[yocto] Compiling different version of IGB as a kernel module

Jonathan Haws Jonathan.Haws at sdl.usu.edu
Sat Jan 30 07:24:25 PST 2016


Chris,

I have, but the problem there is that the driver patch I have apples to version 5.3.2 or newer. I'll look into it more, but I'd like to understand what is going on in the build and why it is falling.

Thanks!
Jon



Sent from my Verizon Wireless 4G LTE smartphone


-------- Original message --------
From: Chris Trobridge <christrobridge at hotmail.com>
Date: 2016/01/30 5:13 AM (GMT-07:00)
To: Jonathan Haws <Jonathan.Haws at sdl.usu.edu>, yocto at yoctoproject.org
Subject: RE: [yocto] Compiling different version of IGB as a kernel module

Hi Jon,

As an alternative, have you considered patching the igb source that already exists within yocto?

Regards,
Chris

From: Jonathan.Haws at sdl.usu.edu
To: yocto at yoctoproject.org
Date: Sat, 30 Jan 2016 00:46:20 +0000
Subject: [yocto] Compiling different version of IGB as a kernel module


Hello all!

I am trying to get the Intel IGB driver compiling as a kernel module
using source downloaded from SF.  I'm doing this because I was given a
patch to enable an external PPS to provide a time source to the PTP
mechanisms in the device.

I've got the recipe setup and it makes it to the do_compile step,
however it fails here with some interesting output.  First, I get a
message from the shell stating that -- is an invalid option.  Second,
cc1 fails with a bad option.  I'm not even sure where to start debugging
this!

I did have to patch the Makefile to get it to find the kernel sources
and build files correctly.  I've attached my recipe and that patch file.

I've also attached the output of do_compile - hopefully someone can look
at this and see what I am doing wrong!

One thing I have tried is I went down into the tmp/work directory where
the IGB driver was being built and tried running the same make command
(without the CC=, AR=, LD= settings, since those are not on my path...)
and it seemed to build just fine, so it seems like it is something to do
with the toolchain, but that's all I can tell.

Thoughts?

Thanks!
Jon


--
Jonathan R. Haws
Embedded Engineer
Space Dynamics Laboratory
(435) 713-3489
jhaws at sdl.usu.edu

-- _______________________________________________ yocto mailing list yocto at yoctoproject.org https://lists.yoctoproject.org/listinfo/yocto



More information about the yocto mailing list