Skip to contentSkip to conversation

net: uninitialized loopback addr leaks to userspace

PATCH New Vegard Nossum 8 participants 11 messages May 30, 2009 – Jun 9, 2009

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

Hi,

It seems that loopback's hardware address is never initialized by the kernel. So if userspace attempts to read this address before it has been set, the kernel will return some uninitialized data (only 6 bytes, though). This can be demonstrated by creating a new network namespace (CLONE_NEWNET), which creates a new loopback device, then call ioctl() with SIOCGIFHWADDR on "lo". If this is done in a loop, with some background load, or by running multiple instances, random data will start to show up in the returned address.

[  406.750329] WARNING: kmemcheck: Caught 16-bit read from
uninitialized memory (ffff880007220974)
[  406.753555] 18a2d7060088ffff18a2d7060088ffff00000000010000000100000003000000
[  406.758862]  i i i i i i i i i i i i i i i i i u u u u u u u u u u u u u u u
[  406.766224]                                          ^
[  406.768792] Modules linked in:
[  406.770416] Pid: 757, comm: ifconfig Not tainted
2.6.30-rc7-next-20090529 #404
[  406.772876] RIP: 0010:[<ffffffff80664789>]  [<ffffffff80664789>]
dev_ioctl+0x5d9/0x600
[  406.804677]  [<ffffffff8064ff75>] sock_ioctl+0x95/0x2a0
[  406.807242]  [<ffffffff802c35eb>] vfs_ioctl+0x1b/0x70
[  406.809348]  [<ffffffff802c36fa>] do_vfs_ioctl+0x8a/0x570
[  406.811419]  [<ffffffff802c3c79>] sys_ioctl+0x99/0xa0
[  406.813400]  [<ffffffff802f3941>] dev_ifsioc+0x81/0x2f0
[  406.815424]  [<ffffffff802f454d>] compat_sys_ioctl+0xed/0x3c0
[  406.817596]  [<ffffffff8022d476>] cstar_dispatch+0x7/0x26
[  406.819978]  [<ffffffffffffffff>] 0xffffffffffffffff

This is the code that triggers the warning, in net/core/dev.c, around line 4150:

    memcpy(ifr->ifr_hwaddr.sa_data, dev->dev_addr,
        min(sizeof ifr->ifr_hwaddr.sa_data, (size_t) dev->addr_len));

So it's dev->dev_addr that is the pointer to the uninitialized data.

I didn't know how to fix it.

Vegard

#1 lore raw
Quoted 4 lines · Vegard Nossum

It seems that loopback's hardware address is never initialized by the kernel. So if userspace attempts to read this address before it has been set, the kernel will return some uninitialized data (only 6 bytes, though).

Thank you for the report, Vegard.

I've been unable to reproduce the problem you describe, using 2.6-30-rc8, this test program and a couple of kernel builds for system load:

------------------------------------------------------------------
#define REPEAT_COUNT 10000

int childTask() {
	struct ifreq ifreq;
	int fd;
	unsigned char allBits;

	fd = socket(AF_INET, SOCK_DGRAM, 0);
	if (fd < 0){
		printf("Error %s from socket()\n", strerror(errno));
		_exit(-1);
	}

	strncpy(ifreq.ifr_name, "lo", sizeof("lo"));
	if (ioctl (fd, SIOCGIFHWADDR, &ifreq) < 0){
		printf("Error %s from ioctl(SIOCGIFHWADDR) for %s.\n", strerror(errno), ifreq.ifr_name);
		_exit(-1);
	}
	
	allBits = ifreq.ifr_hwaddr.sa_data[0] |
			ifreq.ifr_hwaddr.sa_data[1] |
			ifreq.ifr_hwaddr.sa_data[2] |
			ifreq.ifr_hwaddr.sa_data[3] |
			ifreq.ifr_hwaddr.sa_data[4] |
			ifreq.ifr_hwaddr.sa_data[5];

