Continued...
And this really sums up the whole need for
"
magisk: Remove Play Services from DenyList …
The Zygisk module will never load if GMS is in the DenyList..."
I did envisage issues with multiple hiding solutions operating on one process as covered above, but these haven't been demonstrated anywhere...
However while denylist does hide by virtue of 'Magisk getting out of the way', it actually prevents Magisk modifications from affecting selected processes at all, unlike. hiding solutions, so any Zygisk based injection of code into a forked zygote process that loads and runs a new app's code cannot occur for instance.
Since USNF's key function works this way it must fail when gms droidguard process is in (enforced) denylist:
I believe that the various targeted prop changes needed to trigger hardware based verdict enforcement bypasses in S/N and PI also use Zygisk, so these functions will fail also...
I have to agree that it's the nature of denylist that necessitates the removal of at least the gms droidguard process from denylist (although I still think merely instructing users is to do this manually where needed is all that is needed and would actually cause less confusion)...
Anyway, thanks for helping me properly connect modification denial with the reason for
magisk: Remove Play Services from DenyList commit.
I didn't mean to suggest that Shamiko users were dumb enough to enforce denylist... USNF would remove gms processes on next boot anyway...
Shamiko should really note somewhere that hijacked denylist really = hidelist, if only to create an excuse to fall about laughing at the confused gringo-user posts suddenly appearing in LSP TG threads everywhere...
View attachment 5782451
PW