Merge remote-tracking branch 'stefanha/net' into staging

# By Dmitry Fleytman (5) and others
# Via Stefan Hajnoczi
* stefanha/net:
  net: increase buffer size to accommodate Jumbo frame pkts
  VMXNET3 device implementation
  Packet abstraction for VMWARE network devices
  Common definitions for VMWARE devices
  net: iovec checksum calculator
  Checksum-related utility functions
  net: use socket_set_nodelay() for -netdev socket

default-configs/pci.mak

@@ -13,6 +13,7 @@ CONFIG_LSI_SCSI_PCI=y
 CONFIG_MEGASAS_SCSI_PCI=y
 CONFIG_RTL8139_PCI=y
 CONFIG_E1000_PCI=y
+CONFIG_VMXNET3_PCI=y
 CONFIG_IDE_CORE=y
 CONFIG_IDE_QDEV=y
 CONFIG_IDE_PCI=y

hw/Makefile.objs

@@ -119,6 +119,8 @@ common-obj-$(CONFIG_PCNET_PCI) += pcnet-pci.o
 common-obj-$(CONFIG_PCNET_COMMON) += pcnet.o
 common-obj-$(CONFIG_E1000_PCI) += e1000.o
 common-obj-$(CONFIG_RTL8139_PCI) += rtl8139.o
+common-obj-$(CONFIG_VMXNET3_PCI) += vmxnet_tx_pkt.o vmxnet_rx_pkt.o
+common-obj-$(CONFIG_VMXNET3_PCI) += vmxnet3.o
 
 common-obj-$(CONFIG_SMC91C111) += smc91c111.o
 common-obj-$(CONFIG_LAN9118) += lan9118.o

hw/pci/pci.h

@@ -60,6 +60,7 @@
 #define PCI_DEVICE_ID_VMWARE_NET         0x0720
 #define PCI_DEVICE_ID_VMWARE_SCSI        0x0730
 #define PCI_DEVICE_ID_VMWARE_IDE         0x1729
+#define PCI_DEVICE_ID_VMWARE_VMXNET3     0x07B0
 
 /* Intel (0x8086) */
 #define PCI_DEVICE_ID_INTEL_82551IT      0x1209

hw/vmware_utils.h

+/*
+ * QEMU VMWARE paravirtual devices - auxiliary code
+ *
+ * Copyright (c) 2012 Ravello Systems LTD (http://ravellosystems.com)
+ *
+ * Developed by Daynix Computing LTD (http://www.daynix.com)
+ *
+ * Authors:
+ * Dmitry Fleytman <dmitry@daynix.com>
+ * Yan Vugenfirer <yan@daynix.com>
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ *
+ */
+
+#ifndef VMWARE_UTILS_H
+#define VMWARE_UTILS_H
+
+#include "qemu/range.h"
+
+#ifndef VMW_SHPRN
+#define VMW_SHPRN(fmt, ...) do {} while (0)
+#endif
+
+/*
+ * Shared memory access functions with byte swap support
+ * Each function contains printout for reverse-engineering needs
+ *
+ */
+static inline void
+vmw_shmem_read(hwaddr addr, void *buf, int len)
+{
+    VMW_SHPRN("SHMEM r: %" PRIx64 ", len: %d to %p", addr, len, buf);
+    cpu_physical_memory_read(addr, buf, len);
+}
+
+static inline void
+vmw_shmem_write(hwaddr addr, void *buf, int len)
+{
+    VMW_SHPRN("SHMEM w: %" PRIx64 ", len: %d to %p", addr, len, buf);
+    cpu_physical_memory_write(addr, buf, len);
+}
+
+static inline void
+vmw_shmem_rw(hwaddr addr, void *buf, int len, int is_write)
+{
+    VMW_SHPRN("SHMEM r/w: %" PRIx64 ", len: %d (to %p), is write: %d",
+              addr, len, buf, is_write);
+
+    cpu_physical_memory_rw(addr, buf, len, is_write);
+}
+
+static inline void
+vmw_shmem_set(hwaddr addr, uint8 val, int len)
+{
+    int i;
+    VMW_SHPRN("SHMEM set: %" PRIx64 ", len: %d (value 0x%X)", addr, len, val);
+
+    for (i = 0; i < len; i++) {
+        cpu_physical_memory_write(addr + i, &val, 1);
+    }
+}
+
+static inline uint32_t
+vmw_shmem_ld8(hwaddr addr)
+{
+    uint8_t res = ldub_phys(addr);
+    VMW_SHPRN("SHMEM load8: %" PRIx64 " (value 0x%X)", addr, res);
+    return res;
+}
+
+static inline void
+vmw_shmem_st8(hwaddr addr, uint8_t value)
+{
+    VMW_SHPRN("SHMEM store8: %" PRIx64 " (value 0x%X)", addr, value);
+    stb_phys(addr, value);
+}
+
+static inline uint32_t
+vmw_shmem_ld16(hwaddr addr)
+{
+    uint16_t res = lduw_le_phys(addr);
+    VMW_SHPRN("SHMEM load16: %" PRIx64 " (value 0x%X)", addr, res);
+    return res;
+}
+
+static inline void
+vmw_shmem_st16(hwaddr addr, uint16_t value)
+{
+    VMW_SHPRN("SHMEM store16: %" PRIx64 " (value 0x%X)", addr, value);
+    stw_le_phys(addr, value);
+}
+
+static inline uint32_t
+vmw_shmem_ld32(hwaddr addr)
+{
+    uint32_t res = ldl_le_phys(addr);
+    VMW_SHPRN("SHMEM load32: %" PRIx64 " (value 0x%X)", addr, res);
+    return res;
+}
+
+static inline void
+vmw_shmem_st32(hwaddr addr, uint32_t value)
+{
+    VMW_SHPRN("SHMEM store32: %" PRIx64 " (value 0x%X)", addr, value);
+    stl_le_phys(addr, value);
+}
+
+static inline uint64_t
+vmw_shmem_ld64(hwaddr addr)
+{
+    uint64_t res = ldq_le_phys(addr);
+    VMW_SHPRN("SHMEM load64: %" PRIx64 " (value %" PRIx64 ")", addr, res);
+    return res;
+}
+
+static inline void
+vmw_shmem_st64(hwaddr addr, uint64_t value)
+{
+    VMW_SHPRN("SHMEM store64: %" PRIx64 " (value %" PRIx64 ")", addr, value);
+    stq_le_phys(addr, value);
+}
+
+/* Macros for simplification of operations on array-style registers */
+
+/*
+ * Whether <addr> lies inside of array-style register defined by <base>,
+ * number of elements (<cnt>) and element size (<regsize>)
+ *
+*/
+#define VMW_IS_MULTIREG_ADDR(addr, base, cnt, regsize)                 \
+    range_covers_byte(base, cnt * regsize, addr)
+
+/*
+ * Returns index of given register (<addr>) in array-style register defined by
+ * <base> and element size (<regsize>)
+ *
+*/
+#define VMW_MULTIREG_IDX_BY_ADDR(addr, base, regsize)                  \
+    (((addr) - (base)) / (regsize))
+
+#endif

