From patchwork Wed Feb 29 12:43:53 2012 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Dmitry Fleytman X-Patchwork-Id: 143717 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@bilbo.ozlabs.org Received: from lists.gnu.org (lists.gnu.org [208.118.235.17]) (using TLSv1 with cipher AES256-SHA (256/256 bits)) (Client did not present a certificate) by ozlabs.org (Postfix) with ESMTPS id 75934B6FA2 for ; Wed, 29 Feb 2012 23:48:35 +1100 (EST) Received: from localhost ([::1]:53209 helo=lists.gnu.org) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1S2ixM-0007l2-Sc for incoming@patchwork.ozlabs.org; Wed, 29 Feb 2012 07:48:32 -0500 Received: from eggs.gnu.org ([208.118.235.92]:60992) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1S2iwp-0007kB-Ow for qemu-devel@nongnu.org; Wed, 29 Feb 2012 07:48:18 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1S2iwI-0006zi-5E for qemu-devel@nongnu.org; Wed, 29 Feb 2012 07:47:59 -0500 Received: from na3sys010aog101.obsmtp.com ([74.125.245.70]:46888) by eggs.gnu.org with smtp (Exim 4.71) (envelope-from ) id 1S2iwH-0006z2-CY for qemu-devel@nongnu.org; Wed, 29 Feb 2012 07:47:26 -0500 Received: from mail-wi0-f174.google.com ([209.85.212.174]) (using TLSv1) by na3sys010aob101.postini.com ([74.125.244.12]) with SMTP ID DSNKT04eWZipngkZ9gQ/3yTymP2FdmdOXjoG@postini.com; Wed, 29 Feb 2012 04:47:24 PST Received: by mail-wi0-f174.google.com with SMTP id hm11so3420925wib.5 for ; Wed, 29 Feb 2012 04:47:21 -0800 (PST) Received-SPF: pass (google.com: domain of dmitry.fleytman@ravellosystems.com designates 10.180.83.97 as permitted sender) client-ip=10.180.83.97; Authentication-Results: mr.google.com; spf=pass (google.com: domain of dmitry.fleytman@ravellosystems.com designates 10.180.83.97 as permitted sender) smtp.mail=dmitry.fleytman@ravellosystems.com Received: from mr.google.com ([10.180.83.97]) by 10.180.83.97 with SMTP id p1mr436997wiy.19.1330519641457 (num_hops = 1); Wed, 29 Feb 2012 04:47:21 -0800 (PST) MIME-Version: 1.0 Received: by 10.180.83.97 with SMTP id p1mr355714wiy.19.1330519641291; Wed, 29 Feb 2012 04:47:21 -0800 (PST) Received: from localhost.localdomain ([95.35.0.236]) by mx.google.com with ESMTPS id gm1sm36647817wib.7.2012.02.29.04.47.16 (version=TLSv1/SSLv3 cipher=OTHER); Wed, 29 Feb 2012 04:47:19 -0800 (PST) From: Dmitry Fleytman To: qemu-devel@nongnu.org Date: Wed, 29 Feb 2012 14:43:53 +0200 Message-Id: <1330519433-20536-1-git-send-email-dmitry.fleytman@ravellosystems.com> X-Mailer: git-send-email 1.7.7.6 X-Gm-Message-State: ALoCoQnVdt6kM0Tc+fsfLGJ2hFil8MeXBPwGMtayCR1Jldqauh4/q6p9gR3zEwurBZLnlgSkbwwZ X-detected-operating-system: by eggs.gnu.org: GNU/Linux 2.6, seldom 2.4 (older, 4) X-Received-From: 74.125.245.70 Cc: Anthony Liguori , Alex Fishman , "Michael S. Tsirkin" , yvugenfi@redhat.com, Izik Eidus , Yan Vugenfirer , Dmitry Fleytman Subject: [Qemu-devel] (no subject) X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.14 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org From ac51c01697ee2dc2ff2cb8c24346054ae0e3e32a Mon Sep 17 00:00:00 2001 From: Dmitry Fleytman Date: Sun, 26 Feb 2012 18:39:07 +0200 Subject: [PATCH v2] VMXNET3 paravirtual NIC device implementation Implementation of VMWare VMXNET3 paravirtual NIC device. Supports of all the device features including offload capabilties, VLANs and etc. The device is tested on different OSes: Fedora 15 Ubuntu 10.4 Centos 6.2 Windows 2008R2 Windows 2008 64bit Windows 2008 32bit Windows 2003 64bit Windows 2003 32bit Currently live migration is not supported. Signed-off-by: Dmitry Fleytman Signed-off-by: Yan Vugenfirer --- qemu/Makefile.objs | 1 + qemu/default-configs/pci.mak | 1 + qemu/hw/pci.c | 2 + qemu/hw/pci.h | 1 + qemu/hw/virtio-net.h | 13 +- qemu/hw/vmware_utils.h | 131 +++ qemu/hw/vmxnet3.c | 2559 ++++++++++++++++++++++++++++++++++++++++++ qemu/hw/vmxnet3.h | 727 ++++++++++++ qemu/hw/vmxnet3_debug.h | 104 ++ qemu/hw/vmxnet_utils.c | 172 +++ qemu/hw/vmxnet_utils.h | 242 ++++ qemu/net.c | 2 +- qemu/net/checksum.h | 7 + 13 files changed, 3955 insertions(+), 7 deletions(-) create mode 100644 qemu/hw/vmware_utils.h create mode 100644 qemu/hw/vmxnet3.c create mode 100644 qemu/hw/vmxnet3.h create mode 100644 qemu/hw/vmxnet3_debug.h create mode 100644 qemu/hw/vmxnet_utils.c create mode 100644 qemu/hw/vmxnet_utils.h diff --git a/qemu/Makefile.objs b/qemu/Makefile.objs index 808de6a..3f846a6 100644 --- a/qemu/Makefile.objs +++ b/qemu/Makefile.objs @@ -264,6 +264,7 @@ hw-obj-$(CONFIG_PCNET_PCI) += pcnet-pci.o hw-obj-$(CONFIG_PCNET_COMMON) += pcnet.o hw-obj-$(CONFIG_E1000_PCI) += e1000.o hw-obj-$(CONFIG_RTL8139_PCI) += rtl8139.o +hw-obj-$(CONFIG_VMXNET3_PCI) += vmxnet3.o vmxnet_utils.o hw-obj-$(CONFIG_SMC91C111) += smc91c111.o hw-obj-$(CONFIG_LAN9118) += lan9118.o diff --git a/qemu/default-configs/pci.mak b/qemu/default-configs/pci.mak index 21e4ccf..f8e6ee1 100644 --- a/qemu/default-configs/pci.mak +++ b/qemu/default-configs/pci.mak @@ -13,6 +13,7 @@ CONFIG_PCNET_COMMON=y CONFIG_LSI_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 diff --git a/qemu/hw/pci.c b/qemu/hw/pci.c index bf046bf..f0fb1ee 100644 --- a/qemu/hw/pci.c +++ b/qemu/hw/pci.c @@ -1350,6 +1350,7 @@ static const char * const pci_nic_models[] = { "e1000", "pcnet", "virtio", + "vmxnet3", NULL }; @@ -1362,6 +1363,7 @@ static const char * const pci_nic_names[] = { "e1000", "pcnet", "virtio-net-pci", + "vmxnet3", NULL }; diff --git a/qemu/hw/pci.h b/qemu/hw/pci.h index 4f19fdb..fee8250 100644 --- a/qemu/hw/pci.h +++ b/qemu/hw/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 diff --git a/qemu/hw/virtio-net.h b/qemu/hw/virtio-net.h index 4468741..fa3c17b 100644 --- a/qemu/hw/virtio-net.h +++ b/qemu/hw/virtio-net.h @@ -78,13 +78,14 @@ struct virtio_net_config * 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 +#define VIRTIO_NET_HDR_F_NEEDS_CSUM 1 /* Use csum_start, csum_offset */ +#define VIRTIO_NET_HDR_F_DATA_VALID 2 /* Csum is valid */ 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 +#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; diff --git a/qemu/hw/vmware_utils.h b/qemu/hw/vmware_utils.h new file mode 100644 index 0000000..304bb48 --- /dev/null +++ b/qemu/hw/vmware_utils.h @@ -0,0 +1,131 @@ +/* + * 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 + * Yan Vugenfirer + * + * This work is licensed under the terms of the GNU GPL, version 2 or later. + * See the COPYING file in the top-level directory. + * + */ + +/* Shared memory access functions with byte swap support */ +static inline void +vmw_shmem_read(target_phys_addr_t addr, void *buf, int len) +{ + DSHPRINTF("SHMEM r: %" PRIx64 ", len: %d to %p", + (uint64_t) addr, len, buf); + cpu_physical_memory_read(addr, buf, len); +} + +static inline void +vmw_shmem_write(target_phys_addr_t addr, void *buf, int len) +{ + DSHPRINTF("SHMEM w: %" PRIx64 ", len: %d to %p", + (uint64_t) addr, len, buf); + cpu_physical_memory_write(addr, buf, len); +} + +static inline void +vmw_shmem_rw(target_phys_addr_t addr, void *buf, int len, int is_write) +{ + DSHPRINTF("SHMEM r/w: %" PRIx64 ", len: %d (to %p), is write: %d", + (uint64_t) addr, len, buf, is_write); + + cpu_physical_memory_rw(addr, buf, len, is_write); +} + +static inline void +vmw_shmem_set(target_phys_addr_t addr, uint8 val, int len) +{ + int i; + DSHPRINTF("SHMEM set: %" PRIx64 ", len: %d (value 0x%X)", + (uint64_t) addr, len, val); + + for (i = 0; i < len; i++) { + cpu_physical_memory_write(addr + i, &val, 1); + } +} + +static inline uint32_t +vmw_shmem_ld8(target_phys_addr_t addr) +{ + uint8_t res = ldub_phys(addr); + DSHPRINTF("SHMEM load8: %" PRIx64 " (value 0x%X)", + (uint64_t) addr, res); + return res; +} + +static inline void +vmw_shmem_st8(target_phys_addr_t addr, uint8_t value) +{ + DSHPRINTF("SHMEM store8: %" PRIx64 " (value 0x%X)", + (uint64_t) addr, value); + stb_phys(addr, value); +} + +static inline uint32_t +vmw_shmem_ld16(target_phys_addr_t addr) +{ + uint16_t res = lduw_le_phys(addr); + DSHPRINTF("SHMEM load16: %" PRIx64 " (value 0x%X)", + (uint64_t) addr, res); + return res; +} + +static inline void +vmw_shmem_st16(target_phys_addr_t addr, uint16_t value) +{ + DSHPRINTF("SHMEM store16: %" PRIx64 " (value 0x%X)", + (uint64_t) addr, value); + stw_le_phys(addr, value); +} + +static inline uint32_t +vmw_shmem_ld32(target_phys_addr_t addr) +{ + uint32_t res = ldl_le_phys(addr); + DSHPRINTF("SHMEM load32: %" PRIx64 " (value 0x%X)", + (uint64_t) addr, res); + return res; +} + +static inline void +vmw_shmem_st32(target_phys_addr_t addr, uint32_t value) +{ + DSHPRINTF("SHMEM store32: %" PRIx64 " (value 0x%X)", + (uint64_t) addr, value); + stl_le_phys(addr, value); +} + +static inline uint64_t +vmw_shmem_ld64(target_phys_addr_t addr) +{ + uint64_t res = ldq_le_phys(addr); + DSHPRINTF("SHMEM load64: %" PRIx64 " (value %" PRIx64 ")", + (uint64_t) addr, res); + return res; +} + +static inline void +vmw_shmem_st64(target_phys_addr_t addr, uint64_t value) +{ + DSHPRINTF("SHMEM store64: %" PRIx64 " (value %" PRIx64 ")", + (uint64_t) addr, value); + stq_le_phys(addr, value); +} + +/* MACROS for simplification of operations on array-style registers */ +#define IS_MULTIREG_ADDR(addr, base, cnt, regsize) \ + (((addr) >= (base)) && ((addr) < (base) + (cnt) * (regsize))) + +#define MULTIREG_IDX_BY_ADDR(addr, base, regsize) \ + (((addr) - (base)) / (regsize)) + +/* Bitfields */ +#define FLAG_IS_SET(field, flag) (((field) & (flag)) == (flag)) diff --git a/qemu/hw/vmxnet3.c b/qemu/hw/vmxnet3.c new file mode 100644 index 0000000..112d3b9 --- /dev/null +++ b/qemu/hw/vmxnet3.c @@ -0,0 +1,2559 @@ +/* + * QEMU VMWARE VMXNET3 paravirtual NIC + * + * Copyright (c) 2012 Ravello Systems LTD (http://ravellosystems.com) + * + * Developed by Daynix Computing LTD (http://www.daynix.com) + * + * Authors: + * Dmitry Fleytman + * Yan Vugenfirer + * + * This work is licensed under the terms of the GNU GPL, version 2 or later. + * See the COPYING file in the top-level directory. + * + */ + +#define VMXNET3_ENABLE_MSIX +#define VMXNET3_ENABLE_MSI + +/* Define this constant to non-zero to enable IP4 */ +/* fragmentation feature */ + +/* #define VMXNET_MAX_IP_PLOAD_LEN ETH_MAX_IP_PLOAD_LEN */ +#define VMXNET3_MAX_IP_PLOAD_LEN 0 + +#include "hw.h" +#include "pci.h" +#include "net.h" +#include "virtio-net.h" +#include "net/tap.h" +#include "sysemu.h" +#include "iov.h" +#include "bswap.h" +#ifdef VMXNET3_ENABLE_MSIX +#include "msix.h" +#endif +#ifdef VMXNET3_ENABLE_MSI +#include "msi.h" +#endif + +#include "vmxnet3_debug.h" +#include "vmxnet3.h" +#include "vmware_utils.h" +#include "vmxnet_utils.h" + +#define PCI_DEVICE_ID_VMWARE_VMXNET3_REVISION 0x1 +#define VMXNET3_MSIX_BAR_SIZE 0x2000 + +#define VMXNET3_BAR0_IDX (0) +#define VMXNET3_BAR1_IDX (1) +#define VMXNET3_MSIX_BAR_IDX (2) + +/* Link speed in Mbps should be shifted by 16 */ +#define VMXNET3_LINK_SPEED (1000 << 16) + +/* Link status: 1 - up, 0 - down. */ +#define VMXNET3_LINK_STATUS_UP 0x1 + +/* Least significant bit should be set for revision and version */ +#define VMXNET3_DEVICE_VERSION 0x1 +#define VMXNET3_DEVICE_REVISION 0x1 + +/* Macros for rings descriptors access */ +#define VMXNET3_READ_TX_QUEUE_DESCR8(dpa, field) \ + (vmw_shmem_ld8(dpa + offsetof(struct Vmxnet3_TxQueueDesc, field))) + +#define VMXNET3_WRITE_TX_QUEUE_DESCR8(dpa, field, value) \ + (vmw_shmem_st8(dpa + offsetof(struct Vmxnet3_TxQueueDesc, field, value))) + +#define VMXNET3_READ_TX_QUEUE_DESCR32(dpa, field) \ + (vmw_shmem_ld32(dpa + offsetof(struct Vmxnet3_TxQueueDesc, field))) + +#define VMXNET3_WRITE_TX_QUEUE_DESCR32(dpa, field, value) \ + (vmw_shmem_st32(dpa + offsetof(struct Vmxnet3_TxQueueDesc, field), value)) + +#define VMXNET3_READ_TX_QUEUE_DESCR64(dpa, field) \ + (vmw_shmem_ld64(dpa + offsetof(struct Vmxnet3_TxQueueDesc, field))) + +#define VMXNET3_WRITE_TX_QUEUE_DESCR64(dpa, field, value) \ + (vmw_shmem_st64(dpa + offsetof(struct Vmxnet3_TxQueueDesc, field), value)) + +#define VMXNET3_READ_RX_QUEUE_DESCR64(dpa, field) \ + (vmw_shmem_ld64(dpa + offsetof(struct Vmxnet3_RxQueueDesc, field))) + +#define VMXNET3_READ_RX_QUEUE_DESCR32(dpa, field) \ + (vmw_shmem_ld32(dpa + offsetof(struct Vmxnet3_RxQueueDesc, field))) + +#define VMXNET3_WRITE_RX_QUEUE_DESCR64(dpa, field, value) \ + (vmw_shmem_st64(dpa + offsetof(struct Vmxnet3_RxQueueDesc, field), value)) + +#define VMXNET3_WRITE_RX_QUEUE_DESCR8(dpa, field, value) \ + (vmw_shmem_st8(dpa + offsetof(struct Vmxnet3_RxQueueDesc, field), value)) + +/* Macros for guest driver shared area access */ +#define VMXNET3_READ_DRV_SHARED64(shpa, field) \ + (vmw_shmem_ld64(shpa + offsetof(struct Vmxnet3_DriverShared, field))) + +#define VMXNET3_READ_DRV_SHARED32(shpa, field) \ + (vmw_shmem_ld32(shpa + offsetof(struct Vmxnet3_DriverShared, field))) + +#define VMXNET3_WRITE_DRV_SHARED32(shpa, field, val) \ + (vmw_shmem_st32(shpa + offsetof(struct Vmxnet3_DriverShared, field), val)) + +#define VMXNET3_READ_DRV_SHARED16(shpa, field) \ + (vmw_shmem_ld16(shpa + offsetof(struct Vmxnet3_DriverShared, field))) + +#define VMXNET3_READ_DRV_SHARED8(shpa, field) \ + (vmw_shmem_ld8(shpa + offsetof(struct Vmxnet3_DriverShared, field))) + +#define VMXNET3_READ_DRV_SHARED(shpa, field, b, l) \ + (vmw_shmem_read(shpa + offsetof(struct Vmxnet3_DriverShared, field), b, l)) + +/* TX/RX packets abstractions */ +typedef struct Vmxnet3_TxPktMdata { + uint32_t offload_mode; + uint32_t cso_or_gso_size; + uint32_t hdr_length; + eth_pkt_types_e packet_type; +} Vmxnet3_TxPktMdata; + +typedef struct _Vmxnet3_RxPktMdata { + uint32_t tot_len; + uint16_t vlan_tag; + bool vlan_stripped; + bool vhdr_valid; + eth_pkt_types_e packet_type; +} Vmxnet3_RxPktMdata; + +#define