[U-Boot-Users] [PATCH] Make it possible to use a builtin OF tree.

[U-Boot-Users] [PATCH] Make it possible to use a builtin OF tree.

[Joakim Tjernlund]

Signed-off-by: Joakim Tjernlund <Joakim.Tjernlund@transmode.se>
---
 common/cmd_bootm.c |    5 ++++-
 1 files changed, 4 insertions(+), 1 deletions(-)

diff --git a/common/cmd_bootm.c b/common/cmd_bootm.c
index d816349..8405de7 100644
--- a/common/cmd_bootm.c
+++ b/common/cmd_bootm.c
@@ -507,6 +507,9 @@ fixup_silent_linux ()
 #endif /* CONFIG_SILENT_CONSOLE */
 
 #ifdef CONFIG_PPC
+#ifndef DEFAULT_OF_TREE
+ #define DEFAULT_OF_TREE NULL
+#endif
 static void  __attribute__((noinline))
 do_bootm_linux (cmd_tbl_t *cmdtp, int flag,
 		int	argc, char *argv[],
@@ -527,7 +530,7 @@ do_bootm_linux (cmd_tbl_t *cmdtp, int flag,
 	void	(*kernel)(bd_t *, ulong, ulong, ulong, ulong);
 	image_header_t *hdr = &header;
 #if defined(CONFIG_OF_FLAT_TREE) || defined(CONFIG_OF_LIBFDT)
-	char	*of_flat_tree = NULL;
+	char	*of_flat_tree = DEFAULT_OF_TREE;
 	ulong	of_data = 0;
 #endif
 
-- 
1.5.3.6

[U-Boot-Users] [PATCH] Make it possible to use a builtin OF tree.

[Jerry Van Baren]

Joakim Tjernlund wrote:
> Signed-off-by: Joakim Tjernlund <Joakim.Tjernlund@transmode.se>
> ---
>  common/cmd_bootm.c |    5 ++++-
>  1 files changed, 4 insertions(+), 1 deletions(-)
> 
> diff --git a/common/cmd_bootm.c b/common/cmd_bootm.c
> index d816349..8405de7 100644
> --- a/common/cmd_bootm.c
> +++ b/common/cmd_bootm.c
> @@ -507,6 +507,9 @@ fixup_silent_linux ()
>  #endif /* CONFIG_SILENT_CONSOLE */
>  
>  #ifdef CONFIG_PPC
> +#ifndef DEFAULT_OF_TREE
> + #define DEFAULT_OF_TREE NULL
> +#endif
>  static void  __attribute__((noinline))
>  do_bootm_linux (cmd_tbl_t *cmdtp, int flag,
>  		int	argc, char *argv[],
> @@ -527,7 +530,7 @@ do_bootm_linux (cmd_tbl_t *cmdtp, int flag,
>  	void	(*kernel)(bd_t *, ulong, ulong, ulong, ulong);
>  	image_header_t *hdr = &header;
>  #if defined(CONFIG_OF_FLAT_TREE) || defined(CONFIG_OF_LIBFDT)
> -	char	*of_flat_tree = NULL;
> +	char	*of_flat_tree = DEFAULT_OF_TREE;
>  	ulong	of_data = 0;
>  #endif

The loss was most likely my error/misunderstanding way back at the start 
of LIBFDT.  Your change seems reasonable, I'll add it to my list of 
patches to look into and add.

Thanks,
gvb

[U-Boot-Users] [PATCH] Make it possible to use a builtin OF tree.

[gvb.uboot]

Jerry Van Baren wrote:
> Joakim Tjernlund wrote:
>> Signed-off-by: Joakim Tjernlund <Joakim.Tjernlund@transmode.se>
>> ---
>>  common/cmd_bootm.c |    5 ++++-
>>  1 files changed, 4 insertions(+), 1 deletions(-)
>>
>> diff --git a/common/cmd_bootm.c b/common/cmd_bootm.c
>> index d816349..8405de7 100644
>> --- a/common/cmd_bootm.c
>> +++ b/common/cmd_bootm.c
>> @@ -507,6 +507,9 @@ fixup_silent_linux ()
>>  #endif /* CONFIG_SILENT_CONSOLE */
>>  
>>  #ifdef CONFIG_PPC
>> +#ifndef DEFAULT_OF_TREE
>> + #define DEFAULT_OF_TREE NULL
>> +#endif
>>  static void  __attribute__((noinline))
>>  do_bootm_linux (cmd_tbl_t *cmdtp, int flag,
>>  		int	argc, char *argv[],
>> @@ -527,7 +530,7 @@ do_bootm_linux (cmd_tbl_t *cmdtp, int flag,
>>  	void	(*kernel)(bd_t *, ulong, ulong, ulong, ulong);
>>  	image_header_t *hdr = &header;
>>  #if defined(CONFIG_OF_FLAT_TREE) || defined(CONFIG_OF_LIBFDT)
>> -	char	*of_flat_tree = NULL;
>> +	char	*of_flat_tree = DEFAULT_OF_TREE;
>>  	ulong	of_data = 0;
>>  #endif
> 
> The loss was most likely my error/misunderstanding way back at the start 
> of LIBFDT.  Your change seems reasonable, I'll add it to my list of 
> patches to look into and add.
> 
> Thanks,
> gvb

