[PATCH] yocto-bsps: u-boot: Search for rootfs partitions on both SD and eMMC

[PATCH] yocto-bsps: u-boot: Search for rootfs partitions on both SD and eMMC

[Kevin Hao]

From: Jussi Laako <jussi.laako@linux.intel.com>

On BeagleBone Black, first attempt to find rootfs from microSD
and then from eMMC if not available.

Fixes: #9176

Signed-off-by: Jussi Laako <jussi.laako@linux.intel.com>
[Kevin: update to 2016.03, also add subject prefix and Upstream-Status]
Signed-off-by: Kevin Hao <kexin.hao@windriver.com>
---
This is based on the patch [1] posted by Jussi Laako. Add the Upstream-Status
tag as required by Saul, also tweak the subject and u-boot version.

[1] https://lists.yoctoproject.org/pipermail/poky/2016-March/010419.html

 ...r-bootable-partitions-on-both-SD-and-eMMC.patch | 42 ++++++++++++++++++++++
 .../recipes-bsp/u-boot/u-boot_2016.03.bbappend     |  3 ++
 2 files changed, 45 insertions(+)
 create mode 100644 meta-yocto-bsp/recipes-bsp/u-boot/u-boot/0001-Search-for-bootable-partitions-on-both-SD-and-eMMC.patch
 create mode 100644 meta-yocto-bsp/recipes-bsp/u-boot/u-boot_2016.03.bbappend

diff --git a/meta-yocto-bsp/recipes-bsp/u-boot/u-boot/0001-Search-for-bootable-partitions-on-both-SD-and-eMMC.patch b/meta-yocto-bsp/recipes-bsp/u-boot/u-boot/0001-Search-for-bootable-partitions-on-both-SD-and-eMMC.patch
new file mode 100644
index 0000000..6c4c59b
--- /dev/null
+++ b/meta-yocto-bsp/recipes-bsp/u-boot/u-boot/0001-Search-for-bootable-partitions-on-both-SD-and-eMMC.patch
@@ -0,0 +1,42 @@
+From 86cc560ff9a6578326532e277d987e24661e5f04 Mon Sep 17 00:00:00 2001
+From: Jussi Laako <jussi.laako@linux.intel.com>
+Date: Fri, 26 Feb 2016 14:58:32 +0200
+Subject: [PATCH] Search for bootable partitions on both SD and eMMC
+
+On BeagleBone Black, first attempt to boot from microSD and then from
+eMMC if not available.
+
+Upstream-Status: Inappropriate [not author]
+
+Signed-off-by: Jussi Laako <jussi.laako@linux.intel.com>
+[Kevin: Add the Upstream-Status tag]
+Signed-off-by: Kevin Hao <kexin.hao@windriver.com>
+---
+ include/configs/ti_armv7_common.h | 11 ++++++++++-
+ 1 file changed, 10 insertions(+), 1 deletion(-)
+
+diff --git a/include/configs/ti_armv7_common.h b/include/configs/ti_armv7_common.h
+index 199612b..ece87ab 100644
+--- a/include/configs/ti_armv7_common.h
++++ b/include/configs/ti_armv7_common.h
+@@ -60,7 +60,16 @@
+ #define DEFAULT_MMC_TI_ARGS \
+ 	"mmcdev=0\0" \
+ 	"mmcrootfstype=ext4 rootwait\0" \
+-	"finduuid=part uuid mmc 0:2 uuid\0" \
++	"findsduuid=part uuid mmc 0:2 sduuid\0" \
++	"findemmcuuid=part uuid mmc 1:2 emmcuuid\0" \
++	"finduuid=if run findsduuid; then "\
++		"echo \"Boot from microSD\"; "\
++		"setenv uuid ${sduuid}; " \
++	"else " \
++		"run findemmcuuid; " \
++		"echo \"Boot from eMMC\"; " \
++		"setenv uuid ${emmcuuid}; " \
++	"fi;\0" \
+ 	"args_mmc=run finduuid;setenv bootargs console=${console} " \
+ 		"${optargs} " \
+ 		"root=PARTUUID=${uuid} rw " \
+-- 
+2.7.0
+
diff --git a/meta-yocto-bsp/recipes-bsp/u-boot/u-boot_2016.03.bbappend b/meta-yocto-bsp/recipes-bsp/u-boot/u-boot_2016.03.bbappend
new file mode 100644
index 0000000..01c8e68
--- /dev/null
+++ b/meta-yocto-bsp/recipes-bsp/u-boot/u-boot_2016.03.bbappend
@@ -0,0 +1,3 @@
+FILESEXTRAPATHS_prepend := "${THISDIR}/${PN}:"
+
+SRC_URI_append = " file://0001-Search-for-bootable-partitions-on-both-SD-and-eMMC.patch"
-- 
2.8.1