VMXNET3_TXPKT_REBUILT_HDR_LEN (1024) + +typedef struct _Vmxnet3_TxPkt { + Vmxnet3_TxPktMdata mdata; + struct virtio_net_hdr virt_hdr; + bool has_virt_hdr; + + struct iovec *vec; + + uint8_t __l2_hdr[ETH_MAX_L2_HDR_LEN]; + uint8_t __l3_hdr[ETH_MAX_L3_HDR_LEN]; + + uint32_t payload_len; + uint32_t max_payload_len; + + uint32_t payload_frags; + uint32_t max_payload_frags; + + struct { + uint32_t offset; + bool more_frags; + bool orig_more_frags; + } fragmentation; +} Vmxnet3_TxPkt; + +#define VMXNET3_TXPKT_VHDR_FRAG (0) +#define VMXNET3_TXPKT_L2HDR_FRAG (1) +#define VMXNET3_TXPKT_L3HDR_FRAG (2) +#define VMXNET3_TXPKT_PL_START_FRAG (3) + +#define vmxnet3_txpkt_get_mdata(p) (&((p)->mdata)) +#define vmxnet3_txpkt_get_vhdr(p) (&((p)->virt_hdr)) + +#define vmxnet3_txpkt_get_l2hdr(p) \ + ((p)->vec[VMXNET3_TXPKT_L2HDR_FRAG].iov_base) +#define vmxnet3_txpkt_get_l2hdr_len(p) \ + ((p)->vec[VMXNET3_TXPKT_L2HDR_FRAG].iov_len) +#define vmxnet3_txpkt_set_l2hdr_len(p, l) \ + ((p)->vec[VMXNET3_TXPKT_L2HDR_FRAG].iov_len = l) +#define vmxnet3_txpkt_get_l3hdr(p) \ + ((p)->vec[VMXNET3_TXPKT_L3HDR_FRAG].iov_base) +#define vmxnet3_txpkt_get_l3hdr_len(p) \ + ((p)->vec[VMXNET3_TXPKT_L3HDR_FRAG].iov_len) +#define vmxnet3_txpkt_set_l3hdr_len(p, l) \ + ((p)->vec[VMXNET3_TXPKT_L3HDR_FRAG].iov_len = l) +#define vmxnet3_txpkt_get_payload_len(p) \ + ((p)->payload_len) + +#define vmxnet3_txpkt_set_more_frags(p, mf) \ + ((p)->fragmentation.more_frags = mf) +#define vmxnet3_txpkt_get_more_frags(p) \ + ((p)->fragmentation.more_frags | \ + (p)->fragmentation.orig_more_frags) +#define vmxnet3_txpkt_set_frag_off(p, off) \ + ((p)->fragmentation.offset = off) +#define vmxnet3_txpkt_get_frag_off(p) \ + ((p)->fragmentation.offset) +#define vmxnet3_txpkt_advance_frag_off(p, off) \ + ((p)->fragmentation.offset += off) + +static inline size_t +vmxnet3_txpkt_get_total_len(const Vmxnet3_TxPkt *p) +{ + return vmxnet3_txpkt_get_l2hdr_len(p) + + vmxnet3_txpkt_get_l3hdr_len(p) + + vmxnet3_txpkt_get_payload_len(p); +} + +static inline struct iovec* +vmxnet3_txpkt_get_payload_frag(Vmxnet3_TxPkt *p, uint32_t num) +{ + assert(num < p->max_payload_frags); + return &p->vec[num + VMXNET3_TXPKT_PL_START_FRAG]; +} + +static inline void +vmxnet3_txpkt_set_num_pl_frags(Vmxnet3_TxPkt *p, uint32_t num) +{ + assert(num <= p->max_payload_frags); + p->payload_frags = num; +} + +static inline void +vmxnet3_txpkt_reset_payload(Vmxnet3_TxPkt *p) +{ + p->payload_len = 0; +} + +static void vmxnet3_txpkt_reset(Vmxnet3_TxPkt *p) +{ + memset(&p->mdata, 0, sizeof(p->mdata)); + vmxnet3_txpkt_set_num_pl_frags(p, 0); + vmxnet3_txpkt_reset_payload(p); + vmxnet3_txpkt_set_more_frags(p, 0); + vmxnet3_txpkt_set_frag_off(p, 0); + p->max_payload_len = 0; + + if (NULL != p->vec) { + p->vec[VMXNET3_TXPKT_L2HDR_FRAG].iov_len = 0; + p->vec[VMXNET3_TXPKT_L3HDR_FRAG].iov_len = 0; + } +} + +static bool +vmxnet3_txpkt_prealloc(Vmxnet3_TxPkt *p, uint32_t max_frags, bool has_virt_hdr) +{ + if (NULL != p->vec) { + g_free(p->vec); + } + + p->vec = + g_malloc(sizeof(*p->vec) * (max_frags + VMXNET3_TXPKT_PL_START_FRAG)); + if (NULL == p->vec) { + return false; + } + + p->max_payload_frags = max_frags; + p->has_virt_hdr = has_virt_hdr; + p->vec[VMXNET3_TXPKT_VHDR_FRAG].iov_base = &p->virt_hdr; + p->vec[VMXNET3_TXPKT_VHDR_FRAG].iov_len = + p->has_virt_hdr ? sizeof(p->virt_hdr) : 0; + p->vec[VMXNET3_TXPKT_L2HDR_FRAG].iov_base = &p->__l2_hdr; + p->vec[VMXNET3_TXPKT_L3HDR_FRAG].iov_base = &p->__l3_hdr; + vmxnet3_txpkt_reset(p); + return true; +} + +static void vmxnet3_txpkt_init(Vmxnet3_TxPkt *p) +{ + p->vec = NULL; +} + +static void vmxnet3_txpkt_cleanup(Vmxnet3_TxPkt *p) +{ + g_free(p->vec); +} + +static void vmxnet3_txpkt_unmap(Vmxnet3_TxPkt *p, bool is_write) +{ + int i; + + for (i = VMXNET3_TXPKT_PL_START_FRAG; + i < p->payload_frags + VMXNET3_TXPKT_PL_START_FRAG; i++) { + cpu_physical_memory_unmap(p->vec[i].iov_base, p->vec[i].iov_len, + is_write, p->vec[i].iov_len); + } +} + +static void* +vmxnet3_txpkt_map(Vmxnet3_TxPkt *p, uint32_t *mapped_fragments, bool is_write) +{ + int i; + + for (i = VMXNET3_TXPKT_PL_START_FRAG; + i < p->payload_frags + VMXNET3_TXPKT_PL_START_FRAG; i++) { + target_phys_addr_t mapped_len = p->vec[i].iov_len; + size_t orig_len = p->vec[i].iov_len; + p->vec[i].iov_base = + cpu_physical_memory_map((uint64_t) p->vec[i].iov_base, + &mapped_len, is_write); + p->vec[i].iov_len = mapped_len; + + if ((NULL == p->vec[i].iov_base) || (orig_len != mapped_len)) { + p->payload_frags = i + !!p->vec[i].iov_base; + vmxnet3_txpkt_unmap(p, is_write); + return NULL; + } + } + + *mapped_fragments = VMXNET3_TXPKT_PL_START_FRAG + p->payload_frags; + return p->vec; +} + +static inline void +vmxnet3_txpkt_dump(Vmxnet3_TxPkt *p) +{ +#ifdef DEBUG_VMXNET3_PACKETS + Vmxnet3_TxPktMdata *m = vmxnet3_txpkt_get_mdata(p); +#endif + + DPKPRINTF("TXPKT MDATA: om: %d, cso/gso_size: %d, hdr_len: %d, " + "pkt_type: 0x%X, l2hdr_len: %lu l3hdr_len: %lu, payload_len: %u", + m->offload_mode, m->cso_or_gso_size, + m->hdr_length, m->packet_type, + vmxnet3_txpkt_get_l2hdr_len(p), + vmxnet3_txpkt_get_l3hdr_len(p), + vmxnet3_txpkt_get_payload_len(p)); +}; + +/* 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 VMXNET3_MAX_RX_PACKET_FRAGMENTS (2) + +typedef struct _Vmxnet3_RxPkt { + Vmxnet3_RxPktMdata mdata; + struct virtio_net_hdr virt_hdr; + struct eth_header eth_hdr; + struct iovec vec[VMXNET3_MAX_RX_PACKET_FRAGMENTS]; + uint16 vec_len; +} Vmxnet3_RxPkt; + +#define vmxnet3_rxpkt_get_mdata(p) (&((p)->mdata)) +#define vmxnet3_rxpkt_get_ehdr(p) (&((p)->eth_hdr)) +#define vmxnet3_rxpkt_get_vhdr(p) (&((p)->virt_hdr)) +#define vmxnet3_rxpkt_get_frag(p, n) (&((p)->vec[(n)])) +#define vmxnet3_rxpkt_set_num_frags(p, n) ((p)->vec_len = (n)) +#define vmxnet3_rxpkt_get_num_frags(p) ((p)->vec_len) + +static inline void vmxnet3_rxpkt_attach_ehdr(Vmxnet3_RxPkt *p) +{ + vmxnet3_rxpkt_get_frag(p, 0)->iov_base = &p->eth_hdr; + vmxnet3_rxpkt_get_frag(p, 0)->iov_len = sizeof(p->eth_hdr); +} + +static inline void vmxnet3_rxpkt_reset(Vmxnet3_RxPkt *p) +{ + memset(&p->mdata, 0, sizeof(p->mdata)); + memset(&p->virt_hdr, 0, sizeof(p->virt_hdr)); + vmxnet3_rxpkt_set_num_frags(p, 0); +} + +static void vmxnet3_rxpkt_init(Vmxnet3_RxPkt *p) +{ + vmxnet3_rxpkt_reset(p); +} + +static inline void +vmxnet3_rxpkt_dump(Vmxnet3_RxPkt *p) +{ +#ifdef DEBUG_VMXNET3_PACKETS + Vmxnet3_RxPktMdata *m = vmxnet3_rxpkt_get_mdata(p); +#endif + + DPKPRINTF("RXPKT MDATA: tot_len: %d, pkt_type: 0x%X, " + "vlan_stripped: %d, vlan_tag: %d, vhdr_valid: %d", + m->tot_len, m->packet_type, + m->vlan_stripped, m->vlan_tag, m->vhdr_valid); +}; + +/* Cyclic ring abstraction */ +typedef struct _Vmxnet3_Ring { + target_phys_addr_t pa; + size_t size; + size_t cell_size; + size_t next; + uint8_t gen; +} Vmxnet3_Ring; + +static inline void vmxnet3_ring_init(Vmxnet3_Ring *ring, + target_phys_addr_t pa, + size_t size, + size_t cell_size, + bool zero_region) +{ + ring->pa = pa; + ring->size = size; + ring->cell_size = cell_size; + ring->gen = VMXNET3_INIT_GEN; + ring->next = 0; + + if (zero_region) { + vmw_shmem_set(pa, 0, size*cell_size); + } +} + +#define vmxnet3_ring_dump(macro, ring_name, ridx, r) \ + macro("%s#%d: base %" PRIx64 " size %lu cell_size %lu gen %d next %lu", \ + (ring_name), (ridx), (uint64_t) (r)->pa, \ + (r)->size, (r)->cell_size, (r)->gen, (r)->next) + +static inline void vmxnet3_ring_inc(Vmxnet3_Ring *ring) +{ + if (++ring->next >= ring->size) { + ring->next = 0; + ring->gen ^= 1; + } +} + +static inline void vmxnet3_ring_dec(Vmxnet3_Ring *ring) +{ + if (0 == ring->next--) { + ring->next = ring->size - 1; + ring->gen ^= 1; + } +} + +static inline target_phys_addr_t vmxnet3_ring_curr_cell_pa(Vmxnet3_Ring *ring) +{ + return ring->pa + ring->next * ring->cell_size; +} + +static inline void vmxnet3_ring_read_curr_cell(Vmxnet3_Ring *ring, void *buff) +{ + vmw_shmem_read(vmxnet3_ring_curr_cell_pa(ring), buff, ring->cell_size); +} + +static inline void vmxnet3_ring_write_curr_cell(Vmxnet3_Ring *ring, void *buff) +{ + vmw_shmem_write(vmxnet3_ring_curr_cell_pa(ring), buff, ring->cell_size); +} + +static inline size_t vmxnet3_ring_curr_cell_idx(Vmxnet3_Ring *ring) +{ + return ring->next; +} + +static inline uint8_t vmxnet3_ring_curr_gen(Vmxnet3_Ring *ring) +{ + return ring->gen; +} + +/* Debug trace-related functions */ +static inline void +vmxnet3_dump_tx_descr(struct Vmxnet3_TxDesc *descr) +{ + DPKPRINTF("TX DESCR: " + "addr %" PRIx64 ", len: %d, gen: %d, rsvd: %d, " + "dtype: %d, ext1: %d, msscof: %d, hlen: %d, om: %d, " + "eop: %d, cq: %d, ext2: %d, ti: %d, tci: %d", + le64_to_cpu(descr->addr), descr->len, descr->gen, descr->rsvd, + descr->dtype, descr->ext1, descr->msscof, descr->hlen, descr->om, + descr->eop, descr->cq, descr->ext2, descr->ti, descr->tci); +} + +static inline void +vmxnet3_dump_virt_hdr(struct virtio_net_hdr *vhdr) +{ + DPKPRINTF("VHDR: flags 0x%x, gso_type: 0x%x, hdr_len: %d, gso_size: %d, " + "csum_start: %d, csum_offset: %d", + vhdr->flags, vhdr->gso_type, vhdr->hdr_len, vhdr->gso_size, + vhdr->csum_start, vhdr->csum_offset); +} + +static inline void +vmxnet3_dump_rx_descr(struct Vmxnet3_RxDesc *descr) +{ + DPKPRINTF("RX DESCR: addr %" PRIx64 ", len: %d, gen: %d, rsvd: %d, " + "dtype: %d, ext1: %d, btype: %d", + le64_to_cpu(descr->addr), descr->len, descr->gen, + descr->rsvd, descr->dtype, descr->ext1, descr->btype); +} + +/* Device state and helper functions */ +#define VMXNET3_RX_RINGS_PER_QUEUE (2) + +typedef struct _VMXNET3_State { + PCIDevice dev; + NICState *nic; + NICConf conf; + MemoryRegion bar0; + MemoryRegion bar1; + MemoryRegion msix_bar; + +#ifdef VMXNET3_ENABLE_MSIX + /* Whether MSI-X support was installed successfully */ + uint8_t msix_used; +#endif +#ifdef VMXNET3_ENABLE_MSI + /* Whether MSI support was installed successfully */ + uint8_t msi_used; +#endif + + target_phys_addr_t drv_shmem; + target_phys_addr_t temp_shared_guest_driver_memory; + + uint8_t txq_num; + struct { + Vmxnet3_Ring tx_ring; + Vmxnet3_Ring comp_ring; + + uint8_t intr_idx; + target_phys_addr_t tx_stats_pa; + struct UPT1_TxStats txq_stats; + } txq_descr[VMXNET3_DEVICE_MAX_TX_QUEUES]; + + /* This boolean tells whether RX packet being indicated has to */ + /* be split into head and body chunks from different RX rings */ + bool rx_packets_compound; + + bool rx_vlan_stripping; + bool lro_supported; + + uint8_t rxq_num; + struct { + Vmxnet3_Ring rx_ring[VMXNET3_RX_RINGS_PER_QUEUE]; + Vmxnet3_Ring comp_ring; + uint8_t intr_idx; + target_phys_addr_t rx_stats_pa; + struct UPT1_RxStats rxq_stats; + } rxq_descr[VMXNET3_DEVICE_MAX_RX_QUEUES]; + + /* Network MTU */ + uint32_t mtu; + + /* Maximum number of fragments for indicated TX packets */ + uint32_t max_tx_frags; + + /* Maximum number of fragments for indicated RX packets */ + uint16_t max_rx_frags; + + /* Index for events interrupt */ + uint8_t event_int_idx; + + /* Whether automatic interrupts masking enabled */ + uint8_t auto_int_masking; + + bool peer_has_vhdr; + + /* TX packets to QEMU interface */ + Vmxnet3_TxPkt curr_txpkt; + size_t curr_txpkt_pl_frags; + bool curr_txpkt_skip; + bool curr_txpkt_header_processed; + + uint32_t device_active; + uint32_t last_command; + + uint32_t link_status_and_speed; + + struct { + bool is_masked; + bool is_pending; + bool is_asserted; + } interrupt_states[VMXNET3_MAX_INTRS]; + + uint32_t temp_mac; /* To store the low part first */ + + MACAddr perm_mac; + uint32_t vlan_table[VMXNET3_VFT_SIZE]; + uint32_t rx_mode; + MACAddr *mcast_list; + uint16_t mcast_list_len; +} VMXNET3_State; + +/* Interrupt management */ + +/* This function returns sign whether interrupt line is in asserted state */ +/* This depends on the type of interrupt used. For INTX interrupt line will */ +/* be asserted until explicit deassertion, for MSI(X) interrupt line will */ +/* be deasserted automatically due to notifiction symantics of the MSI(X) */ +/* interrupts */ +static bool _vmxnet3_assert_interrupt_line(VMXNET3_State *s, uint32_t int_idx) +{ +#ifdef VMXNET3_ENABLE_MSIX + if (s->msix_used && msix_enabled(&s->dev)) { + DIRPRINTF("Sending MSI-X notification for vector %u", int_idx); + msix_notify(&s->dev, int_idx); + return false; + } +#endif +#ifdef VMXNET3_ENABLE_MSI + if (s->msi_used && msi_enabled(&s->dev)) { + DIRPRINTF("Sending MSI notification for vector %u", int_idx); + msi_notify(&s->dev, int_idx); + return false; + } +#endif + + DIRPRINTF("Asserting line for interrupt %u", int_idx); + qemu_set_irq(s->dev.irq[int_idx], 1); + return true; +} + +static void _vmxnet3_deassert_interrupt_line(VMXNET3_State *s, int lidx) +{ +#ifdef VMXNET3_ENABLE_MSIX + /* This function should never be called for MSI(X) interrupts */ + /* because deassertion never required for message interrupts */ + assert(!s->msix_used || !msix_enabled(&s->dev)); +#endif +#ifdef VMXNET3_ENABLE_MSI + /* This function should never be called for MSI(X) interrupts */ + /* because deassertion never required for message interrupts */ + assert(!s->msi_used || !msi_enabled(&s->dev)); +#endif + + DIRPRINTF("Deasserting line for interrupt %u", lidx); + qemu_set_irq(s->dev.irq[lidx], 0); +} + +static void vmxnet3_update_interrupt_line_state(VMXNET3_State *s, int lidx) +{ + if (!s->interrupt_states[lidx].is_pending && + s->interrupt_states[lidx].is_asserted) { + DIRPRINTF("New interrupt line state for index %d is DOWN", lidx); + _vmxnet3_deassert_interrupt_line(s, lidx); + s->interrupt_states[lidx].is_asserted = false; + return; + } + + if (s->interrupt_states[lidx].is_pending && + !s->interrupt_states[lidx].is_masked && + !s->interrupt_states[lidx].is_asserted) { + DIRPRINTF("New