warning, so we modify it to info.
BUG=none
TEST=build pass
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: I4a930b69bd45d5f0d84c3d269ca721b287dbadea
Reviewed-on: https://review.coreboot.org/c/coreboot/+/62775
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
It's more reasonable to halt when we trigger watchdog reset because
the whole system should be reset afterwards.
BUG=b:222217317
TEST=build pass
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: I726ba1599841f63b37062f9ce2e04840e4f250bb
Reviewed-on: https://review.coreboot.org/c/coreboot/+/62752
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Enable PCIe support for mt8195.
TEST=Build pass and boot up to kernel successfully via SSD on Dojo
board, here is the SSD information in boot log:
== NVME IDENTIFY CONTROLLER DATA ==
PCI VID : 0x15b7
PCI SSVID : 0x15b7
SN : 21517J440114
MN : WDC PC SN530 SDBPTPZ-256G-1006
RAB : 0x4
AERL : 0x7
SQES : 0x66
CQES : 0x44
NN : 0x1
Identified NVMe model WDC PC SN530 SDBPTPZ-256G-1006
BUG=b:178565024
Signed-off-by: Jianjun Wang <jianjun.wang@mediatek.com>
Change-Id: I314572955f1021abe9f2f0f4635670135ed08fff
Reviewed-on: https://review.coreboot.org/c/coreboot/+/56793
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Add a new function 'mtk_pcie_pre_init' to assert the PCIe reset at early
stage to reduce the impact of 100ms delay.
TEST=Build pass and boot up to kernel successfully via SSD on Dojo
board, here is the SSD information in boot log:
== NVME IDENTIFY CONTROLLER DATA ==
PCI VID : 0x15b7
PCI SSVID : 0x15b7
SN : 21517J440114
MN : WDC PC SN530 SDBPTPZ-256G-1006
RAB : 0x4
AERL : 0x7
SQES : 0x66
CQES : 0x44
NN : 0x1
Identified NVMe model WDC PC SN530 SDBPTPZ-256G-1006
BUG=b:178565024
Signed-off-by: Jianjun Wang <jianjun.wang@mediatek.com>
Change-Id: If6799c53b03a33be91157ea088d829beb4272976
Reviewed-on: https://review.coreboot.org/c/coreboot/+/56792
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
The mainboard may not be able to disable the internal cap, so we want
to set 0xe0 for all boards to minimize the internal cap. And a
mainboard implementation may choose XTAL with higher cload if the
frequency requirement is met, and the total capacitance can be tuned
externally for different boards.
BUG=b:218439447
TEST=set capid to 0xe0.
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: I2139e6b3456d7a50e3cdc8fc606e5f6ea3406044
Reviewed-on: https://review.coreboot.org/c/coreboot/+/62563
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Set NOR pin drive to 8mA to comply with HW requirement.
This implementation is according to chapter 5.1, 5.6 and 5.8 in MT8186
Functional Specification.
BUG=b:218775654, b:216462313, b:212375511
TEST=SPI SI tests for AP to NOR pass for both kingler and krabby.
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: I5b6e37b0f7d4207ea35f11394d25ad1e096ac01a
Reviewed-on: https://review.coreboot.org/c/coreboot/+/62472
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Add GPIO driving functions to adjust pin driving.
The value of drive strength is different for each SoC, so we define
GPIO_DRV0 to GPIO_DRV7 which are corresponding to 2/4/6/8/10/12/14/16mA
in MT8186.
This implementation is according to chapter 5.1 in MT8186 Functional
Specification.
BUG=b:218775654, b:216462313, b:212375511
TEST=build pass
Signed-off-by: Guodong Liu <guodong.liu@mediatek.corp-partner.google.com>
Change-Id: I6d987f28be98b515fa5c542222bda08bea1d5118
Reviewed-on: https://review.coreboot.org/c/coreboot/+/62471
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
cb_err_t was meant to be used in place of `enum cb_err` in all
situations, but the choice to use a typedef here seems to be
controversial. We should not be arbitrarily using two different
identifiers for the same thing across the codebase, so since there are
no use cases for serializing enum cb_err at the moment (which would be
the primary reason to typedef a fixed-width integer instead), remove
cb_err_t again for now.
Signed-off-by: Julius Werner <jwerner@chromium.org>
Change-Id: Iaec36210d129db26d51f0a105d3de070c03b686b
Reviewed-on: https://review.coreboot.org/c/coreboot/+/62600
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Reviewed-by: Tim Wawrzynczak <twawrzynczak@chromium.org>
Add PCIe support for MediaTek platform.
Reference:
- MT8195 Register Map V0.3-2, Chapter 3.18 PCIe controller (Page 1250)
- linux/drivers/pci/controller/pcie-mediatek-gen3.c
This code is based on MT8195 platform, but it should be common in each
platform with the same PCIe IP in the future.
TEST=Build pass and boot up to kernel successfully via SSD on Dojo
board, here is the SSD information in boot log:
== NVME IDENTIFY CONTROLLER DATA ==
PCI VID : 0x15b7
PCI SSVID : 0x15b7
SN : 21517J440114
MN : WDC PC SN530 SDBPTPZ-256G-1006
RAB : 0x4
AERL : 0x7
SQES : 0x66
CQES : 0x44
NN : 0x1
Identified NVMe model WDC PC SN530 SDBPTPZ-256G-1006
BUG=b:178565024
Signed-off-by: Jianjun Wang <jianjun.wang@mediatek.com>
Change-Id: Ib9b6adaafa20aeee136372ec9564273f86776da0
Reviewed-on: https://review.coreboot.org/c/coreboot/+/56791
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Move some structures to common folder (CB:61293), so we need to update
header version.
