mirrors/pipewire
1
0
Fork 0
mirror of https://gitlab.freedesktop.org/pipewire/pipewire.git synced 2025-10-31 22:25:38 -04:00
Code Issues Projects Releases Packages Wiki Activity Actions
pipewire/src/modules/module-protocol-native/v0/protocol-native.c

1372 lines
39 KiB
C
Raw Normal View History

protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
/* PipeWire
*
src: change license (LGPL -> MIT) on left-over LGPL files Discussed with and approved by Wim
2021-02-10 13:11:00 +02:00
* Copyright © 2017 Wim Taymans <wim.taymans@gmail.com>
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
*
src: change license (LGPL -> MIT) on left-over LGPL files Discussed with and approved by Wim
2021-02-10 13:11:00 +02:00
* Permission is hereby granted, free of charge, to any person obtaining a
* copy of this software and associated documentation files (the "Software"),
* to deal in the Software without restriction, including without limitation
* the rights to use, copy, modify, merge, publish, distribute, sublicense,
* and/or sell copies of the Software, and to permit persons to whom the
* Software is furnished to do so, subject to the following conditions:
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
*
src: change license (LGPL -> MIT) on left-over LGPL files Discussed with and approved by Wim
2021-02-10 13:11:00 +02:00
* The above copyright notice and this permission notice (including the next
* paragraph) shall be included in all copies or substantial portions of the
* Software.
*
* 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.
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
*/
#include <stdio.h>
#include <errno.h>
#include "spa/pod/parser.h"
avoid including private.h Expose some more methods to make it possible to build without including the private.h header, especially in modules.
2020-01-02 13:48:50 +01:00
#include "spa/pod/builder.h"
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
#include "spa/debug/pod.h"
treewide: replace strcmp() == 0 with spa_streq() This change is only done in source files for now, header files will be done separately.
2021-05-18 11:36:13 +10:00
#include "spa/utils/string.h"
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
#include "pipewire/pipewire.h"
#include "pipewire/private.h"
#include "pipewire/protocol.h"
#include "pipewire/resource.h"
includes: update all references to extensions to point to pipewire/extensions This also brings the advantage that all tools, examples, modules, components can also be compiled standalone out-of-tree using libpipewire from the system
2021-06-18 17:51:24 +03:00
#include "pipewire/extensions/protocol-native.h"
#include "pipewire/extensions/metadata.h"
#include "pipewire/extensions/session-manager.h"
#include "pipewire/extensions/client-node.h"
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
#include "interfaces.h"
#include "typemap.h"
#include "../connection.h"
modules: switch protocol-native to the new logging system This is more complicated than a normal module because we have two logging topics: mod.protocol-native and conn.protocol-native for wire messages. Because the latter use spa_debug (through spa_debug_pod) we need to #define our way around so those too use the right topics. Note that this removes the previous "connection" category, it is now "conn.protocol-native" instead.
2021-09-17 10:12:28 +10:00
PW_LOG_TOPIC_EXTERN(mod_topic);
#define PW_LOG_TOPIC_DEFAULT mod_topic
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
#define PW_PROTOCOL_NATIVE_FLAG_REMAP (1<<0)
SPA_EXPORT
pw_client -> pw_impl_client
2019-12-11 11:21:43 +01:00
uint32_t pw_protocol_native0_find_type(struct pw_impl_client *client, const char *type)
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
{
uint32_t i;
for (i = 0; i < SPA_N_ELEMENTS(type_map); i++) {
treewide: replace !strcmp() with spa_streq() This change is only done in source files for now, header files will be done separately.
2021-05-18 11:40:50 +10:00
if (spa_streq(type_map[i].type, type))
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
return i;
}
return SPA_ID_INVALID;
}
static void
update_types_server(struct pw_resource *resource)
{
struct spa_pod_builder *b;
struct spa_pod_frame f;
uint32_t i;
rename core_proxy -> core Rename core_proxy to core and move the introspect and interface contents to core.h In an effort to promote the proxy API.
2019-12-11 07:46:59 +01:00
b = pw_protocol_native_begin_resource(resource, PW_CORE_V0_EVENT_UPDATE_TYPES, NULL);
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
spa_pod_builder_push_struct(b, &f);
spa_pod_builder_add(b,
"i", 0,
"i", SPA_N_ELEMENTS(type_map), NULL);
for (i = 0; i < SPA_N_ELEMENTS(type_map); i++) {
spa_pod_builder_add(b, "s", type_map[i].type, NULL);
}
spa_pod_builder_pop(b, &f);
pw_protocol_native_end_resource(resource, b);
}
Cleanup argument names in methods and events interfaces First method argument is object, first event argument is data. Closes !963
2021-09-27 15:09:32 +02:00
static void core_marshal_info(void *data, const struct pw_core_info *info)
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
{
Cleanup argument names in methods and events interfaces First method argument is object, first event argument is data. Closes !963
2021-09-27 15:09:32 +02:00
struct pw_resource *resource = data;
improve v2 compat mode Always send the server types when we got the first client types and don't rely on the message sequence number. Add some more types in the typemap.
2020-02-06 16:56:19 +01:00
struct pw_impl_client *client = pw_resource_get_client(resource);
struct protocol_compat_v2 *compat_v2 = client->compat_v2;
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
struct spa_pod_builder *b;
uint32_t i, n_items;
v0: set media class Set the media class on the client-node after we know the number of ports. Add more type mappings Fix change_mask on core_info
2019-12-17 17:34:10 +01:00
uint64_t change_mask = 0;
struct spa_pod_frame f;
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
struct pw_protocol_native_message *msg;
v0: set media class Set the media class on the client-node after we know the number of ports. Add more type mappings Fix change_mask on core_info
2019-12-17 17:34:10 +01:00
#define PW_CORE_V0_CHANGE_MASK_USER_NAME (1 << 0)
#define PW_CORE_V0_CHANGE_MASK_HOST_NAME (1 << 1)
#define PW_CORE_V0_CHANGE_MASK_VERSION (1 << 2)
#define PW_CORE_V0_CHANGE_MASK_NAME (1 << 3)
#define PW_CORE_V0_CHANGE_MASK_COOKIE (1 << 4)
#define PW_CORE_V0_CHANGE_MASK_PROPS (1 << 5)
improve v2 compat mode Always send the server types when we got the first client types and don't rely on the message sequence number. Add some more types in the typemap.
2020-02-06 16:56:19 +01:00
if (compat_v2->send_types) {
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
update_types_server(resource);
v0: set media class Set the media class on the client-node after we know the number of ports. Add more type mappings Fix change_mask on core_info
2019-12-17 17:34:10 +01:00
change_mask |= PW_CORE_V0_CHANGE_MASK_USER_NAME |
PW_CORE_V0_CHANGE_MASK_HOST_NAME |
PW_CORE_V0_CHANGE_MASK_VERSION |
PW_CORE_V0_CHANGE_MASK_NAME |
PW_CORE_V0_CHANGE_MASK_COOKIE;
improve v2 compat mode Always send the server types when we got the first client types and don't rely on the message sequence number. Add some more types in the typemap.
2020-02-06 16:56:19 +01:00
compat_v2->send_types = false;
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
}
improve v2 compat mode Always send the server types when we got the first client types and don't rely on the message sequence number. Add some more types in the typemap.
2020-02-06 16:56:19 +01:00
b = pw_protocol_native_begin_resource(resource, PW_CORE_V0_EVENT_INFO, &msg);
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
n_items = info->props ? info->props->n_items : 0;
v0: set media class Set the media class on the client-node after we know the number of ports. Add more type mappings Fix change_mask on core_info
2019-12-17 17:34:10 +01:00
if (info->change_mask & PW_CORE_CHANGE_MASK_PROPS)
change_mask |= PW_CORE_V0_CHANGE_MASK_PROPS;
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
spa_pod_builder_push_struct(b, &f);
spa_pod_builder_add(b,
"i", info->id,
v0: set media class Set the media class on the client-node after we know the number of ports. Add more type mappings Fix change_mask on core_info
2019-12-17 17:34:10 +01:00
"l", change_mask,
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
"s", info->user_name,
"s", info->host_name,
"s", info->version,
"s", info->name,
"i", info->cookie,
"i", n_items, NULL);
for (i = 0; i < n_items; i++) {
spa_pod_builder_add(b,
"s", info->props->items[i].key,
"s", info->props->items[i].value, NULL);
}
spa_pod_builder_pop(b, &f);
pw_protocol_native_end_resource(resource, b);
}
Cleanup argument names in methods and events interfaces First method argument is object, first event argument is data. Closes !963
2021-09-27 15:09:32 +02:00
static void core_marshal_done(void *data, uint32_t id, int seq)
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
{
Cleanup argument names in methods and events interfaces First method argument is object, first event argument is data. Closes !963
2021-09-27 15:09:32 +02:00
struct pw_resource *resource = data;
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
struct spa_pod_builder *b;
rename core_proxy -> core Rename core_proxy to core and move the introspect and interface contents to core.h In an effort to promote the proxy API.
