03:57nektro: libX11.so: undefined reference to '_XlcGenericLoader'
03:57nektro: would this be caused by an issue in libX11 or the caller? (eg out of date function call)
04:01nektro: google says it mayyy be locale related
04:08nektro: needed to add --enable-loadable-i18n to libx11
11:00gry: nektro: could #xorg-devel help?
15:26DavidHeidelberg[m]: mupuf: Hey, something is broken in Valve farm again: https://gitlab.freedesktop.org/mesa/mesa/-/jobs/40027792
15:29Wallbraker: Can somebody explain to me what ".rules-anchors:" does?
15:30Wallbraker: And how I can easily disable the CI jobs to run in forks of the Monado repo?
15:44mupuf: DavidHeidelberg[m]: I doubt it is the farm. Looks like a regression to me
15:44DavidHeidelberg[m]: seeing symbol lookup errors; but when I looked at daily, it seems it also run some tests with killing the GPU
15:45DavidHeidelberg[m]: I suspect some recent RT changes I seen, but cannot look into it now
15:46mupuf: I reported the regression to radv devs
15:46mupuf: As for the random GPU hang, they are also looking into it
16:05mupuf: DavidHeidelberg[m]: is it possible for the test and build containers to use different libdrm versions?
16:06mupuf: Sorry, still on a paternal leave until Monday
16:06DavidHeidelberg[m]: oh, don't worry about it then.. anyway it's weekend :D
16:06DavidHeidelberg[m]: also. Until Debian 12, it's possible, since we build and supply own libdrm
16:07DavidHeidelberg[m]: but in Debian 12 I partly done and I plan to drop custom libdrm builds
16:10mupuf: Oh...
16:11mupuf: Recipe for disaster... Libdrm should be in install.tar if we want to have custom libdrms
16:11DavidHeidelberg[m]: I think we have it in rootfs
16:13mupuf: I see. Hopefully we can unify these two paths soon, and remove the duplication
16:19DavidHeidelberg[m]: mupuf: working on it as part of bookworm uprev
16:20DavidHeidelberg[m]: also I was thinking we could have some custom Debian repo a top of Bookworm, with manual "uprev" of this type of packages like libdrm
16:20DavidHeidelberg[m]: so we get it distributed across all containers/rootfs etc.
16:21mupuf: Cool!
16:22mupuf: That would be one way to share it, indeed :)
16:25mupuf: As for the rootfs, it would be great if it could just be an extract of the test containers + wget a kernel from gfx-ci/linux
17:06DavidHeidelberg[m]: Where is my priviledges?! "Looks like you don't have enough privileges, please see https://gitlab.freedesktop.org/freedesktop/freedesktop/-/issues/540"
17:08DavidHeidelberg[m]: it just worked 5 minutes ago
17:09DavidHeidelberg[m]: ... my bad, it probably wasn't. I was triggering the pipeline from MR
17:09DavidHeidelberg[m]: but it was working ... day ago or something for sure
17:42daniels: Wallbraker: it does nothing - the section has no effect in and of itself but the individual rules can be referenced from other parts
17:43Wallbraker: Ah thanks, it took me a little while to understand that was going on. Since I had no knowledge about yaml before having to deal with the CI fallout.
17:44Wallbraker: Also YAML is dumb as it doesn't seem to be a way to combine arrays of strings.
20:09daniels: yeah, the [!reference] syntax can work around that