alyssa changed the topic of #panfrost to: Panfrost - FLOSS Mali Midgard & Bifrost - https://gitlab.freedesktop.org/panfrost - Logs https://freenode.irclog.whitequark.org/panfrost - <daniels> avoiding X is a huge feature
megi has quit [Ping timeout: 245 seconds]
_whitelogger has joined #panfrost
fysa has quit [Remote host closed the connection]
vstehle has quit [Ping timeout: 246 seconds]
MistahDarcy has quit [Quit: Leaving]
_whitelogger has joined #panfrost
davidlt has joined #panfrost
davidlt has quit [Read error: Connection reset by peer]
davidlt has joined #panfrost
davidlt has quit [Ping timeout: 245 seconds]
davidlt has joined #panfrost
vstehle has joined #panfrost
davidlt has quit [Ping timeout: 246 seconds]
Elpaulo has quit [Remote host closed the connection]
Elpaulo has joined #panfrost
NeuroScr has joined #panfrost
Elpaulo has quit [Ping timeout: 245 seconds]
Elpaulo has joined #panfrost
chewitt has joined #panfrost
maciejjo has quit [Quit: leaving]
yann has quit [Ping timeout: 272 seconds]
adjtm_ has quit [Remote host closed the connection]
maciejjo has joined #panfrost
adjtm has joined #panfrost
pH5 has joined #panfrost
warpme_ has joined #panfrost
tgall_foo has quit [Ping timeout: 268 seconds]
megi has joined #panfrost
BenG83 has joined #panfrost
NeuroScr has quit [Quit: NeuroScr]
adjtm has quit [Ping timeout: 268 seconds]
yann has joined #panfrost
raster has joined #panfrost
jcureton has quit [Remote host closed the connection]
raster has quit [Remote host closed the connection]
davidlt has joined #panfrost
davidlt has quit [Ping timeout: 246 seconds]
adjtm has joined #panfrost
_whitelogger has joined #panfrost
raster has joined #panfrost
tgall_foo has joined #panfrost
warpme_ has quit [Quit: warpme_]
warpme_ has joined #panfrost
warpme_ has quit [Quit: warpme_]
warpme_ has joined #panfrost
vstehle has quit [Ping timeout: 245 seconds]
warpme_ has quit [Quit: warpme_]
warpme_ has joined #panfrost
warpme_ has quit [Quit: warpme_]
<alyssa> tomeu: Mm, interesting. Maybe we should just do that after all (to eliminate the 16MB upfront)
<alyssa> Still doesn't solve the blend constant issue, of course
<alyssa> (Connor pointed out we should maybe probably be keying instead; I'm just worried about combinatoric exploison)
vstehle has joined #panfrost
afaerber has quit [Quit: Leaving]
raster has quit [Remote host closed the connection]
<tomeu> alyssa: what's the blend constant issue? the additional BO that gets allocated when patching is disposed of in panfrost_delete_blend_state
<alyssa> tomeu: I worry about the perf implications
<alyssa> (When adding -x support)
<alyssa> Actually, not a problem
<alyssa> We just need to make sure the BO cache respects execute/no-execute flags
<tomeu> I don't see much of a perf impact with deqp
<tomeu> ah yes, I had to do that
<alyssa> :+1:
raster has joined #panfrost
raster has quit [Ping timeout: 245 seconds]
raster has joined #panfrost
<alyssa> Besides chromium, does anyone have a list of apps known to be unacceptably buggy with Panfrost?
* alyssa wants to make sure the 19.2 release is smooth
TheKit has quit [Ping timeout: 248 seconds]
TheKit has joined #panfrost
<EmilKarlson> I would not say Xorg with glamor is usable for me
<EmilKarlson> well a wekk ago anyway
<EmilKarlson> but if it works for others, might be just my setup
<EmilKarlson> in general might be related to gtk2 applications
<EmilKarlson> but if you can reproduce chromium, I'd work on that
pH5 has quit [Quit: bye]
shadeslayer has quit [Quit: http://quassel-irc.org - Chat comfortably. Anywhere.]
