freebsd-nq/module/zfs/zfs_fuid.c

795 lines
19 KiB
C
Raw Normal View History

2008-11-20 20:01:55 +00:00
/*
* 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) 2007, 2010, Oracle and/or its affiliates. All rights reserved.
2008-11-20 20:01:55 +00:00
*/
#include <sys/zfs_context.h>
#include <sys/dmu.h>
#include <sys/avl.h>
#include <sys/zap.h>
#include <sys/nvpair.h>
#ifdef _KERNEL
#include <sys/sid.h>
#include <sys/zfs_vfsops.h>
#include <sys/zfs_znode.h>
#endif
#include <sys/zfs_fuid.h>
/*
* FUID Domain table(s).
*
* The FUID table is stored as a packed nvlist of an array
* of nvlists which contain an index, domain string and offset
*
* During file system initialization the nvlist(s) are read and
* two AVL trees are created. One tree is keyed by the index number
* and the other by the domain string. Nodes are never removed from
2009-07-02 22:44:48 +00:00
* trees, but new entries may be added. If a new entry is added then
* the zfsvfs->z_fuid_dirty flag is set to true and the caller will then
2009-07-02 22:44:48 +00:00
* be responsible for calling zfs_fuid_sync() to sync the changes to disk.
*
2008-11-20 20:01:55 +00:00
*/
#define FUID_IDX "fuid_idx"
#define FUID_DOMAIN "fuid_domain"
#define FUID_OFFSET "fuid_offset"
#define FUID_NVP_ARRAY "fuid_nvlist"
typedef struct fuid_domain {
avl_node_t f_domnode;
avl_node_t f_idxnode;
ksiddomain_t *f_ksid;
uint64_t f_idx;
} fuid_domain_t;
static char *nulldomain = "";
2008-11-20 20:01:55 +00:00
/*
* Compare two indexes.
*/
static int
idx_compare(const void *arg1, const void *arg2)
{
Performance optimization of AVL tree comparator functions perf: 2.75x faster ddt_entry_compare() First 256bits of ddt_key_t is a block checksum, which are expected to be close to random data. Hence, on average, comparison only needs to look at first few bytes of the keys. To reduce number of conditional jump instructions, the result is computed as: sign(memcmp(k1, k2)). Sign of an integer 'a' can be obtained as: `(0 < a) - (a < 0)` := {-1, 0, 1} , which is computed efficiently. Synthetic performance evaluation of original and new algorithm over 1G random keys on 2.6GHz Intel(R) Xeon(R) CPU E5-2660 v3: old 6.85789 s new 2.49089 s perf: 2.8x faster vdev_queue_offset_compare() and vdev_queue_timestamp_compare() Compute the result directly instead of using conditionals perf: zfs_range_compare() Speedup between 1.1x - 2.5x, depending on compiler version and optimization level. perf: spa_error_entry_compare() `bcmp()` is not suitable for comparator use. Use `memcmp()` instead. perf: 2.8x faster metaslab_compare() and metaslab_rangesize_compare() perf: 2.8x faster zil_bp_compare() perf: 2.8x faster mze_compare() perf: faster dbuf_compare() perf: faster compares in spa_misc perf: 2.8x faster layout_hash_compare() perf: 2.8x faster space_reftree_compare() perf: libzfs: faster avl tree comparators perf: guid_compare() perf: dsl_deadlist_compare() perf: perm_set_compare() perf: 2x faster range_tree_seg_compare() perf: faster unique_compare() perf: faster vdev_cache _compare() perf: faster vdev_uberblock_compare() perf: faster fuid _compare() perf: faster zfs_znode_hold_compare() Signed-off-by: Gvozden Neskovic <neskovic@gmail.com> Signed-off-by: Richard Elling <richard.elling@gmail.com> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #5033
2016-08-27 18:12:53 +00:00
const fuid_domain_t *node1 = (const fuid_domain_t *)arg1;
const fuid_domain_t *node2 = (const fuid_domain_t *)arg2;
2008-11-20 20:01:55 +00:00
Reduce loaded range tree memory usage This patch implements a new tree structure for ZFS, and uses it to store range trees more efficiently. The new structure is approximately a B-tree, though there are some small differences from the usual characterizations. The tree has core nodes and leaf nodes; each contain data elements, which the elements in the core nodes acting as separators between its children. The difference between core and leaf nodes is that the core nodes have an array of children, while leaf nodes don't. Every node in the tree may be only partially full; in most cases, they are all at least 50% full (in terms of element count) except for the root node, which can be less full. Underfull nodes will steal from their neighbors or merge to remain full enough, while overfull nodes will split in two. The data elements are contained in tree-controlled buffers; they are copied into these on insertion, and overwritten on deletion. This means that the elements are not independently allocated, which reduces overhead, but also means they can't be shared between trees (and also that pointers to them are only valid until a side-effectful tree operation occurs). The overhead varies based on how dense the tree is, but is usually on the order of about 50% of the element size; the per-node overheads are very small, and so don't make a significant difference. The trees can accept arbitrary records; they accept a size and a comparator to allow them to be used for a variety of purposes. The new trees replace the AVL trees used in the range trees today. Currently, the range_seg_t structure contains three 8 byte integers of payload and two 24 byte avl_tree_node_ts to handle its storage in both an offset-sorted tree and a size-sorted tree (total size: 64 bytes). In the new model, the range seg structures are usually two 4 byte integers, but a separate one needs to exist for the size-sorted and offset-sorted tree. Between the raw size, the 50% overhead, and the double storage, the new btrees are expected to use 8*1.5*2 = 24 bytes per record, or 33.3% as much memory as the AVL trees (this is for the purposes of storing metaslab range trees; for other purposes, like scrubs, they use ~50% as much memory). We reduced the size of the payload in the range segments by teaching range trees about starting offsets and shifts; since metaslabs have a fixed starting offset, and they all operate in terms of disk sectors, we can store the ranges using 4-byte integers as long as the size of the metaslab divided by the sector size is less than 2^32. For 512-byte sectors, this is a 2^41 (or 2TB) metaslab, which with the default settings corresponds to a 256PB disk. 4k sector disks can handle metaslabs up to 2^46 bytes, or 2^63 byte disks. Since we do not anticipate disks of this size in the near future, there should be almost no cases where metaslabs need 64-byte integers to store their ranges. We do still have the capability to store 64-byte integer ranges to account for cases where we are storing