On Mon, Nov 24, 2008 at 11:08:02AM +0900, hooanon05@xxxxxxxxxxx wrote:
> Subject: [PATCH 1/1] grow the size of loopback device and its backend file
It seems that the kernel part (LOOP_SET_CAPACITY) is not restricted
and it supports resizing in both directions (increase/decrease).
Right?
> +int grow_file(const char *file, unsigned long long new)
> +{
[...]
> + /*
> + * instread of ftrundate(2),
> + * allocate disk blocks to support ENOSPC
> + * on the loopback mounted filesystem
> + */
> + err = 0;
> + while (append > 0) {
> + if (append < sz)
> + sz = append;
> + ssz = write(fd, p, sz);
> + if (ssz == -1) {
> + if (errno == EAGAIN || errno == EINTR)
> + continue;
> + err = errno;
> + break;
> + }
> + append -= ssz;
> + }
Do we really need to hardcode this functionality to losetup(8)? Maybe
users want to use a different way how to resize the file. For example
the latest coreutils contains a nice user-friendly "truncate" utility
that provides all necessary functionality.
From my point of view everything what we need in losetup(8) is to
call LOOP_SET_CAPACITY ioctl.
Karel
--
Karel Zak <kzak@xxxxxxxxxx>
--
To unsubscribe from this list: send the line "unsubscribe util-linux-ng" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
