2005-07-08 08:57:13 +08:00
|
|
|
/*
|
|
|
|
* Copyright (c) 2005 Topspin Communications. All rights reserved.
|
2006-01-31 06:29:21 +08:00
|
|
|
* Copyright (c) 2005, 2006 Cisco Systems. All rights reserved.
|
2005-08-11 14:03:10 +08:00
|
|
|
* Copyright (c) 2005 Mellanox Technologies. All rights reserved.
|
|
|
|
* Copyright (c) 2005 Voltaire, Inc. All rights reserved.
|
2005-10-15 06:26:04 +08:00
|
|
|
* Copyright (c) 2005 PathScale, Inc. All rights reserved.
|
2005-07-08 08:57:13 +08:00
|
|
|
*
|
|
|
|
* 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.
|
|
|
|
*/
|
|
|
|
|
|
|
|
#include <linux/module.h>
|
|
|
|
#include <linux/init.h>
|
|
|
|
#include <linux/device.h>
|
|
|
|
#include <linux/err.h>
|
|
|
|
#include <linux/fs.h>
|
|
|
|
#include <linux/poll.h>
|
2009-10-04 20:11:37 +08:00
|
|
|
#include <linux/sched.h>
|
2005-07-08 08:57:13 +08:00
|
|
|
#include <linux/file.h>
|
2005-10-29 06:38:26 +08:00
|
|
|
#include <linux/cdev.h>
|
2010-02-25 08:51:20 +08:00
|
|
|
#include <linux/anon_inodes.h>
|
2005-07-08 08:57:13 +08:00
|
|
|
|
|
|
|
#include <asm/uaccess.h>
|
|
|
|
|
|
|
|
#include "uverbs.h"
|
|
|
|
|
|
|
|
MODULE_AUTHOR("Roland Dreier");
|
|
|
|
MODULE_DESCRIPTION("InfiniBand userspace verbs access");
|
|
|
|
MODULE_LICENSE("Dual BSD/GPL");
|
|
|
|
|
|
|
|
enum {
|
|
|
|
IB_UVERBS_MAJOR = 231,
|
|
|
|
IB_UVERBS_BASE_MINOR = 192,
|
|
|
|
IB_UVERBS_MAX_DEVICES = 32
|
|
|
|
};
|
|
|
|
|
|
|
|
#define IB_UVERBS_BASE_DEV MKDEV(IB_UVERBS_MAJOR, IB_UVERBS_BASE_MINOR)
|
|
|
|
|
2005-10-29 06:38:26 +08:00
|
|
|
static struct class *uverbs_class;
|
|
|
|
|
IB/uverbs: Don't serialize with ib_uverbs_idr_mutex
Currently, all userspace verbs operations that call into the kernel
are serialized by ib_uverbs_idr_mutex. This can be a scalability
issue for some workloads, especially for devices driven by the ipath
driver, which needs to call into the kernel even for datapath
operations.
Fix this by adding reference counts to the userspace objects, and then
converting ib_uverbs_idr_mutex into a spinlock that only protects the
idrs long enough to take a reference on the object being looked up.
Because remove operations may fail, we have to do a slightly funky
two-step deletion, which is described in the comments at the top of
uverbs_cmd.c.
This also still leaves ib_uverbs_idr_lock as a single lock that is
possibly subject to contention. However, the lock hold time will only
be a single idr operation, so multiple threads should still be able to
make progress, even if ib_uverbs_idr_lock is being ping-ponged.
Surprisingly, these changes even shrink the object code:
add/remove: 23/5 grow/shrink: 4/21 up/down: 633/-693 (-60)
Signed-off-by: Roland Dreier <rolandd@cisco.com>
2006-06-18 11:44:49 +08:00
|
|
|
DEFINE_SPINLOCK(ib_uverbs_idr_lock);
|
2005-07-08 08:57:13 +08:00
|
|
|
DEFINE_IDR(ib_uverbs_pd_idr);
|
|
|
|
DEFINE_IDR(ib_uverbs_mr_idr);
|
|
|
|
DEFINE_IDR(ib_uverbs_mw_idr);
|
|
|
|
DEFINE_IDR(ib_uverbs_ah_idr);
|
|
|
|
DEFINE_IDR(ib_uverbs_cq_idr);
|
|
|
|
DEFINE_IDR(ib_uverbs_qp_idr);
|
2005-08-19 03:24:13 +08:00
|
|
|
DEFINE_IDR(ib_uverbs_srq_idr);
|
2005-07-08 08:57:13 +08:00
|
|
|
|
2009-09-06 11:24:23 +08:00
|
|
|
static DEFINE_SPINLOCK(map_lock);
|
2005-07-08 08:57:13 +08:00
|
|
|
static DECLARE_BITMAP(dev_map, IB_UVERBS_MAX_DEVICES);
|
|
|
|
|
|
|
|
static ssize_t (*uverbs_cmd_table[])(struct ib_uverbs_file *file,
|
|
|
|
const char __user *buf, int in_len,
|
|
|
|
int out_len) = {
|
2010-02-03 03:08:19 +08:00
|
|
|
[IB_USER_VERBS_CMD_GET_CONTEXT] = ib_uverbs_get_context,
|
|
|
|
[IB_USER_VERBS_CMD_QUERY_DEVICE] = ib_uverbs_query_device,
|
|
|
|
[IB_USER_VERBS_CMD_QUERY_PORT] = ib_uverbs_query_port,
|
|
|
|
[IB_USER_VERBS_CMD_ALLOC_PD] = ib_uverbs_alloc_pd,
|
|
|
|
[IB_USER_VERBS_CMD_DEALLOC_PD] = ib_uverbs_dealloc_pd,
|
|
|
|
[IB_USER_VERBS_CMD_REG_MR] = ib_uverbs_reg_mr,
|
|
|
|
[IB_USER_VERBS_CMD_DEREG_MR] = ib_uverbs_dereg_mr,
|
2005-09-27 04:53:25 +08:00
|
|
|
[IB_USER_VERBS_CMD_CREATE_COMP_CHANNEL] = ib_uverbs_create_comp_channel,
|
2010-02-03 03:08:19 +08:00
|
|
|
[IB_USER_VERBS_CMD_CREATE_CQ] = ib_uverbs_create_cq,
|
|
|
|
[IB_USER_VERBS_CMD_RESIZE_CQ] = ib_uverbs_resize_cq,
|
|
|
|
[IB_USER_VERBS_CMD_POLL_CQ] = ib_uverbs_poll_cq,
|
|
|
|
[IB_USER_VERBS_CMD_REQ_NOTIFY_CQ] = ib_uverbs_req_notify_cq,
|
|
|
|
[IB_USER_VERBS_CMD_DESTROY_CQ] = ib_uverbs_destroy_cq,
|
|
|
|
[IB_USER_VERBS_CMD_CREATE_QP] = ib_uverbs_create_qp,
|
|
|
|
[IB_USER_VERBS_CMD_QUERY_QP] = ib_uverbs_query_qp,
|
|
|
|
[IB_USER_VERBS_CMD_MODIFY_QP] = ib_uverbs_modify_qp,
|
|
|
|
[IB_USER_VERBS_CMD_DESTROY_QP] = ib_uverbs_destroy_qp,
|
|
|
|
[IB_USER_VERBS_CMD_POST_SEND] = ib_uverbs_post_send,
|
|
|
|
[IB_USER_VERBS_CMD_POST_RECV] = ib_uverbs_post_recv,
|
|
|
|
[IB_USER_VERBS_CMD_POST_SRQ_RECV] = ib_uverbs_post_srq_recv,
|
|
|
|
[IB_USER_VERBS_CMD_CREATE_AH] = ib_uverbs_create_ah,
|
|
|
|
[IB_USER_VERBS_CMD_DESTROY_AH] = ib_uverbs_destroy_ah,
|
|
|
|
[IB_USER_VERBS_CMD_ATTACH_MCAST] = ib_uverbs_attach_mcast,
|
|
|
|
[IB_USER_VERBS_CMD_DETACH_MCAST] = ib_uverbs_detach_mcast,
|
|
|
|
[IB_USER_VERBS_CMD_CREATE_SRQ] = ib_uverbs_create_srq,
|
|
|
|
[IB_USER_VERBS_CMD_MODIFY_SRQ] = ib_uverbs_modify_srq,
|
|
|
|
[IB_USER_VERBS_CMD_QUERY_SRQ] = ib_uverbs_query_srq,
|
|
|
|
[IB_USER_VERBS_CMD_DESTROY_SRQ] = ib_uverbs_destroy_srq,
|
2005-07-08 08:57:13 +08:00
|
|
|
};
|
|
|
|
|
|
|
|
static void ib_uverbs_add_one(struct ib_device *device);
|
|
|
|
static void ib_uverbs_remove_one(struct ib_device *device);
|
|
|
|
|
2005-10-29 06:38:26 +08:00
|
|
|
static void ib_uverbs_release_dev(struct kref *ref)
|
|
|
|
{
|
|
|
|
struct ib_uverbs_device *dev =
|
|
|
|
container_of(ref, struct ib_uverbs_device, ref);
|
|
|
|
|
2006-08-04 01:56:42 +08:00
|
|
|
complete(&dev->comp);
|
2005-10-29 06:38:26 +08:00
|
|
|
}
|
|
|
|
|
2007-10-10 10:59:15 +08:00
|
|
|
static void ib_uverbs_release_event_file(struct kref *ref)
|
|
|
|
{
|
|
|
|
struct ib_uverbs_event_file *file =
