cvs commit: src/sys/coda coda_fbsd.c coda_vnops.c src/sys/compat/linux linux_misc.c src/sys/dev/raidframe rf_freebsdkintf.c src/sys/fs/hpfs hpfs_vfsops.c src/sys/fs/msdosfs msdosfs_vfsops.c src/sys/fs/ntfs ntfs_vfsops.c src/sys/fs/nwfs ...

[Alfred Perlstein]

* Poul-Henning Kamp <phk at phk.freebsd.dk> [030726 13:31] wrote:
> In message <20030726091651.GB42924 at elvis.mu.org>, Alfred Perlstein writes:
> >* Poul-Henning Kamp <phk at FreeBSD.org> [030726 00:32] wrote:
> >> phk         2003/07/26 00:32:23 PDT
> >>   Log:
> >>   Add a "int fd" argument to VOP_OPEN() which in the future will
> >>   contain the filedescriptor number on opens from userland.
> >>   
> >>   The index is used rather than a "struct file *" since it conveys a bit
> >>   more information, which may be useful to in particular fdescfs and /dev/fd/*
> >>   
> >>   For now pass -1 all over the place.
> >
> >This seems like a mistake, it forces one to have a struct filedesc instead
> >of merely faking up a struct file.
> >
> >What if one wants to have a cookie without a filedesc?
> >
> >Am I mistaken?
> 
> I'm afraid that I totally fail to even understand your question...
> 
> The objective here is to improve the situation where a vnode open
> results in a non-vnode method vector on the file, /dev/fd[0-2],
> /dev/stream etc.  The index can be used with the (already) passed
> thread pointer to access the relevant file entry.
> 
> BTW: I am not done with this change, the next commit adds the
> identical argument to vn_open().

I was mistaken as to what I thought you were trying to accomplish
here.

I thought you were doing the "each device open gets a cookie" thing
from linux via some hack.  But instead it appears to be some weird
optimization.  Nevermind I guess.

-- 
-Alfred Perlstein [alfred at freebsd.org]
'Instead of asking why a piece of software is using "1970s technology,"
 start asking why software is ignoring 30 years of accumulated wisdom.'