Merge tag 'gfs2-v6.5-rc5-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/gfs2/linux-gfs2
Pull gfs2 updates from Andreas Gruenbacher:
- Fix a glock state (non-)transition bug when a dlm request times out
and is canceled, and we have locking requests that can now be granted
immediately
- Various fixes and cleanups in how the logd and quotad daemons are
woken up and terminated
- Fix several bugs in the quota data reference counting and shrinking.
Free quota data objects synchronously in put_super() instead of
letting call_rcu() run wild
- Make sure not to deallocate quota data during a withdraw; rather,
defer quota data deallocation to put_super(). Withdraws can happen in
contexts in which callers on the stack are holding quota data
references
- Many minor quota fixes and cleanups by Bob
- Update the the mailing list address for gfs2 and dlm. (It's the same
list for both and we are moving it to gfs2@lists.linux.dev)
- Various other minor cleanups
* tag 'gfs2-v6.5-rc5-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/gfs2/linux-gfs2: (51 commits)
MAINTAINERS: Update dlm mailing list
MAINTAINERS: Update gfs2 mailing list
gfs2: change qd_slot_count to qd_slot_ref
gfs2: check for no eligible quota changes
gfs2: Remove useless assignment
gfs2: simplify slot_get
gfs2: Simplify qd2offset
gfs2: introduce qd_bh_get_or_undo
gfs2: Remove quota allocation info from quota file
gfs2: use constant for array size
gfs2: Set qd_sync_gen in do_sync
gfs2: Remove useless err set
gfs2: Small gfs2_quota_lock cleanup
gfs2: move qdsb_put and reduce redundancy
gfs2: improvements to sysfs status
gfs2: Don't try to sync non-changes
gfs2: Simplify function need_sync
gfs2: remove unneeded pg_oflow variable
gfs2: remove unneeded variable done
gfs2: pass sdp to gfs2_write_buf_to_page
...
diff --git a/Documentation/filesystems/gfs2-glocks.rst b/Documentation/filesystems/gfs2-glocks.rst
index bec25c8..8a58429 100644
--- a/Documentation/filesystems/gfs2-glocks.rst
+++ b/Documentation/filesystems/gfs2-glocks.rst
@@ -20,8 +20,7 @@
just the holders) associated with the glock. If there are any
held locks, then they will be contiguous entries at the head
of the list. Locks are granted in strictly the order that they
-are queued, except for those marked LM_FLAG_PRIORITY which are
-used only during recovery, and even then only for journal locks.
+are queued.
There are three lock states that users of the glock layer can request,
namely shared (SH), deferred (DF) and exclusive (EX). Those translate
diff --git a/MAINTAINERS b/MAINTAINERS
index 0c098d5..c8fdd0d 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -6118,7 +6118,7 @@
DISTRIBUTED LOCK MANAGER (DLM)
M: Christine Caulfield <ccaulfie@redhat.com>
M: David Teigland <teigland@redhat.com>
-L: cluster-devel@redhat.com
+L: gfs2@lists.linux.dev
S: Supported
W: http://sources.redhat.com/cluster/
T: git git://git.kernel.org/pub/scm/linux/kernel/git/teigland/linux-dlm.git
@@ -8774,7 +8774,7 @@
GFS2 FILE SYSTEM
M: Bob Peterson <rpeterso@redhat.com>
M: Andreas Gruenbacher <agruenba@redhat.com>
-L: cluster-devel@redhat.com
+L: gfs2@lists.linux.dev
S: Supported
B: https://bugzilla.kernel.org/enter_bug.cgi?product=File%20System&component=gfs2
T: git git://git.kernel.org/pub/scm/linux/kernel/git/gfs2/linux-gfs2.git
diff --git a/fs/gfs2/aops.c b/fs/gfs2/aops.c
index 9c4b26a..c26d483 100644
--- a/fs/gfs2/aops.c
+++ b/fs/gfs2/aops.c
@@ -183,13 +183,13 @@
int ret;
/*
- * Even if we didn't write any pages here, we might still be holding
+ * Even if we didn't write enough pages here, we might still be holding
* dirty pages in the ail. We forcibly flush the ail because we don't
* want balance_dirty_pages() to loop indefinitely trying to write out
* pages held in the ail that it can't find.
*/
ret = iomap_writepages(mapping, wbc, &wpc, &gfs2_writeback_ops);
- if (ret == 0)
+ if (ret == 0 && wbc->nr_to_write > 0)
set_bit(SDF_FORCE_AIL_FLUSH, &sdp->sd_flags);
return ret;
}
@@ -272,8 +272,7 @@
* not be suitable for data integrity
* writeout).
*/
- *done_index = folio->index +
- folio_nr_pages(folio);
+ *done_index = folio_next_index(folio);
ret = 1;
break;
}
diff --git a/fs/gfs2/bmap.c b/fs/gfs2/bmap.c
index f62366b..ef7017f 100644
--- a/fs/gfs2/bmap.c
+++ b/fs/gfs2/bmap.c
@@ -161,7 +161,7 @@
int error;
down_write(&ip->i_rw_mutex);
- page = find_or_create_page(inode->i_mapping, 0, GFP_NOFS);
+ page = grab_cache_page(inode->i_mapping, 0);
error = -ENOMEM;
if (!page)
goto out;
diff --git a/fs/gfs2/glock.c b/fs/gfs2/glock.c
index 1438e74..9cbf8d9 100644
--- a/fs/gfs2/glock.c
+++ b/fs/gfs2/glock.c
@@ -176,7 +176,7 @@
wake_up_glock(gl);
call_rcu(&gl->gl_rcu, gfs2_glock_dealloc);
if (atomic_dec_and_test(&sdp->sd_glock_disposal))
- wake_up(&sdp->sd_glock_wait);
+ wake_up(&sdp->sd_kill_wait);
}
/**
@@ -468,10 +468,10 @@
* do_promote - promote as many requests as possible on the current queue
* @gl: The glock
*
- * Returns: 1 if there is a blocked holder at the head of the list
+ * Returns true on success (i.e., progress was made or there are no waiters).
*/
-static int do_promote(struct gfs2_glock *gl)
+static bool do_promote(struct gfs2_glock *gl)
{
struct gfs2_holder *gh, *current_gh;
@@ -484,10 +484,10 @@
* If we get here, it means we may not grant this
* holder for some reason. If this holder is at the
* head of the list, it means we have a blocked holder
- * at the head, so return 1.
+ * at the head, so return false.
*/
if (list_is_first(&gh->gh_list, &gl->gl_holders))
- return 1;
+ return false;
do_error(gl, 0);
break;
}
@@ -497,7 +497,7 @@
if (!current_gh)
current_gh = gh;
}
- return 0;
+ return true;
}
/**
@@ -591,10 +591,11 @@
if (gh && !test_bit(GLF_DEMOTE_IN_PROGRESS, &gl->gl_flags)) {
/* move to back of queue and try next entry */
if (ret & LM_OUT_CANCELED) {
- if ((gh->gh_flags & LM_FLAG_PRIORITY) == 0)
- list_move_tail(&gh->gh_list, &gl->gl_holders);
+ list_move_tail(&gh->gh_list, &gl->gl_holders);
gh = find_first_waiter(gl);
gl->gl_target = gh->gh_state;
+ if (do_promote(gl))
+ goto out;
goto retry;
}
/* Some error or failed "try lock" - report it */
@@ -679,8 +680,7 @@
gh && !(gh->gh_flags & LM_FLAG_NOEXP))
goto skip_inval;
- lck_flags &= (LM_FLAG_TRY | LM_FLAG_TRY_1CB | LM_FLAG_NOEXP |
- LM_FLAG_PRIORITY);
+ lck_flags &= (LM_FLAG_TRY | LM_FLAG_TRY_1CB | LM_FLAG_NOEXP);
GLOCK_BUG_ON(gl, gl->gl_state == target);
GLOCK_BUG_ON(gl, gl->gl_state == gl->gl_target);
if ((target == LM_ST_UNLOCKED || target == LM_ST_DEFERRED) &&
@@ -834,7 +834,7 @@
} else {
if (test_bit(GLF_DEMOTE, &gl->gl_flags))
gfs2_demote_wake(gl);
- if (do_promote(gl) == 0)
+ if (do_promote(gl))
goto out_unlock;
gh = find_first_waiter(gl);
gl->gl_target = gh->gh_state;
@@ -1022,7 +1022,7 @@
* step entirely.
