From patchwork Fri Jan 30 06:36:17 2015 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Huawei Xie X-Patchwork-Id: 2797 Return-Path: X-Original-To: patchwork@dpdk.org Delivered-To: patchwork@dpdk.org Received: from [92.243.14.124] (localhost [IPv6:::1]) by dpdk.org (Postfix) with ESMTP id 56CC95AAA; Fri, 30 Jan 2015 07:36:48 +0100 (CET) Received: from mga02.intel.com (mga02.intel.com [134.134.136.20]) by dpdk.org (Postfix) with ESMTP id 464555AAA for ; Fri, 30 Jan 2015 07:36:43 +0100 (CET) Received: from orsmga003.jf.intel.com ([10.7.209.27]) by orsmga101.jf.intel.com with ESMTP; 29 Jan 2015 22:36:42 -0800 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.09,490,1418112000"; d="scan'208";a="520051435" Received: from shvmail01.sh.intel.com ([10.239.29.42]) by orsmga003.jf.intel.com with ESMTP; 29 Jan 2015 22:29:23 -0800 Received: from shecgisg003.sh.intel.com (shecgisg003.sh.intel.com [10.239.29.90]) by shvmail01.sh.intel.com with ESMTP id t0U6acfE006251; Fri, 30 Jan 2015 14:36:38 +0800 Received: from shecgisg003.sh.intel.com (localhost [127.0.0.1]) by shecgisg003.sh.intel.com (8.13.6/8.13.6/SuSE Linux 0.8) with ESMTP id t0U6aa0D012058; Fri, 30 Jan 2015 14:36:38 +0800 Received: (from hxie5@localhost) by shecgisg003.sh.intel.com (8.13.6/8.13.6/Submit) id t0U6aabt012054; Fri, 30 Jan 2015 14:36:36 +0800 From: Huawei Xie To: dev@dpdk.org Date: Fri, 30 Jan 2015 14:36:17 +0800 Message-Id: <1422599787-12009-3-git-send-email-huawei.xie@intel.com> X-Mailer: git-send-email 1.7.4.1 In-Reply-To: <1422599787-12009-1-git-send-email-huawei.xie@intel.com> References: <1422599787-12009-1-git-send-email-huawei.xie@intel.com> Subject: [dpdk-dev] [PATCH 02/12] lib/librte_vhost: seperate vhost cuse driver from vhost common logic X-BeenThere: dev@dpdk.org X-Mailman-Version: 2.1.15 Precedence: list List-Id: patches and discussions about DPDK List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: dev-bounces@dpdk.org Sender: "dev" create vhost_cuse directory move vhost-net-cdev.c into vhost_cuse directory vhost-cuse driver will be splitted into two parts: cuse driver specific message handling and common vhost message handling logic. cuse driver specific message handling is in vhost_cuse directory. vhost ioctl message is pre-processed there and then sent to virtio-net(virtio-net.c) module if necessary. Some message handling is terminated in vhost-cuse or vhost-user. virtio-net.c provides common message handling for both vhost-cuse and vhost-user. Signed-off-by: Huawei Xie --- lib/librte_vhost/Makefile | 4 +- lib/librte_vhost/vhost-net-cdev.c | 389 --------------------------- lib/librte_vhost/vhost_cuse/vhost-net-cdev.c | 389 +++++++++++++++++++++++++++ 3 files changed, 391 insertions(+), 391 deletions(-) delete mode 100644 lib/librte_vhost/vhost-net-cdev.c create mode 100644 lib/librte_vhost/vhost_cuse/vhost-net-cdev.c diff --git a/lib/librte_vhost/Makefile b/lib/librte_vhost/Makefile index c008d64..0b2f08f 100644 --- a/lib/librte_vhost/Makefile +++ b/lib/librte_vhost/Makefile @@ -34,10 +34,10 @@ include $(RTE_SDK)/mk/rte.vars.mk # library name LIB = librte_vhost.a -CFLAGS += $(WERROR_FLAGS) -I$(SRCDIR) -O3 -D_FILE_OFFSET_BITS=64 -lfuse +CFLAGS += $(WERROR_FLAGS) -I$(SRCDIR) -I vhost_cuse -O3 -D_FILE_OFFSET_BITS=64 -lfuse LDFLAGS += -lfuse # all source are stored in SRCS-y -SRCS-$(CONFIG_RTE_LIBRTE_VHOST) := vhost-net-cdev.c virtio-net.c vhost_rxtx.c +SRCS-$(CONFIG_RTE_LIBRTE_VHOST) := vhost_cuse/vhost-net-cdev.c virtio-net.c vhost_rxtx.c # install includes SYMLINK-$(CONFIG_RTE_LIBRTE_VHOST)-include += rte_virtio_net.h diff --git a/lib/librte_vhost/vhost-net-cdev.c b/lib/librte_vhost/vhost-net-cdev.c deleted file mode 100644 index 57c76cb..0000000 --- a/lib/librte_vhost/vhost-net-cdev.c +++ /dev/null @@ -1,389 +0,0 @@ -/*- - * BSD LICENSE - * - * Copyright(c) 2010-2014 Intel Corporation. All rights reserved. - * All rights reserved. - * - * Redistribution and use in source and binary forms, with or without - * modification, are permitted provided that the following conditions - * are met: - * - * * Redistributions of source code must retain the above copyright - * notice, this list of conditions and the following disclaimer. - * * Redistributions in binary form must reproduce the above copyright - * notice, this list of conditions and the following disclaimer in - * the documentation and/or other materials provided with the - * distribution. - * * Neither the name of Intel Corporation nor the names of its - * contributors may be used to endorse or promote products derived - * from this software without specific prior written permission. - * - * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS - * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT - * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR - * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT - * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, - * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT - * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, - * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY - * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT - * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE - * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. - */ - -#include -#include -#include -#include -#include -#include -#include - -#include -#include -#include -#include - -#include "vhost-net-cdev.h" - -#define FUSE_OPT_DUMMY "\0\0" -#define FUSE_OPT_FORE "-f\0\0" -#define FUSE_OPT_NOMULTI "-s\0\0" - -static const uint32_t default_major = 231; -static const uint32_t default_minor = 1; -static const char cuse_device_name[] = "/dev/cuse"; -static const char default_cdev[] = "vhost-net"; - -static struct fuse_session *session; -static struct vhost_net_device_ops const *ops; - -/* - * Returns vhost_device_ctx from given fuse_req_t. The index is populated later - * when the device is added to the device linked list. - */ -static struct vhost_device_ctx -fuse_req_to_vhost_ctx(fuse_req_t req, struct fuse_file_info *fi) -{ - struct vhost_device_ctx ctx; - struct