stage2-wasm: packed store/load 128 bits

This commit is contained in:
Pavel Verigo 2025-03-24 14:54:31 +01:00
parent 911f4527f0
commit 33ad2c949e
4 changed files with 86 additions and 60 deletions

View file

@ -759,6 +759,16 @@ fn resolveInst(cg: *CodeGen, ref: Air.Inst.Ref) InnerError!WValue {
return result;
}
fn resolveValue(cg: *CodeGen, val: Value) InnerError!WValue {
const zcu = cg.pt.zcu;
const ty = val.typeOf(zcu);
return if (isByRef(ty, zcu, cg.target))
.{ .uav_ref = .{ .ip_index = val.toIntern() } }
else
try cg.lowerConstant(val, ty);
}
/// NOTE: if result == .stack, it will be stored in .local
fn finishAir(cg: *CodeGen, inst: Air.Inst.Index, result: WValue, operands: []const Air.Inst.Ref) InnerError!void {
assert(operands.len <= Liveness.bpi - 1);
@ -2319,39 +2329,56 @@ fn airStore(cg: *CodeGen, inst: Air.Inst.Index, safety: bool) InnerError!void {
} else {
// at this point we have a non-natural alignment, we must
// load the value, and then shift+or the rhs into the result location.
const int_elem_ty = try pt.intType(.unsigned, ptr_info.packed_offset.host_size * 8);
const host_size = ptr_info.packed_offset.host_size * 8;
const host_ty = try pt.intType(.unsigned, host_size);
const bit_size: u16 = @intCast(ty.bitSize(zcu));
const bit_offset = ptr_info.packed_offset.bit_offset;
if (isByRef(int_elem_ty, zcu, cg.target)) {
return cg.fail("TODO: airStore for pointers to bitfields with backing type larger than 64bits", .{});
const mask_val = try cg.resolveValue(val: {
const limbs = try cg.gpa.alloc(
std.math.big.Limb,
std.math.big.int.calcTwosCompLimbCount(host_size) + 1,
);
defer cg.gpa.free(limbs);
var mask_bigint: std.math.big.int.Mutable = .{ .limbs = limbs, .positive = undefined, .len = undefined };
mask_bigint.setTwosCompIntLimit(.max, .unsigned, host_size);
if (bit_size != host_size) {
mask_bigint.shiftRight(mask_bigint.toConst(), host_size - bit_size);
}
if (bit_offset != 0) {
mask_bigint.shiftLeft(mask_bigint.toConst(), bit_offset);
}
mask_bigint.bitNotWrap(mask_bigint.toConst(), .unsigned, host_size);
var mask = @as(u64, @intCast((@as(u65, 1) << @as(u7, @intCast(ty.bitSize(zcu)))) - 1));
mask <<= @as(u6, @intCast(ptr_info.packed_offset.bit_offset));
mask ^= ~@as(u64, 0);
const shift_val: WValue = if (ptr_info.packed_offset.host_size <= 4)
.{ .imm32 = ptr_info.packed_offset.bit_offset }
else
.{ .imm64 = ptr_info.packed_offset.bit_offset };
const mask_val: WValue = if (ptr_info.packed_offset.host_size <= 4)
.{ .imm32 = @as(u32, @truncate(mask)) }
else
.{ .imm64 = mask };
const wrap_mask_val: WValue = if (ptr_info.packed_offset.host_size <= 4)
.{ .imm32 = @truncate(~@as(u64, 0) >> @intCast(64 - ty.bitSize(zcu))) }
else
.{ .imm64 = ~@as(u64, 0) >> @intCast(64 - ty.bitSize(zcu)) };
break :val try pt.intValue_big(host_ty, mask_bigint.toConst());
});
const shift_val: WValue = if (33 <= host_size and host_size <= 64)
.{ .imm64 = bit_offset }
else
.{ .imm32 = bit_offset };
if (host_size <= 64) {
try cg.emitWValue(lhs);
const loaded = try cg.load(lhs, int_elem_ty, 0);
const anded = try cg.binOp(loaded, mask_val, int_elem_ty, .@"and");
const extended_value = try cg.intcast(rhs, ty, int_elem_ty);
const masked_value = try cg.binOp(extended_value, wrap_mask_val, int_elem_ty, .@"and");
const shifted_value = if (ptr_info.packed_offset.bit_offset > 0) shifted: {
break :shifted try cg.binOp(masked_value, shift_val, int_elem_ty, .shl);
} else masked_value;
const result = try cg.binOp(anded, shifted_value, int_elem_ty, .@"or");
// lhs is still on the stack
try cg.store(.stack, result, int_elem_ty, lhs.offset());
}
const loaded = if (host_size <= 64)
try cg.load(lhs, host_ty, 0)
else
lhs;
const anded = try cg.binOp(loaded, mask_val, host_ty, .@"and");
const extended_value = try cg.intcast(rhs, ty, host_ty);
