friederike: i am seeing http://bugs.freedesktop.org/show_bug.cgi?id=13474 on my toshiba a100-773. libv commented it is fixed except for macbook but it is not fixed for me. :-(
carldani: z3ro: ping
mjt: radeonhd finds TWO monitors connected to a DVI-D output of X1200, one with native resolution (1024x1280) and another with smaller (800x1024), and tries to init them in a endless loop... But the thing is that there's only one monitor connected. Is there anything I can do with this?
mjt: (it's a built-in video)
mjt: it works with d-sub output, but I don't have the cable (borrowed one from a friend of mine just to check things, and returned it back)
mjt: the driver finds 4 output connectors (that's right) -- VGA, TV-Out, DVI-D and HDMI
libv: mjt: hrm
mjt: when i try to start X, the screen flickers, mostly black, with flashes every few seconds - when switching resolutions I think.
libv: mjt: has this been mailed to the ml yet? can you include a log?
libv: mjt: can you then, also run with Option "NoRandR" "True"
mjt: i'm X-less right now (for obvious reason ;), so no email for me (unless I'll be so bored as to install some text-based MUA :). I'm trying to run SOMETHING, -- just replaced a mobo that broke, and nothing works so far, not even vesa and vga drivers.
libv: ok
libv: save the current log
libv: somewhere with a useful so that you can mail it in later :)
mjt: i guess it's better to turn on some debugging ;)
libv: useful name even
libv: right... -logverbose 7
mjt: lemme try with norandr
libv: Xorg -logverbose 7 ought to do it for a good log
libv: try norandr later
mjt: (by the way, `Option "NoRandR"' is enough - default is true)
libv: because we do want to have all the necessary information to provide a proper fix :)
mjt: sure
libv: ah, ok, will save some typing in future :)
libv: does norandr fix this in some way?
da1l6: hi
da1l6: Short question: Is there powerplay support for radeon mobility 2600 cards avalible or planned?
libv: da1l6: not yet available, but it should be high on our list
da1l6: thanks :)
libv: da1l6: we are still in startup mode, in the new year, one of our first jobs is to prioritise what we need to do for our driver in this quarter
da1l6: are the docs for powerplay avalible from AMD?
libv: but the chance is that this will be this quarter, depending on what ATI can provide us with
libv: da1l6: not really, but we should still dig into the information we have available already in respect to this topic
libv: could be that the only reliable option is to call atombios (as it might do board specific things), and that that just magically works in most cases
da1l6: okay, thanks libv for your answers and thanks to all radeonhd devs for your good work so far!
mjt: is back
mjt: whooh. I finally managed to start the damn X, but it's near impossible to see what's in there. Several columns on the screen are much wider than should be, like the screen has been stretched at several vertical lines, so that some chars are duplicated here and there.
mjt: NoRandR doesn't fix the issue
mjt: currently i'm running over dsub cable (got one from a friend again).
mjt: no it's impossible to work at this screen... Switching back to text mode...
mjt: http://www.corpit.ru/mjt/sd.png -- a screenshot of the screen, without NoRandR, DVI output, when booted with monitor attached to D-SUB port. I mean, boot with dsub, when unplug dsub and connect dvi, and run X.
mjt: it recognized 2 monitors, 1280x1024 and 1280x800.
mjt: hence the wrong placement of the task bar etc.
mjt: http://www.corpit.ru/mjt/X-RandR.log - X log with RandR
mjt: http://www.corpit.ru/mjt/X-NoRandR.log - X log without RandR
mjt: should I (try to) mail it to the mailing list?
libv: no, let me read the logs first :)
libv: mjt: is this an embedded board?
mjt: yes
mjt: err
libv: mjt: because of the pci subsystem ids, they are not filled in correctly :)
mjt: it's an onboard video card
libv: but who is the manufacturer of this motherboard?
libv: biostar?
