Improve wait_on_bit interface.
Subsystems (22)
cover Improve wait_on_bit interface.
Hi Linus
(hoping to get through your spam filter :-)
I wonder if you would consider applying these patches directly?
I originally sent them to Peter Zijlstra who was happy with them and they went into "tip" for a while
http://lkml.iu.edu/hypermail/linux/kernel/1405.2/01678.html
however other code (nfs and cifs) added new users of the old wait_on_bit interface causing conflicts, so Ingo had to remove them. At that time I suggested:
Quoted 5 lines
How about you drop my patch for now, we wait for -rc1 to come out, then I submit a new version against -rc1 and we get that into -rc2. That should minimise such conflicts.
Does that work for you?
and Ingo replied
Quoted 2 lines
Sure, that sounds like a good approach, if Linus doesn't object.
which I took to mean that I could forward revised patches though the 'tip' tree at that time.
I did resend after rc2 (missed rc1 - on leave) and have heard nothing from Ingo since despite a ping. So maybe he meant I should submit them directly to you.
I would really like at least the first of these to go in before 3.16 else other people could add calls using the old interface and cause the same problems again. Having both of them go in would make me very happy as I could then submit the change to NFS which needs the new wait_on_bit() functionality to device loop-back NFS deadlocks.
Thanks a lot,
NeilBrown
---
NeilBrown (2):
SCHED: remove proliferation of wait_on_bit action functions.
SCHED: allow wait_on_bit_action functions to support a timeout.
Documentation/filesystems/caching/operations.txt | 2 drivers/md/dm-bufio.c | 41 ++----- drivers/md/dm-snap.c | 10 -- drivers/media/usb/dvb-usb-v2/dvb_usb_core.c | 12 -- fs/btrfs/extent_io.c | 10 -- fs/buffer.c | 11 -- fs/cifs/connect.c | 10 -- fs/cifs/file.c | 9 -- fs/cifs/inode.c | 6 + fs/cifs/misc.c | 2 fs/fs-writeback.c | 3 - fs/fscache/cookie.c | 7 + fs/fscache/internal.h | 2 fs/fscache/main.c | 18 --- fs/fscache/page.c | 4 - fs/gfs2/glock.c | 25 ---- fs/gfs2/lock_dlm.c | 8 - fs/gfs2/ops_fstype.c | 11 -- fs/gfs2/recovery.c | 8 - fs/gfs2/super.c | 8 - fs/inode.c | 7 - fs/jbd2/transaction.c | 10 -- fs/nfs/file.c | 4 - fs/nfs/filelayout/filelayoutdev.c | 4 - fs/nfs/inode.c | 6 + fs/nfs/internal.h | 2 fs/nfs/nfs4state.c | 4 - fs/nfs/pagelist.c | 14 +- fs/nfs/pnfs.c | 2 fs/nfs/write.c | 4 - include/linux/sunrpc/sched.h | 2 include/linux/wait.h | 125 +++++++++++++++++++++- include/linux/writeback.h | 3 - kernel/ptrace.c | 8 - kernel/sched/wait.c | 30 ++++- mm/filemap.c | 20 +--- mm/ksm.c | 8 - net/bluetooth/hci_core.c | 8 - net/sunrpc/sched.c | 4 - security/keys/gc.c | 11 -- security/keys/request_key.c | 23 ---- 41 files changed, 214 insertions(+), 292 deletions(-)
Signature · 1 line
Signature
PATCH 1/2 SCHED: remove proliferation of wait_on_bit action functions.
The current "wait_on_bit" interface requires an 'action' function to be provided which does the actual waiting.
There are over 20 such functions, many of them identical. Most cases can be satisfied by one of just two functions, one which uses io_schedule() and one which just uses schedule().
So:
Rename wait_on_bit and wait_on_bit_lock to
wait_on_bit_action and wait_on_bit_lock_action
to make it explicit that they need an action function.
Introduce new wait_on_bit{,_lock} and wait_on_bit{,_lock}_io which are *not* given an action function but implicitly use a standard one.
The decision to error-out if a signal is pending is now made based on the 'mode' argument rather than being encoded in the action function.
All instances of the old wait_on_bit and wait_on_bit_lock which can use the new version have been changed accordingly and their action functions have been discarded. wait_on_bit{_lock} does not return any specific error code in the event of a signal so the caller must check for non-zero and interpolate their own error code as appropriate.
The wait_on_bit() call in __fscache_wait_on_invalidate() was ambiguous as it specified TASK_UNINTERRUPTIBLE but used fscache_wait_bit_interruptible as an action function.
David Howells confirms this should be uniformly "uninterruptible"
The main remaining user of wait_on_bit{,_lock}_action is NFS which needs to use a freezer-aware schedule() call.
A comment in fs/gfs2/glock.c notes that having multiple 'action' functions is useful as they display differently in the 'wchan' field of 'ps'. (and /proc/$PID/wchan).
As the new bit_wait{,_io} functions are tagged "__sched", they will not show up at all, but something higher in the stack. So the distinction will still be visible, only with different function names (gds2_glock_wait versus gfs2_glock_dq_wait in the gfs2/glock.c case).
Since first version of this patch (against 3.15) two new action functions appeared, on in NFS and one in CIFS. CIFS also now uses an action function that makes the same freezer aware schedule call as NFS.
Signed-off-by NeilBrown <[email protected]>
Cc Oleg Nesterov <[email protected]>
Cc Peter Zijlstra <[email protected]>
--- Documentation/filesystems/caching/operations.txt | 2 drivers/md/dm-bufio.c | 41 +++----- drivers/md/dm-snap.c | 10 -- drivers/media/usb/dvb-usb-v2/dvb_usb_core.c | 12 -- fs/btrfs/extent_io.c | 10 -- fs/buffer.c | 11 -- fs/cifs/connect.c | 10 -- fs/cifs/file.c | 9 -- fs/cifs/inode.c | 4 - fs/cifs/misc.c | 2 fs/fs-writeback.c | 3 - fs/fscache/cookie.c | 7 + fs/fscache/internal.h | 2 fs/fscache/main.c | 18 --- fs/fscache/page.c | 4 - fs/gfs2/glock.c | 25 ----- fs/gfs2/lock_dlm.c | 8 -- fs/gfs2/ops_fstype.c | 11 -- fs/gfs2/recovery.c | 8 -- fs/gfs2/super.c | 8 -- fs/inode.c | 7 - fs/jbd2/transaction.c | 10 -- fs/nfs/file.c | 4 - fs/nfs/filelayout/filelayoutdev.c | 4 - fs/nfs/inode.c | 4 - fs/nfs/nfs4state.c | 4 - fs/nfs/pagelist.c | 12 -- fs/nfs/pnfs.c | 2 fs/nfs/write.c | 4 - include/linux/wait.h | 115 +++++++++++++++++++++- include/linux/writeback.h | 3 - kernel/ptrace.c | 8 -- kernel/sched/wait.c | 18 +++ mm/filemap.c | 20 +--- mm/ksm.c | 8 -- net/bluetooth/hci_core.c | 8 -- security/keys/gc.c | 11 -- security/keys/request_key.c | 23 ---- 38 files changed, 195 insertions(+), 275 deletions(-)
Documentation/filesystems/caching/operations.txt +1 -1
@@ -90,7 +90,7 @@ operations: to be cleared before proceeding: wait_on_bit(&op->flags, FSCACHE_OP_WAITING, - fscache_wait_bit, TASK_UNINTERRUPTIBLE); + TASK_UNINTERRUPTIBLE); (2) The operation may be fast asynchronous (FSCACHE_OP_FAST), in which case it
drivers/md/dm-bufio.c +14 -27
92 lines not on this page · raw message
drivers/md/dm-snap.c +1 -9
@@ -1032,21 +1032,13 @@ static void start_merge(struct dm_snapshot *s)snapshot_merge_next_chunks (s); } -static int wait_schedule (void *ptr) -{ -schedule (); - -return 0 ; -} -/* * Stop the merging process and wait until it finishes. */ static void stop_merge (struct dm_snapshot *s) {set_bit (SHUTDOWN_MERGE , &s->state_bits); -wait_on_bit (&s->state_bits,RUNNING_MERGE , wait_schedule, -TASK_UNINTERRUPTIBLE ); +wait_on_bit (&s->state_bits,RUNNING_MERGE ,TASK_UNINTERRUPTIBLE );clear_bit (SHUTDOWN_MERGE , &s->state_bits); }
drivers/media/usb/dvb-usb-v2/dvb_usb_core.c +2 -10
@@ -253,13 +253,6 @@ static int dvb_usbv2_adapter_stream_exit(struct dvb_usb_adapter *adap)return usb_urb_exitv2 (&adap->stream); } -static int wait_schedule (void *ptr) -{ -schedule (); - -return 0 ; -} -static int dvb_usb_start_feed (struct dvb_demux_feed *dvbdmxfeed) {struct dvb_usb_adapter *adap = dvbdmxfeed->demux->priv;
@@ -273,8 +266,7 @@ static int dvb_usb_start_feed(struct dvb_demux_feed *dvbdmxfeed) dvbdmxfeed->pid, dvbdmxfeed->index);/* wait init is done */ -wait_on_bit (&adap->state_bits,ADAP_INIT , wait_schedule, -TASK_UNINTERRUPTIBLE ); +wait_on_bit (&adap->state_bits,ADAP_INIT ,TASK_UNINTERRUPTIBLE );if (adap->active_fe ==-1 )return -EINVAL ;
@@ -568,7 +560,7 @@ static int dvb_usb_fe_sleep(struct dvb_frontend *fe)if (!adap->suspend_resume_active) {set_bit (ADAP_SLEEP , &adap->state_bits); -wait_on_bit (&adap->state_bits,ADAP_STREAMING , wait_schedule, +wait_on_bit (&adap->state_bits,ADAP_STREAMING ,TASK_UNINTERRUPTIBLE ); }
fs/btrfs/extent_io.c +2 -8
@@ -3437,16 +3437,10 @@ done_unlocked:return 0 ; } -static int eb_wait (void *word) -{ -io_schedule (); -return 0 ; -} -void wait_on_extent_buffer_writeback (struct extent_buffer *eb) { -wait_on_bit (&eb->bflags,EXTENT_BUFFER_WRITEBACK , eb_wait, -TASK_UNINTERRUPTIBLE ); +wait_on_bit_io (&eb->bflags,EXTENT_BUFFER_WRITEBACK , +TASK_UNINTERRUPTIBLE ); }static noinline_for_stack int
fs/buffer.c +2 -9
@@ -61,16 +61,9 @@ inline void touch_buffer(struct buffer_head *bh) }EXPORT_SYMBOL (touch_buffer); -static int sleep_on_buffer (void *word) -{ -io_schedule (); -return 0 ; -} -void __lock_buffer (struct buffer_head *bh) { -wait_on_bit_lock (&bh->b_state, BH_Lock, sleep_on_buffer, -TASK_UNINTERRUPTIBLE ); +wait_on_bit_lock_io (&bh->b_state, BH_Lock,TASK_UNINTERRUPTIBLE ); }EXPORT_SYMBOL (__lock_buffer);
@@ -123,7 +116,7 @@ EXPORT_SYMBOL(buffer_check_dirty_writeback);*/ void __wait_on_buffer (struct buffer_head * bh) { -wait_on_bit (&bh->b_state, BH_Lock, sleep_on_buffer,TASK_UNINTERRUPTIBLE ); +wait_on_bit_io (&bh->b_state, BH_Lock,TASK_UNINTERRUPTIBLE ); }EXPORT_SYMBOL (__wait_on_buffer);
fs/cifs/connect.c +1 -9
