York Sun [Mon, 7 Dec 2015 19:08:58 +0000 (11:08 -0800)]
armv8: fsl-layerscale: Rewrite reserving memory for MC and debug server
MC and debug server are not board-specific. Move reserving memory to SoC
file, using the new board_reserve_ram_top function. Reduce debug server
memory by 2MB to make room for secure memory.
In the system with MC and debug server, the top of u-boot memory
is not the end of memory. PRAM is not used for this reservation.
York Sun [Mon, 7 Dec 2015 19:05:29 +0000 (11:05 -0800)]
common: Rewrite hiding the end of memory
As the name may be confusing, the CONFIG_SYS_MEM_TOP_HIDE reserves
some memory from the end of ram, tracked by gd->ram_size. It is not
always the top of u-boot visible memory. Rewrite the macro with a
weak function to provide flexibility for complex calcuation. Legacy
use of this macro is still supported.
Signed-off-by: York Sun <yorksun@freescale.com> Reviewed-by: Simon Glass <sjg@chromium.org>
York Sun [Fri, 4 Dec 2015 19:57:08 +0000 (11:57 -0800)]
armv8: fsl-layerscape: Make DDR non secure in MMU tables
DDR has been set as secure in MMU tables. Non-secure master such
as SDHC DMA cannot access data correctly. Mixing secure and non-
secure MMU entries requirs the MMU tables themselves in secure
memory. This patch moves MMU tables into a secure DDR area.
Early MMU tables are changed to set DDR as non-secure. A new
table is added into final MMU tables so secure memory can have
2MB granuality.
gd->secure_ram tracks the location of this secure memory. For
ARMv8 SoCs, the RAM base is not zero and RAM is divided into several
banks. gd->secure_ram needs to be maintained before using. This
maintenance is board-specific, depending on the SoC and memory
bank of the secure memory falls into.
York Sun [Fri, 4 Dec 2015 19:57:07 +0000 (11:57 -0800)]
Reserve secure memory
Secure memory is at the end of memory, separated and reserved
from OS, tracked by gd->secure_ram. Secure memory can host
MMU tables, security monitor, etc. This is different from PRAM
used to reserve private memory. PRAM offers memory at the top
of u-boot memory, not necessarily the real end of memory for
systems with very large DDR. Using the end of memory simplifies
MMU setup and avoid memory fragmentation.
"bdinfo" command shows gd->secure_ram value if this memory is
marked as secured.
Tom Rini [Sat, 28 Nov 2015 13:04:42 +0000 (08:04 -0500)]
qbman_portal.c: Update BUG_ON() call in qbman_swp_mc_submit
With gcc-5.x we get a warning about the ambiguity of BUG_ON(!a != b) and
becomes BUG_ON((!a) != b). In this case reading of the function leads to
us wanting to rewrite this as BUG_ON(a != b).
Cc: Prabhakar Kushwaha <prabhakar@freescale.com> Cc: Geoff Thorpe <Geoff.Thorpe@freescale.com> Cc: Haiying Wang <Haiying.Wang@freescale.com> Cc: Roy Pledge <Roy.Pledge@freescale.com> Cc: York Sun <yorksun@freescale.com> Signed-off-by: Tom Rini <trini@konsulko.com> Reviewed-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: York Sun <yorksun@freescale.com>
Tom Rini [Sat, 28 Nov 2015 13:04:41 +0000 (08:04 -0500)]
fsl_*_serdes.c: Modify memset call in serdes_init
GCC 5.x does not like sizeof(array_variable) and errors out. Change these
calls to be instead sizeof(u8) (as that's what serdes_prtcl_map is) *
SERDES_PRCTL_COUNT (the number of array elements).
Cc: York Sun <yorksun@freescale.com> Signed-off-by: Tom Rini <trini@konsulko.com> Reviewed-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: York Sun <yorksun@freescale.com>
Shaohui Xie [Mon, 23 Nov 2015 07:23:48 +0000 (15:23 +0800)]
armv8/ls1043ardb: Add support for >2GB memory
This patch also expose the complete DDR region(s) to Linux.
