alyssa changed the topic of #panfrost to: Panfrost - FLOSS Mali Midgard & Bifrost - Logs https://freenode.irclog.whitequark.org/panfrost - <daniels> avoiding X is a huge feature
Xalius has quit [Remote host closed the connection]
raster has quit [Quit: Gettin' stinky!]
stikonas has quit [Remote host closed the connection]
kaspter has quit [Ping timeout: 260 seconds]
kaspter has joined #panfrost
Xalius_Ph has joined #panfrost
camus has joined #panfrost
kaspter has quit [Ping timeout: 260 seconds]
camus is now known as kaspter
nerdboy has joined #panfrost
camus has joined #panfrost
kaspter has quit [Ping timeout: 260 seconds]
camus is now known as kaspter
warpme_ has quit [Quit: Connection closed for inactivity]
Xalius_Ph has quit [Ping timeout: 260 seconds]
nerdboy has quit [Ping timeout: 240 seconds]
icecream95 has quit [Quit: leaving]
icecream95 has joined #panfrost
davidlt has joined #panfrost
megi has quit [Ping timeout: 265 seconds]
tlwoerner has quit [Quit: Leaving]
<tomeu> narmstrong: was thinking that I could be able to erase the first blocks of the emmc with ./update store
nerdboy has joined #panfrost
tlwoerner has joined #panfrost
davidlt has quit [Ping timeout: 260 seconds]
icecream95 has quit [Ping timeout: 258 seconds]
guillaume_g has joined #panfrost
icecream95 has joined #panfrost
NeuroScr has quit [Quit: NeuroScr]
<tomeu> narmstrong: was able to unbrick it by booting into u-boot with https://github.com/superna9999/aml-usb-load-uboot/blob/master/run-uboot.sh and erasing the first blocks in mmc from uboot
warpme_ has joined #panfrost
yann has quit [Ping timeout: 258 seconds]
davidlt has joined #panfrost
pH5 has joined #panfrost
<icecream95> tomeu: That sounds better than on chromebooks, where you have to manually change the partition flags on the recovery image, replace the install script with one that just runs flashrom or crossystem, then plug the drive in and hope it works...
<narmstrong> tomeu: nice !
<tomeu> icecream95: I think with newer models one can do some stuff from the usb port, with CCD
<narmstrong> tomeu: did you boot linux try panfrost yet ?
<tomeu> I'm copying the fs to emmc right now
<tomeu> so I'm almost there
<icecream95> tomeu: Don't you need a USB-C cable with some of the wires switched round for that?
<tomeu> icecream95: something like that, yeah
<tomeu> smaller than a servo :p
andrey-konovalov has joined #panfrost
stikonas has joined #panfrost
icecream95 has quit [Ping timeout: 265 seconds]
raster has joined #panfrost
yann has joined #panfrost
stikonas has quit [Remote host closed the connection]
<urjaman> alyssa: did someone fix the perf regression you mentioned like a bit over a week ago?
<urjaman> that is, should i update mesa :P (or should I run a bisect? :P)
karolherbst has joined #panfrost
kherbst has joined #panfrost
karolherbst has quit [Disconnected by services]
kherbst is now known as karolherbst
raster has quit [Quit: Gettin' stinky!]
megi has joined #panfrost
NeuroScr has joined #panfrost
<tomeu> alyssa: any ideas about this? http://paste.debian.net/1125922/
<tomeu> started getting it on gnome-shell after I updated this morning
NeuroScr has quit [Quit: NeuroScr]
NeuroScr has joined #panfrost
davidlt has quit [Ping timeout: 268 seconds]
NeuroScr has quit [Quit: NeuroScr]
raster has joined #panfrost
NeuroScr has joined #panfrost
Xalius has joined #panfrost
NeuroScr has quit [Quit: NeuroScr]
<tomeu> alyssa: don't waste time on that, I think my build directory got all messed up
<alyssa> urjaman: I don't think so...? I've been focused on functionality right now.
<Xalius> is there a version of or a trick how to attach apitrace to a running process?
<Xalius> like ltrace for example?
