From: Wu Zongyong <wuzongyong@linux.alibaba.com>
To: Jason Wang <jasowang@redhat.com>
Cc: virtualization <virtualization@lists.linux-foundation.org>,
linux-kernel <linux-kernel@vger.kernel.org>, mst <mst@redhat.com>,
wei.yang1@linux.alibaba.com
Subject: Re: [PATCH v3 7/7] eni_vdpa: add vDPA driver for Alibaba ENI
Date: Sun, 26 Sep 2021 11:24:34 +0800 [thread overview]
Message-ID: <20210926032434.GA32570@L-PF27918B-1352.localdomain> (raw)
In-Reply-To: <CACGkMEt5rQv8DFdsYuJ6SF2YOsh_3YP_yzSsdL3X_n3Mfz3Gag@mail.gmail.com>
On Sun, Sep 26, 2021 at 10:24:21AM +0800, Jason Wang wrote:
> On Wed, Sep 22, 2021 at 8:47 PM Wu Zongyong
> <wuzongyong@linux.alibaba.com> wrote:
> >
> > This patch adds a new vDPA driver for Alibaba ENI(Elastic Network
> > Interface) which is build upon virtio 0.9.5 specification.
> > And this driver doesn't support to run on BE host.
> >
> > Signed-off-by: Wu Zongyong <wuzongyong@linux.alibaba.com>
> > ---
> > drivers/vdpa/Kconfig | 8 +
> > drivers/vdpa/Makefile | 1 +
> > drivers/vdpa/alibaba/Makefile | 3 +
> > drivers/vdpa/alibaba/eni_vdpa.c | 554 ++++++++++++++++++++++++++++++++
> > 4 files changed, 566 insertions(+)
> > create mode 100644 drivers/vdpa/alibaba/Makefile
> > create mode 100644 drivers/vdpa/alibaba/eni_vdpa.c
> >
> > diff --git a/drivers/vdpa/Kconfig b/drivers/vdpa/Kconfig
> > index 3d91982d8371..9587b9177b05 100644
> > --- a/drivers/vdpa/Kconfig
> > +++ b/drivers/vdpa/Kconfig
> > @@ -78,4 +78,12 @@ config VP_VDPA
> > help
> > This kernel module bridges virtio PCI device to vDPA bus.
> >
> > +config ALIBABA_ENI_VDPA
> > + tristate "vDPA driver for Alibaba ENI"
> > + select VIRTIO_PCI_LEGACY_LIB
> > + depends on PCI_MSI
> > + help
> > + VDPA driver for Alibaba ENI(Elastic Network Interface) which is build upon
> > + virtio 0.9.5 specification.
> > +
> > endif # VDPA
> > diff --git a/drivers/vdpa/Makefile b/drivers/vdpa/Makefile
> > index f02ebed33f19..15665563a7f4 100644
> > --- a/drivers/vdpa/Makefile
> > +++ b/drivers/vdpa/Makefile
> > @@ -5,3 +5,4 @@ obj-$(CONFIG_VDPA_USER) += vdpa_user/
> > obj-$(CONFIG_IFCVF) += ifcvf/
> > obj-$(CONFIG_MLX5_VDPA) += mlx5/
> > obj-$(CONFIG_VP_VDPA) += virtio_pci/
> > +obj-$(CONFIG_ALIBABA_ENI_VDPA) += alibaba/
> > diff --git a/drivers/vdpa/alibaba/Makefile b/drivers/vdpa/alibaba/Makefile
> > new file mode 100644
> > index 000000000000..ef4aae69f87a
> > --- /dev/null
> > +++ b/drivers/vdpa/alibaba/Makefile
> > @@ -0,0 +1,3 @@
> > +# SPDX-License-Identifier: GPL-2.0
> > +obj-$(CONFIG_ALIBABA_ENI_VDPA) += eni_vdpa.o
> > +
> > diff --git a/drivers/vdpa/alibaba/eni_vdpa.c b/drivers/vdpa/alibaba/eni_vdpa.c
> > new file mode 100644
> > index 000000000000..b6eef696cec5
> > --- /dev/null
> > +++ b/drivers/vdpa/alibaba/eni_vdpa.c
> > @@ -0,0 +1,554 @@
> > +// SPDX-License-Identifier: GPL-2.0-only
> > +/*
> > + * vDPA bridge driver for Alibaba ENI(Elastic Network Interface)
> > + *
> > + * Copyright (c) 2021, Alibaba Inc. All rights reserved.
> > + * Author: Wu Zongyong <wuzongyong@linux.alibaba.com>
> > + *
> > + */
> > +
> > +#include "linux/bits.h"
> > +#include <linux/interrupt.h>
> > +#include <linux/module.h>
> > +#include <linux/pci.h>
> > +#include <linux/vdpa.h>
> > +#include <linux/virtio.h>
> > +#include <linux/virtio_config.h>
> > +#include <linux/virtio_ring.h>
> > +#include <linux/virtio_pci.h>
> > +#include <linux/virtio_pci_legacy.h>
> > +#include <uapi/linux/virtio_net.h>
> > +
> > +#define ENI_MSIX_NAME_SIZE 256
> > +
> > +#define ENI_ERR(pdev, fmt, ...) \
> > + dev_err(&pdev->dev, "%s"fmt, "eni_vdpa: ", ##__VA_ARGS__)
> > +#define ENI_DBG(pdev, fmt, ...) \
> > + dev_dbg(&pdev->dev, "%s"fmt, "eni_vdpa: ", ##__VA_ARGS__)
> > +#define ENI_INFO(pdev, fmt, ...) \
> > + dev_info(&pdev->dev, "%s"fmt, "eni_vdpa: ", ##__VA_ARGS__)
> > +
> > +struct eni_vring {
> > + void __iomem *notify;
> > + char msix_name[ENI_MSIX_NAME_SIZE];
> > + struct vdpa_callback cb;
> > + int irq;
> > +};
> > +
> > +struct eni_vdpa {
> > + struct vdpa_device vdpa;
> > + struct virtio_pci_legacy_device ldev;
> > + struct eni_vring *vring;
> > + struct vdpa_callback config_cb;
> > + char msix_name[ENI_MSIX_NAME_SIZE];
> > + int config_irq;
> > + int queues;
> > + int vectors;
> > +};
> > +
> > +static struct eni_vdpa *vdpa_to_eni(struct vdpa_device *vdpa)
> > +{
> > + return container_of(vdpa, struct eni_vdpa, vdpa);
> > +}
> > +
> > +static struct virtio_pci_legacy_device *vdpa_to_ldev(struct vdpa_device *vdpa)
> > +{
> > + struct eni_vdpa *eni_vdpa = vdpa_to_eni(vdpa);
> > +
> > + return &eni_vdpa->ldev;
> > +}
> > +
> > +static u64 eni_vdpa_get_features(struct vdpa_device *vdpa)
> > +{
> > + struct virtio_pci_legacy_device *ldev = vdpa_to_ldev(vdpa);
> > + u64 features = vp_legacy_get_features(ldev);
> > +
> > + features |= BIT_ULL(VIRTIO_F_ACCESS_PLATFORM);
> > +
> > + return features;
>
> I wonder if the following can work with ENI:
>
> -device virtio-net-pci,mrg_rxbuf=off
>
> ?
ENI didn't work.
I will remove F_MRG_RXBUF when get_features.
>
> Thanks
>
> > +}
> > +
> > +static int eni_vdpa_set_features(struct vdpa_device *vdpa, u64 features)
> > +{
> > + struct virtio_pci_legacy_device *ldev = vdpa_to_ldev(vdpa);
> > +
> > + vp_legacy_set_features(ldev, (u32)features);
> > +
> > + return 0;
> > +}
> > +
> > +static u8 eni_vdpa_get_status(struct vdpa_device *vdpa)
> > +{
> > + struct virtio_pci_legacy_device *ldev = vdpa_to_ldev(vdpa);
> > +
> > + return vp_legacy_get_status(ldev);
> > +}
> > +
> > +static int eni_vdpa_get_vq_irq(struct vdpa_device *vdpa, u16 idx)
> > +{
> > + struct eni_vdpa *eni_vdpa = vdpa_to_eni(vdpa);
> > + int irq = eni_vdpa->vring[idx].irq;
> > +
> > + if (irq == VIRTIO_MSI_NO_VECTOR)
> > + return -EINVAL;
> > +
> > + return irq;
> > +}
> > +
> > +static void eni_vdpa_free_irq(struct eni_vdpa *eni_vdpa)
> > +{
> > + struct virtio_pci_legacy_device *ldev = &eni_vdpa->ldev;
> > + struct pci_dev *pdev = ldev->pci_dev;
> > + int i;
> > +
> > + for (i = 0; i < eni_vdpa->queues; i++) {
> > + if (eni_vdpa->vring[i].irq != VIRTIO_MSI_NO_VECTOR) {
> > + vp_legacy_queue_vector(ldev, i, VIRTIO_MSI_NO_VECTOR);
> > + devm_free_irq(&pdev->dev, eni_vdpa->vring[i].irq,
> > + &eni_vdpa->vring[i]);
> > + eni_vdpa->vring[i].irq = VIRTIO_MSI_NO_VECTOR;
> > + }
> > + }
> > +
> > + if (eni_vdpa->config_irq != VIRTIO_MSI_NO_VECTOR) {
> > + vp_legacy_config_vector(ldev, VIRTIO_MSI_NO_VECTOR);
> > + devm_free_irq(&pdev->dev, eni_vdpa->config_irq, eni_vdpa);
> > + eni_vdpa->config_irq = VIRTIO_MSI_NO_VECTOR;
> > + }
> > +
> > + if (eni_vdpa->vectors) {
> > + pci_free_irq_vectors(pdev);
> > + eni_vdpa->vectors = 0;
> > + }
> > +}
> > +
> > +static irqreturn_t eni_vdpa_vq_handler(int irq, void *arg)
> > +{
> > + struct eni_vring *vring = arg;
> > +
> > + if (vring->cb.callback)
> > + return vring->cb.callback(vring->cb.private);
> > +
> > + return IRQ_HANDLED;