@@ -3934,13 +3934,6 @@ cifs_sb_master_tcon(struct cifs_sb_info *cifs_sb)return tlink_tcon (cifs_sb_master_tlink (cifs_sb)); } -static int -cifs_sb_tcon_pending_wait (void *unused) -{ -schedule (); -return signal_pending (current) ? -ERESTARTSYS :0 ; -} -/* find and return a tlink with given uid */ static struct tcon_link *tlink_rb_search (struct rb_root *root,kuid_t uid)
@@ -4039,11 +4032,10 @@ cifs_sb_tlink(struct cifs_sb_info *cifs_sb) }else { wait_for_construction: ret =wait_on_bit (&tlink->tl_flags,TCON_LINK_PENDING , - cifs_sb_tcon_pending_wait,TASK_INTERRUPTIBLE );if (ret) {cifs_put_tlink (tlink); -return ERR_PTR (ret); +return ERR_PTR (-ERESTARTSYS ); }/* if it's good, return it */
fs/cifs/file.c +1 -8
@@ -3618,13 +3618,6 @@ static int cifs_launder_page(struct page *page)return rc; } -static int -cifs_pending_writers_wait (void *unused) -{ -schedule (); -return 0 ; -} -void cifs_oplock_break (struct work_struct *work) { struct cifsFileInfo *cfile = container_of(work, struct cifsFileInfo,
@@ -3636,7 +3629,7 @@ void cifs_oplock_break(struct work_struct *work)int rc =0 ;wait_on_bit (&cinode->flags,CIFS_INODE_PENDING_WRITERS , - cifs_pending_writers_wait,TASK_UNINTERRUPTIBLE ); +TASK_UNINTERRUPTIBLE ); server->ops->downgrade_oplock (server, cinode,test_bit (CIFS_INODE_DOWNGRADE_OPLOCK_TO_L2 , &cinode->flags));
fs/cifs/inode.c +2 -2
@@ -1794,8 +1794,8 @@ cifs_revalidate_mapping(struct inode *inode)int rc;unsigned long *flags = &CIFS_I (inode)->flags; - rc =wait_on_bit_lock (flags,CIFS_INO_LOCK , cifs_wait_bit_killable, -TASK_KILLABLE ); + rc =wait_on_bit_lock_action (flags,CIFS_INO_LOCK , cifs_wait_bit_killable, +TASK_KILLABLE );if (rc)return rc;
fs/cifs/misc.c +1 -1
@@ -582,7 +582,7 @@ int cifs_get_writer(struct cifsInodeInfo *cinode) start: rc =wait_on_bit (&cinode->flags,CIFS_INODE_PENDING_OPLOCK_BREAK , - cifs_oplock_break_wait,TASK_KILLABLE ); +TASK_KILLABLE );if (rc)return rc;
fs/fs-writeback.c +2 -1
@@ -342,7 +342,8 @@ static void __inode_wait_for_writeback(struct inode *inode) wqh =bit_waitqueue (&inode->i_state, __I_SYNC);while (inode->i_state &I_SYNC ) {spin_unlock (&inode->i_lock); -__wait_on_bit (wqh, &wq, inode_wait,TASK_UNINTERRUPTIBLE ); +__wait_on_bit (wqh, &wq, bit_wait, +TASK_UNINTERRUPTIBLE );spin_lock (&inode->i_lock); } }
fs/fscache/cookie.c +3 -4
@@ -160,7 +160,7 @@ void __fscache_enable_cookie(struct fscache_cookie *cookie,_enter ("%p" , cookie);wait_on_bit_lock (&cookie->flags,FSCACHE_COOKIE_ENABLEMENT_LOCK , - fscache_wait_bit,TASK_UNINTERRUPTIBLE ); +TASK_UNINTERRUPTIBLE );if (test_bit (FSCACHE_COOKIE_ENABLED , &cookie->flags)) goto out_unlock;
@@ -255,7 +255,7 @@ static int fscache_acquire_non_index_cookie(struct fscache_cookie *cookie)if (!fscache_defer_lookup) {_debug ("non-deferred lookup %p" , &cookie->flags);wait_on_bit (&cookie->flags,FSCACHE_COOKIE_LOOKING_UP , - fscache_wait_bit,TASK_UNINTERRUPTIBLE ); +TASK_UNINTERRUPTIBLE );_debug ("complete" );if (test_bit (FSCACHE_COOKIE_UNAVAILABLE , &cookie->flags)) goto unavailable;
@@ -463,7 +463,6 @@ void __fscache_wait_on_invalidate(struct fscache_cookie *cookie)_enter ("%p" , cookie);wait_on_bit (&cookie->flags,FSCACHE_COOKIE_INVALIDATING , - fscache_wait_bit_interruptible,TASK_UNINTERRUPTIBLE );_leave ("" );
@@ -525,7 +524,7 @@ void __fscache_disable_cookie(struct fscache_cookie *cookie, bool invalidate) }wait_on_bit_lock (&cookie->flags,FSCACHE_COOKIE_ENABLEMENT_LOCK , - fscache_wait_bit,TASK_UNINTERRUPTIBLE ); +TASK_UNINTERRUPTIBLE );if (!test_and_clear_bit (FSCACHE_COOKIE_ENABLED , &cookie->flags)) goto out_unlock_enable;
fs/fscache/internal.h +0 -2
@@ -97,8 +97,6 @@ static inline bool fscache_object_congested(void)return workqueue_congested (WORK_CPU_UNBOUND , fscache_object_wq); } -extern int fscache_wait_bit (void *); -extern int fscache_wait_bit_interruptible (void *);extern int fscache_wait_atomic_t (atomic_t *);/*
fs/fscache/main.c +0 -18
@@ -197,24 +197,6 @@ static void __exit fscache_exit(void)module_exit (fscache_exit);/* -* wait_on_bit() sleep function for uninterruptible waiting -*/ -int fscache_wait_bit (void *flags) -{ -schedule (); -return 0 ; -} - -/* -* wait_on_bit() sleep function for interruptible waiting -*/ -int fscache_wait_bit_interruptible (void *flags) -{ -schedule (); -return signal_pending (current); -} - -/* * wait_on_atomic_t() sleep function for uninterruptible waiting */ int fscache_wait_atomic_t (atomic_t *p)
fs/fscache/page.c +1 -3
@@ -298,7 +298,6 @@ int fscache_wait_for_deferred_lookup(struct fscache_cookie *cookie) jif = jiffies;if (wait_on_bit (&cookie->flags,FSCACHE_COOKIE_LOOKING_UP , - fscache_wait_bit_interruptible,TASK_INTERRUPTIBLE ) !=0 ) {fscache_stat (&fscache_n_retrievals_intr);_leave (" = -ERESTARTSYS" );
@@ -342,7 +341,6 @@ int fscache_wait_for_operation_activation(struct fscache_object *object,if (stat_op_waits)fscache_stat (stat_op_waits);if (wait_on_bit (&op->flags,FSCACHE_OP_WAITING , - fscache_wait_bit_interruptible,TASK_INTERRUPTIBLE ) !=0 ) { ret =fscache_cancel_op (op, do_cancel);if (ret ==0 )
@@ -351,7 +349,7 @@ int fscache_wait_for_operation_activation(struct fscache_object *object,/* it's been removed from the pending queue by another party, * so we should get to run shortly */ wait_on_bit (&op->flags,FSCACHE_OP_WAITING , - fscache_wait_bit,TASK_UNINTERRUPTIBLE ); +TASK_UNINTERRUPTIBLE ); }_debug ("<<< GO" );
fs/gfs2/glock.c +2 -23
@@ -856,27 +856,6 @@ void gfs2_holder_uninit(struct gfs2_holder *gh) }/** -* gfs2_glock_holder_wait -* @word: unused -* -* This function and gfs2_glock_demote_wait both show up in the WCHAN -* field. Thus I've separated these otherwise identical functions in -* order to be more informative to the user. -*/ - -static int gfs2_glock_holder_wait (void *word) -{ -schedule (); -return 0 ; -} - -static int gfs2_glock_demote_wait (void *word) -{ -schedule (); -return 0 ; -} - -/** * gfs2_glock_wait - wait on a glock acquisition * @gh: the glock holder *
@@ -888,7 +867,7 @@ int gfs2_glock_wait(struct gfs2_holder *gh)unsigned long time1 = jiffies;might_sleep (); -wait_on_bit (&gh->gh_iflags,HIF_WAIT , gfs2_glock_holder_wait,TASK_UNINTERRUPTIBLE ); +wait_on_bit (&gh->gh_iflags,HIF_WAIT ,TASK_UNINTERRUPTIBLE );if (time_after (jiffies, time1 +HZ ))/* have we waited > a second? */ /* Lengthen the minimum hold time. */ gh->gh_gl->gl_hold_time = min(gh->gh_gl->gl_hold_time +
@@ -1128,7 +1107,7 @@ void gfs2_glock_dq_wait(struct gfs2_holder *gh)struct gfs2_glock *gl = gh->gh_gl;gfs2_glock_dq (gh);might_sleep (); -wait_on_bit (&gl->gl_flags,GLF_DEMOTE , gfs2_glock_demote_wait,TASK_UNINTERRUPTIBLE ); +wait_on_bit (&gl->gl_flags,GLF_DEMOTE ,TASK_UNINTERRUPTIBLE ); }/**
fs/gfs2/lock_dlm.c +1 -7
@@ -936,12 +936,6 @@ fail:return error; } -static int dlm_recovery_wait (void *word) -{ -schedule (); -return 0 ; -} -static int control_first_done (struct gfs2_sbd *sdp) {struct lm_lockstruct *ls = &sdp->sd_lockstruct;
@@ -976,7 +970,7 @@ restart:fs_info (sdp,"control_first_done wait gen %u\n" , start_gen);wait_on_bit (&ls->ls_recover_flags,DFL_DLM_RECOVERY , - dlm_recovery_wait,TASK_UNINTERRUPTIBLE ); +TASK_UNINTERRUPTIBLE ); goto restart; }
fs/gfs2/ops_fstype.c +2 -9
@@ -1024,20 +1024,13 @@ void gfs2_lm_unmount(struct gfs2_sbd *sdp) lm->lm_unmount (sdp); } -static int gfs2_journalid_wait (void *word) -{ -if (signal_pending (current)) -return -EINTR ; -schedule (); -return 0 ; -} -static int wait_on_journal (struct gfs2_sbd *sdp) {if (sdp->sd_lockstruct.ls_ops->lm_mount ==NULL )return 0 ; -return wait_on_bit (&sdp->sd_flags,SDF_NOJOURNALID , gfs2_journalid_wait,TASK_INTERRUPTIBLE ); +return wait_on_bit (&sdp->sd_flags,SDF_NOJOURNALID ,TASK_INTERRUPTIBLE ) + ? -EINTR :0 ; }void gfs2_online_uevent (struct gfs2_sbd *sdp)
fs/gfs2/recovery.c +1 -7
@@ -591,12 +591,6 @@ done:wake_up_bit (&jd->jd_flags,JDF_RECOVERY ); } -static int gfs2_recovery_wait (void *word) -{ -schedule (); -return 0 ; -} -int gfs2_recover_journal (struct gfs2_jdesc *jd,bool wait) {int rv;
@@ -609,7 +603,7 @@ int gfs2_recover_journal(struct gfs2_jdesc *jd, bool wait)BUG_ON (!rv);if (wait) -wait_on_bit (&jd->jd_flags,JDF_RECOVERY , gfs2_recovery_wait, +wait_on_bit (&jd->jd_flags,JDF_RECOVERY ,TASK_UNINTERRUPTIBLE );return wait ? jd->jd_recover_error :0 ;
fs/gfs2/super.c +1 -7
@@ -864,12 +864,6 @@ static int gfs2_make_fs_ro(struct gfs2_sbd *sdp)return error; } -static int gfs2_umount_recovery_wait (void *word) -{ -schedule (); -return 0 ; -} -/** * gfs2_put_super - Unmount the filesystem * @sb: The VFS superblock
@@ -894,7 +888,7 @@ restart:continue ;spin_unlock (&sdp->sd_jindex_spin);wait_on_bit (&jd->jd_flags,JDF_RECOVERY , - gfs2_umount_recovery_wait,TASK_UNINTERRUPTIBLE ); +TASK_UNINTERRUPTIBLE ); goto restart; }spin_unlock (&sdp->sd_jindex_spin);
fs/inode.c +0 -7
@@ -1695,13 +1695,6 @@ int inode_needs_sync(struct inode *inode) }EXPORT_SYMBOL (inode_needs_sync); -int inode_wait (void *word) -{ -schedule (); -return 0 ; -} -EXPORT_SYMBOL (inode_wait); -/* * If we try to find an inode in the inode hash while it is being * deleted, we have to wait until the filesystem completes its
fs/jbd2/transaction.c +2 -8
@@ -763,12 +763,6 @@ static void warn_dirty_buffer(struct buffer_head *bh) bdevname(bh->b_bdev, b), (unsigned long long)bh->b_blocknr); } -static int sleep_on_shadow_bh (void *word) -{ -io_schedule (); -return 0 ; -} -/* * If the buffer is already part of the current transaction, then there * is nothing we need to do. If it is already part of a prior
