docs/egl.html - platform/external/mesa3d - Gitiles

 <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" "http://www.w3.org/TR/html4/loose.dtd">
 <html lang="en">
 <head>
   <meta http-equiv="content-type" content="text/html; charset=utf-8">
   <title>Mesa EGL</title>
   <link rel="stylesheet" type="text/css" href="mesa.css">
 </head>
 <body>

 <div class="header">
   <h1>The Mesa 3D Graphics Library</h1>
 </div>

 <iframe src="contents.html"></iframe>
 <div class="content">

 <h1>Mesa EGL</h1>

 <p>The current version of EGL in Mesa implements EGL 1.4.  More information
 about EGL can be found at
 <a href="https://www.khronos.org/egl/">
 https://www.khronos.org/egl/</a>.</p>

 <p>The Mesa's implementation of EGL uses a driver architecture.  The main
 library (<code>libEGL</code>) is window system neutral.  It provides the EGL
 API entry points and helper functions for use by the drivers.  Drivers are
 dynamically loaded by the main library and most of the EGL API calls are
 directly dispatched to the drivers.</p>

 <p>The driver in use decides the window system to support.</p>

 <h2>Build EGL</h2>

 <ol>
 <li>
 <p>Run <code>configure</code> with the desired client APIs and enable
 the driver for your hardware.  For example</p>

 <pre>
   $ ./configure --enable-gles1 --enable-gles2 \
                 --with-dri-drivers=... \
                 --with-gallium-drivers=...
 </pre>

 <p>The main library and OpenGL is enabled by default.  The first two options
 above enables <a href="opengles.html">OpenGL ES 1.x and 2.x</a>.  The last two
 options enables the listed classic and Gallium drivers respectively.</p>

 </li>

 <li>Build and install Mesa as usual.</li>
 </ol>

 <p>In the given example, it will build and install <code>libEGL</code>,
 <code>libGL</code>, <code>libGLESv1_CM</code>, <code>libGLESv2</code>, and one
 or more EGL drivers.</p>

 <h3>Configure Options</h3>

 <p>There are several options that control the build of EGL at configuration
 time</p>

 <dl>
 <dt><code>--enable-egl</code></dt>
 <dd>

 <p>By default, EGL is enabled.  When disabled, the main library and the drivers
 will not be built.</p>

 </dd>

 <dt><code>--with-egl-driver-dir</code></dt>
 <dd>

 <p>The directory EGL drivers should be installed to.  If not specified, EGL
 drivers will be installed to <code>${libdir}/egl</code>.</p>

 </dd>

 <dt><code>--with-platforms</code></dt>
 <dd>

 <p>List the platforms (window systems) to support.  Its argument is a comma
 separated string such as <code>--with-platforms=x11,drm</code>.  It decides
 the platforms a driver may support.  The first listed platform is also used by
 the main library to decide the native platform.</p>

 <p>The available platforms are <code>x11</code>, <code>drm</code>,
 <code>wayland</code>, <code>surfaceless</code>, <code>android</code>,
 and <code>haiku</code>.
 The <code>android</code> platform can either be built as a system
 component, part of AOSP, using <code>Android.mk</code> files, or
 cross-compiled using appropriate <code>configure</code> options.
 The <code>haiku</code> platform can only be built with SCons.
 Unless for special needs, the build system should
 select the right platforms automatically.</p>

 </dd>

 <dt><code>--enable-gles1</code></dt>
 <dt><code>--enable-gles2</code></dt>
 <dd>

 <p>These options enable OpenGL ES support in OpenGL.  The result is one big
 internal library that supports multiple APIs.</p>

 </dd>

 <dt><code>--enable-shared-glapi</code></dt>
 <dd>

 <p>By default, <code>libGL</code> has its own copy of <code>libglapi</code>.
 This options makes <code>libGL</code> use the shared <code>libglapi</code>.  This
 is required if applications mix OpenGL and OpenGL ES.</p>

