Hi Thomas, > -----Original Message----- > From: Thomas Monjalon [mailto:thomas.monjalon at 6wind.com] > Sent: Monday, March 16, 2015 4:05 PM > To: Ananyev, Konstantin > Cc: dev at dpdk.org; Qiu, Michael > Subject: Re: [dpdk-dev] [PATCH] common/rte_memcpy: Fix x86intrin.h missed > > 2015-03-13 09:44, Ananyev, Konstantin: > > From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Michael Qiu > > > > > > rte_memcpy.h(46): catastrophic error: cannot open source file > > > "x86intrin.h" > > > > > > For icc and old gcc, this header is not included. > > > > > > Signed-off-by: Michael Qiu <michael.qiu at intel.com> > > > --- > > > lib/librte_eal/common/include/arch/x86/rte_memcpy.h | 20 > > > ++++++++++++++++++++ > > > 1 file changed, 20 insertions(+) > > > > > > diff --git a/lib/librte_eal/common/include/arch/x86/rte_memcpy.h > > > b/lib/librte_eal/common/include/arch/x86/rte_memcpy.h > > > index ac72069..bd10d36 100644 > > > --- a/lib/librte_eal/common/include/arch/x86/rte_memcpy.h > > > +++ b/lib/librte_eal/common/include/arch/x86/rte_memcpy.h > > > @@ -43,7 +43,27 @@ > > > #include <stdio.h> > > > #include <stdint.h> > > > #include <string.h> > > > +#if (defined(__ICC) || (__GNUC__ == 4 && __GNUC_MINOR__ < 4)) > > > + > > > +#ifdef __SSE__ > > > +#include <xmmintrin.h> > > > +#endif > > > + > > > +#ifdef __SSE2__ > > > +#include <emmintrin.h> > > > +#endif > > > + > > > +#if defined(__SSE4_2__) || defined(__SSE4_1__) > > > +#include <smmintrin.h> > > > +#endif > > > + > > > +#if defined(__AVX__) > > > +#include <immintrin.h> > > > +#endif > > > + > > > +#else > > > #include <x86intrin.h> > > > +#endif > > > > > > #ifdef __cplusplus > > > extern "C" { > > > -- > > > 1.9.3 > > > > Wonder why to spread this thing over? > > Why not just #include <rte_common_vec.h>? > > I agree to use rte_common_vec.h. > I pushed this approach to check every tested compilers: > http://dpdk.org/browse/dpdk/commit/?id=9a01c31b94e54384 > > I think that rte_common_vec.h should be moved into > lib/librte_eal/common/include/arch/x86/rte_vec.h as it's x86-specific.
Seems like a good thing to do. Do you want me to submit a patch for 2.0 for that? Konstantin