Signed-off-by: Shaohui Xie <Shaohui.Xie@freescale.com> Signed-off-by: Mingkai Hu <Mingkai.Hu@freescale.com> Reviewed-by: York Sun <yorksun@freescale.com>
Shaohui Xie [Tue, 10 Nov 2015 11:20:16 +0000 (19:20 +0800)]
freescale: fman: make sure phy-handle property is big endian
When creating phy-handle property, an unsigned int value is created by
fdt_create_phandle, and memcpy is used to get the value, since DTS is
big endian, the value cannot be used directly on little endian SoCs,
it should be converted by cpu_to_fdt32.
Signed-off-by: Shaohui Xie <Shaohui.Xie@freescale.com> Reviewed-by: York Sun <yorksun@freescale.com>
Jagan Teki [Sun, 6 Dec 2015 16:03:32 +0000 (21:33 +0530)]
sf: sf_probe: Remove spi_slave pointer argument
Since spi_slave is a spi pointer in spi_flash{} then assign
spi_slave{} pointer to flash->spi and remove spi_slave
pointer argument to
- spi_flash_probe_slave
- spi_flash_scan
Tested-by: Jagan Teki <jteki@openedev.com> Tested-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Simon Glass <sjg@chromium.org> Signed-off-by: Jagan Teki <jteki@openedev.com>
Jagan Teki [Wed, 25 Nov 2015 19:33:33 +0000 (01:03 +0530)]
sf: Remove unneeded SST_BP and SST_WP
SST parts added on sf_params.c supports both SST_WR which consits
of both BP and WP and there is a spi controller ich which supports
only BP so the relevent _write hook set based on "slave->op_mode_tx"
hence there is no respective change required from flash side hance
removed these.
Tested-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Bin Meng <bmeng.cn@gmail.com> Signed-off-by: Jagan Teki <jteki@openedev.com>
Jagan Teki [Tue, 29 Sep 2015 20:31:23 +0000 (02:01 +0530)]
sf: Flash power up read-only based on idcode0
Using macro's for flash power up read-only access code
leads wrong behaviour hence use idcode0 for runtime
detection, hence the flash which require this functionality
gets detected at runtime.
Tested-by: Jagan Teki <jteki@openedev.com> Reviewed-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Simon Glass <sjg@chromium.org> Signed-off-by: Jagan Teki <jteki@openedev.com>
Jagan Teki [Tue, 29 Sep 2015 16:59:33 +0000 (22:29 +0530)]
sf: Use simple name for register access functions
Most of the register access function are static,
so used simple name to represent each.
Tested-by: Jagan Teki <jteki@openedev.com> Reviewed-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Simon Glass <sjg@chromium.org> Signed-off-by: Jagan Teki <jteki@openedev.com>
Jagan Teki [Tue, 3 Nov 2015 18:57:35 +0000 (00:27 +0530)]
sf: probe: Code cleanup
- Move bar read code below the bar write hance both
at once place, hence it easy for #ifdef macro only
once and readable.
- Move read_cmd_array at top
Tested-by: Jagan Teki <jteki@openedev.com> Reviewed-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Simon Glass <sjg@chromium.org> Signed-off-by: Jagan Teki <jteki@openedev.com>
Jagan Teki [Tue, 29 Sep 2015 12:36:04 +0000 (18:06 +0530)]
sf: Move read_id code to sf_ops
read_id code is related to spi_flash stuff
hence moved to sf_ops.