 </dd>

 </dl>

 <h2>Use EGL</h2>

 <h3>Demos</h3>

 <p>There are demos for the client APIs supported by EGL.  They can be found in
 mesa/demos repository.</p>

 <h3>Environment Variables</h3>

 <p>There are several environment variables that control the behavior of EGL at
 runtime</p>

 <dl>
 <dt><code>EGL_PLATFORM</code></dt>
 <dd>

 <p>This variable specifies the native platform.  The valid values are the same
 as those for <code>--with-platforms</code>.  When the variable is not set,
 the main library uses the first platform listed in
 <code>--with-platforms</code> as the native platform.</p>

 <p>Extensions like <code>EGL_MESA_drm_display</code> define new functions to
 create displays for non-native platforms.  These extensions are usually used by
 applications that support non-native platforms.  Setting this variable is
 probably required only for some of the demos found in mesa/demo repository.</p>

 </dd>

 <dt><code>EGL_LOG_LEVEL</code></dt>
 <dd>

 <p>This changes the log level of the main library and the drivers.  The valid
 values are: <code>debug</code>, <code>info</code>, <code>warning</code>, and
 <code>fatal</code>.</p>

 </dd>
 </dl>

 <h2>EGL Drivers</h2>

 <dl>
 <dt><code>egl_dri2</code></dt>
 <dd>

 <p>This driver supports both <code>x11</code> and <code>drm</code> platforms.
 It functions as a DRI driver loader.  For <code>x11</code> support, it talks to
 the X server directly using (XCB-)DRI2 protocol.</p>

 <p>This driver can share DRI drivers with <code>libGL</code>.</p>

 </dd>

 <h2>Packaging</h2>

 <p>The ABI between the main library and its drivers are not stable.  Nor is
 there a plan to stabilize it at the moment.</p>

 <h2>Developers</h2>

 <p>The sources of the main library and drivers can be found at
 <code>src/egl/</code>.</p>

 <h3>Lifetime of Display Resources</h3>

 <p>Contexts and surfaces are examples of display resources.  They might live
 longer than the display that creates them.</p>

 <p>In EGL, when a display is terminated through <code>eglTerminate</code>, all
 display resources should be destroyed.  Similarly, when a thread is released
 through <code>eglReleaseThread</code>, all current display resources should be
 released.  Another way to destroy or release resources is through functions
 such as <code>eglDestroySurface</code> or <code>eglMakeCurrent</code>.</p>

 <p>When a resource that is current to some thread is destroyed, the resource
 should not be destroyed immediately.  EGL requires the resource to live until
 it is no longer current.  A driver usually calls
 <code>eglIs&lt;Resource&gt;Bound</code> to check if a resource is bound
 (current) to any thread in the destroy callbacks.  If it is still bound, the
 resource is not destroyed.</p>

 <p>The main library will mark destroyed current resources as unlinked.  In a
 driver's <code>MakeCurrent</code> callback,
 <code>eglIs&lt;Resource&gt;Linked</code> can then be called to check if a newly
 released resource is linked to a display.  If it is not, the last reference to
 the resource is removed and the driver should destroy the resource.  But it
 should be careful here because <code>MakeCurrent</code> might be called with an
 uninitialized display.</p>

 <p>This is the only mechanism provided by the main library to help manage the
 resources.  The drivers are responsible to the correct behavior as defined by
 EGL.</p>

 <h3><code>EGL_RENDER_BUFFER</code></h3>

 <p>In EGL, the color buffer a context should try to render to is decided by the
 binding surface.  It should try to render to the front buffer if the binding
 surface has <code>EGL_RENDER_BUFFER</code> set to
 <code>EGL_SINGLE_BUFFER</code>;  If the same context is later bound to a
 surface with <code>EGL_RENDER_BUFFER</code> set to
 <code>EGL_BACK_BUFFER</code>, the context should try to render to the back
 buffer.  However, the context is allowed to make the final decision as to which
 color buffer it wants to or is able to render to.</p>

