I have not been able to get 3D (OpenGL) running with either x86_64 or x86 versions of FC6, using both the opensource and proprietary ATI drivers, on a HP dx5150. 2D works out of box.
relevent info: [EMAIL PROTECTED] ~]# lspci ... 01:05.0 VGA compatible controller: ATI Technologies Inc RS480 [Radeon Xpress 200 G Series] 01:05.1 Display controller: ATI Technologies Inc Unknown device 5854 ... [EMAIL PROTECTED] ~]# dmesg |grep agpgart Linux agpgart interface v0.101 (c) Dave Jones [EMAIL PROTECTED] ~]# grep /var/log/Xorg.0.log -e WW -e EE ... (WW) fglrx: No matching Device section for instance (BusID PCI:1:5:1) found (EE) end of block range 0x1fffffff < begin 0xe0000000 (WW) fglrx(0): Failed to open DRM connection (WW) fglrx(0): board is an unknown third party board, chipset is supported (WW) fglrx(0): *********************************************** (WW) fglrx(0): * DRI initialization failed! * (WW) fglrx(0): * (maybe driver kernel module missing or bad) * (WW) fglrx(0): * 2D acceleraton available (MMIO) * (WW) fglrx(0): * no 3D acceleration available * (WW) fglrx(0): ********************************************* * .... >From googling, I realise I have to recompile the kernel,- agpgart is built in >by default, and the amd64-agp is not compiled. (The following modules are >required to be loaded in sequence:) agpgart amd64-agp drm (radeon or fglrx) There seems to be something strange with the new kernels - I've tried 2.6.18-2 and 2.6.19... Options relevant to agpgart, are not visable with "make menuconfig", though the option is listed in the .config file. There is a discussion on this, and the link between agpgart and iommu, the need to disable iommu (without much discussion on the repercussions) and the fact that in the newer kernels, IOMMU is not tog-able by default... ..... # Mark as embedded because too many people got it wrong. # The code disables itself when not needed. config IOMMU .... This hardware has a third-party card (ChipID 5854) which is listed on the ATI windows catalyst drivers. Has anyone got an easier solution - like maybe positive results with an earlier kernel? I still have not got the kernel compiled correctly to get openGl working, and am dreading trying to scale this solution to a few hundred machines. Andrew Lynn Ph.D Asstt. Professor School of Information Technology Jawaharlal Nehru University New Delhi 110067 _______________________________________________ ilugd mailinglist -- ilugd@lists.linux-delhi.org http://frodo.hserus.net/mailman/listinfo/ilugd Archives at: http://news.gmane.org/gmane.user-groups.linux.delhi http://www.mail-archive.com/ilugd@lists.linux-delhi.org/
