Driver model uses a different way to find the SPI bus and slave from the
numbered devices given on the command line. Adjust the code to suit.
We use a generic SPI device, and attach it to the SPI bus before performing
the transaction.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Jagannadha Sutradharudu Teki <jagannadh.teki@gmail.com>
Some files are using SPI functions but not explitly including the SPI
header file. Fix this, since driver model needs it.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Jagannadha Sutradharudu Teki <jagannadh.teki@gmail.com>
Driver model does its own init, so we don't need this.
There is still a call in board_f.c but it is only enabled by CONFIG_HARD_SPI.
It is easy enough to disable that option when converting boards which use
it to driver model.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Jagannadha Sutradharudu Teki <jagannadh.teki@gmail.com>
U-Boot includes a SPI emulation driver already but it is not explicit, and
is hidden in the SPI flash code.
Conceptually with sandbox's SPI implementation we have a layer which
creates SPI bus transitions and a layer which interprets them, currently
only for SPI flash. The latter is actually an emulation, and it should be
possible to add more than one emulation - not just SPI flash.
Add a SPI emulation uclass so that other emulations can be plugged in to
support different types of emulated devices on difference buses/chip
selects.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Jagannadha Sutradharudu Teki <jagannadh.teki@gmail.com>
Add a uclass which provides access to SPI buses and includes operations
required by SPI.
For a time driver model will need to co-exist with the legacy SPI interface
so some parts of the header file are changed depending on which is in use.
The exports are adjusted also since some functions are not available with
driver model.
Boards must define CONFIG_DM_SPI to use driver model for SPI.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Jagannadha Sutradharudu Teki <jagannadh.teki@gmail.com>
(Discussed some follow-up comments which will address in future add-ons)
The sequence number is unique within the uclass, so state this clearly.
Signed-off-by: Simon Glass <sjg@chromium.org>
Acked-by: Jagannadha Sutradharudu Teki <jagannadh.teki@gmail.com>
Buses sometimes want to pass data to their children when they are probed.
For example, a SPI bus may want to tell the slave device about the chip
select it is connected to.
Add a new function to permit the parent data to be supplied to the child.
Signed-off-by: Simon Glass <sjg@chromium.org>
Acked-by: Jagannadha Sutradharudu Teki <jagannadh.teki@gmail.com>
Buses need to iterate through their children in some situations. Add a few
functions to make this easy.
Signed-off-by: Simon Glass <sjg@chromium.org>
Acked-by: Jagannadha Sutradharudu Teki <jagannadh.teki@gmail.com>
Add a SPI device which can be used for testing SPI flash features in
sandbox.
Also add a cros_ec device since with driver model the Chrome OS EC
emulation will not otherwise be available.
Reviewed-by: Jagannadha Sutradharudu Teki <jagannadh.teki@gmail.com>
Signed-off-by: Simon Glass <sjg@chromium.org>
Convert the exynos GPIO driver to driver model. This implements the generic
GPIO interface but not the extra Exynos-specific functions.
Signed-off-by: Simon Glass <sjg@chromium.org>
With driver model GPIOs must be requested before use. Make sure this is
done correctly.
(Note that the soft SPI part of universal is omitted, since this driver
is about to be replaced with a driver-model-aware version)
Signed-off-by: Simon Glass <sjg@chromium.org>
The defines at the top of the GPIO driver use single-character names for
parameters which are not very descriptive.
Improve these to use descriptive parameter names.
Signed-off-by: Simon Glass <sjg@chromium.org>
The wrong header is being included, thus requiring the code to re-declare
the generic GPIO interface in each GPIO header.
Fix this.
Signed-off-by: Simon Glass <sjg@chromium.org>
The generic board deadline is approaching, and we need this feature to
enable driver model. Enable CONFIG_SYS_GENERIC_BOARD for s5p_goni.
Signed-off-by: Simon Glass <sjg@chromium.org>
The generic board deadline is approaching, and we need this feature to
enable driver model. Enable CONFIG_SYS_GENERIC_BOARD for smdkc100.
Signed-off-by: Simon Glass <sjg@chromium.org>
The pinctrl bindings used by Linux are an incomplete description of the
hardware. It is possible in most cases to determine the register address
of each, but not in all cases. By adding an additional property we can
fix this, and avoid adding a table to U-Boot for every single Exynos
SOC.
Signed-off-by: Simon Glass <sjg@chromium.org>
We don't include the pinctrl functions for U-Boot as they use up quite
a bit of space and are not used.
We could instead perhaps eliminate this material with fdtgrep, but so far
this tool has not made it to upstream.
Signed-off-by: Simon Glass <sjg@chromium.org>
Bring in required device tree files for pinctrl from Linux v3.14. These
are initially unchanged and have a number of pieces not needed by U-Boot.
Note that exynos5420 is renamed to exynos54xx here since we want to
support exynos5422 also.