BUG=none
TEST=dram calibration pass
Signed-off-by: Ryan Chuang <ryan.chuang@mediatek.corp-partner.google.com>
Change-Id: Id82cbef9cb10dba71489ea96f67c329de9aadc49
Reviewed-on: https://review.coreboot.org/c/coreboot/+/62550
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Move some structures to common folder (CB:61293), so we need to update
header version for this.
BUG=none
TEST=dram calibration pass
Signed-off-by: Ryan Chuang <ryan.chuang@mediatek.corp-partner.google.com>
Change-Id: I8cf12f4967af116eaef88c1f688567f1da9fa6e4
Reviewed-on: https://review.coreboot.org/c/coreboot/+/62549
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
The XTAL 26MHz test failed on krabby, so we adjust RTC capid from
default value 0x88 to 0xC0 for MT8186. We also add a new log message
to show the capid value which is read from MT6366.
This implementation is according to chapter 5.13 in MT8186 Functional
Specification.
BUG=b:218439447
TEST=set capid to 0xc0.
TEST=XTAL 26MHz test passed.
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: I16ab46a5697d304e8001de231ffc9b7b7a2f9282
Reviewed-on: https://review.coreboot.org/c/coreboot/+/62290
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
RTC_GPIO_USER_MASK is not used in any drivers, so we remove them.
BUG=none
TEST=build pass
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: I0a15d5da142bb38feb595610d69566330e31fedd
Reviewed-on: https://review.coreboot.org/c/coreboot/+/62457
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
The ddr_base_info struct, which stores basic DDR information, should be
platform independent. Currently the struct is defined in each SoC's
dramc_parah.h. To prevent code duplication, move it as well as other
related structs and enums to a common header.
Signed-off-by: Xi Chen <xixi.chen@mediatek.corp-partner.google.com>
Change-Id: I99772427f9b0755dc2c778b5f4150b2f8147bcc3
Reviewed-on: https://review.coreboot.org/c/coreboot/+/61293
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
VSRAM_CORE is not used on kingler/krabby, so we disable it.
This implementation is according to chapter 3.7 in MT8186 Functional
Specification.
BUG=b:220071688
TEST=the rail steadily shows 0V in either S0, S3, and S5.
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: I5256f6a2c0ca5a951dc79f564575b526a84463fd
Reviewed-on: https://review.coreboot.org/c/coreboot/+/62253
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Currently the SPM version string is stored at the end of the blob,
possibly without a trailing '\0'. Therefore, we should be careful not to
print characters beyond the blob size.
BUG=b:211944565
TEST=emerge-corsola coreboot
TEST=SPM version looked good in AP console
BRANCH=asurada,cherry
Change-Id: Icfeb686539dc20cf5b78de77c27bdbb137b5d624
Signed-off-by: Yu-Ping Wu <yupingso@chromium.org>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/61800
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Hung-Te Lin <hungte@chromium.org>
The current SPI NOR speed mainpll_d7_d2 (78MHz) is too fast for MT8186's
HW design, which is capable of up to 52MHz. Therefore, lower the speed
to univpll_d3_d8 (52MHz).
BUG=b:218775654
TEST=emerge-corsola coreboot
TEST=Boot time didn't increase significantly
BRAHCH=none
Change-Id: I5a03e41d4ce47d45b97a805b9b98877ef0dac7b7
Signed-off-by: Yu-Ping Wu <yupingso@chromium.org>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/61796
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Reviewed-by: Hung-Te Lin <hungte@chromium.org>
To prevent to modify original value of wdt_mode, we use setbits32 to
update required bits.
BUG=b:218420108
TEST=build pass
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: I743c1af3583c18ec8500fc1eb89f31cdbce5317c
Reviewed-on: https://review.coreboot.org/c/coreboot/+/61729
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
The implementation of clearing watchdog status is wrong in CB:58835.
The value written to the 'wdt_mode' register should be
'wdt_mode | 0x22000000' instead of 'wdt_status | 0x22000000'.
BUG=b:204229208
TEST=check watchdog status is cleared.
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: I8c5dbaab2ac43d3867037bc4160aa5af2d79284f
Reviewed-on: https://review.coreboot.org/c/coreboot/+/61446
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
For most MediaTek SoCs (MT8183, MT8192, MT8195) we rely on an external
program (e.g., the "DRAM blob") to do the full DRAM calibration first,
then store and and apply the generated parameters to the reference
"fast DRAM calibration" in the vendor/mediatek folder for normal system
boot.
Starting with MT8186 the implementation of fast calibration may need
to be changed, and a "DRAM blob" only path is introduced for devices
that have to do both full and fast calibration using the external blob.
TEST=fast calibration pass on kingler/krabby
BUG=b:204226005
Signed-off-by: Xi Chen <xixi.chen@mediatek.corp-partner.google.com>
Change-Id: If25a7dd6aa6261ecff79a1b4df8b1f2e53d896dc
Reviewed-on: https://review.coreboot.org/c/coreboot/+/61133
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Now that the console system itself will clearly differentiate loglevels,
it is no longer necessary to explicitly add "ERROR: " in front of every
BIOS_ERR message to help it stand out more (and allow automated tooling
to grep for it). Removing all these extra .rodata characters should save
us a nice little amount of binary size.
This patch was created by running
find src/ -type f -exec perl -0777 -pi -e 's/printk\(\s*BIOS_ERR,\s*"ERROR: /printk\(BIOS_ERR, "/gi' '{}' ';'
and doing some cursory review/cleanup on the result. Then doing the same
thing for BIOS_WARN with
's/printk\(\s*BIOS_WARNING,\s*"WARN(ING)?: /printk\(BIOS_WARNING, "/gi'
Signed-off-by: Julius Werner <jwerner@chromium.org>
Change-Id: I3d0573acb23d2df53db6813cb1a5fc31b5357db8
Reviewed-on: https://review.coreboot.org/c/coreboot/+/61309
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: HAOUAS Elyes <ehaouas@noos.fr>
Reviewed-by: Lance Zhao
Reviewed-by: Jason Glenesk <jason.glenesk@gmail.com>
Some registers of PMIC init settings are protected, so we failed to
set the correct value for init_setting. We disable protection before
setting PMIC init setting and enable it afterward.
