Robert,

On Sun, Jan 03, 2016 at 10:27:26AM +0100, Robert Schwebel wrote:
> On Sat, Jan 02, 2016 at 06:02:37PM -0800, Bryan Hundven wrote:
> > > > I'm curious to know if ptxdist/OSELAS.Toolchain plans to support
> > > > uClibc-ng in the future?
> > > > 
> > > > http://www.uclibc-ng.org/
> > > 
> > > with half a year without commit to uClib git I'm considering it a dead
> > > project (I failed to verify mailing list activity as neither Firefox
> > > nor IE (!) allowed me to enter archives). uClibc-ng should be relatively
> > > easy to support as config system remained the same... Just tried and
> > > succeded. But are there any active uClibc users?
> > 
> > Well, I use it for a few platforms. And I get a lot of requests for it's
> > support on crosstool-ng.
> 
> We also use uclibc, especially for uCLinux systems. A few years ago, we
> mainlined Cortex-M3 support. It never gained business traction so far,
> because the BoM for a system is more expensive and less powerful than
> the cheaper systems with mmu (like Atmel 9x25, TI AM335, FSL MX23 etc).
> We still maintain EFM32 in the kernel.

Specifically, I still mess around with the EFM32 port, which is why this
was an interest. Although I realize it is not a focus of pengutronix.

But being able to turn on and off parts of uClibc/uClibc-ng also helps to
shrink the user-space footprint on other systems as well.

> However, these days we are seeing more and more Cortex-A cpus emerging
> with Cortex-M coprocessors. With these devices, running Linux on both
> cores is a real option (as far as I know, this is already implemented
> for Vybrid).
> 
> So uClibc stays relevant and will continue to be in OSELAS.Toolchain.

Indeed.

-Bryan