Hey guys,

quoted

I tried both rt-lts and rt from the repo, not from AUR, for a x86_64
install.

$ pacman -Q linux-rt-lts
linux-rt-lts 3.10.39_rt40-1

I only tried to boot it one time.
Startup finished with a black screen, the display manager wasn't
visible.

$ pacman -Q linux-rt
linux-rt 3.14.3_rt5-1

The first time I booted it, a restart happened during startup.
The second time I booted it, startup hung.

If so I have no idea...  Maybe ask Jordan, as he has been having a
similar problem with AMD.  Maybe try his linux-l-pa kernel from the AUR.

No, what Ralf is describing sounds different to me, for a couple of reasons;

1). I've never had a restart during startup.
2). I never had a blank screen on startup
3). I never had a problem with 3.10-rt [although, i suppose if the 3
RT patches have been backported, then 3.10-lts wouldn't boot for me,
either. - haven't used 3.10-rt since 3.12-rt came out.]

My Phenom just doesn't boot on -rt [3.12.x-rt9+], unless i revert the
3 problematic patches. Ralf seems to have something else happening.  I
know that a couple of weeks ago, Pawel [on this list] also tried
reverting the same patches on one of his problematic AMD systems,
which did not yield a working system for him, either. IIRC, he could
only boot into that kernel by passing some kernel paramters, disabling
SMP, etc. [and like Ralf, the symptoms didn't seem quite the same as
mine, or other AMD users who's systems don't work, without reverting
said patches.].

but everyone [that i know] who couldn't boot on 3.12.x-rt9+ can boot
just fine, when ditching those 3 rt patches. - So again, i think it is
likely that Pawel and Ralf's problem[s] may not be the same, as the
"revert_timers-dont_raise_softirq_unconditionally_and_fixes.patch"
resolves.

Jordan

--

   Joakim