Add kmem_cache explanation to memory allocation section

Introduces the kmem_cache for frequent, fixed-size allocations.

Author: visitorckw

examples/Makefile

@@ -40,6 +40,7 @@ obj-m += devicetree.o
 obj-m += dma.o
 obj-m += blkram.o
 obj-m += vnetloop.o
+obj-m += kmem_cache.o
 
 KDIR ?= /lib/modules/$(shell uname -r)/build
 PWD := $(CURDIR)

examples/kmem_cache.c

@@ -0,0 +1,70 @@
+/*
+ * kmem_cache_example.c - Demonstrates the use of the slab allocator.
+ */
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/slab.h>
+
+struct my_custom_struct {
+    int id;
+    char name[32];
+};
+
+static struct kmem_cache *my_cachep;
+static struct my_custom_struct *obj1, *obj2;
+
+static int __init kmem_cache_example_init(void)
+{
+    pr_info("kmem_cache_example: Initializing module\n");
+
+    my_cachep =
+        kmem_cache_create("my_custom_cache", sizeof(struct my_custom_struct), 0,
+                          SLAB_HWCACHE_ALIGN, NULL);
+
+    if (!my_cachep) {
+        pr_err("kmem_cache_example: Failed to create cache\n");
+        return -ENOMEM;
+    }
+
+    obj1 = kmem_cache_alloc(my_cachep, GFP_KERNEL);
+    obj2 = kmem_cache_alloc(my_cachep, GFP_KERNEL);
+
+    if (!obj1 || !obj2) {
+        pr_err("kmem_cache_example: Failed to allocate objects\n");
+        if (obj1)
+            kmem_cache_free(my_cachep, obj1);
+        if (obj2)
+            kmem_cache_free(my_cachep, obj2);
+        kmem_cache_destroy(my_cachep);
+        return -ENOMEM;
+    }
+
+    obj1->id = 1;
+    obj2->id = 2;
+    pr_info("kmem_cache_example: Allocated obj1 at %p (id: %d)\n", obj1,
+            obj1->id);
+    pr_info("kmem_cache_example: Allocated obj2 at %p (id: %d)\n", obj2,
+            obj2->id);
+
+    return 0;
+}
+
+static void __exit kmem_cache_example_exit(void)
+{
+    if (obj1)
+        kmem_cache_free(my_cachep, obj1);
+    if (obj2)
+        kmem_cache_free(my_cachep, obj2);
+
+    if (my_cachep)
+        kmem_cache_destroy(my_cachep);
+
+    pr_info("kmem_cache_example: Cleaned up and exited\n");
+}
+
+module_init(kmem_cache_example_init);
+module_exit(kmem_cache_example_exit);
+
+MODULE_LICENSE("GPL");
+MODULE_AUTHOR("Kuan-Wei Chiu <visitorckw@gmail.com>");
+MODULE_DESCRIPTION("A simple example of kmem_cache");

lkmpg.tex

@@ -2480,6 +2480,23 @@ \section{Memory Allocation and Ownership}
 Similarly, driver authors should resist the temptation to guess at allocator
 internals or to depend on low-level layout details that are not part of the API.
 
+When a module needs to frequently allocate and free objects of the same size---such 
+as custom data structures, network packet descriptors, or state containers---using 
+\cpp|kmalloc()| repeatedly can lead to memory fragmentation and unnecessary overhead.
+For these highly repetitive allocations, the slab allocator interface is the 
+correct tool. 
+Driver authors should use \cpp|kmem_cache_create()| during module initialization 
+to set up a dedicated cache for their specific structure. 
+Once the cache is created, \cpp|kmem_cache_alloc()| and \cpp|kmem_cache_free()| 
+provide extremely fast, fragmentation-resistant memory management for those objects.
+It is crucial to symmetrically destroy the cache with \cpp|kmem_cache_destroy()| 
+during module cleanup to prevent memory leaks.
+
+For a practical demonstration of creating a custom slab cache, allocating objects from
+it, and ensuring proper cleanup, see the following example.
+
+\samplec{examples/kmem_cache.c}
+
 Ownership is just as important as allocation.
 In Linux 5.10 and later, many drivers benefit from managed-resource helpers such as
 \cpp|devm_kzalloc()| and friends when they are written as proper device-model