February 15, 2018 
< | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | >
February 15, 2018 [00:08:59] *** groton <groton!~groton@unaffiliated/groton> has joined ##OpenGL[00:10:20] *** Asu` <Asu`!~sdelang@92.184.100.134> has quit IRC (Remote host closed the connection)[00:10:51] *** Gama11_ <Gama11_!~quassel@p5090244F.dip0.t-ipconnect.de> has quit IRC (Read error: Connection reset by peer)[00:16:28] *** svp <svp!~svp@unaffiliated/svp> has quit IRC (Quit: My MacBook has gone to sleep. ZZZzzz…)[00:17:52] *** davr0s <davr0s!~textual@host86-147-196-14.range86-147.btcentralplus.com> has quit IRC (Quit: My MacBook Pro has gone to sleep. ZZZzzz…)[00:25:20] *** slidercrank <slidercrank!~slidercra@ircpuzzles/2015/april-fools/fifth/slidercrank> has quit IRC (Ping timeout: 265 seconds)[00:31:01] *** glYoda <glYoda!~MTLYoda@c-73-25-27-206.hsd1.or.comcast.net> has joined ##OpenGL[00:35:16] *** brallan <brallan!~brallan@186.96.85.180> has joined ##OpenGL[00:36:20] *** panda81 <panda81!a2f6d81c@gateway/web/cgi-irc/kiwiirc.com/ip.162.246.216.28> has joined ##OpenGL[00:38:41] *** Albori <Albori!~Albori@64-251-148-158.fidnet.com> has quit IRC (Ping timeout: 248 seconds)[00:40:03] *** castleboar <castleboar!~bes@unaffiliated/castlelore> has quit IRC (Ping timeout: 256 seconds)[00:40:05] *** brallan <brallan!~brallan@186.96.85.180> has quit IRC (Ping timeout: 240 seconds)[00:41:30] *** asecretcat <asecretcat!~allisonze@pool-72-77-42-97.pitbpa.fios.verizon.net> has quit IRC (Quit: let us connect our intestines and mutually digest)[00:43:02] *** Lunatrius <Lunatrius!~Lunatrius@77.38.21.155> has quit IRC (Ping timeout: 255 seconds)[00:43:27] *** acksys_ <acksys_!~fritz@mail12.shutterfly.com> has quit IRC (Ping timeout: 240 seconds)[00:46:02] *** Lunatrius <Lunatrius!~Lunatrius@77.38.21.155> has joined ##OpenGL[00:46:10] *** lasserix <lasserix!~Nona@cpe-142-129-71-216.socal.res.rr.com> has quit IRC (Read error: Connection reset by peer)[00:47:31] *** xerpi <xerpi!~xerpi@206.red-83-45-197.dynamicip.rima-tde.net> has quit IRC (Quit: Leaving)[00:52:08] *** ipe <ipe!~ipe@91-156-244-112.elisa-laajakaista.fi> has quit IRC (Read error: Connection reset by peer)[00:59:53] *** KAHR-Alpha <KAHR-Alpha!~Alpha@2a01cb0d04c90500f11c8d43c346cc6c.ipv6.abo.wanadoo.fr> has quit IRC (Ping timeout: 256 seconds)[01:03:22] *** Albori <Albori!~Albori@64-251-148-158.fidnet.com> has joined ##OpenGL[01:05:21] *** PaulCapestany <PaulCapestany!~PaulCapes@ip72-209-228-50.dc.dc.cox.net> has quit IRC (Ping timeout: 248 seconds)[01:05:51] *** l3dx <l3dx!uid2250@gateway/web/irccloud.com/x-esnkmoeihrwilenq> has quit IRC (Quit: Connection closed for inactivity)[01:07:30] *** jdashg <jdashg!~jdashg@corp-nat.fw1.untrust.mtv2.mozilla.net> has quit IRC (Ping timeout: 256 seconds)[01:08:05] *** PaulCapestany <PaulCapestany!~PaulCapes@ip72-209-228-50.dc.dc.cox.net> has joined ##OpenGL[01:13:11] *** Kingsquee <Kingsquee!~kingsquee@d108-180-237-219.bchsia.telus.net> has quit IRC (Read error: Connection reset by peer)[01:13:38] *** Kingsquee <Kingsquee!~kingsquee@d108-180-237-219.bchsia.telus.net> has joined ##OpenGL[01:25:08] *** MrFlibble <MrFlibble!MrFlibble@90.196.204.40> has quit IRC (Ping timeout: 276 seconds)[01:33:38] *** suhdood <suhdood!~suhdood@2600:380:9914:a394:cc27:30d7:d31e:d29e> has joined ##OpenGL[01:34:27] *** derhass <derhass!~derhass@dslb-094-222-155-209.094.222.pools.vodafone-ip.de> has quit IRC (Quit: leaving)[01:34:28] *** asecretcat <asecretcat!~allisonze@pool-72-77-42-97.pitbpa.fios.verizon.net> has joined ##OpenGL[01:35:53] *** Kingsquee <Kingsquee!~kingsquee@d108-180-237-219.bchsia.telus.net> has quit IRC (Read error: Connection reset by peer)[01:36:21] *** Kingsquee <Kingsquee!~kingsquee@d108-180-237-219.bchsia.telus.net> has joined ##OpenGL[01:37:56] *** upgrdman <upgrdman!~upgrdman@blender/artist/upgrdman> has quit IRC (Quit: Leaving)[01:47:31] *** Kingsquee <Kingsquee!~kingsquee@d108-180-237-219.bchsia.telus.net> has quit IRC (Remote host closed the connection)[01:48:03] *** Kingsquee <Kingsquee!~kingsquee@d108-180-237-219.bchsia.telus.net> has joined ##OpenGL[02:09:55] *** kajika <kajika!~corentin@unaffiliated/kajika> has joined ##OpenGL[02:23:10] *** Fig1024 <Fig1024!~Fig@175.117.3.23> has joined ##OpenGL[02:23:31] *** davr0s <davr0s!~textual@host86-147-196-14.range86-147.btcentralplus.com> has joined ##OpenGL[02:29:10] *** ratchet_freak <ratchet_freak!~ratchetfr@ptr-82s3g7n8nco8egy36si.18120a2.ip6.access.telenet.be> has quit IRC (Ping timeout: 240 seconds)[02:31:42] *** SiN_Bizkit <SiN_Bizkit!~SiN_Bizki@studio.unigine.com> has quit IRC (Remote host closed the connection)[02:33:37] *** ville_ <ville_!~ville@37-136-22-232.rev.dnainternet.fi> has joined ##OpenGL[02:35:35] *** ville <ville!~ville@176-93-117-125.bb.dnainternet.fi> has quit IRC (Ping timeout: 240 seconds)[02:45:14] *** ville <ville!~ville@37-136-26-207.rev.dnainternet.fi> has joined ##OpenGL[02:47:33] *** ville_ <ville_!~ville@37-136-22-232.rev.dnainternet.fi> has quit IRC (Ping timeout: 256 seconds)[02:48:53] *** upgrdman <upgrdman!~upgrdman@blender/artist/upgrdman> has joined ##OpenGL[02:52:27] *** Orion] <Orion]!~H0i@unaffiliated/orion/x-3970838> has quit IRC (Ping timeout: 256 seconds)[02:58:19] *** SiN_Bizkit <SiN_Bizkit!~SiN_Bizki@studio.unigine.com> has joined ##OpenGL[03:14:33] *** YungMoonHodler_ <YungMoonHodler_!~me@unaffiliated/boscop> has joined ##OpenGL[03:17:50] *** YungMoon <YungMoon!~me@unaffiliated/boscop> has quit IRC (Ping timeout: 256 seconds)[03:35:05] *** YungMoonHodler_ is now known as YungMoon[03:38:09] *** Neomex <Neomex!~quassel@net-37-117-5-106.cust.vodafonedsl.it> has quit IRC (Ping timeout: 264 seconds)[03:38:36] *** DarkUranium <DarkUranium!~DarkUrani@77.38.34.248> has quit IRC (Read error: Connection reset by peer)[03:47:34] *** davr0s <davr0s!~textual@host86-147-196-14.range86-147.btcentralplus.com> has quit IRC (Quit: My MacBook Pro has gone to sleep. ZZZzzz…)[03:50:04] *** davr0s <davr0s!~textual@host86-147-196-14.range86-147.btcentralplus.com> has joined ##OpenGL[03:50:07] *** davr0s <davr0s!~textual@host86-147-196-14.range86-147.btcentralplus.com> has quit IRC (Client Quit)[03:51:41] *** fluffrabbit <fluffrabbit!~fluffrabb@173-122-184-174.pools.spcsdns.net> has joined ##OpenGL[03:52:08] *** davr0s <davr0s!~textual@host86-147-196-14.range86-147.btcentralplus.com> has joined ##OpenGL[03:52:31] *** davr0s <davr0s!~textual@host86-147-196-14.range86-147.btcentralplus.com> has quit IRC (Client Quit)[04:03:11] *** akawaka <akawaka!~akawaka@2605:e000:1701:c4d5:211:32ff:fe51:9639> has quit IRC (Excess Flood)[04:03:30] *** akawaka <akawaka!~akawaka@2605:e000:1701:c4d5:211:32ff:fe51:9639> has joined ##OpenGL[04:07:38] *** slime <slime!~slime73@24.215.81.93> has quit IRC (Quit: This computer has gone to sleep)[04:36:02] *** akawaka <akawaka!~akawaka@2605:e000:1701:c4d5:211:32ff:fe51:9639> has quit IRC (Ping timeout: 256 seconds)[04:36:10] *** akawaka <akawaka!~akawaka@2605:e000:1701:c4d5:211:32ff:fe51:9639> has joined ##OpenGL[04:38:46] *** suhdood <suhdood!~suhdood@2600:380:9914:a394:cc27:30d7:d31e:d29e> has quit IRC (Remote host closed the connection)[04:39:38] *** deepio_ <deepio_!~deepio@host-091-097-228-140.ewe-ip-backbone.de> has joined ##OpenGL[04:43:16] *** deepio <deepio!~deepio@host-091-097-228-249.ewe-ip-backbone.de> has quit IRC (Ping timeout: 256 seconds)[04:52:32] *** rizzo <rizzo!~RizzoTheR@p54B6BDA4.dip0.t-ipconnect.de> has quit IRC (Ping timeout: 252 seconds)[04:52:40] *** Fig1024 <Fig1024!~Fig@175.117.3.23> has quit IRC (Ping timeout: 240 seconds)[04:55:37] *** ville <ville!~ville@37-136-26-207.rev.dnainternet.fi> has quit IRC (Ping timeout: 256 seconds)[05:09:28] *** charlie5 <charlie5!~rod@124-170-214-48.dyn.iinet.net.au> has joined ##OpenGL[05:10:51] *** Ekho <Ekho!~Ekho@unaffiliated/ekho> has quit IRC (Remote host closed the connection)[05:19:23] *** Ekho <Ekho!~Ekho@unaffiliated/ekho> has joined ##OpenGL[05:20:14] *** castlelore <castlelore!~bes@cpe-76-173-200-54.hawaii.res.rr.com> has joined ##OpenGL[05:20:32] *** castlelore <castlelore!~bes@cpe-76-173-200-54.hawaii.res.rr.com> has quit IRC (Changing host)[05:20:32] *** castlelore <castlelore!