alyssa changed the topic of #panfrost to: Panfrost - FLOSS Mali Midgard & Bifrost - - Logs - Transientification is terminating. Memory reductions in progress.
anarsoul has joined #panfrost
anarsoul has quit [Client Quit]
anarsoul has joined #panfrost
anarsoul has quit [Read error: Connection reset by peer]
anarsoul has joined #panfrost
anarsoul has quit [Remote host closed the connection]
anarsoul has joined #panfrost
anarsoul has quit [Remote host closed the connection]
anarsoul has joined #panfrost
anarsoul has quit [Remote host closed the connection]
anarsoul has joined #panfrost
_whitelogger has joined #panfrost
anarsoul has quit [Ping timeout: 245 seconds]
BenG83 has quit [Ping timeout: 250 seconds]
BenG83 has quit [Ping timeout: 250 seconds]
anarsoul has joined #panfrost
BenG83 has joined #panfrost
NeuroScr has quit [Quit: NeuroScr]
NeuroScr has quit [Quit: NeuroScr]
_whitelogger has joined #panfrost
raster has joined #panfrost
raster has quit [Read error: Connection reset by peer]
<tomeu> thefloweringash: eballetbo and ezequielg are interested making all those features work with the mainline kernel
<tomeu> you can ask them in #armlinux
raster has joined #panfrost
<tomeu> (that's on the kevin)
pH5 has joined #panfrost
_whitelogger has joined #panfrost
cwabbott has quit [Ping timeout: 250 seconds]
cwabbott has joined #panfrost
cwabbott has quit [Client Quit]
cwabbott has joined #panfrost
_whitelogger has joined #panfrost
klaxa has joined #panfrost
cwabbott has quit [Ping timeout: 252 seconds]
chewitt has joined #panfrost
cwabbott has joined #panfrost
anarsoul has quit [Remote host closed the connection]
anarsoul has joined #panfrost
anarsoul has quit [Remote host closed the connection]
mateo` has joined #panfrost
anarsoul|2 has joined #panfrost
pH5 has quit [Quit: bye]
<Lyude> alyssa: did you get all of your patches reviewed btw
stikonas has joined #panfrost
<HdkR> Lyude: Which ones? The ones I reviewed? :)
NeuroScr has joined #panfrost
<raster> oooh that's soonm up on my todo list.. .get panfost working on my rockpro64 :)
<Lyude> HdkR: I guess so-it looks like that the tomeu merge is leftover as well
<HdkR> Which I haven't looked at that one much. Looked massive :)
<Lyude> yep, lol
<Lyude> lots of rebase from me
<Lyude> I always forget that gitlab actually tells me what I need to review and what I don't need to
<HdkR> I guess if an assignee was assigned?
<Lyude> yeah basically
<Lyude> still though
<Lyude> there are so many patches that get dropped in the projects I work on because of not knowing what still needs review, mail missing someone's inbox, etc. etc.
