zig/tools/gen_spirv_spec.zig
Robin Voetter 074ba69ba6
spirv: assembler
spirv: introduce SpvModule.Fn to generate function code into

spirv: assembler error message setup

spirv: runtime spec info

spirv: inline assembly tokenizer

spirv: inline assembly lhs result/opcode parsing

spirv: forgot to fmt

spirv: tokenize opcodes and assigned result-ids

spirv: operand parsing setup

spirv: assembler string literals

spirv: assembler integer literals

spirv: assembler value enums

spirv: assembler bit masks

spirv: update assembler to new asm air format

spirv: target 1.5 for now

Current vulkan sdk version (1.3.204) ships spirv tools targetting 1.5,
and so these do not work with binaries targetting 1.6 yet. In the
future, this version number should be decided by the target.

spirv: store operands in flat arraylist.

Instead of having dedicated Operand variants for variadic operands,
just flatten them and store them in the normal inst.operands list.
This is a little simpler, but is not easily decodable in the operand
data representation.

spirv: parse variadic assembly operands

spirv: improve assembler result-id tokenization

spirv: begin instruction processing

spirv: only remove decl if it was actually allocated

spirv: work around weird miscompilation

Seems like there are problems with switch in anonymous struct literals.

spirv: begin resolving some types in assembler

spirv: improve instruction processing

spirv: rename some types + process OpTypeInt

spirv: process OpTypeVector

spirv: process OpTypeMatrix and OpTypeSampler

spirv: add opcode class to spec, remove @exclude'd instructions

spirv: process more type instructions

spirv: OpTypeFunction

spirv: OpTypeOpaque

spirv: parse LiteralContextDependentNumber operands

spirv: emit assembly instruction into right section

spirv: parse OpPhi parameters

spirv: inline assembly inputs

spirv: also copy air types

spirv: inline assembly outputs

spirv: spir-v address spaces

spirv: basic vector constants/types and shuffle

spirv: assembler OpTypeImage

spirv: some stuff

spirv: remove spirv address spaces for now
2022-11-23 19:17:58 +01:00

693 lines
22 KiB
Zig

const std = @import("std");
const g = @import("spirv/grammar.zig");
const Allocator = std.mem.Allocator;
const ExtendedStructSet = std.StringHashMap(void);
pub fn main() !void {
var arena = std.heap.ArenaAllocator.init(std.heap.page_allocator);
defer arena.deinit();
const allocator = arena.allocator();
const args = try std.process.argsAlloc(allocator);
if (args.len != 2) {
usageAndExit(std.io.getStdErr(), args[0], 1);
}
const spec_path = args[1];
const spec = try std.fs.cwd().readFileAlloc(allocator, spec_path, std.math.maxInt(usize));
// Required for json parsing.
@setEvalBranchQuota(10000);
var tokens = std.json.TokenStream.init(spec);
var registry = try std.json.parse(g.Registry, &tokens, .{ .allocator = allocator });
const core_reg = switch (registry) {
.core => |core_reg| core_reg,
.extension => return error.TODOSpirVExtensionSpec,
};
var bw = std.io.bufferedWriter(std.io.getStdOut().writer());
try render(bw.writer(), allocator, core_reg);
try bw.flush();
}
/// Returns a set with types that require an extra struct for the `Instruction` interface
/// to the spir-v spec, or whether the original type can be used.
fn extendedStructs(
arena: Allocator,
kinds: []const g.OperandKind,
) !ExtendedStructSet {
var map = ExtendedStructSet.init(arena);
