00:50 karolherbst: ben wtf.. dropping a 150 patch series like nothing happened
00:50 karolherbst: https://patchwork.freedesktop.org/series/132523/
00:53 airlied: I wonder how long it'll take phoronix to spot the email address :-P
00:53 karolherbst: Lyude: fyi^^
00:53 karolherbst: I already made a shitpost on fedi
00:54 karolherbst: though no idea if Michael follows me there with a secret account
00:55 karolherbst: tbh, should have done it on April 1st
00:55 karolherbst: Ben I mean
00:55 dwfreed: 306 files changed, 6653 insertions(+), 10697 deletions(-)
00:55 dwfreed: jesus
00:55 karolherbst: well.. refactoring be like that
00:55 fdobridge: <a​irlied> gets rid of a horrible internal interface
00:55 karolherbst: yeah...
00:56 karolherbst: this should make it way easier to hack on nouveau and actually debug it
00:58 dwfreed: fyi, that patchwork page says it is missing patch 25
00:58 karolherbst: mailman moments
00:59 dwfreed: based on the cover letter, that should be "drm/nouveau/nvkm: remove perfmon"
01:00 karolherbst: yeah.. I think Ben is here on IRC under a cover name...
01:00 karolherbst: so uhm.. Ben one of your patches is missing
01:02 airlied: is it in lore?
01:02 airlied: it might just be delayed in dri-devel moderation due to size
01:03 airlied: oh nouveau probably isn't in lore
01:03 dwfreed: it is
01:04 dwfreed: but I asked lore for that subject across all mailboxes, and have no results other than the cover letter
01:05 dwfreed: (that was an impressively fast query)
01:06 airlied: ah yeah patch 25 is missing there as well
01:49 fdobridge: <c​onan_kudo> So much for Ben retiring into the sunset...
01:50 fdobridge: <c​onan_kudo> (as an aside, patchwork is still terrible)
05:19 fdobridge: <g​fxstrand> I'm hoping to make them available to anyone who wants one. I'm just waiting on my samples to come so I can be reasonably sure they're decent shirts.
12:41 fdobridge: <s​unrise_sky> Isn't this a bit too late for an April Fools joke? https://www.phoronix.com/news/Ben-Skeggs-Joins-NVIDIA
12:43 fdobridge: <k​arolherbst🐧🦀> correct
12:44 fdobridge: <m​tijanic> IIRC there was a suggestion to drop them on April 1st.
12:44 fdobridge: <m​ohamexiety> it would've been glorious on april 1st
12:44 fdobridge: <k​arolherbst🐧🦀> Ben is too much of a perfectionist to drop non perfect code though 😛
12:44 fdobridge: <k​arolherbst🐧🦀> even as a joke
12:45 fdobridge: <s​unrise_sky> Wow
12:45 fdobridge: <s​unrise_sky> So all the drama around Ben leaving was for nothing?
12:45 fdobridge: <k​arolherbst🐧🦀> well.. Ben left RH
14:36 fdobridge: <z​mike.> is there an nvk equivalent to RADV_DEBUG=zerovram or is vram zeroed by default?
14:42 fdobridge: <m​henning> @zmike. I believe it's NVK_DEBUG=zero_memory
14:45 fdobridge: <z​mike.> ah
14:45 fdobridge: <z​mike.> of course
14:46 fdobridge: <z​mike.> now I'm failing consistently
14:49 fdobridge: <k​arolherbst🐧🦀> @gfxstrand I think `cbindgen` needs to be required, no?
