From: Martin Josefsson <gandalf@wlug.westbo.se>
To: Patrick McHardy <kaber@trash.net>
Cc: Thomas Graf <tgraf@suug.ch>, David Miller <davem@davemloft.net>,
jmorris@namei.org, netdev@vger.kernel.org, vnuorval@tcs.hut.fi,
usagi-core@linux-ipv6.org, yoshfuji@linux-ipv6.org,
anttit@tcs.hut.fi
Subject: Re: [RESEND 3/5] [NET]: Protocol Independant Policy Routing Rules Framework
Date: Fri, 28 Jul 2006 11:25:27 +0200 [thread overview]
Message-ID: <1154078728.23563.21.camel@localhost.localdomain> (raw)
In-Reply-To: <44C94529.5080605@trash.net>
[-- Attachment #1: Type: text/plain, Size: 2182 bytes --]
On Fri, 2006-07-28 at 00:58 +0200, Patrick McHardy wrote:
> > +int fib_rules_lookup(struct fib_rules_ops *ops, struct flowi *fl,
> > + int flags, struct fib_lookup_arg *arg)
> > +{
> > + struct fib_rule *rule;
> > + int err;
> > +
> > + rcu_read_lock();
> > +
> > + list_for_each_entry(rule, ops->rules_list, list) {
>
> Shouldn't that be list_for_each_entry_rcu?
Yes that's correct, it should.
> > + if (rule->ifname[0] && (rule->ifindex != fl->iif))
> > + continue;
> > +
> > + if (!ops->match(rule, fl, flags))
> > + continue;
> > +
> > + rcu_read_unlock();
> > +
> > + err = ops->action(rule, fl, flags, arg);
> > + if (err != -EAGAIN) {
> > + fib_rule_get(rule);
> > + arg->rule = rule;
> > + goto out;
> > + }
> > + }
> > +
> > + err = -ENETUNREACH;
> > +out:
> > + rcu_read_unlock();
>
> rcu_read_unlock might get called multiple times in the list iteration
> and once again here.
Yes, the rcu_read_unlock() in the list iteration is misplaced, it
shouldn't be there. Besides the unbalanced lock/unlocks it suffers from
the general issue described below
As a somewhat related note, I've just digged a bit through RCU land,
talked to dipankar and mckenney, and discovered that rcu_read_lock() /
rcu_read_unlock() aren't strictly needed in softirqs since preempt is
already disabled in softirqs. This means that you can use the result of
the rcu read-side critical outside of the rcu_read_lock() /
rcu_read_unlock() section. BUT this changes with the -rt kernel where
softirqs are preemptable and where rcu_read_lock() / rcu_read_unlock()
doesn't disable/enable preempt anymore, which means the rcu read-side
critical section is also preemptable. This means that we can get
preempted in the read-side critical section but the resulting grace
period won't occur until rcu_read_unlock() is called, which means that
using results of an read-side critical section outside of the critical
section is just not going to work in softirqs in -rt kernels.
I'm sure Ingo has reviewed the RCU usage in softirqs but I don't know if
there's been any changes in this area after his review.
--
/Martin
[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 189 bytes --]
next prev parent reply other threads:[~2006-07-28 9:26 UTC|newest]
Thread overview: 53+ messages / expand[flat|nested] mbox.gz Atom feed top
2006-07-26 22:11 [RFC] Multiple IPV6 Routing Tables & Policy Routing Thomas Graf
2006-07-26 22:00 ` [PATCH 1/5] [IPV6]: Remove ndiscs rt6_lock dependency Thomas Graf
2006-07-26 22:28 ` David Miller
2006-07-26 23:34 ` Tushar Gohad
2006-07-26 23:34 ` David Miller
2006-07-31 11:01 ` Ville Nuorvala
2006-07-26 22:00 ` [PATCH 2/5] [IPV6]: Multiple Routing Tables Thomas Graf
2006-07-26 22:39 ` David Miller
2006-07-26 22:48 ` Thomas Graf
2006-07-26 22:55 ` David Miller
2006-07-29 4:13 ` YOSHIFUJI Hideaki / 吉藤英明
2006-07-29 4:14 ` David Miller
2006-07-29 4:28 ` YOSHIFUJI Hideaki / 吉藤英明
2006-07-29 10:29 ` Thomas Graf
2006-07-31 13:55 ` Ville Nuorvala
2006-07-31 14:01 ` Herbert Xu
2006-07-31 14:02 ` Herbert Xu
2006-07-31 15:41 ` Thomas Graf
2006-07-31 20:09 ` David Miller
2006-07-31 15:34 ` Thomas Graf
2006-07-26 22:00 ` [PATCH 3/5] [NET]: Protocol Independant Policy Routing Rules Framework Thomas Graf
2006-07-26 22:41 ` David Miller
2006-07-27 5:58 ` James Morris
2006-07-27 6:02 ` David Miller
2006-07-27 22:39 ` [RESEND " Thomas Graf
2006-07-27 22:58 ` Patrick McHardy
2006-07-27 23:17 ` David Miller
2006-07-27 23:31 ` Patrick McHardy
2006-07-28 9:25 ` Martin Josefsson [this message]
2006-07-29 1:40 ` Patrick McHardy
2006-07-29 7:25 ` Martin Josefsson
2006-07-27 23:30 ` Patrick McHardy
2006-07-28 10:23 ` Thomas Graf
2006-07-31 14:46 ` Ville Nuorvala
2006-07-31 15:24 ` Thomas Graf
2006-07-31 18:01 ` Patrick McHardy
2006-07-31 20:01 ` Thomas Graf
2006-07-26 22:00 ` [PATCH 4/5] [IPV6]: Policy Routing Rules Thomas Graf
2006-07-26 22:42 ` David Miller
2006-07-26 23:26 ` David Miller
2006-07-26 23:33 ` David Miller
2006-07-26 23:40 ` David Miller
2006-07-27 22:40 ` [RESEND " Thomas Graf
2006-07-31 14:55 ` Ville Nuorvala
2006-07-26 22:00 ` [PATCH 5/5] [IPV4]: Use Protocol Independant Policy Routing Rules Framework Thomas Graf
2006-07-26 22:43 ` David Miller
2006-07-26 23:47 ` David Miller
2006-07-27 22:40 ` [RESEND " Thomas Graf
2006-07-28 6:10 ` [RFC] Multiple IPV6 Routing Tables & Policy Routing YOSHIFUJI Hideaki / 吉藤英明
2006-07-28 8:23 ` David Miller
2006-07-28 10:32 ` Thomas Graf
2006-07-29 4:27 ` YOSHIFUJI Hideaki / 吉藤英明
2006-07-31 11:01 ` Ville Nuorvala
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=1154078728.23563.21.camel@localhost.localdomain \
--to=gandalf@wlug.westbo.se \
--cc=anttit@tcs.hut.fi \
--cc=davem@davemloft.net \
--cc=jmorris@namei.org \
--cc=kaber@trash.net \
--cc=netdev@vger.kernel.org \
--cc=tgraf@suug.ch \
--cc=usagi-core@linux-ipv6.org \
--cc=vnuorval@tcs.hut.fi \
--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).