@@ -906,8 +900,8 @@ repeat:if (buffer_shadow (bh)) {JBUFFER_TRACE (jh,"on shadow: sleep" );jbd_unlock_bh_state (bh); -wait_on_bit (&bh->b_state, BH_Shadow, - sleep_on_shadow_bh,TASK_UNINTERRUPTIBLE ); +wait_on_bit_io (&bh->b_state, BH_Shadow, +TASK_UNINTERRUPTIBLE ); goto repeat; }
fs/nfs/file.c +2 -2
@@ -361,8 +361,8 @@ start:* Prevent starvation issues if someone is doing a consistency * sync-to-disk */ - ret =wait_on_bit (&NFS_I (mapping->host)->flags,NFS_INO_FLUSHING , - nfs_wait_bit_killable,TASK_KILLABLE ); + ret =wait_on_bit_action (&NFS_I (mapping->host)->flags,NFS_INO_FLUSHING , + nfs_wait_bit_killable,TASK_KILLABLE );if (ret)return ret;
fs/nfs/filelayout/filelayoutdev.c +2 -2
@@ -783,8 +783,8 @@ nfs4_fl_select_ds_fh(struct pnfs_layout_segment *lseg, u32 j)static void nfs4_wait_ds_connect (struct nfs4_pnfs_ds *ds) {might_sleep (); -wait_on_bit (&ds->ds_state,NFS4DS_CONNECTING , - nfs_wait_bit_killable,TASK_KILLABLE ); +wait_on_bit_action (&ds->ds_state,NFS4DS_CONNECTING , + nfs_wait_bit_killable,TASK_KILLABLE ); }static void nfs4_clear_ds_conn_bit (struct nfs4_pnfs_ds *ds)
fs/nfs/inode.c +2 -2
@@ -1074,8 +1074,8 @@ int nfs_revalidate_mapping(struct inode *inode, struct address_space *mapping)* the bit lock here if it looks like we're going to be doing that. */ for (;;) { - ret =wait_on_bit (bitlock,NFS_INO_INVALIDATING , - nfs_wait_bit_killable,TASK_KILLABLE ); + ret =wait_on_bit_action (bitlock,NFS_INO_INVALIDATING , + nfs_wait_bit_killable,TASK_KILLABLE );if (ret) goto out;spin_lock (&inode->i_lock);
fs/nfs/nfs4state.c +2 -2
@@ -1251,8 +1251,8 @@ int nfs4_wait_clnt_recover(struct nfs_client *clp)might_sleep ();atomic_inc (&clp->cl_count); - res =wait_on_bit (&clp->cl_state,NFS4CLNT_MANAGER_RUNNING , - nfs_wait_bit_killable,TASK_KILLABLE ); + res =wait_on_bit_action (&clp->cl_state,NFS4CLNT_MANAGER_RUNNING , + nfs_wait_bit_killable,TASK_KILLABLE );if (res) goto out;if (clp->cl_cons_state <0 )
fs/nfs/pagelist.c +2 -10
@@ -138,12 +138,6 @@ nfs_iocounter_wait(struct nfs_io_counter *c)return __nfs_iocounter_wait (c); } -static int nfs_wait_bit_uninterruptible (void *word) -{ -io_schedule (); -return 0 ; -} -/* * nfs_page_group_lock - lock the head of the page group * @req - request in group that is to be locked
@@ -158,7 +152,6 @@ nfs_page_group_lock(struct nfs_page *req)WARN_ON_ONCE (head != head->wb_head);wait_on_bit_lock (&head->wb_flags,PG_HEADLOCK , - nfs_wait_bit_uninterruptible,TASK_UNINTERRUPTIBLE ); }
@@ -425,9 +418,8 @@ void nfs_release_request(struct nfs_page *req)int nfs_wait_on_request (struct nfs_page *req) { -return wait_on_bit (&req->wb_flags,PG_BUSY , - nfs_wait_bit_uninterruptible, -TASK_UNINTERRUPTIBLE ); +return wait_on_bit_io (&req->wb_flags,PG_BUSY , +TASK_UNINTERRUPTIBLE ); }/*
fs/nfs/pnfs.c +1 -1
@@ -1885,7 +1885,7 @@ pnfs_layoutcommit_inode(struct inode *inode, bool sync)if (test_and_set_bit (NFS_INO_LAYOUTCOMMITTING , &nfsi->flags)) {if (!sync) goto out; - status =wait_on_bit_lock (&nfsi->flags, + status =wait_on_bit_lock_action (&nfsi->flags,NFS_INO_LAYOUTCOMMITTING , nfs_wait_bit_killable,TASK_KILLABLE );
fs/nfs/write.c +2 -2
@@ -397,7 +397,7 @@ int nfs_writepages(struct address_space *mapping, struct writeback_control *wbc)int err;/* Stop dirtying of new pages while we sync */ - err =wait_on_bit_lock (bitlock,NFS_INO_FLUSHING , + err =wait_on_bit_lock_action (bitlock,NFS_INO_FLUSHING , nfs_wait_bit_killable,TASK_KILLABLE );if (err) goto out_err;
@@ -1475,7 +1475,7 @@ int nfs_commit_inode(struct inode *inode, int how)return error;if (!may_wait) goto out_mark_dirty; - error =wait_on_bit (&NFS_I (inode)->flags, + error =wait_on_bit_action (&NFS_I (inode)->flags,NFS_INO_COMMIT , nfs_wait_bit_killable,TASK_KILLABLE );
include/linux/wait.h +111 -4
147 lines not on this page · raw message
include/linux/writeback.h +1 -2
@@ -90,7 +90,6 @@ struct writeback_control {* fs/fs-writeback.c */ struct bdi_writeback ; -int inode_wait (void *);void writeback_inodes_sb (struct super_block *,enum wb_reason reason);void writeback_inodes_sb_nr (struct super_block *,unsigned long nr,enum wb_reason reason);
@@ -105,7 +104,7 @@ void inode_wait_for_writeback(struct inode *inode);static inline void wait_on_inode (struct inode *inode) {might_sleep (); -wait_on_bit (&inode->i_state, __I_NEW, inode_wait,TASK_UNINTERRUPTIBLE ); +wait_on_bit (&inode->i_state, __I_NEW,TASK_UNINTERRUPTIBLE ); }/*
kernel/ptrace.c +1 -7
@@ -28,12 +28,6 @@#include <linux/compat.h> -static int ptrace_trapping_sleep_fn (void *flags) -{ -schedule (); -return 0 ; -} -/* * ptrace a task: make the debugger its new parent and * move it to the ptrace list.
@@ -371,7 +365,7 @@ unlock_creds: out:if (!retval) {wait_on_bit (&task->jobctl,JOBCTL_TRAPPING_BIT , - ptrace_trapping_sleep_fn,TASK_UNINTERRUPTIBLE ); +TASK_UNINTERRUPTIBLE );proc_ptrace_connector (task,PTRACE_ATTACH ); }
kernel/sched/wait.c +18 -0
@@ -502,3 +502,21 @@ void wake_up_atomic_t(atomic_t *p)__wake_up_bit (atomic_t_waitqueue (p), p,WAIT_ATOMIC_T_BIT_NR ); }EXPORT_SYMBOL (wake_up_atomic_t); + +__sched int bit_wait(void *word) +{ +if (signal_pending_state (current->state, current)) +return 1 ; +schedule (); +return 0 ; +} +EXPORT_SYMBOL (bit_wait); + +__sched int bit_wait_io(void *word) +{ +if (signal_pending_state (current->state, current)) +return 1 ; +io_schedule (); +return 0 ; +} +EXPORT_SYMBOL (bit_wait_io);
mm/filemap.c +4 -16
@@ -241,18 +241,6 @@ void delete_from_page_cache(struct page *page) }EXPORT_SYMBOL (delete_from_page_cache); -static int sleep_on_page (void *word) -{ -io_schedule (); -return 0 ; -} - -static int sleep_on_page_killable (void *word) -{ -sleep_on_page (word); -return fatal_signal_pending (current) ? -EINTR :0 ; -} -static int filemap_check_errors (struct address_space *mapping) {int ret =0 ;
@@ -692,7 +680,7 @@ void wait_on_page_bit(struct page *page, int bit_nr)DEFINE_WAIT_BIT (wait, &page->flags, bit_nr);if (test_bit (bit_nr, &page->flags)) -__wait_on_bit (page_waitqueue (page), &wait, sleep_on_page, +__wait_on_bit (page_waitqueue (page), &wait, bit_wait_io,TASK_UNINTERRUPTIBLE ); }EXPORT_SYMBOL (wait_on_page_bit);
@@ -705,7 +693,7 @@ int wait_on_page_bit_killable(struct page *page, int bit_nr)return 0 ;return __wait_on_bit (page_waitqueue (page), &wait, - sleep_on_page_killable,TASK_KILLABLE ); + bit_wait_io,TASK_KILLABLE ); }/**
@@ -806,7 +794,7 @@ void __lock_page(struct page *page) {DEFINE_WAIT_BIT (wait, &page->flags, PG_locked); -__wait_on_bit_lock (page_waitqueue (page), &wait, sleep_on_page, +__wait_on_bit_lock (page_waitqueue (page), &wait, bit_wait_io,TASK_UNINTERRUPTIBLE ); }EXPORT_SYMBOL (__lock_page);
@@ -816,7 +804,7 @@ int __lock_page_killable(struct page *page)DEFINE_WAIT_BIT (wait, &page->flags, PG_locked);return __wait_on_bit_lock (page_waitqueue (page), &wait, - sleep_on_page_killable,TASK_KILLABLE ); + bit_wait_io,TASK_KILLABLE ); }EXPORT_SYMBOL_GPL (__lock_page_killable);
mm/ksm.c +1 -7
@@ -1978,18 +1978,12 @@ void ksm_migrate_page(struct page *newpage, struct page *oldpage)#endif /* CONFIG_MIGRATION */ #ifdef CONFIG_MEMORY_HOTREMOVE -static int just_wait (void *word) -{ -schedule (); -return 0 ; -} -static void wait_while_offlining (void ) {while (ksm_run &KSM_RUN_OFFLINE ) {mutex_unlock (&ksm_thread_mutex);wait_on_bit (&ksm_run,ilog2 (KSM_RUN_OFFLINE ), - just_wait,TASK_UNINTERRUPTIBLE ); +TASK_UNINTERRUPTIBLE );mutex_lock (&ksm_thread_mutex); } }
net/bluetooth/hci_core.c +1 -7
@@ -2186,12 +2186,6 @@ static void hci_inq_req(struct hci_request *req, unsigned long opt)hci_req_add (req,HCI_OP_INQUIRY ,sizeof (cp), &cp); } -static int wait_inquiry (void *word) -{ -schedule (); -return signal_pending (current); -} -int hci_inquiry (void __user *arg) {__u8 __user *ptr = arg;
@@ -2242,7 +2236,7 @@ int hci_inquiry(void __user *arg)/* Wait until Inquiry procedure finishes (HCI_INQUIRY flag is * cleared). If it is interrupted by a signal, return -EINTR. */ -if (wait_on_bit (&hdev->flags,HCI_INQUIRY , wait_inquiry, +if (wait_on_bit (&hdev->flags,HCI_INQUIRY ,TASK_INTERRUPTIBLE ))return -EINTR ; }
security/keys/gc.c +1 -10
@@ -92,15 +92,6 @@ static void key_gc_timer_func(unsigned long data) }/* -* wait_on_bit() sleep function for uninterruptible waiting -*/ -static int key_gc_wait_bit (void *flags) -{ -schedule (); -return 0 ; -} - -/* * Reap keys of dead type. * * We use three flags to make sure we see three complete cycles of the garbage
@@ -123,7 +114,7 @@ void key_gc_keytype(struct key_type *ktype)schedule_work (&key_gc_work);kdebug ("sleep" ); -wait_on_bit (&key_gc_flags,KEY_GC_REAPING_KEYTYPE , key_gc_wait_bit, +wait_on_bit (&key_gc_flags,KEY_GC_REAPING_KEYTYPE ,TASK_UNINTERRUPTIBLE ); key_gc_dead_keytype =NULL ;
security/keys/request_key.c +2 -21
@@ -21,24 +21,6 @@#define key_negative_timeout 60/* default timeout on a negative key's existence */ -/* -* wait_on_bit() sleep function for uninterruptible waiting -*/ -static int key_wait_bit (void *flags) -{ -schedule (); -return 0 ; -} - -/* -* wait_on_bit() sleep function for interruptible waiting -*/ -static int key_wait_bit_intr (void *flags) -{ -schedule (); -return signal_pending (current) ? -ERESTARTSYS :0 ; -} -/** * complete_request_key - Complete the construction of a key. * @cons: The key construction record.