> > +}
> > +
> > +static irqreturn_t eni_vdpa_config_handler(int irq, void *arg)
> > +{
> > + struct eni_vdpa *eni_vdpa = arg;
> > +
> > + if (eni_vdpa->config_cb.callback)
> > + return eni_vdpa->config_cb.callback(eni_vdpa->config_cb.private);
> > +
> > + return IRQ_HANDLED;
> > +}
> > +
> > +static int eni_vdpa_request_irq(struct eni_vdpa *eni_vdpa)
> > +{
> > + struct virtio_pci_legacy_device *ldev = &eni_vdpa->ldev;
> > + struct pci_dev *pdev = ldev->pci_dev;
> > + int i, ret, irq;
> > + int queues = eni_vdpa->queues;
> > + int vectors = queues + 1;
> > +
> > + ret = pci_alloc_irq_vectors(pdev, vectors, vectors, PCI_IRQ_MSIX);
> > + if (ret != vectors) {
> > + ENI_ERR(pdev,
> > + "failed to allocate irq vectors want %d but %d\n",
> > + vectors, ret);
> > + return ret;
> > + }
> > +
> > + eni_vdpa->vectors = vectors;
> > +
> > + for (i = 0; i < queues; i++) {
> > + snprintf(eni_vdpa->vring[i].msix_name, ENI_MSIX_NAME_SIZE,
> > + "eni-vdpa[%s]-%d\n", pci_name(pdev), i);
> > + irq = pci_irq_vector(pdev, i);
> > + ret = devm_request_irq(&pdev->dev, irq,
> > + eni_vdpa_vq_handler,
> > + 0, eni_vdpa->vring[i].msix_name,
> > + &eni_vdpa->vring[i]);
> > + if (ret) {
> > + ENI_ERR(pdev, "failed to request irq for vq %d\n", i);
> > + goto err;
> > + }
> > + vp_legacy_queue_vector(ldev, i, i);
> > + eni_vdpa->vring[i].irq = irq;
> > + }
> > +
> > + snprintf(eni_vdpa->msix_name, ENI_MSIX_NAME_SIZE, "eni-vdpa[%s]-config\n",
> > + pci_name(pdev));
> > + irq = pci_irq_vector(pdev, queues);
> > + ret = devm_request_irq(&pdev->dev, irq, eni_vdpa_config_handler, 0,
> > + eni_vdpa->msix_name, eni_vdpa);
> > + if (ret) {
> > + ENI_ERR(pdev, "failed to request irq for config vq %d\n", i);
> > + goto err;
> > + }
> > + vp_legacy_config_vector(ldev, queues);
> > + eni_vdpa->config_irq = irq;
> > +
> > + return 0;
> > +err:
> > + eni_vdpa_free_irq(eni_vdpa);
> > + return ret;
> > +}
> > +
> > +static void eni_vdpa_set_status(struct vdpa_device *vdpa, u8 status)
> > +{
> > + struct eni_vdpa *eni_vdpa = vdpa_to_eni(vdpa);
> > + struct virtio_pci_legacy_device *ldev = &eni_vdpa->ldev;
> > + u8 s = eni_vdpa_get_status(vdpa);
> > +
> > + if (status & VIRTIO_CONFIG_S_DRIVER_OK &&
> > + !(s & VIRTIO_CONFIG_S_DRIVER_OK)) {
> > + eni_vdpa_request_irq(eni_vdpa);
> > + }
> > +
> > + vp_legacy_set_status(ldev, status);
> > +
> > + if (!(status & VIRTIO_CONFIG_S_DRIVER_OK) &&
> > + (s & VIRTIO_CONFIG_S_DRIVER_OK))
> > + eni_vdpa_free_irq(eni_vdpa);
> > +}
> > +
> > +static int eni_vdpa_reset(struct vdpa_device *vdpa)
> > +{
> > + struct eni_vdpa *eni_vdpa = vdpa_to_eni(vdpa);
> > + struct virtio_pci_legacy_device *ldev = &eni_vdpa->ldev;
> > + u8 s = eni_vdpa_get_status(vdpa);
> > +
> > + vp_legacy_set_status(ldev, 0);
> > +
> > + if (s & VIRTIO_CONFIG_S_DRIVER_OK)
> > + eni_vdpa_free_irq(eni_vdpa);
> > +
> > + return 0;
> > +}
> > +
> > +static u16 eni_vdpa_get_vq_num_max(struct vdpa_device *vdpa)
> > +{
> > + struct virtio_pci_legacy_device *ldev = vdpa_to_ldev(vdpa);
> > +
> > + return vp_legacy_get_queue_size(ldev, 0);
> > +}
> > +
> > +static u16 eni_vdpa_get_vq_num_min(struct vdpa_device *vdpa)
> > +{
> > + struct virtio_pci_legacy_device *ldev = vdpa_to_ldev(vdpa);
> > +
> > + return vp_legacy_get_queue_size(ldev, 0);
> > +}
> > +
> > +static int eni_vdpa_get_vq_state(struct vdpa_device *vdpa, u16 qid,
> > + struct vdpa_vq_state *state)
> > +{
> > + return -EOPNOTSUPP;
> > +}
> > +
> > +static int eni_vdpa_set_vq_state(struct vdpa_device *vdpa, u16 qid,
> > + const struct vdpa_vq_state *state)
> > +{
> > + struct virtio_pci_legacy_device *ldev = vdpa_to_ldev(vdpa);
> > + const struct vdpa_vq_state_split *split = &state->split;
> > +
> > + /* ENI is build upon virtio-pci specfication which not support
> > + * to set state of virtqueue. But if the state is equal to the
> > + * device initial state by chance, we can let it go.
