[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: probably a general problem concerning latest X-Version



typo, sorry, x IS NOW running


jdaues wrote:
Peter,

Thanks for you comments. Based on them, Xis not running. I have not installed nvidia drivers yet, so I am not taking full advantage of the hw, but at least its up. I am hoping there is a way to install the nividia driver without recompiing linux source.

John

Peter Mattern wrote:
Within the last months I could install several different etch-Versions (beta, RC1) on both a desktop system based on Gigabyte's GA-M55plus-S3G (grapics: nvidia GeForce 6100) and an Fujitsu-Siemens Amilo Pro 2055 (graphics VIA UniChrome) without any problems concerning X. This means, the installer chose, lacking a special driver for the graphics, the vesa driver, enabling me to install the proprietary nvidia driver and the one from openchrome.org in a second step - all without any problems.

Since a couple of days I also get the message about the missing kbd and mouse modules - on both machines, and using exactly the same old RC1 installation media. Furthermore, the debian installer installs only the fglrx (!) driver. The problem also appears the same way using weekly snapshot and daily build of 2007/01/06.

Solution:
Manually install xserver-xorg-input-kbd/mouse, xserver-xorg-video-vesa (the lack of this module after the normal installation is the reason why just entering the module within xorg.conf doesn't work). Run dpkg-reconfigure xserver-xorg, chose the manual recognition of the graphics card and thus urge X to use the vesa driver.
Then, in a second step the appropriate driver can be installed.

at jdaues: on the Gigabyte system, everything worked fine using the proprietary driver from nvidia.com. I first had to install linux-source-2.6.x (there's only one package fiting this scheme available), linux-headers-<result from uname -r>, pkg-config, x-dev. The installation had to be done from a text-console with no X running (which can be achieved easily using kdm, not using gdm, as far as I know).

at sdpatt: I actually think that this is a bug that ought to be reported, isn't it? I'm just not sure about how because it's not obvious to me which packet(s) are concerned ...







Reply to: