linux-kernel - Re: [PATCH 4/7] k3dma: Add cyclic mode for audio

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: <CALAqxLXFv90c94svtfrW56Dh+6bbVJWu6azfbR7v++Bs3OL82w@mail.gmail.com>
Date:	Fri, 29 Jul 2016 15:38:01 -0700
From:	John Stultz <john.stultz@...aro.org>
To:	Vinod Koul <vinod.koul@...el.com>
Cc:	lkml <linux-kernel@...r.kernel.org>,
	Andy Green <andy.green@...aro.org>,
	Zhangfei Gao <zhangfei.gao@...aro.org>,
	Jingoo Han <jg1.han@...sung.com>,
	Krzysztof Kozlowski <k.kozlowski@...sung.com>,
	Maxime Ripard <maxime.ripard@...e-electrons.com>,
	Dan Williams <dan.j.williams@...el.com>,
	Mark Brown <broonie@...nel.org>, Andy Green <andy@...mcat.com>
Subject: Re: [PATCH 4/7] k3dma: Add cyclic mode for audio

On Sun, Jul 24, 2016 at 12:31 AM, Vinod Koul <vinod.koul@...el.com> wrote:
> On Wed, Jul 20, 2016 at 08:53:06PM -0700, John Stultz wrote:
>
>>       ds = kzalloc(sizeof(*ds) + num * sizeof(ds->desc_hw[0]), GFP_ATOMIC);
>> -     if (!ds) {
>> -             dev_dbg(chan->device->dev, "vchan %p: kzalloc fail\n", &c->vc);
>> +     if (!ds)
>
> This is an unrelated change

Ack. And Its fallen out with recent changes in 4.8-rc


>> +
>> +     ds = kzalloc(sizeof(*ds) + num * sizeof(ds->desc_hw[0]), GFP_ATOMIC);
>
> we should use GFP_NOWAIT. And looks like driver doesn't use GFP_NOWAIT< so
> you may fix that up as well

Done.

>> +     if (!ds)
>> +             return NULL;
>> +
>> +     ds->desc_hw_lli = __virt_to_phys((unsigned long)&ds->desc_hw[0]);
>
> why no dma_map_single(()?
>
> Also __api is internal APIs, driver should not use them. Why not plain
> virt_to_phys()

Right. This is reworked by one of the following patches, but I
reordered things in my tree so those fixes land before we add the
cyclic feature.

I'll resend after the merge-window is over (unless you'd like to see it sooner).

thanks
-john