per-vdev (or per-dnode) trees, which could reasonably go above the limits discussed. We also do not store fill information in the compact version of the node, since it is only used for sorted scrub. We also optimized the metaslab loading process in various other ways to offset some inefficiencies in the btree model. While individual operations (find, insert, remove_from) are faster for the btree than they are for the avl tree, remove usually requires a find operation, while in the AVL tree model the element itself suffices. Some clever changes actually caused an overall speedup in metaslab loading; we use approximately 40% less cpu to load metaslabs in our tests on Illumos. Another memory and performance optimization was achieved by changing what is stored in the size-sorted trees. When a disk is heavily fragmented, the df algorithm used by default in ZFS will almost always find a number of small regions in its initial cursor-based search; it will usually only fall back to the size-sorted tree to find larger regions. If we increase the size of the cursor-based search slightly, and don't store segments that are smaller than a tunable size floor in the size-sorted tree, we can further cut memory usage down to below 20% of what the AVL trees store. This also results in further reductions in CPU time spent loading metaslabs. The 16KiB size floor was chosen because it results in substantial memory usage reduction while not usually resulting in situations where we can't find an appropriate chunk with the cursor and are forced to use an oversized chunk from the size-sorted tree. In addition, even if we do have to use an oversized chunk from the size-sorted tree, the chunk would be too small to use for ZIL allocations, so it isn't as big of a loss as it might otherwise be. And often, more small allocations will follow the initial one, and the cursor search will now find the remainder of the chunk we didn't use all of and use it for subsequent allocations. Practical testing has shown little or no change in fragmentation as a result of this change. If the size-sorted tree becomes empty while the offset sorted one still has entries, it will load all the entries from the offset sorted tree and disregard the size floor until it is unloaded again. This operation occurs rarely with the default setting, only on incredibly thoroughly fragmented pools. There are some other small changes to zdb to teach it to handle btrees, but nothing major. Reviewed-by: George Wilson <gwilson@delphix.com> Reviewed-by: Matt Ahrens <matt@delphix.com> Reviewed by: Sebastien Roy seb@delphix.com Reviewed-by: Igor Kozhukhov <igor@dilos.org> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Paul Dagnelie <pcd@delphix.com> Closes #9181
2019-10-09 17:36:03 +00:00
return (TREE_CMP(node1->f_idx, node2->f_idx));
2008-11-20 20:01:55 +00:00
}
/*
* Compare two domain strings.
*/
static int
domain_compare(const void *arg1, const void *arg2)
{
Performance optimization of AVL tree comparator functions perf: 2.75x faster ddt_entry_compare() First 256bits of ddt_key_t is a block checksum, which are expected to be close to random data. Hence, on average, comparison only needs to look at first few bytes of the keys. To reduce number of conditional jump instructions, the result is computed as: sign(memcmp(k1, k2)). Sign of an integer 'a' can be obtained as: `(0 < a) - (a < 0)` := {-1, 0, 1} , which is computed efficiently. Synthetic performance evaluation of original and new algorithm over 1G random keys on 2.6GHz Intel(R) Xeon(R) CPU E5-2660 v3: old 6.85789 s new 2.49089 s perf: 2.8x faster vdev_queue_offset_compare() and vdev_queue_timestamp_compare() Compute the result directly instead of using conditionals perf: zfs_range_compare() Speedup between 1.1x - 2.5x, depending on compiler version and optimization level. perf: spa_error_entry_compare() `bcmp()` is not suitable for comparator use. Use `memcmp()` instead. perf: 2.8x faster metaslab_compare() and metaslab_rangesize_compare() perf: 2.8x faster zil_bp_compare() perf: 2.8x faster mze_compare() perf: faster dbuf_compare() perf: faster compares in spa_misc perf: 2.8x faster layout_hash_compare() perf: 2.8x faster space_reftree_compare() perf: libzfs: faster avl tree comparators perf: guid_compare() perf: dsl_deadlist_compare() perf: perm_set_compare() perf: 2x faster range_tree_seg_compare() perf: faster unique_compare() perf: faster vdev_cache _compare() perf: faster vdev_uberblock_compare() perf: faster fuid _compare() perf: faster zfs_znode_hold_compare() Signed-off-by: Gvozden Neskovic <neskovic@gmail.com> Signed-off-by: Richard Elling <richard.elling@gmail.com> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #5033
2016-08-27 18:12:53 +00:00
const fuid_domain_t *node1 = (const fuid_domain_t *)arg1;
const fuid_domain_t *node2 = (const fuid_domain_t *)arg2;
2008-11-20 20:01:55 +00:00
int val;
val = strcmp(node1->f_ksid->kd_name, node2->f_ksid->kd_name);
Performance optimization of AVL tree comparator functions perf: 2.75x faster ddt_entry_compare() First 256bits of ddt_key_t is a block checksum, which are expected to be close to random data. Hence, on average, comparison only needs to look at first few bytes of the keys. To reduce number of conditional jump instructions, the result is computed as: sign(memcmp(k1, k2)). Sign of an integer 'a' can be obtained as: `(0 < a) - (a < 0)` := {-1, 0, 1} , which is computed efficiently. Synthetic performance evaluation of original and new algorithm over 1G random keys on 2.6GHz Intel(R) Xeon(R) CPU E5-2660 v3: old 6.85789 s new 2.49089 s perf: 2.8x faster vdev_queue_offset_compare() and vdev_queue_timestamp_compare() Compute the result directly instead of using conditionals perf: zfs_range_compare() Speedup between 1.1x - 2.5x, depending on compiler version and optimization level. perf: spa_error_entry_compare() `bcmp()` is not suitable for comparator use. Use `memcmp()` instead. perf: 2.8x faster metaslab_compare() and metaslab_rangesize_compare() perf: 2.8x faster zil_bp_compare() perf: 2.8x faster mze_compare() perf: faster dbuf_compare() perf: faster compares in spa_misc perf: 2.8x faster layout_hash_compare() perf: 2.8x faster space_reftree_compare() perf: libzfs: faster avl tree comparators perf: guid_compare() perf: dsl_deadlist_compare() perf: perm_set_compare() perf: 2x faster range_tree_seg_compare() perf: faster unique_compare() perf: faster vdev_cache _compare() perf: faster vdev_uberblock_compare() perf: faster fuid _compare() perf: faster zfs_znode_hold_compare() Signed-off-by: Gvozden Neskovic <neskovic@gmail.com> Signed-off-by: Richard Elling <richard.elling@gmail.com> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #5033
