[dpdk-dev] [PATCH v9 0/5] link bonding
Hi Declan, I've made some adjustments in your patches. * Some renaming: - RTE_PMD_BOND / PMD_BOND_KVARG -> PMD_BOND_NAME - rte_pmd_bond_init -> bond_init - rte_pmd_bond_drv -> bond_drv - rte_eth_bond_vargs.c -> rte_eth_bond_args.c * Some moves in doxygen configuration * Removed rte_snprintf usages * Reword commit logs bond: new link bonding library ethdev: add unique name to devices eal: support link bonding device initialization bond: unit tests bond: testpmd support But there is still something wrong in the Makefile: the copyright is missing. I cannot add it by myself, so please send a v10 based on this v9 with copyright added. Thanks
[dpdk-dev] [PATCH v9 1/5] bond: new link bonding library
From: Declan Doherty Initial release with support for Mode 0 - Round Robin Mode 1 - Active Backup Mode 2 - Balance -> Supports 3 transmit polices (layer 2, layer 2+3, layer 3+4) Mode 3 - Broadcast Signed-off-by: Declan Doherty Acked-by: Pablo de Lara --- config/common_bsdapp |5 + config/common_linuxapp |5 + doc/doxy-api-index.md |1 + doc/doxy-api.conf |1 + lib/Makefile |1 + lib/librte_pmd_bond/Makefile | 32 + lib/librte_pmd_bond/rte_eth_bond.h | 255 ++ lib/librte_pmd_bond/rte_eth_bond_api.c | 662 +++ lib/librte_pmd_bond/rte_eth_bond_args.c| 252 ++ lib/librte_pmd_bond/rte_eth_bond_pmd.c | 1212 lib/librte_pmd_bond/rte_eth_bond_private.h | 215 + mk/rte.app.mk |4 + 12 files changed, 2645 insertions(+) create mode 100644 lib/librte_pmd_bond/Makefile create mode 100644 lib/librte_pmd_bond/rte_eth_bond.h create mode 100644 lib/librte_pmd_bond/rte_eth_bond_api.c create mode 100644 lib/librte_pmd_bond/rte_eth_bond_args.c create mode 100644 lib/librte_pmd_bond/rte_eth_bond_pmd.c create mode 100644 lib/librte_pmd_bond/rte_eth_bond_private.h diff --git a/config/common_bsdapp b/config/common_bsdapp index d5db4ab..c243b0c 100644 --- a/config/common_bsdapp +++ b/config/common_bsdapp @@ -206,6 +206,11 @@ CONFIG_RTE_PMD_RING_MAX_TX_RINGS=16 CONFIG_RTE_LIBRTE_PMD_PCAP=y # +# Compile link bonding PMD library +# +CONFIG_RTE_LIBRTE_PMD_BOND=y + +# # Do prefetch of packet data within PMD driver receive function # CONFIG_RTE_PMD_PACKET_PREFETCH=y diff --git a/config/common_linuxapp b/config/common_linuxapp index c5c0cb6..6237ce5 100644 --- a/config/common_linuxapp +++ b/config/common_linuxapp @@ -244,6 +244,11 @@ CONFIG_RTE_PMD_RING_MAX_TX_RINGS=16 CONFIG_RTE_LIBRTE_PMD_PCAP=n # +# Compile link bonding PMD library +# +CONFIG_RTE_LIBRTE_PMD_BOND=y + +# # Compile Xen PMD # CONFIG_RTE_LIBRTE_PMD_XENVIRT=n diff --git a/doc/doxy-api-index.md b/doc/doxy-api-index.md index 7b26e98..d72659a 100644 --- a/doc/doxy-api-index.md +++ b/doc/doxy-api-index.md @@ -38,6 +38,7 @@ There are many libraries, so their headers may be grouped by topics: - **device**: [ethdev] (@ref rte_ethdev.h), [devargs](@ref rte_devargs.h), + [bond] (@ref rte_eth_bond.h), [KNI](@ref rte_kni.h), [PCI](@ref rte_pci.h), [PCI IDs](@ref rte_pci_dev_ids.h) diff --git a/doc/doxy-api.conf b/doc/doxy-api.conf index f380d9a..1fd4492 100644 --- a/doc/doxy-api.conf +++ b/doc/doxy-api.conf @@ -47,6 +47,7 @@ INPUT = doc/doxy-api-index.md \ lib/librte_pipeline \ lib/librte_port \ lib/librte_power \ + lib/librte_pmd_bond \ lib/librte_ring \ lib/librte_sched \ lib/librte_table \ diff --git a/lib/Makefile b/lib/Makefile index c58c0c9..10c5bb3 100644 --- a/lib/Makefile +++ b/lib/Makefile @@ -44,6 +44,7 @@ DIRS-$(CONFIG_RTE_LIBRTE_ETHER) += librte_ether DIRS-$(CONFIG_RTE_LIBRTE_E1000_PMD) += librte_pmd_e1000 DIRS-$(CONFIG_RTE_LIBRTE_IXGBE_PMD) += librte_pmd_ixgbe DIRS-$(CONFIG_RTE_LIBRTE_I40E_PMD) += librte_pmd_i40e +DIRS-$(CONFIG_RTE_LIBRTE_PMD_BOND) += librte_pmd_bond DIRS-$(CONFIG_RTE_LIBRTE_PMD_RING) += librte_pmd_ring DIRS-$(CONFIG_RTE_LIBRTE_PMD_PCAP) += librte_pmd_pcap DIRS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD) += librte_pmd_virtio diff --git a/lib/librte_pmd_bond/Makefile b/lib/librte_pmd_bond/Makefile new file mode 100644 index 000..6960551 --- /dev/null +++ b/lib/librte_pmd_bond/Makefile @@ -0,0 +1,32 @@ +# + +include $(RTE_SDK)/mk/rte.vars.mk + +# +# library name +# +LIB = librte_pmd_bond.a + +CFLAGS += -O3 +CFLAGS += $(WERROR_FLAGS) + +# +# all source are stored in SRCS-y +# +SRCS-$(CONFIG_RTE_LIBRTE_PMD_BOND) += rte_eth_bond_api.c +SRCS-$(CONFIG_RTE_LIBRTE_PMD_BOND) += rte_eth_bond_pmd.c +SRCS-$(CONFIG_RTE_LIBRTE_PMD_BOND) += rte_eth_bond_vargs.c + +# +# Export include files +# +SYMLINK-y-include += rte_eth_bond.h + +# this lib depends upon: +DEPDIRS-$(CONFIG_RTE_LIBRTE_PMD_BOND) += lib/librte_mbuf +DEPDIRS-$(CONFIG_RTE_LIBRTE_PMD_BOND) += lib/librte_ether +DEPDIRS-$(CONFIG_RTE_LIBRTE_PMD_BOND) += lib/librte_malloc +DEPDIRS-$(CONFIG_RTE_LIBRTE_PMD_BOND) += lib/librte_eal +DEPDIRS-$(CONFIG_RTE_LIBRTE_PMD_BOND) += lib/librte_kvargs + +include $(RTE_SDK)/mk/rte.lib.mk diff --git a/lib/librte_pmd_bond/rte_eth_bond.h b/lib/librte_pmd_bond/rte_eth_bond.h new file mode 100644 index 000..95a4fa9 --- /dev/null +++ b/lib/librte_pmd_bond/rte_eth_bond.h @@ -0,0 +1,255 @@ +/*- + * BSD LICENSE + * + * Copyright(c) 2010-2014 Intel Corporation. All rights reserved. + * All r
[dpdk-dev] [PATCH v9 2/5] ethdev: add unique name to devices
From: Declan Doherty
Adding support to rte_eth_dev_data structure to support unique
name identifier for ethdevs to support adding slave ethdevs
(specifically virtual devices which have no public unique
identifier) to a link bonding device. This changes the API
rte_eth_dev_allocate() to require a const char *name when
allocating a ethdev, which also verifies that the name is
unique and hasn?t been already used by an existed allocated
rte_eth_dev. Also contains updates to virtual pmd?s to now call
the API with a name parameter.
Signed-off-by: Declan Doherty
Acked-by: Pablo de Lara
---
lib/librte_ether/rte_ethdev.c| 31 ---
lib/librte_ether/rte_ethdev.h| 7 ++-
lib/librte_pmd_pcap/rte_eth_pcap.c | 22 +++---
lib/librte_pmd_ring/rte_eth_ring.c | 32 +---
lib/librte_pmd_ring/rte_eth_ring.h | 3 ++-
lib/librte_pmd_xenvirt/rte_eth_xenvirt.c | 2 +-
6 files changed, 65 insertions(+), 32 deletions(-)
diff --git a/lib/librte_ether/rte_ethdev.c b/lib/librte_ether/rte_ethdev.c
index 8687499..a21cdb9 100644
--- a/lib/librte_ether/rte_ethdev.c
+++ b/lib/librte_ether/rte_ethdev.c
@@ -65,6 +65,7 @@
#include
#include
#include
+#include
#include "rte_ether.h"
#include "rte_ethdev.h"
@@ -153,21 +154,39 @@ rte_eth_dev_data_alloc(void)
RTE_MAX_ETHPORTS * sizeof(*rte_eth_dev_data));
}
+static struct rte_eth_dev *
+rte_eth_dev_allocated(const char *name)
+{
+ unsigned i;
+
+ for (i = 0; i < nb_ports; i++) {
+ if (strcmp(rte_eth_devices[i].data->name, name) == 0)
+ return &rte_eth_devices[i];
+ }
+ return NULL;
+}
+
struct rte_eth_dev *
-rte_eth_dev_allocate(void)
+rte_eth_dev_allocate(const char *name)
{
struct rte_eth_dev *eth_dev;
if (nb_ports == RTE_MAX_ETHPORTS) {
- PMD_DEBUG_TRACE("Reached maximum number of ethernet ports\n");
+ PMD_DEBUG_TRACE("Reached maximum number of Ethernet ports\n");
return NULL;
}
if (rte_eth_dev_data == NULL)
rte_eth_dev_data_alloc();
+ if (rte_eth_dev_allocated(name) != NULL) {
+ PMD_DEBUG_TRACE("Ethernet Device with name %s already
allocated!\n");
+ return NULL;
+ }
+
eth_dev = &rte_eth_devices[nb_ports];
eth_dev->data = &rte_eth_dev_data[nb_ports];
+ snprintf(eth_dev->data->name, sizeof(eth_dev->data->name), "%s", name);
eth_dev->data->port_id = nb_ports++;
return eth_dev;
}
@@ -178,11 +197,17 @@ rte_eth_dev_init(struct rte_pci_driver *pci_drv,
{
struct eth_driver*eth_drv;
struct rte_eth_dev *eth_dev;
+ char ethdev_name[RTE_ETH_NAME_MAX_LEN];
+
int diag;
eth_drv = (struct eth_driver *)pci_drv;
- eth_dev = rte_eth_dev_allocate();
+ /* Create unique Ethernet device name using PCI address */
+ snprintf(ethdev_name, RTE_ETH_NAME_MAX_LEN, "%d:%d.%d",
+ pci_dev->addr.bus, pci_dev->addr.devid,
pci_dev->addr.function);
+
+ eth_dev = rte_eth_dev_allocate(ethdev_name);
if (eth_dev == NULL)
return -ENOMEM;
diff --git a/lib/librte_ether/rte_ethdev.h b/lib/librte_ether/rte_ethdev.h
index 526331a..25f2df7 100644
--- a/lib/librte_ether/rte_ethdev.h
+++ b/lib/librte_ether/rte_ethdev.h
@@ -1497,6 +1497,8 @@ struct rte_eth_dev_sriov {
};
#define RTE_ETH_DEV_SRIOV(dev) ((dev)->data->sriov)
+#define RTE_ETH_NAME_MAX_LEN (32)
+
/**
* @internal
* The data part, with no function pointers, associated with each ethernet
device.