*/
if (gfs2_try_evict(gl)) {
- if (test_bit(SDF_DEACTIVATING, &sdp->sd_flags))
+ if (test_bit(SDF_KILL, &sdp->sd_flags))
goto out;
if (gfs2_queue_verify_evict(gl))
return;
@@ -1035,7 +1035,7 @@
GFS2_BLKST_UNLINKED);
if (IS_ERR(inode)) {
if (PTR_ERR(inode) == -EAGAIN &&
- !test_bit(SDF_DEACTIVATING, &sdp->sd_flags) &&
+ !test_bit(SDF_KILL, &sdp->sd_flags) &&
gfs2_queue_verify_evict(gl))
return;
} else {
@@ -1231,7 +1231,7 @@
out_free:
gfs2_glock_dealloc(&gl->gl_rcu);
if (atomic_dec_and_test(&sdp->sd_glock_disposal))
- wake_up(&sdp->sd_glock_wait);
+ wake_up(&sdp->sd_kill_wait);
out:
return ret;
@@ -1515,27 +1515,20 @@
}
if (test_bit(HIF_HOLDER, &gh2->gh_iflags))
continue;
- if (unlikely((gh->gh_flags & LM_FLAG_PRIORITY) && !insert_pt))
- insert_pt = &gh2->gh_list;
}
trace_gfs2_glock_queue(gh, 1);
gfs2_glstats_inc(gl, GFS2_LKS_QCOUNT);
gfs2_sbstats_inc(gl, GFS2_LKS_QCOUNT);
if (likely(insert_pt == NULL)) {
list_add_tail(&gh->gh_list, &gl->gl_holders);
- if (unlikely(gh->gh_flags & LM_FLAG_PRIORITY))
- goto do_cancel;
return;
}
list_add_tail(&gh->gh_list, insert_pt);
-do_cancel:
gh = list_first_entry(&gl->gl_holders, struct gfs2_holder, gh_list);
- if (!(gh->gh_flags & LM_FLAG_PRIORITY)) {
- spin_unlock(&gl->gl_lockref.lock);
- if (sdp->sd_lockstruct.ls_ops->lm_cancel)
- sdp->sd_lockstruct.ls_ops->lm_cancel(gl);
- spin_lock(&gl->gl_lockref.lock);
- }
+ spin_unlock(&gl->gl_lockref.lock);
+ if (sdp->sd_lockstruct.ls_ops->lm_cancel)
+ sdp->sd_lockstruct.ls_ops->lm_cancel(gl);
+ spin_lock(&gl->gl_lockref.lock);
return;
trap_recursive:
@@ -2195,7 +2188,7 @@
flush_workqueue(glock_workqueue);
glock_hash_walk(clear_glock, sdp);
flush_workqueue(glock_workqueue);
- wait_event_timeout(sdp->sd_glock_wait,
+ wait_event_timeout(sdp->sd_kill_wait,
atomic_read(&sdp->sd_glock_disposal) == 0,
HZ * 600);
glock_hash_walk(dump_glock_func, sdp);
@@ -2227,8 +2220,6 @@
*p++ = 'e';
if (flags & LM_FLAG_ANY)
*p++ = 'A';
- if (flags & LM_FLAG_PRIORITY)
- *p++ = 'p';
if (flags & LM_FLAG_NODE_SCOPE)
*p++ = 'n';
if (flags & GL_ASYNC)
diff --git a/fs/gfs2/glock.h b/fs/gfs2/glock.h
index 1f1ba92..c8685ca 100644
--- a/fs/gfs2/glock.h
+++ b/fs/gfs2/glock.h
@@ -68,14 +68,6 @@
* also be granted in SHARED. The preferred state is whichever is compatible
* with other granted locks, or the specified state if no other locks exist.
*
- * LM_FLAG_PRIORITY
- * Override fairness considerations. Suppose a lock is held in a shared state
- * and there is a pending request for the deferred state. A shared lock
- * request with the priority flag would be allowed to bypass the deferred
- * request and directly join the other shared lock. A shared lock request
- * without the priority flag might be forced to wait until the deferred
- * requested had acquired and released the lock.
- *
* LM_FLAG_NODE_SCOPE
* This holder agrees to share the lock within this node. In other words,
* the glock is held in EX mode according to DLM, but local holders on the
@@ -86,7 +78,6 @@
#define LM_FLAG_TRY_1CB 0x0002
#define LM_FLAG_NOEXP 0x0004
#define LM_FLAG_ANY 0x0008
-#define LM_FLAG_PRIORITY 0x0010
#define LM_FLAG_NODE_SCOPE 0x0020
#define GL_ASYNC 0x0040
#define GL_EXACT 0x0080
diff --git a/fs/gfs2/glops.c b/fs/gfs2/glops.c
index aecdac3..d26759a 100644
--- a/fs/gfs2/glops.c
+++ b/fs/gfs2/glops.c
@@ -637,7 +637,7 @@
struct gfs2_sbd *sdp = gl->gl_name.ln_sbd;
if (!remote || sb_rdonly(sdp->sd_vfs) ||
- test_bit(SDF_DEACTIVATING, &sdp->sd_flags))
+ test_bit(SDF_KILL, &sdp->sd_flags))
return;
if (gl->gl_demote_state == LM_ST_UNLOCKED &&
diff --git a/fs/gfs2/incore.h b/fs/gfs2/incore.h
index 04f2d78..a8c95c5 100644
--- a/fs/gfs2/incore.h
+++ b/fs/gfs2/incore.h
@@ -452,7 +452,7 @@
s64 qd_change_sync;
unsigned int qd_slot;
- unsigned int qd_slot_count;
+ unsigned int qd_slot_ref;
struct buffer_head *qd_bh;
struct gfs2_quota_change *qd_bh_qc;
@@ -537,6 +537,7 @@
#define GFS2_QUOTA_OFF 0
#define GFS2_QUOTA_ACCOUNT 1
#define GFS2_QUOTA_ON 2
+#define GFS2_QUOTA_QUIET 3 /* on but not complaining */
#define GFS2_DATA_DEFAULT GFS2_DATA_ORDERED
#define GFS2_DATA_WRITEBACK 1
@@ -606,7 +607,7 @@
SDF_REMOTE_WITHDRAW = 13, /* Performing remote recovery */
SDF_WITHDRAW_RECOVERY = 14, /* Wait for journal recovery when we are
withdrawing */
- SDF_DEACTIVATING = 15,
+ SDF_KILL = 15,
SDF_EVICTING = 16,
SDF_FROZEN = 17,
};
@@ -716,7 +717,7 @@
struct gfs2_glock *sd_rename_gl;
struct gfs2_glock *sd_freeze_gl;
struct work_struct sd_freeze_work;
- wait_queue_head_t sd_glock_wait;
+ wait_queue_head_t sd_kill_wait;
wait_queue_head_t sd_async_glock_wait;
atomic_t sd_glock_disposal;
struct completion sd_locking_init;
diff --git a/fs/gfs2/inode.c b/fs/gfs2/inode.c
index a21ac41..0eac045 100644
--- a/fs/gfs2/inode.c
+++ b/fs/gfs2/inode.c
@@ -276,10 +276,16 @@
* gfs2_lookup_simple callers expect ENOENT
* and do not check for NULL.
*/
- if (inode == NULL)
- return ERR_PTR(-ENOENT);
- else
- return inode;
+ if (IS_ERR_OR_NULL(inode))
+ return inode ? inode : ERR_PTR(-ENOENT);
+
+ /*
+ * Must not call back into the filesystem when allocating
+ * pages in the metadata inode's address space.
