Date: Wed, 16 Dec 1998 22:23:34 -0500 (EST) From: Brian Feldman <green@unixhelp.org> To: Julian Elischer <julian@whistle.com> Cc: Mike Smith <mike@smith.net.au>, "Richard Seaman, Jr." <lists@tar.com>, Nate Williams <nate@mt.sri.com>, "current@freebsd.org" <current@FreeBSD.ORG> Subject: Re: Linux Threads patches available Message-ID: <Pine.BSF.4.05.9812162217050.19075-100000@janus.syracuse.net> In-Reply-To: <36783B03.446B9B3D@whistle.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Wed, 16 Dec 1998, Julian Elischer wrote:
> Mike Smith wrote:
>
> >
> > For the sake of reestablishing our compatibility with Linux, I think
> > that your current code should be strongly considered. As for the
Yay! Something came of my efforts (even if I stopped working on it myself)!
> > eventual fate of threading in the FreeBSD kernel, we're still really
> > dependant on the availibility of development resources here, either in
> > the form of willing and able volunteers or funding which would allow .
>
> The system patch affects the following files:
>
> Index: bin/ps/keyword.c
> Index: lib/libc/i386/sys/sigsuspend.S
> Index: sys/i386/linux/linux.h
> Index: sys/i386/linux/linux_dummy.c
> Index: sys/i386/linux/linux_misc.c
> Index: sys/i386/linux/linux_proto.h
> Index: sys/i386/linux/linux_syscall.h
> Index: sys/i386/linux/linux_sysent.c
> Index: sys/i386/linux/linux_sysvec.c
> Index: sys/i386/linux/syscalls.master
> Index: sys/kern/imgact_elf.c
> Index: sys/kern/init_main.c
> Index: sys/kern/kern_exit.c
> Index: sys/kern/kern_fork.c
> Index: sys/kern/kern_sig.c
> Index: sys/sys/proc.h
> Index: sys/sys/signalvar.h
> Index: sys/sys/unistd.h
> Index: sys/sys/user.h
> Index: sys/vm/vm_glue.c
>
>
> the sys/i386/linux files are not an issue really.
>
> the other changes relate to the moving of a couple of fields out
> of the proc/user structures to a separate sharable entity. vis.
I tried to keep my patches style(9) compliant, make sure your modifications are too :)
>
> + struct procsig {
> + #define ps_begincopy ps_sigignore
> + sigset_t ps_sigignore; /* Signals being ignored. */
> + sigset_t ps_sigcatch; /* Signals being caught by user. */
> + int ps_flag;
> + struct sigacts ps_sigacts;
> + #define ps_endcopy ps_refcnt
> + int ps_refcnt;
> + int ps_posix;
> + };
> +
>
> As I said in other mail. This may not be perfect but if we don't
> make a start on threads, we'll never get to the end.. This seems like
> a fair place to start. I don't see that implimenting it would be
> detrimental to the existing system..
>
> I would vote for an inclusion to allow others to start experimenting
> with linux-threads based software,
> (note there is now a linux-threads based java JVM.)
>
> thoughts all?
>
I don't think it should quite be included yet, because noone's
really thought of the impact of an EXTRA per-process malloc, have
they? If this is alright to give up an extra 4k per process (let's
see, a max-process-count of 100000, this is a total of 400 mb?!?),
then perhaps this should be done. Other than that, I'm not so sure,
because this could be a Very Bad Thing. Has anyone really thought
about the per-process memory overhead issues that come up nowadays
with struct proc growing, and being added to? It seems like to max
out a system, there's at LEAST 800 MB of RAM dedicated to struct
proc and procsig, right?
> julian
>
> To Unsubscribe: send mail to majordomo@FreeBSD.org
> with "unsubscribe freebsd-current" in the body of the message
>
Brian Feldman _ __ ___ ___ ___
green@unixhelp.org _ __ ___ | _ ) __| \
http://www.freebsd.org/ _ __ ___ ____ | _ \__ \ |) |
FreeBSD: The Power to Serve! _ __ ___ ____ _____ |___/___/___/
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-current" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.BSF.4.05.9812162217050.19075-100000>