d/numam-dpdk
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity
f0ca77fd60
numam-dpdk/drivers/net/vmxnet3/vmxnet3_ethdev.c

1466 lines
42 KiB
C
Raw Normal View History

drivers: use SPDX tag for Intel copyright files Replace the BSD license header with the SPDX tag for files with only an Intel copyright on them. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2017-12-19 15:49:01 +00:00
/* SPDX-License-Identifier: BSD-3-Clause
* Copyright(c) 2010-2015 Intel Corporation
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
*/
#include <sys/queue.h>
#include <stdio.h>
#include <errno.h>
#include <stdint.h>
#include <string.h>
#include <unistd.h>
#include <stdarg.h>
#include <fcntl.h>
#include <inttypes.h>
#include <rte_byteorder.h>
#include <rte_common.h>
#include <rte_cycles.h>
#include <rte_interrupts.h>
#include <rte_log.h>
#include <rte_debug.h>
#include <rte_pci.h>
pci: introduce library and driver The PCI lib defines the types and methods allowing to use PCI elements. The PCI bus implements a bus driver for PCI devices by constructing rte_bus elements using the PCI lib. Move the relevant code out of the EAL to its expected place. Libraries, drivers, unit tests and applications are updated to use the new rte_bus_pci.h header when necessary. Signed-off-by: Gaetan Rivet <gaetan.rivet@6wind.com>
2017-10-26 10:06:08 +00:00
#include <rte_bus_pci.h>
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
#include <rte_branch_prediction.h>
#include <rte_memory.h>
#include <rte_memzone.h>
#include <rte_eal.h>
#include <rte_alarm.h>
#include <rte_ether.h>
ethdev: separate driver APIs Create a rte_ethdev_driver.h file and move PMD specific APIs here. Drivers updated to include this new header file. There is no update in header content and since ethdev.h included by ethdev_driver.h, nothing changed from driver point of view, only logically grouping of APIs. From applications point of view they can't access to driver specific APIs anymore and they shouldn't. More PMD specific data structures still remain in ethdev.h because of inline functions in header use them. Those will be handled separately. Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com> Acked-by: Shreyansh Jain <shreyansh.jain@nxp.com> Acked-by: Andrew Rybchenko <arybchenko@solarflare.com> Acked-by: Thomas Monjalon <thomas@monjalon.net>
2018-01-22 00:16:22 +00:00
#include <rte_ethdev_driver.h>
drivers/net: do not use ethdev driver Signed-off-by: Jan Blunck <jblunck@infradead.org> Signed-off-by: Gaetan Rivet <gaetan.rivet@6wind.com> Acked-by: Stephen Hemminger <stephen@networkplumber.org>
2017-04-11 15:44:24 +00:00
#include <rte_ethdev_pci.h>
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
#include <rte_string_fns.h>
#include <rte_malloc.h>
vmxnet3: convert to use of PMD_REGISTER_DRIVER and fix linking Convert the vmxnet3 pmd driver to use the PMD_REGISTER_DRIVER macro. This means that the test applications now have no reference to the vmxnet3 library when building DSO's and must specify its use on the command line with the -d option. Static linking will still initalize the driver automatically. Signed-off-by: Neil Horman <nhorman@tuxdriver.com> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-04-21 14:59:38 +00:00
#include <rte_dev.h>
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
vmxnet3: move to drivers/net/ Move vmxnet3 PMD to drivers/net directory. As part of the move, rename the "vmxnet3" subdirectory, containing the original FreeBSD drivers, from "vmxnet3" to the more standard name "base", to indicate it contains the base drivers used for the implementation. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com> Acked-by: John McNamara <john.mcnamara@intel.com> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2015-05-15 15:56:59 +00:00
#include "base/vmxnet3_defs.h"
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
#include "vmxnet3_ring.h"
#include "vmxnet3_logs.h"
#include "vmxnet3_ethdev.h"
#define PROCESS_SYS_EVENTS 0
net/vmxnet3: add Tx preparation Signed-off-by: Konstantin Ananyev <konstantin.ananyev@intel.com> Acked-by: Yong Wang <yongwang@vmware.com>
2016-12-23 18:40:52 +00:00
#define VMXNET3_TX_MAX_SEG UINT8_MAX
net/vmxnet3: convert to new Rx offload API Ethdev RX offloads API has changed since: commit ce17eddefc20 ("ethdev: introduce Rx queue offloads API") This patch adopts the new RX Offload API in vmxnet3 driver. Signed-off-by: Louis Luo <llouis@vmware.com> Acked-by: Yong Wang <yongwang@vmware.com>
2018-05-01 21:22:56 +00:00
#define VMXNET3_TX_OFFLOAD_CAP \
(DEV_TX_OFFLOAD_VLAN_INSERT | \
DEV_TX_OFFLOAD_TCP_CKSUM | \
DEV_TX_OFFLOAD_UDP_CKSUM | \
DEV_TX_OFFLOAD_TCP_TSO | \
DEV_TX_OFFLOAD_MULTI_SEGS)
#define VMXNET3_RX_OFFLOAD_CAP \
(DEV_RX_OFFLOAD_VLAN_STRIP | \
net/vmxnet3: add VLAN filter capability The driver supports VLAN filtering, but the option is not included in the advertised list of capabilities. Thus any attempt to enable VLAN filtering always fails. Update the list of offload capabilities to include DEV_RX_OFFLOAD_VLAN_FILTER. Fixes: 95e4a96ccbf1 ("net/vmxnet3: convert to new Rx offload API") Cc: stable@dpdk.org Signed-off-by: Mark Gillott <mgillott@vyatta.att-mail.com> Acked-by: Yong Wang <yongwang@vmware.com>
2019-05-01 15:29:42 +00:00
DEV_RX_OFFLOAD_VLAN_FILTER | \
net/vmxnet3: convert to new Rx offload API Ethdev RX offloads API has changed since: commit ce17eddefc20 ("ethdev: introduce Rx queue offloads API") This patch adopts the new RX Offload API in vmxnet3 driver. Signed-off-by: Louis Luo <llouis@vmware.com> Acked-by: Yong Wang <yongwang@vmware.com>
2018-05-01 21:22:56 +00:00
DEV_RX_OFFLOAD_SCATTER | \
DEV_RX_OFFLOAD_UDP_CKSUM | \
DEV_RX_OFFLOAD_TCP_CKSUM | \
DEV_RX_OFFLOAD_TCP_LRO | \
drivers/net: update Rx RSS hash offload capabilities Add DEV_RX_OFFLOAD_RSS_HASH flag for all PMDs that support RSS hash delivery. Signed-off-by: Pavan Nikhilesh <pbhagavatula@marvell.com> Reviewed-by: Andrew Rybchenko <arybchenko@solarflare.com> Reviewed-by: Hemant Agrawal <hemant.agrawal@nxp.com> Acked-by: Jerin Jacob <jerinj@marvell.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2019-11-11 13:19:08 +00:00
DEV_RX_OFFLOAD_JUMBO_FRAME | \
DEV_RX_OFFLOAD_RSS_HASH)
net/vmxnet3: convert to new Rx offload API Ethdev RX offloads API has changed since: commit ce17eddefc20 ("ethdev: introduce Rx queue offloads API") This patch adopts the new RX Offload API in vmxnet3 driver. Signed-off-by: Louis Luo <llouis@vmware.com> Acked-by: Yong Wang <yongwang@vmware.com>
2018-05-01 21:22:56 +00:00
ethdev: remove useless parameter in init functions The pointer to struct eth_driver is not used and is already set in struct rte_eth_dev. It's a small cleanup in PMD API which probably needs more attention to make clear what is a driver, a PCI driver, an ethernet driver, etc. Signed-off-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2015-03-05 18:24:59 +00:00
static int eth_vmxnet3_dev_init(struct rte_eth_dev *eth_dev);
vmxnet3: support port hotplug Signed-off-by: Bernard Iremonger <bernard.iremonger@intel.com> Acked-by: Yong Wang <yongwang@vmware.com>
2015-10-27 17:12:26 +00:00
static int eth_vmxnet3_dev_uninit(struct rte_eth_dev *eth_dev);
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
static int vmxnet3_dev_configure(struct rte_eth_dev *dev);
static int vmxnet3_dev_start(struct rte_eth_dev *dev);
static void vmxnet3_dev_stop(struct rte_eth_dev *dev);
static void vmxnet3_dev_close(struct rte_eth_dev *dev);
static void vmxnet3_dev_set_rxmode(struct vmxnet3_hw *hw, uint32_t feature, int set);
ethdev: change promiscuous callbacks to return status Enabling/disabling of promiscuous mode is not always successful and it should be taken into account to be able to handle it properly. When correct return status is unclear from driver code, -EAGAIN is used. Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Acked-by: Matan Azrad <matan@mellanox.com> Acked-by: Hyong Youb Kim <hyonkim@cisco.com>
2019-09-14 11:37:24 +00:00
static int vmxnet3_dev_promiscuous_enable(struct rte_eth_dev *dev);
static int vmxnet3_dev_promiscuous_disable(struct rte_eth_dev *dev);
ethdev: change allmulticast callbacks to return status Enabling/disabling of allmulticast mode is not always successful and it should be taken into account to be able to handle it properly. When correct return status is unclear from driver code, -EAGAIN is used. Signed-off-by: Ivan Ilchenko <ivan.ilchenko@oktetlabs.ru> Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Acked-by: Hyong Youb Kim <hyonkim@cisco.com>
2019-09-24 12:56:10 +00:00
static int vmxnet3_dev_allmulticast_enable(struct rte_eth_dev *dev);
static int vmxnet3_dev_allmulticast_disable(struct rte_eth_dev *dev);
net/vmxnet3: generate link-state change notifications Generate link-state change notifications by listening to interrupts generated by the device. Make use of the existing vmxnet3_process_events function that was compiled out, but change it to call vmxnet3_dev_link_update on a VMXNET3_ECR_LINK event and to not be so noisy in its log messages. Enable interrupts on starting the device, using a new helper function, vmxnet3_enable_intr, based on vmxnet3_disable_intr and validated against the FreeBSD driver. Keep track of the number of interrupts registered for to avoid hardcoding these in vmxnet3_enable/disable_intr and to provision for any future rxq intr support. Factor out the guts of vmxnet3_dev_link_update minus the started check to allow the new function to be called from vmxnet3_dev_start in the lsc-enabled case to ensure that the link state is correctly set from the actual state at that point. Signed-off-by: Robert Shearman <rshearma@brocade.com>
2017-06-15 12:16:10 +00:00
static int __vmxnet3_dev_link_update(struct rte_eth_dev *dev,
int wait_to_complete);
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
static int vmxnet3_dev_link_update(struct rte_eth_dev *dev,
net/vmxnet3: coding style changes Signed-off-by: Yong Wang <yongwang@vmware.com> Acked-by: Stephen Hemminger <stephen@networkplumber.org>
2016-08-29 19:18:46 +00:00
int wait_to_complete);
net/vmxnet3: retain counters on restart Most NICs like virtio, igb/ixgbe etc. don't reset counters on dev_start and arguably this helps in monitoring the counters across a longer time span with multiple device start/stops. vmxnet3 behavior is opposite to that and counters are reset by the host side implementation each time the device is restarted. Change the driver to save the counters in its private context before it is reset by writing CMD_ACTIVATE to REG_CMD. Signed-off-by: Nachiketa Prachanda <nprachan@brocade.com> Acked-by: Shrikrishna Khare <skhare@vmware.com>
2017-06-15 12:16:08 +00:00
static void vmxnet3_hw_stats_save(struct vmxnet3_hw *hw);
ethdev: add return value to stats get dev op The stats_get dev op API doesn't include return value, so PMD cannot return an error in case of failure at stats getting process time. Since PCI devices can be removed and there is a time between the physical removal to the RMV interrupt, the user may get invalid stats without any indication. This patch changes the stats_get API return value to be int instead of void. All the net PMDs stats_get dev ops are adjusted by this patch. Signed-off-by: Matan Azrad <matan@mellanox.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2017-10-10 20:20:18 +00:00
static int vmxnet3_dev_stats_get(struct rte_eth_dev *dev,
net/vmxnet3: coding style changes Signed-off-by: Yong Wang <yongwang@vmware.com> Acked-by: Stephen Hemminger <stephen@networkplumber.org>
2016-08-29 19:18:46 +00:00
struct rte_eth_stats *stats);
ethdev: make stats and xstats reset callbacks return int Change return value of the callbacks from void to int. Make implementations across all drivers return negative errno values in case of error conditions. Both callbacks are updated together because a large number of drivers assign the same function to both callbacks. Signed-off-by: Igor Romanov <igor.romanov@oktetlabs.ru> Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2019-09-06 14:34:54 +00:00
static int vmxnet3_dev_stats_reset(struct rte_eth_dev *dev);
net/vmxnet3: implement extended stats Implement xstats_get() to allow a number of driver-specific Tx and Rx stats to be retrieved. Signed-off-by: Robert Shearman <rshearma@brocade.com> Acked-by: Shrikrishna Khare <skhare@vmware.com>
2017-06-15 12:16:09 +00:00
static int vmxnet3_dev_xstats_get_names(struct rte_eth_dev *dev,
struct rte_eth_xstat_name *xstats,
unsigned int n);
static int vmxnet3_dev_xstats_get(struct rte_eth_dev *dev,
struct rte_eth_xstat *xstats, unsigned int n);
ethdev: change device info get callback to return int Change eth_dev_infos_get_t return value from void to int. Make eth_dev_infos_get_t implementations across all drivers to return negative errno values if case of error conditions. Signed-off-by: Ivan Ilchenko <ivan.ilchenko@oktetlabs.ru> Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2019-09-12 16:42:28 +00:00
static int vmxnet3_dev_info_get(struct rte_eth_dev *dev,
struct rte_eth_dev_info *dev_info);
ethdev: query supported packet types Add a new API rte_eth_dev_get_supported_ptypes to query what packet types can be filled by a given device. The device should be already started or its PMD RX burst function already decided, since the packet types supported may vary depending on RX function. Signed-off-by: Jianfeng Tan <jianfeng.tan@intel.com> Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com> Acked-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>
2016-03-14 20:50:50 +00:00
static const uint32_t *
vmxnet3_dev_supported_ptypes_get(struct rte_eth_dev *dev);
net/vmxnet3: support MTU set (Picked up from @Charles Myers patch https://patchwork.dpdk.org/patch/57771/) When the mtu_set() function is not implemented, rte_eth_dev_set_mtu() fails with -ENOTSUP and mtu is not stored in the mtu field in the rte_eth_dev_data. This causes the mtu in Vmxnet3_MiscConf which is shared with hypervisor to always be set to 1500. This may cause issues receiving jumbo frames on Enhanced Data Path N-VDS. Signed-off-by: Eduard Serra <eserra@vmware.com> Acked-by: Yong Wang <yongwang@vmware.com>
2020-03-25 19:18:51 +00:00
static int vmxnet3_dev_mtu_set(struct rte_eth_dev *dev, uint16_t mtu);
vmxnet3: enable vlan filtering Support the VLAN filter functionality of the VMXNET3 interface. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Yong Wang <yongwang@vmware.com>
2015-07-09 18:24:09 +00:00
static int vmxnet3_dev_vlan_filter_set(struct rte_eth_dev *dev,
uint16_t vid, int on);
ethdev: allow returning error on VLAN offload ops Some devices may not support or fail setting VLAN offload configuration based on dynamic circumstances so the vlan_offload_set_t vector is modified to return an int so the caller can determine success or not. rte_eth_dev_set_vlan_offload is updated to return the value provided by the vector when called along with restoring the original offload configs on failure. Existing vlan_offload_set_t vectors are modified to return an int. Majority of cases return 0 but a few that actually can fail now return their failure codes. Finally, a vlan_offload_set_t vector is added to virtio to facilitate dynamically turning VLAN strip on or off. Signed-off-by: David Harton <dharton@cisco.com> Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com>
2017-09-01 02:36:28 +00:00
static int vmxnet3_dev_vlan_offload_set(struct rte_eth_dev *dev, int mask);
ethdev: return diagnostic when setting MAC address Change the prototype and the behavior of dev_ops->eth_mac_addr_set(): a return code is added to notify the caller (librte_ether) if an error occurred in the PMD. The new default MAC address is now copied in dev->data->mac_addrs[0] only if the operation is successful. The patch also updates all the PMDs accordingly. Signed-off-by: Olivier Matz <olivier.matz@6wind.com> Signed-off-by: Ivan Malov <ivan.malov@oktetlabs.ru> Acked-by: Andrew Rybchenko <arybchenko@solarflare.com> Acked-by: Adrien Mazarguil <adrien.mazarguil@6wind.com> Acked-by: Shreyansh Jain <shreyansh.jain@nxp.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com> Acked-by: Thomas Monjalon <thomas@monjalon.net> Acked-by: Nelio Laranjeiro <nelio.laranjeiro@6wind.com>
2018-04-11 16:32:51 +00:00
static int vmxnet3_mac_addr_set(struct rte_eth_dev *dev,
net: add rte prefix to ether structures Add 'rte_' prefix to structures: - rename struct ether_addr as struct rte_ether_addr. - rename struct ether_hdr as struct rte_ether_hdr. - rename struct vlan_hdr as struct rte_vlan_hdr. - rename struct vxlan_hdr as struct rte_vxlan_hdr. - rename struct vxlan_gpe_hdr as struct rte_vxlan_gpe_hdr. Do not update the command line library to avoid adding a dependency to librte_net. Signed-off-by: Olivier Matz <olivier.matz@6wind.com> Reviewed-by: Stephen Hemminger <stephen@networkplumber.org> Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2019-05-21 16:13:03 +00:00
struct rte_ether_addr *mac_addr);
net/vmxnet3: generate link-state change notifications Generate link-state change notifications by listening to interrupts generated by the device. Make use of the existing vmxnet3_process_events function that was compiled out, but change it to call vmxnet3_dev_link_update on a VMXNET3_ECR_LINK event and to not be so noisy in its log messages. Enable interrupts on starting the device, using a new helper function, vmxnet3_enable_intr, based on vmxnet3_disable_intr and validated against the FreeBSD driver. Keep track of the number of interrupts registered for to avoid hardcoding these in vmxnet3_enable/disable_intr and to provision for any future rxq intr support. Factor out the guts of vmxnet3_dev_link_update minus the started check to allow the new function to be called from vmxnet3_dev_start in the lsc-enabled case to ensure that the link state is correctly set from the actual state at that point. Signed-off-by: Robert Shearman <rshearma@brocade.com>
2017-06-15 12:16:10 +00:00
static void vmxnet3_interrupt_handler(void *param);
vmxnet3: enable vlan filtering Support the VLAN filter functionality of the VMXNET3 interface. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Yong Wang <yongwang@vmware.com>
2015-07-09 18:24:09 +00:00
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
/*
* The set of PCI devices this driver supports
*/
net/vmxnet3: move PCI device ids to the driver Moved vmware device ids macro since the driver had no such information. Used RTE_PCI_DEVICE in place of RTE_PCI_DEV_ID_DECL* stuff. Signed-off-by: David Marchand <david.marchand@6wind.com>
2016-07-11 14:40:41 +00:00
#define VMWARE_PCI_VENDOR_ID 0x15AD
#define VMWARE_DEV_ID_VMXNET3 0x07B0
pci: make device id tables const The PCI device id table is immutable and should be made const in all drivers. The pseudo drivers can initialize their local copy as necessary. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Neil Horman <nhorman@tuxdriver.com>
2015-04-16 23:23:39 +00:00
static const struct rte_pci_id pci_id_vmxnet3_map[] = {
net/vmxnet3: move PCI device ids to the driver Moved vmware device ids macro since the driver had no such information. Used RTE_PCI_DEVICE in place of RTE_PCI_DEV_ID_DECL* stuff. Signed-off-by: David Marchand <david.marchand@6wind.com>
2016-07-11 14:40:41 +00:00
{ RTE_PCI_DEVICE(VMWARE_PCI_VENDOR_ID, VMWARE_DEV_ID_VMXNET3) },
{ .vendor_id = 0, /* sentinel */ },
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
};
ethdev: make dev_ops const The ethernet device ops function table should be made const for safety and security. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Neil Horman <nhorman@tuxdriver.com>
2015-04-07 21:21:03 +00:00
static const struct eth_dev_ops vmxnet3_eth_dev_ops = {
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
.dev_configure = vmxnet3_dev_configure,
.dev_start = vmxnet3_dev_start,
.dev_stop = vmxnet3_dev_stop,
.dev_close = vmxnet3_dev_close,
.promiscuous_enable = vmxnet3_dev_promiscuous_enable,
.promiscuous_disable = vmxnet3_dev_promiscuous_disable,
.allmulticast_enable = vmxnet3_dev_allmulticast_enable,
.allmulticast_disable = vmxnet3_dev_allmulticast_disable,
.link_update = vmxnet3_dev_link_update,
.stats_get = vmxnet3_dev_stats_get,
net/vmxnet3: implement extended stats Implement xstats_get() to allow a number of driver-specific Tx and Rx stats to be retrieved. Signed-off-by: Robert Shearman <rshearma@brocade.com> Acked-by: Shrikrishna Khare <skhare@vmware.com>
2017-06-15 12:16:09 +00:00
.xstats_get_names = vmxnet3_dev_xstats_get_names,
.xstats_get = vmxnet3_dev_xstats_get,
net/vmxnet3: support stats reset 'stats_reset()' callback was missing because the device backend doesn't support it. This commit adds a workaround to this and implements the callback by taking a snapshot of the stats (SNAPSHOT) each time 'stats_reset()' is called. When getting stats with 'stats_get()', hw stats which always increase reduce SNAPSHOT stats. That's how we get the "real" stats since the last 'stats_reset()'. Signed-off-by: Yogev Chaimovich <yogev@cgstowernetworks.com> Acked-by: Yong Wang <yongwang@vmware.com>
2018-09-20 06:37:05 +00:00
.stats_reset = vmxnet3_dev_stats_reset,
net/vmxnet3: coding style changes Signed-off-by: Yong Wang <yongwang@vmware.com> Acked-by: Stephen Hemminger <stephen@networkplumber.org>
2016-08-29 19:18:46 +00:00
.mac_addr_set = vmxnet3_mac_addr_set,
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
.dev_infos_get = vmxnet3_dev_info_get,
ethdev: query supported packet types Add a new API rte_eth_dev_get_supported_ptypes to query what packet types can be filled by a given device. The device should be already started or its PMD RX burst function already decided, since the packet types supported may vary depending on RX function. Signed-off-by: Jianfeng Tan <jianfeng.tan@intel.com> Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com> Acked-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>
2016-03-14 20:50:50 +00:00
.dev_supported_ptypes_get = vmxnet3_dev_supported_ptypes_get,
net/vmxnet3: support MTU set (Picked up from @Charles Myers patch https://patchwork.dpdk.org/patch/57771/) When the mtu_set() function is not implemented, rte_eth_dev_set_mtu() fails with -ENOTSUP and mtu is not stored in the mtu field in the rte_eth_dev_data. This causes the mtu in Vmxnet3_MiscConf which is shared with hypervisor to always be set to 1500. This may cause issues receiving jumbo frames on Enhanced Data Path N-VDS. Signed-off-by: Eduard Serra <eserra@vmware.com> Acked-by: Yong Wang <yongwang@vmware.com>
2020-03-25 19:18:51 +00:00
.mtu_set = vmxnet3_dev_mtu_set,
vmxnet3: enable vlan filtering Support the VLAN filter functionality of the VMXNET3 interface. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Yong Wang <yongwang@vmware.com>
2015-07-09 18:24:09 +00:00
.vlan_filter_set = vmxnet3_dev_vlan_filter_set,
.vlan_offload_set = vmxnet3_dev_vlan_offload_set,
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
.rx_queue_setup = vmxnet3_dev_rx_queue_setup,
.rx_queue_release = vmxnet3_dev_rx_queue_release,
.tx_queue_setup = vmxnet3_dev_tx_queue_setup,
.tx_queue_release = vmxnet3_dev_tx_queue_release,
};
net/vmxnet3: implement extended stats Implement xstats_get() to allow a number of driver-specific Tx and Rx stats to be retrieved. Signed-off-by: Robert Shearman <rshearma@brocade.com> Acked-by: Shrikrishna Khare <skhare@vmware.com>
2017-06-15 12:16:09 +00:00
struct vmxnet3_xstats_name_off {
char name[RTE_ETH_XSTATS_NAME_SIZE];
unsigned int offset;
};
/* tx_qX_ is prepended to the name string here */
static const struct vmxnet3_xstats_name_off vmxnet3_txq_stat_strings[] = {
{"drop_total", offsetof(struct vmxnet3_txq_stats, drop_total)},
{"drop_too_many_segs", offsetof(struct vmxnet3_txq_stats, drop_too_many_segs)},
{"drop_tso", offsetof(struct vmxnet3_txq_stats, drop_tso)},
{"tx_ring_full", offsetof(struct vmxnet3_txq_stats, tx_ring_full)},
};
/* rx_qX_ is prepended to the name string here */
static const struct vmxnet3_xstats_name_off vmxnet3_rxq_stat_strings[] = {
{"drop_total", offsetof(struct vmxnet3_rxq_stats, drop_total)},
{"drop_err", offsetof(struct vmxnet3_rxq_stats, drop_err)},
{"drop_fcs", offsetof(struct vmxnet3_rxq_stats, drop_fcs)},
{"rx_buf_alloc_failure", offsetof(struct vmxnet3_rxq_stats, rx_buf_alloc_failure)},
};
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
static const struct rte_memzone *
gpa_zone_reserve(struct rte_eth_dev *dev, uint32_t size,
net/vmxnet3: reallocate shared memzone on re-config When adding a DPDK port to ovs-vswitchd with DPDK, the vmxnet3 device fails to activate due to mismatched magic number. This failure causes following operations to run: start the port, stop the port, reconfigure and re-start the port. During reconfigure, if there is an existing memzone, driver will reuse it. But reconfigure may request different number of Tx/Rx queues. This results in a memzone with wrong size and potential invalid memory access. To fix this, free the memzone if found and reserve a new one. Signed-off-by: Yong Wang <yongwang@vmware.com> Reviewed-by: Guolin Yang <gyang@vmware.com> Reviewed-by: Daniele Di Proietto <ddiproietto@vmware.com> Tested-by: Daniele Di Proietto <ddiproietto@vmware.com> Acked-by: Stephen Hemminger <stephen@networkplumber.org>
2016-08-29 19:18:47 +00:00
const char *post_string, int socket_id,
uint16_t align, bool reuse)
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
{
char z_name[RTE_MEMZONE_NAMESIZE];
const struct rte_memzone *mz;
ethdev: rename memzones allocated for DMA The helper rte_eth_dma_zone_reserve() is called by PMDs when probing a new port. It creates a new memzone with an unique name. The name of this memzone was using the name of the driver doing the probe. In order to avoid assigning the driver before the end of the probing, the driver name is removed from these memzone names. The ethdev name (data->name) is not used because it may be too long and may be not set at this stage of probing. Syntax of old name: <driver>_<ring>_<port>_<queue> Syntax of new name: eth_p<port>_q<queue>_<ring> Signed-off-by: Thomas Monjalon <thomas@monjalon.net> Tested-by: Andrew Rybchenko <arybchenko@solarflare.com>
2018-10-11 12:58:49 +00:00
snprintf(z_name, sizeof(z_name), "eth_p%d_%s",
dev->data->port_id, post_string);
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
mz = rte_memzone_lookup(z_name);
net/vmxnet3: reallocate shared memzone on re-config When adding a DPDK port to ovs-vswitchd with DPDK, the vmxnet3 device fails to activate due to mismatched magic number. This failure causes following operations to run: start the port, stop the port, reconfigure and re-start the port. During reconfigure, if there is an existing memzone, driver will reuse it. But reconfigure may request different number of Tx/Rx queues. This results in a memzone with wrong size and potential invalid memory access. To fix this, free the memzone if found and reserve a new one. Signed-off-by: Yong Wang <yongwang@vmware.com> Reviewed-by: Guolin Yang <gyang@vmware.com> Reviewed-by: Daniele Di Proietto <ddiproietto@vmware.com> Tested-by: Daniele Di Proietto <ddiproietto@vmware.com> Acked-by: Stephen Hemminger <stephen@networkplumber.org>
2016-08-29 19:18:47 +00:00
if (!reuse) {
if (mz)
rte_memzone_free(mz);
return rte_memzone_reserve_aligned(z_name, size, socket_id,
net/vmxnet3: use contiguous allocation for DMA memory All hardware drivers should allocate IOVA-contiguous memzones for their hardware resources. Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com> Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com> Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com> Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>
2018-04-11 12:29:58 +00:00
RTE_MEMZONE_IOVA_CONTIG, align);
net/vmxnet3: reallocate shared memzone on re-config When adding a DPDK port to ovs-vswitchd with DPDK, the vmxnet3 device fails to activate due to mismatched magic number. This failure causes following operations to run: start the port, stop the port, reconfigure and re-start the port. During reconfigure, if there is an existing memzone, driver will reuse it. But reconfigure may request different number of Tx/Rx queues. This results in a memzone with wrong size and potential invalid memory access. To fix this, free the memzone if found and reserve a new one. Signed-off-by: Yong Wang <yongwang@vmware.com> Reviewed-by: Guolin Yang <gyang@vmware.com> Reviewed-by: Daniele Di Proietto <ddiproietto@vmware.com> Tested-by: Daniele Di Proietto <ddiproietto@vmware.com> Acked-by: Stephen Hemminger <stephen@networkplumber.org>
2016-08-29 19:18:47 +00:00
}
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
if (mz)
return mz;
net/vmxnet3: use contiguous allocation for DMA memory All hardware drivers should allocate IOVA-contiguous memzones for their hardware resources. Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com> Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com> Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com> Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>
2018-04-11 12:29:58 +00:00
return rte_memzone_reserve_aligned(z_name, size, socket_id,
RTE_MEMZONE_IOVA_CONTIG, align);
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
}
/*
* This function is based on vmxnet3_disable_intr()
*/
static void
vmxnet3_disable_intr(struct vmxnet3_hw *hw)
{
int i;
PMD_INIT_FUNC_TRACE();
hw->shared->devRead.intrConf.intrCtrl |= VMXNET3_IC_DISABLE_ALL;
net/vmxnet3: generate link-state change notifications Generate link-state change notifications by listening to interrupts generated by the device. Make use of the existing vmxnet3_process_events function that was compiled out, but change it to call vmxnet3_dev_link_update on a VMXNET3_ECR_LINK event and to not be so noisy in its log messages. Enable interrupts on starting the device, using a new helper function, vmxnet3_enable_intr, based on vmxnet3_disable_intr and validated against the FreeBSD driver. Keep track of the number of interrupts registered for to avoid hardcoding these in vmxnet3_enable/disable_intr and to provision for any future rxq intr support. Factor out the guts of vmxnet3_dev_link_update minus the started check to allow the new function to be called from vmxnet3_dev_start in the lsc-enabled case to ensure that the link state is correctly set from the actual state at that point. Signed-off-by: Robert Shearman <rshearma@brocade.com>
2017-06-15 12:16:10 +00:00
for (i = 0; i < hw->num_intrs; i++)
net/vmxnet3: coding style changes Signed-off-by: Yong Wang <yongwang@vmware.com> Acked-by: Stephen Hemminger <stephen@networkplumber.org>
2016-08-29 19:18:46 +00:00
VMXNET3_WRITE_BAR0_REG(hw, VMXNET3_REG_IMR + i * 8, 1);
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
}
net/vmxnet3: generate link-state change notifications Generate link-state change notifications by listening to interrupts generated by the device. Make use of the existing vmxnet3_process_events function that was compiled out, but change it to call vmxnet3_dev_link_update on a VMXNET3_ECR_LINK event and to not be so noisy in its log messages. Enable interrupts on starting the device, using a new helper function, vmxnet3_enable_intr, based on vmxnet3_disable_intr and validated against the FreeBSD driver. Keep track of the number of interrupts registered for to avoid hardcoding these in vmxnet3_enable/disable_intr and to provision for any future rxq intr support. Factor out the guts of vmxnet3_dev_link_update minus the started check to allow the new function to be called from vmxnet3_dev_start in the lsc-enabled case to ensure that the link state is correctly set from the actual state at that point. Signed-off-by: Robert Shearman <rshearma@brocade.com>
2017-06-15 12:16:10 +00:00
static void
vmxnet3_enable_intr(struct vmxnet3_hw *hw)
{
int i;
PMD_INIT_FUNC_TRACE();
hw->shared->devRead.intrConf.intrCtrl &= ~VMXNET3_IC_DISABLE_ALL;
for (i = 0; i < hw->num_intrs; i++)
VMXNET3_WRITE_BAR0_REG(hw, VMXNET3_REG_IMR + i * 8, 0);
}
net/vmxnet3: allow variable length Tx data ring vmxnet3 driver supports transmit data ring viz. a set of fixed size buffers used by the driver to copy packet headers. Small packets that fit these buffers are copied into these buffers entirely. Currently this buffer size of fixed at 128 bytes. This patch extends transmit data ring implementation to allow variable length transmit data ring buffers. The length of the buffer is read from the emulation during initialization. Signed-off-by: Shrikrishna Khare <skhare@vmware.com> Acked-by: Yong Wang <yongwang@vmware.com> Acked-by: Jin Heo <heoj@vmware.com>
2017-03-06 22:55:03 +00:00
/*
* Gets tx data ring descriptor size.