fuse_ctx const *const req_ctx = fuse_req_ctx(req); - - ctx.pid = req_ctx->pid; - ctx.fh = fi->fh; - - return ctx; -} - -/* - * When the device is created in QEMU it gets initialised here and - * added to the device linked list. - */ -static void -vhost_net_open(fuse_req_t req, struct fuse_file_info *fi) -{ - struct vhost_device_ctx ctx = fuse_req_to_vhost_ctx(req, fi); - int err = 0; - - err = ops->new_device(ctx); - if (err == -1) { - fuse_reply_err(req, EPERM); - return; - } - - fi->fh = err; - - RTE_LOG(INFO, VHOST_CONFIG, - "(%"PRIu64") Device configuration started\n", fi->fh); - fuse_reply_open(req, fi); -} - -/* - * When QEMU is shutdown or killed the device gets released. - */ -static void -vhost_net_release(fuse_req_t req, struct fuse_file_info *fi) -{ - int err = 0; - struct vhost_device_ctx ctx = fuse_req_to_vhost_ctx(req, fi); - - ops->destroy_device(ctx); - RTE_LOG(INFO, VHOST_CONFIG, "(%"PRIu64") Device released\n", ctx.fh); - fuse_reply_err(req, err); -} - -/* - * Boilerplate code for CUSE IOCTL - * Implicit arguments: ctx, req, result. - */ -#define VHOST_IOCTL(func) do { \ - result = (func)(ctx); \ - fuse_reply_ioctl(req, result, NULL, 0); \ -} while (0) - -/* - * Boilerplate IOCTL RETRY - * Implicit arguments: req. - */ -#define VHOST_IOCTL_RETRY(size_r, size_w) do { \ - struct iovec iov_r = { arg, (size_r) }; \ - struct iovec iov_w = { arg, (size_w) }; \ - fuse_reply_ioctl_retry(req, &iov_r, \ - (size_r) ? 1 : 0, &iov_w, (size_w) ? 1 : 0);\ -} while (0) - -/* - * Boilerplate code for CUSE Read IOCTL - * Implicit arguments: ctx, req, result, in_bufsz, in_buf. - */ -#define VHOST_IOCTL_R(type, var, func) do { \ - if (!in_bufsz) { \ - VHOST_IOCTL_RETRY(sizeof(type), 0);\ - } else { \ - (var) = *(const type*)in_buf; \ - result = func(ctx, &(var)); \ - fuse_reply_ioctl(req, result, NULL, 0);\ - } \ -} while (0) - -/* - * Boilerplate code for CUSE Write IOCTL - * Implicit arguments: ctx, req, result, out_bufsz. - */ -#define VHOST_IOCTL_W(type, var, func) do { \ - if (!out_bufsz) { \ - VHOST_IOCTL_RETRY(0, sizeof(type));\ - } else { \ - result = (func)(ctx, &(var));\ - fuse_reply_ioctl(req, result, &(var), sizeof(type));\ - } \ -} while (0) - -/* - * Boilerplate code for CUSE Read/Write IOCTL - * Implicit arguments: ctx, req, result, in_bufsz, in_buf. - */ -#define VHOST_IOCTL_RW(type1, var1, type2, var2, func) do { \ - if (!in_bufsz) { \ - VHOST_IOCTL_RETRY(sizeof(type1), sizeof(type2));\ - } else { \ - (var1) = *(const type1*) (in_buf); \ - result = (func)(ctx, (var1), &(var2)); \ - fuse_reply_ioctl(req, result, &(var2), sizeof(type2));\ - } \ -} while (0) - -/* - * The IOCTLs are handled using CUSE/FUSE in userspace. Depending on the type - * of IOCTL a buffer is requested to read or to write. This request is handled - * by FUSE and the buffer is then given to CUSE. - */ -static void -vhost_net_ioctl(fuse_req_t req, int cmd, void *arg, - struct fuse_file_info *fi, __rte_unused unsigned flags, - const void *in_buf, size_t in_bufsz, size_t out_bufsz) -{ - struct