00:00 <HdkR> Could just be that the hardware requires the format to be unpacked? Or does it do something when the empty bits have data in them?

00:00 <alyssa> HdkR: Format is determined based on fmt of the framebuffer itself (specified elsewhere in the struct)

00:00 <alyssa> 1) Likely 2) Possible

00:01 <anarsoul> alyssa: what happens if you set unused zeroes to ones?

00:01 <HdkR> ^

00:01 <alyssa> anarsoul: I have to implement support for RGB565 RTs first, then I can tell you that :p

00:01 <HdkR> Even whiter whites

00:02 <anarsoul> alyssa: is it difficult?

00:02 <anarsoul> or just a matter of changing format in rt descriptor?

00:03 <alyssa> anarsoul: Just the bits

00:03 <alyssa> But I'm not magic :p

00:03 <HdkR> Not yet anyway

00:03 <alyssa> HdkR: I'm working on letting love consume me

00:04 <alyssa> Then I'll be nothing but magic.

00:04 <HdkR> Love of open source

00:04 <alyssa> What else would there be? :)

00:05 <HdkR> Love of closed source

00:05 <HdkR> heathens :D

00:19 <alyssa> Sure feels like a lot of work for a single dEQP test atm ...

00:19 <alyssa> Still failing, too!

00:22 <alyssa> Not even close to passing, either

00:28 <marcodiego> alyssa, these multiplications... don't they must be cast to a type?

00:43 <alyssa> marcodiego: Yeah, but in pseudocode! :)

00:44 <alyssa> ^not

01:52 <alyssa> ~2600 expected ES2 failures, huh? Something should be done about that ... :)

04:53 <tomeu> alyssa: cool, then I guess it's fine to skip liveness analysis as in my original patch

04:54 <tomeu> alyssa: any thoughts on this, btw? https://gitlab.freedesktop.org/tomeu/mesa/commit/06bf97e4a83637302199e31c53b12e12b112ba22

13:40 <tomeu> alyssa: is the plan to use u64 everywhere we have uintptr_t in the cmdstream?

14:09 <alyssa> tomeu: I mean, the skip is fine (I can't think of why that would hurt anything), I'm just not sure why that's necessary and a little worried it's masking a deeper bug

14:11 <alyssa> (The second skip. The -1 skip sounds correct.)

14:11 <alyssa> I don't have a firm plan on how we want to handle cmdstream bitness yet.

15:21 <pendarsoheil> This is probably the wrong place for tech support for ASUS C201 but I was hoping someone could help me out here. I recently bricked my device and reflashed it with ChromeOS Recovery. Problem is... now I can't enable debugging mode. I've tried various ChromeOS versions but same problem.

15:22 <alyssa> urjaman: ^^?

15:22 <pendarsoheil> "something went wrong debugging features were not completely enabled on this chrome device site:www.reddit.com"

15:22 <alyssa> bee

15:23 <daniels> try #chromium-os

15:24 <urjaman> i think i did see it say that to me the last time i recovered one, but i didnt actually need any of those features i think (it's still in dev mode) so i just shrugged and moved on

15:24 <alyssa> pendarsoheil: Come to think of it, I do remember seeing that on a Veyron

15:25 <alyssa> Also shrugged it off

15:25 <urjaman> (that is, the 2GB one :P - the 4GB one no longer has ChromeOS or even depthcharge for that matter...)

15:25 <pendarsoheil> Crap. That may make sense

15:25 <pendarsoheil> I have a few of these bad boys. I may have bought a 4GiB one :(

15:26 <urjaman> ":(" ?

15:26 <pendarsoheil> Wait nvm. I was able to boot from USB prior

15:26 <pendarsoheil> But now I can't use cros_flash to get those same permissions to flash a linux distro

15:26 <urjaman> i have no clue what you're talking about really

15:27 <pendarsoheil> Sorry, if I don't have debugging features enabled is there a way to install a Linux Distro (boot from USB debugging feature)

15:27 <urjaman> crossystem dev_boot_usb=1 and it'll boot from USB

15:28 <urjaman> (with Ctrl-U in the depthcharge nag screen)

15:28 <pendarsoheil> urjaman, running Libreboot. Cannot open lockfile //run/lock/firmware_utility_lock: Permission denied

15:29 <pendarsoheil> waitpid() or mosys error \ Parameter dev_boot_usb is read-only

15:29 <urjaman> run crossystem as root

15:29 <urjaman> but if you're running libreboot, why do you even have ChromeOS on it :P

15:30 <tomeu> hehe

15:30 <pendarsoheil> T_T. TY

15:30 <pendarsoheil> I bricked the device previously haha

15:31 <pendarsoheil> Completely forgot the process. Now a quick question regarding mesa -- I know Panfrost is coming to Mesa in 19.1 -- is 19.1 still on track for May 21st? Also, how would one install Panfrost on Devsus/Devuan?

18:40 <tomeu> alyssa: I think the results in https://gitlab.freedesktop.org/tomeu/mesa/-/jobs/277368 are correct

18:41 <tomeu> if you apply the diff to the expected results file (or download the whole failures file from the job artifacts) and push, the next run should be green

18:41 <tomeu> so we have 2841 failures in master atm

20:25 <alyssa> tomeu: OK, I'll try that this evening. Thank you! :)

22:53 <alyssa> tomeu: dEQP-GLES2.functional.fbo.render.depth.npot_tex2d_rgb_depth_component16 is a flip flop..

22:55 <alyssa> Running it myself flip-flops, so that's totally fun :V

22:56 <alyssa> Memory corruption, GPUside.

22:56 <alyssa> The worst kind of flop is a flip kind of flop.

22:57 <HdkR> Cosmic butterfly flops

22:59 <alyssa> That or vim