mpv/libmpv/render.h

760 lines
34 KiB
C
Raw Permalink Normal View History

/* Copyright (C) 2018 the mpv developers
*
* Permission to use, copy, modify, and/or distribute this software for any
* purpose with or without fee is hereby granted, provided that the above
* copyright notice and this permission notice appear in all copies.
*
* THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
* WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
* MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
* ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
* WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
* ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
* OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
*/
#ifndef MPV_CLIENT_API_RENDER_H_
#define MPV_CLIENT_API_RENDER_H_
#include "client.h"
#ifdef __cplusplus
extern "C" {
#endif
/**
* Overview
* --------
*
* This API can be used to make mpv render using supported graphic APIs (such
* as OpenGL). It can be used to handle video display.
*
* The renderer needs to be created with mpv_render_context_create() before
* you start playback (or otherwise cause a VO to be created). Then (with most
* backends) mpv_render_context_render() can be used to explicitly render the
* current video frame. Use mpv_render_context_set_update_callback() to get
* notified when there is a new frame to draw.
*
* Preferably rendering should be done in a separate thread. If you call
* normal libmpv API functions on the renderer thread, deadlocks can result
* (these are made non-fatal with timeouts, but user experience will obviously
* suffer). See "Threading" section below.
*
* You can output and embed video without this API by setting the mpv "wid"
* option to a native window handle (see "Embedding the video window" section
* in the client.h header). In general, using the render API is recommended,
* because window embedding can cause various issues, especially with GUI
* toolkits and certain platforms.
*
* Supported backends
* ------------------
*
* OpenGL: via MPV_RENDER_API_TYPE_OPENGL, see render_gl.h header.
* Software: via MPV_RENDER_API_TYPE_SW, see section "Software renderer"
*
* Threading
* ---------
*
* You are recommended to do rendering on a separate thread than normal libmpv
* use.
*
* The mpv_render_* functions can be called from any thread, under the
* following conditions:
* - only one of the mpv_render_* functions can be called at the same time
* (unless they belong to different mpv cores created by mpv_create())
* - never can be called from within the callbacks set with
* mpv_set_wakeup_callback() or mpv_render_context_set_update_callback()
* - if the OpenGL backend is used, for all functions the OpenGL context
* must be "current" in the calling thread, and it must be the same OpenGL
* context as the mpv_render_context was created with. Otherwise, undefined
* behavior will occur.
* - the thread does not call libmpv API functions other than the mpv_render_*
* functions, except APIs which are declared as safe (see below). Likewise,
* there must be no lock or wait dependency from the render thread to a
* thread using other libmpv functions. Basically, the situation that your
* render thread waits for a "not safe" libmpv API function to return must
* not happen. If you ignore this requirement, deadlocks can happen, which
* are made non-fatal with timeouts; then playback quality will be degraded,
* and the message
* mpv_render_context_render() not being called or stuck.
* is logged. If you set MPV_RENDER_PARAM_ADVANCED_CONTROL, you promise that
* this won't happen, and must absolutely guarantee it, or a real deadlock
* will freeze the mpv core thread forever.
*
* libmpv functions which are safe to call from a render thread are:
* - functions marked with "Safe to be called from mpv render API threads."
* - client.h functions which don't have an explicit or implicit mpv_handle
* parameter
* - mpv_render_* functions; but only for the same mpv_render_context pointer.
* If the pointer is different, mpv_render_context_free() is not safe. (The
* reason is that if MPV_RENDER_PARAM_ADVANCED_CONTROL is set, it may have
* to process still queued requests from the core, which it can do only for
* the current context, while requests for other contexts would deadlock.
* Also, it may have to wait and block for the core to terminate the video
* chain to make sure no resources are used after context destruction.)
* - if the mpv_handle parameter refers to a different mpv core than the one
* you're rendering for (very obscure, but allowed)
*
client API: fix potential deadlock problems by throwing more shit at it The render API (vo_libmpv) had potential deadlock problems with MPV_RENDER_PARAM_ADVANCED_CONTROL. This required vd-lavc-dr to be enabled (the default). I never observed these deadlocks in the wild (doesn't mean they didn't happen), although I could specifically provoke them with some code changes. The problem was mostly about DR (direct rendering, letting the video decoder write to OpenGL buffer memory). Allocating/freeing a DR image needs to be done on the OpenGL thread, even though _lots_ of threads are involved with handling images. Freeing a DR image is a special case that can happen any time. dr_helper.c does most of the evil magic of achieving this. Unfortunately, there was a (sort of) circular lock dependency: freeing an image while certain internal locks are held would trigger the user's context update callback, which in turn would call mpv_render_context_update(), which processed all pending free requests, and then acquire an internal lock - which the caller might not release until a further DR image could be freed. "Solve" this by making freeing DR images asynchronous. This is slightly risky, but actually not much. The DR images will be free'd eventually. The biggest disadvantage is probably that debugging might get trickier. Any solution to this problem will probably add images to free to some sort of queue, and then process it later. I considered making this more explicit (so there'd be a point where the caller forcibly waits for all queued items to be free'd), but discarded these ideas as this probably would only increase complexity. Another consequence is that freeing DR images on the GL thread is not synchronous anymore. Instead, it mpv_render_context_update() will do it with a delay. This seems roundabout, but doesn't actually change anything, and avoids additional code. This also fixes that the render API required the render API user to remain on the same thread, even though this wasn't documented. As such, it was a bug. OpenGL essentially forces you to do all GL usage on a single thread, but in theory the API user could for example move the GL context to another thread. The API bump is because I think you can't make enough noise about this. Since we don't backport fixes to old versions, I'm specifically stating that old versions are broken, and I'm supplying workarounds. Internally, dr_helper_create() does not use pthread_self() anymore, thus the vo.c change. I think it's better to make binding to the current thread as explicit as possible. Of course it's not sure that this fixes all deadlocks (probably not).