BUG=b:216263707
TEST=PMIC setting value is set correctly.
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: I94d73d9c8a137444988e65c3709d29a3a4c03c5b
Reviewed-on: https://review.coreboot.org/c/coreboot/+/61390
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
There is no wakeup source when we test function of suspend and resume.
The root cause is that the monitor enable bit of PWRAP is not configured
correctly.
BUG=b:213255218, b:214978483
TEST=receive wakeup source from MT6366 successfully
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: I324d18fa5d3cd745c35fcf0f207e1b444b5e898b
Reviewed-on: https://review.coreboot.org/c/coreboot/+/61330
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Fast-k flow may need to re-init header because mrc_cache doesn't
store header. Storing header together with dparam data is better
for data consistancy.
TEST=fast calibration pass on Corsola
BUG=b:204226005
Signed-off-by: Xi Chen <xixi.chen@mediatek.corp-partner.google.com>
Change-Id: I22982923dce06c9e770aa4f20f3dcd2f33685d84
Reviewed-on: https://review.coreboot.org/c/coreboot/+/61294
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
To be shared with different SOCs, move the dramc_param_header struct
as well DRAMC_PARAM_FLAG and DRAMC_PARAM_CONFIG enums to a common
header file dramc_param_common.h.
TEST=fast calibration pass
BUG=b:204226005
Signed-off-by: Xi Chen <xixi.chen@mediatek.corp-partner.google.com>
Change-Id: I087971799803e47e34c30063b2b0bd0cfc5795ac
Reviewed-on: https://review.coreboot.org/c/coreboot/+/61132
Reviewed-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Fix the name of the include guard for
soc/mediatek/common/include/soc/emi.h.
BUG=none
TEST=emerge-corsola coreboot
BRANCH=none
Change-Id: Iddac3467959545b7db141545aaa2a135536f44f1
Signed-off-by: Yu-Ping Wu <yupingso@chromium.org>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/60440
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Hung-Te Lin <hungte@chromium.org>
Reviewed-by: Xi Chen <xixi.chen@mediatek.com>
Reviewed-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
We need to protect debugsys for firmware image without serial console.
Original settings for protecting debugsys is wrong which will cause some
hardware modules to fail to set their registers correctly.
We move the setting from MM_AO_APC to INFRA_AO_APC because the setting
of debugsys is defined in INFRA_AO_APC and set the debugsys index to
correct value of 94.
BUG=b:213125558
TEST=all modules work normally using image without serial console.
Signed-off-by: Runyang Chen <runyang.chen@mediatek.corp-partner.google.com>
Change-Id: Ibce626386ac1f8de42f8717c4ad9ba403640b3ec
Reviewed-on: https://review.coreboot.org/c/coreboot/+/60833
Reviewed-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Reviewed-by: Hung-Te Lin <hungte@chromium.org>
Reviewed-by: Paul Menzel <paulepanter@mailbox.org>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
DFD (Design for Debug) is a debugging tool, which scans flip-flops
and dumps to internal RAM on the WDT reset. After system reboots,
those values can be shown for debugging using MTK internal parsing
tools.
BUG=b:202871018
TEST=build pass
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: I7b711755022b5d9767019611151fea65e71edc66
Reviewed-on: https://review.coreboot.org/c/coreboot/+/60828
Reviewed-by: Hung-Te Lin <hungte@chromium.org>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
The current CBFS mcache size (roughly 7KiB) is insufficient for mt8186,
so we need to increase it by 1KiB (and decrease the stack by 1KiB).
Error logs:
CBFS ERROR: mcache overflow, should increase CBFS_MCACHE size!
CBFS: mcache @0x0010e004 built for 63 files, used 0xde4 of 0xdfc bytes
BUG=b:202871018
TEST=no cbfs error logs.
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: I1e627ede3774665575006f752f89101e3c5bde9f
Reviewed-on: https://review.coreboot.org/c/coreboot/+/60529
Reviewed-by: Hung-Te Lin <hungte@chromium.org>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
To provide power to PS8640, the eDP bridge IC on krabby, add control
of VRF12 and VCN33 to set voltage from MT6366.
TEST=measure 1.2V from VRF12 and 3.3V from VCN33.
BUG=b:210806060
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: I55a9ca16e1e335e9355d0a1b30c278a9969db197
Reviewed-on: https://review.coreboot.org/c/coreboot/+/60446
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Remove emi.h because emi.h is defined in common/include/soc.
Add dramc_param.h and dramc_soc.h to prepare for implementation of
DRAM full calibration.
TEST=build pass.
BUG=b:202871018
Signed-off-by: Ryan Chuang <ryan.chuang@mediatek.corp-partner.google.com>
Change-Id: If8662ed43088ea5aa1fe6cb5b2c4bda2338c4387
Reviewed-on: https://review.coreboot.org/c/coreboot/+/60385
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Previously (before CB:56965 [1]) mtk_dsi_send_init_command() would
ignore NULL initialization commands passed to it. However, in the
current code mipi_panel_parse_init_commands() doesn't check that (see
CB:57150 [2]), so we should check it on the caller side from
mtk_dsi_init().
[1] b2a1480191 device: Move MIPI panel library from
mainboard/google/kukui into common
[2] 4757a7ea33 mipi: Make panel init callback work directly on DSI
transaction types
BUG=b:202871018, b:209930699
TEST=emerge-corsola coreboot
BRANCH=none
Change-Id: I8196e3b135da273325e2e121523abb7fb230a49c
Signed-off-by: Yu-Ping Wu <yupingso@chromium.org>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/60392
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Hung-Te Lin <hungte@chromium.org>
Reviewed-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
PS8640 is a low power MIPI-to-eDP video format converter.