@@ -1505,6 +1507,8 @@ struct rte_eth_dev_sriov {
* processes in a multi-process configuration.
*/
struct rte_eth_dev_data {
+ char name[RTE_ETH_NAME_MAX_LEN]; /**< Unique identifier name */
+
void **rx_queues; /**< Array of pointers to RX queues. */
void **tx_queues; /**< Array of pointers to TX queues. */
uint16_t nb_rx_queues; /**< Number of RX queues. */
@@ -1560,10 +1564,11 @@ extern uint8_t rte_eth_dev_count(void);
* Allocates a new ethdev slot for an ethernet device and returns the pointer
* to that slot for the driver to use.
*
+ * @param nameUnique identifier name for each Ethernet device
* @return
* - Slot in the rte_dev_devices array for a new device;
*/
-struct rte_eth_dev *rte_eth_dev_allocate(void);
+struct rte_eth_dev *rte_eth_dev_allocate(const char *name);
struct eth_driver;
/**
diff --git a/lib/librte_pmd_pcap/rte_eth_pcap.c
b/lib/librte_pmd_pcap/rte_eth_pcap.c
index b3dbbda..12b7e0c 100644
--- a/lib/librte_pmd_pcap/rte_eth_pcap.c
+++ b/lib/librte_pmd_pcap/rte_eth_pcap.c
@@ -534,7 +534,7 @@ open_tx_iface(const char *key __rte_unused, const char
*value, void *extra_args)
static int
-rte_pmd_init_internals(const unsigned nb_rx_queues,
+rte_pmd_init_internals(const char *name, const un
[dpdk-dev] [PATCH v9 3/5] eal: support link bonding device initialization
From: Declan Doherty
Updating functionality in EAL to support adding link bonding
devices via ?vdev option. Link bonding devices will be
initialized after all physical devices have been probed and
initialized.
Signed-off-by: Declan Doherty
Acked-by: Pablo de Lara
---
lib/librte_eal/bsdapp/eal/eal.c | 10 -
lib/librte_eal/common/eal_common_dev.c | 58 -
lib/librte_eal/common/eal_common_pci.c | 3 ++
lib/librte_eal/common/include/eal_private.h | 7
lib/librte_eal/common/include/rte_dev.h | 14 ++-
lib/librte_eal/linuxapp/eal/eal.c | 11 +-
6 files changed, 74 insertions(+), 29 deletions(-)
diff --git a/lib/librte_eal/bsdapp/eal/eal.c b/lib/librte_eal/bsdapp/eal/eal.c
index ce0db03..ec57b26 100644
--- a/lib/librte_eal/bsdapp/eal/eal.c
+++ b/lib/librte_eal/bsdapp/eal/eal.c
@@ -919,7 +919,7 @@ rte_eal_init(int argc, char **argv)
rte_eal_mcfg_complete();
- if (rte_eal_dev_init() < 0)
+ if (rte_eal_dev_init(PMD_INIT_PRE_PCI_PROBE) < 0)
rte_panic("Cannot init pmd devices\n");
RTE_LCORE_FOREACH_SLAVE(i) {
@@ -951,6 +951,14 @@ rte_eal_init(int argc, char **argv)
rte_eal_mp_remote_launch(sync_func, NULL, SKIP_MASTER);
rte_eal_mp_wait_lcore();
+ /* Probe & Initialize PCI devices */
+ if (rte_eal_pci_probe())
+ rte_panic("Cannot probe PCI\n");
+
+ /* Initialize any outstanding devices */
+ if (rte_eal_dev_init(PMD_INIT_POST_PCI_PROBE) < 0)
+ rte_panic("Cannot init pmd devices\n");
+
return fctret;
}
diff --git a/lib/librte_eal/common/eal_common_dev.c
b/lib/librte_eal/common/eal_common_dev.c
index eae5656..1194419 100644
--- a/lib/librte_eal/common/eal_common_dev.c
+++ b/lib/librte_eal/common/eal_common_dev.c
@@ -62,7 +62,7 @@ rte_eal_driver_unregister(struct rte_driver *driver)
}
int
-rte_eal_dev_init(void)
+rte_eal_dev_init(uint8_t init_pri)
{
struct rte_devargs *devargs;
struct rte_driver *driver;
@@ -80,30 +80,52 @@ rte_eal_dev_init(void)
continue;
TAILQ_FOREACH(driver, &dev_driver_list, next) {
- if (driver->type != PMD_VDEV)
- continue;
+ /* RTE_DEVTYPE_VIRTUAL can only be a virtual or bonded
device,
+* virtual devices are initialized pre PCI probing and
bonded
+* device are post pci probing */
+ if ((driver->type == PMD_VDEV && init_pri ==
+ PMD_INIT_PRE_PCI_PROBE) ||
+ (driver->type == PMD_BDEV && init_pri ==
+ PMD_INIT_POST_PCI_PROBE)) {
- /* search a driver prefix in virtual device name */
- if (!strncmp(driver->name, devargs->virtual.drv_name,
- strlen(driver->name))) {
- driver->init(devargs->virtual.drv_name,
- devargs->args);
- break;
+ /* search a driver prefix in virtual device
name */
+ if (!strncmp(driver->name,
devargs->virtual.drv_name,
+ strlen(driver->name))) {
+ printf("init (%u) %s\n", init_pri,
devargs->virtual.drv_name);
+ driver->init(devargs->virtual.drv_name,
+ devargs->args);
+ break;
+ }
}
}
- if (driver == NULL) {
- rte_panic("no driver found for %s\n",
- devargs->virtual.drv_name);
+ /* If initializing pre PCI probe, then we don't expect a bonded
driver
+* to be found */
+ if (init_pri == PMD_INIT_PRE_PCI_PROBE &&
+ strncmp(PMD_BOND_NAME,
devargs->virtual.drv_name,
+ strlen(PMD_BOND_NAME)) != 0) {
+ if (driver == NULL) {
+ rte_panic("no driver found for virtual device
%s\n",
+ devargs->virtual.drv_name);
+ }
+ } else if (init_pri == PMD_INIT_POST_PCI_PROBE &&
+ strncmp(PMD_BOND_NAME,
devargs->virtual.drv_name,
+ strlen(PMD_BOND_NAME)) == 0) {
+ if (driver == NULL) {
+ rte_panic("no driver found for bonded device
%s\n",
+ devargs->virtual.drv_name);
+ }
[dpdk-dev] [PATCH v9 4/5] bond: unit tests
From: Declan Doherty Including: - code to generate packet bursts for testing rx and tx functionality of bonded device - virtual/stubbed out ethdev for use as slave ethdev in testing Signed-off-by: Declan Doherty Acked-by: Pablo de Lara --- app/test/Makefile |4 +- app/test/commands.c |7 + app/test/packet_burst_generator.c | 287 +++ app/test/packet_burst_generator.h | 78 + app/test/test.h |1 + app/test/test_link_bonding.c | 3958 + app/test/virtual_pmd.c| 574 ++ app/test/virtual_pmd.h| 74 + 8 files changed, 4982 insertions(+), 1 deletion(-) create mode 100644 app/test/packet_burst_generator.c create mode 100644 app/test/packet_burst_generator.h create mode 100644 app/test/test_link_bonding.c create mode 100644 app/test/virtual_pmd.c create mode 100644 app/test/virtual_pmd.h diff --git a/app/test/Makefile b/app/test/Makefile index 45d0cf2..cec2f2f 100644 --- a/app/test/Makefile +++ b/app/test/Makefile @@ -102,7 +102,9 @@ SRCS-$(CONFIG_RTE_APP_TEST) += test_ivshmem.c SRCS-$(CONFIG_RTE_APP_TEST) += test_distributor.c SRCS-$(CONFIG_RTE_APP_TEST) += test_distributor_perf.c SRCS-$(CONFIG_RTE_APP_TEST) += test_devargs.c - +SRCS-$(CONFIG_RTE_APP_TEST) += virtual_pmd.c +SRCS-$(CONFIG_RTE_APP_TEST) += packet_burst_generator.c +SRCS-$(CONFIG_RTE_APP_TEST) += test_link_bonding.c ifeq ($(CONFIG_RTE_APP_TEST),y) SRCS-$(CONFIG_RTE_LIBRTE_ACL) += test_acl.c SRCS-$(CONFIG_RTE_LIBRTE_PMD_RING) += test_pmd_ring.c diff --git a/app/test/commands.c b/app/test/commands.c index c9dc085..5f23420 100644 --- a/app/test/commands.c +++ b/app/test/commands.c @@ -159,6 +159,10 @@ static void cmd_autotest_parsed(void *parsed_result, ret = test_timer(); if (!strcmp(res->autotest, "timer_perf_autotest")) ret = test_timer_perf(); +#ifdef RTE_LIBRTE_PMD_BOND + if (!strcmp(res->autotest, "link_bonding_autotest")) + ret = test_link_bonding(); +#endif if (!strcmp(res->autotest, "mempool_autotest")) ret = test_mempool(); if (!strcmp(res->autotest, "mempool_perf_autotest")) @@ -227,6 +231,9 @@ cmdline_parse_token_string_t cmd_autotest_autotest = "alarm_autotest#interrupt_autotest#" "version_autotest#eal_fs_autotest#" "cmdline_autotest#func_reentrancy_autotest#" +#ifdef RTE_LIBRTE_PMD_BOND + "link_bonding_autotest#" +#endif "mempool_perf_autotest#hash_perf_autotest#" "memcpy_perf_autotest#ring_perf_autotest#" "red_autotest#meter_autotest#sched_autotest#" diff --git a/app/test/packet_burst_generator.c b/app/test/packet_burst_generator.c new file mode 100644 index 000..5d539f1 --- /dev/null +++ b/app/test/packet_burst_generator.c @@ -0,0 +1,287 @@ +/*- + * BSD LICENSE + * + * Copyright(c) 2010-2014 Intel Corporation. All rights reserved. + * All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * * Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * * Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * * Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + */ + +#include +#include + +#include "packet_burst_generator.h" + +#define UDP_SRC_PORT 1024 +#define UDP_DST_PORT 1024 + + +#define IP_DEFTTL 64 /* from RFC 1340. */ +#define IP_VERSION 0x40 +#define IP_HDRLEN 0x05 /* default IP header length == five 32-bits words. */ +#define IP_VHL_DEF (IP_V
[dpdk-dev] [PATCH v9 5/5] bond: testpmd support
From: Declan Doherty
- Includes the ability to create new bonded devices.
- Add /remove bonding slave devices.
