superrob1500
/
mirror-linux
mirror of https://github.com/torvalds/linux
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

gfs2 changes 

- Major withdraw / error handling overhaul based on dlm's new
   DLM_RELEASE_RECOVER feature: this allows gfs to treat withdraws like
   node failures.  Make withdraws asynchronous.
 
 - Fix a bug in commit e4a8b5481c that caused 'df' to remain out of
   sync.  ('df' is still allowed to go slightly out of sync for short
   periods of time.)
 
 - Prevent recusive memory reclaim in gfs2_unstuff_dinode().
 
 - Clean up SDF_JOURNAL_LIVE flag handling.
 
 - Fix remote evict for read-only filesystems.
 
 - Fix a misuse of bio_chain().
 
 - Various other minor cleanups.
 -----BEGIN PGP SIGNATURE-----
 
 iQJIBAABCAAyFiEEJZs3krPW0xkhLMTc1b+f6wMTZToFAmkvM/oUHGFncnVlbmJh
 QHJlZGhhdC5jb20ACgkQ1b+f6wMTZTqFLQ//Ra+FUIMYRrU83xd8OB+rXaFixymk
 riPnoXZHrccoRAARc6d/p/nTYmV7FSnTa2jlsUFshsO2jl/1OIRI6xqScL+W5dp9
 ToxHKnezOsRmy5tXvomYy9Teo/ngz6dpv6ixVrn7qTGJJ5zpir0ji2fO+xUvCUQq
 IVGgvQN0O7ECUmmRXxjkRLhoEdus2Ye4jQkpJmqY9I1+H/0E43HlF2zDgoisTkiH
 baNtFkTgE65Hba2xhoTCt3NO17POgqGcRaLVOLv5dqGhLAQ33+jXeYduM6ujbqwk
 1FznGf09NRBJVQkyJvbgZELdX+mWXvWf/2kPmW6TXd7phFiO7cByxN2aHHcu/wBp
 kh0ZgLFtvjyuR7wiyKf0tAKKPh250YiSkBuJU64dlHha2ZQuME6Z4dlYqBKwvPXf
 AwxAu0tB2bAjc2OXvPaHhLHjdpySEWqNZIITbfQPVPonRKed10IqCwlTd+j5/64H
 mnZCCxr2p2akrSYbQjl7h1k4LO7eX3cYJ23JoKIjwQ2hJE+Pd4li9+xhQCEQUVyE
 ou/ezUKc19Xe7uxG1412PrbmOtIMgKfpq2KW151b1KA3pfCUt7GMvpYlvmnf9Wzs
 XGjrQ1p4XdlxzQqe9XTrdO0IrNUNGDcky1Kqishje9SPa8FthwrtTuBSYL8nkcX6
 3KRRzlFCRHfM5ng=
 =AWK2
 -----END PGP SIGNATURE-----

Merge tag 'gfs2-for-6.19' of git://git.kernel.org/pub/scm/linux/kernel/git/gfs2/linux-gfs2

Pull gfs2 updates from Andreas Gruenbacher:

 - Major withdraw / error handling overhaul based on dlm's new
   DLM_RELEASE_RECOVER feature: this allows gfs to treat withdraws like
   node failures. Make withdraws asynchronous

 - Fix a bug in commit e4a8b5481c that caused 'df' to remain out of
   sync. ('df' is still allowed to go slightly out of sync for short
   periods of time)

 - Prevent recusive memory reclaim in gfs2_unstuff_dinode()

 - Clean up SDF_JOURNAL_LIVE flag handling

 - Fix remote evict for read-only filesystems

 - Fix a misuse of bio_chain()

 - Various other minor cleanups

* tag 'gfs2-for-6.19' of git://git.kernel.org/pub/scm/linux/kernel/git/gfs2/linux-gfs2: (35 commits)
  gfs2: Fix use of bio_chain
  gfs2: Clean up SDF_JOURNAL_LIVE flag handling
  gfs2: No longer thaw filesystems during a withdraw
  gfs2: Withdraw immediately in gfs2_trans_add_meta
  gfs2: New gfs2_withdraw_helper
  gfs2: Clean up properly during a withdraw
  gfs2: Rename gfs2_{gl_dq_holders => withdraw_glocks}
  Revert "gfs2: fix infinite loop when checking ail item count before go_inval"
  Revert "gfs2: Allow some glocks to be used during withdraw"
  Revert "gfs2: Check for log write errors before telling dlm to unlock"
  Revert "gfs2: fix a deadlock on withdraw-during-mount"
  Revert "gfs2: Force withdraw to replay journals and wait for it to finish" (6/6)
  Revert "gfs2: Force withdraw to replay journals and wait for it to finish" (5/6)
  Revert "gfs2: Force withdraw to replay journals and wait for it to finish" (4/6)
  Revert "gfs2: Force withdraw to replay journals and wait for it to finish" (3/6)
  Revert "gfs2: Force withdraw to replay journals and wait for it to finish" (2/6)
  Revert "gfs2: Force withdraw to replay journals and wait for it to finish" (1/6)
  Revert "gfs2: don't stop reads while withdraw in progress"
  gfs2: Rename LM_FLAG_{NOEXP -> RECOVER}
  gfs2: Kill gfs2_io_error_bh_wd
  ...
pull/1354/merge
Linus Torvalds 2025-12-03 20:28:50 -08:00
parent 869737543b 8a157e0a0a
commit afcbce74f3
27 changed files with 397 additions and 769 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

0
Documentation/filesystems/gfs2-glocks.rst → Documentation/filesystems/gfs2/glocks.rst
View File

12
Documentation/filesystems/gfs2.rst → Documentation/filesystems/gfs2/index.rst
View File

@ -4,6 +4,9 @@
Global File System 2 Global File System 2
==================== ====================
Overview
========
GFS2 is a cluster file system. It allows a cluster of computers to GFS2 is a cluster file system. It allows a cluster of computers to
simultaneously use a block device that is shared between them (with FC, simultaneously use a block device that is shared between them (with FC,
iSCSI, NBD, etc). GFS2 reads and writes to the block device like a local iSCSI, NBD, etc). GFS2 reads and writes to the block device like a local
@ -50,3 +53,12 @@ The following man pages are available from gfs2-utils:
gfs2_convert to convert a gfs filesystem to GFS2 in-place gfs2_convert to convert a gfs filesystem to GFS2 in-place
mkfs.gfs2 to make a filesystem mkfs.gfs2 to make a filesystem
============ ============================================= ============ =============================================
Implementation Notes
====================
.. toctree::
:maxdepth: 1
glocks
uevents

0
Documentation/filesystems/gfs2-uevents.rst → Documentation/filesystems/gfs2/uevents.rst
View File

4
Documentation/filesystems/index.rst
View File

@ -89,9 +89,7 @@ Documentation for filesystem implementations.
ext3 ext3
ext4/index ext4/index
f2fs f2fs
gfs2 gfs2/index
gfs2-uevents
gfs2-glocks
hfs hfs
hfsplus hfsplus
hpfs hpfs

2
MAINTAINERS
View File

@ -10535,7 +10535,7 @@ L: gfs2@lists.linux.dev
S: Supported S: Supported
B: https://bugzilla.kernel.org/enter_bug.cgi?product=File%20System&component=gfs2 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 T: git git://git.kernel.org/pub/scm/linux/kernel/git/gfs2/linux-gfs2.git
F: Documentation/filesystems/gfs2* F: Documentation/filesystems/gfs2/
F: fs/gfs2/ F: fs/gfs2/
F: include/uapi/linux/gfs2_ondisk.h F: include/uapi/linux/gfs2_ondisk.h

2
fs/gfs2/aops.c
View File

@ -431,7 +431,7 @@ static int gfs2_read_folio(struct file *file, struct folio *folio)
error = mpage_read_folio(folio, gfs2_block_map); error = mpage_read_folio(folio, gfs2_block_map);
} }
if (gfs2_withdrawing_or_withdrawn(sdp)) if (gfs2_withdrawn(sdp))
return -EIO; return -EIO;
return error; return error;

2
fs/gfs2/file.c
View File

@ -1446,7 +1446,7 @@ static int gfs2_lock(struct file *file, int cmd, struct file_lock *fl)
if (!(fl->c.flc_flags & FL_POSIX)) if (!(fl->c.flc_flags & FL_POSIX))
return -ENOLCK; return -ENOLCK;
if (gfs2_withdrawing_or_withdrawn(sdp)) { if (gfs2_withdrawn(sdp)) {
if (lock_is_unlock(fl)) if (lock_is_unlock(fl))
locks_lock_file_wait(file, fl); locks_lock_file_wait(file, fl);
return -EIO; return -EIO;

227
fs/gfs2/glock.c
View File

@ -137,33 +137,6 @@ static void gfs2_glock_dealloc(struct rcu_head *rcu)
kmem_cache_free(gfs2_glock_cachep, gl); kmem_cache_free(gfs2_glock_cachep, gl);
} }
/**
* glock_blocked_by_withdraw - determine if we can still use a glock
* @gl: the glock
*
* We need to allow some glocks to be enqueued, dequeued, promoted, and demoted
* when we're withdrawn. For example, to maintain metadata integrity, we should
* disallow the use of inode and rgrp glocks when withdrawn. Other glocks like
* the iopen or freeze glock may be safely used because none of their
* metadata goes through the journal. So in general, we should disallow all
* glocks that are journaled, and allow all the others. One exception is:
* we need to allow our active journal to be promoted and demoted so others
* may recover it and we can reacquire it when they're done.
*/
static bool glock_blocked_by_withdraw(struct gfs2_glock *gl)
{
struct gfs2_sbd *sdp = gl->gl_name.ln_sbd;
if (!gfs2_withdrawing_or_withdrawn(sdp))
return false;
if (gl->gl_ops->go_flags & GLOF_NONDISK)
return false;
if (!sdp->sd_jdesc ||
gl->gl_name.ln_number == sdp->sd_jdesc->jd_no_addr)
return false;
return true;
}
static void __gfs2_glock_free(struct gfs2_glock *gl) static void __gfs2_glock_free(struct gfs2_glock *gl)
{ {
rhashtable_remove_fast(&gl_hash_table, &gl->gl_node, ht_parms); rhashtable_remove_fast(&gl_hash_table, &gl->gl_node, ht_parms);
@ -270,7 +243,7 @@ static void __gfs2_glock_put(struct gfs2_glock *gl)
GLOCK_BUG_ON(gl, !list_empty(&gl->gl_holders)); GLOCK_BUG_ON(gl, !list_empty(&gl->gl_holders));
if (mapping) { if (mapping) {
truncate_inode_pages_final(mapping); truncate_inode_pages_final(mapping);
if (!gfs2_withdrawing_or_withdrawn(sdp)) if (!gfs2_withdrawn(sdp))
GLOCK_BUG_ON(gl, !mapping_empty(mapping)); GLOCK_BUG_ON(gl, !mapping_empty(mapping));
} }
trace_gfs2_glock_put(gl); trace_gfs2_glock_put(gl);
@ -485,8 +458,14 @@ done:
static void do_promote(struct gfs2_glock *gl) static void do_promote(struct gfs2_glock *gl)
{ {
struct gfs2_sbd *sdp = gl->gl_name.ln_sbd;
struct gfs2_holder *gh, *current_gh; struct gfs2_holder *gh, *current_gh;
if (gfs2_withdrawn(sdp)) {
do_error(gl, LM_OUT_ERROR);
return;
}
current_gh = find_first_holder(gl); current_gh = find_first_holder(gl);
list_for_each_entry(gh, &gl->gl_holders, gh_list) { list_for_each_entry(gh, &gl->gl_holders, gh_list) {
if (test_bit(HIF_HOLDER, &gh->gh_iflags)) if (test_bit(HIF_HOLDER, &gh->gh_iflags))
@ -592,7 +571,6 @@ static void finish_xmote(struct gfs2_glock *gl, unsigned int ret)
state_change(gl, state); state_change(gl, state);
} }
/* Demote to UN request arrived during demote to SH or DF */ /* Demote to UN request arrived during demote to SH or DF */
if (test_bit(GLF_DEMOTE_IN_PROGRESS, &gl->gl_flags) && if (test_bit(GLF_DEMOTE_IN_PROGRESS, &gl->gl_flags) &&
gl->gl_state != LM_ST_UNLOCKED && gl->gl_state != LM_ST_UNLOCKED &&
@ -663,16 +641,6 @@ out:
clear_bit(GLF_LOCK, &gl->gl_flags); clear_bit(GLF_LOCK, &gl->gl_flags);
} }
static bool is_system_glock(struct gfs2_glock *gl)
{
struct gfs2_sbd *sdp = gl->gl_name.ln_sbd;
struct gfs2_inode *m_ip = GFS2_I(sdp->sd_statfs_inode);
if (gl == m_ip->i_gl)
return true;
return false;
}
/** /**
* do_xmote - Calls the DLM to change the state of a lock * do_xmote - Calls the DLM to change the state of a lock
* @gl: The lock state * @gl: The lock state
@ -691,95 +659,47 @@ __acquires(&gl->gl_lockref.lock)
struct lm_lockstruct *ls = &sdp->sd_lockstruct; struct lm_lockstruct *ls = &sdp->sd_lockstruct;
int ret; int ret;
if (target != LM_ST_UNLOCKED && glock_blocked_by_withdraw(gl) && /*
gh && !(gh->gh_flags & LM_FLAG_NOEXP)) * When a filesystem is withdrawing, the remaining cluster nodes will
goto skip_inval; * take care of recovering the withdrawing node's journal. We only
* need to make sure that once we trigger remote recovery, we won't
* write to the shared block device anymore. This means that here,
*
* - no new writes to the filesystem must be triggered (->go_sync()).
*
* - any cached data should be discarded by calling ->go_inval(), dirty
* or not and journaled or unjournaled.
*
* - no more dlm locking operations should be issued (->lm_lock()).
*/
GLOCK_BUG_ON(gl, gl->gl_state == target); GLOCK_BUG_ON(gl, gl->gl_state == target);
GLOCK_BUG_ON(gl, gl->gl_state == gl->gl_target); GLOCK_BUG_ON(gl, gl->gl_state == gl->gl_target);
if (!glops->go_inval || !glops->go_sync) if (!glops->go_inval || !glops->go_sync)
goto skip_inval; goto skip_inval;
spin_unlock(&gl->gl_lockref.lock); spin_unlock(&gl->gl_lockref.lock);
ret = glops->go_sync(gl); if (!gfs2_withdrawn(sdp)) {
/* If we had a problem syncing (due to io errors or whatever, ret = glops->go_sync(gl);
* we should not invalidate the metadata or tell dlm to if (ret) {
* release the glock to other nodes. if (cmpxchg(&sdp->sd_log_error, 0, ret)) {
*/ fs_err(sdp, "Error %d syncing glock\n", ret);
if (ret) { gfs2_dump_glock(NULL, gl, true);
if (cmpxchg(&sdp->sd_log_error, 0, ret)) { gfs2_withdraw(sdp);
fs_err(sdp, "Error %d syncing glock\n", ret); }
gfs2_dump_glock(NULL, gl, true);
} }
spin_lock(&gl->gl_lockref.lock);
goto skip_inval;
} }
if (target == LM_ST_UNLOCKED || target == LM_ST_DEFERRED) { if (target == LM_ST_UNLOCKED || target == LM_ST_DEFERRED)
/*
* The call to go_sync should have cleared out the ail list.
* If there are still items, we have a problem. We ought to
* withdraw, but we can't because the withdraw code also uses
* glocks. Warn about the error, dump the glock, then fall
* through and wait for logd to do the withdraw for us.
*/
if ((atomic_read(&gl->gl_ail_count) != 0) &&
(!cmpxchg(&sdp->sd_log_error, 0, -EIO))) {
gfs2_glock_assert_warn(gl,
!atomic_read(&gl->gl_ail_count));
gfs2_dump_glock(NULL, gl, true);
}
glops->go_inval(gl, target == LM_ST_DEFERRED ? 0 : DIO_METADATA); glops->go_inval(gl, target == LM_ST_DEFERRED ? 0 : DIO_METADATA);
}
spin_lock(&gl->gl_lockref.lock); spin_lock(&gl->gl_lockref.lock);
skip_inval: skip_inval:
/* if (gfs2_withdrawn(sdp)) {
* Check for an error encountered since we called go_sync and go_inval. if (target != LM_ST_UNLOCKED)
* If so, we can't withdraw from the glock code because the withdraw target = LM_OUT_ERROR;
* code itself uses glocks (see function signal_our_withdraw) to goto out;
* change the mount to read-only. Most importantly, we must not call
* dlm to unlock the glock until the journal is in a known good state
* (after journal replay) otherwise other nodes may use the object
* (rgrp or dinode) and then later, journal replay will corrupt the
* file system. The best we can do here is wait for the logd daemon
* to see sd_log_error and withdraw, and in the meantime, requeue the
* work for later.
*
* We make a special exception for some system glocks, such as the
* system statfs inode glock, which needs to be granted before the
* gfs2_quotad daemon can exit, and that exit needs to finish before
* we can unmount the withdrawn file system.
*
* However, if we're just unlocking the lock (say, for unmount, when
* gfs2_gl_hash_clear calls clear_glock) and recovery is complete
* then it's okay to tell dlm to unlock it.
*/
if (unlikely(sdp->sd_log_error) && !gfs2_withdrawing_or_withdrawn(sdp))
gfs2_withdraw_delayed(sdp);
if (glock_blocked_by_withdraw(gl) &&
(target != LM_ST_UNLOCKED ||
test_bit(SDF_WITHDRAW_RECOVERY, &sdp->sd_flags))) {
if (!is_system_glock(gl)) {
request_demote(gl, LM_ST_UNLOCKED, 0, false);
/*
* Ordinarily, we would call dlm and its callback would call
* finish_xmote, which would call state_change() to the new state.
* Since we withdrew, we won't call dlm, so call state_change
* manually, but to the UNLOCKED state we desire.
*/
state_change(gl, LM_ST_UNLOCKED);
/*
* We skip telling dlm to do the locking, so we won't get a
* reply that would otherwise clear GLF_LOCK. So we clear it here.
*/
if (!test_bit(GLF_CANCELING, &gl->gl_flags))
clear_bit(GLF_LOCK, &gl->gl_flags);
clear_bit(GLF_DEMOTE_IN_PROGRESS, &gl->gl_flags);
gl->gl_lockref.count++;
gfs2_glock_queue_work(gl, GL_GLOCK_DFT_HOLD);
return;
}
} }
if (ls->ls_ops->lm_lock) { if (ls->ls_ops->lm_lock) {
@ -795,19 +715,23 @@ skip_inval:
} }
clear_bit(GLF_PENDING_REPLY, &gl->gl_flags); clear_bit(GLF_PENDING_REPLY, &gl->gl_flags);
if (ret == -ENODEV && gl->gl_target == LM_ST_UNLOCKED && if (ret == -ENODEV) {
target == LM_ST_UNLOCKED) {
/* /*
* The lockspace has been released and the lock has * The lockspace has been released and the lock has
* been unlocked implicitly. * been unlocked implicitly.
*/ */
if (target != LM_ST_UNLOCKED) {
target = LM_OUT_ERROR;
goto out;
}
} else { } else {
fs_err(sdp, "lm_lock ret %d\n", ret); fs_err(sdp, "lm_lock ret %d\n", ret);
GLOCK_BUG_ON(gl, !gfs2_withdrawing_or_withdrawn(sdp)); GLOCK_BUG_ON(gl, !gfs2_withdrawn(sdp));
return; return;
} }
} }
out:
/* Complete the operation now. */ /* Complete the operation now. */
finish_xmote(gl, target); finish_xmote(gl, target);
gl->gl_lockref.count++; gl->gl_lockref.count++;
@ -966,14 +890,14 @@ static struct gfs2_inode *gfs2_grab_existing_inode(struct gfs2_glock *gl)
return ip; return ip;
} }
static void gfs2_try_evict(struct gfs2_glock *gl) static void gfs2_try_to_evict(struct gfs2_glock *gl)
{ {
struct gfs2_inode *ip; struct gfs2_inode *ip;
/* /*
* If there is contention on the iopen glock and we have an inode, try * If there is contention on the iopen glock and we have an inode, try
* to grab and release the inode so that it can be evicted. The * to grab and release the inode so that it can be evicted. The
* GIF_DEFER_DELETE flag indicates to gfs2_evict_inode() that the inode * GLF_DEFER_DELETE flag indicates to gfs2_evict_inode() that the inode
* should not be deleted locally. This will allow the remote node to * should not be deleted locally. This will allow the remote node to
* go ahead and delete the inode without us having to do it, which will * go ahead and delete the inode without us having to do it, which will
* avoid rgrp glock thrashing. * avoid rgrp glock thrashing.
@ -1026,8 +950,14 @@ static void delete_work_func(struct work_struct *work)
struct gfs2_sbd *sdp = gl->gl_name.ln_sbd; struct gfs2_sbd *sdp = gl->gl_name.ln_sbd;
bool verify_delete = test_and_clear_bit(GLF_VERIFY_DELETE, &gl->gl_flags); bool verify_delete = test_and_clear_bit(GLF_VERIFY_DELETE, &gl->gl_flags);
/*
* Check for the GLF_VERIFY_DELETE above: this ensures that we won't
* immediately process GLF_VERIFY_DELETE work that the below call to
* gfs2_try_to_evict() queues.
*/
if (test_and_clear_bit(GLF_TRY_TO_EVICT, &gl->gl_flags)) if (test_and_clear_bit(GLF_TRY_TO_EVICT, &gl->gl_flags))
gfs2_try_evict(gl); gfs2_try_to_evict(gl);
if (verify_delete) { if (verify_delete) {
u64 no_addr = gl->gl_name.ln_number; u64 no_addr = gl->gl_name.ln_number;
@ -1211,10 +1141,13 @@ int gfs2_glock_get(struct gfs2_sbd *sdp, u64 number,
mapping = gfs2_glock2aspace(gl); mapping = gfs2_glock2aspace(gl);
if (mapping) { if (mapping) {
gfp_t gfp_mask;
mapping->a_ops = &gfs2_meta_aops; mapping->a_ops = &gfs2_meta_aops;
mapping->host = sdp->sd_inode; mapping->host = sdp->sd_inode;
mapping->flags = 0; mapping->flags = 0;
mapping_set_gfp_mask(mapping, GFP_NOFS); gfp_mask = mapping_gfp_mask(sdp->sd_inode->i_mapping);
mapping_set_gfp_mask(mapping, gfp_mask);
mapping->i_private_data = NULL; mapping->i_private_data = NULL;
mapping->writeback_index = 0; mapping->writeback_index = 0;
} }
@ -1241,7 +1174,7 @@ found:
* @state: the state we're requesting * @state: the state we're requesting
* @flags: the modifier flags * @flags: the modifier flags
* @gh: the holder structure * @gh: the holder structure
* * @ip: caller's return address for debugging
*/ */
void __gfs2_holder_init(struct gfs2_glock *gl, unsigned int state, u16 flags, void __gfs2_holder_init(struct gfs2_glock *gl, unsigned int state, u16 flags,
@ -1539,9 +1472,10 @@ trap_recursive:
int gfs2_glock_nq(struct gfs2_holder *gh) int gfs2_glock_nq(struct gfs2_holder *gh)
{ {
struct gfs2_glock *gl = gh->gh_gl; struct gfs2_glock *gl = gh->gh_gl;
struct gfs2_sbd *sdp = gl->gl_name.ln_sbd;
int error; int error;
if (glock_blocked_by_withdraw(gl) && !(gh->gh_flags & LM_FLAG_NOEXP)) if (gfs2_withdrawn(sdp))
return -EIO; return -EIO;
if (gh->gh_flags & GL_NOBLOCK) { if (gh->gh_flags & GL_NOBLOCK) {
@ -1566,7 +1500,7 @@ unlock:
gh->gh_error = 0; gh->gh_error = 0;
spin_lock(&gl->gl_lockref.lock); spin_lock(&gl->gl_lockref.lock);
add_to_queue(gh); add_to_queue(gh);
if (unlikely((LM_FLAG_NOEXP & gh->gh_flags) && if (unlikely((LM_FLAG_RECOVER & gh->gh_flags) &&
test_and_clear_bit(GLF_HAVE_FROZEN_REPLY, &gl->gl_flags))) { test_and_clear_bit(GLF_HAVE_FROZEN_REPLY, &gl->gl_flags))) {
set_bit(GLF_HAVE_REPLY, &gl->gl_flags); set_bit(GLF_HAVE_REPLY, &gl->gl_flags);
gl->gl_lockref.count++; gl->gl_lockref.count++;
@ -1639,7 +1573,6 @@ static void __gfs2_glock_dq(struct gfs2_holder *gh)
void gfs2_glock_dq(struct gfs2_holder *gh) void gfs2_glock_dq(struct gfs2_holder *gh)
{ {
struct gfs2_glock *gl = gh->gh_gl; struct gfs2_glock *gl = gh->gh_gl;
struct gfs2_sbd *sdp = gl->gl_name.ln_sbd;
spin_lock(&gl->gl_lockref.lock); spin_lock(&gl->gl_lockref.lock);
if (!gfs2_holder_queued(gh)) { if (!gfs2_holder_queued(gh)) {
@ -1666,24 +1599,6 @@ void gfs2_glock_dq(struct gfs2_holder *gh)
goto out; goto out;
} }
/*
* If we're in the process of file system withdraw, we cannot just
* dequeue any glocks until our journal is recovered, lest we introduce
* file system corruption. We need two exceptions to this rule: We need
* to allow unlocking of nondisk glocks and the glock for our own
* journal that needs recovery.
*/
if (test_bit(SDF_WITHDRAW_RECOVERY, &sdp->sd_flags) &&
glock_blocked_by_withdraw(gl) &&
gh->gh_gl != sdp->sd_jinode_gl) {
sdp->sd_glock_dqs_held++;
spin_unlock(&gl->gl_lockref.lock);
might_sleep();
wait_on_bit(&sdp->sd_flags, SDF_WITHDRAW_RECOVERY,
TASK_UNINTERRUPTIBLE);
spin_lock(&gl->gl_lockref.lock);
}
__gfs2_glock_dq(gh); __gfs2_glock_dq(gh);
out: out:
spin_unlock(&gl->gl_lockref.lock); spin_unlock(&gl->gl_lockref.lock);
@ -1871,7 +1786,7 @@ void gfs2_glock_cb(struct gfs2_glock *gl, unsigned int state)
* *
* Glocks are not frozen if (a) the result of the dlm operation is * Glocks are not frozen if (a) the result of the dlm operation is
* an error, (b) the locking operation was an unlock operation or * an error, (b) the locking operation was an unlock operation or
* (c) if there is a "noexp" flagged request anywhere in the queue * (c) if there is a "recover" flagged request anywhere in the queue
* *
* Returns: 1 if freezing should occur, 0 otherwise * Returns: 1 if freezing should occur, 0 otherwise
*/ */
@ -1888,7 +1803,7 @@ static int gfs2_should_freeze(const struct gfs2_glock *gl)
list_for_each_entry(gh, &gl->gl_holders, gh_list) { list_for_each_entry(gh, &gl->gl_holders, gh_list) {
if (test_bit(HIF_HOLDER, &gh->gh_iflags)) if (test_bit(HIF_HOLDER, &gh->gh_iflags))
continue; continue;
if (LM_FLAG_NOEXP & gh->gh_flags) if (LM_FLAG_RECOVER & gh->gh_flags)
return 0; return 0;
} }
@ -2165,18 +2080,26 @@ static void dump_glock_func(struct gfs2_glock *gl)
dump_glock(NULL, gl, true); dump_glock(NULL, gl, true);
} }
static void withdraw_dq(struct gfs2_glock *gl) static void withdraw_glock(struct gfs2_glock *gl)
{ {
spin_lock(&gl->gl_lockref.lock); spin_lock(&gl->gl_lockref.lock);
if (!__lockref_is_dead(&gl->gl_lockref) && if (!__lockref_is_dead(&gl->gl_lockref)) {
glock_blocked_by_withdraw(gl)) /*
* We don't want to write back any more dirty data. Unlock the
* remaining inode and resource group glocks; this will cause
* their ->go_inval() hooks to toss out all the remaining
* cached data, dirty or not.
*/
if (gl->gl_ops->go_inval && gl->gl_state != LM_ST_UNLOCKED)
request_demote(gl, LM_ST_UNLOCKED, 0, false);
do_error(gl, LM_OUT_ERROR); /* remove pending waiters */ do_error(gl, LM_OUT_ERROR); /* remove pending waiters */
}
spin_unlock(&gl->gl_lockref.lock); spin_unlock(&gl->gl_lockref.lock);
} }
void gfs2_gl_dq_holders(struct gfs2_sbd *sdp) void gfs2_withdraw_glocks(struct gfs2_sbd *sdp)
{ {
glock_hash_walk(withdraw_dq, sdp); glock_hash_walk(withdraw_glock, sdp);
} }
/** /**
@ -2237,7 +2160,7 @@ static const char *hflags2str(char *buf, u16 flags, unsigned long iflags)
*p++ = 't'; *p++ = 't';
if (flags & LM_FLAG_TRY_1CB) if (flags & LM_FLAG_TRY_1CB)
*p++ = 'T'; *p++ = 'T';
if (flags & LM_FLAG_NOEXP) if (flags & LM_FLAG_RECOVER)
*p++ = 'e'; *p++ = 'e';
if (flags & LM_FLAG_ANY) if (flags & LM_FLAG_ANY)
*p++ = 'A'; *p++ = 'A';
@ -2324,8 +2247,6 @@ static const char *gflags2str(char *buf, const struct gfs2_glock *gl)
*p++ = 'o'; *p++ = 'o';
if (test_bit(GLF_BLOCKING, gflags)) if (test_bit(GLF_BLOCKING, gflags))
*p++ = 'b'; *p++ = 'b';
if (test_bit(GLF_UNLOCKED, gflags))
*p++ = 'x';
if (test_bit(GLF_INSTANTIATE_NEEDED, gflags)) if (test_bit(GLF_INSTANTIATE_NEEDED, gflags))
*p++ = 'n'; *p++ = 'n';
if (test_bit(GLF_INSTANTIATE_IN_PROG, gflags)) if (test_bit(GLF_INSTANTIATE_IN_PROG, gflags))

12
fs/gfs2/glock.h
View File

@ -58,10 +58,10 @@ enum {
* LM_FLAG_TRY_1CB * LM_FLAG_TRY_1CB
* Send one blocking callback if TRY is set and the lock is not granted. * Send one blocking callback if TRY is set and the lock is not granted.
* *
* LM_FLAG_NOEXP * LM_FLAG_RECOVER
* GFS sets this flag on lock requests it makes while doing journal recovery. * GFS sets this flag on lock requests it makes while doing journal recovery.
* These special requests should not be blocked due to the recovery like * While ordinary requests are blocked until the end of recovery, requests
* ordinary locks would be. * with this flag set do proceed.
* *
* LM_FLAG_ANY * LM_FLAG_ANY
* A SHARED request may also be granted in DEFERRED, or a DEFERRED request may * A SHARED request may also be granted in DEFERRED, or a DEFERRED request may
@ -80,7 +80,7 @@ enum {
#define LM_FLAG_TRY 0x0001 #define LM_FLAG_TRY 0x0001
#define LM_FLAG_TRY_1CB 0x0002 #define LM_FLAG_TRY_1CB 0x0002
#define LM_FLAG_NOEXP 0x0004 #define LM_FLAG_RECOVER 0x0004
#define LM_FLAG_ANY 0x0008 #define LM_FLAG_ANY 0x0008
#define LM_FLAG_NODE_SCOPE 0x0020 #define LM_FLAG_NODE_SCOPE 0x0020
#define GL_ASYNC 0x0040 #define GL_ASYNC 0x0040
@ -136,7 +136,7 @@ struct lm_lockops {
void (*lm_first_done) (struct gfs2_sbd *sdp); void (*lm_first_done) (struct gfs2_sbd *sdp);
void (*lm_recovery_result) (struct gfs2_sbd *sdp, unsigned int jid, void (*lm_recovery_result) (struct gfs2_sbd *sdp, unsigned int jid,
unsigned int result); unsigned int result);
void (*lm_unmount) (struct gfs2_sbd *sdp); void (*lm_unmount) (struct gfs2_sbd *sdp, bool clean);
void (*lm_withdraw) (struct gfs2_sbd *sdp); void (*lm_withdraw) (struct gfs2_sbd *sdp);
void (*lm_put_lock) (struct gfs2_glock *gl); void (*lm_put_lock) (struct gfs2_glock *gl);
int (*lm_lock) (struct gfs2_glock *gl, unsigned int req_state, int (*lm_lock) (struct gfs2_glock *gl, unsigned int req_state,
@ -263,7 +263,7 @@ bool gfs2_queue_verify_delete(struct gfs2_glock *gl, bool later);
void gfs2_cancel_delete_work(struct gfs2_glock *gl); void gfs2_cancel_delete_work(struct gfs2_glock *gl);
void gfs2_flush_delete_work(struct gfs2_sbd *sdp); void gfs2_flush_delete_work(struct gfs2_sbd *sdp);
void gfs2_gl_hash_clear(struct gfs2_sbd *sdp); void gfs2_gl_hash_clear(struct gfs2_sbd *sdp);
void gfs2_gl_dq_holders(struct gfs2_sbd *sdp); void gfs2_withdraw_glocks(struct gfs2_sbd *sdp);
void gfs2_glock_thaw(struct gfs2_sbd *sdp); void gfs2_glock_thaw(struct gfs2_sbd *sdp);
void gfs2_glock_free(struct gfs2_glock *gl); void gfs2_glock_free(struct gfs2_glock *gl);
void gfs2_glock_free_later(struct gfs2_glock *gl); void gfs2_glock_free_later(struct gfs2_glock *gl);

