After a recent fsmark benchmarking run, I observed that the overhead of
parent pointers on file creation and deletion can be a bit high. On a
machine with 20 CPUs, 128G of memory, and an NVME SSD capable of pushing
750000iops, I see the following results:
$ mkfs.xfs -f -l logdev=/dev/nvme1n1,size=1g /dev/nvme0n1 -n parent=0
meta-data=/dev/nvme0n1 isize=512 agcount=40, agsize=
9767586 blks
= sectsz=4096 attr=2, projid32bit=1
= crc=1 finobt=1, sparse=1, rmapbt=1
= reflink=1 bigtime=1 inobtcount=1 nrext64=1
= exchange=0 metadir=0
data = bsize=4096 blocks=
390703440, imaxpct=5
= sunit=0 swidth=0 blks
naming =version 2 bsize=4096 ascii-ci=0, ftype=1, parent=0
log =/dev/nvme1n1 bsize=4096 blocks=262144, version=2
= sectsz=4096 sunit=1 blks, lazy-count=1
realtime =none extsz=4096 blocks=0, rtextents=0
= rgcount=0 rgsize=0 extents
= zoned=0 start=0 reserved=0
So we created 40 AGs, one per CPU. Now we create 40 directories and run
fsmark:
$ time fs_mark -D 10000 -S 0 -n 100000 -s 0 -L 8 -d ...
# Version 3.3, 40 thread(s) starting at Wed Dec 10 14:22:07 2025
# Sync method: NO SYNC: Test does not issue sync() or fsync() calls.
# Directories: Time based hash between directories across 10000 subdirectories with 180 seconds per subdirectory.
# File names: 40 bytes long, (16 initial bytes of time stamp with 24 random bytes at end of name)
# Files info: size 0 bytes, written with an IO size of 16384 bytes per write
# App overhead is time in microseconds spent in the test not doing file writing related system calls.
parent=0 parent=1
================== ==================
real 0m57.573s real 1m2.934s
user 3m53.578s user 3m53.508s
sys 19m44.440s sys 25m14.810s
$ time rm -rf ...
parent=0 parent=1
================== ==================
real 0m59.649s real 1m12.505s
user 0m41.196s user 0m47.489s
sys 13m9.566s sys 20m33.844s
Parent pointers increase the system time by 28% overhead to create 32
million files that are totally empty. Removing them incurs a system
time increase of 56%. Wall time increases by 9% and 22%.
For most filesystems, each file tends to have a single owner and not
that many xattrs. If the xattr structure is shortform, then all xattr
changes are logged with the inode and do not require the the xattr
intent mechanism to persist the parent pointer.
Therefore, we can speed up parent pointer operations by calling the
shortform xattr functions directly if the child's xattr is in short
format. Now the overhead looks like:
$ time fs_mark -D 10000 -S 0 -n 100000 -s 0 -L 8 -d ...
parent=0 parent=1
================== ==================
real 0m58.030s real 1m0.983s
user 3m54.141s user 3m53.758s
sys 19m57.003s sys 21m30.605s
$ time rm -rf ...
parent=0 parent=1
================== ==================
real 0m58.911s real 1m4.420s
user 0m41.329s user 0m45.169s
sys 13m27.857s sys 15m58.564s
Now parent pointers only increase the system time by 8% for creation and
19% for deletion. Wall time increases by 5% and 9% now.
Close the performance gap by creating helpers for the attr set, remove,
and replace operations that will try to make direct shortform updates,
and fall back to the attr intent machinery if that doesn't work. This
works for regular xattrs and for parent pointers.
Signed-off-by: "Darrick J. Wong" <djwong@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
return error;
}
+/*
+ * Decide if it is theoretically possible to try to bypass the attr intent
+ * mechanism for better performance. Other constraints (e.g. available space
+ * in the existing structure) are not considered here.
+ */
+static inline bool
+xfs_attr_can_shortcut(
+ const struct xfs_inode *ip)
+{
+ return xfs_inode_has_attr_fork(ip) && xfs_attr_is_shortform(ip);
+}
+
+/* Try to set an attr in one transaction or fall back to attr intents. */
+int
+xfs_attr_setname(
+ struct xfs_da_args *args,
+ int rmt_blks)
+{
+ int error;
+
+ if (!rmt_blks && xfs_attr_can_shortcut(args->dp)) {
+ args->op_flags |= XFS_DA_OP_ADDNAME;
+
+ error = xfs_attr_try_sf_addname(args);
+ if (error != -ENOSPC)
+ return error;
+ }
+
+ xfs_attr_defer_add(args, XFS_ATTR_DEFER_SET);
+ return 0;
+}
+
+/* Try to remove an attr in one transaction or fall back to attr intents. */
+int
+xfs_attr_removename(
+ struct xfs_da_args *args)
+{
+ if (xfs_attr_can_shortcut(args->dp))
+ return xfs_attr_sf_removename(args);
+
+ xfs_attr_defer_add(args, XFS_ATTR_DEFER_REMOVE);
+ return 0;
+}
+
+/* Try to replace an attr in one transaction or fall back to attr intents. */
+int
+xfs_attr_replacename(
+ struct xfs_da_args *args,
+ int rmt_blks)
+{
+ int error;
+
+ if (rmt_blks || !xfs_attr_can_shortcut(args->dp)) {
+ xfs_attr_defer_add(args, XFS_ATTR_DEFER_REPLACE);
+ return 0;
+ }
+
+ args->op_flags |= XFS_DA_OP_ADDNAME | XFS_DA_OP_REPLACE;
+
+ error = xfs_attr_sf_removename(args);
+ if (error)
+ return error;
+
+ if (args->attr_filter & XFS_ATTR_PARENT) {
+ /*
+ * Move the new name/value to the regular name/value slots and
+ * zero out the new name/value slots because we don't need to
+ * log them for a PPTR_SET operation.
