Bin Meng [Sun, 9 Oct 2016 11:14:17 +0000 (04:14 -0700)]
dm: video: Don't do anything in alloc_fb() when plat->size is zero
With DM VESA driver on x86 boards, plat->base/size/align are all
zeroes and starting address passed to alloc_fb() happens to be 1MB
aligned, so this routine does not trigger any issue. On QEMU with
U-Boot as coreboot payload, the starting address is within 1MB
range (eg: 0x7fb0000), thus causes failure in video_post_bind().
Actually if plat->size is zero, it makes no sense to do anything
in this routine. Add such check there.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Bin Meng [Sun, 9 Oct 2016 11:14:13 +0000 (04:14 -0700)]
x86: Convert to use DM VESA video driver
At present only chromebook boards are converted to DM video. Other
x86 boards are still using the legacy cfb_console driver. This
switches to use DM version drivers.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Bin Meng [Sun, 9 Oct 2016 11:14:10 +0000 (04:14 -0700)]
x86: doc: Document coreboot framebuffer driver issue on QEMU
For some unknown reason, coreboot framebuffer driver never works on
QEMU since day 1. It seems the driver only works on real hardware.
Document this issue.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Stefan Brüns [Tue, 4 Oct 2016 19:46:35 +0000 (21:46 +0200)]
sandbox/fs: Set correct filetype for unknown filetype
The "hostfs ls" command prefixes each directory entry with either DIR,
LNK or " " if it is a directory, symlink resp. regular file, or
"???" for any other or unknown type.
The latter only works if the type is set correctly, as the entry defaults
to OS_FILET_REG and e.g. socket files show up as regular files.
Signed-off-by: Stefan Brüns <stefan.bruens@rwth-aachen.de> Acked-by: Simon Glass <sjg@chromium.org>
Stefan Brüns [Sat, 1 Oct 2016 18:41:40 +0000 (20:41 +0200)]
sandbox/fs: Use correct size path name buffer
The readdir linux manpage explicitly states (quoting POSIX.1) that
sizeof(d_name) is not correct for determining the required size, but to
always use strlen. Grow the buffer if needed.
Signed-off-by: Stefan Brüns <stefan.bruens@rwth-aachen.de> Acked-by: Simon Glass <sjg@chromium.org>
Stefan Brüns [Sat, 1 Oct 2016 18:41:39 +0000 (20:41 +0200)]
sandbox/fs: Make linking of nodes in os_dirent_ls more obvious
Previously, after reading/creating the second dirent, the second entry
would be chained to the first entry and the first entry would be linked
to head. Instead, immediately link the first entry to head.
Signed-off-by: Stefan Brüns <stefan.bruens@rwth-aachen.de> Acked-by: Simon Glass <sjg@chromium.org>
power: regulator: Add ctrl_reg and volt_reg fields for pmic
The ctrl reg contains bit fields to enable and disable regulators,
and volt_reg has the bit fields to configure the voltage values.
The registers are frequently accessed hence make them part
of dm_regulator_uclass_platdata structure.
Signed-off-by: Keerthy <j-keerthy@ti.com> Reviewed-by: Tom Rini <trini@konsulko.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Add support for gpio regulators. As of now this driver caters
to gpio regulators with one gpio. Supports setting voltage values to gpio
regulators and retrieving the values.
Signed-off-by: Keerthy <j-keerthy@ti.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Simon Glass [Mon, 26 Sep 2016 03:33:31 +0000 (21:33 -0600)]
x86: Don't export interrupt handlers with x86_64
We don't have a way of adjusting these at present so it is best to refuse to
export these functions. This can be implemented later if the API is required.
Signed-off-by: Simon Glass <sjg@chromium.org> Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Simon Glass [Thu, 6 Oct 2016 02:42:18 +0000 (20:42 -0600)]
x86: Adjust config to support DM_VIDEO
Update the common configuration so that it works correctly when
CONFIG_DM_VIDEO is enabled. This involves dropping the legacy CONFIG_VIDEO
option and changing the stdio device from "vga" to "vidconsole".
Signed-off-by: Simon Glass <sjg@chromium.org> Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Simon Glass [Thu, 6 Oct 2016 02:42:15 +0000 (20:42 -0600)]
dm: x86: video: Add a driver-model driver for ivybridge graphics
At present we use the legacy vesa driver for graphics. Add a driver which
supports driver model. This can be probed only when needed, removing the
need to start up the display if it is not used.
