I have a test router on which I installed " kmod-fs-ntfs3" that is supposed to support writable ntfs at the kernel level but there is virtually no documentation. Since it´s part the kernel I should be able to mount a NTFS drive with the "mount" command but it always fails:
mount -t "ntfs" /dev/sda2 /mnt/myhd
mount: mounting /dev/sda2 on /mnt/myhd failed: No such device
and before you ask, the usb drive does register correctly and "block info" reports the right information 
Are you actually using the quotes around ntfs ? If so, omit them.
mount -t ntfs3 /dev/sda2 /mnt/myhd it should be like this as far as I know.
I have been using it for quite some time, although it is a backported version for 5.10 but it's running as smooth as it possibly could.
I tried that one too but I keep getting this error message:
mount: mounting /dev/sda2 on /mnt/myhd failed: Invalid argument
Thanks for your help! I won't spend too much time on this, I thought it was something cool to try and compare speeds with ntfs-3g on a low-powered router.
I tried with and without just in case but nothing would do for some reason. Good point though.
Believe me it's definitely worth the effort. With ntfs 3g , my device cpu goes 100% during file transfers and i get like 20-30mbps read with ntfs3 100mbps read and 80mbps write, that too with lower cpu usage.
It's a total game changer, atleast for me.
I remember one thing though, for some reason, the drive won't mount unless I add force as mount option. So do try that along with ntfs3.
mount -t ntfs3 -o force /dev/sda2 /mnt/myhd
Unfortunately I keep getting the same error message without any log message. I'll try again with a full config reset as this router may have accumulated "lab garbage" over time. That will take a few days before I can do that. Thanks again for your valuable input.
For reference, my test router is a GL.iNet 6416 but the target "production" router would be a GL.iNet GL-AR750S (NOR/NAND).
I remember this error now, it's possible that the patch for fixing issues with big endian devices was never implemented with main driver.
Your test unit and other both seem big endian devices and ntfs3 has issues with it.
There was a patch made to solve this problem but I have a feeling that it hasn't made its way to the main code.
The ntfs3 driver in its current condition (if I'm right about the patch) will only work with little endian devices.
I read about that Endian stuff a few month ago but frankly that never crossed my mind while trying that extension but It does makes a lot of sense. Should we report this as a bug?
This is the patch that needs to be implemented in ntfs3 driver source when compiling. https://lore.kernel.org/lkml/20211207102454.576906-4-thomas.kuehnel@avm.de/T/
I don't know if this will even be approved for the push. Or perhaps it's already there? I don't know, but I highly doubt about that it is.
But the same patch made this driver work with my big endian device and I've been using it ever since.
It would make sense to test the newest kernel you can (KERNEL_TESTING would be v6.1 in main) on this device first, before spending any time on debugging the details.