mesa/src/mapi/glapi/gen/gl_SPARC_asm.py

190 lines
6.9 KiB
Python
Raw Normal View History

# (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>
import argparse
import license
import gl_XML, glX_XML
Mammoth update to the Python code generator scripts that live in src/mesa/glapi. Basically, the scripts that did simple things (like gl_offsets.py) were simple, and the scripts that did more complicated things (like glX_proto_send.py) were getting progressively more and more out of control. So, I re-write the foundation classes on which everything is based. One problem with the existing code is that the division between the GL API database representation and the way the output code is generated was either blury or nonexistant. The new code somewhat follows the Model-View-Controller pattern, minus the Controller. There is a distinct set of classes that model the API data, and there is a distinct set of classes that generate code from that data. One big change is in the class that represents GL functions (was glFunction, is now gl_function). There used to be an instance of this calls for each function and for each alias to that function. For example, there was an instance for PointParameterivSGIS, PointParameterivEXT, PointParameterivARB, and PointParameteriv. In the new code, there is one instance. Each instance has a list of entrypoint names for the function. In the next revision, this will allow a couple useful things. The script will be able to verify that the parameters, return type, and GLX protocol for a function and all it's aliases match. It will also allow aliases to be represented in the XML more compactly. Instead of repeating all the information, an alias can be listed as: <function name="PointParameterivARB" alias="PointParameterivEXT"/> Because the data representation was changed, the order that the alias functions are processed by the scripts also changed. This accounts for at least 2,700 of the ~3,600 lines of diffs in the generated code. Most of the remaining ~900 lines of diffs are the result of bugs *fixed* by the new scripts. The old scripts also generated code with some bugs in it. These bugs were discovered while the new code was being written. These changes were discussed on the mesa3d-dev mailing list back at the end of May: http://marc.theaimsgroup.com/?t=111714569000004&r=1&w=2 Xorg bug: 3197, 3208
2005-06-22 00:42:43 +01:00
class PrintGenericStubs(gl_XML.gl_print_base):
def __init__(self):
gl_XML.gl_print_base.__init__(self)
self.name = "gl_SPARC_asm.py (from Mesa)"
self.license = license.bsd_license_template % ( \
"""Copyright (C) 1999-2003 Brian Paul All Rights Reserved.
(C) Copyright IBM Corporation 2004""", "BRIAN PAUL, IBM")
def printRealHeader(self):
print('#ifdef __arch64__')
print('#define GL_OFF(N)\t((N) * 8)')
print('#define GL_LL\t\tldx')
print('#define GL_TIE_LD(SYM)\t%tie_ldx(SYM)')
print('#define GL_STACK_SIZE\t128')
print('#else')
print('#define GL_OFF(N)\t((N) * 4)')
print('#define GL_LL\t\tld')
print('#define GL_TIE_LD(SYM)\t%tie_ld(SYM)')
print('#define GL_STACK_SIZE\t64')
print('#endif')
print('')
print('#define GLOBL_FN(x) .globl x ; .type x, @function')
print('#define HIDDEN(x) .hidden x')
print('')
print('\t.register %g2, #scratch')
print('\t.register %g3, #scratch')
print('')
print('\t.text')
print('')
print('\tGLOBL_FN(__glapi_sparc_icache_flush)')
print('\tHIDDEN(__glapi_sparc_icache_flush)')
print('\t.type\t__glapi_sparc_icache_flush, @function')
print('__glapi_sparc_icache_flush: /* %o0 = insn_addr */')
print('\tflush\t%o0')
print('\tretl')
print('\t nop')
print('')
print('\t.align\t32')
print('')
print('\t.type\t__glapi_sparc_get_pc, @function')
print('__glapi_sparc_get_pc:')
print('\tretl')
print('\t add\t%o7, %g2, %g2')
print('\t.size\t__glapi_sparc_get_pc, .-__glapi_sparc_get_pc')
print('')
print('')
print('\tGLOBL_FN(__glapi_sparc_get_dispatch)')
print('\tHIDDEN(__glapi_sparc_get_dispatch)')
print('__glapi_sparc_get_dispatch:')
print('\tmov\t%o7, %g1')
print('\tsethi\t%hi(_GLOBAL_OFFSET_TABLE_-4), %g2')
print('\tcall\t__glapi_sparc_get_pc')
print('\tadd\t%g2, %lo(_GLOBAL_OFFSET_TABLE_+4), %g2')
print('\tmov\t%g1, %o7')
print('\tsethi\t%tie_hi22(_glapi_tls_Dispatch), %g1')
print('\tadd\t%g1, %tie_lo10(_glapi_tls_Dispatch), %g1')
print('\tGL_LL\t[%g2 + %g1], %g2, GL_TIE_LD(_glapi_tls_Dispatch)')
print('\tretl')
print('\t mov\t%g2, %o0')
print('')
print('\t.data')
print('\t.align\t32')
print('')
print('\t/* --> sethi %hi(_glapi_tls_Dispatch), %g1 */')
print('\t/* --> or %g1, %lo(_glapi_tls_Dispatch), %g1 */')
print('\tGLOBL_FN(__glapi_sparc_tls_stub)')
print('\tHIDDEN(__glapi_sparc_tls_stub)')
print('__glapi_sparc_tls_stub: /* Call offset in %g3 */')
print('\tmov\t%o7, %g1')
print('\tsethi\t%hi(_GLOBAL_OFFSET_TABLE_-4), %g2')
print('\tcall\t__glapi_sparc_get_pc')
print('\tadd\t%g2, %lo(_GLOBAL_OFFSET_TABLE_+4), %g2')
print('\tmov\t%g1, %o7')
print('\tsrl\t%g3, 10, %g3')
print('\tsethi\t%tie_hi22(_glapi_tls_Dispatch), %g1')
print('\tadd\t%g1, %tie_lo10(_glapi_tls_Dispatch), %g1')
print('\tGL_LL\t[%g2 + %g1], %g2, GL_TIE_LD(_glapi_tls_Dispatch)')
print('\tGL_LL\t[%g7+%g2], %g1')
print('\tGL_LL\t[%g1 + %g3], %g1')
print('\tjmp\t%g1')
print('\t nop')
