Agrajag-: yeah i already tried swapping them - they're exactly the same monitor/model, so i wasn't really expecting anything different to happen
agd5f: well, if you get the same result, then you know it's not the monitor :)
rx__: hum..
rx__: regression in of agd5f's commits
rx__: something up with the text in VT
rx__: in one of :P
airlied: possibly the scaler..
kdekorte: airlied, you had mentioned a few weeks ago that you needed to fix something in the xserver to fix odd background colors in nautilus when radeon is running in EXA mode
kdekorte: and updates on that (as to Fedora 8)
rx__: why does the source look like he enabled rather than disable the scaler :)
rx__: reads bug 14001 for more info
rx__: oh.. enabling doesn't work just yet
agd5f: rx__: it only enables it if the USE_RMX bit is set in mode flags
agd5f: enable doesn't work yet anyway. sigh... I suppose we should save those regs as well... tons of them though
rx__: ah i see
rx__: i think that rules scaling out as the culprit
agd5f: rx__: no, it's the scaling. my last commit disables it. your test mode must use it
agd5f: *text
rx__: oh
damentz: agd5f: have you guys worked on mode changing lately?
damentz: anything i can test in git?
rx__: my guess is no from git log :)
damentz: ya i just looked at it now
damentz: RADEON: clean up output handling
damentz: i don't think thats related
agd5f: damentz: goes this patch help? http://www.botchco.com/alex/xorg/damentz.diff
damentz: agd5f: hmm, let me quickly patch mine
damentz: i'll use latest git sources from debian
damentz: its day 9 of january
agd5f: should be fine
damentz: agd5f: i think i found out why the radeon driver does everything other than graphics optimization and implementation better than fglrx
damentz: your code is readable, even to me
damentz: and i'm just a cs3 student
damentz: computer science 3 in a lowly highschool
damentz: and we're force to learn java... what a waste
damentz: alright its building
damentz: agd5f: i didn't run the patch, but i found the file and just commented the line you pointed out
agd5f: k
agd5f: does it help?
damentz: still compiling
damentz: slow laptop
agd5f: cool
damentz: 1.2ghz pentium 4-m
damentz: ya, not even a pentium m
damentz: its like a 600mhz old athlon
damentz: but its practically one of the most inefficient cpus i've ever heard of
damentz: and somehow i managed to get it through a handy down
damentz: agd5f: running it, so far so good - it gets to kdm
damentz: agd5f: it doesn't work
agd5f: damentz: bummer
agd5f: damentz: are you on an XPRESS or an M7?
damentz: m7
agd5f: ok
agd5f: damentz: how about this one: http://www.botchco.com/alex/xorg/damentz.diff
damentz: agd5f: should i start from a clean source or keep the old patch?
agd5f: doesn't matter
damentz: agd5f: its patch -p1 < /directory/to/patch.diff right?
agd5f: yeah -p1
agd5f: from xf86-video-ati
damentz: alright its compiling again
damentz: agd5f: no, still doesn't work
agd5f: damentz: and it only happens with you have noOffscreenPixmaps set?
damentz: ya
damentz: or exa
airlied: agd5f: desktop switching seems to make the cursor wierdness happen..
agd5f: airlied: we could try playing with teh cursor lock bit when we update the image
airlied: I'll play around now ..
agd5f: or when we show/hide
airlied: hmm maybe I was running old tree..
airlied: it seems to be gone away.. will keep an eye out for it ..
damentz: agd5f: the laptop panel shuts off for a brief second when changing resolutions
damentz: with the 6.6 series it snapped to the new resolution
damentz: agd5f: it actually seems to be whenever 3d acceleration is used that the bloom happens
damentz: without the xaanooffscreenpixmaps option, i can change resolutions
damentz: but when i load warsow
damentz: for instance, the screen just goes white
agd5f: damentz: with the randr code we turn off the display before switching modes
damentz: agd5f: ohhh, that might be it
damentz: because the screen also goes off right before switching to a vt on the 6.6 series
damentz: and that causes it to bloom
damentz: pressing the lid button fixes that though
agd5f: doubtful. sounds display fifo related
agd5f: VT switch just saves/restores all the regs
damentz: agd5f: is there a quick patch you can make to prevent it from turning off the monitor?
damentz: my laptop panel is really unstandard and probably can't handle it
damentz: i _have_ to use IgnoreEDID with the 6.6 series to use the correct resolutions
agd5f: damentz: does dpms work ok?
damentz: hmm, haven't tried that with 6.7 yet
damentz: usually it doesn't last long enough
damentz: lets see with an xset ocmmand
damentz: agd5f: just did xset dpms force off
damentz: screen went white
damentz: i think thats it
agd5f: dpms and mode set both use the same function to turn off the outputs
damentz: ya, pressing hte power lid doesn't help
damentz: it just makes hte screen change colors
agd5f: lid is a whole other issue
damentz: agd5f: hmm, it has to be the dpms issue
damentz: or some monitor on/off stuff you might have changed
damentz: dpms worked in 6.6
agd5f: what I don't understand is why it only has a problem when you use exa or xaanooffsceenpixmaps
agd5f: damentz: did it work under 6.6.19x or just 6.6.3?
damentz: 6.6.193
damentz: 6.6.3 is old, i don't think i can even install it anymore
damentz: but i remember it working
damentz: i had to use it before
agd5f: damentz: http://www.botchco.com/alex/xorg/damentz.diff
damentz: agd5f: the laptop panel still switched off and on
damentz: and gave me a white screen again
agd5f: damentz: well, the dpms code w/ that patch is the same between 6.6 and 6.7
agd5f: very strange
damentz: agd5f: hmm, maybe its not dpms completely...
damentz: i just let it idle
damentz: the monitor went off
damentz: moved the touchpad, came back on
damentz: changed a resolution
damentz: white again