linux-kernel - Re: [PATCH] media: v4l2-mem2mem: allow device run without buf

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] [day] [month] [year] [list]

Date:   Sat, 26 Nov 2022 16:28:33 +0800
From:   ayaka <ayaka@...lik.info>
To:     Nicolas Dufresne <nicolas.dufresne@...labora.com>
Cc:     Hsia-Jun Li <randy.li@...aptics.com>, linux-media@...r.kernel.org,
        mchehab@...nel.org, hverkuil-cisco@...all.nl,
        sebastian.fricke@...labora.com,
        kieran.bingham+renesas@...asonboard.com,
        laurent.pinchart@...asonboard.com, ming.qian@....com,
        wenst@...omium.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] media: v4l2-mem2mem: allow device run without buf



> On Nov 25, 2022, at 9:43 PM, Nicolas Dufresne <nicolas.dufresne@...labora.com> wrote:
> 
> CAUTION: Email originated externally, do not click links or open attachments unless you recognize the sender and know the content is safe.
> 
> 
>> Le mercredi 23 novembre 2022 à 17:24 +0800, Hsia-Jun Li a écrit :
>> From: Randy Li <ayaka@...lik.info>
>> 
>> For the decoder supports Dynamic Resolution Change,
>> we don't need to allocate any CAPTURE or graphics buffer
>> for them at inital CAPTURE setup step.
>> 
>> We need to make the device run or we can't get those
>> metadata.
> 
> Nack: This is not how it works. I know the m2m framework make it difficult, but
> it is expected that the driver have a special state for OUTPUT streamon (before
> capture streamon). Please have a look at other drivers.
> 
I have some good reasons here that make dynamic resolution should happen in the device_run().
1. If the CAPTURE is STREAMON, when the resolution changed event should be triggered? Of course it would be in device_run(), there is no reason to make an special case here.

The following reasons may be better applied for encrypted(DRM) video, when no normal video stream parser could be invoked.
We don’t know whether the user input contains valid sequence header, would the sequence header applied to the current frame? Beside the metadata we need may not at the beginning of the buffer.
2. If it would cost lots of device time on parsing it, it even may even need to read more than one OUTPUT buffers, we would be better to fix this into the normal schedule procedure. Or it would block the other running contexts(instances)
3. We need extra methods to wait the other context done their work which breaks the original job queue.

buffered flag is a perfect way to fix these problems, I didn’t see any m2m driver uses them.

> Nicolas
>> 
>> Signed-off-by: Randy Li <ayaka@...lik.info>
>> ---
>> drivers/media/v4l2-core/v4l2-mem2mem.c | 5 +++--
>> 1 file changed, 3 insertions(+), 2 deletions(-)
>> 
>> diff --git a/drivers/media/v4l2-core/v4l2-mem2mem.c b/drivers/media/v4l2-core/v4l2-mem2mem.c
>> index be7fde1ed3ea..cd56d60fad9d 100644
>> --- a/drivers/media/v4l2-core/v4l2-mem2mem.c
>> +++ b/drivers/media/v4l2-core/v4l2-mem2mem.c
>> @@ -301,8 +301,9 @@ static void __v4l2_m2m_try_queue(struct v4l2_m2m_dev *m2m_dev,
>> 
>>      dprintk("Trying to schedule a job for m2m_ctx: %p\n", m2m_ctx);
>> 
>> -     if (!m2m_ctx->out_q_ctx.q.streaming
>> -         || !m2m_ctx->cap_q_ctx.q.streaming) {
>> +     if (!(m2m_ctx->out_q_ctx.q.streaming || m2m_ctx->out_q_ctx.buffered)
>> +         || !(m2m_ctx->cap_q_ctx.q.streaming
>> +              || m2m_ctx->cap_q_ctx.buffered)) {
>>              dprintk("Streaming needs to be on for both queues\n");
>>              return;
>>      }
>