Report data decomp in calcrom

This commit is contained in:
PikalaxALT 2019-08-06 13:00:03 -04:00
parent 1bda9ef2f6
commit 2a0517f0fd

View File

@ -9,6 +9,8 @@ open(my $file, $ARGV[0])
my $src = 0;
my $asm = 0;
my $srcdata = 0;
my $data = 0;
while (my $line = <$file>)
{
if ($line =~ /^ \.(\w+)\s+0x[0-9a-f]+\s+(0x[0-9a-f]+) (\w+)\/.+\.o/)
@ -28,9 +30,22 @@ while (my $line = <$file>)
$asm += $size;
}
}
elsif ($section =~ /rodata/)
{
if ($dir eq 'src')
{
$srcdata += $size;
}
elsif ($dir eq 'data')
{
$data += $size;
}
}
}
}
(my $elffname = $ARGV[0]) =~ s/\.map/.elf/;
# Note that the grep filters out all branch labels. It also requires a minimum
# line length of 5, to filter out a ton of generated symbols (like AcCn). No
# settings to nm seem to remove these symbols. Finally, nm prints out a separate
@ -40,7 +55,7 @@ while (my $line = <$file>)
#
# You'd expect this to take a while, because of uniq. It runs in under a second,
# though. Uniq is pretty fast!
my $base_cmd = "nm pokeemerald.elf | awk '{print \$3}' | grep '^[^_].\\{4\\}' | uniq";
my $base_cmd = "nm $elffname | awk '{print \$3}' | grep '^[^_].\\{4\\}' | uniq";
# This looks for Unknown_, Unknown_, or sub_, followed by just numbers. Note that
# it matches even if stuff precedes the unknown, like sUnknown/gUnknown.
@ -118,3 +133,11 @@ print "$total_syms total symbols\n";
print "$documented symbols documented ($docPct%)\n";
print "$partial_documented symbols partially documented ($partialPct%)\n";
print "$undocumented symbols undocumented ($undocPct%)\n";
print "\n";
my $dataTotal = $srcdata + $data;
my $srcDataPct = sprintf("%.4f", 100 * $srcdata / $dataTotal);
my $dataPct = sprintf("%.4f", 100 * $data / $dataTotal);
print "$dataTotal total bytes of data\n";
print "$srcdata bytes of data in src ($srcDataPct%)\n";
print "$data bytes of data in data ($dataPct%)\n";