genoffsets - zygaena/hammerhead

root / base / usr / src / tools / scripts / genoffsets
genoffsets Plain Text 353 lines 10.8 KB
#!/usr/bin/perl
#
# CDDL HEADER START
#
# The contents of this file are subject to the terms of the
# Common Development and Distribution License, Version 1.0 only
# (the "License").  You may not use this file except in compliance
# with the License.
#
# You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
# or http://www.opensolaris.org/os/licensing.
# See the License for the specific language governing permissions
# and limitations under the License.
#
# When distributing Covered Code, include this CDDL HEADER in each
# file and include the License file at usr/src/OPENSOLARIS.LICENSE.
# If applicable, add the following below this CDDL HEADER, with the
# fields enclosed by brackets "[]" replaced with your own identifying
# information: Portions Copyright [yyyy] [name of copyright owner]
#
# CDDL HEADER END
#
#
# Copyright 2004 Sun Microsystems, Inc.  All rights reserved.
# Use is subject to license terms.
#
# ident	"%Z%%M%	%I%	%E% SMI"
#

#
# ctfstabs requires an object file with CTF data, and a file containing
# directives which indicate the types and members for which offsets are to
# be generated.  The developer provides a single input file containing both
# the #include directives used to generate the object file as well as the
# offset directives.  This script automates the splitting of the master file,
# the generation of the object file, the invocation of ctfstabs, and cleanup.
#

use strict;
use warnings;
use File::Basename;
use Getopt::Std;
use POSIX qw(:sys_wait_h);

# Globals.
our $PROGNAME = basename($0);
our ($CTmp, $OTmp, $GenTmp, $GenPPTmp, $Keep, $Verbose);
our ($SizeofTmp, $SizeofExe);  # Hammerhead: sizeof helper temps

sub usage {
	print STDERR "Usage: $PROGNAME [-k] [-s ctfstabs] [-r ctfconvert] ",
	  "compiler [options]\n";
	print STDERR "  NOTE: compiler options must enable stabs or DWARF as ",
	  "appropriate\n";
	exit(2);
}

sub cleanup {
	return if ($Keep);

	unlink($CTmp) if (defined $CTmp && -f $CTmp);
	unlink($OTmp) if (defined $OTmp && -f $OTmp);
	unlink($GenTmp) if (defined $GenTmp && -f $GenTmp);
	unlink($GenPPTmp) if (defined $GenPPTmp && -f $GenPPTmp);
	# Hammerhead: Clean up sizeof helper files
	unlink($SizeofTmp) if (defined $SizeofTmp && -f $SizeofTmp);
	unlink($SizeofExe) if (defined $SizeofExe && -f $SizeofExe);
}

sub bail {
	print STDERR "$PROGNAME: ", join(" ", @_), "\n";
	cleanup();
	exit(1);
}


sub findprog {
	my ($arg, $name, $default) = @_;

	if (defined $arg) {
		return ($arg);
	} elsif (defined $ENV{$name}) {
		return ($ENV{$name});
	} else {
		return ($default);
	}
}

sub runit {
	my (@argv) = @_;
	my $rc;

	if ($Verbose) {
		print STDERR "+ @argv\n";
	}
	if ((my $rc = system(@argv)) == -1) {
		bail("Failed to execute $argv[0]: $!");
	} elsif (WIFEXITED($rc)) {
		$_ = WEXITSTATUS($rc);
		if ($_ == 0) {
			return;
		} else {
			bail("$argv[0] failed with status $_");
		}
	} elsif (WSIGNALLED($rc)) {
		$_ = WTERMSIG($rc);
		# WCOREDUMP isn't a POSIX macro, do it the non-portable way.
		if ($rc & 0x80) {
			bail("$argv[0] failed with signal $_ (core dumped)");
		} else {
			bail("$argv[0] failed with signal $_");
		}
	}
}

#
# Main.
#

my %opts;
getopts("kr:s:v", \%opts) || usage();
usage() if (@ARGV < 1);

my $ctfstabs = findprog($opts{"s"}, "CTFSTABS", "ctfstabs");
my $ctfconvert = findprog($opts{"r"}, "CTFCONVERT", "ctfconvert");

$Keep = $opts{k};
$Verbose = $opts{k} || $opts{v};
my ($cc, @cflags) = @ARGV;

