01:40 steel01[d]: https://cdn.discordapp.com/attachments/1034184951790305330/1428558313456664728/nice_stairs.jpg?ex=68f2f01b&is=68f19e9b&hm=4cb36e716c9bc9142aab6ca7776a2d15ca444a2cd3f721b187e37b9b80365158&
01:40 steel01[d]: There we go.
01:40 steel01[d]: Beautiful boot ani.
01:41 sonicadvance1[d]: These hyper-stylized car logos are getting out of hand.
01:41 steel01[d]: https://gitlab.incom.co/CM-Shield/android_external_minigbm/-/commits/lineage-23.0
01:41 steel01[d]: Here you go, gfxstrand[d]. Squash those appropriately. And figure out where I screwed stride.
01:43 steel01[d]: Oh, and just to double-check, you did change the minigbm platform in tegra-common from nouveau to tegra, right? I forgot that on my first build. 🙄
03:19 astondiary: So what shows that being a jewish propaganda in the field of compression is: a simple procedure that has nothing to do with fft/dft, kececi prime numbers nor markov-chains. it has a bit to do with collatz however, such as we fix the format a bit, for odd or even numbers, but other than that it does not connect to collatz either. Algarv is the translation of kececi prime numbers, where many
03:19 astondiary: had suggested it, to make the compression you do not need those either. 8-52=44/2=22 8+22=30 you do this at compile time, suggestion is to do it another way around, going from smaller to bigger to avoid division, 52 is a quadlet of 7+11+15+19 that came out from a long semi-complex synthesis that matters few. so so 8+22-26+8+22 is 4 while 8+22-52+30=8, so the selection get's 26's and not
03:19 astondiary: selection get's 52's where as from all 52's subtract yields answers for any bank, there is some index calculations involved, per quadlet and offsets per a word of 32 or 64bit, and do not forget to add +1 to the answer for better decoding without collisions. I do not think such formulas could ever be out of any complexity for employed programmers, so it has to be a scam propaganda. And along
03:19 astondiary: with terror committed to my life it will result into bloody war if continued to be done at me. however in the example any number inside 52 can be chosen except 0, which in case of NaN handling needs to be presented with something else. That is very simple show off how numeral systems actually work, and they are done in science with care to allow high end computation and compression, the fast
03:19 astondiary: computation is a minor extension from there, a set of routines at compile and runtime.
03:20 astondiary: I very much doubt that you could not approve those methods, or can not compile those for FS's cause it's impossible that you should be so stupid, while doing so much more complex work.
03:29 kode54[d]: god dammit
03:35 diddeynumpy: What is being afraid of is the dirty laundry that comes out from store now and decrypt later, since breaking RSA is just a matter of getting minor set of numbers right. And then the tyrans and traitors and obnoxious trash in the world, gets handled by militarians , where our lines gonna really clash where as you were traitors with bunch of other jews eating my cells, we torture them to
03:35 diddeynumpy: death and on spot butcher them alive and put their organs while they watch themselves on a pan, i have talked about it and they deserved those actions and dug their own hole, it's just i now humiliate as much as i can, and soon they are on the pan.
03:35 kode54[d]: I swear they have an eye in the channel already and just send in another connection to spam
03:36 HdkR: Not too hard when it's publicly logged :)
03:39 HdkR: They're inspirational really. I hope that one day I too can be this dedicated to something.
03:42 kode54[d]: dedicated to the bit
03:43 airlied: if they could just use twitter, they'd make 1000s of dollars
03:43 airlied: spewing that level of shit on twitter constantly would get you insta followers
03:44 gfxstrand[d]: steel01[d]: Thinking about this more... I suspect you didn't screw anything up. I suspect this is a bug that's been hidden in the GL driver for years.
03:44 steel01[d]: Oh?
03:44 gfxstrand[d]: It may or may not be fixable. I suspect not. If not, then the solution, once again, is to fix the NVK bugs and delete the GL driver. 🙃
03:44 steel01[d]: ><
03:45 steel01[d]: But hang on a sec. I'm 99.8% sure this happens without mesa loaded.
