/dev/disk0 vs /dev/mmc0

[Jean-Christophe PLAGNIOL-VILLARD]

On 12:39 Sun 06 Oct     , Sascha Hauer wrote:
> On Fri, Oct 04, 2013 at 09:17:39AM +0200, David Jander wrote:
> > 
> > Hi Jean-Christophe,
> > 
> > On Thu, 3 Oct 2013 21:23:49 +0200
> > Jean-Christophe PLAGNIOL-VILLARD <plagnioj at jcrosoft.com> wrote:
> > 
> > > On 17:17 Thu 03 Oct     , David Jander wrote:
> > > > 
> > > > Hi all,
> > > > 
> > > > I am following barebox git closely and noticed a change recently: Device
> > > > names for MMC (MCI) and USB mass-storage devices have changed to the
> > > > generic "/dev/diskX". Earlier an MMC device was named "/dev/mmc0".
> > > > Unfortunately this change breaks my /env/bin/init script and I don't know
> > > > how to fix it. I relied on the existence of certain devices to distinguish
> > > > between USB mass-storage device presence and/or SD-card presence. How can
> > > > I do this with this new device naming convention?
> > > > 
> > > > I used to have these kind of checks in /env/bin/init:
> > > > 
> > > > # Mount MMC (first partition) if available
> > > > if [ -e "/dev/mmc0.0" ]; then
> > > >         mkdir /mmc
> > > >         mount /dev/mmc0.0 /mmc
> > > >         if [ -e "/mmc/uImage" ]; then
> > > >                 boot_target="mmc"
> > > >         fi
> > > > fi
> > > > 
> > > > Booting like this is only allowed from MMC and not from USB, and now it
> > > > seems impossible to distinguish between them anymore.
> > > > 
> > > > Btw, why was this changed in the first place?
> > > 
> > > now you an use devname parameter to specify a specifc name for mmc
> > 
> > My board is fully device-tree based, but your suggestion pointed me in the
> > right direction: the "alias" section seems to have a whole new meaning now ;-)
> > I hope this won't interfere with Linux, since I intend to use a single DT for
> > both Barebox _and_ Linux.
> 
> It doesn't interfere with the kernel. The kernel currently ignores this
> aliases. There are patches floating to let the kernel honor this
> aliases, but then they should simply have the same effect as they have
> in barebox.
> 
> > 
> > > not all the drivers have the platform_data update to support but it will be
> > > easy enough to add it
> > > 
> > > I recently add this to atmel_mci and animeo_ip board
> > 
> > Didn't find that one... probably not in mainline yet? Never mind.
> > 
> > > and I recomment you to switch to defaultenv-2 for this
> > > 
> > > and use boot sequence this will simplify your env hugely
> > 
> > Although I agree defaultenv-2 looks pretty cool and will add a lot of
> > convenience and flexibility for complex systems, IMHO it will not simplify the
> > env of this board the least bit :-)
> > In fact, on a simple embedded system that has to have a strictly predefined
> > boot-behavior with a few simple rules, it is overly complex and if an
> > "outsider" ever needs to understand how it works, that can be a bit of a
> > problem (the default boot path probably going through some 5 different
> > scripts with lots of compound config-variables).
> 
> At least I recently replaced the boot script with a command (not yet
> mainline), so there are fewer scripts involved.

and I hope we can kill the old defaultenv soon
for my board I just need to resend the bootp support

Best Regards,
J.
> 
> Sascha
> 
> -- 
> Pengutronix e.K.                           |                             |
> Industrial Linux Solutions                 | http://www.pengutronix.de/  |
> Peiner Str. 6-8, 31137 Hildesheim, Germany | Phone: +49-5121-206917-0    |
> Amtsgericht Hildesheim, HRA 2686           | Fax:   +49-5121-206917-5555 |