Tested-by: Jagan Teki <jteki@openedev.com> Tested-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Simon Glass <sjg@chromium.org> Signed-off-by: Jagan Teki <jteki@openedev.com>
Jagan Teki [Fri, 11 Dec 2015 16:06:34 +0000 (21:36 +0530)]
sf: Move spi_flash_scan code to sf_ops
Intension is that sf_ops should deals all spi_flash
related stuff and sf_probe (which should renamed future)
should be an interface layer for spi_flash versus spi drivers.
sf_ops => spi_flash interface
sf_probe => interface layer vs spi_flash(sf_probe) to spi drivers
Tested-by: Jagan Teki <jteki@openedev.com> Tested-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Simon Glass <sjg@chromium.org> Signed-off-by: Jagan Teki <jteki@openedev.com>
Jagan Teki [Tue, 29 Sep 2015 11:58:20 +0000 (17:28 +0530)]
sf: spi_flash_validate_params => spi_flash_scan
Rename spi_flash_validate_params to spi_flash_scan
as this code not only deals with params setup but
also configure all spi_flash attributes.
And also moved all flash related code into
spi_flash_scan for future functionality addition.
Tested-by: Jagan Teki <jteki@openedev.com> Tested-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Simon Glass <sjg@chromium.org> Signed-off-by: Jagan Teki <jteki@openedev.com>
Peng Fan [Mon, 30 Nov 2015 09:45:02 +0000 (17:45 +0800)]
imx: mx7dsabresd: Add QSPI support
Support qspi flashes for mx7dsabresd
1. introduce pin mux settings
2. enable qspi clock
3. introduce related macro definitions
Default QSPI is not enabled, since we need hardware rework to use QSPI,
see SPF-28590, page 9:
"
QSPI signals are muxed with EPDC_D[7:0]
When using QSPI: de-populate R388-R391, R396-R399
populate R392-R395, R299, R300
"
After hardware rework, define CONFIG_FSL_QSPI in mx7dsabresd.h. qspi
flashes can be deteced and read/erase/write. Log info:
"
=> sf probe
SF: Detected MX25L51235F with page size 256 Bytes, erase size 64 KiB, total 64 MiB
=> sf read 0x80000000 0 0x4000000
device 0 whole chip
SF: 67108864 bytes @ 0x0 Read: OK
=> sf erase 0 0x4000000
SF: 67108864 bytes @ 0x0 Erased: OK
=> sf write 0x80000000 0 0x4000000
device 0 whole chip
SF: 67108864 bytes @ 0x0 Written: OK
"
Cc: Fabio Estevam <fabio.estevam@freescale.com> Cc: Adrian Alonso <aalonso@freescale.com> Reviewed-by: Stefano Babic <sbabic@denx.de> Reviewed-by: Jagan Teki <jteki@openedev.com> Signed-off-by: Peng Fan <Peng.Fan@freescale.com>
Stefan Roese [Thu, 26 Nov 2015 12:38:43 +0000 (13:38 +0100)]
dm: serial: Minor coding style cleanup of some comments
Fix incorrect comment alignments.
Signed-off-by: Stefan Roese <sr@denx.de> Cc: Simon Glass <sjg@chromium.org> Reviewed-by: Bin Meng <bmeng.cn@gmail.com> Acked-by: Simon Glass <sjg@chromium.org>
Stefan Roese [Thu, 26 Nov 2015 12:38:01 +0000 (13:38 +0100)]
dm: core: Fix Kconfig text to mention SPL in SPL_OF_TRANSLATE
Add a remark about SPL to this Kconfig option. Otherwise its identitcal
to the non-SPL version, which is confusing.
Signed-off-by: Stefan Roese <sr@denx.de> Cc: Simon Glass <sjg@chromium.org> Acked-by: Simon Glass <sjg@chromium.org> Acked-by: Michal Simek <michal.simek@xilinx.com>
Jens Kuske [Thu, 10 Dec 2015 10:41:02 +0000 (11:41 +0100)]
sunxi: Fix H3 DRAM DQ read delay configuration
The read delays were set incorrectly, leading to reliability
issues at higher DRAM clock speeds. This commit adjusts this
to match the vendor boot0 behaviour.
Signed-off-by: Jens Kuske <jenskuske@gmail.com> Tested-by: Siarhei Siamashka <siarhei.siamashka@gmail.com> Reviewed-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Olliver Schinagl <o.schinagl@ultimaker.com> Reviewed-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Commit 6c739c5d added code to enable i2c bus 4 and 5 on the sun7i SoC
but forgot to extend the range check in clock_twi_onoff, resulting in
the clock not getting enabled.
