From: Eric Dumazet <dada1@cosmosbay.com>
To: David Miller <davem@davemloft.net>, mingo@elte.hu
Cc: cl@linux-foundation.org, rjw@sisk.pl,
linux-kernel@vger.kernel.org, kernel-testers@vger.kernel.org,
efault@gmx.de, a.p.zijlstra@chello.nl,
Linux Netdev List <netdev@vger.kernel.org>
Subject: [PATCH] fs: pipe/sockets/anon dentries should not have a parent
Date: Fri, 21 Nov 2008 16:13:38 +0100 [thread overview]
Message-ID: <4926D022.5060008@cosmosbay.com> (raw)
In-Reply-To: <4926AEDB.10007@cosmosbay.com>
[-- Attachment #1: Type: text/plain, Size: 5668 bytes --]
Eric Dumazet a écrit :
> David Miller a écrit :
>> From: Eric Dumazet <dada1@cosmosbay.com>
>> Date: Fri, 21 Nov 2008 09:51:32 +0100
>>
>>> Now, I wish sockets and pipes not going through dcache, not tbench
>>> affair
>>> of course but real workloads...
>>>
>>> running 8 processes on a 8 way machine doing a
>>> for (;;)
>>> close(socket(AF_INET, SOCK_STREAM, 0));
>>>
>>> is slow as hell, we hit so many contended cache lines ...
>>>
>>> ticket spin locks are slower in this case (dcache_lock for example
>>> is taken twice when we allocate a socket(), once in d_alloc(),
>>> another one
>>> in d_instantiate())
>>
>> As you of course know, this used to be a ton worse. At least now
>> these things are unhashed. :)
>
> Well, this is dust compared to what we currently have.
>
> To allocate a socket we :
> 0) Do the usual file manipulation (pretty scalable these days)
> (but recent drop_file_write_access() and co slow down a bit)
> 1) allocate an inode with new_inode()
> This function :
> - locks inode_lock,
> - dirties nr_inodes counter
> - dirties inode_in_use list (for sockets, I doubt it is usefull)
> - dirties superblock s_inodes.
> - dirties last_ino counter
> All these are in different cache lines of course.
> 2) allocate a dentry
> d_alloc() takes dcache_lock,
> insert dentry on its parent list (dirtying sock_mnt->mnt_sb->s_root)
> dirties nr_dentry
> 3) d_instantiate() dentry (dcache_lock taken again)
> 4) init_file() -> atomic_inc on sock_mnt->refcount (in case we want to
> umount this vfs ...)
>
>
>
> At close() time, we must undo the things. Its even more expensive because
> of the _atomic_dec_and_lock() that stress a lot, and because of two
> cache lines that are touched when an element is deleted from a list.
>
> for (i = 0; i < 1000*1000; i++)
> close(socket(socket(AF_INET, SOCK_STREAM, 0));
>
> Cost if run one one cpu :
>
> real 0m1.561s
> user 0m0.092s
> sys 0m1.469s
>
> If run on 8 CPUS :
>
> real 0m27.496s
> user 0m0.657s
> sys 3m39.092s
>
>
[PATCH] fs: pipe/sockets/anon dentries should not have a parent
Linking pipe/sockets/anon dentries to one root 'parent' has no functional
impact at all, but a scalability one.
We can avoid touching a cache line at allocation stage (inside d_alloc(), no need
to touch root->d_count), but also at freeing time (in d_kill, decrementing d_count)
We avoid an expensive atomic_dec_and_lock() call on the root dentry.
If we correct dnotify_parent() and inotify_d_instantiate() to take into account
a NULL d_parent, we can call d_alloc() with a NULL parent instead of root dentry.