libv: hrm
mjt: the mobo is a micro-ATX Biostar TA690G board
libv: this one has messed up pciids on a unichrome board once, so that wouldn't surprise me
libv: right
libv: mjt: crap vendor :p
mjt: purchased yesterday
libv: mjt: return it, it is broken ;p
mjt: well... may be. It looked pretty good with their solid capacitors - a rare thing for this type of mobos.
mjt: i tried geforce-6510-based ones before, from asus, and descided to try an amd-based this time.
mjt: 6150 even
libv: i'm just kidding here
mjt: i understand
libv: there is some problem in how we handle the connector table
libv: ATIs atombios, this version of it, provides DFP1, DFP2, DFP3
mjt: and yes, i don't like it, -- the first thing (which is very important for me) i come across is that the fan control ("smart fan" function) does not work on this mobo - neither built-in nor software-based.
libv: we deduce that DFP1 is TMDSA
libv: DFP2 is LVTMA
libv: and we just continue with DFP3 as LVTMA
mjt: heh. That's chineese writing for me ;)
libv: (TMDSA and LVTMA are two outputs that can be used as TMDS outputs, or DVI)
libv: so your connector table tells us that DFP2 and DFP3 are being used
libv: and our atombios code then thinks that LVTMA is used twice
libv: which, of course, cannot be true.
mjt: by the way, i wasn't able to run fglrx on it, too
libv: mjt: which connectors are attached at the moment?
libv: mjt: hdmi or dvi?
mjt: only d-sub right now
mjt: the logs are from dvi only
mjt: the screenshot i did was when I booted with dsub and turned on dvi before starting X
mjt: s/turned on/switched to/
libv: hrm
mjt: (with dsub it works)
libv: hpd messed up?
libv: mjt: can you run the rhd_conntest utility?
libv: mjt: is there an adapter connected to the dvi connector?
mjt: currently only d-sub is connected
mjt: building conntestg
mjt: -g
libv: mjt: hrm, that's weird
mjt: the logs are taken with DVI connected
mjt: only DVI
mjt: ghrm.. My X build environment is broken at the moment
mjt: because I've a mix of various stable/testing/unstable libs, made when I tried different drivers.
libv: rhd_conntest only needs libpci and libz
mjt: and imake
mjt: ok got it
libv: hrm, heh, the autoconf stuff probably makes it impossible to build all on its own
mjt: Checking connectors on 0x791E, 0x1002, 0x791E (@01:05:00):
mjt: Load Detection: RHD_OUTPUT_DACA
mjt: HotPlug: RHD_HPD_NONE
mjt: DDC: RHD_DDC_0 RHD_DDC_2
mjt: LVDS Info:
mjt: 18bits, single link, LDI Panel found.
mjt: Power Timing: 0x000, 0x000, 0x00, 0x00, 0x000
mjt: Macro: 0x00000000, Clock Pattern: 0x0001
mjt: this is dvi
mjt: err
mjt: this is dsub
mjt: i have to reboot to get dvi info
libv: no
libv: mjt: you can just connect the dvi cable
libv: rhd_conntest pokes the hardware itself
mjt: the thing is that the situation is different whenever i boot with dvi connected or switch at runtime
libv: mjt: it is there to help us debug issues like the one we are seeing with this device
libv: mjt: hrm, really?
mjt: i've only seen real screen when booted with dsub and connecting dvi afterwards
libv: mjt: even with our driver?
libv: hrm.
mjt: the screenshot - it's done this way
mjt: if i boot with dvi, i can't run x
libv: this shouldn't happen, the connector table should not be altered imho
libv: hrm
libv: anyway, you do not need to run X
libv: all you currently need to do is find out which output is actually tied to DVI
mjt: it's DVI-1 on the logs
mjt: connected dvi and dsub
libv: sure, but we need to know what it is in reality
libv: therefor, we need to run conntest and find out
mjt: Checking connectors on 0x791E, 0x1002, 0x791E (@01:05:00):
mjt: Load Detection: RHD_OUTPUT_DACA
mjt: HotPlug: RHD_HPD_0
mjt: DDC: RHD_DDC_0 RHD_DDC_1 RHD_DDC_2
mjt: LVDS Info:
mjt: 18bits, single link, LDI Panel found.