vhost_device_ctx ctx = fuse_req_to_vhost_ctx(req, fi); - struct vhost_vring_file file; - struct vhost_vring_state state; - struct vhost_vring_addr addr; - uint64_t features; - uint32_t index; - int result = 0; - - switch (cmd) { - case VHOST_NET_SET_BACKEND: - LOG_DEBUG(VHOST_CONFIG, - "(%"PRIu64") IOCTL: VHOST_NET_SET_BACKEND\n", ctx.fh); - VHOST_IOCTL_R(struct vhost_vring_file, file, ops->set_backend); - break; - - case VHOST_GET_FEATURES: - LOG_DEBUG(VHOST_CONFIG, - "(%"PRIu64") IOCTL: VHOST_GET_FEATURES\n", ctx.fh); - VHOST_IOCTL_W(uint64_t, features, ops->get_features); - break; - - case VHOST_SET_FEATURES: - LOG_DEBUG(VHOST_CONFIG, - "(%"PRIu64") IOCTL: VHOST_SET_FEATURES\n", ctx.fh); - VHOST_IOCTL_R(uint64_t, features, ops->set_features); - break; - - case VHOST_RESET_OWNER: - LOG_DEBUG(VHOST_CONFIG, - "(%"PRIu64") IOCTL: VHOST_RESET_OWNER\n", ctx.fh); - VHOST_IOCTL(ops->reset_owner); - break; - - case VHOST_SET_OWNER: - LOG_DEBUG(VHOST_CONFIG, - "(%"PRIu64") IOCTL: VHOST_SET_OWNER\n", ctx.fh); - VHOST_IOCTL(ops->set_owner); - break; - - case VHOST_SET_MEM_TABLE: - /*TODO fix race condition.*/ - LOG_DEBUG(VHOST_CONFIG, - "(%"PRIu64") IOCTL: VHOST_SET_MEM_TABLE\n", ctx.fh); - static struct vhost_memory mem_temp; - - switch (in_bufsz) { - case 0: - VHOST_IOCTL_RETRY(sizeof(struct vhost_memory), 0); - break; - - case sizeof(struct vhost_memory): - mem_temp = *(const struct vhost_memory *) in_buf; - - if (mem_temp.nregions > 0) { - VHOST_IOCTL_RETRY(sizeof(struct vhost_memory) + - (sizeof(struct vhost_memory_region) * - mem_temp.nregions), 0); - } else { - result = -1; - fuse_reply_ioctl(req, result, NULL, 0); - } - break; - - default: - result = ops->set_mem_table(ctx, - in_buf, mem_temp.nregions); - if (result) - fuse_reply_err(req, EINVAL); - else - fuse_reply_ioctl(req, result, NULL, 0); - } - break; - - case VHOST_SET_VRING_NUM: - LOG_DEBUG(VHOST_CONFIG, - "(%"PRIu64") IOCTL: VHOST_SET_VRING_NUM\n", ctx.fh); - VHOST_IOCTL_R(struct vhost_vring_state, state, - ops->set_vring_num); - break; - - case VHOST_SET_VRING_BASE: - LOG_DEBUG(VHOST_CONFIG, - "(%"PRIu64") IOCTL: VHOST_SET_VRING_BASE\n", ctx.fh); - VHOST_IOCTL_R(struct vhost_vring_state, state, - ops->set_vring_base); - break; - - case VHOST_GET_VRING_BASE: - LOG_DEBUG(VHOST_CONFIG, - "(%"PRIu64") IOCTL: VHOST_GET_VRING_BASE\n", ctx.fh); - VHOST_IOCTL_RW(uint32_t, index, - struct vhost_vring_state, state, ops->get_vring_base); - break; - - case VHOST_SET_VRING_ADDR: - LOG_DEBUG(VHOST_CONFIG, - "(%"PRIu64") IOCTL: VHOST_SET_VRING_ADDR\n", ctx.fh); - VHOST_IOCTL_R(struct vhost_vring_addr, addr, - ops->set_vring_addr); - break; - - case VHOST_SET_VRING_KICK: - LOG_DEBUG(VHOST_CONFIG, - "(%"PRIu64") IOCTL: VHOST_SET_VRING_KICK\n", ctx.fh); - VHOST_IOCTL_R(struct vhost_vring_file, file, - ops->set_vring_kick); - break; - - case VHOST_SET_VRING_CALL: - LOG_DEBUG(VHOST_CONFIG, - "(%"PRIu64") IOCTL: VHOST_SET_VRING_CALL\n", ctx.fh); - VHOST_IOCTL_R(struct vhost_vring_file, file, - ops->set_vring_call); - break; - - default: - RTE_LOG(ERR, VHOST_CONFIG, - "(%"PRIu64") IOCTL: DOESN NOT EXIST\n", ctx.fh); - result = -1; - fuse_reply_ioctl(req, result, NULL, 0); - } - - if (result < 0) - LOG_DEBUG(VHOST_CONFIG, - "(%"PRIu64") IOCTL: FAIL\n", ctx.fh); - else - LOG_DEBUG(VHOST_CONFIG, - "(%"PRIu64") IOCTL: SUCCESS\n", ctx.fh); -} - -/* - * Structure handling open, release and ioctl function pointers is populated. - */ -static const struct cuse_lowlevel_ops vhost_net_ops = { - .open = vhost_net_open, - .release = vhost_net_release, - .ioctl = vhost_net_ioctl, -}; - -/* - * cuse_info is populated and used to register the cuse device. - * vhost_net_device_ops are also passed when the device is registered in app. - */ -int -rte_vhost_driver_register(const char *dev_name) -{ - struct cuse_info cuse_info; - char device_name[PATH_MAX] = ""; - char char_device_name[PATH_MAX] = ""; - const char *device_argv[] = { device_name }; - - char fuse_opt_dummy[] = FUSE_OPT_DUMMY; - char fuse_opt_fore[] = FUSE_OPT_FORE; - char fuse_opt_nomulti[] = FUSE_OPT_NOMULTI; - char *fuse_argv[] = {fuse_opt_dummy, fuse_opt_fore, fuse_opt_nomulti}; - - if (access(cuse_device_name, R_OK | W_OK) < 0) { - RTE_LOG(ERR, VHOST_CONFIG, - "char device %s can't be accessed, maybe not exist\n", - cuse_device_name); - return -1; - } - - /* - * The device name is created. This is passed to QEMU so that it can - * register the device with our application. - */ - snprintf(device_name, PATH_MAX, "DEVNAME=%s", dev_name); - snprintf(char_device_name, PATH_MAX, "/dev/%s", dev_name); - - /* Check if device already exists. */ - if (access(char_device_name, F_OK) != -1) { - RTE_LOG(ERR, VHOST_CONFIG, - "char device %s already exists\n", char_device_name); - return -1; - } - - memset(&cuse_info, 0, sizeof(cuse_info)); - cuse_info.dev_major = default_major; - cuse_info.dev_minor = default_minor; - cuse_info.dev_info_argc = 1; - cuse_info.dev_info_argv = device_argv; - cuse_info.flags = CUSE_UNRESTRICTED_IOCTL; - - ops = get_virtio_net_callbacks(); - - session = cuse_lowlevel_setup(3, fuse_argv, - &cuse_info, &vhost_net_ops, 0, NULL); - if (session == NULL) - return -1; - - return 0; -} - -/** - * The CUSE session is launched allowing the application to receive open, - * release and ioctl calls. - */ -int -rte_vhost_driver_session_start(void) -{ - fuse_session_loop(session); - - return 0; -} diff --git a/lib/librte_vhost/vhost_cuse/vhost-net-cdev.c b/lib/librte_vhost/vhost_cuse/vhost-net-cdev.c new file mode 100644 index 0000000..57c76cb --- /dev/null +++ b/lib/librte_vhost/vhost_cuse/vhost-net-cdev.c @@ -0,0 +1,389 @@ +/*- + * BSD LICENSE + * + * Copyright(c) 2010-2014 Intel Corporation. All rights reserved. + * All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * * Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * * Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * * Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + */ + +#include +#include +#include +#include +#include +#include +#include + +#include +#include +#include +#include + +#include "vhost-net-cdev.h" + +#define FUSE_OPT_DUMMY "\0\0" +#define FUSE_OPT_FORE "-f\0\0" +#define FUSE_OPT_NOMULTI "-s\0\0" + +static const uint32_t default_major = 231; +static const uint32_t default_minor = 1; +static const char cuse_device_name[] = "/dev/cuse"; +static const char default_cdev[] = "vhost-net"; + +static struct fuse_session *session; +static struct vhost_net_device_ops const *ops; + +/* + * Returns vhost_device_ctx from given fuse_req_t. The index is populated later + * when the device is added to the device linked list. + */ +static struct vhost_device_ctx +fuse_req_to_vhost_ctx(fuse_req_t req, struct fuse_file_info *fi) +{ + struct vhost_device_ctx ctx; + struct fuse_ctx const *const req_ctx = fuse_req_ctx(req); + + ctx.pid = req_ctx->pid; + ctx.fh = fi->fh; + + return ctx; +} + +/* + * When the device is created in QEMU it gets initialised here and + * added to the device linked list. + */ +static void +vhost_net_open(fuse_req_t req, struct fuse_file_info *fi) +{ + struct vhost_device_ctx ctx = fuse_req_to_vhost_ctx(req, fi); + int err = 0; + + err = ops->new_device(ctx); + if (err == -1) { + fuse_reply_err(req, EPERM); + return; + } + + fi->fh = err; + + RTE_LOG(INFO, VHOST_CONFIG, + "(%"PRIu64") Device configuration started\n", fi->fh); + fuse_reply_open(req, fi); +} + +/* + * When QEMU is shutdown or killed the device gets released. + */ +static void +vhost_net_release(fuse_req_t req, struct fuse_file_info *fi) +{ + int err = 0; + struct vhost_device_ctx ctx = fuse_req_to_vhost_ctx(req, fi); + + ops->destroy_device(ctx); + RTE_LOG(INFO, VHOST_CONFIG, "(%"PRIu64") Device released\n", ctx.fh); + fuse_reply_err(req, err); +} + +/* + * Boilerplate code for CUSE IOCTL + * Implicit arguments: ctx, req, result. + */ +#define VHOST_IOCTL(func) do { \ + result = (func)(ctx); \ + fuse_reply_ioctl(req, result, NULL, 0); \ +} while (0) + +/* + * Boilerplate IOCTL RETRY + * Implicit arguments: req. + */ +#define VHOST_IOCTL_RETRY(size_r, size_w) do { \ + struct iovec iov_r = { arg, (size_r) }; \ + struct iovec iov_w = { arg, (size_w) }; \ + fuse_reply_ioctl_retry(req, &iov_r, \ + (size_r) ? 1 : 0, &iov_w, (size_w) ? 1 : 0);\ +} while (0) + +/* + * Boilerplate code for CUSE Read IOCTL + * Implicit arguments: ctx, req, result, in_bufsz, in_buf. + */ +#define VHOST_IOCTL_R(type, var, func) do { \ + if (!in_bufsz) { \ + VHOST_IOCTL_RETRY(sizeof(type), 0);\ + } else { \ + (var) = *(const type*)in_buf; \ + result = func(ctx, &(var)); \ + fuse_reply_ioctl(req, result, NULL, 0);\ + } \ +} while (0) + +/* + * Boilerplate code for CUSE Write IOCTL + * Implicit arguments: ctx, req, result, out_bufsz. + */ +#define VHOST_IOCTL_W(type, var, func) do { \ + if (!out_bufsz) { \ + VHOST_IOCTL_RETRY(0, sizeof(type));\ + } else { \ + result = (func)(ctx, &(var));\ + fuse_reply_ioctl(req, result, &(var), sizeof(type));\ + } \ +} while (0) + +/* + * Boilerplate code for CUSE Read/Write IOCTL + * Implicit arguments: ctx, req, result, in_bufsz, in_buf. + */ +#define VHOST_IOCTL_RW(type1, var1, type2, var2, func) do { \ + if (!in_bufsz) { \ + VHOST_IOCTL_RETRY(sizeof(type1), sizeof(type2));\ + } else { \ + (var1) = *(const type1*) (in_buf); \ + result = (func)(ctx, (var1), &(var2)); \ + fuse_reply_ioctl(req, result, &(var2), sizeof(type2));\ + } \ +} while (0) + +/* + * The IOCTLs are handled using CUSE/FUSE in userspace. Depending on the type + * of IOCTL a buffer is requested to read or to write. This request is handled + * by FUSE and the buffer is then given to CUSE. + */ +static void +vhost_net_ioctl(fuse_req_t req, int cmd, void *arg, + struct fuse_file_info *fi, __rte_unused unsigned flags, + const void *in_buf, size_t in_bufsz, size_t out_bufsz) +{ + struct vhost_device_ctx ctx = fuse_req_to_vhost_ctx(req, fi); + struct vhost_vring_file file; + struct vhost_vring_state state; + struct vhost_vring_addr addr; + uint64_t features; + uint32_t index; + int result = 0; + + switch (cmd) { + case VHOST_NET_SET_BACKEND: + LOG_DEBUG(VHOST_CONFIG, + "(%"PRIu64") IOCTL: VHOST_NET_SET_BACKEND\n", ctx.fh); + VHOST_IOCTL_R(struct vhost_vring_file, file, ops->set_backend); + break; + + case VHOST_GET_FEATURES: + LOG_DEBUG(VHOST_CONFIG, + "(%"PRIu64") IOCTL: VHOST_GET_FEATURES\n", ctx.fh); + VHOST_IOCTL_W(uint64_t, features, ops->get_features); + break; + + case VHOST_SET_FEATURES: + LOG_DEBUG(VHOST_CONFIG, + "(%"PRIu64") IOCTL: VHOST_SET_FEATURES\n", ctx.fh); + VHOST_IOCTL_R(uint64_t, features, ops->set_features); + break; + + case VHOST_RESET_OWNER: + LOG_DEBUG(VHOST_CONFIG, + "(%"PRIu64") IOCTL: VHOST_RESET_OWNER\n", ctx.fh); + VHOST_IOCTL(ops->reset_owner); + break; + + case VHOST_SET_OWNER: + LOG_DEBUG(VHOST_CONFIG, + "(%"PRIu64") IOCTL: VHOST_SET_OWNER\n", ctx.fh); + VHOST_IOCTL(ops->set_owner); + break; + + case VHOST_SET_MEM_TABLE: + /*TODO fix race condition.*/ + LOG_DEBUG(VHOST_CONFIG, + "(%"PRIu64") IOCTL: VHOST_SET_MEM_TABLE\n", ctx.fh); + static struct vhost_memory mem_temp; + + switch (in_bufsz) { + case 0: + VHOST_IOCTL_RETRY(sizeof(struct vhost_memory), 0); + break; + + case sizeof(struct vhost_memory): + mem_temp = *(const struct vhost_memory *) in_buf; + + if (mem_temp.nregions > 0) { + VHOST_IOCTL_RETRY(sizeof(struct vhost_memory) + + (sizeof(struct vhost_memory_region) * + mem_temp.nregions), 0); + } else { + result = -1; + fuse_reply_ioctl(req, result, NULL, 0); + } + break; + + default: + result = ops->set_mem_table(ctx, + in_buf, mem_temp.nregions); + if (result) + fuse_reply_err(req, EINVAL); + else + fuse_reply_ioctl(req, result, NULL, 0); + } + break; + + case VHOST_SET_VRING_NUM: + LOG_DEBUG(VHOST_CONFIG, + "(%"PRIu64") IOCTL: VHOST_SET_VRING_NUM\n", ctx.fh); + VHOST_IOCTL_R(struct vhost_vring_state, state, + ops->set_vring_num); + break; + + case VHOST_SET_VRING_BASE: + LOG_DEBUG(VHOST_CONFIG, + "(%"PRIu64") IOCTL: VHOST_SET_VRING_BASE\n", ctx.fh); + VHOST_IOCTL_R(struct vhost_vring_state, state, + ops->set_vring_base); + break; + + case VHOST_GET_VRING_BASE: + LOG_DEBUG(VHOST_CONFIG, + "(%"PRIu64") IOCTL: VHOST_GET_VRING_BASE\n", ctx.fh); + VHOST_IOCTL_RW(uint32_t, index, + struct