linux-kernel - [PATCH 2/5] bdi: Add bdi->id

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

Message-ID: <20190815195750.GC2263813@devbig004.ftw2.facebook.com>
Date:   Thu, 15 Aug 2019 12:57:50 -0700
From:   Tejun Heo <tj@...nel.org>
To:     axboe@...nel.dk, jack@...e.cz, hannes@...xchg.org,
        mhocko@...nel.org, vdavydov.dev@...il.com
Cc:     cgroups@...r.kernel.org, linux-mm@...ck.org,
        linux-block@...r.kernel.org, linux-kernel@...r.kernel.org,
        kernel-team@...com, guro@...com, akpm@...ux-foundation.org
Subject: [PATCH 2/5] bdi: Add bdi->id

There currently is no way to universally identify and lookup a bdi
without holding a reference and pointer to it.  This patch adds an
non-recycling bdi->id and implements bdi_get_by_id() which looks up
bdis by their ids.  This will be used by memcg foreign inode flushing.

I left bdi_list alone for simplicity and because while rb_tree does
support rcu assignment it doesn't seem to guarantee lossless walk when
walk is racing aginst tree rebalance operations.

Signed-off-by: Tejun Heo <tj@...nel.org>
Reviewed-by: Jan Kara <jack@...e.cz>
---
 include/linux/backing-dev-defs.h |    2 +
 include/linux/backing-dev.h      |    1 
 mm/backing-dev.c                 |   65 +++++++++++++++++++++++++++++++++++++--
 3 files changed, 66 insertions(+), 2 deletions(-)

--- a/include/linux/backing-dev-defs.h
+++ b/include/linux/backing-dev-defs.h
@@ -185,6 +185,8 @@ struct bdi_writeback {
 };
 
 struct backing_dev_info {
+	u64 id;
+	struct rb_node rb_node; /* keyed by ->id */
 	struct list_head bdi_list;
 	unsigned long ra_pages;	/* max readahead in PAGE_SIZE units */
 	unsigned long io_pages;	/* max allowed IO size */
--- a/include/linux/backing-dev.h
+++ b/include/linux/backing-dev.h
@@ -24,6 +24,7 @@ static inline struct backing_dev_info *b
 	return bdi;
 }
 
+struct backing_dev_info *bdi_get_by_id(u64 id);
 void bdi_put(struct backing_dev_info *bdi);
 
 __printf(2, 3)
--- a/mm/backing-dev.c
+++ b/mm/backing-dev.c
@@ -1,6 +1,7 @@
 // SPDX-License-Identifier: GPL-2.0-only
 
 #include <linux/wait.h>
+#include <linux/rbtree.h>
 #include <linux/backing-dev.h>
 #include <linux/kthread.h>
 #include <linux/freezer.h>
@@ -22,10 +23,12 @@ EXPORT_SYMBOL_GPL(noop_backing_dev_info)
 static struct class *bdi_class;
 
 /*
- * bdi_lock protects updates to bdi_list. bdi_list has RCU reader side
- * locking.
+ * bdi_lock protects bdi_tree and updates to bdi_list. bdi_list has RCU
+ * reader side locking.
  */
 DEFINE_SPINLOCK(bdi_lock);
+static u64 bdi_id_cursor;
+static struct rb_root bdi_tree = RB_ROOT;
 LIST_HEAD(bdi_list);
 
 /* bdi_wq serves all asynchronous writeback tasks */
@@ -859,9 +862,58 @@ struct backing_dev_info *bdi_alloc_node(
 }
 EXPORT_SYMBOL(bdi_alloc_node);
 
+static struct rb_node **bdi_lookup_rb_node(u64 id, struct rb_node **parentp)
+{
+	struct rb_node **p = &bdi_tree.rb_node;
+	struct rb_node *parent = NULL;
+	struct backing_dev_info *bdi;
+
+	lockdep_assert_held(&bdi_lock);
+
+	while (*p) {
+		parent = *p;
+		bdi = rb_entry(parent, struct backing_dev_info, rb_node);
+
+		if (bdi->id > id)
+			p = &(*p)->rb_left;
+		else if (bdi->id < id)
+			p = &(*p)->rb_right;
+		else
+			break;
+	}
+
+	if (parentp)
+		*parentp = parent;
+	return p;
+}
+
+/**
+ * bdi_get_by_id - lookup and get bdi from its id
+ * @id: bdi id to lookup
+ *
+ * Find bdi matching @id and get it.  Returns NULL if the matching bdi
+ * doesn't exist or is already unregistered.
+ */
+struct backing_dev_info *bdi_get_by_id(u64 id)
+{
+	struct backing_dev_info *bdi = NULL;
+	struct rb_node **p;
+
+	spin_lock_bh(&bdi_lock);
+	p = bdi_lookup_rb_node(id, NULL);
+	if (*p) {
+		bdi = rb_entry(*p, struct backing_dev_info, rb_node);
+		bdi_get(bdi);
+	}
+	spin_unlock_bh(&bdi_lock);
+
+	return bdi;
+}
+
 int bdi_register_va(struct backing_dev_info *bdi, const char *fmt, va_list args)
 {
 	struct device *dev;
+	struct rb_node *parent, **p;
 
 	if (bdi->dev)	/* The driver needs to use separate queues per device */
 		return 0;
@@ -877,7 +929,15 @@ int bdi_register_va(struct backing_dev_i
 	set_bit(WB_registered, &bdi->wb.state);
 
 	spin_lock_bh(&bdi_lock);
+
+	bdi->id = ++bdi_id_cursor;
+
+	p = bdi_lookup_rb_node(bdi->id, &parent);
+	rb_link_node(&bdi->rb_node, parent, p);
+	rb_insert_color(&bdi->rb_node, &bdi_tree);
+
 	list_add_tail_rcu(&bdi->bdi_list, &bdi_list);
+
 	spin_unlock_bh(&bdi_lock);
 
 	trace_writeback_bdi_register(bdi);
@@ -918,6 +978,7 @@ EXPORT_SYMBOL(bdi_register_owner);
 static void bdi_remove_from_list(struct backing_dev_info *bdi)
 {
 	spin_lock_bh(&bdi_lock);
+	rb_erase(&bdi->rb_node, &bdi_tree);
 	list_del_rcu(&bdi->bdi_list);
 	spin_unlock_bh(&bdi_lock);