Before patch, time to run 8 millions of close(socket()) calls on 8 CPUS was :
real 0m27.496s
user 0m0.657s
sys 3m39.092s
After patch :
real 0m23.997s
user 0m0.682s
sys 3m11.193s
Old oprofile :
CPU: Core 2, speed 3000.11 MHz (estimated)
Counted CPU_CLK_UNHALTED events (Clock cycles when not halted) with a unit mask of 0x00 (Unhalted core cycles) count 100000
samples cum. samples % cum. % symbol name
164257 164257 11.0245 11.0245 init_file
155488 319745 10.4359 21.4604 d_alloc
151887 471632 10.1942 31.6547 _atomic_dec_and_lock
91620 563252 6.1493 37.8039 inet_create
74245 637497 4.9831 42.7871 kmem_cache_alloc
46702 684199 3.1345 45.9216 dentry_iput
46186 730385 3.0999 49.0215 tcp_close
42824 773209 2.8742 51.8957 kmem_cache_free
37275 810484 2.5018 54.3975 wake_up_inode
36553 847037 2.4533 56.8508 tcp_v4_init_sock
35661 882698 2.3935 59.2443 inotify_d_instantiate
32998 915696 2.2147 61.4590 sysenter_past_esp
31442 947138 2.1103 63.5693 d_instantiate
31303 978441 2.1010 65.6703 generic_forget_inode
27533 1005974 1.8479 67.5183 vfs_dq_drop
24237 1030211 1.6267 69.1450 sock_attach_fd
19290 1049501 1.2947 70.4397 __copy_from_user_ll
New oprofile :
CPU: Core 2, speed 3000.24 MHz (estimated)
Counted CPU_CLK_UNHALTED events (Clock cycles when not halted) with a unit mask of 0x00 (Unhalted core cycles) count 100000
samples cum. samples % cum. % symbol name
147287 147287 10.3984 10.3984 new_inode
144884 292171 10.2287 20.6271 inet_create
93670 385841 6.6131 27.2402 init_file
89852 475693 6.3435 33.5837 wake_up_inode
80910 556603 5.7122 39.2959 kmem_cache_alloc
53588 610191 3.7833 43.0792 _atomic_dec_and_lock
44341 654532 3.1305 46.2096 generic_forget_inode
38710 693242 2.7329 48.9425 kmem_cache_free
37605 730847 2.6549 51.5974 tcp_v4_init_sock
37228 768075 2.6283 54.2257 d_alloc
34085 802160 2.4064 56.6321 tcp_close
32550 834710 2.2980 58.9301 sysenter_past_esp
25931 860641 1.8307 60.7608 vfs_dq_drop
24458 885099 1.7267 62.4875 d_kill
22015 907114 1.5542 64.0418 dentry_iput
18877 925991 1.3327 65.3745 __copy_from_user_ll
17873 943864 1.2618 66.6363 mwait_idle
Signed-off-by: Eric Dumazet <dada1@cosmosbay.com>
---
fs/anon_inodes.c | 2 +-
fs/dnotify.c | 2 +-
fs/inotify.c | 2 +-
fs/pipe.c | 2 +-
net/socket.c | 2 +-
5 files changed, 5 insertions(+), 5 deletions(-)
[-- Attachment #2: null_parent.patch --]
[-- Type: text/plain, Size: 2076 bytes --]
diff --git a/fs/anon_inodes.c b/fs/anon_inodes.c
index 3662dd4..22cce87 100644
--- a/fs/anon_inodes.c
+++ b/fs/anon_inodes.c
@@ -92,7 +92,7 @@ int anon_inode_getfd(const char *name, const struct file_operations *fops,
this.name = name;
this.len = strlen(name);
this.hash = 0;
- dentry = d_alloc(anon_inode_mnt->mnt_sb->s_root, &this);
+ dentry = d_alloc(NULL, &this);
