Commit graph

39442 commits

Author SHA1 Message Date
Fabio Estevam
ca4f338e2e warp7: Use PARTUUID to specify the rootfs location
warp7 can run different kernel versions, such as NXP 4.1 or mainline.

Currently the rootfs location is passed via mmcblk number and the
problem with this approach is that the mmcblk number for the eMMC
changes depending on the kernel version.

In order to avoid such issue, use UUID method to specify the rootfs
location.

Succesfully tested booting a NXP 4.1 and also a mainline kernel.

Signed-off-by: Fabio Estevam <fabio.estevam@nxp.com>
2016-09-06 18:22:48 +02:00
Fabio Estevam
375d19c911 warp7: Add a secure mode target
NXP kernel expects to boot in secure mode, so introduce
warp7_secure_defconfig target which selects CONFIG_ARMV7_BOOT_SEC_DEFAULT.

Signed-off-by: Fabio Estevam <fabio.estevam@nxp.com>
2016-09-06 18:22:48 +02:00
Fabio Estevam
ab25f0f69f mx6ul_14x14_ev: Enable the CCGR clocks earlier
To be in the safe side we need to enable the CCGR clocks prior
to calling arch_cpu_init().

Inspired by Tim Harvey's commit d783c2744f ("imx: ventana: fix boot to SD").

Signed-off-by: Fabio Estevam <fabio.estevam@nxp.com>
Reviewed-by: Eric Nelson <eric@nelint.com>
Tested-by: Eric Nelson <eric@nelint.com>
2016-09-06 18:22:48 +02:00
Fabio Estevam
b343417e29 mx6ul_14x14_evk: Adjust SPL DDR3 settings
Adjust DDR3 initialization done in SPL by comparing them against
the NXP DCD table.

Signed-off-by: Fabio Estevam <fabio.estevam@nxp.com>
Reviewed-by: Eric Nelson <eric@nelint.com>
2016-09-06 18:22:48 +02:00
Fabio Estevam
7dbda25ecd mx6ul_14x14_evk: Pass refsel and refr fields to avoid hang
When running a NXP 4.1 kernel with U-Boot mainline on a mx6ul-evk,
we observe a hang when going into the lowest operational point of cpufreq.

This hang issue does not happen on the NXP U-Boot version.

After comparing the SPL DDR initialization against the DCD table
from NXP U-Boot, the key difference that causes the hang is the
MDREF register setting:

DATA 4 0x021B0020 0x00000800

,which means:

REF_SEL = 0 --> Periodic refresh cycle: 64kHz
REFR = 1 ---> Refresh Rate - 2 refreshes

So adjust the MDREF initialization for mx6ul_evk accordingly
to fix the kernel hang issue at low bus frequency.

Reported-by: Eric Nelson <eric@nelint.com>
Signed-off-by: Fabio Estevam <fabio.estevam@nxp.com>
Reviewed-by: Eric Nelson <eric@nelint.com>
2016-09-06 18:22:48 +02:00
Fabio Estevam
edf0093732 mx6: ddr: Allow changing REFSEL and REFR fields
Currently MX6 SPL DDR initialization hardcodes the REF_SEL and
REFR fields of the MDREF register as 1 and 7, respectively for
DDR3 and 0 and 3 for LPDDR2.

Looking at the MDREF initialization done via DCD we see that
boards do need to initialize these fields differently:

$ git grep 0x021b0020 board/
board/bachmann/ot1200/mx6q_4x_mt41j128.cfg:DATA 4 0x021b0020 0x00005800
board/ccv/xpress/imximage.cfg:DATA 4 0x021b0020 0x00000800 /* MMDC0_MDREF */
board/freescale/mx6qarm2/imximage.cfg:DATA 4 0x021b0020 0x7800
board/freescale/mx6qarm2/imximage.cfg:DATA 4 0x021b0020 0x00005800
board/freescale/mx6qarm2/imximage_mx6dl.cfg:DATA 4 0x021b0020 0x00005800
board/freescale/mx6qarm2/imximage_mx6dl.cfg:DATA 4 0x021b0020 0x00005800
board/freescale/mx6qsabreauto/imximage.cfg:DATA 4 0x021b0020 0x00005800
board/freescale/mx6qsabreauto/mx6dl.cfg:DATA 4 0x021b0020 0x00005800
board/freescale/mx6qsabreauto/mx6qp.cfg:DATA 4 0x021b0020 0x00005800
board/freescale/mx6sabresd/mx6dlsabresd.cfg:DATA 4      0x021b0020 0x00005800
board/freescale/mx6sabresd/mx6q_4x_mt41j128.cfg:DATA 4 0x021b0020 0x00005800
board/freescale/mx6slevk/imximage.cfg:DATA 4 0x021b0020 0x00001800
board/freescale/mx6sxsabreauto/imximage.cfg:DATA 4 0x021b0020 0x00000800
board/freescale/mx6sxsabresd/imximage.cfg:DATA 4 0x021b0020 0x00000800
board/warp/imximage.cfg:DATA 4 0x021b0020 0x00001800

