01:00 imirkin: skeggsb: got this for a vdpau client: https://hastebin.com/qetobanada.go
01:01 imirkin: looks like there's an imbalance in falcon_fini/init with certain error paths
17:00 mirko182: hi. is there a solution for low brightness/backlight after suspend?
17:39 nyef`: mirko182: Adjust the brightness post-suspend?
17:40 nyef`: (That'd be more a workaround than a fix, though.)
17:41 mirko182: brightness controls dont work after suspend
17:44 mirko182: even tried to manualy change values i /sys/class/backlight (both acpi_video0 and nv_backlight) but no go
17:47 karolherbst: mirko182: kernel version?
17:48 mirko182: 4.9.39
17:49 karolherbst: I guess there is nothing in dmesg either?
17:50 karolherbst: mirko182: also if you read the values back, are those the same as the ones you wrote?
17:52 mirko182: nothing in dmesg and values are same
17:56 mirko182: it's this bug https://bugs.freedesktop.org/show_bug.cgi?id=39550
17:57 karolherbst: mhh, we had some patches regarding backlight stuff, let me check
17:58 karolherbst: dmesg would be still interesting, maybe something else is odd
18:03 karolherbst: mirko182: was there a kernel version where it actually worked? Or could you check if a newer fixes the problem? Allthough I highly doubt the latter, but maybe it does. Getting your vbios will also help
18:07 mirko182: tried with fedora's 4.11, debian 3.16 same thing. here's dmesg right after my last suspendhttps://pastebin.com/QNM6pfT7 .
18:07 mirko182: bah, https://pastebin.com/QNM6pfT7
18:13 karolherbst: I think the best would be if you create a new bug report, attached with your dmesg and vbios file. And if you have a little time to spare, doing an mmiotrace running nvidia would be perfect
18:39 mirko182: sry was afk. this is my old notebook with quadro FX770M, maybe newer cards donth have similar problems, but dont have machine with never nvidia chip to check. With prop. drivers (340x) brithness after suspend worked ok but machine was hot as hell.. And would lock after suspend on newer versions of kernel (3.18+).
18:41 mirko182: Thnx for help
19:00 RSpliet: karolherbst: why a new bugreport? what's wrong with the old one?
19:02 RSpliet: Doesn't seem like the value for the PWM divider is set in nouveau two years after I proposed that mock-up patch, and I have neither tools nor time to figure this stuff out myself
19:02 RSpliet: Least mirko182 could've done was SSH, do some nvapeek/poke to verify that solution is still valid and chime in so that someone with manpower and money for the right HW can solve this issue properly
19:03 RSpliet: (and he should've been told to try exactly that obviously, not blaiming mirko)
21:46 karolherbst: wuhu, d3d11 games on wine are working
21:48 karolherbst: performance with nouveau ain't that bad either
22:53 karolherbst: imirkin_: if dnz is set on a mul, what would be the result of mul(a, -inf)?
22:56 karolherbst: fun
22:57 karolherbst: ex2(+inf) = +inf, but ex2(-inf) = +0
22:58 karolherbst: wait
22:58 karolherbst: uhh crap
22:58 karolherbst: pow(0, 0) == 1
23:03 karolherbst: currently wondering what happens when we indeed get pow(a, b), where a and b are 0