Add support for multiple microarchitectures in single binary

 * compiler -march= parameter is changed from native to corei7
   so code is always genereted with instructions which are available
   on the Nehalem microarchitecture (up to SSE4.2)

 * compiler -mtune= parameter is added so code is optimized for
   corei7-avx which equals to Sandy Bridge microarchitecture

 * set of macros is added which allows run-time detection of available
   cpu instructions (e.g. clib_cpu_supports_avx())

 * set of macros is added which allows us to clone graph node funcitons
   where cloned function is optmized for different microarchitecture
   Those macros are using following attributes:
     __attribute__((flatten))
     __attribute__((target("arch=core-avx2)))

   I.e. If applied to foo_node_fn() macro will generate cloned
   functions foo_node_fn_avx2() and foo_node_fn_avx512() (future)
   It will also generate function void * foo_node_fn_multiarch_select()
   which detects available instruction set and returns pointer to the
   best matching function clone.

Change-Id: I2dce0ac92a5ede95fcb56f47f3d1f3c4c040bac0
Signed-off-by: Damjan Marion <damarion@cisco.com>

1 files changed, 16 insertions, 0 deletions

diff --git a/vnet/vnet/ip/ip6_forward.c b/vnet/vnet/ip/ip6_forward.c
index a136da3e142..7093c571e1f 100644
--- a/vnet/vnet/ip/ip6_forward.c
+++ b/vnet/vnet/ip/ip6_forward.c
@@ -1260,6 +1260,8 @@ VLIB_REGISTER_NODE (ip6_lookup_node) = {
   .next_nodes = IP6_LOOKUP_NEXT_NODES,
 };
 
+VLIB_NODE_FUNCTION_MULTIARCH (ip6_lookup_node, ip6_lookup)
+
 static uword
 ip6_indirect (vlib_main_t * vm,
 	      vlib_node_runtime_t * node,
@@ -1278,6 +1280,8 @@ VLIB_REGISTER_NODE (ip6_indirect_node) = {
   .next_nodes = IP6_LOOKUP_NEXT_NODES,
 };
 
+VLIB_NODE_FUNCTION_MULTIARCH (ip6_indirect_node, ip6_indirect)
+
 typedef struct {
   /* Adjacency taken. */
   u32 adj_index;
@@ -1454,6 +1458,8 @@ VLIB_REGISTER_NODE (ip6_drop_node,static) = {
   },
 };
 
+VLIB_NODE_FUNCTION_MULTIARCH (ip6_drop_node, ip6_drop)
+
 VLIB_REGISTER_NODE (ip6_punt_node,static) = {
   .function = ip6_punt,
   .name = "ip6-punt",
@@ -1467,6 +1473,8 @@ VLIB_REGISTER_NODE (ip6_punt_node,static) = {
   },
 };
 
+VLIB_NODE_FUNCTION_MULTIARCH (ip6_punt_node, ip6_punt)
+
 VLIB_REGISTER_NODE (ip6_miss_node,static) = {
   .function = ip6_miss,
   .name = "ip6-miss",
@@ -1480,6 +1488,8 @@ VLIB_REGISTER_NODE (ip6_miss_node,static) = {
   },
 };
 
+VLIB_NODE_FUNCTION_MULTIARCH (ip6_miss_node, ip6_miss)
+
 VLIB_REGISTER_NODE (ip6_multicast_node,static) = {
   .function = ip6_drop,
   .name = "ip6-multicast",
@@ -1858,6 +1868,8 @@ VLIB_REGISTER_NODE (ip6_local_node,static) = {
   },
 };
 
+VLIB_NODE_FUNCTION_MULTIARCH (ip6_local_node, ip6_local)
+
 void ip6_register_protocol (u32 protocol, u32 node_index)
 {
   vlib_main_t * vm = vlib_get_main();
@@ -2424,6 +2436,8 @@ VLIB_REGISTER_NODE (ip6_rewrite_node) = {
   },
 };
 
+VLIB_NODE_FUNCTION_MULTIARCH (ip6_rewrite_node, ip6_rewrite_transit)
+
 VLIB_REGISTER_NODE (ip6_rewrite_local_node,static) = {
   .function = ip6_rewrite_local,
   .name = "ip6-rewrite-local",
@@ -2439,6 +2453,8 @@ VLIB_REGISTER_NODE (ip6_rewrite_local_node,static) = {
   },
 };
 
+VLIB_NODE_FUNCTION_MULTIARCH (ip6_rewrite_local_node, ip6_rewrite_local)
+
 /* Global IP6 main. */
 ip6_main_t ip6_main;