Hi Tom,

On Wed, 26 Feb 2025 at 14:13, Tom Rini <tr...@konsulko.com> wrote:
>
> On Wed, Feb 26, 2025 at 09:26:15AM -0700, Simon Glass wrote:
>
> > Snow requires a fixed bloblist to operate, so re-enable this option.
> >
> > Signed-off-by: Simon Glass <s...@chromium.org>
> > Fixes: 864106f3c47 ("bloblist: Make BLOBLIST_ALLOC the default")
> > ---
> >
> > (no changes since v1)
> >
> >  configs/snow_defconfig | 2 ++
> >  1 file changed, 2 insertions(+)
> >
> > diff --git a/configs/snow_defconfig b/configs/snow_defconfig
> > index 7a1dc80b1cc..f9fd76daa67 100644
> > --- a/configs/snow_defconfig
> > +++ b/configs/snow_defconfig
> > @@ -33,7 +33,9 @@ CONFIG_DISTRO_DEFAULTS=y
> >  CONFIG_SYS_PBSIZE=1024
> >  CONFIG_SILENT_CONSOLE=y
> >  CONFIG_BLOBLIST=y
> > +CONFIG_BLOBLIST_FIXED=y
> >  # CONFIG_SPL_BLOBLIST is not set
> > +CONFIG_BLOBLIST_ADDR=0x43d00000
> >  # CONFIG_SPL_FRAMEWORK is not set
> >  CONFIG_SPL_FOOTPRINT_LIMIT=y
> >  CONFIG_SPL_MAX_FOOTPRINT=0x3800
>
> I'm not going to NAK the patch, but:
> - "snow" does not require a bloblist to operate, just to run the
>   bloblist tests on hardware

That's right. 'snow' is the platform I chose to try to run more of the
sandbox tests, although I haven't done much with that.

>   as implied by d985f1dbddb2 ("bootstd: Add a
>   test for the bootstd menu"), which also don't seem to run anyhow?

The bootstd tests all need sandbox at present, since there are images
required. If you were interested in my idea of building the
'image-setup' stuff into the pytests, then we could change that.

> - We could just as easily continue to use an allocated bloblist if
>   CONFIG_SYS_MALLOC_F_LEN is increased to say 0x800 from 0x400 (which is
>   the same size as the requested bloblist allocation).

OK

> - This very tiny SYS_MALLOC_F_LEN is just the old default value from
>   when it was bumped up to 0x2000 for most platforms and is also safe
>   from a quick test.

OK. I didn't get time to look at bloblist last weekend and this one is
not looking good either, but I will get to it.

Regards,
Simon

Reply via email to