On Sat, Jan 14, 2017 at 02:38:02PM +0100, Emmanuel Vadot wrote:
On Wed, 11 Jan 2017 22:08:03 -0700 Simon Glass sjg@chromium.org wrote:
Hi Emmanuel,
On 28 December 2016 at 03:57, Emmanuel Vadot manu@bidouilliste.com wrote:
On Wed, 28 Dec 2016 11:30:10 +0100 Emmanuel Vadot manu@bidouilliste.com wrote:
Hello Simon,
On Fri, 23 Dec 2016 01:39:06 -0700 Simon Glass sjg@chromium.org wrote:
On 18 December 2016 at 19:48, Tom Rini trini@konsulko.com wrote:
On Fri, Nov 25, 2016 at 08:16:01PM -0700, Simon Glass wrote:
> Move sunxi boards to use binman. This involves adding the image definition > to the device tree and using it in the Makefile. > > Signed-off-by: Simon Glass sjg@chromium.org
Reviewed-by: Tom Rini trini@konsulko.com
-- Tom
Applied to u-boot-dm and now in mainline.
Shouldn't all the config for sunxi board defines CONFIG_SPL_OF_PLATDATA=y now so libfdt.so will be built ?
Sorry, I meant the python module.
Well for now an easier solution is to build sandbox_spl once. My plan is to upstream the Python bindings to libfdt, but it's slow going. There should be no major consequence to not having the module. It is just fractionally slower.
Regards, Simon
On FreeBSD we don't have fdtget as part as our base system and it's not available in the ports (well now it is). For now we simply don't depend on CONFIG_SPL_OF_PLATDATA to compile the swig bindings as it's faster to compile/use it than using fdtget. Maybe this patch https://people.freebsd.org/~manu/uboot/2017.01/0001-build-Always-build-the-l... make sense to be upstream in u-boot ? It doesn't take that long to compile the bindings and there is still the check on the presence of the swig binary.
Yes, I think the FreeBSD patch there should be upstream, it seems reasonable enough for portability.