<HdkR> I /think/ I had issues with that feature previously where only the assignee could sign off the PR, other people couldn't if they looked at it instead
<HdkR> That was either gitlab or something very similar
<Lyude> o_O
<Lyude> weird
<Lyude> well if that's an issue gitlab is open source :)
<HdkR> If you could assign multiple people it wouldn't be an issue I guess :P
<Lyude> or assign people by labels
<HdkR> I don't want to stick people in boxes
<HdkR> :D
<Lyude> why not? :p, honestly it helps a ton with reviewing stuff
<Lyude> Normally the people I choose to review stuff I do so because they actually have experience in whatever area my PR is changing, and people seem to do the same for me (for instance, I usually get cc'd by random people when there's an MST change waiting on dri-devel)
<Lyude> it's better then assigning all new bugs to
<HdkR> Sure. I've also not had a gitlab instance with a lot of people using it. Max of 55? With only regular interaction with like 5 people :P
<Lyude> wait, I guess we don't do that last part since this isn't bz anymore
<Lyude> HdkR: mhm, i would hope you could just rely on people to forward it elsewhere if it isn't there issue
<Lyude> *their
<HdkR> That very much depends on project. I mostly expect things to get ignored :D
<Lyude> unfortunately same :(
<Lyude> i'm starting to realize though the problems there lie with not having a good way of making sure the right people get poked by default
<Lyude> or, well i've realized this since forever but you get the idea
<HdkR> "I want documentation on <thing>. I've added a list of ~300 people, sorry for the spam." Zero responses
<Lyude> oh yeah that doesn't work lol
<HdkR> :D
<Lyude> in fact it's so far from working I've been really strongly considering just removing cccmd for most of my projects
<Lyude> since all it does is spam people's emails and very rarely ever actually helps with getting the right people notified because the right people have muted Ccs at that point due to all the spam
<HdkR> Oh yea, of course
<HdkR> I'm slowly accumulating lists of people in various teams that I can just ping directly
<Lyude> i should probably do that
<HdkR> I find it still doesn't help in a lot of cases. So eh
<Lyude> for cases like that I usually just start poking more
<HdkR> I'm learning to berate
<HdkR> Lyude: Oh right. Only one real problem that I've found on this Razer so far. Its Type-C ports sometimes lose connection to the charger and...stop charging
<HdkR> Keep waking up to the laptop completely drained because of it :|
<HdkR> (Also mushy keyboard, but w/e)
<Lyude> oh no
<Lyude> the keyboard on mine is actually really nice, feels like a 2016 mbp
<HdkR> That'd be nice
<Lyude> HdkR: also be suspecious of the charger and maybe try using a different non-razer one
<Lyude> I remember chargers were an issue with this generation of blade stealths too, although I never ran into any charger problems myself
<HdkR> Definitely depends on type-C cable. I have one that very easily loses connection while another is pretty solid
<HdkR> Still pretty bad ports
<Lyude> huh,
<Lyude> usb-c on mine has been flawless, except for the outdated firmware that i'm very lucky to have a lyude-specific workaround for
<HdkR> Yea, other than the connection issues it has been fine as well
<Lyude> that's what I get for jumping on the TB3 docking train too early
<HdkR> Thunderbolt on one side keeps spamming messages in the kernel log though
<Lyude> oh?
<Lyude> do tell
<HdkR> Spams a bunch of those even though I'm not using it atm
<Lyude> hm.
<HdkR> One of the older Thunderbolt controllers, the JHL6340
<Lyude> tbh, I've been wondering myself if someone should make the thunderbolt drivers less spammy
<HdkR> Which is surprising, I expected them to use the new one that was announced early last year
<Lyude> looking at that makes me suspecious we're looking at a situation where your thunderbolt controller keeps getting runtime resumed after being runtime suspended
<HdkR> lol
<Lyude> HdkR: it's not surprising unfortunately :(. implementation wise thunderbolt is an intel mess
<Lyude> I say "intel mess" because it's redoing a ton of mistakes I thought we'd learn not to do from the bios days
<HdkR> I haven't had the pleasure to work with it directly :P
<Lyude> for ex: TB3 requires OEM specific firmware. OEM == the manufacturer of your laptop, not of the thunderbolt controller
<Lyude> which is -bs-
<HdkR> wtf
<HdkR> The TB3 controller is a friggin PCIe device
<Lyude> because if your oem doesn't supply updates, and your oem won't supply updates because when has that literally ever been a thing oems are good at ever?, you don't get fw updates
<HdkR> wtf does it need an OEM specific firmware for? :|
<Lyude> HdkR: yeah, I know all of this becaue of my TB3 docking station adventures
<Lyude> i have two older-rev caldigit ts3 docks in this office, and robclark has one of the newer revs, I know for a fact robclark's literally won't work with my laptop because of the dumb oem firmware issues
<HdkR> Wait, is this the same reason why Thunderbolt add-in cards for some motherboards actually require connecting to a little i2c header on the mobo?
