| Commit message (Collapse) | Author | Age | Lines |
|
|
|
|
|
|
|
|
|
|
|
| |
Add download_mode command in uboot to enter MFG dowload mode ,
you can try download mode command in uboot and enter download mode.
it first set srtc register, then before enter linux,
it will clear these register to prevent the up comming watchdog
reset will enter mfgtool mode.
only add mx53 now.
Signed-off-by: Zhang Jiejing <jiejing.zhang@freescale.com>
|
|
|
|
|
|
|
|
|
|
|
| |
Implement a key press check on recovery mode check.
User can press Vol- key to enter recovery mode when boot.
Idealy, should be a combo key press together, but on SMD
it only can Vol+ or Vol- but it can't press together.
More usuful for user and less bug.
Signed-off-by: Zhang Jiejing <jiejing.zhang@freescale.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
In addition to ensuring that PERCLK remains at least 2.5 times slower
than the AHB clock, certain steps need to be followed to ensure robust
operation of PERCLK when reconfiguring the PERCLK clock source.
To properly configure the PERCLK clock source, the following steps are
required:
1.In the CCGR registers, gate the clocks to all PERCLK-dependent
modules.
2.Select the desired input clock for the PERCLK root clock (to be either
source from the peripherals main source clock or the
lp_apm clock source). Refer to the CMCBR register,
perclk_lp_apm_sel bit.
3.Configure the perclk_pred1, perclk_pred2, and perclk_podf dividers
to the desired setting. Refer to the CBCDR register for details.
4.In the CCGR registers, enable the desired clocks for the
PERCLK-dependent module clocks.
Certain steps are required to reconfigure perclk_root.
If don't follow these steps, GPT timer may stop and the kernel stops
at " "Calibrating delay loop".
Signed-off-by: Terry Lv <r65388@freescale.com>
|
|
|
|
|
|
|
| |
Enable RGBA formated fb buffers.
And increase the VPU pmem to 32MB.
Signed-off-by: Xinyu Chen <xinyu.chen@freescale.com>
|
|
|
|
|
|
|
| |
300MB vmalloc vm area is not enough for GPU driver.
enlarge to 400.
Signed-off-by: Zhang Jiejing <jiejing.zhang@freescale.com>
|
|
|
|
|
|
| |
Incorrect usb string package size assign.
Signed-off-by: Xinyu Chen <xinyu.chen@freescale.com>
|
|
|
|
|
|
|
| |
enlarge max image size to 148MB from 128MB. since android become bigger.
so 128M not enough.
Signed-off-by: Zhang Jiejing <jiejing.zhang@freescale.com>
|
|
|
|
|
|
|
|
| |
The latest kernel with memory optimize needs fb0 reserve
buffer size, pmem size and somehow vmalloc size configuration
in boot command line. Here change the GB/HC uboot to align.
Signed-off-by: Xinyu Chen <xinyu.chen@freescale.com>
|
|
|
|
|
|
|
| |
Increase the DDR size to 128M.
Increase the GPU memory to 128M in default bootargs.
Signed-off-by: Xinyu Chen <xinyu.chen@freescale.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
For DA9053 I2C SDA stuck low issue: the I2C block in DA9053 may not correctly
receive a Power On Reset and device is in unknown state during start-up.
The only way to get the chip into known state before any communication
with the Chip via I2C is to dummy clock the I2C and bring it in a state
where I2C can communicate. Dialog suggested to provide 9 clock on SCL.
Dialog don't know the exact reason for the fault and assume it is because
some random noise or spurious behaviour.
This has to been done in host platform specific I2C driver during
start-up when the I2C is being configured at platform level to supply with
dummy 9 clock on SCL. Dialog I2C driver has no control to provide dummy 9
clock on SCL.
Signed-off-by: Wayne Zou <b36644@freescale.com>
|
|
|
|
|
|
|
|
|
|
|
|
| |
MX53 SMD hangs if reset many times with lower possibility.
If doing I2C access in early time, I2C may cause system hangs.
So moving I2C access to late phase to make system hang issue disappear.
QA Test result: QA raised 6 full rounds of CTS one-round test
Totally ran for 6 rounds about 27 hours, reboot for 56*6=336 times,
no reboot failure occurred.
