From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mail.transmode.se ([83.241.175.147] helo=tmnt04.transmode.se) by bombadil.infradead.org with esmtp (Exim 4.68 #1 (Red Hat Linux)) id 1JMhOA-0002kg-4p for linux-mtd@lists.infradead.org; Wed, 06 Feb 2008 10:20:28 +0000 Subject: Re: [PATCH] [JFFS2] Fix free space leaking From: Joakim Tjernlund To: Damir Shayhutdinov In-Reply-To: <20071113102323.GA5933@damir.rnd.local> References: <20071113102323.GA5933@damir.rnd.local> Content-Type: text/plain Content-Transfer-Encoding: 7bit Date: Wed, 06 Feb 2008 11:20:19 +0100 Message-Id: <1202293219.25864.45.camel@gentoo-jocke.transmode.se> Mime-Version: 1.0 Cc: linux-mtd@lists.infradead.org Reply-To: joakim.tjernlund@transmode.se List-Id: Linux MTD discussion mailing list List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , On Tue, 2007-11-13 at 13:23 +0300, Damir Shayhutdinov wrote: > This patch is addressed to fix very-long-standing problem in JFFS2, > first described in 2004: > http://lists.infradead.org/pipermail/linux-mtd/2004-March/009456.html > > jffs2_link_node_ref() decreases c->free_size by > c->cleanmarker_size but the clean marker space can't > be accounted as free space! So we just compensate > the difference. > > Signed-off-by: Alexander Yurchenko > Signed-off-by: Damir Shayhutdinov > --- > fs/jffs2/erase.c | 11 +++++++++++ > 1 files changed, 11 insertions(+), 0 deletions(-) > > diff --git a/fs/jffs2/erase.c b/fs/jffs2/erase.c > index a1db918..c574fa3 100644 > --- a/fs/jffs2/erase.c > +++ b/fs/jffs2/erase.c > @@ -454,6 +454,17 @@ static void jffs2_mark_erased_block(struct jffs2_sb_info *c, struct jffs2_eraseb > jeb->free_size = c->sector_size; > /* FIXME Special case for cleanmarker in empty block */ > jffs2_link_node_ref(c, jeb, jeb->offset | REF_NORMAL, c->cleanmarker_size, NULL); > + /* > + * XXX: I'm not sure this is correct but it prevents > + * c->free_size from slow leaking under a frequent file > + * overwriting. > + * jffs2_link_node_ref() decreases c->free_size by > + * c->cleanmarker_size but the clean marker space can't > + * be accounted as free space! So we just compensate > + * the difference. > + */ > + c->free_size += c->cleanmarker_size; > + c->used_size -= c->cleanmarker_size; > } > > down(&c->erase_free_sem); What happened to this patch? Just forgotten or was it rejected? Jocke