13:02 zanderover: Is there someone who could help me out with a little issue here??
13:07 zanderover: every millisecond I get the same nouveau warning in dmesg, I don know what is happening.
13:10 gremlink_: Any news on using the GTX 960 (maxwell) with nouveau?
13:12 buhman: any news on nouveau's takeover of nvidiaHQ?
13:13 gremlink_: heh
13:21 zanderover: Everything really just looks fine, only this message is causing me a headache!
13:21 gremlink_: zanderover: What's the message?
13:22 zanderover: It is saying:
13:23 zanderover: nouveau E[ PIBUS][0000:01:00.0] GPC2: 0x5133e4 0xbadf1301 (0x01022215)
13:24 gremlink_: zanderover: That's over my head. But it's working, right?
13:26 zanderover: yeah, but why is this message flooding my journal??
13:27 gremlink_: zanderover: Does putting that error into Google find anything about it?
13:32 zanderover: Yeah I've searched the internet before on this issue, never found anything useful.
13:35 zanderover: It is getting a little difficult to troubleshoot other things with this message around.
14:02 imirkin: zanderover: that's a GM206, right?
14:03 imirkin: zanderover: do you see the message in http://cgit.freedesktop.org/nouveau/linux-2.6/log/?h=linux-4.1
14:08 zanderover: no i believe i've a gk104, thanks for the link!
14:08 imirkin: that's a kepler...
14:09 zanderover: you are correct
14:09 imirkin: lspci -d 10de: should tell you
14:11 zanderover: a little different output: GK110
14:11 imirkin: hah! seriously?
14:12 imirkin: can you paste the vga line of 'lspci -d 10de: -nn' ?
14:12 imirkin: (there will also be an audio subfunction, don't care about that one)
14:12 zanderover: 01:00.0 VGA compatible controller: NVIDIA Corporation GK110 [GeForce GTX 780] (rev a1)
14:13 imirkin: you didn't add the '-nn'
14:13 zanderover: oh wait a sec.
14:14 zanderover: 01:00.0 VGA compatible controller [0300]: NVIDIA Corporation GK110 [GeForce GTX 780] [10de:1004] (rev a1)
14:14 imirkin: hah, and it was marketed as a GTX 960?
14:16 zanderover: No I suppose you're mistaking _Gremlink's Card?
14:16 imirkin: oh, hah, yes i am.
14:17 imirkin: skeggsb_ might have some ideas about that message... iirc there was some compute thing? but newer mesa doesn't even try to instantiate the compute class...
14:18 imirkin: zanderover: at what rate is the flood happening? like 100/s or 1/hr?
14:19 zanderover: It is going crazy here!! 50 a sec I guess.
14:20 imirkin: hm, ok. what kernel?
14:20 imirkin: and what mesa?
14:25 zanderover: 3.17.4-301.fc.21 and 10.4.7-1.20150323.fc21
14:26 imirkin: hm ok
14:26 imirkin: i was hoping i could blame it on some ancient debian build
14:27 zanderover: same problem in 3.19.3-200.fc21
14:28 imirkin: yeah i wouldn't expect things newer than 3.17 to really help
14:30 zanderover: It started indeed with 21.. i have to admit I don't really remember my settings in fc20. I know there was no flood for sure.
14:31 imirkin: gk110 probably wasn't supported with the fedora 20 kernel
14:32 imirkin: and/or mesa
14:32 imirkin: i guess i dunno
14:32 imirkin: iirc gk110 became supported in 3.12 or so
14:33 imirkin: looks like last june is when it happened... mesa 10.2, and whatever kernel happened around then
14:35 zanderover: I got this installed:
14:35 zanderover: mesa-libGLU.x86_64 @koji-override-0/$releasever 9.0.0-7.fc21
14:35 zanderover: rest is all 10.4.7-1
14:37 tobijk: dont worry about libglu that is stuck at version 9
14:58 imirkin: zanderover: there _is_ a small change in that linux-4.1 branch related to gk110 bringup... could make a diff, but unlikely
15:10 zanderover: Could be worth a try.
15:11 imirkin: specifically this one: http://cgit.freedesktop.org/nouveau/linux-2.6/commit/?h=linux-4.1&id=e1fc44fb9dbec4ff4e63c888d206a0bc4ca93f4f
15:11 imirkin: although if things are *basically* working, it's unlikely to change anything
15:12 damo22: is this the dev channel?