*/
static uint16_t
eth_vmxnet3_txdata_get(struct vmxnet3_hw *hw)
{
uint16 txdata_desc_size;
VMXNET3_WRITE_BAR1_REG(hw, VMXNET3_REG_CMD,
VMXNET3_CMD_GET_TXDATA_DESC_SIZE);
txdata_desc_size = VMXNET3_READ_BAR1_REG(hw, VMXNET3_REG_CMD);
return (txdata_desc_size < VMXNET3_TXDATA_DESC_MIN_SIZE ||
txdata_desc_size > VMXNET3_TXDATA_DESC_MAX_SIZE ||
txdata_desc_size & VMXNET3_TXDATA_DESC_SIZE_MASK) ?
sizeof(struct Vmxnet3_TxDataDesc) : txdata_desc_size;
}
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
/*
* It returns 0 on success.
*/
static int
ethdev: remove useless parameter in init functions The pointer to struct eth_driver is not used and is already set in struct rte_eth_dev. It's a small cleanup in PMD API which probably needs more attention to make clear what is a driver, a PCI driver, an ethernet driver, etc. Signed-off-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2015-03-05 18:24:59 +00:00
eth_vmxnet3_dev_init(struct rte_eth_dev *eth_dev)
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
{
struct rte_pci_device *pci_dev;
vmxnet3: remove useless adapter wrapper The adapter struct is just a wrapper around the vmxnet3_hw structure. Eliminate the wrapper and get rid of the macro used to access and needlessly cast the private data. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:39:58 +00:00
struct vmxnet3_hw *hw = eth_dev->data->dev_private;
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
uint32_t mac_hi, mac_lo, ver;
net/vmxnet3: keep consistent link status Bonding may examine the link properties to ensure that matching interfaces are bound together. If the link is going to have fixed properties, these need to remain consistent regardless of the link_status or the state of the adapter. Signed-off-by: Chas Williams <chas3@att.com> Acked-by: Shrikrishna Khare <skhare@vmware.com>
2018-01-17 15:04:57 +00:00
struct rte_eth_link link;
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
PMD_INIT_FUNC_TRACE();
eth_dev->dev_ops = &vmxnet3_eth_dev_ops;
eth_dev->rx_pkt_burst = &vmxnet3_recv_pkts;
eth_dev->tx_pkt_burst = &vmxnet3_xmit_pkts;
net/vmxnet3: add Tx preparation Signed-off-by: Konstantin Ananyev <konstantin.ananyev@intel.com> Acked-by: Yong Wang <yongwang@vmware.com>
2016-12-23 18:40:52 +00:00
eth_dev->tx_pkt_prepare = vmxnet3_prep_pkts;
drivers/net: add generic ethdev macro to get PCI device Instead of many PMD define their own macro, define a generic one in ethdev and use that in PMDs. Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com> Acked-by: Allain Legacy <allain.legacy@windriver.com> Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com> Acked-by: Andrew Rybchenko <arybchenko@solarflare.com>
2017-05-15 10:24:03 +00:00
pci_dev = RTE_ETH_DEV_TO_PCI(eth_dev);
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
remove trailing whitespaces This commit removes trailing whitespace from lines in files. Almost all files are affected, as the BSD license copyright header had trailing whitespace on 4 lines in it [hence the number of files reporting 8 lines changed in the diffstat]. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com> Acked-by: Neil Horman <nhorman@tuxdriver.com> [Thomas: remove spaces before tabs in libs] [Thomas: remove more trailing spaces in non-C files] Signed-off-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-03 23:42:50 +00:00
/*
vmxnet3: fix spelling Fix spelling errors in strings and comments. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: John McNamara <john.mcnamara@intel.com> Acked-by: Yong Wang <yongwang@vmware.com>
2015-06-10 16:09:37 +00:00
* for secondary processes, we don't initialize any further as primary
remove trailing whitespaces This commit removes trailing whitespace from lines in files. Almost all files are affected, as the BSD license copyright header had trailing whitespace on 4 lines in it [hence the number of files reporting 8 lines changed in the diffstat]. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com> Acked-by: Neil Horman <nhorman@tuxdriver.com> [Thomas: remove spaces before tabs in libs] [Thomas: remove more trailing spaces in non-C files] Signed-off-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-03 23:42:50 +00:00
* has already done this work.
*/
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
if (rte_eal_process_type() != RTE_PROC_PRIMARY)
return 0;
drivers: copy PCI device info to ethdev data Use new function rte_eth_copy_pci_info. Copy device info for the following pdevs: bnx2x cxgbe e1000 enic fm10k i40e ixgbe mlx4 mlx5 virtio vmxnet3 Signed-off-by: Bernard Iremonger <bernard.iremonger@intel.com> Acked-by: Bruce Richardson <bruce.richardson@intel.com>
2015-11-03 13:01:56 +00:00
rte_eth_copy_pci_info(eth_dev, pci_dev);
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
/* Vendor and Device ID need to be set before init of shared code */
hw->device_id = pci_dev->id.device_id;
hw->vendor_id = pci_dev->id.vendor_id;
hw->hw_addr0 = (void *)pci_dev->mem_resource[0].addr;
hw->hw_addr1 = (void *)pci_dev->mem_resource[1].addr;
hw->num_rx_queues = 1;
hw->num_tx_queues = 1;
hw->bufs_per_pkt = 1;
/* Check h/w version compatibility with driver. */
vmxnet3: cleanup style and indentation This driver had several style problems, the worst of which was botched indentation. Fix almost all the problems reported by checkpatch. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:37:04 +00:00
ver = VMXNET3_READ_BAR1_REG(hw, VMXNET3_REG_VRRS);
vmxnet3: fix spelling Fix spelling errors in strings and comments. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: John McNamara <john.mcnamara@intel.com> Acked-by: Yong Wang <yongwang@vmware.com>
2015-06-10 16:09:37 +00:00
PMD_INIT_LOG(DEBUG, "Hardware version : %d", ver);
net/vmxnet3: prepare for version 3 changes Cleanup some code in preparation of vmxnet3 version 3 changes. Signed-off-by: Shrikrishna Khare <skhare@vmware.com> Acked-by: Yong Wang <yongwang@vmware.com> Acked-by: Jin Heo <heoj@vmware.com>
2017-03-06 22:55:01 +00:00
net/vmxnet3: add v4 boot and guest UDP RSS config This patch introduces: - VMxnet3 v4 negotiation and, - entirely guest-driven UDP RSS support. VMxnet3 v3 already has UDP RSS support, however it depends on hypervisor provisioning on the VM through ESX specific flags, which are not transparent or known to the guest later on. Vmxnet3 v4 introduces a new API transaction which allows configuring RSS entirely from the guest. This API must be invoked after device shared mem region is initialized. IPv4 ESP RSS (SPI based) is also available, but currently there are no ESP RSS definitions on rte_eth layer to handle that. Signed-off-by: Eduard Serra <eserra@vmware.com> Acked-by: Yong Wang <yongwang@vmware.com>
2019-04-18 20:59:37 +00:00
if (ver & (1 << VMXNET3_REV_4)) {
VMXNET3_WRITE_BAR1_REG(hw, VMXNET3_REG_VRRS,
1 << VMXNET3_REV_4);
hw->version = VMXNET3_REV_4 + 1;
} else if (ver & (1 << VMXNET3_REV_3)) {
net/vmxnet3: update to version 3 With all vmxnet3 version 3 changes incorporated in the vmxnet3 driver, the driver can configure emulation to run at vmxnet3 version 3, provided the emulation advertises support for version 3. This patch also updates release notes. Signed-off-by: Shrikrishna Khare <skhare@vmware.com> Acked-by: Yong Wang <yongwang@vmware.com> Acked-by: Jin Heo <heoj@vmware.com>
2017-03-06 22:55:07 +00:00
VMXNET3_WRITE_BAR1_REG(hw, VMXNET3_REG_VRRS,
1 << VMXNET3_REV_3);
hw->version = VMXNET3_REV_3 + 1;
} else if (ver & (1 << VMXNET3_REV_2)) {
net/vmxnet3: prepare for version 3 changes Cleanup some code in preparation of vmxnet3 version 3 changes. Signed-off-by: Shrikrishna Khare <skhare@vmware.com> Acked-by: Yong Wang <yongwang@vmware.com> Acked-by: Jin Heo <heoj@vmware.com>
2017-03-06 22:55:01 +00:00
VMXNET3_WRITE_BAR1_REG(hw, VMXNET3_REG_VRRS,
1 << VMXNET3_REV_2);
hw->version = VMXNET3_REV_2 + 1;
} else if (ver & (1 << VMXNET3_REV_1)) {
VMXNET3_WRITE_BAR1_REG(hw, VMXNET3_REG_VRRS,
1 << VMXNET3_REV_1);
hw->version = VMXNET3_REV_1 + 1;
} else {
PMD_INIT_LOG(ERR, "Incompatible hardware version: %d", ver);
remove trailing whitespaces This commit removes trailing whitespace from lines in files. Almost all files are affected, as the BSD license copyright header had trailing whitespace on 4 lines in it [hence the number of files reporting 8 lines changed in the diffstat]. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com> Acked-by: Neil Horman <nhorman@tuxdriver.com> [Thomas: remove spaces before tabs in libs] [Thomas: remove more trailing spaces in non-C files] Signed-off-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-03 23:42:50 +00:00
return -EIO;
vmxnet3: cleanup style and indentation This driver had several style problems, the worst of which was botched indentation. Fix almost all the problems reported by checkpatch. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:37:04 +00:00
}
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
net/vmxnet3: prepare for version 3 changes Cleanup some code in preparation of vmxnet3 version 3 changes. Signed-off-by: Shrikrishna Khare <skhare@vmware.com> Acked-by: Yong Wang <yongwang@vmware.com> Acked-by: Jin Heo <heoj@vmware.com>
2017-03-06 22:55:01 +00:00
PMD_INIT_LOG(DEBUG, "Using device version %d\n", hw->version);
vmxnet3: cleanup style and indentation This driver had several style problems, the worst of which was botched indentation. Fix almost all the problems reported by checkpatch. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:37:04 +00:00
/* Check UPT version compatibility with driver. */
ver = VMXNET3_READ_BAR1_REG(hw, VMXNET3_REG_UVRS);
vmxnet3: fix spelling Fix spelling errors in strings and comments. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: John McNamara <john.mcnamara@intel.com> Acked-by: Yong Wang <yongwang@vmware.com>
2015-06-10 16:09:37 +00:00
PMD_INIT_LOG(DEBUG, "UPT hardware version : %d", ver);
vmxnet3: cleanup style and indentation This driver had several style problems, the worst of which was botched indentation. Fix almost all the problems reported by checkpatch. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:37:04 +00:00
if (ver & 0x1)
remove trailing whitespaces This commit removes trailing whitespace from lines in files. Almost all files are affected, as the BSD license copyright header had trailing whitespace on 4 lines in it [hence the number of files reporting 8 lines changed in the diffstat]. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com> Acked-by: Neil Horman <nhorman@tuxdriver.com> [Thomas: remove spaces before tabs in libs] [Thomas: remove more trailing spaces in non-C files] Signed-off-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-03 23:42:50 +00:00
VMXNET3_WRITE_BAR1_REG(hw, VMXNET3_REG_UVRS, 1);
vmxnet3: cleanup style and indentation This driver had several style problems, the worst of which was botched indentation. Fix almost all the problems reported by checkpatch. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:37:04 +00:00
else {
vmxnet3: fix spelling Fix spelling errors in strings and comments. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: John McNamara <john.mcnamara@intel.com> Acked-by: Yong Wang <yongwang@vmware.com>
2015-06-10 16:09:37 +00:00
PMD_INIT_LOG(ERR, "Incompatible UPT version.");
remove trailing whitespaces This commit removes trailing whitespace from lines in files. Almost all files are affected, as the BSD license copyright header had trailing whitespace on 4 lines in it [hence the number of files reporting 8 lines changed in the diffstat]. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com> Acked-by: Neil Horman <nhorman@tuxdriver.com> [Thomas: remove spaces before tabs in libs] [Thomas: remove more trailing spaces in non-C files] Signed-off-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-03 23:42:50 +00:00
return -EIO;
vmxnet3: cleanup style and indentation This driver had several style problems, the worst of which was botched indentation. Fix almost all the problems reported by checkpatch. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:37:04 +00:00
}
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
/* Getting MAC Address */
mac_lo = VMXNET3_READ_BAR1_REG(hw, VMXNET3_REG_MACL);
mac_hi = VMXNET3_READ_BAR1_REG(hw, VMXNET3_REG_MACH);
net/vmxnet3: coding style changes Signed-off-by: Yong Wang <yongwang@vmware.com> Acked-by: Stephen Hemminger <stephen@networkplumber.org>
2016-08-29 19:18:46 +00:00
memcpy(hw->perm_addr, &mac_lo, 4);
memcpy(hw->perm_addr + 4, &mac_hi, 2);
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
/* Allocate memory for storing MAC addresses */
net: add rte prefix to ether defines Add 'RTE_' prefix to defines: - rename ETHER_ADDR_LEN as RTE_ETHER_ADDR_LEN. - rename ETHER_TYPE_LEN as RTE_ETHER_TYPE_LEN. - rename ETHER_CRC_LEN as RTE_ETHER_CRC_LEN. - rename ETHER_HDR_LEN as RTE_ETHER_HDR_LEN. - rename ETHER_MIN_LEN as RTE_ETHER_MIN_LEN. - rename ETHER_MAX_LEN as RTE_ETHER_MAX_LEN. - rename ETHER_MTU as RTE_ETHER_MTU. - rename ETHER_MAX_VLAN_FRAME_LEN as RTE_ETHER_MAX_VLAN_FRAME_LEN. - rename ETHER_MAX_VLAN_ID as RTE_ETHER_MAX_VLAN_ID. - rename ETHER_MAX_JUMBO_FRAME_LEN as RTE_ETHER_MAX_JUMBO_FRAME_LEN. - rename ETHER_MIN_MTU as RTE_ETHER_MIN_MTU. - rename ETHER_LOCAL_ADMIN_ADDR as RTE_ETHER_LOCAL_ADMIN_ADDR. - rename ETHER_GROUP_ADDR as RTE_ETHER_GROUP_ADDR. - rename ETHER_TYPE_IPv4 as RTE_ETHER_TYPE_IPv4. - rename ETHER_TYPE_IPv6 as RTE_ETHER_TYPE_IPv6. - rename ETHER_TYPE_ARP as RTE_ETHER_TYPE_ARP. - rename ETHER_TYPE_VLAN as RTE_ETHER_TYPE_VLAN. - rename ETHER_TYPE_RARP as RTE_ETHER_TYPE_RARP. - rename ETHER_TYPE_QINQ as RTE_ETHER_TYPE_QINQ. - rename ETHER_TYPE_ETAG as RTE_ETHER_TYPE_ETAG. - rename ETHER_TYPE_1588 as RTE_ETHER_TYPE_1588. - rename ETHER_TYPE_SLOW as RTE_ETHER_TYPE_SLOW. - rename ETHER_TYPE_TEB as RTE_ETHER_TYPE_TEB. - rename ETHER_TYPE_LLDP as RTE_ETHER_TYPE_LLDP. - rename ETHER_TYPE_MPLS as RTE_ETHER_TYPE_MPLS. - rename ETHER_TYPE_MPLSM as RTE_ETHER_TYPE_MPLSM. - rename ETHER_VXLAN_HLEN as RTE_ETHER_VXLAN_HLEN. - rename ETHER_ADDR_FMT_SIZE as RTE_ETHER_ADDR_FMT_SIZE. - rename VXLAN_GPE_TYPE_IPV4 as RTE_VXLAN_GPE_TYPE_IPV4. - rename VXLAN_GPE_TYPE_IPV6 as RTE_VXLAN_GPE_TYPE_IPV6. - rename VXLAN_GPE_TYPE_ETH as RTE_VXLAN_GPE_TYPE_ETH. - rename VXLAN_GPE_TYPE_NSH as RTE_VXLAN_GPE_TYPE_NSH. - rename VXLAN_GPE_TYPE_MPLS as RTE_VXLAN_GPE_TYPE_MPLS. - rename VXLAN_GPE_TYPE_GBP as RTE_VXLAN_GPE_TYPE_GBP. - rename VXLAN_GPE_TYPE_VBNG as RTE_VXLAN_GPE_TYPE_VBNG. - rename ETHER_VXLAN_GPE_HLEN as RTE_ETHER_VXLAN_GPE_HLEN. Do not update the command line library to avoid adding a dependency to librte_net. Signed-off-by: Olivier Matz <olivier.matz@6wind.com> Reviewed-by: Stephen Hemminger <stephen@networkplumber.org> Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2019-05-21 16:13:05 +00:00
eth_dev->data->mac_addrs = rte_zmalloc("vmxnet3", RTE_ETHER_ADDR_LEN *
vmxnet3: cleanup style and indentation This driver had several style problems, the worst of which was botched indentation. Fix almost all the problems reported by checkpatch. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:37:04 +00:00
VMXNET3_MAX_MAC_ADDRS, 0);
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
if (eth_dev->data->mac_addrs == NULL) {
PMD_INIT_LOG(ERR,
vmxnet3: cleanup style and indentation This driver had several style problems, the worst of which was botched indentation. Fix almost all the problems reported by checkpatch. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:37:04 +00:00
"Failed to allocate %d bytes needed to store MAC addresses",
net: add rte prefix to ether defines Add 'RTE_' prefix to defines: - rename ETHER_ADDR_LEN as RTE_ETHER_ADDR_LEN. - rename ETHER_TYPE_LEN as RTE_ETHER_TYPE_LEN. - rename ETHER_CRC_LEN as RTE_ETHER_CRC_LEN. - rename ETHER_HDR_LEN as RTE_ETHER_HDR_LEN. - rename ETHER_MIN_LEN as RTE_ETHER_MIN_LEN. - rename ETHER_MAX_LEN as RTE_ETHER_MAX_LEN. - rename ETHER_MTU as RTE_ETHER_MTU. - rename ETHER_MAX_VLAN_FRAME_LEN as RTE_ETHER_MAX_VLAN_FRAME_LEN. - rename ETHER_MAX_VLAN_ID as RTE_ETHER_MAX_VLAN_ID. - rename ETHER_MAX_JUMBO_FRAME_LEN as RTE_ETHER_MAX_JUMBO_FRAME_LEN. - rename ETHER_MIN_MTU as RTE_ETHER_MIN_MTU. - rename ETHER_LOCAL_ADMIN_ADDR as RTE_ETHER_LOCAL_ADMIN_ADDR. - rename ETHER_GROUP_ADDR as RTE_ETHER_GROUP_ADDR. - rename ETHER_TYPE_IPv4 as RTE_ETHER_TYPE_IPv4. - rename ETHER_TYPE_IPv6 as RTE_ETHER_TYPE_IPv6. - rename ETHER_TYPE_ARP as RTE_ETHER_TYPE_ARP. - rename ETHER_TYPE_VLAN as RTE_ETHER_TYPE_VLAN. - rename ETHER_TYPE_RARP as RTE_ETHER_TYPE_RARP. - rename ETHER_TYPE_QINQ as RTE_ETHER_TYPE_QINQ. - rename ETHER_TYPE_ETAG as RTE_ETHER_TYPE_ETAG. - rename ETHER_TYPE_1588 as RTE_ETHER_TYPE_1588. - rename ETHER_TYPE_SLOW as RTE_ETHER_TYPE_SLOW. - rename ETHER_TYPE_TEB as RTE_ETHER_TYPE_TEB. - rename ETHER_TYPE_LLDP as RTE_ETHER_TYPE_LLDP. - rename ETHER_TYPE_MPLS as RTE_ETHER_TYPE_MPLS. - rename ETHER_TYPE_MPLSM as RTE_ETHER_TYPE_MPLSM. - rename ETHER_VXLAN_HLEN as RTE_ETHER_VXLAN_HLEN. - rename ETHER_ADDR_FMT_SIZE as RTE_ETHER_ADDR_FMT_SIZE. - rename VXLAN_GPE_TYPE_IPV4 as RTE_VXLAN_GPE_TYPE_IPV4. - rename VXLAN_GPE_TYPE_IPV6 as RTE_VXLAN_GPE_TYPE_IPV6. - rename VXLAN_GPE_TYPE_ETH as RTE_VXLAN_GPE_TYPE_ETH. - rename VXLAN_GPE_TYPE_NSH as RTE_VXLAN_GPE_TYPE_NSH. - rename VXLAN_GPE_TYPE_MPLS as RTE_VXLAN_GPE_TYPE_MPLS. - rename VXLAN_GPE_TYPE_GBP as RTE_VXLAN_GPE_TYPE_GBP. - rename VXLAN_GPE_TYPE_VBNG as RTE_VXLAN_GPE_TYPE_VBNG. - rename ETHER_VXLAN_GPE_HLEN as RTE_ETHER_VXLAN_GPE_HLEN. Do not update the command line library to avoid adding a dependency to librte_net. Signed-off-by: Olivier Matz <olivier.matz@6wind.com> Reviewed-by: Stephen Hemminger <stephen@networkplumber.org> Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2019-05-21 16:13:05 +00:00
RTE_ETHER_ADDR_LEN * VMXNET3_MAX_MAC_ADDRS);
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
return -ENOMEM;
}
/* Copy the permanent MAC address */
net: add rte prefix to ether functions Add 'rte_' prefix to functions: - rename is_same_ether_addr() as rte_is_same_ether_addr(). - rename is_zero_ether_addr() as rte_is_zero_ether_addr(). - rename is_unicast_ether_addr() as rte_is_unicast_ether_addr(). - rename is_multicast_ether_addr() as rte_is_multicast_ether_addr(). - rename is_broadcast_ether_addr() as rte_is_broadcast_ether_addr(). - rename is_universal_ether_addr() as rte_is_universal_ether_addr(). - rename is_local_admin_ether_addr() as rte_is_local_admin_ether_addr(). - rename is_valid_assigned_ether_addr() as rte_is_valid_assigned_ether_addr(). - rename eth_random_addr() as rte_eth_random_addr(). - rename ether_addr_copy() as rte_ether_addr_copy(). - rename ether_format_addr() as rte_ether_format_addr(). Signed-off-by: Olivier Matz <olivier.matz@6wind.com> Reviewed-by: Stephen Hemminger <stephen@networkplumber.org> Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2019-05-21 16:13:04 +00:00
rte_ether_addr_copy((struct rte_ether_addr *)hw->perm_addr,
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
&eth_dev->data->mac_addrs[0]);
vmxnet3: fix double spacing of log messages The debug log macro's already include newline, no need to double space the output. Note: other drivers have the same problem Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:37:13 +00:00
PMD_INIT_LOG(DEBUG, "MAC Address : %02x:%02x:%02x:%02x:%02x:%02x",
vmxnet3: cleanup style and indentation This driver had several style problems, the worst of which was botched indentation. Fix almost all the problems reported by checkpatch. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:37:04 +00:00
hw->perm_addr[0], hw->perm_addr[1], hw->perm_addr[2],
hw->perm_addr[3], hw->perm_addr[4], hw->perm_addr[5]);
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
drivers/net: set close behaviour flag at probing The ethdev flag RTE_ETH_DEV_CLOSE_REMOVE is set for drivers having migrated to the new behaviour of rte_eth_dev_close(). As any other flag, it can be useful to know about its value as soon as the port is probed. Unfortunately, it was set inside the close operation, just before being erased by memset() in rte_eth_dev_release_port(). The flag assignment is moved to the probing stage, so it can be checked by the application in order to anticipate the behaviour. Fixes: 42603bbdb58e ("net/mlx5: release port on close") Fixes: 6c99085d972b ("net/vmxnet3: fix hot-unplug") Fixes: 4d7877fde2ef ("net/ena: remove resources when port is being closed") Signed-off-by: Thomas Monjalon <thomas@monjalon.net> Acked-by: Luca Boccassi <bluca@debian.org> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2018-11-11 22:46:11 +00:00
/* Flag to call rte_eth_dev_release_port() in rte_eth_dev_close(). */
eth_dev->data->dev_flags |= RTE_ETH_DEV_CLOSE_REMOVE;
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
/* Put device in Quiesce Mode */
VMXNET3_WRITE_BAR1_REG(hw, VMXNET3_REG_CMD, VMXNET3_CMD_QUIESCE_DEV);
vmxnet3: fix VLAN filtering During an MTU change, the adapter is restarted. If hardware VLAN offload is in use, this existing filter table would also be cleared. Instead, setup the shadow table once during device initialization and just update during restart. vmxnet3_dev_vlan_offload_set(dev, mask) was incorrectly treating the mask parameter as the bitmask for vlan_strip and vlan_filter, whereas the mask indicates only what has changed - the values for vlan_stripping and vlan_filter needs to be taken from dev_conf.rxmode. Fixes: f003fc383487 ("vmxnet3: enable vlan filtering") Signed-off-by: Charles (Chas) Williams <ciwillia@brocade.com> Signed-off-by: Nachiketa Prachanda <nprachan@brocade.com> Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Remy Horton <remy.horton@intel.com>
2016-03-04 18:08:01 +00:00
/* allow untagged pkts */
VMXNET3_SET_VFTABLE_ENTRY(hw->shadow_vfta, 0);
net/vmxnet3: allow variable length Tx data ring vmxnet3 driver supports transmit data ring viz. a set of fixed size buffers used by the driver to copy packet headers. Small packets that fit these buffers are copied into these buffers entirely. Currently this buffer size of fixed at 128 bytes. This patch extends transmit data ring implementation to allow variable length transmit data ring buffers. The length of the buffer is read from the emulation during initialization. Signed-off-by: Shrikrishna Khare <skhare@vmware.com> Acked-by: Yong Wang <yongwang@vmware.com> Acked-by: Jin Heo <heoj@vmware.com>
2017-03-06 22:55:03 +00:00
hw->txdata_desc_size = VMXNET3_VERSION_GE_3(hw) ?
