summaryrefslogtreecommitdiff
path: root/lib
AgeCommit message (Collapse)Author
2015-11-04dm: spl: Support device tree when BSS is in a different sectionSimon Glass
At present in SPL we place the device tree immediately after BSS. This avoids needing to copy it out of the way before BSS can be used. However on some boards BSS is not placed with the image - e.g. it can be in RAM if available. Add an option to tell U-Boot that the device tree should be placed at the end of the image binary (_image_binary_end) instead of at the end of BSS. Note: A common reason to place BSS in RAM is to support the FAT filesystem. We should update the code so that it does not use so much BSS. Signed-off-by: Simon Glass <sjg@chromium.org> Signed-off-by: Michal Simek <michal.simek@xilinx.com>
2015-11-04fdt: Correct handling of alias regionsSimon Glass
At present the last four bytes of the alias region are dropped in the case where the last alias is included. This results in a corrupted device tree. Fix this. Signed-off-by: Simon Glass <sjg@chromium.org> Signed-off-by: Michal Simek <michal.simek@xilinx.com>
2015-11-04fdt: Add a function to look up a /chosen propertySimon Glass
It is sometimes useful to find a property in the chosen node. Add a function for this. Signed-off-by: Simon Glass <sjg@chromium.org> Signed-off-by: Michal Simek <michal.simek@xilinx.com>
2015-10-27Merge git://git.denx.de/u-boot-dmTom Rini
2015-10-23fdt: fix fdtdec_get_pci_addr() for CONFIG_PHYS_64BITStephen Warren
PCI addresses are always represented as 3 cells in DT. (one cell for bus and device, and two cells for a 64-bit addres). This does not vary based on either the physical address size of the CPU, nor any #address-cells property in DT (or more precisely, #address-cells must be set to 3 in any PCIe controller's node). Fix fdtdec_get_pci_addr() to use conversion functions that operate on (fixed) cell-sized data rather than (varying) physical-address-sized data, so that the function works on 64-bit systems. Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Simon Glass <sjg@chromium.org> Reviewed-by: Thierry Reding <treding@nvidia.com>
2015-10-23dm: tpm: Drop CONFIG_DM_TPMSimon Glass
Now that all TPM drivers use driver model, we can drop the special driver model CONFIG option. Signed-off-by: Simon Glass <sjg@chromium.org> Acked-by: Christophe Ricard<christophe-h.ricard@st.com>
2015-10-23dm: tpm: Remove every compilation switch for TPM driver modelChristophe Ricard
As every TPM drivers support UCLASS_TPM, we can only rely on DM_TPM functions. This simplify a bit the code. Signed-off-by: Christophe Ricard <christophe-h.ricard@st.com> Reviewed-by: Tom Rini <trini@konsulko.com> Acked-by: Simon Glass <sjg@chromium.org>
2015-10-22dm: implement a Timer uclassThomas Chou
Implement a Timer uclass to work with lib/time.c. Signed-off-by: Thomas Chou <thomas@wytron.com.tw> Acked-by: Simon Glass <sjg@chromium.org>
2015-10-21debug_uart: Adjust the declaration of debug_uart_init()Simon Glass
We want to be able to add other common code to this function. So change the driver's version to have an underscore before it, just like _debug_uart_putc(). Define debug_uart_init() to call this version. Update all drivers to this new method. Signed-off-by: Simon Glass <sjg@chromium.org> Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
2015-10-11Add support for LZ4 decompression algorithmJulius Werner
This patch adds support for LZ4-compressed FIT image contents. This algorithm has a slightly worse compression ration than LZO while being nearly twice as fast to decompress. When loading images from a fast storage medium this usually results in a boot time win. Sandbox-tested only since I don't have a U-Boot development system set up right now. The code was imported unchanged from coreboot where it's proven to work, though. I'm mostly interested in getting this recognized by mkImage for use in a downstream project. Signed-off-by: Julius Werner <jwerner@chromium.org> Acked-by: Simon Glass <sjg@chromium.org>
