This enables configs necessary for using SPI. The environment is saved to
the very end of SPI flash. This is unlikely to be overwritten unless the
entire flash is reprogrammed.

This also supplies a default bootcommand. It loads an image and device tree
from the first partition of the MMC. This is a minimal/least effort
bootcmd, so suggestions (especially in the form of patches) are welcome. I
didn't set up distro boot because I think it is unlikely that any
general-purpose linux distros will ever be ported to this board.
Signed-off-by: NSean Anderson <seanga2@gmail.com>
Tested-by: NChrstopher Obbard <obbardc@gmail.com>
Reviewed-by: NRick Chen <rick@andestech.com>

Latest opensbi uses generic platform for Qemu. Update the build
instructions.
Signed-off-by: NAtish Patra <atish.patra@wdc.com>
Reviewed-by: NBin Meng <bin.meng@windriver.com>
Reviewed-by: NRick Chen <rick@andestech.com>

- Add MBR partition layout writing supporting, clean up code.
- A large number of assorted console/iomux cleanups
- A large number of board_r related cleanups.
- Log enhancements

Add a 'mbr' command to let users create or verify MBR partition layout
based on the provided text description. The partition layout is
alternatively read from the 'mbr_parts' environment variable. This can be
used in scripts to help system image flashing tools to ensure proper
partition layout.

The syntax of the text description of the partition list is similar to
the one used by the 'gpt' command. Supported parameters are: name
(currently ignored), start (partition start offset in bytes), size (in
bytes or '-' to expand it to the whole free area), bootable (boolean
flag) and id (MBR partition type). If one wants to create more than 4
partitions, an 'Extended' primary partition (with 0x05 ID) has to be
explicitely provided as a one of the first 4 entries.

Here is an example how to create a 6 partitions (3 on the 'extended
volume'), some of the predefined sizes:

> setenv mbr_parts 'name=boot,start=4M,size=128M,bootable,id=0x0e;
  name=rootfs,size=3072M,id=0x83;
  name=system-data,size=512M,id=0x83;
  name=[ext],size=-,id=0x05;
  name=user,size=-,id=0x83;
  name=modules,size=100M,id=0x83;
  name=ramdisk,size=8M,id=0x83'
> mbr write mmc 0
Signed-off-by: NMarek Szyprowski <m.szyprowski@samsung.com>

Add a code for creating and writing MBR partition layout. The code generates
similar layout of EBRs (Exteneded Block Records) and logical volumes as
Linux's fdisk utility.
Signed-off-by: NMarek Szyprowski <m.szyprowski@samsung.com>

Make functions not used outside this file static.
Signed-off-by: NMarek Szyprowski <m.szyprowski@samsung.com>
[trini: Use __maybe_unused as there are cases they may not be used]
Signed-off-by: NTom Rini <trini@konsulko.com>

Use a generic helper for reading LE32 integers.
Signed-off-by: NMarek Szyprowski <m.szyprowski@samsung.com>

Add some handy defines for some hardcoded magic numbers related to
extended partition handling.
Signed-off-by: NMarek Szyprowski <m.szyprowski@samsung.com>

write_mbr_partition() function name is a bit misleading, so rename it to
write_mbr_sector(). This is a preparation for adding code for writing a
complete MBR partition layout.
Signed-off-by: NMarek Szyprowski <m.szyprowski@samsung.com>

When at some point environment shrinks we need to stop dropped devices.
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Reviewed-by: NSimon Glass <sjg@chromium.org>

Obviously the following has unnecessary indentation level in 'else' branch.

	if (foo) {
		...
		return;
	} else {
		...
	}

Drop indentation level by removing redundant 'else'.
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Reviewed-by: NSimon Glass <sjg@chromium.org>

Refactor iomux_doenv() a bit in order to increase readability.
There is no change in code generation on x86.
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Reviewed-by: NSimon Glass <sjg@chromium.org>

It's realloc() 101 to avoid `foo = realloc(foo, ...);` call
due to getting a memory leak.

Actually it's not clear why realloc() has been used here.
If we shrink the array, the memcpy() overwrites it anyway
with the contents of a new array. If it becomes bigger,
same story.

Drop useless realloc() for good and thus preserve console list
in case of failed allocation.

