d/numam-dpdk
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity
015f9489c4
numam-dpdk/drivers/net/vmxnet3/vmxnet3_ethdev.c

1377 lines
40 KiB
C
Raw Normal View History

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
/*-
* BSD LICENSE
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: 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
* Copyright(c) 2010-2015 Intel Corporation. All rights reserved.
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
* All rights reserved.
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
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
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
* * Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* * Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in
* the documentation and/or other materials provided with the
* distribution.
* * Neither the name of Intel Corporation nor the names of its
* contributors may be used to endorse or promote products derived
* from this software without specific prior written permission.
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
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
* "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
* LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
* A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
* OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
* SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
* LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
* DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
* THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
* (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
* OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/
#include <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_atomic.h>
#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>
#include <rte_ethdev.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
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);
static void vmxnet3_dev_promiscuous_enable(struct rte_eth_dev *dev);
static void vmxnet3_dev_promiscuous_disable(struct rte_eth_dev *dev);
static void vmxnet3_dev_allmulticast_enable(struct rte_eth_dev *dev);
static void 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);
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);
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 vmxnet3_dev_info_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_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);
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);
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_mac_addr_set(struct rte_eth_dev *dev,
struct 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: 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,
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;
eal: deprecate rte_snprintf The function rte_snprintf serves no useful purpose. It is the same as snprintf() for all valid inputs. Deprecate it and replace all uses in current code. Leave the tests for the deprecated function in place. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2014-06-24 18:15:28 +00:00
snprintf(z_name, sizeof(z_name), "%s_%d_%s",
ethdev: remove driver name from device private data rte_driver->name has the driver name and all physical and virtual devices has access to it. Previously it was not possible for virtual ethernet devices to access rte_driver->name field (because eth_dev used to keep only pci_dev), and it was required to save driver name in the device private struct. After re-works on bus and vdev, it is possible for all bus types to access rte_driver. It is able to remove the driver name from ethdev device private data and use eth_dev->device->driver->name. Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com> Acked-by: Shreyansh Jain <shreyansh.jain@nxp.com> Acked-by: Jan Blunck <jblunck@infradead.org>
2017-06-12 15:25:12 +00:00
dev->device->driver->name, 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,
0, 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
if (mz)
return mz;
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
return rte_memzone_reserve_aligned(z_name, size, socket_id, 0, 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
}
/**
* Atomically reads the link status information from global
* structure rte_eth_dev.
*
* @param dev
* - Pointer to the structure rte_eth_dev to read from.
* - Pointer to the buffer to be saved with the link status.
*
* @return
* - On success, zero.
* - On failure, negative value.
*/
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
static int
vmxnet3_dev_atomic_read_link_status(struct rte_eth_dev *dev,
struct rte_eth_link *link)
{
struct rte_eth_link *dst = link;
struct rte_eth_link *src = &(dev->data->dev_link);
if (rte_atomic64_cmpset((uint64_t *)dst, *(uint64_t *)dst,
*(uint64_t *)src) == 0)
return -1;
return 0;
}
/**
* Atomically writes the link status information into global
* structure rte_eth_dev.
*
* @param dev
* - Pointer to the structure rte_eth_dev to write to.
* - Pointer to the buffer to be saved with the link status.
*
* @return
* - On success, zero.
* - On failure, negative value.
*/
static int
vmxnet3_dev_atomic_write_link_status(struct rte_eth_dev *dev,
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
{
struct rte_eth_link *dst = &(dev->data->dev_link);
struct rte_eth_link *src = link;
if (rte_atomic64_cmpset((uint64_t *)dst, *(uint64_t *)dst,
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
*(uint64_t *)src) == 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
return -1;
return 0;
}
/*
* 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;
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: 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
if (ver & (1 << VMXNET3_REV_3)) {
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 */
eth_dev->data->mac_addrs = rte_zmalloc("vmxnet3", 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",
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 */
ether_addr_copy((struct ether_addr *) hw->perm_addr,
&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
/* 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));
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;
if (hw->adapter_stopped == 0)
vmxnet3_dev_close(eth_dev);
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
rte_free(eth_dev->data->mac_addrs);
eth_dev->data->mac_addrs = NULL;
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();
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;
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
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
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
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 */
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.hw_ip_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: 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
if (dev->data->dev_conf.rxmode.enable_lro) {
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");
}
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) {
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 already closed.");
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: 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
hw->adapter_stopped = 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_dev_clear_queues(dev);
/* Clear recorded link status */
memset(&link, 0, sizeof(link));
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
vmxnet3_dev_atomic_write_link_status(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
}
/*
* Reset and stop device.
