lguest: Handle non-GSO Guests This is a test of feature bits. Signed-off-by: Rusty Russell --- Documentation/lguest/lguest.c | 315 ++++++++++++++++++++++++++---------------- 1 file changed, 199 insertions(+), 116 deletions(-) diff -r 34ccd603999b Documentation/lguest/lguest.c --- a/Documentation/lguest/lguest.c Thu Jan 24 12:46:30 2008 +1100 +++ b/Documentation/lguest/lguest.c Thu Jan 24 12:47:03 2008 +1100 @@ -36,6 +36,7 @@ #include #include #include +#include #include "linux/lguest_launcher.h" #include "linux/virtio_config.h" #include "linux/virtio_net.h" @@ -44,7 +45,7 @@ #include "linux/virtio_rng.h" #include "linux/virtio_ring.h" #include "asm-x86/bootparam.h" -/*L:110 We can ignore the 38 include files we need for this program, but I do +/*L:110 We can ignore the 39 include files we need for this program, but I do * want to draw attention to the use of kernel-style types. * * As Linus said, "C is a Spartan language, and so should your naming be." I @@ -199,6 +200,14 @@ static u8 *get_feature_bits(struct devic { return (u8 *)(dev->desc + 1) + dev->desc->num_vq * sizeof(struct lguest_vqconfig); +} + +static bool has_feature(struct device *dev, unsigned bit) +{ + /* Guest feature bits follow our feature bits. */ + u8 *features = get_feature_bits(dev) + dev->desc->feature_len; + + return features[bit / CHAR_BIT] & (1 << (bit % CHAR_BIT)); } /*L:100 The Launcher code itself takes us out into userspace, that scary place @@ -842,26 +851,151 @@ static void handle_console_output(int fd add_used_and_trigger(fd, vq, head, len); } } +/*:*/ +static u32 str2ip(const char *ipaddr) +{ + unsigned int byte[4]; -/* + sscanf(ipaddr, "%u.%u.%u.%u", &byte[0], &byte[1], &byte[2], &byte[3]); + return (byte[0] << 24) | (byte[1] << 16) | (byte[2] << 8) | byte[3]; +} + +/* This code is "adapted" from libbridge: it attaches the Host end of the + * network device to the bridge device specified by the command line. + * + * This is yet another James Morris contribution (I'm an IP-level guy, so I + * dislike bridging), and I just try not to break it. */ +static void add_to_bridge(int fd, const char *if_name, const char *br_name) +{ + int ifidx; + struct ifreq ifr; + + if (!*br_name) + errx(1, "must specify bridge name"); + + ifidx = if_nametoindex(if_name); + if (!ifidx) + errx(1, "interface %s does not exist!", if_name); + + strncpy(ifr.ifr_name, br_name, IFNAMSIZ); + ifr.ifr_ifindex = ifidx; + if (ioctl(fd, SIOCBRADDIF, &ifr) < 0) + err(1, "can't add %s to bridge %s", if_name, br_name); +} + +/* This sets up the Host end of the network device with an IP address and brings + * it up so packets will flow. */ +static void configure_device(int fd, const char *devname, u32 ipaddr) +{ + struct ifreq ifr; + struct sockaddr_in *sin = (struct sockaddr_in *)&ifr.ifr_addr; + + /* Don't read these incantations. Just cut & paste them like I did! */ + memset(&ifr, 0, sizeof(ifr)); + strcpy(ifr.ifr_name, devname); + sin->sin_family = AF_INET; + sin->sin_addr.s_addr = htonl(ipaddr); + if (ioctl(fd, SIOCSIFADDR, &ifr) != 0) + err(1, "Setting %s interface address", devname); + ifr.ifr_flags = IFF_UP; + if (ioctl(fd, SIOCSIFFLAGS, &ifr) != 0) + err(1, "Bringing interface %s up", devname); +} + +/* Normally, we assume that the Guest uses all the features, so it will be + * using gso. But to demonstrate feature negotiation, we allow it to choose + * not to. Unfortunately, we can't configure the tun/tap device until we can + * tell it whether we handle GSO or not, so we delay that until later. */ +struct net_priv +{ + bool done_setup; + bool use_gso; + struct virtio_net_config conf; + u32 ip; + const char *bridge_name; +}; + +static void complete_net_setup(struct device *dev) +{ + struct