2015-10-03fdtdec: fix parsing 'reg' property with zero value in '#size-cells'Przemyslaw Marczak
After rework of lib/fdtdec.c by: commit: 02464e3 fdt: add new fdt address parsing functions the function fdtdec_get_addr() doesn't work as previous, because the implementation assumes that properties '#address-cells' and '#size-cells' are equal to 1, which can be not true sometimes. The new API introduced fdtdec_get_addr_size_auto_parent() for the 'reg' property parsing, but the implementation assumes, that #size-cells can't be less than 1. This causes that the following children's 'reg' property can't be reached: parent@0x0 { #address-cells = <1>; #size-cells = <0>; children@0x100 { reg = < 0x100 >; }; }; Change the condition value from '1' to '0', which allows parsing property with at least zero #size-cells, fixes the issue. Now, fdtdec_get_addr_size_auto_parent() works properly. Tested on: Odroid U3/X2, Trats, Trats2, Odroid XU3, Snow (by Simon). Signed-off-by: Przemyslaw Marczak <p.marczak@samsung.com> Acked-by: Stephen Warren <swarren@nvidia.com> Acked-by: Simon Glass <sjg@chromium.org> Tested-by: Simon Glass <sjg@chromium.org>
2015-10-03fdt: fix fdtdec_get_addr_size not to require any size cellsStephen Warren
fdtdec_get_addr_size() may be used in two cases: a) With sizep supplied, in which case both an address and a size are parsed from DT. In this case, the DT property must be large enough to contain both values. b) With sizep NULL, in which case only an address is parsed from DT. In this case, the DT property only need be large enough to contain this address value. Commit 02464e386bb5 "fdt: add new fdt address parsing functions" broke this relaxed checking, and required the DT property to contain both an address and a size value in all cases. Fix fdtdec_get_addr_size() to vary ns based on whether the size value is being parsed from the DT or not. This is safe since the function only parses the first entry in the property, so the overall value of (na + ns) need not be accurate, since it is never used to step through the property data to find other entries. Besides, this fixed behaviour essentially matches the original behaviour before the patch this patch fixes. (The original code validated that the property was exactly the length of either na or (na + ns), whereas the current code only validates that the property is at least that long. For non-failure cases, the two behaviours are identical). Cc: Przemyslaw Marczak <p.marczak@samsung.com> Cc: Simon Glass <sjg@chromium.org> Cc: Thierry Reding <treding@nvidia.com> Cc: Bin Meng <bmeng.cn@gmail.com> Cc: Michal Suchanek <hramrach@gmail.com> Fixes: 02464e386bb5 ("fdt: add new fdt address parsing functions") Reported-by: Przemyslaw Marczak <p.marczak@samsung.com> Signed-off-by: Stephen Warren <swarren@nvidia.com> Tested-by: Przemyslaw Marczak <p.marczak@samsung.com> Acked-by: Simon Glass <sjg@chromium.org>
2015-09-16Merge git://git.denx.de/u-boot-fdtTom Rini
2015-09-15env: import: hashtable: Free memory allocated before exiting from himport_r()Lukasz Majewski
ithout this patch memory is not released on early exit. Signed-off-by: Lukasz Majewski <l.majewski@majess.pl>
2015-09-15env: import: hashtable: Prevent buffer overrun when importing environment ↵Lukasz Majewski
from file Lets consider following scenario: - One uses echo -n "key=value" to define environment variable in a file (single variable) - The file content is "key=value" without any terminating byte (e.g. 0x0a or 0x0d). - The file is loaded to u-boot non zero'ed RAM buffer (with load command). - Then "env import -t -r $loadaddr $filesize" is executed. - Due to lack of proper termination byte we have classical example of buffer overrun. This patch prevents from this by allocating one extra byte than size and explicitly null terminate it. There should be no change for normal env import operation after applying this patch. Signed-off-by: Lukasz Majewski <l.majewski@majess.pl>
2015-09-15fdt: add new fdt address parsing functionsStephen Warren