The range-check is not needed at all, since clock_twi_onoff only gets
called with such high indexes when CONFIG_I2C3_ENABLE / CONFIG_I2C4_ENABLE
is set and Kconfig already only allows these on sun6i / sun7i.
This commit removes the range-check all together fixing i2c bus 4 and 5
not working on sun7i.
Signed-off-by: Olliver Schinagl <o.schinagl@ultimaker.com>
[hdegoede@redhat.com] Remove range check instead of extending it Signed-off-by: Hans de Goede <hdegoede@redhat.com>
vishnupatekar [Sat, 28 Nov 2015 17:07:27 +0000 (01:07 +0800)]
sunxi: Add suport for A83T HomletV2 Board by Allwinner
Add dts and defconfig for h8homletv2 board.
H8Homlet Proto v2.0 Board is A83T Dev Board by Allwinner.
It has UART, ethernet, USB, HDMI, etc ports on it.
A83T patches are tested on this board.
Signed-off-by: Vishnu Patekar <vishnupatekar0510@gmail.com> Reviewed-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Hans de Goede <hdegoede@redhat.com>
vishnupatekar [Sat, 28 Nov 2015 17:07:26 +0000 (01:07 +0800)]
sunxi: dts: sun8i: Add Allwinner A83T dtsi
Allwinner A83T is new octa-core cortex-a7 SOC.
This adds the basic dtsi, the clocks differs from
earlier sun8i SOCs.
This is not yet included in kernel.
Signed-off-by: Vishnu Patekar <vishnupatekar0510@gmail.com> Reviewed-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Hans de Goede <hdegoede@redhat.com>
vishnupatekar [Sat, 28 Nov 2015 17:07:25 +0000 (01:07 +0800)]
sunxi: Add support for Allwinner A83T DRAM
Add support for A83T dram. Register are different from sun8i A33.
init code is similar to A33 dram init.
hope we'll shift duplicate code in dram_sun8i_*
to dram helper in future.
Signed-off-by: Vishnu Patekar <vishnupatekar0510@gmail.com> Reviewed-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Hans de Goede <hdegoede@redhat.com>
vishnupatekar [Sat, 28 Nov 2015 17:07:24 +0000 (01:07 +0800)]
sunxi: clk: add basic clocks for A83T
Add basic clocks pll1, pll5, and some default values from allwinner u-boot.
Signed-off-by: Vishnu Patekar <vishnupatekar0510@gmail.com>
[hdegoede@redhat.com] Fix PLL6 init to run at 600 MHz instead of 288 MHz,
fixing the mmc support not working
[hdegoede@redhat.com] Fix PLL init code to properly wait for the PLL-s to
stabilize, fixing cold-booting directly from sdcard not working Signed-off-by: Hans de Goede <hdegoede@redhat.com>
vishnupatekar [Sat, 28 Nov 2015 17:07:22 +0000 (01:07 +0800)]
sunxi: power: enabled support for axp818
Enabled support for AXP818 in SPL and u-boot.
DCDC1, DCDC2, DCDC3 and DCSC5 are enabled.
Signed-off-by: Vishnu Patekar <vishnupatekar0510@gmail.com> Reviewed-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Hans de Goede <hdegoede@redhat.com>
vishnupatekar [Sat, 28 Nov 2015 17:07:21 +0000 (01:07 +0800)]
sunxi: power: axp818: add support for axp818 driver
AXP818 is rsb based PMIC and used on Allwinner A83T H8 Homlet dev board.
It's registers are different and calculating reg config is different than
that of earlier axp power ICs.
DCDC1, DCDC2, DCDC3 and DCDC5 is implemented at the moment.
all other voltages can be added subsequently.
