This is the mail archive of the crossgcc@sourceware.org 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] |
On 2/19/06, Robert Schwebel <robert@schwebel.de> wrote: > Good idea :-) Note I didn't say how to achieve it :-) > I understand the current gcc documentation like this: > > * -mfpu=name: specifies hard floating point. Values for name > are: fpa, fpe2, fpe3, maverick, vfp > > * additional -msoft-float: generate software floating point > for the format specified above. > > As long as the Makefiles specify what they want, everything is file. > When building things like u-boot and kernel/userland you still have to > agree on one variant and people will problably have different ideas > about what is right for them. Thanks for the summary. > I would have no problem to build things in a consistent way, because I > can easily configure PTXdist to do it in such a way. But I understand > that people want to do things like arm-softfloat-linux-gnu-gcc foo.c and > expect a.out to run on "their" systems. > > So the question is: what should toolchains do by default, when no > -m*float* argument is given? Whoever builds the toolchain (in this case, ptxdist) just needs to pick a default value for -mfpu and -msoft-float, right? > Should we build > arm-softvfp-linux-gnu > arm-softfpa-linux-gnu > toolchains? Not a bad idea. Which combinations are likely to get used in the real world? - Dan -- Wine for Windows ISVs: http://kegel.com/wine/isv ------ Want more information? See the CrossGCC FAQ, http://www.objsw.com/CrossGCC/ Want to unsubscribe? Send a note to crossgcc-unsubscribe@sourceware.org
Index Nav: | [Date Index] [Subject Index] [Author Index] [Thread Index] | |
---|---|---|
Message Nav: | [Date Prev] [Date Next] | [Thread Prev] [Thread Next] |