ChanServ changed the topic of #freedesktop to: https://www.freedesktop.org infrastructure and online services || for questions about freedesktop.org projects, please see each project's contact || for discussions about specifications, please use https://gitlab.freedesktop.org/xdg or xdg@lists.freedesktop.org
alanc has quit [Remote host closed the connection]
alanc has joined #freedesktop
Leopold_ has quit [Ping timeout: 480 seconds]
Leopold_ has joined #freedesktop
<bnieuwenhuizen_> Anyone with mesa maintainer access able to delete https://gitlab.freedesktop.org/mesa/usermode-queue ? I did an oopsie since apparently new projects for me defaulted to the mesa namespace instead of personal and I didn't check :(
co1umbarius has joined #freedesktop
columbarius has quit [Ping timeout: 480 seconds]
woodwose has joined #freedesktop
i509vcb has quit [Quit: Connection closed for inactivity]
i-garrison has quit []
chipxxx has quit [Remote host closed the connection]
chipxxx has joined #freedesktop
chipxxx has quit [Remote host closed the connection]
i-garrison has joined #freedesktop
danvet has joined #freedesktop
chipxxx has joined #freedesktop
chipxxx has quit [Remote host closed the connection]
chipxxx has joined #freedesktop
robobub_ has joined #freedesktop
<bentiss> bnieuwenhuizen_: done :)
garrison has joined #freedesktop
i-garrison has quit [Ping timeout: 480 seconds]
rudi_s has joined #freedesktop
<rudi_s> Hi. Could you please enable my gitlab.freedesktop.org account (rudis) to be able to fork projects? I'd like to fix a documentation issue in pipewire.
___nick___ has joined #freedesktop
Haaninjo has joined #freedesktop
Lyude has quit [Read error: Connection reset by peer]
Lyude has joined #freedesktop
AbleBacon has quit [Read error: Connection reset by peer]
Haaninjo has quit [Quit: Ex-Chat]
Leopold_ has quit [Ping timeout: 480 seconds]
Leopold_ has joined #freedesktop
chipxxx has quit [Remote host closed the connection]
chipxxx has joined #freedesktop
chipxxx has quit [Remote host closed the connection]
chipxxx has joined #freedesktop
chipxxx has quit [Remote host closed the connection]
chipxxx has joined #freedesktop
Leopold_ has quit [Remote host closed the connection]
Leopold_ has joined #freedesktop
___nick___ has quit []
___nick___ has joined #freedesktop
___nick___ has quit []
___nick___ has joined #freedesktop
pixelcluster_ has quit []
pixelcluster has joined #freedesktop
<Venemo> hi guys, can someone help me understand how this MR can possibly regress the amd-raven-skqp job? https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/22241
MajorBiscuit has joined #freedesktop
MajorBiscuit has quit []
MajorBiscuit has joined #freedesktop
<DavidHeidelberg[m]> Venemo: not flake?
<Venemo> DavidHeidelberg[m]: it happened twice, so I assumed it isn't a flake
<Venemo> but I don't really see how that job can be affected by radv and aco changes, especially because this MR doesn't change anything, only reorders some includes
<DavidHeidelberg[m]> Venemo: could be rare scenario with 2 flakes in row, but skqp is usually pretty stable
<Venemo> what do you suggest? just assign to marge again and see what happens?
<DavidHeidelberg[m]> Venemo: just retriggered job manually and we'll see
<Venemo> thx
<Venemo> what is skqp? does it really use radv?
<DavidHeidelberg[m]> Venemo: skqp is testing suite for Skia library
<DavidHeidelberg[m]> btw. failed again, so you really breaking something :D
<Venemo> okay. I don't see how rearranging some include can change runtime behaviour, but I'll investigate this
<Venemo> is there a way for me to run that job locally?
<Venemo> against my local copy of mesa?
<DavidHeidelberg[m]> Venemo: .gitlab-ci/container/build-skqp.sh :(
MrCooper_ is now known as MrCooper
<DavidHeidelberg[m]> Venemo: configuration: .gitlab-ci/container/build-skqp_base.gn (need to be supplied)
<MrCooper> Venemo: you could check that stripped binaries are identical before & after your change
<Venemo> DavidHeidelberg[m]: sorry can you give me a bit more details than that? For example what is the test suite that is being used here?
<Venemo> MrCooper: can you tell me how to check that, please?
<DavidHeidelberg[m]> it's run trough deqp; backends are usually gl,gles,vk (each around 250 tests)
<DavidHeidelberg[m]> but you can run it by hand for sure
<Venemo> what does "through deqp" mean?
<DavidHeidelberg[m]> options are here: src/amd/ci/deqp-amd-raven-skqp.toml
<Venemo> this looks like it isn't running deqp but its own test suite?
<DavidHeidelberg[m]> (but ofc that's the location inside CI container)
<Venemo> I mean this is not the gl or vk cts, is it?
