On Tue, Jun 23, 2015 at 3:04 PM, Brian Paul <bri...@vmware.com> wrote: > Otherwise, if we're trying to build a 32-bit Mesa on a 64-bit host > we wind up with -DUSE_X86_64_ASM, which is incorrect. > --- > configure.ac | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/configure.ac b/configure.ac > index ddc757e..b12f5f9 100644 > --- a/configure.ac > +++ b/configure.ac > @@ -605,7 +605,7 @@ if test "x$enable_asm" = xyes -a "x$cross_compiling" = > xyes; then > fi > # check for supported arches > if test "x$enable_asm" = xyes; then > - case "$host_cpu" in > + case "$target_cpu" in > i?86) > case "$host_os" in > linux* | *freebsd* | dragonfly* | *netbsd* | openbsd* | gnu*) > -- > 1.9.1
According to [1], host is "the machine that you are building for" and target is "the machine that GCC will produce code for". In the context of building GCC, I think this means that the resulting GCC binaries will run on $host and will produce code for $target. In the context of Mesa, I can't come up with a way that host != target makes sense. docs/autoconf.html suggests using --build=x86_64-pc-linux-gnu --host=i686-pc-linux-gnu to build on x86_64 for i686. Is that what you're doing? [1] https://gcc.gnu.org/onlinedocs/gccint/Configure-Terms.html _______________________________________________ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
