Hi Arjan, Yes, one old version of eMMC Dekker Algorithm patch broke N900. Later we fixed the issue and verified on N900. During integration testing with SCU FW, some new issues happened. So the patch is delayed on community submission until no new issue reported. We will fix Greg's comments and submit new version to community soon.
But looks like eMMC community is not very active. For example, our eMMC hardware reset patch doesn't get community response yet after the submission 3 weeks ago. We met the same situation on eMMC Trim and enhanced area support patches. Could you give us any guideline on this kind of situation? Shall we continue to wait for community's feedback? Or can we submit to MeeGo when no response in community? Thanks, Henry > -----Original Message----- > From: [email protected] > [mailto:[email protected]] On Behalf Of Arjan van de Ven > Sent: Wednesday, November 17, 2010 11:33 AM > To: Dong, Chuanxiao > Cc: [email protected] > Subject: Re: [Meego-kernel] [PATCH v1 0/3]MMC:implemented Dekker algorithm to > avoid SCU and IA using host controller at the same time > > On 11/17/2010 2:53 AM, Dong, Chuanxiao wrote: > > >> this code has a history of breaking all other systems. > >> I'm not even going to think about taking this until it's accepted by the > >> upstream maintainers. > > The reason this code breaking all other systems is the wrong usage of Dekker > algorithm in MMC block layer. But actually this code is only for MFLD and was > wrapped by MFLD controller quirks. I think the wrong use of any platform > specific > code maybe can also break other systems. Agree? And if you are still not > going to > review these patches, let me know. > > let me know how the upstream review goes. > really. > > this stuff is not even "young" anymore; we're making exceptions for code > that isn't upstream yet for your team, but the hard assumption is that > the code will go upstream within a few (3) months... > ... these patches are older than a few months already..... they really > should have been sent and accepted upstream by now. > And given their troublesome history and touching of generic code, I'm > going to stick to that. > > _______________________________________________ > MeeGo-kernel mailing list > [email protected] > http://lists.meego.com/listinfo/meego-kernel _______________________________________________ MeeGo-kernel mailing list [email protected] http://lists.meego.com/listinfo/meego-kernel
