Re: [RFC PATCH v2 1/3] meminfo_extra: introduce meminfo extra

On 2020년 03월 23일 18:53, Greg KH wrote:

On Mon, Mar 23, 2020 at 05:05:01PM +0900, Jaewon Kim wrote:

quoted

Provide APIs to drivers so that they can show its memory usage on
/proc/meminfo_extra.

int register_meminfo_extra(atomic_long_t *val, int shift,
			   const char *name);
int unregister_meminfo_extra(atomic_long_t *val);

Nit, isn't it nicer to have the subsystem name first:
	meminfo_extra_register()
	meminfo_extra_unregister()
?

OK. Name can be changed.

quoted

Signed-off-by: Jaewon Kim <redacted>
---
v2: move to /proc/meminfo_extra as a new file, meminfo_extra.c
    use rcu to reduce lock overhead
v1: print info at /proc/meminfo
---
 fs/proc/Makefile        |   1 +
 fs/proc/meminfo_extra.c | 123 ++++++++++++++++++++++++++++++++++++++++++++++++
 include/linux/mm.h      |   4 ++
 mm/page_alloc.c         |   1 +
 4 files changed, 129 insertions(+)
 create mode 100644 fs/proc/meminfo_extra.c

diff --git a/fs/proc/Makefile b/fs/proc/Makefile
index bd08616ed8ba..83d2f55591c6 100644
--- a/fs/proc/Makefile
+++ b/fs/proc/Makefile

@@ -19,6 +19,7 @@ proc-y	+= devices.o
 proc-y	+= interrupts.o
 proc-y	+= loadavg.o
 proc-y	+= meminfo.o
+proc-y	+= meminfo_extra.o
 proc-y	+= stat.o
 proc-y	+= uptime.o
 proc-y	+= util.o

diff --git a/fs/proc/meminfo_extra.c b/fs/proc/meminfo_extra.c
new file mode 100644
index 000000000000..bd3f0d2b7fb7
--- /dev/null
+++ b/fs/proc/meminfo_extra.c

@@ -0,0 +1,123 @@
+// SPDX-License-Identifier: GPL-2.0
+#include <linux/mm.h>
+#include <linux/proc_fs.h>
+#include <linux/seq_file.h>
+#include <linux/slab.h>
+
+static void show_val_kb(struct seq_file *m, const char *s, unsigned long num)
+{
+	seq_put_decimal_ull_width(m, s, num << (PAGE_SHIFT - 10), 8);
+	seq_write(m, " kB\n", 4);
+}
+
+static LIST_HEAD(meminfo_head);
+static DEFINE_SPINLOCK(meminfo_lock);
+
+#define NAME_SIZE      15
+#define NAME_BUF_SIZE  (NAME_SIZE + 2) /* ':' and '\0' */
+
+struct meminfo_extra {
+	struct list_head list;
+	atomic_long_t *val;
+	int shift_for_page;
+	char name[NAME_BUF_SIZE];
+	char name_pad[NAME_BUF_SIZE];
+};
+
+int register_meminfo_extra(atomic_long_t *val, int shift, const char *name)
+{
+	struct meminfo_extra *meminfo, *memtemp;
+	int len;
+	int error = 0;
+
+	meminfo = kzalloc(sizeof(*meminfo), GFP_KERNEL);
+	if (!meminfo) {
+		error = -ENOMEM;
+		goto out;
+	}
+
+	meminfo->val = val;
+	meminfo->shift_for_page = shift;
+	strncpy(meminfo->name, name, NAME_SIZE);
+	len = strlen(meminfo->name);
+	meminfo->name[len] = ':';
+	strncpy(meminfo->name_pad, meminfo->name, NAME_BUF_SIZE);
+	while (++len < NAME_BUF_SIZE - 1)
+		meminfo->name_pad[len] = ' ';
+
+	spin_lock(&meminfo_lock);
+	list_for_each_entry_rcu(memtemp, &meminfo_head, list) {
+		if (memtemp->val == val) {
+			error = -EINVAL;
+			break;
+		}
+	}
+	if (!error)
+		list_add_tail_rcu(&meminfo->list, &meminfo_head);
+	spin_unlock(&meminfo_lock);

If you have a lock, why are you needing rcu?

I think _rcu should be removed out of list_for_each_entry_rcu.
But I'm confused about what you meant.
I used rcu_read_lock on __meminfo_extra,
and I think spin_lock is also needed for addition and deletion to handle multiple modifiers.

quoted

+	if (error)
+		kfree(meminfo);
+out:
+
+	return error;
+}
+EXPORT_SYMBOL(register_meminfo_extra);

EXPORT_SYMBOL_GPL()?  I have to ask :)

I can use EXPORT_SYMBOL_GPL.

thanks,

greg k-h

Hello
Thank you for your comment.

By the way there was not resolved discussion on v1 patch as I mentioned on cover page.
I'd like to hear your opinion on this /proc/meminfo_extra node.
Do you think this is meaningful or cannot co-exist with other future sysfs based API.