 <p>For pbuffer surfaces, the render buffer is always
 <code>EGL_BACK_BUFFER</code>.  And for pixmap surfaces, the render buffer is
 always <code>EGL_SINGLE_BUFFER</code>.  Unlike window surfaces, EGL spec
 requires their <code>EGL_RENDER_BUFFER</code> values to be honored.  As a
 result, a driver should never set <code>EGL_PIXMAP_BIT</code> or
 <code>EGL_PBUFFER_BIT</code> bits of a config if the contexts created with the
 config won't be able to honor the <code>EGL_RENDER_BUFFER</code> of pixmap or
 pbuffer surfaces.</p>

 <p>It should also be noted that pixmap and pbuffer surfaces are assumed to be
 single-buffered, in that <code>eglSwapBuffers</code> has no effect on them.  It
 is desirable that a driver allocates a private color buffer for each pbuffer
 surface created.  If the window system the driver supports has native pbuffers,
 or if the native pixmaps have more than one color buffers, the driver should
 carefully attach the native color buffers to the EGL surfaces, re-route them if
 required.</p>

 <p>There is no defined behavior as to, for example, how
 <code>glDrawBuffer</code> interacts with <code>EGL_RENDER_BUFFER</code>.  Right
 now, it is desired that the draw buffer in a client API be fixed for pixmap and
 pbuffer surfaces.  Therefore, the driver is responsible to guarantee that the
 client API renders to the specified render buffer for pixmap and pbuffer
 surfaces.</p>

 <h3><code>EGLDisplay</code> Mutex</h3>

 The <code>EGLDisplay</code> will be locked before calling any of the dispatch
 functions (well, except for GetProcAddress which does not take an
 <code>EGLDisplay</code>).  This guarantees that the same dispatch function will
 not be called with the sample display at the same time.  If a driver has access
 to an <code>EGLDisplay</code> without going through the EGL APIs, the driver
 should as well lock the display before using it.

 </div>
 </body>
 </html>
	<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" "http://www.w3.org/TR/html4/loose.dtd">
	<html lang="en">
	<head>
	<meta http-equiv="content-type" content="text/html; charset=utf-8">
	<title>Mesa EGL</title>
	<link rel="stylesheet" type="text/css" href="mesa.css">
	</head>
	<body>

	<div class="header">
	<h1>The Mesa 3D Graphics Library</h1>
	</div>

	<iframe src="contents.html"></iframe>
	<div class="content">

	<h1>Mesa EGL</h1>

	<p>The current version of EGL in Mesa implements EGL 1.4. More information
	about EGL can be found at
	<a href="https://www.khronos.org/egl/">
	https://www.khronos.org/egl/</a>.</p>

	<p>The Mesa's implementation of EGL uses a driver architecture. The main
	library (<code>libEGL</code>) is window system neutral. It provides the EGL
	API entry points and helper functions for use by the drivers. Drivers are
	dynamically loaded by the main library and most of the EGL API calls are
	directly dispatched to the drivers.</p>

	<p>The driver in use decides the window system to support.</p>

	<h2>Build EGL</h2>

	<ol>
	<li>
	<p>Run <code>configure</code> with the desired client APIs and enable
	the driver for your hardware. For example</p>

	<pre>
	$ ./configure --enable-gles1 --enable-gles2 \
	--with-dri-drivers=... \
	--with-gallium-drivers=...
	</pre>

	<p>The main library and OpenGL is enabled by default. The first two options
	above enables <a href="opengles.html">OpenGL ES 1.x and 2.x</a>. The last two
	options enables the listed classic and Gallium drivers respectively.</p>

	</li>

	<li>Build and install Mesa as usual.</li>
	</ol>

	<p>In the given example, it will build and install <code>libEGL</code>,
	<code>libGL</code>, <code>libGLESv1_CM</code>, <code>libGLESv2</code>, and one
	or more EGL drivers.</p>

