Feat/uint384 hint (#503)

uint_utils

src/hint_processor/builtin_hint_codes.zig

@@ -664,3 +664,189 @@ pub const NONDET_BIGINT3_V2 =
     \\from starkware.cairo.common.cairo_secp.secp_utils import split
     \\segments.write_arg(ids.res.address_, split(value))
 ;
+
+
+// The following hints support the lib https://github.com/NethermindEth/research-basic-Cairo-operations-big-integers/blob/main/lib
+pub const UINT384_UNSIGNED_DIV_REM =
+    \\def split(num: int, num_bits_shift: int, length: int):
+    \\    a = []
+    \\    for _ in range(length):
+    \\        a.append( num & ((1 << num_bits_shift) - 1) )
+    \\        num = num >> num_bits_shift
+    \\    return tuple(a)
+    \\
+    \\def pack(z, num_bits_shift: int) -> int:
+    \\    limbs = (z.d0, z.d1, z.d2)
+    \\    return sum(limb << (num_bits_shift * i) for i, limb in enumerate(limbs))
+    \\
+    \\a = pack(ids.a, num_bits_shift = 128)
+    \\div = pack(ids.div, num_bits_shift = 128)
+    \\quotient, remainder = divmod(a, div)
+    \\
+    \\quotient_split = split(quotient, num_bits_shift=128, length=3)
+    \\assert len(quotient_split) == 3
+    \\
+    \\ids.quotient.d0 = quotient_split[0]
+    \\ids.quotient.d1 = quotient_split[1]
+    \\ids.quotient.d2 = quotient_split[2]
+    \\
+    \\remainder_split = split(remainder, num_bits_shift=128, length=3)
+    \\ids.remainder.d0 = remainder_split[0]
+    \\ids.remainder.d1 = remainder_split[1]
+    \\ids.remainder.d2 = remainder_split[2]
+;
+
+pub const UINT384_SPLIT_128 =
+    \\ids.low = ids.a & ((1<<128) - 1)
+    \\ids.high = ids.a >> 128
+;
+
+pub const ADD_NO_UINT384_CHECK =
+    \\sum_d0 = ids.a.d0 + ids.b.d0
+    \\ids.carry_d0 = 1 if sum_d0 >= ids.SHIFT else 0
+    \\sum_d1 = ids.a.d1 + ids.b.d1 + ids.carry_d0
+    \\ids.carry_d1 = 1 if sum_d1 >= ids.SHIFT else 0
+    \\sum_d2 = ids.a.d2 + ids.b.d2 + ids.carry_d1
+    \\ids.carry_d2 = 1 if sum_d2 >= ids.SHIFT else 0
+;
+
+pub const UINT384_SQRT =
+    \\from starkware.python.math_utils import isqrt
+    \\
+    \\def split(num: int, num_bits_shift: int, length: int):
+    \\    a = []
+    \\    for _ in range(length):
+    \\        a.append( num & ((1 << num_bits_shift) - 1) )
+    \\        num = num >> num_bits_shift
+    \\    return tuple(a)
+    \\
+    \\def pack(z, num_bits_shift: int) -> int:
+    \\    limbs = (z.d0, z.d1, z.d2)
+    \\    return sum(limb << (num_bits_shift * i) for i, limb in enumerate(limbs))
+    \\
+    \\a = pack(ids.a, num_bits_shift=128)
+    \\root = isqrt(a)
+    \\assert 0 <= root < 2 ** 192
+    \\root_split = split(root, num_bits_shift=128, length=3)
+    \\ids.root.d0 = root_split[0]
+    \\ids.root.d1 = root_split[1]
+    \\ids.root.d2 = root_split[2]
+;
+
+pub const SUB_REDUCED_A_AND_REDUCED_B =
+    \\def split(num: int, num_bits_shift: int, length: int):
+    \\    a = []
+    \\    for _ in range(length):
+    \\        a.append( num & ((1 << num_bits_shift) - 1) )
+    \\        num = num >> num_bits_shift
+    \\    return tuple(a)
+    \\
+    \\def pack(z, num_bits_shift: int) -> int:
+    \\    limbs = (z.d0, z.d1, z.d2)
+    \\    return sum(limb << (num_bits_shift * i) for i, limb in enumerate(limbs))
+    \\
+    \\a = pack(ids.a, num_bits_shift = 128)
+    \\b = pack(ids.b, num_bits_shift = 128)
+    \\p = pack(ids.p, num_bits_shift = 128)
+    \\
+    \\res = (a - b) % p
+    \\
+    \\
+    \\res_split = split(res, num_bits_shift=128, length=3)
+    \\
+    \\ids.res.d0 = res_split[0]
+    \\ids.res.d1 = res_split[1]
+    \\ids.res.d2 = res_split[2]
+;
+
+pub const UINT384_SIGNED_NN = "memory[ap] = 1 if 0 <= (ids.a.d2 % PRIME) < 2 ** 127 else 0";