|
|
|
|
container_of(ref, struct ib_uverbs_event_file, ref);
|
|
|
|
|
|
|
|
kfree(file);
|
|
|
|
}
|
|
|
|
|
2005-10-29 06:38:26 +08:00
|
|
|
void ib_uverbs_release_ucq(struct ib_uverbs_file *file,
|
|
|
|
struct ib_uverbs_event_file *ev_file,
|
|
|
|
struct ib_ucq_object *uobj)
|
|
|
|
{
|
|
|
|
struct ib_uverbs_event *evt, *tmp;
|
|
|
|
|
|
|
|
if (ev_file) {
|
|
|
|
spin_lock_irq(&ev_file->lock);
|
|
|
|
list_for_each_entry_safe(evt, tmp, &uobj->comp_list, obj_list) {
|
|
|
|
list_del(&evt->list);
|
|
|
|
kfree(evt);
|
|
|
|
}
|
|
|
|
spin_unlock_irq(&ev_file->lock);
|
|
|
|
|
|
|
|
kref_put(&ev_file->ref, ib_uverbs_release_event_file);
|
|
|
|
}
|
|
|
|
|
|
|
|
spin_lock_irq(&file->async_file->lock);
|
|
|
|
list_for_each_entry_safe(evt, tmp, &uobj->async_list, obj_list) {
|
|
|
|
list_del(&evt->list);
|
|
|
|
kfree(evt);
|
|
|
|
}
|
|
|
|
spin_unlock_irq(&file->async_file->lock);
|
|
|
|
}
|
|
|
|
|
|
|
|
void ib_uverbs_release_uevent(struct ib_uverbs_file *file,
|
|
|
|
struct ib_uevent_object *uobj)
|
|
|
|
{
|
|
|
|
struct ib_uverbs_event *evt, *tmp;
|
|
|
|
|
|
|
|
spin_lock_irq(&file->async_file->lock);
|
|
|
|
list_for_each_entry_safe(evt, tmp, &uobj->event_list, obj_list) {
|
|
|
|
list_del(&evt->list);
|
|
|
|
kfree(evt);
|
|
|
|
}
|
|
|
|
spin_unlock_irq(&file->async_file->lock);
|
|
|
|
}
|
|
|
|
|
2005-11-30 08:57:01 +08:00
|
|
|
static void ib_uverbs_detach_umcast(struct ib_qp *qp,
|
|
|
|
struct ib_uqp_object *uobj)
|
|
|
|
{
|
|
|
|
struct ib_uverbs_mcast_entry *mcast, *tmp;
|
|
|
|
|
|
|
|
list_for_each_entry_safe(mcast, tmp, &uobj->mcast_list, list) {
|
|
|
|
ib_detach_mcast(qp, &mcast->gid, mcast->lid);
|
|
|
|
list_del(&mcast->list);
|
|
|
|
kfree(mcast);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2005-10-29 06:38:26 +08:00
|
|
|
static int ib_uverbs_cleanup_ucontext(struct ib_uverbs_file *file,
|
|
|
|
struct ib_ucontext *context)
|
2005-07-08 08:57:13 +08:00
|
|
|
{
|
|
|
|
struct ib_uobject *uobj, *tmp;
|
|
|
|
|
|
|
|
if (!context)
|
|
|
|
return 0;
|
|
|
|
|
2007-03-05 08:15:11 +08:00
|
|
|
context->closing = 1;
|
|
|
|
|
2005-10-15 06:26:04 +08:00
|
|
|
list_for_each_entry_safe(uobj, tmp, &context->ah_list, list) {
|
IB/uverbs: Don't serialize with ib_uverbs_idr_mutex
Currently, all userspace verbs operations that call into the kernel
are serialized by ib_uverbs_idr_mutex. This can be a scalability
issue for some workloads, especially for devices driven by the ipath
driver, which needs to call into the kernel even for datapath
operations.
Fix this by adding reference counts to the userspace objects, and then
converting ib_uverbs_idr_mutex into a spinlock that only protects the
idrs long enough to take a reference on the object being looked up.
Because remove operations may fail, we have to do a slightly funky
two-step deletion, which is described in the comments at the top of
uverbs_cmd.c.
This also still leaves ib_uverbs_idr_lock as a single lock that is
possibly subject to contention. However, the lock hold time will only
be a single idr operation, so multiple threads should still be able to
make progress, even if ib_uverbs_idr_lock is being ping-ponged.
Surprisingly, these changes even shrink the object code:
add/remove: 23/5 grow/shrink: 4/21 up/down: 633/-693 (-60)
Signed-off-by: Roland Dreier <rolandd@cisco.com>
2006-06-18 11:44:49 +08:00
|
|
|
struct ib_ah *ah = uobj->object;
|
|
|
|
|
|
|
|
idr_remove_uobj(&ib_uverbs_ah_idr, uobj);
|
2005-10-15 06:26:04 +08:00
|
|
|
ib_destroy_ah(ah);
|
|
|
|
kfree(uobj);
|
|
|
|
}
|
2005-07-08 08:57:13 +08:00
|
|
|
|
|
|
|
list_for_each_entry_safe(uobj, tmp, &context->qp_list, list) {
|
IB/uverbs: Don't serialize with ib_uverbs_idr_mutex
Currently, all userspace verbs operations that call into the kernel
are serialized by ib_uverbs_idr_mutex. This can be a scalability
issue for some workloads, especially for devices driven by the ipath
driver, which needs to call into the kernel even for datapath
operations.
Fix this by adding reference counts to the userspace objects, and then
converting ib_uverbs_idr_mutex into a spinlock that only protects the
idrs long enough to take a reference on the object being looked up.
Because remove operations may fail, we have to do a slightly funky
two-step deletion, which is described in the comments at the top of
uverbs_cmd.c.
This also still leaves ib_uverbs_idr_lock as a single lock that is
possibly subject to contention. However, the lock hold time will only
be a single idr operation, so multiple threads should still be able to
make progress, even if ib_uverbs_idr_lock is being ping-ponged.
Surprisingly, these changes even shrink the object code:
add/remove: 23/5 grow/shrink: 4/21 up/down: 633/-693 (-60)
Signed-off-by: Roland Dreier <rolandd@cisco.com>
2006-06-18 11:44:49 +08:00
|
|
|
struct ib_qp *qp = uobj->object;
|
2005-11-30 08:57:01 +08:00
|
|
|
struct ib_uqp_object *uqp =
|
|
|
|
container_of(uobj, struct ib_uqp_object, uevent.uobject);
|
IB/uverbs: Don't serialize with ib_uverbs_idr_mutex
Currently, all userspace verbs operations that call into the kernel
are serialized by ib_uverbs_idr_mutex. This can be a scalability
issue for some workloads, especially for devices driven by the ipath
driver, which needs to call into the kernel even for datapath
operations.
Fix this by adding reference counts to the userspace objects, and then
converting ib_uverbs_idr_mutex into a spinlock that only protects the
idrs long enough to take a reference on the object being looked up.
Because remove operations may fail, we have to do a slightly funky
two-step deletion, which is described in the comments at the top of
uverbs_cmd.c.
This also still leaves ib_uverbs_idr_lock as a single lock that is
possibly subject to contention. However, the lock hold time will only
be a single idr operation, so multiple threads should still be able to
make progress, even if ib_uverbs_idr_lock is being ping-ponged.
