Make a few minor changes to make it easier to add driver-model support.

Signed-off-by: Simon Glass <sjg@chromium.org>
Tested-by: Stephen Warren <swarren@nvidia.com>

Add a uclass for block devices. These provide block-oriented data access,
supporting reading, writing and erasing of whole blocks.

Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Tested-by: Stephen Warren <swarren@nvidia.com>

Adjust a few things so that the addition of driver-models support involved
adding code rather than also changing it. This makes the patches easier to
review.

Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Marek Vasut <marex@denx.de>
Tested-by: Stephen Warren <swarren@nvidia.com>

The limit on storage devices is USB_MAX_STOR_DEV but we use one extra
element while probing to see if a device is a storage device. Avoid this,
since it causes memory corruption.

Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Marek Vasut <marex@denx.de>
Tested-by: Stephen Warren <swarren@nvidia.com>

To ease conversion to driver model, add helper functions which deal with
calling each block device method. With driver model we can reimplement these
functions with the same arguments.

Use inline functions to avoid increasing code size on some boards.

Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Tested-by: Stephen Warren <swarren@nvidia.com>

This is a device number, and we want to use 'dev' to mean a driver model
device. Rename the member.

Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Tested-by: Stephen Warren <swarren@nvidia.com>

Enable these two filesystems to provide better build coverage in sandbox.

Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Tested-by: Stephen Warren <swarren@nvidia.com>

The comment for file_cbfs_type() says that it returns 0 for an invalid type.
The code appears to check for -1, except that it uses an unsigned variable
to store the type. This results in a warning on 64-bit machines.

Adjust it to make the meaning clearer. Continue to handle the -1 case since
it may be needed.

Signed-off-by: Simon Glass <sjg@chromium.org>
Tested-by: Stephen Warren <swarren@nvidia.com>

Rename three partition functions so that they start with part_. This makes
it clear what they relate to.

Signed-off-by: Simon Glass <sjg@chromium.org>
Tested-by: Stephen Warren <swarren@nvidia.com>

We can use linker lists instead of explicitly declaring each function.
This makes the code shorter by avoiding switch() statements and lots of
header file declarations.

While this does clean up the code it introduces a few code issues with SPL.
SPL never needs to print partition information since this all happens from
commands. SPL mostly doesn't need to obtain information about a partition
either, except in a few cases. Add these cases so that the code will be
dropped from each partition driver when not needed. This avoids code bloat.

I think this is still a win, since it is not a bad thing to be explicit
about which features are used in SPL. But others may like to weigh in.

Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Tom Rini <trini@konsulko.com>
Tested-by: Stephen Warren <swarren@nvidia.com>

It is useful to have sandbox build as much code as possible to avoid having
to build every board to detect build errors. Also we may add tests for some
more partition types at some point.

Enable all partition types in sandbox.

Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Tested-by: Stephen Warren <swarren@nvidia.com>

In part_amiga.c the name is unsigned but bcpl_strcpy() requires a signed
pointer. Add a cast to fix the warning.

Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Tested-by: Stephen Warren <swarren@nvidia.com>

Rename this function to blk_get_device_part_str(). This is a better name
because it makes it clear that the function returns a block device and
parses a string.

Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Tested-by: Stephen Warren <swarren@nvidia.com>

The current name is too generic. The function returns a block device based
on a provided string. Rename it to aid searching and make its purpose
clearer. Also add a few comments.

Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Tested-by: Stephen Warren <swarren@nvidia.com>

The current name is too generic. Add a 'blk_' prefix to aid searching and
make its purpose clearer.

Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Tested-by: Stephen Warren <swarren@nvidia.com>

The block interface is not well documented in the code. Pick two important
functions and add comments.

Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Tested-by: Stephen Warren <swarren@nvidia.com>

Since these are sequentially numbered it makes sense to use an enum. It
avoids having to maintain the maximum value, and provides a type we can use
if it is useful.

In fact the maximum value is not used. Rename it to COUNT, since MAX suggests
it is the maximum valid value, but it is not.

Signed-off-by: Simon Glass <sjg@chromium.org>
Tested-by: Stephen Warren <swarren@nvidia.com>

At present block devices are tied up with partitions. But not all block
devices have partitions within them. They are in fact separate concepts.

Create a separate blk.h header file for block devices.

Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Tested-by: Stephen Warren <swarren@nvidia.com>

We should not include <common.h> in header files. Each C file should include
it if needed.

Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Tested-by: Stephen Warren <swarren@nvidia.com>

Adjust the cast to avoid a warning when stdint.h is used.

Signed-off-by: Simon Glass <sjg@chromium.org>
Tested-by: Stephen Warren <swarren@nvidia.com>

Adjust the cast to avoid a warning when stdint.h is used.

Signed-off-by: Simon Glass <sjg@chromium.org>
Tested-by: Stephen Warren <swarren@nvidia.com>

Each region is displayed in almost the same way. Break out this common code
into its own function.

