eal: switch to architecture specific pause function

Remove rte_pause() definition from rte_common.h and
switchover to architecture specific rte_pause.h

Signed-off-by: Jerin Jacob <jerin.jacob@caviumnetworks.com>

examples/distributor/main.c

@@ -43,6 +43,7 @@
 #include <rte_debug.h>
 #include <rte_prefetch.h>
 #include <rte_distributor.h>
+#include <rte_pause.h>
 
 #define RX_RING_SIZE 512
 #define TX_RING_SIZE 512

examples/l2fwd-jobstats/main.c

@@ -67,6 +67,7 @@
 #include <rte_jobstats.h>
 #include <rte_timer.h>
 #include <rte_alarm.h>
+#include <rte_pause.h>
 
 #define RTE_LOGTYPE_L2FWD RTE_LOGTYPE_USER1
 

examples/performance-thread/l3fwd-thread/main.c

@@ -73,6 +73,7 @@
 #include <rte_tcp.h>
 #include <rte_udp.h>
 #include <rte_string_fns.h>
+#include <rte_pause.h>
 
 #include <cmdline_parse.h>
 #include <cmdline_parse_etheraddr.h>

examples/tep_termination/main.c

@@ -50,6 +50,7 @@
 #include <rte_string_fns.h>
 #include <rte_malloc.h>
 #include <rte_vhost.h>
+#include <rte_pause.h>
 
 #include "main.h"
 #include "vxlan.h"

examples/vhost/main.c

@@ -52,6 +52,7 @@
 #include <rte_vhost.h>
 #include <rte_ip.h>
 #include <rte_tcp.h>
+#include <rte_pause.h>
 
 #include "main.h"
 

examples/vhost_xen/main.c

@@ -48,6 +48,7 @@
 #include <rte_ethdev.h>
 #include <rte_log.h>
 #include <rte_string_fns.h>
+#include <rte_pause.h>
 
 #include "main.h"
 #include "virtio-net.h"

lib/librte_distributor/rte_distributor.c

@@ -42,6 +42,8 @@
 #include <rte_string_fns.h>
 #include <rte_eal_memconfig.h>
 #include <rte_compat.h>
+#include <rte_pause.h>
+
 #include "rte_distributor_private.h"
 #include "rte_distributor.h"
 #include "rte_distributor_v20.h"

lib/librte_distributor/rte_distributor_v20.c

@@ -41,6 +41,8 @@
 #include <rte_compat.h>
 #include <rte_string_fns.h>
 #include <rte_eal_memconfig.h>
+#include <rte_pause.h>
+
 #include "rte_distributor_v20.h"
 #include "rte_distributor_private.h"
 

lib/librte_eal/common/eal_common_launch.c

@@ -41,6 +41,7 @@
 #include <rte_memzone.h>
 #include <rte_eal.h>
 #include <rte_atomic.h>
+#include <rte_pause.h>
 #include <rte_per_lcore.h>
 #include <rte_lcore.h>
 

lib/librte_eal/common/eal_common_timer.c

@@ -41,6 +41,7 @@
 #include <rte_common.h>
 #include <rte_log.h>
 #include <rte_cycles.h>
+#include <rte_pause.h>
 
 #include "eal_private.h"
 

lib/librte_eal/common/include/arch/ppc_64/rte_spinlock.h

@@ -38,6 +38,7 @@ extern "C" {
 #endif
 
 #include <rte_common.h>
+#include <rte_pause.h>
 #include "generic/rte_spinlock.h"
 
 /* Fixme: Use intrinsics to implement the spinlock on Power architecture */

lib/librte_eal/common/include/arch/x86/rte_spinlock.h

@@ -43,6 +43,7 @@ extern "C" {
 #include "rte_cpuflags.h"
 #include "rte_branch_prediction.h"
 #include "rte_common.h"
+#include "rte_pause.h"
 
 #define RTE_RTM_MAX_RETRIES (10)
 #define RTE_XABORT_LOCK_BUSY (0xff)

lib/librte_eal/common/include/generic/rte_rwlock.h

@@ -52,6 +52,7 @@ extern "C" {
 
 #include <rte_common.h>
 #include <rte_atomic.h>
+#include <rte_pause.h>
 
 /**
  * The rte_rwlock_t type.

lib/librte_eal/common/include/generic/rte_spinlock.h

@@ -51,6 +51,7 @@
 #ifdef RTE_FORCE_INTRINSICS
 #include <rte_common.h>
 #endif
+#include <rte_pause.h>
 
 /**
  * The rte_spinlock_t type.

lib/librte_eal/common/include/rte_common.h

@@ -304,21 +304,6 @@ rte_align64pow2(uint64_t v)
 