hw/vmxnet_debug.h

+/*
+ * QEMU VMWARE VMXNET* paravirtual NICs - debugging facilities
+ *
+ * Copyright (c) 2012 Ravello Systems LTD (http://ravellosystems.com)
+ *
+ * Developed by Daynix Computing LTD (http://www.daynix.com)
+ *
+ * Authors:
+ * Dmitry Fleytman <dmitry@daynix.com>
+ * Tamir Shomer <tamirs@daynix.com>
+ * Yan Vugenfirer <yan@daynix.com>
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ *
+ */
+
+#ifndef _QEMU_VMXNET_DEBUG_H
+#define _QEMU_VMXNET_DEBUG_H
+
+#define VMXNET_DEVICE_NAME "vmxnet3"
+
+/* #define VMXNET_DEBUG_CB */
+#define VMXNET_DEBUG_WARNINGS
+#define VMXNET_DEBUG_ERRORS
+/* #define VMXNET_DEBUG_INTERRUPTS */
+/* #define VMXNET_DEBUG_CONFIG */
+/* #define VMXNET_DEBUG_RINGS */
+/* #define VMXNET_DEBUG_PACKETS */
+/* #define VMXNET_DEBUG_SHMEM_ACCESS */
+
+#ifdef VMXNET_DEBUG_SHMEM_ACCESS
+#define VMW_SHPRN(fmt, ...)                                                   \
+    do {                                                                      \
+        printf("[%s][SH][%s]: " fmt "\n", VMXNET_DEVICE_NAME, __func__,       \
+            ## __VA_ARGS__);                                                  \
+    } while (0)
+#else
+#define VMW_SHPRN(fmt, ...) do {} while (0)
+#endif
+
+#ifdef VMXNET_DEBUG_CB
+#define VMW_CBPRN(fmt, ...)                                                   \
+    do {                                                                      \
+        printf("[%s][CB][%s]: " fmt "\n", VMXNET_DEVICE_NAME, __func__,       \
+            ## __VA_ARGS__);                                                  \
+    } while (0)
+#else
+#define VMW_CBPRN(fmt, ...) do {} while (0)
+#endif
+
+#ifdef VMXNET_DEBUG_PACKETS
+#define VMW_PKPRN(fmt, ...)                                                   \
+    do {                                                                      \
+        printf("[%s][PK][%s]: " fmt "\n", VMXNET_DEVICE_NAME, __func__,       \
+            ## __VA_ARGS__);                                                  \
+    } while (0)
+#else
+#define VMW_PKPRN(fmt, ...) do {} while (0)
+#endif
+
+#ifdef VMXNET_DEBUG_WARNINGS
+#define VMW_WRPRN(fmt, ...)                                                   \
+    do {                                                                      \
+        printf("[%s][WR][%s]: " fmt "\n", VMXNET_DEVICE_NAME, __func__,       \
+            ## __VA_ARGS__);                                                  \
+    } while (0)
+#else
+#define VMW_WRPRN(fmt, ...) do {} while (0)
+#endif
+
+#ifdef VMXNET_DEBUG_ERRORS
+#define VMW_ERPRN(fmt, ...)                                                   \
+    do {                                                                      \
+        printf("[%s][ER][%s]: " fmt "\n", VMXNET_DEVICE_NAME, __func__,       \
+            ## __VA_ARGS__);                                                  \
+    } while (0)
+#else
+#define VMW_ERPRN(fmt, ...) do {} while (0)
+#endif
+
+#ifdef VMXNET_DEBUG_INTERRUPTS
+#define VMW_IRPRN(fmt, ...)                                                   \
+    do {                                                                      \
+        printf("[%s][IR][%s]: " fmt "\n", VMXNET_DEVICE_NAME, __func__,       \
+            ## __VA_ARGS__);                                                  \
+    } while (0)
+#else
+#define VMW_IRPRN(fmt, ...) do {} while (0)
+#endif
+
+#ifdef VMXNET_DEBUG_CONFIG
+#define VMW_CFPRN(fmt, ...)                                                   \
+    do {                                                                      \
+        printf("[%s][CF][%s]: " fmt "\n", VMXNET_DEVICE_NAME, __func__,       \
+            ## __VA_ARGS__);                                                  \
+    } while (0)
+#else
+#define VMW_CFPRN(fmt, ...) do {} while (0)
+#endif
+
+#ifdef VMXNET_DEBUG_RINGS
+#define VMW_RIPRN(fmt, ...)                                                   \
+    do {                                                                      \
+        printf("[%s][RI][%s]: " fmt "\n", VMXNET_DEVICE_NAME, __func__,       \
+            ## __VA_ARGS__);                                                  \
+    } while (0)
+#else
+#define VMW_RIPRN(fmt, ...) do {} while (0)
+#endif
+
+#define VMXNET_MF       "%02X:%02X:%02X:%02X:%02X:%02X"
+#define VMXNET_MA(a)    (a)[0], (a)[1], (a)[2], (a)[3], (a)[4], (a)[5]
+
+#endif /* _QEMU_VMXNET3_DEBUG_H  */

hw/vmxnet_rx_pkt.c

+/*
+ * QEMU VMWARE VMXNET* paravirtual NICs - RX packets abstractions
+ *
+ * Copyright (c) 2012 Ravello Systems LTD (http://ravellosystems.com)
+ *
+ * Developed by Daynix Computing LTD (http://www.daynix.com)
+ *
+ * Authors:
+ * Dmitry Fleytman <dmitry@daynix.com>
+ * Tamir Shomer <tamirs@daynix.com>
+ * Yan Vugenfirer <yan@daynix.com>
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ *
+ */
+
+#include "vmxnet_rx_pkt.h"
+#include "net/eth.h"
+#include "qemu-common.h"
+#include "qemu/iov.h"
+#include "net/checksum.h"
+#include "net/tap.h"
+
+/*
+ * RX packet may contain up to 2 fragments - rebuilt eth header
+ * in case of VLAN tag stripping
+ * and payload received from QEMU - in any case
+ */
+#define VMXNET_MAX_RX_PACKET_FRAGMENTS (2)
+
+struct VmxnetRxPkt {
+    struct virtio_net_hdr virt_hdr;
+    uint8_t ehdr_buf[ETH_MAX_L2_HDR_LEN];
+    struct iovec vec[VMXNET_MAX_RX_PACKET_FRAGMENTS];
+    uint16_t vec_len;
+    uint32_t tot_len;
+    uint16_t tci;
+    bool vlan_stripped;
+    bool has_virt_hdr;
+    eth_pkt_types_e packet_type;
+
+    /* Analysis results */
+    bool isip4;
+    bool isip6;
+    bool isudp;
+    bool istcp;
+};
+
+void vmxnet_rx_pkt_init(struct VmxnetRxPkt **pkt, bool has_virt_hdr)
+{
+    struct VmxnetRxPkt *p = g_malloc0(sizeof *p);
+    p->has_virt_hdr = has_virt_hdr;
+    *pkt = p;
+}
+
+void vmxnet_rx_pkt_uninit(struct VmxnetRxPkt *pkt)
+{
+    g_free(pkt);
+}
+
+struct virtio_net_hdr *vmxnet_rx_pkt_get_vhdr(struct VmxnetRxPkt *pkt)
+{
+    assert(pkt);
+    return &pkt->virt_hdr;
+}
+
+void vmxnet_rx_pkt_attach_data(struct VmxnetRxPkt *pkt, const void *data,
+                               size_t len, bool strip_vlan)
+{
+    uint16_t tci = 0;
+    uint16_t ploff;
+    assert(pkt);
+    pkt->vlan_stripped = false;
+
+    if (strip_vlan) {
+        pkt->vlan_stripped = eth_strip_vlan(data, pkt->ehdr_buf, &ploff, &tci);
+    }
+
+    if (pkt->vlan_stripped) {
+        pkt->vec[0].iov_base = pkt->ehdr_buf;
+        pkt->vec[0].iov_len = ploff - sizeof(struct vlan_header);
+        pkt->vec[1].iov_base = (uint8_t *) data + ploff;
+        pkt->vec[1].iov_len = len - ploff;
+        pkt->vec_len = 2;
+        pkt->tot_len = len - ploff + sizeof(struct eth_header);
+    } else {
+        pkt->vec[0].iov_base = (void *)data;
+        pkt->vec[0].iov_len = len;
+        pkt->vec_len = 1;
+        pkt->tot_len = len;
+    }
+
+    pkt->tci = tci;
+
+    eth_get_protocols(data, len, &pkt->isip4, &pkt->isip6,
+        &pkt->isudp, &pkt->istcp);
+}
+
+void vmxnet_rx_pkt_dump(struct VmxnetRxPkt *pkt)
+{
+#ifdef VMXNET_RX_PKT_DEBUG
+    VmxnetRxPkt *pkt = (VmxnetRxPkt *)pkt;
+    assert(pkt);
+
+    printf("RX PKT: tot_len: %d, vlan_stripped: %d, vlan_tag: %d\n",
+              pkt->tot_len, pkt->vlan_stripped, pkt->tci);
+#endif
+}
+
+void vmxnet_rx_pkt_set_packet_type(struct VmxnetRxPkt *pkt,
+    eth_pkt_types_e packet_type)
+{
+    assert(pkt);
+
+    pkt->packet_type = packet_type;
+
+}
+
+eth_pkt_types_e vmxnet_rx_pkt_get_packet_type(struct VmxnetRxPkt *pkt)
+{
+    assert(pkt);
+
+    return pkt->packet_type;
+}
+
+size_t vmxnet_rx_pkt_get_total_len(struct VmxnetRxPkt *pkt)
+{
+    assert(pkt);
+
+    return pkt->tot_len;
+}
+
+void vmxnet_rx_pkt_get_protocols(struct VmxnetRxPkt *pkt,
+                                 bool *isip4, bool *isip6,
+                                 bool *isudp, bool *istcp)
+{
+    assert(pkt);
+
+    *isip4 = pkt->isip4;
+    *isip6 = pkt->isip6;
+    *isudp = pkt->isudp;
+    *istcp = pkt->istcp;
+}
+
+struct iovec *vmxnet_rx_pkt_get_iovec(struct VmxnetRxPkt *pkt)
+{
+    assert(pkt);
+
+    return pkt->vec;
+}
+
+void vmxnet_rx_pkt_set_vhdr(struct VmxnetRxPkt *pkt,
+                            struct virtio_net_hdr *vhdr)
+{
+    assert(pkt);
+
+    memcpy(&pkt->virt_hdr, vhdr, sizeof pkt->virt_hdr);
+}
+
+bool vmxnet_rx_pkt_is_vlan_stripped(struct VmxnetRxPkt *pkt)
+{
+    assert(pkt);
+
+    return pkt->vlan_stripped;
+}
+
+bool vmxnet_rx_pkt_has_virt_hdr(struct VmxnetRxPkt *pkt)
+{
+    assert(pkt);
+
+    return pkt->has_virt_hdr;
+}
+
+uint16_t vmxnet_rx_pkt_get_num_frags(struct VmxnetRxPkt *pkt)
+{
+    assert(pkt);
+
+    return pkt->vec_len;
+}
+
+uint16_t vmxnet_rx_pkt_get_vlan_tag(struct VmxnetRxPkt *pkt)
+{
+    assert(pkt);
+
+    return pkt->tci;
+}