+ */
+ mapping_set_gfp_mask(inode->i_mapping, GFP_NOFS);
+
+ return inode;
}
diff --git a/fs/gfs2/lock_dlm.c b/fs/gfs2/lock_dlm.c
index 5491129..59ab18c 100644
--- a/fs/gfs2/lock_dlm.c
+++ b/fs/gfs2/lock_dlm.c
@@ -222,11 +222,6 @@
lkf |= DLM_LKF_NOQUEUEBAST;
}
- if (gfs_flags & LM_FLAG_PRIORITY) {
- lkf |= DLM_LKF_NOORDER;
- lkf |= DLM_LKF_HEADQUE;
- }
-
if (gfs_flags & LM_FLAG_ANY) {
if (req == DLM_LOCK_PR)
lkf |= DLM_LKF_ALTCW;
diff --git a/fs/gfs2/log.c b/fs/gfs2/log.c
index aa56879..e5271ae 100644
--- a/fs/gfs2/log.c
+++ b/fs/gfs2/log.c
@@ -1227,6 +1227,21 @@
gfs2_log_unlock(sdp);
}
+static inline int gfs2_jrnl_flush_reqd(struct gfs2_sbd *sdp)
+{
+ return atomic_read(&sdp->sd_log_pinned) +
+ atomic_read(&sdp->sd_log_blks_needed) >=
+ atomic_read(&sdp->sd_log_thresh1);
+}
+
+static inline int gfs2_ail_flush_reqd(struct gfs2_sbd *sdp)
+{
+ return sdp->sd_jdesc->jd_blocks -
+ atomic_read(&sdp->sd_log_blks_free) +
+ atomic_read(&sdp->sd_log_blks_needed) >=
+ atomic_read(&sdp->sd_log_thresh2);
+}
+
/**
* gfs2_log_commit - Commit a transaction to the log
* @sdp: the filesystem
@@ -1246,9 +1261,7 @@
{
log_refund(sdp, tr);
- if (atomic_read(&sdp->sd_log_pinned) > atomic_read(&sdp->sd_log_thresh1) ||
- ((sdp->sd_jdesc->jd_blocks - atomic_read(&sdp->sd_log_blks_free)) >
- atomic_read(&sdp->sd_log_thresh2)))
+ if (gfs2_ail_flush_reqd(sdp) || gfs2_jrnl_flush_reqd(sdp))
wake_up(&sdp->sd_logd_waitq);
}
@@ -1271,24 +1284,6 @@
gfs2_assert_warn(sdp, list_empty(&sdp->sd_ail2_list));
}
-static inline int gfs2_jrnl_flush_reqd(struct gfs2_sbd *sdp)
-{
- return (atomic_read(&sdp->sd_log_pinned) +
- atomic_read(&sdp->sd_log_blks_needed) >=
- atomic_read(&sdp->sd_log_thresh1));
-}
-
-static inline int gfs2_ail_flush_reqd(struct gfs2_sbd *sdp)
-{
- unsigned int used_blocks = sdp->sd_jdesc->jd_blocks - atomic_read(&sdp->sd_log_blks_free);
-
- if (test_and_clear_bit(SDF_FORCE_AIL_FLUSH, &sdp->sd_flags))
- return 1;
-
- return used_blocks + atomic_read(&sdp->sd_log_blks_needed) >=
- atomic_read(&sdp->sd_log_thresh2);
-}
-
/**
* gfs2_logd - Update log tail as Active Items get flushed to in-place blocks
* @data: Pointer to GFS2 superblock
@@ -1301,14 +1296,11 @@
{
struct gfs2_sbd *sdp = data;
unsigned long t = 1;
- DEFINE_WAIT(wait);
while (!kthread_should_stop()) {
+ if (gfs2_withdrawn(sdp))
+ break;
- if (gfs2_withdrawn(sdp)) {
- msleep_interruptible(HZ);
- continue;
- }
/* Check for errors writing to the journal */
if (sdp->sd_log_error) {
gfs2_lm(sdp,
@@ -1317,7 +1309,7 @@
"prevent further damage.\n",
sdp->sd_fsname, sdp->sd_log_error);
gfs2_withdraw(sdp);
- continue;
+ break;
}
if (gfs2_jrnl_flush_reqd(sdp) || t == 0) {
@@ -1326,7 +1318,9 @@
GFS2_LFC_LOGD_JFLUSH_REQD);
}
- if (gfs2_ail_flush_reqd(sdp)) {
+ if (test_bit(SDF_FORCE_AIL_FLUSH, &sdp->sd_flags) ||
+ gfs2_ail_flush_reqd(sdp)) {
+ clear_bit(SDF_FORCE_AIL_FLUSH, &sdp->sd_flags);
gfs2_ail1_start(sdp);
gfs2_ail1_wait(sdp);
gfs2_ail1_empty(sdp, 0);
@@ -1338,17 +1332,14 @@
try_to_freeze();
- do {
- prepare_to_wait(&sdp->sd_logd_waitq, &wait,
- TASK_INTERRUPTIBLE);
- if (!gfs2_ail_flush_reqd(sdp) &&
- !gfs2_jrnl_flush_reqd(sdp) &&
- !kthread_should_stop())
- t = schedule_timeout(t);
- } while(t && !gfs2_ail_flush_reqd(sdp) &&
- !gfs2_jrnl_flush_reqd(sdp) &&
- !kthread_should_stop());
- finish_wait(&sdp->sd_logd_waitq, &wait);
+ t = wait_event_interruptible_timeout(sdp->sd_logd_waitq,
+ test_bit(SDF_FORCE_AIL_FLUSH, &sdp->sd_flags) ||
+ gfs2_ail_flush_reqd(sdp) ||
+ gfs2_jrnl_flush_reqd(sdp) ||
+ sdp->sd_log_error ||
+ gfs2_withdrawn(sdp) ||
+ kthread_should_stop(),
+ t);
}
return 0;
diff --git a/fs/gfs2/lops.c b/fs/gfs2/lops.c
index 251322b..483f698 100644
--- a/fs/gfs2/lops.c
+++ b/fs/gfs2/lops.c
@@ -456,7 +456,7 @@
* Find the folio with 'index' in the journal's mapping. Search the folio for
* the journal head if requested (cleanup == false). Release refs on the
* folio so the page cache can reclaim it. We grabbed a
- * reference on this folio twice, first when we did a find_or_create_page()
+ * reference on this folio twice, first when we did a grab_cache_page()
* to obtain the folio to add it to the bio and second when we do a
* filemap_get_folio() here to get the folio to wait on while I/O on it is being
* completed.
@@ -481,7 +481,7 @@
if (!*done)
*done = gfs2_jhead_pg_srch(jd, head, &folio->page);
- /* filemap_get_folio() and the earlier find_or_create_page() */
+ /* filemap_get_folio() and the earlier grab_cache_page() */
folio_put_refs(folio, 2);
}
@@ -535,8 +535,7 @@
for (; block < je->lblock + je->blocks; block++, dblock++) {
if (!page) {
- page = find_or_create_page(mapping,
- block >> shift, GFP_NOFS);
+ page = grab_cache_page(mapping, block >> shift);
if (!page) {
ret = -ENOMEM;
done = true;
diff --git a/fs/gfs2/main.c b/fs/gfs2/main.c
index afcb328..66eb98b 100644
--- a/fs/gfs2/main.c
+++ b/fs/gfs2/main.c
@@ -152,9 +152,9 @@
goto fail_shrinker;
error = -ENOMEM;
- gfs_recovery_wq = alloc_workqueue("gfs_recovery",
+ gfs2_recovery_wq = alloc_workqueue("gfs2_recovery",
WQ_MEM_RECLAIM | WQ_FREEZABLE, 0);
- if (!gfs_recovery_wq)
+ if (!gfs2_recovery_wq)
goto fail_wq1;
gfs2_control_wq = alloc_workqueue("gfs2_control",
@@ -162,7 +162,7 @@
if (!gfs2_control_wq)
goto fail_wq2;
- gfs2_freeze_wq = alloc_workqueue("freeze_workqueue", 0, 0);
+ gfs2_freeze_wq = alloc_workqueue("gfs2_freeze", 0, 0);
if (!gfs2_freeze_wq)
goto fail_wq3;
@@ -194,7 +194,7 @@
fail_wq3:
destroy_workqueue(gfs2_control_wq);
fail_wq2:
- destroy_workqueue(gfs_recovery_wq);
+ destroy_workqueue(gfs2_recovery_wq);
fail_wq1:
unregister_shrinker(&gfs2_qd_shrinker);
fail_shrinker:
@@ -234,7 +234,7 @@
gfs2_unregister_debugfs();
unregister_filesystem(&gfs2_fs_type);
