uverbs_cmd.c

来自「linux 内核源代码」· C语言 代码 · 共 2,158 行 · 第 1/4 页

/*
 * Copyright (c) 2005 Topspin Communications.  All rights reserved.
 * Copyright (c) 2005, 2006, 2007 Cisco Systems.  All rights reserved.
 * Copyright (c) 2005 PathScale, Inc.  All rights reserved.
 * Copyright (c) 2006 Mellanox Technologies.  All rights reserved.
 *
 * This software is available to you under a choice of one of two
 * licenses.  You may choose to be licensed under the terms of the GNU
 * General Public License (GPL) Version 2, available from the file
 * COPYING in the main directory of this source tree, or the
 * OpenIB.org BSD license below:
 *
 *     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.
 *
 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
 * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
 * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
 * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
 * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
 * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
 * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
 * SOFTWARE.
 *
 * $Id: uverbs_cmd.c 2708 2005-06-24 17:27:21Z roland $
 */

#include <linux/file.h>
#include <linux/fs.h>

#include <asm/uaccess.h>

#include "uverbs.h"

static struct lock_class_key pd_lock_key;
static struct lock_class_key mr_lock_key;
static struct lock_class_key cq_lock_key;
static struct lock_class_key qp_lock_key;
static struct lock_class_key ah_lock_key;
static struct lock_class_key srq_lock_key;

#define INIT_UDATA(udata, ibuf, obuf, ilen, olen)			\
	do {								\
		(udata)->inbuf  = (void __user *) (ibuf);		\
		(udata)->outbuf = (void __user *) (obuf);		\
		(udata)->inlen  = (ilen);				\
		(udata)->outlen = (olen);				\
	} while (0)

/*
 * The ib_uobject locking scheme is as follows:
 *
 * - ib_uverbs_idr_lock protects the uverbs idrs themselves, so it
 *   needs to be held during all idr operations.  When an object is
 *   looked up, a reference must be taken on the object's kref before
 *   dropping this lock.
 *
 * - Each object also has an rwsem.  This rwsem must be held for
 *   reading while an operation that uses the object is performed.
 *   For example, while registering an MR, the associated PD's
 *   uobject.mutex must be held for reading.  The rwsem must be held
 *   for writing while initializing or destroying an object.
 *
 * - In addition, each object has a "live" flag.  If this flag is not
 *   set, then lookups of the object will fail even if it is found in
 *   the idr.  This handles a reader that blocks and does not acquire
 *   the rwsem until after the object is destroyed.  The destroy
 *   operation will set the live flag to 0 and then drop the rwsem;
 *   this will allow the reader to acquire the rwsem, see that the
 *   live flag is 0, and then drop the rwsem and its reference to
 *   object.  The underlying storage will not be freed until the last
 *   reference to the object is dropped.
 */

static void init_uobj(struct ib_uobject *uobj, u64 user_handle,
		      struct ib_ucontext *context, struct lock_class_key *key)
{
	uobj->user_handle = user_handle;
	uobj->context     = context;
	kref_init(&uobj->ref);
	init_rwsem(&uobj->mutex);
	lockdep_set_class(&uobj->mutex, key);
	uobj->live        = 0;
}

static void release_uobj(struct kref *kref)
{
	kfree(container_of(kref, struct ib_uobject, ref));
}

static void put_uobj(struct ib_uobject *uobj)
{
	kref_put(&uobj->ref, release_uobj);
}

static void put_uobj_read(struct ib_uobject *uobj)
{
	up_read(&uobj->mutex);
	put_uobj(uobj);
}

static void put_uobj_write(struct ib_uobject *uobj)
{
	up_write(&uobj->mutex);
	put_uobj(uobj);
}

static int idr_add_uobj(struct idr *idr, struct ib_uobject *uobj)
{
	int ret;

retry:
	if (!idr_pre_get(idr, GFP_KERNEL))
		return -ENOMEM;

	spin_lock(&ib_uverbs_idr_lock);
	ret = idr_get_new(idr, uobj, &uobj->id);
	spin_unlock(&ib_uverbs_idr_lock);

	if (ret == -EAGAIN)
		goto retry;

	return ret;
}

void idr_remove_uobj(struct idr *idr, struct ib_uobject *uobj)
{
	spin_lock(&ib_uverbs_idr_lock);
	idr_remove(idr, uobj->id);
	spin_unlock(&ib_uverbs_idr_lock);
}

static struct ib_uobject *__idr_get_uobj(struct idr *idr, int id,
					 struct ib_ucontext *context)
{
	struct ib_uobject *uobj;

