mir/ovs
2
0
Fork 0
mirror of https://github.com/openvswitch/ovs synced 2025-08-22 18:07:40 +00:00
Code Issues Releases Wiki Activity
ovs/lib/dpif.h

122 lines
4.2 KiB
C
Raw Normal View History

Import from old repository commit 61ef2b42a9c4ba8e1600f15bb0236765edc2ad45.
2009-07-08 13:19:16 -07:00
/*
dpif: Update dpif interface to match netdev. This brings over some features that were added to the netdev interface, most notably the separation between the name and the type. In addition to being cleaner, this also avoids problems where it is expected that the local port has the same name as the datapath.
2010-01-22 14:37:10 -05:00
* Copyright (c) 2008, 2009, 2010 Nicira Networks.
Import from old repository commit 61ef2b42a9c4ba8e1600f15bb0236765edc2ad45.
2009-07-08 13:19:16 -07:00
*
Update primary code license to Apache 2.0.
2009-06-15 15:11:30 -07:00
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at:
Import from old repository commit 61ef2b42a9c4ba8e1600f15bb0236765edc2ad45.
2009-07-08 13:19:16 -07:00
*
Update primary code license to Apache 2.0.
2009-06-15 15:11:30 -07:00
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
Import from old repository commit 61ef2b42a9c4ba8e1600f15bb0236765edc2ad45.
2009-07-08 13:19:16 -07:00
*/
#ifndef DPIF_H
#define DPIF_H 1
#include <stdbool.h>
#include <stddef.h>
#include <stdint.h>
ofproto: Avoid user->kernel->user round-trip for many controller actions. When an OpenFlow flow says to send packets to the controller, until now ofproto has executed that using dpif_execute(), which passes the packet up to the kernel. The kernel queues the packet into its "action" queue, and then later ofproto pulls the packet back down from the kernel and sends it to the controller. However, this is unnecessary. Open vSwitch can just recognize in advance that it will get the packet back and handle it directly, skipping the round trip. This commit implements this optimization. This generally affects only the first packet in a flow, since generally the rest come directly down from the kernel. It only optimizes the "easy" case where the first action in a flow is to send the packet to the controller, since this seems to be the common case in the flows that I'm looking at now.
2010-08-04 14:08:26 -07:00
#include "openflow/openflow.h"
#include "openvswitch/datapath-protocol.h"
#include "util.h"
Import from old repository commit 61ef2b42a9c4ba8e1600f15bb0236765edc2ad45.
2009-07-08 13:19:16 -07:00
Add extern "C" to more header files. From partner.
2010-02-17 10:36:57 -05:00
#ifdef __cplusplus
extern "C" {
#endif
dpif: Hide the contents of struct dpif. This helps prepare for multiple dpif implementations, and ensures that code outside dpif.c does not depend on its internals.
2009-06-16 10:09:10 -07:00
struct dpif;
datapath: Make adding and attaching a vport a single step. For some time now, Open vSwitch datapaths have internally made a distinction between adding a vport and attaching it to a datapath. Adding a vport just means to create it, as an entity detached from any datapath. Attaching it gives it a port number and a datapath. Similarly, a vport could be detached and deleted separately. After some study, I think I understand why this distinction exists. It is because ovs-vswitchd tries to open all the datapath ports before it tries to create them. However, changing it to create them before it tries to open them is not difficult, so this commit does this. The bulk of this commit, however, changes the datapath interface to one that always creates a vport and attaches it to a datapath in a single step, and similarly detaches a vport and deletes it in a single step. Signed-off-by: Ben Pfaff <blp@nicira.com> Acked-by: Jesse Gross <jesse@nicira.com>
2010-12-03 14:41:38 -08:00
struct netdev;
datapath: Replace "struct odp_action" by Netlink attributes. In the medium term, we plan to migrate the datapath to use Netlink as its communication channel. In the short term, we need to be able to have actions with 64-bit arguments but "struct odp_action" only has room for 48 bits. So this patch shifts to variable-length arguments using Netlink attributes, which starts in on the Netlink transition and makes 64-bit arguments possible at the same time. Signed-off-by: Ben Pfaff <blp@nicira.com> Acked-by: Jesse Gross <jesse@nicira.com>
2010-12-10 10:40:58 -08:00
struct nlattr;
Import from old repository commit 61ef2b42a9c4ba8e1600f15bb0236765edc2ad45.
