Hello,
On 04/30/2011 09:56 AM, Wolfgang Denk wrote:
Dear Valentin Longchamp,
In message 08249e40b548fff1e636cecf980e11adabda14ac.1302272395.git.valentin.longchamp@keymile.com you wrote:
From: Holger Brunck holger.brunck@keymile.com
To be prepared for mgcoge3ne which has a different SDRAM on board. The config was moved from generic code to board specific header.
Signed-off-by: Holger Brunck holger.brunck@keymile.com Acked-by: Heiko Schocher hs@denx.de cc: Wolfgang Denk wd@denx.de cc: Detlev Zundel dzu@denx.de cc: Valentin Longchamp valentin.longchamp@keymile.com Signed-off-by: Valentin Longchamp valentin.longchamp@keymile.com
include/configs/km82xx-common.h | 26 -------------------------- include/configs/mgcoge.h | 26 ++++++++++++++++++++++++++ include/configs/mgcoge2ne.h | 23 +++++++++++++++++++++++ 3 files changed, 49 insertions(+), 26 deletions(-)
If I'm understanding this right, we now duplicate the definitions into two files, because two boards happen to use the same SDRAM. Would it
Yes.
not make more sense to move this into a single copy of a SDRAM specific header, which then gets included by all boards that use this SDRAM type? This way we can avoid multiple copied of the same stuff (which always is a maintenance nightmare) for popular RAM types.
Yes this sounds reasonable. But later on in the patch serie we replace the mgcoge2ne support with the support for mgcoge3ne which has a different SDRAM. And therefore this common file would become obsolete. I try to rebase these patches that the mgcoge3ne support is also part of this patch.
BTW: What would your proposal for a filename for a SDRAM specific header?
Best regards Holger Brunck