hw/vmxnet_rx_pkt.h

+/*
+ * QEMU VMWARE VMXNET* paravirtual NICs - RX packets abstraction
+ *
+ * Copyright (c) 2012 Ravello Systems LTD (http://ravellosystems.com)
+ *
+ * Developed by Daynix Computing LTD (http://www.daynix.com)
+ *
+ * Authors:
+ * Dmitry Fleytman <dmitry@daynix.com>
+ * Tamir Shomer <tamirs@daynix.com>
+ * Yan Vugenfirer <yan@daynix.com>
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ *
+ */
+
+#ifndef VMXNET_RX_PKT_H
+#define VMXNET_RX_PKT_H
+
+#include "stdint.h"
+#include "stdbool.h"
+#include "net/eth.h"
+
+/* defines to enable packet dump functions */
+/*#define VMXNET_RX_PKT_DEBUG*/
+
+struct VmxnetRxPkt;
+
+/**
+ * Clean all rx packet resources
+ *
+ * @pkt:            packet
+ *
+ */
+void vmxnet_rx_pkt_uninit(struct VmxnetRxPkt *pkt);
+
+/**
+ * Init function for rx packet functionality
+ *
+ * @pkt:            packet pointer
+ * @has_virt_hdr:   device uses virtio header
+ *
+ */
+void vmxnet_rx_pkt_init(struct VmxnetRxPkt **pkt, bool has_virt_hdr);
+
+/**
+ * returns total length of data attached to rx context
+ *
+ * @pkt:            packet
+ *
+ * Return:  nothing
+ *
+ */
+size_t vmxnet_rx_pkt_get_total_len(struct VmxnetRxPkt *pkt);
+
+/**
+ * fetches packet analysis results
+ *
+ * @pkt:            packet
+ * @isip4:          whether the packet given is IPv4
+ * @isip6:          whether the packet given is IPv6
+ * @isudp:          whether the packet given is UDP
+ * @istcp:          whether the packet given is TCP
+ *
+ */
+void vmxnet_rx_pkt_get_protocols(struct VmxnetRxPkt *pkt,
+                                 bool *isip4, bool *isip6,
+                                 bool *isudp, bool *istcp);
+
+/**
+ * returns virtio header stored in rx context
+ *
+ * @pkt:            packet
+ * @ret:            virtio header
+ *
+ */
+struct virtio_net_hdr *vmxnet_rx_pkt_get_vhdr(struct VmxnetRxPkt *pkt);
+
+/**
+ * returns packet type
+ *
+ * @pkt:            packet
+ * @ret:            packet type
+ *
+ */
+eth_pkt_types_e vmxnet_rx_pkt_get_packet_type(struct VmxnetRxPkt *pkt);
+
+/**
+ * returns vlan tag
+ *
+ * @pkt:            packet
+ * @ret:            VLAN tag
+ *
+ */
+uint16_t vmxnet_rx_pkt_get_vlan_tag(struct VmxnetRxPkt *pkt);
+
+/**
+ * tells whether vlan was stripped from the packet
+ *
+ * @pkt:            packet
+ * @ret:            VLAN stripped sign
+ *
+ */
+bool vmxnet_rx_pkt_is_vlan_stripped(struct VmxnetRxPkt *pkt);
+
+/**
+ * notifies caller if the packet has virtio header
+ *
+ * @pkt:            packet
+ * @ret:            true if packet has virtio header, false otherwize
+ *
+ */
+bool vmxnet_rx_pkt_has_virt_hdr(struct VmxnetRxPkt *pkt);
+
+/**
+ * returns number of frags attached to the packet
+ *
+ * @pkt:            packet
+ * @ret:            number of frags
+ *
+ */
+uint16_t vmxnet_rx_pkt_get_num_frags(struct VmxnetRxPkt *pkt);
+
+/**
+ * attach data to rx packet
+ *
+ * @pkt:            packet
+ * @data:           pointer to the data buffer
+ * @len:            data length
+ * @strip_vlan:     should the module strip vlan from data
+ *
+ */
+void vmxnet_rx_pkt_attach_data(struct VmxnetRxPkt *pkt, const void *data,
+    size_t len, bool strip_vlan);
+
+/**
+ * returns io vector that holds the attached data
+ *
+ * @pkt:            packet
+ * @ret:            pointer to IOVec
+ *
+ */
+struct iovec *vmxnet_rx_pkt_get_iovec(struct VmxnetRxPkt *pkt);
+
+/**
+ * prints rx packet data if debug is enabled
+ *
+ * @pkt:            packet
+ *
+ */
+void vmxnet_rx_pkt_dump(struct VmxnetRxPkt *pkt);
+
+/**
+ * copy passed vhdr data to packet context
+ *
+ * @pkt:            packet
+ * @vhdr:           VHDR buffer
+ *
+ */
+void vmxnet_rx_pkt_set_vhdr(struct VmxnetRxPkt *pkt,
+    struct virtio_net_hdr *vhdr);
+
+/**
+ * save packet type in packet context
+ *
+ * @pkt:            packet
+ * @packet_type:    the packet type
+ *
+ */
+void vmxnet_rx_pkt_set_packet_type(struct VmxnetRxPkt *pkt,
+    eth_pkt_types_e packet_type);
+
+#endif