2016-08-27 18:12:53 +00:00
Reduce loaded range tree memory usage This patch implements a new tree structure for ZFS, and uses it to store range trees more efficiently. The new structure is approximately a B-tree, though there are some small differences from the usual characterizations. The tree has core nodes and leaf nodes; each contain data elements, which the elements in the core nodes acting as separators between its children. The difference between core and leaf nodes is that the core nodes have an array of children, while leaf nodes don't. Every node in the tree may be only partially full; in most cases, they are all at least 50% full (in terms of element count) except for the root node, which can be less full. Underfull nodes will steal from their neighbors or merge to remain full enough, while overfull nodes will split in two. The data elements are contained in tree-controlled buffers; they are copied into these on insertion, and overwritten on deletion. This means that the elements are not independently allocated, which reduces overhead, but also means they can't be shared between trees (and also that pointers to them are only valid until a side-effectful tree operation occurs). The overhead varies based on how dense the tree is, but is usually on the order of about 50% of the element size; the per-node overheads are very small, and so don't make a significant difference. The trees can accept arbitrary records; they accept a size and a comparator to allow them to be used for a variety of purposes. The new trees replace the AVL trees used in the range trees today. Currently, the range_seg_t structure contains three 8 byte integers of payload and two 24 byte avl_tree_node_ts to handle its storage in both an offset-sorted tree and a size-sorted tree (total size: 64 bytes). In the new model, the range seg structures are usually two 4 byte integers, but a separate one needs to exist for the size-sorted and offset-sorted tree. Between the raw size, the 50% overhead, and the double storage, the new btrees are expected to use 8*1.5*2 = 24 bytes per record, or 33.3% as much memory as the AVL trees (this is for the purposes of storing metaslab range trees; for other purposes, like scrubs, they use ~50% as much memory). We reduced the size of the payload in the range segments by teaching range trees about starting offsets and shifts; since metaslabs have a fixed starting offset, and they all operate in terms of disk sectors, we can store the ranges using 4-byte integers as long as the size of the metaslab divided by the sector size is less than 2^32. For 512-byte sectors, this is a 2^41 (or 2TB) metaslab, which with the default settings corresponds to a 256PB disk. 4k sector disks can handle metaslabs up to 2^46 bytes, or 2^63 byte disks. Since we do not anticipate disks of this size in the near future, there should be almost no cases where metaslabs need 64-byte integers to store their ranges. We do still have the capability to store 64-byte integer ranges to account for cases where we are storing per-vdev (or per-dnode) trees, which could reasonably go above the limits discussed. We also do not store fill information in the compact version of the node, since it is only used for sorted scrub. We also optimized the metaslab loading process in various other ways to offset some inefficiencies in the btree model. While individual operations (find, insert, remove_from) are faster for the btree than they are for the avl tree, remove usually requires a find operation, while in the AVL tree model the element itself suffices. Some clever changes actually caused an overall speedup in metaslab loading; we use approximately 40% less cpu to load metaslabs in our tests on Illumos. Another memory and performance optimization was achieved by changing what is stored in the size-sorted trees. When a disk is heavily fragmented, the df algorithm used by default in ZFS will almost always find a number of small regions in its initial cursor-based search; it will usually only fall back to the size-sorted tree to find larger regions. If we increase the size of the cursor-based search slightly, and don't store segments that are smaller than a tunable size floor in the size-sorted tree, we can further cut memory usage down to below 20% of what the AVL trees store. This also results in further reductions in CPU time spent loading metaslabs. The 16KiB size floor was chosen because it results in substantial memory usage reduction while not usually resulting in situations where we can't find an appropriate chunk with the cursor and are forced to use an oversized chunk from the size-sorted tree. In addition, even if we do have to use an oversized chunk from the size-sorted tree, the chunk would be too small to use for ZIL allocations, so it isn't as big of a loss as it might otherwise be. And often, more small allocations will follow the initial one, and the cursor search will now find the remainder of the chunk we didn't use all of and use it for subsequent allocations. Practical testing has shown little or no change in fragmentation as a result of this change. If the size-sorted tree becomes empty while the offset sorted one still has entries, it will load all the entries from the offset sorted tree and disregard the size floor until it is unloaded again. This operation occurs rarely with the default setting, only on incredibly thoroughly fragmented pools. There are some other small changes to zdb to teach it to handle btrees, but nothing major. Reviewed-by: George Wilson <gwilson@delphix.com> Reviewed-by: Matt Ahrens <matt@delphix.com> Reviewed by: Sebastien Roy seb@delphix.com Reviewed-by: Igor Kozhukhov <igor@dilos.org> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Paul Dagnelie <pcd@delphix.com> Closes #9181
2019-10-09 17:36:03 +00:00
return (TREE_ISIGN(val));
2008-11-20 20:01:55 +00:00
}
2009-07-02 22:44:48 +00:00
void
zfs_fuid_avl_tree_create(avl_tree_t *idx_tree, avl_tree_t *domain_tree)
{
avl_create(idx_tree, idx_compare,
sizeof (fuid_domain_t), offsetof(fuid_domain_t, f_idxnode));
avl_create(domain_tree, domain_compare,
sizeof (fuid_domain_t), offsetof(fuid_domain_t, f_domnode));
}
2008-11-20 20:01:55 +00:00
/*
* load initial fuid domain and idx trees. This function is used by
* both the kernel and zdb.
