Python script to generate various bits of client-side GLX protocol code.

1 files changed, 981 insertions, 0 deletions

diff --git a/src/mesa/glapi/glX_proto_send.py b/src/mesa/glapi/glX_proto_send.py
new file mode 100644
index 0000000000..77d54b6ef1
--- /dev/null
+++ b/src/mesa/glapi/glX_proto_send.py
@@ -0,0 +1,981 @@
+#!/usr/bin/python2
+
+# (C) Copyright IBM Corporation 2004
+# All Rights Reserved.
+#
+# 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
+# on the rights to use, copy, modify, merge, publish, distribute, sub
+# license, 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:
+#
+# 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 NON-INFRINGEMENT.  IN NO EVENT SHALL
+# IBM AND/OR ITS SUPPLIERS 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.
+#
+# Authors:
+#    Ian Romanick <idr@us.ibm.com>
+
+from xml.sax import saxutils
+from xml.sax import make_parser
+from xml.sax.handler import feature_namespaces
+
+import gl_XML
+import license
+import sys, getopt
+
+
+def printPure():
+	print """#  if __GNUC__ > 2 || (__GNUC__ == 2 && __GNUC_MINOR__ >= 96)
+#    define PURE __attribute__((pure))
+#  else
+#    define PURE
+#  endif"""
+
+def printFastcall():
+	print """#  if defined(__i386__) && defined(__GNUC__)
+#    define FASTCALL __attribute__((fastcall))
+#  else
+#    define FASTCALL
+#  endif"""
+
+def printVisibility(S, s):
+	print """#  if __GNUC__ > 3 || (__GNUC__ == 3 && __GNUC_MINOR__ >= 3)
+#    define %s  __attribute__((visibility("%s")))
+#  else
+#    define %s
+#  endif""" % (S, s, S)
+
+def printNoinline():
+	print """#  if defined(__GNUC__)
+#    define NOINLINE __attribute__((noinline))
+#  else
+#    define NOINLINE
+#  endif"""
+
+
+class glXItemFactory(gl_XML.glItemFactory):
+	"""Factory to create GLX protocol oriented objects derived from glItem."""
+    
+	def create(self, context, name, attrs):
+		if name == "function":
+			return glXFunction(context, name, attrs)
+		elif name == "enum":
+			return glXEnum(context, name, attrs)
+		elif name == "param":
+			return glXParameter(context, name, attrs)
+		else:
+			return gl_XML.glItemFactory.create(self, context, name, attrs)
+
+class glXEnumFunction:
+	def __init__(self, name):
+		self.name = name
+		
+		# "enums" is a set of lists.  The element in the set is the
+		# value of the enum.  The list is the list of names for that
+		# value.  For example, [0x8126] = {"POINT_SIZE_MIN",
+		# "POINT_SIZE_MIN_ARB", "POINT_SIZE_MIN_EXT",
+		# "POINT_SIZE_MIN_SGIS"}.
+
+		self.enums = {}
+
+		# "count" is indexed by count values.  Each element of count
+		# is a list of index to "enums" that have that number of
+		# associated data elements.  For example, [4] = 
+		# {GL_AMBIENT, GL_DIFFUSE, GL_SPECULAR, GL_EMISSION,
+		# GL_AMBIENT_AND_DIFFUSE} (the enum names are used here,
+		# but the actual hexadecimal values would be in the array).
+
+		self.count = {}
+
+
+	def append(self, count, value, name):
+		if self.enums.has_key( value ):
+			self.enums[value].append(name)
+		else:
+			if not self.count.has_key(count):
+				self.count[count] = []
+
+			self.enums[value] = []
+			self.enums[value].append(name)
+			self.count[count].append(value)
+
+
+	def signature( self ):
+		sig = ""
+		for i in self.count:
+			for e in self.count[i]:
+				sig += "%04x,%u," % (e, i)
+	
+		return sig;
+
+
+	def PrintUsingTable(self):
+		"""Emit the body of the __gl*_size function using a pair
+		of look-up tables and a mask.  The mask is calculated such
+		that (e & mask) is unique for all the valid values of e for
+		this function.  The result of (e & mask) is used as an index
+		into the first look-up table.  If it matches e, then the
+		same entry of the second table is returned.  Otherwise zero
+		is returned.