fdtdec_get_addr_size() hard-codes the number of cells used to represent an address or size in DT. This is incorrect in many cases depending on the DT binding for a particular node or property (e.g. it is incorrect for the "reg" property). In most cases, DT parsing code must use the properties #address-cells and #size-cells to parse addres properties. This change splits up the implementation of fdtdec_get_addr_size() so that the core logic can be used for both hard-coded and non-hard-coded cases. Various wrapper functions are implemented that support cases where hard-coded cell counts should or should not be used, and where the client does and doesn't know the parent node ID that contains the properties #address-cells and #size-cells. dev_get_addr() is updated to use the new functions. Core functionality in fdtdec_get_addr_size_fixed() is widely tested via fdtdec_get_addr_size(). I tested fdtdec_get_addr_size_auto_noparent() and dev_get_addr() by manually modifying the Tegra I2C driver to invoke them. Much of the core implementation of fdtdec_get_addr_size_fixed(), fdtdec_get_addr_size_auto_parent(), and fdtdec_get_addr_size_auto_noparent() comes from Thierry Reding's previous commit "fdt: Fix fdtdec_get_addr_size() for 64-bit". Based-on-work-by: Thierry Reding <treding@nvidia.com> Cc: Thierry Reding <treding@nvidia.com> Cc: Simon Glass <sjg@chromium.org> Cc: Michal Suchanek <hramrach@gmail.com> Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Simon Glass <sjg@chromium.org> Dropped #define DEBUG at the top of fdtdec.c: Signed-off-by: Simon Glass <sjg@chromium.org>
2015-09-11Move malloc_cache_aligned() to its own headerSimon Glass
At present malloc.h is included everywhere since it recently was added to common.h in this commit: 4519668 mtd/nand/ubi: assortment of alignment fixes This seems wasteful and unnecessary. We have been trying to trim down common.h and put separate functions into separate header files and that change goes in the opposite direction. Move malloc_cache_aligned() to a new header so that this can be avoided. The header would perhaps be better named as alignmem.h but it needs to be included after common.h and people might be confused by this. With the name memalign.h it fits nicely after malloc() in most cases. Signed-off-by: Simon Glass <sjg@chromium.org> Acked-by: Marcel Ziswiler <marcel.ziswiler@toradex.com>
2015-09-04mmc: dw_mmc: Probe the MMC from OFMarek Vasut
Rework the driver to probe the MMC controller from Device Tree and make it mandatory. There is no longer support for probing from the ancient qts-generated header files. This patch now also removes previous temporary workaround. Signed-off-by: Marek Vasut <marex@denx.de> Cc: Dinh Nguyen <dinguyen@opensource.altera.com> Cc: Pantelis Antoniou <panto@antoniou-consulting.com> Cc: Tom Rini <trini@konsulko.com>
2015-08-31tpm: Add functions to access flags and permissionsSimon Glass
Add a few new functions which will be used by the test command in a future patch. Signed-off-by: Simon Glass <sjg@chromium.org> Acked-by: Christophe Ricard<christophe-h.ricard@st.com> Reviewed-by: Heiko Schocher <hs@denx.de>
2015-08-31dm: tpm: Convert I2C driver to driver modelSimon Glass
Convert the tpm_tis_i2c driver to use driver model and update boards which use it. Signed-off-by: Simon Glass <sjg@chromium.org> Acked-by: Christophe Ricard<christophe-h.ricard@st.com> Reviewed-by: Heiko Schocher <hs@denx.de>
2015-08-31dm: tpm: Convert the TPM command and library to driver modelSimon Glass
Add driver model support to the TPM command and the TPM library. Both support only a single TPM at present. Signed-off-by: Simon Glass <sjg@chromium.org> Acked-by: Christophe Ricard<christophe-h.ricard@st.com> Reviewed-by: Heiko Schocher <hs@denx.de>
2015-08-31tpm: Add Kconfig options for TPMsSimon Glass
Add new Kconfig options for TPMs in preparation for moving boards to use Kconfig for TPM configuration. Signed-off-by: Simon Glass <sjg@chromium.org> Acked-by: Christophe Ricard<christophe-h.ricard@st.com> Reviewed-by: Heiko Schocher <hs@denx.de>