hw/vmxnet_tx_pkt.c

+/*
+ * QEMU VMWARE VMXNET* paravirtual NICs - TX packets abstractions
+ *
+ * Copyright (c) 2012 Ravello Systems LTD (http://ravellosystems.com)
+ *
+ * Developed by Daynix Computing LTD (http://www.daynix.com)
+ *
+ * Authors:
+ * Dmitry Fleytman <dmitry@daynix.com>
+ * Tamir Shomer <tamirs@daynix.com>
+ * Yan Vugenfirer <yan@daynix.com>
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ *
+ */
+
+#include "vmxnet_tx_pkt.h"
+#include "net/eth.h"
+#include "qemu-common.h"
+#include "qemu/iov.h"
+#include "net/checksum.h"
+#include "net/tap.h"
+#include "net/net.h"
+#include "exec/cpu-common.h"
+
+enum {
+    VMXNET_TX_PKT_VHDR_FRAG = 0,
+    VMXNET_TX_PKT_L2HDR_FRAG,
+    VMXNET_TX_PKT_L3HDR_FRAG,
+    VMXNET_TX_PKT_PL_START_FRAG
+};
+
+/* TX packet private context */
+struct VmxnetTxPkt {
+    struct virtio_net_hdr virt_hdr;
+    bool has_virt_hdr;
+
+    struct iovec *raw;
+    uint32_t raw_frags;
+    uint32_t max_raw_frags;
+
+    struct iovec *vec;
+
+    uint8_t l2_hdr[ETH_MAX_L2_HDR_LEN];
+
+    uint32_t payload_len;
+
+    uint32_t payload_frags;
+    uint32_t max_payload_frags;
+
+    uint16_t hdr_len;
+    eth_pkt_types_e packet_type;
+    uint8_t l4proto;
+};
+
+void vmxnet_tx_pkt_init(struct VmxnetTxPkt **pkt, uint32_t max_frags,
+    bool has_virt_hdr)
+{
+    struct VmxnetTxPkt *p = g_malloc0(sizeof *p);
+
+    p->vec = g_malloc((sizeof *p->vec) *
+        (max_frags + VMXNET_TX_PKT_PL_START_FRAG));
+
+    p->raw = g_malloc((sizeof *p->raw) * max_frags);
+
+    p->max_payload_frags = max_frags;
+    p->max_raw_frags = max_frags;
+    p->has_virt_hdr = has_virt_hdr;
+    p->vec[VMXNET_TX_PKT_VHDR_FRAG].iov_base = &p->virt_hdr;
+    p->vec[VMXNET_TX_PKT_VHDR_FRAG].iov_len =
+        p->has_virt_hdr ? sizeof p->virt_hdr : 0;
+    p->vec[VMXNET_TX_PKT_L2HDR_FRAG].iov_base = &p->l2_hdr;
+    p->vec[VMXNET_TX_PKT_L3HDR_FRAG].iov_base = NULL;
+    p->vec[VMXNET_TX_PKT_L3HDR_FRAG].iov_len = 0;
+
+    *pkt = p;
+}
+
+void vmxnet_tx_pkt_uninit(struct VmxnetTxPkt *pkt)
+{
+    if (pkt) {
+        g_free(pkt->vec);
+        g_free(pkt->raw);
+        g_free(pkt);
+    }
+}
+
+void vmxnet_tx_pkt_update_ip_checksums(struct VmxnetTxPkt *pkt)
+{
+    uint16_t csum;
+    uint32_t ph_raw_csum;
+    assert(pkt);
+    uint8_t gso_type = pkt->virt_hdr.gso_type & ~VIRTIO_NET_HDR_GSO_ECN;
+    struct ip_header *ip_hdr;
+
+    if (VIRTIO_NET_HDR_GSO_TCPV4 != gso_type &&
+        VIRTIO_NET_HDR_GSO_UDP != gso_type) {
+        return;
+    }
+
+    ip_hdr = pkt->vec[VMXNET_TX_PKT_L3HDR_FRAG].iov_base;
+
+    if (pkt->payload_len + pkt->vec[VMXNET_TX_PKT_L3HDR_FRAG].iov_len >
+        ETH_MAX_IP_DGRAM_LEN) {
+        return;
+    }
+
+    ip_hdr->ip_len = cpu_to_be16(pkt->payload_len +
+        pkt->vec[VMXNET_TX_PKT_L3HDR_FRAG].iov_len);
+
+    /* Calculate IP header checksum                    */
+    ip_hdr->ip_sum = 0;
+    csum = net_raw_checksum((uint8_t *)ip_hdr,
+        pkt->vec[VMXNET_TX_PKT_L3HDR_FRAG].iov_len);
+    ip_hdr->ip_sum = cpu_to_be16(csum);
+
+    /* Calculate IP pseudo header checksum             */
+    ph_raw_csum = eth_calc_pseudo_hdr_csum(ip_hdr, pkt->payload_len);
+    csum = cpu_to_be16(~net_checksum_finish(ph_raw_csum));
+    iov_from_buf(&pkt->vec[VMXNET_TX_PKT_PL_START_FRAG], pkt->payload_frags,
+                 pkt->virt_hdr.csum_offset, &csum, sizeof(csum));
+}
+
+static void vmxnet_tx_pkt_calculate_hdr_len(struct VmxnetTxPkt *pkt)
+{
+    pkt->hdr_len = pkt->vec[VMXNET_TX_PKT_L2HDR_FRAG].iov_len +
+        pkt->vec[VMXNET_TX_PKT_L3HDR_FRAG].iov_len;
+}
+
+static bool vmxnet_tx_pkt_parse_headers(struct VmxnetTxPkt *pkt)
+{
+    struct iovec *l2_hdr, *l3_hdr;
+    size_t bytes_read;
+    size_t full_ip6hdr_len;
+    uint16_t l3_proto;
+
+    assert(pkt);
+
+    l2_hdr = &pkt->vec[VMXNET_TX_PKT_L2HDR_FRAG];
+    l3_hdr = &pkt->vec[VMXNET_TX_PKT_L3HDR_FRAG];
+
+    bytes_read = iov_to_buf(pkt->raw, pkt->raw_frags, 0, l2_hdr->iov_base,
+                            ETH_MAX_L2_HDR_LEN);
+    if (bytes_read < ETH_MAX_L2_HDR_LEN) {
+        l2_hdr->iov_len = 0;
+        return false;
+    } else {
+        l2_hdr->iov_len = eth_get_l2_hdr_length(l2_hdr->iov_base);
+    }
+
+    l3_proto = eth_get_l3_proto(l2_hdr->iov_base, l2_hdr->iov_len);
+
+    switch (l3_proto) {
+    case ETH_P_IP:
+        l3_hdr->iov_base = g_malloc(ETH_MAX_IP4_HDR_LEN);
+
+        bytes_read = iov_to_buf(pkt->raw, pkt->raw_frags, l2_hdr->iov_len,
+                                l3_hdr->iov_base, sizeof(struct ip_header));
+
+        if (bytes_read < sizeof(struct ip_header)) {
+            l3_hdr->iov_len = 0;
+            return false;
+        }
+
+        l3_hdr->iov_len = IP_HDR_GET_LEN(l3_hdr->iov_base);
+        pkt->l4proto = ((struct ip_header *) l3_hdr->iov_base)->ip_p;
+
+        /* copy optional IPv4 header data */
+        bytes_read = iov_to_buf(pkt->raw, pkt->raw_frags,
+                                l2_hdr->iov_len + sizeof(struct ip_header),
+                                l3_hdr->iov_base + sizeof(struct ip_header),
+                                l3_hdr->iov_len - sizeof(struct ip_header));
+        if (bytes_read < l3_hdr->iov_len - sizeof(struct ip_header)) {
+            l3_hdr->iov_len = 0;
+            return false;
+        }
+        break;
+
+    case ETH_P_IPV6:
+        if (!eth_parse_ipv6_hdr(pkt->raw, pkt->raw_frags, l2_hdr->iov_len,
+                               &pkt->l4proto, &full_ip6hdr_len)) {
+            l3_hdr->iov_len = 0;
+            return false;
+        }
+
+        l3_hdr->iov_base = g_malloc(full_ip6hdr_len);
+
+        bytes_read = iov_to_buf(pkt->raw, pkt->raw_frags, l2_hdr->iov_len,
+                                l3_hdr->iov_base, full_ip6hdr_len);
+
+        if (bytes_read < full_ip6hdr_len) {
+            l3_hdr->iov_len = 0;
+            return false;
+        } else {
+            l3_hdr->iov_len = full_ip6hdr_len;
+        }
+        break;
+
+    default:
+        l3_hdr->iov_len = 0;
+        break;
+    }
+
+    vmxnet_tx_pkt_calculate_hdr_len(pkt);
+    pkt->packet_type = get_eth_packet_type(l2_hdr->iov_base);
+    return true;
+}
+
+static bool vmxnet_tx_pkt_rebuild_payload(struct VmxnetTxPkt *pkt)
+{
+    size_t payload_len = iov_size(pkt->raw, pkt->raw_frags) - pkt->hdr_len;
+
+    pkt->payload_frags = iov_copy(&pkt->vec[VMXNET_TX_PKT_PL_START_FRAG],
+                                pkt->max_payload_frags,
+                                pkt->raw, pkt->raw_frags,
+                                pkt->hdr_len, payload_len);
+
+    if (pkt->payload_frags != (uint32_t) -1) {
+        pkt->payload_len = payload_len;
+        return true;
+    } else {
+        return false;
+    }
+}
+
+bool vmxnet_tx_pkt_parse(struct VmxnetTxPkt *pkt)
+{
+    return vmxnet_tx_pkt_parse_headers(pkt) &&
+           vmxnet_tx_pkt_rebuild_payload(pkt);
+}
+
+struct virtio_net_hdr *vmxnet_tx_pkt_get_vhdr(struct VmxnetTxPkt *pkt)
+{
+    assert(pkt);
+    return &pkt->virt_hdr;
+}
+
+static uint8_t vmxnet_tx_pkt_get_gso_type(struct VmxnetTxPkt *pkt,
+                                          bool tso_enable)
+{
+    uint8_t rc = VIRTIO_NET_HDR_GSO_NONE;
+    uint16_t l3_proto;
+
+    l3_proto = eth_get_l3_proto(pkt->vec[VMXNET_TX_PKT_L2HDR_FRAG].iov_base,
+        pkt->vec[VMXNET_TX_PKT_L2HDR_FRAG].iov_len);
+
+    if (!tso_enable) {
+        goto func_exit;
+    }
+
+    rc = eth_get_gso_type(l3_proto, pkt->vec[VMXNET_TX_PKT_L3HDR_FRAG].iov_base,
+                          pkt->l4proto);
+
+func_exit:
+    return rc;
+}
+
+void vmxnet_tx_pkt_build_vheader(struct VmxnetTxPkt *pkt, bool tso_enable,
+    bool csum_enable, uint32_t gso_size)
+{
+    struct tcp_hdr l4hdr;
+    assert(pkt);
+
+    /* csum has to be enabled if tso is. */
+    assert(csum_enable || !tso_enable);
+
+    pkt->virt_hdr.gso_type = vmxnet_tx_pkt_get_gso_type(pkt, tso_enable);
+
+    switch (pkt->virt_hdr.gso_type & ~VIRTIO_NET_HDR_GSO_ECN) {
+    case VIRTIO_NET_HDR_GSO_NONE:
+        pkt->virt_hdr.hdr_len = 0;
+        pkt->virt_hdr.gso_size = 0;
+        break;
+
+    case VIRTIO_NET_HDR_GSO_UDP:
+        pkt->virt_hdr.gso_size = IP_FRAG_ALIGN_SIZE(gso_size);
+        pkt->virt_hdr.hdr_len = pkt->hdr_len + sizeof(struct udp_header);
+        break;
+
+    case VIRTIO_NET_HDR_GSO_TCPV4:
+    case VIRTIO_NET_HDR_GSO_TCPV6:
+        iov_to_buf(&pkt->vec[VMXNET_TX_PKT_PL_START_FRAG], pkt->payload_frags,