So introduce a mechanism for users to be able to configure
REFSEL and REFR fields as needed.

Keep all the mx6 SPL users in their current REF_SEL and REFR values,
so no functional changes for the existing users.

Signed-off-by: Fabio Estevam <fabio.estevam@nxp.com>
Reviewed-by: Eric Nelson <eric@nelint.com>
2016-09-06 18:22:48 +02:00
Fabio Estevam
946db0cbd0 mx7dsabresd: Directly write to register LDOGCTL
Register LDOGCTL contains only bit 0 as a valid bit, so there is no need
to do a read-modify-write operation.

Simplify the code by writing directly to this register.

Signed-off-by: Fabio Estevam <fabio.estevam@nxp.com>
2016-09-06 18:22:48 +02:00
Fabio Estevam
78eed0a6d5 mx7dsabresd: Directly write to register LDOGCTL
Register LDOGCTL contains only bit 0 as a valid bit, so there is no need
to do a read-modify-write operation.

Simplify the code by writing directly to this register.

Signed-off-by: Fabio Estevam <fabio.estevam@nxp.com>
2016-09-06 18:22:48 +02:00
Fabio Estevam
938076efa9 pico-imx6ul: Directly write to register LDOGCTL
Register LDOGCTL contains only bit 0 as a valid bit, so there is no need
to do a read-modify-write operation.

Simplify the code by writing directly to this register.

Signed-off-by: Fabio Estevam <fabio.estevam@nxp.com>
2016-09-06 18:22:48 +02:00
Christopher Spinrath
f8de60bd58 ARM: board: cm_fx6: fix mtd partition fixup
ft_board_setup may return early in the case that the board revision
cannot be obtained. In that case it is assumed that no revision
specific correction in the fdt is neccessary. But the mtd partitions
will not be fixed up either altough they are not revision specific.

Move the call to fdt_fixup_mtdparts in front of the revision specific
part to ensure that the partitions are fixed up even if the board
revision cannot be obtained.

While on it, fix a spelling mistake in a comment introduced by the
same commit.

Fixes: 62d6bac660 ("ARM: board: cm_fx6: fixup mtd partitions in the fdt")
Signed-off-by: Christopher Spinrath <christopher.spinrath@rwth-aachen.de>
Reviewed-by: Stefano Babic <sbabic@denx.de>
Reviewed-by: Nikita Kiryanov <nikita@compulab.co.il>
2016-09-06 18:22:48 +02:00
Eric Nelson
eb3813ad1a mx6ul_14x14_evk: don't use array for SD2 card detect pad
Only a single pad is changed to change sdhc2_dat3 from an
SDIO pin to and from GPIO4:5, so remove the array and use
the imx_iomux_v3_setup_pad() routine.

Signed-off-by: Eric Nelson <eric@nelint.com>
Reviewed-by: Fabio Estevam <fabio.estevam@nxp.com>
2016-09-06 18:22:48 +02:00
Breno Lima
ed39522680 warp7: Modify fdt_file environment variable
Use imx7s-warp.dts as fdt_file because this is the name that upstream
kernel will deploy.

Signed-off-by: Breno Lima <breno.lima@nxp.com>
Acked-by: Fabio Estevam <fabio.estevam@nxp.com>
2016-09-06 18:22:48 +02:00
Fabio Estevam
693779e371 warp: Fix RAM size runtime detection
Since commit a13d3757f7 ("warp: Use imx_ddr_size() for calculating the
DDR size") warp board no longer boots.

