09:14tizbac: has anyone experienced a bug with nouveau where with gnome-shell the stuff behind the cursor square would flicker?
09:15tizbac: i mean mouse cursor
12:17mlankhorst: not me, what card ?
12:26vnd: mupuf: take a look of this: https://bugs.freedesktop.org/show_bug.cgi?id=84721
12:27vnd: I've found the commit which broken the nvc0 fan speed
12:30tobijk: mh lets have a look... :)
12:33tobijk: how convenient, i already had a look ath this one and thought, that this can't possible the one as it was just to plain simple to find :o
12:36vnd: anyway I just think that this commit revealed the bug that existed before
12:36vnd: I mean - when you switch to manual mode by echo 1 > pwm1_enable then the fan speeds up
12:37vnd: but I'm not the expert :)
12:37vnd: you guys are, it would be great if you were able to fix it or at least give me some work-around
12:37tobijk: i'm not one as well, sorry :/
12:38epdv: Hello, I've got problems with a GTX860M - from dmesg: "Direct firmware load failed with error -2"
12:38vnd: epdv: do you have firmware installed?
12:38tobijk: vnd, can you just try something as a workaround?
12:39epdv: Using arch linux, installed nouveau-fw from AUR
12:39tobijk: find /* starting from fermi, fan management is always linear */ and just comment the next 3 lines
12:39vnd: tobijk: ok, give me 5 minutes
12:39tobijk: epdv: the new cards need a signed firmware from nvidia, which they havent provided until now
12:40tobijk: you are left with modesetting (with kernel 3.19rcX)
12:41epdv: tobijk: NVidia has to provide the Firmware? Any chance to extract it myself?
12:42tobijk: hm i dont think nouveau could load it, but i'm not sure
12:44tobijk: epdv: can you provide the chips exact number? i'd like to add it to our codeName page
12:44tobijk: (or just hand me a dmesg and i'll look for myself :) )
12:49vnd: tobijk: works like a charm
12:50tobijk: can you try the manual mode?
12:50vnd: some progress finally :)
12:50tobijk: just to make sure it is still broken :D
12:50epdv: tobijk: http://ix.io/g2L
12:51tobijk: epdv: oh that actually is not a maxwell chip
12:51tobijk: forget my previous statement, novueau should really load you firmware
12:52epdv: Hm ... how can I verify the firmware is installed correctly?
12:52vnd: tobijk: after changing to manual mode the fan speeds up
12:52tobijk: ke, as expected
12:53tobijk: vnd: that was the "easy" part, let mupuf handle the hard one ;-)
12:53epdv: where should fw reside?
12:53tobijk: can you please update the bugreport with the change you made?
12:54tobijk: epdv: there are different sources
12:54tobijk: normally it is in pramin
12:54tobijk: but yours is in prom
12:54tobijk: dont ask me about the difference though :/
12:54vnd: no problem, now having the work-around I'm not in a hurry :)
12:56epdv: Hmm, found "/usr/lib/firmware/nouveau/"
13:03tobijk: epdv: wait, the firmware is loaded just fine (i really should not do something related to nouveau today)
13:07epdv: tobijk: Just found this - from the comments it seems, PRAMIN etc. are loading methods: http://lists.freedesktop.org/archives/nouveau/2010-April/005583.html
13:08tobijk: its strange the pramin source is broken but youf firmware gets loaded just fine
13:09tobijk: source is just prom not pramin
13:13epdv: But there's still also an error (MMIO ...)
13:14tobijk: wait, is there a backtrace in the logs you cut away?
13:14tobijk: please get me the full dmesg
13:14epdv: "[ 1.399511] nouveau E[ PBUS][0000:01:00.0] MMIO read of 0x00000000 FAULT at 0x500c30 [ IBUS ]"
13:15tobijk: -> pci_pm_runtime_suspend(): nouveau_pmops_runtime_suspend+0x0/0xf0 [nouveau] returns -16
13:15tobijk: looks lie a part of a backtrace
13:16epdv: Full dmesg: http://ix.io/g2N
13:16tobijk: yep a backtrace
13:19tobijk: epdv: try booting with nouveau.runpm=1
13:19tobijk: this disables powermanagement and may save you from the errors
13:19tobijk: but it will consome more power
13:20epdv: Hm, what will happen then to my graphics card? Hopefully it will not overheat?
13:21tobijk: it will just stay not try to autosuspend while not in use
13:21RSpliet: epdv: don't worry, I reverse-engineered the clock tree without casualties
13:21epdv: Ah, okay ... I just wanted to try switching off MSI, as I've found it may work - or not
13:23epdv: Thanks - I'll try that switch tomorrow, as I've to reboot after setting it
13:24tobijk: RSpliet: this looks just like this one https://bugs.freedesktop.org/show_bug.cgi?id=70354
13:24tobijk: any ideas on it? :>
13:26epdv: Wow, rather old ...
13:28tobijk: there are not many developers having the knowledge to solve those bugs, and there are not that many nouveau developers :/
13:29epdv: Of course, You need to spend a lot of time, even if You're an experienced embedded systems developer - and I'm neither
13:29epdv: Good night, bye!
13:43mupuf: vnd: thanks a lot!
13:46tobijk: mupuf: i guess there is just another vbios entry to determine the right mode for nvc1
13:47mupuf: Possible ... Will need to have a close look into it!
20:20mark4: can anyone tell me why i no longer have a visible mouse pointer after switching to nouveau from nvidia drivers?
20:41mark4: not sure if anyone answered me peviously, tried to start enlightenment and my screen went black and stayed there. in windowmaker i get a desktop but no visible mouse pointer
20:41mark4: can anyone tell me why?
22:00fling: Is it getting much better about reclocking? :P