Re: [PATCH] IPSec: IPv6 source address not set correctly in xfrm_state

Re: [PATCH] IPSec: IPv6 source address not set correctly in xfrm_state

[Tom Lendacky]

> >                      atomic_inc(&x0->refcnt);
> >          } else if (create && (x0 = xfrm_state_alloc()) != NULL) {
> > -                    memcpy(x0->sel.daddr.a6, daddr, sizeof(struct
in6_addr));
> > -                    memcpy(x0->sel.saddr.a6, saddr, sizeof(struct
in6_addr));
> > +                    memcpy(&x0->sel.daddr.a6, daddr, sizeof(struct
in6_addr));
> > +                    memcpy(&x0->sel.saddr.a6, saddr, sizeof(struct
in6_addr));
> >                      x0->sel.prefixlen_d = 128;
> >                      x0->sel.prefixlen_s = 128;
> > -                    memcpy(x0->props.saddr.a6, saddr, sizeof(struct
in6_addr));
> > +                    memcpy(&x0->props.saddr.a6, saddr, sizeof(struct
in6_addr));
> >                      x0->km.state = XFRM_STATE_ACQ;
> > -                    memcpy(x0->id.daddr.a6, daddr, sizeof(struct
in6_addr));
> > +                    memcpy(&x0->id.daddr.a6, daddr, sizeof(struct
in6_addr));
> >                      x0->id.proto = proto;
> >                      x0->props.family = AF_INET6;
> >                      x0->props.mode = mode;

> these are not correct.

Ok, I see, because "a6" is defined as an array.  Maybe all of the memcpy's
should be the same (include the .a6 or don't include the .a6) so that it is
consistent in the code?

David, would you like me to resubmit the patch or do you just want to
remove the last part of it?

Thanks,
Tom

Re: [PATCH] IPSec: IPv6 source address not set correctly in xfrm_state

[Jeff Garzik]

On Fri, Mar 21, 2003 at 10:47:22AM -0600, Tom Lendacky wrote:
> David, would you like me to resubmit the patch or do you just want to
> remove the last part of it?

Not speaking for David, but in general you should update and resend your
patches.  David, myself and others get tons of patches and we don't have
the time to babysit and hand-edit each one.

	Jeff

[PATCH] IPSec: IPv6 source address not set correctly in xfrm_state

[Tom Lendacky]

Here is a patch that fixes the source address in an xfrm_state
structure.  I found this when the incorrect address was supplied
in a pfkey ACQUIRE message.  Also, I wasn't able to test the
xfrm6_find_acq path, but I believe the memcpy's in that function
were not correct and fixed those also.

Thanks, 
Tom 

diff -ur linux-2.5.65-orig/net/ipv4/xfrm_state.c linux-2.5.65/net/ipv4/xfrm_state.c
--- linux-2.5.65-orig/net/ipv4/xfrm_state.c	2003-03-17 15:44:21.000000000 -0600
+++ linux-2.5.65/net/ipv4/xfrm_state.c	2003-03-21 08:48:58.000000000 -0600
@@ -404,7 +404,7 @@
 			memcpy(&x->id.daddr, daddr, sizeof(x->sel.daddr));
 		memcpy(&x->props.saddr, &tmpl->saddr, sizeof(x->props.saddr));
 		if (ipv6_addr_any((struct in6_addr*)&x->props.saddr))
-			memcpy(&x->props.saddr, &saddr, sizeof(x->sel.saddr));
+			memcpy(&x->props.saddr, saddr, sizeof(x->props.saddr));
 		x->props.mode = tmpl->mode;
 		x->props.reqid = tmpl->reqid;
 		x->props.family = AF_INET6;
@@ -642,13 +642,13 @@
 	if (x0) {
 		atomic_inc(&x0->refcnt);
 	} else if (create && (x0 = xfrm_state_alloc()) != NULL) {
-		memcpy(x0->sel.daddr.a6, daddr, sizeof(struct in6_addr));
-		memcpy(x0->sel.saddr.a6, saddr, sizeof(struct in6_addr));
+		memcpy(&x0->sel.daddr.a6, daddr, sizeof(struct in6_addr));
+		memcpy(&x0->sel.saddr.a6, saddr, sizeof(struct in6_addr));
 		x0->sel.prefixlen_d = 128;
 		x0->sel.prefixlen_s = 128;
-		memcpy(x0->props.saddr.a6, saddr, sizeof(struct in6_addr));
+		memcpy(&x0->props.saddr.a6, saddr, sizeof(struct in6_addr));
 		x0->km.state = XFRM_STATE_ACQ;
-		memcpy(x0->id.daddr.a6, daddr, sizeof(struct in6_addr));
+		memcpy(&x0->id.daddr.a6, daddr, sizeof(struct in6_addr));
 		x0->id.proto = proto;
 		x0->props.family = AF_INET6;
 		x0->props.mode = mode;

Re: [PATCH] IPSec: IPv6 source address not set correctly in xfrm_state

[YOSHIFUJI Hideaki / 吉藤英明]

Hello!

In article <1048258764.1214.28.camel@tomlt2.tomloffice.austin.ibm.com> (at 21 Mar 2003 08:59:23 -0600), Tom Lendacky <toml@us.ibm.com> says:

>  			memcpy(&x->id.daddr, daddr, sizeof(x->sel.daddr));
>  		memcpy(&x->props.saddr, &tmpl->saddr, sizeof(x->props.saddr));
>  		if (ipv6_addr_any((struct in6_addr*)&x->props.saddr))
> -			memcpy(&x->props.saddr, &saddr, sizeof(x->sel.saddr));
> +			memcpy(&x->props.saddr, saddr, sizeof(x->props.saddr));

this fix is ok. however

>  		atomic_inc(&x0->refcnt);
>  	} else if (create && (x0 = xfrm_state_alloc()) != NULL) {
> -		memcpy(x0->sel.daddr.a6, daddr, sizeof(struct in6_addr));
> -		memcpy(x0->sel.saddr.a6, saddr, sizeof(struct in6_addr));
> +		memcpy(&x0->sel.daddr.a6, daddr, sizeof(struct in6_addr));
> +		memcpy(&x0->sel.saddr.a6, saddr, sizeof(struct in6_addr));
>  		x0->sel.prefixlen_d = 128;
>  		x0->sel.prefixlen_s = 128;
> -		memcpy(x0->props.saddr.a6, saddr, sizeof(struct in6_addr));
> +		memcpy(&x0->props.saddr.a6, saddr, sizeof(struct in6_addr));
>  		x0->km.state = XFRM_STATE_ACQ;
> -		memcpy(x0->id.daddr.a6, daddr, sizeof(struct in6_addr));
> +		memcpy(&x0->id.daddr.a6, daddr, sizeof(struct in6_addr));
>  		x0->id.proto = proto;
>  		x0->props.family = AF_INET6;
>  		x0->props.mode = mode;

these are not correct.

-- 
Hideaki YOSHIFUJI @ USAGI Project <yoshfuji@linux-ipv6.org>
GPG FP: 9022 65EB 1ECF 3AD1 0BDF  80D8 4807 F894 E062 0EEA