- 21 Feb, 2023 38 commits
-
-
Jason Wang authored
This patch introduces a new method to query the dma device that is use for a specific virtqueue. Reviewed-by:
Eli Cohen <elic@nvidia.com> Tested-by:
Jason Wang <jasowang@redhat.com> Message-Id: <20230119061525.75068-3-jasowang@redhat.com> Signed-off-by:
Michael S. Tsirkin <mst@redhat.com>
-
Jason Wang authored
This patch introduces a per virtqueue dma device. This will be used for virtio devices whose virtqueue are backed by different underlayer devices. One example is the vDPA that where the control virtqueue could be implemented through software mediation. Some of the work are actually done before since the helper like vring_dma_device(). This work left are: - Let vring_dma_device() return the per virtqueue dma device instead of the vdev's parent. - Allow passing a dma_device when creating the virtqueue through a new helper, old vring creation helper will keep using vdev's parent. Reviewed-by:
-
Liming Wu authored
"enabled" is defined in vhost_init_device_iotlb, but it is never used. Let's remove it. Signed-off-by:
Liming Wu <liming.wu@jaguarmicro.com> Message-Id: <20230110024445.303-1-liming.wu@jaguarmicro.com> Signed-off-by:
Stefano Garzarella <sgarzare@redhat.com> Reviewed-by:
Eugenio Pérez <eperezma@redhat.com>
-
Liming Wu authored
remove printk as it is meaningless. Signed-off-by:
-
Jason Wang authored
We used to (ab)use the DMA ops for setting up identical mappings in the IOTLB. This patch tries to get rid of the those unnecessary DMA ops by maintaining a simple identical/passthrough mappings by default. When bound to virtio_vdpa driver, DMA API will simply use PA as the IOVA and we will be all fine. When the vDPA bus tries to setup customized mapping (e.g when bound to vhost-vDPA), the identical/passthrough mapping will be removed. Signed-off-by:
Christoph Hellwig <hch@lst.de>
-
Jason Wang authored
This patch adds support for basic vendor stats that include counters for tx, rx and cvq. Acked-by:
-
Jason Wang authored
This patch adds a new config ops callback to allow individual simulator to implement the vendor stats callback. Acked-by:
-
Jason Wang authored
Allow individual simulator to customize the allocation size. Reviewed-by:
-
Jason Wang authored
This allows us to control the allocation size of the structure. Reviewed-by:
-
Jason Wang authored
vDPA simulators are software emulated device, so let's switch to use weak barriers to avoid extra overhead in the driver. Signed-off-by:
-
Suwan Kim authored
This patch adds completion batching to the IRQ path. It reuses batch completion code of virtblk_poll(). It collects requests to io_comp_batch and processes them all at once. It can boost up the performance by 2%. To validate the performance improvement and stabilty, I did fio test with 4 vCPU VM and 12 vCPU VM respectively. Both VMs have 8GB ram and the same number of HW queues as vCPU. The fio cammad is as follows and I ran the fio 5 times and got IOPS average. (io_uring, randread, direct=1, bs=512, iodepth=64 numjobs=2,4) Test result shows about 2% improvement. 4 vcpu VM | numjobs=2 | numjobs=4 ----------------------------------------------------------- fio without patch | 367.2K IOPS | 397.6K IOPS ----------------------------------------------------------- fio with patch | 372.8K IOPS | 407.7K IOPS 12 vcpu VM | numjobs=2 | numjobs=4 ----------------------------------------------------------- fio without patch | 363.6K IOPS | 374.8K IOPS ----------------------------------------------------------- fio with patch | 373.8K IOPS | 385.3K IOPS Signed-off-by:Suwan Kim <suwan.kim027@gmail.com> Reviewed-by:
-
Suwan Kim authored
Driver should set req->state to MQ_RQ_COMPLETE after it finishes to process req. But virtio-blk doesn't set MQ_RQ_COMPLETE after virtblk_poll() handles req and req->state still remains MQ_RQ_IN_FLIGHT. Fortunately so far there is no issue about it because blk_mq_end_request_batch() sets req->state to MQ_RQ_IDLE. In this patch, virblk_poll() calls blk_mq_complete_request_remote() to set req->state to MQ_RQ_COMPLETE before it adds req to a batch completion list. So it properly sets req->state after polling I/O is finished. Fixes: 4e040052 ("virtio-blk: support polling I/O") Signed-off-by:
-
Bagas Sanjaya authored
A sentence that checks for later spec version is meant for developers hacking the documentation source. Make it comment block (hidden from actual output). Signed-off-by:
Bagas Sanjaya <bagasdotme@gmail.com> Message-Id: <20221220095828.27588-4-bagasdotme@gmail.com> Signed-off-by:
