[U-Boot] [PATCH 0/3] Fixes for edb9315

[Matthias Kaehlcke <matthias@kaehlcke.net>]

Hi Alessandro,

El Sat, Feb 06, 2010 at 08:53:30PM +0100 Alessandro Rubini ha dit:

> I'm porting current u-boot to a board similar to EP9315A, so I'm using
> arm/master as a basis, as it includes the patches by Matthias
> Kaehlcke.  I'm currently running from RAM (SKIP_LOWLEVEL_INIT), after
> setting up sdram and pll elsewhere (older vendor u-boot code, still to
> be ported).
> 
> "fix syscon_regs definition" is needed to access any software-locked
> syscon register from C code (e.g., in reset_cpu() and devicecfg used
> by me in patch 3).

you are right, i missed the gap between ChipID and SysCfg

> "change calculation un early_udelay.h" is needed at least for eldk-4.2
> (gcc-4.2.2), as without this patch it will use the stack before setting SP.
> Actually, we could use the ether buffer as a stack, if needed, but it's not
> really needed here.

i feared this could happen with different compiler versions, your fix
certainly reduces the risk

> "enable the uart in devicecfg register" prevents u-boot from freezing
> at least with SKIP_LOWLEVEL_INIT set, but I'm pretty sure lowlevel_setup
> assembly code doesn't enable the uart, either.

my ep9301 and ep9307 based boards for some reason boot with the uart
enabled, though according to the datasheet it is disabled until enabled
explicitly. therefore your fix is necessary for all cpus behaving as
expected.

thanks a lot for your patches!

-- 
Matthias Kaehlcke
Embedded Linux Developer
Barcelona

      The assumption that what currently exists must necessarily
        exist is the acid that corrodes all visionary thinking
                                                                 .''`.
    using free software / Debian GNU/Linux | http://debian.org  : :'  :
                                                                `. `'`
gpg --keyserver pgp.mit.edu --recv-keys 47D8E5D4                  `-