2019-12-11 07:46:59 +01:00
b = pw_protocol_native_begin_resource(resource, PW_CORE_V0_EVENT_DONE, NULL);
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
spa_pod_builder_add_struct(b, "i", seq);
pw_protocol_native_end_resource(resource, b);
}
Cleanup argument names in methods and events interfaces First method argument is object, first event argument is data. Closes !963
2021-09-27 15:09:32 +02:00
static void core_marshal_error(void *data, uint32_t id, int seq, int res, const char *error)
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
{
Cleanup argument names in methods and events interfaces First method argument is object, first event argument is data. Closes !963
2021-09-27 15:09:32 +02:00
struct pw_resource *resource = data;
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
struct spa_pod_builder *b;
rename core_proxy -> core Rename core_proxy to core and move the introspect and interface contents to core.h In an effort to promote the proxy API.
2019-12-11 07:46:59 +01:00
b = pw_protocol_native_begin_resource(resource, PW_CORE_V0_EVENT_ERROR, NULL);
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
spa_pod_builder_add_struct(b,
"i", id,
"i", res,
"s", error);
pw_protocol_native_end_resource(resource, b);
}
Cleanup argument names in methods and events interfaces First method argument is object, first event argument is data. Closes !963
2021-09-27 15:09:32 +02:00
static void core_marshal_remove_id(void *data, uint32_t id)
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
{
Cleanup argument names in methods and events interfaces First method argument is object, first event argument is data. Closes !963
2021-09-27 15:09:32 +02:00
struct pw_resource *resource = data;
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
struct spa_pod_builder *b;
rename core_proxy -> core Rename core_proxy to core and move the introspect and interface contents to core.h In an effort to promote the proxy API.
2019-12-11 07:46:59 +01:00
b = pw_protocol_native_begin_resource(resource, PW_CORE_V0_EVENT_REMOVE_ID, NULL);
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
spa_pod_builder_add_struct(b, "i", id);
pw_protocol_native_end_resource(resource, b);
}
static int core_demarshal_client_update(void *object, const struct pw_protocol_native_message *msg)
{
struct pw_resource *resource = object;
avoid including private.h Expose some more methods to make it possible to build without including the private.h header, especially in modules.
2020-01-02 13:48:50 +01:00
struct pw_impl_client *client = pw_resource_get_client(resource);
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
struct spa_dict props;
struct spa_pod_parser prs;
struct spa_pod_frame f;
uint32_t i;
spa_pod_parser_init(&prs, msg->data, msg->size);
if (spa_pod_parser_push_struct(&prs, &f) < 0 ||
spa_pod_parser_get(&prs,
"i", &props.n_items, NULL) < 0)
return -EINVAL;
props.items = alloca(props.n_items * sizeof(struct spa_dict_item));
for (i = 0; i < props.n_items; i++) {
if (spa_pod_parser_get(&prs,
"s", &props.items[i].key,
"s", &props.items[i].value,
NULL) < 0)
return -EINVAL;
}
avoid including private.h Expose some more methods to make it possible to build without including the private.h header, especially in modules.
2020-01-02 13:48:50 +01:00
pw_impl_client_update_properties(client, &props);
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
return 0;
}
protocol-native: implement v2 update_permissions Convert the old permission update to new permission events.
2020-02-06 16:57:24 +01:00
static uint32_t parse_perms(const char *str)
{
uint32_t perms = 0;
while (*str != '\0') {
switch (*str++) {
case 'r':
perms |= PW_PERM_R;
break;
case 'w':
perms |= PW_PERM_W;
break;
case 'x':
perms |= PW_PERM_X;
break;
}
}
return perms;
}
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
static int core_demarshal_permissions(void *object, const struct pw_protocol_native_message *msg)
{
protocol-native: implement v2 update_permissions Convert the old permission update to new permission events.
2020-02-06 16:57:24 +01:00
struct pw_resource *resource = object;
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
struct spa_dict props;
struct spa_pod_parser prs;
struct spa_pod_frame f;
protocol-native: implement v2 update_permissions Convert the old permission update to new permission events.
2020-02-06 16:57:24 +01:00
uint32_t i, n_permissions;
struct pw_permission *permissions, defperm = { 0, };
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
spa_pod_parser_init(&prs, msg->data, msg->size);
if (spa_pod_parser_push_struct(&prs, &f) < 0 ||
spa_pod_parser_get(&prs, "i", &props.n_items, NULL) < 0)
return -EINVAL;
props.items = alloca(props.n_items * sizeof(struct spa_dict_item));
protocol-native: implement v2 update_permissions Convert the old permission update to new permission events.
2020-02-06 16:57:24 +01:00
n_permissions = 0;
permissions = alloca(props.n_items * sizeof(struct pw_permission));
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
for (i = 0; i < props.n_items; i++) {
protocol-native: implement v2 update_permissions Convert the old permission update to new permission events.
2020-02-06 16:57:24 +01:00
uint32_t id, perms;
const char *str;
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
if (spa_pod_parser_get(&prs,
"s", &props.items[i].key,
"s", &props.items[i].value,
NULL) < 0)
return -EINVAL;
protocol-native: implement v2 update_permissions Convert the old permission update to new permission events.
2020-02-06 16:57:24 +01:00
str = props.items[i].value;
/* first set global permissions */
treewide: replace strcmp() == 0 with spa_streq() This change is only done in source files for now, header files will be done separately.
2021-05-18 11:36:13 +10:00
if (spa_streq(props.items[i].key, PW_CORE_PERMISSIONS_GLOBAL)) {
protocol-native: implement v2 update_permissions Convert the old permission update to new permission events.
2020-02-06 16:57:24 +01:00
size_t len;
/* <global-id>:[r][w][x] */
len = strcspn(str, ":");
if (len == 0)
continue;
id = atoi(str);
perms = parse_perms(str + len);
permissions[n_permissions++] = PW_PERMISSION_INIT(id, perms);
treewide: replace strcmp() == 0 with spa_streq() This change is only done in source files for now, header files will be done separately.
2021-05-18 11:36:13 +10:00
} else if (spa_streq(props.items[i].key, PW_CORE_PERMISSIONS_DEFAULT)) {
protocol-native: implement v2 update_permissions Convert the old permission update to new permission events.
2020-02-06 16:57:24 +01:00
perms = parse_perms(str);
defperm = PW_PERMISSION_INIT(PW_ID_ANY, perms);
}
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
}
protocol-native: implement v2 update_permissions Convert the old permission update to new permission events.
2020-02-06 16:57:24 +01:00
/* add default permission if set */
if (defperm.id == PW_ID_ANY)
permissions[n_permissions++] = defperm;
for (i = 0; i < n_permissions; i++) {
pw_log_debug("%d: %d: %08x", i, permissions[i].id, permissions[i].permissions);
}
return pw_impl_client_update_permissions(resource->client,
n_permissions, permissions);
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
}
static int core_demarshal_hello(void *object, const struct pw_protocol_native_message *msg)
{
struct pw_resource *resource = object;
struct spa_pod_parser prs;
void *ptr;
spa_pod_parser_init(&prs, msg->data, msg->size);
if (spa_pod_parser_get_struct(&prs,
"P", &ptr) < 0)
return -EINVAL;
rename core_proxy -> core Rename core_proxy to core and move the introspect and interface contents to core.h In an effort to promote the proxy API.
2019-12-11 07:46:59 +01:00
return pw_resource_notify(resource, struct pw_core_methods, hello, 0, 2);
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
}
static int core_demarshal_sync(void *object, const struct pw_protocol_native_message *msg)
{
struct pw_resource *resource = object;
struct spa_pod_parser prs;
uint32_t seq;
spa_pod_parser_init(&prs, msg->data, msg->size);
if (spa_pod_parser_get_struct(&prs, "i", &seq) < 0)
return -EINVAL;
rename core_proxy -> core Rename core_proxy to core and move the introspect and interface contents to core.h In an effort to promote the proxy API.
2019-12-11 07:46:59 +01:00
return pw_resource_notify(resource, struct pw_core_methods, sync, 0, 0, seq);
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
}
static int core_demarshal_get_registry(void *object, const struct pw_protocol_native_message *msg)
{
struct pw_resource *resource = object;
struct spa_pod_parser prs;
int32_t version, new_id;
spa_pod_parser_init(&prs, msg->data, msg->size);
if (spa_pod_parser_get_struct(&prs,
"i", &version,
"i", &new_id) < 0)
return -EINVAL;
rename core_proxy -> core Rename core_proxy to core and move the introspect and interface contents to core.h In an effort to promote the proxy API.
