linux-kernel - Re: [PATCH 1/2] xen: privcmd: Switch from mutex to spinlock for irqfds

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <20240702101727.nytmrgdzcebtinbt@vireshk-i7>
Date: Tue, 2 Jul 2024 15:47:27 +0530
From: Viresh Kumar <viresh.kumar@...aro.org>
To: Juergen Gross <jgross@...e.com>
Cc: Stefano Stabellini <sstabellini@...nel.org>,
	Oleksandr Tyshchenko <oleksandr_tyshchenko@...m.com>,
	Vincent Guittot <vincent.guittot@...aro.org>,
	Alex Bennée <alex.bennee@...aro.org>,
	Manos Pitsidianakis <manos.pitsidianakis@...aro.org>,
	Paolo Bonzini <pbonzini@...hat.com>,
	Al Viro <viro@...iv.linux.org.uk>, xen-devel@...ts.xenproject.org,
	linux-kernel@...r.kernel.org
Subject: Re: [PATCH 1/2] xen: privcmd: Switch from mutex to spinlock for
 irqfds

On 02-07-24, 12:12, Juergen Gross wrote:
> On 18.06.24 11:42, Viresh Kumar wrote:
> > irqfd_wakeup() gets EPOLLHUP, when it is called by
> > eventfd_release() by way of wake_up_poll(&ctx->wqh, EPOLLHUP), which
> > gets called under spin_lock_irqsave(). We can't use a mutex here as it
> > will lead to a deadlock.
> > 
> > Fix it by switching over to a spin lock.
> > 
> > Reported-by: Al Viro <viro@...iv.linux.org.uk>
> > Signed-off-by: Viresh Kumar <viresh.kumar@...aro.org>
> > ---
> >   drivers/xen/privcmd.c | 26 +++++++++++++++-----------
> >   1 file changed, 15 insertions(+), 11 deletions(-)
> > 
> > diff --git a/drivers/xen/privcmd.c b/drivers/xen/privcmd.c
> > index 67dfa4778864..5ceb6c56cf3e 100644
> > --- a/drivers/xen/privcmd.c
> > +++ b/drivers/xen/privcmd.c
> > @@ -13,7 +13,6 @@
> >   #include <linux/file.h>
> >   #include <linux/kernel.h>
> >   #include <linux/module.h>
> > -#include <linux/mutex.h>
> 
> I don't think you can drop that. There is still the ioreq_lock mutex.

You are right. The header got included from somewhere else I believe
since the build didn't fail for me.

> I can fix that up while committing, with that:
> 
> Reviewed-by: Juergen Gross <jgross@...e.com>

Thanks.

-- 
viresh