utils/conv_gen.py: use shared tables if possible

This change introduces the memory usage optimization, mentioned
in d2d9760c08. The aim is to make
code generator able to detect, whether the same tables are used
by several convolutional code definitions, and prevent one from
writing these tables multiple times.

For now, the detection process isn't fully automatic, so all
shared polynomials should be placed inside the 'shared_polys'
dictionary, for example:

shared_polys = {
	"xcch" : [
		( G0, 1 ),
		( G1, 1 ),
	],
	"mcs" : [
		( G4, 1 ),
		( G7, 1 ),
		( G5, 1 ),
	],
}

Change-Id: I84760f5cdfdaece376b801d2e6cb2954ee875a3b
This commit is contained in:
Vadim Yanitskiy 2016-10-29 00:00:57 +07:00 committed by Harald Welte
parent c014f606d0
commit 6431adde78
1 changed files with 61 additions and 37 deletions

View File

@ -164,7 +164,7 @@ class ConvolutionalCode(object):
# Up to 12 numbers should be placed per line
print_formatted(self.puncture, "%3d, ", 12, fi)
def gen_tables(self, pref, fi):
def print_state_and_output(self, fi):
pack = lambda n: \
sum([x << (self.rate_inv - i - 1) for i, x in enumerate(n)])
num_states = 1 << (self.k - 1)
@ -186,6 +186,14 @@ class ConvolutionalCode(object):
self._print_term(fi, num_states, pack)
fi.write("};\n\n")
def gen_tables(self, pref, fi, shared_tables = None):
# Do not print shared tables
if shared_tables is None:
self.print_state_and_output(fi)
table_pref = self.name
else:
table_pref = shared_tables
if len(self.puncture):
fi.write("static const int %s_puncture[] = {\n" % self.name)
self._print_puncture(fi)
@ -203,15 +211,15 @@ class ConvolutionalCode(object):
fi.write("\t.N = %d,\n" % self.rate_inv)
fi.write("\t.K = %d,\n" % self.k)
fi.write("\t.len = %d,\n" % self.block_len)
fi.write("\t.next_output = %s_output,\n" % self.name)
fi.write("\t.next_state = %s_state,\n" % self.name)
fi.write("\t.next_output = %s_output,\n" % table_pref)
fi.write("\t.next_state = %s_state,\n" % table_pref)
if self.term_type is not None:
fi.write("\t.term = %s,\n" % self.term_type)
if self.recursive:
fi.write("\t.next_term_output = %s_term_output,\n" % self.name)
fi.write("\t.next_term_state = %s_term_state,\n" % self.name)
fi.write("\t.next_term_output = %s_term_output,\n" % table_pref)
fi.write("\t.next_term_state = %s_term_state,\n" % table_pref)
if len(self.puncture):
fi.write("\t.puncture = %s_puncture,\n" % self.name)
@ -239,6 +247,12 @@ def print_formatted(items, format, count, fi):
fi.write("\n")
def print_shared(fi):
for (name, polys) in shared_polys.items():
# HACK
code = ConvolutionalCode(0, polys, name = name)
code.print_state_and_output(fi)
# Polynomials according to 3GPP TS 05.03 Annex B
G0 = poly(0, 3, 4)
G1 = poly(0, 1, 3, 4)
@ -249,22 +263,23 @@ G5 = poly(0, 1, 4, 6)
G6 = poly(0, 1, 2, 3, 4, 6)
G7 = poly(0, 1, 2, 3, 6)
CCH_poly = [
( G0, 1 ),
( G1, 1 ),
]
MCS_poly = [
( G4, 1 ),
( G7, 1 ),
( G5, 1 ),
]
shared_polys = {
"xcch" : [
( G0, 1 ),
( G1, 1 ),
],
"mcs" : [
( G4, 1 ),
( G7, 1 ),
( G5, 1 ),
],
}
conv_codes = [
# xCCH definition
ConvolutionalCode(
224,
CCH_poly,
shared_polys["xcch"],
name = "xcch",
description = [
"xCCH convolutional code:",
@ -277,7 +292,7 @@ conv_codes = [
# RACH definition
ConvolutionalCode(
14,
CCH_poly,
shared_polys["xcch"],
name = "rach",
description = ["RACH convolutional code"]
),
@ -285,7 +300,7 @@ conv_codes = [
# SCH definition
ConvolutionalCode(
35,
CCH_poly,
shared_polys["xcch"],
name = "sch",
description = ["SCH convolutional code"]
),
@ -293,7 +308,7 @@ conv_codes = [
# CS2 definition
ConvolutionalCode(
290,
CCH_poly,
shared_polys["xcch"],
puncture = [
15, 19, 23, 27, 31, 35, 43, 47, 51, 55, 59, 63, 67, 71,
75, 79, 83, 91, 95, 99, 103, 107, 111, 115, 119, 123, 127, 131,
@ -317,7 +332,7 @@ conv_codes = [