@@ -592,10 +574,9 @@ int wait_for_key_construction(struct key *key, bool intr)int ret; ret =wait_on_bit (&key->flags,KEY_FLAG_USER_CONSTRUCT , - intr ? key_wait_bit_intr : key_wait_bit, intr ?TASK_INTERRUPTIBLE :TASK_UNINTERRUPTIBLE ); -if (ret <0 ) -return ret; +if (ret) +return -ERESTARTSYS ;if (test_bit (KEY_FLAG_NEGATIVE , &key->flags)) {smp_rmb ();return key->type_data.reject_error;
tip-bot for NeilBrown
merge bot[tip:sched/core] sched: Remove proliferation of wait_on_bit() action functions
sched: Remove proliferation of wait_on_bit() action functions
Commit-ID: 743162013d40ca612b4cb53d3a200dff2d9ab26e Gitweb: http://git.kernel.org/tip/743162013d40ca612b4cb53d3a200dff2d9ab26e Author: NeilBrown <[email protected]> AuthorDate: Mon, 7 Jul 2014 15:16:04 +1000 Committer: Ingo Molnar <[email protected]> CommitDate: Wed, 16 Jul 2014 15:10:39 +0200
sched: Remove proliferation of wait_on_bit() action functions
The current "wait_on_bit" interface requires an 'action' function to be provided which does the actual waiting. There are over 20 such functions, many of them identical. Most cases can be satisfied by one of just two functions, one which uses io_schedule() and one which just uses schedule().
So:
Rename wait_on_bit and wait_on_bit_lock to
wait_on_bit_action and wait_on_bit_lock_action
to make it explicit that they need an action function.
Introduce new wait_on_bit{,_lock} and wait_on_bit{,_lock}_io which are *not* given an action function but implicitly use a standard one.
The decision to error-out if a signal is pending is now made based on the 'mode' argument rather than being encoded in the action function.
All instances of the old wait_on_bit and wait_on_bit_lock which can use the new version have been changed accordingly and their action functions have been discarded. wait_on_bit{_lock} does not return any specific error code in the event of a signal so the caller must check for non-zero and interpolate their own error code as appropriate.
The wait_on_bit() call in __fscache_wait_on_invalidate() was ambiguous as it specified TASK_UNINTERRUPTIBLE but used fscache_wait_bit_interruptible as an action function. David Howells confirms this should be uniformly "uninterruptible"
The main remaining user of wait_on_bit{,_lock}_action is NFS which needs to use a freezer-aware schedule() call.
A comment in fs/gfs2/glock.c notes that having multiple 'action' functions is useful as they display differently in the 'wchan' field of 'ps'. (and /proc/$PID/wchan).
As the new bit_wait{,_io} functions are tagged "__sched", they will not show up at all, but something higher in the stack. So the distinction will still be visible, only with different function names (gds2_glock_wait versus gfs2_glock_dq_wait in the gfs2/glock.c case).
Since first version of this patch (against 3.15) two new action functions appeared, on in NFS and one in CIFS. CIFS also now uses an action function that makes the same freezer aware schedule call as NFS.
Signed-off-by NeilBrown <[email protected]>
Acked-by Peter Zijlstra <[email protected]>
Cc Oleg Nesterov <[email protected]>
Cc Steve French <[email protected]>
Cc Linus Torvalds <[email protected]>
Signed-off-by Ingo Molnar <[email protected]>
--- Documentation/filesystems/caching/operations.txt | 2 +- drivers/md/dm-bufio.c | 41 +++----- drivers/md/dm-snap.c | 10 +- drivers/media/usb/dvb-usb-v2/dvb_usb_core.c | 12 +-- fs/btrfs/extent_io.c | 10 +- fs/buffer.c | 11 +-- fs/cifs/connect.c | 10 +- fs/cifs/file.c | 9 +- fs/cifs/inode.c | 4 +- fs/cifs/misc.c | 2 +- fs/fs-writeback.c | 3 +- fs/fscache/cookie.c | 7 +- fs/fscache/internal.h | 2 - fs/fscache/main.c | 18 ---- fs/fscache/page.c | 4 +- fs/gfs2/glock.c | 25 +---- fs/gfs2/lock_dlm.c | 8 +- fs/gfs2/ops_fstype.c | 11 +-- fs/gfs2/recovery.c | 8 +- fs/gfs2/super.c | 8 +- fs/inode.c | 7 -- fs/jbd2/transaction.c | 10 +- fs/nfs/file.c | 4 +- fs/nfs/filelayout/filelayoutdev.c | 4 +- fs/nfs/inode.c | 4 +- fs/nfs/nfs4state.c | 4 +- fs/nfs/pagelist.c | 12 +-- fs/nfs/pnfs.c | 2 +- fs/nfs/write.c | 4 +- include/linux/wait.h | 115 ++++++++++++++++++++++- include/linux/writeback.h | 3 +- kernel/ptrace.c | 8 +- kernel/sched/wait.c | 18 ++++ mm/filemap.c | 20 +--- mm/ksm.c | 8 +- net/bluetooth/hci_core.c | 8 +- security/keys/gc.c | 11 +-- security/keys/request_key.c | 23 +---- 38 files changed, 195 insertions(+), 275 deletions(-)
Documentation/filesystems/caching/operations.txt +1 -1
@@ -90,7 +90,7 @@ operations: to be cleared before proceeding: wait_on_bit(&op->flags, FSCACHE_OP_WAITING, - fscache_wait_bit, TASK_UNINTERRUPTIBLE); + TASK_UNINTERRUPTIBLE); (2) The operation may be fast asynchronous (FSCACHE_OP_FAST), in which case it
drivers/md/dm-bufio.c +14 -27
92 lines not on this page · raw message
drivers/md/dm-snap.c +1 -9
@@ -1032,21 +1032,13 @@ static void start_merge(struct dm_snapshot *s)snapshot_merge_next_chunks (s); } -static int wait_schedule (void *ptr) -{ -schedule (); - -return 0 ; -} -/* * Stop the merging process and wait until it finishes. */ static void stop_merge (struct dm_snapshot *s) {set_bit (SHUTDOWN_MERGE , &s->state_bits); -wait_on_bit (&s->state_bits,RUNNING_MERGE , wait_schedule, -TASK_UNINTERRUPTIBLE ); +wait_on_bit (&s->state_bits,RUNNING_MERGE ,TASK_UNINTERRUPTIBLE );clear_bit (SHUTDOWN_MERGE , &s->state_bits); }
drivers/media/usb/dvb-usb-v2/dvb_usb_core.c +2 -10
@@ -253,13 +253,6 @@ static int dvb_usbv2_adapter_stream_exit(struct dvb_usb_adapter *adap)return usb_urb_exitv2 (&adap->stream); } -static int wait_schedule (void *ptr) -{ -schedule (); - -return 0 ; -} -static int dvb_usb_start_feed (struct dvb_demux_feed *dvbdmxfeed) {struct dvb_usb_adapter *adap = dvbdmxfeed->demux->priv;
@@ -273,8 +266,7 @@ static int dvb_usb_start_feed(struct dvb_demux_feed *dvbdmxfeed) dvbdmxfeed->pid, dvbdmxfeed->index);/* wait init is done */ -wait_on_bit (&adap->state_bits,ADAP_INIT , wait_schedule, -TASK_UNINTERRUPTIBLE ); +wait_on_bit (&adap->state_bits,ADAP_INIT ,TASK_UNINTERRUPTIBLE );if (adap->active_fe ==-1 )return -EINVAL ;
@@ -568,7 +560,7 @@ static int dvb_usb_fe_sleep(struct dvb_frontend *fe)if (!adap->suspend_resume_active) {set_bit (ADAP_SLEEP , &adap->state_bits); -wait_on_bit (&adap->state_bits,ADAP_STREAMING , wait_schedule, +wait_on_bit (&adap->state_bits,ADAP_STREAMING ,TASK_UNINTERRUPTIBLE ); }
fs/btrfs/extent_io.c +2 -8
@@ -3437,16 +3437,10 @@ done_unlocked:return 0 ; } -static int eb_wait (void *word) -{ -io_schedule (); -return 0 ; -} -void wait_on_extent_buffer_writeback (struct extent_buffer *eb) { -wait_on_bit (&eb->bflags,EXTENT_BUFFER_WRITEBACK , eb_wait, -TASK_UNINTERRUPTIBLE ); +wait_on_bit_io (&eb->bflags,EXTENT_BUFFER_WRITEBACK , +TASK_UNINTERRUPTIBLE ); }static noinline_for_stack int
fs/buffer.c +2 -9
@@ -61,16 +61,9 @@ inline void touch_buffer(struct buffer_head *bh) }EXPORT_SYMBOL (touch_buffer); -static int sleep_on_buffer (void *word) -{ -io_schedule (); -return 0 ; -} -void __lock_buffer (struct buffer_head *bh) { -wait_on_bit_lock (&bh->b_state, BH_Lock, sleep_on_buffer, -TASK_UNINTERRUPTIBLE ); +wait_on_bit_lock_io (&bh->b_state, BH_Lock,TASK_UNINTERRUPTIBLE ); }EXPORT_SYMBOL (__lock_buffer);
@@ -123,7 +116,7 @@ EXPORT_SYMBOL(buffer_check_dirty_writeback);*/ void __wait_on_buffer (struct buffer_head * bh) { -wait_on_bit (&bh->b_state, BH_Lock, sleep_on_buffer,TASK_UNINTERRUPTIBLE ); +wait_on_bit_io (&bh->b_state, BH_Lock,TASK_UNINTERRUPTIBLE ); }EXPORT_SYMBOL (__wait_on_buffer);
fs/cifs/connect.c +1 -9
@@ -3934,13 +3934,6 @@ cifs_sb_master_tcon(struct cifs_sb_info *cifs_sb)return tlink_tcon (cifs_sb_master_tlink (cifs_sb)); } -static int -cifs_sb_tcon_pending_wait (void *unused) -{ -schedule (); -return signal_pending (current) ? -ERESTARTSYS :0 ; -} -/* find and return a tlink with given uid */ static struct tcon_link *tlink_rb_search (struct rb_root *root,kuid_t uid)
@@ -4039,11 +4032,10 @@ cifs_sb_tlink(struct cifs_sb_info *cifs_sb) }else { wait_for_construction: ret =wait_on_bit (&tlink->tl_flags,TCON_LINK_PENDING , - cifs_sb_tcon_pending_wait,TASK_INTERRUPTIBLE );if (ret) {cifs_put_tlink (tlink); -return ERR_PTR (ret); +return ERR_PTR (-ERESTARTSYS ); }/* if it's good, return it */
fs/cifs/file.c +1 -8
@@ -3618,13 +3618,6 @@ static int cifs_launder_page(struct page *page)return rc; } -static int -cifs_pending_writers_wait (void *unused) -{ -schedule (); -return 0 ; -} -void cifs_oplock_break (struct work_struct *work) { struct cifsFileInfo *cfile = container_of(work, struct cifsFileInfo,
@@ -3636,7 +3629,7 @@ void cifs_oplock_break(struct work_struct *work)int rc =0 ;wait_on_bit (&cinode->flags,CIFS_INODE_PENDING_WRITERS , - cifs_pending_writers_wait,TASK_UNINTERRUPTIBLE ); +TASK_UNINTERRUPTIBLE ); server->ops->downgrade_oplock (server, cinode,test_bit (CIFS_INODE_DOWNGRADE_OPLOCK_TO_L2 , &cinode->flags));
