[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Preparing the m68k port for the future.



On Fri, Jan 13, 2006 at 02:09:19PM +0100, Wouter Verhelst wrote:

> Additionally, Ingo told me when the mail about that meeting had come out
> that he'd already tried such a setup in the past (I didn't know that
> when we were in Helsinki, but it was before that), and that his setup,
> IIRC, was in fact slower than a native build because of the overhead of
> the network call to start the compiler--At least that's how I recall his
> explanation. I don't remember the details; if you have any questions,
> please ask him.

Uhm, well, yes, I tried a distcc setup on m68k. Actually it's the opposite: 
Of course there's a overhead via network (I was using two m68ks with 450 km
and 2 DSL dialups inbetween), but the test (kernel compile) showed approx. a
total of 40-50% gain in build time. M68ks (well, Amigas at least) only have
10 Mbps NICs, so you won't have much faster transfers than 300-500 kB/s
between hosts anyway. 
The m68k distcc worked quite well, but when I tried to setup a cross-distcc
it failed on the crosscc side. Somehow the generation of the need cross
files failed. Can't remember the specifics anymore. 
 
> Now it might be that with faster hardware (I don't know the specifics of
> Ingo's setup; I seem to recall he mentioned a Pentium III-class system

It was a PowerPC G4 1 GHz. ;)

> as the distcc host, but I might be delirious) the distcc stuff will
> indeed work better, but if such a setup with a not /that/ old system is
> already slower than a native build, then I'm not very hopeful that a
> distcc setup is going to get us much benefit, while it _will_ give us a
> huge overhead.

I think, distcc will be nice for such things like qt-x11 and other long
building stuff. For packages with shorter build times (<1 hour or so) I
don't see much gain, because the installation and removal of build-deps will
take the most time and distcc certainly can't help there. ;)
The benefit of distcc would be, that we can add flawlessly more CPU power
whenever needed, but don't need to when there's no backlog, so we can build
natively. 

The main showblocker with that is that package building doesn't support "make
-jX" yet. I think other archs with SMP support might benefit as well when
there would be a way to support this feature... 

-- 
Ciao...                //        Fon: 0381-2744150 
      Ingo           \X/         SIP: 2744150@sipgate.de

gpg pubkey: http://www.juergensmann.de/ij/public_key.asc



Reply to: