c/mirror_zfs
1
0
Fork 0
mirror of https://git.proxmox.com/git/mirror_zfs.git synced 2024-11-17 18:11:00 +03:00
Code Issues Packages Projects Releases Wiki Activity
701b1f8168
mirror_zfs/module/zfs/zfs_sa.c
Brian Behlendorf 5484965ab6 Drop HAVE_XVATTR macros 
When I began work on the Posix layer it immediately became clear to
me that to integrate cleanly with the Linux VFS certain Solaris
specific things would have to go.  One of these things was to elimate
as many Solaris specific types from the ZPL layer as possible.  They
would be replaced with their Linux equivalents.  This would not only
be good for performance, but for the general readability and health of
the code.  The Solaris and Linux VFS are different beasts and should
be treated as such.  Most of the code remains common for constructing
transactions and such, but there are subtle and important differenced
which need to be repsected.

This policy went quite for for certain types such as the vnode_t,
and it initially seemed to be working out well for the vattr_t.  There
was a relatively small amount of related xvattr_t code I was forced to
comment out with HAVE_XVATTR.  But it didn't look that hard to come
back soon and replace it all with a native Linux type.

However, after going doing this path with xvattr some distance it
clear that this code was woven in the ZPL more deeply than I thought.
In particular its hooks went very deep in to the ZPL replay code
and replacing it would not be as easy as I originally thought.

Rather than continue persuing replacing and removing this code I've
taken a step back and reevaluted things.  This commit reverts many of
my previous commits which removed xvattr related code.  It restores
much of the code to its original upstream state and now relies on
improved xvattr_t support in the zfs package itself.

The result of this is that much of the code which I had commented
out, which accidentally broke things like replay, is now back in
place and working.  However, there may be a small performance
impact for getattr/setattr operations because they now require
a translation from native Linux to Solaris types.  For now that's
a price I'm willing to pay.  Once everything is completely functional
we can revisting the issue of removing the vattr_t/xvattr_t types.

