| 25 Jun 2024 |
 Alex | If you're not building any Haskell packages on purpose, you should try disabling ShellCheck (by default, it is used to check all shell scripts). | 13:21:46 |
 Mic92 | In reply to @thefossguy:matrix.org
I realize that. What I'm asking is if you had a list of blockers for bootstrapping riscv64 GHC on x86 Ah. There it's basically ghc not publishing bootstrap binaries afaik. | 13:22:10 |
| Mic92 | Yes. Disabling shellcheck is the way to go. | 13:22:31 |
| Alex | In reply to @thefossguy:matrix.org
I realize that. What I'm asking is if you had a list of blockers for bootstrapping riscv64 GHC on x86 Nixpkgs generally relies on upstream bindists, but there are a few (as-of-yet unmerged?) PRs to also make it possible to bootstrap via cross. | 13:22:33 |
| Mic92 | Can we not also upload our own cross-compiled ghc binary tarball? | 13:23:19 |
| Mic92 | to tarballs.nixos.org? | 13:23:24 |
 Pratham Patel (you can mention me) | I'm not sure. Only now starting to understand Nix after being 1+ years in the Nix ecosystem lol | 13:23:57 |
| Pratham Patel (you can mention me) | Okay, my custom minimal ISO started building :) | 13:24:25 |
| Alex | In reply to @joerg:thalheim.io
Can we not also upload our own cross-compiled ghc binary tarball? That is definitely worth discussing with the Nixpkgs maintainers, but it'd be easier once the changes that enable cross-compilation of GHC are merged.
The PRs that enable it also add x86_64 -> riscv64 cross-compilation to Hydra to catch regressions. | 13:24:49 |
| Alex | So once the changes are merged, the Hydra build could be downloaded and packaged into a tarball to bootstrap native builds (or, more likely, used to package a complete native build). | 13:26:05 |
| Mic92 | Also sounds good to me. | 13:26:24 |
| Alex | The newest PR for this is: https://github.com/NixOS/nixpkgs/pull/305392
(It works already, but I think sterni might want to make a few changes before it is merged?) | 13:27:50 |
|  axiomss joined the room. | 21:57:12 |
| 27 Jun 2024 |
|  BurntKrispe joined the room. | 16:20:48 |
 fgaz | https://github.com/openresty/luajit2/pull/236 | 19:59:36 |
| 28 Jun 2024 |
| axiomss left the room. | 04:12:59 |
 atka | risc-v boards have arrived! | 17:35:16 |
|  Janik (they/them) left the room. | 18:20:53 |
 hexa | all of them? | 19:30:40 |
| atka | well the boards I ordered :), first time hands on with real hardware | 20:23:30 |
 [0x4A6F] | Nice, have fun. Which boards do you hav? | 20:54:54 |
| [0x4A6F] | * Nice, have fun. Which boards do you have? | 20:55:02 |
| atka | [0x4A6F]: two milk-v duo 64M, 1x milk-v duo 256M, and an ethernet/io board | 21:30:13 |
| atka | converting some old rpi and microcontroller projects over, playing around with porting a drm driver to risc | 21:31:03 |
| atka | dabbling in some assembly | 21:31:22 |
| 29 Jun 2024 |
|  maka_77x joined the room. | 00:48:18 |
| BurntKrispe changed their display name from burntkrispe to BurntKrispe. | 05:28:39 |
| atka | Nick Cao: thanks for the work on the Duo64 and Duo256! | 21:30:26 |
| atka | is ION size supposed to be set to zero by default for these images? | 21:39:56 |
| atka | [root@nixos-duo:~]# cat /sys/kernel/debug/ion/cvi_carveout_heap_dump/summary
Summary:
[0] carveout heap size:28102656 bytes, used:0 bytes
usage rate:0%, memory usage peak 0 bytes
Details:
heap_id alloc_buf_size phy_addr kmap_cnt buffer name
minimum ion allocate unit = 4096
free memory regions:
heap_id start end length
0 8e333000 8fe00000 28102656
| 21:40:10 |