2019-12-11 07:46:59 +01:00
return pw_resource_notify(resource, struct pw_core_methods, get_registry, 0, version, new_id);
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
}
SPA_EXPORT
pw_client -> pw_impl_client
2019-12-11 11:21:43 +01:00
uint32_t pw_protocol_native0_type_from_v2(struct pw_impl_client *client, uint32_t type)
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
{
void *t;
uint32_t index;
struct protocol_compat_v2 *compat_v2 = client->compat_v2;
if ((t = pw_map_lookup(&compat_v2->types, type)) == NULL)
return SPA_ID_INVALID;
index = PW_MAP_PTR_TO_ID(t);
if (index >= SPA_N_ELEMENTS(type_map))
return SPA_ID_INVALID;
return type_map[index].id;
}
Make interface types a string This is more in line with wayland and it allows us to create new interfaces in modules without having to add anything to the type enum. It also removes some lookups to map type_id to readable name in debug.
2019-12-19 13:15:10 +01:00
SPA_EXPORT
const char * pw_protocol_native0_name_from_v2(struct pw_impl_client *client, uint32_t type)
{
void *t;
uint32_t index;
struct protocol_compat_v2 *compat_v2 = client->compat_v2;
if ((t = pw_map_lookup(&compat_v2->types, type)) == NULL)
return NULL;
index = PW_MAP_PTR_TO_ID(t);
if (index >= SPA_N_ELEMENTS(type_map))
return NULL;
return type_map[index].name;
}
v0: improve compatibility Implement RequestClockUpdate event and ClockUpdate message for old clients.
2020-02-11 11:23:28 +01:00
SPA_EXPORT
uint32_t pw_protocol_native0_name_to_v2(struct pw_impl_client *client, const char *name)
{
uint32_t i;
/* match name to type table and return index */
for (i = 0; i < SPA_N_ELEMENTS(type_map); i++) {
treewide: replace !strcmp() with spa_streq() This change is only done in source files for now, header files will be done separately.
2021-05-18 11:40:50 +10:00
if (type_map[i].name != NULL && spa_streq(type_map[i].name, name))
v0: improve compatibility Implement RequestClockUpdate event and ClockUpdate message for old clients.
2020-02-11 11:23:28 +01:00
return i;
}
return SPA_ID_INVALID;
}
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
SPA_EXPORT
pw_client -> pw_impl_client
2019-12-11 11:21:43 +01:00
uint32_t pw_protocol_native0_type_to_v2(struct pw_impl_client *client,
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
const struct spa_type_info *info, uint32_t type)
{
const char *name;
/** find full name of type in type_info */
if ((name = spa_debug_type_find_name(info, type)) == NULL)
return SPA_ID_INVALID;
v0: improve compatibility Implement RequestClockUpdate event and ClockUpdate message for old clients.
2020-02-11 11:23:28 +01:00
return pw_protocol_native0_name_to_v2(client, name);
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
}
struct spa_pod_prop_body0 {
uint32_t key;
#define SPA_POD_PROP0_RANGE_NONE 0 /**< no range */
#define SPA_POD_PROP0_RANGE_MIN_MAX 1 /**< property has range */
#define SPA_POD_PROP0_RANGE_STEP 2 /**< property has range with step */
#define SPA_POD_PROP0_RANGE_ENUM 3 /**< property has enumeration */
#define SPA_POD_PROP0_RANGE_FLAGS 4 /**< property has flags */
#define SPA_POD_PROP0_RANGE_MASK 0xf /**< mask to select range type */
#define SPA_POD_PROP0_FLAG_UNSET (1 << 4) /**< property value is unset */
#define SPA_POD_PROP0_FLAG_OPTIONAL (1 << 5) /**< property value is optional */
#define SPA_POD_PROP0_FLAG_READONLY (1 << 6) /**< property is readonly */
#define SPA_POD_PROP0_FLAG_DEPRECATED (1 << 7) /**< property is deprecated */
#define SPA_POD_PROP0_FLAG_INFO (1 << 8) /**< property is informational and is not
* used when filtering */
uint32_t flags;
struct spa_pod value;
/* array with elements of value.size follows,
* first element is value/default, rest are alternatives */
};
/* v2 iterates object as containing spa_pod */
#define SPA_POD_OBJECT_BODY_FOREACH0(body, size, iter) \
spa/buffer: rename SPA_MEMBER to SPA_PTROFF SPA_MEMBER is misleading, all we're doing here is pointer+offset and a type-casting the result. Rename to SPA_PTROFF which is more expressive (and has the same number of characters so we don't need to re-indent).
2021-05-06 13:41:44 +10:00
for ((iter) = SPA_PTROFF((body), sizeof(struct spa_pod_object_body), struct spa_pod); \
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
spa_pod_is_inside(body, size, iter); \
(iter) = spa_pod_next(iter))
#define SPA_POD_PROP_ALTERNATIVE_FOREACH0(body, _size, iter) \
spa/buffer: rename SPA_MEMBER to SPA_PTROFF SPA_MEMBER is misleading, all we're doing here is pointer+offset and a type-casting the result. Rename to SPA_PTROFF which is more expressive (and has the same number of characters so we don't need to re-indent).
2021-05-06 13:41:44 +10:00
for ((iter) = SPA_PTROFF((body), (body)->value.size + \
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
sizeof(struct spa_pod_prop_body0), __typeof__(*iter)); \
spa/buffer: rename SPA_MEMBER to SPA_PTROFF SPA_MEMBER is misleading, all we're doing here is pointer+offset and a type-casting the result. Rename to SPA_PTROFF which is more expressive (and has the same number of characters so we don't need to re-indent).
2021-05-06 13:41:44 +10:00
(iter) <= SPA_PTROFF((body), (_size)-(body)->value.size, __typeof__(*iter)); \
(iter) = SPA_PTROFF((iter), (body)->value.size, __typeof__(*iter)))
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
protocol: add workaround for bad v0 clients Fix up clients that say they have fixed properties while in fact they are not. Assume that when there are alternatives, the property was in fact not fixed.
2020-06-09 13:22:07 +02:00
#define SPA0_POD_PROP_N_VALUES(b,size) ((size - sizeof(struct spa_pod_prop_body0)) / (b)->value.size)
pw_client -> pw_impl_client
2019-12-11 11:21:43 +01:00
static int remap_from_v2(uint32_t type, void *body, uint32_t size, struct pw_impl_client *client,
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
struct spa_pod_builder *builder)
{
int res;
switch (type) {
case SPA_TYPE_Id:
spa_pod_builder_id(builder, pw_protocol_native0_type_from_v2(client, *(int32_t*) body));
break;
/** choice was props in v2 */
case SPA_TYPE_Choice:
{
struct spa_pod_prop_body0 *b = body;
struct spa_pod_frame f;
void *alt;
uint32_t key = pw_protocol_native0_type_from_v2(client, b->key);
enum spa_choice_type type;
spa_pod_builder_prop(builder, key, 0);
switch (b->flags & SPA_POD_PROP0_RANGE_MASK) {
default:
case SPA_POD_PROP0_RANGE_NONE:
type = SPA_CHOICE_None;
break;
case SPA_POD_PROP0_RANGE_MIN_MAX:
type = SPA_CHOICE_Range;
break;
case SPA_POD_PROP0_RANGE_STEP:
type = SPA_CHOICE_Step;
break;
case SPA_POD_PROP0_RANGE_ENUM:
type = SPA_CHOICE_Enum;
break;
case SPA_POD_PROP0_RANGE_FLAGS:
type = SPA_CHOICE_Flags;
break;
}
protocol: add workaround for bad v0 clients Fix up clients that say they have fixed properties while in fact they are not. Assume that when there are alternatives, the property was in fact not fixed.