Surprisingly, these changes even shrink the object code:
add/remove: 23/5 grow/shrink: 4/21 up/down: 633/-693 (-60)
Signed-off-by: Roland Dreier <rolandd@cisco.com>
2006-06-18 11:44:49 +08:00
|
|
|
|
|
|
|
idr_remove_uobj(&ib_uverbs_qp_idr, uobj);
|
2005-11-30 08:57:01 +08:00
|
|
|
ib_uverbs_detach_umcast(qp, uqp);
|
2005-07-08 08:57:13 +08:00
|
|
|
ib_destroy_qp(qp);
|
2005-11-30 08:57:01 +08:00
|
|
|
ib_uverbs_release_uevent(file, &uqp->uevent);
|
|
|
|
kfree(uqp);
|
2005-07-08 08:57:13 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
list_for_each_entry_safe(uobj, tmp, &context->cq_list, list) {
|
IB/uverbs: Don't serialize with ib_uverbs_idr_mutex
Currently, all userspace verbs operations that call into the kernel
are serialized by ib_uverbs_idr_mutex. This can be a scalability
issue for some workloads, especially for devices driven by the ipath
driver, which needs to call into the kernel even for datapath
operations.
Fix this by adding reference counts to the userspace objects, and then
converting ib_uverbs_idr_mutex into a spinlock that only protects the
idrs long enough to take a reference on the object being looked up.
Because remove operations may fail, we have to do a slightly funky
two-step deletion, which is described in the comments at the top of
uverbs_cmd.c.
This also still leaves ib_uverbs_idr_lock as a single lock that is
possibly subject to contention. However, the lock hold time will only
be a single idr operation, so multiple threads should still be able to
make progress, even if ib_uverbs_idr_lock is being ping-ponged.
Surprisingly, these changes even shrink the object code:
add/remove: 23/5 grow/shrink: 4/21 up/down: 633/-693 (-60)
Signed-off-by: Roland Dreier <rolandd@cisco.com>
2006-06-18 11:44:49 +08:00
|
|
|
struct ib_cq *cq = uobj->object;
|
2005-10-29 06:38:26 +08:00
|
|
|
struct ib_uverbs_event_file *ev_file = cq->cq_context;
|
|
|
|
struct ib_ucq_object *ucq =
|
|
|
|
container_of(uobj, struct ib_ucq_object, uobject);
|
IB/uverbs: Don't serialize with ib_uverbs_idr_mutex
Currently, all userspace verbs operations that call into the kernel
are serialized by ib_uverbs_idr_mutex. This can be a scalability
issue for some workloads, especially for devices driven by the ipath
driver, which needs to call into the kernel even for datapath
operations.
Fix this by adding reference counts to the userspace objects, and then
converting ib_uverbs_idr_mutex into a spinlock that only protects the
idrs long enough to take a reference on the object being looked up.
Because remove operations may fail, we have to do a slightly funky
two-step deletion, which is described in the comments at the top of
uverbs_cmd.c.
This also still leaves ib_uverbs_idr_lock as a single lock that is
possibly subject to contention. However, the lock hold time will only
be a single idr operation, so multiple threads should still be able to
make progress, even if ib_uverbs_idr_lock is being ping-ponged.
Surprisingly, these changes even shrink the object code:
add/remove: 23/5 grow/shrink: 4/21 up/down: 633/-693 (-60)
Signed-off-by: Roland Dreier <rolandd@cisco.com>
2006-06-18 11:44:49 +08:00
|
|
|
|
|
|
|
idr_remove_uobj(&ib_uverbs_cq_idr, uobj);
|
2005-07-08 08:57:13 +08:00
|
|
|
ib_destroy_cq(cq);
|
2005-10-29 06:38:26 +08:00
|
|
|
ib_uverbs_release_ucq(file, ev_file, ucq);
|
|
|
|
kfree(ucq);
|
2005-07-08 08:57:13 +08:00
|
|
|
}
|
|
|
|
|
2005-08-19 03:24:13 +08:00
|
|
|
list_for_each_entry_safe(uobj, tmp, &context->srq_list, list) {
|
IB/uverbs: Don't serialize with ib_uverbs_idr_mutex
Currently, all userspace verbs operations that call into the kernel
are serialized by ib_uverbs_idr_mutex. This can be a scalability
issue for some workloads, especially for devices driven by the ipath
driver, which needs to call into the kernel even for datapath
operations.
Fix this by adding reference counts to the userspace objects, and then
converting ib_uverbs_idr_mutex into a spinlock that only protects the
idrs long enough to take a reference on the object being looked up.
Because remove operations may fail, we have to do a slightly funky
two-step deletion, which is described in the comments at the top of
uverbs_cmd.c.
This also still leaves ib_uverbs_idr_lock as a single lock that is
possibly subject to contention. However, the lock hold time will only
be a single idr operation, so multiple threads should still be able to
make progress, even if ib_uverbs_idr_lock is being ping-ponged.
Surprisingly, these changes even shrink the object code:
add/remove: 23/5 grow/shrink: 4/21 up/down: 633/-693 (-60)
Signed-off-by: Roland Dreier <rolandd@cisco.com>
2006-06-18 11:44:49 +08:00
|
|
|
struct ib_srq *srq = uobj->object;
|
2005-10-29 06:38:26 +08:00
|
|
|
struct ib_uevent_object *uevent =
|
|
|
|
container_of(uobj, struct ib_uevent_object, uobject);
|
IB/uverbs: Don't serialize with ib_uverbs_idr_mutex
Currently, all userspace verbs operations that call into the kernel
are serialized by ib_uverbs_idr_mutex. This can be a scalability
issue for some workloads, especially for devices driven by the ipath
driver, which needs to call into the kernel even for datapath
operations.
Fix this by adding reference counts to the userspace objects, and then
converting ib_uverbs_idr_mutex into a spinlock that only protects the
idrs long enough to take a reference on the object being looked up.
Because remove operations may fail, we have to do a slightly funky
two-step deletion, which is described in the comments at the top of
uverbs_cmd.c.
This also still leaves ib_uverbs_idr_lock as a single lock that is
possibly subject to contention. However, the lock hold time will only
be a single idr operation, so multiple threads should still be able to
make progress, even if ib_uverbs_idr_lock is being ping-ponged.
Surprisingly, these changes even shrink the object code:
add/remove: 23/5 grow/shrink: 4/21 up/down: 633/-693 (-60)
Signed-off-by: Roland Dreier <rolandd@cisco.com>
2006-06-18 11:44:49 +08:00
|
|
|
|
|
|
|
idr_remove_uobj(&ib_uverbs_srq_idr, uobj);
|
2005-08-19 03:24:13 +08:00
|
|
|
ib_destroy_srq(srq);
|
2005-10-29 06:38:26 +08:00
|
|
|
ib_uverbs_release_uevent(file, uevent);
|
|
|
|
kfree(uevent);
|
2005-08-19 03:24:13 +08:00
|
|
|
}
|
|
|
|
|
2005-07-08 08:57:13 +08:00
|
|
|
/* XXX Free MWs */
|
|
|
|
|
|
|
|
list_for_each_entry_safe(uobj, tmp, &context->mr_list, list) {
|
IB/uverbs: Don't serialize with ib_uverbs_idr_mutex
Currently, all userspace verbs operations that call into the kernel
are serialized by ib_uverbs_idr_mutex. This can be a scalability
issue for some workloads, especially for devices driven by the ipath
driver, which needs to call into the kernel even for datapath
operations.
Fix this by adding reference counts to the userspace objects, and then
converting ib_uverbs_idr_mutex into a spinlock that only protects the
idrs long enough to take a reference on the object being looked up.
Because remove operations may fail, we have to do a slightly funky
two-step deletion, which is described in the comments at the top of
uverbs_cmd.c.
This also still leaves ib_uverbs_idr_lock as a single lock that is
possibly subject to contention. However, the lock hold time will only
be a single idr operation, so multiple threads should still be able to
make progress, even if ib_uverbs_idr_lock is being ping-ponged.
Surprisingly, these changes even shrink the object code:
add/remove: 23/5 grow/shrink: 4/21 up/down: 633/-693 (-60)
Signed-off-by: Roland Dreier <rolandd@cisco.com>
2006-06-18 11:44:49 +08:00
|
|
|
struct ib_mr *mr = uobj->object;
|
2005-07-08 08:57:13 +08:00
|
|
|
|
IB/uverbs: Don't serialize with ib_uverbs_idr_mutex
Currently, all userspace verbs operations that call into the kernel
are serialized by ib_uverbs_idr_mutex. This can be a scalability
issue for some workloads, especially for devices driven by the ipath
driver, which needs to call into the kernel even for datapath
operations.
