[Qemu-devel] [6073] virtio-net support

[Anthony Liguori <anthony@codemonkey.ws>]

Revision: 6073
          http://svn.sv.gnu.org/viewvc/?view=rev&root=qemu&revision=6073
Author:   aliguori
Date:     2008-12-17 19:13:11 +0000 (Wed, 17 Dec 2008)

Log Message:
-----------
virtio-net support

This adds virtio-net support.  This is based on the virtio-net driver 
that exists in kvm-userspace.  This also adds a new qemu_sendv_packet 
which virtio-net requires.

Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

Modified Paths:
--------------
    trunk/Makefile.target
    trunk/hw/pci.c
    trunk/net.c
    trunk/net.h

Added Paths:
-----------
    trunk/hw/virtio-net.c
    trunk/hw/virtio-net.h

Modified: trunk/Makefile.target
===================================================================
--- trunk/Makefile.target	2008-12-17 19:00:18 UTC (rev 6072)
+++ trunk/Makefile.target	2008-12-17 19:13:11 UTC (rev 6073)
@@ -637,7 +637,7 @@
 OBJS+= cirrus_vga.o apic.o parallel.o acpi.o piix_pci.o
 OBJS+= usb-uhci.o vmmouse.o vmport.o vmware_vga.o
 # virtio support
-OBJS+= virtio.o virtio-blk.o virtio-balloon.o
+OBJS+= virtio.o virtio-blk.o virtio-balloon.o virtio-net.o
 CPPFLAGS += -DHAS_AUDIO -DHAS_AUDIO_CHOICE
 endif
 ifeq ($(TARGET_BASE_ARCH), ppc)
@@ -664,7 +664,7 @@
 OBJS+= kvm_ppc.o
 endif
 # virtio support
-OBJS+= virtio.o virtio-blk.o virtio-balloon.o
+OBJS+= virtio.o virtio-blk.o virtio-balloon.o virtio-net.o
 endif
 ifeq ($(TARGET_BASE_ARCH), mips)
 OBJS+= mips_r4k.o mips_jazz.o mips_malta.o mips_mipssim.o

Modified: trunk/hw/pci.c
===================================================================
--- trunk/hw/pci.c	2008-12-17 19:00:18 UTC (rev 6072)
+++ trunk/hw/pci.c	2008-12-17 19:13:11 UTC (rev 6073)
@@ -25,6 +25,7 @@
 #include "pci.h"
 #include "console.h"
 #include "net.h"
+#include "virtio-net.h"
 
 //#define DEBUG_PCI
 
@@ -654,9 +655,11 @@
         pci_e1000_init(bus, nd, devfn);
     } else if (strcmp(nd->model, "pcnet") == 0) {
         pci_pcnet_init(bus, nd, devfn);
+    } else if (strcmp(nd->model, "virtio") == 0) {
+        virtio_net_init(bus, nd, devfn);
     } else if (strcmp(nd->model, "?") == 0) {
         fprintf(stderr, "qemu: Supported PCI NICs: i82551 i82557b i82559er"
-                        " ne2k_pci pcnet rtl8139 e1000\n");
+                        " ne2k_pci pcnet rtl8139 e1000 virtio\n");
         exit (1);
     } else {
         fprintf(stderr, "qemu: Unsupported NIC: %s\n", nd->model);

Added: trunk/hw/virtio-net.c
===================================================================
--- trunk/hw/virtio-net.c	                        (rev 0)
+++ trunk/hw/virtio-net.c	2008-12-17 19:13:11 UTC (rev 6073)
@@ -0,0 +1,329 @@
+/*
+ * Virtio Network Device
+ *
+ * Copyright IBM, Corp. 2007
+ *
+ * Authors:
+ *  Anthony Liguori   <aliguori@us.ibm.com>
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2.  See
+ * the COPYING file in the top-level directory.
+ *
+ */
+
+#include "virtio.h"
+#include "net.h"
+#include "qemu-timer.h"
+#include "virtio-net.h"
+
+typedef struct VirtIONet
+{
+    VirtIODevice vdev;
+    uint8_t mac[6];
+    VirtQueue *rx_vq;
+    VirtQueue *tx_vq;
+    VLANClientState *vc;
+    QEMUTimer *tx_timer;
+    int tx_timer_active;
+    int mergeable_rx_bufs;
+} VirtIONet;
+
+/* TODO
+ * - we could suppress RX interrupt if we were so inclined.
+ */
+
+static VirtIONet *to_virtio_net(VirtIODevice *vdev)
+{
+    return (VirtIONet *)vdev;
+}
+
+static void virtio_net_update_config(VirtIODevice *vdev, uint8_t *config)
+{
+    VirtIONet *n = to_virtio_net(vdev);
+    struct virtio_net_config netcfg;
+
+    memcpy(netcfg.mac, n->mac, 6);
+    memcpy(config, &netcfg, sizeof(netcfg));
+}
+
+static uint32_t virtio_net_get_features(VirtIODevice *vdev)
+{
+    uint32_t features = (1 << VIRTIO_NET_F_MAC);
+
+    return features;
+}
+
+static void virtio_net_set_features(VirtIODevice *vdev, uint32_t features)
+{
+    VirtIONet *n = to_virtio_net(vdev);
+
+    n->mergeable_rx_bufs = !!(features & (1 << VIRTIO_NET_F_MRG_RXBUF));
+}
+
+/* RX */
+
+static void virtio_net_handle_rx(VirtIODevice *vdev, VirtQueue *vq)
+{
+}
+
+static int do_virtio_net_can_receive(VirtIONet *n, int bufsize)
+{
+    if (!virtio_queue_ready(n->rx_vq) ||
+        !(n->vdev.status & VIRTIO_CONFIG_S_DRIVER_OK))
+        return 0;
+
+    if (virtio_queue_empty(n->rx_vq) ||
+        (n->mergeable_rx_bufs &&
+         !virtqueue_avail_bytes(n->rx_vq, bufsize, 0))) {
+        virtio_queue_set_notification(n->rx_vq, 1);
+        return 0;
+    }
+
+    virtio_queue_set_notification(n->rx_vq, 0);
+    return 1;
+}
+
+static int virtio_net_can_receive(void *opaque)
+{
+    VirtIONet *n = opaque;
+
+    return do_virtio_net_can_receive(n, VIRTIO_NET_MAX_BUFSIZE);
+}
+
+static int iov_fill(struct iovec *iov, int iovcnt, const void *buf, int count)
+{
+    int offset, i;
+
+    offset = i = 0;
+    while (offset < count && i < iovcnt) {
+        int len = MIN(iov[i].iov_len, count - offset);
+        memcpy(iov[i].iov_base, buf + offset, len);
+        offset += len;
+        i++;
+    }
+
+    return offset;
+}
+
+static int receive_header(VirtIONet *n, struct iovec *iov, int iovcnt,
+                          const void *buf, int size, int hdr_len)
+{
+    struct virtio_net_hdr *hdr = iov[0].iov_base;
+    int offset = 0;
+
+    hdr->flags = 0;
+    hdr->gso_type = VIRTIO_NET_HDR_GSO_NONE;
+
+    /* We only ever receive a struct virtio_net_hdr from the tapfd,
+     * but we may be passing along a larger header to the guest.
+     */
+    iov[0].iov_base += hdr_len;
+    iov[0].iov_len  -= hdr_len;
+
+    return offset;
+}
+
+static void virtio_net_receive(void *opaque, const uint8_t *buf, int size)
+{
+    VirtIONet *n = opaque;
+    struct virtio_net_hdr_mrg_rxbuf *mhdr = NULL;
+    int hdr_len, offset, i;
+
+    if (!do_virtio_net_can_receive(n, size))
+        return;
+
+    /* hdr_len refers to the header we supply to the guest */
+    hdr_len = n->mergeable_rx_bufs ?
+        sizeof(struct virtio_net_hdr_mrg_rxbuf) : sizeof(struct virtio_net_hdr);
+
+    offset = i = 0;
+
+    while (offset < size) {
+        VirtQueueElement elem;
+        int len, total;
+        struct iovec sg[VIRTQUEUE_MAX_SIZE];
+
+        len = total = 0;
+
+        if ((i != 0 && !n->mergeable_rx_bufs) ||
+            virtqueue_pop(n->rx_vq, &elem) == 0) {
+            if (i == 0)
+                return;
+            fprintf(stderr, "virtio-net truncating packet\n");
+            exit(1);
+        }
+
+        if (elem.in_num < 1) {
+            fprintf(stderr, "virtio-net receive queue contains no in buffers\n");
+            exit(1);
+        }
+
+        if (!n->mergeable_rx_bufs && elem.in_sg[0].iov_len != hdr_len) {
+            fprintf(stderr, "virtio-net header not in first element\n");
+            exit(1);
+        }
+
+        memcpy(&sg, &elem.in_sg[0], sizeof(sg[0]) * elem.in_num);
+
+        if (i == 0) {
+            if (n->mergeable_rx_bufs)
+                mhdr = (struct virtio_net_hdr_mrg_rxbuf *)sg[0].iov_base;
+
+            offset += receive_header(n, sg, elem.in_num,
+                                     buf + offset, size - offset, hdr_len);
+            total += hdr_len;
+        }
+
+        /* copy in packet.  ugh */
+        len = iov_fill(sg, elem.in_num,
+                       buf + offset, size - offset);
+        total += len;
+
+        /* signal other side */
+        virtqueue_fill(n->rx_vq, &elem, total, i++);
+
+        offset += len;
+    }
+
+    if (mhdr)
+        mhdr->num_buffers = i;
+
+    virtqueue_flush(n->rx_vq, i);
+    virtio_notify(&n->vdev, n->rx_vq);
+}
+
+/* TX */
+static void virtio_net_flush_tx(VirtIONet *n, VirtQueue *vq)
+{
+    VirtQueueElement elem;
+    int has_vnet_hdr = 0;
+
+    if (!(n->vdev.status & VIRTIO_CONFIG_S_DRIVER_OK))
+        return;
+
+    while (virtqueue_pop(vq, &elem)) {
+        ssize_t len = 0;
+        unsigned int out_num = elem.out_num;
+        struct iovec *out_sg = &elem.out_sg[0];
+        unsigned hdr_len;
+
+        /* hdr_len refers to the header received from the guest */
+        hdr_len = n->mergeable_rx_bufs ?
+            sizeof(struct virtio_net_hdr_mrg_rxbuf) :
+            sizeof(struct virtio_net_hdr);
+
+        if (out_num < 1 || out_sg->iov_len != hdr_len) {
+            fprintf(stderr, "virtio-net header not in first element\n");
+            exit(1);
+        }
+
+        /* ignore the header if GSO is not supported */
+        if (!has_vnet_hdr) {
+            out_num--;
+            out_sg++;
+            len += hdr_len;
+        } else if (n->mergeable_rx_bufs) {
+            /* tapfd expects a struct virtio_net_hdr */
+            hdr_len -= sizeof(struct virtio_net_hdr);
+            out_sg->iov_len -= hdr_len;
+            len += hdr_len;
+        }
+
+        len += qemu_sendv_packet(n->vc, out_sg, out_num);
+
+        virtqueue_push(vq, &elem, len);
+        virtio_notify(&n->vdev, vq);
+    }
+}
+
+static void virtio_net_handle_tx(VirtIODevice *vdev, VirtQueue *vq)
+{
+    VirtIONet *n = to_virtio_net(vdev);
+
+    if (n->tx_timer_active) {
+        virtio_queue_set_notification(vq, 1);
+        qemu_del_timer(n->tx_timer);
+        n->tx_timer_active = 0;
+        virtio_net_flush_tx(n, vq);
+    } else {
+        qemu_mod_timer(n->tx_timer,
+                       qemu_get_clock(vm_clock) + TX_TIMER_INTERVAL);
+        n->tx_timer_active = 1;
+        virtio_queue_set_notification(vq, 0);
+    }
+}
+
+static void virtio_net_tx_timer(void *opaque)
+{
+    VirtIONet *n = opaque;
+
+    n->tx_timer_active = 0;
+
+    /* Just in case the driver is not ready on more */
+    if (!(n->vdev.status & VIRTIO_CONFIG_S_DRIVER_OK))
+        return;
+
+    virtio_queue_set_notification(n->tx_vq, 1);
+    virtio_net_flush_tx(n, n->tx_vq);
+}
+
+static void virtio_net_save(QEMUFile *f, void *opaque)
+{
+    VirtIONet *n = opaque;
+
+    virtio_save(&n->vdev, f);
+
+    qemu_put_buffer(f, n->mac, 6);
+    qemu_put_be32(f, n->tx_timer_active);
+}
+
+static int virtio_net_load(QEMUFile *f, void *opaque, int version_id)
+{
+    VirtIONet *n = opaque;
+
+    if (version_id != 1)
+        return -EINVAL;
+
+    virtio_load(&n->vdev, f);
+
+    qemu_get_buffer(f, n->mac, 6);
+    n->tx_timer_active = qemu_get_be32(f);
+
+    if (n->tx_timer_active) {
+        qemu_mod_timer(n->tx_timer,
+                       qemu_get_clock(vm_clock) + TX_TIMER_INTERVAL);
+    }
+
+    return 0;
+}
+
+PCIDevice *virtio_net_init(PCIBus *bus, NICInfo *nd, int devfn)
+{
+    VirtIONet *n;
+    static int virtio_net_id;
+
+    n = (VirtIONet *)virtio_init_pci(bus, "virtio-net", 6900, 0x1000,
+                                     0, VIRTIO_ID_NET,
+                                     0x02, 0x00, 0x00,
+                                     6, sizeof(VirtIONet));
+    if (!n)
+        return NULL;
+
+    n->vdev.get_config = virtio_net_update_config;
+    n->vdev.get_features = virtio_net_get_features;
+    n->vdev.set_features = virtio_net_set_features;
+    n->rx_vq = virtio_add_queue(&n->vdev, 256, virtio_net_handle_rx);
+    n->tx_vq = virtio_add_queue(&n->vdev, 256, virtio_net_handle_tx);
+    memcpy(n->mac, nd->macaddr, 6);
+    n->vc = qemu_new_vlan_client(nd->vlan, virtio_net_receive,
+                                 virtio_net_can_receive, n);
+
+    n->tx_timer = qemu_new_timer(vm_clock, virtio_net_tx_timer, n);
+    n->tx_timer_active = 0;
+    n->mergeable_rx_bufs = 0;
+
+    register_savevm("virtio-net", virtio_net_id++, 1,
+                    virtio_net_save, virtio_net_load, n);
+
+    return (PCIDevice *)n;
+}