2009-07-08 13:19:16 -07:00
struct ofpbuf;
vswitch: Avoid knowledge of details specific to Linux datapaths. At startup, the vswitch needs to delete datapaths that are not configured by the administrator. Until now this was done by knowing the possible names of Linux datapaths. This commit cleans up by allowing each datapath class to enumerate its existing datapaths and their names.
2009-07-06 11:06:36 -07:00
struct svec;
dpif: Allow providers to be managed at runtime. The list of datapath providers was previously staticly defined at compile time. This allows new providers to be added and removed at runtime.
2010-02-01 11:36:01 -05:00
struct dpif_class;
Import from old repository commit 61ef2b42a9c4ba8e1600f15bb0236765edc2ad45.
2009-07-08 13:19:16 -07:00
dpif: Add new functions dp_run() and dp_wait(). The upcoming netdev-based dpif needs a hook where it can process packets and throw them against the flow table, and this provides a suitable place.
2009-06-19 14:09:09 -07:00
void dp_run(void);
void dp_wait(void);
dpif: Allow providers to be managed at runtime. The list of datapath providers was previously staticly defined at compile time. This allows new providers to be added and removed at runtime.
2010-02-01 11:36:01 -05:00
int dp_register_provider(const struct dpif_class *);
int dp_unregister_provider(const char *type);
dpif: Update dpif interface to match netdev. This brings over some features that were added to the netdev interface, most notably the separation between the name and the type. In addition to being cleaner, this also avoids problems where it is expected that the local port has the same name as the datapath.
2010-01-22 14:37:10 -05:00
void dp_enumerate_types(struct svec *types);
dpif: Allow providers to be managed at runtime. The list of datapath providers was previously staticly defined at compile time. This allows new providers to be added and removed at runtime.
2010-02-01 11:36:01 -05:00
dpif: Update dpif interface to match netdev. This brings over some features that were added to the netdev interface, most notably the separation between the name and the type. In addition to being cleaner, this also avoids problems where it is expected that the local port has the same name as the datapath.
2010-01-22 14:37:10 -05:00
int dp_enumerate_names(const char *type, struct svec *names);
void dp_parse_name(const char *datapath_name, char **name, char **type);
dpif: Add new functions dp_run() and dp_wait(). The upcoming netdev-based dpif needs a hook where it can process packets and throw them against the flow table, and this provides a suitable place.
2009-06-19 14:09:09 -07:00
dpif: Update dpif interface to match netdev. This brings over some features that were added to the netdev interface, most notably the separation between the name and the type. In addition to being cleaner, this also avoids problems where it is expected that the local port has the same name as the datapath.
2010-01-22 14:37:10 -05:00
int dpif_open(const char *name, const char *type, struct dpif **);
int dpif_create(const char *name, const char *type, struct dpif **);
int dpif_create_and_open(const char *name, const char *type, struct dpif **);
Import from old repository commit 61ef2b42a9c4ba8e1600f15bb0236765edc2ad45.
2009-07-08 13:19:16 -07:00
void dpif_close(struct dpif *);
dpif: Replace dpif_id() by dpif_name(). dpif_id() is often used in error messages, e.g. "dp%u: screwed up". But soon we will be generalizing the concept of a datapath, so it is better to have a function that returns a full name, e.g. "%s: screwed up". Accordingly, this commit replaces dpif_id() by a new function dpif_name() that does so.
2009-06-16 11:00:22 -07:00
const char *dpif_name(const struct dpif *);
dpif: Update dpif interface to match netdev. This brings over some features that were added to the netdev interface, most notably the separation between the name and the type. In addition to being cleaner, this also avoids problems where it is expected that the local port has the same name as the datapath.
2010-01-22 14:37:10 -05:00
const char *dpif_base_name(const struct dpif *);
vswitch: Avoid knowledge of details specific to Linux datapaths. At startup, the vswitch needs to delete datapaths that are not configured by the administrator. Until now this was done by knowing the possible names of Linux datapaths. This commit cleans up by allowing each datapath class to enumerate its existing datapaths and their names.
2009-07-06 11:06:36 -07:00
int dpif_get_all_names(const struct dpif *, struct svec *);
Import from old repository commit 61ef2b42a9c4ba8e1600f15bb0236765edc2ad45.
