Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=80601
Reported-by: David Binderman <[email protected]>
Signed-off-by: Andrey Utkin <[email protected]>
---
net/netfilter/ipvs/ip_vs_ctl.c | 2 --
1 file changed, 2 deletions(-)
diff --git a/net/netfilter/ipvs/ip_vs_ctl.c b/net/netfilter/ipvs/ip_vs_ctl.c
index 581a658..4ed7b59 100644
--- a/net/netfilter/ipvs/ip_vs_ctl.c
+++ b/net/netfilter/ipvs/ip_vs_ctl.c
@@ -2338,8 +2338,6 @@ do_ip_vs_set_ctl(struct sock *sk, int cmd, void __user *user, unsigned int len)
if (cmd < IP_VS_BASE_CTL || cmd > IP_VS_SO_SET_MAX)
return -EINVAL;
- if (len < 0 || len > MAX_ARG_LEN)
- return -EINVAL;
if (len != set_arglen[SET_CMDID(cmd)]) {
pr_err("set_ctl: len %u != %u\n",
len, set_arglen[SET_CMDID(cmd)]);
--
1.8.5.5
Hello,
On Fri, 18 Jul 2014, Andrey Utkin wrote:
> Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=80601
> Reported-by: David Binderman <[email protected]>
> Signed-off-by: Andrey Utkin <[email protected]>
> ---
> net/netfilter/ipvs/ip_vs_ctl.c | 2 --
> 1 file changed, 2 deletions(-)
>
> diff --git a/net/netfilter/ipvs/ip_vs_ctl.c b/net/netfilter/ipvs/ip_vs_ctl.c
> index 581a658..4ed7b59 100644
> --- a/net/netfilter/ipvs/ip_vs_ctl.c
> +++ b/net/netfilter/ipvs/ip_vs_ctl.c
> @@ -2338,8 +2338,6 @@ do_ip_vs_set_ctl(struct sock *sk, int cmd, void __user *user, unsigned int len)
>
> if (cmd < IP_VS_BASE_CTL || cmd > IP_VS_SO_SET_MAX)
> return -EINVAL;
> - if (len < 0 || len > MAX_ARG_LEN)
> - return -EINVAL;
The above check ensures the set_arglen[] value (some
struct size) does not exceed the arg[MAX_ARG_LEN] space. You can check
commit 04bcef2a83f40c ("ipvs: Add boundary check on ioctl arguments")
for more info. Still, check can be reduced to
if (len > MAX_ARG_LEN)... Also, len is unsigned, so len < 0 is
useless even for this reason.
> if (len != set_arglen[SET_CMDID(cmd)]) {
> pr_err("set_ctl: len %u != %u\n",
> len, set_arglen[SET_CMDID(cmd)]);
> --
> 1.8.5.5
Regards
--
Julian Anastasov <[email protected]>
2014-07-18 23:48 GMT+03:00 Julian Anastasov <[email protected]>:
> The above check ensures the set_arglen[] value (some
> struct size) does not exceed the arg[MAX_ARG_LEN] space. You can check
> commit 04bcef2a83f40c ("ipvs: Add boundary check on ioctl arguments")
> for more info.
Thanks for info.
What about static check at compilation time?
#if (DAEMON_ARG_LEN > MAX_ARG_LEN) \
|| (SERVICE_ARG_LEN > MAX_ARG_LEN) \
|| (SVCDEST_ARG_LEN > MAX_ARG_LEN)
#error MAX_ARG_LEN exceeded in set_arglen table
#endif
--
Andrey Utkin
2014-07-19 0:06 GMT+03:00 Andrey Utkin <[email protected]>:
> What about static check at compilation time?
>
> #if (DAEMON_ARG_LEN > MAX_ARG_LEN) \
> || (SERVICE_ARG_LEN > MAX_ARG_LEN) \
> || (SVCDEST_ARG_LEN > MAX_ARG_LEN)
> #error MAX_ARG_LEN exceeded in set_arglen table
> #endif
This approach doesn't work, looks because sizeof() values are not
calculated at preprocessing stage.
--
Andrey Utkin
2014-07-20 1:59 GMT+03:00 Andrey Utkin <[email protected]>:
> This approach doesn't work, looks because sizeof() values are not
> calculated at preprocessing stage.
What about adding equivalent BUG_ON() statement to __init
ip_vs_register_nl_ioctl(), which "registers" do_ip_vs_set_ctl()
callback?
--
Andrey Utkin
Hello,
On Sat, 19 Jul 2014, Andrey Utkin wrote:
> 2014-07-18 23:48 GMT+03:00 Julian Anastasov <[email protected]>:
> > The above check ensures the set_arglen[] value (some
> > struct size) does not exceed the arg[MAX_ARG_LEN] space. You can check
> > commit 04bcef2a83f40c ("ipvs: Add boundary check on ioctl arguments")
> > for more info.
