[PATCH 0/3] of: dts: enable memory at 0 quirk for PPC32 only
leif.lindholm at linaro.org
Fri Apr 18 22:48:22 EST 2014
On Thu, Apr 17, 2014 at 07:43:13PM -0500, Rob Herring wrote:
> On Thu, Apr 17, 2014 at 12:41 PM, Leif Lindholm
> <leif.lindholm at linaro.org> wrote:
> > drivers/of/fdt.c contains a workaround for a missing memory type
> > entry on longtrail firmware. Make that quirk PPC32 only, and while
> > at it - fix up the .dts files in the tree currently working only
> > because of that quirk.
> But why do you need this?
Apart from the current code permitting recreating a 15+ year old
firmware bug into completely new platform ports?
Because the UEFI stub for arm/arm64 needs to delete all of the "memory"
nodes from the DT. And it would be nice to at least not have to compile
the "and also delete anything called memory at 0" into the arm64 image. Or
any image not including support for affected platforms.
> > arch/arm/boot/dts/ste-ccu8540.dts | 1 +
> > arch/mips/lantiq/dts/easy50712.dts | 1 +
> > arch/mips/ralink/dts/mt7620a_eval.dts | 1 +
> > arch/mips/ralink/dts/rt2880_eval.dts | 1 +
> > arch/mips/ralink/dts/rt3052_eval.dts | 1 +
> > arch/mips/ralink/dts/rt3883_eval.dts | 1 +
> I'm not worried about these MIPS dts files because they are all
> built-in, but you are breaking compatibility with old dtbs for this
> ARM board.
Yeah, sorry. Sending out a v2 of part 3 shortly.
More information about the Linuxppc-dev