On 6/17/2018 7:19 PM, Jerin Jacob wrote:
-----Original Message-----
Date: Fri, 8 Jun 2018 23:45:18 +0530
From: Nikhil Rao <nikhil....@intel.com>
To: jerin.ja...@caviumnetworks.com
CC: dev@dpdk.org, Nikhil Rao <nikhil....@intel.com>
Subject: [PATCH v1 4/4] eventdev: add interrupt driven queues in Rx event
adapter
X-Mailer: git-send-email 1.8.3.1
Add support for interrupt driven queues when eth device is
configured for rxq interrupts and servicing weight for the
queue is configured to be zero.
A interrupt driven packet received counter has been added to
rte_event_eth_rx_adapter_stats.
Signed-off-by: Nikhil Rao <nikhil....@intel.com>
---
lib/librte_eventdev/rte_event_eth_rx_adapter.h | 5 +-
lib/librte_eventdev/rte_event_eth_rx_adapter.c | 1049 +++++++++++++++++++-
test/test/test_event_eth_rx_adapter.c | 261 ++++-
Please move the testcase to separate patch.
.../prog_guide/event_ethernet_rx_adapter.rst | 24 +
config/common_base | 1 +
This patch creates build issue with meson build.
command to reproduce:
--------------------
export MESON_PARAMS='-Dwerror=true -Dexamples=all'
CC="ccache gcc" meson --default-library=shared $MESON_PARAMS gcc-shared-build
ninja -C gcc-shared-build
log:
---
../lib/librte_eventdev/rte_event_eth_rx_adapter.c: In function
‘rxa_intr_ring_check_avail’:
../lib/librte_eventdev/rte_event_eth_rx_adapter.c:916:5: error:
‘RTE_EVENT_ETH_INTR_RING_SIZE’ undeclared (first use in this function);
did you mean ‘RTE_EVENT_DEV_XSTATS_NAME_SIZE’?
RTE_EVENT_ETH_INTR_RING_SIZE) {
^~~~~~~~~~~~~~~~~~~~~~~~~~~~
RTE_EVENT_DEV_XSTATS_NAME_SIZE
../lib/librte_eventdev/rte_event_eth_rx_adapter.c:916:5: note: each
undeclared identifier is reported only once for each function it appears
in
../lib/librte_eventdev/rte_event_eth_rx_adapter.c: In function
‘rxa_intr_thread’:
../lib/librte_eventdev/rte_event_eth_rx_adapter.c:971:8: error:
‘RTE_EVENT_ETH_INTR_RING_SIZE’ undeclared (first use in this function);
did you mean ‘RTE_EVENT_DEV_XSTATS_NAME_SIZE’?
RTE_EVENT_ETH_INTR_RING_SIZE + 1, -1);
^~~~~~~~~~~~~~~~~~~~~~~~~~~~
6 files changed, 1296 insertions(+), 48 deletions(-)
diff --git a/lib/librte_eventdev/rte_event_eth_rx_adapter.h
b/lib/librte_eventdev/rte_event_eth_rx_adapter.h
index 307b2b5..97f25e9 100644
--- a/lib/librte_eventdev/rte_event_eth_rx_adapter.h
+++ b/lib/librte_eventdev/rte_event_eth_rx_adapter.h
@@ -64,8 +64,7 @@
* the service function ID of the adapter in this case.
*
* Note:
- * 1) Interrupt driven receive queues are currently unimplemented.
- * 2) Devices created after an instance of rte_event_eth_rx_adapter_create
+ * 1) Devices created after an instance of rte_event_eth_rx_adapter_create
* should be added to a new instance of the rx adapter.
Can we remove this NOTE and add this check in the code if it is not the
case?
OK.
*/
@@ -199,6 +198,8 @@ struct rte_event_eth_rx_adapter_stats {
* block cycles can be used to compute the percentage of
* cycles the service is blocked by the event device.
*/
+ uint64_t rx_intr_packets;
+ /**< Received packet count for interrupt mode Rx queues */
};
/**
diff --git a/lib/librte_eventdev/rte_event_eth_rx_adapter.c
b/lib/librte_eventdev/rte_event_eth_rx_adapter.c
index 40e9bc9..d038ee4 100644
--- a/lib/librte_eventdev/rte_event_eth_rx_adapter.c
+++ b/lib/librte_eventdev/rte_event_eth_rx_adapter.c
@@ -2,6 +2,8 @@
* Copyright(c) 2017 Intel Corporation.