	if (allBits != 0)
		printf("Device %s -> Ethernet %02x:%02x:%02x:%02x:%02x:%02x\n", ifreq.ifr_name,
			(int) ((unsigned char *) &ifreq.ifr_hwaddr.sa_data)[0],
			(int) ((unsigned char *) &ifreq.ifr_hwaddr.sa_data)[1],
			(int) ((unsigned char *) &ifreq.ifr_hwaddr.sa_data)[2],
			(int) ((unsigned char *) &ifreq.ifr_hwaddr.sa_data)[3],
			(int) ((unsigned char *) &ifreq.ifr_hwaddr.sa_data)[4],
			(int) ((unsigned char *) &ifreq.ifr_hwaddr.sa_data)[5]);
}

int main(int argc, char *argv[]) {

   	void **child_stack;
	int pid, i, status;

	child_stack = (void **) malloc(16384);

	for (i = 0; i < REPEAT_COUNT; i++){

		pid = clone(childTask, child_stack, CLONE_NEWNET, NULL);
		if (pid < 0){
			printf("Error %s from clone()\n", strerror(errno));
			_exit(-1);
		}
		
		pid = waitpid(pid, &status, __WCLONE);
		if (pid < 0){
			printf("Error %s from waitpid()\n", strerror(errno));
			_exit(-1);
		}
	}   
      
	return 0;
}

------------------------------------------------------------------

Looking at the kernel code, it appears that all bytes of struct net_device, including the L2 address, are initialized to zeros at interface creation time.

Can you spot a difference between your test procedures and mine that would enable me to reproduce the problem?

-- John

#2 lore raw

2009/6/7 John Dykstra <[email protected]>:

Quoted 11 lines · Vegard Nossum
Quoted 4 lines

It seems that loopback's hardware address is never initialized by the kernel. So if userspace attempts to read this address before it has been set, the kernel will return some uninitialized data (only 6 bytes, though).

Thank you for the report, Vegard.

I've been unable to reproduce the problem you describe, using 2.6-30-rc8, this test program and a couple of kernel builds for system load:

[...]

Quoted 8 lines

------------------------------------------------------------------

Looking at the kernel code, it appears that all bytes of struct net_device, including the L2 address, are initialized to zeros at interface creation time.

Can you spot a difference between your test procedures and mine that would enable me to reproduce the problem?

Hi,

I just tried your test program on a linux-next kernel, it works beautifully :-)

(I made one change: The stack grows downwards on x86, so I think you should put child_stack + 16386 as the stack to clone()?)

As I wrote in reply to Stephen Hemminger, this problem seems to be caused by a particular patch in linux-next:

commit f001fde5eadd915f4858d22ed70d7040f48767cf

Author: Jiri Pirko <[email protected]>
Date:   Tue May 5 02:48:28 2009 +0000

net: introduce a list of device addresses dev_addr_list (v6)

Thanks for testing.

Vegard

Signature · 4 lines
"The animistic metaphor of the bug that maliciously sneaked in while
the programmer was not looking is intellectually dishonest as it
disguises that the error is the programmer's own creation."
	-- E. W. Dijkstra, EWD1036
#3 lore raw

PATCH 1/1 net: loopback device dev->addr_len fix

Quoted 28 lines · Vegard Nossum

2009/6/7 John Dykstra <[email protected]>:

Quoted 10 lines · Vegard Nossum
Quoted 4 lines

It seems that loopback's hardware address is never initialized by the kernel. So if userspace attempts to read this address before it has been set, the kernel will return some uninitialized data (only 6 bytes, though).

Thank you for the report, Vegard.

I've been unable to reproduce the problem you describe, using 2.6-30-rc8, this test program and a couple of kernel builds for system load:

[...]

Quoted 8 lines

------------------------------------------------------------------

Looking at the kernel code, it appears that all bytes of struct net_device, including the L2 address, are initialized to zeros at interface creation time.

Can you spot a difference between your test procedures and mine that would enable me to reproduce the problem?

Hi,

I just tried your test program on a linux-next kernel, it works beautifully :-)

(I made one change: The stack grows downwards on x86, so I think you should put child_stack + 16386 as the stack to clone()?)

