xfs
[Top] [All Lists]

Re: BLKGETSIZE64, BLKBSZSET, BLKSSZGET definition in xfsprogs.

To: "Kevin P. Fleming" <kpfleming@xxxxxxxxxxxxxxxxxxxx>
Subject: Re: BLKGETSIZE64, BLKBSZSET, BLKSSZGET definition in xfsprogs.
From: Jan Derfinak <ja@xxxxxxxxxxxx>
Date: Tue, 7 Oct 2003 16:02:46 +0200 (CEST)
Cc: linux-xfs@xxxxxxxxxxx
In-reply-to: <3F82C06A.1020808@xxxxxxxxxxxxxxxxxxxx>
References: <Pine.LNX.4.53.0310060739360.16790@xxxxxxxxxxxxxxxxxx> <20031006055957.GC1001@frodo> <Pine.LNX.4.53.0310071107220.1331@xxxxxxxxxxxxxxxxxx> <3F82C06A.1020808@xxxxxxxxxxxxxxxxxxxx>
Sender: linux-xfs-bounce@xxxxxxxxxxx
On Tue, 7 Oct 2003, Kevin P. Fleming wrote:

> Jan Derfinak wrote:
>
> > There was change in 2.6.0-test5 and linux/fs.h isn't included in
> > xfsprogs so xfsprogs uses old definition.
> >
> >
> >>bzcat patch-2.6.0-test5.bz2 | grep -e BLKBSZSET -e BLKGETSIZE64
> >
> > -#define BLKBSZSET  _IOW(0x12,113,sizeof(int))
> > -#define BLKGETSIZE64 _IOR(0x12,114,sizeof(u64))        /* return device 
> > size in bytes (u64 *arg) */
> > +#define BLKBSZSET  _IOW(0x12,113,size_t)
> > +#define BLKGETSIZE64 _IOR(0x12,114,size_t)     /* return device size in 
> > bytes (u64 *arg) */
> >
>
> Actually, those definitions will produce the same ioctl number, so
> effectively there is no change. The kernel can't really change ioctl
> numbers once userspace begins using them, because that's an
> incompatible change, so it's fine if a userspace package copies the
> definition of some ioctls and then uses its own copies.
>

But there is difference in third parameter. And this cause error in
_IOC_TYPECHECK macro in asm/ioctl.h.

                                jan

-- 


<Prev in Thread] Current Thread [Next in Thread>