98
fs/gfs2/glops.c
View File

@ -30,8 +30,6 @@
struct workqueue_struct *gfs2_freeze_wq; struct workqueue_struct *gfs2_freeze_wq;
extern struct workqueue_struct *gfs2_control_wq;
static void gfs2_ail_error(struct gfs2_glock *gl, const struct buffer_head *bh) static void gfs2_ail_error(struct gfs2_glock *gl, const struct buffer_head *bh)
{ {
struct gfs2_sbd *sdp = gl->gl_name.ln_sbd; struct gfs2_sbd *sdp = gl->gl_name.ln_sbd;
@ -45,7 +43,7 @@ static void gfs2_ail_error(struct gfs2_glock *gl, const struct buffer_head *bh)
gl->gl_name.ln_type, gl->gl_name.ln_number, gl->gl_name.ln_type, gl->gl_name.ln_number,
gfs2_glock2aspace(gl)); gfs2_glock2aspace(gl));
gfs2_lm(sdp, "AIL error\n"); gfs2_lm(sdp, "AIL error\n");
gfs2_withdraw_delayed(sdp); gfs2_withdraw(sdp);
} }
/** /**
@ -83,9 +81,6 @@ static void __gfs2_ail_flush(struct gfs2_glock *gl, bool fsync,
GLOCK_BUG_ON(gl, !fsync && atomic_read(&gl->gl_ail_count)); GLOCK_BUG_ON(gl, !fsync && atomic_read(&gl->gl_ail_count));
spin_unlock(&sdp->sd_ail_lock); spin_unlock(&sdp->sd_ail_lock);
gfs2_log_unlock(sdp); gfs2_log_unlock(sdp);
if (gfs2_withdrawing(sdp))
gfs2_withdraw(sdp);
} }
@ -178,7 +173,7 @@ static int gfs2_rgrp_metasync(struct gfs2_glock *gl)
filemap_fdatawrite_range(metamapping, start, end); filemap_fdatawrite_range(metamapping, start, end);
error = filemap_fdatawait_range(metamapping, start, end); error = filemap_fdatawait_range(metamapping, start, end);
WARN_ON_ONCE(error && !gfs2_withdrawing_or_withdrawn(sdp)); WARN_ON_ONCE(error && !gfs2_withdrawn(sdp));
mapping_set_error(metamapping, error); mapping_set_error(metamapping, error);
if (error) if (error)
gfs2_io_error(sdp); gfs2_io_error(sdp);
@ -237,6 +232,7 @@ static void rgrp_go_inval(struct gfs2_glock *gl, int flags)
end = PAGE_ALIGN((rgd->rd_addr + rgd->rd_length) * bsize) - 1; end = PAGE_ALIGN((rgd->rd_addr + rgd->rd_length) * bsize) - 1;
gfs2_rgrp_brelse(rgd); gfs2_rgrp_brelse(rgd);
WARN_ON_ONCE(!(flags & DIO_METADATA)); WARN_ON_ONCE(!(flags & DIO_METADATA));
gfs2_assert_withdraw(sdp, !atomic_read(&gl->gl_ail_count));
truncate_inode_pages_range(mapping, start, end); truncate_inode_pages_range(mapping, start, end);
} }
@ -363,6 +359,8 @@ static void inode_go_inval(struct gfs2_glock *gl, int flags)
{ {
struct gfs2_inode *ip = gfs2_glock2inode(gl); struct gfs2_inode *ip = gfs2_glock2inode(gl);
gfs2_assert_withdraw(gl->gl_name.ln_sbd, !atomic_read(&gl->gl_ail_count));
if (flags & DIO_METADATA) { if (flags & DIO_METADATA) {
struct address_space *mapping = gfs2_glock2aspace(gl); struct address_space *mapping = gfs2_glock2aspace(gl);
truncate_inode_pages(mapping, 0); truncate_inode_pages(mapping, 0);
@ -608,10 +606,10 @@ static int freeze_go_xmote_bh(struct gfs2_glock *gl)
j_gl->gl_ops->go_inval(j_gl, DIO_METADATA); j_gl->gl_ops->go_inval(j_gl, DIO_METADATA);
error = gfs2_find_jhead(sdp->sd_jdesc, &head); error = gfs2_find_jhead(sdp->sd_jdesc, &head);
if (gfs2_assert_withdraw_delayed(sdp, !error)) if (gfs2_assert_withdraw(sdp, !error))
return error; return error;
if (gfs2_assert_withdraw_delayed(sdp, head.lh_flags & if (gfs2_assert_withdraw(sdp, head.lh_flags &
GFS2_LOG_HEAD_UNMOUNT)) GFS2_LOG_HEAD_UNMOUNT))
return -EIO; return -EIO;
gfs2_log_pointers_init(sdp, &head); gfs2_log_pointers_init(sdp, &head);
} }
@ -630,8 +628,7 @@ static void iopen_go_callback(struct gfs2_glock *gl, bool remote)
struct gfs2_inode *ip = gl->gl_object; struct gfs2_inode *ip = gl->gl_object;
struct gfs2_sbd *sdp = gl->gl_name.ln_sbd; struct gfs2_sbd *sdp = gl->gl_name.ln_sbd;
if (!remote || sb_rdonly(sdp->sd_vfs) || if (!remote || test_bit(SDF_KILL, &sdp->sd_flags))
test_bit(SDF_KILL, &sdp->sd_flags))
return; return;
if (gl->gl_demote_state == LM_ST_UNLOCKED && if (gl->gl_demote_state == LM_ST_UNLOCKED &&
@ -642,76 +639,8 @@ static void iopen_go_callback(struct gfs2_glock *gl, bool remote)
} }
} }
/**
* inode_go_unlocked - wake up anyone waiting for dlm's unlock ast
* @gl: glock being unlocked
*
* For now, this is only used for the journal inode glock. In withdraw
* situations, we need to wait for the glock to be unlocked so that we know
* other nodes may proceed with recovery / journal replay.
*/
static void inode_go_unlocked(struct gfs2_glock *gl)
{
/* Note that we cannot reference gl_object because it's already set
* to NULL by this point in its lifecycle. */
if (!test_bit(GLF_UNLOCKED, &gl->gl_flags))
return;
clear_bit_unlock(GLF_UNLOCKED, &gl->gl_flags);
wake_up_bit(&gl->gl_flags, GLF_UNLOCKED);
}
/**
* nondisk_go_callback - used to signal when a node did a withdraw
* @gl: the nondisk glock
* @remote: true if this came from a different cluster node
*
*/
static void nondisk_go_callback(struct gfs2_glock *gl, bool remote)
{
struct gfs2_sbd *sdp = gl->gl_name.ln_sbd;
/* Ignore the callback unless it's from another node, and it's the
live lock. */
if (!remote || gl->gl_name.ln_number != GFS2_LIVE_LOCK)
return;
/* First order of business is to cancel the demote request. We don't
* really want to demote a nondisk glock. At best it's just to inform
* us of another node's withdraw. We'll keep it in SH mode. */
clear_bit(GLF_DEMOTE, &gl->gl_flags);
clear_bit(GLF_PENDING_DEMOTE, &gl->gl_flags);
/* Ignore the unlock if we're withdrawn, unmounting, or in recovery. */
if (test_bit(SDF_NORECOVERY, &sdp->sd_flags) ||
test_bit(SDF_WITHDRAWN, &sdp->sd_flags) ||
test_bit(SDF_REMOTE_WITHDRAW, &sdp->sd_flags))
return;
/* We only care when a node wants us to unlock, because that means
* they want a journal recovered. */
if (gl->gl_demote_state != LM_ST_UNLOCKED)
return;
if (sdp->sd_args.ar_spectator) {
fs_warn(sdp, "Spectator node cannot recover journals.\n");
return;
}
fs_warn(sdp, "Some node has withdrawn; checking for recovery.\n");
set_bit(SDF_REMOTE_WITHDRAW, &sdp->sd_flags);
/*
* We can't call remote_withdraw directly here or gfs2_recover_journal
* because this is called from the glock unlock function and the
* remote_withdraw needs to enqueue and dequeue the same "live" glock
* we were called from. So we queue it to the control work queue in
* lock_dlm.
*/
queue_delayed_work(gfs2_control_wq, &sdp->sd_control_work, 0);
}
const struct gfs2_glock_operations gfs2_meta_glops = { const struct gfs2_glock_operations gfs2_meta_glops = {
.go_type = LM_TYPE_META, .go_type = LM_TYPE_META,
.go_flags = GLOF_NONDISK,
}; };
const struct gfs2_glock_operations gfs2_inode_glops = { const struct gfs2_glock_operations gfs2_inode_glops = {
@ -722,7 +651,6 @@ const struct gfs2_glock_operations gfs2_inode_glops = {
.go_dump = inode_go_dump, .go_dump = inode_go_dump,
.go_type = LM_TYPE_INODE, .go_type = LM_TYPE_INODE,
.go_flags = GLOF_ASPACE | GLOF_LVB, .go_flags = GLOF_ASPACE | GLOF_LVB,
.go_unlocked = inode_go_unlocked,
}; };
const struct gfs2_glock_operations gfs2_rgrp_glops = { const struct gfs2_glock_operations gfs2_rgrp_glops = {
@ -738,36 +666,30 @@ const struct gfs2_glock_operations gfs2_freeze_glops = {
.go_xmote_bh = freeze_go_xmote_bh, .go_xmote_bh = freeze_go_xmote_bh,
.go_callback = freeze_go_callback, .go_callback = freeze_go_callback,
.go_type = LM_TYPE_NONDISK, .go_type = LM_TYPE_NONDISK,
.go_flags = GLOF_NONDISK,
}; };
const struct gfs2_glock_operations gfs2_iopen_glops = { const struct gfs2_glock_operations gfs2_iopen_glops = {
.go_type = LM_TYPE_IOPEN, .go_type = LM_TYPE_IOPEN,
.go_callback = iopen_go_callback, .go_callback = iopen_go_callback,
.go_dump = inode_go_dump, .go_dump = inode_go_dump,
.go_flags = GLOF_NONDISK,
.go_subclass = 1, .go_subclass = 1,
}; };
const struct gfs2_glock_operations gfs2_flock_glops = { const struct gfs2_glock_operations gfs2_flock_glops = {
.go_type = LM_TYPE_FLOCK, .go_type = LM_TYPE_FLOCK,
.go_flags = GLOF_NONDISK,
}; };
const struct gfs2_glock_operations gfs2_nondisk_glops = { const struct gfs2_glock_operations gfs2_nondisk_glops = {
.go_type = LM_TYPE_NONDISK, .go_type = LM_TYPE_NONDISK,
.go_flags = GLOF_NONDISK,
.go_callback = nondisk_go_callback,
}; };
const struct gfs2_glock_operations gfs2_quota_glops = { const struct gfs2_glock_operations gfs2_quota_glops = {
.go_type = LM_TYPE_QUOTA, .go_type = LM_TYPE_QUOTA,
.go_flags = GLOF_LVB | GLOF_NONDISK, .go_flags = GLOF_LVB,
}; };
const struct gfs2_glock_operations gfs2_journal_glops = { const struct gfs2_glock_operations gfs2_journal_glops = {
.go_type = LM_TYPE_JOURNAL, .go_type = LM_TYPE_JOURNAL,
.go_flags = GLOF_NONDISK,
}; };
const struct gfs2_glock_operations *gfs2_glops_list[] = { const struct gfs2_glock_operations *gfs2_glops_list[] = {

22
fs/gfs2/incore.h
View File

@ -223,13 +223,11 @@ struct gfs2_glock_operations {
void (*go_dump)(struct seq_file *seq, const struct gfs2_glock *gl, void (*go_dump)(struct seq_file *seq, const struct gfs2_glock *gl,
const char *fs_id_buf); const char *fs_id_buf);
void (*go_callback)(struct gfs2_glock *gl, bool remote); void (*go_callback)(struct gfs2_glock *gl, bool remote);
void (*go_unlocked)(struct gfs2_glock *gl);
const int go_subclass; const int go_subclass;
const int go_type; const int go_type;
const unsigned long go_flags; const unsigned long go_flags;
#define GLOF_ASPACE 1 /* address space attached */ #define GLOF_ASPACE 1 /* address space attached */
#define GLOF_LVB 2 /* Lock Value Block attached */ #define GLOF_LVB 2 /* Lock Value Block attached */
#define GLOF_NONDISK 8 /* not I/O related */
}; };
enum { enum {
@ -326,7 +324,6 @@ enum {
GLF_LRU = 13, GLF_LRU = 13,
GLF_OBJECT = 14, /* Used only for tracing */ GLF_OBJECT = 14, /* Used only for tracing */
GLF_BLOCKING = 15, GLF_BLOCKING = 15,
GLF_UNLOCKED = 16, /* Wait for glock to be unlocked */
GLF_TRY_TO_EVICT = 17, /* iopen glocks only */ GLF_TRY_TO_EVICT = 17, /* iopen glocks only */
GLF_VERIFY_DELETE = 18, /* iopen glocks only */ GLF_VERIFY_DELETE = 18, /* iopen glocks only */
GLF_PENDING_REPLY = 19, GLF_PENDING_REPLY = 19,
@ -520,8 +517,6 @@ struct gfs2_jdesc {
struct list_head jd_revoke_list; struct list_head jd_revoke_list;
unsigned int jd_replay_tail; unsigned int jd_replay_tail;
u64 jd_no_addr;
}; };
struct gfs2_statfs_change_host { struct gfs2_statfs_change_host {
@ -542,8 +537,7 @@ struct gfs2_statfs_change_host {
#define GFS2_ERRORS_DEFAULT GFS2_ERRORS_WITHDRAW #define GFS2_ERRORS_DEFAULT GFS2_ERRORS_WITHDRAW
#define GFS2_ERRORS_WITHDRAW 0 #define GFS2_ERRORS_WITHDRAW 0
#define GFS2_ERRORS_CONTINUE 1 /* place holder for future feature */ #define GFS2_ERRORS_DEACTIVATE 1
#define GFS2_ERRORS_RO 2 /* place holder for future feature */
#define GFS2_ERRORS_PANIC 3 #define GFS2_ERRORS_PANIC 3
struct gfs2_args { struct gfs2_args {
@ -559,7 +553,7 @@ struct gfs2_args {
unsigned int ar_data:2; /* ordered/writeback */ unsigned int ar_data:2; /* ordered/writeback */
unsigned int ar_meta:1; /* mount metafs */ unsigned int ar_meta:1; /* mount metafs */
unsigned int ar_discard:1; /* discard requests */ unsigned int ar_discard:1; /* discard requests */
unsigned int ar_errors:2; /* errors=withdraw | panic */ unsigned int ar_errors:2; /* errors=withdraw | deactivate | panic */
unsigned int ar_nobarrier:1; /* do not send barriers */ unsigned int ar_nobarrier:1; /* do not send barriers */
unsigned int ar_rgrplvb:1; /* use lvbs for rgrp info */ unsigned int ar_rgrplvb:1; /* use lvbs for rgrp info */
unsigned int ar_got_rgrplvb:1; /* Was the rgrplvb opt given? */ unsigned int ar_got_rgrplvb:1; /* Was the rgrplvb opt given? */
@ -585,6 +579,7 @@ struct gfs2_tune {
unsigned int gt_complain_secs; unsigned int gt_complain_secs;
unsigned int gt_statfs_quantum; unsigned int gt_statfs_quantum;
unsigned int gt_statfs_slow; unsigned int gt_statfs_slow;
unsigned int gt_withdraw_helper_timeout;
}; };
enum { enum {
@ -599,11 +594,6 @@ enum {
SDF_SKIP_DLM_UNLOCK = 8, SDF_SKIP_DLM_UNLOCK = 8,
SDF_FORCE_AIL_FLUSH = 9, SDF_FORCE_AIL_FLUSH = 9,
SDF_FREEZE_INITIATOR = 10, SDF_FREEZE_INITIATOR = 10,
SDF_WITHDRAWING = 11, /* Will withdraw eventually */
SDF_WITHDRAW_IN_PROG = 12, /* Withdraw is in progress */
SDF_REMOTE_WITHDRAW = 13, /* Performing remote recovery */
SDF_WITHDRAW_RECOVERY = 14, /* Wait for journal recovery when we are
withdrawing */
SDF_KILL = 15, SDF_KILL = 15,
SDF_EVICTING = 16, SDF_EVICTING = 16,
SDF_FROZEN = 17, SDF_FROZEN = 17,
@ -716,11 +706,13 @@ struct gfs2_sbd {
struct gfs2_glock *sd_rename_gl; struct gfs2_glock *sd_rename_gl;
struct gfs2_glock *sd_freeze_gl; struct gfs2_glock *sd_freeze_gl;
struct work_struct sd_freeze_work; struct work_struct sd_freeze_work;
struct work_struct sd_withdraw_work;
wait_queue_head_t sd_kill_wait; wait_queue_head_t sd_kill_wait;
wait_queue_head_t sd_async_glock_wait; wait_queue_head_t sd_async_glock_wait;
atomic_t sd_glock_disposal; atomic_t sd_glock_disposal;
struct completion sd_locking_init; struct completion sd_locking_init;
struct completion sd_wdack; struct completion sd_withdraw_helper;
int sd_withdraw_helper_status;
struct delayed_work sd_control_work; struct delayed_work sd_control_work;
/* Inode Stuff */ /* Inode Stuff */
@ -761,7 +753,6 @@ struct gfs2_sbd {
struct gfs2_jdesc *sd_jdesc; struct gfs2_jdesc *sd_jdesc;
struct gfs2_holder sd_journal_gh; struct gfs2_holder sd_journal_gh;
struct gfs2_holder sd_jinode_gh; struct gfs2_holder sd_jinode_gh;
struct gfs2_glock *sd_jinode_gl;
struct gfs2_holder sd_sc_gh; struct gfs2_holder sd_sc_gh;
struct buffer_head *sd_sc_bh; struct buffer_head *sd_sc_bh;
@ -846,7 +837,6 @@ struct gfs2_sbd {
unsigned long sd_last_warning; unsigned long sd_last_warning;
struct dentry *debugfs_dir; /* debugfs directory */ struct dentry *debugfs_dir; /* debugfs directory */
unsigned long sd_glock_dqs_held;
}; };
#define GFS2_BAD_INO 1 #define GFS2_BAD_INO 1