try map.ensureTotalCapacity(@intCast(u32, kinds.len));
for (kinds) |kind| {
const enumerants = kind.enumerants orelse continue;
for (enumerants) |enumerant| {
if (enumerant.parameters.len > 0) {
break;
}
} else continue;
map.putAssumeCapacity(kind.kind, {});
}
return map;
}
// Return a score for a particular priority. Duplicate instruction/operand enum values are
// removed by picking the tag with the lowest score to keep, and by making an alias for the
// other. Note that the tag does not need to be just a tag at this point, in which case it
// gets the lowest score automatically anyway.
fn tagPriorityScore(tag: []const u8) usize {
if (tag.len == 0) {
return 1;
} else if (std.mem.eql(u8, tag, "EXT")) {
return 2;
} else if (std.mem.eql(u8, tag, "KHR")) {
return 3;
} else {
return 4;
}
}
fn render(writer: anytype, allocator: Allocator, registry: g.CoreRegistry) !void {
try writer.writeAll(
\\//! This file is auto-generated by tools/gen_spirv_spec.zig.
\\
\\const Version = @import("std").builtin.Version;
\\
\\pub const Word = u32;
\\pub const IdResultType = struct{
\\ id: Word,
\\ pub fn toRef(self: IdResultType) IdRef {
\\ return .{.id = self.id};
\\ }
\\};
\\pub const IdResult = struct{
\\ id: Word,
\\ pub fn toRef(self: IdResult) IdRef {
\\ return .{.id = self.id};
\\ }
\\ pub fn toResultType(self: IdResult) IdResultType {
\\ return .{.id = self.id};
\\ }
\\};
\\pub const IdRef = struct{ id: Word };
\\
\\pub const IdMemorySemantics = IdRef;
\\pub const IdScope = IdRef;
\\
\\pub const LiteralInteger = Word;
\\pub const LiteralString = []const u8;
\\pub const LiteralContextDependentNumber = union(enum) {
\\ int32: i32,
\\ uint32: u32,
\\ int64: i64,
\\ uint64: u64,
\\ float32: f32,
\\ float64: f64,
\\};
\\pub const LiteralExtInstInteger = struct{ inst: Word };
\\pub const LiteralSpecConstantOpInteger = struct { opcode: Opcode };
\\pub const PairLiteralIntegerIdRef = struct { value: LiteralInteger, label: IdRef };
\\pub const PairIdRefLiteralInteger = struct { target: IdRef, member: LiteralInteger };
\\pub const PairIdRefIdRef = [2]IdRef;
\\
\\pub const Quantifier = enum {
\\ required,
\\ optional,
\\ variadic,
\\};
\\
\\pub const Operand = struct {
\\ kind: OperandKind,
\\ quantifier: Quantifier,
\\};
\\
\\pub const OperandCategory = enum {
\\ bit_enum,
\\ value_enum,
\\ id,
\\ literal,
\\ composite,
\\};
\\
\\pub const Enumerant = struct {
\\ name: []const u8,
\\ value: Word,
\\ parameters: []const OperandKind,
\\};
\\
\\
);
try writer.print(
\\pub const version = Version{{ .major = {}, .minor = {}, .patch = {} }};
\\pub const magic_number: Word = {s};
\\
\\
,
.{ registry.major_version, registry.minor_version, registry.revision, registry.magic_number },
);
const extended_structs = try extendedStructs(allocator, registry.operand_kinds);
try renderClass(writer, allocator, registry.instructions);
try renderOperandKind(writer, registry.operand_kinds);
try renderOpcodes(writer, allocator, registry.instructions, extended_structs);
try renderOperandKinds(writer, allocator, registry.operand_kinds, extended_structs);
}
fn renderClass(writer: anytype, allocator: Allocator, instructions: []const g.Instruction) !void {
var class_map = std.StringArrayHashMap(void).init(allocator);
for (instructions) |inst| {
if (std.mem.eql(u8, inst.class.?, "@exclude")) {
continue;
}
try class_map.put(inst.class.?, {});
}
try writer.writeAll("pub const Class = enum {\n");
for (class_map.keys()) |class| {
try renderInstructionClass(writer, class);
try writer.writeAll(",\n");
}
try writer.writeAll("};\n");
}
fn renderInstructionClass(writer: anytype, class: []const u8) !void {
// Just assume that these wont clobber zig builtin types.
var prev_was_sep = true;
for (class) |c| {
switch (c) {
'-', '_' => prev_was_sep = true,
else => if (prev_was_sep) {
try writer.writeByte(std.ascii.toUpper(c));
prev_was_sep = false;
} else {
try writer.writeByte(std.ascii.toLower(c));
},
}
}
}
fn renderOperandKind(writer: anytype, operands: []const g.OperandKind) !void {
try writer.writeAll("pub const OperandKind = enum {\n");
for (operands) |operand| {
try writer.print("{},\n", .{std.zig.fmtId(operand.kind)});
}
try writer.writeAll(
\\
\\pub fn category(self: OperandKind) OperandCategory {
\\return switch (self) {
\\
);
for (operands) |operand| {
const cat = switch (operand.category) {
.BitEnum => "bit_enum",
.ValueEnum => "value_enum",
.Id => "id",
.Literal => "literal",
.Composite => "composite",
};
try writer.print(".{} => .{s},\n", .{ std.zig.fmtId(operand.kind), cat });
}
try writer.writeAll(
\\};
\\}
\\pub fn enumerants(self: OperandKind) []const Enumerant {
\\return switch (self) {
\\
);
for (operands) |operand| {
switch (operand.category) {
.BitEnum, .ValueEnum => {},
else => {
try writer.print(".{} => unreachable,\n", .{std.zig.fmtId(operand.kind)});
continue;
},
}
try writer.print(".{} => &[_]Enumerant{{", .{std.zig.fmtId(operand.kind)});
for (operand.enumerants.?) |enumerant| {
if (enumerant.value == .bitflag and std.mem.eql(u8, enumerant.enumerant, "None")) {
continue;
}
try renderEnumerant(writer, enumerant);
try writer.writeAll(",");
}
try writer.writeAll("},\n");
}
try writer.writeAll("};\n}\n};\n");
}
fn renderEnumerant(writer: anytype, enumerant: g.Enumerant) !void {
try writer.print(".{{.name = \"{s}\", .value = ", .{enumerant.enumerant});
switch (enumerant.value) {
.bitflag => |flag| try writer.writeAll(flag),
.int => |int| try writer.print("{}", .{int}),
}
try writer.writeAll(", .parameters = &[_]OperandKind{");
for (enumerant.parameters) |param, i| {
if (i != 0)
try writer.writeAll(", ");
// Note, param.quantifier will always be one.
try writer.print(".{}", .{std.zig.fmtId(param.kind)});
}
try writer.writeAll("}}");
}
fn renderOpcodes(
writer: anytype,
allocator: Allocator,
instructions: []const g.Instruction,
extended_structs: ExtendedStructSet,
) !void {
var inst_map = std.AutoArrayHashMap(u32, usize).init(allocator);
try inst_map.ensureTotalCapacity(instructions.len);
var aliases = std.ArrayList(struct { inst: usize, alias: usize }).init(allocator);
try aliases.ensureTotalCapacity(instructions.len);
for (instructions) |inst, i| {
if (std.mem.eql(u8, inst.class.?, "@exclude")) {
continue;
}
const result = inst_map.getOrPutAssumeCapacity(inst.opcode);
if (!result.found_existing) {
result.value_ptr.* = i;
continue;
}
const existing = instructions[result.value_ptr.*];
const tag_index = std.mem.indexOfDiff(u8, inst.opname, existing.opname).?;
const inst_priority = tagPriorityScore(inst.opname[tag_index..]);
const existing_priority = tagPriorityScore(existing.opname[tag_index..]);
if (inst_priority < existing_priority) {
aliases.appendAssumeCapacity(.{ .inst = result.value_ptr.*, .alias = i });
result.value_ptr.* = i;
} else {
aliases.appendAssumeCapacity(.{ .inst = i, .alias = result.value_ptr.* });
}
}
const instructions_indices = inst_map.values();
try writer.writeAll("pub const Opcode = enum(u16) {\n");
