Using vesafb-tng with xf86-video-i810 video drivers, hibernate support using suspend2 (now tuxonice) worked fine. Upgrading to uvesafb, resume from S3 causes the system to hang with a garbled graphics output. Switching to vesa video driver alleviates the problem, as does using alternate framebuffer drivers. Reproducible: Always Steps to Reproduce: 1. Install tuxonice-sources (2.6.23) configured to use uvesafb on an intel 945GM system using xf86-video-i810 driver in X. 2. Suspend to S3 using hibernate-ram (hibernate script). 3. Attempt to resume from S3. Actual Results: System attempts to resume from S3, a garbled display is shown and the system locks up. Expected Results: Resumed from S3 in a functional state. Vesafb-tng works perfectly and supplies the 1280x800 resolution support required for the laptop. Uvesafb works fine except for the conflict with xf86-video-i810. No other framebuffers support the resolution required. As a possible "solution", we could put vesafb-tng back into the gentoo patchset to allow a choice between the framebuffers.
xf86-video-i810 version? emerge --info? Any relevant dmesg output? Also note that tuxonice-sources is not a supported kernel. [1] [1] http://www.gentoo.org/doc/en/gentoo-kernel.xml#doc_chap2
Yes, I have this problem too. I use uvesafb and xf86-video-intel. kernel is tuxonice-sources-2.6.28-r1 And I found tuxonice has nothing to do with this problem. The main reason is uvesafb conflict with intel video driver belonging to X, and the reason of conflict is MTRR. You can refer to this page: http://www.nabble.com/-Cooker--Re:--RPM--cooker-main-release-x11-driver-video-intel-2.5.1-2mdv2009.1-td20759691.html If I set uvesafb like this in kernel command line: video=uvesafb:1024x768-32,mtrr=0 or like this video=uvesafb:off # (use text mode) this problem gone. I can resume to X. But if I set this: video=uvesafb:1024x768,mtrr=3 I got the problem.