00:02 <bwidawsk> can I control ulimit in the docker image?

00:11 <daniels> bwidawsk: what for? you can run ulimit within the container context, or if you're invoking it yourself you can pass various options to set ulimits as it's invoked

00:12 <bwidawsk> daniels: Hmm, I thought ulimit is typically not allowed. I have a case where file descriptors are leaked and it's too much work to fix at the moment

00:12 <bwidawsk> so I just want to raise the limit

00:12 <bwidawsk> I can try it though, maybe I was overly cautious

00:12 <daniels> ah, raising might not work indeed

00:14 <bwidawsk> I realize it's not the ideal reasoning, the higher fd limit... but it is what it is

05:56 <airlied> hey mesa issue 3860 is a spam

06:05 <liyi__> hi

06:05 <liyi__> hi All, I send Linux kernel patches to dri-devel@lists.freedesktop.org using liyi@loongson.cn, but all the emails is refused by the server, the Bounce reason is Can not connect to recipient's server because of unstable network or firewall filter. rcpt handle timeout,last handle info: Can not connect to lists.freedesktop.org:2610:10:20:722:a800:ff:fe36:1795:25, the Bounce Code is can not connect to , but my other emils is OK. I don't know Who I

06:05 <liyi__> <liyi__> Shoud Contact, anyone can help me? Thank you

06:25 <mupuf> airlied: the whole issue?

06:26 <mupuf> I see 2 obvious spamming users, I'll ban them

06:26 <mupuf> s/2/3/

06:31 <mupuf> I removed the last 3 comments and their associated users which had no other activity

06:32 <airlied> mupuf: oh maybe just the last 3 comments I saw in my email

07:36 <daniels> liyi__: if I had to guess, it’s an IPv6 issue and it would be fixed by asking the Loongson admin to force IPv4

07:37 <daniels> airlied: please hit the report spam button on the comment with that happens

13:24 <Venemo> random freedreno CI fail: https://gitlab.freedesktop.org/mesa/mesa/-/jobs/36501435

17:14 <anholt_> I've been seeing intermittent container build fails recently https://gitlab.freedesktop.org/mesa/mesa/-/jobs/36481788#L34

17:34 <daniels> anholt_: yeah, /app/cbuild contains ‘Internal server error’ after the fetch (note the 30s duration) then we try to execute it

17:44 <eric_engestrom> feels like there should be multiple ways to detect this and retry the download (`--retry-all-errors` and friends, or at least check the resulting file)

17:48 <daniels> yeah, we did that with Mesa; it requires using curl since that's the only one which can be robust against everything (DNS fail, connect fail, HTTP errors)

17:51 <eric_engestrom> ah, this is from the template, isn't it? I was confused by your "with mesa" when I was looking at a job in mesa/mesa 😅

17:55 <daniels> yeah, that's from ci-templates core

18:17 <alanc> "Something went wrong during merge pre-receive hook. Internal API unreachable. Try again." - haven't seen that before when clicking merge on an MR

18:17 <alanc> seems to have worked on the second try though

23:38 <DavidHeidelberg[m]> daniels: bentiss how much pain is having uploaded extra ~ 500 M (~ 250M unstripped mesa, ~ 250M debug symbols) per Mesa build x7 (arm64, armv7 x86-64, *asan, CL build)... that would be around 3.5G per pipeline run extra.

23:39 <DavidHeidelberg[m]> I guess some builds are bit smaller, lets say 2.5 - 3G