Hi,
While waiting for the TTG return. I'm trying to start working on the platform ds3617xs.
You will find on github a first test version for very basic modifications of redpill-klm with small draft of the development of this platform
https://github.com/jimmyGALLAND/redpill-lkm/tree/develop
And also the loader configurations on https://github.com/jimmyGALLAND/redpill-load/tree/develop
I specify here that I am far from having the level of team thor in kernel development (and english !).
So I wouldn't be able to go too far in increasing compatibility with 3617xs.
Regarding the compatibles extensions I only put online at https://github.com/jimmyGALLAND/redpill-ext, the virtio (without loading virtio_balloon which crach !),
The acpi management module, and intel e1000 net driver.
Note: The bromolow 3.10.105 modules from pocopico may work, but i haven't tested. However, given that archi is broadwell not bromolow, I think that it is preferable to compile extensions modules specifically for this platform.
I have not tested the USB boot! However the SATA boot works fine, and install ds3617xs-25556xs.pat package work !
Note that if you want to update to ds3617xs-25556xs update2, DO NOT NOW ! then a new loader will be needed!
In the ds3617xs-25556-2xs root package, there is no rd.gz, and zimage: They are contained in a .deb archive in the archive tree. I would try to modify the loader for support criticals updates.
Regarding the development for DSM 7.0.1 with ds3617xs, I saw you coming;).
What is interesting about this "broadwell" platform is that in DSM 7 version the kernel goes to version 4.4.180+, like the ds918+. (work with apache>2.46 by example. With kernel 3.1: "Function not implemented: AH00141: Could not initialize random number generator")
However we can always notice the lack of publication of the GPL sources for this 4.4.180+ kernel "sadness and cry"
The problem is, all modules i compiled with the vanilla kernel 4.4.180+ have gone into crach or not load.
I need more investigation on this. However the ds3617xs 7.0.1 boot, is ALIVE !!!, and I have the /dev/synoboot[1-3] in my sights! I would post this in few days.
Have a good day