> > + */
> > + if (!vp_legacy_get_queue_enable(ldev, qid)
> > + && split->avail_index == 0)
> > + return 0;
> > +
> > + return -EOPNOTSUPP;
> > +}
> > +
> > +
> > +static void eni_vdpa_set_vq_cb(struct vdpa_device *vdpa, u16 qid,
> > + struct vdpa_callback *cb)
> > +{
> > + struct eni_vdpa *eni_vdpa = vdpa_to_eni(vdpa);
> > +
> > + eni_vdpa->vring[qid].cb = *cb;
> > +}
> > +
> > +static void eni_vdpa_set_vq_ready(struct vdpa_device *vdpa, u16 qid,
> > + bool ready)
> > +{
> > + struct virtio_pci_legacy_device *ldev = vdpa_to_ldev(vdpa);
> > +
> > + /* ENI is a legacy virtio-pci device. This is not supported
> > + * by specification. But we can disable virtqueue by setting
> > + * address to 0.
> > + */
> > + if (!ready)
> > + vp_legacy_set_queue_address(ldev, qid, 0);
> > +}
> > +
> > +static bool eni_vdpa_get_vq_ready(struct vdpa_device *vdpa, u16 qid)
> > +{
> > + struct virtio_pci_legacy_device *ldev = vdpa_to_ldev(vdpa);
> > +
> > + return vp_legacy_get_queue_enable(ldev, qid);
> > +}
> > +
> > +static void eni_vdpa_set_vq_num(struct vdpa_device *vdpa, u16 qid,
> > + u32 num)
> > +{
> > + struct virtio_pci_legacy_device *ldev = vdpa_to_ldev(vdpa);
> > + struct pci_dev *pdev = ldev->pci_dev;
> > + u16 n = vp_legacy_get_queue_size(ldev, qid);
> > +
> > + /* ENI is a legacy virtio-pci device which not allow to change
> > + * virtqueue size. Just report a error if someone tries to
> > + * change it.
> > + */
> > + if (num != n)
> > + ENI_ERR(pdev,
> > + "not support to set vq %u fixed num %u to %u\n",
> > + qid, n, num);
> > +}
> > +
> > +static int eni_vdpa_set_vq_address(struct vdpa_device *vdpa, u16 qid,
> > + u64 desc_area, u64 driver_area,
> > + u64 device_area)
> > +{
> > + struct virtio_pci_legacy_device *ldev = vdpa_to_ldev(vdpa);
> > + u32 pfn = desc_area >> VIRTIO_PCI_QUEUE_ADDR_SHIFT;
> > +
> > + vp_legacy_set_queue_address(ldev, qid, pfn);
> > +
> > + return 0;
> > +}
> > +
> > +static void eni_vdpa_kick_vq(struct vdpa_device *vdpa, u16 qid)
> > +{
> > + struct eni_vdpa *eni_vdpa = vdpa_to_eni(vdpa);
> > +
> > + iowrite16(qid, eni_vdpa->vring[qid].notify);
> > +}
> > +
> > +static u32 eni_vdpa_get_device_id(struct vdpa_device *vdpa)
> > +{
> > + struct virtio_pci_legacy_device *ldev = vdpa_to_ldev(vdpa);
> > +
> > + return ldev->id.device;
> > +}
> > +
> > +static u32 eni_vdpa_get_vendor_id(struct vdpa_device *vdpa)
> > +{
> > + struct virtio_pci_legacy_device *ldev = vdpa_to_ldev(vdpa);
> > +
> > + return ldev->id.vendor;
> > +}
> > +
> > +static u32 eni_vdpa_get_vq_align(struct vdpa_device *vdpa)
> > +{
> > + return PAGE_SIZE;
> > +}
> > +
> > +static size_t eni_vdpa_get_config_size(struct vdpa_device *vdpa)
> > +{
> > + return sizeof(struct virtio_net_config);
> > +}
> > +
> > +
> > +static void eni_vdpa_get_config(struct vdpa_device *vdpa,
> > + unsigned int offset,
> > + void *buf, unsigned int len)
> > +{
> > + struct eni_vdpa *eni_vdpa = vdpa_to_eni(vdpa);
> > + struct virtio_pci_legacy_device *ldev = &eni_vdpa->ldev;
> > + void __iomem *ioaddr = ldev->ioaddr +
> > + VIRTIO_PCI_CONFIG_OFF(eni_vdpa->vectors) +
> > + offset;
> > + u8 *p = buf;
> > + int i;
> > +
> > + for (i = 0; i < len; i++)
> > + *p++ = ioread8(ioaddr + i);
> > +}
> > +
> > +static void eni_vdpa_set_config(struct vdpa_device *vdpa,
> > + unsigned int offset, const void *buf,
> > + unsigned int len)
> > +{
> > + struct eni_vdpa *eni_vdpa = vdpa_to_eni(vdpa);
> > + struct virtio_pci_legacy_device *ldev = &eni_vdpa->ldev;
