On 04/27/2017 12:52 PM, Greg Kroah-Hartman wrote: > On Thu, Apr 27, 2017 at 11:03:08AM +0200, Ulf Hansson wrote: >> +Greg >> -trimmed cc >> >> On 26 April 2017 at 23:44, Andy Gross <andy.gross@xxxxxxxxxx> wrote: >>> On Mon, Nov 21, 2016 at 12:07:17PM +0530, Ritesh Harjani wrote: >>>> This add get_min_clock() and get_max_clock() callback >>>> for sdhci-msm. sdhci-msm min/max clocks may be different >>>> hence implement these callbacks. >>>> >>>> Signed-off-by: Sahitya Tummala <stummala@xxxxxxxxxxxxxx> >>>> Signed-off-by: Ritesh Harjani <riteshh@xxxxxxxxxxxxxx> >>>> Acked-by: Adrian Hunter <adrian.hunter@xxxxxxxxx> >>> >>> This patch needs to be picked up for stable as 4.9 is currently broken w.r.t. >>> MMC on Qualcomm platforms. >>> >>> Fixes: d84be51d1c1d ("mmc: sdhci-msm: Enable few quirks") >>> >>> Regards, >>> >>> Andy Gross >> >> I did a quick investigation around this. >> >> So for some reason Greg decided to pick up the original commit, >> a0e3142869d2 ("mmc: sdhci-msm: Enable few quirks"), for 4.9 stable >> [1], even if that patch wasn't sent or tagged for stable. >> It was a mistake, as simple as that - and I should also have spotted >> it when Greg posted his stable-review series, sorry. >> >> For clarification, the original commit a0e3142869d2 ("mmc: sdhci-msm: >> Enable few quirks"), is present in 4.10 together with $subject patch. >> So the regression is present only in 4.9 stable. >> >> Greg, can you pick up @subject patch for stable according to Andy's >> suggestion or you want some of us to re-post it to stable? > > Sorry about this. What git commit should I take for this from Linus's > tree for the 4.9-stable tree? Or if there isn't one, I'll be glad to > just revert this. > Thanks Greg and Ulf, A revert of d84be51d1c1d ("mmc: sdhci-msm: Enable few quirks") would be the best solution. The other option is to pick 80031bdeb764 ("mmc: sdhci-msm: Add get_min_clock() and get_max_clock() callback"). BR, Georgi