fs/cifs/inode.c +2 -2
@@ -1794,8 +1794,8 @@ cifs_revalidate_mapping(struct inode *inode)int rc;unsigned long *flags = &CIFS_I (inode)->flags; - rc =wait_on_bit_lock (flags,CIFS_INO_LOCK , cifs_wait_bit_killable, -TASK_KILLABLE ); + rc =wait_on_bit_lock_action (flags,CIFS_INO_LOCK , cifs_wait_bit_killable, +TASK_KILLABLE );if (rc)return rc;
fs/cifs/misc.c +1 -1
@@ -582,7 +582,7 @@ int cifs_get_writer(struct cifsInodeInfo *cinode) start: rc =wait_on_bit (&cinode->flags,CIFS_INODE_PENDING_OPLOCK_BREAK , - cifs_oplock_break_wait,TASK_KILLABLE ); +TASK_KILLABLE );if (rc)return rc;
fs/fs-writeback.c +2 -1
@@ -342,7 +342,8 @@ static void __inode_wait_for_writeback(struct inode *inode) wqh =bit_waitqueue (&inode->i_state, __I_SYNC);while (inode->i_state &I_SYNC ) {spin_unlock (&inode->i_lock); -__wait_on_bit (wqh, &wq, inode_wait,TASK_UNINTERRUPTIBLE ); +__wait_on_bit (wqh, &wq, bit_wait, +TASK_UNINTERRUPTIBLE );spin_lock (&inode->i_lock); } }
fs/fscache/cookie.c +3 -4
@@ -160,7 +160,7 @@ void __fscache_enable_cookie(struct fscache_cookie *cookie,_enter ("%p" , cookie);wait_on_bit_lock (&cookie->flags,FSCACHE_COOKIE_ENABLEMENT_LOCK , - fscache_wait_bit,TASK_UNINTERRUPTIBLE ); +TASK_UNINTERRUPTIBLE );if (test_bit (FSCACHE_COOKIE_ENABLED , &cookie->flags)) goto out_unlock;
@@ -255,7 +255,7 @@ static int fscache_acquire_non_index_cookie(struct fscache_cookie *cookie)if (!fscache_defer_lookup) {_debug ("non-deferred lookup %p" , &cookie->flags);wait_on_bit (&cookie->flags,FSCACHE_COOKIE_LOOKING_UP , - fscache_wait_bit,TASK_UNINTERRUPTIBLE ); +TASK_UNINTERRUPTIBLE );_debug ("complete" );if (test_bit (FSCACHE_COOKIE_UNAVAILABLE , &cookie->flags)) goto unavailable;
@@ -463,7 +463,6 @@ void __fscache_wait_on_invalidate(struct fscache_cookie *cookie)_enter ("%p" , cookie);wait_on_bit (&cookie->flags,FSCACHE_COOKIE_INVALIDATING , - fscache_wait_bit_interruptible,TASK_UNINTERRUPTIBLE );_leave ("" );
@@ -525,7 +524,7 @@ void __fscache_disable_cookie(struct fscache_cookie *cookie, bool invalidate) }wait_on_bit_lock (&cookie->flags,FSCACHE_COOKIE_ENABLEMENT_LOCK , - fscache_wait_bit,TASK_UNINTERRUPTIBLE ); +TASK_UNINTERRUPTIBLE );if (!test_and_clear_bit (FSCACHE_COOKIE_ENABLED , &cookie->flags)) goto out_unlock_enable;
fs/fscache/internal.h +0 -2
@@ -97,8 +97,6 @@ static inline bool fscache_object_congested(void)return workqueue_congested (WORK_CPU_UNBOUND , fscache_object_wq); } -extern int fscache_wait_bit (void *); -extern int fscache_wait_bit_interruptible (void *);extern int fscache_wait_atomic_t (atomic_t *);/*
fs/fscache/main.c +0 -18
@@ -197,24 +197,6 @@ static void __exit fscache_exit(void)module_exit (fscache_exit);/* -* wait_on_bit() sleep function for uninterruptible waiting -*/ -int fscache_wait_bit (void *flags) -{ -schedule (); -return 0 ; -} - -/* -* wait_on_bit() sleep function for interruptible waiting -*/ -int fscache_wait_bit_interruptible (void *flags) -{ -schedule (); -return signal_pending (current); -} - -/* * wait_on_atomic_t() sleep function for uninterruptible waiting */ int fscache_wait_atomic_t (atomic_t *p)
fs/fscache/page.c +1 -3
@@ -298,7 +298,6 @@ int fscache_wait_for_deferred_lookup(struct fscache_cookie *cookie) jif = jiffies;if (wait_on_bit (&cookie->flags,FSCACHE_COOKIE_LOOKING_UP , - fscache_wait_bit_interruptible,TASK_INTERRUPTIBLE ) !=0 ) {fscache_stat (&fscache_n_retrievals_intr);_leave (" = -ERESTARTSYS" );
@@ -342,7 +341,6 @@ int fscache_wait_for_operation_activation(struct fscache_object *object,if (stat_op_waits)fscache_stat (stat_op_waits);if (wait_on_bit (&op->flags,FSCACHE_OP_WAITING , - fscache_wait_bit_interruptible,TASK_INTERRUPTIBLE ) !=0 ) { ret =fscache_cancel_op (op, do_cancel);if (ret ==0 )
@@ -351,7 +349,7 @@ int fscache_wait_for_operation_activation(struct fscache_object *object,/* it's been removed from the pending queue by another party, * so we should get to run shortly */ wait_on_bit (&op->flags,FSCACHE_OP_WAITING , - fscache_wait_bit,TASK_UNINTERRUPTIBLE ); +TASK_UNINTERRUPTIBLE ); }_debug ("<<< GO" );
fs/gfs2/glock.c +2 -23
@@ -856,27 +856,6 @@ void gfs2_holder_uninit(struct gfs2_holder *gh) }/** -* gfs2_glock_holder_wait -* @word: unused -* -* This function and gfs2_glock_demote_wait both show up in the WCHAN -* field. Thus I've separated these otherwise identical functions in -* order to be more informative to the user. -*/ - -static int gfs2_glock_holder_wait (void *word) -{ -schedule (); -return 0 ; -} - -static int gfs2_glock_demote_wait (void *word) -{ -schedule (); -return 0 ; -} - -/** * gfs2_glock_wait - wait on a glock acquisition * @gh: the glock holder *
@@ -888,7 +867,7 @@ int gfs2_glock_wait(struct gfs2_holder *gh)unsigned long time1 = jiffies;might_sleep (); -wait_on_bit (&gh->gh_iflags,HIF_WAIT , gfs2_glock_holder_wait,TASK_UNINTERRUPTIBLE ); +wait_on_bit (&gh->gh_iflags,HIF_WAIT ,TASK_UNINTERRUPTIBLE );if (time_after (jiffies, time1 +HZ ))/* have we waited > a second? */ /* Lengthen the minimum hold time. */ gh->gh_gl->gl_hold_time = min(gh->gh_gl->gl_hold_time +
@@ -1128,7 +1107,7 @@ void gfs2_glock_dq_wait(struct gfs2_holder *gh)struct gfs2_glock *gl = gh->gh_gl;gfs2_glock_dq (gh);might_sleep (); -wait_on_bit (&gl->gl_flags,GLF_DEMOTE , gfs2_glock_demote_wait,TASK_UNINTERRUPTIBLE ); +wait_on_bit (&gl->gl_flags,GLF_DEMOTE ,TASK_UNINTERRUPTIBLE ); }/**
fs/gfs2/lock_dlm.c +1 -7
@@ -936,12 +936,6 @@ fail:return error; } -static int dlm_recovery_wait (void *word) -{ -schedule (); -return 0 ; -} -static int control_first_done (struct gfs2_sbd *sdp) {struct lm_lockstruct *ls = &sdp->sd_lockstruct;
@@ -976,7 +970,7 @@ restart:fs_info (sdp,"control_first_done wait gen %u\n" , start_gen);wait_on_bit (&ls->ls_recover_flags,DFL_DLM_RECOVERY , - dlm_recovery_wait,TASK_UNINTERRUPTIBLE ); +TASK_UNINTERRUPTIBLE ); goto restart; }
fs/gfs2/ops_fstype.c +2 -9
@@ -1024,20 +1024,13 @@ void gfs2_lm_unmount(struct gfs2_sbd *sdp) lm->lm_unmount (sdp); } -static int gfs2_journalid_wait (void *word) -{ -if (signal_pending (current)) -return -EINTR ; -schedule (); -return 0 ; -} -static int wait_on_journal (struct gfs2_sbd *sdp) {if (sdp->sd_lockstruct.ls_ops->lm_mount ==NULL )return 0 ; -return wait_on_bit (&sdp->sd_flags,SDF_NOJOURNALID , gfs2_journalid_wait,TASK_INTERRUPTIBLE ); +return wait_on_bit (&sdp->sd_flags,SDF_NOJOURNALID ,TASK_INTERRUPTIBLE ) + ? -EINTR :0 ; }void gfs2_online_uevent (struct gfs2_sbd *sdp)
fs/gfs2/recovery.c +1 -7
@@ -591,12 +591,6 @@ done:wake_up_bit (&jd->jd_flags,JDF_RECOVERY ); } -static int gfs2_recovery_wait (void *word) -{ -schedule (); -return 0 ; -} -int gfs2_recover_journal (struct gfs2_jdesc *jd,bool wait) {int rv;
@@ -609,7 +603,7 @@ int gfs2_recover_journal(struct gfs2_jdesc *jd, bool wait)BUG_ON (!rv);if (wait) -wait_on_bit (&jd->jd_flags,JDF_RECOVERY , gfs2_recovery_wait, +wait_on_bit (&jd->jd_flags,JDF_RECOVERY ,TASK_UNINTERRUPTIBLE );return wait ? jd->jd_recover_error :0 ;
fs/gfs2/super.c +1 -7
@@ -864,12 +864,6 @@ static int gfs2_make_fs_ro(struct gfs2_sbd *sdp)return error; } -static int gfs2_umount_recovery_wait (void *word) -{ -schedule (); -return 0 ; -} -/** * gfs2_put_super - Unmount the filesystem * @sb: The VFS superblock
@@ -894,7 +888,7 @@ restart:continue ;spin_unlock (&sdp->sd_jindex_spin);wait_on_bit (&jd->jd_flags,JDF_RECOVERY , - gfs2_umount_recovery_wait,TASK_UNINTERRUPTIBLE ); +TASK_UNINTERRUPTIBLE ); goto restart; }spin_unlock (&sdp->sd_jindex_spin);
fs/inode.c +0 -7
@@ -1695,13 +1695,6 @@ int inode_needs_sync(struct inode *inode) }EXPORT_SYMBOL (inode_needs_sync); -int inode_wait (void *word) -{ -schedule (); -return 0 ; -} -EXPORT_SYMBOL (inode_wait); -/* * If we try to find an inode in the inode hash while it is being * deleted, we have to wait until the filesystem completes its
fs/jbd2/transaction.c +2 -8
@@ -763,12 +763,6 @@ static void warn_dirty_buffer(struct buffer_head *bh) bdevname(bh->b_bdev, b), (unsigned long long)bh->b_blocknr); } -static int sleep_on_shadow_bh (void *word) -{ -io_schedule (); -return 0 ; -} -/* * If the buffer is already part of the current transaction, then there * is nothing we need to do. If it is already part of a prior
@@ -906,8 +900,8 @@ repeat:if (buffer_shadow (bh)) {JBUFFER_TRACE (jh,"on shadow: sleep" );jbd_unlock_bh_state (bh); -wait_on_bit (&bh->b_state, BH_Shadow, - sleep_on_shadow_bh,TASK_UNINTERRUPTIBLE ); +wait_on_bit_io (&bh->b_state, BH_Shadow, +TASK_UNINTERRUPTIBLE ); goto repeat; }
fs/nfs/file.c +2 -2
@@ -361,8 +361,8 @@ start:* Prevent starvation issues if someone is doing a consistency * sync-to-disk */ - ret =wait_on_bit (&NFS_I (mapping->host)->flags,NFS_INO_FLUSHING , - nfs_wait_bit_killable,TASK_KILLABLE ); + ret =wait_on_bit_action (&NFS_I (mapping->host)->flags,NFS_INO_FLUSHING , + nfs_wait_bit_killable,TASK_KILLABLE );if (ret)return ret;