Signed-off-by: Simon Glass <sjg@chromium.org>
We should be consistent about this. The kernel has moved to #include
which breaks error reporting to some extent but does allow us to include
binding files.
Signed-off-by: Simon Glass <sjg@chromium.org>
Add proper initialization of GPIO pins used by software i2c.
Signed-off-by: Robert Baldyga <r.baldyga@samsung.com>
Acked-by: Simon Glass <sjg@chromium.org>
The driver model supports two ways for passing device parameters;
Device Tree and platform_data (board file).
Each driver should generally support both of them because some
popular IPs are used on various platforms.
Assume the following scenario:
- The driver Foo is used on SoC Bar and SoC Baz
- The SoC Bar uses Device Tree control (CONFIG_OF_CONTROL=y)
- The SoC Baz does not support Device Tree; uses a board file
In this situation, the device driver Foo should work with/without
the device tree control. The driver should have .of_match and
.ofdata_to_platdata members for SoC Bar, while they are meaningless
for SoC Baz; therefore those device-tree control code should go
inside #ifdef CONFIG_OF_CONTROL.
The driver code will be like this:
#ifdef CONFIG_OF_CONTROL
static const struct udevice_id foo_of_match = {
{ .compatible = "foo_driver" },
{},
}
static int foo_ofdata_to_platdata(struct udevice *dev)
{
...
}
#endif
U_BOOT_DRIVER(foo_driver) = {
...
.of_match = of_match_ptr(foo_of_match),
.ofdata_to_platdata = of_match_ptr(foo_ofdata_to_platdata),
...
}
This idea has been borrowed from Linux.
(In Linux, this macro is defined in include/linux/of.h)
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Acked-by: Simon Glass <sjg@chromium.org>
The header files include/dm/platdata.h and include/dm/uclass.h
use ll_entry_declare(); therefore they depend on
include/linker_lists.h.
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Acked-by: Simon Glass <sjg@chromium.org>
The header file include/linker_lists.h uses __aligned();
therefore it depends on include/linux/compiler.h
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Acked-by: Simon Glass <sjg@chromium.org>
if (strncmp(name, entry->name, len))
continue;
/* Full match */
if (len == strlen(entry->name))
return entry;
is equivalent to:
if (!strcmp(name, entry->name))
return entry;
The latter is simpler.
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Acked-by: Simon Glass <sjg@chromium.org>
Acked-by: Igor Grinberg <grinberg@compulab.co.il>
The function uclass_add() checks uc_drv->ops as follows:
if (uc_drv->ops) {
dm_warn("No ops for uclass id %d\n", id);
return -EINVAL;
}
It seems odd because it warns "No ops" when uc_drv->ops has
non-NULL pointer. (Looks opposite.)
Anyway, most of UCLASS_DRIVER entries have no .ops member.
This check makes no sense.
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Acked-by: Simon Glass <sjg@chromium.org>
The struct udevice stands for a device, not a driver.
The driver_info.name is a driver's name, which is referenced
to bind devices.
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Acked-by: Simon Glass <sjg@chromium.org>
Update Apalis T30 as per the following commits
c369139234
tegra: dts: Add serial port details
461be2f96e
kconfig: remove redundant "string" type in arch and board Kconfigs
f1ef2b6233
kconfig: move CONFIG_DEFAULT_DEVICE_TREE to kconfig
Signed-off-by: Marcel Ziswiler <marcel@ziswiler.com>
Signed-off-by: Tom Warren <twarren@nvidia.com>
Add missing chosen stdout-path device tree node. This got missed by
commit
c369139234
tegra: dts: Add serial port details
Signed-off-by: Marcel Ziswiler <marcel@ziswiler.com>
Signed-off-by: Tom Warren <twarren@nvidia.com>
Clean-up a spurious new line which got introduced resp. left behind by
commit
f1ef2b6233
kconfig: move CONFIG_DEFAULT_DEVICE_TREE to kconfig
Signed-off-by: Marcel Ziswiler <marcel@ziswiler.com>
Acked-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Tom Warren <twarren@nvidia.com>
On popular request this now completes the Warren's work started for
TK1:
aeb3fcb359
ARM: tegra: Use mem size from MC rather than ODMDATA
In addition to the move of using the Tegra memory controller (MC)
register rather than ODMDATA for T20, T30 and T114 as well it further
uses the generic get_ram_size() function (see "common/memsize.c")
<supposed to be used in each and every U-Boot port>TM. Added benefit is
that it should <catch 99% of hardware related (i. e. reliably
reproducible) memory errors> as well.
Thoroughly tested on the various Toradex line of Tegra modules
available which unfortunately does not include T114 and T124 (yet at
least) plus on the Jetson TK1.