Closes #111
2011-03-02 11:44:34 -08:00

338 lines
10 KiB
C
Raw Blame History

/*
* CDDL HEADER START
*
* The contents of this file are subject to the terms of the
* Common Development and Distribution License (the "License").
* You may not use this file except in compliance with the License.
*
* You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
* or http://www.opensolaris.org/os/licensing.
* See the License for the specific language governing permissions
* and limitations under the License.
*
* When distributing Covered Code, include this CDDL HEADER in each
* file and include the License file at usr/src/OPENSOLARIS.LICENSE.
* If applicable, add the following below this CDDL HEADER, with the
* fields enclosed by brackets "[]" replaced with your own identifying
* information: Portions Copyright [yyyy] [name of copyright owner]
*
* CDDL HEADER END
*/
/*
* Copyright (c) 2010, Oracle and/or its affiliates. All rights reserved.
*/
#include <sys/types.h>
#include <sys/param.h>
#include <sys/vnode.h>
#include <sys/sa.h>
#include <sys/zfs_acl.h>
#include <sys/zfs_sa.h>
/*
* ZPL attribute registration table.
* Order of attributes doesn't matter
* a unique value will be assigned for each
* attribute that is file system specific
*
* This is just the set of ZPL attributes that this
* version of ZFS deals with natively. The file system
* could have other attributes stored in files, but they will be
* ignored. The SA framework will preserve them, just that
* this version of ZFS won't change or delete them.
*/
sa_attr_reg_t zfs_attr_table[ZPL_END+1] = {
{"ZPL_ATIME", sizeof (uint64_t) * 2, SA_UINT64_ARRAY, 0},
{"ZPL_MTIME", sizeof (uint64_t) * 2, SA_UINT64_ARRAY, 1},
{"ZPL_CTIME", sizeof (uint64_t) * 2, SA_UINT64_ARRAY, 2},
{"ZPL_CRTIME", sizeof (uint64_t) * 2, SA_UINT64_ARRAY, 3},
{"ZPL_GEN", sizeof (uint64_t), SA_UINT64_ARRAY, 4},
{"ZPL_MODE", sizeof (uint64_t), SA_UINT64_ARRAY, 5},
{"ZPL_SIZE", sizeof (uint64_t), SA_UINT64_ARRAY, 6},
{"ZPL_PARENT", sizeof (uint64_t), SA_UINT64_ARRAY, 7},
{"ZPL_LINKS", sizeof (uint64_t), SA_UINT64_ARRAY, 8},
{"ZPL_XATTR", sizeof (uint64_t), SA_UINT64_ARRAY, 9},
{"ZPL_RDEV", sizeof (uint64_t), SA_UINT64_ARRAY, 10},
{"ZPL_FLAGS", sizeof (uint64_t), SA_UINT64_ARRAY, 11},
{"ZPL_UID", sizeof (uint64_t), SA_UINT64_ARRAY, 12},
{"ZPL_GID", sizeof (uint64_t), SA_UINT64_ARRAY, 13},
{"ZPL_PAD", sizeof (uint64_t) * 4, SA_UINT64_ARRAY, 14},
{"ZPL_ZNODE_ACL", 88, SA_UINT8_ARRAY, 15},
{"ZPL_DACL_COUNT", sizeof (uint64_t), SA_UINT64_ARRAY, 0},
{"ZPL_SYMLINK", 0, SA_UINT8_ARRAY, 0},
{"ZPL_SCANSTAMP", 32, SA_UINT8_ARRAY, 0},
{"ZPL_DACL_ACES", 0, SA_ACL, 0},
{NULL, 0, 0, 0}
};
#ifdef _KERNEL
int
zfs_sa_readlink(znode_t *zp, uio_t *uio)
{
dmu_buf_t *db = sa_get_db(zp->z_sa_hdl);
size_t bufsz;
int error;
bufsz = zp->z_size;
if (bufsz + ZFS_OLD_ZNODE_PHYS_SIZE <= db->db_size) {
error = uiomove((caddr_t)db->db_data +
ZFS_OLD_ZNODE_PHYS_SIZE,
MIN((size_t)bufsz, uio->uio_resid), UIO_READ, uio);
} else {
dmu_buf_t *dbp;
if ((error = dmu_buf_hold(ZTOZSB(zp)->z_os, zp->z_id,
0, FTAG, &dbp, DMU_READ_NO_PREFETCH)) == 0) {
error = uiomove(dbp->db_data,
MIN((size_t)bufsz, uio->uio_resid), UIO_READ, uio);
dmu_buf_rele(dbp, FTAG);
}
}
return (error);
}
void
zfs_sa_symlink(znode_t *zp, char *link, int len, dmu_tx_t *tx)
{
dmu_buf_t *db = sa_get_db(zp->z_sa_hdl);
if (ZFS_OLD_ZNODE_PHYS_SIZE + len <= dmu_bonus_max()) {
VERIFY(dmu_set_bonus(db,