+		
+		It seems like this should cause better code to be generated.
+		However, on x86 at least, the resulting .o file is about 20%
+		larger then the switch-statment version.  I am leaving this
+		code in because the results may be different on other
+		platforms (e.g., PowerPC or x86-64)."""
+
+		return 0
+		count = 0
+		for a in self.enums:
+			count += 1
+
+		# Determine if there is some mask M, such that M = (2^N) - 1,
+		# that will generate unique values for all of the enums.
+
+		mask = 0
+		for i in [1, 2, 3, 4, 5, 6, 7, 8]:
+			mask = (1 << i) - 1
+
+			fail = 0;
+			for a in self.enums:
+				for b in self.enums:
+					if a != b:
+						if (a & mask) == (b & mask):
+							fail = 1;
+
+			if not fail:
+				break;
+			else:
+				mask = 0
+
+		if (mask != 0) and (mask < (2 * count)):
+			masked_enums = {}
+			masked_count = {}
+
+			for i in range(0, mask):
+				masked_enums[i] = "0";
+				masked_count[i] = 0;
+
+			for c in self.count:
+				for e in self.count[c]:
+					i = e & mask
+					masked_enums[i] = '0x%04x /* %s */' % (e, self.enums[e][0])
+					masked_count[i] = c
+
+
+			print '    static const GLushort a[%u] = {' % (mask + 1)
+			for e in masked_enums:
+				print '        %s, ' % (masked_enums[e])
+			print '    };'
+
+			print '    static const GLubyte b[%u] = {' % (mask + 1)
+			for c in masked_count:
+				print '        %u, ' % (masked_count[c])
+			print '    };'
+
+			print '    const unsigned idx = (e & 0x%02xU);' % (mask)
+			print ''
+			print '    return (e == a[idx]) ? (GLint) b[idx] : 0;'
+			return 1;
+		else:
+			return 0;
+
+	def PrintUsingSwitch(self):
+		"""Emit the body of the __gl*_size function using a 
+		switch-statement."""
+
+		print '    switch( e ) {'
+
+		for c in self.count:
+			for e in self.count[c]:
+				first = 1
+
+				# There may be multiple enums with the same
+				# value.  This happens has extensions are
+				# promoted from vendor-specific or EXT to
+				# ARB and to the core.  Emit the first one as
+				# a case label, and emit the others as
+				# commented-out case labels.
+
+				for j in self.enums[e]:
+					if first:
+						print '        case %s:' % (j)
+						first = 0
+					else:
+						print '/*      case %s:*/' % (j)
+					
+			print '            return %u;' % (c)
+					
+		print '        default: return 0;'
+		print '    }'
+
+
+	def Print(self, name):
+		print 'INTERNAL PURE FASTCALL GLint'
+		print '__gl%s_size( GLenum e )' % (name)
+		print '{'
+
+		if not self.PrintUsingTable():
+			self.PrintUsingSwitch()
+
+		print '}'
+		print ''
+
+
+
+class glXEnum(gl_XML.glEnum):
+	def __init__(self, context, name, attrs):
+		gl_XML.glEnum.__init__(self, context, name, attrs)
+		self.glx_functions = []
+
+	def startElement(self, name, attrs):
+		if name == "size":
+			n = attrs.get('name', None)
+			if not self.context.glx_enum_functions.has_key( n ):
+				f = glXEnumFunction( n )
+				self.context.glx_enum_functions[ f.name ] = f
+
+			temp = attrs.get('count', None)
+			try:
+				c = int(temp)
+			except Exception,e:
+				raise RuntimeError('Invalid count value "%s" for enum "%s" in function "%s" when an integer was expected.' % (temp, self.name, n))
+
+			self.context.glx_enum_functions[ n ].append( c, self.value, self.name )
+		else:
+			gl_XML.glEnum.startElement(self, context, name, attrs)
+		return
+
+
+class glXParameter(gl_XML.glParameter):
+	def __init__(self, context, name, attrs):
+		self.order = 1;
+		gl_XML.glParameter.__init__(self, context, name, attrs);
+
+
+class glXFunction(gl_XML.glFunction):
+	glx_rop = 0
+	glx_sop = 0
+	glx_vendorpriv = 0
+
+	# If this is set to true, it means that GLdouble parameters should be
+	# written to the GLX protocol packet in the order they appear in the
+	# prototype.  This is different from the "classic" ordering.  In the
+	# classic ordering GLdoubles are written to the protocol packet first,
+	# followed by non-doubles.  NV_vertex_program was the first extension
+	# to break with this tradition.
+
+	glx_doubles_in_order = 0
+
+	vectorequiv = None
+	handcode = 0
+	ignore = 0
+	can_be_large = 0
+
+	def __init__(self, context, name, attrs):
+		self.vectorequiv = attrs.get('vectorequiv', None)
+		self.count_parameters = None
+		self.counter = None
+		self.output = None
+		self.can_be_large = 0
+		self.reply_always_array = 0
+
+		gl_XML.glFunction.__init__(self, context, name, attrs)
+		return
+
+	def startElement(self, name, attrs):
+		"""Process elements within a function that are specific to GLX."""
+
+		if name == "glx":
+			self.glx_rop = int(attrs.get('rop', "0"))
+			self.glx_sop = int(attrs.get('sop', "0"))
+			self.glx_vendorpriv = int(attrs.get('vendorpriv', "0"))
+
+			if attrs.get('handcode', "false") == "true":
+				self.handcode = 1
+			else:
+				self.handcode = 0
+
+			if attrs.get('ignore', "false") == "true":
+				self.ignore = 1
+			else:
+				self.ignore = 0
+
+			if attrs.get('large', "false") == "true":
+				self.can_be_large = 1
+			else:
+				self.can_be_large = 0
+
+			if attrs.get('doubles_in_order', "false") == "true":
+				self.glx_doubles_in_order = 1
+			else:
+				self.glx_doubles_in_order = 0
+
+			if attrs.get('always_array', "false") == "true":
+				self.reply_always_array = 1
+			else:
+				self.reply_always_array = 0
+
+		else:
+			gl_XML.glFunction.startElement(self, name, attrs)
+
+
+	def append(self, tag_name, p):
+		gl_XML.glFunction.append(self, tag_name, p)
+
+		if p.is_variable_length_array():
+			p.order = 2;
+		elif not self.glx_doubles_in_order and p.p_type.size == 8:
+			p.order = 0;
+
+		if p.p_count_parameters != None:
+			self.count_parameters = p.p_count_parameters
+		
+		if p.is_counter:
+			self.counter = p.name
+			
+		if p.is_output:
+			self.output = p
+
+		return
+
+	def variable_length_parameter(self):
+		for param in self.fn_parameters:
+			if param.is_variable_length_array():
+				return param
+			
+		return None
+
+
+	def command_payload_length(self):
+		size = 0
+		size_string = ""
+		for p in self:
+			if p.is_output: continue
+			temp = p.size_string()
+			try:
+				s = int(temp)
+				size += s
+			except Exception,e:
+				size_string = size_string + " + __GLX_PAD(%s)" % (temp)
+
+		return [size, size_string]
+
+	def command_length(self):
+		[size, size_string] = self.command_payload_length()
+
+		if self.glx_rop != 0:
+			size += 4
+
+		size = ((size + 3) & ~3)
+		return "%u%s" % (size, size_string)
+
+
+	def opcode_value(self):
+		if self.glx_rop != 0:
+			return self.glx_rop
+		elif self.glx_sop != 0:
+			return self.glx_sop
