* [PATCH v3 0/2] ceph: fix cephfs rsync kworker high load issue
@ 2022-02-23 1:59 xiubli
2022-02-23 1:59 ` [PATCH v3 1/2] ceph: allocate capsnap memory outside of ceph_queue_cap_snap() xiubli
` (2 more replies)
0 siblings, 3 replies; 4+ messages in thread
From: xiubli @ 2022-02-23 1:59 UTC (permalink / raw)
To: jlayton; +Cc: idryomov, vshankar, ceph-devel, Xiubo Li
From: Xiubo Li <xiubli@redhat.com>
V3:
- switch to use the kmem_cache_zalloc() to zero the memory.
- rebase to the latest code in testing branch.
V2:
- allocate the capsnap memory ourside of ceph_queue_cap_snap() from
Jeff's advice.
- fix the code style and logs to make the logs to be more readable
Xiubo Li (2):
ceph: allocate capsnap memory outside of ceph_queue_cap_snap()
ceph: misc fix for code style and logs
fs/ceph/snap.c | 168 ++++++++++++++++++++++++++-----------------------
1 file changed, 90 insertions(+), 78 deletions(-)
--
2.27.0
^ permalink raw reply [flat|nested] 4+ messages in thread
* [PATCH v3 1/2] ceph: allocate capsnap memory outside of ceph_queue_cap_snap()
2022-02-23 1:59 [PATCH v3 0/2] ceph: fix cephfs rsync kworker high load issue xiubli
@ 2022-02-23 1:59 ` xiubli
2022-02-23 1:59 ` [PATCH v3 2/2] ceph: misc fix for code style and logs xiubli
2022-02-23 12:49 ` [PATCH v3 0/2] ceph: fix cephfs rsync kworker high load issue Jeff Layton
2 siblings, 0 replies; 4+ messages in thread
From: xiubli @ 2022-02-23 1:59 UTC (permalink / raw)
To: jlayton; +Cc: idryomov, vshankar, ceph-devel, Xiubo Li
From: Xiubo Li <xiubli@redhat.com>
This will reduce very possible but unnecessary frequently memory
allocate/free in this loop.
URL: https://tracker.ceph.com/issues/44100
Signed-off-by: Xiubo Li <xiubli@redhat.com>
---
fs/ceph/snap.c | 46 ++++++++++++++++++++++++++++------------------
1 file changed, 28 insertions(+), 18 deletions(-)
diff --git a/fs/ceph/snap.c b/fs/ceph/snap.c
index cc9097c27052..b75dcc9a36b6 100644
--- a/fs/ceph/snap.c
+++ b/fs/ceph/snap.c
@@ -522,23 +522,15 @@ static bool has_new_snaps(struct ceph_snap_context *o,
* Caller must hold snap_rwsem for read (i.e., the realm topology won't
* change).
*/
-static void ceph_queue_cap_snap(struct ceph_inode_info *ci)
+static void ceph_queue_cap_snap(struct ceph_inode_info *ci,
+ struct ceph_cap_snap **pcapsnap)
{
struct inode *inode = &ci->vfs_inode;
- struct ceph_cap_snap *capsnap;
struct ceph_snap_context *old_snapc, *new_snapc;
+ struct ceph_cap_snap *capsnap = *pcapsnap;
struct ceph_buffer *old_blob = NULL;
int used, dirty;
- capsnap = kmem_cache_zalloc(ceph_cap_snap_cachep, GFP_NOFS);
- if (!capsnap) {
- pr_err("ENOMEM allocating ceph_cap_snap on %p\n", inode);
- return;
- }
- capsnap->cap_flush.is_capsnap = true;
- INIT_LIST_HEAD(&capsnap->cap_flush.i_list);
- INIT_LIST_HEAD(&capsnap->cap_flush.g_list);
-
spin_lock(&ci->i_ceph_lock);
used = __ceph_caps_used(ci);
dirty = __ceph_caps_dirty(ci);
