-
-
QNX drivers were user processes. I really liked that model.
-
Ability to reuse existing widespread drivers is a must for any OS project if they want to support even moderate amount of COTS hardware.
-
Absolutely. It's utterly stupid for an OS/kernel project to try to NIH drivers for everything.
-
Unfortunately this leaves us stuck emulating the Linux kernel-internal driver APIs or similar.
-
That's the route Genode took with dde_linux right? https://genode.org/documentation/developer-resources/porting_device_drivers …
-
Thanks for the link! Even better that source compat might be loading the .ko ELF files in user processes tho.
End of conversation
New conversation -
-
-
There's also some work in progress to "rump-ify" the Linux kernel via LKL, but IMO that will take a long time /cc
@thehajimeThanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
This is how
@GenodeLabs ported several Linux drivers to Genode, e.g. Intel KMS see http://genode.org/documentation/release-notes/15.11#Intel-KMS_framebuffer_driver …Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
http://rumpkernel.org plus some glue to run in $SANDBOX_OF_CHOICE does what you want for existing drivers, without the NIH.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
me too! any concrete plans and timelines so far?
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.