<EmilKarlson> when is the release btw.?
<EmilKarlson> my best reproducer so far is to start claws-mail, no internet connection or setup is required
<EmilKarlson> possibly workspace switches or something
shadeslayer has joined #panfrost
shadeslayer has quit [Remote host closed the connection]
shadeslayer has joined #panfrost
<alyssa> EmilKarlson: Release is tomorrow..
<alyssa> Xorg+glamor has been okayish to me? what DE?
<EmilKarlson> lxde
<EmilKarlson> with xmonad
* alyssa can't support xmonad ;P
<alyssa> Hello from MATE + marco
* alyssa can probably try LXDE
<EmilKarlson> thanks for the good work
<alyssa> EmilKarlson: Hey, this is my baby :)
<alyssa> EmilKarlson: Hello from LXDE + whatever its default is (openbox looks like)
<alyssa> Performance is very nice, I'm pleasantly surprised!
<alyssa> Uh oh STK rendering somehow regressed maybe ummm that will be on today's todo
<alyssa> EmilKarlson: claws-mail is open, seems fine to me
<alyssa> No faults, no obvious visual glitches,...
<alyssa> EmilKarlson: Mind trying something other than xmonad?
<alyssa> GNOME Terminal is broken under GNOME+X11 umm
<alyssa> (But it's fine under GNOME+Wayland or under MATE+X11)
<EmilKarlson> nothing in kernel log?
raster has quit [Remote host closed the connection]
<shadeslayer> alyssa: does this look more sensible? http://paste.ubuntu.com/p/j5kFDqnmMg/
stikonas has joined #panfrost
<alyssa> EmilKarlson: Not that I saw
<EmilKarlson> ok
<EmilKarlson> anyway, if sway works with panfrost, I'll probably migrate
<alyssa> shadeslayer: That looks alright, but I'm going to hold off merging until after the 19.2 release since we're on feature freeze today ;)
<alyssa> Do we have results from index bias CTS stuff?
<EmilKarlson> that was going to happen eventually
<alyssa> shadeslayer: You should be able to run: gles31 ./deqp-gles31 -n dEQP-GLES31.functional.draw_base_vertex.*
<alyssa> in VK-GL-CTS/modules/gles31
<alyssa> gles3 = MESA_GLES_VERSION_OVERRIDE=3.0 PAN_MESA_DEBUG=deqp
<alyssa> er no
<alyssa> gles31 = MESA_GLES_VERSION_OVERRIDE=3.1 PAN_MESA_DEBUG=deqp
<alyssa> 8 steps of bisection good golly
raster has joined #panfrost
<alyssa> 5 steps to go
<alyssa> 1 step left
<alyssa> "pan/midgard: Add post-RA move elimination" Hngrh
* alyssa isn't seeing the problem
<alyssa> Oh, I see... something..
<alyssa> Here lies the bug but how did my change cause that lol
yann has quit [Ping timeout: 272 seconds]
<alyssa> SuperTuxKart fixed, there we rae.
Stenzek has quit [Ping timeout: 264 seconds]
Stenzek has joined #panfrost
<alyssa> Trying to debug glamor issues but running out of ideas
<alyssa> flttening c/f doesn't help
adjtm has quit [Quit: Leaving]
<enunes> alyssa: hey, I recently saw https://gitlab.freedesktop.org/mesa/mesa/blob/master/src/panfrost/midgard/midgard_nir_algebraic.py#L70 and realized we can probably reuse that in lima, we currently have a much bigger and uglier chunk of code to do the same thing
<enunes> I thought about moving it to the panfrost/shared directory, rename it to panfrost
<enunes> probably turn it into a mul 1/pi which is easier to implmeent
<enunes> any thoughts?