# CS3 definition
ConvolutionalCode(
334,
CCH_poly,
shared_polys["xcch"],
puncture = [
15, 17, 21, 23, 27, 29, 33, 35, 39, 41, 45, 47, 51, 53,
57, 59, 63, 65, 69, 71, 75, 77, 81, 83, 87, 89, 93, 95,
@ -579,7 +594,7 @@ conv_codes = [
# TCH_FR definition
ConvolutionalCode(
185,
CCH_poly,
shared_polys["xcch"],
name = "tch_fr",
description = ["TCH/F convolutional code"]
),
@ -724,7 +739,7 @@ conv_codes = [
# EDGE MCS1_DL_HDR definition
ConvolutionalCode(
36,
MCS_poly,
shared_polys["mcs"],
name = "mcs1_dl_hdr",
term_type = "CONV_TERM_TAIL_BITING",
description = [
@ -739,7 +754,7 @@ conv_codes = [
# EDGE MCS1_UL_HDR definition
ConvolutionalCode(
39,
MCS_poly,
shared_polys["mcs"],
name = "mcs1_ul_hdr",
term_type = "CONV_TERM_TAIL_BITING",
description = [
@ -754,7 +769,7 @@ conv_codes = [
# EDGE MCS1 definition
ConvolutionalCode(
190,
MCS_poly,
shared_polys["mcs"],
name = "mcs1",
description = [
"EDGE MCS-1 data convolutional code:",
@ -768,7 +783,7 @@ conv_codes = [
# EDGE MCS2 definition
ConvolutionalCode(
238,
MCS_poly,
shared_polys["mcs"],
name = "mcs2",
description = [
"EDGE MCS-2 data convolutional code:",
@ -782,7 +797,7 @@ conv_codes = [
# EDGE MCS3 definition
ConvolutionalCode(
310,
MCS_poly,
shared_polys["mcs"],
name = "mcs3",
description = [
"EDGE MCS-3 data convolutional code:",
@ -796,7 +811,7 @@ conv_codes = [
# EDGE MCS4 definition
ConvolutionalCode(
366,
MCS_poly,
shared_polys["mcs"],
name = "mcs4",
description = [
"EDGE MCS-4 data convolutional code:",
@ -810,7 +825,7 @@ conv_codes = [
# EDGE MCS5_DL_HDR definition
ConvolutionalCode(
33,
MCS_poly,
shared_polys["mcs"],
name = "mcs5_dl_hdr",
term_type = "CONV_TERM_TAIL_BITING",
description = [
@ -825,7 +840,7 @@ conv_codes = [
# EDGE MCS5_UL_HDR definition
ConvolutionalCode(
45,
MCS_poly,
shared_polys["mcs"],
name = "mcs5_ul_hdr",
term_type = "CONV_TERM_TAIL_BITING",
description = [
@ -840,7 +855,7 @@ conv_codes = [
# EDGE MCS5 definition
ConvolutionalCode(
462,
MCS_poly,
shared_polys["mcs"],
name = "mcs5",
description = [
"EDGE MCS-5 data convolutional code:",
@ -854,7 +869,7 @@ conv_codes = [
# EDGE MCS6 definition
ConvolutionalCode(
606,
MCS_poly,
shared_polys["mcs"],
name = "mcs6",
description = [
"EDGE MCS-6 data convolutional code:",
@ -868,7 +883,7 @@ conv_codes = [
# EDGE MCS7_DL_HDR definition
ConvolutionalCode(
45,
MCS_poly,
shared_polys["mcs"],
name = "mcs7_dl_hdr",
term_type = "CONV_TERM_TAIL_BITING",
description = [
@ -883,7 +898,7 @@ conv_codes = [
# EDGE MCS7_UL_HDR definition
ConvolutionalCode(
54,
MCS_poly,
shared_polys["mcs"],
name = "mcs7_ul_hdr",
term_type = "CONV_TERM_TAIL_BITING",
description = [
@ -898,7 +913,7 @@ conv_codes = [
# EDGE MCS7 definition
ConvolutionalCode(
462,
MCS_poly,
shared_polys["mcs"],
name = "mcs7",
description = [
"EDGE MCS-7 data convolutional code:",
@ -912,7 +927,7 @@ conv_codes = [
# EDGE MCS8 definition
ConvolutionalCode(
558,
MCS_poly,
shared_polys["mcs"],
name = "mcs8",
description = [
"EDGE MCS-8 data convolutional code:",
@ -926,7 +941,7 @@ conv_codes = [
# EDGE MCS9 definition
ConvolutionalCode(
606,
MCS_poly,
shared_polys["mcs"],
name = "mcs9",
description = [
"EDGE MCS-9 data convolutional code:",
@ -949,10 +964,19 @@ if __name__ == '__main__':
f.write(mod_license + "\n")
f.write("#include <stdint.h>\n")
f.write("#include <osmocom/core/conv.h>\n\n")
print_shared(f)
# Generate the tables one by one
for code in conv_codes:
sys.stderr.write("Generate '%s' definition\n" % code.name)
code.gen_tables(prefix, f)
# Check whether shared polynomials are used
shared = None
for (name, polys) in shared_polys.items():
if code.polys == polys:
shared = name
break
code.gen_tables(prefix, f, shared_tables = shared)
sys.stderr.write("Generation complete.\n")