linux-kernel - Re: [PATCH v5 2/3] media: venus: Populate video encoder/decoder nodename entries

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: <47d81240-2717-48f6-89c5-f64f7bbd7505@linaro.org>
Date: Fri, 13 Dec 2024 22:46:35 +0000
From: Bryan O'Donoghue <bryan.odonoghue@...aro.org>
To: Nicolas Dufresne <nicolas@...fresne.ca>,
 Stanimir Varbanov <stanimir.k.varbanov@...il.com>,
 Vikash Garodia <quic_vgarodia@...cinc.com>,
 Mauro Carvalho Chehab <mchehab@...nel.org>, Rob Herring <robh@...nel.org>,
 Krzysztof Kozlowski <krzk+dt@...nel.org>, Conor Dooley <conor+dt@...nel.org>
Cc: quic_renjiang@...cinc.com, quic_vnagar@...cinc.com,
 quic_dikshita@...cinc.com, konradybcio@...nel.org,
 linux-media@...r.kernel.org, linux-arm-msm@...r.kernel.org,
 linux-kernel@...r.kernel.org, devicetree@...r.kernel.org,
 Stanimir Varbanov <stanimir.varbanov@...aro.org>
Subject: Re: [PATCH v5 2/3] media: venus: Populate video encoder/decoder
 nodename entries

On 13/12/2024 20:00, Nicolas Dufresne wrote:
> Hope this hardcoding of node name is historical ? 

Hardcoding is historical in dts.

We need to add two more chips into venus before iris is merged and at 
feature parity for HFI_6XX and above - HFI_GEN2

Something like this.

enum {
	HFI_1XX
	..
	HFI_6XX
	HFI_GEN2
	..
};

 > And not done for newer chips ?

HFI_6XX and above will be fully supported in "iris" with encoder/decoder 
selection done at session creation time.

Iris is being added phased. Basic decoder with one format, followed by 
decoder and additional formats.

Once we get to feature parity HFI_6XX and above will be supported in 
Iris and removed from venus.

Leaving HFI_4XX and below.

That's a long winded way of saying new chips minted from the fab will 
either be HFI_GEN2+ or HFI_6XX.

> We discourage userspace on relying on node names cause it always leads to
> complication and non-portable code.

Writing this driver from scratch - basically what HFI_6XX in Iris does, 
you'd select encoder/decoder when you create the initial session - the 
initial state.

For venus that's an unknown amount of work to do.

What we _could_ certainly do is make the static assignment in this 
series assignable via a kernel parameter.

I'd say though that's an additional series on top of this.

First pass here is just to fix up the original sin, not to improve 
selectivity, just yet.

---
bod