From: ebiederm-aS9lmoZGLiVWk0Htik3J/w@public.gmane.org (Eric W. Biederman)
To: Herbert Xu <herbert-lOAM2aK0SrRLBo1qDEOMRrpzq4S04n8Q@public.gmane.org>
Cc: lguest-mnsaURCQ41sdnm+yROfE0A@public.gmane.org,
Christian Borntraeger
<borntraeger-tA70FqPdS9bQT0dZR+AlfA@public.gmane.org>,
"David S. Miller" <davem-fT/PcQaiUtIeIZ0/mPfg9Q@public.gmane.org>,
virtualization-qjLDD68F18O7TbgM5vRIOg@public.gmane.org,
Matias Zabaljauregui
<zabaljauregui-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>,
netdev-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
Patrick McHardy <kaber-dcUjhNyLwpNeoWH0uzbU5w@public.gmane.org>
Subject: Re: [PATCH 4/5] lguest: use KVM hypercalls
Date: Wed, 15 Apr 2009 06:35:58 -0700 [thread overview]
Message-ID: <m1skkaox8h.fsf@fess.ebiederm.org> (raw)
In-Reply-To: <20090415132802.GA11408-lOAM2aK0SrRLBo1qDEOMRrpzq4S04n8Q@public.gmane.org> (Herbert Xu's message of "Wed\, 15 Apr 2009 21\:28\:02 +0800")
Herbert Xu <herbert-lOAM2aK0SrRLBo1qDEOMRrpzq4S04n8Q@public.gmane.org> writes:
> On Wed, Apr 15, 2009 at 06:23:29AM -0700, Eric W. Biederman wrote:
>>
>> There is a GIGANTIC reason to have the wait queue on tfile.
>>
>> If you open a file, and do ip link del tapN you can still
>> be blocked waiting in poll.
>>
>> The problem is specifically free_poll_entry, where we call
>> remove_wait_queue and fput without calling any file methods.
>> So all of this happens without struct tun_file's count being
>> elevated. Which means tun_net_uninit can detach before we get
>> off of the stupid poll wait queue.
>
> What about taking a netdev refcount before calling poll_wait?
Because as far as I can tell we would just leak that refcount.
The poll code does not appear to call back into any of the file
methods when it frees itself from the wait queue.
Eric
next prev parent reply other threads:[~2009-04-15 13:35 UTC|newest]
Thread overview: 27+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <200903271022.38244.rusty@rustcorp.com.au>
[not found] ` <1238709324.5823.8.camel@odie.local>
[not found] ` <1239043798.27826.93.camel@zetabook>
[not found] ` <200904081021.39877.rusty@rustcorp.com.au>
[not found] ` <1239224319.17844.16.camel@zetabook>
[not found] ` <49DDE91A.8060603@trash.net>
[not found] ` <49DDF614.1060909@trash.net>
[not found] ` <m1bpr6hqrm.fsf@fess.ebiederm.org>
[not found] ` <49E47976.8020005@trash.net>
2009-04-15 8:36 ` [Lguest] [PATCH 4/5] lguest: use KVM hypercalls Herbert Xu
[not found] ` <20090415083610.GA8579-lOAM2aK0SrRLBo1qDEOMRrpzq4S04n8Q@public.gmane.org>
2009-04-15 8:47 ` Herbert Xu
2009-04-15 9:07 ` [Lguest] " Christian Borntraeger
2009-04-15 11:07 ` Patrick McHardy
2009-04-15 13:23 ` Eric W. Biederman
[not found] ` <m18wm2rqy6.fsf-+imSwln9KH6u2/kzUuoCbdi2O/JbrIOy@public.gmane.org>
2009-04-15 13:28 ` Herbert Xu
