00:42mlankhorst: can I set a watchdog timer on the tegra?
00:47mupuf: mlankhorst: for shader execution time?
00:49mlankhorst: no, it hard locks
01:22mlankhorst: and now with serial attached it fails to hard lock :/
01:26mlankhorst: oh didn't have the tv on so no output was reported for hdmi-1
01:57 < ahellquist|work> pmoreau: imirkin http://fpaste.org/199367/ zbook15 G2 optimus enabled, boot with console log
02:46pmoreau: ahellquist|work: Thanks
02:47pmoreau: So, it is still experiencing the HUB_TIMEOUT error, implying the NVidia card wasn't suspended.
02:47pmoreau: ahellquist|work: Would it be possible to have the full logs, including Nouveau initialisation of the card?
02:49 < ahellquist|work> pmoreau: anything is possible :)
02:50pmoreau: Cool :D
02:50 < ahellquist|work> pmoreau: what do I need change ?
02:52pmoreau: Possibly add debug to the command line (it helped me to get more messages iirc), and disable nouveau at boot to load it manually later on
02:53 < ahellquist|work> pmoreau: I'll give it a try
02:53pmoreau: The debug thing shouldn't be necessary, except if you need Nouveau debug level messages
02:54pmoreau: But you don't seem to display them anyway
02:55 < ahellquist|work> pmoreau: I do not know whats needed but I guess the more the merrier
03:07 < ahellquist|work> pmoreau: HUB_Timeout was showing up in the previous paste, yes
03:08 < ahellquist|work> pmoreau: tested booting with nouveau.modeset=0 and console and can load/unload nouveau but loading nouveau does not show any info in dmesg
03:09 < ahellquist|work> pmoreau: do I need some fancy dri.*=too:?
03:12pmoreau: ahellquist|work: You shouldn't need that
03:13pmoreau: But you probably need to load nouveau with `modprobe nouveau modeset=1` to override the initial modeset=0
03:13 < ahellquist|work> pmoreau: missed the debug, testing again.
03:18mlankhorst: imirkin: hm with uboot it's trying to do a tftp boot without any aid..
03:22dardevelin: skeggsb, I prefer to have completely free software, I don't demand it. Just want to make my informed decisions. Hence the questions.
03:22dardevelin: skeggsb, btw sorry for coming back to you so late.
03:23mlankhorst: Using RTL8169#0 device
03:23mlankhorst: TFTP from server 192.168.178.250; our IP address is 192.168.178.118
03:23mlankhorst: Filename 'C0A8B276.img'.
03:23mlankhorst: Load address: 0x80408000
03:23mlankhorst: which is my dhcp server
03:24mlankhorst: if that fails it tries to obtain pxelinux.cfg/default-arm-tegra124 eventually
03:25 < ahellquist|work> pmoreau: new paste, hopefully all debug info is there.. http://fpaste.org/199391/
03:26pmoreau: ahellquist|work: It seems to, thanks! :)
03:26huehner: imirkin: http://lists.freedesktop.org/archives/nouveau/2015-February/020144.html looks like my ddx patch for gm206 was overlooked
03:26 < ahellquist|work> pmoreau: wc
03:28pmoreau: Oh, it's missing the part when the NVidia card is supposed to power down
03:28pmoreau: Which should be a bit later after you did the dmesg command
03:29pmoreau: But at least we know that the Optimus DSM is detected, and vga_switcheroo is enabled
03:31 < ahellquist|work> pmoreau: talking to me ?
03:31pmoreau: ahellquist|work: Yep
03:31 < ahellquist|work> pmoreau: I am still connected to that box so i can grab mor log info
03:33pmoreau: Did you get the HUB_TIMEOUT later after getting the dmesg or not?
03:36 < ahellquist|work> pmoreau: Yes, here is the tail of dmesg including HUB_TIMEOUT http://fpaste.org/199395/
03:40pmoreau: ahellquist|work: So the card doesn't go to sleep for some reason... :/
03:43pmoreau: ahellquist|work: I'll try to have a look at the code tonight.
03:44 < ahellquist|work> pmoreau: sounds great, let me know if you need anything. I have no hurry and can assist with feedback and logs
03:47pmoreau: ahellquist|work: Ok, will do if needed :)
04:11abinader: hi folks :-) I wonder what is the current status for GM107 (GeForce GTX 750) device? proprietary nvidia 346.47 does not recognize my card (even though it says it should support), and it seems I need to generate a MMIO tracing to get it working on nouveau
04:12abinader: however, it seems like a "chicken vs. egg" issue - I can't get the proprietary "nvidia" kernel module to load, thus I can't get the MMIO tracing needed for nouveau (if I understood it correctly)
05:44abinader: ok, just a quick update: I managed to get it working by disabling PCI host bridge window info from ACPI
09:27Omar007: So I tried experimenting again with Nouveau on my 970 and I'm now getting some interesting output
09:27Omar007: Basically it renders the screen twice on 1 display, side by side
09:28Omar007: all squished vertically and without cleaning the previously drawn CLI output
09:28imirkin: abinader: there's also early support for GM107 ctxsw fw in nouveau in ben's branch
09:29imirkin: Omar007: 4K displayport display?
09:29Omar007: imirking: nope 1920x1200
09:29imirkin: hm, that's much less likely to be DP MST
09:29Omar007: imirkin: I do have 3 of them plugged in atm
09:29Omar007: All on their own DP though so no chaining
09:30imirkin: abinader: http://cgit.freedesktop.org/~darktama/nouveau
09:31Omar007: Oh and it has a line accross the top like so: | | | | | | | | | | | | | | | | | | etc
09:31Omar007: in I think 4 repeating colors
09:35Omar007: imirkin: http://imgur.com/iDYkmba
09:35Omar007: Bad quality image though
09:35Omar007: but should help with what I'm trying to explain xD
09:38imirkin: Omar007: what kernel are you on?
09:39imirkin: Omar007: there were a couple of fixes to the dcb logic recently... dunno if they'd account for your issues though
09:40imirkin: Omar007: you might give http://cgit.freedesktop.org/nouveau/linux-2.6/log/?h=linux-4.0 a shot
09:40Omar007: I also tried to build the master branch from git using the following AUR package: https://aur.archlinux.org/packages/xf86-video-nouveau-git/
09:40imirkin: xf86-video-nouveau has nothing to do with it
09:40Omar007: 4.0 :O
09:58Omar007: Connection dropped :V
09:58Omar007: Seems like compiling the kernel is going to be quite an adventure :O
10:01imirkin: if you're not used to compiling kernels, don't worry about it
10:01imirkin: it's not like the GM20x support in nouveau is particularly useful -- only kms
10:03Omar007: imirkin, if it can run my desktop normally I'm fine with that.
10:04Omar007: I can always boot the standard kernel again if I really need to right?
10:04imirkin: Omar007: depends what your desktop is. if it's a bunch of opengl stuff, that'll end up running on your cpu
10:04Omar007: I don't think my CPU would mind much haha
10:04Omar007: got 6 cores @ 3.5GHz so a desktop should be fine
10:05Omar007: Heavy 3D is ofcourse a different story
10:05Omar007: But then I can always fall back to the normal kernel + nvidia or something
10:06imirkin: ok. well, i'll bbl
10:19Omar007: Ok I first reset my system back to a working state and now I'm building the kernel. And so we wait....
10:48Omar007: Yay atleast it reaches the CLI with this kernel :D
10:48imirkin_: and it's all displaying properly?
10:48Omar007: GUI I need to work on first
10:48imirkin_: huehner: sounds like your patch fixes everything. nicely found :)
10:48imirkin_: Omar007: just use the modesetting driver. xf86-video-nouveau won't load
10:48Omar007: Yea I know
10:49Omar007: I figured that one out while I was waiting for it to compile :P
10:50Omar007: damn ok I celebrated to early
10:50Omar007: I appareantly had forgotten to un-blacklist nouveau
10:51imirkin_: and it fails in the same way when you load it?
10:51Omar007: But instead of throwing all kinds of weird output, it now freezes ones it tries to go into KMS
10:51Omar007: *once >.>
10:51imirkin_: can you ssh in and get dmesg?