2015-08-28mtd/nand/ubi: assortment of alignment fixesMarcel Ziswiler
Various U-Boot adoptions/extensions to MTD/NAND/UBI did not take buffer alignment into account which led to failures of the following form: ERROR: v7_dcache_inval_range - start address is not aligned - 0x1f7f0108 ERROR: v7_dcache_inval_range - stop address is not aligned - 0x1f7f1108 Signed-off-by: Marcel Ziswiler <marcel.ziswiler@toradex.com> Reviewed-by: Simon Glass <sjg@chromium.org> Acked-by: Scott Wood <scottwood@freescale.com> [trini: Add __UBOOT__ hunk to lib/zlib/zutil.c due to malloc.h in common.h] Signed-off-by: Tom Rini <trini@konsulko.com>
2015-08-28lib/display_options: Fix print_freqSuriyan Ramasami
Build without CONFIG_SPL_SERIAL_SUPPORT does not print the cpu freq. I have seen this in the odroid U3 board, where on boot one sees this: CPU: Exynos4412 @ GHz instead of: CPU: Exynos4412 @ 1 GHz I am assuming that this change was done to get rid of compiler warnings related to unused variables when building with CONFIG_SPL_SERIAL_SUPPORT not being defined in an SPL build. Signed-off-by: Suriyan Ramasami <suriyan.r@gmail.com> Acked-by: Simon Glass <sjg@chromium.org> Acked-by: Heiko Schocher <hs@denx.de> Reviewed-by: Joe Hershberger <joe.hershberger@ni.com>
2015-08-26fdtdec: Fix possible infinite loop in fdtdec_get_pci_vendev()Bin Meng
When there is no valid compatible string in current list, we should advance to next one in the compatible string list. Signed-off-by: Bin Meng <bmeng.cn@gmail.com> Acked-by: Simon Glass <sjg@chromium.org>
2015-08-18of: clean up OF_CONTROL ifdef conditionalsMasahiro Yamada
We have flipped CONFIG_SPL_DISABLE_OF_CONTROL. We have cleansing devices, $(SPL_) and CONFIG_IS_ENABLED(), so we are ready to clear away the ugly logic in include/fdtdec.h: #ifdef CONFIG_OF_CONTROL # if defined(CONFIG_SPL_BUILD) && !defined(SPL_OF_CONTROL) # define OF_CONTROL 0 # else # define OF_CONTROL 1 # endif #else # define OF_CONTROL 0 #endif Now CONFIG_IS_ENABLED(OF_CONTROL) is the substitute. It refers to CONFIG_OF_CONTROL for U-boot proper and CONFIG_SPL_OF_CONTROL for SPL. Also, we no longer have to cancel CONFIG_OF_CONTROL in include/config_uncmd_spl.h and scripts/Makefile.spl. Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com> Reviewed-by: Tom Rini <trini@konsulko.com> Reviewed-by: Simon Glass <sjg@chromium.org> Acked-by: Linus Walleij <linus.walleij@linaro.org>
2015-08-18of: flip CONFIG_SPL_DISABLE_OF_CONTROL into CONFIG_SPL_OF_CONTROLMasahiro Yamada
As we discussed a couple of times, negative CONFIG options make our life difficult; CONFIG_SYS_NO_FLASH, CONFIG_SYS_DCACHE_OFF, ... and here is another one. Now, there are three boards enabling OF_CONTROL on SPL: - socfpga_arria5_defconfig - socfpga_cyclone5_defconfig - socfpga_socrates_defconfig This commit adds CONFIG_SPL_OF_CONTROL for them and deletes CONFIG_SPL_DISABLE_OF_CONTROL from the other boards to invert the logic. Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com> Reviewed-by: Tom Rini <trini@konsulko.com> Reviewed-by: Simon Glass <sjg@chromium.org>
2015-08-14Revert "fdt: Fix fdtdec_get_addr_size() for 64-bit"Simon Glass
This reverts commit 5b34436035fc862b5e8d0d2c3eab74ba36f1a7f4. This function has a few problems. It calls fdt_parent_offset() which as mentioned in code review is very slow. https://patchwork.ozlabs.org/patch/499482/ https://patchwork.ozlabs.org/patch/452604/ It also happens to break SPI flash on Minnowboard max which is how I noticed that this was applied. I can send a patch to tidy that up, but in any case I think we should consider a revert until the function is better implemented. Signed-off-by: Simon Glass <sjg@chromium.org>
2015-08-14x86: baytrail: Configure FSP UPD from device treeAndrew Bradford