*/
uint64_t
zfs_fuid_table_load(objset_t *os, uint64_t fuid_obj, avl_tree_t *idx_tree,
avl_tree_t *domain_tree)
{
dmu_buf_t *db;
uint64_t fuid_size;
2009-07-02 22:44:48 +00:00
ASSERT(fuid_obj != 0);
VERIFY(0 == dmu_bonus_hold(os, fuid_obj,
FTAG, &db));
2008-11-20 20:01:55 +00:00
fuid_size = *(uint64_t *)db->db_data;
dmu_buf_rele(db, FTAG);
if (fuid_size) {
nvlist_t **fuidnvp;
nvlist_t *nvp = NULL;
uint_t count;
char *packed;
int i;
packed = kmem_alloc(fuid_size, KM_SLEEP);
2009-07-02 22:44:48 +00:00
VERIFY(dmu_read(os, fuid_obj, 0,
fuid_size, packed, DMU_READ_PREFETCH) == 0);
2008-11-20 20:01:55 +00:00
VERIFY(nvlist_unpack(packed, fuid_size,
&nvp, 0) == 0);
VERIFY(nvlist_lookup_nvlist_array(nvp, FUID_NVP_ARRAY,
&fuidnvp, &count) == 0);
for (i = 0; i != count; i++) {
fuid_domain_t *domnode;
char *domain;
uint64_t idx;
VERIFY(nvlist_lookup_string(fuidnvp[i], FUID_DOMAIN,
&domain) == 0);
VERIFY(nvlist_lookup_uint64(fuidnvp[i], FUID_IDX,
&idx) == 0);
domnode = kmem_alloc(sizeof (fuid_domain_t), KM_SLEEP);
domnode->f_idx = idx;
domnode->f_ksid = ksid_lookupdomain(domain);
avl_add(idx_tree, domnode);
avl_add(domain_tree, domnode);
}
nvlist_free(nvp);
kmem_free(packed, fuid_size);
}
return (fuid_size);
}
void
zfs_fuid_table_destroy(avl_tree_t *idx_tree, avl_tree_t *domain_tree)
{
fuid_domain_t *domnode;
void *cookie;
cookie = NULL;
while ((domnode = avl_destroy_nodes(domain_tree, &cookie)))
2008-11-20 20:01:55 +00:00
ksiddomain_rele(domnode->f_ksid);
avl_destroy(domain_tree);
cookie = NULL;
while ((domnode = avl_destroy_nodes(idx_tree, &cookie)))
2008-11-20 20:01:55 +00:00
kmem_free(domnode, sizeof (fuid_domain_t));
avl_destroy(idx_tree);
}
char *
zfs_fuid_idx_domain(avl_tree_t *idx_tree, uint32_t idx)
{
fuid_domain_t searchnode, *findnode;
avl_index_t loc;
searchnode.f_idx = idx;
findnode = avl_find(idx_tree, &searchnode, &loc);
return (findnode ? findnode->f_ksid->kd_name : nulldomain);
2008-11-20 20:01:55 +00:00
}
#ifdef _KERNEL
/*
* Load the fuid table(s) into memory.
*/
static void
zfs_fuid_init(zfsvfs_t *zfsvfs)
2008-11-20 20:01:55 +00:00
{
rw_enter(&zfsvfs->z_fuid_lock, RW_WRITER);
2008-11-20 20:01:55 +00:00
if (zfsvfs->z_fuid_loaded) {
rw_exit(&zfsvfs->z_fuid_lock);
2008-11-20 20:01:55 +00:00
return;
}
zfs_fuid_avl_tree_create(&zfsvfs->z_fuid_idx, &zfsvfs->z_fuid_domain);
2008-11-20 20:01:55 +00:00
(void) zap_lookup(zfsvfs->z_os, MASTER_NODE_OBJ,
ZFS_FUID_TABLES, 8, 1, &zfsvfs->z_fuid_obj);
if (zfsvfs->z_fuid_obj != 0) {
zfsvfs->z_fuid_size = zfs_fuid_table_load(zfsvfs->z_os,
zfsvfs->z_fuid_obj, &zfsvfs->z_fuid_idx,
&zfsvfs->z_fuid_domain);
}
2008-11-20 20:01:55 +00:00
zfsvfs->z_fuid_loaded = B_TRUE;
rw_exit(&zfsvfs->z_fuid_lock);
2009-07-02 22:44:48 +00:00
}
/*
* sync out AVL trees to persistent storage.
*/
void
zfs_fuid_sync(zfsvfs_t *zfsvfs, dmu_tx_t *tx)
2009-07-02 22:44:48 +00:00
{
nvlist_t *nvp;
nvlist_t **fuids;
size_t nvsize = 0;
char *packed;
dmu_buf_t *db;
fuid_domain_t *domnode;
int numnodes;
int i;
if (!zfsvfs->z_fuid_dirty) {
2009-07-02 22:44:48 +00:00
return;
}
rw_enter(&zfsvfs->z_fuid_lock, RW_WRITER);
2009-07-02 22:44:48 +00:00
/*
* First see if table needs to be created?
*/
if (zfsvfs->z_fuid_obj == 0) {
zfsvfs->z_fuid_obj = dmu_object_alloc(zfsvfs->z_os,
2009-07-02 22:44:48 +00:00
DMU_OT_FUID, 1 << 14, DMU_OT_FUID_SIZE,
sizeof (uint64_t), tx);
VERIFY(zap_add(zfsvfs->z_os, MASTER_NODE_OBJ,
2009-07-02 22:44:48 +00:00
ZFS_FUID_TABLES, sizeof (uint64_t), 1,
&zfsvfs->z_fuid_obj, tx) == 0);
2009-07-02 22:44:48 +00:00
}
VERIFY(nvlist_alloc(&nvp, NV_UNIQUE_NAME, KM_SLEEP) == 0);
numnodes = avl_numnodes(&zfsvfs->z_fuid_idx);
2009-07-02 22:44:48 +00:00
fuids = kmem_alloc(numnodes * sizeof (void *), KM_SLEEP);
for (i = 0, domnode = avl_first(&zfsvfs->z_fuid_domain); domnode; i++,
domnode = AVL_NEXT(&zfsvfs->z_fuid_domain, domnode)) {
2009-07-02 22:44:48 +00:00
VERIFY(nvlist_alloc(&fuids[i], NV_UNIQUE_NAME, KM_SLEEP) == 0);
VERIFY(nvlist_add_uint64(fuids[i], FUID_IDX,
domnode->f_idx) == 0);
VERIFY(nvlist_add_uint64(fuids[i], FUID_OFFSET, 0) == 0);
VERIFY(nvlist_add_string(fuids[i], FUID_DOMAIN,
domnode->f_ksid->kd_name) == 0);
}
VERIFY(nvlist_add_nvlist_array(nvp, FUID_NVP_ARRAY,
fuids, numnodes) == 0);
for (i = 0; i != numnodes; i++)
nvlist_free(fuids[i]);
kmem_free(fuids, numnodes * sizeof (void *));
VERIFY(nvlist_size(nvp, &nvsize, NV_ENCODE_XDR) == 0);
packed = kmem_alloc(nvsize, KM_SLEEP);
VERIFY(nvlist_pack(nvp, &packed, &nvsize,
NV_ENCODE_XDR, KM_SLEEP) == 0);
nvlist_free(nvp);
zfsvfs->z_fuid_size = nvsize;
dmu_write(zfsvfs->z_os, zfsvfs->z_fuid_obj, 0,
zfsvfs->z_fuid_size, packed, tx);
kmem_free(packed, zfsvfs->z_fuid_size);
VERIFY(0 == dmu_bonus_hold(zfsvfs->z_os, zfsvfs->z_fuid_obj,
2009-07-02 22:44:48 +00:00
FTAG, &db));
dmu_buf_will_dirty(db, tx);
*(uint64_t *)db->db_data = zfsvfs->z_fuid_size;
2009-07-02 22:44:48 +00:00
dmu_buf_rele(db, FTAG);
zfsvfs->z_fuid_dirty = B_FALSE;
rw_exit(&zfsvfs->z_fuid_lock);
2008-11-20 20:01:55 +00:00
}
/*
* Query domain table for a given domain.
