zig/tools/update_glibc.zig
2021-02-27 13:11:47 +02:00

342 lines
13 KiB
Zig

const std = @import("std");
const fs = std.fs;
const fmt = std.fmt;
const assert = std.debug.assert;
// Example abilist path:
// ./sysdeps/unix/sysv/linux/aarch64/libc.abilist
const AbiList = struct {
targets: []const ZigTarget,
path: []const u8,
};
const ZigTarget = struct {
arch: std.Target.Cpu.Arch,
abi: std.Target.Abi,
};
const lib_names = [_][]const u8{
"c",
"dl",
"m",
"pthread",
"rt",
"ld",
"util",
};
// fpu/nofpu are hardcoded elsewhere, based on .gnueabi/.gnueabihf with an exception for .arm
// n64/n32 are hardcoded elsewhere, based on .gnuabi64/.gnuabin32
const abi_lists = [_]AbiList{
AbiList{
.targets = &[_]ZigTarget{
ZigTarget{ .arch = .aarch64, .abi = .gnu },
ZigTarget{ .arch = .aarch64_be, .abi = .gnu },
},
.path = "aarch64",
},
AbiList{
.targets = &[_]ZigTarget{ZigTarget{ .arch = .s390x, .abi = .gnu }},
.path = "s390/s390-64",
},
AbiList{
.targets = &[_]ZigTarget{
ZigTarget{ .arch = .arm, .abi = .gnueabi },
ZigTarget{ .arch = .armeb, .abi = .gnueabi },
ZigTarget{ .arch = .arm, .abi = .gnueabihf },
ZigTarget{ .arch = .armeb, .abi = .gnueabihf },
},
.path = "arm",
},
AbiList{
.targets = &[_]ZigTarget{
ZigTarget{ .arch = .sparc, .abi = .gnu },
ZigTarget{ .arch = .sparcel, .abi = .gnu },
},
.path = "sparc/sparc32",
},
AbiList{
.targets = &[_]ZigTarget{ZigTarget{ .arch = .sparcv9, .abi = .gnu }},
.path = "sparc/sparc64",
},
AbiList{
.targets = &[_]ZigTarget{
ZigTarget{ .arch = .mips64el, .abi = .gnuabi64 },
ZigTarget{ .arch = .mips64, .abi = .gnuabi64 },
},
.path = "mips/mips64",
},
AbiList{
.targets = &[_]ZigTarget{
ZigTarget{ .arch = .mips64el, .abi = .gnuabin32 },
ZigTarget{ .arch = .mips64, .abi = .gnuabin32 },
},
.path = "mips/mips64",
},
AbiList{
.targets = &[_]ZigTarget{
ZigTarget{ .arch = .mipsel, .abi = .gnueabihf },
ZigTarget{ .arch = .mips, .abi = .gnueabihf },
},
.path = "mips/mips32",
},
AbiList{
.targets = &[_]ZigTarget{
ZigTarget{ .arch = .mipsel, .abi = .gnueabi },
ZigTarget{ .arch = .mips, .abi = .gnueabi },
},
.path = "mips/mips32",
},
AbiList{
.targets = &[_]ZigTarget{ZigTarget{ .arch = .x86_64, .abi = .gnu }},
.path = "x86_64/64",
},
AbiList{
.targets = &[_]ZigTarget{ZigTarget{ .arch = .x86_64, .abi = .gnux32 }},
.path = "x86_64/x32",
},
AbiList{
.targets = &[_]ZigTarget{ZigTarget{ .arch = .i386, .abi = .gnu }},
.path = "i386",
},
AbiList{
.targets = &[_]ZigTarget{ZigTarget{ .arch = .powerpc64le, .abi = .gnu }},
.path = "powerpc/powerpc64/le",
},
AbiList{
.targets = &[_]ZigTarget{ZigTarget{ .arch = .powerpc64, .abi = .gnu }},
.path = "powerpc/powerpc64/be",
},
AbiList{
.targets = &[_]ZigTarget{
ZigTarget{ .arch = .powerpc, .abi = .gnueabi },
ZigTarget{ .arch = .powerpc, .abi = .gnueabihf },
},
.path = "powerpc/powerpc32",
},
};
const FunctionSet = struct {
list: std.ArrayList(VersionedFn),
fn_vers_list: FnVersionList,
};
const FnVersionList = std.StringHashMap(std.ArrayList(usize));
const VersionedFn = struct {
ver: []const u8, // example: "GLIBC_2.15"
name: []const u8, // example: "puts"
};
const Function = struct {
name: []const u8, // example: "puts"
lib: []const u8, // example: "c"
index: usize,
};
pub fn main() !void {
var arena = std.heap.ArenaAllocator.init(std.heap.page_allocator);
const allocator = &arena.allocator;
const args = try std.process.argsAlloc(allocator);
const in_glibc_dir = args[1]; // path to the unzipped tarball of glibc, e.g. ~/downloads/glibc-2.25
const zig_src_dir = args[2]; // path to the source checkout of zig, lib dir, e.g. ~/zig-src/lib
