Okay, so that simply appears to be rtcwake (or the actual rtc) screwing the pooch (i.e., it suspends, but never wakes up).
Unfortunately, I don't think I have a better option for you than basically rewriting the suspend and rtc code yourself, which is... non trivial (c.f., KOReader & Plato for viable implementations of said madness).
Alternatively, you *could* try with a real rtcwake binary cross-compiled (or pulled from Debian) from util-linux instead of the old busybox version Kobo is shipping.
|