02:39 hakzsam: imirkin, okay, makes more sense
06:20 pmoreau: karolherbst: Still stable on the pmoreau branch; I get a score of 68 in Unigine Heaven… --"
06:20 karolherbst: uhh nice
06:20 karolherbst: pmoreau: but the clock is higher now, right?
06:21 pmoreau: It is
06:21 pmoreau: Oh, you mean with boost?
06:21 pmoreau: Haven't tried boost
06:21 karolherbst: the boost stuff isn't on your branch
06:21 pmoreau: Ah, k
06:21 karolherbst: besides, it only lowers clocks
06:22 pmoreau: Well, I still have the 925MHz core clock, like I did yesterday IIRC
06:22 karolherbst: ohh
06:22 karolherbst: sure you switched the branch?
06:22 pmoreau: Let me check
06:22 karolherbst: dmesg | grep nouveau | grep MHz
06:23 karolherbst: if it shows boost/base clocks, you are still on the kepler branch
06:23 karolherbst: we should move some stuff out of init into oneinit ...
06:23 pmoreau: The grep returns nothing
06:23 karolherbst: resuming the gpu prints some stuff we don't need
06:23 karolherbst: pmoreau: mhh then let me check
06:24 karolherbst: pmoreau: cat pstate
06:25 karolherbst: the boost table says your base/boost clock is 925MHz, which is weird, because they are the same :D
06:25 karolherbst: but that has no effect whatsoever
06:25 pmoreau: 07: core 270-405 MHz memory 838 MHz AC DC * 0a: core 270-925 MHz memory 1560 MHz
06:25 pmoreau: 0e: core 270-925 MHz memory 4000 MHz 0f: core 270-925 MHz memory 5016 MHz
06:25 pmoreau: DC: core 405 MHz memory 837 MHz
06:25 karolherbst: why are the pstates limited...
06:26 karolherbst: ohh i know
06:26 karolherbst: stupid boost tables
06:26 karolherbst: this vbios is just a mess
06:26 karolherbst: your gpu has not GPUBoost :p
06:26 karolherbst: it behaves completly like a tesla with cstates :D
06:27 pmoreau: lol
06:27 karolherbst: even the normal fermi has more boost stuff
06:27 karolherbst: they really ruined this gpu
06:28 karolherbst: pmoreau: can you run gputest furmark and see how hot your gpu gets?
06:28 pmoreau: (BTW, do you know which identifier I need to use for the Intel driver in the xorg.conf? I'd like to set it to use DRI3)
06:28 pmoreau: I'll check that
06:29 karolherbst: pmoreau: use mine xorg.conf https://gist.github.com/karolherbst/1f1bdd1a3822df74097f
06:29 karolherbst: this way you can unload nouveau while X is running
06:29 karolherbst: you only need the dummy driver
06:30 pmoreau: Thanks
06:30 karolherbst: though it might that you also need a X patch
06:31 karolherbst: ohh no, it was merged in 1.18
06:31 pmoreau: I'm on 1.18.x
06:31 pmoreau: Should be good then
06:31 karolherbst: yeah
06:31 karolherbst: with 1.17 loading/unloading nouveau hanged my X server
06:31 karolherbst: if nouveau was added lateR? no idea
06:31 karolherbst: but there was an issue :)
06:35 pmoreau: 85C it seems
06:37 karolherbst: mhh
06:37 karolherbst: k, this is hotter than mine
06:38 karolherbst: so we shouldn't play around witht he clocks then :D
06:38 karolherbst: as I said, like a tesla :D
06:39 pmoreau: But high is 95C and crit is 105C, so still a bit of room there
06:39 karolherbst: pmoreau: yeah, but you only do rendering now
06:39 karolherbst: imagine you also do video decoding and encoding
06:40 karolherbst: pmoreau: anyway, there is one point left which bothers me a bit. the vbios declares 5.0 PCIe speed on the higher pstates
06:40 karolherbst: it would be interessting to know what nvidia does
06:40 karolherbst: and if nvidia changes the width
06:41 karolherbst: LnkSta is the important line in lspci
06:41 karolherbst: if nvidia changes the width, then it would be nice if you would do a mmiotrace of that
06:42 karolherbst: just so that we have at least one for kepler
06:42 karolherbst: and if it changes the link to 8.0 we might read it out wrongly out of the vbios
06:44 karolherbst: pmoreau: I am curious why you get only 68 points though
06:44 karolherbst: with mine I get like 340
06:45 karolherbst: pmoreau: which resolution?