net_priv *priv = dev->priv; + int ipfd; + struct ifreq ifr; + + if (priv->done_setup) + return; + + memset(&ifr, 0, sizeof(ifr)); + ifr.ifr_flags = IFF_TAP | IFF_NO_PI; + + /* If Guest can handle everything, we can enable IFF_GSO_HDR */ + if (has_feature(dev, VIRTIO_NET_F_CSUM) + && has_feature(dev, VIRTIO_NET_F_GSO)) { + priv->use_gso = true; + ifr.ifr_flags |= IFF_VIRTIO_HDR|IFF_RECV_CSUM|IFF_RECV_GSO; + } else + priv->use_gso = false; + + strcpy(ifr.ifr_name, "tap%d"); + if (ioctl(dev->fd, TUNSETIFF, &ifr) != 0) + err(1, "configuring /dev/net/tun"); + + /* SIOC stands for Socket I/O Control. S means Set. IF means + * Interface, and HWADDR is hardware address. Simple! */ + memcpy(ifr.ifr_hwaddr.sa_data, priv->conf.mac, ETH_ALEN); + ifr.ifr_hwaddr.sa_family = AF_LOCAL; + if (ioctl(dev->fd, SIOCSIFHWADDR, &ifr) != 0) + err(1, "setting hw address for %s", ifr.ifr_name); + + /* We don't need checksums calculated for packets coming in this + * device: trust us! */ + ioctl(dev->fd, TUNSETNOCSUM, 1); + + /* We need a socket to perform the magic network ioctls to bring up the + * tap interface, connect to the bridge etc. Any socket will do! */ + ipfd = socket(PF_INET, SOCK_DGRAM, IPPROTO_IP); + if (ipfd < 0) + err(1, "opening IP socket"); + + /* If the command line was --tunnet=bridge: do bridging. */ + if (priv->bridge_name) + add_to_bridge(ipfd, ifr.ifr_name, priv->bridge_name); + + /* Set up the tun device, and get the mac address for the interface. */ + configure_device(ipfd, ifr.ifr_name, priv->ip); + + priv->done_setup = true; +} + +/*L:201 * The Network * * Handling output for network is also simple: we get all the output buffers - * and write them (ignoring the first element) to this device's file descriptor - * (stdout). */ + * and write them to this device's file descriptor (the tap device). */ static void handle_net_output(int fd, struct virtqueue *vq) { unsigned int head, out, in; int len; + struct net_priv *priv = vq->dev->priv; struct iovec iov[vq->vring.num]; + + /* We might not know whether this Guest speaks GSO until now. */ + complete_net_setup(vq->dev); /* Keep getting output buffers from the Guest until we run out. */ while ((head = get_vq_desc(vq, iov, &out, &in)) != vq->vring.num) { if (in) errx(1, "Input buffers in output queue?"); + /* With IFF_GSO_HDR, tap takes the same format header as - * virtio_net, so we just pass it through. */ - len = writev(vq->dev->fd, iov, out); + * virtio_net, so we just pass it through. Otherwise, + * skip the header. */ + if (priv->use_gso) + len = writev(vq->dev->fd, iov, out); + else { + len = writev(vq->dev->fd, iov+1, out-1); + if (len >= 0) + len += iov[0].iov_len; + } add_used_and_trigger(fd, vq, head, len); } } @@ -872,27 +1006,40 @@ static bool handle_tun_input(int fd, str { unsigned int head, in_num, out_num; int len; + struct net_priv *priv = dev->priv; struct iovec iov[dev->vq->vring.num]; + + /* We might not know whether this Guest speaks GSO until now. */ + complete_net_setup(dev); /* First we need a network buffer from the Guests's recv virtqueue. */ head = get_vq_desc(dev->vq, iov, &out_num, &in_num); if (head == dev->vq->vring.num) { - /* Now, it's expected that if we try to send a packet too - * early, the Guest won't be ready yet. Wait until the device - * status says it's ready. */ - /* FIXME: Actually want DRIVER_ACTIVE here. */ - if (dev->desc->status & VIRTIO_CONFIG_S_DRIVER_OK) - warn("network: no dma buffer!"); + /* FIXME: Only do this if DRIVER_ACTIVE. */ + warn("network: no dma buffer!"); /* We'll turn this back on if input buffers are registered. */ return false; } else if (out_num) errx(1, "Output buffers in network recv queue?"); - /* Read the GSO header and packet from the