Re: [PATCH 06/14] ASoC: SOF: add a power status IPC

[Mark Brown <broonie@kernel.org>]

[-- Attachment #1: Type: text/plain, Size: 1010 bytes --]

On Fri, Mar 20, 2020 at 04:07:27PM +0100, Guennadi Liakhovetski wrote:
> On Fri, Mar 20, 2020 at 02:27:32PM +0100, Guennadi Liakhovetski wrote:

> > No, this isn't a completion - it's a counter. I've used atomic variables 
> > before, I cannot remember seeing any difficulties with their correct use 
> > described. Do you have a pointer?

> Actually I'd even say this isn't a problem. I think it's safe to say, you 
> won't suspend and resume your audio interface more often than every 10 
> seconds. That makes under 3200000 cycles per year. Even with 31 bits for a 
> signed integer that makes more than 600 years. I think we're safe.

The problem is that atomics are just incredibly error prone - for
example they're just a plain number, they're usually counting something
which is not being locked so you have to be careful any time you do
anything around them.  Their lack of structure makes them harder to
reason about than most other locking primitives, often harder than it's
worth.

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]