- Interogate bonded device stats/configuration
- Change bonding modes and select balance transmit polices
Signed-off-by: Declan Doherty
Acked-by: Pablo de Lara
---
app/test-pmd/cmdline.c| 579 ++
app/test-pmd/config.c | 4 +-
app/test-pmd/parameters.c | 3 +
app/test-pmd/testpmd.c| 40 +++-
app/test-pmd/testpmd.h| 2 +
5 files changed, 619 insertions(+), 9 deletions(-)
diff --git a/app/test-pmd/cmdline.c b/app/test-pmd/cmdline.c
index 45ce343..5cf93bb 100644
--- a/app/test-pmd/cmdline.c
+++ b/app/test-pmd/cmdline.c
@@ -84,6 +84,9 @@
#include
#include
#include
+#ifdef RTE_LIBRTE_PMD_BOND
+#include
+#endif
#include "testpmd.h"
@@ -404,6 +407,31 @@ static void cmd_help_long_parsed(void *parsed_result,
" Show the bypass configuration for a bypass enabled
NIC"
" using the lowest port on the NIC.\n\n"
#endif
+#ifdef RTE_LIBRTE_PMD_BOND
+ "create bonded device (mode) (socket)\n"
+ " Create a new bonded device with specific
bonding mode and socket.\n\n"
+
+ "add bonding slave (slave_id) (port_id)\n"
+ " Add a slave device to a bonded device.\n\n"
+
+ "remove bonding slave (slave_id) (port_id)\n"
+ " Remove a slave device from a bonded device.\n\n"
+
+ "set bonding mode (value) (port_id)\n"
+ " Set the bonding mode on a bonded device.\n\n"
+
+ "set bonding primary (slave_id) (port_id)\n"
+ " Set the primary slave for a bonded device.\n\n"
+
+ "show bonding config (port_id)\n"
+ " Show the bonding config for port_id.\n\n"
+
+ "set bonding mac_addr (port_id) (address)\n"
+ " Set the MAC address of a bonded device.\n\n"
+
+ "set bonding xmit_balance_policy (port_id)
(l2|l23|l34)\n"
+ " Set the transmit balance policy for bonded
device running in balance mode.\n\n"
+#endif
, list_pkt_forwarding_modes()
);
@@ -3031,6 +3059,547 @@ cmdline_parse_inst_t cmd_show_bypass_config = {
};
#endif
+#ifdef RTE_LIBRTE_PMD_BOND
+/* *** SET BONDING MODE *** */
+struct cmd_set_bonding_mode_result {
+ cmdline_fixed_string_t set;
+ cmdline_fixed_string_t bonding;
+ cmdline_fixed_string_t mode;
+ uint8_t value;
+ uint8_t port_id;
+};
+
+static void cmd_set_bonding_mode_parsed(void *parsed_result,
+ __attribute__((unused)) struct cmdline *cl,
+ __attribute__((unused)) void *data)
+{
+ struct cmd_set_bonding_mode_result *res = parsed_result;
+ portid_t port_id = res->port_id;
+
+ /* Set the bonding mode for the relevant port. */
+ if (0 != rte_eth_bond_mode_set(port_id, res->value))
+ printf("\t Failed to set bonding mode for port = %d.\n",
port_id);
+}
+
+cmdline_parse_token_string_t cmd_setbonding_mode_set =
+TOKEN_STRING_INITIALIZER(struct cmd_set_bonding_mode_result,
+ set, "set");
+cmdline_parse_token_string_t cmd_setbonding_mode_bonding =
+TOKEN_STRING_INITIALIZER(struct cmd_set_bonding_mode_result,
+ bonding, "bonding");
+cmdline_parse_token_string_t cmd_setbonding_mode_mode =
+TOKEN_STRING_INITIALIZER(struct cmd_set_bonding_mode_result,
+ mode, "mode");
+cmdline_parse_token_num_t cmd_setbonding_mode_value =
+TOKEN_NUM_INITIALIZER(struct cmd_set_bonding_mode_result,
+ value, UINT8);
+cmdline_parse_token_num_t cmd_setbonding_mode_port =
+TOKEN_NUM_INITIALIZER(struct cmd_set_bonding_mode_result,
+ port_id, UINT8);
+
+cmdline_parse_inst_t cmd_set_bonding_mode = {
+ .f = cmd_set_bonding_mode_parsed,
+ .help_str = "set bonding mode (mode_value) (port_id): Set the
bonding mode for port_id",
+ .data = NULL,
+ .tokens = {
+ (void *) &cmd_setbonding_mode_set,
+ (void *) &cmd_setbonding_mode_bonding,
+ (void *) &cmd_setbonding_mode_mode,
+ (void *) &cmd_setbonding_mode_value,
+ (void *) &cmd_setbonding_mode_port,
+ NULL
+ }
+};
+
+/* *** SET BALANCE XMIT POLICY *** */
+struct cmd_set_bonding_balance_xmit_policy_result {
+ cmdline_fixed_string_t set;
+ cmdline_fixed_string_t bonding;
+ cmdline_fixed_string_t balance_xmit_policy;
+ uint8_t port_id;
+ cmdline_fixed_string_t policy;
+};
+
+static void cmd_set_bonding_balan
[dpdk-dev] [PATCH] ethdev: read link state interrupt without link_update service
> > It is now possible to read link status updated by interrupt without > > having manual link_update() service provided by the PMD. > > Indeed link_update() is useless in interrupt case. > > > > Signed-off-by: Thomas Monjalon > > Acked-by: Olivier Matz Applied for version 1.7.0. -- Thomas
[dpdk-dev] [PATCH] rte_memory.h: include stdio.h for FILE
> The below commit requires stdio FILE structure. > > commit 591a9d7985c1230652d9f7ea1f9221e8c66ec188 > Author: Stephen Hemminger > Date: Fri May 2 16:42:56 2014 -0700 > > add FILE argument to debug functions > > Application which includes rte_memory.h without stdio.h will be hit > compilation failure. > > /path/to/include/rte_memory.h:146:30: error: unknown type name ?FILE? > void rte_dump_physmem_layout(FILE *f); > > Signed-off-by: Hiroshi Shimamoto > Reviewed-by: Hayato Momma Acked-by: Thomas Monjalon Applied for version 1.7.0. Thanks -- Thomas
[dpdk-dev] [PATCH v2] rte_ethdev: add link support flag
2014-06-20 15:06, Thomas Monjalon:
> 2014-06-19 15:12, Stephen Hemminger:
> > Only some devices support the link state interrupt configuration option.
> > Link state control does not work in virtual drivers
> > (virtio, vmxnet3, igbvf, and ixgbevf). Instead of having the application
> > try and guess whether it will work or not provide a driver flag that
> > can be checked instead.
> >
> > Note: if device driver doesn't support link state control, what
> > would happen previously is that the code would never detect link
> > transitions. This prevents that.
>
> [...]
>
> > @@ -197,6 +197,8 @@ struct rte_pci_driver {
> >
> > #define RTE_PCI_DRV_MULTIPLE 0x0002
> > /** Device needs to be unbound even if no module is provided */
> > #define RTE_PCI_DRV_FORCE_UNBIND 0x0004
> >
> > +/** Device driver supports link state interrupt */
> > +#define RTE_PCI_DRV_LSC0x0008
>
> I feel RTE_PCI_DRV_INTR_LSC would be easier to understand.
> Do you agree?
>
> Note that related event is RTE_ETH_EVENT_INTR_LSC
> and configuration is intr_conf.lsc.
Applied with flag renamed.
Thanks
--
Thomas
[dpdk-dev] [PATCH 1/3] stringfns: remove rte_snprintf
2014-06-26 13:20, Aaron Campbell: > On Jun 26, 2014, at 12:09 PM, Richardson, Bruce intel.com> wrote: > >> I agree we should try to use the "deprecated" attribute when possible. > >> So application porting effort will be smoother. > >> > >> But in this case, there is something different: as Stephen wrote, > >> rte_snprintf is useless. It's useless inside the DPDK so it's even more > >> useless for user applications. > >> As it's really useless, it has no sense to keep it as deprecated. > >> Please, let's simply remove it. > > > > The reason to keep it as deprecated is so that those customers who don't > > want to do a huge amount of search-replace immediately can get things > > working again temporarily using -Wno-deprecated. It provides a simple > > temporary fallback cushion, and then we can completely remove the > > function later. So, I'd like to see us remove all our usage of the > > function internally in 1.7, along with marking as deprecated, and then > > completely remove in 1.8, (i.e. in a week's time or so) :-) > > As a DPDK user, I?d vote to kill it now. I doubt it is widely used in any > external applications. Such usage would be mostly from copy/pasting the > sample code, is my guess. I think everybody understood the idea: we'll try to use deprecation model when possible. In this case, it's probably useless but we close the discussion by deprecating it. We'll remove it soon, don't worry :) -- Thomas
[dpdk-dev] [PATCH v9 1/5] bond: new link bonding library
2014-06-27 01:57, Thomas Monjalon: > +SRCS-$(CONFIG_RTE_LIBRTE_PMD_BOND) += rte_eth_bond_vargs.c I forgot to rename the file here to rte_eth_bond_args.c. -- Thomas
[dpdk-dev] [PATCH 0/2] minor compile fixes for FreeBSD 10
2014-06-24 01:23, Bruce Richardson: > These two small patches fix minor compilation issues found when testing > compilation on a FreeBSD 10 system. > > > Bruce Richardson (2): > mk: overriding CC also overrides HOSTCC > bsdapp: disable ACL library compilation Acked and applied for version 1.7.0. Thanks -- Thomas
[dpdk-dev] [PATCH v3] EAL: fix usage of printf-like functions
> Mark the rte_log, cmdline_printf and rte_snprintf functions as > being printf-style functions. This causes compilation errors > due to mis-matched parameter types, so the parameter types are > fixed where appropriate. > > Changes in V2: > * Additional fixes for ivshmem-target compilation > > Changes in V3: > * additional printf-format fix for i40e driver as pointed out by Pablo De > Lara > > Signed-off-by: Bruce Richardson Acked-by: Thomas Monjalon Applied for version 1.7.0. Thanks -- Thomas
[dpdk-dev] [PATCH 1/2 v2] string: deprecate rte_snprintf
> The function rte_snprintf serves no useful purpose. It is the > same as snprintf() for all valid inputs. Deprecate it and > replace all uses in current code. > > Leave the tests for the deprecated function in place. > > Signed-off-by: Stephen Hemminger Acked-by: Thomas Monjalon Applied for version 1.7.0. Thanks -- Thomas
[dpdk-dev] [PATCH 1/2 v2] fix incorrect snprintf args
2014-06-24 11:14, Stephen Hemminger: > Now that snprintf is used, Gcc finds more uses of unsafe arguments. > Fix where found. Some of these may have already been fixed by > other patches on the mailing list. > > Signed-off-by: Stephen Hemminger It was already fixed by Bruce's patch: http://dpdk.org/browse/dpdk/commit/?id=e8ed6c78177 Thanks -- Thomas
[dpdk-dev] [PATCH] eal: fix invalid memory read as reported by valgrind
2014-06-26 10:54, Aaron Campbell: > ==29880== Invalid read of size 1 > ==29880==at 0x56FF9A5: cpu_socket_id (eal_lcore.c:101) > ==29880==by 0x56FFAE9: rte_eal_cpu_init (eal_lcore.c:168) > ==29880==by 0x56F944A: rte_eal_init (eal.c:975) > > The problem is that endptr points to memory allocated underneath the DIR > handle, which has already been freed. So move the closedir() call lower. > > Signed-off-by: Aaron Campbell Good catch! Acked-by: Thomas Monjalon Applied for version 1.7.0. Thanks -- Thomas
[dpdk-dev] [PATCH 0/3] fix flow control in testpmd
2014-06-24 15:06, David Marchand: > Here is a patchset that addresses the problem reported by Jijiang Liu and > introduces "partial" commands for setting flow control parameters in > testpmd. > > By the way, I tried to factorise the code in testpmd when handling flow > control "commands" so that we have only one parser even if multiple > commands are supported. Acked-by: Thomas Monjalon Applied for version 1.7.0. Thanks -- Thomas
[dpdk-dev] [PATCH] dpdk_nic_bind: allow status query without igb_uio
> Allow the nic bind/unbind script to print out its status messages even > if the igb_uio driver is not loaded. For binding and unbinding NICs, the > behaviour is the same, and the igb_uio driver still needs to be loaded. > > Signed-off-by: Bruce richardson Acked-by: Thomas Monjalon Applied for version 1.7.0. Thanks -- Thomas
[dpdk-dev] [PATCH] dpdk_nic_bind: unbind ports that were erroneously bound
> When binding devices to a generic driver (i.e. one that doesn't have a > PCI ID table, some devices that are not bound to any other driver could > be bound even if no one has asked them to. hence, we check the list of > drivers again, and see if some of the previously-unbound devices were > erroneously bound. if such devices are found, they are unbound back. > > Signed-off-by: Anatoly Burakov Acked-by: Thomas Monjalon Applied for version 1.7.0. Thanks -- Thomas
[dpdk-dev] [dpdk-announce] release candidate 1.7.0-rc2
The second release candidate of version 1.7.0 can be downloaded here: http://dpdk.org/browse/dpdk/tag/?id=v1.7.0-rc2 Please test it. The release 1.7.0 should be ready in few days if things go well. You can check the log here: http://dpdk.org/browse/dpdk/log A short changelog will be provided with the release. Some patches are pending for integration in next release candidate: - link bonding - KNI fixes - socket id detection fallback - new core set option Please help to review them. Thank you to everyone, -- Thomas
[dpdk-dev] Intel I350 | Jumbo Frame support | Segmentation fault
Hai all, When I try to send packets of sizes > 512 bytes (in a burst of 64), the program exits with a segmentation fault. I'm using Intel I350 1GbE card. Moreover if I try to enable jumbo frame support , it exits with an error saying 'Could not configure port1 (-22)' . Any help would be appreciated. Thanks in advance.