<tomeu> I would be surprised if that was possible
<Xalius> I attached ltrace to xwayland and that works
<Xalius> but that didn't really get me much further
<Xalius> I would like to apitrace between xwayland and mesa
<Xalius> it works with Xephyr
<Xalius> but using Xephyr and glamor works on panfrost
<Xalius> whereas xwayland and glamor cause issues for me
<alyssa> Xalius: welcome to glamor hell, there is no return past this point
* alyssa shakes head sadly
<Xalius> it doesn't really matter what I run in xwayland, 5 line gtk hello world is enough to trigger the issues
<alyssa> Which WM is this, by the way?
<alyssa> (or desktop environment)
<Xalius> sway
<alyssa> Alright
<Xalius> native applications work fine
<alyssa> Does native X (not wayland) work fine with those apps as well?
<Xalius> but anything in xwayland causes kernel side issues in panfrost
<alyssa> (since native X is at least marginally easy to debug than xwayland I think)
<Xalius> in X those applications just draw their UI very slow
<Xalius> and with some articfacts
<Xalius> like Gimp
<Xalius> disabling glamor doesn't make much difference in X
<Xalius> but it does in xwayland
<Xalius> xwayland without glamor works fine
<alyssa> Gotcha
<alyssa> Which chip did you say you're using?
<Xalius> rk3399
<Xalius> pinebook pro
<alyssa> Hrm, same here.
<Xalius> can you run stuff in Xwayland?
<Xalius> I have mesa-git from 2 days ago, xorg is still release
<alyssa> Yeah, I did `GDK_BACKEND=x11 [...]` with various gnome apps which I assume are gtk
<alyssa> Xalius: by the way, can I say your theme is amazing? :)
<Xalius> :)
<alyssa> Oh, hm, I wonder..
<alyssa> Xalius: Did you try the disable tiling patch floating around?
<Xalius> yes
<alyssa> Aight
<Xalius> glamor works in Xephyr
<Xalius> without causing any issues
<Xalius> I tried both glamor with GL and GLES context
<Xalius> that's how it looks, https://imgur.com/a/xnNSrSv
<alyssa> I suppose "use Krita" isn't a good answer ;)
<Xalius> KiCad and freecad are the same ;)
<Xalius> I mean I can live without glamor
<Xalius> those applications are totally usable with glamor turned off
<alyssa> Krita can't do CAD, no...
<Xalius> but I was mainly interested how the problems on the kernel side are triggered just by that specific set of cirumstances
<Xalius> dosbox, lzdoom, openarena, native wayland stuff all run fine
<alyssa> Xalius: GIMP+sway+Xwayland is working just fine for me, though slow.
<Xalius> interesting
<alyssa> Whch kernel do you run?
<Xalius> maybe something else is brooken on manjaro-arm
<Xalius> 5.5-rc3
<Xalius> so the drm_sched fixes are already in
<alyssa> tomeu: ^ Maybe something regressed kernel side
<Xalius> it was already there on 5.4.x
<Xalius> but there mixed in with the drm_sched stuff
<Xalius> maybe my xorg is broken?
<Xalius> it's just 1.20.6 stable
<urjaman> i'm running Xorg fine on 5.4.11 right now (archlinux-arm tho, not manjaro)
<Xalius> Xorg works for me too
<Xalius> it's only stuff in xwayland
<Xalius> manjaro-arm is mostly archlinux-arm
<urjaman> ah so you're not running Xwayland because Xorg is broken
<Xalius> no
<urjaman> i was like why are you doing that ...
<Xalius> xwayland mostly because sway
<Xalius> and it works in Xephyr
<urjaman> i didnt know Xephyr actually ran glamor
<Xalius> it has a -glamor and -glamor_gles2 switch at least
<urjaman> not that i have used it in half a decade :P
<tomeu> narmstrong: finally got to test STK under gnome-shell on the s912, and feels quite fast
<tomeu> not slower than last time, when I tested on rk3399
<Xalius> I suspected something in cairo first
<narmstrong> tomeu: nice !
<Xalius> but my 5 line empty gtk window triggers the issue too
<Xalius> even xterm does
<urjaman> i'm pretty sure i ran a bunch of gtk things in sway back when i dabbled with it...
<urjaman> xterm is not gtk btw
<urjaman> afaik atleast
<Xalius> yeah
<Xalius> I have no clue how to narrow that down from the kernel side
<narmstrong> i should take time to test STK on VIM2...