	<h3>Configure Options</h3>

	<p>There are several options that control the build of EGL at configuration
	time</p>

	<dl>
	<dt><code>--enable-egl</code></dt>
	<dd>

	<p>By default, EGL is enabled. When disabled, the main library and the drivers
	will not be built.</p>

	</dd>

	<dt><code>--with-egl-driver-dir</code></dt>
	<dd>

	<p>The directory EGL drivers should be installed to. If not specified, EGL
	drivers will be installed to <code>${libdir}/egl</code>.</p>

	</dd>

	<dt><code>--with-platforms</code></dt>
	<dd>

	<p>List the platforms (window systems) to support. Its argument is a comma
	separated string such as <code>--with-platforms=x11,drm</code>. It decides
	the platforms a driver may support. The first listed platform is also used by
	the main library to decide the native platform.</p>

	<p>The available platforms are <code>x11</code>, <code>drm</code>,
	<code>wayland</code>, <code>surfaceless</code>, <code>android</code>,
	and <code>haiku</code>.
	The <code>android</code> platform can either be built as a system
	component, part of AOSP, using <code>Android.mk</code> files, or
	cross-compiled using appropriate <code>configure</code> options.
	The <code>haiku</code> platform can only be built with SCons.
	Unless for special needs, the build system should
	select the right platforms automatically.</p>

	</dd>

	<dt><code>--enable-gles1</code></dt>
	<dt><code>--enable-gles2</code></dt>
	<dd>

	<p>These options enable OpenGL ES support in OpenGL. The result is one big
	internal library that supports multiple APIs.</p>

	</dd>

	<dt><code>--enable-shared-glapi</code></dt>
	<dd>

	<p>By default, <code>libGL</code> has its own copy of <code>libglapi</code>.
	This options makes <code>libGL</code> use the shared <code>libglapi</code>. This
	is required if applications mix OpenGL and OpenGL ES.</p>

	</dd>

	</dl>

	<h2>Use EGL</h2>

	<h3>Demos</h3>

	<p>There are demos for the client APIs supported by EGL. They can be found in
	mesa/demos repository.</p>

	<h3>Environment Variables</h3>

	<p>There are several environment variables that control the behavior of EGL at
	runtime</p>

	<dl>
	<dt><code>EGL_PLATFORM</code></dt>
	<dd>

	<p>This variable specifies the native platform. The valid values are the same
	as those for <code>--with-platforms</code>. When the variable is not set,
	the main library uses the first platform listed in
	<code>--with-platforms</code> as the native platform.</p>

	<p>Extensions like <code>EGL_MESA_drm_display</code> define new functions to
	create displays for non-native platforms. These extensions are usually used by
	applications that support non-native platforms. Setting this variable is
	probably required only for some of the demos found in mesa/demo repository.</p>

	</dd>

	<dt><code>EGL_LOG_LEVEL</code></dt>
	<dd>

	<p>This changes the log level of the main library and the drivers. The valid
	values are: <code>debug</code>, <code>info</code>, <code>warning</code>, and
	<code>fatal</code>.</p>

	</dd>
	</dl>

	<h2>EGL Drivers</h2>

	<dl>
	<dt><code>egl_dri2</code></dt>
	<dd>

	<p>This driver supports both <code>x11</code> and <code>drm</code> platforms.
	It functions as a DRI driver loader. For <code>x11</code> support, it talks to
	the X server directly using (XCB-)DRI2 protocol.</p>

	<p>This driver can share DRI drivers with <code>libGL</code>.</p>

	</dd>

	<h2>Packaging</h2>

	<p>The ABI between the main library and its drivers are not stable. Nor is
	there a plan to stabilize it at the moment.</p>

	<h2>Developers</h2>

	<p>The sources of the main library and drivers can be found at
	<code>src/egl/</code>.</p>