~bes@unaffiliated/castlelore> has joined ##OpenGL[05:20:33] *** castlelore is now known as castleboar[05:41:48] *** bookmark <bookmark!~IceChat77@47.196.110.8> has joined ##OpenGL[05:41:51] <bookmark> hi[05:42:08] <bookmark> does anyone know where i can find information on glm absolute value?[05:42:22] <bookmark> there is nothing in the manual though it seems there are a few functions[05:42:27] <bookmark> or templates[05:42:31] <bookmark> or whatever its called[05:42:50] <bookmark> i'm having a compilation error about abs()[05:45:23] <bookmark> nevermind[05:45:28] <bookmark> ill just stop using namespace glm[05:48:33] *** Kingsquee <Kingsquee!~kingsquee@d108-180-237-219.bchsia.telus.net> has quit IRC (Read error: Connection reset by peer)[05:49:23] *** Kingsquee <Kingsquee!~kingsquee@d108-180-237-219.bchsia.telus.net> has joined ##OpenGL[05:49:43] *** Fig1024 <Fig1024!~Fig@1.209.170.47> has joined ##OpenGL[06:04:10] *** tambre <tambre!~tambre@7886-b93f-04c4-3840-ab80-8a0a-07d0-2001.dyn.estpak.ee> has joined ##OpenGL[06:07:05] *** upgrdman <upgrdman!~upgrdman@blender/artist/upgrdman> has quit IRC (Ping timeout: 240 seconds)[06:17:00] *** suhdood <suhdood!~suhdood@2601:281:c502:25d9:49a5:2630:d1f1:a2c9> has joined ##OpenGL[06:17:51] *** davr0s <davr0s!~textual@host86-147-196-14.range86-147.btcentralplus.com> has joined ##OpenGL[06:28:58] *** immibis <immibis!~chatzilla@122-59-200-50.jetstream.xtra.co.nz> has joined ##OpenGL[06:32:56] *** krystcich_ <krystcich_!~krystcich@178235146009.dynamic-zab-01.vectranet.pl> has joined ##OpenGL[06:37:34] <bookmark> ill just stop using namespace glm[06:37:38] <bookmark> whoops[06:37:43] <bookmark> anyone know how to set identity on a glm::mat4 that is part of a structure?[06:38:25] <bookmark> like normally i'd just put glm::mat4 myMat(1.0);[06:38:35] <bookmark> but i can't really do that in a structure can i?[06:40:34] <bookmark> with and equal sign probably duh[06:41:11] *** BitPuffin <BitPuffin!~isak@c83-248-113-111.bredband.comhem.se> has joined ##OpenGL[06:41:55] <bookmark> sorry for the vague questions it's just that i have a clear screen rendering, no level, no characters :([06:44:38] <Stragus> Sorry, never used glm, good luck[06:45:12] <bookmark> it seems to be putting how higher fps[06:45:18] <bookmark> when i do get it working[06:45:32] <bookmark> but right now its busted[06:53:27] *** zeduckmaster <zeduckmaster!~zeduckmas@85.106.1.181> has joined ##OpenGL[07:06:45] *** spear2 <spear2!~spear2@71-95-117-135.dhcp.mdfd.or.charter.com> has joined ##OpenGL[07:13:00] *** krystcich_ <krystcich_!~krystcich@178235146009.dynamic-zab-01.vectranet.pl> has quit IRC (Quit: Going offline, see ya! (www.adiirc.com))[07:19:01] *** jdashg <jdashg!~jdashg@c-71-198-195-159.hsd1.ca.comcast.net> has joined ##OpenGL[07:20:22] *** suhdood <suhdood!~suhdood@2601:281:c502:25d9:49a5:2630:d1f1:a2c9> has quit IRC (Remote host closed the connection)[07:24:03] <charlie5> shiny balls are best ? ... (when testing lighting, i mean)[07:36:24] <bookmark> sweet i did it![07:37:59] *** jdashg <jdashg!~jdashg@c-71-198-195-159.hsd1.ca.comcast.net> has quit IRC (Ping timeout: 260 seconds)[07:49:56] *** qeed <qeed!~qeed@pool-108-44-253-40.clppva.fios.verizon.net> has quit IRC (Ping timeout: 276 seconds)[07:54:10] *** groton_ <groton_!~groton@unaffiliated/groton> has joined ##OpenGL[07:55:01] *** qeed <qeed!~qeed@pool-98-118-249-109.clppva.fios.verizon.net> has joined ##OpenGL[07:55:03] *** groton <groton!~groton@unaffiliated/groton> has quit IRC (Ping timeout: 256 seconds)[07:55:03] *** groton_ is now known as groton[08:16:39] *** davr0s <davr0s!~textual@host86-147-196-14.range86-147.btcentralplus.com> has quit IRC (Quit: My MacBook Pro has gone to sleep. ZZZzzz…)[08:19:48] *** davr0s <davr0s!~textual@host86-147-196-14.range86-147.btcentralplus.com> has joined ##OpenGL[08:35:53] *** slvn_ <slvn_!~slvn_@c2s31-1-78-245-90-111.fbx.proxad.net> has joined ##OpenGL[08:36:48] *** slidercrank <slidercrank!~slidercra@ircpuzzles/2015/april-fools/fifth/slidercrank> has joined ##OpenGL[08:38:39] *** Fig1024 <Fig1024!~Fig@1.209.170.47> has quit IRC (Ping timeout: 248 seconds)[08:42:06] *** watered_ <watered_!~dagger@gateway/tor-sasl/watered> has joined ##OpenGL[08:43:58] *** slidercrank <slidercrank!~slidercra@ircpuzzles/2015/april-fools/fifth/slidercrank> has quit IRC (Remote host closed the connection)[08:44:05] *** Matthijs <Matthijs!~quassel@unaffiliated/matthijs> has joined ##OpenGL[08:46:39] <Zero_Walker> gotta hate the vendor specific compilation for the shaders. Tried to run my stuff on a friends and had to try to solve it as the shaders complained when compiling[08:46:48] *** stefkos <stefkos!~Pawel@82.177.144.226> has joined ##OpenGL[08:47:31] <Zero_Walker> even though glslangvalidator seemed to work when creating Spirv (note: i do not use Spirv in the application itself though, but would have hoped it would be strict enough to validate it)[08:47:37] <bookmark> did you get it fixed?[08:48:12] <karalaine> what kind of error did you get? would cool to know common pitfalls between vendors[08:53:50] *** l3dx <l3dx!uid2250@gateway/web/irccloud.com/x-cyqywgmluqsffhqq> has joined ##OpenGL[08:54:18] *** watered_ <watered_!~dagger@gateway/tor-sasl/watered> has quit IRC (Remote host closed the connection)[08:56:38] *** watered_ <watered_!~dagger@gateway/tor-sasl/watered> has joined ##OpenGL[08:57:39] *** deepio_ <deepio_!~deepio@host-091-097-228-140.ewe-ip-backbone.de> has quit IRC ()[09:02:41] *** davr0s <davr0s!~textual@host86-147-196-14.range86-147.btcentralplus.com> has quit IRC (Quit: My MacBook Pro has gone to sleep. ZZZzzz…)[09:05:21] *** groton <groton!~groton@unaffiliated/groton> has quit IRC (Quit: groton)[09:18:26] *** moser <moser!~moser@223.73.115.71> has quit IRC (Ping timeout: 276 seconds)[09:23:29] *** Fig1024 <Fig1024!~Fig@175.117.3.23> has joined ##OpenGL[09:32:15] *** eto <eto!~eto@unaffiliated/eto> has quit IRC (Ping timeout: 268 seconds)[09:33:52] *** xerpi <xerpi!~xerpi@98.red-83-45-194.dynamicip.rima-tde.net> has joined ##OpenGL[09:33:54] <exDM69> Zero_Walker: I so much agree with you[09:34:17] <exDM69> Zero_Walker: trying to write GLSL that passes glslangValidator and apple's glsl compiler is pretty impossible[09:34:45] *** xerpi <xerpi!~xerpi@98.red-83-45-194.dynamicip.rima-tde.net> has quit IRC (Remote host closed the connection)[09:35:07] *** xerpi <xerpi!~xerpi@98.red-83-45-194.dynamicip.rima-tde.net> has joined ##OpenGL[09:36:58] *** Keniyal <Keniyal!~Keniyal@unaffiliated/keniyal> has joined ##OpenGL[09:48:24] *** jzeus <jzeus!~jzeus@68.21.148.158> has joined ##OpenGL[09:52:55] *** slidercrank <slidercrank!~slidercra@ircpuzzles/2015/april-fools/fifth/slidercrank> has joined ##OpenGL[09:53:09] <neure> Zero_Walker, there is spir-v for opengl too[09:53:57] <neure> ARB_gl_spirv[09:58:20] *** karab44 <karab44!~karab44@unaffiliated/karab44> has joined ##OpenGL[09:59:46] *** eto <eto!~eto@unaffiliated/eto> has joined ##OpenGL[10:00:01] <exDM69> unfortunately, spirv is not that widely available[10:01:33] *** spear2 <spear2!~spear2@71-95-117-135.dhcp.mdfd.or.charter.com> has quit IRC (Quit: Leaving)[10:02:00] *** xissburg <xissburg!~xissburg@unaffiliated/xissburg> has quit IRC (Quit: ZZZzzz…)[10:09:06] *** Keniyal <Keniyal!~Keniyal@unaffiliated/keniyal> has quit IRC (Ping timeout: 256 seconds)[10:09:29] *** Keniyal <Keniyal!~Keniyal@unaffiliated/keniyal> has joined ##OpenGL[10:12:43] *** ratchetfreak <ratchetfreak!c351a8d8@gateway/web/freenode/ip.195.81.168.216> has joined ##OpenGL[10:16:28] <neure> the extension is not supported? or something else?[10:18:06] <exDM69> the extension is not supported on a lot of platforms that are still relevant[10:18:09] <exDM69> apple, mobile, etc[10:18:35] <exDM69> basically if you have GL_ARB_spirv available, you probably also have vulkan[10:21:46] *** slidercrank <slidercrank!~slidercra@ircpuzzles/2015/april-fools/fifth/slidercrank> has quit IRC (Ping timeout: 265 seconds)[10:25:49] *** Gama11 <Gama11!~quassel@p5DCDB9A1.dip0.t-ipconnect.de> has joined ##OpenGL[10:31:44] *** kasper^ <kasper^!~safaf@82.137.10.92> has quit IRC ()[10:32:53] *** kasper^ <kasper^!~safaf@82.137.15.191> has joined ##OpenGL[10:34:04] *** ipe <ipe!~ipe@91-156-244-112.elisa-laajakaista.fi> has joined ##OpenGL[10:45:34] *** karab44 <karab44!