<div class="gmail_quote">On Sun, Sep 12, 2010 at 9:20 PM, Christopher James Halse Rogers <<a href="mailto:raof@ubuntu.com">raof@ubuntu.com</a>> wrote: <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;"> Hello again! For Lucid we asked the kernel team to blacklist KMS on i830, i845 & i855 to try to work around the instability in the intel drivers on those cards. This instability (while, apparently, slightly reduced by the 2.12 DDX) is still present in Maverick, so I asked the kernel team to re-enable the blacklist in a recent upload. Since we've got the 2.12 DDX with no userspace modesetting support, that means that these chips should get vesa. There's an issue¹ with X actually loading vesa rather than fbdev or intel which I need to investigate which is preventing this from happening now. The other option for these cards is to use KMS + the fbdev driver². There have been a couple of concerns raised about vesa: * Possibly less capable of proper suspend/resume than KMS. * Won't drive the display in the native mode for laptops with a poorly-written video bios. fbdev+KMS will work where ever KMS works, and will set the native mode where ever KMS sets the native mode. It is not capable of changing resolution, however. I've been persuaded that fbdev+KMS will provide a better experience for more people than kms-blacklist+vesa. This is a final call for comment before I ask the kernel team to revert the blacklist for these cards and flick the switch in xorg-server to use fbdev rather than intel. ¹: Bug #633593: <a href="https://bugs.edge.launchpad.net/ubuntu/+source/xserver-xorg-video-intel/+bug/633593" target="_blank">https://bugs.edge.launchpad.net/ubuntu/+source/xserver-xorg-video-intel/+bug/633593</a> ²: This is similar in some respects to the upstream “Shadow branch” work. </blockquote><div>This seems an acceptable resolution to me. </div></div> -- Greg Bair<div><a href="mailto:gregbair@gmail.com" target="_blank">gregbair@gmail.com</a></div>