RTEMS on ARMv8
Jeff Kubascik
Jeff.Kubascik at dornerworks.com
Tue Apr 2 13:21:32 UTC 2019
On 4/2/2019 8:34 AM, Joel Sherrill wrote:
> On Tue, Apr 2, 2019, 7:21 AM Sebastian Huber <sebastian.huber at embedded-brains.de
> <mailto:sebastian.huber at embedded-brains.de>> wrote:
>
> On 02/04/2019 14:15, Jeff Kubascik wrote:
> > How capable is the device tree support in RTEMS? I did notice some device tree
> > code in the IMX BSP layer, but it appeared to reference fixed nodes. Does
> RTEMS
> > perform driver probing based on the device tree?
>
> The device tree support in RTEMS is quite rudimentary. However, in the
> BSP you don't need much. You just have to deal with the interrupt
> controller, the clock driver and an UART driver. Maybe also I2C and SPI,
> but these are simple things. The more complex drivers are in libbsd
> (USB, MMC/SDCard, Ethernet). Here we have the device tree support from
> FreeBSD. It is not as good as the device tree support from Linux
> especially if it comes down to pins and clocks since this is a real mess
> usually.
>
>
> Neither of us mentioned licence. We do not accept GPL code in RTEMS because it
> is statically linked with applications and this would result in applications
> needing to follow the GPL requirements.
>
> The 2 paragraph BSD license is not the preferred license. The historical RTEMS
> license is GPLv2 with a linking exclusion. It was based on the GCC runtimes.
>
> --joel
>
>
> --
> Sebastian Huber, embedded brains GmbH
>
> Address : Dornierstr. 4, D-82178 Puchheim, Germany
> Phone : +49 89 189 47 41-16
> Fax : +49 89 189 47 41-09
> E-Mail : sebastian.huber at embedded-brains.de
> <mailto:sebastian.huber at embedded-brains.de>
> PGP : Public key available on request.
>
> Diese Nachricht ist keine geschäftliche Mitteilung im Sinne des EHUG.
>
That is good information to know. My plan is to use the RTEMS license for the
MPSoC BSP to keep things simple.
However, the Xen support could be tricky, as the majority of the kernel and
library code is GPLv2. We have ported other operating systems to Xen in the
past, and from what I remember, the Xen headers that were copied into the guest
OS code do have a MIT license.
-Jeff Kubascik
More information about the devel
mailing list