2019-09-26 14:07:33 +02:00
* Note about old libmpv version:
*
* Before API version 1.105 (basically in mpv 0.29.x), simply enabling
* MPV_RENDER_PARAM_ADVANCED_CONTROL could cause deadlock issues. This can
* be worked around by setting the "vd-lavc-dr" option to "no".
* In addition, you were required to call all mpv_render*() API functions
* from the same thread on which mpv_render_context_create() was originally
* run (for the same the mpv_render_context). Not honoring it led to UB
2023-10-21 04:55:41 +02:00
* (deadlocks, use of invalid mp_thread handles), even if you moved your GL
client API: fix potential deadlock problems by throwing more shit at it The render API (vo_libmpv) had potential deadlock problems with MPV_RENDER_PARAM_ADVANCED_CONTROL. This required vd-lavc-dr to be enabled (the default). I never observed these deadlocks in the wild (doesn't mean they didn't happen), although I could specifically provoke them with some code changes. The problem was mostly about DR (direct rendering, letting the video decoder write to OpenGL buffer memory). Allocating/freeing a DR image needs to be done on the OpenGL thread, even though _lots_ of threads are involved with handling images. Freeing a DR image is a special case that can happen any time. dr_helper.c does most of the evil magic of achieving this. Unfortunately, there was a (sort of) circular lock dependency: freeing an image while certain internal locks are held would trigger the user's context update callback, which in turn would call mpv_render_context_update(), which processed all pending free requests, and then acquire an internal lock - which the caller might not release until a further DR image could be freed. "Solve" this by making freeing DR images asynchronous. This is slightly risky, but actually not much. The DR images will be free'd eventually. The biggest disadvantage is probably that debugging might get trickier. Any solution to this problem will probably add images to free to some sort of queue, and then process it later. I considered making this more explicit (so there'd be a point where the caller forcibly waits for all queued items to be free'd), but discarded these ideas as this probably would only increase complexity. Another consequence is that freeing DR images on the GL thread is not synchronous anymore. Instead, it mpv_render_context_update() will do it with a delay. This seems roundabout, but doesn't actually change anything, and avoids additional code. This also fixes that the render API required the render API user to remain on the same thread, even though this wasn't documented. As such, it was a bug. OpenGL essentially forces you to do all GL usage on a single thread, but in theory the API user could for example move the GL context to another thread. The API bump is because I think you can't make enough noise about this. Since we don't backport fixes to old versions, I'm specifically stating that old versions are broken, and I'm supplying workarounds. Internally, dr_helper_create() does not use pthread_self() anymore, thus the vo.c change. I think it's better to make binding to the current thread as explicit as possible. Of course it's not sure that this fixes all deadlocks (probably not).
2019-09-26 14:07:33 +02:00
* context to a different thread correctly.
* These problems were addressed in API version 1.105 (mpv 0.30.0).
*
* Context and handle lifecycle
* ----------------------------
*
* Video initialization will fail if the render context was not initialized yet
* (with mpv_render_context_create()), or it will revert to a VO that creates
* its own window.
*
* Currently, there can be only 1 mpv_render_context at a time per mpv core.
*
* Calling mpv_render_context_free() while a VO is using the render context is
* active will disable video.
*
* You must free the context with mpv_render_context_free() before the mpv core
* is destroyed. If this doesn't happen, undefined behavior will result.
*
* Software renderer
* -----------------
*
* MPV_RENDER_API_TYPE_SW provides an extremely simple (but slow) renderer to
* memory surfaces. You probably don't want to use this. Use other render API
* types, or other methods of video embedding.
*
* Use mpv_render_context_create() with MPV_RENDER_PARAM_API_TYPE set to
* MPV_RENDER_API_TYPE_SW.
*
* Call mpv_render_context_render() with various MPV_RENDER_PARAM_SW_* fields
* to render the video frame to an in-memory surface. The following fields are
* required: MPV_RENDER_PARAM_SW_SIZE, MPV_RENDER_PARAM_SW_FORMAT,
* MPV_RENDER_PARAM_SW_STRIDE, MPV_RENDER_PARAM_SW_POINTER.
*
* This method of rendering is very slow, because everything, including color
* conversion, scaling, and OSD rendering, is done on the CPU, single-threaded.