> > + void __iomem *ioaddr = ldev->ioaddr +
> > + VIRTIO_PCI_CONFIG_OFF(eni_vdpa->vectors) +
> > + offset;
> > + const u8 *p = buf;
> > + int i;
> > +
> > + for (i = 0; i < len; i++)
> > + iowrite8(*p++, ioaddr + i);
> > +}
> > +
> > +static void eni_vdpa_set_config_cb(struct vdpa_device *vdpa,
> > + struct vdpa_callback *cb)
> > +{
> > + struct eni_vdpa *eni_vdpa = vdpa_to_eni(vdpa);
> > +
> > + eni_vdpa->config_cb = *cb;
> > +}
> > +
> > +static const struct vdpa_config_ops eni_vdpa_ops = {
> > + .get_features = eni_vdpa_get_features,
> > + .set_features = eni_vdpa_set_features,
> > + .get_status = eni_vdpa_get_status,
> > + .set_status = eni_vdpa_set_status,
> > + .reset = eni_vdpa_reset,
> > + .get_vq_num_max = eni_vdpa_get_vq_num_max,
> > + .get_vq_num_min = eni_vdpa_get_vq_num_min,
> > + .get_vq_state = eni_vdpa_get_vq_state,
> > + .set_vq_state = eni_vdpa_set_vq_state,
> > + .set_vq_cb = eni_vdpa_set_vq_cb,
> > + .set_vq_ready = eni_vdpa_set_vq_ready,
> > + .get_vq_ready = eni_vdpa_get_vq_ready,
> > + .set_vq_num = eni_vdpa_set_vq_num,
> > + .set_vq_address = eni_vdpa_set_vq_address,
> > + .kick_vq = eni_vdpa_kick_vq,
> > + .get_device_id = eni_vdpa_get_device_id,
> > + .get_vendor_id = eni_vdpa_get_vendor_id,
> > + .get_vq_align = eni_vdpa_get_vq_align,
> > + .get_config_size = eni_vdpa_get_config_size,
> > + .get_config = eni_vdpa_get_config,
> > + .set_config = eni_vdpa_set_config,
> > + .set_config_cb = eni_vdpa_set_config_cb,
> > + .get_vq_irq = eni_vdpa_get_vq_irq,
> > +};
> > +
> > +
> > +static u16 eni_vdpa_get_num_queues(struct eni_vdpa *eni_vdpa)
> > +{
> > + struct virtio_pci_legacy_device *ldev = &eni_vdpa->ldev;
> > + u32 features = vp_legacy_get_features(ldev);
> > + u16 num = 2;
> > +
> > + if (features & BIT_ULL(VIRTIO_NET_F_MQ)) {
> > + __virtio16 max_virtqueue_pairs;
> > +
> > + eni_vdpa_get_config(&eni_vdpa->vdpa,
> > + offsetof(struct virtio_net_config, max_virtqueue_pairs),
> > + &max_virtqueue_pairs,
> > + sizeof(max_virtqueue_pairs));
> > + num = 2 * __virtio16_to_cpu(virtio_legacy_is_little_endian(),
> > + max_virtqueue_pairs);
> > + }
> > +
> > + if (features & BIT_ULL(VIRTIO_NET_F_CTRL_VQ))
> > + num += 1;
> > +
> > + return num;
> > +}
> > +
> > +static void eni_vdpa_free_irq_vectors(void *data)
> > +{
> > + pci_free_irq_vectors(data);
> > +}
> > +
> > +#ifdef __LITTLE_ENDIAN
> > +static int eni_vdpa_probe(struct pci_dev *pdev, const struct pci_device_id *id)
> > +{
> > + struct device *dev = &pdev->dev;
> > + struct eni_vdpa *eni_vdpa;
> > + struct virtio_pci_legacy_device *ldev;
> > + int ret, i;
> > +
> > + ret = pcim_enable_device(pdev);
> > + if (ret)
> > + return ret;
> > +
> > + eni_vdpa = vdpa_alloc_device(struct eni_vdpa, vdpa,
> > + dev, &eni_vdpa_ops, NULL, false);
> > + if (IS_ERR(eni_vdpa)) {
> > + ENI_ERR(pdev, "failed to allocate vDPA structure\n");
> > + return PTR_ERR(eni_vdpa);
> > + }
> > +
> > + ldev = &eni_vdpa->ldev;
> > + ldev->pci_dev = pdev;
> > +
> > + ret = vp_legacy_probe(ldev);
> > + if (ret) {
> > + ENI_ERR(pdev, "failed to probe legacy PCI device\n");
> > + goto err;
> > + }
> > +
> > + pci_set_master(pdev);
> > + pci_set_drvdata(pdev, eni_vdpa);
> > +
> > + eni_vdpa->vdpa.dma_dev = &pdev->dev;
> > + eni_vdpa->queues = eni_vdpa_get_num_queues(eni_vdpa);
> > +
> > + ret = devm_add_action_or_reset(dev, eni_vdpa_free_irq_vectors, pdev);
> > + if (ret) {
> > + ENI_ERR(pdev,
> > + "failed for adding devres for freeing irq vectors\n");
> > + goto err;
> > + }
> > +