15:13 imirkin: i suppose
15:14 damo22: i am wondering if anyone has had trouble making nouveau run on certain hardware, like everything seems good with the driver but it refuses to work
15:14 damo22: i think it is related to PAVP
15:14 imirkin: for a question _that_ generic, it'd be a complete lie if anyone claimed otherwise
15:15 imirkin: there is definitely hardware out there on which nouveau doesn't work :)
15:15 damo22: let me try to be more specific
15:15 damo22: is there a list of hardware that nouveau is supposed to work on, but for some reason it just wont initialise or something
15:16 imirkin: i am unaware of such a list. you can look through the list of bugs at bugs.freedesktop.org
15:16 imirkin: under xorg -> Driver/nouveau
15:16 damo22: thanks
15:16 imirkin: (even though it has nothing to do with xorg, but it's an old bugtracker)
15:19 damo22: thing is, there is this thing called the Intel Management Engine, and it has a PCI device that authenticates with the binary driver
15:21 damo22: im not sure if it is relevant to nvidia cards or not
15:21 mjg59: It's not
15:22 damo22: mjg59: okay thanks, by the way, do you know what it relates to?
15:23 mjg59: damo22: THe Intel Management Engine?
15:23 damo22: no, i mean this authentication thing with PAVP
15:23 damo22: and some GPUs
15:24 specing: the management engine is just a hardware backdoor, it should not prevent nouveau from running
15:24 mjg59: I don't really understand what you're asking
15:28 damo22: im talking about this: http://i.imgur.com/B2c4smX.png
15:29 damo22: it implements hardware DRM
15:31 mjg59: I understand what you're talking about, but not what you're asking
15:33 damo22: can this feature be omitted and still have video working?
15:33 mjg59: Yes
15:34 damo22: even if the content it is trying to play is encrypted and requested to be played via the PAVP?
15:35 mjg59: If the player you're using asks for PAVP and doesn't get it, it's up to the player what happens next
15:35 damo22: ok
15:36 damo22: so we're all going to refuse to use PAVP?
15:36 imirkin: damo22: it sounds like you're having a very concrete problem, but you're asking very generic and high-level questions
15:36 imirkin: damo22: you may be better off asking the concrete question
15:36 mjg59: damo22: No open drivers implement any form of PAVP, to the best of my knowledge
15:37 damo22: i hate the management engine because it is going to be the end of x86 freedom
15:37 damo22: that is my problem
15:37 imirkin: ok, and you're asking about nouveau because...
15:38 damo22: because i like to use video
15:38 imirkin: so... use it
15:38 damo22: and i dont want others to be restricted in the future
15:39 imirkin: what does any of this have to do with encryption or hdcp or any of that?
15:39 imirkin: just don't use it
15:40 damo22: okay yes, i can decide not to use it myself
15:40 damo22: but if many people continue buying x86 with restrictions built into the hardware, we're all doomed
15:41 artag: damo22: don't worry. If x86 gets too restricted to be useful to developers, it will die
15:41 artag: and good riddance
15:42 imirkin: ... and i want laptop screens that are designed for work rather than movies. but guess what -- majority doesn't care about the things you (or i) care about.
15:42 damo22: artag: ok, lets build some opensparcs
15:42 artag: exactly
15:43 artag: fpgas are one solution but there are other commercial architectures too
15:43 damo22: sorry to bother the channel, i just read a book on management engine and i got a shock
15:43 artag: x86 is an ancient bodge. It has the advantage of performance as a result of how much investment it has received, but it's not an architecture anyoone would choose
15:44 artag: when it dies, it will be a good thing
15:44 imirkin: artag: and yet... everyone chooses it.
15:44 artag: of course. it's cheap and fast and the world's software runs on it
15:45 artag: everyone bouhgt IBM once. Not now. The world moves on.
15:45 damo22: but we're always playing catch up, and my argument is that its gotten to the point of being too hard to reverse engineer
15:45 damo22: and now near impossible to maintain computing freedom
15:46 damo22: but no one actually cares
15:47 artag: they do care, but not enough to give up their cheap computers
15:47 damo22: yes
15:47 artag: when the restrictions get worse and the performance of the alternatives gets better, they'll change
15:48 artag: if you want to be ahead of the game, invest time in opensparc or one of the other open architectures
15:48 damo22: oh god, i dont want to be surrounded by people with worse restrictions
15:48 damo22: its bad enough
15:49 artag: but don't bother kicking against the vested interests directly. find an alternative, like RMS did
15:49 damo22: what did he find?
15:49 artag: he started developing free software
15:51 artag: he didn't pressure AT&T to give away Unix, he just did it himself. and many others followed.
15:52 imirkin: artag: unfortunately hardware and entertainment content is more difficult to create by yourself
15:52 artag: and now commercial unixes are all but dead, and the many free ones proliferate
15:52 damo22: artag: that is why i try to port more old hardware to libreboot
15:53 artag: hardware is not difficult. it's getting easier, and things like kickstarter provide the means to make large investments if needed
15:53 artag: I don't care about entertainment. Good entertainment to me is a live band, not a commercial blockbuster movie
15:53 damo22: hehehe me too
15:53 artag: if someone else cares about them, maybe they can solve that
15:55 artag: 20 years ago it would be difficult to contemplate something as complex as Linux being built by enthusiasts, but not now. It happened.
