On Tue, Oct 2, 2018 at 1:11 AM AKASHI, Takahiro
<takahiro.akashi@xxxxxxxxxx> wrote:
>
> Rob,
>
> On Fri, Sep 28, 2018 at 07:49:13AM -0500, Rob Herring wrote:
> > On Wed, Sep 26, 2018 at 12:57 AM AKASHI Takahiro
> > <takahiro.akashi@xxxxxxxxxx> wrote:
> > >
> > > There are two files that are part of dtc/libfdt, but don't appear
> > > in lib. This patch fixes it.
> > > Please note that strtoul() used in fdt_overlay.c is now faked using
> > > kstrtoul() (not simple_strtoul() which is obsolute).
> > >
> > > Signed-off-by: AKASHI Takahiro <takahiro.akashi@xxxxxxxxxx>
> > > Cc: Rob Herring <robh+dt@xxxxxxxxxx>
> > > Cc: Frank Rowand <frowand.list@xxxxxxxxx>
> > > ---
> > > lib/Makefile | 2 +-
> > > lib/fdt_addresses.c | 3 +++
> > > lib/fdt_overlay.c | 23 +++++++++++++++++++++++
> > > 3 files changed, 27 insertions(+), 1 deletion(-)
> > > create mode 100644 lib/fdt_addresses.c
> > > create mode 100644 lib/fdt_overlay.c
> > >
> > > diff --git a/lib/Makefile b/lib/Makefile
> > > index ca3f7ebb900d..444c82413a3c 100644
> > > --- a/lib/Makefile
> > > +++ b/lib/Makefile
> > > @@ -202,7 +202,7 @@ KASAN_SANITIZE_stackdepot.o := n
> > > KCOV_INSTRUMENT_stackdepot.o := n
> > >
> > > libfdt_files = fdt.o fdt_ro.o fdt_wip.o fdt_rw.o fdt_sw.o fdt_strerror.o \
> > > - fdt_empty_tree.o
> > > + fdt_empty_tree.o fdt_addresses.o fdt_overlay.o
> >
> > Nothing in the kernel needs fdt_overlay.o, so we shouldn't add it
> > until we do. It used to bloat the kernel size, but maybe that changed
> > with using archive files now.
> >
> > > $(foreach file, $(libfdt_files), \
> > > $(eval CFLAGS_$(file) = -I$(src)/../scripts/dtc/libfdt))
> > > lib-$(CONFIG_LIBFDT) += $(libfdt_files)
> > > diff --git a/lib/fdt_addresses.c b/lib/fdt_addresses.c
> > > new file mode 100644
> > > index 000000000000..241780d09882
> > > --- /dev/null
> > > +++ b/lib/fdt_addresses.c
> > > @@ -0,0 +1,3 @@
> > > +// SPDX-License-Identifier: GPL-2.0
> > > +#include <linux/libfdt_env.h>
> > > +#include "../scripts/dtc/libfdt/fdt_addresses.c"
> > > diff --git a/lib/fdt_overlay.c b/lib/fdt_overlay.c
> > > new file mode 100644
> > > index 000000000000..1def7268d313
> > > --- /dev/null
> > > +++ b/lib/fdt_overlay.c
> > > @@ -0,0 +1,23 @@
> > > +// SPDX-License-Identifier: GPL-2.0
> > > +#include <linux/kernel.h>
> > > +#include <linux/libfdt_env.h>
> > > +
> > > +static unsigned long strtoul(const char *nptr, char **endptr, int base)
> > > +{
> > > + unsigned long res;
> > > + int ret;
> > > +
> > > + /*
> > > + * FIXME: no way to return a correct value of endptr.
> > > + * The values here would be still good for use in fdt_overlay
> > > + */
> > > + *endptr = (char *)nptr;
> >
> > endptr could be NULL.
>
> No, it couldn't at overlay_fixup_phandle(), which is the only caller
> of strtoul() in fdt_overlay.c, is concerned.
The definition of strtoul allows endptr to be NULL, so your
implementation should support that. Otherwise, when a new user is
added we won't find it until run-time.
> > > + ret = kstrtoul(nptr, base, &res);
> >
> > kstrtoul doesn't handle conversions if there are non-convertible
> > characters after the number while strtoul (including simple_strtoul)
> > will still do the conversion.
>
> I don't believe so.
The same problem exists here. IMO, we should use simple_strtoul
because kstrtoul is not a compatible implementation of strtoul.
Or libfdt upstream should add a strtoul wrapper that has the same
traits as kstrtoul.
> > We should make sure we don't need that
> > behavior. That's doubtful if callers pass non-NULL endptr.
> >
> > > + if (ret < 0)
> > > + return ULONG_MAX;
> > > +
> > > + (*endptr)++;
> > > + return res;
> > > +}
> > > +
> > > +#include "../scripts/dtc/libfdt/fdt_overlay.c"
> > > --
> > > 2.18.0
> > >
>
> Having said that, I would simply drop 'fdt_overlay' part from my patch
> as it won't be used anywhere at this stage.
> Are you happy with this change?
That's fine by me.
Rob