if (!dentry)
goto err_put_unused_fd;
diff --git a/fs/dnotify.c b/fs/dnotify.c
index 676073b..66066a3 100644
--- a/fs/dnotify.c
+++ b/fs/dnotify.c
@@ -173,7 +173,7 @@ void dnotify_parent(struct dentry *dentry, unsigned long event)
spin_lock(&dentry->d_lock);
parent = dentry->d_parent;
- if (parent->d_inode->i_dnotify_mask & event) {
+ if (parent && parent->d_inode->i_dnotify_mask & event) {
dget(parent);
spin_unlock(&dentry->d_lock);
__inode_dir_notify(parent->d_inode, event);
diff --git a/fs/inotify.c b/fs/inotify.c
index 7bbed1b..9f051bb 100644
--- a/fs/inotify.c
+++ b/fs/inotify.c
@@ -270,7 +270,7 @@ void inotify_d_instantiate(struct dentry *entry, struct inode *inode)
spin_lock(&entry->d_lock);
parent = entry->d_parent;
- if (parent->d_inode && inotify_inode_watched(parent->d_inode))
+ if (parent && parent->d_inode && inotify_inode_watched(parent->d_inode))
entry->d_flags |= DCACHE_INOTIFY_PARENT_WATCHED;
spin_unlock(&entry->d_lock);
}
diff --git a/fs/pipe.c b/fs/pipe.c
index 7aea8b8..4b961bc 100644
--- a/fs/pipe.c
+++ b/fs/pipe.c
@@ -926,7 +926,7 @@ struct file *create_write_pipe(int flags)
goto err;
err = -ENOMEM;
- dentry = d_alloc(pipe_mnt->mnt_sb->s_root, &name);
+ dentry = d_alloc(NULL, &name);
if (!dentry)
goto err_inode;
diff --git a/net/socket.c b/net/socket.c
index e9d65ea..b84de7d 100644
--- a/net/socket.c
+++ b/net/socket.c
@@ -373,7 +373,7 @@ static int sock_attach_fd(struct socket *sock, struct file *file, int flags)
struct dentry *dentry;
struct qstr name = { .name = "" };
- dentry = d_alloc(sock_mnt->mnt_sb->s_root, &name);
+ dentry = d_alloc(NULL, &name);
if (unlikely(!dentry))
return -ENOMEM;
next prev parent reply other threads:[~2008-11-21 15:14 UTC|newest]
Thread overview: 185+ messages / expand[flat|nested] mbox.gz Atom feed top
2008-11-16 17:38 2.6.28-rc5: Reported regressions 2.6.26 -> 2.6.27 Rafael J. Wysocki
2008-11-16 17:38 ` [Bug #11207] VolanoMark regression with 2.6.27-rc1 Rafael J. Wysocki
2008-11-16 17:40 ` [Bug #11215] INFO: possible recursive locking detected ps2_command Rafael J. Wysocki
2008-11-16 17:40 ` [Bug #11308] tbench regression on each kernel release from 2.6.22 -> 2.6.28 Rafael J. Wysocki
2008-11-17 9:06 ` Ingo Molnar
2008-11-17 9:14 ` David Miller
2008-11-17 11:01 ` Ingo Molnar
2008-11-17 11:20 ` Eric Dumazet
2008-11-17 16:11 ` Ingo Molnar
2008-11-17 16:35 ` Eric Dumazet
2008-11-17 17:08 ` Ingo Molnar
2008-11-17 17:25 ` Ingo Molnar
2008-11-17 17:33 ` Eric Dumazet
2008-11-17 17:38 ` Linus Torvalds
2008-11-17 17:42 ` Eric Dumazet
2008-11-17 18:23 ` Ingo Molnar
2008-11-17 18:33 ` Linus Torvalds
2008-11-17 18:49 ` Ingo Molnar
2008-11-17 19:30 ` Eric Dumazet
2008-11-17 19:39 ` David Miller
2008-11-17 19:43 ` Eric Dumazet