2020-06-09 13:22:07 +02:00
if (!SPA_FLAG_IS_SET(b->flags, SPA_POD_PROP0_FLAG_UNSET) &&
SPA0_POD_PROP_N_VALUES(b, size) == 1)
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
type = SPA_CHOICE_None;
spa_pod_builder_push_choice(builder, &f, type, 0);
if (b->value.type == SPA_TYPE_Id) {
uint32_t id;
if ((res = spa_pod_get_id(&b->value, &id)) < 0)
return res;
spa_pod_builder_id(builder, pw_protocol_native0_type_from_v2(client, id));
SPA_POD_PROP_ALTERNATIVE_FOREACH0(b, size, alt)
if ((res = remap_from_v2(b->value.type, alt, b->value.size, client, builder)) < 0)
return res;
} else {
spa_pod_builder_raw(builder, &b->value, size - sizeof(struct spa_pod));
}
spa_pod_builder_pop(builder, &f);
break;
}
case SPA_TYPE_Object:
{
struct spa_pod_object_body *b = body;
struct spa_pod *p;
struct spa_pod_frame f;
uint32_t type, count = 0;
/* type and id are switched */
type = pw_protocol_native0_type_from_v2(client, b->id),
spa_pod_builder_push_object(builder, &f, type,
pw_protocol_native0_type_from_v2(client, b->type));
/* object contained pods in v2 */
SPA_POD_OBJECT_BODY_FOREACH0(b, size, p) {
if (type == SPA_TYPE_OBJECT_Format && count < 2) {
uint32_t id;
if (spa_pod_get_id(p, &id) < 0)
continue;
id = pw_protocol_native0_type_from_v2(client, id);
if (count == 0) {
spa_pod_builder_prop(builder, SPA_FORMAT_mediaType, 0);
spa_pod_builder_id(builder, id);
}
if (count == 1) {
spa_pod_builder_prop(builder, SPA_FORMAT_mediaSubtype, 0);
spa_pod_builder_id(builder, id);
}
count++;
continue;
}
if ((res = remap_from_v2(p->type,
SPA_POD_BODY(p),
p->size,
client, builder)) < 0)
return res;
}
spa_pod_builder_pop(builder, &f);
break;
}
case SPA_TYPE_Struct:
{
struct spa_pod *b = body, *p;
struct spa_pod_frame f;
spa_pod_builder_push_struct(builder, &f);
SPA_POD_FOREACH(b, size, p)
if ((res = remap_from_v2(p->type, SPA_POD_BODY(p), p->size, client, builder)) < 0)
return res;
spa_pod_builder_pop(builder, &f);
break;
}
default:
break;
}
return 0;
}
pw_client -> pw_impl_client
2019-12-11 11:21:43 +01:00
static int remap_to_v2(struct pw_impl_client *client, const struct spa_type_info *info,
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
uint32_t type, void *body, uint32_t size,
struct spa_pod_builder *builder)
{
int res;
switch (type) {
case SPA_TYPE_Id:
spa_pod_builder_id(builder, pw_protocol_native0_type_to_v2(client, info, *(int32_t*) body));
break;
case SPA_TYPE_Object:
{
struct spa_pod_object_body *b = body;
struct spa_pod_prop *p;
struct spa_pod_frame f[2];
uint32_t type;
const struct spa_type_info *ti, *ii;
ti = spa_debug_type_find(info, b->type);
ii = ti ? spa_debug_type_find(ti->values, 0) : NULL;
v0: improve compatibility Implement RequestClockUpdate event and ClockUpdate message for old clients.
2020-02-11 11:23:28 +01:00
if (b->type == SPA_TYPE_COMMAND_Node ||
b->type == SPA_TYPE_EVENT_Node) {
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
spa_pod_builder_push_object(builder, &f[0], 0,
pw_protocol_native0_type_to_v2(client, ii ? ii->values : NULL, b->id));
} else {
ii = ii ? spa_debug_type_find(ii->values, b->id) : NULL;
/* type and id are switched */
type = pw_protocol_native0_type_to_v2(client, info, b->type),
spa_pod_builder_push_object(builder, &f[0],
pw_protocol_native0_type_to_v2(client, ii ? ii->values : NULL, b->id), type);
}
info = ti ? ti->values : info;
SPA_POD_OBJECT_BODY_FOREACH(b, size, p) {
uint32_t key, flags;
uint32_t n_vals, choice;
struct spa_pod *values;
ii = spa_debug_type_find(info, p->key);
values = spa_pod_get_values(&p->value, &n_vals, &choice);
if (b->type == SPA_TYPE_OBJECT_Format &&
(p->key == SPA_FORMAT_mediaType ||
p->key == SPA_FORMAT_mediaSubtype)) {
uint32_t val;
if (spa_pod_get_id(values, &val) < 0)
continue;
spa_pod_builder_id(builder,
pw_protocol_native0_type_to_v2(client, ii ? ii->values : NULL, val));
continue;
}
flags = 0;
switch(choice) {
case SPA_CHOICE_None:
flags |= SPA_POD_PROP0_RANGE_NONE;
break;
case SPA_CHOICE_Range:
flags |= SPA_POD_PROP0_RANGE_MIN_MAX | SPA_POD_PROP0_FLAG_UNSET;
break;
case SPA_CHOICE_Step:
flags |= SPA_POD_PROP0_RANGE_STEP | SPA_POD_PROP0_FLAG_UNSET;
break;
case SPA_CHOICE_Enum:
flags |= SPA_POD_PROP0_RANGE_ENUM | SPA_POD_PROP0_FLAG_UNSET;
break;
case SPA_CHOICE_Flags:
flags |= SPA_POD_PROP0_RANGE_FLAGS | SPA_POD_PROP0_FLAG_UNSET;
break;
}
key = pw_protocol_native0_type_to_v2(client, info, p->key);
spa_pod_builder_push_choice(builder, &f[1], key, flags);
if (values->type == SPA_TYPE_Id) {
uint32_t i, *id = SPA_POD_BODY(values);
for (i = 0; i < n_vals; i++) {
spa_pod_builder_id(builder,
pw_protocol_native0_type_to_v2(client, ii ? ii->values : NULL, id[i]));
}
} else {
spa_pod_builder_raw(builder, values, sizeof(struct spa_pod) + n_vals * values->size);
}
spa_pod_builder_pop(builder, &f[1]);
}
spa_pod_builder_pop(builder, &f[0]);
break;
}
case SPA_TYPE_Struct:
{
struct spa_pod *b = body, *p;
struct spa_pod_frame f;
spa_pod_builder_push_struct(builder, &f);
SPA_POD_FOREACH(b, size, p)
if ((res = remap_to_v2(client, info, p->type, SPA_POD_BODY(p), p->size, builder)) < 0)
return res;
spa_pod_builder_pop(builder, &f);
break;
}
default:
break;
}
return 0;
}
SPA_EXPORT
pw_client -> pw_impl_client
2019-12-11 11:21:43 +01:00
struct spa_pod * pw_protocol_native0_pod_from_v2(struct pw_impl_client *client, const struct spa_pod *pod)
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
{
uint8_t buffer[4096];
struct spa_pod *copy;
struct spa_pod_builder b = SPA_POD_BUILDER_INIT(buffer, 4096);
int res;
if (pod == NULL)
return NULL;
if ((res = remap_from_v2(SPA_POD_TYPE(pod),
SPA_POD_BODY(pod),
SPA_POD_BODY_SIZE(pod),
client, &b)) < 0) {
errno = -res;
return NULL;
}
copy = spa_pod_copy(b.data);
return copy;
}
SPA_EXPORT
pw_client -> pw_impl_client
2019-12-11 11:21:43 +01:00
int pw_protocol_native0_pod_to_v2(struct pw_impl_client *client, const struct spa_pod *pod,
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
struct spa_pod_builder *b)
{
int res;
if (pod == NULL) {
spa_pod_builder_none(b);
return 0;
}
if ((res = remap_to_v2(client, pw_type_info(),
SPA_POD_TYPE(pod),
SPA_POD_BODY(pod),
SPA_POD_BODY_SIZE(pod),
b)) < 0) {
return -res;
}
return 0;
}
static int core_demarshal_create_object(void *object, const struct pw_protocol_native_message *msg)
{
struct pw_resource *resource = object;
pw_client -> pw_impl_client
2019-12-11 11:21:43 +01:00
struct pw_impl_client *client = pw_resource_get_client(resource);
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
struct spa_pod_parser prs;
struct spa_pod_frame f;
uint32_t version, type, new_id, i;
Make interface types a string This is more in line with wayland and it allows us to create new interfaces in modules without having to add anything to the type enum. It also removes some lookups to map type_id to readable name in debug.
2019-12-19 13:15:10 +01:00
const char *factory_name, *type_name;
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
struct spa_dict props;
spa_pod_parser_init(&prs, msg->data, msg->size);
if (spa_pod_parser_push_struct(&prs, &f) < 0 ||
spa_pod_parser_get(&prs,
"s", &factory_name,
"I", &type,
"i", &version,
"i", &props.n_items, NULL) < 0)
return -EINVAL;
props.items = alloca(props.n_items * sizeof(struct spa_dict_item));
for (i = 0; i < props.n_items; i++) {
if (spa_pod_parser_get(&prs,
"s", &props.items[i].key,
"s", &props.items[i].value, NULL) < 0)
return -EINVAL;
}
if (spa_pod_parser_get(&prs, "i", &new_id, NULL) < 0)
return -EINVAL;
Make interface types a string This is more in line with wayland and it allows us to create new interfaces in modules without having to add anything to the type enum. It also removes some lookups to map type_id to readable name in debug.
2019-12-19 13:15:10 +01:00
type_name = pw_protocol_native0_name_from_v2(client, type);
if (type_name == NULL)
return -EINVAL;
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
rename core_proxy -> core Rename core_proxy to core and move the introspect and interface contents to core.h In an effort to promote the proxy API.
2019-12-11 07:46:59 +01:00
return pw_resource_notify(resource, struct pw_core_methods, create_object, 0, factory_name,
Make interface types a string This is more in line with wayland and it allows us to create new interfaces in modules without having to add anything to the type enum. It also removes some lookups to map type_id to readable name in debug.