+
+pub const UNSIGNED_DIV_REM_UINT768_BY_UINT384 =
+    \\def split(num: int, num_bits_shift: int, length: int):
+    \\    a = []
+    \\    for _ in range(length):
+    \\        a.append( num & ((1 << num_bits_shift) - 1) )
+    \\        num = num >> num_bits_shift 
+    \\    return tuple(a)
+    \\
+    \\def pack(z, num_bits_shift: int) -> int:
+    \\    limbs = (z.d0, z.d1, z.d2)
+    \\    return sum(limb << (num_bits_shift * i) for i, limb in enumerate(limbs))
+    \\    
+    \\def pack_extended(z, num_bits_shift: int) -> int:
+    \\    limbs = (z.d0, z.d1, z.d2, z.d3, z.d4, z.d5)
+    \\    return sum(limb << (num_bits_shift * i) for i, limb in enumerate(limbs))
+    \\
+    \\a = pack_extended(ids.a, num_bits_shift = 128)
+    \\div = pack(ids.div, num_bits_shift = 128)
+    \\
+    \\quotient, remainder = divmod(a, div)
+    \\
+    \\quotient_split = split(quotient, num_bits_shift=128, length=6)
+    \\
+    \\ids.quotient.d0 = quotient_split[0]
+    \\ids.quotient.d1 = quotient_split[1]
+    \\ids.quotient.d2 = quotient_split[2]
+    \\ids.quotient.d3 = quotient_split[3]
+    \\ids.quotient.d4 = quotient_split[4]
+    \\ids.quotient.d5 = quotient_split[5]
+    \\
+    \\remainder_split = split(remainder, num_bits_shift=128, length=3)
+    \\ids.remainder.d0 = remainder_split[0]
+    \\ids.remainder.d1 = remainder_split[1]
+    \\ids.remainder.d2 = remainder_split[2]
+;
+
+// equal to UNSIGNED_DIV_REM_UINT768_BY_UINT384 but with some whitespace removed
+// in the `num = num >> num_bits_shift` and between `pack` and `pack_extended`
+pub const UNSIGNED_DIV_REM_UINT768_BY_UINT384_STRIPPED =
+    \\def split(num: int, num_bits_shift: int, length: int):
+    \\    a = []
+    \\    for _ in range(length):
+    \\        a.append( num & ((1 << num_bits_shift) - 1) )
+    \\        num = num >> num_bits_shift
+    \\    return tuple(a)
+    \\
+    \\def pack(z, num_bits_shift: int) -> int:
+    \\    limbs = (z.d0, z.d1, z.d2)
+    \\    return sum(limb << (num_bits_shift * i) for i, limb in enumerate(limbs))
+    \\
+    \\def pack_extended(z, num_bits_shift: int) -> int:
+    \\    limbs = (z.d0, z.d1, z.d2, z.d3, z.d4, z.d5)
+    \\    return sum(limb << (num_bits_shift * i) for i, limb in enumerate(limbs))
+    \\
+    \\a = pack_extended(ids.a, num_bits_shift = 128)
+    \\div = pack(ids.div, num_bits_shift = 128)
+    \\
+    \\quotient, remainder = divmod(a, div)
+    \\
+    \\quotient_split = split(quotient, num_bits_shift=128, length=6)
+    \\
+    \\ids.quotient.d0 = quotient_split[0]
+    \\ids.quotient.d1 = quotient_split[1]
+    \\ids.quotient.d2 = quotient_split[2]
+    \\ids.quotient.d3 = quotient_split[3]
+    \\ids.quotient.d4 = quotient_split[4]
+    \\ids.quotient.d5 = quotient_split[5]
+    \\
+    \\remainder_split = split(remainder, num_bits_shift=128, length=3)
+    \\ids.remainder.d0 = remainder_split[0]
+    \\ids.remainder.d1 = remainder_split[1]
+    \\ids.remainder.d2 = remainder_split[2]
+;
+
+pub const INV_MOD_P_UINT512 =
+    \\def pack_512(u, num_bits_shift: int) -> int:
+    \\    limbs = (u.d0, u.d1, u.d2, u.d3)
+    \\    return sum(limb << (num_bits_shift * i) for i, limb in enumerate(limbs))
+    \\
+    \\x = pack_512(ids.x, num_bits_shift = 128)
+    \\p = ids.p.low + (ids.p.high << 128)
+    \\x_inverse_mod_p = pow(x,-1, p)
+    \\
+    \\x_inverse_mod_p_split = (x_inverse_mod_p & ((1 << 128) - 1), x_inverse_mod_p >> 128)
+    \\
+    \\ids.x_inverse_mod_p.low = x_inverse_mod_p_split[0]
+    \\ids.x_inverse_mod_p.high = x_inverse_mod_p_split[1]
+;
+