Signed-off-by: Wayne Zou <b36644@freescale.com>
|
|
|
|
|
|
| |
remove printf() because serial interface is not ready in board_init()
Signed-off-by: Wayne Zou <b36644@freescale.com>
|
|
|
|
|
|
|
|
|
| |
when board boots up, during the iMX53 SOC does DA9053 Read/Write
operation, it writes slave address and wait for ACK . Instead of ACK
PMIC sends NAK. A workaround fix is provided as a part of retries to
fix I2C NAK for very first access.
Signed-off-by: Wayne Zou <b36644@freescale.com>
|
|
|
|
|
|
|
| |
since bootargs_base already removed.
also remove it in recovery boot cmdline.
Signed-off-by: Zhang Jiejing <jiejing.zhang@freescale.com>
|
|
|
|
|
|
|
| |
In mx51 configuration, CONFIG_BOOT_PARTITION_ACCESS is not defined.
This cause build error to fastboot.c
Signed-off-by: Xinyu Chen <xinyu.chen@freescale.com>
|
|
|
|
|
|
| |
Change DCDC_3V15's GPIO setting for REV-D.
Signed-off-by: Anson Huang <b20788@freescale.com>
|
|
|
|
|
|
|
|
|
|
| |
This patch is used to support watchdog timeout in SMD RevA, RevB
board.
1. Revert "ENGR00143469 mx53 smd: pull down GPIO_9 to reset the
board".
2. Force warm reset as cold reset.
Signed-off-by: Lily Zhang <r58066@freescale.com>
|
|
|
|
|
|
|
|
|
|
|
| |
Updated mx53 ddr3 script according to MX53_TO2_DDR3_LCB_SMD_ARDb_v1.inc
from Michael J Kjar on July 8, 2011:
-change ESDREF[REF_SEL]=01 (for 32KHz), from incorrect setting of 00 (64KHz)
-change DDR3 MR0 write to "setmem /32 0x63fd901c = 0x052080b0" from
"0x092080b0". This chagned write recovery from 8 clocks to 6 clocks
(in line with ESDCFG1[tWR])
Signed-off-by: Lily Zhang <r58066@freescale.com>
|
|
|
|
|
|
|
| |
uboot image cannot be burned to boot partition for eMMC 4.3. This
patch will fix it.
Signed-off-by: Sammy He <r62914@freescale.com>
|
|
|
|
|
|
|
|
|
| |
Use simple enviroment to implement the default boot command.
The original one is too complex, and not readable.
For MX51BBG, only SD card boot env is supportd by default.
For MX53SMD, only eMMC boot env is supportd by default.
Signed-off-by: Xinyu Chen <xinyu.chen@freescale.com>
|
|
|
|
|
|
|
| |
Drop NAND/SPI boot support.
Enable fastboot.
Signed-off-by: Xinyu Chen <xinyu.chen@freescale.com>
|
|
|
|
|
|
| |
update default cmdline to align with Document.
Signed-off-by: Zhang Jiejing <jiejing.zhang@freescale.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
ROM requires DCD table instead of plugin to initialize DRAM if emmc fastboot
mode is to be used. Therefore, switched the DRAM script from plugin to
DCD table. The DCD table created is based on the following RVD script:
Arik_init_DDR3_528MHz_002.inc found at
http://compass.freescale.net/livelink/livelink?func=ll&objId=222928845
When fastboot mode is used by ROM, the MMC_BOOT register of USDHC does not
get reset when RSTA bit is set by uboot driver. Therefore, need to write 0
to it manually during driver init. This brings USDHC out of fastboot mode,
allowing normal communication with emmc to proceed in uboot.
Changed comments for DLL delay to be more accurate.
Signed-off-by: Anish Trivedi <anish@freescale.com>
|
|
|
|
|
|
|
|
|
|
|
| |
New bit definitions in USDHC.
Added is_usdhc variable to fsl_esdhc_cfg to distinguish between ESDHC
and USDHC.
Enabled DDR mode support in USDHC.
Created a config to customize target delay for DDR mode.
Modified USDHC pad settings to make DDR mode work for all emmcs at 50 MHz.
Signed-off-by: Anish Trivedi <anish@freescale.com>
|
|
|
|
|
|
| |
Make sure the PLL workaround is done only for PLL1.
Signed-off-by: Ranjani Vaidyanathan <ra5478@freescale.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
In precode, PHY forced to work at 100M even connect to
1G switch.