Signed-off-by: Simon Glass <sjg@chromium.org>
Tested-by: Stephen Warren <swarren@nvidia.com>

Use 'struct' instead of a typdef. Also since 'struct block_dev_desc' is long
and causes 80-column violations, rename it to struct blk_desc.

Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Tested-by: Stephen Warren <swarren@nvidia.com>

The serial output from the debug UART carries on going far to the
right in the console.

Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Reviewed-by: Stefan Roese <sr@denx.de>
Reviewed-by: Simon Glass <sjg@chromium.org>

Simple MFD devices can bind children without special bus configuration.
Like Linux, let's handle "simple-mfd" in the same way as "simple-bus".

Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Acked-by: Simon Glass <sjg@chromium.org>

Use this new function in places where it simplifies the code.

Signed-off-by: Simon Glass <sjg@chromium.org>

A common pattern is to call uclass_first_device() and then check if it
actually returns a device. Add a new function which does this, returning
an error if there are no devices in that uclass.

Signed-off-by: Simon Glass <sjg@chromium.org>

Signed-off-by: Tom Rini <trini@konsulko.com>

GPIO4_21 is the LAN8720 power pin, not the LAN8720 reset pin.

Fix that, so that we can have Ethernet functional again.

Reviewed-by: Tom Rini <trini@konsulko.com>
Signed-off-by: Fabio Estevam <fabio.estevam@nxp.com>

The stm_is_locked_sr() function is picked from Linux kernel. For reason
unknown, the 64bit data types used by the function and present in Linux
were replaced with 32bit unsigned ones, which causes trouble.

The testcase performed was done using ST M25P80 chip.
The command used was:
 => sf protect unlock 0 0x10000

The call chain starts in stm_unlock(), which calls stm_is_locked_sr()
with negative ofs argument. This works fine in Linux, where the "ofs"
is loff_t, which is signed long long, while this fails in U-Boot, where
"ofs" is u32 (unsigned int). Because of this signedness problem, the
expression past the return statement to be incorrectly evaluated to 1,
which in turn propagates back to stm_unlock() and results in -EINVAL.

The correction is very simple, just use the correctly sized data types
with correct signedness in the function to make it work as intended.

Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Simon Glass <sjg@chromium.org>
Reviewed-by: Jagan Teki <jteki@openedev.com>

TI QSPI driver directly typecasts fdt_addr_t to a pointer. This is
not strictly correct, as it gives a build warning when fdt_addr_t is u64.
So, use map_physmem for a proper typecasts.

This is inspired by commit 167efe01 ("dm: ns16550: Use an address
instead of a pointer for the uart base")

Signed-off-by: Lokesh Vutla <lokeshvutla@ti.com>
Reviewed-by: Jagan Teki <jteki@openedev.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Mugunthan V N <mugunthanvnm@ti.com>

Enable CONFIG_USB_ETHER_RTL8152 support for Odroid XU4 which
has support for RTL8153-CG gigabit Ethernet adapter,
connected over USB 3.0.

commit 9dc8ba19 added support
for Realtek 8152/8153 driver.

Signed-off-by: Anand Moon <linux.amoon@gmail.com>
Signed-off-by: Minkyu Kang <mk7.kang@samsung.com>

The DMC driver in v3.14 kernel[0] get the ddr setting from PMU_SYS_REG2,
and it expects uboot to store the value using a same protocol. But now
the ddr setting value is different with DMC, so if you enable the DMC,
system would crash in kernel. Correct the sdram setting here, according
to the requirements of kernel.

[0]
https://chromium.googlesource.com/chromiumos/third_party/kernel/+/


chromeos-3.14/drivers/clk/rockchip/clk-rk3288-dmc.c

Signed-off-by: Chris Zhong <zyw@rock-chips.com>
Acked-by: Simon Glass <sjg@chromium.org>

on v2016.03-rc3, size of SPL image compiled by gcc 5.3.0 is too large for
Firefly-RK3288. (it's fine for Rock2)

$ gcc --version
gcc (Ubuntu/Linaro 5.3.0-3ubuntu1~14.04) 5.3.0 20151204
Copyright (C) 2015 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

$ ./tools/mkimage -n rk3288 -T rksd -d spl/u-boot-spl-dtb.bin u-boot-spl-dtb.img
Warning: SPL image is too large (size 0x80d0) and will not boot

to reduce size of SPL image, this patch makes configure_emmc() empty for
Firefly-RK3288 as same as Rock2.

Signed-off-by: FUKAUMI Naoki <naobsd@gmail.com>
Acked-by: Simon Glass <sjg@chromium.org>
Tested-By: Vagrant Cascadian <vagrant@debian.org>

emac may use dpll as clock parent, and it request the clock frequency
multiples of 50, so change ddr frequency to 400M.