Fix this by adding reference counts to the userspace objects, and then
converting ib_uverbs_idr_mutex into a spinlock that only protects the
idrs long enough to take a reference on the object being looked up.
Because remove operations may fail, we have to do a slightly funky
two-step deletion, which is described in the comments at the top of
uverbs_cmd.c.
This also still leaves ib_uverbs_idr_lock as a single lock that is
possibly subject to contention. However, the lock hold time will only
be a single idr operation, so multiple threads should still be able to
make progress, even if ib_uverbs_idr_lock is being ping-ponged.
Surprisingly, these changes even shrink the object code:
add/remove: 23/5 grow/shrink: 4/21 up/down: 633/-693 (-60)
Signed-off-by: Roland Dreier <rolandd@cisco.com>
2006-06-18 11:44:49 +08:00
|
|
|
idr_remove_uobj(&ib_uverbs_mr_idr, uobj);
|
2005-07-08 08:57:13 +08:00
|
|
|
ib_dereg_mr(mr);
|
2007-03-05 08:15:11 +08:00
|
|
|
kfree(uobj);
|
2005-07-08 08:57:13 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
list_for_each_entry_safe(uobj, tmp, &context->pd_list, list) {
|
IB/uverbs: Don't serialize with ib_uverbs_idr_mutex
Currently, all userspace verbs operations that call into the kernel
are serialized by ib_uverbs_idr_mutex. This can be a scalability
issue for some workloads, especially for devices driven by the ipath
driver, which needs to call into the kernel even for datapath
operations.
Fix this by adding reference counts to the userspace objects, and then
converting ib_uverbs_idr_mutex into a spinlock that only protects the
idrs long enough to take a reference on the object being looked up.
Because remove operations may fail, we have to do a slightly funky
two-step deletion, which is described in the comments at the top of
uverbs_cmd.c.
This also still leaves ib_uverbs_idr_lock as a single lock that is
possibly subject to contention. However, the lock hold time will only
be a single idr operation, so multiple threads should still be able to
make progress, even if ib_uverbs_idr_lock is being ping-ponged.
Surprisingly, these changes even shrink the object code:
add/remove: 23/5 grow/shrink: 4/21 up/down: 633/-693 (-60)
Signed-off-by: Roland Dreier <rolandd@cisco.com>
2006-06-18 11:44:49 +08:00
|
|
|
struct ib_pd *pd = uobj->object;
|
|
|
|
|
|
|
|
idr_remove_uobj(&ib_uverbs_pd_idr, uobj);
|
2005-07-08 08:57:13 +08:00
|
|
|
ib_dealloc_pd(pd);
|
|
|
|
kfree(uobj);
|
|
|
|
}
|
|
|
|
|
|
|
|
return context->device->dealloc_ucontext(context);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void ib_uverbs_release_file(struct kref *ref)
|
|
|
|
{
|
|
|
|
struct ib_uverbs_file *file =
|
|
|
|
container_of(ref, struct ib_uverbs_file, ref);
|
|
|
|
|
|
|
|
module_put(file->device->ib_dev->owner);
|
2005-10-29 06:38:26 +08:00
|
|
|
kref_put(&file->device->ref, ib_uverbs_release_dev);
|
|
|
|
|
2005-07-08 08:57:13 +08:00
|
|
|
kfree(file);
|
|
|
|
}
|
|
|
|
|
|
|
|
static ssize_t ib_uverbs_event_read(struct file *filp, char __user *buf,
|
|
|
|
size_t count, loff_t *pos)
|
|
|
|
{
|
|
|
|
struct ib_uverbs_event_file *file = filp->private_data;
|
2005-09-10 06:55:08 +08:00
|
|
|
struct ib_uverbs_event *event;
|
2005-07-08 08:57:13 +08:00
|
|
|
int eventsz;
|
|
|
|
int ret = 0;
|
|
|
|
|
|
|
|
spin_lock_irq(&file->lock);
|
|
|
|
|
2005-09-27 04:53:25 +08:00
|
|
|
while (list_empty(&file->event_list)) {
|
2005-07-08 08:57:13 +08:00
|
|
|
spin_unlock_irq(&file->lock);
|
|
|
|
|
|
|
|
if (filp->f_flags & O_NONBLOCK)
|
|
|
|
return -EAGAIN;
|
|
|
|
|
|
|
|
if (wait_event_interruptible(file->poll_wait,
|
2005-09-27 04:53:25 +08:00
|
|
|
!list_empty(&file->event_list)))
|
2005-07-08 08:57:13 +08:00
|
|
|
return -ERESTARTSYS;
|
|
|
|
|
|
|
|
spin_lock_irq(&file->lock);
|
|
|
|
}
|
|
|
|
|
2005-09-10 06:55:08 +08:00
|
|
|
event = list_entry(file->event_list.next, struct ib_uverbs_event, list);
|
|
|
|
|
|
|
|
if (file->is_async)
|
2005-07-08 08:57:13 +08:00
|
|
|
eventsz = sizeof (struct ib_uverbs_async_event_desc);
|
2005-09-10 06:55:08 +08:00
|
|
|
else
|
2005-07-08 08:57:13 +08:00
|
|
|
eventsz = sizeof (struct ib_uverbs_comp_event_desc);
|
|
|
|
|
|
|
|
if (eventsz > count) {
|
|
|
|
ret = -EINVAL;
|
|
|
|
event = NULL;
|
2005-09-10 06:55:08 +08:00
|
|
|
} else {
|
2005-07-08 08:57:13 +08:00
|
|
|
list_del(file->event_list.next);
|
2005-09-10 06:55:08 +08:00
|
|
|
if (event->counter) {
|
|
|
|
++(*event->counter);
|
|
|
|
list_del(&event->obj_list);
|
|
|
|
}
|
|
|
|
}
|
2005-07-08 08:57:13 +08:00
|
|
|
|
|
|
|
spin_unlock_irq(&file->lock);
|
|
|
|
|
|
|
|
if (event) {
|
|
|
|
if (copy_to_user(buf, event, eventsz))
|
|
|
|
ret = -EFAULT;
|
|
|
|
else
|
|
|
|
ret = eventsz;
|
|
|
|
}
|
|
|
|
|
|
|
|
kfree(event);
|
|
|
|
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
static unsigned int ib_uverbs_event_poll(struct file *filp,
|
|
|
|
struct poll_table_struct *wait)
|
|
|
|
{
|
|
|
|
unsigned int pollflags = 0;
|
|
|
|
struct ib_uverbs_event_file *file = filp->private_data;
|
|
|
|
|
|
|
|
poll_wait(filp, &file->poll_wait, wait);
|
|
|
|
|
|
|
|
spin_lock_irq(&file->lock);
|
2005-09-27 04:53:25 +08:00
|
|
|
if (!list_empty(&file->event_list))
|
2005-07-08 08:57:13 +08:00
|
|
|
pollflags = POLLIN | POLLRDNORM;
|
|
|
|
spin_unlock_irq(&file->lock);
|
|
|
|
|
|
|
|
return pollflags;
|
|
|
|
}
|
|
|
|
|
2005-07-28 05:40:00 +08:00
|
|
|
static int ib_uverbs_event_fasync(int fd, struct file *filp, int on)
|
|
|
|
{
|
|
|