unregister_filesystem(&gfs2meta_fs_type);
- destroy_workqueue(gfs_recovery_wq);
+ destroy_workqueue(gfs2_recovery_wq);
destroy_workqueue(gfs2_control_wq);
destroy_workqueue(gfs2_freeze_wq);
list_lru_destroy(&gfs2_qd_lru);
diff --git a/fs/gfs2/ops_fstype.c b/fs/gfs2/ops_fstype.c
index 8a27957..33ca0473 100644
--- a/fs/gfs2/ops_fstype.c
+++ b/fs/gfs2/ops_fstype.c
@@ -87,7 +87,7 @@
set_bit(SDF_NOJOURNALID, &sdp->sd_flags);
gfs2_tune_init(&sdp->sd_tune);
- init_waitqueue_head(&sdp->sd_glock_wait);
+ init_waitqueue_head(&sdp->sd_kill_wait);
init_waitqueue_head(&sdp->sd_async_glock_wait);
atomic_set(&sdp->sd_glock_disposal, 0);
init_completion(&sdp->sd_locking_init);
@@ -1103,29 +1103,49 @@
struct task_struct *p;
int error = 0;
- p = kthread_run(gfs2_logd, sdp, "gfs2_logd");
+ p = kthread_create(gfs2_logd, sdp, "gfs2_logd/%s", sdp->sd_fsname);
if (IS_ERR(p)) {
error = PTR_ERR(p);
- fs_err(sdp, "can't start logd thread: %d\n", error);
+ fs_err(sdp, "can't create logd thread: %d\n", error);
return error;
}
+ get_task_struct(p);
sdp->sd_logd_process = p;
- p = kthread_run(gfs2_quotad, sdp, "gfs2_quotad");
+ p = kthread_create(gfs2_quotad, sdp, "gfs2_quotad/%s", sdp->sd_fsname);
if (IS_ERR(p)) {
error = PTR_ERR(p);
- fs_err(sdp, "can't start quotad thread: %d\n", error);
+ fs_err(sdp, "can't create quotad thread: %d\n", error);
goto fail;
}
+ get_task_struct(p);
sdp->sd_quotad_process = p;
+
+ wake_up_process(sdp->sd_logd_process);
+ wake_up_process(sdp->sd_quotad_process);
return 0;
fail:
kthread_stop(sdp->sd_logd_process);
+ put_task_struct(sdp->sd_logd_process);
sdp->sd_logd_process = NULL;
return error;
}
+void gfs2_destroy_threads(struct gfs2_sbd *sdp)
+{
+ if (sdp->sd_logd_process) {
+ kthread_stop(sdp->sd_logd_process);
+ put_task_struct(sdp->sd_logd_process);
+ sdp->sd_logd_process = NULL;
+ }
+ if (sdp->sd_quotad_process) {
+ kthread_stop(sdp->sd_quotad_process);
+ put_task_struct(sdp->sd_quotad_process);
+ sdp->sd_quotad_process = NULL;
+ }
+}
+
/**
* gfs2_fill_super - Read in superblock
* @sb: The VFS superblock
@@ -1276,12 +1296,7 @@
if (error) {
gfs2_freeze_unlock(&sdp->sd_freeze_gh);
- if (sdp->sd_quotad_process)
- kthread_stop(sdp->sd_quotad_process);
- sdp->sd_quotad_process = NULL;
- if (sdp->sd_logd_process)
- kthread_stop(sdp->sd_logd_process);
- sdp->sd_logd_process = NULL;
+ gfs2_destroy_threads(sdp);
fs_err(sdp, "can't make FS RW: %d\n", error);
goto fail_per_node;
}
@@ -1381,6 +1396,7 @@
{"off", GFS2_QUOTA_OFF},
{"account", GFS2_QUOTA_ACCOUNT},
{"on", GFS2_QUOTA_ON},
+ {"quiet", GFS2_QUOTA_QUIET},
{}
};
@@ -1786,9 +1802,9 @@
/*
* Flush and then drain the delete workqueue here (via
* destroy_workqueue()) to ensure that any delete work that
- * may be running will also see the SDF_DEACTIVATING flag.
+ * may be running will also see the SDF_KILL flag.
*/
- set_bit(SDF_DEACTIVATING, &sdp->sd_flags);
+ set_bit(SDF_KILL, &sdp->sd_flags);
gfs2_flush_delete_work(sdp);
destroy_workqueue(sdp->sd_delete_wq);
diff --git a/fs/gfs2/quota.c b/fs/gfs2/quota.c
index aa5fd06..171b271 100644
--- a/fs/gfs2/quota.c
+++ b/fs/gfs2/quota.c
@@ -109,38 +109,44 @@
static void gfs2_qd_dealloc(struct rcu_head *rcu)
{
struct gfs2_quota_data *qd = container_of(rcu, struct gfs2_quota_data, qd_rcu);
+ struct gfs2_sbd *sdp = qd->qd_sbd;
+
kmem_cache_free(gfs2_quotad_cachep, qd);
+ if (atomic_dec_and_test(&sdp->sd_quota_count))
+ wake_up(&sdp->sd_kill_wait);
}
-static void gfs2_qd_dispose(struct list_head *list)
+static void gfs2_qd_dispose(struct gfs2_quota_data *qd)
+{
+ struct gfs2_sbd *sdp = qd->qd_sbd;
+
+ spin_lock(&qd_lock);
+ list_del(&qd->qd_list);
+ spin_unlock(&qd_lock);
+
+ spin_lock_bucket(qd->qd_hash);
+ hlist_bl_del_rcu(&qd->qd_hlist);
+ spin_unlock_bucket(qd->qd_hash);
+
+ if (!gfs2_withdrawn(sdp)) {
+ gfs2_assert_warn(sdp, !qd->qd_change);
+ gfs2_assert_warn(sdp, !qd->qd_slot_ref);
+ gfs2_assert_warn(sdp, !qd->qd_bh_count);
+ }
+
+ gfs2_glock_put(qd->qd_gl);
+ call_rcu(&qd->qd_rcu, gfs2_qd_dealloc);
+}
+
+static void gfs2_qd_list_dispose(struct list_head *list)
{
struct gfs2_quota_data *qd;
- struct gfs2_sbd *sdp;
while (!list_empty(list)) {
qd = list_first_entry(list, struct gfs2_quota_data, qd_lru);
- sdp = qd->qd_gl->gl_name.ln_sbd;
-
list_del(&qd->qd_lru);
- /* Free from the filesystem-specific list */
- spin_lock(&qd_lock);
- list_del(&qd->qd_list);
- spin_unlock(&qd_lock);
-
- spin_lock_bucket(qd->qd_hash);
- hlist_bl_del_rcu(&qd->qd_hlist);
- spin_unlock_bucket(qd->qd_hash);
-
- gfs2_assert_warn(sdp, !qd->qd_change);
- gfs2_assert_warn(sdp, !qd->qd_slot_count);
- gfs2_assert_warn(sdp, !qd->qd_bh_count);
-
- gfs2_glock_put(qd->qd_gl);
- atomic_dec(&sdp->sd_quota_count);
-
- /* Delete it from the common reclaim list */
- call_rcu(&qd->qd_rcu, gfs2_qd_dealloc);
+ gfs2_qd_dispose(qd);
}
}
@@ -149,18 +155,22 @@
struct list_lru_one *lru, spinlock_t *lru_lock, void *arg)
{
struct list_head *dispose = arg;
- struct gfs2_quota_data *qd = list_entry(item, struct gfs2_quota_data, qd_lru);
+ struct gfs2_quota_data *qd =
+ list_entry(item, struct gfs2_quota_data, qd_lru);
+ enum lru_status status;
if (!spin_trylock(&qd->qd_lockref.lock))
return LRU_SKIP;
+ status = LRU_SKIP;
if (qd->qd_lockref.count == 0) {
lockref_mark_dead(&qd->qd_lockref);
list_lru_isolate_move(lru, &qd->qd_lru, dispose);
+ status = LRU_REMOVED;
}
spin_unlock(&qd->qd_lockref.lock);
- return LRU_REMOVED;
+ return status;
}
static unsigned long gfs2_qd_shrink_scan(struct shrinker *shrink,
@@ -175,7 +185,7 @@
freed = list_lru_shrink_walk(&gfs2_qd_lru, sc,
gfs2_qd_isolate, &dispose);
- gfs2_qd_dispose(&dispose);
+ gfs2_qd_list_dispose(&dispose);
return freed;
}
@@ -203,12 +213,7 @@
static u64 qd2offset(struct gfs2_quota_data *qd)
{