Allow for configuration of FSP UPD from the device tree which will override any settings which the FSP was built with itself. Modify the MinnowMax and BayleyBay boards to transfer sensible UPD settings from the Intel FSPv4 Gold release to the respective dts files, with the condition that the memory-down parameters for MinnowMax are also used. Signed-off-by: Andrew Bradford <andrew.bradford@kodakalaris.com> Reviewed-by: Bin Meng <bmeng.cn@gmail.com> Tested-by: Bin Meng <bmeng.cn@gmail.com> Removed fsp,mrc-debug-msg and fsp,enable-xhci for minnowmax, bayleybay Fixed lines >80col Signed-off-by: Simon Glass <sjg@chromium.org>
2015-08-13spl, common, serial: build SPL without serial supportHeiko Schocher
This patch enables building SPL without CONFIG_SPL_SERIAL_SUPPORT support. Signed-off-by: Heiko Schocher <hs@denx.de> [trini: Ensure we build arch/arm/imx-common on mx28] Signed-off-by: Tom Rini <trini@konsulko.com>
2015-08-08arm: socfpga: misc: Reset ethernet from OFMarek Vasut
Reset the GMAC ethernets based on the "resets" OF node instead of ad-hoc hardcoded values in the U-Boot code. Since we don't have a proper reset framework in place yet, we have to do this slightly ad-hoc parsing of the OF tree instead. Signed-off-by: Marek Vasut <marex@denx.de> Cc: Dinh Nguyen <dinguyen@opensource.altera.com> Cc: Joe Hershberger <joe.hershberger@ni.com>
2015-08-06dts: Drop unused compatible ID for the NXP video bridgeSimon Glass
This has moved to driver model so we can drop the fdtdec support. Signed-off-by: Simon Glass <sjg@chromium.org>
2015-08-06video: Remove the old parade driverSimon Glass
We have a new one which uses driver model and device tree configuration. Remove the old one. Signed-off-by: Simon Glass <sjg@chromium.org>
2015-08-06power: Remove old TPS65090 driversSimon Glass
Remove the old drivers (both the normal one and the cros_ec one) now that we have new drivers that use driver model. Signed-off-by: Simon Glass <sjg@chromium.org>
2015-08-05efi: Add functions for decoding the EFI tablesSimon Glass
The EFI stub can pass a table to U-Boot with information about the memory map Potentially other things will follow. Add a way to access this table. Signed-off-by: Simon Glass <sjg@chromium.org> Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
2015-08-05efi: Add 64-bit payload supportSimon Glass
Most EFI implementations use 64-bit. Add a way to build U-Boot as a 64-bit EFI payload. The payload unpacks a (32-bit) U-Boot and starts it. This can be enabled for x86 boards at present. Signed-off-by: Simon Glass <sjg@chromium.org> Improvements to how the payload is built: Signed-off-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Bin Meng <bmeng.cn@gmail.com> Tested-by: Bin Meng <bmeng.cn@gmail.com>
2015-08-05efi: Add support for loading U-Boot through an EFI stubSimon Glass
It is useful to be able to load U-Boot onto a board even if is it already running EFI. This can allow access to the U-Boot command interface, flexible booting options and easier development. The easiest way to do this is to build U-Boot as a binary blob and have an EFI stub copy it into RAM. Add support for this feature, targeting 32-bit initially. Also add a way to detect when U-Boot has been loaded via a stub. This goes in common.h since it needs to be widely available so that we avoid redoing initialisation that should be skipped. Signed-off-by: Simon Glass <sjg@chromium.org> Improvements to how the payload is built: Signed-off-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Bin Meng <bmeng.cn@gmail.com> Tested-by: Bin Meng <bmeng.cn@gmail.com>
2015-08-05efi: Display the correct initcall pre-relocation valuesSimon Glass
With EFI the start address of U-Boot is specified differently. We could consider just setting GD_FLG_RELOC and then setting up reloc_off. But that flag has other implementations and we are not able to use U-Boot relocation which this flag implies. Instead, just add a special case for EFI. Signed-off-by: Simon Glass <sjg@chromium.org> Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
2015-08-05efi: Add start-up library codeSimon Glass
