xiaomi-sm8450/android_kernel_xiaomi_sm8450
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

ceph: reorganize __send_cap for less spinlock abuse

Browse Source 
Get rid of the __releases annotation by breaking it up into two
functions: __prep_cap which is done under the spinlock and __send_cap
that is done outside it. Add new fields to cap_msg_args for the wake
boolean and old_xattr_buf pointer.

Nothing checks the return value from __send_cap, so make it void
return.

Signed-off-by: Jeff Layton <jlayton@kernel.org>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
This commit is contained in:
Jeff Layton
2020-03-17 08:47:31 -04:00
committed by Ilya Dryomov
parent 70c948206f
commit 0a454bdd50
1 changed files with 86 additions and 79 deletions
Download Patch File Download Diff File Expand all files Collapse all files

165
fs/ceph/caps.c
View File

@@ -1181,6 +1181,7 @@ struct cap_msg_args {
u64 xattr_version; u64 xattr_version;
u64 change_attr; u64 change_attr;
struct ceph_buffer *xattr_buf; struct ceph_buffer *xattr_buf;
struct ceph_buffer *old_xattr_buf;
struct timespec64 atime, mtime, ctime, btime; struct timespec64 atime, mtime, ctime, btime;
int op, caps, wanted, dirty; int op, caps, wanted, dirty;
u32 seq, issue_seq, mseq, time_warp_seq; u32 seq, issue_seq, mseq, time_warp_seq;
@@ -1189,6 +1190,7 @@ struct cap_msg_args {
kgid_t gid; kgid_t gid;
umode_t mode; umode_t mode;
bool inline_data; bool inline_data;
bool wake;
}; };
/* /*
@@ -1318,44 +1320,29 @@ void __ceph_remove_caps(struct ceph_inode_info *ci)
} }
/* /*
* Send a cap msg on the given inode. Update our caps state, then * Prepare to send a cap message to an MDS. Update the cap state, and populate
* drop i_ceph_lock and send the message. * the arg struct with the parameters that will need to be sent. This should
* be done under the i_ceph_lock to guard against changes to cap state.
* *
* Make note of max_size reported/requested from mds, revoked caps * Make note of max_size reported/requested from mds, revoked caps
* that have now been implemented. * that have now been implemented.
*
* Return non-zero if delayed release, or we experienced an error
* such that the caller should requeue + retry later.
*
* called with i_ceph_lock, then drops it.
* caller should hold snap_rwsem (read), s_mutex.
*/ */
static int __send_cap(struct ceph_mds_client *mdsc, struct ceph_cap *cap, static void __prep_cap(struct cap_msg_args *arg, struct ceph_cap *cap,
int op, int flags, int used, int want, int retain, int op, int flags, int used, int want, int retain,
int flushing, u64 flush_tid, u64 oldest_flush_tid) int flushing, u64 flush_tid, u64 oldest_flush_tid)
__releases(cap->ci->i_ceph_lock)
{ {
struct ceph_inode_info *ci = cap->ci; struct ceph_inode_info *ci = cap->ci;
struct inode *inode = &ci->vfs_inode; struct inode *inode = &ci->vfs_inode;
struct ceph_buffer *old_blob = NULL;
struct cap_msg_args arg;
int held, revoking; int held, revoking;
int wake = 0;
int ret;
/* Don't send anything if it's still being created. Return delayed */ lockdep_assert_held(&ci->i_ceph_lock);
if (ci->i_ceph_flags & CEPH_I_ASYNC_CREATE) {
spin_unlock(&ci->i_ceph_lock);
dout("%s async create in flight for %p\n", __func__, inode);
return 1;
}
held = cap->issued | cap->implemented; held = cap->issued | cap->implemented;
revoking = cap->implemented & ~cap->issued; revoking = cap->implemented & ~cap->issued;
retain &= ~revoking; retain &= ~revoking;
dout("__send_cap %p cap %p session %p %s -> %s (revoking %s)\n", dout("%s %p cap %p session %p %s -> %s (revoking %s)\n",
inode, cap, cap->session, __func__, inode, cap, cap->session,
ceph_cap_string(held), ceph_cap_string(held & retain), ceph_cap_string(held), ceph_cap_string(held & retain),
ceph_cap_string(revoking)); ceph_cap_string(revoking));
BUG_ON((retain & CEPH_CAP_PIN) == 0); BUG_ON((retain & CEPH_CAP_PIN) == 0);
@@ -1363,60 +1350,58 @@ static int __send_cap(struct ceph_mds_client *mdsc, struct ceph_cap *cap,
ci->i_ceph_flags &= ~CEPH_I_FLUSH; ci->i_ceph_flags &= ~CEPH_I_FLUSH;
cap->issued &= retain; /* drop bits we don't want */ cap->issued &= retain; /* drop bits we don't want */
if (cap->implemented & ~cap->issued) { /*
/* * Wake up any waiters on wanted -> needed transition. This is due to
* Wake up any waiters on wanted -> needed transition. * the weird transition from buffered to sync IO... we need to flush
* This is due to the weird transition from buffered * dirty pages _before_ allowing sync writes to avoid reordering.