VRF12 does not provide power to PS8640 on krabby.
In original patch, VRF12 is not used, and is set to hardware control
for low power. We change the setting to remove hardware control.
Therefore, if we want to control VRF12 by software, we can control
it directly.
BUG=b:210806060
TEST=build pass
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: I87d6a94b6fb343590d563ac1554ff87b11c01549
Reviewed-on: https://review.coreboot.org/c/coreboot/+/60340
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Add basic devapc (device access permission control) drivers.
DAPC driver is used to set up bus fabric security and data protection
among hardwares. DAPC driver groups the master hardwares into different
domains and gives secure and non-secure property. The slave hardware can
configure different access permissions for different domains via DAPC
driver.
1. Initialize devapc.
2. Set master domain and secure side band.
3. Set default permission.
BUG=b:202871018
TEST=build pass
Signed-off-by: Runyang Chen <runyang.chen@mediatek.corp-partner.google.com>
Change-Id: I5dad4f342eef3136c24c38259ad176dc86b7c0d7
Reviewed-on: https://review.coreboot.org/c/coreboot/+/60317
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
We use parts of SRAM_L2C as the memory of PRERAM_CBMEM_CONSOLE before
DRAM calibration. When we check cbmem, we found the content of this
memory is unreadable.
The L3 (can be used as SRAM_L2C) is 1MB in total. However the BootROM
has configured only half of L2/L3 cache as SRAM. Therefore, decrease
the size of each SRAM region to fit into the first half of the cache.
BUG=b:207725851
TEST=Bootblock log looked good in `cbmem -c`
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: I6041767a1ac0a48ecdda29a0c35d90acf6ad0ef2
Reviewed-on: https://review.coreboot.org/c/coreboot/+/60316
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
There is a design issue of bit shift which will drop a bit for
USB3 phy on MT8195. Therefore, we add this patch to set USB phy
registers from value of efuse.
BUG=b:211528577
TEST=build pass
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Signed-off-by: Tianping Fang <tianping.fang@mediatek.corp-partner.google.com>
Tested-by: Tianping Fang <tianping.fang@mediatek.corp-partner.google.com>
Change-Id: I43cb6c1c795dd181d6eba7f3bc52e4eb1a602081
Reviewed-on: https://review.coreboot.org/c/coreboot/+/60312
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
To save the S3 power, USB3_HUB_RST_L is externally pulled up to a weak
resistor, so we have to reset the hub as early as possible.
Otherwise the USB3 hub may be not usable. Therefore, move USB3 HUB
reset function to bootblock.
BUG=b:210065282
TEST=build pass
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: I92feb2316302fda32478b24c014bcd380d0ac55d
Reviewed-on: https://review.coreboot.org/c/coreboot/+/60088
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Tracker is a debugging tool, and MT8186 only supports AP tracker.
When bus timeout occurs, the system reboots and latches some values
which could be used for debugging.
This function will be triggered only when it encounters the bug
hanging issue.
BUG=b:202871018
TEST=range of registers are dumped as expected.
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: Ie023de2a6f7421a16b2516baa0bf0bf6fff589e2
Reviewed-on: https://review.coreboot.org/c/coreboot/+/59990
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
There are two versions for tracker system:
Version 1 for MT8186, and version 2 for MT8192 and MT8195.
Reference document:
MT8169_bus_dbg_tracker_cfg_reg.xls from MediaTek internal.
BUG=b:202871018
TEST=build pass
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: Idb146974da118b1cf5a349370bf7b2fa13f1aba8
Reviewed-on: https://review.coreboot.org/c/coreboot/+/59989
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
The checking register will be cleared after EC resets, so we move
bustracker dump from ramstage to bootblock, before triggering EC reset.
TEST=bustracker shows status before watchdog resets
BUG=b:207743045
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: Ic18dc9742cd9f657a035a374e28371dfc5f04ac3
Reviewed-on: https://review.coreboot.org/c/coreboot/+/59667
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
There will be no log in cbmem if we trigger ec reset on bootblock
stage. Therefore, call dcache_clean_all() before triggering ec
reset to flush cache to store logs on cbmem.
BUG=b:207743045
TEST=show logs on cbmem
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: I1bd900beb4cc84f7121c5fb66907fa73b62517fa
Reviewed-on: https://review.coreboot.org/c/coreboot/+/59683
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Raise little CPU to 2GHz at romstage.
TEST=check little core cpu frequency is 2GHz
BUG=b:202871018
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: If4c983d15beb2b588230f3db7416cb767b29978d
Reviewed-on: https://review.coreboot.org/c/coreboot/+/59569
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
The types of pwrap_read_field()'s return value and pwrap_write_field()'s
`val` argument are u16, so correct the usage in MT6366.
TEST=build pass
BUG=b:202871018
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: Ie05ab65ecd9b8ea1379ef74393285c4f5d2db8a4
Reviewed-on: https://review.coreboot.org/c/coreboot/+/59567
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
To raise little CPU frequency, add support for VPROC12 and VSRAM_PROC12 of MT6366.
TEST=build pass
BUG=b:202871018
Signed-off-by: James Lo <james.lo@mediatek.corp-partner.google.com>
Change-Id: I718fdf36d34969a6e21ddc8c1ec6f525e0e20904
Reviewed-on: https://review.coreboot.org/c/coreboot/+/59566
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
This patch adds support for loading SPM firmware from CBFS to SPM SRAM.
SPM needs its own firmware to enable SPM suspend/resume function which
turns off several resources such as DRAM/mainpll/26M clk when linux
system suspend.
