This is the mail archive of the crossgcc@sources.redhat.com mailing list for the crossgcc project.
See the CrossGCC FAQ for lots more information.
Index Nav: | [Date Index] [Subject Index] [Author Index] [Thread Index] | |
---|---|---|
Message Nav: | [Date Prev] [Date Next] | [Thread Prev] [Thread Next] |
Other format: | [Raw text] |
> -----Original Message----- > From: David Korn [mailto:dkorn@pixelpower.com] > Sent: dinsdag 26 februari 2002 11:22 > > >-----Original Message----- > >From: Vermeulen Jan [mailto:Jan.Vermeulen@siemens.atea.be] > >Sent: 26 February 2002 07:30 > > >So, what does the "und" stand for and what will I be missing from my > >crosscompiler because I manually removed it? > > > >There were others too which build fine, called "ca", "nof", > "le" etc. I can > >guess the use of some of them ("le" = little endian libs?, "nof" = no > >floating point?) But not "und" and "ca". > > Yep, you're spot on for nof and le, they exist in 2.95.x > with the same> meaning. I haven't heard of the others, but the way to find out is to > find out which one of the $srcdir/gcc/config/rs6000/t-* files > is being > selected in the top level configure, and read it to see which > flags are > passed to the build for that multilib variant. In 2.95.x the file > will be t-ppcgas, and (digging it up) aha: "ca" means AIX calling > conventions: > > ---begin quote--- > MULTILIB_OPTIONS = msoft-float \ > mlittle/mbig \ > mcall-sysv/mcall-aix/mcall-linux > > MULTILIB_DIRNAMES = nof \ > le be \ > cs ca lin > ----end quote---- > > What's in the v3 equivalent then? Ah, thanks for the pointer... I checked it out and I see: <quoting> MULTILIB_OPTIONS = msoft-float \ mlittle/mbig \ mcall-sysv/mcall-aix \ fleading-underscore MULTILIB_DIRNAMES = nof \ le be \ cs ca \ und </quoting> So, i guess that some libraries will not have been made. But which ones? The ones containing leading underscores for the functions or those without them? I have a feeling this isn't a big problem, because GCC can work with both types. The -fleading-underscore and -fno-leading-underscore flags will link against the 'und' libraries or others. My guess is that, since the normal way of compiling C is to add an underscore before the symbol (from the GCC point of view), those missing 'und'-type libraries are those which symbols do not have an underscore. The only use of these libs are to mix them with assembly code... Is this somewhat correct? Is there a simple way of testing which libs are actually missing? Jan ------ Want more information? See the CrossGCC FAQ, http://www.objsw.com/CrossGCC/ Want to unsubscribe? Send a note to crossgcc-unsubscribe@sources.redhat.com
Index Nav: | [Date Index] [Subject Index] [Author Index] [Thread Index] | |
---|---|---|
Message Nav: | [Date Prev] [Date Next] | [Thread Prev] [Thread Next] |