+		elif self.glx_vendorpriv != 0:
+			return self.glx_vendorpriv
+		else:
+			return -1
+	
+	def opcode_rop_basename(self):
+		if self.vectorequiv == None:
+			return self.name
+		else:
+			return self.vectorequiv
+
+	def opcode_name(self):
+		if self.glx_rop != 0:
+			return "X_GLrop_%s" % (self.opcode_rop_basename())
+		elif self.glx_sop != 0:
+			return "X_GLsop_%s" % (self.name)
+		elif self.glx_vendorpriv != 0:
+			return "X_GLvop_%s" % (self.name)
+		else:
+			return "ERROR"
+
+	def opcode_real_name(self):
+		if self.glx_vendorpriv != 0:
+			if self.needs_reply():
+				return "X_GLXVendorPrivateWithReply"
+			else:
+				return "X_GLXVendorPrivate"
+		else:
+			return self.opcode_name()
+
+
+	def return_string(self):
+		if self.fn_return_type != 'void':
+			return "return retval;"
+		else:
+			return "return;"
+
+
+	def needs_reply(self):
+		return self.fn_return_type != 'void' or self.output != None
+
+
+class GlxProto(gl_XML.FilterGLAPISpecBase):
+	name = "glX_proto_send.py (from Mesa)"
+
+	def __init__(self):
+		gl_XML.FilterGLAPISpecBase.__init__(self)
+		self.factory = glXItemFactory()
+		self.glx_enum_functions = {}
+
+
+	def endElement(self, name):
+		if name == 'OpenGLAPI':
+			# Once all the parsing is done, we have to go back and
+			# fix-up some cross references between different
+			# functions.
+
+			for k in self.functions:
+				f = self.functions[k]
+				if f.vectorequiv != None:
+					equiv = self.find_function(f.vectorequiv)
+					if equiv != None:
+						f.glx_doubles_in_order = equiv.glx_doubles_in_order
+						f.glx_rop = equiv.glx_rop
+					else:
+						raise RuntimeError("Could not find the vector equiv. function %s for %s!" % (f.name, f.vectorequiv))
+		else:
+			gl_XML.FilterGLAPISpecBase.endElement(self, name)
+		return
+
+
+class PrintGlxProtoStubs(GlxProto):
+	def __init__(self):
+		GlxProto.__init__(self)
+		self.last_category = ""
+		self.license = license.bsd_license_template % ( "(C) Copyright IBM Corporation 2004", "IBM")
+		self.generic_sizes = [3, 4, 6, 8, 12, 16, 24, 32]
+		return
+
+	def printRealHeader(self):
+		print ''
+		print '#include <GL/gl.h>'
+		print '#include "indirect.h"'
+		print '#include "glxclient.h"'
+		print '#include "size.h"'
+		print '#include <GL/glxproto.h>'
+		print ''
+		print '#define __GLX_PAD(n) (((n) + 3) & ~3)'
+		print ''
+		printFastcall()
+		printNoinline()
+		print ''
+		print '/* If the size and opcode values are known at compile-time, this will, on'
+		print ' * x86 at least, emit them with a single instruction.'