fs/nfs/filelayout/filelayoutdev.c +2 -2
@@ -783,8 +783,8 @@ nfs4_fl_select_ds_fh(struct pnfs_layout_segment *lseg, u32 j)static void nfs4_wait_ds_connect (struct nfs4_pnfs_ds *ds) {might_sleep (); -wait_on_bit (&ds->ds_state,NFS4DS_CONNECTING , - nfs_wait_bit_killable,TASK_KILLABLE ); +wait_on_bit_action (&ds->ds_state,NFS4DS_CONNECTING , + nfs_wait_bit_killable,TASK_KILLABLE ); }static void nfs4_clear_ds_conn_bit (struct nfs4_pnfs_ds *ds)
fs/nfs/inode.c +2 -2
@@ -1074,8 +1074,8 @@ int nfs_revalidate_mapping(struct inode *inode, struct address_space *mapping)* the bit lock here if it looks like we're going to be doing that. */ for (;;) { - ret =wait_on_bit (bitlock,NFS_INO_INVALIDATING , - nfs_wait_bit_killable,TASK_KILLABLE ); + ret =wait_on_bit_action (bitlock,NFS_INO_INVALIDATING , + nfs_wait_bit_killable,TASK_KILLABLE );if (ret) goto out;spin_lock (&inode->i_lock);
fs/nfs/nfs4state.c +2 -2
@@ -1251,8 +1251,8 @@ int nfs4_wait_clnt_recover(struct nfs_client *clp)might_sleep ();atomic_inc (&clp->cl_count); - res =wait_on_bit (&clp->cl_state,NFS4CLNT_MANAGER_RUNNING , - nfs_wait_bit_killable,TASK_KILLABLE ); + res =wait_on_bit_action (&clp->cl_state,NFS4CLNT_MANAGER_RUNNING , + nfs_wait_bit_killable,TASK_KILLABLE );if (res) goto out;if (clp->cl_cons_state <0 )
fs/nfs/pagelist.c +2 -10
@@ -138,12 +138,6 @@ nfs_iocounter_wait(struct nfs_io_counter *c)return __nfs_iocounter_wait (c); } -static int nfs_wait_bit_uninterruptible (void *word) -{ -io_schedule (); -return 0 ; -} -/* * nfs_page_group_lock - lock the head of the page group * @req - request in group that is to be locked
@@ -158,7 +152,6 @@ nfs_page_group_lock(struct nfs_page *req)WARN_ON_ONCE (head != head->wb_head);wait_on_bit_lock (&head->wb_flags,PG_HEADLOCK , - nfs_wait_bit_uninterruptible,TASK_UNINTERRUPTIBLE ); }
@@ -425,9 +418,8 @@ void nfs_release_request(struct nfs_page *req)int nfs_wait_on_request (struct nfs_page *req) { -return wait_on_bit (&req->wb_flags,PG_BUSY , - nfs_wait_bit_uninterruptible, -TASK_UNINTERRUPTIBLE ); +return wait_on_bit_io (&req->wb_flags,PG_BUSY , +TASK_UNINTERRUPTIBLE ); }/*
fs/nfs/pnfs.c +1 -1
@@ -1885,7 +1885,7 @@ pnfs_layoutcommit_inode(struct inode *inode, bool sync)if (test_and_set_bit (NFS_INO_LAYOUTCOMMITTING , &nfsi->flags)) {if (!sync) goto out; - status =wait_on_bit_lock (&nfsi->flags, + status =wait_on_bit_lock_action (&nfsi->flags,NFS_INO_LAYOUTCOMMITTING , nfs_wait_bit_killable,TASK_KILLABLE );
fs/nfs/write.c +2 -2
@@ -397,7 +397,7 @@ int nfs_writepages(struct address_space *mapping, struct writeback_control *wbc)int err;/* Stop dirtying of new pages while we sync */ - err =wait_on_bit_lock (bitlock,NFS_INO_FLUSHING , + err =wait_on_bit_lock_action (bitlock,NFS_INO_FLUSHING , nfs_wait_bit_killable,TASK_KILLABLE );if (err) goto out_err;
@@ -1475,7 +1475,7 @@ int nfs_commit_inode(struct inode *inode, int how)return error;if (!may_wait) goto out_mark_dirty; - error =wait_on_bit (&NFS_I (inode)->flags, + error =wait_on_bit_action (&NFS_I (inode)->flags,NFS_INO_COMMIT , nfs_wait_bit_killable,TASK_KILLABLE );
include/linux/wait.h +111 -4
147 lines not on this page · raw message
include/linux/writeback.h +1 -2
@@ -90,7 +90,6 @@ struct writeback_control {* fs/fs-writeback.c */ struct bdi_writeback ; -int inode_wait (void *);void writeback_inodes_sb (struct super_block *,enum wb_reason reason);void writeback_inodes_sb_nr (struct super_block *,unsigned long nr,enum wb_reason reason);
@@ -105,7 +104,7 @@ void inode_wait_for_writeback(struct inode *inode);static inline void wait_on_inode (struct inode *inode) {might_sleep (); -wait_on_bit (&inode->i_state, __I_NEW, inode_wait,TASK_UNINTERRUPTIBLE ); +wait_on_bit (&inode->i_state, __I_NEW,TASK_UNINTERRUPTIBLE ); }/*
kernel/ptrace.c +1 -7
@@ -28,12 +28,6 @@#include <linux/compat.h> -static int ptrace_trapping_sleep_fn (void *flags) -{ -schedule (); -return 0 ; -} -/* * ptrace a task: make the debugger its new parent and * move it to the ptrace list.
@@ -371,7 +365,7 @@ unlock_creds: out:if (!retval) {wait_on_bit (&task->jobctl,JOBCTL_TRAPPING_BIT , - ptrace_trapping_sleep_fn,TASK_UNINTERRUPTIBLE ); +TASK_UNINTERRUPTIBLE );proc_ptrace_connector (task,PTRACE_ATTACH ); }
kernel/sched/wait.c +18 -0
@@ -502,3 +502,21 @@ void wake_up_atomic_t(atomic_t *p)__wake_up_bit (atomic_t_waitqueue (p), p,WAIT_ATOMIC_T_BIT_NR ); }EXPORT_SYMBOL (wake_up_atomic_t); + +__sched int bit_wait(void *word) +{ +if (signal_pending_state (current->state, current)) +return 1 ; +schedule (); +return 0 ; +} +EXPORT_SYMBOL (bit_wait); + +__sched int bit_wait_io(void *word) +{ +if (signal_pending_state (current->state, current)) +return 1 ; +io_schedule (); +return 0 ; +} +EXPORT_SYMBOL (bit_wait_io);
mm/filemap.c +4 -16
@@ -241,18 +241,6 @@ void delete_from_page_cache(struct page *page) }EXPORT_SYMBOL (delete_from_page_cache); -static int sleep_on_page (void *word) -{ -io_schedule (); -return 0 ; -} - -static int sleep_on_page_killable (void *word) -{ -sleep_on_page (word); -return fatal_signal_pending (current) ? -EINTR :0 ; -} -static int filemap_check_errors (struct address_space *mapping) {int ret =0 ;
@@ -692,7 +680,7 @@ void wait_on_page_bit(struct page *page, int bit_nr)DEFINE_WAIT_BIT (wait, &page->flags, bit_nr);if (test_bit (bit_nr, &page->flags)) -__wait_on_bit (page_waitqueue (page), &wait, sleep_on_page, +__wait_on_bit (page_waitqueue (page), &wait, bit_wait_io,TASK_UNINTERRUPTIBLE ); }EXPORT_SYMBOL (wait_on_page_bit);
@@ -705,7 +693,7 @@ int wait_on_page_bit_killable(struct page *page, int bit_nr)return 0 ;return __wait_on_bit (page_waitqueue (page), &wait, - sleep_on_page_killable,TASK_KILLABLE ); + bit_wait_io,TASK_KILLABLE ); }/**
@@ -806,7 +794,7 @@ void __lock_page(struct page *page) {DEFINE_WAIT_BIT (wait, &page->flags, PG_locked); -__wait_on_bit_lock (page_waitqueue (page), &wait, sleep_on_page, +__wait_on_bit_lock (page_waitqueue (page), &wait, bit_wait_io,TASK_UNINTERRUPTIBLE ); }EXPORT_SYMBOL (__lock_page);
@@ -816,7 +804,7 @@ int __lock_page_killable(struct page *page)DEFINE_WAIT_BIT (wait, &page->flags, PG_locked);return __wait_on_bit_lock (page_waitqueue (page), &wait, - sleep_on_page_killable,TASK_KILLABLE ); + bit_wait_io,TASK_KILLABLE ); }EXPORT_SYMBOL_GPL (__lock_page_killable);
mm/ksm.c +1 -7
@@ -1978,18 +1978,12 @@ void ksm_migrate_page(struct page *newpage, struct page *oldpage)#endif /* CONFIG_MIGRATION */ #ifdef CONFIG_MEMORY_HOTREMOVE -static int just_wait (void *word) -{ -schedule (); -return 0 ; -} -static void wait_while_offlining (void ) {while (ksm_run &KSM_RUN_OFFLINE ) {mutex_unlock (&ksm_thread_mutex);wait_on_bit (&ksm_run,ilog2 (KSM_RUN_OFFLINE ), - just_wait,TASK_UNINTERRUPTIBLE ); +TASK_UNINTERRUPTIBLE );mutex_lock (&ksm_thread_mutex); } }
net/bluetooth/hci_core.c +1 -7
@@ -2186,12 +2186,6 @@ static void hci_inq_req(struct hci_request *req, unsigned long opt)hci_req_add (req,HCI_OP_INQUIRY ,sizeof (cp), &cp); } -static int wait_inquiry (void *word) -{ -schedule (); -return signal_pending (current); -} -int hci_inquiry (void __user *arg) {__u8 __user *ptr = arg;
@@ -2242,7 +2236,7 @@ int hci_inquiry(void __user *arg)/* Wait until Inquiry procedure finishes (HCI_INQUIRY flag is * cleared). If it is interrupted by a signal, return -EINTR. */ -if (wait_on_bit (&hdev->flags,HCI_INQUIRY , wait_inquiry, +if (wait_on_bit (&hdev->flags,HCI_INQUIRY ,TASK_INTERRUPTIBLE ))return -EINTR ; }
security/keys/gc.c +1 -10
@@ -92,15 +92,6 @@ static void key_gc_timer_func(unsigned long data) }/* -* wait_on_bit() sleep function for uninterruptible waiting -*/ -static int key_gc_wait_bit (void *flags) -{ -schedule (); -return 0 ; -} - -/* * Reap keys of dead type. * * We use three flags to make sure we see three complete cycles of the garbage
@@ -123,7 +114,7 @@ void key_gc_keytype(struct key_type *ktype)schedule_work (&key_gc_work);kdebug ("sleep" ); -wait_on_bit (&key_gc_flags,KEY_GC_REAPING_KEYTYPE , key_gc_wait_bit, +wait_on_bit (&key_gc_flags,KEY_GC_REAPING_KEYTYPE ,TASK_UNINTERRUPTIBLE ); key_gc_dead_keytype =NULL ;
security/keys/request_key.c +2 -21
@@ -21,24 +21,6 @@#define key_negative_timeout 60/* default timeout on a negative key's existence */ -/* -* wait_on_bit() sleep function for uninterruptible waiting -*/ -static int key_wait_bit (void *flags) -{ -schedule (); -return 0 ; -} - -/* -* wait_on_bit() sleep function for interruptible waiting -*/ -static int key_wait_bit_intr (void *flags) -{ -schedule (); -return signal_pending (current) ? -ERESTARTSYS :0 ; -} -/** * complete_request_key - Complete the construction of a key. * @cons: The key construction record.