eth_vmxnet3_txdata_get(hw) : sizeof(struct Vmxnet3_TxDataDesc);
net/vmxnet3: support receive data ring vmxnet3 driver preallocates buffers for receiving packets and posts the buffers to the emulation. In order to deliver a received packet to the guest, the emulation must map buffer(s) and copy the packet into it. To avoid this memory mapping overhead, this patch introduces the receive data ring - a set of small sized buffers that are always mapped by the emulation. If a packet fits into the receive data ring buffer, the emulation delivers the packet via the receive data ring (which must be copied by the guest driver), or else the usual receive path is used. Signed-off-by: Shrikrishna Khare <skhare@vmware.com> Acked-by: Yong Wang <yongwang@vmware.com> Acked-by: Jin Heo <heoj@vmware.com>
2017-03-06 22:55:04 +00:00
hw->rxdata_desc_size = VMXNET3_VERSION_GE_3(hw) ?
VMXNET3_DEF_RXDATA_DESC_SIZE : 0;
RTE_ASSERT((hw->rxdata_desc_size & ~VMXNET3_RXDATA_DESC_SIZE_MASK) ==
hw->rxdata_desc_size);
net/vmxnet3: retain counters on restart Most NICs like virtio, igb/ixgbe etc. don't reset counters on dev_start and arguably this helps in monitoring the counters across a longer time span with multiple device start/stops. vmxnet3 behavior is opposite to that and counters are reset by the host side implementation each time the device is restarted. Change the driver to save the counters in its private context before it is reset by writing CMD_ACTIVATE to REG_CMD. Signed-off-by: Nachiketa Prachanda <nprachan@brocade.com> Acked-by: Shrikrishna Khare <skhare@vmware.com>
2017-06-15 12:16:08 +00:00
/* clear shadow stats */
memset(hw->saved_tx_stats, 0, sizeof(hw->saved_tx_stats));
memset(hw->saved_rx_stats, 0, sizeof(hw->saved_rx_stats));
net/vmxnet3: support stats reset 'stats_reset()' callback was missing because the device backend doesn't support it. This commit adds a workaround to this and implements the callback by taking a snapshot of the stats (SNAPSHOT) each time 'stats_reset()' is called. When getting stats with 'stats_get()', hw stats which always increase reduce SNAPSHOT stats. That's how we get the "real" stats since the last 'stats_reset()'. Signed-off-by: Yogev Chaimovich <yogev@cgstowernetworks.com> Acked-by: Yong Wang <yongwang@vmware.com>
2018-09-20 06:37:05 +00:00
/* clear snapshot stats */
memset(hw->snapshot_tx_stats, 0, sizeof(hw->snapshot_tx_stats));
memset(hw->snapshot_rx_stats, 0, sizeof(hw->snapshot_rx_stats));
net/vmxnet3: keep consistent link status Bonding may examine the link properties to ensure that matching interfaces are bound together. If the link is going to have fixed properties, these need to remain consistent regardless of the link_status or the state of the adapter. Signed-off-by: Chas Williams <chas3@att.com> Acked-by: Shrikrishna Khare <skhare@vmware.com>
2018-01-17 15:04:57 +00:00
/* set the initial link status */
memset(&link, 0, sizeof(link));
link.link_duplex = ETH_LINK_FULL_DUPLEX;
link.link_speed = ETH_SPEED_NUM_10G;
link.link_autoneg = ETH_LINK_FIXED;
net/vmxnet3: use link status helper functions Use new rte_eth_linkstatus_get/set helper functions to handle link status update. Also remove no longer necessary include of rte_atomic.h Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2018-01-26 02:01:40 +00:00
rte_eth_linkstatus_set(eth_dev, &link);
net/vmxnet3: keep consistent link status Bonding may examine the link properties to ensure that matching interfaces are bound together. If the link is going to have fixed properties, these need to remain consistent regardless of the link_status or the state of the adapter. Signed-off-by: Chas Williams <chas3@att.com> Acked-by: Shrikrishna Khare <skhare@vmware.com>
2018-01-17 15:04:57 +00:00
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
return 0;
}
vmxnet3: support port hotplug Signed-off-by: Bernard Iremonger <bernard.iremonger@intel.com> Acked-by: Yong Wang <yongwang@vmware.com>
2015-10-27 17:12:26 +00:00
static int
eth_vmxnet3_dev_uninit(struct rte_eth_dev *eth_dev)
{
struct vmxnet3_hw *hw = eth_dev->data->dev_private;
PMD_INIT_FUNC_TRACE();
if (rte_eal_process_type() != RTE_PROC_PRIMARY)
return 0;
net/vmxnet3: fix hot-unplug The vmxnet3 driver can't call back into dev_close(), and possibly dev_stop(), in dev_uninit(). When dev_uninit() is called, anything that those routines would want to clean up has already been released. Further, for complete cleanup, it is necessary to release any of the queue resources during dev_close(). This allows a vmxnet3 device to be hot-unplugged without leaking queues. Also set RTE_ETH_DEV_CLOSE_REMOVE on close so that the port resources can be deallocated. Return EBUSY if remove is called before stop. Fixes: dfaff37fc46d ("vmxnet3: import new vmxnet3 poll mode driver implementation") Cc: stable@dpdk.org Signed-off-by: Brian Russell <brussell@brocade.com> Signed-off-by: Luca Boccassi <bluca@debian.org>
2018-10-31 18:39:44 +00:00
if (hw->adapter_stopped == 0) {
PMD_INIT_LOG(DEBUG, "Device has not been closed.");
return -EBUSY;
}
vmxnet3: support port hotplug Signed-off-by: Bernard Iremonger <bernard.iremonger@intel.com> Acked-by: Yong Wang <yongwang@vmware.com>
2015-10-27 17:12:26 +00:00
eth_dev->dev_ops = NULL;
eth_dev->rx_pkt_burst = NULL;
eth_dev->tx_pkt_burst = NULL;
net/vmxnet3: add Tx preparation Signed-off-by: Konstantin Ananyev <konstantin.ananyev@intel.com> Acked-by: Yong Wang <yongwang@vmware.com>
2016-12-23 18:40:52 +00:00
eth_dev->tx_pkt_prepare = NULL;
vmxnet3: support port hotplug Signed-off-by: Bernard Iremonger <bernard.iremonger@intel.com> Acked-by: Yong Wang <yongwang@vmware.com>
2015-10-27 17:12:26 +00:00
return 0;
}
drivers/net: do not use ethdev driver Signed-off-by: Jan Blunck <jblunck@infradead.org> Signed-off-by: Gaetan Rivet <gaetan.rivet@6wind.com> Acked-by: Stephen Hemminger <stephen@networkplumber.org>
2017-04-11 15:44:24 +00:00
static int eth_vmxnet3_pci_probe(struct rte_pci_driver *pci_drv __rte_unused,
struct rte_pci_device *pci_dev)
{
return rte_eth_dev_pci_generic_probe(pci_dev,
sizeof(struct vmxnet3_hw), eth_vmxnet3_dev_init);
}
static int eth_vmxnet3_pci_remove(struct rte_pci_device *pci_dev)
{
return rte_eth_dev_pci_generic_remove(pci_dev, eth_vmxnet3_dev_uninit);
}
static struct rte_pci_driver rte_vmxnet3_pmd = {
.id_table = pci_id_vmxnet3_map,
net/vmxnet3: generate link-state change notifications Generate link-state change notifications by listening to interrupts generated by the device. Make use of the existing vmxnet3_process_events function that was compiled out, but change it to call vmxnet3_dev_link_update on a VMXNET3_ECR_LINK event and to not be so noisy in its log messages. Enable interrupts on starting the device, using a new helper function, vmxnet3_enable_intr, based on vmxnet3_disable_intr and validated against the FreeBSD driver. Keep track of the number of interrupts registered for to avoid hardcoding these in vmxnet3_enable/disable_intr and to provision for any future rxq intr support. Factor out the guts of vmxnet3_dev_link_update minus the started check to allow the new function to be called from vmxnet3_dev_start in the lsc-enabled case to ensure that the link state is correctly set from the actual state at that point. Signed-off-by: Robert Shearman <rshearma@brocade.com>
2017-06-15 12:16:10 +00:00
.drv_flags = RTE_PCI_DRV_NEED_MAPPING | RTE_PCI_DRV_INTR_LSC,
drivers/net: do not use ethdev driver Signed-off-by: Jan Blunck <jblunck@infradead.org> Signed-off-by: Gaetan Rivet <gaetan.rivet@6wind.com> Acked-by: Stephen Hemminger <stephen@networkplumber.org>
2017-04-11 15:44:24 +00:00
.probe = eth_vmxnet3_pci_probe,
.remove = eth_vmxnet3_pci_remove,
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
};
static int
vmxnet3_dev_configure(struct rte_eth_dev *dev)
{
const struct rte_memzone *mz;
vmxnet3: remove useless adapter wrapper The adapter struct is just a wrapper around the vmxnet3_hw structure. Eliminate the wrapper and get rid of the macro used to access and needlessly cast the private data. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:39:58 +00:00
struct vmxnet3_hw *hw = dev->data->dev_private;
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
size_t size;
PMD_INIT_FUNC_TRACE();
drivers/net: fix RSS hash offload flag if no RSS By default RSS hash delivery (offload) is bound to RSS mode and it is incorrect to advertise it as enabled if Rx multi-queue mode has no RSS. Fixes: 8b945a7f7dcb ("drivers/net: update Rx RSS hash offload capabilities") Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2019-11-14 16:40:50 +00:00
if (dev->data->dev_conf.rxmode.mq_mode & ETH_MQ_RX_RSS_FLAG)
dev->data->dev_conf.rxmode.offloads |= DEV_RX_OFFLOAD_RSS_HASH;
drivers/net: update Rx RSS hash offload capabilities Add DEV_RX_OFFLOAD_RSS_HASH flag for all PMDs that support RSS hash delivery. Signed-off-by: Pavan Nikhilesh <pbhagavatula@marvell.com> Reviewed-by: Andrew Rybchenko <arybchenko@solarflare.com> Reviewed-by: Hemant Agrawal <hemant.agrawal@nxp.com> Acked-by: Jerin Jacob <jerinj@marvell.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2019-11-11 13:19:08 +00:00
net/vmxnet3: improve error checks and return values Signed-off-by: Yong Wang <yongwang@vmware.com> Reviewed-by: Juho Snellman <jsnell@iki.fi> Acked-by: Stephen Hemminger <stephen@networkplumber.org>
2016-08-29 19:18:45 +00:00
if (dev->data->nb_tx_queues > VMXNET3_MAX_TX_QUEUES ||
dev->data->nb_rx_queues > VMXNET3_MAX_RX_QUEUES) {
PMD_INIT_LOG(ERR, "ERROR: Number of queues not supported");
vmxnet3: cleanup style and indentation This driver had several style problems, the worst of which was botched indentation. Fix almost all the problems reported by checkpatch. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:37:04 +00:00
return -EINVAL;
net/vmxnet3: improve error checks and return values Signed-off-by: Yong Wang <yongwang@vmware.com> Reviewed-by: Juho Snellman <jsnell@iki.fi> Acked-by: Stephen Hemminger <stephen@networkplumber.org>
2016-08-29 19:18:45 +00:00
}
if (!rte_is_power_of_2(dev->data->nb_rx_queues)) {
PMD_INIT_LOG(ERR, "ERROR: Number of rx queues not power of 2");
return -EINVAL;
}
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
size = dev->data->nb_rx_queues * sizeof(struct Vmxnet3_TxQueueDesc) +
vmxnet3: cleanup style and indentation This driver had several style problems, the worst of which was botched indentation. Fix almost all the problems reported by checkpatch. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:37:04 +00:00
dev->data->nb_tx_queues * sizeof(struct Vmxnet3_RxQueueDesc);
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
if (size > UINT16_MAX)
vmxnet3: cleanup style and indentation This driver had several style problems, the worst of which was botched indentation. Fix almost all the problems reported by checkpatch. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:37:04 +00:00
return -EINVAL;
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
hw->num_rx_queues = (uint8_t)dev->data->nb_rx_queues;
hw->num_tx_queues = (uint8_t)dev->data->nb_tx_queues;
remove trailing whitespaces This commit removes trailing whitespace from lines in files. Almost all files are affected, as the BSD license copyright header had trailing whitespace on 4 lines in it [hence the number of files reporting 8 lines changed in the diffstat]. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com> Acked-by: Neil Horman <nhorman@tuxdriver.com> [Thomas: remove spaces before tabs in libs] [Thomas: remove more trailing spaces in non-C files] Signed-off-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-03 23:42:50 +00:00
/*
* Allocate a memzone for Vmxnet3_DriverShared - Vmxnet3_DSDevRead
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
* on current socket
*/
vmxnet3: cleanup style and indentation This driver had several style problems, the worst of which was botched indentation. Fix almost all the problems reported by checkpatch. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:37:04 +00:00
mz = gpa_zone_reserve(dev, sizeof(struct Vmxnet3_DriverShared),
net/vmxnet3: reallocate shared memzone on re-config When adding a DPDK port to ovs-vswitchd with DPDK, the vmxnet3 device fails to activate due to mismatched magic number. This failure causes following operations to run: start the port, stop the port, reconfigure and re-start the port. During reconfigure, if there is an existing memzone, driver will reuse it. But reconfigure may request different number of Tx/Rx queues. This results in a memzone with wrong size and potential invalid memory access. To fix this, free the memzone if found and reserve a new one. Signed-off-by: Yong Wang <yongwang@vmware.com> Reviewed-by: Guolin Yang <gyang@vmware.com> Reviewed-by: Daniele Di Proietto <ddiproietto@vmware.com> Tested-by: Daniele Di Proietto <ddiproietto@vmware.com> Acked-by: Stephen Hemminger <stephen@networkplumber.org>
2016-08-29 19:18:47 +00:00
"shared", rte_socket_id(), 8, 1);
remove trailing whitespaces This commit removes trailing whitespace from lines in files. Almost all files are affected, as the BSD license copyright header had trailing whitespace on 4 lines in it [hence the number of files reporting 8 lines changed in the diffstat]. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com> Acked-by: Neil Horman <nhorman@tuxdriver.com> [Thomas: remove spaces before tabs in libs] [Thomas: remove more trailing spaces in non-C files] Signed-off-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-03 23:42:50 +00:00
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
if (mz == NULL) {
vmxnet3: fix double spacing of log messages The debug log macro's already include newline, no need to double space the output. Note: other drivers have the same problem Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:37:13 +00:00
PMD_INIT_LOG(ERR, "ERROR: Creating shared zone");
vmxnet3: cleanup style and indentation This driver had several style problems, the worst of which was botched indentation. Fix almost all the problems reported by checkpatch. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:37:04 +00:00
return -ENOMEM;
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
}
memset(mz->addr, 0, mz->len);
hw->shared = mz->addr;
memzone: rename address from physical to IOVA The struct rte_memzone field .phys_addr is renamed to .iova. The deprecated name is kept in an anonymous union to avoid breaking the API. Signed-off-by: Thomas Monjalon <thomas@monjalon.net> Acked-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
2017-11-04 01:22:28 +00:00
hw->sharedPA = mz->iova;
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
remove trailing whitespaces This commit removes trailing whitespace from lines in files. Almost all files are affected, as the BSD license copyright header had trailing whitespace on 4 lines in it [hence the number of files reporting 8 lines changed in the diffstat]. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com> Acked-by: Neil Horman <nhorman@tuxdriver.com> [Thomas: remove spaces before tabs in libs] [Thomas: remove more trailing spaces in non-C files] Signed-off-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-03 23:42:50 +00:00
/*
vmxnet3: cleanup style and indentation This driver had several style problems, the worst of which was botched indentation. Fix almost all the problems reported by checkpatch. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:37:04 +00:00
* Allocate a memzone for Vmxnet3_RxQueueDesc - Vmxnet3_TxQueueDesc
net/vmxnet3: reallocate shared memzone on re-config When adding a DPDK port to ovs-vswitchd with DPDK, the vmxnet3 device fails to activate due to mismatched magic number. This failure causes following operations to run: start the port, stop the port, reconfigure and re-start the port. During reconfigure, if there is an existing memzone, driver will reuse it. But reconfigure may request different number of Tx/Rx queues. This results in a memzone with wrong size and potential invalid memory access. To fix this, free the memzone if found and reserve a new one. Signed-off-by: Yong Wang <yongwang@vmware.com> Reviewed-by: Guolin Yang <gyang@vmware.com> Reviewed-by: Daniele Di Proietto <ddiproietto@vmware.com> Tested-by: Daniele Di Proietto <ddiproietto@vmware.com> Acked-by: Stephen Hemminger <stephen@networkplumber.org>
2016-08-29 19:18:47 +00:00
* on current socket.