const shifted_value = if (bit_offset > 0)
try cg.binOp(extended_value, shift_val, host_ty, .shl)
else
extended_value;
const result = try cg.binOp(anded, shifted_value, host_ty, .@"or");
if (host_size <= 64) {
try cg.store(.stack, result, host_ty, lhs.offset());
} else {
try cg.store(lhs, result, host_ty, lhs.offset());
}
}
return cg.finishAir(inst, .none, &.{ bin_op.lhs, bin_op.rhs });
@ -2494,22 +2521,30 @@ fn airLoad(cg: *CodeGen, inst: Air.Inst.Index) InnerError!void {
}
if (ptr_info.packed_offset.host_size == 0) {
break :result try cg.load(operand, ty, 0);
const loaded = try cg.load(operand, ty, 0);
const ty_size = ty.abiSize(zcu);
if (ty.isAbiInt(zcu) and ty_size * 8 > ty.bitSize(zcu)) {
const int_elem_ty = try pt.intType(.unsigned, @intCast(ty_size * 8));
break :result try cg.trunc(loaded, ty, int_elem_ty);
} else {
break :result loaded;
}
// at this point we have a non-natural alignment, we must
// shift the value to obtain the correct bit.
} else {
const int_elem_ty = try pt.intType(.unsigned, ptr_info.packed_offset.host_size * 8);
const shift_val: WValue = if (ptr_info.packed_offset.host_size <= 4)
.{ .imm32 = ptr_info.packed_offset.bit_offset }
else if (ptr_info.packed_offset.host_size <= 8)
.{ .imm64 = ptr_info.packed_offset.bit_offset }
else
return cg.fail("TODO: airLoad where ptr to bitfield exceeds 64 bits", .{});
.{ .imm32 = ptr_info.packed_offset.bit_offset };
const stack_loaded = try cg.load(operand, int_elem_ty, 0);
const stack_loaded = if (ptr_info.packed_offset.host_size <= 8)
try cg.load(operand, int_elem_ty, 0)
else
operand;
const shifted = try cg.binOp(stack_loaded, shift_val, int_elem_ty, .shr);
break :result try cg.trunc(shifted, ty, int_elem_ty);
}
};
return cg.finishAir(inst, result, &.{ty_op.operand});
}
@ -3857,15 +3892,12 @@ fn airStructFieldVal(cg: *CodeGen, inst: Air.Inst.Index) InnerError!void {
const packed_struct = zcu.typeToPackedStruct(struct_ty).?;
const offset = pt.structPackedFieldBitOffset(packed_struct, field_index);
const backing_ty = Type.fromInterned(packed_struct.backingIntTypeUnordered(ip));
const wasm_bits = toWasmBits(backing_ty.intInfo(zcu).bits) orelse {
return cg.fail("TODO: airStructFieldVal for packed structs larger than 128 bits", .{});
};
const const_wvalue: WValue = if (wasm_bits == 32)
.{ .imm32 = offset }
else if (wasm_bits == 64)
const host_bits = backing_ty.intInfo(zcu).bits;
const const_wvalue: WValue = if (33 <= host_bits and host_bits <= 64)
.{ .imm64 = offset }
else
return cg.fail("TODO: airStructFieldVal for packed structs larger than 64 bits", .{});
.{ .imm32 = offset };
// for first field we don't require any shifting
const shifted_value = if (offset == 0)
@ -4043,7 +4075,7 @@ fn airSwitchBr(cg: *CodeGen, inst: Air.Inst.Index, is_dispatch_loop: bool) Inner
if (use_br_table) {
const width = width_maybe.?;
const br_value_original = try cg.binOp(target, try cg.resolveInst(Air.internedToRef(min.?.toIntern())), target_ty, .sub);
const br_value_original = try cg.binOp(target, try cg.resolveValue(min.?), target_ty, .sub);
_ = try cg.intcast(br_value_original, target_ty, Type.u32);
const jump_table: Mir.JumpTable = .{ .length = width + 1 };
@ -5232,7 +5264,7 @@ fn airAggregateInit(cg: *CodeGen, inst: Air.Inst.Index) InnerError!void {
}
}
if (sentinel) |s| {
const val = try cg.resolveInst(Air.internedToRef(s.toIntern()));
const val = try cg.resolveValue(s);
try cg.store(offset, val, elem_ty, 0);
}
} else {
@ -5243,7 +5275,7 @@ fn airAggregateInit(cg: *CodeGen, inst: Air.Inst.Index) InnerError!void {
offset += elem_size;
}
if (sentinel) |s| {
const val = try cg.resolveInst(Air.internedToRef(s.toIntern()));
const val = try cg.resolveValue(s);
try cg.store(result, val, elem_ty, offset);
}
}

View file

@ -480,7 +480,6 @@ test "@bitCast of packed struct of bools all true" {
if (builtin.zig_backend == .stage2_aarch64) return error.SkipZigTest; // TODO
if (builtin.zig_backend == .stage2_arm) return error.SkipZigTest; // TODO
if (builtin.zig_backend == .stage2_c) return error.SkipZigTest; // TODO
if (builtin.zig_backend == .stage2_wasm) return error.SkipZigTest; // TODO
if (builtin.zig_backend == .stage2_spirv64) return error.SkipZigTest; // TODO
const P = packed struct {
@ -501,7 +500,6 @@ test "@bitCast of packed struct of bools all false" {
if (builtin.zig_backend == .stage2_aarch64) return error.SkipZigTest; // TODO
if (builtin.zig_backend == .stage2_arm) return error.SkipZigTest; // TODO
if (builtin.zig_backend == .stage2_c) return error.SkipZigTest; // TODO
if (builtin.zig_backend == .stage2_wasm) return error.SkipZigTest; // TODO
if (builtin.zig_backend == .stage2_spirv64) return error.SkipZigTest; // TODO
const P = packed struct {

View file

@ -1321,7 +1321,6 @@ test "packed struct with signed field" {
test "assign packed struct initialized with RLS to packed struct literal field" {
if (builtin.zig_backend == .stage2_llvm and builtin.cpu.arch.isWasm()) return error.SkipZigTest;
if (builtin.zig_backend == .stage2_riscv64) return error.SkipZigTest;
if (builtin.zig_backend == .stage2_wasm) return error.SkipZigTest;
if (builtin.zig_backend == .stage2_spirv64) return error.SkipZigTest;
const Inner = packed struct { x: u17 };

View file

@ -421,9 +421,7 @@ const Foo96Bits = packed struct {
test "packed struct 24bits" {
if (builtin.zig_backend == .stage2_aarch64) return error.SkipZigTest;
if (builtin.zig_backend == .stage2_spirv64) return error.SkipZigTest;
if (builtin.zig_backend == .stage2_wasm) return error.SkipZigTest; // TODO
if (builtin.zig_backend == .stage2_arm) return error.SkipZigTest; // TODO
if (builtin.cpu.arch.isWasm()) return error.SkipZigTest; // TODO
if (builtin.cpu.arch.isArm()) return error.SkipZigTest; // TODO
if (builtin.zig_backend == .stage2_sparc64) return error.SkipZigTest; // TODO
@ -763,7 +761,6 @@ const S0 = struct {
var g_foo: S0 = S0.init();
test "packed struct with fp fields" {
if (builtin.zig_backend == .stage2_wasm) return error.SkipZigTest; // TODO
if (builtin.zig_backend == .stage2_arm) return error.SkipZigTest; // TODO
if (builtin.zig_backend == .stage2_aarch64) return error.SkipZigTest; // TODO
if (builtin.zig_backend == .stage2_sparc64) return error.SkipZigTest; // TODO