Conversation
SUSE #hackweek going great!
1
1
6
With the help of serial console I now know the problem is... insufficient memory for malloc(). Great, I thought I would be doing something different than usual for Hackweek, and here I am.
2
0
6

@vbabka at least you've got the relevant skillset

1
0
1

@vbabka or is this a "when what you have is a hammer..." situation?

1
0
1
this thing has only 256k RAM... time to resurrect SLOB?
2
0
4
@vbabka Linux on Atari 8-bit was not a joke?
1
0
2

@vbabka time to implement pageable slabs

1
0
1
@jann @vbabka time to implement that april 1st patch that removed all slabs
0
0
1
@pavel it's a RPi Pico in the picopad device, so no Linux or other OS ;) hacking on the C64 emulator for it.
3
0
2

HAMMER SMASHED FILESYSTEM πŸ‡ΊπŸ‡¦

@vbabka @pavel oh, so 264kb ram actually *fish shriek* 🌊πŸͺ“πŸŸ

0
0
2
@vbabka @pavel it's sort of heroic and shit to support these museum pieces, but it's a constant pain in the arse for work on heavily-in-use arches which actually really need a constantly moving kernel.

If people can maintain this stuff, they can do it outside of mainline and make whatever compromises they want.

I think it's just absolutely unfair for core kernel devs to be expected to do the work.
1
0
1
@ljs @vbabka Maybe your employer can fork the kernel if that's so unfair? :-(.
1
0
0
@vbabka I'd really like to see picopad emulation. I tried to hack qemu to do that, but it is a lot of work.
0
0
0
@pavel @vbabka I speak only for myself (look in my profile) and this represents the views of many kernel developers.

I find this response unpleasant, so I'm blocking you.

EDIT: Reduced spice.
1
0
0
@ljs @pavel @vbabka With my admin hat on, I am putting an end to this and invite you both to take a breath and cool off, before I start revoking accounts.
1
0
1
@admin @pavel @vbabka I'm not sure why I deserve a warning when I'm getting facetious and mean comments aimed at me out of the blue?... I blocked Pavel in response.

This is after a lot of very unpleasant comments I let go.

I can edit the reply to be less spicy but honestly that's a little unfair.
1
0
0
@ljs @pavel @vbabka "insulting/derogatory comments, and personal attacks" -- straight out of the code of conduct, which you've all agreed to adhere to on this instance.
2
0
0
@vbabka @pavel i need to check for fun what are the lowest specs for atari m68k. i possess STE with 4 megs and ultrasatan ;-) maybe i could create a config for it to my buildroot based shenanigans.
1
0
0
@admin @pavel @vbabka ok apologies, you are right about that.

I overreacted perhaps, because it was an unwarranted, mean-spirited and unpleasant attack out of the blue and part of a series of unpleasant comments I've received for no (apparent) reason.

I will refrain from being quite so on the nose like that again.

EDIT: I edited the original toot to remove this bit too.
0
0
1
@admin @ljs @vbabka For the record, would you mind explaining which part of my comments do you find objectionable? I believe "no regressions" is a long term kernel rule, and while it may be inconvenient to companies, it is existing policy. I'd prefer that rule to stay.
0
0
1
@jarkko @vbabka I feel 4MB is too little for anything other than busybox these days.
1
0
0
@pavel @vbabka that's about what i run in my test images. actually i do have bash because kselftest requires it so that might be problem :-) it can't be worse performer than running RISC-V on FPGA, which I did for work purposes at some point...
1
0
1
@jarkko @vbabka It may be good enough for that with some luck. I actually have 9MB 386sx here. Too bad Linux stopped supporting 386 some time ago :-(.
1
0
0
@pavel @vbabka i've tried to give feedback in some reviews to make sure that test script would be busybox ash (i.e. #!/usr/bin/env sh) compatible but some others have downplayed that feedback ;-) at least i've tried...

especially with preproduction hardware it is not uncommon start with fpga so it would somewhat feasible to reduce image size for kernel + kselftest... because loading is SLOW
1
0
1
small extra trouble which would widen embedded coverage so much more. just don't get it tbh. data center infrastructure mob is dominating the linux upstream i guess ;-)
0
0
0