$CTmp = "ctfstabs.tmp.$$.c";		# The C file used to generate CTF
$OTmp = "ctfstabs.tmp.$$.o";		# Object file with CTF
$GenTmp = "ctfstabs.tmp.$$.gen.c";	# genassym directives
$GenPPTmp = "ctfstabs.tmp.$$.genpp";	# Post-processed genassym directives
$SizeofTmp = "ctfstabs.tmp.$$.sizeof.c";	# Hammerhead: sizeof helper C file
$SizeofExe = "ctfstabs.tmp.$$.sizeof";	# Hammerhead: sizeof helper executable

my ($cfile, $genfile);
open($cfile, '>', $CTmp) || bail("failed to create $CTmp: $!");
open($genfile, '>', $GenTmp) || bail("failed to create $GenTmp: $!");

if ($Verbose) {
	print STDERR "Splitting from stdin to $CTmp and $GenTmp\n";
}

# Hammerhead: Collect type names to force them into debug info
my @typenames;

# Hammerhead: Track types that only need sizeof (no member offsets)
# Format: { typename => SIZEOF_MACRO }
my %sizeof_only_types;
my @input_lines;  # Save all input lines for parsing

# Hammerhead: Add __EXTENSIONS__ to enable all type definitions (e.g., sigjmp_buf)
print $cfile "#define __EXTENSIONS__ 1\n";

while (<STDIN>) {
	push @input_lines, $_;
	# #includes go to the C file.  All other preprocessor directives
	# go to both the C file and the offsets input file.  Anything
	# that's not a preprocessor directive goes into the offsets input
	# file.  Also strip comments from the genfile, as they can confuse
	# the preprocessor.
	if (/^#include/) {
		print $cfile $_;
	} elsif (/^#/) {
		print $cfile $_;
		print $genfile $_;
	} elsif (/^\\#/) {
		print $genfile $_;
	} elsif (!/^\\/) {
		# Hammerhead: Preserve blank lines using a marker that survives preprocessing.
		# ctfstabs uses blank lines to separate type stanzas.
		if (/^\s*$/) {
			print $genfile "__GENOFFSETS_BLANK__\n";
		} else {
			print $genfile $_;
			# Hammerhead: Capture type names (lines starting with non-whitespace
			# that aren't preprocessor directives). Type names are at the start
			# of lines, optionally followed by a size macro name.
			if (/^([A-Za-z_][A-Za-z0-9_]*)(\s|$)/) {
				push @typenames, $1;
			}
		}
	}
}

# Hammerhead: Identify types that only need sizeof (no member fields following).
# These are types followed by a blank line or EOF, with only a size macro.
# Parse the input to find patterns like:
#   typename  SIZEOF_TYPENAME
#   <blank line or EOF>
for (my $i = 0; $i < @input_lines; $i++) {
	my $line = $input_lines[$i];
	# Look for type lines with a size macro: "typename SIZEOF_*"
	if ($line =~ /^([A-Za-z_][A-Za-z0-9_]*)\s+(SIZEOF_[A-Za-z0-9_]+)\s*$/) {
		my ($typename, $sizemacro) = ($1, $2);
		# Check if next non-comment, non-preprocessor line is blank or starts a new type
		my $next_is_member = 0;
		for (my $j = $i + 1; $j < @input_lines; $j++) {
			my $nextline = $input_lines[$j];
			next if $nextline =~ /^\\/ || $nextline =~ /^#/;  # Skip comments/preprocessor
			if ($nextline =~ /^\s+\S/) {
				# Indented line = member field
				$next_is_member = 1;
			}
			last;  # Only check first relevant line
		}
		if (!$next_is_member) {
			$sizeof_only_types{$typename} = $sizemacro;
			if ($Verbose) {
				print STDERR "Detected sizeof-only type: $typename -> $sizemacro\n";
			}
		}
	}
}

# Note: We don't need to force types into debug info with dummy declarations.
# The types are already present in the CTF data from the compiled object,
# and ctfstabs's find_type() function handles struct/union/enum prefixes.

close($cfile) || bail("can't close $CTmp: $!");
close($genfile) || bail("can't close $GenTmp: $!");

# Compile the C file.
runit($cc, @cflags, '-c', '-o', $OTmp, $CTmp);

# Convert the debugging information to CTF.
runit($ctfconvert, '-l', 'ctfstabs', $OTmp);

# Run ctfstabs on the resulting mess.
# Hammerhead: Use -x assembler-with-cpp for preprocessing. GCC 14 otherwise fails
# on bare type names in the genassym directives (e.g. "proc PROCSIZE").
runit($cc, @cflags, "-x", "assembler-with-cpp", "-E", "-P", "-o", "$GenPPTmp.tmp", $GenTmp);