Fixes: 16a28ef2 ("IOMUX: Add console multiplexing support.")
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Reviewed-by: NSimon Glass <sjg@chromium.org>

Provide a documentation for console_search_dev().
Suggested-by: NSimon Glass <sjg@chromium.org>
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>

Rename search_device() to console_search_dev() since it's in console.h.
Suggested-by: NSimon Glass <sjg@chromium.org>
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Reviewed-by: NSimon Glass <sjg@chromium.org>

search_device() is defined in console.c. Move its declaration
to an appropriate header file.
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Reviewed-by: NSimon Glass <sjg@chromium.org>

There is no need to call ->start() for already started device. All the same,
there is no need to call ->stop() for devices still in use.

For now enforce this only for IOMUX case.

Cc: Simon Glass <sjg@chromium.org>
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Reviewed-by: NSimon Glass <sjg@chromium.org>

In the future we would like to stop unused consoles and
also add a reference counting to avoid imbalanced calls
to ->start() and ->stop() in some cases.
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Reviewed-by: NSimon Glass <sjg@chromium.org>

Add the helper function console_has_tstc() and replace the test
#if CONFIG_IS_ENABLED(CONSOLE_MUX) to a simple 'if' test to
respect the U-Boot coding rule.

No functional change.
Signed-off-by: NPatrick Delaunay <patrick.delaunay@st.com>
Signed-off-by: NPatrick Delaunay <patrick.delaunay@foss.st.com>
Reviewed-by: NSimon Glass <sjg@chromium.org>

Add helper functions to access to gd->console_out and gd->console_in
with membuff API and replace the #ifdef CONFIG_CONSOLE_RECORD test
by if (IS_ENABLED(CONFIG_CONSOLE_RECORD)) to respect the U-Boot
coding rule.
Signed-off-by: NPatrick Delaunay <patrick.delaunay@st.com>
Signed-off-by: NPatrick Delaunay <patrick.delaunay@foss.st.com>
Reviewed-by: NSimon Glass <sjg@chromium.org>

Add a new function to access to console_devices only defined if
CONFIG_IS_ENABLED(CONSOLE_MUX).

This path allows to remove #if CONFIG_IS_ENABLED(CONSOLE_MUX)
in console_getc function.
Signed-off-by: NPatrick Delaunay <patrick.delaunay@st.com>
Reviewed-by: NSimon Glass <sjg@chromium.org>
Signed-off-by: NPatrick Delaunay <patrick.delaunay@foss.st.com>

Remove #ifdef or #ifndef for CONFIG when it is possible to simplify
the console.c code and respect the U-Boot coding rules.
Signed-off-by: NPatrick Delaunay <patrick.delaunay@st.com>
Signed-off-by: NPatrick Delaunay <patrick.delaunay@foss.st.com>
Reviewed-by: NSimon Glass <sjg@chromium.org>

spl_handoff should only be enabled when CONFIG_HANDOFF is set. Drop the
nested ifdefs and check for CONFIG_HANDOFF instead.
Signed-off-by: NOvidiu Panait <ovidiu.panait@windriver.com>
Reviewed-by: NSimon Glass <sjg@chromium.org>

CONFIG_HANDOFF is used in u-boot proper to locate handoff info from SPL
during pre-relocation init (in setup_spl_handoff). Add explicit dependency
on CONFIG_SPL, to fix the following build error when CONFIG_HANDOFF &&
!CONFIG_SPL:

common/board_f.c: In function ‘setup_spl_handoff’:
common/board_f.c:283:4: error: ‘gd_t {aka struct global_data}’
has no member named ‘spl_handoff’
  gd->spl_handoff = bloblist_find(BLOBLISTT_SPL_HANDOFF,
    ^~
Signed-off-by: NOvidiu Panait <ovidiu.panait@windriver.com>
Reviewed-by: NSimon Glass <sjg@chromium.org>

In order to remove the arch-specific ifdefs around initr_trap, introduce
arch_initr_trap weak initcall. Implementations for ppc/m68k/mips have
been moved to arch/<arch>/lib/traps.c

Default implementation is a nop stub.
Signed-off-by: NOvidiu Panait <ovidiu.panait@windriver.com>
Reviewed-by: NSimon Glass <sjg@chromium.org>
Reviewed-by: NDaniel Schwierzeck <daniel.schwierzeck@gmail.com>