* to sync IO... we need to flush dirty pages _before_ */
* allowing sync writes to avoid reordering. arg->wake = cap->implemented & ~cap->issued;
*/
wake = 1;
}
cap->implemented &= cap->issued | used; cap->implemented &= cap->issued | used;
cap->mds_wanted = want; cap->mds_wanted = want;
arg.session = cap->session; arg->session = cap->session;
arg.ino = ceph_vino(inode).ino; arg->ino = ceph_vino(inode).ino;
arg.cid = cap->cap_id; arg->cid = cap->cap_id;
arg.follows = flushing ? ci->i_head_snapc->seq : 0; arg->follows = flushing ? ci->i_head_snapc->seq : 0;
arg.flush_tid = flush_tid; arg->flush_tid = flush_tid;
arg.oldest_flush_tid = oldest_flush_tid; arg->oldest_flush_tid = oldest_flush_tid;
arg.size = inode->i_size; arg->size = inode->i_size;
ci->i_reported_size = arg.size; ci->i_reported_size = arg->size;
arg.max_size = ci->i_wanted_max_size; arg->max_size = ci->i_wanted_max_size;
if (cap == ci->i_auth_cap) if (cap == ci->i_auth_cap)
ci->i_requested_max_size = arg.max_size; ci->i_requested_max_size = arg->max_size;
if (flushing & CEPH_CAP_XATTR_EXCL) { if (flushing & CEPH_CAP_XATTR_EXCL) {
old_blob = __ceph_build_xattrs_blob(ci); arg->old_xattr_buf = __ceph_build_xattrs_blob(ci);
arg.xattr_version = ci->i_xattrs.version; arg->xattr_version = ci->i_xattrs.version;
arg.xattr_buf = ci->i_xattrs.blob; arg->xattr_buf = ci->i_xattrs.blob;
} else { } else {
arg.xattr_buf = NULL; arg->xattr_buf = NULL;
arg->old_xattr_buf = NULL;
} }
arg.mtime = inode->i_mtime; arg->mtime = inode->i_mtime;
arg.atime = inode->i_atime; arg->atime = inode->i_atime;
arg.ctime = inode->i_ctime; arg->ctime = inode->i_ctime;
arg.btime = ci->i_btime; arg->btime = ci->i_btime;
arg.change_attr = inode_peek_iversion_raw(inode); arg->change_attr = inode_peek_iversion_raw(inode);
arg.op = op; arg->op = op;
arg.caps = cap->implemented; arg->caps = cap->implemented;
arg.wanted = want; arg->wanted = want;
arg.dirty = flushing; arg->dirty = flushing;
arg.seq = cap->seq; arg->seq = cap->seq;
arg.issue_seq = cap->issue_seq; arg->issue_seq = cap->issue_seq;
arg.mseq = cap->mseq; arg->mseq = cap->mseq;
arg.time_warp_seq = ci->i_time_warp_seq; arg->time_warp_seq = ci->i_time_warp_seq;
arg.uid = inode->i_uid; arg->uid = inode->i_uid;
arg.gid = inode->i_gid; arg->gid = inode->i_gid;
arg.mode = inode->i_mode; arg->mode = inode->i_mode;
arg.inline_data = ci->i_inline_version != CEPH_INLINE_NONE; arg->inline_data = ci->i_inline_version != CEPH_INLINE_NONE;
if (!(flags & CEPH_CLIENT_CAPS_PENDING_CAPSNAP) && if (!(flags & CEPH_CLIENT_CAPS_PENDING_CAPSNAP) &&
!list_empty(&ci->i_cap_snaps)) { !list_empty(&ci->i_cap_snaps)) {
struct ceph_cap_snap *capsnap; struct ceph_cap_snap *capsnap;
@@ -1429,27 +1414,35 @@ static int __send_cap(struct ceph_mds_client *mdsc, struct ceph_cap *cap,
} }
} }
} }
arg.flags = flags; arg->flags = flags;
}
spin_unlock(&ci->i_ceph_lock); /*
* Send a cap msg on the given inode.
