Andreas Färber [Mon, 9 Jan 2017 15:08:02 +0000 (16:08 +0100)]
cmd/fdt: Make fdt get value endian-safe for single-cell properties
On a Raspberry Pi 2 disagreements on cell endianness can be observed:
U-Boot> fdt print /soc/gpio@7e200000 phandle
phandle = <0x0000000d>
U-Boot> fdt get value myvar /soc/gpio@7e200000 phandle; printenv myvar
myvar=0x0D000000
Fix this by always treating the pointer as BE and converting it in
fdt_value_setenv(), like its counterpart fdt_parse_prop() already does.
Consistently use fdt32_t, fdt32_to_cpu() and cpu_to_fdt32().
Fixes: bc80295 ("fdt: Add get commands to fdt") Cc: Joe Hershberger <joe.hershberger@ni.com> Cc: Gerald Van Baren <gvb@unssw.com> Signed-off-by: Andreas Färber <afaerber@suse.de> Acked-by: Simon Glass <sjg@chromium.org>
Stefan Agner [Tue, 20 Dec 2016 14:58:45 +0000 (15:58 +0100)]
cmd: fdt: Print error message when fdt application fails
There are lots of reason why a FDT application might fail, the
error code might give an indication. Let the error code translate
in a error string so users can try to understand what went wrong.
Signed-off-by: Stefan Agner <stefan.agner@toradex.com> Acked-by: Maxime Ripard <maxime.ripard@free-electrons.com> Acked-by: Simon Glass <sjg@chromium.org>
David Gibson [Tue, 20 Dec 2016 14:58:44 +0000 (15:58 +0100)]
libfdt: Correct fdt handling of overlays without fixups and base trees without symbols
The fdt_overlay_apply() function purports to support the edge cases where
an overlay has no fixups to be applied, or a base tree which has no
symbols (the latter can only work if the former is also true). However it
gets it wrong in a couple of small ways:
* In the no fixups case, it doesn't fail immediately, but will attempt
fdt_for_each_property_offset() giving -FDT_ERR_NOTFOUND as the node
offset, which will fail. Instead it should succeed immediately, since
there's nothing to do.
* In the case of no symbols, it again doesn't fail immediately. However
if there is an actual fixup it will fail with an unexpected error,
because -FDT_ERR_NOTFOUND is passed to fdt_getprop() when attempting to
look up the symbols. We should instead return -FDT_ERR_NOTFOUND
directly.
Both of these errors lead to the code returning misleading error codes in
failing cases.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Stefan Agner <stefan.agner@toradex.com> Acked-by: Simon Glass <sjg@chromium.org>
Masahiro Yamada [Thu, 12 Jan 2017 03:16:15 +0000 (12:16 +0900)]
mmc: sunxi: revive depends on UART0_PORT_F
Commit f401e907fcbc ("ARM: sunxi: remove bare default for
CONFIG_MMC") dropped "depends on UART0_PORT_F", but it is still
needed. Revive it as a prerequisite of CONFIG_MMC_SUNXI.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com> Acked-by: Maxime Ripard <maxime.ripard@free-electrons.com>
Masahiro Yamada [Fri, 13 Jan 2017 02:51:51 +0000 (11:51 +0900)]
mmc: sdhci: fix NULL pointer access when host->ops is not set
Until recently, sdhci_ops was used only for overriding IO accessors.
(so, host->ops was not set by any drivers except bcm2835_sdhci.c)
Now, we have more optional callbacks, get_cd, set_control_reg, and
set_clock. However, the code
if (host->ops->get_cd)
host->ops->get_cd(host);
... expects host->ops is set for all drivers.
Commit 5e96217f0434 ("mmc: pic32_sdhci: move the code to
pic32_sdhci.c") and commit 62226b68631b ("mmc: sdhci: move the
callback function into sdhci_ops") added sdhci_ops for pic32_sdhci.c
and s5p_sdhci.c, but the other drivers still do not (need not) set
host->ops because all callbacks in sdhci_ops are optional.
host->ops must be checked to avoid the system crash caused by NULL
pointer access.
Fabio Estevam [Thu, 15 Dec 2016 18:00:13 +0000 (16:00 -0200)]
cmd: mem: Use memcpy for 'cp' command
Simplify the 'cp' command implementation by using the memcpy() function,
which brings the additional benefit of performance gain for those who have
CONFIG_USE_ARCH_MEMCPY selected.
