I wonder if someone has this problem ? /var/log/Xorg.0.log says this: (--) RADEON(0): Chipset: "ATI Radeon Mobility M6 LY (AGP)" (ChipID = 0x4c59) (--) RADEON(0): Linear framebuffer at 0xd8000000 (--) RADEON(0): VideoRAM: 16384 kByte (32 bit DDR SDRAM) (II) RADEON(0): AGP card detected I have blinking triangles in descent3, even in the title screen. It happens like if everything were rendered to the front buffer, thus showing intermediate steps in the screen refresh. **But** I have the double-buffer extension: (II) LoadModule: "dbe" (II) Loading /usr/lib/modules/extensions/libdbe.so (II) Module dbe: vendor="X.Org Foundation" compiled for 6.8.2, module version = 1.0.0 Module class: X.Org Server Extension ABI class: X.Org Server Extension, version 0.2 (II) Loading extension DOUBLE-BUFFER Plus, quake3 and tuxracer work perfectly, no blinking at all.. Does someone know what could be done ? Thanks
I need: emerge info emerge -vp xorg-x11 I don't have the game
Created attachment 52867 [details] emerge info
emerge -pv xorg-x11 These are the packages that I would merge, in order: Calculating dependencies ...done! [ebuild R ] x11-base/xorg-x11-6.8.2-r1 -3dfx -3dnow +bitmap-fonts -cjk -debug +dlloader -dmx +doc -font-server -hardened +insecure-drivers -ipv6 -minimal +mmx +nls +opengl +pam -sdk +sse -static +truetype-fonts +type1-fonts (-uclibc) +xprint +xv 0 kB
I'm betting it's due to the silly CFLAGS.
Very possible. I'll try again after rebuilding org-x11 with -O2 :-)
What packages have you *ever* had on your system from breakmygentoo? Usage of BMG packages is pretty much a guarantee to not getting support from us on packages, as BMG stuff is of very poor quality in most cases and has been proven time and again to break things.
Yeah, most stuff is b0rked there. But I only include some of it: joel ~ $ ls -l /usr/local/bmg-main-filtered/* total 0 lrwxrwxrwx 1 root root 29 Jan 12 20:06 gimp -> ../../bmg-main/media-gfx/gimp I used to have gimp 2.1.x and some evolution betas.. so nothing worrying for opengl. Besides, I just tried something for descent3: switching to 24-bit color (at 1400x1050) needs too much memory for my 16MB board, so DRI gets disabled. Then when I start descent3, it's very slow (rendered in software) but it displays correctly. So it may be radeon-7500-specific.
Just out of curiosity, can you make a new xorg.conf with 24-bit color and 640x480 as your video res, then setup a .xinitrc to "exec descent3" rather than your normal window manager and see if it still has the triangles?
Chris, I just tried what you said, but it still has triangles. Don't get me wrong: it's not funny-colored triangles or anything. Rather, it seems like incomplete frames are rendered, when some front objects still miss a few triangles. It even happens in the main menu, and when triangles disappear, one can see the background through the holes in the menu.
Well, it definitely sounds like render problems with your driver. I would be willing to bet that your CFLAGS are causing it. One in particular is the -ffast-math, as that tells gcc to use imprecise math when doing calculations. It is faster, yes, provided you don't need precision (and nearly everything does).
actually Mesa (and thus the DRI drivers) expect -ffast-math and have code in place to handle it properly. however this doesn't rule out driver bugs. any chance you could try a DRI driver snapshot? http://dri.freedesktop.org/wiki/Download
Adam, I just tried the one from radeon-20050225-linux.i386.tar.bz2. I replaced /usr/lib/modules/dri/radeon_dri.so by the new one and restarted X, but unfortunately, saw no difference. Would it be useful that I try recompiling xorg-x11 with -O2 ?
Ok.. I recompiled xorg-x11 6.8.2-r1 with: CFLAGS="-pipe -O2" CXXFLAGS="-pipe -O2" LDFLAGS="" but it's still the same...
Have there been any more current releases of the drivers that might resolve this issue? Unfortunately, I'm not very up on anything ATI-related. Also, try using your own compiled libSDL and see if that helps any.
Unfortunately my previous laptop (Vaio Z1 with ATI Radeon Mobility M6) completely died, and my new one has an ATI X700. For this reason I can't experience/reproduce the problem anymore.. Sorry.
Alright... since nobody else has reported or confirmed this bug, I'm just going to mark it INVALID and claim it was a driver glitch... ;] If someone else experiences this, feel free to comment and we can REOPEN it then when we can get more information.