print('\t.size\t__glapi_sparc_tls_stub, .-__glapi_sparc_tls_stub')
print('')
print('#define GL_STUB(fn, off)\t\t\t\t\\')
print('\tGLOBL_FN(fn);\t\t\t\t\t\\')
print('fn:\tba\t__glapi_sparc_tls_stub;\t\t\t\\')
print('\t sethi\tGL_OFF(off), %g3;\t\t\t\\')
print('\t.size\tfn,.-fn;')
print('')
print('#define GL_STUB_ALIAS(fn, alias) \\')
print(' .globl fn; \\')
print(' .set fn, alias')
print('')
print('\t.text')
print('\t.align\t32')
print('')
print('\t.globl\tgl_dispatch_functions_start')
print('\tHIDDEN(gl_dispatch_functions_start)')
print('gl_dispatch_functions_start:')
print('')
return
Mammoth update to the Python code generator scripts that live in src/mesa/glapi. Basically, the scripts that did simple things (like gl_offsets.py) were simple, and the scripts that did more complicated things (like glX_proto_send.py) were getting progressively more and more out of control. So, I re-write the foundation classes on which everything is based. One problem with the existing code is that the division between the GL API database representation and the way the output code is generated was either blury or nonexistant. The new code somewhat follows the Model-View-Controller pattern, minus the Controller. There is a distinct set of classes that model the API data, and there is a distinct set of classes that generate code from that data. One big change is in the class that represents GL functions (was glFunction, is now gl_function). There used to be an instance of this calls for each function and for each alias to that function. For example, there was an instance for PointParameterivSGIS, PointParameterivEXT, PointParameterivARB, and PointParameteriv. In the new code, there is one instance. Each instance has a list of entrypoint names for the function. In the next revision, this will allow a couple useful things. The script will be able to verify that the parameters, return type, and GLX protocol for a function and all it's aliases match. It will also allow aliases to be represented in the XML more compactly. Instead of repeating all the information, an alias can be listed as: <function name="PointParameterivARB" alias="PointParameterivEXT"/> Because the data representation was changed, the order that the alias functions are processed by the scripts also changed. This accounts for at least 2,700 of the ~3,600 lines of diffs in the generated code. Most of the remaining ~900 lines of diffs are the result of bugs *fixed* by the new scripts. The old scripts also generated code with some bugs in it. These bugs were discovered while the new code was being written. These changes were discussed on the mesa3d-dev mailing list back at the end of May: http://marc.theaimsgroup.com/?t=111714569000004&r=1&w=2 Xorg bug: 3197, 3208
2005-06-22 00:42:43 +01:00
def printRealFooter(self):
print('')
print('\t.globl\tgl_dispatch_functions_end')
print('\tHIDDEN(gl_dispatch_functions_end)')
print('gl_dispatch_functions_end:')
return
Mammoth update to the Python code generator scripts that live in src/mesa/glapi. Basically, the scripts that did simple things (like gl_offsets.py) were simple, and the scripts that did more complicated things (like glX_proto_send.py) were getting progressively more and more out of control. So, I re-write the foundation classes on which everything is based. One problem with the existing code is that the division between the GL API database representation and the way the output code is generated was either blury or nonexistant. The new code somewhat follows the Model-View-Controller pattern, minus the Controller. There is a distinct set of classes that model the API data, and there is a distinct set of classes that generate code from that data. One big change is in the class that represents GL functions (was glFunction, is now gl_function). There used to be an instance of this calls for each function and for each alias to that function. For example, there was an instance for PointParameterivSGIS, PointParameterivEXT, PointParameterivARB, and PointParameteriv. In the new code, there is one instance. Each instance has a list of entrypoint names for the function. In the next revision, this will allow a couple useful things. The script will be able to verify that the parameters, return type, and GLX protocol for a function and all it's aliases match. It will also allow aliases to be represented in the XML more compactly. Instead of repeating all the information, an alias can be listed as: <function name="PointParameterivARB" alias="PointParameterivEXT"/> Because the data representation was changed, the order that the alias functions are processed by the scripts also changed. This accounts for at least 2,700 of the ~3,600 lines of diffs in the generated code. Most of the remaining ~900 lines of diffs are the result of bugs *fixed* by the new scripts. The old scripts also generated code with some bugs in it. These bugs were discovered while the new code was being written. These changes were discussed on the mesa3d-dev mailing list back at the end of May: http://marc.theaimsgroup.com/?t=111714569000004&r=1&w=2 Xorg bug: 3197, 3208
2005-06-22 00:42:43 +01:00
def printBody(self, api):
for f in api.functionIterateByOffset():
name = f.dispatch_name()
