https://github.com/maxsteeel/nomount

NoMount is still in a beta/testing phase, If you find any bugs, please let me know 🙃

For now, until they announce the update, NoMount is broken

NoMount is currently in a usable state, according to tests done by @YGood2, but it still has some problems with files like framework.jar.
Once those problems are resolved, I will update the patches for everyone

I updated the test patches but forgot to update the kernel patches with susfs:

https://github.com/maxsteeel/nomount/tree/experimental/patches

Maybe I'll update the others later when I have time; right now I'm tired 🫩

Thanks to @YGood2 for their great help and collaboration 🫂

re-updated test patches, to add this changes:

nomount@ba71cddcb
nomount@09833c04a

credits to @YGood2
(susfs patches not updated yet)

Right now I can say that NoMount is completely stable, although some testing is still needed (especially on kernels like 6.6), so I will update the test patches as soon as I can

I've updated all test patches (including susfs and experimental)

https://github.com/maxsteeel/nomount/tree/experimental/patches

I also added a patch for 4.14, although I haven't tested it, and it might not work
this version of nomount is stable in the tests i've performed, but there might still be some bugs

latest ci module version:
https://nightly.link/maxsteeel/nomount/actions/runs/21773323042/NoMount-CI_33

I've refactor the nomount code, because atp it was causing too much overheat if you had too many injections. All the logic for resolving paths, inodes, and so on, was eliminated; now the nm binary handles collecting that data and sending it to the kernel, making all operations extremely efficient since it only has to search among the data it already has ( O(1) ), So whether it's 8 files or 800, performance remains the same ( tested with 1805 files, zero overheat )

refs:
nomount@a0d6add
nm.c

NoMount works in 4.14! thanks @Medusa659 for test!

refactor complete (you can see commit for more details)