for (instructions_indices) |i| {
const inst = instructions[i];
try writer.print("{} = {},\n", .{ std.zig.fmtId(inst.opname), inst.opcode });
}
try writer.writeByte('\n');
for (aliases.items) |alias| {
try writer.print("pub const {} = Opcode.{};\n", .{
std.zig.fmtId(instructions[alias.inst].opname),
std.zig.fmtId(instructions[alias.alias].opname),
});
}
try writer.writeAll(
\\
\\pub fn Operands(comptime self: Opcode) type {
\\return switch (self) {
\\
);
for (instructions_indices) |i| {
const inst = instructions[i];
try renderOperand(writer, .instruction, inst.opname, inst.operands, extended_structs);
}
try writer.writeAll(
\\};
\\}
\\pub fn operands(self: Opcode) []const Operand {
\\return switch (self) {
\\
);
for (instructions_indices) |i| {
const inst = instructions[i];
try writer.print(".{} => &[_]Operand{{", .{std.zig.fmtId(inst.opname)});
for (inst.operands) |operand| {
const quantifier = if (operand.quantifier) |q|
switch (q) {
.@"?" => "optional",
.@"*" => "variadic",
}
else
"required";
try writer.print(".{{.kind = .{s}, .quantifier = .{s}}},", .{ operand.kind, quantifier });
}
try writer.writeAll("},\n");
}
try writer.writeAll(
\\};
\\}
\\pub fn class(self: Opcode) Class {
\\return switch (self) {
\\
);
for (instructions_indices) |i| {
const inst = instructions[i];
try writer.print(".{} => .", .{std.zig.fmtId(inst.opname)});
try renderInstructionClass(writer, inst.class.?);
try writer.writeAll(",\n");
}
try writer.writeAll("};\n}\n};\n");
}
fn renderOperandKinds(
writer: anytype,
allocator: Allocator,
kinds: []const g.OperandKind,
extended_structs: ExtendedStructSet,
) !void {
for (kinds) |kind| {
switch (kind.category) {
.ValueEnum => try renderValueEnum(writer, allocator, kind, extended_structs),
.BitEnum => try renderBitEnum(writer, allocator, kind, extended_structs),
else => {},
}
}
}
fn renderValueEnum(
writer: anytype,
allocator: Allocator,
enumeration: g.OperandKind,
extended_structs: ExtendedStructSet,
) !void {
const enumerants = enumeration.enumerants orelse return error.InvalidRegistry;
var enum_map = std.AutoArrayHashMap(u32, usize).init(allocator);
try enum_map.ensureTotalCapacity(enumerants.len);
var aliases = std.ArrayList(struct { enumerant: usize, alias: usize }).init(allocator);
try aliases.ensureTotalCapacity(enumerants.len);
for (enumerants) |enumerant, i| {
const result = enum_map.getOrPutAssumeCapacity(enumerant.value.int);
if (!result.found_existing) {
result.value_ptr.* = i;
continue;
}
const existing = enumerants[result.value_ptr.*];
const tag_index = std.mem.indexOfDiff(u8, enumerant.enumerant, existing.enumerant).?;
const enum_priority = tagPriorityScore(enumerant.enumerant[tag_index..]);
const existing_priority = tagPriorityScore(existing.enumerant[tag_index..]);
if (enum_priority < existing_priority) {
aliases.appendAssumeCapacity(.{ .enumerant = result.value_ptr.*, .alias = i });
result.value_ptr.* = i;
} else {
aliases.appendAssumeCapacity(.{ .enumerant = i, .alias = result.value_ptr.* });
}
}
const enum_indices = enum_map.values();
try writer.print("pub const {s} = enum(u32) {{\n", .{std.zig.fmtId(enumeration.kind)});
for (enum_indices) |i| {
const enumerant = enumerants[i];
if (enumerant.value != .int) return error.InvalidRegistry;
try writer.print("{} = {},\n", .{ std.zig.fmtId(enumerant.enumerant), enumerant.value.int });
}
try writer.writeByte('\n');
for (aliases.items) |alias| {
try writer.print("pub const {} = {}.{};\n", .{
