On 1/8/26 08:36, Daniel P. Smith wrote: > It was simply that it was the tip of torvalds/master at the time the > patch series was prepared. As far as we could find, there is not a lot > of guidance on commit selection in the patch guide. If there is a > generally accepted convention we should follow, we would be glad to > follow it.
Barring a specific dependency, pick something that is more likely to have been tested. -rc's are fine. Linus's releases are fine. Random snapshots in time are not likely to be widely tested and shouldn't be used.
