Linux-bcache Archive on lore.kernel.org
 help / color / Atom feed
WARNING: multiple messages have this Message-ID
From: Coly Li <colyli@suse.de>
To: linux-bcache@vger.kernel.org
Cc: linux-block@vger.kernel.org, Coly Li <colyli@suse.de>
Subject: [RFC PATCH 09/16] bcache: introduce meta_bucket_pages() related helper routines
Date: Sun,  5 Jul 2020 23:55:54 +0800
Message-ID: <20200705155601.5404-10-colyli@suse.de> (raw)
In-Reply-To: <20200705155601.5404-1-colyli@suse.de>

Currently the in-memory meta data like c->uuids or c->disk_buckets
are allocated by alloc_bucket_pages(). The macro alloc_bucket_pages()
calls __get_free_pages() to allocated continuous pages with order
indicated by ilog2(bucket_pages(c)),
 #define alloc_bucket_pages(gfp, c)                      \
     ((void *) __get_free_pages(__GFP_ZERO|gfp, ilog2(bucket_pages(c))))

The maximum order is defined as MAX_ORDER, the default value is 11 (and
can be overwritten by CONFIG_FORCE_MAX_ZONEORDER). In bcache code the
maximum bucket size width is 16bits, this is restricted both by KEY_SIZE
size and bucket_size size from struct cache_sb_disk. The maximum 16bits
width and power-of-2 value is (1<<15) in unit of sector (512byte). It
means the maximum value of bucket size in bytes is (1<<24) bytes a.k.a
4096 pages.

When the bucket size is set to maximum permitted value, ilog2(4096) is
12, which exceeds the default maximum order __get_free_pages() can
accepted, the failed pages allocation will fail cache set registration
procedure and print a kernel oops message for the exceeded pages order.

This patch introduces meta_bucket_pages(), meta_bucket_bytes(), and
alloc_bucket_pages() helper routines. meta_bucket_pages() indicates the
maximum pages can be allocated to meta data bucket, meta_bucket_bytes()
indicates the according maximum bytes, and alloc_bucket_pages() does
the pages allocation for meta bucket. Because meta_bucket_pages()
chooses the smaller value among the bucket size and MAX_ORDER_NR_PAGES,
it still works when MAX_ORDER overwritten by CONFIG_FORCE_MAX_ZONEORDER.

Following patches will use these helper routines to decide maximum pages
can be allocated for different meta data buckets. If the bucket size is
larger than meta_bucket_bytes(), the bcache registration can continue to
success, just the space more than meta_bucket_bytes() inside the bucket
is wasted. Comparing bcache failed for large bucket size, wasting some
space for meta data buckets is acceptable at this moment.

Signed-off-by: Coly Li <colyli@suse.de>
---
 drivers/md/bcache/bcache.h | 20 ++++++++++++++++++++
 drivers/md/bcache/super.c  |  3 +++
 2 files changed, 23 insertions(+)

diff --git a/drivers/md/bcache/bcache.h b/drivers/md/bcache/bcache.h
index 221e0191b687..2bfbb80841af 100644
--- a/drivers/md/bcache/bcache.h
+++ b/drivers/md/bcache/bcache.h
@@ -762,6 +762,26 @@ struct bbio {
 #define bucket_bytes(c)		((c)->sb.bucket_size << 9)
 #define block_bytes(c)		((c)->sb.block_size << 9)
 
+static inline unsigned int meta_bucket_pages(struct cache_sb *sb)
+{
+	unsigned int n, max_pages;
+
+	max_pages = min_t(unsigned int,
+			  __rounddown_pow_of_two(USHRT_MAX) / PAGE_SECTORS,
+			  MAX_ORDER_NR_PAGES);
+
+	n = sb->bucket_size / PAGE_SECTORS;
+	if (n > max_pages)
+		n = max_pages;
+
+	return n;
+}
+
+static inline unsigned int meta_bucket_bytes(struct cache_sb *sb)
+{
+	return meta_bucket_pages(sb) << PAGE_SHIFT;
+}
+
 #define prios_per_bucket(c)				\
 	((bucket_bytes(c) - sizeof(struct prio_set)) /	\
 	 sizeof(struct bucket_disk))
diff --git a/drivers/md/bcache/super.c b/drivers/md/bcache/super.c
index b3248a7b0e20..ec60145277c4 100644
--- a/drivers/md/bcache/super.c
+++ b/drivers/md/bcache/super.c
@@ -1821,6 +1821,9 @@ void bch_cache_set_unregister(struct cache_set *c)
 #define alloc_bucket_pages(gfp, c)			\
 	((void *) __get_free_pages(__GFP_ZERO|gfp, ilog2(bucket_pages(c))))
 
