ENI drive compilation fails

[andsens]

Copy pasta from comment on PR #162 by @daveloyall

So, did you end up using the host's build environment for this? I think so, because today (using HEAD of master) I got this:

Error! Your kernel headers for kernel 3.2.0-4-amd64 cannot be found. Please install the linux-headers-3.2.0-4-amd64 package, or use the --kernelsourcedir option to tell DKMS where it's located Command 'chroot /target/c670114b/root dkms build -m ixgbevf -v 2.16.1' returned non-zero exit status 1 The host is:

hobbes@somemachine:~/bootstrap-vz$ uname -a Linux someothermachine 3.2.0-4-amd64 #1 SMP Debian 3.2.41-2 x86_64 GNU/Linux The guest (target new AMI) is (this is from the bootstrap-vz's --debug console output):

...
Determining kernel version Kernel version is 4.0.0-2-amd64 ... Wait a minute... Here's the host again:

hobbes@somemachine:~/bootstrap-vz$ uname -v

1 SMP Debian 3.2.41-2

Where does ${kernelver} come from? https://github.com/andsens/bootstrap-vz/blob/master/bootstrapvz/providers/ec2/tasks/network.py#L56

To be clear: it appears that dkms is attempting to target a kernel version that is neither that of the target nor that of the host. (Or, I'm reading this wrong! :) )

It seems the kernel version is not detected correctly. We can fix that though. There's a task for that :-)

[daveloyall]

Thanks @andsens!

If you ask me to run some commands and give your the output, or checkout some branch and try it, etc, I'm happy to oblige.

[myhro]

I've always thought that this piece of code were somewhat duplicated (and badly located, as it's part of a task which has another purpose) as DetermineKernelVersion exists. But this makes sense, as when Jonas' PR was merged I didn't knew/remembered/thought about it. :-(

Edit: worth to mention that his PR was based on my code that was sitting on a temporary branch, so it's not his fault. :-)