~karab44@unaffiliated/karab44> has quit IRC ()[10:47:05] *** kasper^ <kasper^!~safaf@82.137.15.191> has quit IRC (Ping timeout: 240 seconds)[10:49:37] *** davr0s <davr0s!~textual@host86-147-196-14.range86-147.btcentralplus.com> has joined ##OpenGL[10:52:31] *** davr0s <davr0s!~textual@host86-147-196-14.range86-147.btcentralplus.com> has quit IRC (Client Quit)[10:55:40] *** ville <ville!~ville@87-93-19-74.bb.dnainternet.fi> has joined ##OpenGL[11:01:37] *** hgoel <hgoel!uid175521@gateway/web/irccloud.com/x-utfyglcknrcowsex> has quit IRC (Quit: Connection closed for inactivity)[11:05:38] *** bookmark <bookmark!~IceChat77@47.196.110.8> has quit IRC (Quit: Easy as 3.14159265358979323846...)[11:06:26] *** Serpent7776 <Serpent7776!~Serpent77@90-156-68-175.internetia.net.pl> has joined ##OpenGL[11:15:55] *** Elysion <Elysion!~Elysion@118.211.15.153> has quit IRC (Read error: Connection reset by peer)[11:16:20] *** Elysion <Elysion!~Elysion@118.211.15.153> has joined ##OpenGL[11:27:30] *** AlexSvideniuk <AlexSvideniuk!~oleksandr@195.234.75.138> has joined ##OpenGL[11:32:40] *** svp <svp!~svp@unaffiliated/svp> has joined ##OpenGL[11:34:53] *** lucaswang <lucaswang!~lucaswang@112.17.242.139> has joined ##OpenGL[11:43:35] *** charlie5 <charlie5!~rod@124-170-214-48.dyn.iinet.net.au> has quit IRC (Ping timeout: 240 seconds)[11:44:42] *** zzzoid <zzzoid!~zzzoid@85-195-242-83.fiber7.init7.net> has joined ##OpenGL[11:46:29] *** lucaswang <lucaswang!~lucaswang@112.17.242.139> has quit IRC (Ping timeout: 260 seconds)[11:53:47] *** davr0s <davr0s!~textual@host86-147-196-14.range86-147.btcentralplus.com> has joined ##OpenGL[11:53:49] *** akawaka <akawaka!~akawaka@2605:e000:1701:c4d5:211:32ff:fe51:9639> has quit IRC (Ping timeout: 256 seconds)[11:59:53] *** rizzo <rizzo!~RizzoTheR@p54B6BDA4.dip0.t-ipconnect.de> has joined ##OpenGL[12:03:05] <Zero_Walker> precisely, that's the problem that SPIRV is not widely available. It looks like the perfect solution though as then the standard stands for the compiler if i understand things correctly[12:09:09] *** xerpi <xerpi!~xerpi@98.red-83-45-194.dynamicip.rima-tde.net> has quit IRC (Quit: Leaving)[12:20:07] *** targz_ <targz_!984296e0@gateway/web/freenode/ip.152.66.150.224> has joined ##OpenGL[12:25:59] *** Lunatrius` <Lunatrius`!~Lunatrius@77.38.21.155> has joined ##OpenGL[12:26:56] *** Lunatrius <Lunatrius!~Lunatrius@77.38.21.155> has quit IRC (Ping timeout: 248 seconds)[12:28:35] *** bnka <bnka!~niklas@c-176be555.08-890-6e6b701.cust.bredbandsbolaget.se> has joined ##OpenGL[12:30:54] *** Lunatrius` <Lunatrius`!~Lunatrius@77.38.21.155> has quit IRC (Ping timeout: 256 seconds)[12:33:09] *** Lunatrius <Lunatrius!~Lunatrius@77.38.21.155> has joined ##OpenGL[12:38:57] *** SiN_Bizkit <SiN_Bizkit!~SiN_Bizki@studio.unigine.com> has quit IRC (Ping timeout: 240 seconds)[12:40:03] *** SiN_Bizkit <SiN_Bizkit!~SiN_Bizki@studio.unigine.com> has joined ##OpenGL[12:41:06] *** akawaka <akawaka!~akawaka@2605:e000:1701:c4d5:211:32ff:fe51:9639> has joined ##OpenGL[12:45:45] *** charlie5 <charlie5!~rod@124-170-214-48.dyn.iinet.net.au> has joined ##OpenGL[12:46:35] *** immibis <immibis!~chatzilla@122-59-200-50.jetstream.xtra.co.nz> has quit IRC (Ping timeout: 240 seconds)[12:47:17] *** bnka <bnka!~niklas@c-176be555.08-890-6e6b701.cust.bredbandsbolaget.se> has quit IRC (Quit: leaving)[12:47:35] *** bnka <bnka!~niklas@c-176be555.08-890-6e6b701.cust.bredbandsbolaget.se> has joined ##OpenGL[12:49:09] *** svp <svp!~svp@unaffiliated/svp> has quit IRC (Ping timeout: 256 seconds)[12:49:17] *** Lunatrius` <Lunatrius`!~Lunatrius@77.38.21.155> has joined ##OpenGL[12:50:25] *** Lunatrius <Lunatrius!~Lunatrius@77.38.21.155> has quit IRC (Ping timeout: 248 seconds)[12:50:25] *** Lunatrius` is now known as Lunatrius[12:54:33] *** Kingsquee <Kingsquee!~kingsquee@d108-180-237-219.bchsia.telus.net> has quit IRC (Quit: Just Monica.)[12:56:17] *** Borkr <Borkr!~Borkr@mail.seaonics.com> has joined ##OpenGL[12:57:25] *** kajika <kajika!~corentin@unaffiliated/kajika> has quit IRC (Quit: Konversation terminated!)[12:58:39] *** _Borkr_ <_Borkr_!~Borkr@mail.seaonics.com> has joined ##OpenGL[12:59:03] *** _Borkr_ <_Borkr_!~Borkr@mail.seaonics.com> has quit IRC (Remote host closed the connection)[12:59:13] *** Borkr <Borkr!~Borkr@mail.seaonics.com> has quit IRC (Client Quit)[13:01:35] *** davr0s <davr0s!~textual@host86-147-196-14.range86-147.btcentralplus.com> has quit IRC (Quit: My MacBook Pro has gone to sleep. ZZZzzz…)[13:02:43] *** Lunatrius` <Lunatrius`!~Lunatrius@77.38.21.155> has joined ##OpenGL[13:03:53] *** Lunatrius <Lunatrius!~Lunatrius@77.38.21.155> has quit IRC (Ping timeout: 276 seconds)[13:03:53] *** Lunatrius` is now known as Lunatrius[13:04:11] *** slime <slime!~slime73@blk-215-81-93.eastlink.ca> has joined ##OpenGL[13:06:24] *** ASvideniuk <ASvideniuk!~oleksandr@195.234.75.138> has joined ##OpenGL[13:07:39] *** bnka <bnka!~niklas@c-176be555.08-890-6e6b701.cust.bredbandsbolaget.se> has quit IRC (Quit: leaving)[13:07:56] *** bnka <bnka!~niklas@c-176be555.08-890-6e6b701.cust.bredbandsbolaget.se> has joined ##OpenGL[13:09:18] *** AlexSvideniuk <AlexSvideniuk!~oleksandr@195.234.75.138> has quit IRC (Ping timeout: 240 seconds)[13:10:30] *** bnka <bnka!~niklas@c-176be555.08-890-6e6b701.cust.bredbandsbolaget.se> has quit IRC (Client Quit)[13:10:46] *** bnka <bnka!~niklas@c-176be555.08-890-6e6b701.cust.bredbandsbolaget.se> has joined ##OpenGL[13:11:29] *** Fr0stBit <Fr0stBit!5150@2a02:2149:8167:1b00:922b:34ff:fe36:b288> has joined ##OpenGL[13:12:54] <Fr0stBit> Hi guys! I'm want to generate an image histogram. How can i do it with compute shaders? Can i find any example?[13:15:12] <Fig1024> I'm sure google has examples[13:15:55] <Fr0stBit> Fig1024: I can't seem to find any[13:17:55] <Fig1024> https://www.opengl.org/discussion_boards/showthread.php/180692-OpenGL-Compute-Shaders-vs-OpenCL[13:19:26] <Fr0stBit> Fig1024: i don't know how i missed that! Title probably mislead me[13:20:04] <Fig1024> https://github.com/maknoll/GPU/blob/master/exercise06/Histogram/histogram.cpp[13:20:28] <Fr0stBit> ^ That one is not with compute shaders[13:21:37] *** kasper^ <kasper^!~safaf@82.137.10.67> has joined ##OpenGL[13:22:24] *** slime <slime!~slime73@blk-215-81-93.eastlink.ca> has quit IRC (Quit: This computer has gone to sleep)[13:25:20] *** ville_ <ville_!~ville@87-95-87-194.bb.dnainternet.fi> has joined ##OpenGL[13:26:09] *** kasumi-owari <kasumi-owari!~kasumi-ow@ftth-213-233-237-007.solcon.nl> has quit IRC (Ping timeout: 264 seconds)[13:26:35] *** bnka <bnka!~niklas@c-176be555.08-890-6e6b701.cust.bredbandsbolaget.se> has quit IRC (Quit: leaving)[13:26:58] *** DarkUranium <DarkUranium!~DarkUrani@77.38.34.248> has joined ##OpenGL[13:27:10] *** ville <ville!~ville@87-93-19-74.bb.dnainternet.fi> has quit IRC (Ping timeout: 240 seconds)[13:27:21] *** ville_ is now known as ville[13:27:24] *** charlie5 <charlie5!~rod@124-170-214-48.dyn.iinet.net.au> has quit IRC (Ping timeout: 260 seconds)[13:32:23] *** kasumi-owari <kasumi-owari!~kasumi-ow@ftth-213-233-237-007.solcon.nl> has joined ##OpenGL[13:34:58] *** zzzoid <zzzoid!~zzzoid@85-195-242-83.fiber7.init7.net> has quit IRC (Quit: Leaving)[13:34:58] *** ASvideniuk <ASvideniuk!~oleksandr@195.234.75.138> has quit IRC (Remote host closed the connection)[13:35:50] *** ASvideniuk <ASvideniuk!~oleksandr@195.234.75.138> has joined ##OpenGL[13:41:54] *** targz_ <targz_!984296e0@gateway/web/freenode/ip.152.66.150.224> has quit IRC (Quit: Page closed)[13:47:08] *** raynold <raynold!uid201163@gateway/web/irccloud.com/x-dlbhbiymgwqubcho> has quit IRC (Quit: Connection closed for inactivity)[13:49:55] *** jdashg <jdashg!~jdashg@c-71-198-195-159.hsd1.ca.comcast.net> has joined ##OpenGL[13:55:04] *** AlexSvideniuk <AlexSvideniuk!~oleksandr@195.234.75.138> has joined ##OpenGL[13:57:46] *** ASvideniuk <ASvideniuk!~oleksandr@195.234.75.138> has quit IRC (Ping timeout: 252 seconds)[14:01:04] *** Quetzal2 <Quetzal2!~Quetzal2@unaffiliated/quetzal2> has joined ##OpenGL[14:06:33] *** hellozee <hellozee!~hellozee@45.118.157.211> has joined ##OpenGL[14:07:33] *** ville <ville!~ville@87-95-87-194.bb.dnainternet.fi> has quit IRC (Ping timeout: 256 seconds)[14:09:08] *** jdashg <jdashg!