posix_fallocate(2)

[mdf at FreeBSD.org]

Mailing list ate my attachment.

See http://people.freebsd.org/~mdf/check_posix_fallocate.c .

As an aside, when I post from gmail.com to the mailing lists, I don't
see a copy of my message.  Is this a feature of the litserv, where it
doesn't mail it back to the originator, or a "feature" of gmail where
it tries to be clever?  The problem is that when the list eats an
attachment I don't know until someone takes the time to email me, and
I'd rather be able to tell immediately.

Thanks,
matthew

On Tue, Apr 19, 2011 at 10:08 AM,  <mdf at freebsd.org> wrote:
> The syscall and default VOP are in CURRENT now.
>
> I'm attaching the testcase I was using to make sure it was correct, so
> that work doesn't need to be re-done if anyone wants to work on an
> implementation of VOP_ALLOCATE for ufs or zfs.  I'd like to integrate
> this into tools/test/posix_fallocate/ but honestly I'm not likely to
> have the time soon.  I intend to try and take care of this at BSDCan
> when my job responsibilities for the week are to hack on FreeBSD. :-)
>
> FYI, the odd format to the code is mostly due to using a few #defines
> to get this test code that was originally written for $WORK's hack of
> libcheck to compile and run as a stand-alone.  For FreeBSD I've been
> testing as:
>
> cc -o falloc check_posix_fallocate.c -lsbuf
> ./falloc
>
> At $WORK we have a test.mk make addendum, and a few variables set in a
> make file, and the test is run as:
>
> make check
>
> Cheers,
> matthew
>