<daniels> no, skia is a library which is used to perform various graphics operations: https://skia.org/
<daniels> skqp is the set of skia conformance tests, i.e. it tests the behaviour observed when you run particular workloads via skia (like piglit, except it tests skia - which can use GPU drivers to accelerate its work - rather than using gl/es directly)
<daniels> so the tests being run are showing that things that used to work when you ran workloads through skia, no longer work
<Venemo> thanks, I'll look into it. maybe I'll start with fossilize and see if there are any unintended changes there
<daniels> given that you've changed a quad-state enum into a dual-state bool, even though all the direct uses got changed, maybe you have a memcmp() on the entire struct somewhere which is affected
agd5f has joined #freedesktop
<Venemo> yeah, I too suspect the first patch
MajorBiscuit has quit [Ping timeout: 480 seconds]
lynxis has quit [Ping timeout: 480 seconds]
Leopold_ has quit [Ping timeout: 480 seconds]
MajorBiscuit has joined #freedesktop
garrison has quit []
krushia has quit [Ping timeout: 480 seconds]
Leopold_ has joined #freedesktop
AbleBacon has joined #freedesktop
i509vcb has joined #freedesktop
Leopold__ has joined #freedesktop
Leopold__ has quit []
Leopold_ has quit [Ping timeout: 480 seconds]
miracolix has joined #freedesktop
<__tim> I have a spam issue here that's seemingly not being picked up by the damspam bot and I'm not sure why: https://gitlab.freedesktop.org/gstreamer/gst-plugins-base/-/issues/993 (I tagged this 3 days before I registered gst-plugins-base with the bot but I wouldn't expect that to matter, don't think it did in another project)
i-garrison has joined #freedesktop
<slomo> __tim: it only works if you add the label afterwards
<__tim> but for gstreamer-rs we added the label before as well, no?
<slomo> i re-added the label and then the bot did its things immediately
<__tim> ah
<slomo> the way how it works afaiu is that it gets notified when the label is added and then works based on that
<slomo> it doesn't check for existing things with a label
<__tim> oh, of course, d'oh
<__tim> it's hook based not query-based
<__tim> Thanks
Leopold_ has joined #freedesktop
garrison has joined #freedesktop
thaller has joined #freedesktop
i-garrison has quit [Read error: Connection reset by peer]
jarthur has joined #freedesktop
miracolix has quit []
<Venemo> DavidHeidelberg[m]: apparently this breaks skqp too: https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/22393
<DavidHeidelberg[m]> Venemo: ok, first idea: let's see if any MR got merged when Collabora farm was offline which could break this ?
<Venemo> apparently they just happen to break the exact same test
<Venemo> it would seem that the last merged MR did pass the skqp job
<DavidHeidelberg[m]> https://gitlab.freedesktop.org/mesa/mesa/-/issues/8800 s/skqp doesn't look good, probably the fail was introduced during Collabora farm outage
<Venemo> just earlier today
<DavidHeidelberg[m]> hmm, it's flake
<DavidHeidelberg[m]> but probably 90% flake, 10% pass and first seen 1 day ago
<DavidHeidelberg[m]> which could make sense?
<Venemo> in both MR 22393 and MR 22241 the failing test is called gles_lcdblendmodes, does this even use Vulkan? its name would suggest not
<DavidHeidelberg[m]> no, I guess 99% it's not related to your MR
<DavidHeidelberg[m]> it's failing on gles test, probably MR 19234 introduced it, but only as flake, so it passed
<DavidHeidelberg[m]> I pinged Marek on #dri-devel about it
<DavidHeidelberg[m]> let me test it
<DavidHeidelberg[m]> running the test with reverted c7da969f8fe1b78d2479623e2510e52827c3035a, in 10-20 minutes we'll see
<Venemo> thx
lynxis has joined #freedesktop
<DavidHeidelberg[m]> Venemo: " nir: make nir_fisnan helper exact " just passed, it's 100% flake
ximion has joined #freedesktop
krushia has joined #freedesktop
Haaninjo has joined #freedesktop
<DavidHeidelberg[m]> Venemo: just had few runs with reverted patch and after 2 runs got failed without reverted patch. So, it's for sure. Here is fix, including some CI adjusting to current situation: https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/22403
<Venemo> thank you DavidHeidelberg[m]
<Wallbraker> Running into a little strange issue, from what I can tell you can't merge a change to the CI script from a different repo? The source branch needs to be in the same repo. Is this correct? https://gitlab.freedesktop.org/monado/monado/-/merge_requests/1762#note_1862127
<Wallbraker> When I spell it out like that I see that's probably it, and also why.
thaller has quit [Ping timeout: 480 seconds]
___nick___ has quit [Ping timeout: 480 seconds]
<daniels> Wallbraker: it's definitely not correct; the issue is that if you look at https://gitlab.freedesktop.org/monado/monado/-/pipelines/853062 you can see that none of the container-build jobs are scheduled to run, which points towards an issue with the rules that govern how they're run
<Wallbraker> Hmm okay thanks
<Wallbraker> Could it have something to do with running it in a repo first or something?
<daniels> nah, not at all
<daniels> at a guess, you never rebuilt a container since you added the workflow: rules: stuff to do builds in MR context
<daniels> and now you have both rules and only applying, which are mutually exclusive, so you need to rewrite your only: changes: to be based on rules instead
<Wallbraker> Okay
<Wallbraker> Is there some documentation for this, other then "gitlab"?
<Wallbraker> Thanks for the info!
<daniels> https://docs.gitlab.com/ee/ci/yaml/ covers the CI syntax p well, including the incompatibility between rules vs. only/except, and some examples for using rules
<Venemo> DavidHeidelberg[m]: one of my MRs got merged at the next attempt now
<Venemo> so you are right, it does seem to be a flake after all
trinitronx has joined #freedesktop
trinitronx has quit []
trinitronx has joined #freedesktop
<Wallbraker> Thanks again.
<daniels> np
MajorBiscuit has quit [Quit: WeeChat 3.6]
danvet has quit [Ping timeout: 480 seconds]
<alanc> https://lists.freedesktop.org/archives/freedesktop/2023-March/000491.html seems like a gitlab admin problem mailed to the wrong list
damian_ has joined #freedesktop
Haaninjo has quit [Quit: Ex-Chat]
damian has quit [Read error: Connection reset by peer]
damian has joined #freedesktop
damian_ has quit [Read error: Connection reset by peer]
Haaninjo has joined #freedesktop