linux-kernel - Re: /proc/sys/vm/drop_caches: add error handling

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <20090204182645.GB4797@elf.ucw.cz>
Date:	Wed, 4 Feb 2009 19:26:45 +0100
From:	Pavel Machek <pavel@...e.cz>
To:	Andrew Morton <akpm@...ux-foundation.org>
Cc:	KOSAKI Motohiro <kosaki.motohiro@...fujitsu.com>,
	linux-kernel@...r.kernel.org, linux-mm@...ck.org
Subject: Re: /proc/sys/vm/drop_caches: add error handling

Hi!

> > I think following is clarify more.
> > 
> > 	res = proc_dointvec_minmax(table, write, file, buffer, length, ppos);
> > 	if (res)
> > 		return res;
> > 	if (!write)
> > 		return 0;
> > 	if (sysctl_drop_caches & ~3)
> > 		return -EINVAL;
> > 	if (sysctl_drop_caches & 1)
> > 		drop_pagecache();
> > 	if (sysctl_drop_caches & 2)
> > 		drop_slab();
> > 	return 0;
> > 
> > 
> > otherthings, _very_ looks good to me. :)
> 
> For better or for worse, my intent here was to be
> future-back-compatible.  So if we later add new flags, and people write
> code which uses those new flags, that code won't break on old kernels.

Well, should I move test for &~3 to the end? You'd still get
compatibility, but you'd also get expected error reports.

> Probably that wasn't a very good idea, and such userspace code isn't
> very good.

--- 

Report errors in drop_caches, and document stuff a bit better. 

Signed-off-by: Pavel Machek <pavel@...e.cz>

diff --git a/Documentation/sysctl/vm.txt b/Documentation/sysctl/vm.txt
index 3197fc8..b90f050 100644
--- a/Documentation/sysctl/vm.txt
+++ b/Documentation/sysctl/vm.txt
@@ -132,6 +132,10 @@ To free pagecache, dentries and inodes:
 As this is a non-destructive operation and dirty objects are not freeable, the
 user should run `sync' first.
 
+Note that calling this causes some serious latencies, and that this is
+a debug feature; it should not be used for production as it does not
+contain neccessary locking to guarantee safe operation.
+
 ==============================================================
 
 hugepages_treat_as_movable
diff --git a/fs/drop_caches.c b/fs/drop_caches.c
index 3e5637f..f913a64 100644
--- a/fs/drop_caches.c
+++ b/fs/drop_caches.c
@@ -9,8 +9,9 @@
 #include <linux/sysctl.h>
 #include <linux/gfp.h>
 
-/* A global variable is a bit ugly, but it keeps the code simple */
-int sysctl_drop_caches;
+/* A global variable is a bit ugly, and has locking problems,
+   but it keeps the code simple */
+unsigned int sysctl_drop_caches;
 
 static void drop_pagecache_sb(struct super_block *sb)
 {
@@ -65,12 +66,17 @@ static void drop_slab(void)
 int drop_caches_sysctl_handler(ctl_table *table, int write,
 	struct file *file, void __user *buffer, size_t *length, loff_t *ppos)
 {
-	proc_dointvec_minmax(table, write, file, buffer, length, ppos);
-	if (write) {
-		if (sysctl_drop_caches & 1)
-			drop_pagecache();
-		if (sysctl_drop_caches & 2)
-			drop_slab();
-	}
+	int res;
+	res = proc_dointvec_minmax(table, write, file, buffer, length, ppos);
+	if (res)
+		return res;
+	if (!write)
+		return res;
+	if (sysctl_drop_caches & 1)
+		drop_pagecache();
+	if (sysctl_drop_caches & 2)
+		drop_slab();
+	if (sysctl_drop_caches & ~3)
+		return -EINVAL;
 	return 0;
 }


-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/