std.zig.fmtId(enumerants[alias.enumerant].enumerant),
std.zig.fmtId(enumeration.kind),
std.zig.fmtId(enumerants[alias.alias].enumerant),
});
}
if (!extended_structs.contains(enumeration.kind)) {
try writer.writeAll("};\n");
return;
}
try writer.print("\npub const Extended = union({}) {{\n", .{std.zig.fmtId(enumeration.kind)});
for (enum_indices) |i| {
const enumerant = enumerants[i];
try renderOperand(writer, .@"union", enumerant.enumerant, enumerant.parameters, extended_structs);
}
try writer.writeAll("};\n};\n");
}
fn renderBitEnum(
writer: anytype,
allocator: Allocator,
enumeration: g.OperandKind,
extended_structs: ExtendedStructSet,
) !void {
try writer.print("pub const {s} = packed struct {{\n", .{std.zig.fmtId(enumeration.kind)});
var flags_by_bitpos = [_]?usize{null} ** 32;
const enumerants = enumeration.enumerants orelse return error.InvalidRegistry;
var aliases = std.ArrayList(struct { flag: usize, alias: u5 }).init(allocator);
try aliases.ensureTotalCapacity(enumerants.len);
for (enumerants) |enumerant, i| {
if (enumerant.value != .bitflag) return error.InvalidRegistry;
const value = try parseHexInt(enumerant.value.bitflag);
if (value == 0) {
continue; // Skip 'none' items
}
std.debug.assert(@popCount(value) == 1);
var bitpos = std.math.log2_int(u32, value);
if (flags_by_bitpos[bitpos]) |*existing| {
const tag_index = std.mem.indexOfDiff(u8, enumerant.enumerant, enumerants[existing.*].enumerant).?;
const enum_priority = tagPriorityScore(enumerant.enumerant[tag_index..]);
const existing_priority = tagPriorityScore(enumerants[existing.*].enumerant[tag_index..]);
if (enum_priority < existing_priority) {
aliases.appendAssumeCapacity(.{ .flag = existing.*, .alias = bitpos });
existing.* = i;
} else {
aliases.appendAssumeCapacity(.{ .flag = i, .alias = bitpos });
}
} else {
flags_by_bitpos[bitpos] = i;
}
}
for (flags_by_bitpos) |maybe_flag_index, bitpos| {
if (maybe_flag_index) |flag_index| {
try writer.print("{}", .{std.zig.fmtId(enumerants[flag_index].enumerant)});
} else {
try writer.print("_reserved_bit_{}", .{bitpos});
}
try writer.writeAll(": bool = false,\n");
}
try writer.writeByte('\n');
for (aliases.items) |alias| {
try writer.print("pub const {}: {} = .{{.{} = true}};\n", .{
std.zig.fmtId(enumerants[alias.flag].enumerant),
std.zig.fmtId(enumeration.kind),
std.zig.fmtId(enumerants[flags_by_bitpos[alias.alias].?].enumerant),
});
}
if (!extended_structs.contains(enumeration.kind)) {
try writer.writeAll("};\n");
return;
}
try writer.print("\npub const Extended = struct {{\n", .{});
for (flags_by_bitpos) |maybe_flag_index, bitpos| {
const flag_index = maybe_flag_index orelse {
try writer.print("_reserved_bit_{}: bool = false,\n", .{bitpos});
continue;
};
const enumerant = enumerants[flag_index];
try renderOperand(writer, .mask, enumerant.enumerant, enumerant.parameters, extended_structs);
}
try writer.writeAll("};\n};\n");
}
fn renderOperand(
writer: anytype,
kind: enum {
@"union",
instruction,
mask,
},
field_name: []const u8,
parameters: []const g.Operand,
extended_structs: ExtendedStructSet,
) !void {
if (kind == .instruction) {
try writer.writeByte('.');
}
try writer.print("{}", .{std.zig.fmtId(field_name)});
if (parameters.len == 0) {
switch (kind) {
.@"union" => try writer.writeAll(",\n"),
.instruction => try writer.writeAll(" => void,\n"),
.mask => try writer.writeAll(": bool = false,\n"),
}
return;
}
if (kind == .instruction) {