|
struct ib_uverbs_event_file *file = filp->private_data;
|
|
|
|
|
|
|
|
return fasync_helper(fd, filp, on, &file->async_queue);
|
|
|
|
}
|
|
|
|
|
2005-07-08 08:57:13 +08:00
|
|
|
static int ib_uverbs_event_close(struct inode *inode, struct file *filp)
|
|
|
|
{
|
|
|
|
struct ib_uverbs_event_file *file = filp->private_data;
|
2005-09-27 04:53:25 +08:00
|
|
|
struct ib_uverbs_event *entry, *tmp;
|
|
|
|
|
|
|
|
spin_lock_irq(&file->lock);
|
2008-04-17 12:01:08 +08:00
|
|
|
file->is_closed = 1;
|
2005-09-27 04:53:25 +08:00
|
|
|
list_for_each_entry_safe(entry, tmp, &file->event_list, list) {
|
|
|
|
if (entry->counter)
|
|
|
|
list_del(&entry->obj_list);
|
|
|
|
kfree(entry);
|
|
|
|
}
|
|
|
|
spin_unlock_irq(&file->lock);
|
2005-07-08 08:57:13 +08:00
|
|
|
|
2005-09-27 04:53:25 +08:00
|
|
|
if (file->is_async) {
|
|
|
|
ib_unregister_event_handler(&file->uverbs_file->event_handler);
|
|
|
|
kref_put(&file->uverbs_file->ref, ib_uverbs_release_file);
|
|
|
|
}
|
|
|
|
kref_put(&file->ref, ib_uverbs_release_event_file);
|
2005-07-08 08:57:13 +08:00
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2007-02-12 16:55:32 +08:00
|
|
|
static const struct file_operations uverbs_event_fops = {
|
2005-09-27 04:53:25 +08:00
|
|
|
.owner = THIS_MODULE,
|
2010-02-03 03:08:19 +08:00
|
|
|
.read = ib_uverbs_event_read,
|
2005-07-08 08:57:13 +08:00
|
|
|
.poll = ib_uverbs_event_poll,
|
2005-07-28 05:40:00 +08:00
|
|
|
.release = ib_uverbs_event_close,
|
|
|
|
.fasync = ib_uverbs_event_fasync
|
2005-07-08 08:57:13 +08:00
|
|
|
};
|
|
|
|
|
|
|
|
void ib_uverbs_comp_handler(struct ib_cq *cq, void *cq_context)
|
|
|
|
{
|
2005-09-27 04:53:25 +08:00
|
|
|
struct ib_uverbs_event_file *file = cq_context;
|
|
|
|
struct ib_ucq_object *uobj;
|
|
|
|
struct ib_uverbs_event *entry;
|
|
|
|
unsigned long flags;
|
|
|
|
|
|
|
|
if (!file)
|
|
|
|
return;
|
|
|
|
|
|
|
|
spin_lock_irqsave(&file->lock, flags);
|
2008-04-17 12:01:08 +08:00
|
|
|
if (file->is_closed) {
|
2005-09-27 04:53:25 +08:00
|
|
|
spin_unlock_irqrestore(&file->lock, flags);
|
|
|
|
return;
|
|
|
|
}
|
2005-07-08 08:57:13 +08:00
|
|
|
|
|
|
|
entry = kmalloc(sizeof *entry, GFP_ATOMIC);
|
2005-10-12 06:39:38 +08:00
|
|
|
if (!entry) {
|
|
|
|
spin_unlock_irqrestore(&file->lock, flags);
|
2005-07-08 08:57:13 +08:00
|
|
|
return;
|
2005-10-12 06:39:38 +08:00
|
|
|
}
|
2005-07-08 08:57:13 +08:00
|
|
|
|
2005-09-10 06:55:08 +08:00
|
|
|
uobj = container_of(cq->uobject, struct ib_ucq_object, uobject);
|
|
|
|
|
|
|
|
entry->desc.comp.cq_handle = cq->uobject->user_handle;
|
|
|
|
entry->counter = &uobj->comp_events_reported;
|
2005-07-08 08:57:13 +08:00
|
|
|
|
2005-09-27 04:53:25 +08:00
|
|
|
list_add_tail(&entry->list, &file->event_list);
|
2005-09-10 06:55:08 +08:00
|
|
|
list_add_tail(&entry->obj_list, &uobj->comp_list);
|
2005-09-27 04:53:25 +08:00
|
|
|
spin_unlock_irqrestore(&file->lock, flags);
|
2005-07-08 08:57:13 +08:00
|
|
|
|
2005-09-27 04:53:25 +08:00
|
|
|
wake_up_interruptible(&file->poll_wait);
|
|
|
|
kill_fasync(&file->async_queue, SIGIO, POLL_IN);
|
2005-07-08 08:57:13 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
static void ib_uverbs_async_handler(struct ib_uverbs_file *file,
|
2005-09-10 06:55:08 +08:00
|
|
|
__u64 element, __u64 event,
|
|
|
|
struct list_head *obj_list,
|
|
|
|
u32 *counter)
|
2005-07-08 08:57:13 +08:00
|
|
|
{
|
2005-09-10 06:55:08 +08:00
|
|
|
struct ib_uverbs_event *entry;
|
2005-07-08 08:57:13 +08:00
|
|
|
unsigned long flags;
|
|
|
|
|
2005-09-27 04:53:25 +08:00
|
|
|
spin_lock_irqsave(&file->async_file->lock, flags);
|
2008-06-19 06:36:38 +08:00
|
|
|
if (file->async_file->is_closed) {
|
2005-09-27 04:53:25 +08:00
|
|
|
spin_unlock_irqrestore(&file->async_file->lock, flags);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
2005-07-08 08:57:13 +08:00
|
|
|
entry = kmalloc(sizeof *entry, GFP_ATOMIC);
|
2005-10-12 06:39:38 +08:00
|
|
|
if (!entry) {
|
|
|
|
spin_unlock_irqrestore(&file->async_file->lock, flags);
|
2005-07-08 08:57:13 +08:00
|
|
|
return;
|
2005-10-12 06:39:38 +08:00
|
|
|
}
|
2005-07-08 08:57:13 +08:00
|
|
|
|
2005-09-10 06:55:08 +08:00
|
|
|
entry->desc.async.element = element;
|
|
|
|
entry->desc.async.event_type = event;
|
|
|
|
entry->counter = counter;
|
2005-07-08 08:57:13 +08:00
|
|
|
|
2005-09-27 04:53:25 +08:00
|
|
|
list_add_tail(&entry->list, &file->async_file->event_list);
|
2005-09-10 06:55:08 +08:00
|
|
|
if (obj_list)
|
|
|
|
list_add_tail(&entry->obj_list, obj_list);
|
2005-09-27 04:53:25 +08:00
|
|
|
spin_unlock_irqrestore(&file->async_file->lock, flags);
|
2005-07-08 08:57:13 +08:00
|
|
|
|
2005-09-27 04:53:25 +08:00
|
|
|
wake_up_interruptible(&file->async_file->poll_wait);
|
|
|
|
kill_fasync(&file->async_file->async_queue, SIGIO, POLL_IN);
|
2005-07-08 08:57:13 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
void ib_uverbs_cq_event_handler(struct ib_event *event, void *context_ptr)
|
|
|
|
{
|
2005-10-31 01:50:04 +08:00
|
|
|
struct ib_ucq_object *uobj = container_of(event->element.cq->uobject,
|
|
|
|
struct ib_ucq_object, uobject);
|
2005-09-10 06:55:08 +08:00
|
|
|
|
2005-10-31 01:50:04 +08:00
|
|
|
ib_uverbs_async_handler(uobj->uverbs_file, uobj->uobject.user_handle,
|
2005-09-10 06:55:08 +08:00
|
|
|
event->event, &uobj->async_list,
|
|
|
|
&uobj->async_events_reported);
|
2005-07-08 08:57:13 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
void ib_uverbs_qp_event_handler(struct ib_event *event, void *context_ptr)
