udovdh: good morning
udovdh: stupid question:
udovdh: I was trying to build radeonhd on my second box. git clone was OK.
udovdh: then this:
udovdh: # ./autogen.sh
udovdh: autoreconf: Entering directory `.'
udovdh: autoreconf: configure.ac: not using Gettext
udovdh: autoreconf: running: aclocal
udovdh: autoreconf: configure.ac: tracing
udovdh: autoreconf: running: libtoolize --copy
udovdh: autoreconf: running: /usr/bin/autoconf
udovdh: configure.ac:41: error: possibly undefined macro: XORG_DRIVER_CHECK_EXT
udovdh: If this token and others are legitimate, please use m4_pattern_allow.
udovdh: See the Autoconf documentation.
udovdh: autoreconf: /usr/bin/autoconf failed with exit status: 1
udovdh: any ideas?
arekm: udovdh: install xserver-devel
udovdh: ah, OK, thanks
udovdh: will try
udovdh: yup, xorg-x11-server-sdk did the trick as it looks...
gavinreid: guys is it possible to get compiz fusion effect running on radeonhd?
gavinreid: !compiz
dli: gavinreid, no
dli: gavinreid, I wonder why you would ask still
gavinreid: dli I've had problems getting the drivers to work in the first place I wasnt sure if it was something I did, or just that the driver wasnt capable
libvde: udovdh: no xorg sdk installed?
udovdh: yes, taht fixed it
udovdh: I overlooked, it was even in the wiki...
libvde: :)
udovdh: new box, same error.... :-)
libvde: i hadn't scrolled fully down yet :)
udovdh: (or rather: old box, different contents; same board and gfx card)
ndim: udovdh: It is eaven in the README.
ndim: s/eaven/even/
udovdh: haha oh well, I know now
udovdh: and I won't build new boxes for the next near future
ndim: Yes. Just RTFM next time. :)
udovdh: I installed a software development selection from the fedora dvd
udovdh: but not enough development then
the-me: is the new EXA acceleration available on my X1200 pro mobile gpu?
udovdh: ototh: audacity buult without wxGTK problems
udovdh: is that r5xx?
udovdh: only r5xx is supported for exa etc afaik
libvde: the-me: without compositing, exa is rather slow :)
libvde: "rather"
the-me: hmm, I've made a little modification at rhd_driver.c to hardcode it to use 97x95 dpi: http://nopaste.linux-dev.org/?378
the-me: but I've got still 98x98 active here.
the-me: hmm and in the xorg log stands: (==) RADEONHD(0): DPI set to (96, 96)
the-me: what could be wrong there?
ndim: the-me: Possibly "32d1b4d9: If no EDID is available, assume 96dpi on the first (preferred) mode"
the-me: ndim, no, cant find anything like this
the-me: and how ever it says its setted to 96x06 but it's using 98x98
ndim: the-me: http://gitweb.freedesktop.org/?p=xorg/driver/xf86-video-radeonhd;a=commit;h=32d1b4d99afdda3bc402a7ada0a833d650da4433
ndim: That's my guess.
ndim: is building new snapshot RPMs.
the-me: hmm no, it isn't in the log, too
ndim: the-me: Which log?
the-me: the xorg log
ndim: Why should it be?
the-me: because as the code says it will be logged if this is the case
ndim: Oh. Indeed.
ndim: Sorry. I need to properly wake up :)
tulcod: hey there
tulcod: reading that radeonhd is going to support 3d and 2d acceleration within some term, it seems like investing in ati cards is safe/
tulcod: right?
xAFFE: yeah, I'm using the radeonhd for all my work and I cant seem to miss a bit. 3d would be nice, but if it takes a year thats ok for me
tulcod: xAFFE: well yeah, I know it should work at all, but I kinda need the 3d support and I'm really curious about the open source decisions amd took, cause I really want open source drivers
tulcod: but I wasn't sure whether radeonhd would actually get anywhere soon, but it seems like it will now
wirechief: would like to know if there is a graphic benchmarking program that one could use to see how the new driver performs ?
retrry: gtkperf
wirechief: ok
retrry: phoronix did bencharks with it in their article about radeonhd performance :)
udovdh: hmmm
udovdh: gtkperf:
udovdh: 6.56s at 1GHz
udovdh: 5.81s at 2.1GHz
Zkyez: just came to say thank you. radeonhd accelerated works perfect with firefox while scrolling pages with pictures (although i had to disable smooth scrolling option in ff).
Zkyez: so thanks again libv
Zkyez: & co
Temujin_: hey Zkyez, where is the smooth scrolling option, is it a key in about:config?
Zkyez: yep
Zkyez: general.smoothscroll
Temujin_: cool, i shall try it, thanks
Zkyez: npc
johnficca: any 2d for Radeon mobility x1400?
pechanoSAURUZ: actually I just joined this channel to say thank you for developing a great open source driver and keep up the good work etc.
pechanoSAURUZ: much less of a hassle to make it work, compared to the fglrx driver. lack of 3d acceleration doesnt bother me too much, as long as I can get the right resolutions with the right frequencies.
xAFFE: pechanoSAURUZ: ACK, works like a charm here too :)
pechanoSAURUZ: i've had so much trouble with the fglrx driver
tulcod: lol, I'm using r300, so I can't judge :p
pechanoSAURUZ: with radeonHD its like: change "ati" to "radeonhd" and everything is perfect.
pechanoSAURUZ: really good work
tulcod: but yeah, I hope the 2d/3d acceleration is implemented soon
tulcod: so I can finally decide to get an ati card if I'm buying a new gpu
tulcod: cause I herad so much bad stories about nvidia too
tulcod: it's just no fun
Temujin_: i'd rather see the 3D acceleration done right than done quickly, we've seen what happens when you push complex code out faster than it should be (*cough Catalyst)
pechanoSAURUZ: I know for sure that I will buy nvidia for a secondary computer, because it seems work better with linuxMCE, but this radeonHD driver weighs a lot when it comes to making the decision for the main computer
pechanoSAURUZ: yes, Temujin_ I agree on that
pechanoSAURUZ: and I have a feeling it will be done right :)
tulcod: Temujin_: agreed, but it has to be done too
tulcod: Temujin_: so the current double iteration is fine with me
Temujin_: yeah
pechanoSAURUZ: looks promising yes
libvde: johnficca: x1400 is r5xx
xsdnyd: hi
z3ro: I got the atombios parser doing both data table and command table parsing in a external program... :)
z3ro: although the asic init command table gets into a loop (probably because register read/write is a no-op right now, and I'm on r350. :p
airlied: z3ro: if you integrated it into radeontool or avivotool it would be most useful :)
airlied: z3ro: then we could do suspend/resume with it like vbetool
z3ro: airlied: yeah I might do that once I rewrite it a bit more. right now it's an ugly mix of amd, radeonhd, and my own code. :p
airlied: z3ro: avivotool links to libpciaccesss..
airlied: z3ro: so you should be able to rip that code outta one of the drivers.
z3ro: I'll probably do something for (dis)assembling the bytecode (aka scripts aka command tables) too.
z3ro: airlied: right now I just ripped the atombios code and rhd_atom* from radeonhd, and wrote my own main.c on top of that.
z3ro: so it's pretty damn ugly, but it (mostly) works.
airlied: z3ro: I started porting to avivotool last week but LCA kept me distracted..