@@ -595,9 +587,6 @@ static void ceph_queue_cap_snap(struct ceph_inode_info *ci)
capsnap->need_flush ? "" : "no_flush");
ihold(inode);
- refcount_set(&capsnap->nref, 1);
- INIT_LIST_HEAD(&capsnap->ci_item);
-
capsnap->follows = old_snapc->seq;
capsnap->issued = __ceph_caps_issued(ci, NULL);
capsnap->dirty = dirty;
@@ -635,7 +624,7 @@ static void ceph_queue_cap_snap(struct ceph_inode_info *ci)
/* note mtime, size NOW. */
__ceph_finish_cap_snap(ci, capsnap);
}
- capsnap = NULL;
+ *pcapsnap = NULL;
old_snapc = NULL;
update_snapc:
@@ -651,8 +640,6 @@ static void ceph_queue_cap_snap(struct ceph_inode_info *ci)
spin_unlock(&ci->i_ceph_lock);
ceph_buffer_put(old_blob);
- if (capsnap)
- kmem_cache_free(ceph_cap_snap_cachep, capsnap);
ceph_put_snap_context(old_snapc);
}
@@ -720,6 +707,7 @@ static void queue_realm_cap_snaps(struct ceph_snap_realm *realm)
{
struct ceph_inode_info *ci;
struct inode *lastinode = NULL;
+ struct ceph_cap_snap *capsnap = NULL;
dout("queue_realm_cap_snaps %p %llx inodes\n", realm, realm->ino);
@@ -731,12 +719,34 @@ static void queue_realm_cap_snaps(struct ceph_snap_realm *realm)
spin_unlock(&realm->inodes_with_caps_lock);
iput(lastinode);
lastinode = inode;
- ceph_queue_cap_snap(ci);
+
+ /*
+ * Allocate the capsnap memory outside of ceph_queue_cap_snap()
+ * to reduce very possible but unnecessary frequently memory
+ * allocate/free in this loop.
+ */
+ if (!capsnap) {
+ capsnap = kmem_cache_zalloc(ceph_cap_snap_cachep, GFP_NOFS);
+ if (!capsnap) {
+ pr_err("ENOMEM allocating ceph_cap_snap on %p\n",
+ inode);
+ return;
+ }
+ }
+ capsnap->cap_flush.is_capsnap = true;
+ refcount_set(&capsnap->nref, 1);
+ INIT_LIST_HEAD(&capsnap->cap_flush.i_list);
+ INIT_LIST_HEAD(&capsnap->cap_flush.g_list);
+ INIT_LIST_HEAD(&capsnap->ci_item);
+
+ ceph_queue_cap_snap(ci, &capsnap);
spin_lock(&realm->inodes_with_caps_lock);
}
spin_unlock(&realm->inodes_with_caps_lock);
iput(lastinode);
+ if (capsnap)
+ kmem_cache_free(ceph_cap_snap_cachep, capsnap);
dout("queue_realm_cap_snaps %p %llx done\n", realm, realm->ino);
}
--
2.27.0
^ permalink raw reply related [flat|nested] 4+ messages in thread
* [PATCH v3 2/2] ceph: misc fix for code style and logs
2022-02-23 1:59 [PATCH v3 0/2] ceph: fix cephfs rsync kworker high load issue xiubli
2022-02-23 1:59 ` [PATCH v3 1/2] ceph: allocate capsnap memory outside of ceph_queue_cap_snap() xiubli
@ 2022-02-23 1:59 ` xiubli
2022-02-23 12:49 ` [PATCH v3 0/2] ceph: fix cephfs rsync kworker high load issue Jeff Layton
2 siblings, 0 replies; 4+ messages in thread
From: xiubli @ 2022-02-23 1:59 UTC (permalink / raw)
To: jlayton; +Cc: idryomov, vshankar, ceph-devel, Xiubo Li
From: Xiubo Li <xiubli@redhat.com>
To make the logs more readable such as for log likes:
ceph: will move 00000000a42b796b to split realm 100000003ed 000000007146df45
With this it will always show the inode numbers instead the inode
addresses.