- u64 offset;
-
- offset = qd2index(qd);
- offset *= sizeof(struct gfs2_quota);
-
- return offset;
+ return qd2index(qd) * sizeof(struct gfs2_quota);
}
static struct gfs2_quota_data *qd_alloc(unsigned hash, struct gfs2_sbd *sdp, struct kqid qid)
@@ -221,7 +226,7 @@
return NULL;
qd->qd_sbd = sdp;
- qd->qd_lockref.count = 1;
+ qd->qd_lockref.count = 0;
spin_lock_init(&qd->qd_lockref.lock);
qd->qd_id = qid;
qd->qd_slot = -1;
@@ -283,6 +288,7 @@
spin_lock_bucket(hash);
*qdp = qd = gfs2_qd_search_bucket(hash, sdp, qid);
if (qd == NULL) {
+ new_qd->qd_lockref.count++;
*qdp = new_qd;
list_add(&new_qd->qd_list, &sdp->sd_quota_list);
hlist_bl_add_head_rcu(&new_qd->qd_hlist, &qd_hash_table[hash]);
@@ -302,20 +308,31 @@
static void qd_hold(struct gfs2_quota_data *qd)
{
- struct gfs2_sbd *sdp = qd->qd_gl->gl_name.ln_sbd;
+ struct gfs2_sbd *sdp = qd->qd_sbd;
gfs2_assert(sdp, !__lockref_is_dead(&qd->qd_lockref));
lockref_get(&qd->qd_lockref);
}
static void qd_put(struct gfs2_quota_data *qd)
{
+ struct gfs2_sbd *sdp;
+
if (lockref_put_or_lock(&qd->qd_lockref))
return;
+ BUG_ON(__lockref_is_dead(&qd->qd_lockref));
+ sdp = qd->qd_sbd;
+ if (unlikely(!test_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags))) {
+ lockref_mark_dead(&qd->qd_lockref);
+ spin_unlock(&qd->qd_lockref.lock);
+
+ gfs2_qd_dispose(qd);
+ return;
+ }
+
qd->qd_lockref.count = 0;
list_lru_add(&gfs2_qd_lru, &qd->qd_lru);
spin_unlock(&qd->qd_lockref.lock);
-
}
static int slot_get(struct gfs2_quota_data *qd)
@@ -325,20 +342,19 @@
int error = 0;
spin_lock(&sdp->sd_bitmap_lock);
- if (qd->qd_slot_count != 0)
- goto out;
-
- error = -ENOSPC;
- bit = find_first_zero_bit(sdp->sd_quota_bitmap, sdp->sd_quota_slots);
- if (bit < sdp->sd_quota_slots) {
+ if (qd->qd_slot_ref == 0) {
+ bit = find_first_zero_bit(sdp->sd_quota_bitmap,
+ sdp->sd_quota_slots);
+ if (bit >= sdp->sd_quota_slots) {
+ error = -ENOSPC;
+ goto out;
+ }
set_bit(bit, sdp->sd_quota_bitmap);
qd->qd_slot = bit;
- error = 0;
-out:
- qd->qd_slot_count++;
}
+ qd->qd_slot_ref++;
+out:
spin_unlock(&sdp->sd_bitmap_lock);
-
return error;
}
@@ -347,8 +363,8 @@
struct gfs2_sbd *sdp = qd->qd_sbd;
spin_lock(&sdp->sd_bitmap_lock);
- gfs2_assert(sdp, qd->qd_slot_count);
- qd->qd_slot_count++;
+ gfs2_assert(sdp, qd->qd_slot_ref);
+ qd->qd_slot_ref++;
spin_unlock(&sdp->sd_bitmap_lock);
}
@@ -357,8 +373,8 @@
struct gfs2_sbd *sdp = qd->qd_sbd;
spin_lock(&sdp->sd_bitmap_lock);
- gfs2_assert(sdp, qd->qd_slot_count);
- if (!--qd->qd_slot_count) {
+ gfs2_assert(sdp, qd->qd_slot_ref);
+ if (!--qd->qd_slot_ref) {
BUG_ON(!test_and_clear_bit(qd->qd_slot, sdp->sd_quota_bitmap));
qd->qd_slot = -1;
}
@@ -367,7 +383,7 @@
static int bh_get(struct gfs2_quota_data *qd)
{
- struct gfs2_sbd *sdp = qd->qd_gl->gl_name.ln_sbd;
+ struct gfs2_sbd *sdp = qd->qd_sbd;
struct inode *inode = sdp->sd_qc_inode;
struct gfs2_inode *ip = GFS2_I(inode);
unsigned int block, offset;
@@ -421,7 +437,7 @@
static void bh_put(struct gfs2_quota_data *qd)
{
- struct gfs2_sbd *sdp = qd->qd_gl->gl_name.ln_sbd;
+ struct gfs2_sbd *sdp = qd->qd_sbd;
mutex_lock(&sdp->sd_quota_mutex);
gfs2_assert(sdp, qd->qd_bh_count);
@@ -451,6 +467,20 @@
return 1;
}
+static int qd_bh_get_or_undo(struct gfs2_sbd *sdp, struct gfs2_quota_data *qd)
+{
+ int error;
+
+ error = bh_get(qd);
+ if (!error)
+ return 0;
+
+ clear_bit(QDF_LOCKED, &qd->qd_flags);
+ slot_put(qd);
+ qd_put(qd);
+ return error;
+}
+
static int qd_fish(struct gfs2_sbd *sdp, struct gfs2_quota_data **qdp)
{
struct gfs2_quota_data *qd = NULL, *iter;
@@ -473,30 +503,29 @@
spin_unlock(&qd_lock);
if (qd) {
- error = bh_get(qd);
- if (error) {
- clear_bit(QDF_LOCKED, &qd->qd_flags);
- slot_put(qd);
- qd_put(qd);
+ error = qd_bh_get_or_undo(sdp, qd);
+ if (error)
return error;
- }
+ *qdp = qd;
}
- *qdp = qd;
-
return 0;
}
-static void qd_unlock(struct gfs2_quota_data *qd)
+static void qdsb_put(struct gfs2_quota_data *qd)
{
- gfs2_assert_warn(qd->qd_gl->gl_name.ln_sbd,
- test_bit(QDF_LOCKED, &qd->qd_flags));
- clear_bit(QDF_LOCKED, &qd->qd_flags);
bh_put(qd);
slot_put(qd);
qd_put(qd);
}
+static void qd_unlock(struct gfs2_quota_data *qd)
+{
+ gfs2_assert_warn(qd->qd_sbd, test_bit(QDF_LOCKED, &qd->qd_flags));
+ clear_bit(QDF_LOCKED, &qd->qd_flags);
+ qdsb_put(qd);
+}
+
static int qdsb_get(struct gfs2_sbd *sdp, struct kqid qid,
struct gfs2_quota_data **qdp)
{
@@ -523,13 +552,6 @@
return error;
}
-static void qdsb_put(struct gfs2_quota_data *qd)
-{
- bh_put(qd);
- slot_put(qd);
- qd_put(qd);
-}
-
/**
* gfs2_qa_get - make sure we have a quota allocations data structure,
* if necessary
@@ -666,7 +688,7 @@
static void do_qc(struct gfs2_quota_data *qd, s64 change, int qc_type)
{
- struct gfs2_sbd *sdp = qd->qd_gl->gl_name.ln_sbd;
+ struct gfs2_sbd *sdp = qd->qd_sbd;
struct gfs2_inode *ip = GFS2_I(sdp->sd_qc_inode);
struct gfs2_quota_change *qc = qd->qd_bh_qc;
s64 x;
@@ -708,30 +730,29 @@
mutex_unlock(&sdp->sd_quota_mutex);
}
-static int gfs2_write_buf_to_page(struct gfs2_inode *ip, unsigned long index,
+static int gfs2_write_buf_to_page(struct gfs2_sbd *sdp, unsigned long index,
unsigned off, void *buf, unsigned bytes)
{
+ struct gfs2_inode *ip = GFS2_I(sdp->sd_quota_inode);
struct inode *inode = &ip->i_inode;
- struct gfs2_sbd *sdp = GFS2_SB(inode);
struct address_space *mapping = inode->i_mapping;
struct page *page;
struct buffer_head *bh;
u64 blk;
unsigned bsize = sdp->sd_sb.sb_bsize, bnum = 0, boff = 0;
unsigned to_write = bytes, pg_off = off;
- int done = 0;
blk = index << (PAGE_SHIFT - sdp->sd_sb.sb_bsize_shift);
boff = off % bsize;
- page = find_or_create_page(mapping, index, GFP_NOFS);
+ page = grab_cache_page(mapping, index);
if (!page)
return -ENOMEM;
if (!page_has_buffers(page))
