• Home
  • Line#
  • Scopes#
  • Navigate#
  • Raw
  • Download
1GL Dispatch
2===========
3
4Several factors combine to make efficient dispatch of OpenGL functions
5fairly complicated. This document attempts to explain some of the issues
6and introduce the reader to Mesa's implementation. Readers already
7familiar with the issues around GL dispatch can safely skip ahead to the
8:ref:`overview of Mesa's implementation <overview>`.
9
101. Complexity of GL Dispatch
11----------------------------
12
13Every GL application has at least one object called a GL *context*. This
14object, which is an implicit parameter to every GL function, stores all
15of the GL related state for the application. Every texture, every buffer
16object, every enable, and much, much more is stored in the context.
17Since an application can have more than one context, the context to be
18used is selected by a window-system dependent function such as
19``glXMakeContextCurrent``.
20
21In environments that implement OpenGL with X-Windows using GLX, every GL
22function, including the pointers returned by ``glXGetProcAddress``, are
23*context independent*. This means that no matter what context is
24currently active, the same ``glVertex3fv`` function is used.
25
26This creates the first bit of dispatch complexity. An application can
27have two GL contexts. One context is a direct rendering context where
28function calls are routed directly to a driver loaded within the
29application's address space. The other context is an indirect rendering
30context where function calls are converted to GLX protocol and sent to a
31server. The same ``glVertex3fv`` has to do the right thing depending on
32which context is current.
33
34Highly optimized drivers or GLX protocol implementations may want to
35change the behavior of GL functions depending on current state. For
36example, ``glFogCoordf`` may operate differently depending on whether or
37not fog is enabled.
38
39In multi-threaded environments, it is possible for each thread to have a
40different GL context current. This means that poor old ``glVertex3fv``
41has to know which GL context is current in the thread where it is being
42called.
43
44.. _overview:
45
462. Overview of Mesa's Implementation
47------------------------------------
48
49Mesa uses two per-thread pointers. The first pointer stores the address
50of the context current in the thread, and the second pointer stores the
51address of the *dispatch table* associated with that context. The
52dispatch table stores pointers to functions that actually implement
53specific GL functions. Each time a new context is made current in a
54thread, these pointers are updated.
55
56The implementation of functions such as ``glVertex3fv`` becomes
57conceptually simple:
58
59-  Fetch the current dispatch table pointer.
60-  Fetch the pointer to the real ``glVertex3fv`` function from the
61   table.
62-  Call the real function.
63
64This can be implemented in just a few lines of C code. The file
65``src/mesa/glapi/glapitemp.h`` contains code very similar to this.
66
67.. code-block:: c
68   :caption: Sample dispatch function
69
70   void glVertex3f(GLfloat x, GLfloat y, GLfloat z)
71   {
72       const struct _glapi_table * const dispatch = GET_DISPATCH();
73
74       (*dispatch->Vertex3f)(x, y, z);
75   }
76
77The problem with this simple implementation is the large amount of
78overhead that it adds to every GL function call.
79
80In a multithreaded environment, a naive implementation of
81``GET_DISPATCH`` involves a call to ``pthread_getspecific`` or a similar
82function. Mesa provides a wrapper function called
83``_glapi_get_dispatch`` that is used by default.
84
853. Optimizations
86----------------
87
88A number of optimizations have been made over the years to diminish the
89performance hit imposed by GL dispatch. This section describes these
90optimizations. The benefits of each optimization and the situations
91where each can or cannot be used are listed.
92
933.1. Dual dispatch table pointers
94~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
95
96The vast majority of OpenGL applications use the API in a single
97threaded manner. That is, the application has only one thread that makes
98calls into the GL. In these cases, not only do the calls to
99``pthread_getspecific`` hurt performance, but they are completely
100unnecessary! It is possible to detect this common case and avoid these
101calls.
102
103Each time a new dispatch table is set, Mesa examines and records the ID
104of the executing thread. If the same thread ID is always seen, Mesa
105knows that the application is, from OpenGL's point of view, single
106threaded.
107
108As long as an application is single threaded, Mesa stores a pointer to
109the dispatch table in a global variable called ``_glapi_Dispatch``. The
110pointer is also stored in a per-thread location via
111``pthread_setspecific``. When Mesa detects that an application has
112become multithreaded, ``NULL`` is stored in ``_glapi_Dispatch``.
113
114Using this simple mechanism the dispatch functions can detect the
115multithreaded case by comparing ``_glapi_Dispatch`` to ``NULL``. The
116resulting implementation of ``GET_DISPATCH`` is slightly more complex,
117but it avoids the expensive ``pthread_getspecific`` call in the common
118case.
119
120.. code-block:: c
121   :caption: Improved ``GET_DISPATCH`` Implementation
122
123   #define GET_DISPATCH() \
124       (_glapi_Dispatch != NULL) \
125           ? _glapi_Dispatch : pthread_getspecific(&_glapi_Dispatch_key)
126
1273.2. ELF TLS
128~~~~~~~~~~~~
129
130Starting with the 2.4.20 Linux kernel, each thread is allocated an area
131of per-thread, global storage. Variables can be put in this area using
132some extensions to GCC. By storing the dispatch table pointer in this
133area, the expensive call to ``pthread_getspecific`` and the test of
134``_glapi_Dispatch`` can be avoided.
