PLEX86  x86- Virtual Machine (VM) Program
 Plex86  |  CVS  |  Mailing List  |  Download  |  Linux  |  Newsgroups

Shutting down User Mode Linux


Your Ad Here

Your Ad Here

Because people have poor expression and comprehension, in general.

Please do not quote out of context! I

modutils on crack" followupmostly solved
On Jan 24, 2005, I posted "2.6.10--modutils on crack? also 15-pin joystick" in Message-ID I've done a great deal of fiddling around with the problems described in the original...

If you want to ask a sensible question, try something like "my uml kernel process enters state D when I run shutdown -h. How can I best find what is it waiting on?". And post to the UML list.

You are not understanding. When a process does not die immediately when signalled with -9, it is not because the process is at fault, but because the kernel is broken-stuck-whatever. The process has not even received the signal. The kernel has not acted on it. As you have said, the process is in kernel code. The process itself, the userspace code, is working perfectly correctly.

No you don't get it! That's OK.

What do you call

If you want to ask a sensible question, try something like "my uml kernel process enters state D when I run shutdown -h. How can I best find what is it waiting on?". And post to the UML list.

??

I even gave him this hint:

I too have experienced lock ups-processes entering state D in user level applications (not uml, of course) in the early kernel 2.6 series through BADF from a mmap call, but I have not investigated.

There is no "holy wrath"! And no misleading information! Honestly, I think a misperception arises from a preexisting fixation of yours that people in general tend to say that user processes must die when -9 is sent, when in fact there may be an arbitrarily long temporal delay (perhaps forever) until they "die"-are removed from the process table.

I think the misperception arises from a peculiar and distinct lack of understanding of your own of the logic and its consequences. You understand that the process is in the kernel, and the kernel code it is in is uninterruptible, but you don't make the logical jump that the process is therefore not at fault for not receiving a signal, indeed, doesn't even exist at that point - it isn't stuck in its own code! Anybody using that bit of kernel code would stick. The kernel is at fault, not the process. The correct question to ask is about the kernel, not about the process.

Linux and Windows XP on the same PC
I have to get Linux and Windows XP running in the same PC. For a number of reasons that are not relevant here, I have to start from Linux...

Peter



Your Ad Here

List | Previous | Next

modutils on crack" followupmostly solved

Linux groups from Newsgroups

The #1 Usenet Provider on the Internet

off topic: LCD Terminal 2928