03:45 steel01[d]: I can explicitly verify that in a bit.
03:46 steel01[d]: Cause tegra-drm can drive the display hardware with the gpu being on at all. And that backend only uses tegra-drm ioctls.
03:47 gfxstrand[d]: Basically, the GL driver totally ignores the modifier on imported buffers. It instead takes whatever is set on the BO itself. The only reason NVK->GL import works is because NVK has a special case to set the tiling on the BO for modifiers, even though it shouldn't be necessary. (The modifier is supposed to carry that information.) The problem is that when the BO is imported from the tegra driver
03:47 gfxstrand[d]: into the nouveau driver in the kernel, the extra BO metadata is discarded. (All BO imports are dumb if they cross driver boundaries.) This means that nouveau always sees a linear BO and a modifier, which it ignores.
03:50 steel01[d]: Mmm. That sounds fragile. Once my current pipeline of builds finishes, I'll run a build with swiftshader (no mesa at all) and this minigbm backend. See what it does. I *think* I did that when I was originally working on this and saw the same stride issue. But I can't swear to it right now.
03:51 gfxstrand[d]: The solution is to not ignore the modifier information. However, some of that information is the PTE kind, which the userspace driver can't supply on its own because it goes in the page tables, which aren't something it controls. We've fixed this for NVK with the new kernel API but nouveau still uses the old API and the two are incompatible. So even if we did fix the tile mode, we'd still have PTE
03:51 gfxstrand[d]: kind to contend with and pre-Turing, the GPU will ignore tiling if you don't use one of the tiled PTE kinds.
03:52 gfxstrand[d]: In short, NGL can only tile on BOs allocated from nouveau.ko and it ignores the modifier provided by gralloc. That second bit is fixable, the first really isn't.
03:54 gfxstrand[d]: But there might be a hack! I might be able to make my gralloc backend enumerate on the tegra device so that's the one SurfaceFlinger sees but then secretly open the nouveau device behind everyone's back and allocate BOs off of that.
03:54 gfxstrand[d]: But long term the solution is to delete the GL driver. :frog_upside_down:
03:54 steel01[d]: Not sure that'd even be a hack. It'd probably be better performance.
03:56 gfxstrand[d]: But first we need to figure out how to get SurfaceFlinger to stop segfaulting.
03:56 steel01[d]: Well, my build with the broken stride booted to launcher.
03:57 steel01[d]: So once you get a build with that, you should be good on that front.
03:57 gfxstrand[d]: 👍🏻
03:58 gfxstrand[d]: I may not have pulled enough from your branch.
03:58 gfxstrand[d]: Or maybe I screwed it up trying to add the service stuff
03:58 gfxstrand[d]: Oh, you have more hacks. Will try tomorrow
04:01 steel01[d]: 'hacks' The main issue was that backend_tegra didn't get added to drv.c It wasn't in the commit you picked, because it had been previously added in a commit adding tegra support to the dumb buffer backend.
04:02 steel01[d]: And since tegra apparently shouldn't support linear buffers or whatever that was, that dumb driver support commit shouldn't be grabbed.
04:10 gfxstrand[d]: Right
04:45 cubanismo[d]: gfxstrand[d]: Isn't that what the Tegra driver does with every other operation anyway?
05:18 steel01[d]: steel01[d]: Well, I get to eat this. I thought I had checked that case before, but apparently I never did. The backend renders correctly with swiftshader. The boot ani seems notably slower than with dumb buffers though, which is kinda weird. The normal ui seems about the same, though. So yeah, the issue is with the interaction with nouveau.
