s/Elf/ELF/, I think p_elf_brandinfo should be saved and cleared before img activator
work, and restored on error like p_osrel.

asserts fixed, tinderboxed

done

done

done, nice, thank you)))

fixed, i missied that KERN_SUCCESS != 0

indeed, fixed, now its good.

'all rights' removed, adapted to the previous changes

fixed, thanks!

fixed copyright, arm64 write()

fixed, thank you,
one question, in D30900 I'll change this, sv_onexec() will return an error,
should I call vm_object_deallocate(obj) or somthing other. seems that
on if vm_map_stack() return error no resources are released

here was a mistake, now it's ok, runtime tested.

fixed, thanks!

comment added

move linux_on_exec() call below, after the stack is created

something like this?

In D30899#695137, @kib wrote:

Does the move after process event handler breaks existing linux_on_exec()?

In D30900#695126, @emaste wrote:

ifunc can't be used

For any cases where we would want ifuncs we could "just" select the appropriate implementation in kernel, right? (Subject to the caveat that this might require a lot of additional infrastructure, might not be worth the effort, etc.)

In D30903#695122, @emaste wrote:

process VA layout has changed

This is a recent change? Can you link the rev or review for it?

In D30332#691452, @pitwuu_gmail.com wrote:

In D30332#691205, @dchagin wrote:

Im don't understand what do you mean, I mean that musl it's a separate brand, see Elf_Brandinfo and linux_sysvec.c,
and only for musl brand we should change cmp_requeue, for glibc brand cmp_requeue should return EINVAL

Well. No. Musl doesn't have different brands for the elf files, only the path of the interp differs, and in any case we are a Linuxulator, not a glibculator, it's better to just remain agnostic of the userland and let it sort itself out like is expected of the kernel.

If there's a quirk for some ancient glibc version then make that a knob somewhere or break it in favour of modern (read from this decade) versions.

In D30751#691414, @rpokala wrote:

To make sure I understand this: this change is to no longer allow Linuxulator to be built into the kernel, and only allow it to be built as a module?

fixed Ravi's comments. thanks!

emaste@, what are you think about using fget_unlocked() in linux_to_bsd_poll_events?

Im don't understand what do you mean, I mean that musl it's a separate brand, see Elf_Brandinfo and linux_sysvec.c,
and only for musl brand we should change cmp_requeue, for glibc brand cmp_requeue should return EINVAL

move conversion routines declaration to linux_common.h as
linux.h does not intended for MD things

fix grammar, rate limit diagnostics.

In D30597#690019, @trasz wrote:

(Tinderboxed.)