~jdashg@c-71-198-195-159.hsd1.ca.comcast.net> has quit IRC (Ping timeout: 268 seconds)[14:12:32] *** delicado <delicado!~delicado@110.54.183.54> has joined ##OpenGL[14:24:28] *** ville <ville!~ville@87-95-60-183.bb.dnainternet.fi> has joined ##OpenGL[14:27:51] *** bnka <bnka!~niklas@c-89-47.eduroam.liu.se> has joined ##OpenGL[14:28:38] *** Fr0stBit <Fr0stBit!5150@2a02:2149:8167:1b00:922b:34ff:fe36:b288> has quit IRC (Quit: WeeChat 2.0.1)[14:29:35] *** delicado <delicado!~delicado@110.54.183.54> has quit IRC (Ping timeout: 240 seconds)[14:49:35] *** bnka <bnka!~niklas@c-89-47.eduroam.liu.se> has quit IRC (Quit: leaving)[14:49:51] *** bnka <bnka!~niklas@c-89-47.eduroam.liu.se> has joined ##OpenGL[14:57:24] *** bnka <bnka!~niklas@c-89-47.eduroam.liu.se> has quit IRC (Quit: leaving)[14:57:41] *** bnka <bnka!~niklas@c-89-47.eduroam.liu.se> has joined ##OpenGL[15:18:34] *** davr0s <davr0s!~textual@host86-147-196-14.range86-147.btcentralplus.com> has joined ##OpenGL[15:25:40] *** zeduckmaster <zeduckmaster!~zeduckmas@85.106.1.181> has quit IRC (Quit: Leaving)[15:29:47] *** ESphynx <ESphynx!~ESphynx@modemcable234.227-82-70.mc.videotron.ca> has joined ##OpenGL[15:29:58] <ESphynx> hey guys, why would I be having trouble reading from a depth texture?[15:30:11] <ESphynx> I fill it up with 1.0 but it seems to be all 0 :S[15:33:05] *** watered_ <watered_!~dagger@gateway/tor-sasl/watered> has quit IRC (Ping timeout: 255 seconds)[15:41:57] *** delicado <delicado!~delicado@110.54.183.54> has joined ##OpenGL[15:43:16] <Dave13h> ESphynx: actually 0 or very close to 0? could be your depth range is very big[15:44:09] *** rizzo <rizzo!~RizzoTheR@p54B6BDA4.dip0.t-ipconnect.de> has quit IRC (Ping timeout: 264 seconds)[15:44:11] <ESphynx> Dave13h: well the point is it's supposed to be 1[15:44:19] <ESphynx> I've filled hte texture with 1, i've tried to glClear it with 1's[15:44:33] <Dave13h> ie if your near is 0.0001, far is 100000.0 and most of your depth might be really close to 0.0001[15:45:09] <ESphynx> Dave13h: but I'm testing with a 'cleared' texture, not rendering output[15:46:51] <Dave13h> using glClearDepth(1.0); and glClear(GL_DEPTH_BUFFER_BIT); yeah?[15:47:00] <ESphynx> yes[15:47:42] <Dave13h> That should do it afaik, are you checking your framebuffer status is complete?[15:48:10] <ESphynx> i am[15:48:23] <ESphynx> and a glClear with color bit DOES clear the color buffer, which samples fine[15:49:40] <ESphynx> I also initialize the texture with 1.0 with glTexImage()[15:49:43] <ESphynx> I really don't get it[15:50:11] <Dave13h> Not sure I'm afraid, sorry =([15:50:38] <ESphynx> hmm :|[15:50:43] <Dave13h> How are you checking that its not working? reading it back or drawing it?[15:51:17] <ESphynx> drawing it[15:51:19] <ESphynx> texture( texDepth, fragcoord )[15:51:23] <ESphynx> .r[15:54:27] <Dave13h> Absolutely sure you've bound the texture uniform?[15:54:30] * Dave13h runs out of ideas[15:54:34] <ESphynx> positive! :P[15:55:53] <Dave13h> Not fiddling with the depth range or anything are you?[15:56:23] <ESphynx> I am not[15:58:12] *** ville <ville!~ville@87-95-60-183.bb.dnainternet.fi> has quit IRC (Quit:)[15:58:42] <ESphynx> it's set up as glTexImage2D(GL_TEXTURE_2D, 0, GL_DEPTH_COMPONENT32F, size.w, size.h, 0, GL_DEPTH_COMPONENT, GL_FLOAT, null);[16:00:21] *** davr0s <davr0s!~textual@host86-147-196-14.range86-147.btcentralplus.com> has quit IRC (Quit: My MacBook Pro has gone to sleep. ZZZzzz…)[16:06:11] <Dave13h> Maybe try another format? GL_DEPTH_COMPONENT instead of 32F?[16:07:00] <karalaine> you could also check with renderdoc how is the actual texture[16:08:30] <ESphynx> GL_DEPTH_COMPONENT is that an actual format?[16:09:27] *** hellozee <hellozee!~hellozee@45.118.157.211> has quit IRC (Quit: Leaving)[16:09:28] <Dave13h> yeah, with GL_UNSIGNED_BYTE I believe[16:11:52] *** gareppa <gareppa!~gareppa@unaffiliated/gareppa> has joined ##OpenGL[16:17:01] <ESphynx> trying to understand this RenderDoc thing, never tried that before![16:17:07] <ESphynx> don't even think I heard of it before :P[16:17:20] <Dave13h> What does your attachment / glFramebufferTexture2D code look like?[16:17:30] <ESphynx> how can I look at the texture?[16:17:45] <ESphynx> it looks like glFramebufferTexture2D(GL_DRAW_FRAMEBUFFER, GL_DEPTH_ATTACHMENT, GL_TEXTURE_2D, texDepth, 0);[16:20:30] <Dave13h> What about if you try GL_FRAMEBUFFER instead?[16:21:06] <Dave13h> Does GL_DRAW_FRAMEBUFFER give it some weird hint that prevents you from reading it? or is that only if you are reading it back from that framebuffer?[16:22:39] *** gareppa <gareppa!~gareppa@unaffiliated/gareppa> has quit IRC (Quit: Leaving)[16:24:16] <ESphynx> Renderdoc seems to be telling me something's unbound[16:24:49] *** ImQ009 <ImQ009!~ImQ009@unaffiliated/imq009> has joined ##OpenGL[16:26:01] <ESphynx> though I don't really understand it all :P[16:30:46] <ESphynx> OK I'm sort of starting to understand Renderdoc[16:30:51] <ESphynx> the depth texture definite has 0's all over[16:36:33] <ESphynx> what does texture2d dsv 130 mean[16:44:02] <Zero_Walker> guten tag[16:44:47] <Dave13h> is 130 the GLSL version?[16:46:44] *** fluffrabbit <fluffrabbit!~fluffrabb@173-122-184-174.pools.spcsdns.net> has quit IRC (Quit: Leaving)[16:47:30] <ESphynx> most likely yes :P[16:50:56] <ESphynx> grrr[16:51:00] <ESphynx> I hate OpenGL.[16:51:14] <ESphynx> either glDepthMask(GL_TRUE); or glEnable(GL_DEPTH_TEST); did it[16:51:55] <ESphynx> glDepthMask(GL_TRUE); -- apparently[16:52:34] <ESphynx> you're explicitly passing GL_DEPTH_BUFFER_BIT to glClear() , why would DepthMask affect that :S[16:54:35] <ESphynx> it loos like my initial array passed to glTexImage doesn't work either[16:55:29] <ESphynx> ah now I know why lol[16:56:31] *** Asu <Asu!~sdelang@80.12.63.0> has joined ##OpenGL[16:56:33] *** delicado <delicado!~delicado@110.54.183.54> has quit IRC ()[16:58:02] <Dave13h> I would imagine it would NOP that part of the pipeline if the mask is false?[16:58:08] *** acksys_ <acksys_!~fritz@mail12.shutterfly.com> has joined ##OpenGL[16:58:14] <ESphynx> seems like it does :P[16:58:39] <ESphynx> so is it possible to to a blitframebuffer from my mulsisampled target (the window back buffer?) to my FBO with rgb/depth ?[16:58:41] *** krystcich_ <krystcich_!~krystcich@178235146015.dynamic-zab-01.vectranet.pl> has joined ##OpenGL[16:58:50] <ESphynx> it keeps telling me the depth formats don't match[17:04:55] *** upgrdman <upgrdman!~upgrdman@blender/artist/upgrdman> has joined ##OpenGL[17:04:58] *** xerpi <xerpi!~xerpi@98.red-83-45-194.dynamicip.rima-tde.net> has joined ##OpenGL[17:04:59] <ESphynx> I used WGL_DEPTH_BITS_ARB,16,[17:05:18] <ESphynx> even without multisampling it seems I get this error that my depth format do not match?[17:05:33] <ESphynx> even with GL_DEPTH_COMPONENT16[17:05:56] *** xerpi <xerpi!~xerpi@98.red-83-45-194.dynamicip.rima-tde.net> has quit IRC (Remote host closed the connection)[17:06:09] <ESphynx> oh with GL_DEPTH_COMPONENT it wants to copy![17:06:16] *** xerpi <xerpi!~xerpi@98.red-83-45-194.dynamicip.rima-tde.net> has joined ##OpenGL[17:08:32] *** grim001 <grim001!~grim001@ip70-173-209-228.lv.lv.cox.net> has quit IRC (Quit: Leaving)[17:14:25] *** Neomex <Neomex!~quassel@net-37-117-5-106.cust.vodafonedsl.it> has joined ##OpenGL[17:23:54] *** Matthijs <Matthijs!~quassel@unaffiliated/matthijs> has quit IRC (Quit: *waves frantically*)[17:34:52] *** bloops <bloops!~bloops@203.110.242.18> has joined ##OpenGL[17:35:17] *** groton <groton!~groton@unaffiliated/groton> has joined ##OpenGL[17:36:05] *** acksys_ <acksys_!~fritz@mail12.shutterfly.com> has quit IRC (Ping timeout: 240 seconds)[17:36:40] *** acksys_ <acksys_!~fritz@mail12.shutterfly.com> has joined ##OpenGL[17:37:32] *** upgrdman <upgrdman!~upgrdman@blender/artist/upgrdman> has quit IRC (Ping timeout: 260 seconds)[17:41:01] *** stefkos <stefkos!~Pawel@82.177.144.226> has quit IRC (Remote host closed the connection)[17:42:06] *** ASvideniuk <ASvideniuk!~oleksandr@195.234.75.138> has joined ##OpenGL[17:45:28] *** AlexSvideniuk <AlexSvideniuk!