-
Bagas Sanjaya authored
"It's at this stage that" means "At this point", so use the latter as it is more effective. Signed-off-by:
-
Bagas Sanjaya authored
Parenthesize targets to links in "References" section to distinguish them from remaining texts. While at it, describe the second target. Signed-off-by:
-
Sebastien Boeuf authored
Implement resume operation for vdpa_sim devices, so vhost-vdpa will offer that backend feature and userspace can effectively resume the device. Signed-off-by:
Sebastien Boeuf <sebastien.boeuf@intel.com> Message-Id: <15a4566826033c5dd9a2167e5cfb0ef4d90cea49.1672742878.git.sebastien.boeuf@intel.com> Signed-off-by:
-
Sebastien Boeuf authored
This new ioctl adds support for resuming the device from userspace. This is required when trying to restore the device in a functioning state after it's been suspended. It is already possible to reset a suspended device, but that means the device must be reconfigured and all the IOMMU/IOTLB mappings must be recreated. This new operation allows the device to be resumed without going through a full reset. This is particularly useful when trying to perform offline migration of a virtual machine (also known as snapshot/restore) as it allows the VMM to resume the virtual machine back to a running state after the snapshot is performed. Acked-by:
-
Sebastien Boeuf authored
Userspace knows if the device can be resumed or not by checking this feature bit. It's only exposed if the vdpa driver backend implements the resume() operation callback. Userspace trying to negotiate this feature when it hasn't been exposed will result in an error. Acked-by:
-
Sebastien Boeuf authored
Add a new operation to allow a vDPA device to be resumed after it has been suspended. Trying to resume a device that wasn't suspended will result in a no-op. This operation is optional. If it's not implemented, the associated backend feature bit will not be exposed. And if the feature bit is not exposed, invoking this operation will return an error. Acked-by:
-
Alvaro Karsz authored
This commit includes: 1) The driver to manage the controlplane over vDPA bus. 2) A HW monitor device to read health values from the DPU. Signed-off-by:
Alvaro Karsz <alvaro.karsz@solid-run.com> Acked-by:
-
Alvaro Karsz authored
This patch fixes a FLR bug on the SNET DPU rev 1 by setting the PCI_DEV_FLAGS_NO_FLR_RESET flag. As there is a quirk to avoid FLR (quirk_no_flr), I added a new quirk to check the rev ID before calling to quirk_no_flr. Without this patch, a SNET DPU rev 1 may hang when FLR is applied. Signed-off-by:
Bjorn Helgaas <bhelgaas@google.com> Message-Id: <20230110165638.123745-3-alvaro.karsz@solid-run.com> Signed-off-by:
-
Alvaro Karsz authored
Add SolidRun vendor ID to pci_ids.h The vendor ID is used in 2 different source files, the SNET vDPA driver and PCI quirks. Signed-off-by:
-
Eugenio Pérez authored
VIRTIO_NET_S_LINK_UP is already returned in config reads since vdpasim creation, but the feature bit was not offered to the driver. Tested modifying VIRTIO_NET_S_LINK_UP and different values of "status" in qemu virtio-net options, using vhost_vdpa. Not considering as a fix, because there should be no driver trusting in this config read before the feature flag. Signed-off-by:
Lei Yang <leiyang@redhat.com>
-
Zhu Lingshan authored
This commit implements features provisioning for ifcvf, that means: 1)checkk whether the provisioned features are supported by the management device 2)vDPA device only presents selected feature bits Examples: a)The management device supported features: $ vdpa mgmtdev show pci/0000:01:00.5 pci/0000:01:00.5: supported_classes net max_supported_vqs 9 dev_features MTU MAC MRG_RXBUF CTRL_VQ MQ ANY_LAYOUT VERSION_1 ACCESS_PLATFORM b)Provision a vDPA device with all supported features: $ vdpa dev add name vdpa0 mgmtdev pci/0000:01:00.5 $ vdpa/vdpa dev config show vdpa0 vdpa0: mac 00:e8:ca:11:be:05 link up link_announce false max_vq_pairs 4 mtu 1500 negotiated_features MRG_RXBUF CTRL_VQ MQ VERSION_1 ACCESS_PLATFORM c)Provision a vDPA device with a subset of the supported features: $ vdpa dev add name vdpa0 mgmtdev pci/0000:01:00.5 device_features 0x300020020 $ vdpa dev config show vdpa0 mac 00:e8:ca:11:be:05 link up link_announce false negotiated_features CTRL_VQ VERSION_1 ACCESS_PLATFORM Signed-off-by:
Zhu Lingshan <lingshan.zhu@intel.com> Message-Id: <20221125145724.1129962-13-lingshan.zhu@intel.com> Signed-off-by:
-