Re: [PATCH] yocto-bsps: u-boot: Search for rootfs partitions on both SD and eMMC

[Jussi Kukkonen]

[-- Attachment #1: Type: text/plain, Size: 4042 bytes --]

On 2 June 2016 at 10:51, Kevin Hao <kexin.hao@windriver.com> wrote:
>
> From: Jussi Laako <jussi.laako@linux.intel.com>
>
> On BeagleBone Black, first attempt to find rootfs from microSD
> and then from eMMC if not available.
>
> Fixes: #9176
>
> Signed-off-by: Jussi Laako <jussi.laako@linux.intel.com>
> [Kevin: update to 2016.03, also add subject prefix and Upstream-Status]
> Signed-off-by: Kevin Hao <kexin.hao@windriver.com>
> ---
> This is based on the patch [1] posted by Jussi Laako. Add the
Upstream-Status
> tag as required by Saul, also tweak the subject and u-boot version.
>
> [1] https://lists.yoctoproject.org/pipermail/poky/2016-March/010419.html
>
>  ...r-bootable-partitions-on-both-SD-and-eMMC.patch | 42
++++++++++++++++++++++
>  .../recipes-bsp/u-boot/u-boot_2016.03.bbappend     |  3 ++
>  2 files changed, 45 insertions(+)
>  create mode 100644
meta-yocto-bsp/recipes-bsp/u-boot/u-boot/0001-Search-for-bootable-partitions-on-both-SD-and-eMMC.patch
>  create mode 100644
meta-yocto-bsp/recipes-bsp/u-boot/u-boot_2016.03.bbappend
>
> diff --git
a/meta-yocto-bsp/recipes-bsp/u-boot/u-boot/0001-Search-for-bootable-partitions-on-both-SD-and-eMMC.patch
b/meta-yocto-bsp/recipes-bsp/u-boot/u-boot/0001-Search-for-bootable-partitions-on-both-SD-and-eMMC.patch
> new file mode 100644
> index 0000000..6c4c59b
> --- /dev/null
> +++
b/meta-yocto-bsp/recipes-bsp/u-boot/u-boot/0001-Search-for-bootable-partitions-on-both-SD-and-eMMC.patch
> @@ -0,0 +1,42 @@
> +From 86cc560ff9a6578326532e277d987e24661e5f04 Mon Sep 17 00:00:00 2001
> +From: Jussi Laako <jussi.laako@linux.intel.com>
> +Date: Fri, 26 Feb 2016 14:58:32 +0200
> +Subject: [PATCH] Search for bootable partitions on both SD and eMMC
> +
> +On BeagleBone Black, first attempt to boot from microSD and then from
> +eMMC if not available.
> +
> +Upstream-Status: Inappropriate [not author]
> +
> +Signed-off-by: Jussi Laako <jussi.laako@linux.intel.com>
> +[Kevin: Add the Upstream-Status tag]

I realise there are cases when upstreaming is not a good choice but ... I
don't think this was what Saul was hoping for when he recommended
upstreaming and asked for an Upstream-Status the first time this patch was
sent.

Jussi

> +Signed-off-by: Kevin Hao <kexin.hao@windriver.com>
> +---
> + include/configs/ti_armv7_common.h | 11 ++++++++++-
> + 1 file changed, 10 insertions(+), 1 deletion(-)
> +
> +diff --git a/include/configs/ti_armv7_common.h
b/include/configs/ti_armv7_common.h
> +index 199612b..ece87ab 100644
> +--- a/include/configs/ti_armv7_common.h
> ++++ b/include/configs/ti_armv7_common.h
> +@@ -60,7 +60,16 @@
> + #define DEFAULT_MMC_TI_ARGS \
> +       "mmcdev=0\0" \
> +       "mmcrootfstype=ext4 rootwait\0" \
> +-      "finduuid=part uuid mmc 0:2 uuid\0" \
> ++      "findsduuid=part uuid mmc 0:2 sduuid\0" \
> ++      "findemmcuuid=part uuid mmc 1:2 emmcuuid\0" \
> ++      "finduuid=if run findsduuid; then "\
> ++              "echo \"Boot from microSD\"; "\
> ++              "setenv uuid ${sduuid}; " \
> ++      "else " \
> ++              "run findemmcuuid; " \
> ++              "echo \"Boot from eMMC\"; " \
> ++              "setenv uuid ${emmcuuid}; " \
> ++      "fi;\0" \
> +       "args_mmc=run finduuid;setenv bootargs console=${console} " \
> +               "${optargs} " \
> +               "root=PARTUUID=${uuid} rw " \
> +--
> +2.7.0
> +
> diff --git a/meta-yocto-bsp/recipes-bsp/u-boot/u-boot_2016.03.bbappend
b/meta-yocto-bsp/recipes-bsp/u-boot/u-boot_2016.03.bbappend
> new file mode 100644
> index 0000000..01c8e68
> --- /dev/null
> +++ b/meta-yocto-bsp/recipes-bsp/u-boot/u-boot_2016.03.bbappend
> @@ -0,0 +1,3 @@
> +FILESEXTRAPATHS_prepend := "${THISDIR}/${PN}:"
> +
> +SRC_URI_append = "
file://0001-Search-for-bootable-partitions-on-both-SD-and-eMMC.patch"
> --
> 2.8.1
>
> --
> _______________________________________________
> poky mailing list
> poky@yoctoproject.org
> https://lists.yoctoproject.org/listinfo/poky