~oleksandr@195.234.75.138> has quit IRC (Ping timeout: 265 seconds)[17:49:08] *** Fr0stBit <Fr0stBit!5150@2a02:2149:8167:1b00:922b:34ff:fe36:b288> has joined ##OpenGL[17:50:26] <Fr0stBit> What is the best way to copy the current framebuffer's texture into some different smaller texture? Can it be done without a new framebuffer and a quad render that samples the previous framebuffer into the current one?[17:52:46] *** bnka <bnka!~niklas@c-89-47.eduroam.liu.se> has quit IRC (Quit: leaving)[17:53:43] *** Orion] <Orion]!~H0i@unaffiliated/orion/x-3970838> has joined ##OpenGL[17:55:00] <Fr0stBit> Does glCopyImageSubData crop or resize the image if the destination is smaller than the source?[17:56:08] *** groton <groton!~groton@unaffiliated/groton> has quit IRC (Quit: groton)[17:57:21] <Yaniel> looks like it returns an error in that case[17:57:26] <Yaniel> http://docs.gl/gl4/glCopyTexSubImage2D[17:57:54] <Yaniel> GL_INVALID_VALUE is generated if [rect is larger than] texture image being modified.[17:58:30] *** raynold <raynold!uid201163@gateway/web/irccloud.com/x-ztesgjmqzhqgnjij> has joined ##OpenGL[17:58:33] <Fr0stBit> damn[17:58:38] <Yaniel> wait wrong function[17:58:46] <Yaniel> > GL_INVALID_VALUE is generated if the dimensions of the either subregion exceeds the boundaries of the corresponding image object,[17:58:51] <Yaniel> same situation though[17:58:55] <Yaniel> http://docs.gl/gl4/glCopyImageSubData[17:59:35] <Fr0stBit> Seems like glBlitFramebuffer might do[18:00:08] <Yaniel> yeah that even has a filter parameter[18:00:14] *** davr0s <davr0s!~textual@host86-147-196-14.range86-147.btcentralplus.com> has joined ##OpenGL[18:00:15] <Fr0stBit> But on the filter param it says: Specifies the interpolation to be applied if the image is stretched. Must be GL_NEAREST or GL_LINEAR.[18:00:25] <Fr0stBit> Does stretching mean downsizing too?[18:01:48] <Yaniel> > If the sizes of the source and destination rectangles are not equal, filter specifies the interpolation method that will be applied to resize the source image[18:01:56] <Yaniel> http://docs.gl/gl4/glBlitFramebuffer[18:02:03] <Fr0stBit> Nice[18:02:06] <Yaniel> I'd consider that to count both up- and downscaling[18:02:19] <Yaniel> (of course, that is the manpage and not the spec)[18:02:50] *** Murii <Murii!~Murii@79.113.205.114> has joined ##OpenGL[18:02:52] <Yaniel> but if it's filtering to begin with I'd expect it to use the same path as texture sampling[18:06:57] *** _borkr_ <_borkr_!~borkr@static130-244.mimer.net> has joined ##OpenGL[18:09:46] *** hellozee <hellozee!~hellozee@45.118.157.211> has joined ##OpenGL[18:10:46] *** upgrdman <upgrdman!~upgrdman@blender/artist/upgrdman> has joined ##OpenGL[18:20:48] *** ImQ009 <ImQ009!~ImQ009@unaffiliated/imq009> has quit IRC (Quit: Leaving)[18:21:08] *** ImQ009 <ImQ009!~ImQ009@unaffiliated/imq009> has joined ##OpenGL[18:24:33] *** No0n3Left <No0n3Left!~No0n3Left@2600:100d:b113:50e:d87e:cbb1:a2dc:b4c9> has joined ##OpenGL[18:28:34] *** ShadowIce <ShadowIce!~pyoro@unaffiliated/shadowice-x841044> has joined ##OpenGL[18:33:23] *** M6HZ <M6HZ!~M6HZ@mar75-4-82-230-46-11.fbx.proxad.net> has quit IRC (Read error: Connection reset by peer)[18:36:17] *** davr0s <davr0s!~textual@host86-147-196-14.range86-147.btcentralplus.com> has quit IRC (Quit: My MacBook Pro has gone to sleep. ZZZzzz…)[18:36:36] *** Serpent7776 <Serpent7776!~Serpent77@90-156-68-175.internetia.net.pl> has quit IRC (Quit: Leaving)[18:36:50] *** rizzo <rizzo!~RizzoTheR@p54B6BDA4.dip0.t-ipconnect.de> has joined ##OpenGL[18:37:06] *** ImQ009_ <ImQ009_!~ImQ009@unaffiliated/imq009> has joined ##OpenGL[18:37:34] *** ratchetfreak <ratchetfreak!c351a8d8@gateway/web/freenode/ip.195.81.168.216> has quit IRC (Ping timeout: 260 seconds)[18:40:22] *** ImQ009 <ImQ009!~ImQ009@unaffiliated/imq009> has quit IRC (Ping timeout: 256 seconds)[18:42:37] *** bloops <bloops!~bloops@203.110.242.18> has quit IRC (Quit: Leaving)[18:47:47] *** No0n3Left <No0n3Left!~No0n3Left@2600:100d:b113:50e:d87e:cbb1:a2dc:b4c9> has quit IRC (Quit: Quiting)[18:47:52] *** Asu <Asu!~sdelang@80.12.63.0> has quit IRC (Ping timeout: 268 seconds)[18:48:11] *** Asu <Asu!~sdelang@80.12.63.143> has joined ##OpenGL[18:49:11] *** No0n3Left <No0n3Left!~No0n3Left@2600:100d:b113:50e:d87e:cbb1:a2dc:b4c9> has joined ##OpenGL[18:57:28] *** No0n3Left <No0n3Left!~No0n3Left@2600:100d:b113:50e:d87e:cbb1:a2dc:b4c9> has quit IRC (Quit: Quiting)[18:57:34] *** echotangoecho <echotangoecho!~echotango@92-108-127-60.cable.dynamic.v4.ziggo.nl> has joined ##OpenGL[19:00:25] *** xissburg <xissburg!~xissburg@unaffiliated/xissburg> has joined ##OpenGL[19:00:50] *** jzeus <jzeus!~jzeus@68.21.148.158> has quit IRC (Read error: Connection reset by peer)[19:01:09] *** Asu <Asu!~sdelang@80.12.63.143> has quit IRC (Remote host closed the connection)[19:01:18] *** Asu <Asu!~sdelang@80.12.63.143> has joined ##OpenGL[19:02:21] *** No0n3Left <No0n3Left!~No0n3Left@2600:100d:b113:50e:d87e:cbb1:a2dc:b4c9> has joined ##OpenGL[19:12:13] *** nidefawl <nidefawl!~nidefawl@p2E58DD99.dip0.t-ipconnect.de> has quit IRC (Ping timeout: 256 seconds)[19:13:14] *** ratchetfreak <ratchetfreak!~ratchetfr@ptr-82s3g7n8nco8egy36si.18120a2.ip6.access.telenet.be> has joined ##OpenGL[19:28:41] *** xerpi <xerpi!~xerpi@98.red-83-45-194.dynamicip.rima-tde.net> has quit IRC (Quit: Leaving)[19:30:55] *** No0n3Left <No0n3Left!~No0n3Left@2600:100d:b113:50e:d87e:cbb1:a2dc:b4c9> has quit IRC (Quit: Quiting)[19:30:56] *** slidercrank <slidercrank!~slidercra@ircpuzzles/2015/april-fools/fifth/slidercrank> has joined ##OpenGL[19:32:23] *** No0n3Left <No0n3Left!~No0n3Left@2600:100d:b113:50e:d87e:cbb1:a2dc:b4c9> has joined ##OpenGL[19:35:14] *** No0n3Left <No0n3Left!~No0n3Left@2600:100d:b113:50e:d87e:cbb1:a2dc:b4c9> has quit IRC (Client Quit)[19:37:18] *** No0n3Left <No0n3Left!~No0n3Left@2600:100d:b113:50e:d87e:cbb1:a2dc:b4c9> has joined ##OpenGL[19:37:49] *** fatalhalt <fatalhalt!~fatalhalt@c-67-163-60-93.hsd1.il.comcast.net> has quit IRC (Remote host closed the connection)[19:38:24] *** Thaelim <Thaelim!~suhdood@2601:281:c502:25d9:8003:d196:44fb:37fe> has joined ##OpenGL[19:40:09] *** fatalhalt <fatalhalt!~fatalhalt@c-67-163-60-93.hsd1.il.comcast.net> has joined ##OpenGL[19:41:09] *** kasper_ <kasper_!~safaf@82.137.10.92> has joined ##OpenGL[19:43:45] *** kasper^ <kasper^!~safaf@82.137.10.67> has quit IRC (Ping timeout: 248 seconds)[19:45:05] *** suhdood_ <suhdood_!~suhdood@2600:380:9914:a394:8003:d196:44fb:37fe> has joined ##OpenGL[19:46:40] *** suhdood_ <suhdood_!~suhdood@2600:380:9914:a394:8003:d196:44fb:37fe> has quit IRC (Remote host closed the connection)[19:47:15] *** Thaelimm <Thaelimm!~Thaelim@2600:380:9914:a394:8003:d196:44fb:37fe> has joined ##OpenGL[19:47:56] *** Thaelim <Thaelim!~suhdood@2601:281:c502:25d9:8003:d196:44fb:37fe> has quit IRC (Ping timeout: 252 seconds)[19:49:11] *** MrBusiness <MrBusiness!~ArcMrBism@2600:1700:6620:5660:dc9c:3733:a49c:c985> has quit IRC (Ping timeout: 255 seconds)[19:49:49] *** No0n3Left <No0n3Left!~No0n3Left@2600:100d:b113:50e:d87e:cbb1:a2dc:b4c9> has quit IRC (Remote host closed the connection)[19:50:58] *** No0n3Left <No0n3Left!~No0n3Left@2600:100d:b113:50e:d87e:cbb1:a2dc:b4c9> has joined ##OpenGL[19:52:47] *** acksys_ <acksys_!~fritz@mail12.shutterfly.com> has quit IRC (Ping timeout: 248 seconds)[19:59:25] *** Thaelimmm <Thaelimmm!~Thaelim@2601:281:c502:25d9:8003:d196:44fb:37fe> has joined ##OpenGL[20:00:21] *** No0n3Left <No0n3Left!~No0n3Left@2600:100d:b113:50e:d87e:cbb1:a2dc:b4c9> has quit IRC (Quit: Quiting)[20:01:26] *** upgrdman <upgrdman!~upgrdman@blender/artist/upgrdman> has quit IRC (Quit: Leaving)[20:02:29] *** Thaelimm <Thaelimm!~Thaelim@2600:380:9914:a394:8003:d196:44fb:37fe> has quit IRC (Ping timeout: 276 seconds)[20:02:35] *** No0n3Left <No0n3Left!~No0n3Left@2600:100d:b113:50e:d87e:cbb1:a2dc:b4c9> has joined ##OpenGL[20:05:42] *** derhass <derhass!~derhass@dslb-094-222-155-209.094.222.pools.vodafone-ip.de> has joined ##OpenGL[20:06:01] *** Silwer <Silwer!~SilwerKop@dslb-188-104-038-134.188.104.pools.vodafone-ip.de> has joined ##OpenGL[20:06:50] *** No0n3Left <No0n3Left!~No0n3Left@2600:100d:b113:50e:d87e:cbb1:a2dc:b4c9> has quit IRC (Client Quit)[20:21:10] *** jdashg <jdashg!~jdashg@corp-nat.fw1.untrust.mtv2.mozilla.net> has joined ##OpenGL[20:21:19] *** Thaelimmm <Thaelimmm!