[PATCH net-next] ipcomp: Convert struct xt_ipcomp spis into 16bits

[PATCH net-next] ipcomp: Convert struct xt_ipcomp spis into 16bits

[Fan Du]

sparse warnings: (new ones prefixed by >>)

>> >> net/netfilter/xt_ipcomp.c:63:26: sparse: restricted __be16 degrades to integer
>> >> net/netfilter/xt_ipcomp.c:63:26: sparse: cast to restricted __be32

Fix this by using 16bits long spi, as IPcomp CPI is only valid for 16bits.

Signed-off-by: Fan Du <fan.du@windriver.com>
---
 include/uapi/linux/netfilter/xt_ipcomp.h |    2 +-
 net/netfilter/xt_ipcomp.c                |    4 ++--
 2 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/include/uapi/linux/netfilter/xt_ipcomp.h b/include/uapi/linux/netfilter/xt_ipcomp.h
index 45c7e40..ca82ebb 100644
--- a/include/uapi/linux/netfilter/xt_ipcomp.h
+++ b/include/uapi/linux/netfilter/xt_ipcomp.h
@@ -4,7 +4,7 @@
 #include <linux/types.h>
 
 struct xt_ipcomp {
-	__u32 spis[2];	/* Security Parameter Index */
+	__u16 spis[2];	/* Security Parameter Index */
 	__u8 invflags;	/* Inverse flags */
 	__u8 hdrres;	/* Test of the Reserved Filed */
 };
diff --git a/net/netfilter/xt_ipcomp.c b/net/netfilter/xt_ipcomp.c
index a4c7561..5542cb2 100644
--- a/net/netfilter/xt_ipcomp.c
+++ b/net/netfilter/xt_ipcomp.c
@@ -29,7 +29,7 @@ MODULE_DESCRIPTION("Xtables: IPv4/6 IPsec-IPComp SPI match");
 
 /* Returns 1 if the spi is matched by the range, 0 otherwise */
 static inline bool
-spi_match(u_int32_t min, u_int32_t max, u_int32_t spi, bool invert)
+spi_match(u_int16_t min, u_int16_t max, u_int16_t spi, bool invert)
 {
 	bool r;
 	pr_debug("spi_match:%c 0x%x <= 0x%x <= 0x%x\n",
@@ -60,7 +60,7 @@ static bool comp_mt(const struct sk_buff *skb, struct xt_action_param *par)
 	}
 
 	return spi_match(compinfo->spis[0], compinfo->spis[1],
-			 ntohl(chdr->cpi << 16),
+			 ntohl(chdr->cpi),
 			 !!(compinfo->invflags & XT_IPCOMP_INV_SPI));
 }
 
-- 
1.7.9.5

Re: [PATCH net-next] ipcomp: Convert struct xt_ipcomp spis into 16bits

[Pablo Neira Ayuso]

On Sat, Jan 18, 2014 at 10:16:14AM +0800, Fan Du wrote:
> sparse warnings: (new ones prefixed by >>)
> 
> >> >> net/netfilter/xt_ipcomp.c:63:26: sparse: restricted __be16 degrades to integer
> >> >> net/netfilter/xt_ipcomp.c:63:26: sparse: cast to restricted __be32
> 
> Fix this by using 16bits long spi, as IPcomp CPI is only valid for 16bits.
> 
> Signed-off-by: Fan Du <fan.du@windriver.com>
> ---
>  include/uapi/linux/netfilter/xt_ipcomp.h |    2 +-
>  net/netfilter/xt_ipcomp.c                |    4 ++--
>  2 files changed, 3 insertions(+), 3 deletions(-)
> 
> diff --git a/include/uapi/linux/netfilter/xt_ipcomp.h b/include/uapi/linux/netfilter/xt_ipcomp.h
> index 45c7e40..ca82ebb 100644
> --- a/include/uapi/linux/netfilter/xt_ipcomp.h
> +++ b/include/uapi/linux/netfilter/xt_ipcomp.h
> @@ -4,7 +4,7 @@
>  #include <linux/types.h>
>  
>  struct xt_ipcomp {
> -	__u32 spis[2];	/* Security Parameter Index */
> +	__u16 spis[2];	/* Security Parameter Index */

This changes the binary interface so it break userspace (iptables
needs to be recompiled), we're still in time to make such change as
this is net-next stuff, but what I understand from the patch
description is that this aims to fix a sparse warning, which is a bit
of intrusive change.

Didn't you find any way to fix this without change the layout of
xt_ipcomp?

>  	__u8 invflags;	/* Inverse flags */
>  	__u8 hdrres;	/* Test of the Reserved Filed */
>  };
> diff --git a/net/netfilter/xt_ipcomp.c b/net/netfilter/xt_ipcomp.c
> index a4c7561..5542cb2 100644
> --- a/net/netfilter/xt_ipcomp.c
> +++ b/net/netfilter/xt_ipcomp.c
> @@ -29,7 +29,7 @@ MODULE_DESCRIPTION("Xtables: IPv4/6 IPsec-IPComp SPI match");
>  
>  /* Returns 1 if the spi is matched by the range, 0 otherwise */
>  static inline bool
> -spi_match(u_int32_t min, u_int32_t max, u_int32_t spi, bool invert)
> +spi_match(u_int16_t min, u_int16_t max, u_int16_t spi, bool invert)
>  {
>  	bool r;
>  	pr_debug("spi_match:%c 0x%x <= 0x%x <= 0x%x\n",
> @@ -60,7 +60,7 @@ static bool comp_mt(const struct sk_buff *skb, struct xt_action_param *par)
>  	}
>  
>  	return spi_match(compinfo->spis[0], compinfo->spis[1],
> -			 ntohl(chdr->cpi << 16),
> +			 ntohl(chdr->cpi),
>  			 !!(compinfo->invflags & XT_IPCOMP_INV_SPI));
>  }
>  
> -- 
> 1.7.9.5
> 