+		print ' */'
+		print '#define emit_header(dest, op, size)            \\'
+		print '    do { union { short s[2]; int i; } temp;    \\'
+		print '         temp.s[0] = (size); temp.s[1] = (op); \\'
+		print '         *((int *)(dest)) = temp.i; } while(0)'
+		print ''
+		print """static NOINLINE CARD32
+read_reply( Display *dpy, size_t size, void * dest, GLboolean reply_is_always_array )
+{
+    xGLXSingleReply reply;
+    
+    (void) _XReply(dpy, (xReply *) & reply, 0, False);
+    if (size != 0) {
+	if ((reply.size >= 1) || reply_is_always_array) {
+	    const GLint bytes = (reply_is_always_array) 
+	      ? (4 * reply.length) : (reply.size * size);
+	    const GLint extra = 4 - (bytes & 3);
+
+	    _XRead(dpy, dest, bytes);
+	    if ( extra != 0 ) {
+		_XEatData(dpy, extra);
+	    }
+	}
+	else {
+	    (void) memcpy( dest, &(reply.pad3), size);
+	}
+    }
+
+    return reply.retval;
+}
+
+#define X_GLXSingle 0
+
+static NOINLINE GLubyte *
+setup_single_request( __GLXcontext * gc, GLint sop, GLint cmdlen )
+{
+    xGLXSingleReq * req;
+    Display * const dpy = gc->currentDpy;
+
+    (void) __glXFlushRenderBuffer(gc, gc->pc);
+    LockDisplay(dpy);
+    GetReqExtra(GLXSingle, cmdlen, req);
+    req->reqType = gc->majorOpcode;
+    req->contextTag = gc->currentContextTag;
+    req->glxCode = sop;
+    return (GLubyte *)(req) + sz_xGLXSingleReq;
+}
+
+static NOINLINE GLubyte *
+setup_vendor_request( __GLXcontext * gc, GLint code, GLint vop, GLint cmdlen )
+{
+    xGLXVendorPrivateReq * req;
+    Display * const dpy = gc->currentDpy;
+
+    (void) __glXFlushRenderBuffer(gc, gc->pc);
+    LockDisplay(dpy);
+    GetReqExtra(GLXVendorPrivate, cmdlen, req);
+    req->reqType = gc->majorOpcode;
+    req->glxCode = code;
+    req->vendorCode = vop;
+    req->contextTag = gc->currentContextTag;
+    return (GLubyte *)(req) + sz_xGLXVendorPrivateReq;
+}
+"""
+
+		for size in self.generic_sizes:
+			self.print_generic_function(size)
+		return
+
+	def printFunction(self, f):
+		if f.fn_offset < 0 or f.handcode or f.ignore: return
+
+		if f.glx_rop != 0 or f.vectorequiv != None:
+			self.printRenderFunction(f)
+		elif f.glx_sop != 0 or f.glx_vendorpriv != 0:
+			self.printSingleFunction(f)
+		else:
+			print "/* Missing GLX protocol for %s. */" % (f.name)
+
+	def print_generic_function(self, n):
+		print """static FASTCALL NOINLINE void
+generic_%u_byte( GLint rop, const void * ptr )
+{
+    __GLXcontext * const gc = __glXGetCurrentContext();
+    const GLuint cmdlen = %u;
+
+    emit_header(gc->pc, rop, cmdlen);
+    (void) memcpy((void *)(gc->pc + 4), ptr, %u);
+    gc->pc += cmdlen;
+    if (gc->pc > gc->limit) { (void) __glXFlushRenderBuffer(gc, gc->pc); }
+}
+""" % (n, n + 4, n)
+
+
+	def common_emit_one_arg(self, p, offset, pc, indent, adjust):
+		if p.is_output: return
+
+		t = p.p_type
+		if p.is_array():
+			src_ptr = p.name
+		else:
+			src_ptr = "&" + p.name
+
+		print '%s    (void) memcpy((void *)(%s + %u), (void *)(%s), %s);' \
+			% (indent, pc, offset + adjust, src_ptr, p.size_string() )
+
+	def common_emit_args(self, f, pc, indent, adjust, skip_vla):
+		# First emit all of the fixed-length 8-byte (i.e., GLdouble)
+		# parameters.