~Thaelim@2601:281:c502:25d9:8003:d196:44fb:37fe> has quit IRC (Remote host closed the connection)[20:21:52] *** Thaelimmm <Thaelimmm!~Thaelim@2601:281:c502:25d9:8003:d196:44fb:37fe> has joined ##OpenGL[20:23:43] *** BearishMushroom <BearishMushroom!~BearishMu@82-209-154-59.cust.bredband2.com> has joined ##OpenGL[20:24:39] *** upgrdman <upgrdman!~upgrdman@blender/artist/upgrdman> has joined ##OpenGL[20:25:53] *** davr0s <davr0s!~textual@host86-147-196-14.range86-147.btcentralplus.com> has joined ##OpenGL[20:27:23] *** Thaelimmm <Thaelimmm!~Thaelim@2601:281:c502:25d9:8003:d196:44fb:37fe> has quit IRC (Remote host closed the connection)[20:28:33] *** _borkr_ <_borkr_!~borkr@static130-244.mimer.net> has quit IRC (Ping timeout: 248 seconds)[20:30:09] *** rburkholder <rburkholder!~overonthe@199.68.193.62> has joined ##OpenGL[20:36:18] *** echotangoecho <echotangoecho!~echotango@92-108-127-60.cable.dynamic.v4.ziggo.nl> has quit IRC (Ping timeout: 240 seconds)[20:38:21] *** hellozee <hellozee!~hellozee@45.118.157.211> has quit IRC (Remote host closed the connection)[20:40:33] *** jdashg <jdashg!~jdashg@corp-nat.fw1.untrust.mtv2.mozilla.net> has quit IRC (Ping timeout: 264 seconds)[20:48:55] *** KAHR-Alpha <KAHR-Alpha!~Alpha@LFbn-1-15527-64.w86-225.abo.wanadoo.fr> has joined ##OpenGL[20:50:09] *** Neomex <Neomex!~quassel@net-37-117-5-106.cust.vodafonedsl.it> has quit IRC (Read error: Connection reset by peer)[20:52:20] *** Neomex <Neomex!~quassel@net-37-117-5-106.cust.vodafonedsl.it> has joined ##OpenGL[20:53:41] *** echotangoecho <echotangoecho!~echotango@92-108-127-60.cable.dynamic.v4.ziggo.nl> has joined ##OpenGL[20:58:11] *** acksys_ <acksys_!~fritz@mail12.shutterfly.com> has joined ##OpenGL[21:04:04] *** easzero <easzero!~quassel@85.203.15.20> has joined ##OpenGL[21:07:30] *** echotangoecho <echotangoecho!~echotango@92-108-127-60.cable.dynamic.v4.ziggo.nl> has quit IRC (Ping timeout: 265 seconds)[21:07:37] *** davr0s <davr0s!~textual@host86-147-196-14.range86-147.btcentralplus.com> has quit IRC (Quit: My MacBook Pro has gone to sleep. ZZZzzz…)[21:09:47] *** stefkos <stefkos!~Pawel@pc11-226.chomiczowka.waw.pl> has joined ##OpenGL[21:14:32] *** tambre <tambre!~tambre@7886-b93f-04c4-3840-ab80-8a0a-07d0-2001.dyn.estpak.ee> has quit IRC (Ping timeout: 255 seconds)[21:18:00] *** jdashg <jdashg!~jdashg@corp-nat.fw1.untrust.mtv2.mozilla.net> has joined ##OpenGL[21:18:13] *** Foritus <Foritus!~buggery@cpc91390-watf11-2-0-cust41.15-2.cable.virginm.net> has quit IRC (Read error: Connection reset by peer)[21:19:32] *** Foritus <Foritus!~buggery@cpc91390-watf11-2-0-cust41.15-2.cable.virginm.net> has joined ##OpenGL[21:27:08] *** immibis <immibis!~chatzilla@122-59-200-50.jetstream.xtra.co.nz> has joined ##OpenGL[21:29:51] *** salamanderrake <salamanderrake!~quassel@cpe-65-189-208-111.neo.res.rr.com> has joined ##OpenGL[21:34:22] *** easzero <easzero!~quassel@85.203.15.20> has quit IRC (Remote host closed the connection)[21:35:36] *** slvn_ <slvn_!~slvn_@c2s31-1-78-245-90-111.fbx.proxad.net> has quit IRC (Remote host closed the connection)[21:38:03] <Fr0stBit> Ok guys, what i'm doing wrong? I get only values on the first bin after i read the histogram ssbo back http://lpaste.net/6777308777902243840[21:39:01] *** ville <ville!~ville@188-67-10-70.bb.dnainternet.fi> has joined ##OpenGL[21:43:56] *** echotangoecho <echotangoecho!~echotango@92-108-127-60.cable.dynamic.v4.ziggo.nl> has joined ##OpenGL[21:47:31] *** echotangoecho <echotangoecho!~echotango@92-108-127-60.cable.dynamic.v4.ziggo.nl> has quit IRC (Client Quit)[21:48:35] *** mij <mij!~mijowh@75.97.236.147.res-cmts.leh.ptd.net> has joined ##OpenGL[21:49:41] *** davr0s <davr0s!~textual@host86-147-196-14.range86-147.btcentralplus.com> has joined ##OpenGL[21:59:02] *** Amadiro <Amadiro!~Amadiro@v220100555393154.yourvserver.net> has quit IRC (Ping timeout: 268 seconds)[21:59:31] <chrisf> Fr0stBit: let's see your host code too.[22:00:21] <Fr0stBit> chrisf: http://lpaste.net/1030351568999809024[22:00:24] <Fr0stBit> Here you go[22:00:30] *** ignatiz- <ignatiz-!~scott@82.138.221.18> has joined ##OpenGL[22:00:55] <Xeek> dav1d: whats khrplatform.h for?[22:03:57] *** immibis <immibis!~chatzilla@122-59-200-50.jetstream.xtra.co.nz> has quit IRC (Ping timeout: 264 seconds)[22:07:51] <chrisf> Fr0stBit: i see 2 issues: your glMemoryBarrier in the host code is wrong. the flags you pass are for the access /after/ the barrier. the access before the barrier is assumed to be shader writes. the other issue is your shader only orders memory operations, not execution. you need barrier() calls as well.[22:08:19] *** ShadowIce <ShadowIce!~pyoro@unaffiliated/shadowice-x841044> has quit IRC (Quit: Leaving)[22:09:22] <Fr0stBit> chrisf: I suppose you mean a barrier() call after the initialization stage? Cause the other steps do atomicAdds[22:10:08] <Stragus> After your write to shared memory, yes[22:10:17] <Stragus> You want all the shared memory to be loaded before you continue[22:10:48] <Fr0stBit> chrisf: Also what do i do wrong about host memory barrier?[22:10:56] *** bwidawsk <bwidawsk!~bwidawsk@zangief.bwidawsk.net> has quit IRC (Quit: Always remember, and never forget; I'll be back.)[22:11:25] *** bwidawsk <bwidawsk!~bwidawsk@zangief.bwidawsk.net> has joined ##OpenGL[22:14:05] <Fr0stBit> I suppose i need to do memory barrier with GL_CLIENT_MAPPED_BUFFER_BARRIER_BIT ?[22:15:04] <chrisf> if you're mapping afterward you shouldnt need any bits[22:15:44] *** easzero <easzero!~quassel@85.203.15.20> has joined ##OpenGL[22:16:20] <mangelis> are those memoryBarrierShared() and groupMemoryBarrier() calls right? shouldn't both be calls to same function?[22:17:12] <chrisf> hang on, you need BUFFER_UPDATE_BARRIER_BIT[22:18:13] <chrisf> been forever since i did any readbacks that werent via persistently mapped buffers :)[22:19:01] <chrisf> Fr0stBit: why are you pulling the histogram back to the host?[22:19:54] <chrisf> Fr0stBit: use a CS to massage it into the shape you need for your next pass, but leave it on the GPU.[22:19:56] <Fr0stBit> chrisf: I need to use it for tonemapping[22:20:27] <Fr0stBit> chrisf: Yeah, thats right too. I need to know if the calculations are correct though[22:20:29] <Stragus> Why not compute that on the GPU?[22:20:32] <Stragus> Ah, debugging[22:20:48] <chrisf> for debug only; this is a brutal stall once it works :)[22:21:05] <Fr0stBit> I mean the output is like: {0 <repeats 21 times>, 33555, 31981, 0 <repeats 41 times>}[22:21:08] <Stragus> For debugging results back to CPU, just pure glFinish() everywhere[22:21:11] <Fr0stBit> That is not logical at all[22:21:24] <chrisf> is this after putting correct barriers in?[22:21:32] <Fr0stBit> yep.[22:21:39] <Fr0stBit> Let me update with a current paste[22:21:54] <mangelis> usually it's a good idea to solve problems even though they are irrelevant to the final implementation. the process of solving often forces one to learn a lot of useful information[22:22:03] *** ImQ009_ <ImQ009_!~ImQ009@unaffiliated/imq009> has quit IRC (Quit: Leaving)[22:22:34] <Fr0stBit> Current paste: http://lpaste.net/4489771993045925888[22:24:31] <Fr0stBit> Stragus: tried adding glFinish() after glDispatchCompute(). Still the same result.[22:25:59] <mangelis> what happens if you do final atomicadds with constant values?[22:26:19] <Fr0stBit> mangelis: let me try[22:26:55] <mangelis> what if you just use = operator instead of atomicadd?[22:28:00] <Fr0stBit> mangelis: Would it work? It accumulates the results of each work group, but is it guaranteed that the workgroups are being run in sequence?[22:29:05] <chrisf> no such guarantee[22:29:18] <chrisf> but interesting experiments to determine which step is fouled.[22:29:23] *** xerpi <xerpi!~xerpi@98.red-83-45-194.dynamicip.rima-tde.net> has joined ##OpenGL[22:29:52] <Fr0stBit> Just changed the final atomicAdd call from atomicAdd(values[i], group_histogram[i]) to atomicAdd(values[i], 1)[22:29:58] <mangelis> why would that need atomic operations? the final for loop should run in only one thread[22:30:03] <Fr0stBit> And the histogram i get now is: {256 <repeats 64 times>} wtf[22:30:15] <Fr0stBit> Shouldn't it be all ones?[22:30:26] <chrisf> Fr0stBit: you're dispatching 256 groups.