* In particular, large video or display sizes, as well as presence of OSD or
* subtitles can make it too slow for realtime. As with other software rendering
* VOs, setting "sw-fast" may help. Enabling or disabling zimg may help,
* depending on the platform.
*
* In addition, certain multimedia job creation measures like HDR may not work
* properly, and will have to be manually handled by for example inserting
* filters.
*
* This API is not really suitable to extract individual frames from video etc.
* (basically non-playback uses) - there are better libraries for this. It can
* be used this way, but it may be clunky and tricky.
*
* Further notes:
* - MPV_RENDER_PARAM_FLIP_Y is currently ignored (unsupported)
* - MPV_RENDER_PARAM_DEPTH is ignored (meaningless)
*/
/**
* Opaque context, returned by mpv_render_context_create().
*/
typedef struct mpv_render_context mpv_render_context;
/**
* Parameters for mpv_render_param (which is used in a few places such as
* mpv_render_context_create().
*
* Also see mpv_render_param for conventions and how to use it.
*/
typedef enum mpv_render_param_type {
/**
* Not a valid value, but also used to terminate a params array. Its value
* is always guaranteed to be 0 (even if the ABI changes in the future).
*/
MPV_RENDER_PARAM_INVALID = 0,
/**
* The render API to use. Valid for mpv_render_context_create().
*
* Type: char*
*
* Defined APIs:
*
* MPV_RENDER_API_TYPE_OPENGL:
* OpenGL desktop 2.1 or later (preferably core profile compatible to
* OpenGL 3.2), or OpenGLES 2.0 or later.
* Providing MPV_RENDER_PARAM_OPENGL_INIT_PARAMS is required.
* It is expected that an OpenGL context is valid and "current" when
* calling mpv_render_* functions (unless specified otherwise). It
* must be the same context for the same mpv_render_context.
*/
MPV_RENDER_PARAM_API_TYPE = 1,
/**
* Required parameters for initializing the OpenGL renderer. Valid for
* mpv_render_context_create().
* Type: mpv_opengl_init_params*
*/
MPV_RENDER_PARAM_OPENGL_INIT_PARAMS = 2,
/**
* Describes a GL render target. Valid for mpv_render_context_render().
* Type: mpv_opengl_fbo*
*/
MPV_RENDER_PARAM_OPENGL_FBO = 3,
/**
* Control flipped rendering. Valid for mpv_render_context_render().
* Type: int*
* If the value is set to 0, render normally. Otherwise, render it flipped,
* which is needed e.g. when rendering to an OpenGL default framebuffer
* (which has a flipped coordinate system).
*/
MPV_RENDER_PARAM_FLIP_Y = 4,
/**
* Control surface depth. Valid for mpv_render_context_render().
* Type: int*
* This implies the depth of the surface passed to the render function in
* bits per channel. If omitted or set to 0, the renderer will assume 8.
* Typically used to control dithering.
*/
MPV_RENDER_PARAM_DEPTH = 5,
/**
* ICC profile blob. Valid for mpv_render_context_set_parameter().
* Type: mpv_byte_array*
* Set an ICC profile for use with the "icc-profile-auto" option. (If the
* option is not enabled, the ICC data will not be used.)
*/
MPV_RENDER_PARAM_ICC_PROFILE = 6,
/**
* Ambient light in lux. Valid for mpv_render_context_set_parameter().
* Type: int*
* This can be used for automatic gamma correction.
*/
MPV_RENDER_PARAM_AMBIENT_LIGHT = 7,
client API: add a new way to pass X11 Display etc. to render API Hardware decoding things often need access to additional handles from the windowing system, such as the X11 or Wayland display when using vaapi. The opengl-cb had nothing dedicated for this, and used the weird GL_MP_MPGetNativeDisplay GL extension (which was mpv specific and not officially registered with OpenGL). This was awkward, and a pain due to having to emulate GL context behavior (like needing a TLS variable to store context for the pseudo GL extension function). In addition (and not inherently due to this), we could pass only one resource from mpv builtin context backends to hwdecs. It was also all GL specific. Replace this with a newer mechanism. It works for all RA backends, not just GL. the API user can explicitly pass the objects at init time via mpv_render_context_create(). Multiple resources are naturally possible. The API uses MPV_RENDER_PARAM_* defines, but internally we use strings. This is done for 2 reasons: 1. trying to leave libmpv and internal mechanisms decoupled, 2. not having to add public API for some of the internal resource types (especially D3D/GL interop stuff). To remain sane, drop support for obscure half-working opengl-cb things, like the DRM interop (was missing necessary things), the RPI window thing (nobody used it), and obscure D3D interop things (not needed with ANGLE, others were undocumented). In order not to break ABI and the C API, we don't remove the associated structs from opengl_cb.h. The parts which are still needed (in particular DRM interop) needs to be ported to the render API.
