On 2014-06-17 13:59:17 +0200, Diego Biurrun wrote: > On Tue, Jun 10, 2014 at 10:43:33PM +0200, Janne Grunau wrote: > > --- a/configure > > +++ b/configure > > @@ -1361,6 +1361,7 @@ HAVE_LIST_PUB=" > > HEADERS_LIST=" > > alsa_asoundlib_h > > altivec_h > > + arm_neon_h > > arpa_inet_h > > cdio_paranoia_h > > cdio_paranoia_paranoia_h > > @@ -1477,6 +1478,7 @@ TOOLCHAIN_FEATURES=" > > gnu_as > > ibm_asm > > inline_asm_labels > > + neon_intrinsics > > pragma_deprecated > > symver_asm_label > > symver_gnu_asm > > @@ -1639,6 +1641,7 @@ armv6_deps="arm" > > armv6t2_deps="arm" > > armv8_deps="aarch64" > > neon_deps_any="aarch64 arm" > > +neon_intrinsics_deps="arm_neon_h neon" > > vfp_deps_any="aarch64 arm" > > vfpv3_deps="vfp" > > > > @@ -3900,6 +3903,11 @@ EOF > > > > +if enabled neon ; then > > + check_header arm_neon.h > > + check_code cc arm_neon.h "int64x2_t test" && enable neon_intrinsics > > +fi > > I wonder why you check for arm_neon.h separately, the result of that > check is never used separately.
yes, not really needed, will be removed > neon_intrinsics depends on neon, so the if-block should be unnecessary. that's there to skip the check an archs except arm and aarch64, so not really necessary but nice to have. with arm_neon_h removed I can change it to: enabled neon && check_code cc arm_neon.h "int64x2_t test" || disable neon_intrinsics which get's rid of the if-block and fixes the logic disable/enable logic Janne _______________________________________________ libav-devel mailing list [email protected] https://lists.libav.org/mailman/listinfo/libav-devel