2019-12-19 13:15:10 +01:00
type_name, version,
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
&props, new_id);
}
static int core_demarshal_destroy(void *object, const struct pw_protocol_native_message *msg)
{
struct pw_resource *resource = object, *r;
pw_client -> pw_impl_client
2019-12-11 11:21:43 +01:00
struct pw_impl_client *client = pw_resource_get_client(resource);
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
struct spa_pod_parser prs;
uint32_t id;
spa_pod_parser_init(&prs, msg->data, msg->size);
if (spa_pod_parser_get_struct(&prs,
"i", &id, NULL) < 0)
return -EINVAL;
pw_log_debug("client %p: destroy resource %u", client, id);
pw_client -> pw_impl_client
2019-12-11 11:21:43 +01:00
if ((r = pw_impl_client_find_resource(client, id)) == NULL)
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
goto no_resource;
rename core_proxy -> core Rename core_proxy to core and move the introspect and interface contents to core.h In an effort to promote the proxy API.
2019-12-11 07:46:59 +01:00
return pw_resource_notify(resource, struct pw_core_methods, destroy, 0, r);
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
no_resource:
Run codespell on the entire codebase This tool detects and fixes common English spelling mistakes, with generally very few mistakes. Here is the command I used to generate this commit. There were a few changes that had to be done manually, and of course adding the ignore file: ```shell codespell -I .codespell-ignore -x .codespell-ignore -w ``` I didn’t add it to the CI, but this would be a good place for it.
2020-07-22 20:54:06 +02:00
pw_log_error("client %p: unknown resource %u op:%u", client, id, msg->opcode);
return ENOENT for unknown resources Fix some other errors as well, ENOSYS for invalid interface, ENOTSUP for unimplemented methods.
2020-11-20 12:11:37 +01:00
pw_resource_errorf(resource, -ENOENT, "unknown resource %d op:%u", id, msg->opcode);
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
return 0;
}
static int core_demarshal_update_types_server(void *object, const struct pw_protocol_native_message *msg)
{
struct pw_resource *resource = object;
pw_client -> pw_impl_client
2019-12-11 11:21:43 +01:00
struct pw_impl_client *client = pw_resource_get_client(resource);
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
struct protocol_compat_v2 *compat_v2 = client->compat_v2;
struct spa_pod_parser prs;
uint32_t first_id, n_types;
struct spa_pod_frame f;
const char **types;
uint32_t i;
spa_pod_parser_init(&prs, msg->data, msg->size);
if (spa_pod_parser_push_struct(&prs, &f) < 0 ||
spa_pod_parser_get(&prs,
"i", &first_id,
"i", &n_types,
NULL) < 0)
return -EINVAL;
improve v2 compat mode Always send the server types when we got the first client types and don't rely on the message sequence number. Add some more types in the typemap.
2020-02-06 16:56:19 +01:00
if (first_id == 0)
compat_v2->send_types = true;
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
types = alloca(n_types * sizeof(char *));
for (i = 0; i < n_types; i++) {
if (spa_pod_parser_get(&prs, "s", &types[i], NULL) < 0)
return -EINVAL;
}
for (i = 0; i < n_types; i++, first_id++) {
protocol-native: improve v0 compatibility
2020-02-10 18:15:24 +01:00
uint32_t type_id = pw_protocol_native0_find_type(client, types[i]);
if (type_id == SPA_ID_INVALID)
continue;
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
if (pw_map_insert_at(&compat_v2->types, first_id, PW_MAP_ID_TO_PTR(type_id)) < 0)
pw_log_error("can't add type %d->%d for client", first_id, type_id);
}
return 0;
}
Cleanup argument names in methods and events interfaces First method argument is object, first event argument is data. Closes !963
2021-09-27 15:09:32 +02:00
static void registry_marshal_global(void *data, uint32_t id, uint32_t permissions,
Make interface types a string This is more in line with wayland and it allows us to create new interfaces in modules without having to add anything to the type enum. It also removes some lookups to map type_id to readable name in debug.
2019-12-19 13:15:10 +01:00
const char *type, uint32_t version, const struct spa_dict *props)
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
{
Cleanup argument names in methods and events interfaces First method argument is object, first event argument is data. Closes !963
2021-09-27 15:09:32 +02:00
struct pw_resource *resource = data;
avoid including private.h Expose some more methods to make it possible to build without including the private.h header, especially in modules.
2020-01-02 13:48:50 +01:00
struct pw_impl_client *client = pw_resource_get_client(resource);
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
struct spa_pod_builder *b;
struct spa_pod_frame f;
uint32_t i, n_items, parent_id;
Make interface types a string This is more in line with wayland and it allows us to create new interfaces in modules without having to add anything to the type enum. It also removes some lookups to map type_id to readable name in debug.
2019-12-19 13:15:10 +01:00
uint32_t type_id;
protocol-native: improve v0 compatibility
2020-02-10 18:15:24 +01:00
const char *str;
type_id = pw_protocol_native0_find_type(client, type);
if (type_id == SPA_ID_INVALID)
return;
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
pw_registry_proxy -> pw_registry
2019-12-11 09:44:48 +01:00
b = pw_protocol_native_begin_resource(resource, PW_REGISTRY_V0_EVENT_GLOBAL, NULL);
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
n_items = props ? props->n_items : 0;
parent_id = 0;
avoid following NULL pointers
2020-05-20 13:41:48 +02:00
if (props) {
treewide: replace strcmp() == 0 with spa_streq() This change is only done in source files for now, header files will be done separately.
2021-05-18 11:36:13 +10:00
if (spa_streq(type, PW_TYPE_INTERFACE_Port)) {
avoid following NULL pointers
2020-05-20 13:41:48 +02:00
if ((str = spa_dict_lookup(props, "node.id")) != NULL)
parent_id = atoi(str);
treewide: replace strcmp() == 0 with spa_streq() This change is only done in source files for now, header files will be done separately.
2021-05-18 11:36:13 +10:00
} else if (spa_streq(type, PW_TYPE_INTERFACE_Node)) {
avoid following NULL pointers
2020-05-20 13:41:48 +02:00
if ((str = spa_dict_lookup(props, "device.id")) != NULL)
parent_id = atoi(str);
treewide: replace strcmp() == 0 with spa_streq() This change is only done in source files for now, header files will be done separately.
2021-05-18 11:36:13 +10:00
} else if (spa_streq(type, PW_TYPE_INTERFACE_Client) ||
spa_streq(type, PW_TYPE_INTERFACE_Device) ||
spa_streq(type, PW_TYPE_INTERFACE_Factory)) {
avoid following NULL pointers
2020-05-20 13:41:48 +02:00
if ((str = spa_dict_lookup(props, "module.id")) != NULL)
parent_id = atoi(str);
}
protocol-native: improve v0 compatibility
2020-02-10 18:15:24 +01:00
}
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
version = 0;
spa_pod_builder_push_struct(b, &f);
spa_pod_builder_add(b,
"i", id,
"i", parent_id,
"i", permissions,
Make interface types a string This is more in line with wayland and it allows us to create new interfaces in modules without having to add anything to the type enum. It also removes some lookups to map type_id to readable name in debug.
2019-12-19 13:15:10 +01:00
"I", type_id,
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
"i", version,
"i", n_items, NULL);
for (i = 0; i < n_items; i++) {
spa_pod_builder_add(b,
"s", props->items[i].key,
"s", props->items[i].value, NULL);
}
spa_pod_builder_pop(b, &f);
pw_protocol_native_end_resource(resource, b);
}
Cleanup argument names in methods and events interfaces First method argument is object, first event argument is data. Closes !963
2021-09-27 15:09:32 +02:00
static void registry_marshal_global_remove(void *data, uint32_t id)
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
{
Cleanup argument names in methods and events interfaces First method argument is object, first event argument is data. Closes !963
2021-09-27 15:09:32 +02:00
struct pw_resource *resource = data;
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
struct spa_pod_builder *b;
pw_registry_proxy -> pw_registry
2019-12-11 09:44:48 +01:00
b = pw_protocol_native_begin_resource(resource, PW_REGISTRY_V0_EVENT_GLOBAL_REMOVE, NULL);
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
spa_pod_builder_add_struct(b, "i", id);
pw_protocol_native_end_resource(resource, b);
}
static int registry_demarshal_bind(void *object, const struct pw_protocol_native_message *msg)
{
struct pw_resource *resource = object;
avoid including private.h Expose some more methods to make it possible to build without including the private.h header, especially in modules.
2020-01-02 13:48:50 +01:00
struct pw_impl_client *client = pw_resource_get_client(resource);
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
struct spa_pod_parser prs;
uint32_t id, version, type, new_id;
Make interface types a string This is more in line with wayland and it allows us to create new interfaces in modules without having to add anything to the type enum. It also removes some lookups to map type_id to readable name in debug.
2019-12-19 13:15:10 +01:00
const char *type_name;
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
spa_pod_parser_init(&prs, msg->data, msg->size);
if (spa_pod_parser_get_struct(&prs,
"i", &id,
"I", &type,
"i", &version,
"i", &new_id) < 0)
return -EINVAL;
avoid including private.h Expose some more methods to make it possible to build without including the private.h header, especially in modules.