2018-03-22 17:05:01 +01:00
/**
* X11 Display, sometimes used for hwdec. Valid for
* mpv_render_context_create(). The Display must stay valid for the lifetime
* of the mpv_render_context.
client API: add a new way to pass X11 Display etc. to render API Hardware decoding things often need access to additional handles from the windowing system, such as the X11 or Wayland display when using vaapi. The opengl-cb had nothing dedicated for this, and used the weird GL_MP_MPGetNativeDisplay GL extension (which was mpv specific and not officially registered with OpenGL). This was awkward, and a pain due to having to emulate GL context behavior (like needing a TLS variable to store context for the pseudo GL extension function). In addition (and not inherently due to this), we could pass only one resource from mpv builtin context backends to hwdecs. It was also all GL specific. Replace this with a newer mechanism. It works for all RA backends, not just GL. the API user can explicitly pass the objects at init time via mpv_render_context_create(). Multiple resources are naturally possible. The API uses MPV_RENDER_PARAM_* defines, but internally we use strings. This is done for 2 reasons: 1. trying to leave libmpv and internal mechanisms decoupled, 2. not having to add public API for some of the internal resource types (especially D3D/GL interop stuff). To remain sane, drop support for obscure half-working opengl-cb things, like the DRM interop (was missing necessary things), the RPI window thing (nobody used it), and obscure D3D interop things (not needed with ANGLE, others were undocumented). In order not to break ABI and the C API, we don't remove the associated structs from opengl_cb.h. The parts which are still needed (in particular DRM interop) needs to be ported to the render API.
2018-03-22 17:05:01 +01:00
* Type: Display*
*/
MPV_RENDER_PARAM_X11_DISPLAY = 8,
/**
* Wayland display, sometimes used for hwdec. Valid for
* mpv_render_context_create(). The wl_display must stay valid for the
* lifetime of the mpv_render_context.
client API: add a new way to pass X11 Display etc. to render API Hardware decoding things often need access to additional handles from the windowing system, such as the X11 or Wayland display when using vaapi. The opengl-cb had nothing dedicated for this, and used the weird GL_MP_MPGetNativeDisplay GL extension (which was mpv specific and not officially registered with OpenGL). This was awkward, and a pain due to having to emulate GL context behavior (like needing a TLS variable to store context for the pseudo GL extension function). In addition (and not inherently due to this), we could pass only one resource from mpv builtin context backends to hwdecs. It was also all GL specific. Replace this with a newer mechanism. It works for all RA backends, not just GL. the API user can explicitly pass the objects at init time via mpv_render_context_create(). Multiple resources are naturally possible. The API uses MPV_RENDER_PARAM_* defines, but internally we use strings. This is done for 2 reasons: 1. trying to leave libmpv and internal mechanisms decoupled, 2. not having to add public API for some of the internal resource types (especially D3D/GL interop stuff). To remain sane, drop support for obscure half-working opengl-cb things, like the DRM interop (was missing necessary things), the RPI window thing (nobody used it), and obscure D3D interop things (not needed with ANGLE, others were undocumented). In order not to break ABI and the C API, we don't remove the associated structs from opengl_cb.h. The parts which are still needed (in particular DRM interop) needs to be ported to the render API.
2018-03-22 17:05:01 +01:00
* Type: struct wl_display*
*/
MPV_RENDER_PARAM_WL_DISPLAY = 9,
/**
* Better control about rendering and enabling some advanced features. Valid
* for mpv_render_context_create().
*
* This conflates multiple requirements the API user promises to abide if
* this option is enabled:
*
* - The API user's render thread, which is calling the mpv_render_*()
* functions, never waits for the core. Otherwise deadlocks can happen.
* See "Threading" section.
* - The callback set with mpv_render_context_set_update_callback() can now
* be called even if there is no new frame. The API user should call the
* mpv_render_context_update() function, and interpret the return value
* for whether a new frame should be rendered.
* - Correct functionality is impossible if the update callback is not set,
* or not set soon enough after mpv_render_context_create() (the core can
* block while waiting for you to call mpv_render_context_update(), and
* if the update callback is not correctly set, it will deadlock, or
* block for too long).
*
* In general, setting this option will enable the following features (and
* possibly more):
*
* - "Direct rendering", which means the player decodes directly to a
* texture, which saves a copy per video frame ("vd-lavc-dr" option
* needs to be enabled, and the rendering backend as well as the
* underlying GPU API/driver needs to have support for it).
* - Rendering screenshots with the GPU API if supported by the backend
* (instead of using a suboptimal software fallback via libswscale).
*
* Warning: do not just add this without reading the "Threading" section
* above, and then wondering that deadlocks happen. The
* requirements are tricky. But also note that even if advanced
* control is disabled, not adhering to the rules will lead to
* playback problems. Enabling advanced controls simply makes
* violating these rules fatal.
*
* Type: int*: 0 for disable (default), 1 for enable
*/
MPV_RENDER_PARAM_ADVANCED_CONTROL = 10,
/**
* Return information about the next frame to render. Valid for
* mpv_render_context_get_info().
*
* Type: mpv_render_frame_info*
*
* It strictly returns information about the _next_ frame. The implication
* is that e.g. mpv_render_context_update()'s return value will have
* MPV_RENDER_UPDATE_FRAME set, and the user is supposed to call
* mpv_render_context_render(). If there is no next frame, then the
* return value will have is_valid set to 0.
*/
MPV_RENDER_PARAM_NEXT_FRAME_INFO = 11,
/**
* Enable or disable video timing. Valid for mpv_render_context_render().
