On Mon, Aug 13, 2007 at 11:01:00PM +0200, Samuel Thibault wrote:
> Package: e2fsprogs
> Version: 1.39+1.40-WIP-2006.11.14+dfsg-2
> Severity: important
>
> libdevmapper-dev, libselinux1-dev and libsepol1-dev are Linux-specific,
> please make these build-deps
> [!hurd-i386 !kfreebsd-i386 !kfreebsd-amd64]
> and disabled --enable-blkid-devmapper on these archs.
What is DEB_HOST_GNU_TYPE_SYSTEM set to on the hurd? Is it "hurd"
(caps significant)?
Interestingly, the config.guess and config.sub files (from the FSF,
dated 2006-02-23) don't have the string "hurd" anywhere in them. That
can't be right, is it?
I'm on a plane at the moment heading over the pond en route to
Stockholm, so I can't check but iirc there isn't a GNU/hurd system
available for DD's to login to. If there isn't such a system, it will
be a lot better if you send me patches, as opposed to me just guessing
whether a set of patches will cause to build without bugs on the Hurd.
> Also, the Hurd (and maybe kfreebsd too) doesn't (yet) have the ioctls
> for performing online resizing. A crude solution for now is to just
> disable the resizing facility (in all of the STD, BF flags and the udeb
> rules), but there could probably be a ./configure option to just disable
> compiling online.c and error out instead of calling online_resize_fs().
No, the right thing to do is to just do an #ifdef in online.c to check
to see if the ioctl's are defined. Resize2fs is still useful on
non-Linux systems since it also handles off-line resizing.
Regards,
- Ted
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