|
|
|
|
{
|
2005-09-10 06:55:08 +08:00
|
|
|
struct ib_uevent_object *uobj;
|
|
|
|
|
|
|
|
uobj = container_of(event->element.qp->uobject,
|
|
|
|
struct ib_uevent_object, uobject);
|
|
|
|
|
|
|
|
ib_uverbs_async_handler(context_ptr, uobj->uobject.user_handle,
|
|
|
|
event->event, &uobj->event_list,
|
|
|
|
&uobj->events_reported);
|
2005-07-08 08:57:13 +08:00
|
|
|
}
|
|
|
|
|
2005-08-19 03:24:13 +08:00
|
|
|
void ib_uverbs_srq_event_handler(struct ib_event *event, void *context_ptr)
|
|
|
|
{
|
2005-09-10 06:55:08 +08:00
|
|
|
struct ib_uevent_object *uobj;
|
|
|
|
|
|
|
|
uobj = container_of(event->element.srq->uobject,
|
|
|
|
struct ib_uevent_object, uobject);
|
|
|
|
|
|
|
|
ib_uverbs_async_handler(context_ptr, uobj->uobject.user_handle,
|
|
|
|
event->event, &uobj->event_list,
|
|
|
|
&uobj->events_reported);
|
2005-08-19 03:24:13 +08:00
|
|
|
}
|
|
|
|
|
2005-09-27 04:53:25 +08:00
|
|
|
void ib_uverbs_event_handler(struct ib_event_handler *handler,
|
|
|
|
struct ib_event *event)
|
2005-07-08 08:57:13 +08:00
|
|
|
{
|
|
|
|
struct ib_uverbs_file *file =
|
|
|
|
container_of(handler, struct ib_uverbs_file, event_handler);
|
|
|
|
|
2005-09-10 06:55:08 +08:00
|
|
|
ib_uverbs_async_handler(file, event->element.port_num, event->event,
|
|
|
|
NULL, NULL);
|
2005-07-08 08:57:13 +08:00
|
|
|
}
|
|
|
|
|
2005-09-27 04:53:25 +08:00
|
|
|
struct file *ib_uverbs_alloc_event_file(struct ib_uverbs_file *uverbs_file,
|
2010-01-18 14:38:00 +08:00
|
|
|
int is_async)
|
2005-07-08 08:57:13 +08:00
|
|
|
{
|
2005-09-27 04:53:25 +08:00
|
|
|
struct ib_uverbs_event_file *ev_file;
|
2005-07-08 08:57:13 +08:00
|
|
|
struct file *filp;
|
|
|
|
|
2005-09-27 04:53:25 +08:00
|
|
|
ev_file = kmalloc(sizeof *ev_file, GFP_KERNEL);
|
|
|
|
if (!ev_file)
|
|
|
|
return ERR_PTR(-ENOMEM);
|
|
|
|
|
|
|
|
kref_init(&ev_file->ref);
|
|
|
|
spin_lock_init(&ev_file->lock);
|
|
|
|
INIT_LIST_HEAD(&ev_file->event_list);
|
|
|
|
init_waitqueue_head(&ev_file->poll_wait);
|
|
|
|
ev_file->uverbs_file = uverbs_file;
|
|
|
|
ev_file->async_queue = NULL;
|
|
|
|
ev_file->is_async = is_async;
|
2008-04-17 12:01:08 +08:00
|
|
|
ev_file->is_closed = 0;
|
2005-09-27 04:53:25 +08:00
|
|
|
|
2010-01-18 14:38:00 +08:00
|
|
|
filp = anon_inode_getfile("[infinibandevent]", &uverbs_event_fops,
|
2010-02-25 08:51:20 +08:00
|
|
|
ev_file, O_RDONLY);
|
2010-01-18 14:38:00 +08:00
|
|
|
if (IS_ERR(filp))
|
|
|
|
kfree(ev_file);
|
2008-04-17 12:01:08 +08:00
|
|
|
|
2005-09-27 04:53:25 +08:00
|
|
|
return filp;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Look up a completion event file by FD. If lookup is successful,
|
|
|
|
* takes a ref to the event file struct that it returns; if
|
|
|
|
* unsuccessful, returns NULL.
|
|
|
|
*/
|
|
|
|
struct ib_uverbs_event_file *ib_uverbs_lookup_comp_file(int fd)
|
|
|
|
{
|
|
|
|
struct ib_uverbs_event_file *ev_file = NULL;
|
|
|
|
struct file *filp;
|
|
|
|
|
|
|
|
filp = fget(fd);
|
|
|
|
if (!filp)
|
|
|
|
return NULL;
|
|
|
|
|
|
|
|
if (filp->f_op != &uverbs_event_fops)
|
|
|
|
goto out;
|
|
|
|
|
|
|
|
ev_file = filp->private_data;
|
|
|
|
if (ev_file->is_async) {
|
|
|
|
ev_file = NULL;
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
|
|
|
kref_get(&ev_file->ref);
|
|
|
|
|
|
|
|
out:
|
|
|
|
fput(filp);
|
|
|
|
return ev_file;
|
2005-07-08 08:57:13 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
static ssize_t ib_uverbs_write(struct file *filp, const char __user *buf,
|
|
|
|
size_t count, loff_t *pos)
|
|
|
|
{
|
|
|
|
struct ib_uverbs_file *file = filp->private_data;
|
|
|
|
struct ib_uverbs_cmd_hdr hdr;
|
|
|
|
|
|
|
|
if (count < sizeof hdr)
|
|
|
|
return -EINVAL;
|
|
|
|
|
|
|
|
if (copy_from_user(&hdr, buf, sizeof hdr))
|
|
|
|
return -EFAULT;
|
|
|
|
|
|
|
|
if (hdr.in_words * 4 != count)
|
|
|
|
return -EINVAL;
|
|
|
|
|
2005-09-27 04:01:03 +08:00
|
|
|
if (hdr.command < 0 ||
|
|
|
|
hdr.command >= ARRAY_SIZE(uverbs_cmd_table) ||
|
2009-09-06 11:24:24 +08:00
|
|
|
!uverbs_cmd_table[hdr.command])
|
2005-07-08 08:57:13 +08:00
|
|
|
return -EINVAL;
|
|
|
|
|
2005-09-27 04:53:25 +08:00
|
|
|
if (!file->ucontext &&
|
2005-07-08 08:57:13 +08:00
|
|
|
hdr.command != IB_USER_VERBS_CMD_GET_CONTEXT)
|
|
|
|
return -EINVAL;
|
|
|
|
|
2009-09-06 11:24:24 +08:00
|
|
|
if (!(file->device->ib_dev->uverbs_cmd_mask & (1ull << hdr.command)))
|
|
|
|
return -ENOSYS;
|
|
|
|
|
2005-07-08 08:57:13 +08:00
|
|
|
return uverbs_cmd_table[hdr.command](file, buf + sizeof hdr,
|
|
|
|
hdr.in_words * 4, hdr.out_words * 4);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int ib_uverbs_mmap(struct file *filp, struct vm_area_struct *vma)
|
|
|
|
{
|
|
|
|
struct ib_uverbs_file *file = filp->private_data;
|
|
|
|
|
|
|
|
if (!file->ucontext)
|
|
|
|
return -ENODEV;
|
|
|
|
else
|
|
|
|
return file->device->ib_dev->mmap(file->ucontext, vma);
|
|
|
|
}
|
|
|
|
|
2008-06-28 05:43:20 +08:00
|
|
|
/*
|
|
|
|
* ib_uverbs_open() does not need the BKL:
|
|
|
|
*
|
2010-02-03 03:07:54 +08:00
|
|
|
* - the ib_uverbs_device structures are properly reference counted and
|
2008-06-28 05:43:20 +08:00
|
|
|
* everything else is purely local to the file being created, so
|
|
|
|
* races against other open calls are not a problem;
|
|
|
|
* - there is no ioctl method to race against;
|
2010-02-03 03:07:54 +08:00
|
|
|
* - the open method will either immediately run -ENXIO, or all
|
|
|
|
* required initialization will be done.
