01:09 <hanetzer> second trig exam down

01:09 <hanetzer> one more, then the final for this semester

04:53 <alyssa> A few more things I want to do before I'm happy enough to send it off to the ML, but I spent the evening working on a series of refactors relating to framebuffers

05:26 <tomeu> chewitt, robher: indeed, I missed a little -a...

05:26 <tomeu> chewitt: btw, I haven't been ignoring your questions, but you have never been online at the same time as me before :)

05:28 <chewitt> no worries .. I travelling a lot with work at the moment so not online a lot

05:28 <tomeu> alyssa: please tell me if I can help with the BO refactor in mesa

05:29 <alyssa> tomeu: Focus on the kernel stuff, I can handle pure-userspace ;P

05:29 <alyssa> (Sorry, that came off rude. Sentiment being, you're needed on kernel, don't get distracted by my shenanigans? :P)

05:34 <tomeu> np :)

05:36 <alyssa> tomeu: While I have your attn though -- do you have a preference on merging the earlier parts of the DRM series (the prep stuff that doesn't touch the kernel abstraction yet)? It might ease your load a bit and I don't see a reasont o delay it, but maybe you wanted to keep the series coherent

05:36 <tomeu> alyssa: that would be great, though there might be some feedback that I needed to address?

05:37 <alyssa> I think the first few I just R-b'd as is

05:47 <tomeu> there's good stuff in panfrost/framebuffer-bits

05:47 <tomeu> alyssa: can you push them as they are, or better to have them in a MR?

05:48 <alyssa> tomeu: That's my BO refactor WIP; I'm sending it off when it's, you know, ready :P

05:48 <alyssa> ("Alyssa, why are you still awake? :V")

05:50 <tomeu> robher: the rebase is looking good, I'm testing now

13:00 <tomeu> alyssa: oh, I think there was a misunderstanding, I asked you about pushing those patches from me that have already a r-b

13:00 <tomeu> not your WIP branch :)

15:18 <alyssa> tomeu: Oh, gotcha

15:20 <tomeu> alyssa: with the DRM driver, I'm getting DATA_INVALID_FAULT in the job dispatcher irq when a frame is submitted due to a flush when the context gets destroyed

15:20 <tomeu> but that doesn't happen with kbase

15:21 <alyssa> Hint: all of all flushing logic is deeply flawed. All of it.

15:21 <tomeu> do you have any ideas?

15:21 <alyssa> :P

15:21 <alyssa> tomeu: Why are we flushing on context destroy? Like, is that also the case on kbase?

15:21 <tomeu> well, by flushing I mean that the gallium pipe_flush cb gets called

15:22 <alyssa> (Aside: DATA_INVALID_FAULT is the most common -- and annoying -- to deal with it...)

15:22 <tomeu> guess we could just not submit any jobs on pipe_flush based on what flags are passed

15:23 <alyssa> tomeu: That's sort of the hack we're doing with kbase as it is

15:23 <tomeu> but I would like to know why with kbase we don't get any faults logged

15:23 <alyssa> Quick question, this "extra" flush, are there any draws?

15:23 <tomeu> that's what I wanted to check next :)

15:23 <alyssa> :)

15:23 <alyssa> Since I think there have been historical bugs for test-clear type jobs

15:25 <alyssa> Aside: I should've mentioned the motivation for the last night's patch series

15:25 <alyssa> All of this is a very long-winded build up to slimming down set_framebuffer_state

15:26 <alyssa> In particular, rather than converting a linear/tiled BO into AFBC, at BO creation time we decide the layout depending on the usage flags (which also fixes the memory leak thing you mentioned..?)

15:27 <alyssa> Also, just as a sanity check, I'd like to test rendering and sampling from linear render targets

15:27 <tomeu> yeah, I'm all for refactoring that stuff

15:27 <alyssa> Obviously that already works conceptually (Weston does that across processes) but we don't have any tests for that for FBOs

15:27 <tomeu> no draws on context destruction, btw

15:27 <alyssa> tomeu: Alright, let me look at your code closer then

15:28 <tomeu> but I do get some frames submitted on context creation with no draws and no JS faults

15:28 <tomeu> so I guess

15:28 <tomeu> it's not a simple problem with job dependencies

15:28 <alyssa> I recall getting faults when there were no draws and no clears and just nothing

15:28 <alyssa> Which is why for a while I inserted a bunch of dummy clears

15:28 <alyssa> Which was so egregiously wrong but for some reason it helped :P

15:30 <tomeu> hmm, maybe the job descriptor was "too empty"?

15:30 <alyssa> Yeah

15:30 <alyssa> If such a hack really is necessary, it can now be done non-invasively in whatchamacallit

15:30 <alyssa> clear_mfbd or one of those

15:30 <tomeu> but then, I don't know why with kbase we don't get that fault

15:30 <alyssa> (The one that actually does, like, "fragment_rts[0]->clear_color_1 = job->clear_color"

15:30 <alyssa> kbase, I think, had a "frame cleared?" flag that we check for; maybe you removed that...?

15:34 <tomeu> checking

15:36 <tomeu> cannot find anything related to that

15:36 <tomeu> but I see this in panfrost_flush:

15:36 <tomeu> /* Nothing to do! */

15:36 <tomeu> if (!ctx->draw_count && !job->clear) return;

15:37 <tomeu> hmm, job->clear is forever ever 5

15:48 <tomeu> alyssa: this avoids the problem, but I don't know if is correct, and I still would like to know what I don't get those faults with kbase :/ http://paste.debian.net/1072095/

16:08 <alyssa> tomeu: job->clear... should be getting cleared every frame.. anyway?

16:09 <alyssa> panfrost_job is supposed to be reset after every frame

19:33 <davidlt> https://www.phoronix.com/scan.php?page=news_item&px=GraphicsFuzz-1.2-Released

19:34 <davidlt> GraphicsFuzz is the 3D GPU driver fuzzer that was born out of academia research for finding GPU driver bugs and ended up being acquired by Google and then open-sourced.

20:52 <HdkR> Interestingly the mesa-dev email about that project went directly to my spam folder

21:06 <hanetzer> davidlt: ooh, kewl

22:20 <alyssa> davidlt: GPU drivers are chockful of security bugs IME

22:20 <alyssa> Which is why WebGL is such an awful idea tbh

22:20 <HdkR> WHLSL and WebGPU to save the day!

22:20 <HdkR> :P

22:21 <alyssa> I think that's worse, dear

22:21 <alyssa> :P

22:21 <alyssa> ("Read in mom voice")

23:44 <hanetzer> wew, we goawt kernel 5.0