March 24, 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 | >
March 24, 2019 [03:08:16] *** LunarJetman <LunarJetman!LunarJetma@5ec16e1a.skybroadband.com> has quit IRC (Ping timeout: 255 seconds)[03:23:47] *** ratchet_freak <ratchet_freak!~ratchetfr@ptr-82s3g7l9r42569r41f4.18120a2.ip6.access.telenet.be> has quit IRC (Ping timeout: 240 seconds)[04:30:30] *** penguin42 <penguin42!~dg@cpc109017-salf6-2-0-cust428.10-2.cable.virginm.net> has quit IRC (Remote host closed the connection)[04:40:59] *** davr0s <davr0s!~textual@host81-147-72-11.range81-147.btcentralplus.com> has quit IRC (Quit: My MacBook Pro has gone to sleep. ZZZzzz…)[07:18:22] *** RicardasSim <RicardasSim!~RicardasS@78-63-172-44.static.zebra.lt> has quit IRC (Remote host closed the connection)[07:27:41] *** RicardasSim <RicardasSim!~RicardasS@78-63-172-44.static.zebra.lt> has joined ##vulkan[09:21:11] *** ratchet_freak <ratchet_freak!~ratchetfr@ptr-82s3g7l9r42569r41f4.18120a2.ip6.access.telenet.be> has joined ##vulkan[09:21:16] *** alexge50 <alexge50!~alexge50@unaffiliated/alexge50> has joined ##vulkan[09:25:52] *** ratchet_freak <ratchet_freak!~ratchetfr@ptr-82s3g7l9r42569r41f4.18120a2.ip6.access.telenet.be> has quit IRC (Ping timeout: 257 seconds)[10:23:52] *** glYoda <glYoda!~MTLYoda@c-73-25-27-206.hsd1.or.comcast.net> has quit IRC (Ping timeout: 255 seconds)[10:56:34] *** ratchet_freak <ratchet_freak!~ratchetfr@ptr-82s3g7l9r42569r41f4.18120a2.ip6.access.telenet.be> has joined ##vulkan[11:10:40] *** Deluxe <Deluxe!~Deluxe@212.4.150.151> has joined ##vulkan[11:17:05] *** emersont1 <emersont1!~emersont1@2a00:23c5:2e19:c400:977:f486:fb4d:791f> has joined ##vulkan[11:41:11] *** emersont1 <emersont1!~emersont1@2a00:23c5:2e19:c400:977:f486:fb4d:791f> has quit IRC (Remote host closed the connection)[12:16:39] *** borkr <borkr!~borkr@static130-244.mimer.net> has joined ##vulkan[12:18:47] *** borkr <borkr!~borkr@static130-244.mimer.net> has quit IRC (Remote host closed the connection)[12:24:26] *** emersont1 <emersont1!~emersont1@host81-155-32-5.range81-155.btcentralplus.com> has joined ##vulkan[13:25:12] *** borkr <borkr!~borkr@static130-244.mimer.net> has joined ##vulkan[13:25:21] *** RicardasSim <RicardasSim!~RicardasS@78-63-172-44.static.zebra.lt> has quit IRC (Quit: Leaving)[13:32:47] *** davr0s <davr0s!~textual@host81-147-72-11.range81-147.btcentralplus.com> has joined ##vulkan[13:53:46] *** penguin42 <penguin42!~dg@cpc109017-salf6-2-0-cust428.10-2.cable.virginm.net> has joined ##vulkan[14:32:42] *** LunarJetman <LunarJetman!LunarJetma@5ec16e1a.skybroadband.com> has joined ##vulkan[14:35:07] *** Fats <Fats!fats@gateway/vpn/privateinternetaccess/fats> has quit IRC (Remote host closed the connection)[14:36:15] *** Fats <Fats!fats@gateway/vpn/privateinternetaccess/fats> has joined ##vulkan[14:37:35] *** slime <slime!~slime73@156.57.127.44> has joined ##vulkan[15:03:15] *** davr0s <davr0s!~textual@host81-147-72-11.range81-147.btcentralplus.com> has quit IRC (Quit: My MacBook Pro has gone to sleep. ZZZzzz…)[15:25:21] *** Deluxe <Deluxe!~Deluxe@212.4.150.151> has quit IRC (Ping timeout: 244 seconds)[15:55:40] *** davr0s <davr0s!~textual@host81-147-72-11.range81-147.btcentralplus.com> has joined ##vulkan[16:06:26] *** davr0s <davr0s!~textual@host81-147-72-11.range81-147.btcentralplus.com> has quit IRC (Quit: My MacBook Pro has gone to sleep. ZZZzzz…)[17:00:40] *** RicardasSim <RicardasSim!~RicardasS@78-63-172-44.static.zebra.lt> has joined ##vulkan[17:29:12] *** mluser-home <mluser-home!~mluser-ho@ip68-0-67-199.tu.ok.cox.net> has joined ##vulkan[17:41:22] *** davr0s <davr0s!~textual@host81-147-72-11.range81-147.btcentralplus.com> has joined ##vulkan[17:57:12] *** cocholate <cocholate!~cppfag@179.180.81.142> has joined ##vulkan[18:10:51] *** emersont1 <emersont1!~emersont1@host81-155-32-5.range81-155.btcentralplus.com> has quit IRC (Remote host closed the connection)[18:11:19] *** emersont1 <emersont1!~emersont1@host81-155-32-5.range81-155.btcentralplus.com> has joined ##vulkan[18:19:31] *** cocholate <cocholate!~cppfag@179.180.81.142> has quit IRC (Quit: cocholate)[18:48:27] *** emersont1 <emersont1!~emersont1@host81-155-32-5.range81-155.btcentralplus.com> has quit IRC (Remote host closed the connection)[18:54:23] *** glYoda <glYoda!~MTLYoda@c-73-25-27-206.hsd1.or.comcast.net> has joined ##vulkan[19:13:23] *** slime <slime!~slime73@156.57.127.44> has quit IRC (Quit: This computer has gone to sleep)[19:53:17] *** davr0s <davr0s!~textual@host81-147-72-11.range81-147.btcentralplus.com> has quit IRC (Quit: My MacBook Pro has gone to sleep. ZZZzzz…)[20:09:00] *** borkr <borkr!~borkr@static130-244.mimer.net> has quit IRC (Ping timeout: 244 seconds)[20:11:06] *** marcheu <marcheu!~marcheu@annarchy.freedesktop.org> has quit IRC (Ping timeout: 250 seconds)[20:40:38] *** Layl <Layl!545740b6@gateway/web/freenode/ip.84.87.64.182> has joined ##vulkan[20:43:24] <Layl> I'm trying to use VMA in my project, I'm using Rust + Ash (for Vulkan) + vk-mem (for VMA). When I try to create a buffer though, I get VK_ERROR_FEATURE_NOT_PRESENT. Here's my code for reference, it should match the index staging buffer from the sample. https://gist.github.com/LaylConway/7868a7ef417cda9c63933c805a3103c2[20:43:49] *** LunarJetman2 <LunarJetman2!LunarJetma@5ec16e1a.skybroadband.com> has joined ##vulkan[20:43:54] <Layl> For some reason I can only create buffers that are GPU only, not CPU to GPU, GPU to CPU or CPU only[20:44:05] <Layl> I'm probably making a silly beginner's mistake, but I've run out of ideas to fix this[20:46:12] *** LunarJetman <LunarJetman!LunarJetma@5ec16e1a.skybroadband.com> has quit IRC (Ping timeout: 244 seconds)[20:49:06] *** emersont1 <emersont1!~emersont1@host81-155-32-5.range81-155.btcentralplus.com> has joined ##vulkan[20:52:59] *** emersont1 <emersont1!~emersont1@host81-155-32-5.range81-155.btcentralplus.com> has quit IRC (Remote host closed the connection)[20:54:46] <Ralith> it sounds like you need to enable some device feature for that lib to work[20:55:10] <Ralith> are you running with the validation layers?[20:57:34] <Layl> Ralith: I am using the validation layers. The docs seem to indicate this happens when VMA can't find a memory type for what you're requesting[20:58:01] <Layl> If there's any specific feature that must be enabled, VMA's docs don't mention it[20:58:14] <Ralith> is the error coming from the Vulkan impl or from VMA?[20:59:22] <Layl> it's crashing in vk-mem, it's receiving an error from VMA and panicking on that[20:59:55] <Layl> I don't know if the error is coming from vulkan or VMA itself, as there's no information other than that error code[21:00:04] <Ralith> what memory type is it looking for that it's not finding?[21:00:17] <Layl> Oh! I did miss a validation message that was printed before the stack trace[21:00:42] <Layl> Says somewhere vkCreateBuffer is receiving VK_NULL_HANDLE[21:01:03] <Layl> Oh nevermind that's a try I did where I removed the usage[21:01:22] <Layl> Ralith: I don't know what memory type it's looking for as VMA resolves that, I just tell it what I need[21:01:37] <Layl> Running it again with the usage again it doesn't print any validation errors[21:01:52] <Ralith> you'll probably need to step into it to work out what's going on, then[21:03:24] <Layl> Can't really do easily that when it's going across a language barrier into an external library[21:03:55] * penguin42 did end up rebuilding vulkano on rust adding a few extra prints to figure out where errors were coming from[21:05:28] *** penguin42 <penguin42!~dg@cpc109017-salf6-2-0-cust428.10-2.cable.virginm.net> has quit IRC (Remote host closed the connection)[21:08:15] <Layl> Ralith: to clarify, I can get it working when it's a GPU only allocation, so it's not like there's fundamentally a device feature I'm missing for VMA to work[21:08:29] <Layl> Just can't do much with only GPU accessible buffers[21:08:39] <Ralith> yes, that was a red herring[21:08:46] <Ralith> gdb steps across language barriers fine[21:08:54] <turol> Layl: what gpu?[21:09:01] <Layl> GTX 980[21:09:54] <Layl> I don't think I've got gdb installed, I'm on windows and on the MSVC rust toolchain, I can try if I can debug it with VS2018[21:10:08] <turol> so, nvidia-style memory areas[21:13:17] <Layl> Oh right VS2017, 2018 doesn't exist[21:16:20] <turol> are you sure you have the right usage?[21:16:37] <turol> for cpu to gpu transfer you must use VMA_MEMORY_USAGE_CPU_TO_GPU[21:16:45] <turol> you seem to be using CPU_ONLY[21:16:55] <Layl> turol: both CPU_ONLY and CPU_TO_GPU give the same result[21:17:21] <Layl> I decided to try CPU_ONLY based on the sample: https://github.com/GPUOpen-LibrariesAndSDKs/VulkanMemoryAllocator/blob/master/src/VulkanSample.cpp#L380[21:22:03] <Layl> I am genuinely surprised that, yes indeed you can debug Rust in VS2017 without any plugins[21:23:51] <Ralith> rust looks pretty much the same as C to a debugger[21:23:54] <Ralith> this is not an accident[21:25:18] <Layl> Alright, it seems the error is coming from vkCreateBuffer[21:26:12] <Layl> Oh nevermind, I confused some control flow, one moment[21:27:04] <Layl> Okay, in the create buffer function, it's coming from AllocateMemory[21:29:35] <turol> looking at the code (might not be the same version) that could be because of vmaFindMemoryTypeIndex or AllocateMemoryOfType[21:29:40] <turol> which one?[21:31:33] <Layl> Yea, it seems to be coming from vmaFindMemoryTypeIndex[21:35:24] <Layl> So I guess it can't find anything[21:35:48] <turol> have you check with vulkaninfo that your driver is sane?[21:36:01] <turol> nvidia should advertise a dozen memory types and two heaps[21:39:11] <Layl> As far as I can tell it looks sane: https://i.imgur.com/HjUJ7kB.png[21:39:35] <turol> yes, that's what it's supposed to look like[21:39:53] <turol> the other reason findmemorytype could fail if the flags are not sane[21:40:14] <turol> look at them with the debugger or looks at the rust binding source code[21:42:15] <Layl> Required flags are `VK_MEMORY_PROPERTY_HOST_VISIBLE_BIT | VK_MEMORY_PROPERTY_HOST_COHERENT_BIT`[21:44:30] <Layl> Huh actually, preferredFlags shows up as 7, not 6, so it also has the DEVICE_LOCAL_BIT[21:44:37] <Layl> that's definitely not sane for CPU_ONLY memory[21:44:47] <Layl> *requiredFlags[21:45:44] <Layl> `pAllocationCreateInfo->requiredFlags` has that flag[21:46:23] <turol> well there's you problem...[21:46:29] <turol> sounds like some problem with the rust binding[21:47:42] <Layl> This is all the bindings do: https://github.com/gwihlidal/vk-mem-rs/blob/master/src/lib.rs#L1787[21:48:42] *** alexge50 <alexge50!~alexge50@unaffiliated/alexge50> has quit IRC (Ping timeout: 246 seconds)[21:49:08] <turol> what does "..Default::default()" in your code do?[21:49:16] <turol> i don't speak rust[21:49:49] <Ralith> in this case, zeroes out the unnamed members[21:50:26] <Ralith> except for type fields, which are set appropriately[21:58:44] <Layl> Right, it is correct that the struct coming from the bindings has the wrong flags[22:00:34] <Layl> What the heck[22:00:56] <Layl> Ralith: check out this line: https://github.com/gwihlidal/vk-mem-rs/blob/master/src/lib.rs#L563[22:01:04] <Layl> This is not what I call a sane implementation of Default[22:01:08] <Layl> Oh my god[22:01:29] <Ralith> submit a PR[22:01:34] <Layl> Yea, I'll go do that[22:01:54] <Layl> Glad I at least found the issue[22:02:12] <Layl> Easy mistake to make though, if you're quickly implementing it, since it's the first entry so it would be sane if it were an enum[22:03:19] <Ralith> on a totally unrelated subject: is it legal to bind a texture array whose layers are not all in the declared layout, so long as your actual accesses are restricted to the layers that are?[22:07:06] <Layl> 'ere we go: https://github.com/gwihlidal/vk-mem-rs/pull/4[22:21:19] *** Layl <Layl!545740b6@gateway/web/freenode/ip.84.87.64.182> has quit IRC (Ping timeout: 256 seconds)[22:47:04] *** marcheu <marcheu!~marcheu@annarchy.freedesktop.org> has joined ##vulkan[23:20:35] *** slime <slime!~slime73@24.215.81.93> has joined ##vulkan