On Wed, Aug 29, 2018 at 12:58 AM Sascha Hauer <s.hauer@xxxxxxxxxxxxxx> wrote:
>
> On Wed, Aug 29, 2018 at 12:21:52AM -0700, Andrey Smirnov wrote:
> > Convert barebox_update_handler_exists() to treat struct bbu_data with
> > both "handler_name" and "devicefile" set to NULL as a specifier for
> > default update handler in order to support such use-case for
> > "barebox_update" and fix a recent regression.
> >
> > This change shouldn't affect another user of
> > barebox_update_handler_exists(), cb_flash() in
> > drivers/usb/gadget/f_fastboot.c, since that function explicitly
> > specifies "devicefile".
> >
> > Fixes 0ac96ab6e ("bbu: command: Make sure specified update handler exists")
> >
> > Signed-off-by: Andrey Smirnov <andrew.smirnov@xxxxxxxxx>
> > ---
> > common/bbu.c | 2 +-
> > 1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > diff --git a/common/bbu.c b/common/bbu.c
> > index 3974bf672..ffe5e15a2 100644
> > --- a/common/bbu.c
> > +++ b/common/bbu.c
> > @@ -148,7 +148,7 @@ bool barebox_update_handler_exists(struct bbu_data *data)
> > if (handler)
> > return true;
> >
> > - if (!data->handler_name)
> > + if (!data->handler_name && data->devicefile)
> > return false;
> >
> > return bbu_find_handler(data->handler_name) != NULL;
>
> This function is confusing. We should split this up into two functions:
> barebox_update_handler_exists_for_name() and
> barebox_update_handler_exists_for_devpath() *). Fastboot would only need
> the latter and the barebox_update command would use both, possibly
> barking when both a name and a device is given (as otherwise we would
> have to check for name <-> devpath conflicts)
>
What if we just expose already existing bbu_find_handler (potentially
renaming it to bbu_find_handler_by_name()) and
bbu_find_handler_by_device() instead and use them as you propose?
Thanks,
Andrey Smirnov
_______________________________________________
barebox mailing list
barebox@xxxxxxxxxxxxxxxxxxx
http://lists.infradead.org/mailman/listinfo/barebox
