From: Nikolay Aleksandrov <nikolay@redhat.com>
To: netdev@vger.kernel.org
Cc: Florian Westphal <fw@strlen.de>,
Nikolay Aleksandrov <nikolay@redhat.com>,
"David S. Miller" <davem@davemloft.net>,
Alexey Kuznetsov <kuznet@ms2.inr.ac.ru>,
James Morris <jmorris@namei.org>,
Hideaki YOSHIFUJI <yoshfuji@linux-ipv6.org>,
Patrick McHardy <kaber@trash.net>,
Alexander Aring <alex.aring@gmail.com>,
Eric Dumazet <eric.dumazet@gmail.com>
Subject: [PATCH net-next 9/9] inet: frag: set limits and make init_net's high_thresh limit global
Date: Thu, 24 Jul 2014 16:50:37 +0200 [thread overview]
Message-ID: <1406213437-6155-10-git-send-email-nikolay@redhat.com> (raw)
In-Reply-To: <1406213437-6155-1-git-send-email-nikolay@redhat.com>
This patch makes init_net's high_thresh limit to be the maximum for all
namespaces, thus introducing a global memory limit threshold equal to the
sum of the individual high_thresh limits which are capped.
It also introduces some sane minimums for low_thresh as it shouldn't be
able to drop below 0 (or > high_thresh in the unsigned case), and
overall low_thresh should not ever be above high_thresh, so we make the
following relations for a namespace:
init_net:
high_thresh - max(not capped), min(init_net low_thresh)
low_thresh - max(init_net high_thresh), min (0)
all other namespaces:
high_thresh = max(init_net high_thresh), min(namespace's low_thresh)
low_thresh = max(namespace's high_thresh), min(0)
The major issue with having low_thresh > high_thresh is that we'll
schedule eviction but never evict anything and thus rely only on the
timers.
Signed-off-by: Nikolay Aleksandrov <nikolay@redhat.com>
---
Documentation/networking/ip-sysctl.txt | 3 ++-
net/ieee802154/reassembly.c | 12 ++++++++++--
net/ipv4/ip_fragment.c | 10 ++++++++--
net/ipv6/netfilter/nf_conntrack_reasm.c | 12 ++++++++++--
net/ipv6/reassembly.c | 12 ++++++++++--
5 files changed, 40 insertions(+), 9 deletions(-)
diff --git a/Documentation/networking/ip-sysctl.txt b/Documentation/networking/ip-sysctl.txt
index e8c304e37831..29a93518bf18 100644
--- a/Documentation/networking/ip-sysctl.txt
+++ b/Documentation/networking/ip-sysctl.txt
@@ -101,7 +101,8 @@ ipfrag_high_thresh - INTEGER
Maximum memory used to reassemble IP fragments. When
ipfrag_high_thresh bytes of memory is allocated for this purpose,
the fragment handler will toss packets until ipfrag_low_thresh
- is reached.
+ is reached. This also serves as a maximum limit to namespaces
+ different from the initial one.
ipfrag_low_thresh - INTEGER
Maximum memory used to reassemble IP fragments before the kernel
diff --git a/net/ieee802154/reassembly.c b/net/ieee802154/reassembly.c
index 8da635d92a58..f13d4f32e207 100644
--- a/net/ieee802154/reassembly.c
+++ b/net/ieee802154/reassembly.c
@@ -386,20 +386,25 @@ err:
EXPORT_SYMBOL(lowpan_frag_rcv);
#ifdef CONFIG_SYSCTL
+static int zero;
+
static struct ctl_table lowpan_frags_ns_ctl_table[] = {
{
.procname = "6lowpanfrag_high_thresh",
.data = &init_net.ieee802154_lowpan.frags.high_thresh,
.maxlen = sizeof(int),
.mode = 0644,
- .proc_handler = proc_dointvec
+ .proc_handler = proc_dointvec_minmax,
+ .extra1 = &init_net.ieee802154_lowpan.frags.low_thresh
},
{
.procname = "6lowpanfrag_low_thresh",