The reason for the breakage is that the warp board is using the DDR
configuration from mx6slevk. A fundamental difference between warp and
mx6slevk is that warp only uses one DDR chip select while mx6slevk uses two.

The imx_ddr() function calculates the RAM size in runtime by reading the
values of registers MDCTL and MDMISC.

So in order to fix this warp boot issue, create a imximage DDR file specific
to warp, where the MDCTL register is configured to only activates a single
chip select.

Reported-by: Breno Lima <breno.lima@nxp.com>
Signed-off-by: Fabio Estevam <fabio.estevam@nxp.com>
Tested-by: Breno Lima <breno.lima@nxp.com>
2016-09-06 18:22:48 +02:00
Vanessa Maegima
7d301a594d warp7: Add PMIC support
Add PMIC support. Tested by command "pmic PFUZE3000 dump".

Signed-off-by: Vanessa Maegima <vanessa.maegima@nxp.com>
Reviewed-by: Fabio Estevam <fabio.estevam@nxp.com>
Reviewed-by: Stefano Babic <sbabic@denx.de>
2016-09-06 18:22:48 +02:00
Akshay Bhat
ff3832205e arm: imx: Add support for Advantech DMS-BA16 board
Add support for Advantech DMS-BA16 board. The board is based on Advantech
BA16 module which has a i.MX6D processor. The board supports:
 - FEC Ethernet
 - USB Ports
 - SDHC and MMC boot
 - SPI NOR
 - LVDS and HDMI display

Basic information about the module:
 - Module manufacturer: Advantech
 - CPU: Freescale ARM Cortex-A9 i.MX6D
 - SPECS:
     Up to 2GB Onboard DDR3 Memory;
     Up to 16GB Onboard eMMC NAND Flash
     Supports OpenGL ES 2.0 and OpenVG 1.1
     HDMI, 24-bit LVDS
     1x UART, 2x I2C, 8x GPIO,
     4x Host USB 2.0 port, 1x USB OTG port,
     1x micro SD (SDHC),1x SDIO, 1x SATA II,
     1x 10/100/1000 Mbps Ethernet, 1x PCIe X1 Gen2

Signed-off-by: Akshay Bhat <akshay.bhat@timesys.com>
Cc: u-boot@lists.denx.de
Cc: sbabic@denx.de
2016-09-06 18:22:48 +02:00
Fabio Estevam
76b21efd55 mx7dsabresd: Print secure/non-secure mode info
mx7dsabresd has two targets:

- mx7dsabresd_defconfig: boots in non-secure mode
- mx7dsabresd_secure_defconfig: boots in secure mode

Print the mode that is being used to help users to easily identify
which target is running on the board.

Signed-off-by: Fabio Estevam <fabio.estevam@nxp.com>
2016-09-06 18:22:48 +02:00
Stefan Agner
2a83c95fdb mtd: nand: mxs: fix cache alignment for cache lines >32
Currently the command buffer gets allocated with a size of 32 bytes.
This causes warning messages on systems with cache lines bigger than
32 bytes:
CACHE: Misaligned operation at range [9df17a00, 9df17a20]

Define command buffer to be at least 32 bytes, but more if cache
line is bigger.

Signed-off-by: Stefan Agner <stefan.agner@toradex.com>
Reviewed-by: Stefano Babic <sbabic@denx.de>
2016-09-06 18:22:48 +02:00
Tim Harvey
0a22c7f0dc imx: ventana: enable splashscreen support
Signed-off-by: Tim Harvey <tharvey@gateworks.com>
2016-09-06 18:22:48 +02:00
Soeren Moch
8ce747fcff board: tbs2910: fix HDMI pre-console buffer
HDMI output must be enabled very early to also enable the pre-console buffer

Signed-off-by: Soeren Moch <smoch@web.de>
Reviewed-by: Stefano Babic <sbabic@denx.de>
2016-09-06 18:22:48 +02:00
Soeren Moch
8741a374f5 board: tbs2910: always enable usbkbd
'usb start' is much faster now, so always enable usb keyboard

Signed-off-by: Soeren Moch <smoch@web.de>
Reviewed-by: Stefano Babic <sbabic@denx.de>
2016-09-06 18:22:48 +02:00
Tom Rini
c0afcb5889 Merge branch 'master' of http://git.denx.de/u-boot-sunxi 2016-09-06 11:28:42 -04:00
Tom Rini
57288e3d95 Merge git://git.denx.de/u-boot-nand-flash 2016-09-06 11:28:37 -04:00
Mugunthan V N
0068dd687d ARM: dts: dra72-evm: fix broken ethernet
With commit ceec08f50b, phy is connected to slave 0, but
changing the phy node was missed, fix it by populating the
phy node to proper cpsw slave node.