14:50 fdobridge: <k​arolherbst🐧🦀> otherwise you get `../src/nouveau/nil/meson.build:104:9: ERROR: Tried to use not-found external program in "command"`
14:50 fdobridge: <g​fxstrand> Yes it does
14:50 fdobridge: <k​arolherbst🐧🦀> because atm it's `required : false`
14:53 fdobridge: <m​ohamexiety> yeah I needed to install it standalone, it wasn't autoinstalled by meson
14:55 fdobridge: <g​fxstrand> Auto-RB if you care to fix it
15:07 fdobridge: <p​homes_> Could be this issue with zero_memory? https://gitlab.freedesktop.org/mesa/mesa/-/issues/10800
15:15 fdobridge: <z​mike.> no idea, I'm just using it for cts debugging
16:50 fdobridge: <z​mike.> incredible, that GL test splitting found a zink bug
16:50 fdobridge: <z​mike.> regrettable, the zink bug was not causing all the nvk sparse hangs
18:08 bl4ckb0ne: what would be the best approach to fix the issue in meson where cbindgen is not found? a graceful handle with a user friendly message ou just put it as required: true and call it a day
18:11 fdobridge: <p​avlo_kozlenko> this is a special operation for obtaining PMU, firmware, and architecture documentation. It's all planned
18:12 fdobridge: <p​avlo_kozlenko> https://cdn.discordapp.com/attachments/1034184951790305330/1230219262518624338/zIuHUe2h_400x400.png?ex=663285f8&is=662010f8&hm=79bc5a7743ae098a3bec63290e1c98805f3a93435ecaf39c25ee561f2ba23ab4&
18:12 fdobridge: <p​avlo_kozlenko> 😂
18:21 fdobridge: <p​avlo_kozlenko> https://cdn.discordapp.com/attachments/1034184951790305330/1230221702819086427/image.png?ex=6632883e&is=6620133e&hm=a84ba37ab9bae816c8e61d145f4c6d74563831a0997f4125441476ac1554bc22&
18:43 fdobridge: <b​abblebones> Ben on that triple agent insider biz
18:56 Lyude: i swear every single time i come back to try something on my testing desktop someone has caused a regression somewhere in the kernel's early boot code
18:56 Lyude: it's -always- just this machine too
18:58 Lyude: rc3 too yikes
20:42 Lyude: skeggsb: welcome back ♥
20:44 skeggsb: Lyude: hey! thanks :)
20:45 RSpliet: skeggsb: yeah what Lyude says!
20:46 RSpliet: I defo feel bamboozled
20:46 RSpliet: But glad to see you contributing to the project again!
20:46 Lyude: as long as I still get to write a display driver I'm happy :)
20:46 Lyude: and rust
20:47 skeggsb: RSpliet: it was a bit of an unexpected turn of events from my perspective too :P
20:48 RSpliet: haha is this story time? or should I be trying to attend XDC again for a change and feed you beer to try and pry stories from you?
20:49 karolherbst: skeggsb: welcome back :)
20:49 karolherbst: and here I assumed you wanted to take a longer break :D
20:50 skeggsb: RSpliet: there's not much of a story really
20:51 skeggsb: karolherbst: me too haha :P
20:52 skeggsb: a friend and i were chatting about what my next steps might be on nye, somehow the possibility of nvidia came up, and i decided to send an email a day or two later. skip forward a couple of months, and here we are!
20:53 RSpliet: did they get you on the payroll for nouveau/open source drivers? or more of a mixed bag of work?
20:53 karolherbst: heh :D what a turn of events anyhow. I know that nvidia was more reluctant in the past on hiring like developers working on nouveau tbh
20:55 RSpliet: karolherbst: I could imagine that the change of heart could in part be by NVIDIA now becoming an IP licenser rather than just building their own SoCs
20:55 RSpliet: did you read https://corp.mediatek.com/news-events/press-releases/mediatek-brings-advanced-ai-capabilities-to-vehicles-with-new-dimensity-auto-cockpit-chipsets-enabled-by-nvidia-technology ?
20:57 RSpliet: (although technically it could be a chiplets solution, the announcement doesn't specify)
21:02 fdobridge: <r​edsheep> That would be really surprising if that means IP licensing. It doesn't really read that way to me though, there are a few ways this could work. It could even just be attached with pcie.
21:03 fdobridge: <r​edsheep> If anything if it's one chip it would sound more likely to me that it's Nvidia doing the actual design, maybe incorporating some mediatek ip, not the other way around
21:07 fdobridge: <r​edsheep> Like what AMD does with semi-custom
21:09 fdobridge: <g​fxstrand> bl4ckb0ne: Yeah, we just need to switch the `find_program()` in `meson.build` to `required : true`. I'll do that when I get home.
21:09 fdobridge: <g​fxstrand> Or someone else can make the MR and I can merge it.
21:14 bl4ckb0ne: im cooking dinner, i can take care of that tomorrow
21:16 fdobridge: <g​fxstrand> In that case, I'll fix it up tonight. I think I'm probably going to leave the coffee shop soon
21:20 fdobridge: <g​fxstrand> Sorry stuff is so slow going in right now. Having my test machine tied up with GL 4.6 is making it hard to do anything that nees more than build testing.
21:21 fdobridge: <g​fxstrand> So I've been farting around with other stuff while I wait.
21:22 Lyude: yay, regression found
21:23 bl4ckb0ne: dont worry, i just caught that while trying to build mesa on bookworm
21:23 skeggsb: RSpliet: my initial focus will be towards nouveau/nova, and i'll see where things go from there! lots to learn and explore, hopefully an exciting time ahead!
21:40 Lyude: oh - the early boot regression my little persnickity desktop found if anyone was wondering: https://lore.kernel.org/lkml/3d77cb89857ee43a9c31249f4eab7196013bc4b4.camel@redhat.com/T/#u