July 21, 2019 
< | 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 | 29 | 30 | 31 | >
July 21, 2019 [00:23:45] *** Deluxe <Deluxe!~Deluxe@212.4.150.151> has quit IRC (Read error: Connection reset by peer)[01:09:06] *** LunarJetman <LunarJetman!LunarJetma@176.248.197.112> has quit IRC (Quit: LunarJetman)[01:19:16] *** felipealmeida <felipealmeida!~user@191.191.10.198> has quit IRC (Remote host closed the connection)[02:57:10] *** felipealmeida <felipealmeida!~user@191.191.10.198> has joined ##vulkan[02:57:40] <felipealmeida> hi, I'm passing a fence to VkQueueSubmit[02:57:55] <felipealmeida> but after a few iterations (~5-20) loops[02:58:03] <felipealmeida> the wait on fence never finishes[02:58:14] <felipealmeida> I'm waiting just after the queue submit[04:01:33] <felipealmeida> hi, I have a swapChain with two images. I'm using two commandbuffers pre-recorded per framebuffer[04:02:25] <felipealmeida> I'm getting messages that my commandbuffer is not recorded with VK_COMMAND_BUFFER_USAGE_SIMULTANEOUS_USE_BIT, however I can't see how it can be used simultaneously[04:02:43] <felipealmeida> if I acquire image 0, I use commandbuffer[0], if I acquire image 1 I use commandbuffer[1][04:04:22] <Ralith> you're probably not synchronizing access to them correctly[04:05:00] <felipealmeida> well, but I can only acquire a image that is no longer in use, right?[04:05:15] <felipealmeida> so I can only acquire a image where the render has already finished, don't I?[04:05:55] <felipealmeida> or can the rendering still be happening after the second acquire of the same image?[04:06:31] <Ralith> I strongly encourage you to read the spec for vkAcquireNextImageKHR[04:06:34] <felipealmeida> I do use the semaphores from acquire to queue submit[04:07:12] <felipealmeida> well, I have a few times, but I'll look again[04:08:41] <felipealmeida> well, it is not really the most descriptive function in the spec[04:13:35] <felipealmeida> can't find anything...[04:26:59] *** ratchetfreak <ratchetfreak!~ratchetfr@ptr-82s3g7l9r42569r41f4.18120a2.ip6.access.telenet.be> has quit IRC (Ping timeout: 252 seconds)[04:40:22] *** felipealmeida <felipealmeida!~user@191.191.10.198> has quit IRC (Remote host closed the connection)[04:41:54] *** penguin42 <penguin42!~dg@cpc109017-salf6-2-0-cust428.10-2.cable.virginm.net> has quit IRC (Remote host closed the connection)[04:49:53] *** saint_go <saint_go!~textual@135-23-210-183.cpe.pppoe.ca> has joined ##vulkan[05:30:37] *** tristianc <tristianc!~tristianc@pool-72-83-128-193.washdc.fios.verizon.net> has joined ##vulkan[07:54:36] *** wildlander <wildlander!~wildlande@unaffiliated/wildlander> has joined ##vulkan[08:07:46] *** salamanderrake <salamanderrake!~quassel@2605:a000:122a:2139:3dbe:c8d6:f55b:7577> has quit IRC (Ping timeout: 276 seconds)[08:26:06] *** ciaala <ciaala!~crypt@2a02:120b:2c1f:5f40:6ef0:49ff:feee:4777> has joined ##vulkan[08:39:13] *** wildlander <wildlander!~wildlande@unaffiliated/wildlander> has quit IRC (Ping timeout: 245 seconds)[09:06:50] *** wildlander <wildlander!~wildlande@unaffiliated/wildlander> has joined ##vulkan[09:42:53] *** slime <slime!~slime73@24.215.81.93> has quit IRC (Quit: This computer has gone to sleep)[09:55:44] *** Deluxe <Deluxe!~Deluxe@212.4.150.151> has joined ##vulkan[10:25:16] *** ImQ009 <ImQ009!~ImQ009@unaffiliated/imq009> has joined ##vulkan[11:19:08] *** bpmedley <bpmedley!~bpm@2600:1700:eb20:6050:8ee:42b5:38bc:425b> has joined ##vulkan[11:22:13] *** ciaala <ciaala!~crypt@2a02:120b:2c1f:5f40:6ef0:49ff:feee:4777> has quit IRC (Quit: Konversation terminated!)[11:27:19] *** bpmedley <bpmedley!~bpm@2600:1700:eb20:6050:8ee:42b5:38bc:425b> has quit IRC (Ping timeout: 276 seconds)[12:52:11] *** ratchetfreak <ratchetfreak!~ratchetfr@ptr-82s3g7l9r42569r41f4.18120a2.ip6.access.telenet.be> has joined ##vulkan[13:37:25] *** penguin42 <penguin42!~dg@cpc109017-salf6-2-0-cust428.10-2.cable.virginm.net> has joined ##vulkan[14:56:21] *** slime <slime!~slime73@24.215.81.93> has joined ##vulkan[14:56:43] *** fling <fling!~fling@fsf/member/fling> has quit IRC (Ping timeout: 245 seconds)[15:33:56] *** ryp <ryp!ryp@gateway/vpn/privateinternetaccess/ryp> has joined ##vulkan[15:49:29] *** LunarJetman <LunarJetman!LunarJetma@176.248.197.112> has joined ##vulkan[15:50:49] *** ciaala <ciaala!~crypt@2a02:120b:2c1f:5f40:6ef0:49ff:feee:4777> has joined ##vulkan[15:54:12] *** ryp <ryp!ryp@gateway/vpn/privateinternetaccess/ryp> has quit IRC (Ping timeout: 245 seconds)[15:56:23] *** RicardasSim <RicardasSim!~RicardasS@78-63-172-44.static.zebra.lt> has quit IRC (Quit: Leaving)[16:16:18] *** fling <fling!~fling@fsf/member/fling> has joined ##vulkan[16:21:30] *** ryp <ryp!ryp@gateway/vpn/privateinternetaccess/ryp> has joined ##vulkan[17:15:22] *** LunarJetman <LunarJetman!LunarJetma@176.248.197.112> has left ##vulkan[17:30:50] *** ciaala <ciaala!~crypt@2a02:120b:2c1f:5f40:6ef0:49ff:feee:4777> has quit IRC (Quit: Konversation terminated!)[17:48:22] *** ryp <ryp!ryp@gateway/vpn/privateinternetaccess/ryp> has quit IRC (Ping timeout: 245 seconds)[17:52:30] *** penguin42 <penguin42!~dg@cpc109017-salf6-2-0-cust428.10-2.cable.virginm.net> has quit IRC (Read error: Connection timed out)[17:52:56] *** penguin42 <penguin42!~dg@cpc109017-salf6-2-0-cust428.10-2.cable.virginm.net> has joined ##vulkan[18:03:21] *** Deluxe_ <Deluxe_!~Deluxe@212.4.150.151> has joined ##vulkan[18:22:02] *** pa <pa!~pa@unaffiliated/pa> has quit IRC (Ping timeout: 272 seconds)[18:23:43] *** pa <pa!~pa@unaffiliated/pa> has joined ##vulkan[18:49:35] *** salamanderrake <salamanderrake!~quassel@2605:a000:122a:2139:50b5:d668:b233:7be2> has joined ##vulkan[19:40:10] *** bpmedley <bpmedley!~bpm@2600:1700:eb20:6050:41bf:c680:1ac1:cc4f> has joined ##vulkan[19:46:41] *** bpmedley <bpmedley!~bpm@2600:1700:eb20:6050:41bf:c680:1ac1:cc4f> has quit IRC (Quit: Textual IRC Client: www.textualapp.com)[19:47:58] *** bpmedley <bpmedley!~bpm@2600:1700:eb20:6050:1003:60b:7744:40ad> has joined ##vulkan[19:52:39] *** Deluxe_ <Deluxe_!~Deluxe@212.4.150.151> has quit IRC (Remote host closed the connection)[19:54:15] *** Deluxe_ <Deluxe_!~Deluxe@212.4.150.151> has joined ##vulkan[19:54:49] *** Deluxe_ <Deluxe_!~Deluxe@212.4.150.151> has quit IRC (Remote host closed the connection)[20:30:21] *** soulz <soulz!~soulz@unaffiliated/soulz> has quit IRC (Quit: leaving)[20:51:04] *** penguin42 <penguin42!~dg@cpc109017-salf6-2-0-cust428.10-2.cable.virginm.net> has quit IRC (Remote host closed the connection)[20:52:57] *** soulz <soulz!~soulz@unaffiliated/soulz> has joined ##vulkan[21:34:49] *** wildlander <wildlander!~wildlande@unaffiliated/wildlander> has quit IRC (Quit: Konversation terminated!)[21:34:51] *** mandeep <mandeep!mandeep@gateway/vpn/privateinternetaccess/mandeepb> has joined ##vulkan[21:37:38] *** felipealmeida <felipealmeida!~user@191.191.10.198> has joined ##vulkan[22:11:23] *** ImQ009 <ImQ009!~ImQ009@unaffiliated/imq009> has quit IRC (Quit: Leaving)[22:16:12] <Yatekii> folks how would you transmit a huge array of tile transforms to the vertex shader? uniform buffer seems to be bad ... using a texture has been an advice but that seems like a hack to me[22:18:17] *** ryp <ryp!ryp@gateway/vpn/privateinternetaccess/ryp> has joined ##vulkan[22:18:36] <derhass> what's bad about UBOs? size limit?[22:35:53] *** Deluxe <Deluxe!~Deluxe@212.4.150.151> has quit IRC (Read error: Connection reset by peer)[22:36:45] *** Deluxe <Deluxe!~Deluxe@212.4.150.151> has joined ##vulkan[22:37:32] *** Deluxe <Deluxe!~Deluxe@212.4.150.151> has quit IRC (Remote host closed the connection)[22:38:08] <Yatekii> derhass: well on the GTX970 I am good. all works nice and dandy. on the iGPU, after like 28transforms I just read all 0 for the transforms, so I guess it's hitting the size limit. but the buffer contents are actually correct. just the shader does not read porperly[22:38:41] <bnieuwenhuizen> igpu being intel? linux or windows?[22:39:45] <turol> Yatekii: SSBO[22:40:07] <Yatekii> bnieuwenhuizen: iGPU being integrated GPU ofc[22:40:15] <Yatekii> and yes intel, sorry[22:40:22] <Yatekii> HD 620[22:40:25] <Yatekii> linux[22:41:28] <bnieuwenhuizen> Per http://vulkan.gpuinfo.org/displayreport.php?id=6241#limits I think UBOs are supported up to 128M?[22:42:06] <bnieuwenhuizen> If 29 transforms are more than 128 megabyte, that is quite a lot[22:44:27] <bnieuwenhuizen> Yatekii: please check the range limit above. Given that you are reading all 0, I think it is more likely that you specify the wrong UBO size when binding to the descriptor, or maybe in the shader[22:44:46] <Yatekii> bnieuwenhuizen: I am 100% not specifying the wrong buffer size[22:46:06] <Yatekii> the buffer size in the shader is bigger than it actually is (because I would have to know at compiletime how many items I have, which I ofc don't know, so I just make it a max size)[22:46:21] <Yatekii> and the size on the CPU is corect. if it wasn't it would not contain the correct data[22:47:16] <Yatekii> I mean it has to be some stupid GPU related thing[22:47:25] <Yatekii> otherwise it wouldn't run on the GTX[22:47:54] *** LunarJetman <LunarJetman!LunarJetma@176.248.197.112> has joined ##vulkan[22:56:27] <Yatekii> yeah, changing to SSBO doesn't change an inch :([22:58:08] <bnieuwenhuizen> Yatekii: remember you also have to set the size in vkUpdateDescriptorSets[23:01:19] <Yatekii> bnieuwenhuizen: wgpu is handling that part for me[23:01:36] <Yatekii> I can see that the size is correct in renderdoc ;)[23:03:41] <fazias> just because it runs on one vendor, doesn't mean it's correct.[23:04:29] <mirv_> I don't suppose anyone has run across problems calling vkGetPhysicalDeviceImageFormatProperties?[23:04:37] <bnieuwenhuizen> Have you run the validation layers?[23:04:48] <mirv_> of course, it's spitting out invalid physical device handle[23:05:07] <mirv_> but it's most definitely not invalid (I called it just before on something else, and was fine)[23:05:09] <bnieuwenhuizen> Reason I still suspect it is your app is that I cannot believe any games works if UBOs are broken[23:05:19] <bnieuwenhuizen> mirv_: sorry was not meant for you[23:05:28] <mirv_> bnieuwenhuizen: my bad, I realised too late[23:05:35] <mirv_> and should read up more often[23:05:46] <bnieuwenhuizen> mirv_: what is your issue with the function?[23:06:42] <mirv_> segfault, after validation error about invalid physical device handler[23:07:35] <mirv_> I'll keep on poking[23:07:49] <fazias> doesn't the vulkan sdk come with a program which basically calls that function?[23:07:58] <bnieuwenhuizen> yeah, vulkaninfo[23:08:06] <fazias> if it works, you have some problem in your app[23:08:36] <fazias> if it doesn't, sounds like strange ICD configuration. maybe check that with the... there was app for that also[23:09:03] <fazias> vkvia[23:09:19] <fazias> is good to run for checking if you got sane environment[23:09:36] <bnieuwenhuizen> wow, did not know about that tool, thanks![23:10:01] <fazias> it doesn't do as much as you would hope, but it does bunch of tests and makes some sort of html report[23:11:50] <Yatekii> bnieuwenhuizen: I agree, still, I am intrigued by the fact that it runs on the GTX970.[23:12:04] <Yatekii> bnieuwenhuizen: I just installed the validation layers. or rather it's compiling them[23:12:11] <Yatekii> will need to figure how to use them :/[23:12:40] <bnieuwenhuizen> Yatekii: would not surprise me. Not every driver is consistent in doing bounds checks[23:12:50] <Yatekii> kk[23:13:07] <Yatekii> then I wonder why renderdoc shows the right number :/[23:13:08] <fazias> validation layer is your best friend catching your own mistakes[23:13:17] <Yatekii> anyways, i'm trying to run the validation layers[23:20:21] <Yatekii> lol endless validation errors[23:20:26] * Yatekii guilty[23:20:31] <Yatekii> but for the stencil actually[23:20:34] <Yatekii> imma patch that up[23:22:07] <mirv_> pro tip: don't try get an instance function pointer using device function pointer querying[23:22:22] <mirv_> (I just found my error)[23:23:08] <bnieuwenhuizen> mirv_: validation layers did not catch that?[23:23:55] <bnieuwenhuizen> oh, might actually be hard to check that[23:24:45] <mirv_> yeah[23:25:20] <mirv_> well, to be fair, I should write less shitty code and check for a nullptr return[23:29:46] <fazias> \o/