+#define alloc_meta_bucket_pages(gfp, sb)		\
+	((void *) __get_free_pages(__GFP_ZERO|gfp, ilog2(meta_bucket_pages(sb))))
+
 struct cache_set *bch_cache_set_alloc(struct cache_sb *sb)
 {
 	int iter_size;
-- 
2.26.2

From: Coly Li <colyli@suse.de>
To: linux-bcache@vger.kernel.org
Cc: linux-block@vger.kernel.org, Coly Li <colyli@suse.de>
Subject: [RFC PATCH 09/16] bcache: introduce meta_bucket_pages() related helper routines
Date: Sun,  5 Jul 2020 23:55:54 +0800
Message-ID: <20200705155601.5404-10-colyli@suse.de> (raw)
Message-ID: <20200705155554.2ced-n9wpRUiR2T-ieThEHN2tDKB0EwqPybs1Q2uLZ0@z> (raw)
In-Reply-To: <20200705155601.5404-1-colyli@suse.de>

Currently the in-memory meta data like c->uuids or c->disk_buckets
are allocated by alloc_bucket_pages(). The macro alloc_bucket_pages()
calls __get_free_pages() to allocated continuous pages with order
indicated by ilog2(bucket_pages(c)),
 #define alloc_bucket_pages(gfp, c)                      \
     ((void *) __get_free_pages(__GFP_ZERO|gfp, ilog2(bucket_pages(c))))

The maximum order is defined as MAX_ORDER, the default value is 11 (and
can be overwritten by CONFIG_FORCE_MAX_ZONEORDER). In bcache code the
maximum bucket size width is 16bits, this is restricted both by KEY_SIZE
size and bucket_size size from struct cache_sb_disk. The maximum 16bits
width and power-of-2 value is (1<<15) in unit of sector (512byte). It
means the maximum value of bucket size in bytes is (1<<24) bytes a.k.a
4096 pages.

When the bucket size is set to maximum permitted value, ilog2(4096) is
12, which exceeds the default maximum order __get_free_pages() can
accepted, the failed pages allocation will fail cache set registration
procedure and print a kernel oops message for the exceeded pages order.

This patch introduces meta_bucket_pages(), meta_bucket_bytes(), and
alloc_bucket_pages() helper routines. meta_bucket_pages() indicates the
maximum pages can be allocated to meta data bucket, meta_bucket_bytes()
indicates the according maximum bytes, and alloc_bucket_pages() does
the pages allocation for meta bucket. Because meta_bucket_pages()
chooses the smaller value among the bucket size and MAX_ORDER_NR_PAGES,
it still works when MAX_ORDER overwritten by CONFIG_FORCE_MAX_ZONEORDER.

Following patches will use these helper routines to decide maximum pages
can be allocated for different meta data buckets. If the bucket size is
larger than meta_bucket_bytes(), the bcache registration can continue to
success, just the space more than meta_bucket_bytes() inside the bucket
is wasted. Comparing bcache failed for large bucket size, wasting some
space for meta data buckets is acceptable at this moment.

Signed-off-by: Coly Li <colyli@suse.de>
---
 drivers/md/bcache/bcache.h | 20 ++++++++++++++++++++
 drivers/md/bcache/super.c  |  3 +++
 2 files changed, 23 insertions(+)

diff --git a/drivers/md/bcache/bcache.h b/drivers/md/bcache/bcache.h
index 221e0191b687..2bfbb80841af 100644
--- a/drivers/md/bcache/bcache.h
+++ b/drivers/md/bcache/bcache.h
@@ -762,6 +762,26 @@ struct bbio {
 #define bucket_bytes(c)		((c)->sb.bucket_size << 9)
 #define block_bytes(c)		((c)->sb.block_size << 9)
 
