Re: [PATCH v2 20/34] media: iris: add video hardware internal buffer count and size calculation

[Dikshita Agarwal]

On 12/19/2023 5:36 PM, Bryan O'Donoghue wrote:
> On 18/12/2023 11:32, Dikshita Agarwal wrote:
>> drivers/media/platform/qcom/vcodec/iris/Makefile | 1 +
>> .../media/platform/qcom/vcodec/iris/iris_buffer.c | 48 ++
>> .../media/platform/qcom/vcodec/iris/iris_common.h | 1 +
>> .../media/platform/qcom/vcodec/iris/iris_core.h | 2 +
>> .../media/platform/qcom/vcodec/iris/iris_helpers.c | 13 +
>> .../media/platform/qcom/vcodec/iris/iris_helpers.h | 4 +
>> .../platform/qcom/vcodec/iris/iris_instance.h | 2 +
>> .../media/platform/qcom/vcodec/iris/iris_vdec.c | 1 +
>> .../media/platform/qcom/vcodec/iris/vpu_common.h | 8 +
>> .../media/platform/qcom/vcodec/iris/vpu_iris3.c | 6 +
>> .../platform/qcom/vcodec/iris/vpu_iris3_buffer.c | 201 +++++
>> .../platform/qcom/vcodec/iris/vpu_iris3_buffer.h | 845 +++++++++++++++++++++
> 
> So, unless the code you are adding effects existing upstream venus, I think
> it would be a bit easier to review if you squashed down changes that
> pertain to Iris only.
> 
> For example this patch seems to relate to Iris only, so why is it a
> progressive change within your series.
> 
> Similar comment for "add vb2 streaming and buffer ops" and other "add
> feature x" patches in this series.
> 
> If the change is contained to your own codebase, then progressive changes
> are more noise than content.
> 
> Please try to squash down changes - to reduce the number of patches and the
> total LOC being proposed here.
> 
As requested in V1, we separated out the iris driver code in functional
patches. Also, as commented in 1st patch, we are OK to have the common
helpers as separate series.

Patch 6 onward all the code is contained to iris driver only.
Could you please elaborate, what do you mean by squashing down the changes?
which patches you want us to squash?
> ---
> bod