[not found] ` <20090415132802.GA11408-lOAM2aK0SrRLBo1qDEOMRrpzq4S04n8Q@public.gmane.org>
2009-04-15 13:35 ` Eric W. Biederman [this message]
[not found] ` <m1skkaox8h.fsf-+imSwln9KH6u2/kzUuoCbdi2O/JbrIOy@public.gmane.org>
2009-04-15 13:46 ` Herbert Xu
[not found] ` <20090415134610.GA11683-lOAM2aK0SrRLBo1qDEOMRrpzq4S04n8Q@public.gmane.org>
2009-04-15 13:55 ` Herbert Xu
[not found] ` <20090415135502.GA11827-lOAM2aK0SrRLBo1qDEOMRrpzq4S04n8Q@public.gmane.org>
2009-04-15 14:10 ` Eric W. Biederman
[not found] ` <m1ocuynh2f.fsf-+imSwln9KH6u2/kzUuoCbdi2O/JbrIOy@public.gmane.org>
2009-04-15 14:12 ` Herbert Xu
2009-04-15 14:06 ` [Lguest] " Eric W. Biederman
[not found] ` <m11vruovu5.fsf-+imSwln9KH6u2/kzUuoCbdi2O/JbrIOy@public.gmane.org>
2009-04-15 14:08 ` Herbert Xu
[not found] ` <20090415140819.GA11991-lOAM2aK0SrRLBo1qDEOMRrpzq4S04n8Q@public.gmane.org>
2009-04-15 14:18 ` Eric W. Biederman
[not found] ` <m1iql6m24b.fsf-+imSwln9KH6u2/kzUuoCbdi2O/JbrIOy@public.gmane.org>
2009-04-15 14:23 ` Herbert Xu
2009-04-15 14:38 ` Herbert Xu
[not found] ` <20090415143834.GA12384-lOAM2aK0SrRLBo1qDEOMRrpzq4S04n8Q@public.gmane.org>
2009-04-15 14:56 ` Eric W. Biederman
[not found] ` <m1zleiklsl.fsf-+imSwln9KH6u2/kzUuoCbdi2O/JbrIOy@public.gmane.org>
2009-04-15 22:27 ` Herbert Xu
2009-04-16 11:08 ` [1/2] tun: Only free a netdev when all tun descriptors are closed Herbert Xu
[not found] ` <20090416110818.GA20950-lOAM2aK0SrRLBo1qDEOMRrpzq4S04n8Q@public.gmane.org>
2009-04-16 11:09 ` [2/2] tun: Fix sk_sleep races when attaching/detaching Herbert Xu
[not found] ` <20090416110952.GB20950-lOAM2aK0SrRLBo1qDEOMRrpzq4S04n8Q@public.gmane.org>
2009-04-20 8:35 ` Herbert Xu
2009-04-20 9:26 ` David Miller
2009-04-20 9:35 ` Herbert Xu
2009-04-20 10:02 ` David Miller
2009-04-24 8:55 ` [1/2] tun: Only free a netdev when all tun descriptors are closed Christian Borntraeger
[not found] ` <200904241055.49794.borntraeger-tA70FqPdS9bQT0dZR+AlfA@public.gmane.org>
2009-04-24 12:11 ` Herbert Xu
[not found] ` <20090424121156.GA28039-lOAM2aK0SrRLBo1qDEOMRrpzq4S04n8Q@public.gmane.org>
2009-04-24 12:40 ` Christian Borntraeger
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=m1skkaox8h.fsf@fess.ebiederm.org \
--to=ebiederm-as9lmozglivwk0htik3j/w@public.gmane.org \
--cc=borntraeger-tA70FqPdS9bQT0dZR+AlfA@public.gmane.org \
--cc=davem-fT/PcQaiUtIeIZ0/mPfg9Q@public.gmane.org \
--cc=herbert-lOAM2aK0SrRLBo1qDEOMRrpzq4S04n8Q@public.gmane.org \
--cc=kaber-dcUjhNyLwpNeoWH0uzbU5w@public.gmane.org \
--cc=lguest-mnsaURCQ41sdnm+yROfE0A@public.gmane.org \
--cc=netdev-u79uwXL29TY76Z2rM5mHXA@public.gmane.org \
--cc=virtualization-qjLDD68F18O7TbgM5vRIOg@public.gmane.org \
--cc=zabaljauregui-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.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).