*
* Type: int*: 0 for disable, 1 for enable (default)
*
* When video is timed to audio, the player attempts to render video a bit
* ahead, and then do a blocking wait until the target display time is
* reached. This blocks mpv_render_context_render() for up to the amount
* specified with the "video-timing-offset" global option. You can set
* this parameter to 0 to disable this kind of waiting. If you do, it's
* recommended to use the target time value in mpv_render_frame_info to
* wait yourself, or to set the "video-timing-offset" to 0 instead.
*
* Disabling this without doing anything in addition will result in A/V sync
* being slightly off.
*/
MPV_RENDER_PARAM_BLOCK_FOR_TARGET_TIME = 12,
/**
* Use to skip rendering in mpv_render_context_render().
*
* Type: int*: 0 for rendering (default), 1 for skipping
*
* If this is set, you don't need to pass a target surface to the render
* function (and if you do, it's completely ignored). This can still call
* into the lower level APIs (i.e. if you use OpenGL, the OpenGL context
* must be set).
*
* Be aware that the render API will consider this frame as having been
* rendered. All other normal rules also apply, for example about whether
* you have to call mpv_render_context_report_swap(). It also does timing
* in the same way.
*/
MPV_RENDER_PARAM_SKIP_RENDERING = 13,
/**
* Deprecated. Not supported. Use MPV_RENDER_PARAM_DRM_DISPLAY_V2 instead.
* Type : struct mpv_opengl_drm_params*
*/
MPV_RENDER_PARAM_DRM_DISPLAY = 14,
/**
* DRM draw surface size, contains draw surface dimensions.
* Valid for mpv_render_context_create().
* Type : struct mpv_opengl_drm_draw_surface_size*
*/
MPV_RENDER_PARAM_DRM_DRAW_SURFACE_SIZE = 15,
/**
* DRM display, contains drm display handles.
* Valid for mpv_render_context_create().
* Type : struct mpv_opengl_drm_params_v2*
*/
MPV_RENDER_PARAM_DRM_DISPLAY_V2 = 16,
/**
* MPV_RENDER_API_TYPE_SW only: rendering target surface size, mandatory.
* Valid for MPV_RENDER_API_TYPE_SW & mpv_render_context_render().
* Type: int[2] (e.g.: int s[2] = {w, h}; param.data = &s[0];)
*
* The video frame is transformed as with other VOs. Typically, this means
* the video gets scaled and black bars are added if the video size or
* aspect ratio mismatches with the target size.
*/
MPV_RENDER_PARAM_SW_SIZE = 17,
/**
* MPV_RENDER_API_TYPE_SW only: rendering target surface pixel format,
* mandatory.
* Valid for MPV_RENDER_API_TYPE_SW & mpv_render_context_render().
* Type: char* (e.g.: char *f = "rgb0"; param.data = f;)
*
* Valid values are:
* "rgb0", "bgr0", "0bgr", "0rgb"
* 4 bytes per pixel RGB, 1 byte (8 bit) per component, component bytes
* with increasing address from left to right (e.g. "rgb0" has r at
* address 0), the "0" component contains uninitialized garbage (often
* the value 0, but not necessarily; the bad naming is inherited from
* FFmpeg)
* Pixel alignment size: 4 bytes
* "rgb24"
* 3 bytes per pixel RGB. This is strongly discouraged because it is
* very slow.
* Pixel alignment size: 1 bytes
* other
* The API may accept other pixel formats, using mpv internal format
* names, as long as it's internally marked as RGB, has exactly 1
* plane, and is supported as conversion output. It is not a good idea
* to rely on any of these. Their semantics and handling could change.
*/
MPV_RENDER_PARAM_SW_FORMAT = 18,
/**
* MPV_RENDER_API_TYPE_SW only: rendering target surface bytes per line,
* mandatory.
* Valid for MPV_RENDER_API_TYPE_SW & mpv_render_context_render().
* Type: size_t*
*
* This is the number of bytes between a pixel (x, y) and (x, y + 1) on the
* target surface. It must be a multiple of the pixel size, and have space
* for the surface width as specified by MPV_RENDER_PARAM_SW_SIZE.
*
* Both stride and pointer value should be a multiple of 64 to facilitate
* fast SIMD operation. Lower alignment might trigger slower code paths,
* and in the worst case, will copy the entire target frame. If mpv is built
* with zimg (and zimg is not disabled), the performance impact might be
* less.
* In either cases, the pointer and stride must be aligned at least to the
* pixel alignment size. Otherwise, crashes and undefined behavior is
* possible on platforms which do not support unaligned accesses (either
* through normal memory access or aligned SIMD memory access instructions).
*/
MPV_RENDER_PARAM_SW_STRIDE = 19,
/*
* MPV_RENDER_API_TYPE_SW only: rendering target surface pixel data pointer,
* mandatory.
* Valid for MPV_RENDER_API_TYPE_SW & mpv_render_context_render().
* Type: void*
*
* This points to the first pixel at the left/top corner (0, 0). In
* particular, each line y starts at (pointer + stride * y). Upon rendering,
* all data between pointer and (pointer + stride * h) is overwritten.