vhost_vring_state, state, ops->get_vring_base); + break; + + case VHOST_SET_VRING_ADDR: + LOG_DEBUG(VHOST_CONFIG, + "(%"PRIu64") IOCTL: VHOST_SET_VRING_ADDR\n", ctx.fh); + VHOST_IOCTL_R(struct vhost_vring_addr, addr, + ops->set_vring_addr); + break; + + case VHOST_SET_VRING_KICK: + LOG_DEBUG(VHOST_CONFIG, + "(%"PRIu64") IOCTL: VHOST_SET_VRING_KICK\n", ctx.fh); + VHOST_IOCTL_R(struct vhost_vring_file, file, + ops->set_vring_kick); + break; + + case VHOST_SET_VRING_CALL: + LOG_DEBUG(VHOST_CONFIG, + "(%"PRIu64") IOCTL: VHOST_SET_VRING_CALL\n", ctx.fh); + VHOST_IOCTL_R(struct vhost_vring_file, file, + ops->set_vring_call); + break; + + default: + RTE_LOG(ERR, VHOST_CONFIG, + "(%"PRIu64") IOCTL: DOESN NOT EXIST\n", ctx.fh); + result = -1; + fuse_reply_ioctl(req, result, NULL, 0); + } + + if (result < 0) + LOG_DEBUG(VHOST_CONFIG, + "(%"PRIu64") IOCTL: FAIL\n", ctx.fh); + else + LOG_DEBUG(VHOST_CONFIG, + "(%"PRIu64") IOCTL: SUCCESS\n", ctx.fh); +} + +/* + * Structure handling open, release and ioctl function pointers is populated. + */ +static const struct cuse_lowlevel_ops vhost_net_ops = { + .open = vhost_net_open, + .release = vhost_net_release, + .ioctl = vhost_net_ioctl, +}; + +/* + * cuse_info is populated and used to register the cuse device. + * vhost_net_device_ops are also passed when the device is registered in app. + */ +int +rte_vhost_driver_register(const char *dev_name) +{ + struct cuse_info cuse_info; + char device_name[PATH_MAX] = ""; + char char_device_name[PATH_MAX] = ""; + const char *device_argv[] = { device_name }; + + char fuse_opt_dummy[] = FUSE_OPT_DUMMY; + char fuse_opt_fore[] = FUSE_OPT_FORE; + char fuse_opt_nomulti[] = FUSE_OPT_NOMULTI; + char *fuse_argv[] = {fuse_opt_dummy, fuse_opt_fore, fuse_opt_nomulti}; + + if (access(cuse_device_name, R_OK | W_OK) < 0) { + RTE_LOG(ERR, VHOST_CONFIG, + "char device %s can't be accessed, maybe not exist\n", + cuse_device_name); + return -1; + } + + /* + * The device name is created. This is passed to QEMU so that it can + * register the device with our application. + */ + snprintf(device_name, PATH_MAX, "DEVNAME=%s", dev_name); + snprintf(char_device_name, PATH_MAX, "/dev/%s", dev_name); + + /* Check if device already exists. */ + if (access(char_device_name, F_OK) != -1) { + RTE_LOG(ERR, VHOST_CONFIG, + "char device %s already exists\n", char_device_name); + return -1; + } + + memset(&cuse_info, 0, sizeof(cuse_info)); + cuse_info.dev_major = default_major; + cuse_info.dev_minor = default_minor; + cuse_info.dev_info_argc = 1; + cuse_info.dev_info_argv = device_argv; + cuse_info.flags = CUSE_UNRESTRICTED_IOCTL; + + ops = get_virtio_net_callbacks(); + + session = cuse_lowlevel_setup(3, fuse_argv, + &cuse_info, &vhost_net_ops, 0, NULL); + if (session == NULL) + return -1; + + return 0; +} + +/** + * The CUSE session is launched allowing the application to receive open, + * release and ioctl calls. + */ +int +rte_vhost_driver_session_start(void) +{ + fuse_session_loop(session); + + return 0; +}