06:45 Tom^: does he still have that .drirc crap that got merged a few months ago
06:45 karolherbst: I doubt it
06:45 Tom^: because well, once the .drirc has it added in ~ it wont be removed unless you do it :p
06:45 karolherbst: most likly he runs it at an insane resolution
06:46 karolherbst: Tom^: why should that be?
06:46 karolherbst: usually you don't have any .drirc
06:46 karolherbst: only the system-wide one
06:46 Tom^: no idea why i had one
06:47 Tom^: perhaps driconf or something like that creates one
06:47 karolherbst: yeah
06:48 karolherbst: pmoreau: your gpu is just too slow... 750M was it?
06:49 karolherbst: mhh
06:49 karolherbst: it should be clocked at 967MHz boost I think
06:50 karolherbst: yeah well, most likely the apple guys just remove some high perf stuff, so that the gpu stays a bit cooler
06:57 pmoreau: karolherbst: At 2880x1800
06:57 karolherbst: pmoreau: ohh k
06:57 karolherbst: then the perf difference is to be expected
06:58 pmoreau: Finally on the Intel chipset! Thanks for your Xorg.conf
06:58 karolherbst: no problems :)
07:00 pmoreau: Yeah! And with DRI3, so I can switch between both!
07:00 pmoreau: Well, not really switch, but well
07:00 karolherbst: yeah I know :p
07:00 karolherbst: but
07:00 karolherbst: and this is a big bug
07:00 karolherbst: 5.0 pcie is a bit slow..
07:00 karolherbst: pmoreau: try fullscreen glxgears or something
07:00 karolherbst: you will be disappointed
07:01 karolherbst: prime offloaded I mean
07:02 pmoreau: 2866FPS -> 58FPS
07:02 karolherbst: pmoreau: I get around 520 fps with 8.0 at fullhd
07:04 pmoreau: And vulkaninfo correctly founds the Intel chipset!
07:05 pmoreau: Time to run some VUlkan demo then
07:06 karolherbst: pmoreau: ohh seems like 5.0 doesn't make much of a difference here
07:06 karolherbst: 344/540-550/550-590 on 2.5/5.0/8.0
07:26 karolherbst: pmoreau: ohh and it would be also good to know what nvidia clocks to at full load
07:29 pmoreau: Ok, I put it on my todo list
07:34 karolherbst: thanks a lot
08:18 karolherbst: pmoreau: do you have anything against if I put tested-by tags ont he patches?
08:18 pmoreau: Not at all, please do
08:19 karolherbst: and I really don't like journald if you need to debug errors...
08:19 karolherbst: the only way to get stuff is pstate
08:19 karolherbst: ..
08:19 karolherbst: pstore
09:05 imirkin: pmoreau: karolherbst: do either of you have a minute to trace a couple of dEQP tests?
09:05 karolherbst: I would have to setup dEQP first
09:07 imirkin: oh, i thought you had that done already.
09:08 karolherbst: do I have to do something special or just grab the deqp sources and build it?
09:10 karolherbst: imirkin: which DEQP_TARGET should I set?
09:10 karolherbst: x11_glx or x11_egl_glx?
09:17 imirkin: karolherbst: either
09:17 imirkin: but not both :)
09:17 imirkin: it'll fail with some idiot libpng thing
09:18 imirkin: so you need this patch: http://hastebin.com/ehetupikix.avrasm
09:18 imirkin: also, if you don't have xorg 1.18.1 you need another patch, but i suspect you're on the latest and greatest
09:19 karolherbst: I am on 1.18.1
09:19 imirkin: ok you're good on that front then
09:19 karolherbst: there is some really annoying freeze with 1.18.2 and chromium
09:20 karolherbst: yeah, already stumbled across the libpng thing
09:20 karolherbst: this is super annoying, they could just use the system one
09:20 imirkin: i sent a patch
09:20 imirkin: they've allegedly integrated it
09:20 imirkin: which means it should be publicly visible in the next 5 years or so.
