ipc/mqueue: optimize msg_get()

Our msg priorities became an rbtree as of d6629859b3 ("ipc/mqueue:
improve performance of send/recv").  However, consuming a msg in
msg_get() remains logarithmic (still being better than the case before
of course).  By applying well known techniques to cache pointers we can
have the node with the highest priority in O(1), which is specially nice
for the rt cases.  Furthermore, some callers can call msg_get() in a
loop.

A new msg_tree_erase() helper is also added to encapsulate the tree
removal and node_cache game.  Passes ltp mq testcases.

Link: http://lkml.kernel.org/r/20190321190216.1719-2-dave@stgolabs.net
Signed-off-by: Davidlohr Bueso <dbueso@suse.de>
Cc: Manfred Spraul <manfred@colorfullife.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This commit is contained in:
Davidlohr Bueso 2019-05-14 15:46:26 -07:00 committed by Linus Torvalds
parent 0ecb58210b
commit a5091fda4e
1 changed files with 35 additions and 25 deletions

View File

@ -76,6 +76,7 @@ struct mqueue_inode_info {
wait_queue_head_t wait_q;
struct rb_root msg_tree;
struct rb_node *msg_tree_rightmost;
struct posix_msg_tree_node *node_cache;
struct mq_attr attr;
@ -131,6 +132,7 @@ static int msg_insert(struct msg_msg *msg, struct mqueue_inode_info *info)
{
struct rb_node **p, *parent = NULL;
struct posix_msg_tree_node *leaf;
bool rightmost = true;
p = &info->msg_tree.rb_node;
while (*p) {
@ -139,9 +141,10 @@ static int msg_insert(struct msg_msg *msg, struct mqueue_inode_info *info)
if (likely(leaf->priority == msg->m_type))
goto insert_msg;
else if (msg->m_type < leaf->priority)
else if (msg->m_type < leaf->priority) {
p = &(*p)->rb_left;
else
rightmost = false;
} else
p = &(*p)->rb_right;
}
if (info->node_cache) {
@ -154,6 +157,10 @@ static int msg_insert(struct msg_msg *msg, struct mqueue_inode_info *info)
INIT_LIST_HEAD(&leaf->msg_list);
}
leaf->priority = msg->m_type;
if (rightmost)
info->msg_tree_rightmost = &leaf->rb_node;
rb_link_node(&leaf->rb_node, parent, p);
rb_insert_color(&leaf->rb_node, &info->msg_tree);
insert_msg:
@ -163,23 +170,35 @@ insert_msg:
return 0;
}
static inline void msg_tree_erase(struct posix_msg_tree_node *leaf,
struct mqueue_inode_info *info)
{
struct rb_node *node = &leaf->rb_node;
if (info->msg_tree_rightmost == node)
info->msg_tree_rightmost = rb_prev(node);
rb_erase(node, &info->msg_tree);
if (info->node_cache) {
kfree(leaf);
} else {
info->node_cache = leaf;
}
}
static inline struct msg_msg *msg_get(struct mqueue_inode_info *info)
{
struct rb_node **p, *parent = NULL;
struct rb_node *parent = NULL;
struct posix_msg_tree_node *leaf;
struct msg_msg *msg;
try_again:
p = &info->msg_tree.rb_node;
while (*p) {
parent = *p;
/*
* During insert, low priorities go to the left and high to the
* right. On receive, we want the highest priorities first, so
* walk all the way to the right.
*/
p = &(*p)->rb_right;
}
parent = info->msg_tree_rightmost;
if (!parent) {
if (info->attr.mq_curmsgs) {
pr_warn_once("Inconsistency in POSIX message queue, "
@ -194,24 +213,14 @@ try_again:
pr_warn_once("Inconsistency in POSIX message queue, "
"empty leaf node but we haven't implemented "
"lazy leaf delete!\n");
rb_erase(&leaf->rb_node, &info->msg_tree);
if (info->node_cache) {
kfree(leaf);
} else {
info->node_cache = leaf;
}
msg_tree_erase(leaf, info);
goto try_again;
} else {
msg = list_first_entry(&leaf->msg_list,
struct msg_msg, m_list);
list_del(&msg->m_list);
if (list_empty(&leaf->msg_list)) {
rb_erase(&leaf->rb_node, &info->msg_tree);
if (info->node_cache) {
kfree(leaf);
} else {
info->node_cache = leaf;
}
msg_tree_erase(leaf, info);
}
}
info->attr.mq_curmsgs--;
@ -254,6 +263,7 @@ static struct inode *mqueue_get_inode(struct super_block *sb,
info->qsize = 0;
info->user = NULL; /* set when all is ok */
info->msg_tree = RB_ROOT;
info->msg_tree_rightmost = NULL;
info->node_cache = NULL;
memset(&info->attr, 0, sizeof(info->attr));
info->attr.mq_maxmsg = min(ipc_ns->mq_msg_max,