On Mon, Apr 01, 2024 at 09:59:10PM +0100, Russell King (Oracle) wrote:
On Mon, Apr 01, 2024 at 04:17:53PM +0200, Oswald Buddenhagen wrote:
i think that speculative/rfc patches are a perfectly fine way to get things clarified, and the linux kernel is no exception to that.
This wasn't a "speculative/rfc" patch. Such patches get marked with "RFC" in the tag.
putting an obvious disclaimer/question section after a three-dash line is a perfectly sufficient way to mark such a patch. at least if the receiver is actually interested in cooperation rather than harping on formalities.
Comments are not always correct.
so how about taking the opportunity to fix this one?
I don't think this comment is incorrect.
"ALSA wants the byte-size of the FIFOs."
That is a fact when the flag you refer to is not set.
yet the formulation also suggests that this is something that just is, rather than something that the code has control over.
[...] At some point, knowledge has to be assumed.
the problem is the omission of specific information that is in this context at least as pertinent as the information that _was_ supplied.
the code is also somewhat special in that it implements an interface, which makes it more likely to be "visited" by outsiders than some implementation details. it makes sense to take that into account in related comments.