2009-07-08 13:19:16 -07:00
int dpif_delete(struct dpif *);
int dpif_get_dp_stats(const struct dpif *, struct odp_stats *);
int dpif_get_drop_frags(const struct dpif *, bool *drop_frags);
int dpif_set_drop_frags(struct dpif *, bool drop_frags);
datapath: Make adding and attaching a vport a single step. For some time now, Open vSwitch datapaths have internally made a distinction between adding a vport and attaching it to a datapath. Adding a vport just means to create it, as an entity detached from any datapath. Attaching it gives it a port number and a datapath. Similarly, a vport could be detached and deleted separately. After some study, I think I understand why this distinction exists. It is because ovs-vswitchd tries to open all the datapath ports before it tries to create them. However, changing it to create them before it tries to open them is not difficult, so this commit does this. The bulk of this commit, however, changes the datapath interface to one that always creates a vport and attaches it to a datapath in a single step, and similarly detaches a vport and deletes it in a single step. Signed-off-by: Ben Pfaff <blp@nicira.com> Acked-by: Jesse Gross <jesse@nicira.com>
2010-12-03 14:41:38 -08:00
int dpif_port_add(struct dpif *, struct netdev *, uint16_t *port_nop);
Import from old repository commit 61ef2b42a9c4ba8e1600f15bb0236765edc2ad45.
2009-07-08 13:19:16 -07:00
int dpif_port_del(struct dpif *, uint16_t port_no);
int dpif_port_query_by_number(const struct dpif *, uint16_t port_no,
struct odp_port *);
int dpif_port_query_by_name(const struct dpif *, const char *devname,
struct odp_port *);
dpif: Rename dpif_get_name() to dpif_port_get_name(), update interface. With multiple kinds of datapaths, code should not just use "dp%u" along with dpif_minor() to print a datapath name, because not all datapaths can sensibly be named that way. We want to use a function with a name like dpif_get_name() to retrieve a datapath name for printing to the user, in which case the existing dpif_get_name() function would be confusing. So rename the existing one to something more explicit.
2009-06-15 16:51:46 -07:00
int dpif_port_get_name(struct dpif *, uint16_t port_no,
char *name, size_t name_size);
Import from old repository commit 61ef2b42a9c4ba8e1600f15bb0236765edc2ad45.
2009-07-08 13:19:16 -07:00
int dpif_port_list(const struct dpif *, struct odp_port **, size_t *n_ports);
Introduce general-purpose ways to wait for dpif and netdev changes. The dpif and netdev code has had various ways to check for changes to dpifs and netdevs over the course of Open vSwitch development. All of these have been thus far fairly specific to the Linux implementation. This commit is the start of a more general API for watching for such changes. The dpif-related parts seem fairly mature and so they are documented, the netdev parts will probably need to change somewhat and so they are not documented yet.
2009-06-24 10:24:09 -07:00
int dpif_port_poll(const struct dpif *, char **devnamep);
void dpif_port_poll_wait(const struct dpif *);
Import from old repository commit 61ef2b42a9c4ba8e1600f15bb0236765edc2ad45.
2009-07-08 13:19:16 -07:00
int dpif_flow_flush(struct dpif *);
int dpif_flow_put(struct dpif *, struct odp_flow_put *);
int dpif_flow_del(struct dpif *, struct odp_flow *);
int dpif_flow_get(const struct dpif *, struct odp_flow *);
int dpif_flow_get_multiple(const struct dpif *, struct odp_flow[], size_t n);
datapath: Change listing flows to use an iterator concept. One of the goals for Open vSwitch is to decouple kernel and userspace software, so that either one can be upgraded or rolled back independent of the other. To do this in full generality, it must be possible to change the kernel's idea of the flow key separately from the userspace version. In turn, that means that flow keys must become variable-length. This does not, however, fit in well with the ODP_FLOW_LIST ioctl in its current form, because that would require userspace to know how much space to allocate for each flow's key in advance, or to allocate as much space as could possibly be needed. Neither choice is very attractive. This commit prepares for a different solution, by replacing ODP_FLOW_LIST by a new ioctl ODP_FLOW_DUMP that retrieves a single flow from the datapath on each call. It is much cleaner to allocate the maximum amount of space for a single flow key than to do so for possibly a very large number of flow keys. As a side effect, this patch also fixes a race condition that sometimes made "ovs-dpctl dump-flows" print an error: previously, flows were listed and then their actions were retrieved, which left a window in which ovs-vswitchd could delete the flow. Now dumping a flow and its actions is a single step, closing that window. Dumping all of the flows in a datapath is no longer an atomic step, so now it is possible to miss some flows or see a single flow twice during iteration, if the flow table is modified by another process. It doesn't look like this should be a problem for ovs-vswitchd. It would be faster to retrieve a number of flows in batch instead of just one at a time, but that will naturally happen later when the kernel datapath interface is changed to use Netlink, so this patch does not bother with it. Signed-off-by: Ben Pfaff <blp@nicira.com> Acked-by: Jesse Gross <jesse@nicira.com>
2010-12-28 10:39:52 -08:00
struct dpif_flow_dump {
const struct dpif *dpif;
int error;
void *state;
};
void dpif_flow_dump_start(struct dpif_flow_dump *, const struct dpif *);
bool dpif_flow_dump_next(struct dpif_flow_dump *, struct odp_flow *);
int dpif_flow_dump_done(struct dpif_flow_dump *);
Import from old repository commit 61ef2b42a9c4ba8e1600f15bb0236765edc2ad45.