15
fs/gfs2/inode.c
View File

@ -89,6 +89,19 @@ static int iget_set(struct inode *inode, void *opaque)
return 0; return 0;
} }
void gfs2_setup_inode(struct inode *inode)
{
gfp_t gfp_mask;
/*
* Ensure all page cache allocations are done from GFP_NOFS context to
* prevent direct reclaim recursion back into the filesystem and blowing
* stacks or deadlocking.
*/
gfp_mask = mapping_gfp_mask(inode->i_mapping);
mapping_set_gfp_mask(inode->i_mapping, gfp_mask & ~__GFP_FS);
}
/** /**
* gfs2_inode_lookup - Lookup an inode * gfs2_inode_lookup - Lookup an inode
* @sb: The super block * @sb: The super block
@ -132,6 +145,7 @@ struct inode *gfs2_inode_lookup(struct super_block *sb, unsigned int type,
struct gfs2_glock *io_gl; struct gfs2_glock *io_gl;
int extra_flags = 0; int extra_flags = 0;
gfs2_setup_inode(inode);
error = gfs2_glock_get(sdp, no_addr, &gfs2_inode_glops, CREATE, error = gfs2_glock_get(sdp, no_addr, &gfs2_inode_glops, CREATE,
&ip->i_gl); &ip->i_gl);
if (unlikely(error)) if (unlikely(error))
@ -752,6 +766,7 @@ static int gfs2_create_inode(struct inode *dir, struct dentry *dentry,
error = -ENOMEM; error = -ENOMEM;
if (!inode) if (!inode)
goto fail_gunlock; goto fail_gunlock;
gfs2_setup_inode(inode);
ip = GFS2_I(inode); ip = GFS2_I(inode);
error = posix_acl_create(dir, &mode, &default_acl, &acl); error = posix_acl_create(dir, &mode, &default_acl, &acl);

1
fs/gfs2/inode.h
View File

@ -86,6 +86,7 @@ err:
return -EIO; return -EIO;
} }
void gfs2_setup_inode(struct inode *inode);
struct inode *gfs2_inode_lookup(struct super_block *sb, unsigned type, struct inode *gfs2_inode_lookup(struct super_block *sb, unsigned type,
u64 no_addr, u64 no_formal_ino, u64 no_addr, u64 no_formal_ino,
unsigned int blktype); unsigned int blktype);

57
fs/gfs2/lock_dlm.c
View File

@ -15,9 +15,6 @@
#include <linux/sched/signal.h> #include <linux/sched/signal.h>
#include "incore.h" #include "incore.h"
#include "glock.h"
#include "glops.h"
#include "recovery.h"
#include "util.h" #include "util.h"
#include "sys.h" #include "sys.h"
#include "trace_gfs2.h" #include "trace_gfs2.h"
@ -139,8 +136,6 @@ static void gdlm_ast(void *arg)
switch (gl->gl_lksb.sb_status) { switch (gl->gl_lksb.sb_status) {
case -DLM_EUNLOCK: /* Unlocked, so glock can be freed */ case -DLM_EUNLOCK: /* Unlocked, so glock can be freed */
if (gl->gl_ops->go_unlocked)
gl->gl_ops->go_unlocked(gl);
gfs2_glock_free(gl); gfs2_glock_free(gl);
return; return;
case -DLM_ECANCEL: /* Cancel while getting lock */ case -DLM_ECANCEL: /* Cancel while getting lock */
@ -399,7 +394,6 @@ static void gdlm_cancel(struct gfs2_glock *gl)
/* /*
* dlm/gfs2 recovery coordination using dlm_recover callbacks * dlm/gfs2 recovery coordination using dlm_recover callbacks
* *
* 0. gfs2 checks for another cluster node withdraw, needing journal replay
* 1. dlm_controld sees lockspace members change * 1. dlm_controld sees lockspace members change
* 2. dlm_controld blocks dlm-kernel locking activity * 2. dlm_controld blocks dlm-kernel locking activity
* 3. dlm_controld within dlm-kernel notifies gfs2 (recover_prep) * 3. dlm_controld within dlm-kernel notifies gfs2 (recover_prep)
@ -657,28 +651,6 @@ static int control_lock(struct gfs2_sbd *sdp, int mode, uint32_t flags)
&ls->ls_control_lksb, "control_lock"); &ls->ls_control_lksb, "control_lock");
} }
/**
* remote_withdraw - react to a node withdrawing from the file system
* @sdp: The superblock
*/
static void remote_withdraw(struct gfs2_sbd *sdp)
{
struct gfs2_jdesc *jd;
int ret = 0, count = 0;
list_for_each_entry(jd, &sdp->sd_jindex_list, jd_list) {
if (jd->jd_jid == sdp->sd_lockstruct.ls_jid)
continue;
ret = gfs2_recover_journal(jd, true);
if (ret)
break;
count++;
}
/* Now drop the additional reference we acquired */
fs_err(sdp, "Journals checked: %d, ret = %d.\n", count, ret);
}
static void gfs2_control_func(struct work_struct *work) static void gfs2_control_func(struct work_struct *work)
{ {
struct gfs2_sbd *sdp = container_of(work, struct gfs2_sbd, sd_control_work.work); struct gfs2_sbd *sdp = container_of(work, struct gfs2_sbd, sd_control_work.work);
@ -689,13 +661,6 @@ static void gfs2_control_func(struct work_struct *work)
int recover_size; int recover_size;
int i, error; int i, error;
/* First check for other nodes that may have done a withdraw. */
if (test_bit(SDF_REMOTE_WITHDRAW, &sdp->sd_flags)) {
remote_withdraw(sdp);
clear_bit(SDF_REMOTE_WITHDRAW, &sdp->sd_flags);
return;
}
spin_lock(&ls->ls_recover_spin); spin_lock(&ls->ls_recover_spin);
/* /*
* No MOUNT_DONE means we're still mounting; control_mount() * No MOUNT_DONE means we're still mounting; control_mount()
@ -1195,7 +1160,7 @@ static void gdlm_recover_prep(void *arg)
struct gfs2_sbd *sdp = arg; struct gfs2_sbd *sdp = arg;
struct lm_lockstruct *ls = &sdp->sd_lockstruct; struct lm_lockstruct *ls = &sdp->sd_lockstruct;
if (gfs2_withdrawing_or_withdrawn(sdp)) { if (gfs2_withdrawn(sdp)) {
fs_err(sdp, "recover_prep ignored due to withdraw.\n"); fs_err(sdp, "recover_prep ignored due to withdraw.\n");
return; return;
} }
@ -1221,7 +1186,7 @@ static void gdlm_recover_slot(void *arg, struct dlm_slot *slot)
struct lm_lockstruct *ls = &sdp->sd_lockstruct; struct lm_lockstruct *ls = &sdp->sd_lockstruct;
int jid = slot->slot - 1; int jid = slot->slot - 1;
if (gfs2_withdrawing_or_withdrawn(sdp)) { if (gfs2_withdrawn(sdp)) {
fs_err(sdp, "recover_slot jid %d ignored due to withdraw.\n", fs_err(sdp, "recover_slot jid %d ignored due to withdraw.\n",
jid); jid);
return; return;
@ -1250,7 +1215,7 @@ static void gdlm_recover_done(void *arg, struct dlm_slot *slots, int num_slots,
struct gfs2_sbd *sdp = arg; struct gfs2_sbd *sdp = arg;
struct lm_lockstruct *ls = &sdp->sd_lockstruct; struct lm_lockstruct *ls = &sdp->sd_lockstruct;
if (gfs2_withdrawing_or_withdrawn(sdp)) { if (gfs2_withdrawn(sdp)) {
fs_err(sdp, "recover_done ignored due to withdraw.\n"); fs_err(sdp, "recover_done ignored due to withdraw.\n");
return; return;
} }
@ -1281,7 +1246,7 @@ static void gdlm_recovery_result(struct gfs2_sbd *sdp, unsigned int jid,
{ {
struct lm_lockstruct *ls = &sdp->sd_lockstruct; struct lm_lockstruct *ls = &sdp->sd_lockstruct;
if (gfs2_withdrawing_or_withdrawn(sdp)) { if (gfs2_withdrawn(sdp)) {
fs_err(sdp, "recovery_result jid %d ignored due to withdraw.\n", fs_err(sdp, "recovery_result jid %d ignored due to withdraw.\n",
jid); jid);
return; return;
@ -1438,7 +1403,15 @@ static void gdlm_first_done(struct gfs2_sbd *sdp)
fs_err(sdp, "mount first_done error %d\n", error); fs_err(sdp, "mount first_done error %d\n", error);
} }
static void gdlm_unmount(struct gfs2_sbd *sdp) /*
* gdlm_unmount - release our lockspace
* @sdp: the superblock
* @clean: Indicates whether or not the remaining nodes in the cluster should
* perform recovery. Recovery is necessary when a node withdraws and
* its journal remains dirty. Recovery isn't necessary when a node
* cleanly unmounts a filesystem.
*/
static void gdlm_unmount(struct gfs2_sbd *sdp, bool clean)
{ {
struct lm_lockstruct *ls = &sdp->sd_lockstruct; struct lm_lockstruct *ls = &sdp->sd_lockstruct;
@ -1456,7 +1429,9 @@ static void gdlm_unmount(struct gfs2_sbd *sdp)
release: release:
down_write(&ls->ls_sem); down_write(&ls->ls_sem);
if (ls->ls_dlm) { if (ls->ls_dlm) {
dlm_release_lockspace(ls->ls_dlm, DLM_RELEASE_NORMAL); dlm_release_lockspace(ls->ls_dlm,
clean ? DLM_RELEASE_NORMAL :
DLM_RELEASE_RECOVER);
ls->ls_dlm = NULL; ls->ls_dlm = NULL;
} }
up_write(&ls->ls_sem); up_write(&ls->ls_sem);

59
fs/gfs2/log.c
View File

@ -112,13 +112,11 @@ __acquires(&sdp->sd_ail_lock)
&tr->tr_ail2_list); &tr->tr_ail2_list);
continue; continue;
} }
if (!cmpxchg(&sdp->sd_log_error, 0, -EIO)) { if (!cmpxchg(&sdp->sd_log_error, 0, -EIO))
gfs2_io_error_bh(sdp, bh); gfs2_io_error_bh(sdp, bh);
gfs2_withdraw_delayed(sdp);
}
} }
if (gfs2_withdrawing_or_withdrawn(sdp)) { if (gfs2_withdrawn(sdp)) {
gfs2_remove_from_ail(bd); gfs2_remove_from_ail(bd);
continue; continue;
} }
@ -324,10 +322,8 @@ static int gfs2_ail1_empty_one(struct gfs2_sbd *sdp, struct gfs2_trans *tr,
continue; continue;
} }
if (!buffer_uptodate(bh) && if (!buffer_uptodate(bh) &&
!cmpxchg(&sdp->sd_log_error, 0, -EIO)) { !cmpxchg(&sdp->sd_log_error, 0, -EIO))
gfs2_io_error_bh(sdp, bh); gfs2_io_error_bh(sdp, bh);
gfs2_withdraw_delayed(sdp);
}
/* /*
* If we have space for revokes and the bd is no longer on any * If we have space for revokes and the bd is no longer on any
* buf list, we can just add a revoke for it immediately and * buf list, we can just add a revoke for it immediately and
@ -807,9 +803,6 @@ void gfs2_flush_revokes(struct gfs2_sbd *sdp)
gfs2_log_lock(sdp); gfs2_log_lock(sdp);
gfs2_ail1_empty(sdp, max_revokes); gfs2_ail1_empty(sdp, max_revokes);
gfs2_log_unlock(sdp); gfs2_log_unlock(sdp);
if (gfs2_withdrawing(sdp))
gfs2_withdraw(sdp);
} }
/** /**
@ -837,7 +830,7 @@ void gfs2_write_log_header(struct gfs2_sbd *sdp, struct gfs2_jdesc *jd,
struct super_block *sb = sdp->sd_vfs; struct super_block *sb = sdp->sd_vfs;
u64 dblock; u64 dblock;
if (gfs2_withdrawing_or_withdrawn(sdp)) if (gfs2_withdrawn(sdp))
return; return;
page = mempool_alloc(gfs2_page_pool, GFP_NOIO); page = mempool_alloc(gfs2_page_pool, GFP_NOIO);
@ -984,12 +977,9 @@ static void empty_ail1_list(struct gfs2_sbd *sdp)
gfs2_ail1_wait(sdp); gfs2_ail1_wait(sdp);
empty = gfs2_ail1_empty(sdp, 0); empty = gfs2_ail1_empty(sdp, 0);
if (gfs2_withdrawing_or_withdrawn(sdp)) if (gfs2_withdrawn(sdp))
break; break;
} }
if (gfs2_withdrawing(sdp))
gfs2_withdraw(sdp);
} }
/** /**
@ -1050,7 +1040,7 @@ repeat:
* Do this check while holding the log_flush_lock to prevent new * Do this check while holding the log_flush_lock to prevent new
* buffers from being added to the ail via gfs2_pin() * buffers from being added to the ail via gfs2_pin()
*/ */
if (gfs2_withdrawing_or_withdrawn(sdp) || if (gfs2_withdrawn(sdp) ||
!test_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags)) !test_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags))
goto out; goto out;
@ -1071,7 +1061,7 @@ repeat:
sdp->sd_log_tr = NULL; sdp->sd_log_tr = NULL;
tr->tr_first = first_log_head; tr->tr_first = first_log_head;
if (unlikely(frozen)) { if (unlikely(frozen)) {
if (gfs2_assert_withdraw_delayed(sdp, if (gfs2_assert_withdraw(sdp,
!tr->tr_num_buf_new && !tr->tr_num_databuf_new)) !tr->tr_num_buf_new && !tr->tr_num_databuf_new))
goto out_withdraw; goto out_withdraw;
} }
@ -1096,18 +1086,18 @@ repeat:
clear_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags); clear_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags);
if (unlikely(frozen)) if (unlikely(frozen))
if (gfs2_assert_withdraw_delayed(sdp, !reserved_revokes)) if (gfs2_assert_withdraw(sdp, !reserved_revokes))
goto out_withdraw; goto out_withdraw;
gfs2_ordered_write(sdp); gfs2_ordered_write(sdp);
if (gfs2_withdrawing_or_withdrawn(sdp)) if (gfs2_withdrawn(sdp))
goto out_withdraw; goto out_withdraw;
lops_before_commit(sdp, tr); lops_before_commit(sdp, tr);
if (gfs2_withdrawing_or_withdrawn(sdp)) if (gfs2_withdrawn(sdp))
goto out_withdraw; goto out_withdraw;
if (sdp->sd_jdesc) if (sdp->sd_jdesc)
gfs2_log_submit_bio(&sdp->sd_jdesc->jd_log_bio, REQ_OP_WRITE); gfs2_log_submit_bio(&sdp->sd_jdesc->jd_log_bio, REQ_OP_WRITE);
if (gfs2_withdrawing_or_withdrawn(sdp)) if (gfs2_withdrawn(sdp))
goto out_withdraw; goto out_withdraw;
if (sdp->sd_log_head != sdp->sd_log_flush_head) { if (sdp->sd_log_head != sdp->sd_log_flush_head) {
@ -1115,7 +1105,7 @@ repeat:
} else if (sdp->sd_log_tail != sdp->sd_log_flush_tail && !sdp->sd_log_idle) { } else if (sdp->sd_log_tail != sdp->sd_log_flush_tail && !sdp->sd_log_idle) {
log_write_header(sdp, flags); log_write_header(sdp, flags);
} }
if (gfs2_withdrawing_or_withdrawn(sdp)) if (gfs2_withdrawn(sdp))
goto out_withdraw; goto out_withdraw;
lops_after_commit(sdp, tr); lops_after_commit(sdp, tr);
@ -1133,7 +1123,7 @@ repeat:
if (!(flags & GFS2_LOG_HEAD_FLUSH_NORMAL)) { if (!(flags & GFS2_LOG_HEAD_FLUSH_NORMAL)) {
if (!sdp->sd_log_idle) { if (!sdp->sd_log_idle) {
empty_ail1_list(sdp); empty_ail1_list(sdp);
if (gfs2_withdrawing_or_withdrawn(sdp)) if (gfs2_withdrawn(sdp))
goto out_withdraw; goto out_withdraw;
log_write_header(sdp, flags); log_write_header(sdp, flags);
} }
@ -1151,13 +1141,11 @@ out_end:
reserved_blocks += (reserved_revokes - sdp->sd_ldptrs) / sdp->sd_inptrs; reserved_blocks += (reserved_revokes - sdp->sd_ldptrs) / sdp->sd_inptrs;
out: out:
if (used_blocks != reserved_blocks) { if (used_blocks != reserved_blocks) {
gfs2_assert_withdraw_delayed(sdp, used_blocks < reserved_blocks); gfs2_assert_withdraw(sdp, used_blocks < reserved_blocks);
gfs2_log_release(sdp, reserved_blocks - used_blocks); gfs2_log_release(sdp, reserved_blocks - used_blocks);
} }
up_write(&sdp->sd_log_flush_lock); up_write(&sdp->sd_log_flush_lock);
gfs2_trans_free(sdp, tr); gfs2_trans_free(sdp, tr);
if (gfs2_withdrawing(sdp))
gfs2_withdraw(sdp);
trace_gfs2_log_flush(sdp, 0, flags); trace_gfs2_log_flush(sdp, 0, flags);
return; return;
@ -1304,20 +1292,9 @@ int gfs2_logd(void *data)
set_freezable(); set_freezable();
while (!kthread_should_stop()) { while (!kthread_should_stop()) {
if (gfs2_withdrawing_or_withdrawn(sdp)) if (gfs2_withdrawn(sdp))
break; break;
/* Check for errors writing to the journal */
if (sdp->sd_log_error) {
gfs2_lm(sdp,
"GFS2: fsid=%s: error %d: "
"withdrawing the file system to "
"prevent further damage.\n",
sdp->sd_fsname, sdp->sd_log_error);
gfs2_withdraw(sdp);
break;
}
if (gfs2_jrnl_flush_reqd(sdp) || t == 0) { if (gfs2_jrnl_flush_reqd(sdp) || t == 0) {
gfs2_ail1_empty(sdp, 0); gfs2_ail1_empty(sdp, 0);
gfs2_log_flush(sdp, NULL, GFS2_LOG_HEAD_FLUSH_NORMAL | gfs2_log_flush(sdp, NULL, GFS2_LOG_HEAD_FLUSH_NORMAL |
@ -1340,15 +1317,11 @@ int gfs2_logd(void *data)
test_bit(SDF_FORCE_AIL_FLUSH, &sdp->sd_flags) || test_bit(SDF_FORCE_AIL_FLUSH, &sdp->sd_flags) ||
gfs2_ail_flush_reqd(sdp) || gfs2_ail_flush_reqd(sdp) ||
gfs2_jrnl_flush_reqd(sdp) || gfs2_jrnl_flush_reqd(sdp) ||
sdp->sd_log_error || gfs2_withdrawn(sdp) ||
gfs2_withdrawing_or_withdrawn(sdp) ||
kthread_should_stop(), kthread_should_stop(),
t); t);
} }
if (gfs2_withdrawing(sdp))
gfs2_withdraw(sdp);
return 0; return 0;
} }