*
2009-07-02 22:44:48 +00:00
* If domain isn't found and addok is set, it is added to AVL trees and
* the zfsvfs->z_fuid_dirty flag will be set to TRUE. It will then be
2009-07-02 22:44:48 +00:00
* necessary for the caller or another thread to detect the dirty table
* and sync out the changes.
2008-11-20 20:01:55 +00:00
*/
int
zfs_fuid_find_by_domain(zfsvfs_t *zfsvfs, const char *domain,
2009-07-02 22:44:48 +00:00
char **retdomain, boolean_t addok)
2008-11-20 20:01:55 +00:00
{
fuid_domain_t searchnode, *findnode;
avl_index_t loc;
krw_t rw = RW_READER;
2008-11-20 20:01:55 +00:00
/*
* If the dummy "nobody" domain then return an index of 0
* to cause the created FUID to be a standard POSIX id
* for the user nobody.
*/
if (domain[0] == '\0') {
2009-07-02 22:44:48 +00:00
if (retdomain)
*retdomain = nulldomain;
2008-11-20 20:01:55 +00:00
return (0);
}
searchnode.f_ksid = ksid_lookupdomain(domain);
2009-07-02 22:44:48 +00:00
if (retdomain)
2008-11-20 20:01:55 +00:00
*retdomain = searchnode.f_ksid->kd_name;
if (!zfsvfs->z_fuid_loaded)
zfs_fuid_init(zfsvfs);
2008-11-20 20:01:55 +00:00
retry:
rw_enter(&zfsvfs->z_fuid_lock, rw);
findnode = avl_find(&zfsvfs->z_fuid_domain, &searchnode, &loc);
2008-11-20 20:01:55 +00:00
if (findnode) {
rw_exit(&zfsvfs->z_fuid_lock);
2008-11-20 20:01:55 +00:00
ksiddomain_rele(searchnode.f_ksid);
return (findnode->f_idx);
2009-07-02 22:44:48 +00:00
} else if (addok) {
2008-11-20 20:01:55 +00:00
fuid_domain_t *domnode;
uint64_t retidx;
if (rw == RW_READER && !rw_tryupgrade(&zfsvfs->z_fuid_lock)) {
rw_exit(&zfsvfs->z_fuid_lock);
rw = RW_WRITER;
goto retry;
}
2008-11-20 20:01:55 +00:00
domnode = kmem_alloc(sizeof (fuid_domain_t), KM_SLEEP);
domnode->f_ksid = searchnode.f_ksid;
retidx = domnode->f_idx = avl_numnodes(&zfsvfs->z_fuid_idx) + 1;
2008-11-20 20:01:55 +00:00
avl_add(&zfsvfs->z_fuid_domain, domnode);
avl_add(&zfsvfs->z_fuid_idx, domnode);
zfsvfs->z_fuid_dirty = B_TRUE;
rw_exit(&zfsvfs->z_fuid_lock);
2008-11-20 20:01:55 +00:00
return (retidx);
2009-07-02 22:44:48 +00:00
} else {
rw_exit(&zfsvfs->z_fuid_lock);
2009-07-02 22:44:48 +00:00
return (-1);
2008-11-20 20:01:55 +00:00
}
}
/*
* Query domain table by index, returning domain string
*
* Returns a pointer from an avl node of the domain string.
*
*/
2009-07-02 22:44:48 +00:00
const char *
zfs_fuid_find_by_idx(zfsvfs_t *zfsvfs, uint32_t idx)
2008-11-20 20:01:55 +00:00
{
char *domain;
if (idx == 0 || !zfsvfs->z_use_fuids)
2008-11-20 20:01:55 +00:00
return (NULL);
if (!zfsvfs->z_fuid_loaded)
zfs_fuid_init(zfsvfs);
2008-11-20 20:01:55 +00:00
rw_enter(&zfsvfs->z_fuid_lock, RW_READER);
if (zfsvfs->z_fuid_obj || zfsvfs->z_fuid_dirty)
domain = zfs_fuid_idx_domain(&zfsvfs->z_fuid_idx, idx);
else
domain = nulldomain;
rw_exit(&zfsvfs->z_fuid_lock);
2008-11-20 20:01:55 +00:00
ASSERT(domain);
return (domain);
}
void
zfs_fuid_map_ids(znode_t *zp, cred_t *cr, uid_t *uidp, uid_t *gidp)
{
*uidp = zfs_fuid_map_id(ZTOZSB(zp), KUID_TO_SUID(ZTOUID(zp)),
cr, ZFS_OWNER);
*gidp = zfs_fuid_map_id(ZTOZSB(zp), KGID_TO_SGID(ZTOGID(zp)),
cr, ZFS_GROUP);
2008-11-20 20:01:55 +00:00
}
uid_t
zfs_fuid_map_id(zfsvfs_t *zfsvfs, uint64_t fuid,
2008-11-20 20:01:55 +00:00
cred_t *cr, zfs_fuid_type_t type)
{
#ifdef HAVE_KSID
2008-11-20 20:01:55 +00:00
uint32_t index = FUID_INDEX(fuid);
2009-07-02 22:44:48 +00:00
const char *domain;
2008-11-20 20:01:55 +00:00
uid_t id;
if (index == 0)
return (fuid);
domain = zfs_fuid_find_by_idx(zfsvfs, index);
2008-11-20 20:01:55 +00:00
ASSERT(domain != NULL);
if (type == ZFS_OWNER || type == ZFS_ACE_USER) {
(void) kidmap_getuidbysid(crgetzone(cr), domain,
FUID_RID(fuid), &id);
} else {
(void) kidmap_getgidbysid(crgetzone(cr), domain,
FUID_RID(fuid), &id);
}
return (id);
#else
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
2011-03-01 20:24:09 +00:00
/*
* The Linux port only supports POSIX IDs, use the passed id.
*/
return (fuid);
#endif /* HAVE_KSID */
2008-11-20 20:01:55 +00:00
}
/*
* Add a FUID node to the list of fuid's being created for this
* ACL
*
* If ACL has multiple domains, then keep only one copy of each unique
* domain.