	spin_lock(&ib_uverbs_idr_lock);
	uobj = idr_find(idr, id);
	if (uobj) {
		if (uobj->context == context)
			kref_get(&uobj->ref);
		else
			uobj = NULL;
	}
	spin_unlock(&ib_uverbs_idr_lock);

	return uobj;
}

static struct ib_uobject *idr_read_uobj(struct idr *idr, int id,
					struct ib_ucontext *context, int nested)
{
	struct ib_uobject *uobj;

	uobj = __idr_get_uobj(idr, id, context);
	if (!uobj)
		return NULL;

	if (nested)
		down_read_nested(&uobj->mutex, SINGLE_DEPTH_NESTING);
	else
		down_read(&uobj->mutex);
	if (!uobj->live) {
		put_uobj_read(uobj);
		return NULL;
	}

	return uobj;
}

static struct ib_uobject *idr_write_uobj(struct idr *idr, int id,
					 struct ib_ucontext *context)
{
	struct ib_uobject *uobj;

	uobj = __idr_get_uobj(idr, id, context);
	if (!uobj)
		return NULL;

	down_write(&uobj->mutex);
	if (!uobj->live) {
		put_uobj_write(uobj);
		return NULL;
	}

	return uobj;
}

static void *idr_read_obj(struct idr *idr, int id, struct ib_ucontext *context,
			  int nested)
{
	struct ib_uobject *uobj;

	uobj = idr_read_uobj(idr, id, context, nested);
	return uobj ? uobj->object : NULL;
}

static struct ib_pd *idr_read_pd(int pd_handle, struct ib_ucontext *context)
{
	return idr_read_obj(&ib_uverbs_pd_idr, pd_handle, context, 0);
}

static void put_pd_read(struct ib_pd *pd)
{
	put_uobj_read(pd->uobject);
}

static struct ib_cq *idr_read_cq(int cq_handle, struct ib_ucontext *context, int nested)
{
	return idr_read_obj(&ib_uverbs_cq_idr, cq_handle, context, nested);
}

static void put_cq_read(struct ib_cq *cq)
{
	put_uobj_read(cq->uobject);
}

static struct ib_ah *idr_read_ah(int ah_handle, struct ib_ucontext *context)
{
	return idr_read_obj(&ib_uverbs_ah_idr, ah_handle, context, 0);
}

static void put_ah_read(struct ib_ah *ah)
{
	put_uobj_read(ah->uobject);
}

static struct ib_qp *idr_read_qp(int qp_handle, struct ib_ucontext *context)
{
	return idr_read_obj(&ib_uverbs_qp_idr, qp_handle, context, 0);
}

static void put_qp_read(struct ib_qp *qp)
{
	put_uobj_read(qp->uobject);
}

static struct ib_srq *idr_read_srq(int srq_handle, struct ib_ucontext *context)
{
	return idr_read_obj(&ib_uverbs_srq_idr, srq_handle, context, 0);
}

static void put_srq_read(struct ib_srq *srq)
{
	put_uobj_read(srq->uobject);
}

ssize_t ib_uverbs_get_context(struct ib_uverbs_file *file,
			      const char __user *buf,
			      int in_len, int out_len)
{
	struct ib_uverbs_get_context      cmd;
	struct ib_uverbs_get_context_resp resp;
	struct ib_udata                   udata;
	struct ib_device                 *ibdev = file->device->ib_dev;
	struct ib_ucontext		 *ucontext;
	struct file			 *filp;
	int ret;

	if (out_len < sizeof resp)
		return -ENOSPC;

	if (copy_from_user(&cmd, buf, sizeof cmd))
		return -EFAULT;

	mutex_lock(&file->mutex);

	if (file->ucontext) {
		ret = -EINVAL;
		goto err;
	}

	INIT_UDATA(&udata, buf + sizeof cmd,
		   (unsigned long) cmd.response + sizeof resp,
		   in_len - sizeof cmd, out_len - sizeof resp);

	ucontext = ibdev->alloc_ucontext(ibdev, &udata);
	if (IS_ERR(ucontext)) {
		ret = PTR_ERR(file->ucontext);
		goto err;
	}

	ucontext->device = ibdev;
	INIT_LIST_HEAD(&ucontext->pd_list);
	INIT_LIST_HEAD(&ucontext->mr_list);
	INIT_LIST_HEAD(&ucontext->mw_list);
	INIT_LIST_HEAD(&ucontext->cq_list);
	INIT_LIST_HEAD(&ucontext->qp_list);
	INIT_LIST_HEAD(&ucontext->srq_list);
	INIT_LIST_HEAD(&ucontext->ah_list);
	ucontext->closing = 0;