mjt: Power Timing: 0x000, 0x000, 0x00, 0x00, 0x000
mjt: Macro: 0x00000000, Clock Pattern: 0x0001
libv: mjt: even with the dvi connected?
mjt: here we go.
mjt: the last one is dsub+dvi
mjt: first was with dsub only
mjt: RHD_DDC_1 added
libv: ok, so hpd0, ddc1 and lvtma
mjt: i guess RHD_DDC_2 is the wrong one
libv: do you have the ability to test the hdmi?
mjt: no
mjt: lemme try dvi only...
libv: no point :)
libv: mjt: can you attach the vga cable through an adapter to the dvi connector?
mjt: oops
libv: because on the pictures, it is a DVI-I and not a DVI-D
mjt: DDC: RHD_DDC_1 RHD_DDC_2
libv: mjt: DACA and HPD0 again?
mjt: and the screen.. err.. the lines become 2x higher (it's not 80x25, it looks like 80x12)
mjt: x13 ;)
mjt: Checking connectors on 0x791E, 0x1002, 0x791E (@01:05:00):
mjt: Load Detection: RHD_OUTPUT_NONE
mjt: HotPlug: RHD_HPD_0
mjt: DDC: RHD_DDC_1 RHD_DDC_2
mjt: LVDS Info:
mjt: 18bits, single link, LDI Panel found.
mjt: Power Timing: 0x000, 0x000, 0x00, 0x00, 0x000
mjt: Macro: 0x00000000, Clock Pattern: 0x0001
mjt: that's dvi only
mjt: i have to reboot - now even text screen is impossible to use (the 13 lines, that is)
mjt: ..back
mjt: by the way, this mobo is very popular here. A nice working horse for an office PC.
libv: hrm, the lines do what?
libv: when you run rhd_conntest?
mjt: no, it was when I turned off d-sub, connected dvi, run rhd_conntest and plugged d-sub back
mjt: the screen switched from 80x25 to 80x13 mode
mjt: (it turned off when i disconnected dsub ofcourse, and when i plugged it back again, i've seen 80x13)
mjt: quite.. awful, I'd say - all the chars.. ;)
libv: and it still is in 80x13?
libv: hrm
mjt: no, i rebooted
libv: ah, ok, why?
mjt: because it's painy to look at such an ugly chars ;)
mjt: i probably may try to switch the mode, but... simpler to reboot ;)
libv: so it was rhd_conntest that caused this?
mjt: well.. I can try again ;)
libv: really, don't reboot even if the current mode is messed up
libv: X should still run correctly, even with a messed up text mode
mjt: ghrm.
libv: and we need to get to the point where X does run correctly
mjt: x does not work at all ;)
mjt: but.. hmm.
mjt: now it works
libv: right, but you're logged in remotely anyway, aren't you :)
mjt: yes
mjt: heh
mjt: well, the irc client is running at our office, hundreds miles away
libv: ok?
mjt: the machine in question is here @home
mjt: X works now, when I booted with D-SUB only
mjt: it still thinks there are 2 monitors
mjt: but at least it works.
libv: those conntest results, were they using a DVI-VGA adapter on the dvi connector?
libv: because i am hoping to see DACB pop up there
mjt: right now I've two cables - dvi, which is mine, and d-sub, which is from my friend. Both cables are real. The monitor has 2 connectors - DVI and D-SUB. I always used DVI before (with another motherboard), hence I don't have D-DUB cable.
libv: so no
libv: no adapter
libv: hrm
libv: well, i am going to guess and create a connector table which puts DACB there and which ignores hdmi, for now
mjt: DVI cable come together with the monitor (which is TFT - it's in the logs), so I guess it's the "right" one (there are many variants of DVI-* cables/connectors)
libv: there are three
mjt: nope - there are more than 3 of them ;)
libv: DVI-I (both digital and analog), DVI-D (only digital), DVI-A (only analog)
mjt: (if counting various "adaptors" too)
libv: you might be talking about things like DMS-60, DMS-59 and whatever
mjt: I've yet to see a DVI-A thing
libv: on a card, yes
mjt: don't think such a cable ever exists.