*
* We cannot reuse this memzone from previous allocation as its size
* depends on the number of tx and rx queues, which could be different
* from one config to another.
vmxnet3: cleanup style and indentation This driver had several style problems, the worst of which was botched indentation. Fix almost all the problems reported by checkpatch. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:37:04 +00:00
*/
net/vmxnet3: reallocate shared memzone on re-config When adding a DPDK port to ovs-vswitchd with DPDK, the vmxnet3 device fails to activate due to mismatched magic number. This failure causes following operations to run: start the port, stop the port, reconfigure and re-start the port. During reconfigure, if there is an existing memzone, driver will reuse it. But reconfigure may request different number of Tx/Rx queues. This results in a memzone with wrong size and potential invalid memory access. To fix this, free the memzone if found and reserve a new one. Signed-off-by: Yong Wang <yongwang@vmware.com> Reviewed-by: Guolin Yang <gyang@vmware.com> Reviewed-by: Daniele Di Proietto <ddiproietto@vmware.com> Tested-by: Daniele Di Proietto <ddiproietto@vmware.com> Acked-by: Stephen Hemminger <stephen@networkplumber.org>
2016-08-29 19:18:47 +00:00
mz = gpa_zone_reserve(dev, size, "queuedesc", rte_socket_id(),
VMXNET3_QUEUE_DESC_ALIGN, 0);
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
if (mz == NULL) {
vmxnet3: fix double spacing of log messages The debug log macro's already include newline, no need to double space the output. Note: other drivers have the same problem Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:37:13 +00:00
PMD_INIT_LOG(ERR, "ERROR: Creating queue descriptors zone");
vmxnet3: cleanup style and indentation This driver had several style problems, the worst of which was botched indentation. Fix almost all the problems reported by checkpatch. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:37:04 +00:00
return -ENOMEM;
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
}
memset(mz->addr, 0, mz->len);
hw->tqd_start = (Vmxnet3_TxQueueDesc *)mz->addr;
hw->rqd_start = (Vmxnet3_RxQueueDesc *)(hw->tqd_start + hw->num_tx_queues);
memzone: rename address from physical to IOVA The struct rte_memzone field .phys_addr is renamed to .iova. The deprecated name is kept in an anonymous union to avoid breaking the API. Signed-off-by: Thomas Monjalon <thomas@monjalon.net> Acked-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
2017-11-04 01:22:28 +00:00
hw->queueDescPA = mz->iova;
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
hw->queue_desc_len = (uint16_t)size;
vmxnet3: cleanup style and indentation This driver had several style problems, the worst of which was botched indentation. Fix almost all the problems reported by checkpatch. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:37:04 +00:00
if (dev->data->dev_conf.rxmode.mq_mode == ETH_MQ_RX_RSS) {
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
/* Allocate memory structure for UPT1_RSSConf and configure */
net/vmxnet3: reallocate shared memzone on re-config When adding a DPDK port to ovs-vswitchd with DPDK, the vmxnet3 device fails to activate due to mismatched magic number. This failure causes following operations to run: start the port, stop the port, reconfigure and re-start the port. During reconfigure, if there is an existing memzone, driver will reuse it. But reconfigure may request different number of Tx/Rx queues. This results in a memzone with wrong size and potential invalid memory access. To fix this, free the memzone if found and reserve a new one. Signed-off-by: Yong Wang <yongwang@vmware.com> Reviewed-by: Guolin Yang <gyang@vmware.com> Reviewed-by: Daniele Di Proietto <ddiproietto@vmware.com> Tested-by: Daniele Di Proietto <ddiproietto@vmware.com> Acked-by: Stephen Hemminger <stephen@networkplumber.org>
2016-08-29 19:18:47 +00:00
mz = gpa_zone_reserve(dev, sizeof(struct VMXNET3_RSSConf),
"rss_conf", rte_socket_id(),
RTE_CACHE_LINE_SIZE, 1);
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
if (mz == NULL) {
vmxnet3: cleanup style and indentation This driver had several style problems, the worst of which was botched indentation. Fix almost all the problems reported by checkpatch. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:37:04 +00:00
PMD_INIT_LOG(ERR,
vmxnet3: fix double spacing of log messages The debug log macro's already include newline, no need to double space the output. Note: other drivers have the same problem Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:37:13 +00:00
"ERROR: Creating rss_conf structure zone");
vmxnet3: cleanup style and indentation This driver had several style problems, the worst of which was botched indentation. Fix almost all the problems reported by checkpatch. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:37:04 +00:00
return -ENOMEM;
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
}
memset(mz->addr, 0, mz->len);
hw->rss_conf = mz->addr;
memzone: rename address from physical to IOVA The struct rte_memzone field .phys_addr is renamed to .iova. The deprecated name is kept in an anonymous union to avoid breaking the API. Signed-off-by: Thomas Monjalon <thomas@monjalon.net> Acked-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
2017-11-04 01:22:28 +00:00
hw->rss_confPA = mz->iova;
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
}
return 0;
}
vmxnet3: support setting MAC address Allow overriding the base mac address of the device. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Remy Horton <remy.horton@intel.com>
2016-03-04 18:08:02 +00:00
static void
vmxnet3_write_mac(struct vmxnet3_hw *hw, const uint8_t *addr)
{
uint32_t val;
PMD_INIT_LOG(DEBUG,
"Writing MAC Address : %02x:%02x:%02x:%02x:%02x:%02x",
addr[0], addr[1], addr[2],
addr[3], addr[4], addr[5]);
net/vmxnet3: fix MAC address set Updated vmxnet3_mac_addr_set() to store the newly set MAC address. Modified vmxnet3_write_mac() so the h/w is updated in an endian neutral manner. Fixes: 139f39a97859 ("vmxnet3: support setting MAC address") Cc: stable@dpdk.org Signed-off-by: David Harton <dharton@cisco.com> Acked-by: Shrikrishna Khare <skhare@vmware.com>
2017-08-25 18:54:57 +00:00
memcpy(&val, addr, 4);
vmxnet3: support setting MAC address Allow overriding the base mac address of the device. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Remy Horton <remy.horton@intel.com>
2016-03-04 18:08:02 +00:00
VMXNET3_WRITE_BAR1_REG(hw, VMXNET3_REG_MACL, val);
net/vmxnet3: fix MAC address set Updated vmxnet3_mac_addr_set() to store the newly set MAC address. Modified vmxnet3_write_mac() so the h/w is updated in an endian neutral manner. Fixes: 139f39a97859 ("vmxnet3: support setting MAC address") Cc: stable@dpdk.org Signed-off-by: David Harton <dharton@cisco.com> Acked-by: Shrikrishna Khare <skhare@vmware.com>
2017-08-25 18:54:57 +00:00
memcpy(&val, addr + 4, 2);
vmxnet3: support setting MAC address Allow overriding the base mac address of the device. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Remy Horton <remy.horton@intel.com>
2016-03-04 18:08:02 +00:00
VMXNET3_WRITE_BAR1_REG(hw, VMXNET3_REG_MACH, val);
}
net/vmxnet3: add cmd to register memory region In vmxnet3 version 3, the emulation added support for the vmxnet3 driver to communicate information about the memory regions the driver will use for rx/tx buffers. The driver can also indicate which rx/tx queue the memory region is applicable for. If this information is communicated to the emulation, the emulation will always keep these memory regions mapped, thereby avoiding the mapping/unmapping overhead for every packet. Signed-off-by: Shrikrishna Khare <skhare@vmware.com> Signed-off-by: Guolin Yang <gyang@vmware.com> Acked-by: Yong Wang <yongwang@vmware.com> Acked-by: Jin Heo <heoj@vmware.com>
2017-03-06 22:55:06 +00:00
static int
vmxnet3_dev_setup_memreg(struct rte_eth_dev *dev)
{
struct vmxnet3_hw *hw = dev->data->dev_private;
Vmxnet3_DriverShared *shared = hw->shared;
Vmxnet3_CmdInfo *cmdInfo;
struct rte_mempool *mp[VMXNET3_MAX_RX_QUEUES];
uint8_t index[VMXNET3_MAX_RX_QUEUES + VMXNET3_MAX_TX_QUEUES];
uint32_t num, i, j, size;
if (hw->memRegsPA == 0) {
const struct rte_memzone *mz;
size = sizeof(Vmxnet3_MemRegs) +
(VMXNET3_MAX_RX_QUEUES + VMXNET3_MAX_TX_QUEUES) *
sizeof(Vmxnet3_MemoryRegion);
mz = gpa_zone_reserve(dev, size, "memRegs", rte_socket_id(), 8,
1);
if (mz == NULL) {
PMD_INIT_LOG(ERR, "ERROR: Creating memRegs zone");
return -ENOMEM;
}
memset(mz->addr, 0, mz->len);
hw->memRegs = mz->addr;
memzone: rename address from physical to IOVA The struct rte_memzone field .phys_addr is renamed to .iova. The deprecated name is kept in an anonymous union to avoid breaking the API. Signed-off-by: Thomas Monjalon <thomas@monjalon.net> Acked-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
2017-11-04 01:22:28 +00:00
hw->memRegsPA = mz->iova;
net/vmxnet3: add cmd to register memory region In vmxnet3 version 3, the emulation added support for the vmxnet3 driver to communicate information about the memory regions the driver will use for rx/tx buffers. The driver can also indicate which rx/tx queue the memory region is applicable for. If this information is communicated to the emulation, the emulation will always keep these memory regions mapped, thereby avoiding the mapping/unmapping overhead for every packet. Signed-off-by: Shrikrishna Khare <skhare@vmware.com> Signed-off-by: Guolin Yang <gyang@vmware.com> Acked-by: Yong Wang <yongwang@vmware.com> Acked-by: Jin Heo <heoj@vmware.com>
2017-03-06 22:55:06 +00:00
}
num = hw->num_rx_queues;
for (i = 0; i < num; i++) {
vmxnet3_rx_queue_t *rxq = dev->data->rx_queues[i];
mp[i] = rxq->mp;
index[i] = 1 << i;
}
/*
* The same mempool could be used by multiple queues. In such a case,
* remove duplicate mempool entries. Only one entry is kept with
* bitmask indicating queues that are using this mempool.
*/
for (i = 1; i < num; i++) {
for (j = 0; j < i; j++) {
if (mp[i] == mp[j]) {
mp[i] = NULL;
index[j] |= 1 << i;
break;
}
}
}
j = 0;
for (i = 0; i < num; i++) {
if (mp[i] == NULL)
continue;
Vmxnet3_MemoryRegion *mr = &hw->memRegs->memRegs[j];
mr->startPA =
mempool: rename addresses from physical to IOVA The struct fields phys_addr_t rte_mempool_objhdr.physaddr and rte_mempool_memhdr.phys_addr are renamed to rte_iova_t iova. The deprecated names are kept in an anonymous union to avoid breaking the API. Signed-off-by: Thomas Monjalon <thomas@monjalon.net> Acked-by: Santosh Shukla <santosh.shukla@caviumnetworks.com> Acked-by: Olivier Matz <olivier.matz@6wind.com>
2017-10-20 12:31:31 +00:00
(uintptr_t)STAILQ_FIRST(&mp[i]->mem_list)->iova;
net/vmxnet3: add cmd to register memory region In vmxnet3 version 3, the emulation added support for the vmxnet3 driver to communicate information about the memory regions the driver will use for rx/tx buffers. The driver can also indicate which rx/tx queue the memory region is applicable for. If this information is communicated to the emulation, the emulation will always keep these memory regions mapped, thereby avoiding the mapping/unmapping overhead for every packet. Signed-off-by: Shrikrishna Khare <skhare@vmware.com> Signed-off-by: Guolin Yang <gyang@vmware.com> Acked-by: Yong Wang <yongwang@vmware.com> Acked-by: Jin Heo <heoj@vmware.com>
2017-03-06 22:55:06 +00:00
mr->length = STAILQ_FIRST(&mp[i]->mem_list)->len <= INT32_MAX ?
STAILQ_FIRST(&mp[i]->mem_list)->len : INT32_MAX;
mr->txQueueBits = index[i];
mr->rxQueueBits = index[i];
PMD_INIT_LOG(INFO,
"index: %u startPA: %" PRIu64 " length: %u, "
"rxBits: %x",
j, mr->startPA, mr->length, mr->rxQueueBits);
j++;
}
hw->memRegs->numRegs = j;
PMD_INIT_LOG(INFO, "numRegs: %u", j);
size = sizeof(Vmxnet3_MemRegs) +
(j - 1) * sizeof(Vmxnet3_MemoryRegion);
cmdInfo = &shared->cu.cmdInfo;
cmdInfo->varConf.confVer = 1;
cmdInfo->varConf.confLen = size;
cmdInfo->varConf.confPA = hw->memRegsPA;
return 0;
}
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
static int
vmxnet3_setup_driver_shared(struct rte_eth_dev *dev)
{
struct rte_eth_conf port_conf = dev->data->dev_conf;
vmxnet3: remove useless adapter wrapper The adapter struct is just a wrapper around the vmxnet3_hw structure. Eliminate the wrapper and get rid of the macro used to access and needlessly cast the private data. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:39:58 +00:00
struct vmxnet3_hw *hw = dev->data->dev_private;
vmxnet3: support jumbo frames Add support for linking multi-segment buffers together to handle Jumbo packets. The vmxnet3 API supports having header and body buffer types. What this patch does is fill the primary ring completely with header buffers and the secondary ring with body buffers. This allows for non-jumbo frames to only use one mbuf (from primary ring); and jumbo frames will have first mbuf from primary ring and following mbufs from other ring. This could be optimized in future if the DPDK had API to supply different sized mbufs (two pools) into driver. Signed-off-by: Stephen Hemminger <shemming@brocade.com> Acked-by: Remy Horton <remy.horton@intel.com> Release note addition: Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2016-03-04 18:08:00 +00:00
uint32_t mtu = dev->data->mtu;
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
Vmxnet3_DriverShared *shared = hw->shared;
Vmxnet3_DSDevRead *devRead = &shared->devRead;
net/vmxnet3: convert to new Rx offload API Ethdev RX offloads API has changed since: commit ce17eddefc20 ("ethdev: introduce Rx queue offloads API") This patch adopts the new RX Offload API in vmxnet3 driver. Signed-off-by: Louis Luo <llouis@vmware.com> Acked-by: Yong Wang <yongwang@vmware.com>
2018-05-01 21:22:56 +00:00
uint64_t rx_offloads = dev->data->dev_conf.rxmode.offloads;
vmxnet3: support setting MAC address Allow overriding the base mac address of the device. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Remy Horton <remy.horton@intel.com>
2016-03-04 18:08:02 +00:00
uint32_t i;
vmxnet3: fix VLAN filtering During an MTU change, the adapter is restarted. If hardware VLAN offload is in use, this existing filter table would also be cleared. Instead, setup the shadow table once during device initialization and just update during restart. vmxnet3_dev_vlan_offload_set(dev, mask) was incorrectly treating the mask parameter as the bitmask for vlan_strip and vlan_filter, whereas the mask indicates only what has changed - the values for vlan_stripping and vlan_filter needs to be taken from dev_conf.rxmode. Fixes: f003fc383487 ("vmxnet3: enable vlan filtering") Signed-off-by: Charles (Chas) Williams <ciwillia@brocade.com> Signed-off-by: Nachiketa Prachanda <nprachan@brocade.com> Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Remy Horton <remy.horton@intel.com>
2016-03-04 18:08:01 +00:00
int ret;
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
net/vmxnet3: guess MSS if not provided in LRO mode Not so old variants of vmxnet3 do not provide MSS value along with LRO packet. When this case happens, try to guess MSS value with information at hand. Signed-off-by: Didier Pallard <didier.pallard@6wind.com> Acked-by: Yong Wang <yongwang@vmware.com>
2018-03-28 15:43:47 +00:00
hw->mtu = mtu;
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
shared->magic = VMXNET3_REV1_MAGIC;
devRead->misc.driverInfo.version = VMXNET3_DRIVER_VERSION_NUM;
/* Setting up Guest OS information */
devRead->misc.driverInfo.gos.gosBits = sizeof(void *) == 4 ?
net/vmxnet3: coding style changes Signed-off-by: Yong Wang <yongwang@vmware.com> Acked-by: Stephen Hemminger <stephen@networkplumber.org>
2016-08-29 19:18:46 +00:00
VMXNET3_GOS_BITS_32 : VMXNET3_GOS_BITS_64;
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
devRead->misc.driverInfo.gos.gosType = VMXNET3_GOS_TYPE_LINUX;
devRead->misc.driverInfo.vmxnet3RevSpt = 1;
devRead->misc.driverInfo.uptVerSpt = 1;
vmxnet3: support jumbo frames Add support for linking multi-segment buffers together to handle Jumbo packets. The vmxnet3 API supports having header and body buffer types. What this patch does is fill the primary ring completely with header buffers and the secondary ring with body buffers. This allows for non-jumbo frames to only use one mbuf (from primary ring); and jumbo frames will have first mbuf from primary ring and following mbufs from other ring. This could be optimized in future if the DPDK had API to supply different sized mbufs (two pools) into driver. Signed-off-by: Stephen Hemminger <shemming@brocade.com> Acked-by: Remy Horton <remy.horton@intel.com> Release note addition: Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2016-03-04 18:08:00 +00:00
devRead->misc.mtu = rte_le_to_cpu_32(mtu);
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
devRead->misc.queueDescPA = hw->queueDescPA;
devRead->misc.queueDescLen = hw->queue_desc_len;
devRead->misc.numTxQueues = hw->num_tx_queues;
devRead->misc.numRxQueues = hw->num_rx_queues;
/*
vmxnet3: cleanup style and indentation This driver had several style problems, the worst of which was botched indentation. Fix almost all the problems reported by checkpatch. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:37:04 +00:00
* Set number of interrupts to 1
net/vmxnet3: generate link-state change notifications Generate link-state change notifications by listening to interrupts generated by the device. Make use of the existing vmxnet3_process_events function that was compiled out, but change it to call vmxnet3_dev_link_update on a VMXNET3_ECR_LINK event and to not be so noisy in its log messages. Enable interrupts on starting the device, using a new helper function, vmxnet3_enable_intr, based on vmxnet3_disable_intr and validated against the FreeBSD driver. Keep track of the number of interrupts registered for to avoid hardcoding these in vmxnet3_enable/disable_intr and to provision for any future rxq intr support. Factor out the guts of vmxnet3_dev_link_update minus the started check to allow the new function to be called from vmxnet3_dev_start in the lsc-enabled case to ensure that the link state is correctly set from the actual state at that point. Signed-off-by: Robert Shearman <rshearma@brocade.com>
2017-06-15 12:16:10 +00:00
* PMD by default disables all the interrupts but this is MUST
* to activate device. It needs at least one interrupt for
* link events to handle
vmxnet3: cleanup style and indentation This driver had several style problems, the worst of which was botched indentation. Fix almost all the problems reported by checkpatch. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:37:04 +00:00
*/
net/vmxnet3: generate link-state change notifications Generate link-state change notifications by listening to interrupts generated by the device. Make use of the existing vmxnet3_process_events function that was compiled out, but change it to call vmxnet3_dev_link_update on a VMXNET3_ECR_LINK event and to not be so noisy in its log messages. Enable interrupts on starting the device, using a new helper function, vmxnet3_enable_intr, based on vmxnet3_disable_intr and validated against the FreeBSD driver. Keep track of the number of interrupts registered for to avoid hardcoding these in vmxnet3_enable/disable_intr and to provision for any future rxq intr support. Factor out the guts of vmxnet3_dev_link_update minus the started check to allow the new function to be called from vmxnet3_dev_start in the lsc-enabled case to ensure that the link state is correctly set from the actual state at that point. Signed-off-by: Robert Shearman <rshearma@brocade.com>
2017-06-15 12:16:10 +00:00
hw->num_intrs = devRead->intrConf.numIntrs = 1;
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
devRead->intrConf.intrCtrl |= VMXNET3_IC_DISABLE_ALL;
for (i = 0; i < hw->num_tx_queues; i++) {
Vmxnet3_TxQueueDesc *tqd = &hw->tqd_start[i];
vmxnet3: leverage data ring on Tx path Data_ring is a pre-mapped guest ring buffer that vmxnet3 backend has access to directly without a need for buffer address mapping and unmapping during packet transmission. It is useful in reducing device emulation cost on the tx path. There are some additional cost though on the guest driver for packet copy and overall it's a win. This patch leverages the data_ring for packets with a length less than or equal to the data_ring entry size (128B). For larger packet, we won't use the data_ring as that requires one extra tx descriptor and it's not clear if doing this will be beneficial. Performance results show that this patch significantly boosts vmxnet3 64B tx performance (pkt rate) for l2fwd application on a Ivy Bridge server by >20% at which point we start to hit some bottleneck on the rx side. Signed-off-by: Yong Wang <yongwang@vmware.com>
2014-11-05 01:49:43 +00:00
vmxnet3_tx_queue_t *txq = dev->data->tx_queues[i];
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
net/vmxnet3: set the queue shared buffer at start If a reconfiguration happens, queuedesc is reallocated. Any queues that are preserved point to the previous queuedesc since the queues are only configured during queue setup. Delay configuration of the shared queue pointers until device start when queuedesc is no longer changing. Fixes: 8618d19b52b1 ("net/vmxnet3: reallocate shared memzone on re-config") Cc: stable@dpdk.org Signed-off-by: Chas Williams <chas3@att.com> Acked-by: Shrikrishna Khare <skhare@vmware.com>
2018-01-17 15:04:56 +00:00
txq->shared = &hw->tqd_start[i];
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
tqd->ctrl.txNumDeferred = 0;
tqd->ctrl.txThreshold = 1;
tqd->conf.txRingBasePA = txq->cmd_ring.basePA;
tqd->conf.compRingBasePA = txq->comp_ring.basePA;
vmxnet3: leverage data ring on Tx path Data_ring is a pre-mapped guest ring buffer that vmxnet3 backend has access to directly without a need for buffer address mapping and unmapping during packet transmission. It is useful in reducing device emulation cost on the tx path. There are some additional cost though on the guest driver for packet copy and overall it's a win. This patch leverages the data_ring for packets with a length less than or equal to the data_ring entry size (128B). For larger packet, we won't use the data_ring as that requires one extra tx descriptor and it's not clear if doing this will be beneficial. Performance results show that this patch significantly boosts vmxnet3 64B tx performance (pkt rate) for l2fwd application on a Ivy Bridge server by >20% at which point we start to hit some bottleneck on the rx side. Signed-off-by: Yong Wang <yongwang@vmware.com>
2014-11-05 01:49:43 +00:00
tqd->conf.dataRingBasePA = txq->data_ring.basePA;
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
tqd->conf.txRingSize = txq->cmd_ring.size;
tqd->conf.compRingSize = txq->comp_ring.size;
vmxnet3: leverage data ring on Tx path Data_ring is a pre-mapped guest ring buffer that vmxnet3 backend has access to directly without a need for buffer address mapping and unmapping during packet transmission. It is useful in reducing device emulation cost on the tx path. There are some additional cost though on the guest driver for packet copy and overall it's a win. This patch leverages the data_ring for packets with a length less than or equal to the data_ring entry size (128B). For larger packet, we won't use the data_ring as that requires one extra tx descriptor and it's not clear if doing this will be beneficial. Performance results show that this patch significantly boosts vmxnet3 64B tx performance (pkt rate) for l2fwd application on a Ivy Bridge server by >20% at which point we start to hit some bottleneck on the rx side. Signed-off-by: Yong Wang <yongwang@vmware.com>
2014-11-05 01:49:43 +00:00
tqd->conf.dataRingSize = txq->data_ring.size;
net/vmxnet3: allow variable length Tx data ring vmxnet3 driver supports transmit data ring viz. a set of fixed size buffers used by the driver to copy packet headers. Small packets that fit these buffers are copied into these buffers entirely. Currently this buffer size of fixed at 128 bytes. This patch extends transmit data ring implementation to allow variable length transmit data ring buffers. The length of the buffer is read from the emulation during initialization. Signed-off-by: Shrikrishna Khare <skhare@vmware.com> Acked-by: Yong Wang <yongwang@vmware.com> Acked-by: Jin Heo <heoj@vmware.com>
2017-03-06 22:55:03 +00:00
tqd->conf.txDataRingDescSize = txq->txdata_desc_size;
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
tqd->conf.intrIdx = txq->comp_ring.intr_idx;
tqd->status.stopped = TRUE;
tqd->status.error = 0;
memset(&tqd->stats, 0, sizeof(tqd->stats));
}
for (i = 0; i < hw->num_rx_queues; i++) {
Vmxnet3_RxQueueDesc *rqd = &hw->rqd_start[i];
vmxnet3: leverage data ring on Tx path Data_ring is a pre-mapped guest ring buffer that vmxnet3 backend has access to directly without a need for buffer address mapping and unmapping during packet transmission. It is useful in reducing device emulation cost on the tx path. There are some additional cost though on the guest driver for packet copy and overall it's a win. This patch leverages the data_ring for packets with a length less than or equal to the data_ring entry size (128B). For larger packet, we won't use the data_ring as that requires one extra tx descriptor and it's not clear if doing this will be beneficial. Performance results show that this patch significantly boosts vmxnet3 64B tx performance (pkt rate) for l2fwd application on a Ivy Bridge server by >20% at which point we start to hit some bottleneck on the rx side. Signed-off-by: Yong Wang <yongwang@vmware.com>
2014-11-05 01:49:43 +00:00
vmxnet3_rx_queue_t *rxq = dev->data->rx_queues[i];
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
net/vmxnet3: set the queue shared buffer at start If a reconfiguration happens, queuedesc is reallocated. Any queues that are preserved point to the previous queuedesc since the queues are only configured during queue setup. Delay configuration of the shared queue pointers until device start when queuedesc is no longer changing. Fixes: 8618d19b52b1 ("net/vmxnet3: reallocate shared memzone on re-config") Cc: stable@dpdk.org Signed-off-by: Chas Williams <chas3@att.com> Acked-by: Shrikrishna Khare <skhare@vmware.com>
2018-01-17 15:04:56 +00:00
rxq->shared = &hw->rqd_start[i];
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
rqd->conf.rxRingBasePA[0] = rxq->cmd_ring[0].basePA;
rqd->conf.rxRingBasePA[1] = rxq->cmd_ring[1].basePA;
rqd->conf.compRingBasePA = rxq->comp_ring.basePA;
rqd->conf.rxRingSize[0] = rxq->cmd_ring[0].size;
rqd->conf.rxRingSize[1] = rxq->cmd_ring[1].size;
rqd->conf.compRingSize = rxq->comp_ring.size;
rqd->conf.intrIdx = rxq->comp_ring.intr_idx;
net/vmxnet3: support receive data ring vmxnet3 driver preallocates buffers for receiving packets and posts the buffers to the emulation. In order to deliver a received packet to the guest, the emulation must map buffer(s) and copy the packet into it. To avoid this memory mapping overhead, this patch introduces the receive data ring - a set of small sized buffers that are always mapped by the emulation. If a packet fits into the receive data ring buffer, the emulation delivers the packet via the receive data ring (which must be copied by the guest driver), or else the usual receive path is used. Signed-off-by: Shrikrishna Khare <skhare@vmware.com> Acked-by: Yong Wang <yongwang@vmware.com> Acked-by: Jin Heo <heoj@vmware.com>
2017-03-06 22:55:04 +00:00
if (VMXNET3_VERSION_GE_3(hw)) {
rqd->conf.rxDataRingBasePA = rxq->data_ring.basePA;
rqd->conf.rxDataRingDescSize = rxq->data_desc_size;
}
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
rqd->status.stopped = TRUE;
rqd->status.error = 0;
memset(&rqd->stats, 0, sizeof(rqd->stats));
}
/* RxMode set to 0 of VMXNET3_RXM_xxx */
devRead->rxFilterConf.rxMode = 0;
/* Setting up feature flags */
net/vmxnet3: convert to new Rx offload API Ethdev RX offloads API has changed since: commit ce17eddefc20 ("ethdev: introduce Rx queue offloads API") This patch adopts the new RX Offload API in vmxnet3 driver. Signed-off-by: Louis Luo <llouis@vmware.com> Acked-by: Yong Wang <yongwang@vmware.com>
2018-05-01 21:22:56 +00:00
if (rx_offloads & DEV_RX_OFFLOAD_CHECKSUM)
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
devRead->misc.uptFeatures |= VMXNET3_F_RXCSUM;
net/vmxnet3: convert to new Rx offload API Ethdev RX offloads API has changed since: commit ce17eddefc20 ("ethdev: introduce Rx queue offloads API") This patch adopts the new RX Offload API in vmxnet3 driver. Signed-off-by: Louis Luo <llouis@vmware.com> Acked-by: Yong Wang <yongwang@vmware.com>
2018-05-01 21:22:56 +00:00
if (rx_offloads & DEV_RX_OFFLOAD_TCP_LRO) {
net/vmxnet3: enable LRO This change enables device LRO if requested. The current implementation of jumbo frame Rx can be used for LRO directly without changes. Note that since jumbo frame uses both ring0 and ring1, it cannot be enabled in UPT (VMDirectPath) mode. Signed-off-by: Yong Wang <yongwang@vmware.com> Acked-by: Stephen Hemminger <stephen@networkplumber.org>
2016-08-29 19:18:50 +00:00
devRead->misc.uptFeatures |= VMXNET3_F_LRO;
devRead->misc.maxNumRxSG = 0;
}
vmxnet3: cleanup style and indentation This driver had several style problems, the worst of which was botched indentation. Fix almost all the problems reported by checkpatch. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:37:04 +00:00
if (port_conf.rxmode.mq_mode == ETH_MQ_RX_RSS) {
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
ret = vmxnet3_rss_configure(dev);
vmxnet3: cleanup style and indentation This driver had several style problems, the worst of which was botched indentation. Fix almost all the problems reported by checkpatch. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:37:04 +00:00
if (ret != VMXNET3_SUCCESS)
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
return ret;
vmxnet3: cleanup style and indentation This driver had several style problems, the worst of which was botched indentation. Fix almost all the problems reported by checkpatch. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:37:04 +00:00
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
devRead->misc.uptFeatures |= VMXNET3_F_RSS;
devRead->rssConfDesc.confVer = 1;
devRead->rssConfDesc.confLen = sizeof(struct VMXNET3_RSSConf);
devRead->rssConfDesc.confPA = hw->rss_confPA;
}
ethdev: allow returning error on VLAN offload ops Some devices may not support or fail setting VLAN offload configuration based on dynamic circumstances so the vlan_offload_set_t vector is modified to return an int so the caller can determine success or not. rte_eth_dev_set_vlan_offload is updated to return the value provided by the vector when called along with restoring the original offload configs on failure. Existing vlan_offload_set_t vectors are modified to return an int. Majority of cases return 0 but a few that actually can fail now return their failure codes. Finally, a vlan_offload_set_t vector is added to virtio to facilitate dynamically turning VLAN strip on or off. Signed-off-by: David Harton <dharton@cisco.com> Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com>
2017-09-01 02:36:28 +00:00
ret = vmxnet3_dev_vlan_offload_set(dev,
ETH_VLAN_STRIP_MASK | ETH_VLAN_FILTER_MASK);
if (ret)
return ret;
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
net/vmxnet3: preserve configured MAC address When starting a vmxnet3 device, it is always writing the permanent MAC address, even if a different MAC address was configured. Write from the device data instead which holds the current one. Signed-off-by: George Wilkie <gwilkie@brocade.com> Acked-by: Shrikrishna Khare <skhare@vmware.com>
2017-06-15 12:17:42 +00:00
vmxnet3_write_mac(hw, dev->data->mac_addrs->addr_bytes);
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
return VMXNET3_SUCCESS;
}
/*
* Configure device link speed and setup link.
