mmc: Read eMMC partition access bits before card reset

[Tom Rini]

On Mon, Jun 12, 2023 at 07:39:20PM +0200, Pali Rohár wrote:
> On Monday 12 June 2023 11:22:12 Stefan Roese wrote:
> > Hi Pali,
> > 
> > (added Peng for PowerPC and Jaehoon for MMC to Cc)
> > 
> > On 6/11/23 15:03, Pali Rohár wrote:
> > 
> > <snip>
> > 
> > > > Pali, I will try to pull these not applied patches - if there are not
> > > > issues of course. Could you please send me a list of these patches so
> > > > that I can take look? Please note that I will probably not be able to
> > > > work on this this week. But next week should be possible.
> > > > 
> > > > Thanks,
> > > > Stefan
> > > 
> > > Hello!
> > > 
> > > These patches are on the list for year:
> > > https://patchwork.ozlabs.org/project/uboot/patch/20220516090119.20217-1-pali@kernel.org/
> > 
> > This patch fails with multiple problems. Reason is, that it's very "old"
> > and many of the Kconfig stuff was changed, which is of course not your
> > fault. Still, I'm not in the right position to make these changes I'm
> > afraid.
> > 
> > Peng, this patch is now assigned to you in patchwork. Could you please
> > also take a look? And if Pali (or someone else) sends an updated
> > version, could you please make sure that it's reviewed (and applied if
> > no issues arise) in due time?
> 
> Updated version was already sent but again no response...
> https://patchwork.ozlabs.org/project/uboot/patch/20220817205622.26644-1-pali@kernel.org/

I was expecting Marek Behún to follow up and take this as he's the one
that volunteered to take over one of the powerpc custodian trees and
push this through.  I've replied now with more feedback.

> > > https://patchwork.ozlabs.org/project/uboot/patch/20220828151929.32588-1-kabel@kernel.org/
> > 
> > AFAICT, this patch has some open issues still.
> 
> There were suggestion to make (that new kconfig option) enabled for
> everything. But nobody respond if it is possible to have it
> unconditionally enabled for every one board. Hence new config option in
> that patch was introduced.

Yes, the feedback was to change things so that the behavior is the
default, since it looks like what we're doing now is wrong. And you have
access to a few platforms to test things on to see if it breaks other
platforms, and then putting it in to -next or early in the merge cycle
on master would shake out any other problems with the approach. So if
you're still interested in that platform, please make it a generic
change or report back where it does fail so we can see what's going on
elsewhere. Making the init_sequence_f table even more difficult to read
is something we do want to avoid.

-- 
Tom
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 659 bytes
Desc: not available
URL: <https://lists.denx.de/pipermail/u-boot/attachments/20230612/f42664cb/attachment.sig>