# Hammerhead: Restore blank lines from __GENOFFSETS_BLANK__ markers.
# GCC preprocessor strips blank lines, but ctfstabs needs them to separate type stanzas.
open(my $ppin, '<', "$GenPPTmp.tmp") || bail("can't open $GenPPTmp.tmp: $!");
open(my $ppout, '>', $GenPPTmp) || bail("can't create $GenPPTmp: $!");
while (<$ppin>) {
	s/^__GENOFFSETS_BLANK__\s*$/\n/;
	print $ppout $_;
}
close($ppin);
close($ppout);
unlink("$GenPPTmp.tmp") unless $Keep;

# Hammerhead: Generate sizeof values for types that don't have members.
# CTF/ctfstabs can't handle array typedefs (like sigjmp_buf), so we
# compute their sizes directly via a helper program.
my %sizeof_values;
if (%sizeof_only_types) {
	if ($Verbose) {
		print STDERR "Generating sizeof helper for: ", join(", ", keys %sizeof_only_types), "\n";
	}

	# Generate sizeof helper C program
	open(my $sizeoffile, '>', $SizeofTmp) || bail("can't create $SizeofTmp: $!");
	print $sizeoffile "#define __EXTENSIONS__ 1\n";
	print $sizeoffile "#include <stdio.h>\n";
	# Copy the #include lines from the original input
	foreach my $line (@input_lines) {
		if ($line =~ /^#include/) {
			print $sizeoffile $line;
		}
	}
	print $sizeoffile "int main(void) {\n";
	foreach my $type (keys %sizeof_only_types) {
		my $macro = $sizeof_only_types{$type};
		print $sizeoffile "\tprintf(\"$macro\\t0x%lx\\n\", (unsigned long)sizeof($type));\n";
	}
	print $sizeoffile "\treturn 0;\n}\n";
	close($sizeoffile);

	# Compile and run the sizeof helper (use native compiler, not cross-compile flags)
	# Filter out cross-compilation flags that would prevent running the helper
	my @native_flags = grep { !/^-m64$/ && !/^-xarch/ && !/^-xmodel/ } @cflags;
	# Remove -g flags (not needed for sizeof helper)
	@native_flags = grep { !/^-g/ } @native_flags;
	# Keep only include paths and defines
	@native_flags = grep { /^-I/ || /^-D/ } @native_flags;

	if ($Verbose) {
		print STDERR "+ $cc @native_flags -o $SizeofExe $SizeofTmp\n";
	}
	my $rc = system($cc, @native_flags, '-o', $SizeofExe, $SizeofTmp);
	if ($rc == 0) {
		if ($Verbose) {
			print STDERR "+ ./$SizeofExe\n";
		}
		open(my $sizeout, '-|', "./$SizeofExe") || bail("can't run $SizeofExe: $!");
		while (<$sizeout>) {
			chomp;
			if (/^(\S+)\t(\S+)$/) {
				$sizeof_values{$1} = $2;
				if ($Verbose) {
					print STDERR "Got sizeof: $1 = $2\n";
				}
			}
		}
		close($sizeout);
	} else {
		print STDERR "$PROGNAME: warning: sizeof helper failed, trying ctfstabs for all types\n";
	}

	# Clean up sizeof helper files
	unless ($Keep) {
		unlink($SizeofTmp) if -f $SizeofTmp;
		unlink($SizeofExe) if -f $SizeofExe;
	}
}

# Hammerhead: If we have sizeof values for types that ctfstabs can't handle,
# filter them out of the genpp file before running ctfstabs.
my $GenPPFiltered = $GenPPTmp;
if (%sizeof_values) {
	$GenPPFiltered = "ctfstabs.tmp.$$.genpp.filtered";
	open(my $gpin, '<', $GenPPTmp) || bail("can't open $GenPPTmp: $!");
	open(my $gpout, '>', $GenPPFiltered) || bail("can't create $GenPPFiltered: $!");
	my $skip_type = 0;
	while (<$gpin>) {
		# Check if this line starts a sizeof-only type that we've handled
		if (/^([A-Za-z_][A-Za-z0-9_]*)\s+(SIZEOF_[A-Za-z0-9_]+)\s*$/) {
			if (exists $sizeof_values{$2}) {
				$skip_type = 1;
				next;
			}
		}
		# Blank line ends any type section
		if (/^\s*$/) {
			$skip_type = 0;
		}
		print $gpout $_ unless $skip_type;
	}
	close($gpin);
	close($gpout);
}

# Run ctfstabs
runit($ctfstabs, "-t", "genassym", "-i", $GenPPFiltered, $OTmp);

# Hammerhead: Output sizeof values we computed directly
foreach my $macro (sort keys %sizeof_values) {
	print "#define\t$macro\t$sizeof_values{$macro}\n";
}

# Clean up filtered file
unlink($GenPPFiltered) if ($GenPPFiltered ne $GenPPTmp && !$Keep);

cleanup();

exit (0);