* Whether the padding between (w, y) and (0, y + 1) is overwritten is left
* unspecified (it should not be, but unfortunately some scaler backends
* will do it anyway). It is assumed that even the padding after the last
* line (starting at bytepos(w, h) until (pointer + stride * h)) is
* writable.
*
* See MPV_RENDER_PARAM_SW_STRIDE for alignment requirements.
*/
MPV_RENDER_PARAM_SW_POINTER = 20,
} mpv_render_param_type;
/**
* For backwards compatibility with the old naming of
* MPV_RENDER_PARAM_DRM_DRAW_SURFACE_SIZE
*/
#define MPV_RENDER_PARAM_DRM_OSD_SIZE MPV_RENDER_PARAM_DRM_DRAW_SURFACE_SIZE
/**
* Used to pass arbitrary parameters to some mpv_render_* functions. The
* meaning of the data parameter is determined by the type, and each
* MPV_RENDER_PARAM_* documents what type the value must point to.
*
* Each value documents the required data type as the pointer you cast to
* void* and set on mpv_render_param.data. For example, if MPV_RENDER_PARAM_FOO
* documents the type as Something* , then the code should look like this:
*
* Something foo = {...};
* mpv_render_param param;
* param.type = MPV_RENDER_PARAM_FOO;
* param.data = & foo;
*
* Normally, the data field points to exactly 1 object. If the type is char*,
* it points to a 0-terminated string.
*
* In all cases (unless documented otherwise) the pointers need to remain
* valid during the call only. Unless otherwise documented, the API functions
* will not write to the params array or any data pointed to it.
*
* As a convention, parameter arrays are always terminated by type==0. There
* is no specific order of the parameters required. The order of the 2 fields in
* this struct is guaranteed (even after ABI changes).
*/
typedef struct mpv_render_param {
enum mpv_render_param_type type;
void *data;
} mpv_render_param;
/**
* Predefined values for MPV_RENDER_PARAM_API_TYPE.
*/
// See render_gl.h
#define MPV_RENDER_API_TYPE_OPENGL "opengl"
// See section "Software renderer"
#define MPV_RENDER_API_TYPE_SW "sw"
/**
* Flags used in mpv_render_frame_info.flags. Each value represents a bit in it.
*/
typedef enum mpv_render_frame_info_flag {
/**
* Set if there is actually a next frame. If unset, there is no next frame
* yet, and other flags and fields that require a frame to be queued will
* be unset.
*
* This is set for _any_ kind of frame, even for redraw requests.
*
* Note that when this is unset, it simply means no new frame was
* decoded/queued yet, not necessarily that the end of the video was
* reached. A new frame can be queued after some time.
*
* If the return value of mpv_render_context_render() had the
* MPV_RENDER_UPDATE_FRAME flag set, this flag will usually be set as well,
* unless the frame is rendered, or discarded by other asynchronous events.
*/
MPV_RENDER_FRAME_INFO_PRESENT = 1 << 0,
/**
* If set, the frame is not an actual new video frame, but a redraw request.
* For example if the video is paused, and an option that affects video
* rendering was changed (or any other reason), an update request can be
* issued and this flag will be set.
*
* Typically, redraw frames will not be subject to video timing.
*
* Implies MPV_RENDER_FRAME_INFO_PRESENT.
*/
MPV_RENDER_FRAME_INFO_REDRAW = 1 << 1,
/**
* If set, this is supposed to reproduce the previous frame perfectly. This
* is usually used for certain "video-sync" options ("display-..." modes).
* Typically the renderer will blit the video from a FBO. Unset otherwise.
*
* Implies MPV_RENDER_FRAME_INFO_PRESENT.
*/
MPV_RENDER_FRAME_INFO_REPEAT = 1 << 2,
/**
* If set, the player timing code expects that the user thread blocks on
* vsync (by either delaying the render call, or by making a call to
* mpv_render_context_report_swap() at vsync time).
*
* Implies MPV_RENDER_FRAME_INFO_PRESENT.
*/
MPV_RENDER_FRAME_INFO_BLOCK_VSYNC = 1 << 3,
} mpv_render_frame_info_flag;
/**
* Information about the next video frame that will be rendered. Can be
* retrieved with MPV_RENDER_PARAM_NEXT_FRAME_INFO.
*/
typedef struct mpv_render_frame_info {
/**
* A bitset of mpv_render_frame_info_flag values (i.e. multiple flags are
* combined with bitwise or).
*/
uint64_t flags;
/**
* Absolute time at which the frame is supposed to be displayed. This is in
* the same unit and base as the time returned by mpv_get_time_us(). For
* frames that are redrawn, or if vsync locked video timing is used (see
* "video-sync" option), then this can be 0. The "video-timing-offset"
* option determines how much "headroom" the render thread gets (but a high
* enough frame rate can reduce it anyway). mpv_render_context_render() will
* normally block until the time is elapsed, unless you pass it
* MPV_RENDER_PARAM_BLOCK_FOR_TARGET_TIME = 0.
