[PATCH 11/11] bootm: Support string substitution in bootargs

[Simon Glass]

Hi Rasmus,

On Mon, 19 Oct 2020 at 08:43, Rasmus Villemoes
<rasmus.villemoes at prevas.dk> wrote:
>
> On 19/10/2020 15.56, Simon Glass wrote:
> > In some cases it is necessary to pass parameters to Linux so that it will
> > boot correctly. For example, the rootdev parameter is often used to
> > specify the root device. However the root device may change depending on
> > whence U-Boot loads the kernel. At present it is necessary to build up
> > the command line by adding device strings to it one by one.
> >
> > It is often more convenient to provide a template for bootargs, with
> > U-Boot doing the substitution from other environment variables.
> >
> > Add a way to substitute strings in the bootargs variable. This allows
> > things like "rootdev=%U" to be used in bootargs, with the %U substitution
> > providing the UUID of the root device.
> >
> > For example, to substitute the GUID of the kernel partition:
> >
> >   setenv bootargs "console=/dev/ttyS0 rootdev=%U/PARTNROFF=1 kern_guid=%U"
> >   part uuid mmc 2:2 uuid
> >   setenv bootargs_U $uuid
> >   bootm
> >
> > This is particularly useful when the command line from another place. For
> > example, Chrome OS stores the command line next to the kernel itself. It
> > depends on the kernel version being used as well as the hardware features,
> > so it is extremely difficult to devise a U-Boot script that works on all
> > boards and kernel versions. With this feature, the command line can be
> > read from disk and used directly, with a few substitutions set up.
> >
> > Signed-off-by: Simon Glass <sjg at chromium.org>
> > ---
> >
> >  README              |  16 +++++++
> >  arch/Kconfig        |   1 +
> >  common/Kconfig.boot |  20 ++++++++
> >  common/bootm.c      |  72 +++++++++++++++++++++++++++--
> >  include/bootm.h     |  14 ++++--
> >  test/bootm.c        | 110 ++++++++++++++++++++++++++++++++++++++++++--
> >  6 files changed, 221 insertions(+), 12 deletions(-)
> >
> > diff --git a/README b/README
> > index 91c5a1a8fa3..263e31ab7f6 100644
> > --- a/README
> > +++ b/README
> > @@ -3229,6 +3229,22 @@ List of environment variables (most likely not complete):
> >
> >    bootargs   - Boot arguments when booting an RTOS image
> >
> > +  bootargs_subst - Substitution parameters for bootargs. These are applied after
> > +               the commandline has been built. The format is:
> > +
> > +                 <key>=<value>[!<key>=<value>...]
> > +
> > +               where
> > +                 <key> is a string to replace
> > +                 <value> is the value to replace it with (either a simple
> > +                    string or an environment variable starting with $
> > +
> > +               One use for this is to insert the root-disk UUID into the
> > +               command line where bootargs contains "root=%U"
> > +
> > +                 part uuid mmc 2:2 uuid
> > +                 setenv cmdline_repl %U=$uuid
>
> cmdline_repl seems to be stale, it should be bootargs_subst. But the
> whole paragraph seems stale, as you actually implement and test the
> bootargs_X approach.

Yes, thanks, will fix.

>
> Anyway, this does seem useful, but I really question the choice of
> leaving %A in there if bootargs_A does not exist. It's hard if not
> impossible to create an environment variable whose value is empty, and I
> can easily imagine it would be useful to allow a %A to expand to
> nothing. So why not just use the usual semantics of requiring a double
> %% to put a single % in the output? It's probably quite rare that one
> would need that anyway.

I did wonder about the empty env var thing. IMO it would be nice to
support empty variables, so we can distinguish between an empty one
and a missing one.

My concern with removing the var is if people have % in the string.
This way I don't have to worry about quoting, etc. See Wolgang's email
in this thread too.

Regards,
Simon