	<h3>Lifetime of Display Resources</h3>

	<p>Contexts and surfaces are examples of display resources. They might live
	longer than the display that creates them.</p>

	<p>In EGL, when a display is terminated through <code>eglTerminate</code>, all
	display resources should be destroyed. Similarly, when a thread is released
	through <code>eglReleaseThread</code>, all current display resources should be
	released. Another way to destroy or release resources is through functions
	such as <code>eglDestroySurface</code> or <code>eglMakeCurrent</code>.</p>

	<p>When a resource that is current to some thread is destroyed, the resource
	should not be destroyed immediately. EGL requires the resource to live until
	it is no longer current. A driver usually calls
	<code>eglIs<Resource>Bound</code> to check if a resource is bound
	(current) to any thread in the destroy callbacks. If it is still bound, the
	resource is not destroyed.</p>

	<p>The main library will mark destroyed current resources as unlinked. In a
	driver's <code>MakeCurrent</code> callback,
	<code>eglIs<Resource>Linked</code> can then be called to check if a newly
	released resource is linked to a display. If it is not, the last reference to
	the resource is removed and the driver should destroy the resource. But it
	should be careful here because <code>MakeCurrent</code> might be called with an
	uninitialized display.</p>

	<p>This is the only mechanism provided by the main library to help manage the
	resources. The drivers are responsible to the correct behavior as defined by
	EGL.</p>

	<h3><code>EGL_RENDER_BUFFER</code></h3>

	<p>In EGL, the color buffer a context should try to render to is decided by the
	binding surface. It should try to render to the front buffer if the binding
	surface has <code>EGL_RENDER_BUFFER</code> set to
	<code>EGL_SINGLE_BUFFER</code>; If the same context is later bound to a
	surface with <code>EGL_RENDER_BUFFER</code> set to
	<code>EGL_BACK_BUFFER</code>, the context should try to render to the back
	buffer. However, the context is allowed to make the final decision as to which
	color buffer it wants to or is able to render to.</p>

	<p>For pbuffer surfaces, the render buffer is always
	<code>EGL_BACK_BUFFER</code>. And for pixmap surfaces, the render buffer is
	always <code>EGL_SINGLE_BUFFER</code>. Unlike window surfaces, EGL spec
	requires their <code>EGL_RENDER_BUFFER</code> values to be honored. As a
	result, a driver should never set <code>EGL_PIXMAP_BIT</code> or
	<code>EGL_PBUFFER_BIT</code> bits of a config if the contexts created with the
	config won't be able to honor the <code>EGL_RENDER_BUFFER</code> of pixmap or
	pbuffer surfaces.</p>

	<p>It should also be noted that pixmap and pbuffer surfaces are assumed to be
	single-buffered, in that <code>eglSwapBuffers</code> has no effect on them. It
	is desirable that a driver allocates a private color buffer for each pbuffer
	surface created. If the window system the driver supports has native pbuffers,
	or if the native pixmaps have more than one color buffers, the driver should
	carefully attach the native color buffers to the EGL surfaces, re-route them if
	required.</p>

	<p>There is no defined behavior as to, for example, how
	<code>glDrawBuffer</code> interacts with <code>EGL_RENDER_BUFFER</code>. Right
	now, it is desired that the draw buffer in a client API be fixed for pixmap and
	pbuffer surfaces. Therefore, the driver is responsible to guarantee that the
	client API renders to the specified render buffer for pixmap and pbuffer
	surfaces.</p>

	<h3><code>EGLDisplay</code> Mutex</h3>

	The <code>EGLDisplay</code> will be locked before calling any of the dispatch
	functions (well, except for GetProcAddress which does not take an
	<code>EGLDisplay</code>). This guarantees that the same dispatch function will
	not be called with the sample display at the same time. If a driver has access
	to an <code>EGLDisplay</code> without going through the EGL APIs, the driver
	should as well lock the display before using it.

	</div>
	</body>
	</html>