tun device directly into the - * Guest's buffer. */ - len = readv(dev->fd, iov, in_num); - if (len <= 0) - err(1, "reading network"); + if (!priv->use_gso) { + /* Give them a noop header. */ + struct virtio_net_hdr hdr = { .flags = 0 }; + if (iov[0].iov_len != sizeof(hdr)) + errx(1, "Bad net header len %u", iov[0].iov_len); + memcpy(iov[0].iov_base, &hdr, sizeof(hdr)); + + len = readv(dev->fd, iov+1, in_num-1); + if (len <= 0) + err(1, "reading network"); + len += sizeof(hdr); + } else { + /* Read the GSO header and packet from the tun device directly + * into the Guest's buffer. */ + len = readv(dev->fd, iov, in_num); + if (len <= 0) + err(1, "reading network"); + } /* Tell the Guest about the new packet. */ add_used_and_trigger(fd, dev->vq, head, len); @@ -997,17 +1144,20 @@ static void handle_input(int fd) /* Otherwise, call the device(s) which have readable * file descriptors and a method of handling them. */ for (i = devices.dev; i; i = i->next) { + bool ok = i->desc->status&VIRTIO_CONFIG_S_ACKNOWLEDGE; if (i->handle_input && FD_ISSET(i->fd, &fds)) { int dev_fd; - if (i->handle_input(fd, i)) + + /* Don't bother the Guest before it's ready. */ + if (ok && i->handle_input(fd, i)) continue; - /* If handle_input() returns false, it means we - * should no longer service it. Networking and - * console do this when there's no input - * buffers to deliver into. Console also uses - * it when it discovers that stdin is - * closed. */ + /* If handle_input() returns false or driver + * isn't ready, it means we should no longer + * service it. Networking and console do this + * when there's no input buffers to deliver + * into. Console also uses it when it + * discovers that stdin is closed. */ FD_CLR(i->fd, &devices.infds); /* Tell waker to ignore it too, by sending a * negative fd number (-1, since 0 is a valid @@ -1223,63 +1373,14 @@ static void setup_console(void) * * Finally, we could implement a virtio network switch in the kernel. :*/ -static u32 str2ip(const char *ipaddr) +static void random_ether_addr(u8 *mac) { - unsigned int byte[4]; - - sscanf(ipaddr, "%u.%u.%u.%u", &byte[0], &byte[1], &byte[2], &byte[3]); - return (byte[0] << 24) | (byte[1] << 16) | (byte[2] << 8) | byte[3]; -} - -/* This code is "adapted" from libbridge: it attaches the Host end of the - * network device to the bridge device specified by the command line. - * - * This is yet another James Morris contribution (I'm an IP-level guy, so I - * dislike bridging), and I just try not to break it. */ -static void add_to_bridge(int fd, const char *if_name, const char *br_name) -{ - int ifidx; - struct ifreq ifr; - - if (!*br_name) - errx(1, "must specify bridge name"); - - ifidx = if_nametoindex(if_name); - if (!ifidx) - errx(1, "interface %s does not exist!", if_name); - - strncpy(ifr.ifr_name, br_name, IFNAMSIZ); - ifr.ifr_ifindex = ifidx; - if (ioctl(fd, SIOCBRADDIF, &ifr) < 0) - err(1, "can't add %s to bridge %s", if_name, br_name); -} - -/* This sets up the Host end of the network device with an IP address, brings - * it up so packets will flow, the copies the MAC address into the hwaddr - * pointer. */ -static void configure_device(int fd, const char *devname, u32 ipaddr, - unsigned char hwaddr[6]) -{ - struct ifreq ifr; - struct sockaddr_in *sin = (struct sockaddr_in *)&ifr.ifr_addr; - - /* Don't read these incantations. Just cut & paste them like I did! */ - memset(&ifr, 0, sizeof(ifr)); - strcpy(ifr.ifr_name, devname); - sin->sin_family = AF_INET; - sin->sin_addr.s_addr = htonl(ipaddr); - if (ioctl(fd, SIOCSIFADDR, &ifr) != 0) - err(1, "Setting %s interface address", devname); - ifr.ifr_flags = IFF_UP; - if (ioctl(fd, SIOCSIFFLAGS, &ifr) != 0) - err(1, "Bringing interface %s up", devname); - - /* SIOC stands for Socket I/O Control. G means Get (vs S for Set - * above). IF means Interface, and HWADDR is hardware address. - * Simple! */ - if (ioctl(fd, SIOCGIFHWADDR, &ifr) != 0) - err(1, "getting hw address for %s", devname); - memcpy(hwaddr, ifr.ifr_hwaddr.sa_data, 6); + int randfd = open_or_die("/dev/urandom", O_RDONLY); + if (read(randfd, mac, ETH_ALEN) != ETH_ALEN) + err(1, "Reading /dev/urandom"); + close(randfd); + mac[0] &= 0xfe; /* clear multicast bit */ + mac[0] |= 0x02; /* set local assignment bit (IEEE802) */ } /*L:195 Our network is a Host<->Guest network. This can either use bridging or @@ -1289,26 +1390,14 @@ static void setup_tun_net(const char *ar static void setup_tun_net(const char *arg) { struct device *dev; - struct ifreq ifr; - int netfd, ipfd; - u32 ip; - const char *br_name = NULL; - struct virtio_net_config conf; + int netfd; + struct net_priv *priv = malloc(sizeof(*priv)); /* We open the /dev/net/tun device and tell it we want a tap device. A * tap device is like a tun device, only somehow different. To tell * the truth, I completely blundered my way through this code, but it * works now! */ netfd = open_or_die("/dev/net/tun", O_RDWR); - memset(&ifr, 0, sizeof(ifr)); - ifr.ifr_flags = (IFF_TAP | IFF_NO_PI | IFF_VIRTIO_HDR - | IFF_RECV_CSUM | IFF_RECV_GSO); - strcpy(ifr.ifr_name, "tap%d"); - if (ioctl(netfd, TUNSETIFF, &ifr) != 0) - err(1, "configuring /dev/net/tun"); - /* We don't need checksums calculated for packets coming in this - * device: trust us! */ - ioctl(netfd, TUNSETNOCSUM, 1); /* First we create a new network device. */ dev = new_device("net", VIRTIO_ID_NET, netfd, handle_tun_input); @@ -1318,38 +1407,32 @@ static void setup_tun_net(const char *ar add_virtqueue(dev, VIRTQUEUE_NUM, enable_fd); add_virtqueue(dev, VIRTQUEUE_NUM, handle_net_output); - /* We need a socket to perform the magic network ioctls to bring up the - * tap interface, connect to the bridge etc. Any socket will do! */ - ipfd = socket(PF_INET, SOCK_DGRAM, IPPROTO_IP); - if (ipfd < 0) - err(1, "opening IP socket"); + dev->priv = priv; + + /* We have to do setup later. */ + priv->done_setup = false; /* If the command line was --tunnet=bridge: do bridging. */ if (!strncmp(BRIDGE_PFX, arg, strlen(BRIDGE_PFX))) { - ip = INADDR_ANY; - br_name = arg + strlen(BRIDGE_PFX); - add_to_bridge(ipfd, ifr.ifr_name, br_name); + priv->ip = INADDR_ANY; + priv->bridge_name = arg + strlen(BRIDGE_PFX); } else /* It is an IP address to set up the device with */ - ip = str2ip(arg); + priv->ip = str2ip(arg); - /* Set up the tun device, and get the mac address for the interface. */ - configure_device(ipfd, ifr.ifr_name, ip, conf.mac); + /* Get a random mac address, and tell the Guest to use it */ + random_ether_addr(priv->conf.mac); + add_feature(dev, VIRTIO_NET_F_MAC); + set_config(dev, sizeof(priv->conf), &priv->conf); - /* Tell Guest what MAC address to use. */ - add_feature(dev, VIRTIO_NET_F_MAC); - set_config(dev, sizeof(conf), &conf); - /* Tap device can handle csums, and all types of GSO. */ + /* Tap device can handle no csum, and all types of GSO. */ add_feature(dev, VIRTIO_NET_F_CSUM); add_feature(dev, VIRTIO_NET_F_GSO); - /* We don't need the socket any more; setup is done. */ - close(ipfd); - verbose("device %u: tun net %u.%u.%u.%u\n", - devices.device_num++, - (u8)(ip>>24),(u8)(ip>>16),(u8)(ip>>8),(u8)ip); - if (br_name) - verbose("attached to bridge: %s\n", br_name); + devices.device_num++, (u8)(priv->ip>>24), (u8)(priv->ip>>16), + (u8)(priv->ip>>8), (u8)priv->ip); + if (priv->bridge_name) + verbose("attached to bridge: %s\n", priv->bridge_name); } /* Our block (disk) device should be really simple: the Guest asks for a block