len + ZFS_OLD_ZNODE_PHYS_SIZE, tx) == 0);
if (len) {
bcopy(link, (caddr_t)db->db_data +
ZFS_OLD_ZNODE_PHYS_SIZE, len);
}
} else {
dmu_buf_t *dbp;
zfs_grow_blocksize(zp, len, tx);
VERIFY(0 == dmu_buf_hold(ZTOZSB(zp)->z_os,
zp->z_id, 0, FTAG, &dbp, DMU_READ_NO_PREFETCH));
dmu_buf_will_dirty(dbp, tx);
ASSERT3U(len, <=, dbp->db_size);
bcopy(link, dbp->db_data, len);
dmu_buf_rele(dbp, FTAG);
}
}
void
zfs_sa_get_scanstamp(znode_t *zp, xvattr_t *xvap)
{
zfs_sb_t *zsb = ZTOZSB(zp);
xoptattr_t *xoap;
ASSERT(MUTEX_HELD(&zp->z_lock));
VERIFY((xoap = xva_getxoptattr(xvap)) != NULL);
if (zp->z_is_sa) {
if (sa_lookup(zp->z_sa_hdl, SA_ZPL_SCANSTAMP(zsb),
&xoap->xoa_av_scanstamp,
sizeof (xoap->xoa_av_scanstamp)) != 0)
return;
} else {
dmu_object_info_t doi;
dmu_buf_t *db = sa_get_db(zp->z_sa_hdl);
int len;
if (!(zp->z_pflags & ZFS_BONUS_SCANSTAMP))
return;
sa_object_info(zp->z_sa_hdl, &doi);
len = sizeof (xoap->xoa_av_scanstamp) +
ZFS_OLD_ZNODE_PHYS_SIZE;
if (len <= doi.doi_bonus_size) {
(void) memcpy(xoap->xoa_av_scanstamp,
(caddr_t)db->db_data + ZFS_OLD_ZNODE_PHYS_SIZE,
sizeof (xoap->xoa_av_scanstamp));
}
}
XVA_SET_RTN(xvap, XAT_AV_SCANSTAMP);
}
void
zfs_sa_set_scanstamp(znode_t *zp, xvattr_t *xvap, dmu_tx_t *tx)
{
zfs_sb_t *zsb = ZTOZSB(zp);
xoptattr_t *xoap;
ASSERT(MUTEX_HELD(&zp->z_lock));
VERIFY((xoap = xva_getxoptattr(xvap)) != NULL);
if (zp->z_is_sa)
VERIFY(0 == sa_update(zp->z_sa_hdl, SA_ZPL_SCANSTAMP(zsb),
&xoap->xoa_av_scanstamp,
sizeof (xoap->xoa_av_scanstamp), tx));
else {
dmu_object_info_t doi;
dmu_buf_t *db = sa_get_db(zp->z_sa_hdl);
int len;
sa_object_info(zp->z_sa_hdl, &doi);
len = sizeof (xoap->xoa_av_scanstamp) +
ZFS_OLD_ZNODE_PHYS_SIZE;
if (len > doi.doi_bonus_size)
VERIFY(dmu_set_bonus(db, len, tx) == 0);
(void) memcpy((caddr_t)db->db_data + ZFS_OLD_ZNODE_PHYS_SIZE,
xoap->xoa_av_scanstamp, sizeof (xoap->xoa_av_scanstamp));
zp->z_pflags |= ZFS_BONUS_SCANSTAMP;
VERIFY(0 == sa_update(zp->z_sa_hdl, SA_ZPL_FLAGS(zsb),
&zp->z_pflags, sizeof (uint64_t), tx));
}
}
/*
* I'm not convinced we should do any of this upgrade.
* since the SA code can read both old/new znode formats
* with probably little to know performance difference.
*
* All new files will be created with the new format.
*/
void
zfs_sa_upgrade(sa_handle_t *hdl, dmu_tx_t *tx)
{
dmu_buf_t *db = sa_get_db(hdl);
znode_t *zp = sa_get_userdata(hdl);
zfs_sb_t *zsb = ZTOZSB(zp);
int count = 0;
sa_bulk_attr_t *bulk, *sa_attrs;
zfs_acl_locator_cb_t locate = { 0 };
uint64_t uid, gid, mode, rdev, xattr, parent;
uint64_t crtime[2], mtime[2], ctime[2];
zfs_acl_phys_t znode_acl;
char scanstamp[AV_SCANSTAMP_SZ];
boolean_t drop_lock = B_FALSE;
/*
* No upgrade if ACL isn't cached
* since we won't know which locks are held
* and ready the ACL would require special "locked"
* interfaces that would be messy
*/
if (zp->z_acl_cached == NULL || S_ISLNK(ZTOI(zp)->i_mode))
return;
/*
* If the z_lock is held and we aren't the owner
* the just return since we don't want to deadlock
* trying to update the status of z_is_sa. This
* file can then be upgraded at a later time.
*
* Otherwise, we know we are doing the
* sa_update() that caused us to enter this function.
*/
if (mutex_owner(&zp->z_lock) != curthread) {
if (mutex_tryenter(&zp->z_lock) == 0)
return;
else
drop_lock = B_TRUE;
}
/* First do a bulk query of the attributes that aren't cached */
bulk = kmem_alloc(sizeof(sa_bulk_attr_t) * 20, KM_SLEEP);