>
> Thanks for info.
> What about static check at compilation time?
>
> #if (DAEMON_ARG_LEN > MAX_ARG_LEN) \
> || (SERVICE_ARG_LEN > MAX_ARG_LEN) \
> || (SVCDEST_ARG_LEN > MAX_ARG_LEN)
> #error MAX_ARG_LEN exceeded in set_arglen table
> #endif
I prefer this to be fixed once and for all.
Here is my attempt, it appears to work. I use union
to get the max size to use for stack space. checkpatch.pl
is noisy... Let me know if you have a better idea.
ipvs: reduce stack usage for sockopt data
Use macros and union to reserve the required stack space for
sockopt data. Now the tables for commands should be more safe
to extend. The checks added for readability are optimized by
compiler or warn at compile time if command uses too much stack.
Signed-off-by: Julian Anastasov <[email protected]>
---
net/netfilter/ipvs/ip_vs_ctl.c | 98 ++++++++++++++++++++++++------------------
1 file changed, 55 insertions(+), 43 deletions(-)
diff --git a/net/netfilter/ipvs/ip_vs_ctl.c b/net/netfilter/ipvs/ip_vs_ctl.c
index 581a658..08c3389 100644
--- a/net/netfilter/ipvs/ip_vs_ctl.c
+++ b/net/netfilter/ipvs/ip_vs_ctl.c
@@ -2266,28 +2266,34 @@ static int ip_vs_set_timeout(struct net *net, struct ip_vs_timeout_user *u)
}
-#define SET_CMDID(cmd) (cmd - IP_VS_BASE_CTL)
-#define SERVICE_ARG_LEN (sizeof(struct ip_vs_service_user))
-#define SVCDEST_ARG_LEN (sizeof(struct ip_vs_service_user) + \
- sizeof(struct ip_vs_dest_user))
-#define TIMEOUT_ARG_LEN (sizeof(struct ip_vs_timeout_user))
-#define DAEMON_ARG_LEN (sizeof(struct ip_vs_daemon_user))
-#define MAX_ARG_LEN SVCDEST_ARG_LEN
+#define SET_CMDID(cmd) (cmd - IP_VS_BASE_CTL)
+#define IP_VS_SET_CMDID(c, t) [SET_CMDID(c)] = sizeof(t),
+#define IP_VS_SET_CMDID_LEN(c, t) t field_ ## c;
+
+struct ip_vs_svcdest_user {
+ struct ip_vs_service_user s;
+ struct ip_vs_dest_user d;
+};
+
+#define IP_VS_SET_CMDID_TABLE(e) \
+ e(IP_VS_SO_SET_ADD, struct ip_vs_service_user) \
+ e(IP_VS_SO_SET_EDIT, struct ip_vs_service_user) \
+ e(IP_VS_SO_SET_DEL, struct ip_vs_service_user) \
+ e(IP_VS_SO_SET_ADDDEST, struct ip_vs_svcdest_user) \
+ e(IP_VS_SO_SET_DELDEST, struct ip_vs_svcdest_user) \
+ e(IP_VS_SO_SET_EDITDEST, struct ip_vs_svcdest_user) \
+ e(IP_VS_SO_SET_TIMEOUT, struct ip_vs_timeout_user) \
+ e(IP_VS_SO_SET_STARTDAEMON, struct ip_vs_daemon_user) \
+ e(IP_VS_SO_SET_STOPDAEMON, struct ip_vs_daemon_user) \
+ e(IP_VS_SO_SET_ZERO, struct ip_vs_service_user)
static const unsigned char set_arglen[SET_CMDID(IP_VS_SO_SET_MAX)+1] = {
- [SET_CMDID(IP_VS_SO_SET_ADD)] = SERVICE_ARG_LEN,
- [SET_CMDID(IP_VS_SO_SET_EDIT)] = SERVICE_ARG_LEN,
- [SET_CMDID(IP_VS_SO_SET_DEL)] = SERVICE_ARG_LEN,
- [SET_CMDID(IP_VS_SO_SET_FLUSH)] = 0,
- [SET_CMDID(IP_VS_SO_SET_ADDDEST)] = SVCDEST_ARG_LEN,
- [SET_CMDID(IP_VS_SO_SET_DELDEST)] = SVCDEST_ARG_LEN,
- [SET_CMDID(IP_VS_SO_SET_EDITDEST)] = SVCDEST_ARG_LEN,
- [SET_CMDID(IP_VS_SO_SET_TIMEOUT)] = TIMEOUT_ARG_LEN,
- [SET_CMDID(IP_VS_SO_SET_STARTDAEMON)] = DAEMON_ARG_LEN,
- [SET_CMDID(IP_VS_SO_SET_STOPDAEMON)] = DAEMON_ARG_LEN,
- [SET_CMDID(IP_VS_SO_SET_ZERO)] = SERVICE_ARG_LEN,