Added: trunk/hw/virtio-net.h
===================================================================
--- trunk/hw/virtio-net.h	                        (rev 0)
+++ trunk/hw/virtio-net.h	2008-12-17 19:13:11 UTC (rev 6073)
@@ -0,0 +1,80 @@
+/*
+ * Virtio Network Device
+ *
+ * Copyright IBM, Corp. 2007
+ *
+ * Authors:
+ *  Anthony Liguori   <aliguori@us.ibm.com>
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2.  See
+ * the COPYING file in the top-level directory.
+ *
+ */
+
+#ifndef _QEMU_VIRTIO_NET_H
+#define _QEMU_VIRTIO_NET_H
+
+#include "virtio.h"
+#include "net.h"
+#include "pci.h"
+
+/* from Linux's virtio_net.h */
+
+/* The ID for virtio_net */
+#define VIRTIO_ID_NET   1
+
+/* The feature bitmap for virtio net */
+#define VIRTIO_NET_F_CSUM       0       /* Host handles pkts w/ partial csum */
+#define VIRTIO_NET_F_GUEST_CSUM 1       /* Guest handles pkts w/ partial csum */
+#define VIRTIO_NET_F_MAC        5       /* Host has given MAC address. */
+#define VIRTIO_NET_F_GSO        6       /* Host handles pkts w/ any GSO type */
+#define VIRTIO_NET_F_GUEST_TSO4 7       /* Guest can handle TSOv4 in. */
+#define VIRTIO_NET_F_GUEST_TSO6 8       /* Guest can handle TSOv6 in. */
+#define VIRTIO_NET_F_GUEST_ECN  9       /* Guest can handle TSO[6] w/ ECN in. */
+#define VIRTIO_NET_F_GUEST_UFO  10      /* Guest can handle UFO in. */
+#define VIRTIO_NET_F_HOST_TSO4  11      /* Host can handle TSOv4 in. */
+#define VIRTIO_NET_F_HOST_TSO6  12      /* Host can handle TSOv6 in. */
+#define VIRTIO_NET_F_HOST_ECN   13      /* Host can handle TSO[6] w/ ECN in. */
+#define VIRTIO_NET_F_HOST_UFO   14      /* Host can handle UFO in. */
+#define VIRTIO_NET_F_MRG_RXBUF  15      /* Host can merge receive buffers. */
+
+#define TX_TIMER_INTERVAL 150000 /* 150 us */
+
+/* Maximum packet size we can receive from tap device: header + 64k */
+#define VIRTIO_NET_MAX_BUFSIZE (sizeof(struct virtio_net_hdr) + (64 << 10))
+
+/* The config defining mac address (6 bytes) */
+struct virtio_net_config
+{
+    uint8_t mac[6];
+} __attribute__((packed));
+
+/* This is the first element of the scatter-gather list.  If you don't
+ * specify GSO or CSUM features, you can simply ignore the header. */
+struct virtio_net_hdr
+{
+#define VIRTIO_NET_HDR_F_NEEDS_CSUM     1       // Use csum_start, csum_offset
+    uint8_t flags;
+#define VIRTIO_NET_HDR_GSO_NONE         0       // Not a GSO frame
+#define VIRTIO_NET_HDR_GSO_TCPV4        1       // GSO frame, IPv4 TCP (TSO)
+#define VIRTIO_NET_HDR_GSO_UDP          3       // GSO frame, IPv4 UDP (UFO)
+#define VIRTIO_NET_HDR_GSO_TCPV6        4       // GSO frame, IPv6 TCP
+#define VIRTIO_NET_HDR_GSO_ECN          0x80    // TCP has ECN set
+    uint8_t gso_type;
+    uint16_t hdr_len;
+    uint16_t gso_size;
+    uint16_t csum_start;
+    uint16_t csum_offset;
+};
+
+/* This is the version of the header to use when the MRG_RXBUF
+ * feature has been negotiated. */
+struct virtio_net_hdr_mrg_rxbuf
+{
+    struct virtio_net_hdr hdr;
+    uint16_t num_buffers;   /* Number of merged rx buffers */
+};
+
+PCIDevice *virtio_net_init(PCIBus *bus, NICInfo *nd, int devfn);
+
+#endif