*/
static void
vmxnet3_dev_close(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: 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_dev_stop(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
hw->adapter_stopped = 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: 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);
#undef VMXNET3_UPDATE_RX_STATS
}
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: 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
vmxnet3_hw_tx_stats_get(hw, i, &txStats);
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: 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
vmxnet3_hw_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;
stats->rx_nombuf += 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
}
static void
ethdev: move info filling of PCI into drivers Only the drivers itself can decide if it could fill PCI information fields of dev_info. Signed-off-by: Jan Blunck <jblunck@infradead.org> Acked-by: Shreyansh Jain <shreyansh.jain@nxp.com>
2016-12-23 15:58:10 +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
{
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
dev_info->pci_dev = RTE_ETH_DEV_TO_PCI(dev);
ethdev: move info filling of PCI into drivers Only the drivers itself can decide if it could fill PCI information fields of dev_info. Signed-off-by: Jan Blunck <jblunck@infradead.org> Acked-by: Shreyansh Jain <shreyansh.jain@nxp.com>
2016-12-23 15:58:10 +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
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: 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
vmxnet3: fix Tx flags check Now that vmxnet3 supports TCP/UDP checksum offload, let's update the default txq flags to allow such offloads. Also fixed the tx queue setup check to allow TCP/UDP checksum and only error out if SCTP checksum is requested. Fixes: f598fd063bb1 ("vmxnet3: add Tx L4 checksum offload") Reported-by: Heng Ding <hengx.ding@intel.com> Signed-off-by: Yong Wang <yongwang@vmware.com>
2016-03-28 22:35:55 +00:00
dev_info->default_txconf.txq_flags = ETH_TXQ_FLAGS_NOXSUMSCTP;
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
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
dev_info->rx_offload_capa =
DEV_RX_OFFLOAD_VLAN_STRIP |
DEV_RX_OFFLOAD_UDP_CKSUM |
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
DEV_RX_OFFLOAD_TCP_CKSUM |
DEV_RX_OFFLOAD_TCP_LRO;
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
dev_info->tx_offload_capa =
DEV_TX_OFFLOAD_VLAN_INSERT |
DEV_TX_OFFLOAD_TCP_CKSUM |
DEV_TX_OFFLOAD_UDP_CKSUM |
DEV_TX_OFFLOAD_TCP_TSO;
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;
}
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_mac_addr_set(struct rte_eth_dev *dev, struct ether_addr *mac_addr)
{
struct vmxnet3_hw *hw = dev->data->dev_private;
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
ether_addr_copy(mac_addr, (struct ether_addr *)(hw->perm_addr));
ether_addr_copy(mac_addr, &dev->data->mac_addrs[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_write_mac(hw, mac_addr->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 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: fix build with gcc 7 GCC 7 flags a value as uninitialized before used. While it's a false positive, there is little harm in providing an initial value for the variable. Fixes: bb1d14b87fc3 ("vmxnet3: fix link state handling") Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2017-05-04 15:38:17 +00:00
struct rte_eth_link old = { 0 }, 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_dev_atomic_read_link_status(dev, &old);
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);
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;
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
link.link_duplex = ETH_LINK_FULL_DUPLEX;
ethdev: rename link speed constants The speed numbers ETH_LINK_SPEED_ are renamed ETH_SPEED_NUM_. The prefix ETH_LINK_SPEED_ is kept for AUTONEG and will be used for bit flags in next patch. Signed-off-by: Marc Sune <marcdevel@gmail.com>
2016-03-31 22:12:27 +00:00
link.link_speed = ETH_SPEED_NUM_10G;
ethdev: redesign link speed config This patch redesigns the API to set the link speed/s configuration of an ethernet port. Specifically: - it allows to define a set of advertised speeds for auto-negociation. - it allows to disable link auto-negociation (single fixed speed). - default: auto-negociate all supported speeds. A flag autoneg in struct rte_eth_link indicates if link speed was a result of auto-negociation or was fixed by configuration. Signed-off-by: Marc Sune <marcdevel@gmail.com> Tested-by: Nelio Laranjeiro <nelio.laranjeiro@6wind.com> Signed-off-by: Thomas Monjalon <thomas.monjalon@6wind.com> Tested-by: Beilei Xing <beilei.xing@intel.com> Tested-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>
2016-03-31 22:12:29 +00:00
link.link_autoneg = ETH_LINK_SPEED_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
}
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
vmxnet3_dev_atomic_write_link_status(dev, &link);
return (old.link_status == link.link_status) ? -1 : 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
}
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 */
static void
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);
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 */
static void
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;
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: 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
if (dev->data->dev_conf.rxmode.hw_vlan_filter)
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);
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 */
static void
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);
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 */
static void
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);
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;
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) {
if (dev->data->dev_conf.rxmode.hw_vlan_strip)
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) {
if (dev->data->dev_conf.rxmode.hw_vlan_filter)
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,
NULL, NULL);
}
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);
if (rte_intr_enable(&pci_dev->intr_handle) < 0)
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");
Reference in New Issue Copy Permalink