* Must be called after eth_vmxnet3_dev_init. Other wise it might fail
* It returns 0 on success.
*/
static int
vmxnet3_dev_start(struct rte_eth_dev *dev)
{
net/vmxnet3: improve error checks and return values Signed-off-by: Yong Wang <yongwang@vmware.com> Reviewed-by: Juho Snellman <jsnell@iki.fi> Acked-by: Stephen Hemminger <stephen@networkplumber.org>
2016-08-29 19:18:45 +00:00
int ret;
vmxnet3: remove useless adapter wrapper The adapter struct is just a wrapper around the vmxnet3_hw structure. Eliminate the wrapper and get rid of the macro used to access and needlessly cast the private data. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:39:58 +00:00
struct vmxnet3_hw *hw = dev->data->dev_private;
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
PMD_INIT_FUNC_TRACE();
net/vmxnet3: retain counters on restart Most NICs like virtio, igb/ixgbe etc. don't reset counters on dev_start and arguably this helps in monitoring the counters across a longer time span with multiple device start/stops. vmxnet3 behavior is opposite to that and counters are reset by the host side implementation each time the device is restarted. Change the driver to save the counters in its private context before it is reset by writing CMD_ACTIVATE to REG_CMD. Signed-off-by: Nachiketa Prachanda <nprachan@brocade.com> Acked-by: Shrikrishna Khare <skhare@vmware.com>
2017-06-15 12:16:08 +00:00
/* Save stats before it is reset by CMD_ACTIVATE */
vmxnet3_hw_stats_save(hw);
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
ret = vmxnet3_setup_driver_shared(dev);
vmxnet3: cleanup style and indentation This driver had several style problems, the worst of which was botched indentation. Fix almost all the problems reported by checkpatch. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:37:04 +00:00
if (ret != VMXNET3_SUCCESS)
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
return ret;
net/vmxnet3: generate link-state change notifications Generate link-state change notifications by listening to interrupts generated by the device. Make use of the existing vmxnet3_process_events function that was compiled out, but change it to call vmxnet3_dev_link_update on a VMXNET3_ECR_LINK event and to not be so noisy in its log messages. Enable interrupts on starting the device, using a new helper function, vmxnet3_enable_intr, based on vmxnet3_disable_intr and validated against the FreeBSD driver. Keep track of the number of interrupts registered for to avoid hardcoding these in vmxnet3_enable/disable_intr and to provision for any future rxq intr support. Factor out the guts of vmxnet3_dev_link_update minus the started check to allow the new function to be called from vmxnet3_dev_start in the lsc-enabled case to ensure that the link state is correctly set from the actual state at that point. Signed-off-by: Robert Shearman <rshearma@brocade.com>
2017-06-15 12:16:10 +00:00
/* check if lsc interrupt feature is enabled */
if (dev->data->dev_conf.intr_conf.lsc) {
struct rte_pci_device *pci_dev = RTE_DEV_TO_PCI(dev->device);
/* Setup interrupt callback */
rte_intr_callback_register(&pci_dev->intr_handle,
vmxnet3_interrupt_handler, dev);
if (rte_intr_enable(&pci_dev->intr_handle) < 0) {
PMD_INIT_LOG(ERR, "interrupt enable failed");
return -EIO;
}
}
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
/* Exchange shared data with device */
vmxnet3: cleanup style and indentation This driver had several style problems, the worst of which was botched indentation. Fix almost all the problems reported by checkpatch. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:37:04 +00:00
VMXNET3_WRITE_BAR1_REG(hw, VMXNET3_REG_DSAL,
VMXNET3_GET_ADDR_LO(hw->sharedPA));
VMXNET3_WRITE_BAR1_REG(hw, VMXNET3_REG_DSAH,
VMXNET3_GET_ADDR_HI(hw->sharedPA));
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
vmxnet3: cleanup style and indentation This driver had several style problems, the worst of which was botched indentation. Fix almost all the problems reported by checkpatch. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:37:04 +00:00
/* Activate device by register write */
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
VMXNET3_WRITE_BAR1_REG(hw, VMXNET3_REG_CMD, VMXNET3_CMD_ACTIVATE_DEV);
net/vmxnet3: improve error checks and return values Signed-off-by: Yong Wang <yongwang@vmware.com> Reviewed-by: Juho Snellman <jsnell@iki.fi> Acked-by: Stephen Hemminger <stephen@networkplumber.org>
2016-08-29 19:18:45 +00:00
ret = VMXNET3_READ_BAR1_REG(hw, VMXNET3_REG_CMD);
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
net/vmxnet3: improve error checks and return values Signed-off-by: Yong Wang <yongwang@vmware.com> Reviewed-by: Juho Snellman <jsnell@iki.fi> Acked-by: Stephen Hemminger <stephen@networkplumber.org>
2016-08-29 19:18:45 +00:00
if (ret != 0) {
vmxnet3: remove redundant function names in log Signed-off-by: Huawei Xie <huawei.xie@intel.com> Acked-by: Yuanhan Liu <yuanhan.liu@linux.intel.com> Acked-by: Stephen Hemminger <stephen@networkplumber.org>
2016-02-10 16:12:39 +00:00
PMD_INIT_LOG(ERR, "Device activation: UNSUCCESSFUL");
net/vmxnet3: improve error checks and return values Signed-off-by: Yong Wang <yongwang@vmware.com> Reviewed-by: Juho Snellman <jsnell@iki.fi> Acked-by: Stephen Hemminger <stephen@networkplumber.org>
2016-08-29 19:18:45 +00:00
return -EINVAL;
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
}
net/vmxnet3: add cmd to register memory region In vmxnet3 version 3, the emulation added support for the vmxnet3 driver to communicate information about the memory regions the driver will use for rx/tx buffers. The driver can also indicate which rx/tx queue the memory region is applicable for. If this information is communicated to the emulation, the emulation will always keep these memory regions mapped, thereby avoiding the mapping/unmapping overhead for every packet. Signed-off-by: Shrikrishna Khare <skhare@vmware.com> Signed-off-by: Guolin Yang <gyang@vmware.com> Acked-by: Yong Wang <yongwang@vmware.com> Acked-by: Jin Heo <heoj@vmware.com>
2017-03-06 22:55:06 +00:00
/* Setup memory region for rx buffers */
ret = vmxnet3_dev_setup_memreg(dev);
if (ret == 0) {
VMXNET3_WRITE_BAR1_REG(hw, VMXNET3_REG_CMD,
VMXNET3_CMD_REGISTER_MEMREGS);
ret = VMXNET3_READ_BAR1_REG(hw, VMXNET3_REG_CMD);
if (ret != 0)
PMD_INIT_LOG(DEBUG,
"Failed in setup memory region cmd\n");
ret = 0;
} else {
PMD_INIT_LOG(DEBUG, "Failed to setup memory region\n");
}
net/vmxnet3: fix RSS setting on v4 When calling to setup RSS on v4 API, ESX will expect IPv4/6 TCP RSS to be set/requested mandatory. This patch will: - Set IPv4/6 TCP RSS when these have not been set. A warning message is thrown to make sure we warn the application we are setting IPv4/6 TCP RSS when not set. - An additional check has been added to dodge RSS configuration altogether unless MQ_RSS has been requested, similar to v3. The alternative (returning error) was considered, the intent is to ease the task of setting up and running vmxnet3 in situations where it's supposed to be most straightforward (testpmd, pktgen). Bugzilla ID: 400 Fixes: 643fba770705 ("net/vmxnet3: add v4 boot and guest UDP RSS config") Cc: stable@dpdk.org Signed-off-by: Eduard Serra <eserra@vmware.com> Acked-by: Yong Wang <yongwang@vmware.com>
2020-03-04 02:35:56 +00:00
if (VMXNET3_VERSION_GE_4(hw) &&
dev->data->dev_conf.rxmode.mq_mode == ETH_MQ_RX_RSS) {
net/vmxnet3: add v4 boot and guest UDP RSS config This patch introduces: - VMxnet3 v4 negotiation and, - entirely guest-driven UDP RSS support. VMxnet3 v3 already has UDP RSS support, however it depends on hypervisor provisioning on the VM through ESX specific flags, which are not transparent or known to the guest later on. Vmxnet3 v4 introduces a new API transaction which allows configuring RSS entirely from the guest. This API must be invoked after device shared mem region is initialized. IPv4 ESP RSS (SPI based) is also available, but currently there are no ESP RSS definitions on rte_eth layer to handle that. Signed-off-by: Eduard Serra <eserra@vmware.com> Acked-by: Yong Wang <yongwang@vmware.com>
2019-04-18 20:59:37 +00:00
/* Check for additional RSS */
ret = vmxnet3_v4_rss_configure(dev);
if (ret != VMXNET3_SUCCESS) {
PMD_INIT_LOG(ERR, "Failed to configure v4 RSS");
return ret;
}
}
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
/* Disable interrupts */
vmxnet3_disable_intr(hw);
/*
* Load RX queues with blank mbufs and update next2fill index for device
* Update RxMode of the device
*/
ret = vmxnet3_dev_rxtx_init(dev);
vmxnet3: cleanup style and indentation This driver had several style problems, the worst of which was botched indentation. Fix almost all the problems reported by checkpatch. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:37:04 +00:00
if (ret != VMXNET3_SUCCESS) {
net/vmxnet3: improve error checks and return values Signed-off-by: Yong Wang <yongwang@vmware.com> Reviewed-by: Juho Snellman <jsnell@iki.fi> Acked-by: Stephen Hemminger <stephen@networkplumber.org>
2016-08-29 19:18:45 +00:00
PMD_INIT_LOG(ERR, "Device queue init: UNSUCCESSFUL");
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
return ret;
}
net/vmxnet3: add cmd to register memory region In vmxnet3 version 3, the emulation added support for the vmxnet3 driver to communicate information about the memory regions the driver will use for rx/tx buffers. The driver can also indicate which rx/tx queue the memory region is applicable for. If this information is communicated to the emulation, the emulation will always keep these memory regions mapped, thereby avoiding the mapping/unmapping overhead for every packet. Signed-off-by: Shrikrishna Khare <skhare@vmware.com> Signed-off-by: Guolin Yang <gyang@vmware.com> Acked-by: Yong Wang <yongwang@vmware.com> Acked-by: Jin Heo <heoj@vmware.com>
2017-03-06 22:55:06 +00:00
hw->adapter_stopped = FALSE;
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
/* Setting proper Rx Mode and issue Rx Mode Update command */
vmxnet3: initialize receive mode for broadcast The driver must listen to broadcast packets, like other devices. Otherwise protocols like ARP won't work! Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-07-25 17:50:37 +00:00
vmxnet3_dev_set_rxmode(hw, VMXNET3_RXM_UCAST | VMXNET3_RXM_BCAST, 1);
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
net/vmxnet3: generate link-state change notifications Generate link-state change notifications by listening to interrupts generated by the device. Make use of the existing vmxnet3_process_events function that was compiled out, but change it to call vmxnet3_dev_link_update on a VMXNET3_ECR_LINK event and to not be so noisy in its log messages. Enable interrupts on starting the device, using a new helper function, vmxnet3_enable_intr, based on vmxnet3_disable_intr and validated against the FreeBSD driver. Keep track of the number of interrupts registered for to avoid hardcoding these in vmxnet3_enable/disable_intr and to provision for any future rxq intr support. Factor out the guts of vmxnet3_dev_link_update minus the started check to allow the new function to be called from vmxnet3_dev_start in the lsc-enabled case to ensure that the link state is correctly set from the actual state at that point. Signed-off-by: Robert Shearman <rshearma@brocade.com>
2017-06-15 12:16:10 +00:00
if (dev->data->dev_conf.intr_conf.lsc) {
vmxnet3_enable_intr(hw);
/*
* Update link state from device since this won't be
* done upon starting with lsc in use. This is done
* only after enabling interrupts to avoid any race
* where the link state could change without an
* interrupt being fired.
*/
__vmxnet3_dev_link_update(dev, 0);
}
net/vmxnet3: improve error checks and return values Signed-off-by: Yong Wang <yongwang@vmware.com> Reviewed-by: Juho Snellman <jsnell@iki.fi> Acked-by: Stephen Hemminger <stephen@networkplumber.org>
2016-08-29 19:18:45 +00:00
return VMXNET3_SUCCESS;
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
}
/*
* Stop device: disable rx and tx functions to allow for reconfiguring.
*/
static void
vmxnet3_dev_stop(struct rte_eth_dev *dev)
{
struct rte_eth_link link;
vmxnet3: remove useless adapter wrapper The adapter struct is just a wrapper around the vmxnet3_hw structure. Eliminate the wrapper and get rid of the macro used to access and needlessly cast the private data. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:39:58 +00:00
struct vmxnet3_hw *hw = dev->data->dev_private;
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
PMD_INIT_FUNC_TRACE();
vmxnet3: support port hotplug Signed-off-by: Bernard Iremonger <bernard.iremonger@intel.com> Acked-by: Yong Wang <yongwang@vmware.com>
2015-10-27 17:12:26 +00:00
if (hw->adapter_stopped == 1) {
net/vmxnet3: fix hot-unplug The vmxnet3 driver can't call back into dev_close(), and possibly dev_stop(), in dev_uninit(). When dev_uninit() is called, anything that those routines would want to clean up has already been released. Further, for complete cleanup, it is necessary to release any of the queue resources during dev_close(). This allows a vmxnet3 device to be hot-unplugged without leaking queues. Also set RTE_ETH_DEV_CLOSE_REMOVE on close so that the port resources can be deallocated. Return EBUSY if remove is called before stop. Fixes: dfaff37fc46d ("vmxnet3: import new vmxnet3 poll mode driver implementation") Cc: stable@dpdk.org Signed-off-by: Brian Russell <brussell@brocade.com> Signed-off-by: Luca Boccassi <bluca@debian.org>
2018-10-31 18:39:44 +00:00
PMD_INIT_LOG(DEBUG, "Device already stopped.");
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
return;
}
/* disable interrupts */
vmxnet3_disable_intr(hw);
net/vmxnet3: generate link-state change notifications Generate link-state change notifications by listening to interrupts generated by the device. Make use of the existing vmxnet3_process_events function that was compiled out, but change it to call vmxnet3_dev_link_update on a VMXNET3_ECR_LINK event and to not be so noisy in its log messages. Enable interrupts on starting the device, using a new helper function, vmxnet3_enable_intr, based on vmxnet3_disable_intr and validated against the FreeBSD driver. Keep track of the number of interrupts registered for to avoid hardcoding these in vmxnet3_enable/disable_intr and to provision for any future rxq intr support. Factor out the guts of vmxnet3_dev_link_update minus the started check to allow the new function to be called from vmxnet3_dev_start in the lsc-enabled case to ensure that the link state is correctly set from the actual state at that point. Signed-off-by: Robert Shearman <rshearma@brocade.com>
2017-06-15 12:16:10 +00:00
if (dev->data->dev_conf.intr_conf.lsc) {
struct rte_pci_device *pci_dev = RTE_DEV_TO_PCI(dev->device);
rte_intr_disable(&pci_dev->intr_handle);
rte_intr_callback_unregister(&pci_dev->intr_handle,
vmxnet3_interrupt_handler, dev);
}
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
/* quiesce the device first */
VMXNET3_WRITE_BAR1_REG(hw, VMXNET3_REG_CMD, VMXNET3_CMD_QUIESCE_DEV);
VMXNET3_WRITE_BAR1_REG(hw, VMXNET3_REG_DSAL, 0);
VMXNET3_WRITE_BAR1_REG(hw, VMXNET3_REG_DSAH, 0);
/* reset the device */
VMXNET3_WRITE_BAR1_REG(hw, VMXNET3_REG_CMD, VMXNET3_CMD_RESET_DEV);
vmxnet3: fix double spacing of log messages The debug log macro's already include newline, no need to double space the output. Note: other drivers have the same problem Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:37:13 +00:00
PMD_INIT_LOG(DEBUG, "Device reset.");
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
vmxnet3_dev_clear_queues(dev);
/* Clear recorded link status */
memset(&link, 0, sizeof(link));
net/vmxnet3: keep consistent link status Bonding may examine the link properties to ensure that matching interfaces are bound together. If the link is going to have fixed properties, these need to remain consistent regardless of the link_status or the state of the adapter. Signed-off-by: Chas Williams <chas3@att.com> Acked-by: Shrikrishna Khare <skhare@vmware.com>
2018-01-17 15:04:57 +00:00
link.link_duplex = ETH_LINK_FULL_DUPLEX;
link.link_speed = ETH_SPEED_NUM_10G;
link.link_autoneg = ETH_LINK_FIXED;
net/vmxnet3: use link status helper functions Use new rte_eth_linkstatus_get/set helper functions to handle link status update. Also remove no longer necessary include of rte_atomic.h Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2018-01-26 02:01:40 +00:00
rte_eth_linkstatus_set(dev, &link);
net/vmxnet3: fix hot-unplug The vmxnet3 driver can't call back into dev_close(), and possibly dev_stop(), in dev_uninit(). When dev_uninit() is called, anything that those routines would want to clean up has already been released. Further, for complete cleanup, it is necessary to release any of the queue resources during dev_close(). This allows a vmxnet3 device to be hot-unplugged without leaking queues. Also set RTE_ETH_DEV_CLOSE_REMOVE on close so that the port resources can be deallocated. Return EBUSY if remove is called before stop. Fixes: dfaff37fc46d ("vmxnet3: import new vmxnet3 poll mode driver implementation") Cc: stable@dpdk.org Signed-off-by: Brian Russell <brussell@brocade.com> Signed-off-by: Luca Boccassi <bluca@debian.org>
2018-10-31 18:39:44 +00:00
hw->adapter_stopped = 1;
}
static void
vmxnet3_free_queues(struct rte_eth_dev *dev)
{
int i;
PMD_INIT_FUNC_TRACE();
for (i = 0; i < dev->data->nb_rx_queues; i++) {
void *rxq = dev->data->rx_queues[i];
vmxnet3_dev_rx_queue_release(rxq);
}
dev->data->nb_rx_queues = 0;
for (i = 0; i < dev->data->nb_tx_queues; i++) {
void *txq = dev->data->tx_queues[i];
vmxnet3_dev_tx_queue_release(txq);
}
dev->data->nb_tx_queues = 0;
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
}
/*
* Reset and stop device.