Fixes: ceec08f50b ("ARM: dts: dra72-evm: Add mode-gpios entry for mac node")
Signed-off-by: Mugunthan V N <mugunthanvnm@ti.com>
Cc: Vignesh R <vigneshr@ti.com>
Tested-by: Tom Rini <trini@konsulko.com>
2016-09-06 11:28:27 -04:00
Andre Przywara
5a74a39129 sunxi: fix 64-bit compiler warning for SPL header parsing
Casting "int"s to pointers is only valid for 32-bit systems.
Add the appropriate pointer type cast to avoid a compiler warning
when compiling for AArch64.

Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Siarhei Siamashka <siarhei.siamashka@gmail.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
2016-09-06 13:35:52 +02:00
Andre Przywara
fa855d3d55 sunxi: Kconfig: rename non-existent SUN50I_A64 config symbol
There is no "CONFIG_MACH_SUN50I_A64" in upstream U-Boot, so fix
the name to prevent the option to be enabled.

Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
2016-09-06 13:35:52 +02:00
Andre Przywara
eb504fa13f Revert "sunxi: Move the SPL stack top to 0x1A000 on Allwinner A64/A80"
This commit moved the SPL stack into SRAM C, which worked when the SPL
set the AHB1 clock down to 100 MHz to cope with the flaky SRAM C access
from the CPU.
However booting with boot0 (and thus not using SPL at all) we still run
with a 200 MHz AHB1, so any access to SRAM C is prone to fail.
Since this commit does _not_ only affect the SPL code, but also the
U-Boot proper, we fail when booting with boot0.

As the introduction of tiny-printf reduced the size of the SPL, we
can afford to have the SPL stack in SRAM A1.

This reverts commit 1a83fb4a17
and fixes booting the Pine64 when using boot0.

Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Siarhei Siamashka <siarhei.siamashka@gmail.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
2016-09-06 13:35:52 +02:00
Hans de Goede
de300ea5db sunxi: Add defconfig and dts file for the Orange Pi Plus2E SBC
The Orange Pi Plus2E is an extended version of the Orange Pi Pc Plus,
with 2G RAM and an external gbit ethernet phy.

The dts file is identical to the one submitted to the upstream kernel,
except that it has the pending patch to enable the ethernet controller
squashed in, as u-boot already has sun8i-emac support.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
2016-09-03 13:05:43 +02:00
Hans de Goede
1c145c39dc sunxi: Enable emac on H3 orangepi boards
The Orange Pi 2 and Orange Pi Plus also come with ethernet, enable
support for this.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
2016-09-03 10:57:00 +02:00
Hans de Goede
019731a88f sunxi: Sync h3-orangepi dts files with kernel
This adds an emac node to the orangepi-2 dts (not yet merged upstream,
but in u-boot we already have emac support); fixes the alphetically
sorting of nodes in sun8i-h3-orangepi-plus.dts and disables some
usb controllers in sun8i-h3-orangepi-plus.dts which are only used
on the plus2e, as upstream has decided to do a separate dts files
for the plus2e.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
2016-09-03 10:57:00 +02:00
Chen-Yu Tsai
68871efe1d sunxi: Fix H3 EMAC syscon register address
The sun8i-emac driver follows an old version of the proposed DT
bindings, where the EMAC clock and EPHY control register range is
listed directly, rather than through a syscon phandle.

Add back the syscon register range to avoid an invalid data access.
We should fix the driver once the Linux kernel bindings have been
finalized.

Signed-off-by: Chen-Yu Tsai <wens@csie.org>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
2016-09-03 10:33:54 +02:00
Stefan Mavrodiev
ca5c37026b sunxi: Add support for A33-OLinuXino board
A33-OLinuXino is A33 development board designed by Olimex LTD.