> > + eni_vdpa->vring = devm_kcalloc(&pdev->dev, eni_vdpa->queues,
> > + sizeof(*eni_vdpa->vring),
> > + GFP_KERNEL);
> > + if (!eni_vdpa->vring) {
> > + ret = -ENOMEM;
> > + ENI_ERR(pdev, "failed to allocate virtqueues\n");
> > + goto err;
> > + }
> > +
> > + for (i = 0; i < eni_vdpa->queues; i++) {
> > + eni_vdpa->vring[i].irq = VIRTIO_MSI_NO_VECTOR;
> > + eni_vdpa->vring[i].notify = ldev->ioaddr + VIRTIO_PCI_QUEUE_NOTIFY;
> > + }
> > + eni_vdpa->config_irq = VIRTIO_MSI_NO_VECTOR;
> > +
> > + ret = vdpa_register_device(&eni_vdpa->vdpa, eni_vdpa->queues);
> > + if (ret) {
> > + ENI_ERR(pdev, "failed to register to vdpa bus\n");
> > + goto err;
> > + }
> > +
> > + return 0;
> > +
> > +err:
> > + put_device(&eni_vdpa->vdpa.dev);
> > + return ret;
> > +}
> > +#else
> > +static int eni_vdpa_probe(struct pci_dev *pdev, const struct pci_device_id *id)
> > +{
> > + ENI_ERR(pdev, "this driver not supported on BE host\n");
> > + return -ENODEV;
> > +}
> > +#endif
> > +
> > +static void eni_vdpa_remove(struct pci_dev *pdev)
> > +{
> > + struct eni_vdpa *eni_vdpa = pci_get_drvdata(pdev);
> > +
> > + vdpa_unregister_device(&eni_vdpa->vdpa);
> > + vp_legacy_remove(&eni_vdpa->ldev);
> > +}
> > +
> > +static struct pci_device_id eni_pci_ids[] = {
> > + { PCI_DEVICE_SUB(PCI_VENDOR_ID_REDHAT_QUMRANET,
> > + VIRTIO_TRANS_ID_NET,
> > + PCI_SUBVENDOR_ID_REDHAT_QUMRANET,
> > + VIRTIO_ID_NET) },
> > + { 0 },
> > +};
> > +
> > +static struct pci_driver eni_vdpa_driver = {
> > + .name = "alibaba-eni-vdpa",
> > + .id_table = eni_pci_ids,
> > + .probe = eni_vdpa_probe,
> > + .remove = eni_vdpa_remove,
> > +};
> > +
> > +module_pci_driver(eni_vdpa_driver);
> > +
> > +MODULE_AUTHOR("Wu Zongyong <wuzongyong@linux.alibaba.com>");
> > +MODULE_DESCRIPTION("Alibaba ENI vDPA driver");
> > +MODULE_LICENSE("GPL v2");
> > --
> > 2.31.1
> >
next prev parent reply other threads:[~2021-09-26 3:24 UTC|newest]
Thread overview: 151+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-09-08 12:20 [PATCH 0/6] vDPA driver for legacy virtio-pci device Wu Zongyong
2021-09-08 12:20 ` [PATCH 1/6] virtio-pci: introduce legacy device module Wu Zongyong
2021-09-08 12:20 ` [PATCH 2/6] vdpa: fix typo Wu Zongyong
2021-09-08 12:20 ` [PATCH 3/6] vp_vdpa: add vq irq offloading support Wu Zongyong
2021-09-09 2:53 ` Jason Wang
2021-09-08 12:20 ` [PATCH 4/6] vp_vdpa: split out reusable and device specific codes to separate file Wu Zongyong
2021-09-08 12:20 ` [PATCH 5/6] vdpa: add get_vq_num_unchangeable callback in vdpa_config_ops Wu Zongyong
2021-09-09 2:55 ` Jason Wang
2021-09-09 8:01 ` Wu Zongyong
2021-09-09 9:28 ` Jason Wang
2021-09-09 9:57 ` Wu Zongyong
2021-09-10 1:45 ` Jason Wang
2021-09-10 7:32 ` Wu Zongyong
2021-09-10 8:25 ` Cindy Lu
2021-09-10 9:20 ` Wu Zongyong
2021-09-10 15:10 ` Cindy Lu
2021-09-13 1:43 ` Jason Wang
2021-09-13 2:59 ` Wu Zongyong
2021-09-13 3:13 ` Jason Wang
2021-09-13 6:22 ` Cindy Lu
2021-09-09 9:18 ` Michael S. Tsirkin
2021-09-09 9:32 ` Jason Wang
2021-09-08 12:20 ` [PATCH 6/6] vp_vdpa: introduce legacy virtio pci driver Wu Zongyong
2021-09-09 13:57 ` Michael S. Tsirkin
2021-09-10 2:28 ` Wu Zongyong
2021-09-10 9:57 ` Michael S. Tsirkin
2021-09-10 10:01 ` Wu Zongyong
2021-10-22 9:37 ` Michael S. Tsirkin
2021-09-09 3:05 ` [PATCH 0/6] vDPA driver for legacy virtio-pci device Jason Wang
2021-09-09 3:20 ` Jason Wang
2021-09-09 8:12 ` Wu Zongyong
2021-09-09 9:29 ` Jason Wang
2021-09-09 9:21 ` Michael S. Tsirkin
2021-09-09 9:31 ` Jason Wang
2021-09-09 12:53 ` Wu Zongyong