try writer.writeAll(" => ");
} else {
try writer.writeAll(": ");
}
if (kind == .mask) {
try writer.writeByte('?');
}
try writer.writeAll("struct{");
for (parameters) |param, j| {
if (j != 0) {
try writer.writeAll(", ");
}
try renderFieldName(writer, parameters, j);
try writer.writeAll(": ");
if (param.quantifier) |q| {
switch (q) {
.@"?" => try writer.writeByte('?'),
.@"*" => try writer.writeAll("[]const "),
}
}
try writer.print("{}", .{std.zig.fmtId(param.kind)});
if (extended_structs.contains(param.kind)) {
try writer.writeAll(".Extended");
}
if (param.quantifier) |q| {
switch (q) {
.@"?" => try writer.writeAll(" = null"),
.@"*" => try writer.writeAll(" = &.{}"),
}
}
}
try writer.writeAll("}");
if (kind == .mask) {
try writer.writeAll(" = null");
}
try writer.writeAll(",\n");
}
fn renderFieldName(writer: anytype, operands: []const g.Operand, field_index: usize) !void {
const operand = operands[field_index];
// Should be enough for all names - adjust as needed.
var name_buffer = std.BoundedArray(u8, 64){
.buffer = undefined,
};
derive_from_kind: {
// Operand names are often in the json encoded as "'Name'" (with two sets of quotes).
// Additionally, some operands have ~ in them at the end (D~ref~).
const name = std.mem.trim(u8, operand.name, "'~");
if (name.len == 0) {
break :derive_from_kind;
}
// Some names have weird characters in them (like newlines) - skip any such ones.
// Use the same loop to transform to snake-case.
for (name) |c| {
switch (c) {
'a'...'z', '0'...'9' => try name_buffer.append(c),
'A'...'Z' => try name_buffer.append(std.ascii.toLower(c)),
' ', '~' => try name_buffer.append('_'),
else => break :derive_from_kind,
}
}
// Assume there are no duplicate 'name' fields.
try writer.print("{}", .{std.zig.fmtId(name_buffer.slice())});
return;
}
// Translate to snake case.
name_buffer.len = 0;
for (operand.kind) |c, i| {
switch (c) {
'a'...'z', '0'...'9' => try name_buffer.append(c),
'A'...'Z' => if (i > 0 and std.ascii.isLower(operand.kind[i - 1])) {
try name_buffer.appendSlice(&[_]u8{ '_', std.ascii.toLower(c) });
} else {
try name_buffer.append(std.ascii.toLower(c));
},
else => unreachable, // Assume that the name is valid C-syntax (and contains no underscores).
}
}
try writer.print("{}", .{std.zig.fmtId(name_buffer.slice())});
// For fields derived from type name, there could be any amount.
// Simply check against all other fields, and if another similar one exists, add a number.
const need_extra_index = for (operands) |other_operand, i| {
if (i != field_index and std.mem.eql(u8, operand.kind, other_operand.kind)) {
break true;
}
} else false;
if (need_extra_index) {
try writer.print("_{}", .{field_index});
}
}
fn parseHexInt(text: []const u8) !u31 {
const prefix = "0x";
if (!std.mem.startsWith(u8, text, prefix))
return error.InvalidHexInt;
return try std.fmt.parseInt(u31, text[prefix.len..], 16);
}
fn usageAndExit(file: std.fs.File, arg0: []const u8, code: u8) noreturn {
file.writer().print(
\\Usage: {s} <spirv json spec>
\\
\\Generates Zig bindings for a SPIR-V specification .json (either core or
\\extinst versions). The result, printed to stdout, should be used to update
\\files in src/codegen/spirv. Don't forget to format the output.
\\
\\The relevant specifications can be obtained from the SPIR-V registry:
\\https://github.com/KhronosGroup/SPIRV-Headers/blob/master/include/spirv/unified1/
\\
, .{arg0}) catch std.process.exit(1);
std.process.exit(code);
}