.data = &init_net.ieee802154_lowpan.frags.low_thresh,
.maxlen = sizeof(int),
.mode = 0644,
- .proc_handler = proc_dointvec
+ .proc_handler = proc_dointvec_minmax,
+ .extra1 = &zero,
+ .extra2 = &init_net.ieee802154_lowpan.frags.high_thresh
},
{
.procname = "6lowpanfrag_time",
@@ -446,7 +451,10 @@ static int __net_init lowpan_frags_ns_sysctl_register(struct net *net)
goto err_alloc;
table[0].data = &ieee802154_lowpan->frags.high_thresh;
+ table[0].extra1 = &ieee802154_lowpan->frags.low_thresh;
+ table[0].extra2 = &init_net.ieee802154_lowpan.frags.high_thresh;
table[1].data = &ieee802154_lowpan->frags.low_thresh;
+ table[1].extra2 = &ieee802154_lowpan->frags.high_thresh;
table[2].data = &ieee802154_lowpan->frags.timeout;
table[3].data = &ieee802154_lowpan->max_dsize;
diff --git a/net/ipv4/ip_fragment.c b/net/ipv4/ip_fragment.c
index ccee68dffd6e..634fc31aa243 100644
--- a/net/ipv4/ip_fragment.c
+++ b/net/ipv4/ip_fragment.c
@@ -700,14 +700,17 @@ static struct ctl_table ip4_frags_ns_ctl_table[] = {
.data = &init_net.ipv4.frags.high_thresh,
.maxlen = sizeof(int),
.mode = 0644,
- .proc_handler = proc_dointvec
+ .proc_handler = proc_dointvec_minmax,
+ .extra1 = &init_net.ipv4.frags.low_thresh
},
{
.procname = "ipfrag_low_thresh",
.data = &init_net.ipv4.frags.low_thresh,
.maxlen = sizeof(int),
.mode = 0644,
- .proc_handler = proc_dointvec
+ .proc_handler = proc_dointvec_minmax,
+ .extra1 = &zero,
+ .extra2 = &init_net.ipv4.frags.high_thresh
},
{
.procname = "ipfrag_time",
@@ -752,7 +755,10 @@ static int __net_init ip4_frags_ns_ctl_register(struct net *net)
goto err_alloc;
table[0].data = &net->ipv4.frags.high_thresh;
+ table[0].extra1 = &net->ipv4.frags.low_thresh;
+ table[0].extra2 = &init_net.ipv4.frags.high_thresh;
table[1].data = &net->ipv4.frags.low_thresh;
+ table[1].extra2 = &net->ipv4.frags.high_thresh;
table[2].data = &net->ipv4.frags.timeout;
/* Don't export sysctls to unprivileged users */
diff --git a/net/ipv6/netfilter/nf_conntrack_reasm.c b/net/ipv6/netfilter/nf_conntrack_reasm.c
index 4d9da1e35f8c..3d4bccf6d67d 100644
--- a/net/ipv6/netfilter/nf_conntrack_reasm.c
+++ b/net/ipv6/netfilter/nf_conntrack_reasm.c
@@ -63,6 +63,8 @@ struct nf_ct_frag6_skb_cb
static struct inet_frags nf_frags;
#ifdef CONFIG_SYSCTL
+static int zero;
+
static struct ctl_table nf_ct_frag6_sysctl_table[] = {
{
.procname = "nf_conntrack_frag6_timeout",
@@ -76,14 +78,17 @@ static struct ctl_table nf_ct_frag6_sysctl_table[] = {
.data = &init_net.nf_frag.frags.low_thresh,
.maxlen = sizeof(unsigned int),
.mode = 0644,
- .proc_handler = proc_dointvec,
+ .proc_handler = proc_dointvec_minmax,
+ .extra1 = &zero,
+ .extra2 = &init_net.nf_frag.frags.high_thresh
},
{
.procname = "nf_conntrack_frag6_high_thresh",
.data = &init_net.nf_frag.frags.high_thresh,
.maxlen = sizeof(unsigned int),
.mode = 0644,
- .proc_handler = proc_dointvec,
+ .proc_handler = proc_dointvec_minmax,
+ .extra1 = &init_net.nf_frag.frags.low_thresh
},
{ }
};
@@ -102,7 +107,10 @@ static int nf_ct_frag6_sysctl_register(struct net *net)
table[0].data = &net->nf_frag.frags.timeout;
table[1].data = &net->nf_frag.frags.low_thresh;
+ table[1].extra2 = &net->nf_frag.frags.high_thresh;
table[2].data = &net->nf_frag.frags.high_thresh;
+ table[2].extra1 = &net->nf_frag.frags.low_thresh;