10:51Omar007: Yea I can SSH
10:51imirkin_: (is it actually frozen, or do you just see a black screen?)
10:52Omar007: Was about to go and check the journal
10:52Omar007: No the screen still shows what was on the screen before it went to KMS
10:52imirkin_: systemd throws all sorts of bs in there... just pastebin dmesg output
10:52imirkin_: hm =/
10:53Omar007: so basically it's like: blablabla->KMS->Screen drawing stops
10:53imirkin_: yeah, evo channel probably hangs
10:53Omar007: not even the cursor is blinking
10:53huehner: imirkin_: maybe you can get the ddx part committed?
10:53imirkin_: huehner: did you see my reply?
10:53huehner: probably not ;)
10:53imirkin_: (from a few hours ago)
10:54imirkin_: long story short, i dunno if we should push that out
10:54huehner: imirkin_: i agree makes sense
10:54huehner: imirkin_: i think i saw note that you were working on maxwell exa in general, missing vbo conversion part
10:55Omar007: imirkin_: http://goo.gl/DIKf81
10:55imirkin_: huehner: yeah. i dunno if "working on" is the best word to use... perhaps "spent a few hours about 6 months ago" might be better
10:55imirkin_: which i never pushed to any repo it would seem
10:56imirkin_: but basically i just need to figure out a way to implement the scratch buffer
10:56imirkin_: since there's only one exa operation at a time
10:56imirkin_: i should just make it a GART-mapped buffer and update it in place + draw
10:56imirkin_: huehner: but that still won't help you, since it'll rely on acceleration working :)
10:56imirkin_: 2D isn't magically different from 3D... it's all the same thing
10:57imirkin_: Omar007: that says 3.18.6
10:57imirkin_: Omar007: i would have expected 4.0-rc4 or something along those lines.
10:58Omar007: Yea I noticed that too
10:58imirkin_: i.e. you booted the wrong kernel
10:58huehner: imirkin_: clearer now, thx, i'm still happy with working console as fallback without black screen (which is what i get from the 2 kernel patches) :)
10:59Omar007: Maybe I messed up my boot config lol
10:59imirkin_: huehner: i'm personally betting that accel on GM20x will be worked out in about 6-12 months in upstream. we'll see, i guess.
11:00Omar007: nouveau one also says 3.18
11:00huehner: imirkin_: and even then reclocking with will next missing thing most likely, assuming boot-clock is as low as it was i.e. on fermi
11:01imirkin_: huehner: actually if we end up taking a path that involves using nvidia's firmware, reclocking will get a boost, since their firmware will presumably implement it all already
11:01imirkin_: huehner: and they will have pmu firmware monitoring counters and changing levels or whatnot
11:02imirkin_: huehner: but this is all theory. we'll see how it shakes out.
11:03Omar007: I build the wrong branch
11:05Omar007: sooo..... how do you remove a kernel? Just delete the files from /boot and /lib?
11:05imirkin_: edit /boot/grub/menu.lst?
11:05imirkin_: i dunno how all this "make install" jazz works... i've never used it, and assume it does hoorid things
11:06imirkin_: [for the kernel, that is]
11:06Omar007: I think I just need to remove the old entries from /lib/modules/
11:06Omar007: and the boot files + boot entry
11:06imirkin_: really just the boot entry
11:07Omar007: Well yea but that leaves all the files
11:07imirkin_: no harm in leaving old files around
11:07imirkin_: esp when you mess things up bigtime, and they're there to save you
11:07imirkin_:is a bit of a pack-rat though
11:07Omar007: I still have the core kernel
11:07Omar007: so I'm good on that part
11:07imirkin_: yeah, until you accidentally "clean it up"
11:08Omar007: In that case I can only blame myself for being a turd and deleting the wrong files :P
11:09imirkin_: and determining blame is a lot more important than getting a working system...
11:24minniebot: I have a question about the jetson tk1, is this an appropriate channel?
11:25imirkin_: perhaps, depending on the question
11:25imirkin_: i question your question!