	resp.num_comp_vectors = file->device->num_comp_vectors;

	filp = ib_uverbs_alloc_event_file(file, 1, &resp.async_fd);
	if (IS_ERR(filp)) {
		ret = PTR_ERR(filp);
		goto err_free;
	}

	if (copy_to_user((void __user *) (unsigned long) cmd.response,
			 &resp, sizeof resp)) {
		ret = -EFAULT;
		goto err_file;
	}

	file->async_file = filp->private_data;

	INIT_IB_EVENT_HANDLER(&file->event_handler, file->device->ib_dev,
			      ib_uverbs_event_handler);
	ret = ib_register_event_handler(&file->event_handler);
	if (ret)
		goto err_file;

	kref_get(&file->async_file->ref);
	kref_get(&file->ref);
	file->ucontext = ucontext;

	fd_install(resp.async_fd, filp);

	mutex_unlock(&file->mutex);

	return in_len;

err_file:
	put_unused_fd(resp.async_fd);
	fput(filp);

err_free:
	ibdev->dealloc_ucontext(ucontext);

err:
	mutex_unlock(&file->mutex);
	return ret;
}

ssize_t ib_uverbs_query_device(struct ib_uverbs_file *file,
			       const char __user *buf,
			       int in_len, int out_len)
{
	struct ib_uverbs_query_device      cmd;
	struct ib_uverbs_query_device_resp resp;
	struct ib_device_attr              attr;
	int                                ret;

	if (out_len < sizeof resp)
		return -ENOSPC;

	if (copy_from_user(&cmd, buf, sizeof cmd))
		return -EFAULT;

	ret = ib_query_device(file->device->ib_dev, &attr);
	if (ret)
		return ret;

	memset(&resp, 0, sizeof resp);

	resp.fw_ver 		       = attr.fw_ver;
	resp.node_guid 		       = file->device->ib_dev->node_guid;
	resp.sys_image_guid 	       = attr.sys_image_guid;
	resp.max_mr_size 	       = attr.max_mr_size;
	resp.page_size_cap 	       = attr.page_size_cap;
	resp.vendor_id 		       = attr.vendor_id;
	resp.vendor_part_id 	       = attr.vendor_part_id;
	resp.hw_ver 		       = attr.hw_ver;
	resp.max_qp 		       = attr.max_qp;
	resp.max_qp_wr 		       = attr.max_qp_wr;
	resp.device_cap_flags 	       = attr.device_cap_flags;
	resp.max_sge 		       = attr.max_sge;
	resp.max_sge_rd 	       = attr.max_sge_rd;
	resp.max_cq 		       = attr.max_cq;
	resp.max_cqe 		       = attr.max_cqe;
	resp.max_mr 		       = attr.max_mr;
	resp.max_pd 		       = attr.max_pd;
	resp.max_qp_rd_atom 	       = attr.max_qp_rd_atom;
	resp.max_ee_rd_atom 	       = attr.max_ee_rd_atom;
	resp.max_res_rd_atom 	       = attr.max_res_rd_atom;
	resp.max_qp_init_rd_atom       = attr.max_qp_init_rd_atom;
	resp.max_ee_init_rd_atom       = attr.max_ee_init_rd_atom;
	resp.atomic_cap 	       = attr.atomic_cap;
	resp.max_ee 		       = attr.max_ee;
	resp.max_rdd 		       = attr.max_rdd;
	resp.max_mw 		       = attr.max_mw;
	resp.max_raw_ipv6_qp 	       = attr.max_raw_ipv6_qp;
	resp.max_raw_ethy_qp 	       = attr.max_raw_ethy_qp;
	resp.max_mcast_grp 	       = attr.max_mcast_grp;
	resp.max_mcast_qp_attach       = attr.max_mcast_qp_attach;
	resp.max_total_mcast_qp_attach = attr.max_total_mcast_qp_attach;
	resp.max_ah 		       = attr.max_ah;
	resp.max_fmr 		       = attr.max_fmr;
	resp.max_map_per_fmr 	       = attr.max_map_per_fmr;
	resp.max_srq 		       = attr.max_srq;
	resp.max_srq_wr 	       = attr.max_srq_wr;
	resp.max_srq_sge 	       = attr.max_srq_sge;
	resp.max_pkeys 		       = attr.max_pkeys;
	resp.local_ca_ack_delay        = attr.local_ca_ack_delay;
	resp.phys_port_cnt	       = file->device->ib_dev->phys_port_cnt;