Based-on-work-by: Stephen Warren <swarren@nvidia.com>
Based-on-work-by: Tom Warren <twarren@nvidia.com>
Signed-off-by: Marcel Ziswiler <marcel@ziswiler.com>
Acked-by: Stephen Warren <swarren@nvidia.com>
Signed-off-by: Tom Warren <twarren@nvidia.com>
Wondering what exactly that one should bring (;-p). Looks like a remnant of the last commit
783e6a72b8
kconfig: move CONFIG_OF_* to Kconfig
Signed-off-by: Marcel Ziswiler <marcel@ziswiler.com>
Acked-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Tom Warren <twarren@nvidia.com>
Fix ASIX USB to Ethernet reset which due to the new driver model Tegra
GPIO driver changes now requires a label string to be provided
otherwise the reservation and subsequent direction/value calls will
fail.
This fixes a regression introduced by commit:
2fccd2d96b
tegra: Convert tegra GPIO driver to use driver model
Signed-off-by: Marcel Ziswiler <marcel@ziswiler.com>
Acked-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Tom Warren <twarren@nvidia.com>
Fix Tegra GPIO driver to not crash resp. misbehave upon requesting
GPIOs with an empty aka NULL label. As the driver uses exclusively the
label to check for reservation status actually supplying one is
mandatory!
This fixes a regression introduced by commit:
2fccd2d96b
tegra: Convert tegra GPIO driver to use driver model
Signed-off-by: Marcel Ziswiler <marcel@ziswiler.com>
Acked-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Tom Warren <twarren@nvidia.com>
During rigorous testing of our latest update infrastructure I came
across quite consistent timeouts on certain eMMC parts (e.g. Hynix
H26M21001ECR) when writing big (e.g. in excess of 400 MB) file system
images:
MMC write: dev # 0, block # 40960, count 944128 ...
mmc_send_cmd_bounced: MMC Timeout
Interrupt status 0x00000001
Interrupt status enable 0xdfff003b
Interrupt signal enable 0xdfff0002
Present status 0x01870106
mmc write failed
Comparing the various data sheets I came across the following timeout
specification:
Secure Erase/TRIM Timeout=300ms*2*10=6000ms
Unfortunately empirical testing still failed albeit much more rarely.
Increasing the timeout to 8000ms made it finally disappear entirely.
This patch allows us writing various eMMC parts without seeing any
further issues.
Signed-off-by: Marcel Ziswiler <marcel@ziswiler.com>
Signed-off-by: Tom Warren <twarren@nvidia.com>
Use architecture specific memcpy to speed up things.
Signed-off-by: Marcel Ziswiler <marcel@ziswiler.com>
Tested-by: Stephen Warren <swarren@nvidia.com>
Signed-off-by: Tom Warren <twarren@nvidia.com>
This patch adds board support for the Toradex Apalis T30 a computer on
module which can be used on different carrier boards.
For the sake of ease of use we do not distinguish between different
carrier boards for now as the base module features are deemed
sufficient enough for regular booting.
The following functionality is working so far:
- eMMC boot and environment storage
- Gigabit Ethernet (once Thierry's PCIe as well as my E1000 resp. i210
fixes hit mainline)
- MMC/SD cards (both 8-bit as well as 4-bit slot)
- USB client/host (dual role port as client e.g. for DFU/UMS, other two
ports as host)
Signed-off-by: Marcel Ziswiler <marcel@ziswiler.com>
Acked-by: Stephen Warren <swarren@nvidia.com>
Signed-off-by: Tom Warren <twarren@nvidia.com>
This pinmux tables currently omit any configuration for PCIe clk_req,
wake, and rst pins, which in turn causes intermittent failures in
U-Boot's PCIe support. Import an updated version of the pinmux tables
which rectifies this.
Signed-off-by: Stephen Warren <swarren@nvidia.com>
Signed-off-by: Tom Warren <twarren@nvidia.com>
When building U-Boot with CONFIG_X86_RESET_VECTOR, the linking
process misses the resetvec.o and start16.o so it cannot generate
the rom version of U-Boot. The arch/x86/cpu/Makefile is updated to
pull them into the final linking process.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Acked-by: Simon Glass <sjg@chromium.org>
Add a new setup@ section to the FIT which can be used to provide a setup
binary for booting Linux on x86. This makes it possible to boot x86 from
a FIT.
Signed-off-by: Simon Glass <sjg@chromium.org>
Since sandbox is used for testing, it should be able to 'boot' an image
from any archhitecture. This allows us to test an image by loading it in
sandbox.
Signed-off-by: Simon Glass <sjg@chromium.org>
The x86 bootm code is quite special, and geared to zimage. Adjust it
to support device tree and make it more like the ARM code, with
separate bootm stages and functions for each stage.
Create a function announce_and_cleanup() to handle printing the
"Starting kernel ..." message and put it in bootm so it is in one
place and can be used by any loading code. Also move the
board_final_cleanup() function into bootm.
Signed-off-by: Simon Glass <sjg@chromium.org>