2008-11-17 19:55 ` Linus Torvalds
2008-11-17 20:16 ` David Miller
2008-11-17 20:30 ` Linus Torvalds
2008-11-17 20:58 ` David Miller
2008-11-18 9:44 ` Nick Piggin
2008-11-18 15:58 ` Linus Torvalds
2008-11-19 4:31 ` Nick Piggin
2008-11-20 9:14 ` David Miller
2008-11-20 9:06 ` David Miller
2008-11-18 12:29 ` Mike Galbraith
2008-11-17 19:57 ` Ingo Molnar
2008-11-17 20:20 ` (avc_has_perm_noaudit()) " Ingo Molnar
2008-11-17 20:32 ` ip_queue_xmit(): " Ingo Molnar
2008-11-17 20:57 ` Eric Dumazet
2008-11-18 9:12 ` Nick Piggin
2008-11-17 20:47 ` Ingo Molnar
2008-11-17 20:56 ` Eric Dumazet
2008-11-17 20:55 ` skb_release_head_state(): " Ingo Molnar
2008-11-17 21:01 ` David Miller
2008-11-17 21:04 ` Eric Dumazet
2008-11-17 21:34 ` Linus Torvalds
2008-11-17 21:38 ` Ingo Molnar
2008-11-17 21:09 ` tcp_ack(): " Ingo Molnar
2008-11-17 21:19 ` tcp_recvmsg(): " Ingo Molnar
2008-11-17 21:26 ` eth_type_trans(): " Ingo Molnar
2008-11-17 21:40 ` Eric Dumazet
2008-11-17 23:41 ` Eric Dumazet
2008-11-18 0:01 ` Linus Torvalds
2008-11-18 8:35 ` Eric Dumazet
2008-11-17 21:52 ` Linus Torvalds
2008-11-18 5:16 ` David Miller
2008-11-18 5:35 ` Eric Dumazet
2008-11-18 7:00 ` David Miller
2008-11-18 8:30 ` Ingo Molnar
2008-11-18 8:49 ` Eric Dumazet
2008-11-17 21:35 ` __inet_lookup_established(): " Ingo Molnar
2008-11-17 22:14 ` Eric Dumazet
2008-11-17 21:59 ` system_call() - " Ingo Molnar
2008-11-17 22:09 ` Linus Torvalds
2008-11-17 22:08 ` Ingo Molnar
2008-11-17 22:15 ` Eric Dumazet
2008-11-17 22:26 ` Ingo Molnar
2008-11-17 22:39 ` Eric Dumazet
2008-11-18 5:23 ` David Miller
2008-11-18 8:45 ` Ingo Molnar
2008-11-17 22:14 ` tcp_transmit_skb() - " Ingo Molnar
2008-11-17 22:19 ` Ingo Molnar
2008-11-17 19:36 ` David Miller
2008-11-17 19:31 ` David Miller
2008-11-17 19:47 ` Linus Torvalds
2008-11-17 19:51 ` David Miller
2008-11-17 19:53 ` Ingo Molnar
2008-11-17 22:47 ` Ingo Molnar
2008-11-17 19:21 ` David Miller
2008-11-17 19:48 ` Linus Torvalds
2008-11-17 19:52 ` David Miller
2008-11-17 19:57 ` Linus Torvalds
2008-11-17 20:18 ` David Miller
2008-11-19 19:43 ` Christoph Lameter
2008-11-19 20:14 ` Ingo Molnar
2008-11-20 23:52 ` Christoph Lameter
2008-11-21 8:30 ` Ingo Molnar
2008-11-21 8:51 ` Eric Dumazet
2008-11-21 9:05 ` David Miller
2008-11-21 12:51 ` Eric Dumazet
2008-11-21 15:13 ` Eric Dumazet [this message]
2008-11-21 15:21 ` [PATCH] fs: pipe/sockets/anon dentries should not have a parent Ingo Molnar
2008-11-21 15:28 ` Eric Dumazet
2008-11-21 15:34 ` Ingo Molnar
2008-11-26 23:27 ` [PATCH 0/6] fs: Scalability of sockets/pipes allocation/deallocation on SMP Eric Dumazet
2008-11-27 1:37 ` Christoph Lameter
2008-11-27 6:27 ` Eric Dumazet
2008-11-27 14:44 ` Christoph Lameter
2008-11-27 9:39 ` Christoph Hellwig
