Hi,
I was wondering about how far are we with implementing a RT kernel in
Debian... Some progress here? Would be nice.
The patch I created that "fits" on Debian's vanilla kernel creates
conflicts on the sources with the Debian patches.
I hope to be able to clean that up by minimizing the -rt series (e.g.
the first broken out patch consists usually of various bits from the
-tip tree that are (AFAIK) not all needed.)
So just have some more patience.
How are things going? Just interest...
Hhhm, well, I spottet a problem. The thing is that linux-rt does many
deep changes in the kernel and I won't be able to support the harder
problems. And upstream probably won't help because the Debian rt-kernel
isn't a vanilla rt-kernel. Moreover even the broken out rt-patch isn't
nicely sorted (e.g. bisectable, some patches undo changes of other
patches earier in the series etc. pp), so I fear the Debian kernel team
isn't filled with enthusiasm when asked to add an rt variant.
I already thought about packaging debian-kernel + rt for Debian and
vanilla-kernel + rt for a non-Debian package repository such that it's
easy for bug reporters to try out a vanilla kernel. But that still needs
more work.
I currently investigate how the Debian kernel packages are created. Any
help is welcome. Probably the first step will be to create a vanilla-rt
package, but that wont be accepted to go into Debian main.