*/
static void
vmxnet3_dev_close(struct rte_eth_dev *dev)
{
PMD_INIT_FUNC_TRACE();
vmxnet3_dev_stop(dev);
net/vmxnet3: fix hot-unplug The vmxnet3 driver can't call back into dev_close(), and possibly dev_stop(), in dev_uninit(). When dev_uninit() is called, anything that those routines would want to clean up has already been released. Further, for complete cleanup, it is necessary to release any of the queue resources during dev_close(). This allows a vmxnet3 device to be hot-unplugged without leaking queues. Also set RTE_ETH_DEV_CLOSE_REMOVE on close so that the port resources can be deallocated. Return EBUSY if remove is called before stop. Fixes: dfaff37fc46d ("vmxnet3: import new vmxnet3 poll mode driver implementation") Cc: stable@dpdk.org Signed-off-by: Brian Russell <brussell@brocade.com> Signed-off-by: Luca Boccassi <bluca@debian.org>
2018-10-31 18:39:44 +00:00
vmxnet3_free_queues(dev);
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
}
net/vmxnet3: retain counters on restart Most NICs like virtio, igb/ixgbe etc. don't reset counters on dev_start and arguably this helps in monitoring the counters across a longer time span with multiple device start/stops. vmxnet3 behavior is opposite to that and counters are reset by the host side implementation each time the device is restarted. Change the driver to save the counters in its private context before it is reset by writing CMD_ACTIVATE to REG_CMD. Signed-off-by: Nachiketa Prachanda <nprachan@brocade.com> Acked-by: Shrikrishna Khare <skhare@vmware.com>
2017-06-15 12:16:08 +00:00
static void
vmxnet3_hw_tx_stats_get(struct vmxnet3_hw *hw, unsigned int q,
struct UPT1_TxStats *res)
{
#define VMXNET3_UPDATE_TX_STAT(h, i, f, r) \
((r)->f = (h)->tqd_start[(i)].stats.f + \
(h)->saved_tx_stats[(i)].f)
VMXNET3_UPDATE_TX_STAT(hw, q, ucastPktsTxOK, res);
VMXNET3_UPDATE_TX_STAT(hw, q, mcastPktsTxOK, res);
VMXNET3_UPDATE_TX_STAT(hw, q, bcastPktsTxOK, res);
VMXNET3_UPDATE_TX_STAT(hw, q, ucastBytesTxOK, res);
VMXNET3_UPDATE_TX_STAT(hw, q, mcastBytesTxOK, res);
VMXNET3_UPDATE_TX_STAT(hw, q, bcastBytesTxOK, res);
VMXNET3_UPDATE_TX_STAT(hw, q, pktsTxError, res);
VMXNET3_UPDATE_TX_STAT(hw, q, pktsTxDiscard, res);
#undef VMXNET3_UPDATE_TX_STAT
}
static void
vmxnet3_hw_rx_stats_get(struct vmxnet3_hw *hw, unsigned int q,
struct UPT1_RxStats *res)
{
#define VMXNET3_UPDATE_RX_STAT(h, i, f, r) \
((r)->f = (h)->rqd_start[(i)].stats.f + \
(h)->saved_rx_stats[(i)].f)
VMXNET3_UPDATE_RX_STAT(hw, q, ucastPktsRxOK, res);
VMXNET3_UPDATE_RX_STAT(hw, q, mcastPktsRxOK, res);
VMXNET3_UPDATE_RX_STAT(hw, q, bcastPktsRxOK, res);
VMXNET3_UPDATE_RX_STAT(hw, q, ucastBytesRxOK, res);
VMXNET3_UPDATE_RX_STAT(hw, q, mcastBytesRxOK, res);
VMXNET3_UPDATE_RX_STAT(hw, q, bcastBytesRxOK, res);
VMXNET3_UPDATE_RX_STAT(hw, q, pktsRxError, res);
VMXNET3_UPDATE_RX_STAT(hw, q, pktsRxOutOfBuf, res);
net/vmxnet3: support stats reset 'stats_reset()' callback was missing because the device backend doesn't support it. This commit adds a workaround to this and implements the callback by taking a snapshot of the stats (SNAPSHOT) each time 'stats_reset()' is called. When getting stats with 'stats_get()', hw stats which always increase reduce SNAPSHOT stats. That's how we get the "real" stats since the last 'stats_reset()'. Signed-off-by: Yogev Chaimovich <yogev@cgstowernetworks.com> Acked-by: Yong Wang <yongwang@vmware.com>
2018-09-20 06:37:05 +00:00
#undef VMXNET3_UPDATE_RX_STAT
}
static void
vmxnet3_tx_stats_get(struct vmxnet3_hw *hw, unsigned int q,
struct UPT1_TxStats *res)
{
vmxnet3_hw_tx_stats_get(hw, q, res);
#define VMXNET3_REDUCE_SNAPSHOT_TX_STAT(h, i, f, r) \
((r)->f -= (h)->snapshot_tx_stats[(i)].f)
VMXNET3_REDUCE_SNAPSHOT_TX_STAT(hw, q, ucastPktsTxOK, res);
VMXNET3_REDUCE_SNAPSHOT_TX_STAT(hw, q, mcastPktsTxOK, res);
VMXNET3_REDUCE_SNAPSHOT_TX_STAT(hw, q, bcastPktsTxOK, res);
VMXNET3_REDUCE_SNAPSHOT_TX_STAT(hw, q, ucastBytesTxOK, res);
VMXNET3_REDUCE_SNAPSHOT_TX_STAT(hw, q, mcastBytesTxOK, res);
VMXNET3_REDUCE_SNAPSHOT_TX_STAT(hw, q, bcastBytesTxOK, res);
VMXNET3_REDUCE_SNAPSHOT_TX_STAT(hw, q, pktsTxError, res);
VMXNET3_REDUCE_SNAPSHOT_TX_STAT(hw, q, pktsTxDiscard, res);
#undef VMXNET3_REDUCE_SNAPSHOT_TX_STAT
}
static void
vmxnet3_rx_stats_get(struct vmxnet3_hw *hw, unsigned int q,
struct UPT1_RxStats *res)
{
vmxnet3_hw_rx_stats_get(hw, q, res);
#define VMXNET3_REDUCE_SNAPSHOT_RX_STAT(h, i, f, r) \
((r)->f -= (h)->snapshot_rx_stats[(i)].f)
VMXNET3_REDUCE_SNAPSHOT_RX_STAT(hw, q, ucastPktsRxOK, res);
VMXNET3_REDUCE_SNAPSHOT_RX_STAT(hw, q, mcastPktsRxOK, res);
VMXNET3_REDUCE_SNAPSHOT_RX_STAT(hw, q, bcastPktsRxOK, res);
VMXNET3_REDUCE_SNAPSHOT_RX_STAT(hw, q, ucastBytesRxOK, res);
VMXNET3_REDUCE_SNAPSHOT_RX_STAT(hw, q, mcastBytesRxOK, res);
VMXNET3_REDUCE_SNAPSHOT_RX_STAT(hw, q, bcastBytesRxOK, res);
VMXNET3_REDUCE_SNAPSHOT_RX_STAT(hw, q, pktsRxError, res);
VMXNET3_REDUCE_SNAPSHOT_RX_STAT(hw, q, pktsRxOutOfBuf, res);
#undef VMXNET3_REDUCE_SNAPSHOT_RX_STAT
net/vmxnet3: retain counters on restart Most NICs like virtio, igb/ixgbe etc. don't reset counters on dev_start and arguably this helps in monitoring the counters across a longer time span with multiple device start/stops. vmxnet3 behavior is opposite to that and counters are reset by the host side implementation each time the device is restarted. Change the driver to save the counters in its private context before it is reset by writing CMD_ACTIVATE to REG_CMD. Signed-off-by: Nachiketa Prachanda <nprachan@brocade.com> Acked-by: Shrikrishna Khare <skhare@vmware.com>
2017-06-15 12:16:08 +00:00
}
static void
vmxnet3_hw_stats_save(struct vmxnet3_hw *hw)
{
unsigned int i;
VMXNET3_WRITE_BAR1_REG(hw, VMXNET3_REG_CMD, VMXNET3_CMD_GET_STATS);
RTE_BUILD_BUG_ON(RTE_ETHDEV_QUEUE_STAT_CNTRS < VMXNET3_MAX_TX_QUEUES);
for (i = 0; i < hw->num_tx_queues; i++)
vmxnet3_hw_tx_stats_get(hw, i, &hw->saved_tx_stats[i]);
for (i = 0; i < hw->num_rx_queues; i++)
vmxnet3_hw_rx_stats_get(hw, i, &hw->saved_rx_stats[i]);
}
net/vmxnet3: implement extended stats Implement xstats_get() to allow a number of driver-specific Tx and Rx stats to be retrieved. Signed-off-by: Robert Shearman <rshearma@brocade.com> Acked-by: Shrikrishna Khare <skhare@vmware.com>
2017-06-15 12:16:09 +00:00
static int
vmxnet3_dev_xstats_get_names(struct rte_eth_dev *dev,
struct rte_eth_xstat_name *xstats_names,
unsigned int n)
{
unsigned int i, t, count = 0;
unsigned int nstats =
dev->data->nb_tx_queues * RTE_DIM(vmxnet3_txq_stat_strings) +
dev->data->nb_rx_queues * RTE_DIM(vmxnet3_rxq_stat_strings);
if (!xstats_names || n < nstats)
return nstats;
for (i = 0; i < dev->data->nb_rx_queues; i++) {
if (!dev->data->rx_queues[i])
continue;
for (t = 0; t < RTE_DIM(vmxnet3_rxq_stat_strings); t++) {
snprintf(xstats_names[count].name,
sizeof(xstats_names[count].name),
"rx_q%u_%s", i,
vmxnet3_rxq_stat_strings[t].name);
count++;
}
}
for (i = 0; i < dev->data->nb_tx_queues; i++) {
if (!dev->data->tx_queues[i])
continue;
for (t = 0; t < RTE_DIM(vmxnet3_txq_stat_strings); t++) {
snprintf(xstats_names[count].name,
sizeof(xstats_names[count].name),
"tx_q%u_%s", i,
vmxnet3_txq_stat_strings[t].name);
count++;
}
}
return count;
}
static int
vmxnet3_dev_xstats_get(struct rte_eth_dev *dev, struct rte_eth_xstat *xstats,
unsigned int n)
{
unsigned int i, t, count = 0;
unsigned int nstats =
dev->data->nb_tx_queues * RTE_DIM(vmxnet3_txq_stat_strings) +
dev->data->nb_rx_queues * RTE_DIM(vmxnet3_rxq_stat_strings);
if (n < nstats)
return nstats;
for (i = 0; i < dev->data->nb_rx_queues; i++) {
struct vmxnet3_rx_queue *rxq = dev->data->rx_queues[i];
if (rxq == NULL)
continue;
for (t = 0; t < RTE_DIM(vmxnet3_rxq_stat_strings); t++) {
xstats[count].value = *(uint64_t *)(((char *)&rxq->stats) +
vmxnet3_rxq_stat_strings[t].offset);
xstats[count].id = count;
count++;
}
}
for (i = 0; i < dev->data->nb_tx_queues; i++) {
struct vmxnet3_tx_queue *txq = dev->data->tx_queues[i];
if (txq == NULL)
continue;
for (t = 0; t < RTE_DIM(vmxnet3_txq_stat_strings); t++) {
xstats[count].value = *(uint64_t *)(((char *)&txq->stats) +
vmxnet3_txq_stat_strings[t].offset);
xstats[count].id = count;
count++;
}
}
return count;
}
ethdev: add return value to stats get dev op The stats_get dev op API doesn't include return value, so PMD cannot return an error in case of failure at stats getting process time. Since PCI devices can be removed and there is a time between the physical removal to the RMV interrupt, the user may get invalid stats without any indication. This patch changes the stats_get API return value to be int instead of void. All the net PMDs stats_get dev ops are adjusted by this patch. Signed-off-by: Matan Azrad <matan@mellanox.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2017-10-10 20:20:18 +00:00
static int
vmxnet3: cleanup style and indentation This driver had several style problems, the worst of which was botched indentation. Fix almost all the problems reported by checkpatch. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:37:04 +00:00
vmxnet3_dev_stats_get(struct rte_eth_dev *dev, struct rte_eth_stats *stats)
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
{
unsigned int i;
vmxnet3: remove useless adapter wrapper The adapter struct is just a wrapper around the vmxnet3_hw structure. Eliminate the wrapper and get rid of the macro used to access and needlessly cast the private data. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:39:58 +00:00
struct vmxnet3_hw *hw = dev->data->dev_private;
net/vmxnet3: retain counters on restart Most NICs like virtio, igb/ixgbe etc. don't reset counters on dev_start and arguably this helps in monitoring the counters across a longer time span with multiple device start/stops. vmxnet3 behavior is opposite to that and counters are reset by the host side implementation each time the device is restarted. Change the driver to save the counters in its private context before it is reset by writing CMD_ACTIVATE to REG_CMD. Signed-off-by: Nachiketa Prachanda <nprachan@brocade.com> Acked-by: Shrikrishna Khare <skhare@vmware.com>
2017-06-15 12:16:08 +00:00
struct UPT1_TxStats txStats;
struct UPT1_RxStats rxStats;
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
VMXNET3_WRITE_BAR1_REG(hw, VMXNET3_REG_CMD, VMXNET3_CMD_GET_STATS);
vmxnet3: add per-queue stats Update per-queue statistics and add missing multicast into statistics. Also, no need to zero statistics since they are already cleared in rte_stats_get. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:38:47 +00:00
RTE_BUILD_BUG_ON(RTE_ETHDEV_QUEUE_STAT_CNTRS < VMXNET3_MAX_TX_QUEUES);
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
for (i = 0; i < hw->num_tx_queues; i++) {
net/vmxnet3: support stats reset 'stats_reset()' callback was missing because the device backend doesn't support it. This commit adds a workaround to this and implements the callback by taking a snapshot of the stats (SNAPSHOT) each time 'stats_reset()' is called. When getting stats with 'stats_get()', hw stats which always increase reduce SNAPSHOT stats. That's how we get the "real" stats since the last 'stats_reset()'. Signed-off-by: Yogev Chaimovich <yogev@cgstowernetworks.com> Acked-by: Yong Wang <yongwang@vmware.com>
2018-09-20 06:37:05 +00:00
vmxnet3_tx_stats_get(hw, i, &txStats);
net/vmxnet3: retain counters on restart Most NICs like virtio, igb/ixgbe etc. don't reset counters on dev_start and arguably this helps in monitoring the counters across a longer time span with multiple device start/stops. vmxnet3 behavior is opposite to that and counters are reset by the host side implementation each time the device is restarted. Change the driver to save the counters in its private context before it is reset by writing CMD_ACTIVATE to REG_CMD. Signed-off-by: Nachiketa Prachanda <nprachan@brocade.com> Acked-by: Shrikrishna Khare <skhare@vmware.com>
2017-06-15 12:16:08 +00:00
stats->q_opackets[i] = txStats.ucastPktsTxOK +
txStats.mcastPktsTxOK +
txStats.bcastPktsTxOK;
vmxnet3: add per-queue stats Update per-queue statistics and add missing multicast into statistics. Also, no need to zero statistics since they are already cleared in rte_stats_get. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:38:47 +00:00
net/vmxnet3: retain counters on restart Most NICs like virtio, igb/ixgbe etc. don't reset counters on dev_start and arguably this helps in monitoring the counters across a longer time span with multiple device start/stops. vmxnet3 behavior is opposite to that and counters are reset by the host side implementation each time the device is restarted. Change the driver to save the counters in its private context before it is reset by writing CMD_ACTIVATE to REG_CMD. Signed-off-by: Nachiketa Prachanda <nprachan@brocade.com> Acked-by: Shrikrishna Khare <skhare@vmware.com>
2017-06-15 12:16:08 +00:00
stats->q_obytes[i] = txStats.ucastBytesTxOK +
txStats.mcastBytesTxOK +
txStats.bcastBytesTxOK;
vmxnet3: add per-queue stats Update per-queue statistics and add missing multicast into statistics. Also, no need to zero statistics since they are already cleared in rte_stats_get. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:38:47 +00:00
stats->opackets += stats->q_opackets[i];
stats->obytes += stats->q_obytes[i];
net/vmxnet3: retain counters on restart Most NICs like virtio, igb/ixgbe etc. don't reset counters on dev_start and arguably this helps in monitoring the counters across a longer time span with multiple device start/stops. vmxnet3 behavior is opposite to that and counters are reset by the host side implementation each time the device is restarted. Change the driver to save the counters in its private context before it is reset by writing CMD_ACTIVATE to REG_CMD. Signed-off-by: Nachiketa Prachanda <nprachan@brocade.com> Acked-by: Shrikrishna Khare <skhare@vmware.com>
2017-06-15 12:16:08 +00:00
stats->oerrors += txStats.pktsTxError + txStats.pktsTxDiscard;
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
}
vmxnet3: add per-queue stats Update per-queue statistics and add missing multicast into statistics. Also, no need to zero statistics since they are already cleared in rte_stats_get. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:38:47 +00:00
RTE_BUILD_BUG_ON(RTE_ETHDEV_QUEUE_STAT_CNTRS < VMXNET3_MAX_RX_QUEUES);
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
for (i = 0; i < hw->num_rx_queues; i++) {
net/vmxnet3: support stats reset 'stats_reset()' callback was missing because the device backend doesn't support it. This commit adds a workaround to this and implements the callback by taking a snapshot of the stats (SNAPSHOT) each time 'stats_reset()' is called. When getting stats with 'stats_get()', hw stats which always increase reduce SNAPSHOT stats. That's how we get the "real" stats since the last 'stats_reset()'. Signed-off-by: Yogev Chaimovich <yogev@cgstowernetworks.com> Acked-by: Yong Wang <yongwang@vmware.com>
2018-09-20 06:37:05 +00:00
vmxnet3_rx_stats_get(hw, i, &rxStats);
vmxnet3: add per-queue stats Update per-queue statistics and add missing multicast into statistics. Also, no need to zero statistics since they are already cleared in rte_stats_get. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:38:47 +00:00
net/vmxnet3: retain counters on restart Most NICs like virtio, igb/ixgbe etc. don't reset counters on dev_start and arguably this helps in monitoring the counters across a longer time span with multiple device start/stops. vmxnet3 behavior is opposite to that and counters are reset by the host side implementation each time the device is restarted. Change the driver to save the counters in its private context before it is reset by writing CMD_ACTIVATE to REG_CMD. Signed-off-by: Nachiketa Prachanda <nprachan@brocade.com> Acked-by: Shrikrishna Khare <skhare@vmware.com>
2017-06-15 12:16:08 +00:00
stats->q_ipackets[i] = rxStats.ucastPktsRxOK +
rxStats.mcastPktsRxOK +
rxStats.bcastPktsRxOK;
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
net/vmxnet3: retain counters on restart Most NICs like virtio, igb/ixgbe etc. don't reset counters on dev_start and arguably this helps in monitoring the counters across a longer time span with multiple device start/stops. vmxnet3 behavior is opposite to that and counters are reset by the host side implementation each time the device is restarted. Change the driver to save the counters in its private context before it is reset by writing CMD_ACTIVATE to REG_CMD. Signed-off-by: Nachiketa Prachanda <nprachan@brocade.com> Acked-by: Shrikrishna Khare <skhare@vmware.com>
2017-06-15 12:16:08 +00:00
stats->q_ibytes[i] = rxStats.ucastBytesRxOK +
rxStats.mcastBytesRxOK +
rxStats.bcastBytesRxOK;
vmxnet3: add per-queue stats Update per-queue statistics and add missing multicast into statistics. Also, no need to zero statistics since they are already cleared in rte_stats_get. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:38:47 +00:00
stats->ipackets += stats->q_ipackets[i];
stats->ibytes += stats->q_ibytes[i];
net/vmxnet3: retain counters on restart Most NICs like virtio, igb/ixgbe etc. don't reset counters on dev_start and arguably this helps in monitoring the counters across a longer time span with multiple device start/stops. vmxnet3 behavior is opposite to that and counters are reset by the host side implementation each time the device is restarted. Change the driver to save the counters in its private context before it is reset by writing CMD_ACTIVATE to REG_CMD. Signed-off-by: Nachiketa Prachanda <nprachan@brocade.com> Acked-by: Shrikrishna Khare <skhare@vmware.com>
2017-06-15 12:16:08 +00:00
stats->q_errors[i] = rxStats.pktsRxError;
stats->ierrors += rxStats.pktsRxError;
net/vmxnet3: fill imissed stat This counter comes from a "hardware" register of the vmxnet3 device and seems to behave like the MPC (Missed Packet Count) register of the Intel NICs. So I think this data belongs in the imissed field rather than the rx_nombuf field. Signed-off-by: Jon DeVree <nuxi@vault24.org> Acked-by: Yong Wang <yongwang@vmware.com>
2018-04-07 17:29:52 +00:00
stats->imissed += rxStats.pktsRxOutOfBuf;
vmxnet3: add per-queue stats Update per-queue statistics and add missing multicast into statistics. Also, no need to zero statistics since they are already cleared in rte_stats_get. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:38:47 +00:00
}
ethdev: add return value to stats get dev op The stats_get dev op API doesn't include return value, so PMD cannot return an error in case of failure at stats getting process time. Since PCI devices can be removed and there is a time between the physical removal to the RMV interrupt, the user may get invalid stats without any indication. This patch changes the stats_get API return value to be int instead of void. All the net PMDs stats_get dev ops are adjusted by this patch. Signed-off-by: Matan Azrad <matan@mellanox.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2017-10-10 20:20:18 +00:00
return 0;
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
}
ethdev: make stats and xstats reset callbacks return int Change return value of the callbacks from void to int. Make implementations across all drivers return negative errno values in case of error conditions. Both callbacks are updated together because a large number of drivers assign the same function to both callbacks. Signed-off-by: Igor Romanov <igor.romanov@oktetlabs.ru> Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2019-09-06 14:34:54 +00:00
static int
net/vmxnet3: support stats reset 'stats_reset()' callback was missing because the device backend doesn't support it. This commit adds a workaround to this and implements the callback by taking a snapshot of the stats (SNAPSHOT) each time 'stats_reset()' is called. When getting stats with 'stats_get()', hw stats which always increase reduce SNAPSHOT stats. That's how we get the "real" stats since the last 'stats_reset()'. Signed-off-by: Yogev Chaimovich <yogev@cgstowernetworks.com> Acked-by: Yong Wang <yongwang@vmware.com>
2018-09-20 06:37:05 +00:00
vmxnet3_dev_stats_reset(struct rte_eth_dev *dev)
{
unsigned int i;
struct vmxnet3_hw *hw = dev->data->dev_private;
net/vmxnet3: fix uninitialized variable Coverity issue: 323479 Fixes: 30f77abecd38 ("net/vmxnet3: support stats reset") Cc: stable@dpdk.org Signed-off-by: Yogev Chaimovich <yogev@cgstowernetworks.com> Acked-by: Yong Wang <yongwang@vmware.com>
2019-05-16 07:10:38 +00:00
struct UPT1_TxStats txStats = {0};
struct UPT1_RxStats rxStats = {0};
net/vmxnet3: support stats reset 'stats_reset()' callback was missing because the device backend doesn't support it. This commit adds a workaround to this and implements the callback by taking a snapshot of the stats (SNAPSHOT) each time 'stats_reset()' is called. When getting stats with 'stats_get()', hw stats which always increase reduce SNAPSHOT stats. That's how we get the "real" stats since the last 'stats_reset()'. Signed-off-by: Yogev Chaimovich <yogev@cgstowernetworks.com> Acked-by: Yong Wang <yongwang@vmware.com>
2018-09-20 06:37:05 +00:00
VMXNET3_WRITE_BAR1_REG(hw, VMXNET3_REG_CMD, VMXNET3_CMD_GET_STATS);
RTE_BUILD_BUG_ON(RTE_ETHDEV_QUEUE_STAT_CNTRS < VMXNET3_MAX_TX_QUEUES);
for (i = 0; i < hw->num_tx_queues; i++) {
vmxnet3_hw_tx_stats_get(hw, i, &txStats);
memcpy(&hw->snapshot_tx_stats[i], &txStats,
sizeof(hw->snapshot_tx_stats[0]));
}
for (i = 0; i < hw->num_rx_queues; i++) {
vmxnet3_hw_rx_stats_get(hw, i, &rxStats);
memcpy(&hw->snapshot_rx_stats[i], &rxStats,
sizeof(hw->snapshot_rx_stats[0]));
}
ethdev: make stats and xstats reset callbacks return int Change return value of the callbacks from void to int. Make implementations across all drivers return negative errno values in case of error conditions. Both callbacks are updated together because a large number of drivers assign the same function to both callbacks. Signed-off-by: Igor Romanov <igor.romanov@oktetlabs.ru> Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2019-09-06 14:34:54 +00:00
return 0;
net/vmxnet3: support stats reset 'stats_reset()' callback was missing because the device backend doesn't support it. This commit adds a workaround to this and implements the callback by taking a snapshot of the stats (SNAPSHOT) each time 'stats_reset()' is called. When getting stats with 'stats_get()', hw stats which always increase reduce SNAPSHOT stats. That's how we get the "real" stats since the last 'stats_reset()'. Signed-off-by: Yogev Chaimovich <yogev@cgstowernetworks.com> Acked-by: Yong Wang <yongwang@vmware.com>
2018-09-20 06:37:05 +00:00
}
ethdev: change device info get callback to return int Change eth_dev_infos_get_t return value from void to int. Make eth_dev_infos_get_t implementations across all drivers to return negative errno values if case of error conditions. Signed-off-by: Ivan Ilchenko <ivan.ilchenko@oktetlabs.ru> Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2019-09-12 16:42:28 +00:00
static int
net/vmxnet3: remove unused variable attribute No functional change, just removing unused variable tag from a used variable. Fixes: 643fba770705 ("net/vmxnet3: add v4 boot and guest UDP RSS config") Cc: stable@dpdk.org Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com> Acked-by: Yong Wang <yongwang@vmware.com>
2019-09-13 18:49:10 +00:00
vmxnet3_dev_info_get(struct rte_eth_dev *dev,
vmxnet3: announce offload capabilities Signed-off-by: Yong Wang <yongwang@vmware.com> Acked-by: Stephen Hemminger <stephen@networkplumber.org>
2016-01-13 02:08:37 +00:00
struct rte_eth_dev_info *dev_info)
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
{
net/vmxnet3: add v4 boot and guest UDP RSS config This patch introduces: - VMxnet3 v4 negotiation and, - entirely guest-driven UDP RSS support. VMxnet3 v3 already has UDP RSS support, however it depends on hypervisor provisioning on the VM through ESX specific flags, which are not transparent or known to the guest later on. Vmxnet3 v4 introduces a new API transaction which allows configuring RSS entirely from the guest. This API must be invoked after device shared mem region is initialized. IPv4 ESP RSS (SPI based) is also available, but currently there are no ESP RSS definitions on rte_eth layer to handle that. Signed-off-by: Eduard Serra <eserra@vmware.com> Acked-by: Yong Wang <yongwang@vmware.com>
2019-04-18 20:59:37 +00:00
struct vmxnet3_hw *hw = dev->data->dev_private;
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
dev_info->max_rx_queues = VMXNET3_MAX_RX_QUEUES;
dev_info->max_tx_queues = VMXNET3_MAX_TX_QUEUES;
dev_info->min_rx_bufsize = 1518 + RTE_PKTMBUF_HEADROOM;
dev_info->max_rx_pktlen = 16384; /* includes CRC, cf MAXFRS register */
net/vmxnet3: support MTU set (Picked up from @Charles Myers patch https://patchwork.dpdk.org/patch/57771/) When the mtu_set() function is not implemented, rte_eth_dev_set_mtu() fails with -ENOTSUP and mtu is not stored in the mtu field in the rte_eth_dev_data. This causes the mtu in Vmxnet3_MiscConf which is shared with hypervisor to always be set to 1500. This may cause issues receiving jumbo frames on Enhanced Data Path N-VDS. Signed-off-by: Eduard Serra <eserra@vmware.com> Acked-by: Yong Wang <yongwang@vmware.com>
2020-03-25 19:18:51 +00:00
dev_info->min_mtu = VMXNET3_MIN_MTU;
dev_info->max_mtu = VMXNET3_MAX_MTU;
net/vmxnet3: add speed capability Signed-off-by: Ido Barnea <ibarnea@cisco.com> Acked-by: Yong Wang <yongwang@vmware.com>
2017-02-02 12:08:22 +00:00
dev_info->speed_capa = ETH_LINK_SPEED_10G;
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
dev_info->max_mac_addrs = VMXNET3_MAX_MAC_ADDRS;
vmxnet3: fix default Tx configuration Since commit fbde27f19ab8f "get default Rx/Tx configuration from dev info", a default RX/TX configuration can be used for all PMDs. In case of vmxnet3, the whole structure was zeroed and not filled out. The PMD does not support multi segments or offload functions, so txq_flags should have those flags set. Test report: http://dpdk.org/ml/archives/dev/2014-December/009933.html Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com> Tested-by: Xiaonan Zhang <xiaonanx.zhang@intel.com>
2014-12-11 21:18:40 +00:00
ethdev: unification of RSS offload types RSS offload types were defined separately for 1/10G and 40G NICs, and have no relationship with flow types. The modifications are to unify all RSS offload types for all PMDs. Unified RSS offload types have new and common names which can be used for any PMD or applications, and decouple from specific hardwares. Signed-off-by: Helin Zhang <helin.zhang@intel.com> Acked-by: Jingjing Wu <jingjing.wu@intel.com> [Thomas: merge with fm10k] Signed-off-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2015-02-04 07:16:32 +00:00
dev_info->flow_type_rss_offloads = VMXNET3_RSS_OFFLOAD_ALL;
vmxnet3: get descriptor limits Signed-off-by: Konstantin Ananyev <konstantin.ananyev@intel.com> Acked-by: Yong Wang <yongwang@vmware.com> Acked-by: Remy Horton <remy.horton@intel.com>
2015-10-27 12:51:49 +00:00
net/vmxnet3: add v4 boot and guest UDP RSS config This patch introduces: - VMxnet3 v4 negotiation and, - entirely guest-driven UDP RSS support. VMxnet3 v3 already has UDP RSS support, however it depends on hypervisor provisioning on the VM through ESX specific flags, which are not transparent or known to the guest later on. Vmxnet3 v4 introduces a new API transaction which allows configuring RSS entirely from the guest. This API must be invoked after device shared mem region is initialized. IPv4 ESP RSS (SPI based) is also available, but currently there are no ESP RSS definitions on rte_eth layer to handle that. Signed-off-by: Eduard Serra <eserra@vmware.com> Acked-by: Yong Wang <yongwang@vmware.com>
2019-04-18 20:59:37 +00:00
if (VMXNET3_VERSION_GE_4(hw)) {
dev_info->flow_type_rss_offloads |= VMXNET3_V4_RSS_MASK;
}
vmxnet3: get descriptor limits Signed-off-by: Konstantin Ananyev <konstantin.ananyev@intel.com> Acked-by: Yong Wang <yongwang@vmware.com> Acked-by: Remy Horton <remy.horton@intel.com>
2015-10-27 12:51:49 +00:00
dev_info->rx_desc_lim = (struct rte_eth_desc_lim) {
