Mike A. Harris wrote:

fedora@gxconcepts.com wrote:

...

I installed Fedora Core 3, but seem to be having lots of issues with video display and lockups. I'm currently using a Dell Inspiron 2500 laptop, which has the Intel 810 video chipset.

I first installed FC3 with SELinux Active and then installed a second istalled a second time with SELinux Disabled. I am working with a fresh install on both cases.

I'm unable to track down the source of the problem, has anyone seen this?

The i810 driver in X.Org X11 6.8.x is broken on i810 and i815 video hardware. The exact cause of the problem is not yet known. It could be a bug in the i810 video driver, the core X server code, or in the kernel side of the driver (DRM).

Someone using a nvidia driver reported a refresh problem that is similar to the problem that those with the 810/815 video hardware have really noticed. Is it possible that there is another xorg-x11 component causing the refresh error, other than the driver?

The bug has been filed in Red Hat bugzilla already, and is also filed in X.Org bugzilla and officially being tracked there. Anyone experiencing problems with i810 or i815 video on Fedora Core 3 can add themselves to the CC of the upstream bug report if they would like to track the issue and any progress being made on it by upstream X.Org developers/maintainers. The upstream bug report is locate at:

http://freedesktop.org/bugzilla/show_bug.cgi?id=1817

Thanks! I'll be tracking the bug and hopefully not commenting too much on the issue there.

Red Hat is tracking the issue in X.Org bugzilla above, and once a patch is available that fixes this issue, we will review it for consideration for any future FC3 X.Org update releases. In the mean time, any users who may be experiencing this issue can temporarily work around the issue by editing their /etc/X11/xorg.conf by hand, and adding the following line to the driver section:

Option "noaccel"

This is what works for me, but with "NoAccel". Are these options case sesitive?

This will disable 2D acceleration in the driver, which should result in slower but working video. Alternatively you can try using the "vesa" driver, but it is also unaccelerated and will likely be even slower.

For my particular 815 video, vesa causes diagonal lines across the display and is unusable as a driver selection.

Once X.Org has tracked down the cause of the issue and a patch is made available that fixes it, we will review it for future updates as mentioned above.

Looking around on the server that the belo link refers to, I noticed an i915 driver that is listed. Would stagnating the i810 driver at a level that worked for the 810/815 and adding a new class of drivers be better for keeping the old video hardware working and not hinder the newer similar, but different enough to cause problems drivers evolving? It seems that the first 810 /815 breakage was due to improvements for the 830 video hardware. I know that the 865G does not randomly crash any longer because of improvements made to the i810 driver that make these later video harware more reliable.

Thanks! This issue concerns me, since one video type was made more reliable and the other was made less functional and both use the same driver. One is bound to be killed by revisions to make the other videocard work more reliably.

Jim

Hope this helps.