const prefix = try fs.path.join(allocator, &[_][]const u8{ in_glibc_dir, "sysdeps", "unix", "sysv", "linux" });
const glibc_out_dir = try fs.path.join(allocator, &[_][]const u8{ zig_src_dir, "libc", "glibc" });
var global_fn_set = std.StringHashMap(Function).init(allocator);
var global_ver_set = std.StringHashMap(usize).init(allocator);
var target_functions = std.AutoHashMap(usize, FunctionSet).init(allocator);
for (abi_lists) |*abi_list| {
const target_funcs_gop = try target_functions.getOrPut(@ptrToInt(abi_list));
if (!target_funcs_gop.found_existing) {
target_funcs_gop.entry.value = FunctionSet{
.list = std.ArrayList(VersionedFn).init(allocator),
.fn_vers_list = FnVersionList.init(allocator),
};
}
const fn_set = &target_funcs_gop.entry.value.list;
for (lib_names) |lib_name, lib_name_index| {
const lib_prefix = if (std.mem.eql(u8, lib_name, "ld")) "" else "lib";
const basename = try fmt.allocPrint(allocator, "{s}{s}.abilist", .{ lib_prefix, lib_name });
const abi_list_filename = blk: {
const is_c = std.mem.eql(u8, lib_name, "c");
const is_m = std.mem.eql(u8, lib_name, "m");
const is_ld = std.mem.eql(u8, lib_name, "ld");
if (abi_list.targets[0].abi == .gnuabi64 and (is_c or is_ld)) {
break :blk try fs.path.join(allocator, &[_][]const u8{ prefix, abi_list.path, "n64", basename });
} else if (abi_list.targets[0].abi == .gnuabin32 and (is_c or is_ld)) {
break :blk try fs.path.join(allocator, &[_][]const u8{ prefix, abi_list.path, "n32", basename });
} else if (abi_list.targets[0].arch != .arm and
abi_list.targets[0].abi == .gnueabihf and
(is_c or (is_m and abi_list.targets[0].arch == .powerpc)))
{
break :blk try fs.path.join(allocator, &[_][]const u8{ prefix, abi_list.path, "fpu", basename });
} else if (abi_list.targets[0].arch != .arm and
abi_list.targets[0].abi == .gnueabi and
(is_c or (is_m and abi_list.targets[0].arch == .powerpc)))
{
break :blk try fs.path.join(allocator, &[_][]const u8{ prefix, abi_list.path, "nofpu", basename });
} else if (abi_list.targets[0].arch == .arm) {
break :blk try fs.path.join(allocator, &[_][]const u8{ prefix, abi_list.path, "le", basename });
} else if (abi_list.targets[0].arch == .armeb) {
break :blk try fs.path.join(allocator, &[_][]const u8{ prefix, abi_list.path, "be", basename });
}
break :blk try fs.path.join(allocator, &[_][]const u8{ prefix, abi_list.path, basename });
};
const max_bytes = 10 * 1024 * 1024;
const contents = std.fs.cwd().readFileAlloc(allocator, abi_list_filename, max_bytes) catch |err| {
std.debug.warn("unable to open {s}: {}\n", .{ abi_list_filename, err });
std.process.exit(1);
};
var lines_it = std.mem.tokenize(contents, "\n");
while (lines_it.next()) |line| {
var tok_it = std.mem.tokenize(line, " ");
const ver = tok_it.next().?;
const name = tok_it.next().?;
const category = tok_it.next().?;
if (!std.mem.eql(u8, category, "F") and
!std.mem.eql(u8, category, "D"))
{
continue;
}
if (std.mem.startsWith(u8, ver, "GCC_")) continue;
try global_ver_set.put(ver, undefined);
const gop = try global_fn_set.getOrPut(name);
if (gop.found_existing) {
if (!std.mem.eql(u8, gop.entry.value.lib, "c")) {
gop.entry.value.lib = lib_name;
}
} else {
gop.entry.value = Function{
.name = name,
.lib = lib_name,
.index = undefined,
};
}
try fn_set.append(VersionedFn{
.ver = ver,
.name = name,
});
}
}
}
const global_fn_list = blk: {
var list = std.ArrayList([]const u8).init(allocator);
var it = global_fn_set.iterator();
while (it.next()) |entry| try list.append(entry.key);
std.sort.sort([]const u8, list.items, {}, strCmpLessThan);
break :blk list.items;
};
const global_ver_list = blk: {
var list = std.ArrayList([]const u8).init(allocator);