11:27minniebot: I'm trying to install Fedora on my dev board
11:27minniebot: And Ive been following this https://fedoraarm.wordpress.com/2014/08/14/installing-fedora-21-pre-release-on-the-nvidia-jetson-k1/
11:27imirkin_: if it's a non-nouveau-related question you may have better luck in #tegra
11:27minniebot: But I'm confused as to why I should flash U-boot on it, if it's already on there...
11:27imirkin_: afaik u-boot is not there by default
11:28minniebot: Is #tegra usually empty? :|
11:28imirkin_: doesn't _seem_ empty
11:28imirkin_: but note that it's evening in europe and not morning yet in japan
11:28minniebot: hmm ok ill try there, thanks
11:29imirkin_: you have to use that crazy tegrarcm thing with dubious quantities of arguments
11:29imirkin_: but you can eventually get u-boot on there if you work hard and believe in yourself
11:30imirkin_: or get one of the #tegra guys make a "here is a script that will do exactly what you want" script. i got lucky :)
11:31imirkin_: i think the basic deal is that if you want to boot upstream linux, you have to use u-boot. but u-boot doesn't support e.g. hdmi output which makes it a bit annoying.
11:31imirkin_: i think mlankhorst went through this recently, perhaps he has some comments
11:35Omar007: imirkin_, ok I think it worked but it did not change res. dmesg: http://goo.gl/BwpTWh
11:36imirkin_: Omar007: nouveau is not loaded
11:36mlankhorst: I just followed the instructions..
11:37mlankhorst: got a working root with nouveau installed that way
11:38Omar007: imirkin_, Ah yea I see. Module could not be found... :S
11:42Omar007: imirkin_, Is there something in make nconfig I shoud've set differently?
11:42imirkin_: Omar007: presumably CONFIG_DRM_NOUVEAU=m or something :)
11:42imirkin_: Omar007: perhaps you forgot to "make modules_install"? dunno
11:43Omar007: No I did do that one
11:43Omar007: I did not use the CONFIG_DRM_NOUVEAU=m I think though.. But I didn't the first time either...
11:44mlankhorst: finally moved my rootfs to sata
11:44Omar007: I think that option should be somewhere in the nconfig as well then though..
11:44imirkin_: Omar007: yeah, "nouveau" under drivers -> drm
11:45mlankhorst: an arm quad core with 2 gb memory and a ssd, should be more than capable to rebuild itself. :P
11:46Omar007: drivers -> gpu -> drm -> drm -> nouveau?
11:46imirkin_: sure, why not
11:46Omar007: ok its on M now :P
11:52Omar007: Booting again....
11:52Omar007: black screen :D
11:52Omar007:goes checking dmesg
11:53Omar007: YO! Errors galore :D
11:53Omar007: I'll post the output up again ;)
11:53imirkin_: file a bug + attach your vbios
11:54imirkin_: you can get vbios from /sys/kernel/debug/dri/0/vbios.rom (assuming you built debugfs)
11:56Omar007: my dmesg is missing output I think :(
11:56Omar007: Yea... it is
11:56Omar007: Journalctl does have all the entries
11:57Omar007: >20000 lines
11:57Omar007: all MMIO read errors
12:22Omar007: imirkin_, how is this? https://bugs.freedesktop.org/show_bug.cgi?id=89664
12:23imirkin_: you forgot to include dmesg
12:24Omar007: 0.o I put that one in when I created it... Wonder where it went...
12:24Omar007: Oh well I'll just add it again
12:25imirkin_: the ui is confusing iirc... easy to THINK you added a file when you really didn't
12:25Omar007: Oh I see why
12:25Omar007: It's over the limit
12:25Omar007: It's a whopping 5MB of errors
12:26imirkin_: Omar007: if a boatload of errors are the same
12:26imirkin_: you can remove them and just put like [...] in their place
12:26Omar007: problem is the target address for each seems to be different
12:26imirkin_: can you paste 2?
12:27imirkin_: actually, just attach the first 15K lines
12:27imirkin_: that should be enough.
12:27imirkin_: and fit within the limits
12:27Omar007: I can strip the first 1000 boot lines as well I guess
12:27imirkin_: no, leave those.