*
* Caller should hold snap_rwsem (read), s_mutex.
*/
static void __send_cap(struct ceph_mds_client *mdsc, struct cap_msg_args *arg,
struct ceph_inode_info *ci)
{
struct inode *inode = &ci->vfs_inode;
int ret;
ceph_buffer_put(old_blob); ret = send_cap_msg(arg);
ret = send_cap_msg(&arg);
if (ret < 0) { if (ret < 0) {
pr_err("error sending cap msg, ino (%llx.%llx) " pr_err("error sending cap msg, ino (%llx.%llx) "
"flushing %s tid %llu, requeue\n", "flushing %s tid %llu, requeue\n",
ceph_vinop(inode), ceph_cap_string(flushing), ceph_vinop(inode), ceph_cap_string(arg->dirty),
flush_tid); arg->flush_tid);
spin_lock(&ci->i_ceph_lock); spin_lock(&ci->i_ceph_lock);
__cap_delay_requeue(mdsc, ci); __cap_delay_requeue(mdsc, ci);
spin_unlock(&ci->i_ceph_lock); spin_unlock(&ci->i_ceph_lock);
} }
if (wake) ceph_buffer_put(arg->old_xattr_buf);
wake_up_all(&ci->i_cap_wq);
return ret; if (arg->wake)
wake_up_all(&ci->i_cap_wq);
} }
static inline int __send_flush_snap(struct inode *inode, static inline int __send_flush_snap(struct inode *inode,
@@ -1470,6 +1463,7 @@ static inline int __send_flush_snap(struct inode *inode,
arg.max_size = 0; arg.max_size = 0;
arg.xattr_version = capsnap->xattr_version; arg.xattr_version = capsnap->xattr_version;
arg.xattr_buf = capsnap->xattr_blob; arg.xattr_buf = capsnap->xattr_blob;
arg.old_xattr_buf = NULL;
arg.atime = capsnap->atime; arg.atime = capsnap->atime;
arg.mtime = capsnap->mtime; arg.mtime = capsnap->mtime;
@@ -1493,6 +1487,7 @@ static inline int __send_flush_snap(struct inode *inode,
arg.inline_data = capsnap->inline_data; arg.inline_data = capsnap->inline_data;
arg.flags = 0; arg.flags = 0;
arg.wake = false;
return send_cap_msg(&arg); return send_cap_msg(&arg);
} }
@@ -1967,6 +1962,8 @@ retry_locked:
} }
for (p = rb_first(&ci->i_caps); p; p = rb_next(p)) { for (p = rb_first(&ci->i_caps); p; p = rb_next(p)) {
struct cap_msg_args arg;
cap = rb_entry(p, struct ceph_cap, ci_node); cap = rb_entry(p, struct ceph_cap, ci_node);
/* avoid looping forever */ /* avoid looping forever */
@@ -2094,9 +2091,12 @@ ack:
mds = cap->mds; /* remember mds, so we don't repeat */ mds = cap->mds; /* remember mds, so we don't repeat */
/* __send_cap drops i_ceph_lock */ __prep_cap(&arg, cap, CEPH_CAP_OP_UPDATE, 0, cap_used, want,
__send_cap(mdsc, cap, CEPH_CAP_OP_UPDATE, 0, cap_used, want,
retain, flushing, flush_tid, oldest_flush_tid); retain, flushing, flush_tid, oldest_flush_tid);