+                   0, &l4hdr, sizeof(l4hdr));
+        pkt->virt_hdr.hdr_len = pkt->hdr_len + l4hdr.th_off * sizeof(uint32_t);
+        pkt->virt_hdr.gso_size = IP_FRAG_ALIGN_SIZE(gso_size);
+        break;
+
+    default:
+        assert(false);
+    }
+
+    if (csum_enable) {
+        switch (pkt->l4proto) {
+        case IP_PROTO_TCP:
+            pkt->virt_hdr.flags = VIRTIO_NET_HDR_F_NEEDS_CSUM;
+            pkt->virt_hdr.csum_start = pkt->hdr_len;
+            pkt->virt_hdr.csum_offset = offsetof(struct tcp_hdr, th_sum);
+            break;
+        case IP_PROTO_UDP:
+            pkt->virt_hdr.flags = VIRTIO_NET_HDR_F_NEEDS_CSUM;
+            pkt->virt_hdr.csum_start = pkt->hdr_len;
+            pkt->virt_hdr.csum_offset = offsetof(struct udp_hdr, uh_sum);
+            break;
+        default:
+            break;
+        }
+    }
+}
+
+void vmxnet_tx_pkt_setup_vlan_header(struct VmxnetTxPkt *pkt, uint16_t vlan)
+{
+    bool is_new;
+    assert(pkt);
+
+    eth_setup_vlan_headers(pkt->vec[VMXNET_TX_PKT_L2HDR_FRAG].iov_base,
+        vlan, &is_new);
+
+    /* update l2hdrlen */
+    if (is_new) {
+        pkt->hdr_len += sizeof(struct vlan_header);
+        pkt->vec[VMXNET_TX_PKT_L2HDR_FRAG].iov_len +=
+            sizeof(struct vlan_header);
+    }
+}
+
+bool vmxnet_tx_pkt_add_raw_fragment(struct VmxnetTxPkt *pkt, hwaddr pa,
+    size_t len)
+{
+    hwaddr mapped_len = 0;
+    struct iovec *ventry;
+    assert(pkt);
+    assert(pkt->max_raw_frags > pkt->raw_frags);
+
+    if (!len) {
+        return true;
+     }
+
+    ventry = &pkt->raw[pkt->raw_frags];
+    mapped_len = len;
+
+    ventry->iov_base = cpu_physical_memory_map(pa, &mapped_len, false);
+    ventry->iov_len = mapped_len;
+    pkt->raw_frags += !!ventry->iov_base;
+
+    if ((ventry->iov_base == NULL) || (len != mapped_len)) {
+        return false;
+    }
+
+    return true;
+}
+
+eth_pkt_types_e vmxnet_tx_pkt_get_packet_type(struct VmxnetTxPkt *pkt)
+{
+    assert(pkt);
+
+    return pkt->packet_type;
+}
+
+size_t vmxnet_tx_pkt_get_total_len(struct VmxnetTxPkt *pkt)
+{
+    assert(pkt);
+
+    return pkt->hdr_len + pkt->payload_len;
+}
+
+void vmxnet_tx_pkt_dump(struct VmxnetTxPkt *pkt)
+{
+#ifdef VMXNET_TX_PKT_DEBUG
+    assert(pkt);
+
+    printf("TX PKT: hdr_len: %d, pkt_type: 0x%X, l2hdr_len: %lu, "
+        "l3hdr_len: %lu, payload_len: %u\n", pkt->hdr_len, pkt->packet_type,
+        pkt->vec[VMXNET_TX_PKT_L2HDR_FRAG].iov_len,
+        pkt->vec[VMXNET_TX_PKT_L3HDR_FRAG].iov_len, pkt->payload_len);
+#endif
+}
+
+void vmxnet_tx_pkt_reset(struct VmxnetTxPkt *pkt)
+{
+    int i;
+
+    /* no assert, as reset can be called before tx_pkt_init */
+    if (!pkt) {
+        return;
+    }
+
+    memset(&pkt->virt_hdr, 0, sizeof(pkt->virt_hdr));
+
+    g_free(pkt->vec[VMXNET_TX_PKT_L3HDR_FRAG].iov_base);
+    pkt->vec[VMXNET_TX_PKT_L3HDR_FRAG].iov_base = NULL;
+
+    assert(pkt->vec);
+    for (i = VMXNET_TX_PKT_L2HDR_FRAG;
+         i < pkt->payload_frags + VMXNET_TX_PKT_PL_START_FRAG; i++) {
+        pkt->vec[i].iov_len = 0;
+    }
+    pkt->payload_len = 0;
+    pkt->payload_frags = 0;
+
+    assert(pkt->raw);
+    for (i = 0; i < pkt->raw_frags; i++) {
+        assert(pkt->raw[i].iov_base);
+        cpu_physical_memory_unmap(pkt->raw[i].iov_base, pkt->raw[i].iov_len,
+                                  false, pkt->raw[i].iov_len);
+        pkt->raw[i].iov_len = 0;
+    }
+    pkt->raw_frags = 0;
+
+    pkt->hdr_len = 0;
+    pkt->packet_type = 0;
+    pkt->l4proto = 0;
+}
+
+static void vmxnet_tx_pkt_do_sw_csum(struct VmxnetTxPkt *pkt)
+{
+    struct iovec *iov = &pkt->vec[VMXNET_TX_PKT_L2HDR_FRAG];
+    uint32_t csum_cntr;
+    uint16_t csum = 0;
+    /* num of iovec without vhdr */
+    uint32_t iov_len = pkt->payload_frags + VMXNET_TX_PKT_PL_START_FRAG - 1;
+    uint16_t csl;
+    struct ip_header *iphdr;
+    size_t csum_offset = pkt->virt_hdr.csum_start + pkt->virt_hdr.csum_offset;
+
+    /* Put zero to checksum field */
+    iov_from_buf(iov, iov_len, csum_offset, &csum, sizeof csum);
+
+    /* Calculate L4 TCP/UDP checksum */
+    csl = pkt->payload_len;
+
+    /* data checksum */
+    csum_cntr =
+        net_checksum_add_iov(iov, iov_len, pkt->virt_hdr.csum_start, csl);
+    /* add pseudo header to csum */
+    iphdr = pkt->vec[VMXNET_TX_PKT_L3HDR_FRAG].iov_base;
+    csum_cntr += eth_calc_pseudo_hdr_csum(iphdr, csl);
+
+    /* Put the checksum obtained into the packet */
+    csum = cpu_to_be16(net_checksum_finish(csum_cntr));
+    iov_from_buf(iov, iov_len, csum_offset, &csum, sizeof csum);
+}
+
+enum {
+    VMXNET_TX_PKT_FRAGMENT_L2_HDR_POS = 0,
+    VMXNET_TX_PKT_FRAGMENT_L3_HDR_POS,
+    VMXNET_TX_PKT_FRAGMENT_HEADER_NUM
+};
+
+#define VMXNET_MAX_FRAG_SG_LIST (64)
+
+static size_t vmxnet_tx_pkt_fetch_fragment(struct VmxnetTxPkt *pkt,
+    int *src_idx, size_t *src_offset, struct iovec *dst, int *dst_idx)
+{
+    size_t fetched = 0;
+    struct iovec *src = pkt->vec;
+
+    *dst_idx = VMXNET_TX_PKT_FRAGMENT_HEADER_NUM;
+
+    while (fetched < pkt->virt_hdr.gso_size) {
+
+        /* no more place in fragment iov */
+        if (*dst_idx == VMXNET_MAX_FRAG_SG_LIST) {
+            break;
+        }
+
+        /* no more data in iovec */
+        if (*src_idx == (pkt->payload_frags + VMXNET_TX_PKT_PL_START_FRAG)) {
+            break;
+        }
+
+
+        dst[*dst_idx].iov_base = src[*src_idx].iov_base + *src_offset;
+        dst[*dst_idx].iov_len = MIN(src[*src_idx].iov_len - *src_offset,
+            pkt->virt_hdr.gso_size - fetched);
+
+        *src_offset += dst[*dst_idx].iov_len;
+        fetched += dst[*dst_idx].iov_len;
+
+        if (*src_offset == src[*src_idx].iov_len) {
+            *src_offset = 0;
+            (*src_idx)++;
+        }
+
+        (*dst_idx)++;
+    }
+
+    return fetched;
+}
+
+static bool vmxnet_tx_pkt_do_sw_fragmentation(struct VmxnetTxPkt *pkt,
+    NetClientState *nc)
+{
+    struct iovec fragment[VMXNET_MAX_FRAG_SG_LIST];
+    size_t fragment_len = 0;
+    bool more_frags = false;
+
+    /* some pointers for shorter code */
+    void *l2_iov_base, *l3_iov_base;
+    size_t l2_iov_len, l3_iov_len;
+    int src_idx =  VMXNET_TX_PKT_PL_START_FRAG, dst_idx;
+    size_t src_offset = 0;
+    size_t fragment_offset = 0;
+
+    l2_iov_base = pkt->vec[VMXNET_TX_PKT_L2HDR_FRAG].iov_base;
+    l2_iov_len = pkt->vec[VMXNET_TX_PKT_L2HDR_FRAG].iov_len;
+    l3_iov_base = pkt->vec[VMXNET_TX_PKT_L3HDR_FRAG].iov_base;
+    l3_iov_len = pkt->vec[VMXNET_TX_PKT_L3HDR_FRAG].iov_len;
+
+    /* Copy headers */
+    fragment[VMXNET_TX_PKT_FRAGMENT_L2_HDR_POS].iov_base = l2_iov_base;
+    fragment[VMXNET_TX_PKT_FRAGMENT_L2_HDR_POS].iov_len = l2_iov_len;
+    fragment[VMXNET_TX_PKT_FRAGMENT_L3_HDR_POS].iov_base = l3_iov_base;
+    fragment[VMXNET_TX_PKT_FRAGMENT_L3_HDR_POS].iov_len = l3_iov_len;
+
+
+    /* Put as much data as possible and send */
+    do {
+        fragment_len = vmxnet_tx_pkt_fetch_fragment(pkt, &src_idx, &src_offset,
+            fragment, &dst_idx);
+
+        more_frags = (fragment_offset + fragment_len < pkt->payload_len);
+
+        eth_setup_ip4_fragmentation(l2_iov_base, l2_iov_len, l3_iov_base,
+            l3_iov_len, fragment_len, fragment_offset, more_frags);
+
+        eth_fix_ip4_checksum(l3_iov_base, l3_iov_len);
+
+        qemu_sendv_packet(nc, fragment, dst_idx);
+
+        fragment_offset += fragment_len;
+
+    } while (more_frags);
+
+    return true;
+}
+
+bool vmxnet_tx_pkt_send(struct VmxnetTxPkt *pkt, NetClientState *nc)
+{
+    assert(pkt);
+
+    if (!pkt->has_virt_hdr &&
+        pkt->virt_hdr.flags & VIRTIO_NET_HDR_F_NEEDS_CSUM) {
+        vmxnet_tx_pkt_do_sw_csum(pkt);
+    }
+
+    /*
+     * Since underlying infrastructure does not support IP datagrams longer
+     * than 64K we should drop such packets and don't even try to send
+     */
+    if (VIRTIO_NET_HDR_GSO_NONE != pkt->virt_hdr.gso_type) {
+        if (pkt->payload_len >
+            ETH_MAX_IP_DGRAM_LEN -
+            pkt->vec[VMXNET_TX_PKT_L3HDR_FRAG].iov_len) {
+            return false;
+        }
+    }
+
+    if (pkt->has_virt_hdr ||
+        pkt->virt_hdr.gso_type == VIRTIO_NET_HDR_GSO_NONE) {
+        qemu_sendv_packet(nc, pkt->vec,
+            pkt->payload_frags + VMXNET_TX_PKT_PL_START_FRAG);
+        return true;
+    }
+
+    return vmxnet_tx_pkt_do_sw_fragmentation(pkt, nc);
+}