Signed-off-by: Simon Glass <sjg@chromium.org> Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Simon Glass [Thu, 6 Oct 2016 02:42:17 +0000 (20:42 -0600)]
dm: video: Add driver-model support to vesa graphics
Provide a function to run the Vesa BIOS for a given PCI device and obtain
the resulting configuration (e.g. display size) for use by the video
uclass. This makes it easier to write a video driver that uses vesa and
supports driver model.
Signed-off-by: Simon Glass <sjg@chromium.org> Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Simon Glass [Thu, 6 Oct 2016 02:42:16 +0000 (20:42 -0600)]
dm: stdio: Allow lazy probing of video devices
At present all video devices are probed on start-up. It would be better to
probe a device only when it is needed. This can happen if it is referenced
in the stdout environment variable, for example.
Add support for this by searching for a suitable device when needed, probing
it, and finding the stdio device it creates.
Signed-off-by: Simon Glass <sjg@chromium.org> Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Simon Glass [Thu, 6 Oct 2016 02:42:11 +0000 (20:42 -0600)]
Fix return value in trailing_strtoln()
This function should return -1 if there is no trailing integer in the
string. Instead it returns 0. Fix it by checking for this condition at the
start.
Signed-off-by: Simon Glass <sjg@chromium.org> Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Simon Glass [Thu, 6 Oct 2016 02:42:09 +0000 (20:42 -0600)]
Revert "x86: broadwell: gpio: Remove the codes to set up pin control"
This makes the assumption that setting up pinctrl in cpu_init_r() is safe.
On samus we need GPIOs before relocation in order to support power control.
This commit fixes the following message on boot:
initcall sequence ffe5c6f4 failed at call ffe01d3d (err=-1)
### ERROR ### Please RESET the board ###
In any case it seems better to leave init to driver model, so that it can
pick up the GPIO driver when it needs it. Since pinctrl is a dependency of
the GPIO driver, we may as well put the dependency there and avoid these
problems.
Stefan Roese [Fri, 30 Sep 2016 07:15:59 +0000 (09:15 +0200)]
x86: Fix Linux v4.7+ zimage booting (update bootparam.h)
Booting Linux kernel v4.7+ does not work since Linux kernel commit 974f221c
"x86/boot: Move compressed kernel to the end of the decompression buffer".
This patch adds the latest version of the setup_header struct, adding
"init_size" which is needed since this commit referenced above. With this
patch, booting Linux v4.8-rc8 does work again on x86 boards.
Signed-off-by: Stefan Roese <sr@denx.de> Reviewed-by: Simon Glass <sjg@chromium.org> Reviewed-by: Bin Meng <bmeng.cn@gmail.com> Tested-by: Bin Meng <bmeng.cn@gmail.com>
Stefan Roese [Mon, 26 Sep 2016 08:18:07 +0000 (10:18 +0200)]
x86: baytrail: Add 2nd eMMC controller to the PCI probe list
With this addition, the eMMC device available on the congatec and DFI
BayTrail SoM is detected correctly.
Signed-off-by: Stefan Roese <sr@denx.de> Cc: Simon Glass <sjg@chromium.org> Cc: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Stephen Warren [Mon, 10 Oct 2016 15:50:55 +0000 (09:50 -0600)]
ARM: tegra: reduce DRAM size mapped into MMU on ARM64
ARM CPUs can architecturally (speculatively) prefetch completely arbitrary
normal memory locations, as defined by the current translation tables. The
current MMU configuration for 64-bit Tegras maps an extremely large range
of addresses as DRAM, well beyond the actual physical maximum DRAM window,
even though U-Boot only needs access to the first 2GB of DRAM; the Tegra
port of U-Boot deliberately limits itself to 2GB of RAM since some HW
modules on at least some 64-bit Tegra SoCs can only access a 32-bit
physical address space. This change reduces the amount of RAM mapped via
the MMU to disallow the CPU from ever speculatively accessing RAM that
U-Boot will definitely not access. This avoids the possibility of the HW
raising SError due to accesses to always-invalid physical addresses.