+
+		offset = 0
+
+		if skip_vla:
+			r = [0, 1]
+		else:
+			r = [0, 1, 2]
+
+		for order in r:
+			for p in f:
+				if p.is_output or p.order != order: continue
+
+				self.common_emit_one_arg(p, offset, pc, indent, adjust)
+				offset += p.size()
+
+
+		return offset
+
+
+	def common_func_print_just_header(self, f):
+		print '#define %s %d' % (f.opcode_name(), f.opcode_value())
+
+		print '%s' % (f.fn_return_type)
+		print '__indirect_gl%s(%s)' % (f.name, f.get_parameter_string())
+		print '{'
+
+
+	def common_func_print_header(self, f):
+		self.common_func_print_just_header(f)
+
+		print '    __GLXcontext * const gc = __glXGetCurrentContext();'
+		print '    Display * const dpy = gc->currentDpy;'
+			
+		if f.fn_return_type != 'void':
+			print '    %s retval = (%s) 0;' % (f.fn_return_type, f.fn_return_type)
+
+		if f.count_parameters != None:
+			print '    const GLuint compsize = __gl%s_size(%s);' % (f.name, f.count_parameters)
+
+		print '    const GLuint cmdlen = %s;' % (f.command_length())
+
+		if f.counter != None:
+			print '    if (%s < 0) %s' % (f.counter, f.return_string())
+
+		if f.can_be_large:
+			print '    if (dpy == NULL) return;'
+			print '    if ( ((gc->pc + cmdlen) > gc->bufEnd)'
+			print '         || (cmdlen > gc->maxSmallRenderCommandSize)) {'
+			print '        (void) __glXFlushRenderBuffer(gc, gc->pc);'
+			print '    }'
+		else:
+			print '    (void) dpy;'
+
+		return
+
+
+	def printSingleFunction(self, f):
+		self.common_func_print_header(f)
+
+		print '    if (dpy != NULL) {'
+
+		if f.fn_parameters != []:
+			pc_decl = "GLubyte const * pc ="
+		else:
+			pc_decl = "(void)"
+
+		if f.glx_vendorpriv != 0:
+			print '        %s setup_vendor_request(gc, %s, %s, cmdlen);' % (pc_decl, f.opcode_real_name(), f.opcode_name())
+		else:
+			print '        %s setup_single_request(gc, %s, cmdlen);' % (pc_decl, f.opcode_name())
+
+		self.common_emit_args(f, "pc", "    ", 0, 0)
+
+		if f.needs_reply():
+			if f.output != None:
+				output_size = f.output.p_type.size
+				output_str = f.output.name
+			else:
+				output_size = 0
+				output_str = "NULL"
+
+			if f.fn_return_type != 'void':
+				return_str = " retval = (%s)" % (f.fn_return_type)
+			else:
+				return_str = " (void)"
+
+			if f.reply_always_array:
+				aa = "GL_TRUE"
+			else:
+				aa = "GL_FALSE"
+
+			print "       %s read_reply(gc->currentDpy, %s, %s, %s);" % (return_str, output_size, output_str, aa)
+
+		print '        UnlockDisplay(gc->currentDpy); SyncHandle();'
+		print '    }'
+		print '    %s' % f.return_string()
+		print '}'
+		print ''
+		return
+
+
+	def printRenderFunction(self, f):
+		if f.variable_length_parameter() == None and len(f.fn_parameters) == 1:
+			p = f.fn_parameters[0]
+			if p.is_pointer:
+				[cmdlen, size_string] = f.command_payload_length()
+				if cmdlen in self.generic_sizes:
+					self.common_func_print_just_header(f)
+					print '    generic_%u_byte( %s, %s );' % (cmdlen, f.opcode_real_name(), p.name)
+					print '}'
+					print ''
+					return
+
+		self.common_func_print_header(f)
+
+		if f.can_be_large:
+			print '    if (cmdlen <= gc->maxSmallRenderCommandSize) {'
+			indent = "    "
+		else:
+			indent = ""
+
+		print '%s    emit_header(gc->pc, %s, cmdlen);' % (indent, f.opcode_real_name())
+
+		self.common_emit_args(f, "gc->pc", indent, 4, 0)
+		print '%s    gc->pc += cmdlen;' % (indent)
+		print '%s    if (gc->pc > gc->limit) { (void) __glXFlushRenderBuffer(gc, gc->pc); }' % (indent)
+
+		if f.can_be_large:
+			print '    }'
+			print '    else {'
+			print '        const GLint op = %s;' % (f.opcode_real_name())
+			print '        const GLuint cmdlenLarge = cmdlen + 4;'
+			print '        (void) memcpy((void *)(gc->pc + 0), (void *)(&op), 4);'
+			print '        (void) memcpy((void *)(gc->pc + 4), (void *)(&cmdlenLarge), 4);'
+			offset = self.common_emit_args(f, "gc->pc", indent, 8, 1)
+			
+			p = f.variable_length_parameter()
+			print '        __glXSendLargeCommand(gc, gc->pc, %u, %s, %s);' % (offset + 8, p.name, p.size_string())
+			print '    }'
+
+		print '}'
+		print ''
+		return
+
+
+class PrintGlxProtoInit_c(GlxProto):
+	def __init__(self):
+		GlxProto.__init__(self)
+		self.last_category = ""
+		self.license = license.bsd_license_template % ( \
+"""Copyright 1998-1999 Precision Insight, Inc., Cedar Park, Texas.