15:56 damo22: ok time to go, thanks artag
17:13 storrgie: I just got a Lenovo W541, which has an nvidia chip within it
17:13 storrgie: I'm trying to boot Fedora 22, but I'm getting nouveau debug messages at boot time
17:15 imirkin: you're going to have to provide actual info if you want actual help
17:18 storrgie: My first pass was to see if anyone has that particular machine and has tried it
17:18 storrgie: I'm going to try F21 really quick to see if its something with the beta, I doubt it, but I'll report back once I peek at it
17:19 airlied: storrgie: a copy of the boot msgs would be best place to start
17:20 imirkin: i think you underestimate the number of diff laptop models and overestimate how much we keep track of the wide variety of hw out there
17:21 storrgie: Oh.. I know the herculean efforts of this project
17:21 imirkin: for all i know ben has one. and/or lenovo is sending him one that has the GK106 that refuses to come up... perhaps it's even that one :)
17:21 storrgie: I'm thrown off that its defaulting to the nvidia adapter, this has an intel chip also
17:21 storrgie: imirkin, I'm desperately hoping not. Thats why I cant use my M4800 :(
17:21 airlied: I think we've been promised one from Lenovo
17:22 storrgie: i changed something with the bios and now it just blank screens on boot
17:22 storrgie: im quite upset right now
17:22 storrgie: like, the screen doesn't have a backlight
17:23 storrgie: I changed the bios to turn off secure boot and turn UEFI off (in favor of legacy)
17:23 storrgie: booted to the F22 disk, and now I have no backlight after I had to power it off with the button
17:25 storrgie: whats the best way to get you guys data, I could take a picture of the screen and show you
17:26 storrgie: I feel like booting (and failing) has put the graphics adapter into a state that it's not working properly on boot now
17:39 storrgie: I guess... I'm just flored
17:40 storrgie: these awful... awful nvidia chips are ruining everything
17:43 storrgie: airlied, how can I exfiltrate boot messages if I can't get it to boot?
17:44 airlied: storrgie: that was a response to you saying to were seeing debug messages at boot time
17:44 airlied: not I changed a bunch of BIOS options and now it won't boot
17:44 airlied: does it get to the bootloader?
17:47 storrgie: I'm getting back into that now
17:48 storrgie: I had a problem where backlight wasn't turning on
17:48 storrgie: I got that 'resolved' by loading up the boot menu (f12) which turned it back on.
17:48 storrgie: Literally, no backlight when you should see the lenovo logo
17:49 storrgie: I'm able to boot to F20 or F21
17:49 storrgie: but it won't make it through the boot
17:49 airlied: add nomodeset to the kernel command line, it should at lesat boot
17:50 storrgie: with UEFI can I do that?
17:50 storrgie: it doesn't look like I can pass args
17:51 storrgie: oh nvm, I can edit the settings
17:55 storrgie: airlied, now that I'm in, what should I snatch?
17:55 storrgie: the laptop is getting really, really hot
17:56 storrgie: GK106GLM :(
17:57 airlied: yeah same Gk106 then
17:57 storrgie: so, no options right?
18:01 storrgie: realistically, if Lenovo sent one to you guys... would it even be possible to address?
18:01 storrgie: This has been a prevalent issue on the M4800, mine has been in a box since I got it
18:03 storrgie: I have been talking to an executive rep at lenovo, due to the method that we've purchased these (we bought... many), so I think I could assist on this front. My other option is to demand that they replace these with T540/T550 which are purely intel
18:05 airlied: I suppose the nvidia driver might work
18:12 storrgie: not sure if i can even get it installed
18:12 storrgie: what I fundamentally don't understand is: why can't we just disable the nvidia adapter completely? It seems like this generation forces you to use it.
18:15 storrgie: airlied, is the binary even an option with fedora moving to wayland? I thought nvidia was giving tghe finger to waylan
18:35 airlied: storrgie: should work for now, wayland isn't that much used yet
18:35 airlied: still need a driver for the chip to turn it off
18:52 storrgie: I feel like swan diving on a pile of legos
18:53 storrgie: I've only used my M4800 to kill spiders over the entire time i've owned it
18:53 storrgie: if there is a nouveau dev that would like to figure out this garbage chip using an m4800 I would be interested in being contacted
18:53 storrgie: I'll ship it to them at my own cost
19:33 airlied: hopefully will have a w541 in the same place as the main dev next week