Tested on a mx6qsabreauto board where a 5x gain in performance is seen
when reading 10MB from the parallel NOR memory.
Romain Perier [Wed, 11 Jan 2017 10:46:14 +0000 (11:46 +0100)]
Enable DISTRO_DEFAULTS for Rockchip platforms
This enables suitable commands needed for booting general purpose
Linux distribution. This is required for example if we want to use PXE
or DHCP as default boot targets, symbols no longer enabled by
config_distro_defaults.h .
Simon Glass [Wed, 11 Jan 2017 10:46:13 +0000 (11:46 +0100)]
rockchip: evb-rk3339: Enable DHCP
This is the only RK3399 device without DHCP. Enable it so that we
can use a common BOOT_TARGET_DEVICES setting. It is likely useful to be
able to use USB networking, at least. Full networking can be enabled when
a suitable platform needs it.
Sjoerd Simons [Wed, 11 Jan 2017 10:46:11 +0000 (11:46 +0100)]
net: gmac_rockchip: Add Rockchip GMAC driver
Add a new driver for the GMAC ethernet interface present in Rockchip
RK3288 SOCs. This driver subclasses the generic design-ware driver to
add the glue needed specifically for Rockchip.
Signed-off-by: Sjoerd Simons <sjoerd.simons@collabora.co.uk> Signed-off-by: Simon Glass <sjg@chromium.org> Signed-off-by: Romain Perier <romain.perier@collabora.com> Acked-by: Joe Hershberger <joe.hershberger@ni.com>
Simon Glass [Wed, 11 Jan 2017 10:46:09 +0000 (11:46 +0100)]
net: designware: Split the link init into a separate function
With rockchip we need to make adjustments after the link speed is set but
before enabling received/transmit. In preparation for this, split these
two pieces into separate functions.
Signed-off-by: Simon Glass <sjg@chromium.org> Signed-off-by: Romain Perier <romain.perier@collabora.com> Acked-by: Joe Hershberger <joe.hershberger@ni.com>
Isp-camera preview image will be broken when dual screen display mode.
This patch set isp/vop qos level higher to solve this problem.
We have verified this patch on rk3288-miniarm board.
Signed-off-by: Nickey Yang <nickey.yang@rock-chips.com>
Kever Yang [Thu, 24 Nov 2016 07:29:52 +0000 (15:29 +0800)]
rockchip: dts: popmetal: add usb host power supply node
The popmetal board using a HOST_VBUS_DRV gpio signal to control the
USB host port 5V power, add a fix regulator and pinctrl for it, and
enable the USB host1 controller with the vbus-supply.
Signed-off-by: Kever Yang <kever.yang@rock-chips.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Added rockchip: tag: Signed-off-by: Simon Glass <sjg@chromium.org>
Kever Yang [Thu, 24 Nov 2016 07:29:51 +0000 (15:29 +0800)]
rockchip: config: popmetal: enable the USB host controller and function
RK3288 using the dwc2 USB host controller, enable it and other usb host
funtion like storage and ether.
Signed-off-by: Kever Yang <kever.yang@rock-chips.com> Acked-by: Simon Glass <sjg@chromium.org>
Added rockchip: tag: Signed-off-by: Simon Glass <sjg@chromium.org>
Kever Yang [Thu, 24 Nov 2016 07:29:50 +0000 (15:29 +0800)]
rockchip: board: popmetal: de-assert the host rst pin in board init
The PopMetal board have a on board FE1.1 usb 2.0 hub which connect to
the usb host port, we need to de-assert its reset pin to enable it.
Signed-off-by: Kever Yang <kever.yang@rock-chips.com> Acked-by: Simon Glass <sjg@chromium.org>
Added rockchip: tag: Signed-off-by: Simon Glass <sjg@chromium.org>
Masahiro Yamada [Tue, 10 Jan 2017 04:32:07 +0000 (13:32 +0900)]
mmc: move more driver config options to Kconfig
Move (and rename) the following CONFIG options to Kconfig:
CONFIG_DAVINCI_MMC (renamed to CONFIG_MMC_DAVINCI)
CONFIG_OMAP_HSMMC (renamed to CONFIG_MMC_OMAP_HS)
CONFIG_MXC_MMC (renamed to CONFIG_MMC_MXC)
CONFIG_MXS_MMC (renamed to CONFIG_MMC_MXS)
CONFIG_TEGRA_MMC (renamed to CONFIG_MMC_SDHCI_TEGRA)
CONFIG_SUNXI_MMC (renamed to CONFIG_MMC_SUNXI)
They are the same option names as used in Linux.