2020-01-02 13:48:50 +01:00
type_name = pw_protocol_native0_name_from_v2(client, type);
Make interface types a string This is more in line with wayland and it allows us to create new interfaces in modules without having to add anything to the type enum. It also removes some lookups to map type_id to readable name in debug.
2019-12-19 13:15:10 +01:00
if (type_name == NULL)
return -EINVAL;
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
Make interface types a string This is more in line with wayland and it allows us to create new interfaces in modules without having to add anything to the type enum. It also removes some lookups to map type_id to readable name in debug.
2019-12-19 13:15:10 +01:00
return pw_resource_notify(resource, struct pw_registry_methods, bind, 0, id, type_name, version, new_id);
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
}
Cleanup argument names in methods and events interfaces First method argument is object, first event argument is data. Closes !963
2021-09-27 15:09:32 +02:00
static void module_marshal_info(void *data, const struct pw_module_info *info)
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
{
Cleanup argument names in methods and events interfaces First method argument is object, first event argument is data. Closes !963
2021-09-27 15:09:32 +02:00
struct pw_resource *resource = data;
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
struct spa_pod_builder *b;
struct spa_pod_frame f;
uint32_t i, n_items;
pw_module_proxy -> pw_module
2019-12-11 15:16:30 +01:00
b = pw_protocol_native_begin_resource(resource, PW_MODULE_V0_EVENT_INFO, NULL);
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
n_items = info->props ? info->props->n_items : 0;
spa_pod_builder_push_struct(b, &f);
spa_pod_builder_add(b,
"i", info->id,
"l", info->change_mask,
"s", info->name,
"s", info->filename,
"s", info->args,
"i", n_items, NULL);
for (i = 0; i < n_items; i++) {
spa_pod_builder_add(b,
"s", info->props->items[i].key,
"s", info->props->items[i].value, NULL);
}
spa_pod_builder_pop(b, &f);
pw_protocol_native_end_resource(resource, b);
}
Cleanup argument names in methods and events interfaces First method argument is object, first event argument is data. Closes !963
2021-09-27 15:09:32 +02:00
static void factory_marshal_info(void *data, const struct pw_factory_info *info)
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
{
Cleanup argument names in methods and events interfaces First method argument is object, first event argument is data. Closes !963
2021-09-27 15:09:32 +02:00
struct pw_resource *resource = data;
avoid including private.h Expose some more methods to make it possible to build without including the private.h header, especially in modules.
2020-01-02 13:48:50 +01:00
struct pw_impl_client *client = pw_resource_get_client(resource);
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
struct spa_pod_builder *b;
struct spa_pod_frame f;
uint32_t i, n_items, type, version;
protocol-native: improve v0 compatibility
2020-02-10 18:15:24 +01:00
type = pw_protocol_native0_find_type(client, info->type);
if (type == SPA_ID_INVALID)
return;
pw_factory_proxy -> pw_factory
2019-12-11 15:06:00 +01:00
b = pw_protocol_native_begin_resource(resource, PW_FACTORY_V0_EVENT_INFO, NULL);
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
n_items = info->props ? info->props->n_items : 0;
version = 0;
spa_pod_builder_push_struct(b, &f);
spa_pod_builder_add(b,
"i", info->id,
"l", info->change_mask,
"s", info->name,
"I", type,
"i", version,
"i", n_items, NULL);
for (i = 0; i < n_items; i++) {
spa_pod_builder_add(b,
"s", info->props->items[i].key,
"s", info->props->items[i].value, NULL);
}
spa_pod_builder_pop(b, &f);
pw_protocol_native_end_resource(resource, b);
}
Cleanup argument names in methods and events interfaces First method argument is object, first event argument is data. Closes !963
2021-09-27 15:09:32 +02:00
static void node_marshal_info(void *data, const struct pw_node_info *info)
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
{
Cleanup argument names in methods and events interfaces First method argument is object, first event argument is data. Closes !963
2021-09-27 15:09:32 +02:00
struct pw_resource *resource = data;
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
struct spa_pod_builder *b;
struct spa_pod_frame f;
uint32_t i, n_items;
pw_node_proxy -> pw_node
2019-12-11 15:26:11 +01:00
b = pw_protocol_native_begin_resource(resource, PW_NODE_V0_EVENT_INFO, NULL);
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
n_items = info->props ? info->props->n_items : 0;
spa_pod_builder_push_struct(b, &f);
spa_pod_builder_add(b,
"i", info->id,
"l", info->change_mask,
"s", "node.name",
"i", info->max_input_ports,
"i", info->n_input_ports,
"i", info->max_output_ports,
"i", info->n_output_ports,
"i", info->state,
"s", info->error,
"i", n_items, NULL);
for (i = 0; i < n_items; i++) {
spa_pod_builder_add(b,
"s", info->props->items[i].key,
"s", info->props->items[i].value, NULL);
}
spa_pod_builder_pop(b, &f);
pw_protocol_native_end_resource(resource, b);
}
Cleanup argument names in methods and events interfaces First method argument is object, first event argument is data. Closes !963
2021-09-27 15:09:32 +02:00
static void node_marshal_param(void *data, int seq, uint32_t id, uint32_t index, uint32_t next,
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
const struct spa_pod *param)
{
Cleanup argument names in methods and events interfaces First method argument is object, first event argument is data. Closes !963
2021-09-27 15:09:32 +02:00
struct pw_resource *resource = data;
protocol-native: improve v0 compatibility
2020-02-10 18:15:24 +01:00
struct pw_impl_client *client = pw_resource_get_client(resource);
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
struct spa_pod_builder *b;
protocol-native: improve v0 compatibility
2020-02-10 18:15:24 +01:00
struct spa_pod_frame f;
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
pw_node_proxy -> pw_node
2019-12-11 15:26:11 +01:00
b = pw_protocol_native_begin_resource(resource, PW_NODE_V0_EVENT_PARAM, NULL);
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
protocol-native: improve v0 compatibility
2020-02-10 18:15:24 +01:00
id = pw_protocol_native0_type_to_v2(client, spa_type_param, id),
spa_pod_builder_push_struct(b, &f);
spa_pod_builder_add(b,
"I", id,
"i", index,
"i", next,
NULL);
pw_protocol_native0_pod_to_v2(client, (struct spa_pod *)param, b);
spa_pod_builder_pop(b, &f);
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
pw_protocol_native_end_resource(resource, b);
}
static int node_demarshal_enum_params(void *object, const struct pw_protocol_native_message *msg)
{
struct pw_resource *resource = object;
avoid including private.h Expose some more methods to make it possible to build without including the private.h header, especially in modules.
2020-01-02 13:48:50 +01:00
struct pw_impl_client *client = pw_resource_get_client(resource);
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
struct spa_pod_parser prs;
uint32_t id, index, num;
struct spa_pod *filter;
spa_pod_parser_init(&prs, msg->data, msg->size);
if (spa_pod_parser_get_struct(&prs,
"I", &id,
"i", &index,
"i", &num,
"P", &filter) < 0)
return -EINVAL;
id = pw_protocol_native0_type_from_v2(client, id);
filter = NULL;
pw_node_proxy -> pw_node
2019-12-11 15:26:11 +01:00
return pw_resource_notify(resource, struct pw_node_methods, enum_params, 0,
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
0, id, index, num, filter);
}
Cleanup argument names in methods and events interfaces First method argument is object, first event argument is data. Closes !963
2021-09-27 15:09:32 +02:00
static void port_marshal_info(void *data, const struct pw_port_info *info)
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
{
Cleanup argument names in methods and events interfaces First method argument is object, first event argument is data. Closes !963
2021-09-27 15:09:32 +02:00
struct pw_resource *resource = data;
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
struct spa_pod_builder *b;
struct spa_pod_frame f;
uint32_t i, n_items;
protocol-native: improve v0 compatibility
2020-02-10 18:15:24 +01:00
uint64_t change_mask = 0;
const char *port_name;
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
pw_port_proxy -> pw_port
2019-12-11 15:59:26 +01:00
b = pw_protocol_native_begin_resource(resource, PW_PORT_V0_EVENT_INFO, NULL);
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
n_items = info->props ? info->props->n_items : 0;
protocol-native: improve v0 compatibility
2020-02-10 18:15:24 +01:00
#define PW_PORT_V0_CHANGE_MASK_NAME (1 << 0)
#define PW_PORT_V0_CHANGE_MASK_PROPS (1 << 1)
#define PW_PORT_V0_CHANGE_MASK_ENUM_PARAMS (1 << 2)
change_mask |= PW_PORT_V0_CHANGE_MASK_NAME;
if (info->change_mask & PW_PORT_CHANGE_MASK_PROPS)
change_mask |= PW_PORT_V0_CHANGE_MASK_PROPS;
if (info->change_mask & PW_PORT_CHANGE_MASK_PARAMS)
change_mask |= PW_PORT_V0_CHANGE_MASK_ENUM_PARAMS;
port_name = NULL;
if (info->props != NULL)
port_name = spa_dict_lookup(info->props, "port.name");
if (port_name == NULL)
port_name = "port.name";
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
spa_pod_builder_push_struct(b, &f);
spa_pod_builder_add(b,
"i", info->id,
protocol-native: improve v0 compatibility
2020-02-10 18:15:24 +01:00
"l", change_mask,
"s", port_name,
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
"i", n_items, NULL);
for (i = 0; i < n_items; i++) {
spa_pod_builder_add(b,
"s", info->props->items[i].key,
"s", info->props->items[i].value, NULL);
}
spa_pod_builder_pop(b, &f);
pw_protocol_native_end_resource(resource, b);
}
Cleanup argument names in methods and events interfaces First method argument is object, first event argument is data. Closes !963
2021-09-27 15:09:32 +02:00
static void port_marshal_param(void *data, int seq, uint32_t id, uint32_t index, uint32_t next,
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
const struct spa_pod *param)
{
Cleanup argument names in methods and events interfaces First method argument is object, first event argument is data. Closes !963
2021-09-27 15:09:32 +02:00
struct pw_resource *resource = data;
avoid including private.h Expose some more methods to make it possible to build without including the private.h header, especially in modules.