@@ -592,10 +574,9 @@ int wait_for_key_construction(struct key *key, bool intr)int ret; ret =wait_on_bit (&key->flags,KEY_FLAG_USER_CONSTRUCT , - intr ? key_wait_bit_intr : key_wait_bit, intr ?TASK_INTERRUPTIBLE :TASK_UNINTERRUPTIBLE ); -if (ret <0 ) -return ret; +if (ret) +return -ERESTARTSYS ;if (test_bit (KEY_FLAG_NEGATIVE , &key->flags)) {smp_rmb ();return key->type_data.reject_error;
PATCH 2/2 SCHED: allow wait_on_bit_action functions to support a timeout.
In reply to NeilBrownIt is currently not possible for various wait_on_bit functions to implement a timeout.
While the "action" function that is called to do the waiting could certainly use schedule_timeout(), there is no way to carry forward the remaining timeout after a false wake-up.
As false-wakeups a clearly possible at least due to possible hash collisions in bit_waitqueue(), this is a real problem.
The 'action' function is currently passed a pointer to the word containing the bit being waited on. No current action functions use this pointer. So changing it to something else will be a little noisy but will have no immediate effect.
This patch changes the 'action' function to take a pointer to the "struct wait_bit_key", which contains a pointer to the word containing the bit so nothing is really lost.
It also adds a 'private' field to "struct wait_bit_key", which is initialized to zero.
An action function can now implement a timeout with something like
static int timed_out_waiter(struct wait_bit_key *key)
{
unsigned long waited;
if (key->private == 0) {
key->private = jiffies;
if (key->private == 0)
key->private -= 1;
}
waited = jiffies - key->private;
if (waited > 10 * HZ)
return -EAGAIN;
schedule_timeout(waited - 10 * HZ);
return 0;
}If any other need for context in a waiter were found it would be easy to use ->private for some other purpose, or even extend "struct wait_bit_key".
My particular need is to support timeouts in nfs_release_page() to avoid deadlocks with loopback mounted NFS.
While wait_on_bit_timeout() would be a cleaner interface, it will not meet my need. I need the timeout to be sensitive to the state of the connection with the server, which could change. So I need to use an 'action' interface.
Signed-off-by NeilBrown <[email protected]>
--- fs/cifs/inode.c | 2 +- fs/nfs/inode.c | 2 +- fs/nfs/internal.h | 2 +- fs/nfs/pagelist.c | 2 +- include/linux/sunrpc/sched.h | 2 +- include/linux/wait.h | 18 ++++++++++-------- kernel/sched/wait.c | 16 ++++++++-------- net/sunrpc/sched.c | 4 ++-- 8 files changed, 25 insertions(+), 23 deletions(-)
fs/cifs/inode.c +1 -1
@@ -1780,7 +1780,7 @@ cifs_invalidate_mapping(struct inode *inode)* @word: long word containing the bit lock */ static int -cifs_wait_bit_killable (void *word) +cifs_wait_bit_killable (struct wait_bit_key *key) {if (fatal_signal_pending (current))return -ERESTARTSYS ;
fs/nfs/inode.c +1 -1
@@ -75,7 +75,7 @@ nfs_fattr_to_ino_t(struct nfs_fattr *fattr)* nfs_wait_bit_killable - helper for functions that are sleeping on bit locks * @word: long word containing the bit lock */ -int nfs_wait_bit_killable (void *word) +int nfs_wait_bit_killable (struct wait_bit_key *key) {if (fatal_signal_pending (current))return -ERESTARTSYS ;
fs/nfs/internal.h +1 -1
@@ -347,7 +347,7 @@ extern int nfs_drop_inode(struct inode *);extern void nfs_clear_inode (struct inode *);extern void nfs_evict_inode (struct inode *);void nfs_zap_acl_cache (struct inode *inode); -extern int nfs_wait_bit_killable (void *word); +extern int nfs_wait_bit_killable (struct wait_bit_key *key);/* super.c */ extern const struct super_operations nfs_sops;
fs/nfs/pagelist.c +1 -1
@@ -117,7 +117,7 @@ __nfs_iocounter_wait(struct nfs_io_counter *c)set_bit (NFS_IO_INPROGRESS , &c->flags);if (atomic_read (&c->io_count) ==0 )break ; - ret =nfs_wait_bit_killable (&c->flags); + ret =nfs_wait_bit_killable (&q.key); }while (atomic_read (&c->io_count) !=0 );finish_wait (wq, &q.wait);return ret;
include/linux/sunrpc/sched.h +1 -1
@@ -236,7 +236,7 @@ void * rpc_malloc(struct rpc_task *, size_t);void rpc_free (void *);int rpciod_up (void );void rpciod_down (void ); -int __rpc_wait_for_completion_task (struct rpc_task *task,int (*)(void *)); +int __rpc_wait_for_completion_task (struct rpc_task *task,wait_bit_action_f *);#ifdef RPC_DEBUG struct net ;void rpc_show_tasks (struct net *);
include/linux/wait.h +10 -8
@@ -25,6 +25,7 @@ struct wait_bit_key {void *flags;int bit_nr;#define WAIT_ATOMIC_T_BIT_NR -1 +unsigned long private; };struct wait_bit_queue {
@@ -141,18 +142,19 @@ __remove_wait_queue(wait_queue_head_t *head, wait_queue_t *old)list_del (&old->task_list); } +typedef int wait_bit_action_f (struct wait_bit_key *);void __wake_up (wait_queue_head_t *q,unsigned int mode,int nr,void *key);void __wake_up_locked_key (wait_queue_head_t *q,unsigned int mode,void *key);void __wake_up_sync_key (wait_queue_head_t *q,unsigned int mode,int nr,void *key);void __wake_up_locked (wait_queue_head_t *q,unsigned int mode,int nr);void __wake_up_sync (wait_queue_head_t *q,unsigned int mode,int nr);void __wake_up_bit (wait_queue_head_t *,void *,int ); -int __wait_on_bit (wait_queue_head_t *,struct wait_bit_queue *,int (*)(void *),unsigned ); -int __wait_on_bit_lock (wait_queue_head_t *,struct wait_bit_queue *,int (*)(void *),unsigned ); +int __wait_on_bit (wait_queue_head_t *,struct wait_bit_queue *,wait_bit_action_f *,unsigned ); +int __wait_on_bit_lock (wait_queue_head_t *,struct wait_bit_queue *,wait_bit_action_f *,unsigned );void wake_up_bit (void *,int );void wake_up_atomic_t (atomic_t *); -int out_of_line_wait_on_bit (void *,int ,int (*)(void *),unsigned ); -int out_of_line_wait_on_bit_lock (void *,int ,int (*)(void *),unsigned ); +int out_of_line_wait_on_bit (void *,int ,wait_bit_action_f *,unsigned ); +int out_of_line_wait_on_bit_lock (void *,int ,wait_bit_action_f *,unsigned );int out_of_line_wait_on_atomic_t (atomic_t *,int (*)(atomic_t *),unsigned );wait_queue_head_t *bit_waitqueue (void *,int );
@@ -855,8 +857,8 @@ int wake_bit_function(wait_queue_t *wait, unsigned mode, int sync, void *key); }while (0 ) -extern int bit_wait(void *); -extern int bit_wait_io (void *); +extern int bit_wait(struct wait_bit_key *); +extern int bit_wait_io (struct wait_bit_key *);/** * wait_on_bit - wait for a bit to be cleared
@@ -925,7 +927,7 @@ wait_on_bit_io(void *word, int bit, unsigned mode)* on that signal. */ static inline int -wait_on_bit_action (void *word,int bit,int (*action)(void *),unsigned mode) +wait_on_bit_action (void *word,int bit,wait_bit_action_f *action,unsigned mode) {if (!test_bit (bit, word))return 0 ;
@@ -1000,7 +1002,7 @@ wait_on_bit_lock_io(void *word, int bit, unsigned mode)* the @mode allows that signal to wake the process. */ static inline int -wait_on_bit_lock_action (void *word,int bit,int (*action)(void *),unsigned mode) +wait_on_bit_lock_action (void *word,int bit,wait_bit_action_f *action,unsigned mode) {if (!test_and_set_bit (bit, word))return 0 ;
kernel/sched/wait.c +8 -8
@@ -319,14 +319,14 @@ EXPORT_SYMBOL(wake_bit_function);*/ int __sched__wait_on_bit (wait_queue_head_t *wq,struct wait_bit_queue *q, -int (*action)(void *),unsigned mode) +wait_bit_action_f *action,unsigned mode) {int ret =0 ;do {prepare_to_wait (wq, &q->wait, mode);if (test_bit (q->key.bit_nr, q->key.flags)) - ret = (*action)(q->key.flags); + ret = (*action)(&q->key); }while (test_bit (q->key.bit_nr, q->key.flags) && !ret);finish_wait (wq, &q->wait);return ret;
@@ -334,7 +334,7 @@ __wait_on_bit(wait_queue_head_t *wq, struct wait_bit_queue *q,EXPORT_SYMBOL (__wait_on_bit); int __sched out_of_line_wait_on_bit(void *word, int bit, -int (*action)(void *),unsigned mode) +wait_bit_action_f *action,unsigned mode) {wait_queue_head_t *wq =bit_waitqueue (word, bit);DEFINE_WAIT_BIT (wait, word, bit);
@@ -345,7 +345,7 @@ EXPORT_SYMBOL(out_of_line_wait_on_bit); int __sched__wait_on_bit_lock (wait_queue_head_t *wq,struct wait_bit_queue *q, -int (*action)(void *),unsigned mode) +wait_bit_action_f *action,unsigned mode) { do { int ret;
@@ -353,7 +353,7 @@ __wait_on_bit_lock(wait_queue_head_t *wq, struct wait_bit_queue *q,prepare_to_wait_exclusive (wq, &q->wait, mode);if (!test_bit (q->key.bit_nr, q->key.flags))continue ; - ret =action (q->key.flags); + ret =action (&q->key);if (!ret)continue ;abort_exclusive_wait (wq, &q->wait, mode, &q->key);
@@ -365,7 +365,7 @@ __wait_on_bit_lock(wait_queue_head_t *wq, struct wait_bit_queue *q,EXPORT_SYMBOL (__wait_on_bit_lock); int __sched out_of_line_wait_on_bit_lock(void *word, int bit, -int (*action)(void *),unsigned mode) +wait_bit_action_f *action,unsigned mode) {wait_queue_head_t *wq =bit_waitqueue (word, bit);DEFINE_WAIT_BIT (wait, word, bit);
@@ -503,7 +503,7 @@ void wake_up_atomic_t(atomic_t *p) }EXPORT_SYMBOL (wake_up_atomic_t); -__sched int bit_wait(void *word) +__sched int bit_wait(struct wait_bit_key *word) {if (signal_pending_state (current->state, current))return 1 ;
@@ -512,7 +512,7 @@ __sched int bit_wait(void *word) }EXPORT_SYMBOL (bit_wait); -__sched int bit_wait_io(void *word) +__sched int bit_wait_io(struct wait_bit_key *word) {if (signal_pending_state (current->state, current))return 1 ;
net/sunrpc/sched.c +2 -2
@@ -250,7 +250,7 @@ void rpc_destroy_wait_queue(struct rpc_wait_queue *queue) }EXPORT_SYMBOL_GPL (rpc_destroy_wait_queue); -static int rpc_wait_bit_killable (void *word) +static int rpc_wait_bit_killable (struct wait_bit_key *key) {if (fatal_signal_pending (current))return -ERESTARTSYS ;
@@ -309,7 +309,7 @@ static int rpc_complete_task(struct rpc_task *task)* to enforce taking of the wq->lock and hence avoid races with * rpc_complete_task(). */ -int __rpc_wait_for_completion_task (struct rpc_task *task,int (*action)(void *)) +int __rpc_wait_for_completion_task (struct rpc_task *task,wait_bit_action_f *action) {if (action ==NULL ) action = rpc_wait_bit_killable;
tip-bot for NeilBrown
merge bot[tip:sched/core] sched: Allow wait_on_bit_action() functions to support a timeout
sched: Allow wait_on_bit_action() functions to support a timeout
Commit-ID: c1221321b7c25b53204447cff9949a6d5a7ddddc Gitweb: http://git.kernel.org/tip/c1221321b7c25b53204447cff9949a6d5a7ddddc Author: NeilBrown <[email protected]> AuthorDate: Mon, 7 Jul 2014 15:16:04 +1000 Committer: Ingo Molnar <[email protected]> CommitDate: Wed, 16 Jul 2014 15:10:41 +0200
sched: Allow wait_on_bit_action() functions to support a timeout
It is currently not possible for various wait_on_bit functions to implement a timeout.