09:20 karolherbst: ohh
09:21 karolherbst: where is the patch? :D
09:21 imirkin: see above :p
09:21 karolherbst: ahhhh
09:21 karolherbst: right
09:21 karolherbst: my fault
09:22 imirkin: you're grabbing from https://android.googlesource.com/platform/external/deqp/ right?
09:22 karolherbst: yeah
09:26 imirkin: ok that's right then
09:27 pmoreau: imirkin: Not right now sorry.
10:08 karolherbst: imirkin: I get this error: "print "Both --name and --id must be specified""
10:08 karolherbst: ohhh
10:08 karolherbst: python2....
10:10 mupuf: karolherbst: hey, could you give me the output dmesg when running your kernel?
10:10 karolherbst: mupuf: all?
10:10 mupuf: just the nouveau part
10:10 mupuf:is reviewing your patch series
10:11 mupuf: sunday is patch review day, isn't it?
10:11 karolherbst: mupuf: https://gist.github.com/karolherbst/50834676785ad7206f70
10:11 karolherbst: ahh nice
10:12 karolherbst: mupuf: pmoreau only needs 5 patches of those: https://github.com/karolherbst/nouveau/commits/pmoreau
10:12 karolherbst: mupuf: and I was thinking we should at least get those merged for the next release, because they are trivial and already help somebody
10:12 mupuf: yeah, not a big fan of how you print the boost and base clocks
10:12 mupuf: well, talk to ben :D
10:13 karolherbst: mupuf: his vbios doesn't contain the coefficients stuff, and info.min == info.max, so there is nothing we can do wrong
10:13 mupuf: My work is done right? I already reviewed these patches
10:13 mupuf: but seriously, don't try to hurry it
10:13 karolherbst: only some of them
10:13 karolherbst: these are all now
10:13 mupuf: pmoreau is a grown up
10:13 mupuf: he can recompile his kernel :p
10:13 karolherbst: :D
10:22 Yoshimo: so karol, if the 4.6 merge window is over, will there be any patches on your github left that aren't upstream yet?
10:34 karolherbst: Yoshimo: many
10:34 karolherbst: like most of them
11:15 jayhost: karolherbst you still want me to keep gm107 on high clocks?
11:16 karolherbst: jayhost: if nothing happened until now, then it is all fine
11:16 karolherbst: you can stop now :D
11:27 jayhost: Alright.
12:30 karolherbst: mupuf: should I have one debugfs file for each clock? base_mhz, boost_mhz? And should I put in there the real or the doubled clock (or is it real and halfed?)? I would got for each clock one file, because everybody else seems to do that
12:35 mupuf: hmm, you are confusing me
12:35 karolherbst: which part is confusing?
12:37 mupuf: sorry, back
12:38 mupuf: why would you want to have as many files as clock?
12:38 mupuf: I would propose pstate to actually state the boost level
12:38 mupuf: like
12:39 mupuf: boost level: 0 (1123 MHz), *1 (1175 MHz), 2 (1200 MHz)
12:40 karolherbst: mhhh
12:40 mupuf: and create a debugfs file to change the boost level you want also
12:40 mupuf: or ... just create this boost level file
12:40 mupuf: and make it like pstate
12:40 karolherbst: well there is no difference between selecting a cstate
12:40 mupuf: makes sense?
12:41 karolherbst: boosting on nvidia gpus is only a thing, because we have a base clock entry inside the vbios, without it, there would be no such thing as "boosting", because in the end it is just reclocking depending on the load
12:42 karolherbst: (+ some volt/temperature stuff, but that doesn't make much of a difference)
12:42 karolherbst: in the end at any given time we have a list of cstates we can reclock to
12:43 mupuf: yeah
12:43 karolherbst: and this list may have more or less entries depending on the current temperature
12:43 mupuf: so, what about: max_cstate ?
12:43 karolherbst: well I could also just revice my cstate debugfs file
12:43 karolherbst: and we just push out a list with all cstates
12:43 mupuf: not really no, please
12:43 mupuf: I just want to expose your NvBoost parameter
12:44 karolherbst: k
12:44 karolherbst: what we could do though is this
12:44 mupuf: does that make sense to you?
