linux-kernel - Re: [PATCH 2/2] media: usb: pwc: Don't use coherent DMA buffers for ISO transfer

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]

Date:   Mon, 30 Jul 2018 13:07:02 -0300
From:   Mauro Carvalho Chehab <mchehab+samsung@...nel.org>
To:     Ezequiel Garcia <ezequiel@...labora.com>
Cc:     "Matwey V. Kornilov" <matwey@....msu.ru>,
        Alan Stern <stern@...land.harvard.edu>,
        Hans de Goede <hdegoede@...hat.com>, hverkuil@...all.nl,
        mchehab@...nel.org,
        Laurent Pinchart <laurent.pinchart@...asonboard.com>,
        rostedt@...dmis.org, mingo@...hat.com, isely@...ox.com,
        bhumirks@...il.com, colin.king@...onical.com,
        linux-media@...r.kernel.org,
        open list <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH 2/2] media: usb: pwc: Don't use coherent DMA buffers for
 ISO transfer

Em Tue, 17 Jul 2018 17:10:22 -0300
Ezequiel Garcia <ezequiel@...labora.com> escreveu:

> Yeah, and not setting URB_NO_TRANSFER_DMA_MAP makes the USB core
> create DMA mappings and use the streaming API. Which makes more
> sense in hardware without hardware coherency.
>
> The only thing that bothers me with this patch is that it's not
> really something specific to this driver. If this fix is valid
> for pwc, then it's valid for all the drivers allocating coherent
> memory.

We're actually doing this change on other drivers:
	https://git.linuxtv.org/media_tree.git/commit/?id=d571b592c6206

I suspect that the reason why all USB media drivers were using
URB_NO_TRANSFER_DMA_MAP is just because the first media USB driver
upstream used it.

On that time, I remember I tried once to not use this flag, but there
was something that broke (perhaps I just didn't know enough about the
USB layer - or perhaps some fixes happened at USB core - allowing it
to be used with ISOC transfers).

Anyway, nowadays, I fail to see a reason why not let the USB core
do the DMA maps. On my tests after this patch, at the boards I tested
(arm and x86), I was unable to see any regressions.

Thanks,
Mauro