Re: [PATCH v2 1/2] tty: serial: pch_uart: replace chan->device->dev with dmaengine_get_provider_device()

[Andy Shevchenko]

On Thu, Nov 05, 2026 at 12:04:25AM +0100, Krzysztof Kozlowski wrote:
> On 04/11/2026 21:49, Andy Shevchenko wrote:
> > On Wed, Nov 04, 2026 at 09:52:06AM -0500, Frank.Li@xxxxxxxxxxx wrote:
> >>
> >> Use the dmaengine_get_provider_device() API instead of directly
> >> accessing chan->device->dev to get the DMA provider device.
> >>
> >> No functional change. Prepare make to struct dma_device private to the DMA
> >> engine.
> > 
> > Krzysztof, this patch was a split from the bundle by a reviewer request (not me
> > FWIW), and author obeyed. So, you can do the same in cases when you got something
> > you want to review independently on other content which most likely is a noise
> > (to you).
> 
> Do you propose that I read all patches and complain that partial reviews
> should be split? How does this proposal solve my complain that:
> 1. Incomplete/partial reviews will be marked as full review of the commit.

You mean the problem of tooling? Why is it problem of the tag?

> 2. People get same credit even if they intentionally did not read 95% of
> the patch as the people who read 100% of the patch.

This doesn't stand the same case when patch is split. I will give a Rb for the
full (but isolated) patch in comparison the same code being reviewed in the big
bundle. From my perspective (as a reviewer) I have done almost the same work.
So it's an equivalent.

-- 
With Best Regards,
Andy Shevchenko