Re: [Minios-devel] [PATCH ARM v10 2/2] arm: build system

[Wei Liu <wei.liu2@xxxxxxxxxx>]

On Mon, Aug 03, 2015 at 03:05:08PM +0100, Ian Campbell wrote:
> On Thu, 2015-07-23 at 11:08 +0100, Thomas Leonard wrote:
> [...]
> >  .gitmodules          |  3 +++
> [...]
> >  dtc                  |  1 +
> > 
> [...]
> > diff --git a/.gitmodules b/.gitmodules
> > new file mode 100644
> > index 0000000..221102d
> > --- /dev/null
> > +++ b/.gitmodules
> > @@ -0,0 +1,3 @@
> > +[submodule "dtc"]
> > +   path = dtc
> > +   url = git://git.kernel.org/pub/scm/utils/dtc/dtc.git
> [...]
> > diff --git a/dtc b/dtc
> > new file mode 160000
> > index 0000000..64c46b0
> > --- /dev/null
> > +++ b/dtc
> > @@ -0,0 +1 @@
> > +Subproject commit 64c46b098b969502a74c8b0fd97e6f5e4aa07e21
> 
> I was about to apply when I noticed this and I just wanted to raise the use
> of git submodules here explicitly to check folks were OK with it.
> 
> The (or, one) alternative is to import libfdt into the mini-os code based,
> which is how both Xen and Linux have approached this issue for in-kernel
> use of libfdt (Xen still depends on the normal packaged library for
> userspace use).
> 
> Thoughts?

FWIW I prefer importing libfdt into mini-os source tree because of the
issues you mentioned in next paragraph.

But in the end it's mini-os maintainers' call.

Wei.

> 
> I think the addition of the submodule should be called out in the commit
> message. In particular I expect it means that the build rune quoted there
> won't actually work without having done some other stuff first, perhaps
> this is even worthy of mention in the README?
> 


> Ian.

_______________________________________________
Minios-devel mailing list
Minios-devel@xxxxxxxxxxxxxxxxxxxx
http://lists.xenproject.org/cgi-bin/mailman/listinfo/minios-devel