<Xalius> so I was going down the stack from the userland side...
<tomeu> bbrezillon: is the HUD still broken?
<tomeu> Xalius: what's an easy way of reproducing that if I have gnome-shell with xwayland already running?
<urjaman> oh correction, i'm still running 5.4.10, tho i suppose it doesnt matter
<urjaman> (i thought i had compiled that update but nah...)
<Xalius> tomeu: for me, starting xterm or some gtk2 app works to trigger the issue
<Xalius> I just get a lot of
<Xalius> [ 6457.946569] panfrost ff9a0000.gpu: gpu sched timeout, js=1, config=0x3301, status=0x8, head=0x21a8800, tail=0x21a3980, sched_job=00000000588ae8ea
<Xalius> [ 6457.946596] panfrost ff9a0000.gpu: gpu sched timeout, js=0, config=0x3300, status=0x8, head=0x2402a00, tail=0x2402a00, sched_job=000000009ac51f17
<Xalius> [ 6457.949814] panfrost ff9a0000.gpu: Unhandled Page fault in AS0 at VA 0x0000000002202E80
<Xalius> Reason: TODO
<Xalius> raw fault status: 0x8002C3
<Xalius> decoded fault status: SLAVE FAULT
<Xalius> exception type 0xC3: TRANSLATION_FAULT_LEVEL3
<Xalius> access type 0x2: READ
<Xalius> source id 0x80
<urjaman> please dont spam
<Xalius> [ 6457.962913] panfrost ff9a0000.gpu: Unhandled Page fault in AS0 at VA 0x0000000002215500
<Xalius> Reason: TODO
<Xalius> raw fault status: 0x48002C3
<Xalius> decoded fault status: SLAVE FAULT
<Xalius> exception type 0xC3: TRANSLATION_FAULT_LEVEL3
<Xalius> access type 0x2: READ
<Xalius> source id 0x480
<Xalius> [ 6458.064309] panfrost ff9a0000.gpu: Unhandled Page fault in AS0 at VA 0x0000000002213600
<Xalius> Reason: TODO
<Xalius> raw fault status: 0x48002C3
<Xalius> decoded fault status: SLAVE FAULT
<Xalius> exception type 0xC3: TRANSLATION_FAULT_LEVEL3
<Xalius> access type 0x2: READ
<Xalius> source id 0x480
<Xalius> [ 6458.128752] panfrost ff9a0000.gpu: Unhandled Page fault in AS0 at VA 0x0000000002215500
<Xalius> Reason: TODO
<Xalius> raw fault status: 0x48002C3
<Xalius> decoded fault status: SLAVE FAULT
<Xalius> exception type 0xC3: TRANSLATION_FAULT_LEVEL3
<Xalius> access type 0x2: READ
<Xalius> source id 0x480
<Xalius> [ 6458.193074] panfrost ff9a0000.gpu: Unhandled Page fault in AS0 at VA 0x0000000002203E80
<Xalius> Reason: TODO
<Xalius> raw fault status: 0x28002C3
<Xalius> decoded fault status: SLAVE FAULT
<Xalius> exception type 0xC3: TRANSLATION_FAULT_LEVEL3
<Xalius> access type 0x2: READ
<Xalius> source id 0x280
<Xalius> [ 6458.314303] panfrost ff9a0000.gpu: Unhandled Page fault in AS0 at VA 0x0000000002202E80
<Xalius> Reason: TODO
<Xalius> raw fault status: 0x8002C3
<Xalius> decoded fault status: SLAVE FAULT
<Xalius> exception type 0xC3: TRANSLATION_FAULT_LEVEL3
<Xalius> access type 0x2: READ
<Xalius> source id 0x80
<Xalius> [ 6458.378718] panfrost ff9a0000.gpu: Unhandled Page fault in AS0 at VA 0x0000000002202E80
<Xalius> Reason: TODO
<Xalius> raw fault status: 0x8002C3
<Xalius> decoded fault status: SLAVE FAULT
<Xalius> exception type 0xC3: TRANSLATION_FAULT_LEVEL3
<Xalius> access type 0x2: READ
<Xalius> source id 0x80
<Xalius> [ 6458.453016] panfrost ff9a0000.gpu: Unhandled Page fault in AS0 at VA 0x000000000220F880
<Xalius> Reason: TODO
<Xalius> raw fault status: 0x68002C3
<Xalius> decoded fault status: SLAVE FAULT
<Xalius> exception type 0xC3: TRANSLATION_FAULT_LEVEL3
<Xalius> access type 0x2: READ
<Xalius> source id 0x680
<Xalius> [ 6458.456532] panfrost ff9a0000.gpu: gpu sched timeout, js=0, config=0x3300, status=0x8, head=0x2402a00, tail=0x2402a00, sched_job=00
<Xalius> ah sorry
<Xalius> that was supposed to be pastebin
<robmur01> "sched timeouts and page faults" would have sufficed :(
<Xalius> yeah, sorry I had the wrong thing in the paste buffer :/
<robmur01> there's not really much to diagnose without the contents of memory to hand
<robmur01> and the timeouts are probably caused by the page faults anyway ;)
<tomeu> Xalius: works here :/
<Xalius> on what distro are you running?
