KS2009: Staging, linux-next, and the development process

Posted Oct 21, 2009 6:38 UTC (Wed) by dlang (guest, #313)
Parent article: KS2009: Staging, linux-next, and the development process

I wonder if it would make sense to have some sort of -pending tree that could contain patches that maintainers think may fix a regression, but are not sure about and gets the latest -rc merged into it after each release. that way people who run into regressions could pull the -pending tree for a quick test to see if the proposed fixes solve the problem or not.

currently you have to troll mailing lists, watch for the regression summary messages, or track down the maintainer of the specific component and ask if there is a fix (and that maintainer may or may not recognize your problem as being the same as the one the patch was written for)

KS2009: Staging, linux-next, and the development process

Posted Oct 21, 2009 10:02 UTC (Wed) by mingo (guest, #31122) [Link]

At least for the subsystems maintained in the -tip tree there's a consistent pattern you can use to recover fixlets: the tip:*/urgent branches.

The current branches are:

 
     core/urgent
      irq/urgent
     perf/urgent
    sched/urgent
   timers/urgent
  tracing/urgent
      x86/urgent

All the stuff separated out in the -tip urgent branches is directed towards Linus's current tree - and if it has a Cc: stable tag then it's directed towards earlier versions as well.