This commit was created as follows:
[1] Rename the options with the following command:
Masahiro Yamada [Tue, 10 Jan 2017 04:32:05 +0000 (13:32 +0900)]
mmc: move DesignWare-based drivers to Kconfig
Move (and rename) the following CONFIG options to Kconfig:
CONFIG_EXYNOS_DWMMC (renamed to CONFIG_MMC_DW_EXYNOS)
CONFIG_HIKEY_DWMMC (renamed to CONFIG_MMC_DW_K3)
CONFIG_SOCFPGA_DWMMC (renamed to CONFIG_MMC_DW_SOCFPGA)
The "HIKEY" is a board name, so it is not suitable for the MMC
controller name. I am following the name used in Linux.
This commit was generated as follows:
[1] Rename the config options with the following command:
find . -name .git -prune -o ! -path ./scripts/config_whitelist.txt \
-type f -print | xargs sed -i -e '
s/CONFIG_EXYNOS_DWMMC/CONFIG_MMC_DW_EXYNOS/g
s/CONFIG_HIKEY_DWMMC/CONFIG_MMC_DW_K3/g
s/CONFIG_SOCFPGA_DWMMC/CONFIG_MMC_DW_SOCFPGA/g
'
[2] Commit the changes
[3] Create the entries in drivers/mmc/Kconfig
(with default y for EXYNOS and SOCFPGA)
[4] Run the following:
tools/moveconfig.py -y -r HEAD MMC_DW_EXYNOS MMC_DW_K3 MMC_DW_SOCFPGA
[5] Sort and align drivers/mmc/Makefile for readability
[6] Clean-up doc/README.socfpga by hand
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com> Reviewed-by: Marek Vasut <marex@denx.de>
Masahiro Yamada [Tue, 10 Jan 2017 04:32:04 +0000 (13:32 +0900)]
mmc: move CONFIG_DWMMC to Kconfig, renaming to CONFIG_MMC_DW
This commit was created as follows:
[1] Rename the option with the following command:
find . -name .git -prune -o ! -path ./scripts/config_whitelist.txt \
-type f -print | xargs sed -i -e 's/CONFIG_DWMMC/CONFIG_MMC_DW/g'
[2] create the entry for MMC_DW in drivers/mmc/Kconfig
(the prompt and help were copied from Linux)
[3] run "tools/moveconfig.py -y MMC_DW"
[4] add "depends on MMC_DW" to the MMC_DW_ROCKCHIP entry
[5] Clean-up doc/README.socfpga by hand
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com> Reviewed-by: Marek Vasut <marex@denx.de>
Masahiro Yamada [Tue, 10 Jan 2017 04:32:02 +0000 (13:32 +0900)]
ARM: socfpga: remove unused CONFIG option and cleanup README.socfpga
CONFIG_SOCFPGA_DWMMC_FIFO_DEPTH is defined in the socfpga_common.h,
but not referenced at all. Remove.
Also, clean-up the README.socfpga. CONFIG_MMC should not be defined
in the header since it was moved to Kconfig by commit c27269953b94
("mmc: complete unfinished move of CONFIG_MMC"). I see no grep hit
for the others.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com> Reviewed-by: Marek Vasut <marex@denx.de>
Masahiro Yamada [Fri, 30 Dec 2016 14:20:14 +0000 (23:20 +0900)]
mmc: uniphier-sd: fix Kconfig dependency
Some MMC drivers describe operations with the DM_MMC_OPS form, but
there are still several drivers with older implementation. We can
not compile drivers from different groups at the same time because
the core framework is shared with #ifdef CONFIG_DM_MMC_OPS.
Every driver should have "depends on DM_MMC_OPS" (or !DM_MMC_OPS)
explicitly to express which framework it is based on. This will
avoid enabling drivers with incompatible interface at the same time.
It is incorrect to make a driver "select DM_MMC_OPS".
While we are here, add "depends on OF_CONTROL" as well because this
driver can be configured only by Device Tree.
Jaehoon Chung [Fri, 30 Dec 2016 06:30:11 +0000 (15:30 +0900)]
mmc: sdhci: disable the 8bit mode when host doesn't support it
Buswidth is depeneded on Hardware schematic.
