| seb128 | cpaelzer_, hey, sounds like that libaio delta could be forwarded to Debian, it shouldn't make a difference for them unless they decide to start using lto in which case it makes sense to have there | 08:04 |
|---|---|---|
| cpaelzer_ | seb128: I was already working on that :-) | 08:09 |
| seb128 | cpaelzer_, š | 08:10 |
| cpaelzer_ | seb128: but their update speed on this is slow, so I didn't expect much nor block on it | 08:10 |
| seb128 | right, it's just annoying when we carry delta for things like that :-) | 08:10 |
| === cpaelzer_ is now known as cpaelzer | ||
| seb128 | not your fault, you are not the one who added it, those should be forwarded to Debian from the start | 08:10 |
| cpaelzer | TBH I'd need to check if optimize=-lto is a noop for them or if we need dpkg-vendor based switch for this | 08:11 |
| cpaelzer | I feel I might have checked this before but forgotten the outcome over the sprint | 08:11 |
| seb128 | cpaelzer, right, I still think we should have used the blacklist for those packages to avoid adding delta and extra work but that's another topic | 08:14 |
| cpaelzer | I can see why doko_ felt that if we do that even for main packages no one will care to resolve them | 08:14 |
| cpaelzer | and I think that is right | 08:14 |
| cpaelzer | althought it indeed triggers some busy work here and there | 08:15 |
| seb128 | I don't see any sign of forcing us to have delta pushing people to do extra work | 08:19 |
| cpaelzer | the pain to merge it? | 08:19 |
| seb128 | we just merge rules changes to optimize=-lto | 08:19 |
| cpaelzer | as I said i was working on upstreaming it already to get rid of seeing it next cycle | 08:19 |
| seb128 | well, for desktop we got the -lto changes in Debian and are back in sync | 08:19 |
| cpaelzer | since I wasn't initially adding it this moment is when I'd pick it up | 08:20 |
| cpaelzer | seb128: you are right after submitting it there it is kind of off-the-table for us | 08:20 |
| seb128 | my point is that I didn't see any of those lto delta turned into working with upstream to fix the lto builds | 08:20 |
| cpaelzer | agreed, only some bug filing/tracking but not adding LTO support there | 08:20 |
| didrocks | slyon: hey, Iām unsure if you had time to look at my answers on bug #1948392 and if so, you have time to ack/nack it :) | 08:51 |
| ubottu | Bug 1948392 in wsl-setup (Ubuntu) "[MIR] wsl-setup" [Undecided, New] https://launchpad.net/bugs/1948392 | 08:51 |
| slyon | didrocks: I was out a few days. Will look into it soon, thanks for the heads up! | 08:51 |
| didrocks | thank you :) | 09:01 |
| === doko_ is now known as doko | ||
| paride | juliank, hi! In comment 3 here: https://bugs.launchpad.net/ubuntu/+source/grub2/+bug/1935659 did you mean "source-copyable"? | 10:56 |
| ubottu | Launchpad bug 1935659 in grub2-unsigned (Ubuntu Impish) "grub2 FTBFS when built with libfuse3-dev" [Undecided, New] | 10:56 |
| juliank | paride: i meant what I wrote | 10:57 |
| juliank | paride: the world changed though | 10:57 |
| paride | juliank, ok :-) then I have to think again. | 10:57 |
| juliank | paride: the goal of grub2-unsigned was to be able to just binary-copy it back like shim, but then things changed because bionic needed special work | 10:58 |
| juliank | paride: binary copy is easy for grub2-unsigned, it does not use libfuse at all | 10:58 |
| juliank | because grub2-unsigned only builds the EFI binaries, not userspace | 10:59 |
| paride | yeah that's the bit I was missing. I think I understand now, thanks. | 10:59 |
| === mfo_ is now known as mfo | ||
| === genii-core is now known as genii | ||
| === cpaelzer_ is now known as cpaelzer | ||
| === genii is now known as genii-core | ||
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!