12:28imirkin_: just grab the first 15K lines of your dmesg
12:28imirkin_: or boot log or whatever
12:28Omar007: Yea total is almost 50k :D
12:30Omar007: imirkin_, done
12:30Omar007: There does seem to be a repeating pattern..
12:31Omar007: But at the same time, it also changes at times
12:31imirkin_: wait, you have 2 of them??
12:32Omar007: Second one is doing nothing atm though
12:32Omar007: Except for sitting there
12:32imirkin_: i can tell you what it's not doing -- spewing messages to dmesg :p
12:33Omar007: Probably because there is nothing in there
12:33imirkin_: this is very odd
12:33Omar007: The 01 is the one that has the monitors in it
12:33Omar007: 02 is doing nothing
12:33imirkin_: neither card says that it executed the vbios
12:34imirkin_: but generally the option rom would have only been run on one of them
12:34imirkin_: Omar007: can you boot with nouveau.config=NvForcePost=1
12:34Omar007: as boot parameter?
12:35Omar007: but I first need to go away for dinner
12:35Omar007: Back in a bit
13:58mlankhorst: I wonder whether I should make a ppa for the tegra stuff
13:58mlankhorst: in case there are other lazy people like me
13:58imirkin_: seems far-fetched... people as lazy as you?
14:00mlankhorst: well once it works i want to nuke l4t and try a real ubuntu install
14:14Omar007: imirkin_ NvForcePost seems to actually work mostly
14:14imirkin_: Omar007: ok cool. so that means we're misdetecting one of the cards as init'd but it's really not
14:14Omar007: It went into KMS; second screen activated and clones the first
14:15imirkin_: that's expected
14:15Omar007: Third screen still idle
14:15imirkin_: that's less expected... but i dunno how fbcon works
14:15imirkin_: what happens if you run X?
14:17Omar007: I think I saw something about HEAD3 pass by in the Nouveau output
14:17Omar007: Oh I may want to disable GDM...
14:18Omar007: Server terminated with error 1. Going to check the logs now
14:23Omar007: imirkin_; so far I can see it fails to find the nv, and nouveau drivers, as I'd expect, then it loads the KMS driver. Later on it starts setting up the screen and finds the DRM driver
14:23Omar007: So so far so good
14:23imirkin_: it should be using the modesetting driver
14:24Omar007: That's what I said right? nv and nouveau drivers fail so it goes to the KMS one
14:24Omar007: It also seems to properly see each screen
14:24imirkin_: just confirming it...
14:25Omar007: imirkin_, ok when it reads the EDID things get weird. Lots of WW about the fact that the HorizSync range would exclude the HorizSync range for the given resolutions
14:26imirkin_: pastebin xorg log
14:34Omar007: imirkin_, http://goo.gl/Mbsccq
14:35imirkin_: Omar007: it's using the nvidia driver
14:36imirkin_: Omar007: and it's a 3.18.6 boot without NvForcePost=1
14:37Omar007: idk why it says that because it certainly is not :/
14:37Omar007: I can do it again
14:37imirkin_: it clearly is.
14:37imirkin_: and you did something wrong. perhaps it's the wrong Xorg log file.
14:38Omar007: Then it did not create one :(
14:38imirkin_: or you're using arch
14:38Omar007: That was the latest
14:38imirkin_: and systemd
14:38imirkin_: and it's start X as you and the log goes somewhere in your homedir
14:39imirkin_: well, all i can say is that that's not a log from a nouveau setup.
14:42Omar007: ok I think I've found the correct file
14:42Omar007: Sorry it takes so long but I've only been on Linux for a month now or so xD
14:44imirkin_: np. you've wandered into a maze of annoyingness with systemd.
14:46Omar007: imirkin_, http://goo.gl/muu18y
14:47Omar007: that one seems to be the correct one then
14:47imirkin_: [ 29.842] (II) NVIDIA GLX Module 346.47 Thu Feb 19 18:09:07 PST 2015
14:47imirkin_: you need to switch back to mesa libraries
14:48imirkin_: i wonder why it's trying to load glamor... hopefully it'll have an exception for swrast drivers
14:49Omar007: ugh stupid! Since I didnt install the xorg-video-nouveau I also forgot to install mesa D:
14:50Omar007: ok I now have 2 bright white screens
14:51imirkin_: Omar007: pastebin xorg log :)
14:51Omar007: It's still doing stuff
14:51Omar007: I'm waiting for that to finish first
14:51imirkin_: i bet it's trying to use glamor. ugh.
