linux-kernel - Re: [RFC PATCH] usb: dwc3: ep0: Fix mem corruption on OUT transfers of more than 512 bytes

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: <5577CC47.1000706@ti.com>
Date:	Wed, 10 Jun 2015 11:03:59 +0530
From:	Kishon Vijay Abraham I <kishon@...com>
To:	Alan Stern <stern@...land.harvard.edu>
CC:	Michael Trimarchi <michael@...rulasolutions.com>,
	Felipe Balbi <balbi@...com>, <gregkh@...uxfoundation.org>,
	<linux-omap@...r.kernel.org>, <nsekhar@...com>,
	<linux-kernel@...r.kernel.org>, <linux-usb@...r.kernel.org>
Subject: Re: [RFC PATCH] usb: dwc3: ep0: Fix mem corruption on OUT transfers
 of more than 512 bytes

Hi,

On Tuesday 09 June 2015 10:54 PM, Alan Stern wrote:
> On Tue, 9 Jun 2015, Kishon Vijay Abraham I wrote:
>
>>> But with a bounce buffer that's only 512 bytes long, you can never send
>>> an entire packet's worth of data.  If the bounce buffer is 1024 bytes
>>
>> for control endpoint, 512 bytes should be sufficient to send entire packet right?
>
> Yes, you're right.  I had confused control endpoints with bulk
> endpoints, where the maxpacket size is 1024.  Sorry for the mistake.

no problem.
>
>>> then you can send the entire first packet.  When that's done, you can
>>> send the second packet.  And so on.  It wouldn't be quite as fast, but
>>> for ep0 that shouldn't matter.
>>
>> right! this is a variant of what I tried to implement in chained TRB [1].
>> $subject tries just to avoid memory corruption instead of actually trying to
>> receive all the data.
>
> Okay.  If you take the $SUBJECT approach, I think it would be better
> for an URB submission to fail than for the host controller to send only
> part of the data.

Could be but we also want to prevent mem corruption in the case of a faulty 
host to be more robust.

Thanks
Kishon
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/