The goal: compile and install a new kernel (compiled from a different source) alongside
the existing (stock) one.
I have been out of touch with kernel building for the last few years. In the good old days, I would compile a new kernel package, install it, and it would show up alongside my existing kernels when booting.
I did essentially the same, but I ended up with more packages that I used to:
make deb-pkg
linux-headers....deb
linux-image....deb
linux-firmware-image....deb
linux-libc-dev....deb // <---
Upon checking the contents of these packages, most of them have a versioned subdir. linux-libc-dev....deb
, however, just spits its content into /usr
, without an special path prefix. Meaning it will overwrite whatever the stock package / the custom build before it has put in there...
The question: how can I use multiple kernels on th same system?
Is it somehow safe to install a dozen versions of linux-libc-dev....deb
over-and-over?
Shall I just forgo installing, and pretend this is not there, and the stock version linux-libc-dev....deb
will do a perfect job supporting another kernel?
Not sure, if its relevant, I am building a newer kernel than I have as stock, going from 4.9.144
to 4.9.160
, and 4.20.11
. (I am trying to test out an error with different kernels.)
I DID see a lot of posts on "what is linux-libc-dev", none of which answered my question.
0 Answers