14:51Omar007: Ok there we go. One last flash and we're back at the CLI
14:51Omar007: I'll grab the log now
14:51imirkin_: airlied: will modesetting end up using glamor even though there's only swrast?
14:53Omar007: imirkin_, http://goo.gl/BaZjuJ
14:53imirkin_: oh good. glamor init failed. that's positive.
14:55mlankhorst: weird, now nouveau is trying to open a tegra device :/
14:55imirkin_: Omar007: well that all looks perfectly happy.... what was the problem?
14:56Omar007: It just crashed back to the CLI
14:56Omar007: I have 1 message in the CLI
14:56Omar007: link training failed
14:56imirkin_: mlankhorst: mind glancing at that log?
14:57imirkin_: Omar007: that's probably why the third monitor doesn't turn on
14:57Omar007: Well since I ran startx, the second also turned off
14:58imirkin_: in either case, link trianing is in relation to displayport link training
14:58Omar007: After the whiteness, just before crashing bac kto the CLI
14:59Omar007: I'll give the gdm a go in the meantime to see what/if anything happens :P
14:59imirkin_: what were you running before?
15:00imirkin_: ah... you probably don't have stuff configured for that to work
15:01Omar007: This stuff is making me so confused right now lol
15:01Omar007: GDM does nothing. I'll grab the log and see what it contains
15:01imirkin_: define 'nothing'
15:02Omar007: Black screen with a single _ in the upper left
15:02Omar007: Xorg logs didn't change
15:03imirkin_: so it switched VT's but X didn't actually start
15:03imirkin_: or it switched to the wrong vt
15:03Omar007: VT = 1
15:04imirkin_: unlikely... it normally siwtches to vt 7
15:04imirkin_: but if you've messed around with stuff, sometimes it'll bt vt 8
15:04Omar007: It's always 1 afaik. Atleast on Arch
15:04imirkin_: where does the regular vt go?
15:04imirkin_: anyways... wtvr
15:05Omar007: it runs gdm through systemctl though
15:05Omar007: Maybe that changes things?
15:05Omar007: wtvr is a command you want me to run?
15:05imirkin_: running X creates a new tty
15:05imirkin_: er, new vt
15:06Omar007: oh then I'm not sure about VT. I think I confused it with tty then
15:07imirkin_: VT = virtual terminal
15:07imirkin_: tty implies a line discipline (ldisc)
15:07Omar007: TTY is the thing you can usually switch between with Alt+Ctrl+F* or is that VT?
15:07imirkin_: that stuff is crazy-confusing btw
15:07imirkin_: and dates back to how computers worked in 1960
15:08imirkin_: that's a vt
15:08imirkin_: which has a tty attached to it, usually
15:08imirkin_: but not necessarily :)
15:08Omar007: Well afaik on Arch that always runs under F1
15:08imirkin_: that's very surprising
15:08imirkin_: X usually switches to 7
15:09Omar007: "About the tty, yes that is normal, and is due to switching to Systemd, only starts one getty by default."
15:09imirkin_: but then X should be on 2...
15:09Omar007: From the arch forum: https://bbs.archlinux.org/viewtopic.php?id=169903
15:09imirkin_: or does it reuse the same vt?
15:09Omar007: Idk man D:
15:10Omar007: I think so
15:10Omar007: Because I can't switch to CLI either
15:10Omar007: no wait
15:11Omar007: further in the topic it is X behaviour
15:11imirkin_: anyways... wtvr
15:11imirkin_: can you pastebin the log from when you ran it with gdm?
15:11imirkin_: or perhaps your system is in a weird state
15:12imirkin_: and you should just reboot and try running gdm again
15:12Omar007: Still with the flag I assume?
15:12Omar007: Or try again without?
15:16pmoreau: I'd like some help to fix this issue: http://pastebin.com/YVD4ZmSh
15:17imirkin_: Omar007: what flag?