spin_unlock(&ci->i_ceph_lock);
__send_cap(mdsc, &arg, ci);
goto retry; /* retake i_ceph_lock and restart our cap scan. */ goto retry; /* retake i_ceph_lock and restart our cap scan. */
} }
@@ -2135,6 +2135,7 @@ retry:
retry_locked: retry_locked:
if (ci->i_dirty_caps && ci->i_auth_cap) { if (ci->i_dirty_caps && ci->i_auth_cap) {
struct ceph_cap *cap = ci->i_auth_cap; struct ceph_cap *cap = ci->i_auth_cap;
struct cap_msg_args arg;
if (session != cap->session) { if (session != cap->session) {
spin_unlock(&ci->i_ceph_lock); spin_unlock(&ci->i_ceph_lock);
@@ -2162,11 +2163,13 @@ retry_locked:
flush_tid = __mark_caps_flushing(inode, session, true, flush_tid = __mark_caps_flushing(inode, session, true,
&oldest_flush_tid); &oldest_flush_tid);
/* __send_cap drops i_ceph_lock */ __prep_cap(&arg, cap, CEPH_CAP_OP_FLUSH, CEPH_CLIENT_CAPS_SYNC,
__send_cap(mdsc, cap, CEPH_CAP_OP_FLUSH, CEPH_CLIENT_CAPS_SYNC,
__ceph_caps_used(ci), __ceph_caps_wanted(ci), __ceph_caps_used(ci), __ceph_caps_wanted(ci),
(cap->issued | cap->implemented), (cap->issued | cap->implemented),
flushing, flush_tid, oldest_flush_tid); flushing, flush_tid, oldest_flush_tid);
spin_unlock(&ci->i_ceph_lock);
__send_cap(mdsc, &arg, ci);
} else { } else {
if (!list_empty(&ci->i_cap_flush_list)) { if (!list_empty(&ci->i_cap_flush_list)) {
struct ceph_cap_flush *cf = struct ceph_cap_flush *cf =
@@ -2368,15 +2371,19 @@ static void __kick_flushing_caps(struct ceph_mds_client *mdsc,
first_tid = cf->tid + 1; first_tid = cf->tid + 1;
if (cf->caps) { if (cf->caps) {
struct cap_msg_args arg;
dout("kick_flushing_caps %p cap %p tid %llu %s\n", dout("kick_flushing_caps %p cap %p tid %llu %s\n",
inode, cap, cf->tid, ceph_cap_string(cf->caps)); inode, cap, cf->tid, ceph_cap_string(cf->caps));
__send_cap(mdsc, cap, CEPH_CAP_OP_FLUSH, __prep_cap(&arg, cap, CEPH_CAP_OP_FLUSH,
(cf->tid < last_snap_flush ? (cf->tid < last_snap_flush ?
CEPH_CLIENT_CAPS_PENDING_CAPSNAP : 0), CEPH_CLIENT_CAPS_PENDING_CAPSNAP : 0),
__ceph_caps_used(ci), __ceph_caps_used(ci),
__ceph_caps_wanted(ci), __ceph_caps_wanted(ci),
(cap->issued | cap->implemented), (cap->issued | cap->implemented),
cf->caps, cf->tid, oldest_flush_tid); cf->caps, cf->tid, oldest_flush_tid);
spin_unlock(&ci->i_ceph_lock);
__send_cap(mdsc, &arg, ci);
} else { } else {
struct ceph_cap_snap *capsnap = struct ceph_cap_snap *capsnap =
container_of(cf, struct ceph_cap_snap, container_of(cf, struct ceph_cap_snap,