var it = global_ver_set.iterator();
while (it.next()) |entry| try list.append(entry.key);
std.sort.sort([]const u8, list.items, {}, versionLessThan);
break :blk list.items;
};
{
const vers_txt_path = try fs.path.join(allocator, &[_][]const u8{ glibc_out_dir, "vers.txt" });
const vers_txt_file = try fs.cwd().createFile(vers_txt_path, .{});
defer vers_txt_file.close();
var buffered = std.io.bufferedWriter(vers_txt_file.writer());
const vers_txt = buffered.writer();
for (global_ver_list) |name, i| {
global_ver_set.put(name, i) catch unreachable;
try vers_txt.print("{s}\n", .{name});
}
try buffered.flush();
}
{
const fns_txt_path = try fs.path.join(allocator, &[_][]const u8{ glibc_out_dir, "fns.txt" });
const fns_txt_file = try fs.cwd().createFile(fns_txt_path, .{});
defer fns_txt_file.close();
var buffered = std.io.bufferedWriter(fns_txt_file.writer());
const fns_txt = buffered.writer();
for (global_fn_list) |name, i| {
const entry = global_fn_set.getEntry(name).?;
entry.value.index = i;
try fns_txt.print("{s} {s}\n", .{ name, entry.value.lib });
}
try buffered.flush();
}
// Now the mapping of version and function to integer index is complete.
// Here we create a mapping of function name to list of versions.
for (abi_lists) |*abi_list, abi_index| {
const entry = target_functions.getEntry(@ptrToInt(abi_list)).?;
const fn_vers_list = &entry.value.fn_vers_list;
for (entry.value.list.items) |*ver_fn| {
const gop = try fn_vers_list.getOrPut(ver_fn.name);
if (!gop.found_existing) {
gop.entry.value = std.ArrayList(usize).init(allocator);
}
const ver_index = global_ver_set.getEntry(ver_fn.ver).?.value;
if (std.mem.indexOfScalar(usize, gop.entry.value.items, ver_index) == null) {
try gop.entry.value.append(ver_index);
}
}
}
{
const abilist_txt_path = try fs.path.join(allocator, &[_][]const u8{ glibc_out_dir, "abi.txt" });
const abilist_txt_file = try fs.cwd().createFile(abilist_txt_path, .{});
defer abilist_txt_file.close();
var buffered = std.io.bufferedWriter(abilist_txt_file.writer());
const abilist_txt = buffered.writer();
// first iterate over the abi lists
for (abi_lists) |*abi_list, abi_index| {
const fn_vers_list = &target_functions.getEntry(@ptrToInt(abi_list)).?.value.fn_vers_list;
for (abi_list.targets) |target, it_i| {
if (it_i != 0) try abilist_txt.writeByte(' ');
try abilist_txt.print("{s}-linux-{s}", .{ @tagName(target.arch), @tagName(target.abi) });
}
try abilist_txt.writeByte('\n');
// next, each line implicitly corresponds to a function
for (global_fn_list) |name| {
const entry = fn_vers_list.getEntry(name) orelse {
try abilist_txt.writeByte('\n');
continue;
};
for (entry.value.items) |ver_index, it_i| {
if (it_i != 0) try abilist_txt.writeByte(' ');
try abilist_txt.print("{d}", .{ver_index});
}
try abilist_txt.writeByte('\n');
}
}
try buffered.flush();
}
}
pub fn strCmpLessThan(context: void, a: []const u8, b: []const u8) bool {
return std.mem.order(u8, a, b) == .lt;
}
pub fn versionLessThan(context: void, a: []const u8, b: []const u8) bool {
const sep_chars = "GLIBC_.";
var a_tokens = std.mem.tokenize(a, sep_chars);
var b_tokens = std.mem.tokenize(b, sep_chars);
while (true) {
const a_next = a_tokens.next();
const b_next = b_tokens.next();
if (a_next == null and b_next == null) {
return false; // equal means not less than
} else if (a_next == null) {
return true;
} else if (b_next == null) {
return false;
}
const a_int = fmt.parseInt(u64, a_next.?, 10) catch unreachable;
const b_int = fmt.parseInt(u64, b_next.?, 10) catch unreachable;
if (a_int < b_int) {
return true;
} else if (a_int > b_int) {
return false;
}
}
}