<tomeu> debian
<Xalius> I have swapped around most components from arch-arm/manjaro-arm without success
<tomeu> kernel 5.3
<Xalius> xorg stable or xorg master also produced the same
<urjaman> hmm i do still have the not-yet-upstreamed part of bbrezillon's patch set in my kernel - that dealt with some of the page faults i was seeing before
<Xalius> I haven't tried kernel 5.3
<urjaman> (some still happen occasionally but it's not that easy to track down)
<Xalius> 5.4 and 5.5-rcX
<Xalius> I will try it on Gentoo next
<Xalius> that image is cooking atm
<Xalius> rarely I see https://hastebin.com/raw/cihumuyibo but that is probably also a consecutive error?
<urjaman> yeah likely
* urjaman has a C201 for a lap warmer right now
<urjaman> umm yeah.
<urjaman> i decided that i needed some sort of a comparison on this mesa update so i did x11perf -shmputxy100 before and after
<urjaman> from about 70/sec to 27/sec
<urjaman> i'ma bisect
davidlt has joined #panfrost
buzzmarshall has joined #panfrost
robher has quit [Remote host closed the connection]
daniels has quit [Remote host closed the connection]
robher has joined #panfrost
daniels has joined #panfrost
Xalius has quit []
davidlt has quit [Remote host closed the connection]
davidlt has joined #panfrost
Lyude has quit [Read error: Connection reset by peer]
Lyude has joined #panfrost
daniels has quit [Remote host closed the connection]
robher has quit [Remote host closed the connection]
robher has joined #panfrost
daniels has joined #panfrost
<urjaman> alyssa: "32851ff71521d203c08e26b3198d96d0f6467bb5 is the first bad commit"
<urjaman> that is "panfrost: Cleanup tiling selection logic"
<raster> aaaaaah "cleanup" commits breaking things... always happens over the decades. you learn to beware of "cleanup" commits :)
<urjaman> the commit does claim to have a non-zero effect (fixing dEQP case(s)), so it's not really a pure "cleanup" commit in that sense anyways
<raster> the classic sneaking real changes into a "shouldn't change anything" cleanup :)
pH5 has quit [Quit: bye]
karolherbst has quit [Ping timeout: 268 seconds]
yann has quit [Ping timeout: 265 seconds]
<alyssa> urjaman: Ohhh, interesting, okay.
<alyssa> urjaman: Thank you so much for bisecting that :)
enunes has quit [Ping timeout: 248 seconds]
<alyssa> Mistake #1 in that commit is that the checksumming logic changed for no good reason
<alyssa> Though checksumming hasn't made a big difference in perf when I was poking at it
<alyssa> Really though we just need the tiling routine to be 1000x faster, grr
<anarsoul> :)
<anarsoul> that'd be nice
<anarsoul> I guess some arm assembly should do the trick?
<anarsoul> maybe not 1000x
<anarsoul> but 2x should be possible
<alyssa> It's ... a lot harder than you might hope, unfortunately.
<alyssa> I might take another stab at it once the GLES3 drowning slows down
<alyssa> Let's uh investigate to see what the issue is these days
karolherbst has joined #panfrost
<alyssa> Problem #1 is that it's annoyingly difficult to instrument this code sanely since there's a subtle dependency on Gallium
<alyssa> Just need pipe_box, probably better to eliminate that entirely..