[22:30:50] <Fr0stBit> chrisf: yeah sorry brainfart[22:31:01] <chrisf> ^ is exactly what you expect :)[22:32:05] *** Keniyal <Keniyal!~Keniyal@unaffiliated/keniyal> has quit IRC (Remote host closed the connection)[22:32:27] *** msitton <msitton!~matthew@45.76.237.165> has quit IRC (Quit: leaving)[22:32:56] *** Nimmy <Nimmy!~Nimmy@86.93.7.212> has joined ##OpenGL[22:33:11] <chrisf> groupMemoryBarrier should be called before barrier[22:33:16] <chrisf> not after[22:33:36] <chrisf> read it as "make my writes visible"[22:33:39] <Fr0stBit> Ok new data: I changed the output bin from the histogram_bin_index output to the constant 7[22:33:46] <Fr0stBit> I got: {0, 0, 0, 0, 0, 0, 0, 65536, 0 <repeats 56 times>}[22:33:47] <mangelis> (ah sorry, didn't see that the last atomicadd did something else than just assignment :))[22:33:50] *** slime <slime!~slime73@blk-215-81-93.eastlink.ca> has joined ##OpenGL[22:33:57] *** Kingsquee <Kingsquee!~kingsquee@d108-180-237-219.bchsia.telus.net> has joined ##OpenGL[22:35:19] <Fr0stBit> Shouldn't that 65536 be 1024*1024[22:37:00] <chrisf> it's hard to guess exactly how you've hacked it, paste again[22:37:05] <mangelis> your local size is 16*16=256, you have 16*16=256 workgroups[22:37:51] <mangelis> but things like that might change timing, so the original barrier problem may still persist[22:38:49] <Fr0stBit> mangelis: yeah i just noticed that to. Let me confirm a sec[22:39:51] <Fr0stBit> OK guys, that was it. I now get: {0, 0, 0, 0, 0, 0, 0, 368, 11139, 15393, 14820, 14980, 15791, 16374, 17671, 21134, 26632, 35079, 63887, 54985, 73423, 145019, 115939, 87034, 85536, 65964, 62545, 11934, 5816, 4524, 1350, 435, 0, 7288, 0, 0, 494, 11359, 0,[22:39:53] <Fr0stBit> 0, 0, 12, 0, 434, 12626, 18840, 4450, 0, 17886, 5972, 882, 155, 0, 0, 185, 221, 0, 0, 0, 0, 0, 0, 0, 0}[22:40:08] <chrisf> Fr0stBit: the barrier being fouled was the last issue?[22:40:15] <Fr0stBit> That is sensible knowing that the input luminance range is about 0.18~1.8[22:40:37] <Fr0stBit> No, the lesser workgroups -.-[22:40:57] <Fr0stBit> The next step is how do i make it run in less than 60ms[22:41:09] <Fr0stBit> Lets remove the glFinish calls first[22:41:57] <chrisf> can only really measure its cost without the readback[22:42:26] <chrisf> (that MapBufferRange may as well be a Finish)[22:42:48] <Fr0stBit> Ok, removed the readbacks and it dropped to 8ms[22:43:18] <chrisf> still brutal :)[22:43:25] <Fr0stBit> Yep.[22:43:37] <Fr0stBit> Are all these barrier() needed?[22:44:01] <chrisf> assuming your final version has 2, yes[22:44:26] <Fr0stBit> chrisf: If i have barrier(), arent memoryBarrierShared uneccessary?[22:44:44] <mangelis> (and if memoryBarrierShared() is used, why is barrier() necessary?)[22:45:13] *** easzero <easzero!~quassel@85.203.15.20> has quit IRC (Read error: Connection reset by peer)[22:45:24] <chrisf> execution barrier and memory ordering are separate.[22:46:26] *** toblorone <toblorone!~toblorone@162.246.216.28> has joined ##OpenGL[22:46:42] <chrisf> consider how this executes on a 64-wide machine like a GCN[22:47:18] <chrisf> your 16x16 local group is 4 wavefronts[22:48:56] <Fr0stBit> Also what if i need to zero out the final ssbo values before the final add? Should i do it in the first thread of the first group?[22:49:49] <chrisf> perhaps, or you might do it in another earlier CS dispatch[22:50:30] <chrisf> actually, it's no good to do it in the first group -- you cant guarantee the ordering of that group wrt the other groups.[22:50:44] <Stragus> Definitely in a previous kernel launch[22:51:18] <mangelis> chrisf: doesn't barrier() after zeroing guarantee the ordering?[22:51:43] <Stragus> barrier() is only a sync point within the workground, not for the whole launch[22:51:56] <Stragus> workgroup*[22:52:15] <chrisf> mangelis: it has to be that way -- there's no requirement that all the threads in the whole launch even /fit/ on the device simultaneously :)[22:52:46] *** krystcich_ <krystcich_!~krystcich@178235146015.dynamic-zab-01.vectranet.pl> has quit IRC (Quit: Going offline, see ya! (www.adiirc.com))[22:53:05] <Fr0stBit> Could i add it along with the group's local array zeroing? I mean sync shouldn't be an issue if all the threads in all the groups just want to write a zero in the final array[22:53:25] <Stragus> There is no other way than a previous launch[22:53:38] <Stragus> Workgroups can run in any order without any synchronization betweem them[22:54:07] <chrisf> well, you could ask the command processor to do it (glClearBufferData).[22:54:33] <chrisf> YMMV on how the driver chooses to implement that :)[22:55:17] *** Zero_Walker <Zero_Walker!~Zerowalke@78-67-154-24-no268.tbcn.telia.com> has quit IRC (Quit: Leaving)[22:55:23] <Fr0stBit> So the best solution is to make another cs with local size x 64 and a single workgroup launch, that clears out the buffer?[22:55:56] <chrisf> cant say best without measuring, but it's good.[22:56:38] <chrisf> dont forget the glMemoryBarrier between the two[22:56:41] *** stefkos <stefkos!~Pawel@pc11-226.chomiczowka.waw.pl> has quit IRC (Quit: Going offline, see ya! (www.adiirc.com))[22:57:12] *** asecretcat <asecretcat!~allisonze@pool-72-77-42-97.pitbpa.fios.verizon.net> has quit IRC (Quit: let us connect our intestines and mutually digest)[22:57:23] *** Murii <Murii!~Murii@79.113.205.114> has quit IRC (Quit: WeeChat 1.4)[22:57:29] <Stragus> chrisf, I'm a bit confused by GLSL's memoryBarrierShared(). Am I correct to assume it's implied by CUDA's __syncthreads(), which would otherwise be the equivalent of GLSL's barrier()? GLSL added memorySharedBarrier() for... other GPUs which could somehow have barrier() without visibly updated shared memory?[22:58:27] <chrisf> Stragus: on gen, waiting on a barrier and flushing shared writes are separate operations[22:58:29] <Fr0stBit> chrisf: What barrier should i set? I suppose this time GL_SHADER_STORAGE_BARRIER_BIT?[22:58:37] <chrisf> Fr0stBit: sounds correct[22:59:03] <Stragus> chrisf: Interesting, that explains it then. There's no flush of shared memory on Nvidia[22:59:40] *** BitPuffin <BitPuffin!~isak@c83-248-113-111.bredband.comhem.se> has quit IRC (Remote host closed the connection)[22:59:56] <chrisf> Stragus: where does nvidia shared live? on gen, it's done by stealing ways from the gfx L3$.[23:00:00] *** Silwer <Silwer!~SilwerKop@dslb-188-104-038-134.188.104.pools.vodafone-ip.de> has quit IRC (Quit: Leaving)[23:00:17] <mangelis> okay. interesting conversation, thanks[23:00:50] <Stragus> On Nvidia, it's repurposed L1. It's blazzingly fast[23:01:33] <chrisf> neat :)[23:01:52] <chrisf> mangelis: apologies for wandering off into gpu guts :)[23:02:17] *** asecretcat <asecretcat!~allisonze@pool-72-77-42-97.pitbpa.fios.verizon.net> has joined ##OpenGL[23:02:42] <chrisf> Stragus: does that imply groups cant span SMs?[23:02:44] <Stragus> Amazingly fast on-chip kilobytes of memory is wonderful for many algorithms, I thought AMD would also have some dedicated chunk for shared memory[23:03:06] <Stragus> A single group resides on a single SM on Nvidia[23:03:28] <chrisf> GCN has fast ways to do this too, but im not an expert on their architecture :)[23:03:49] *** madsy <madsy!~madsy@fu/coder/madsy> has quit IRC (Quit: leaving)[23:03:51] <Stragus> It's like having 48kb of memory that's as fast as registers[23:04:11] *** madsy <madsy!~madsy@cm-84.215.249.56.getinternet.no> has joined ##OpenGL[23:04:12] *** madsy <madsy!~madsy@cm-84.215.249.56.getinternet.no> has quit IRC (Changing host)[23:04:12] *** madsy <madsy!~madsy@fu/coder/madsy> has joined ##OpenGL[23:04:19] <Stragus> (well, almost as fast, there are complicated concerns regarding shared memory bank conflict access)[23:08:16] <Fr0stBit> Aren't 7ms for the whole operation a little to many?[23:08:51] <Stragus> For 65536 threads doing about nothing? What are you measuring?[23:09:03] <chrisf> Fr0stBit: half a frame? yes, this is proabbly way too slow.[23:09:12] <Fr0stBit> http://lpaste.net/670762523133214720[23:09:28] <Fr0stBit> I get 7ms increase when i get eyeadapt_luminance_hist to run[23:10:29] *** davr0s <davr0s!~textual@host86-147-196-14.range86-147.btcentralplus.com> has quit IRC (Quit: My MacBook Pro has gone to sleep. ZZZzzz…)[23:12:37] <Stragus> You should always avoid all glGet[23:13:13] <chrisf> upsets the threaded drivers :)[23:13:31] <Fr0stBit> Stragus: Ok, i suppose they are driver calls, but how much harm can they cost from the 7ms?