I've got to the bottom of the issue with mSD on the BeagleBone.

Basically the GPIO_OMAP option has changed [1] in the 4.2 cycle and
this is used for card detection of the mSD card. Prior to this we
didn't have it explicitly in our config [2] so it was being selected
by something else and it was being built in rather than modular as
only with 4.2 can it now be built as modular, and as a result it came
up on our radar and was enabled as a module [3].

If I explicitly put this in the initrd it works fine, we have the
omap_hsmmc modular too, but it's not pulled in automatically as
there's no module dependy by omap_hsmmc on gpio_omap.

In the short term I've built it back into the kernel [4] as that's the
way it's been for likely ever so it'll appear in the next kernel build
after rc8.git0.1

If you want to use any of the other 4.2 kernels you can regenerate the
initrd withe the following line in the /etc/dracut.conf file

add_drivers+="gpio_omap"

This is fixed in kernel-4.2.0-0.rc8.git1.1.fc24, latest is f-23 GA is kernel-4.2.0-0.rc8.git0.1.fc23

Proposed as a Freeze Exception for 23-beta by Fedora user pbrobinson using the blocker tracking app because:

 The 4.2.0 GA kernel is already in updates testing, it would be useful to have the BeagleBone Black working as it's a widely available device for testing ARMv7. (I suspect getting 4.2 GA will be useful for other fixes too)

Discussed at 2015-09-10 freeze exception review meeting: https://meetbot-raw.fedoraproject.org/fedora-blocker-review/2015-09-10/f23-blocker-review.2015-09-10-16.00.log.txt . It sounds like the required bits have in fact already made it to stable and nothing more should be needed here; Peter, could you please close this bug if everything is OK, or update it appropriately if something else is still needed? Thanks.

I was awaiting a working TC/RC before I closed it.

Confirmed it's working and looks fine with Beta TC5