12
fs/gfs2/lops.c
View File

@ -49,7 +49,7 @@ void gfs2_pin(struct gfs2_sbd *sdp, struct buffer_head *bh)
if (test_set_buffer_pinned(bh)) if (test_set_buffer_pinned(bh))
gfs2_assert_withdraw(sdp, 0); gfs2_assert_withdraw(sdp, 0);
if (!buffer_uptodate(bh)) if (!buffer_uptodate(bh))
gfs2_io_error_bh_wd(sdp, bh); gfs2_io_error_bh(sdp, bh);
bd = bh->b_private; bd = bh->b_private;
/* If this buffer is in the AIL and it has already been written /* If this buffer is in the AIL and it has already been written
* to in-place disk block, remove it from the AIL. * to in-place disk block, remove it from the AIL.
@ -209,10 +209,7 @@ static void gfs2_end_log_write(struct bio *bio)
if (!cmpxchg(&sdp->sd_log_error, 0, err)) if (!cmpxchg(&sdp->sd_log_error, 0, err))
fs_err(sdp, "Error %d writing to journal, jid=%u\n", fs_err(sdp, "Error %d writing to journal, jid=%u\n",
err, sdp->sd_jdesc->jd_jid); err, sdp->sd_jdesc->jd_jid);
gfs2_withdraw_delayed(sdp); gfs2_withdraw(sdp);
/* prevent more writes to the journal */
clear_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags);
wake_up(&sdp->sd_logd_waitq);
} }
bio_for_each_segment_all(bvec, bio, iter_all) { bio_for_each_segment_all(bvec, bio, iter_all) {
@ -487,7 +484,7 @@ static struct bio *gfs2_chain_bio(struct bio *prev, unsigned int nr_iovecs)
new = bio_alloc(prev->bi_bdev, nr_iovecs, prev->bi_opf, GFP_NOIO); new = bio_alloc(prev->bi_bdev, nr_iovecs, prev->bi_opf, GFP_NOIO);
bio_clone_blkg_association(new, prev); bio_clone_blkg_association(new, prev);
new->bi_iter.bi_sector = bio_end_sector(prev); new->bi_iter.bi_sector = bio_end_sector(prev);
bio_chain(new, prev); bio_chain(prev, new);
submit_bio(prev); submit_bio(prev);
return new; return new;
} }
@ -562,8 +559,7 @@ int gfs2_find_jhead(struct gfs2_jdesc *jd, struct gfs2_log_header_host *head)
bio = gfs2_log_alloc_bio(sdp, dblock, gfs2_end_log_read); bio = gfs2_log_alloc_bio(sdp, dblock, gfs2_end_log_read);
bio->bi_opf = REQ_OP_READ; bio->bi_opf = REQ_OP_READ;
add_block_to_new_bio: add_block_to_new_bio:
if (!bio_add_folio(bio, folio, bsize, off)) bio_add_folio_nofail(bio, folio, bsize, off);
BUG();
block_added: block_added:
off += bsize; off += bsize;
if (off == folio_size(folio)) if (off == folio_size(folio))

13
fs/gfs2/meta_io.c
View File

@ -263,8 +263,7 @@ int gfs2_meta_read(struct gfs2_glock *gl, u64 blkno, int flags,
struct buffer_head *bh, *bhs[2]; struct buffer_head *bh, *bhs[2];
int num = 0; int num = 0;
if (gfs2_withdrawing_or_withdrawn(sdp) && if (gfs2_withdrawn(sdp)) {
!gfs2_withdraw_in_prog(sdp)) {
*bhp = NULL; *bhp = NULL;
return -EIO; return -EIO;
} }
@ -303,7 +302,7 @@ int gfs2_meta_read(struct gfs2_glock *gl, u64 blkno, int flags,
if (unlikely(!buffer_uptodate(bh))) { if (unlikely(!buffer_uptodate(bh))) {
struct gfs2_trans *tr = current->journal_info; struct gfs2_trans *tr = current->journal_info;
if (tr && test_bit(TR_TOUCHED, &tr->tr_flags)) if (tr && test_bit(TR_TOUCHED, &tr->tr_flags))
gfs2_io_error_bh_wd(sdp, bh); gfs2_io_error_bh(sdp, bh);
brelse(bh); brelse(bh);
*bhp = NULL; *bhp = NULL;
return -EIO; return -EIO;
@ -322,8 +321,7 @@ int gfs2_meta_read(struct gfs2_glock *gl, u64 blkno, int flags,
int gfs2_meta_wait(struct gfs2_sbd *sdp, struct buffer_head *bh) int gfs2_meta_wait(struct gfs2_sbd *sdp, struct buffer_head *bh)
{ {
if (gfs2_withdrawing_or_withdrawn(sdp) && if (gfs2_withdrawn(sdp))
!gfs2_withdraw_in_prog(sdp))
return -EIO; return -EIO;
wait_on_buffer(bh); wait_on_buffer(bh);
@ -331,11 +329,10 @@ int gfs2_meta_wait(struct gfs2_sbd *sdp, struct buffer_head *bh)
if (!buffer_uptodate(bh)) { if (!buffer_uptodate(bh)) {
struct gfs2_trans *tr = current->journal_info; struct gfs2_trans *tr = current->journal_info;
if (tr && test_bit(TR_TOUCHED, &tr->tr_flags)) if (tr && test_bit(TR_TOUCHED, &tr->tr_flags))
gfs2_io_error_bh_wd(sdp, bh); gfs2_io_error_bh(sdp, bh);
return -EIO; return -EIO;
} }
if (gfs2_withdrawing_or_withdrawn(sdp) && if (gfs2_withdrawn(sdp))
!gfs2_withdraw_in_prog(sdp))
return -EIO; return -EIO;
return 0; return 0;

37
fs/gfs2/ops_fstype.c
View File

@ -60,6 +60,7 @@ static void gfs2_tune_init(struct gfs2_tune *gt)
gt->gt_new_files_jdata = 0; gt->gt_new_files_jdata = 0;
gt->gt_max_readahead = BIT(18); gt->gt_max_readahead = BIT(18);
gt->gt_complain_secs = 10; gt->gt_complain_secs = 10;
gt->gt_withdraw_helper_timeout = 5;
} }
void free_sbd(struct gfs2_sbd *sdp) void free_sbd(struct gfs2_sbd *sdp)
@ -92,7 +93,7 @@ static struct gfs2_sbd *init_sbd(struct super_block *sb)
init_waitqueue_head(&sdp->sd_async_glock_wait); init_waitqueue_head(&sdp->sd_async_glock_wait);
atomic_set(&sdp->sd_glock_disposal, 0); atomic_set(&sdp->sd_glock_disposal, 0);
init_completion(&sdp->sd_locking_init); init_completion(&sdp->sd_locking_init);
init_completion(&sdp->sd_wdack); init_completion(&sdp->sd_withdraw_helper);
spin_lock_init(&sdp->sd_statfs_spin); spin_lock_init(&sdp->sd_statfs_spin);
spin_lock_init(&sdp->sd_rindex_spin); spin_lock_init(&sdp->sd_rindex_spin);
@ -370,7 +371,7 @@ static int init_locking(struct gfs2_sbd *sdp, struct gfs2_holder *mount_gh,
error = gfs2_glock_nq_num(sdp, error = gfs2_glock_nq_num(sdp,
GFS2_MOUNT_LOCK, &gfs2_nondisk_glops, GFS2_MOUNT_LOCK, &gfs2_nondisk_glops,
LM_ST_EXCLUSIVE, LM_ST_EXCLUSIVE,
LM_FLAG_NOEXP | GL_NOCACHE | GL_NOPID, LM_FLAG_RECOVER | GL_NOCACHE | GL_NOPID,
mount_gh); mount_gh);
if (error) { if (error) {
fs_err(sdp, "can't acquire mount glock: %d\n", error); fs_err(sdp, "can't acquire mount glock: %d\n", error);
@ -380,7 +381,7 @@ static int init_locking(struct gfs2_sbd *sdp, struct gfs2_holder *mount_gh,
error = gfs2_glock_nq_num(sdp, error = gfs2_glock_nq_num(sdp,
GFS2_LIVE_LOCK, &gfs2_nondisk_glops, GFS2_LIVE_LOCK, &gfs2_nondisk_glops,
LM_ST_SHARED, LM_ST_SHARED,
LM_FLAG_NOEXP | GL_EXACT | GL_NOPID, LM_FLAG_RECOVER | GL_EXACT | GL_NOPID,
&sdp->sd_live_gh); &sdp->sd_live_gh);
if (error) { if (error) {
fs_err(sdp, "can't acquire live glock: %d\n", error); fs_err(sdp, "can't acquire live glock: %d\n", error);
@ -542,8 +543,6 @@ static int gfs2_jindex_hold(struct gfs2_sbd *sdp, struct gfs2_holder *ji_gh)
mutex_lock(&sdp->sd_jindex_mutex); mutex_lock(&sdp->sd_jindex_mutex);
for (;;) { for (;;) {
struct gfs2_inode *jip;
error = gfs2_glock_nq_init(dip->i_gl, LM_ST_SHARED, 0, ji_gh); error = gfs2_glock_nq_init(dip->i_gl, LM_ST_SHARED, 0, ji_gh);
if (error) if (error)
break; break;
@ -584,8 +583,6 @@ static int gfs2_jindex_hold(struct gfs2_sbd *sdp, struct gfs2_holder *ji_gh)
d_mark_dontcache(jd->jd_inode); d_mark_dontcache(jd->jd_inode);
spin_lock(&sdp->sd_jindex_spin); spin_lock(&sdp->sd_jindex_spin);
jd->jd_jid = sdp->sd_journals++; jd->jd_jid = sdp->sd_journals++;
jip = GFS2_I(jd->jd_inode);
jd->jd_no_addr = jip->i_no_addr;
list_add_tail(&jd->jd_list, &sdp->sd_jindex_list); list_add_tail(&jd->jd_list, &sdp->sd_jindex_list);
spin_unlock(&sdp->sd_jindex_spin); spin_unlock(&sdp->sd_jindex_spin);
} }
@ -745,7 +742,7 @@ static int init_journal(struct gfs2_sbd *sdp, int undo)
error = gfs2_glock_nq_num(sdp, sdp->sd_lockstruct.ls_jid, error = gfs2_glock_nq_num(sdp, sdp->sd_lockstruct.ls_jid,
&gfs2_journal_glops, &gfs2_journal_glops,
LM_ST_EXCLUSIVE, LM_ST_EXCLUSIVE,
LM_FLAG_NOEXP | GL_NOCACHE | GL_NOPID, LM_FLAG_RECOVER | GL_NOPID,
&sdp->sd_journal_gh); &sdp->sd_journal_gh);
if (error) { if (error) {
fs_err(sdp, "can't acquire journal glock: %d\n", error); fs_err(sdp, "can't acquire journal glock: %d\n", error);
@ -753,9 +750,8 @@ static int init_journal(struct gfs2_sbd *sdp, int undo)
} }
ip = GFS2_I(sdp->sd_jdesc->jd_inode); ip = GFS2_I(sdp->sd_jdesc->jd_inode);
sdp->sd_jinode_gl = ip->i_gl;
error = gfs2_glock_nq_init(ip->i_gl, LM_ST_SHARED, error = gfs2_glock_nq_init(ip->i_gl, LM_ST_SHARED,
LM_FLAG_NOEXP | GL_EXACT | LM_FLAG_RECOVER | GL_EXACT |
GL_NOCACHE | GL_NOPID, GL_NOCACHE | GL_NOPID,
&sdp->sd_jinode_gh); &sdp->sd_jinode_gh);
if (error) { if (error) {
@ -821,13 +817,10 @@ static int init_journal(struct gfs2_sbd *sdp, int undo)
fail_statfs: fail_statfs:
uninit_statfs(sdp); uninit_statfs(sdp);
fail_jinode_gh: fail_jinode_gh:
/* A withdraw may have done dq/uninit so now we need to check it */ if (!sdp->sd_args.ar_spectator)
if (!sdp->sd_args.ar_spectator &&
gfs2_holder_initialized(&sdp->sd_jinode_gh))
gfs2_glock_dq_uninit(&sdp->sd_jinode_gh); gfs2_glock_dq_uninit(&sdp->sd_jinode_gh);
fail_journal_gh: fail_journal_gh:
if (!sdp->sd_args.ar_spectator && if (!sdp->sd_args.ar_spectator)
gfs2_holder_initialized(&sdp->sd_journal_gh))
gfs2_glock_dq_uninit(&sdp->sd_journal_gh); gfs2_glock_dq_uninit(&sdp->sd_journal_gh);
fail_jindex: fail_jindex:
gfs2_jindex_free(sdp); gfs2_jindex_free(sdp);
@ -1040,8 +1033,8 @@ hostdata_error:
void gfs2_lm_unmount(struct gfs2_sbd *sdp) void gfs2_lm_unmount(struct gfs2_sbd *sdp)
{ {
const struct lm_lockops *lm = sdp->sd_lockstruct.ls_ops; const struct lm_lockops *lm = sdp->sd_lockstruct.ls_ops;
if (!gfs2_withdrawing_or_withdrawn(sdp) && lm->lm_unmount) if (!gfs2_withdrawn(sdp) && lm->lm_unmount)
lm->lm_unmount(sdp); lm->lm_unmount(sdp, true);
} }
static int wait_on_journal(struct gfs2_sbd *sdp) static int wait_on_journal(struct gfs2_sbd *sdp)
@ -1183,7 +1176,7 @@ static int gfs2_fill_super(struct super_block *sb, struct fs_context *fc)
mapping = gfs2_aspace(sdp); mapping = gfs2_aspace(sdp);
mapping->a_ops = &gfs2_rgrp_aops; mapping->a_ops = &gfs2_rgrp_aops;
mapping_set_gfp_mask(mapping, GFP_NOFS); gfs2_setup_inode(sdp->sd_inode);
error = init_names(sdp, silent); error = init_names(sdp, silent);
if (error) if (error)
@ -1215,6 +1208,8 @@ static int gfs2_fill_super(struct super_block *sb, struct fs_context *fc)
if (error) if (error)
goto fail_debug; goto fail_debug;
INIT_WORK(&sdp->sd_withdraw_work, gfs2_withdraw_func);
error = init_locking(sdp, &mount_gh, DO); error = init_locking(sdp, &mount_gh, DO);
if (error) if (error)
goto fail_lm; goto fail_lm;
@ -1401,12 +1396,14 @@ static const struct constant_table gfs2_param_data[] = {
}; };
enum opt_errors { enum opt_errors {
Opt_errors_withdraw = GFS2_ERRORS_WITHDRAW, Opt_errors_withdraw = GFS2_ERRORS_WITHDRAW,
Opt_errors_panic = GFS2_ERRORS_PANIC, Opt_errors_deactivate = GFS2_ERRORS_DEACTIVATE,
Opt_errors_panic = GFS2_ERRORS_PANIC,
}; };
static const struct constant_table gfs2_param_errors[] = { static const struct constant_table gfs2_param_errors[] = {
{"withdraw", Opt_errors_withdraw }, {"withdraw", Opt_errors_withdraw },
{"deactivate", Opt_errors_deactivate },
{"panic", Opt_errors_panic }, {"panic", Opt_errors_panic },
{} {}
}; };