Evne though host can support the 8bit buswidth, if hardware doesn't
support 8bit mode, it doesn't work fine.
So the buswidth mode selection leaves a matter in each SoC drivers.
On the contrary to this, hardware supports 8bit mode, but host doesn't
support it. then controller has to disable the MMC_MODE_8BIT.
(Host can check whether 8bit mode is supported or not, since V3.0)
Kamensky Ivan [Tue, 27 Dec 2016 16:12:23 +0000 (19:12 +0300)]
xilinx_phy: Pass correct pointer to fdtdec_get_int()
This patch fixes incorrect pointer on offset device in device tree blob.
When using with the component "Ethernet 1G/2.5G BASE-X PCS/PMA or SGMII"
it does not understand what type is XAE_PHY_TYPE_1000BASE_X and trying
to change frequency.
Signed-off-by: Kamensky Ivan <kamensky.ivan@mail.ru> Signed-off-by: Michal Simek <michal.simek@xilinx.com> Acked-by: Joe Hershberger <joe.hershberger@ni.com>
Michal Simek [Mon, 9 Jan 2017 09:05:16 +0000 (10:05 +0100)]
ARM64: zynqmp: Generate handoff structure for ATF
Xilinx ATF extending options for passing images from BL2(FSBL)
to BL31. U-Boot SPL is FSBL replacement that's why it should generate
handoff structure the same. Support only one entry which is U-Boot in
EL2 itself. When FIT image is adopted structure generate should be data
driven.
Currently ATF is placing this structure at the beggining of OCM which is
rewriting early parts of ATF which should be unused at that time.
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
Moritz Fischer [Thu, 22 Dec 2016 17:36:11 +0000 (09:36 -0800)]
ARM64: zynqmp: Fix i2c node's compatible string
The Zynq Ultrascale MP uses version 1.4 of the Cadence IP core
which fixes some silicon bugs that needed software workarounds
in Version 1.0 that was used on Zynq systems.
Signed-off-by: Moritz Fischer <moritz.fischer@ettus.com> Cc: Michal Simek <michal.simek@xilinx.com> Cc: Heiko Schocher <hs@denx.de> Signed-off-by: Michal Simek <michal.simek@xilinx.com>
ARM64: zynqmp: Enable fastboot for first SD/MMC/EMMC device
DNL numbers are not changed that's why fastboot needs to be called with
-i parameter (Xilinx vendor id).
- Show available devices
sudo fastboot -i 0x03fd devices
xilinx_zynqmp_zcu100 fastboot
- Stop fastboot and go back to U-Boot prompt
sudo fastboot -i 0x03fd continue
- Reboot the board
sudo fastboot -i 0x03fd reboot
- Get internal variables
sudo fastboot -i 0x3fd getvar bootloader-version
bootloader-version: U-Boot 2016.07-00026-g19bd53044817
sudo fastboot -i 0x3fd getvar downloadsize
downloadsize: 0x06000000
sudo fastboot -i 0x3fd getvar version
version: 0.4
(regular variables needs to have fastboot. prefix - there is also
serialno variable which should be define as serial#)
- Format SD/MMC/EMMC card
sudo fastboot -i 0x3fd oem format
- Write images to boot and Linux partition
sudo fastboot -i 0x3fd flash boot sd.img
sudo fastboot -i 0x3fd flash Linux os.img
Added the idle-states node to describe zynqmp idle states. Only cpu-sleep-0
idle state is added in this patch. References to the idle-states node are
added in all CPU nodes. Time values: entry/exit latencies and min-residency,
needs to be tuned. arm,psci-suspend-param is selected to comply with PSCIv1.0
and Extended StateID format.
Signed-off-by: Stefan Krsmanovic <stefan.krsmanovic@aggios.com> Acked-by: Will Wong <willw@xilinx.com> Signed-off-by: Michal Simek <michal.simek@xilinx.com>
Sudeep Holla [Wed, 21 Oct 2015 10:10:16 +0000 (11:10 +0100)]
ARM: dts: zynq: replace gpio-key,wakeup with wakeup-source property
Though the keyboard driver for GPIO buttons(gpio-keys) will continue to
check for/support the legacy "gpio-key,wakeup" boolean property to
enable gpio buttons as wakeup source, "wakeup-source" is the new
standard binding.
This patch replaces the legacy "gpio-key,wakeup" with the unified
"wakeup-source" property in order to avoid any futher copy-paste
duplication.