+static inline unsigned int meta_bucket_pages(struct cache_sb *sb)
+{
+	unsigned int n, max_pages;
+
+	max_pages = min_t(unsigned int,
+			  __rounddown_pow_of_two(USHRT_MAX) / PAGE_SECTORS,
+			  MAX_ORDER_NR_PAGES);
+
+	n = sb->bucket_size / PAGE_SECTORS;
+	if (n > max_pages)
+		n = max_pages;
+
+	return n;
+}
+
+static inline unsigned int meta_bucket_bytes(struct cache_sb *sb)
+{
+	return meta_bucket_pages(sb) << PAGE_SHIFT;
+}
+
 #define prios_per_bucket(c)				\
 	((bucket_bytes(c) - sizeof(struct prio_set)) /	\
 	 sizeof(struct bucket_disk))
diff --git a/drivers/md/bcache/super.c b/drivers/md/bcache/super.c
index b3248a7b0e20..ec60145277c4 100644
--- a/drivers/md/bcache/super.c
+++ b/drivers/md/bcache/super.c
@@ -1821,6 +1821,9 @@ void bch_cache_set_unregister(struct cache_set *c)
 #define alloc_bucket_pages(gfp, c)			\
 	((void *) __get_free_pages(__GFP_ZERO|gfp, ilog2(bucket_pages(c))))
 
+#define alloc_meta_bucket_pages(gfp, sb)		\
+	((void *) __get_free_pages(__GFP_ZERO|gfp, ilog2(meta_bucket_pages(sb))))
+
 struct cache_set *bch_cache_set_alloc(struct cache_sb *sb)
 {
 	int iter_size;
-- 
2.26.2


  parent reply index

Thread overview: 23+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-07-05 15:55 [RFC PATCH 00/16] bcache: extend bucket size to 32bit width Coly Li
2020-07-05 15:55 ` [RFC PATCH 01/16] bcache: add comments to mark member offset of struct cache_sb_disk Coly Li
2020-07-05 15:55 ` [RFC PATCH 02/16] bcache: add read_super_basic() to read major part of super block Coly Li
2020-07-05 15:55   ` Coly Li
2020-07-05 15:55 ` [RFC PATCH 03/16] bcache: add more accurate error information in read_super_basic() Coly Li
2020-07-05 15:55   ` Coly Li
2020-07-05 15:55 ` [RFC PATCH 04/16] bcache: disassemble the big if() checks in bch_cache_set_alloc() Coly Li
2020-07-05 15:55   ` Coly Li
2020-07-05 15:55 ` [RFC PATCH 05/16] bcache: fix super block seq numbers comparision in register_cache_set() Coly Li
2020-07-05 15:55 ` [RFC PATCH 06/16] bcache: increase super block version for cache device and backing device Coly Li
2020-07-05 15:55 ` [RFC PATCH 07/16] bcache: move bucket related code into read_super_basic() Coly Li
2020-07-05 15:55 ` [RFC PATCH 08/16] bcache: struct cache_sb is only for in-memory super block now Coly Li
2020-07-05 15:55 ` Coly Li [this message]
2020-07-05 15:55   ` [RFC PATCH 09/16] bcache: introduce meta_bucket_pages() related helper routines Coly Li
2020-07-05 15:55 ` [RFC PATCH 10/16] bcache: handle c->uuids properly for bucket size > 8MB Coly Li
2020-07-05 15:55 ` [RFC PATCH 11/16] bcache: handle cache prio_buckets and disk_buckets " Coly Li
2020-07-05 15:55 ` [RFC PATCH 12/16] bcache: handle cache set verify_ondisk " Coly Li
2020-07-05 15:55   ` Coly Li
2020-07-05 15:55 ` [RFC PATCH 13/16] bcache: handle btree node memory allocation " Coly Li
2020-07-05 15:55   ` Coly Li
2020-07-05 15:55 ` [RFC PATCH 14/16] bcache: add bucket_size_hi into struct cache_sb_disk for large bucket Coly Li
2020-07-05 15:56 ` [RFC PATCH 15/16] bcache: avoid extra memory allocation from mempool c->fill_iter Coly Li
2020-07-05 15:56 ` [RFC PATCH 16/16] bcache: avoid extra memory consumption in struct bbio for large bucket size Coly Li

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20200705155601.5404-10-colyli@suse.de \
    --to=colyli@suse.de \
    --cc=linux-bcache@vger.kernel.org \
    --cc=linux-block@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

Linux-bcache Archive on lore.kernel.org

Archives are clonable:
	git clone --mirror https://lore.kernel.org/linux-bcache/0 linux-bcache/git/0.git

	# If you have public-inbox 1.1+ installed, you may
	# initialize and index your mirror using the following commands:
	public-inbox-init -V2 linux-bcache linux-bcache/ https://lore.kernel.org/linux-bcache \
		linux-bcache@vger.kernel.org
	public-inbox-index linux-bcache

Example config snippet for mirrors

Newsgroup available over NNTP:
	nntp://nntp.lore.kernel.org/org.kernel.vger.linux-bcache


AGPL code for this site: git clone https://public-inbox.org/public-inbox.git