|
2008-06-28 05:43:20 +08:00
|
|
|
*/
|
2005-07-08 08:57:13 +08:00
|
|
|
static int ib_uverbs_open(struct inode *inode, struct file *filp)
|
|
|
|
{
|
2005-10-29 06:38:26 +08:00
|
|
|
struct ib_uverbs_device *dev;
|
2005-07-08 08:57:13 +08:00
|
|
|
struct ib_uverbs_file *file;
|
2005-10-29 06:38:26 +08:00
|
|
|
int ret;
|
2005-07-08 08:57:13 +08:00
|
|
|
|
2010-02-03 03:07:54 +08:00
|
|
|
dev = container_of(inode->i_cdev, struct ib_uverbs_device, cdev);
|
2005-10-29 06:38:26 +08:00
|
|
|
if (dev)
|
|
|
|
kref_get(&dev->ref);
|
2010-02-03 03:07:54 +08:00
|
|
|
else
|
2005-10-29 06:38:26 +08:00
|
|
|
return -ENXIO;
|
|
|
|
|
|
|
|
if (!try_module_get(dev->ib_dev->owner)) {
|
|
|
|
ret = -ENODEV;
|
|
|
|
goto err;
|
|
|
|
}
|
2005-07-08 08:57:13 +08:00
|
|
|
|
2005-09-27 04:53:25 +08:00
|
|
|
file = kmalloc(sizeof *file, GFP_KERNEL);
|
2005-09-27 04:01:03 +08:00
|
|
|
if (!file) {
|
2005-10-29 06:38:26 +08:00
|
|
|
ret = -ENOMEM;
|
|
|
|
goto err_module;
|
2005-09-27 04:01:03 +08:00
|
|
|
}
|
2005-07-08 08:57:13 +08:00
|
|
|
|
2005-10-29 06:38:26 +08:00
|
|
|
file->device = dev;
|
|
|
|
file->ucontext = NULL;
|
|
|
|
file->async_file = NULL;
|
2005-07-08 08:57:13 +08:00
|
|
|
kref_init(&file->ref);
|
2006-01-14 06:51:39 +08:00
|
|
|
mutex_init(&file->mutex);
|
2005-07-08 08:57:13 +08:00
|
|
|
|
|
|
|
filp->private_data = file;
|
|
|
|
|
|
|
|
return 0;
|
2005-10-29 06:38:26 +08:00
|
|
|
|
|
|
|
err_module:
|
|
|
|
module_put(dev->ib_dev->owner);
|
|
|
|
|
|
|
|
err:
|
|
|
|
kref_put(&dev->ref, ib_uverbs_release_dev);
|
|
|
|
return ret;
|
2005-07-08 08:57:13 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
static int ib_uverbs_close(struct inode *inode, struct file *filp)
|
|
|
|
{
|
|
|
|
struct ib_uverbs_file *file = filp->private_data;
|
|
|
|
|
2005-10-29 06:38:26 +08:00
|
|
|
ib_uverbs_cleanup_ucontext(file, file->ucontext);
|
|
|
|
|
|
|
|
if (file->async_file)
|
|
|
|
kref_put(&file->async_file->ref, ib_uverbs_release_event_file);
|
2005-07-08 08:57:13 +08:00
|
|
|
|
|
|
|
kref_put(&file->ref, ib_uverbs_release_file);
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2007-02-12 16:55:32 +08:00
|
|
|
static const struct file_operations uverbs_fops = {
|
2010-02-03 03:08:19 +08:00
|
|
|
.owner = THIS_MODULE,
|
|
|
|
.write = ib_uverbs_write,
|
|
|
|
.open = ib_uverbs_open,
|
2005-07-08 08:57:13 +08:00
|
|
|
.release = ib_uverbs_close
|
|
|
|
};
|
|
|
|
|
2007-02-12 16:55:32 +08:00
|
|
|
static const struct file_operations uverbs_mmap_fops = {
|
2010-02-03 03:08:19 +08:00
|
|
|
.owner = THIS_MODULE,
|
|
|
|
.write = ib_uverbs_write,
|
2005-07-08 08:57:13 +08:00
|
|
|
.mmap = ib_uverbs_mmap,
|
2010-02-03 03:08:19 +08:00
|
|
|
.open = ib_uverbs_open,
|
2005-07-08 08:57:13 +08:00
|
|
|
.release = ib_uverbs_close
|
|
|
|
};
|
|
|
|
|
|
|
|
static struct ib_client uverbs_client = {
|
|
|
|
.name = "uverbs",
|
|
|
|
.add = ib_uverbs_add_one,
|
|
|
|
.remove = ib_uverbs_remove_one
|
|
|
|
};
|
|
|
|
|
2008-02-22 07:13:36 +08:00
|
|
|
static ssize_t show_ibdev(struct device *device, struct device_attribute *attr,
|
|
|
|
char *buf)
|
2005-07-08 08:57:13 +08:00
|
|
|
{
|
2008-02-22 07:13:36 +08:00
|
|
|
struct ib_uverbs_device *dev = dev_get_drvdata(device);
|
2005-10-29 06:38:26 +08:00
|
|
|
|
|
|
|
if (!dev)
|
|
|
|
return -ENODEV;
|
2005-07-08 08:57:13 +08:00
|
|
|
|
|
|
|
return sprintf(buf, "%s\n", dev->ib_dev->name);
|
|
|
|
}
|
2008-02-22 07:13:36 +08:00
|
|
|
static DEVICE_ATTR(ibdev, S_IRUGO, show_ibdev, NULL);
|
2005-07-08 08:57:13 +08:00
|
|
|
|
2008-02-22 07:13:36 +08:00
|
|
|
static ssize_t show_dev_abi_version(struct device *device,
|
|
|
|
struct device_attribute *attr, char *buf)
|
2005-09-30 05:17:48 +08:00
|
|
|
{
|
2008-02-22 07:13:36 +08:00
|
|
|
struct ib_uverbs_device *dev = dev_get_drvdata(device);
|
2005-10-29 06:38:26 +08:00
|
|
|
|
|
|
|
if (!dev)
|
|
|
|
return -ENODEV;
|
2005-09-30 05:17:48 +08:00
|
|
|
|
|
|
|
return sprintf(buf, "%d\n", dev->ib_dev->uverbs_abi_ver);
|
|
|
|
}
|
2008-02-22 07:13:36 +08:00
|
|
|
static DEVICE_ATTR(abi_version, S_IRUGO, show_dev_abi_version, NULL);
|
2005-09-30 05:17:48 +08:00
|
|
|
|
2010-01-05 19:48:09 +08:00
|
|
|
static CLASS_ATTR_STRING(abi_version, S_IRUGO,
|
|
|
|
__stringify(IB_USER_VERBS_ABI_VERSION));
|
2005-07-08 08:57:13 +08:00
|
|
|
|
2010-02-03 03:08:09 +08:00
|
|
|
static dev_t overflow_maj;
|
|
|
|
static DECLARE_BITMAP(overflow_map, IB_UVERBS_MAX_DEVICES);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* If we have more than IB_UVERBS_MAX_DEVICES, dynamically overflow by
|
|
|
|
* requesting a new major number and doubling the number of max devices we
|
|
|
|
* support. It's stupid, but simple.
|
|
|
|
*/
|
|
|
|
static int find_overflow_devnum(void)
|
|
|
|
{
|
|
|
|
int ret;
|
|
|
|
|
|
|
|
if (!overflow_maj) {
|
|
|
|
ret = alloc_chrdev_region(&overflow_maj, 0, IB_UVERBS_MAX_DEVICES,
|
|
|
|
"infiniband_verbs");
|
|
|
|
if (ret) {
|
|
|
|
printk(KERN_ERR "user_verbs: couldn't register dynamic device number\n");
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
ret = find_first_zero_bit(overflow_map, IB_UVERBS_MAX_DEVICES);
|
|
|
|
if (ret >= IB_UVERBS_MAX_DEVICES)
|
|
|
|
return -1;
|
|
|
|
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2005-07-08 08:57:13 +08:00
|
|
|
static void ib_uverbs_add_one(struct ib_device *device)
|
|
|
|
{
|
2010-02-03 03:07:59 +08:00
|
|
|
int devnum;
|
2010-02-03 03:08:04 +08:00
|
|
|
dev_t base;
|
2005-07-08 08:57:13 +08:00
|
|
|
struct ib_uverbs_device *uverbs_dev;
|
|
|
|
|
|
|
|
if (!device->alloc_ucontext)
|
|
|
|
return;
|
|
|
|
|
2005-11-02 23:23:14 +08:00
|
|
|
uverbs_dev = kzalloc(sizeof *uverbs_dev, GFP_KERNEL);
|
2005-07-08 08:57:13 +08:00
|
|
|
if (!uverbs_dev)
|
|
|
|
return;
|
|
|
|
|
2005-10-29 06:38:26 +08:00
|
|
|
kref_init(&uverbs_dev->ref);
|
2006-08-04 01:56:42 +08:00
|
|
|
init_completion(&uverbs_dev->comp);
|
2005-10-29 06:38:26 +08:00
|
|
|
|
2005-07-08 08:57:13 +08:00
|
|
|
spin_lock(&map_lock);
|
2010-02-03 03:07:59 +08:00
|
|
|
devnum = find_first_zero_bit(dev_map, IB_UVERBS_MAX_DEVICES);
