Just for the records, next info, glint/PGX32 (TechSource Raptor 8P): The glint driver is most sparc-compatible, if it is built inside the old monolithic Xorg 6.9.0 tree.
I tried the following ,(modifying the src only comes at the end of all of this, because I first need to know what and why I would need to change ... ) : copying over all .c and .h files from 6.9.0 to http://xorg.freedesktop.org/releases/individual/driver/xf86-video-glint-1.1.0.tar.gz copying over all .c and .h files from 6.9.0 to ftp://ftp.freedesktop.org/pub/xorg/X11R7.2/src/driver/xf86-video-glint-X11R7.1-1.1.1.tar.gz taking all CFLAGS and -D defines from 6.9.0's glint dir into 7.x's configured Makefiles Moving it vice versa from modular gmake based 7.x into monolithic imake based 6.9.0 (except Makefile) etc. etc. ... Summary: The glint driver doesn't crash like the ati driver did, also rarely freezes any platform model/config. But the screen just remains blank (says "no signal on vga). Verified on SF280R and U60. The verbose Xorg log output doesn't indicate anything (from another machine via ssh). Running xorgconfig manually, choosing most careful/conservative screen settings sync rate setting does not help at all. However, just compile the driver with almost arbitrary flags inside the 6.9.0 tree, and it will just work. Odd !! Include files?? Linking?? Currently no idea. I will co-bundle a "good" 6.9.0-tree built driver with the binary packages. The modular glint src itself otherwise builds (and runs) fine, but is not too useful: For now we need the old binary, too, until we know what is affecting the new way of building the driver [even with identical source]. 'worked whole night on Xorg. Full tree automatically builds now, and results actually work! X and drv's now gcc built, via alan's pre-prepared good work. Still some things on todo list for (and before) release, but step forward after dumb testing over last weekend. Good night ~m This message posted from opensolaris.org
