Attention is currently required from: Nico Huber, Edward O'Callaghan.
1 comment:
Patchset:
I wanted to say: let's not bikeshed, we won't need it for long. […]
There are two reasons why I thought of default shutdown function.
First reason is illustrated in this patch: instead of adding the simplest (and identical code) shutdown function into individual masters, have it added once and reuse. There are ~10 more par masters in the same situation.
I am afraid we may need it for longer, the more I think about it. Strictly speaking, it matters when freeing the data (not allocation) moves to common code. But looking at how it currently works, shutdown function is needed to be registered (even if it does nothing), as a signal to free data. It all can be re-written, but that's what I mean "we may need it for longer".
Second reason, completely independent is that some time ago when we were discussing CB:51676, there was an idea that shutdown function should be required. I think, if it is required, maybe to have default function available? There is `default_spi_send_multicommand` for example.
To view, visit change 54890. To unsubscribe, or for help writing mail filters, visit settings.