Zhu Lingshan authored
This commit retires ifcvf_private_to_vf, because the vf is already a member of the adapter, so it could be easily addressed by adapter->vf. Signed-off-by:
-
Zhu Lingshan authored
The adapter is the container of the vdpa_device, this commits allocate the adapter in dev_add() rather than in probe(). So that the vdpa_device() could be re-created when the userspace creates the vdpa device, and free-ed in dev_del() Signed-off-by:
-
Zhu Lingshan authored
This commit allocates the hw structure in the management device structure. So the hardware can be initialized once the management device is allocated in probe. Signed-off-by:
-
Zhu Lingshan authored
All ifcvf_request_irq's callees are refactored to work on ifcvf_hw, so it should be decoupled from the adapter as well Signed-off-by:
-
Zhu Lingshan authored
This commit decouples the config irq requester, the device shared irq requester and the MSI vectors allocator from the adapter. So they can be safely invoked since probe before the adapter is allocated. Signed-off-by:
-
Zhu Lingshan authored
This commit decouples the vq irq requester from the adapter, so that these functions can be invoked since probe. Signed-off-by:
-
Zhu Lingshan authored
This commit decouples config IRQ releaser from the adapter, so that it could be invoked once probe or in err handlers. ifcvf_free_irq() works on ifcvf_hw in this commit Signed-off-by:
-
Zhu Lingshan authored
This commit decouples the IRQ releasers from the adapter, so that these functions could be safely invoked once probe Signed-off-by:
-
Zhu Lingshan authored
This commit reverses the order of allocating the management device and the adapter. So that it would be possible to move the allocation of the adapter to dev_add(). Signed-off-by:
-
Zhu Lingshan authored
This commit decopules the config space ops from the adapter layer, so these functions can be invoked once the device is probed. Signed-off-by:
-
Zhu Lingshan authored
This commit gets rid of ifcvf_adapter in hw features related functions in ifcvf_base. Then these functions are more rubust and de-coupling from the ifcvf_adapter layer. So these functions could be invoded once the device is probed, even before the adapter is allocaed. Signed-off-by:
-
Eli Cohen authored
For each interface, either VLAN tagged or untagged, add two hardware counters: one for unicast and another for multicast. The counters count RX packets and bytes and can be read through debugfs: $ cat /sys/kernel/debug/mlx5/mlx5_core.sf.1/vdpa-0/rx/untagged/mcast/packets $ cat /sys/kernel/debug/mlx5/mlx5_core.sf.1/vdpa-0/rx/untagged/ucast/bytes This feature is controlled via the config option MLX5_VDPA_STEERING_DEBUG. It is off by default as it may have some impact on performance. includes a fixup By Yang Yingliang <yangyingliang@huawei.com>: vdpa/mlx5: fix check wrong pointer in mlx5_vdpa_add_mac_vlan_rules() The local variable 'rule' is not used anymore, fix return value check after calling mlx5_add_flow_rules(). Signed-off-by:
Yang Yingliang <yangyingliang@huawei.com> Message-Id: <20230104074418.1737510-1-yangyingliang@huawei.com> Signed-off-by:
-
Eli Cohen authored
Add debugfs subtree and expose flow table ID and TIR number. This information can be used by external tools to do extended troubleshooting. The information can be retrieved like so: $ cat /sys/kernel/debug/mlx5/mlx5_core.sf.1/vdpa-0/rx/table_id $ cat /sys/kernel/debug/mlx5/mlx5_core.sf.1/vdpa-0/rx/tirn Reviewed-by:
Si-Wei Liu <si-wei.liu@oracle.com> Acked-by:
-
Eli Cohen authored
Move some definitions from mlx5_vnet.c to newly added header file mlx5_vnet.h. We need these definitions for the following patches that add debugfs tree to expose information vital for debug. Reviewed-by:
-
- 15 Feb, 2023 2 commits
-
-
Michael S. Tsirkin authored
zone is a virtio 1.x feature so all fields are LE, they are handled as such, but have mistakenly been labeled __virtioXX in the header. This results in a bunch of sparse warnings. Use the __leXX tags to make sparse happy. Message-Id: <20221222193214.55146-1-mst@redhat.com> Signed-off-by: -
Michael S. Tsirkin authored
virtio blk returns a 64 bit append_sector in an input buffer, in LE format. This field is not tagged as LE correctly, so even though the generated code is ok, we get warnings from sparse: drivers/block/virtio_blk.c:332:33: sparse: sparse: cast to restricted __le64 Make sparse happy by using the correct type. Message-Id: <20221220125154.564265-1-mst@redhat.com> Signed-off-by:
-