2009-07-08 13:19:16 -07:00
datapath: Replace "struct odp_action" by Netlink attributes. In the medium term, we plan to migrate the datapath to use Netlink as its communication channel. In the short term, we need to be able to have actions with 64-bit arguments but "struct odp_action" only has room for 48 bits. So this patch shifts to variable-length arguments using Netlink attributes, which starts in on the Netlink transition and makes 64-bit arguments possible at the same time. Signed-off-by: Ben Pfaff <blp@nicira.com> Acked-by: Jesse Gross <jesse@nicira.com>
2010-12-10 10:40:58 -08:00
int dpif_execute(struct dpif *, const struct nlattr *actions,
size_t actions_len, const struct ofpbuf *);
Import from old repository commit 61ef2b42a9c4ba8e1600f15bb0236765edc2ad45.
2009-07-08 13:19:16 -07:00
ofproto: Avoid user->kernel->user round-trip for many controller actions. When an OpenFlow flow says to send packets to the controller, until now ofproto has executed that using dpif_execute(), which passes the packet up to the kernel. The kernel queues the packet into its "action" queue, and then later ofproto pulls the packet back down from the kernel and sends it to the controller. However, this is unnecessary. Open vSwitch can just recognize in advance that it will get the packet back and handle it directly, skipping the round trip. This commit implements this optimization. This generally affects only the first packet in a flow, since generally the rest come directly down from the kernel. It only optimizes the "easy" case where the first action in a flow is to send the packet to the controller, since this seems to be the common case in the flows that I'm looking at now.
2010-08-04 14:08:26 -07:00
/* Minimum number of bytes of headroom for a packet returned by dpif_recv()
* member function. This headroom allows "struct odp_msg" to be replaced by
* "struct ofp_packet_in" without copying the buffer. */
Expand tunnel IDs from 32 to 64 bits. We have a need to identify tunnels with keys longer than 32 bits. This commit adds basic datapath and OpenFlow support for such keys. It doesn't actually add any tunnel protocols that support 64-bit keys, so this is not very useful yet. The 'arg' member of struct odp_msg had to be expanded to 64-bits also, because it sometimes contains a tunnel ID. This member also contains the argument passed to ODPAT_CONTROLLER, so I expanded that action's argument to 64 bits also so that it can use the full width of the expanded 'arg'. Userspace doesn't take advantage of the new space though (it was only using 16 bits anyhow). This commit has been tested only to the extent that it doesn't disrupt basic Open vSwitch operation. I have not tested it with tunnel traffic. Signed-off-by: Ben Pfaff <blp@nicira.com> Acked-by: Jesse Gross <jesse@nicira.com> Feature #3976.
2010-12-10 10:42:42 -08:00
#define DPIF_RECV_MSG_PADDING \
ROUND_UP(sizeof(struct ofp_packet_in) - sizeof(struct odp_msg), 8)
ofproto: Avoid user->kernel->user round-trip for many controller actions. When an OpenFlow flow says to send packets to the controller, until now ofproto has executed that using dpif_execute(), which passes the packet up to the kernel. The kernel queues the packet into its "action" queue, and then later ofproto pulls the packet back down from the kernel and sends it to the controller. However, this is unnecessary. Open vSwitch can just recognize in advance that it will get the packet back and handle it directly, skipping the round trip. This commit implements this optimization. This generally affects only the first packet in a flow, since generally the rest come directly down from the kernel. It only optimizes the "easy" case where the first action in a flow is to send the packet to the controller, since this seems to be the common case in the flows that I'm looking at now.