Quoted 9 lines

As I wrote in reply to Stephen Hemminger, this problem seems to be caused by a particular patch in linux-next:

commit f001fde5eadd915f4858d22ed70d7040f48767cf

Author: Jiri Pirko <[email protected]>
Date:   Tue May 5 02:48:28 2009 +0000

net: introduce a list of device addresses dev_addr_list (v6)

I believe following patch should fix this problem.

Thank you

[PATCH net-next-2.6] net: loopback device dev->addr_len fix

commit f001fde5eadd915f4858d22ed70d7040f48767cf (net: introduce a list of device addresses dev_addr_list (v6)) added one regression Vegard Nossum found in its testings.

loopback device doesnt have a hw address, we should set its dev->addr_len to 0, not ETH_ALEN.

---

drivers/net/loopback.c +1 -1
index da472c6..40ded4e 100644
@@ -166,7 +166,7 @@ static void loopback_setup(struct net_device *dev)
 {
 	dev->mtu		= (16 * 1024) + 20 + 20 + 12;
 	dev->hard_header_len	= ETH_HLEN;	/* 14	*/
-	dev->addr_len		= ETH_ALEN;	/* 6	*/
+	dev->addr_len		= 0;
 	dev->tx_queue_len	= 0;
 	dev->type		= ARPHRD_LOOPBACK;	/* 0x0001*/
 	dev->flags		= IFF_LOOPBACK;
#4 lore raw

PATCH 1/1 net: dev_addr_init() fix

Quoted 48 lines · Eric Dumazet
Quoted 27 lines · Vegard Nossum

2009/6/7 John Dykstra <[email protected]>:

Quoted 10 lines · Vegard Nossum
Quoted 4 lines

It seems that loopback's hardware address is never initialized by the kernel. So if userspace attempts to read this address before it has been set, the kernel will return some uninitialized data (only 6 bytes, though).

Thank you for the report, Vegard.

I've been unable to reproduce the problem you describe, using 2.6-30-rc8, this test program and a couple of kernel builds for system load:

[...]

Quoted 8 lines

------------------------------------------------------------------

Looking at the kernel code, it appears that all bytes of struct net_device, including the L2 address, are initialized to zeros at interface creation time.

Can you spot a difference between your test procedures and mine that would enable me to reproduce the problem?

Hi,

I just tried your test program on a linux-next kernel, it works beautifully :-)

(I made one change: The stack grows downwards on x86, so I think you should put child_stack + 16386 as the stack to clone()?)

Quoted 9 lines

As I wrote in reply to Stephen Hemminger, this problem seems to be caused by a particular patch in linux-next:

commit f001fde5eadd915f4858d22ed70d7040f48767cf

Author: Jiri Pirko <[email protected]>
Date:   Tue May 5 02:48:28 2009 +0000

net: introduce a list of device addresses dev_addr_list (v6)

I believe following patch should fix this problem.

Thank you

[PATCH net-next-2.6] net: loopback device dev->addr_len fix

commit f001fde5eadd915f4858d22ed70d7040f48767cf (net: introduce a list of device addresses dev_addr_list (v6)) added one regression Vegard Nossum found in its testings.

Quoted 5 lines

loopback device doesnt have a hw address, we should set its dev->addr_len to 0, not ETH_ALEN.

Oh well, following is probably even more appropriate

[PATCH net-next-2.6] net: dev_addr_init() fix

commit f001fde5eadd915f4858d22ed70d7040f48767cf (net: introduce a list of device addresses dev_addr_list (v6)) added one regression Vegard Nossum found in its testings.

dev_addr_init() incorrectly uses sizeof() operator

---

net/core/dev.c +2 -2
index 1f38401..65387d9 100644
@@ -3655,8 +3655,8 @@ static int dev_addr_init(struct net_device *dev)
 	/* rtnl_mutex must be held here */
 
 	INIT_LIST_HEAD(&dev->dev_addr_list);
-	memset(addr, 0, sizeof(*addr));
-	err = __hw_addr_add(&dev->dev_addr_list, NULL, addr, sizeof(*addr),
+	memset(addr, 0, sizeof(addr));
+	err = __hw_addr_add(&dev->dev_addr_list, NULL, addr, sizeof(addr),
 			    NETDEV_HW_ADDR_T_LAN);
 	if (!err) {
 		/*
#5 lore raw

