Re: [Boost-users] boost on embedded target such as ARM

Subject: Re: [Boost-users] boost on embedded target such as ARM
From: Michael Powell (mwpowellhtx_at_[hidden])
Date: 2015-04-23 12:04:37

On Thu, Apr 23, 2015 at 5:03 AM, Trek <trek830_at_[hidden]> wrote:
> how is boost doing on embedded environment in terms of efficiency?
> thanks

The last time I targeted the ARM with Boost, it works pretty well.

Much of this is contingent on the host OS, processor and supporting
hardware arch, and so on, of course. But by and large, I couldn't
complain; the high-res timer operations did very well, threading and
futures worked splendidly, even signals/slots did pretty good.

This was for embedded only, not real-time, per se. I might dare to
call it soft-real-time; definitely the OS (ArchLinux) was not of the
real-time variety.

We had other issues going on, a kernel design philosophy that failed
to regard application code (i.e. for things like resource file
handling), multiple failures in duplicate I2C addressing, and things
of this nature. Nothing that a disciplined, diligent, if vigilant,
design process couldn't remedy. Nothing to do with Boost however.

These aren't metrics of course, and you should run your own. My
experience there was, by and large, technically very good.

HTH

> _______________________________________________
> Boost-users mailing list
> Boost-users_at_[hidden]
> http://lists.boost.org/mailman/listinfo.cgi/boost-users