.nb_max = VMXNET3_RX_RING_MAX_SIZE,
.nb_min = VMXNET3_DEF_RX_RING_SIZE,
.nb_align = 1,
};
dev_info->tx_desc_lim = (struct rte_eth_desc_lim) {
.nb_max = VMXNET3_TX_RING_MAX_SIZE,
.nb_min = VMXNET3_DEF_TX_RING_SIZE,
.nb_align = 1,
net/vmxnet3: add Tx preparation Signed-off-by: Konstantin Ananyev <konstantin.ananyev@intel.com> Acked-by: Yong Wang <yongwang@vmware.com>
2016-12-23 18:40:52 +00:00
.nb_seg_max = VMXNET3_TX_MAX_SEG,
.nb_mtu_seg_max = VMXNET3_MAX_TXD_PER_PKT,
vmxnet3: get descriptor limits Signed-off-by: Konstantin Ananyev <konstantin.ananyev@intel.com> Acked-by: Yong Wang <yongwang@vmware.com> Acked-by: Remy Horton <remy.horton@intel.com>
2015-10-27 12:51:49 +00:00
};
vmxnet3: announce offload capabilities Signed-off-by: Yong Wang <yongwang@vmware.com> Acked-by: Stephen Hemminger <stephen@networkplumber.org>
2016-01-13 02:08:37 +00:00
net/vmxnet3: convert to new Rx offload API Ethdev RX offloads API has changed since: commit ce17eddefc20 ("ethdev: introduce Rx queue offloads API") This patch adopts the new RX Offload API in vmxnet3 driver. Signed-off-by: Louis Luo <llouis@vmware.com> Acked-by: Yong Wang <yongwang@vmware.com>
2018-05-01 21:22:56 +00:00
dev_info->rx_offload_capa = VMXNET3_RX_OFFLOAD_CAP;
dev_info->rx_queue_offload_capa = 0;
dev_info->tx_offload_capa = VMXNET3_TX_OFFLOAD_CAP;
dev_info->tx_queue_offload_capa = 0;
ethdev: change device info get callback to return int Change eth_dev_infos_get_t return value from void to int. Make eth_dev_infos_get_t implementations across all drivers to return negative errno values if case of error conditions. Signed-off-by: Ivan Ilchenko <ivan.ilchenko@oktetlabs.ru> Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2019-09-12 16:42:28 +00:00
return 0;
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
}
ethdev: query supported packet types Add a new API rte_eth_dev_get_supported_ptypes to query what packet types can be filled by a given device. The device should be already started or its PMD RX burst function already decided, since the packet types supported may vary depending on RX function. Signed-off-by: Jianfeng Tan <jianfeng.tan@intel.com> Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com> Acked-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>
2016-03-14 20:50:50 +00:00
static const uint32_t *
vmxnet3_dev_supported_ptypes_get(struct rte_eth_dev *dev)
{
static const uint32_t ptypes[] = {
RTE_PTYPE_L3_IPV4_EXT,
RTE_PTYPE_L3_IPV4,
RTE_PTYPE_UNKNOWN
};
if (dev->rx_pkt_burst == vmxnet3_recv_pkts)
return ptypes;
return NULL;
}
net/vmxnet3: support MTU set (Picked up from @Charles Myers patch https://patchwork.dpdk.org/patch/57771/) When the mtu_set() function is not implemented, rte_eth_dev_set_mtu() fails with -ENOTSUP and mtu is not stored in the mtu field in the rte_eth_dev_data. This causes the mtu in Vmxnet3_MiscConf which is shared with hypervisor to always be set to 1500. This may cause issues receiving jumbo frames on Enhanced Data Path N-VDS. Signed-off-by: Eduard Serra <eserra@vmware.com> Acked-by: Yong Wang <yongwang@vmware.com>
2020-03-25 19:18:51 +00:00
static int
vmxnet3_dev_mtu_set(struct rte_eth_dev *dev, __rte_unused uint16_t mtu)
{
if (dev->data->dev_started) {
PMD_DRV_LOG(ERR, "Port %d must be stopped to configure MTU",
dev->data->port_id);
return -EBUSY;
}
return 0;
}
ethdev: return diagnostic when setting MAC address Change the prototype and the behavior of dev_ops->eth_mac_addr_set(): a return code is added to notify the caller (librte_ether) if an error occurred in the PMD. The new default MAC address is now copied in dev->data->mac_addrs[0] only if the operation is successful. The patch also updates all the PMDs accordingly. Signed-off-by: Olivier Matz <olivier.matz@6wind.com> Signed-off-by: Ivan Malov <ivan.malov@oktetlabs.ru> Acked-by: Andrew Rybchenko <arybchenko@solarflare.com> Acked-by: Adrien Mazarguil <adrien.mazarguil@6wind.com> Acked-by: Shreyansh Jain <shreyansh.jain@nxp.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com> Acked-by: Thomas Monjalon <thomas@monjalon.net> Acked-by: Nelio Laranjeiro <nelio.laranjeiro@6wind.com>
2018-04-11 16:32:51 +00:00
static int
net: add rte prefix to ether structures Add 'rte_' prefix to structures: - rename struct ether_addr as struct rte_ether_addr. - rename struct ether_hdr as struct rte_ether_hdr. - rename struct vlan_hdr as struct rte_vlan_hdr. - rename struct vxlan_hdr as struct rte_vxlan_hdr. - rename struct vxlan_gpe_hdr as struct rte_vxlan_gpe_hdr. Do not update the command line library to avoid adding a dependency to librte_net. Signed-off-by: Olivier Matz <olivier.matz@6wind.com> Reviewed-by: Stephen Hemminger <stephen@networkplumber.org> Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2019-05-21 16:13:03 +00:00
vmxnet3_mac_addr_set(struct rte_eth_dev *dev, struct rte_ether_addr *mac_addr)
vmxnet3: support setting MAC address Allow overriding the base mac address of the device. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Remy Horton <remy.horton@intel.com>
2016-03-04 18:08:02 +00:00
{
struct vmxnet3_hw *hw = dev->data->dev_private;
net: add rte prefix to ether functions Add 'rte_' prefix to functions: - rename is_same_ether_addr() as rte_is_same_ether_addr(). - rename is_zero_ether_addr() as rte_is_zero_ether_addr(). - rename is_unicast_ether_addr() as rte_is_unicast_ether_addr(). - rename is_multicast_ether_addr() as rte_is_multicast_ether_addr(). - rename is_broadcast_ether_addr() as rte_is_broadcast_ether_addr(). - rename is_universal_ether_addr() as rte_is_universal_ether_addr(). - rename is_local_admin_ether_addr() as rte_is_local_admin_ether_addr(). - rename is_valid_assigned_ether_addr() as rte_is_valid_assigned_ether_addr(). - rename eth_random_addr() as rte_eth_random_addr(). - rename ether_addr_copy() as rte_ether_addr_copy(). - rename ether_format_addr() as rte_ether_format_addr(). Signed-off-by: Olivier Matz <olivier.matz@6wind.com> Reviewed-by: Stephen Hemminger <stephen@networkplumber.org> Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2019-05-21 16:13:04 +00:00
rte_ether_addr_copy(mac_addr, (struct rte_ether_addr *)(hw->perm_addr));
vmxnet3: support setting MAC address Allow overriding the base mac address of the device. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Remy Horton <remy.horton@intel.com>
2016-03-04 18:08:02 +00:00
vmxnet3_write_mac(hw, mac_addr->addr_bytes);
ethdev: return diagnostic when setting MAC address Change the prototype and the behavior of dev_ops->eth_mac_addr_set(): a return code is added to notify the caller (librte_ether) if an error occurred in the PMD. The new default MAC address is now copied in dev->data->mac_addrs[0] only if the operation is successful. The patch also updates all the PMDs accordingly. Signed-off-by: Olivier Matz <olivier.matz@6wind.com> Signed-off-by: Ivan Malov <ivan.malov@oktetlabs.ru> Acked-by: Andrew Rybchenko <arybchenko@solarflare.com> Acked-by: Adrien Mazarguil <adrien.mazarguil@6wind.com> Acked-by: Shreyansh Jain <shreyansh.jain@nxp.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com> Acked-by: Thomas Monjalon <thomas@monjalon.net> Acked-by: Nelio Laranjeiro <nelio.laranjeiro@6wind.com>
2018-04-11 16:32:51 +00:00
return 0;
vmxnet3: support setting MAC address Allow overriding the base mac address of the device. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Remy Horton <remy.horton@intel.com>
2016-03-04 18:08:02 +00:00
}
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
/* return 0 means link status changed, -1 means not changed */
static int
net/vmxnet3: generate link-state change notifications Generate link-state change notifications by listening to interrupts generated by the device. Make use of the existing vmxnet3_process_events function that was compiled out, but change it to call vmxnet3_dev_link_update on a VMXNET3_ECR_LINK event and to not be so noisy in its log messages. Enable interrupts on starting the device, using a new helper function, vmxnet3_enable_intr, based on vmxnet3_disable_intr and validated against the FreeBSD driver. Keep track of the number of interrupts registered for to avoid hardcoding these in vmxnet3_enable/disable_intr and to provision for any future rxq intr support. Factor out the guts of vmxnet3_dev_link_update minus the started check to allow the new function to be called from vmxnet3_dev_start in the lsc-enabled case to ensure that the link state is correctly set from the actual state at that point. Signed-off-by: Robert Shearman <rshearma@brocade.com>
2017-06-15 12:16:10 +00:00
__vmxnet3_dev_link_update(struct rte_eth_dev *dev,
__rte_unused int wait_to_complete)
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
{
vmxnet3: remove useless adapter wrapper The adapter struct is just a wrapper around the vmxnet3_hw structure. Eliminate the wrapper and get rid of the macro used to access and needlessly cast the private data. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:39:58 +00:00
struct vmxnet3_hw *hw = dev->data->dev_private;
net/vmxnet3: use link status helper functions Use new rte_eth_linkstatus_get/set helper functions to handle link status update. Also remove no longer necessary include of rte_atomic.h Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2018-01-26 02:01:40 +00:00
struct rte_eth_link link;
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
uint32_t ret;
vmxnet3: fix link state handling The Intel version of VMXNET3 driver does not handle link state properly. The VMXNET3 API returns 1 if connected and 0 if disconnected. Also need to return correct value to indicate state change. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Yong Wang <yongwang@vmware.com>
2015-07-09 18:24:13 +00:00
memset(&link, 0, sizeof(link));
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
VMXNET3_WRITE_BAR1_REG(hw, VMXNET3_REG_CMD, VMXNET3_CMD_GET_LINK);
ret = VMXNET3_READ_BAR1_REG(hw, VMXNET3_REG_CMD);
net/vmxnet3: keep consistent link status Bonding may examine the link properties to ensure that matching interfaces are bound together. If the link is going to have fixed properties, these need to remain consistent regardless of the link_status or the state of the adapter. Signed-off-by: Chas Williams <chas3@att.com> Acked-by: Shrikrishna Khare <skhare@vmware.com>
2018-01-17 15:04:57 +00:00
if (ret & 0x1)
ethdev: use constants for link state Define and use ETH_LINK_UP and ETH_LINK_DOWN where appropriate. Signed-off-by: Marc Sune <marcdevel@gmail.com> Signed-off-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2016-03-31 22:12:24 +00:00
link.link_status = ETH_LINK_UP;
net/vmxnet3: keep consistent link status Bonding may examine the link properties to ensure that matching interfaces are bound together. If the link is going to have fixed properties, these need to remain consistent regardless of the link_status or the state of the adapter. Signed-off-by: Chas Williams <chas3@att.com> Acked-by: Shrikrishna Khare <skhare@vmware.com>
2018-01-17 15:04:57 +00:00
link.link_duplex = ETH_LINK_FULL_DUPLEX;
link.link_speed = ETH_SPEED_NUM_10G;
net/vmxnet3: keep link state consistent The vmxnet3 never attempts link speed negotiation. As a virtual device the link speed is vague at best. However, it is important for certain applications, like bonding, to see a consistent link_status. 802.3ad requires that only links of the same cost (link speed) be enslaved. Keeping the link status consistent in vmxnet3 avoids races with bonding enslavement. Fixes: 1e3a958f40b3 ("ethdev: fix link autonegotiation value") Cc: stable@dpdk.org Signed-off-by: Chas Williams <chas3@att.com> Acked-by: Yong Wang <yongwang@vmware.com>
2018-03-18 01:45:52 +00:00
link.link_autoneg = ETH_LINK_FIXED;
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
net/vmxnet3: use link status helper functions Use new rte_eth_linkstatus_get/set helper functions to handle link status update. Also remove no longer necessary include of rte_atomic.h Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2018-01-26 02:01:40 +00:00
return rte_eth_linkstatus_set(dev, &link);
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
}
net/vmxnet3: generate link-state change notifications Generate link-state change notifications by listening to interrupts generated by the device. Make use of the existing vmxnet3_process_events function that was compiled out, but change it to call vmxnet3_dev_link_update on a VMXNET3_ECR_LINK event and to not be so noisy in its log messages. Enable interrupts on starting the device, using a new helper function, vmxnet3_enable_intr, based on vmxnet3_disable_intr and validated against the FreeBSD driver. Keep track of the number of interrupts registered for to avoid hardcoding these in vmxnet3_enable/disable_intr and to provision for any future rxq intr support. Factor out the guts of vmxnet3_dev_link_update minus the started check to allow the new function to be called from vmxnet3_dev_start in the lsc-enabled case to ensure that the link state is correctly set from the actual state at that point. Signed-off-by: Robert Shearman <rshearma@brocade.com>
2017-06-15 12:16:10 +00:00
static int
vmxnet3_dev_link_update(struct rte_eth_dev *dev, int wait_to_complete)
{
/* Link status doesn't change for stopped dev */
if (dev->data->dev_started == 0)
return -1;
return __vmxnet3_dev_link_update(dev, wait_to_complete);
}
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
/* Updating rxmode through Vmxnet3_DriverShared structure in adapter */
static void
net/vmxnet3: coding style changes Signed-off-by: Yong Wang <yongwang@vmware.com> Acked-by: Stephen Hemminger <stephen@networkplumber.org>
2016-08-29 19:18:46 +00:00
vmxnet3_dev_set_rxmode(struct vmxnet3_hw *hw, uint32_t feature, int set)
{
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
struct Vmxnet3_RxFilterConf *rxConf = &hw->shared->devRead.rxFilterConf;
vmxnet3: cleanup style and indentation This driver had several style problems, the worst of which was botched indentation. Fix almost all the problems reported by checkpatch. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:37:04 +00:00
if (set)
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
rxConf->rxMode = rxConf->rxMode | feature;
else
rxConf->rxMode = rxConf->rxMode & (~feature);
VMXNET3_WRITE_BAR1_REG(hw, VMXNET3_REG_CMD, VMXNET3_CMD_UPDATE_RX_MODE);
}
/* Promiscuous supported only if Vmxnet3_DriverShared is initialized in adapter */
ethdev: change promiscuous callbacks to return status Enabling/disabling of promiscuous mode is not always successful and it should be taken into account to be able to handle it properly. When correct return status is unclear from driver code, -EAGAIN is used. Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Acked-by: Matan Azrad <matan@mellanox.com> Acked-by: Hyong Youb Kim <hyonkim@cisco.com>
2019-09-14 11:37:24 +00:00
static int
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
vmxnet3_dev_promiscuous_enable(struct rte_eth_dev *dev)
{
vmxnet3: remove useless adapter wrapper The adapter struct is just a wrapper around the vmxnet3_hw structure. Eliminate the wrapper and get rid of the macro used to access and needlessly cast the private data. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:39:58 +00:00
struct vmxnet3_hw *hw = dev->data->dev_private;
vmxnet3: enable vlan filtering Support the VLAN filter functionality of the VMXNET3 interface. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Yong Wang <yongwang@vmware.com>
2015-07-09 18:24:09 +00:00
uint32_t *vf_table = hw->shared->devRead.rxFilterConf.vfTable;
vmxnet3: cleanup style and indentation This driver had several style problems, the worst of which was botched indentation. Fix almost all the problems reported by checkpatch. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:37:04 +00:00
vmxnet3: enable vlan filtering Support the VLAN filter functionality of the VMXNET3 interface. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Yong Wang <yongwang@vmware.com>
2015-07-09 18:24:09 +00:00
memset(vf_table, 0, VMXNET3_VFT_TABLE_SIZE);
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
vmxnet3_dev_set_rxmode(hw, VMXNET3_RXM_PROMISC, 1);
vmxnet3: enable vlan filtering Support the VLAN filter functionality of the VMXNET3 interface. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Yong Wang <yongwang@vmware.com>
2015-07-09 18:24:09 +00:00
VMXNET3_WRITE_BAR1_REG(hw, VMXNET3_REG_CMD,
VMXNET3_CMD_UPDATE_VLAN_FILTERS);
ethdev: change promiscuous callbacks to return status Enabling/disabling of promiscuous mode is not always successful and it should be taken into account to be able to handle it properly. When correct return status is unclear from driver code, -EAGAIN is used. Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Acked-by: Matan Azrad <matan@mellanox.com> Acked-by: Hyong Youb Kim <hyonkim@cisco.com>
2019-09-14 11:37:24 +00:00
return 0;
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
}
/* Promiscuous supported only if Vmxnet3_DriverShared is initialized in adapter */
ethdev: change promiscuous callbacks to return status Enabling/disabling of promiscuous mode is not always successful and it should be taken into account to be able to handle it properly. When correct return status is unclear from driver code, -EAGAIN is used. Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Acked-by: Matan Azrad <matan@mellanox.com> Acked-by: Hyong Youb Kim <hyonkim@cisco.com>
2019-09-14 11:37:24 +00:00
static int
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
vmxnet3_dev_promiscuous_disable(struct rte_eth_dev *dev)
{
vmxnet3: remove useless adapter wrapper The adapter struct is just a wrapper around the vmxnet3_hw structure. Eliminate the wrapper and get rid of the macro used to access and needlessly cast the private data. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:39:58 +00:00
struct vmxnet3_hw *hw = dev->data->dev_private;
vmxnet3: enable vlan filtering Support the VLAN filter functionality of the VMXNET3 interface. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Yong Wang <yongwang@vmware.com>
2015-07-09 18:24:09 +00:00
uint32_t *vf_table = hw->shared->devRead.rxFilterConf.vfTable;
net/vmxnet3: convert to new Rx offload API Ethdev RX offloads API has changed since: commit ce17eddefc20 ("ethdev: introduce Rx queue offloads API") This patch adopts the new RX Offload API in vmxnet3 driver. Signed-off-by: Louis Luo <llouis@vmware.com> Acked-by: Yong Wang <yongwang@vmware.com>
2018-05-01 21:22:56 +00:00
uint64_t rx_offloads = dev->data->dev_conf.rxmode.offloads;
vmxnet3: cleanup style and indentation This driver had several style problems, the worst of which was botched indentation. Fix almost all the problems reported by checkpatch. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:37:04 +00:00
net/vmxnet3: convert to new Rx offload API Ethdev RX offloads API has changed since: commit ce17eddefc20 ("ethdev: introduce Rx queue offloads API") This patch adopts the new RX Offload API in vmxnet3 driver. Signed-off-by: Louis Luo <llouis@vmware.com> Acked-by: Yong Wang <yongwang@vmware.com>
2018-05-01 21:22:56 +00:00
if (rx_offloads & DEV_RX_OFFLOAD_VLAN_FILTER)
net/vmxnet3: fix filtering on promiscuous disabling We should only restore shadow_vfta when hw_vlan_filter is active. Otherwise, we should restore the previous filtering behavior. Fixes: f003fc383487 ("vmxnet3: enable vlan filtering") Cc: stable@dpdk.org Signed-off-by: Chas Williams <ciwillia@brocade.com> Acked-by: Shrikrishna Khare <skhare@vmware.com>
2017-07-24 14:22:42 +00:00
memcpy(vf_table, hw->shadow_vfta, VMXNET3_VFT_TABLE_SIZE);
else
memset(vf_table, 0xff, VMXNET3_VFT_TABLE_SIZE);
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
vmxnet3_dev_set_rxmode(hw, VMXNET3_RXM_PROMISC, 0);
vmxnet3: enable vlan filtering Support the VLAN filter functionality of the VMXNET3 interface. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Yong Wang <yongwang@vmware.com>
2015-07-09 18:24:09 +00:00
VMXNET3_WRITE_BAR1_REG(hw, VMXNET3_REG_CMD,
VMXNET3_CMD_UPDATE_VLAN_FILTERS);
ethdev: change promiscuous callbacks to return status Enabling/disabling of promiscuous mode is not always successful and it should be taken into account to be able to handle it properly. When correct return status is unclear from driver code, -EAGAIN is used. Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Acked-by: Matan Azrad <matan@mellanox.com> Acked-by: Hyong Youb Kim <hyonkim@cisco.com>
2019-09-14 11:37:24 +00:00
return 0;
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
}
/* Allmulticast supported only if Vmxnet3_DriverShared is initialized in adapter */
ethdev: change allmulticast callbacks to return status Enabling/disabling of allmulticast mode is not always successful and it should be taken into account to be able to handle it properly. When correct return status is unclear from driver code, -EAGAIN is used. Signed-off-by: Ivan Ilchenko <ivan.ilchenko@oktetlabs.ru> Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Acked-by: Hyong Youb Kim <hyonkim@cisco.com>
2019-09-24 12:56:10 +00:00
static int
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
vmxnet3_dev_allmulticast_enable(struct rte_eth_dev *dev)
{
vmxnet3: remove useless adapter wrapper The adapter struct is just a wrapper around the vmxnet3_hw structure. Eliminate the wrapper and get rid of the macro used to access and needlessly cast the private data. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:39:58 +00:00
struct vmxnet3_hw *hw = dev->data->dev_private;
vmxnet3: cleanup style and indentation This driver had several style problems, the worst of which was botched indentation. Fix almost all the problems reported by checkpatch. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:37:04 +00:00
vmxnet3: fix multicast enabling The driver was incorrectly enabling/disabling promiscious mode when it should have be setting/clearing all multicast mode. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:36:56 +00:00
vmxnet3_dev_set_rxmode(hw, VMXNET3_RXM_ALL_MULTI, 1);
ethdev: change allmulticast callbacks to return status Enabling/disabling of allmulticast mode is not always successful and it should be taken into account to be able to handle it properly. When correct return status is unclear from driver code, -EAGAIN is used. Signed-off-by: Ivan Ilchenko <ivan.ilchenko@oktetlabs.ru> Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Acked-by: Hyong Youb Kim <hyonkim@cisco.com>
2019-09-24 12:56:10 +00:00
return 0;
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
}
/* Allmulticast supported only if Vmxnet3_DriverShared is initialized in adapter */
ethdev: change allmulticast callbacks to return status Enabling/disabling of allmulticast mode is not always successful and it should be taken into account to be able to handle it properly. When correct return status is unclear from driver code, -EAGAIN is used. Signed-off-by: Ivan Ilchenko <ivan.ilchenko@oktetlabs.ru> Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Acked-by: Hyong Youb Kim <hyonkim@cisco.com>
2019-09-24 12:56:10 +00:00
static int
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
vmxnet3_dev_allmulticast_disable(struct rte_eth_dev *dev)
{
vmxnet3: remove useless adapter wrapper The adapter struct is just a wrapper around the vmxnet3_hw structure. Eliminate the wrapper and get rid of the macro used to access and needlessly cast the private data. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:39:58 +00:00
struct vmxnet3_hw *hw = dev->data->dev_private;
vmxnet3: cleanup style and indentation This driver had several style problems, the worst of which was botched indentation. Fix almost all the problems reported by checkpatch. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:37:04 +00:00
vmxnet3: fix multicast enabling The driver was incorrectly enabling/disabling promiscious mode when it should have be setting/clearing all multicast mode. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:36:56 +00:00
vmxnet3_dev_set_rxmode(hw, VMXNET3_RXM_ALL_MULTI, 0);
ethdev: change allmulticast callbacks to return status Enabling/disabling of allmulticast mode is not always successful and it should be taken into account to be able to handle it properly. When correct return status is unclear from driver code, -EAGAIN is used. Signed-off-by: Ivan Ilchenko <ivan.ilchenko@oktetlabs.ru> Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Acked-by: Hyong Youb Kim <hyonkim@cisco.com>
2019-09-24 12:56:10 +00:00
return 0;
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
}
vmxnet3: enable vlan filtering Support the VLAN filter functionality of the VMXNET3 interface. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Yong Wang <yongwang@vmware.com>
2015-07-09 18:24:09 +00:00
/* Enable/disable filter on vlan */
static int
vmxnet3_dev_vlan_filter_set(struct rte_eth_dev *dev, uint16_t vid, int on)
{
struct vmxnet3_hw *hw = dev->data->dev_private;
struct Vmxnet3_RxFilterConf *rxConf = &hw->shared->devRead.rxFilterConf;
uint32_t *vf_table = rxConf->vfTable;
/* save state for restore */
if (on)
VMXNET3_SET_VFTABLE_ENTRY(hw->shadow_vfta, vid);
else
VMXNET3_CLEAR_VFTABLE_ENTRY(hw->shadow_vfta, vid);
vmxnet3: clean up typos and unused code Signed-off-by: Yong Wang <yongwang@vmware.com> Acked-by: Stephen Hemminger <stephen@networkplumber.org>
2016-01-13 02:08:32 +00:00
/* don't change active filter if in promiscuous mode */
vmxnet3: enable vlan filtering Support the VLAN filter functionality of the VMXNET3 interface. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Yong Wang <yongwang@vmware.com>
2015-07-09 18:24:09 +00:00
if (rxConf->rxMode & VMXNET3_RXM_PROMISC)
return 0;
/* set in hardware */
if (on)
VMXNET3_SET_VFTABLE_ENTRY(vf_table, vid);
else
VMXNET3_CLEAR_VFTABLE_ENTRY(vf_table, vid);
VMXNET3_WRITE_BAR1_REG(hw, VMXNET3_REG_CMD,
VMXNET3_CMD_UPDATE_VLAN_FILTERS);
return 0;
}
ethdev: allow returning error on VLAN offload ops Some devices may not support or fail setting VLAN offload configuration based on dynamic circumstances so the vlan_offload_set_t vector is modified to return an int so the caller can determine success or not. rte_eth_dev_set_vlan_offload is updated to return the value provided by the vector when called along with restoring the original offload configs on failure. Existing vlan_offload_set_t vectors are modified to return an int. Majority of cases return 0 but a few that actually can fail now return their failure codes. Finally, a vlan_offload_set_t vector is added to virtio to facilitate dynamically turning VLAN strip on or off. Signed-off-by: David Harton <dharton@cisco.com> Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com>
2017-09-01 02:36:28 +00:00
static int
vmxnet3: fix VLAN filtering During an MTU change, the adapter is restarted. If hardware VLAN offload is in use, this existing filter table would also be cleared. Instead, setup the shadow table once during device initialization and just update during restart. vmxnet3_dev_vlan_offload_set(dev, mask) was incorrectly treating the mask parameter as the bitmask for vlan_strip and vlan_filter, whereas the mask indicates only what has changed - the values for vlan_stripping and vlan_filter needs to be taken from dev_conf.rxmode. Fixes: f003fc383487 ("vmxnet3: enable vlan filtering") Signed-off-by: Charles (Chas) Williams <ciwillia@brocade.com> Signed-off-by: Nachiketa Prachanda <nprachan@brocade.com> Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Remy Horton <remy.horton@intel.com>
2016-03-04 18:08:01 +00:00
vmxnet3_dev_vlan_offload_set(struct rte_eth_dev *dev, int mask)
vmxnet3: enable vlan filtering Support the VLAN filter functionality of the VMXNET3 interface. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Yong Wang <yongwang@vmware.com>
2015-07-09 18:24:09 +00:00
{
struct vmxnet3_hw *hw = dev->data->dev_private;
Vmxnet3_DSDevRead *devRead = &hw->shared->devRead;
uint32_t *vf_table = devRead->rxFilterConf.vfTable;