TEST=program counter of SPM is correct value(0x250) after booting up
BUG=b:202871018
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: Ia13e5a2ecf09561856b7e958128cd2f045c39f33
Reviewed-on: https://review.coreboot.org/c/coreboot/+/59341
Reviewed-by: Rex-BC Chen <rex-bc.chen@mediatek.corp-partner.google.com>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
SSPM is "Secure System Power Manager" that provides power control in
secure domain. The initialization flow is to load SSPM firmware to
its SRAM space and then enable.
TEST=build pass
BUG=b:202871018
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: I92eb501a1e48dd02d2f94ff392933261e6a42391
Reviewed-on: https://review.coreboot.org/c/coreboot/+/59340
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Setting of MSDC is defined by soc, so we move them to soc folder.
TEST=emerge-cherry coreboot; emerge-asurada coreboot
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: I84ad8a4cde120c97024870ebf750d44b36c2284d
Reviewed-on: https://review.coreboot.org/c/coreboot/+/59339
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
DCM (dynamic clock management) can dynamically slow down or gate clocks
during CPU or bus idle. Enable DCM settings on the MT8186 platform.
TEST=build pass and check register ok
BUG=b:202871018
Signed-off-by: Edward-JW Yang <edward-jw.yang@mediatek.corp-partner.google.com>
Change-Id: I82add5ae629d59f7d6773e26ac9cba9d54ab8caf
Reviewed-on: https://review.coreboot.org/c/coreboot/+/59338
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Some eMMCs need 80+ms for CMD1 to complete. And the payload may need to
access eMMC in the very early stage (for example, depthcharge needs it
20ms after started) so we have to start initialization in coreboot.
TEST=boot kernel from eMMC ok
BUG=b:202871018
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: I3bc06b1fc506b1d6f54f7f456117d22477a87e29
Reviewed-on: https://review.coreboot.org/c/coreboot/+/59294
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
The Corsola reference design has both eMMC and SD Card interfaces
so we have to configure both in RAM stage.
TEST=build pass
BUG=b:202871018
Signed-off-by: Wenbin Mei <wenbin.mei@mediatek.com>
Change-Id: I2f26a8a11edd29a80a7195e3a324151d66ecb293
Reviewed-on: https://review.coreboot.org/c/coreboot/+/59292
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
The _Static_assert without message string is only available
since C++17. Add the message to avoid build fail in the macro.
BUG=b:203145462
BRANCH=cherry
TEST=build pass and boot pass
Signed-off-by: Flora Fu <flora.fu@mediatek.com>
Change-Id: Ib146ffafc21b9dbb9d383c9343a9ec1d7c478faf
Reviewed-on: https://review.coreboot.org/c/coreboot/+/59298
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Add support for VMCH and VMC of MT6366.
TEST=measure voltage 3.3V for VMCH and VMC
BUG=b:202871018
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: Id8d98b6d827abd4713ee5c216941a9621422c7eb
Reviewed-on: https://review.coreboot.org/c/coreboot/+/59254
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
1. Enable and setup USB drivers.
2. Pull up to a weak resistor for USB3_HUB_RST_L and we reset
the hub via GPIO149.
TEST=boot kernel from USB ok
BUG=b:202871018
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: Ifcc11d51b0c1e495477957111e6021ef8275f629
Reviewed-on: https://review.coreboot.org/c/coreboot/+/59251
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
1. Turn off L2C SRAM and reconfigure as L2 cache
Mediatek SoC uses part of the L2 cache as SRAM before DRAM is ready.
After DRAM is ready, we should invoke disable_l2c_sram to reconfigure
the L2C SRAM as L2 cache.
2. Configure DMA buffer in DRAM
Set DRAM DMA to be non-cacheable to load blob correctly.
TEST=build pass
BUG=b:202871018
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: If56d29cdd7d9dfaed05e129754aa1f887a581482
Reviewed-on: https://review.coreboot.org/c/coreboot/+/59247
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Move mtk_soc_disable_l2c_sram and mtk_soc_after_dram to common folder
which are the same between MT8192, MT8195 and MT8186.
TEST=build pass
BUG=b:202871018
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: I8f49214b932a8d28ed2ca0d764dc745fa8ad330d
Reviewed-on: https://review.coreboot.org/c/coreboot/+/59246
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Change help text to "dual IO read mode" to reduce noun confusion.
Suggestion from this comment:
https://review.coreboot.org/c/coreboot/+/58837/comment/40a98af1_dce6bb2b/
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: I54b81cdeba3b693451f66e003fb470c9f8c19ad9
Reviewed-on: https://review.coreboot.org/c/coreboot/+/59244
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Reviewed-by: Felix Held <felix-coreboot@felixheld.de>
Add APU device apc driver and set up permissions.
APU has its own device apc for control access by domains.
For Domain 0, the access to the following slaves are restricted to
security read and write:
apusys_ao-2, apusys_ao-4, apusys_ao-5, apu_sctrl_reviser,
apu_iommu0_r1 apu_iommu0_r2, apu_iommu0_r3, apu_iommu0_r4
apu_iommu1_r1, apu_iommu1_r2, apu_iommu1_r3,apu_iommu1_r4
For VPU, D0/D5 are set as no protection, other domains are forbidden.
For other slaves, the D0 is no protection, other domains are forbidden.
BUG=b:203145462
BRANCH=cherry
TEST=boot cherry, check dump log and test permissions
Signed-off-by: Flora Fu <flora.fu@mediatek.com>
Change-Id: If92d3b02ac4966332315b85d68e0f48c6a9fce85
Reviewed-on: https://review.coreboot.org/c/coreboot/+/58969
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
EINT event mask register is used to mask EINT wakeup source.
All wakeup sources are masked by default. Since most MediaTek SoCs do
not have this design, we can't modify the kernel EINT upstream driver to
solve the issue 'Can't wake using power button (cros_ec) or touchpad'.