2008-11-28 18:03 ` Ingo Molnar
2008-11-28 18:47 ` Peter Zijlstra
2008-11-29 6:38 ` Christoph Hellwig
2008-11-29 8:07 ` Eric Dumazet
2008-11-29 8:43 ` [PATCH v2 0/5] " Eric Dumazet
2008-12-11 22:38 ` [PATCH v3 0/7] " Eric Dumazet
2008-12-11 22:38 ` [PATCH v3 1/7] fs: Use a percpu_counter to track nr_dentry Eric Dumazet
2007-07-24 1:24 ` Nick Piggin
2008-12-16 21:04 ` Paul E. McKenney
2008-12-11 22:39 ` [PATCH v3 2/7] fs: Use a percpu_counter to track nr_inodes Eric Dumazet
2007-07-24 1:30 ` Nick Piggin
2008-12-12 5:11 ` Eric Dumazet
2008-12-16 21:10 ` Paul E. McKenney
2008-12-11 22:39 ` [PATCH v3 3/7] fs: Introduce a per_cpu last_ino allocator Eric Dumazet
2007-07-24 1:34 ` Nick Piggin
2008-12-16 21:26 ` Paul E. McKenney
2008-12-11 22:39 ` [PATCH v3 4/7] fs: Introduce SINGLE dentries for pipes, socket, anon fd Eric Dumazet
2008-12-16 21:40 ` Paul E. McKenney
2008-12-11 22:40 ` [PATCH v3 5/7] fs: new_inode_single() and iput_single() Eric Dumazet
2008-12-16 21:41 ` Paul E. McKenney
2008-12-11 22:40 ` [PATCH v3 6/7] fs: struct file move from call_rcu() to SLAB_DESTROY_BY_RCU Eric Dumazet
2007-07-24 1:13 ` Nick Piggin
2008-12-12 2:50 ` Nick Piggin
2008-12-12 4:45 ` Eric Dumazet
2008-12-12 16:48 ` Eric Dumazet
2008-12-13 2:07 ` Christoph Lameter
2008-12-17 20:25 ` Eric Dumazet
2008-12-13 1:41 ` Christoph Lameter
2008-12-11 22:41 ` [PATCH v3 7/7] fs: MS_NOREFCOUNT Eric Dumazet
2008-11-29 8:43 ` [PATCH v2 1/5] fs: Use a percpu_counter to track nr_dentry Eric Dumazet
2008-11-29 8:43 ` [PATCH v2 2/5] fs: Use a percpu_counter to track nr_inodes Eric Dumazet
2008-11-29 8:44 ` [PATCH v2 3/5] fs: Introduce a per_cpu last_ino allocator Eric Dumazet
2008-11-29 8:44 ` [PATCH v2 4/5] fs: Introduce SINGLE dentries for pipes, socket, anon fd Eric Dumazet
2008-11-29 10:38 ` Jörn Engel
2008-11-29 11:14 ` Eric Dumazet
2008-11-29 8:45 ` [PATCH v2 5/5] fs: new_inode_single() and iput_single() Eric Dumazet
2008-11-29 11:14 ` Jörn Engel
2008-11-26 23:30 ` [PATCH 1/6] fs: Introduce a per_cpu nr_dentry Eric Dumazet
2008-11-27 9:41 ` Christoph Hellwig
2008-11-26 23:32 ` [PATCH 3/6] fs: Introduce a per_cpu last_ino allocator Eric Dumazet
2008-11-27 9:46 ` Christoph Hellwig
2008-11-26 23:32 ` [PATCH 4/6] fs: Introduce a per_cpu nr_inodes Eric Dumazet
2008-11-27 9:32 ` Peter Zijlstra
2008-11-27 9:39 ` Peter Zijlstra
2008-11-27 9:48 ` Christoph Hellwig
2008-11-27 10:01 ` Eric Dumazet
2008-11-27 10:07 ` Andi Kleen
2008-11-27 14:46 ` Christoph Lameter
2008-11-26 23:32 ` [PATCH 5/6] fs: Introduce special inodes Eric Dumazet
2008-11-27 8:20 ` David Miller
2008-11-26 23:32 ` [PATCH 6/6] fs: Introduce kern_mount_special() to mount special vfs Eric Dumazet
2008-11-27 8:21 ` David Miller
2008-11-27 9:53 ` Christoph Hellwig