When running as an EFI application, U-Boot must request memory from EFI, and provide access to the boot services U-Boot needs. Add library code to perform these tasks. This includes efi_main() which is the entry point from EFI. U-Boot is built as a shared library. Signed-off-by: Simon Glass <sjg@chromium.org> Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
2015-08-05lib: fdt: fix indent of #ifdef..#endif conditionalMasahiro Yamada
Match the depth of indentation between #ifdef and #endif for better readability. Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com> Acked-by: Simon Glass <sjg@chromium.org>
2015-07-28ARM: Tegra210: Add support to common Tegra source/config filesTom Warren
Derived from Tegra124, modified as appropriate during T210 board bringup. Cleaned up debug statements to conserve string space, too. This also adds misc 64-bit changes from Thierry Reding/Stephen Warren. Signed-off-by: Tom Warren <twarren@nvidia.com> Signed-off-by: Thierry Reding <treding@nvidia.com> Signed-off-by: Stephen Warren <swarren@nvidia.com>
2015-07-28fdt: Fix fdtdec_get_addr_size() for 64-bitThierry Reding
Signed-off-by: Thierry Reding <treding@nvidia.com> Signed-off-by: Tom Warren <twarren@nvidia.com> Signed-off-by: Stephen Warren <swarren@nvidia.com>
2015-07-22linux_compat: handle __GFP_ZERO in kmalloc()Masahiro Yamada
Currently, kzalloc() returns zero-filled memory, while kmalloc() simply ignores the second argument and never fills the memory area with zeros. I want kmalloc(size, __GFP_ZERO) to behave as kzalloc() does, which will make it easier to add more memory allocator variants. With the introduction of __GFP_ZERO flag, going forward, kzmalloc() variants can fall back to kmalloc() enabling the __GFP_ZERO flag. Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com> Reviewed-by: Heiko Schocher <hs@denx.de> Acked-by: Simon Glass <sjg@chromium.org> Reviewed-by: Lukasz Majewski <l.majewski@samsung.com>
2015-07-22linux_compat: move vzalloc() to header file as an inline functionMasahiro Yamada
The vzalloc(size) is equivalent to kzalloc(size, 0). Move it to include/linux/compat.h as an inline function in order to avoid the function call overhead. Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com> Reviewed-by: Heiko Schocher <hs@denx.de> Acked-by: Simon Glass <sjg@chromium.org>
2015-07-21fdt: Provide debug info when a device tree cannot be foundSimon Glass
It can be quite confusing with a new platform to figure out why the device tree cannot be located. Add some debug information for this case. Signed-off-by: Simon Glass <sjg@chromium.org>
2015-07-21lib: Add function to extract a number from the end of a stringSimon Glass
Split out the code in fdtdec which finds a number at the end of a string. It can be useful in other situations. Signed-off-by: Simon Glass <sjg@chromium.org>
2015-07-21Add rivest cipher 4 (rc4) implementationSimon Glass
Add an implementation of RC4. This will be used by Rockchip booting but may be useful in other situations. Signed-off-by: Simon Glass <sjg@chromium.org>
2015-07-21Drop CONFIG_ERRNO_STR from SPLSimon Glass
This bloats the code size quite a bit and is less useful in SPL where there is no command line. Avoid including this code in SPL. Signed-off-by: Simon Glass <sjg@chromium.org>
2015-07-21fdt: Add fdt_first/next_region() functionsSimon Glass
These have been sent upstream but not accepted to libfdt. For now, bring these into U-Boot to enable fdtgrep to operate. We will use fdtgrep to cut device tree files down for SPL. Signed-off-by: Simon Glass <sjg@chromium.org>
2015-07-21fdt: Add a function to remove unused strings from a device treeSimon Glass
Property names are stored in a string table. When a node property is removed, the string table is not updated since other nodes may have a property with the same name. Thus it is possible for the string table to build up a number of unused strings. Add a function to remove these. This works by building a new device tree from the old one, adding strings one by one as needed. Signed-off-by: Simon Glass <sjg@chromium.org>