Re: [PATCH net-next-2.6] net: dev_addr_init() fix

Quoted 49 lines · Mon, Jun 08, 2009 at 02:13:32PM CEST, [email protected] wrote:
Quoted 48 lines · Eric Dumazet
Quoted 27 lines · Vegard Nossum

2009/6/7 John Dykstra <[email protected]>:

Quoted 10 lines · Vegard Nossum
Quoted 4 lines

It seems that loopback's hardware address is never initialized by the kernel. So if userspace attempts to read this address before it has been set, the kernel will return some uninitialized data (only 6 bytes, though).

Thank you for the report, Vegard.

I've been unable to reproduce the problem you describe, using 2.6-30-rc8, this test program and a couple of kernel builds for system load:

[...]

Quoted 8 lines

------------------------------------------------------------------

Looking at the kernel code, it appears that all bytes of struct net_device, including the L2 address, are initialized to zeros at interface creation time.

Can you spot a difference between your test procedures and mine that would enable me to reproduce the problem?

Hi,

I just tried your test program on a linux-next kernel, it works beautifully :-)

(I made one change: The stack grows downwards on x86, so I think you should put child_stack + 16386 as the stack to clone()?)

Quoted 9 lines

As I wrote in reply to Stephen Hemminger, this problem seems to be caused by a particular patch in linux-next:

commit f001fde5eadd915f4858d22ed70d7040f48767cf

Author: Jiri Pirko <[email protected]>
Date:   Tue May 5 02:48:28 2009 +0000

net: introduce a list of device addresses dev_addr_list (v6)

I believe following patch should fix this problem.

Thank you

[PATCH net-next-2.6] net: loopback device dev->addr_len fix

commit f001fde5eadd915f4858d22ed70d7040f48767cf (net: introduce a list of device addresses dev_addr_list (v6)) added one regression Vegard Nossum found in its testings.

Quoted 5 lines
Quoted 5 lines

loopback device doesnt have a hw address, we should set its dev->addr_len to 0, not ETH_ALEN.

oops, sorry for this...

Quoted 30 lines

Oh well, following is probably even more appropriate

[PATCH net-next-2.6] net: dev_addr_init() fix

commit f001fde5eadd915f4858d22ed70d7040f48767cf (net: introduce a list of device addresses dev_addr_list (v6)) added one regression Vegard Nossum found in its testings.

dev_addr_init() incorrectly uses sizeof() operator

---

net/core/dev.c +2 -2
index 1f38401..65387d9 100644
@@ -3655,8 +3655,8 @@ static int dev_addr_init(struct net_device *dev)
 	/* rtnl_mutex must be held here */
 
 	INIT_LIST_HEAD(&dev->dev_addr_list);
-	memset(addr, 0, sizeof(*addr));
-	err = __hw_addr_add(&dev->dev_addr_list, NULL, addr, sizeof(*addr),
+	memset(addr, 0, sizeof(addr));
+	err = __hw_addr_add(&dev->dev_addr_list, NULL, addr, sizeof(addr),
 			    NETDEV_HW_ADDR_T_LAN);
 	if (!err) {
 		/*
#6 lore raw

Re: [PATCH net-next-2.6] net: dev_addr_init() fix

In reply to Eric Dumazet
Quoted 59 lines · Eric Dumazet
Quoted 48 lines · Eric Dumazet
Quoted 27 lines · Vegard Nossum

2009/6/7 John Dykstra <[email protected]>:

Quoted 10 lines · Vegard Nossum
Quoted 4 lines

It seems that loopback's hardware address is never initialized by the kernel. So if userspace attempts to read this address before it has been set, the kernel will return some uninitialized data (only 6 bytes, though).

Thank you for the report, Vegard.

I've been unable to reproduce the problem you describe, using 2.6-30-rc8, this test program and a couple of kernel builds for system load:

[...]

Quoted 8 lines

------------------------------------------------------------------

Looking at the kernel code, it appears that all bytes of struct net_device, including the L2 address, are initialized to zeros at interface creation time.