interrupt line state for index %d is UP", lidx); + s->interrupt_states[lidx].is_asserted = + _vmxnet3_assert_interrupt_line(s, lidx); + s->interrupt_states[lidx].is_pending = false; + return; + } +} + +static void vmxnet3_trigger_interrupt(VMXNET3_State *s, int lidx) +{ + s->interrupt_states[lidx].is_pending = true; + vmxnet3_update_interrupt_line_state(s, lidx); + +#ifdef VMXNET3_ENABLE_MSIX + if (s->msix_used && msix_enabled(&s->dev) && s->auto_int_masking) { + goto do_automask; + } +#endif + +#ifdef VMXNET3_ENABLE_MSI + if (s->msi_used && msi_enabled(&s->dev) && s->auto_int_masking) { + goto do_automask; + } +#endif + + return; + +do_automask: + s->interrupt_states[lidx].is_masked = true; + vmxnet3_update_interrupt_line_state(s, lidx); +} + +static bool vmxnet3_interrupt_asserted(VMXNET3_State *s, int lidx) +{ + return s->interrupt_states[lidx].is_asserted; +} + +static void vmxnet3_clear_interrupt(VMXNET3_State *s, int int_idx) +{ + s->interrupt_states[int_idx].is_pending = false; + if (s->auto_int_masking) { + s->interrupt_states[int_idx].is_masked = true; + } + vmxnet3_update_interrupt_line_state(s, int_idx); +} + +static void +vmxnet3_on_interrupt_mask_changed(VMXNET3_State *s, int lidx, bool is_masked) +{ + s->interrupt_states[lidx].is_masked = is_masked; + vmxnet3_update_interrupt_line_state(s, lidx); +} + +static bool vmxnet3_verify_driver_magic(target_phys_addr_t dshmem) +{ + return (VMXNET3_REV1_MAGIC == VMXNET3_READ_DRV_SHARED32(dshmem, magic)); +} + +#define _GET_BYTE(x, byte_num) (((x) >> (byte_num)*8) & 0xFF) +#define _MAKE_BYTE(byte_num, val) (((uint32_t)((val) & 0xFF)) << (byte_num)*8) + +static void vmxnet3_set_variable_mac(VMXNET3_State *s, uint32_t h, uint32_t l) +{ + s->conf.macaddr.a[0] = _GET_BYTE(l, 0); + s->conf.macaddr.a[1] = _GET_BYTE(l, 1); + s->conf.macaddr.a[2] = _GET_BYTE(l, 2); + s->conf.macaddr.a[3] = _GET_BYTE(l, 3); + s->conf.macaddr.a[4] = _GET_BYTE(h, 0); + s->conf.macaddr.a[5] = _GET_BYTE(h, 1); + + DCFPRINTF("Variable MAC: " MAC_FMT, MAC_ARG(s->conf.macaddr.a)); + + qemu_format_nic_info_str(&s->nic->nc, s->conf.macaddr.a); +} + +static uint64_t vmxnet3_get_mac_low(MACAddr *addr) +{ + return _MAKE_BYTE(0, addr->a[0]) | + _MAKE_BYTE(1, addr->a[1]) | + _MAKE_BYTE(2, addr->a[2]) | + _MAKE_BYTE(3, addr->a[3]); +} + +static uint64_t vmxnet3_get_mac_high(MACAddr *addr) +{ + return _MAKE_BYTE(0, addr->a[4]) | + _MAKE_BYTE(1, addr->a[5]); +} + +static void +vmxnet3_inc_tx_consumption_counter(VMXNET3_State *s, int qidx) +{ + vmxnet3_ring_inc(&s->txq_descr[qidx].tx_ring); +} + +static inline void +vmxnet3_inc_rx_consumption_counter(VMXNET3_State *s, int qidx, int ridx) +{ + vmxnet3_ring_inc(&s->rxq_descr[qidx].rx_ring[ridx]); +} + +static inline void +vmxnet3_inc_tx_completion_counter(VMXNET3_State *s, int qidx) +{ + vmxnet3_ring_inc(&s->txq_descr[qidx].comp_ring); +} + +static void +vmxnet3_inc_rx_completion_counter(VMXNET3_State *s, int qidx) +{ + vmxnet3_ring_inc(&s->rxq_descr[qidx].comp_ring); +} + +static void +vmxnet3_dec_rx_completion_counter(VMXNET3_State *s, int qidx) +{ + vmxnet3_ring_dec(&s->rxq_descr[qidx].comp_ring); +} + +static inline void vmxnet3_flush_shmem_changes(void) +{ + /* Flush shared memory changes */ + /* Needed before transferring comntrol to guest */ + smp_wmb(); +} + +static void vmxnet3_complete_packet(VMXNET3_State *s, int qidx, uint32 tx_ridx) +{ + struct Vmxnet3_TxCompDesc txcq_descr; + + vmxnet3_ring_dump(DRIPRINTF, "TXC", qidx, &s->txq_descr[qidx].comp_ring); + + txcq_descr.txdIdx = tx_ridx; + txcq_descr.gen = vmxnet3_ring_curr_gen(&s->txq_descr[qidx].comp_ring); + + vmxnet3_ring_write_curr_cell(&s->txq_descr[qidx].comp_ring, &txcq_descr); + vmxnet3_inc_tx_completion_counter(s, qidx); + + vmxnet3_flush_shmem_changes(); + vmxnet3_trigger_interrupt(s, s->txq_descr[qidx].intr_idx); +} + +static bool +vmxnet3_setup_tx_offloads(Vmxnet3_TxPkt *pkt) +{ + Vmxnet3_TxPktMdata *mdata = vmxnet3_txpkt_get_mdata(pkt); + struct virtio_net_hdr *vhdr = vmxnet3_txpkt_get_vhdr(pkt); + bool res = true; + + vhdr->hdr_len = mdata->hdr_length; + + switch (mdata->offload_mode) { + case VMXNET3_OM_NONE: + vhdr->flags = 0; + vhdr->gso_type = VIRTIO_NET_HDR_GSO_NONE; + vhdr->gso_size = 0; + break; + + case VMXNET3_OM_CSUM: { + DPKPRINTF("L4 CSO requested data_offset: %d, csoff: %d", + mdata->hdr_length, mdata->cso_or_gso_size); + vhdr->flags = VIRTIO_NET_HDR_F_NEEDS_CSUM; + vhdr->gso_type = VIRTIO_NET_HDR_GSO_NONE; + vhdr->gso_size = 0; + vhdr->csum_start = mdata->hdr_length; + vhdr->csum_offset = mdata->cso_or_gso_size - mdata->hdr_length; + } + break; + + case VMXNET3_OM_TSO: { + uint8_t *l2hdr = vmxnet3_txpkt_get_l2hdr(pkt); + size_t l2hdr_len = vmxnet3_txpkt_get_l2hdr_len(pkt); + uint8_t *l3hdr = vmxnet3_txpkt_get_l3hdr(pkt); + size_t l3hdr_len = vmxnet3_txpkt_get_l3hdr_len(pkt); + uint16_t l3_proto = eth_get_l3_proto(l2hdr, l2hdr_len); + size_t payload_len = vmxnet3_txpkt_get_payload_len(pkt); + + vhdr->gso_type = eth_get_gso_type(l3_proto, l3hdr); + vhdr->gso_size = mdata->cso_or_gso_size; + res = eth_setup_tx_offloads(l3hdr, l3hdr_len, l2hdr_len, + payload_len, vhdr, + vmxnet3_txpkt_get_more_frags(pkt), + vmxnet3_txpkt_get_frag_off(pkt)); + DPKPRINTF("GSO offload type %d requested.", vhdr->gso_type); + } + break; + + default: + assert(false); + } + + return res; +} + +static size_t +vmxnet3_txpkt_adopt_headers(Vmxnet3_TxPkt *pkt, + size_t data_len, + target_phys_addr_t data_pa, + bool needs_tso) +{ + /* Copy L2 header */ + uint8_t *l2hdr = vmxnet3_txpkt_get_l2hdr(pkt); + uint8_t *l3hdr = vmxnet3_txpkt_get_l3hdr(pkt); + size_t l2hdr_len = 0; + size_t l3hdr_len = 0; + + assert(data_len >= ETH_MAX_L2_HDR_LEN); + cpu_physical_memory_read(data_pa, l2hdr, ETH_MAX_L2_HDR_LEN); + l2hdr_len = eth_get_l2_hdr_length(l2hdr); + vmxnet3_txpkt_set_l2hdr_len(pkt, l2hdr_len); + + /* If packet requires offload - copy L3 header */ + if (needs_tso) { + switch (eth_get_l3_proto(l2hdr, l2hdr_len)) { + case ETH_P_IP: { + target_phys_addr_t ip_opt_pa; + struct ip_header *iphdr = (struct ip_header *) l3hdr; + assert(data_len >= l2hdr_len + sizeof(struct ip_header)); + cpu_physical_memory_read(data_pa + l2hdr_len, + l3hdr, sizeof(struct ip_header)); + l3hdr_len = IP_HDR_GET_LEN(l3hdr); + ip_opt_pa = data_pa + l2hdr_len + sizeof(struct ip_header); + cpu_physical_memory_read(ip_opt_pa, + l3hdr + sizeof(struct ip_header), + l3hdr_len - sizeof(struct ip_header)); + pkt->max_payload_len = + IP_FRAG_ALIGN_SIZE(VMXNET3_MAX_IP_PLOAD_LEN); + pkt->fragmentation.orig_more_frags = + FLAG_IS_SET(be16_to_cpu(iphdr->ip_off), IP_MF); + } + break; + + case ETH_P_IPV6: { + target_phys_addr_t l3hdr_pa = data_pa + l2hdr_len; + l3hdr_len = sizeof(struct ip6_header); + assert(data_len >= l2hdr_len + l3hdr_len); + cpu_physical_memory_read(l3hdr_pa, l3hdr, l3hdr_len); + pkt->max_payload_len = 0; + } + break; + + default: { + l3hdr_len = 0; + pkt->max_payload_len = 0; + } + break; + } + } + + vmxnet3_txpkt_set_l3hdr_len(pkt, l3hdr_len); + + /* Return amount of data adopted */ + return l2hdr_len + l3hdr_len; +} + +static void +vmxnet3_tx_retrieve_metadata(Vmxnet3_TxPkt *pkt, + const struct Vmxnet3_TxDesc *txd) +{ + struct eth_header *ehdr = + (struct eth_header *) vmxnet3_txpkt_get_l2hdr(pkt); + + Vmxnet3_TxPktMdata *mdata = vmxnet3_txpkt_get_mdata(pkt); + mdata->offload_mode = txd->om; + mdata->cso_or_gso_size = txd->msscof; + mdata->hdr_length = txd->hlen; + mdata->packet_type = get_eth_packet_type(ehdr); + + if (txd->ti) { + eth_setup_vlan_headers(ehdr, txd->tci); + } +} + +static size_t +vmxnet3_txpkt_adopt_data_fragment(Vmxnet3_TxPkt *pkt, + target_phys_addr_t data_pa, + size_t data_len, + uint32_t fragment_num) +{ + struct iovec *v = vmxnet3_txpkt_get_payload_frag(pkt, fragment_num); + size_t bytes_to_adopt = data_len; + + if ((0 != pkt->max_payload_len) && + (pkt->payload_len + data_len > pkt->max_payload_len)) { + bytes_to_adopt = pkt->max_payload_len - pkt->payload_len; + } + + v->iov_base = (void *) (uint64_t) data_pa; + v->iov_len = bytes_to_adopt; + pkt->payload_len += bytes_to_adopt; + return bytes_to_adopt; +} + +typedef enum { + VMXNET3_SUCCEEDED = 0xBEEFBEEF, + VMXNET3_OUT_OF_BUF, + VMXNET3_PKT_ERROR +} Vmxnet3_PktStatus; + +static void +vmxnet3_on_tx_done_update_stats(VMXNET3_State *s, + Vmxnet3_TxPkt *pkt, + int qidx, + Vmxnet3_PktStatus status) +{ + Vmxnet3_TxPktMdata *mdata = vmxnet3_txpkt_get_mdata(&s->curr_txpkt); + size_t tot_len = vmxnet3_txpkt_get_total_len(pkt); + struct UPT1_TxStats *stats = &s->txq_descr[qidx].txq_stats; + + switch (status) { + case VMXNET3_SUCCEEDED: { + switch (mdata->packet_type) { + case VMXNET3_PKT_BCAST: + stats->bcastPktsTxOK++; + stats->bcastBytesTxOK += tot_len; + break; + case VMXNET3_PKT_MCAST: + stats->mcastPktsTxOK++; + stats->mcastBytesTxOK += tot_len; + break; + case VMXNET3_PKT_UCAST: + stats->ucastPktsTxOK++; + stats->ucastBytesTxOK += tot_len; + break; + default: + assert(false); + } + + if (VMXNET3_OM_TSO == mdata->offload_mode) { + /* According to VMWARE headers this statistic is a number */ + /* of packets after segmentation but since we don't have */ + /* this information in QEMU model, the best we can do is to */ + /* provide number of non-segmented packets */ + stats->TSOPktsTxOK++; + stats->TSOBytesTxOK += tot_len; + } + } + break; + + case VMXNET3_PKT_ERROR: { + stats->pktsTxDiscard++; + } + break; + + case VMXNET3_OUT_OF_BUF: { + stats->pktsTxError++; + } + break; + + default: + assert(false); + } +} + +static void +vmxnet3_on_rx_done_update_stats(VMXNET3_State *s, + Vmxnet3_RxPkt *pkt, + int qidx, + Vmxnet3_PktStatus status) +{ + Vmxnet3_RxPktMdata *mdata = vmxnet3_rxpkt_get_mdata(pkt); + struct UPT1_RxStats *stats = &s->rxq_descr[qidx].rxq_stats; + + switch (status) { + case VMXNET3_OUT_OF_BUF: + stats->pktsRxOutOfBuf++; + break; + + case VMXNET3_PKT_ERROR: + stats->pktsRxError++; + break; + + case VMXNET3_SUCCEEDED: + switch (mdata->packet_type) { + case VMXNET3_PKT_BCAST: + stats->bcastPktsRxOK++; + stats->bcastBytesRxOK += mdata->tot_len; + break; + case VMXNET3_PKT_MCAST: + stats->mcastPktsRxOK++; + stats->mcastBytesRxOK += mdata->tot_len; + break; + case VMXNET3_PKT_UCAST: + stats->ucastPktsRxOK++; + stats->ucastBytesRxOK += mdata->tot_len; + break; + default: + assert(false); + } + + if (mdata->tot_len > s->mtu) { + stats->LROPktsRxOK++; + stats->LROBytesRxOK += mdata->tot_len; + } + break; + + default: + assert(false); + } +} + +static inline bool +vmxnet3_pop_next_tx_descr(VMXNET3_State *s, + int qidx, + struct Vmxnet3_TxDesc *txd, + uint32_t *descr_idx) +{ + Vmxnet3_Ring *ring = &s->txq_descr[qidx].tx_ring; + + vmxnet3_ring_read_curr_cell(ring, txd); + if (txd->gen == vmxnet3_ring_curr_gen(ring)) { + vmxnet3_ring_dump(DRIPRINTF, "TX", qidx, ring); + *descr_idx = vmxnet3_ring_curr_cell_idx(ring); + vmxnet3_inc_tx_consumption_counter(s, qidx); + return true; + } + + return false; +} + +static bool +vmxnet3_send_packet(VMXNET3_State *s, Vmxnet3_TxPkt *pkt, uint32_t qidx) +{ + uint32_t mapped_fragments; + void *mapped = vmxnet3_txpkt_map(&s->curr_txpkt, &mapped_fragments, false); + if (NULL != mapped) { + bool res; + if (vmxnet3_setup_tx_offloads(pkt)) { + vmxnet3_dump_virt_hdr(vmxnet3_txpkt_get_vhdr(pkt)); + vmxnet3_txpkt_dump(pkt); + qemu_sendv_packet(&s->nic->nc, mapped, mapped_fragments); + vmxnet3_on_tx_done_update_stats(s, pkt, qidx, VMXNET3_SUCCEEDED); + res = true; + } else { + vmxnet3_on_tx_done_update_stats(s, pkt, qidx, VMXNET3_PKT_ERROR); + res = false; + } + vmxnet3_txpkt_unmap(pkt, false); + return res; + } else { + vmxnet3_on_tx_done_update_stats(s, &s->curr_txpkt, + qidx, VMXNET3_OUT_OF_BUF); + return false; + } +} + +static void vmxnet3_process_tx_queue(VMXNET3_State *s, int qidx) +{ + struct Vmxnet3_TxDesc txd; + uint32_t txd_idx; + uint32_t data_len; + target_phys_addr_t data_pa; + size_t bytes_adopted; + + for (;;) { + if (!vmxnet3_pop_next_tx_descr(s, qidx, &txd, &txd_idx)) { + break; + } + + vmxnet3_dump_tx_descr(&txd); + + if (!s->curr_txpkt_skip) { + data_len = (txd.len > 0) ? txd.len : VMXNET3_MAX_TX_BUF_SIZE; + data_pa = le64_to_cpu(txd.addr); + + if (!s->curr_txpkt_header_processed) { + bool needs_tso = (VMXNET3_OM_TSO == txd.om); + bytes_adopted = vmxnet3_txpkt_adopt_headers(&s->curr_txpkt, + data_len, + data_pa, + needs_tso); + vmxnet3_tx_retrieve_metadata(&s->curr_txpkt, &txd); + data_pa += bytes_adopted; + data_len -= bytes_adopted; + s->curr_txpkt_header_processed = true; + } + + do { + if (0 != data_len) { + int frag_num = s->curr_txpkt_pl_frags++; + bytes_adopted = + vmxnet3_txpkt_adopt_data_fragment(&s->curr_txpkt, + data_pa, + data_len, + frag_num); + data_pa += bytes_adopted; + data_len -= bytes_adopted; + } + + if ((0 != data_len) || txd.eop) { + size_t frag_off; + + vmxnet3_txpkt_set_num_pl_frags(&s->curr_txpkt, + s->curr_txpkt_pl_frags); + + vmxnet3_txpkt_set_more_frags(&s->curr_txpkt, + (0 != data_len)); + + s->curr_txpkt_skip = + !vmxnet3_send_packet(s, &s->curr_txpkt, qidx); + + frag_off = vmxnet3_txpkt_get_payload_len(&s->curr_txpkt) / + IP_FRAG_UNIT_SIZE; + + vmxnet3_txpkt_advance_frag_off(&s->curr_txpkt, frag_off); + + vmxnet3_txpkt_reset_payload(&s->curr_txpkt); + s->curr_txpkt_pl_frags = 0; + } + } while (0 != data_len); + } + + if (txd.eop) { + vmxnet3_complete_packet(s, qidx, txd_idx); + vmxnet3_txpkt_reset(&s->curr_txpkt); + s->curr_txpkt_skip = false; + s->curr_txpkt_header_processed = false; + } + } +} + +static inline void +vmxnet3_read_next_rx_descr(VMXNET3_State *s, int qidx, int ridx, + struct Vmxnet3_RxDesc *dbuf, uint32_t *didx) +{ + Vmxnet3_Ring *ring = &s->rxq_descr[qidx].rx_ring[ridx]; + *didx = vmxnet3_ring_curr_cell_idx(ring); + vmxnet3_ring_read_curr_cell(ring, dbuf); +} + +static inline uint8_t +vmxnet3_get_rx_ring_gen(VMXNET3_State *s, int qidx, int ridx) +{ + return s->rxq_descr[qidx].rx_ring[ridx].gen; +} + +static inline target_phys_addr_t +vmxnet3_pop_rxc_descr(VMXNET3_State *s, int qidx, uint32_t *descr_gen) +{ + uint8_t ring_gen; + struct Vmxnet3_RxCompDesc rxcd; + + target_phys_addr_t daddr = + vmxnet3_ring_curr_cell_pa(&s->rxq_descr[qidx].comp_ring); + + cpu_physical_memory_read(daddr, &rxcd, sizeof(struct Vmxnet3_RxCompDesc)); + ring_gen = vmxnet3_ring_curr_gen(&s->rxq_descr[qidx].comp_ring); + + if (rxcd.gen != ring_gen) { + *descr_gen = ring_gen; + vmxnet3_inc_rx_completion_counter(s, qidx); + return daddr; + } + + return 0; +} + +static inline void +vmxnet3_revert_rxc_descr(VMXNET3_State *s, int qidx) +{ + vmxnet3_dec_rx_completion_counter(s, qidx); +} + +#define RXQ_IDX (0) +#define RX_HEAD_BODY_RING (0) +#define RX_BODY_ONLY_RING (1) + +static bool +vmxnet3_get_next_head_rx_descr(VMXNET3_State *s, + struct Vmxnet3_RxDesc *descr_buf, + uint32_t *descr_idx, + uint32_t *ridx) +{ + for (;;) { + uint32_t ring_gen; + vmxnet3_read_next_rx_descr(s, RXQ_IDX, RX_HEAD_BODY_RING, + descr_buf, descr_idx); + + /* If no more free descriptors - return */ + ring_gen = vmxnet3_get_rx_ring_gen(s, RXQ_IDX, RX_HEAD_BODY_RING); + if (descr_buf->gen != ring_gen) { + return false; + } + + /* Mark current descriptor as used/skipped */ + vmxnet3_inc_rx_consumption_counter(s, RXQ_IDX, RX_HEAD_BODY_RING); + + /* If this is what we are looking for - return */ + if (VMXNET3_RXD_BTYPE_HEAD == descr_buf->btype) { + *ridx = RX_HEAD_BODY_RING; + return true; + } + } +} + +static bool +vmxnet3_get_next_body_rx_descr(VMXNET3_State *s, + struct Vmxnet3_RxDesc *dbuf, + uint32_t *didx, + uint32_t *ridx) +{ + vmxnet3_read_next_rx_descr(s, RXQ_IDX, RX_HEAD_BODY_RING, dbuf, didx); + + /* Try to find corresponding descriptor in head/body ring */ + if ((dbuf->gen == vmxnet3_get_rx_ring_gen(s, RXQ_IDX, RX_HEAD_BODY_RING)) && + (VMXNET3_RXD_BTYPE_BODY == dbuf->btype)) { + vmxnet3_inc_rx_consumption_counter(s, RXQ_IDX, RX_HEAD_BODY_RING); + *ridx = RX_HEAD_BODY_RING; + return true; + } + + /* If there is no free descriptors on head/body ring or next free */ + /* descriptor is a head descriptor switch to body only ring */ + vmxnet3_read_next_rx_descr(s, RXQ_IDX, RX_BODY_ONLY_RING, dbuf, didx); + + /* If no more free descriptors - return */ + if (dbuf->gen == vmxnet3_get_rx_ring_gen(s, RXQ_IDX, RX_BODY_ONLY_RING)) { + assert(VMXNET3_RXD_BTYPE_BODY == dbuf->btype); + *ridx = RX_BODY_ONLY_RING; + vmxnet3_inc_rx_consumption_counter(s, RXQ_IDX, RX_BODY_ONLY_RING); + return true; + } + + return false; +} + +static inline bool +vmxnet3_get_next_rx_descr(VMXNET3_State *s, bool is_head, + struct Vmxnet3_RxDesc *descr_buf, + uint32_t *descr_idx, + uint32_t *ridx) +{ + if (is_head || !s->rx_packets_compound) { + return vmxnet3_get_next_head_rx_descr(s, descr_buf, descr_idx, ridx); + } else { + return vmxnet3_get_next_body_rx_descr(s, descr_buf, descr_idx, ridx); + } +} + +static void +vmxnet3_rx_put_metadata_to_descr(Vmxnet3_RxPkt *pkt, + struct Vmxnet3_RxCompDesc *rxcd) +{ + int csum_correct, is_gso; + bool isip4, isip6, istcp, isudp; + uint8_t headers[ETH_MAX_L2_HDR_LEN + ETH_MAX_L3_HDR_LEN]; + Vmxnet3_RxPktMdata *mdata = vmxnet3_rxpkt_get_mdata(pkt); + struct virtio_net_hdr *vhdr; + uint8_t offload_type; + + if (mdata->vlan_stripped) { + rxcd->ts = 1; + rxcd->tci = mdata->vlan_tag; + } + + if (!mdata->vhdr_valid) { + goto nocsum; + } + + vhdr = vmxnet3_rxpkt_get_vhdr(pkt); + /* Checksum is valid when lower level tell so or when lower level */ + /* requires checksum offload telling that packet produced/bridged */ + /* locally and did travel over network after last checksum calculation */ + /* or production */ + csum_correct = FLAG_IS_SET(vhdr->flags, VIRTIO_NET_HDR_F_DATA_VALID) || + FLAG_IS_SET(vhdr->flags, VIRTIO_NET_HDR_F_NEEDS_CSUM); + + offload_type = vhdr->gso_type & ~VIRTIO_NET_HDR_GSO_ECN; + is_gso = (offload_type != VIRTIO_NET_HDR_GSO_NONE) ? 1 : 0; + + if (!csum_correct && !is_gso) { + goto nocsum; + } + + /* Make linear copy of L2-L3 headers and parse it */ + iov_to_buf(vmxnet3_rxpkt_get_frag(pkt, 0), + vmxnet3_rxpkt_get_num_frags(pkt), + headers, 0, MIN(sizeof(headers), mdata->tot_len)); + + eth_get_protocols(headers, MIN(sizeof(headers), mdata->tot_len), + &isip4, &isip6, &isudp, &istcp); + if ((!istcp && !isudp) || (!isip4 && !isip6)) { + goto nocsum; + } + + rxcd->cnc = 0; + rxcd->v4 = isip4 ? 1 : 0; + rxcd->v6 = isip6 ? 1 : 0; + rxcd->tcp = istcp ? 1 : 0; + rxcd->udp = isudp ? 1 : 0; + rxcd->fcs = rxcd->tuc = rxcd->ipc = 1; + return; + +nocsum: + rxcd->cnc = 1; + return; +} + +static void +vmxnet3_physical_memory_writev(const struct iovec *iov, + size_t start_iov_off, + target_phys_addr_t target_addr, + size_t bytes_to_copy) +{ + size_t curr_off = 0; + size_t copied = 0; + + while (bytes_to_copy) { + if (start_iov_off < (curr_off + iov->iov_len)) { + size_t chunk_len = + MIN((curr_off + iov->iov_len) - start_iov_off, bytes_to_copy); + + cpu_physical_memory_write(target_addr + copied, + iov->iov_base + start_iov_off - curr_off, + chunk_len); + + copied += chunk_len; + start_iov_off += chunk_len; + curr_off = start_iov_off; + bytes_to_copy -= chunk_len; + } else { + curr_off += iov->iov_len; + } + iov++; + } +} + +static bool +vmxnet3_indicate_packet(VMXNET3_State *s, Vmxnet3_RxPkt *pkt) +{ + struct Vmxnet3_RxDesc rxd; + bool is_head = true; + uint32_t rxd_idx; + uint32_t rx_ridx; + + struct Vmxnet3_RxCompDesc rxcd; + uint32_t new_rxcd_gen = VMXNET3_INIT_GEN; + target_phys_addr_t new_rxcd_pa = 0; + target_phys_addr_t ready_rxcd_pa = 0; + struct iovec *data = vmxnet3_rxpkt_get_frag(pkt, 0); + Vmxnet3_RxPktMdata *mdata = vmxnet3_rxpkt_get_mdata(pkt); + size_t bytes_copied = 0; + size_t bytes_left = mdata->tot_len; + uint16_t num_frags = 0; + + vmxnet3_rxpkt_dump(pkt); + + while ((num_frags < s->max_rx_frags) && + (bytes_left > 0) && + (new_rxcd_pa = vmxnet3_pop_rxc_descr(s, RXQ_IDX, &new_rxcd_gen)) && + vmxnet3_get_next_rx_descr(s, is_head, &rxd, &rxd_idx, &rx_ridx)) { + size_t chunk_size = MIN(bytes_left, rxd.len); + vmxnet3_physical_memory_writev(data, bytes_copied, + le64_to_cpu(rxd.addr), chunk_size); + bytes_copied += chunk_size; + bytes_left -= chunk_size; + + vmxnet3_dump_rx_descr(&rxd); + + if (0 != ready_rxcd_pa) { + cpu_physical_memory_write(ready_rxcd_pa, &rxcd, sizeof(rxcd)); + } + + memset(&rxcd, 0, sizeof(struct Vmxnet3_RxCompDesc)); + rxcd.rxdIdx = rxd_idx; + rxcd.len = chunk_size; + rxcd.sop = is_head; + rxcd.gen = new_rxcd_gen; + rxcd.rqID = RXQ_IDX + rx_ridx*s->rxq_num; + + if (0 == bytes_left) { + vmxnet3_rx_put_metadata_to_descr(pkt, &rxcd); + } + + DRIPRINTF("RX Completion descriptor: rxRing: %lu rxIdx %lu len %lu " + "sop %d csum_correct %lu", + (unsigned long) rx_ridx, + (unsigned long) rxcd.rxdIdx, + (unsigned long) rxcd.len, + (int) rxcd.sop, + (unsigned long) rxcd.tuc); + + is_head = false; + ready_rxcd_pa = new_rxcd_pa; + new_rxcd_pa = 0; + } + + if (0 != ready_rxcd_pa) { + rxcd.eop = 1; + rxcd.err = (0 != bytes_left); + cpu_physical_memory_write(ready_rxcd_pa, &rxcd, sizeof(rxcd)); + vmxnet3_flush_shmem_changes(); + } + + if (0 != new_rxcd_pa) { + vmxnet3_revert_rxc_descr(s, RXQ_IDX); + } + + vmxnet3_trigger_interrupt(s, s->rxq_descr[RXQ_IDX].intr_idx); + + if (bytes_left == 0) { + vmxnet3_on_rx_done_update_stats(s, pkt, RXQ_IDX, VMXNET3_SUCCEEDED); + return true; + } else if (num_frags == s->max_rx_frags) { + vmxnet3_on_rx_done_update_stats(s, pkt, RXQ_IDX, VMXNET3_PKT_ERROR); + return false; + } else { + vmxnet3_on_rx_done_update_stats(s, pkt, RXQ_IDX, VMXNET3_OUT_OF_BUF); + return false; + } +} + +static void +vmxnet3_io_bar0_write(void *opaque, target_phys_addr_t addr, + uint64_t val, unsigned size) +{ + VMXNET3_State *s = opaque; + + if (IS_MULTIREG_ADDR(addr, VMXNET3_REG_TXPROD, + VMXNET3_DEVICE_MAX_TX_QUEUES, VMXNET3_REG_ALIGN)) { + int tx_queue_idx = + MULTIREG_IDX_BY_ADDR(addr, VMXNET3_REG_TXPROD, VMXNET3_REG_ALIGN); + assert(tx_queue_idx <= s->txq_num); + vmxnet3_process_tx_queue(s, tx_queue_idx); + return; + } + +#pragma GCC diagnostic push +#pragma GCC diagnostic ignored "-Wtype-limits" + if (IS_MULTIREG_ADDR(addr, VMXNET3_REG_IMR, + VMXNET3_MAX_INTRS, VMXNET3_REG_ALIGN)) { + int l = MULTIREG_IDX_BY_ADDR(addr, VMXNET3_REG_IMR, VMXNET3_REG_ALIGN); + + DCBPRINTF("Interrupt mask for line %d written: 0x%" PRIx64, l, val); + + vmxnet3_on_interrupt_mask_changed(s, l, val); + return; + } +#pragma GCC diagnostic pop + + if (IS_MULTIREG_ADDR(addr, VMXNET3_REG_RXPROD, + VMXNET3_DEVICE_MAX_RX_QUEUES, VMXNET3_REG_ALIGN) || + IS_MULTIREG_ADDR(addr, VMXNET3_REG_RXPROD2, + VMXNET3_DEVICE_MAX_RX_QUEUES, VMXNET3_REG_ALIGN)) { + return; + } + + DWRPRINTF("BAR0 unknown write [%" PRIx64 "] = %" PRIx64 ", size %d", + (uint64_t) addr, val, size); +} + +static uint64_t +vmxnet3_io_bar0_read(void *opaque, target_phys_addr_t addr, unsigned size) +{ +#pragma GCC diagnostic push +#pragma GCC diagnostic ignored "-Wtype-limits" + if (IS_MULTIREG_ADDR(addr, VMXNET3_REG_IMR, + VMXNET3_MAX_INTRS, VMXNET3_REG_ALIGN)) { + assert(false); + } +#pragma GCC diagnostic pop + + DCBPRINTF("BAR0 unknown read [%" PRIx64 "], size %d", + (uint64_t) addr, size); + return 0; +} + +static void vmxnet3_reset(VMXNET3_State *s) +{ + DCBPRINTF("Resetting vmxnet3..."); + + vmxnet3_txpkt_reset(&s->curr_txpkt); + s->curr_txpkt_pl_frags = 0; + s->curr_txpkt_skip = false; + s->curr_txpkt_header_processed = false; +} + +static void vmxnet3_deactivate_device(VMXNET3_State *s) +{ + DCBPRINTF("Deactivating vmxnet3..."); + s->device_active = false; +} + +static void vmxnet3_update_rx_mode(VMXNET3_State *s) +{ + s->rx_mode = VMXNET3_READ_DRV_SHARED32(s->drv_shmem, + devRead.rxFilterConf.rxMode); + DCFPRINTF("RX mode: 0x%08X", s->rx_mode); +} + +static void vmxnet3_update_vlan_filters(VMXNET3_State *s) +{ + int i; + + /* Copy configuration from shared memory */ + VMXNET3_READ_DRV_SHARED(s->drv_shmem, + devRead.rxFilterConf.vfTable, + s->vlan_table, + sizeof(s->vlan_table)); + + /* Invert byte order when needed */ + for (i = 0; i < ARRAY_SIZE(s->vlan_table); i++) { + s->vlan_table[i] = le32_to_cpu(s->vlan_table[i]); + } + + /* Dump configuration for debugging purposes */ + DCFPRINTF("Configured VLANs:"); + for (i = 0; i < sizeof(s->vlan_table) * 8; i++) { + if (VMXNET3_VFTABLE_ENTRY_IS_SET(s->vlan_table, i)) { + DCFPRINTF("\tVLAN %d is present", i); + } + } +} + +static void vmxnet3_update_mcast_filters(VMXNET3_State *s) +{ + uint16_t list_bytes = + VMXNET3_READ_DRV_SHARED16(s->drv_shmem, + devRead.rxFilterConf.mfTableLen); + + s->mcast_list_len = list_bytes / sizeof(s->mcast_list[0]); + + s->mcast_list = g_realloc(s->mcast_list, list_bytes); + if (NULL == s->mcast_list) { + if (0 == s->mcast_list_len) { + DCFPRINTF("Current multicast list is empty"); + } else { + DERPRINTF("Failed to allocate multicast list of %d elements", + s->mcast_list_len); + } + s->mcast_list_len = 0; + } else { + int i; + target_phys_addr_t mcast_list_pa = + VMXNET3_READ_DRV_SHARED64(s->drv_shmem, + devRead.rxFilterConf.mfTablePA); + + cpu_physical_memory_read(mcast_list_pa, s->mcast_list, list_bytes); + DCFPRINTF("Current multicast list len is %d:", s->mcast_list_len); + for (i = 0; i < s->mcast_list_len; i++) { + DCFPRINTF("\t" MAC_FMT, MAC_ARG(s->mcast_list[i].a)); + } + } +} + +static void vmxnet3_setup_rx_filtering(VMXNET3_State *s) +{ + vmxnet3_update_rx_mode(s); + vmxnet3_update_vlan_filters(s); + vmxnet3_update_mcast_filters(s); +} + +static uint32_t vmxnet3_get_interrupt_config(VMXNET3_State *s) +{ + uint32_t interrupt_mode = VMXNET3_IT_AUTO | (VMXNET3_IMM_AUTO << 2); + DCFPRINTF("Interrupt config is 0x%X", interrupt_mode); + return interrupt_mode; +} + +static void vmxnet3_fill_stats(VMXNET3_State *s) +{ + int i; + for (i = 0; i < s->txq_num; i++) { + cpu_physical_memory_write(s->txq_descr[i].tx_stats_pa, + &s->txq_descr[i].txq_stats, + sizeof(s->txq_descr[i].txq_stats)); + } + + for (i = 0; i < s->rxq_num; i++) { + cpu_physical_memory_write(s->rxq_descr[i].rx_stats_pa, + &s->rxq_descr[i].rxq_stats, + sizeof(s->rxq_descr[i].rxq_stats)); + } +} + +static void vmxnet3_adjust_by_guest_type(VMXNET3_State *s) +{ + struct Vmxnet3_GOSInfo gos; + + VMXNET3_READ_DRV_SHARED(s->drv_shmem, devRead.misc.driverInfo.gos, + &gos, sizeof(gos)); + s->rx_packets_compound = + (VMXNET3_GOS_TYPE_WIN == gos.gosType) ? false : true; + + DCFPRINTF("Guest type specifics: RXCOMPOUND: %d", s->rx_packets_compound); +} + +static void +vmxnet3_dump_conf_descr(const char *name, + struct Vmxnet3_VariableLenConfDesc *pm_descr) +{ + DCFPRINTF("%s descriptor dump: Version %u, Length %u", + name, pm_descr->confVer, pm_descr->confLen); + +}; + +static void vmxnet3_update_pm_state(VMXNET3_State *s) +{ + struct Vmxnet3_VariableLenConfDesc pm_descr; + + pm_descr.confLen = + VMXNET3_READ_DRV_SHARED32(s->drv_shmem, devRead.pmConfDesc.confLen); + pm_descr.confVer = + VMXNET3_READ_DRV_SHARED32(s->drv_shmem, devRead.pmConfDesc.confVer); + pm_descr.confPA = + VMXNET3_READ_DRV_SHARED64(s->drv_shmem, devRead.pmConfDesc.confPA); + + vmxnet3_dump_conf_descr("PM State", &pm_descr); +} + +static void vmxnet3_update_features(VMXNET3_State *s) +{ + uint32_t guest_features; + int rxcsum_offload_supported; + + guest_features = VMXNET3_READ_DRV_SHARED32(s->drv_shmem, + devRead.misc.uptFeatures); + + rxcsum_offload_supported = FLAG_IS_SET(guest_features, UPT1_F_RXCSUM); + s->rx_vlan_stripping = FLAG_IS_SET(guest_features, UPT1_F_RXVLAN); + s->lro_supported = FLAG_IS_SET(guest_features, UPT1_F_LRO); + + DCFPRINTF("Features