[U-Boot-Users] FW: Uboot-V2: ARM: towards a common start-arm.S

[Peter Pearse]

Copy to list 

> -----Original Message-----
> From: Peter Pearse [mailto:peter.pearse at arm.com] 
> Sent: 18 March 2008 09:32
> To: 'Nishanth Menon'
> Subject: RE: [U-Boot-Users] Uboot-V2: ARM: towards a common 
> start-arm.S
> 
>  
> 
> > -----Original Message-----
> > From: u-boot-users-bounces at lists.sourceforge.net
> > [mailto:u-boot-users-bounces at lists.sourceforge.net] On Behalf Of 
> > Nishanth Menon
> > Sent: 17 March 2008 22:32
> > To: s.hauer at pengutronix.de; u-boot-users at lists.sourceforge.net
> > Subject: [U-Boot-Users] Uboot-V2: ARM: towards a common start-arm.S
> > 
> > Hi Sascha,
> > 
> > I think we dont have a seperate list for uboot-v2, so, just 
> wanted to 
> > know the opinion of having a single 
> arch/arm/cpu/start-arm.S instead 
> > of multiple start-arm files.
> > essentially providing hooks to arch,soc,board specific preinits can 
> > allow us to reuse major chunk of code..
> > I have attached a sample start-arm.S (modified from
> > start-arm920t.S) as reference.
> 
> Great idea.
> 
> Two trivial points.
> 
> a) In the past confusion has arisen because the reset vector 
> is so called. Some users thought the code after reset: 
> caused a reset, rather than running after the cpu has been reset.
> Could it be renamed start_code, as per current arm920t/start.S?
> 
> b) I dont believe that relocating the vector table to the 
> reset address is ARCH or SOC dependent. It depends on 
> TEXT_BASE and whether memory at the reset location is 
> writeable. So it should be in your start.S somewhere, as per 
> current arm920t/start.S.
> 
> Regards
> 
> Peter
>