	if (copy_to_user((void __user *) (unsigned long) cmd.response,
			 &resp, sizeof resp))
		return -EFAULT;

	return in_len;
}

ssize_t ib_uverbs_query_port(struct ib_uverbs_file *file,
			     const char __user *buf,
			     int in_len, int out_len)
{
	struct ib_uverbs_query_port      cmd;
	struct ib_uverbs_query_port_resp resp;
	struct ib_port_attr              attr;
	int                              ret;

	if (out_len < sizeof resp)
		return -ENOSPC;

	if (copy_from_user(&cmd, buf, sizeof cmd))
		return -EFAULT;

	ret = ib_query_port(file->device->ib_dev, cmd.port_num, &attr);
	if (ret)
		return ret;

	memset(&resp, 0, sizeof resp);

	resp.state 	     = attr.state;
	resp.max_mtu 	     = attr.max_mtu;
	resp.active_mtu      = attr.active_mtu;
	resp.gid_tbl_len     = attr.gid_tbl_len;
	resp.port_cap_flags  = attr.port_cap_flags;
	resp.max_msg_sz      = attr.max_msg_sz;
	resp.bad_pkey_cntr   = attr.bad_pkey_cntr;
	resp.qkey_viol_cntr  = attr.qkey_viol_cntr;
	resp.pkey_tbl_len    = attr.pkey_tbl_len;
	resp.lid 	     = attr.lid;
	resp.sm_lid 	     = attr.sm_lid;
	resp.lmc 	     = attr.lmc;
	resp.max_vl_num      = attr.max_vl_num;
	resp.sm_sl 	     = attr.sm_sl;
	resp.subnet_timeout  = attr.subnet_timeout;
	resp.init_type_reply = attr.init_type_reply;
	resp.active_width    = attr.active_width;
	resp.active_speed    = attr.active_speed;
	resp.phys_state      = attr.phys_state;

	if (copy_to_user((void __user *) (unsigned long) cmd.response,
			 &resp, sizeof resp))
		return -EFAULT;

	return in_len;
}

ssize_t ib_uverbs_alloc_pd(struct ib_uverbs_file *file,
			   const char __user *buf,
			   int in_len, int out_len)
{
	struct ib_uverbs_alloc_pd      cmd;
	struct ib_uverbs_alloc_pd_resp resp;
	struct ib_udata                udata;
	struct ib_uobject             *uobj;
	struct ib_pd                  *pd;
	int                            ret;

	if (out_len < sizeof resp)
		return -ENOSPC;

	if (copy_from_user(&cmd, buf, sizeof cmd))
		return -EFAULT;

	INIT_UDATA(&udata, buf + sizeof cmd,
		   (unsigned long) cmd.response + sizeof resp,
		   in_len - sizeof cmd, out_len - sizeof resp);

	uobj = kmalloc(sizeof *uobj, GFP_KERNEL);
	if (!uobj)
		return -ENOMEM;

	init_uobj(uobj, 0, file->ucontext, &pd_lock_key);
	down_write(&uobj->mutex);

	pd = file->device->ib_dev->alloc_pd(file->device->ib_dev,
					    file->ucontext, &udata);
	if (IS_ERR(pd)) {
		ret = PTR_ERR(pd);
		goto err;
	}

	pd->device  = file->device->ib_dev;
	pd->uobject = uobj;
	atomic_set(&pd->usecnt, 0);

	uobj->object = pd;
	ret = idr_add_uobj(&ib_uverbs_pd_idr, uobj);
	if (ret)
		goto err_idr;

	memset(&resp, 0, sizeof resp);
	resp.pd_handle = uobj->id;

	if (copy_to_user((void __user *) (unsigned long) cmd.response,
			 &resp, sizeof resp)) {
		ret = -EFAULT;
		goto err_copy;
	}

	mutex_lock(&file->mutex);
	list_add_tail(&uobj->list, &file->ucontext->pd_list);
	mutex_unlock(&file->mutex);

	uobj->live = 1;

	up_write(&uobj->mutex);

	return in_len;

err_copy:
	idr_remove_uobj(&ib_uverbs_pd_idr, uobj);

err_idr:
	ib_dealloc_pd(pd);

err:
	put_uobj_write(uobj);
	return ret;
}

ssize_t ib_uverbs_dealloc_pd(struct ib_uverbs_file *file,
			     const char __user *buf,