Signed-off-by: Stephen Warren <swarren@nvidia.com> Signed-off-by: Tom Warren <twarren@nvidia.com>
Without this patch the following error will be shown:
stdio_add_devices: Video device failed (ret=-22)
As commit ec5507707a1d1e84056a6c864338f95f6118d3ca (video: tegra: Move
to using simple-panel and pwm-backlight) states the Colibri T20 needs
updating too which this patch finally attempts doing.
Signed-off-by: Marcel Ziswiler <marcel.ziswiler@toradex.com> Acked-by: Stephen Warren <swarren@nvidia.com> Signed-off-by: Tom Warren <twarren@nvidia.com>
regulator: fixed: honour optionality of enable gpio
According to the binding documentation the fixed regulator enable GPIO
is optional. However so far registration thereof failed if no enable
GPIO was specified. Fix this by making it entirely optional whether an
enable GPIO is used.
Signed-off-by: Marcel Ziswiler <marcel.ziswiler@toradex.com> Acked-by: Stephen Warren <swarren@nvidia.com> Signed-off-by: Tom Warren <twarren@nvidia.com>
sunxi: OLinuXino Lime A20 boards: Use 384 MHz DRAM clock
We have a number of OlinuXino Lime2 boards (both NAND and eMMC versions)
which were experiencing sporadic hangs. After testing with some heavy
benchmarking and help from the Armbian forum, it was pinned down as the
DRAM settings for the board. The default is 480MHz, but this is unstable,
and even the build instructions from the vendor Olimex themselves say to
set the DRAM clock to 384. See line 96 at:
https://github.com/OLIMEX/OLINUXINO/blob/master/SOFTWARE/A20/A20-build-3.4.103-release-2/BUILD_DESCRIPTION_A20_Olimex_kernel_3.4.103%2B_Jessie_rel_2.txt
Reviewed-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Hans de Goede <hdegoede@redhat.com>
sunxi: Fix H3 DRAM impedance calibration on rev. A chips
H3 seems to have a silicon bug breaking the impedance calibration.
This is currently worked around in software by multiple steps
combining the results to replace the wrong values.
Revision A chips need a different workaround, which is present in
the vendor bootloader too, but got overlooked in lack of
information and affected boards till now.
This commit adds a simplified version without correction factor,
which would be 1.00 for all known boards anyway.
Signed-off-by: Jens Kuske <jenskuske@gmail.com> Reviewed-by: Jagan Teki <jteki@openedev.com> Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Masahiro Yamada [Sat, 8 Oct 2016 04:25:30 +0000 (13:25 +0900)]
reset: declare fdtdec_phandle_args as struct to fix warning
The of_xlate() callback needs to know fdtdec_phandle_args is struct.
Otherwise, the following warning is displayed.
include/reset-uclass.h:40:11: warning: 'struct fdtdec_phandle_args'
declared inside parameter list
struct fdtdec_phandle_args *args);
^
include/reset-uclass.h:40:11: warning: its scope is only this
definition or declaration, which is probably not what you want
Masahiro Yamada [Sat, 8 Oct 2016 04:25:29 +0000 (13:25 +0900)]
ARM: uniphier: insert udelay() just before support_card_reset_deassert()
As for LD11/LD20, we can no longer rely on the udelay() in the PLL
init functions. udelay(200) is needed here to keep the ethernet
device in the reset state for enough time. Anyway, 200 usec is
quite short for humans, so nobody cares it.
Masahiro Yamada [Sat, 8 Oct 2016 04:25:27 +0000 (13:25 +0900)]
ARM: uniphier: enable CONFIG_SYS_NO_FLASH if no CONFIG_MICRO_SUPPORT_CARD
NOR flash devices are seldom used on UniPhier platforms these days.
The only use case I see is the Micro Support Card is connected.
Otherwise, define CONFIG_SYS_NO_FLASH to disable NOR FLASH.
Masahiro Yamada [Sat, 8 Oct 2016 04:25:24 +0000 (13:25 +0900)]
ARM: uniphier: update DRAM init code for LD20 SoC (2nd)
- Do not reference CONFIG_DDR_FREQ; now the DDR frequency
is passed from the uniphier_board_data structure
- Constify parameter arrays
- Tidy up cluttered macros
- Lots of code cleanups
- Lots of coding style fixes
ARM: uniphier: do not setup pins for System Bus on NAND boot mode
For LD11 and LD20 SoCs, the System Bus and NAND are multiplexed
in the same I/O pins. When booting from a NAND device, pin-mux
for the System Bus must not be set-up because they are exclusive
with each other.
