Add /proc/kpagemap interface
This makes physical page flags and counts available to userspace.
Together with /proc/pid/pagemap and /proc/pid/clear_refs, this can be
used to measure memory usage on a per-page basis.
Signed-off-by: Matt Mackall <[EMAIL PROTECTED]>
Index: mm/fs/proc/proc_misc.c
===================================================================
--- mm.orig/fs/proc/proc_misc.c 2007-04-05 14:18:49.000000000 -0500
+++ mm/fs/proc/proc_misc.c 2007-04-05 14:26:23.000000000 -0500
@@ -46,6 +46,8 @@
#include <linux/vmalloc.h>
#include <linux/crash_dump.h>
#include <linux/pid_namespace.h>
+#include <linux/ptrace.h>
+#include <linux/bootmem.h>
#include <asm/uaccess.h>
#include <asm/pgtable.h>
#include <asm/io.h>
@@ -733,6 +735,91 @@ static struct file_operations proc_page_
};
#endif
+#ifdef CONFIG_PROC_KPAGEMAP
+#define KPMSIZE (sizeof(unsigned long) * 2)
+#define KPMMASK (KPMSIZE - 1)
+/* /proc/kpagemap - an array exposing page flags and counts
+ *
+ * Each entry is a pair of unsigned longs representing the
+ * corresponding physical page, the first containing the page flags
+ * and the second containing the page use count.
+ *
+ * The first 4 bytes of this file form a simple header:
+ *
+ * first byte: 0 for big endian, 1 for little
+ * second byte: page shift (eg 12 for 4096 byte pages)
+ * third byte: entry size in bytes (currently either 4 or 8)
+ * fourth byte: header size
+ */
+static ssize_t kpagemap_read(struct file *file, char __user *buf,
+ size_t count, loff_t *ppos)
+{
+ unsigned long *page;
+ struct page *ppage;
+ unsigned long src = *ppos;
+ unsigned long pfn;
+ ssize_t ret = 0;
+ int chunk, i;
+
+ pfn = src / KPMSIZE - 1;
+ count = min_t(size_t, count, ((max_pfn + 1) * KPMSIZE) - src);
+ if (src & KPMMASK || count & KPMMASK)
+ return -EIO;
+
+ page = (unsigned long *)__get_free_page(GFP_USER);
+ if (!page)
+ return -ENOMEM;
+
+ while (count > 0) {
+ chunk = min_t(size_t, count, PAGE_SIZE);
+ i = 0;
+
+ if (pfn == -1) {
+ page[0] = 0;
+ page[1] = 0;
+ ((char *)page)[0] = (ntohl(1) != 1);
+ ((char *)page)[1] = PAGE_SHIFT;
+ ((char *)page)[2] = sizeof(unsigned long);
+ ((char *)page)[3] = KPMSIZE;
+ i = 2;
+ pfn++;
+ }
+
+ for (; i < 2 * chunk / KPMSIZE; i += 2, pfn++) {
+ ppage = pfn_to_page(pfn);
+ if (!ppage) {
+ page[i] = 0;
+ page[i + 1] = 0;
+ } else {
+ page[i] = ppage->flags;
+ page[i + 1] = atomic_read(&ppage->_count);
+ }
+ }
+ chunk = (i / 2) * KPMSIZE;
+
+ if (copy_to_user(buf, page, chunk)) {
+ ret = -EFAULT;
+ break;
+ }
+ ret += chunk;
+ src += chunk;
+ buf += chunk;
+ count -= chunk;
+ cond_resched();
+ }
+ *ppos = src;
+
+ free_page((unsigned long)page);
+ return ret;
+}
+
+struct proc_dir_entry *proc_kpagemap;
+static struct file_operations proc_kpagemap_operations = {
+ .llseek = mem_lseek,
+ .read = kpagemap_read,
+};
+#endif
+
struct proc_dir_entry *proc_root_kcore;
void create_seq_entry(char *name, mode_t mode, const struct file_operations *f)
@@ -812,6 +899,11 @@ void __init proc_misc_init(void)
(size_t)high_memory - PAGE_OFFSET + PAGE_SIZE;
}
#endif
+#ifdef CONFIG_PROC_KPAGEMAP
+ proc_kpagemap = create_proc_entry("kpagemap", S_IRUSR, NULL);
+ if (proc_kpagemap)
+ proc_kpagemap->proc_fops = &proc_kpagemap_operations;
+#endif
#ifdef CONFIG_PROC_VMCORE
proc_vmcore = create_proc_entry("vmcore", S_IRUSR, NULL);
if (proc_vmcore)
Index: mm/init/Kconfig
===================================================================
--- mm.orig/init/Kconfig 2007-04-05 14:18:49.000000000 -0500
+++ mm/init/Kconfig 2007-04-05 14:26:23.000000000 -0500
@@ -612,6 +612,15 @@ config PROC_PAGEMAP
with other processes. Disabling this interface will reduce the
size of the kernel for small machines.
+config PROC_KPAGEMAP
+ default y
+ bool "Enable /proc/kpagemap support" if EMBEDDED && PROC_FS
+ help
+ The /proc/pid/kpagemap interface allows reading the
+ kernel's per-page flag and usage counts to gather precise
+ information on page-level memory usage. Disabling this interface
+ will reduce the size of the kernel for small machines.
+
endmenu # General setup
config RT_MUTEXES
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [EMAIL PROTECTED]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