2021-09-10 1:44 ` Jason Wang
2021-09-14 12:24 ` [PATCH v2 0/5] vDPA driver for Alibaba ENI Wu Zongyong
2021-09-14 12:24 ` Wu Zongyong
2021-09-14 12:24 ` [PATCH v2 1/5] virtio-pci: introduce legacy device module Wu Zongyong
2021-09-14 22:37 ` Randy Dunlap
2021-09-14 12:24 ` [PATCH v2 2/5] vdpa: fix typo Wu Zongyong
2021-09-15 3:15 ` Jason Wang
2021-09-14 12:24 ` [PATCH v2 3/5] vp_vdpa: add vq irq offloading support Wu Zongyong
2021-09-15 3:16 ` Jason Wang
2021-09-15 3:31 ` Wu Zongyong
2021-09-15 4:13 ` Jason Wang
2021-09-14 12:24 ` [PATCH v2 4/5] vdpa: add new vdpa attribute VDPA_ATTR_DEV_F_VERSION_1 Wu Zongyong
2021-09-14 12:58 ` Michael S. Tsirkin
2021-09-15 3:18 ` Jason Wang
2021-09-15 7:38 ` Michael S. Tsirkin
2021-09-15 8:06 ` Jason Wang
2021-09-15 11:08 ` Michael S. Tsirkin
2021-09-15 12:12 ` Wu Zongyong
2021-09-16 1:11 ` Jason Wang
2021-09-16 1:05 ` Jason Wang
2021-09-17 2:34 ` Wu Zongyong
2021-09-17 7:51 ` Jason Wang
2021-09-15 3:24 ` Wu Zongyong
2021-09-15 7:30 ` Michael S. Tsirkin
2021-09-15 8:05 ` Jason Wang
2021-09-15 11:10 ` Michael S. Tsirkin
2021-09-14 12:24 ` [PATCH v2 5/5] eni_vdpa: add vDPA driver for Alibaba ENI Wu Zongyong
2021-09-14 22:35 ` Randy Dunlap
2021-09-15 3:14 ` Jason Wang
2021-09-15 3:47 ` Wu Zongyong
2021-09-15 13:36 ` kernel test robot
2021-09-22 12:46 ` [PATCH v3 0/7] " Wu Zongyong
2021-09-22 12:46 ` [PATCH v3 1/7] virtio-pci: introduce legacy device module Wu Zongyong
2021-09-22 12:46 ` [PATCH v3 2/7] vdpa: fix typo Wu Zongyong
2021-09-22 12:46 ` [PATCH v3 3/7] vp_vdpa: add vq irq offloading support Wu Zongyong
2021-09-22 12:46 ` [PATCH v3 4/7] vdpa: add new callback get_vq_num_min in vdpa_config_ops Wu Zongyong
2021-09-22 12:46 ` [PATCH v3 5/7] virtio_vdpa: setup correct vq size with callbacks get_vq_num_{max,min} Wu Zongyong
2021-09-22 12:46 ` [PATCH v3 6/7] vdpa: add new attribute VDPA_ATTR_DEV_MIN_VQ_SIZE Wu Zongyong
2021-09-22 12:46 ` [PATCH v3 7/7] eni_vdpa: add vDPA driver for Alibaba ENI Wu Zongyong
2021-09-26 2:24 ` Jason Wang
2021-09-26 3:24 ` Wu Zongyong [this message]
2021-09-26 4:18 ` Jason Wang
2021-09-27 10:36 ` Michael S. Tsirkin
2021-09-28 2:17 ` Jason Wang
2021-09-26 2:26 ` Jason Wang
2021-09-26 3:27 ` Wu Zongyong
2021-09-26 4:19 ` Jason Wang
2021-09-29 12:54 ` kernel test robot
2021-09-29 6:11 ` [PATCH v4 0/7] " Wu Zongyong
2021-09-29 6:11 ` [PATCH v4 1/7] virtio-pci: introduce legacy device module Wu Zongyong
2021-10-11 3:00 ` Jason Wang
2021-10-11 3:02 ` Jason Wang
2021-09-29 6:11 ` [PATCH v4 2/7] vdpa: fix typo Wu Zongyong
2021-10-11 3:03 ` Jason Wang
2021-09-29 6:11 ` [PATCH v4 3/7] vp_vdpa: add vq irq offloading support Wu Zongyong
2021-09-29 6:11 ` [PATCH v4 4/7] vdpa: add new callback get_vq_num_min in vdpa_config_ops Wu Zongyong
2021-10-11 3:04 ` Jason Wang
2021-09-29 6:11 ` [PATCH v4 5/7] virtio_vdpa: setup correct vq size with callbacks get_vq_num_{max,min} Wu Zongyong
2021-10-11 3:06 ` Jason Wang
2021-09-29 6:11 ` [PATCH v4 6/7] vdpa: add new attribute VDPA_ATTR_DEV_MIN_VQ_SIZE Wu Zongyong
2021-10-11 3:10 ` Jason Wang
2021-09-29 6:11 ` [PATCH v4 7/7] eni_vdpa: add vDPA driver for Alibaba ENI Wu Zongyong
2021-10-11 3:18 ` Jason Wang
2021-10-15 7:14 ` [PATCH v5 0/8] " Wu Zongyong
2021-10-15 7:14 ` [PATCH v5 1/8] virtio-pci: introduce legacy device module Wu Zongyong
2021-10-15 8:22 ` Jason Wang
2021-10-15 7:14 ` [PATCH v5 2/8] vdpa: fix typo Wu Zongyong
2021-10-15 7:14 ` [PATCH v5 3/8] vp_vdpa: add vq irq offloading support Wu Zongyong
