linux-kernel - Re: [patch/rfc] s390: get rid of own uid16 compat system calls

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [thread-next>] [day] [month] [year] [list]

Message-Id: <200607270303.42959.arnd.bergmann@de.ibm.com>
Date:	Thu, 27 Jul 2006 03:03:42 +0200
From:	Arnd Bergmann <arnd.bergmann@...ibm.com>
To:	Heiko Carstens <heiko.carstens@...ibm.com>
Cc:	linux-kernel@...r.kernel.org,
	Martin Schwidefsky <schwidefsky@...ibm.com>
Subject: Re: [patch/rfc] s390: get rid of own uid16 compat system calls

On Monday 10 July 2006 10:51, Heiko Carstens wrote:
> "Only" thing is that we unfortunately have different sizes for
> __kernel_old_[uid|gid]_t (16 bit on s390, 32 on s390x). I was tempted to
> change these just to find out that there are other users as well:
> 
> include/linux/ncp_fs.h:
>  #define NCP_IOC_GETMOUNTUID _IOW('n', 2, __kernel_old_uid_t)
> include/linux/smb_fs.h:
>  #define SMB_IOC_GETMOUNTUID _IOR('u', 1, __kernel_old_uid_t)
> 
> So, this is no option. Would anybody know of something to get this work?
> Or is this just a stupid idea?

Ok, I don't know exactly what you're talking about, but I have in the
past tried to hack that area as well. It's probably a good idea to
pick up my old patch and work from there, by making these two file systems
understand all possible ways:

#ifdef __KERNEL__
/* anything the user may be passing to us */
#define NCP_IOC_GETMOUNTUID16 _IOW('n', 2, u16)
#define NCP_IOC_GETMOUNTUID32 _IOW('n', 2, u32)
#define NCP_IOC_GETMOUNTUID64 _IOW('n', 2, u64)
#else
/* what the user is _supposed_ to pass */
#define NCP_IOC_GETMOUNTUID _IOW('n', 2, __kernel_old_uid_t) /* 16/32 bit */
#define NCP_IOC_GETMOUNTUID2 _IOW('n', 2, unsigned long) /* 32 or 64 bit */
#endif

int ncp_ioctl(struct inode *inode, struct file *filp, ...)
{
	...
	switch (arg) {
		...
	case NCP_IOC_GETMOUNTUID16:
	case NCP_IOC_GETMOUNTUID32:
	case NCP_IOC_GETMOUNTUID64:
		{
			unsigned long tmp = server->m.mounted_uid;

			if ((file_permission(filp, MAY_READ) != 0)
			    && (current->uid != server->m.mounted_uid))	{
				return -EACCES;
			}
			switch (arg) {
#ifdef CONFIG_UID16
			case NCP_IOC_GETMOUNTUID16:
				SET_UID(tmp, server->m.mounted_uid);
				if (put_user(tmp, (u16 __user *)argp)) 
					return -EFAULT;
#endif
			case NCP_IOC_GETMOUNTUID32:
				if (put_user(tmp, (u32 __user *)argp)) 
					return -EFAULT;
#ifdef CONFIG_64BIT
			case NCP_IOC_GETMOUNTUID64:
				if (put_user(tmp, (u64 __user *)argp)) 
					return -EFAULT;
			}
#endif
			return 0;
		}
	}
	...
}

I'm also posting the two patches I made a long time ago as a reference.

	Arnd <><
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/