12:44 karolherbst: debugfs file name "boost" or "gpu_boost"
12:44 karolherbst: 0: base MHz
12:44 karolherbst: 1*: boost MHz
12:44 karolherbst: 2: max MHz
12:44 karolherbst: and we just put the value in there
12:46 karolherbst: on mine it would be 0: 705 MHz 1: 797 MHz 2: 862 MHz
12:46 karolherbst: where the max one might not be even reached, because of the voltage calculations
12:47 karolherbst: ohhh
12:47 karolherbst: wait
12:47 karolherbst: I think this was what you meant :D
12:55 mupuf: * 1: boost MHz
12:55 mupuf: yes
12:55 mupuf: this is what I meant by : max_cstate
12:55 mupuf: not a guarantee, just a limiter :)
12:56 karolherbst: yeah I think it would make actually sense to have
12:56 karolherbst: *this
12:56 mupuf: yep :)
12:57 mupuf: The fact that you wanted to expose it in the kernel logs pretty much said it already
12:58 karolherbst: well there are cards where this file would be of noe use though :D
13:02 mupuf: no worries
13:02 mupuf: at least, no need to reboot to test the boost clocks, that's a win!
13:03 karolherbst: I think we need a field in nvkm_clock to remember the cstate we _want_ to have, because the one we actually have might be lower, and in our volt adjustmen code we could then clock up if possible
13:03 mupuf: oh, right
13:04 pmoreau: God it was complicated to install Linux on the computer: first manage to connect a screen to one of the GPU, then the ISO had old version of the GPG keys and didn't wanted to install the packages, once I had the new version, internet didn't wanted to work… --"
13:04 karolherbst: but currently there are all kind of states saved there
13:04 karolherbst: and I really want to clean up that mess
13:04 karolherbst: astate, tstate, dstate?
13:04 karolherbst: ...
13:04 karolherbst: and then ustate_ac and ustate_dc
13:05 karolherbst: pmoreau: you should try to install gentoo for once :D
13:05 karolherbst: or better: LFS
13:06 pmoreau: :-D
13:06 pmoreau: It's more a "it was complicated" because everything was working against me, not the task per se was complicated.
13:06 karolherbst: but the gentoo bootstrap isos are crap and garbage
13:06 karolherbst: they don't have any efi stuff on it
13:07 karolherbst: uhh
13:07 karolherbst: pmoreau: never forget, I installed gentoo once on my mbp, from scratch
13:07 karolherbst: it was not a retina one, but basically the same issues
13:07 pmoreau: I'm considering installing gentoo one day
13:07 karolherbst: bios emulation: only radeon card, configureing efi crap, moving to intel as main ;) much fun
13:08 pmoreau: :-)
13:08 karolherbst: and this was before vga_switcheroo supported the gmux
13:09 karolherbst: pmoreau: I suggest you to use a live CD that you trust for this :D
13:09 pmoreau: Trying to install Fedora/Debian/Ubuntu in nconfig on my old mbp was kinda fun as well
13:09 pmoreau: s/nconfig/ncurvers
13:09 pmoreau: ncurves
13:09 karolherbst: at some point I shall try out these non APT based distributions... never did that actually
13:13 mupuf: pmoreau: ncurses?
13:13 mupuf: funny. arch decided to do away with any kind of GUI to install
13:14 mupuf: to be fair, it is not like it is needed
13:15 pmoreau: mupuf: Yes, ncurses, thanks
13:17 Tom^: mupuf: it was just a blue menu driven ncurses menu that started fdisk and nano for you. it was a bit pointless imo :P
13:17 mupuf: Tom^: agreed
13:46 kugel: karolherbst: pong
13:46 kugel: sorry, very busy currently
13:47 karolherbst: kugel: ahh, k, just wanted to know if you still have any reclocking issues with nouveau
18:01 karolherbst: mupuf: still there?
21:50 t4nk270: someone else is having too much lag even with the mouse using gtx960 with wayland?
21:51 t4nk270: i think there isn't hardware accel.
22:05 imirkin: t4nk270: that's right - there isn't. GM20x requires signed firmware in order to have acceleration, and nvidia has only very recently released proper information about it
22:06 imirkin: t4nk270: 4.6-rc1 will have some acceleration support
22:08 t4nk270: great news
22:10 t4nk270: thank you :)