2021-10-15 7:14 ` [PATCH v5 4/8] vdpa: add new callback get_vq_num_min in vdpa_config_ops Wu Zongyong
2021-10-15 8:22 ` Jason Wang
2021-10-15 7:14 ` [PATCH v5 5/8] vdpa: min vq num of vdpa device cannot be greater than max vq num Wu Zongyong
2021-10-15 8:24 ` Jason Wang
2021-10-15 7:14 ` [PATCH v5 6/8] virtio_vdpa: setup correct vq size with callbacks get_vq_num_{max,min} Wu Zongyong
2021-10-15 8:30 ` Jason Wang
2021-10-15 7:15 ` [PATCH v5 7/8] vdpa: add new attribute VDPA_ATTR_DEV_MIN_VQ_SIZE Wu Zongyong
2021-10-15 7:15 ` [PATCH v5 8/8] eni_vdpa: add vDPA driver for Alibaba ENI Wu Zongyong
2021-10-15 8:18 ` Jason Wang
2021-10-15 19:13 ` Randy Dunlap
2021-10-22 2:44 ` [PATCH v6 0/8] " Wu Zongyong
2021-10-22 2:44 ` [PATCH v6 1/8] virtio-pci: introduce legacy device module Wu Zongyong
2021-10-25 1:30 ` Jason Wang
2021-10-22 2:44 ` [PATCH v6 2/8] vdpa: fix typo Wu Zongyong
2021-10-22 2:44 ` [PATCH v6 3/8] vp_vdpa: add vq irq offloading support Wu Zongyong
2021-10-22 2:44 ` [PATCH v6 4/8] vdpa: add new callback get_vq_num_min in vdpa_config_ops Wu Zongyong
2021-10-25 2:17 ` Jason Wang
2021-10-22 2:44 ` [PATCH v6 5/8] vdpa: min vq num of vdpa device cannot be greater than max vq num Wu Zongyong
2021-10-25 2:18 ` Jason Wang
2021-10-22 2:44 ` [PATCH v6 6/8] virtio_vdpa: setup correct vq size with callbacks get_vq_num_{max,min} Wu Zongyong
2021-10-25 2:22 ` Jason Wang
[not found] ` <20211025024403.GA3684@L-PF27918B-1352.localdomain>
2021-10-25 4:45 ` Jason Wang
[not found] ` <20211025062454.GA4832@L-PF27918B-1352.localdomain>
2021-10-26 4:46 ` Jason Wang
2021-10-22 2:44 ` [PATCH v6 7/8] vdpa: add new attribute VDPA_ATTR_DEV_MIN_VQ_SIZE Wu Zongyong
2021-10-25 2:23 ` Jason Wang
2021-10-22 2:44 ` [PATCH v6 8/8] eni_vdpa: add vDPA driver for Alibaba ENI Wu Zongyong
2021-10-25 2:27 ` Jason Wang
[not found] ` <20211025032146.GC3684@L-PF27918B-1352.localdomain>
2021-10-25 4:40 ` Jason Wang
2021-10-27 2:47 ` Wu Zongyong
2021-10-27 3:55 ` Jason Wang
2021-10-27 7:04 ` Wu Zongyong
2021-10-29 9:14 ` [PATCH v7 0/9] " Wu Zongyong
2021-10-29 9:14 ` [PATCH v7 1/9] virtio-pci: introduce legacy device module Wu Zongyong
2021-10-29 9:14 ` [PATCH v7 2/9] vdpa: fix typo Wu Zongyong
2021-10-29 9:14 ` [PATCH v7 3/9] vp_vdpa: add vq irq offloading support Wu Zongyong
2021-10-29 9:14 ` [PATCH v7 4/9] vdpa: add new callback get_vq_num_min in vdpa_config_ops Wu Zongyong
2021-10-29 9:14 ` [PATCH v7 5/9] vdpa: min vq num of vdpa device cannot be greater than max vq num Wu Zongyong
2021-10-29 9:14 ` [PATCH v7 6/9] virtio_vdpa: setup correct vq size with callbacks get_vq_num_{max,min} Wu Zongyong
2021-10-29 9:14 ` [PATCH v7 7/9] vdpa: add new attribute VDPA_ATTR_DEV_MIN_VQ_SIZE Wu Zongyong
2021-10-29 9:14 ` [PATCH v7 8/9] eni_vdpa: add vDPA driver for Alibaba ENI Wu Zongyong
2021-10-29 9:14 ` [PATCH v7 9/9] eni_vdpa: alibaba: fix Kconfig typo Wu Zongyong
2021-11-01 8:18 ` Michael S. Tsirkin
2021-11-01 3:31 ` [PATCH v7 0/9] vDPA driver for Alibaba ENI Jason Wang
2021-11-01 6:22 ` Wu Zongyong
2021-11-01 7:02 ` Jason Wang
2021-11-01 8:11 ` Wu Zongyong
2021-11-01 11:12 ` Michael S. Tsirkin
2021-11-01 8:19 ` Michael S. Tsirkin
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20210926032434.GA32570@L-PF27918B-1352.localdomain \
--to=wuzongyong@linux.alibaba.com \
--cc=jasowang@redhat.com \
--cc=linux-kernel@vger.kernel.org \
--cc=mst@redhat.com \
--cc=virtualization@lists.linux-foundation.org \
--cc=wei.yang1@linux.alibaba.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).