conv_acc: Our code requires SSSE3, not just SSE3
The accelerated convolutional decoder uses SSSE3 instructions such
as PSIGNW (via _mm_sign_epi16) which go beyond what SSE3 offers. So
let's make sure we use the right compiler flag (-mssse3) and also the
right runtime check.
Without this patch, we would use illegal instructions e.g. on Opteron
Gen3 such as Opteron 2427, which are also used as build.opensuse.org
build hosts (build31 through build36) where we wouldn't pass "make
check" as a result.
Change-Id: I2754164384109f2821fd98ffb48f625893f2923d
Fixes: OS#2386
diff --git a/src/Makefile.am b/src/Makefile.am
index e7f94ce..3d6e6f7 100644
--- a/src/Makefile.am
+++ b/src/Makefile.am
@@ -23,20 +23,20 @@
macaddr.c stat_item.c stats.c stats_statsd.c prim.c \
conv_acc.c conv_acc_generic.c sercomm.c prbs.c
-if HAVE_SSE3
+if HAVE_SSSE3
libosmocore_la_SOURCES += conv_acc_sse.c
if HAVE_SSE4_1
-conv_acc_sse.lo : AM_CFLAGS += -msse3 -msse4.1
+conv_acc_sse.lo : AM_CFLAGS += -mssse3 -msse4.1
else
-conv_acc_sse.lo : AM_CFLAGS += -msse3
+conv_acc_sse.lo : AM_CFLAGS += -mssse3
endif
if HAVE_AVX2
libosmocore_la_SOURCES += conv_acc_sse_avx.c
if HAVE_SSE4_1
-conv_acc_sse_avx.lo : AM_CFLAGS += -msse3 -mavx2 -msse4.1
+conv_acc_sse_avx.lo : AM_CFLAGS += -mssse3 -mavx2 -msse4.1
else
-conv_acc_sse_avx.lo : AM_CFLAGS += -msse3 -mavx2
+conv_acc_sse_avx.lo : AM_CFLAGS += -mssse3 -mavx2
endif
endif
endif