On Mon, 2006-05-15 at 20:50 +0200, Henrik Nordstrom wrote:
> tis 2006-05-16 klockan 02:01 +0800 skrev Adrian Chadd:
>
> > I'll eventually turn all of this into compile-time options, with
> > relevant warnings as to what works/what doesn't. I'd like to hope
> > the emulation of pread/pwrite is 'correct' but if you're right it
> > will be emulated by userspace lseek/read and lseek/write calls - which
> > must be made atomic somehow.
>
> it is quite trivial to emulate, thanks to dup() creating a new file
> pointer..
>
> dup
> lseek
> read/write
> close
Would it be that hard to just use multiple FDs to the same storage file
and then leave the OS to schedule the I/O?
Kinkie