It has AXP223 PMU, 1GB DRAM, a micro SD card, one USB-OTG connector,
headphone and mic jacks, connector for LiPo battery and optional
4GB NAND Flash.

It has two 40-pin headers. One for LCD panel, and one for
additional modules. Also there is CSI/DSI connector.

The dts files are identical to the ones submitted to the upstream kernel.

Signed-off-by: Stefan Mavrodiev <stefan.mavrodiev@gmail.com>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
2016-09-03 10:33:44 +02:00
Icenowy Zheng
6d973ad9d2 sunxi: Add iNet D978 rev2 defconfig
The iNet D978 rev2 is a tablet board designed by iNet, which is intended to
use on 10" tablets with a appearance like Apple iPad. It has A33 SoC, 1GB
RAM, 8GB/16GB NAND, SDIO Wi-Fi, a MicroUSB port and a MicroSD slot.

Signed-off-by: Icenowy Zheng <icenowy@aosc.xyz>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
2016-09-03 10:04:15 +02:00
Icenowy Zheng
8e71a7ebdc sunxi: add proper device tree for iNet D978 rev2 boards
Add a proper dts for the iNet D978 rev2 based A33 tablets.

Signed-off-by: Icenowy Zheng <icenowy@aosc.xyz>
Acked-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
2016-09-03 10:04:15 +02:00
Scott Wood
8b7d51249e nand: Fix some more NULL name tests
Now that nand_info[] is an array of pointers we need to test the
pointer itself rather than using name as a proxy for NULLness.

Fixes: b616d9b0a7 ("nand: Embed mtd_info in struct nand_chip")
Signed-off-by: Scott Wood <oss@buserror.net>
Cc: Lukasz Majewski <l.majewski@samsung.com>
Cc: Tony Lindgren <tony@atomide.com>
Acked-by: Tony Lindgren <tony@atomide.com>
2016-09-01 20:08:48 -05:00
Tony Lindgren
4004a81828 nand: Fix nand info for no device
Looks like we have few more places where we're testing for
nand_info[i]->name. We can now use just test for nand_info[i]
instead.

This fixes a data abort on devices with no NAND when doing
nand info.

Signed-off-by: Tony Lindgren <tony@atomide.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
2016-09-01 17:30:25 -05:00
Chris Packham
91395b5d4e mtd: nand: pxa3xx: use nand_set_controller_data
In commit 17cb4b8f32 ("mtd: nand: Add+use mtd_to/from_nand and
nand_get/set_controller_data") the assignment of mtd->priv was removed
but was not replaced. This adds the required nand_set_controller_data()
call.

Signed-off-by: Chris Packham <judge.packham@gmail.com>
2016-09-01 17:30:11 -05:00
Peter Chubb
b615267633 ARM: tegra: Add support for TK1-SOM board from Colorado Engineering
The Colorado TK1 SOM is a small form factor board similar to the
Jetson TK1.  The main differences lie in the pinmux, and in that the
PCIe controller is set to use in 4lanes+1lane, rather than 2+2.

The pinmux header here was generated from a spreadsheet provided by
Colorado Engineering using the tegra-pinmux scripts.  The spreadsheet
was converted from v09 to v11 by me.

Signed-off-by: Peter Chubb <peter.chubb@data61.csiro.au>
Acked-by: Stephen Warren <swarren@nvidia.com>
Signed-off-by: Tom Warren <twarren@nvidia.com>
2016-09-01 09:24:30 -07:00
Stephen Warren
7932d3e4a7 ARM: tegra: use numeric versioning for p2771-0000
The board ID EEPROM and board ID stickers on p2771-0000 will use a numeric
versioning scheme, with version numbers such as 000/100/200/300/400/500.
Within NVIDIA, these versions are also known as A00/A01/A02/A03/A04/B00.
However, that numbering scheme is not easily visible outside of NVIDIA,
and so does not make much sense to use. Convert U-Boot to use the readily
visible numeric scheme.

Also, it turns out that the current A02 DT actually applies to board
versions 000/100/200 (A00..A02). Consequently rename this to 000 not 200
so that all U-Boot builds are named after the first version of the HW they
support.