[dpdk-dev] [PATCH] app/test: fix build switches to enable cmdline tests
There were 2 typos since these commits (in 1.6.0 releases):
21a7f4e264 fix build without librte_cmdline
cac6d08c8b replace --use-device option by --pci-whitelist and --vdev
In makefiles, the build options are prefixed with CONFIG_RTE_
but in .c file, it is only RTE_.
These typos were disabling cmdline unit tests and test of "--vdev eth_ring"
option.
Signed-off-by: Thomas Monjalon
---
app/test/test_cmdline.c | 2 +-
app/test/test_eal_flags.c | 2 +-
2 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/app/test/test_cmdline.c b/app/test/test_cmdline.c
index 77475c4..10a3f77 100644
--- a/app/test/test_cmdline.c
+++ b/app/test/test_cmdline.c
@@ -39,7 +39,7 @@
int
test_cmdline(void)
{
-#ifdef CONFIG_RTE_LIBRTE_CMDLINE
+#ifdef RTE_LIBRTE_CMDLINE
printf("Testind parsing ethernet addresses...\n");
if (test_parse_etheraddr_valid() < 0)
return -1;
diff --git a/app/test/test_eal_flags.c b/app/test/test_eal_flags.c
index a862654..1b80b80 100644
--- a/app/test/test_eal_flags.c
+++ b/app/test/test_eal_flags.c
@@ -317,7 +317,7 @@ test_whitelist_flag(void)
const char *wlval3[] = {prgname, prefix, mp_flag, "-n", "1", "-c", "1",
pci_whitelist, "09:0B.3,type=test",
pci_whitelist, "08:00.1,type=normal",
-#ifdef CONFIG_RTE_LIBRTE_PMD_RING
+#ifdef RTE_LIBRTE_PMD_RING
vdev, "eth_ring,arg=test",
#endif
};
--
2.0.0
[dpdk-dev] [PATCH] app/test: fix build switches to enable cmdline tests
> -Original Message- > From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Thomas Monjalon > Sent: Friday, June 27, 2014 10:47 AM > To: dev at dpdk.org > Subject: [dpdk-dev] [PATCH] app/test: fix build switches to enable cmdline > tests > > There were 2 typos since these commits (in 1.6.0 releases): > 21a7f4e264 fix build without librte_cmdline > cac6d08c8b replace --use-device option by --pci-whitelist and --vdev > In makefiles, the build options are prefixed with CONFIG_RTE_ > but in .c file, it is only RTE_. > > These typos were disabling cmdline unit tests and test of "--vdev eth_ring" > option. > > Signed-off-by: Thomas Monjalon > --- Acked-by: Pablo de Lara
[dpdk-dev] [PATCH v10 0/5] link bonding
Fixed copyright in link bonding library Makefile Declan Doherty (5): bond: new link bonding library ethdev: add unique name to devices eal: support link bonding device initialization bond: unit tests bond: testpmd support app/test-pmd/cmdline.c | 579 app/test-pmd/config.c |4 +- app/test-pmd/parameters.c |3 + app/test-pmd/testpmd.c | 40 +- app/test-pmd/testpmd.h |2 + app/test/Makefile |4 +- app/test/commands.c |7 + app/test/packet_burst_generator.c | 287 ++ app/test/packet_burst_generator.h | 78 + app/test/test.h |1 + app/test/test_link_bonding.c| 3958 +++ app/test/virtual_pmd.c | 574 app/test/virtual_pmd.h | 74 + config/common_bsdapp|5 + config/common_linuxapp |5 + doc/doxy-api-index.md |1 + doc/doxy-api.conf |1 + lib/Makefile|1 + lib/librte_eal/bsdapp/eal/eal.c | 10 +- lib/librte_eal/common/eal_common_dev.c | 58 +- lib/librte_eal/common/eal_common_pci.c |3 + lib/librte_eal/common/include/eal_private.h |7 - lib/librte_eal/common/include/rte_dev.h | 14 +- lib/librte_eal/linuxapp/eal/eal.c | 11 +- lib/librte_ether/rte_ethdev.c | 31 +- lib/librte_ether/rte_ethdev.h |7 +- lib/librte_pmd_bond/Makefile| 61 + lib/librte_pmd_bond/rte_eth_bond.h | 255 ++ lib/librte_pmd_bond/rte_eth_bond_api.c | 662 + lib/librte_pmd_bond/rte_eth_bond_args.c | 252 ++ lib/librte_pmd_bond/rte_eth_bond_pmd.c | 1212 lib/librte_pmd_bond/rte_eth_bond_private.h | 215 ++ lib/librte_pmd_pcap/rte_eth_pcap.c | 22 +- lib/librte_pmd_ring/rte_eth_ring.c | 32 +- lib/librte_pmd_ring/rte_eth_ring.h |3 +- lib/librte_pmd_xenvirt/rte_eth_xenvirt.c|2 +- mk/rte.app.mk |4 + 37 files changed, 8414 insertions(+), 71 deletions(-) create mode 100644 app/test/packet_burst_generator.c create mode 100644 app/test/packet_burst_generator.h create mode 100644 app/test/test_link_bonding.c create mode 100644 app/test/virtual_pmd.c create mode 100644 app/test/virtual_pmd.h create mode 100644 lib/librte_pmd_bond/Makefile create mode 100644 lib/librte_pmd_bond/rte_eth_bond.h create mode 100644 lib/librte_pmd_bond/rte_eth_bond_api.c create mode 100644 lib/librte_pmd_bond/rte_eth_bond_args.c create mode 100644 lib/librte_pmd_bond/rte_eth_bond_pmd.c create mode 100644 lib/librte_pmd_bond/rte_eth_bond_private.h
[dpdk-dev] [PATCH v10 3/5] eal: support link bonding device initialization
Updating functionality in EAL to support adding link bonding
devices via ?vdev option. Link bonding devices will be
initialized after all physical devices have been probed and
initialized.
Signed-off-by: Declan Doherty
Acked-by: Pablo de Lara
---
lib/librte_eal/bsdapp/eal/eal.c | 10 -
lib/librte_eal/common/eal_common_dev.c | 58 ++
lib/librte_eal/common/eal_common_pci.c |3 +
lib/librte_eal/common/include/eal_private.h |7 ---
lib/librte_eal/common/include/rte_dev.h | 14 ++-
lib/librte_eal/linuxapp/eal/eal.c | 11 +-
6 files changed, 74 insertions(+), 29 deletions(-)
diff --git a/lib/librte_eal/bsdapp/eal/eal.c b/lib/librte_eal/bsdapp/eal/eal.c
index a1f014f..c53f63e 100644
--- a/lib/librte_eal/bsdapp/eal/eal.c
+++ b/lib/librte_eal/bsdapp/eal/eal.c
@@ -874,7 +874,7 @@ rte_eal_init(int argc, char **argv)
rte_eal_mcfg_complete();
- if (rte_eal_dev_init() < 0)
+ if (rte_eal_dev_init(PMD_INIT_PRE_PCI_PROBE) < 0)
rte_panic("Cannot init pmd devices\n");
RTE_LCORE_FOREACH_SLAVE(i) {
@@ -906,6 +906,14 @@ rte_eal_init(int argc, char **argv)
rte_eal_mp_remote_launch(sync_func, NULL, SKIP_MASTER);
rte_eal_mp_wait_lcore();
+ /* Probe & Initialize PCI devices */
+ if (rte_eal_pci_probe())
+ rte_panic("Cannot probe PCI\n");
+
+ /* Initialize any outstanding devices */
+ if (rte_eal_dev_init(PMD_INIT_POST_PCI_PROBE) < 0)
+ rte_panic("Cannot init pmd devices\n");
+
return fctret;
}
diff --git a/lib/librte_eal/common/eal_common_dev.c
b/lib/librte_eal/common/eal_common_dev.c
index eae5656..1194419 100644
--- a/lib/librte_eal/common/eal_common_dev.c
+++ b/lib/librte_eal/common/eal_common_dev.c
@@ -62,7 +62,7 @@ rte_eal_driver_unregister(struct rte_driver *driver)
}
int
-rte_eal_dev_init(void)
+rte_eal_dev_init(uint8_t init_pri)
{
struct rte_devargs *devargs;
struct rte_driver *driver;
@@ -80,30 +80,52 @@ rte_eal_dev_init(void)
continue;
TAILQ_FOREACH(driver, &dev_driver_list, next) {
- if (driver->type != PMD_VDEV)
- continue;
+ /* RTE_DEVTYPE_VIRTUAL can only be a virtual or bonded
device,
+* virtual devices are initialized pre PCI probing and
bonded
+* device are post pci probing */
+ if ((driver->type == PMD_VDEV && init_pri ==
+ PMD_INIT_PRE_PCI_PROBE) ||
+ (driver->type == PMD_BDEV && init_pri ==
+ PMD_INIT_POST_PCI_PROBE)) {
- /* search a driver prefix in virtual device name */
- if (!strncmp(driver->name, devargs->virtual.drv_name,
- strlen(driver->name))) {
- driver->init(devargs->virtual.drv_name,
- devargs->args);
- break;
+ /* search a driver prefix in virtual device
name */
+ if (!strncmp(driver->name,
devargs->virtual.drv_name,
+ strlen(driver->name))) {
+ printf("init (%u) %s\n", init_pri,
devargs->virtual.drv_name);
+ driver->init(devargs->virtual.drv_name,
+ devargs->args);
+ break;
+ }
}
}
- if (driver == NULL) {
- rte_panic("no driver found for %s\n",
- devargs->virtual.drv_name);
+ /* If initializing pre PCI probe, then we don't expect a bonded
driver
+* to be found */
+ if (init_pri == PMD_INIT_PRE_PCI_PROBE &&
+ strncmp(PMD_BOND_NAME,
devargs->virtual.drv_name,
+ strlen(PMD_BOND_NAME)) != 0) {
+ if (driver == NULL) {
+ rte_panic("no driver found for virtual device
%s\n",
+ devargs->virtual.drv_name);
+ }
+ } else if (init_pri == PMD_INIT_POST_PCI_PROBE &&
+ strncmp(PMD_BOND_NAME,
devargs->virtual.drv_name,
+ strlen(PMD_BOND_NAME)) == 0) {
+ if (driver == NULL) {
+ rte_panic("no driver found for bonded device
%s\n",
+ devargs->virtual.drv_name);
+ }
}
[dpdk-dev] [PATCH v10 2/5] ethdev: add unique name to devices
Adding support to rte_eth_dev_data structure to support unique
name identifier for ethdevs to support adding slave ethdevs
(specifically virtual devices which have no public unique
identifier) to a link bonding device. This changes the API
rte_eth_dev_allocate() to require a const char *name when
allocating a ethdev, which also verifies that the name is
unique and hasn?t been already used by an existed allocated
rte_eth_dev. Also contains updates to virtual pmd?s to now call
the API with a name parameter.