* All rights reserved.
*/
+#include <unistd.h>
+#include <sys/epoll.h>
#include <rte_cycles.h>
#include <rte_common.h>
#include <rte_dev.h>
@@ -11,6 +13,7 @@
#include <rte_malloc.h>
#include <rte_service_component.h>
#include <rte_thash.h>
+#include <rte_interrupts.h>
#include "rte_eventdev.h"
#include "rte_eventdev_pmd.h"
@@ -24,6 +27,36 @@
#define ETH_RX_ADAPTER_MEM_NAME_LEN 32
+static void *
+rxa_intr_thread(void *arg)
+{
+ struct rte_event_eth_rx_adapter *rx_adapter = arg;
+ struct rte_epoll_event *epoll_events = rx_adapter->epoll_events;
+ int n, i;
+ uint8_t val;
+ ssize_t bytes_read;
+
+ while (1) {
+ n = rte_epoll_wait(rx_adapter->epd, epoll_events,
+ RTE_EVENT_ETH_INTR_RING_SIZE + 1, -1);
Can you check with FreeBSD if everything is fine or not?
Interrupt functionality works only on Linux (rte_epoll_wait() etc are
implemented only in Linux.) or I am not understanding your question
correctly ?
+ if (unlikely(n < 0))
+ RTE_EDEV_LOG_ERR("rte_epoll_wait returned error %d",
+ n);
+ for (i = 0; i < n; i++) {
+ if (epoll_events[i].fd == rx_adapter->intr_pipe.readfd)
+ goto done;
+ rxa_intr_ring_enqueue(rx_adapter,
+ epoll_events[i].epdata.data);
+ }
+ }
+
+done:
+
+static int
+rxa_create_intr_thread(struct rte_event_eth_rx_adapter *rx_adapter)
+{
+ int err;
+ uint8_t val;
+ char thread_name[RTE_MAX_THREAD_NAME_LEN];
+
+ if (rx_adapter->intr_ring)
+ return 0;
+
+ rx_adapter->intr_ring = rte_ring_create("intr_ring",
+ RTE_EVENT_ETH_INTR_RING_SIZE,
+ rte_socket_id(), 0);
+ if (!rx_adapter->intr_ring)
+ return -ENOMEM;
+
+ rx_adapter->epoll_events = rte_zmalloc_socket(rx_adapter->mem_name,
+ (RTE_EVENT_ETH_INTR_RING_SIZE + 1) *
+ sizeof(struct rte_epoll_event),
+ RTE_CACHE_LINE_SIZE,
+ rx_adapter->socket_id);
+ if (!rx_adapter->epoll_events) {
+ err = -ENOMEM;
+ goto error;
+ }
+
+ rte_spinlock_init(&rx_adapter->intr_ring_lock);
+
+ snprintf(thread_name, RTE_MAX_THREAD_NAME_LEN,
+ "rx-intr-thread");
+ err = pthread_create(&rx_adapter->rx_intr_thread, NULL,
+ rxa_intr_thread, rx_adapter);
Can you replace the pthread_* with new rte_ctrl_thread_create()
abstraction ?
OK.
#
diff --git a/lib/librte_eventdev/Makefile b/lib/librte_eventdev/Makefile
index b3e2546..e269357 100644
--- a/lib/librte_eventdev/Makefile
+++ b/lib/librte_eventdev/Makefile
@@ -8,14 +8,16 @@ include $(RTE_SDK)/mk/rte.vars.mk
LIB = librte_eventdev.a
# library version
-LIBABIVER := 4
+LIBABIVER := 5
# build flags
CFLAGS += -DALLOW_EXPERIMENTAL_API
+CFLAGS += -D_GNU_SOURCE
CFLAGS += -O3
CFLAGS += $(WERROR_FLAGS)
LDLIBS += -lrte_eal -lrte_ring -lrte_ethdev -lrte_hash -lrte_mempool
-lrte_timer
LDLIBS += -lrte_mbuf -lrte_cryptodev
+LDLIBS += -lpthread
This may not be required if we add rte_ctrl_thread library.
# library source files
SRCS-y += rte_eventdev.c
--
1.8.3.1
Thanks for the review.
Nikhil