Hi Joakim,

I looked for the definition of DEFAULT_OF_TREE and it doesn't exist in 
the current u-boot.  I used the gitweb "pickaxe" to look for 
DEFAULT_OF_TREE and didn't find it that way either.  Can you identify 
when we lost the definition?

The problem with your patch is that, if DEFAULT_OF_TREE isn't defined by 
somebody, the builds will break.  Since it is undefined at the moment, 
that means *all* of the OF_FLAT_TREE and OF_LIBFDT builds would break. :-(

Thanks,
gvb

[U-Boot-Users] [PATCH] Make it possible to use a builtin OF tree.

[Grant Likely]

On 11/25/07, Joakim Tjernlund <Joakim.Tjernlund@transmode.se> wrote:
>
> Signed-off-by: Joakim Tjernlund <Joakim.Tjernlund@transmode.se>
> ---
>  common/cmd_bootm.c |    5 ++++-
>  1 files changed, 4 insertions(+), 1 deletions(-)
>
> diff --git a/common/cmd_bootm.c b/common/cmd_bootm.c
> index d816349..8405de7 100644
> --- a/common/cmd_bootm.c
> +++ b/common/cmd_bootm.c
> @@ -507,6 +507,9 @@ fixup_silent_linux ()
>  #endif /* CONFIG_SILENT_CONSOLE */
>
>  #ifdef CONFIG_PPC
> +#ifndef DEFAULT_OF_TREE
> + #define DEFAULT_OF_TREE NULL
> +#endif

Where is DEFAULT_OF_TREE intended to be defined?  I don't want to get
into defining it in any of the include/configs/*.h files since the
goal is to move over to a Kconfig system and eliminate the per-board
config files entirely.

It might be better to do this with two linker symbols around a .dtb
section.  If the two symbols are identical, then no dtb was provided.
Otherwise, go ahead and use the section as the default dtb.

Cheers,
g.

-- 
Grant Likely, B.Sc., P.Eng.
Secret Lab Technologies Ltd.
grant.likely at secretlab.ca
(403) 399-0195

[U-Boot-Users] [PATCH] Make it possible to use a builtin OF tree.

[Joakim Tjernlund]

On Mon, 2007-11-26 at 21:32 -0500, gvb.uboot wrote:
> Jerry Van Baren wrote:
> > Joakim Tjernlund wrote:
> >> Signed-off-by: Joakim Tjernlund <Joakim.Tjernlund@transmode.se>
> >> ---
> >>  common/cmd_bootm.c |    5 ++++-
> >>  1 files changed, 4 insertions(+), 1 deletions(-)
> >>
> >> diff --git a/common/cmd_bootm.c b/common/cmd_bootm.c
> >> index d816349..8405de7 100644
> >> --- a/common/cmd_bootm.c
> >> +++ b/common/cmd_bootm.c
> >> @@ -507,6 +507,9 @@ fixup_silent_linux ()
> >>  #endif /* CONFIG_SILENT_CONSOLE */
> >>  
> >>  #ifdef CONFIG_PPC
> >> +#ifndef DEFAULT_OF_TREE
> >> + #define DEFAULT_OF_TREE NULL
> >> +#endif
> >>  static void  __attribute__((noinline))
> >>  do_bootm_linux (cmd_tbl_t *cmdtp, int flag,
> >>  		int	argc, char *argv[],
> >> @@ -527,7 +530,7 @@ do_bootm_linux (cmd_tbl_t *cmdtp, int flag,
> >>  	void	(*kernel)(bd_t *, ulong, ulong, ulong, ulong);
> >>  	image_header_t *hdr = &header;
> >>  #if defined(CONFIG_OF_FLAT_TREE) || defined(CONFIG_OF_LIBFDT)
> >> -	char	*of_flat_tree = NULL;
> >> +	char	*of_flat_tree = DEFAULT_OF_TREE;
> >>  	ulong	of_data = 0;
> >>  #endif
> > 
> > The loss was most likely my error/misunderstanding way back at the start 
> > of LIBFDT.  Your change seems reasonable, I'll add it to my list of 
> > patches to look into and add.
> > 
> > Thanks,
> > gvb
> 
> Hi Joakim,
> 
> I looked for the definition of DEFAULT_OF_TREE and it doesn't exist in 
> the current u-boot.  I used the gitweb "pickaxe" to look for 
> DEFAULT_OF_TREE and didn't find it that way either.  Can you identify 
> when we lost the definition?