Signed-off-by: Xiubo Li <xiubli@redhat.com>
---
fs/ceph/snap.c | 122 +++++++++++++++++++++++++------------------------
1 file changed, 62 insertions(+), 60 deletions(-)
diff --git a/fs/ceph/snap.c b/fs/ceph/snap.c
index b75dcc9a36b6..322ee5add942 100644
--- a/fs/ceph/snap.c
+++ b/fs/ceph/snap.c
@@ -137,7 +137,7 @@ static struct ceph_snap_realm *ceph_create_snap_realm(
__insert_snap_realm(&mdsc->snap_realms, realm);
mdsc->num_snap_realms++;
- dout("create_snap_realm %llx %p\n", realm->ino, realm);
+ dout("%s %llx %p\n", __func__, realm->ino, realm);
return realm;
}
@@ -161,7 +161,7 @@ static struct ceph_snap_realm *__lookup_snap_realm(struct ceph_mds_client *mdsc,
else if (ino > r->ino)
n = n->rb_right;
else {
- dout("lookup_snap_realm %llx %p\n", r->ino, r);
+ dout("%s %llx %p\n", __func__, r->ino, r);
return r;
}
}
@@ -189,7 +189,7 @@ static void __destroy_snap_realm(struct ceph_mds_client *mdsc,
{
lockdep_assert_held_write(&mdsc->snap_rwsem);
- dout("__destroy_snap_realm %p %llx\n", realm, realm->ino);
+ dout("%s %p %llx\n", __func__, realm, realm->ino);
rb_erase(&realm->node, &mdsc->snap_realms);
mdsc->num_snap_realms--;
@@ -302,9 +302,8 @@ static int adjust_snap_realm_parent(struct ceph_mds_client *mdsc,
if (IS_ERR(parent))
return PTR_ERR(parent);
}
- dout("adjust_snap_realm_parent %llx %p: %llx %p -> %llx %p\n",
- realm->ino, realm, realm->parent_ino, realm->parent,
- parentino, parent);
+ dout("%s %llx %p: %llx %p -> %llx %p\n", __func__, realm->ino,
+ realm, realm->parent_ino, realm->parent, parentino, parent);
if (realm->parent) {
list_del_init(&realm->child_item);
ceph_put_snap_realm(mdsc, realm->parent);
@@ -360,9 +359,8 @@ static int build_snap_context(struct ceph_snap_realm *realm,
realm->cached_context->seq == realm->seq &&
(!parent ||
realm->cached_context->seq >= parent->cached_context->seq)) {
- dout("build_snap_context %llx %p: %p seq %lld (%u snaps)"
- " (unchanged)\n",
- realm->ino, realm, realm->cached_context,
+ dout("%s %llx %p: %p seq %lld (%u snaps) (unchanged)\n",
+ __func__, realm->ino, realm, realm->cached_context,
realm->cached_context->seq,
(unsigned int)realm->cached_context->num_snaps);
return 0;
@@ -401,9 +399,8 @@ static int build_snap_context(struct ceph_snap_realm *realm,
sort(snapc->snaps, num, sizeof(u64), cmpu64_rev, NULL);
snapc->num_snaps = num;
- dout("build_snap_context %llx %p: %p seq %lld (%u snaps)\n",
- realm->ino, realm, snapc, snapc->seq,
- (unsigned int) snapc->num_snaps);
+ dout("%s %llx %p: %p seq %lld (%u snaps)\n", __func__, realm->ino,
+ realm, snapc, snapc->seq, (unsigned int) snapc->num_snaps);
ceph_put_snap_context(realm->cached_context);
realm->cached_context = snapc;
@@ -420,8 +417,7 @@ static int build_snap_context(struct ceph_snap_realm *realm,
ceph_put_snap_context(realm->cached_context);
realm->cached_context = NULL;
}
- pr_err("build_snap_context %llx %p fail %d\n", realm->ino,
- realm, err);
+ pr_err("%s %llx %p fail %d\n", __func__, realm->ino, realm, err);
return err;
}
@@ -455,7 +451,7 @@ static void rebuild_snap_realms(struct ceph_snap_realm *realm,
}
last = build_snap_context(_realm, &realm_queue, dirty_realms);
- dout("rebuild_snap_realms %llx %p, %s\n", _realm->ino, _realm,
+ dout("%s %llx %p, %s\n", __func__, _realm->ino, _realm,
last > 0 ? "is deferred" : !last ? "succeeded" : "failed");
/* is any child in the list ? */
@@ -551,12 +547,14 @@ static void ceph_queue_cap_snap(struct ceph_inode_info *ci,
