On Mon, 23 Sep 2013 15:57:33 +0200 Jan-Simon Möller <[email protected]> said:

> > 
> > 64bit os's come with a major cost in memory footprint for pointers... it's
> > quite measurable. and that doesn't change the fact that:
> > 
> > 1. arm is STILL 32bit and will be for a long time still. if tizen can't be
> > built ON an arm platform (which is 32bit) then it can't self-host on arm...
> > which is embarrassing.
> 
> We have to differ here ...
> 
> a)
> The request was to drop the tools for x86@32bit and just use the tools on 
> x86@64bit.  Those tools are built/shipped for non-tizen distributions.
> That being said, I'm not in favour of dropping it. 
> I know it is a maintainance burden - but that actually _is_ part of the game.
> And if we cannot do this for our own tools ... 

as per my mail - i believe the issue is at least partly qemu memory mapping
issue on 32bit (some kind of collission of mappings i think), and if it hapens
on x86 it possibly/likely happens on arm too, so same issue if such a 32bit
issue is dropped. :)

> b) 
> To my best knownledge, the set of tools was never built for arm native (other 
> distros or tizen). Actually there was no reason to up to now; as we have more 
> powerful native hw, that _can_ be considered. But it means to check another 
> target.
> 
> c)
> Running the tizen tools on tizen iself  is not yet done. Well, we should be 
> able to eat our own dogfood out of our own 3d-printed bowl.

that was the point - we should be dogfooding... and we are not. and dropping
32bit support moves a step away from that, not towards it. :)

-- 
Carsten Haitzler (The Rasterman) <[email protected]>
_______________________________________________
General mailing list
[email protected]
https://lists.tizen.org/listinfo/general

Reply via email to