diff --git a/lib/std/heap/SmpAllocator.zig b/lib/std/heap/SmpAllocator.zig
index 50593687ec..63f1e38a2a 100644
--- a/lib/std/heap/SmpAllocator.zig
+++ b/lib/std/heap/SmpAllocator.zig
@@ -51,10 +51,6 @@ const slab_len: usize = @max(std.heap.page_size_max, 64 * 1024);
 /// Because of storing free list pointers, the minimum size class is 3.
 const min_class = math.log2(@sizeOf(usize));
 const size_class_count = math.log2(slab_len) - min_class;
-/// When a freelist length exceeds this number, a `free` will rotate up to
-/// `max_free_search` times before pushing.
-const max_freelist_len: u8 = 16;
-const max_free_search = 1;
 /// Before mapping a fresh page, `alloc` will rotate this many times.
 const max_alloc_search = 1;
 
@@ -73,8 +69,6 @@ const Thread = struct {
     next_addrs: [size_class_count]usize = @splat(0),
     /// For each size class, points to the freed pointer.
     frees: [size_class_count]usize = @splat(0),
-    /// For each size class, tracks the number of items in the freelist.
-    freelist_lens: [size_class_count]u8 = @splat(0),
 
     fn lock() *Thread {
         var index = thread_index;
@@ -142,7 +136,6 @@ fn alloc(context: *anyopaque, len: usize, alignment: mem.Alignment, ra: usize) ?
             defer t.unlock();
             const node: *usize = @ptrFromInt(top_free_ptr);
             t.frees[class] = node.*;
-            t.freelist_lens[class] -|= 1;
             return @ptrFromInt(top_free_ptr);
         }
 
@@ -160,7 +153,6 @@ fn alloc(context: *anyopaque, len: usize, alignment: mem.Alignment, ra: usize) ?
             // slab alignment here ensures the % slab len earlier catches the end of slots.
             const slab = PageAllocator.map(slab_len, .fromByteUnits(slab_len)) orelse return null;
             t.next_addrs[class] = @intFromPtr(slab) + slot_size;
-            t.freelist_lens[class] = 0;
             return slab;
         }
 
@@ -214,42 +206,12 @@ fn free(context: *anyopaque, memory: []u8, alignment: mem.Alignment, ra: usize)
     }
 
     const node: *usize = @alignCast(@ptrCast(memory.ptr));
-    var search_count: u8 = 0;
 
-    var t = Thread.lock();
+    const t = Thread.lock();
+    defer t.unlock();
 
-    outer: while (true) {
-        const freelist_len = t.freelist_lens[class];
-        if (freelist_len < max_freelist_len) {
-            @branchHint(.likely);
-            defer t.unlock();
-            node.* = t.frees[class];
-            t.frees[class] = @intFromPtr(node);
-            return;
-        }
-
-        if (search_count >= max_free_search) {
-            defer t.unlock();
-            t.freelist_lens[class] = freelist_len +| 1;
-            node.* = t.frees[class];
-            t.frees[class] = @intFromPtr(node);
-            return;
-        }
-
-        t.unlock();
-        const cpu_count = getCpuCount();
-        assert(cpu_count != 0);
-        var index = thread_index;
-        while (true) {
-            index = (index + 1) % cpu_count;
-            t = &global.threads[index];
-            if (t.mutex.tryLock()) {
-                thread_index = index;
-                search_count += 1;
-                continue :outer;
-            }
-        }
-    }
+    node.* = t.frees[class];
+    t.frees[class] = @intFromPtr(node);
 }
 
 fn sizeClassIndex(len: usize, alignment: mem.Alignment) usize {