Re: binaries not portable from Debian to RedHat

From: Kasper Dupont (kasperd_at_daimi.au.dk)
Date: 07/16/05 

Date: Sat, 16 Jul 2005 16:04:58 +0200

David Schwartz wrote:
>
> That, or you can't have a 'ps'.

su -c 'rm /bin/ps'

Problem solved. Somehow I doubt any standard says
you are not allowed to have a ps command, if there
isn't a way to tell a difference between threads
and processes. In fact in Linux they are the same.

>
> >> It's based on the definition of "process". Modern "ps"
> >> implementations
> >> are smart enough to know that they're supposed to show *processes*, not
> >> KSEs. Older ones aren't so smart.
>
> > Try an old ps on a new kernel, and try a new ps on
> > an old kernel.
>
> Eww. I hope they didn't 'fix' the kernel, because there really wasn't
> anything wrong with it.

No, there was nothing wrong with it. It worked
perfectly fine, it just wasn't complying with any
standards. So a few changes were made to make it
compliant with Posix.

> The ideal solution would be for the kernel to
> identify which KSEs share vm and for 'ps' to use this information to prepare
> the correct display.

Could be handy, but AFAIK the kernel does not any
way to tell if two tasks share vm.

> Configuring the kernel to suppress what threads is,
> IMO, an inferior solution.

Seems like you actually don't know how it works. 

-- 
Kasper Dupont -- der bruger for meget tid på usenet.
Note to self: Don't try to allocate 256000 pages
with GFP_KERNEL on x86.

Relevant Pages

  • Re: Accessing memory at 0xb800
    ... FYI, "10h" isn't a standard hardware interrupt, it's a BIOS interrupt. ... Unless you're hacking your kernel, kiss your BIOS interrupts goodbye, ... a privileged mode, you basically have to write your own kernel (or hack ...
    (alt.comp.lang.learn.c-cpp)
  • Re: Accessing memory at 0xb800
    ... FYI, "10h" isn't a standard hardware interrupt, it's a BIOS interrupt. ... Unless you're hacking your kernel, kiss your BIOS interrupts goodbye, ... a privileged mode, you basically have to write your own kernel (or hack ...
    (comp.lang.cpp)
  • Re: [PATCH 6/6] sched: disabled rt-bandwidth by default
    ... For this, if this time limit does kick in, we should at the very least ... The kernel should not set policy but provide capabilities. ... really good reason to show why the standard is wrong. ...
    (Linux-Kernel)
  • Re: Why bigForth is Linux only
    ... different operating systems. ... They differ _very_ much at the kernel ... Because some 20 year old standard mandates that? ...
    (comp.lang.forth)
  • Re: Setting up KVM?
    ... I did that with the Standard PC, but there's several drivers it can't find. ... In the 2.6.22 kernel they moved rtc to hpet, and I still get the same ... Once you get the machine running and can log into windows goto ... something to do with an acpi interrupt that windows xp uses and cause ...
    (Fedora)