as no new writes are allowed to start when pending, so any
writes in progress now were started before the previous
cap_snap. lucky us. */
- dout("queue_cap_snap %p already pending\n", inode);
+ dout("%s %p %llx.%llx already pending\n",
+ __func__, inode, ceph_vinop(inode));
goto update_snapc;
}
if (ci->i_wrbuffer_ref_head == 0 &&
!(dirty & (CEPH_CAP_ANY_EXCL|CEPH_CAP_FILE_WR))) {
- dout("queue_cap_snap %p nothing dirty|writing\n", inode);
+ dout("%s %p %llx.%llx nothing dirty|writing\n",
+ __func__, inode, ceph_vinop(inode));
goto update_snapc;
}
@@ -576,15 +574,15 @@ static void ceph_queue_cap_snap(struct ceph_inode_info *ci,
} else {
if (!(used & CEPH_CAP_FILE_WR) &&
ci->i_wrbuffer_ref_head == 0) {
- dout("queue_cap_snap %p "
- "no new_snap|dirty_page|writing\n", inode);
+ dout("%s %p %llx.%llx no new_snap|dirty_page|writing\n",
+ __func__, inode, ceph_vinop(inode));
goto update_snapc;
}
}
- dout("queue_cap_snap %p cap_snap %p queuing under %p %s %s\n",
- inode, capsnap, old_snapc, ceph_cap_string(dirty),
- capsnap->need_flush ? "" : "no_flush");
+ dout("%s %p %llx.%llx cap_snap %p queuing under %p %s %s\n",
+ __func__, inode, ceph_vinop(inode), capsnap, old_snapc,
+ ceph_cap_string(dirty), capsnap->need_flush ? "" : "no_flush");
ihold(inode);
capsnap->follows = old_snapc->seq;
@@ -616,8 +614,8 @@ static void ceph_queue_cap_snap(struct ceph_inode_info *ci,
list_add_tail(&capsnap->ci_item, &ci->i_cap_snaps);
if (used & CEPH_CAP_FILE_WR) {
- dout("queue_cap_snap %p cap_snap %p snapc %p"
- " seq %llu used WR, now pending\n", inode,
+ dout("%s %p %llx.%llx cap_snap %p snapc %p seq %llu used WR,"
+ " now pending\n", __func__, inode, ceph_vinop(inode),
capsnap, old_snapc, old_snapc->seq);
capsnap->writing = 1;
} else {
@@ -628,12 +626,12 @@ static void ceph_queue_cap_snap(struct ceph_inode_info *ci,
old_snapc = NULL;
update_snapc:
- if (ci->i_wrbuffer_ref_head == 0 &&
- ci->i_wr_ref == 0 &&
- ci->i_dirty_caps == 0 &&
- ci->i_flushing_caps == 0) {
- ci->i_head_snapc = NULL;
- } else {
+ if (ci->i_wrbuffer_ref_head == 0 &&
+ ci->i_wr_ref == 0 &&
+ ci->i_dirty_caps == 0 &&
+ ci->i_flushing_caps == 0) {
+ ci->i_head_snapc = NULL;
+ } else {
ci->i_head_snapc = ceph_get_snap_context(new_snapc);
dout(" new snapc is %p\n", new_snapc);
}
@@ -668,27 +666,28 @@ int __ceph_finish_cap_snap(struct ceph_inode_info *ci,
capsnap->truncate_size = ci->i_truncate_size;
capsnap->truncate_seq = ci->i_truncate_seq;
if (capsnap->dirty_pages) {
- dout("finish_cap_snap %p cap_snap %p snapc %p %llu %s s=%llu "
- "still has %d dirty pages\n", inode, capsnap,
- capsnap->context, capsnap->context->seq,
- ceph_cap_string(capsnap->dirty), capsnap->size,
- capsnap->dirty_pages);
+ dout("%s %p %llx.%llx cap_snap %p snapc %p %llu %s s=%llu "
+ "still has %d dirty pages\n", __func__, inode,
+ ceph_vinop(inode), capsnap, capsnap->context,
+ capsnap->context->seq, ceph_cap_string(capsnap->dirty),
+ capsnap->size, capsnap->dirty_pages);
return 0;
}
/* Fb cap still in use, delay it */
if (ci->i_wb_ref) {
- dout("finish_cap_snap %p cap_snap %p snapc %p %llu %s s=%llu "
- "used WRBUFFER, delaying\n", inode, capsnap,
- capsnap->context, capsnap->context->seq,
- ceph_cap_string(capsnap->dirty), capsnap->size);
+ dout("%s %p %llx.%llx cap_snap %p snapc %p %llu %s s=%llu "
+ "used WRBUFFER, delaying\n", __func__, inode,
+ ceph_vinop(inode), capsnap, capsnap->context,
+ capsnap->context->seq, ceph_cap_string(capsnap->dirty),
