[RFC] [PATCH 0/2] OrangeFS in-kernel lookup and attribute cache

linux-fsdevel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed

* [RFC] [PATCH 0/2] OrangeFS in-kernel lookup and attribute cache
@ 2016-07-05 19:13 Martin Brandenburg
  2016-07-05 19:13 ` [PATCH 1/2] orangefs: Use d_time to avoid excessive lookups Martin Brandenburg
  2016-07-05 19:13 ` [PATCH 2/2] orangefs: Cache getattr results Martin Brandenburg
  0 siblings, 2 replies; 3+ messages in thread
From: Martin Brandenburg @ 2016-07-05 19:13 UTC (permalink / raw)
  To: linux-fsdevel; +Cc: Martin Brandenburg

This patch introduces caching for OrangeFS server lookups and getattrs.
The userspace OrangeFS client has done this in the past. We see a
significant (3 to 4 times) performance improvement with these patches
at a cost of consistency across the cluster, which OrangeFS doesn't
really guarantee anyway.

Martin Brandenburg (2):
  orangefs: Use d_time to avoid excessive lookups
  orangefs: Cache getattr results.

 fs/orangefs/dcache.c          |  4 ++++
 fs/orangefs/inode.c           |  6 +++---
 fs/orangefs/namei.c           | 12 ++++++++++++
 fs/orangefs/orangefs-kernel.h |  4 +++-
 fs/orangefs/orangefs-utils.c  | 38 +++++++++++++++++++++-----------------
 fs/orangefs/protocol.h        |  8 --------
 6 files changed, 43 insertions(+), 29 deletions(-)

-- 
2.7.4

^ permalink raw reply	[flat|nested] 3+ messages in thread

* [PATCH 1/2] orangefs: Use d_time to avoid excessive lookups
  2016-07-05 19:13 [RFC] [PATCH 0/2] OrangeFS in-kernel lookup and attribute cache Martin Brandenburg
@ 2016-07-05 19:13 ` Martin Brandenburg
  2016-07-05 19:13 ` [PATCH 2/2] orangefs: Cache getattr results Martin Brandenburg
  1 sibling, 0 replies; 3+ messages in thread
From: Martin Brandenburg @ 2016-07-05 19:13 UTC (permalink / raw)
  To: linux-fsdevel; +Cc: Martin Brandenburg

From: Martin Brandenburg <martin@omnibond.com>

Signed-off-by: Martin Brandenburg <martin@omnibond.com>
---
 fs/orangefs/dcache.c | 4 ++++
 fs/orangefs/namei.c  | 5 +++++
 2 files changed, 9 insertions(+)

diff --git a/fs/orangefs/dcache.c b/fs/orangefs/dcache.c
index 5dfc4f3..dcb55bb 100644
--- a/fs/orangefs/dcache.c
+++ b/fs/orangefs/dcache.c
@@ -73,6 +73,7 @@ static int orangefs_revalidate_lookup(struct dentry *dentry)
 		}
 	}
 
+	dentry->d_time = jiffies + HZ;
 	ret = 1;
 out_release_op:
 	op_release(new_op);
@@ -94,6 +95,9 @@ static int orangefs_d_revalidate(struct dentry *dentry, unsigned int flags)
 {
 	int ret;
 
+	if (dentry->d_time > jiffies)
+		return 1;
+
 	if (flags & LOOKUP_RCU)
 		return -ECHILD;
 
diff --git a/fs/orangefs/namei.c b/fs/orangefs/namei.c
index 5a60c50..5e32e0e 100644
--- a/fs/orangefs/namei.c
+++ b/fs/orangefs/namei.c
@@ -72,6 +72,7 @@ static int orangefs_create(struct inode *dir,
 
 	d_instantiate(dentry, inode);
 	unlock_new_inode(inode);
+	dentry->d_time = jiffies + HZ;
 
 	gossip_debug(GOSSIP_NAME_DEBUG,
 		     "%s: dentry instantiated for %s\n",
@@ -181,6 +182,8 @@ static struct dentry *orangefs_lookup(struct inode *dir, struct dentry *dentry,
 		goto out;
 	}
 