<anarsoul> alyssa: maybe it would make sense to utilize GPU for tiling/untiling?
<alyssa> anarsoul: That's a last restort imho
<anarsoul> why?
<anarsoul> it should be able to do that faster
<alyssa> I mean
<alyssa> hrm
<anarsoul> *faster than CPU
<alyssa> I suppose it's still an option worth looking into
<robmur01> cf. the "higher throughput" vs. "significantly larger setup overhead" from crypto and pretty much every other hardware offload argument ever :P
<alyssa> robmur01: A bug in libgcrypt's implementation of AES-GCM via armv8 crypto prevented me from chatting with a friend on XMPP
<alyssa> Software makes me happier sometime s:V
<alyssa> anarsoul: We agreed to drop RGB (24bpp) , yeah?
<alyssa> Although wait 6bpp formats are a thing in GL
raster has quit [Quit: Gettin' stinky!]
<alyssa> and 12bpp
<anarsoul> alyssa: yeah
<anarsoul> I dropped 24bpp in lima
<alyssa> *nods*
Ke has joined #panfrost
<alyssa> Really we need better templating etc
<alyssa> like etna has
<alyssa> Er no there are, agh
<alyssa> This code is at the mix of complex and slow alyssa is not happy
<anarsoul> you mean code in etnaviv?
<alyssa> Yeah
<alyssa> Okay, so, job#1 is to cleanup this code since there's a lot of half-baked opts here and it's just making things hard to work with
<alyssa> Supporting different bpp is annoying..
megi has quit [Ping timeout: 240 seconds]
yann has joined #panfrost
<anarsoul> alyssa: sure it is
<anarsoul> but we need that
<anarsoul> see the positive side - you don't have to deal with separate channels
* robmur01 recalls struggling to comprehend weird planar PC video modes (EGA, maybe?) as a youngster
* anarsoul started with zx spectrum so is used to weird framebuffer layouts
<alyssa> anarsoul: true!
stikonas has joined #panfrost
enunes has joined #panfrost
indy has quit [Quit: ZNC - http://znc.sourceforge.net]
buzzmarshall has quit [Remote host closed the connection]
guillaume_g has quit [Quit: Konversation terminated!]
raster has joined #panfrost
indy has joined #panfrost
icecream95 has joined #panfrost
davidlt has quit [Ping timeout: 265 seconds]
<urjaman> https://urja.dev/IMG_20200114_223104.jpg this is my 25 Mhz 486SL
<urjaman> https://urja.dev/IMG_20200114_223215.jpg this is my 1.8Ghz RK3288
krh has quit [Quit: leaving]
<icecream95> tomeu: /time
icecream95 has quit [Quit: leaving]
icecream95 has joined #panfrost
<icecream95> tomeu: HUD is still broken. The cause is known, but nobody ever got round to doing anything about it...
<anarsoul> go for it? :)
megi has joined #panfrost
<daniels> ^ :)
<daniels> if nothing else, understanding colourspaces will guarantee you a job for life
<icecream95> I could fix the HUD in about 5 minutes, but then it would break again when AA support gets added...
<icecream95> The `fire` Mesa demo looks quite different with Panfrost than llvmpipe...
<icecream95> And with llvmpipe it doesn't cause kernel faults. Time to compile the latest 5.5 rc now...
stikonas has quit [Remote host closed the connection]
icecream95 has quit [Ping timeout: 260 seconds]
stikonas has joined #panfrost
stikonas has quit [Remote host closed the connection]
stikonas has joined #panfrost
<alyssa> icecream95: Well, does it look better in Panfrost than llvmpipe? then it's a feature, not a bug! :D
icecream95 has joined #panfrost
raster has quit [Quit: Gettin' stinky!]
stikonas has quit [Remote host closed the connection]
stikonas has joined #panfrost
<anarsoul> alyssa: ouch, lima spends up to 30% of time in panfrost_access_tiled_image_generic() in dEQP-GLES2.functional.texture.*
<icecream95> Is anyone ever going to fix glquake ports with Panfrost?
<HdkR> ever? probably