Signed-off-by: Stephen Warren <swarren@nvidia.com>
Signed-off-by: Tom Warren <twarren@nvidia.com>
2016-08-30 11:14:53 -07:00
Bin Meng
cb1cbdd969 x86: qemu: efi: Add two boards for EFI 32-bit and 64-bit payload
This introduces two board defconfig files for generating EFI 32-bit
and 64-bit payloads, to run on QEMU x86 target.

With these in place, hopefully buildman will catch any build error
with EFI payload support on x86.

Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
2016-08-30 09:26:05 +08:00
Bin Meng
3e6cc35f4e x86: efi: Fix EFI 64-bit payload build warnings
There are lots of warnings when building EFI 64-bit payload.

include/asm-generic/bitops/__fls.h:17:2:
  warning: left shift count >= width of type
  	if (!(word & (~0ul << 32))) {
			^

In fact, U-Boot itself as EFI payload is running in 32-bit mode.
So BITS_PER_LONG needs to still be 32, but EFI status codes are
64-bit when booting from 64-bit EFI. Introduce EFI_BITS_PER_LONG
to bridge those status codes with U-Boot's BITS_PER_LONG.

Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
2016-08-30 09:26:05 +08:00
Bin Meng
3dc51ab0e1 x86: efi: payload: Make EFI payload build again
Since commit 73c5c39 "Makefile: Drop unnecessary -dtb suffixes",
EFI payload does not build anymore. This fixes the build.

Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
2016-08-30 09:26:05 +08:00
Simon Glass
4cc00f0611 x86: Add debugging when cpu_common_init() fails
Add a debug() at this point to help figure out what is wrong.

Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
2016-08-30 09:26:05 +08:00
Simon Glass
e6294e0579 x86: ivybridge: Allow microcode to be collated
Generally the microcode is combined into a single block only (and removed
from the device tree) when there are multiple blocks. But this is not a
requirement.

Adjust the ivybridge code to avoid assuming this.

Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
2016-08-30 09:26:05 +08:00
Simon Glass
fda4fa8195 x86: Add debugging when a microcode update fails
Add a debug() at this point to help figure out what is wrong.

Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Heiko Schocher<hs@denx.de>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
2016-08-30 09:26:05 +08:00
Tom Rini
ff62bdfbd5 Merge branch 'master' of git://git.denx.de/u-boot-uniphier 2016-08-28 10:36:20 -04:00
Masahiro Yamada
8d11f80413 ARM: uniphier: enable CONFIG_CMD_CACHE
This will be useful, for example, to load firmware to DRAM and make
it visible to other agents.

Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
2016-08-28 13:12:18 +09:00
Masahiro Yamada
85dc2fe119 ARM: uniphier: change UNIPHIER_SERIAL to default y option
This is very likely to be necessary for normal use cases.
Set its default to 'y' for shorter defconfig files.

Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
2016-08-28 13:11:35 +09:00
Masahiro Yamada
f0633533d5 ARM: dts: uniphier: add u-boot, dm-pre-reloc to use eMMC boot on sLD3
The eMMC on sLD3 is assigned with dedicated pins (only multiplexed
with GPIO), so it shouldn't hurt to enable eMMC on SPL all the time.

Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
2016-08-28 13:11:34 +09:00
Masahiro Yamada
e8811fc06c ARM: uniphier: increase CONFIG_SYS_MALLOC_F_LEN for sLD3
Commit 76c52ce29f ("ARM: uniphier: increase CONFIG_SYS_MALLOC_F_LEN
to bind all nodes") missed to increase this config for sLD3.

This change is needed to add "u-boot,dm-pre-reloc" to some nodes;
more devices are bound, more malloc memory is needed.

Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
2016-08-28 13:11:34 +09:00
Masahiro Yamada
499c8679be ARM: uniphier: display revision of Micro Support Card 3.6.x kindly
The revision of the original support card (rev 3.5, rev 3.6) fits in
the 8 bit width revision register.  When it was extended in a weird
way, it was versioned in the format of "3.6.x" (where it should have
been "3.7", of course).  What is worse, only the sub-level version
"6.x" was recorded in the 8 bit width register, completely ignoring
the compatibility of the revision register format.

This patch saves madly-versioned support cards by assuming the major
version "3" when the MSB 4 bit of the register is read as "6".  With
this, the support card revision that were displayed as "6.10" is now
corrected to "3.6.10".

Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
2016-08-28 13:11:31 +09:00