2010-08-04 14:08:26 -07:00
BUILD_ASSERT_DECL(sizeof(struct ofp_packet_in) > sizeof(struct odp_msg));
Expand tunnel IDs from 32 to 64 bits. We have a need to identify tunnels with keys longer than 32 bits. This commit adds basic datapath and OpenFlow support for such keys. It doesn't actually add any tunnel protocols that support 64-bit keys, so this is not very useful yet. The 'arg' member of struct odp_msg had to be expanded to 64-bits also, because it sometimes contains a tunnel ID. This member also contains the argument passed to ODPAT_CONTROLLER, so I expanded that action's argument to 64 bits also so that it can use the full width of the expanded 'arg'. Userspace doesn't take advantage of the new space though (it was only using 16 bits anyhow). This commit has been tested only to the extent that it doesn't disrupt basic Open vSwitch operation. I have not tested it with tunnel traffic. Signed-off-by: Ben Pfaff <blp@nicira.com> Acked-by: Jesse Gross <jesse@nicira.com> Feature #3976.
2010-12-10 10:42:42 -08:00
BUILD_ASSERT_DECL(DPIF_RECV_MSG_PADDING % 8 == 0);
ofproto: Avoid user->kernel->user round-trip for many controller actions. When an OpenFlow flow says to send packets to the controller, until now ofproto has executed that using dpif_execute(), which passes the packet up to the kernel. The kernel queues the packet into its "action" queue, and then later ofproto pulls the packet back down from the kernel and sends it to the controller. However, this is unnecessary. Open vSwitch can just recognize in advance that it will get the packet back and handle it directly, skipping the round trip. This commit implements this optimization. This generally affects only the first packet in a flow, since generally the rest come directly down from the kernel. It only optimizes the "easy" case where the first action in a flow is to send the packet to the controller, since this seems to be the common case in the flows that I'm looking at now.
2010-08-04 14:08:26 -07:00
dpif: Rename odp_msg related functions for more consistency. This seems like a more consistent naming scheme, since all of these functions are related but none of them were named similarly or grouped together.
2009-06-17 14:18:10 -07:00
int dpif_recv_get_mask(const struct dpif *, int *listen_mask);
int dpif_recv_set_mask(struct dpif *, int listen_mask);
Initial implementation of sFlow. Tested very slightly with "ping" and "sflowtool -t | tcpdump -r -".
2010-01-04 13:08:37 -08:00
int dpif_get_sflow_probability(const struct dpif *, uint32_t *probability);
int dpif_set_sflow_probability(struct dpif *, uint32_t probability);
Import from old repository commit 61ef2b42a9c4ba8e1600f15bb0236765edc2ad45.
2009-07-08 13:19:16 -07:00
int dpif_recv(struct dpif *, struct ofpbuf **);
dpif: Rename odp_msg related functions for more consistency. This seems like a more consistent naming scheme, since all of these functions are related but none of them were named similarly or grouped together.
2009-06-17 14:18:10 -07:00
int dpif_recv_purge(struct dpif *);
Import from old repository commit 61ef2b42a9c4ba8e1600f15bb0236765edc2ad45.
2009-07-08 13:19:16 -07:00
void dpif_recv_wait(struct dpif *);
dpif: New function dpif_get_netflow_ids(). The 'minor' member of struct dpif is used for two different purposes: for printing in log messages and for encapsulating in NetFlow messages. The needs in each case are different, so we should break up these uses. This commit does half of that, by introducing a new function to retrieve NetFlow ids and using it where appropriate.
2009-06-16 10:59:43 -07:00
void dpif_get_netflow_ids(const struct dpif *,
uint8_t *engine_type, uint8_t *engine_id);
Import from old repository commit 61ef2b42a9c4ba8e1600f15bb0236765edc2ad45.
2009-07-08 13:19:16 -07:00
dpif: Abstract translation from OpenFlow queue ID into ODP priority value. When the QoS code was integrated, I didn't yet know how to abstract the translation from a queue ID in an OpenFlow OFPAT_ENQUEUE action into a priority value for an ODP ODPAT_SET_PRIORITY action. This commit is a first attempt that works OK for Linux, so far. It's possible that in fact this translation needs the 'netdev' as an argument too, but it's not needed yet.
2010-07-20 11:23:21 -07:00
int dpif_queue_to_priority(const struct dpif *, uint32_t queue_id,
uint32_t *priority);
Add extern "C" to more header files. From partner.
2010-02-17 10:36:57 -05:00
#ifdef __cplusplus
}
#endif
Import from old repository commit 61ef2b42a9c4ba8e1600f15bb0236765edc2ad45.
2009-07-08 13:19:16 -07:00
#endif /* dpif.h */
Reference in New Issue Copy Permalink