print('\tGL_STUB(gl%s, %d)' % (name, f.offset))
if not f.is_static_entry_point(f.name):
print('\tHIDDEN(gl%s)' % (name))
for f in api.functionIterateByOffset():
name = f.dispatch_name()
Mammoth update to the Python code generator scripts that live in src/mesa/glapi. Basically, the scripts that did simple things (like gl_offsets.py) were simple, and the scripts that did more complicated things (like glX_proto_send.py) were getting progressively more and more out of control. So, I re-write the foundation classes on which everything is based. One problem with the existing code is that the division between the GL API database representation and the way the output code is generated was either blury or nonexistant. The new code somewhat follows the Model-View-Controller pattern, minus the Controller. There is a distinct set of classes that model the API data, and there is a distinct set of classes that generate code from that data. One big change is in the class that represents GL functions (was glFunction, is now gl_function). There used to be an instance of this calls for each function and for each alias to that function. For example, there was an instance for PointParameterivSGIS, PointParameterivEXT, PointParameterivARB, and PointParameteriv. In the new code, there is one instance. Each instance has a list of entrypoint names for the function. In the next revision, this will allow a couple useful things. The script will be able to verify that the parameters, return type, and GLX protocol for a function and all it's aliases match. It will also allow aliases to be represented in the XML more compactly. Instead of repeating all the information, an alias can be listed as: <function name="PointParameterivARB" alias="PointParameterivEXT"/> Because the data representation was changed, the order that the alias functions are processed by the scripts also changed. This accounts for at least 2,700 of the ~3,600 lines of diffs in the generated code. Most of the remaining ~900 lines of diffs are the result of bugs *fixed* by the new scripts. The old scripts also generated code with some bugs in it. These bugs were discovered while the new code was being written. These changes were discussed on the mesa3d-dev mailing list back at the end of May: http://marc.theaimsgroup.com/?t=111714569000004&r=1&w=2 Xorg bug: 3197, 3208
2005-06-22 00:42:43 +01:00
if f.is_static_entry_point(f.name):
for n in f.entry_points:
if n != f.name:
text = '\tGL_STUB_ALIAS(gl%s, gl%s)' % (n, f.name)
if f.has_different_protocol(n):
print('#ifndef GLX_INDIRECT_RENDERING')
print(text)
print('#endif')
else:
print(text)
return
Mammoth update to the Python code generator scripts that live in src/mesa/glapi. Basically, the scripts that did simple things (like gl_offsets.py) were simple, and the scripts that did more complicated things (like glX_proto_send.py) were getting progressively more and more out of control. So, I re-write the foundation classes on which everything is based. One problem with the existing code is that the division between the GL API database representation and the way the output code is generated was either blury or nonexistant. The new code somewhat follows the Model-View-Controller pattern, minus the Controller. There is a distinct set of classes that model the API data, and there is a distinct set of classes that generate code from that data. One big change is in the class that represents GL functions (was glFunction, is now gl_function). There used to be an instance of this calls for each function and for each alias to that function. For example, there was an instance for PointParameterivSGIS, PointParameterivEXT, PointParameterivARB, and PointParameteriv. In the new code, there is one instance. Each instance has a list of entrypoint names for the function. In the next revision, this will allow a couple useful things. The script will be able to verify that the parameters, return type, and GLX protocol for a function and all it's aliases match. It will also allow aliases to be represented in the XML more compactly. Instead of repeating all the information, an alias can be listed as: <function name="PointParameterivARB" alias="PointParameterivEXT"/> Because the data representation was changed, the order that the alias functions are processed by the scripts also changed. This accounts for at least 2,700 of the ~3,600 lines of diffs in the generated code. Most of the remaining ~900 lines of diffs are the result of bugs *fixed* by the new scripts. The old scripts also generated code with some bugs in it. These bugs were discovered while the new code was being written. These changes were discussed on the mesa3d-dev mailing list back at the end of May: http://marc.theaimsgroup.com/?t=111714569000004&r=1&w=2 Xorg bug: 3197, 3208
2005-06-22 00:42:43 +01:00
def _parser():
"""Parse arguments and return a namespace."""
parser = argparse.ArgumentParser()
parser.add_argument('-f',
dest='filename',
default='gl_API.xml',
help='An XML description of an API.')
return parser.parse_args()
def main():
"""Main function."""
args = _parser()
printer = PrintGenericStubs()
api = gl_XML.parse_GL_API(args.filename, glX_XML.glx_item_factory())
printer.Print(api)
if __name__ == '__main__':
main()