Can you spot a difference between your test procedures and mine that would enable me to reproduce the problem?

Hi,

I just tried your test program on a linux-next kernel, it works beautifully :-)

(I made one change: The stack grows downwards on x86, so I think you should put child_stack + 16386 as the stack to clone()?)

Quoted 9 lines

As I wrote in reply to Stephen Hemminger, this problem seems to be caused by a particular patch in linux-next:

commit f001fde5eadd915f4858d22ed70d7040f48767cf

Author: Jiri Pirko <[email protected]>
Date:   Tue May 5 02:48:28 2009 +0000

net: introduce a list of device addresses dev_addr_list (v6)

I believe following patch should fix this problem.

Thank you

[PATCH net-next-2.6] net: loopback device dev->addr_len fix

commit f001fde5eadd915f4858d22ed70d7040f48767cf (net: introduce a list of device addresses dev_addr_list (v6)) added one regression Vegard Nossum found in its testings.

Quoted 5 lines

loopback device doesnt have a hw address, we should set its dev->addr_len to 0, not ETH_ALEN.

Oh well, following is probably even more appropriate

[PATCH net-next-2.6] net: dev_addr_init() fix

Quoted 8 lines

commit f001fde5eadd915f4858d22ed70d7040f48767cf (net: introduce a list of device addresses dev_addr_list (v6)) added one regression Vegard Nossum found in its testings.

dev_addr_init() incorrectly uses sizeof() operator

Could you please put the word 'kmemcheck' somewhere into the changelog, to make git-grepping and historic comparisons easier?

Thanks,

Ingo

Acked-by Jiri Pirko
#7 lore raw

Re: [PATCH net-next-2.6] net: dev_addr_init() fix

Quoted 63 lines · Ingo Molnar
Quoted 61 lines · Eric Dumazet
Quoted 47 lines · Eric Dumazet
Quoted 27 lines · Vegard Nossum

2009/6/7 John Dykstra <[email protected]>:

Quoted 10 lines · Vegard Nossum
Quoted 4 lines

It seems that loopback's hardware address is never initialized by the kernel. So if userspace attempts to read this address before it has been set, the kernel will return some uninitialized data (only 6 bytes, though).

Thank you for the report, Vegard.

I've been unable to reproduce the problem you describe, using 2.6-30-rc8, this test program and a couple of kernel builds for system load:

[...]

Quoted 8 lines

------------------------------------------------------------------

Looking at the kernel code, it appears that all bytes of struct net_device, including the L2 address, are initialized to zeros at interface creation time.

Can you spot a difference between your test procedures and mine that would enable me to reproduce the problem?

Hi,

I just tried your test program on a linux-next kernel, it works beautifully :-)

(I made one change: The stack grows downwards on x86, so I think you should put child_stack + 16386 as the stack to clone()?)

Quoted 9 lines

As I wrote in reply to Stephen Hemminger, this problem seems to be caused by a particular patch in linux-next:

commit f001fde5eadd915f4858d22ed70d7040f48767cf

Author: Jiri Pirko <[email protected]>
Date:   Tue May 5 02:48:28 2009 +0000

net: introduce a list of device addresses dev_addr_list (v6)

I believe following patch should fix this problem.

Thank you

[PATCH net-next-2.6] net: loopback device dev->addr_len fix

commit f001fde5eadd915f4858d22ed70d7040f48767cf (net: introduce a list of device addresses dev_addr_list (v6)) added one regression Vegard Nossum found in its testings.

Quoted 5 lines

loopback device doesnt have a hw address, we should set its dev->addr_len to 0, not ETH_ALEN.

Oh well, following is probably even more appropriate

[PATCH net-next-2.6] net: dev_addr_init() fix

commit f001fde5eadd915f4858d22ed70d7040f48767cf (net: introduce a list of device addresses dev_addr_list (v6)) added one regression Vegard Nossum found in its testings.

Quoted 8 lines
Quoted 4 lines

dev_addr_init() incorrectly uses sizeof() operator

Could you please put the word 'kmemcheck' somewhere into the changelog, to make git-grepping and historic comparisons easier?