AXP datasheet is uploaded to wiki:
http://linux-sunxi.org/File:AXP818_datasheet_Revision1.0.pdf
Signed-off-by: Vishnu Patekar <vishnupatekar0510@gmail.com> Reviewed-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Hans de Goede <hdegoede@redhat.com>
vishnupatekar [Sat, 28 Nov 2015 17:07:20 +0000 (01:07 +0800)]
sunxi: Add support for UART0 in PB pin group on A83T
On A83T, PB9,PB10 are UART0 pins.
On allwinner A83T Dev board(h8homlet), this uart0 serial connector
is exposed.
Signed-off-by: Vishnu Patekar <vishnupatekar0510@gmail.com> Reviewed-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Hans de Goede <hdegoede@redhat.com>
vishnupatekar [Sat, 28 Nov 2015 17:07:19 +0000 (01:07 +0800)]
sunxi: Add Machine Support for A83T SOC
Allwinner A83T is octa-core cortex-a7 SOC.
This enables support for A83T.
SMP is not yet supported.
Signed-off-by: Vishnu Patekar <vishnupatekar0510@gmail.com> Reviewed-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans de Goede [Fri, 20 Nov 2015 18:29:49 +0000 (19:29 +0100)]
sunxi: Set AHB1 clock to PLL6/3 on all clock_sun6i.h using SoCs
According to the datasheets the max speed of AHB1 is 276 MHz, so
setting it to PLL6 / 3 which gives us 200MHz everywhere is fine,
and gives us a nice speed-up in certain workloads.
Suggested-by: Chen-Yu Tsai <wens@csie.org> Signed-off-by: Hans de Goede <hdegoede@redhat.com> Acked-by: Ian Campbell <ijc@hellion.org.uk> Tested-by: Chen-Yu Tsai <wens@csie.org>
Bin Meng [Thu, 26 Nov 2015 01:46:08 +0000 (17:46 -0800)]
x86: Remove CPU_INTEL_SOCKET_RPGA989
This Kconfig option name indicates it has something to do with cpu
socket, however it is actually not the case. Remove it and move
options inside it to NORTHBRIDGE_INTEL_IVYBRIDGE.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com> Acked-by: Simon Glass <sjg@chromium.org> Tested-by: Simon Glass <sjg@chromium.org>
Bin Meng [Wed, 9 Dec 2015 01:31:39 +0000 (17:31 -0800)]
x86: Move i8254_init() to x86_cpu_init_f()
Right now i8254_init() is called from timer_init() in the tsc timer
driver. But actually i8254 and tsc are completely different things.
Since tsc timer has been converted to driver model, we should find
a new place that is appropriate for U-Boot to call i8254_init(),
which is now x86_cpu_init_f().
Signed-off-by: Bin Meng <bmeng.cn@gmail.com> Acked-by: Simon Glass <sjg@chromium.org>
Bin Meng [Mon, 7 Dec 2015 13:28:13 +0000 (05:28 -0800)]
x86: Fix PCI UART compatible string for crownbay and galileo
With recent ns16550 driver changes, we only changed the legacy UART
(at I/O port 0x3f8) compatible string, but forgot to change the PCI
UART compatible string. Now fix it.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com> Acked-by: Simon Glass <sjg@chromium.org>
Bin Meng [Mon, 7 Dec 2015 09:39:47 +0000 (01:39 -0800)]
fdt: Change OF_BAD_ADDR to FDT_ADDR_T_NONE
Currently OF_BAD_ADDR is always -1ULL. When using OF_BAD_ADDR as the
return value of dev_get_addr(), it creates potential size mismatch
as dev_get_addr() uses FDT_ADDR_T_NONE as the return value which can
be either -1U or -1ULL depending on CONFIG_PHYS_64BIT. Now we change
OF_BAD_ADDR to FDT_ADDR_T_NONE to avoid such case.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com> Acked-by: Simon Glass <sjg@chromium.org> Reviewed-by: Stefan Roese <sr@denx.de>
Marek Vasut [Mon, 7 Dec 2015 17:01:54 +0000 (18:01 +0100)]
mkimage: Fix warning from fix for generating multi and script images again
Seems 6ae6e160 broke creating images in certain cases, there
are two problems with that patch.