*/
void
2008-11-20 20:01:55 +00:00
zfs_fuid_node_add(zfs_fuid_info_t **fuidpp, const char *domain, uint32_t rid,
uint64_t idx, uint64_t id, zfs_fuid_type_t type)
{
zfs_fuid_t *fuid;
zfs_fuid_domain_t *fuid_domain;
zfs_fuid_info_t *fuidp;
uint64_t fuididx;
boolean_t found = B_FALSE;
if (*fuidpp == NULL)
*fuidpp = zfs_fuid_info_alloc();
fuidp = *fuidpp;
/*
* First find fuid domain index in linked list
*
* If one isn't found then create an entry.
*/
for (fuididx = 1, fuid_domain = list_head(&fuidp->z_domains);
fuid_domain; fuid_domain = list_next(&fuidp->z_domains,
fuid_domain), fuididx++) {
if (idx == fuid_domain->z_domidx) {
found = B_TRUE;
break;
}
}
if (!found) {
fuid_domain = kmem_alloc(sizeof (zfs_fuid_domain_t), KM_SLEEP);
fuid_domain->z_domain = domain;
fuid_domain->z_domidx = idx;
list_insert_tail(&fuidp->z_domains, fuid_domain);
fuidp->z_domain_str_sz += strlen(domain) + 1;
fuidp->z_domain_cnt++;
}
if (type == ZFS_ACE_USER || type == ZFS_ACE_GROUP) {
2009-07-02 22:44:48 +00:00
2008-11-20 20:01:55 +00:00
/*
* Now allocate fuid entry and add it on the end of the list
*/
fuid = kmem_alloc(sizeof (zfs_fuid_t), KM_SLEEP);
fuid->z_id = id;
fuid->z_domidx = idx;
fuid->z_logfuid = FUID_ENCODE(fuididx, rid);
list_insert_tail(&fuidp->z_fuids, fuid);
fuidp->z_fuid_cnt++;
} else {
if (type == ZFS_OWNER)
fuidp->z_fuid_owner = FUID_ENCODE(fuididx, rid);
else
fuidp->z_fuid_group = FUID_ENCODE(fuididx, rid);
}
}
#ifdef HAVE_KSID
2008-11-20 20:01:55 +00:00
/*
* Create a file system FUID, based on information in the users cred
*
* If cred contains KSID_OWNER then it should be used to determine
* the uid otherwise cred's uid will be used. By default cred's gid
* is used unless it's an ephemeral ID in which case KSID_GROUP will
* be used if it exists.
2008-11-20 20:01:55 +00:00
*/
uint64_t
zfs_fuid_create_cred(zfsvfs_t *zfsvfs, zfs_fuid_type_t type,
2009-07-02 22:44:48 +00:00
cred_t *cr, zfs_fuid_info_t **fuidp)
2008-11-20 20:01:55 +00:00
{
uint64_t idx;
ksid_t *ksid;
uint32_t rid;
Prototype/structure update for Linux I appologize in advance why to many things ended up in this commit. When it could be seperated in to a whole series of commits teasing that all apart now would take considerable time and I'm not sure there's much merrit in it. As such I'll just summerize the intent of the changes which are all (or partly) in this commit. Broadly the intent is to remove as much Solaris specific code as possible and replace it with native Linux equivilants. More specifically: 1) Replace all instances of zfsvfs_t with zfs_sb_t. While the type is largely the same calling it private super block data rather than a zfsvfs is more consistent with how Linux names this. While non critical it makes the code easier to read when your thinking in Linux friendly VFS terms. 2) Replace vnode_t with struct inode. The Linux VFS doesn't have the notion of a vnode and there's absolutely no good reason to create one. There are in fact several good reasons to remove it. It just adds overhead on Linux if we were to manage one, it conplicates the code, and it likely will lead to bugs so there's a good change it will be out of date. The code has been updated to remove all need for this type. 3) Replace all vtype_t's with umode types. Along with this shift all uses of types to mode bits. The Solaris code would pass a vtype which is redundant with the Linux mode. Just update all the code to use the Linux mode macros and remove this redundancy. 4) Remove using of vn_* helpers and replace where needed with inode helpers. The big example here is creating iput_aync to replace vn_rele_async. Other vn helpers will be addressed as needed but they should be be emulated. They are a Solaris VFS'ism and should simply be replaced with Linux equivilants. 5) Update znode alloc/free code. Under Linux it's common to embed the inode specific data with the inode itself. This removes the need for an extra memory allocation. In zfs this information is called a znode and it now embeds the inode with it. Allocators have been updated accordingly. 6) Minimal integration with the vfs flags for setting up the super block and handling mount options has been added this code will need to be refined but functionally it's all there. This will be the first and last of these to large to review commits.
2011-02-08 19:16:06 +00:00
char *kdomain;
2008-11-20 20:01:55 +00:00
const char *domain;
uid_t id;
VERIFY(type == ZFS_OWNER || type == ZFS_GROUP);
ksid = crgetsid(cr, (type == ZFS_OWNER) ? KSID_OWNER : KSID_GROUP);
if (!zfsvfs->z_use_fuids || (ksid == NULL)) {
id = (type == ZFS_OWNER) ? crgetuid(cr) : crgetgid(cr);
if (IS_EPHEMERAL(id))
return ((type == ZFS_OWNER) ? UID_NOBODY : GID_NOBODY);
return ((uint64_t)id);
}
2008-11-20 20:01:55 +00:00
/*
* ksid is present and FUID is supported
*/
id = (type == ZFS_OWNER) ? ksid_getid(ksid) : crgetgid(cr);
if (!IS_EPHEMERAL(id))
2008-11-20 20:01:55 +00:00
return ((uint64_t)id);
if (type == ZFS_GROUP)
id = ksid_getid(ksid);
2008-11-20 20:01:55 +00:00
rid = ksid_getrid(ksid);
domain = ksid_getdomain(ksid);
idx = zfs_fuid_find_by_domain(zfsvfs, domain, &kdomain, B_TRUE);
2008-11-20 20:01:55 +00:00
zfs_fuid_node_add(fuidp, kdomain, rid, idx, id, type);
return (FUID_ENCODE(idx, rid));
Add `zfs allow` and `zfs unallow` support ZFS allows for specific permissions to be delegated to normal users with the `zfs allow` and `zfs unallow` commands. In addition, non- privileged users should be able to run all of the following commands: * zpool [list | iostat | status | get] * zfs [list | get] Historically this functionality was not available on Linux. In order to add it the secpolicy_* functions needed to be implemented and mapped to the equivalent Linux capability. Only then could the permissions on the `/dev/zfs` be relaxed and the internal ZFS permission checks used. Even with this change some limitations remain. Under Linux only the root user is allowed to modify the namespace (unless it's a private namespace). This means the mount, mountpoint, canmount, unmount, and remount delegations cannot be supported with the existing code. It may be possible to add this functionality in the future. This functionality was validated with the cli_user and delegation test cases from the ZFS Test Suite. These tests exhaustively verify each of the supported permissions which can be delegated and ensures only an authorized user can perform it. Two minor bug fixes were required for test-running.py. First, the Timer() object cannot be safely created in a `try:` block when there is an unconditional `finally` block which references it. Second, when running as a normal user also check for scripts using the both the .ksh and .sh suffixes. Finally, existing users who are simulating delegations by setting group permissions on the /dev/zfs device should revert that customization when updating to a version with this change. Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Tony Hutter <hutter2@llnl.gov> Closes #362 Closes #434 Closes #4100 Closes #4394 Closes #4410 Closes #4487
2016-06-07 16:16:52 +00:00
}
#endif /* HAVE_KSID */
2008-11-20 20:01:55 +00:00
/*
* Create a file system FUID for an ACL ace
* or a chown/chgrp of the file.