05:36 jon_hamstad: It's a total fullest tyranny faced, by humans in the nature who reversed the good will somehow, it is technically very long story, I have no mental illness to speak of. I won a tournament in style that none had did before straight 12set or 14set wins without suffering a loss, just a one time success in Tallinn. I started to notice compromised health after throat surgery and vast conspiracy
05:36 jon_hamstad: against me, my bones no longer grew, as the osgood-schlatter was largest ever recorded, the bumps started to form near the growth plate etc. On fingers on hands on the head, so someone was controlling my height as well as tendons like penis, where i could see the degradation after youth clearly. I have lot of hard tissue, so they used xray fluorescence on me, and started to get pure
05:36 jon_hamstad: substances deposited out of my organs clearly, which they consume at other sites for others favour, however there were so to speak years of humiliations at my identity such as , ET ON NUKUKE NII PIKUKE SUL songs, and abortion leftovers celebrating their success around estonia. But it did never prove to be enough of terror they committed, so at surgery i was completly butchered etc to become
05:36 jon_hamstad: first time entirely injured , and even that made no difference on their bad performance, so the terror has continued ever since, and I am glad to say, i had enough of that with my alliance and justice is being served soon. Mentality and brain is the only one not injured, so thinking is my strongest side and only side i can rely on, but they just want to make the full house or royal flush
05:36 jon_hamstad: with the scam to cause me more trouble to title me with mental illness, but it will only add to their pentalty which is realized with more torture, since they already have several death pentalties assigned.
07:30 leilasuarez: the other way around is a better method, but you figure it out yourself, I have talked with americans every where i go, they are quite intelligent ones and very friendly, it's just that their rulers are not as good as Putin in fighting jew terror, and hence i had chosen Russians for my allies, and USA is not as powerful as they think about themselves with their tomahawks either, Russia is
07:30 leilasuarez: full of natural resources and with my technology i know that no other nation can threatten me or them after i join the military. No hard feelings for americans however, i have liked so many of them. Current world of science is at the tops, when you execute the code directly it will kill bastards without problems. And united states ain't gonna hold us back in any ways, we do not need to leap
07:30 leilasuarez: to surpass them, we already have the leap done, and manufacturing and alliance in place. So with fpga's another methods does not need division in compiling the trees only + and -, which is only slihtly more complex to get right, CGRA's FPGA's and asics are all good to wipe out monsters. however former two perform the best, tensor cores that llvm synths are prolly as good as those former two
07:30 leilasuarez: too.
08:29 jja2000[d]: This would go hard ob a Death Grips beat
08:57 leilasuarez: The conflicts are hard always, if the tyrans would be beatable by other means i would had done it, but I have not been able to live my life in any enjoyable way in anywhere where monsters would not annoy me, and at that state i have nothing to lose either, i am nearly destroyed altogether by weirdos, and Europe's policy caused that, so this was net loss to risk my health and such policy i
08:57 leilasuarez: can not tolerate, so it could be only Russians or Americans to ally with. Why it's a net loss to table tennis sports and life, i speant 9years training twtice during summers, and i am very confident to say that those monsters never played but dealt inside their intrigues only and terrored me with their doctor friends. There isn't much to do with such people, they can mucnh steroids yes, but
08:57 leilasuarez: that alone does not make them capable as i was promising to be.
09:14 ermine1716[d]: misyltoad[d]: Grave alloc
09:22 leilasuarez: So yes, i became very knowledgeable in technology, cause i backed up my research as of late, for an example, the hotel my dad financed with type of dark moneys, had very okay mid or entry level growatt rectifiers, IOT devices very easy to handle, there was nothing wrong with the tech, constantly cracked was everything there, cause tyrans are idiots , not handling any work, nor sufficient
09:22 leilasuarez: results etc.
09:30 leilasuarez: it's not like americans do not know that RLE compression is not a way to go when lives are in danger in a conflict zone, it's not like i try to teach anyone here, but those people really annoyed me few, i never published anything terror like described is not allowed, but indeed the real code is bit too high profile at the same time, so americans faces were also very worried, as are my
09:30 leilasuarez: facelines.
09:35 leilasuarez: but what i say regardless of the authority or current regime or order, life is a derivative about what decisions and useful things you do, and that is wrong, when the regime changes from russia to europe and some monkeys who were never even capable declude people who were in soviet unions lines, it's wrong, because soviets took the best of estonians to their lines.
09:35 leilasuarez: and in europe the worst morons got a chanche to commit more crimes
09:38 leilasuarez: i just you know figured knowing their scams, that once they say i am a russian puppet and spying and agent for Russians, to claim this is also very shortsighted and incredibly wrong
09:39 leilasuarez: no matter what the regime is , i gotta take the best out of my body, and during europes tyranny i have not able to do that, so my decision or hopes are simple
09:53 leilasuarez: those people who saw what happened to me, i do not know if they can be relied on, sincei queried the video as to what the hell happened again, i lack the memory after the beer bong shot, and my allies were not there, when i got knocked down, no memory whatsoever from a moment, where calimingly a retarded jew assaulted me. This is just one over hundreds of this, where Fidel Castro has had
09:53 leilasuarez: more such things. RIP fidel.
09:55 leilasuarez: only thing i know is again i have a chip on the neck, that is the only thing i do know , and back to square one, it was already told before i got it, that you will be getting one, so soon i am dead anyhow, but claimingly we know who to go after, my alliance knows that.
09:57 leilasuarez: Elon Musk says a lot of things, and i was not ever sure if the neuralink is real, but something like this i have on the neck, i have had many before alike.
09:57 leilasuarez: but not on the neck
10:00 leilasuarez: once again i report that my last memory over that incident was incoherent talks what happened totally during another day where i tried to kiss some italian jew, which was partly only truth, and kaboom, it was a setup anyways, and days later a trashbag attacked me, but if they wanted to put a chip on me, this was all decided before , and they told me before that kissing scene happened, my
10:00 leilasuarez: last memory was during beerpong, and i was recklessly drunk, cause no friend i had with me, i was homeless alone
10:01 leilasuarez: and there was none to guard me, cause all were dickheads criminals working against me
10:02 leilasuarez: There is no surgeon in this country who wants to remove that chip
10:02 leilasuarez: and i am not sure if i can find anyone in the world
10:03 leilasuarez: offshore or active licensed or current whatever, so this is a knock down to me a final one, i have no ransom requests as of yet, besides this trash on e-mails pegasus surveillance software extorting to publish my naked pics etc.
10:04 leilasuarez: i do not care about this garbage
14:07 snowycoder[d]: Crashing the GPU from userspace is distressingly easy.
14:07 snowycoder[d]: Your buffer addres calcs are wrong? Boom, need a reboot
14:08 gfxstrand[d]: 😭
14:23 mohamexiety[d]: hmm that’s a bit odd
14:23 mohamexiety[d]: the only stuff I found that needed hard reboots was when I was doing compute mme tests and messing up the methods was basically a DoS attack
14:24 mohamexiety[d]: otherwise it was just the usual channel killed/device lost stuff with the app crashing but not the system
14:25 snowycoder[d]: Maybe it's because I'm doing spinlocks on strange addresses?
14:26 mohamexiety[d]: oh yeah lmao that could do it
14:33 karolherbst[d]: there are certain types of errors that nouveau fails to recover from
14:34 karolherbst[d]: spinning infinitely in a shader usually causes more severe issues
14:40 karolherbst[d]: `dEQP-VK.spirv_assembly.instruction.compute.untyped_pointers.vulkan_memory_model.cooperative_matrix.mixed.load.a.row_major.float16` hits us with `deqp-vk: ../src/compiler/nir/nir_lower_explicit_io.c:1150: explicit_io_offset_from_deref: Assertion stride > 0 failed.` 🥲
14:40 karolherbst[d]: and it's a pain
14:47 karolherbst[d]: oh nvm the cmat lowering should be more explicit about casts I think...
14:49 karolherbst[d]: actually why is the stride 0...
14:59 karolherbst[d]: mhh it comes straight out of vtn like this
15:00 karolherbst[d]: gfxstrand[d]: if I have a cast with ptr_stride = 0, do I need to cast it to something with a stride before I can use it with `deref_ptr_as_array`? Though then the question is what stride should I assume...
15:03 gfxstrand[d]: yes
15:03 gfxstrand[d]: ptr_as_array requires a stride
15:03 karolherbst[d]: right...
15:04 karolherbst[d]: ohh wait on cmat_load/store I actually get a stride
15:04 karolherbst[d]: mhhh
15:04 karolherbst[d]: I need to think about it
15:05 karolherbst[d]: I wonder if the stride is required to be constant, that would make a lot of things a lot simpler
15:11 karolherbst[d]: yeah.. just casting it to the element type should be doable there
15:12 karolherbst[d]: maybe I should play around doing the lowering based on derefs and to not manually do the array offsetting...
15:38 snowycoder[d]: We don't have `nir_intrinsic_printf` wired up, do we have any plans for that?
15:38 gfxstrand[d]: gfxstrand[d]: I'm also slightly tempted to teach NVK to allocate coherent memory from Tegra just to work around the nouveau.ko bug. Probably better to fix the bug, though. 😂
15:38 gfxstrand[d]: snowycoder[d]: We should. When the time comes to do more CL and RT stuff, it'll come in handy.
16:52 snowycoder[d]: Test case 'dEQP-VK.fragment_shader_interlock.basic.nodiscard.ssbo.pixel_unordered.1xaa.no_sample_shading.16x16'..
16:52 snowycoder[d]: Pass (Pass)
16:52 snowycoder[d]: https://cdn.discordapp.com/attachments/1034184951790305330/1428787793055715448/0tk8rfe38at81.png?ex=68f3c5d3&is=68f27453&hm=fd0638cbc861247dcc3c67ccb5411a498e6c6a2def7b2b93a62fe755afd3e094&
16:53 snowycoder[d]: Somehow initializing the interlock buffer to -1 does... something?
16:55 gfxstrand[d]: \o/
19:15 Lynne: I can with 100% reliability hard reset my machine by dispatching a compute shader, just in case anyone's interested
19:57 snowycoder[d]: Lynne: I really wonder how nvidia handles that. In servers it could be really bad
20:17 gfxstrand[d]: I've not seen that with modern GPUs running real firmware
20:53 mohamexiety[d]: ...and doesn't walk out (the gpu hanged, can't exit without a hard reset)
21:32 cubanismo[d]: Does locking up the GPU reset the machine at the HW level, or just something nouveau is doing? We have test cases for like, infinite loop shaders.
21:32 karolherbst[d]: GPU reset has a few gaps afaik
21:33 cubanismo[d]: If you find something where you submit commands and the GPU becomes unreachable or causes some hardware-level reset, I'm sure people here would be interested to hear about that.
21:33 cubanismo[d]: Unless it was like, you programmed the GPU PTEs to point to kernel code pages. Then, don't do that.
21:34 cubanismo[d]: Definitely never not even once did anything that stupid myself.
21:35 mhenning[d]: I think there are specifically some known issues with recovery in the nouveau kernel module. So the hardware might recover okay but then leave the kernel in a weird state.
21:37 karolherbst[d]: yeah.. the kernel gets some interrupts and some certain types of those nouveau can't properly reply to or something
21:37 karolherbst[d]: there is a sequence of things one needs to do based on the interrupt
22:46 ermine1716[d]: GPU resets are hard
22:48 ermine1716[d]: If you need that, i have some vulkan code that messes up memory transfer queue indices
22:50 mohamexiety[d]: cubanismo[d]: This does happen when you submit some invalid commands but idk if it’s a nouveau thing or a NV thing
22:51 mohamexiety[d]: I ran into this while I was doing the compute mme unit tests. There are only 8 registers for compute, when I would try and access more than 8 it was an instant freeze and I’d need to hard reset
22:54 mohamexiety[d]: I am suspecting it’s nouveau because I once had this pop up:
22:54 mohamexiety[d]: [ 73.880144] nouveau 0000:01:00.0: gsp: intr badf5620
22:54 mohamexiety[d]: Which from what I gathered is an indication that the GPU did recover and this is something from the reset process