[23:13:40] <chrisf> Fr0stBit: stick some timestamp queries around the actual dispatches[23:15:17] <chrisf> (or get your gpu profiler to do it)[23:15:37] <Fr0stBit> chrisf: I ain't got any gpu profiler. Can you suggest one?[23:16:18] <chrisf> Fr0stBit: what are you running this on?[23:16:36] <Stragus> Why do you need that blit anyway?[23:16:41] <Fr0stBit> chrisf: Linux[23:16:47] <chrisf> Fr0stBit: the GPU :)[23:17:08] <Fr0stBit> Stragus: in order to make the buffer a round 1024x1024[23:17:17] <Fr0stBit> chrisf: NVIDIA[23:17:22] <Fr0stBit> chrisf: GT640[23:17:26] <chrisf> use nsight then[23:17:26] <Stragus> That's rather wasteful for nothing[23:17:52] <Stragus> Read the real FBO just once, adjust your workgroup counts and discard threads out of bounds[23:18:44] <chrisf> 640 is pretty miserable, but not THAT slow.[23:19:14] <Stragus> That's one pretty old and low-end GPU...[23:20:12] <chrisf> Stragus: what does kepler like for group sizes?[23:20:39] <Stragus> 256 is good[23:20:51] <chrisf> Stragus: if Fr0stBit were to match to the SM size, would the barriers all evaporate for a nice win?[23:21:27] <Stragus> The SM size is multiple warps of 32 threads running concurrently, pausing whenever they feel like it[23:21:51] <Stragus> 256 is a good number to allow multiple resident blocks on a same SM[23:22:00] <Stragus> Err, block is CUDA speak, workgroup[23:22:30] <chrisf> is kepler likely to like unrolling this a few times?[23:23:09] <Stragus> The or (uint i = 0; i < NUM_HISTOGRAM_BINS; ++i) only on thread 0 is pretty bad[23:23:14] <Fr0stBit> Stragus: No blit version: down to 5ms![23:23:20] <Stragus> Each thread should add a separate value[23:23:23] <Fr0stBit> (Still seems pretty slow though...)[23:23:53] <Stragus> Same for initial clear of shared memory[23:24:17] *** madsy <madsy!~madsy@fu/coder/madsy> has quit IRC (Quit: leaving)[23:24:58] <chrisf> work out the speed of light number for just memory in / memory out.[23:25:17] <chrisf> suspect it's <1ms even on a 640[23:25:52] *** madsy <madsy!~madsy@cm-84.215.249.56.getinternet.no> has joined ##OpenGL[23:25:52] *** madsy <madsy!~madsy@cm-84.215.249.56.getinternet.no> has quit IRC (Changing host)[23:25:52] *** madsy <madsy!~madsy@fu/coder/madsy> has joined ##OpenGL[23:26:03] <Fr0stBit> Stragus: you mean, i should not make the first thread do all the clears?[23:26:24] <Stragus> Fr0stBit: First thread clears value 0, second thread clears value 1, etc.[23:26:32] <Fr0stBit> Stragus: also what do you mean by 'each thread should add a spearate value'[23:26:54] <Stragus> Same thing, have all threads perform an atomic add, not a single thread perform all of them[23:27:59] <chrisf> or at least #bins threads do it[23:28:29] <chrisf> Stragus: is there trouble on kepler with atomics contention?[23:28:59] <Stragus> Kind of, shared memory atomics are rather bad[23:29:59] <chrisf> and/or with a single atomic insn that hits 64 different addresses[23:30:23] <chrisf> i know gen's perf behavior here is pretty nasty unless you get it just right[23:30:52] *** madsy <madsy!~madsy@fu/coder/madsy> has quit IRC (Quit: leaving)[23:31:14] <Fr0stBit> You mean something like this?[23:31:15] <Fr0stBit> http://lpaste.net/3469071164020621312[23:31:30] *** madsy <madsy!~madsy@cm-84.215.249.56.getinternet.no> has joined ##OpenGL[23:31:30] *** madsy <madsy!~madsy@cm-84.215.249.56.getinternet.no> has quit IRC (Changing host)[23:31:30] *** madsy <madsy!~madsy@fu/coder/madsy> has joined ##OpenGL[23:31:46] <Fr0stBit> So far on 4.5~5ms[23:32:29] <Stragus> That's better[23:33:27] *** madsy <madsy!~madsy@fu/coder/madsy> has quit IRC (Client Quit)[23:33:59] <Stragus> Next, I would try to put more work per thread, accumulate more pixels in order to reduce the count of global atomics[23:34:51] <Fr0stBit> Stragus: can i somehow configure the local size from outside the shader?[23:35:17] <Stragus> You can in CUDA... I'll let chrisf answer that one for GLSL[23:35:58] * Stragus has been writing CUDA for 10 years, but only wrote about 100 lines of GLSL compute shaders :|[23:36:54] <Spark> christ, it has been 10 years hasn't it[23:36:56] * Spark feels old[23:37:21] <Fr0stBit> Stragus: increased local size to 32 from 16, got from 4.5ms up to 7ms[23:37:24] <Fr0stBit> ...[23:37:44] <Stragus> 1024 threads per block is way too much, too few resident blocks per SM[23:37:59] <Stragus> Your 16x16 blocks are good[23:38:48] <Fr0stBit> Stragus: replaced imageLoad with const vec(0.0). Got to 2.5ms[23:39:24] <Stragus> Okay, that can be misleading because the compiler can be smart enough to remove your += 0.0;[23:40:22] <Fr0stBit> Stragus: what if i make it another constant, like 5.3[23:40:31] <Stragus> That's better[23:41:14] <Stragus> Get rid of the FBO blit, read the real FBO directly. Blocks of 16x16, as many blocks as you need, discard threads that are out of range in edge blocks[23:41:55] <Stragus> Next, perform more work per thread to reduce global atomics, aim for something like 16384 concurrent threads (rather than over a million)[23:42:50] <Fr0stBit> Stragus: I got everything but the last[23:43:14] <Fr0stBit> How do i 'throttle' the threads? Multiple dispatch calls?[23:43:17] *** moser <moser!~moser@223.73.115.211> has joined ##OpenGL[23:43:35] <Stragus> Please don't do that, why would you do that?[23:44:47] <Fr0stBit> you said i should aim for 16384 concurrent threads rather than over a million[23:44:51] <chrisf> Fr0stBit: you can configure it from outside -- ARB_compute_variable_group_size[23:45:01] <Fr0stBit> chrisf: Fantastic![23:45:01] <Stragus> Okay, let me rephrase, 16384 total threads for the job[23:45:04] <chrisf> you shouldnt need it though. find the right point.[23:45:14] * Stragus believes 16x16 is the right point[23:45:41] <chrisf> my gut is telling me 16x16, and each doing 4-8px[23:45:51] <Fr0stBit> Also weird thing. I removed the imageLoad call and replaced it with a vec3(6.2) constant (random) and when i run it it hangs the whole desktop[23:45:57] <Fr0stBit> Like it gets too much work[23:46:10] *** jdashg <jdashg!~jdashg@corp-nat.fw1.untrust.mtv2.mozilla.net> has quit IRC (Ping timeout: 240 seconds)[23:46:29] *** l3dx <l3dx!uid2250@gateway/web/irccloud.com/x-cyqywgmluqsffhqq> has quit IRC (Quit: Connection closed for inactivity)[23:46:32] <Fr0stBit> Ah it must be some kind of overflow in my buckets, my bad[23:47:14] <Fr0stBit> Ok so without the imageLoad() i get from 4.5ms to 2.0ms[23:48:14] <chrisf> can go as fast as you like if it doesnt have to work[23:49:24] <chrisf> let's see what you've got now[23:49:25] <Fr0stBit> chrisf: Just trying to find out how much time each thing eats[23:50:15] <Fr0stBit> chrisf: http://lpaste.net/3469071164020621312[23:51:08] <chrisf> the game now is to get more work into each thread, and the total number of threads down[23:51:36] <Fr0stBit> chrisf: I see.. Like i should do more imageLoads in each thread?[23:51:40] <chrisf> yep[23:52:53] <Fr0stBit> chrisf: Should i get the number of threads per group down? Or the number of groups[23:52:58] <chrisf> number of groups[23:54:03] <chrisf> Stragus: is imageSize reasonable on kepler?[23:54:21] *** madsy <madsy!~madsy@cm-84.215.249.56.getinternet.no> has joined ##OpenGL[23:54:21] *** madsy <madsy!~madsy@cm-84.215.249.56.getinternet.no> has quit IRC (Changing host)[23:54:21] *** madsy <madsy!~madsy@fu/coder/madsy> has joined ##OpenGL[23:54:44] <Stragus> I have no idea, I never trust these functions. I would pass it explicitely[23:55:15] <chrisf> on some bindless hw it's "free" if you're already fetching the descriptor[23:55:26] <Stragus> Do these imageLoad() go through the texture cache?...[23:55:42] <Stragus> On Kepler, direct memory loads are *SLOW*, everything must go through the texture cache to have decent performance[23:56:06] <Stragus> Perhaps a texture() would be safer than imageLoad()[23:56:11] <chrisf> texelFetch?[23:56:15] <Stragus> That, yes[23:57:06] *** Asu <Asu!~sdelang@80.12.63.143> has quit IRC (Remote host closed the connection)[23:57:15] <Stragus> On Kepler, direct memory loads are basically 2.5x as slow as texture fetches