66
fs/gfs2/quota.c
View File

@ -125,7 +125,7 @@ static void gfs2_qd_dispose(struct gfs2_quota_data *qd)
hlist_bl_del_rcu(&qd->qd_hlist); hlist_bl_del_rcu(&qd->qd_hlist);
spin_unlock_bucket(qd->qd_hash); spin_unlock_bucket(qd->qd_hash);
if (!gfs2_withdrawing_or_withdrawn(sdp)) { if (!gfs2_withdrawn(sdp)) {
gfs2_assert_warn(sdp, !qd->qd_change); gfs2_assert_warn(sdp, !qd->qd_change);
gfs2_assert_warn(sdp, !qd->qd_slot_ref); gfs2_assert_warn(sdp, !qd->qd_slot_ref);
gfs2_assert_warn(sdp, !qd->qd_bh_count); gfs2_assert_warn(sdp, !qd->qd_bh_count);
@ -1551,27 +1551,13 @@ static void quotad_error(struct gfs2_sbd *sdp, const char *msg, int error)
{ {
if (error == 0 || error == -EROFS) if (error == 0 || error == -EROFS)
return; return;
if (!gfs2_withdrawing_or_withdrawn(sdp)) { if (!gfs2_withdrawn(sdp)) {
if (!cmpxchg(&sdp->sd_log_error, 0, error)) if (!cmpxchg(&sdp->sd_log_error, 0, error))
fs_err(sdp, "gfs2_quotad: %s error %d\n", msg, error); fs_err(sdp, "gfs2_quotad: %s error %d\n", msg, error);
wake_up(&sdp->sd_logd_waitq); wake_up(&sdp->sd_logd_waitq);
} }
} }
static void quotad_check_timeo(struct gfs2_sbd *sdp, const char *msg,
int (*fxn)(struct super_block *sb, int type),
unsigned long t, unsigned long *timeo,
unsigned int *new_timeo)
{
if (t >= *timeo) {
int error = fxn(sdp->sd_vfs, 0);
quotad_error(sdp, msg, error);
*timeo = gfs2_tune_get_i(&sdp->sd_tune, new_timeo) * HZ;
} else {
*timeo -= t;
}
}
void gfs2_wake_up_statfs(struct gfs2_sbd *sdp) { void gfs2_wake_up_statfs(struct gfs2_sbd *sdp) {
if (!sdp->sd_statfs_force_sync) { if (!sdp->sd_statfs_force_sync) {
sdp->sd_statfs_force_sync = 1; sdp->sd_statfs_force_sync = 1;
@ -1589,36 +1575,46 @@ void gfs2_wake_up_statfs(struct gfs2_sbd *sdp) {
int gfs2_quotad(void *data) int gfs2_quotad(void *data)
{ {
struct gfs2_sbd *sdp = data; struct gfs2_sbd *sdp = data;
struct gfs2_tune *tune = &sdp->sd_tune; unsigned long now = jiffies;
unsigned long statfs_timeo = 0; unsigned long statfs_deadline = now;
unsigned long quotad_timeo = 0; unsigned long quotad_deadline = now;
unsigned long t = 0;
set_freezable(); set_freezable();
while (!kthread_should_stop()) { while (!kthread_should_stop()) {
if (gfs2_withdrawing_or_withdrawn(sdp)) unsigned long t;
if (gfs2_withdrawn(sdp))
break; break;
/* Update the master statfs file */ now = jiffies;
if (sdp->sd_statfs_force_sync) { if (sdp->sd_statfs_force_sync ||
int error = gfs2_statfs_sync(sdp->sd_vfs, 0); time_after(now, statfs_deadline)) {
unsigned int quantum;
int error;
/* Update the master statfs file */
error = gfs2_statfs_sync(sdp->sd_vfs, 0);
quotad_error(sdp, "statfs", error); quotad_error(sdp, "statfs", error);
statfs_timeo = gfs2_tune_get(sdp, gt_statfs_quantum) * HZ;
quantum = gfs2_tune_get(sdp, gt_statfs_quantum);
statfs_deadline = now + quantum * HZ;
} }
else if (time_after(now, quotad_deadline)) {
quotad_check_timeo(sdp, "statfs", gfs2_statfs_sync, t, unsigned int quantum;
&statfs_timeo, int error;
&tune->gt_statfs_quantum);
/* Update quota file */ /* Update the quota file */
quotad_check_timeo(sdp, "sync", gfs2_quota_sync, t, error = gfs2_quota_sync(sdp->sd_vfs, 0);
&quotad_timeo, &tune->gt_quota_quantum); quotad_error(sdp, "sync", error);
t = min(quotad_timeo, statfs_timeo); quantum = gfs2_tune_get(sdp, gt_quota_quantum);
quotad_deadline = now + quantum * HZ;
}
t = wait_event_freezable_timeout(sdp->sd_quota_wait, t = min(statfs_deadline - now, quotad_deadline - now);
wait_event_freezable_timeout(sdp->sd_quota_wait,
sdp->sd_statfs_force_sync || sdp->sd_statfs_force_sync ||
gfs2_withdrawing_or_withdrawn(sdp) || gfs2_withdrawn(sdp) ||
kthread_should_stop(), kthread_should_stop(),
t); t);

8
fs/gfs2/recovery.c
View File

@ -408,7 +408,7 @@ void gfs2_recover_func(struct work_struct *work)
int error = 0; int error = 0;
int jlocked = 0; int jlocked = 0;
if (gfs2_withdrawing_or_withdrawn(sdp)) { if (gfs2_withdrawn(sdp)) {
fs_err(sdp, "jid=%u: Recovery not attempted due to withdraw.\n", fs_err(sdp, "jid=%u: Recovery not attempted due to withdraw.\n",
jd->jd_jid); jd->jd_jid);
goto fail; goto fail;
@ -424,7 +424,8 @@ void gfs2_recover_func(struct work_struct *work)
error = gfs2_glock_nq_num(sdp, jd->jd_jid, &gfs2_journal_glops, error = gfs2_glock_nq_num(sdp, jd->jd_jid, &gfs2_journal_glops,
LM_ST_EXCLUSIVE, LM_ST_EXCLUSIVE,
LM_FLAG_NOEXP | LM_FLAG_TRY | GL_NOCACHE, LM_FLAG_RECOVER | LM_FLAG_TRY |
GL_NOCACHE,
&j_gh); &j_gh);
switch (error) { switch (error) {
case 0: case 0:
@ -440,7 +441,8 @@ void gfs2_recover_func(struct work_struct *work)
} }
error = gfs2_glock_nq_init(ip->i_gl, LM_ST_SHARED, error = gfs2_glock_nq_init(ip->i_gl, LM_ST_SHARED,
LM_FLAG_NOEXP | GL_NOCACHE, &ji_gh); LM_FLAG_RECOVER | GL_NOCACHE,
&ji_gh);
if (error) if (error)
goto fail_gunlock_j; goto fail_gunlock_j;
} else { } else {

35
fs/gfs2/super.c
View File

@ -137,7 +137,7 @@ int gfs2_make_fs_rw(struct gfs2_sbd *sdp)
int error; int error;
j_gl->gl_ops->go_inval(j_gl, DIO_METADATA); j_gl->gl_ops->go_inval(j_gl, DIO_METADATA);
if (gfs2_withdrawing_or_withdrawn(sdp)) if (gfs2_withdrawn(sdp))
return -EIO; return -EIO;
if (sdp->sd_log_sequence == 0) { if (sdp->sd_log_sequence == 0) {
@ -147,7 +147,7 @@ int gfs2_make_fs_rw(struct gfs2_sbd *sdp)
} }
error = gfs2_quota_init(sdp); error = gfs2_quota_init(sdp);
if (!error && gfs2_withdrawing_or_withdrawn(sdp)) if (!error && gfs2_withdrawn(sdp))
error = -EIO; error = -EIO;
if (!error) if (!error)
set_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags); set_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags);
@ -351,7 +351,7 @@ static int gfs2_lock_fs_check_clean(struct gfs2_sbd *sdp)
gfs2_freeze_unlock(sdp); gfs2_freeze_unlock(sdp);
error = gfs2_glock_nq_init(sdp->sd_freeze_gl, LM_ST_EXCLUSIVE, error = gfs2_glock_nq_init(sdp->sd_freeze_gl, LM_ST_EXCLUSIVE,
LM_FLAG_NOEXP | GL_NOPID, LM_FLAG_RECOVER | GL_NOPID,
&sdp->sd_freeze_gh); &sdp->sd_freeze_gh);
if (error) if (error)
goto relock_shared; goto relock_shared;
@ -491,7 +491,7 @@ static void gfs2_dirty_inode(struct inode *inode, int flags)
if (unlikely(!ip->i_gl)) if (unlikely(!ip->i_gl))
return; return;
if (gfs2_withdrawing_or_withdrawn(sdp)) if (gfs2_withdrawn(sdp))
return; return;
if (!gfs2_glock_is_locked_by_me(ip->i_gl)) { if (!gfs2_glock_is_locked_by_me(ip->i_gl)) {
ret = gfs2_glock_nq_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, &gh); ret = gfs2_glock_nq_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, &gh);
@ -597,13 +597,13 @@ restart:
if (!sb_rdonly(sb)) if (!sb_rdonly(sb))
gfs2_make_fs_ro(sdp); gfs2_make_fs_ro(sdp);
else { else {
if (gfs2_withdrawing_or_withdrawn(sdp)) if (gfs2_withdrawn(sdp))
gfs2_destroy_threads(sdp); gfs2_destroy_threads(sdp);
gfs2_quota_cleanup(sdp); gfs2_quota_cleanup(sdp);
} }
WARN_ON(gfs2_withdrawing(sdp)); flush_work(&sdp->sd_withdraw_work);
/* At this point, we're through modifying the disk */ /* At this point, we're through modifying the disk */
@ -749,9 +749,7 @@ static int gfs2_freeze_super(struct super_block *sb, enum freeze_holder who,
break; break;
} }
error = gfs2_do_thaw(sdp, who, freeze_owner); (void)gfs2_do_thaw(sdp, who, freeze_owner);
if (error)
goto out;
if (error == -EBUSY) if (error == -EBUSY)
fs_err(sdp, "waiting for recovery before freeze\n"); fs_err(sdp, "waiting for recovery before freeze\n");
@ -778,7 +776,7 @@ static int gfs2_freeze_fs(struct super_block *sb)
if (test_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags)) { if (test_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags)) {
gfs2_log_flush(sdp, NULL, GFS2_LOG_HEAD_FLUSH_FREEZE | gfs2_log_flush(sdp, NULL, GFS2_LOG_HEAD_FLUSH_FREEZE |
GFS2_LFC_FREEZE_GO_SYNC); GFS2_LFC_FREEZE_GO_SYNC);
if (gfs2_withdrawing_or_withdrawn(sdp)) if (gfs2_withdrawn(sdp))
return -EIO; return -EIO;
} }
return 0; return 0;
@ -819,20 +817,6 @@ static int gfs2_thaw_super(struct super_block *sb, enum freeze_holder who,
return error; return error;
} }
void gfs2_thaw_freeze_initiator(struct super_block *sb)
{
struct gfs2_sbd *sdp = sb->s_fs_info;
mutex_lock(&sdp->sd_freeze_mutex);
if (!test_bit(SDF_FREEZE_INITIATOR, &sdp->sd_flags))
goto out;
gfs2_freeze_unlock(sdp);
out:
mutex_unlock(&sdp->sd_freeze_mutex);
}
/** /**
* statfs_slow_fill - fill in the sg for a given RG * statfs_slow_fill - fill in the sg for a given RG
* @rgd: the RG * @rgd: the RG
@ -1147,6 +1131,9 @@ static int gfs2_show_options(struct seq_file *s, struct dentry *root)
case GFS2_ERRORS_WITHDRAW: case GFS2_ERRORS_WITHDRAW:
state = "withdraw"; state = "withdraw";
break; break;
case GFS2_ERRORS_DEACTIVATE:
state = "deactivate";
break;
case GFS2_ERRORS_PANIC: case GFS2_ERRORS_PANIC:
state = "panic"; state = "panic";
break; break;

1
fs/gfs2/super.h
View File

@ -47,7 +47,6 @@ void gfs2_statfs_change_out(const struct gfs2_statfs_change_host *sc,
void update_statfs(struct gfs2_sbd *sdp, struct buffer_head *m_bh); void update_statfs(struct gfs2_sbd *sdp, struct buffer_head *m_bh);
int gfs2_statfs_sync(struct super_block *sb, int type); int gfs2_statfs_sync(struct super_block *sb, int type);
void gfs2_freeze_func(struct work_struct *work); void gfs2_freeze_func(struct work_struct *work);
void gfs2_thaw_freeze_initiator(struct super_block *sb);
void free_local_statfs_inodes(struct gfs2_sbd *sdp); void free_local_statfs_inodes(struct gfs2_sbd *sdp);
struct inode *find_local_statfs_inode(struct gfs2_sbd *sdp, struct inode *find_local_statfs_inode(struct gfs2_sbd *sdp,