Signed-off-by: Declan Doherty
Acked-by: Pablo de Lara
---
lib/librte_ether/rte_ethdev.c| 31 ++--
lib/librte_ether/rte_ethdev.h|7 +-
lib/librte_pmd_pcap/rte_eth_pcap.c | 22 ++--
lib/librte_pmd_ring/rte_eth_ring.c | 32 +++--
lib/librte_pmd_ring/rte_eth_ring.h |3 +-
lib/librte_pmd_xenvirt/rte_eth_xenvirt.c |2 +-
6 files changed, 65 insertions(+), 32 deletions(-)
diff --git a/lib/librte_ether/rte_ethdev.c b/lib/librte_ether/rte_ethdev.c
index ef81a76..0ebb8fb 100644
--- a/lib/librte_ether/rte_ethdev.c
+++ b/lib/librte_ether/rte_ethdev.c
@@ -65,6 +65,7 @@
#include
#include
#include
+#include
#include "rte_ether.h"
#include "rte_ethdev.h"
@@ -153,21 +154,39 @@ rte_eth_dev_data_alloc(void)
RTE_MAX_ETHPORTS * sizeof(*rte_eth_dev_data));
}
+static struct rte_eth_dev *
+rte_eth_dev_allocated(const char *name)
+{
+ unsigned i;
+
+ for (i = 0; i < nb_ports; i++) {
+ if (strcmp(rte_eth_devices[i].data->name, name) == 0)
+ return &rte_eth_devices[i];
+ }
+ return NULL;
+}
+
struct rte_eth_dev *
-rte_eth_dev_allocate(void)
+rte_eth_dev_allocate(const char *name)
{
struct rte_eth_dev *eth_dev;
if (nb_ports == RTE_MAX_ETHPORTS) {
- PMD_DEBUG_TRACE("Reached maximum number of ethernet ports\n");
+ PMD_DEBUG_TRACE("Reached maximum number of Ethernet ports\n");
return NULL;
}
if (rte_eth_dev_data == NULL)
rte_eth_dev_data_alloc();
+ if (rte_eth_dev_allocated(name) != NULL) {
+ PMD_DEBUG_TRACE("Ethernet Device with name %s already
allocated!\n");
+ return NULL;
+ }
+
eth_dev = &rte_eth_devices[nb_ports];
eth_dev->data = &rte_eth_dev_data[nb_ports];
+ snprintf(eth_dev->data->name, sizeof(eth_dev->data->name), "%s", name);
eth_dev->data->port_id = nb_ports++;
return eth_dev;
}
@@ -178,11 +197,17 @@ rte_eth_dev_init(struct rte_pci_driver *pci_drv,
{
struct eth_driver*eth_drv;
struct rte_eth_dev *eth_dev;
+ char ethdev_name[RTE_ETH_NAME_MAX_LEN];
+
int diag;
eth_drv = (struct eth_driver *)pci_drv;
- eth_dev = rte_eth_dev_allocate();
+ /* Create unique Ethernet device name using PCI address */
+ snprintf(ethdev_name, RTE_ETH_NAME_MAX_LEN, "%d:%d.%d",
+ pci_dev->addr.bus, pci_dev->addr.devid,
pci_dev->addr.function);
+
+ eth_dev = rte_eth_dev_allocate(ethdev_name);
if (eth_dev == NULL)
return -ENOMEM;
diff --git a/lib/librte_ether/rte_ethdev.h b/lib/librte_ether/rte_ethdev.h
index 2406e45..50df654 100644
--- a/lib/librte_ether/rte_ethdev.h
+++ b/lib/librte_ether/rte_ethdev.h
@@ -1497,6 +1497,8 @@ struct rte_eth_dev_sriov {
};
#define RTE_ETH_DEV_SRIOV(dev) ((dev)->data->sriov)
+#define RTE_ETH_NAME_MAX_LEN (32)
+
/**
* @internal
* The data part, with no function pointers, associated with each ethernet
device.
@@ -1505,6 +1507,8 @@ struct rte_eth_dev_sriov {
* processes in a multi-process configuration.
*/
struct rte_eth_dev_data {
+ char name[RTE_ETH_NAME_MAX_LEN]; /**< Unique identifier name */
+
void **rx_queues; /**< Array of pointers to RX queues. */
void **tx_queues; /**< Array of pointers to TX queues. */
uint16_t nb_rx_queues; /**< Number of RX queues. */
@@ -1560,10 +1564,11 @@ extern uint8_t rte_eth_dev_count(void);
* Allocates a new ethdev slot for an ethernet device and returns the pointer
* to that slot for the driver to use.
*
+ * @param nameUnique identifier name for each Ethernet device
* @return
* - Slot in the rte_dev_devices array for a new device;
*/
-struct rte_eth_dev *rte_eth_dev_allocate(void);
+struct rte_eth_dev *rte_eth_dev_allocate(const char *name);
struct eth_driver;
/**
diff --git a/lib/librte_pmd_pcap/rte_eth_pcap.c
b/lib/librte_pmd_pcap/rte_eth_pcap.c
index b3dbbda..12b7e0c 100644
--- a/lib/librte_pmd_pcap/rte_eth_pcap.c
+++ b/lib/librte_pmd_pcap/rte_eth_pcap.c
@@ -534,7 +534,7 @@ open_tx_iface(const char *key __rte_unused, const char
*value, void *extra_args)
static int
-rte_pmd_init_internals(const unsigned nb_rx_queues,
+rte_pmd_init_internals(const char *name, const unsigned nb_rx_queues,
[dpdk-dev] [PATCH v10 5/5] bond: testpmd support
- Includes the ability to create new bonded devices.
- Add /remove bonding slave devices.
- Interogate bonded device stats/configuration
- Change bonding modes and select balance transmit polices
Signed-off-by: Declan Doherty
Acked-by: Pablo de Lara
---
app/test-pmd/cmdline.c| 579 +
app/test-pmd/config.c |4 +-
app/test-pmd/parameters.c |3 +
app/test-pmd/testpmd.c| 40 +++-
app/test-pmd/testpmd.h|2 +
5 files changed, 619 insertions(+), 9 deletions(-)
diff --git a/app/test-pmd/cmdline.c b/app/test-pmd/cmdline.c
index f61a31c..345be11 100644
--- a/app/test-pmd/cmdline.c
+++ b/app/test-pmd/cmdline.c
@@ -84,6 +84,9 @@
#include
#include
#include
+#ifdef RTE_LIBRTE_PMD_BOND
+#include
+#endif
#include "testpmd.h"
@@ -412,6 +415,31 @@ static void cmd_help_long_parsed(void *parsed_result,
" Show the bypass configuration for a bypass enabled
NIC"
" using the lowest port on the NIC.\n\n"
#endif
+#ifdef RTE_LIBRTE_PMD_BOND
+ "create bonded device (mode) (socket)\n"
+ " Create a new bonded device with specific
bonding mode and socket.\n\n"
+
+ "add bonding slave (slave_id) (port_id)\n"
+ " Add a slave device to a bonded device.\n\n"
+
+ "remove bonding slave (slave_id) (port_id)\n"
+ " Remove a slave device from a bonded device.\n\n"
+
+ "set bonding mode (value) (port_id)\n"
+ " Set the bonding mode on a bonded device.\n\n"
+
+ "set bonding primary (slave_id) (port_id)\n"
+ " Set the primary slave for a bonded device.\n\n"
+
+ "show bonding config (port_id)\n"
+ " Show the bonding config for port_id.\n\n"
+
+ "set bonding mac_addr (port_id) (address)\n"
+ " Set the MAC address of a bonded device.\n\n"
+
+ "set bonding xmit_balance_policy (port_id)
(l2|l23|l34)\n"
+ " Set the transmit balance policy for bonded
device running in balance mode.\n\n"
+#endif
, list_pkt_forwarding_modes()
);
@@ -3039,6 +3067,547 @@ cmdline_parse_inst_t cmd_show_bypass_config = {
};
#endif
+#ifdef RTE_LIBRTE_PMD_BOND
+/* *** SET BONDING MODE *** */
+struct cmd_set_bonding_mode_result {
+ cmdline_fixed_string_t set;
+ cmdline_fixed_string_t bonding;
+ cmdline_fixed_string_t mode;
+ uint8_t value;
+ uint8_t port_id;
+};
+
+static void cmd_set_bonding_mode_parsed(void *parsed_result,
+ __attribute__((unused)) struct cmdline *cl,
+ __attribute__((unused)) void *data)
+{
+ struct cmd_set_bonding_mode_result *res = parsed_result;
+ portid_t port_id = res->port_id;
+
+ /* Set the bonding mode for the relevant port. */
+ if (0 != rte_eth_bond_mode_set(port_id, res->value))
+ printf("\t Failed to set bonding mode for port = %d.\n",
port_id);
+}
+
+cmdline_parse_token_string_t cmd_setbonding_mode_set =
+TOKEN_STRING_INITIALIZER(struct cmd_set_bonding_mode_result,
+ set, "set");
+cmdline_parse_token_string_t cmd_setbonding_mode_bonding =
+TOKEN_STRING_INITIALIZER(struct cmd_set_bonding_mode_result,
+ bonding, "bonding");
+cmdline_parse_token_string_t cmd_setbonding_mode_mode =
+TOKEN_STRING_INITIALIZER(struct cmd_set_bonding_mode_result,
+ mode, "mode");
+cmdline_parse_token_num_t cmd_setbonding_mode_value =
+TOKEN_NUM_INITIALIZER(struct cmd_set_bonding_mode_result,
+ value, UINT8);
+cmdline_parse_token_num_t cmd_setbonding_mode_port =
+TOKEN_NUM_INITIALIZER(struct cmd_set_bonding_mode_result,
+ port_id, UINT8);
+
+cmdline_parse_inst_t cmd_set_bonding_mode = {
+ .f = cmd_set_bonding_mode_parsed,
+ .help_str = "set bonding mode (mode_value) (port_id): Set the
bonding mode for port_id",
+ .data = NULL,
+ .tokens = {
+ (void *) &cmd_setbonding_mode_set,
+ (void *) &cmd_setbonding_mode_bonding,
+ (void *) &cmd_setbonding_mode_mode,
+ (void *) &cmd_setbonding_mode_value,
+ (void *) &cmd_setbonding_mode_port,
+ NULL
+ }
+};
+
+/* *** SET BALANCE XMIT POLICY *** */
+struct cmd_set_bonding_balance_xmit_policy_result {
+ cmdline_fixed_string_t set;
+ cmdline_fixed_string_t bonding;
+ cmdline_fixed_string_t balance_xmit_policy;
+ uint8_t port_id;
+ cmdline_fixed_string_t policy;