15:17Omar007: haha ok I rebooted and now the 2nd one is black and the 3rd monitor is blinking (cloning monitor 1)
15:17imirkin_: you still want ForcePost
15:17Omar007: They just all came online
15:17imirkin_: until we figure out why we think your cards are init'd when they're not
15:18Omar007: going to log into the CLI now then I'll try GDM again
15:18pmoreau: It could be due to the channel switching, or to the base channel - I tried doing some stuff on the overlay channel and it also failed, so not sure..
15:18Omar007: All screens are now white
15:19Omar007: (my eyes! D:)
15:19pmoreau: And it's probaly not PFIFO nor PFB related.
15:19imirkin_: Omar007: that's with gdm right?
15:19Omar007: so it does the same as startx at first
15:19Omar007: but now it is stuck on a black screen with a _ in the top left
15:19Omar007: I'll go grab the logs
15:21Omar007: imirkin_: http://goo.gl/UGNCLZ
15:23imirkin_: Omar007: dunno, that all seems fine. maybe gdm requires acceleration and that's somehow messed up?
15:23Omar007: But shouldn't startx atleast work then?
15:23Omar007: Or does that also require acceleration?
15:23imirkin_: well, depends what it starts :)
15:23imirkin_: let's try something simpler
15:23imirkin_: what happens when you just run 'xinit'
15:24imirkin_: ideally you should see a black screen and an xterm
15:24Omar007: Well I grabbed xorg-xinit specially for this ;)
15:24Omar007: Ok i'll try
15:24Omar007: let me reboot first
15:24imirkin_: (no window manager or anything of that sort, so you can't move/resize the window... at least not easily)
15:27Omar007: It's like a roullete; you never know which monitor it will turn on; the 2nd or the 3rd or both xD
15:28Omar007: xinit didn't even make the screens white aww :(
15:28Omar007: Grabbing logs
15:29Omar007: Cannot open /dev/tty0 (No such file or directory)
15:29imirkin_: ugh. something dumb about systemd/arch/something
15:30imirkin_: i haven't kept up with their stupidities... sorry, can't really help more
15:30imirkin_: as far as i can tell, X is working perfectly fine
15:30Omar007: I'll look into running xinit with Arch
15:30Omar007: Probably just need to supply a paramete or something
15:32Omar007: trying xinit -- :1 -nolisten tcp vt$XDG_VTNR
15:33Omar007: And that does the same as startx
15:33imirkin_: do you not see a terminal?
15:33Omar007: it just drops back to the CLI
15:33imirkin_: well i dunno why it exits
15:33imirkin_: oh, what happens if you do...
15:33imirkin_: xinit xterm -- the-stuff-you-had
15:34Omar007: Gotta reboot first now D:
15:34Omar007: this killed my second monitor and keyboard or something
15:35Omar007: or alteast the output of letters to the screen
15:37Omar007: Aww tis time it gave me only 1 out of 3 screens :'(
15:46Omar007: I haz terminal! :D
15:46Omar007: So that works
15:48Omar007: Also, it seems every time I reboot when nothing x-related crashed, it enables an extra screen lol
15:48Omar007: But I feel like that is just coincidence :P
15:49Omar007: imirkin_; but that still leaves the question; why doesn't the rest work? D:
15:57imirkin_: Omar007: soon you'll have like 10 screens
16:01Omar007: FYI, on exit it threw me "EGL_MESA_drm_image required"
16:01Omar007: xinit: connection to server lost
16:02Omar007: and link training failed is now flowing into the CLI
16:26Omar007: imirkin_: weston works almost fine it would seem. Only thing I'm missing is my mouse cursor. It is there just not rendered
16:27imirkin_: something wrong with the cursor overlay i guess? any errors in dmesg about EVO?
16:30Omar007: imirkin_: oh wait when I zoom in a bit the mouse shows fine xD
16:30Omar007: Though the top bar gets quite a bit of artifacts when I move around
16:31Omar007: well since weston works... I could give gnome 3.15.9 a go I guess and see if that changes anything,..
16:36Omar007: Ok it has 1 difference; black screens without a _ in the upper left! xD