Re: execute a user process in the kernel

2004-09-25 Thread Bruce M Simpson
On Fri, Sep 24, 2004 at 06:21:25PM +0200, Dag-Erling Smørgrav wrote: You could, of course, write a kernel API for creating processes from scratch. They'd still need a parent, but you can use init(8) (pid 1) for that. If we were to implement POSIX spawn(), we'd need something like this. So

Re: execute a user process in the kernel

2004-09-25 Thread Dag-Erling Smørgrav
Bruce M Simpson [EMAIL PROTECTED] writes: On Fri, Sep 24, 2004 at 06:21:25PM +0200, Dag-Erling Smørgrav wrote: You could, of course, write a kernel API for creating processes from scratch. They'd still need a parent, but you can use init(8) (pid 1) for that. If we were to implement POSIX

Re: execute a user process in the kernel

2004-09-25 Thread Bruce M Simpson
On Sat, Sep 25, 2004 at 01:35:07PM +0200, Dag-Erling Smørgrav wrote: If we were to implement POSIX spawn(), we'd need something like this. Uh, no. posix_spawn(3) can be implemented entirely in userland and does not require any special kernel support. Discussions with peter@ and others on

Re: Strange FTPD behavior

2004-09-25 Thread Ralph Huntington
Is the user's shell listed in /etc/shells? It must be there for ftpd to let them in. On Mon, 26 Nov 2001, Vladimir Terziev wrote: I run FreeBSD 4.3-STABLE machine. I use ftpd for ftp server daemon. It has very strange behavior with one of user accounts on my machine. Every one user

Re: Freebsd 5.0-RELEASE named pipes

2004-09-25 Thread Borut Kurnik
Hi! The system doesn't wait for me to open the reader, I get the message instantly. I worked on 4.7 still does (also on linux, netbsd, ... :-) ) This is NetBSD 1.6 [EMAIL PROTECTED] root]# echo test f [1] 2383 [EMAIL PROTECTED] root]# cat f test [1]+ Doneecho test f

Re: execute a user process in the kernel

2004-09-25 Thread Dag-Erling Smørgrav
Bruce M Simpson [EMAIL PROTECTED] writes: On Sat, Sep 25, 2004 at 01:35:07PM +0200, Dag-Erling Smørgrav wrote: If we were to implement POSIX spawn(), we'd need something like this. Uh, no. posix_spawn(3) can be implemented entirely in userland and does not require any special kernel

SMP on quad xeons

2004-09-25 Thread jromero
Posted to freebsd-smp but didn't get too many replies, so I apologize for cross posting ahead of time. Need to configure groupware server and multiprotocol wireless proxy for aproximatly 2500 accounts. Application is heavily multi threaded and willrequire alot of CPU power. The OS will be FreeBSD

Re: SMP on quad xeons

2004-09-25 Thread Erich Dollansky
Hi, [EMAIL PROTECTED] wrote: Posted to freebsd-smp but didn't get too many replies, so I apologize for I am on that list too but I can't remember your posting. cross posting ahead of time. Need to configure groupware server and multiprotocol wireless proxy for aproximatly 2500 accounts.

Re: SMP on quad xeons

2004-09-25 Thread Steven Hartland
My experience with the current SMP kernel is it works very well on a 4 CPU opteron. That was 5.x-CURRENT a few months back at that point the new ULE scheduler was unusable as in it caused major delays in processing server apps. switched back to the old BSD one and all was good. Not sure it these