+};
+
+static void cmd_set_bonding_balance_xmit_policy_pars
[dpdk-dev] [PATCH v10 1/5] bond: new link bonding library
Initial release with support for Mode 0 - Round Robin Mode 1 - Active Backup Mode 2 - Balance -> Supports 3 transmit polices (layer 2, layer 2+3, layer 3+4) Mode 3 - Broadcast Signed-off-by: Declan Doherty Acked-by: Pablo de Lara --- config/common_bsdapp |5 + config/common_linuxapp |5 + doc/doxy-api-index.md |1 + doc/doxy-api.conf |1 + lib/Makefile |1 + lib/librte_pmd_bond/Makefile | 61 ++ lib/librte_pmd_bond/rte_eth_bond.h | 255 ++ lib/librte_pmd_bond/rte_eth_bond_api.c | 662 +++ lib/librte_pmd_bond/rte_eth_bond_args.c| 252 ++ lib/librte_pmd_bond/rte_eth_bond_pmd.c | 1212 lib/librte_pmd_bond/rte_eth_bond_private.h | 215 + mk/rte.app.mk |4 + 12 files changed, 2674 insertions(+), 0 deletions(-) create mode 100644 lib/librte_pmd_bond/Makefile create mode 100644 lib/librte_pmd_bond/rte_eth_bond.h create mode 100644 lib/librte_pmd_bond/rte_eth_bond_api.c create mode 100644 lib/librte_pmd_bond/rte_eth_bond_args.c create mode 100644 lib/librte_pmd_bond/rte_eth_bond_pmd.c create mode 100644 lib/librte_pmd_bond/rte_eth_bond_private.h diff --git a/config/common_bsdapp b/config/common_bsdapp index d5db4ab..c243b0c 100644 --- a/config/common_bsdapp +++ b/config/common_bsdapp @@ -206,6 +206,11 @@ CONFIG_RTE_PMD_RING_MAX_TX_RINGS=16 CONFIG_RTE_LIBRTE_PMD_PCAP=y # +# Compile link bonding PMD library +# +CONFIG_RTE_LIBRTE_PMD_BOND=y + +# # Do prefetch of packet data within PMD driver receive function # CONFIG_RTE_PMD_PACKET_PREFETCH=y diff --git a/config/common_linuxapp b/config/common_linuxapp index c5c0cb6..6237ce5 100644 --- a/config/common_linuxapp +++ b/config/common_linuxapp @@ -244,6 +244,11 @@ CONFIG_RTE_PMD_RING_MAX_TX_RINGS=16 CONFIG_RTE_LIBRTE_PMD_PCAP=n # +# Compile link bonding PMD library +# +CONFIG_RTE_LIBRTE_PMD_BOND=y + +# # Compile Xen PMD # CONFIG_RTE_LIBRTE_PMD_XENVIRT=n diff --git a/doc/doxy-api-index.md b/doc/doxy-api-index.md index 7b26e98..d72659a 100644 --- a/doc/doxy-api-index.md +++ b/doc/doxy-api-index.md @@ -38,6 +38,7 @@ There are many libraries, so their headers may be grouped by topics: - **device**: [ethdev] (@ref rte_ethdev.h), [devargs](@ref rte_devargs.h), + [bond] (@ref rte_eth_bond.h), [KNI](@ref rte_kni.h), [PCI](@ref rte_pci.h), [PCI IDs](@ref rte_pci_dev_ids.h) diff --git a/doc/doxy-api.conf b/doc/doxy-api.conf index f380d9a..1fd4492 100644 --- a/doc/doxy-api.conf +++ b/doc/doxy-api.conf @@ -47,6 +47,7 @@ INPUT = doc/doxy-api-index.md \ lib/librte_pipeline \ lib/librte_port \ lib/librte_power \ + lib/librte_pmd_bond \ lib/librte_ring \ lib/librte_sched \ lib/librte_table \ diff --git a/lib/Makefile b/lib/Makefile index c58c0c9..10c5bb3 100644 --- a/lib/Makefile +++ b/lib/Makefile @@ -44,6 +44,7 @@ DIRS-$(CONFIG_RTE_LIBRTE_ETHER) += librte_ether DIRS-$(CONFIG_RTE_LIBRTE_E1000_PMD) += librte_pmd_e1000 DIRS-$(CONFIG_RTE_LIBRTE_IXGBE_PMD) += librte_pmd_ixgbe DIRS-$(CONFIG_RTE_LIBRTE_I40E_PMD) += librte_pmd_i40e +DIRS-$(CONFIG_RTE_LIBRTE_PMD_BOND) += librte_pmd_bond DIRS-$(CONFIG_RTE_LIBRTE_PMD_RING) += librte_pmd_ring DIRS-$(CONFIG_RTE_LIBRTE_PMD_PCAP) += librte_pmd_pcap DIRS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD) += librte_pmd_virtio diff --git a/lib/librte_pmd_bond/Makefile b/lib/librte_pmd_bond/Makefile new file mode 100644 index 000..21998f7 --- /dev/null +++ b/lib/librte_pmd_bond/Makefile @@ -0,0 +1,61 @@ +# BSD LICENSE +# +# Copyright(c) 2010-2014 Intel Corporation. All rights reserved. +# All rights reserved. +# +# Redistribution and use in source and binary forms, with or without +# modification, are permitted provided that the following conditions +# are met: +# +# * Redistributions of source code must retain the above copyright +# notice, this list of conditions and the following disclaimer. +# * Redistributions in binary form must reproduce the above copyright +# notice, this list of conditions and the following disclaimer in +# the documentation and/or other materials provided with the +# distribution. +# * Neither the name of Intel Corporation nor the names of its +# contributors may be used to endorse or promote products derived +# from this software without specific prior written permission. +# +# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS +# "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT +# LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR +#
[dpdk-dev] [PATCH v10 4/5] bond: unit tests
Including: - code to generate packet bursts for testing rx and tx functionality of bonded device - virtual/stubbed out ethdev for use as slave ethdev in testing Signed-off-by: Declan Doherty Acked-by: Pablo de Lara --- app/test/Makefile |4 +- app/test/commands.c |7 + app/test/packet_burst_generator.c | 287 +++ app/test/packet_burst_generator.h | 78 + app/test/test.h |1 + app/test/test_link_bonding.c | 3958 + app/test/virtual_pmd.c| 574 ++ app/test/virtual_pmd.h| 74 + 8 files changed, 4982 insertions(+), 1 deletions(-) create mode 100644 app/test/packet_burst_generator.c create mode 100644 app/test/packet_burst_generator.h create mode 100644 app/test/test_link_bonding.c create mode 100644 app/test/virtual_pmd.c create mode 100644 app/test/virtual_pmd.h diff --git a/app/test/Makefile b/app/test/Makefile index 45d0cf2..cec2f2f 100644 --- a/app/test/Makefile +++ b/app/test/Makefile @@ -102,7 +102,9 @@ SRCS-$(CONFIG_RTE_APP_TEST) += test_ivshmem.c SRCS-$(CONFIG_RTE_APP_TEST) += test_distributor.c SRCS-$(CONFIG_RTE_APP_TEST) += test_distributor_perf.c SRCS-$(CONFIG_RTE_APP_TEST) += test_devargs.c - +SRCS-$(CONFIG_RTE_APP_TEST) += virtual_pmd.c +SRCS-$(CONFIG_RTE_APP_TEST) += packet_burst_generator.c +SRCS-$(CONFIG_RTE_APP_TEST) += test_link_bonding.c ifeq ($(CONFIG_RTE_APP_TEST),y) SRCS-$(CONFIG_RTE_LIBRTE_ACL) += test_acl.c SRCS-$(CONFIG_RTE_LIBRTE_PMD_RING) += test_pmd_ring.c diff --git a/app/test/commands.c b/app/test/commands.c index c9dc085..5f23420 100644 --- a/app/test/commands.c +++ b/app/test/commands.c @@ -159,6 +159,10 @@ static void cmd_autotest_parsed(void *parsed_result, ret = test_timer(); if (!strcmp(res->autotest, "timer_perf_autotest")) ret = test_timer_perf(); +#ifdef RTE_LIBRTE_PMD_BOND + if (!strcmp(res->autotest, "link_bonding_autotest")) + ret = test_link_bonding(); +#endif if (!strcmp(res->autotest, "mempool_autotest")) ret = test_mempool(); if (!strcmp(res->autotest, "mempool_perf_autotest")) @@ -227,6 +231,9 @@ cmdline_parse_token_string_t cmd_autotest_autotest = "alarm_autotest#interrupt_autotest#" "version_autotest#eal_fs_autotest#" "cmdline_autotest#func_reentrancy_autotest#" +#ifdef RTE_LIBRTE_PMD_BOND + "link_bonding_autotest#" +#endif "mempool_perf_autotest#hash_perf_autotest#" "memcpy_perf_autotest#ring_perf_autotest#" "red_autotest#meter_autotest#sched_autotest#" diff --git a/app/test/packet_burst_generator.c b/app/test/packet_burst_generator.c new file mode 100644 index 000..5d539f1 --- /dev/null +++ b/app/test/packet_burst_generator.c @@ -0,0 +1,287 @@ +/*- + * BSD LICENSE + * + * Copyright(c) 2010-2014 Intel Corporation. All rights reserved. + * All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * * Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * * Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * * Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + */ + +#include +#include + +#include "packet_burst_generator.h" + +#define UDP_SRC_PORT 1024 +#define UDP_DST_PORT 1024 + + +#define IP_DEFTTL 64 /* from RFC 1340. */ +#define IP_VERSION 0x40 +#define IP_HDRLEN 0x05 /* default IP header length == five 32-bits words. */ +#define IP_VHL_DEF (IP_VERSION | IP_HDRLEN) +
[dpdk-dev] [PATCH] string: fix potential seg fault on snprintf
From: Pablo de Lara
Several functions did not check if destination buffer, used
in snprintf was a non-NULL pointer.