Cc: Michal Simek <michal.simek@xilinx.com> Cc: "Sören Brinkmann" <soren.brinkmann@xilinx.com> Signed-off-by: Sudeep Holla <sudeep.holla@arm.com> Signed-off-by: Michal Simek <michal.simek@xilinx.com>
Add support to save environment as a file of FAT filesystem
on to SD card. The file will be saved with name uEnv.txt.
This environment will be retrieved during boot.
Signed-off-by: Siva Durga Prasad Paladugu <sivadur@xilinx.com> Signed-off-by: Michal Simek <michal.simek@xilinx.com>
Michal Simek [Mon, 2 Jan 2017 08:40:09 +0000 (09:40 +0100)]
scsi: dm: Unbind all scsi based block devices before new scan
New scan should unbind all block devices not to be listed again.
Without this patch if scsi reset or scan is called new block devices are
created which point to the same id and lun.
For example:
ZynqMP> scsi scan
scsi_scan: if_type=2, devnum=0: sdhci@ff170000.blk, 6, 0
scsi_scan: if_type=2, devnum=0: ahci@fd0c0000.id1lun0, 2, 0
scsi_scan: if_type=2, devnum=0: ahci@fd0c0000.id1lun0, 2, 1
scsi_scan: if_type=2, devnum=0: ahci@fd0c0000.id1lun0, 2, 2
scsi_scan: if_type=2, devnum=0: ahci@fd0c0000.id1lun0, 2, 3
scsi_scan: if_type=2, devnum=0: ahci@fd0c0000.id1lun0, 2, 4
scanning bus for devices...
Device 0: (1:0) Vendor: ATA Prod.: KINGSTON SVP200S Rev: 501A
Type: Hard Disk
Capacity: 57241.8 MB = 55.9 GB (117231408 x 512)
Reported-by: Ken Ma <make@marvell.com> Signed-off-by: Michal Simek <michal.simek@xilinx.com> Acked-by: Simon Glass <sjg@chromium.org>
Jagan Teki [Thu, 5 Jan 2017 14:32:50 +0000 (15:32 +0100)]
mx6ullevk: Add missing MAINTAINERS for mx6ull_14x14_evk_plugin_defconfig
Add 'Peng Fan' as MAINTAINERS of configs/mx6ull_14x14_evk_plugin_defconfig
which is missing in below commit
"imx: mx6ull_14x14_evk: add plugin defconfig"
(sha1: b90ebf49bb8f74afe68f696f59a0e24cc79f2031)
Cc: Stefano Babic <sbabic@denx.de> Reviewed-by: Peng Fan <peng.fan@nxp.com> Signed-off-by: Jagan Teki <jagan@openedev.com>
Andrew F. Davis [Fri, 6 Jan 2017 22:32:12 +0000 (16:32 -0600)]
am335x: configs: Use ISW_ENTRY_ADDR to set SPL_TEXT_BASE
The SPL load address changes based on boot type in HS devices,
ISW_ENTRY_ADDR is used to set this address for AM43xx based SoCs
for similar reasons. Add this same logic for AM33xx devices.
Also make the default value for ISW_ENTRY_ADDR correct for GP
devices based on SoC, HS devices already pick the correct
value in their defconfig.
Andrew F. Davis [Fri, 6 Jan 2017 22:20:02 +0000 (16:20 -0600)]
arm: mach-omap2: Fix secure file generation
When TI_SECURE_DEV_PKG is not defined we warn that the file '*_HS' was
not generated but generate an unsigned one anyway, first fix this
warning to say that it was generated but not secured.
When the user then exports TI_SECURE_DEV_PKG after getting this warning,
and tries to re-build, 'make' will detect the build artifacts as
unchanged and so assume they do not need to be re-generated. This causes
it to fail to sign the files and it will pack unsigned files into the
final image, even though TI_SECURE_DEV_PKG is now correctly defined and
working.
Fix this by using FORCE on the targets causes them to be re-run even if
the dependent files have not changed.
This then causes another issue. We currently rename the signed dtb files
to overwrite the non-signed ones. We do this so the 'mkimage' tool gives
the packaged dtb sections the correct name. If we do not rename the files
then SPL will not find them during boot.
Fix this by renaming the dtb files by appending _HS to the end of the
filename, after the ".dtb", this causes them to still be named correctly
in the FIT blob.
projects / u-boot.git / log