[bug#51663] [PATCH] gnu: u-boot: Fix rk3399 boot from emmc.

[Pierre Langlois <pierre.langlois@gmx.com>]

[-- Attachment #1: Type: text/plain, Size: 2010 bytes --]

Hi,

Vagrant Cascadian <vagrant@debian.org> writes:

> [[PGP Signed Part:Undecided]]
> On 2021-11-07, Pierre Langlois wrote:
>> I'm afraid the last u-boot update broke booting the rockpro64 and
>> pinebook-pro from the eMMC card :-/.  Going through the u-boot ML, I saw
>> that it was reported [0], and this fix [1] should be applied.
>>
>> So here's a patch!  I tested it on on both the pinebook-pro and
>> rockpro64.
>>
>> OK to apply?
>>
>> Thanks,
>> Pierre
>>
>> [0]: https://lists.denx.de/pipermail/u-boot/2021-November/466329.html
>> [1]: https://patchwork.ozlabs.org/project/uboot/patch/20211101044347.17822-1-yifeng.zhao@rock-chips.com/
> ...
>> diff --git a/gnu/packages/patches/u-boot-rk3399-enable-emmc-phy.patch b/gnu/packages/patches/u-boot-rk3399-enable-emmc-phy.patch
>> new file mode 100644
>> index 0000000000..f14a9ce104
>> --- /dev/null
>> +++ b/gnu/packages/patches/u-boot-rk3399-enable-emmc-phy.patch
>> @@ -0,0 +1,26 @@
>> +adapting commit ac804143cf ("mmc: rockchip_sdhci: add phy and clock
>> +config for rk3399") to fix the issue "Not found emmc phy device".
>
> I'm a little confused about this comment in the patch... is ac804143cf
> the commit which fixed the issue or broke it? is there a reference to an
> upstream commit and/or issue that fixes it?

Yeah it's not very clear to me either, I think ac804143cf is the commit
that introduced the regression indeed. Doing a bit of digging, I don't
think the fix was merged upstream yet though :-(.

Instead, I see Fedora opted to revert the problematic patch, along with
another dependent one. Probably before the fix was posted to the list:

  https://src.fedoraproject.org/rpms/uboot-tools/c/37df227bc0961f9f0dc4dafa9e983290dbdb2bc3
  https://bugzilla.redhat.com/show_bug.cgi?id=2014182#c3

Given it's not upstream, it could be safer to do what Fedora did and
revert a couple of patches, what do you think? The fix is from the
original author so I'm inclined to trust it, I'd be surprised if it
doesn't get merged.

Thanks,
Pierre

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 519 bytes --]