First is that the expression "!x == 4 || !x == 6" is ambiguous. The
intention here was "!(x == 4) || !(x == 6)" based on reading further in
the file, where this was borrowed from. This however is interpreted by
gcc as "(!x) == 4 || (!x) == 6" and always false. gcc-5.x will warn
about this case.
The second problem is that we do not want to test for the case of "(NOT x
is 4) OR (NOT x is 6)" but instead "(x is not equal to 4) AND (x is not
equal to 6)". This is because in those two cases we already execute the
code question in another part of the file. Rewrite the expression and
add parenthesis for clarity.
Signed-off-by: Marek Vasut <marex@denx.de> Cc: Tom Rini <trini@konsulko.com> Cc: Philippe De Swert <philippedeswert@gmail.com> Cc: Simon Glass <sjg@chromium.org>
[trini: Re-word Marek's explanation]
Tom Rini [Mon, 7 Dec 2015 13:23:29 +0000 (08:23 -0500)]
CONFIG_NEEDS_MANUAL_RELOC: Fix warnings when not set
Now that we may compile (but not link) code calling fixup_cmdtable when
this is not set, we need to always have the declaration available. We
should also make sure that anyone calling the function includes
<command.h> as that's where the function declaration is.
Michal Simek [Tue, 1 Dec 2015 07:37:16 +0000 (08:37 +0100)]
dm: core: Enable SPL_SIMPLE_BUS by default
This option is needed for all SoCs which have nodes on bus. Without
enabling this drivers are not found and probed.
Issue was found on Zynq MMC probe.
Enable this option by default.
Signed-off-by: Michal Simek <michal.simek@xilinx.com> Reviewed-by: Stefan Roese <sr@denx.de> Reviewed-by: Simon Glass <sjg@chromium.org>
zynq: sdhci: Calculate minimum frequency based on max frequency
Calculate the minimum sd clock based on max clock. This will
be done by add_sdhci() if we pass minimum clock as zero.
It also does based on SD host contoller version.
Signed-off-by: Siva Durga Prasad Paladugu <sivadur@xilinx.com> Signed-off-by: Michal Simek <michal.simek@xilinx.com>
Michal Simek [Mon, 23 Nov 2015 15:27:38 +0000 (16:27 +0100)]
ARM: zynq: Fix location of stack and malloc areas
The patch
"board_init: Change the logic to setup malloc_base"
(sha1: 9ac4fc82071ce346e3885118242ff45d22f69b82)
breaks SPL for Zynq because it puts early alloc area on the stack which
caused that stack was decreased by CONFIG_SYS_MALLOC_F_LEN (0x400)
and there was not enough space for regular stack.
This patch changes memory layout to better utilize the last 64k OCM
block.
0xffff0000 - 0xfff1000 - Full malloc space
0xffff1000 - 0xffff300 - Stack location
0xfffff300 - CONFIG_SYS_MALLOC_F_LEN - Early malloc space
0xfffffd00 - sizeof(GD) - GD
0xfffffe00 - 0xffffffff - SoC specific boot code
Signed-off-by: Michal Simek <michal.simek@xilinx.com> Tested-by: Moritz Fischer <moritz.fischer@ettus.com>
Michal Simek [Mon, 30 Nov 2015 13:14:56 +0000 (14:14 +0100)]
net: gem: Move driver to DM
- Enable DM_ETH by default for Zynq and ZynqMP
- Remove board_eth_init code
- Change miiphy_read function to return value instead of error code
based on DM requirement
- Do not enable EMIO DT support by default
Signed-off-by: Michal Simek <michal.simek@xilinx.com> Reviewed-by: Jagan Teki <jteki@openedev.com> Reviewed-by: Simon Glass <sjg@chromium.org> Reviewed-by: Bin Meng <bmeng.cn@gmail.com>