create_empty_buffers(page, bsize, 0);
bh = page_buffers(page);
- while (!done) {
+ for(;;) {
/* Find the beginning block within the page */
if (pg_off >= ((bnum * bsize) + bsize)) {
bh = bh->b_this_page;
@@ -751,10 +772,7 @@
set_buffer_uptodate(bh);
if (bh_read(bh, REQ_META | REQ_PRIO) < 0)
goto unlock_out;
- if (gfs2_is_jdata(ip))
- gfs2_trans_add_data(ip->i_gl, bh);
- else
- gfs2_ordered_add_inode(ip);
+ gfs2_trans_add_data(ip->i_gl, bh);
/* If we need to write to the next block as well */
if (to_write > (bsize - boff)) {
@@ -763,7 +781,7 @@
boff = pg_off % bsize;
continue;
}
- done = 1;
+ break;
}
/* Write to the page, now that we have setup the buffer(s) */
@@ -780,12 +798,12 @@
return -EIO;
}
-static int gfs2_write_disk_quota(struct gfs2_inode *ip, struct gfs2_quota *qp,
+static int gfs2_write_disk_quota(struct gfs2_sbd *sdp, struct gfs2_quota *qp,
loff_t loc)
{
unsigned long pg_beg;
unsigned pg_off, nbytes, overflow = 0;
- int pg_oflow = 0, error;
+ int error;
void *ptr;
nbytes = sizeof(struct gfs2_quota);
@@ -794,17 +812,15 @@
pg_off = offset_in_page(loc);
/* If the quota straddles a page boundary, split the write in two */
- if ((pg_off + nbytes) > PAGE_SIZE) {
- pg_oflow = 1;
+ if ((pg_off + nbytes) > PAGE_SIZE)
overflow = (pg_off + nbytes) - PAGE_SIZE;
- }
ptr = qp;
- error = gfs2_write_buf_to_page(ip, pg_beg, pg_off, ptr,
+ error = gfs2_write_buf_to_page(sdp, pg_beg, pg_off, ptr,
nbytes - overflow);
/* If there's an overflow, write the remaining bytes to the next page */
- if (!error && pg_oflow)
- error = gfs2_write_buf_to_page(ip, pg_beg + 1, 0,
+ if (!error && overflow)
+ error = gfs2_write_buf_to_page(sdp, pg_beg + 1, 0,
ptr + nbytes - overflow,
overflow);
return error;
@@ -812,7 +828,7 @@
/**
* gfs2_adjust_quota - adjust record of current block usage
- * @ip: The quota inode
+ * @sdp: The superblock
* @loc: Offset of the entry in the quota file
* @change: The amount of usage change to record
* @qd: The quota data
@@ -824,12 +840,12 @@
* Returns: 0 or -ve on error
*/
-static int gfs2_adjust_quota(struct gfs2_inode *ip, loff_t loc,
+static int gfs2_adjust_quota(struct gfs2_sbd *sdp, loff_t loc,
s64 change, struct gfs2_quota_data *qd,
struct qc_dqblk *fdq)
{
+ struct gfs2_inode *ip = GFS2_I(sdp->sd_quota_inode);
struct inode *inode = &ip->i_inode;
- struct gfs2_sbd *sdp = GFS2_SB(inode);
struct gfs2_quota q;
int err;
u64 size;
@@ -846,7 +862,6 @@
return err;
loc -= sizeof(q); /* gfs2_internal_read would've advanced the loc ptr */
- err = -EIO;
be64_add_cpu(&q.qu_value, change);
if (((s64)be64_to_cpu(q.qu_value)) < 0)
q.qu_value = 0; /* Never go negative on quota usage */
@@ -866,7 +881,7 @@
}
}
- err = gfs2_write_disk_quota(ip, &q, loc);
+ err = gfs2_write_disk_quota(sdp, &q, loc);
if (!err) {
size = loc + sizeof(struct gfs2_quota);
if (size > inode->i_size)
@@ -881,7 +896,7 @@
static int do_sync(unsigned int num_qd, struct gfs2_quota_data **qda)
{
- struct gfs2_sbd *sdp = (*qda)->qd_gl->gl_name.ln_sbd;
+ struct gfs2_sbd *sdp = (*qda)->qd_sbd;
struct gfs2_inode *ip = GFS2_I(sdp->sd_quota_inode);
struct gfs2_alloc_parms ap = { .aflags = 0, };
unsigned int data_blocks, ind_blocks;
@@ -893,18 +908,12 @@
unsigned int nalloc = 0, blocks;
int error;
- error = gfs2_qa_get(ip);
- if (error)
- return error;
-
gfs2_write_calc_reserv(ip, sizeof(struct gfs2_quota),
&data_blocks, &ind_blocks);
ghs = kmalloc_array(num_qd, sizeof(struct gfs2_holder), GFP_NOFS);
- if (!ghs) {
- error = -ENOMEM;
- goto out;
- }
+ if (!ghs)
+ return -ENOMEM;
sort(qda, num_qd, sizeof(struct gfs2_quota_data *), sort_qd, NULL);
inode_lock(&ip->i_inode);
@@ -953,7 +962,8 @@
for (x = 0; x < num_qd; x++) {
qd = qda[x];
offset = qd2offset(qd);
- error = gfs2_adjust_quota(ip, offset, qd->qd_change_sync, qd, NULL);
+ error = gfs2_adjust_quota(sdp, offset, qd->qd_change_sync, qd,
+ NULL);
if (error)
goto out_end_trans;
@@ -961,8 +971,6 @@
set_bit(QDF_REFRESH, &qd->qd_flags);
}
- error = 0;
-
out_end_trans:
gfs2_trans_end(sdp);
out_ipres:
@@ -976,8 +984,10 @@
kfree(ghs);
gfs2_log_flush(ip->i_gl->gl_name.ln_sbd, ip->i_gl,
GFS2_LOG_HEAD_FLUSH_NORMAL | GFS2_LFC_DO_SYNC);
-out:
- gfs2_qa_put(ip);
+ if (!error) {
+ for (x = 0; x < num_qd; x++)
+ qda[x]->qd_sync_gen = sdp->sd_quota_sync_gen;
+ }
return error;
}
@@ -1009,11 +1019,12 @@
static int do_glock(struct gfs2_quota_data *qd, int force_refresh,
struct gfs2_holder *q_gh)
{
- struct gfs2_sbd *sdp = qd->qd_gl->gl_name.ln_sbd;
+ struct gfs2_sbd *sdp = qd->qd_sbd;
struct gfs2_inode *ip = GFS2_I(sdp->sd_quota_inode);
struct gfs2_holder i_gh;
int error;
+ gfs2_assert_warn(sdp, sdp == qd->qd_gl->gl_name.ln_sbd);
restart:
error = gfs2_glock_nq_init(qd->qd_gl, LM_ST_SHARED, 0, q_gh);
if (error)
@@ -1059,9 +1070,10 @@
struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
struct gfs2_quota_data *qd;
u32 x;
- int error = 0;
+ int error;
- if (sdp->sd_args.ar_quota != GFS2_QUOTA_ON)
+ if (sdp->sd_args.ar_quota != GFS2_QUOTA_ON &&
+ sdp->sd_args.ar_quota != GFS2_QUOTA_QUIET)
return 0;
error = gfs2_quota_hold(ip, uid, gid);
@@ -1089,16 +1101,15 @@
return error;
}
-static int need_sync(struct gfs2_quota_data *qd)
+static bool need_sync(struct gfs2_quota_data *qd)
{
- struct gfs2_sbd *sdp = qd->qd_gl->gl_name.ln_sbd;
+ struct gfs2_sbd *sdp = qd->qd_sbd;
struct gfs2_tune *gt = &sdp->sd_tune;
s64 value;
unsigned int num, den;
- int do_sync = 1;
if (!qd->qd_qb.qb_limit)
- return 0;
+ return false;
spin_lock(&qd_lock);
value = qd->qd_change;
@@ -1109,26 +1120,26 @@
den = gt->gt_quota_scale_den;
spin_unlock(>->gt_spin);
- if (value < 0)
- do_sync = 0;
+ if (value <= 0)
+ return false;
else if ((s64)be64_to_cpu(qd->qd_qb.qb_value) >=
(s64)be64_to_cpu(qd->qd_qb.qb_limit))