64
fs/gfs2/sys.c
View File

@ -59,7 +59,7 @@ static struct kset *gfs2_kset;
static ssize_t id_show(struct gfs2_sbd *sdp, char *buf) static ssize_t id_show(struct gfs2_sbd *sdp, char *buf)
{ {
return snprintf(buf, PAGE_SIZE, "%u:%u\n", return sysfs_emit(buf, "%u:%u\n",
MAJOR(sdp->sd_vfs->s_dev), MINOR(sdp->sd_vfs->s_dev)); MAJOR(sdp->sd_vfs->s_dev), MINOR(sdp->sd_vfs->s_dev));
} }
@ -68,7 +68,7 @@ static ssize_t status_show(struct gfs2_sbd *sdp, char *buf)
unsigned long f = sdp->sd_flags; unsigned long f = sdp->sd_flags;
ssize_t s; ssize_t s;
s = snprintf(buf, PAGE_SIZE, s = sysfs_emit(buf,
"Journal Checked: %d\n" "Journal Checked: %d\n"
"Journal Live: %d\n" "Journal Live: %d\n"
"Journal ID: %d\n" "Journal ID: %d\n"
@ -84,10 +84,6 @@ static ssize_t status_show(struct gfs2_sbd *sdp, char *buf)
"Force AIL Flush: %d\n" "Force AIL Flush: %d\n"
"FS Freeze Initiator: %d\n" "FS Freeze Initiator: %d\n"
"FS Frozen: %d\n" "FS Frozen: %d\n"
"Withdrawing: %d\n"
"Withdraw In Prog: %d\n"
"Remote Withdraw: %d\n"
"Withdraw Recovery: %d\n"
"Killing: %d\n" "Killing: %d\n"
"sd_log_error: %d\n" "sd_log_error: %d\n"
"sd_log_flush_lock: %d\n" "sd_log_flush_lock: %d\n"
@ -117,10 +113,6 @@ static ssize_t status_show(struct gfs2_sbd *sdp, char *buf)
test_bit(SDF_FORCE_AIL_FLUSH, &f), test_bit(SDF_FORCE_AIL_FLUSH, &f),
test_bit(SDF_FREEZE_INITIATOR, &f), test_bit(SDF_FREEZE_INITIATOR, &f),
test_bit(SDF_FROZEN, &f), test_bit(SDF_FROZEN, &f),
test_bit(SDF_WITHDRAWING, &f),
test_bit(SDF_WITHDRAW_IN_PROG, &f),
test_bit(SDF_REMOTE_WITHDRAW, &f),
test_bit(SDF_WITHDRAW_RECOVERY, &f),
test_bit(SDF_KILL, &f), test_bit(SDF_KILL, &f),
sdp->sd_log_error, sdp->sd_log_error,
rwsem_is_locked(&sdp->sd_log_flush_lock), rwsem_is_locked(&sdp->sd_log_flush_lock),
@ -140,7 +132,7 @@ static ssize_t status_show(struct gfs2_sbd *sdp, char *buf)
static ssize_t fsname_show(struct gfs2_sbd *sdp, char *buf) static ssize_t fsname_show(struct gfs2_sbd *sdp, char *buf)
{ {
return snprintf(buf, PAGE_SIZE, "%s\n", sdp->sd_fsname); return sysfs_emit(buf, "%s\n", sdp->sd_fsname);
} }
static ssize_t uuid_show(struct gfs2_sbd *sdp, char *buf) static ssize_t uuid_show(struct gfs2_sbd *sdp, char *buf)
@ -150,7 +142,7 @@ static ssize_t uuid_show(struct gfs2_sbd *sdp, char *buf)
buf[0] = '\0'; buf[0] = '\0';
if (uuid_is_null(&s->s_uuid)) if (uuid_is_null(&s->s_uuid))
return 0; return 0;
return snprintf(buf, PAGE_SIZE, "%pUB\n", &s->s_uuid); return sysfs_emit(buf, "%pUB\n", &s->s_uuid);
} }
static ssize_t freeze_show(struct gfs2_sbd *sdp, char *buf) static ssize_t freeze_show(struct gfs2_sbd *sdp, char *buf)
@ -158,7 +150,7 @@ static ssize_t freeze_show(struct gfs2_sbd *sdp, char *buf)
struct super_block *sb = sdp->sd_vfs; struct super_block *sb = sdp->sd_vfs;
int frozen = (sb->s_writers.frozen == SB_UNFROZEN) ? 0 : 1; int frozen = (sb->s_writers.frozen == SB_UNFROZEN) ? 0 : 1;
return snprintf(buf, PAGE_SIZE, "%d\n", frozen); return sysfs_emit(buf, "%d\n", frozen);
} }
static ssize_t freeze_store(struct gfs2_sbd *sdp, const char *buf, size_t len) static ssize_t freeze_store(struct gfs2_sbd *sdp, const char *buf, size_t len)
@ -193,8 +185,8 @@ static ssize_t freeze_store(struct gfs2_sbd *sdp, const char *buf, size_t len)
static ssize_t withdraw_show(struct gfs2_sbd *sdp, char *buf) static ssize_t withdraw_show(struct gfs2_sbd *sdp, char *buf)
{ {
unsigned int b = gfs2_withdrawing_or_withdrawn(sdp); unsigned int b = gfs2_withdrawn(sdp);
return snprintf(buf, PAGE_SIZE, "%u\n", b); return sysfs_emit(buf, "%u\n", b);
} }
static ssize_t withdraw_store(struct gfs2_sbd *sdp, const char *buf, size_t len) static ssize_t withdraw_store(struct gfs2_sbd *sdp, const char *buf, size_t len)
@ -397,7 +389,7 @@ static struct kobj_type gfs2_ktype = {
static ssize_t proto_name_show(struct gfs2_sbd *sdp, char *buf) static ssize_t proto_name_show(struct gfs2_sbd *sdp, char *buf)
{ {
const struct lm_lockops *ops = sdp->sd_lockstruct.ls_ops; const struct lm_lockops *ops = sdp->sd_lockstruct.ls_ops;
return sprintf(buf, "%s\n", ops->lm_proto_name); return sysfs_emit(buf, "%s\n", ops->lm_proto_name);
} }
static ssize_t block_show(struct gfs2_sbd *sdp, char *buf) static ssize_t block_show(struct gfs2_sbd *sdp, char *buf)
@ -408,7 +400,7 @@ static ssize_t block_show(struct gfs2_sbd *sdp, char *buf)
if (test_bit(DFL_BLOCK_LOCKS, &ls->ls_recover_flags)) if (test_bit(DFL_BLOCK_LOCKS, &ls->ls_recover_flags))
val = 1; val = 1;
ret = sprintf(buf, "%d\n", val); ret = sysfs_emit(buf, "%d\n", val);
return ret; return ret;
} }
@ -433,33 +425,27 @@ static ssize_t block_store(struct gfs2_sbd *sdp, const char *buf, size_t len)
return len; return len;
} }
static ssize_t wdack_show(struct gfs2_sbd *sdp, char *buf) static ssize_t withdraw_helper_status_store(struct gfs2_sbd *sdp,
{ const char *buf,
int val = completion_done(&sdp->sd_wdack) ? 1 : 0; size_t len)
return sprintf(buf, "%d\n", val);
}
static ssize_t wdack_store(struct gfs2_sbd *sdp, const char *buf, size_t len)
{ {
int ret, val; int ret, val;
ret = kstrtoint(buf, 0, &val); ret = kstrtoint(buf, 0, &val);
if (ret) if (ret)
return ret; return ret;
if (val < 0 || val > 1)
if ((val == 1) &&
!strcmp(sdp->sd_lockstruct.ls_ops->lm_proto_name, "lock_dlm"))
complete(&sdp->sd_wdack);
else
return -EINVAL; return -EINVAL;
sdp->sd_withdraw_helper_status = val;
complete(&sdp->sd_withdraw_helper);
return len; return len;
} }
static ssize_t lkfirst_show(struct gfs2_sbd *sdp, char *buf) static ssize_t lkfirst_show(struct gfs2_sbd *sdp, char *buf)
{ {
struct lm_lockstruct *ls = &sdp->sd_lockstruct; struct lm_lockstruct *ls = &sdp->sd_lockstruct;
return sprintf(buf, "%d\n", ls->ls_first); return sysfs_emit(buf, "%d\n", ls->ls_first);
} }
static ssize_t lkfirst_store(struct gfs2_sbd *sdp, const char *buf, size_t len) static ssize_t lkfirst_store(struct gfs2_sbd *sdp, const char *buf, size_t len)
@ -492,7 +478,7 @@ out:
static ssize_t first_done_show(struct gfs2_sbd *sdp, char *buf) static ssize_t first_done_show(struct gfs2_sbd *sdp, char *buf)
{ {
struct lm_lockstruct *ls = &sdp->sd_lockstruct; struct lm_lockstruct *ls = &sdp->sd_lockstruct;
return sprintf(buf, "%d\n", !!test_bit(DFL_FIRST_MOUNT_DONE, &ls->ls_recover_flags)); return sysfs_emit(buf, "%d\n", !!test_bit(DFL_FIRST_MOUNT_DONE, &ls->ls_recover_flags));
} }
int gfs2_recover_set(struct gfs2_sbd *sdp, unsigned jid) int gfs2_recover_set(struct gfs2_sbd *sdp, unsigned jid)
@ -550,18 +536,18 @@ out:
static ssize_t recover_done_show(struct gfs2_sbd *sdp, char *buf) static ssize_t recover_done_show(struct gfs2_sbd *sdp, char *buf)
{ {
struct lm_lockstruct *ls = &sdp->sd_lockstruct; struct lm_lockstruct *ls = &sdp->sd_lockstruct;
return sprintf(buf, "%d\n", ls->ls_recover_jid_done); return sysfs_emit(buf, "%d\n", ls->ls_recover_jid_done);
} }
static ssize_t recover_status_show(struct gfs2_sbd *sdp, char *buf) static ssize_t recover_status_show(struct gfs2_sbd *sdp, char *buf)
{ {
struct lm_lockstruct *ls = &sdp->sd_lockstruct; struct lm_lockstruct *ls = &sdp->sd_lockstruct;
return sprintf(buf, "%d\n", ls->ls_recover_jid_status); return sysfs_emit(buf, "%d\n", ls->ls_recover_jid_status);
} }
static ssize_t jid_show(struct gfs2_sbd *sdp, char *buf) static ssize_t jid_show(struct gfs2_sbd *sdp, char *buf)
{ {
return sprintf(buf, "%d\n", sdp->sd_lockstruct.ls_jid); return sysfs_emit(buf, "%d\n", sdp->sd_lockstruct.ls_jid);
} }
static ssize_t jid_store(struct gfs2_sbd *sdp, const char *buf, size_t len) static ssize_t jid_store(struct gfs2_sbd *sdp, const char *buf, size_t len)
@ -599,7 +585,7 @@ static struct gfs2_attr gdlm_attr_##_name = __ATTR(_name,_mode,_show,_store)
GDLM_ATTR(proto_name, 0444, proto_name_show, NULL); GDLM_ATTR(proto_name, 0444, proto_name_show, NULL);
GDLM_ATTR(block, 0644, block_show, block_store); GDLM_ATTR(block, 0644, block_show, block_store);
GDLM_ATTR(withdraw, 0644, wdack_show, wdack_store); GDLM_ATTR(withdraw, 0200, NULL, withdraw_helper_status_store);
GDLM_ATTR(jid, 0644, jid_show, jid_store); GDLM_ATTR(jid, 0644, jid_show, jid_store);
GDLM_ATTR(first, 0644, lkfirst_show, lkfirst_store); GDLM_ATTR(first, 0644, lkfirst_show, lkfirst_store);
GDLM_ATTR(first_done, 0444, first_done_show, NULL); GDLM_ATTR(first_done, 0444, first_done_show, NULL);
@ -626,7 +612,7 @@ static struct attribute *lock_module_attrs[] = {
static ssize_t quota_scale_show(struct gfs2_sbd *sdp, char *buf) static ssize_t quota_scale_show(struct gfs2_sbd *sdp, char *buf)
{ {
return snprintf(buf, PAGE_SIZE, "%u %u\n", return sysfs_emit(buf, "%u %u\n",
sdp->sd_tune.gt_quota_scale_num, sdp->sd_tune.gt_quota_scale_num,
sdp->sd_tune.gt_quota_scale_den); sdp->sd_tune.gt_quota_scale_den);
} }
@ -679,7 +665,7 @@ static struct gfs2_attr tune_attr_##name = __ATTR(name, 0644, show, store)
#define TUNE_ATTR_2(name, store) \ #define TUNE_ATTR_2(name, store) \
static ssize_t name##_show(struct gfs2_sbd *sdp, char *buf) \ static ssize_t name##_show(struct gfs2_sbd *sdp, char *buf) \
{ \ { \
return snprintf(buf, PAGE_SIZE, "%u\n", sdp->sd_tune.gt_##name); \ return sysfs_emit(buf, "%u\n", sdp->sd_tune.gt_##name); \
} \ } \
TUNE_ATTR_3(name, name##_show, store) TUNE_ATTR_3(name, name##_show, store)
@ -698,6 +684,7 @@ TUNE_ATTR(statfs_slow, 0);
TUNE_ATTR(new_files_jdata, 0); TUNE_ATTR(new_files_jdata, 0);
TUNE_ATTR(statfs_quantum, 1); TUNE_ATTR(statfs_quantum, 1);
TUNE_ATTR_3(quota_scale, quota_scale_show, quota_scale_store); TUNE_ATTR_3(quota_scale, quota_scale_show, quota_scale_store);
TUNE_ATTR(withdraw_helper_timeout, 1);
static struct attribute *tune_attrs[] = { static struct attribute *tune_attrs[] = {
&tune_attr_quota_warn_period.attr, &tune_attr_quota_warn_period.attr,
@ -708,6 +695,7 @@ static struct attribute *tune_attrs[] = {
&tune_attr_statfs_quantum.attr, &tune_attr_statfs_quantum.attr,
&tune_attr_quota_scale.attr, &tune_attr_quota_scale.attr,
&tune_attr_new_files_jdata.attr, &tune_attr_new_files_jdata.attr,
&tune_attr_withdraw_helper_timeout.attr,
NULL, NULL,
}; };

1
fs/gfs2/trace_gfs2.h
View File

@ -59,7 +59,6 @@
{(1UL << GLF_LRU), "L" }, \ {(1UL << GLF_LRU), "L" }, \
{(1UL << GLF_OBJECT), "o" }, \ {(1UL << GLF_OBJECT), "o" }, \
{(1UL << GLF_BLOCKING), "b" }, \ {(1UL << GLF_BLOCKING), "b" }, \
{(1UL << GLF_UNLOCKED), "x" }, \
{(1UL << GLF_INSTANTIATE_NEEDED), "n" }, \ {(1UL << GLF_INSTANTIATE_NEEDED), "n" }, \
{(1UL << GLF_INSTANTIATE_IN_PROG), "N" }, \ {(1UL << GLF_INSTANTIATE_IN_PROG), "N" }, \
{(1UL << GLF_TRY_TO_EVICT), "e" }, \ {(1UL << GLF_TRY_TO_EVICT), "e" }, \

32
fs/gfs2/trans.c
View File

@ -49,7 +49,7 @@ int __gfs2_trans_begin(struct gfs2_trans *tr, struct gfs2_sbd *sdp,
} }
BUG_ON(blocks == 0 && revokes == 0); BUG_ON(blocks == 0 && revokes == 0);
if (!test_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags)) if (gfs2_withdrawn(sdp))
return -EROFS; return -EROFS;
tr->tr_ip = ip; tr->tr_ip = ip;
@ -85,25 +85,30 @@ int __gfs2_trans_begin(struct gfs2_trans *tr, struct gfs2_sbd *sdp,
*/ */
down_read(&sdp->sd_log_flush_lock); down_read(&sdp->sd_log_flush_lock);
if (unlikely(!test_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags)))
goto out_not_live;
if (gfs2_log_try_reserve(sdp, tr, &extra_revokes)) if (gfs2_log_try_reserve(sdp, tr, &extra_revokes))
goto reserved; goto reserved;
up_read(&sdp->sd_log_flush_lock); up_read(&sdp->sd_log_flush_lock);
gfs2_log_reserve(sdp, tr, &extra_revokes); gfs2_log_reserve(sdp, tr, &extra_revokes);
down_read(&sdp->sd_log_flush_lock); down_read(&sdp->sd_log_flush_lock);
if (unlikely(!test_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags))) {
revokes = tr->tr_revokes + extra_revokes;
gfs2_log_release_revokes(sdp, revokes);
gfs2_log_release(sdp, tr->tr_reserved);
goto out_not_live;
}
reserved: reserved:
gfs2_log_release_revokes(sdp, extra_revokes); gfs2_log_release_revokes(sdp, extra_revokes);
if (unlikely(!test_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags))) {
gfs2_log_release_revokes(sdp, tr->tr_revokes);
up_read(&sdp->sd_log_flush_lock);
gfs2_log_release(sdp, tr->tr_reserved);
sb_end_intwrite(sdp->sd_vfs);
return -EROFS;
}
current->journal_info = tr; current->journal_info = tr;
return 0; return 0;
out_not_live:
up_read(&sdp->sd_log_flush_lock);
sb_end_intwrite(sdp->sd_vfs);
return -EROFS;
} }
int gfs2_trans_begin(struct gfs2_sbd *sdp, unsigned int blocks, int gfs2_trans_begin(struct gfs2_sbd *sdp, unsigned int blocks,
@ -255,7 +260,6 @@ void gfs2_trans_add_meta(struct gfs2_glock *gl, struct buffer_head *bh)
struct gfs2_bufdata *bd; struct gfs2_bufdata *bd;
struct gfs2_meta_header *mh; struct gfs2_meta_header *mh;
struct gfs2_trans *tr = current->journal_info; struct gfs2_trans *tr = current->journal_info;
bool withdraw = false;
lock_buffer(bh); lock_buffer(bh);
if (buffer_pinned(bh)) { if (buffer_pinned(bh)) {
@ -289,14 +293,14 @@ void gfs2_trans_add_meta(struct gfs2_glock *gl, struct buffer_head *bh)
(unsigned long long)bd->bd_bh->b_blocknr); (unsigned long long)bd->bd_bh->b_blocknr);
BUG(); BUG();
} }
if (gfs2_withdrawing_or_withdrawn(sdp)) { if (gfs2_withdrawn(sdp)) {
fs_info(sdp, "GFS2:adding buf while withdrawn! 0x%llx\n", fs_info(sdp, "GFS2:adding buf while withdrawn! 0x%llx\n",
(unsigned long long)bd->bd_bh->b_blocknr); (unsigned long long)bd->bd_bh->b_blocknr);
goto out_unlock; goto out_unlock;
} }
if (unlikely(sb->s_writers.frozen == SB_FREEZE_COMPLETE)) { if (unlikely(sb->s_writers.frozen == SB_FREEZE_COMPLETE)) {
fs_info(sdp, "GFS2:adding buf while frozen\n"); fs_info(sdp, "GFS2:adding buf while frozen\n");
withdraw = true; gfs2_withdraw(sdp);
goto out_unlock; goto out_unlock;
} }
gfs2_pin(sdp, bd->bd_bh); gfs2_pin(sdp, bd->bd_bh);
@ -306,8 +310,6 @@ void gfs2_trans_add_meta(struct gfs2_glock *gl, struct buffer_head *bh)
tr->tr_num_buf_new++; tr->tr_num_buf_new++;
out_unlock: out_unlock:
gfs2_log_unlock(sdp); gfs2_log_unlock(sdp);
if (withdraw)
gfs2_assert_withdraw(sdp, 0);
out: out:
unlock_buffer(bh); unlock_buffer(bh);
} }