2020-01-02 13:48:50 +01:00
struct pw_impl_client *client = pw_resource_get_client(resource);
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
struct spa_pod_builder *b;
protocol-native: improve v0 compatibility
2020-02-10 18:15:24 +01:00
struct spa_pod_frame f;
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
pw_port_proxy -> pw_port
2019-12-11 15:59:26 +01:00
b = pw_protocol_native_begin_resource(resource, PW_PORT_V0_EVENT_PARAM, NULL);
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
protocol-native: improve v0 compatibility
2020-02-10 18:15:24 +01:00
id = pw_protocol_native0_type_to_v2(client, spa_type_param, id),
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
protocol-native: improve v0 compatibility
2020-02-10 18:15:24 +01:00
spa_pod_builder_push_struct(b, &f);
spa_pod_builder_add(b,
"I", id,
"i", index,
"i", next,
NULL);
pw_protocol_native0_pod_to_v2(client, (struct spa_pod *)param, b);
spa_pod_builder_pop(b, &f);
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
pw_protocol_native_end_resource(resource, b);
}
static int port_demarshal_enum_params(void *object, const struct pw_protocol_native_message *msg)
{
struct pw_resource *resource = object;
avoid including private.h Expose some more methods to make it possible to build without including the private.h header, especially in modules.
2020-01-02 13:48:50 +01:00
struct pw_impl_client *client = pw_resource_get_client(resource);
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
struct spa_pod_parser prs;
uint32_t id, index, num;
struct spa_pod *filter;
spa_pod_parser_init(&prs, msg->data, msg->size);
if (spa_pod_parser_get_struct(&prs,
"I", &id,
"i", &index,
"i", &num,
"P", &filter) < 0)
return -EINVAL;
id = pw_protocol_native0_type_from_v2(client, id);
filter = NULL;
pw_port_proxy -> pw_port
2019-12-11 15:59:26 +01:00
return pw_resource_notify(resource, struct pw_port_methods, enum_params, 0,
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
0, id, index, num, filter);
}
Cleanup argument names in methods and events interfaces First method argument is object, first event argument is data. Closes !963
2021-09-27 15:09:32 +02:00
static void client_marshal_info(void *data, const struct pw_client_info *info)
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
{
Cleanup argument names in methods and events interfaces First method argument is object, first event argument is data. Closes !963
2021-09-27 15:09:32 +02:00
struct pw_resource *resource = data;
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
struct spa_pod_builder *b;
struct spa_pod_frame f;
uint32_t i, n_items;
pw_client_proxy -> pw_client
2019-12-11 14:53:39 +01:00
b = pw_protocol_native_begin_resource(resource, PW_CLIENT_V0_EVENT_INFO, NULL);
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
n_items = info->props ? info->props->n_items : 0;
spa_pod_builder_push_struct(b, &f);
spa_pod_builder_add(b,
"i", info->id,
"l", info->change_mask,
"i", n_items, NULL);
for (i = 0; i < n_items; i++) {
spa_pod_builder_add(b,
"s", info->props->items[i].key,
"s", info->props->items[i].value, NULL);
}
spa_pod_builder_pop(b, &f);
pw_protocol_native_end_resource(resource, b);
}
Cleanup argument names in methods and events interfaces First method argument is object, first event argument is data. Closes !963
2021-09-27 15:09:32 +02:00
static void client_marshal_permissions(void *data, uint32_t index, uint32_t n_permissions,
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
const struct pw_permission *permissions)
{
}
Cleanup argument names in methods and events interfaces First method argument is object, first event argument is data. Closes !963
2021-09-27 15:09:32 +02:00
static void link_marshal_info(void *data, const struct pw_link_info *info)
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
{
Cleanup argument names in methods and events interfaces First method argument is object, first event argument is data. Closes !963
2021-09-27 15:09:32 +02:00
struct pw_resource *resource = data;
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
struct spa_pod_builder *b;
struct spa_pod_frame f;
uint32_t i, n_items;
pw_link_proxy -> pw_link
2019-12-11 15:11:56 +01:00
b = pw_protocol_native_begin_resource(resource, PW_LINK_V0_EVENT_INFO, NULL);
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
n_items = info->props ? info->props->n_items : 0;
spa_pod_builder_push_struct(b, &f);
spa_pod_builder_add(b,
"i", info->id,
"l", info->change_mask,
"i", info->output_node_id,
"i", info->output_port_id,
"i", info->input_node_id,
"i", info->input_port_id,
"P", info->format,
"i", n_items, NULL);
for (i = 0; i < n_items; i++) {
spa_pod_builder_add(b,
"s", info->props->items[i].key,
"s", info->props->items[i].value, NULL);
}
spa_pod_builder_pop(b, &f);
pw_protocol_native_end_resource(resource, b);
}
rename core_proxy -> core Rename core_proxy to core and move the introspect and interface contents to core.h In an effort to promote the proxy API.
2019-12-11 07:46:59 +01:00
static const struct pw_protocol_native_demarshal pw_protocol_native_core_method_demarshal[PW_CORE_V0_METHOD_NUM] = {
[PW_CORE_V0_METHOD_HELLO] = { &core_demarshal_hello, 0, },
[PW_CORE_V0_METHOD_UPDATE_TYPES] = { &core_demarshal_update_types_server, 0, },
[PW_CORE_V0_METHOD_SYNC] = { &core_demarshal_sync, 0, },
[PW_CORE_V0_METHOD_GET_REGISTRY] = { &core_demarshal_get_registry, 0, },
[PW_CORE_V0_METHOD_CLIENT_UPDATE] = { &core_demarshal_client_update, 0, },
[PW_CORE_V0_METHOD_PERMISSIONS] = { &core_demarshal_permissions, 0, },
[PW_CORE_V0_METHOD_CREATE_OBJECT] = { &core_demarshal_create_object, 0, PW_PROTOCOL_NATIVE_FLAG_REMAP, },
[PW_CORE_V0_METHOD_DESTROY] = { &core_demarshal_destroy, 0, }
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
};
rename core_proxy -> core Rename core_proxy to core and move the introspect and interface contents to core.h In an effort to promote the proxy API.
2019-12-11 07:46:59 +01:00
static const struct pw_core_events pw_protocol_native_core_event_marshal = {
PW_VERSION_CORE_EVENTS,
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
.info = &core_marshal_info,
.done = &core_marshal_done,
.error = &core_marshal_error,
.remove_id = &core_marshal_remove_id,
};
static const struct pw_protocol_marshal pw_protocol_native_core_marshal = {
PW_TYPE_INTERFACE_Core,
PW_VERSION_CORE_V0,
rename core_proxy -> core Rename core_proxy to core and move the introspect and interface contents to core.h In an effort to promote the proxy API.
2019-12-11 07:46:59 +01:00
PW_CORE_V0_METHOD_NUM,
PW_CORE_EVENT_NUM,
protocol: add flags to protocol marshal Allows us to to select marshallers for implementation and binding.