Modified: trunk/net.c
===================================================================
--- trunk/net.c	2008-12-17 19:00:18 UTC (rev 6072)
+++ trunk/net.c	2008-12-17 19:13:11 UTC (rev 6073)
@@ -369,6 +369,50 @@
     }
 }
 
+static ssize_t vc_sendv_compat(VLANClientState *vc, const struct iovec *iov,
+                               int iovcnt)
+{
+    uint8_t buffer[4096];
+    size_t offset = 0;
+    int i;
+
+    for (i = 0; i < iovcnt; i++) {
+        size_t len;
+
+        len = MIN(sizeof(buffer) - offset, iov[i].iov_len);
+        memcpy(buffer + offset, iov[i].iov_base, len);
+        offset += len;
+    }
+
+    vc->fd_read(vc->opaque, buffer, offset);
+
+    return offset;
+}
+
+ssize_t qemu_sendv_packet(VLANClientState *vc1, const struct iovec *iov,
+                          int iovcnt)
+{
+    VLANState *vlan = vc1->vlan;
+    VLANClientState *vc;
+    ssize_t max_len = 0;
+
+    for (vc = vlan->first_client; vc != NULL; vc = vc->next) {
+        ssize_t len = 0;
+
+        if (vc == vc1)
+            continue;
+
+        if (vc->fd_readv)
+            len = vc->fd_readv(vc->opaque, iov, iovcnt);
+        else if (vc->fd_read)
+            len = vc_sendv_compat(vc, iov, iovcnt);
+
+        max_len = MAX(max_len, len);
+    }
+
+    return max_len;
+}
+
 #if defined(CONFIG_SLIRP)
 
 /* slirp network adapter */

Modified: trunk/net.h
===================================================================
--- trunk/net.h	2008-12-17 19:00:18 UTC (rev 6072)
+++ trunk/net.h	2008-12-17 19:13:11 UTC (rev 6073)
@@ -1,12 +1,17 @@
 #ifndef QEMU_NET_H
 #define QEMU_NET_H
 
+#include "qemu-common.h"
+
 /* VLANs support */
 
+typedef ssize_t (IOReadvHandler)(void *, const struct iovec *, int);
+
 typedef struct VLANClientState VLANClientState;
 
 struct VLANClientState {
     IOReadHandler *fd_read;
+    IOReadvHandler *fd_readv;
     /* Packets may still be sent if this returns zero.  It's used to
        rate-limit the slirp code.  */
     IOCanRWHandler *fd_can_read;
@@ -30,6 +35,8 @@
                                       void *opaque);
 void qemu_del_vlan_client(VLANClientState *vc);
 int qemu_can_send_packet(VLANClientState *vc);
+ssize_t qemu_sendv_packet(VLANClientState *vc, const struct iovec *iov,
+                          int iovcnt);
 void qemu_send_packet(VLANClientState *vc, const uint8_t *buf, int size);
 void qemu_handler_true(void *opaque);