[oi-dev] MAJOR BREAKTHROUGH: On Solaris 11.0 and even Illumos it works now!!! __/__ Re[2]: [OpenIndiana-discuss] Sun/Oracle China's DRM//KMS Sol11.2 port backported to function on old-style gfxp_private from pre-2010 era but still immediatedly PANICS

[Мартин Бохниг]

 Results: I post test bins and code in a few hours.
New finding: The same GPU errors haven on a vanilla completely untouched Sol11.1 LiveDVD.

After having attempted all somehow possible settings in BIOS (sizes of stolen memory for gtt, lowest gpu clock etc) I replaced the 2700K with the original Celeron G530 and the same still happens, even with a GPU clock of only 400MHz.
I also ran memtest  and even replaced the main memory, no change.

Now, to make a long story short: 

A) Either it is a bug in my H61M-S2V-B3
BIOS Configuration: Award Software International, Inc. F8 03/27/2012

or

B) it is a bug in Sun/Oracle's DRM port 


or
 
C) a bug in DRM itself.
But whatever I changed in the backport (including among many other things using a different vmem arena), it never helped.
But as said: The Sol11.1 DVD has the same problem on my system.
It runs a long while but then - if DRM and compiz are enabled - it tries to dump rings and times out.
On the other hand I cannot believe that Oracle shipped such code which is that instable everywhere.
Therefore I can only assume that my Mainboard (BIOS??) is the reason.

Today I give you test bins and code.
Cleanup can come later.


As for the reurn to console problem: Even in the most verbose Debug output everything looks good.
But the console doesn't get restored (kbd yes, but not stdout).
You can logout from gnome on OI and re-login as often as you want. You could even disable gdm and enter commands blindly (which works!).

Also you could prepare 2 scripts, one to start Xorg via vesa and fone using a different xorg.conf (or no xorg.conf) for intel-drm/gem/kms.
So assuming you first start intel, exit X11 and enter /v (your vesa script) blindly, then exit X11 again, your console will be there again.

But so far I didn't find the right way (which function for .lastclose ???) to do this directly.
Well - most folks won't need this anyway. Much more problematic is the instability on my hardware.


This was just written in a hurry to write at least _any_thing.
No, I must really upload something today. Otherwise it;s not polished and not published even in a year (if somebody can still surf some Internet by that time, which is of course another question).


It was good that I replaced the CPU with the old Celeron and old mem (down from 16GB back to 8GB), because I'm forced to sell that stuff urgently no matter what. The cpu is already listed (199 EUR).




-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://openindiana.org/pipermail/oi-dev/attachments/20160106/7b62de0b/attachment.html>