linux-kernel - Re: [RFC PATCH] usb: host: xhci: allow __GFP

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

Message-ID: <Pine.LNX.4.44L0.1905221055190.1410-100000@iolanthe.rowland.org>
Date:   Wed, 22 May 2019 10:56:01 -0400 (EDT)
From:   Alan Stern <stern@...land.harvard.edu>
To:     Oliver Neukum <oneukum@...e.com>
cc:     Jaewon Kim <jaewon31.kim@...il.com>,
        Christoph Hellwig <hch@...radead.org>, <linux-mm@...ck.org>,
        <gregkh@...uxfoundation.org>,
        Jaewon Kim <jaewon31.kim@...sung.com>,
        <m.szyprowski@...sung.com>, <ytk.lee@...sung.com>,
        <linux-kernel@...r.kernel.org>, <linux-usb@...r.kernel.org>
Subject: Re: [RFC PATCH] usb: host: xhci: allow __GFP_FS in dma allocation

On Wed, 22 May 2019, Oliver Neukum wrote:

> On Di, 2019-05-21 at 10:00 -0400, Alan Stern wrote:
> > 
> > Changing configurations amounts to much the same as disconnecting,
> > because both operations destroy all the existing interfaces.
> > 
> > Disconnect can arise in two different ways.
> > 
> >         Physical hot-unplug: All I/O operations will fail.
> > 
> >         Rmmod or unbind: I/O operations will succeed.
> > 
> > The second case is probably okay.  The first we can do nothing about.  
> > However, in either case we do need to make sure that memory allocations
> > do not require any writebacks.  This suggests that we need to call
> > memalloc_noio_save() from within usb_unbind_interface().
> 
> I agree with the problem, but I fail to see why this issue would be
> specific to USB. Shouldn't this be done in the device core layer?

Only for drivers that are on the block-device writeback path.  The 
device core doesn't know which drivers these are.

Alan Stern