<alyssa> enunes: \o
<alyssa> I mean, if you don't already have a lima-specific nir_algebraic.py, you should definitely start one; it's great :)
<alyssa> It looks like you need 1/tau instead of 1/pi?
<enunes> hmm right, we have 1/2pi...
<alyssa> enunes: Anyway, it doesn't sound like we need to share code per se
<alyssa> Just do your own lima_nir_algebraic.py with the pass you need :)
<enunes> yeah that makes more sense for now
<alyssa> (There's delightfully little boilerplate needed... just the python which you can cargo cult from us / freedreno, some meson magic, and then you just #include and call it as a NIR opt)
<alyssa> Let me know if you have any questions! :)
<enunes> sure, at my first glance it seemed like it was the same
<alyssa> enunes: BTW, looking forward to meeting y'all at XDC? :)
<enunes> alyssa: yeah I'll attend this year, looking forward too
<alyssa> \o/
<alyssa> Now, back to our usual channel of ALYSSA BANGING HER HEAD AGAINST A WALL (aka opencl):
<alyssa> I added a pass to string masks thorugh so addresses are only vec2 now
<alyssa> Unfortunately, it's still totally broken since addresses are 64-bit and need to be 64-bit aligned, which RA doesn't know about yet
<alyssa> So we'll need to teach RA about alignment for 64-bit.. but first we'll need to string 64-bit support through in the first place, good golly.
<alyssa> I'm thinking I should rebase and try to cleanup first but.. hmm
<alyssa> I'm in for a world of pain when I need to support 8/16 bit RA, huh :(
* alyssa added a hack
<alyssa> We'll deal with this properly later; I just want to make forward progress rn
<alyssa> (Not a hack to be pushed to master, ever, just so I can proceed on my local branch)
<alyssa> DOH~!
<alyssa> I just realized the trick for load/store
<alyssa> Those extra bits aren't specifying register *size*, they're specifying a register *shift*
<alyssa> This is Arm we're talking about, of course they'll add barrel shifters to random places ;D
<alyssa> HdkR: ^^
* alyssa verifies
<alyssa> Hrm
<alyssa> I see it generating shifts/muls/etc, despite there being no real reason to... so maybe it's not a shift
<alyssa> Or maybe the compiler just didn't anticipate this pattern
<alyssa> Thinking the latter tbh
<alyssa> Regardless that's defn what it is
<alyssa> 3 bits up there
<alyssa> Hm, case not totally closed..
yann has joined #panfrost
<alyssa> Because for some reason, the addresses themselves are being shifted too...
<alyssa> ------Oh
<alyssa> I have to think *even more Arm*
<alyssa> Barrel shifters only work on the second source.
<alyssa> Right.
NeuroScr has joined #panfrost
adjtm has joined #panfrost
afaerber has joined #panfrost
<alyssa> Headmashing succcessful.
<alyssa> Indirect SSBO accesses work now.
<HdkR> alyssa: offset scale is very useful :)
<alyssa> HdkR: You think useful, I think another 150 lines of backend opt pass to make use of it :(
<HdkR> :)
<alyssa> HdkR: It occurs to me there's probably someone out there who's doing compute stuff in vertex shaders with SSBO + rasterizer discard
<alyssa> Tho TBF vertex and compute shaders are almost identical on Midgard so maybe it's reasonable :P
<HdkR> Could do
<HdkR> Not sure what you would be testing in the vertex stage that regular culling doesn't solve though
<HdkR> Also AAA games typically do a compute pass first before passing data over to VSPS these days :P
<HdkR> But this is just because AMD's geometry pipeline has been relatively low performing
<alyssa> Blop
<HdkR> (Navi fixes that problem)
<HdkR> Supposedly 4x throughput improvement in best case
<alyssa> Hmmm
<bnieuwenhuizen> HdkR: woah, source?
stikonas has quit [Remote host closed the connection]
<HdkR> ah. Some random AMD engineer who probably shouldn't have said that :P
<bnieuwenhuizen> :P
raster has quit [Remote host closed the connection]