| 26 May 2021 |
 baloo | :D | 00:02:15 |
| baloo | NOPE | 00:02:21 |
| baloo | not a solution either, not touching those makefiles | 00:02:32 |
 andi- | perhaps just use clang :P | 00:02:41 |
| baloo | https://github.com/gcc-mirror/gcc/blob/master/Makefile.in | 00:03:01 |
| baloo | :D | 00:03:02 |
| baloo | like ... there is a lot of makefile | 00:03:11 |
| andi- | I touched them when I was working on my own nix based package set.. You can do that... it just isn't pretty | 00:03:23 |
| andi- | baloo: > # Makefile.in is generated from Makefile.tpl by 'autogen Makefile.def'. | 00:04:10 |
| andi- | So perhaps it is even easier | 00:04:19 |
| baloo | ho | 00:04:20 |
| baloo | yeah, could be a bit easier | 00:04:28 |
| baloo | but it's still a beast | 00:04:35 |
| baloo | I'll do it eventually, I want to | 00:05:26 |
| andi- | https://github.com/gcc-mirror/gcc/blob/master/Makefile.tpl#L1184-L1208 | 00:05:47 |
| andi- | At the end there is our option to write cp $whatever $debug/[+id+] | 00:06:13 |
| baloo | that could work | 00:07:24 |
| baloo | I completely missed the template | 00:08:06 |
|  samueldr changed their display name from samueldr to . | 00:46:03 |
| samueldr changed their display name from to samueldr. | 00:46:32 |
|  matthewcroughan joined the room. | 01:33:02 |
 Gytis Ivaskevicius | Just droping this off here, may spike up few discussions: https://github.com/NixOS/nixpkgs/issues/123467
I'd imagine that it is very much possible to add additional stage which would take less than 15min to compile | 07:30:52 |
 Jonas Chevalier | Gytis Ivaskevicius: my impression is that you want to bump the boostrap tools to GCC10 | 07:54:10 |
| Gytis Ivaskevicius | At this point there is no point in doing so | 07:54:33 |
| Gytis Ivaskevicius | I am interested in "proper fix" noted in the issue | 07:54:46 |
| Jonas Chevalier | instead of adding another stage, maybe the bump could be simplified | 07:54:48 |
| Jonas Chevalier | that would have the advantage of not adding more build time | 07:55:10 |
| Gytis Ivaskevicius | In reply to @zimbatm:numtide.com
instead of adding another stage, maybe the bump could be simplified I am not sure if I get what you mean? | 07:55:13 |
| Jonas Chevalier | the reason why the bootstrap tools are not to the latest version is that they require some manual intervention | 07:55:53 |
| Gytis Ivaskevicius | In case we are performing "quick fix" (bootstrap bump) - that would mean that it should get bumped every single release (assuming each release uses different gcc) | 07:55:59 |