+ capsnap->size);
capsnap->writing = 1;
return 0;
}
ci->i_ceph_flags |= CEPH_I_FLUSH_SNAPS;
- dout("finish_cap_snap %p cap_snap %p snapc %p %llu %s s=%llu\n",
- inode, capsnap, capsnap->context,
+ dout("%s %p %llx.%llx cap_snap %p snapc %p %llu %s s=%llu\n",
+ __func__, inode, ceph_vinop(inode), capsnap, capsnap->context,
capsnap->context->seq, ceph_cap_string(capsnap->dirty),
capsnap->size);
@@ -709,7 +708,7 @@ static void queue_realm_cap_snaps(struct ceph_snap_realm *realm)
struct inode *lastinode = NULL;
struct ceph_cap_snap *capsnap = NULL;
- dout("queue_realm_cap_snaps %p %llx inodes\n", realm, realm->ino);
+ dout("%s %p %llx inode\n", __func__, realm, realm->ino);
spin_lock(&realm->inodes_with_caps_lock);
list_for_each_entry(ci, &realm->inodes_with_caps, i_snap_realm_item) {
@@ -747,7 +746,7 @@ static void queue_realm_cap_snaps(struct ceph_snap_realm *realm)
if (capsnap)
kmem_cache_free(ceph_cap_snap_cachep, capsnap);
- dout("queue_realm_cap_snaps %p %llx done\n", realm, realm->ino);
+ dout("%s %p %llx done\n", __func__, realm, realm->ino);
}
/*
@@ -773,7 +772,7 @@ int ceph_update_snap_trace(struct ceph_mds_client *mdsc,
lockdep_assert_held_write(&mdsc->snap_rwsem);
- dout("update_snap_trace deletion=%d\n", deletion);
+ dout("%s deletion=%d\n", __func__, deletion);
more:
rebuild_snapcs = 0;
ceph_decode_need(&p, e, sizeof(*ri), bad);
@@ -802,7 +801,7 @@ int ceph_update_snap_trace(struct ceph_mds_client *mdsc,
rebuild_snapcs += err;
if (le64_to_cpu(ri->seq) > realm->seq) {
- dout("update_snap_trace updating %llx %p %lld -> %lld\n",
+ dout("%s updating %llx %p %lld -> %lld\n", __func__,
realm->ino, realm, realm->seq, le64_to_cpu(ri->seq));
/* update realm parameters, snap lists */
realm->seq = le64_to_cpu(ri->seq);
@@ -826,11 +825,11 @@ int ceph_update_snap_trace(struct ceph_mds_client *mdsc,
rebuild_snapcs = 1;
} else if (!realm->cached_context) {
- dout("update_snap_trace %llx %p seq %lld new\n",
+ dout("%s %llx %p seq %lld new\n", __func__,
realm->ino, realm, realm->seq);
rebuild_snapcs = 1;
} else {
- dout("update_snap_trace %llx %p seq %lld unchanged\n",
+ dout("%s %llx %p seq %lld unchanged\n", __func__,
realm->ino, realm, realm->seq);
}
@@ -883,7 +882,7 @@ int ceph_update_snap_trace(struct ceph_mds_client *mdsc,
ceph_put_snap_realm(mdsc, realm);
if (first_realm)
ceph_put_snap_realm(mdsc, first_realm);
- pr_err("update_snap_trace error %d\n", err);
+ pr_err("%s error %d\n", __func__, err);
return err;
}
@@ -900,7 +899,7 @@ static void flush_snaps(struct ceph_mds_client *mdsc)
struct inode *inode;
struct ceph_mds_session *session = NULL;
- dout("flush_snaps\n");
+ dout("%s\n", __func__);
spin_lock(&mdsc->snap_flush_lock);
while (!list_empty(&mdsc->snap_flush_list)) {
ci = list_first_entry(&mdsc->snap_flush_list,
@@ -915,7 +914,7 @@ static void flush_snaps(struct ceph_mds_client *mdsc)
spin_unlock(&mdsc->snap_flush_lock);
ceph_put_mds_session(session);
- dout("flush_snaps done\n");
+ dout("%s done\n", __func__);
}
/**
@@ -997,8 +996,8 @@ void ceph_handle_snap(struct ceph_mds_client *mdsc,
trace_len = le32_to_cpu(h->trace_len);
p += sizeof(*h);
- dout("handle_snap from mds%d op %s split %llx tracelen %d\n", mds,
- ceph_snap_op_name(op), split, trace_len);
+ dout("%s from mds%d op %s split %llx tracelen %d\n", __func__,
+ mds, ceph_snap_op_name(op), split, trace_len);
mutex_lock(&session->s_mutex);
inc_session_sequence(session);
@@ -1058,13 +1057,13 @@ void ceph_handle_snap(struct ceph_mds_client *mdsc,
*/
if (ci->i_snap_realm->created >
le64_to_cpu(ri->created)) {
- dout(" leaving %p in newer realm %llx %p\n",
- inode, ci->i_snap_realm->ino,
+ dout(" leaving %p %llx.%llx in newer realm %llx %p\n",
+ inode, ceph_vinop(inode), ci->i_snap_realm->ino,
ci->i_snap_realm);
goto skip_inode;
}
- dout(" will move %p to split realm %llx %p\n",
- inode, realm->ino, realm);
+ dout(" will move %p %llx.%llx to split realm %llx %p\n",
+ inode, ceph_vinop(inode), realm->ino, realm);
ceph_get_snap_realm(mdsc, realm);
ceph_change_snap_realm(inode, realm);
@@ -1107,7 +1106,7 @@ void ceph_handle_snap(struct ceph_mds_client *mdsc,
return;
bad:
- pr_err("corrupt snap message from mds%d\n", mds);
+ pr_err("%s corrupt snap message from mds%d\n", __func__, mds);
ceph_msg_dump(msg);
out:
if (locked_rwsem)
@@ -1140,7 +1139,8 @@ struct ceph_snapid_map* ceph_get_snapid_map(struct ceph_mds_client *mdsc,
}
spin_unlock(&mdsc->snapid_map_lock);
if (exist) {
- dout("found snapid map %llx -> %x\n", exist->snap, exist->dev);
+ dout("%s found snapid map %llx -> %x\n", __func__,
+ exist->snap, exist->dev);
return exist;
}
@@ -1184,11 +1184,13 @@ struct ceph_snapid_map* ceph_get_snapid_map(struct ceph_mds_client *mdsc,
if (exist) {
free_anon_bdev(sm->dev);
kfree(sm);
- dout("found snapid map %llx -> %x\n", exist->snap, exist->dev);
+ dout("%s found snapid map %llx -> %x\n", __func__,
+ exist->snap, exist->dev);
return exist;
}
- dout("create snapid map %llx -> %x\n", sm->snap, sm->dev);
+ dout("%s create snapid map %llx -> %x\n", __func__,
+ sm->snap, sm->dev);
return sm;
}
--
2.27.0
^ permalink raw reply related [flat|nested] 4+ messages in thread
* Re: [PATCH v3 0/2] ceph: fix cephfs rsync kworker high load issue
2022-02-23 1:59 [PATCH v3 0/2] ceph: fix cephfs rsync kworker high load issue xiubli
2022-02-23 1:59 ` [PATCH v3 1/2] ceph: allocate capsnap memory outside of ceph_queue_cap_snap() xiubli
2022-02-23 1:59 ` [PATCH v3 2/2] ceph: misc fix for code style and logs xiubli
@ 2022-02-23 12:49 ` Jeff Layton
2 siblings, 0 replies; 4+ messages in thread
From: Jeff Layton @ 2022-02-23 12:49 UTC (permalink / raw)
To: xiubli; +Cc: idryomov, vshankar, ceph-devel
On Wed, 2022-02-23 at 09:59 +0800, xiubli@redhat.com wrote:
> From: Xiubo Li <xiubli@redhat.com>
>
> V3:
> - switch to use the kmem_cache_zalloc() to zero the memory.
> - rebase to the latest code in testing branch.
>
> V2:
> - allocate the capsnap memory ourside of ceph_queue_cap_snap() from
> Jeff's advice.
> - fix the code style and logs to make the logs to be more readable
>
> Xiubo Li (2):
> ceph: allocate capsnap memory outside of ceph_queue_cap_snap()
> ceph: misc fix for code style and logs
>
> fs/ceph/snap.c | 168 ++++++++++++++++++++++++++-----------------------
> 1 file changed, 90 insertions(+), 78 deletions(-)
>
LGTM!
Reviewed-by: Jeff Layton <jlayton@kernel.org>
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2022-02-23 12:49 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2022-02-23 1:59 [PATCH v3 0/2] ceph: fix cephfs rsync kworker high load issue xiubli
2022-02-23 1:59 ` [PATCH v3 1/2] ceph: allocate capsnap memory outside of ceph_queue_cap_snap() xiubli
2022-02-23 1:59 ` [PATCH v3 2/2] ceph: misc fix for code style and logs xiubli
2022-02-23 12:49 ` [PATCH v3 0/2] ceph: fix cephfs rsync kworker high load issue Jeff Layton
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox