On Sat, May 28, 2011 at 2:57 PM, Markus M. May <mmay@javafreedom.org> wrote:
On 28.05.2011 21:41, Rogutės Sparnuotos wrote:
C Anthony Risinger (2011-05-28 00:02):
does anyone else experience<title>?
i've been getting this for sometime -- on intel, nouveau, and radeon KMS drivers -- and it looks terrible. after the module loads within initramfs, the space *within* the old resolution is pure white, surrounded by normal black. i'm pretty sure the "whitebox" corresponds to the size of the previous resolution because it's different sizes on all my machines, and pretty small on my 1920x1280 laptop (@ 15.5in ... awesome :-)
anyways, long story short it turns back 2 black when written over (systemd/initscripts output) or cleared, so:
echo -en '\e[H\e[2J'
... takes care of it. i basically just did a `clear> out`, looked at the codes, and made a small initramfs hook from the above.
it works fine as is, but is anyone else seeing this? it annoys me because i don't run getty on tty1, so it's there forever unless i clear right away. not really a big deal, as it took longer to write this msg than fix it, but it started happening a few kernels ago IIRC and i'm wondering if we can just throw something like this into mkinitcpio (if `quiet`, or before output), because it looks terrible :-(
i'll open a report unless someone feels like throwing it in there, it's a pretty simple change.
No white boxes here with radeon KMS (built-in) under 2.6.38.3. Actually, I can even see the "PCI Device Listing ..." table from the BIOS when I switch to tty1.
Hm, me is having the same problems, using Grub 2. This problem is only seen, if I do use the external Graphic card in the laptop I am using.
No white boxes on another PC with radeon, Arch's kernel26 and initramfs.
thanks for input all ... i see it on these cards: (nouveau) nVidia Corporation G86M [Quadro FX 360M] (rev a1) (radeon) ATI Technologies Inc RV770 [Radeon HD 4850] (radeon) ATI Technologies Inc RS880 [Radeon HD 4200] (i915) Intel Corporation Mobile 945GME Express Integrated Graphics Controller (rev 03) ... just seemed weird to me that most others don't, and on such a wide array of cards. after installing arch under KVM for other reasons i realized i'm running a little bit older version of syslinux -- as i didn't actually reinstall it after pacamn upgrades -- so ill try that too. it doesnt happen under qemu, but im not using KMS in qemu (though one does exist now: http://lists.freedesktop.org/archives/dri-devel/2011-April/010396.html :haven't been able to try it out yet tho). meh, it doesn't matter in the end, i really just want to know if i'm doing something to cause it ... since it literally happens on every machine i own, and some at work. if anyone is interested in the solution i use: /lib/initcpio/install/clear ----------------------------- # vim:set ft=sh: install() { SCRIPT="clear" } help () { cat <<HELPEOF Clear the console immediately HELPEOF } ----------------------------- /lib/initcpio/hooks/clear ----------------------------- # vim: set ft=sh: run_hook () { echo -en '\e[H\e[2J' } ----------------------------- ... then add the `clear` hook before everything else ... yes, everything: # grep ^HOOKS /etc/mkinitcpio.conf HOOKS="clear base consolefont udev autodetect plymouth pata scsi sata filesystems" -- C Anthony