Signed-off-by: Pablo de Lara
---
examples/cmdline/parse_obj_list.c|3 +++
lib/librte_cmdline/cmdline_parse_etheraddr.c |3 +++
lib/librte_cmdline/cmdline_parse_num.c |2 +-
lib/librte_cmdline/cmdline_parse_portlist.c |4
lib/librte_eal/linuxapp/eal/eal_pci_uio.c|3 +++
5 files changed, 14 insertions(+), 1 deletions(-)
diff --git a/examples/cmdline/parse_obj_list.c
b/examples/cmdline/parse_obj_list.c
index 2625ca3..97dfa09 100644
--- a/examples/cmdline/parse_obj_list.c
+++ b/examples/cmdline/parse_obj_list.c
@@ -157,6 +157,9 @@ int complete_get_elt_obj_list(cmdline_parse_token_hdr_t *tk,
int get_help_obj_list(__attribute__((unused)) cmdline_parse_token_hdr_t *tk,
char *dstbuf, unsigned int size)
{
+ if (!dstbuf)
+ return -1;
+
snprintf(dstbuf, size, "Obj-List");
return 0;
}
diff --git a/lib/librte_cmdline/cmdline_parse_etheraddr.c
b/lib/librte_cmdline/cmdline_parse_etheraddr.c
index 5285c40..774b167 100644
--- a/lib/librte_cmdline/cmdline_parse_etheraddr.c
+++ b/lib/librte_cmdline/cmdline_parse_etheraddr.c
@@ -170,6 +170,9 @@ cmdline_get_help_etheraddr(__attribute__((unused))
cmdline_parse_token_hdr_t *tk
{
int ret;
+ if (!dstbuf)
+ return -1;
+
ret = snprintf(dstbuf, size, "Ethernet address");
if (ret < 0)
return -1;
diff --git a/lib/librte_cmdline/cmdline_parse_num.c
b/lib/librte_cmdline/cmdline_parse_num.c
index 0b9e4d0..26ba5ef 100644
--- a/lib/librte_cmdline/cmdline_parse_num.c
+++ b/lib/librte_cmdline/cmdline_parse_num.c
@@ -349,7 +349,7 @@ cmdline_get_help_num(cmdline_parse_token_hdr_t *tk, char
*dstbuf, unsigned int s
struct cmdline_token_num_data nd;
int ret;
- if (!tk)
+ if (!tk || !dstbuf)
return -1;
memcpy(&nd, &((struct cmdline_token_num *)tk)->num_data, sizeof(nd));
diff --git a/lib/librte_cmdline/cmdline_parse_portlist.c
b/lib/librte_cmdline/cmdline_parse_portlist.c
index 7eac05c..8e3521a 100644
--- a/lib/librte_cmdline/cmdline_parse_portlist.c
+++ b/lib/librte_cmdline/cmdline_parse_portlist.c
@@ -163,6 +163,10 @@ cmdline_get_help_portlist(__attribute__((unused))
cmdline_parse_token_hdr_t *tk,
char *dstbuf, unsigned int size)
{
int ret;
+
+ if (!dstbuf)
+ return -1;
+
ret = snprintf(dstbuf, size, "range of ports as 3,4-6,8-19,20");
if (ret < 0)
return -1;
diff --git a/lib/librte_eal/linuxapp/eal/eal_pci_uio.c
b/lib/librte_eal/linuxapp/eal/eal_pci_uio.c
index 7e62266..ec0ba42 100644
--- a/lib/librte_eal/linuxapp/eal/eal_pci_uio.c
+++ b/lib/librte_eal/linuxapp/eal/eal_pci_uio.c
@@ -211,6 +211,9 @@ pci_get_uio_dev(struct rte_pci_device *dev, char *dstbuf,
DIR *dir;
char dirname[PATH_MAX];
+ if (!dstbuf)
+ return -1;
+
/* depending on kernel version, uio can be located in uio/uioX
* or uio:uioX */
--
1.7.0.7
[dpdk-dev] [PATCH] string: fix potential seg fault on snprintf
Hello Pablo, On 06/27/2014 01:04 PM, Pablo de Lara wrote: > From: Pablo de Lara > > Several functions did not check if destination buffer, used > in snprintf was a non-NULL pointer. Did you noticed any issue without this patch? It seems that all the get_help() cmdline functions are never called with a NULL destination buffer (see in cmdline_parse.c). I think it is useless to add this test as there is no good reason to give a NULL argument. It is like testing that the arguments of strcpy() are non-NULL. I would say the same for pci_get_uio_dev(). Regards, Olivier
[dpdk-dev] [PATCH] string: fix potential seg fault on snprintf
Hi Olivier > -Original Message- > From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Olivier MATZ > Sent: Friday, June 27, 2014 12:30 PM > To: De Lara Guarch, Pablo; dev at dpdk.org > Subject: Re: [dpdk-dev] [PATCH] string: fix potential seg fault on snprintf > > Hello Pablo, > > On 06/27/2014 01:04 PM, Pablo de Lara wrote: > > From: Pablo de Lara > > > > Several functions did not check if destination buffer, used > > in snprintf was a non-NULL pointer. > > Did you noticed any issue without this patch? With last Thomas' patch, cmdline unit test does not pass due to this problem (basically it tests this situation). After test passed, by fixing this issue in several functions, I looked for other places where this happened. > It seems that all the get_help() cmdline functions are never called > with a NULL destination buffer (see in cmdline_parse.c). I think it > is useless to add this test as there is no good reason to give a NULL > argument. It is like testing that the arguments of strcpy() are > non-NULL. > > I would say the same for pci_get_uio_dev(). > So, if people prefer to discard this patch, then we should modify the cmdline unit test , or we could just include the test only in the places needed for the unit test (so, remove it in pci_get_uio_dev, for instance). Regards, Pablo
[dpdk-dev] [PATCH] string: fix potential seg fault on snprintf
Hi Pablo, On 06/27/2014 02:13 PM, De Lara Guarch, Pablo wrote: > With last Thomas' patch, cmdline unit test does not pass due to this problem > (basically it tests this situation). > After test passed, by fixing this issue in several functions, I looked for > other places where this happened. Indeed I missed the unit test, thanks. I think that testing the NULL case is not required. To me, it is like testing snprintf(NULL, ...) Regards, Olivier
[dpdk-dev] [PATCH] test: Ring PMD unit test rollback
From: Pablo de Lara
New ring PMD unit test requires extra EAL option (vdev),
in order to pass, which I believe is incorrect, as this
test should be focused on testing the API, without needing
to pass any argument.
Added again all functions that create all the ring devices
necessary for the test, and remove the last two tests
(test_pmd_ring_pair_create and test_pmd_ring_pair_attach),
as they call functions that are deprecated.
Signed-off-by: Pablo de Lara
---
app/test/test_pmd_ring.c | 217 +++---
1 files changed, 72 insertions(+), 145 deletions(-)
diff --git a/app/test/test_pmd_ring.c b/app/test/test_pmd_ring.c
index 0d3d95c..6988277 100644
--- a/app/test/test_pmd_ring.c
+++ b/app/test/test_pmd_ring.c
@@ -42,6 +42,7 @@
/* two test rings, r1 is used by two ports, r2 just by one */
static struct rte_ring *r1[2], *r2;
+static struct rte_ring *nullring = NULL;
static struct rte_mempool *mp;
static uint8_t start_idx; /* will store the port id of the first of our new
ports */
@@ -49,8 +50,6 @@ static uint8_t start_idx; /* will store the port id of the
first of our new port
#define RX_PORT (uint8_t)(start_idx + 2)
#define RXTX_PORT (uint8_t)(start_idx + 3)
#define RXTX_PORT2 (uint8_t)(start_idx + 4)
-#define RXTX_PORT4 (uint8_t)(start_idx + 6)
-#define RXTX_PORT5 (uint8_t)(start_idx + 7)
#define SOCKET0 0
#define RING_SIZE 256
@@ -58,6 +57,58 @@ static uint8_t start_idx; /* will store the port id of the
first of our new port
#define MBUF_SIZE (2048 + sizeof(struct rte_mbuf) + RTE_PKTMBUF_HEADROOM)
#define NB_MBUF 512
+
+static int
+test_ring_ethdev_create(void)
+{
+ int retval;
+ printf("Testing ring pmd create\n");
+
+ retval = rte_eth_from_rings(NULL, 0, NULL, 0, SOCKET0);
+ if (retval < 0) {
+ printf("Failure, failed to create zero-sized RXTX ring pmd\n");
+ return -1;
+ }
+
+ retval = rte_eth_from_rings(NULL, 0, NULL, 0, RTE_MAX_NUMA_NODES);
+ if (retval >= 0) {
+ printf("Failure, can create ring pmd on socket %d\n",
RTE_MAX_NUMA_NODES);
+ return -1;
+ }
+
+ retval = rte_eth_from_rings(NULL, 1, &r2, 1, SOCKET0);
+ if (retval >= 0) {
+ printf("Failure, can create pmd with null rx rings\n");
+ return -1;
+ }
+
+ retval = rte_eth_from_rings(r1, 1, NULL, 1, SOCKET0);
+ if (retval >= 0) {
+ printf("Failure, can create pmd with null tx rings\n");
+ return -1;
+ }
+
+ retval = rte_eth_from_rings(&nullring, 1, r1, 2, SOCKET0);
+ if (retval < 0) {
+ printf("Failure, failed to create TX-only ring pmd\n");
+ return -1;
+ }
+
+ retval = rte_eth_from_rings(r1, 1, &nullring, 1, SOCKET0);
+ if (retval < 0) {
+ printf("Failure, failed to create RX-only ring pmd\n");
+ return -1;
+ }
+
+ retval = rte_eth_from_rings(&r2, 1, &r2, 1, SOCKET0);
+ if (retval < 0) {
+ printf("Failure, failed to create RXTX ring pmd\n");
+ return -1;
+ }
+
+ return 0;
+}
+
static int
test_ethdev_configure(void)
{
@@ -252,12 +303,26 @@ test_stats_reset(void)
static int
test_pmd_ring_init(void)
{
+ const char * name1 = "R3";
+ const char * name2 = "R4";
+ const char * params_null = NULL;
+ const char * params = "PARAMS";
struct rte_eth_stats stats;
struct rte_mbuf buf, *pbuf = &buf;
struct rte_eth_conf null_conf;
printf("Testing ring pmd init\n");
+ if (rte_pmd_ring_devinit(name1, params_null) < 0) {
+ printf("Testing ring pmd init fail\n");
+ return -1;
+ }
+
+ if (rte_pmd_ring_devinit(name2, params) < 0) {
+ printf("Testing ring pmd init fail\n");
+ return -1;
+ }
+
if (RXTX_PORT2 >= RTE_MAX_ETHPORTS) {
printf(" TX/RX port exceed max eth ports\n");
return -1;
@@ -304,144 +369,8 @@ test_pmd_ring_init(void)
rte_eth_dev_stop(RXTX_PORT2);
- return 0;
-}
-
-static int
-test_pmd_ring_pair_create(void)
-{
- struct rte_eth_stats stats, stats2;
- struct rte_mbuf buf, *pbuf = &buf;
- struct rte_eth_conf null_conf;
-
- if ((RXTX_PORT4 >= RTE_MAX_ETHPORTS) || (RXTX_PORT5 >=
RTE_MAX_ETHPORTS)) {
- printf(" TX/RX port exceed max eth ports\n");
- return -1;
- }
- if ((rte_eth_dev_configure(RXTX_PORT4, 1, 1, &null_conf) < 0)
- || (rte_eth_dev_configure(RXTX_PORT5, 1, 1, &null_conf) < 0)) {
- printf("Configure failed for RXTX port\n");
- return -1;
- }
-
- if ((rte_eth_tx_queue_setup(RXTX_PORT4, 0, RING_SIZE, SOCKET0, NULL) <
0)
- || (rte_eth_tx_queue_setup(RXTX_PORT5, 0, RING_SIZE, SOCKET0,
NULL) < 0)) {
- printf("TX queu
[dpdk-dev] [PATCH] test: Ring PMD unit test rollback
On Fri, Jun 27, 2014 at 03:46:39PM +0100, Pablo de Lara wrote:
> From: Pablo de Lara
>
> New ring PMD unit test requires extra EAL option (vdev),
> in order to pass, which I believe is incorrect, as this
> test should be focused on testing the API, without needing
> to pass any argument.
>
> Added again all functions that create all the ring devices
> necessary for the test, and remove the last two tests
> (test_pmd_ring_pair_create and test_pmd_ring_pair_attach),
> as they call functions that are deprecated.
>
> Signed-off-by: Pablo de Lara
Nak, this also re-adds the need to link the pmd to the application, either
statically or at runtime (i.e. this implicitly obviates the -d option in
rte_eal_parse_args). If you really want to re-add these api calls, I recommend
that you split them into a separate ring library and ring_pmd library.
Neil
P.S. you did remember to use the -d option in your testing, right? Without it
the pmd won't load and the vdev option won't find the proper pmd to parse it.
> ---
> app/test/test_pmd_ring.c | 217
> +++---
> 1 files changed, 72 insertions(+), 145 deletions(-)
>
> diff --git a/app/test/test_pmd_ring.c b/app/test/test_pmd_ring.c
> index 0d3d95c..6988277 100644
> --- a/app/test/test_pmd_ring.c
> +++ b/app/test/test_pmd_ring.c
> @@ -42,6 +42,7 @@
> /* two test rings, r1 is used by two ports, r2 just by one */
> static struct rte_ring *r1[2], *r2;
>
> +static struct rte_ring *nullring = NULL;
> static struct rte_mempool *mp;
> static uint8_t start_idx; /* will store the port id of the first of our new
> ports */
>
> @@ -49,8 +50,6 @@ static uint8_t start_idx; /* will store the port id of the
> first of our new port
> #define RX_PORT (uint8_t)(start_idx + 2)
> #define RXTX_PORT (uint8_t)(start_idx + 3)
> #define RXTX_PORT2 (uint8_t)(start_idx + 4)
> -#define RXTX_PORT4 (uint8_t)(start_idx + 6)
> -#define RXTX_PORT5 (uint8_t)(start_idx + 7)
> #define SOCKET0 0
>
> #define RING_SIZE 256
> @@ -58,6 +57,58 @@ static uint8_t start_idx; /* will store the port id of the
> first of our new port
> #define MBUF_SIZE (2048 + sizeof(struct rte_mbuf) + RTE_PKTMBUF_HEADROOM)
> #define NB_MBUF 512
>
> +
> +static int
> +test_ring_ethdev_create(void)
> +{
> + int retval;
> + printf("Testing ring pmd create\n");
> +
> + retval = rte_eth_from_rings(NULL, 0, NULL, 0, SOCKET0);
> + if (retval < 0) {
> + printf("Failure, failed to create zero-sized RXTX ring pmd\n");
> + return -1;
> + }
> +
> + retval = rte_eth_from_rings(NULL, 0, NULL, 0, RTE_MAX_NUMA_NODES);
> + if (retval >= 0) {
> + printf("Failure, can create ring pmd on socket %d\n",
> RTE_MAX_NUMA_NODES);
> + return -1;
> + }
> +
> + retval = rte_eth_from_rings(NULL, 1, &r2, 1, SOCKET0);
> + if (retval >= 0) {
> + printf("Failure, can create pmd with null rx rings\n");
> + return -1;
> + }
> +
> + retval = rte_eth_from_rings(r1, 1, NULL, 1, SOCKET0);
> + if (retval >= 0) {
> + printf("Failure, can create pmd with null tx rings\n");
> + return -1;
> + }
> +
> + retval = rte_eth_from_rings(&nullring, 1, r1, 2, SOCKET0);
> + if (retval < 0) {
> + printf("Failure, failed to create TX-only ring pmd\n");
> + return -1;
> + }
> +
> + retval = rte_eth_from_rings(r1, 1, &nullring, 1, SOCKET0);
> + if (retval < 0) {
> + printf("Failure, failed to create RX-only ring pmd\n");
> + return -1;
> + }
> +
> + retval = rte_eth_from_rings(&r2, 1, &r2, 1, SOCKET0);
> + if (retval < 0) {
> + printf("Failure, failed to create RXTX ring pmd\n");
> + return -1;
> + }
> +
> + return 0;
> +}
> +
> static int
> test_ethdev_configure(void)
> {
> @@ -252,12 +303,26 @@ test_stats_reset(void)
> static int
> test_pmd_ring_init(void)
> {
> + const char * name1 = "R3";
> + const char * name2 = "R4";
> + const char * params_null = NULL;
> + const char * params = "PARAMS";
> struct rte_eth_stats stats;
> struct rte_mbuf buf, *pbuf = &buf;
> struct rte_eth_conf null_conf;
>
> printf("Testing ring pmd init\n");
>
> + if (rte_pmd_ring_devinit(name1, params_null) < 0) {
> + printf("Testing ring pmd init fail\n");
> + return -1;
> + }
> +
> + if (rte_pmd_ring_devinit(name2, params) < 0) {
> + printf("Testing ring pmd init fail\n");
> + return -1;
> + }
> +
> if (RXTX_PORT2 >= RTE_MAX_ETHPORTS) {
> printf(" TX/RX port exceed max eth ports\n");
> return -1;
> @@ -304,144 +369,8 @@ test_pmd_ring_init(void)
>
> rte_eth_dev_stop(RXTX_PORT2);
>
> - return 0;
> -}
> -
> -static int
> -test_pmd_ring_pair_create(void)
> -{
> - struct rte_eth_stats stats, stats2;
> - struct rte_mbuf bu
[dpdk-dev] [PATCH] test: Ring PMD unit test rollback
Hi Neil, > -Original Message- > From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Neil Horman > Sent: Friday, June 27, 2014 4:03 PM > To: De Lara Guarch, Pablo > Cc: dev at dpdk.org > Subject: Re: [dpdk-dev] [PATCH] test: Ring PMD unit test rollback > > On Fri, Jun 27, 2014 at 03:46:39PM +0100, Pablo de Lara wrote: > > From: Pablo de Lara > > > > New ring PMD unit test requires extra EAL option (vdev), > > in order to pass, which I believe is incorrect, as this > > test should be focused on testing the API, without needing > > to pass any argument. > > > > Added again all functions that create all the ring devices > > necessary for the test, and remove the last two tests > > (test_pmd_ring_pair_create and test_pmd_ring_pair_attach), > > as they call functions that are deprecated. > > > > Signed-off-by: Pablo de Lara > > Nak, this also re-adds the need to link the pmd to the application, either > statically or at runtime (i.e. this implicitly obviates the -d option in > rte_eal_parse_args). If you really want to re-add these api calls, I > recommend > that you split them into a separate ring library and ring_pmd library. > > Neil > > P.S. you did remember to use the -d option in your testing, right? Without it > the pmd won't load and the vdev option won't find the proper pmd to parse > it. > Why does it need the -d? I could use the vdev option without loading any dynamic pmds, and it created the devices, but it is just difficult to run a simple unit test, by having to know which ethdevs you have to create, instead of letting the test itself to create the ones that it needs, if you know what I mean. So, if this solution is not good enough, another idea could be to let the unit test make calls to the application, creating the virtual devices needed. What I don't like is the idea of having to pass the vdev to the generic test application, as it is intended to run other tests and not just the ring pmd (we could do something like in the eal_flags test, where several test app instances are run, with different parameters, although I am not sure, if this is possible for this kind of test). Thanks, Pablo
[dpdk-dev] [PATCH] test: Ring PMD unit test rollback
On Fri, Jun 27, 2014 at 03:47:07PM +, De Lara Guarch, Pablo wrote: > Hi Neil, > > > -Original Message- > > From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Neil Horman > > Sent: Friday, June 27, 2014 4:03 PM > > To: De Lara Guarch, Pablo > > Cc: dev at dpdk.org > > Subject: Re: [dpdk-dev] [PATCH] test: Ring PMD unit test rollback > > > > On Fri, Jun 27, 2014 at 03:46:39PM +0100, Pablo de Lara wrote: > > > From: Pablo de Lara > > > > > > New ring PMD unit test requires extra EAL option (vdev), > > > in order to pass, which I believe is incorrect, as this > > > test should be focused on testing the API, without needing > > > to pass any argument. > > > > > > Added again all functions that create all the ring devices > > > necessary for the test, and remove the last two tests > > > (test_pmd_ring_pair_create and test_pmd_ring_pair_attach), > > > as they call functions that are deprecated. > > > > > > Signed-off-by: Pablo de Lara > > > > Nak, this also re-adds the need to link the pmd to the application, either > > statically or at runtime (i.e. this implicitly obviates the -d option in > > rte_eal_parse_args). If you really want to re-add these api calls, I > > recommend > > that you split them into a separate ring library and ring_pmd library. > > > > Neil > > > > P.S. you did remember to use the -d option in your testing, right? Without > > it > > the pmd won't load and the vdev option won't find the proper pmd to parse > > it. > > > > Why does it need the -d? I could use the vdev option without loading any > dynamic pmds, and You don't need it only if you build the application statically. If you build it as shared objects you need to specify which pmds to load at run time using the -d option. If you're building statically then no, you don't need it. > it created the devices, but it is just difficult to run a simple unit test, > by having to know which > ethdevs you have to create, instead of letting the test itself to create the > ones that it needs, > if you know what I mean. So, if this solution is not good enough, another > idea could be to let > the unit test make calls to the application, creating the virtual devices > needed. What I don't > like is the idea of having to pass the vdev to the generic test application, > as it is intended to > run other tests and not just the ring pmd (we could do something like in the > eal_flags test, > where several test app instances are run, with different parameters, although > I am not sure, > if this is possible for this kind of test). > Alternatively to restoring the API, you can hardcode the command line strings into the test pmd, so you don't have to specify them at run time, and just pass them directly to rte_eal_init. The problem with just blindly adding back the API is that it requires any application that might want to use the ring pmd to load it at run time unilaterally. Regards Neil > Thanks, > Pablo >
[dpdk-dev] [PATCH] string: fix potential seg fault on snprintf
> -Original Message- > From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Olivier MATZ > Sent: Friday, June 27, 2014 5:34 AM > To: De Lara Guarch, Pablo; dev at dpdk.org > Subject: Re: [dpdk-dev] [PATCH] string: fix potential seg fault on snprintf > > Hi Pablo, > > On 06/27/2014 02:13 PM, De Lara Guarch, Pablo wrote: > > With last Thomas' patch, cmdline unit test does not pass due to this problem > (basically it tests this situation). > > After test passed, by fixing this issue in several functions, I looked for > > other > places where this happened. > > Indeed I missed the unit test, thanks. > > I think that testing the NULL case is not required. To me, it is like > testing snprintf(NULL, ...) > Famous last words include "that could never happen!" :-) Since this is not a performance critical piece of code, it does not hurt to leave the Null-check in, and get the additional safety of checking for invalid inputs.
[dpdk-dev] [PATCH v10 0/5] link bonding
2014-06-27 11:18, Declan Doherty: > Fixed copyright in link bonding library Makefile A small fix is needed in Makefile for renaming of file rte_eth_bond_args.c. And more important, it cannot be built as an optional shared library as test applications call functions from the specific API. The best design would be to have a base library always linked and an optional PMD layer. -- Thomas