So we add a driver here to unmask all wakeup sources.
TEST=build pass
BUG=b:202871018
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: I84946c2c74dd233419cb94f013a42c734363baf7
Reviewed-on: https://review.coreboot.org/c/coreboot/+/58938
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
The NOR-Flash can be configured on SPI0 or TDM-RX GPIOs so we have to
provide an init function in SoC for the mainboard to select right
configuration.
TEST=boot to romstage
BUG=b:202871018
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: I285ec64ace8b72a48ef1d481d366bd67cb9b0337
Reviewed-on: https://review.coreboot.org/c/coreboot/+/58935
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
MT8186 requires writing speical value to mode register to clear
status register. The flow of clear status is different from other
platforms, so we override mtk_wdt_clr_status() for MT8186.
TEST=build pass
BUG=b:202871018
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: I290b69573a8e58db76814e16b5c17c23413f1108
Reviewed-on: https://review.coreboot.org/c/coreboot/+/58835
Reviewed-by: Rex-BC Chen <rex-bc.chen@mediatek.corp-partner.google.com>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
mtk_wdt_clr_status is different for MT8186 and MT8195,
so we move this function to soc folder.
TEST=build pass
BUG=b:202871018
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: Ia8697ffdca1e2d1443f2259713c4ab6fdf1b1a9e
Reviewed-on: https://review.coreboot.org/c/coreboot/+/58834
Reviewed-by: Rex-BC Chen <rex-bc.chen@mediatek.corp-partner.google.com>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Add new folder and basic drivers for Mediatek SoC 'MT8186'.
Difference of modules including in this patch between MT8186 and existing SoCs:
Timer:
Similar to MT8195, MT8186 uses v2 timer.
EMI/PLL/SPI:
Different from existing SoCs.
TEST=boot from SPI-NOR and show uart log on MT8186 EVB
BUG=b:200134633
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: I579f79c15f4bf5e1daf6b35c70cfd00a985a0b81
Reviewed-on: https://review.coreboot.org/c/coreboot/+/58640
Reviewed-by: Rex-BC Chen <rex-bc.chen@mediatek.corp-partner.google.com>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Some enum variables of timer v2 are the same between MT8195 and MT8186,
so we move them to common timer_v2.h.
TEST=emerge-cherry coreboot
BUG=b:200134633
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: I89891a19e622aa24783025e73c38c4ffa43aa166
Reviewed-on: https://review.coreboot.org/c/coreboot/+/58829
Reviewed-by: Rex-BC Chen <rex-bc.chen@mediatek.corp-partner.google.com>
Reviewed-by: Hung-Te Lin <hungte@chromium.org>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Tracker is a debugging tool, include AP/INFRA/PERI tracker.
When bus timeout occurs, the system reboots and latches some
values which could be used for debug.
Signed-off-by: Zhenguo Li <ot_zhenguo.li@mediatek.corp-partner.google.com>
Change-Id: I82f8e6e5f8ccb7f8246cae45a01a3ddd5f2966f2
Reviewed-on: https://review.coreboot.org/c/coreboot/+/58244
Reviewed-by: Rex-BC Chen <rex-bc.chen@mediatek.corp-partner.google.com>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Tracker is a debugging tool, include AP/INFRA/PERI tracker.
When bus timeout occurs, the system reboots and latches some
values which could be used for debug.
Signed-off-by: Zhenguo Li <ot_zhenguo.li@mediatek.corp-partner.google.com>
Change-Id: If457f4a096cd63038bf6b40552aa3caaba33d5fd
Reviewed-on: https://review.coreboot.org/c/coreboot/+/58243
Reviewed-by: Rex-BC Chen <rex-bc.chen@mediatek.corp-partner.google.com>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
These issues were found and fixed by codespell, a useful tool for
finding spelling errors.
Signed-off-by: Martin Roth <martin@coreboot.org>
Change-Id: Ieafbc93e49fcef198ac6e31fc8a3b708c395e08e
Reviewed-on: https://review.coreboot.org/c/coreboot/+/58082
Reviewed-by: Felix Held <felix-coreboot@felixheld.de>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
ltiming and clock_div are not support for MT8173, so we separate them
to weak function: mtk_i2c_dump_more_info()
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: I3228c6953be5fac18a76029702b878a34c7563f7
Reviewed-on: https://review.coreboot.org/c/coreboot/+/58074
Reviewed-by: Rex-BC Chen <rex-bc.chen@mediatek.corp-partner.google.com>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
1. The original algorithm for I2C speed cannot always make the
timing meet I2C specification so a new algorithm is introduced
to calculate the timing parameters more correctly.
2. Some I2C buses should be initialized in a different speed while
the original implementation was fixed at fast mode (400Khz).
So the mtk_i2c_bus_init is now also taking an extra speed
parameter.
There is an equivalent change in kernel side:
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/drivers/i2c/busses/i2c-mt65xx.c?h=v5.15-rc3&id=be5ce0e97cc7a5c0d2da45d617b7bc567c3d3fa1
BUG=b:189899864
TEST=Test on Tomato, boot pass and timing pass
at 100/300/400/500/800/1000Khz.
Signed-off-by: Daolong Zhu <jg_daolongzhu@mediatek.corp-partner.google.com>
Change-Id: Id25b7bb3a76908a7943b940eb5bee799e80626a0
Reviewed-on: https://review.coreboot.org/c/coreboot/+/58053
Reviewed-by: Rex-BC Chen <rex-bc.chen@mediatek.corp-partner.google.com>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
DFD (Design for Debug) is a debugging tool, which scans
flip-flops and dumps to internal RAM on the WDT reset.
After system reboots, those values could be showed for
debugging.