- do_sync = 0;
+ return false;
else {
value *= gfs2_jindex_size(sdp) * num;
value = div_s64(value, den);
value += (s64)be64_to_cpu(qd->qd_qb.qb_value);
if (value < (s64)be64_to_cpu(qd->qd_qb.qb_limit))
- do_sync = 0;
+ return false;
}
- return do_sync;
+ return true;
}
void gfs2_quota_unlock(struct gfs2_inode *ip)
{
struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
- struct gfs2_quota_data *qda[4];
+ struct gfs2_quota_data *qda[2 * GFS2_MAXQUOTAS];
unsigned int count = 0;
u32 x;
int found;
@@ -1138,7 +1149,7 @@
for (x = 0; x < ip->i_qadata->qa_qd_num; x++) {
struct gfs2_quota_data *qd;
- int sync;
+ bool sync;
qd = ip->i_qadata->qa_qd[x];
sync = need_sync(qd);
@@ -1154,15 +1165,8 @@
if (!found)
continue;
- gfs2_assert_warn(sdp, qd->qd_change_sync);
- if (bh_get(qd)) {
- clear_bit(QDF_LOCKED, &qd->qd_flags);
- slot_put(qd);
- qd_put(qd);
- continue;
- }
-
- qda[count++] = qd;
+ if (!qd_bh_get_or_undo(sdp, qd))
+ qda[count++] = qd;
}
if (count) {
@@ -1178,12 +1182,13 @@
static int print_message(struct gfs2_quota_data *qd, char *type)
{
- struct gfs2_sbd *sdp = qd->qd_gl->gl_name.ln_sbd;
+ struct gfs2_sbd *sdp = qd->qd_sbd;
- fs_info(sdp, "quota %s for %s %u\n",
- type,
- (qd->qd_id.type == USRQUOTA) ? "user" : "group",
- from_kqid(&init_user_ns, qd->qd_id));
+ if (sdp->sd_args.ar_quota != GFS2_QUOTA_QUIET)
+ fs_info(sdp, "quota %s for %s %u\n",
+ type,
+ (qd->qd_id.type == USRQUOTA) ? "user" : "group",
+ from_kqid(&init_user_ns, qd->qd_id));
return 0;
}
@@ -1269,7 +1274,8 @@
u32 x;
struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
- if (sdp->sd_args.ar_quota != GFS2_QUOTA_ON ||
+ if ((sdp->sd_args.ar_quota != GFS2_QUOTA_ON &&
+ sdp->sd_args.ar_quota != GFS2_QUOTA_QUIET) ||
gfs2_assert_warn(sdp, change))
return;
if (ip->i_diskflags & GFS2_DIF_SYSTEM)
@@ -1288,6 +1294,24 @@
}
}
+static bool qd_changed(struct gfs2_sbd *sdp)
+{
+ struct gfs2_quota_data *qd;
+ bool changed = false;
+
+ spin_lock(&qd_lock);
+ list_for_each_entry(qd, &sdp->sd_quota_list, qd_list) {
+ if (test_bit(QDF_LOCKED, &qd->qd_flags) ||
+ !test_bit(QDF_CHANGE, &qd->qd_flags))
+ continue;
+
+ changed = true;
+ break;
+ }
+ spin_unlock(&qd_lock);
+ return changed;
+}
+
int gfs2_quota_sync(struct super_block *sb, int type)
{
struct gfs2_sbd *sdp = sb->s_fs_info;
@@ -1297,6 +1321,9 @@
unsigned int x;
int error = 0;
+ if (!qd_changed(sdp))
+ return 0;
+
qda = kcalloc(max_qd, sizeof(struct gfs2_quota_data *), GFP_KERNEL);
if (!qda)
return -ENOMEM;
@@ -1318,10 +1345,6 @@
if (num_qd) {
if (!error)
error = do_sync(num_qd, qda);
- if (!error)
- for (x = 0; x < num_qd; x++)
- qda[x]->qd_sync_gen =
- sdp->sd_quota_sync_gen;
for (x = 0; x < num_qd; x++)
qd_unlock(qda[x]);
@@ -1423,7 +1446,7 @@
set_bit(QDF_CHANGE, &qd->qd_flags);
qd->qd_change = qc_change;
qd->qd_slot = slot;
- qd->qd_slot_count = 1;
+ qd->qd_slot_ref = 1;
spin_lock(&qd_lock);
BUG_ON(test_and_set_bit(slot, sdp->sd_quota_bitmap));
@@ -1455,36 +1478,35 @@
void gfs2_quota_cleanup(struct gfs2_sbd *sdp)
{
- struct list_head *head = &sdp->sd_quota_list;
struct gfs2_quota_data *qd;
+ LIST_HEAD(dispose);
+ int count;
+
+ BUG_ON(test_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags));
spin_lock(&qd_lock);
- while (!list_empty(head)) {
- qd = list_last_entry(head, struct gfs2_quota_data, qd_list);
+ list_for_each_entry(qd, &sdp->sd_quota_list, qd_list) {
+ spin_lock(&qd->qd_lockref.lock);
+ if (qd->qd_lockref.count != 0) {
+ spin_unlock(&qd->qd_lockref.lock);
+ continue;
+ }
+ lockref_mark_dead(&qd->qd_lockref);
+ spin_unlock(&qd->qd_lockref.lock);
- list_del(&qd->qd_list);
-
- /* Also remove if this qd exists in the reclaim list */
list_lru_del(&gfs2_qd_lru, &qd->qd_lru);
- atomic_dec(&sdp->sd_quota_count);
- spin_unlock(&qd_lock);
-
- spin_lock_bucket(qd->qd_hash);
- hlist_bl_del_rcu(&qd->qd_hlist);
- spin_unlock_bucket(qd->qd_hash);
-
- gfs2_assert_warn(sdp, !qd->qd_change);
- gfs2_assert_warn(sdp, !qd->qd_slot_count);
- gfs2_assert_warn(sdp, !qd->qd_bh_count);
-
- gfs2_glock_put(qd->qd_gl);
- call_rcu(&qd->qd_rcu, gfs2_qd_dealloc);
-
- spin_lock(&qd_lock);
+ list_add(&qd->qd_lru, &dispose);
}
spin_unlock(&qd_lock);
- gfs2_assert_warn(sdp, !atomic_read(&sdp->sd_quota_count));
+ gfs2_qd_list_dispose(&dispose);
+
+ wait_event_timeout(sdp->sd_kill_wait,
+ (count = atomic_read(&sdp->sd_quota_count)) == 0,
+ HZ * 60);
+
+ if (count != 0)
+ fs_err(sdp, "%d left-over quota data objects\n", count);
kvfree(sdp->sd_quota_bitmap);
sdp->sd_quota_bitmap = NULL;
@@ -1536,12 +1558,11 @@
unsigned long statfs_timeo = 0;
unsigned long quotad_timeo = 0;
unsigned long t = 0;
- DEFINE_WAIT(wait);
while (!kthread_should_stop()) {
-
if (gfs2_withdrawn(sdp))
- goto bypass;
+ break;
+
/* Update the master statfs file */
if (sdp->sd_statfs_force_sync) {
int error = gfs2_statfs_sync(sdp->sd_vfs, 0);
@@ -1559,15 +1580,16 @@
try_to_freeze();
-bypass:
t = min(quotad_timeo, statfs_timeo);
- prepare_to_wait(&sdp->sd_quota_wait, &wait, TASK_INTERRUPTIBLE);
- if (!sdp->sd_statfs_force_sync)
- t -= schedule_timeout(t);
- else
+ t = wait_event_interruptible_timeout(sdp->sd_quota_wait,
+ sdp->sd_statfs_force_sync ||
+ gfs2_withdrawn(sdp) ||
+ kthread_should_stop(),
+ t);
+
+ if (sdp->sd_statfs_force_sync)
t = 0;
- finish_wait(&sdp->sd_quota_wait, &wait);
}
return 0;
@@ -1580,6 +1602,8 @@
memset(state, 0, sizeof(*state));
switch (sdp->sd_args.ar_quota) {
+ case GFS2_QUOTA_QUIET:
+ fallthrough;
case GFS2_QUOTA_ON:
state->s_state[USRQUOTA].flags |= QCI_LIMITS_ENFORCED;
state->s_state[GRPQUOTA].flags |= QCI_LIMITS_ENFORCED;
@@ -1726,7 +1750,7 @@
goto out_release;
/* Apply changes */
- error = gfs2_adjust_quota(ip, offset, 0, qd, fdq);
+ error = gfs2_adjust_quota(sdp, offset, 0, qd, fdq);