src/hint_processor/builtin_hint_processor/secp/bigint_utils.zig

@@ -52,16 +52,22 @@ pub fn BigIntN(comptime NUM_LIMBS: usize) type {
             return .{ .limbs = limbs };
         }
 
-        pub fn insertFromVarName(self: *Self, allocator: std.mem.allocator, var_name: []const u8, vm: *CairoVM, ids_data: std.StringHashMap(HintReference), ap_tracking: ApTracking) !void {
+        pub fn insertFromVarName(
+            self: *const Self,
+            allocator: std.mem.Allocator,
+            var_name: []const u8,
+            vm: *CairoVM,
+            ids_data: std.StringHashMap(HintReference),
+            ap_tracking: ApTracking,
+        ) !void {
             const addr = try hint_utils.getRelocatableFromVarName(var_name, vm, ids_data, ap_tracking);
             inline for (0..NUM_LIMBS) |i| {
-                try vm.insertInMemory(allocator, addr + i, self.limbs[i]);
+                try vm.insertInMemory(allocator, try addr.addUint(i), MaybeRelocatable.fromFelt(self.limbs[i]));
             }
         }
 
         pub fn pack(self: *const Self, allocator: std.mem.Allocator) !Int {
-            const result = packBigInt(allocator, NUM_LIMBS, self.limbs, 128);
-            return result;
+            return packBigInt(allocator, NUM_LIMBS, self.limbs, 128);
         }
 
         pub fn pack86(self: *const Self, allocator: std.mem.Allocator) !Int {
@@ -80,9 +86,9 @@ pub fn BigIntN(comptime NUM_LIMBS: usize) type {
             return result;
         }
 
-        pub fn split(self: *Self, num: Int) Self {
-            const limbs = splitBigInt(std.mem.Allocator, num, self.limbs.len, 128);
-            return self.fromValues(limbs);
+
+        pub fn split(allocator: std.mem.Allocator, num: Int) !Self {
+            return Self.fromValues(try splitBigInt(allocator, num, NUM_LIMBS, 128));
         }
 
         // @TODO: implement from. It is dependent on split function.

src/hint_processor/hint_processor_def.zig

@@ -38,6 +38,9 @@ const segments = @import("segments.zig");
 
 const bigint_utils = @import("../hint_processor/builtin_hint_processor/secp/bigint_utils.zig");
 const bigint = @import("bigint.zig");
+const uint384 = @import("uint384.zig");
+const inv_mod_p_uint512 = @import("vrf/inv_mod_p_uint512.zig");
+
 
 const deserialize_utils = @import("../parser/deserialize_utils.zig");
 
