diff --git a/lib/compiler_rt/divmodei4.zig b/lib/compiler_rt/divmodei4.zig
index f30eded317..55c7ec5792 100644
--- a/lib/compiler_rt/divmodei4.zig
+++ b/lib/compiler_rt/divmodei4.zig
@@ -10,45 +10,39 @@ const symbol = @import("../compiler_rt.zig").symbol;
 comptime {
     symbol(&__divei4, "__divei4");
     symbol(&__modei4, "__modei4");
+    symbol(&__divei5, "__divei5");
+    symbol(&__modei5, "__modei5");
 }
 
-inline fn limb(x: []u32, i: usize) *u32 {
-    return if (endian == .little) &x[i] else &x[x.len - 1 - i];
+inline fn limb(i: usize, len: usize) usize {
+    return if (endian == .little) i else len - 1 - i;
 }
 
-inline fn neg(x: []u32) void {
+inline fn neg(out: []u32, in: []const u32) void {
     var ov: u1 = 1;
-    for (0..x.len) |limb_index| {
-        const l = limb(x, limb_index);
-        l.*, ov = @addWithOverflow(~l.*, ov);
+    for (0..in.len) |limb_index| {
+        const new, ov = @addWithOverflow(~in[limb(limb_index, in.len)], ov);
+        out[limb(limb_index, out.len)] = new;
     }
 }
 
-const max_limbs = std.math.divCeil(usize, 65535, 32) catch unreachable;
-
-fn divmod(q: ?[]u32, r: ?[]u32, u: []const u32, v: []const u32) !void {
-    const u_sign: i32 = @bitCast(u[u.len - 1]);
-    const v_sign: i32 = @bitCast(v[v.len - 1]);
-    var ua: [max_limbs]u32 = undefined;
-    const us = ua[0..u.len];
-    @memcpy(us, u);
-    var va: [max_limbs]u32 = undefined;
-    const vs = va[0..v.len];
-    @memcpy(vs, v);
-    if (u_sign < 0) neg(us);
-    if (v_sign < 0) neg(vs);
-    try @call(.always_inline, udivmod, .{ q, r, us, vs });
-    if (q) |x| if (u_sign ^ v_sign < 0) neg(x);
-    if (r) |x| if (u_sign < 0) neg(x);
+fn divmod(q: ?[]u32, r: ?[]u32, u: []const u32, v: []const u32, tu: []u32, tv: []u32) !void {
+    const u_sign: i32 = @bitCast(u[limb(u.len - 1, u.len)]);
+    const v_sign: i32 = @bitCast(v[limb(v.len - 1, v.len)]);
+    if (u_sign < 0) neg(tu, u);
+    if (v_sign < 0) neg(tv, v);
+    try @call(.always_inline, udivmod, .{ q, r, if (u_sign < 0) tu else u, if (v_sign < 0) tv else v });
+    if (q) |x| if (u_sign ^ v_sign < 0) neg(x, x);
+    if (r) |x| if (u_sign < 0) neg(x, x);
 }
 
-pub fn __divei4(q_p: [*]u8, u_p: [*]const u8, v_p: [*]const u8, bits: usize) callconv(.c) void {
+pub fn __divei4(q_p: [*]u8, u_p: [*]u8, v_p: [*]u8, bits: usize) callconv(.c) void {
     @setRuntimeSafety(compiler_rt.test_safety);
     const byte_size = std.zig.target.intByteSize(&builtin.target, @intCast(bits));
     const q: []u32 = @ptrCast(@alignCast(q_p[0..byte_size]));
-    const u: []const u32 = @ptrCast(@alignCast(u_p[0..byte_size]));
-    const v: []const u32 = @ptrCast(@alignCast(v_p[0..byte_size]));
-    @call(.always_inline, divmod, .{ q, null, u, v }) catch unreachable;
+    const u: []u32 = @ptrCast(@alignCast(u_p[0..byte_size]));
+    const v: []u32 = @ptrCast(@alignCast(v_p[0..byte_size]));
+    @call(.always_inline, divmod, .{ q, null, u, v, u, v }) catch unreachable;
 }
 
 pub fn __modei4(r_p: [*]u8, u_p: [*]u8, v_p: [*]u8, bits: usize) callconv(.c) void {
@@ -57,5 +51,27 @@ pub fn __modei4(r_p: [*]u8, u_p: [*]u8, v_p: [*]u8, bits: usize) callconv(.c) vo
     const r: []u32 = @ptrCast(@alignCast(r_p[0..byte_size]));
     const u: []u32 = @ptrCast(@alignCast(u_p[0..byte_size]));
     const v: []u32 = @ptrCast(@alignCast(v_p[0..byte_size]));
-    @call(.always_inline, divmod, .{ null, r, u, v }) catch unreachable;
+    @call(.always_inline, divmod, .{ null, r, u, v, u, v }) catch unreachable;
+}
+
+pub fn __divei5(q_p: [*]u8, u_p: [*]const u8, v_p: [*]const u8, t_p: [*]u8, bits: usize) callconv(.c) void {
+    @setRuntimeSafety(compiler_rt.test_safety);
+    const byte_size = std.zig.target.intByteSize(&builtin.target, @intCast(bits));
+    const q: []u32 = @ptrCast(@alignCast(q_p[0..byte_size]));
+    const u: []const u32 = @ptrCast(@alignCast(u_p[0..byte_size]));
+    const v: []const u32 = @ptrCast(@alignCast(v_p[0..byte_size]));
+    const tu: []u32 = @ptrCast(@alignCast(t_p[0..byte_size]));
+    const tv: []u32 = @ptrCast(@alignCast(t_p[byte_size..][0..byte_size]));
+    @call(.always_inline, divmod, .{ q, null, u, v, tu, tv }) catch unreachable;
+}
+
+pub fn __modei5(r_p: [*]u8, u_p: [*]const u8, v_p: [*]const u8, t_p: [*]u8, bits: usize) callconv(.c) void {
+    @setRuntimeSafety(compiler_rt.test_safety);
+    const byte_size = std.zig.target.intByteSize(&builtin.target, @intCast(bits));
+    const r: []u32 = @ptrCast(@alignCast(r_p[0..byte_size]));
+    const u: []const u32 = @ptrCast(@alignCast(u_p[0..byte_size]));
+    const v: []const u32 = @ptrCast(@alignCast(v_p[0..byte_size]));
+    const tu: []u32 = @ptrCast(@alignCast(t_p[0..byte_size]));
+    const tv: []u32 = @ptrCast(@alignCast(t_p[byte_size..][0..byte_size]));
+    @call(.always_inline, divmod, .{ null, r, u, v, tu, tv }) catch unreachable;
 }
diff --git a/lib/compiler_rt/udivmodei4.zig b/lib/compiler_rt/udivmodei4.zig
index f037639c72..5aa8421bdb 100644
--- a/lib/compiler_rt/udivmodei4.zig
+++ b/lib/compiler_rt/udivmodei4.zig
@@ -13,6 +13,8 @@ const max_limbs = std.math.divCeil(usize, 65535, 32) catch unreachable; // max s
 comptime {
     symbol(&__udivei4, "__udivei4");
     symbol(&__umodei4, "__umodei4");
+    symbol(&__udivei5, "__udivei5");
+    symbol(&__umodei5, "__umodei5");
 }
 
 /// Get the value of a limb.
@@ -132,6 +134,32 @@ pub fn __umodei4(r_p: [*]u8, u_p: [*]const u8, v_p: [*]const u8, bits: usize) ca
     @call(.always_inline, divmod, .{ null, r, u, v }) catch unreachable;
 }
 
+pub fn __udivei5(q_p: [*]u8, u_p: [*]const u8, v_p: [*]const u8, t_p: [*]u8, bits: usize) callconv(.c) void {
+    @setRuntimeSafety(compiler_rt.test_safety);
+    const byte_size = std.zig.target.intByteSize(&builtin.target, @intCast(bits));
+    const q: []u32 = @ptrCast(@alignCast(q_p[0..byte_size]));
+    const u: []const u32 = @ptrCast(@alignCast(u_p[0..byte_size]));
+    const v: []const u32 = @ptrCast(@alignCast(v_p[0..byte_size]));
+    const tu: []u32 = @ptrCast(@alignCast(t_p[0..byte_size]));
+    _ = tu;
+    const tv: []u32 = @ptrCast(@alignCast(t_p[byte_size..][0..byte_size]));
+    _ = tv;
+    @call(.always_inline, divmod, .{ q, null, u, v }) catch unreachable;
+}
+
+pub fn __umodei5(r_p: [*]u8, u_p: [*]const u8, v_p: [*]const u8, t_p: [*]u8, bits: usize) callconv(.c) void {
+    @setRuntimeSafety(compiler_rt.test_safety);
+    const byte_size = std.zig.target.intByteSize(&builtin.target, @intCast(bits));
+    const r: []u32 = @ptrCast(@alignCast(r_p[0..byte_size]));
+    const u: []const u32 = @ptrCast(@alignCast(u_p[0..byte_size]));
+    const v: []const u32 = @ptrCast(@alignCast(v_p[0..byte_size]));
+    const tu: []u32 = @ptrCast(@alignCast(t_p[0..byte_size]));
+    _ = tu;
+    const tv: []u32 = @ptrCast(@alignCast(t_p[byte_size..][0..byte_size]));
+    _ = tv;
+    @call(.always_inline, divmod, .{ null, r, u, v }) catch unreachable;
+}
+
 test "__udivei4/__umodei4" {
     if (builtin.zig_backend == .stage2_aarch64) return error.SkipZigTest;
     if (builtin.zig_backend == .stage2_c) return error.SkipZigTest;
diff --git a/src/codegen/wasm/CodeGen.zig b/src/codegen/wasm/CodeGen.zig
index 82b4c21e0b..12e8826a92 100644
--- a/src/codegen/wasm/CodeGen.zig
+++ b/src/codegen/wasm/CodeGen.zig
@@ -2530,11 +2530,24 @@ fn intDiv(cg: *CodeGen, ty: IntType, lhs: WValue, rhs: WValue) InnerError!WValue
         },
         else => {
             const result = try cg.allocInt(ty);
+            const bits = cg.intBackingBits(ty.bits);
+            var tmp = try cg.allocInt(.{ .is_signed = false, .bits = bits * 2 });
             if (ty.is_signed) {
-                _ = try cg.callIntrinsic(.__divei4, &.{ .usize_type, .usize_type, .usize_type, .usize_type }, .void, &.{ result, lhs, rhs, .{ .imm32 = ty.bits } });
+                _ = try cg.callIntrinsic(
+                    .__divei5,
+                    &.{ .usize_type, .usize_type, .usize_type, .usize_type, .usize_type },
+                    .void,
+                    &.{ result, lhs, rhs, tmp, .{ .imm32 = ty.bits } },
+                );
             } else {
-                _ = try cg.callIntrinsic(.__udivei4, &.{ .usize_type, .usize_type, .usize_type, .usize_type }, .void, &.{ result, lhs, rhs, .{ .imm32 = ty.bits } });
+                _ = try cg.callIntrinsic(
+                    .__udivei5,
+                    &.{ .usize_type, .usize_type, .usize_type, .usize_type, .usize_type },
+                    .void,
+                    &.{ result, lhs, rhs, tmp, .{ .imm32 = ty.bits } },
+                );
             }
+            tmp.free(cg);
             return result;
         },
     }
@@ -2631,11 +2644,24 @@ fn intRem(cg: *CodeGen, ty: IntType, lhs: WValue, rhs: WValue) InnerError!WValue
         },
         else => {
             const result = try cg.allocInt(ty);
+            const bits = cg.intBackingBits(ty.bits);
+            var tmp = try cg.allocInt(.{ .is_signed = false, .bits = bits * 2 });
             if (ty.is_signed) {
-                _ = try cg.callIntrinsic(.__modei4, &.{ .usize_type, .usize_type, .usize_type, .usize_type }, .void, &.{ result, lhs, rhs, .{ .imm32 = ty.bits } });
+                _ = try cg.callIntrinsic(
+                    .__modei5,
+                    &.{ .usize_type, .usize_type, .usize_type, .usize_type, .usize_type },
+                    .void,
+                    &.{ result, lhs, rhs, tmp, .{ .imm32 = ty.bits } },
+                );
             } else {
-                _ = try cg.callIntrinsic(.__umodei4, &.{ .usize_type, .usize_type, .usize_type, .usize_type }, .void, &.{ result, lhs, rhs, .{ .imm32 = ty.bits } });
+                _ = try cg.callIntrinsic(
+                    .__umodei5,
+                    &.{ .usize_type, .usize_type, .usize_type, .usize_type, .usize_type },
+                    .void,
+                    &.{ result, lhs, rhs, tmp, .{ .imm32 = ty.bits } },
+                );
             }
+            tmp.free(cg);
             return result;
         },
     }
diff --git a/src/codegen/wasm/Mir.zig b/src/codegen/wasm/Mir.zig
index d2d48e9bee..11bf63bdfd 100644
--- a/src/codegen/wasm/Mir.zig
+++ b/src/codegen/wasm/Mir.zig
@@ -825,7 +825,7 @@ pub const Intrinsic = enum(u32) {
     __ceilx,
     __cosh,
     __cosx,
-    __divei4,
+    __divei5,
     __divhf3,
     __divtf3,
     __divti3,
@@ -951,7 +951,7 @@ pub const Intrinsic = enum(u32) {
     __lshrti3,
     __lttf2,
     __ltxf2,
-    __modei4,
+    __modei5,
     __modti3,
     __mulhf3,
     __mulodi4,
@@ -982,9 +982,9 @@ pub const Intrinsic = enum(u32) {
     __truncxfdf2,
     __truncxfhf2,
     __truncxfsf2,
-    __udivei4,
+    __udivei5,
     __udivti3,
-    __umodei4,
+    __umodei5,
     __umodti3,
     ceilq,
     cos,