2019-11-03 10:21:57 +01:00
0,
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
NULL,
pw_protocol_native_core_method_demarshal,
&pw_protocol_native_core_event_marshal,
NULL
};
static const struct pw_protocol_native_demarshal pw_protocol_native_registry_method_demarshal[] = {
pw_registry_proxy -> pw_registry
2019-12-11 09:44:48 +01:00
[PW_REGISTRY_V0_METHOD_BIND] = { &registry_demarshal_bind, 0, PW_PROTOCOL_NATIVE_FLAG_REMAP, },
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
};
pw_registry_proxy -> pw_registry
2019-12-11 09:44:48 +01:00
static const struct pw_registry_events pw_protocol_native_registry_event_marshal = {
PW_VERSION_REGISTRY_EVENTS,
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
.global = &registry_marshal_global,
.global_remove = &registry_marshal_global_remove,
};
static const struct pw_protocol_marshal pw_protocol_native_registry_marshal = {
PW_TYPE_INTERFACE_Registry,
PW_VERSION_REGISTRY_V0,
pw_registry_proxy -> pw_registry
2019-12-11 09:44:48 +01:00
PW_REGISTRY_V0_METHOD_NUM,
PW_REGISTRY_EVENT_NUM,
protocol: add flags to protocol marshal Allows us to to select marshallers for implementation and binding.
2019-11-03 10:21:57 +01:00
0,
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
NULL,
pw_protocol_native_registry_method_demarshal,
&pw_protocol_native_registry_event_marshal,
NULL
};
pw_module_proxy -> pw_module
2019-12-11 15:16:30 +01:00
static const struct pw_module_events pw_protocol_native_module_event_marshal = {
PW_VERSION_MODULE_EVENTS,
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
.info = &module_marshal_info,
};
static const struct pw_protocol_marshal pw_protocol_native_module_marshal = {
PW_TYPE_INTERFACE_Module,
PW_VERSION_MODULE_V0,
0,
pw_module_proxy -> pw_module
2019-12-11 15:16:30 +01:00
PW_MODULE_EVENT_NUM,
protocol: add flags to protocol marshal Allows us to to select marshallers for implementation and binding.
2019-11-03 10:21:57 +01:00
0,
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
NULL, NULL,
&pw_protocol_native_module_event_marshal,
NULL
};
pw_factory_proxy -> pw_factory
2019-12-11 15:06:00 +01:00
static const struct pw_factory_events pw_protocol_native_factory_event_marshal = {
PW_VERSION_FACTORY_EVENTS,
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
.info = &factory_marshal_info,
};
static const struct pw_protocol_marshal pw_protocol_native_factory_marshal = {
PW_TYPE_INTERFACE_Factory,
PW_VERSION_FACTORY_V0,
0,
pw_factory_proxy -> pw_factory
2019-12-11 15:06:00 +01:00
PW_FACTORY_EVENT_NUM,
protocol: add flags to protocol marshal Allows us to to select marshallers for implementation and binding.
2019-11-03 10:21:57 +01:00
0,
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
NULL, NULL,
&pw_protocol_native_factory_event_marshal,
NULL,
};
static const struct pw_protocol_native_demarshal pw_protocol_native_node_method_demarshal[] = {
pw_node_proxy -> pw_node
2019-12-11 15:26:11 +01:00
[PW_NODE_V0_METHOD_ENUM_PARAMS] = { &node_demarshal_enum_params, 0, PW_PROTOCOL_NATIVE_FLAG_REMAP, },
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
};
pw_node_proxy -> pw_node
2019-12-11 15:26:11 +01:00
static const struct pw_node_events pw_protocol_native_node_event_marshal = {
PW_VERSION_NODE_EVENTS,
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
.info = &node_marshal_info,
.param = &node_marshal_param,
};
static const struct pw_protocol_marshal pw_protocol_native_node_marshal = {
PW_TYPE_INTERFACE_Node,
PW_VERSION_NODE_V0,
pw_node_proxy -> pw_node
2019-12-11 15:26:11 +01:00
PW_NODE_V0_METHOD_NUM,
PW_NODE_EVENT_NUM,
protocol: add flags to protocol marshal Allows us to to select marshallers for implementation and binding.
2019-11-03 10:21:57 +01:00
0,
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
NULL,
pw_protocol_native_node_method_demarshal,
&pw_protocol_native_node_event_marshal,
NULL
};
static const struct pw_protocol_native_demarshal pw_protocol_native_port_method_demarshal[] = {
pw_port_proxy -> pw_port
2019-12-11 15:59:26 +01:00
[PW_PORT_V0_METHOD_ENUM_PARAMS] = { &port_demarshal_enum_params, 0, PW_PROTOCOL_NATIVE_FLAG_REMAP, },
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
};
pw_port_proxy -> pw_port
2019-12-11 15:59:26 +01:00
static const struct pw_port_events pw_protocol_native_port_event_marshal = {
PW_VERSION_PORT_EVENTS,
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
.info = &port_marshal_info,
.param = &port_marshal_param,
};
static const struct pw_protocol_marshal pw_protocol_native_port_marshal = {
PW_TYPE_INTERFACE_Port,
PW_VERSION_PORT_V0,
pw_port_proxy -> pw_port
2019-12-11 15:59:26 +01:00
PW_PORT_V0_METHOD_NUM,
PW_PORT_EVENT_NUM,
protocol: add flags to protocol marshal Allows us to to select marshallers for implementation and binding.
2019-11-03 10:21:57 +01:00
0,
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
NULL,
pw_protocol_native_port_method_demarshal,
&pw_protocol_native_port_event_marshal,
NULL
};
pw_client_proxy -> pw_client
2019-12-11 14:53:39 +01:00
static const struct pw_client_events pw_protocol_native_client_event_marshal = {
PW_VERSION_CLIENT_EVENTS,
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
.info = &client_marshal_info,
.permissions = &client_marshal_permissions,
};
static const struct pw_protocol_marshal pw_protocol_native_client_marshal = {
PW_TYPE_INTERFACE_Client,
PW_VERSION_CLIENT_V0,
0,
pw_client_proxy -> pw_client
2019-12-11 14:53:39 +01:00
PW_CLIENT_EVENT_NUM,
protocol: add flags to protocol marshal Allows us to to select marshallers for implementation and binding.
2019-11-03 10:21:57 +01:00
0,
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
NULL, NULL,
&pw_protocol_native_client_event_marshal,
NULL,
};
pw_link_proxy -> pw_link
2019-12-11 15:11:56 +01:00
static const struct pw_link_events pw_protocol_native_link_event_marshal = {
PW_VERSION_LINK_EVENTS,
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
.info = &link_marshal_info,
};
static const struct pw_protocol_marshal pw_protocol_native_link_marshal = {
PW_TYPE_INTERFACE_Link,
PW_VERSION_LINK_V0,
0,
pw_link_proxy -> pw_link
2019-12-11 15:11:56 +01:00
PW_LINK_EVENT_NUM,
protocol: add flags to protocol marshal Allows us to to select marshallers for implementation and binding.
2019-11-03 10:21:57 +01:00
0,
protocol: add v0 compatibility For flatpaks we need to be able to support older v0 protocol clients. To handle this we have: - the connection detects an old client when it receives the first message. It can do this by checking the sequence number, on old versions it contains the message size and is never 0, on new clients the sequence number is 0. - We add a new signal at the start of the connection with the detected version number. This installs the right version of the core proxy. We also move the binding of the client until the hello message is received. This way we can have a new client connect (portal), hand over the connection to an old client, which then removes the client binding again in the hello request with a v0 version. There are some changes to the passing of fds in v0 vs v3 which need to investigated some more. - bump version of our interfaces to 3. This makes it possible to have v0 and v3 protocol marshal functions. - Add version number in the proxy. This is mostly automatically done internally based on the version numbers the library is compiled with. Where the version number was in the API before, it is now actually used to look up the right protocol marshal functions. For Proxies there is usually just 1 version, the current one. It is the server that will support different versions. - Add v0 compat marshal functions to convert from and to v0 format. This has some complications. v0 has a type map it keeps in sync with the server. For this we have a static type map with mappings to our own v3 types. Pods are mostly the same except for objects that used to have arbitrary pods in v0 vs spa_pod_prop in v3. Also convert between v0 spa_pod_prop and v3 spa_pod_choice. Formats and commands are also slightly different so handle those mappings as well. We only have marshal functions for the server side (resource) v0 functions. - Add v0 compatible client-node again. It's a bit tricky to map, v0 client-node basically lets the server to the mixing and teeing and just does the processing of the internal node.
2019-10-08 22:52:25 +02:00
NULL, NULL,
&pw_protocol_native_link_event_marshal,
NULL
};
void pw_protocol_native0_init(struct pw_protocol *protocol)
{
pw_protocol_add_marshal(protocol, &pw_protocol_native_core_marshal);
pw_protocol_add_marshal(protocol, &pw_protocol_native_registry_marshal);
pw_protocol_add_marshal(protocol, &pw_protocol_native_module_marshal);
pw_protocol_add_marshal(protocol, &pw_protocol_native_node_marshal);
pw_protocol_add_marshal(protocol, &pw_protocol_native_port_marshal);
pw_protocol_add_marshal(protocol, &pw_protocol_native_factory_marshal);
pw_protocol_add_marshal(protocol, &pw_protocol_native_client_marshal);
pw_protocol_add_marshal(protocol, &pw_protocol_native_link_marshal);
}
Reference in a new issue Copy permalink