libv: on a monitor, no
libv: there were some CRT monitors which claimed to be DVI
libv: they had DVI-A
mjt: heh
mjt: lame ;)
mjt: also, there are different bandwidth of DVI cables/connectors (forgot how it's called) - in short, how many wires are in there, like with PCI-Express lanes.
mjt: but ok
libv: right, but this is rather unimportant here :)
mjt: aieee. Now my X screen is also 2x higher
libv: duallink
mjt: yeah
libv: so at your office, a lot of people do use these motherboards?
mjt: no
mjt: we used gf6150-based ones
mjt: many of those
libv: hrm
mjt: now i tried amd/ati, for.. comparision, so to say
libv: ah, right
mjt: (neither of them really works with linux because of the drivers issues. But at least I was able to run X on geforce, even with xvideo stuff)
mjt: and nvidia (which is on line with this 690 thing) is significantly hotter than this amd one.
libv: well, this here is all because of a seriously broken connector table in the bios :)
libv: and i am rather amazed at other issues you are seeing beyond that
mjt: I already upgraded the bios (they said "fixed issues with video") - nope, didn't help ;)
mjt: for now i don't see other issues. At least for me all the issues is a single one - "problems with the video card/driver" :)
mjt: lol
mjt: now I can't STOP X ;)
mjt: it doesn't want to go to text mode - black screen and nothing more
mjt: unplugged dvi, ran conntest again, and the screen is now back normal.
libv: mjt: this is the bios that was released yesterday, right?
mjt: no, last year
mjt: lemme see...
mjt: heh
mjt: i used the one released 2007-11-22. Lemme try current (yesterday's)...
mjt: ..updated
libv: are the pci subsystem ids still broken?
mjt: where it was broken?
libv: look at your Xorg.0.log
libv: where it lists pci ids
libv: (II) PCI: 01:05:0: chip 1002,791e card 1002,791e rev 00 class 03,00,00 hdr 80
libv: first pair is ati, rs690
libv: second pair, should read biostar whatever
libv: but it reads ati rs690
mjt: it's the same
libv: ok, is the connector table listed in the log changed in any way?
mjt: not as far as i can see
libv: crap vendor ;)
mjt: ;)
libv: ok, let me dish up a patch for this
mjt: I wonder what has changed in this bios, to start with ;)
mjt: there's NO DAMN LOG!.. ;)
mjt: they fixed alot of PCI resource overlaps
mjt: well. it definitely changed.. something
mjt: lol
mjt: -(II) RADEONHD(0): Manufacturer: BNQ Model: 76bc Serial#: 10459
mjt: +(II) RADEONHD(0): Manufacturer: BNQ Model: 76bb Serial#: 10459
mjt: (that's due to analog vs digita output)
libv: ;)
libv: can you dump the bios through rhd_conntest -d
libv: and put it online as well?
libv: egbert wants a look at it up close
mjt: i switched from d-sub to dvi without reboot but now on a new bios
mjt: got that half-screen taskbar again (as on the screenshot)
mjt: the only relevant differences in the log:
mjt: -(II) RADEONHD(0): Mapped IO at 0xf7ac7000 (size 0x00010000)
mjt: +(II) RADEONHD(0): Mapped IO at 0xf7b29000 (size 0x00010000)
mjt: -(II) RADEONHD(0): D1CRTCDisable: 0 loops
mjt: +(WW) RADEONHD(0): D1CRTCDisable: Failed to Unsync CRTC 1
mjt: (the last one is ok because of removed (at wrong moment) d-sub cable)
mjt: nothing else has changed
libv: the last one is usually ok
libv: i'm not sure why the unsyncing times out there, but it often does
libv: it has nothing to do with the d-sub either
libv: CRTC != CRT
mjt: there's also a difference in clock - but that's because i experimented with the bious before, and underclocked it from 400mhz to 200
mjt: -(II) RADEONHD(0): Default Engine Clock: 200000
mjt: +(II) RADEONHD(0): Default Engine Clock: 400000
libv: this should not affect us at the moment
mjt: yes
mjt: and a funny one:
mjt: (II) RADEONHD(0): ATOM BIOS Rom:
mjt: - Filename: 690_F_EFUSE.