* This is similar to zfs_fuid_create_cred, except that
* we can't find the domain + rid information in the
* cred. Instead we have to query Winchester for the
* domain and rid.
*
* During replay operations the domain+rid information is
* found in the zfs_fuid_info_t that the replay code has
* attached to the zfsvfs of the file system.
2008-11-20 20:01:55 +00:00
*/
uint64_t
zfs_fuid_create(zfsvfs_t *zfsvfs, uint64_t id, cred_t *cr,
2009-07-02 22:44:48 +00:00
zfs_fuid_type_t type, zfs_fuid_info_t **fuidpp)
2008-11-20 20:01:55 +00:00
{
#ifdef HAVE_KSID
2008-11-20 20:01:55 +00:00
const char *domain;
char *kdomain;
uint32_t fuid_idx = FUID_INDEX(id);
uint32_t rid;
idmap_stat status;
uint64_t idx = 0;
2008-11-20 20:01:55 +00:00
zfs_fuid_t *zfuid = NULL;
zfs_fuid_info_t *fuidp = NULL;
2008-11-20 20:01:55 +00:00
/*
* If POSIX ID, or entry is already a FUID then
* just return the id
*
* We may also be handed an already FUID'ized id via
* chmod.
*/
if (!zfsvfs->z_use_fuids || !IS_EPHEMERAL(id) || fuid_idx != 0)
2008-11-20 20:01:55 +00:00
return (id);
if (zfsvfs->z_replay) {
fuidp = zfsvfs->z_fuid_replay;
2008-11-20 20:01:55 +00:00
/*
* If we are passed an ephemeral id, but no
* fuid_info was logged then return NOBODY.
* This is most likely a result of idmap service
* not being available.
*/
if (fuidp == NULL)
return (UID_NOBODY);
VERIFY3U(type, >=, ZFS_OWNER);
VERIFY3U(type, <=, ZFS_ACE_GROUP);
2008-11-20 20:01:55 +00:00
switch (type) {
case ZFS_ACE_USER:
case ZFS_ACE_GROUP:
zfuid = list_head(&fuidp->z_fuids);
rid = FUID_RID(zfuid->z_logfuid);
idx = FUID_INDEX(zfuid->z_logfuid);
break;
case ZFS_OWNER:
rid = FUID_RID(fuidp->z_fuid_owner);
idx = FUID_INDEX(fuidp->z_fuid_owner);
break;
case ZFS_GROUP:
rid = FUID_RID(fuidp->z_fuid_group);
idx = FUID_INDEX(fuidp->z_fuid_group);
break;
};
domain = fuidp->z_domain_table[idx - 1];
2008-11-20 20:01:55 +00:00
} else {
if (type == ZFS_OWNER || type == ZFS_ACE_USER)
status = kidmap_getsidbyuid(crgetzone(cr), id,
&domain, &rid);
else
status = kidmap_getsidbygid(crgetzone(cr), id,
&domain, &rid);
if (status != 0) {
/*
* When returning nobody we will need to
* make a dummy fuid table entry for logging
* purposes.
*/
rid = UID_NOBODY;
domain = nulldomain;
2008-11-20 20:01:55 +00:00
}
}
idx = zfs_fuid_find_by_domain(zfsvfs, domain, &kdomain, B_TRUE);
2008-11-20 20:01:55 +00:00
if (!zfsvfs->z_replay)
2009-07-02 22:44:48 +00:00
zfs_fuid_node_add(fuidpp, kdomain,
rid, idx, id, type);
2008-11-20 20:01:55 +00:00
else if (zfuid != NULL) {
list_remove(&fuidp->z_fuids, zfuid);
kmem_free(zfuid, sizeof (zfs_fuid_t));
}
return (FUID_ENCODE(idx, rid));
#else
/*
* The Linux port only supports POSIX IDs, use the passed id.
*/
return (id);
#endif
2008-11-20 20:01:55 +00:00
}
void
zfs_fuid_destroy(zfsvfs_t *zfsvfs)
2008-11-20 20:01:55 +00:00
{
rw_enter(&zfsvfs->z_fuid_lock, RW_WRITER);
if (!zfsvfs->z_fuid_loaded) {
rw_exit(&zfsvfs->z_fuid_lock);
2008-11-20 20:01:55 +00:00
return;
}
zfs_fuid_table_destroy(&zfsvfs->z_fuid_idx, &zfsvfs->z_fuid_domain);
rw_exit(&zfsvfs->z_fuid_lock);
2008-11-20 20:01:55 +00:00
}
/*
* Allocate zfs_fuid_info for tracking FUIDs created during
* zfs_mknode, VOP_SETATTR() or VOP_SETSECATTR()
*/
zfs_fuid_info_t *
zfs_fuid_info_alloc(void)
{
zfs_fuid_info_t *fuidp;
fuidp = kmem_zalloc(sizeof (zfs_fuid_info_t), KM_SLEEP);
list_create(&fuidp->z_domains, sizeof (zfs_fuid_domain_t),
offsetof(zfs_fuid_domain_t, z_next));
list_create(&fuidp->z_fuids, sizeof (zfs_fuid_t),
offsetof(zfs_fuid_t, z_next));
return (fuidp);
}
/*
* Release all memory associated with zfs_fuid_info_t
*/
void
zfs_fuid_info_free(zfs_fuid_info_t *fuidp)
{
zfs_fuid_t *zfuid;
zfs_fuid_domain_t *zdomain;
while ((zfuid = list_head(&fuidp->z_fuids)) != NULL) {
list_remove(&fuidp->z_fuids, zfuid);
kmem_free(zfuid, sizeof (zfs_fuid_t));
}
if (fuidp->z_domain_table != NULL)
kmem_free(fuidp->z_domain_table,
(sizeof (char *)) * fuidp->z_domain_cnt);
2008-11-20 20:01:55 +00:00
while ((zdomain = list_head(&fuidp->z_domains)) != NULL) {
list_remove(&fuidp->z_domains, zdomain);
kmem_free(zdomain, sizeof (zfs_fuid_domain_t));
}
kmem_free(fuidp, sizeof (zfs_fuid_info_t));
}
/*
* Check to see if id is a groupmember. If cred
* has ksid info then sidlist is checked first
* and if still not found then POSIX groups are checked
*
* Will use a straight FUID compare when possible.
