On Tue, 2010-06-22 at 08:23 +0300, Martin-Éric Racine wrote: > On Tue, Jun 22, 2010 at 4:11 AM, Mitch Davis > <[email protected]> wrote: > > On Tue, Jun 22, 2010 at 10:54 AM, Gaetan Nadon <[email protected]> wrote: > >> When doing full X.Org builds, many waste time investigating > >> compile errors, not knowing anything about Geode. > >> > >> Only build on i*86 cpu. The configuration will not abort, > >> a warning message is issued and no C code is compiled. > > > > I haven't tried it yet to know whether it's a problem, but it's > > possible to *build* it on ia64, just not *run* it. > > > > (I am using a cross-compiler to compile x86 code on am ia64 machine) > > > > Even if the software still builds, I think the messages should be > > changed to indicate it can't be run on a non x86 machine. > > We already know that it cannot run on non-x86 architectures and that > it can only compile on x86_32 or x86_64 (IIRC the build scripts even > change some flags if the host is x86_64). Where would you put the > message then? >
Sorry, I read your post after I re-submitted the patch. Bottom line is I am trying to prevent build failure for those, mainly new comers to X.Org, who build the whole system (approx 240 packages). By making use of the existing I386ARCH Automake conditional, I can almost eliminate the build failures and the posts and the bug reports that come with it. An alternative is to fix or conditional compile the assembler code so it compiles on AMD64. I thought this would harder to justify for Geode developers. You may review the the second edition of the patch, as I am not familiar of the Geode architecture. Thanks > Martin-Éric
signature.asc
Description: This is a digitally signed message part
_______________________________________________ Xorg-driver-geode mailing list [email protected] http://lists.x.org/mailman/listinfo/xorg-driver-geode