In this commit, let PHY auto negotiate it working speed. Enet tx
work at store-and-forward mode.
BTW, AR8031 take quite a long time, about 1.6s from negotiation to link up.
we have to wait and then set ENET correctly.
Signed-off-by: Zeng Zhaoming <b32542@freescale.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Add ENET and AR8031 PHY support to uboot.
To make it works on sabreauto, need do following changes:
1. rework phy to output 125M clock from CLK_25M signal,
and the 125M clock input to SoC as reference clock to generate
RGMII_TXC clock.
2. Enable TXC delay in PHY debug register.
3. set ENET working in RMII mode.
4. set ENET working at 1000M or 100M/10M.
5. set ENET TX fifo to maximum to avoid underrun error.
6. force AR8031 PHY working at 100M
Signed-off-by: Anson Huang <b20788@freescale.com>
Signed-off-by: Zeng Zhaoming <b32542@freescale.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Use 528M DDR script
Disable L2 cache because rom enable L2 cache when use plug-in
Fix usdhc pad settings
Remove mac address hardcode
Signed-off-by: Anson Huang <b20788@freescale.com>
Signed-off-by: Richard Zhu <r65037@freescale.com>
Signed-off-by: Terry Lv <r65388@freescale.com>
Signed-off-by: Anish Trivedi <anish@freescale.com>
Signed-off-by: Jason Liu <r64343@freescale.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Apply the following SW workaround to fix the PLL unlock issue.
1.Move all the clock sources which are currently running
on PLL1 from PLL1 to PLL2
2.Clear AREN bit in PLL1 (to avoid restart during MFN change)
3.Program the PLL1 to the next settings:
a. MFI = 8
b. MFD = 179
c. MFN = 180
d. PLM = 1
4.Manually restart the PLL1
5.Wait to PLL1 to lock
6.Reprogram the PLL1 to the next settings:
a. MFI = 60, others keep same
7.Load the MFN
8.Wait for LDREQ and delay ~4.6us
9.Switch the clocks which were previously moved from PLL1 to PLL2 back to PLL1
Signed-off-by: Ranjani Vaidyanathan <ra5478@freescale.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
PLL1 workaround to prevent it from losing lock:
(1) Disable AREN bit to avoid PLL1 restart during MFN change
(2) set PLL1 to ~864Mhz with MFI = 8, MFN = 180, MFD = 179, PDF = 0
(3) Manual restart PLL1
(4) Wait PLL1 lock
(5) Set PLL1 to 800Mhz with only change MFN to 60, others keep
(6) Set LDREQ bit to load new MFN
(7) Poll on LDREQ bit for MFN update to be completed
(8) Delay at least 4 us to avoid PLL1 instability window
(9) Switch ARM back to PLL1
Signed-off-by: Anish Trivedi <anish@freescale.com>
|
|
|
|
|
|
|
| |
After reseting in stop mode, the VUSB_2V5 voltage is disable by pmic.
It needs to be enable manually in u-boot.
Signed-off-by: Wayne Zou <b36644@freescale.com>
|
|
|
|
|
|
|
| |
Not all peripherals are mapped in MMU.
Thus we add those missed mapped area.
Signed-off-by: Terry Lv <r65388@freescale.com>
|
|
|
|
|
|
| |
Change VCC from 1.35V to 1.3V QS Ripley board
Signed-off-by: Wayne Zou <b36644@freescale.com>
|
|
|
|
|
|
| |
Change the recovery boot for MX53_SMD to emmc 's device 1.
Signed-off-by: Zhang Jiejing <jiejing.zhang@freescale.com>
|
|
|
|
|
|
|
|
|
| |
Enable NAND gpio, recovery mode detect after boot from spi nor.
Change default env for loading kernel and uramdisk from NAND,
disabling elcdif lcd driver to support EPDC eink panel as default.
Enable recovery mode support and NAND/UBI/UBIFS command.
Signed-off-by: Xinyu Chen <xinyu.chen@freescale.com>
|
|
|
|
|
|
| |
Nand oobsize is wrong in some nand chips.
Signed-off-by: Terry Lv <r65388@freescale.com>
|
|
|
|
|
|
|
| |
For clk command always make console output mess characters,
here we reinitilize it after clock is changed.