mjt: + Filename:
mjt: ;)
libv: crap vendor (in case that wasn't mentioned before :p)
mjt: http://www.corpit.ru/mjt/biostar-ta690g-videorom.bin
mjt: 54784 bytes
mjt: crap. the screen periodically turns off for a few seconds.
mjt: hmm. conntest -s - is it supposed to do its things for so long time?
mjt: it runs (or, rather, sleeps) for about 20 secs
libv: mjt: it scans the ddc channels too i think
mjt: funny thing
mjt: i wonder if i can just replace the video bios from another mobo... ;)
libv: don't
libv: that will only make your problems even worse
mjt: don't do that != can't do that ;)
mjt: I used to edit bioses somewhat
mjt: like, adding a bios for scsi card or sata card to the main bios code
mjt: (had a bunch of bios-less adaptec cards - linux recognizes them just fine but not the main bios, so they aren't bootable - had to find the proper bios code for the card and to add it to the main bios proper)
mjt: or just get the d-sub cable
mjt: ;)
mjt: the picture looks more stable and colorful with dvi compared to d-sub. At least on several nvidia boards. Not a great difference, but still noticeable.
libv: mjt: http://people.freedesktop.org/~libv/radeonhd_connectortable_biostar_ta690g_am2.diff
mjt: it really is dvi-d, not dvi-i. I think anyway
mjt: .. so bacically you switched DVI and HDMI connector positions
libv: mjt: the picture shows a dvi-i
mjt: manual says it's dvi-d
mjt: like, can't connect analog display to the dvi port.
mjt: where is xf86Crtc.h file?
agd5f: mjt: in the xserver
mjt: sure thing
mjt: xserver is... large :)
mjt: found it
agd5f: hw/xfree86/modes
mjt: heh. with monolitic X stuff things were easier to find
mjt: xserver-xorg-dev
ajax: use zsh. vim **/xf86Crtc.h.
mjt: i had to find which debian package contains that file
mjt: +(II) RADEONHD(0): ConnectorInfo from quirk table:(II) RADEONHD(0): Connector[0] {RHD_CONNECTOR_VGA, "VGA", RHD_DDC_0, RHD_HPD_NONE, { RHD_O
mjt: UTPUT_DACA, RHD_OUTPUT_NONE } }
mjt: someone put that newline there please ;)
mjt: nothing changed on the X front with this patch applied
mjt: --- rhd_connector.c.orig 2007-12-19 17:29:18.000000000 +0300
mjt: +++ rhd_connector.c 2008-01-04 21:28:56.000000000 +0300
mjt: @@ -205 +205 @@ RHDConnectorsInit(RHDPtr rhdPtr, struct
mjt: - "ConnectorInfo from quirk table:");
mjt: + "ConnectorInfo from quirk table:\n");
mjt: this newline please ;)
mjt: but i wonder. If it's so messed up, should 'doze not work on it too?
mjt: (i haven't tried 'doze on this mobo, just because i don't have 'doze around)
libv: nothing changed?
libv: get me a log please
mjt: nothing as far as i can see on the screen.
mjt: moment
mjt: http://www.corpit.ru/mjt/rhd/3.log
mjt: that's with NoRandR
mjt: lemme try with randr as well...