+	dentry->d_time = jiffies + HZ;
+
 	inode = orangefs_iget(dir->i_sb, &new_op->downcall.resp.lookup.refn);
 	if (IS_ERR(inode)) {
 		gossip_debug(GOSSIP_NAME_DEBUG,
@@ -316,6 +319,7 @@ static int orangefs_symlink(struct inode *dir,
 
 	d_instantiate(dentry, inode);
 	unlock_new_inode(inode);
+	dentry->d_time = jiffies + HZ;
 
 	gossip_debug(GOSSIP_NAME_DEBUG,
 		     "Inode (Symlink) %pU -> %s\n",
@@ -378,6 +382,7 @@ static int orangefs_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode
 
 	d_instantiate(dentry, inode);
 	unlock_new_inode(inode);
+	dentry->d_time = jiffies + HZ;
 
 	gossip_debug(GOSSIP_NAME_DEBUG,
 		     "Inode (Directory) %pU -> %s\n",
-- 
2.7.4


^ permalink raw reply related	[flat|nested] 3+ messages in thread

* [PATCH 2/2] orangefs: Cache getattr results.
  2016-07-05 19:13 [RFC] [PATCH 0/2] OrangeFS in-kernel lookup and attribute cache Martin Brandenburg
  2016-07-05 19:13 ` [PATCH 1/2] orangefs: Use d_time to avoid excessive lookups Martin Brandenburg
@ 2016-07-05 19:13 ` Martin Brandenburg
  1 sibling, 0 replies; 3+ messages in thread
From: Martin Brandenburg @ 2016-07-05 19:13 UTC (permalink / raw)
  To: linux-fsdevel; +Cc: Martin Brandenburg

From: Martin Brandenburg <martin@omnibond.com>

The userspace component attempts to do this, but this will prevent
us from even needing to go into userspace to satisfy certain getattr
requests.

Signed-off-by: Martin Brandenburg <martin@omnibond.com>
---
 fs/orangefs/inode.c           |  6 +++---
 fs/orangefs/namei.c           |  7 +++++++
 fs/orangefs/orangefs-kernel.h |  4 +++-
 fs/orangefs/orangefs-utils.c  | 38 +++++++++++++++++++++-----------------
 fs/orangefs/protocol.h        |  8 --------
 5 files changed, 34 insertions(+), 29 deletions(-)

diff --git a/fs/orangefs/inode.c b/fs/orangefs/inode.c
index 85640e9..6bf7a19 100644
--- a/fs/orangefs/inode.c
+++ b/fs/orangefs/inode.c
@@ -265,7 +265,7 @@ int orangefs_getattr(struct vfsmount *mnt,
 		     "orangefs_getattr: called on %s\n",
 		     dentry->d_name.name);
 
-	ret = orangefs_inode_getattr(inode, 0, 1);
+	ret = orangefs_inode_getattr(inode, 0, 0);
 	if (ret == 0) {
 		generic_fillattr(inode, kstat);
 
@@ -387,7 +387,7 @@ struct inode *orangefs_iget(struct super_block *sb, struct orangefs_object_kref
 	if (!inode || !(inode->i_state & I_NEW))
 		return inode;
 
-	error = orangefs_inode_getattr(inode, 1, 0);
+	error = orangefs_inode_getattr(inode, 1, 1);
 	if (error) {
 		iget_failed(inode);
 		return ERR_PTR(error);
@@ -432,7 +432,7 @@ struct inode *orangefs_new_inode(struct super_block *sb, struct inode *dir,
 	orangefs_set_inode(inode, ref);
 	inode->i_ino = hash;	/* needed for stat etc */
 
-	error = orangefs_inode_getattr(inode, 1, 0);
+	error = orangefs_inode_getattr(inode, 1, 1);
 	if (error)
 		goto out_iput;
 
diff --git a/fs/orangefs/namei.c b/fs/orangefs/namei.c
index 5e32e0e..ffe6377 100644
--- a/fs/orangefs/namei.c
+++ b/fs/orangefs/namei.c
@@ -73,6 +73,7 @@ static int orangefs_create(struct inode *dir,
 	d_instantiate(dentry, inode);
 	unlock_new_inode(inode);
 	dentry->d_time = jiffies + HZ;
+	ORANGEFS_I(inode)->getattr_time = 0;
 
 	gossip_debug(GOSSIP_NAME_DEBUG,
 		     "%s: dentry instantiated for %s\n",
@@ -192,6 +193,8 @@ static struct dentry *orangefs_lookup(struct inode *dir, struct dentry *dentry,
 		goto out;
 	}
 
