Linux 2.6.36-rc3 Kernel Released

[phoronix]

Phoronix: Linux 2.6.36-rc3 Kernel Released

Linus Torvalds has just done a Sunday afternoon release of the Linux 2.6.36-rc3 kernel. With the merge window for the Linux 2.6.36 kernel having closed a few weeks ago, the third 2.6.36 release candidate isn't too exciting unless you were affected by one of the kernel's outstanding bugs...

http://www.phoronix.com/vr.php?view=ODU1NA

[Pfanne]

how is the radeon powermanagement in this kernel doing?

[Melcar]

Supposedly, super great on r5xx and lower, though as of rc2 "dynpm" mode still causes my laptop (RS480) to sometimes hard lock when freq. switching (same deal with the "auto" profile). Haven't had a chance to test my r6/7xx parts.

[allquixotic]

With Linus heading down to South America next week for LinuxCon Brazil, he won't be pulling any code in the coming days into what will eventually be the Linux 2.6.36-rc4 kernel and it may be over a week before we see this -rc4 tagged.

So another reminder that we need a Linus delegate. No need to replace Linus, but he needs to be able to give the keys to the castle to someone else for a while. Maybe he would want to take a break, and that doesn't mean the Linux release cycle should also take a break, as distros and IHVs and tons of other people depend on the release cycle. Or maybe Linus will get sick (hey, it's inevitable, we are all mortal). Would be better to plan ahead than to ask him to communicate his wishes with eyelid blinking at the end of the road...

The kernel is more important than Linus or anybody else imagined. One guy taking a business trip shouldn't hold up the cycle -- what if the trip were longer, and it caused a distro to have to ship an older kernel rather than the latest? I'm sure this is the kind of thing that keeps Mark Shuttleworth up at night, lol.

[gentoofu]

So another reminder that we need a Linus delegate. No need to replace Linus, but he needs to be able to give the keys to the castle to someone else for a while. Maybe he would want to take a break, and that doesn't mean the Linux release cycle should also take a break, as distros and IHVs and tons of other people depend on the release cycle. Or maybe Linus will get sick (hey, it's inevitable, we are all mortal). Would be better to plan ahead than to ask him to communicate his wishes with eyelid blinking at the end of the road...

The kernel is more important than Linus or anybody else imagined. One guy taking a business trip shouldn't hold up the cycle -- what if the trip were longer, and it caused a distro to have to ship an older kernel rather than the latest? I'm sure this is the kind of thing that keeps Mark Shuttleworth up at night, lol.

What's wrong with longer testing and regression fixing? RC5 may be the last release candidate for all we know.

[Melcar]

The point is that it´s just one guy you´re relying on; with a project of this magnitude and importance that situation is simple bad. We need at the very least an extra hand that can move things forward if shit happens. It would be interesting to know if Linus is the one not willing to share in his responsibilities/work or just that no one is willing to step up.