Solar Designer
Solar Designer
> I re-ran those two jobs after your previous comment, and only 1 of them failed. I re-ran the one still failing job since, and it failed again. So far...
> Curious why we don't do full verification on `schedule()` just pCFI, @solardiz do you happened to know/remember? I don't remember such discussions (nor whether we had any), but I...
> looks like the failing build we have is on the code related to your NET module. Can you check this? Hmm, yes. I wonder if ALT is somehow quicker...
Thank you @Adam-pi3 for investigating this! > corner case with IRQ and WQ where a last frame (what is interesting it looks like a valid frame) has IP/FP set to...
> I wonder if ALT is somehow quicker to update to a newer mainline rc kernel than Fedora is? Looks so. The Fedora test is still with 6.18.0-65.fc44.x86_64.
> @solardiz can you merge this PR if there is no more feedback? :) Yes, I can merge it now, unless you want to remove `P_KERNEL_LAST_UNUSED_HOLE` from the header first?
@Adam-pi3 Thanks. Let's merge this. If this were my PR, I think I'd squash these 4 commits into 1 or maybe 2, but I don't mind keeping them as-is. BTW,...
Thank you @vt-alt! I am now dealing with 6.19 portability issues in #452.
Reverting d265b3cf097510c26e80bd0d18b5779605b34c8e repairs it.
This hack repairs it (just to narrow down where the problem is): ```diff +++ b/src/rules.c @@ -1863,8 +1863,11 @@ char *rules_process_stack_all(char *key, rule_stack *ctx) rules_stacked_after = 0; while (ctx->rule) {...