+(C) Copyright IBM Corporation 2004""", "PRECISION INSIGHT, IBM")
+
+
+	def printRealHeader(self):
+		print """/**
+ * \\file indirect_init.c
+ * Initialize indirect rendering dispatch table.
+ *
+ * \\author Kevin E. Martin <kevin@precisioninsight.com>
+ * \\author Brian Paul <brian@precisioninsight.com>
+ * \\author Ian Romanick <idr@us.ibm.com>
+ */
+
+#include "indirect_init.h"
+#include "indirect.h"
+#include "glapi.h"
+
+
+/**
+ * No-op function used to initialize functions that have no GLX protocol
+ * support.
+ */
+static int NoOp(void)
+{
+    return 0;
+}
+
+/**
+ * Create and initialize a new GL dispatch table.  The table is initialized
+ * with GLX indirect rendering protocol functions.
+ */
+__GLapi * __glXNewIndirectAPI( void )
+{
+    __GLapi *glAPI;
+    GLuint entries;
+
+    entries = _glapi_get_dispatch_table_size();
+    glAPI = (__GLapi *) Xmalloc(entries * sizeof(void *));
+
+    /* first, set all entries to point to no-op functions */
+    {
+       int i;
+       void **dispatch = (void **) glAPI;
+       for (i = 0; i < entries; i++) {
+          dispatch[i] = (void *) NoOp;
+       }
+    }
+
+    /* now, initialize the entries we understand */"""
+
+	def printRealFooter(self):
+		print """
+    return glAPI;
+}
+"""
+
+	def printFunction(self, f):
+		if f.fn_offset < 0 or f.ignore: return
+		
+		if f.category != self.last_category:
+			self.last_category = f.category
+			print ''
+			print '    /* %s */' % (self.last_category)
+			print ''
+			
+		print '    glAPI->%s = __indirect_gl%s;' % (f.name, f.name)
+
+
+class PrintGlxProtoInit_h(GlxProto):
+	def __init__(self):
+		GlxProto.__init__(self)
+		self.last_category = ""
+		self.license = license.bsd_license_template % ( \
+"""Copyright 1998-1999 Precision Insight, Inc., Cedar Park, Texas.
+(C) Copyright IBM Corporation 2004""", "PRECISION INSIGHT, IBM")
+
+
+	def printRealHeader(self):
+		print """
+/**
+ * \\file
+ * Prototypes for indirect rendering functions.