BUG=b:192429713
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: Ied63913db94b2e52ab394a66c70f7edfd507c99b
Reviewed-on: https://review.coreboot.org/c/coreboot/+/57980
Reviewed-by: Rex-BC Chen <rex-bc.chen@mediatek.corp-partner.google.com>
Reviewed-by: Hung-Te Lin <hungte@chromium.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Our MIPI panel initialization framework differentiates between DCS and
GENERIC commands, but the exact interpretation of those terms is left to
the platform drivers. In practice, the MIPI DSI transaction codes for
these are standardized and platforms always need to do the same
operation of combining the command length and transfer type into a
correct DSI protocol code. This patch factors out the various
platform-specific DSI protocol definitions into a single global one and
moves the transaction type calculation into the common panel framework.
The Qualcomm SC7180 implementation which previously only supported DCS
commands is enhanced to (hopefully? untested for now...) also support
GENERIC commands. While we're rewriting that whole section also fix some
other issues about how exactly long and short commands need to be passed
to that hardware which we identified in the meantime.
Signed-off-by: Julius Werner <jwerner@chromium.org>
Change-Id: I09ade7857ca04e89d286cf538b1a5ebb1eeb8c04
Reviewed-on: https://review.coreboot.org/c/coreboot/+/57150
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Hung-Te Lin <hungte@chromium.org>
The eventlog requires RTC to provide correct timestamps, so we have to
turn on the config and add the common drivers.
BUG=b:199003609
TEST=check timestamp in 'mosys eventlog list'
BRANCH=none
Signed-off-by: Chen-Tsung Hsieh <chentsung@chromium.org>
Change-Id: Ia382cd023fcbfdf2c1efeb7b32c0b99feb71effa
Reviewed-on: https://review.coreboot.org/c/coreboot/+/57403
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Hung-Te Lin <hungte@chromium.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
1. Disable external output reset signal in first WDT reset
to preserve WDT original reset reason for WDT issue in kernel stage.
2. After preserved WDT reset reason, do fully reset again by sending
external output reset signal.
BUG=b:194025005
TEST=boot to kernel ok and function test pass
Signed-off-by: Fengquan Chen <fengquan.chen@mediatek.corp-partner.google.com>
Change-Id: I5887a8312f4daab3cbd0a30fea0195670a932e52
Reviewed-on: https://review.coreboot.org/c/coreboot/+/57270
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Rex-BC Chen <rex-bc.chen@mediatek.corp-partner.google.com>
Reviewed-by: Hung-Te Lin <hungte@chromium.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Sounds like we prefer to have this under drivers/ instead of device/.
Also move all MIPI-related headers out from device/ into their own
directory.
Signed-off-by: Julius Werner <jwerner@chromium.org>
Change-Id: Ib3e66954b8f0cf85b28d8d186b09d7846707559d
Reviewed-on: https://review.coreboot.org/c/coreboot/+/57128
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Hung-Te Lin <hungte@chromium.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Reviewed-by: Nico Huber <nico.h@gmx.de>
Move bcd2bin() / bin2bcd() functions to commonlib/bsd/include/
Also, the license is changed from GPL to BSD.
This is because it is needed from "utils" (see CL in the chain).
For reference bin2bcd() & bcd2bin() are very simple functions.
There are already BSD implementations, like these ones (just to
name a few):
https://chromium.googlesource.com/chromiumos/platform/mosys/+/refs/heads/main/include/lib/math.h#67http://web.mit.edu/freebsd/head/sys/contrib/octeon-sdk/cvmx-cn3010-evb-hs5.c
BUG=b:172210863
TEST=make (everything compiled Ok).
Change-Id: If2eba82da35838799bcbcf38303de6bd53f7eb72
Signed-off-by: Ricardo Quesada <ricardoq@google.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/56904
Reviewed-by: Furquan Shaikh <furquan@google.com>
Reviewed-by: Paul Menzel <paulepanter@mailbox.org>
Reviewed-by: Jack Rosenthal <jrosenth@chromium.org>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
To reduce suspend power consumption,
1. Disable unused CLKSQ2.
2. Set CLKSQ_EN to sleep control for SPM 26M sleep control.
No bus clock when enter 26m sleep control, and only control
clock square by side band.
Signed-off-by: Chun-Jie Chen <chun-jie.chen@mediatek.com>
Change-Id: Ia9a1735d6f508ce35b9af2d67831a3474255198b
Reviewed-on: https://review.coreboot.org/c/coreboot/+/57043
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
All boards that are trying to use MIPI panels eventually run into the
problem that they need to store physical parameters and a list of DCS
initialization commands for each panel, and these commands can be very
different (e.g. a large amount of very short commands, a few very large
commands, etc.). Finding a data format to fit all these different cases
efficiently into the same structures keeps being a challenge, and the
Kukui mainboard already once put a lot of effort into designing a
clean, flexible and efficient solution for this. This patch moves that
framework into a common src/device/mipi/ library where it can be used by
other boards as well. (Also, this will hopefully allow us to save some
duplicated work when using the same panel on different boards at some
point.)
Signed-off-by: Julius Werner <jwerner@chromium.org>
Change-Id: I877f2b0c7ab984412b288e2ed27f37cd93c70863
Reviewed-on: https://review.coreboot.org/c/coreboot/+/56965
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Reviewed-by: Hung-Te Lin <hungte@chromium.org>
Move DFD driver to common folder so MT8195 can also use it.
BUG=b:192429713
TEST=emerge-asurada coreboot
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: I7937cddf5f3a66f9269a94301d3134e6f4f9f22e
Reviewed-on: https://review.coreboot.org/c/coreboot/+/56908
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
The MEDIATEK_DRAM_DVFS_LIMIT_FREQ_CNT only makes sense if DVFS
is enabled (e.g., MEDIATEK_DRAM_DVFS) so we should change it to
depend on that instead of selecting DVFS.
