[Mageia-dev] cleaning old architecutres in glibc and elsewhere

Arnaud Patard (Rtp) arnaud.patard at rtp-net.org
Thu Feb 24 17:33:25 CET 2011


Michael Scherer <misc at zarb.org> writes:

> Hi,

Hi,

>
> while trying to fix some missing packages ( the locales-* ), I have
> found myself looking at gibc packages.
>
> We still have support for various arch :
> - alpha

this arch is broken for quite some time. afaik alpha support is done
through glibc-ports and our spec doesn't handle that atm (yet, as arm &
mips need glibc-ports). If we ignore this, the rest of the support for
alpha is not a big part of the glibc spec. Grepping for alpha in the
patches didn't returned a lot of answer too.

> - sparc*

no idea about sparc support.

> - ppc

I guess that last time ppc support was tested was when some people
were testing cell ppc systems. It probably needs more care to work
nowadays. I'm seeing some ppc32 support in glibc-ports but I've no idea
if it's really needed (eg ppc64 may not need that).

>
> Since alpha is dead ( last proc, 2003 ), I think we can all agree that
> we can clean this. 
>
> Now, for sparc and ppc, is there someone here who is willing to work on
> a port ? If not, I will clean the support ( mainly removing %ifdef and
> stuff like that ) from glibc.spec in 2 weeks. 

It's a matter of weighing the cost of keeping them. If the cost is high
and there's no one stepping up to maintain sparc or ppc, it's better to
remove them. It can be added back later if someone jumps in.
If there's no cost, it's more a subjective question and the people who
are best entitled to answer to this question are the people who will
maintain the glibc.

fwiw, I hope I'll be able to merge at least the arm bits needed for
support in the glibc within theses 2 weeks but if I don't manage to do
it, please ping me on irc when you'll do the cleaning in order to make
sure it won't make merging my work more difficult :)

Arnaud


More information about the Mageia-dev mailing list