Merge branch 'for-next/nv'

command: nv: rewrap help text lines ...

Just let the help text appear a bit more aligned.

Signed-off-by: Enrico Jorns <ejo@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

commands: nv: adjust command help ...

Help did not point out that -r option requires n arguments and did not
mention support for removing variables in help text.

Also introduce and consistently reuse 'NV' as abbreviation for 'non volatile'.

Signed-off-by: Enrico Jorns <ejo@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

commands: nv: assure error code will be returned when an error occurred ...

Due to iteration over possibly multiple variables, the return value of
an individual call to nvvar_remove() / nvvar_add() gets lost.

This will result in do_nv() returning success (0) if any variable
processing failed as long as processing of the last variable succeeded.

Processing will not be aborted on individual errors, such as the 'cp'
handles it for invalid files.

Signed-off-by: Enrico Jorns <ejo@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

commands: nv: call nvvar_print() only if no argument is given ...

Printing the nv variables (before performing any modification!) is useless
and irritating.

Signed-off-by: Enrico Jorns <ejo@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

commands: nv: fail with verbose message if invoked without args ...

Signed-off-by: Enrico Jorns <ejo@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

commands: spi: do not restrict spi's data to be hex-formatted ...

Change the default base to 0 to be consistent with the "mw" command and
particularly allow decimally formatted data.

The price to pay is needing the usual prefix "0x" for each hex-formatted number.

Signed-off-by: Ulrich Ölmann <u.oelmann@pengutronix.de>
Acked-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

commands: i2c: do not restrict i2c_write's data to be hex-formatted ...

Change the default base to 0 to be consistent with the "mw" command and
particularly allow decimally formatted data.

The price to pay is needing the usual prefix "0x" for each hex-formatted number.

Signed-off-by: Ulrich Ölmann <u.oelmann@pengutronix.de>
Acked-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

Merge branch 'for-next/usbgadget'

fbtest: add gradients pattern ...

This pattern draws red, green, blue, and white color gradients, together with
3 anchor rectangles in corners.

To be used with automated screen testing via computer vision methods.

Suggested-by: Chris Healy <cphealy@gmail.com>
Signed-off-by: Nikita Yushchenko <nikita.yoush@cogentembedded.com>
Tested-by: Chris Healy <[1]cphealy@gmail.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

fbtest: add solid pattern ...

This just fills entire screen with single color, as passed in -c
argument in RRGGBB format.

Suggested-by: Chris Healy <cphealy@gmail.com>
Signed-off-by: Nikita Yushchenko <nikita.yoush@cogentembedded.com>
Tested-by: Chris Healy <[1]cphealy@gmail.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

fastboot command: Add -b option to export bbu handlers ...

When -b is given the usbgadget command will automatically export the
registered bbu handlers via fastboot.

Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

usbgadget command: catch errors when parsing the file list ...

file_list_parse() can fail. Print an error message in this case instead
of crashing.

Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

Merge branch 'for-next/state'

state command: extend help part II ...

Since commit 0c74367 the 'state' command
supports authentification and a related parameter. Honor this in the help
text.

Signed-off-by: Juergen Borleis <jbe@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

state command: extend help part I ...

The loading parameter and its help text was removed in commit
446f3d0 and the loading parameter
was re-added in commit 3a5ddf3 but the
help was forgotten. Re-add the help as well.

Signed-off-by: Juergen Borleis <jbe@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

loadxy: add missing ) to help text ...

Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Lucas Stach <l.stach@pengutronix.de>

Merge branch 'for-next/mmc'

mmc_extcsd command: rework ...

This patch is a rework of the mmc_extcsd command:

- Always print registers. Previously we never printed registers that
  were already present in an older version of the spec
- Put register names, access type and width into an array indexed by the
  register number
- Print multibyte registers only once with the resulting value, and not
  bytewise.

There's still more to cleanup, like for example we want to write
multibyte registers once with the complete value, not bytewise. Anyway,
this is a start.

Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Signed-off-by: Lucas Stach <l.stach@pengutronix.de>

mmc_extcsd command: Fix multiline prints ...

The multiline prints lack the indention at the beginning of the second
line. Add these.

Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Signed-off-by: Lucas Stach <l.stach@pengutronix.de>

readlink: Improve -f handling ...

For `readlink -f FILE VARIABLE`, set VARIABLE to the absolute, canonical
file name of FILE by following symbolic links.  All but the final
component of FILE must exist.

Prior to commit a602bebcf7e4 ("fs: Implement links to directories") the
above worked in the limited case where FILE was an absolute path, the
final component was a symbolic link, and all preceding components were
directories.  If FILE was a relative path with the final component being
a symbolic link, and all preceding paths being directories, the command
sort of worked, but produced a relative path as the result.  Both of
these cases were broken by the above commit.

Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Lucas Stach <l.stach@pengutronix.de>

Merge branch 'for-next/misc'

commands: mount: mention file loop mounts in the help text ...

Signed-off-by: Philipp Zabel <p.zabel@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

commands: mount: combine dev and devstr, which are the same ...

Unify the device path used by the default mount and mount with specified
mount point paths.

Signed-off-by: Philipp Zabel <p.zabel@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

true: make true_aliases[] const ...

Although the elements of `true_aliases[]` are of type `const char *`,
the elements themselves are not const-qualified.  Make them `const`.

Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

test: make test_aliases[] const ...

Although the elements of `test_aliases[]` are of type `const char *`,
the elements themselves are not const-qualified.  Make them `const`.

Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

help: make help_aliases[] const ...

Although the elements of `help_aliases[]` are of type `const char *`,
the elements themselves are not const-qualified.  Make them `const`.

Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

edit: make edit_aliases[] const ...

Although the elements of `edit_aliases[]` are of type `const char *`,
the elements themselves are not const-qualified.  Make them `const`.

Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

commands: add ':' as alias for 'true' ...

':' is the null command, sometimes used in shell scripts (especially
after 'then'), but it is not yet supported by barebox.  It is supposed
to do nothing, successfully, just like the 'true' command, so add it as
an alias.

Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

ls: Fix showing links to directories ...

With links to directories we have to do some adjustments in the
printout. In ls_one we have to use lstat() because we want to
show informations about the file or link. When determing if it's
a file or directory that we show we have to use stat() instead.

Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>