York Sun [Tue, 4 Oct 2016 21:33:50 +0000 (14:33 -0700)]
tools: buildman: Remove duplicated code
Signed-off-by: York Sun <york.sun@nxp.com> CC: Simon Glass <sjg@chromium.org>
Fixed commit subject: Signed-off-by: Simon Glass <sjg@chromium.org> Acked-by: Simon Glass <sjg@chromium.org>
Paul Burton [Tue, 27 Sep 2016 15:03:57 +0000 (16:03 +0100)]
dtoc: Decode strings for struct.unpack on python 3.x
On python 3.x struct.unpack will complain if we provide it with a
string since it expects to operate on a bytes object. In order to
satisfy this requirement, encode the string to a bytes object when
running on python 3.x.
Signed-off-by: Paul Burton <paul.burton@imgtec.com> Acked-by: Simon Glass <sjg@chromium.org>
Paul Burton [Tue, 27 Sep 2016 15:03:56 +0000 (16:03 +0100)]
dtoc: Use items() to iterate over dictionaries in python 3.x
In python 3.x the iteritems() method has been removed from dictionaries,
and the items() method does effectively the same thing. On python 2.x
using items() is a little less efficient since it involves copying data,
but as speed isn't a concern in the affected code switch to using
items() anyway for simplicity.
Signed-off-by: Paul Burton <paul.burton@imgtec.com> Acked-by: Simon Glass <sjg@chromium.org>
Paul Burton [Tue, 27 Sep 2016 15:03:55 +0000 (16:03 +0100)]
patman: Fix doctest StringIO import for python 3.x
In python 3.x StringIO is no longer a module, and the class can instead
be found in the io module. Adjust the code in the doctest input to
account for both.
Signed-off-by: Paul Burton <paul.burton@imgtec.com> Acked-by: Simon Glass <sjg@chromium.org>
Paul Burton [Tue, 27 Sep 2016 15:03:54 +0000 (16:03 +0100)]
patman: Use items() to iterate over dictionaries
In python 3.x the iteritems() method has been removed from dictionaries,
and the items() method does effectively the same thing. On python 2.x
using items() is a little less efficient since it involves copying data,
but as speed isn't a concern in this code switch to using items() anyway
for simplicity.
Signed-off-by: Paul Burton <paul.burton@imgtec.com> Acked-by: Simon Glass <sjg@chromium.org>
Paul Burton [Tue, 27 Sep 2016 15:03:52 +0000 (16:03 +0100)]
patman: Import 'configparser' lower case to be python 3.x safe
In python 3.x module names used in import statements are case sensitive,
and the configparser module is named in all lower-case. Import it as such
in order to avoid errors when running with python 3.x.
Signed-off-by: Paul Burton <paul.burton@imgtec.com> Acked-by: Simon Glass <sjg@chromium.org>
Paul Burton [Tue, 27 Sep 2016 15:03:50 +0000 (16:03 +0100)]
patman: Make print statements python 3.x safe
In python 3.x, print must be used as a function call. Convert all print
statements to the function call style, importing from __future__ where
we print with no trailing newline or print to a file object.
Signed-off-by: Paul Burton <paul.burton@imgtec.com> Acked-by: Simon Glass <sjg@chromium.org>
Paul Burton [Tue, 27 Sep 2016 15:03:49 +0000 (16:03 +0100)]
patman: Replace tabs with spaces
In preparation for running on python 3.x, which will refuse to run
scripts which mix tabs & spaces for indentation, replace 2 tab
characters present in series.py with spaces.
Signed-off-by: Paul Burton <paul.burton@imgtec.com> Acked-by: Simon Glass <sjg@chromium.org>
Simon Glass [Sun, 25 Sep 2016 21:52:19 +0000 (15:52 -0600)]
dtoc: Add a way for tests to request the fallback library
We need to test both the normal (Python libfdt module) and fallback (fdtget)
implementations of the Fdt class. Add a way to select which implementation
to use.
projects / u-boot / log