configuration: LRO: %d, RXCSUM: %d, VLANSTRIP: %d", + s->lro_supported, rxcsum_offload_supported, + s->rx_vlan_stripping); + + tap_set_offload(s->nic->nc.peer, + rxcsum_offload_supported, + s->lro_supported, + s->lro_supported, + 0, + 0); +} + +static void vmxnet3_activate_device(VMXNET3_State *s) +{ + int i; + static const uint32_t VMXNET3_DEF_TX_THRESHOLD = 1; + target_phys_addr_t qdescr_table_pa; + uint64_t pa; + uint32_t size; + + /* Verify configuration consistency */ + if (!vmxnet3_verify_driver_magic(s->drv_shmem)) { + DERPRINTF("Device configuration received from driver is invalid"); + return; + } + + vmxnet3_adjust_by_guest_type(s); + vmxnet3_update_features(s); + vmxnet3_update_pm_state(s); + vmxnet3_setup_rx_filtering(s); + /* Cache fields from shared memory */ + s->mtu = VMXNET3_READ_DRV_SHARED32(s->drv_shmem, devRead.misc.mtu); + DCFPRINTF("MTU is %u", s->mtu); + + s->max_rx_frags = + VMXNET3_READ_DRV_SHARED16(s->drv_shmem, devRead.misc.maxNumRxSG); + + DCFPRINTF("Max RX fragments is %u", s->max_rx_frags); + + s->event_int_idx = + VMXNET3_READ_DRV_SHARED8(s->drv_shmem, devRead.intrConf.eventIntrIdx); + DCFPRINTF("Events interrupt line is %u", s->event_int_idx); + + s->auto_int_masking = + VMXNET3_READ_DRV_SHARED8(s->drv_shmem, devRead.intrConf.autoMask); + DCFPRINTF("Automatic interrupt masking is %d", (int)s->auto_int_masking); + + s->txq_num = + VMXNET3_READ_DRV_SHARED8(s->drv_shmem, devRead.misc.numTxQueues); + s->rxq_num = + VMXNET3_READ_DRV_SHARED8(s->drv_shmem, devRead.misc.numRxQueues); + + DCFPRINTF("Number of TX/RX queues %u/%u", s->txq_num, s->rxq_num); + assert(s->txq_num <= VMXNET3_DEVICE_MAX_TX_QUEUES); + + qdescr_table_pa = + VMXNET3_READ_DRV_SHARED64(s->drv_shmem, devRead.misc.queueDescPA); + DCFPRINTF("TX queues descriptors table is at 0x%" PRIx64, + (uint64_t) qdescr_table_pa); + + /* Worst-case scenario is a packet that holds all TX rings space so */ + /* we calculate total size of all TX rings for max TX fragments number */ + s->max_tx_frags = 0; + + /* TX queues */ + for (i = 0; i < s->txq_num; i++) { + target_phys_addr_t qdescr_pa = + qdescr_table_pa + i*sizeof(struct Vmxnet3_TxQueueDesc); + + /* Read interrupt number for this TX queue */ + s->txq_descr[i].intr_idx = + VMXNET3_READ_TX_QUEUE_DESCR8(qdescr_pa, conf.intrIdx); + + DCFPRINTF("TX Queue %d interrupt: %d", i, s->txq_descr[i].intr_idx); + + /* Read rings memory locations for TX queues */ + pa = VMXNET3_READ_TX_QUEUE_DESCR64(qdescr_pa, conf.txRingBasePA); + size = VMXNET3_READ_TX_QUEUE_DESCR32(qdescr_pa, conf.txRingSize); + + vmxnet3_ring_init(&s->txq_descr[i].tx_ring, pa, size, + sizeof(struct Vmxnet3_TxDesc), false); + vmxnet3_ring_dump(DCFPRINTF, "TX", i, &s->txq_descr[i].tx_ring); + + s->max_tx_frags += size; + + /* TXC ring */ + pa = VMXNET3_READ_TX_QUEUE_DESCR64(qdescr_pa, conf.compRingBasePA); + size = VMXNET3_READ_TX_QUEUE_DESCR32(qdescr_pa, conf.compRingSize); + vmxnet3_ring_init(&s->txq_descr[i].comp_ring, pa, size, + sizeof(struct Vmxnet3_TxCompDesc), true); + vmxnet3_ring_dump(DCFPRINTF, "TXC", i, &s->txq_descr[i].comp_ring); + + s->txq_descr[i].tx_stats_pa = + qdescr_pa + offsetof(struct Vmxnet3_TxQueueDesc, stats); + + memset(&s->txq_descr[i].txq_stats, 0, + sizeof(s->txq_descr[i].txq_stats)); + + /* Fill device-managed parameters for queues */ + VMXNET3_WRITE_TX_QUEUE_DESCR32(qdescr_pa, + ctrl.txThreshold, + VMXNET3_DEF_TX_THRESHOLD); + } + + /* Preallocate TX packet wrapper */ + DCFPRINTF("Max TX fragments is %u", s->max_tx_frags); + if (!vmxnet3_txpkt_prealloc(&s->curr_txpkt, s->max_tx_frags, + s->peer_has_vhdr)) { + hw_error("TX rings configuration problem"); + } + + /* Read rings memory locations for RX queues */ + for (i = 0; i < s->rxq_num; i++) { + int j; + target_phys_addr_t qd_pa = + qdescr_table_pa + s->txq_num*sizeof(struct Vmxnet3_TxQueueDesc) + + i*sizeof(struct Vmxnet3_RxQueueDesc); + + /* Read interrupt number for this RX queue */ + s->rxq_descr[i].intr_idx = + VMXNET3_READ_TX_QUEUE_DESCR8(qd_pa, conf.intrIdx); + + DCFPRINTF("RX Queue %d interrupt: %d", i, s->rxq_descr[i].intr_idx); + + /* Read rings memory locations */ + for (j = 0; j < VMXNET3_RX_RINGS_PER_QUEUE; j++) { + /* RX rings */ + pa = VMXNET3_READ_RX_QUEUE_DESCR64(qd_pa, conf.rxRingBasePA[j]); + size = VMXNET3_READ_RX_QUEUE_DESCR32(qd_pa, conf.rxRingSize[j]); + vmxnet3_ring_init(&s->rxq_descr[i].rx_ring[j], pa, size, + sizeof(struct Vmxnet3_RxDesc), false); + DCFPRINTF("RX queue %d:%d: Base: %" PRIx64 ", Size: %d", + i, j, pa, size); + } + + /* RXC ring */ + pa = VMXNET3_READ_RX_QUEUE_DESCR64(qd_pa, conf.compRingBasePA); + size = VMXNET3_READ_RX_QUEUE_DESCR32(qd_pa, conf.compRingSize); + vmxnet3_ring_init(&s->rxq_descr[i].comp_ring, pa, size, + sizeof(struct Vmxnet3_RxCompDesc), true); + DCFPRINTF("RXC queue %d: Base: %" PRIx64 ", Size: %d", i, pa, size); + + s->rxq_descr[i].rx_stats_pa = + qd_pa + offsetof(struct Vmxnet3_RxQueueDesc, stats); + memset(&s->rxq_descr[i].rxq_stats, 0, + sizeof(s->rxq_descr[i].rxq_stats)); + } + + vmxnet3_flush_shmem_changes(); + s->device_active = true; +} + +static void vmxnet3_handle_command(VMXNET3_State *s, uint64_t cmd) +{ + s->last_command = cmd; + + switch (cmd) { + case VMXNET3_CMD_GET_PERM_MAC_HI: + DCBPRINTF("Set: Get upper part of permanent MAC"); + break; + + case VMXNET3_CMD_GET_PERM_MAC_LO: + DCBPRINTF("Set: Get lower part of permanent MAC"); + break; + + case VMXNET3_CMD_GET_STATS: + DCBPRINTF("Set: Get device statistics"); + vmxnet3_fill_stats(s); + break; + + case VMXNET3_CMD_ACTIVATE_DEV: + DCBPRINTF("Set: Activating vmxnet3 device"); + vmxnet3_activate_device(s); + break; + + case VMXNET3_CMD_UPDATE_RX_MODE: + DCBPRINTF("Set: Update rx mode"); + vmxnet3_update_rx_mode(s); + break; + + case VMXNET3_CMD_UPDATE_VLAN_FILTERS: + DCBPRINTF("Set: Update VLAN filters"); + vmxnet3_update_vlan_filters(s); + break; + + case VMXNET3_CMD_UPDATE_MAC_FILTERS: + DCBPRINTF("Set: Update MAC filters"); + vmxnet3_update_mcast_filters(s); + break; + + case VMXNET3_CMD_UPDATE_FEATURE: + DCBPRINTF("Set: Update features"); + vmxnet3_update_features(s); + break; + + case VMXNET3_CMD_UPDATE_PMCFG: + DCBPRINTF("Set: Update power management config"); + vmxnet3_update_pm_state(s); + break; + + case VMXNET3_CMD_GET_LINK: + DCBPRINTF("Set: Get link"); + break; + + case VMXNET3_CMD_RESET_DEV: + DCBPRINTF("Set: Reset device"); + vmxnet3_reset(s); + break; + + case VMXNET3_CMD_QUIESCE_DEV: + DCBPRINTF("Set: VMXNET3_CMD_QUIESCE_DEV - pause the device"); + vmxnet3_deactivate_device(s); + break; + + case VMXNET3_CMD_GET_CONF_INTR: + DCBPRINTF("Set: VMXNET3_CMD_GET_CONF_INTR - interrupt configuration"); + break; + + default: + DCBPRINTF("Received unknown command: %" PRIx64, cmd); + break; + } +} + +static uint64_t vmxnet3_get_command_status(VMXNET3_State *s) +{ + uint64_t ret; + + switch (s->last_command) { + case VMXNET3_CMD_ACTIVATE_DEV: + ret = (s->device_active) ? 0 : -1; + DCFPRINTF("Device active: %" PRIx64, ret); + break; + + case VMXNET3_CMD_GET_LINK: + ret = s->link_status_and_speed; + DCFPRINTF("Link and speed: %" PRIx64, ret); + break; + + case VMXNET3_CMD_GET_PERM_MAC_LO: + ret = vmxnet3_get_mac_low(&s->perm_mac); + break; + + case VMXNET3_CMD_GET_PERM_MAC_HI: + ret = vmxnet3_get_mac_high(&s->perm_mac); + break; + + case VMXNET3_CMD_GET_CONF_INTR: + ret = vmxnet3_get_interrupt_config(s); + break; + + default: + DWRPRINTF("Received request for unknown command: %x", s->last_command); + ret = -1; + break; + } + + return ret; +} + +static void vmxnet3_set_events(VMXNET3_State *s, uint32_t val) +{ + uint32_t events; + + DCBPRINTF("Setting events: 0x%x", val); + events = VMXNET3_READ_DRV_SHARED32(s->drv_shmem, ecr) | val; + VMXNET3_WRITE_DRV_SHARED32(s->drv_shmem, ecr, events); +} + +static void vmxnet3_ack_events(VMXNET3_State *s, uint32_t val) +{ + uint32_t events; + + DCBPRINTF("Clearing events: 0x%x", val); + events = VMXNET3_READ_DRV_SHARED32(s->drv_shmem, ecr) & ~val; + VMXNET3_WRITE_DRV_SHARED32(s->drv_shmem, ecr, events); +} + +static void +vmxnet3_io_bar1_write(void *opaque, + target_phys_addr_t addr, + uint64_t val, + unsigned size) +{ + VMXNET3_State *s = opaque; + + switch (addr) { + /* Vmxnet3 Revision Report Selection */ + case VMXNET3_REG_VRRS: + DCBPRINTF("Write BAR1 [VMXNET3_REG_VRRS] = %" PRIx64 ", size %d", + val, size); + break; + + /* UPT Version Report Selection */ + case VMXNET3_REG_UVRS: + DCBPRINTF("Write BAR1 [VMXNET3_REG_UVRS] = %" PRIx64 ", size %d", + val, size); + break; + + /* Driver Shared Address Low */ + case VMXNET3_REG_DSAL: + DCBPRINTF("Write BAR1 [VMXNET3_REG_DSAL] = %" PRIx64 ", size %d", + val, size); + /* Guest driver will first write the low part of the shared */ + /* memory address. We save it to temp variable and set the */ + /* shared address only after we get the high part */ + if (0 == val) { + s->device_active = false; + } + s->temp_shared_guest_driver_memory = val; + s->drv_shmem = 0; + break; + + /* Driver Shared Address High */ + case VMXNET3_REG_DSAH: + DCBPRINTF("Write BAR1 [VMXNET3_REG_DSAH] = %" PRIx64 ", size %d", + val, size); + /* Set the shared memory between guest driver and device. */ + /* We already should have low address part. */ + s->drv_shmem = s->temp_shared_guest_driver_memory | (val << 32); + break; + + /* Command */ + case VMXNET3_REG_CMD: + DCBPRINTF("Write BAR1 [VMXNET3_REG_CMD] = %" PRIx64 ", size %d", + val, size); + vmxnet3_handle_command(s, val); + break; + + /* MAC Address Low */ + case VMXNET3_REG_MACL: + DCBPRINTF("Write BAR1 [VMXNET3_REG_MACL] = %" PRIx64 ", size %d", + val, size); + s->temp_mac = val; + break; + + /* MAC Address High */ + case VMXNET3_REG_MACH: + DCBPRINTF("Write BAR1 [VMXNET3_REG_MACH] = %" PRIx64 ", size %d", + val, size); + vmxnet3_set_variable_mac(s, val, s->temp_mac); + break; + + /* Interrupt Cause Register */ + case VMXNET3_REG_ICR: + DCBPRINTF("Write BAR1 [VMXNET3_REG_ICR] = %" PRIx64 ", size %d", + val, size); + assert(false); + break; + + /* Event Cause Register */ + case VMXNET3_REG_ECR: + DCBPRINTF("Write BAR1 [VMXNET3_REG_ECR] = %" PRIx64 ", size %d", + val, size); + vmxnet3_ack_events(s, val); + break; + + default: + DCBPRINTF("Unknown Write to BAR1 [%" PRIx64 "] = %" PRIx64 ", size %d", + (uint64_t) addr, val, size); + break; + } +} + +static uint64_t +vmxnet3_io_bar1_read(void *opaque, target_phys_addr_t addr, unsigned size) +{ + VMXNET3_State *s = opaque; + uint64_t ret = 0; + + switch (addr) { + /* Vmxnet3 Revision Report Selection */ + case VMXNET3_REG_VRRS: + DCBPRINTF("Read BAR1 [VMXNET3_REG_VRRS], size %d", size); + ret = VMXNET3_DEVICE_REVISION; + break; + + /* UPT Version Report Selection */ + case VMXNET3_REG_UVRS: + DCBPRINTF("Read BAR1 [VMXNET3_REG_UVRS], size %d", size); + ret = VMXNET3_DEVICE_VERSION; + break; + + /* Command */ + case VMXNET3_REG_CMD: + DCBPRINTF("Read BAR1 [VMXNET3_REG_CMD], size %d", size); + ret = vmxnet3_get_command_status(s); + break; + + /* MAC Address Low */ + case VMXNET3_REG_MACL: + DCBPRINTF("Read BAR1 [VMXNET3_REG_MACL], size %d", size); + ret = vmxnet3_get_mac_low(&s->conf.macaddr); + break; + + /* MAC Address High */ + case VMXNET3_REG_MACH: + DCBPRINTF("Read BAR1 [VMXNET3_REG_MACH], size %d", size); + ret = vmxnet3_get_mac_high(&s->conf.macaddr); + break; + + /* Interrupt Cause Register */ + /* Used for legacy interrupts only so interrupt index always 0 */ + case VMXNET3_REG_ICR: + DCBPRINTF("Read BAR1 [VMXNET3_REG_ICR], size %d", size); + if (vmxnet3_interrupt_asserted(s, 0)) { + vmxnet3_clear_interrupt(s, 0); + ret = true; + } else { + ret = false; + } + break; + + default: + DCBPRINTF("Unknow read BAR1[%" PRIx64 "], %d bytes", + (uint64_t) addr, size); + break; + } + + return ret; +} + +static int +vmxnet3_can_receive(VLANClientState *nc) +{ + VMXNET3_State *s = DO_UPCAST(NICState, nc, nc)->opaque; + return s->device_active && + FLAG_IS_SET(s->link_status_and_speed, VMXNET3_LINK_STATUS_UP); +} + +static inline bool +vmxnet3_is_registered_vlan(VMXNET3_State *s, const void *data) +{ + uint16_t vlan_tag = eth_get_pkt_vlan_tag(data) & VLAN_VID_MASK; + + if (IS_SPECIAL_VLAN_ID(vlan_tag)) { + return true; + } + + return VMXNET3_VFTABLE_ENTRY_IS_SET(s->vlan_table, vlan_tag); +} + +static bool +vmxnet3_is_allowed_mcast_group(VMXNET3_State *s, const uint8_t *group_mac) +{ + int i; + for (i = 0; i < s->mcast_list_len; i++) { + if (memcmp(group_mac, s->mcast_list[i].a, sizeof(s->mcast_list[i]))) { + return true; + } + } + return false; +} + +static bool +vmxnet3_rx_filter_may_indicate(VMXNET3_State *s, const void *data, + size_t size, eth_pkt_types_e packet_type) +{ + struct eth_header *ehdr = PKT_GET_ETH_HDR(data); + + if (FLAG_IS_SET(s->rx_mode, VMXNET3_RXM_PROMISC)) { + return true; + } + + switch (packet_type) { + case VMXNET3_PKT_UCAST: + if (!FLAG_IS_SET(s->rx_mode, VMXNET3_RXM_UCAST)) { + return false; + } + if (memcmp(s->conf.macaddr.a, ehdr->h_dest, ETH_ALEN)) { + return false; + } + break; + + case VMXNET3_PKT_BCAST: + if (!FLAG_IS_SET(s->rx_mode, VMXNET3_RXM_BCAST)) { + return false; + } + break; + + case VMXNET3_PKT_MCAST: + if (FLAG_IS_SET(s->rx_mode, VMXNET3_RXM_ALL_MULTI)) { + return true; + } + if (!FLAG_IS_SET(s->rx_mode, VMXNET3_RXM_MCAST)) { + return false; + } + if (!vmxnet3_is_allowed_mcast_group(s, ehdr->h_dest)) { + return false; + } + break; + + default: + assert(false); + } + + return vmxnet3_is_registered_vlan(s, data); +} + +static void +vmxnet3_rxpkt_attach_data(VMXNET3_State *s, Vmxnet3_RxPkt *pkt, + const void *data, size_t len) +{ + uint16_t vtag = 0; + uint16_t ploff; + Vmxnet3_RxPktMdata *mdata = vmxnet3_rxpkt_get_mdata(pkt); + + if (s->rx_vlan_stripping) { + mdata->vlan_stripped = + eth_strip_vlan(&data, vmxnet3_rxpkt_get_ehdr(pkt), &ploff, &vtag); + } else { + mdata->vlan_stripped = false; + } + + if (mdata->vlan_stripped) { + vmxnet3_rxpkt_attach_ehdr(pkt); + vmxnet3_rxpkt_get_frag(pkt, 1)->iov_base = (uint8_t *) data + ploff; + vmxnet3_rxpkt_get_frag(pkt, 1)->iov_len = len - ploff; + vmxnet3_rxpkt_set_num_frags(pkt, 2); + mdata->tot_len = len - ploff + sizeof(struct eth_header); + } else { + vmxnet3_rxpkt_get_frag(pkt, 0)->iov_base = (void *) data; + vmxnet3_rxpkt_get_frag(pkt, 0)->iov_len = len; + vmxnet3_rxpkt_set_num_frags(pkt, 1); + mdata->tot_len = len; + } + + mdata->vlan_tag = vtag; +} + +static ssize_t +vmxnet3_receive(VLANClientState *nc, const uint8_t *buf, size_t size) +{ + VMXNET3_State *s = DO_UPCAST(NICState, nc, nc)->opaque; + Vmxnet3_RxPkt pkt; + size_t bytes_indicated; + Vmxnet3_RxPktMdata *mdata; + + if (!vmxnet3_can_receive(&s->nic->nc)) { + DPKPRINTF("Cannot receive now"); + return -1; + } + + vmxnet3_rxpkt_init(&pkt); + mdata = vmxnet3_rxpkt_get_mdata(&pkt); + + if (s->peer_has_vhdr) { + buf += sizeof(struct virtio_net_hdr); + size -= sizeof(struct virtio_net_hdr); + *vmxnet3_rxpkt_get_vhdr(&pkt) = *(const struct virtio_net_hdr *) buf; + mdata->vhdr_valid = true; + } else { + mdata->vhdr_valid = false; + } + + mdata->packet_type = get_eth_packet_type(PKT_GET_ETH_HDR(buf)); + + if (vmxnet3_rx_filter_may_indicate(s, buf, size, mdata->packet_type)) { + vmxnet3_rxpkt_attach_data(s, &pkt, buf, size); + bytes_indicated = vmxnet3_indicate_packet(s, &pkt) ? size : -1; + if (bytes_indicated < size) { + DPKPRINTF("RX: %lu of %lu bytes indicated", bytes_indicated, size); + } + } else { + DPKPRINTF("Packet dropped by RX filter"); + bytes_indicated = size; + } + + assert(size > 0); + assert(bytes_indicated != 0); + return bytes_indicated; +} + +static void vmxnet3_cleanup(VLANClientState *nc) +{ + VMXNET3_State *s = DO_UPCAST(NICState, nc, nc)->opaque; + s->nic = NULL; +} + +static void vmxnet3_set_link_status(VLANClientState *nc) +{ + VMXNET3_State *s = DO_UPCAST(NICState, nc, nc)->opaque; + + if (nc->link_down) { + s->link_status_and_speed &= ~VMXNET3_LINK_STATUS_UP; + } else { + s->link_status_and_speed |= VMXNET3_LINK_STATUS_UP; + } + + vmxnet3_set_events(s, VMXNET3_ECR_LINK); + vmxnet3_trigger_interrupt(s, s->event_int_idx); +} + +static NetClientInfo net_vmxnet3_info = { + .type = NET_CLIENT_TYPE_NIC, + .size = sizeof(NICState), + .can_receive = vmxnet3_can_receive, + .receive = vmxnet3_receive, + .cleanup = vmxnet3_cleanup, + .link_status_changed = vmxnet3_set_link_status, +}; + +static bool vmxnet3_peer_has_vnet_hdr(VMXNET3_State *s) +{ + VLANClientState *peer = s->nic->nc.peer; + + if ((NULL != peer) && + (NET_CLIENT_TYPE_TAP == peer->info->type) && + tap_has_vnet_hdr(peer)) { + return true; + } + + DWRPRINTF("Peer has no virtio extension. Task offloads will not work."); + return false; +} + +static void vmxnet3_net_uninit(VMXNET3_State *s) +{ + if (NULL != s->mcast_list) { + g_free(s->mcast_list); + } + + vmxnet3_txpkt_cleanup(&s->curr_txpkt); +} + +static void vmxnet3_net_init(VMXNET3_State *s) +{ + DCBPRINTF("vmxnet3_net_init called..."); + + vmxnet3_txpkt_init(&s->curr_txpkt); + + qemu_macaddr_default_if_unset(&s->conf.macaddr); + + /* Windows guest will query the address that was set on init */ + memcpy(&s->perm_mac.a, &s->conf.macaddr.a, sizeof(s->perm_mac.a)); + + s->mcast_list = NULL; + s->mcast_list_len = 0; + + s->link_status_and_speed = VMXNET3_LINK_SPEED | VMXNET3_LINK_STATUS_UP; + + DCFPRINTF("Permanent MAC: " MAC_FMT, MAC_ARG(s->perm_mac.a)); + + s->nic = qemu_new_nic(&net_vmxnet3_info, &s->conf, + object_get_typename(OBJECT(s)), + s->dev.qdev.id, s); + + s->peer_has_vhdr = vmxnet3_peer_has_vnet_hdr(s); + s->curr_txpkt_pl_frags = 0; + s->curr_txpkt_skip = false; + s->curr_txpkt_header_processed = false; + + if (s->peer_has_vhdr) { + tap_set_vnet_hdr_len(s->nic->nc.peer, sizeof(struct virtio_net_hdr)); + tap_using_vnet_hdr(s->nic->nc.peer, 1); + } + + qemu_format_nic_info_str(&s->nic->nc, s->conf.macaddr.a); +} + +#ifdef VMXNET3_ENABLE_MSIX + +static void +vmxnet3_unuse_msix_vectors(VMXNET3_State *s, int num_vectors) +{ + int i; + for (i = 0; i < num_vectors; i++) { + msix_vector_unuse(&s->dev, i); + } +} + +static bool +vmxnet3_use_msix_vectors(VMXNET3_State *s, int num_vectors) +{ + int i; + for (i = 0; i < num_vectors; i++) { + int res = msix_vector_use(&s->dev, i); + if (0 > res) { + DWRPRINTF("Failed to use MSI-X vector %d, error %d", i, res); + vmxnet3_unuse_msix_vectors(s, i); + return false; + } + } + return true; +} + +static bool +vmxnet3_init_msix(VMXNET3_State *s) +{ + int res = msix_init(&s->dev, VMXNET3_MAX_INTRS, + &s->msix_bar, VMXNET3_MSIX_BAR_IDX, 0); + if (0 > res) { + DWRPRINTF("Failed to initialize MSI-X, error %d", res); + s->msix_used = false; + } else { + if (!vmxnet3_use_msix_vectors(s, VMXNET3_MAX_INTRS)) { + DWRPRINTF("Failed to use MSI-X vectors, error %d", res); + msix_uninit(&s->dev, &s->msix_bar); + s->msix_used = false; + } else { + s->msix_used = true; + } + } + return s->msix_used; +} + +static void +vmxnet3_cleanup_msix(VMXNET3_State *s) +{ + if (s->msix_used) { + msix_vector_unuse(&s->dev, VMXNET3_MAX_INTRS); + msix_uninit(&s->dev, &s->msix_bar); + } +} +#endif + +#ifdef VMXNET3_ENABLE_MSI + +static bool +vmxnet3_init_msi(VMXNET3_State *s) +{ +#define VMXNET3_MSI_NUM_VECTORS (1) +#define VMXNET3_MSI_OFFSET (0x50) +#define VMXNET3_USE_64BIT (true) +#define VMXNET3_PER_VECTOR_MASK (false) + + int res; + res = msi_init(&s->dev, VMXNET3_MSI_OFFSET, VMXNET3_MSI_NUM_VECTORS, + VMXNET3_USE_64BIT, VMXNET3_PER_VECTOR_MASK); + if (0 > res) { + DWRPRINTF("Failed to initialize MSI, error %d", res); + s->msi_used = false; + } else { + s->msi_used = true; + } + + return s->msi_used; +} + +static void +vmxnet3_cleanup_msi(VMXNET3_State *s) +{ + if (s->msi_used) { + msi_uninit(&s->dev); + } +} +#endif + +static int vmxnet3_pci_init(PCIDevice *dev) +{ + static const MemoryRegionOps b0_ops = { + .read = vmxnet3_io_bar0_read, + .write = vmxnet3_io_bar0_write, + .endianness = DEVICE_LITTLE_ENDIAN, + .impl = { + .min_access_size = 4, + .max_access_size = 4, + }, + }; + + static const MemoryRegionOps b1_ops = { + .read = vmxnet3_io_bar1_read, + .write = vmxnet3_io_bar1_write, + .endianness = DEVICE_LITTLE_ENDIAN, + .impl = { + .min_access_size = 4, + .max_access_size = 4, + }, + }; + + VMXNET3_State *s = DO_UPCAST(VMXNET3_State, dev, dev); + int i; + + DCBPRINTF("Starting init..."); + + memory_region_init_io(&s->bar0, &b0_ops, s, + "vmxnet3-b0", VMXNET3_PT_REG_SIZE); + pci_register_bar(&s->dev, VMXNET3_BAR0_IDX, + PCI_BASE_ADDRESS_SPACE_MEMORY, &s->bar0); + + memory_region_init_io(&s->bar1, &b1_ops, s, + "vmxnet3-b1", VMXNET3_VD_REG_SIZE); + pci_register_bar(&s->dev, VMXNET3_BAR1_IDX, + PCI_BASE_ADDRESS_SPACE_MEMORY, &s->bar1); + + memory_region_init(&s->msix_bar, "vmxnet3-msix-bar", + VMXNET3_MSIX_BAR_SIZE); + pci_register_bar(&s->dev, VMXNET3_MSIX_BAR_IDX, + PCI_BASE_ADDRESS_SPACE_MEMORY, &s->msix_bar); + + for (i = 0; i < ARRAY_SIZE(s->interrupt_states); i++) { + s->interrupt_states[i].is_asserted = false; + s->interrupt_states[i].is_pending = false; + s->interrupt_states[i].is_masked = true; + } + + /* Interrupt pin A */ + s->dev.config[PCI_INTERRUPT_PIN] = 0x01; + +#ifdef VMXNET3_ENABLE_MSIX + if (!vmxnet3_init_msix(s)) { + hw_error("Failed to initialize MSI-X, configuration is inconsistent."); + } +#endif + +#ifdef VMXNET3_ENABLE_MSI + if (!vmxnet3_init_msi(s)) { + hw_error("Failed to initialize MSI, configuration is inconsistent."); + } +#endif + + vmxnet3_net_init(s); + add_boot_device_path(s->conf.bootindex, &dev->qdev, "/ethernet-phy@0"); + + return 0; +} + + +static int vmxnet3_pci_uninit(PCIDevice *dev) +{ + VMXNET3_State *s = DO_UPCAST(VMXNET3_State, dev, dev); + + DCBPRINTF("Starting uninit..."); + + vmxnet3_net_uninit(s); + +#ifdef VMXNET3_ENABLE_MSIX + vmxnet3_cleanup_msix(s); +#endif + +#ifdef VMXNET3_ENABLE_MSI + vmxnet3_cleanup_msi(s); +#endif + + memory_region_destroy(&s->bar0); + memory_region_destroy(&s->bar1); + memory_region_destroy(&s->msix_bar); + + return 0; +} + +static void vmxnet3_qdev_reset(DeviceState *dev) +{ + VMXNET3_State *s = DO_UPCAST(VMXNET3_State, dev.qdev, dev); + DCBPRINTF("Starting QDEV reset..."); + vmxnet3_reset(s); +} + +static const VMStateDescription vmstate_vmxnet3 = { + .name = "vmxnet3", + .version_id = 1, + .minimum_version_id = 1, + .minimum_version_id_old = 1, + .fields = (VMStateField[]) { + VMSTATE_PCI_DEVICE(dev, VMXNET3_State), + VMSTATE_END_OF_LIST() + } +}; + +#if defined(VMXNET3_ENABLE_MSI) || defined(VMXNET3_ENABLE_MSIX) +static void +vmxnet3_write_config(PCIDevice *pci, uint32_t addr, uint32_t val, int len) +{ + pci_default_write_config(pci, addr, val, len); +#if defined(VMXNET3_ENABLE_MSIX) + msix_write_config(pci, addr, val, len); +#endif +#if defined(VMXNET3_ENABLE_MSI) + msi_write_config(pci, addr, val, len); +#endif +} +#endif + +static Property vmxnet3_properties[] = { + DEFINE_NIC_PROPERTIES(VMXNET3_State, conf), + DEFINE_PROP_END_OF_LIST(), +}; + +static void vmxnet3_class_init(ObjectClass *class, void *data) +{ + DeviceClass *dc = DEVICE_CLASS(class); + PCIDeviceClass *c = PCI_DEVICE_CLASS(class); + + c->init = vmxnet3_pci_init; + c->exit = vmxnet3_pci_uninit; + c->romfile = "pxe-e1000.rom"; + c->vendor_id = PCI_VENDOR_ID_VMWARE; + c->device_id = PCI_DEVICE_ID_VMWARE_VMXNET3; + c->revision = PCI_DEVICE_ID_VMWARE_VMXNET3_REVISION; + c->class_id = PCI_CLASS_NETWORK_ETHERNET; + c->subsystem_vendor_id = PCI_VENDOR_ID_VMWARE; + c->subsystem_id = PCI_DEVICE_ID_VMWARE_VMXNET3; +#if defined(VMXNET3_ENABLE_MSI) || defined(VMXNET3_ENABLE_MSIX) + c->config_write = vmxnet3_write_config, +#endif + dc->desc = "VMWare Paravirtualized Ethernet v3"; + dc->reset = vmxnet3_qdev_reset; + dc->vmsd = &vmstate_vmxnet3; + dc->props = vmxnet3_properties; +} + +static TypeInfo vmxnet3_info = { + .name = "vmxnet3", + .parent = TYPE_PCI_DEVICE, + .instance_size = sizeof(VMXNET3_State), + .class_init = vmxnet3_class_init, +}; + +static void vmxnet3_register_types(void) +{ + DCBPRINTF("vmxnet3_register_types called..."); + type_register_static(&vmxnet3_info); +} + +type_init(vmxnet3_register_types) diff --git a/qemu/hw/vmxnet3.h b/qemu/hw/vmxnet3.h new file mode 100644 index 0000000..6ec3fd5 --- /dev/null +++ b/qemu/hw/vmxnet3.h @@ -0,0 +1,727 @@ +/* + * QEMU VMWARE VMXNET3 paravirtual NIC + * + * Copyright (c) 2012 Ravello Systems LTD (http://ravellosystems.com) + * + * Developed by Daynix Computing LTD (http://www.daynix.com) + * + * Authors: + * Dmitry Fleytman + * Yan Vugenfirer + * + * 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_VMXNET3_H +#define _QEMU_VMXNET3_H + +#define VMXNET3_DEVICE_MAX_TX_QUEUES 8 +#define VMXNET3_DEVICE_MAX_RX_QUEUES 8 /* Keep this value as a power of 2 */ + +/* Defines needed to integrate VMWARE headers */ +#define u64 uint64_t +#define u32 uint32_t +#define u16 uint16_t +#define u8 uint8_t +#define __le16 uint16_t +#define __le32 uint32_t +#define __le64 uint64_t +#define __packed QEMU_PACKED + +#if defined(HOST_WORDS_BIGENDIAN) +#define const_cpu_to_le64(x) bswap_64(x) +#define __BIG_ENDIAN_BITFIELD +#else +#define const_cpu_to_le64(x) (x) +#endif + +/* Following is an interface definition for */ +/* VMXNET3 device as provided by VMWARE */ +/* Original file and copyright is available */ +/* in Linux kernel v3.2.8 at */ +/* drivers/net/vmxnet3/vmxnet3_defs.h */ + +struct UPT1_TxStats { + u64 TSOPktsTxOK; /* TSO pkts post-segmentation */ + u64 TSOBytesTxOK; + u64 ucastPktsTxOK; + u64 ucastBytesTxOK; + u64 mcastPktsTxOK; + u64 mcastBytesTxOK; + u64 bcastPktsTxOK; + u64 bcastBytesTxOK; + u64 pktsTxError; + u64 pktsTxDiscard; +}; + +struct UPT1_RxStats { + u64 LROPktsRxOK; /* LRO pkts */ + u64 LROBytesRxOK; /* bytes from LRO pkts */ + /* the following counters are for pkts from the wire, i.e., pre-LRO */ + u64 ucastPktsRxOK; + u64 ucastBytesRxOK; + u64 mcastPktsRxOK; + u64 mcastBytesRxOK; + u64 bcastPktsRxOK; + u64 bcastBytesRxOK; + u64 pktsRxOutOfBuf; + u64 pktsRxError; +}; + +/* interrupt moderation level */ +enum { + UPT1_IML_NONE = 0, /* no interrupt moderation */ + UPT1_IML_HIGHEST = 7, /* least intr generated */ + UPT1_IML_ADAPTIVE = 8, /* adpative intr moderation */ +}; +/* values for UPT1_RSSConf.hashFunc */ +enum { + UPT1_RSS_HASH_TYPE_NONE = 0x0, + UPT1_RSS_HASH_TYPE_IPV4 = 0x01, + UPT1_RSS_HASH_TYPE_TCP_IPV4 = 0x02, + UPT1_RSS_HASH_TYPE_IPV6 = 0x04, + UPT1_RSS_HASH_TYPE_TCP_IPV6 = 0x08, +}; + +enum { + UPT1_RSS_HASH_FUNC_NONE = 0x0, + UPT1_RSS_HASH_FUNC_TOEPLITZ = 0x01, +}; + +#define UPT1_RSS_MAX_KEY_SIZE 40 +#define UPT1_RSS_MAX_IND_TABLE_SIZE 128 + +struct UPT1_RSSConf { + u16 hashType; + u16 hashFunc; + u16 hashKeySize; + u16 indTableSize; + u8 hashKey[UPT1_RSS_MAX_KEY_SIZE]; + u8 indTable[UPT1_RSS_MAX_IND_TABLE_SIZE]; +}; + +/* features */ +enum { + UPT1_F_RXCSUM = const_cpu_to_le64(0x0001), /* rx csum verification */ + UPT1_F_RSS = const_cpu_to_le64(0x0002), + UPT1_F_RXVLAN = const_cpu_to_le64(0x0004), /* VLAN tag stripping */ + UPT1_F_LRO = const_cpu_to_le64(0x0008), +}; + +/* all registers are 32 bit wide */ +/* BAR 1 */ +enum { + VMXNET3_REG_VRRS = 0x0, /* Vmxnet3 Revision Report Selection */ + VMXNET3_REG_UVRS = 0x8, /* UPT Version Report Selection */ + VMXNET3_REG_DSAL = 0x10, /* Driver Shared Address Low */ + VMXNET3_REG_DSAH = 0x18, /* Driver Shared Address High */ + VMXNET3_REG_CMD = 0x20, /* Command */ + VMXNET3_REG_MACL = 0x28, /* MAC Address Low */ + VMXNET3_REG_MACH = 0x30, /* MAC Address High */ + VMXNET3_REG_ICR = 0x38, /* Interrupt Cause Register */ + VMXNET3_REG_ECR = 0x40 /* Event Cause Register */ +}; + +/* BAR 0 */ +enum { + VMXNET3_REG_IMR = 0x0, /* Interrupt Mask Register */ + VMXNET3_REG_TXPROD = 0x600, /* Tx Producer Index */ + VMXNET3_REG_RXPROD = 0x800, /* Rx Producer Index for ring 1 */ + VMXNET3_REG_RXPROD2 = 0xA00 /* Rx Producer Index for ring 2 */ +}; + +#define VMXNET3_PT_REG_SIZE 4096 /* BAR 0 */ +#define VMXNET3_VD_REG_SIZE 4096 /* BAR 1 */ + +#define VMXNET3_REG_ALIGN 8 /* All registers are 8-byte aligned. */ +#define VMXNET3_REG_ALIGN_MASK 0x7 + +/* I/O Mapped access to registers */ +#define VMXNET3_IO_TYPE_PT 0 +#define VMXNET3_IO_TYPE_VD 1 +#define VMXNET3_IO_ADDR(type, reg) (((type) << 24) | ((reg) & 0xFFFFFF)) +#define VMXNET3_IO_TYPE(addr) ((addr) >> 24) +#define VMXNET3_IO_REG(addr) ((addr) & 0xFFFFFF) + +enum { + VMXNET3_CMD_FIRST_SET = 0xCAFE0000, + VMXNET3_CMD_ACTIVATE_DEV = VMXNET3_CMD_FIRST_SET, /* 0xCAFE0000 */ + VMXNET3_CMD_QUIESCE_DEV, /* 0xCAFE0001 */ + VMXNET3_CMD_RESET_DEV, /* 0xCAFE0002 */ + VMXNET3_CMD_UPDATE_RX_MODE, /* 0xCAFE0003 */ + VMXNET3_CMD_UPDATE_MAC_FILTERS, /* 0xCAFE0004 */ + VMXNET3_CMD_UPDATE_VLAN_FILTERS, /* 0xCAFE0005 */ + VMXNET3_CMD_UPDATE_RSSIDT, /* 0xCAFE0006 */ + VMXNET3_CMD_UPDATE_IML, /* 0xCAFE0007 */ + VMXNET3_CMD_UPDATE_PMCFG, /* 0xCAFE0008 */ + VMXNET3_CMD_UPDATE_FEATURE, /* 0xCAFE0009 */ + VMXNET3_CMD_LOAD_PLUGIN, /* 0xCAFE000A */ + + VMXNET3_CMD_FIRST_GET = 0xF00D0000, + VMXNET3_CMD_GET_QUEUE_STATUS = VMXNET3_CMD_FIRST_GET, /* 0xF00D0000 */ + VMXNET3_CMD_GET_STATS, /* 0xF00D0001 */ + VMXNET3_CMD_GET_LINK, /* 0xF00D0002 */ + VMXNET3_CMD_GET_PERM_MAC_LO, /* 0xF00D0003 */ + VMXNET3_CMD_GET_PERM_MAC_HI, /* 0xF00D0004 */ + VMXNET3_CMD_GET_DID_LO, /* 0xF00D0005 */ + VMXNET3_CMD_GET_DID_HI, /* 0xF00D0006 */ + VMXNET3_CMD_GET_DEV_EXTRA_INFO, /* 0xF00D0007 */ + VMXNET3_CMD_GET_CONF_INTR /* 0xF00D0008 */ +}; + +/* + * Little Endian layout of bitfields - + * Byte 0 : 7.....len.....0 + * Byte 1 : rsvd gen 13.len.8 + * Byte 2 : 5.msscof.0 ext1 dtype + * Byte 3 : 13...msscof...6 + * + * Big Endian layout of bitfields - + * Byte 0: 13...msscof...6 + * Byte 1 : 5.msscof.0 ext1 dtype + * Byte 2 : rsvd gen 13.len.8 + * Byte 3 : 7.....len.....0 + * + * Thus, le32_to_cpu on the dword will allow the big endian driver to read + * the bit fields correctly. And cpu_to_le32 will convert bitfields + * bit fields written by big endian driver to format required by device. + */ + +struct Vmxnet3_TxDesc { + __le64 addr; + +#ifdef __BIG_ENDIAN_BITFIELD + u32 msscof:14; /* MSS, checksum offset, flags */ + u32 ext1:1; + u32 dtype:1; /* descriptor type */ + u32 rsvd:1; + u32 gen:1; /* generation bit */ + u32 len:14; +#else + u32 len:14; + u32 gen:1; /* generation bit */ + u32 rsvd:1; + u32 dtype:1; /* descriptor type */ + u32 ext1:1; + u32 msscof:14; /* MSS, checksum offset, flags */ +#endif /* __BIG_ENDIAN_BITFIELD */ + +#ifdef __BIG_ENDIAN_BITFIELD + u32 tci:16; /* Tag to Insert */ + u32 ti:1; /* VLAN Tag Insertion */ + u32 ext2:1; + u32 cq:1; /* completion request */ + u32 eop:1; /* End Of Packet */ + u32 om:2; /* offload mode */ + u32 hlen:10; /* header len */ +#else + u32 hlen:10; /* header len */ + u32 om:2; /* offload mode */ + u32 eop:1; /* End Of Packet */ + u32 cq:1; /* completion request */ + u32 ext2:1; + u32 ti:1; /* VLAN Tag Insertion */ + u32 tci:16; /* Tag to Insert */ +#endif /* __BIG_ENDIAN_BITFIELD */ +}; + +/* TxDesc.OM values */ +#define VMXNET3_OM_NONE 0 +#define VMXNET3_OM_CSUM 2 +#define VMXNET3_OM_TSO 3 + +/* fields in TxDesc we access w/o using bit fields */ +#define VMXNET3_TXD_EOP_SHIFT 12 +#define VMXNET3_TXD_CQ_SHIFT 13 +#define VMXNET3_TXD_GEN_SHIFT 14 +#define VMXNET3_TXD_EOP_DWORD_SHIFT 3 +#define VMXNET3_TXD_GEN_DWORD_SHIFT 2 + +#define VMXNET3_TXD_CQ (1 << VMXNET3_TXD_CQ_SHIFT) +#define VMXNET3_TXD_EOP (1 << VMXNET3_TXD_EOP_SHIFT) +#define VMXNET3_TXD_GEN (1 << VMXNET3_TXD_GEN_SHIFT) + +#define VMXNET3_HDR_COPY_SIZE 128 + + +struct Vmxnet3_TxDataDesc { + u8 data[VMXNET3_HDR_COPY_SIZE]; +}; + +#define VMXNET3_TCD_GEN_SHIFT 31 +#define VMXNET3_TCD_GEN_SIZE 1 +#define VMXNET3_TCD_TXIDX_SHIFT 0 +#define VMXNET3_TCD_TXIDX_SIZE 12 +#define VMXNET3_TCD_GEN_DWORD_SHIFT 3 + +struct Vmxnet3_TxCompDesc { + u32 txdIdx:12; /* Index of the EOP TxDesc */ + u32 ext1:20; + + __le32 ext2; + __le32 ext3; + + u32 rsvd:24; + u32 type:7; /* completion type */ + u32 gen:1; /* generation bit */ +}; + +struct Vmxnet3_RxDesc { + __le64 addr; + +#ifdef __BIG_ENDIAN_BITFIELD + u32 gen:1; /* Generation bit */ + u32 rsvd:15; + u32 dtype:1; /* Descriptor type */ + u32 btype:1; /* Buffer Type */ + u32 len:14; +#else + u32 len:14; + u32 btype:1; /* Buffer Type */ + u32 dtype:1; /* Descriptor type */ + u32 rsvd:15; + u32 gen:1; /* Generation bit */ +#endif + u32 ext1; +}; + +/* values of RXD.BTYPE */ +#define VMXNET3_RXD_BTYPE_HEAD 0 /* head only */ +#define VMXNET3_RXD_BTYPE_BODY 1 /* body only */ + +/* fields in RxDesc we access w/o using bit fields */ +#define VMXNET3_RXD_BTYPE_SHIFT 14 +#define VMXNET3_RXD_GEN_SHIFT 31 + +struct Vmxnet3_RxCompDesc { +#ifdef __BIG_ENDIAN_BITFIELD + u32 ext2:1; + u32 cnc:1; /* Checksum Not Calculated */ + u32 rssType:4; /* RSS hash type used */ + u32 rqID:10; /* rx queue/ring ID */ + u32 sop:1; /* Start of Packet */ + u32 eop:1; /* End of Packet */ + u32 ext1:2; + u32 rxdIdx:12; /* Index of the RxDesc */ +#else + u32 rxdIdx:12; /* Index of the RxDesc */ + u32 ext1:2; + u32 eop:1; /* End of Packet */ + u32 sop:1; /* Start of Packet */ + u32 rqID:10; /* rx queue/ring ID */ + u32 rssType:4; /* RSS hash type used */ + u32 cnc:1; /* Checksum Not Calculated */ + u32 ext2:1; +#endif /* __BIG_ENDIAN_BITFIELD */ + + __le32 rssHash; /* RSS hash value */ + +#ifdef __BIG_ENDIAN_BITFIELD + u32 tci:16; /* Tag stripped */ + u32 ts:1; /* Tag is stripped */ + u32 err:1; /* Error */ + u32 len:14; /* data length */ +#else + u32 len:14; /* data length */ + u32 err:1; /* Error */ + u32 ts:1; /* Tag is stripped */ + u32 tci:16; /* Tag stripped */ +#endif /* __BIG_ENDIAN_BITFIELD */ + + +#ifdef __BIG_ENDIAN_BITFIELD + u32 gen:1; /* generation bit */ + u32 type:7; /* completion type */ + u32 fcs:1; /* Frame CRC correct */ + u32 frg:1; /* IP Fragment */ + u32 v4:1; /* IPv4 */ + u32 v6:1; /* IPv6 */ + u32 ipc:1; /* IP Checksum Correct */ + u32 tcp:1; /* TCP packet */ + u32 udp:1; /* UDP packet */ + u32 tuc:1; /* TCP/UDP Checksum Correct */ + u32 csum:16; +#else + u32 csum:16; + u32 tuc:1; /* TCP/UDP Checksum Correct */ + u32 udp:1; /* UDP packet */ + u32 tcp:1; /* TCP packet */ + u32 ipc:1; /* IP Checksum Correct */ + u32 v6:1; /* IPv6 */ + u32 v4:1; /* IPv4 */ + u32 frg:1; /* IP Fragment */ + u32 fcs:1; /* Frame CRC correct */ + u32 type:7; /* completion type */ + u32 gen:1; /* generation bit */ +#endif /* __BIG_ENDIAN_BITFIELD */ +}; + +/* fields in RxCompDesc we access via Vmxnet3_GenericDesc.dword[3] */ +#define VMXNET3_RCD_TUC_SHIFT 16 +#define VMXNET3_RCD_IPC_SHIFT 19 + +/* fields in RxCompDesc we access via Vmxnet3_GenericDesc.qword[1] */ +#define VMXNET3_RCD_TYPE_SHIFT 56 +#define VMXNET3_RCD_GEN_SHIFT 63 + +/* csum OK for TCP/UDP pkts over IP */ +#define VMXNET3_RCD_CSUM_OK (1 << VMXNET3_RCD_TUC_SHIFT | \ + 1 << VMXNET3_RCD_IPC_SHIFT) +#define VMXNET3_TXD_GEN_SIZE 1 +#define VMXNET3_TXD_EOP_SIZE 1 + +/* value of RxCompDesc.rssType */ +enum { + VMXNET3_RCD_RSS_TYPE_NONE = 0, + VMXNET3_RCD_RSS_TYPE_IPV4 = 1, + VMXNET3_RCD_RSS_TYPE_TCPIPV4 = 2, + VMXNET3_RCD_RSS_TYPE_IPV6 = 3, + VMXNET3_RCD_RSS_TYPE_TCPIPV6 = 4, +}; + + +/* a union for accessing all cmd/completion descriptors */ +union Vmxnet3_GenericDesc { + __le64 qword[2]; + __le32 dword[4]; + __le16 word[8]; + struct Vmxnet3_TxDesc txd; + struct Vmxnet3_RxDesc rxd; + struct Vmxnet3_TxCompDesc tcd; + struct Vmxnet3_RxCompDesc rcd; +}; + +#define VMXNET3_INIT_GEN 1 + +/* Max size of a single tx buffer */ +#define VMXNET3_MAX_TX_BUF_SIZE (1 << 14) + +/* # of tx desc needed for a tx buffer size */ +#define VMXNET3_TXD_NEEDED(size) (((size) + VMXNET3_MAX_TX_BUF_SIZE - 1) / \ + VMXNET3_MAX_TX_BUF_SIZE) + +/* max # of tx descs for a non-tso pkt */ +#define VMXNET3_MAX_TXD_PER_PKT 16 + +/* Max size of a single rx buffer */ +#define VMXNET3_MAX_RX_BUF_SIZE ((1 << 14) - 1) +/* Minimum size of a type 0 buffer */ +#define VMXNET3_MIN_T0_BUF_SIZE 128 +#define VMXNET3_MAX_CSUM_OFFSET 1024 + +/* Ring base address alignment */ +#define VMXNET3_RING_BA_ALIGN 512 +#define VMXNET3_RING_BA_MASK (VMXNET3_RING_BA_ALIGN - 1) + +/* Ring size must be a multiple of 32 */ +#define VMXNET3_RING_SIZE_ALIGN 32 +#define VMXNET3_RING_SIZE_MASK (VMXNET3_RING_SIZE_ALIGN - 1) + +/* Max ring size */ +#define VMXNET3_TX_RING_MAX_SIZE 4096 +#define VMXNET3_TC_RING_MAX_SIZE 4096 +#define VMXNET3_RX_RING_MAX_SIZE 4096 +#define VMXNET3_RC_RING_MAX_SIZE 8192 + +/* a list of reasons for queue stop */ + +enum { + VMXNET3_ERR_NOEOP = 0x80000000, /* cannot find the EOP desc of a pkt */ + VMXNET3_ERR_TXD_REUSE = 0x80000001, /* reuse TxDesc before tx completion */ + VMXNET3_ERR_BIG_PKT = 0x80000002, /* too many TxDesc for a pkt */ + VMXNET3_ERR_DESC_NOT_SPT = 0x80000003, /* descriptor type not supported */ + VMXNET3_ERR_SMALL_BUF = 0x80000004, /* type 0 buffer too small */ + VMXNET3_ERR_STRESS = 0x80000005, /* stress option firing in vmkernel */ + VMXNET3_ERR_SWITCH = 0x80000006, /* mode switch failure */ + VMXNET3_ERR_TXD_INVALID = 0x80000007, /* invalid TxDesc */ +}; + +/* completion descriptor types */ +#define VMXNET3_CDTYPE_TXCOMP 0 /* Tx Completion Descriptor */ +#define VMXNET3_CDTYPE_RXCOMP 3 /* Rx Completion Descriptor */ + +enum { + VMXNET3_GOS_BITS_UNK = 0, /* unknown */ + VMXNET3_GOS_BITS_32 = 1, + VMXNET3_GOS_BITS_64 = 2, +}; + +#define VMXNET3_GOS_TYPE_UNK 0 /* unknown */ +#define VMXNET3_GOS_TYPE_LINUX 1 +#define VMXNET3_GOS_TYPE_WIN 2 +#define VMXNET3_GOS_TYPE_SOLARIS 3 +#define VMXNET3_GOS_TYPE_FREEBSD 4 +#define VMXNET3_GOS_TYPE_PXE 5 + +struct Vmxnet3_GOSInfo { +#ifdef __BIG_ENDIAN_BITFIELD + u32 gosMisc:10; /* other info about gos */ + u32 gosVer:16; /* gos version */ + u32 gosType:4; /* which guest */ + u32 gosBits:2; /* 32-bit or 64-bit? */ +#else + u32 gosBits:2; /* 32-bit or 64-bit? */ + u32 gosType:4; /* which guest */ + u32 gosVer:16; /* gos version */ + u32 gosMisc:10; /* other info about gos */ +#endif /* __BIG_ENDIAN_BITFIELD */ +}; + +struct Vmxnet3_DriverInfo { + __le32 version; + struct Vmxnet3_GOSInfo gos; + __le32 vmxnet3RevSpt; + __le32 uptVerSpt; +}; + + +#define VMXNET3_REV1_MAGIC 0xbabefee1 + +/* + * QueueDescPA must be 128 bytes aligned. It points to an array of + * Vmxnet3_TxQueueDesc followed by an array of Vmxnet3_RxQueueDesc. + * The number of Vmxnet3_TxQueueDesc/Vmxnet3_RxQueueDesc are specified by + * Vmxnet3_MiscConf.numTxQueues/numRxQueues, respectively. + */ +#define VMXNET3_QUEUE_DESC_ALIGN 128 + + +struct Vmxnet3_MiscConf { + struct Vmxnet3_DriverInfo driverInfo; + __le64 uptFeatures; + __le64 ddPA; /* driver data PA */ + __le64 queueDescPA; /* queue descriptor table PA */ + __le32 ddLen; /* driver data len */ + __le32 queueDescLen; /* queue desc. table len in bytes */ + __le32 mtu; + __le16 maxNumRxSG; + u8 numTxQueues; + u8 numRxQueues; + __le32 reserved[4]; +}; + + +struct Vmxnet3_TxQueueConf { + __le64 txRingBasePA; + __le64 dataRingBasePA; + __le64 compRingBasePA; + __le64 ddPA; /* driver data */ + __le64 reserved; + __le32 txRingSize; /* # of tx desc */ + __le32 dataRingSize; /* # of data desc */ + __le32 compRingSize; /* # of comp desc */ + __le32 ddLen; /* size of driver data */ + u8 intrIdx; + u8 _pad[7]; +}; + + +struct Vmxnet3_RxQueueConf { + __le64 rxRingBasePA[2]; + __le64 compRingBasePA; + __le64 ddPA; /* driver data */ + __le64 reserved; + __le32 rxRingSize[2]; /* # of rx desc */ + __le32 compRingSize; /* # of rx comp desc */ + __le32 ddLen; /* size of driver data */ + u8 intrIdx; + u8 _pad[7]; +}; + + +enum vmxnet3_intr_mask_mode { + VMXNET3_IMM_AUTO = 0, + VMXNET3_IMM_ACTIVE = 1, + VMXNET3_IMM_LAZY = 2 +}; + +enum vmxnet3_intr_type { + VMXNET3_IT_AUTO = 0, + VMXNET3_IT_INTX = 1, + VMXNET3_IT_MSI = 2, + VMXNET3_IT_MSIX = 3 +}; + +#define VMXNET3_MAX_TX_QUEUES 8 +#define VMXNET3_MAX_RX_QUEUES 16 +/* addition 1 for events */ +#define VMXNET3_MAX_INTRS 25 + +/* value of intrCtrl */ +#define VMXNET3_IC_DISABLE_ALL 0x1 /* bit 0 */ + + +struct Vmxnet3_IntrConf { + bool autoMask; + u8 numIntrs; /* # of interrupts */ + u8 eventIntrIdx; + u8 modLevels[VMXNET3_MAX_INTRS]; /* moderation level for + * each intr */ + __le32 intrCtrl; + __le32 reserved[2]; +}; + +/* one bit per VLAN ID, the size is in the units of u32 */ +#define VMXNET3_VFT_SIZE (4096/(sizeof(uint32_t)*8)) + + +struct Vmxnet3_QueueStatus { + bool stopped; + u8 _pad[3]; + __le32 error; +}; + + +struct Vmxnet3_TxQueueCtrl { + __le32 txNumDeferred; + __le32 txThreshold; + __le64 reserved; +}; + + +struct Vmxnet3_RxQueueCtrl { + bool updateRxProd; + u8 _pad[7]; + __le64 reserved; +}; + +enum { + VMXNET3_RXM_UCAST = 0x01, /* unicast only */ + VMXNET3_RXM_MCAST = 0x02, /* multicast passing the filters */ + VMXNET3_RXM_BCAST = 0x04, /* broadcast only */ + VMXNET3_RXM_ALL_MULTI = 0x08, /* all multicast */ + VMXNET3_RXM_PROMISC = 0x10 /* promiscuous */ +}; + +struct Vmxnet3_RxFilterConf { + __le32 rxMode; /* VMXNET3_RXM_xxx */ + __le16 mfTableLen; /* size of the multicast filter table */ + __le16 _pad1; + __le64 mfTablePA; /* PA of the multicast filters table */ + __le32 vfTable[VMXNET3_VFT_SIZE]; /* vlan filter */ +}; + + +#define VMXNET3_PM_MAX_FILTERS 6 +#define VMXNET3_PM_MAX_PATTERN_SIZE 128 +#define VMXNET3_PM_MAX_MASK_SIZE (VMXNET3_PM_MAX_PATTERN_SIZE / 8) + +#define VMXNET3_PM_WAKEUP_MAGIC cpu_to_le16(0x01) /* wake up on magic pkts */ +#define