+ IP_VS_SET_CMDID_TABLE(IP_VS_SET_CMDID)
};
+union ip_vs_set_arglen { IP_VS_SET_CMDID_TABLE(IP_VS_SET_CMDID_LEN) };
+#define MAX_SET_ARGLEN sizeof(union ip_vs_set_arglen)
+
static void ip_vs_copy_usvc_compat(struct ip_vs_service_user_kern *usvc,
struct ip_vs_service_user *usvc_compat)
{
@@ -2325,7 +2331,7 @@ do_ip_vs_set_ctl(struct sock *sk, int cmd, void __user *user, unsigned int len)
{
struct net *net = sock_net(sk);
int ret;
- unsigned char arg[MAX_ARG_LEN];
+ unsigned char arg[MAX_SET_ARGLEN];
struct ip_vs_service_user *usvc_compat;
struct ip_vs_service_user_kern usvc;
struct ip_vs_service *svc;
@@ -2333,13 +2339,12 @@ do_ip_vs_set_ctl(struct sock *sk, int cmd, void __user *user, unsigned int len)
struct ip_vs_dest_user_kern udest;
struct netns_ipvs *ipvs = net_ipvs(net);
+ BUILD_BUG_ON(sizeof(arg) > 256);
if (!ns_capable(sock_net(sk)->user_ns, CAP_NET_ADMIN))
return -EPERM;
if (cmd < IP_VS_BASE_CTL || cmd > IP_VS_SO_SET_MAX)
return -EINVAL;
- if (len < 0 || len > MAX_ARG_LEN)
- return -EINVAL;
if (len != set_arglen[SET_CMDID(cmd)]) {
pr_err("set_ctl: len %u != %u\n",
len, set_arglen[SET_CMDID(cmd)]);
@@ -2606,49 +2611,56 @@ __ip_vs_get_timeouts(struct net *net, struct ip_vs_timeout_user *u)
}
-#define GET_CMDID(cmd) (cmd - IP_VS_BASE_CTL)
-#define GET_INFO_ARG_LEN (sizeof(struct ip_vs_getinfo))
-#define GET_SERVICES_ARG_LEN (sizeof(struct ip_vs_get_services))
-#define GET_SERVICE_ARG_LEN (sizeof(struct ip_vs_service_entry))
-#define GET_DESTS_ARG_LEN (sizeof(struct ip_vs_get_dests))
-#define GET_TIMEOUT_ARG_LEN (sizeof(struct ip_vs_timeout_user))
-#define GET_DAEMON_ARG_LEN (sizeof(struct ip_vs_daemon_user) * 2)
+#define GET_CMDID(cmd) (cmd - IP_VS_BASE_CTL)
+#define IP_VS_GET_CMDID(c, t) [GET_CMDID(c)] = sizeof(t),
+#define IP_VS_GET_CMDID_LEN(c, t) t field_ ## c;
+
+struct ip_vs_version_user {
+ char v[64];
+};
+
+struct ip_vs_daemon_user2 {
+ struct ip_vs_daemon_user d1, d2;
+};
+
+#define IP_VS_GET_CMDID_TABLE(e) \
+ e(IP_VS_SO_GET_VERSION, struct ip_vs_version_user) \
+ e(IP_VS_SO_GET_INFO, struct ip_vs_getinfo) \
+ e(IP_VS_SO_GET_SERVICES, struct ip_vs_get_services) \
+ e(IP_VS_SO_GET_SERVICE, struct ip_vs_service_entry) \
+ e(IP_VS_SO_GET_DESTS, struct ip_vs_get_dests) \
+ e(IP_VS_SO_GET_TIMEOUT, struct ip_vs_timeout_user) \
+ e(IP_VS_SO_GET_DAEMON, struct ip_vs_daemon_user2)
static const unsigned char get_arglen[GET_CMDID(IP_VS_SO_GET_MAX)+1] = {
- [GET_CMDID(IP_VS_SO_GET_VERSION)] = 64,
- [GET_CMDID(IP_VS_SO_GET_INFO)] = GET_INFO_ARG_LEN,
- [GET_CMDID(IP_VS_SO_GET_SERVICES)] = GET_SERVICES_ARG_LEN,
- [GET_CMDID(IP_VS_SO_GET_SERVICE)] = GET_SERVICE_ARG_LEN,
- [GET_CMDID(IP_VS_SO_GET_DESTS)] = GET_DESTS_ARG_LEN,
- [GET_CMDID(IP_VS_SO_GET_TIMEOUT)] = GET_TIMEOUT_ARG_LEN,
- [GET_CMDID(IP_VS_SO_GET_DAEMON)] = GET_DAEMON_ARG_LEN,
+ IP_VS_GET_CMDID_TABLE(IP_VS_GET_CMDID)
};
+union ip_vs_get_arglen { IP_VS_GET_CMDID_TABLE(IP_VS_GET_CMDID_LEN) };
+#define MAX_GET_ARGLEN sizeof(union ip_vs_get_arglen)
+
static int
do_ip_vs_get_ctl(struct sock *sk, int cmd, void __user *user, int *len)
{
