linux-kernel - Re: is avoiding compat ioctls possible?

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

Message-ID: <20091029083415.GA6639@osiris.boeblingen.de.ibm.com>
Date:	Thu, 29 Oct 2009 09:34:16 +0100
From:	Heiko Carstens <heiko.carstens@...ibm.com>
To:	David Miller <davem@...emloft.net>
Cc:	arndbergmann@...glemail.com, airlied@...ux.ie,
	dri-devel@...ts.sourceforge.net, andi@...stfloor.org,
	linux-kernel@...r.kernel.org, schwidefsky@...ibm.com,
	Ankit Jain <me@...itjain.org>, Christoph Hellwig <hch@....de>,
	Al Viro <viro@...iv.linux.org.uk>
Subject: Re: is avoiding compat ioctls possible?

On Wed, Oct 28, 2009 at 10:41:57PM -0700, David Miller wrote:
> From: Arnd Bergmann <arndbergmann@...glemail.com>
> Date: Wed, 28 Oct 2009 16:40:18 +0100
> 
> > I'm pretty sure it was ok when we started adding the compat_ioctl
> > handlers years ago. I think most people just ignored these for
> > the majority of drivers that can't possibly run on s390. Even
> > on s390, gcc will always do the right thing if you call call ioctl
> > with a pointer to a normal object in the .data section, heap or stack,
> > but hand-written assembly or other compilers may not.
> 
> Arnd, even compat_sys_ioctl() itself has constructs like:
> 
> 	case FS_IOC_RESVSP:
> 	case FS_IOC_RESVSP64:
> 		error = ioctl_preallocate(filp, (void __user *)arg);
> 		goto out_fput;

That's broken, but it's quite new code. In general it looks like we don't
have many compat ioctl problems on s390. At least I don't remember when
we faced the last bug.
We did have some compat syscall issues when SLES11 testing started.
The lack of bug reports is probably just a lack of 32 bit userspace ;)

This should fix at least the bug above:

Subject: [PATCH] fs: add missing compat_ptr handling for FS_IOC_RESVSP ioctl

From: Heiko Carstens <heiko.carstens@...ibm.com>

For FS_IOC_RESVSP and FS_IOC_RESVSP64 compat_sys_ioctl() uses its
arg argument as a pointer to userspace. However it is missing a
a call to compat_ptr() which will do a proper pointer conversion.

This was introduced with 3e63cbb1 "fs: Add new pre-allocation ioctls
to vfs for compatibility with legacy xfs ioctls".

Cc: Ankit Jain <me@...itjain.org>
Cc: Christoph Hellwig <hch@....de>
Cc: Al Viro <viro@...iv.linux.org.uk>
Cc: Arnd Bergmann <arndbergmann@...glemail.com>
Reported-by: David Miller <davem@...emloft.net>
Signed-off-by: Heiko Carstens <heiko.carstens@...ibm.com>
---
 fs/compat_ioctl.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/fs/compat_ioctl.c b/fs/compat_ioctl.c
index f91fd51..d84e705 100644
--- a/fs/compat_ioctl.c
+++ b/fs/compat_ioctl.c
@@ -1800,7 +1800,7 @@ struct space_resv_32 {
 /* just account for different alignment */
 static int compat_ioctl_preallocate(struct file *file, unsigned long arg)
 {
-	struct space_resv_32	__user *p32 = (void __user *)arg;
+	struct space_resv_32	__user *p32 = compat_ptr(arg);
 	struct space_resv	__user *p = compat_alloc_user_space(sizeof(*p));
 
 	if (copy_in_user(&p->l_type,	&p32->l_type,	sizeof(s16)) ||
@@ -2802,7 +2802,7 @@ asmlinkage long compat_sys_ioctl(unsigned int fd, unsigned int cmd,
 #else
 	case FS_IOC_RESVSP:
 	case FS_IOC_RESVSP64:
-		error = ioctl_preallocate(filp, (void __user *)arg);
+		error = ioctl_preallocate(filp, compat_ptr(arg));
 		goto out_fput;
 #endif
 
--
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/