[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <CAJedcCy3fCLxcPP3a5iSXecX-dpKmsePiJ45HYKfJE2CZXq=6A@mail.gmail.com>
Date: Sat, 11 Mar 2023 00:45:19 +0800
From: Zheng Hacker <hackerzheng666@...il.com>
To: Yoshihiro Shimoda <yoshihiro.shimoda.uh@...esas.com>
Cc: Zheng Wang <zyytlz.wz@....com>,
"gregkh@...uxfoundation.org" <gregkh@...uxfoundation.org>,
"p.zabel@...gutronix.de" <p.zabel@...gutronix.de>,
Biju Das <biju.das.jz@...renesas.com>,
"phil.edworthy@...esas.com" <phil.edworthy@...esas.com>,
"linux-usb@...r.kernel.org" <linux-usb@...r.kernel.org>,
"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
"1395428693sheep@...il.com" <1395428693sheep@...il.com>,
"alex000young@...il.com" <alex000young@...il.com>
Subject: Re: [PATCH] usb: gadget: udc: renesas_usb3: Fix use after free bug in
renesas_usb3_remove due to race condition
Yoshihiro Shimoda <yoshihiro.shimoda.uh@...esas.com> 于2023年3月10日周五 12:56写道:
>
> Hi Zheng,
>
> > From: Zheng Wang, Sent: Friday, March 10, 2023 4:30 AM
> >
> > In renesas_usb3_probe, &usb3->role_work is bound with
> > renesas_usb3_role_work. renesas_usb3_start will be called
> > to start the work.
> >
> > If we remove the module which will call renesas_usb3_remove
> > to make cleanup, there may be a unfinished work. The possible
> > sequence is as follows:
>
> Thank you for the patch!
> I think we should remove double spaces like below. (Also "an" unfinished):
>
Thank you for your kind reminder.Will correct in the next version of patch.
> -----
> If we remove the module which will call renesas_usb3_remove
> to make cleanup, there may be an unfinished work. The possible
> sequence is as follows:
> -----
>
> > Fix it by canceling the work before cleanup in the renesas_usb3_remove
> >
> > CPU0 CPUc1
>
> s/CPUc1/CPU1/
>
>
> > |renesas_usb3_role_work
> > renesas_usb3_remove |
> > usb_role_switch_unregister |
> > device_unregister |
> > kfree(sw); |
> > free usb3->role_sw |
> > | usb_role_switch_set_role
> > | //use usb3->role_sw
>
> Adding a blank line here is better, I think.
Agree with that :)
>
> > Fixes: 39facfa01c9f ("usb: gadget: udc: renesas_usb3: Add register of usb role switch")
> > Signed-off-by: Zheng Wang <zyytlz.wz@....com>
> > ---
> > drivers/usb/gadget/udc/renesas_usb3.c | 2 ++
> > 1 file changed, 2 insertions(+)
> >
> > diff --git a/drivers/usb/gadget/udc/renesas_usb3.c b/drivers/usb/gadget/udc/renesas_usb3.c
> > index bee6bceafc4f..23b5f1706d25 100644
> > --- a/drivers/usb/gadget/udc/renesas_usb3.c
> > +++ b/drivers/usb/gadget/udc/renesas_usb3.c
> > @@ -2658,6 +2658,8 @@ static int renesas_usb3_remove(struct platform_device *pdev)
> > {
> > struct renesas_usb3 *usb3 = platform_get_drvdata(pdev);
> >
> > + cancel_work_sync(&usb3->extcon_work);
>
> IIUC, this work is not related to the issue.
> But, what do you think?
> If so, adding this should be a separate patch with the following Fixes tag:
>
> Fixes: 3b68e7ca3888 ("usb: gadget: udc: renesas_usb3: add extcon support")
>
Yes, I didn't find the uaf part for this work. As I didn't see any
code to cancel this work,
I added it then. I think you are right. In the next version of patch,
we'll only keep the
cancel of related work.
Best regards,
Zheng
>
> > + cancel_work_sync(&usb3->role_work);
> > debugfs_remove_recursive(usb3->dentry);
> > device_remove_file(&pdev->dev, &dev_attr_role);
> >
> > --
> > 2.25.1
>
Powered by blists - more mailing lists