All of lore.kernel.org
 help / color / mirror / Atom feed
From: Mingming Cao <cmm@us.ibm.com>
To: linux-ext4@vger.kernel.org
Subject: [RFC][PATCH 3/6] delalloc ENOSPC: reserve blocks
Date: Sun, 01 Jun 2008 16:35:55 -0700	[thread overview]
Message-ID: <1212363355.4368.66.camel@localhost.localdomain> (raw)

ext4: delalloc ENOSPC handling core

From: Mingming cao <cmm@us.ibm.com>

Core part of delaloc ENOSPC (block reservation.) 
data/meta blocks are reserved on write_begin(), and per-inode reserved counters
are updated  after block allocation.
 
Signed-off-by: Mingming cao <cmm@us.ibm.com>

---
 fs/ext4/inode.c |   69 +++++++++++++++++++++++++++++++++++++++++++++++++++++---
 1 file changed, 66 insertions(+), 3 deletions(-)

Index: linux-2.6.26-rc4/fs/ext4/inode.c
===================================================================
--- linux-2.6.26-rc4.orig/fs/ext4/inode.c	2008-06-01 14:26:13.000000000 -0700
+++ linux-2.6.26-rc4/fs/ext4/inode.c	2008-06-01 15:04:06.000000000 -0700
@@ -38,6 +38,7 @@
 #include "ext4_jbd2.h"
 #include "xattr.h"
 #include "acl.h"
+#include "ext4_extents.h"
 
 static void ext4_invalidatepage(struct page *page, unsigned long offset);
 
@@ -1410,6 +1411,61 @@ static int ext4_journalled_write_end(str
 	return ret ? ret : copied;
 }
 
+static int ext4_da_reserve_space(struct inode *inode, int nrblocks)
+{
+       struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
+       unsigned long md_needed, mdblocks, total = 0;
+
+       /*
+        * calculate the amount of metadata blocks to reserve
+	* in order to allocate nrblocks
+	* worse case is one extent per block
+	*/
+	total = EXT4_I(inode)->i_reserved_data_blocks + nrblocks;
+	mdblocks = ext4_ext_calc_metadata_amount(inode, total);
+	BUG_ON(mdblocks < EXT4_I(inode)->i_reserved_meta_blocks);
+
+	md_needed = mdblocks - EXT4_I(inode)->i_reserved_meta_blocks;
+	total = md_needed + nrblocks;
+
+	if (ext4_has_free_blocks(sbi, total) < total)
+		return -ENOSPC;
+
+	/* reduce fs free blocks counter */
+	percpu_counter_sub(&sbi->s_freeblocks_counter, total);
+
+	EXT4_I(inode)->i_reserved_data_blocks += nrblocks;
+	EXT4_I(inode)->i_reserved_meta_blocks += md_needed;
+
+	return 0;       /* success */
+}
+
+static void ext4_da_release_space(struct inode *inode, int used, int to_free)
+{
+	struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
+	int total, mdb, release;
+
+	/* calculate the number of metablocks still need to be reserved */
+	total = EXT4_I(inode)->i_reserved_data_blocks - used - to_free;
+	mdb = ext4_ext_calc_metadata_amount(inode, total);
+
+	/* figure out how many metablocks to release */
+	BUG_ON(mdb > EXT4_I(inode)->i_reserved_meta_blocks);
+	mdb = EXT4_I(inode)->i_reserved_meta_blocks - mdb;
+
+	release = to_free + mdb;
+
+	/* update fs free blocks counter for truncate case */
+	percpu_counter_add(&sbi->s_freeblocks_counter, release);
+
+	/* update per-inode reservations */
+	BUG_ON(used + to_free > EXT4_I(inode)->i_reserved_data_blocks);
+	EXT4_I(inode)->i_reserved_data_blocks -= used + to_free;
+
+	BUG_ON(mdb > EXT4_I(inode)->i_reserved_meta_blocks);
+	EXT4_I(inode)->i_reserved_meta_blocks -= mdb;
+}
+
 /*
  * this is a special callback for ->write_begin() only
  * it's intention is to return mapped block or reserve space
@@ -1428,13 +1484,17 @@ static int ext4_da_get_block_prep(struct
 	 * the same as allocated blocks.
 	 */
 	ret = ext4_get_blocks_wrap(NULL, inode, iblock, 1,  bh_result, 0, 0);
-	if (ret == 0) {
-		/* the block isn't allocated yet, let's reserve space */
-		/* XXX: call reservation here */
+	if ((ret == 0)&& !buffer_delay(bh_result)) {
+		/* the block isn't (pre)allocated yet, let's reserve space */
 		/*
 		 * XXX: __block_prepare_write() unmaps passed block,
 		 * is it OK?
 		 */
+		ret = ext4_da_reserve_space(inode, 1);
+		if (ret)
+			/* not enough space to reserve */
+			return ret;
+
 		map_bh(bh_result, inode->i_sb, 0);
 		set_buffer_new(bh_result);
 		set_buffer_delay(bh_result);
@@ -1463,6 +1523,9 @@ static int ext4_da_get_block_write(struc
 	if (ret > 0) {
 		bh_result->b_size = (ret << inode->i_blkbits);
 
+		/* release reserved-but-unused meta blocks */
+		ext4_da_release_space(inode, ret, 0);
+
 		/*
 		 * Update on-disk size along with block allocation
 		 * we don't use 'extend_disksize' as size may change



             reply	other threads:[~2008-06-01 23:36 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2008-06-01 23:35 Mingming Cao [this message]
2008-06-02  6:44 ` [RFC][PATCH 3/6] delalloc ENOSPC: reserve blocks Aneesh Kumar K.V
2008-06-02  6:55   ` Aneesh Kumar K.V
2008-06-02  8:32 ` Aneesh Kumar K.V

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=1212363355.4368.66.camel@localhost.localdomain \
    --to=cmm@us.ibm.com \
    --cc=linux-ext4@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
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.