While the "action" function that is called to do the waiting could certainly use schedule_timeout(), there is no way to carry forward the remaining timeout after a false wake-up. As false-wakeups a clearly possible at least due to possible hash collisions in bit_waitqueue(), this is a real problem.
The 'action' function is currently passed a pointer to the word containing the bit being waited on. No current action functions use this pointer. So changing it to something else will be a little noisy but will have no immediate effect.
This patch changes the 'action' function to take a pointer to the "struct wait_bit_key", which contains a pointer to the word containing the bit so nothing is really lost.
It also adds a 'private' field to "struct wait_bit_key", which is initialized to zero.
An action function can now implement a timeout with something like
static int timed_out_waiter(struct wait_bit_key *key)
{
unsigned long waited;
if (key->private == 0) {
key->private = jiffies;
if (key->private == 0)
key->private -= 1;
}
waited = jiffies - key->private;
if (waited > 10 * HZ)
return -EAGAIN;
schedule_timeout(waited - 10 * HZ);
return 0;
}If any other need for context in a waiter were found it would be easy to use ->private for some other purpose, or even extend "struct wait_bit_key".
My particular need is to support timeouts in nfs_release_page() to avoid deadlocks with loopback mounted NFS.
While wait_on_bit_timeout() would be a cleaner interface, it will not meet my need. I need the timeout to be sensitive to the state of the connection with the server, which could change.
So I need to use an 'action' interface.
Signed-off-by NeilBrown <[email protected]>
Acked-by Peter Zijlstra <[email protected]>
Cc Oleg Nesterov <[email protected]>
Cc Steve French <[email protected]>
Cc David Howells <[email protected]>
Cc Steven Whitehouse <[email protected]>
Cc Linus Torvalds <[email protected]>
Signed-off-by Ingo Molnar <[email protected]>
--- fs/cifs/inode.c | 2 +- fs/nfs/inode.c | 2 +- fs/nfs/internal.h | 2 +- fs/nfs/pagelist.c | 2 +- include/linux/sunrpc/sched.h | 2 +- include/linux/wait.h | 18 ++++++++++-------- kernel/sched/wait.c | 16 ++++++++-------- net/sunrpc/sched.c | 4 ++-- 8 files changed, 25 insertions(+), 23 deletions(-)
fs/cifs/inode.c +1 -1
@@ -1780,7 +1780,7 @@ cifs_invalidate_mapping(struct inode *inode)* @word: long word containing the bit lock */ static int -cifs_wait_bit_killable (void *word) +cifs_wait_bit_killable (struct wait_bit_key *key) {if (fatal_signal_pending (current))return -ERESTARTSYS ;
fs/nfs/inode.c +1 -1
@@ -75,7 +75,7 @@ nfs_fattr_to_ino_t(struct nfs_fattr *fattr)* nfs_wait_bit_killable - helper for functions that are sleeping on bit locks * @word: long word containing the bit lock */ -int nfs_wait_bit_killable (void *word) +int nfs_wait_bit_killable (struct wait_bit_key *key) {if (fatal_signal_pending (current))return -ERESTARTSYS ;
fs/nfs/internal.h +1 -1
@@ -347,7 +347,7 @@ extern int nfs_drop_inode(struct inode *);extern void nfs_clear_inode (struct inode *);extern void nfs_evict_inode (struct inode *);void nfs_zap_acl_cache (struct inode *inode); -extern int nfs_wait_bit_killable (void *word); +extern int nfs_wait_bit_killable (struct wait_bit_key *key);/* super.c */ extern const struct super_operations nfs_sops;
fs/nfs/pagelist.c +1 -1
@@ -117,7 +117,7 @@ __nfs_iocounter_wait(struct nfs_io_counter *c)set_bit (NFS_IO_INPROGRESS , &c->flags);if (atomic_read (&c->io_count) ==0 )break ; - ret =nfs_wait_bit_killable (&c->flags); + ret =nfs_wait_bit_killable (&q.key); }while (atomic_read (&c->io_count) !=0 );finish_wait (wq, &q.wait);return ret;
include/linux/sunrpc/sched.h +1 -1
@@ -236,7 +236,7 @@ void * rpc_malloc(struct rpc_task *, size_t);void rpc_free (void *);int rpciod_up (void );void rpciod_down (void ); -int __rpc_wait_for_completion_task (struct rpc_task *task,int (*)(void *)); +int __rpc_wait_for_completion_task (struct rpc_task *task,wait_bit_action_f *);#ifdef RPC_DEBUG struct net ;void rpc_show_tasks (struct net *);
include/linux/wait.h +10 -8
@@ -25,6 +25,7 @@ struct wait_bit_key {void *flags;int bit_nr;#define WAIT_ATOMIC_T_BIT_NR -1 +unsigned long private; };struct wait_bit_queue {
@@ -141,18 +142,19 @@ __remove_wait_queue(wait_queue_head_t *head, wait_queue_t *old)list_del (&old->task_list); } +typedef int wait_bit_action_f (struct wait_bit_key *);void __wake_up (wait_queue_head_t *q,unsigned int mode,int nr,void *key);void __wake_up_locked_key (wait_queue_head_t *q,unsigned int mode,void *key);void __wake_up_sync_key (wait_queue_head_t *q,unsigned int mode,int nr,void *key);void __wake_up_locked (wait_queue_head_t *q,unsigned int mode,int nr);void __wake_up_sync (wait_queue_head_t *q,unsigned int mode,int nr);void __wake_up_bit (wait_queue_head_t *,void *,int ); -int __wait_on_bit (wait_queue_head_t *,struct wait_bit_queue *,int (*)(void *),unsigned ); -int __wait_on_bit_lock (wait_queue_head_t *,struct wait_bit_queue *,int (*)(void *),unsigned ); +int __wait_on_bit (wait_queue_head_t *,struct wait_bit_queue *,wait_bit_action_f *,unsigned ); +int __wait_on_bit_lock (wait_queue_head_t *,struct wait_bit_queue *,wait_bit_action_f *,unsigned );void wake_up_bit (void *,int );void wake_up_atomic_t (atomic_t *); -int out_of_line_wait_on_bit (void *,int ,int (*)(void *),unsigned ); -int out_of_line_wait_on_bit_lock (void *,int ,int (*)(void *),unsigned ); +int out_of_line_wait_on_bit (void *,int ,wait_bit_action_f *,unsigned ); +int out_of_line_wait_on_bit_lock (void *,int ,wait_bit_action_f *,unsigned );int out_of_line_wait_on_atomic_t (atomic_t *,int (*)(atomic_t *),unsigned );wait_queue_head_t *bit_waitqueue (void *,int );
@@ -855,8 +857,8 @@ int wake_bit_function(wait_queue_t *wait, unsigned mode, int sync, void *key); }while (0 ) -extern int bit_wait(void *); -extern int bit_wait_io (void *); +extern int bit_wait(struct wait_bit_key *); +extern int bit_wait_io (struct wait_bit_key *);/** * wait_on_bit - wait for a bit to be cleared
@@ -925,7 +927,7 @@ wait_on_bit_io(void *word, int bit, unsigned mode)* on that signal. */ static inline int -wait_on_bit_action (void *word,int bit,int (*action)(void *),unsigned mode) +wait_on_bit_action (void *word,int bit,wait_bit_action_f *action,unsigned mode) {if (!test_bit (bit, word))return 0 ;
@@ -1000,7 +1002,7 @@ wait_on_bit_lock_io(void *word, int bit, unsigned mode)* the @mode allows that signal to wake the process. */ static inline int -wait_on_bit_lock_action (void *word,int bit,int (*action)(void *),unsigned mode) +wait_on_bit_lock_action (void *word,int bit,wait_bit_action_f *action,unsigned mode) {if (!test_and_set_bit (bit, word))return 0 ;
kernel/sched/wait.c +8 -8
@@ -319,14 +319,14 @@ EXPORT_SYMBOL(wake_bit_function);*/ int __sched__wait_on_bit (wait_queue_head_t *wq,struct wait_bit_queue *q, -int (*action)(void *),unsigned mode) +wait_bit_action_f *action,unsigned mode) {int ret =0 ;do {prepare_to_wait (wq, &q->wait, mode);if (test_bit (q->key.bit_nr, q->key.flags)) - ret = (*action)(q->key.flags); + ret = (*action)(&q->key); }while (test_bit (q->key.bit_nr, q->key.flags) && !ret);finish_wait (wq, &q->wait);return ret;
@@ -334,7 +334,7 @@ __wait_on_bit(wait_queue_head_t *wq, struct wait_bit_queue *q,EXPORT_SYMBOL (__wait_on_bit); int __sched out_of_line_wait_on_bit(void *word, int bit, -int (*action)(void *),unsigned mode) +wait_bit_action_f *action,unsigned mode) {wait_queue_head_t *wq =bit_waitqueue (word, bit);DEFINE_WAIT_BIT (wait, word, bit);
@@ -345,7 +345,7 @@ EXPORT_SYMBOL(out_of_line_wait_on_bit); int __sched__wait_on_bit_lock (wait_queue_head_t *wq,struct wait_bit_queue *q, -int (*action)(void *),unsigned mode) +wait_bit_action_f *action,unsigned mode) { do { int ret;
@@ -353,7 +353,7 @@ __wait_on_bit_lock(wait_queue_head_t *wq, struct wait_bit_queue *q,prepare_to_wait_exclusive (wq, &q->wait, mode);if (!test_bit (q->key.bit_nr, q->key.flags))continue ; - ret =action (q->key.flags); + ret =action (&q->key);if (!ret)continue ;abort_exclusive_wait (wq, &q->wait, mode, &q->key);
@@ -365,7 +365,7 @@ __wait_on_bit_lock(wait_queue_head_t *wq, struct wait_bit_queue *q,EXPORT_SYMBOL (__wait_on_bit_lock); int __sched out_of_line_wait_on_bit_lock(void *word, int bit, -int (*action)(void *),unsigned mode) +wait_bit_action_f *action,unsigned mode) {wait_queue_head_t *wq =bit_waitqueue (word, bit);DEFINE_WAIT_BIT (wait, word, bit);
@@ -503,7 +503,7 @@ void wake_up_atomic_t(atomic_t *p) }EXPORT_SYMBOL (wake_up_atomic_t); -__sched int bit_wait(void *word) +__sched int bit_wait(struct wait_bit_key *word) {if (signal_pending_state (current->state, current))return 1 ;
@@ -512,7 +512,7 @@ __sched int bit_wait(void *word) }EXPORT_SYMBOL (bit_wait); -__sched int bit_wait_io(void *word) +__sched int bit_wait_io(struct wait_bit_key *word) {if (signal_pending_state (current->state, current))return 1 ;
net/sunrpc/sched.c +2 -2
@@ -250,7 +250,7 @@ void rpc_destroy_wait_queue(struct rpc_wait_queue *queue) }EXPORT_SYMBOL_GPL (rpc_destroy_wait_queue); -static int rpc_wait_bit_killable (void *word) +static int rpc_wait_bit_killable (struct wait_bit_key *key) {if (fatal_signal_pending (current))return -ERESTARTSYS ;
@@ -309,7 +309,7 @@ static int rpc_complete_task(struct rpc_task *task)* to enforce taking of the wq->lock and hence avoid races with * rpc_complete_task(). */ -int __rpc_wait_for_completion_task (struct rpc_task *task,int (*action)(void *)) +int __rpc_wait_for_completion_task (struct rpc_task *task,wait_bit_action_f *action) {if (action ==NULL ) action = rpc_wait_bit_killable;