VMXNET3_PM_WAKEUP_FILTER cpu_to_le16(0x02) /* wake up on pkts matching + * filters */ + + +struct Vmxnet3_PM_PktFilter { + u8 maskSize; + u8 patternSize; + u8 mask[VMXNET3_PM_MAX_MASK_SIZE]; + u8 pattern[VMXNET3_PM_MAX_PATTERN_SIZE]; + u8 pad[6]; +}; + + +struct Vmxnet3_PMConf { + __le16 wakeUpEvents; /* VMXNET3_PM_WAKEUP_xxx */ + u8 numFilters; + u8 pad[5]; + struct Vmxnet3_PM_PktFilter filters[VMXNET3_PM_MAX_FILTERS]; +}; + + +struct Vmxnet3_VariableLenConfDesc { + __le32 confVer; + __le32 confLen; + __le64 confPA; +}; + + +struct Vmxnet3_TxQueueDesc { + struct Vmxnet3_TxQueueCtrl ctrl; + struct Vmxnet3_TxQueueConf conf; + + /* Driver read after a GET command */ + struct Vmxnet3_QueueStatus status; + struct UPT1_TxStats stats; + u8 _pad[88]; /* 128 aligned */ +}; + + +struct Vmxnet3_RxQueueDesc { + struct Vmxnet3_RxQueueCtrl ctrl; + struct Vmxnet3_RxQueueConf conf; + /* Driver read after a GET commad */ + struct Vmxnet3_QueueStatus status; + struct UPT1_RxStats stats; + u8 __pad[88]; /* 128 aligned */ +}; + + +struct Vmxnet3_DSDevRead { + /* read-only region for device, read by dev in response to a SET cmd */ + struct Vmxnet3_MiscConf misc; + struct Vmxnet3_IntrConf intrConf; + struct Vmxnet3_RxFilterConf rxFilterConf; + struct Vmxnet3_VariableLenConfDesc rssConfDesc; + struct Vmxnet3_VariableLenConfDesc pmConfDesc; + struct Vmxnet3_VariableLenConfDesc pluginConfDesc; +}; + +/* All structures in DriverShared are padded to multiples of 8 bytes */ +struct Vmxnet3_DriverShared { + __le32 magic; + /* make devRead start at 64bit boundaries */ + __le32 pad; + struct Vmxnet3_DSDevRead devRead; + __le32 ecr; + __le32 reserved[5]; +}; + + +#define VMXNET3_ECR_RQERR (1 << 0) +#define VMXNET3_ECR_TQERR (1 << 1) +#define VMXNET3_ECR_LINK (1 << 2) +#define VMXNET3_ECR_DIC (1 << 3) +#define VMXNET3_ECR_DEBUG (1 << 4) + +/* flip the gen bit of a ring */ +#define VMXNET3_FLIP_RING_GEN(gen) ((gen) = (gen) ^ 0x1) + +/* only use this if moving the idx won't affect the gen bit */ +#define VMXNET3_INC_RING_IDX_ONLY(idx, ring_size) \ + do {\ + (idx)++;\ + if (unlikely((idx) == (ring_size))) {\ + (idx) = 0;\ + } \ + } while (0) + +#define VMXNET3_SET_VFTABLE_ENTRY(vfTable, vid) \ + (vfTable[vid >> 5] |= (1 << (vid & 31))) +#define VMXNET3_CLEAR_VFTABLE_ENTRY(vfTable, vid) \ + (vfTable[vid >> 5] &= ~(1 << (vid & 31))) + +#define VMXNET3_VFTABLE_ENTRY_IS_SET(vfTable, vid) \ + ((vfTable[vid >> 5] & (1 << (vid & 31))) != 0) + +#define VMXNET3_MAX_MTU 9000 +#define VMXNET3_MIN_MTU 60 + +#define VMXNET3_LINK_UP (10000 << 16 | 1) /* 10 Gbps, up */ +#define VMXNET3_LINK_DOWN 0 + +#undef u64 +#undef u32 +#undef u16 +#undef u8 +#undef __le16 +#undef __le32 +#undef __le64 +#undef __packed +#undef const_cpu_to_le64 +#if defined(HOST_WORDS_BIGENDIAN) +#undef __BIG_ENDIAN_BITFIELD +#endif + +#endif diff --git a/qemu/hw/vmxnet3_debug.h b/qemu/hw/vmxnet3_debug.h new file mode 100644 index 0000000..8383c22 --- /dev/null +++ b/qemu/hw/vmxnet3_debug.h @@ -0,0 +1,104 @@ +/* + * QEMU VMWARE VMXNET3 paravirtual NIC - debugging facilities + * + * Copyright (c) 2012 Ravello Systems LTD (http://ravellosystems.com) + * + * Developed by Daynix Computing LTD (http://www.daynix.com) + * + * Authors: + * Dmitry Fleytman + * Yan Vugenfirer + * + * 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_VMXNET3_DEBUG_H +#define _QEMU_VMXNET3_DEBUG_H + +/* #define DEBUG_VMXNET3_CB */ +#define DEBUG_VMXNET3_WARNINGS +#define DEBUG_VMXNET3_ERRORS +/* #define DEBUG_VMXNET3_INTERRUPTS */ +/* #define DEBUG_VMXNET3_CONFIG */ +/* #define DEBUG_VMXNET3_SHMEM_ACCESS */ +/* #define DEBUG_VMXNET3_RINGS */ +/* #define DEBUG_VMXNET3_PACKETS */ + +#ifdef DEBUG_VMXNET3_SHMEM_ACCESS +#define DSHPRINTF(fmt, ...) \ + do { \ + printf("[vmxnet3][SH][%s]: " fmt "\n", __func__, ## __VA_ARGS__); \ + } while (0) +#else +#define DSHPRINTF(fmt, ...) do {} while (0) +#endif + +#ifdef DEBUG_VMXNET3_CB +#define DCBPRINTF(fmt, ...) \ + do { \ + printf("[vmxnet3][CB][%s]: " fmt "\n", __func__, ## __VA_ARGS__); \ + } while (0) +#else +#define DCBPRINTF(fmt, ...) do {} while (0) +#endif + +#ifdef DEBUG_VMXNET3_PACKETS +#define DPKPRINTF(fmt, ...) \ + do { \ + printf("[vmxnet3][PK][%s]: " fmt "\n", __func__, ## __VA_ARGS__); \ + } while (0) +#else +#define DPKPRINTF(fmt, ...) do {} while (0) +#endif + +#ifdef DEBUG_VMXNET3_WARNINGS +#define DWRPRINTF(fmt, ...) \ + do { \ + printf("[vmxnet3][WR][%s]: " fmt "\n", __func__, ## __VA_ARGS__); \ + } while (0) +#else +#define DWRPRINTF(fmt, ...) do {} while (0) +#endif + +#ifdef DEBUG_VMXNET3_ERRORS +#define DERPRINTF(fmt, ...) \ + do { \ + printf("[vmxnet3][ER][%s]: " fmt "\n", __func__, ## __VA_ARGS__); \ + } while (0) +#else +#define DERPRINTF(fmt, ...) do {} while (0) +#endif + +#ifdef DEBUG_VMXNET3_INTERRUPTS +#define DIRPRINTF(fmt, ...) \ + do { \ + printf("[vmxnet3][IR][%s]: " fmt "\n", __func__, ## __VA_ARGS__); \ + } while (0) +#else +#define DIRPRINTF(fmt, ...) do {} while (0) +#endif + +#ifdef DEBUG_VMXNET3_CONFIG +#define DCFPRINTF(fmt, ...) \ + do { \ + printf("[vmxnet3][CF][%s]: " fmt "\n", __func__, ## __VA_ARGS__); \ + } while (0) +#else +#define DCFPRINTF(fmt, ...) do {} while (0) +#endif + +#ifdef DEBUG_VMXNET3_RINGS +#define DRIPRINTF(fmt, ...) \ + do { \ + printf("[vmxnet3][RI][%s]: " fmt "\n", __func__, ## __VA_ARGS__); \ + } while (0) +#else +#define DRIPRINTF(fmt, ...) do {} while (0) +#endif + +#define MAC_FMT "%02X:%02X:%02X:%02X:%02X:%02X" +#define MAC_ARG(a) (a)[0], (a)[1], (a)[2], (a)[3], (a)[4], (a)[5] + +#endif /* _QEMU_VMXNET3_DEBUG_H */ diff --git a/qemu/hw/vmxnet_utils.c b/qemu/hw/vmxnet_utils.c new file mode 100644 index 0000000..e310828 --- /dev/null +++ b/qemu/hw/vmxnet_utils.c @@ -0,0 +1,172 @@ +/* + * QEMU VMWARE paravirtual devices - network auxiliary code + * + * Copyright (c) 2012 Ravello Systems LTD (http://ravellosystems.com) + * + * Developed by Daynix Computing LTD (http://www.daynix.com) + * + * Authors: + * Dmitry Fleytman + * Yan Vugenfirer + * + * 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 "hw.h" +#include "virtio-net.h" +#include "vmxnet_utils.h" +#include "net/checksum.h" + +void eth_setup_vlan_headers(struct eth_header *ehdr, uint16_t vlan_tag) +{ + switch (be16_to_cpu(ehdr->h_proto)) { + case ETH_P_VLAN: + case ETH_P_DVLAN: { + /* Header already present, just put proper VLAN tag */ + struct vlan_header *vhdr = PKT_GET_VLAN_HDR(ehdr); + vhdr->h_tci = cpu_to_be16(vlan_tag); + } + default: { + /* No VLAN header, put a new one */ + struct vlan_header *vhdr = PKT_GET_VLAN_HDR(ehdr); + vhdr->h_proto = ehdr->h_proto; + ehdr->h_proto = cpu_to_be16(ETH_P_VLAN); + vhdr->h_tci = cpu_to_be16(vlan_tag); + } + } +} + +bool eth_setup_tx_offloads(uint8_t *l3hdr, + size_t l3hdr_len, + size_t l3hdr_off, + uint32_t l3payload_len, + struct virtio_net_hdr *vhdr, + bool more_frags, + uint16_t fragmentation_offset) +{ + uint16_t csum; + + switch (vhdr->gso_type & ~VIRTIO_NET_HDR_GSO_ECN) { + case VIRTIO_NET_HDR_GSO_TCPV4: + case VIRTIO_NET_HDR_GSO_UDP: { + struct ip_header *iphdr = (struct ip_header *) l3hdr; + uint16_t new_ip_off; + + vhdr->flags = VIRTIO_NET_HDR_F_NEEDS_CSUM; + vhdr->csum_start = l3hdr_off; + vhdr->csum_offset = offsetof(struct ip_header, ip_sum); + + if (l3payload_len + l3hdr_len > ETH_MAX_IP_DGRAM_LEN) { + /* This must never happen with fragmentation enabled */ + assert(0 == more_frags); + assert(0 == fragmentation_offset); + return false; + } + + iphdr->ip_len = cpu_to_be16(l3payload_len + l3hdr_len); + assert(0 == (~IP_OFFMASK & fragmentation_offset)); + + new_ip_off = fragmentation_offset | (more_frags ? IP_MF : 0) | + (be16_to_cpu(iphdr->ip_off) & ~(IP_OFFMASK|IP_MF)); + + iphdr->ip_off = cpu_to_be16(new_ip_off); + + /* Due to Linux bridge bugs/features IP header checksum */ + /* must be calculated in order to make it process */ + /* packet with segmentation requirements successfully */ + eth_put_csum(l3hdr, vhdr->csum_offset, 0); + csum = net_raw_checksum(l3hdr, l3hdr_len); + eth_put_csum(l3hdr, vhdr->csum_offset, csum); + } + break; + + case VIRTIO_NET_HDR_GSO_TCPV6: + default: + vhdr->flags = 0; + break; + } + + return true; +} + +uint8_t +eth_get_gso_type(uint16_t l3_proto, uint8_t *l3_hdr) +{ + uint8_t ecn_state = 0; + + if (ETH_P_IP == l3_proto) { + struct ip_header *iphdr = (struct ip_header *) l3_hdr; + + if (IP_HEADER_VERSION_4 == IP_HEADER_VERSION(iphdr)) { + if (IPTOS_ECN_CE == IPTOS_ECN(iphdr->ip_tos)) { + ecn_state = VIRTIO_NET_HDR_GSO_ECN; + } + if (IP_PROTO_TCP == iphdr->ip_p) { + return VIRTIO_NET_HDR_GSO_TCPV4 | ecn_state; + } else if (IP_PROTO_UDP == iphdr->ip_p) { + return VIRTIO_NET_HDR_GSO_UDP | ecn_state; + } + } + } else if (ETH_P_IPV6 == l3_proto) { + struct ip6_header *ip6hdr = (struct ip6_header *) l3_hdr; + + if (IP6_ECN_CE == IP6_ECN(ip6hdr->ip6_ecn_acc)) { + ecn_state = VIRTIO_NET_HDR_GSO_ECN; + } + + if (IP_PROTO_TCP == ip6hdr->ip6_nxt) { + return VIRTIO_NET_HDR_GSO_TCPV6 | ecn_state; + } + } + + /* Unsupported offload */ + assert(false); + + return VIRTIO_NET_HDR_GSO_NONE | ecn_state; +} + +void eth_get_protocols(const uint8_t *headers, + uint32_t hdr_length, + bool *isip4, bool *isip6, + bool *isudp, bool *istcp) +{ + int proto; + size_t l2hdr_len = eth_get_l2_hdr_length(headers); + assert(hdr_length >= eth_get_l2_hdr_length(headers)); + *isip4 = *isip6 = *isudp = *istcp = false; + + proto = eth_get_l3_proto(headers, l2hdr_len); + if (ETH_P_IP == proto) { + *isip4 = true; + + struct ip_header *iphdr; + + assert(hdr_length >= + eth_get_l2_hdr_length(headers) + sizeof(struct ip_header)); + + iphdr = PKT_GET_IP_HDR(headers); + + if (IP_HEADER_VERSION_4 == IP_HEADER_VERSION(iphdr)) { + if (IP_PROTO_TCP == iphdr->ip_p) { + *istcp = true; + } else if (IP_PROTO_UDP == iphdr->ip_p) { + *isudp = true; + } + } + } else if (ETH_P_IPV6 == proto) { + *isip6 = true; + + struct ip6_header *ip6hdr; + assert(hdr_length >= + eth_get_l2_hdr_length(headers) + sizeof(struct ip6_header)); + ip6hdr = PKT_GET_IP6_HDR(headers); + + if (IP_PROTO_TCP == ip6hdr->ip6_nxt) { + *istcp = true; + } else if (IP_PROTO_UDP == ip6hdr->ip6_nxt) { + *isudp = true; + } + } +} diff --git a/qemu/hw/vmxnet_utils.h b/qemu/hw/vmxnet_utils.h new file mode 100644 index 0000000..f5e79dd --- /dev/null +++ b/qemu/hw/vmxnet_utils.h @@ -0,0 +1,242 @@ +/* + * QEMU VMWARE paravirtual devices - network auxiliary code + * + * Copyright (c) 2012 Ravello Systems LTD (http://ravellosystems.com) + * + * Developed by Daynix Computing LTD (http://www.daynix.com) + * + * Authors: + * Dmitry Fleytman + * Yan Vugenfirer + * + * This work is licensed under the terms of the GNU GPL, version 2 or later. + * See the COPYING file in the top-level directory. + * + */ + +#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 dest address */ +}; + +/* 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 */ +}; +#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_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) \ + ((0 == (x)) || (0xFFF == (x))) + +#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_IP6_HDR_LEN \ + (sizeof(struct ip6_header)) +#define ETH_MAX_L3_HDR_LEN \ + (MAX(ETH_MAX_IP4_HDR_LEN, ETH_MAX_IP6_HDR_LEN)) +#define ETH_MAX_IP_DGRAM_LEN (0xFFFF) +#define ETH_MAX_IP_PLOAD_LEN \ + (ETH_MAX_IP_DGRAM_LEN - ETH_MAX_IP4_HDR_LEN - ETH_MAX_L2_HDR_LEN) + +#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 */ + + +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 { + VMXNET3_PKT_UCAST = 0xAABBCC00, + VMXNET3_PKT_BCAST, + VMXNET3_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 VMXNET3_PKT_BCAST; + } else if (is_multicast_ether_addr(ehdr->h_dest)) { + return VMXNET3_PKT_MCAST; + } else { /* unicast */ + return VMXNET3_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 (ETH_P_VLAN == hvlan->h_proto) { + 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_vlan_tag(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_proto); + default: + return 0; + } +} + +static inline bool +eth_strip_vlan(const void *p, struct eth_header *new_ehdr, + uint16_t *payload_offset, uint16_t *vlan_tag) +{ + 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: + 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; + *vlan_tag = be16_to_cpu(hvlan->h_tci); + *payload_offset = + sizeof(struct eth_header) + 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); +} + +static inline void +eth_put_csum(uint8_t *buf, uint32_t cso, uint16_t csum) +{ + cpu_to_be16wu((uint16_t *)(buf + cso), csum); +} + +void eth_setup_vlan_headers(struct eth_header *ehdr, uint16_t vlan_tag); + + +bool eth_setup_tx_offloads(uint8_t *l3hdr, + size_t l3hdr_len, + size_t l3hdr_off, + uint32_t l3payload_len, + struct virtio_net_hdr *vhdr, + bool more_frags, + uint16_t fragmentation_offset); + +uint8_t eth_get_gso_type(uint16_t l3_proto, uint8_t *l3_hdr); + +void eth_get_protocols(const uint8_t *headers, + uint32_t hdr_length, + bool *isip4, bool *isip6, + bool *isudp, bool *istcp); diff --git a/qemu/net.c b/qemu/net.c index c34474f..e2f586c 100644 --- a/qemu/net.c +++ b/qemu/net.c @@ -857,7 +857,7 @@ static const struct { }, { .name = "model", .type = QEMU_OPT_STRING, - .help = "device model (e1000, rtl8139, virtio etc.)", + .help = "device model (e1000, rtl8139, virtio, vmxnet3 etc.)", }, { .name = "addr", .type = QEMU_OPT_STRING, diff --git a/qemu/net/checksum.h b/qemu/net/checksum.h index 1f05298..5f42a02 100644 --- a/qemu/net/checksum.h +++ b/qemu/net/checksum.h @@ -26,4 +26,11 @@ 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 uint16_t +net_raw_checksum(uint8_t *data, int length) +{ + return net_checksum_finish(net_checksum_add(length, data)); +} + + #endif /* QEMU_NET_CHECKSUM_H */