- unsigned char arg[128];
+ unsigned char arg[MAX_GET_ARGLEN];
int ret = 0;
unsigned int copylen;
struct net *net = sock_net(sk);
struct netns_ipvs *ipvs = net_ipvs(net);
BUG_ON(!net);
+ BUILD_BUG_ON(sizeof(arg) > 256);
if (!ns_capable(sock_net(sk)->user_ns, CAP_NET_ADMIN))
return -EPERM;
if (cmd < IP_VS_BASE_CTL || cmd > IP_VS_SO_GET_MAX)
return -EINVAL;
- if (*len < get_arglen[GET_CMDID(cmd)]) {
- pr_err("get_ctl: len %u < %u\n",
- *len, get_arglen[GET_CMDID(cmd)]);
- return -EINVAL;
- }
-
copylen = get_arglen[GET_CMDID(cmd)];
- if (copylen > 128)
+ if (*len < (int) copylen || *len < 0) {
+ pr_err("get_ctl: len %d < %u\n", *len, copylen);
return -EINVAL;
+ }
if (copy_from_user(arg, user, copylen) != 0)
return -EFAULT;
--
1.9.0
Regards
--
Julian Anastasov <[email protected]>
On Mon, Jul 21, 2014 at 11:01:56PM +0300, Julian Anastasov wrote:
> @@ -2333,13 +2339,12 @@ do_ip_vs_set_ctl(struct sock *sk, int cmd, void __user *user, unsigned int len)
> struct ip_vs_dest_user_kern udest;
> struct netns_ipvs *ipvs = net_ipvs(net);
>
> + BUILD_BUG_ON(sizeof(arg) > 256);
256 is off-by-one because u8 ranges from 0-255 so we are never able to
copy 256 bytes into the "arg" buffer.
> - if (copylen > 128)
> + if (*len < (int) copylen || *len < 0) {
> + pr_err("get_ctl: len %d < %u\n", *len, copylen);
Don't let users flood dmesg. Just return an error. (This can be
triggered by non-root as well).
> return -EINVAL;
> + }
regards,
dan carpenter
Hello,
On Tue, 22 Jul 2014, Dan Carpenter wrote:
> On Mon, Jul 21, 2014 at 11:01:56PM +0300, Julian Anastasov wrote:
> > @@ -2333,13 +2339,12 @@ do_ip_vs_set_ctl(struct sock *sk, int cmd, void __user *user, unsigned int len)
> > struct ip_vs_dest_user_kern udest;
> > struct netns_ipvs *ipvs = net_ipvs(net);
> >
> > + BUILD_BUG_ON(sizeof(arg) > 256);
>
> 256 is off-by-one because u8 ranges from 0-255 so we are never able to
> copy 256 bytes into the "arg" buffer.
I'll change it to >= 256, to catch that we can
not hold such big size in unsigned char [gs]et_arglen and
also as an indication that we have to use allocated buffer
instead of stack.
> > - if (copylen > 128)
> > + if (*len < (int) copylen || *len < 0) {
> > + pr_err("get_ctl: len %d < %u\n", *len, copylen);
>
> Don't let users flood dmesg. Just return an error. (This can be
> triggered by non-root as well).
For now both set and get are privileged operations,
so we can keep it, it can catch if something wrong happens
with the structure sizes.
Regards
--
Julian Anastasov <[email protected]>
On Tue, Jul 22, 2014 at 11:52:02AM +0300, Julian Anastasov wrote:
> > > - if (copylen > 128)
> > > + if (*len < (int) copylen || *len < 0) {
> > > + pr_err("get_ctl: len %d < %u\n", *len, copylen);
> >
> > Don't let users flood dmesg. Just return an error. (This can be
> > triggered by non-root as well).
>
> For now both set and get are privileged operations,
> so we can keep it, it can catch if something wrong happens
> with the structure sizes.
If you have namespaces enabled then it's not *that* privaleged.
regards,
dan carpenter