Sure I can do that, giving me opportunity to use my current email address, since [email protected] will disappear shortly.

Thank you

[PATCH net-next-2.6] net: dev_addr_init() fix

commit f001fde5eadd915f4858d22ed70d7040f48767cf (net: introduce a list of device addresses dev_addr_list (v6)) added one regression Vegard Nossum found in its testings.

With kmemcheck help, Vegard found some uninitialized memory was read and reported to user, potentialy leaking kernel data. ( thread can be found on http://lkml.org/lkml/2009/5/30/177 )

dev_addr_init() incorrectly uses sizeof() operator. We were initializing one byte instead of MAX_ADDR_LEN bytes.

---

net/core/dev.c +2 -2
index 1f38401..65387d9 100644
@@ -3655,8 +3655,8 @@ static int dev_addr_init(struct net_device *dev)
 	/* rtnl_mutex must be held here */
 
 	INIT_LIST_HEAD(&dev->dev_addr_list);
-	memset(addr, 0, sizeof(*addr));
-	err = __hw_addr_add(&dev->dev_addr_list, NULL, addr, sizeof(*addr),
+	memset(addr, 0, sizeof(addr));
+	err = __hw_addr_add(&dev->dev_addr_list, NULL, addr, sizeof(addr),
 			    NETDEV_HW_ADDR_T_LAN);
 	if (!err) {
 		/*
maintainer Applied
#8 lore raw

Re: [PATCH net-next-2.6] net: dev_addr_init() fix

From: Eric Dumazet <[email protected]>
Date: Mon, 08 Jun 2009 15:49:24 +0200

Quoted 10 lines

[PATCH net-next-2.6] net: dev_addr_init() fix

commit f001fde5eadd915f4858d22ed70d7040f48767cf (net: introduce a list of device addresses dev_addr_list (v6)) added one regression Vegard Nossum found in its testings.

With kmemcheck help, Vegard found some uninitialized memory was read and reported to user, potentialy leaking kernel data. ( thread can be found on http://lkml.org/lkml/2009/5/30/177 )

Quoted 6 lines

dev_addr_init() incorrectly uses sizeof() operator. We were initializing one byte instead of MAX_ADDR_LEN bytes.

Applied, thanks.

#9 lore raw
In reply to Vegard Nossum
Quoted 30 lines · On Sat, 30 May 2009 22:23:24 +0200 Vegard Nossum <[email protected]>…

Hi,

It seems that loopback's hardware address is never initialized by the kernel. So if userspace attempts to read this address before it has been set, the kernel will return some uninitialized data (only 6 bytes, though). This can be demonstrated by creating a new network namespace (CLONE_NEWNET), which creates a new loopback device, then call ioctl() with SIOCGIFHWADDR on "lo". If this is done in a loop, with some background load, or by running multiple instances, random data will start to show up in the returned address.

[  406.750329] WARNING: kmemcheck: Caught 16-bit read from
uninitialized memory (ffff880007220974)
[  406.753555] 18a2d7060088ffff18a2d7060088ffff00000000010000000100000003000000
[  406.758862]  i i i i i i i i i i i i i i i i i u u u u u u u u u u u u u u u
[  406.766224]                                          ^
[  406.768792] Modules linked in:
[  406.770416] Pid: 757, comm: ifconfig Not tainted
2.6.30-rc7-next-20090529 #404
[  406.772876] RIP: 0010:[<ffffffff80664789>]  [<ffffffff80664789>]
dev_ioctl+0x5d9/0x600
[  406.804677]  [<ffffffff8064ff75>] sock_ioctl+0x95/0x2a0
[  406.807242]  [<ffffffff802c35eb>] vfs_ioctl+0x1b/0x70
[  406.809348]  [<ffffffff802c36fa>] do_vfs_ioctl+0x8a/0x570
[  406.811419]  [<ffffffff802c3c79>] sys_ioctl+0x99/0xa0
[  406.813400]  [<ffffffff802f3941>] dev_ifsioc+0x81/0x2f0
[  406.815424]  [<ffffffff802f454d>] compat_sys_ioctl+0xed/0x3c0
[  406.817596]  [<ffffffff8022d476>] cstar_dispatch+0x7/0x26
[  406.819978]  [<ffffffffffffffff>] 0xffffffffffffffff
Quoted 9 lines

This is the code that triggers the warning, in net/core/dev.c, around line 4150:

    memcpy(ifr->ifr_hwaddr.sa_data, dev->dev_addr,
        min(sizeof ifr->ifr_hwaddr.sa_data, (size_t) dev->addr_len));

So it's dev->dev_addr that is the pointer to the uninitialized data.