mjt: http://www.corpit.ru/mjt/rhd/4.log
mjt: -- with RandR
mjt: err
mjt: scratch that 4
mjt: moment
mjt: and 3, too
mjt: that's off d-sub
mjt: http://www.corpit.ru/mjt/rhd/biostar-patched-NoRandR.log
mjt: http://www.corpit.ru/mjt/rhd/biostar-patched-RandR.log
mjt: booted with dvi only connected, ran X with d-sub once, removed dsub, and ran x twice off dvi - with and w/o randr
mjt: with randr, screen flickers every few secs - showing a picture for a moment, next black for a few secs again
mjt: without randr, it's just black, no flicker
mjt: icewm says "Xinerama: 2 screens" still
libv: mjt: but at least the connector seems to be handled ok, right?
mjt: well, maybe in the log only... ;)
mjt: the behavior didn't change, as far as i can see
mjt: and there's no 1280x800 mode anymroe
mjt: I moved previous logs into tthat directory as well
libv: these are seperate issues
mjt: oh well.
libv: mjt: this driver version you are using... this seems to have no git version in there
libv: mjt: is this our released driver?
mjt: see my first logs
mjt: (II) RADEONHD: version 1.1.0, built from dist of git branch master, commit f213db06
mjt: after I applied the patch and recompiled it, it now thinks it's "no git source"
mjt: but nothing has changed except of rhd_id.c
libv: right, release it is
mjt: it's a debian package really - from unstable
libv: can you grab our git and patch it with the connector stuff
mjt: yes
libv: hrm, no should have no effect in this case
mjt: xf86-video-radeonhd it is, right?
libv: right
libv: but i do not think it will change much for you
mjt: ;)
AndyBurnsUK: I've read the phoronix stuff about impending r500/r600 docs, so are X1950 the meatiest cards that will (in the fullness of time) support 2D/3D/video accel?
mjt: by the way, should i change it to DVI-D still (not DVI-I?)
libv: mjt: look at the connector itself
mjt: well, it's a dvi-i connector
mjt: but the analog wires aren't solderfed
mjt: soldered
libv: ok
libv: this is rather useless :)
mjt: the driver's built
mjt: ...and it locked up hard when I tried to replace a driver file... Looks like a kernel bug.
mjt: hmm
mjt: now it works
mjt: X WORKS!!! :)
mjt: at least with NoRandr
mjt: http://www.corpit.ru/mjt/rhd/rhd-git4e488a34+biostar-norandr.log
mjt: ..and with randr, too
mjt: http://www.corpit.ru/mjt/rhd/rhd-git4e488a34+biostar-randr.log
mjt: lemme try w/o the patch, too...
mjt: ..and w/o a patch, I see that 1280x800 taskbar-on-the-middle-of-the-screen thing
mjt: aha
mjt: I booted with d-sub
mjt: i'm losing track of what's going on. I have to go do something else for a while... ;)
agd5f: libv: DFP2 is dvo/sdvo
libv: agd5f: which means: external tmds encoder?
agd5f: yeah
libv: agd5f: so, in this case, it further points to a bogus connector table
agd5f: libv: probably the vendor offers a pcie addin card with hdmi or dvi or some such
agd5f: similar to intel ADD(2) cards
libv: agd5f: there are 4 connectors in the atombios provided table, there are 4 connectors on the board
libv: hdmi is available on the board
agd5f: well then the DFP 2 is probably one of the connectors then
libv: or it could be a bogus connector table
agd5f: or does the table have 5 connectors listed?
libv: ddc is switched, hpd is not listed for dvi
libv: 20:30 < libv> agd5f: there are 4 connectors in the atombios provided table, there are 4 connectors on the board
libv: biostar has seriously messed up with this board
libv: pci subsystem ids are not appropriately set, they use a DVI-I connector while not exporting analog lines to this connector
agd5f: idiots
mjt: i wonder if it's worth the effort to contact them about all this mess
mjt: it shouldn't work with doze, should it?
mjt: (another reboot, now with dvi only, to test things again...)