hw/vmxnet_tx_pkt.h

+/*
+ * QEMU VMWARE VMXNET* paravirtual NICs - TX packets abstraction
+ *
+ * Copyright (c) 2012 Ravello Systems LTD (http://ravellosystems.com)
+ *
+ * Developed by Daynix Computing LTD (http://www.daynix.com)
+ *
+ * Authors:
+ * Dmitry Fleytman <dmitry@daynix.com>
+ * Tamir Shomer <tamirs@daynix.com>
+ * Yan Vugenfirer <yan@daynix.com>
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ *
+ */
+
+#ifndef VMXNET_TX_PKT_H
+#define VMXNET_TX_PKT_H
+
+#include "stdint.h"
+#include "stdbool.h"
+#include "net/eth.h"
+#include "exec/hwaddr.h"
+
+/* define to enable packet dump functions */
+/*#define VMXNET_TX_PKT_DEBUG*/
+
+struct VmxnetTxPkt;
+
+/**
+ * Init function for tx packet functionality
+ *
+ * @pkt:            packet pointer
+ * @max_frags:      max tx ip fragments
+ * @has_virt_hdr:   device uses virtio header.
+ */
+void vmxnet_tx_pkt_init(struct VmxnetTxPkt **pkt, uint32_t max_frags,
+    bool has_virt_hdr);
+
+/**
+ * Clean all tx packet resources.
+ *
+ * @pkt:            packet.
+ */
+void vmxnet_tx_pkt_uninit(struct VmxnetTxPkt *pkt);
+
+/**
+ * get virtio header
+ *
+ * @pkt:            packet
+ * @ret:            virtio header
+ */
+struct virtio_net_hdr *vmxnet_tx_pkt_get_vhdr(struct VmxnetTxPkt *pkt);
+
+/**
+ * build virtio header (will be stored in module context)
+ *
+ * @pkt:            packet
+ * @tso_enable:     TSO enabled
+ * @csum_enable:    CSO enabled
+ * @gso_size:       MSS size for TSO
+ *
+ */
+void vmxnet_tx_pkt_build_vheader(struct VmxnetTxPkt *pkt, bool tso_enable,
+    bool csum_enable, uint32_t gso_size);
+
+/**
+ * updates vlan tag, and adds vlan header in case it is missing
+ *
+ * @pkt:            packet
+ * @vlan:           VLAN tag
+ *
+ */
+void vmxnet_tx_pkt_setup_vlan_header(struct VmxnetTxPkt *pkt, uint16_t vlan);
+
+/**
+ * populate data fragment into pkt context.
+ *
+ * @pkt:            packet
+ * @pa:             physical address of fragment
+ * @len:            length of fragment
+ *
+ */
+bool vmxnet_tx_pkt_add_raw_fragment(struct VmxnetTxPkt *pkt, hwaddr pa,
+    size_t len);
+
+/**
+ * fix ip header fields and calculate checksums needed.
+ *
+ * @pkt:            packet
+ *
+ */
+void vmxnet_tx_pkt_update_ip_checksums(struct VmxnetTxPkt *pkt);
+
+/**
+ * get length of all populated data.
+ *
+ * @pkt:            packet
+ * @ret:            total data length
+ *
+ */
+size_t vmxnet_tx_pkt_get_total_len(struct VmxnetTxPkt *pkt);
+
+/**
+ * get packet type
+ *
+ * @pkt:            packet
+ * @ret:            packet type
+ *
+ */
+eth_pkt_types_e vmxnet_tx_pkt_get_packet_type(struct VmxnetTxPkt *pkt);
+
+/**
+ * prints packet data if debug is enabled
+ *
+ * @pkt:            packet
+ *
+ */
+void vmxnet_tx_pkt_dump(struct VmxnetTxPkt *pkt);
+
+/**
+ * reset tx packet private context (needed to be called between packets)
+ *
+ * @pkt:            packet
+ *
+ */
+void vmxnet_tx_pkt_reset(struct VmxnetTxPkt *pkt);
+
+/**
+ * Send packet to qemu. handles sw offloads if vhdr is not supported.
+ *
+ * @pkt:            packet
+ * @nc:             NetClientState
+ * @ret:            operation result
+ *
+ */
+bool vmxnet_tx_pkt_send(struct VmxnetTxPkt *pkt, NetClientState *nc);
+
+/**
+ * parse raw packet data and analyze offload requirements.
+ *
+ * @pkt:            packet
+ *
+ */
+bool vmxnet_tx_pkt_parse(struct VmxnetTxPkt *pkt);
+
+#endif

include/net/checksum.h

@@ -20,10 +20,34 @@
 
 #include <stdint.h>
 
-uint32_t net_checksum_add(int len, uint8_t *buf);
+uint32_t net_checksum_add_cont(int len, uint8_t *buf, int seq);
 uint16_t net_checksum_finish(uint32_t sum);
 uint16_t net_checksum_tcpudp(uint16_t length, uint16_t proto,
                              uint8_t *addrs, uint8_t *buf);
 void net_checksum_calculate(uint8_t *data, int length);
 
+static inline uint32_t
+net_checksum_add(int len, uint8_t *buf)
+{
+    return net_checksum_add_cont(len, buf, 0);
+}
+
+static inline uint16_t
+net_raw_checksum(uint8_t *data, int length)
+{
+    return net_checksum_finish(net_checksum_add(length, data));
+}
+
+/**
+ * net_checksum_add_iov: scatter-gather vector checksumming
+ *
+ * @iov: input scatter-gather array
+ * @iov_cnt: number of array elements
+ * @iov_off: starting iov offset for checksumming
+ * @size: length of data to be checksummed
+ */
+uint32_t net_checksum_add_iov(const struct iovec *iov,
+                              const unsigned int iov_cnt,
+                              uint32_t iov_off, uint32_t size);
+
 #endif /* QEMU_NET_CHECKSUM_H */

include/net/eth.h

+/*
+ * QEMU network structures definitions and helper functions
+ *
+ * Copyright (c) 2012 Ravello Systems LTD (http://ravellosystems.com)
+ *
+ * Developed by Daynix Computing LTD (http://www.daynix.com)
+ *
+ * Portions developed by Free Software Foundation, Inc
+ * Copyright (C) 1991-1997, 2001, 2003, 2006 Free Software Foundation, Inc.
+ * See netinet/ip6.h and netinet/in.h (GNU C Library)
+ *
+ * Portions developed by Igor Kovalenko
+ * Copyright (c) 2006 Igor Kovalenko
+ * See hw/rtl8139.c (QEMU)
+ *
+ * Authors:
+ * Dmitry Fleytman <dmitry@daynix.com>
+ * Tamir Shomer <tamirs@daynix.com>
+ * Yan Vugenfirer <yan@daynix.com>
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ *
+ */
+
+#ifndef QEMU_ETH_H
+#define QEMU_ETH_H
+
+#include <sys/types.h>
+#include <string.h>
+#include "qemu/bswap.h"
+#include "qemu/iov.h"
+
+#define ETH_ALEN 6
+
+struct eth_header {
+    uint8_t  h_dest[ETH_ALEN];   /* destination eth addr */
+    uint8_t  h_source[ETH_ALEN]; /* source ether addr    */
+    uint16_t h_proto;            /* packet type ID field */
+};
+
+struct vlan_header {
+    uint16_t  h_tci;     /* priority and VLAN ID  */
+    uint16_t  h_proto;   /* encapsulated protocol */
+};
+
+struct ip_header {
+    uint8_t  ip_ver_len;     /* version and header length */
+    uint8_t  ip_tos;         /* type of service */
+    uint16_t ip_len;         /* total length */
+    uint16_t ip_id;          /* identification */
+    uint16_t ip_off;         /* fragment offset field */
+    uint8_t  ip_ttl;         /* time to live */
+    uint8_t  ip_p;           /* protocol */
+    uint16_t ip_sum;         /* checksum */
+    uint32_t ip_src, ip_dst; /* source and destination address */
+};
+
+typedef struct tcp_header {
+    uint16_t th_sport;          /* source port */
+    uint16_t th_dport;          /* destination port */
+    uint32_t th_seq;            /* sequence number */
+    uint32_t th_ack;            /* acknowledgment number */
+    uint16_t th_offset_flags;   /* data offset, reserved 6 bits, */
+                                /* TCP protocol flags */
+    uint16_t th_win;            /* window */
+    uint16_t th_sum;            /* checksum */
+    uint16_t th_urp;            /* urgent pointer */
+} tcp_header;
+
+typedef struct udp_header {
+    uint16_t uh_sport; /* source port */
+    uint16_t uh_dport; /* destination port */
+    uint16_t uh_ulen;  /* udp length */
+    uint16_t uh_sum;   /* udp checksum */
+} udp_header;
+
+typedef struct ip_pseudo_header {
+    uint32_t ip_src;
+    uint32_t ip_dst;
+    uint8_t  zeros;
+    uint8_t  ip_proto;
+    uint16_t ip_payload;
+} ip_pseudo_header;
+
+/* IPv6 address */
+struct in6_addr {
+    union {
+        uint8_t __u6_addr8[16];
+    } __in6_u;
+};
+
+struct ip6_header {
+    union {
+        struct ip6_hdrctl {
+            uint32_t ip6_un1_flow; /* 4 bits version, 8 bits TC,
+                                      20 bits flow-ID */
+            uint16_t ip6_un1_plen; /* payload length */
+            uint8_t  ip6_un1_nxt;  /* next header */
+            uint8_t  ip6_un1_hlim; /* hop limit */
+        } ip6_un1;
+        uint8_t ip6_un2_vfc;       /* 4 bits version, top 4 bits tclass */
+        struct ip6_ecn_access {
+            uint8_t  ip6_un3_vfc;  /* 4 bits version, top 4 bits tclass */
+            uint8_t  ip6_un3_ecn;  /* 2 bits ECN, top 6 bits payload length */
+        } ip6_un3;
+    } ip6_ctlun;
+    struct in6_addr ip6_src;     /* source address */
+    struct in6_addr ip6_dst;     /* destination address */
+};
+
+struct ip6_ext_hdr {
+    uint8_t        ip6r_nxt;   /* next header */
+    uint8_t        ip6r_len;   /* length in units of 8 octets */
+};
+
+struct udp_hdr {
+  uint16_t uh_sport;           /* source port */
+  uint16_t uh_dport;           /* destination port */
+  uint16_t uh_ulen;            /* udp length */
+  uint16_t uh_sum;             /* udp checksum */
+};
+
+struct tcp_hdr {
+    u_short     th_sport;   /* source port */
+    u_short     th_dport;   /* destination port */
+    uint32_t    th_seq;     /* sequence number */
+    uint32_t    th_ack;     /* acknowledgment number */
+#ifdef HOST_WORDS_BIGENDIAN
+    u_char  th_off : 4,     /* data offset */
+        th_x2:4;            /* (unused) */
+#else
+    u_char  th_x2 : 4,      /* (unused) */
+        th_off:4;           /* data offset */
+#endif
+
+#define TH_ELN  0x1 /* explicit loss notification */
+#define TH_ECN  0x2 /* explicit congestion notification */
+#define TH_FS   0x4 /* fast start */
+
+    u_char  th_flags;
+#define TH_FIN  0x01
+#define TH_SYN  0x02
+#define TH_RST  0x04
+#define TH_PUSH 0x08
+#define TH_ACK  0x10
+#define TH_URG  0x20
+    u_short th_win;      /* window */
+    u_short th_sum;      /* checksum */
+    u_short th_urp;      /* urgent pointer */
+};
+
+#define ip6_nxt      ip6_ctlun.ip6_un1.ip6_un1_nxt
+#define ip6_ecn_acc  ip6_ctlun.ip6_un3.ip6_un3_ecn
+
+#define PKT_GET_ETH_HDR(p)        \
+    ((struct eth_header *)(p))
+#define PKT_GET_VLAN_HDR(p)       \
+    ((struct vlan_header *) (((uint8_t *)(p)) + sizeof(struct eth_header)))
+#define PKT_GET_DVLAN_HDR(p)       \
+    (PKT_GET_VLAN_HDR(p) + 1)
+#define PKT_GET_IP_HDR(p)         \
+    ((struct ip_header *)(((uint8_t *)(p)) + eth_get_l2_hdr_length(p)))
+#define IP_HDR_GET_LEN(p)         \
+    ((((struct ip_header *)p)->ip_ver_len & 0x0F) << 2)
+#define PKT_GET_IP_HDR_LEN(p)     \
+    (IP_HDR_GET_LEN(PKT_GET_IP_HDR(p)))
+#define PKT_GET_IP6_HDR(p)        \
+    ((struct ip6_header *) (((uint8_t *)(p)) + eth_get_l2_hdr_length(p)))
+#define IP_HEADER_VERSION(ip)     \
+    ((ip->ip_ver_len >> 4)&0xf)
+
+#define ETH_P_IP                  (0x0800)
+#define ETH_P_IPV6                (0x86dd)
+#define ETH_P_VLAN                (0x8100)
+#define ETH_P_DVLAN               (0x88a8)
+#define VLAN_VID_MASK             0x0fff
+#define IP_HEADER_VERSION_4       (4)
+#define IP_HEADER_VERSION_6       (6)
+#define IP_PROTO_TCP              (6)
+#define IP_PROTO_UDP              (17)
+#define IPTOS_ECN_MASK            0x03
+#define IPTOS_ECN(x)              ((x) & IPTOS_ECN_MASK)
+#define IPTOS_ECN_CE              0x03
+#define IP6_ECN_MASK              0xC0
+#define IP6_ECN(x)                ((x) & IP6_ECN_MASK)
+#define IP6_ECN_CE                0xC0
+#define IP4_DONT_FRAGMENT_FLAG    (1 << 14)
+
+#define IS_SPECIAL_VLAN_ID(x)     \
+    (((x) == 0) || ((x) == 0xFFF))
+
+#define ETH_MAX_L2_HDR_LEN  \
+    (sizeof(struct eth_header) + 2 * sizeof(struct vlan_header))
+
+#define ETH_MAX_IP4_HDR_LEN   (60)
+#define ETH_MAX_IP_DGRAM_LEN  (0xFFFF)
+
+#define IP_FRAG_UNIT_SIZE     (8)
+#define IP_FRAG_ALIGN_SIZE(x) ((x) & ~0x7)
+#define IP_RF                 0x8000           /* reserved fragment flag */
+#define IP_DF                 0x4000           /* don't fragment flag */
+#define IP_MF                 0x2000           /* more fragments flag */
+#define IP_OFFMASK            0x1fff           /* mask for fragmenting bits */
+
+#define IP6_EXT_GRANULARITY   (8)  /* Size granularity for
+                                      IPv6 extension headers */
+
+/* IP6 extension header types */
+#define IP6_HOP_BY_HOP        (0)
+#define IP6_ROUTING           (43)
+#define IP6_FRAGMENT          (44)
+#define IP6_ESP               (50)
+#define IP6_AUTHENTICATION    (51)
+#define IP6_NONE              (59)
+#define IP6_DESTINATON        (60)
+#define IP6_MOBILITY          (135)
+
+static inline int is_multicast_ether_addr(const uint8_t *addr)
+{
+    return 0x01 & addr[0];
+}
+
+static inline int is_broadcast_ether_addr(const uint8_t *addr)
+{
+    return (addr[0] & addr[1] & addr[2] & addr[3] & addr[4] & addr[5]) == 0xff;
+}
+
+static inline int is_unicast_ether_addr(const uint8_t *addr)
+{
+    return !is_multicast_ether_addr(addr);
+}
+
+typedef enum {
+    ETH_PKT_UCAST = 0xAABBCC00,
+    ETH_PKT_BCAST,
+    ETH_PKT_MCAST
+} eth_pkt_types_e;
+
+static inline eth_pkt_types_e
+get_eth_packet_type(const struct eth_header *ehdr)
+{
+    if (is_broadcast_ether_addr(ehdr->h_dest)) {
+        return ETH_PKT_BCAST;
+    } else if (is_multicast_ether_addr(ehdr->h_dest)) {
+        return ETH_PKT_MCAST;
+    } else { /* unicast */
+        return ETH_PKT_UCAST;
+    }
+}
+
+static inline uint32_t
+eth_get_l2_hdr_length(const void *p)
+{
+    uint16_t proto = be16_to_cpu(PKT_GET_ETH_HDR(p)->h_proto);
+    struct vlan_header *hvlan = PKT_GET_VLAN_HDR(p);
+    switch (proto) {
+    case ETH_P_VLAN:
+        return sizeof(struct eth_header) + sizeof(struct vlan_header);
+    case ETH_P_DVLAN:
+        if (hvlan->h_proto == ETH_P_VLAN) {
+            return sizeof(struct eth_header) + 2 * sizeof(struct vlan_header);
+        } else {
+            return sizeof(struct eth_header) + sizeof(struct vlan_header);
+        }
+    default:
+        return sizeof(struct eth_header);
+    }
+}
+
+static inline uint16_t
+eth_get_pkt_tci(const void *p)
+{
+    uint16_t proto = be16_to_cpu(PKT_GET_ETH_HDR(p)->h_proto);
+    struct vlan_header *hvlan = PKT_GET_VLAN_HDR(p);
+    switch (proto) {
+    case ETH_P_VLAN:
+    case ETH_P_DVLAN:
+        return be16_to_cpu(hvlan->h_tci);
+    default:
+        return 0;
+    }
+}
+
+static inline bool
+eth_strip_vlan(const void *p, uint8_t *new_ehdr_buf,
+               uint16_t *payload_offset, uint16_t *tci)
+{
+    uint16_t proto = be16_to_cpu(PKT_GET_ETH_HDR(p)->h_proto);
+    struct vlan_header *hvlan = PKT_GET_VLAN_HDR(p);
+    struct eth_header *new_ehdr = (struct eth_header *) new_ehdr_buf;
+
+    switch (proto) {
+    case ETH_P_VLAN:
+    case ETH_P_DVLAN:
+        memcpy(new_ehdr->h_source, PKT_GET_ETH_HDR(p)->h_source, ETH_ALEN);
+        memcpy(new_ehdr->h_dest, PKT_GET_ETH_HDR(p)->h_dest, ETH_ALEN);
+        new_ehdr->h_proto = hvlan->h_proto;
+        *tci = be16_to_cpu(hvlan->h_tci);
+        *payload_offset =
+            sizeof(struct eth_header) + sizeof(struct vlan_header);
+        if (be16_to_cpu(new_ehdr->h_proto) == ETH_P_VLAN) {
+            memcpy(PKT_GET_VLAN_HDR(new_ehdr),
+                   PKT_GET_DVLAN_HDR(p),
+                   sizeof(struct vlan_header));
+            *payload_offset += sizeof(struct vlan_header);
+        }
+        return true;
+    default:
+        return false;
+    }
+}
+
+static inline uint16_t
+eth_get_l3_proto(const void *l2hdr, size_t l2hdr_len)
+{
+    uint8_t *proto_ptr = (uint8_t *) l2hdr + l2hdr_len - sizeof(uint16_t);
+    return be16_to_cpup((uint16_t *)proto_ptr);
+}
+
+void eth_setup_vlan_headers(struct eth_header *ehdr, uint16_t vlan_tag,
+    bool *is_new);
+
+uint8_t eth_get_gso_type(uint16_t l3_proto, uint8_t *l3_hdr, uint8_t l4proto);
+
+void eth_get_protocols(const uint8_t *headers,
+                       uint32_t hdr_length,
+                       bool *isip4, bool *isip6,
+                       bool *isudp, bool *istcp);
+
+void eth_setup_ip4_fragmentation(const void *l2hdr, size_t l2hdr_len,
+                                 void *l3hdr, size_t l3hdr_len,
+                                 size_t l3payload_len,
+                                 size_t frag_offset, bool more_frags);
+
+void
+eth_fix_ip4_checksum(void *l3hdr, size_t l3hdr_len);
+
+uint32_t
+eth_calc_pseudo_hdr_csum(struct ip_header *iphdr, uint16_t csl);
+
+bool
+eth_parse_ipv6_hdr(struct iovec *pkt, int pkt_frags,
+                   size_t ip6hdr_off, uint8_t *l4proto,
+                   size_t *full_hdr_len);
+
+#endif