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: Ib81e4e48e863616ed1e36cd5c0000f4e2cfb5456
Reviewed-on: https://review.coreboot.org/c/coreboot/+/56906
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Add mt6359p vcore set/get support.
To adjust frequency of little core, we need to adjust voltage of vcore.
Signed-off-by: James Lo <james.lo@mediatek.corp-partner.google.com>
Change-Id: Ibf49390ba78870b834c6d0b64e3f0f30f3494f18
Reviewed-on: https://review.coreboot.org/c/coreboot/+/56843
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Rex-BC Chen <rex-bc.chen@mediatek.corp-partner.google.com>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
DFD (Design for Debug) is a debugging tool, which scans
flip-flops and dumps to internal RAM on the WDT reset.
After system reboots, those values could be showed for
debugging.
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: I39a4391c1d1e832d77b709f8f899bb1c6dcacd69
Reviewed-on: https://review.coreboot.org/c/coreboot/+/56797
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Add basic devapc (device access permission control) drivers.
DAPC driver is used to set up bus fabric security and data protection
among hardwares. DAPC driver groups the master hardwares into different
domains and gives secure and non-secure property. The slave hardware can
configure different access permissions for different domains via DAPC
driver.
1. Initialize devapc.
2. Set master domain and secure side band.
3. Set domain remap.
4. Set default permission.
Change-Id: I3677657a117caed0d73526f78b0ebe8180148335
Signed-off-by: Nina Wu <nina-cm.wu@mediatek.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/56764
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
The GENMASK is defined in multiple files (with various names such as
MASKBIT), which sets certain consecutive bits to 1 and leaves the others
to 0. To avoid duplicate macros, add GENMASK macro to helpers.h.
GENMASK(high, low) sets bits from `high` to `low` (inclusive) to 1. For
example, GENMASK(39, 21) gives us the 64-bit vector 0x000000ffffe00000.
Remove duplicate macro definitions. Also utilize GENMASK for _BF_MASK in
mmio.h.
BUG=none
TEST=make tests/commonlib/bsd/helpers-test
TEST=emerge-cherry coreboot
BRANCH=none
Change-Id: If2e7c4827d8a7d27688534593b556a72f16f0c2b
Signed-off-by: Yu-Ping Wu <yupingso@chromium.org>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/56543
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Julius Werner <jwerner@chromium.org>
With the new definition of mt6360_regulator_id,
merge the MT6360 LDO and PMIC interfaces into one.
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: I7ccc32cb0a9481d5f55349c152267a44fe09d20a
Reviewed-on: https://review.coreboot.org/c/coreboot/+/56435
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Hung-Te Lin <hungte@chromium.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
On MT8195 platforms with BC1.2, we have to use EC to control
MT6360 so the mt6360_regulator_id is redefined to match the
numbers defined in EC driver.
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: I9437edb9776442759ce04c31d315c3760078ffb3
Reviewed-on: https://review.coreboot.org/c/coreboot/+/56434
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
The dramc_param.h defines the header version,
structure and APIs for the DRAM calibration parameters
stored on the flash, and should be platform independent.
Signed-off-by: Ryan Chuang <ryan.chuang@mediatek.corp-partner.google.com>
Change-Id: Ib8a6ea1b6cf1538854890b653d5d9a934f7f687e
Reviewed-on: https://review.coreboot.org/c/coreboot/+/56098
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Enable DCM settings on the MT8195 platform.
DCM means dynamic clock management, and it can dynamically
slow down or gate clocks during CPU or bus idle.
Change-Id: Ib431a0334c157d440d6e89dcb154241d980d97ce
Signed-off-by: Garmin Chang <garmin.chang@mediatek.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/56097
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Add mt6691 buck control for DRAM to run fast calibration test.
It is needed to get and set voltage during testing.
Signed-off-by: Henry Chen <henryc.chen@mediatek.com>
Change-Id: I4fb9f7245d44383a6a3a0cf8d00f7f503cbdeb06
Reviewed-on: https://review.coreboot.org/c/coreboot/+/55575
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Reviewed-by: Hung-Te Lin <hungte@chromium.org>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
This patch adds support for loading SPM firmware from CBFS to SPM SRAM.
SPM needs its own firmware to enable SPM suspend/resume function which
turns off several resources such as DRAM/mainpll/26M clk when linux
system suspend.
TEST=program counter of SPM is correct value after booting up.
Change-Id: Ia0f9b9f86e44b293c1cc47213946304c64aea75e
Signed-off-by: Dawei Chien <dawei.chien@mediatek.com>
Signed-off-by: Edward-JW Yang <edward-jw.yang@mediatek.corp-partner.google.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/55140
Reviewed-by: Rex-BC Chen <rex-bc.chen@mediatek.corp-partner.google.com>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Load SSPM firmware and boot up SSPM in ramstage.
This adds 23ms to the boot time.
TEST=Load SSPM blob ok, and we can see some logs of SSPM from AP.
Signed-off-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Change-Id: Ia227ea9f7d58129068cb36ec2de7d9feb677006b
Reviewed-on: https://review.coreboot.org/c/coreboot/+/55051
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Initialize DRAM in romstage and configure to support fast calibration.
Change-Id: I89b87be62c8e88ae4a620d56aa7a35e47f97952d
Signed-off-by: Yidi Lin <yidi.lin@mediatek.com>
Signed-off-by: Ryan Chuang <ryan.chuang@mediatek.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/54229
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
vpp_sel and ethdr_sel are vdosys clock source select mux.
Steps to change to support 4K source.
1. Change vpp_sel source to mainpll_d4 to run at 546MHz
2. Change ethdr_sel source to univpll_d6 to run at 416MHz
Signed-off-by: Nancy Lin <nancy.lin@mediatek.com>
Change-Id: Ib6518ed6204528489c41e7161534bbd3734ac851
Reviewed-on: https://review.coreboot.org/c/coreboot/+/54082
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