net/vmxnet3: convert to new Rx offload API Ethdev RX offloads API has changed since: commit ce17eddefc20 ("ethdev: introduce Rx queue offloads API") This patch adopts the new RX Offload API in vmxnet3 driver. Signed-off-by: Louis Luo <llouis@vmware.com> Acked-by: Yong Wang <yongwang@vmware.com>
2018-05-01 21:22:56 +00:00
uint64_t rx_offloads = dev->data->dev_conf.rxmode.offloads;
vmxnet3: enable vlan filtering Support the VLAN filter functionality of the VMXNET3 interface. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Yong Wang <yongwang@vmware.com>
2015-07-09 18:24:09 +00:00
vmxnet3: fix VLAN filtering During an MTU change, the adapter is restarted. If hardware VLAN offload is in use, this existing filter table would also be cleared. Instead, setup the shadow table once during device initialization and just update during restart. vmxnet3_dev_vlan_offload_set(dev, mask) was incorrectly treating the mask parameter as the bitmask for vlan_strip and vlan_filter, whereas the mask indicates only what has changed - the values for vlan_stripping and vlan_filter needs to be taken from dev_conf.rxmode. Fixes: f003fc383487 ("vmxnet3: enable vlan filtering") Signed-off-by: Charles (Chas) Williams <ciwillia@brocade.com> Signed-off-by: Nachiketa Prachanda <nprachan@brocade.com> Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Remy Horton <remy.horton@intel.com>
2016-03-04 18:08:01 +00:00
if (mask & ETH_VLAN_STRIP_MASK) {
net/vmxnet3: convert to new Rx offload API Ethdev RX offloads API has changed since: commit ce17eddefc20 ("ethdev: introduce Rx queue offloads API") This patch adopts the new RX Offload API in vmxnet3 driver. Signed-off-by: Louis Luo <llouis@vmware.com> Acked-by: Yong Wang <yongwang@vmware.com>
2018-05-01 21:22:56 +00:00
if (rx_offloads & DEV_RX_OFFLOAD_VLAN_STRIP)
vmxnet3: fix VLAN filtering During an MTU change, the adapter is restarted. If hardware VLAN offload is in use, this existing filter table would also be cleared. Instead, setup the shadow table once during device initialization and just update during restart. vmxnet3_dev_vlan_offload_set(dev, mask) was incorrectly treating the mask parameter as the bitmask for vlan_strip and vlan_filter, whereas the mask indicates only what has changed - the values for vlan_stripping and vlan_filter needs to be taken from dev_conf.rxmode. Fixes: f003fc383487 ("vmxnet3: enable vlan filtering") Signed-off-by: Charles (Chas) Williams <ciwillia@brocade.com> Signed-off-by: Nachiketa Prachanda <nprachan@brocade.com> Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Remy Horton <remy.horton@intel.com>
2016-03-04 18:08:01 +00:00
devRead->misc.uptFeatures |= UPT1_F_RXVLAN;
else
devRead->misc.uptFeatures &= ~UPT1_F_RXVLAN;
vmxnet3: enable vlan filtering Support the VLAN filter functionality of the VMXNET3 interface. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Yong Wang <yongwang@vmware.com>
2015-07-09 18:24:09 +00:00
vmxnet3: fix VLAN filtering During an MTU change, the adapter is restarted. If hardware VLAN offload is in use, this existing filter table would also be cleared. Instead, setup the shadow table once during device initialization and just update during restart. vmxnet3_dev_vlan_offload_set(dev, mask) was incorrectly treating the mask parameter as the bitmask for vlan_strip and vlan_filter, whereas the mask indicates only what has changed - the values for vlan_stripping and vlan_filter needs to be taken from dev_conf.rxmode. Fixes: f003fc383487 ("vmxnet3: enable vlan filtering") Signed-off-by: Charles (Chas) Williams <ciwillia@brocade.com> Signed-off-by: Nachiketa Prachanda <nprachan@brocade.com> Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Remy Horton <remy.horton@intel.com>
2016-03-04 18:08:01 +00:00
VMXNET3_WRITE_BAR1_REG(hw, VMXNET3_REG_CMD,
VMXNET3_CMD_UPDATE_FEATURE);
vmxnet3: enable vlan filtering Support the VLAN filter functionality of the VMXNET3 interface. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Yong Wang <yongwang@vmware.com>
2015-07-09 18:24:09 +00:00
}
vmxnet3: fix VLAN filtering During an MTU change, the adapter is restarted. If hardware VLAN offload is in use, this existing filter table would also be cleared. Instead, setup the shadow table once during device initialization and just update during restart. vmxnet3_dev_vlan_offload_set(dev, mask) was incorrectly treating the mask parameter as the bitmask for vlan_strip and vlan_filter, whereas the mask indicates only what has changed - the values for vlan_stripping and vlan_filter needs to be taken from dev_conf.rxmode. Fixes: f003fc383487 ("vmxnet3: enable vlan filtering") Signed-off-by: Charles (Chas) Williams <ciwillia@brocade.com> Signed-off-by: Nachiketa Prachanda <nprachan@brocade.com> Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Remy Horton <remy.horton@intel.com>
2016-03-04 18:08:01 +00:00
if (mask & ETH_VLAN_FILTER_MASK) {
net/vmxnet3: convert to new Rx offload API Ethdev RX offloads API has changed since: commit ce17eddefc20 ("ethdev: introduce Rx queue offloads API") This patch adopts the new RX Offload API in vmxnet3 driver. Signed-off-by: Louis Luo <llouis@vmware.com> Acked-by: Yong Wang <yongwang@vmware.com>
2018-05-01 21:22:56 +00:00
if (rx_offloads & DEV_RX_OFFLOAD_VLAN_FILTER)
vmxnet3: fix VLAN filtering During an MTU change, the adapter is restarted. If hardware VLAN offload is in use, this existing filter table would also be cleared. Instead, setup the shadow table once during device initialization and just update during restart. vmxnet3_dev_vlan_offload_set(dev, mask) was incorrectly treating the mask parameter as the bitmask for vlan_strip and vlan_filter, whereas the mask indicates only what has changed - the values for vlan_stripping and vlan_filter needs to be taken from dev_conf.rxmode. Fixes: f003fc383487 ("vmxnet3: enable vlan filtering") Signed-off-by: Charles (Chas) Williams <ciwillia@brocade.com> Signed-off-by: Nachiketa Prachanda <nprachan@brocade.com> Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Remy Horton <remy.horton@intel.com>
2016-03-04 18:08:01 +00:00
memcpy(vf_table, hw->shadow_vfta, VMXNET3_VFT_TABLE_SIZE);
else
memset(vf_table, 0xff, VMXNET3_VFT_TABLE_SIZE);
vmxnet3: enable vlan filtering Support the VLAN filter functionality of the VMXNET3 interface. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Yong Wang <yongwang@vmware.com>
2015-07-09 18:24:09 +00:00
vmxnet3: fix VLAN filtering During an MTU change, the adapter is restarted. If hardware VLAN offload is in use, this existing filter table would also be cleared. Instead, setup the shadow table once during device initialization and just update during restart. vmxnet3_dev_vlan_offload_set(dev, mask) was incorrectly treating the mask parameter as the bitmask for vlan_strip and vlan_filter, whereas the mask indicates only what has changed - the values for vlan_stripping and vlan_filter needs to be taken from dev_conf.rxmode. Fixes: f003fc383487 ("vmxnet3: enable vlan filtering") Signed-off-by: Charles (Chas) Williams <ciwillia@brocade.com> Signed-off-by: Nachiketa Prachanda <nprachan@brocade.com> Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Remy Horton <remy.horton@intel.com>
2016-03-04 18:08:01 +00:00
VMXNET3_WRITE_BAR1_REG(hw, VMXNET3_REG_CMD,
VMXNET3_CMD_UPDATE_VLAN_FILTERS);
}
ethdev: allow returning error on VLAN offload ops Some devices may not support or fail setting VLAN offload configuration based on dynamic circumstances so the vlan_offload_set_t vector is modified to return an int so the caller can determine success or not. rte_eth_dev_set_vlan_offload is updated to return the value provided by the vector when called along with restoring the original offload configs on failure. Existing vlan_offload_set_t vectors are modified to return an int. Majority of cases return 0 but a few that actually can fail now return their failure codes. Finally, a vlan_offload_set_t vector is added to virtio to facilitate dynamically turning VLAN strip on or off. Signed-off-by: David Harton <dharton@cisco.com> Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com>
2017-09-01 02:36:28 +00:00
return 0;
vmxnet3: enable vlan filtering Support the VLAN filter functionality of the VMXNET3 interface. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Yong Wang <yongwang@vmware.com>
2015-07-09 18:24:09 +00:00
}
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
static void
net/vmxnet3: generate link-state change notifications Generate link-state change notifications by listening to interrupts generated by the device. Make use of the existing vmxnet3_process_events function that was compiled out, but change it to call vmxnet3_dev_link_update on a VMXNET3_ECR_LINK event and to not be so noisy in its log messages. Enable interrupts on starting the device, using a new helper function, vmxnet3_enable_intr, based on vmxnet3_disable_intr and validated against the FreeBSD driver. Keep track of the number of interrupts registered for to avoid hardcoding these in vmxnet3_enable/disable_intr and to provision for any future rxq intr support. Factor out the guts of vmxnet3_dev_link_update minus the started check to allow the new function to be called from vmxnet3_dev_start in the lsc-enabled case to ensure that the link state is correctly set from the actual state at that point. Signed-off-by: Robert Shearman <rshearma@brocade.com>
2017-06-15 12:16:10 +00:00
vmxnet3_process_events(struct rte_eth_dev *dev)
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
{
net/vmxnet3: generate link-state change notifications Generate link-state change notifications by listening to interrupts generated by the device. Make use of the existing vmxnet3_process_events function that was compiled out, but change it to call vmxnet3_dev_link_update on a VMXNET3_ECR_LINK event and to not be so noisy in its log messages. Enable interrupts on starting the device, using a new helper function, vmxnet3_enable_intr, based on vmxnet3_disable_intr and validated against the FreeBSD driver. Keep track of the number of interrupts registered for to avoid hardcoding these in vmxnet3_enable/disable_intr and to provision for any future rxq intr support. Factor out the guts of vmxnet3_dev_link_update minus the started check to allow the new function to be called from vmxnet3_dev_start in the lsc-enabled case to ensure that the link state is correctly set from the actual state at that point. Signed-off-by: Robert Shearman <rshearma@brocade.com>
2017-06-15 12:16:10 +00:00
struct vmxnet3_hw *hw = dev->data->dev_private;
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
uint32_t events = hw->shared->ecr;
vmxnet3: cleanup style and indentation This driver had several style problems, the worst of which was botched indentation. Fix almost all the problems reported by checkpatch. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:37:04 +00:00
net/vmxnet3: make event processing less noisy Make vmxnet3_process_events less noisy by removing logging when there are no events to process and by making link, device-change and debug events DEBUG level rather than ERR. Change these to use PMD_DRV_LOG instead of PMD_INIT_LOG since they don't happen at device init. Signed-off-by: Robert Shearman <rshearma@brocade.com> Acked-by: Shrikrishna Khare <skhare@vmware.com>
2017-06-15 12:16:11 +00:00
if (!events)
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
return;
remove trailing whitespaces This commit removes trailing whitespace from lines in files. Almost all files are affected, as the BSD license copyright header had trailing whitespace on 4 lines in it [hence the number of files reporting 8 lines changed in the diffstat]. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com> Acked-by: Neil Horman <nhorman@tuxdriver.com> [Thomas: remove spaces before tabs in libs] [Thomas: remove more trailing spaces in non-C files] Signed-off-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-03 23:42:50 +00:00
/*
vmxnet3: cleanup style and indentation This driver had several style problems, the worst of which was botched indentation. Fix almost all the problems reported by checkpatch. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:37:04 +00:00
* ECR bits when written with 1b are cleared. Hence write
* events back to ECR so that the bits which were set will be reset.
*/
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
VMXNET3_WRITE_BAR1_REG(hw, VMXNET3_REG_ECR, events);
/* Check if link state has changed */
net/vmxnet3: generate link-state change notifications Generate link-state change notifications by listening to interrupts generated by the device. Make use of the existing vmxnet3_process_events function that was compiled out, but change it to call vmxnet3_dev_link_update on a VMXNET3_ECR_LINK event and to not be so noisy in its log messages. Enable interrupts on starting the device, using a new helper function, vmxnet3_enable_intr, based on vmxnet3_disable_intr and validated against the FreeBSD driver. Keep track of the number of interrupts registered for to avoid hardcoding these in vmxnet3_enable/disable_intr and to provision for any future rxq intr support. Factor out the guts of vmxnet3_dev_link_update minus the started check to allow the new function to be called from vmxnet3_dev_start in the lsc-enabled case to ensure that the link state is correctly set from the actual state at that point. Signed-off-by: Robert Shearman <rshearma@brocade.com>
2017-06-15 12:16:10 +00:00
if (events & VMXNET3_ECR_LINK) {
net/vmxnet3: make event processing less noisy Make vmxnet3_process_events less noisy by removing logging when there are no events to process and by making link, device-change and debug events DEBUG level rather than ERR. Change these to use PMD_DRV_LOG instead of PMD_INIT_LOG since they don't happen at device init. Signed-off-by: Robert Shearman <rshearma@brocade.com> Acked-by: Shrikrishna Khare <skhare@vmware.com>
2017-06-15 12:16:11 +00:00
PMD_DRV_LOG(DEBUG, "Process events: VMXNET3_ECR_LINK event");
net/vmxnet3: generate link-state change notifications Generate link-state change notifications by listening to interrupts generated by the device. Make use of the existing vmxnet3_process_events function that was compiled out, but change it to call vmxnet3_dev_link_update on a VMXNET3_ECR_LINK event and to not be so noisy in its log messages. Enable interrupts on starting the device, using a new helper function, vmxnet3_enable_intr, based on vmxnet3_disable_intr and validated against the FreeBSD driver. Keep track of the number of interrupts registered for to avoid hardcoding these in vmxnet3_enable/disable_intr and to provision for any future rxq intr support. Factor out the guts of vmxnet3_dev_link_update minus the started check to allow the new function to be called from vmxnet3_dev_start in the lsc-enabled case to ensure that the link state is correctly set from the actual state at that point. Signed-off-by: Robert Shearman <rshearma@brocade.com>
2017-06-15 12:16:10 +00:00
if (vmxnet3_dev_link_update(dev, 0) == 0)
_rte_eth_dev_callback_process(dev,
RTE_ETH_EVENT_INTR_LSC,
ethdev: remove useless parameter in callback process The pointer to the user parameter of the callback registration is automatically pass to the callback function. There is no point to allow changing this user parameter by a caller. That's why this parameter is always set to NULL by PMDs and set only in ethdev layer before calling the callback function. The history is that the user parameter was initially used by the callback implementation to pass some information between the application and the driver: c1ceaf3ad056 ("ethdev: add an argument to internal callback function") Then a new parameter has been added to leave the user parameter to its standard usage of context given at registration: d6af1a13d7a1 ("ethdev: add return values to callback process API") The NULL parameter in the internal callback processing function is now removed. It makes clear that the callback parameter is user managed and opaque from a DPDK point of view. Signed-off-by: Thomas Monjalon <thomas@monjalon.net> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2018-01-04 16:01:08 +00:00
NULL);
net/vmxnet3: generate link-state change notifications Generate link-state change notifications by listening to interrupts generated by the device. Make use of the existing vmxnet3_process_events function that was compiled out, but change it to call vmxnet3_dev_link_update on a VMXNET3_ECR_LINK event and to not be so noisy in its log messages. Enable interrupts on starting the device, using a new helper function, vmxnet3_enable_intr, based on vmxnet3_disable_intr and validated against the FreeBSD driver. Keep track of the number of interrupts registered for to avoid hardcoding these in vmxnet3_enable/disable_intr and to provision for any future rxq intr support. Factor out the guts of vmxnet3_dev_link_update minus the started check to allow the new function to be called from vmxnet3_dev_start in the lsc-enabled case to ensure that the link state is correctly set from the actual state at that point. Signed-off-by: Robert Shearman <rshearma@brocade.com>
2017-06-15 12:16:10 +00:00
}
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
/* Check if there is an error on xmit/recv queues */
if (events & (VMXNET3_ECR_TQERR | VMXNET3_ECR_RQERR)) {
net/vmxnet3: coding style changes Signed-off-by: Yong Wang <yongwang@vmware.com> Acked-by: Stephen Hemminger <stephen@networkplumber.org>
2016-08-29 19:18:46 +00:00
VMXNET3_WRITE_BAR1_REG(hw, VMXNET3_REG_CMD,
VMXNET3_CMD_GET_QUEUE_STATUS);
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
if (hw->tqd_start->status.stopped)
net/vmxnet3: make event processing less noisy Make vmxnet3_process_events less noisy by removing logging when there are no events to process and by making link, device-change and debug events DEBUG level rather than ERR. Change these to use PMD_DRV_LOG instead of PMD_INIT_LOG since they don't happen at device init. Signed-off-by: Robert Shearman <rshearma@brocade.com> Acked-by: Shrikrishna Khare <skhare@vmware.com>
2017-06-15 12:16:11 +00:00
PMD_DRV_LOG(ERR, "tq error 0x%x",
hw->tqd_start->status.error);
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
if (hw->rqd_start->status.stopped)
net/vmxnet3: make event processing less noisy Make vmxnet3_process_events less noisy by removing logging when there are no events to process and by making link, device-change and debug events DEBUG level rather than ERR. Change these to use PMD_DRV_LOG instead of PMD_INIT_LOG since they don't happen at device init. Signed-off-by: Robert Shearman <rshearma@brocade.com> Acked-by: Shrikrishna Khare <skhare@vmware.com>
2017-06-15 12:16:11 +00:00
PMD_DRV_LOG(ERR, "rq error 0x%x",
vmxnet3: cleanup style and indentation This driver had several style problems, the worst of which was botched indentation. Fix almost all the problems reported by checkpatch. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:37:04 +00:00
hw->rqd_start->status.error);
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
vmxnet3: cleanup style and indentation This driver had several style problems, the worst of which was botched indentation. Fix almost all the problems reported by checkpatch. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-13 01:37:04 +00:00
/* Reset the device */
/* Have to reset the device */
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
}
if (events & VMXNET3_ECR_DIC)
net/vmxnet3: make event processing less noisy Make vmxnet3_process_events less noisy by removing logging when there are no events to process and by making link, device-change and debug events DEBUG level rather than ERR. Change these to use PMD_DRV_LOG instead of PMD_INIT_LOG since they don't happen at device init. Signed-off-by: Robert Shearman <rshearma@brocade.com> Acked-by: Shrikrishna Khare <skhare@vmware.com>
2017-06-15 12:16:11 +00:00
PMD_DRV_LOG(DEBUG, "Device implementation change event.");
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
if (events & VMXNET3_ECR_DEBUG)
net/vmxnet3: make event processing less noisy Make vmxnet3_process_events less noisy by removing logging when there are no events to process and by making link, device-change and debug events DEBUG level rather than ERR. Change these to use PMD_DRV_LOG instead of PMD_INIT_LOG since they don't happen at device init. Signed-off-by: Robert Shearman <rshearma@brocade.com> Acked-by: Shrikrishna Khare <skhare@vmware.com>
2017-06-15 12:16:11 +00:00
PMD_DRV_LOG(DEBUG, "Debug event generated by device.");
vmxnet3: import new vmxnet3 poll mode driver implementation Poll Mode Driver for Paravirtual VMXNET3 NIC. As a PMD, the VMXNET3 driver provides the packet reception and transmission callbacks, vmxnet3_recv_pkts and vmxnet3_xmit_pkts. It does not support scattered packet reception as part of vmxnet3_recv_pkts and vmxnet3_xmit_pkts. Also, it does not support scattered packet reception as part of the device operations supported. The VMXNET3 PMD handles all the packet buffer memory allocation and resides in guest address space and it is solely responsible to free that memory when not needed. The packet buffers and features to be supported are made available to hypervisor via VMXNET3 PCI configuration space BARs. During RX/TX, the packet buffers are exchanged by their GPAs, and the hypervisor loads the buffers with packets in the RX case and sends packets to vSwitch in the TX case. The VMXNET3 PMD is compiled with vmxnet3 device headers. The interface is similar to that of the other PMDs available in the Intel(R) DPDK API. The driver pre-allocates the packet buffers and loads the command ring descriptors in advance. The hypervisor fills those packet buffers on packet arrival and write completion ring descriptors, which are eventually pulled by the PMD. After reception, the Intel(R) DPDK application frees the descriptors and loads new packet buffers for the coming packets. The interrupts are disabled and there is no notification required. This keeps performance up on the RX side, even though the device provides a notification feature. In the transmit routine, the Intel(R) DPDK application fills packet buffer pointers in the descriptors of the command ring and notifies the hypervisor. In response the hypervisor takes packets and passes them to the vSwitch. It writes into the completion descriptors ring. The rings are read by the PMD in the next transmit routine call and the buffers and descriptors are freed from memory. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2014-02-10 15:27:26 +00:00
}
net/vmxnet3: generate link-state change notifications Generate link-state change notifications by listening to interrupts generated by the device. Make use of the existing vmxnet3_process_events function that was compiled out, but change it to call vmxnet3_dev_link_update on a VMXNET3_ECR_LINK event and to not be so noisy in its log messages. Enable interrupts on starting the device, using a new helper function, vmxnet3_enable_intr, based on vmxnet3_disable_intr and validated against the FreeBSD driver. Keep track of the number of interrupts registered for to avoid hardcoding these in vmxnet3_enable/disable_intr and to provision for any future rxq intr support. Factor out the guts of vmxnet3_dev_link_update minus the started check to allow the new function to be called from vmxnet3_dev_start in the lsc-enabled case to ensure that the link state is correctly set from the actual state at that point. Signed-off-by: Robert Shearman <rshearma@brocade.com>
2017-06-15 12:16:10 +00:00
static void
vmxnet3_interrupt_handler(void *param)
{
struct rte_eth_dev *dev = param;
struct rte_pci_device *pci_dev = RTE_DEV_TO_PCI(dev->device);
vmxnet3_process_events(dev);
drivers/net: use ack in interrupt handlers Replace rte_intr_enable() with rte_intr_ack() API for acking an interrupt in interrupt handlers and rx_queue_intr_enable() callbacks of PMD's. This is inline with original intent of this change in PMDs to ack interrupts after handling is completed if device is backed by UIO, IGB_UIO or VFIO(with INTx). Signed-off-by: Nithin Dabilpuram <ndabilpuram@marvell.com> Signed-off-by: Jerin Jacob <jerinj@marvell.com> Acked-by: Shahed Shaikh <shshaikh@marvell.com> Tested-by: Shahed Shaikh <shshaikh@marvell.com> Signed-off-by: David Marchand <david.marchand@redhat.com>
2019-07-23 08:04:19 +00:00
if (rte_intr_ack(&pci_dev->intr_handle) < 0)
net/vmxnet3: generate link-state change notifications Generate link-state change notifications by listening to interrupts generated by the device. Make use of the existing vmxnet3_process_events function that was compiled out, but change it to call vmxnet3_dev_link_update on a VMXNET3_ECR_LINK event and to not be so noisy in its log messages. Enable interrupts on starting the device, using a new helper function, vmxnet3_enable_intr, based on vmxnet3_disable_intr and validated against the FreeBSD driver. Keep track of the number of interrupts registered for to avoid hardcoding these in vmxnet3_enable/disable_intr and to provision for any future rxq intr support. Factor out the guts of vmxnet3_dev_link_update minus the started check to allow the new function to be called from vmxnet3_dev_start in the lsc-enabled case to ensure that the link state is correctly set from the actual state at that point. Signed-off-by: Robert Shearman <rshearma@brocade.com>
2017-06-15 12:16:10 +00:00
PMD_DRV_LOG(ERR, "interrupt enable failed");
}
vmxnet3: convert to use of PMD_REGISTER_DRIVER and fix linking Convert the vmxnet3 pmd driver to use the PMD_REGISTER_DRIVER macro. This means that the test applications now have no reference to the vmxnet3 library when building DSO's and must specify its use on the command line with the -d option. Static linking will still initalize the driver automatically. Signed-off-by: Neil Horman <nhorman@tuxdriver.com> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-04-21 14:59:38 +00:00
drivers/net: do not use ethdev driver Signed-off-by: Jan Blunck <jblunck@infradead.org> Signed-off-by: Gaetan Rivet <gaetan.rivet@6wind.com> Acked-by: Stephen Hemminger <stephen@networkplumber.org>
2017-04-11 15:44:24 +00:00
RTE_PMD_REGISTER_PCI(net_vmxnet3, rte_vmxnet3_pmd);
drivers: rename register macro prefix All macros related to driver registeration renamed from DRIVER_* to RTE_PMD_* This includes: DRIVER_REGISTER_PCI -> RTE_PMD_REGISTER_PCI DRIVER_REGISTER_PCI_TABLE -> RTE_PMD_REGISTER_PCI_TABLE DRIVER_REGISTER_VDEV -> RTE_PMD_REGISTER_VDEV DRIVER_REGISTER_PARAM_STRING -> RTE_PMD_REGISTER_PARAM_STRING DRIVER_EXPORT_* -> RTE_PMD_EXPORT_* Fix PMDINFOGEN tool to look for matches of RTE_PMD_REGISTER_*. Signed-off-by: Shreyansh Jain <shreyansh.jain@nxp.com> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2016-10-10 05:43:15 +00:00
RTE_PMD_REGISTER_PCI_TABLE(net_vmxnet3, pci_id_vmxnet3_map);
drivers/net: fix vfio kmod dependency vfio is the kernel framework used by the vfio-pci kernel driver. DPDK drivers do not rely solely on vfio, but rather on vfio-pci to gain access to pci resources. Fixes: 0880c40113ef ("drivers: advertise kmod dependencies in pmdinfo") Cc: stable@dpdk.org Signed-off-by: David Marchand <david.marchand@6wind.com> Acked-by: Olivier Matz <olivier.matz@6wind.com>
2017-05-20 13:12:37 +00:00
RTE_PMD_REGISTER_KMOD_DEP(net_vmxnet3, "* igb_uio | uio_pci_generic | vfio-pci");
log: introduce logtype register macro Introduce the RTE_LOG_REGISTER macro to avoid the code duplication in the logtype registration process. It is a wrapper macro for declaring the logtype, registering it and setting its level in the constructor context. Signed-off-by: Jerin Jacob <jerinj@marvell.com> Acked-by: Adam Dybkowski <adamx.dybkowski@intel.com> Acked-by: Sachin Saxena <sachin.saxena@nxp.com> Acked-by: Akhil Goyal <akhil.goyal@nxp.com>
2020-07-01 12:33:35 +00:00
RTE_LOG_REGISTER(vmxnet3_logtype_init, pmd.net.vmxnet3.init, NOTICE);
RTE_LOG_REGISTER(vmxnet3_logtype_driver, pmd.net.vmxnet3.driver, NOTICE);
Reference in New Issue Copy Permalink