[-- Attachment #2: Type: text/html, Size: 5348 bytes --]

Re: [PATCH] yocto-bsps: u-boot: Search for rootfs partitions on both SD and eMMC

[Kevin Hao]

[-- Attachment #1: Type: text/plain, Size: 1263 bytes --]

On Thu, Jun 02, 2016 at 11:57:25AM +0300, Jussi Kukkonen wrote:
> > +Upstream-Status: Inappropriate [not author]
> > +
> > +Signed-off-by: Jussi Laako <jussi.laako@linux.intel.com>
> > +[Kevin: Add the Upstream-Status tag]
> 
> I realise there are cases when upstreaming is not a good choice but ... I don't
> think this was what Saul was hoping for when he recommended upstreaming and
> asked for an Upstream-Status the first time this patch was sent.

I only saw that Saul asked to add the Upstream-Status tag in the patch in order
to be compatible with the yocto requirement. I didn't noticed that he also
asked to submit this patch to upstream.

For upstreaming:
1) This is not a patch from me.
2) The macro DEFAULT_MMC_TI_ARGS is also used by other TI boards. Some boards
   may not have the mmc 1 device at all. So this change makes no sense to
   these boards.
3) We have many ways to overwrite the default value of "finduuid" such as
   (boot.scr, uEnv.txt). The default env for this board is already very
   complex, I don't like to make it even more complex.

Give the above reasons, I am fine to integrate this patch to yocto-bsps for
our convenient. But I will not take the responsibility to upstreaming it.

Thanks,
Kevin

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 484 bytes --]

Re: [PATCH] yocto-bsps: u-boot: Search for rootfs partitions on both SD and eMMC

[Paul Eggleton]

On Thu, 02 Jun 2016 17:40:12 Kevin Hao wrote:
> On Thu, Jun 02, 2016 at 11:57:25AM +0300, Jussi Kukkonen wrote:
> > > +Upstream-Status: Inappropriate [not author]
> > > +
> > > +Signed-off-by: Jussi Laako <jussi.laako@linux.intel.com>
> > > +[Kevin: Add the Upstream-Status tag]
> > 
> > I realise there are cases when upstreaming is not a good choice but ... I
> > don't think this was what Saul was hoping for when he recommended
> > upstreaming and asked for an Upstream-Status the first time this patch
> > was sent.
> 
> I only saw that Saul asked to add the Upstream-Status tag in the patch in
> order to be compatible with the yocto requirement. I didn't noticed that he
> also asked to submit this patch to upstream.
> 
> For upstreaming:
> 1) This is not a patch from me.
> 2) The macro DEFAULT_MMC_TI_ARGS is also used by other TI boards. Some
> boards may not have the mmc 1 device at all. So this change makes no sense
> to these boards.
> 3) We have many ways to overwrite the default value of "finduuid" such as
>    (boot.scr, uEnv.txt). The default env for this board is already very
>    complex, I don't like to make it even more complex.
> 
> Give the above reasons, I am fine to integrate this patch to yocto-bsps for
> our convenient. But I will not take the responsibility to upstreaming it.

I hadn't realised until I checked, but we do actually officially list 
"Inappropriate [not author]" as a reason that can be given, so as things stand 
right now officially there's nothing wrong with using this status.

I don't think that's ideal though. I'm not suggesting requiring submitters to 
take responsibility for upstreaming the patch, but I think we do need a better 
distinction between "this patch isn't appropriate to be sent upstream [for 
this reason relating to the patch itself]" and "this patch looks like it could 
go upstream if someone took the responsibility of doing it in the future". It 
would be nice if when scanning through for patches for upstreaming at least on 
the first pass we could filter out the "Inappropriate" ones without filtering 
out patches that may in fact be appropriate for upstreaming based on the patch 
content alone.

I wonder if it's time to revisit our Upstream-Status values and check if we 
want to do an overhaul - besides the above I've always felt that "Denied" was 
poorly named and I know "Pending" has raised some complaints about being too 
vague also.

Cheers,
Paul

-- 

Paul Eggleton
Intel Open Source Technology Centre