Signed-off-by: Lin Huang <hl@rock-chips.com>
Signed-off-by: Jeffy Chen <jeffy.chen@rock-chips.com>
Acked-by: Simon Glass <sjg@chromium.org>

MIPS EL boards should define CONFIG_USE_PRIVATE_LIBGCC=y to work
with EB-only toolchains like the one from kernel.org. If one do
not globally set CONFIG_USE_PRIVATE_LIBGCC=y, the build fails with:

/opt/gcc-4.9.0-nolibc/mips-linux/bin/mips-linux-ld.bfd: /opt/gcc-4.9.0-nolibc/mips-linux/bin/../lib/gcc/mips-linux/4.9.0/libgcc.a(_lshrdi3.o): compiled for a big endian system and target is little endian
/opt/gcc-4.9.0-nolibc/mips-linux/bin/mips-linux-ld.bfd: /opt/gcc-4.9.0-nolibc/mips-linux/bin/../lib/gcc/mips-linux/4.9.0/libgcc.a(_lshrdi3.o): endianness incompatible with that of the selected emulation
/opt/gcc-4.9.0-nolibc/mips-linux/bin/mips-linux-ld.bfd: failed to merge target specific data of file /opt/gcc-4.9.0-nolibc/mips-linux/bin/../lib/gcc/mips-linux/4.9.0/libgcc.a(_lshrdi3.o)
/opt/gcc-4.9.0-nolibc/mips-linux/bin/mips-linux-ld.bfd: /opt/gcc-4.9.0-nolibc/mips-linux/bin/../lib/gcc/mips-linux/4.9.0/libgcc.a(_ashldi3.o): compiled for a big endian system and target is little endian
/opt/gcc-4.9.0-nolibc/mips-linux/bin/mips-linux-ld.bfd: /opt/gcc-4.9.0-nolibc/mips-linux/bin/../lib/gcc/mips-linux/4.9.0/libgcc.a(_ashldi3.o): endianness incompatible with that of the selected emulation
/opt/gcc-4.9.0-nolibc/mips-linux/bin/mips-linux-ld.bfd: failed to merge target specific data of file /opt/gcc-4.9.0-nolibc/mips-linux/bin/../lib/gcc/mips-linux/4.9.0/libgcc.a(_ashldi3.o)
/work/git-trees/u-boot-mips/Makefile:1171: recipe for target 'u-boot' failed

One example for a failing build is Travis CI.

Signed-off-by: Daniel Schwierzeck <daniel.schwierzeck@gmail.com>
Reviewed-by: Purna Chandra Mandal <purna.mandal@microchip.com>

The "R" constraint supplies the address of an variable in a register. Use
"r" instead and adjust asm to supply the content of addr in a register
instead.

Fixes: 2b8bcc5a ("MIPS: avoid .set ISA for cache operations")
Signed-off-by: Matthias Schiffer <mschiffer@universe-factory.net>
Cc: Paul Burton <paul.burton@imgtec.com>
Cc: Daniel Schwierzeck <daniel.schwierzeck@gmail.com>

Following the previous patch, malloc() is never called before gd is set,
so we can remove the special-case check for this condition.

This reverts commit 854d2b97 "dlmalloc: ensure gd is set for early
alloc".

Cc: Rabin Vincent <rabin@rab.in>
Signed-off-by: Stephen Warren <swarren@wwwdotorg.org>
Reviewed-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Simon Glass <sjg@chromium.org>

When running sandbox, the following phases occur, each with different
malloc implementations or behaviors:

1) Dynamic linker execution, using the dynamic linker's own malloc()
implementation. This is fully functional.

2) After U-Boot's malloc symbol has been hooked into the GOT, but before
any U-Boot code has run. This phase is entirely non-functional, since
U-Boot's gd symbol is NULL and U-Boot's initf_malloc() and
mem_malloc_init() have not been called.

At least on Ubuntu Xenial, the dynamic linker does make both malloc() and
free() calls during this phase. Currently these free() calls crash since
they dereference gd, which is NULL.

U-Boot itself makes no use of malloc() during this phase.

3) U-Boot execution after gd is set and initf_malloc() has been called.
This is fully functional, albeit via a very simple malloc()
implementation.

4) U-Boot execution after mem_malloc_init() has been called. This is fully
functional with a complete malloc() implementation.

Furthermore, if code that called malloc() during phase 1 calls free() in
phase 3 or later, it is likely that heap corruption will occur, since
U-Boot's malloc implementation will assume the pointer is part of its own
heap, although it isn't. I have not actively observed this happening.

To prevent phase 2 from happening, this patch makes all of U-Boot's malloc
library public symbols have hidden visibility. This prevents them from
being hooked into the GOT, so only code in the U-Boot binary itself
actually calls them; any other code will call into the standard C library
malloc(). This also avoids the "furthermore" issue mentioned above.

I have seen references to this GCC pragma in blog posts from 2008, and
RHEL5's ancient gcc appears to accept it fine, so I believe it's quite
safe to use it without checking gcc version.

Cc: Rabin Vincent <rabin@rab.in>
Signed-off-by: Stephen Warren <swarren@wwwdotorg.org>
Reviewed-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Simon Glass <sjg@chromium.org>