Signed-off-by: Terry Lv <r65388@freescale.com>
|
|
|
|
|
|
| |
Add mc34708 pmic support on loco/Ripley board
Signed-off-by: Zou Weihua -wayne zou <b36644@freescale.com>
|
|
|
|
|
|
|
| |
Change the default environment setting as sd boot for mx53
loco, mx53 smd and mx53 ard boards.
Signed-off-by: Lily Zhang <r58066@freescale.com>
|
|
|
|
|
|
|
| |
update the ramdisk load address due to android kernel size enlarge.
the ramdisk memory load address is 5MB offset to kernel.
Signed-off-by: Xinyu Chen <xinyu.chen@freescale.com>
|
|
|
|
|
|
|
|
|
| |
In mx53 smd, to type "reset" command in u-boot console can
not reset the system. It hangs in ROM with unknown reason.
This patch adds one workaround to configure GPIO_9 (WDT_OUTPUT_B)
as GPIO and pull down it to reset DA9053 PMIC.
Signed-off-by: Lily Zhang <r58066@freescale.com>
|
|
|
|
|
|
|
|
| |
The default VDDGP output voltage is 1.05V in mx51 evk board
According to mx51 datasheet (Rev 0.4), the VDDGP for 800MHZ
should be 1.1v for 800MHZ
Signed-off-by: Lily Zhang <r58066@freescale.com>
|
|
|
|
|
|
|
|
|
|
|
| |
Don't need let ROM copy the .bss section since it
will all be zeroed by u-boot at start up, thus it
can speed up the boot up time.
Need add CONFIG_FLASH_HEADER_OFFSET to the size since
ROM will copy from the beginning of the MMC card.
Signed-off-by: Jason Liu <r64343@freescale.com>
|
|
|
|
|
|
|
|
|
|
|
|
| |
Add mx50_rd3_android default config file
Add basic support for UBI partition mount and UBIFS file read for recovery
Add gpmi nand enable in MFG kernel commandline by uboot configure,
which enable MFG tool to flash system images on NAND.
The total NAND boot and NAND recovery has been disabled.
They will be enabled later.
Signed-off-by: Xinyu Chen <xinyu.chen@freescale.com>
|
|
|
|
|
|
| |
Switch to use SATA internal clock in mx53 ARD RevB board.
Signed-off-by: Lily Zhang <r58066@freescale.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
On SD3 on MX50, there is an option to choose eSDHC or uSDHC controller.
By default eSDHC is selected. However, eSDHC shows some borderline timing
in SDR mode at 50 MHz, whereas uSDHC shows borderline timing in DDR mode
at 50 MHz. Therefore, add a compile time option to uboot for MX50 to
select uSDHC in SDR mode or eSDHC in DDR mode on SD3 port.
By default the compile time option, CONFIG_MX50_ENABLE_USDHC_SDR,
is commented out in the include/configs/mx50_<board>.h file to
select eSDHC with DDR mode enabled. Uncomment the define to select
uSDHC with only SDR mode enabled.
Also increased max frequency supported by ESDHC to 52 MHz instead
of 50 MHz.
Signed-off-by: Anish Trivedi <anish@freescale.com>
|
|
|
|
|
|
| |
Add MFG tool support for MX53 ARD DDR3 board.
Signed-off-by: Lily Zhang <r58066@freescale.com>
|
|
|
|
|
|
|
|
| |
As android make UBIFS as default file system on MX508 RDP,
we must enable MFG tool to support UBIFS image update.
So enable GPMI nand in default commandline to detect nand devices.
Signed-off-by: Xinyu Chen <xinyu.chen@freescale.com>
|
|
|
|
|
|
|
|
|
|
| |
Check eMMC and SD cards recovery file, if it exist, enter recovery mode.
original code only check SD card, since we already change main storage
to eMMC, so we check it both, since most of customer still test it under
SD card, check them to avoid support effert.
Signed-off-by: Zhang Jiejing <jiejing.zhang@freescale.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Spi nor can't erase 0x200000 size.
There are two issues in this CR.
1. Spi nor can't erase 0x200000 size.
2. Whole chip erase don't work.
The solution will be:
1. Delay more time for WIP check.
2. Use normal erase for whole chip erase.
Signed-off-by: Terry Lv <r65388@freescale.com>
|