Re: [PATCH v3] regmap: add cache validity to REGCACHE_FLAT

[Sander Vanheule <sander@svanheule.net>]

Hi Mark,

On Fri, 2025-10-17 at 14:56 +0100, Mark Brown wrote:
> On Thu, Oct 16, 2025 at 10:49:40PM +0200, Sander Vanheule wrote:
> > Some quick grepping turned up 35 drivers which use REGCACHE_FLAT without any
> > cache initialization (see below), thus defaulting to the all-zero cache. This
> > includes the driver I think you found issues with on the pine64: sun8i-codec.c
> > If this driver (accidentally) relies on the zero-initialization, I would expect
> > the spurious interrupts to also pop up when switching to a maple cache, or when
> > using the num_reg_defaults_raw workaround.
> 
> Yes, that driver is relying on the current behaviour - I'd expect at
> least some of the other drivers are too.  
> 
> There's two separate things here, there's what the driver you're working
> on should do (which is to use a maple tree cache from the sounds of it)
> and there's potential improvements to the flat cache, or adding a new
> sparse flat cache.

For my driver I went with a maple cache, as that's the one I can get functional.

Meanwhile, I also noticed others on the linux-gpio list being unaware of the pitfalls of
the flat cache [1]. So I decided to send a v4 [2] of this patch, albeit split up in a new
flat-sparse variant and a warning for the current flat cache. Even with clearer
documentation on the limitations of the flat cache, devs/reviewers/maintainers would need
to be aware. I feel like it's better to just have less surprises.

[1] https://lore.kernel.org/linux-gpio/e461ca08-ad28-44fe-85f1-afe332c1d43d@topic.nl/
[2] https://lore.kernel.org/lkml/20251022200408.63027-1-sander@svanheule.net/T/

Best,
Sander