+	ORANGEFS_I(inode)->getattr_time = 0;
+
 	gossip_debug(GOSSIP_NAME_DEBUG,
 		     "%s:%s:%d "
 		     "Found good inode [%lu] with count [%d]\n",
@@ -320,6 +323,7 @@ static int orangefs_symlink(struct inode *dir,
 	d_instantiate(dentry, inode);
 	unlock_new_inode(inode);
 	dentry->d_time = jiffies + HZ;
+	ORANGEFS_I(inode)->getattr_time = 0;
 
 	gossip_debug(GOSSIP_NAME_DEBUG,
 		     "Inode (Symlink) %pU -> %s\n",
@@ -383,6 +387,7 @@ static int orangefs_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode
 	d_instantiate(dentry, inode);
 	unlock_new_inode(inode);
 	dentry->d_time = jiffies + HZ;
+	ORANGEFS_I(inode)->getattr_time = 0;
 
 	gossip_debug(GOSSIP_NAME_DEBUG,
 		     "Inode (Directory) %pU -> %s\n",
@@ -417,6 +422,8 @@ static int orangefs_rename(struct inode *old_dir,
 		     new_dentry->d_name.name,
 		     d_count(new_dentry));
 
+	ORANGEFS_I(new_dentry->d_parent->d_inode)->getattr_time = 0;
+
 	new_op = op_alloc(ORANGEFS_VFS_OP_RENAME);
 	if (!new_op)
 		return -EINVAL;
diff --git a/fs/orangefs/orangefs-kernel.h b/fs/orangefs/orangefs-kernel.h
index a9925e2..f6f58b7 100644
--- a/fs/orangefs/orangefs-kernel.h
+++ b/fs/orangefs/orangefs-kernel.h
@@ -257,6 +257,8 @@ struct orangefs_inode_s {
 	 * with this object
 	 */
 	unsigned long pinode_flags;
+
+	unsigned long getattr_time;
 };
 
 #define P_ATIME_FLAG 0
@@ -540,7 +542,7 @@ int orangefs_inode_setxattr(struct inode *inode,
 			 size_t size,
 			 int flags);
 
-int orangefs_inode_getattr(struct inode *inode, int new, int size);
+int orangefs_inode_getattr(struct inode *inode, int new, int bypass);
 
 int orangefs_inode_check_changed(struct inode *inode);
 
diff --git a/fs/orangefs/orangefs-utils.c b/fs/orangefs/orangefs-utils.c
index 2d129b5..1eccc0c 100644
--- a/fs/orangefs/orangefs-utils.c
+++ b/fs/orangefs/orangefs-utils.c
@@ -251,7 +251,7 @@ static int orangefs_inode_is_stale(struct inode *inode, int new,
 	return 0;
 }
 
-int orangefs_inode_getattr(struct inode *inode, int new, int size)
+int orangefs_inode_getattr(struct inode *inode, int new, int bypass)
 {
 	struct orangefs_inode_s *orangefs_inode = ORANGEFS_I(inode);
 	struct orangefs_kernel_op_s *new_op;
@@ -261,12 +261,16 @@ int orangefs_inode_getattr(struct inode *inode, int new, int size)
 	gossip_debug(GOSSIP_UTILS_DEBUG, "%s: called on inode %pU\n", __func__,
 	    get_khandle_from_ino(inode));
 
+	if (!new && !bypass) {
+		if (orangefs_inode->getattr_time > jiffies)
+			return 0;
+	}
+
 	new_op = op_alloc(ORANGEFS_VFS_OP_GETATTR);
 	if (!new_op)
 		return -ENOMEM;
 	new_op->upcall.req.getattr.refn = orangefs_inode->refn;
-	new_op->upcall.req.getattr.mask = size ?
-	    ORANGEFS_ATTR_SYS_ALL_NOHINT : ORANGEFS_ATTR_SYS_ALL_NOHINT_NOSIZE;
+	new_op->upcall.req.getattr.mask = ORANGEFS_ATTR_SYS_ALL_NOHINT;
 