Not really, it was possible in u-boot 1.1.4 and now it is not.

> 
> The problem with your patch is that, if DEFAULT_OF_TREE isn't defined by 
> somebody, the builds will break.  Since it is undefined at the moment, 
> that means *all* of the OF_FLAT_TREE and OF_LIBFDT builds would break. :-(

ehh, doesn't the 
#ifndef DEFAULT_OF_TREE
 #define DEFAULT_OF_TREE NULL
#endif
take care of that?

 Jocke

[U-Boot-Users] [PATCH] Make it possible to use a builtin OF tree.

[Joakim Tjernlund]

On Mon, 2007-11-26 at 20:06 -0700, Grant Likely wrote:
> On 11/25/07, Joakim Tjernlund <Joakim.Tjernlund@transmode.se> wrote:
> >
> > Signed-off-by: Joakim Tjernlund <Joakim.Tjernlund@transmode.se>
> > ---
> >  common/cmd_bootm.c |    5 ++++-
> >  1 files changed, 4 insertions(+), 1 deletions(-)
> >
> > diff --git a/common/cmd_bootm.c b/common/cmd_bootm.c
> > index d816349..8405de7 100644
> > --- a/common/cmd_bootm.c
> > +++ b/common/cmd_bootm.c
> > @@ -507,6 +507,9 @@ fixup_silent_linux ()
> >  #endif /* CONFIG_SILENT_CONSOLE */
> >
> >  #ifdef CONFIG_PPC
> > +#ifndef DEFAULT_OF_TREE
> > + #define DEFAULT_OF_TREE NULL
> > +#endif
> 
> Where is DEFAULT_OF_TREE intended to be defined?  I don't want to get
> into defining it in any of the include/configs/*.h files since the
> goal is to move over to a Kconfig system and eliminate the per-board
> config files entirely.

You define it to whatever you like in your board config file. Can be a
char array or a function returning a char ptr, sorry :(

> 
> It might be better to do this with two linker symbols around a .dtb
> section.  If the two symbols are identical, then no dtb was provided.
> Otherwise, go ahead and use the section as the default dtb.

I feel my current way is more flexible as it allows you build/modify the
tree in a function.

 Jocke

[U-Boot-Users] [PATCH] Make it possible to use a builtin OF tree.

[Kumar Gala]

>>> The loss was most likely my error/misunderstanding way back at the  
>>> start
>>> of LIBFDT.  Your change seems reasonable, I'll add it to my list of
>>> patches to look into and add.
>>>
>>> Thanks,
>>> gvb
>>
>> Hi Joakim,
>>
>> I looked for the definition of DEFAULT_OF_TREE and it doesn't exist  
>> in
>> the current u-boot.  I used the gitweb "pickaxe" to look for
>> DEFAULT_OF_TREE and didn't find it that way either.  Can you identify
>> when we lost the definition?
>
> Not really, it was possible in u-boot 1.1.4 and now it is not.

This was dropped because Wolfgang was exist embedding device tree's in  
the u-boot source tree.

- k

[U-Boot-Users] [PATCH] Make it possible to use a builtin OF tree.

[Joakim Tjernlund]

On Tue, 2007-11-27 at 01:54 -0600, Kumar Gala wrote:
> >>> The loss was most likely my error/misunderstanding way back at the  
> >>> start
> >>> of LIBFDT.  Your change seems reasonable, I'll add it to my list of
> >>> patches to look into and add.
> >>>
> >>> Thanks,
> >>> gvb
> >>
> >> Hi Joakim,
> >>
> >> I looked for the definition of DEFAULT_OF_TREE and it doesn't exist  
> >> in
> >> the current u-boot.  I used the gitweb "pickaxe" to look for
> >> DEFAULT_OF_TREE and didn't find it that way either.  Can you identify
> >> when we lost the definition?
> >
> > Not really, it was possible in u-boot 1.1.4 and now it is not.
> 
> This was dropped because Wolfgang was exist embedding device tree's in  
> the u-boot source tree.

hmm, I got a vague memory about that too. As I recall Wolfgang didn't
motivate why that did make sense to me. Wolfgang, can you recap.?

 Jocke