In <https://chromium-review.googlesource.com/c/chromiumos/third_party/kernel/+/297826> we landed a CHROMIUM patch to make certain hardware devices running Hynix eMMC resume more quickly.

We carried this patch forward to the Chrome OS 4.4 and Chrome OS 4.14 kernels.  

I'm proposing dropping it from 4.19 kernel.

If this patch is actually important then someone needs to come up with a solution that upstream Linux is happy with.  Carrying forward a patch in the Chrome OS kernel like this makes very little sense.

--

NOTE: During the 4.14 rebase I asked Gwendal about this patch.  He said:

> I don't understand this patch. If the controller intent to shutdown 
> power in S3, then it has MMC_CAP2_FULL_PWR_CYCLE set. What is 
> incorrect is when we are suspending, we are sending a POWER_OFF_SHORT, 
> while we are suppose to send a SLEEP_NOTIFICATION (see eMMC 5.1 section 7.4.110).

--

It looks like there was some attempt to upstream the patch in the past:

v1: https://patchwork.kernel.org/patch/9360633/
v2: https://patchwork.kernel.org/patch/9434051/
v3: https://patchwork.kernel.org/patch/9452209/

...that seems to have stalled.  We need an owner to pick this up and figure out how to move forward.

---

From looking at the internal bug (b/35556267), if relevant hardware gets uprevved to a newer kernel for some reason then the worst case is a resume time regression.  Presumably that will get caught in our tests since resume time is definitely something we look at.