|
|
|
|
if (devnum >= IB_UVERBS_MAX_DEVICES) {
|
2005-07-08 08:57:13 +08:00
|
|
|
spin_unlock(&map_lock);
|
2010-02-03 03:08:09 +08:00
|
|
|
devnum = find_overflow_devnum();
|
|
|
|
if (devnum < 0)
|
|
|
|
goto err;
|
|
|
|
|
|
|
|
spin_lock(&map_lock);
|
|
|
|
uverbs_dev->devnum = devnum + IB_UVERBS_MAX_DEVICES;
|
|
|
|
base = devnum + overflow_maj;
|
|
|
|
set_bit(devnum, overflow_map);
|
|
|
|
} else {
|
|
|
|
uverbs_dev->devnum = devnum;
|
|
|
|
base = devnum + IB_UVERBS_BASE_DEV;
|
|
|
|
set_bit(devnum, dev_map);
|
2005-07-08 08:57:13 +08:00
|
|
|
}
|
|
|
|
spin_unlock(&map_lock);
|
|
|
|
|
2005-09-27 04:53:25 +08:00
|
|
|
uverbs_dev->ib_dev = device;
|
2007-05-03 18:48:47 +08:00
|
|
|
uverbs_dev->num_comp_vectors = device->num_comp_vectors;
|
2005-07-08 08:57:13 +08:00
|
|
|
|
2010-02-03 03:07:49 +08:00
|
|
|
cdev_init(&uverbs_dev->cdev, NULL);
|
|
|
|
uverbs_dev->cdev.owner = THIS_MODULE;
|
|
|
|
uverbs_dev->cdev.ops = device->mmap ? &uverbs_mmap_fops : &uverbs_fops;
|
|
|
|
kobject_set_name(&uverbs_dev->cdev.kobj, "uverbs%d", uverbs_dev->devnum);
|
2010-02-03 03:08:04 +08:00
|
|
|
if (cdev_add(&uverbs_dev->cdev, base, 1))
|
2005-10-29 06:38:26 +08:00
|
|
|
goto err_cdev;
|
2005-07-08 08:57:13 +08:00
|
|
|
|
2008-07-22 11:03:34 +08:00
|
|
|
uverbs_dev->dev = device_create(uverbs_class, device->dma_device,
|
2010-02-03 03:07:49 +08:00
|
|
|
uverbs_dev->cdev.dev, uverbs_dev,
|
2008-07-22 11:03:34 +08:00
|
|
|
"uverbs%d", uverbs_dev->devnum);
|
2008-02-22 07:13:36 +08:00
|
|
|
if (IS_ERR(uverbs_dev->dev))
|
2005-07-08 08:57:13 +08:00
|
|
|
goto err_cdev;
|
|
|
|
|
2008-02-22 07:13:36 +08:00
|
|
|
if (device_create_file(uverbs_dev->dev, &dev_attr_ibdev))
|
2005-07-08 08:57:13 +08:00
|
|
|
goto err_class;
|
2008-02-22 07:13:36 +08:00
|
|
|
if (device_create_file(uverbs_dev->dev, &dev_attr_abi_version))
|
2005-09-30 05:17:48 +08:00
|
|
|
goto err_class;
|
2005-07-08 08:57:13 +08:00
|
|
|
|
|
|
|
ib_set_client_data(device, &uverbs_client, uverbs_dev);
|
|
|
|
|
|
|
|
return;
|
|
|
|
|
|
|
|
err_class:
|
2010-02-03 03:07:49 +08:00
|
|
|
device_destroy(uverbs_class, uverbs_dev->cdev.dev);
|
2005-07-08 08:57:13 +08:00
|
|
|
|
|
|
|
err_cdev:
|
2010-02-03 03:07:49 +08:00
|
|
|
cdev_del(&uverbs_dev->cdev);
|
2010-02-03 03:08:09 +08:00
|
|
|
if (uverbs_dev->devnum < IB_UVERBS_MAX_DEVICES)
|
|
|
|
clear_bit(devnum, dev_map);
|
|
|
|
else
|
|
|
|
clear_bit(devnum, overflow_map);
|
2005-07-08 08:57:13 +08:00
|
|
|
|
|
|
|
err:
|
2005-10-29 06:38:26 +08:00
|
|
|
kref_put(&uverbs_dev->ref, ib_uverbs_release_dev);
|
2006-08-04 01:56:42 +08:00
|
|
|
wait_for_completion(&uverbs_dev->comp);
|
|
|
|
kfree(uverbs_dev);
|
2005-07-08 08:57:13 +08:00
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void ib_uverbs_remove_one(struct ib_device *device)
|
|
|
|
{
|
|
|
|
struct ib_uverbs_device *uverbs_dev = ib_get_client_data(device, &uverbs_client);
|
|
|
|
|
|
|
|
if (!uverbs_dev)
|
|
|
|
return;
|
|
|
|
|
2008-02-22 07:13:36 +08:00
|
|
|
dev_set_drvdata(uverbs_dev->dev, NULL);
|
2010-02-03 03:07:49 +08:00
|
|
|
device_destroy(uverbs_class, uverbs_dev->cdev.dev);
|
|
|
|
cdev_del(&uverbs_dev->cdev);
|
2005-10-29 06:38:26 +08:00
|
|
|
|
2010-02-03 03:08:09 +08:00
|
|
|
if (uverbs_dev->devnum < IB_UVERBS_MAX_DEVICES)
|
|
|
|
clear_bit(uverbs_dev->devnum, dev_map);
|
|
|
|
else
|
|
|
|
clear_bit(uverbs_dev->devnum - IB_UVERBS_MAX_DEVICES, overflow_map);
|
2006-08-04 01:56:42 +08:00
|
|
|
|
2005-10-29 06:38:26 +08:00
|
|
|
kref_put(&uverbs_dev->ref, ib_uverbs_release_dev);
|
2006-08-04 01:56:42 +08:00
|
|
|
wait_for_completion(&uverbs_dev->comp);
|
|
|
|
kfree(uverbs_dev);
|
2005-07-08 08:57:13 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
static int __init ib_uverbs_init(void)
|
|
|
|
{
|
|
|
|
int ret;
|
|
|
|
|
|
|
|
ret = register_chrdev_region(IB_UVERBS_BASE_DEV, IB_UVERBS_MAX_DEVICES,
|
|
|
|
"infiniband_verbs");
|
|
|
|
if (ret) {
|
|
|
|
printk(KERN_ERR "user_verbs: couldn't register device number\n");
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
2005-10-29 06:38:26 +08:00
|
|
|
uverbs_class = class_create(THIS_MODULE, "infiniband_verbs");
|
|
|
|
if (IS_ERR(uverbs_class)) {
|
|
|
|
ret = PTR_ERR(uverbs_class);
|
2005-07-08 08:57:13 +08:00
|
|
|
printk(KERN_ERR "user_verbs: couldn't create class infiniband_verbs\n");
|
|
|
|
goto out_chrdev;
|
|
|
|
}
|
|
|
|
|
2010-01-05 19:48:09 +08:00
|
|
|
ret = class_create_file(uverbs_class, &class_attr_abi_version.attr);
|
2005-07-08 08:57:13 +08:00
|
|
|
if (ret) {
|
|
|
|
printk(KERN_ERR "user_verbs: couldn't create abi_version attribute\n");
|
|
|
|
goto out_class;
|
|
|
|
}
|
|
|
|
|
|
|
|
ret = ib_register_client(&uverbs_client);
|
|
|
|
if (ret) {
|
|
|
|
printk(KERN_ERR "user_verbs: couldn't register client\n");
|
2010-02-25 08:51:20 +08:00
|
|
|
goto out_class;
|
2005-07-08 08:57:13 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
out_class:
|
2005-10-29 06:38:26 +08:00
|
|
|
class_destroy(uverbs_class);
|
2005-07-08 08:57:13 +08:00
|
|
|
|
|
|
|
out_chrdev:
|
|
|
|
unregister_chrdev_region(IB_UVERBS_BASE_DEV, IB_UVERBS_MAX_DEVICES);
|
|
|
|
|
|
|
|
out:
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void __exit ib_uverbs_cleanup(void)
|
|
|
|
{
|
|
|
|
ib_unregister_client(&uverbs_client);
|
2005-10-29 06:38:26 +08:00
|
|
|
class_destroy(uverbs_class);
|
2005-07-08 08:57:13 +08:00
|
|
|
unregister_chrdev_region(IB_UVERBS_BASE_DEV, IB_UVERBS_MAX_DEVICES);
|
2010-02-03 03:08:09 +08:00
|
|
|
if (overflow_maj)
|
|
|
|
unregister_chrdev_region(overflow_maj, IB_UVERBS_MAX_DEVICES);
|
2005-10-25 01:53:25 +08:00
|
|
|
idr_destroy(&ib_uverbs_pd_idr);
|
|
|
|
idr_destroy(&ib_uverbs_mr_idr);
|
|
|
|
idr_destroy(&ib_uverbs_mw_idr);
|
|
|
|
idr_destroy(&ib_uverbs_ah_idr);
|
|
|
|
idr_destroy(&ib_uverbs_cq_idr);
|
|
|
|
idr_destroy(&ib_uverbs_qp_idr);
|
|
|
|
idr_destroy(&ib_uverbs_srq_idr);
|
2005-07-08 08:57:13 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
module_init(ib_uverbs_init);
|
|
|
|
module_exit(ib_uverbs_cleanup);
|