[U-Boot] [PATCH 23/25] sunxi: Use mmc_bootdev=2 for MMC2 boot

From: Maxime Ripard <maxime.ripard@bootlin.com>
To: u-boot@lists.denx.de
Subject: [U-Boot] [PATCH 23/25] sunxi: Use mmc_bootdev=2 for MMC2 boot
Date: Tue, 17 Jul 2018 13:55:28 +0200	[thread overview]
Message-ID: <20180717115528.yiv7jltag76nh3gx@flea.home> (raw)
In-Reply-To: <CAD6G_RRuPqXxqCDz5ZzQHFpaiGVzUvmGwzmN_9G-XVYcT-WZpA@mail.gmail.com>

On Mon, Jul 16, 2018 at 03:51:33PM +0530, Jagan Teki wrote:
> On Mon, Jul 16, 2018 at 3:18 PM, Maxime Ripard
> <maxime.ripard@bootlin.com> wrote:
> > On Mon, Jul 16, 2018 at 01:49:54PM +0530, Jagan Teki wrote:
> >> With DM_MMC the mmc devices are probed as per dt status
> >> and eMMC can probed maximum device of 2, if all nodes like
> >> mmc0, mmc1 and mmc2 status are 'okay'.
> >>
> >> So update mmc_bootdev to 2 and add boot order as 2, 1, 0
> >> devices. so-that it can to boot any identified device even
> >> if the respective device status disabled or unsupported.
> >>
> >> Signed-off-by: Jagan Teki <jagan@amarulasolutions.com>
> >> ---
> >>  board/sunxi/board.c            | 2 +-
> >>  include/configs/sunxi-common.h | 4 +++-
> >>  2 files changed, 4 insertions(+), 2 deletions(-)
> >>
> >> diff --git a/board/sunxi/board.c b/board/sunxi/board.c
> >> index 5ed1b8bae1..7edc468185 100644
> >> --- a/board/sunxi/board.c
> >> +++ b/board/sunxi/board.c
> >> @@ -768,7 +768,7 @@ int misc_init_r(void)
> >>       } else if (boot == BOOT_DEVICE_MMC1) {
> >>               env_set("mmc_bootdev", "0");
> >>       } else if (boot == BOOT_DEVICE_MMC2) {
> >> -             env_set("mmc_bootdev", "1");
> >> +             env_set("mmc_bootdev", "2");
> >
> > Isn't that broken too? I guess we should return MMC3 in that case.
> 
> Yes, A10 has maximum of mmc3 will fix.
> 
> >
> > Also, what about bisectability?
> 
> are you referring order check while booting, ie reason I've update the
> ordering with device 2, 1, 0

No, I'm referring to the fact that each and every commit of your serie
should be bootable and fully working without any regression. Selecting
DM_MMC and then fixing the IDs in that case breaks that assumption.

Maxime

-- 
Maxime Ripard, Bootlin (formerly Free Electrons)
Embedded Linux and Kernel engineering
https://bootlin.com
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: not available
URL: <http://lists.denx.de/pipermail/u-boot/attachments/20180717/53d1cee1/attachment.sig>