SA_ADD_BULK_ATTR(bulk, count, SA_ZPL_MTIME(zsb), NULL, &mtime, 16);
SA_ADD_BULK_ATTR(bulk, count, SA_ZPL_CTIME(zsb), NULL, &ctime, 16);
SA_ADD_BULK_ATTR(bulk, count, SA_ZPL_CRTIME(zsb), NULL, &crtime, 16);
SA_ADD_BULK_ATTR(bulk, count, SA_ZPL_MODE(zsb), NULL, &mode, 8);
SA_ADD_BULK_ATTR(bulk, count, SA_ZPL_PARENT(zsb), NULL, &parent, 8);
SA_ADD_BULK_ATTR(bulk, count, SA_ZPL_XATTR(zsb), NULL, &xattr, 8);
SA_ADD_BULK_ATTR(bulk, count, SA_ZPL_RDEV(zsb), NULL, &rdev, 8);
SA_ADD_BULK_ATTR(bulk, count, SA_ZPL_UID(zsb), NULL, &uid, 8);
SA_ADD_BULK_ATTR(bulk, count, SA_ZPL_GID(zsb), NULL, &gid, 8);
SA_ADD_BULK_ATTR(bulk, count, SA_ZPL_ZNODE_ACL(zsb), NULL,
&znode_acl, 88);
if (sa_bulk_lookup_locked(hdl, bulk, count) != 0) {
kmem_free(bulk, sizeof(sa_bulk_attr_t) * 20);
goto done;
}
/*
* While the order here doesn't matter its best to try and organize
* it is such a way to pick up an already existing layout number
*/
count = 0;
sa_attrs = kmem_zalloc(sizeof(sa_bulk_attr_t) * 20, KM_SLEEP);
SA_ADD_BULK_ATTR(sa_attrs, count, SA_ZPL_MODE(zsb), NULL, &mode, 8);
SA_ADD_BULK_ATTR(sa_attrs, count, SA_ZPL_SIZE(zsb), NULL,
&zp->z_size, 8);
SA_ADD_BULK_ATTR(sa_attrs, count, SA_ZPL_GEN(zsb),
NULL, &zp->z_gen, 8);
SA_ADD_BULK_ATTR(sa_attrs, count, SA_ZPL_UID(zsb), NULL, &uid, 8);
SA_ADD_BULK_ATTR(sa_attrs, count, SA_ZPL_GID(zsb), NULL, &gid, 8);
SA_ADD_BULK_ATTR(sa_attrs, count, SA_ZPL_PARENT(zsb),
NULL, &parent, 8);
SA_ADD_BULK_ATTR(sa_attrs, count, SA_ZPL_FLAGS(zsb), NULL,
&zp->z_pflags, 8);
SA_ADD_BULK_ATTR(sa_attrs, count, SA_ZPL_ATIME(zsb), NULL,
zp->z_atime, 16);
SA_ADD_BULK_ATTR(sa_attrs, count, SA_ZPL_MTIME(zsb), NULL,
&mtime, 16);
SA_ADD_BULK_ATTR(sa_attrs, count, SA_ZPL_CTIME(zsb), NULL,
&ctime, 16);
SA_ADD_BULK_ATTR(sa_attrs, count, SA_ZPL_CRTIME(zsb), NULL,
&crtime, 16);
SA_ADD_BULK_ATTR(sa_attrs, count, SA_ZPL_LINKS(zsb), NULL,
&zp->z_links, 8);
if (S_ISBLK(ZTOI(zp)->i_mode) || S_ISCHR(ZTOI(zp)->i_mode))
SA_ADD_BULK_ATTR(sa_attrs, count, SA_ZPL_RDEV(zsb), NULL,
&rdev, 8);
SA_ADD_BULK_ATTR(sa_attrs, count, SA_ZPL_DACL_COUNT(zsb), NULL,
&zp->z_acl_cached->z_acl_count, 8);
if (zp->z_acl_cached->z_version < ZFS_ACL_VERSION_FUID)
zfs_acl_xform(zp, zp->z_acl_cached, CRED());
locate.cb_aclp = zp->z_acl_cached;
SA_ADD_BULK_ATTR(sa_attrs, count, SA_ZPL_DACL_ACES(zsb),
zfs_acl_data_locator, &locate, zp->z_acl_cached->z_acl_bytes);
if (xattr)
SA_ADD_BULK_ATTR(sa_attrs, count, SA_ZPL_XATTR(zsb),
NULL, &xattr, 8);
/* if scanstamp then add scanstamp */
if (zp->z_pflags & ZFS_BONUS_SCANSTAMP) {
bcopy((caddr_t)db->db_data + ZFS_OLD_ZNODE_PHYS_SIZE,
scanstamp, AV_SCANSTAMP_SZ);
SA_ADD_BULK_ATTR(sa_attrs, count, SA_ZPL_SCANSTAMP(zsb),
NULL, scanstamp, AV_SCANSTAMP_SZ);
zp->z_pflags &= ~ZFS_BONUS_SCANSTAMP;
}
VERIFY(dmu_set_bonustype(db, DMU_OT_SA, tx) == 0);
VERIFY(sa_replace_all_by_template_locked(hdl, sa_attrs,
count, tx) == 0);
if (znode_acl.z_acl_extern_obj)
VERIFY(0 == dmu_object_free(zsb->z_os,
znode_acl.z_acl_extern_obj, tx));
zp->z_is_sa = B_TRUE;
kmem_free(sa_attrs, sizeof(sa_bulk_attr_t) * 20);
kmem_free(bulk, sizeof(sa_bulk_attr_t) * 20);
done:
if (drop_lock)
mutex_exit(&zp->z_lock);
}
void
zfs_sa_upgrade_txholds(dmu_tx_t *tx, znode_t *zp)
{
if (!ZTOZSB(zp)->z_use_sa || zp->z_is_sa)
return;
dmu_tx_hold_sa(tx, zp->z_sa_hdl, B_TRUE);
if (zfs_external_acl(zp)) {
dmu_tx_hold_free(tx, zfs_external_acl(zp), 0,
DMU_OBJECT_END);
}
}
#endif
Reference in New Issue View Git Blame Copy Permalink