@@ -384,6 +387,24 @@ pub const CairoVMHintProcessor = struct {
             try bigint.bigintPackDivModHint(allocator, vm, exec_scopes, hint_data.ids_data, hint_data.ap_tracking);
         } else if (std.mem.eql(u8, hint_codes.BIGINT_SAFE_DIV, hint_data.code)) {
             try bigint.bigIntSafeDivHint(allocator, vm, exec_scopes, hint_data.ids_data, hint_data.ap_tracking);
+        } else if (std.mem.eql(u8, hint_codes.UINT384_UNSIGNED_DIV_REM, hint_data.code)) {
+            try uint384.uint384UnsignedDivRem(allocator, vm, hint_data.ids_data, hint_data.ap_tracking);
+        } else if (std.mem.eql(u8, hint_codes.UINT384_SPLIT_128, hint_data.code)) {
+            try uint384.uint384Split128(allocator, vm, hint_data.ids_data, hint_data.ap_tracking);
+        } else if (std.mem.eql(u8, hint_codes.ADD_NO_UINT384_CHECK, hint_data.code)) {
+            try uint384.addNoUint384Check(allocator, vm, hint_data.ids_data, hint_data.ap_tracking, constants);
+        } else if (std.mem.eql(u8, hint_codes.UINT384_SQRT, hint_data.code)) {
+            try uint384.uint384Sqrt(allocator, vm, hint_data.ids_data, hint_data.ap_tracking);
+        } else if (std.mem.eql(u8, hint_codes.UINT384_SIGNED_NN, hint_data.code)) {
+            try uint384.uint384SignedNn(allocator, vm, hint_data.ids_data, hint_data.ap_tracking);
+        } else if (std.mem.eql(u8, hint_codes.SUB_REDUCED_A_AND_REDUCED_B, hint_data.code)) {
+            try uint384.subReducedAAndReducedB(allocator, vm, hint_data.ids_data, hint_data.ap_tracking);
+        } else if (std.mem.eql(u8, hint_codes.UNSIGNED_DIV_REM_UINT768_BY_UINT384, hint_data.code)) {
+            try uint384.unsignedDivRemUint768ByUint384(allocator, vm, hint_data.ids_data, hint_data.ap_tracking);
+        } else if (std.mem.eql(u8, hint_codes.UNSIGNED_DIV_REM_UINT768_BY_UINT384_STRIPPED, hint_data.code)) {
+            try uint384.unsignedDivRemUint768ByUint384(allocator, vm, hint_data.ids_data, hint_data.ap_tracking);
+        } else if (std.mem.eql(u8, hint_codes.INV_MOD_P_UINT512, hint_data.code)) {
+            try inv_mod_p_uint512.invModPUint512(allocator, vm, hint_data.ids_data, hint_data.ap_tracking);
         } else {
             std.log.err("not implemented: {s}\n", .{hint_data.code});
             return HintError.HintNotImplemented;

src/hint_processor/testing_utils.zig

@@ -48,6 +48,10 @@ pub fn checkMemory(mem: *Memory, comptime rows: anytype) !void {
 pub fn checkMemoryAddress(mem: *Memory, data: anytype) !void {
     const expected = if (data[1].len == 2) MaybeRelocatable.fromRelocatable(Relocatable.init(data[1][0], data[1][1])) else MaybeRelocatable.fromInt(u256, data[1][0]);
 
+    errdefer {
+        std.log.err("failed expect: {any}, got: {any}\n", .{ expected, mem.get(Relocatable.init(data[0][0], data[0][1])) });
+    }
+
     try std.testing.expectEqual(expected, mem.get(Relocatable.init(data[0][0], data[0][1])));
 }
 

src/hint_processor/uint256_utils.zig

@@ -12,6 +12,7 @@ const ApTracking = @import("../vm/types/programjson.zig").ApTracking;
 const HintData = @import("hint_processor_def.zig").HintData;
 const ExecutionScopes = @import("../vm/types/execution_scopes.zig").ExecutionScopes;
 
+const Int = @import("std").math.big.int.Managed;
 const helper = @import("../math/fields/helper.zig");
 const MathError = @import("../vm/error.zig").MathError;
 const HintError = @import("../vm/error.zig").HintError;
@@ -55,6 +56,15 @@ pub const Uint256 = struct {
     pub fn split(comptime T: type, num: T) Self {
         return Self.init(Felt252.fromInt(T, num & std.math.maxInt(u128)), Felt252.fromInt(T, num >> 128));
     }
+
+    pub fn pack(self: Self, allocator: std.mem.Allocator) !Int {
+        var result = try Int.initSet(allocator, self.high.toInteger());
+        errdefer result.deinit();
+
+        try result.shiftLeft(&result, 128);
+        try result.addScalar(&result, self.low.toInteger());
+        return result;
+    }
     // converting self to biguint value
     // optimize by using biguint
     // right now using u512, so to not use allocator with big int