 /*********** Other general functions / macros ********/
 
-#ifdef __SSE2__
-#include <emmintrin.h>
-/**
- * PAUSE instruction for tight loops (avoid busy waiting)
- */
-static inline void
-rte_pause (void)
-{
-	_mm_pause();
-}
-#else
-static inline void
-rte_pause(void) {}
-#endif
-
 /**
  * Searches the input parameter for the least significant set bit
  * (starting from zero).

lib/librte_eal/common/include/rte_eal_memconfig.h

@@ -39,6 +39,7 @@
 #include <rte_memzone.h>
 #include <rte_malloc_heap.h>
 #include <rte_rwlock.h>
+#include <rte_pause.h>
 
 #ifdef __cplusplus
 extern "C" {

lib/librte_eal/linuxapp/eal/eal_interrupts.c

@@ -64,6 +64,7 @@
 #include <rte_malloc.h>
 #include <rte_errno.h>
 #include <rte_spinlock.h>
+#include <rte_pause.h>
 
 #include "eal_private.h"
 #include "eal_vfio.h"

lib/librte_hash/rte_cuckoo_hash.c

@@ -57,6 +57,7 @@
 #include <rte_spinlock.h>
 #include <rte_ring.h>
 #include <rte_compat.h>
+#include <rte_pause.h>
 
 #include "rte_hash.h"
 #include "rte_cuckoo_hash.h"

lib/librte_ring/rte_ring.h

@@ -101,6 +101,7 @@ extern "C" {
 #include <rte_atomic.h>
 #include <rte_branch_prediction.h>
 #include <rte_memzone.h>
+#include <rte_pause.h>
 
 #define RTE_TAILQ_RING_NAME "RTE_RING"
 

lib/librte_timer/rte_timer.c

@@ -51,6 +51,7 @@
 #include <rte_branch_prediction.h>
 #include <rte_spinlock.h>
 #include <rte_random.h>
+#include <rte_pause.h>
 
 #include "rte_timer.h"
 

test/test/test_common.c

@@ -35,6 +35,7 @@
 #include <string.h>
 #include <rte_common.h>
 #include <rte_hexdump.h>
+#include <rte_pause.h>
 
 #include "test.h"
 

test/test/test_cryptodev.c

@@ -35,6 +35,7 @@
 #include <rte_mbuf.h>
 #include <rte_malloc.h>
 #include <rte_memcpy.h>
+#include <rte_pause.h>
 
 #include <rte_crypto.h>
 #include <rte_cryptodev.h>

test/test/test_cryptodev_blockcipher.c

@@ -35,6 +35,7 @@
 #include <rte_mbuf.h>
 #include <rte_malloc.h>
 #include <rte_memcpy.h>
+#include <rte_pause.h>
 
 #include <rte_crypto.h>
 #include <rte_cryptodev.h>

test/test/test_distributor_perf.c

@@ -40,6 +40,7 @@
 #include <rte_common.h>
 #include <rte_mbuf.h>
 #include <rte_distributor.h>
+#include <rte_pause.h>
 
 #define ITER_POWER_CL 25 /* log 2 of how many iterations  for Cache Line test */
 #define ITER_POWER 21 /* log 2 of how many iterations we do when timing. */

test/test/test_eventdev_sw.c

@@ -47,8 +47,9 @@
 #include <rte_debug.h>
 #include <rte_ethdev.h>
 #include <rte_cycles.h>
-
 #include <rte_eventdev.h>
+#include <rte_pause.h>
+
 #include "test.h"
 
 #define MAX_PORTS 16

test/test/test_ring_perf.c

@@ -37,6 +37,7 @@
 #include <rte_ring.h>
 #include <rte_cycles.h>
 #include <rte_launch.h>
+#include <rte_pause.h>
 
 #include "test.h"
 

test/test/test_timer.c

@@ -136,6 +136,7 @@
 #include <rte_timer.h>
 #include <rte_random.h>
 #include <rte_malloc.h>
+#include <rte_pause.h>
 
 #define TEST_DURATION_S 1 /* in seconds */
 #define NB_TIMER 4

test/test/test_timer_perf.c

@@ -42,6 +42,7 @@
 #include <rte_lcore.h>
 #include <rte_random.h>
 #include <rte_malloc.h>
+#include <rte_pause.h>
 
 #define MAX_ITERATIONS 1000000
 

test/test/test_timer_racecond.c

@@ -42,6 +42,7 @@
 #include <rte_lcore.h>
 #include <rte_random.h>
 #include <rte_malloc.h>
+#include <rte_pause.h>
 
 #undef TEST_TIMER_RACECOND_VERBOSE