*/
int64_t target_time;
} mpv_render_frame_info;
/**
* Initialize the renderer state. Depending on the backend used, this will
* access the underlying GPU API and initialize its own objects.
*
* You must free the context with mpv_render_context_free(). Not doing so before
* the mpv core is destroyed may result in memory leaks or crashes.
*
* Currently, only at most 1 context can exists per mpv core (it represents the
* main video output).
*
* You should pass the following parameters:
* - MPV_RENDER_PARAM_API_TYPE to select the underlying backend/GPU API.
* - Backend-specific init parameter, like MPV_RENDER_PARAM_OPENGL_INIT_PARAMS.
* - Setting MPV_RENDER_PARAM_ADVANCED_CONTROL and following its rules is
* strongly recommended.
* - If you want to use hwdec, possibly hwdec interop resources.
*
* @param res set to the context (on success) or NULL (on failure). The value
* is never read and always overwritten.
* @param mpv handle used to get the core (the mpv_render_context won't depend
* on this specific handle, only the core referenced by it)
* @param params an array of parameters, terminated by type==0. It's left
* unspecified what happens with unknown parameters. At least
* MPV_RENDER_PARAM_API_TYPE is required, and most backends will
* require another backend-specific parameter.
* @return error code, including but not limited to:
* MPV_ERROR_UNSUPPORTED: the OpenGL version is not supported
* (or required extensions are missing)
* MPV_ERROR_NOT_IMPLEMENTED: an unknown API type was provided, or
* support for the requested API was not
* built in the used libmpv binary.
* MPV_ERROR_INVALID_PARAMETER: at least one of the provided parameters was
* not valid.
*/
MPV_EXPORT int mpv_render_context_create(mpv_render_context **res, mpv_handle *mpv,
mpv_render_param *params);
/**
* Attempt to change a single parameter. Not all backends and parameter types
* support all kinds of changes.
*
* @param ctx a valid render context
* @param param the parameter type and data that should be set
* @return error code. If a parameter could actually be changed, this returns
* success, otherwise an error code depending on the parameter type
* and situation.
*/
MPV_EXPORT int mpv_render_context_set_parameter(mpv_render_context *ctx,
mpv_render_param param);
/**
* Retrieve information from the render context. This is NOT a counterpart to
* mpv_render_context_set_parameter(), because you generally can't read
* parameters set with it, and this function is not meant for this purpose.
* Instead, this is for communicating information from the renderer back to the
* user. See mpv_render_param_type; entries which support this function
* explicitly mention it, and for other entries you can assume it will fail.
*
* You pass param with param.type set and param.data pointing to a variable
* of the required data type. The function will then overwrite that variable
* with the returned value (at least on success).
*
* @param ctx a valid render context
* @param param the parameter type and data that should be retrieved
* @return error code. If a parameter could actually be retrieved, this returns
* success, otherwise an error code depending on the parameter type
* and situation. MPV_ERROR_NOT_IMPLEMENTED is used for unknown
* param.type, or if retrieving it is not supported.
*/
MPV_EXPORT int mpv_render_context_get_info(mpv_render_context *ctx,
mpv_render_param param);
typedef void (*mpv_render_update_fn)(void *cb_ctx);
/**
* Set the callback that notifies you when a new video frame is available, or
* if the video display configuration somehow changed and requires a redraw.
* Similar to mpv_set_wakeup_callback(), you must not call any mpv API from
* the callback, and all the other listed restrictions apply (such as not
* exiting the callback by throwing exceptions).
*
* This can be called from any thread, except from an update callback. In case
* of the OpenGL backend, no OpenGL state or API is accessed.
*
* Calling this will raise an update callback immediately.
*
* @param callback callback(callback_ctx) is called if the frame should be
* redrawn
* @param callback_ctx opaque argument to the callback
*/
MPV_EXPORT void mpv_render_context_set_update_callback(mpv_render_context *ctx,
mpv_render_update_fn callback,
void *callback_ctx);
/**
* The API user is supposed to call this when the update callback was invoked
* (like all mpv_render_* functions, this has to happen on the render thread,
* and _not_ from the update callback itself).
*
* This is optional if MPV_RENDER_PARAM_ADVANCED_CONTROL was not set (default).
* Otherwise, it's a hard requirement that this is called after each update
* callback. If multiple update callback happened, and the function could not
* be called sooner, it's OK to call it once after the last callback.
*
* If an update callback happens during or after this function, the function
* must be called again at the soonest possible time.
*
* If MPV_RENDER_PARAM_ADVANCED_CONTROL was set, this will do additional work
* such as allocating textures for the video decoder.
*
* @return a bitset of mpv_render_update_flag values (i.e. multiple flags are
* combined with bitwise or). Typically, this will tell the API user
* what should happen next. E.g. if the MPV_RENDER_UPDATE_FRAME flag is
* set, mpv_render_context_render() should be called. If flags unknown
* to the API user are set, or if the return value is 0, nothing needs
* to be done.
*/
MPV_EXPORT uint64_t mpv_render_context_update(mpv_render_context *ctx);
/**
* Flags returned by mpv_render_context_update(). Each value represents a bit
* in the function's return value.
