Hello Simon,
On Tue, 18 Feb 2014 21:45:12 -0800
Simon Glass <s...@chromium.org> wrote:
> Hi Masahiro,
>
> On 18 February 2014 01:27, Masahiro Yamada <yamad...@jp.panasonic.com> wrote:
> > Hello Simon,
> >
> >
> >
> >> > Useful rules in scripts/Makefile.lib allows us to easily
> >> > generate a device tree blob and wrap it in assembly code.
> >> >
> >> > We do not need to parse a linker script to get output format and arch.
> >> >
> >> > This commit deletes ./u-boot.dtb since it is a copy of dts/dt.dtb
> >>
> >> I'd rather have the former and delete the latter temporary file as
> >> part of the build. u-boot.dtb is currently a build output.
> >
> >
> > This is what I have said in the thread
> > Re: [PATCH] dts: re-write dts/Makefile more simply with Kbuild
> >
> > We must keep dts/dt.dtb to suppress the re-generation of dts/dt.dtb.
>
> OK. From memory when I did this, it was just a temporary file which
> was used to create dts/dt.o. Make deleted it at the end of the build,
> I think. But I may remember wrong, and it doesn't matter anyway.
I think dts/dt.o was deleted at the end
only when CONFIG_OF_SEPARATE=y
by the following rule.
$(obj)u-boot.dtb: checkdtc $(obj)u-boot
$(MAKE) $(build) dts binary
mv $(obj)dts/dt.dtb $@
It moves, not copies dts/dt.dtb.
> >
> > Besides, dts/dt.dtb is a prerequisite of dts/dt.dtb.S
> > when CONFIG_OF_EMBED is enabled.
> >
> > I believe keeping dts/dt.dtb is reasonable enough.
> >
> > Better to keep both?
>
> Yes I think so.
OK.
I will revive ./u-boot.dtb and post a new version.
And I will send it as a single patch
dropping 2/3 and 3/3.
Do you think it's better?
Many kbuild-related patches are being stuck on patchwork
and my local branch is getting messed up.
I don't want to delay this patch any more.
Best Regards
Masahiro Yamada
_______________________________________________
U-Boot mailing list
U-Boot@lists.denx.de
http://lists.denx.de/mailman/listinfo/u-boot
