LWN: Comments on "Alternatives to fibrils"

This isn't slashdot.

tlw — 2007-02-28T20:25:54+00:00

... so that just leaves the third "it's"...

> I'm liberal by nature, and I think
> freedom will always find it's way.

which is incorrect. Freedom probably won't "find it is way".

Stupid question - that Linus anticipated

huaz — 2007-02-22T21:02:31+00:00

You are right, it's indeed unacceptable.

This isn't slashdot.

irios — 2007-02-22T08:33:39+00:00

Right about "Wiki's", but both the "It's", meaning "It is" rather than the possesive, are correct.

Cancellation of an operation

mikov — 2007-02-20T02:09:27+00:00

Has there been any discussion of canceling an asynchronous operation ? It seems to me that there may not be a structured way to do it using fibrils/kernel_threads because there are no formally defined states where a request sits and can be canceled.

Stupid question - that Linus anticipated

mikov — 2007-02-20T00:28:50+00:00

How could it _not_ matter to callers ? If the thread id can change arbitrarily based on factors outside of the applications control - e.g. if some driver buffer is empty - then the thread id becomes completely pointless.

To me this seems completely unacceptable. Am I missing something ?

Stupid question - that Linus anticipated

ds2horner — 2007-02-17T02:05:09+00:00

In his explanation it was so that the fork code could be reused with no modifications. He implies it could be done if it matters to callers.

Linus:
Now, I agree that this is a bit ugly in some of the details: in
particular, it means that if the system call blocks, we will literally
return as a *different* thread to user space. If you care, you shouldn't
use this interface, or come up with some way to make it work nicely (doing
it this way meant that I could just re-use all the clone/fork code as-is).

Alternatives to fibrils

proski — 2007-02-16T20:46:26+00:00

This is why free software is special - he just assumes it can get into the kernel, even tough Linus doesn't like it.

Who is "he"? Ingo or our editor? If Ingo, he wrote the patch before Linus commented on it. If it's the editor, please note the word "eventually". I don't think anybody assumes that Ingo's patch can go to the kernel as is.

I think you are generalizing too much. Other projects are run in a different way. Consensus doesn't always work. Although if developers are motivated to stick together, they will look for a solution that doesn't alienate any of them.

Stupid question

spitzak — 2007-02-16T19:47:42+00:00

In both cases the system call is done by the parent process and the return is the child process. Why can't this be done the other way around, where the asyncrhonous call is being done by the child and it returns immediately to the parent? That would make a lot more sense, so I assumme there is a sensible reason that I can't figure out for how they are doing this.

This isn't slashdot.

ncm — 2007-02-16T04:55:39+00:00

The possessive for "it" is "its". The plural of "wiki" is "wikis". That is all.

Alternatives to fibrils

jospoortvliet — 2007-02-15T19:11:43+00:00

>But the initial reaction to syslets appears to be
>positive (though Linus hates it); syslets might
>just point to the form of the fibril idea which
>eventually makes it into the mainline kernel.

This is why free software is special - he just assumes it can get into the
kernel, even tough Linus doesn't like it. Consensus is so 'normal' in FOSS
development you don't notice it all the time, but it's a great thing. It's
the reason for creativity and the 'free spirit'. I love it. I wish more
things worked this way... Politics anyone? I'm liberal by nature, and I
think freedom will always find it's way. As long as there is a community
watching it and protecting it. Wiki's, blogs, podcasts, they are
transforming the world. I hope we can protect it from you-know-who (from
evil governments to RIAA, from terrorists and other fanatics to greedy
companies).