Date: Wed, 03 Mar 2004 00:48:21 -0500 From: Jason <jason@ec.rr.com> To: Tobias Aigner <lists@mangosuechtig.org> Cc: freebsd-questions@freebsd.org Subject: Re: DRI troubles with radeon Message-ID: <404571A5.5030809@ec.rr.com> In-Reply-To: <20040302151414.2528806c.lists@mangosuechtig.org> References: <20040302151414.2528806c.lists@mangosuechtig.org>
next in thread | previous in thread | raw e-mail | index | archive | help
Tobias Aigner wrote: >Hi there, > >I'm having some troubles getting DRI to work. I'm runnig an >ASUS A7N8X-Deluxe Board with NForce2 and a ATI Radeon 8500. First of all >I started using the 5.2.1-RELEASE with following kernel-options: > >device agp >device radeondrm > >this caused XFree86 (4.3.0 and 4.4.0) to freeze on startx when DRI was >enabled. Then I removed agp and radeondrm from my kernel-config, now >dmesg showed this when startx was launched: > >error: [drm:pid580:radeon_cp_init] *ERROR* radeon_cp_init called >without lock held >error: [drm:pid580:radeon_unlock] *ERROR* Process 580 >using kernel context 0 > >but X actually started (glxinfo sayed 'direct rendering: no'). After >some reading I started to put these options in my XF86Config: > >Option "ForcePCIMode" "true" >Option "AGPMode" "4" > >well, ok now glxinfo reported 'direct rendering: yes' but my performance >with glxgears should be much better: > >1167 frames in 5.0 seconds = 233.400 FPS >1403 frames in 5.0 seconds = 280.600 FPS > >dmesg showed this: > >drm0: <ATI Radeon QL R200 8500 LE> port 0xb000-0xb0ff mem >0xc2000000-0xc200ffff,0xb0000000-0xbfffffff irq 5 at device 0.0 on pci2 >info: [drm] Initialized radeon 1.10.0 20020828 on minor 0 info: [drm] >Loading R200 Microcode > >This problem is obviously caused by my mainboard, because my radeon used >to work with my old Gigabyte VIA mainboard. > >So I'm slowly runnig out of ideas what to do. > > > I have the same problem. I have to use force pci mode or my system or it freezes on startx. It seems to me to be a driver issue. I preload all the agp and video stuff yet there is no driver for the memory controllers. pci0: <memory, RAM> at device 0.1 (no driver attached) pci0: <memory, RAM> at device 0.2 (no driver attached) pci0: <memory, RAM> at device 0.3 (no driver attached) pci0: <memory, RAM> at device 0.4 (no driver attached) pci0: <memory, RAM> at device 0.5 (no driver attached) The agp chipset driver was ported from linux and unlike most other drivers has mc0, mc1, and a agp bridge device in the driver code. Since these seem unique to the nforce platform that could be why, but having a bridge and memory controller is not and no other drivers have it. So now I am trying to rewrite it. I must say it is a very hard thing for me to do. Its my first attempt at any real programing outside of class. If you check your dmesg you will also see 5 memory controllers, but the only documents I can get from nvidia only show 2 in use. The people at nvidia seem to be of no use, so you might not bother asking but enough people asking might get results. Sorry but if you want 3d today I have to recommend you use nvidia cards on nvidia boards, or any supported card on any other support board(like the via). By the way I think option agpmode 4x is irrelavent with pci mode. Let me know if you got anymore ideas. Jason
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?404571A5.5030809>