+ table[2].extra2 = &init_net.nf_frag.frags.high_thresh;
}
hdr = register_net_sysctl(net, "net/netfilter", table);
diff --git a/net/ipv6/reassembly.c b/net/ipv6/reassembly.c
index 57a9707b2032..f1709c4a289a 100644
--- a/net/ipv6/reassembly.c
+++ b/net/ipv6/reassembly.c
@@ -578,20 +578,25 @@ static const struct inet6_protocol frag_protocol =
};
#ifdef CONFIG_SYSCTL
+static int zero;
+
static struct ctl_table ip6_frags_ns_ctl_table[] = {
{
.procname = "ip6frag_high_thresh",
.data = &init_net.ipv6.frags.high_thresh,
.maxlen = sizeof(int),
.mode = 0644,
- .proc_handler = proc_dointvec
+ .proc_handler = proc_dointvec_minmax,
+ .extra1 = &init_net.ipv6.frags.low_thresh
},
{
.procname = "ip6frag_low_thresh",
.data = &init_net.ipv6.frags.low_thresh,
.maxlen = sizeof(int),
.mode = 0644,
- .proc_handler = proc_dointvec
+ .proc_handler = proc_dointvec_minmax,
+ .extra1 = &zero,
+ .extra2 = &init_net.ipv6.frags.high_thresh
},
{
.procname = "ip6frag_time",
@@ -628,7 +633,10 @@ static int __net_init ip6_frags_ns_sysctl_register(struct net *net)
goto err_alloc;
table[0].data = &net->ipv6.frags.high_thresh;
+ table[0].extra1 = &net->ipv6.frags.low_thresh;
+ table[0].extra2 = &init_net.ipv6.frags.high_thresh;
table[1].data = &net->ipv6.frags.low_thresh;
+ table[1].extra2 = &net->ipv6.frags.high_thresh;
table[2].data = &net->ipv6.frags.timeout;
/* Don't export sysctls to unprivileged users */
--
1.9.3
next prev parent reply other threads:[~2014-07-24 14:52 UTC|newest]
Thread overview: 15+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-07-24 14:50 [PATCH net-next 0/9] inet: frag: cleanup and update Nikolay Aleksandrov
2014-07-24 14:50 ` [PATCH net-next 1/9] inet: frag: constify match, hashfn and constructor arguments Nikolay Aleksandrov
2014-07-24 14:50 ` [PATCH net-next 2/9] inet: frag: remove hash size assumptions from callers Nikolay Aleksandrov
2014-07-24 14:50 ` [PATCH net-next 3/9] inet: frag: move evictor calls into frag_find function Nikolay Aleksandrov
2014-07-24 14:50 ` [PATCH net-next 4/9] inet: frag: move eviction of queues to work queue Nikolay Aleksandrov
2014-07-24 14:50 ` [PATCH net-next 5/9] inet: frag: don't account number of fragment queues Nikolay Aleksandrov
2014-07-25 6:15 ` David Miller
2014-07-25 8:32 ` Florian Westphal
2014-07-24 14:50 ` [PATCH net-next 6/9] inet: frag: remove lru list Nikolay Aleksandrov
2014-07-24 14:50 ` [PATCH net-next 7/9] inet: frag: remove periodic secret rebuild timer Nikolay Aleksandrov
2014-07-24 14:50 ` [PATCH net-next 8/9] inet: frag: use seqlock for hash rebuild Nikolay Aleksandrov
2014-07-24 14:50 ` Nikolay Aleksandrov [this message]
2014-07-28 5:37 ` [PATCH net-next 0/9] inet: frag: cleanup and update David Miller
2014-07-28 8:00 ` Florian Westphal
2014-07-28 6:43 ` Alexander Aring
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=1406213437-6155-10-git-send-email-nikolay@redhat.com \
--to=nikolay@redhat.com \
--cc=alex.aring@gmail.com \
--cc=davem@davemloft.net \
--cc=eric.dumazet@gmail.com \
--cc=fw@strlen.de \
--cc=jmorris@namei.org \
--cc=kaber@trash.net \
--cc=kuznet@ms2.inr.ac.ru \
--cc=netdev@vger.kernel.org \
--cc=yoshfuji@linux-ipv6.org \
/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).