Re: Bug#218980: parted: FTBFS : probably due to new glibc and 2.6.0-test linux kernel headers.

[Sven Luther <sven.luther@wanadoo.fr>]

On Mon, Nov 10, 2003 at 01:01:09AM +0000, Philip Blundell wrote:
> On Sun, 2003-11-09 at 19:46, Sven Luther wrote:
> > Well, it should not at least. I will investigate and see if the
> > linux/fs.h is where it comes from or not. I was not able to find where
> > the size_t was defined though.
> 
> This block of definitions in libparted/linux.c looks kind of
> suspicious.  Particularly the last line:
> 
> /* from <linux/fs.h> */
> #define BLKRRPART  _IO(0x12,95) /* re-read partition table */
> #define BLKGETSIZE _IO(0x12,96) /* return device size */
> #define BLKFLSBUF  _IO(0x12,97) /* flush buffer cache */
> #define BLKSSZGET  _IO(0x12,104) /* get block device sector size */
> #define BLKGETLASTSECT  _IO(0x12,108) /* get last sector of block device
> */
> #define BLKSETLASTSECT  _IO(0x12,109) /* set last sector of block device
> */
> #define BLKGETSIZE64 _IOR(0x12,114,sizeof(uint64_t))    /* return device
> size in bytes (u64 *arg) */

Ok, changing the sizeof(uint64_t) to size_t fixes one of the problem,
still the warning concerning the signed/unsigned comparison of
__invalid_size_argument_for_IOC and sizeof(size_t) persists, and since
dpkg-buildpackage will die on such warnings, i still cannot build the
package.

I have cloned the bug report, assigning the one above to parted, which
i will soon fix in an NMU, and assigning the one about signed/unsigned
to linux-kernel-headers. I have looked at the last 2 month of lkml
archives, and the consensus there seems to be to change the int
definition of __invalid_size_argument_for_IOC to unsigned int, not sure
though. I also did find a post of a guy having build problems of debian
stuff about exactly this for a bunch of packages, so it is not just me.

I hope to have helped solve this thing, and again, sorry for not having
noticed the above defines in linux.c and thanks for the help in this. I
have also mailed the parted mailing list about this problem.

BTW, forwarding to debian-glibc@lists.debian.org, as i don't know if it
will reach you now that i have reassigned the bug to parted, and not
sure i did the bug cloning right.

Friendly,

Sven Luther