Simpler syslets [LWN.net]

Simpler syslets

Posted Dec 10, 2007 23:01 UTC (Mon) by josh (subscriber, #17465)
Parent article: Simpler syslets

This new architecture seems to remove one of the key advantages of the previous syslet
interfaces: the ability to run many system calls without incurring system call overhead for
each one.

(Log in to post comments)

Simpler syslets

Posted Dec 10, 2007 23:10 UTC (Mon) by nix (subscriber, #2304) [Link]

Also, because the completion function is called in the original 
process --- well, what do the old and new tasks share? If they share as 
little as POSIX processes, the completion function is going to have to 
resort to something nasty like a pipe read by the other process just to 
communicate with itself.

Simpler syslets

Posted Dec 10, 2007 23:31 UTC (Mon) by i3839 (guest, #31386) [Link]

I think that's what the ugly syslet_ring_wait() is meant for.

I see a lot of unnecessary complexity, where should I drop my feedback, lkml or lwn? I'll mail
it to lkml tomorrow...

Simpler syslets

Posted Dec 11, 2007 6:39 UTC (Tue) by butlerm (subscriber, #13312) [Link]

In the current patch, the parent and child share virtually everything (VM, FS, files, etc) - i.e. the child task is essentially a thread, not an independent POSIX process. I can hardly imagine why anyone would want to do anything else.

Also, the more sophisticated functionality is not being abandoned, it is just being put off in the name of initial simplicity. The announcement has the details.