 	ret = service_operation(new_op, __func__,
 	    get_interruptible_flag(inode));
@@ -287,20 +291,18 @@ int orangefs_inode_getattr(struct inode *inode, int new, int size)
 	case S_IFREG:
 		inode->i_flags = orangefs_inode_flags(&new_op->
 		    downcall.resp.getattr.attributes);
-		if (size) {
-			inode_size = (loff_t)new_op->
-			    downcall.resp.getattr.attributes.size;
-			rounded_up_size =
-			    (inode_size + (4096 - (inode_size % 4096)));
-			inode->i_size = inode_size;
-			orangefs_inode->blksize =
-			    new_op->downcall.resp.getattr.attributes.blksize;
-			spin_lock(&inode->i_lock);
-			inode->i_bytes = inode_size;
-			inode->i_blocks =
-			    (unsigned long)(rounded_up_size / 512);
-			spin_unlock(&inode->i_lock);
-		}
+	       	inode_size = (loff_t)new_op->
+	       	    downcall.resp.getattr.attributes.size;
+	       	rounded_up_size =
+	       	    (inode_size + (4096 - (inode_size % 4096)));
+	       	inode->i_size = inode_size;
+       		orangefs_inode->blksize =
+       		    new_op->downcall.resp.getattr.attributes.blksize;
+       		spin_lock(&inode->i_lock);
+       		inode->i_bytes = inode_size;
+       		inode->i_blocks =
+       		    (unsigned long)(rounded_up_size / 512);
+       		spin_unlock(&inode->i_lock);
 		break;
 	case S_IFDIR:
 		inode->i_size = PAGE_SIZE;
@@ -345,6 +347,7 @@ int orangefs_inode_getattr(struct inode *inode, int new, int size)
 	inode->i_mode = type | (is_root_handle(inode) ? S_ISVTX : 0) |
 	    orangefs_inode_perms(&new_op->downcall.resp.getattr.attributes);
 
+	orangefs_inode->getattr_time = jiffies + HZ;
 	ret = 0;
 out:
 	op_release(new_op);
@@ -418,6 +421,7 @@ int orangefs_inode_setattr(struct inode *inode, struct iattr *iattr)
 		ClearMtimeFlag(orangefs_inode);
 		ClearCtimeFlag(orangefs_inode);
 		ClearModeFlag(orangefs_inode);
+		orangefs_inode->getattr_time = 0;
 	}
 
 	return ret;
diff --git a/fs/orangefs/protocol.h b/fs/orangefs/protocol.h
index 1efc6f8..3d7418c 100644
--- a/fs/orangefs/protocol.h
+++ b/fs/orangefs/protocol.h
@@ -207,14 +207,6 @@ typedef __s64 ORANGEFS_offset;
 	 ORANGEFS_ATTR_SYS_DIRENT_COUNT		|	\
 	 ORANGEFS_ATTR_SYS_BLKSIZE)
 
-#define ORANGEFS_ATTR_SYS_ALL_NOHINT_NOSIZE		\
-	(ORANGEFS_ATTR_SYS_COMMON_ALL		|	\
-	 ORANGEFS_ATTR_SYS_LNK_TARGET		|	\
-	 ORANGEFS_ATTR_SYS_DFILE_COUNT		|	\
-	 ORANGEFS_ATTR_SYS_MIRROR_COPIES_COUNT	|	\
-	 ORANGEFS_ATTR_SYS_DIRENT_COUNT		|	\
-	 ORANGEFS_ATTR_SYS_BLKSIZE)
-
 #define ORANGEFS_XATTR_REPLACE 0x2
 #define ORANGEFS_XATTR_CREATE  0x1
 #define ORANGEFS_MAX_SERVER_ADDR_LEN 256
-- 
2.7.4


^ permalink raw reply related	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2016-07-05 19:13 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2016-07-05 19:13 [RFC] [PATCH 0/2] OrangeFS in-kernel lookup and attribute cache Martin Brandenburg
2016-07-05 19:13 ` [PATCH 1/2] orangefs: Use d_time to avoid excessive lookups Martin Brandenburg
2016-07-05 19:13 ` [PATCH 2/2] orangefs: Cache getattr results Martin Brandenburg

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).