include/net/net.h

@@ -11,6 +11,11 @@
 
 #define MAX_QUEUE_NUM 1024
 
+/* Maximum GSO packet size (64k) plus plenty of room for
+ * the ethernet and virtio_net headers
+ */
+#define NET_BUFSIZE (4096 + 65536)
+
 struct MACAddr {
     uint8_t a[6];
 };

net/Makefile.objs

 common-obj-y = net.o queue.o checksum.o util.o hub.o
 common-obj-y += socket.o
 common-obj-y += dump.o
+common-obj-y += eth.o
 common-obj-$(CONFIG_POSIX) += tap.o
 common-obj-$(CONFIG_LINUX) += tap-linux.o
 common-obj-$(CONFIG_WIN32) += tap-win32.o

net/net.c

@@ -497,7 +497,7 @@ ssize_t qemu_send_packet_raw(NetClientState *nc, const uint8_t *buf, int size)
 static ssize_t nc_sendv_compat(NetClientState *nc, const struct iovec *iov,
                                int iovcnt)
 {
-    uint8_t buffer[4096];
+    uint8_t buffer[NET_BUFSIZE];
     size_t offset;
 
     offset = iov_to_buf(iov, iovcnt, 0, buffer, sizeof(buffer));

net/vde.c

@@ -39,7 +39,7 @@ typedef struct VDEState {
 static void vde_to_qemu(void *opaque)
 {
     VDEState *s = opaque;
-    uint8_t buf[4096];
+    uint8_t buf[NET_BUFSIZE];
     int size;
 
     size = vde_recv(s->vde, (char *)buf, sizeof(buf), 0);