Hi, (I'm cc:-ing the suspend-devel list, maybe somebody has an idea about the 3d problems).
On Sun, Jun 25, 2006 at 09:33:57PM +0100, Richard Neill wrote: > Here are the results for a thinkpad X22. > > 1)Identification s2ram -i > > This machine can be identified by: > sys_vendor = "IBM" > sys_product = "26629BG" > sys_version = "Not Available" > bios_version = "1DET51WW (1.12 )" > > > 2)Results from CONSOLE (init=/bin/bash) > > -f -a1 Machine hangs on resume (totally dead) I'd have expected this one to result in "backlight on, but VGA console not switched back to framebuffer", but well, it might be different on different BIOSes. > -f -a2 Works > > -f a3 Works "-a3" (s3_bios,s3_mode) is the way that works on almost all thinkpads, so that's the one i will add to the whitelist. > -f -p -s Works (but we must run udevstart to create /dev/zero) This is strange, since it _should_ be almost equivalent to "-a3", just that it is done a little bit later, when control is already handed back to userspace. > Also, creates some green vertical lines on the top of > the display This is normal with e.g. vesafb and bootsplash, we lack the proper save / restore of the framebuffer contents. A switch to vt2 and back to vt1 usually fixes that. It is only a cosmetic problem, that's why nobody bothered fixing it yet :-) > -f -s Works > > -f -p Partial: machine is alive; graphics unusable should be roughtly equivalent to > -f -a1 -s Works. Same as -f -p -s Expected, although i wonder why it does not hang the same as "-a1" above. > 3)Note: we need to use -r in order to turn off the backlight. Aha, so there are actually more machines out there that need this one than we would have expected... > 4)Results once X has started (with 3D acceleration enabled; this is the > ati driver in xorg 6.9, with the dri module loaded). We have already > done chvt 1. Somehow, the fact that the dri module has ever been loaded, > even if X is then stopped, and the radeon and drm kernel modules are > removed(!) is enough to make the machine hang. > > ALL of the above leave the machine totally hung. I'd say this is a Kernel and/or X server bug, nothing we (suspend-"crew" :-) can do much about. Reporting this bug on the linux kernel mailing list and maybe in the X.org bug tracking system might be a good idea. > 5)Results with X and appending vga=0 to kernel commandline: > > -f -a3 -r Works if kdm is started; hangs if user has logged in. > [vga=0] > > -f -p -s -r Hangs (totally dead) - even if user hasn't logged in. > [vga=0] > > -f -a2 -r Hangs (dead) > > > Tried the old trick: > killall -STOP X; s2ram -f -a2 -r; killall -CONT X > Result: it still hangs totally. > > > > 6)Without dri/drm: > The machine is fine (with -f -a3 -r) IFF I disable the dri module in > Xorg.conf. This works OK with vga=0 and also works fine with vga=791 > > [Besides which, even if we do use vga=0, this is really ugly. Is there > any way to get vga=791 back?] Well, this machine should work well with framebuffer (vga=791) AFAICS, there are only few thinkpads that don't. OTOH, machines that need vga=0 to function (usually newer HPs or DELLs with ATI graphics AFAICT) usually lock up or reboot hard on resume before even returning into kernelspace, and i do not think there is much we can do about that. In the end, the BIOS was not decigned for an OS with a VESA framebuffer console ;-) -- Stefan Seyfried \ "I didn't want to write for pay. I QA / R&D Team Mobile Devices \ wanted to be paid for what I write." SUSE LINUX Products GmbH, Nürnberg \ -- Leonard Cohen Using Tomcat but need to do more? Need to support web services, security? Get stuff done quickly with pre-integrated technology to make your job easier Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642 _______________________________________________ Suspend-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/suspend-devel