2008-11-27 10:04 ` Eric Dumazet
2008-11-27 10:10 ` Christoph Hellwig
2008-11-28 9:26 ` Al Viro
2008-11-28 9:34 ` Al Viro
2008-11-28 18:02 ` Ingo Molnar
2008-11-28 18:58 ` Ingo Molnar
2008-11-28 22:20 ` Eric Dumazet
2008-11-28 22:37 ` Eric Dumazet
2008-11-28 22:43 ` Eric Dumazet
2008-11-21 15:36 ` [PATCH] fs: pipe/sockets/anon dentries should not have a parent Christoph Hellwig
2008-11-21 17:58 ` [PATCH] fs: pipe/sockets/anon dentries should have themselves as parent Eric Dumazet
2008-11-21 18:43 ` Matthew Wilcox
2008-11-23 3:53 ` Eric Dumazet
2008-11-21 9:18 ` [Bug #11308] tbench regression on each kernel release from 2.6.22 -> 2.6.28 Ingo Molnar
2008-11-21 9:03 ` David Miller
2008-11-21 16:11 ` Christoph Lameter
2008-11-21 18:06 ` Christoph Lameter
2008-11-21 18:16 ` Eric Dumazet
2008-11-21 18:19 ` Eric Dumazet
2008-11-16 17:40 ` [Bug #11664] acpi errors and random freeze on sony vaio sr Rafael J. Wysocki
2008-11-16 17:40 ` [Bug #11698] 2.6.27-rc7, freezes with > 1 s2ram cycle Rafael J. Wysocki
2008-11-16 17:40 ` [Bug #11404] BUG: in 2.6.23-rc3-git7 in do_cciss_intr Rafael J. Wysocki
2008-11-17 16:19 ` Randy Dunlap
2008-11-16 17:40 ` [Bug #11569] Panic stop CPUs regression Rafael J. Wysocki
2008-11-16 17:40 ` [Bug #11543] kernel panic: softlockup in tick_periodic() ??? Rafael J. Wysocki
2008-11-16 17:40 ` [Bug #11836] Scheduler on C2D CPU and latest 2.6.27 kernel Rafael J. Wysocki
2008-11-16 17:40 ` [Bug #11805] mounting XFS produces a segfault Rafael J. Wysocki
2008-11-17 14:44 ` Christoph Hellwig
2008-11-16 17:40 ` [Bug #11795] ks959-sir dongle no longer works under 2.6.27 (REGRESSION) Rafael J. Wysocki
2008-11-16 17:40 ` [Bug #11865] WOL for E100 Doesn't Work Anymore Rafael J. Wysocki
2008-11-16 17:40 ` [Bug #11843] usb hdd problems with 2.6.27.2 Rafael J. Wysocki
2008-11-16 21:37 ` Luciano Rocha
2008-11-16 17:40 ` [Bug #11876] RCU hang on cpu re-hotplug with 2.6.27rc8 Rafael J. Wysocki
2008-11-16 17:40 ` [Bug #11886] without serial console system doesn't poweroff Rafael J. Wysocki
2008-11-16 17:41 ` [Bug #12039] Regression: USB/DVB 2.6.26.8 --> 2.6.27.6 Rafael J. Wysocki
2008-11-16 17:41 ` [Bug #11983] iwlagn: wrong command queue 31, command id 0x0 Rafael J. Wysocki
2008-11-16 17:41 ` [Bug #12048] Regression in bonding between 2.6.26.8 and 2.6.27.6 Rafael J. Wysocki
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=4926D022.5060008@cosmosbay.com \
--to=dada1@cosmosbay.com \
--cc=a.p.zijlstra@chello.nl \
--cc=cl@linux-foundation.org \
--cc=davem@davemloft.net \
--cc=efault@gmx.de \
--cc=kernel-testers@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@elte.hu \
--cc=netdev@vger.kernel.org \
--cc=rjw@sisk.pl \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).