<Lyude> HdkR: i have no idea
<HdkR> wow
<Lyude> i don't work with thunderbolt much, i've just struggled with getting it to work on here before
<Lyude> HdkR: yeah it's definitely an intel design
<HdkR> I was thinking about getting the Razer eGPU dock for fun
<HdkR> :D
<Lyude> but I kinda hoped that even intel would realize putting OEMs in charge of updating /literally anything/ == basically just dropping updates entirely
<HdkR> yea
<Lyude> HdkR: oooh, that will definitely be Hell
<Lyude> but there's been a good bit of work on it iirc
<HdkR> Might be fun
<Lyude> hotplugging is /always/ 'fun'
<Lyude> anyway, HdkR: might be worth turning off runtime PM control on that thunderbolt controller and seeing if the spam stops
<Lyude> i have a feeling it will
<HdkR> Could try
<Lyude> imo though I really think the thunderbolt driver goes overboard with it's dmesg output :s
<Lyude> we stopped printing suspend/resume messages with nouveau entirely, and I want to do the same for amdgpu at some point
<HdkR> Ah, runtime pm is already disabled for it
<Lyude> you /sure/? check all the devs under that pci address
<Lyude> sometimes the RPM attribute isn't on the one that you would expect
<Lyude> if not, probably worth filing a bug
<Lyude> fuck-I recently managed to finally come up with a gnome-shell bug I've been living with for over a year that caused my laptop to lag like heck when docked after having suspended/resumed before, and now my life is perfect
<HdkR> Looks like the only thing changing is PCIe link width
<HdkR> I could probably skim the kernel source later and find what these messages are about
<HdkR> I recommend using i3 instead of gnome-shell :P
<alyssa> tomeu: oh, wb!
<Lyude> HdkR: I'm fine with having no page tearing :)
<alyssa> Lyude: Yeah, everything but the massive tomeu merge which we have to figure out what has to actually happen
<Lyude> alyssa: fair enough! i'll get to reviewing next chance I get
<Lyude> should help that I am now in boston again and not in florida
<HdkR> Lyude: Ah. I hear sway will fix that :D
<alyssa> Woop
<Lyude> HdkR: hehe, honestly (keep in mind too-I was a gnome3 refugee at one point) g-s has very much won my heart back
<Lyude> it's so smooth and, surprisingly light on resources compared to how it used to be
<alyssa> Memes
<HdkR> Interesting
<alyssa> Tell me when g-s is tiling..
<alyssa> And works with Panfrost ("Pretty sure that's your problem, Alyssa")
<HdkR> That's my main issue as well. I fell in love with the keyboard shortcuts for i3
<Lyude> also: I kinda need to dogfood g-s for fedora :)
<Lyude> since I have a weird knack for finding edge case bugs by accident
<HdkR> Always a good trait for being annoyed more often than everyone else
<alyssa> :P
<Lyude> hehe, true that...
<alyssa> My problem is I have a knack for finding inane workarounds and then forgetting to fix the bug since "It works fine for me"
<alyssa> Like my literal rube goldberg machine for getting headphones to work on kevin
<Lyude> alyssa: that's not just you, I do that too! :P
<alyssa> :D
<alyssa> Lyude: BTW, do you know if the latest CrOS kernel has the mali_kbase we want?
<HdkR> Oh hey. I do a bluetooth pairing dance every time I want to use A2DP on my headphones
<Lyude> alyssa: no idea, sorry. but it's better to get the repos we've already got (like my mali_kbase repo which I need to add to panfrost's page) working imo
<Lyude> less variance between people's platforms
<alyssa> Mmyeah
<alyssa> I'll be a lot happier once we have X11/Wayland properly working in tomeu's branch
<Lyude> i think we all will :)
<Lyude> myself inclued
<alyssa> :D
<Lyude> *included
<alyssa> (And then I can use non-GBM apps on the mainline machine)
<Lyude> maybe I'll finally have panfrost running on one of my systems!
<alyssa> ^_^
<alyssa> Anyway,
<Lyude> narmstrong: ^ btw, I still have not managed to do that yet
<Lyude> there's something dumb I'm missing, I'm sure of it
* alyssa poofs for homework and other code projects and whatever, will poof back once Wayland is doing its thing :P
NeuroScr has quit [Quit: NeuroScr]
raster has quit [Read error: Connection reset by peer]
stikonas has quit [Remote host closed the connection]
stikonas has joined #panfrost