328
fs/gfs2/util.c
View File

@ -58,7 +58,7 @@ int check_journal_clean(struct gfs2_sbd *sdp, struct gfs2_jdesc *jd,
struct gfs2_inode *ip; struct gfs2_inode *ip;
ip = GFS2_I(jd->jd_inode); ip = GFS2_I(jd->jd_inode);
error = gfs2_glock_nq_init(ip->i_gl, LM_ST_SHARED, LM_FLAG_NOEXP | error = gfs2_glock_nq_init(ip->i_gl, LM_ST_SHARED, LM_FLAG_RECOVER |
GL_EXACT | GL_NOCACHE, &j_gh); GL_EXACT | GL_NOCACHE, &j_gh);
if (error) { if (error) {
if (verbose) if (verbose)
@ -99,7 +99,7 @@ out_unlock:
*/ */
int gfs2_freeze_lock_shared(struct gfs2_sbd *sdp) int gfs2_freeze_lock_shared(struct gfs2_sbd *sdp)
{ {
int flags = LM_FLAG_NOEXP | GL_EXACT; int flags = LM_FLAG_RECOVER | GL_EXACT;
int error; int error;
error = gfs2_glock_nq_init(sdp->sd_freeze_gl, LM_ST_SHARED, flags, error = gfs2_glock_nq_init(sdp->sd_freeze_gl, LM_ST_SHARED, flags,
@ -115,182 +115,32 @@ void gfs2_freeze_unlock(struct gfs2_sbd *sdp)
gfs2_glock_dq_uninit(&sdp->sd_freeze_gh); gfs2_glock_dq_uninit(&sdp->sd_freeze_gh);
} }
static void signal_our_withdraw(struct gfs2_sbd *sdp) static void do_withdraw(struct gfs2_sbd *sdp)
{ {
struct gfs2_glock *live_gl = sdp->sd_live_gh.gh_gl; down_write(&sdp->sd_log_flush_lock);
struct inode *inode; if (!test_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags)) {
struct gfs2_inode *ip; up_write(&sdp->sd_log_flush_lock);
struct gfs2_glock *i_gl;
u64 no_formal_ino;
int ret = 0;
int tries;
if (test_bit(SDF_NORECOVERY, &sdp->sd_flags) || !sdp->sd_jdesc)
return; return;
}
clear_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags);
up_write(&sdp->sd_log_flush_lock);
gfs2_ail_drain(sdp); /* frees all transactions */ gfs2_ail_drain(sdp); /* frees all transactions */
inode = sdp->sd_jdesc->jd_inode;
ip = GFS2_I(inode);
i_gl = ip->i_gl;
no_formal_ino = ip->i_no_formal_ino;
/* Prevent any glock dq until withdraw recovery is complete */ wake_up(&sdp->sd_logd_waitq);
set_bit(SDF_WITHDRAW_RECOVERY, &sdp->sd_flags); wake_up(&sdp->sd_quota_wait);
/*
* Don't tell dlm we're bailing until we have no more buffers in the
* wind. If journal had an IO error, the log code should just purge
* the outstanding buffers rather than submitting new IO. Making the
* file system read-only will flush the journal, etc.
*
* During a normal unmount, gfs2_make_fs_ro calls gfs2_log_shutdown
* which clears SDF_JOURNAL_LIVE. In a withdraw, we must not write
* any UNMOUNT log header, so we can't call gfs2_log_shutdown, and
* therefore we need to clear SDF_JOURNAL_LIVE manually.
*/
clear_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags);
if (!sb_rdonly(sdp->sd_vfs)) {
bool locked = mutex_trylock(&sdp->sd_freeze_mutex);
wake_up(&sdp->sd_logd_waitq); wait_event_timeout(sdp->sd_log_waitq,
wake_up(&sdp->sd_quota_wait); gfs2_log_is_empty(sdp),
HZ * 5);
wait_event_timeout(sdp->sd_log_waitq, sdp->sd_vfs->s_flags |= SB_RDONLY;
gfs2_log_is_empty(sdp),
HZ * 5);
sdp->sd_vfs->s_flags |= SB_RDONLY;
if (locked)
mutex_unlock(&sdp->sd_freeze_mutex);
/*
* Dequeue any pending non-system glock holders that can no
* longer be granted because the file system is withdrawn.
*/
gfs2_gl_dq_holders(sdp);
}
if (sdp->sd_lockstruct.ls_ops->lm_lock == NULL) { /* lock_nolock */
if (!ret)
ret = -EIO;
clear_bit(SDF_WITHDRAW_RECOVERY, &sdp->sd_flags);
goto skip_recovery;
}
/*
* Drop the glock for our journal so another node can recover it.
*/
if (gfs2_holder_initialized(&sdp->sd_journal_gh)) {
gfs2_glock_dq_wait(&sdp->sd_journal_gh);
gfs2_holder_uninit(&sdp->sd_journal_gh);
}
sdp->sd_jinode_gh.gh_flags |= GL_NOCACHE;
gfs2_glock_dq(&sdp->sd_jinode_gh);
gfs2_thaw_freeze_initiator(sdp->sd_vfs);
wait_on_bit(&i_gl->gl_flags, GLF_DEMOTE, TASK_UNINTERRUPTIBLE);
/* /*
* holder_uninit to force glock_put, to force dlm to let go * Dequeue any pending non-system glock holders that can no
* longer be granted because the file system is withdrawn.
*/ */
gfs2_holder_uninit(&sdp->sd_jinode_gh); gfs2_withdraw_glocks(sdp);
/*
* Note: We need to be careful here:
* Our iput of jd_inode will evict it. The evict will dequeue its
* glock, but the glock dq will wait for the withdraw unless we have
* exception code in glock_dq.
*/
iput(inode);
sdp->sd_jdesc->jd_inode = NULL;
/*
* Wait until the journal inode's glock is freed. This allows try locks
* on other nodes to be successful, otherwise we remain the owner of
* the glock as far as dlm is concerned.
*/
if (i_gl->gl_ops->go_unlocked) {
set_bit(GLF_UNLOCKED, &i_gl->gl_flags);
wait_on_bit(&i_gl->gl_flags, GLF_UNLOCKED, TASK_UNINTERRUPTIBLE);
}
/*
* Dequeue the "live" glock, but keep a reference so it's never freed.
*/
gfs2_glock_hold(live_gl);
gfs2_glock_dq_wait(&sdp->sd_live_gh);
/*
* We enqueue the "live" glock in EX so that all other nodes
* get a demote request and act on it. We don't really want the
* lock in EX, so we send a "try" lock with 1CB to produce a callback.
*/
fs_warn(sdp, "Requesting recovery of jid %d.\n",
sdp->sd_lockstruct.ls_jid);
gfs2_holder_reinit(LM_ST_EXCLUSIVE,
LM_FLAG_TRY_1CB | LM_FLAG_NOEXP | GL_NOPID,
&sdp->sd_live_gh);
msleep(GL_GLOCK_MAX_HOLD);
/*
* This will likely fail in a cluster, but succeed standalone:
*/
ret = gfs2_glock_nq(&sdp->sd_live_gh);
gfs2_glock_put(live_gl); /* drop extra reference we acquired */
clear_bit(SDF_WITHDRAW_RECOVERY, &sdp->sd_flags);
/*
* If we actually got the "live" lock in EX mode, there are no other
* nodes available to replay our journal.
*/
if (ret == 0) {
fs_warn(sdp, "No other mounters found.\n");
/*
* We are about to release the lockspace. By keeping live_gl
* locked here, we ensure that the next mounter coming along
* will be a "first" mounter which will perform recovery.
*/
goto skip_recovery;
}
/*
* At this point our journal is evicted, so we need to get a new inode
* for it. Once done, we need to call gfs2_find_jhead which
* calls gfs2_map_journal_extents to map it for us again.
*
* Note that we don't really want it to look up a FREE block. The
* GFS2_BLKST_FREE simply overrides a block check in gfs2_inode_lookup
* which would otherwise fail because it requires grabbing an rgrp
* glock, which would fail with -EIO because we're withdrawing.
*/
inode = gfs2_inode_lookup(sdp->sd_vfs, DT_UNKNOWN,
sdp->sd_jdesc->jd_no_addr, no_formal_ino,
GFS2_BLKST_FREE);
if (IS_ERR(inode)) {
fs_warn(sdp, "Reprocessing of jid %d failed with %ld.\n",
sdp->sd_lockstruct.ls_jid, PTR_ERR(inode));
goto skip_recovery;
}
sdp->sd_jdesc->jd_inode = inode;
d_mark_dontcache(inode);
/*
* Now wait until recovery is complete.
*/
for (tries = 0; tries < 10; tries++) {
ret = check_journal_clean(sdp, sdp->sd_jdesc, false);
if (!ret)
break;
msleep(HZ);
fs_warn(sdp, "Waiting for journal recovery jid %d.\n",
sdp->sd_lockstruct.ls_jid);
}
skip_recovery:
if (!ret)
fs_warn(sdp, "Journal recovery complete for jid %d.\n",
sdp->sd_lockstruct.ls_jid);
else
fs_warn(sdp, "Journal recovery skipped for jid %d until next "
"mount.\n", sdp->sd_lockstruct.ls_jid);
fs_warn(sdp, "Glock dequeues delayed: %lu\n", sdp->sd_glock_dqs_held);
sdp->sd_glock_dqs_held = 0;
wake_up_bit(&sdp->sd_flags, SDF_WITHDRAW_RECOVERY);
} }
void gfs2_lm(struct gfs2_sbd *sdp, const char *fmt, ...) void gfs2_lm(struct gfs2_sbd *sdp, const char *fmt, ...)
@ -309,43 +159,104 @@ void gfs2_lm(struct gfs2_sbd *sdp, const char *fmt, ...)
va_end(args); va_end(args);
} }
void gfs2_withdraw(struct gfs2_sbd *sdp) /**
* gfs2_offline_uevent - run gfs2_withdraw_helper
* @sdp: The GFS2 superblock
*/
static bool gfs2_offline_uevent(struct gfs2_sbd *sdp)
{ {
struct lm_lockstruct *ls = &sdp->sd_lockstruct;
long timeout;
/* Skip protocol "lock_nolock" which doesn't require shared storage. */
if (!ls->ls_ops->lm_lock)
return false;
/*
* The gfs2_withdraw_helper replies by writing one of the following
* status codes to "/sys$DEVPATH/lock_module/withdraw":
*
* 0 - The shared block device has been marked inactive. Future write
* operations will fail.
*
* 1 - The shared block device may still be active and carry out
* write operations.
*
* If the "offline" uevent isn't reacted upon in time, the event
* handler is assumed to have failed.
*/
sdp->sd_withdraw_helper_status = -1;
kobject_uevent(&sdp->sd_kobj, KOBJ_OFFLINE);
timeout = gfs2_tune_get(sdp, gt_withdraw_helper_timeout) * HZ;
wait_for_completion_timeout(&sdp->sd_withdraw_helper, timeout);
if (sdp->sd_withdraw_helper_status == -1) {
fs_err(sdp, "%s timed out\n", "gfs2_withdraw_helper");
} else {
fs_err(sdp, "%s %s with status %d\n",
"gfs2_withdraw_helper",
sdp->sd_withdraw_helper_status == 0 ?
"succeeded" : "failed",
sdp->sd_withdraw_helper_status);
}
return sdp->sd_withdraw_helper_status == 0;
}
void gfs2_withdraw_func(struct work_struct *work)
{
struct gfs2_sbd *sdp = container_of(work, struct gfs2_sbd, sd_withdraw_work);
struct lm_lockstruct *ls = &sdp->sd_lockstruct; struct lm_lockstruct *ls = &sdp->sd_lockstruct;
const struct lm_lockops *lm = ls->ls_ops; const struct lm_lockops *lm = ls->ls_ops;
bool device_inactive;
if (sdp->sd_args.ar_errors == GFS2_ERRORS_WITHDRAW) { if (test_bit(SDF_KILL, &sdp->sd_flags))
unsigned long old = READ_ONCE(sdp->sd_flags), new; return;
do { BUG_ON(sdp->sd_args.ar_debug);
if (old & BIT(SDF_WITHDRAWN)) {
wait_on_bit(&sdp->sd_flags,
SDF_WITHDRAW_IN_PROG,
TASK_UNINTERRUPTIBLE);
return;
}
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); * Try to deactivate the shared block device so that no more I/O will
* go through. If successful, we can immediately trigger remote
* recovery. Otherwise, we must first empty out all our local caches.
*/
signal_our_withdraw(sdp); device_inactive = gfs2_offline_uevent(sdp);
kobject_uevent(&sdp->sd_kobj, KOBJ_OFFLINE); if (sdp->sd_args.ar_errors == GFS2_ERRORS_DEACTIVATE && !device_inactive)
panic("GFS2: fsid=%s: panic requested\n", sdp->sd_fsname);
if (!strcmp(sdp->sd_lockstruct.ls_ops->lm_proto_name, "lock_dlm")) if (lm->lm_unmount) {
wait_for_completion(&sdp->sd_wdack); if (device_inactive) {
lm->lm_unmount(sdp, false);
if (lm->lm_unmount) { do_withdraw(sdp);
fs_err(sdp, "telling LM to unmount\n"); } else {
lm->lm_unmount(sdp); do_withdraw(sdp);
lm->lm_unmount(sdp, false);
} }
fs_err(sdp, "File system withdrawn\n"); } else {
do_withdraw(sdp);
}
fs_err(sdp, "file system withdrawn\n");
}
void gfs2_withdraw(struct gfs2_sbd *sdp)
{
if (sdp->sd_args.ar_errors == GFS2_ERRORS_WITHDRAW ||
sdp->sd_args.ar_errors == GFS2_ERRORS_DEACTIVATE) {
if (test_and_set_bit(SDF_WITHDRAWN, &sdp->sd_flags))
return;
dump_stack(); dump_stack();
clear_bit(SDF_WITHDRAW_IN_PROG, &sdp->sd_flags); /*
smp_mb__after_atomic(); * There is no need to withdraw when the superblock hasn't been
wake_up_bit(&sdp->sd_flags, SDF_WITHDRAW_IN_PROG); * fully initialized, yet.
*/
if (!(sdp->sd_vfs->s_flags & SB_BORN))
return;
fs_err(sdp, "about to withdraw this file system\n");
schedule_work(&sdp->sd_withdraw_work);
return;
} }
if (sdp->sd_args.ar_errors == GFS2_ERRORS_PANIC) if (sdp->sd_args.ar_errors == GFS2_ERRORS_PANIC)
@ -357,10 +268,9 @@ void gfs2_withdraw(struct gfs2_sbd *sdp)
*/ */
void gfs2_assert_withdraw_i(struct gfs2_sbd *sdp, char *assertion, void gfs2_assert_withdraw_i(struct gfs2_sbd *sdp, char *assertion,
const char *function, char *file, unsigned int line, const char *function, char *file, unsigned int line)
bool delayed)
{ {
if (gfs2_withdrawing_or_withdrawn(sdp)) if (gfs2_withdrawn(sdp))
return; return;
fs_err(sdp, fs_err(sdp,
@ -368,17 +278,7 @@ void gfs2_assert_withdraw_i(struct gfs2_sbd *sdp, char *assertion,
"function = %s, file = %s, line = %u\n", "function = %s, file = %s, line = %u\n",
assertion, function, file, line); assertion, function, file, line);
/* gfs2_withdraw(sdp);
* If errors=panic was specified on mount, it won't help to delay the
* withdraw.
*/
if (sdp->sd_args.ar_errors == GFS2_ERRORS_PANIC)
delayed = false;
if (delayed)
gfs2_withdraw_delayed(sdp);
else
gfs2_withdraw(sdp);
dump_stack(); dump_stack();
} }
@ -520,22 +420,18 @@ void gfs2_io_error_i(struct gfs2_sbd *sdp, const char *function, char *file,
} }
/* /*
* gfs2_io_error_bh_i - Flag a buffer I/O error * gfs2_io_error_bh_i - Flag a buffer I/O error and withdraw
* @withdraw: withdraw the filesystem
*/ */
void gfs2_io_error_bh_i(struct gfs2_sbd *sdp, struct buffer_head *bh, void gfs2_io_error_bh_i(struct gfs2_sbd *sdp, struct buffer_head *bh,
const char *function, char *file, unsigned int line, const char *function, char *file, unsigned int line)
bool withdraw)
{ {
if (gfs2_withdrawing_or_withdrawn(sdp)) if (gfs2_withdrawn(sdp))
return; return;
fs_err(sdp, "fatal: I/O error - " fs_err(sdp, "fatal: I/O error - "
"block = %llu, " "block = %llu, "
"function = %s, file = %s, line = %u\n", "function = %s, file = %s, line = %u\n",
(unsigned long long)bh->b_blocknr, function, file, line); (unsigned long long)bh->b_blocknr, function, file, line);
if (withdraw) gfs2_withdraw(sdp);
gfs2_withdraw(sdp);
} }

56
fs/gfs2/util.h
View File

@ -37,24 +37,14 @@ do { \
void gfs2_assert_withdraw_i(struct gfs2_sbd *sdp, char *assertion, void gfs2_assert_withdraw_i(struct gfs2_sbd *sdp, char *assertion,
const char *function, char *file, unsigned int line, const char *function, char *file, unsigned int line);
bool delayed);
#define gfs2_assert_withdraw(sdp, assertion) \ #define gfs2_assert_withdraw(sdp, assertion) \
({ \ ({ \
bool _bool = (assertion); \ bool _bool = (assertion); \
if (unlikely(!_bool)) \ if (unlikely(!_bool)) \
gfs2_assert_withdraw_i((sdp), #assertion, \ gfs2_assert_withdraw_i((sdp), #assertion, \
__func__, __FILE__, __LINE__, false); \ __func__, __FILE__, __LINE__); \
!_bool; \
})
#define gfs2_assert_withdraw_delayed(sdp, assertion) \
({ \
bool _bool = (assertion); \
if (unlikely(!_bool)) \
gfs2_assert_withdraw_i((sdp), #assertion, \
__func__, __FILE__, __LINE__, true); \
!_bool; \ !_bool; \
}) })
@ -161,14 +151,10 @@ gfs2_io_error_i((sdp), __func__, __FILE__, __LINE__)
void gfs2_io_error_bh_i(struct gfs2_sbd *sdp, struct buffer_head *bh, void gfs2_io_error_bh_i(struct gfs2_sbd *sdp, struct buffer_head *bh,
const char *function, char *file, unsigned int line, const char *function, char *file, unsigned int line);
bool withdraw);
#define gfs2_io_error_bh_wd(sdp, bh) \
gfs2_io_error_bh_i((sdp), (bh), __func__, __FILE__, __LINE__, true)
#define gfs2_io_error_bh(sdp, bh) \ #define gfs2_io_error_bh(sdp, bh) \
gfs2_io_error_bh_i((sdp), (bh), __func__, __FILE__, __LINE__, false) gfs2_io_error_bh_i((sdp), (bh), __func__, __FILE__, __LINE__)
extern struct kmem_cache *gfs2_glock_cachep; extern struct kmem_cache *gfs2_glock_cachep;
@ -193,38 +179,12 @@ static inline unsigned int gfs2_tune_get_i(struct gfs2_tune *gt,
} }
/** /**
* gfs2_withdraw_delayed - withdraw as soon as possible without deadlocks * gfs2_withdrawn - test whether the file system is withdrawn
* @sdp: the superblock * @sdp: the superblock
*/ */
static inline void gfs2_withdraw_delayed(struct gfs2_sbd *sdp) static inline bool gfs2_withdrawn(struct gfs2_sbd *sdp)
{ {
set_bit(SDF_WITHDRAWING, &sdp->sd_flags); return unlikely(test_bit(SDF_WITHDRAWN, &sdp->sd_flags));
}
/**
* gfs2_withdrawing_or_withdrawn - test whether the file system is withdrawing
* or withdrawn
* @sdp: the superblock
*/
static inline bool gfs2_withdrawing_or_withdrawn(struct gfs2_sbd *sdp)
{
return unlikely(test_bit(SDF_WITHDRAWN, &sdp->sd_flags) ||
test_bit(SDF_WITHDRAWING, &sdp->sd_flags));
}
/**
* gfs2_withdrawing - check if a withdraw is pending
* @sdp: the superblock
*/
static inline bool gfs2_withdrawing(struct gfs2_sbd *sdp)
{
return unlikely(test_bit(SDF_WITHDRAWING, &sdp->sd_flags) &&
!test_bit(SDF_WITHDRAWN, &sdp->sd_flags));
}
static inline bool gfs2_withdraw_in_prog(struct gfs2_sbd *sdp)
{
return unlikely(test_bit(SDF_WITHDRAW_IN_PROG, &sdp->sd_flags));
} }
#define gfs2_tune_get(sdp, field) \ #define gfs2_tune_get(sdp, field) \
@ -232,6 +192,8 @@ gfs2_tune_get_i(&(sdp)->sd_tune, &(sdp)->sd_tune.field)
__printf(2, 3) __printf(2, 3)
void gfs2_lm(struct gfs2_sbd *sdp, const char *fmt, ...); void gfs2_lm(struct gfs2_sbd *sdp, const char *fmt, ...);
void gfs2_withdraw_func(struct work_struct *work);
void gfs2_withdraw(struct gfs2_sbd *sdp); void gfs2_withdraw(struct gfs2_sbd *sdp);
#endif /* __UTIL_DOT_H__ */ #endif /* __UTIL_DOT_H__ */