*/
boolean_t
zfs_groupmember(zfsvfs_t *zfsvfs, uint64_t id, cred_t *cr)
2008-11-20 20:01:55 +00:00
{
#ifdef HAVE_KSID
2008-11-20 20:01:55 +00:00
ksid_t *ksid = crgetsid(cr, KSID_GROUP);
2009-07-02 22:44:48 +00:00
ksidlist_t *ksidlist = crgetsidlist(cr);
2008-11-20 20:01:55 +00:00
uid_t gid;
2009-07-02 22:44:48 +00:00
if (ksid && ksidlist) {
Prototype/structure update for Linux I appologize in advance why to many things ended up in this commit. When it could be seperated in to a whole series of commits teasing that all apart now would take considerable time and I'm not sure there's much merrit in it. As such I'll just summerize the intent of the changes which are all (or partly) in this commit. Broadly the intent is to remove as much Solaris specific code as possible and replace it with native Linux equivilants. More specifically: 1) Replace all instances of zfsvfs_t with zfs_sb_t. While the type is largely the same calling it private super block data rather than a zfsvfs is more consistent with how Linux names this. While non critical it makes the code easier to read when your thinking in Linux friendly VFS terms. 2) Replace vnode_t with struct inode. The Linux VFS doesn't have the notion of a vnode and there's absolutely no good reason to create one. There are in fact several good reasons to remove it. It just adds overhead on Linux if we were to manage one, it conplicates the code, and it likely will lead to bugs so there's a good change it will be out of date. The code has been updated to remove all need for this type. 3) Replace all vtype_t's with umode types. Along with this shift all uses of types to mode bits. The Solaris code would pass a vtype which is redundant with the Linux mode. Just update all the code to use the Linux mode macros and remove this redundancy. 4) Remove using of vn_* helpers and replace where needed with inode helpers. The big example here is creating iput_aync to replace vn_rele_async. Other vn helpers will be addressed as needed but they should be be emulated. They are a Solaris VFS'ism and should simply be replaced with Linux equivilants. 5) Update znode alloc/free code. Under Linux it's common to embed the inode specific data with the inode itself. This removes the need for an extra memory allocation. In zfs this information is called a znode and it now embeds the inode with it. Allocators have been updated accordingly. 6) Minimal integration with the vfs flags for setting up the super block and handling mount options has been added this code will need to be refined but functionally it's all there. This will be the first and last of these to large to review commits.
2011-02-08 19:16:06 +00:00
int i;
2008-11-20 20:01:55 +00:00
ksid_t *ksid_groups;
uint32_t idx = FUID_INDEX(id);
uint32_t rid = FUID_RID(id);
ksid_groups = ksidlist->ksl_sids;
for (i = 0; i != ksidlist->ksl_nsid; i++) {
if (idx == 0) {
if (id != IDMAP_WK_CREATOR_GROUP_GID &&
id == ksid_groups[i].ks_id) {
return (B_TRUE);
}
} else {
2009-07-02 22:44:48 +00:00
const char *domain;
2008-11-20 20:01:55 +00:00
domain = zfs_fuid_find_by_idx(zfsvfs, idx);
2008-11-20 20:01:55 +00:00
ASSERT(domain != NULL);
if (strcmp(domain,
IDMAP_WK_CREATOR_SID_AUTHORITY) == 0)
return (B_FALSE);
if ((strcmp(domain,
ksid_groups[i].ks_domain->kd_name) == 0) &&
rid == ksid_groups[i].ks_rid)
return (B_TRUE);
}
}
}
/*
* Not found in ksidlist, check posix groups
*/
gid = zfs_fuid_map_id(zfsvfs, id, cr, ZFS_GROUP);
2008-11-20 20:01:55 +00:00
return (groupmember(gid, cr));
#else
return (B_TRUE);
#endif
2008-11-20 20:01:55 +00:00
}
2009-07-02 22:44:48 +00:00
void
zfs_fuid_txhold(zfsvfs_t *zfsvfs, dmu_tx_t *tx)
2009-07-02 22:44:48 +00:00
{
if (zfsvfs->z_fuid_obj == 0) {
2009-07-02 22:44:48 +00:00
dmu_tx_hold_bonus(tx, DMU_NEW_OBJECT);
dmu_tx_hold_write(tx, DMU_NEW_OBJECT, 0,
FUID_SIZE_ESTIMATE(zfsvfs));
2009-07-02 22:44:48 +00:00
dmu_tx_hold_zap(tx, MASTER_NODE_OBJ, FALSE, NULL);
} else {
dmu_tx_hold_bonus(tx, zfsvfs->z_fuid_obj);
dmu_tx_hold_write(tx, zfsvfs->z_fuid_obj, 0,
FUID_SIZE_ESTIMATE(zfsvfs));
2009-07-02 22:44:48 +00:00
}
}
/*
* buf must be big enough (eg, 32 bytes)
*/
int
zfs_id_to_fuidstr(zfsvfs_t *zfsvfs, const char *domain, uid_t rid,
char *buf, size_t len, boolean_t addok)
{
uint64_t fuid;
int domainid = 0;
if (domain && domain[0]) {
domainid = zfs_fuid_find_by_domain(zfsvfs, domain, NULL, addok);
if (domainid == -1)
return (SET_ERROR(ENOENT));
}
fuid = FUID_ENCODE(domainid, rid);
(void) snprintf(buf, len, "%llx", (longlong_t)fuid);
return (0);
}
2008-11-20 20:01:55 +00:00
#endif