Re: more power5 problems

Hi Daniel,

   in principle you should take the Power PC 64 linux version from the download page. I did built that version on an IBM PowerPC 64 running Suse.

If you really get stuck I can try if you give me an ssh login.

Cheers, Fons.

Daniel Cano wrote:
> Hi,
> I'm getting more errors while compiling on this machines.
> After executing make command, and after a while, I start getting lots of
> warnings like this
>
> g++ -m64 -O -o bin/proofserv main/src/pmain.o \
> -Llib -lCore -lCint -lHist -lGraf -lGraf3d -lGpad -lTree
> -lMatrix -lProof -lTreePlayer -lThread -lm -ldl
>
> /usr/bin/ld: Warning: type of symbol
> `G__MethodArgInfo::Init(G__MethodInfo&)' changed from 1 to 2 in
> lib/libCint.so
> /usr/bin/ld: Warning: type of symbol `G__defined_tagname' changed from 1
> to 2 in lib/libCint.so
> /usr/bin/ld: Warning: type of symbol `G__CallFunc::SetArg(double)'
> changed from 1 to 2 in lib/libCint.so
> /usr/bin/ld: Warning: type of symbol `G__scratch_globals_upto' changed
> from 1 to 2 in lib/libCint.so
> /usr/bin/ld: Warning: type of symbol `G__search_typename2' changed from
> 1 to 2 in lib/libCint.so
> /usr/bin/ld: Warning: type of symbol `G__Longlongref' changed from 1 to
> 2 in lib/libCint.so
> /usr/bin/ld: Warning: type of symbol `G__add_compiledheader' changed
> from 1 to 2 in lib/libCint.so
> /usr/bin/ld: Warning: type of symbol `G__SourceFileInfo::Init(char
> const*)' changed from 1 to 2 in lib/libCint.so
> /usr/bin/ld: Warning: type of symbol `G__MethodArgInfo::Next()' changed
> from 1 to 2 in lib/libCint.so
> /usr/bin/ld: Warning: type of symbol `G__MethodInfo::NArg()' changed
> from 1 to 2 in lib/libCint.so
>
> and so on. On the other hand, I'm not sure which precompiled version I
> should try, or even if should use a precompiled one. I need some help
> and it's a great pity not using this machines because they are optimal
> for ROOT (well for anything in fact), they have 4cpu's with the IBM
> hyperthreading and 12 and 24 GB RAM. This must run ROOT at lightspeed.
>
> Thanks in advance
>
> Dani

-- 
Org:    CERN, European Laboratory for Particle Physics.
Mail:   1211 Geneve 23, Switzerland
E-Mail: Fons.Rademakers_at_cern.ch              Phone: +41 22 7679248
WWW:    http://www.rademakers.org/fons/      Fax:   +41 22 7679480