if (!error)
clear_bit(QDF_QMSG_QUIET, &qd->qd_flags);
diff --git a/fs/gfs2/recovery.c b/fs/gfs2/recovery.c
index 9c7a9f6..5aae026 100644
--- a/fs/gfs2/recovery.c
+++ b/fs/gfs2/recovery.c
@@ -27,7 +27,7 @@
#include "util.h"
#include "dir.h"
-struct workqueue_struct *gfs_recovery_wq;
+struct workqueue_struct *gfs2_recovery_wq;
int gfs2_replay_read_block(struct gfs2_jdesc *jd, unsigned int blk,
struct buffer_head **bh)
@@ -570,7 +570,7 @@
return -EBUSY;
/* we have JDF_RECOVERY, queue should always succeed */
- rv = queue_work(gfs_recovery_wq, &jd->jd_work);
+ rv = queue_work(gfs2_recovery_wq, &jd->jd_work);
BUG_ON(!rv);
if (wait)
diff --git a/fs/gfs2/recovery.h b/fs/gfs2/recovery.h
index 0d30f8e..7a0c9d0 100644
--- a/fs/gfs2/recovery.h
+++ b/fs/gfs2/recovery.h
@@ -9,7 +9,7 @@
#include "incore.h"
-extern struct workqueue_struct *gfs_recovery_wq;
+extern struct workqueue_struct *gfs2_recovery_wq;
static inline void gfs2_replay_incr_blk(struct gfs2_jdesc *jd, u32 *blk)
{
diff --git a/fs/gfs2/super.c b/fs/gfs2/super.c
index 2f70133..02d93da 100644
--- a/fs/gfs2/super.c
+++ b/fs/gfs2/super.c
@@ -546,20 +546,10 @@
{
int log_write_allowed = test_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags);
- if (!test_bit(SDF_DEACTIVATING, &sdp->sd_flags))
+ if (!test_bit(SDF_KILL, &sdp->sd_flags))
gfs2_flush_delete_work(sdp);
- if (!log_write_allowed && current == sdp->sd_quotad_process)
- fs_warn(sdp, "The quotad daemon is withdrawing.\n");
- else if (sdp->sd_quotad_process)
- kthread_stop(sdp->sd_quotad_process);
- sdp->sd_quotad_process = NULL;
-
- if (!log_write_allowed && current == sdp->sd_logd_process)
- fs_warn(sdp, "The logd daemon is withdrawing.\n");
- else if (sdp->sd_logd_process)
- kthread_stop(sdp->sd_logd_process);
- sdp->sd_logd_process = NULL;
+ gfs2_destroy_threads(sdp);
if (log_write_allowed) {
gfs2_quota_sync(sdp->sd_vfs, 0);
@@ -580,15 +570,8 @@
gfs2_log_is_empty(sdp),
HZ * 5);
gfs2_assert_warn(sdp, gfs2_log_is_empty(sdp));
- } else {
- wait_event_timeout(sdp->sd_log_waitq,
- gfs2_log_is_empty(sdp),
- HZ * 5);
}
gfs2_quota_cleanup(sdp);
-
- if (!log_write_allowed)
- sdp->sd_vfs->s_flags |= SB_RDONLY;
}
/**
@@ -622,6 +605,10 @@
if (!sb_rdonly(sb)) {
gfs2_make_fs_ro(sdp);
}
+ if (gfs2_withdrawn(sdp)) {
+ gfs2_destroy_threads(sdp);
+ gfs2_quota_cleanup(sdp);
+ }
WARN_ON(gfs2_withdrawing(sdp));
/* At this point, we're through modifying the disk */
@@ -1134,6 +1121,9 @@
case GFS2_QUOTA_ON:
state = "on";
break;
+ case GFS2_QUOTA_QUIET:
+ state = "quiet";
+ break;
default:
state = "unknown";
break;
diff --git a/fs/gfs2/super.h b/fs/gfs2/super.h
index bba5862..ab9c831 100644
--- a/fs/gfs2/super.h
+++ b/fs/gfs2/super.h
@@ -36,6 +36,7 @@
extern int gfs2_make_fs_rw(struct gfs2_sbd *sdp);
extern void gfs2_make_fs_ro(struct gfs2_sbd *sdp);
extern void gfs2_online_uevent(struct gfs2_sbd *sdp);
+extern void gfs2_destroy_threads(struct gfs2_sbd *sdp);
extern int gfs2_statfs_init(struct gfs2_sbd *sdp);
extern void gfs2_statfs_change(struct gfs2_sbd *sdp, s64 total, s64 free,
s64 dinodes);
diff --git a/fs/gfs2/sys.c b/fs/gfs2/sys.c
index c60bc7f..60a0206 100644
--- a/fs/gfs2/sys.c
+++ b/fs/gfs2/sys.c
@@ -98,7 +98,10 @@
"sd_log_flush_head: %d\n"
"sd_log_flush_tail: %d\n"
"sd_log_blks_reserved: %d\n"
- "sd_log_revokes_available: %d\n",
+ "sd_log_revokes_available: %d\n"
+ "sd_log_pinned: %d\n"
+ "sd_log_thresh1: %d\n"
+ "sd_log_thresh2: %d\n",
test_bit(SDF_JOURNAL_CHECKED, &f),
test_bit(SDF_JOURNAL_LIVE, &f),
(sdp->sd_jdesc ? sdp->sd_jdesc->jd_jid : 0),
@@ -118,7 +121,7 @@
test_bit(SDF_WITHDRAW_IN_PROG, &f),
test_bit(SDF_REMOTE_WITHDRAW, &f),
test_bit(SDF_WITHDRAW_RECOVERY, &f),
- test_bit(SDF_DEACTIVATING, &f),
+ test_bit(SDF_KILL, &f),
sdp->sd_log_error,
rwsem_is_locked(&sdp->sd_log_flush_lock),
sdp->sd_log_num_revoke,
@@ -128,7 +131,10 @@
sdp->sd_log_flush_head,
sdp->sd_log_flush_tail,
sdp->sd_log_blks_reserved,
- atomic_read(&sdp->sd_log_revokes_available));
+ atomic_read(&sdp->sd_log_revokes_available),
+ atomic_read(&sdp->sd_log_pinned),
+ atomic_read(&sdp->sd_log_thresh1),
+ atomic_read(&sdp->sd_log_thresh2));
return s;
}
diff --git a/fs/gfs2/util.c b/fs/gfs2/util.c
index dac22b1..da29faf 100644
--- a/fs/gfs2/util.c
+++ b/fs/gfs2/util.c
@@ -9,6 +9,7 @@
#include <linux/spinlock.h>
#include <linux/completion.h>
#include <linux/buffer_head.h>
+#include <linux/kthread.h>
#include <linux/crc32.h>
#include <linux/gfs2_ondisk.h>
#include <linux/delay.h>
@@ -150,7 +151,14 @@
if (!sb_rdonly(sdp->sd_vfs)) {
bool locked = mutex_trylock(&sdp->sd_freeze_mutex);
- gfs2_make_fs_ro(sdp);
+ wake_up(&sdp->sd_logd_waitq);
+ wake_up(&sdp->sd_quota_wait);
+
+ wait_event_timeout(sdp->sd_log_waitq,
+ gfs2_log_is_empty(sdp),
+ HZ * 5);
+
+ sdp->sd_vfs->s_flags |= SB_RDONLY;
if (locked)
mutex_unlock(&sdp->sd_freeze_mutex);
@@ -315,19 +323,19 @@
struct lm_lockstruct *ls = &sdp->sd_lockstruct;
const struct lm_lockops *lm = ls->ls_ops;
- if (sdp->sd_args.ar_errors == GFS2_ERRORS_WITHDRAW &&
- test_and_set_bit(SDF_WITHDRAWN, &sdp->sd_flags)) {
- if (!test_bit(SDF_WITHDRAW_IN_PROG, &sdp->sd_flags))
- return -1;
-
- wait_on_bit(&sdp->sd_flags, SDF_WITHDRAW_IN_PROG,
- TASK_UNINTERRUPTIBLE);
- return -1;
- }
-
- set_bit(SDF_WITHDRAW_IN_PROG, &sdp->sd_flags);
-
if (sdp->sd_args.ar_errors == GFS2_ERRORS_WITHDRAW) {
+ unsigned long old = READ_ONCE(sdp->sd_flags), new;
+
+ do {
+ if (old & BIT(SDF_WITHDRAWN)) {
+ wait_on_bit(&sdp->sd_flags,
+ SDF_WITHDRAW_IN_PROG,
+ TASK_UNINTERRUPTIBLE);
+ return -1;
+ }
+ new = old | BIT(SDF_WITHDRAWN) | BIT(SDF_WITHDRAW_IN_PROG);
+ } while (unlikely(!try_cmpxchg(&sdp->sd_flags, &old, new)));
+
fs_err(sdp, "about to withdraw this file system\n");
BUG_ON(sdp->sd_args.ar_debug);
