re: problem with starting X in -current.
matthew green <[email protected]> Sat, 05 Jan 2019 08:17:24 +1100
| Newsgroups | gmane.os.netbsd.ports.ofppc |
|---|---|
| Message-ID | <[email protected]> |
Frank Wille writes: > matthew green wrote: > > > actually, it hangs on _todays_ netbsd-8. i'll try netbsd 8.0 > > release kernel -- nope, the netbsd-GENERIC in the 8.0 ofppc > > release hangs the same :-( > > Cannot confirm that. 8.0 GENERIC works here (also Xorg from 8.0beta > userland). Where does it hang? it hangs when starting X itself. > I also tried current from nyftp HEAD/latest. It hangs! So I cannot test > Xorg. Did this kernel really work for you? > > It hangs after configuring fwohci0, ieee1394if0, uhub0 and uhub1. Then > follows "Kernel PGM trap @ ...". > The backtrace from the kernel debuger shows: > idle_loop > cpu_lwp_bootstrap i haven't tried those kernels, but my own ones work OK mostly as long asi don't start X :-) NetBSD 8.99.30 (_pegi_) #8: Fri Jan 4 01:25:24 PST 2019 > > there are the old and the new drivers. the new one ONLY supports > > DRM with KMS, so it's so far is only useful for x86. > > But... for that reason I would expect that DRM/KMS drivers won't be part of > the xserver.tgz release for any other port than i386 or amd64? right. > Or does "so far" mean that it is planned to make those drivers useful for > all ports, even with restrictions (like hardware-accelerated GL, etc.)? well, it would be nice if the drm code work elsewhere. the old radeon driver is abandonware now, so we have to maintain it ourselves. > > last time i tried to make the old driver work on amd64 it was > > more effort than i could remember how to do. :-( i was able > > to get X started, but not anything like Xvideo or GL. > > So it was only a dumb frame buffer? Ok, Xvideo would be nice, but I have > never seen GL working on any of my NetBSD machines anway. ;) i forget. i was just trying to have a working hdmi-output on my old laptop, where drm2 (in -7, -8, or -current) crashes when i plug into the hdmi port and i can't figure out why (something is corrupted, and a later fault is seen.)