*/
typedef enum mpv_render_update_flag {
/**
* A new video frame must be rendered. mpv_render_context_render() must be
* called.
*/
MPV_RENDER_UPDATE_FRAME = 1 << 0,
} mpv_render_context_flag;
/**
* Render video.
*
* Typically renders the video to a target surface provided via mpv_render_param
* (the details depend on the backend in use). Options like "panscan" are
* applied to determine which part of the video should be visible and how the
* video should be scaled. You can change these options at runtime by using the
* mpv property API.
*
* The renderer will reconfigure itself every time the target surface
* configuration (such as size) is changed.
*
* This function implicitly pulls a video frame from the internal queue and
* renders it. If no new frame is available, the previous frame is redrawn.
* The update callback set with mpv_render_context_set_update_callback()
* notifies you when a new frame was added. The details potentially depend on
* the backends and the provided parameters.
*
* Generally, libmpv will invoke your update callback some time before the video
* frame should be shown, and then lets this function block until the supposed
* display time. This will limit your rendering to video FPS. You can prevent
* this by setting the "video-timing-offset" global option to 0. (This applies
* only to "audio" video sync mode.)
*
* You should pass the following parameters:
* - Backend-specific target object, such as MPV_RENDER_PARAM_OPENGL_FBO.
* - Possibly transformations, such as MPV_RENDER_PARAM_FLIP_Y.
*
* @param ctx a valid render context
* @param params an array of parameters, terminated by type==0. Which parameters
* are required depends on the backend. It's left unspecified what
* happens with unknown parameters.
* @return error code
*/
MPV_EXPORT int mpv_render_context_render(mpv_render_context *ctx, mpv_render_param *params);
/**
* Tell the renderer that a frame was flipped at the given time. This is
* optional, but can help the player to achieve better timing.
*
* Note that calling this at least once informs libmpv that you will use this
* function. If you use it inconsistently, expect bad video playback.
*
* If this is called while no video is initialized, it is ignored.
*
* @param ctx a valid render context
*/
MPV_EXPORT void mpv_render_context_report_swap(mpv_render_context *ctx);
/**
* Destroy the mpv renderer state.
*
* If video is still active (e.g. a file playing), video will be disabled
* forcefully.
*
* @param ctx a valid render context. After this function returns, this is not
* a valid pointer anymore. NULL is also allowed and does nothing.
*/
MPV_EXPORT void mpv_render_context_free(mpv_render_context *ctx);
cplugin: allow loading mpv_* symbols dynamically Defining MPV_CPLUGIN_DYNAMIC_SYM during plugin compilation will replace mpv_* functions with function pointers. Those pointer will be initialized when loading the plugin. It is recommended to use this symbol table when targeting Windows. The loader does not have notion of global symbols. Loading cplugin into mpv process will not allow this plugin to call any of the symbols that may be available in other modules. Instead cplugin has to link explicitly to specific PE binary, libmpv-2.dll/mpv.exe or any other binary that may have linked mpv statically. This limits portability of cplugin as it would need to be compiled separately for each of target PE binary that includes mpv's symbols. Which in practice is unrealictis, as we want one cplugin to be loaded without those restrictions. Instead of linking to any PE binary, we create function pointer for all mpv's exported symbols. For convinience names of entrypoints are redefined to those pointer so no changes are required in cplugin source code, except defining MPV_CPLUGIN_DYNAMIC_SYM. Those function pointer are exported to make them available for mpv to init with correct values during runtime, before calling `mpv_open_cplugin`. Note that those pointer are decorated with `selectany` attribute, so no need to worry about multiple definitions, linker will keep only single instance. This fixes cplugin usability on Windows. Without any API changes, only recompilation with -DMPV_CPLUGIN_DYNAMIC_SYM is needed.
2023-08-17 16:19:26 +02:00
#ifdef MPV_CPLUGIN_DYNAMIC_SYM
MPV_DEFINE_SYM_PTR(mpv_render_context_create)
#define mpv_render_context_create pfn_mpv_render_context_create
MPV_DEFINE_SYM_PTR(mpv_render_context_set_parameter)
#define mpv_render_context_set_parameter pfn_mpv_render_context_set_parameter
MPV_DEFINE_SYM_PTR(mpv_render_context_get_info)
#define mpv_render_context_get_info pfn_mpv_render_context_get_info
MPV_DEFINE_SYM_PTR(mpv_render_context_set_update_callback)
#define mpv_render_context_set_update_callback pfn_mpv_render_context_set_update_callback
MPV_DEFINE_SYM_PTR(mpv_render_context_update)
#define mpv_render_context_update pfn_mpv_render_context_update
MPV_DEFINE_SYM_PTR(mpv_render_context_render)
#define mpv_render_context_render pfn_mpv_render_context_render
MPV_DEFINE_SYM_PTR(mpv_render_context_report_swap)
#define mpv_render_context_report_swap pfn_mpv_render_context_report_swap
MPV_DEFINE_SYM_PTR(mpv_render_context_free)
#define mpv_render_context_free pfn_mpv_render_context_free
#endif
#ifdef __cplusplus
}
#endif
#endif