+ *
+ * \\author Kevin E. Martin <kevin@precisioninsight.com>
+ * \\author Ian Romanick <idr@us.ibm.com>
+ */
+
+#if !defined( _INDIRECT_H_ )
+#  define _INDIRECT_H_
+
+"""
+		printVisibility( "HIDDEN", "hidden" )
+
+
+	def printRealFooter(self):
+		print "#  undef HIDDEN"
+		print "#endif /* !defined( _INDIRECT_H_ ) */"
+
+	def printFunction(self, f):
+		if f.fn_offset < 0 or f.ignore: return
+		print 'extern HIDDEN %s __indirect_gl%s(%s);' % (f.fn_return_type, f.name, f.get_parameter_string())
+
+
+class PrintGlxSizeStubs(GlxProto):
+	def __init__(self):
+		GlxProto.__init__(self)
+		self.license = license.bsd_license_template % ( "(C) Copyright IBM Corporation 2004", "IBM")
+		self.aliases = []
+		self.glx_enum_sigs = {}
+
+	def printRealHeader(self):
+		print ''
+		print '#include <GL/gl.h>'
+		print '#include "indirect_size.h"'
+		
+		print ''
+		printPure()
+		print ''
+		printFastcall()
+		print ''
+		printVisibility( "INTERNAL", "internal" )
+		print ''
+		print ''
+		print '#ifdef HAVE_ALIAS'
+		print '#  define ALIAS2(from,to) \\'
+		print '    INTERNAL PURE FASTCALL GLint __gl ## from ## _size( GLenum e ) \\'
+		print '        __attribute__ ((alias( # to )));'
+		print '#  define ALIAS(from,to) ALIAS2( from, __gl ## to ## _size )'
+		print '#else'
+		print '#  define ALIAS(from,to) \\'
+		print '    INTERNAL PURE FASTCALL GLint __gl ## from ## _size( GLenum e ) \\'
+		print '    { return __gl ## to ## _size( e ); }'
+		print '#endif'
+		print ''
+		print ''
+
+	def printRealFooter(self):
+		for a in self.aliases:
+			print a
+
+	def printFunction(self, f):
+		if self.glx_enum_functions.has_key(f.name):
+			ef = self.glx_enum_functions[f.name]
+
+			sig = ef.signature();
+			if self.glx_enum_sigs.has_key(sig):
+				n = self.glx_enum_sigs[sig];
+				a = 'ALIAS( %s, %s )' % (f.name, n)
+				self.aliases.append(a)
+			else:
+				ef.Print( f.name )
+				self.glx_enum_sigs[sig] = f.name;
+
+
+				
+class PrintGlxSizeStubs_h(GlxProto):
+	def __init__(self):
+		GlxProto.__init__(self)
+		self.license = license.bsd_license_template % ( "(C) Copyright IBM Corporation 2004", "IBM")
+		self.aliases = []
+		self.glx_enum_sigs = {}
+
+	def printRealHeader(self):
+		print """
+/**
+ * \\file
+ * Prototypes for functions used to determine the number of data elements in
+ * various GLX protocol messages.
+ *
+ * \\author Ian Romanick <idr@us.ibm.com>
+ */
+
+#if !defined( _GLXSIZE_H_ )
+#  define _GLXSIZE_H_
+
+"""
+		printPure();
+		print ''
+		printFastcall();
+		print ''
+		printVisibility( "INTERNAL", "internal" );
+		print ''
+
+	def printRealFooter(self):
+		print ''
+		print "#  undef INTERNAL"
+		print "#  undef PURE"
+		print "#  undef FASTCALL"
+		print "#endif /* !defined( _GLXSIZE_H_ ) */"
+
+
+	def printFunction(self, f):
+		if self.glx_enum_functions.has_key(f.name):
+			ef = self.glx_enum_functions[f.name]
+			print 'extern INTERNAL GLint __gl%s_size(GLenum) PURE FASTCALL;' % (f.name)
+
+
+def show_usage():
+	print "Usage: %s [-f input_file_name] [-m output_mode]" % sys.argv[0]
+	sys.exit(1)
+
+
+if __name__ == '__main__':
+	file_name = "gl_API.xml"
+
+	try:
+		(args, trail) = getopt.getopt(sys.argv[1:], "f:m:")
+	except Exception,e:
+		show_usage()
+
+	mode = "proto"
+	for (arg,val) in args:
+		if arg == "-f":
+			file_name = val
+		elif arg == "-m":
+			mode = val
+
+	if mode == "proto":
+		dh = PrintGlxProtoStubs()
+	elif mode == "init_c":
+		dh = PrintGlxProtoInit_c()
+	elif mode == "init_h":
+		dh = PrintGlxProtoInit_h()
+	elif mode == "size_c":
+		dh = PrintGlxSizeStubs()
+	elif mode == "size_h":
+		dh = PrintGlxSizeStubs_h()
+	else:
+		show_usage()
+
+	parser = make_parser()
+	parser.setFeature(feature_namespaces, 0)
+	parser.setContentHandler(dh)
+
+	f = open(file_name)
+
+	dh.printHeader()
+	parser.parse(f)
+	dh.printFooter()