Add a return value to bb_miiphy_init and use it directly in the
post-relocation init sequence, rather than using a wrapper stub.
Signed-off-by: NOvidiu Panait <ovidiu.panait@windriver.com>
Reviewed-by: NSimon Glass <sjg@chromium.org>

Add a return value to api_init and use it directly in the
post-relocation init sequence, rather than using a wrapper stub.
Signed-off-by: NOvidiu Panait <ovidiu.panait@windriver.com>
Reviewed-by: NSimon Glass <sjg@chromium.org>

Add a return value to jumptable_init and use it directly in the
post-relocation init sequence, rather than using a wrapper stub.
Signed-off-by: NOvidiu Panait <ovidiu.panait@windriver.com>
Reviewed-by: NSimon Glass <sjg@chromium.org>

Add a return value to xen_init and use it directly in the
post-relocation init sequence, rather than using a wrapper stub.
Signed-off-by: NOvidiu Panait <ovidiu.panait@windriver.com>
Reviewed-by: NSimon Glass <sjg@chromium.org>

Add a return value to noncached_init and use it directly in the
post-relocation init sequence, rather than using a wrapper stub.
Signed-off-by: NOvidiu Panait <ovidiu.panait@windriver.com>
Reviewed-by: NSimon Glass <sjg@chromium.org>

Add a return value to pci_init and use it directly in the post-relocation
init sequence, rather than using a wrapper stub.
Signed-off-by: NOvidiu Panait <ovidiu.panait@windriver.com>
Reviewed-by: NSimon Glass <sjg@chromium.org>

Add a return value to pci_ep_init and use it directly in the
post-relocation init sequence, rather than using a wrapper stub.
Signed-off-by: NOvidiu Panait <ovidiu.panait@windriver.com>
Reviewed-by: NSimon Glass <sjg@chromium.org>

Add a return value to post_output_backlog and use it directly in the
post-relocation init sequence, rather than using a wrapper stub.
Signed-off-by: NOvidiu Panait <ovidiu.panait@windriver.com>
Reviewed-by: NSimon Glass <sjg@chromium.org>

Add a return value to cpu_secondary_init_r and use it directly in the
post-relocation init sequence, rather than using a wrapper stub.
Signed-off-by: NOvidiu Panait <ovidiu.panait@windriver.com>
Reviewed-by: NSimon Glass <sjg@chromium.org>

Drop initr_console_record wrapper and call console_record_init directly.
Signed-off-by: NOvidiu Panait <ovidiu.panait@windriver.com>
Reviewed-by: NSimon Glass <sjg@chromium.org>

Use IS_ENABLED(CONFIG_OF_EMBED) in instead of #ifdefs in reserve_fdt,
reloc_fdt functions.
Signed-off-by: NOvidiu Panait <ovidiu.panait@windriver.com>
Reviewed-by: NSimon Glass <sjg@chromium.org>

setup_bdinfo is used to populate various bdinfo fields, so move
setup_machine code there, as all it does is setting
gd->bd->bi_arch_number.
Signed-off-by: NOvidiu Panait <ovidiu.panait@windriver.com>
Reviewed-by: NSimon Glass <sjg@chromium.org>

Use IS_ENABLED(CONFIG_TIMER_EARLY) instead of #ifdef in initf_dm. Also,
move timer code to the main ifdef, so that ret is defined.
Signed-off-by: NOvidiu Panait <ovidiu.panait@windriver.com>
Reviewed-by: NSimon Glass <sjg@chromium.org>

Drop initf_console_record wrapper and call console_record_init directly.
Signed-off-by: NOvidiu Panait <ovidiu.panait@windriver.com>
Reviewed-by: NSimon Glass <sjg@chromium.org>

Currently, the following #ifdef construct is used to check whether to run
console_record_init() during pre-relocation init:
 defined(CONFIG_CONSOLE_RECORD) && CONFIG_VAL(SYS_MALLOC_F_LEN)

Introduce CONFIG_CONSOLE_RECORD_INIT_F Kconfig option to get rid of the
complex ifdef check. Also, use IS_ENABLED() instead of #ifdef.
Signed-off-by: NOvidiu Panait <ovidiu.panait@windriver.com>
Reviewed-by: NSimon Glass <sjg@chromium.org>