I now have my doubts too. In fact, more than doubts. The pread()/pwrite()
which are currently (2.2.0-pre8 I know it is ancient...) in Linux are
wrong. If you look close you will see that they are nothing than just
read()/write() with the extra argument (pos) passed as position to the
fs-specific layer, instead of file->f_pos. Most of the required
(by Single UNIX v2) error checking is missing. Hence I cc:linux-kernel so
that it is fixed. (I will do it if I have time, but if someone is already
doing it, this mail adds even more reason to do so).
Also, in write() the use of struct inode * inode makes things tidier than
not using it in pwrite(). So, that can be tidied up as well.
Regards,
------
Tigran A. Aivazian | http://www.sco.com
Escalations Research Group | tel: +44-(0)1923-813796
Santa Cruz Operation Ltd | http://www.aivazian.demon.co.uk
On Sun, 24 Jan 1999, Mike Jagdis wrote:
> On Fri, 22 Jan 1999, Tigran Aivazian wrote:
>
> > Looking in linux/fs/read_write.c I see that pread(2) is already there, but
> > no pread64(). So, am I right in assuming that it is relatively easy for
> > iBCS to support pread(2) now, but not pread64() yet?
>
> I have my doubts. Looking at the Linux pread() it seems to
> take an loff_t as the last argument so it may not be a
> straight forward mapping.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/