I didn't know how to fix it.

The whole dev structure is zeroed in alloc_netdev(), kmemcheck is giving bogus warning.

#10 lore raw

2009/6/8 Stephen Hemminger <[email protected]>:

Quoted 38 lines
Quoted 35 lines · On Sat, 30 May 2009 22:23:24 +0200 Vegard Nossum <[email protected]>…

Hi,

It seems that loopback's hardware address is never initialized by the kernel. So if userspace attempts to read this address before it has been set, the kernel will return some uninitialized data (only 6 bytes, though). This can be demonstrated by creating a new network namespace (CLONE_NEWNET), which creates a new loopback device, then call ioctl() with SIOCGIFHWADDR on "lo". If this is done in a loop, with some background load, or by running multiple instances, random data will start to show up in the returned address.

[  406.750329] WARNING: kmemcheck: Caught 16-bit read from uninitialized memory (ffff880007220974)
[  406.753555] 18a2d7060088ffff18a2d7060088ffff00000000010000000100000003000000 [  406.758862]  i i i i i i i i i i i i i i i i i u u u u u u u u u u u u u u u [  406.766224]                                          ^
[  406.768792] Modules linked in:
[  406.770416] Pid: 757, comm: ifconfig Not tainted 2.6.30-rc7-next-20090529 #404
[  406.772876] RIP: 0010:[<ffffffff80664789>]  [<ffffffff80664789>]

dev_ioctl+0x5d9/0x600
[  406.804677]  [<ffffffff8064ff75>] sock_ioctl+0x95/0x2a0
[  406.807242]  [<ffffffff802c35eb>] vfs_ioctl+0x1b/0x70
[  406.809348]  [<ffffffff802c36fa>] do_vfs_ioctl+0x8a/0x570
[  406.811419]  [<ffffffff802c3c79>] sys_ioctl+0x99/0xa0
[  406.813400]  [<ffffffff802f3941>] dev_ifsioc+0x81/0x2f0
[  406.815424]  [<ffffffff802f454d>] compat_sys_ioctl+0xed/0x3c0
[  406.817596]  [<ffffffff8022d476>] cstar_dispatch+0x7/0x26

[  406.819978]  [<ffffffffffffffff>] 0xffffffffffffffff

This is the code that triggers the warning, in net/core/dev.c, around line 4150:

    memcpy(ifr->ifr_hwaddr.sa_data, dev->dev_addr, min(sizeof ifr->ifr_hwaddr.sa_data, (size_t) dev->addr_len));

Quoted 8 lines
Quoted 4 lines

So it's dev->dev_addr that is the pointer to the uninitialized data.

I didn't know how to fix it.

The whole dev structure is zeroed in alloc_netdev(), kmemcheck is giving bogus warning.

Hi -- and sorry for being unclear. If I hadn't been sure that this was a real error, I would have said so (or not reported it at all).

I investigated it now, and as can be seen in the report above, I am using a -next kernel. It seems that the error was introduced in:

commit f001fde5eadd915f4858d22ed70d7040f48767cf

Author: Jiri Pirko <[email protected]>
Date:   Tue May 5 02:48:28 2009 +0000

net: introduce a list of device addresses dev_addr_list (v6)

So the error does not, as you say, exist in mainline Linux, but it's not a bogus warning either :-)

Adding to Cc.

Vegard

Signature · 4 lines
"The animistic metaphor of the bug that maliciously sneaked in while
the programmer was not looking is intellectually dishonest as it
disguises that the error is the programmer's own creation."
	-- E. W. Dijkstra, EWD1036