135
136The dispatch table pointer is stored in a new variable called
137``_glapi_tls_Dispatch``. A new variable name is used so that a single
138libGL can implement both interfaces. This allows the libGL to operate
139with direct rendering drivers that use either interface. Once the
140pointer is properly declared, ``GET_DISPACH`` becomes a simple variable
141reference.
142
143.. code-block:: c
144   :caption: TLS ``GET_DISPATCH`` Implementation
145
146   extern __THREAD_INITIAL_EXEC struct _glapi_table *_glapi_tls_Dispatch;
147
148   #define GET_DISPATCH() _glapi_tls_Dispatch
149
150Use of this path is controlled by the preprocessor define
151``USE_ELF_TLS``. Any platform capable of using ELF TLS should use this
152as the default dispatch method.
153
154Windows has a similar concept, and beginning with Windows Vista, shared
155libraries can take advantage of compiler-assisted TLS. This TLS data
156has no fixed size and does not compete with API-based TLS (``TlsAlloc``)
157for the limited number of slots available there, and so ``USE_ELF_TLS`` can
158be used on Windows too, even though it's not truly ELF.
159
1603.3. Assembly Language Dispatch Stubs
161~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
162
163Many platforms have difficulty properly optimizing the tail-call in the
164dispatch stubs. Platforms like x86 that pass parameters on the stack
165seem to have even more difficulty optimizing these routines. All of the
166dispatch routines are very short, and it is trivial to create optimal
167assembly language versions. The amount of optimization provided by using
168assembly stubs varies from platform to platform and application to
169application. However, by using the assembly stubs, many platforms can
170use an additional space optimization (see :ref:`below <fixedsize>`).
171
172The biggest hurdle to creating assembly stubs is handling the various
173ways that the dispatch table pointer can be accessed. There are four
174different methods that can be used:
175
176#. Using ``_glapi_Dispatch`` directly in builds for non-multithreaded
177   environments.
178#. Using ``_glapi_Dispatch`` and ``_glapi_get_dispatch`` in
179   multithreaded environments.
180#. Using ``_glapi_Dispatch`` and ``pthread_getspecific`` in
181   multithreaded environments.
182#. Using ``_glapi_tls_Dispatch`` directly in TLS enabled multithreaded
183   environments.
184
185People wishing to implement assembly stubs for new platforms should
186focus on #4 if the new platform supports TLS. Otherwise, implement #2
187followed by #3. Environments that do not support multithreading are
188uncommon and not terribly relevant.
189
190Selection of the dispatch table pointer access method is controlled by a
191few preprocessor defines.
192
193-  If ``USE_ELF_TLS`` is defined, method #3 is used.
194-  If ``HAVE_PTHREAD`` is defined, method #2 is used.
195-  If none of the preceding are defined, method #1 is used.
196
197Two different techniques are used to handle the various different cases.
198On x86 and SPARC, a macro called ``GL_STUB`` is used. In the preamble of
199the assembly source file different implementations of the macro are
200selected based on the defined preprocessor variables. The assembly code
201then consists of a series of invocations of the macros such as:
202
203.. code-block:: c
204   :caption: SPARC Assembly Implementation of ``glColor3fv``
205
206   GL_STUB(Color3fv, _gloffset_Color3fv)
207
208The benefit of this technique is that changes to the calling pattern
209(i.e., addition of a new dispatch table pointer access method) require
210fewer changed lines in the assembly code.
211
212However, this technique can only be used on platforms where the function
213implementation does not change based on the parameters passed to the
214function. For example, since x86 passes all parameters on the stack, no
215additional code is needed to save and restore function parameters around
216a call to ``pthread_getspecific``. Since x86-64 passes parameters in
217registers, varying amounts of code needs to be inserted around the call
218to ``pthread_getspecific`` to save and restore the GL function's
219parameters.
220
221The other technique, used by platforms like x86-64 that cannot use the
222first technique, is to insert ``#ifdef`` within the assembly
223implementation of each function. This makes the assembly file
224considerably larger (e.g., 29,332 lines for ``glapi_x86-64.S`` versus
2251,155 lines for ``glapi_x86.S``) and causes simple changes to the
226function implementation to generate many lines of diffs. Since the
227assembly files are typically generated by scripts, this isn't a
228significant problem.
229
230Once a new assembly file is created, it must be inserted in the build
231system. There are two steps to this. The file must first be added to
232``src/mesa/sources``. That gets the file built and linked. The second
233step is to add the correct ``#ifdef`` magic to
234``src/mesa/glapi/glapi_dispatch.c`` to prevent the C version of the
235dispatch functions from being built.
236
237.. _fixedsize:
238
2393.4. Fixed-Length Dispatch Stubs
240~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
241
242To implement ``glXGetProcAddress``, Mesa stores a table that associates
243function names with pointers to those functions. This table is stored in
244``src/mesa/glapi/glprocs.h``. For different reasons on different
245platforms, storing all of those pointers is inefficient. On most
246platforms, including all known platforms that support TLS, we can avoid
247this added overhead.
248
249If the assembly stubs are all the same size, the pointer need not be
250stored for every function. The location of the function can instead be
251calculated by multiplying the size of the dispatch stub by the offset of
252the function in the table. This value is then added to the address of
253the first dispatch stub.
254
255This path is activated by adding the correct ``#ifdef`` magic to
256``src/mesa/glapi/glapi.c`` just before ``glprocs.h`` is included.
257