Re: [PATCH net-next] ipcomp: Convert struct xt_ipcomp spis into 16bits

[Fan Du]

On 2014年01月18日 20:24, Pablo Neira Ayuso wrote:
> On Sat, Jan 18, 2014 at 10:16:14AM +0800, Fan Du wrote:
>> >  sparse warnings: (new ones prefixed by>>)
>> >
>>>>>> >  >>  >>  net/netfilter/xt_ipcomp.c:63:26: sparse: restricted __be16 degrades to integer
>>>>>> >  >>  >>  net/netfilter/xt_ipcomp.c:63:26: sparse: cast to restricted __be32
>> >
>> >  Fix this by using 16bits long spi, as IPcomp CPI is only valid for 16bits.
>> >
>> >  Signed-off-by: Fan Du<fan.du@windriver.com>
>> >  ---
>> >    include/uapi/linux/netfilter/xt_ipcomp.h |    2 +-
>> >    net/netfilter/xt_ipcomp.c                |    4 ++--
>> >    2 files changed, 3 insertions(+), 3 deletions(-)
>> >
>> >  diff --git a/include/uapi/linux/netfilter/xt_ipcomp.h b/include/uapi/linux/netfilter/xt_ipcomp.h
>> >  index 45c7e40..ca82ebb 100644
>> >  --- a/include/uapi/linux/netfilter/xt_ipcomp.h
>> >  +++ b/include/uapi/linux/netfilter/xt_ipcomp.h
>> >  @@ -4,7 +4,7 @@
>> >    #include<linux/types.h>
>> >
>> >    struct xt_ipcomp {
>> >  -	__u32 spis[2];	/* Security Parameter Index */
>> >  +	__u16 spis[2];	/* Security Parameter Index */
> This changes the binary interface so it break userspace (iptables
> needs to be recompiled), we're still in time to make such change as
> this is net-next stuff, but what I understand from the patch
> description is that this aims to fix a sparse warning, which is a bit
> of intrusive change.
>
> Didn't you find any way to fix this without change the layout of
> xt_ipcomp?
>

My bad for not catching this in the initial ipcomp version.
It would be easier to just use ntohs here to ease sparse checking.


 From dbf796e63d600256fd40000669b83227ddccebc4 Mon Sep 17 00:00:00 2001
From: Fan Du <fan.du@windriver.com>
Date: Mon, 20 Jan 2014 09:42:12 +0800
Subject: [PATCHv2 net-next] ipcomp: Use ntohs to ease sparse warning

0-DAY kernel build testing backend reported:

sparse warnings: (new ones prefixed by >>)

 >> >> net/netfilter/xt_ipcomp.c:63:26: sparse: restricted __be16 degrades to integer
 >> >> net/netfilter/xt_ipcomp.c:63:26: sparse: cast to restricted __be32

Fix this by using ntohs without shifting.

Tested with: make C=1 CF=-D__CHECK_ENDIAN__

Signed-off-by: Fan Du <fan.du@windriver.com>
---
  net/netfilter/xt_ipcomp.c |    2 +-
  1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/net/netfilter/xt_ipcomp.c b/net/netfilter/xt_ipcomp.c
index a4c7561..89d5310 100644
--- a/net/netfilter/xt_ipcomp.c
+++ b/net/netfilter/xt_ipcomp.c
@@ -60,7 +60,7 @@ static bool comp_mt(const struct sk_buff *skb, struct xt_action_param *par)
  	}

  	return spi_match(compinfo->spis[0], compinfo->spis[1],
-			 ntohl(chdr->cpi << 16),
+			 ntohs(chdr->cpi),
  			 !!(compinfo->invflags & XT_IPCOMP_INV_SPI));
  }

-- 
1.7.9.5


-- 
浮沉随浪只记今朝笑

--fan

Re: [PATCH net-next] ipcomp: Convert struct xt_ipcomp spis into 16bits

[Pablo Neira Ayuso]

On Mon, Jan 20, 2014 at 09:55:42AM +0800, Fan Du wrote:
> On 2014年01月18日 20:24, Pablo Neira Ayuso wrote:
> >Didn't you find any way to fix this without change the layout of
> >xt_ipcomp?
> 
> My bad for not catching this in the initial ipcomp version.
> It would be easier to just use ntohs here to ease sparse checking.
> 
> 
> From dbf796e63d600256fd40000669b83227ddccebc4 Mon Sep 17 00:00:00 2001
> From: Fan Du <fan.du@windriver.com>
> Date: Mon, 20 Jan 2014 09:42:12 +0800
> Subject: [PATCHv2 net-next] ipcomp: Use ntohs to ease sparse warning
> 
> 0-DAY kernel build testing backend reported:
> 
> sparse warnings: (new ones prefixed by >>)
> 
> >> >> net/netfilter/xt_ipcomp.c:63:26: sparse: restricted __be16 degrades to integer
> >> >> net/netfilter/xt_ipcomp.c:63:26: sparse: cast to restricted __be32
> 
> Fix this by using ntohs without shifting.
> 
> Tested with: make C=1 CF=-D__CHECK_ENDIAN__

Applied to nf-next, thanks.