agd5f: connector table looks fine to me: crt1 (vga port), tv1 (tv port), dfp2 (dvi, probably sdvo), dfp3 (hdmi)
libv: http://www.biostar.com.tw/picture/Mb/17/_TA690G-AM2-IO.jpg
libv: agd5f: plus, ddc and hpd trouble
agd5f: yeah the ddc is screwy. I bet you have to toggle a gpio bit somewhere to to switch that line for dvi vs hdmi
bahadunn: does hdmi work with radeonhd driver?
libv: bahadunn: there is no special handling for it
bahadunn: what does that mean?
libv: bahadunn: but it should generally function like dvi
bahadunn: ah ok
libv: there's a bit in there we haven't played with yet
mjt: blah. new bios does not recheck dvi/dsub when rebooting
bahadunn: amd drop those new docs yet?
mjt: i had to turn off power to restore screen
libv: bridgman thinks it might enable the sound data in the blanking
bahadunn: I guess next week we hope for if they have not dropped by now
libv: bahadunn: agd5f put them online earlier
bahadunn: oh
bahadunn: cool
mjt: and when booting with DVI attached, driver from current git with the patch applied doesn't work too
bahadunn: is that on the topic or web?
libv: bahadunn: it is modesetting and very basic chipset stuff really
agd5f: bahadunn: http://www.x.org/docs/AMD/
libv: bahadunn: there's nothing magical in there :)
bahadunn: what do these docs give you guys?
bahadunn: 2d,3d stuff?
marcheu: if you download it 10 times, the last one has the 3D information
libv: marcheu: evil!
bahadunn: hehe
bahadunn: libv: nevermind I see what you said above
libv: bahadunn: it provides us with most of what we needed for our radeonhd driver so far
bahadunn: libv: my eyes had missed it beofre
bahadunn: before
bahadunn: libv: so its just docs for other chipsets?
mjt: ok. With pure DVI, it still doesn't work. It ONLY works when booting with d-sub, and switching to DVI right before running X - last git with the biostar-fix applied.
marcheu: libv: wait until I post that as a slashdot comment
mjt: http://www.corpit.ru/mjt/rhd-git4e488a34+biostar-randr-dviboot.log
mjt: http://www.corpit.ru/mjt/rhd-git4e488a34+biostar-norandr-dviboot.log
mjt: (same files but without -dviboot are when booting with d-sub - when it actually works)
mjt: the patch definitely fixed 1280x800 issue
mjt: by the way, is it difficult to implement xvideo?
agd5f: mjt: requires 3D
mjt: so the f*** vendors do everything using 3D nowadays.... Oh well, I miss the ol'good 2D days... ;)
mjt: heh. No wonder fglrx didn't work
libv: mjt: there is an overlay, which does some colour space conversion
yangman: it's more that it's more economical to just let the 3d pipeline handle it, as opposed to having a separate 2d acceleration
libv: mjt: but we believe it is just there for running opengl apps in overlays
libv: mjt: something which might still be a requirement for the market firegl is for
mjt: i think it's worth the effort to post something to the mailinglist about all this mess
mjt: Handle 0x0001, DMI type 1, 27 bytes
mjt: System Information
mjt: Serial Number: Unknow
libv: mjt: can you provide me with a copy of your bios? rhd_conntest -d should dump this
mjt: it's in there
mjt: lemme move it to the same dir, too...
libv: ah, ok, i missed it then
mjt: ~20:48 < mjt> http://www.corpit.ru/mjt/biostar-ta690g-videorom.bin
mjt: now at http://www.corpit.ru/mjt/rhd/biostar-ta690g-videorom.bin
mjt: that's from their latest bios
libv: ah, yes, i see it now, sorry
mjt: n/p! ;)
mjt: hmm. It looks like i have an old install of winxp here.
mjt: wonder if it will ever boot after mobo change.
mjt: lemme try...
bahadunn: is mirroring of the documentation needed?
mjt: ..doze worked
mjt: with either dvi or d-sub
mjt: so.. umm... i wonder if they will ever look at all this...
mjt: (surprizingly it booted - I installed it several years ago on a VIA C3 system - now it's a dualcore AMD 64, with VERY different hardware and stuff - only the hard drive is the same)