+ */
+ xfs_attr_update_pptr_replace_args(args);
+ args->new_name = NULL;
+ args->new_namelen = 0;
+ args->new_value = NULL;
+ args->new_valuelen = 0;
+ }
+ args->op_flags &= ~XFS_DA_OP_REPLACE;
+
+ error = xfs_attr_try_sf_addname(args);
+ if (error != -ENOSPC)
+ return error;
+
+ xfs_attr_defer_add(args, XFS_ATTR_DEFER_SET);
+ return 0;
+}
+
/*
* Make a change to the xattr structure.
*
case -EEXIST:
if (op == XFS_ATTRUPDATE_REMOVE) {
/* if no value, we are performing a remove operation */
- xfs_attr_defer_add(args, XFS_ATTR_DEFER_REMOVE);
+ error = xfs_attr_removename(args);
+ if (error)
+ goto out_trans_cancel;
break;
}
/* Pure create fails if the attr already exists */
if (op == XFS_ATTRUPDATE_CREATE)
goto out_trans_cancel;
- xfs_attr_defer_add(args, XFS_ATTR_DEFER_REPLACE);
+
+ error = xfs_attr_replacename(args, rmt_blks);
+ if (error)
+ goto out_trans_cancel;
break;
case -ENOATTR:
/* Can't remove what isn't there. */
/* Pure replace fails if no existing attr to replace. */
if (op == XFS_ATTRUPDATE_REPLACE)
goto out_trans_cancel;
- xfs_attr_defer_add(args, XFS_ATTR_DEFER_SET);
+
+ error = xfs_attr_setname(args, rmt_blks);
+ if (error)
+ goto out_trans_cancel;
break;
default:
goto out_trans_cancel;
*/
static inline bool
xfs_attr_is_shortform(
- struct xfs_inode *ip)
+ const struct xfs_inode *ip)
{
return ip->i_af.if_format == XFS_DINODE_FMT_LOCAL ||
(ip->i_af.if_format == XFS_DINODE_FMT_EXTENTS &&
int xfs_attr_sf_totsize(struct xfs_inode *dp);
int xfs_attr_add_fork(struct xfs_inode *ip, int size, int rsvd);
+int xfs_attr_setname(struct xfs_da_args *args, int rmt_blks);
+int xfs_attr_removename(struct xfs_da_args *args);
+int xfs_attr_replacename(struct xfs_da_args *args, int rmt_blks);
+
#endif /* __XFS_ATTR_H__ */
#include "xfs_trans_space.h"
#include "xfs_attr_item.h"
#include "xfs_health.h"
+#include "xfs_attr_leaf.h"
struct kmem_cache *xfs_parent_args_cache;
xfs_inode_to_parent_rec(&ppargs->rec, dp);
xfs_parent_da_args_init(&ppargs->args, tp, &ppargs->rec, child,
child->i_ino, parent_name);
- xfs_attr_defer_add(&ppargs->args, XFS_ATTR_DEFER_SET);
- return 0;
+
+ return xfs_attr_setname(&ppargs->args, 0);
}
/* Remove a parent pointer to reflect a dirent removal. */
xfs_inode_to_parent_rec(&ppargs->rec, dp);
xfs_parent_da_args_init(&ppargs->args, tp, &ppargs->rec, child,
child->i_ino, parent_name);
- xfs_attr_defer_add(&ppargs->args, XFS_ATTR_DEFER_REMOVE);
- return 0;
+
+ return xfs_attr_removename(&ppargs->args);
}
/* Replace one parent pointer with another to reflect a rename. */
child->i_ino, old_name);
xfs_inode_to_parent_rec(&ppargs->new_rec, new_dp);
+
ppargs->args.new_name = new_name->name;
ppargs->args.new_namelen = new_name->len;
ppargs->args.new_value = &ppargs->new_rec;
ppargs->args.new_valuelen = sizeof(struct xfs_parent_rec);
- xfs_attr_defer_add(&ppargs->args, XFS_ATTR_DEFER_REPLACE);
- return 0;
+
+ return xfs_attr_replacename(&ppargs->args, 0);
}
/*
projects / thirdparty / kernel / linux.git / commitdiff
