[Buildroot] [PATCH 6/7] support/scripts/qemu-boot: gitlab tests for Qemu
arnout at mind.be
Tue Apr 30 11:20:47 UTC 2019
[Please don't top-post, but reply inline like I do below.]
On 30/04/2019 10:54, Jugurtha BELKALEM wrote:
> Thank you for you comments Arnout, I have corrected the indentation, timeout
> and function declaration issues.
> However; I think that duplicating the content of readme.txt of each qemu
> architecture would be difficult to maintain (even if it is more convenient to
> use), this is why a project like toolchain builder :
> https://github.com/bootlin/toolchains-builder/blob/master/build.sh is reading
> directly from readme.txt.
The idea would not be to duplicate it, but to move it. I.e., readme.txt just
says "launch with board/qemu/x86_64/launch.sh", and the actual command is in
that shell script.
But maybe we should see what other Buildroot maintainers have to say about it.
Thomas, Peter, Yann, Romain?
> Concerning the naming of x86_64 folder in board/qemu, it is the only one that
> uses underscore instead of hyphen. Changing the directory name may cause issues
> with other works as maybe other scripts as relying on this name (especially that
> it exists since a long time ago).
OK, true. Let's keep that like you have it then.
> I want to thank you also Max for your answer (We need to investigate more).
If it doesn't work on gitlab-ci, I would just not enable the build of host-qemu
for those (and put some intelligence in the script that it only runs if
host-qemu has been built).
> On Tue, Apr 30, 2019 at 1:38 AM Max Filippov <jcmvbkbc at gmail.com
> <mailto:jcmvbkbc at gmail.com>> wrote:
> On Mon, Apr 29, 2019 at 3:32 PM Arnout Vandecappelle <arnout at mind.be
> <mailto:arnout at mind.be>> wrote:
> > > + elif [ $device_name == "xtensa-lx60" ] || [ $device_name ==
> "xtensa-lx60-nommu" ]; then
> > > + echo "xtensa cannot be tested"
> > If that is the case, why do we even have these qemu defconfigs? Max?
> Tested both configurations just now, with the current buildroot tip
> (2019.02-959-gd54b0e22aef3) and QEMU-4.0. Both build and boot for me
> with QEMU command line from the readme.
Is this with the system-intalled qemu? I guess so, because Buildroot still has
3.1.0... That might be the issue :-)
More information about the buildroot