[U-Boot] Non-OMAP/DaVinci ARM SPL? (Was: Re: [PATCH V9 2/4] omap-common: Add NAND SPL linux booting)
On Wed, Dec 7, 2011 at 8:39 AM, Stefano Babic sbabic@denx.de wrote: [snip]
The function you write are in omap-common/spl_nand.c, but they are not specific for OMAP, and any architecture can profit from your work. What about to move common functions away from TI related directory ? I do not know which is the best place, maybe a new directory under the root as spllib ?
There's a lot of thinking and reworking we should do for the SPL code we have today under arch/arm, both in terms of sharing between davinci and OMAP(ish) and non-TI parts. My question is, for non-TI stuff, is anyone using or about to start using this framework?
On 12/07/11 18:00, Tom Rini wrote:
On Wed, Dec 7, 2011 at 8:39 AM, Stefano Babic sbabic@denx.de wrote: [snip]
The function you write are in omap-common/spl_nand.c, but they are not specific for OMAP, and any architecture can profit from your work. What about to move common functions away from TI related directory ? I do not know which is the best place, maybe a new directory under the root as spllib ?
There's a lot of thinking and reworking we should do for the SPL code we have today under arch/arm, both in terms of sharing between davinci and OMAP(ish) and non-TI parts. My question is, for non-TI stuff, is anyone using or about to start using this framework?
I think NVIDIA Tegra can benefit from this. Added Simon Glass to Cc.
Hi,
On Wed, Dec 7, 2011 at 8:17 AM, Igor Grinberg grinberg@compulab.co.il wrote:
On 12/07/11 18:00, Tom Rini wrote:
On Wed, Dec 7, 2011 at 8:39 AM, Stefano Babic sbabic@denx.de wrote: [snip]
The function you write are in omap-common/spl_nand.c, but they are not specific for OMAP, and any architecture can profit from your work. What about to move common functions away from TI related directory ? I do not know which is the best place, maybe a new directory under the root as spllib ?
There's a lot of thinking and reworking we should do for the SPL code we have today under arch/arm, both in terms of sharing between davinci and OMAP(ish) and non-TI parts. My question is, for non-TI stuff, is anyone using or about to start using this framework?
I think NVIDIA Tegra can benefit from this. Added Simon Glass to Cc.
The Tegra boot rom can load the whole U-Boot in from various media types so we haven't had to use SPL to date. I have been trying out a way of moving a good chunk of the code (that is useful to devs but not used in a normal boot) into the back end of the image to load later. But this is just a link trick and not really related to SPL (although I think it might have application if we want to generalise SPL).
Samsung and Freescale use it I think. I agree that SPL code should be as generic as possible.
Regards, Simon
-- Regards, Igor.
On 07/12/2011 17:00, Tom Rini wrote:
On Wed, Dec 7, 2011 at 8:39 AM, Stefano Babic sbabic@denx.de wrote: [snip]
The function you write are in omap-common/spl_nand.c, but they are not specific for OMAP, and any architecture can profit from your work. What about to move common functions away from TI related directory ? I do not know which is the best place, maybe a new directory under the root as spllib ?
There's a lot of thinking and reworking we should do for the SPL code we have today under arch/arm, both in terms of sharing between davinci and OMAP(ish) and non-TI parts. My question is, for non-TI stuff, is anyone using or about to start using this framework?
I am still using with TI, but my goal is to have it available for the IMX SOCs, too. Some of them boot with old spl code (MX3x), and some of them (MX5 and in future MX6) need it and it would be nice to reuse the code.
Stefano
participants (4)
-
Igor Grinberg -
Simon Glass -
Stefano Babic -
Tom Rini